PLACE IN RETURN BOX to removo thlo checkout from you: record.
TO AVOID FINES mum on or before duo duo.

L

DATE DUE DATE DUE DATE DUE
prmij'i

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

=—:———____l

 

MSU I: An Afirmdlvo ActioNEqual Opportunity Institution
Wanna-u 1/

,___,

 

IA“.—

 

GROUP COMPOSITION AND THE ROLE OF UNIQUE
RAW MATERIALS IN THE TERMINAL WOODLAND SUBSTAGE
OF THE LAKE SUPERIOR BASIN.

By

Caven Peter Clark

A DISSERTATION
' Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Anthropology

1991

‘1‘)

ABSTRACT
GROUP COMPOSITION AND THE ROLE OF UNIQUE
RAW MATERIALS IN THE TERMINAL WOODLAND SUBSTAGE
OF THE LAKE SUPERIOR BASIN.
By

Caven Peter Clark

Descriptions of Terminal Woodland substage ceramic heterogeneity in the Upper
Great Lakes region vary from chaotic to cosmopolitan. In this analysis the
relationship between ethnographic and ethnohistoric groups is contrasted with
archaeological cultures to provide a model of group composition which assumes that
group membership is not restricted to a single archaeological culture. A range of
potential interaction alternatives is defined on the basis of ethnographic and
ethnohistoric accounts and is evaluated with respect to the anticipated archaeological
cognates with which interaction must be inferred from the archaeological record.
Using Isle Royale in Lake Superior as the focus of study, the distribution of
geological copper is contrasted with that of archaeological copper within the
parameters of the Terminal Woodland substage (AD. 700 to AD. 1650). The
associations of waste copper, and copper tools and ornaments with ceramics related
to specific archaeological cultures are used as an indication of differential access to
or interest in native copper. Neutron activation analysis of trace elements in 100
geological and archaeological samples of clays derived from Terminal Woodland
ceramics representing Blackduck, Juntunen, Huron, Sand Point, and Oneota

archaeological cultures is used to assess the movement of pots and pottcrs within the

Lake Superior basin. Zones of clay procurement are identified which, in part,
correspond to the distribution of archaeological cultures. The combination of the
copper distribution data and the trace element data suggest that in spite of
widespread associations of diverse ceramics styles on archaeological sites, there may
have been differential access to copper resources indicative of a type of cultural

boundary not generally observed in a prehistoric setting.

COPyright by
CAVEN PETER CLARK
1991

ACKNOWLEDGMENTS

The production of this dissertation involved the assistance of a large number
of individuals and institutions. The neutron activation analysis was supported by a
grant from the National Science Foundation (ENS-8801707) to the University of
Missouri Reactor Facility, Columbia, MO, and by the National Park Service, Midwest
Archeological Center, Lincoln, NE. The following individuals and institutions made
collections available for this analysis: Dave Arthurs (formerly Heritage Branch,
Ministry of Culture and Communications, Thunder Bay, Ontario); Mark Lynott
(NPS-MWAC); John O’Shea (Museum of Anthropology, University of Michigan);
Betsy Garland (Department of Anthropology, Western Michigan University); Pat
Martin (Department of Social Sciences, Michigan Technological University); and
John Williams (Arcadia, Michigan). Technical services and advise were obtained from
Michael Glascock, Hector Neff, and Gar Harbottle. Discussions of regional cultural
historical problems with Claire McHale Milner and Dave Arthurs were most helpful
and the input and support of my committee, Bill Lovis, Susan Martin, Larry Robbins,
and Norm Sauer are greatfully acknowledged. My many families provided the
impetus, momentum, and perseverance that brought this to completion: the Clarks,
Thurbers, Keiths, and Petersons all made substantive material and spiritual
contributions. The staff at Isle Royale National Park, especially, warrants many
thanks for their friendship, accommodation, logistical support throughout my stay in
the north. The BLACKDUCK let me down only when I mistreated her, otherwise
seeing us through the fogs, shoals, and rough seas of Lake Superior. Stellar
individuals who stood by during the long winters and all too short summers, Jo
Thurber, Tim and Jean Cochrane, Edith Green, Stu and Ellen Croll, Uncle Bob, and
of course, Sam, Spider, and Twister, are scattered throughout these pages to remind
me of the years I spent in paradise. Buiadh na Bas.

List of Tables ............................................
List of Figures ...........................................
Chapter 1. Introduction .................................
Research Design ..............................
Chapter 2. Ethnicity and Interaction
Introduction .................................
Ethnicity ....................................

Chapter 3.

Chapter 4.

TABLE OF CONTENTS

Interaction Alternatives: Ethnohistorical and

Ethnographic Basis ............................

Archaeological Cultures and Archaeological Data

Introduction .................................
Archaeological Cultures .........................
Style .......................................
Range and Territory ...........................
Archaeological Evidence of Interaction .............
Archaeological Implications of Interaction ...........
Artifact Analysis ..............................
Site Classification .............................

Archaeological Cultures and Cultural-Historical
Overview

Introduction .................................

Archaeological Cultures of the Terminal Woodland

Substage ....................................

vi

Page

viii

xii

14
14

14

40

48
51
55
59
73

78

88

Chapter 5.

Chapter 6.

Chapter 7.

Chapter 8.
Appendix A.

Appendix B.

Appendix C.

Appendix D.

References

Isle Royale Archaeology

Introduction ..................................
Environmental Setting ...........................
Previous Research ..............................
Overview of the Terminal Woodland Substage on

Isle Royale ...................................
Archaeological Cultures and Interaction ..............

Copper and Copper Mining in the Terminal

Woodland Substage

Introduction ..................................
Geological DistrIbution of Copper ..................
Distnbution of Aboriginal Copper Mines .............
Distribution of Archaeological Copper in the Terminal
Woodland Substage .............................
Examples of Other Unique Exotic Raw Materials .......
Value .......................................

The Movement of Pots and Potters in the Terminal Woodland
Substage: Neutron Activation Analysis

Introduction ..................................
Assumptions ..................................
Conditions and Implications of Covariation ............
The Sample ...................................
Cultural and Chronological Dimensions of the Sample . . .
Geographical Dimensions of the Sample .............
Results ......................................
Hypotheses ...................................

Conclusions ...................................
Terminal Woodland Sites on Isle Royale .............

Fauna] Remains from Terminal Woodland Sites on
Isle Royale ...................................

Neutron Activation Analysis of Prehistoric
Ceramics from Isle Royale and the Upper Great
Lakes (Michael D. Glascock and Hector Neff) .........

NAA Samples .................................

vii

112
112
118

123
137

147
149
151

165
174
178

183
184
186
189
191
193
194
195

201

212

242

250

280

287

Table 1.

Table 2.

Table 3.

Table 4.
Table 5.
Table 6.
Table 7.
Table 8.

Table 9.

Table 10.
Table 11.
Table 12.
Table 13.
Table 14.
Table 15.
Table 16.

Table 17.

LIST OF TABLES

Summary of Terminal Woodland Sites at Isle Royale
National Park ..................................

Calibrated Terminal Woodland Substage Radiocarbon
Dates ........................................

Summary of The Terminal Woodland Archaeological
Cultures at Isle Royale National Park ................

NAA Sample Locations ...........................
Artifact and Raw Material Summary, 201R1 ............
Artifact and Raw Material Summary, 201R5 ............
Artifact and Raw Material Summary, 201R14 ...........
Artifact and Raw Material Summary, 201R17 ...........
Artifact and Raw Material Summary, 201R18 ...........
Artifact and Raw Material Summary, 2OIR28 ...........
Artifact and Raw Material Summary, 2OIR29 ...........
Artifact and Raw Material Summary, 201R31 ...........
Artifact and Raw Material Summary, 201R41 ...........
Artifact and Raw Material Summary, 2OIR42 ...........
Artifact and Raw Material Summary, 2OIR45 ...........
Artifact and Raw Material Summary, 201R52 ...........

Artifact and Raw Material Summary, 201R53 ...........

viii

124

130

139

194

212

213

215

215

216

217

218

219

220

221

222

222

223

(List of Tables cont’d)

Table 18.
Table 19.
Table 20.
Table 21.
Table 22.
Table 23.
Table 24.
Table 25.
Table 26.
Table 27.
Table 28.
Table 29.
Table 30.
Table 31.
Table 32.
Table 33.
Table 34.
Table 35.
Table 36.
Table 37.
Table 38.

Table 39.

Artifact and Raw Material Summary, 201R56 ...........
Artifact and Raw Material Summary, 201R64 ...........
Artifact and Raw Material Summary, 201R65 ...........
Artifact and Raw Material Summary, 2OIR73 ...........
Artifact and Raw Material Summary, 201R78 ...........
Artifact and Raw Material Summary, 201R79 ...........
Artifact and Raw Material Summary, 201R107 ..........
Artifact and Raw Material Summary, 201R111 ..........
Artifact and Raw Material Summary, 201R114 ..........
Artifact and Raw Material Summary, 2OIR116 ..........
Artifact and Raw Material Summary, 201R120 ..........
Artifact and Raw Material Summary, 201R124 ..........
Artifact and Raw Material Summary, 201R127 ..........
Artifact and Raw Material Summary, 201R128 ..........
Artifact and Raw Material Summary, 201R134 ..........
Artifact and Raw Material Summary, 201le ..........
Artifact and Raw Material Summary, 201R142 ..........
Artifact and Raw Material Summary, 201R143 ..........
Artifact and Raw Material Summary, 201R144 ..........
Artifact and Raw Material Summary, 2OIRl47 ..........
Artifact and Raw Material Summary, 201R148 ..........
Artifact and Raw Material Summary, 201R149 ..........

ix

223

224

224

225

225

226

226

227

228

228

229

229

230

230

231

231

232

233

233

234

234

234

(List of Tables cont’d)

Table 40.
Table 41.
Table 42.
Table 43.
Table 44.
Table 45.
Table 46.
Table 47.
Table 48.
Table 49.
Table 50.
Table 51.
Table 52.
Table 53.
Table 54.
Table 55.
Table 56.
Table 57.
Table 58.
Table 59.
Table 60.

Table 61.

Artifact and Raw Material Summary, 201R150 ..........
Artifact and Raw Material Summary, 201R152 ..........
Artifact and Raw Material Summary, 201R153 ..........
Artifact and Raw Material Summary, 201R154 ..........
Artifact and Raw Material Summary, 201R160 ..........
Artifact and Raw Material Summary, 201R174 ..........
Artifact and Raw Material Summary, 201R175 ..........
Artifact and Raw Material Summary, 2OIRl79 ..........
Artifact and Raw Material Summary, 201R180 ..........
Artifact and Raw Material Summary, 201R181 ..........
Artifact and Raw Material Summary, 201R183 ..........
201R1, Chippewa Harbor #1 .......................
201R111, Rock Harbor Lighthouse ..................
201R41, Siskiwit Mine ............................
2OIR114, Lone Tree Cove .........................
201R29, Belle Isle #1 ............................
201R31, Grass Point .............................
20IR45, Daisy Farm .............................
201R149, McCargoe Cove #2 ......................
201R144, Pickerel Cove #1 ........................
201R124, Lake Ritchie #1 .........................
201R142, Belle Isle #2 ...........................

X

235
235
236
236
237
238
238
239
240
240
241
243
244
245
245
246
246
247
247
248
248

248

(List of Tables cont’d)

Table 62.
Table 63.
Table 64.

Table 65.

Table 66.

Table 67.

Table 68.
Table 69.
Table 70.
Table 71.

Table 72.

Table 73.

Table 74.

Table 75.

Table 76.

Table 77.

Table 78.

201R143, Belle Isle #3 ...........................
201R153, Cove #3 ..............................
201R17, Grace Island ............................
201R18, Grace Point .............................
2OIR64, Barnum Island ...........................
201R65, Washington Island #1 .....................
2OIR81, Phelps .................................
2OIR116, Threemile #1 ...........................
201R120, Chickenbone Lake #1 ....................
201le, Washington Island #2 ....................

Element Concentrations in Ceramic Samples from Isle
Royale National Park ............................

Principal Components Analysis for Logged Data
on All 100 Specimens ............................

Fina] Group Assignments and Other Data for Entire
Isle Royale Area Data Set ........................

Mahalaobis Distance Calculation and Posterior
Classification for Reference Groups, Based on First
Four Principal Components ........................

Mahalanobis Distance Calculation and Posterior
Classification for 5 Reference Groups, Based on First
Three Principal Components .......................

Principal Components Analysis for J untunen Reference
Group .......................................

Mahalanobis Distance Calculation and Posterior
Classification for Two or More Groups ...............

248

248

248

249

249

249

249

249

249

249

258

267

268

270

272

274

275

Figure 1.

Figure 2.

Figure 3.
Figure 4.
Figure 5.
Figure 6.
Figure 7.
Figure 8.

Figure 9.

Figure 10.
Figure 11.
Figure 12.
Figure 13.

Figure 14.

Figure 15.
Figure 16.

Figure 17.

LIST OF FIGURES

Lithic Raw Material Sources .......................

Distribution of Terminal Woodland Archaeological ......
Cultures in the Upper Great Lakes ..................

Blackduck Pottery ...............................
Selkirk Pottery .................................
Sandy Lake and Mackinac Pottery ...................
Juntunen Pottery ...............................
Juntunen Pottery ...............................
Huron Pottery .................................
Sand Point (Lakes Phase) Pottery ...................
Oneota (Mero Phase) Pottery ......................
Location of Isle Royale and the Upper Great Lakes .....
Terminal Woodland Sites on Isle Royale, East End .....
Terminal Woodland Sites on Isle Royale, West End .....

Terminal Woodland Sites on Isle Royale, McCargoe
Cove Area ....................................

Distribution of Geological Copper Sources ............
Distribution of Prehistoric Copper Mines ..............

Location of Geological and Archaeological NAA
Samples ......................................

xii

61

90

91

94

96

99

100

105

107

110

113

127

128

129

150

153

190

(List of Figures cont’d)

Figure 18.
Figure 19.
Figure 20.

Figure 21.

Bivariate Plot: Principal Components #1 and #2 ........
Bivariate Plot: Principal Components #1 and #3 ........
Bivariate Plot: Principal Components #3 and #4 ........

Bivariate Plot: Europeum and Lanthanum .............

xiii

276

277

278

279

CHAPTER 1

INTRODUCTION

Archaeology assumes a linkage between archaeological data and the living
people who, produced them. The linkage sought in this dissertation is between the
material remains of late prehistoric archaeological cultures that occupied the Lake
Superior basin and the information it encodes concerning the nature of interactions
among these cultures over a period of about 700 years. Establishing this linkage
entails an examination of differences among material representations of
archaeological cultures through the vehicle of stylistic variation in ceramics and the
evidence for the physical movement of ceramics across a cultural and physical
landscape through trace element analysis of clays used by prehistoric potters. This
dissertation explores the composition of sociocultural groups, the relationships
between sociocultural groups as they are manifested in the archaeological record, and
how a single resource, native copper, may have served as an incentive for and a
medium of social interaction during the late prehistory of Isle Royale and the Lake
Superior region of North America.

Isle Royale serves as the geographical focus of this analysis for a variety of
reasons. As an island in one of the world’s largest bodies of fresh water, in a climate
which imposes definite limitations on access to it as well as the resources it is able
to offer its human inhabitants, the archaeological record of its use in prehistory
should reflect patterns of a less than casual nature. Effort and risks were required

to get there and to live there. The presence of native copper sources there are

2

believed to have constituted an additional rationale for visiting the island, and it is
this resource that may have been one of the factors which contributed to the mixing
of archaeological cultures on the island in the Terminal Woodland substage. From
a practical perspective, the selection of Isle Royale was also determined by the
opportunity of the author to oversee a four-year program of survey and testing of
archaeological sites on the island with the questions addressed by this study in mind.

Trigger (1978) points out the importance of the dialectic between historically
oriented ideographic goals and nomothetic goals:

A better understanding of the past, as
distinguished from formulating timeless laws about
human behaviour, is itself of value and therefore a
worthy goal for archaeologists. Equally important,
however, is my conviction that such an understanding of
prehistory is an integral aspect of the scientific
investigation of human behaviour. [1978:xii]

There is one major nomothetic goal in this dissertation: to elucidate the effects
of a unique, localized, and stable nonsubsistence resource on interaction among
archaeological cultures that operated at equivalent levels of sociopolitical
organization, and that lacked distinct geographical boundaries.

Nomothetic goals are borne on a pyramid of contextually specific, or historical,
ideographic goals which attempt to reconstruct with as much precision as possible the
relations among late prehistoric archaeological cultures as they revolve around the
use of unique localized resources. Prehistoric settlement and subsistence are

considered secondary to the central theme of interaction, and while they are

appropriate areas of investigation in their own right, they are used here as points of

3

departure for the larger issues being considered. In this analytical hierarchy
ideographic goals are the prerequisites for the formulation of nomothetic statements.

As one focus of this study, the role of copper as a stable, durable resource will
be addressed as a vehicle by which interactive/integrative alternatives can be
evaluated. The information derived from stylistic and source analysis of material
culture will make possible statements regarding differential access, value, and mode
of acquisition of copper. The use of trade and exchange as explanatory factors in the
movement of goods may not fully address the implications of interaction between
groups as a manifestation of culture contact. Resorting to trade and exchange as the
explanations for the presence of exotic goods establishes that locally unobtainable
material Objects move between different cultural settings. However, this explanation

does little to assess the context of interaction between parties, the substance of

 

interaction, or the part interaction plays in the larger arena of regional cultural
development and change.

The sociocultural and ethnohistoric anthropology of culture contact situations
is often embedded in a colonial setting, which emphasizes the interactions between
technologically advanced states and their neighboring bands and/or tribal societies.
This approach is one of contrasting extremes in not only the realm of technological
superiority of one side over another, but of the political organization which wields
that superiority.

The dichotomy between directed and nondirected contacts, drawn early in the
anthropology of culture contact situations (Linton 1940, Spicer 1961), reduces

interactive types to two basic forms. Directed contact assumes a

4

superordinate/subordinate relationship between interacting parties where sanctions
of the former are brought to bear on the latter, and where the process is an
intentional effort on the part of the superordinate group. Non-directed contact, on
the other hand, is defined on the basis of the absence of these criteria for directed
contact: that is, in situations where there is no hierarchical relationship, and where
there is no overt anticipation of imposed change on the part of either party.

Such a dichotomy directed toward investigation of colonial situations, is
inappropriate for the study of the subtle and complex interactions between groups
with a less clear agenda for assirnilating or manipulating their neighbors beyond the
personal level. It is also directed to the study of the interactions between whole
corporate bodies (e.g., tribes or states), and lacks the sensitivity to examine the range
of variation which occurs between smaller units within societies. Similar shortcomings
can be found in the area of frontier theory where definitions tend to assume the
existence of a state organization and its interaction with various types of subordinate
societies (Casagrande, Thompson, and Young 1964, Prescott 1965). Directed contact
is really the only type of interaction considered in frontier theory.

Boundary theory comes much closer to the types of interaction considered in
this thesis. Barth’s (1969) edited volume on ethnic groups and boundaries presents
the relational structures of ethnicity relative to other groups among which interaction
is contrasted with the ongoing definition and redefinition of the corporate identity.
All definitions emphasize the highly relativistic nature of boundaries which are
contingent upon the context and constituents involved (Despres 1975, Justeson and

Hampson 1985, Leach 1954).

5

There is, in addition, the coincidence of literacy vs. nonliteracy in this arena
of culture contact which provides written documentation, albeit often biased, of the
context and nature of interaction. We know what the French thought of the Huron,
for example, but have a less clear idea of how the Huron perceived the French (e.g.,
Trigger 1976). What is clear even from this type of early, observationally biased,
documentation is that there is not just one form of interaction repeated between
different groups, but that in spite of an official policy which attempts to condition
appropriate social responses to other groups, interaction varies with its context. The
numbers of individuals involved, the composition of groups, the influence of ties to
parties not present, the primary activities being undertaken at the time of contact, all
influence both the predisposition and outcome of a given episode of interaction. The
study of interaction between native groups prior to the influence of non-native
economic and colonial influences should be similarly structured if more subtle
(Richards, Bruhy, and Goldstein 1987).

How, then, does non-directed contact differ from the interactive design
offered in this dissertation? In a hierarchically structured model of interactive types
which has as its primary variables social structure, subsistence, seasonal demographics,
and geography, different levels of interaction will manifest, in varying amounts,
elements of both directed and non-directed contact. It is not simply a matter of
examining the interaction between whole societies, but one which contrasts and
establishes systematic relationships between the different segments of each society in

their particular forms with those of other societies. This places emphasis on the

6

distinction between the sentiment and content of culture contact situations, and seeks
to define the criteria for each in a given context.

