Read Tveskov_HayesSite.pdf text version

The Hayes Site: Oral Tradition, Ethnohistory, and Archaeology of the South Fork Coquille River

by Mark Tveskov with contributions by Nick Halousek, Nicole Norris, & Christian Solfisburg

SOULA Research Report 2004-1

Southern Oregon University Laboratory of Anthropology Ashland, Oregon 2004

The Hayes Site Oral Tradition, Ethnohistory, and Archaeology of the South Fork Coquille River

A Report to the Coquille Indian Tribe P.O. Box 783 3050 Tremont North Bend, Oregon 97459 by Mark Tveskov Department of Sociology and Anthropology Southern Oregon University With contributions by Nick Halousek, Nicole Norris, & Christian Solfisburg

April 22, 2004 Laboratory of Anthropology Southern Oregon University Ashland, Oregon 97520 SOULA Research Report 2004-1

© 2004 Southern Oregon University Laboratory of Anthropology Department of Sociology and Anthropology Southern Oregon University Ashland, Oregon 97520

ii

Abstract

Archaeological investigations were conducted by archaeologists from Southern Oregon University at the Hayes site (35CS196), located on a terrace overlooking the South Fork Coquille River in Coos County, Oregon. This work was conducted in collaboration with the Coquille Indian Tribe and was part of a larger effort to understand the distribution of archaeological sites in the South Fork Coquille River drainage and to place these sites into a historical, ethnohistorical, and oral tradition context. The Hayes site project was also designed in light of the severe looting that continues to take place at the site; in addition to learning about the size, depth, age, and cultural significance of the site, we hoped to document the extent of the damage to the cultural deposits and to use the project as a means to raise public and professional awareness about the impact of looting on archaeological sites such as the Hayes site. Work at the site took place for four weeks in the summer of 2002 during a Southern Oregon University archaeological field school. The Hayes site sits on an uplifted terrace on the west side of the South Fork of the Coquille River upstream from the modern-day community of Powers. The site is part of a large complex of archaeological sites that lies across Powers prairie to the north and Dingbat Flats to the south, all oriented around the confluence of Salmon Creek (flowing from the west) and Mill Creek (flowing from the east) with the South Fork. The local environment includes an extensive system of riffles immediately below the site ideal for salmon and eel fishing, an oak and grassy prairie/meadow complex on the site itself and in the surrounding foothills that was likely once rich in botanical resources such as camas and acorns. Finally, the site affords access--via ridgeline trails--to upland forests and meadows in the mountains surrounding the site. Ethnohistoric and ethnographic information suggests that Powers prairie in general and the Hayes site in particular sat at a hub of a major trail system that linked the ancient Athapaskan landscape of southwest Oregon. Trails passing through Powers led to the southern Oregon coast, the lower Rogue River region, the interior Rogue River Valley, and the Interior Umpqua Valley. Oral historic and ethnohistoric sources suggest that a local group of Coquille Indians known as the Nati jí dunne ("People Who Lived by the Big Fish Dam") lived in villages in the Powers/Hayes site area, and these villages were highly populated and wealthy prior to the incursion of Euro-Americans in the early 19th century. Archaeological surveys conducted by Southern Oregon University (Tveskov 2001) and interviews with local artifact collectors attest to the heavy archaeological footprint on the prairies in and around Powers. Southern Oregon University's work at the Hayes site included the excavation of 16 individual 50 cm x 50 cm shovel test pits, 12 individual 1 m x 1m excavation units as well as the extensive mapping of the site's natural and cultural features. This work resulted in the excavation of 9.25 m3 of sediment covering less than 0.1% of the site. A rich assemblage of artifacts was recovered during this project, including 5,615 pieces of lithic debitage, 81 pieces iii

of animal bone, 100 individual chipped stone tools, 19 individual groundstone artifacts, and the identification of two large cultural features comprised of burned river cobbles that likely represent the remains of roasting ovens. No organic material of unquestionable cultural association was recovered from the site for radiocarbon dating. However, the recovery of several artifacts from the site by local collectors reminiscent of Gunther Pattern assemblages from elsewhere in southwest Oregon suggests that the Hayes site was in use during the last 1,700 years (c.f. Connolly 1991). However, the excavated assemblage of leaf-shaped and Coquille Series projectile points similar to other Glade Tradition assemblages in southwest Oregon (including sites on the Coquille River such as the Standley site) suggest the possibility that the site was in use during earlier millennia. Given that the immediate site area was described by early White pioneers as being used by Indian people as late as the early 1850s, it is perhaps not unreasonable to surmise that the Hayes site was in use either continuously or intermittently for at least the last 2,000 years. Surface observation and the shovel test pit excavations indicate that the Hayes site covers an estimated 10,000 square meters. Cultural deposits are spread across several stepped terraces from the foothills to the west to the river on the east. Based on our excavations and the testimony of local residents, the richest and most diverse concentrations of artifacts are found on the higher (and, during the winter, dryer) terraces, especially on the north end of the current airport runway. Unfortunately, this portion of the site has been most severely damaged, as the runway has been extensively machine graded. Most of the Southern Oregon University excavations were concentrated on the terrace immediately below and east of the runway, and here, scatters of cultural material are found clustered around large burned rock oven features. This area has also been badly damaged, this time by extensive looting that in some areas appears to have completely obliterated cultural features, as well as by bioturbation (particularly rodent activity), plowing, and cliff face erosion. The Hayes site, like many archaeological sites on the South Fork Coquille River in particular and southwest Oregon in general, has been and continues to be severely looted. The site is easily accessible to the public and appears to be a favored location for causal artifact collecting as well as for more concentrated efforts to excavate artifacts for sale. Based on our excavations and information provided by local residents, I would estimate that conservatively, more than half of this site has been destroyed by looting, construction activities, bioturbation, and erosion. Nevertheless, conservation of the cultural material still intact at the site remains a concern: the Hayes site was looted immediately prior, during, and after our excavation. While cultural material was abundant across the entire site, our excavations suggest that in comparison with other southwest Oregon archaeological sites, there is not, overall, a very high density of artifacts, and the assemblage is likewise relatively simple. The most common artifacts were very small pieces of lithic debitage that, based on their size, material, and morphology, are indicative of the final preparation and repair of chipped stone tools such as projectile points. Very little lithic quarrying or early stage core reduction, if any at all, was occurring at the site. Locally available crypto-crystalline silicate (i.e. chert) was the most common material used, although significantly smaller quantities of imported obsidian and iv

some coarser materials such as basalt were used as well. The chipped stone tool assemblage is dominated by expediently used flakes and unifacial end scrapers, with smaller numbers of bifaces fragments and fragmentary projectile points (usually the bases) found as well. The assemblage of ground and pecked stone artifacts is likewise not very large, and is limited to expediently-used or minimally modified cobbles used as pestles, hammerstones, hopper mortar bases, and grinding slabs. It should be noted, however, that formal artifacts have been found at the site by collectors, particularly on the upper, western terraces at the site. These artifacts include an extremely elaborated groundstone adze handle and a decorated antler pipe collected from the site by Jackie Hofsess. The characteristics of the Hayes site, at least in the relatively intact areas on Terrace III around the burned rock oven features, are indicative of a relatively narrow range of activities. The lithic debitage and chipped stone tool assemblage suggest that the site was used as a base from which big game hunting took place. This assemblage is comprised primarily of broken projectile points, debitage resulting from the finishing and repair of stone tools, and a relatively limited range of bifaces and scrapers likely used to process killed game. The ground and pecked stone artifact artifacts would have been used in the processing of acorns, camas, and other plants, and these resources, in accordance with southwest Oregon oral tradition, would have been roasted on large earth ovens such as represented by the two burned rock features. The assemblage of artifacts recovered by both artifact collectors and the Southern University excavations fits the model of settlement and subsistence developed from our archaeological survey and oral history research (e.g. Chapter 2 and Tveskov 2001). Resource-rich river bank areas such as the confluence of Mill Creek, Salmon Creek, and the South Fork Coquille River were the focus of settlement and subsistence; there, permanent villages were established on upper terraces, above lower surfaces that were subject to seasonal flooding. The most artifact-rich (but most damaged) part of the Hayes site is on the airport runway terrace (Terrace IV) and the knoll to the west of the runway (Terrace V), and if there was a permanent village in this area, this is likely where it was. Such a village would have been owned by one or more families that would play host to their relatives at certain seasons of the year when the local resources that they controlled became available. In the case of the Hayes site vicinity, such seasonal activities likely included at least the spring and fall salmon fishing, summer eel fishing, spring camas harvesting, and fall acorn harvesting. At such times visiting households would camp around the host village, including on the lower terraces where most of our excavations took place. The archaeological remains in this area of the site are indicative of repeated use on a temporary, task specific basis.

v

vi

Acknowledgements

Our work on the South Fork Coquille River was completed and facilitated through the hard work, enthusiasm, and support of many people. The field work was completed by a Southern Oregon University archaeological field school directed by Mark Tveskov during the summer of 2002, and the recovered material was cataloged and analyzed over the following two school years at the Southern Oregon University Laboratory of Anthropology in Ashland under the direction of Nicole Norris (now a graduate student at the University of Oregon), Christian Solfisburg, and Eirik Thorsgard. The field crew included Nicole Norris (crew chief) and Barbara Buettner, Nick Halousek, Rory Jones, Mysti Kelly, Jessie Plueard, and Shawna Rider. Beth Hyman and Darby Keenan contributed to the final completion of this monograph. I would also like to thank Terry Knowles and Valerie Dean of SOU for their administrative support. The Hayes site project could not have been completed without the assistance, hospitality, and encouragement of the residents of Powers, Oregon. I would like to acknowledge in particular Jackie Hofsess, Gordie Hayes, Nancy Dupaquier, Jean LeCuyer, Joanie Bedwell, Peggy Davis, Sally Brock, and Kami Harter. Jackie in particular shared her considerable knowledge of the archaeology and landscape of the South Fork region with us, and her insights were invaluable to the project. I would also like to thank Janet Joyer of the Rogue River National Forest for the logistical assistance she provided for this project. The Project could not have been completed without the support of the Coquille Indian Tribe Cultural Resources Program and the Coquille Tribal Council. I would like to thank Don Ivy and Scott Byram in particular, who assisted, encouraged, and participated in the Hayes site work. The project was funded by the Coquille Indian Tribe and by a grant from the National Park Service Historic Preservation Fund to the Coquille Indian Tribe.

left to right, back: Rory Jones, Nick Halousek, Shawna Rider, Mysti Kelly, & Nicole Norris. Front: Jessie Plueard, Barbara Buettner, Pugsly, & Jackie Hofsess.

vii

viii

Table of Contents

Abstract.................................................................................................iii Acknowledgements....................................................................................v Chapter 1. Introduction...............................................................................1 Chapter 2. Environmental and Cultural Context...................................................5 Regional Geology and Biota.................................................................5 Ethnographic Background...................................................................7 Prehistoric Background..........................................................................10 South Fork Coquille River..................................................................15 Chapter 3. Site Setting, Field Methods, and Site Structure.......................................37 Site Setting....................................................................................37 Degradation of the Hayes Site......................................................................41 Sampling Strategies and Procedures......................................................44 Site Stratigraphy.............................................................................47 Vertical and Horizontal Patterning of Cultural Material.................................49 Cultural Features............................................................................49 Summary.....................................................................................55 Chapter 4. Chipped Stone Tools..................................................................58 Bifaces.......................................................................................58 Unifaces......................................................................................62 Summary.....................................................................................64 Chapter 5. Lithic Debitage.........................................................................65 Results.......................................................................................66 Summary.....................................................................................70 Chapter 6. Ground and Pecked Stone Tools......................................................71 Handstones..................................................................................71 Netherstones.................................................................................74 Pestles........................................................................................76 Manuports....................................................................................78 Unknown Artifacts..........................................................................78 Chapter 7. Faunal Remains.........................................................................81 Results........................................................................................81 Summary.....................................................................................83 Chapter 8. Summary and Discussion...............................................................85 Discussion...................................................................................88 References Cited.....................................................................................93

ix

List of Tables

Table 1. Debitage frequency and density from 50 cm x 50 cm.................................54 Table 2. Debitage frequency and density from 1 m x 1 m........................................54 Table 3. Metric Data for the Hayes site...........................................................58 Table 4. Metric attributes of bifaces form the Hayes site........................................60 Table 5. Metric attributes of uniface endscrapers from the Hayes site.........................62 Table 6. Metric attributes of retouched flakes and utilized flakes..............................63 Table 7. Material distribution by site area.........................................................66 Table 8. Lithic debitage grouped by material and reduction type...............................67 Table 9. Lithic debitage grouped by material and debitage class................................68 Table 10. Hayes site debitage frequency by raw material and flake size........................68 Table 11. Platform morphology by material......................................................69 Table 12. Attributes of Hayes site flakes with platforms.........................................69 Table 13. Faunal remains recovered from the Hayes site.......................................82 Table 14. Density of artifacts from southwest Oregon sites....................................87

List of Figures

Figure 1. The location of the Hayes site............................................................2 Figure 2. Groundstone grinding stone in a pothunter's back dirt..................................3 Figure 3. Riffle in the South Fork of the Coquille River..........................................6 Figure 4. Dingbat Flats...............................................................................7 Figure 5. The southern Cascadian region show location of major tribal groups.................9 Figure 6. Powers, Oregon, summer 2002.........................................................17 Figure 7. Possibly the first detailed map of the interior of southwest Oregon................20 Figure 8. Stanton's and Williamson's map of the Coquille River...............................21 Figure 9. Detail map of 1853 of northern California.............................................23 Figure 10. Map of southwest Oregon made by Lt. Kautz.......................................24 Figure 11. Chinese bronze vessel..................................................................25 Figure 12. Bottom of the Chinese bronze vessel..................................................25 Figure 13. Groundstone adze handle..............................................................29 Figure 14. Antler pipe stem........................................................................29 Figure 15. Basalt hammer...........................................................................30 Figure 16. Obsidian blade...........................................................................30 Figure 17. Basalt maul...............................................................................30 Figure 18. Obsidian knife............................................................................31 Figure 19. Idealized model of site locations......................................................32 x

Figure 20. Figure 21. Figure 22. Figure 23. Figure 24. Figure 25. Figure 26. Figure 27. Figure 28. Figure 29. Figure 30. Figure 21. Figure 32. Figure 33. Figure 34. Figure 35. Figure 36. Figure 37. Figure 38. Figure 39. Figure 40. Figure 41. Figure 42. Figure 43. Figure 44. Figure 45. Figure 46. Figure 47. Figure 48. Figure 49.

The South Fork Coquille River watershed............................................33 Artifact collector demonstrating digging technique.................................34 The South Fork of the Coquille River................................................38 View south from the airport terrace...................................................39 View west over looking Terrace IV...................................................40 View southwest across the airport runway............................................40 View north across Terrace III from feature 2 area...................................41 Looters holes located on the leading edge of Terrace IV...........................42 Artifacts collected from the Hayes site................................................43 Extensive krotovina and active rodent burrows.....................................43 Plan map of the Hayes site.............................................................45 Excavation underway in the Feature 2 excavation block............................46 Excavation in progress in the Feature 2 excavation block............................47 Stratigraphic plan of Unit 11, Unit 9, Unit 2, and Unit 10...........................48 Feature 1 excavation block, view south...............................................50 Plan of the Feature 1 excavation block................................................51 Feature 2 excavation block, view south...............................................52 Detail of Feature 2, Unit 9, view south................................................52 Feature 2 excavation block.............................................................53 Projectile Points from the Hayes site..................................................59 Bifaces from the Hayes site............................................................61 Distribution of material type for the Hayes site debitage.............................67 Size distribution of lithic debitage from the Hayes site..............................68 Hammerstone from the Hayes site (SOU-02.03-78).................................72 Abraders from the Hayes site............................................................73 Hopper mortar bases from the Hayes site..............................................75 Grinding slab recovered from pot hunter's hole at the Hayes site..................76 Pestles from the Hayes site..............................................................77 Possible net sinker or pendant from the Hayes site...................................79 Incised cobble from Feature 2 at the Hayes site.....................................79

xi

xii

Chapter 1. Introduction

The Hayes Site is located in Coos County, Oregon in the foothills of the Klamath Mountains just south of the present day community of Powers (Figure 1). The site sits on an uplifted alluvial terrace just south of where Salmon Creek (flowing from the west) and Mill Creek (flowing from the east) enter the South Fork of the Coquille River. The Hayes site was identified during a larger survey of the South Fork Coquille River area conducted by the Laboratory of Anthropology at Southern Oregon University and the Coquille Indian Tribe Cultural Resources Program, a project that included pedestrian survey of selected parcels of public and private lands, interviews with local residents knowledgeable about the history and archaeology of the vicinity, and reviews of ethnographic and ethnohistoric accounts of the South Fork area (Tveskov 2001). The site lies in an around Hayes Field, a small airport with a grass runway that serves the community of Powers, and was first recorded in the summer of 2000 by Mark Tveskov of Southern Oregon University and Don Ivy of the Coquille Indian Tribe, who had learned about it after talking with local artifact collectors. During their first visit to the site, Tveskov and Ivy observed lithic debitage, chipped, cobble, and ground stone artifacts, and burned rock in the spoil piles of a large number of looter's holes spread over the western end of the airport runway and in the grassy field to the north and east. Many of these holes appeared to have been dug recently, and our subsequent work at the site suggested that, like many other sites in the Powers/South Fork Coquille River area, the Hayes Site has long been and continues to be a target for the recreational and for-profit digging of artifacts (Figure 2). At that time, the site was given the designation 35CS196 by the Oregon State Historic Preservation Office (Tveskov 2001). In the report of our survey work in the South Fork area during 2000-2001, we concluded that the apparent size of the site, the observed density of cultural material, and the numerous artifacts found there by local people indicated that the Hayes site was highly significant and was likely to be eligible for inclusion on the National Register of Historic Places. The previous and ongoing looting at the site, however, has and continues to severely damage these archaeological deposits. To further assess the cultural significance of the site, thoroughly document the extent of the damage through looting as a means to monitor future impacts, and to provide a forum for raising awareness about looting for the general public, Tveskov conducted additional archaeological research at the site in the summer of 2002 through his Southern Oregon University archaeological field school. This project included the excavation of twelve individual 1 m x 1 m excavation units and sixteen 50 cm x 50 cm shovel test pits as well as extensive mapping of the site's

Figure 1. The location of the Hayes site in relation to the larger Pacific Northwest and northern California region.

2

Figure 2. Grinding stone in a pot hunter's back dirt at the Hayes site. topography and cultural and natural features (including the extensive evidence of illicit pot hunting). This work resulted in the excavation of 9.25 m3 of sediment covering less than 0.1% of the surface of the site. A rich assemblage of artifacts was recovered during this project, including 5,615 pieces of lithic debitage, 81 pieces of animal bone, 100 individual chipped stone tools, 19 individual groundstone artifacts, and the identification of at least two cultural features. These cultural materials were subsequently returned to the Laboratory of Anthropology at Southern Oregon University for analysis, where they are now curated under accession SOU-2002.03. This monograph is the final report on the Southern Oregon University investigations. In Chapter 2, I review the environment of the South Fork Coquille River/Klamath Mountain region, review our current understandings of the ethnography of the Indian people of southwest Oregon in general and the South Fork area in particular, outline archaeological models of the region's prehistory, and summarize our previous survey work that together guided our analysis of the Hayes site material. In Chapter 3, we describe the natural setting of the site and its immediate environment, the field methods used in our investigations, and the horizontal and vertical (stratigraphic) structure of the site deposits, including the extent to which the site has been impacted by looting. In subsequent chapters we consider the different classes of artifacts and ecofacts that were recovered from our excavations: Chapter 4 discusses the chipped stone artifacts, Chapter 5 examines the lithic debitage, Chapter 6 presents an analysis of the ground- and pecked- stone

3

material, and Chapter 7 considers the site's small faunal assemblage. Finally, in Chapter 8, we summarize our findings and consider how our work at the Hayes Site can inform our current readings of the prehistory of Southwest Oregon. The Hayes site is a rich deposit of cultural material that appears to represent the remains of repeated Native American use of this location over a long period of time. The site is a palimpsest of material remains left over from many small occupations, and the profusion of small bits of chipped stone debris, broken arrow and spear points, and expediently manufactured heavy groundstone tools bespeak how members of individual households traveled to this spot on a seasonal basis to gather plants and hunt in the meadows and forests that surround the site and perhaps fish from the South Fork Coquille River itself. The exact chronology of these occupations remains unknown; the site is heavily disturbed by rodent activity and illegal digging, and we were unable to obtain samples of organic material of undisputed cultural origin suitable for radiocarbon dating. However, artifacts diagnostic of Gunther Pattern assemblages (c.f. Connolly 1991, 1986) collected from the site by local residents suggest that the site was in use during the late Holocene period after 1,700 A.D. When people first occupied the site remains unknown, but our recovery of chipped stone tools diagnostic of the Glade Tradition could indicate that the earliest use of the site began several thousand years ago. The Southern Oregon University and Coquille Tribe work at the Hayes Site contributes to ongoing research in southwest Oregon that offers a growing appreciation of the dynamics of household social experience in the prehistory of the region's Indian people (Tveskov 2003; Tveskov and Erlandson 2003; Byram 2002; Tveskov et al. 2002; Tveskov 2000; Winthrop 1993; Connolly 1991, 1986). Oral traditions, ethnohistoric accounts, and archaeological data indicate that the productive activities and the social experiences of the Coquille people and their neighbors focused on individual households. These activities and experiences have a deep time depth in southwest Oregon and northern California, and remained fundamental to the cultural landscape even after the dramatic cultural and sociological transformations that occurred during the late Holocene. After 2,000 years ago, southwest Oregon and northern Californian societies developed complex and ranked social relations, settlement patterns centered around large permanent plank house villages, elaborate trade networks, subsistence practices centered on the seasonal mass harvest, storage, and public redistribution of salmon and sea mammals, and possibly higher population densities (Tveskov 2003; Winthrop 1993; Connolly 1991; Lyman 1991). Although this was a local manifestation of similar processes at work up and down the West Coast (described as the evolution of the Developed Northwest Coast Pattern by Matson and Coupland 1995; see also Ames and Maschner 1999; Chartkoff and Chartkoff 1984), the causes of this development remain unclear. While possibly a response to rising populations regionally, the immigration of Athapaskan- and Algic- speaking people into southwest Oregon and northern California after 1,700 B.P. may also have been a factor.

