From the Arctic to Avalon: Papers in Honour of Jim Tuck: Proceedings of the conference "From the Arctic to Avalon: Transforming the History of Northeastern North America", St. John's, Newfoundland, October 14-16 2004 9781841717432, 9781407329604

Table of contents :
Front Cover
Copyright
Dedication
Table of Contents
Foreword
Abstracts
Inukshuk: Caribou Drive Lanes on Southern Victoria Island, Nunavut, Canada
Defense in an Iroquois Village
But Once the Twain Did Meet: A Speculation About Iroquois Origins
A Labrador Archaic Longhouse Site at Sandy Cove, Labrador: Life on the South Side of Groswater Bay
"Their House is the Best I Have Seen on the Labrador": A Nineteenth-Century Jersey Dwelling at L'Anse au Cotard
Cultures, Borders, and Basques: Archaeological Surveys on Quebec's Lower North Shore
A Microscopic Approach to Palaeoeskimo Plant Use
Public Archaeology and the Public Good
Looking in the Right Places: Maritime Adaptations in Northeastern North America and the Central Andes
The Tuckerton Shell Mound: A Lasting Fertilizer
St. Lawrence Iroquoians in Northern New England: Intruders in the "Dawnland"?
"Maryland:'...Where the Winters be Shorter and Less Rigorous'."
Aviation Archaeology of Pre-Confederation Newfoundland
Archaeological Investigations in Renews; Newfoundland and Labrador
The Kirke House at Ferryland: Gentry Life in Seventeenth-Century Newfoundland
References Cited

Citation preview

From the Arctic to Avalon Papers in Honour of Jim Tuck

Edited by

Lisa Rankin & Peter Ramsden Proceedings of the conference: "From the Arctic to Avalon: Transforming the History of Northeastern North America"

St. John's, Newfoundland, October 14-16 2004

BAR International Series 1507 2006

Published in 2019 by BAR Publishing, Oxford BAR International Series 1507 From the Arctic to Avalon © The editors and contributors severally and the Publisher 2006 The authors’ moral rights under the 1988 UK Copyright, Designs and Patents Act are hereby expressly asserted. All rights reserved. No part of this work may be copied, reproduced, stored, sold, distributed, scanned, saved in any form of digital format or transmitted in any form digitally, without the written permission of the Publisher. ISBN 9781841717432 paperback ISBN 9781407329604 e-book DOI https://doi.org/10.30861/9781841717432 A catalogue record for this book is available from the British Library This book is available at www.barpublishing.com BAR Publishing is the trading name of British Archaeological Reports (Oxford) Ltd. British Archaeological Reports was first incorporated in 1974 to publish the BAR Series, International and British. In 1992 Hadrian Books Ltd became part of the BAR group. This volume was originally published by John and Erica Hedges in conjunction with British Archaeological Reports (Oxford) Ltd / Hadrian Books Ltd, the Series principal publisher, in 2006. This present volume is published by BAR Publishing, 2019.

BAR

PUBLISHING BAR titles are available from:

E MAIL P HONE F AX

BAR Publishing 122 Banbury Rd, Oxford, OX2 7BP, UK [email protected] +44 (0)1865 310431 +44 (0)1865 316916 www.barpublishing.com

To the Memory of JIM PETERSEN 1954-2005

iii

iv

TABLE OF CONTENTS

DEDICATION ...............................................................................................................................................................

iii

FOREWORD ................................................................................................................................................................

vii

ABSTRACTS .......................................................................................................................

.......................................... ix

INUKSHUK: CARIBOU DRIVE LANES ON SOUTHERN VICTORIA ISLAND, NUNAVUT. JACK BRINK ..................................................................................................................................................................

1

DEFENSE IN AN IROQUOIS VILLAGE. WILLIAM ENGELBRECHT ........................................................................................................................................

19

BUT ONCE THE TWAIN DID MEET: A SPECULATION ABOUT IROQUOIS ORIGINS. PETER RAMSDEN ............................................................................................................... ........................................

27

A LABRADOR ARCHAIC LONGHOUSE SITE AT SANDY COVE, LABRADOR: LIFE ON THE SOUTH SIDE OF GROSWATER BAY. LISA RANKIN ....................................................................................................................... "THEIR HOUSE IS THE BEST I HA VE SEEN ON THE LABRADOR": A NINETEENTH-CENTURY JERSEY DWELLING AT L'ANSE AU COTARD BLAIR TEMPLE .................................................................................................................

....................................... 33

..................................... ..... 43

BORDERS, BIFACES AND BASQUES: ARCHAEOLOGICAL SURVEYS ON QUEBEC'S LOWER NORTH SHORE. WILLIAM W. FITZHUGH ...........................................................................................................................................

53

A MICROSCOPIC APPROACH TO PALAEOESKIMO PLANT USE. LATONIA HARTERY ...................................................................... ............................................................................

71

PUBLIC ARCHAEOLOGY AND THE PUBLIC GOOD W. J. BYRNE .................................................................................................................................................................

79

LOOKING IN THE RIGHT PLACES: MARITIME ADAPT ATIO NS IN NORTHEASTERN NORTH AMERICA AND THE CENTRAL ANDES. JAMES L. RI CHA.RD SON III ...................................................... .................... ............................ ................................. 83 THE TUCKERTON SHELL MOUND: A LASTING FERTILIZER R. ALAN MOUNIER .................................................................................................................................................... ST. LAWRENCE IROQUOIANS IN NORTHERN NEW ENGLAND: INVADERS IN THE DA WNLAND? JAMES B. PETERSEN .............................................................................................................

99

................................... 109

MARYLAND: WHERE THE WINTERS BE SHORTER AND LESS RIGOROUS SILAS HURRY ...........................................................................................................................................................

129

AVIA TI ON ARCHAEOLOGY IN PRE-CONFEDERATION NEWFOUNDLAND MICHAEL DEAL ...................................................... ................................................ ..................................................

139

ARCHAEOLOGICAL INVESTIGATIONS IN RENEWS, NEWFOUNDLAND AND LABRADOR STEPHEN MILLS .......................................................................................................................................................

157

THE KIRKE HOUSE AT FERRYLAND: GENTRY LIFE IN SEVENTEENTH-CENTURY NEWFOUNDLAND BARRY GAULTON ....................................................................................................................................................

165

V

vi

FOREWORD This volume is the result of a conference held in St. John's in October 2004 to mark the retirement of Jim Tuck from the Archaeology Unit at Memorial University. The conference was entitled "From the Arctic to Avalon: Transforming the History of Northeastern North America", and the wide geographical and topical scope of the papers reflected Jim's many and diverse contributions to the archaeology of eastern North America, from the Arctic to the Great Lakes, from the Archaic to European colonization. The conference was generously supported by the Social Sciences and Humanities Research Council, the Atlantic Canada Opportunities Agency, The Institute for Social and Economic Research, The office of the Dean of Arts of Memorial University and the Provincial Archaeology Office of Newfoundland and Labrador. Not all the papers presented at the conference were submitted for publication. In some of these cases, the highly pictorial nature of the presentation made it unsuitable for transformation into a printed paper. On the other hand, a couple of papers included in this volume were not able to be presented at the conference for one reason or another. Nevertheless, the papers included here faithfully reflect the range of topics dealt with at the two-day conference in St. John's, and certainly reflect the variety and significance of Jim Tuck's contributions to understanding human history in northeastem North America, as well as the tremendous respect that Jim's friends and colleagues have for both him and his work. The editors would like to thank all of those who attended the conference, both presenters and others, for making it such an enjoyable and stimulating experience. We are grateful to all the volunteers who made the conference possible in a variety of ways, particularly MUNARCH (the Memorial University student archaeology society), and the graduate students in the Archaeology Unit at Memorial University. We also thank those who contributed papers to this volume for their timely submissions and their tolerance of our editorial demands, and those who generously agreed to review the papers. We are humbly grateful to John Hedges of BAR, first, for agreeing to publish this volume, and equally for his patience, help and advice while we were preparing the manuscript for submission. The one sad note to this whole process was the tragic and untimely death of Jim Petersen in August, 2005 in Brazil. Jim was a participant in the conference, and a good colleague and close friend of many of those who attended. We received his paper for this volume soon after the conference, and we are honoured to be able to include it. Lisa Rankin Peter Ramsden St. John's, March, 2006

vii

viii

ABSTRACTS Jack W. Brink

Inukshuk: Caribou Drive Lanes on Southern Victoria Island, Nunavut, Canada. Caribou drive systems made of stone lines and cairns [inuksuit] are a common feature of the far north but have been little studied by archaeologists. Two communal caribou kill sites from southern Victoria Island, Nunavut, Canada are discussed and illustrated. The Eggington site is a single-line drive where herds of caribou were directed through a saddle between two hills and killed from shooting pits. The POD site is a V-shaped funnel with two prominent lines of cairns and stone walls ending with opposing shooting pits. The sites, of uncertain age, are similar to those described by Jenness for the historic Caribou Inuit. Caribou drives, though often devoid of artifacts, have the power to reveal the sophisticated systems of knowledge that enabled successful communal kills. W.J. Byrne

Public Archaeology and The Public Good. There remains today a serious disconnect between the discipline of archaeology in North America and the public. This disconnect is most evident in the consistent denial by aboriginal North Americans of many of the key findings of archaeology, and at times their active opposition both to the practice of the discipline and its practitioners. Compounding the problem is the fact that while non-aboriginal members of the public are generally interested in the findings of archaeology, there is a strong proclivity to accept aboriginal, alternative interpretations of many of the findings of scientific archaeology rather than the archaeologists' own views. Unless serious efforts are made to fully engage both the aboriginal and non-aboriginal public in the discipline, there is a real danger of archaeology becoming irrelevant before long. Michael Deal

Aviation Archaeology of Pre-Confederation Newfoundland. Pre-Confederation Newfoundland played a prominent role in the history of aviation as a nexus in the advancement of transatlantic flight and the ferrying of military aircraft to Europe during World War II. This paper considers the potential for aviation archaeology in Newfoundland and Labrador, in terms of the surviving pre-Confederation aviation infrastructure, significant aircraft crash sites, and current government policy relevant to aviation heritage resources. Emphasis is placed here on the responsibilities of the government and archaeologists for the protection and recovery of these resources. William Engelbrecht

Defense in an Iroquois Village. Iroquois oral tradition states that fighting was once common. Since fighting involved social and political units of variable scale, it is not surprising that an Iroquois concern with defense can be inferred archaeologically at both the regional and village level. Regionally, the development of Iroquois nations is reflected in the location of contemporaneous villages near one another. At the village level, the development of large palisaded hilltop communities typical of the sixteenth century Iroquois is commonly interpreted as a defensive response to endemic warfare. This paper explores the possibility that features within the village also reflect a concern with defense. William W. Fitzhugh

Cultures, Borders, and Basques: Archaeological Surveys on Quebec's Lower North Shore. Research in Labrador since 1968 has been concerned with developing a culture history and environmental framework of sufficient detail and geographic scope that processes of ethnic and cultural boundary change and relationships with environment through time could be revealed. This paper expands the scope of that inquiry into the Quebec Lower North Shore where surveys and excavations since 2001 have begun to produce data that can be related to previous research in Labrador, Newfoundland, and the Strait of Belle Isle. Goals of the St. Lawrence Gateways Project are outlined; site survey data are presented; and results are advanced for studies of cultural boundaries and culture change. Particular attention is given to Maritime Archaic, Paleoeskimo, and Neoeskimo cultures, and preliminary reporting on a Basque site at Petit Mecatina. Barry Gaulton

The Kirke House at Ferryland: Gentry Life in Seventeenth-Century Newfoundland. Sir David Kirke and his family were amongst the most successful merchants and entrepreneurs in seventeenth-century Newfoundland. Excavation of the Kirke house (ca. 1640s) at Ferryland revealed a large timber-framed dwelling, several outbuildings and tens of thousands of associated artifacts. The architectural features and material culture reveal the daily life of this gentry family, and show how it differed from that of the more numerous planter families that shared this colonial settlement. ix

Latonia Hartery

A Microscopic Approach to Palaeoeskimo Plant Use. Palaeoeskimo plant use is often considered unworthy of archaeological investigation. Poor macrobotanical preservation and the opinion that Arctic-adapted cultures infrequently used plants contribute substantially to this lack of inquiry. When research designs do include ethnobotany, the standard technique involves separating macrobotanical remains from soil by flotation. Unless a site has excellent preservation, these experiments often yield poor results. An alternative strategy is phytolith research. This microscopic approach was applied to the Middle Dorset Paleaoeskimo site of Peat Garden North, Bird Cove, NL to reveal evidence about plants and their relationship to tool function. Silas Hurry

Maryland: ' ... Where the Winters be Shorter and Less Rigorous.' After a winter in Newfoundland, George Calvert, the first Lord Baltimore, wrote to Charles I that he wished to seek a new venue for his colonial endeavors. A mere five years later George was dead and his son Cecil was Lord Baltimore and Lord Proprietor of Maryland. St. Mary's City was founded as the capital of the Calvert colony of Maryland and remained so until late in the 17th century when the Calvert family lost control of the colony and the capital was moved to Annapolis. For more that thirty years, Historic St. Mary's City has investigated the archaeological remains of this second Calvert colony and worked to interpret its story to the visiting public. This essay will review some of the findings and how they speak to the Calvert vision of the New World. Steven Mills

Archaeological Investigations in Renews, Newfoundland and Labrador. Renews is a small harbour on the southeast coast of Newfoundland's Avalon Peninsula. Since the early sixteenth century explorers have visited it, fishermen have worked and lived in it, propagandists have written about it and soldiers have fought over it. Since the 1990s archaeological investigations have uncovered artifacts, houses and features dating from the sixteenth century onwards. Centuries old cultural landscape features on the shores and in the meadows that surround the harbour have also been documented. This research illuminates the simple, yet cosmopolitan, lives of the fishers, settlers and soldiers who once visited and lived on the eastern edge of the New World. R. Alan Mounier

The Tuckerton Shell Mound: a Lasting Fertilizer. The Tuckerton Shell Mound is the most imposing of the few remaining aboriginal shell mounds on the Atlantic coast of New Jersey. Consisting almost entirely of quahog (hard-shell clam) shells, the mound lies in the midst of an extensive tidal meadow. As early as 1868 the mound was recognized as the product of aboriginal shellfishing. At that time, the ancient shells were reported to make "a lasting fertilizer" for agricultural purposes. Subsequently, the mound was mined as a source of lime, and its shells were used for fill as well as for paving material. Archaeological investigation of the mound has gone on since the early 20 th century, most recently by the Archaeological Society of New Jersey. The age of the mound is indicated by two carbon dates of ca. 1500 B.P. and 1940±90 B.P. Investigation of the shell deposits, surrounding marsh sediments and underlying geological strata should illuminate the processes involving the origin and development of the mound through time. Although the shell mound has great explanatory potential from a theoretical perspective---for example, from the standpoint of landscape or gender issues- the present work is essentially empirical. .Research at the Tuckerton Shell Mound may prove to be a lasting fertilizer for nurturing knowledge of human interactions with the ancient environment. James B. Petersen

St. Lawrence Iroquoians in Northern New England: Invaders in the Dawnland? One of the most remarkable developments in northeastern North America during the long span of Native American occupation was the appearance of Iroquoian-speaking peoples, along side non-Iroquoian Algonquian speakers. One major Iroquoian group, the St. Lawrence Iroquoians, became archaeologically recognizable some time after ca. A.D. 1200-1300, primarily along the Great Lakes and St. Lawrence River in Ontario, Quebec, and northern New York. Recent research establishes that the St. Lawrence Iroquoians were also present in various areas of northern New England over a substantial period, minimally ca. A.D. 1300-1580 or later, especially in western Vermont. This paper reviews the available evidence from Maine, New Hampshire, and Vermont, including pottery find spots and recently studied habitation sites. The significance of early discoveries , a subsequent hypothesis about a Champlain-Richelieu "cluster" for the St. Lawrence Iroquoians, and recent discoveries are discussed, along with different mechanisms that help account for the cumulative evidence. This research demonstrates the vitality of current Iroquoian archaeology within broader northeastern contexts and points toward avenues for future exploration.

X

Peter Ramsden

But Once the Twain Did Meet: A Speculation about Iroquois Origins. The longest-standing puzzle in Iroquoian studies has been the apparent contradiction between the obvious archaeological fact that Iroquois culture developed in situ in the lower Great Lakes, and the linguistic fact that the Iroquois seem to be relatively recent arrivals in a vast region of Algonkian speakers. This paper attempts a reconciliation by suggesting, in essence, that it was Iroquois language that arrived recently in the lower Great Lakes, not Iroquois people. Possible reasons for and mechanisms of the adoption of Iroquois language by groups of Algonkian speakers are offered. Lisa Rankin

A Late Archaic Longhouse Site at Sandy Cove, Labrador: Life on the South Side of Groswater Bay. In 2002 and 2003 the Porcupine Archaeology Project collected and excavated portions of a nearly 6000 year old Labrador Archaic site on the southern shore of Groswater Bay. This paper describes and examines this Sandy Cove complex site and compares it to similar sites in central Labrador. While in many ways the site can be considered typical of the complex, the presence of three boulder-ringed longhouses is unique and marks the earliest appearance of these structures in the region. James B. Richardson III

Looking in the Right Places: Maritime Adaptations in Northeastern North America and the Central Andes. In this paper I will compare and contrast maritime adaptations of Northeast North America with that of Peru. This discussion will focus on the pre-5,000 BP evidence for the use of marine resources in both regions which will include the changing marine environments, the archaeological evidence for early coastal and drowned sites, and the rise of cultural complexity in the Preceramic.