This is not to imply that the study of relations between groups has been
limited to those falling strictly into the directed contact category, although these have
certainly been emphasized. It is possrble to treat specific dimensions of such
interactions, such as trade and exchange, diffusion, or warfare. Such approaches
cannot, however, be expected to explain independently the diversity of interactive
types which occur between group segments because in any given situation there may
be several potentially viable alternative choices of interactive forms.

Within the study of subsistence and settlement there has been a tendency to
concentrate on food resources, appropriately enough, because they are important
factors influencing cultural behavior, and because they leave material remains which
are recovered from archaeological sites. In the present study material
(nonsubsistence) resources are also considered, for they are no less important as
elements around which interactive patterns may revolve. Resources such as chert and
copper are given equal weight in this analysis. Unlike most food resources, these two
material resources are stable and predictable on the natural landscape, but may not
be available to every 81’ cup in equal amounts. The articulation of the combined
strategies for acquisition of subsistence and material resources is important if our
goal is to make generalizations about culture contact, for it is in this arena that
virtually all episodes of contact must be framed.

The role of information exchange between groups may have a significant

influence on decisions regarding group movement or site location selection (Moore

7

1981). Often there is an explicit relationship between mobility and subsistence
interests, such as information regarding localized resource depletion or abundance.
There can also be an economic or prestige-based stimulus for trading Opportunities
both within and among various groups, coordinated by this exchange of information.
While not entirely unconnected from material and subsistence interests, social and
religious concerns will also influence the decision making process, such as the desire
to visit relatives living with other groups and the use of divination as a directive for
planning, respectively.

Visiting and the sharing of resources and information is a characteristic of
small related groups (Sahlins 1972). The exchange of information is a critical aspect
of interaction because the amount and kind of information, transmitted, received, or
symboled, is a function of perceived distance between groups. Information permits
decisions to be made concerning future movements and allocation of energy and
subsistence activities (e.g., Leacock 1954, Moore 1981). Misinformation, or lack of
information, plays an equally important role as a manipulative agency.

It is necessary to have a working definition of the "group" as a meaningful
analytical unit. The term must remain a fluid one with the ability to be redefined
contextually, rather like Sahlins’ (1961) concept of structural relativity as a means of
creating a telescoping structure of social organization relative to a specific set of
circumstances. This is not to advocate a loose definition, but one which is redefined
on a contextual basis, depending on the variables outlined above. Many of the
smaller group segments anticipated in this analysis fall short of current definitions of

ethnicity (e.g., Aronson 1976, Barth 1969, Despres 1975).

8

Ethnic interactions as a subject of study may suffer from the same problems
as have been previously discussed with respect to directed contact. Total societies
become the minimal units of analysis at the expense of the many smaller
permutations of the social structure. In addition, the assumption that such things as
so-called pure ethnic groups exist, now or in the past, is contradicted by the evidence
for multiethnic associations of families and individuals in a variety of contexts (e.g.,
Sharrock 1974). This analysis will consider the archaeological implications of such
multiethnic associations as they relate to traditional explanations of trade and
exchange, and the ability to account for assemblage heterogeneity.

Since this application is archaeological in nature, the maximal interactive units
of this analysis are defined as archaeological cultures as they are manifested by their
material remains. This definition is accomplished by a delineation of stylistically
oriented bodies of material representations in ceramic artifacts with boundaries in
space and time. Once defined, these archaeological cultures will be examined with
respect to their smaller socioeconomic/political segments which correspond to our
understanding of seasonal demographics and their social cognates.

It is assumed here that style encodes some level of corporate membership, and
that the commingling or discreteness of style in a given assemblage will reflect a level
of interaction or integration (Longacre 1968, Michlovic 1981, Weissner1985, Whallon
1968). This is, admittedly, a problematic area. Stylistic heterogeneity in ceramics,
for example, has invoked explanations of a group’s marriage and residence rules (e.g.,
McPherron 1967, Whallon 1968), raiding and warfare (e.g., Trigger 1984, Ramsden

1977), stylistic mimicry (e.g., Mason 1976), and trade and exchange (e.g., Fitting

9

1975), all of which are potentially viable causal factors which are considered in this

analysis.

RESEARCH DESIGN

The sequence of analysis begins with the definition of key analytical units.
In Chapter Two the concept of ethnicity and its relation to archaeological cultures
with an emphasis on the use of style is discussed. Using ethnographic and
ethnohistoric sources examples of alternative interactive options are presented.
With these examples it is possrble to identify grades or stages in interactive
possibilities from animosity to amicability. It is asserted that no single type of
interaction is the exclusive domain of any particular group since, as noted earlier,
unique situational factors may condition the nature of any episode of interaction.
This notwithstanding, there is an important dimension of interactive classification
that is contingent upon group size, composition, and level of sociopolitical
complexity that circumscrrbes the possible choices in interactive types on the part
of the participants. As will be shown, four general types of interaction encompass
most of the variability seen in contact situations.

However varied are the alternatives seen in interaction possibilities of
ethnographic and ethnohistoric groups, the translation of these into archaeological
interpretation poses severe limitations. In Chapter Three the translation of
ethnographic/ethnohistoric interaction into a form which lends itself to
archaeological interpretation is made. Of importance to this study is the extent to

which style is used to symbol group identity, and the extent to which style is

10

utilized by archaeologists in the creation of archaeological cultures. The
implications of style for an archaeologically based definition of group range and
territory are also discussed. Finally, the archaeological cognates of group
interaction are specified, addressing the thorny issue of assemblage variability and
its explanation via archeological data at the levels of artifacts, sites, and the
region.

In Chapter Four, archaeological cultures are defined at a theoretical level
and for the Lake Superior region specifically. A cultural-historical overview gives
the context for the definition of Terminal Woodland substage cultures based on
stylistic variation expressed in material culture. This, in turn, provides the basis
for examining patterns of interaction in the Lake Superior basin. The precontact
archaeological cultures of the Upper Great Lakes region are identified by their
ceramics, some of which, in turn, have been linked to corresponding postcontact
linguistic groups. These hypothesized equivalencies approximate ethnic groups
and/or archaeological cultures, and serve as the maximal units of analysis. While
they form discrete analytical entities, the artifacts which are used to identify and
define these cultures are often found in mixed assemblages. In some cases
assemblages contain materials which appear to reflect a combination of stylistic
elements derived from two distinct traditions. The extent of this mixing and/or
hybridization is assumed to reflect actual or potential contact situations.

In Chapter Five, the Terminal Woodland substage on Isle Royale is
examined with respect to 1) the archaeological cultures represented, 2) the degree

of homogeneity/heterogeneity in the ceramic assemblage, 3) the subsistence base,

11

seasonal indicators, 4) the inferred demographic content, and 5) evidence for
technological activities involving discrete raw materials. The aggregate of these
sites constitutes the basis for suggesting patterns of cultural interaction/integration
in the Terminal Woodland substage. This provides the initial phase of identifying
archaeological examples of potential contact situations.

The Terminal Woodland substage on Isle Royale in the Lake Superior
Basin has been selected as the temporal and areal universe of this analysis. This
framework offers several advantages to this type of study in the form of
reasonably discrete archaeological cultures whose regime of subsistence and
settlement is fairly well documented on the mainland. The Terminal Woodland
was chosen primarily due to the presence and abundance of ceramics which, in
this analytic context, is essential to a meaningful analysis of interaction/integration
among groups. The Terminal Woodland substage has the advantage of being
better documented than any other period of prehistory in the Great Lakes area
(as well as the concomitant disadvantage of being very complex) in the Upper
Great Lakes area.

Chapter Six considers unique resources within a variety of cultural settings,
examining the systems for acquisition, processing, redistribution, and disposal of
the resource. This involves sociocultural, technological, and logistical dimensions
and serves as the point of departure for the discussion of Lake Superior copper.
The role of Lake Superior copper, the techniques used in its extraction and

fabrication, and its regional distribution on archaeological sites are discussed as a

12

necessary prerequisite to the consideration of interaction revolving around copper
resources.

In Chapter Seven, neutron activation analysis is used to test a series of
hypotheses concerning relationships among a sample of Terminal Woodland
ceramics from across the Upper Great Lakes region. These sample groups,
selected from Isle Royale, northern Minnesota, northwestern Ontario, the Straits
of Mackinac, and the Keweenaw Peninsula, reflect mobility of either the pots as
objects of trade or exchange, or the movement of the potters. These data are
then discussed in light of the evidence of cultural interaction on Isle Royale and
across the Lake Superior basin. The availability of ethnohistoric literature from
the Upper Great Lakes enhances the potential for the construction of models
which reflect regional ecology and geography. Within the time frame embraced by
the Terminal Woodland period (ca. A.D. 800-1450), the Lake Superior basin acts
concurrently as a barrier and facility to mobility (Little 1987). Although canoe
trips out of sight of land are reported, voyages across the lake involve risks so
great so as to preclude them as a viable option on a regular basis. Travel along
the lake shore, on the other hand, increases mobility and makes possrble the
transportation of quantities of goods that would be otherwise impossrble. It is
suggested that goods and persons made frequent use of the east-west routes
throughout prehistory, both on the north shore and on the south, resulting in the
interaction between groups of Algonkian, Iroquoian, and Siouan speaking peoples.

There may be a gulf between an interactive study based on ethnographic

and ethnohistorical data, and one based solely on archaeological evidence. The

13

patterns documented by the former must be reduced greatly in order to render
them meaningful to the archaeological context. The interpretation of stylistic
diversity on sites has shifted, historically, from theories of migration and diffusion
to explanations of group exogamy and direct trade or exchange involving ceramics.
This shift is largely due to the increasing emphasis on models which employ a
relatively narrow scope, either of single sites or of catchment areas, and which
insist on viewing archaeological cultures or sites as isolated phenomena. The
limitations of single site analysis are such that equally important issues of
information exchange, mobility beyond catchment areas, and intergroup
interaction/integration are effectively precluded. The extent to which this analysis
is successful will be measured in its capacity to demonstrate that (1) the
examination of nonsubsistence variables is a both desirable and fruitful
undertaking for prehistoric archaeology, and (2) archaeological cultures are fluid
associations which are highly variable in composition, and capable of a wide range

of interactive possrbilities with other groups.

CHAPTER 2

ETHNICITY AND INTERACTION

INTRODUCTION

In this chapter the concept of ethnicity is discussed with respect to its
relevance to an operative definition of archaeological cultures and to the potential
forms of interaction open to ethnographic and ethnohistorical groups. Examples of
interaction drawn from ethnographic and ethnohistorical groups are used to
demonstrate the complexity of the content of group interaction, and the contextually
specific variation that may occur in interaction between two groups. The
classification of interactive types offered is not a model of interaction, per se, but a
means of organizing the continuum of interaction alternatives in anticipation of a
study of archaeological cultures in which a major transformation of the expectations
from an ethnographic/ethnohistoric reality to an archaeological approximation is

required.

ETHNICITY

The use of the classification of sociocultural organization which originated in
the neoevolutionary school of anthropology (Service 1962, 1971) is not a requisite of
this analysis. The catholic use of the definitions of bands, trrbes, and chiefdoms has
repeatedly demonstrated the inherent weaknesses of that system of classification
(Fried 1966), although as broad parameters by which levels of sociocultural
complexity may be contrasted, the nomenclature continues to have heuristic value.

In the Upper Great Lakes region there is a concern with group dynamics which not

14

15

only include the concepts of bands and tribes, but many smaller internal partitions
such as hunting units or trapping partnerships. The notion of an ethnically pure band
or tribe cannot be outright discounted insofar as this idea is the essential foundation
for the archaeological interpretation of style. However, it serves only as a model, and
as a point of departure for an examination of the ragged margins of these concepts
where groups blend and boundaries become indistinct.

Barth (1969) is an oft cited source for definitions of ethnicity in the
archaeological literature. His four dimensions of ethnicity include the requirements
that; 1) the ethnic group must be biologically self-perpetuating, 2) the group shares
fundamental cultural values, 3) the ethnic group makes up a field of communication
and interaction, and 4) the group has a membership which identifies itself, and is
identified by others, as constituting a category distinguishable from other categories
of the same order (1969:10ff). Barth acknowledges the flow of contact and
communication across ethnic boundaries, stating that the nature of the interaction is
structured to preserve the categories of ascription and cultural differences which
constitute the basis for self-identification.

In an elaboration of Barth’s contribution Despres (1975) investigates the
genesis and persistence of ethnic boundaries, the political incorporation of ethnic
populations, and the organization of interethnic relations. In his critique of Barth,
Despres finds the former’s conception of ethnicity highly subjective. However,
Despres himself underscores the subjective nature of ethnicity by pointing out that
cultural categories are relativistic. Much of the argument which follows his definition

and critique is directed towards ethnicity in the context of modern multiethnic

16

communities or states. Despres also finds important the role of competition for
resources in explaining ethnicity: "ethnic boundaries give issue to social strategies
which are designed to monopolize particular resources, or in some instances, entire
resource domains" (1975:200).

A more appealing approach to ethnicity proposes a cognitive/ideational
definition. Aronson (1976:14) states that ethnicity is, "an ideology of and for value
dissensus and disengagement from an inclusive sociopolitical arena, that is, for
pursuing major values deemed not shared by others in the arena." Following Geertz,
Aronson further explains that the "function of [ethnic] ideology is to make an
autonomous politic possible by providing the authoritative concepts that render it
meaningful, the suasive images by means of which it can be sensrbly grasped."
Aronson goes on to point out that the ethnic group is not the only type of group in
the sociopolitical sphere. For example, class is not necessarily synonymous with
ethnicity, and that the use of cultural differences by competing groups for political
ends is the basis for ethnicity. Like Despres, Aronson’s concern is primarily in the
realm of modern social pluralism and multiethnic nation-states.

Using a case study of the interaction between ethnic groups Foster and White
(1982) operationalize a definition of ethnicity which emphasizes the cognitive process
in perceived distance between ethnic categories. Utilizing data from Thailand they
contrast relations between Thai-Muslim (Tai educated), and Thai-Muslim (Islamic
educated), and in another study they examine perceived distance between Chinese,
Malay, Indian, European, and other foreigners in Singapore. Their third and final

case study evaluates the relations between local and immigrant Filipinos in Hawaii.

17

Among their conclusions they find that ethnic categories become integrated into
symbolic systems which are given affective dimensions, and that these categories
become associated with rights and obligations of ethnic groups. Of particular
importance is the finding of differences between behavioral and perceived distance
between groups (I have chosen to express this as the difference between content and
sentiment of relations, respectively), and further, that identity at any given point is
a "dynamic cognitive process involving the social-situational manipulation of ethnic
identity" (Foster and White 1982:121).

The underlying principle shared by these definitions of ethnicity is one of
structural relativity, a term employed by Sahlins (1961) to descrIbe the changing
nature of lineage segmentation but one which has much broader applicability without
compromising its intended meaning. Structural relativity allows for transformations
between categories of identity, dependent on contextual variables. In the following
section, ethnographic and ethnohistoric examples provide us with these contextual
variables and permit a classification of modes of interaction which reflect the

vacillating structure of intergroup/interethnic relations.

INTERACTION ALTERNATIVES: ETHNOHISTORICAL AND ETHNOGRAPHIC
BASIS

Ethnohistoric and ethnographic sources provide the basis for the generation
of a range of interactive types utilized in this analysis. The sources are not intended
to have any historical connection to the archaeological case study. Rather, four
essential interactive options described below are found to operate within a wide

variety of sociocultural situations and in a wide range of settings. The first examples

18

are drawn from historically documented contacts between Caribou Inuit and
Caribou-Eater Chipewyan which form the basis of the interaction classification. This
is briefly applied to a consideration of patterns of warfare among East African
pastoralists. Finally, some examples of interaction are drawn from the Upper Great
Lakes region’s ethnohistoric sources. These examples are only a few of many
possible choices which would be equally appropriate to demonstrate the interaction
options open to groups.

In some respects the Chipewyan-Inuit example contains more extremes in
oppositional categories than will be the case for the archaeological study.
Historically, the two groups had a relatively brief period of contact which may have
curtailed a developing long term relationship in contrast to the prehistoric
Algonkians, Iroquoians, and Siouan groups which had been neighbors for centuries.
The environmental constraints and subsistence options are also more pronounced
than was the case in the prehistoric Lake Superior basin and likely contributed to the
types of interaction documented for the region.

In considering relationships between two ethnic groups and their relations to
their environment, the bOreal forest and barrenland ecology provides the requisite
background. A canbou-hunting subsistence economy characterizes both the Inuit and
Chipewyan with antecedents in the prehistory of the region; a pattern which among
the Chipewyan persists in a modified form to the present.

The prehistoric record of human-caribou interrelationships is an appropriate
point of introduction to the diachronic trends in adaptive strategies and human

interaction examined below. Gordon (1975) has developed the "discrete band/discrete

19

herd" hypothesis based on archaeological investigations in the barrenlands. Arctic
Small Tool and Dorset sites dating between 1500 and 700 BC. produced a
configuration of internally homogeneous stylistic complexes which appear to
correspond to the distributions of the Kaminuriak, Beverly, and Bathurst herds. The
discrete band/discrete herd hypothesis is suggested for a Paleo Eskimo archaeological
culture, yet the model is derived in part form the historic Chipewyan, indicating a
highly successful adaptation transcending chronological or cultural differences. The
Chipewyan pattern may be projected back as early as 200 BC. with the Hennessey
complex of the Taltheilei Shale Tradition (Gordon 1975). The adaptive strategies of
both the Chipewyan and Inuit are remarkably similar. Over time, that is, with the
cultural evolution of the Caribou Inuit, the Chipewyan and Inuit grow to resemble
each other more closely. For both groups social organization encodes the strategy
and tactics for optimizing the target resource: caribou.

Helm’s (1968) model of Athabaskan social organization defines the regional
band, the local band, and the hunting unit. The Chipewyan bands which exploit the
Kaminuriak herd and those which exploit the Beverly herd assume their group
identity from some geographical feature within their range (Smith 1978:76). Band
membership is fluid with recruitment by residence and joint cooperation in hunting.
This corresponds closely to the definition of the band range which is defined by the
regular use of a discrete area by a particular band. The regional band rarely if ever
coalesces into a single cooperative unit.

The local band functions as a "microcosm of the regional band" (Smith

1978:76). Identity is derived from a smaller geographical feature within the band’s

20

range. Winter sites of the local band are located within the transitional zone near
frozen lakes where yarding caribou are anticipated (Irimoto 1981, Smith 1978).

Hunting (and later in the eighteenth century, trapping) groups are the smallest
unit of organization among the Chipewyan and are typically comprised of three to
eight nuclear families (Smith 1978:77). Irimoto (1981) recognizes the nuclear family
as the minimal unit of production. Both are accurate interpretations when applied
to specific activities, emphasizing corporate or individual strategies. Units smaller
that hunting groups are usually formed within the context of the hunting group. By
contrast, caribou hunting almost always involves the entire hunting or local band,
depending on season and strategy.

Major harvests of carbon are possrble twice a year, during the spring and fall
migrations. Since a herd’s migration corridor may vary from year to year Chipewyan
hunting units are distrrbuted with respect to the anticipated distribution of caribou
(Smith 1978:82). This is a linear arrangement of kin-based communications networks
which crosscut the paths of the Beverly and Kaminuriak herds along the southern
margins of the tree line (Gordon 1975, Smith 1978). Herd movement is monitored
in all season by small hunting units, families, or by the occasional all male hunting
parties that venture well into the barrens (Smith 1978:82). This monitoring and
subsequent reporting of herd locations and movement provides the basis for
anticipating which locations will intersect the migration conidors. Water crossings
figure importantly in this respect, since spearing from kayaks or canoes provides large
returns. A second period of caribou hunting occurs between migrations and is

conducted by individuals or small hunting units. Summertime hunts followed the

21

caribou along their migration corridor but this technique was not used to generate
a large surplus. Winter hunting was conducted within the relatively restricted winter
range with camps typically located near a yarding area.

In contrast to the longevity of the Chipewyan pattern, the Caribou Inuit
arrived in the barrenlands early in the seventeenth century. Their origin has been
variously hypothesized: Birket-Smith (1929) viewed them as a "proto-Eskimo" relic
of the earliest form of Eskimo culture prior to the development of a coastal
adaptation. Mathiassen (1930) suggested an opposing idea, that the Caribou Inuit
split off from a Thule group and gradually adopted an interior subsistence pattern.
Burch (1972, 1979) utilizes climatic data to support Mattiason’s hypothesis, suggesting
that an interior carrbou centered adaptation evolved out of an episode of climatic
deterioration in the "Little Ice Age."

The Chipewyan sentiment for the Inuit is perhaps best expressed by their
taxonomy of other groups. Groups with which the Chipewyan have frequent contact
(i.e., other Athabaskans) are called M according to the cardinal directions (e.g.,
saaisa dene or eastern people, yodai dene or western people). The Inuit are referred
to as Ma (barrenland enemy), the Cree as 91313 or enemy (Irimoto 1981:25).
Viewed as the object and subject of contempt by the Chipewyan, the Inuit provided
the model for the hot’elna det’oi, a supernatural people who resemble Eskimos but
have massive body hair resembling a bear. Food is eaten raw, and in times of need,
so are their own children. They are found only in the range of the Inuit (Irimoto
1981:24). Clearly, the use of the mythical image served to reinforce the distance

between Chipewyan and Inuit by promoting inimical attitudes.