4

Chapter 2. Environmental and Cultural Context

Regional Geology and Biota

The Hayes site is located in southwest Oregon on an uplifted alluvial terrace overlooking the South Fork Coquille River, one of the four major tributaries of the Coquille River (Figure 3). The South Fork Coquille River originates high in the Klamath Mountains near the divide with the Rogue River drainage to the south and flows north, eventually entering the main stem of the Coquille River at the confluence of the Middle Fork Coquille River near the present day community of Myrtle Point, Oregon. The upper portions of the South Fork Coquille River lie in the Siskiyou National Forest, and there the river runs first southwest, paralleling the Rogue River and cleaving through the Klamath Mountains before turning a sharp corner north around Eden Ridge. Along this portion of the river, the canyon is steeply incised, often rising over 80 meters within a single kilometer to either side. The Klamath Mountains, forested with Douglas fir (Pseudotsuga menziesii), cedar (Thuja spp.), hemlock (Tsuga spp.), and alder (Alnus spp.), are the remains of Jurassic and Paleozoic landscapes that have been faulted and up-ended by tectonic activity (Orr et al. 1992). These terranes are interspersed by igneous and metamorphic formations that include serpentine, crypto-crystalline silicate, schist, coal, gold, and other rocks and minerals that Indian, Chinese, and Euro-American people have found useful over the millennia. These are often found in colluvial and alluvial deposits in and around the South Fork Coquille River itself. Just south of the town of Powers, the river leaves federal land and runs north through a series of fairly wide prairies (Figure 4). These prairies, which often are the sites of modern-day communities such as Powers, Gaylord, Warner, and Broadbent, are large deposits of Pleistocene- or Holocene-age alluvial sediments that often tower 20 meters or so above the river (Tveskov 2000a:30; Personius 1993). In some areas, such as at the Hayes site, these alluvial deposits sit on bedrock that is itself dramatically faulted and uplifted, resulting in steep and at times vertical canyon walls of rock and stratified sediment. Secondary terraces of mixed colluvial and alluvial sediments are also found on the prairies that border the South Fork Coquille River, which grade back into the surrounding foothills and mountains. Today, these prairies are open and grassy, although research into 19th century General Land Office survey notes suggest that such flood plains along the Coquille River were wetter and more heavily vegetated prior to the immigration of Euro-Americans and the introduction of farming and ranching into the area. The secondary terraces and surrounding hillsides typically contain oak prairies, reflecting the more mesic conditions of these rainshadowed valleys, the larger transitional position of the South Fork area between the Northwest Coast and California bioregions, as well as the likely relic influence of traditional Native American burning practices (Tveskov et al. 2002). Prior to Euro-American contact, the flood plains, secondary terraces, and surrounding hillsides would have supported acorn

Figure 3. Riffle on the South Fork of the Coquille River immediately upstream from the confluence of Salmon Creek. The Hayes site is located on the river bank on the right behind the trees. bearing oak trees (Quercus spp.), camas (Camassia quamash), tarweed (Madia spp.), epos (Perideridia spp.), elk (Cervus elaphus), deer (Odocoileus spp.), and a wide variety of other flora and fauna used by Indian people of Southwest Oregon. The South Fork Coquille River at one time contained abundant runs of salmon1. Coho (Onchorhynchus tshawytscha) and Chinook (O. kisutch) salmon ran in the fall, but the South Fork is unique in the Coquille drainage for having a substantial spring run of Chinook salmon as well. This may be because the South Fork, in comparison to the Middle, East, and North Forks that drain from the Coast Range, emerges from the taller Klamath Mountains and has a correspondingly higher spring run off. Steelhead (O. mykiss) are also found in the South Fork, particularly in the winter. Eels (Lampetra) were also common, particularly in the summer when they ascended the river to spawn.

Salmon and eels still run on the South Fork; one local resident explained to one of the crew members on the 2002 Archaeological Field School that a preferred local method of catching salmon was "with a pitchfork."

1

6

Figure 4. Dingbat Flats--the prairie located upstream and on the opposite bank of the South Fork Coquille River from the Hayes site. The south end of the airport runway lies on the bluff visible in the middle left of the photograph; note the elevation difference between this terrace and the site terrace on the opposite bank.

Ethnographic Background

Southwest Oregon and northern California lie at the southernmost part of the Northwest Coast culture area. However, until recently, despite the grand tradition of anthropological research in the larger region, ethnographic understandings about the Indian people of the region in general and the Coquille River drainage in particular remained at a basic level (Wasson 2001; Tveskov 2000; Drucker 1937). This was, in part, a result of the death of the majority of the region's people through introduced Old World diseases and by the subsequent removal of most of the survivors to the Coast Reservation on the northern Oregon coast in 1856 following the Rogue River War (see Tveskov 2000). Decades later, anthropologists interviewed a bare handful of Indian people, mostly on the coast Reservation or its splinters--the Siletz and Grand Ronde Reservations--resulting in a handful of primary ethnographic analyses. Of these publications, Barnett (1937), Drucker (1937), Jacobs (1940, 1939), Harrington (1942), and Waterman (1925) are the most detailed. 7

In recent years, University and Tribal scholars have revisited the field notes left behind by this earlier generation of anthropologists and have provided more detailed ethnographic and historical analyses. At the University of Oregon, the Coquille Indian Tribe and the Native American Program at the Smithsonian Institution have collaborated on the Southwest Oregon Research Project (SWORP), sending Tribal and non-Tribal scholars to Washington D.C. to the National Archives and the National Anthropological Archives to identify primary documents from the BIA, the U.S. Military, and the Bureau of Ethnology relevant to the Tribes of northern California and western Oregon. Since 1995, over 100,000 pages of primary documents have been recovered. Over the last eight years, several historical, ethnographic, and prehistoric analyses utilizing this archive have appeared and shed light on the history and culture of the people of southwest Oregon (e.g. Younker 2003; Byram 2002; Wasson 2001; Tveskov 2000; Hall 1999). My own research considers the traditional social relations, economy, and cultural geography of the Coos and Coquille people and their neighbors (e.g. Tveskov 2003, 2002, 2000; Tveskov et al. 2002). Like other societies of northwest California or the Northwest Coast, the Coquille people were traditionally characterized by their complex social arrangements, high population densities, and maritime hunting, fishing, and gathering ecology. They lived in often large plank house villages located along major rivers, estuaries, and coastal headlands, and this individual village and its constituent households served as the fundamental political, social, and economic unit (Hall 1992:182; see also Tveskov 2002; Barnett 1937:185; Drucker 1937). Despite their basic independence, villages were linked into larger webs of cultural, political, and economic social relations that started with individual households and included loose associations of local villages (known as Yetl-ein in Tolowa, see Lewis 2001) as well as far-flung networks of trade and intermarriage that stretched across and beyond the entire region (Figure 5)(Tveskov 2002; Tveskov et al. 2002; Drucker 1937). Southwest Oregon and northern California Indian societies were transegalitarian, with each village headed by one or more powerful families, each led by a chief. The largest seasonal aggregations of many households took place at large villages associated with exceptionally rich resource patches such as sea mammal rookeries on the outer coast, major fisheries along estuaries, or upriver at major salmon fisheries. While the periodic mass harvest of such resources was crucial in the ecology of the people of the region, their importance was also social: They were public arenas where individuals and families participated in social relations that went beyond the individual household. There, households negotiated, confirmed, challenged, and otherwise negotiated their identity, rank, and status through arranging suitable marriages for their children, organizing, directing, or participating in the gathering and distribution of surplus commodities, sponsoring or participating in gift giving or feasting rituals, engaging in trade for non-local or exotic goods, or negotiating political or spiritual alliances (Tveskov 2002, 2000; Tveskov et al. 2002, c.f. Swezey and Heizer 1993, Hayden 1995). Large villages and seasonal gatherings at these villages were public arenas for political action among and between multiple households. However, in the social landscape of the Coquille and their neighbors, the individual household maintained a high degree of political independence and economic mobility (Byram 2002; Tveskov et al. 2002; Tveskov 2002, 2000). In this they differed to some degree from the powerful chieftaincies of the far northern Northwest Coast or the Columbia River region (e.g. Ames 1994, 1996). Among 8

Figure 5. The southern Cascadia region showing the location of major tribal groups (shown in larger font) and the yetl-ein (local groups, shown in smaller font) of the Pacific Coast Athapaskans of southwest Oregon.

9

the Chinook, for example, chiefs wielded power over large estates and many households, tied together under one roof in large, multi-family plank houses of grand architectural scale (Ames and Maschner 1999). In contrast, southwest Oregon families usually owned their own houses, even when living in a large village nominally controlled by a particular powerful chief or household. This primacy placed on household independence was described by Drucker (1937:243-244) for the Tolowa, where the basic social `unit' was: extremely simple, consisting of a group of paternal kindred, who inhabited a town and shared in the right to exploit resources of the surrounding territory. Before the destructive effect of Caucasian culture made itself felt, genealogies show that nearly all the male inhabitants of a town were related . . . . If a lineage prospered and multiplied for some generations, the kinship bond weakened and the people gravitated around the stronger subgroups, or rather around the leading men of the branches . . . . Or a man might move, after a family unpleasantness, establishing a settlement at some camp site. . . . Once established, a division or suburb might thrive, coming to rival the parent town in importance. The political and social independence of individual families was also maintained by the economic importance of household subsistence practices (Tveskov and Erlandson 2003; Byram 2002; Tveskov et al. 2002; Lewis 2001; Tveskov 2000; Hannon 1990). Anchored by permanent plank house villages located at rich resources patches on the outer coast, on the estuaries, or on terraces overlooking the main stems of the Coquille River, families followed a complex seasonal round of activities across the landscape to harvest camas and other geophytes, acorns, berries, epos, tarweed, lamprey eels, salmon, deer, elk, and other resources. Where these resources were particularly or seasonally abundant--such as at fisheries on the main river trunks, in estuaries, adjacent to pinniped haulouts on the outer coast, or on prairies rich in camas or acorns--these seasonal camps were often located adjacent to permanent villages owned by kin or other allied families. At other times of the year, household camps were located in distant areas adjacent to hunting grounds, fishing and shellfish gathering areas, or upland meadows particular rich in plant resources.

Prehistoric Background

The prehistory and origins of the people of southwest Oregon and northern California are poorly understood. Research has historically been biased toward the excavation of large village sites or highly visible shell middens at the expense of upland or interior valley sites, and archaeological research is further challenged by dramatic geological processes that hide or obliterate many sites predating the late Holocene (Tveskov and Erlandson 2003; Tveskov et al. 2002; Tveskov 2000; Erlandson and Moss 1999; Lyman 1991). Analyses of the region's prehistory have focused on building cultural chronologies based on diagnostic artifacts (Connolly 1986; Pettigrew and Lebow 1987; Pullen 1982), finding very ancient sites (Erlandson and Moss 1999), explaining the timing and nature of the Athapaskan and Algic entrée into the region (Tveskov et al. 2002; Connolly 1991, 1986), reconstructing diachronic changes in settlement patterns over the Holocene (Connolly et al. 1994; Winthrop 1993; Connolly 1991; Lyman 1991; O'Neill 1989; Pettigrew and Lebow

10

1987), and reconciling the archaeological record with oral traditions of the Native people of the region (Tveskov and Erlandson 2003; Tveskov et al. 2002; Byram 2002; Lewis 2001; Tveskov 2000). The late Holocene archaeological record of the region is the most comprehensive, and provides at least a 1,700 year time depth for cultural practices largely similar to those described in oral traditions, ethnohistoric documents, and ethnographies, as summarized above (see also Younker 2003; Byram 2002; Tveskov et al. 2002; Tveskov 2000). Oral traditions regarding traditional plank house construction as well as the excavations of lateprehistoric or early historic-era villages at Point St. George in northern California (Gould 1966), Lone Ranch on the outer coast of Curry County, Oregon (Berreman 1944), on Flores Lake on the coast south of the Coquille River (Newman 1959), the Ritsch Site (Wilson 1979) and the Marthaller site (Steele 1984) in the interior Rogue River valley and at site 35JA42 on the Applegate River (Brauner and Lebow 1983) show that, in contrast to the large multifamily houses common along the northern Oregon coast and the lower Columbia River. The Athapaskan- and Penutian-speaking people of northern California and southwest Oregon constructed small, single family plank houses. These sites, as well many other late Holocene and early historic-era sites in the region, contain a suite of artifacts described either as the Gunther Pattern along the coast or the Siskiyou Pattern in the interior valleys (Connolly 1991, 1986; Frederickson 1973). As described by Connolly (1991:12-13), these late Holocene assemblages include: barbed projectile points, bone harpoon points, triangular concave base harpoon tips of chipped stone, shaped stone mauls, flanged and offset pestles, notched and grooved net weights, shallow steatite grease bowls or lamps, zoomorphic stone clubs, large red and black ceremonial obsidian blades, baked clay figurines, and a variety of bone and shell ornaments. Such artistically elaborated artifacts, often recovered from permanent villages located at environmentally rich ecotones--promontories on the outer coast that overlook rich shellfish beds and pinniped haulouts, the margins of estuaries, or on dry terraces overlooking rich salmon fisheries in up river locations--bespeak their role in the feasting, gift giving, and other aggrandizing behaviors that took place within the social experience of seasonal, multihousehold aggregations at such sites, as described in oral histories and ethnohistoric sources. Obsidian debitage is also common in late Holocene archaeological sites, and its presence bespeaks the development of extensive trade networks between households of southwest Oregon and northern California and their neighbors across the Cascade Range to the east. The ethnographic pattern of trade is also reflected in the assemblage from the Gold Hill site, the only large cemetery excavated in southwest Oregon. In 1930, Luther Cressman of the University of Oregon recovered 39 whole or partial human skeletons from a natural mound of alluvial sediments located adjacent to the Rogue River in the interior Rogue River valley. Bodies were interred on their left side, head to the south, facing west, with their legs flexed against their chest (Cressman 1933:119). Grave goods included projectile points, broken mortars and pestles, schist and steatite tobacco pipes, as well as other items manufactured from material obtained through trade west from the coast or east across the Cascade Range including obsidian wealth blades and Olivella shell and pine nut beads. 11

In recent decades, archaeological surveys and excavations, usually conducted in the course of cultural resource management work, have identified an abundance of smaller, task specific, or seasonal camp sites, often located in upland settings. This work underscores the importance of household mobility and economic practices, as described in oral traditions (e.g. Tveskov and Erlandson 2003; Tveskov et al. 2002; Byram 2002; Lewis 2001; Tveskov 2000; Connolly 1991; Winthrop 1993). Household middens along the coast or along estuary margins contain enormous quantities of shellfish remains and the bones of herring, smelt, and other small fish. As determined through fine-grained analysis of bulk midden samples and a technological analysis of nearby estuarine fish weirs, household activities at such sites supplied the majority of the subsistence requirements of a family over the course of a year. Household camps served additional and important social functions. Unlike village sites that were often occupied for relatively short periods of time, household camps and wood stake fish weirs--the domain of household social and economic experience-- frequently demonstrate persistent use and reuse over millennia. The repeated use of such sites and the physical manipulation of the surrounding landscapes created a stability of meanings that served to reify and confirm the importance of a household's social identity. Recent excavations and analyses of the Windom site, located in the western foothills of the Cascade Range in the upper Rogue River valley, for example, suggest that the site was continually reused for at least the last 2,000 years for the hunting of big game and the gathering camas, acorns, and other plant resources (Tveskov et al. 2002). The site is situated on the edge of an upland oak prairie that was created and maintained by anthropogenic fire. The significance of the site for the households that used it was clearly tied to the economic importance of the resources that were fostered and harvested there, but over successive generations, this significance was recursively amplified through the habitus of day-to-day life away from the larger, more public villages. Over time, the place itself became wrapped into the biography and tenure of specific groups of people. Given the increasing complexity of social and political relations in southwest Oregon over the last 2,000 years, the role of household camps and social experiences likely became ever more important as social forces were mounted against that sense of identity and the fundamental autonomy enjoyed by southwest Oregon and northern California households. It is tempting to view the Gunther and Siskiyou Patterns as evidence of the final evolution of the elaborate social relations described historically for the region. Indeed the material culture, settlement patterns, and architecture represented by these archaeological cultures are essentially indistinguishable from those described ethnographically. Ceremonial obsidian blades, zoomorphic stone clubs, Dentalium shell beads, and other `fancy' artifacts were certainly used to negotiate the social relations involved in building and maintaining an individual's or family's status, as was considered vital to a wealthy family according to oral traditions. Likewise, the sea mammal hunting, mass fish harvesting, and extra-regional trade in obsidian inferred from these late Holocene archaeological assemblages provide a time depth for the economic basis of the historically described aggrandizing behavior associated with the social experience at permanent villages. Why and exactly when this local manifestation of the Developed Northwest Coast Pattern (after Matson and Coupland 1995) evolved remains unclear. Connolly (1991:12-13, 1989, 1986) suggested that the relatively sudden appearance of Gunther and Siskiyou Pattern 12

"without any identifiable local ancestor" was the result of the immigration of new, i.e. Algicor Athapaskan-speaking people into the region. The linguistic evidence, however, is not completely congruent with the appearance of these assemblages after 1,700 B.P. Hoijer (1956, 1960) suggested on the basis of glottochronology that the Pacific Coast Athapaskan languages separated from northern Athapaskan only 1,300 years ago. Similarly, the Algic languages of the Wiyot and Yurok of northern California likely split from their eastern relatives after A.D. 900. Alternatively, Winthrop (1993) modeled Holocene cultural changes in southwest Oregon from a processual point of view: Drawing on Binford's (1980) distinction between foraging and collecting settlement systems, she typed 92 archaeological site assemblages and components from southwest Oregon into `task specific,' `seasonal camp,' and `village' categories and ordered them in time based on their constituent styles of projectile points. Based on this analysis, she suggested that settlement patterns were at first highly mobile, but with a hypothesized steady growth in the population, the region's cultures began to approach their historic configuration as semi-permanent villages appeared along the main stems of the region after 4,000 years ago and along tributary streams after 2,000 years ago. Resolution of the timing and nature of Holocene cultural change is further confounded by the dramatic geomorphology of the region. The relatively `sudden' appearance of the Gunther and Siskiyou Patterns and the relative dearth of middle and early Holocene sites, may, to some extent, be the result of taphonomic effects. The archaeological record prior to the late Holocene is difficult to access; along the coast, tectonic uplift periodic subsidence, and dramatic erosion combine to destroy ancient surfaces, and in the interior valleys, ancient terraces are often buried deeply beneath meters of alluvium (Tveskov and Erlandson 2003; Erlandson and Moss 1999; Personius 1993). Although a handful of early and middle Holocene sites have been found, they remain rare. Furthermore, the frequent occurrence of natural and anthropogenic fire as well as local soil conditions that foster bioturbation from root, worm, and rodent activity hamper the preservation and obscure the context of organic material in archaeological sites. Archaeologists working in southwest Oregon and northern California are typically confronted by palimpsests of undifferentiated material, more often than not of late Holocene age. Consequently there are often questions about the cultural association of the burned wood used to radiocarbon date archaeological components.

Continuity and Persistence

Despite their differences, Winthrop's and Connolly's models are not completely incompatible. Whenever the migration of Athapaskan- and Algic-speaking people into southwest Oregon and northern California occurred, these events must have had a dramatic impact on the social landscape. At the present level of our knowledge of the archaeological record, it is likely best to envision these migrations as one of many possible factors that, during the late Holocene, spurred on local societies to their historical configurations, a process that was occurring at the same time in generally similar ways across the Northwest Coast and California culture areas (Ames and Maschner 1990; Matson and Coupland 1995; Chartkoff and Chartkoff 1984). It is perhaps significant to note that despite the acknowledged cultural differences between the Penutian-, Athapaskan-, Hokan- and Algicspeaking people that lived in southwest Oregon and northern California, they indeed shared

13

many cultural similarities. By the time of European contact, they had blended together to a large degree, particularly in regards to economic practices and social relations centered on fundamental household independence. This blending is also evident archaeologically. Connolly (1991, 1989, 1986; see also Connolly et al. 1994) points out that archaeological assemblages recovered from the few well-dated middle and early Holocene assemblages from southwest Oregon and northern California are characterized by broad-necked side notched and stemmed projectile points, foliate or contracting stem points, edge-ground cobbles, thick unifacial endscrapers, hammer and anvil stones, and stone bowls. Connolly classified these assemblages as the Glade Tradition, and further points out that such assemblages appear to continue to occur in late Holocene contexts, often in the same components as Gunther or Siskiyou Pattern material. The long temporal span of the Glade Tradition suggests that despite the immigration of Algic- and Athapaskan-speaking people in the late Holocene, the historically described societies of the region maintained very deep cultural roots. Hence, according to Connolly (Connolly et al. 1994:165): It appears from present evidence that much of the region's human history can be characterized by relatively low population densities and a reasonably stable and abundant food supply. The anthropological literature frequently attributes cultural changes to causal factors such as population circumscription, environmental pressure, or contact with other cultural systems. It is possible that in the absence of such factors, the natural inertia of successful cultural traditions may go towards explaining the unusual degree of cultural stability and conservativism in evidence in southwest Oregon. The formal similarities in cultural assemblages of widely disparate ages (from Pre-Mazama to late prehistoric times) from the region has been identified . . .as the Glade Tradition. It is noted here that acknowledgment of these formal similarities through time points to a strong element of technical continuity and stability in the region; recognizing this continuity should not inhibit our delineation of important cultural patterns including changing settlement modes and technological innovations. The emergence of the Gunther Pattern and the Siskiyou Pattern after 2,000 years ago, inferred to possibly be the result of Athapaskan and Algic immigration into the region, is one such important cultural and historical change that perhaps led to the evolution of the ethnographically described cultural pattern. New cultural traits (both material--barbed projectile points and obsidian wealth blades--and social--intensive trade and the mass harvest, storage, and redistribution of salmon) appear to have been added onto those of the Glade Tradition. The historically described people of southwest Oregon and northern California, despite the likely immigration of new peoples from the outside, maintained indigenous roots from deep in the Holocene. If, as seems likely, Glade Tradition sites continue well into the late Holocene and are found contemporaneously with Gunther or Siskiyou Pattern assemblages, it is at least possible that it is the ethnographically well-described mobility of individual households and the value put on their independence that is the element of stability posed by Connolly. 14

Given the dynamics of household social experience described in oral traditions and in the archaeological analysis of wood stake fish weirs, household middens and hunting and gathering camps (e.g. Tveskov and Erlandson 2003; Byram 2002; Tveskov et al. 2002; Tveskov 2000), it would have been the actions of individual members of such households acting on that cultural value that was the agency of its long term conservatism. Indeed, it is often household social experience, rather than outward political contact, that is the agent of social stability over the longue durée witnessed by the archaeological record (e.g. Lightfoot et al. 1998). The long term stability of household mobility and social experience is supported by functional interpretations of Glade Tradition sites that suggest they represent the material remains left by small groups of highly mobile hunter gatherers who do not appear to have intensively utilized salmon fisheries or (on the coast) sea mammals (Connolly 1991, 1986; Lyman 1991).