Blair Temple

"Their House is the Best I Have Seen on the Labrador": A Nineteenth-Century Jersey Dwelling in L'Anse au Cotard. Partial excavation of an abandoned community in southern Labrador have uncovered a nineteenth-century structure containing a large quantity and variety of material culture. The community was settled by fishers from Jersey during the early part of the nineteenth-century , and was abandoned during the later part of the century. Preliminary analysis of the artifact assemblage indicates that the structure was a dwelling with an attached pantry and storage room. The inhabitants were engaged primarily in the fishery as a livelihood, but also engaged in sealing.

xi

xii

INUKSHUK: CARIBOU DRIVE LANES ON SOUTHERN VICTORIA ISLAND, NUNAVUT, CANADA Jack W. Brink Provincial Museum of Alberta

The importance of caribou (Rangifer tarandus) to many Aboriginal groups of the far north can hardly be overstated. Birket-Smith (1929:47) referred to caribou as the "axis on which everything turns in the existence of the Barren Ground Eskimo." For the Caribou Inuit of the central Arctic, Steenhoven (1962:25) said, "it is fair to say that these Eskimos wake up and go to sleep with the word tuktoo [caribou] on their minds ... [it is] the very pulse of their lives ... tuktoo to them is not just game, but it is the focus of their cultural existence." In addition, Burch (1972:339) stated that caribou might be the species of single greatest importance in the anthropological literature of hunting societies. Caribou provided many of the basic necessities of life, including food, clothing, shelter and materials for tools, and was the source of much spiritual belief and ceremony (Arima 1984; Spiess 1979). It is not surprising that northern peoples developed a wide array of ingenious and sophisticated methods to hunt the great herds of caribou. Perhaps the most compelling of these methods are the communal drives that included elaborate stone structures built on the landscape.

1981 and Taylor 1972 being notable exceptions). The reasons for this paucity of study have been stated by Fitzhugh ( 1981:188) and include: caribou drive sites are difficult to find, hard or impossible to date, and difficult to interpret; the sites are of indeterminate cultural origin; and archaeologists have failed to consult with Native elders and to spend a significant amount of time in the regions where these sites are found. To this can be added that caribou drive sites generally lack any significant amount of cultural material. As a result, elaborate caribou drive structures have been repeatedly noted in the archaeological literature for the north but have seldom been the target of specific investigation. This paper describes two caribou drive structures of southern Victoria Island, Nunavut, Canada, and examines how Aboriginal knowledge of caribou allowed for the manipulation and harvesting of large numbers of animals.

BACKGROUND AND SETTING

Victoria Island is a large island in the west-central Canadian Arctic (Figure 1). The eastern halfofthe island lies in the territory of Nunavut. The southern shore of the island--of concern here-is separated from the mainland of Nunavut by Dolphin and Union Strait in the southwest, Coronation Gulf to the south, and Dease Strait and Victoria Strait to the southeast. The environment of the region is one of open, treeless arctic tundra; permafrost country dotted with countless small ponds, rivers, and many lakes (Jenness 1922:13-27). Damas (1972:7), perhaps a little too harshly, described much of Victoria Island as a "virtual wasteland;" but it is true to say that

Some researchers have argued that communal hunting of caribou was more productive than solitary hunting (Blehr 1990:321-322), which was considered successful if a single animal was killed (Balikci 1970:41). As Rasmussen (1927:73) noted when discussing communal caribou hunts, the "ingenious methods ... give so rich a yield as to cover also the dead seasons when no game was to be had ... " The critical challenge in the vast, open tundra regions of the Arctic and Subarctic is to know where caribou will be located, and to be able to predict and control their path of flight when approached by human hunters (Blehr 1990:310). In the treeless Arctic, lines of stone structures were the solution to this problem, as their purpose was to make caribou flight predictable. Given the extraordinary environmental constraints of the far north, it is not surprising that such solutions were remarkably uniform across the Arctic. Caribou drive systems built of stone structures are known from across the Canadian Arctic, Alaska, and Greenland (Arima 1975; Balikci 1970; Birket-Smith 1929; Boas 1964; Fitzhugh 1981; Gordon 1990; Gubser 1965; Jenness 1922; Morrison 1981; see Spiess 1979 for an overview). Birket-Smith (1929:11) reports that these features "are spread all over the whole country, wherefore a record of them all is impossible." Despite their ubiquity, archaeological study of these simple-looking yet complex structures has been lacking. Most archaeologists have done little more than note their presence (Fitzhugh 1981, Gnmnow et al. 1983, Morrison

'

NOlm!Wl $T \ TERRITORY \

Figure 1. Lake.

1

NUNAVUT

Location of Victoria Island and Ferguson

FROM THE ARCTIC TO AV ALON: PAPERS IN HONOUR OF JIM TUCK

the central Arctic is one of the most challenging inhabited places on earth. Oceans, lakes, and rivers are frozen for eight to nine months of the year. Whales, the mainstay of many Inuit cultures, are absent from the seas of Coronation Gulf and Dease Strait (Taylor 1965: 12). Only the small ringed seal is found in abundance in the seas, while caribou, and to a much lesser extent musk ox, are the primary large land mammals (Morrison 1992: 15). Fish are abundant, but can only be taken in significant numbers during the short open water season (Damas 1984a). Vegetation is limited to reindeer moss and various Arctic flowers and grasses.

comprehensive ethnography of any Arctic people (Collins 1984; Damas 1984a). The sites discussed here lie to the north side of the western end of Ferguson Lake, a long, thin, inland lake which trends some 75 km east to west on southern Victoria Island, about 50 km north of the community of Cambridge Bay, Nunavut (Figure 1). The west end of Ferguson Lake ends about 2 km from the coast and is drained by the Ekalluk River; a short, fast river that empties into Wellington Bay on the north side of Dease Strait. Importantly, two hills lie just north of the end of the lake and the beginning of the Ekalluk River (Figure 2). These hills play an important role in the configuration of the two caribou drive sites. The tops and sides of the hills are rocky and barren, and are composed of a series of downward cascading beach ridges, created through wave action as the land has continuously risen through isostatic rebound since the time of deglaciation (Fyles 1963). The lowlands surrounding the hills, and the saddle between them, are vegetated with moss and grass, and are often wet just below the surface. Numerous small ponds occupy the lowlands.

Caribou have become rare on Victoria Island during the past half-century (Banfield 1954 Pt. 1:68-69) but the island was once the summer home to vast herds (Banfield 1954 Pt. 2:10-11; Stefansson 1914:39). Bathurst Inlet, just south of Victoria Island, was a major calving ground (Kelsall 1968: 110). Herds came north from the Coronation Gulf region, crossing Dease Strait and Dolphin and Union Strait in late spring before the ice melted (Kelsall 1968: 16). After summering on Victoria Island they re grouped on the southern shore until freezeup, and then crossed back to the mainland (Jenness 1922:125; Stefansson 1914:41). In 1910 and 1911 Stefansson (1951 :224) reports traveling with huge herds of caribou on Victoria Island, and that hundreds of thousands, possibly millions, were seen headed south from this island across frozen Dease Strait. A few years later Hoare (cited in Banfield 1954 Pt. I: 10) observed these same herds and estimated their numbers in the millions.

PREVIOUS RESEARCH

The first archaeological work in the Ferguson Lake region of Victoria Island was conducted in 1963 and 1965 by William Taylor (1967, 1972) who recorded and tested sites on the north and south sides of the west end of Ferguson Lake and along the Ekalluk River. Pre-Dorset, Dorset and Thule components were identified. Two sites

In historic times these vast herds were systematically hunted by the residents of southern Victoria Island, the Copper Inuit, whose territory extended onto the adjacent portions of the mainland in the vicinity of Coronation Gulf (Damas 1984a). The Copper Inuit are the westernmost group of the Central Inuit, an Arctic adapted people who subsisted almost exclusively on foods obtained from land and sea mammal hunting, fish, fowl and small mammals (Damas 1984b). How long these people and their ancestors have occupied the region is unclear, but McGhee (1972) argues that the Copper Inuit are direct descendents of the earlier Thule culture, which would place them in the region for at least the past 800 years. Owing to their central Arctic location, distant from both eastern and western entrances to the northern oceans, the Copper Inuit were among the last northern people to sustain frequent and prolonged contact with European culture (Damas 1984a). Indeed, so poorly known were the Copper Inuit that in 1888 Boas completely omitted them from his listing of Central Inuit groups (Damas 1984b). As a result, they continued to live a relatively traditional lifestyle into the early portions of the twentieth century. Fortunately, excellent ethnographic data exist for the Copper Inuit, primarily through the work of Diamond Jenness, who lived among them for two years ( 1914-1916) and whose account of their life (1922) is widely regarded as the finest and most

Figure 2. Topographic map of the Ferguson Lake region showing the two hills to the north side of the Ekalluk River, and the saddle between. Two sites reported here are in the centre, two reported by Taylor (1972) are near the bottom

2

BRINK: CARIBOU DRIVE LANES ON SOUTHERN VICTORIA ISLAND .

identified by Taylor on the north side of the Ekalluk River included extensive caribou game drive complexes (Figure 2). One, NiNg-17, was situated on generally level terrain near the coast of Wellington Bay and consisted of a series of simple stone cairns, often in multiple rows, extending in roughly a straight line for a distance of about 1.6 km. The other drive complex, NiNg-4, lies about 2 km to the southeast of the first, and consists of a crescent-shaped arrangement of stone cairns that conform to the contour of an elevated gravel ridge. This second drive is said to be about 1.2 km long and is situated to the south side of the southernmost of the two hills, with a low, wet area dotted with small ponds lying between the south side of the hill and the elevated ridge on which the stone structures occur (Taylor 1972:72-73). This is much like the setting of the two sites discussed here, which are situated just to the northeast side of the same hill, some 4 km to the northeast of the sites described by Taylor. The archaeological richness of the Ekalluk River region led Taylor to return in the summer of 1988. McGhee ( 1971, 1972) also carried out archaeological survey and excavation on western Victoria Island and recorded sites attributed to the Pre-Dorset, Dorset, Thule and Historic periods. At the Kunana site in Prince Albert Sound, some 300 km northwest of Ferguson Lake, McGhee (1972:71) reports a complex series of caribou drive lanes on raised beach ridges north and west of a historic period habitation site. That 96% of the more than 3,000 identifiable fauna! remains from the Kunana site were caribou suggests that the primary purpose of the site was for hunting and butchering migrating herds of caribou (McGhee 1972:71). McGhee (1972:100) believes that the Kunana site was occupied primarily during the nineteenth century, probably by the historic Copper Inuit.

Figure 3. Example of classic form of inukshuk with large basal stone, overhanging upper rocks weighted down and balanced stone on top. From the Eggington site. human beings. The basic type of inukshuk stone structure is generally similar in size, shape, and method of construction for both sites reported in this paper and consists of three or four rocks stacked on top of each other to form a small cairn some 30 to 40 cm high (Figure 3). Basal stones are often the largest in the cairn , with both large (20-30 cm) and small (5-10 cm) rocks added to the top. Often, rocks intentionally overhang the edge of the basal stone and are held in place with the weight of additional stones on top. This overhang gives the structure an appearance of having wings or arms. Note in Figure 3 how stones of specific size and shape were carefully selected to balance on the top of the inukshuk. While this describes a "classic" inukshuk, most structures are simply piles of a few rocks, or consist of nothing more than a single boulder propped up on its long axis. In some cases it could be determined that upper stones had fallen from the basal rock; but in others it appeared that no upper rocks had ever been added. Inuksuit at the two sites discussed here were also built using high points of local bedrock as the basal stone to which one or two rocks were added.

Taylor (1965: 17) also tested a site on western Victoria Island, at Lady Franklin Point, where 82% of the nearly 6,000 bones recovered were caribou. Taylor ( 1965: 17) concluded that the local Inuit cultures had successfully adapted to life without the typical Thule dependence on hunting of large sea mammals. Thus, caribou hunting using stone structures on southern Victoria Island can be traced from ethnographic times (Jenness 1922), through the historic period (McGhee 1972), back to at least the Thule culture (Taylor 1965). It may, of course, extend much further in time, as game drive structures built in a productive location may be used over a great time span by a succession of different cultures.

THE EGGINGTON SITE The Eggington site was the first site with stone features discovered, lying between the two hills on the north side of the west end of Ferguson Lake. This is a complex network of many lanes of stones and inuksuit, rock walls and shooting pits. While it seems likely that this structure represents a single caribou hunting system, as opposed to

The term "inukshuk" (plural "inuksuit" - both spelled in a variety of ways) is an Inuit word that translates as "acting like a man" or "acting like a person" (Hallendy 1994); the implication being that these structures were intended to resemble, and perhaps take the place of,

3

FROM THE ARCTIC TO AV ALON : PAPERS IN HONOUR OF JIM TUCK

upslope and others cross over level ground following the contour of beach ridges. Designating the upslope portion of the lane as the beginning assumes a downslope direction to caribou drives that cannot be proven. However, as discussed below, there are aspects of wind direction and caribou biology and behavior that would tend to make downhill drives more likely to succeed.

multiple, independent hunting components, this cannot be proven with the evidence at hand.

Location

Between the two hills to the northwest side of Ferguson Lake, at an elevation of about 80 m ASL, is a saddle containing one large lake and many small ponds (Figure 2). It is the north and east facing sides of the southernmost hill, and the saddle between the hills, that are of concern to the sites reported here. The Eggington site is situated in the saddle between the two hills to the east side of the large lake and small ponds. It begins at an elevation of about 80 m ASL and ends about 300 m from the northwestern shore of Ferguson Lake at an elevation of about 30 m ASL (Figures 2, 4). The total length of the stone lines of the Eggington drive is about 2,400 m, although undiscovered portions may exist at either the east or west ends.

The Eggington Drive System

Mapping and recording of the site was complicated by the convoluted nature of the drive system. A line of inuksuit splits into many lanes, parallel lanes rejoin, some lanes taper out while others cross each other, and sections of the stone walls are interspersed with individual cairns. Walking what appeared to be the main lane of inuksuit frequently resulted in spotting additional lanes located both upslope and downslope , leading to uncertainty as to which, if any, could be considered the primary lane. Undoubtedly some stone features went unrecorded. Tape measures , compass, and pacing were used to record the Eggington site; therefore , all site maps must be regarded as approximate.

The beach ridge gravels of the southern hill cascade downslope towards the northeast. As the steep slope of the hillside eases into the saddle, the ground covers become more mosaic; an interfingering of ridge gravels, vegetation consisting of Arctic grasses and moss, and exposed bedrock. Noteworthy is the fact that the stone lanes mostly follow areas of exposed bedrock and beach ridge gravels, seldom crossing vegetated areas. The west end of the lane begins on an outcrop of ridge gravels in the central part of the saddle, and the east end is near the lakeshore where grasses again appear. The general orientation of the entire lane system is WNW to ESE, though . there are divergences from this trend. Likewise, most of the Eggington lane system precedes downslope towards Ferguson Lake, though some sections angle

The long, complex stone alignment is conveniently described in western and eastern sections. Most of the western half of the Eggington site trends northwest to southeast and follows distinct, parallel beach ridges that are oriented roughly perpendicular to the lakeshore (Figure 4). Thus, the western portion of the lane traverses mostly level ground. The first evidence of stone features was found in the central part of the saddle between the two hills, to the east of a large pond where vegetation is rich. The system begins with three parallel lanes that traverse two ridges of beach gravels exposed near the bottom of the saddle (Figure 5). Rock cairns in these lanes were generally small, seldom consisting of more that two or three stacked rocks. Occasionally, small rocks were placed on in situ larger rocks to produce a higher cairn. The three lanes converged to one but soon split again into three or possibly four separate lines of inuksuit leading to a cluster of small shooting pits. Five to seven pits were noted, an exact count being hampered by the shallowness of the pits. These pits were typically 1.5 min diameter , and formed by a slight scooping out of beach ridge gravels and the construction of a small berm round the lip of the pit. The berm was often no more than a single line of stones and the total depth of the pits was little more than 15 to 20 cm. No pattern could be detected to the positioning of the shooting pits; they were placed inside and outside of the parallel lanes. Past the cluster of shooting pits the southeast-trending line, with numerous splits and side branches , continues several hundred metres along roughly parallel beach ridges. In places this portion of the lane becomes sinuous and crosses up or down to adjacent parallel beach ridges (Figure 6). The lane then rises up a slight slope crossing an elevated area of rugged gravels. This portion of the

Figure 4. Aerial photograph of the Eggington and POD caribou drive sites illustrating the location of the drives in the saddle situated between two hills on the north side of Ferguson Lake.

4

BRINK: CARIBOU DRIVE LANES ON SOUTHERN VICTORIA ISLAND.

Nothing remotely similar to this single inukshuk was observed anywhere in the west Ferguson Lake area.

,,f·> ~!/ ,l•

,-7.'~··..,I Cluster of Shooting Pits /

Located some 5 m from the large inukshuk was an exceptionally large stone feature interpreted as a hunting blind (Figure 7). This feature differed from shooting pit features in that it had only a negligible amount of excavation onto the basal gravels, but rather was constructed through placement of large boulders to form a crescent-shaped enclosure. In places the walls of this feature reached nearly 1 m in height, with the outside diameter being about 2 m. The closed, high part of the wall faced towards the northwest, away from Ferguson Lake and overlooking the saddle between the two hills. The open end of the crescent faced to the southeast. The rocks used in construction of this feature were exceptionally large for the area, much larger than those used in typical inukshuk construction. Rocks of this size were observed in the area sporadically basis; the concentration of large rocks used in this feature suggests transport from the surrounding region.

~/-

~_;

,/; Ii,7 : ., II y

I

, Large B Large lnukshuk

'\

··-

Large lnukshuk

Cairnline

Cluster ofTiny Shooting Pits

/\

J(

o Shooting pit

*

/

Western

Section

\

\

0

30m

30mGap to Eastern Section

Figure 5. Sketch map of the western portion of the Eggington game drive site.

lane is considered a critical part in the game drive system. Visibility is excellent in all directions. Stone features found here include one extraordinary inukshuk, one very large blind structure, a cluster of at least seven tiny pits and the single lane of small inukshuk cairns (Figure 5). The single extraordinary inukshuk (Figure 7) consists of one large, rectangular rock standing on end with a smaller, sub-rounded boulder placed on top. Together, these two rocks reach a height of 1.5 m and form a very human-like appearance. While use of naturally upright stones to construct inuksuit was a common occurrence in the west Ferguson Lake region, the huge bottom boulder used in this single inukshuk has clearly been purposely erected into this position and the capstone placed on top of it. Whether or not the massive bottom boulder had been moved to this location for this purpose is unknown.

Figure 7. Large shooting pit or hunting blind and single tall inukshuk along the western portion of the ERRinRtonsite.