22

The Caribou Inuit are divided into five maximal bands consisting of 50 to 300
persons each (Arima 1984). These are autonomous and lack formal internal political
organization. There is consciousness of group identity among the maximal bands with
particular affectations of dress, name, and tool styles. Maintenance of maximal band
identity is seen in the tendency for endogamy within Caribou Inuit society, in contrast
to marriages with coastal Inuit groups.

Spearing at water crossings and bow and arrow killing at drive lanes meet the
caribou in May. Hunting units intercept the herds to prepare a store of dried meat
for summer use (Arima 1984:448). If necessary the summer stores are augmented
by fishing. Maximal band aggregates coalesce in late summer and remain together
throughout the winter. There is a major investment of corporate energy directed to
the fall harvest of caribou since it must serve the requirements for food and shelter
until the following spring. Small groups of carrbou often wintered in the wooded
river valleys of the barrens. These were hunted by stalking but this technique was
met with low returns until the introduction of the rifle made long range shooting
possible (Arima 1984:451). Winter camps were concentrated in the interior with
dispersed coastal camps in the summer.

The settlement and subsistence strategies of the Caribou Inuit and the Caribou
Eater Chipewyan are very similar. The tenure of the Caribou Inuit is barely 300
years long, but within this brief time one witnesses the rapid development of a
subsistence system highly reminiscent of its Athabaskan neighbors to the south, while
in other domains of culture it remained uniquely Inuit. Sharp (1977) examines

Chipewyan social organization throughout the historic period, finding that despite the

23

adoption of the Algonkian style canoe, use of dog teams, and firearms, it remains
essentially unchanged. Modifications were made in seasonal scheduling to meet the
added responsibilities of provisioning fur trading posts with carrbou meat. Fur
trapping was undertaken unenthusiastically (Sharp 1977:37).

Secondary sources based on historical documents and direct observations allow
examination of the patterns of interaction between these two groups.
Chipewyan-Inuit interaction is considered below with respect to four questions:

1) Is the content of the interaction representative of the sentiment

expressed by one group for another?

2) What motivational factors can be identified to explain the various types
of interaction?

3) To what extent can the types of interaction between these groups be
attributed to general principles of interaction between groups with
similar adaptive strategies and technological capabilities?

4) How do interactions document themselves materially?

Using examples of interaction cited by Arima (1984), J anes (1973), and Smith
and Burch (1979) largely taken from Hudson’s Bay Company records and those of
the Thule Expedition, four categories or types of interactive options have been
abstracted:ambivalent,amicable/cooperative/integrative,antagonistic/ noncooperative,
aggressive.

Smith and Burch (1979) have divided the history of Inuit-Chipewyan

interaction into four periods: Early Historic (1613-1715), Period of Intervention

24
(1715-1782), Period of Peaceful Coexistence (1782-1950), and the Period of

Nucleated Settlement (1950-1977).

The early historic period finds the Chipewyan hunting caribou well north in
the barrens. Inuit groups were restricted to the coast where they were actively
engaged in sealing. This early period is inadequately documented for our purposes
although according to Smith and Burch (1979) there are vague references to hostile
sentiments between the two groups.

Documented interaction begins to gain momentum with the presence of
French and British traders and officials. The establishment of posts at York Factory
and Churchill marks the beginning of the Period of Intervention at which time the
Chipewyan were encouraged to pursue furbearers for the trade. Hudson’s Bay
Company (HBC) records note that in 1717 a party of Inuit ambushed and killed six
of a party of nine Chipewyan. No motivation is mentioned. Documents for the
period between 1717 and 1721 give indications of alternating trade and hostile
relations as Inuit groups began to impinge on the Chipewyan summer hunting range
(Smith and Burch 1979:78). Until the Inuit received firearms in 1770 they were kept
well to the north of Churchill by the flintlock-armed Chipewyan seeking to restrict
access to the posts (Arima 1984:459).

A document from 1725 records a Chipewyan impingement some 150 km into
the Inuit range (Smith and Burch 1979:78). Apparently there were no consequences
of this action. What is interesting about this account is the recognition on the part
of the HBC recorder of an Inuit range or territory with a characterization of the

Chipewyan hunting unit as intrusive. By the middle of this period a pattern of

25

variable interaction seems well established. HBC records from 1748 note that there
are hostile sentiments among the Chipewyan who tend to blame the Inuit for their
own misfortunes. However, it is the same year in which the first recorded example
of coresidence occurs.

The regular trading visits to Eskimo Point by the HBC beginning in 1750
stimulated feelings of jealousy among the Chipewyan who were losing their
monopoly over access to trade centers. Accusations of preferential treatment towards
the Inuit may have given rise to the planned killing of an Inuit trading party camped
at Eskimo Point in 1755. The HBC trading sloop had only just sailed out of sight
when the Inuit were set upon by a party of Chipewyan. The reasons for this episode
are twofold. The sloop had apparently failed or chosen not to respond to the smoke
signals of the Chipewyan who desired to trade. Insulted, they prepared to retaliate
against the Inuit who had just completed their trade (Smith and Burch 1979:51).
Hearne (1958:217) suggests the killings were revenge, albeit in the context Of
opportunistic circumstance, for the magical slaying of two Chipewyan in the previous
season. The explanations are not mutually exclusive and presuppose an atmosphere
of hostility, real or imagined.

A peace agreement between the Caribou Inuit and the Chipewyan was
negotiated by the HBC in 1764. Over the winter of 1767-68 a second case of
coresidence was noted. In 1771, while peaceful trading was underway among the
coastal Inuit and Chipewyan on Hudson’s Bay, Yellowknife and Chipewyan
massacred the Inuit of the Coppermine River at Bloody Falls (Hearne 1958).

Hearne’s account makes no mistake about the organization and intended outcome

26

of this episode. His own compassion for some of the unfortunate Inuit was viewed
as pathological by his Chipewyan hosts.

The period of Peaceful Coexistence is subdivided into two subunits by Burch
and Smith (1979:81ff): Inuit Expansion (1782-1890) and Territorial Stabilization
(1890-1950). At the outset of this period several factors crippled the Chipewyan
position in the region. The winter of 1781-82 brought smallpox to many local band
camps. In 1782 the French burned the HBC post at Churchill, which only served to
encourage a shift to the post on Lake Athabaska established in 1778. The population
decline and reduction in trading opportunities resulted in diminished opportunities
for contact with the Inuit (Smith and Burch 1979:82-83).

By contrast, the Caribou Inuit appear to have undergone a population increase
during the latter part of the eighteenth century. Continued encroachment into the
Chipewyan summer range by the Inuit took place in a mixed atmosphere of amicable
trading relations and small scale depredations between 1807 and 1818. HBC records
continue to note the jealousy of the Chipewyan over signs of favoritism toward the
Inuit. Inuit incursion westward and southward into the barrens took place as many
Chipewyan bands were moving into the full boreal forest recently vacated by the
Cree, themselves decimated by smallpox and measles (Yerbury 1976). Coresidential
fall groups including both Inuit and Chipewyan were an annual occurrence on
Yath-kyed Lake (Smith and Burch 1979:85). Cooperative efforts took on a new form
in 1840 when the HBC uncovered a Chipewyan-Inuit conspiracy to defraud the
Company in the sale of furs. Permanent occupation of the interior barrens by the

Inuit was established by 1860. Friendly relations persisted without the exclusion of

27

the Chipewyan from their summer hunting range in southern Keewatin, and by 1870
discrete ranges ceased to exist in an area of 24,000 km square.

Observations in 1868 by a Catholic missionary among the Chipewyan
underscore the friendly content of Chipewyan-Inuit relations. There were
expectations of regular contact with the Inuit on the part of Chipewyan hosts. Some
of the Chipewyan were bilingual to a limited extent, although there is no indication
that the reverse was true. A greeting ritual was developed and there was enough
knowledge of Chipewyan food preferences so that an Inuit host knew to offer his
guest dried instead of raw meat (Smith and Burch 1979:85). These observations,
assuming they reflect the general content of relations for this period, indicate a
tendency toward cooperation in trading and amicability in other forms of interaction.

The end of Inuit expansion is marked by the establishment of multiethnic
communities around the trading posts. HBC census show that in 1881 the Churchill
post community consisted of 515 coastal Inuit and 157 Chipewyan. The Brochet post
in the same year included 386 Chipewyan, 217 Inuit, and 29 Cree. The transition
into the following period also sees the crystallization of ethnic stereotypes which were
to influence employment opportunities up to the present. Inuit workers were hired
into the whaling industry in the 1830’s where they gained a reputation for
industriousness as opposed to the indolent Chipewyan, as they were perceived by
whites. It was also this association with American whalers that introduced repeating
rifles to the Caribou Inuit (Arima 1984).

The subperiod of Territorial Stabilization (1890-1950) sees the emergence of

Inuit dominance in the local trade. However, the benefits of this advantage were

28

shortlived. The year 1917 brought a severe decrease in the caribou population,
resulting in starvation among the Inuit. This catastrophe became a trend, persisting
with dramatic declines among both Inuit and Chipewyan populations. A famine more
devastating than the 1917 disaster occurred in 1950. The reasons for this particular
famine are unknown, although Arima (1984) suggests that a combination of
overhunting and erratic carrbou population decline are likely factors.

Interaction characteristic of earlier periods continued in the context of
seasonal hunts in the interior. The establishment of a post on the northern end of
Nueltin Lake in 1917 for the Kazan River Inuit brought them into sporadic but
friendly contact with the Hatchet Lake Chipewyan. In 1930 a single Chipewyan
household camped across a river from an Inuit settlement of four or five families
where visiting, gaming, and dancing took place, "yet the establishment of regular
interethnic social relations failed, leaving mutual suspicion and contempt" (Irimoto
1981:21-22). For all intents and purposes, opportunities for even this type of limited
interaction ceased in 1961 when, in the face of a famine which threatened the
remnants of the Caribou Inuit, they were removed to government camps on Hudson’s
Bay.

The only documented case of intermarriage occurred in 1938. Unfortunately,
both parents died shortly after the birth of their first child, underscoring the
Chipewyan belief that it is impossrble to breed successfully with the Eskimo (Irimoto
1981, Smith and Burch 1979).

Post-1950 culture change increases in momentum and in complexity.

Chipewyan and Inuit contact, which was becoming less frequent, was limited to the

29

trade center at Churchill. Caribou Inuit and other Inuit groups who had no history
of interaction with the Chipewyan began to emigrate to this port, creating a
pronounced discontinuity in Inuit-Chipewyan relations up to this point (Smith and
Burch 1979:90). Ethnically segregated neighborhoods and a ranked ethnic hierarchy
virtually isolated the two groups in spite of their close physical proximity.

Smith offers an optimistic view of recent changes in Chipewyan social
structure. Rather than assuming negatively that a process of "deculturalization" is
underway, the shift towards increased sedentism is described in terms of the "the
emergence of new cultural adaptations based in part on upon traditional patterns"
(1978:38). In time, kin-based institutions tended to become more difficult to
maintain: the larger postcentered villages may have strained interpersonal relations
previously avoided by greater mobility and a diffuse settlement pattern. However,
traditional values continue to be preserved, including an emphasis on egalitarianism,
decision-making by consensus, and principles of reciprocity (Mattiason 1975).

Smith and Burch (1979) conclude that despite a sentiment of hostile
relationships, few instances of violence can be documented for the Chipewyan and
Caribou Inuit. Their use of HBC records suggests that the sentiment of relations
became less overtly hostile through time. Janes (1973) also finds the content of
historical relations essentially peaceful, but colored by a sentiment of mutual
animosity.

Organized, premeditated violence occurred in the 1755 incident at Eskimo
Point and possibly earlier in the 1717 ambush. The extent to which these events can

be considered organized is debatable. In all likelihood the incidents were

30

spontaneous in origin and acephalous in leadership, despite the necessity in the case
of the 1755 killings of some level of premeditation and predisposition on the part of
the Chipewyan. The Bloody Falls massacre of 1771 did not involve the Canbou Inuit
but represents an extreme case of organized violence.

Antagonistic and noncooperative interaction includes the numerous accounts
of small scale depredations. The 1717 incident may belong in this category.
Spontaneous conflicts may have arisen out of chance meetings, although the stimulus
for violence is impossrble to characterize in a general way. This form of essentially
hostile relations persisted until about 1820.

Both amicable/cooperative/assertive and ambivalent interaction functioned
simultaneously with antagonistic examples, although the general tendency was toward
more amicable relations. The extent to which these relations are conditioned by
sanctions of the HBC and the desire to preserve economic relations with whites (or
more importantly, access to trade goods) has not been developed here. Cooperation
in hunting may have evolved later as shifting ranges brought Chipewyan and Inuit
into contact at strategic hunting localities. Although avoidance through scheduling
was suggested for the earliest Chipewyan-Inuit relations, it will be a difficult
proposition to support empirically since, by definition, it leaves no tangrble evidence.

While it is obvious that both the Chipewyan and Inuit were in competition for
caribou, this competition was spatially and temporally isolated, usually without
confrontation at the time the actual hunts were in progress. Contrary to expectations
that competition over resources leads to violent or antagonistic behavior, caribou

hunting appears to have been the context for coresidence and cooperation. Only

31

control of access to trade centers and the jealousy related to the trade promoted
antagonistic relations. In this setting it is likely that a synergism between the negative
sentiments of the corporate body of Chipewyan and threats to an individual’s avenue
for prestige (i.e., trading) reinforced the possibility of violent interaction. Ironically,
it is the same desire to control access to trade goods (and prestige) that at other
junctures promoted amicable, or at least cooperative, exchanges between the
Chipewyan and Inuit. Technological advantage in the form of firearms was briefly
held by the Chipewyan until 1750 when regular trade between Inuit and the HBC
bypassed the Chipewyan middlemen, effectively destroying their monopoly on access
to Churchill. Otherwise, the two groups operated in a state of parity. As the number
of sled dogs increased as a concomitant to an increased involvement in the fur trade,
hunting pressure on the carrbou increased. This, combined with the effectiveness of
the repeating rifle and a predisposition for over-harvesting, placed both human
groups in a precarious disadvantage with respect to their resource base. While there
is no documentation for it, one might predict that under this type of stress and given
what is known about caribou behavior and the hunting practices of both the
Chipewyan and Inuit, each group may have sought to blame the other for resource
scarcity and the ensuing collapse of the traditional hunting system.

The range of interactive options available to the Chipewyan and Inuit are not
artifacts of their sociocultural organization or of any historical phenomenon. Looking
at interaction among East African pastoralist groups one finds the same range of

options available. As with the former example adaptive strategies and technology are

32

viewed as constants. The critical natural resources are the distrrbution and
availability of water and grazing lands.

In contrast to the Chipewyan and Inuit, the frequency of violent interaction
is quite high. The need to maintain herds instead of hunt them places obvious
constraints on the pastoralists. Typically, they operate in a range with a finite
number of water sources, many of which are seasonal. Similarly, there is a
nonuniform distrrbution of grazing lands which is also seasonal. As a consequence,
pastoralists view territory rather differently than do hunters. There are finite limits
to options in coping with drought, one of which is impingement on a neighboring
group’s waterholes. There is a cognition of territory among pastoralists that sets
spatial limits relative to herd and camp location and needs. These may not
encompass the same physical territory all year but fluctuate seasonally.

Violent interaction among pastoralists is largely reciprocal. Among them is
found a similar juxtaposition between the sentiment and content of interaction
(Jacobs 1977). That there is a predisposition towards violent interaction is, however,
indicated by the social institutions which operate to satisfy certain needs such as
economic and spiritual motivations. Cattle raids may or may not involve violence
between groups but raiding often provides the impetus for retribution and escalation
of hostilities. Cattle raiding is conducted to acquire bridewealth (Evans-Pritchard
1940), status (Baxter 1977, Fukui 1977, Tomay 1977), and/or revenge or recover
stolen property (Larry Robbins, personal communication 1991). Economic
opportunism may stimulate violent interaction as in the case of the Maasai’s regional

dominance in the ivory and slave trade (Jacobs 1977).

33

Among the Bodi intertnbal homicide is generated by the death of a morare

 

or "favorite ox" (Fukui 1977). Each adult has a morare, the color patterns of which

are a source of personal spiritual identity. If the morare sickens or dies, the death

 

of a member of a neighboring non Bodi group is sought as recompense. There is no
reward other than spiritual implied in this form of violence, although individual
prestige is a factor.

There is no central decision making process and usually no long term strategic
objectives in mind when raiding is undertaken by pastoralists (although there are
tactics). To this extent there is some similarity with the Chipewyan and Inuit.
However, the role of age-grades indicates an institutionalized aspect of raiding in
East African pastoralist societies (Baxter 1977, Evans-Pritchard 1940, Jacobs 1977).
With respect to its functioning as a defensive network, Baxter (1977) notes some
similarities with bilateral kin structure. Bilaterality exists in the context of contrasting
pairs of age sets. The younger unmarried males constitute the active and aggressive
role complemented by the older males who typically advocate passive resolutions to
intergroup hostility.

There is proportionately more violence in intergroup interaction among East
African pastoralists than between the Chipewyan and Inuit. Dramatically different
perceptions and realities of territory and resources prevail with greater limitations on
the pastoralist’s mobility and interactive options. Among horticulturalists there is an
even more emphatic definition of territory (Fukui and Turton 1977).

Before turning to the archaeological cultures and interaction in the next

chapter some ethnohistoric examples from the study area require mention. Although

34

this analysis downplays the immediate relations between a late prehistoric
archaeological culture and the ethnohistorically documented groups of contact and
the early historic period, the circumstances of interaction during this time are
important as the closest geographical and cultural-historical analogs germane to the
question of interaction.

The ethnohistorian may be as guilty of oversimplification of ethnic
composition as the archaeologist by making blanket identification of a group without
considering the origins of each individual of the group: a Huron village may contain
a number of individuals or families from neighboring Algonkian or other Iroquois
groups. Sharrock’s (1974) analysis of ethnohistoric documents relating to the Cree
and Assiniboine are highly instructive in this regard. Apart from the motives which
may be directly attrrbutable to the fur trade in the region, she documents the
multiethnic association of the two groups in coresidence units ("any unified
cooperative aggregate of any size whose members, for social, political, and/or
economic purposes, form a single local group for any length of time," 1974:98).

Ethnohistoric accounts descrrbe avoidance behavior (Innis 1970:24) and the
presence of no-man’s-land or buffer zones between villages as a means of
ameliorating the potential for violent interaction. An atmosphere of reciprocal blood
feuds acted as a backdrop for relations between the Huron and their neighbors to the
south and east (Trigger 1976:319), and between the eastern Sioux and Ojrbwa
(Hickerson 1970). But beneath the very real potential for violence there existed a
variety of mitigating techniques which belie the less antagonistic relations between

interactive parties. In spite of officially antagonistic relations with the Mohawk many

35

Huron maintained amicable relations with them (Trigger 1976:349). The use of real
and fictive kinship terminology coupled with the widespread adoption of captives also
formed tangible links between groups. Finally, cooperative ritual was used to
establish and reinforce relations: the Feast of the Dead included representatives of
the Huron, Nipissing, and Algonkian and physically merged the bones of the dead as
a strong symbolic merger of common interests (Cleland 1971, Hickerson 1960,
Trigger 1976).

The distinction between the corporate feelings of animosity and the actual
sentiment of interacting individuals is well expressed in an observation made by
Joseph Nicollet, August 6, 1863, on a trip up the Mississippi River:

The Chippewa from Sandy Lake and the Sioux from the lakes around

St. Anthony Falls [modern St. Paul] often live on good terms and part

only in case of general belligerency between the two nations to which

they belong. So it was with the two parties I met successively on my

way here. In the course of their hunting expedition, now interrupted,

they had fratemized every evening in those abandoned carnps...I found

out that they had danced every evening, but I arrived too late to see

the dances. They separated following the news of crimes committed
at large in the land, and, not knowing what consequences these events

might bring. [Bray 1970:45]

Thus, in spite of overall good relations, the Ojrbwa and Sioux were
compelled by the anticipated need to align with their respective corporate interests
when events suggested an outbreak of hostility.

Based on the foregoing ethnohistoric and ethnographic examples, four

major modes of interactive options are seen operating in varying degrees among

36

human groups which perceive themselves as discrete and nonrelated to those with
whom they are interacting:

AMBIVALENT: A pattern of small group interaction, ambivalent
relations may involve some of the characteristics of cooperative relations but is
tempered with mutual distrust and suspicion. Scheduled interaction is less likely
to occur although it is not precluded (e.g., Ford 1972). The development of
interpersonal relations may remain unaffected by this larger atmosphere of
unfriendly relations (Trigger 1976). Ethnic distance as it is perceived by respective
group members conditions ambivalent interaction, potentially transforming it in
either less or more amicable relations.