Contact Period Culture Change

Anthropologists have also considered the interaction between Indian people and European or Euro-American explorers and immigrants in southwest Oregon (Erlandson et al. 2000; Tveskov 2000; Draper 1988). Europeans first explored the coast of southwest Oregon in the 16th century, but sustained interactions between Indians and Whites did not occur until after the opening of the Pacific Northwest maritime fur trade in the mid-18th century. Even then, the Indian people of the region may have participated only tangentially in this trade. The northern California and Oregon coasts, without deep or sheltered anchorages, largely defeated the European maritime technology prior to the mid-19th century, and it was until after 1825 that land-based fur trading expeditions entered the region on a regular basis. Nevertheless, the early effect of European contact was profound. Oral traditions and ethnohistoric sources suggest that introduced disease epidemics such as small pox, measles, and malaria swept the region as early as the first decades of the 19th century. Archaeological evidence suggests the possibility that even earlier epidemics occurred. During the proto-historic period (1492-1790 A.D.), major village sites across southwest Oregon and northern California were abandoned, their plank houses often burned down, perhaps as a result of the death of their inhabitants through epidemic disease (Erlandson et al. 2000; Tveskov 2000; Draper 1988). It appears that upland and outer coast villages may have been abandoned during this time, and the survivors amalgamated into fewer, but perhaps larger villages located along the main river trunks in interior valleys and along estuaries on the coast.

The South Fork Coquille River Oral Tradition and Ethnohistory

The South Fork Coquille River area is traditionally considered home of a portion of the Athapaskan-speaking Mishikwut´me-dunne ("People living along the River Mishi") or Upper Coquille people (Tveskov 2000; Hall 1992; Harrington 1942). They were thus part of a larger distribution of Pacific Coast Athapaskan speakers that included the Tolowa of the northern California coast, the Tututni of the northern California and southern Oregon south of the mouth of the Coquille River, The Galice and Applegate groups that lived along the

15

middle riches of the Rogue River up to the mouth of the Applegate River, and the Upper Umpqua who lived along the central portion of the Umpqua River, from the present day town of Scottsburg to the confluence of Cow Creek and the South Fork Umpqua River (Tveskov 2000; Connolly 1986; Miller and Seaburg 1990). The Pacific Coast Athapaskans shared southwest Oregon and northern California with the Penutian-speaking Hanis and Miluk Coos and Lower Coquille who lived along the coast on Coos Bay and at the mouth of the Coquille River, The Kalowatsett and Cow Creek Umpqua that lived on the Umpqua River, the Penutian-speaking Takelma of the interior Rogue River valley, and the Algicspeaking Yurok and Wiyot and the Hokan-speaking Shasta that lived along the middle reaches of the Klamath River in northern California. Information about the Indian people that lived along the South Fork Coquille River is limited to testimony of Coquille Thompson and several other Coquille elders interviewed by anthropologists during the late 19th and early 20th centuries (Tveskov 2002, 2000). The area appears to have been the home of the Chochrela dunne who lived at and around the confluence of the South and Middle Fork Coquille River and the Nati jí dunne, who lived up stream around the vicinity of present-day Powers. The Chochrela dunne and the Nati jí dunne, along with the Set`luim´dunne who lived in Camas Valley on the upper reaches of the Middle Fork Coquille River, the Xwe´c dunne who lived between the present-day town of Arago and the mouth of the North Fork Coquille River, and the Lænxas dunne who lived further down stream around the present-day town of Coquille, were the five yetl-ein (the Tolowa word for a loose cultural, political, and economic association of villages) of the predominantly Athapaskan-speaking Upper Coquille (Mishikwut´me-dunne) people (Tveskov 2000:65; see also Harrington 1942[26]:202-216; M. Jacobs 1931-34[104]; E. Jacobs 1933-34[119]:120-131; Dorsey 1890:232). Today, the town of Powers sits at the end of a winding highway that is often closed by mud flows, a road that is unimproved as it crosses over the divide to the Rogue River (Figure 6). The southerly part of this road is often closed in the winter. From our contemporary perspective, the South Fork Coquille River is hard to access and is at the end of a dead-end road. Powers is consequently a small community located far from the more populated towns of Coos County. This marginal position, however, belies the importance of the area in the traditional cultural, political, and economic landscape of the Pacific Coast Athapaskans and their neighbors. The Nati jí dunne villages appear to have been at one time heavily populated and prosperous. Coquille Thompson recalled that his own people referred to the Nati jí dunne as the Dæns´h`dunne, or "Little Higher People," a double entendre referring to both their relative position on the Coquille River as well as their power and wealth. Indeed, Coquille Thompson also recalls that the Chochrela dunne, who in the mid19th were led by Thompson's father Chief Washington--one of the most powerful of the pre-conquest Coquille chiefs--often married their daughters into Nati jí dunne families. The name Nati jí dun itself means "Big Fish Dam Place," suggesting that this yetl-ein collectively controlled access to productive salmon fisheries. One of these fisheries is described by Wasson (1991: see also Harrington 1942[26]:188), who presents a story told by Coquille Thompson about Thetsuwulletslu-dun ("Spherical Stones Besides the Water"), one of the villages of the Nati jí dunne yetl-ein on the South Fork Coquille River. In this story, Thompson relates how Coyote attempted to ram through a fish dam, placing two large stones in his canoe for added momentum. He was unsuccessful, and the village thus obtained its name as his canoe overturns and the rocks are dumped in the river. 16

Figure 6. Powers, Oregon, summer 2002. It is possible that the Nati jí dunne and the Chochrela dunne owed their wealth to not only the rich salmon and eel fisheries that they controlled--including the only sizable spring run of Chinook salmon in the Coquille River system--but also to their position at a central hub of a number of trails that linked together the various Pacific Coast Athapaskan groups that lived along the middle and lower reaches of the Rogue River, the Oregon and northern California coasts south of the Coquille River mouth, and the interior Umpqua River valley. Such a trail system is described in oral traditions (e.g. Harrington 1942[26]:185188), but are best detailed in ethnohistoric sources that describe the European and Euro American use of these trails. These sources also provide a context for understanding the interaction between the Nati jí dunne and these newcomers. The first White visitors to the South Fork Coquille River were members of an expedition led by Alexander McLeod, who explored southwest Oregon during the winter of 1826-27 for the English Hudson's Bay Company (Tveskov 2000; Davies 1961). From a base camp near the present-day town of Coquille on the main trunk of the Coquille River, McLeod and his men ascended the South Fork, and his journal entry of December 11, 1826, describes his arrival at the Nati jí dunne villages on Powers prairie (Davies 1961:196-197): As soon as day light enabled us to see our way we moved forward, after passing a short belt of wood we opened into a fine plain at the extremity of

17

which, we came to a village of five dwellings rather unexpectedly. Our sudden appearance amazed the inhabitants who had not observed us, till we reached their door. Their fear was soon dissipated; we obtained some dried salmon, indifferently cured for which they got in return a few trinkets. My men had took their breakfast and by means of canoes we forded the river, about 50 yards wide--continued on our journey on the east bank about five miles and reached another village greater and more populous than the last. Here, the river assumes a different aspect; it becomes rocky, with many cataracts, some perpendicular falls, that afford the means of spearing the salmon trout, our guide made some objections for going further alleging that we could not reach the great river [i.e. the Rogue River] (for by that time, it appeared evident to him, our intention led that way) owing to the high state of the water, all the others supported the argument, yet persisted in going forward, tho' aware we could not go far, not being prepared for such difficulties the natives presented, but still no impediment as yet afforded to obstruct our progress, with a great deal of persuasion our guide assented to proceed as we gave him to understand, our intention was merely to see the country, we would defer till a ore favorable period to visit the Great River, which these people are in the habit of visiting from hence in summer reached in two days . . . . From this account, we have `populous' villages located on the South Fork Coquille River in 1826, villages whose inhabitants maintained contact with neighbors living south over the divide on the Rogue River. These were likely permanent villages; as the time of McLeod's visit, the fall fishery would have been completed, and the Coquille people would have been settling into the plank houses of their home villages for the winter. The trail over the divide to Rogue River described by McLeod was encountered from the other direction by the American T'Vault expedition of 1851. William T'Vault and his companions were attempting to blaze a trail up the Rogue River from the coast to the Oregon-California trail (that ran north-south through the interior Umpqua and Rogue river valleys), as part of William Tichenor's ambition to turn the newly established American settlement of Port Orford into a supply port for the gold mining operations then sprouting up in the interior (Tveskov 2000; T'Vault 1851; Williams n.d.). T'Vault's party started eastward from the coast on ridge lines overlooking the north bank of the Rogue River, but soon became lost in the "rough, rugged, heavy timbered, bushy terrain" (Williams n.d.:12; see also T'Vault n.d.:18). Eventually, however, the party encountered the trail to Powers prairie, which they described as a "plain Indian trail" running north-south (T'Vault 1851:1; Williams n.d.:18). After following this path for a few days, they "descended a long spur of the mountain to a narrow valley of prairie and timbered land, alternating through which a stream of seventy-five to eighty feet in width was flowing northwesterly" (Williams n.d.:1819). This was the South Fork Coquille River, and the Americans likely encountered the trail on the Iron Mountain ridge line, and descended that long spur that approaches Powers prairie and meets the South Fork at the mouth of Salmon Creek, immediately north of the Hayes site. Once on the South Fork, the T'Vault party arrived at one of the Nati jí dunne villages, where they observed: 18

An abundance of fresh Indian signs, and numerous hard beaten trails, and an old camp, a new fishery nearby indicated that this was a favorite resort for hunting and fishing purposes (Williams n.d.:19). They proceeded downstream, eventually portaging over to the Middle Fork Coquille River on an Indian trail, perhaps via Yellow Creek to the vicinity of the present-day community of Bridge. The T'Vault expedition failed to open a trail to the interior Rogue River valley. The gorge separating the lower reaches of the Rogue River from the interior valley was too rugged to traverse east-to-west. To travel from the mouth of Rogue River/Port Orford vicinity to the interior Rogue River valley, one would take trails that led through the Powers prairie area on the South Fork Coquille River. This was demonstrated by later American military and civilian expeditions. Later in 1851, following the failure of the T'Vault expedition, a detachment of U.S. Army engineers let by Lt. Williamson attempted to blaze a trail up the Rogue River (Beckham 1971:67). Like T'Vault, Williamson failed, but did map a trail that led up the coast from Port Orford and then east on a ridgeline located north of Flores Creek thence over the divide to the South Fork Coquille River (Figure 7). This expedition failed to reach to Oregon-California in 1851, but a second attempt led by Lt. Williamson and Lt. Stanton finally succeeded in the summer of 1852. According to their report and accompanying map (Figure 8; Stanton 1852), the most practical route to the upper Coquille River valley was a trail that traversed the ridge lines from the coast just north of Flores Creek (likely the same trail followed by Williamson the year before, a route that probably followed closely today's Langlois Mountain Road that crossed the divide into the Coquille Valley near Bennett Butte), and from there travels down Catching Creek to the "Forks"--the confluence of the Middle and South Fork Coquille River. From there, the expedition traveled up the South Fork Coquille River "through a succession of grassy slopes" (Stanton 1852:107) and established what they called a `depot camp' from which they could base themselves while exploring the Coquille River headwaters. This camp, by their own reckoning, was located on the Coquille River at 42° 57 03 north latitude, approximately the location of present-day Gaylord on the South Fork, a location near the portage via Yellow Creek to the Middle Fork Coquille River followed by T'Vault in 1851. Over the summer, Stanton and Williamson explored and blazed various trails to find the most practical route to the Oregon-California trail to the east. In June, they traveled south from their camp, and after "crossing the river twice to arrive at high bluffs, we reached a small prairie, situated like the others we had passed through, on the slopes of the mountain and here we encamped" (Stanton 1852:107). This was likely Powers prairie, as the following day, when the expedition attempted to ascend the river, they were:

19

Figure 7. Possibly the first detailed map of the interior of southwest Oregon, produced in 1851 by Lt. Williamson while I attached to Col. Casey and Major Kearny (National Archives RG 77-Central Map File: Civil Works Map File: Oregon W-12). The map reflects Williamson's attempt to succeed where T'Vault failed and map a route from Port Orford to the Oregon-California Trail. Note the accurate detail of the Coquille River relative to the Rogue River, a product of Casey's military expedition against the Coquille Indians that fall (Tveskov 2000).

Overtaken by a party of Indians who gave us to understanding by signs that we were not to be able to follow the course we were going on account of the high bluffs [i.e. directly up the South Fork Coquille River as it enters a deep and narrow canyon in the Siskiyou National Forest]. They were going to Rogue River . . . . I decided to accompany the Indians. Under their guidance, we crossed the Coquille again, and traveling south and southwest on a high ridge along a well defined trail, we encamped for the night at a spring . . . [In two days] we reached the Rogue River . . . . The last part of the descent to the river was through open tall timber, and at the foot of the slope an open grassy bottom about a mile along the river . . . . Here our Indians refused to go any further, as they had come for the purpose of catching salmon (Stanton 1852:107-108).

20

Figure 8. Stanton's and Williamson's map of the Coquille River based on their explorations during the summer of 1852. The trails on their map and the location of their "depot camp"--near present-day Warner between Dement and Catching Creek--are highlighted. This trail, shown on the expeditions map, is likely the same "plain Indian trail" followed in the opposite direction by the T'Vault expedition the previous year (see above), that (heading south) leaves Powers prairie at the mouth of Salmon Creek just north of the Hayes site, ascends a long mountain spur over Iron Mountain, and eventually arrives at Rogue River near present-day Illahee. There, in the ethnographically described territory of the Athapaskan-speaking Chasta Costa yetl-ein of the Tututni (Miller and Seaburg 1990), Stanton (1852:109) described an Indian village that "contain[ed] a good many huts, and having extensive fishing dams in its vicinity . . . [and was inhabited by] several hundred Indians." 21

This trail was not an ideal route to the interior Rogue River valley, as Stanton and Williamson were stymied by the precipitous bluffs on either side of the Rogue River heading up stream. Rather, this route appears to have been a commonly used trail linking the lower Rogue River (from Illahee and downstream) to the interior Coquille River valley. After returning to their camp down the South Fork, the army engineers finally managed to blaze a trail to the Oregon-California Trail in at least three places. The most northerly route started in Chochrela dunne territory at the confluence of the South Fork and Middle Fork Coquille Rivers and followed the ridge line north of the Middle Fork through Camas Valley to the Umpqua River near present-day Roseburg. This trail is described extensively by Coquille Thompson and it was clearly an important part of his own social identity and ancestral biography, as, in conversations with anthropologists such as Dorsey, Drucker, Elizabeth Jacobs, and Harrington, he consistently described how it linked his father's village at `The Forks' of the Middle and South Fork Coquille Rivers with the interior Umpqua Valley, and that this was the trail his paternal grandfather (Tcæmt`at`adæn) followed when moving from the Umpqua to the Coquille River (Harrington 1942[26]:185-188, 1942[26]:190; see also Tveskov 2000:64; Drucker 1937:279-280; E. Jacobs 1933-34[101]:52-58; Dorsey 1890:232). At least two other trails connected the South Fork Coquille River with the interior valleys. From ridgelines that rose from the prairies around Gaylord and Powers these trails passed eastward over Six Mile Ridge and Eden Ridge, with the northerly route intersecting the Oregon-California on Cow Creek and the southerly reaching the Rogue River valley in the Galice/Grave Creek area (Figure 9). Neither of these trails, however, ever saw significant use due to the high relief of the terrain. Even today, only minimally improved roads follow these routes. By 1853, Williamson's and Stanton's most northerly route became the established trail for travel between the coast and the interior. An expedition of 40 men led by Perry B. Marple departed Jacksonville in the interior Rogue River valley with the intention of establishing a new settlement on Coos Bay (Tveskov 2000:424). The "Coos Bay Company" traveled north along the Oregon-California trail into the interior Umpqua River Valley near present-day Roseburg, where they hired an Umpqua woman to serve as a guide. They then followed the trail west, traveling via Lookinglass Creek and Ten Mile Creek to pass over the divide into Camas Valley, eventually arriving at the Chochrela dunne villages at `The Forks,' where they were greeted by Coquille Thompson's father, Chief Washington (Harris n.d.:4-5). This route, which generally corresponds to today's Highway 42, remains the only improved road between the interior and the coast in southwest Oregon (Figure 10). The wealth and status of the Nati jí dunne and Chochrela dunne can thus be considered not only due to their control over particularly valuable fisheries and other resources, but also by virtue of their location astride a major trail system that linked together the pre-contact Athapaskan cultural, political, and economic landscape. Trails linked these yetl-ein east and northeast to the interior Umpqua River valley--the home of the Athapaskan-speaking T`ln dunne--south to the Mikono dunne, Chasta Costa, Galice, and Applegate Athapaskan yetl-ein of the middle and lower Rogue River, and west over Flores Creek and the Sixes River to the Kwatami, Yukiche dunne, Tututni, Chetleshin dunne, Khwaishtunne dunne, and Chetco that are the yetl-ein of the Tututni of the southern Oregon Coast. 22

Figure 9. Detail of 1853 map of northern California and southern Oregon published by Thomas Tennant of San Francisco (National Archives RG 77, Central Map file: Civil Works map file: Oregon, W-606). The trail north from Port Orford follows Flores Creek over the South Fork Coquille River and on to the Oregon-California Trail is shown. Note that "Table Rock" is inaccurately shown as located on the upper Bear Creek.

23

Figure 10. Map of southwest Oregon made by Lt. Kautz, stationed at Fort Orford, in 1856 (National Archives II: RG 77: Central Map File: Civil Works Map File U.S. 324-38). This map shows the trail route over Flores Creek via the South Fork Coquille River to the Oregon-California Trail on Cow Creek as well as the trail running north of the Middle Fork Coquille River via Camas Valley to the interior Umpqua Valley--today's Highway 42. Despite their numbers and apparent wealth, the Nati jí dunne were apparently devastated by the immigration of American settlers into the region. Their specific fate is unknown; unlike their relatives, the Nati jí dunne are almost never described in written accounts of the turbulent years of the 1850s that saw the near genocide and eventual forced removal of Indian people from southwest Oregon (Tveskov 2000). One possibility is that their villages were overcome by introduced diseases such as small pox, measles, and malaria the swept the region in the 1830s, shortly after McLeod's visit. Or perhaps, as was common in that era, they were attacked by the Americans in an as-yet undocumented massacre. The oral traditions and ethnohistoric data lends some circumstantial support for this scenario: Coquille Thompson recalled family stories that describe the Nati jí dunne as a thriving community in his grandfather's and father's time, and in 1826, McLeod observed people living in permanent plank house villages in the Powers area. However, T'Vault, Williamson, and Stanton, writing in the early 1850s, describe Indian people visiting the area to fish, but did not specifically mention permanent villages (see above). In either case, survivors could have incorporated into the villages of their relatives down river. By the time Powers prairie was settled by Whites in the 1870s, only a few Indians were living in or using the area (Peterson and Powers 1952:126).

24

Although these first settlers encountered few Indians, they did find the South Fork Coquille River already inhabited by Chinese (Figure 11 and Figure 12). The fate of these immigrants, as that of the Nati jí dunne, remains undocumented, but clearly relates to the arrival of the Americans. According to Mr. and Mrs. Max Dement, two early settlers of the region:

Figure 11. Chinese bronze vessel recovered from the gravel floor of the South Fork Coquille River during archaeological survey near Broadbent (Tveskov 2001).

Figure 12. Bottom of the Chinese bronze vessel.

25

There were Chinese mining here when the settlers came. Those miners bought their supplies from the ranchers. After they had made a purchase, they would go shuffling off to their mines. They were panning out `real paydirt.' Most of the Indians had left the country before the Chinese came. The few Indians that were left were very peaceful and quiet. The settlers were more afraid of the Chinese than they were of the Indians. Finally, they decided they could get along very well without the Chinese who were very superstitious, so it was an easy task for the settlers to scare them out. The settlers could not get along nearly so well with the Chinamen gone as they did while they were here. There seems to be a difference of opinion as to when this happened, some say 1884, others say 1887 (quoted in Peterson and Powers 1952:126-127). With the removal of both the Indians and the Chinese, the South Fork Coquille River was inhabited primarily by a small number of White ranchers. Their homesteads, spread across Powers and Rowland prairies, were collectively known as the "North Carolina Settlement" after the state of origin of some of the first settlers (Douthit 1999:100; Peterson and Powers 1952:126; Wooldridge 1971:40). By the 1890s, with demand for timber growing back east, timber barons such as Albert Powers and C.A. Smith began to amass--often by clandestine and illegal means--large amounts of property across the South Fork watershed and elsewhere in Coos and Curry Counties (Robbins 1988). By the turn of the century the old growth timber along the lower reaches of the river began to give out, and engineers from the Smith-Powers Logging Company surveyed and built a logging railroad up the South Fork Coquille River. By the 1920s, Powers was an established `company' town of an innovative sort, as described by Douthit (1999:102; see also Robbins 1988:45-46). While the tracks were being laid, workers and family members lived in tents and railroad shacks. But afterwards, families moved into the new town of Powers. The company had laid out lots, sewer, and water lines. Rather than run a company town, lots were sold to workers and businesses. The town planning commission insisted that houses in the town had to have double walls and a toilet. Building by workers, businesses, and the company continued through 1914 and 1915, the year that the post office's name officially was changed from Rural to Powers. By the 1920s, Powers was "the center of one of the largest logging operations in the United States," and for a brief time mimicked the wealth and power of the Nati jí dunne villages that preceded it (Robbins 1988:46). With subsidiary camps at China Flat and Eden Valley and elsewhere linked by railroad spurs, four trains each day transported the logs to the "Big Mill" on Coos Bay owned by C.A. Smith, where they were milled for export to San Francisco and eastern markets. Eventually, however, even the enormous stands of old growth timber on Eden Ridge, the South Fork/Rogue River divide and elsewhere in the vicinity were exhausted, and the town began a long slow decline. As described by Robbins (1988:46):

26

Although isolated from the major routes of travel, Powers has been the home for hundreds of loggers and trainmen over the years. It has long since lost its cast as a company town, but the influences of its years as a busy center of logging activity weighs heavily. It suffers, like many rural communities in America, from the forces of modern capitalism that have concentrated wealth in the larger urban centers. From a bustling center of lumbering activity in the second decade of this century, Powers has become the home of retired people and a few logging families who survive on uncertain employment. The trains that traveled regularly between Myrtle Point and Powers stopped years ago; today the log trucks that rumble down the South Fork of the Coquille pass by the small town and its graying buildings on their way to the mills on the lower Coquille or on Coos Bay.