A final discovery at this seemingly significant point in the drive was a cluster of at least seven tiny shooting pits. Exact counting of these pits was hindered by their extreme shallowness and hence minimal intrusion on the landscape. These pits were all crescent-shaped , with no more than 1/3 to 1/2 of a circle represented by the outer berm or rock. Unlike other pits noted in the region, these were conspicuous by their minimal construction. Each pit was 1.0 to 1.5 m in diameter and was formed by a very slight scooping out of beach ridge gravels combined with construction of a tiny berm at the lip of each pit. These berms consisted of no more than one or two small rocks placed around the edges of the depressions, forming a total depth to the pits of no more than 10 to 15 cm. These tiny pits were scattered within a 15 m radius of the tall inukshuk and large blind feature. The pits would have done little to hide a waiting hunter.

Figure 6. Sinuous portion of the Eggington drive lane, following and cross-cutting beach ridges in the upper, western section of the drive.

The single lane of cairns that passes through the position of the tall inukshuk and pit cluster continues on a WNWESE orientation for several hundred metres before 5

FROM THE ARCTIC TO AVALON: PAPERS IN HONOUR OF JIM TUCK

, --~ 30mGap ·• ' to Western Section 15mGap ,

. \ }

N

-----

,••\,o

(*Cache

~..'. Cluster o~ \~ 0 Shooting pits : '~

o Shooting pit

\

*

/

•···-

0:

:

Cache Cairn line Faint cairn line

••• -

"-J

...

o\

\ / WallSm '·\, o\ \

Double lane 0

4m X 15m

Figure 9. Eastern section of the Eggington lane where cairns are constructed on exposed bedrock. Note shooting pit in foreground and cairns silhouetted against sky in background.

40mGap

~1 \ Eastern Section

\o

···1 \0 o\

\ O

funnel (Figure 8). A cluster of at least six shooting pits was located near the apex of this funnel, and a few more along the wings of the funnel. At the end of the northern wing of the V was a large, deep, rock-walled structure believed to be a cache pit. It seems likely that these wing-like attachments to the main inukshuk lane were intended to direct caribou movement in much the same fashion as at the POD site discussed below.

Wall15m

30m

Raised dome o f beach gravels

\\, ,/

I

' ~

f.;tr i \'' ... ,

Downslope from the funnel the lane continued as one prominent row of inuksuit essentially until its end. Noteworthy features observed over this distance include two solid rock walls, one 5 m long, the second 15 m long (Figure 8). Both walls were about 40 cm high and were made of relatively small rocks piled three or four feet high. Why walls appeared in these tow sections of the lane is unknown, but they may indicate greater need to contain or control caribou. Inuksuit adjacent to the walls tended to be very close together, often no more than 0.5 m to 1 m apart. It may be that walls are formed slowly over time by gradually filling in the voids between very closely spaced cairns.

Figure 8. Sketch of the eastern portion of the E~~in~ton site. making a distinct turn to an easterly direction. This eastern portion of the Eggington site (Figure 8) is characterized by lanes of rock cairns that crosscut (rather than parallel) beach ridges, and that run primarily downslope (rather than horizontally). The inukshuk lane that turns east and runs downslope consists primarily of a single line of cairns, but with a number of conspicuous variations. As with the western section, lanes appear, merge and disappear; double lanes appear and then end, as do gaps and short sections of rock walls.

Between the two walls a distinct double row of inuksuit, 15 m long and 4 m wide, was recorded. This double lane was constructed of small rocks spaced so closely as to resemble low walls rather than individual inuksuit. A shooting pit was noted on the south side of the beginning of the double line and another on the north side at the end of the two low walls. The function of this double lane with pits is unknown. It would seem unrealistic to expect fleeing caribou to confine their run to this space. Immediately following this double line there is a gap of some 40 m before the main lane picks up again, with shooting pits located at each end of the gap. The gap and the shooting pits suggest another killing station.

Following the turn to the east the lane crosses an area of exposed bedrock (Figure 9). The surface here is so rocky that it is hard to imagine caribou running through this area. Instead it seems that the inuksuit must have been intended to direct caribou downslope onto the finer gravels of the beach ridges. After about 200 m there is a 30 m gap in the lane with shooting pits at both ends of the gap. It seems likely that the gap formed part of the original design and served as one of many killing stations. The lane then becomes a double line of cairns, followed by a short 15 m gap, and then a single line for hundreds of metres. The lane exits from the bedrock and crosses diagonally down across beach ridge gravels.

The final portion of the Eggington lane crosses bedrock and beach ridge materials towards the shore of Ferguson Lake. Several shooting pits were noted near the terminus of the drive. Eventually, some 300 m from the shore of

As the lane traversed the beach gravels a V-shaped funnel of lanes was noted, split to the north and south sides of the main lane, which continues through the center of the 6

BRINK: CARIBOU DRIVE LANES ON SOUTHERN VICTORIA ISLAND.

the lake, stone cairns become sporadic until none could be detected. The end of the drive appears uneventful; the lane climbs a slight rise caused by a circular dome of exposed bedrock (Figures 4 and 8). The final cairns of the Eggington site are found along the slope of this low dome and then end. No cairns were observed linking the Eggington drive system to the lake shore. Consideration of the Eggington site configuration suggests the following. Inuksuit begin in the centre of the saddle between the two hills, effectively deflecting herds in the saddle region to the southeast. This western portion of the lane roughly parallels the 80 m contour line that represents a notable break between the level area of the plain between the two hills and a steeper drop downslope towards the lake (Figure 2). By paralleling this contour a channel is created between the steep hillside to the southwest and the inuksuit of the Eggington drive to the northeast. This channel opens directly into the centre of the saddle; that is, to the northwest, into the prevailing winds (see below). Herds grazing in the low, vegetated areas of the saddle could be moved to the southeast, opposite the prevailing winds, with the hillside forming a barrier on one side and the line of inuksuit following the 80 contour line on the other. Shooting pits are located at many places along the lane and at the end, suggesting that killing was planned for many locations not just at the terminus, as seems to be the case at the POD site.

Figure 10. Aerial view of POD site. Arrows indicate vosition of inuksuit lanes beyond the solid stone walls. terminates at the narrow end of the funnel) is about 200 m long. However, there is a second line of inuksuit outside of the primary west lane that runs roughly parallel to the west lane about 50 m further to the west (upslope) (Figure 4). This outside lane, about 100 m long, may have been an early version of the game drive system, reflecting revisions over time, or this uphill lane may be connected with attempts to bring animals located further upslope into the funnel system. The distance between the farthest ends of the two primary lanes that form the funnel is about 365 m. A conspicuous feature of the caribou drive structure is at the narrow end where individual inuksuit have merged to form solid rock walls (Figures 11, 12). The north (downslope) wing has a wall of rock over the final 17 m of its course, and the west wing has a 42 m long rock wall. The considerably longer wall on the uphill side of the funnel may have been in response to a greater concern for animals escaping in this direction or for harvesting animals off the side of the hill. This may also explain the outer lane of inuksuit to the west. The transition from solid wall to individual inukshuk is abrupt at 17 m from the terminus of the north wing, but gradual in the west wing. That is, at a distance of 42 m from the end of the

THE POD SITE

The POD site lies on the east-facing side of the southernmost of the two hills that rise at the northwest end of Ferguson Lake (Figures 2 and 4). The funnel sits at the base of a northeast-projecting spur of the hill, with the wide end of the funnel opening to the northwest , facing into the same saddle between the hills as the Eggington site. Thus, the west wing of the funnel angles up the slope of the hill crossing beach ridge gravels, while the opposite (eastern) arm of the funnel angles into the saddle across more level terrain. the entire area of the funnel is rockiy and void of most vegtetation save lichen and moss. Although the funnel does not point towards the lake, it was constructed in an area where the land dips downslope from the side end to the narrow point of the funnel. The main part of the structure is some 2.5 km from Ferguson Lake and at an elevation of about 350 m asl All rocks used in inukshuk construction are consistent with the size and lithology of the local sandstone beach ridge rocks (Thorsteinsson and Tozer 1962). The primary rock features of the POD site are two converging lines of inuksuit and shooting pits located at the end and along the length of the funnel (Figure 10). The wings of the funnel are essentially symmetrical forming a clear V- shape, although the lines are not entirely straight. The north line of the funnel is about 250 m long, and the primary west lane (the one that

Figure 11. Terminus of the north wing of caribou funnel with shooting pit (foreground), 17 m wall of rock followed bv individual inuksuit.

7

FROM THEARCTIC TO AVALON: PAPERSIN HONOUR OF JIM TUCK

built near the terminus of the drive are very closely spaced, averaging just 1 m apart. At the farthest end, cairns are 4.5 m apart. Almost certainly this relates to a greater need for control of the animals near the final portion of the drive. Generally, cairns at the wider end are higher and built with more rocks than those at the narrow end. This may reflect the fact that it was more important for cairns at the wide end of the funnel to be seen and recognized from a greater distance, so as to direct the herds towards the final kill location. However , given that most of the cairns at the narrow end were interpreted as having collapsed, it is difficult to know if these data would be consistent with the condition of the inuksuit during the time they were in use.

,~

3m I O. C I2,sm SLOT

The north wing of the funnel is essentially continuous, with no gaps greater than about 6 m. It is also essentially straight with a few minor bends but no significant turns. Beyond the 17 m long wall inuksuit traverse fairly level beach ridge rock, and then cross a stretch of exposed bedrock where high points appear to have been incorporated into cairn construction. While crossing the bedrock the lane bifurcates into two side branches, the functions of which are unknown, but which may reflect an early phase of construction (Figure 13). The final portion of the north wing crosses beach ridge gravels and consists of cairns more widely spaced and more sinuous than at the narrow end of the funnel. A pond is located just to the east of the end portions of the north lane.

SLOT 0

,om

Figure 12. Sketch map of final portion of the caribou funnel with rock walls and shoo tin~ pits.

west wing there are single inuksuit, interspersed with short sections of wall, followed by more inuksuit. This may reflect an evolution of the building of the west wing, where separate cairns were, over time, linked with others to form a longer wall. Both walls range between 30 and 40 cm in height and are constructed of dry-piled contiguous rocks. The north lane of the funnel (Figure 13) was arbitrarily selected for more detailed study. Data collected included the distance between rock cairns, the height of cairns and the number of rocks used in cairn construction. A sample of 39 cairns was studied, split between the wide and narrow ends of the funnel. Data recording was complicated by uncertainty over the integrity of individual inuksuit. It was not always clear whether rock cairns collapsed over time or were still at their original size. While impossible to determine absolutely, searching the surrounding area sometimes revealed rocks that had apparently once been part of the cairns. It appeared that more of the rock piles at the narrow end of the funnel had collapsed than at the wider end. It may be that this area suffered greater stress (human and animal interaction) than the far end of the wing. Alternatively, the difference in the condition of the cairns may reflect the desire of the hunters to maintain the condition of the more distant elements of the wing.

7

C shooting pit

Table 1 presents data for inuksuit at the beginning Gust past the 17 m wall) and end of the north wing. Cairns that were presumed to have collapsed have not been used in the computation of averages. Noteworthy is that inuksuit

0

30mt

Figure 13. Sketch map of north wing of POD site.

8

BRINK: CARIBOUDRIVE LANES ON SOUTHERNVJCTORIA ISLAND.

Table 1. Data for inuksuit at the narrow and wide ends of north wing of POD site. Asterisk indicates data presumed · lete d ue to cairn co llapse. to be mcomp Narrow End of Funnel Wide end of Funnel Hei2ht (cm) Distance (m) No. of Rocks Distance ( m) No. of Rocks Heh?ht (cm) 1.5 5 6 6.5 5.5 3.5 2 4.5 6 4.50

3 3 2 1 3 3 9 7

35 27 42 37 30 37 45 30

3.88

35.38

0.7 0.7 0.9 0.9 0.5 1.5 1.2 1.3 0.5 1.3 0.9 1.2 1.6 0.9 1.5

3 2 2 2 4 1 1 1 1 4 1 2 3 3 2

10* 20* 18* 20* 15* 10* 10* 10* 20 30 35 10* 20 35* 30

1.04

2.31

26.65

which to shoot or spear the stampeding caribou as they ran through the narrowest confines of the trap.

The Shooting Pits Shallow indentations, interpreted as shooting pits, were a common feature of the caribou funnel. All such pits are roughly circular in plan view and range in diameter from about 2 to 3 m. Pits were constructed by scooping out beach ridge rocks and gravel to form a depression , and using the fill to form a berm around the outer edge of the pit. Generally the berm is highest at the side of the pit facing towards the funnel, and may be absent entirely from the opposite side. Shooting pits are located primarily along both wings of the funnel, although a few were noted inside of the two lanes. Most pits are found near the wide end of the funnel.

Some 25 m downslope from the end of the funnel lies a third shooting pit (Figure 12). This pit was situated directly in the middle of the path leading out of the funnel, equidistant from the upper two pits. It is of similar size and shape to the other two except that it lacks a slot for placement of a weapon . This distant pit is also a semi-circle with the open end facing downslope away from the funnel. The function of the pit seems clear -- to allow killing of animals that had escaped the confines of the funnel. Placement of the pit was probably critical: too close to the end of the funnel and it would become a visual and physical obstacle to the final drive down the funnel, too distant and it would permit a wider area for the surviving animals to disperse.

The two most substantial and conspicuous shooting pits are at the terminus of the funnel where the opposing rock walls at the end of the drive curve around forming a 3/4 circle (Figures 11, 12). Thus, these are not separate shooting pits but rather a continuation of the two rock walls. The pits are about 2.5 in diameter and 45 cm deep. Both crescent-shaped pits are open to the direction facing away from the funnel and most highly bermed at the edge facing into the funnel. Both have a narrow slot in the berm wall that faces into the end of the funnel (Figure 12). These are presumably for the positioning of a weapon, either bow and arrow or spear , allowing the hunters to remain in a lying position and thus completely hidden. The slots are not at right angles to the end of the funnel, but rather face slightly downslope past the end of the funnel, eliminating the possibility that the hunters could accidentally injure each other with loosed arrows. These are the only two shooting pits where these slots were observed, suggesting that concealment was most important at the termination of the drive. Clearly these two pits were precisely built to function as places from

Many other shooting pits were noted in the area of the funnel and it was not possible to count and record them all. Research attention was again directed to pits placed along the north wing where a total of five additional shooting pits were recorded (Figure 13). Measuring from the narrow end of the funnel, pits were recorded at distances of 100, 150, 157, 230 and 265 m, the latter being beyond, but in line with , the end of the inuksuit. Pits ranged from 40 to 60 cm in depth and were placed just outside or just inside the inukshuk line. Shooting pits placed along the lanes of the funnel may have functioned as either places for hunters to contain the stampeding caribou and keep them on the proper path , to kill those that came too close to the inuksuit and threatened to escape or both. However , it is interesting that all additional shooting pits along the north wing are 100 m and further from the terminus of the funnel. If the main purpose of the pits was to kill animals caught in the drive

9

FROM THE ARCTIC TO AVALON: PAPERS IN HONOUR OF JIM TUCK

of the second funnel and to assess its role in caribou hunting on the north side of Ferguson Lake.

then it seems that more pits should have been located closer to the end of the funnel. Placement of the pits a considerable distance from the end of the drive suggests that their primary purpose may have been as strategic positions from which the hunters could direct the course of the stampede. The two pits located at the very end of the funnel, and the one some 25 m beyond the funnel, are those presumed to be most directly associated with killing of caribou.

DISCUSSION OF THE EGGINGTON AND POD SITES Cultural Materials and Dating No faunal remains or definite artifacts were observed at either of the West Ferguson Lake sites. A few small, short, thin pieces of wood were noted. As wood is not locally available , these may have been portions of arrow shafts, several of which were found by Taylor (1968) at the nearby NiNg-4 site. Small lithic artifacts, typical of the caribou hunting assemblages of many Paleo-Eskimo cultures (McGhee 1978, 1996), would have been elusive among the bedrock and beach ridge gravels without a dedicated search, which time did not permit. Fauna! remains should have a higher degree of visibility yet none were observed. This is not unusual. Taylor (1965, 1972) commented on the absence of any caribou bone at the two inuksuit drives recorded a few kilometres east of the West Ferguson sites. At the Mantic caribou drive site on the upper Thelon River, Northwest Territories, Morrison (1981) noted the absence of any faunal remains and found only a single stone artifact. Fitzhugh (1981) did not report any fauna! materials and only a few stone artifacts associated with the Williams Harbour caribou drive in Labrador. Surface inspection of a number of extant wood fence caribou drives in the northern Yukon also revealed no obvious fauna! remains (R. Le Blanc, personal communication 2003). Carcasses may have been removed from the kills and taken to nearby camps. Major camp sites would not be expected in the immediate vicinity of a drive complex owing to the disruption a camp would cause to the efforts to drive game. Furthermore, among the Caribou Inuit, caribou kill sites employing water crossings were regarded as sacred places (Steenhoven 1962:27) , a fact that might also explain the general paocity of cultural materials.

Consideration of the POD site configuration suggests the following. It is possible that the Eggington and POD sites may have been used in concert - the POD funnel helping to trap animals that tried an uphill escape from the Eggington drive, and the latter acting as a barrier to animals that may have avoided the POD funnel after being driven off the hillside. However, the need to station people at one or the other kill site, and the presumably small population of available hunters , suggests that only one site would be the focus of a particular hunting event. Nestled against the base of the hill, this trap was constructed to take advantage of caribou herds from two sources: those moving southeast along the gravel ridges of the hill, and herds moving southeast out of the saddle. Note that the wide end of the funnel opens roughly parallel to the trend of a series of higher contour lines representing elevated ridges (Figure 4). These ridges continue for some 1500 m on a northwest-southeast trend-ideally suited for a downwind drive across the excellent running terrain of the ridge gravels. Herds may have frequented these ridges seeking insect relief or to cool down on warm days. Hunters or beaters, coming from above or behind, could intercept herds moving along the elevated ridges, driving them downslope towards the wide end of the funnel. The run to the funnel was again downslope, adding to the instability and vulnerability of the caribou. Some animals may have been killed as they passed shooting pits distant from the narrow end, but most were probably shot by archers lying in wait in the final two opposing pits and by those stationed beyond the end of the funnel.