Ambivalent interaction may take the form of avoidance. Scheduling of
movements may, in part, be designed to minimize contact with other nonrelated
groups or to avoid a concentration of people where subsistence resources may be
limited. Avoidance is used to explain the mutually exclusive use of the northern
barrens by Chipewyan in the summer and Inuit in the winter (Smith and Burch
1979:78). If such a balance use of the hunting range did occur in prehistory it was
soon overshadowed by other forms of interaction. Avoidance may be
underrepresented in historical records since it may have gone unnoticed.

AMICABLEZCOOPERATIVEQNTEQRATIVE: This type of contact is
peaceful and mutually beneficial to involved parties, typically found in the context
of bands, smaller kin-based groups, and individuals. Shared language is not a
requisite but is a conditioning factor. Mutually beneficial activities may include

interpersonal trading arrangements, exchange of group members on a limited and

37

temporary basis, cooperative resource procurement, and formalized scheduled
contact. In addition, multiethnic associations may form to become integrative
functioning units for any length of time.

In varying degrees, sharing, exchange, and formal trade characterize
relations of mutual benefit which in turn reflect perceived degrees of relatedness.
Sharing of food is a significant measure of relation, according to Sahlins (1972),
representing close ties of mutual consideration modeled after relations between
kin. The movement of goods and persons between interactive groups is
reciprocal, based on mutually determined equivalencies in exchange value. Parity
in values is a function of perceived distance: close relations exhibit little or no
concern for parity in exchange value and the reciprocity can be delayed
indefinitely. With increased distance a more rigid reckoning of value is imposed
and the returns are expected to obtain within a shorter span of time.

Forming trade arrangements will often involve a ritual dimension. Ritual
greetings and games may anticipate the exchange, indicating the formal quality of
interaction. Ritual shared on a larger corporate level is an even stronger
statement of perceived distance, such as Algonkian participation in a
uron-sponsored Feast of the Dead where the commingling of remains symbolically
expressed the closeness between the two groups while at the same time
underscoring the economic obligations held by each party.

ANTAGONISTIQN ONCOOPERATIVE: This pattern of interaction is
characteristic of competing, ethnically distinct groups. Size is not an absolute

factor but the tendency is toward larger and more sociopolitically complex groups

38

which, unlike small groups or individuals, embody corporate sentiments that may
amplify the negative aspects of ambivalent relations. Relations may be
characterized by avoidance or periodic spontaneous aggression. Conflicts arise as
a response to overt displays of animosity at confrontations over resource
competition, often perceived as territorial transgressions or threats of usurpation
of economic prerogatives. The juxtaposition of competing groups will stimulate a
corporate response defined in terms of corporate sentiment and perceived distance
between the opposing groups.

AGGRESSIVE: Violent interaction occurs at all levels of sociopolitical
complexity, although the capacity to orchestrate planned aggression is facilitated
by the presence of more formally structured ideas of corporateness defined in
oppositional terms, and in the existence of integrative structures within a society
which bring together and operationalize planned acts against other groups. The
use of violence or, to invoke a Hobbesian sense of "warre" (a state of anticipation
of being a victim of another’s violence), may involve specialized social institutions
such as warrior sodalities. These may become institutionalized to the extent that
violence assumes a ritual form and seasonal context. Violence involving killing
assumes a spiritually charged nature at all levels. The consequences of violent
relations may result in a conscious restructuring of spatial relations with contested
areas and/or a no-man’s-land phenomenon. Ethnic relations may be played out in
the treatment of captives, for example, adoption and incorporation will most often
occur within linguistically related groups while more distant groups will be given a

nonhuman status.

39

The extent to which the above consideration is applicable to archaeological
examples is contingent upon the quality of the archaeological data. The level of
specificity regarding group size, composition, and other operating factors seen in
the ethnographic and ethnohistoric examples is almost unknown in archaeology.
We may know from isolated examples that violent interaction took place in
prehistory, but knowledge of the conditioning factors or other forms of interaction
that may have occurred simultaneously remain obscure in the archaeological
record. The transformation of the above examples into a form recognizable in an
archaeological setting is one of scale. The fine scale of the ethnographic and
ethnohistoric record requires generalization to a broader scale in which the
operating units, loosely referred to as the "group," embrace larger aggregates of
members in which patterns of interaction are manifest. These larger aggregates

are defined as archaeological cultures.

CHAPTER 3

ARCHAEOLOGICAL CULTURES AND ARCHAEOLOGICAL DATA

INTRODUCTION

We have explored the complexity of interaction as it is played out among
living groups, both distant and recent in time. It is now necessary to make the
transformation that forms a bridge between the living and the dead in order to
further examine the form and variability in interaction among archaeological cultures.
This transformation becomes evident in the definition of archaeological cultures as
entities based upon but vastly different from ethnographic and ethnohistorical ones.
In this chapter we will see how groups are defined and how interaction is viewed
through the archaeological record before moving into the more tangible realm of the

late prehistoric archaeological cultures of the Lake Superior region.

ARCHAEOLOGICAL CULTURES

In the discussion that follows, which adheres to Trigger’s usage of the concept
of archaeological cultures, it will become clear that, for the archaeologist, the
conceptualization of an archaeological culture is almost exclusively dependent on the
configuration of traits (largely normative, nominally behavioral) associated with a
geographic distribution on the physical landscape.

An archaeological culture may be defined as a geographically

contiguous set of artefact types that may occur in differing

combinations in different functional contexts and that together form

the surviving material expression of a distinctive way of life sufficiently

comprehensive to permit its bearers to perpetuate themselves and their
behavioral patterns over successive generations. [Trigger 1978:76]

40

41

Archaeological cultures are large in areal scope, long in temporal duration,
and exclusive in terms of membership and are, therefore, not synonymous with
component or community (Trigger 19782116). Components and communities are
often multiethnic, reflecting the dialectic among archaeological cultures (Syms 1977).
Trigger (1978:117) cautions that, "a uniform material culture does not constitute
proof that the people associated with it had a strong sense of common identity, any
more than differences in material culture prove that lack of such a sense of identity."
Thus, it is accurate to state that a given site has a Terminal Woodland substage
component that includes elements Of the Huron, Blackduck, and Juntunen
archaeological cultures. Conversely, sites cannot, by the criteria set forth here, be
described as having Huron, Blackduck, and Juntunen components, unless
chronological or stratigraphic distinctions are discernable.

Certainly there are examples of sites where stylistically homogeneous ceramics
suggest that only one archaeological culture was present. In this scheme such a site
is considered an example of the absence of mixing of potters from discrete groups,
but not necessarily one of the absence of interaction. Examples of stylistically pure
components tend to occur in small collections where representativeness is in question
and/or at small sites where the implication of a small number of people diminishes
the probability that potters representing discrete stylistic backgrounds will be present
at the same time. Interaction must be measured on a scale larger than a single site.

One attempt to measure complex interactions is Syms’ (1977) Co-Influence

Sphere model. This model is a response to the traditionally simplistic chronological

42

model in which a one-to-one correlation of archaeological cultures and surviving
ethnohistoric groups is assumed. The chronological model is unilineal, that is,
cultures are largely viewed as having evolved in place over long periods of time (e.g.,
from the Shield Archaic to Laurel to Blackduck to historic Cree). It is also an
assumption that there is no geographical or temporal overlap between archaeological
cultures, with little movement within their territories. Under this model, change in
material culture is explainable by invention or diffusion, all transitional forms
attributable to in situ evolution.

Syms reaction to this takes the form of a complex model which attempts to
explain the diversity of ceramic styles so often found on archaeological sites in a
region. His model was stimulated by evidence of contemporaneous use of
sites/regions by a variety of ethnographic groups exhrbiting both positive forms of
interaction as well as avoidance behavior and "emphasizes the importance of a
constant consideration of more than one ethnic group at any season of year in a
region" (1977:5). Syms’ model assumes that groups utilized large areas of diverse
ecological composition by a strategy of high mobility and technological adaptation.

Ceramics in Syms’ model provide the source of boundary information
reflecting a societal level of concern as more sensitive indicators of cultural partitions
(1977:12). Within these partitions are three levels of intensity of range/resource
utilization: core, secondary, and tertiary. The use of ceramics as the defining
elements of cultural boundaries is based on assumptions taken from ethnographic
analysis: i.e., that ceramics are sensitive indications of cultural boundaries produced

in a familial setting which reflects a societal level of concern for the use of style and

43
symbols (1977:12). Important to Syms’ model is the use of microstyle in which, for

example, Blackduck cultural materials represent the remains of a number of "related
but autonomous" groups (1977:106). For Syms (1977 :73) "ceramic variability...reflects
numerous co-existing groups during specific time periods as a result of evolutionary
change in stylistic impressions and movements of people."

For Syms, then, a ceramic tradition as it is perceived by archaeologists,
represents the maximum aggregate of an archaeological culture. However, on the
level of the site or subregion in which his core, secondary, and tertiary areas are in
operation, the archaeological record of diverse ceramics reflects the mobility of
various groups whose composition may include potters from more than one tradition
or who may be influenced by stylistic practices from more than one tradition
(accidental association in the archaeological record, notwithstanding).

Syms’ model is realistic and precise, but as a practical matter very difficult to
operationalize. It is necessary to assume, as Syms does, some level of correlation
between the uniformity of the material culture and a corresponding archaeological
culture as a first step in identifying the situations that deviate from it. Cautionary
notes often include warnings that archaeological cultures cannot be automatically or
mechanically linked with any ethnohistorical group (Trigger 1978:116, 117). This is
true, but it does not mean that there will be no correlation; obviously, there must be
some, or the use of style in archaeology is so inept that it invalidates all prior work.
This cautionary finger shaking has, according to Sampson (1988), resulted in a poor
environment in which to address questions of style and territory. Half of the issue

is addressed by the question: are there patterns which suggest the existence of

44

boundaries? The other half of the issue is framed by the question: what does the

absence of boundaries tell us about intergroup interaction?

STYLE

The issue of style is central to the study of archaeological cultures. Style, in
all its manifestations and definitions, permits the archaeologist to examine groups as
they identify and distinguish themselves from nonmembers. The blending of style or
the association of artifacts of different styles on the same site indicates rates and
kinds of interaction/integration among groups. Subtle messages encoded in the
stylistic elements of design may indicate the nature of social or personal relationships
within and between groups, as well as to serve as a measure of the balance of
personal and social identity through time (Michlovic 1981, Wiessner 1985).

Much of this has been learned from the study of living groups and, as such,
has been found applicable to archaeological examples, albeit with certain limitations
(Hodder 1979, Wiessner 1985). Formal definitions of what we call style in a broad
sense have been offered to partition more explicitly the concept into meaningful units
for analysis. Sackett (1982) considers the standard, isochrestic, and iconographic
levels of stylistic variation with respect to style. The standard approach is a
normative one focused on the "diagnostic" attributes of the artifact, where
morphological similarities are taken to indicate cultural-historical relationships. His
isochrestic approach defines style as a set of alternative ways of accomplishing the
same ends, or as Wiessner has chosen to characterize it, "choosing specific lines of

procedure from the nearly infinite arc of possrbility and sticking to them" (1985 :160).

45

Iconographic variation is defined as "specific elements of nonutilitarian formal
variation which function symbolically as a kind of social iconology to identify human
groups" (Sackett 1982:80). In other words, symbols contain information which
structures formalized relations.

Wiessner commented critically on Sackett’s classification of stylistic variability
(1984, 1985). She acknowledges that there is such a thing as isochrestic style,
particularly in the realm of technology. For example, ceramics may be constructed
from either the coil, paddle and anvil, or some other technique. The choice of one
technique should represent a conscious decision to replicate the previous choice of
other potters in the community, but either way one ends up with a pot. Sackett
clearly sees isochrestic variation as something larger than does Wiessner. The
difference is made up in Wiessner’s definition of stylistic variability where there is

truly a "nearly infinite arc" of potential variation:

Style, as I see it, has a behavioral basis in the fundamental
human cognitive process of personal and social identification through
stylistic and social comparison. In this process, people compare their
ways of making and decorating artifacts with those of others and then
imitate, differentiate, ignore, or in some way comment on how aspects
of the maker or bearer relate to their own social and personal
identities. Style is thus not acquired and developed through routine
duplication of certain standard types but through dynamic comparison
of artifacts and corresponding social attributes of their makers.
Stylistic outcomes project positive images of identity to others to obtain
social recognition. While isochrestic behavior functions to make life
predictable and orderly, stylistic behavior presents information about
similarities and differences that can help reproduce, alter, disrupt, or
create social relations. [1985:161]

46

Isocrestic variation, then, tends towards conservatism in technological aspects
of artifact fabrication while stylistic variation reflects dimensions of the social
atmosphere at the time of manufacture. Wiessner’s informants were not able, or
anxious to discuss specifically isochrestic variation but were eager to comment on
questions of style (1985:161).

Style is used to express corporate membership in some form of group,
theoretically at levels ranging from the immediate family to the maximal unit of
corporate membership. Messages are recognized by group members and outsiders.
Hodder (1979:446) suggests that the use of style as a means of asserting
simultaneously the inclusive and exclusive aspects of a society will intensify during
periods of stress. Although Sampson (1988) notes that the reverse relation has yet
to be demonstrated, sharply defined boundaries which separate stylistically
homogeneous populations of artifacts reflect more emphatic distinctions between
groups while heterogeneity and the absence of clear boundaries suggests a high
degree of permeability. Hodder also questions the applicability of a simple fall-off
model to explain the distributions of certain styles, citing the need to consider
"values" (e.g., social, cultural) in addition to distance and probability of interaction
(1979:447).

Wiessner’s definition of style and stylistic variability, which is based on her
study of metal projectile points of the lKung, directs this analysis of Terminal
Woodland interaction in the Lake Superior basin. In Wiessner’s work, style is seen
as a direct consequence of immediate social conditions which obtain from ongoing

relations with other human groups. The implications of this line of reasoning for the

47

Lake Superior basin, characterized by the widespread distribution of a variety of
ceramic styles, is the permeability of social boundaries. In this instance the relevant
categories of material culture are limited to stone, copper, ceramics, and rarely bone
and shell. Of these, ceramics, by virtue of plasticity and widespread appearance on
archaeological sites, are by far the most significant source of information on group
interaction.

Isochrestic variation would be an appropriate avenue if the temporal scope of
the analysis was extended to include such things as the shift from coil construction
of ceramics by the Laurel culture to the paddle and anvil technique by Terminal
Woodland groups in the same area, or to detect differences in copperworking
techniques among groups in different areas or at different periods. However, within
the time frame of the Terminal Woodland substage, there is no relevant application
of this dimension; isochrestic variation will be held as a constant in this analysis.

What could be construed as a major impediment to this analysis is the use of
archaeological cultures based almost exclusively on a cultural-historical definition of
style, and the use of these to examine relationships among them using Wiessner’s
definition of style. Fortunately, the two approaches are not exclusive. Wiessner
(1985:162) states that style is influenced by social contact, and that "exactly how it
[contact] will affect style depends on history, cultural context, and the nature of
relations." Unfortunately, the history of an archaeological culture, that is to say,
when and by whom the archaeological culture was first identified and its subsequent
ontogeny through excavation, analysis, synthesis, etc., presents a unique and possrbly

enormous source of bias (cf., Syms 1977). When, for example, does one

48

archaeological culture represent something approximating ethnic "purity," and when
does it represent a cultural hybrid?

In the archaeological case study of Isle Royale and the Lake Superior basin,
it is the maximal group or tradition that is reflected in the nomenclature of the
ceramic analysis. Therefore, it is the archaeological cultures that are the primary
focus. Ceramic subtypes will not be used, in spite of the fact that they are a better
representation of the subgroups within major archaeological cultures, and area a
better approximation of the groups discussed in the ethnographic and ethnohistoric
examples. This is due to the nature of the data base of the region which is extensive
in its distribution and diffuse in terms of its content. Therefore, this study is satisfied
to identify an item as belonging to the Huron, or Selkirk style group; an identification

that carries intended information on cultural-historical context.

RANGE AND TERRITORY

A consideration of boundaries as defined by the distribution of style groups
requires alike consideration of group range or territory. The two terms as they are
employed in this analysis reflect two dimensions of one phenomenon. Range
(sometimes referred to as "home range"; e.g., Jochirn 1981) is defined by the area
exploited on a regular basis in a group’s subsistence and settlement system. This
definition is based on the presumed correlation between the structure of sociocultural
groups as they are conditioned by the natural environment (i.e., distribution and types
of resources). In concert with an ecological definition of range, territory is seen as

an inclusive concept in which different levels of corporate membership are structured

49

on the social landscape. Range and territory are not coterminous; any given territory
includes a number of ranges reflecting in a parallel fashion the distrrbution of small
groups (i.e., families, hunting partnerships) that constitute the larger social entity
which, in turn, defines the territory to which a group feels affinity or from which it
derives part of its identity. Archaeological definitions which grapple with the
territory/ range issue are generally confined what is called here "range," as a set of
physical relations among other sites with or without similar artifacts (Sampson 1988).
Developing a cognitive application of territory with archaeological data is more
complex. Range is a significant concept as it relates to the relations of resource
distribution and the cultural means by which they are harvested, and while it would
not be a first choice in the study of a living group, it must be seriously considered in
the analysis of sociocultural boundaries and intergroup interaction. The range
concept is used to model territory when using archaeological data. An aggregate of
ranges which share stylistic affinities may approximate the territory archaeologically.
But even this is incomplete insofar as more than one group may claim a sacred or
special resource area as part of its cognitive territory.

The scope of a range or territory is delineated in terms of its corresponding
unit of sociocultural organization. Families, hunting units, macrobands, etc., will all
have some notion of territory, the definition of which is contingent upon actual or
anticipated needs of the immediate or anticipated corporate group. In this sense, the
definition of territory is analogous to that of group identity (J ochim 1981).
Therefore, the level of specificity in the analysis of artifacts will, in turn, determine

the corresponding level of corporateness reflected archaeologically. Within a single

50

tradition individual stylistic varieties, or subtypes, may represent internal social
partitions of the larger group. If, archaeologically, one is only able to assign artifacts
to a tradition and not the smaller subtypes, it follows that one cannot test
propositions regarding relations among the subtypes and how they relate to the
tradition as a whole, or to their counterparts outside the tradition. My argument
applies to this thesis in the decision to follow whole archaeological cultures,
synonymous with ceramic traditions, rather than the smaller subdivisions of ceramic
classification which, while more specific in detail, are less reliable indications of group
interaction by virtue of the problems inherent in ceramic classification at the ware
level.

In a highly successful look at band territories in South Africa, Sampson (1988)
is able to distinguish among several distinctive decorative techniques and designs on
pottery. For Sampson, a definition of range/territory is a prerequisite to the
delineation of seasonal settlement and subsistence. The more typical approach which
begins with a boundary imposed by a model (e.g., a hexagonal lattice), "opens up the
awkward possibility that the archaeologist has built two or more fragments of
seasonal rounds of neighboring bands into a spurious model of a single, but
nonexistent group" (1988:13). It is clear that, in some areas, the "spurious models"
are a necessary prelude to a more realistic view of sociocultural groups if anything
is to be said of intercultural dynamics. Sampson, with the benefit of a relatively
complete data base and tight spatial and chronological controls, has great control

over his data and is able to discern very fine distinctions of stylistic variation within

g!

'5».
IQ

“I:

51

what would appear under less controlled conditions as internal noise within a larger

stylistic community.

ARCHAEOLOGICAL EVIDENCE OF INTERACTION

What is true of the relative scope of regional stylistic traditions and their
smaller internal partitions is equally true of the scope of analysis permitted by
individual archaeological sites versus a number of sites distrrbuted over a large area.
Usually, it is impossrble to characterize the content of the interaction that occurred
at a given site which resulted in the commingling of artifacts of two or more
stylistically discrete archaeological cultures. However, at the next level of focus, the
subregional or regional distribution of these comminglings may produce evidence
suggestive of patterned relationships among groups. Only in the happy event of clear
cut superimposition or stratigraphic distinction may it be possible to state that certain

types of interaction did not take place, and that the presence of both styles is the

 

result of two events unrelated in time. In areas of pronounced social stratification
where artifacts take on affectations of class membership (e.g., Hopewellian and
Mississippian cultures), it may be possible to determine social access within a single
community. The potters from the various archaeological cultures of the Terminal
Woodland may have symbolized various types of corporate statements, made
comments on others within the group, or to those outside, but it is limited in content

to statements which are essentially egalitarian in nature and directed to a likeminded

audience.