Archaeology

Prior to the archaeological survey conducted by Southern Oregon University and the Coquille Indian Tribe in the summers of 2000 and 2001, virtually nothing was known about archaeological sites in the South Fork Coquille River area (Tveskov 2001). Our work on the South Fork was part of a larger effort to catalog cultural resources throughout the Coquille River watershed through a grant to the Coquille Indian Tribe Cultural Resources Program from the National Park Service (Byram and Ivy 2001; Tveskov 2001). The goals of our survey were to synthesize what was known, both professionally and locally, about the archaeological record on the South Fork watershed, establish productive ties with the local community, and familiarize ourselves with the South Fork environment. The project included a search of the Oregon State Historic Preservation Office archaeological site files to identify previously recorded sites in the South Fork Coquille River watershed, networking and conducting informant interviews with long-standing local residents with knowledge of cultural sites and local history, and revisiting previously recorded sites and sites that we learned about from informants to confirm their existence and assess their current condition. We also conducted a pedestrian archaeological survey of publicly owned lands and selected privately owned lands along the South Fork Coquille River. A total of eighteen archaeological sites were confirmed to exist or are reliably reported by local informants along the South Fork Coquille River. This survey provided a glimpse into the considerable archaeological potential of the South Fork Coquille River area, particularly on the prairies and foothills on either side of the river north of where the South Fork exits the steeply incised gorge in the Siskiyou National Forest. The testimony of local residents and the results of our survey suggest that these prairies were indeed heavily occupied. One person described Rowland and Powers prairies as each being "one big archaeological site," and in fact all but two of the surveyed public land parcels north of the Siskiyou National Forest boundary contained archaeological sites, and 15 of the 18 identified sites lie on these prairies. There is clearly a heavy archaeological imprint on these alluvial floodplains and the surrounding terraces. On the prairies, sites are most often found on the primary alluvial plain, less frequently on the secondary terrace, and sometimes spread across both. This is generally consistent with the pattern of sites described by Connolly (1986:196) based on his pedestrian

27

survey of Camas Valley on the Middle Fork Coquille River. Impressionistically, it appears that along the large prairies, the largest sites are located along the leading edge of the secondary terraces, particularly where such terraces approach the South Fork. At such locations, as well as in the gorges that separate the prairies, the secondary terraces and primary floodplain often blend together in a slope towards the river, and cultural material is found across both land forms. Sites located exclusively on the alluvium appear to be smaller, with less rich archaeological assemblages. Our survey was biased against upland sites, as little survey work was conducted in the mountains that rim the South Fork watershed. On the other hand, we were able to examine many terraces and small floodplains along the South Fork within the steep gorge south of the National Forest boundary, and found no sites at all. This absence is possibly a product of site visibility, as cultural material may be buried deep in the alluvium that characterized the surveyed landforms. Still, this could well be a real pattern, as ethnohistoric sources suggest that the main fisheries were further downstream on the prairies, and that Indian people, rather than travel up this steeply incised gorge, followed trails over high elevation ridge lines to the east and west. Another outcome of our survey was the recognition of the intense interest in the ancient cultures and history of the South Fork Coquille River held by local residents as well as their sense of ownership and stewardship of that history. In particular, Jackie Hofsess, a long time Powers resident, shared her extensive knowledge with us, including the location of several significant sites, including the Hayes site. Jackie's extensive collection of artifacts was recently acquired by the Coquille Indian Tribe, and includes thousands of specimens, many of which are provenienced to (now) recorded archaeological sites. While a full description and analysis of this collection remains to be undertaken, a preliminary look provides some context to our survey work and our excavations at the Hayes site. The collection, originating mostly from sites along the South Fork Coquille River and upper Sixes River watersheds, contains numerous exotic pieces, including flanged mauls, steatite and basalt bowls, stone grooved net sinkers, large obsidian blades, and schist adze blades commonly attributed to the late Holocene Gunther Pattern archaeological culture (Figure 13, Figure 14, Figure 15, Figure 16, Figure 17, & Figure 18). Curiously, among the hundreds of chipped stone projectile points in the collection, only a handful are the triangular barbed "Gunther Barbed" or triangular concave-based types commonly found in late Holocene contexts throughout southwest Oregon and northern California. The majority of the projectile points are lanceolate, foliate, or side-notched forms commonly associated with the middle and late Holocene Glade Tradition. Many artifacts in this assemblage were collected from the Hayes site over the years, including several projectile points as well as a fragment of a carved antler smoking pipe and a groundstone basalt adze handle with an elaborate finely-formed hooked shaped proximal end. Similarly elaborated groundstone adze handles, pestles, and clubs (sometimes referred to as "slave killers" by collectors, see Stewart 1996:72) are often found in late Holocene village sites in the region, and at least one other hook-shaped example originating in the upper Coquille River watershed have been recovered from private artifact collections by the Coquille Indian Tribe.

28

Figure 13. Groundstone adze handle recovered from the Hayes site by Jackie Hofsess (Hofsess Collection, Coquille Indian Tribe).

Figure 14. Antler pipe stem recovered from the Hayes site by Jackie Hofsess (Hofsess Collection, Coquille Indian Tribe).

29

Figure 15. Basalt hammer collected from the South Fork Coquille River region (Hofsess Collection, Coquille Indian Tribe).

Figure 16. Obsidian blade collected from the South Fork Coquille River region (Hofsess Collection, Coquille Indian Tribe).

Figure 17. Basalt maul collected from the South Fork Coquille River region (Hofsess Collection, Coquille Indian Tribe).

30

Figure 18. Obsidian knife collected from the South Fork Coquille River region (Hofsess Collection, Coquille Indian Tribe). The information collected during our survey work on the South Fork Coquille River reconciled with analyses of oral traditions and ethnohistory as well as with models of late prehistoric settlement patterns generated elsewhere in the region (e.g. Connolly 1986; Winthrop 1993). These studies suggest that the largest sites--i.e. permanent plank house villages--were located on secondary terraces such that they were above seasonal flooding while still providing the most convenient access to a variety of habitats (Figure 19). The edge of secondary terraces, particularly where they approach the river on a mountain spur or at the north or south ends of the prairies where the surrounding hillsides come together to form a gorge ideal for salmon or eel fishing, are ecotones or boundary areas from which the bounty of the river (i.e. salmon, eels), the floodplain (camas and other wetland plants), the surrounding oak prairies (acorns) would be easily accessible. Such a location would also be the terminus of trails leading down mountain spurs from adjacent valleys or upland gathering or hunting camps. Ease of access to the South Fork itself may have been a criterion for site location, given that even the primary flood plain often raises steeply 20 meters above the water, and in fact many of the sites we examined (including the Hayes site) were located along slopes leading down to the river that in several cases contained trails used by modernday fishermen. Primary ethnographic data and ethnohistoric accounts suggest that the South Fork was a central place in the traditional landscape of the Pacific Coast Athapaskans and their neighbors rather than the peripheral area it has become over the last 150 years. The wealth of the South Fork fisheries as well as the location of the South Fork prairies astride paths linking the coast and lower Rogue River with the interior Rogue River and Umpqua River valleys provided a basis for the wealth and prosperity attested to in oral traditions (Figure 20). Hosting seasonal visits by near and distant friends and relations would have been key to establishing, maintaining, and legitimizing the status of the Nati jí dunne and Chochrela dunne. Sites on the floodplain would therefore likely be seasonal camp used on a temporary basis by visitors from elsewhere or by the inhabitants of permanent villages themselves.

31

Figure 19. Idealized model of site locations on the South Fork Coquille River based on southwest Oregon Native American oral traditions and archaeological survey data (Tveskov 2001).

32

Figure 20. The South Fork Coqulle River watershed, view south. The trail descending the mountain spur on the upper right of the map is the trail over Iron Mountain from the Illahee area on the Rogue River that terminates at the Hayes site.

Archaeological Site Looting

Despite the enthusiasm shown by many local residents for our work, and their wonderful contributions to our understanding of the archaeological record of the region, it was also the case that many of the sites we visited during our survey either had been or continued to be severely looted by illegal digging. Pot hunting on the South Fork Coquille River watershed is severe and ongoing, even by southwest Oregon standards (see Tveskov et al. 2002; Tveskov 2001). At Orchard Park, a city-owned public camp ground located 2 km upstream from the Hayes Site, an extensive and significant archaeological site is being actively looted, apparently by campers who can conveniently dig within the perimeter of their own camp site. The Hayes site itself has been and continues to be severely looted, as described in more detail in Chapter 3 (Figure 21). The issue of archaeological site looting in southwest Oregon in general and in the Powers area in particular is a complex one. Why is looting so common and widespread? Since the initial settlement of southwest Oregon by Euro-Americans, Whites have

33

commonly looked for and even excavated significant cultural places. Up until the last 20 years or so, however, pot hunting seems to have been a recreational, even family-oriented activity undertaken by individuals who felt a strong connection to and curiosity about their landscape and history. In recent years, owners of many of these often extensive collections appear to be interested in their ongoing legacy, resulting in the transfer of several significant collections to, for example, the Coquille Indian Tribe.

Figure 21. Artifact collector demonstrating digging technique at the Hayes site, summer 2002. Unfortunately, however, pot hunting on a dramatic scale continues, and discussions with local residents (including some active pot hunters) suggest that causes are many and complex. Communities like Powers are isolated from the larger towns of southwest Oregon, and individuals living there are intensely independent and feel a sense of proprietary ownership and stewardship of the local history and culture. Some residents were actively hostile to our project, fearing that the disclosure of information about archaeological sites-- particularly to Indian tribes and university archaeologists--would adversely impact their rights as property owners. One local collector--who actively digs to this day--explained that he had a right to dig based on his spiritual connection to the Indian people who manufactured the artifacts he was excavating, explaining that he often conversed with their spirits while he "dug at night with a headlamp." The common abuse and pernicious effects of methamphetamines in rural areas of southwest Oregon contribute to the problem; unlike the family hobbyists of earlier generations, addicts frequently excavate archaeological sites primarily for profit, selling artifacts locally or on the internet. One visibly intoxicated individual arrived at the Hayes site one morning during the dig, and approached me saying "I will give you $200 for each arrowhead ya got." Several significant sites in southwest Oregon (such as the Windom Site, see Tveskov et al. 2002) have been brought to the attention of archaeologists as a result of plea bargains between individuals under arrest on drug charges and federal law enforcement authorities. The significance and value of artifacts to local people, national and international collectors of Native American antiquities, members of federally recognized Indian Tribes,

34

local people who identify themselves as Native American, and archaeologists are often different and contentious, and the methodological, social, and theoretical dimensions of pot hunting need to be further explored. In regards to the Powers area, clearly, many local residents feel an intense connection to their landscape, and their desire to find and collect artifacts, as well as their distrust and resentment of `outsiders' is understandable. On the other hand, the position taken by most professional archaeologists is that digging an archaeological site, whether for fun, curiosity, profit, or science, is inherently destructive, and should not be taken lightly. Regardless of how the value of an archaeological site and the artifacts it contains are measured, that value will be inherently tied to the context and integrity of that site and those artifacts. Given that archaeological sites along the South Fork of the Coquille River often contain the material legacy of perhaps thousands of years of human use, their destruction through widespread excavation is indeed appalling.

35

36

Chapter 3. Site Setting, Field Methods, and Site Structure

By Mark Tveskov and Nick Halousek

The Hayes site lies at an elevation of ~85 meters on the northern face of the Siskiyou Mountains, a portion of the Klamath Mountains that divide the Coquille River drainage from the Rogue River drainage. The site is located on an uplifted terrace overlooking the South Fork of the Coquille River, just south of a major confluence where Salmon Creek enters from the west and Mill Creek enters from the east (Figure 22). North of this confluence, the South Fork curves through Powers prairie, a three-kilometer long and one-kilometer wide alluvial flood plain. To the south, the South Fork Coquille River meanders through a two-kilometer long and one kilometer wide prairie that includes, on the northwestern side of the river (and immediately south of the Hayes site), the prairie on the south end of the airport and on the southeastern side of the river known locally as "Dingbat Flats." In the vicinity of the site, the South Fork Coquille River apparently runs along a fault line, as the airport prairie is uplifted some 30-40 meters and has a sheer bluff overlooking the river, while across the way, the Dingbat Flats prairie is very low, grading gently down to the river bank (Figure 23). The South Fork is constricted again at the south end of Dingbat Flats, where mountain spurs pinch off the prairies at Orchard Park. The mountains around Powers prairie, the airport, and Dingbat Flats are fringed with foothills, but rise precipitously--up to 840 meters high within three kilometers of the river--to the east and west. A long southwest-northeast-running mountain spur terminates immediately west of the Hayes site, and was likely the ridge line followed by a trail to the Illahee area on Rogue River used for generations by Indian people as well as by the American explorers such as T'Vault, Williamson, and Stanton in the 1850s (see Chapter 2 and Tveskov 2001).

Site Setting

The Hayes site is located on the property containing Powers Airport (Hayes Field), land administered by the Port of Coquille. The archaeological site lies on and around several stepped terraces around the northern end of the airport's single grass runway. At least five separate north-south running terraces (numbered, from east to west, terrace I, II, III, IV, & V), and these grade into a peninsula with a southwest-northeast axis that outlines a bend in the South Fork Coquille River (Figure 24, Figure 25, & Figure 26). The southeastern edge of this peninsula has a high (~40m) and actively eroding bluff face that overlooks Dingbat Flats that tumbles down almost vertically to the South Fork immediately below. The immediate site vicinity is comprised of sediments and clasts of alluvial origin; the southern and eastern edge of the peninsula are comprised of ~2 m of compacted very fine silty sand alluvium capping metamorphic bedrock. The western and northern edges of the peninsula grade gently down and provide access to the river, and this area is comprised of fine silty sand alluvium as well as bars of well-sorted river cobbles and gravels. The sediment at the site is poorly drained, and local residents report that water pools up on the site for long periods of time in the winter.

Figure 22. The South Fork Coquille River showing the location of major communities, parks, and the Hayes site.

38

The Hayes site lies at a transition between habitats. The flat terraces of the site itself and the foothills to the west contain the remnants of an ancient prairie that at one time likely contained camas, acorn-bearing oak trees, and other plant resources important to Indian people. Today, however, the site is covered with introduced grasses and is used as a horse pasture and airport runway. The South Fork Coquille River immediately adjacent to the site runs through a relatively shallow and rocky riffle that would have been (and is still today) an ideal and accessible fishery (see Figure 3), and the high foothills and mountain slopes to the west that today contain tightly managed stands of second or third growth Douglas fir (Pseudotsuga menziesii) would have been, in ancient times, more patchy, with diverse stands of fir, cedar (Thuja spp.), and hemlock (Tsuga spp.) trees interspersed with human-managed meadows rich in berry yielding shrubs (particularly Rubus spp.) and other traditionally important resources.

Figure 23. View south from the airport terrace over the South Fork Coquille River to Dingbat Flats. Note the elevation difference from the site terrace (where the photographer stands) down to the floodplain on the opposite bank. The Hayes site is part of a large complex of archaeological sites in the immediate vicinity of the Salmon Creek and Mill Creek confluence with the South Fork Coquille River (see Chapter 2 and Tveskov 2001). Lithic debris is spread across the leading edge of the secondary terrace throughout Dingbat Flats and into Orchard Park, and archaeological material is likewise present throughout Powers prairie, including within the city limits itself. An archaeological site lies on the terraces on either side of the mouth of Mill Creek directly across the South Fork of the Coquille River from the Hayes site. Local residents report

39

finding numerous artifacts throughout this area, and visitors to our excavations showed us or described projectile points, large crypto-crystalline silicate or obsidian bifaces (including an obsidian "ceremonial blade"), and groundstone artifacts that they had found. One person somewhat cryptically alluded that human remains had been found at one point in Dingbat Flats. Many artifacts were recovered by collectors from the Hayes site itself. Mrs. Hofsess reported finding many artifacts there over the years, including an elaborate groundstone adze handle and a carved antler pipe fragment (see Chapter 2).

Figure 24. View west overlooking Terrace IV to the foothills located west of the site.

Figure 25. View southwest across the airport runway (Terrace IV).

40

Figure 26. View north across Terrace III from the Feature 2 area. Note the leading edge of Terrace IV visible just beneath the fog line on the right side of the photograph.

Degradation of the Hayes Site

The Hayes site vicinity is an area of exceptional cultural resource values. The testimony from ethnohistory and Coquille Indian oral traditions provide a rich cultural and historical context, a context complemented by the artifacts found in local collections and by the results of our archaeological survey that confirmed a heavy archaeological imprint across the Powers prairie/airport/Dingbat Flats/Orchard Park area. Unfortunately, however, the Hayes Site, like many of the archaeological sites in the Powers area, has been and continues to be dramatically impacted and disturbed by illegal digging and other forces (Figure 27). The terraces around the north end of the runway extending into the peninsula are literally pockmarked with shovel scrapes and holes that attest to the frequency and intensity of recreational digging. Although found throughout the site, pot hunting is particularly evident along the southeastern edge of the peninsula, where several large scrapes, some over three meters in diameter and up to 40 cm deep, were observed, and also on the leading, eastern edge of the runway terrace (Terrace IV). In this area, located just south of the airport parking area, collectors excavate into what is apparently sediments pushed over this edge by a bulldozer when the runway was graded. The Hayes site was looted immediately before, during, and after our excavations. On our first visit to the site in the summer of 2001, several new holes were observed immediately east and below the runway. These were surrounded by a scatter of lithic debitage, as well as a grinding slab that onto which the letters "R.I.P." had been inscribed with a metal object (see Figure 2 in Chapter 1, and Chapter 7 for a more detailed description 41

of this artifact). Visitors to the site during our excavations described how a favored way to collect artifacts was to visit the site following a heavy rain when artifacts would be newly exposed, particularly around the roots of the Himalayan blackberry bushes (Rubus discolor, a non-native species) that dot the field. One individual, in apparent impatience with the slow pace of archaeological excavation, provided a demonstration of a favored digging technique: using a tool specifically manufactured for the task--a heavy knife blade set into a 1.5 m long aluminum pole--he turned the top soil over, breaking through the root zone to a depth of approximately 15-20 cm. Working back and forth in a semi-circular fashion, he was able to scrape open a swath some 4 m in diameter within 15 minutes (see Figure 19 in Chapter 2). In the fall of 2002, after the completion of our field work, the site was looted again. Several individuals, apparently from Charleston on Coos Bay, showed Jennifer Viksne, then an employee of the Coquille Indian Tribe, a number of projectile points and bifaces they had excavated from the "airport up at Powers" (Figure 28).

Figure 27. Looter's holes located on the leading (northeastern) edge of Terrace IV. This area is comprised of artifact-rich sediment bladed off the runway. In addition to illicit digging, agricultural activities, the grading of the runway, and rodent activity have also dramatically impacted the archaeological deposits at the Hayes Site. According to many local residents, many artifacts were found after the horse pasture on Terrace II and Terrace III and was disked to encourage the growth of grass. A disc is an apparatus pulled behind a tractor and is composed of two rows of round metal blades that cut into the ground approximately 20 cm deep to turn the sediment to mix and aerate soil and organic matter. Likewise, the north end of the runway on Terrace IV was graded with a bulldozer, creating a small berm of sediment in a spoil pile along the leading, eastern edge of Terrace IV. This is unfortunate, as apparently this activity destroyed one of the richest areas of the site. Local residents report that this area is the favored location for collecting, an

42

assertion evidenced by the numerous pot holes in this spot. The hook-shaped groundstone adze handle in Jackie Hofsess' collection was found here, and our shovel test pit and excavation units in these areas recovered many artifacts from disturbed sediments (see below).

Figure 28. Artifacts collected from the Hayes site following the completion of our excavation in 2002. Photographed by Jennifer Viksne, Coquille Indian Tribe.

Figure 29. Extensive krotovina and active rodent burrows in the Feature 1 excavation block.

43

Bioturbation from rodent activity also seems to have severely impacted the site (Figure 29). All of our excavations exposed sediments laced with active burrows and krotovina, and the cultural features at the site appear to be exploded and homogenized by the cumulative effect of rodent burrowing. The culprit is possibly the Beechy ground squirrel (Spermophilius beecheyi), a critter found along the Oregon coast that prefers to burrow in wet pastureland (Maser et al. 1984:152).

Sampling Strategies and Procedures

The Southern Oregon University excavations at the Hayes site took place during the SOU summer archaeological field school held between June 24 and July 19, 2002. For horizontal and vertical control, a datum was established on eastern edge of the peninsula on the bluff edge. The location of this datum was established with a GPS unit, and it subsequently served as an anchor point from which the site was mapped using a total station transit. Cultural and natural features of the site, evidence of past looting, and all the Southern Oregon University excavations were recorded by 143 mapping points (Figure 30). The boundaries and horizontal extent of the Hayes site were determined by pedestrian survey of the entire airport area and the excavation of eleven 50 x 50 cm2 shovel test pits (Table 1). Cultural material at the Hayes site was observed on all the terraces around the north end of the runway in an area running ~100 m north-south and 100m eastwest, covering an estimated area of 10,000 m2. Transects of shovel test pits, placed at ~2025 m intervals, were excavated running east-west across the horse pasture on Terrace III, north-south across a heavily looted area along the bluff edge on the eastern edge of the peninsula, and east west along the eastern edge of the runway on Terrace IV, across another heavily looted area. A single shovel test pit was excavated on a knoll (Terrace V) to the west of the runway, and yielded significant numbers of artifacts and a red crypto-crystalline silicate biface fragment. All sediment from the shovel test pits was screened through ¼" hardware cloth. Artifacts are continuously distributed across the area around the north end of the runway from the foothills to the river bank. However, based on surface observations, discussions with local residents, and the excavation of our shovel test pits, cultural material seems to be most heavily concentrated along the eroding bank along the eastern side of the peninsula, in the graded spoils along the eastern edge of the north end of the runway, and on the Terrace V knoll to the west. In addition to the shovel test pits, 12 individual 1 m x 1 m square units, each identified by a number, were excavated. Each one meter square unit was divided into four 50 cm square quadrants, designated A (northwest), B (northeast), C (southwest), and D (southeast). Each unit was excavated in arbitrary 10 cm levels reckoned below a datum and line level established for each. All significant features were mapped on standardized forms, as were the location of artifacts, charcoal samples, etc. found in situ. Each excavation unit was trowel shaved, and sediment from quads A, B, & C were screened over ¼-inch hardware cloth while sediment from quad D was screened over -inch hardware cloth. In all, the 12 excavation units and the eleven shovel test pits covered ~17.5 square meters, resulting in a less than 0.2% sample of the site. A total of 9.25 m3 of sediment were

44

excavated. Despite this relatively small sample, a large, if not particularly diverse assemblage of artifacts was recovered, including 5,615 pieces of lithic debitage, 81 pieces of burned animal bone, 100 individual chipped stone tools, 19 individual ground stone artifacts, and the identification of at least two cultural features.