In the absence of fauna! remains , artifacts or other temporally diagnostic material it is not possible to assign any age to the caribou drive complexes on West Ferguson Lake. Based on the assumption that people generally lived near the sea, isostatic rebound and the consequent formation of beach ridges has proven a useful tool for relative dating of archaeological sites in south-west Victoria Island (Savelle and Dyke 2002). Placement of the drive lanes, however, was determined by requirements for gathering and directing the movement of caribou. As such, drive structures invariably crosscut slopes and ancient beach ridges , and thus could not be dated using the chronologies established for the emergence of beach ridges over time. Within a few kilometres of the West Ferguson Lake sites Taylor (1967, 1972) has recorded sites of Pro-Dorset, Dorset, and Thule age. Any or all of these archaeological cultures could have built and used the game drive systems. Taylor 's

A SECOND FUNNEL? Helicopter scheduling placed limits on the time available to examine the area surrounding the POD site. Just prior to departure we observed what appeared to be a second complete caribou funnel located downslope from the first; that is, to the southeast, past the narrow end of the POD funnel. We had no opportunity to examine this apparent feature and were not able to obtain a map reference. However, cursory inspection suggested that this was a very similar structure consisting of two converging lanes of cairns ending at a narrow opening. If correct, this combination of traps suggests an even more complex process of caribou hunting, with the second trap possibly intended to capture animals that escaped the POD trap. Additional field research is needed to verify the existence

JO

BRINK: CARIBOU DRIVE LANES ON SOUTHERN VICTORIA ISLAND.

( 1965: 16) report of a Thule style antler arrowhead associated with the nearby NiNg-4 caribou drive suggests that the West Ferguson sites were also used in Thule times, as the sites appear to form one large game-trapping network. Judging from lichen cover observed on the upfacing surfaces of stones used in inukshuk construction it would appear that the features are of considerable antiquity. As noted above, once built the structures could have been used and modified by a number of successive cultures.

Comparisons The Eggington Site

Single lines of inuksuit like the Eggington site are found all over the Arctic and Subarctic world (see Anell 1969; Blehr 1990; Gordon 1990; Spiess 1979 for reviews). Boas (1964:93), who sailed into Davis Strait in 1883 to do research among the Central Inuit, commented: "Such monuments are found all over the country, most of them having the appearance of being very old." Single stone lines are said to be more common than funnel-shaped arrangements (Arima 1975:148; Birket-Smith 1929:111). Typically, single lines of cairns led to water crossings where swimming caribou were easily killed (BirketSmith 1929:106; Boas 1964:93; Jenness 1922:124, 149; Morrison 1981:176). Less commonly, lanes of stone cairns were designed, like the Eggington site, to capture caribou on land by shooting or spearing them over the course of the drive (Balikci 1970:41; Boas 1964:100; Hearne 1971:320-321; Jenness 1922:148-149). In fact Gordon (1990:285) notes that single lanes of rock cairns were used by the Caribou Inuit to prevent animals from entering water in areas where kayaks were not used or not suitable. Jenness (1922:149) states that, at least in historic times, the Copper Inuit of Victoria Island seldom used kayaks in killing caribou and that drives using stone lines ending in shooting pits were the norm. The fact that other inukshuk drive complexes recorded for Victoria Island (McGhee 1971, 1972; Taylor 1972) are likewise land-based (though near the coast) supports Jenness's statement.

Season of Use

The probable season of use of the two West Ferguson sites can be posited. Historically, limited numbers of caribou did occasionally over-winter on Victoria Island (Banfield 1954 Pt. 1:15, 68-69), but the numbers were very low and not dependable. The great migration of herds across the frozen straits separating Victoria Island from the mainland occurred in the spring and consisted primarily of cow/calf herds, with bulls preferring to remain throughout the summer in the barren grounds (Banfield 1952 Pt. 2:21). The cow/calf herds returned to the mainland in autumn. In historic times, the Copper Inuit of Victoria Island turned almost exclusively to sealing for the winter, typically living on the sea ice for up to six months at a time (Damas 1984a:398; Jenness 1922:111). Thus, the West Ferguson sites could, realistically, only have operated in the summer and fall seasons. This time limit can be further refined. Since cow/calf groups made up the majority of the population, studies have shown that cows nursing calves in the spring are severely depleted in muscle and especially fat tissue (Adamczewski et al. 1987; Dauphine 1976). Cow condition improves over summer and is prime in the early fall (Reimers and Ringberg 1983; Ringberg et al. 1981). Also, spring and summer hides are rendered almost useless to humans owing to holes caused by emerging fly larvae (Balikci 1970:42; Damas 1972:13; Stefansson 1951:334). By late summer these holes have healed and hides are prime. Caribou hides, essential winter clothing of the Central Inuit, were as much the objective of the hunts as was food (Blehr 1990:320; Stefansson 1951:337). It can be suggested that the primary, though not necessarily only, use of the West Ferguson caribou kills would have been in the late summer and fall seasons. This is consistent with ethnographic records that document a strong tendency for communal kills in autumn, especially for the purpose of acquiring hides and surplus meat for winter (Arima 1984:449; Balikci 1964: 12, 1970:43; Birket-Smith 1929: 11O; Blehr 1990:309; Damas 1972:13; Gordon 1990:289; Jenness 1922:102; Steenhoven 1962:25; Stefansson 1951:337).

Taylor's (1972) record of a major inukshuk system, NiNg-4, just a few kilometres east of the Eggington site may well be the most comparable example to the Eggington site. It employs the same configuration of landscape, utilizing beach ridges in the hill and valley country north of the Ekalluk River, is primarily a single line of cairns and contains generally similar stone structures in the form of inuksuit and shooting pits. The NiNg-4 inukshuk system is a 1.2 km long crescentshaped series of stone cairns and shooting pits that hugs the top of a beach ridge situated between the Ekalluk River and the south-eastern flank of the same hill that figures into the operation of the Eggington and POD sites (Figure 2). Taylor (1972:73-75) proposes that caribou were rounded up from the southern, vegetated, low-lying end of the crescent near the Ekalluk River, then driven though the inner curve of the crescent to a series of gaps in the inukshuk line where most of the shooting pits are located. Taylor's interpretation has the animals running south to north, into rather than with the prevailing wind, and the drive may have gone in the other direction. Also, it's possible that the parabola functioned the other way, designed to channel the migrating herds between the outer curve of the crescent and the southeast side of the adjacent hill. Unlike the Eggington site, NiNg-4 appears to have been constructed with continuous inuksuit at both ends of the drive, with gaps for escape in the middle. The

11

FROM THE ARCTIC TO AVALON : PAPERS IN HONOUR OF JIM TUCK

Eggington drive is more complex, with many gaps, walls, parallel lines and shooting pits spread over a greater distance. If shooting pits are associated with killing locations (as opposed to herd control), then the Eggington site has a number of kill stations, while NiNg-4 apparently has a single, central one. The presence of many caribou drives on the north side of Ferguson Lake suggests that all these features were built by the same people, possibly at the same time, the location of the herds determining which drive system was used at any one time.

though some have clearly collapsed. This inuksuit line leads towards a lakeshore where the row of cairns changes to what the authors describe as a "fence." This clearly represents an area of greater reinforcement, equivalent to the stone walls documented along certain portions of the Eggington drive and at the end of the POD funnel. The fence is described as being 70 m long, made of stone blocks ranging in height from 20 to 60 cm (Grnnnow et al. 1983:41). The final fence continues downslope towards the lakeshore; however, placement of some 35 shooting blinds along the drive complex suggests, as at Eggington, that the primary killing of the game occurred on land, not in the water.

Jenness provides a reminiscent description of a Copper Inuit caribou drive on Victoria Island where a single line of inuksuit ends short of a nearby lake:

One of the most thoroughly studied single-line stone cairn caribou drives comes not from the Arctic, but from Morrison's (1981) description of the Mantic River drive complex in the Subarctic. Most Subarctic caribou drives, located where wood was accessible, employed wooden posts rather than stone structures (see Spiess 1979: 106118). The Mantic River stone drive site, a type almost always associated with Inuit culture in northern Canada, is located in the homeland of historic Chipewyan people, and Morrison (1981: 172) believes they, not Inuit, built and used the drive. The site, situated on the upper Thelon River, Northwest Territories, consists of some 3.3 km of stone lines and cairns complete with 36 shooting pits (Morrison 1981: 180). Reminiscent of the Eggington site, Morrison ( 1981) proposes a number of killing locations staggered along the course of the lines based on the presence of shooting pits and gaps in the rock lines. Many individual lines are single rows of stone, but some parallel avenues are noted as is one long V-shaped convergence of two lines that ends, not with shooting pits as at the POD site, but a single rock line leading to the river. However, most lane construction and all shooting pits are situated away from the water edge, suggesting that killing of caribou was intended to be conducted inland rather than while the animals were crossing the river.

These drives call for a considerable amount of strategy and the careful utilization of topographical features. The caribou may be grazing at the end of a plain a quarter of a mile wide, bounded by a low ridge on one side and a lake on the other. The hunters will set up their turf-capped stones at intervals of thirty or forty yards along the top of the ridge, and probably swing the line across the plain to within a hundred yards of the water's edge. (Jenness 1922:149) Except for his more distant spacing of the cairns, Jemess' s description applies exceedingly well to the Eggington site. A description equally reminiscent of the Eggington drive is provided by Balikci for the Netsilik Inuit: A somewhat more complicated and productive method of hunting caribou with bow and arrow involved the co-operation of "beaters" and "archers." Stone cairns designed to frighten the caribou were erected in a row on top of a ridge leading to a lake. At the end of the line of cairns and near the lake the archers lay in ambush, concealed behind low piles of stones. The beaters, using wolf cries, drove the caribou down the line of cairns towards the concealed archers, who then attacked the caribou with a barrage of arrows. Several caribou could be killed in this manner. (Balikci 1970:41)

The POD Site

In contrast to the common single stone lines of the north, V-shaped funnels of stone roughly similar to that of the POD site are rarely reported in the literature. Similar forms of V-shaped funnels were observed by Vorren (1965:516-517) in the 1930s on Norway's Varanger Peninsula. Many converging stone lines were reported, which, like the POD site, were often located on ridges of loose stones that rise up from the valleys (Vorren 1965:516). Stone cairns were likewise of minimal construction, averaging 30 to 70 cm in height. Vorren (1965:517-518) notes that as the wings of the funnel radiate out from the end of the trap, the stone cairns "nearest the enclosure form solid walls, but which soon take the form of small heaps of stones or sharp upright stones at a distance of 3-4 m from one another." The switch from rock walls to individual cairns and the close spacing of the cairns are strikingly similar to the POD

The single stone line caribou drive at the Aasivissuit site in west Greenland is strikingly similar to the Eggington site and is one of the few other instances where details of stone inuksuit have been reported. Some 100 cairns are found over a 4 km distance along a mountainside, giving an average spacing of 39 m (Grnnnow et al. 1983). The majority of the cairns are composite, made up of 6 to 10 stones, while a minority are single stones raised in prominent places (Grnnnow et al. 1983:41). As at the Eggington site, many cairns at Aasivissuit are built on bedrock outcrops, and others consist of single rocks propped up on a fortuitously placed in situ rock (Grnnnow et al. 1983 :41). Average cairn height is about 50 cm,

12

BRINK : CARIBOU DRIVE LANES ON SOUTHERN VICTORIA ISLAND .

archers ... At Talurkuarq [the big hiding place] , now Spence Bay settlement ... there existed a remarkable structure consisting of two slightlycurved stone walls over 70 feet long leading to a narrow passage only 5 or 6 feet wide. At their opposite end the continuous walls were prolonged by some stone cairns. A small lake was to be found in front of the system. A single man armed with a heavy spear ... waited at the upper ends of the walls. Qipingajuk, an old Pelly Bay Eskimo who witnessed a hunt at that place after the tum of the century, informed me that a large number of beaters were necessary for this hunt, usually up to fifteen individuals , who would drive the herd around the lake towards the large opening of the stone walls. At the end of the taalun the caribou were easily speared ... This taalun was believed to have been built by the Tunit. (Balikci 1964:13)

site. Some of the Norwegian drives led to actual stone enclosures. These seem rare in the Canadian Arctic, although Savelle (1987, plate 7b) illustrates what may be a stone corral- like structure used for trapping caribou on the Adelaide Peninsula , Nunavut. For the Canadian Arctic, there are a number of references in the historic and ethnographic literature to converging stone lines (Anell 1969:11; Arima 1975:148; BirketSmith 1929:110, 111; Rasmussen 1927:74; Steenhoven 1962:27; Stefansson 1914:58). Steenhoven (1962:27) describes Caribou Inuit using two long, converging rows of stone cairns with chunks of sod on top, spaced as much as 90 m apart, leading to water crossings. In what seems generally similar to the set up and presumed operation of the POD funnel, Stefansson describes a V-shaped drive he observed while with the Central Eskimo of the Coronation Gulf region:

The term "Tunit" embraces a number of concepts in Inuit culture, one of which is the name given by modem Inuit to the unrelated group of Arctic people who came before them; that is, people of the Dorset culture (McGhee 1996:136). McGhee (1996:135) cites a Netsilingmuit elder as saying that the Tunit people made the Arctic habitable by leaving them such things as stone drives for killing caribou . This reinforces the point that drive lane structures, if built at locations that possessed the essential requirements for gathering , directing, and killing caribou , may have been used for many generations by successive cultural groups.

When a band [of caribou] is discovered feeding, a V-shaped "fence" is constructed somewhere beyond their line of vision, generally to leeward. The "fence " consists of straight lines of stones or pieces of sod raised on end and set twenty to forty yards apart ... These stones, sod or blocks of snow are often not over eighteen inches high and no particular pain need be taken as to their shape or appearance , though a dab of earth is usually put on a block of snow or light colored stone so as to make sure that animals will see it... When the fence is completed , a half dozen men make one a mile long in an hour, the men conceal themselves in the angle of the V, the women and children with the dogs , go to windward of the deer to drive them ... If they attempt to pass outside the wings of the V, someone is there to tum them, and usually the band moving in single file along side the V shaped fence, much like horses along a barbedwire fence, arrives at a walk or slow trot to the point where the angle of the fence becom es so narrow , about one hundred yards, that they begin to notice the fence on the other side and to see there is no opening. Then they bunch up irresolutely and give the hunters good opportunity to shoot. (Stefansson 1914:58)

Certainly the best-documented and most similar site to the POD site in the archaeological literature is that of Williams Harbour in northern Labrador (Fitzhugh 1981). It consists of two converging lanes of rock walls and inuksuit situated on a broad plain that is bounded by a harbour on one side and a high hill on the other (Fitzhugh 1981: 193). Caribou moving south on the peninsula would have to pass between the edge of the harbour and the side of the hill, and an old caribou trail was observed traversing the center of the funnel. Like the POD site, the narrow end of the funnel consists of solid rock walls rather than single cairns. The two walls are 40 m and 50 m long, with individual inuksuit extending this distance , in the one reported case, approximately another 100 m (Fitzhugh 1981: 196). A number of crescent-shaped shooting pits, excavated into the gravel, are scattered along the Williams Harbour drive system; some along the two walls, others inside, still others extending well beyond the ends of the walls. The narrow end of the funnel, about 7 m wide, lacks the two opposing shooting pits of the POD site. Rather, it terminates next to a huge, naturally placed boulder on one side that would have acted as an impressive barrier. Several shooting pits are located near the boulder. No bone was observed , but two diagnostic artifacts suggest use of the trap during middle and late Dorset times, some 800 to 1,600 years ago (Fitzhugh 1981 :202).

Stefansson (1914:58) witnessed eleven caribou killed in such a trap by a group of four male hunters and six women plus children. This was regarded as a relatively poor catch for a drive operation. Consistent with the POD site, Arima (197 5: 148) makes the observation that converging lines of stones tended to lead to archers hidden near the narrow end rather than to water crossings. Of the more detailed reports , Balikci provides the following account of caribou hunting by the Netsilik: Most interesting was the one associated with the taalun, stone constructions to conceal the

13

FROM THE ARCTIC TO AV ALON: PAPERS IN HONOUR OF JIM TUCK

blinds are not constructed as to the size of a person, but rather with regard to concealment of a bow held horizontally (Grnnnow et al. 1983:29).

Shooting Pits and Large Blind Structure It would be expected that the shooting pits recorded by Taylor (1965, 1972) at the NiNg-4 site on the north side of the Ekalluk River would be most similar to those at the West Ferguson sites just a few kilometres to the east. Yet those recorded by Taylor appear to be considerably more substantial. Taylor counted 24 pits over the 1.2 km distance of the crescent-shaped line of inuksuit. Pits are described as gouged out of gravel and often faced with a low balustrade of stones (Taylor 1965: 16), as are the West Ferguson pits. However, Taylor reports pits up to 3.7 m across (12 feet) and 0.61 m (2 feet) deep. None equal to this size, and hardly any half this size, were noted at the Eggington or POD sites. Taylor ( 1965: 16) reports that even the smallest pit would hold two archers. Very few features observed at the two West Ferguson sites would have concealed two archers; the large blind structure at Eggington and the three final shooting pits at the POD site being the exceptions. Why shooting pits are substantially larger at NiNg-4 is unknown.

The seven tiny shooting pits recorded near the tall inukshuk and large blind/pit feature (Figure 7) are unusual in the West Ferguson region and seem to add to the apparent significance of this point along the Eggington drive complex. The pits, consisting of an almost imperceptible scooping of gravel and placement of a lip of single small stones, seem too small to have been of any practical use as either shooting pits or blinds. Perhaps the tiny pits were more symbolic than practical, built with the intent of making the area resemble a hunting or killing location. The imposing presence of the single tall inukshuk certainly lends an unusual atmosphere to the location. However, another possibility is derived from a journal entry by Jenness (1922:138) for August 15, 1915. Summering with the Copper Inuit on Victoria Island he records that while men practiced their archery, the children imitated them and shot at imaginary deer "from miniature pits." The tiny pits at the location of the tall inukshuk and large 3/4 circle structure may well have been used by children while adults watched for caribou.

Probably the most detailed and relevant description of the making of shooting pits comes from Jenness during his time among the Copper Inuit:

The relatively massive, semi-circular stone blind structure also found at this location is unlike any others recorded in the West Ferguson region. Balikci ( 1964: 11) reports that N etsilik caribou bunters, located to the southeast of the Copper Inuit, concealed themselves behind large, onemetre high semi-circular stone walls situated on hill tops for the purpose of observing game. The location of the large Eggington stone structure , on an upslope beach ridge with a commanding view of the region, and its size and shape suggest that it may have served a similar purpose. Indeed, Jenness, while hunting caribou with the Copper Inuit on Victoria Island, describes an identical sounding feature: "During the first days of our stay lkpakhuak made a semi-circular stone shelter on top of a ridge near the camp which commanded a wide view of the surrounding country; every day one or other of us would spend several hours there watching for caribou" (Jenness 1922:141).