52

It now remains to show the linkage between the questions asked of the data
and the archaeological record. How is interaction represented in an archaeological
site, and is it possible to determine the nature of the interaction? This question may
be hierarchically addressed from the perspective of artifacts, individual sites, and
populations of sites within the region as a whole.

An objective of most archaeological studies is interpretation of the behavior
of prehistoric human populations which occupied a given area. For example, lithic
debitage represents an important and durable record of prehistoric behavior. Analysis
of lithic waste is designed to collect data which can be used to make inferences about
the activities of subsistence, tool manufacture and use, and interaction with other
areas and groups in the form of long distance acquisition of discrete lithic raw
materials (Luedtke 1976). Ceramic artifacts encode complex elements of style which
are used to interpret the use of symboling as a means of distinguishing among
archaeological cultures (Syms 1977). The association of certain ceramic types and
certain lithic raw materials provide a measure of a prehistoric group, indicating its
territory/range, sphere of potential interaction with other groups, and something of
how it perceives itself relative to other groups.

The analysis of the prehistoric artifact assemblage is accomplished through
several lines of inquiry, one of which is the delineation of stylistically oriented bodies
of material representations in ceramic and lithic artifacts with boundaries in space
and time. As a model these archaeological cultures can be further broken down into
smaller social segments (e.g., bands, families, task groups) which correspond to our

understanding of seasonal demographics and their social cognates. Interpretation of

53

the nature of interaction is based on the aggregate evidence of stylistic attributes and
discrete raw materials from the region as a whole as well as the more specific
approach of interassemblage/intersite variability. Domestic activities are represented
by ceramics, by evidence of tool maintenance and production, by the presence of
discarded tools and tool fragments, and by the presence of features and food remains
indicative of subsistence activities and season of occupation. More specialized
activities are indicated by a narrow range of artifact types, suggesting only a few
specific activities at a given site. For example, copper mining is recognized by the
association of hammerstones, flakes from hammerstones, and mine pits. Copper
fabrication, on the other hand, is represented by copper waste and discarded or lost
copper objects in various stages of production. It is the configuration of the artifact
assemblage that permits the analyst to address 1) the activities undertaken at the site,
2) the composition of the group, and 3) their subsistence practices. It also allows the
generation of general propositions concerning the relationship between Isle Royale
and coeval cultural groups elsewhere in the Lake Superior basin.

Regarding the relationship between copper and subsistence activities, a stance
of parity between them will be assumed. It is unlikely that anything can be gained
by trying to determine if people mined copper and did a little fishing on the side or
if people fished and mined a little copper. The relation is a dynamic one, the
important aspect being the systematics of the relationship. In the case of Isle Royale
copper, some estimation of the intensity of copper mining in the Terminal Woodland
is necessary to assess the allocation of labor and group composition (demographic,

not ethnic) in an intensive vs. nonintensive strategy for copper acquisition.

54

Evidence of interaction comes from assemblage variability which involves the
stylistic attributes of artifacts and raw materials believed to originate outside of a
perceived range or territory. Style is used to identify geographically and temporally
bounded populations of artifacts believed to be collectively understood bodies of
symbols of group membership-- whether internally or externally directed. The extent
to which the symbolic representations of style are internally coherent and exclusive
is one measure of a group’s membership (of potters, minimally). Style (and
technique) may, however, be introduced from outside the group by means of several
agencies including band exogamy and outright copying, or the items displaying the
stylistic attributes may be obtained from a producer outside the group. One example
is the hypothesis that what is stylistically Huron in design in the Lake Superior basin
is actually copied from the Huron by Ottawa and Ojibwa (Mason 1976, Ramsden
1989, Wright 1968). The introduction of pots and potters from outside the group or
area by a variety of vehicles is an important facet of assemblage variability and is
assessed in light of the many competing hypotheses regarding this phenomenon.

Ceramic heterogeneity has invoked explanations of 1) group marriage and
residence rules (e.g., McPherron 1967, Whallon 1968, J .V. Wright 1972), 2) raiding
and warfare (e.g., Ramsden 1977, Trigger 1984), 3) stylistic mimicry (e.g., Fitting
1975, Mason 1976, Ramsden 1988), and 4) trade and exchange (e.g., Fitting 1975,
J .V. Wright 1972a), all of which are potentially viable causal factors. All presuppose
some level of mobility. Mobility of group segments is required for adjustments to
perturbations in the subsistence base, or as a response to changes in the sociopolitical

environment. Mobility can be juxtaposed with residential stability in order to

55

establish an internal dialectic between questions of range or territoriality and culture
contact which can occur within, at the edges of, and outside of any sociogeographical

COIlStht.

ARCHAEOLOGICAL IMPLICATIONS OF INTERACTION

Given the wide range of interaction alternatives, what are the implications for
the material evidence of interaction as preserved on an archaeological site?

Ambivalent relations should not leave any archaeological evidence because
physical interaction is generally lacking. Archaeological deposits resulting from
parties exercising ambivalent relations should reflect only the membership of their
respective groups. However, the relation implies indifference manifest in avoidance
without the necessity of range or territorial closure. No marked cultural boundaries
will be expressed. It is likely that small ambivalent groups will exhibit a pattern of
site use that is, at least in part, inclusive, resulting in the mixing of occupations in a
single site. This effect mimics relations more friendly and more interactive than may
have actually been the case. Trade and/or exchange are not precluded in ambivalent
relations, particularly when the items involved serve to buffer the local scarcity of
essential resources (Luedtke 1976).

In amicablelcooperativelintegrative interaction artifacts from more than one
archaeological culture are associated on a site as a consequence of mixed group
composition (i.e., a co-residence unit), or as evidence of trade/exchange between
groups. Mixed assemblages are, under these conditions, widespread indicating

permeability of whatever boundaries may be present. Diffusion is likely to occur

56

under these circumstances with potential hybridization or mimicry of style elements
and techniques. Raw materials may also be quite diverse, reflecting the resources of
more than one territory. Trade and exchange are predicted under conditions of
amicability, as are patterns of marriage between groups and transmittal of stylistic
elements or techniques. Extreme diversity in material remains would be predicted,
assuming that the interacting groups are distinctive in their material expressions. A
lack of distinctiveness in style or material between groups may cloak the actual nature
of the interaction.

Although trade is known to exist between hostile parties (e.g., Ford 1972), it
is expected that, on a regional scale, evidence for antagonisticlnoncooperative
interaction is limited in scope. Physical evidence occurs at the territorial margins as
an indication of exclusivity. The actual evidence of violence, well documented
historically, will be virtually invisible archaeologically, except in the case of sites which
contain the physical remains of humans with forensic evidence. At the level of
individual sites this type of interaction will be reflected in a suite of artifacts that is
derived solely from the group membership, and at the regional level, from the
inference of pronounced boundaries. This does not imply that such assemblages are
necessarily expected to be homogeneous in stylistic or raw material content: group
membership may be simultaneously multiethnic in one dimension and exclusive in
another. It is expected that assemblages will tend towards homogeneity as a function
of a heightened sense of corporate exclusivity. Skeletal remains notwithstanding, the
archaeological evidence for antagonistic/noncooperative interaction will not be

apparent on individual sites, but must be inferred at a regional level since the

57

essential characteristics manifest in the material remains will mimic certain aspects
of ambivalent and amicable relations.

Regionally, aggressive interaction is readily recognized in a distribution of
fortified villages and in human physical remains in which direct evidence of the cause
of violent death is preserved. Territories and ranges should be firmly drawn with
possible use of buffer zones (e.g., Hickerson 1970). Sites in a buffer or contested
zone are limited to camps of short duration which would likely be interpreted as
small hunting camps by most archaeologists. On the surface of it, one might expect
absolute exclusivity of assemblage raw materials and stylistic components, formally
indistinguishable from the conditions predicted for ambivalent or amicable relations.
However, if the aggressive interaction involves the taking of captives, especially
female potters, and their subsequent adoption into a group, one’s expectations shift
to accommodate an influx of stylistic variation. Raw materials, on the other hand,
should continue to remain exclusive.

Except in the rare cases where the archaeological record leaves no doubt, each
interaction alternative has the potential for creating an archaeological assemblage
with mixed stylistic elements. Raw materials could be an indication of territorial
boundedness, were it not for the confounding influences of trade/exchange across
boundaries. A more parsimonious explanation of discrete distrrbution of raw
materials is that of a core area (after Syms 1977) in which certain raw materials are
not part of an exchange or trade system with neighboring groups.

A fundamental difficulty with this application of interactive alternatives, and

with all models which attempt to deal with the interpretation of the archaeological

58

record of a region which has several archaeological cultures, is the dimension of time
and site formation processes. Collapsed stratigraphy, as it has been called by
Canadian archaeologists (Hamilton 1988, Reid 1988, Syms 1977), is caused by the
repeated occupation of a site over a long period of time, resulting in the apparent
association of items which, in their systemic context had no immediate relationship
to each other. Excavation strategies which are designed to accommodate this
phenomenon by utilizing 2 cm vertical control, or by piece-plotting each item as it is
excavated, only partially succeed in clarifying the site formation processes. In this
analysis, the associations of artifacts on a site are interpreted as evidence of actual
or potential interaction, irrespective of possible variations in the systemic processes
responsible for their entry into the archaeological record.

The lack of a more fine-tuned chronological partitioning of a data base creates
the possibility of a serious misreading of the archaeological record. For example, in
the 700-800 year time span involved in this analysis, the presence of more than one
style of ceramics on a site or in a region can be explained by a wide variety of factors
not necessarily contingent on the types Of interaction being considered here.
Environmental determinants which condition the movement of groups are not
invoked, for example. The lack of more precise chronological controls could be
compensated for by the use of ceramic varieties which reflect smaller groups, smaller
areas, and smaller increments of time. However, the ability to make accurate
determinations at this level are suspect. Further, the assessment of potential

interaction among Terminal Woodland substage groups requires a larger data base

59

than is permitted by a more particularistic approach; we are studying the forest, not

the trees.

ARTIFACI‘ ANALYSIS

For the archaeologist it is the configuration of suites of associated artifacts
that suggest relationships among archaeological cultures. The raw materials from
which artifacts are made have sources that are almost always restricted in space. The
physical treatment of these materials through a variety of technological operations
transforms it from a raw substance into something useful such as a tool or ornament.
The handling and use of the finished item and its movement within a single group,
or its movement between individuals or groups as an item of exchange or trade
implies, on the one hand, an estimation of value of the item in its systemic context,
and the nature of relationship between interacting parties on the other.

The following section specifies the types of raw materials, and some of their
technological transformations which permit the definition of the material assemblage
specific to an area and, by extension, an archaeological culture.

LITHIC ARTIFACI‘ S: The raw material of lithic artifacts reflects the
movement of people and/or goods in the Lake Superior basin. Identification of lithic
raw materials is observed for all lithic artifacts with the understanding that the
resulting data are based on macroscopic observations. We are fortunate that in the
western Lake Superior basin major sources of lithic raw materials (Figure 1) are
relatively few and these tend to have discrete qualitative characteristics which

minimize the potential for confusion between source identification made

60

macroscopically. The lithic raw materials, including trace-cherts, encountered on Isle
Royale and the Lake Superior basin are described below.

Bois Blanc chert (BBL): Primary sources for BBL are found on Bois Blanc
Island (Mackinac Co., Michigan), Bruce County, and Haldimand County (Ontario).
The geological association is the Lower Devonian Bois Blane Formation where it is
found as nodules and lenses in primary contexts. It is also known from secondary
deposits in the form of pebbles. Luedtke (1976:210) descrrbes BBL from the
Juntunen site as having "large mottles and areas of limestone within the chert."
Colors range from pale gray to cream N8/ to 2.5Y7/2 to 2.5Y8/2 to N7/; occasionally
medium to darker gray N6/ to 5Y6/1 (Munsell 1975). The texture is fine to coarse,
luster is dull, material is opaque; drusy crystals may occur in cavities and along
planes. Luedtke also notes the "striking similarity between Bois Blanc, Onondaga,
and Amherstberg cherts," (1976:211). There may also be some visual confusion with
the Fossil Hill/Collingwood cherts from northern Lake Huron.

Qunflint Silica (GFS): GFS is found in primary outcrops from the
Ontario-Minnesota border to the east of Lake Nipigon. Kakabeka Falls on the
Karninistikwia River near Thunder Bay is the source of a banded variety. The
geological association of GFS is the Gunflint Formation of the Anirnikie group where
it occurs in tabular form (Pye 1969). It is also known from secondary deposits as
till. McLeod (1978:168) describes it as having a variety of colors: clear to white, light
gray, dark gray, blue, and black. Mottled or granular inclusions and thin dark bands
are known. Dave Arthurs (personal communication 1987) describes it as "a milk white

or virtually transparent at its best, but can also be a near opaque grey or blue. The

61

 

HBL

 

HBL

 

 

Figure 1. Lithic raw material sources.

 

62

identifying characteristics are the waxy lustre and the presence of black spherical
inclusions that give it a peppery appearance." It is an excellent quality material,
common and widespread on archaeological sites on the north shore of Lake Superior,
Isle Royale, Lake of the Woods, and Rainy River. McLeod (1978:168) suggests that
GFS was used primarily in Woodland substage sites. On Isle Royale GF S is known
from most Woodland substage sites as debitage and unifacial tools. Rossport chert
(RSPT), a typically homogeneous black raw material, can physically intergrade with
GFS.

Hudson By Lowlands chert (HBL): HBL is found in a wide variety of
secondary deposits as beach and river cobbles and pebbles. The primary source(s) is
far to the north in the Hudson Bay lowlands whence this material was glacially
transported. In the Lake Superior area HBL is strictly a secondary source, exhibiting
highly variable characteristics in quality and color. A tan or yellow cortical rind is
common but not diagnostic. Internal flaws are common but many pieces are
homogenous and of excellent quality, albeit small in size. The bipolar reduction
technique is often applied to HBL which satisfies most requirements for small tools
such as scrapers. Few pieces of HBL are of sufficient size for larger tools, however.
Occasional pieces of agate may be included in this category. A brown translucent
chert often identified as Knife River flint (KRF ) is believed to originate in northern
Ontario, north of Lake Superior (Griffin and Quimby 1961:98). Steinbring
(1974:68-70), in his discussion of Pickerel Lake site artifacts fiom Quetico Provincial
Park, suggests that the "pre-Cambrian algal domes in the vicinity of Schrerber,

Ontario, northeast of Thunder Bay, might well yield comparable variations." Indeed,

63

chemical characterization of some of the KRF mimics by J ulig, Pavlish, and Hancock
(1988) indicates that much of what has been identified as KRF clusters more closely
with the HBL samples than with the KRF. To assess this visual similarity chert
samples from Isle Royale were analyzed at the University of Toronto SLOWPOKE
reactor facility. Findings indicate that the KRF look-alikes from Isle Royale are also
chemically similar to other HBL samples (Julig, Pavlish, and Hancock, personal
communication 1989). On Isle Royale this variety of HBL frequently co-occurs with
J ST and GFS, suggesting that its point of entry to the island is from the Thunder Bay
area to the north. This KRF-like chert has been found at several Terminal
Woodland sites on Isle Royale. While in proportion it falls behind the other
Thunder Bay area sources, it is often found in association with them.

Jasper Taconite (JST): Jasper taconite has a distrrbution similar to that of
GFS; from St. Louis County, Minnesota east to Rossport, Ontario (Steinbring
1974:67-68). It is also found in the Animikie group of the Gunflint Formation where
it occurs in tabular deposits (Pye 1969). J ST is also available from secondary sources
in the western Lake Superior basin where it occurs as beach cobbles. McLeod
(1978:167) descrrbes it as a "reddish granular chert, which, depending on the iron
content (up to 28%) and the degree of oxidation of the iron present, may vary in
colour from bright red to almost black. When held to the light the granules are
visible." Dawson (1983:8) reserves the term "jasper taconite" for the red specimens
only, distinguishing between them and green, gray, and black taconite. Gradations
between colors may occur on single specimens, and it is not uncommon on Isle

Royale to find specimens of J ST that grade into GFS. Confusion between J ST and

64

cherts from other sources is highly unlikely. J ST is a common raw material on Isle
Royale sites and has been found at most sites dating to the Woodland stage.

Kakabeka Chert (KAK): Another material of the Animikie group in the
Kakabeka formation of the Thunder Bay/Port Arthur area, KAK occurs in river cut
gorges of the Kaministikwia River. Pye (1969:15) describes it as closely spaced chert
layers with banded carbonate rocks, forming chert-carbonates. With equal access to
GFS and J ST, KAK may have been a less desirable raw material, owing in part to its
cleavage plane fracture properties and relative lack of homogeneity.

Onondaga Chert (ONON): Onondaga chert has its source in the Onondaga
formation of the Niagara escarpment of New York and southern Ontario where it
occurs in both primary and secondary contexts. Luedtke (1976:273) descrrbes it as
"mottled and streaked and ranges from very dark grey, N3/ to N4/, to lighter greys,
N5/, N6/ and 10YR5/1 to 10YR8/1. Texture is fine to medium, the luster is shiny,
and the material is opaque. Fossils, bits of limestone, and crystal filled veins and
cavities occur as inclusions." There is only one incidence of Onondaga chert on Isle
Royale at the F inn Point site (201R5) where a notched projectile point was found
(Bastian 1963a:367-368).

Portage Lake martzite (PLO): The only primary source of chippable stone
known on Isle Royale comes from fine-grained arkosic quartzites interbedded with
the Portage Lake volcanic series (Terry Keith, personal communication 1990).
Primary and secondary sources are localized and of variable quality. Primary
outcrops examined yield material with pronounced planar fracture properties.

Secondary sources in the form of water-worn beach cobbles are more homogeneous

65

with good conchoidal fracture. PLQ is a dense, micaceous quartzite with very fine
particle size and some bedding. Color is dusky red 2.5YR3/2, homogeneous or with
lighter bands. On Isle Royale PLO has been found at both shoreline and interior
sites, but exhibits a higher absolute and proportionate frequency at Archaic sites on
interior locations. A similar and probably related lithic material called Srbley
sandstone is found on the north shore of Lake Superior.

Quartz (QTZ): In one form or another, quartz is ubiquitous in the Superior
basin, supplying some areas (e.g., the western Upper Peninsula of Michigan) with
significant amounts of chippable stone. QTZ can be found in primary igneous,
sedimentary, and metamorphic deposits, and secondarily in river, lake shore, and till
contexts. On Isle Royale, however, QTZ is not typically found in a state of sufficient
homogeneity rendering it suitable for tool production. No tools made of QTZ are
known from Isle Royale and it is rare as debitage, in contrast to lithic assemblages
from the south shore of Lake Superior and interior sites where it is common.

Rgsspgg (RSPT): Rossport chert occurs in the Gunflint Formation where it
intergrades with GFS and J ST. Known exposures of RSPT are found near the town
of Rossport, Ontario west to Whitefish Lake (Dawson 1980:59). Dawson (1974:69)
describes RSPT as "a black chert with a slightly shiny surface, opaque diaphaneity
and of massive structure." As noted by Dawson (1974:69), RSPT is "reminiscent of
siliceous coal." J .V. Wright named this raw material type (1967:31) and has identified
it in the assemblage from the Terminal Woodland levels at the Michipicoten site
(1968). On Isle Royale RSPT is a common constituent on Woodland stage sites

where other of the Thunder Bay area raw materials occur. It should be remembered

66
that RSPT is an archaeological taxon for a raw material type which physically

intergrades with GFS and JST, and it is not uncommon to find artifacts which
demonstrate this feature.

COPPER ANALYSIS: There has been little interest in the analysis of copper
waste categories, attention being directed towards the formal artifact categories.
Bastian (1963a, 1963b) distinguishes between worked and unworked scrap, and
identifies as intermediate categories bent strips and bars. Hoxie (1980) makes a
distinction between flat hammered chips and nuggets, but found no items which he
recognized as intermediate forms. Most of the literature dealing with copper artifacts
is concerned with finished products, usually in contexts removed from the sites of
their original production (e.g., Penman 1977, Wittry 1957).

A notable exception is Vernon’s (1985) paper on the archaeometallurgy of
the Old Copper industry. In it he studies the microstructure of a series of copper
items including finished tools, ornaments, intermediate categories, and unworked
pieces. A substantial contribution of his work is the identification of stages in
fabrication represented morphologically and through the microstructural attrrbutes.
Cold hammering and annealing as conjoined techniques of copper working were
employed in the manufacture of blanks which anticipate a range of final products,
and in some of the formal tool categories. Vernon’s findings indicate that
cold-working of the metal results in a greater hardness, while annealing has a
softening effect. The cold-hammering of previously annealed pieces obscures the
structural evidence of annealing and returns the copper to a work-hardened state

(1985:156). Although rarely mentioned in the context of aboriginal copper working,

67
the technique of slaking hot metal probably followed the annealing. Slaking, or the

immediate immersion of the metal into a cool liquid bath, renders the object
available for immediate hammering.