Figure 30. Plan map of the Hayes site.

45

The excavation units were placed in three clusters. Six contiguous units, forming a 2 x 3 m excavation block oriented north-south, were placed around STP E on the eastern side of the horse pasture on Terrace III. These were placed to expose a burned rock feature (Feature 1) exposed during the excavation of STP E (see below). An additional excavation unit (Unit 5) was placed four meters south of this block. Four contiguous units (Units 2, 9, 10, and 11) were placed on Terrace III along the edge of the bluff on the eastern edge of the peninsula around another burned rock feature (Feature 2) exposed by the excavation of STP G, located immediately south of a particularly large looters hole that many residents reported yielding numerous artifacts. Finally, a single excavation unit (Unit 12) was placed in the disturbed but artifact-rich sediments along the edge of Terrace IV east of the north end of the runway (Figure 31 and Figure 32).

Figure 31. Excavation underway in the Feature 2 excavation block. View southwest.

46

Figure 32. Excavation in progress in the Feature 2 excavation block. View west looking over the Feature 1 block (next to the van). The airport runway (Terrace IV) is visible beyond the barn.

Site Stratigraphy

The excavations along the bluff edge and around Feature 1 yielded roughly consistent stratigraphic profiles (Figure 33). Three major stratigraphic units, all oriented more or less horizontally, were observed: beneath a very thin (usually less than 1 cm) duff/Ao layer of compacted silty sand and grass roots was Stratum I, a weakly developed Ahorizon some 10-30 cm thick, comprised of silt with less than 5% sub-angular coarse sands and significant quantities of burned river cobbles, rootlets, and extensive krotovina. Stratum II and Stratum III are similar in composition as Stratum I, save that the degree of soil development and amount of organic material diminishes significantly, and the overall color of the sediment gets lighter and redder. The contacts between these units are extremely diffuse, and the sediment profile is best considered a continuum of heavily compacted silts, with some soil development at the top. Given how uplifted the Hayes site terrace is above the river, the sediment at the site likely represents very ancient alluvium mixed with some wind blown and colluvial sands that has been heavily compacted by agricultural activity. Unit 12 was a 1 m x 1 m unit excavated into the sediments on the eastern, leading edge of the runway (Terrace IV) terrace. Consistent with local resident's testimony, this was a berm of sediment graded off the surface of the runway; this unit showed a very disturbed stratigraphic profile. Although the sedimentary matrix was similar to that revealed by the units on Terrace III, the stratigraphic units were jumbled and contained a fair amount of buried organic matter and burned and unburned wood. This unit contained a large number of artifacts, and likewise confirms local resident's assertions that the north end of the runway was once one of the richest areas of the site. These cultural deposits, however, have been almost completely destroyed by the grading of the runway.

47

Figure 33. Stratigraphic plan of Unit 11, Unit 9, Unit 2, and Unit 10 in the Feature 2 excavation block.

48

Vertical and Horizontal Patterning of Cultural Material

Cultural artifacts were abundant in all the excavation units at the Hayes site. The vertical distribution of lithic debitage (the most common artifact type) by arbitrary 10-cm level for all the excavation units, shown in Table 1 and Table 2, indicates that while the frequency of cultural artifacts generally diminishes towards the bottom of the excavation, there is no apparent vertical stratigraphy to the cultural deposits at the site; artifacts are found commonly found fairly evenly distributed within each unit. While this could be indicative of generally continuous use of the site since it was first occupied, it is also possible that any vertical patterning of artifacts has been destroyed by rodent activity. There is little overall difference in the density of artifacts across the surface of the site, as reckoned by the shovel test pits and the excavation units. One exception is along the eastern edge of Terrace V in the airport spoil pile; shovel test pits and the single 1 m x 1 m excavation unit (Unit 12) in this area had significantly higher densities of artifacts, confirming one of the richest areas of the site originally sat on what is now the north end of the airport runway. The overall density of artifacts at the Hayes site--~886 pieces of lithic debitage per cubic meter of excavated sediment--is not exceptionally high in comparison with other southwest Oregon sites (Tveskov et al. 2002: 123; Winthrop 1993:105-108).

Cultural Features

Two cultural features--both large clusters of burned river cobbles--were identified at the Hayes site and were designated Feature 1 and Feature2. Feature 1 was initially identified during the excavation of shovel test pit STP E on Terrace III. This test pit was subsequently expanded into a 2 x 3 m excavation unit, oriented north-south, comprised of Unit 1, Unit 3, Unit 4, Unit 7, and Unit 8 (Figure 34 and Figure 35). These units were excavated to a depth of 30 cm below the ground surface. Feature 1 is a cluster of over 150 burned river cobbles, ranging from five- to over 30-cm in diameter, spread across all six excavation units but concentrated mostly in an east-west running band across Unit 3 and Unit 4. No distinctive soil staining was noted, although there was a band of slightly lighter, more compacted silt in the center of the feature, and very small flecks of charcoal were noted throughout the excavation. The entire feature was not uncovered, as the cluster of rocks appears to continue to both the north and south into the side walls of the excavation. Lithic debitage was common across all of the excavation units in and around Feature 1, and several formal artifacts were recovered as well. Artifacts found in association with this feature included a pestle fragment (02.03-11), two bifaces (02.03-02 & 02.03-23), a projectile point (02.03-27) and a projectile point tip (02.03-04). A chunk of oxidized red mineralized sediment was found in association with the feature, possibly a piece of red ochre. Feature 1 likely represents a large oven for roasting food--particularly plants such as camas. Unfortunately however, the cumulative affect of rodent and agricultural activity as served to `explode' the feature over the years, dispersing the rocks and homogenizing the soil. Feature 2 was initially identified during the excavation of shovel test pit STP G near the eastern edge of the Terrace III (Figure 36, Figure 37, and Figure 38). This test pit was subsequently expanded into an "L"-shaped excavation comprised of four 1 m x 1 m

49

excavation units (including Unit 9, Unit 10, and Unit 11 as well as Unit 2 that subsumed STP G). Unit 2 was excavated completely through the feature while the remaining excavation units were excavated to expose the feature. Also similar to Feature 1, Feature 2 was left in situ and backfilled following our excavation. Like Feature 1, this feature is a cluster of several hundred burned river cobbles lying between 10 and 30 cm below the surface. The exact dimensions of the feature are unknown, as it appears to continue into the side walls of the excavation units in all directions.

Figure 34. Feature 1 excavation block, view south. Note the krotovina visible in the foreground.

50

Figure 35. Plan of the Feature 1 excavation block.

51

Figure 36. Feature 2 excavation block, view south. Note the krotovina in the side wall.

Figure 37. Detail of Feature 2, Unit 9, view south.

52

Figure 38. Feature 2 excavation block.

53

Table 1. Debitage frequency and density from 50 cm x 50 cm shovel test pits at the Hayes site.

Level 1 2 3 4 5 6 7 8 9 10 11 12 total

3

A 20 33 10 16 79

B 12 4 1 7 7 23 4 8 66

C 20 25 3 9 3 5 65

D 14 1 21 6 2 44

E 36 16 9 3 64

F 14 9 20 10 7 13 1 7 81

G 14 33 16 1 12 1 1 10 10 14 1 10 123

H 24 5 5 4 2 2 42 0.15 280

I 9 7 1 3 1 11 1 3 1 37 0.22 168

J 8 1 8 10 25 18 1 5 76 0.2 380

K 30 3 30 1 22 8 1 4 6 1 106 0.25 424

L 16 1 38 2 35 1 27 1 22 28 1 172 0.27 637

M 21 13 18 2 41 2 35 23 1 156 0.22 709

N 20 9 1 30 0.7 43

P 44 50 21 41 2 0 0.12 1317

Total 302 210 202 115 159 84 71 61 40 43 2 1289 3.35 388

m 0.15 0.2 0.15 0.12 0.1 0.2 0.3 nisp/ m3 527 330 433 367 640 405 410 nisp=number of individual specimens. m3=cubic meters excavated

Table 2. Debitage frequency and density from 1 m x 1 m excavation units at the Hayes site.

level 1 2 3 4 5 6 7 8 9 10 11 12 Other* total m3 nisp/ m3

Unit 1 165 94 30 37 326 0.4 815

Unit 2 153 154 104 84 64 58 42 28 51 21 20 13 13 805 1.2 671

Unit 3 151 119 41 25 336 0.4 840

Unit 4 107 80 29 216 0.3 720

Unit 5 22 110 99 96 68 50 1 446 0.6 743

Unit 6 132 109 53 70 364 0.4 910

Unit 7 113 62 52 42 269 0.4 923

Unit 8 8 163 60 52 283 0.4 708

Unit 9 154 78 72 56 53 34 447 0.6 745

Unit 10 318 70 388 0.2 1940

Unit 11 143 91 70 70 374 0.4 935

Unit 12 209 178 242 183 100 52 9 973 0.6 1622

Total 1675 1308 852 715 285 194 42 28 51 21 20 13 23 5227 5.9 886

nisp=number of individual specimens. m3=cubic meters excavated.

*includes flakes recovered from wall scraping/cleaning.

54

Lithic debitage was common across all of the excavation units in and around Feature 2, and several formal artifacts were recovered as well: Two projectile points (02.03-29 & 02.03-64), a projectile point fragment (02.03-35), two bifaces (02.03-28 & 02.03-37), and several unifaces were found in association, as was a hand-sized stone decorated with eleven parallel incisions (02.03-09, see Chapter 7). Like Feature 1, Feature 2 likely represents a large roasting oven that has likewise been exploded and homogenized by rodent and agricultural activity.

Summary

The cultural material at the Hayes site appears to be a nearly continuous spread of debris over several terraces overlooking a riffle on the South Fork Coquille River near the its confluence with Salmon Creek and Mill Creek. It is part of a large complex of archaeological sites found on virtually all of the terraces surrounding this confluence. The site itself appears to have some spatial structure, with varying concentrations of cultural debris found in different areas. Local collectors and our excavations on the leading edge of Terrace IV suggest that that the richest part of the site was originally the northern end of the runway, but this area has been almost completely obliterated by machine grading. Much cultural material from this area has been redeposited in a berm located on the northeastern, leading edge of Terrace IV, an area that is a popular spot for recreational digging to this day. It was from the runway area that Mrs. Hofsess found the hook-shaped adze handle and the antler pipe stem. The shovel test pits excavations indicate that there is a patchy and perhaps less rich and diverse distribution of cultural material found on Terrace III. Two clusters of artifacts explored with 1 m x 1 m excavation units suggest that these concentrations were focused around large oven features of burned river cobbles. This area, like the Terrace IV area, has been heavily disturbed--this time not only by looting, but also by rodent activity and the plowing of the area for agricultural purposes.

55

56

Chapter 4. Chipped Stone Tools

By Nick Halousek and Mark Tveskov

An assemblage of 100 chipped stone tools was recovered from the Southern Oregon University excavations at the Hayes site. Most of the assemblage was manufactured from locally available crypto-crystalline silicate, and included 28 projectile point or projectile point fragments, nineteen bifaces or biface fragments, and 53 formally manufactured or expedient unifacial chipped stone tools.

Bifaces

Bifaces are chipped stone tools that were bifacially modified (flaked on both the dorsal and ventral side of the original flake or core). As classified here, bifaces include projectile points, formed bifaces, performs, and blanks.

Projectile Points

Projectile points are small chipped stone bifaces with finished or nearly finished bases suitable for hafting onto a knife handle, arrow, dart, or harpoon shaft. A total of 28 complete or fragmentary projectile points were recovered from the Hayes site (Figure 39 and Table 3). Thirteen of these of these were classified into four morphological types: Small Foliate Points (n=2), Large Foliate Points (n=4), Foliate Side-Notched Points (n=4), and Broad Side-Notched Points (n=3). The majority of all the projectile points and projectile point fragments (n=21) were finely pressure flaked from preforms of locally available crypto-crystalline silicate, but five specimens manufactured from obsidian imported from over the Cascade Range were also recovered. Two of the projectile points from the Hayes site were classified as Small Foliate Points. These are small (less than 24 mm long), pumpkin seed-shaped points with no shoulders or side notches. One of the specimens (SOU-02.03-107) is manufactured from obsidian and the other (SOU-02.03-55) from crypto-crystalline silicate. Five projectile points were classified as Large Foliate points. They differ from the Small Foliate Point type not only based on their greater length, but also from their overall slender or lanceolate shape. All of the examples from the Hayes site are serrated, and three are slightly shouldered, resulting in a relatively long and pointy or lobate base. All of the Large Foliate Points from the Hayes site are manufactured from crypto-crystalline silicate. Although having a similar lanceolate shape, the four specimens from the Hayes site classified as Foliate Side-Notched points differ from the Large Foliate points in that they are notched along their sides for hafting and they have a straight or convex, rather than pointed

or lobate, base. Like the Large Foliate Points, the Foliate Side-Notched points from the Hayes site all have serrated edges and all are manufactured from crypto-crystalline silicate. Three projectile points from the Hayes site were classified as Broad Side-Notched points. Two of these (SOU-02.03-6 and SOU-02.03-29) were manufactured from obsidian and one (SOU-02.03-106) was manufactured from white crypto-crystalline silicate. Broad Side-Notched points differ from Foliate Side Notched points in that their necks and bases in general are wider, and their blades are shorter. However, all three of these specimens likely were much longer at one point, and each appears to have been heavily resharpened. Finally, there were 15 specimens from the Hayes site that likely represent broken fragments of projectile points. Seven of these were point tips, four were base fragments, and two were midshaft fragments. Eleven of these were manufactured from crypto-crystalline silicate, and the remaining two from obsidian. Table 3. Metric Data for the Hayes site chipped stone projectile point assemblage.

Small Foliate Points (n=2): specimen material fragment length width thickness 02.03-55 CCS 1.15 20.2 5.8 02.03-107 obsidian 23.2 15.6 6.5 Large Foliate Points (n=4): specimen material fragment length width thickness 02.03-33 ccs broken base n/a 15.0 9.2 02.03-65 ccs missing tip n/a 22.9 8.3 02.03-93 ccs broken base & tip n/a 17.7 6.2 02.03-90 ccs 43.9 13.8 6.9 Foliate Side Notched (n=4): specimen material fragment length Width thickness neck width 02.03-28 ccs 49.1 16.3 8.7 n/a 02.03-44 ccs split laterally 26.3 n/a n/a n/a 02.03-64 ccs broken base 34.2 15.5 7.8 10.9 02.03-73 ccs 34.8 14.0 7.0 10.2 Broad Side Notched (n=3): specimen material fragment length Width thickness neck width 02.03-6 obsidian 25.1 18.9 6.0 16.5 02.03-29 obsidian 27.6 18.4 7.4 13.4 02.03-106 ccs 25.0 16.7 5.5 9.6 Projectile Point Fragments (N=15): CCS obsidian total points and barbs 7 0 7 midshaft fragments 1 1 2 base fragments 5 1 4 total 13 2 15 Notes: all measurements in mm. CCS=crypto-crystalline silicate comment unfinished on ventral side comment slight shoulder serrated, slight shoulder serrated, slight shoulder serrated comment serrated serrated serrated serrated comment serrated, heavily resharpened heavily resharpened heavily resharpened

58

Figure 39. Projectile Points from the Hayes site (actual size). Top row: small foliate points: 02.03-55, 02.03-107; Second row: broad side notched points: 2.03-29, 02.03-106, 02.03-06; Third row: foliate side notched points: 02.03-73, 02.03-64, 02.03-28, 02.03-44; Fourth Row: large foliate points: 02.03-93, 02.03-33, 02.03-65, 02.03-90.

59

Formed Bifaces

An assemblage of 19 lenticular-shaped bifaces or biface fragments were recovered from the Hayes site (Figure 40; Table 4). All but one of these represent relatively small (less than 7 cm long) crypto-crystalline chipped stone tools or preforms rather than larger bifacial cores, and all save one are broken and heavily worked. Most have relatively sharp edge angles suggesting their use as knives or scrapers on relatively soft material. One specimen (SOU-02.03-32) is a mid-shaft fragment of a pumpkin seed-shaped biface manufactured from burgundy red crypto-crystalline silicate that was recovered from Unit 12 in the bulldozer spoils at the edge of the runway, and has the distal end of a channel flake scar running longitudinally on one side. Although this flake scar likely represents an attempt by the maker to thin the biface, it is a possibility that this tool is a large spear point. If so, the flake scar is possibly the result of the flake being removed during the impact of the projectile against something hard. There is also a single large bifacial chopper manufactured from dark gray or black crypto-crystalline silicate (SOU-02.03-94). This tool, while bifacially worked, is square in plan and exhibits use wear around its edges, suggesting its use on relatively heavy material such as wood. Table 4. Metric attributes of bifaces from the Hayes site.

specimen length width thickness comment 02.03-2 n/a 47.3 19.6 02.03-13 n/a 29.2 12.8 02.03-23 n/a 30.8 10.4 02.03-30 n/a 19.7 7.2 02.03-32 n/a 38.1 11.7 02.03-37 n/a 21.5 11.6 slender blade; possible knife or spear point 02.03-70 43.5 21.8 7.8 possibly a drill or projectile point preform 02.03-79 n/a n/a 8.8 02.03-80 n/a 23.4 5.2 finely finished tip 02.03-83 n/a 31.3 14.7 02.03-84 n/a 36.5 12.6 02.03-89 n/a 22.5 4.7 finely finished tip or base 02.03-94 93.6 71.5 36.2 large bifacial chopper 02.03-95 58.3 34.5 13.3 02.03-97 n/a 28.2 9.5 slender blade; possible knife or spear point 02.03-102 n/a 26.8 7.2 02.03-103 n/a 22.0 7.2 02.03-104 n/a n/a 7.7 02.03-108 n/a n/a 7.0 02.03-109 n/a 38.0 15.4 note: all bifaces are manufactured from crypto-crystalline silicate. Measurements in mm.

60

Figure 40. Bifaces from the Hayes site (actual size). Top row: 02.03-02, 02.03-13, 02.03-23. Middle Row: 02.03-37, 02.03-32, 02.03-83. Bottom Row: 02.03-95, 02.03-97, 02.03-102.

61

Unifaces

A total of 53 unifaces--tools manufactured from a flake chipped on a single side-- were recovered from the Hayes site (Table 5). These were classified into three groups: endscrapers, retouched flakes, and utilized flakes.

End Scrapers

End scrapers are manufactured from a small but relatively thick flake, with the working end--the thicker margin of the flake--pressure flaked by using the ventral surface as a platform. The end product is a small, often steep-angled scraping edge ideal for controlled or detailed "scraping, shaving, or adzing: of tough material such as wood, bone or hide (Connolly 1991:173). Such scrapers were sometimes hafted onto a wood handle ethnographically, and are commonly found in archaeological sites throughout southwest Oregon. A total of 13 such endscrapers were found at the Hayes site, and save for a single obsidian example, all were manufactured from crypto-crystalline silicate. Connolly (1991) further distinguishes three different kinds of endscrapers for southwest Oregon, based on his excavations at the Standley site in Camas Valley on the Middle Fork of the Coquille River. He describes Type 1 endscrapers that are manufactured from thick flakes that exhibit a bit thickness >7.5 mm, Type 2 endscrapers that have bit thickness <7.5 mm and lateral edges that are proximally tapered, resulting in a tear dropshaped tool ideal for hafting; and Type 3 endscrapers that have a bit thickness <7.5 mm and have their lateral edges unmodified. The Hayes site assemblage contained seven Type 1 examples, three Type 2 examples, and three Type 3 examples.

Table 5. Metric attributes of uniface endscrapers from the Hayes site.

endscrapers (n=13): specimen length width 02.03-1 19.6 25.5 02.03-16 32.3 32.8 02.03-36 24.2 20.2 02.03-57 26.6 19.9 02.03-76 46.4 32.6 02.03-81 29.9 24.0 02.03-87 29.5 25.3 02.03-21 27.5 29.0 02.03-82 31.2 28.5 02.03-38 28.8 18.3 02.03-19 36.5 18.6 02.03-74 23.2 15.2 02.03-115 37.3 18.4 note: Measurements in mm.