Between the end of the barricade and the lake each hunter digs a shallow pit, using for his adze a sharpened antler. He stabs this into the turf, pulls the clod up with his hands and lays it around the edge. In a few minutes he has made a saucer-shaped depression faced with turf and stones or snow to make it as inconspicuous as possible. Here he lies, face downwards, with his bow and arrows by his side, waiting for the deer to be driven within range. (Jenness 1922:149) Jenness ( 1922: 148) further notes that most caribou are shot within 20 paces of the shooting pits, as this is about the effective range of the Inuit bow and arrow. Most of the shooting pits observed at the West Ferguson sites were of the type described by Jenness, only excavated into gravel rather than turf. Pits, or shooting blinds, at the Williams Harbour site in northern Labrador are crescent- shaped, scooped out of gravel, average some 50 to 75 cm in depth, and each is presumed to have held one hunter (Fitzhugh 1981: 198, 201). Except for the greater depth, these are quite similar to those found at the West Ferguson sites. Balikci (1964:14) reports that shooting pits among the Netsilik were simple semi-circular store walls behind which one or two archers lay in ambush. Some 35 shooting coverts or "hides" are reported over the 4 km distance of the inukshuk system reported at the Aasivissuit site in west Greenland (Grnnnow et al. 1983:45). These appear to be more of a low stone wall construction for concealment rather than an excavated pit. They range in height from 13 to 45 cm, with an average distance across of 1.5 m. The authors make the interesting observation that these

CONCLUSION

Benedict (1996) has made an eloquent case for the significance of communal game drive structures. He points out that at one time much of North America would have furnished evidence of communal driving of a wide variety of animals, but that most has been removed by settlement, agriculture, development and decay of perishable materials (Benedict 1996:2-4). The remote, rugged, and largely uninhabited Arctic region is a notable exception. As such the importance of caribou drives is magnified. Not because they are unique--communal drives were once ubiquitous--but because the Arctic is one of the few places left in the world where these

14

BRINK: CARIBOU DRIVE LANES ON SOUTHERN VICTORIA ISLAND.

impressive features are preserved and can still be studied (Benedict 1996:2-4). This fact lends urgency to the need to record these structures. Despite the remoteness of Arctic regions, development is a threat to inuksuit and other ancient stone structures (Heyes 2002). It is unlikely that these sites will provide rich returns of artifacts, but they have the power to reveal evidence of the complex relationships between environment , hunting techniques and knowledge of animal behaviour; information that cannot be garnered from any other type of archaeological site. Such structures are thus ideally suited to build bridges between contemporary Aboriginal elders, who still possess knowledge of the construction and operation of these structures, and archaeologists who seek to understand them. Such work has recently been conducted by Friesen (2002) in the Ekalluk River region Most importantly, these stone features are a testimony to a skillful adaptation that not only eased the course of life, but perhaps permitted it, in one of the most difficult and demanding environments in which humans have lived.

Anell, Bengt 1969 Running Down and Driving of Game in North America. Studia Ethnographica Upsaliensia XXX. Lund: Berlingska Boktryckeriet. Arima, Eugene Y. 1975 A Contextual Study of the Caribou Eskimo Kayak. Canadian Ethnology Service; Mercury Series, 25. Ottawa: National Museums of Canada. 1984 Caribou Eskimo. In Handbook of the North American Indians, vol. 5, Arctic, edited by David Damas,.:447-462. Institution Press.

Washington,

D.C.:

Smithsonian

Balikci, Asen 1964 Development of Basic Socio-Economic Units in Two Eskimo Communities. Bulletin 202, National Museums of Canada, Anthropological Series , 69. Ottawa: Queen's Printer. 1970 The Netsilik Eskimo. American Museum of Natural History. New York: The Natural History Press.

ACKNOWLEDGEMENTS. I am honoured to include this work in recognition of the distinguished and productive career of friend and colleague James A. Tuck. I thank the University of Wisconsin Press and the Editor of Arctic Anthropology , Susan Kaplan, for permission to reproduce this paper, part of a longer article that appeared in that journal. Research on Victoria Island by the late William E. Taylor Jr. was supported by a grant from the Social Sciences and Humanities Research Council, and by the Polar Continental Shelf Project. James Helmer of the University of Calgary helped with all aspects of data collection in the field and I am grateful for his company and advice. Carolyn Lilgert helped with several of the figures. I acknowledge overall support for my work provided by the Royal Alberta Museum. Comments on an early draft by James Benedict, T. Max Friesen and Raymond Le Blanc corrected errors and helped focus the paper. My thanks to three anonymous reviewers for their comments, and to the Editor for assistance with the manuscript. Bill Taylor was one of the pioneers of Arctic archaeology. I am grateful that he included me on his return to the north. His premature passing is lamented by all who knew him.

Banfield, A. W. F. 1954 Preliminary investigation of the barren ground caribou. Part 1: Former and present distribution, migration, and status. Part 2: Life history , ecology, and utilization. Wildlife Management Bulletin Series, 1:IOA. Ottawa: Queen's Printer. Benedict , James B. I 996 The Game Drives of Rocky Mountain National Park. Research Report , 7. Ward: Center for Mountain Archaeology. Birket-Smith , Kaj 1929 The Caribou Eskimos, Material and Social Life and their Cultural Position. Report of the Fifth Thule Expedition 1921-24 , vol. 5, nos.I and 2. Copenhagen: Gyldedalske Boghandel Nordisk Forlag. Blehr , Otto 1990 Communal hunting as a prereqms1te for caribou (wild reindeer) as a human resource. In Hunters of the Recent Past. L.B. Davis. and B. 0. K. Reeves, eds.:304-326. London: Unwin Hyman.

REFERENCES CITED

Boas, Franz 1964 The Central Eskimos. Lincoln: Nebraska Press.

Adamczewski, J. Z., C. C. Gates, R. J. Hudson and M.A. Price 1987 Seasonal changes in body composition of mature female caribou and calves (Rangifer tarandus groenlandicus). Canadian Journal of Zoology 65: 1149-1157.

University of

Burch , Ernest S. Jr. 1972 The caribou/wild reindeer as a human resource. American Antiquity 37:339 -368.

15

FROM THE ARCTIC TO AV ALON: PAPERS IN HONOUR OF JIM TUCK

Collins, Henry B. 1984 History ofresearch before 1945. In Handbook of North American Indians, vol. 5, Arctic. edited by David Damas: 8-16. Washington, D.C. Smithsonian Institution Press.

Gubser, N. 1965 The Nunamiut Eskimo Hunters of Caribou. New Haven: Yale University Press. Hallendy, N. 1994 Inuksuit: semalithic figures constructed by Inuit in the Canadian Arctic. In Threads of Arctic

Damas, David 1972 The

Copper Eskimo. In Hunters and Gatherers Today, edited by M. G. Bicchieri:3-50.

Prehistory: Papers in Honour of William E. Taylor Jr, edited by D. Morrison and J-L. Pilon:387-408.

New York: Holt, Rinehart, and Winston.

Mercury Series Paper , 149, Archaeological Survey of Canada. Hull: Canadian Museum of Civilization.

1984a Copper Eskimo. In Handbook of the North American Indians, vol. 5, Arctic, edited by David Damas:397-414. Washington, D.C.: Smithsonian Institution Press .

Hearne , Samuel 1971 A Journey from Prince of Wales's Fort in

Hudson 's Bay to the Northern Ocean in the years 1769, 1770, 1771, and 1772. Edmonton: M . G.

1984b Central Eskimo: Introduction. In Handbook of the North American Indians, vol. 5, Arc tic, edited by David Damas:391-396. Washington , D.C.: Smithsonian Institution Press.

Hurtig Ltd. Heyes, Scott 2002 Protecting the authenticity and integrity of lnuksuit within the Arctic milieu. Etude/lnuit/ Studies 26(2):133-156.

Dauphine, T. C., Jr. 1976 Biology of the Kaminuriak Population of Barren Ground Caribou. Report Series, 38. Ottawa: Canadian Wildlife Service.

Jenness, Diamond 1922 Report of the Canadian Arctic Expedition

Fitzhugh, William W. 1981 A prehistoric caribou fence from Williams Harbour, Northern Labrador. In Megaliths to

1913-18. Volume XII: The Life of the Copper Eskimos. Ottawa: F. A. Acland.

Medicine Wheels: Boulder Structures in Archaeology,

Kelsall, John P. 1968 The Migratory Barren-Ground Caribou of Canada. Canadian Wildlife Service, Department of Indian Affairs and Northern Development.

edited by Michael Wilson, Kathie L. Road, and Kenneth J. Hardy: 187-206. Proceedings of the Eleventh Annual Chacmool Conference. Archaeological Association, Calgary: University of Calgary.

McGhee, Robert 1971 An archaeological survey of western Victoria Island, N.W.T., Canada. Contributions to

Friesen, T. Max 2002 Analogues at lqaluktuuq: the social context of archaeological inference in Nunavut, Arctic Canada. World Archaeology 34(2):330-345.

Anthropology VII: Archaeology and Physical Anthropology: 157-191. National Museum of Canada Bulletin, 232. Ottawa: National Museum of Canada. 1972 Copper Eskimo Prehistory. Publications in Archaeology, 2. Ottawa: National Museums of Canada.

Fyles, J. G. 1963 Surficial Geology of Victoria and Stefansson Islands, District of Franklin. Geological Survey of Canada Bulletin, 101. Ottawa: Geological Survey of Canada Department of Mines and Technical Surveys. Gordon, Bryan C. 1990 World Rangifer

communal

hunting.

1978 Canadian Arctic Prehistory. Toronto: Van Nostrand Reinhold Ltd. 1996 Ancient People of the Arctic. Vancouver: University of British Columbia Press.

In

Hunters of the Recent Past. L. B. Davis and B. 0. K. Reeves, eds.:277-303. London: Unwin Hyman.

Morrison, David 1981 Chipewyan drift fences and shooting blinds in the central Barren Grounds. In Megaliths to Medicine Wheels: Boulder Structures in Archaeology, edited by Michael Wilson, Kathie L. Road , and Kenneth J. Hardy:171-186. Proceedings of the Eleventh Annual Chacmool Conference. University of Calgary. Archaeological Association. Calgary:

Grnnnow, Bjarne, Morten Meldgaard, and forn Berglund Nielsen 1983 Aasivissuit - The Great Summer Camp.

Archaeological, Ethnographic, and ZooArchaeological Studies of a Caribou Hunting Site in West Greenland. Meddelelser om Grnnland. Man and Society , 5. Copenhagen: Copenhagen : Commission for Scientific Research in Greenland.

16

BRINK: CARIBOU DRIVE LANES ON SOUTHERN VJCTO RIA ISLAND.

1992 Arctic Hunters: The Inuit and Diamond Jenness. Hull: Canadian Museum of Civilization.

Stefansson, Vilhjalmur 1914 The Stefansson - Anderson Arctic Expedition of the American Museum: Preliminary Ethnological Report. Anthropological Papers of the American Museum ofNatural History, vol. 14, pt. 1. New York: American Museum ofNatural History.

Rasmussen, Knud 1927 Across Arctic America, Narrative of the Fifth Thule Expedition. New York: G. P. Putnam's Sons. Reimers, E. andT. Ringberg 1983 Seasonal changes in body weight of Svalbard reindeer from birth to maturity. Acta Zoologica Fennica 175:69-72.

1951 My Life with the Eskimo. New York: The Macmillan Company. Taylor, William E. Jr. 1965 The fragments of Eskimo prehistory. The Beaver Outfit 295:4-17.

Ringberg, T. M., R. G. White, D. F. Hollerman, and J. R. Luick 1981 Body growth and carcass composition in lean reindeer (Rangifer tarandus tarandus L.) from birth to sexual maturity. Canadian Journal of Zoology 59: 1040-1044.

1967 Summary of archaeological field work on Banks and Victoria Islands, Arctic Canada, 1965. Arctic Anthropology 4(1):221-243. 1968 Of soldiers and Inukshuks. North May­ June: 10-15

Savelle, James M. 1987 Subsistence-Settlement System Changes in the Central Canadian Arctic, A.D. 1000-1960. BAR International Series, 358. Oxford: BAR Publishing.

1972 An Archaeological Survey Between Cape Parry and Cambridge Bay, N W.T., Canada in 1963. National Museum of Man. Archaeological Survey of Canada Paper, 1. Ottawa: National Museums of Canada.

Savelle, James M. and Arthur S. Dyke 2002 Variability in Palaeoeskimo occupation on south-western Victoria Island, Arctic Canada: causes and consequences. World Archaeology 33:508-522.

Thorsteinsson, R. and E.T.Tozer 1962 Banks, Victoria, and Stefansson Islands, Arctic Archipelago. Geological Survey of Canada, Memoir, 330. Ottawa: Queen's Printer.

Spiess, Arthur E. 1979 Reindeer and Caribou Hunters: An Archaeological Study. New York: Academic Press.

Vorren, 0mulv 1965 Researches on wild-reindeer catching constructions in the Norwegian Lapp area. In Hunting and Fishing, edited by Harald Hvarfner:513-536. Lulea: Norrbottens Museum.

Steenhoven, Geert Van Den 1962 Leadership and Law among the Eskimos of the Keewatin District, Northwest Territories. The Hague: Uitgeverij Excelsior.

17

FROM THE ARCTIC TO AV ALON: PAPERS IN HONOUR OF JIM TUCK

18

DEFENSE IN AN IROQUOIS VILLAGE William Engelbrecht Buffalo State College

been considerable variation in palisade height, even around the same community, but it is thought that the majority of palisades were somewhere between 4 and 10 metres in height (Prezzano 1992:248). Champlain observed that the palisade around the Iroquois village he attacked in 1615 was some 9 metres high, while he estimated a triple palisade around a Huron village at 11 metres (Grant 1907: 283,292).

Iroquois oral tradition states that fighting was once common. "Feuds with outer nations, feuds with brother nations, feuds of sister towns and feuds of families and of clans made every warrior a stealthy man who liked to kill ...."(Parker 1968a: 17). Since fighting involved social and political units of variable scale, it is not surprising that an Iroquois concern with defense can be inferred archaeologically at both the regional and village level. Regionally, the development of Iroquois nations is reflected in the location of contemporaneous villages near one another, an insight first articulated by Jim Tuck (Tuck 1971:215; Engelbrecht 2003:112-126). It is assumed that warriors within a village would defend nearby villages in the event of an attack. At the village level, the development of large palisaded hilltop communities typi cal of the sixteenth century Iroquois is commonly interpreted as a defensive response to endemic warfare. This paper explores the possibility that features within the vil lage also reflect a concern with defense.

In some cases a ditch was dug in front of the palisade with the dirt piled at the base, creating what today is known as an earth ring. Gabriel Sagard observed large tree trunks laid along the base of a palisade to strengthen it, and branches and saplings were woven between the upright posts to create a wall 2.5 or 3 metres high (Sagard-Theodat 1968: 91-95). These palisades must have involved a great deal of labor to construct. While palisades may have had secondary functions as snow fences or windbreaks, or served various symbolic functions, their sturdiness and the obvious effort involved in their construction suggests that defense was their primary function.

In this paper , a feature is considered to have a defensive property if it serves to restrict physical access. The location of many Iroquois sites on a hilltop or point of land with a steep bank on one or more sides provides well known examples. A palisade typically protected accessible routes of entry, channeling movement through a narrow passageway. While this palisade was the first and most important line of defense, on occasion enemies still gained entry to a village. After briefly reviewing information on palisades, additional points at which enemy movement within a village might have been restricted will be considered.

Palisades with more than three rows of posts are sometimes encountered in excavation (Figure 1). There are no ethnographic descriptions of such palisades, and a palisade of four or five rows of posts would seem unnecessary. Gary Warrick (1988) has suggested that village duration might be inferred from the number of post molds observed along a longhouse wall, many post molds suggesting decay of the original posts and repairs using new posts . In similar fashion, the presence of more than three rows of palisade posts suggests rebuilding. This is a dif-

PALISADED VILLAGES

N

When Europeans first encountered Iroquois palisaded villages they referred to them as "castles". This may strike the modem reader as a quaint misnomer, but originally "castle" meant fortified residence. Both European Medieval castles and Iroquois palisaded villages were built for defense and restricted access to the interior. Also, a village on a hill surrounded by a 6 or 9 metre high palisade would have appeared imposing even if it were not made of stone in the fashion of a walled European castle. Finally, population density would have been high in both European castles and Iroquois palisaded villages.

t

88N 5W

•• 84 N '--

-

-

-+----,-.,,_..

-i-

--

--+- -

-

~

-

-~

4W

..·.

2metr es

82 N 3E

Iroquois palisades evolved over time, becoming increasingly effective as defensive structures (Keener 1999:782). While the diameter of palisade post molds varies within sites, over time there is an increase in size (Ritchie and Funk 1973:363; Prezzano 1992:242). There may have

SON 5E

•••• •

..: .

• •

Figure 1. Five row palisade at the Eaton site, West Seneca, New York.

19

FROM THE AR CTIC TO AV AL ON: PAPERS IN HONOU R OF JIM TUCK

ferent situation from that of multiple palisade rows resulting from village expansion or contraction.

Finlayson 1998:20; see Figure 2). While Rankin (2000) has demonstrated that the occupation of this site is more complex than originally thought, these lines can still be interpreted as cordons to restrict access. A consideration of other village plans suggests that placement of houses within a village may also have been motivated by considerations of defense. The late Dr. Robert Funk recently produced a new plan of the houses at the Garoga site (Funk and Kuhn 2003:84). While the arrangement of houses maximized the available space (Funk 1967:81) the position of the middle longhouses stretching across almost the entire width of the village would have presented barriers to an enemy attempting to move through the village (see Fig. 3).

There were generally only one or two gateways in a palisade. These were often passageways created by overlapping palisade lines that required people to walk single file. These could be easily closed off. Such an entryway was described for the Huron (Sagard-Theodat 1968:9195) and examples have been recovered archaeologically from a number of sites. In 1634, van den Bogaert observed a relatively wide main entryway of 1.1 metres into an Oneida village, but a second entryway was only 0.6 metres across (Gehring and Starna 1988:12). The Iroquois appear to have borrowed "gates" from Europeans, for during the winter of 1692-93 Frontenac encountered Mohawk towns "closed with gates" (O'Callaghan 1858 IX: 550, Abler 1970:28).