In his sample of 15 blanks from an unspecified site in Houghton, Michigan,
Vernon found that most were made "by flattening a piece of copper, folding it over
and flattening it again, and repeating the process in a buildup fashion to the
completion of the blank" (1985:156-157). Observations made on finished tools
suggest that awls were finished in a cold-hammered state, whereas the knives, points,
and adze were finished in a fully annealed state (1985:158-159). He goes on to
consider the cultural implications of his findings, stating the possrbility that items used
for technical purposes, such as the awls, were cold-hammered to improve their
functional qualities while the other artifacts, such as ornaments or large spear points,
were employed in symbolic roles not requiring a final hardening.

Vernon’s contribution to the understanding of the relationship between
cold-hammering and annealing has important implications for copper working of later
periods in prehistory. In the absence of specialized microstructural analysis we must
employ a classification based on macromorphological observations.

In this analysis the extraction and fabrication of copper is modeled after
reduction trajectories for lithic debris, and is derived in part from the replication of
several tool and ornament types. Waste categories analogous to lithic debitage,
intermediate stages of fabrication, and final products are identified in native copper
assemblages. This classification permits the copper industry to be examined relative

to other activities, such as flintknapping or subsistence related pursuits. Proportions

68

of the various categories should indicate the intensity of initial copper working versus
tool or ornament fabrication, and proximity to copper mines. Waste/initial
fabrication categories (Clark 1989a:156-158) defined below are based on a
combination of replication by the author and the examination of archaeological
specimens:

1) RAW WITH MATRIX: These items show no evidence of hammering.
Copper is present with basalt, calcite, epidote, or other nonmetallic matrix. This
category represents an initial stage of processing or extraction. These pieces are of
little value for further modification if proliferated with matrix due to the difficulty of
bonding small pieces of copper and will likely be discarded at or near the site of
origin. If larger amounts of copper are present, the piece will likely enter the flat,
vesicular category.

2) RAW WITHOUT MATRIX: Evidence of hammering is absent but the
piece is homogeneous, lacking nonmetallic matrix. The surfaces of items in this
category may be smooth (e.g., a nugget) or rough. Homogeneity is a desirable quality
and even a small homogeneous piece of copper can provide the basis for a variety of
small tools or ornaments. Transport of homogeneous pieces to fabrication loci is
considered likely.

3) N UQQET, HAMMERED: Items in this group are homogeneous with
relatively smooth surfaces, and exhibit evidence of hammering without being
conspicuously flattened. Folds or laminated features are absent. This category

represents an initial stage of manufacture and is likely to be found at fabrication loci.

69
4) FLAT VESICULAR: This category represents the hammering of the "raw

 

with matrix" stage. The pieces are hammered flat causing the nonmetallic matrix to
fall out, resulting in a "swiss cheese" effect. Depending on the initial size of the piece
being worked and the degree of homogeneity the resultant product may be discarded
due to the problems of bonding small pieces of copper. Items in this group may be
found at both extractive sites where initial fabrication takes place, or at later stage
fabrication loci to which raw pieces with matrix have been transported.

5) FLAT HOMOGENEOUS: These are pieces which are hammered flat but

 

which lack laminated structure. At this stage many will exhrbit the beginnings of a
rectangular profile with a bipolar and quadrilateral orientation to the hammered
edges. This is a highly desirable stage which can anticipate any final product, size
and internal structure being the only limiting factors. This category is more likely to
be recovered at fabrication loci. This group is differentiated from bars (see below)
on the basis of the more formal quality of the latter (of, Vernon’s "blanks";
1985 :155).

6) FLAT LAMINATED: Items in this category can be either single sheets

 

of thin hammered copper or layered sheets. Some are foil-like while others may be
more dense: in either case they represent the limits of hammering single pieces and
must be folded and/or annealed if they are to be used further. These items are used
in the production of rolled tubular beads and small projectile points. It is also
common to find awls and other larger items built up from a few flat sheets. Small

flat laminated pieces represent copper debitage produced in the process of

70

hammering larger homogeneous pieces of copper and should be anticipated as a
numerically abundant category on all sites

where fabrication is undertaken in addition to extractive sites involving the removal
of copper from large masses.

7) ROLLED, LAMINATED: This category may represent the attempt to
build up one or more flat pieces by sequential folding and hammering. A successful
outcome utilizing flat sheets requires annealing since cold hammering of thin sheets
will result in a brittle product. Discarded rolled tubular bead failures may also be
included in this group, especially if crushed postdepositionally. This group is
anticipated only at fabrication loci.

8) MRS: Bars are frequently encountered and represent a general preform
or blank for the production of virtually all final products other than small rolled
beads. They are typically rectangular and may be homogeneous or laminated in
structure. Copper bars can be found at fabrication loci as well as at sites removed
from the copper source areas since it is in this state that much of the copper could
be stored or transported without committing it to a final form.

Formal tool and ornament categories follow the extant literature (e.g., Wittry
1957). Awls, beads, projectile points, crescentics, and "butterknives" form the bulk of
the finished copper products in the region, although other categories including hooks
and gaffs are known.

An important aspect of the analysis of copper artifacts is the consideration of
temporal/cultural variability in formal types of finished items. For example, is there

any significant difference between awls with round versus quadrilateral profiles, or

71

between rolled tubular beads and heavier bent strip beads? Such questions may not
be answerable based on data from Isle Royale alone and need to be addressed on a
regional scale.

The role of copper as a unique resource with limited distribution may have
been an important factor in the interaction between Superior basin groups. It is
necessary to examine the value of copper in the broader context of the total material
culture to determine the extent to which it was an item required for subsistence
technology, a vehicle for attaining or demonstrating status, or a relatively unimportant
luxury. Except for its use as ornaments, virtually all tools manufactured from copper
could be duplicated in stone, bone, and wood. Yet copper is known from widespread
Terminal Woodland sites. Some consideration of its geographic distribution around
the Superior basin is necessary to determine the extent to which copper may have
"drawn" residents of the north shore to Isle Royale. Lake Superior copper is treated
in detail in Chapter 6.

OERAMIC ANALYSIS: The attributes utilized in the analysis of ceramic
artifacts are selected to classify sherds and the vessels they represent into a larger
taxonomic category or type. Materials utilized in manufacture, vessel shape, surface
treatment including decoration are believed to encode the stylistic information
resulting from the cultural context of a ceramic tradition and the temporal placement
of the site from which the pottery was recovered. Further, neutron activation analysis
(NAA) is used to determine the relationships between certain style groups and pots

from archaeological sites around the Lake Superior basin. Specific hypotheses

72

regarding the anticipated outcomes of the analysis are presented together with the
results of the analysis in Chapter 7.

Since whole vessels are rarely recovered the analysis of prehistoric ceramics
necessarily involves a classification of vessel fragments or sherds. In this analysis
sherds are identified as to type, referring to the portion of the vessel from which they
came. Rim, neck, shoulder, body, and basal sherds represent all possible types of
vessel fragments. Where two areas are present on a single sherd, such as one from
the rim and neck, or from the rim, neck, and shoulder of a vessel, the sherd is
classified in the uppermost category, in this case as a rim sherd. Counts are recorded
for each category. Thickness is measured in millimeters where both the interior and
exterior surfaces are preserved intact. Tempering material is identified as to mafic or
felsitic origin and a maximum measurement is recorded in millimeters. Temper
density is estimated for each vessel. These descriptive data are recorded elsewhere
(Clark n.d.) and are not presented in this thesis.

Vessel shape must be inferred from fi'agmentary evidence as well. The
morphological characteristics are thus described for those items which exhibit
morphological attrrbutes, especially the relation between the shape of the lip, and the
attitude of the rim relative to the neck and body. These observations are possible in
a relatively small number of cases.

Surface treatment is recorded for both exteriors and interiors of sherds. In
the Isle Royale assemblage interiors are uniformly smoothed; exteriors, on the other
hand, exhibit more variability. These may be smoothed, polished, cord-marked,

smoothed over cord-marked, or fabric impressed. Surface treatment is considered an

73

aspect of pottery manufacture and is treated apart from decoration. Additional
observations relating to surface conditions include the extent of erosion and the
placement and extent of cooking residue.

Decorative attributes are variable and, taken in concert with the morphological
variables, provide the basis for typological analogy to extant ceramic traditions in the
Upper Great Lakes region. Type of decoration, its placement on the vessel, and its
placement relative to other decorative motifs on the same vessel help to define
individual vessels. The use of incising, impression by various means (e.g.,
cord-wrapped stick or push-pull), and punctations or bosses constitute most of the
decorative technology seen on the Isle Royale ceramics. Characteristic ceramic
attributes are described in Chapter 5 for each Terminal Woodland archaeological
culture.

The functional definition of "vessel" used in this analysis is a sherd or group
of sherds which can be confidently demonstrated to represent a single pot. Typically,
vessels are defined on the basis of rim to shoulder sherds where decorative elements
indicate unique attributes. Occasionally, undecorated or decorated body sherds can
be assigned an individual vessel status on the basis of discrete traits such as temper

or surface treatment.

SITE CLASSIFICATION
In order to understand the organization of demographics and labor in
prehistoric cultures it is necessary to classify sites in terms of the activities undertaken

there. The evidence for activities is inferred from both direct and indirect sources.

74

Artifacts, food remains, and features are a direct indication of the different activities
undertaken at a given site. Site locations may offer some indications of potential
activities by virtue of their proximity to specific resources or their relation to logistical
features of the landscape.

There is not always a direct correspondence between the evidence and the
actual activities conducted at the site. The multiple reoccupation of the same site by
the same or different groups gives the impression of higher density of artifacts and,
hence, of activities conducted there. In addition to this problem of site formation
processes is the extreme variability in data recovery. Single artifacts found on the
surface during a survey often reveal more complex deposits when subjected to a more
intensive investigation involving subsurface testing. It has been the experience of
many archaeologists to find their initial interpretations radically altered by subsequent
investigation on the same site or region, even where a comprehensive sampling
strategy was employed (e.g., Flannery 1976).

As a consequence, any site classification must be seen as a heuristic device by
which broader generalizations regarding the issues mentioned above can be
addressed. The classification is tentative in an absolute sense, but represents an
organizational scheme that incorporates assumptions about land use that is drawn
from ethnohistoric and ethnographic studies of how human groups allocate their
energies and organize themselves on the physical landscape.

Nine site types and their criteria for classification are presented below. These

terms are applied to both surface finds and subsurface deposits as descriptive devices.

75
0) Oncertg'n Stags: This category is required to account for a site whose

location or content is not adequately known. It applies to sites recorded in the past
for which no adequate documentation exists.

1) Sm; Find: The discovery of a single artifact or feature of any kind
constitutes a spot find. The finding of a single artifact should be accompanied
whenever possrble by close examination of surrounding areas for additional materials,
both surface and subsurface. Single artifacts represent the smallest of the "limited
activity sites." This is often a tentative classification, since a revisit to a site is likely
to encounter additional artifacts.

2) Lithic Scatter: The finding of an array of lithic artifacts relating to
flintknapping activities without any associated domestic materials constitutes a lithic
scatter. Artifacts may include any number of flakes, cores, and small (non-mining)
hammerstones. Whole, broken, or exhausted tools may indicate an acerarnic
occupation site and are, therefore, not included in the lithic scatter category. This is
a somewhat discretionary classification, requiring that a discrimination be made
between an acerarnic occupation and a lithic workshop.

3) Pot Break: As the name implies, this is a type of spot find, identified by
the recovery of numerous sherds which represent the breaking of a single vessel.
Single sherds fall under the category of spot find. Any additional associations of
lithics, copper, features, or sherds from a second pot constitute an occupation site.
Note that the term "pot break" can be used in the context of larger occupation sites

to refer to a discrete ceramic feature.

76

4) Occupation: This broad category includes the association of lithics and/or
copper and/or ceramics and/or features. Unless the evidence indicates a highly
specialized activity, the assumption is that occupation sites are the focus of domestic
and ancillary activities relating to habitation, subsistence procurement/processing, and
related technologically oriented functions. Occupation sites are differentiated from
mines, workshops which specialize in a limited range of technical activities, and
mortuary sites. However, any and all potential activities may be associated with an
occupation site.

5) MineZOuaLry: Copper mines and lithic quarries constitute a specialized site
type. Although they usually are discrete in terms of activities represented by artifacts,
they do occur in conjunction with occupation sites in some instances (see below).
Artifacts consist of hammerstones and waste materials in association with relevant
geological features, in the case of Isle Royale, copper deposits.

6) m: Unless associated with occupation sites, burial sites are seen as
strictly the site of the deposition of human remains and associated grave furniture.

7) Ogrpatioleine: This is a composite site type where evidence for
domestic activity supporting mining operations is present.

We have, then, specific expectations of the archaeological record as it pertains
to interaction among groups with distinctive material expressions and raw materials.
These have been discussed in terms the classification and analysis of lithic, copper,
and ceramic artifacts. Through this analysis archaeology defines its archaeological

cultures which, while different from ethnographic cultures, permit the pursuit of

77

interaction data at a general level which recognizes not the smallest but the largest
aggregate of social membership.

In the following chapter, archaeological cultures as analytical entities are
defined. A cultural-historical overview for the Upper Great Lakes and the
archaeological cultures of the Terminal Woodland substage are presented as an
introduction to the evaluation of interaction among groups, and the role of copper

as a potentially mediating resource, on Isle Royale in Lake Superior.

CHAPTER 4

ARCHAEOLOGICAL CULTURES AND CULTURAL-HISTORICAL OVERVIEW

INTRODUCTION

The following discussion is concerned with archaeological cultures and their
relationships in space and through time. As we have seen, archaeological cultures
differ from historically documented cultures in several important ways. Smaller
internal subdivisions including single or associated extended families are among the
most important economic units in the historic ethnic populations of this area, but are
generally below the level of archaeological visrbility. As a consequence,
archaeological cultures are typically much larger, more inclusive, and of longer
temporal duration than their historic counterparts. As is evident in the discussion of
the ceramic artifacts in particular, the attention paid to style, and its association with
time and place, is the primary means by which the identification of archaeological
cultures is pursued.

Apart from the obvious need to exercise control over the spatial and temporal
parameters of the data, the definition of archaeological cultures is an organizational
prelude to the study of interaction among prehistoric groups. In other words, it is
necessary to have defined analytical units in order to assess the extent and nature of
interactions among them. For Isle Royale, this study of interaction is restricted to
the Terminal Woodland substage.

Although the documented prehistory of Isle Royale begins over 4000 years
ago, sites dating as early as 8000 BR dot the Lake Superior basin. Following the
retreat of the last glaciation prehistoric hunters found ample prey among the caribou,

78

79

bison, elk, and moose in an environment then in transition from a periglacial tundra
to the present southern boreal forest.

The archaeological remains are largely confined to stone tools and waste
material left over from stone tool production. Large bifacial projectile points with
distinctive flaking patterns are typical of these early cultures. The classic fluted
points, often found associated with the remains of mastodon, mammoth, or bison are
absent from this area. The first recognizable archaeological culture is the Plano
Tradition, noted for the use of collateral flake scars on finished tools. The absence
of Plano Tradition artifacts from Isle Royale is attributed to the high lake levels
which prevailed at the time (ca. 8000-5000 B.P.). On the north shore of Lake
Superior Plano artifacts have been found at a number of sites preserved, in part, by
the effects of isostatic rebound of the shoreline which have elevated the already high
beaches well above the current levels of Lake Superior. Notable are the Brohm and
Cummins sites near Thunder Bay, Ontario (Dawson 1983). Both are quarry sites and
have extensive areas of flintknapping debris of jasper-taconite, a material which
figures prominently in the later periods of Isle Royale’s prehistory.

The south shore of Lake Superior is less well known than the north, but finds
in recent years have shown that the antiquity of this area extends at least as far back
as 8000 BR The Flambeau (ca. 7000 BC.) and Minocqua (ca. 6000-5000 BC.)
phases were initially defined for sites and artifacts from northern Wisconsin (Salzer
1974) which, like the Plano Tradition artifacts to the north, are characterized by
distinctive flaking patterns and a persistent use of Hixton silicified sandstone as the

preferred raw material. Hixton has its source in south-central Wisconsin and found

80

its way at least as far east as Marquette County, Michigan where it occurs in some
abundance on finished tools (Buckmaster and Paquette 1988, Clark 1989b). The best
documented association of copper with this early period comes from the Itasca site
in northern Minnesota which has been dated between 7600 and 5500 BC. (Shay
1971).

The evidence for a human presence in the Superior basin during this early
period is gradually becoming better known, but the interest of the Plano hunters in
copper has yet to be demonstrated. Our knowledge of the Plano Tradition in this
area is very rudimentary in contrast to subsequent periods where we have not only
artifacts but the food remains and living sites to attest to the past life-ways of these
people.

Two cultural traditions germane to the prehistory of Isle Royale are
recognized for the Archaic stage: the Shield Archaic, and the "Old Copper Culture."
Generalizations which distinguish these from the earlier Plano Tradition include a
more diversified material culture, a broader subsistence base, and larger populations.
Fitzhugh (1972:1) has characterized the Archaic stage in eastern North America as
having "widely flung and locally variable expressions." Such is the case with the
Superior basin cultures among which can be seen similarities in overall patterns of
subsistence and organization, as well as contrasting stylistic expressions.

Some of the differences may be attributed to the local variability in
environmental factors which affords scope to the parameters of archaeological
cultures. Other differences may obtain from nonenvironmental factors which have

more to do with the evolution of culture and society apart from the environmental

81

constraints. Whatever the causes of dissimilarity, there is ample evidence for the
interaction among prehistoric groups seen in the long distance movement of raw
materials, including copper, across cultural boundaries. It is during the Late Archaic
substage that we perceive an intensification of local subsistence strategies which give
structure to the relationships between neighboring groups, thus facilitating the
enhancement of exchange of both goods and information.

The Shield Archaic occupied an area on the Canadian shield from Keewatin
District to Cape Breton, Nova Scotia (Wright 1972b). The distrrbution of their sites
suggests extensive use of watercraft and primary exploitation of moose, caribou, fish,
and beaver. Bone preservation on Shield sites is generally poor and the food
resources are inferred from locational data. In the material culture of the Shield
Archaic one finds no clear cut stylistic patterns, but rather an array of small
utilitarian tools made of locally available resources.

Disagreement exists regarding the pedigree of the Shield Archaic. It is
Wright’s belief that cultural evolution in the Canadian Shield is an in situ transition
from Plano to Shield Archaic to Laurel (1972b). According to Buchner (1979) the
definition of the Shield Archaic is much too large and inclusive, and it is likely that
it includes a number of smaller units which thusfar have fallen outside of our ability
to identify them. Buchner also feels that the origins of Laurel lie not with an
indigenous Archaic population, but came instead from south of the boreal forest. In
either case, the archaeological identification of Shield Archaic, and the discrimination

between it and aceramic Woodland assemblages has been difficult owing to the

82

longevity in stone tool forms that clearly begin in the Archaic and continue into the
Woodland stage.

For reasons of vagueness in its usage, the second Archaic archaeological
culture is equally unpopular with respect to its title, and uncertain with respect to
its meaning. The Old Copper Culture, Old Copper complex, or simply Old Copper
begins as early as 3000 BC. and continues until around 1200 BC. (Stoltman 1986).
The geographical distribution of Old Copper has its heart in northeastern Wisconsin,
proximate to the copper district of the Keweenaw and Ontonagon on Lake Superior’s
south shore. Artifacts attributed to Old Copper are found well away from this core,
however, and extend around the west end of Lake Superior, and into the Lake
Michigan basin, into the northern reaches of the midwest riverine, west to the plains
periphery, and east as far as the Ottawa River between Ontario and Quebec (Mason
1981).

The diagnostic artifact forms are fashioned of copper and represent the most
varied assortment Of copper items at any time in Upper Great Lakes prehistory.
Only during the Middle Woodland in the Hopewell culture, and later in Mississippian
culture does copperworking assume comparable proportions, although in much
different expressions. Archaic artifacts include large (by Terminal Woodland
standards) spearheads, knives, gaffs, adzes, as well as an array of forms more familiar
later in time, the awls, tubular and discoidal beads, and hooks (Wittry 1957).

The boundary between the Archaic and Woodland stages is archaeologically
defined by the introduction of ceramic technology. This analysis uses Initial and

Terminal Woodland for consistency with nomenclature used by some archaeologists

83

in Canada, northern Minnesota, and northern Michigan, as well as for substantive
cultural-historical reasons which reflect the southern tier-northern tier construct of
Mason (1981).