Thickness 10.6 12.2 8.2 11.6 17.2 9.2 11.5 6.6 6.9 5.5 10.0 5.3 5.6

comment type 1 type 1 type 1 type 1 type 1 type 1 type 1 type 2 type 2 type 2 type 3 type 3 obsidian type 3

62

Retouched and Utilized Flakes

Retouched flakes (n=20) and utilized flakes (n=20) were the most common chipped stone tool recovered from the Hayes site (Table 6). Both utilized flakes and retouched flakes are pieces of lithic debitage that do not have a definable form but display evidence of modification such as obvious retouch, edge damage or abrading possibly caused by the flake's use as knife or scraper. Retouched flakes differ from utilized flakes in that they exhibit at least one edge that appears to have been deliberately flaked to improve or sharpen an edge. Both retouched and utilized flakes, however, likely represent the relatively expedient use of a piece of lithic debitage. Table 6. Metric attributes of retouched flakes and utilized flakes from the Hayes site.

retouched flakes (n=20): specimen length width 02.03-5 18.2 27.5 02.03-10 22.4 23.8 02.03-15 n/a 28.5 02.03-24 45.0 26.5 02.03-25 34.9 16.5 02.03-35 n/a 19.1 02.03-45 n/a 19.8 02.03-49 n/a 15.3 02.03-51 n/a 10.3 02.03-54 n/a 14.0 02.03-59 21.2 19.7 02.03-60 n/a n/a 02.03-69 61.6 38.2 02.03-98 26.2 15.2 02.03-100 n/a n/a 02.03-105 39.3 28.4 02.03-110 56.1 38.4 02.03-117 n/a n/a 02.03-118 24.7 10.8 02.03-124 n/a n/a utilized flakes (n=20): specimen length width 02.03-3 31.3 26.5 02.03-12 n/a 32.2 02.03-14 33.9 20.8 02.03-48 34.2 22.5 02.03-52 36.0 22.4 02.03-53 n/a n/a 02.03-54 n/a 14.0 02.03-58 n/a 9.5 02.03-61 n/a n/a 02.03-66 41.8 16.8 02.03-88 22.5 16.8 02.03-101 n/a n/a 02.03-116 42.1 17.6 02.03-119 n/a n/a 02.03-120 14.4 25.4 02.03-121 36.8 25.5 02.03-122 40.3 19.1 02.03-123 23.5 6.3 02.03-125 n/a n/a 02.03-129 31.8 14.9 note: Measurements in mm.

thickness 5.1 6.4 15.0 9.8 6.1 6.6 5.6 4.3 2.9 4.0 3.4 4.6 17.5 6.2 5.4 9.3 6.2 4.3 3.7 n/a thickness 11.8 6.9 6.5 11.0 5.1 5.5 4.0 5.2 5.2 12.8 3.8 3.6 8.4 1.2 4.8 5.8 5.3 5.1 n/a 5.5

comment

thin, beautifully made expedient end scraper

comment

obsidian

63

Summary

The chipped stone tool assemblage from the Hayes site, while relatively large given the small excavated sample, is indicative of a fairly limited range of activities. The assemblage is notable for the complete absence of leaf-shaped bifacial cores, suggesting that chipped stone artifacts were, for the most part, transported to the site in finished or nearly finished forms. The absence of quarrying or initial lithic reduction activity is confirmed by the large number of heavily worked or broken tools--once at the site, tools were heavily used, resharpened, and eventually discarded there. The majority of tools are manufactured from crypto-crystalline silicate that would have been available locally, either in the Coquille River watershed itself or from elsewhere in Coos or Curry counties. A few chipped stone artifacts were manufactured from obsidian that would have been traded from across the Cascade Range from eastern Oregon or northern California. The Hayes site chipped stone assemblage is dominated by expediently used or minimally retouched flakes that appear to have been used as knives or scrapers. Other common artifacts include small bifaces, also used as knives and scrapers, unifacial endscrapers, and projectile points. With the possible exception of the endscrapers, which could have been used to smooth or work wooden implements, there is a noticeable absence of heavy wood working tools such as large choppers, wedges, or adzes. The hunting and processing of big game animals is the primary activity to be inferred from this assemblage. The projectile point assemblage is notable for the absence of Gunther Barbed, Triangular Concave Based, and other forms usually found in southwest Oregon archaeological sites that are associated with the Gunther Pattern, a late Holocene cultural tradition (see Chapter 2). Rather, the site is characterized by foliate and side-notched forms typically assigned to the Glade Tradition, a cultural tradition found in the late Holocene but that also dates as far back as the early Holocene. The absence of Gunther Pattern projectile points appears to be a real pattern rather than a product of our sample size; local residents knowledgeable about projectile points avow that few Gunther Barbed points are found at the airport or elsewhere along the South Fork Coquille River in general. Furthermore, examination of projectile point assemblages belonging to local collectors show that while such points are occasionally found, they are indeed rare in contrast to the foliate and sidenotched forms. The dating of the Hayes site, like for many sites in southwest Oregon, remains unclear. The presence of Gunther Pattern artifacts, such as the antler pipe and groundstone adze handle found at the site, attests to a late Holocene occupation, at least on the Terrace IV runway area, but neither local collectors or our excavations in the spoils from the runway and on Terrace III, recovered Gunther Pattern projectile points. It seems possible, therefore, that the Glade Tradition tools recovered during our excavations were made and used during the Late Holocene, although it is certainly possible that portions of the site date to older than two or three thousand years ago.

64

Chapter 5. Lithic Debitage

By Mark Tveskov and Christian Solfisburg

A total of 5,615 pieces of lithic debitage--the debris left over from the manufacture, finishing, and repair of chipped stone tools--was collected from the Hayes site during the Southern Oregon University excavations. This included 388 individual specimens recovered from fifteen 50 cm x 50 cm shovel test pits and 5,615 individual pieces recovered from twelve 1 m x 1 m excavation units. These artifacts were recovered from a total of 9.25 m3 of excavated sediment. Debitage was recovered from sediment screened over ¼-inch hardware cloth during the excavation of the shovel test pits and the northwest, northeast, and southwest 50 cm2 quads of the excavation units, and from sediment screened over -inch mesh during the excavation of the southeast quad. All debitage recovered from the Hayes site was counted and sorted by raw material, and all the debitage from the -inch sample was further sorted according to size, presence or absence of cortex, and by technological categories modified from Sullivan and Rozen (1985). From this sample, those flakes that exhibited a striking platform were further analyzed according to their platform and dorsal surface morphology (Tveskov et al. 2002; Tveskov et al. 1996; Byram and Connolly 1995; Musil and Minor 1995). The Hayes site assemblage could not be subdivided into distinct temporal or spatial components (see Chapter 3), and the debitage from all units was treated as a single assemblage. Since lithic debitage is the byproduct of the reduction of stone into useful implements, debitage analysis can provide insights into the function of a site. Activities such as raw material acquisition, initial bifacial reduction, final tool preparation, and tool repair are activities related to the availability of raw material, the place of the site in a larger settlement pattern, and cultural and technological traditions (Connolly et al. 1994:127; Sullivan and Rozen 1985). The analysis of the lithic debitage from a site can assist in determining the kinds of lithic reduction that occurred there. The Sullivan and Rozen method, employed for the Hayes site in a modified form, is commonly used as an initial level of categorization, although the information it provides regarding reduction technology is somewhat limited. Nonetheless, this classification method allows for comparability between data sets and is therefore used here as a first level of classification and analysis (e.g. Tveskov et al. 2002; Tveskov et al. 1996; Connolly et al. 1994; Musil and Minor 1995). Sullivan and Rozen's method views lithic reduction as a continuum rather than a set of discrete stages. Their analysis involves classifying debitage according to whether or not a specimen exhibits a single interior surface; specimens without a single surface are grouped as debris. Flakes with a single interior surface are then grouped according to the presence of absence of a striking platform. In our simplified scheme, we classed all flakes with a platform as platform flakes (i.e. subsuming the complete flake and broken flake categories proposed by Sullivan and Rozen) and those without platforms as flakes without platforms (i.e.

Sullivan and Rozen's flake fragments). These categories can be combined with other data drawn from the assemblage. Since greater amounts of cortex can be expected on flakes produced earlier in the reduction process, specimens with cortex on 100% of their dorsal surface were classified as primary flakes (Ahler 1990). Flakes with cortex covering less than 100% of their dorsal surface were classified as secondary flakes. Finally, flakes exhibiting no cortex were typed as interior flakes. The Hayes site debitage was also grouped into classes based on size--measured by maximum diameter. Although small flakes can be produced at any time in the reduction sequence, flake size generally decreases from the initial stages of tool manufacture to the final thinning and shaping of a finished tool, or the sharpening and repair of a broken tool (Stahle and Dunn 1982; Patterson and Sollberger 1978). Finally, platform and dorsal surface morphology were noted for platform flakes. Generally, the degree of platform preparation (abrading) and the number of facets on a platform (the result of removing flakes from a previously worked edge) will increase towards the later stages of lithic reduction (Byram and Connolly 1995; Musil and Minor 1995; Raymond 1989). Likewise, while initial core reduction debitage will generally exhibit fewer flake scars on their dorsal surface relative to their area, while flakes resulting from later stage reduction will have a higher dorsal flake scar-to-surface area ratio.

Results

Like most sites in southwest Oregon, the most common raw material represented in the Hayes site debitage assemblage is crypto-crystalline silicate, accounting for 96% (nisp=5013) of the flakes recovered from the excavation units (Table 7, Figure 41). Although a systematic study of the sources of crypto-crystalline silicate material in southwest Oregon is lacking, water worn nodules are commonly found in river gravel bars throughout the region, and in tabular form in metamorphic outcrops, such as at Coquille Point near the mouth of the Coquille River (Tveskov et al. 1996; Musil and Minor 1995). Smaller amounts (nisp=12) of other local material (mostly basalt or schist) was also recovered. Finally, 4% (nisp=202) of the Hayes site debitage assemblage were obsidian pieces that would have been imported from volcanic sources located east of the Cascade Range. Table 7. Material distribution by site area.

material ccs obsidian other Total Oven Area nisp 2142 95 3 2240 % 95.6 4.2 0.2 100 Bluff Area nisp % 1963 97.8 47 2.1 4 0.1 2014 100 Runway nisp 908 60 5 973 % 93.3 6.2 0.5 100 Total Nisp 5013 202 12 5227 % 95.9 3.8 0.3 100

66

other (<1%) obsidian (4%)

CCS (96%)

Figure 41. Distribution of material type for the Hayes site debitage. A total of 2,302 flakes--the -inch sample recovered from the southwest quad of the excavation units--was categorized according to their type and class (Table 8 and Table 9). Interior flakes dominate the assemblage, accounting for 97% of the crypto-crystalline silicate and 100% of the obsidian specimens. Most (57%) of the crypto-crystalline silicate specimens were classified as flakes without platforms, with 29% of the remainder being flakes with platforms and 14% of the assemblage comprised of the debris. Conversely, most of the obsidian debitage (70%) were flakes with platforms, with 29% of the remainder being flakes without platforms. Only 1% (nisp=1) of the obsidian debitage was classified as debris. The debitage were grouped into the following size classes: less than .5 cm in maximum diameter, 0.5-1 cm, 1-2 cm, 2-3 cm, 3-4 cm, and greater than 4 cm (Table 10, Figure 42). For the most common material classes (i.e. obsidian and crypto-crystalline silicate), the flakes were generally very small. For both obsidian and crypto-crystalline silicate, the majority of the flakes (82% and 69%, respectively) were under 1 cm in size, and fully 100% of the obsidian and 95% of the crypto-crystalline silicate were under 2 cm in maximum diameter. Table 8. Lithic debitage grouped by material and reduction type.

material CCS obsidian other Total Interior nisp 2108 115 5 2228 % 96.7 100 100 96.8 Secondary nisp % 59 2.7 Primary nisp % 15 0.6 Total nisp 2182 115 5 2302 % 100 100 100 100

59

2.6

15

0.6

67

Table 9. Lithic Debitage grouped by material and debitage class.

flakes with platforms nisp % 636 29.1 81 70.5 1 20 718 31.2 flakes w/o platforms nisp % 1246 57.1 33 28.7 3 60 1282 55.6 debris nisp % 300 13.8 1 0.8 1 20 302 13.2 Total nisp 2182 115 5 2302

material CCS obsidian other Total

% 100 100 100 100

Table 10. Hayes site debitage frequency by raw material and flake size (maximum diameter).

Material CCS obsidian other Total nisp 405 32 1 438 <.5 % 18.5 27.9 20 19 .5-1cm nisp % 1096 50.3 74 64.4 0 0 1170 50.9 1-2cm nisp % 561 25.8 8 6.9 4 80 573 24.9 2-3cm nisp % 96 4.3 0 0 0 0 96 4.1 3-4cm nisp % 19 0.8 0 0 0 0 19 0.8 >4 cm nisp 4 1 0 5

% 0.2 0.8 0 0.2

70 % within material class 60 50 40 30 20 10 0 <.5 .5-1cm 1-2cm CCS 2-3cm Obsidian 3-4cm 4-5cm

Figure 42. Size distribution of lithic debitage from the Hayes site.

68

Finally the platform morphology and the number of flake scars on the dorsal surface of all the platform flakes from the -inch sample were examined. A simple typology was used to classify platform morphology, and flakes were classified as having either a complex (multi-faceted or abraded) or simple (single faceted) platform (Table 11 and Table 12). This analysis shows some contrast between material types, with 38% of the obsidian flakes and 29% of the crypto-crystalline silicate flakes having complex platforms. To some degree, the obsidian flakes from the site were being struck from more intensely prepared and worked cores. This difference is also noted in measuring the ratio of the number of negative flake scars on the flake's dorsal surface to the overall flake size. For flakes with both complex and simple platforms, the obsidian had a much higher ratio in comparison to the cryptocrystalline silicate.

Table 11. Platform morphology by material.

material CCS obsidian Other total nisp 493 81 1 # with complex platforms 143 31 0 % 29 38.2 0

Table 12. Attributes of Hayes site flakes with platforms.

flake size 0-.5 cm 0.5-1 cm 1-2 cm 2-3 cm 3+ cm total 0-.5 cm 0.5-1 cm 1-2 cm 2-3 cm 3+ cm total 0-.5 cm 0.5-1 cm 1-2 cm total 0-.5 cm 0.5-1 cm 1-2 cm total nisp 66 266 130 31 4 497 29 74 32 6 2 143 7 39 4 50 11 18 2 31 mean dorsal scar/area 4.12±2.67 2.34±1.24 1.25±0.72 0.91±0.59 0.93±0.43 1.91±1.36 5.01±2.30 2.42±1.29 1.53±0.73 0.78±0.78 1.00±1.06 2.15±1.72 5.43±1.51 2.9±1.90 1.62±0.63 3.31±1.94 6.73±1.35 3.94±1.61 2.00±0.71 4.22±2.38

CCS, simple platform

CCS, complex platform

obsidian, simple platform

obsidian, complex platform

69

Summary

According to Sullivan and Rozen (1985), higher percentages of flake fragments and complete platform flakes might be expected to be produced by final tool preparation from blanks or bifaces or from the repair of damaged tools, whereas reducing large cores into bifaces or blanks results in the production of larger proportions of debris as well as primary and secondary flakes. The Hayes site debitage assemblage is relatively small and of a low density in comparison to other southwest Oregon sites (e.g. summary information in Tveskov et al. 2002:123; Winthrop et al. 1993:105-108) and is characterized by a relatively low proportion of large debris or cores in favor of relatively small interior flakes. The quarrying of lithic material or the initial reduction of stone into bifaces or blanks are clearly not activities indicated by the sample analyzed here. The size, type, class, and platform and dorsal surface morphology of the Hayes site assemblage is thus most indicative of the final stages of the lithic reduction and the repair of chipped stone tools. On the basis of our analyzed debitage assemblage, it seems reasonable to suggest that the occupants of the areas of the Hayes site examined by our excavation were bringing finished or nearly finished bifaces or blanks to the site, and the debitage that we recovered is largely the result of the subsequent finishing, use, and repair of these tools while on the site. There are minor differences, however, in how the two most common material classes--crypto-crystalline silicate and obsidian--were used at the site. The obsidian debitage is generally smaller, and was apparently worked more intensely. This pattern, common in southwest Oregon sites, is likely a reflection of the greater value of this tool stone and a measure of its relative scarcity. While crypto-crystalline silicate would have been available within the immediate range of seasonal round of people living on the South Fork Coquille River, obsidian would have had to have obtained form a great distance--from sources east of the Cascade Range.

70

Chapter 6. Ground and Pecked Stone Tools

By Nicole Norris The category of Ground and Pecked Stone Tools include all cobble- or pebble-sized rocks that have been "manufactured through mechanisms of abrasion, polish, or impaction," or are themselves used to "grind, abrade, polish, or impact" (Adams 2002:1). These types of wear patterns occur as a result of deliberate modifications to the stone or "from attrition due to use-wear" (Connolly 1991:101). A total of 19 pieces of groundstone were recovered and analyzed from the Hayes site. This small assemblage includes a number of food preparation tools including pestle fragments, hopper mortar bases, and a grinding slab, as well as handor hammer-stone pieces and polishing rocks utilized in the production of other tools. Also included in this assemblage are a number of unusual groundstone pieces with characteristics that do not match known ground and pecked stone tool types. The groundstone tools were manufactured from locally available materials such as granite, basalt, and sandstone with a high occurrence of aptly shaped river rocks utilized as tools or in tool production. One aspect of the Hayes site important to note was the significant number of fire-cracked river rock spread across the entirety of the site. Many of these had slight but noticeable characteristics associated with the manufacture of groundstone tools, but had been so modified by fire that they were unidentifiable as specific groundstone tools. The majority of the groundstone pieces recovered from the Hayes-Airport site appear to have been expediently manufactured and resemble many of the groundstone tool assemblages recovered from other Oregon interior valley sites (e.g. Tveskov et al. 2002; Connolly 1991). The ground and pecked stone tools from the Hayes site were classified according to a typology developed from Adams' (1996, 2002) study of groundstone tools in the American southwest and local artifact descriptions by Connolly (1991; Connolly et al. 1994) and Norris (Tveskov et al. 2002: Chapter 6). The Hayes-Airport site groundstone collection was broadly classified into five categories, Hand-Stone, Netherstone, Pestle, Manuport, and Unknown Artifacts.

Handstones

The category of Handstone is a generic term for all hand-held stones with a variety of non-specific attributes such as abrasive wear, faceted sides or edges, and pitting due to impact. This category includes hammerstones or mauls, manos, and abraders/polishing stones. Surface areas of these artifacts have been damaged by moderate or heavy use which has resulted in pitting or flaking of the ends or sides. Also characteristic of handstones are tools with one or more sides smoothed and flattened with surface striations that indicate

their use with lateral or medial reciprocal strokes (Adams 2002). Each characteristic may occur singularly or in any combination on a single tool (e.g. Tveskov et al. 2002:103), as evidenced in all of the artifacts in this groundstone assemblage. The Hayes site collection contains four complete and three partial handstones. All partial handstones have at least one end faceted or one surface flattened, while the complete tools have damage to both distal and proximal ends with one or more surface smoothed with wear.

Hammerstones

Hammerstones are irregularly shaped rocks of "hand gripping size," selected for their useful size and shape, and are characterized by localized percussion damage on one or more end or edge (Stewart 1994:42). The use-wear on a hammerstone suggests that primarily the edges, but sometimes the broad surfaces as well, were struck forcefully against another object. The action of a hammerstone was meant to chip and smash away unwanted portions of the other item's surface, as well as to act as a hammer. Damage apparent on hammerstones includes impact fractures and chipping to the ends and sides. The Hayes site groundstone assemblage contains only two clear examples of hammerstones, while many of the other groundstone tools, such as the pestles and abraders, exhibit signs of wear associated with hammerstones. This is a common observation during the analysis of an expediently manufactured collection of groundstone artifacts. Specimen 02.03-78 is a broken river-worn cobble with deeply pitted proximal and battered and broken distal ends. The ventral surface has also been smoothened with a use-wear pattern consistent with manos and abraders. The other hammerstone specimen (02.03-111) can be described as a `maul'--a heavy, complete river cobble with smoothed handle impressions on either side and a medial indent with pitting consistent with hammer activities (Figure 43).

Figure 43. Hammerstone from the Hayes site (SOU-02.03-78).

72

Abraders

Abraders are small cobbles that fit into the hand and have a texture useful for shaping or polishing the surface of other items. These tools are used to alter the appearance of other objects by creating either a roughly abraded or a smoothed and shiny surface. The texture of an abrader determines the type of damage done to the opposing surface, so that a finer textured stone may be used more to polish than to abrade (Adams 1996:11, 32). Abraders are generally associated with the manufacture of pottery, bone, or wood artifacts, but are also used in the manufacture of chipped and groundstone tools as well. No pottery, bone, or wood artifacts were found in association with the Hayes site. There are five examples of abraders, three complete (02.03-68, 02.03-92, and 02.03126) and two broken (02.03-46 and 02.03-8), in the Hayes site groundstone assemblage (Figure 44). Specimen 02.03-68 is an elongated river cobble shaped like a pestle with flattened ventral and dorsal surfaces and wear striations consistent with abrasive reciprocal hand strokes. This tool exhibits signs of pitting on both distal and proximal ends, as well as minimal pecking scars on both lateral surfaces. The remaining examples of Hayes site abraders vary in shape and size, from round flat river cobbles to triangular basaltic andesite, with at least one surface exhibiting use-wear patterns consistent with abrasive actions.

Figure 44. Abraders from the Hayes site. Top to bottom: 02.03-68, 02.03-126, 02.03-92.

73

Netherstones

Netherstones serve as bases upon which other artifacts were shaped or where food substances such as berries, grains, nuts, roots, and seeds were processed with a handstone (Tveskov et al. 2002:105). They are generally too large to be easily transported, and are modified or smoothed by some sort of grinding or percussive action. At least one surface of all netherstone tools has been modified through abrasion or percussion. In general, the term `netherstone' can be used to "subsume all bottom stones against which something else has been worked" (Adams 1996:28). In regards to the Hayes site, this category includes hopper mortar bases and grinding slabs. Surface observation and collection at the Hayes Site resulted in the recovery of three netherstone artifacts; two hopper mortar bases and one grinding slab.

Hopper Mortars

Hopper mortars are groundstone implements characterized by shallow basins within which substances were processed or reduced through grinding and crushing actions (Adams 1996:26). Typically, a woven basket was placed in the basin of the mortar and a handstone or pestle was used to process and crush the contents in the basket. The size and form of hopper mortar bases can vary greatly, from large bedrock mortars to small pebble mortars (Tveskov et al: 2002:106). Hopper mortars were a main component in the reduction and processing of food materials such as grains, seeds, nuts, and roots and are common lateHolocene groundstone tools throughout Oregon's interior valleys. The two examples of hopper mortar bases from the Hayes site resemble the type generally found in southwestern Oregon (Figure 45). They are made from locally available granite or basaltic river rocks and the use-wear and damage patterns include abrasive scratches, slight pitting, small impact fractures, and basin discoloration. The basin of a hopper mortar base is typically shallow and smoothed from the pounding action of a pestle against the materials within the hopper or mortar basket. The Hayes site hopper mortar bases exhibit signs of these traits. During excavation at the Hayes-Airport site, two hopper mortar bases were collected; one complete specimen (02.03-114) and one broken (02.03-113). The complete hopper mortar base is a large and heavy basaltic river cobble with dimensions of 33 cm x 26 cm. The size and weight of the mortar suggest that it was placed on the ground and used with a basket and pestle to crush food materials. The entirety of the dorsal surface is discolored most likely due to weathering processes, while the basin exhibits clear discoloration due to wear patterns. The broken hopper mortar base (02.03-113) shows clear signs of abrasive wear in the shallow basin and as with the complete Hayes site mortar, was formed from a heavy granite/basaltic river cobble and probably set on the ground and used with a basket and pestle.