VILLAGE STRUCTURE

While palisades served as the first line of a village's defense, they were not always successful. Even with a sturdy palisade, there was always the possibility that the enemy might gain entry into an Iroquois village in a surprise attack. The Jesuit Relation (29:253) of 1646 contains an account of two Huron sentries who fell asleep and were killed by Iroquois warriors. Abler (1970:28) notes that there are numerous accounts in the Jesuit Relations of enemy warriors gaining entry into villages. For the Huron, a warrior gained great prestige if he successfully entered an enemy village at night and captured or killed someone (Finlayson 1998: 19). H_OUSEno. 3

1r rt.

While the Iroquoians seem to have gone to much trouble to build these palisades, to the modern reader it seems almost useless because of their reluctance to use sentries at night (Abler 1970:28).

Figure 2. Longhouse with associated lines of posts from the Nodwell site, Port Elgin , Ontario (from Wright 1974 : 26).

When an enemy was within the village, secondary interior defenses would have become critical.

THE LONGHOUSE

The plans that we have from excavated palisaded villages suggest that houses were often located close to one another. Arranging houses close together would have minimized the effort required to construct a palisade. The restricted area of many hilltop village locations would also have encouraged community nucleation .

It is assumed that female members of a matrilineage lived

in a longhouse with their husbands and children. Nuclear families were arranged along the sides. In addition to construction efficiency, there are advantages to having a household with many members. Iroquois patterns of hunting, fishing, trading, and warfare took men far from the village for extended periods, leaving women, children, and old people behind to carry out other activities, a situation encouraging the formation of both matrilocal residence and extended families (Pasternak et al. 1976; Trigger 1978; Harris 1979:97). In an extended household, many individuals may cooperate in a task, or a variety of tasks may be performed at the same time as needed (Wilk and Rathje 1982, Coupland and Banning 1996:23). Iroquois women cooperated in planting, harvesting, processing, and storing crops in their longhouses (Quain 1961:250, Parker 1968:22-24). In particular, large quan-

Though space was at a premium within these palisaded villages, Sagard stated that an open space was left between the palisade and Huron longhouses to facilitate defense (Sagard-Theodat 1968:92). William Finlayson (1985:407; 1998:20) argues that at the Draper site in Ontario, houses were placed so as to create defensible corridors and reduce access to the center of the village in the event that the palisade was breached. The excavation plan of the Nodwell site, also in Ontario, exhibits lines of posts linking longhouses to palisades (Wright 1974:5;

20

ENGELBRECHT: DEFENSE IN AN IROQUOIS VILLA GE

N

1

100 feet

Figure 3. The Garoga site, Fulton County, New York. This illustration is based on a map drawn by the late Robert Funk (see Funk and Kuhn 2003:84). The dimensions of some lon~houses are estimates. The overlappin~ of some houses with the palisade and with one another suggests rebuilding during occupation of the village. A few 1 foot contour intervals are sketched to indicate the hilltop location. The slope falls off steeply to a creek approximately 45 metres below. gether in a single house with limited access (Rowlands 1972: 456).

tities of seasonably available foodstuffs (crops, fish, venison) were stored within the longhouses. As Lisa Rankin observes " ...rectilinear houses are the easiest to expand when populations increase or more storage area is required" (2000:44).

Gary Warrick (1996:19) considered the possibility that longhouses might be related to defense, but his emphasis was on correlating longhouse length with warfare intensity. After reviewing the evidence, he concluded that: " ...change in Iroquoian house size in prehistory is independent of warfare intensity" (Warrick 1996:19). The argument presented here emphasizes longhouse form rather than length in defensive considerations.

Despite the economic and social advantages longhouses provided, from a Western perspective a longhouse sacrifices both privacy and ease of use (Blanton 1994:32). The existence of a single door at either end and the absence of windows create limited access to the interior. Individuals living toward the center of a longhouse had to regularly pass by other families living closer to a door. Yet more doorways could have been incorporated easily into longhouse design, and some very long longhouses did have side entrances such as the 100 metre long house excavated by Jim Tuck at the Onondaga Howlett Hill site (Tuck 1971: 79). Having more doorways would not have negated the economic or social advantages of longhouse life.

Iroquois men were typically absent from the village for long periods of time: hunting, fishing, trading, conducting diplomacy or waging war. If the Iroquois had lived in nuclear family dwellings, many houses would have lacked an adult male member for much of the year. Even if village residents suddenly became aware of intruders, it is unlikely that women and children or old people scattered in nuclear family residences could have effectively defended their homes against enemy warriors. However, with many families living together, a few men positioned at either longhouse door could defend many families. To get inside a longhouse, enemy warriors would have to enter the outer door, get through the outer storage compartment or vestibule, and then enter the inner door. Once inside, they would still be faced with a series of living compartments. While the primary reason for these compartments was no doubt privacy, they too would also have served to restrict movement. The security a longhouse afforded its inhabitants is reflected in the following oral tradition. "At night none dared leave their doorways lest they be struck down by an enemy's war club. Such

I see three potential advantages to having limited access to longhouses. First, it restricts access to food stored there. However, if this were the critical factor one would expect food to be stored in the center of a longhouse, not in storage compartments at the ends as described in the historic literature. Second, a limited number of doorways would have reduced drafts in winter. Initial entry to a longhouse was into an unheated storage area, which would have served a secondary insulation function. One then entered another doorway to the living area. This arrangement makes sense in winter. Third, there is a defensive advantage for a number of families to live to-

21

FROM THE ARCTI C TO AV ALON: PAPERS IN HONOUR OF JIM TUCK

was the condition when there was no Great Law" (Parker 1968a: 17).

Thus, by the third and fourth decades of the seventeenth century, Iroquoian groups as geographically separated as the Neutral, Seneca, and Mohawk were securing their longhouse doors in some fashion, presumably with European hardware. Gone were the moveable sheets of bark of the traditional longhouse described by Lafitau.

Lafitau, a Jesuit missionary stationed along the St. Lawrence at the Mohawk settlement of Caughnawaga (17121717) describes traditional longhouse doors as follows:

Stealth must have been an important element in surprise attacks within a village. Contrary to the above passage from the Jesuit Relations, it seems likely that chopping through a palisade or the side of a longhouse would have broken "the silence of night". Though chopping would have been made easier by the introduction of the metal axe (Keener 1999:785), the element of surprise would be lost. The same could be said of the use of fire by attackers. Since no suitable non-flammable building alternatives were available to the Iroquois, fire was an everpresent concern for inhabitants of wood and bark longhouses. Under the right conditions, a fire could have spread quickly, but resident warriors in the village would still have had time to rally to the defense after a fire started. In some cases, village dogs may have provided an early warning system preventing intruders from getting close enough to either chop through or bum down a structure.

The doors of the lodges are moveable sheets of bark hung from above, with neither key nor lock. In the past, nothing was closed in Indian houses . When they were gone a long time on a campaign, they contented themselves with fastening their doors with wooden bars to protect them from the village dogs (Lafitau II 1977:22). Lafitau then noted changes that had occurred in the historic period . Others strengthen their lodges at the gables with grossly made planks and install in them wooden doors with bolts bought from the Europeans... (Lafitau II 1977:22). He attributes these changes to the thieving of nearby Europeans, but these changes can also be interpreted as relating to increased security from attack. Such changes were not new for the Mohawk. In 1634, van den Bogaert observed interior longhouse doors made of split planks with iron hinges along with iron chains and bolts (Gehring and Starna 1988:4, 30-31). Kurt Jordan (2002) has suggested that historic period longhouses may be thought of as a hybrid artifact form and the Mohawk evidence suggests that early changes were in the direction of greater security.

Following the epidemic of 1634, longhouses decreased in length (Snow 1994:111). This trend continued through the century as the Iroquois lost population to disease, warfare and out-migration (Guldenzopf 1984:83). Lineages and clan segments responded to these losses by adopting captives and taking in refugees. As a consequence, Snow (1995:304,362) argues that residents of longhouses increasingly formed ad hoc social units. For example, at Caughnawaga in the Mohawk Valley, longhouses are of standardized lengths, suggesting that houses defined the social units, rather than the opposite (Snow 1995:363, 430). The form remained, but the composition changed (Snow 1995:363). Building on this observation, I suggest that the defensive advantage of longhouse life was a consideration in retaining the traditional form in the late seventeenth century. There was still safety in numbers.

Some six or seven years after van den Bogaert 's observation of non-traditional Mohawk doors, the Jesuits found the Neutral fastening their doors against them (J.R. 21:219; Abler 1970:30). Indeed, they no sooner approached a village than from all sides was screamed: "These are the Agwa who are coming" (this is the name they give to their greatest enemies) "fasten your doors;" so that the fathers coming to cabins in order to enter them, according to the rule and custom of the country, found there generally only closed doors; for they were looked upon as sorcerers who carried death and misfortunes everywhere (J. R. 21 :21).

During the mid- ighteenth century, the Iroquois in New York were less involved in inter-group hostilities than previously , and Iroquois settlement assumed a more dispersed character with nuclear family structures scattered across the landscape (Jordan 2002). A few longhouses remained. Moravian missionaries visiting a Seneca village were told by a chief that his house was the largest in town, serving as " ...the meeting place for the Council as well as their fortress" (Beauchamp 1916:74). To use European imagery, the chiefs house was his castle. In the mid-eighteenth century the Iroquois were still aware of the defensive advantages of a longhouse. The longhouse also no doubt continued to be used for its traditional ritual, social, and political functions even though more and more individual families opted for what was

In 1646, the Seneca were also securing their doors. Three Hurons set out to avenge the death of the two sentries mentioned earlier. After twenty day's march, they arrive at Sonnontouans , the most populous of the hostile villages; finding these the cabins closed, they break into one of them at the side, and enter it in the silence and darkness of the night (J.R. 29:253).

22

ENGELBRECHT: DEFENSE IN AN IROQUOIS VILLAGE

probably a healthier and more economically efficient nuclear family home next to their fields and orchards.

ACKNOWLEDGEMENTS I wish to thank Lisa Rankin and Peter Ramsden who invited me to participate in a conference they organized at Memorial University honouring Jim Tuck on his retirement. Jim Tuck's work on the New York Iroquois has long been an inspiration to me and I was honored to be invited. An early version of this paper was given at a session organized by Laurie Miroff and Tim Knapp at the annual meeting of the Society for American Archaeology in Denver in 2002. It was they who suggested looking at the same phenomenon at different spatial scales. Thanks also go to Kurt Jordan, Craig Keener, Keith Otterbein, and Tony Wonderley who commented on an early draft of this paper.

Finally, ancient metaphors may contain wisdom that has been lost. The League of the Iroquois was described by the metaphor of the longhouse, with the Seneca and Mohawk being the keepers of the western and eastern doors, respectively. If the presence of only two longhouse doors was related to defensive concerns, then the title of doorkeeper takes on a military dimension that is generally overlooked.

CONCLUSIONS Iroquoian village structure and longhouse form evolved during a period characterized by inter-group hostility. Modem Iroquois refer to this period as the dark times before the Great Law. It should not be surprising, therefore, that an examination of Iroquois villages and longhouses reveals defensive features. An enemy warrior seeking to enter a longhouse within an Iroquois village would have found his access restricted at a number of junctures. The palisade was the first line of defense, and in the sixteenth and early seventeenth centuries it was often formidable. The placement of structures within some villages would have channeled the movement of enemy warriors through restricted spaces, providing a second line of defense. The absence of windows and the existence of only two doors in a longhouse served to restrict access to the interior providing a third line of defense. An inner longhouse door would have provided a fourth point of restricted access and interior partitions within a longhouse provided still more restrictions to physical movement.

REFERENCES CITED Abler, Thomas S. 1970 Longhouse and palisade: northeastern Iroquoian villages of the seventeenth century. Ontario History LXII: 17-40. Beauchamp, William 1916 Moravian Journals Relating to Central New York, 1745-66. Syracuse, NY: Dehler Press. Blanton, R. E. 1994 Houses and Households. New York: Plenum. Coupland, Gary and E. B. Banning 1996 Introduction: the archaeology of big houses.

People Who Lived in Big Houses: Archaeological Perspectives on Large Domestic Structures, edited by Gary Coupland and E. B. Banning: 1-9. Monographs in World Archaeology no. 27. Madison , Wisconsin: Prehistory Press.

This paper does not argue that defense is the sole explanation for Iroquois village structure and longhouse form. The tight clustering of houses within a palisade was efficient in that it minimized the circumference of this feature. It is also more space- and labor-efficient to construct a single longhouse for many nuclear families than to build many separate houses. The many social, political, and economic advantages of longhouses have long been noted and are not disputed here. Iroquois village and house form met a variety of needs and may be thought of as a compromise between conflicting alternatives. This paper simply argues that defensive concerns played a role in structuring Iroquois village and house form. In the eighteenth century when defense was of less concern, the Iroquois adopted a more dispersed pattern of settlement.

Engelbrecht, William E. 2003 lroquoia: The Development of a Native World. Syracuse, New York: Syracuse University Press. Finlayson, William D. 1985 The 1975 and 1978 Rescue Excavations at the

Draper Site: Introduction and Settlement Patterns. National Museum of Man, Mercury Series, Archaeological Survey of Canada, Paper no. 130. 1998 Iroquoian Peoples of the Land of Rocks and Water, A.D. 1000-1650: A Study in Settlement Archaeology. Vols. 1-4. London, Ontario: London Museum of Archaeology.

23

FROM THE ARCTIC TO AVALON: PAPERS IN HONOUR OF JIM TUCK

1968b Iroquois uses of maize and other food plants. Parker on the Iroquois, edited by William Fenton: 1119. 1910. Reprint. Syracuse University Press.

Funk, Robert E. 1967 Garoga: a Late Prehistoric Iroquois village in the Mohawk valley. Iroquois Culture, History, and Prehistory: Proceedings of the 1965 Conference on Iroquois Research, edited by Elizabeth Tooker: 81-84. Albany: New York State Museum and Science Ser­ vice.

Pasternak, B., Ember, C.R. and M. Ember 1976 On the conditions favoring extended family households." Journal of Anthropological Research 32(2) : 109-123.

Funk, Robert E. and Robert D. Kuhn 2003 Three Sixteenth- Century Mohawk Iroquois Vil­ lage Sites. New York State Museum Bulletin 503: Albany: The State Education Department.

Prezzano, Susan C. 1992 Longhouse, village, and palisade: community patterns at the Iroquois southern door. Ph.D. disser­ tation, SUNY Binghamton.

Gehring, Charles T. and William A. Starna 1988 A Journey into Mohawk and Oneida Country, 1634- 1635: The Journal of Harman Meyndertsz van den Bogaert. Syracuse: Syracuse University Press.

Quain, Buell 1961 The Iroquois. Cooperation and Competition among Primitive Peoples, edited by Margaret Mead :240-281. Boston: Beacon Press.

Grant, W. L. 1907 Voyages of Samuel De Champlain: 1604-1618. New York: Barnes and Noble. 1959. Reprint.

Rankin, Lisa 2000 Interpreting Long -term Trends in the Transi­ tion to Farming: Reconsidering the Nodwell Site, On­ tario Canada. BAR International Series 830 Oxford, England: BAR Publishing.

Guldenzopf, David. 1984 Frontier demography and settlement patterns of the Mohawk Iroquois. Man in the Northeast 27: 79-94.

Ritchie, William and Robert Funk 1973 Aboriginal Settlement Patterns in the North­ east. New York State Museum and Science Service, memoir 20. Albany: The State University of New York, The State Education Department.

Harris, Marvin 1979 Cultural Materialism: The Struggle for a Sci­ ence of Culture. New York: Random House.

Rowlands, M. 1972 Defence: a factor in theorganization of settle­ ments. Man, Settlement and Urbanism, edited by Pe­ ter J. Ucko, Ruth Tringham, and G. W. Dimbleby: 447-462. London: Duckworth.

Jordan, Kurt A. 2002 The Archaeology of the Iroquois Restoration: Settlement, Housing, and Economy at a Dispersed Se­ neca Community, ea. A.D. 1715�1754" Ph.D. disser­ tation, Department of Anthropology, Columbia Uni­ versity.

Sagard-Theodat, G. 1968 Sagard's Long Journey to the Country of the Hurons (1632). Edited by G. M. Wrong. New York: Greenwood Press.

Keener, Craig S. 1999 An ethnohistorical analysis of Iroquois assault tactics used against fortified settlements of the North­ east in the seventeenth century. Ethnohistory 46( 4): 777-807.

Snow, Dean 1994 The Iroquois. Publishers

Lafitau, Father Joseph Fran9ois 1974, 1977 Customs of the American Indians Com­ pared with the Customs of Primitive Times .. 2 vols., edited and translated by William Fenton and Elizabeth Moore. [1724]. Toronto: The Champlain Society.

Cambridge, MA:

Blackwell

1995 Mohawk Valley Archaeology: The Sites. The Institute For Archaeological Studies, University at Albany, SUNY.

O'Callaghan, E. B. 1853-1887 Documents Relative to the Colonial History of the State of New York. 15 vols. Albany: Weed, Parsons, and Co.

Thwaites, Reuben Gold ( editor) 1896-1901 The Jesuit Relations and Allied Docu­ ments: Travels and Explorations of the Jesuit Mis­ sionaries in New France, 1610 -1791. 73 vols. Cleve­ land: Burrows Brothers.

Parker, Arthur C. 1968a The constitution of the Five Nations or the Iro­ quois book of the Great Law. Parker on the Iroquois, edited by William Fenton. 1916. Reprint. Syracuse University Press.

Trigger, Bruce 1978 Iroquoian matriliny. ologist 48, nos. 1-2: 55-65.

24

Pennsylvania Archae­

ENGELBRECHT: DEFENSE IN AN IROQUOIS VILLA GE

Tuck, James A. 1971 Onondaga Iroquois Prehistory: A Study in Settlement Archaeology. Syracuse: Syracuse University Press.

Wilk, R. and W. Rathje 1982 Archaeology of the household: building a prehistory of domestic life. American Behavioral Scientist 25:617-40.

Warrick, Gary 1988 Estimating Ontario Iroquoian village duration. Man In The Northeast 36: 21-60.