In eastern North America south of the boreal forest the first ceramic
producing cultures are referred to as Early Woodland, bearing dates as early as 1000
BC. Early Woodland pottery is characteristically coil constructed and thick with
cordmarked interiors and exteriors. By the time we recognize a Middle Woodland
cultural pattern, beginning about 300 BC, ceramics have a wider variety of shapes,
surface treatments, and types of decoration (Mason 1981). It is at this juncture that
ceramics make their appearance in the material culture of the Shield Archaic (Wright
1972b), or are brought into the region around and north of Lake Superior by a
different culture to the south (Buchner 1979). In either event, the result is the
taxonomic transformation of the resident Archaic population into the Woodland
stage, albeit with little or no effect on the subsistence practices of those dwelling in
the boreal forest. Echoing an earlier sentiment of Wright (1968:47) that, "The
evidence from the sites under consideration...strongly suggests that ceramics are not
an indigenous part of Ojibwa material culture," Mason summarizes the ambivalence
with which he believes pots were accepted in the north:

There is a high probability that there were some people in the

Laurel country who did not manufacture or use earthenware but who

were fully contemporaneous with those who did and that the only

empirical difference between their respective sites would be the
presence or absence of sherds. [1981:286]

84

In practice, there is synonymy in the terms "Initial Woodland" and "Laurel"
in the Lake Superior basin. Other Initial/Early Woodland archaeological cultures
(e.g., Nutimik and Saugeen) are peripheral to this analysis, although they represent
analogous cultural developments as marginal manifestations of more westerly or
southerly Woodland stage cultural evolution (Mason 1981). The Laurel focus was
defined by Wilford (1941) from his work in northern Minnesota. The definition was
subsequently expanded into southern Manitoba by MacNeish (1958). Laurel sites are
now known from a large area north of the Great Lakes and between Saskatchewan
and central Quebec.

Other than the apparent deletion of large sized copper items from the
inventory and a change in raw material preference, there were no fundamental
changes in the lithic and copper technology and their products from previous Archaic
material culture. Net sinkers appear to be new in the region at this time (Wright
1967), and trade items such as Saugeen pottery, shell from Manitoba, and
Yellowstone obsidian occur on both sides of Lake Superior (J anzen 1968, Wright
1967). There is a proliferation of small endscrapers in the Laurel culture, which is
attributed to the functional needs of an economy devoted to fishing (J anzen 1968).
Copper tools and ornaments in Laurel assemblages include the typical range of small
forms common throughout prehistory. There are no copper tools or ornaments
which are considered distinctively Laurel in style. It is assumed that copper was one
of the items used by the Laurel people in the context of gift exchange or trade in the

acquisition of nonlocal commodities. The degree of interregional exchange appears

85

not to be too great, however, since nonlocal items rarely occur in appreciable
numbers.

It is Laurel pottery that identifies this culture. Vessel construction is by
coiling and there is a modest use of temper. Vessels are hard and thin-walled,
usually conoidal with straight rims and squared lips. Decoration is by dentate
stamping, linear stamping, push-pull, incising, pseudo-scallop shell impression, and
use of punctates and bosses confined to the upper 1/2 to 1/3 of the vessel which is
otherwise smooth. Cordmarked exteriors and the use of cord-wrapped sticks (CWS)
for decoration do not occur in Laurel (Mason 1981).

A provisional seriational sequence for Laurel ceramics begins with the Pike
Bay phase in which most decoration is by linear stamp and push-pull. In the
McKinstry phase most decoration is with pseudo-scallop shell (PSS) impressions
which is the hallmark of Laurel pottery. Finally, in the Smith phase, there is a return
to the decorative styles that were popular in the Pike Bay phase (Mason 1981:289ff).
In addition to this chronological dimension there is a geographical distn’bution of
design elements that underscores the internal variation within the Laurel culture: PS8
has a higher frequency on northern sites while the linear stamped and push-pull
decorations are more common at sites along the southern border of the Laurel range.
Similarities among Laurel ceramics and contemporaneous North Bay (Mason 1967,
1969), Saugeen (Wright 1967), and Point Peninsula (Stoltman 1973, Wright 1967)
styles have been noted.

Also along the southern periphery are the largest Laurel sites and the only

Laurel sites with burial mounds, principally in the Rainy, Nipigon, and Pigeon rivers

86

along the international border. At the Smith Mound in northern Minnesota the
mode of interment was by bundle burial in small ossuaries. There was also evidence
for preburial ritual treatment in the form of dismemberment and defleshing (Mason
1981).

There is no sudden change in artifact style to mark the beginning of the
Terminal Woodland substage. Instead it is defined in terms of the trends which set
it apart from its antecedent (Gibbon and Caine 1980). With the Terminal Woodland
there is increased localized cultural differentiation measurable in increments of
stylistic variability and raw material use. Subsistence practices become highly
specialized in areas with unique resources, such as the wild rice district in northern
Wisconsin, Minnesota and adjacent portions of Canada, or the fisheries at Sault Ste.
Marie. There is evidence for an increase in population size in the Terminal
Woodland in the form of a higher density of sites and larger site size (Fitting 1975,
Mason 1981).

The archaeological recognition of the onset of the Terminal Woodland
substage lies between AD. 600 and AD. 700. From this point to recorded history
the Upper Great Lakes was the scene of a complex interplay of archaeological
cultures representing three linguistic groups which were, in turn, partitioned into what
are traditionally thought of by archaeologists as culturally discrete and autonomous
units. It has reached proportions of a litany to state that one of the characteristic
qualities of sites in this region is the consistently heterogeneous ceramic artifacts. At
virtually any site it is possrble to find a constellation of Terminal Woodland wares

including Blackduck, Lakes phase, Huron, and Straits of Mackinac varieties. Brose

87

(1978:577) refers to the general condition as "chaotic" in a substantive sense, citing
widespread population movements, wholesale abandonment of areas, and decimation
by disease. There is increasing evidence of endemic warfare in the Lower Great
Lakes (Trigger 1976) and midwest riverine (Milner, Anderson, and Smith 1988) areas,
primarily among the horticultural societies.

On Isle Royale the Terminal Woodland was a time which saw the most
intensive use of the island’s resources. Here the array of ceramic styles are as diverse
as anywhere on the mainland. As is made clear in the analysis which follows, a great
deal of uncertainty remains regarding the composition, both demographic and ethnic,
of the groups responsible for the archaeological sites of the Terminal Woodland.

Specific accounts of Native Americans or Europeans on Isle Royale are
uncommon in the early documentation of the French missionaries, explorers, and
traders. We know that the first French in Lake Superior, the traders Brule in 1622
and Nicolet in 1634, and the Jesuits Raymbault and Jogues in 1641 had little or no
immediate consequences for native culture. However, the trends which began in the
early and middle seventeenth century in the Lower Great Lakes became amplified
as the incorporation of European material culture and participation in the fur trade
spread to the Upper Country (Stone and Chaput 1978).

The acquisition of European material culture preceded the arrival of the first
whites, filtering through a down-the-line exchange network extending from the St.
Lawrence River across the Great Lakes to the Mississippi Valley. While there is no
consensus regarding the extent and magnitude of sociocultural change brought on by

contact and the fur trade, at a broader scale certain fundamental shifts in native

88

culture can be traced, including an increase not only in the incidence of intergroup
hostility, but in the motivation behind it. Low level endemic warfare existed to
address the reciprocal animosity among groups and as a means to acquire prestige
with lines Often drawn along linguistic boundaries. Blood feuds required exchanges
in which redress and compensation were the ultimate goal.

But by the mid 16008 the motivation for violent interactions, even if founded
on traditional blood feud idioms, were directed towards the acquisition of furs, fur
bearing territories, and/or an economically strategic position as middlemen between
the French and those groups lacking direct access. By the mid seventeenth century
Neutral and Five Nation Iroquois attacks on the Assisteranon (a generic term for
nonIroquois speakers) in Michigan’s Lower Peninsula caused the Sauk, Fox,
Kickapoo, Mascouten, Miami, and Potowatomi to move west of Lake Michigan,
leaving the land between lakes Huron and Michigan virtually uninhabited. It was also
during this period, between the 1660s and 16903 that the French increased their
presence in the Upper Great Lakes with the building of missions, forts, and trading
centers. Changing social configurations found the development of multiethnic
populations around the social and economic nucleus of these French establishments

(Mason 1981, Ouirnby 1966, Ray 1974).

ARCHAEOLOGICAL CULTURES OF THE TERMINAL WOODLAND SUBSTAGE

Anticipating the need to differentiateamong archaeological cultures in order

to address questions of interaction, the archaeological cultures of the Terminal

89

Woodland substage are defined both in terms of normative and, where evidence exists
for it, behavioral characteristics.

BLAOKDUCK: The Blackduck culture was defined initially by Wilford (1941,
1955) with a number of subsequent augmentations (Evans 1961, Hlady 1970,
MacNeish 1958, Wright 1965). The core area of Blackduck is in the Rainy River
area between northern Minnesota, Ontario, and Manitoba, but may be found across
the north shore of Lake Superior to the Straits of Mackinac and into the northern
reaches of the Lake Michigan drainage (Figure 2).

Blackduck dates from AD. 700-800 to around AD. 1100 (Lugenbeal 1979)
or AD. 1750 (Lynott, Richner, and Thompson 1986; Syms 1977), by which time it is,
by one account (Arthurs 1986), replaced by ceramics representing the Selkirk and
Sandy Lake archaeological cultures. The replacement theory is not universally
accepted, and there is evidence to suggest that Blackduck, Selkirk, and Sandy Lake
are contemporaneous (Syms 1977).

Although there is little consensus regarding the origins and ultimate ethnic
identification of Blackduck, or the dates which bracket it, Blackduck pottery is
distinctive, although it shares similarities with Mackinac, Heins Creek, Madison,
Kathio, and Clam River ceramic traditions (Afinson 1979). Typical Blackduck vessels
are globular in shape with triangular or wedge shaped rims decorated with complex
CWS impressions and exterior punctates. Bodies are cordmarked, often with
brushing or combing on the neck (Figure 3).

Subsistence practices emphasize local resources without horticulture, wild rice

being particularly important in the Blackduck core area. Gibbon and Caine (1980)

 

 

BLACKDUCK-
- SELKIRK

 
    
 
    

JUNTUNEN
SAN DY
LAKE

LAKES -
- ONEOTA

 

 

Figure 2.

Distribution of Terminal Woodland archaeological cultures in the
Upper Great Lakes.

91

«f; and}!

' ’4 Lift; . '41'¢’:[:3
_ a”. 4

— 'it
I ,, '. ‘
£' " . ~»- 4
'. -' . 1‘
- “2 . raw! ’2‘“ w
1- (I‘ll %‘ (”4.0
l p '1 Uu '
., . ' 1“
"I \
" . . LA
. ’4 it

 

 

Figure 3. Blackduck Pottery (A, DeIp-3; B, 201R72)

92
identify several related trends from Middle (Initial) to Late (Terminal) Woodland

which are germane to the Blackduck culture, including an increase in rice processing
features, in site numbers and size, and in small limited activity sites.

On Isle Royale, Blackduck ceramics are no more or less common than other
Terminal Woodland types. Pure Blackduck components are fewer than anticipated,
given the proximity to major areas of Blackduck activity. Copper artifacts attest to
some level of interaction with the copper sources of the north shore and Isle Royale,
probably by direct acquisition, although like the other Terminal Woodland copper
industries, needs for the metal were largely nonessential small tools and ornaments.

SELKIRK: Selkirk is known from the area to the north and west of Lake
Superior, dating between AD. 700 and 1750 (Arthurs 1978, Rajnovich 1983, Wright
1981). Initially defined by MacNeish (1958), Selkirk was subsequently expanded to
include a number of related ceramic style groups (Hlady 1970, 1971). The area]
distribution of Selkirk is embrace by an area bounded by northern Saskatchewan, the
Hudson Bay Lowlands, the northern shore of Lake Superior and northern Minnesota
(Rajnovich 1983:52). In a reanalysis of Selkirk ceramics Syms (1977) has brought
together a number of complexes which he believes are part of a larger "Selkirk
Composite" (1977:71). The individual complexes exhibit sufficient integrity of style
and regional distribution to suggest that they formed an internal partition within the
composite. The composite represents the maximal aggregation of related style groups
which constitute Selkirk as an archaeological culture.

According to Rajnovich (1983) subsistence and settlement of Selkirk is difficult

to ascertain due to the archaeological mixing of ceramics from more than one style

93

group. At Voyageurs National Park in northern Minnesota Lynott, Richner, and
Thompson (1986:279) found it impossrble to differentiate the subsistence and
settlement practices of Selkirk from those of Blackduck with which it was consistently
associated. It is likely that the perceived need for so-called pure components for the
elucidation of cultural practices has systematically sidestepped the potential for
examining the meaning of mixed assemblages.

The Selkirk occupation at the Spruce Point site in northwestern Ontario was
found to have structural remains of two oval lodges measuring 6-7 m long by 4 m
wide with outside activity areas (Rajnovich 1983:61). Proximity to water is used as
a criterion for determining seasonality of occupations; using Rogers’ (1967)
ethnographic work among the Mistassini Rajnovich believes that sites occurring
within 10 to 30 m from a shoreline reflect warm season encampments while sites 100
to 200 meters away from water are likely winter occupations (1983:63-64). At
Voyageurs National Park small ephemeral occupations were identified, indicating
short term camps aimed at the exploitation of beaver (Lynott, Richner, and
Thompson 1986:279).

Selkirk is identified archaeologically by its fabric impressed ceramics with little
or no decoration (Figure 4). According to Rajnovich (1983) the earliest Selkirk
pottery is found in northern Manitoba where conical bases on vessels strongly imply
a Laurel origin for Selkirk with a subsequent diffusion to the south and east. Closer
to Lake Superior in southeastern Manitoba and the Lake of the Woods area of
northwestern Ontario the earliest Selkirk material is "simple undecorated ceramics

(Alexander Fabric Impressed)...followed later by decorated types influenced by

94

 

 

B, Mason Collection, Thunder Bay

: A, 201R];

Selkirk Pottery

Figure 4.

95
Blackduck" (1983:58). It is thought that Blackduck is replaced by Selkirk and Sandy

Lake in the mid eighteenth century (Arthurs 1986). Wright (1968) sees Selkirk as
ancestral to the historic Cree in northwestern Ontario and adjacent Manitoba. On
Isle Royale Selkirk has been identified at a number of the recently recorded sites, but
prior collections have not been systematically examined for this type of pottery.

SANDY LAKEZWANIKAN: The Wanikan culture was defined by Birk
(1977), although Sandy Lake pottery, the diagnostic artifact of the Wanikan culture,
was identified earlier by Cooper and Johnson (1964) who suggested that the Wanikan
culture evolved out of the Siouan speaking Clam River focus.

Like Selkirk, Wanikan is identified primarily by its pottery which is often
difficult to distinguish from its late prehistoric cordmarked contemporaries. It is
described as having, "thin-walled, globular pots with straight, thin rims, exterior
surface treatment of vertical cording or smoothed exterior, and occasional interior or
exterior punctates [Figure 5A]. Decoration is confined to interior lip notching,
although some vessels show influence from other wares, such as Oneota, with trailing
or stamping" (Lake Superior Basin Workshop 1988). Temper is predominantly shell
in the southern part of the range of Sandy Lake pottery, changing to grit in the north
(L. Peterson 1986).

Chronologically, the Wanikan culture falls between AD. 1000 and AD. 1700.
The Norway Lake site in northern Minnesota contained a substantial Wanikan
component associated with Blackduck ceramics. A burial mound with a single flexed
primary inhumation contained Sandy Lake pottery exclusively, although Birk (1977)

indicates that simple burials in shallow graves are more typical. Settlement and

96

 

Figure 5. Sandy Lake and Mackinac Pottery (A, 201R28; B, 20MK1)

97

subsistence is not well known but is believed to reflect a predictable focus on wild
rice and other aquatic resources of the region. Sandy Lake pottery has a maximum
distribution from the Mississippi headwaters area east to Lake Nipigon, north and
west as far as eastern Manitoba, with its area of primary concentration in the Lake
of the Woods and Rainy Lake area on the international border (Arthurs 1978).

STRAITS OF MACKINAC: The Terminal Woodland sequence determined
from excavations at the Juntunen site (McPherron 1967) includes three cultural
phases: Mackinac phase (ea. A.D. 800-1000), Bois Blane phase (ea. A.D. 1000-1200),
and the Juntunen phase (ca. AD. 1200-1450). Subsequent reanalysis of the
collection raises some doubt as to the validity of the middle Bois Blane phase, and
has broadened the temporal parameters somewhat (Clair MeHale Milner, personal
communication 1989). For the purposes of the Isle Royale analysis the Bois Blane
phase is divided arbitrarily between the Mackinac and Juntunen phases, ca. A.D.
800-1100 and ca. A.D. 1100-1450, respectively.

The Mackinac phase levels of the J untunen site are interpreted as the remains
of large warm season fisheries (McPherron 1967). There is little evidence of the use
of native copper in this phase. Pottery styles favor cordmarked and fabric impressed
exteriors on short squat vessels with round bases and square lips (Figure 5B).
Decoration is by the use of punctations, and geometric designs on rims and necks
executed with a CWS. Mackinac ware is broadly similar to most contemporaneous
ceramics from surrounding areas, including Blackduck, Wayne, Princess Point, Heins

Creek, and Canton wares. Bois Blane ware as defined by McPherron (1967) has

98

many close stylistic ties with Blackduck and possibly represents the easternmost
manifestation of this culture.

Juntunen phase pottery, on the other hand, has close stylistic affinities with
the Ontario Iroquois Middleport and Uren stage of the Lake Huron basin. J untunen
ware includes castellated collars, and extensive use of linear punctations and the
push-pull technique for decoration (Figures 6 and 7). Nested chevrons and bands are
common motifs. Subsistence emphasizes fishing with augmentation with maize.
There is, according to McPherron (1967) a resurgence in the use of native copper for
small types of ornaments and implements during the Juntunen phase, while lithic
tools do not appear to have made any significant changes since the Archaic, with the
exception of the adoption of the bow and arrow by ea. AD. 900.

In McPherron’s (1967) initial analysis of the cultural/ chronological sequence
at the Juntunen site he described a shift in interaction sphere from west in the
Mackinac phase to the east in the later Juntunen phase. At that time it was
suggested that, on the basis of stylistic similarities between the Juntunen phase
ceramics and the Middleport and Uren substages of the Ontario Iroquois Tradition,
and the suggestion of a longhouse structure, that the Juntunen phase was best

considered the product of an Iroquoian group.

99

 

 

Figure 6. Juntunen Pottery (A and B, 20MK1)

100

 

Figure 7. Juntunen Pottery (A, DeIK-l; B, 201R1)

101

Wright (1968, 1972a) takes exception to this interpretation, stating that:

Although a number of ceramic attribute equivalents exist
between certain Late Woodland ceramics in Michigan and the ceramics

of the Ontario Iroquois Tradition, it is my opinion that the parallels

are of such a general nature that proposal of origin are placed in a

very hazardous position. In short, I cannot see a direct relationship

between the Ontario Iroquois Tradition ceramics of Southern Ontario

and the push-pull ceramics of Michigan and Northern Ontario.

[1968:49]

Wright was, at that time, looking from the perspective of the Michipicoten site
in the eastern Lake Superior basin where his excavation revealed a discontinuous
stratigraphic record, dating between AD. 1100 and AD. 1600. In nine strata Wright
found a record of mixed ceramic styles, representing archaeological cultures from the
south and southeast but not the west. Four groups of ceramics identified at
Michipicoten include Huron, Peninsular Woodland, stamped, and push-pull. The
latter two groups were to soon bear more formal appellations in the Juntunen site
ceramic analysis (McPherron 1967). The missing pottery from the west, Blackduck,
was found at the Pie River west of Michipicoten and the Montreal River to the south,
filling in the regional picture of the complex array of ceramic style. Of all the
published reports on Terminal Woodland material culture, the Michipicoten and Pie
River sites most nearly replicate the Isle Royale assemblage.

Wright asserts that since pottery was not an indigenous part of Ojibwa culture
north of Lake Superior, the heterogeneity in style observed in the prehistoric record

is explainable through trade/exchange. A cognate of this argument is his

interpretation of the lithic industry as highly conservative, showing little variation

102
through time or space. Wright (1968:47-48) believes that this is an indication of local

stability in prehistoric Ojibwa material culture. It could be argued that nowhere in
the region was ceramic technology indigenous. The stance assumed in this thesis is
that ceramic technology was at some point acquired by indigenous groups who used
it, modified it, or rejected it to suit their needs. Further, as it will be argued in
Chapter 7, the notion of ceramics being primarily introduced as trade items is
rejected; virtually all ceramics in the region are locally produced on locally available
clays.

Fitting (1975) favors the Chippewa (Ojibwa) as the probable identity of the
Juntunen phase, based on the fauna] remains which he interpreted as representing
the "Chippewa adaptive pattern" (1975:185). Stylistic similarities in ceramics,
according to Fitting, were the result of influence from interacting Iroquoian groups.
Compelling evidence for an Algonkian identity for Juntunen was found at four
J untunen phase sites in northern Lower Michigan where shale disks bearing

iconographic representations of Algonkian origin have been excavated (Cleland 1985).