74

Figure 45. Hopper mortar bases from the Hayes site. Left: 02.03-114, right: 02.03-113.

Grinding Slabs

Grinding slabs are stable stone platforms on which food material was ground or pounded with a hand-held stone. Damage wear on grinding slabs show distinct areas of polish, abrasive striations, pecking, or food residue (Adams 1996, 2002; Stewart 1994). Generally, a grinding slab is chosen for the flat shape of the stone cobble and is further ground and flattened through wear-use. Grinding slabs differ from hopper mortar bases in that they retain their flat and broad surfaces and are used to process food materials without the use of a woven basket (Tveskov et. al. 2002: 106). In the North American southwest, grinding slabs are often referred to as `metates' and are found in association with `manos' (Adams 1996:23). In the Hayes site groundstone typology, all non-hopper mortar netherstones are referred to as `grinding slabs.' There is only one example of a grinding slab in the Hayes site groundstone assemblage. The collected specimen (02.03-112) was manufactured from a granite/basaltic river cobble and exhibits a minimally worn ventral surface (Figure 46). The use-wear patterns on the ventral side show a smooth surface with slight pitting and abrasive scratches, indicative of wear formed with the reciprocal stroke of a handstone. As with the Hayes site hopper mortar bases, the grinding slab is heavy and was probably intended to be utilized as a sedentary tool.

75

Figure 46. Grinding slab recovered from pot hunter's hole at the Hayes site. This artifact was inscribed "R.I.P." when it was dug up.

Pestles

Pestles are typically classified as elongated handstones used to crush, pulverize, or grind materials placed in a bowl, mortar, or on a stone or wood base (Adams 1996:30). Pestles were utilized to crush both food and non food items such as pigments, medicines, and tobacco mixtures (Stewart 1994:64). The use-wear patterns on a pestle include pitting, chips, abrasions, scrapes, scars, and impact fractures with the nature of the damage a reflection of the type of stroke used with the pestle. When the weight of the stone pestle supplies most of the force used to crush materials, small evenly spaced impact pits are formed. Deep impact fractures, chips, and distal end breakage indicate the pestle was used with additional force or a plunging stroke from a distance above the impact point (Tveskov et al 2002:107). In general pestles can vary greatly in their design and manufacture, from highly formalized and decorated to less elaborate and expediently manufactured examples. With expediently manufactured pestles, appropriately sized cobbles were selected and used with minimal design modification. Like many of the other pestles collected in the interior valleys of southwestern Oregon, all of the pestles recovered from the Hayes site were expediently manufactured. The Hayes collection contains one complete (02.03-75) and two broken pestles (02.03-11, 02.02-96) (Figure 47). Unlike larger cylindrical pestles, the unbroken Hayes pestle is small (11 cm x 6 cm) and has a triangular cross section with three sides utilized for grinding or abrading purposes. Both the distal and proximal ends exhibit slight pitting and abrasive wear. The two broken pestles are both cylindrical with the distal ends severely

76

battered and broken with heavy pitting of the proximal ends. Specimen 02.03-96 shows pitting along the sagittal spine of the ventral surface as if it was utilized as a hammer-stone as well as a pestle. With many expediently manufactured groundstone tools, these pestles may have been used for tasks other than those typically associated with more elaborately formed pestles. These tasks would explain the intense hammer and abrasive wear patterns evident on all three of the pestles collected from the Hayes site.

Figure 47. Pestles from the Hayes site. Top to bottom: 02.03-96, 02.03-11, 02.03-75.

77

Manuports

Manuports are all river-worn cobbles which have been transported by humans from another location. Manuports are classified as such when there is no natural (or non-human) explanation for their occurrence in a specific site. The Hayes site is located approximately 30 meters above the south fork of the Coquille River. "Geomorphological standards dictate that cobbles worn by water movement are transported downstream and downhill" (Tveskov et al. 2002:108). The Hayes site had a great many river-worn cobbles spread across its entirety. Only two were collected and catalogued. The occurrence of these river-worn cobbles at the Hayes site indicates human action in the uphill transportation of said cobbles from the river below. Therefore, all river rocks collected from the Hayes site had to have been carried uphill by some means, presumably human. As previously mentioned, there were a great many river-worn cobbles identified in and around the river area, yet only two were collected from the Hayes site.

Unknown Artifacts

The Unknown Artifact category is a catch-all for those groundstone artifacts that show some signs of human modification, yet their uses remains unknown. The Hayes site groundstone assemblage contains four examples (Figure 48 and Figure 49). The first specimen (02.03-31) was first thought to be a small weight or net sinker. The small riverworn cobble is shaped like a kidney with a groove surrounding the proximal end. After further analysis, similar partial grooves on the distal end were identified, which indicated the possibility that the grooves were naturally formed inclusions and not shaped by human hands. The second specimen (02.03-9) is a small, flat, oval river rock with a smooth, flat ventral surface. On this flat surface, there are eleven uniform etchings that lie horizontally across the face of the stone. This artifact was recovered in situ in association with Feature 2--the burned rock `oven' feature near the bluff edge on Terrace III. The exact meaning and function of this decorated piece can not be known, but it is at least possible that this stone was some sort of personal marker or counting stone as it fits comfortably in the palm of the hand. Specimen 02.03-7 is a possible pinch pot fragment or the start to a small bowl. The piece was manufactured from granite, is small (2.1 cm x 3 cm), and contains a slight rim around the clearly etched concave dorsal surface. The fourth unknown artifact is specimen 02.03-18 and differentiated from the other pieces in that it was manufactured from firehardened clay. This artifact is small with dimensions of 3.7cm x 2.9 cm and exhibits an elongated concave shape on the dorsal surface. Initially, the excavators suspected this artifact was a pipe fragment, but on closer inspection, there appear to be slight rims on either end of the concave shape, and a thick groove on the convex ventral surface. These characteristics do not match those associated with known pipe fragments.

78

Figure 48. Possible net sinker or pendant from the Hayes site (SOU-03.02-31).

Figure 49. Incised cobble from Feature 2 at the Hayes site (SOU-02.03-09)

79

Summary Surface survey and excavations at the Hayes-Airport site resulted in the analysis of 19 groundstone artifacts, all of which were identified on site and collected. The assemblage is dominated by expediently manufactured abraders, pestle fragments, hopper mortar bases, and hammerstones; all manufactured from locally available materials. The frequency of these tool types spread across the site indicates that the processing of plant materials-- possibly camas or acorns--was an activity that occurred at the Hayes site. The remainder of the small groundstone assemblage is indicative of utilitarian activities such as hammering or flint knapping. Two somewhat enigmatic pieces--a palm-sized stone purposefully incised with eleven uniform parallel lines and a piece of baked clay that may or may not represent a pipe fragment--were also recovered.

80

Chapter 7. Faunal Remains

The Southern Oregon University excavations at the Hayes site resulted in the recovery of 81 individual faunal specimens weighing a total of 74 grams. These were recovered from sediment excavated from the three quads of the 1 m x 1 m excavation units and the 50 cm x 50 cm shovel test pits that were screened in the field over ¼-inch hardware cloth and the southwest quads of the 1 m x 1 m excavation units that were screened in the field over ¼-inch hardware cloth. This small collection was catalogued, counted, weighed, and analyzed to identify the lowest possible taxonomic classification using the standards collection at the Southern Oregon University Laboratory of Anthropology, the U.S. Fish and Wildlife Service Forensics Lab in Ashland, Oregon, and several zooarchaeological reference manuals (e.g. Olsen 1964, 1968; Hess and Wapnish 1985). All identified specimens were quantified to the number of individual specimens (NISP) and were weighed to the nearest 0.01 gm. In addition, the faunal remains were each examined for evidence of human or natural modification (Binford 1981; Lyman 1994).

Results

Table 13 summarizes the small and highly fragmented and burned assemblage of faunal remains recovered from the Hayes site. The majority of the faunal remains come from two locations. First, the contiguous excavations of shovel test pit STP E and excavation units Unit 1 and Unit 6 over the Feature 1 oven feature all encountered, near the surface, a scatter juvenile cervid remains. I doubt these remains are associated with the archaeological deposits at the site, and most likely represent the remains of a single juvenile deer (Odocoileus sp.) that had died or was killed on the site in the not too distant past. The majority of the remaining faunal assemblage is comprised of tiny fragments of heavily calcined bone that were recovered from the disturbed sediment on the leading edge of the Terrace IV in Unit 12. A few other small pieces of calcined bone were recovered from Unit 7 and Unit 8 in association with Feature 1, and from Unit 10 in association with Feature 2. Among the few remains likely of ancient cultural origin, heat treatment was the most commonly observed modification. In fact, setting aside the recently deceased deer, all of the recovered faunal remains from the Hayes site were heavily calcined, the result of either human modification (i.e. cooking or landscape burning) or natural processes such as wildfires in the area. No cut marks or polishing were observed on any specimen.

Table 13. Faunal remains recovered from the Hayes Site.

unit 1 1 4 4 4 4 4 6 6 6 6 6 7 12 12 12 12 12 12 2 2 8 10 stp F stp E stp E stp E stp E stp E stp E

level I I I I I I I I I I I I I I II III V I I II II IV I I I I I I I I

quad A D B B B D D A B B C D D D B B B B A B D C D

taxon species element weight nisp comment mammal medium vertebra 4.87 2 mammal medium centrum epiphysis 0.01 1 mammal medium sternum 1.59 1 mammal cervid carpal/tarsal 0.84 1 probably deer mammal cervid hoof 3.11 1 probably deer mammal medium rib diaphysis 4.43 4 mammal medium proximal rib fragment 0.62 mammal medium vertebra centrum 10.37 1 mammal medium rib 10.77 4 mammal medium nid 0.21 1 mammal medium rib 5.06 2 mammal nid nid 0.2 2 nid nid nid 1.28 1 nid nid nid 0.05 13 calcined nid nid nid 0.06 3 calcined nid nid nid 0.02 8 calcined nid nid nid 0.13 1 calcined nid nid nid 0.07 1 calcined nid nid nid 0.1 3 calcined nid nid nid 0.36 4 calcined nid nid nid 0.45 2 calcined nid nid nid 0.55 1 calcined nid nid nid 0.01 3 calcined nid nid nid 0.12 1 calcined mammal cervid vertbra 11.88 4 probably deer mammal medium rib 10.48 5 mammal medium centrum epiphysis 0.91 4 mammal medium sternum 3.69 2 mammal medium rib 1.75 2 mammal nid nid 0.03 2 notes: weight in grams, nisp=number of individual specimens, nid=not identified

82

Summary

The excavations at the Hayes site resulted in the recovery of a small and highly fragmentary and mostly burned and calcined assemblage of animal bone. With the exception of the scattered carcass of a recently deceased deer, no specimen could assuredly be identified to a taxonomic category. No cut marks were observed on any faunal specimen. It is therefore difficult, if not impossible, to use these data to assess the significance of hunting activities It is tempting to argue that local soil conditions impede the preservation of faunal remains, and that much of the bone originally deposited at the site has simply burned away, leaving just a few meager remains, as is common in many interior or non-shell bearing southwest Oregon archaeological sites (e.g. Tveskov et al. 2002:119). Additionally, the use of ¼-inch screen during most of the excavation may have additionally impeded the recovery of faunal remains that were still extant. At the very least, the presence of at least some bone at the site considered with the recovery of chipped stone projectile points suggests that hunting was an activity undertaken by the inhabitants of the Hayes site.

83

84

Chapter 8. Summary and Discussion

Archaeological investigations were conducted by archaeologists from Southern Oregon University at the Hayes site (35CS196), located on a terrace overlooking the South Fork Coquille River in Coos County, Oregon. This work was conducted in collaboration with the Coquille Indian Tribe and was part of a larger effort to understand the distribution of archaeological sites in the South Fork Coquille River drainage and to place these sites into a historical, ethnohistorical, and oral tradition context. The Hayes site project was also designed in light of the severe looting that continues to take place at the site; in addition to learning about the size, depth, age, and cultural significance of the site, we hoped to document the extent of the damage to the site deposits and to use the project as a means to raise public and professional awareness about the impact of looting on archaeological sites such as the Hayes site. Work at the site took place for four weeks in the summer of 2002 during a Southern Oregon University archaeological field school. The Hayes site sits on an uplifted terrace on the west side of the South Fork of the Coquille River, upstream from the modern-day community of Powers. The site is part of a large complex of archaeological sites that lies across Powers prairie to the north and Dingbat Flats to the south, all oriented around the confluence of Salmon Creek (flowing from the west) and Mill Creek (flowing from the east) with the South Fork. The local environment includes an extensive system of riffles immediately below the site ideal for salmon and eel fishing, an oak and grassy prairie/meadow complex on the site itself and in the surrounding foothills that was likely once rich in botanical resources such as camas and acorns, and access--via ridgeline trails--to upland forests and meadows in the mountains surrounding the site. Ethnohistoric and ethnographic information suggests that Powers prairie in general and the Hayes site in particular sat at a hub of a major trail system that linked the ancient Athapaskan landscape of southwest Oregon; trails passing through Powers led to the southern Oregon coast, the lower Rogue River region, the interior Rogue River Valley, and the Interior Umpqua Valley. Oral historic and ethnohistoric sources suggest that a local group of Coquille Indians known as the Nati jí dunne (People Who Lived by the Big Fish Dam" lived in villages in the Powers/Hayes site area, and these villages were highly populated and wealthy prior to the incursion of Euro-Americans in the early 19th century. Archaeological surveys conducted by Southern Oregon University (Tveskov 2001) and interviews with local artifact collectors attest to the heavy archaeological footprint on the prairies in and around Powers. Southern Oregon University's work at the Hayes site included the excavation of 16 individual 50 cm x 50 cm shovel test pits, 12 individual 1 m x 1m excavation units as well as the extensive mapping of the site's natural and cultural features. This work resulted in the excavation of 9.25 m3 of sediment covering less than 0.1% of the site. A rich assemblage of

artifacts was recovered during this project, including 5,615 pieces of lithic debitage, 81 pieces of animal bone, 100 individual chipped stone tools, 19 individual groundstone artifacts, and the identification of two large cultural features comprised of burned river cobbles that likely the represent the remains of roasting ovens. No organic material of unquestionable cultural association was recovered from the site for radiocarbon dating. However, the recovery of several artifacts from the site by local collectors reminiscent of Gunther Pattern assemblages from elsewhere in southwest Oregon suggests that the Hayes site was in use during the last 1,700 years (c.f. Connolly 1991). However, the excavated assemblage of leaf-shaped and Coquille Series projectile points similar to other Glade Tradition assemblages in southwest Oregon (including sites on the Coquille River such as the Standley site) suggest the possibility that the site was in use during earlier millennia. Given that the immediate site area was described by early White pioneers as being used by Indian people as late as the early 1850s, it is perhaps not unreasonable to surmise that the Hayes site was in use either continuously or intermittently for at least the last 2,000 years, and perhaps for several thousand years prior to that. Surface observation and the shovel test pit excavations indicate that the Hayes site covers an estimated 10,000 square meters. Cultural deposits are spread across several stepped terraces from the foothills to the west to the river on the east. Based on our excavations and the testimony of local residents, the richest and most diverse concentrations of artifacts were found on the higher (and, during the winter, dryer) terraces, especially on the north end of the current airport runway. Unfortunately, this portion of the site has been most severely damaged, as the runway has been extensively machine graded. Most of the Southern Oregon University excavations were concentrated on the terrace immediately below and east of the runway, and here, scatters of cultural material are found clustered around large burned rock oven features. This area too has been badly damaged, this time by extensive looting that in some areas appears to have completely obliterated cultural features, as well as by bioturbation (particularly rodent activity), plowing, and cliff face erosion. The Hayes site, like many archaeological sites on the South Fork Coquille River in particular and southwest Oregon in general, has been and continues to be severely looted. The site, which is easily accessible to the public, appears to be a favored location for causal artifact collecting as well as for more concentrated efforts to excavate artifacts for sale. Based on our excavations and information provided by local residents, I would estimate that conservatively, more than half of this site has been destroyed by looting, construction activities, bioturbation, and erosion. Nevertheless, conservation of the cultural material still intact at the site remains a concern: the Hayes site was looted immediately prior, during, and after our excavation. While cultural material was abundant across the entire site, our excavations suggest that in comparison with other southwest Oregon archaeological sites, there is not, overall, a very high density of artifacts (Table 14), and the assemblage is likewise relatively simple. The most common artifacts were very small pieces of lithic debitage that, based on their size, material, and morphology, are indicative of the final preparation and repair of chipped stone tools such as projectile point. Very little lithic quarrying or early stage core reduction, if any at all, was occurring at the site. Locally available crypto-crystalline silicate (i.e. chert) was the most common material used, although significantly smaller quantities of imported obsidian 86

and some coarser materials such as basalt were used as well. The chipped stone tool assemblage is dominated by expediently used flakes and unifacial end scrapers, with smaller numbers of bifaces fragments and fragmentary projectile points (usually the bases) found as well. The assemblage of ground and pecked stone artifacts is likewise not very large, and is limited to expediently-used or minimally modified cobbles used as pestles, hammerstones, hopper mortar bases, and grinding slabs. It should be noted, however, that formal artifacts have been found at the site by collectors, particularly on the upper, western terraces at the site. These artifacts include an extremely elaborated groundstone adze handle and a decorated antler pipe collected from the site by Jackie Hofsess.

Table 14. Density of lithic debitage, projectile points, chipped stone tools, and groundstone artifacts from the Hayes site compared against eleven components from Douglas and Jackson County archaeological sites with high debitage densities based on data assembled by Winthrop (1993:105-108; see also Tveskov et al. 2000:123).

amount debitage/ excavated cubic meter Site Windom site 7 5484.00 Section Creek 22.3 3147.17 DO421 1.1 2993.64 DO205 upper 4.9 1562.65 DO389 2.25 1377.78 JA23 14.94 1204.28 DO205 lower 3.5 1091.71 Elk Creek 2 159 1040.50 JA27A 24 1020.00 DO401 2.8 985.71 DO36 Crispen 5.4 981.48 Hayes 9.25 607.023 Note: amount excavated: cubic meters. projectile points/ cubic meter 12.57 4.98 0.91 9.18 42.67 4.95 30.00 7.17 7.92 1.79 6.11 3.03 total tools/cubic meter 24.43 39.87 25.45 43.88 98.22 26.64 75.71 41.00 64.92 20.00 27.96 10.81 groundstone/ cubic meter 4.14 0.49 0.00 0.00 4.88 0.20 0.57 0.50 3.62 0.00 0.93 2.05

The characteristics of the Hayes site, at least in the relatively intact areas on Terrace III around the burned rock oven features, are indicative of a relatively narrow range of activities. The lithic debitage and chipped stone tool assemblage suggest that the site was used as a base from which big game hunting took place this assemblage is comprised primarily of broken projectile points, debitage resulting from the finishing and repair of stone tools, and a relatively limited range of bifaces and scrapers likely used to process killed game. The ground and pecked stone artifact artifacts would have been used in the processing of acorns, camas, and other plants, and these resources, in accordance with southwest Oregon oral tradition, would have been roasted on large earth ovens such as represented by Feature 1 and Feature 2. The assemblage of artifacts recovered by both artifact collectors and the Southern University excavations fits the model of settlement and subsistence developed from our

87

archaeological survey and oral history research (e.g. Chapter 2 and Tveskov 2001). Resource-rich river bank areas such as the confluence of Mill Creek, Salmon Creek, and the South Fork Coquille River were the focus of settlement and subsistence; there, permanent villages were established on upper terraces, above lower surfaces that were subject to seasonal flooding. The most artifact rich (but most damaged) part of the Hayes site is on the airport runway terrace (Terrace IV) and the knoll to the west of the runway (Terrace V), and if there was a permanent village in this area, this is likely where it was. Such a village would have been owned by one or more families that would play host to their relatives at certain seasons of the year when the local resources that they controlled became available. In the case of the Hayes site vicinity, such seasonal activities likely included at least the spring and fall salmon fishing, summer eel fishing, spring camas harvesting, and fall acorn harvesting. At such times visiting households would camp around the host village, including on the lower terraces such as near the Feature 1 and Feature 2 excavations. The archaeological remains in this area of the site are indicative of repeated use on a temporary, task specific basis.