Wright, J. V. 1974 The Nodwell Site. National Museum of Man, Archaeological Survey of Canada, Mercury Series Paper no. 22. Ottawa

1996 Evolution of the Iroquoian longhouse. People Who Lived in Big Houses: Archaeological Perspectives on Large Domestic Structures, edited by Gary Coupland and E.B. Banning: 11-26. Monographs in World Archaeology No. 27. Madison, Wisconsin: Prehistory Press.

25

FROM THE ARCTIC TO AVALON: PAPERS IN HONOUR OF JIM TUCK

26

BUT ONCE THE TWAIN DID MEET: A SPECULATION ABOUT IROQUOIS ORIGINS Peter Ramsden Memorial University

American anthropologists drew from this is that the Iroquois must therefore have come into the otherwise Algonkian-speaking Northeast from somewhere else, relatively recently. When this was first proposed, American archaeology was in its infancy, and archaeologists confidently expected that they would be able to find the exotic homeland of the ancestral Iroquois, and trace their migration into the lower Great Lakes. Unfortunately, archaeologists have done exactly the opposite, and each of the last 6 decades has only made it more and more clear that Iroquois culture has its roots in the lower Great Lakes, and that nothing resembling it can be found anywhere else.

DISCLAIMER

I must emphasize at the outset that this paper is a highly speculative consideration of the issue of Iroquoian origins: long on outlandish ideas, very short on rigorous analysis. In a sense, it represents a kind of 'coming out', in that it is a more explicit statement of some ideas that I have been hinting at for years (Ramsden 1978, 1988, 1991, 1992c). It is offered as something along the lines of "The Paper I've Always Wanted to Give", and also in the spirit of a tribute to a great archaeologist and great friend, Jim Tuck. It is the kind of thing that Jim and I sometimes end up talking about somewhere around about a quarter to six on a Friday, and is perhaps indicative of his ability to inspire people to think in highly creative, if sometimes inconclusive ways about what happened in the past. A great deal of my enthusiasm for Iroquoian archaeology was inspired by reading Jim's Onondaga Iroquois Prehistory, and, whether he likes it or not, that enthusiasm led to some of the thoughts presented here.

Now, having demonstrated the archaeological origins of the Iroquois to their own satisfaction, archaeologists generally felt free to dismiss or ignore the original linguistic problem posed by the Iroquois as being obviously wrong, or obviously unimportant. I have never believed that it was either. I think the linguistic position of the Iroquois does need to be explained, and the in situ theory fails to do this. In fact, it is almost an axiom of the in situ theory that it does not need to do it. I also think, in the light of the lack of Iroquoian linguistic similarities to their neighbours, that the obvious common archaeological antecedents that they share with those neighbours also needs to be explained. The in situ theory does not do this either.

IDEAS OF IROQUOIS ORIGINS

The origin of this paper is a conviction I developed when I was still a Ph.D. student, that the position of the Iroquois in the Northeast is neither as anomalous as the linguists used to maintain, nor as simple and predictable as the promoters of the archaeological "in situ'' theory would have us believe. The obvious archaeological continuity inherent in the picture of the development of the Iroquois does not explain their apparent linguistic isolation in a sea of Algonkian speakers; nor can that linguistic isolation contradict the obvious fact that Iroquois culture has ancient roots in the historic Iroquois homeland, and nowhere else.

I am not trying here to demolish the in situ theory, or even to do it any serious damage. What the in situ theory can explain, it explains perfectly well: that is, the origin and development of an archaeological culture through time. What it cannot explain, however, and what I have always maintained that archaeology can never explain, is the behaviour of language. And this is my essential point: what is needed to solve the old linguistic puzzle of the Iroquois is a new linguistic explanation. An archaeological explanation simply cannot do it.

The explanation of Iroquois origins advanced by the "in situ" theory is a simple one: they have always been here. 'Always' in this case meaning as far back as it matters. However, I think that the only credible reading of the archaeological record of the northeast makes it clear that the northeastern part of the continent, to well south of the Iroquois area, was settled in the post-glacial period by the ancestors of the Algonkian language family, and they are still here. If the Iroquois had always been here, and that were the whole story, they would be Algonkians. Either the Iroquois are recent arrivals as the linguists once maintained, or they have always been here but that isn't the whole story.

BECOMING IROQUOIS

My Iroquoian work has focused on the Hurons of southcentral Ontario, and it is with considerable trepidation that I even attempt to broaden my scope to include the New York Iroquois (see Figure 1). In the course of my th work on late 16 century Hurons in the Balsam Lake area of southern Ontario, I became convinced that the only clear demarcation between the Hurons and their Algonkian neighbours was a linguistic one; the cultural, political, economic and social boundaries were very blurred indeed (Ramsden 1988). Taking this further, I

Algonkian and Iroquois are members of entirely different linguistic families, and the conclusion that early 27

FROM THE ARCTIC TO AVALON : PAPERS IN HONOUR OF JIM TUCK

plan to argue in a forthcoming book that the Hurons of ethnohistory were in fact an ethnic amalgam, comprising people who were ethnically and linguistically Huron, some Algonkians from slightly further north in the Canadian Shield, a few Neutrals from southwestern Ontario, and St. Lawrence Iroquoian refugees from the east end of Lake Ontario and the St. Lawrence valley (Ramsden 1992b). The important point about this, for present purposes, is that this ethnically and linguistically heterogeneous society presented itself to outsiders, and became known to history, as a Huron society, unified within a Huron territory, and speaking Huron language. The explanation for this, in this case, is reasonably simple: for unique historical reasons, the politically and economically dominant members of the Huron confederacy were in fact Hurons, and thus the confederacy became the Huron confederacy, and the language of the confederacy, particularly in connection with politics and trade, was Huron (Ramsden 1992a).

So what relevance does this have to the original problem of the origins of the Iroquois as a whole? In a nutshell, I believe that this reconstruction of the 'huronization ' of the Hurons can be used as an example in miniature of how the Iroquois may have become the Iroquois. The New York Iroquois are by no means homogeneous, but what has impressed anthropologists and historians from the very beginning has been the high degree of political and social integration among them. And in spite of the cultural and linguistic differences between the five tribes, they were content, at certain political and economic levels, to present themselves to outsiders as "The People of the Longhouse", and other native groups in the Northeast , and European colonists, saw fit to refer to them as a collective whole: The Iroquois. And this view of the Iroquois as a unified whole, which they themselves were at pains to construct and project , has created a thorny archaeological problem.

The Huron confederacy, as known in historic Huronia, lasted a mere 50 years, but even in that short time its cosmopolitan nature became masked, or at least effectively hidden from outsiders. And given another 50 years, it seems doubtful that there would have been anybody in the Huron confederacy who spoke anything other than Huron, or considered themselves to be anything but Hurons. This much we can know, or at least can surmise, because there are historical records pertaining to the Hurons at this time. If the Hurons had been contacted by Europeans 100 years later than they actually were, we might very well have no inkling of the diversity that once existed among them, and it would be a very astute archaeologist who might guess at it based on the archaeological evidence alone.

The Iroquois, whether in the smaller sense of the Five Nations or in the broader sense of the Iroquoian speaking people of the lower Great Lakes, have traditionally been regarded as constituting a single cultural entity. Prior to the 1950s, several migration theories had been advanced to account for the presence of the Iroquois in the Northeast. These theories cited several quite disparate sources of Iroquoian migration, and generally each was based on a piece of ethnohistorical, oral or mythological evidence. But strangely, the proponents of each theory believed that their particular theory must explain the origin of all Iroquois, in spite of the equally good cases to be made for each of the other theories. It doesn't seem that anybody thought it was important to consider just where any particular piece of information came from: which Iroquoian tribe, which village, or which sociopolitical group within a village.

.--.. N 1pp1ssmgs

Mohawk Delaware

Figure 1. Lower Great Lakes and Eas tern seaboard, showing groups mentioned in the text. 28

RAMSDEN: BUT ONCE THE TWAIN DID MEET

and somewhat subjective, but I believe it nonetheless has some validity on a general level. A similar generalization would be that in western Iroquoia, pottery motifs are predominantly executed by incising, whereas in eastern Iroquoia there is a much more frequent use of other decorative techniques, such as dentate stamp and cordwrapped stick impression. In a variety of ways, the pottery of the western Iroquoians has a Mississippian look about it, and the pottery of the eastern Iroquoians much less so.

Similarly, when archaeologists began seriously to investigate the problem of Iroquois origins, it is not unreasonable that they should have presumed that all the Iroquoians would have a single origin, be that in the lower Great Lakes or elsewhere. MacNeish took this as an a priori assumption in his famous book on Iroquois pottery types, so it was in fact impossible for him to come to any other conclusion. But even a brilliant and insightful young scholar from Syracuse University, working a decade and a half after MacNeish, felt that, having demonstrated in detail the in situ origins of the Onondaga, he had, in his words, "once and for all", demonstrated the in situ origins of "the Iroquois" as a whole (Tuck 1971). This is not to criticize in any way the work of MacNeish or Tuck. It is simply to point out the terribly constraining effect that a 300 year old perception of Iroquoian ethnicity has had on our ability to think archaeologically about Iroquoian origins.

My second point was that Iroquoian material culture is not unique to the Iroquois. It has been long noted that the Mohawk and their Algonkian neighbours, the Munsee Delaware, share aspects of their archaeological cultures, and appear to be the products of a common late Owasco ancestry, dating back some thousand years (Ritchie 1969:300). What this really amounts to is that the eastern Iroquois share pottery styles and other cultural traits with their more easterly Algonkian neighbours, and have done for much of their history (see Petersen: this volume). Now it is not unduly strange that two neighbouring peoples should be culturally similar, but what is puzzling is why they should be so linguistically distinct given that they can be regarded as having a common archaeological origin, and have shared so much of the rest of their culture throughout prehistory.

Very little consideration has been given to the possibility that the Iroquois were not originally a single group, or that they might have multiple origins. Early on in my own thinking about the Iroquois, it dawned on me that they could be divided into two broad groups, one in the west, essentially distributed around Lake Erie, and another in the East, distributed around Lake Ontario and the lower St. Lawrence Valley. This hazy distinction that I drew in my mind was based on a few archaeological traits, including some pottery characteristics and burial patterns. Alanson B. Skinner had, unbeknownst to me at the time, already proposed a similar idea in 1921. I still believe that the idea is worth reviving, and reconsidering, and I also believe that it has vital implications for the investigation of Iroquoian origins.

A similar example is that of the Hurons and their Algonkian neighbours and trading partners, the Nippissing. In Jim Wright's classic formulation of the Ontario Iroquois (Wright 1966), he identified certain strata at the Frank Bay site on Lake Nippissing as being (pre-contact) Southern Division Huron, and its upper levels as historic Huron, in which he was in agreement with Ridley (1954). His reason for doing this was that the material culture of the Frank Bay site was indistinguishable from that of the Hurons and their ancestors since the very beginning of the Iroquoian period, at least insofar as that material culture was understood at the time. Wright's explanation for this was that the Frank Bay site was a far-flung Huron fishing camp. More recent, and almost certainly more correct, accounts of the site interpret it as the summer camp or village of the Nippissings (e.g. Fox 1990b). So it looks as though this is another example of two peoples who were linguistically quite distinct in the historic period, but archaeologically indistinguishable throughout their history. I should point out in fairness that more recent analyses recognize that the assemblages of the Nippissings and Hurons can be distinguished on a detailed level, as can those of the Mohawk and the Delaware. On the other hand, so can the assemblages of different Iroquoian groups be distinguished from each other, and by using similar kinds of criteria. The important point here is that the degree of archaeological difference between the Iroquois and their Algonkian neighbours is not of a higher order than that between different Iroquoian groups, such as between the Hurons and the Five Nations.

There are two important points with regard to the archaeological pattern of the Iroquois. The first is that the Iroquois are not archaeologically homogeneous. The second is that no element of Iroquois archaeological culture is unique to the Iroquois. A few examples will illustrate these points. One of the archaeological attributes that seems to me to divide the Iroquois into eastern and western branches is the presence of high-collared pottery. This is a characteristic of the St. Lawrence Iroquois, the Mohawk, the Cayuga, some of the Hurons, and to a lesser extent the Onondaga. In general, high collars are an eastern feature in Iroquois material culture. On the other hand, the pottery of the Erie, Neutral and Seneca is characterized almost exclusively by low collars. A difference in vessel collar height may not seem like much to write home about, but its geographic distribution can hardly be coincidental. Staying with pottery for a moment, I would suggest that it is also true that ceramics in the eastern half of Iroquoia tend to be decorated with more complex motifs, and with a more precise style of execution, than those in western Iroquoia. No doubt this is a gross over-simplification,

29

FROM THE ARCTIC TO AVALON : PAPERS IN HONOUR OF JIM TUCK

By contrast, in southwestern Ontario, the far western part of Iroquoia, the Neutrals are archaeologically distinct from their Algonkian speaking neighbours , and have recognizably separate archaeological origins. Moreover, in Southwestern Ontario, the evidence for the derivation of the early Iroquois Glen Meyer phase from the preceding Middle Woodland is equivocal (see Fox 1990a; Ferris 1999). This is in contrast to more easterly parts of Iroquoia, where the initial Iroquoian Pickering and Owasco phases have clear antecedents in the Middle Woodland Point Peninsula.

economic transactions were conducted was Iroquoian, as it would certainly be if the newcomers were the more prestigious or more powerful group, then Iroquoian is what the Algonkian neighbours would have to speak. While a new lingua franca might exist alongside native languages for centuries, it need not take more than a generation or two for tre language of trade to become the language of choice, as has happened in several other instances. The historian Christopher Ehret has documented this process among horticultural small-scale societies in eastern Africa (Ehret 1988). Ehret notes that in such small-scale societies, ethnicity and its concomitant identification with a particular language are not the abstract compulsives that they are in more complex societies. He states that ethnicity in these societies is "a situational choice ... a fluid reality, able to shift and change" (1988:569).

So perhaps it is becoming clear where I am going with all of this. I propose that the several peoples subsumed under the term "Iroquois", or "Iroquoians" do not all have the same origin and history. In a nutshell, I would propose that a large eastern group of Iroquoians, minimally including the Mohawk, the St. Lawrence Iroquois, the Onondaga, and those Hurons who lived prehistorically in the eastern part of the Huron territory, really do have an in situ origin, and were originally part of the Algonkian-speaking Northeast, as their archaeological affinities to other northeastern Algonkian peoples attest. The western group of Iroquoians, including the Neutral, the Erie, and perhaps the Seneca, may be more recent arrivals, or at least are culturally dominated by a group of more recent arrivals. I believe that the origins of the newcomers lie in the middle or upper Mississippi valley, or at least in the northern part of the vast Mississippian empire. I further believe that they were the instrument of the introduction of Iroquoian language, and many aspects of Iroquoian culture, into the Northeast, in relatively recent times.

Ehret goes on to describe the kind of process I envision for the 'iroquoianization ' of the Lower Great Lakes. The impetus for the wholesale adoption of a foreign language can be as simple as a slight demographic or prestige inequality between neighbouring communities (Ehret 1988:570). Moreover, it need not involve the relationship between whole communities; ethnic and language shifts in a community can be precipitated by "the movement of a band of high-rank men and their followers into another region because the chances for gaining power over the local cultivators seem favourable there" (Ehret 1988:573). The process, moreover, is self-perpetuating. The addition of a neighbouring community to the new ethnic-linguistic group further increases its perceived power and prestige, and thus increases its appeal to further-flung communities:" ... [the new group] now has a tradition of expansion its offspring may seek to build on, and a set of expectations among its neighbours as to where their relations with that society may lead" (Ehret 1988:570). In other words, by attracting surrounding communities to join it, the new 'ethnic' group comes to be perceived by others as powerful, fearsome, and worth joining with. This is beginning to sound like the Iroquois of the historic period.

THE LANGUAGE QUESTION

The problem of how and why the bulk of the Iroquoians became transformed from Middle Woodland Algonkians into Late Woodland Iroquoians is, I admit, a complicated one, but not an insurmountable one. In essence, I believe that the Iroquoian language introduced or first adopted by the western Iroquoians was later gradually adopted by those peoples who eventually became the eastern Iroquoians. Examples of this process are not difficult to find around the world, and perhaps the language in which I am now writing is the easiest one to point to. The question of why this should have happened is the ticklish one, and at this point I can only offer speculations. A group of people arriving in the Lake Erie basin from the Mississippi valley might, by their very foreignness and their ties to the Mississippian empire or to 'foreign ' regions, be viewed with awe, or at least enjoy a degree of prestige, perhaps even of power. This would be even more true if they brought with them knowledge of a new mode of economic production, such as maize horticulture. And if they also brought with them trade ties to the Mississippi and its hinterland, neighbouring groups of Late Woodland Algonkians might be anxious to have a part of that action. If , moreover, the language in which

In any kind of exchange network where the ultimate benefit is prestige and power, rather than mere material wealth, the crucial thing is to maintain its exclusivity. The important point about such a network is that some people are in it (the lucky and prestigious ones) and everybody else is out of it (the unlucky and pathetic ones). The boundaries of this network cannot be allowed to become too blurred, or the benefit is lost. And if speaking Iroquoian was one of the outward signs of membership, then that boundary would have to be maintained. Ehret also emphasizes that the resultant changes in ethnicity and language usually take place without any major changes in everyday material culture "except those

30

RAMSDEN: BUT ONCE THE TWAIN DID MEET

1990b The Odawa. In The Archaeology of Southern Ontario to 1650 A.D., edited by Chris J. Ellis and Neal Ferris: 457-474. London Chapter, Ontario Archaeological Society, Occasional Publication Number 5.

items which symbolize an essential acceptance of or claim to the prestigious or desirable ethnicity" (1988:571). In the latter context, the Iroquoian pipe and smoking complex comes immediately to mind. Ehret also notes that a frequent archaeological indicator of such ethnic and language shifts is a change in ceramic decoration (1988:572), and he specifically mentions one instance in which this took the form of the adoption of a new decorative tool. Here, again, one cannot help but think of the preponderance of incising on Iroquoian pottery in contrast to that of neighbouring groups.

Ramsden, Peter 1978 The In-Situ Theory Revisited - Again: If the In Situ Theory is the Answer, What was the Question? Paper delivered to the McMaster Anthropology Society Annual Symposium. Hamilton.

The result of this hypothetical process would look very much like the Iroquois: a group of societies that are united primarily by a common language, political ties, and some binding ritual practices; a small group of whom may be recently arrived foreigners, but the majority of whom are in every other respect just like their now linguistically distinct neighbours with whom they share a long cultural tradition.

1988 The Huron-Algonkian Interface: the View from Haliburton and the Kawartha Lakes. Paper delivered at the Annual Meeting of the Canadian Archaeological Association, Whistler, B.C. 1991 Reconciling Archaeology, History and Anthropology: An Example from Southern Ontario. Seminar, Department of Archaeology, University of Cambridge, UK 1992a L'Archeologie d'une Federation. Dept. d'Anthropologie, Univ. de Montreal.

ACKNOWLEDGEMENTS

As I indicated at the beginning, I have to thank Jim Tuck for a great deal of inspiration, about a great many things, over the years. Needless to say, I do not hold Jim responsible for any of the specific ideas advanced here, although I would like to be able to. I also thank all my other archaeological colleagues for all the discussions and inspiration: in the context of this paper I should particularly mention, although they may not understand why, Neal Ferris, Allen Tyyska, Bruce Trigger, and Marek Zvelebil. Special thanks to Lisa Rankin for encouraging me to revise the paper for this volume.

Seminar,

1992b A Re-Examination of Huron Ethnicity. Paper delivered at the annual meeting of the Society for American Archaeology, Pittsburgh. 1992c Regards sur l 'hypothese de l' origine in situ des Iroquoiens. Recherches Amerindiennes au Quebec XXIl(4):19 -23 Ridley, Frank 1954 The Frank Bay site. 20(1):40-50

American Antiquity

Ritchie, William A. 1969 The Archaeology of New York State. Natural History Press.

REFERENCES CITED

Ehret, Christopher 1988 Language change and the material correlates of language and ethnic shift. Antiquity 62(236):564-573.

Skinner, Alanson B. 1921 Notes on Iroquois Archaeology . Indian Notes and Monographs. Museum of the American Indian, Heye Foundation: New York

Ferris, Neal 1999 Telling tales: interpretive trends in southern Ontario Late Woodland archaeology. Ontario Archaeology 68: 1-62

Tuck, James A. 1971 Onondaga Iroquois Prehistory : A Study in Settlement Archaeology. Syracuse University Press , Syracuse, NY.

Fox, William A. 1990a The Middle to Late Woodland transition. In The Archaeology of Southern Ontario to 1650 A.D., edited by Chris J. Ellis and Neal Ferris:171-188. London Chapter, Ontario Archaeological Society, Occasional Publication Number 5.

Wright, J. V. 1966 The Ontario Iroquois Tradition. Museum of Canada, Bulletin 210.

31

National

FROM THE ARCTIC TO AV ALON: PAPERS IN HONOUR OF JIM TUCK

32

A LABRADOR ARCHAIC LONGHOUSE SITE AT SANDY COVE, LABRADOR: LIFE ON THE SOUTH SIDE OF GROSW ATER BAY Lisa Rankin Memorial University

In 2001 work commenced on the Porcupine Strand Archaeology Project located in south-central Labrador. The regional focus of the research stretched from Sandwich Bay in the south to the southern shore of Groswater Bay in the north, an area nearly 90km long that contains a number of unique landscapes that had never been subject to any systematic archaeological investigation (Figure 1) (Fitzhugh 1981, 1984, 1986; Stopp 1997). A regional study of the area afforded the opportunity to develop a culture history of south-central Labrador which could then be compared to the extensive work already undertaken in the Labrador Straits much further south (McGhee and Tuck 1975; Tuck and McGhee 1976) and to the comprehensive study of the north shore of Groswater Bay undertaken by Fitzhugh (1972, 1975).

site of Upper Sandy Cove 3 (GbBi-7) (Figure 1). Although neither the oldest nor most recent of the Labrador Archaic sites located by the project, Upper Sandy Cove 3 offered an opportunity to examine a rare feature on the Porcupine Strand and in Labrador generally - house structures dating from the Sandy Cove complex of the Labrador Archaic (6000-4700 BP).

THE LABRADOR ARCHAIC SANDY COVE COMPLEX

The Sandy Cove complex of the Labrador Archaic tradition was first identified by Fitzhugh (1972), and its description was expanded by him in several subsequent publications (1972, 1975a, 1977, 1978, 1984, In Press). He originally defined the Sandy Cove Complex after collecting and testing five sites on the north shore of Groswater Bay during his 1969 research in Hamilton Inlet (Figure 1). By 1973, after further research on the north shore, he had identified twelve sites belonging to the complex situated near the mouth of Groswater Bay (Fitzhugh 1978). Although extensive research on the Labrador Archaic has been undertaken in other regions of Labrador, no Sandy Cove complex sites have been identi-

The first phase of the project focused on the systematic pedestrian survey of the mainland shoreline (inland from the present shore to as much as two kilometres), as well as the survey of adjacent islands. In total, over 125 sites were located and indicated the substantial use of the Porcupine Strand region by various populations over a 7000 year period. Beginning in 2003 sites dating to different periods and associated with different culture groups were selected for excavation, including the Labrador Archaic

t

10km

Figure 1. Map of Groswater Bay showinf;{locations of Fitzhuf;{h's Sandy Cove Complex sites, and Upper Sandy Cove 3 site. Inset shows area of Porcupine Strand Project in Labrador.

33

FROM THE ARCTIC TO AV ALON: PAPERS IN HONOUR OF JIM TUCK

From this new data he concluded that the regular placement of hearths and artifacts was indicative of long rectangular structures ranging in length from 12 to 16m (Fitzhugh 1984). In a very recent article Fitzhugh (In Press) notes that these rectangular structures contain two to three hearths and average four metres in width, and suggests that they were probably multi-family dwellings.

fied outside of Groswater Bay, suggesting that this complex is a distinct regional variant. However, both chipped and ground stone artifacts associated with these sites suggest origins in the earlier Strait of Belle Isle Archaic cultures (Tuck n.d). Given that the Sandy Cove complex sites date between 6000 and 4700 BP (Fitzhugh 1975a:122), this complex may represent a continuity from the central Labrador Hound Pond complex of the Labrador Archaic (7500-7200 BP) that migrated north from the Labrador Straits (Fitzhugh 1977, 1978; Tuck n.d.).

The landscape inhabited by the Sandy Cove population was probably forested by white spruce, fir and paper birch as pollen diagrams show that these species were present along the north shore of Groswater Bay between 5500 and 6000 BP, (Fitzhugh 1975b, 1978, In Press; Fitzhugh and Lamb 1984; Jordan 1975). The outer bay locations of these sites are considered optimal for both caribou and sea mammal hunting. Fitzhugh and Lamb (1984:363) indicate that the central coast of Labrador probably had an abundant and diverse species population at this time given the species rich and structurally diverse vegetation. The sites themselves provide further clues to their seasonality and function. None of the original Sandy Cove complex sites had formal house structures or burial features, leading Fitzhugh to suggest brief summertime occupations, perhaps annual (1972, 1975a, 1975b). No faunal remains were present at the sites, but the large size of many of the points suggests that the occupants of the site were hunting large game (Fitzhugh 1972). Given their proximity to the shoreline it is reasonable to suspect that fish and perhaps sea mammals could be taken from the Sandy Cove sites. Faunal evidence has been found at the somewhat more recent Rattler's Bight site located 20 km away. There, seal, bear, duck, beaver and otter were all hunted and there is no reason to believe that these resources were not similarly available at the Sandy Cove sites (Fitzhugh 1975a). In general, Fitzhugh (1972, 1975b) categorizes the Sandy Cove complex sites as belonging to a relatively mobile hunting and gathering population that had a specialized maritime adaptation in summer months. He further speculates that the population traveled further inland for wint r hunting (Fitzhugh 1972, 1975b). No interior sites associated with this complex have been located to date, but the movement of the population to other locations might be suggested by the fine grained cherts from northern Labrador and other sources that have been recovered from most of the sites.

Only seven Sandy Cove complex site collections have been published in detail (Fitzhugh 1972, 1975a). Assemblages range in size from one to fifty artifacts, and are characterized by tapered stem points, convex based bifaces, leaf shaped bifaces (sometimes referred to as knives), pieces esquillees, discoidal hammerstones, unifacial flake knives, and ground stone tools including tapered points and bipoints, single edged knives and celts (Fitzhugh 1972, 1975a, 1975b; Hood 1981; Tuck n. d.). Chipped stone tools far outnumber ground stone tools. Locally available quartz, red quartzite and slate are the dominant raw materials, but small quantities of fine grained cherts from unknown quarry sources, and even smaller quantities of Ramah chert from northern Labrador are present. The use of Ramah chert appears to be restricted to point production (Fitzhugh 1972). Other raw materials are also associated with specific industries. For example, wedges tend to be made of quartz while slate is used for celt production (Fitzhugh 1975a:122). No bone or hide working tools have been associated with the complex and, probably due to factors of preservation, no organic tools have been recovered. The earliest sites in the Sandy Cove complex are found on beach terraces with elevations greater than 15m above sea level, while the most recent sites can be found on terraces of at least 1Om asl. In light of a recently constructed sea level curve for northern Groswater Bay (Clarke and Fitzhugh 1990), the absence of Sandy Cove sites below 10m (Fitzhugh 1975a) suggests that these sites were 1 to 3 m above the active beach during their occupation. Conifer charcoal occurs in abundance at these sites and indicates a local spruce forest cover as early as 6000 BP that was much thicker than that of the present day (Fitzhugh In Press). No significant house structures are associated with any of the twelve sites recorded by Fitzhugh, but his 1972 publication does refer to a circular arrangement of flat stones with a possible hearth feature in the centre at Sandy Cove 1 (GcBk-1). Furthermore, areas of charcoal mixed with red ochre were present at some of the other sites (Fitzhugh 1972:92-93). Later research concluded that hearth pits were often centrally located within lithic scatters at Sandy Cove sites (1975a:119). In 1983 Fitzhugh (1984:13) returned to several Sandy Cove complex sites on the north shore of Groswater Bay and noted that many of the sites had closely spaced artifact and debitage loci surrounding shallow rock-lined hearths and paralleling the terrace fronts.

Since 1984 Fitzhugh (1984, 1985, In Press) and, more recently, Hood (1993, 1995, 2002) have drawn on the data documented at Sandy Cove complex sites to develop models examining changes in Labrador Archaic social relations over time. The rectangular alignment of hearth features and associated material culture found at Sandy Cove complex sites has been assigned a significant placement in a trajectory of increasingly complex social relationships. These features mark a shift (if a seasonal one) from the small, circular single family boulder pit dwellings used exclusively by earlier groups (and only sporadically later) to longer communal house forms, suggesting an increase in the size of local groups from nuclear families to aggregated kin groups (Fitzhugh In

34

RANKIN: A LABRADOR ARCHAIC LONGHOUSE SITE

Press; Hood 1993). This difference in house form may represent a difference in seasonal activities, but given that Labrador Archaic longhouses continued to increase in size and number of occupants up until ca. 4000-3500 BP, it may also mark a distinct transition in the way people organized themselves and viewed the world (Hood 1993, 2002). Hood (1993) has proposed that the segmented structure of the Labrador Archaic longhouse preserves the autonomy of the nuclear family while at the same time creating the sense of a collective group. The physical separation of families within these structures provides a mechanism by which families are able to fission throughout the year. Nevertheless, the act of creating the longhouse form can be viewed as a way of creating a meaningful space that links families into a corporate unit in a material fashion through the sharing of domestic space (Hood 2002:248). Both Fitzhugh (1985) and Hood (1993) suggest that our understanding of these household relationships could be expanded through the spatial inve stigation of tool use and activity areas within houses.

Cove sites on the southern shore of Groswater Bay (Figure 1). The site was discovered in 2002 during a pedestrian survey of the northern segment of the Porcupine Strand research area. It is located on a narrow bedrock peninsula extending northward from the southern shore of Groswater Bay, less than 10km west of the present bay mouth (Figure 2). It is situated on a raised beach terrace 15 m asl in an area of active aeolian deposits of large sand dunes and basin shaped blowouts. The sand dunes of Upper Sandy Cove overlie raised beach sediments, indicating that the ancient beach terrace was formed and vegetated prior to the onset of aeolian activity in the area. The dunes and blowouts are therefore not related to the timing of sea level fall in this region (Smith 2005 :79-80). The Upper Sandy Cove Peninsula emerged first as an island approximately 7000 years ago when recently deglaciated land rose and sea levels fell (Smith 2005). A Radiocarbon AMS date of 5150±40 (Beta 198381) (cal. 5760-5940 B.P.) was obtained from a hearth feature. So by the time Upper Sandy Cove 3 was occupied by the Labrador Archaic the peninsula would have had much the same form as that visible today (Rankin 2005; Smith 2005). The site's occupants would have been camped on the surface of a stable, vegetated, marine sediment beach approximately two to three metres above the shoreline (Clarke and Fitzhugh 1992:299). The aeolian dune activ-

UPPER SANDY COVE 3

Despite its name, the Upper Sandy Cove 3 site is located approximately 30 km to the south of the original Sandy

longhouse

•

artifact or artifact cluster

0

boulder Edge of blowout

· House 1

..

.......... ..

l

.. . . ...

. ..... ..

--~ --

0

10

m

Figure 2. Map of Upper Sandy Cove 3 showing blowout areas, surface artifact scatter and longhouse structures. Aerial photograph (inset) shows general location of Upper Sandy Cove 3 (black dot) and quartz quarry (white dot).

35

FROM THE ARCTI C TO AV ALON: PAPERS IN HONOUR OF JIM TUCK

flated) or partitions, but was littered with fire-cracked rock. House 2 was also 9m x 4m and contained a single central hearth feature that was almost entirely deflated, and large quantities of fire-cracked rock. House 3 was much bigger, measuring 20m x 4m in total. It contained two rock lined hearths, one near each end of the house, and these were excavated in 2003.

ity which made Upper Sandy Cove 3 so visible to the survey team is a much more recent landscape development that occurred long after site abandonment. It is probably related to the cyclical deforestation of the region which may have been brought about by forest fires, decreased precipitation or animal grazing in the region. Examination of buried paleosols in the Upper Sandy Cove dunes suggests several periods of dune formation followed by surface stabilization beginning 2590±60 BP and continuing to the present day (Smith 2005).

Artifacts and debitage were widely scattered throughout the site. Houses 1 and 2 were almost devoid of cultural material (one artifact was collected from House 1 and three from House 2), and in this respect were similar to the one formal structure identified by Fitzhugh (1972). The highest density of remains was recovered between Houses 1 and 2. Within this area were several clusters of quartz, slate and Ramah chert debitage. The area was also peppered with large amounts of fire cracked rock, but no hearth features were discemable.

Nevertheless , the aeolian history of the region has had an impact on the preservation of the site. Much of the site (approximately 80%) is presently situated at the bottom of a deflated blowout feature (Figure 2). As a result, artifacts originally deposited on the relict beach may have shifted through time with the erosion of underlying sands. As well, artifacts and features have been either severely deflated or have tumbled into the dune basin from the eroding edges as much as two metres above the dune basin. Without doubt, the resulting shifts in artifact position pose problems when attempting to establish firm contexts for, and associations between, artifacts and features at the site. However, there is a clear association of material types from specific lithic production events throughout the dune area that suggest that most of the movement has been minimal, or at the very least comprehensible. A further impact to the site is much more recent. During the later half of the twentieth century a US Army camp was established within one kilometre of the site and spent ammunition and tank tracks are visible around its southern perimeter. This activity may well have destroyed or displaced portions of the site.

A total of 132 artifacts and 712 flakes were surface collected from the site in 2002. Approximately 40% of these were collected from House 3, and 45% of the remainder between Houses 1 and 2. In 2003 we returned to excavate House 3. It contained 96 flakes and 20 artifacts, bringing the total numbers of items recovered from the site to 808 flakes and 152 artifacts. This is the highest number of artifacts ever collected from a Sandy Cove component. We also collected two nodules of limonite, three small caribou bones, fragmented marine shell and the carbon sample used to date the site from a hearth feature in House 3. The percentages of debitage raw material (Table 1) are very similar to those recorded by Fitzhugh at other Sandy Cove complex sites (Fitzhugh 1972:227-230, 1975:122).

The site scatter is approximately 140m by 65m and was mapped in two zones - Area A and Area B (Figure 2). Area A, which formed more than 80% of the overall site area, was located on the floor of a blowout 13m above sea level. The much smaller Area B, located in the northwest comer of the site, is situated two metres above Area A in a shallow dune no more than 50cm deep. Ground cover of spruce, partridgeberry and caribou moss around Area B had protected this portion of the site from further erosion. During the surface collection it became apparent that there were three houses features present. The three features were aligned over a 45 m distance along the northern section of the site from east to west, at right angles to the closest ancient beach terrace. Unfortunately, Houses 1 and 2 were positioned within the deflated blowout. House 1 now sits two metres below the terrace surface on the blowout floor. House 2, which is somewhat less deflated, is lm below the terrace surface. House 3, however, is located in Area B and has only just begun to erode, sitting no more than 50cm below the surface. Even with such dramatic deflation the houses remained very visible due to their boulder outlines. Each house was rectangular in shape and was outlined by a series of boulders 30-50cm in diameter . House 1 measured 9m x 4m and contained no visible internal features such as hearths (which would certainly have been de-

TableJ. Debitagefrom UvperSandyCove3 Material Type Quartz Unknown Chert Ramah Chert Slate Red Quartzite Iceberg Chert

Mica

Total Numbers 545 135 70 50 5 2

Percentages 67.5 16.7 8.6 6.2 0.6 0.3

O1

Quartz is by far the most common material, no doubt due to its local availability. A quartz outcrop, designated GbBi 8, is located 400m from the site on a high rocky outcrop (Figure 2). This outcrop has evidence of heavy battering and is littered with flakes and cores. Furthermore, quartz is prone to shatter while being worked, and thus produces a large number of flakes. Fine-grained cherts from unknown sources are also well represented and include the grey and purple varieties mentioned by Fitzhugh (1972, 1975). Other materials including Ramah chert from northern Labrador and slate are present in smaller quantities as they are at other Sandy Cove complex sites. The presence of the Ramah chert indicates at least a limited contact with northern Labrador. Red

36

RANKIN: A LABRADOR ARCHAIC LONGHOUSE SITE

Table2. Artifactsfrom UvperSandy Cove3 Artifact Type Cores

n

%

93

61.2

Material 88 Quartz 3 Chert

artzite 5

3.3

4

2

1.3

1 Chert

hlc.eus>