Discussion

The limited archaeological work at the Hayes site supports in a general way the models of social experience and settlement and subsistence practices derived from oral history, ethnohistory, and archaeological survey. I have argued that a distinguishing feature of southwest Oregon Indian culture and sociology is the primacy put on household independence within a complex web of local and regional social relations (e.g. Chapter 2 above, Tveskov 2003, 2002, 2000; Tveskov and Erlandson 2003; Tveskov et al. 2002; see also Byram 2002; Connolly 1991, 1986). In contrast to Indian people of the northern Oregon coast or the lower Columbia River, families maintained a fundamental political and economic independence, and this independence has roots deep in the Holocene. This independence was reinforced alternatively through the social experience of household economic and spiritual activities in the isolation of seasonal camps in hinterland locations-- where particular aspects of household identity, including patterns of social relations of gender and authority--were recursively reinforced in the habitus of day-to-day life--as well as in the public arena of permanent villages, where individual households and their constituent members actively negotiated their status and identity against other households through participating or hosting seasonal gatherings for economic, social, or spiritual purposes. The South Fork Coquille River provides an ideal setting to examine these processes. Oral traditions and ethnohistoric data suggest that the region, while isolated and marginalized today--was previously a central fixture in the political and social landscape of southwest Oregon. Trade in surplus commodities, exchanging daughters in marriage to other families, and hosting and visiting for spiritual, social, and economic regions were essential social practices in the negotiation of status for southwest Oregon households, and the South Fork Coquille River, by virtue of its geography and biota, was ideally situated. The prominent families of the two local groups of Coquille Indians that traditionally lived along the South Fork--the Chochrela dunne and the Nati jí dunne--were able to negotiate their status and wealth by virtue of their access and control over the economic riches of the South Fork Coquille River as well as through their central geographic position astride trails that

88

linked together all corners of the pre-contact Athapaskan landscape of southwest Oregon. The archaeological record at the Hayes site in particular and the South Fork Coquille River in general is replete with the remains of substantial villages located on elevated terraces that afford relatively easy access to both ideal fisheries and to trail heads that lead beyond the valley, as well as with the scattered remains of untold numbers of camp sites. Such sites were likely used on a seasonal basis by households traveling both within and from without the South Fork drainage for seasonal economic activities that were imbedded in complex sets of social and cultural relationships. The excavations at the Hayes site, however, underscore some basic and somewhat problematic aspects of the archaeology and prehistory of southwest Oregon. Two interrelated issues include the difficulty in obtaining reliable absolute dates from cultural deposits and the truly fascinating and as yet unresolved relationship between Gunther Pattern and Glade Tradition archaeological assemblages (see Tveskov et al. 2002; Connolly 1991, 1986 for discussion). With regards to the first, it remains the case that southwest Oregon sites in general--particularly away from the coast--are highly bioturbated and infused with large quantities of charred organic material undoubtedly of non-cultural origin. Coupled with a local environment that is geologically highly active, archaeologists are most often left with stratigraphically undifferentiated palimpsests of material, more often than not of late Holocene age. To date, only a single archaeological site--the Marial Site on the middle reaches of the Rogue River--has yielded a convincingly long sequence of relatively undisturbed stratigraphic layers of cultural origin reaching back through the middle Holocene into the early Holocene. Clearly, more geoarchaeological research is required in the interior valleys of southwest Oregon to identify both older surfaces as well as stratified sites that could contain long cultural sequences of material. Regarding the culture history of southwest Oregon, research at the Hayes site, the Windom site in the upper Rogue River valley, the Marial site, the Standley site on the Middle Fork of the Coquille River (Connolly 1991), the Narrows and Martin Creek sites on the Umpqua River (O'Neill 1989), and many others provide convincing evidence of the persistence of very `old' looking artifacts--particularly diamond- or leaf-shaped `Coquille Series' projectile points--into the late Holocene period. Connolly (1991, 1986) and I (Tveskov 2003; Tveskov et al. 2002) have argued that the persistence of such assemblages is a reflection of the persistence and importance of household independence and mobility throughout the Holocene period. The oldest archaeological sites in southwest Oregon are the remains of small groups of highly mobile people inhabiting seasonal camps utilizing Glade Tradition artifacts--assemblages principally defined by edge-ground cobbles and leafshaped, diamond-shaped, and side-notched projectile points. Such sites continue to be used in the late Holocene, but additional artifact classes are generally added to their assemblages. In hinterland locations, these include in particular barbed projectile points generally referred to as Gunther Barbed points. Late Holocene sites along the coast and the major river trunks include additionally a wider range of more formal artifacts such as obsidian blades, flanged pestles, steatite grease bowls, and concave-base projectile points. What processes can account for these changes in the late prehistoric period? Winthrop (1993) created a functional and temporal typology for a large number of archaeological components in southwest Oregon, and concluded that over the course of the Holocene, the human population of southwest Oregon gradually increased. She suggests 89

that settlement patterns were at first highly mobile, but after 4,000 years ago, semipermanent villages began to appear, at first along the main river trunks and then, after 2,000 years ago, along major tributary streams as well. In this `processual' scenario, the appearance of permanent villages and elaborate artifacts can be seen as a material manifestation of the development of social ranking in a more crowded and politically complex social landscape. Alternatively, Connolly (1991, 1986) argues that the immigration of Athapaskan-speaking and Algic-speaking people into southwest Oregon and northern California after 1,700 years ago played a significant role in the relatively sudden appearance of the Gunther Pattern assemblages and the reconfiguration of the regional social landscape. In this more `culture historical' scenario, these newcomers brought with them a Northwest Coast-style maritime/riverine adaptation that included settlement patterns anchored by permanent villages as well as ranked social relations. Within this discussion lies the particularly interesting (even vexing) problematic distribution of the Gunther Barbed projectile point. The geographic and temporal range of this artifact class maps closely--but not exactly--to the geographic and temporal range of Athapaskan and Algic peoples. Is it then an artifact diagnostic to culture in the traditional `culture history" sense? Such points are indeed clearly associated with historically described Athapaskan villages (particularly along the coast), but they do not simply `replace' sidenotched, leaf-shaped, or diamond-shaped points across southwest Oregon. Why then, are there almost no Gunther Barbed points found along the upper forks of the Coquille River? Virtually no such points are found in the Hayes site, the Standley site, or in the extensive amateur collections I examined from the South Fork Coquille River. These are areas that were, according to oral tradition, occupied by Athapaskans. Simultaneously, similar barbed points are found in the upper Rogue River valley in areas demonstrably occupied by Hokan or Penutian speakers (e.g. Tveskov et al. 2002). Perhaps the Gunther Barbed point represents a functional artifact class, and is used side-by-side in some sites or alternatively in other sites with foliate points depending the particular task at hand. The exact function of the Gunther Barbed point is unclear. Its presence in upland sites in the interior Rogue River and Umpqua River valleys suggests its use as an arrow point associated with big game hunting, but its prevalence (often along with concave-base points) on river-edge, estuarine, and outer coast village sites suggests the possibility that it could have been used as a harpoon point as well. Once again, however, the apparent absence of either type of point along the South Fork and Middle Fork Coquille Rivers throws this interpretation into doubt. Presumably, people in these areas were also hunting elk and deer and spearing fish. It is reasonable to suppose there is an element of truth in both Connolly's and Winthrop's models of southwest Oregon prehistory. The development of social ranking and "Northwest Coast" style cultures in southwest Oregon was part of a larger trend up and down the west coast of North America towards the evolution of more socially complex societies--a trend possibly driven by some combination of demographic and environmental pressure--but the immigration of the Athapaskans and Algics was a historical event that clearly would have had a major transformative affect on local adaptations, material culture, and social relations. The archaeological record, oral traditions, and ethnohistory of southwest Oregon all indicate that Indian people of the region did indeed begin to arrange themselves in more socially complex ways during the late Holocene. At the same time, 90

however, they maintained a primacy on household independence in a historically and culturally particular way, a primacy on household independence that was driven, in part, on the precedent and legacy of the millennia of social practice that emphasized such independence and mobility (Tveskov 2003; Tveskov et al. 2002. Clearly, more archaeological work needed to examine some of the particulars of Winthrop's and Connolly's models. As described above, coping with the dramatic geomorphology of the region and finding additional deeply stratified sites would be a good place to start. Additionally, more work on non-village sites such as seasonal camps in hinterland locations, lithic quarries, estuarine fish weirs--the domain of household social and economic experience--could help flesh out our understanding of the complex interplay of the persistence of long standing cultural traditions against the innovation of new sets of social arrangements and adaptations that occurred over the last several thousand years.

91

References Cited

Adams, Jenny L. 2002 Groundstone Analysis: A Technological Approach. University of Utah Press. 1996 Manual for a Technological Approach to Grindstone Analysis. Center for Desert Archaeology. Tuscon, Arizona.

Ahler, Stanley A. 1990 Mass Analysis of Flaking Debris: Studying the Forest Rather Than the Tree. In Alternative Approaches to Lithic Analysis, edited by Donald O. Henry and George H. Odell, pp. 85-118. Archaeological papers of the American Anthropological Association 1. Ames, Kenneth M. 1996 Chiefly Power and Household Production on the Northwest Coast. InFoundations of Inequality, edited by T.D. Price and G.M. Feinman, pp. 155187. Plenum Press, New York. 1994 The Northwest Coast: Complex Hunter-Gatherers, Ecology, and Social Evolution. Annual Reviews of Anthropology 23:209-229.

Ames, Kenneth M. and Herbert D.G. Maschner 1999 Peoples of the Northwest Coast: Their Archaeology and Prehistory. Thames and Hudson, London Barnett, Homer G. 1937 Culture Element Distributions, IX: Gulf Of Georgia Salish. University of California Anthropological Records 1 (5):221-295. Beckham, Stephen Dow 1971 Requiem for a People: The Rogue River Indians and the Frontiersmen. University of Oklahoma Press, Norman. Berreman, Joel V. 1944 Chetco Archaeology: A Report of the Lone Ranch Creek Shell Mound on the Coast of Southern Oregon. American Anthropological Association. General Series in Anthropology 11. Binford, Lewis R. 1980 Willow Smoke and Dog's Tails: Hunter-Gatherer Settlement Systems and Archaeological Site Formation. American Antiquity 45(1):4-20.

Brauner, David R. and Clayton G. Lebow 1983 A Reevaluation of Cultural Resources Within the Proposed Elk Creek Lake Project Area, Jackson County, Oregon. Phase II: Site Evaluation. Report by the Department of Anthropology, Oregon State University, Corvallis. Submitted to the U.S. Army Corps of Engineers, Portland District. Byram, Scott 2002 Brush Fences and Basket Traps: The Archaeology and Ethnohistory of Tidewater Weir Fishing on the Oregon Coast. Unpublished Ph.D. Dissertation, Department of Anthropology, University of Oregon. Byram, Scott and Thomas J. Connolly 1995 Newberry Crater Debitage Analysis. In Human and Environmental Holocene Chronology in Newberry Crater, Central Oregon, by Thomas J. Connolly, pp. 197232. State Museum of Anthropology, University of Oregon. Draft report submitted to the Oregon Department of Transportation. Bryam, Scott, and Don Ivy 2001 The Upper Coquille River Archaeological Survey 2000-2001. Coquille Indian Tribe Cultural Resources Program. North Bend, Oregon. Chartkoff, Joseph L. and Kerry K. Chartkoff 1984 The Archaeology of California. Stanford University Press, Stanford. Connolly, Thomas J., Jane E. Benjamin, Brian L. O'Neill, and Dennis L. Jenkins 1994 Archaeological Investigations at Two Sites on the Upper Rogue River (35JA189 and 35JA190), Southwest Oregon. University of Oregon Anthropological Papers No. 48. Department of Anthropology and Oregon State Museum of Anthropology, University of Oregon, Eugene. Connolly, Thomas J. 1991 The Standley Site (35DO182): Investigations into the Prehistory of Camas Valley, Southwest Oregon. University of Oregon Anthropological Papers No. 43, Department of Anthropology and Oregon State Museum of Anthropology, University of Oregon, Eugene 1989 1986 Points, Patterns, and Prehistory. Table Rock Sentinel 9(5):2-11. Cultural Stability and Change in the Prehistory of Southwest Oregon and Northern California. Unpublished Ph.D. Dissertation, Department of Anthropology, University of Oregon, Eugene, Oregon.

Cressman, Luther S. 1933 Aboriginal Buriels in Southwestern Oregon. American Anthropologist 35:11630.

94

Davies, D.G., editor 1961 Peter Skene Ogden's Snake Country Journal.: 1826-1827. The Hudson's Bay Record Society, London. Dorsey, James O. 1890 The Gentile System of the Siletz Tribes. Journal of American Folk-Lore 3(10):227-237. Douthit, Nathan 1999 A Guide to South Coast History: Traveling the Jedediah Smith Trail. Oregon State University Press, Corvallis. Draper, John A. 1988 A Proposed Model of Late Prehistoric Settlement Systems on the Southern Northwest Coast, Coos and Curry Counties, Oregon. Unpublished Ph.D. Dissertation, Washington State University, Pullman, Washington. Drucker, Philip 1937 The Tolowa and Their Southwest Oregon Kin. University of California Publications in American Archaeology and Ethnology 36:221-300. Erlandson, Jon M., Mark A. Tveskov, Madonna L. Moss, and George B. Wasson 2000 Riverine Erosion and Oregon Coast Archaeology: A Pistol River Case Study. In Changing Landscapes: Proceedings of the 3rd Annual Coquille Cultural Preservation Conference, 1999 edited by Robert J. Losey, pp. 3-18. The Coquille Indian Tribe, North Bend, Oregon. Erlandson, Jon M. and Madonna L. Moss 1999 The Systematic Use of Radiocarbon Dating in Archaeological Surveys in Coastal and Other Erosional Environments. American Antiquity 64(3):431443. Frederickson, David A. 1973 Early Cultures of the North Coast Ranges, California. Unpublished Ph.D. dissertation, University of California, Davis. Hall, Roberta L. 1992 Language and Cultural Affiliation of Natives Residing Near the Mouth of the Coquille River before 1851. Journal of Anthropological Research 48:165-184. Hannon, Nan 1990 Hand to Mouth: Plant Food Resources and Prehistoric People in Southwest Oregon's Bear Creek Valley. Unpublished M.S. Paper, Southern Oregon State College, Ashland, Oregon.

95

Harrington, John P. 1942 Alsea, Siuslaw, Coos, Southwest Oregon Athapaskan: Vocabularies, Linguistic Notes, Ethnographic and Historical notes. John Peabody Harrington Papers, Alaska/Northwest Coast, in National Anthropological Archives, Smithsonian Institution, Washington. Hayden, Brian 1995 Pathways to Power: Principles for Creating Socioeconomic Inequalities. In Foundations of Social Inequality, edited by T.D. Price and G.M. Feinman, pp. 1586. Plenum Press, New York. Hesse, Brian and Paula Wapnish 1985 Animal Bone Archaeology: From Objectives to Analysis. Taraxacum, Washington. Hoijer, Harry 1960 Athapaskan Languages of the Pacific Coast. In Cultures in History: Essays in Honor of Paul Radin, edited by Standley Diamond, pp. 960-976. Columbia University, New York. 1956 The Chronology of Athapaskan Languages. International Journal of American Linguistics 22(4):219-232.

Jacobs, Elizabeth 1933-34 Upper Coquille Athapaskan Linguistic and Ethnographic Data, Melville Jacobs Collection, University of Washington Libraries, Seattle. Jacobs, Melville 1940 Coos Myth Texts. University of Washington Publications in Anthropology 8(2):127259. 1939 Coos Narrative and Ethnologic Texts. University of Washington Publications in Anthropology 8(1):1-125. Hanis and Miluk Coosan Texts and Linguistic and Ethnographic Data, Melville Jacobs Collection, University of Washington Libraries, Seattle.

1931-34

Lewis, David G. 2001 Tolowa Deeni Fish Camp Ethnographies. In Changing Landscapes: Proceedings of the 4th Annual Coquille Cultural Preservation Conference, 2000, edited by Jason Yonker and Mark A. Tveskov, pp. 93-104. The Coquille Indian Tribe, North Bend, Oregon.

96

Lightfoot, Kent G., Antoinette Martinez, and Ann M. Schiff 1998 Daily Practice and Material Culture in Pluralistic Social Settings: An archaeological Study of Culture Change and Persistence from Fort Ross, California. American Antiquity 63(2):199-222. Lyman, R. Lee 1991 Prehistory of the Oregon Coast. Academic Press, San Diego. Maser, Chris, Bruce R. Mate, Jerry F. Franklin, and C.T. Dyrness 1984 Natural History of Oregon Coast Mammals. Museum of Natural History, University of Oregon, Eugene. Matson, R.G. and Gary Coupland 1995 The Prehistory of the Northwest Coast. Academic Press, San Diego. Miller, Jay and William R Seaburg 1990 Athapaskans of Southwestern Oregon. In Handbook of North American Indians, vol. 7, Northwest Coast, edited by Wayne Suttles, pp. 580-588. Smithsonian Institution, Washington, D.C. Musil, Robert R. and Rick Minor 1995 The Bravo Creek Site: A Lithic Workshop in the Klamath Mountain Foothills, Curry County, Oregon. Cultural Resources Series 10. Bureau of Land Management, Portland. Newman, Thomas M. 1959 Tillamook Prehistory and its Relation to the Northwest Coast Culture Area. Unpublished Ph.D. Dissertation, Department of Anthropology, University of Oregon, Eugene. O'Neill, Brian L. 1989 Archaeological Investigations at the Narrows and Martin Creek Sites, Douglas County, Oregon. Cultural Resource Series No. 4. U.S. Department of Interior, Bureau of Land Management, Portland, Oregon Olsen, Stanley J. 1968 Fish, Amphibian and Reptile Remains from Archaeological Sites. Papers of the Peabody Museum of Archaeology and Ethnology 56(2). 1964 Mammal Remains from Archaeological Sites. Papers of the Peabody Museum of Archaeology and Ethnology 56(1).

Orr, Elizabeth, William Orr, and Ewart M. Baldwin 1992 Geology of Oregon, 4th Edition. Kendall/Hunt Publishing Company, Dubuque, Iowa.

97

Patterson, L.W. and J.B. Sollberger 1978 Replication and Classification of Small Size Lithic Debitage. Plains Anthropologist 23:103-112. Personius, Stephen F. 1993 Age and Origin of Fluvial Terraces in the Central Coast Range, Western Oregon. U.S. Geological Survey Bulletin 2038. United States Government Printing Office, Washington D.C. Peterson, Emil R. and Alfred Powers 1952 A Century of Coos and Curry: History of Southwest Oregon. Coos-Curry Pioneer Association, Coquille. Pettigrew, Richard and Clayton G. Lebow 1987 Data Recovery at Sites 35JA27, 35JA59, and 35JA100, Elk Creek Lake Project, Jackson County, Oregon: Report by INFOTEC Research Inc., Eugene, Oregon, for the U.S. Army Corps of Engineers, Portland, Oregon. Pullen, Reginald J. 1981 The Identification of Early Prehistoric Settlement Patterns along the Coast of Southwest Oregon: A Survey Based Upon Amateur Collections. Unpublished M.A. Thesis in Interdisciplinary Studies, Oregon State University, Corvallis. Raymond, Anan 1989 Flaked Stone Technology at the East Bug-A-Boo Site, Linn County, Oregon. In Contributions to the Archaeology of Oregon, 1987-1988, edited by Rick Minor, pp.77-112. Association of Oregon Archaeologists Occasional Papers No. 4, Portland, Oregon. Robbins, William G. 1988 Hard Times in Paradise: Coos Bay, Oregon, 1850-1986. University of Washington Press, Seattle and London. Stahle, David W. and James E. Dunn 1982 An Analysis and Application of the Size Distribution of Waste Flakes from the Manufacture of Bifacial Stone Tools. World Archaeology 14(1):84-97. Stanton, Lt. Henry W. 1852 Headquarters Records, Selected Letters and Telegrams Received 1849-1861, Records of the Division and Department of the Pacific, Record Group 393 Records of the U.S. Army Continental Commands 1820-1920. Microfilm 284-8, SWORP, Knight Library, University of Oregon, Eugene. Steele, Harvey 1984 The Marthhaller Site. Report No. 9, Oregon Archaeological Society. Report on file at the Medford District of the Bureau of Land Management, Medford, Oregon.

98

Stewart, Hilary 1994 Stone, Bone, Antler, and Shell: Artifacts of the Northwest Coast. University of Washington Press, Seattle. Sullivan, Alan P. and Kenneth C. Rozen 1985 Debitage Analysis and Archaeological Interpretation. American Antiquity 50(4):755-779. Swezey, Sean L. and Robert F. Heizer 1993 Ritual Management of Salmonid Fish Resources in California. In Before the Wilderness, edited by Thomas C. Blackburn and Kat Anderson, pp.299-328. Ballena Press, Menlo Park, California. T'Vault, William G. 1851 Massacre of a Port Orford Party. Alta Californian, October 14, 1851. Tveskov, Mark A. 2003 Household, Landscape, and Persistent Places: The Pacific Coast Athapaskans. In

Proceeding of the 6th Biennial Rocky Mountain Anthropological Conference. Edited by R.G. Matson and M. Magne.

2002

The Cultural Geography of the Coos and Coquille. In Changing Landscapes: Proceedings of the 5th and 6th Annual Coquille Cultural Preservation Conferences, edited by Donald B. Ivy and R. Scott Byram, pp. 139-160. The Coquille Indian Tribe, North Bend, Oregon. Archaeological Survey of the South Fork Coquille River. SOULA Research Report 2001-2. Southern Oregon University, Department of Sociology and Anthropology, Ashland, Oregon. The Bandon Sandspit Site: The Archaeology of a Proto-Historic Coquille Indian Village. In Changing Landscapes: Proceedings of the 3rd Annual Coquille Cultural Preservation Conference, 1999, edited by Robert J. Losey, pp. 43-59. The Coquille Indian Tribe, North Bend, Oregon.

2001

2000

Tveskov, Mark A. and Jon M. Erlandson n.d. The Haynes Inlet Weirs: Estuarine Fishing and Archaeological Site Visibility on the Southern Cascadia Coast. Journal of Archaeological Science. In Press. Tveskov, Mark A., Jon M. Erlandson, and Madonna Moss 1996 Archaeological Investigations at the Coquille Point Site (35CS136), Coos County, Oregon. Coastal Prehistory Program, Department of Anthropology, University of Oregon, Eugene.

99

Tveskov, Mark A., Nicole Norris, and Amy Sobiech 2003 The Windom Site: A Persistent Place in the Cascade Foothills of Southwest Oregon. Report to the Bureau of Land Management, Medford District. Southern Oregon University Laboratory Anthropology Research Report 2002-1, Ashland, Oregon. Wasson, George B. Jr. 2001 Growing Up Indian: An Emic Perspective. Unpublished Ph.D. Dissertation, Department of Anthropology, University of Oregon. 1994 The Coquille Indians and the Cultural "Black Hole"of the Southwest Oregon Coast. Unpublished M.S. Paper, Department of Anthropology, University of Oregon, Eugene.

Waterman, Thomas T. 1925 The Village Sites in Tolowa and Neighboring Areas in Northwestern California. American Anthropologist 27(4):528-543. Williams, Loren L. n.d. Captain L.L. Williams and the Exploring Expedition of 1851. In Historical Sketches of Oregon's Southern Coast. Ms. On file, OSMA Anthropology, University of Oregon, Eugene. Wilson, Bart 1979 Excavations at the Ritsch Site. Master's thesis, Oregon State University, Corvallis, Oregon.

Winthrop, Kathryn 1993 Prehistoric Settlement Pattern in Southwest Oregon. Ph.D. dissertation, Department of Anthropology, University of Oregon, Eugene, Oregon. Wooldridge, Alice H., editor 1971 Pioneers and Incidents of the Upper Coquille Valley: 1890-1940. The Mail Printers, Myrtle Creek, Oregon. Younker, Jason 2003 Coquille/K´kwel, a Southwest Oregon Indiand Tribe: Understanding History, Culture, and Context in Revitalization. Unpublished Ph.D. Dissertation, Department of Anthropology, University of Oregon.

100

Information

112 pages

Report File (DMCA)

Our content is added by our users. We aim to remove reported files within 1 working day. Please use this link to notify us:

Report this file as copyright or inappropriate

1250481


You might also be interested in

BETA
Microsoft Word - SFC_MainRept_Part1.doc