HURON: The evolution of the prehistoric Iroquoian-speakers into their
historic counterparts is much better known than for their Algonkian or Siouan
speaking neighbors. This is due, in part, to the continuity in geographical setting of
the Iroquois groups, and to their early interactions with the French. In particular,
the documentation left by the Jesuits is an especially valuable source of information
on those aspects of their traditional society usually unavailable in the form of

archaeological data.

103

The historic Iroquois are divided into the Saint Lawrence, Five Nation, Eric,
Neutral, Susquehannoek, Petun, and Huron. It is primarily the latter two groups,
archaeologically collapsed into one, that are relevant to the prehistory and early
history of Isle Royale, although the effects brought about by the wars of the Iroquois
and Huron diaspora in the mid seventeenth century are significant regionwide. These
people were residentially stable with fortified villages, outside of which were extensive
fields where corn, beans, and tobacco were grown. Hunting, fishing, and gathering
were practiced but the primary subsistence was based on horticultural produce.
Mortuary preference was for ossuary burials, often at intervals of seven to ten years
at which time all the dead which had accumulated since the last interment were
gathered for a large "Feast of the Dead." In addition to the obvious function of
disposal of the dead, the accompanying rituals reinforced mutual ties of kinship or
acknowledged friendship among both Huron and Algonkian groups. Goods were
redistributed and trading partnerships for the following years negotiated, often
assuming the idioms of kinship terminology (Hickerson 1960).

Prior to 1660 the Northern Division of the Huron Branch of the Ontario
Iroquois Tradition (Wright 1966) was located in an area referred to as Huronia,
between northern Lake Huron and Lake Simcoe. Village locations shifted in
response to a variety of factors, but overall there was little movement outside of this
area except by small task-specific groups undertaking long-distance trading
expeditions. After 1660 when relations with the Five Nation Iroquois forced the
Hurons into their historical diaspora across the Great Lakes and into the Mississippi

Valley, the remnant segments of Huron society took on a modified identity as

104

Wyandotts and, with their Ottawa counterparts, continued to practice their trading
across the upper lakes.

Huron pottery is characteristically well fired, thin-walled, burnished or
smoothed, with distinctive decoration on the rims and necks of vessels (Figure 8).
Lines made by incising or by tool impression are most common, although punctates
and push-pull continue to be employed. Vessel shapes are round and globular, often
with squared collars and/or castellated rims. While Huron pottery is "distinctive" it
has been suggested (Fitting 1975, Mason 1976, Ramsden 1988) that what is Huron
in style in the Lake Superior region is, in fact, Algonkian mimicry. Huron pottery
is common on Isle Royale, but not in numbers indicating any more or less activity
there than for other non Iroquoian groups.

The traditional view of the Huron as a relatively unitary cultural entity has
been questioned. Ramsden (1977, 1988) suggests as an alternative that it is the
French documentation that has erroneously resulted in this view, and that in late
prehistory what later became a Huron confederacy as a response to the stressful
conditions of trade and warfare, is better characterized as a number of sociopolitically
and economically independent villages or village clusters which shared a similar
cultural pattern. He reviews the ontogeny of the archaeological Huron, pointing out
that its evolution has strongly colored its use and, concomitantly, our understanding
of group interaction in this area. The origin of the historic Huron and their relations
to their Iroquoian neighbors is irrelevant to the issue at hand, but of moment is
Ramsden’s (1988) investigation of the nature of Huron-Algonkian boundaries to the

north of Huronia.

105

 

 

Figure 8. Huron Pottery (A, 201R28; B, 201R52)

106
LAKES PHASE: It is thought that the Lakes phase of the Late Woodland

evolved directly from its Middle Woodland Nakomis phase ancestor, appearing
archaeologically as a population explosion in an area of northern Wisconsin only
sparsely inhabited previously (Salzer 1974, 1986). The Lakes phase is believed to
date between AD. 600 or 700 and AD. 1400. Subsistence reflects local variation and
constraints: a seasonally mobile strategy of small mammal hunting, ricing, and fowling
was followed in this area of extensive marsh and lakes. Mortuary practices included
burial of bundled secondary, flexed primary, and cremations in simple effigy, linear,
or conical mounds.

Lakes phase pottery is cordmarked with decoration by cord wrapped stick,
cord, and punctation (Figure 9). The problem of the lack of stylistic boundaries is
evident with the Lakes phase which in its early stage is similar to Heins Creek and
Madison of the Lake Michigan and Door Peninsula area while in its late stage it is
similar to Point Sauble, Aztalan, and Mackinac ceramics (Salzer 1986). Mason
(1981:312) summarizes the Lakes phase: "This country was then, as today, far more
a cul-de-sac than a crossroads, and it seems likewise to have encouraged a
conservative tendency which helped secure life even if at a pace behind those of some
other areas."

Copper tools and ornaments are common among the Lakes phase sites.
Immediate access to drift sources in northern Wisconsin and/or primary copper
sources in the Upper Peninsula undoubtedly supplied the Lakes phase with all their

needs.

107

 

 

Figure 9. Sand Point (Lakes Phase) Pottery (A and B, 20BGl4)

108

The Sand Point site at the southern end of Keweenaw Bay in Lake Superior
is the only major site in this part of Michigan which has been extensively excavated.
The site is a combination of occupation and burial mounds, dating between A.D.
1100-1400. Although Sand Point ware was defined as a discrete taxonomic entity
(Dorothy 1980), the ceramics exhibit an expected amount of stylistic merging with the
Lakes phase with which it is combined for the purposes of this report. Associated
ceramics include Blackduck, Oneota, and Juntunen to form a typical mix of Terminal
Woodland wares.

Lakes phase ceramics are not sufficiently distinctive to allow consistent
identification in analysis, and disagreement or uncertainty regarding Lake phase and
other Wisconsin ceramics (e.g., Madison, Heins Creek) is likely to be a problem.

ONEOTA: Oneota is a late prehistoric cultural development on the margins
of the Lake Superior basin which only nominally affected human interaction on Isle
Royale. The geographical core of Oneota cultural development is in the Upper
Mississippi headwaters and east to Lake Michigan. It is believed that Oneota is
ancestral to the historic eastern Siouan speaking groups (Iowa, Oto, Omaha, and
probably Winnebago).

The Oneota cultural-historical chronology employed here is divided into
Emergent (A.D. 900-1000), Developmental (A.D. 1000-1300), Classic (A.D.
1300-1634), and Historic Oneota (post AD. 1634) (Mason 1981:362). Origins of
Oneota include strong ties to Middle Mississippian cultures to the south, reflected in
unmistakable similarities in stylistic qualities of ceramics. The northward diffusion

of maize horticulture into the upper Mississippi valley was adopted by these northern

109

Mississippians. Depending on the configuration of the local resource base, Oneota
adaptive strategies included a well balanced blend of horticulture, hunting, and
collecting which emphasized the resources of the abundant lakes in the region, fish,
mussels, and wild rice.

Villages were semipermanent with circular or rectangular bark covered
wigwams. A defensive stockade, reflecting the increasing atmosphere of endemic
warfare, encircled many settlements. Garden beds for maize and other crops were
located outside the village. Burial of the dead was extended and primary in
cemeteries or infrequently flexed and primary in mounds. Oneota pottery is
distinctive in its shell temper, smoothed surfaces, use of trailing in simple but bold
decorative motifs, and typical Mississippian vessel shapes of strongly outflaring rims
and acute shoulders (Figure 10).

Within the Oneota archaeological culture, several localized variants are
recognized. The one with immediate similarities to Oneota on the south shore of
Lake Superior is the Mero Focus or Complex (Mason 1966) in the Green Bay and
Door Peninsula areas of Wisconsin. In a review of sites along the Menominee River
between Michigan and Wisconsin, Buckmaster (1979) identified a number of sites
with Oneota ceramics, and at the Sand Point site on Keweenaw Bay Oneota sherds
were found in association with an otherwise Algonkian Woodland assemblage
(Dorothy 1980). These northern variants of Oneota did not confine themselves to
the use of shell temper often using grit instead. However, Oneota potters tended to

adhere to more normative uses of decoration and vessel shape.

110

 

 

Figure 10. Oneota (Mero Phase) Pottery (A and B, 20KE15)

111

Copper is found on Oneota sites. It is likely that their requirements for raw
copper were met by the Keweenaw and Ontonagon sources which were within their
immediate reach. Although Bastian (1963a) identifies one sherd from Chippewa
Harbor #1 as Mississippian, it is an inconclusive example of Oneota on Isle Royale.
An explanation of the absence of Oneota in the late prehistory of Isle Royale may
be found in the combination of cultural boundaries, differential access to copper
sources and alternative sources, and Lake Superior as a natural barrier. It seems
clear that the Oneota did not need Isle Royale’s copper, nor does it seem that the
people of Isle Royale needed any Oneota pottery.

Specific hypotheses derived from the above consideration of the archaeological
manifestations of interaction and what is known of the late prehistoric archaeological
cultures of the Lake Superior region are presented as the basis for the neutron
activation analysis of ceramics in Chapter 7. In general, we can conclude from the
overview of Terminal Woodland archaeological cultures that, in spite of differences
in material culture, the essential behavioral characteristics relating to settlement and
subsistence practices are so similar as to preclude differentiation at this level, Huron
and Oneota horticulture notwithstanding. An examination of the distribution of the
different material representations of these cultures, especially ceramics, is used to
assess the possibility of differential access to copper and subsistence resources on Isle

Royale.

CHAPTER 5

ISLE ROYALE ARCHAEOLOGY

INTRODUCTION

The core of this analysis is provided by archaeological data from Isle Royale
National Park. Having provided a regionwide cultural-historical background, the
environmental background and resource base of the island are now examined. In
order to appreciate the evolution of the archaeological data base on the island the
previous and recent research efforts are reviewed. Data representing 51 Terminal
Woodland sites are used to determine the nature and extent of the use of the island’s
resources, and an assessment of the relative intensity of this use by the different

archaeological cultures from the Lake Superior region is made.

ENVIRONMENTAL SETTING

Isle Royale is located in northwestern Lake Superior, 24 km southeast of the
province of Ontario and 72 km northwest of Michigan’s Keweenaw Peninsula (Figure
11). The main island is 72 km long and 14 km at its widest point. The total area of
the island is 544 square km (210 square miles). Approximately 200 smaller islands
occupy the periphery and the interior of the main island has 83 lakes. The island’s
topography is characterized by ridge and swale contours oriented southwest-northeast
along the lines of uplifted geological features. The highest elevation is 425 m (1394
t) arnsl or 241 m above the modern level of Lake Superior at 183 m (602 ft).

Keweenawan Volcanics dominate the geology of Isle Royale, with interbedded

sediments exposed in the upwarping of the deposits which tilt towards the southeast

112

113

 

 

 

 

 

N50 T \ T 5
‘K
0“
”--.\_ 0
"'~- ISLE
%YALE MICHIPICOTEN
ISLAND
4?;
90
xi" . v
\\\ 4‘
‘2. g LAKE SUPERIOR
. .45" “(as
' 3:: {f0 91 “h
A
o . o
E \_ MICHIGAN
; WISCONSIN \.
j ‘3
Figure 11. Location Of Isle Royale and the Upper Great Lakes.

 

 

114

and mirror the formations in the Keweenaw Peninsula which tilt towards the
northwest. The southwestern end Of the island is comprised of sedementary deposits
which also have corresponding features on the south shore Of Lake Superior as
sandstones and conglomerates. The availability of native copper, around which much
of the economic history of Isle Royale revolves, is typically greater in the sedimentary
deposits where it occurs as lodes. F issure deposits are found in transverse faults
which cut across the beds. Although the former are typically more productive, as on
the south shore of Lake Superior, the fissures on Isle Royale provided the bulk of
the mineral for both prehistoric and historic miners (Huber 1975:56-57). As a
consequence, most of the evidence for copper exploitation is found in the
northeastern part of the island. It is significant to note that while there are few
sources of comparable quality on the north shore of Lake Superior, copper is
distributed across the north shore from Duluth to north of Sault Ste. Marie. Some
prehistoric mines are found near Point Mamainse at the eastern end of the basin
(Griffin and Quimby 1961:77-82). Packard (1893:179), citing early geological surveys,
states that some copper is available in the trap rock of Michipicoten and St. Ignace
islands in northern Lake Superior (see Chapter 6).

The Pleistocene history of Isle Royale begins with the recession of the Valders
ice mass ca. 11,000 HP. The recession first exposed the southwestern end of the
island and, punctuated by a series of minor fluctuations, continued to contract
towards the northeast leaving in its wake many of the typical periglacial features such
as drumlins and moraines. The postglacial stages of Lake Superior beginning with

Lake Nipissing at ca. 5000 BR, are relevent to this study since it is here that the

115

cultural history of the island begins. The elevation of Lake Nipissing between 184 m
(605 ft) and 198 m (650 ft) amsl in combination with the effects of isostatic rebound
created shoreline features now well above the present level of Lake Superior
(Farrand 1969, Stanley 1932).

The combination of volcanic and sedementary morphogenesis and subsequent
glacial and lacustrine effects explains the very thin and poorly developed soils
throughout the island. Soils are somewhat better developed on the southwestern end
where the decomposition of sediments has facilitated the process. To a large extent
these edaphic factors influence the forest communities on the island. Two major
upland forest types have their interface on Isle Royale: the southern boreal forest,
dominated by spruce, fir, and birch, and a temperate deciduous community of sugar
maple, birch, and red oak. Lowlands are occupied by communities of cedar and
spruce. Modern vegetational communities are greatly influenced by fire and moose
browsing (Slavick and J anke 1987).

Plants relevent to aboriginal subsistence include blueberries, strawberries, and
thirnbleberries which are widely abundant, and cranberries which are found locally
in their bog habitat. Aquatic lily tubers are available in many of the inland waters.
Roses, high-bush cranberry, currants and a variety of herbaceous plants with edible
greens are also locally abundant. Sugar maple was exploited for its sap in the early
historic period. Other plants of economic value include birch for containers and
canoes, and cedar and nettle for fibers.

Wild rice is not known to occur on Isle Royale, either from cultural sites or

in natural stands. Its presence on a prehistoric site would not be surprising given the

116

proximity to the rice district of northern Minnesota and its documented use by
Blackduck and other Terminal Woodland cultures. Similarly, maize has not been
recorded at any location on Isle Royale. While it is certain that maize horticulture
was impossible on the island, its introduction via a cultural agency such as groups
from the eastern Superior basin is considered highly likely.

Faunal resources are members of the southern boreal forest biome and
included, prior to the extirpation of several species, woodland caribou, beaver,
muskrat, otter, mink, matren, hare, coyote, fox, wolf, and lynx (Meech 1966).

Aquatic resources may have constituted the most reliable source of subsistence
in prehistory. Major economic species include lake trout, Whitefish, suckers, sturgeon,
pike, walleye, yellow perch, and brook trout. Fisheries located on the many inlets and
bays of Lake Superior would have been the most productive, although seasonal
abundance of spawning suckers and sturgeon at strategic locations would have also
been highly desirable. Warmer water species (pike, yellow perch) are found in the
interior lakes and the few streams on the island contain brook trout and seasonal
spawners from Lake Superior (Koelz 1929, Sharp and Nord 1960).

Avian resources are limited to birds with an aquatic orientation such as loons,
grebes, mergansers, and a relatively small variety of ducks, although a wide variety
are available during migration. Geese are abundant only for a brief period during
migration and many choose to over-fly the island. The only specie of gallinaceous
bird recorded on Isle Royale was the sharp-tailed grouse which apparently flew over
from the north shore to colonize a burned over area in historic times (C. Martin

1988). The population has since become locally extirpated. The economic value of

117

this and other terrestrial species was likely insignificant in prehistory. Historical
sources document the presence of passenger pigeons which were harvested on the
island at least as recently as the 18703 (C. Martin, personal communication 1988).

The detailed study of the modern populations of moose and wolf on Isle
Royale indicate a delicate balance of predator-prey relationships. Moose, if not
subjected to the culling and strengthening effects of predation, will quickly
overpopulate the island and strip its resources with a resulting crash in population
(Meech 1966, Peterson 1977). The peaks and valleys of wolf/moose demographics
exhibited in recent times do not take into account the effects Of human predation
which tends not to exercise the same selective discrimination of the wolves. The
implications of human predation for the reliability of major terrestrial faunal
resources (i.e., carrbou) in an insular context where recruitment was possrbly random
and anything but regular, produced a highly unpredictable and capricous subsistence
base. Fish and smaller mammals (especially beaver) undoubtedly constituted an
important buffer or occasionally were the primary subsistence base for the prehistoric
occupants.

Nowhere in the survey area, or on the island as a whole, is one very far from
either copper sources or a wide range of subsistence resources. The shoreline areas
of Rock Harbor and Tobin Harbor are immediately proximate to copper and fishing
resources, as is Chippewa Harbor. Riparian and terrestrial resources are available
a short distance inland from virtually any shoreline site. In addition, the logistical
features Offered by the transverse faults and embayments, such as at Chippewa

Harbor can be considered significant apart from purely economic or subsistence

118

considerations. Interior lakes, while relatively poor in fishing, Offer a combination
of sheltered and warmer environments away from the Superior shoreline with the
benefit of molluscs, beaver, and a few aquatic plant foods. The series of lakes covered
in the 1987 survey offer, in addition, the most expedient means of crossing Isle
Royale at any point along its entire length, and thus represents the potential for

logistically placed sites along portage routes.

PREVIOUS RESEARCH

The advent of American mining on Isle Royale and the Keweenaw and
Ontonagon districts to the south began a period in the awareness of and speculation
about the identity and antiquity of the prehistoric copper miners. From 1846 to
about 1910 attention was focussed on the identity of the miners and their relation to
the mysterious Mound Builders to the south in whose elaborate burial mounds
copper artifacts were often found. Much of this speculation was highly conjectural
and imaginative. The involvement of the Smithsonian Institution brought the first
serious consideration of the historical accounts pertaining to copper in the Great
Lakes region and was a proponent in the identity of the Mound Builders and copper
miners as the ancestors of surviving Native Americans (e.g., Packard 1893, Winchell
1881). The emphasis was on mining, however, and although excavations Of mining pits
were undertaken with descriptions of associated artifacts, there was virtually no
interest in occupation sites in the copper districts.

The emphasis on copper mines continued to dominate the analysis of

prehistoric remains on Isle Royale well into the twentieth century but, beginning with

119
the McDonald-Massee Expedition to Isle Royale in 1928, was tempered with the

recognition of occupation and burial sites on the island. The expedition reported by
West (1929) describes the Massee Rockshelter (201R14), Chippewa Harbor (201R1),
Birch Island (201R27), and McCargoe Cove/Minong Ridge (201R73/201R24)
prehistoric sites and alludes to others on Chickenbone, Liverrnore, and La Sage lakes
in the interior. The artifacts were described by McKem of the Milwaukee Public
Museum and compared to Effigy Mound Tradition materials from Wisconsin, thus
creating the first tangible relationship between the island and prehistoric Indian
cultures on the mainland.

In the 1920s and 30s Fred Dustin applied himself to the prehistory of Isle
Royale, bringing with him a well-tuned eye for occupation sites developed in the
Saginaw Valley of lower Michigan. In addition to traversing much of the island,
Dustin tested the Chippewa Harbor #1 site and several mine locations, publishing
and presenting his findings in the years between 1929 and 1957 (Dustin 1929, 1930,
1957). Among his contributions is a more realistic view of the cultural dynamics and
systematics of the prehistoric miners than was held by his predecessors. Four of
Dustin’s general observations represented in Griffin (1961:30) continue to act as
viable research hypotheses:

1. The probable route by the miners was from the Keweenaw Peninsula

to the Island and back, rather than from the north shore.

2. The ancient miners were probably the ancestors of the Algonquian and

perhaps Iroquois of the Upper Great Lakes.

3. Isle Royale did not offer much inducement for permanent settlement.

120

4. The major mining area was McCargoe Cove.

Carl Guthe of the University of Michigan made limited test excavations at
Birch Island and Chippewa Harbor in 1930 (Guthe 1930). In the same summer tests
at the Indian Point site were undertaken with the assistance of Fred Dustin, adding
to the growing interest in the island’s prehistory.

George Quimby was, perhaps, the first "modern" archaeologist to work on Isle
Royale. In his 1939 publication on aboriginal camp sites on the island he presents
data on several occupation sites with the identification of Iroquoian and Woodland
ceramic types, and fauna] and botanical remains. The interpretation was directed
towards relations between cultures in the Soviet Far East and similar developments
in the New World reflecting hypotheses of interest at that time.

Roy Drier of the Michigan College of Mining and Technology conducted test
excavations at the Minong Mine, Indian Point, and Chippewa Harbor in 1953-54
(Drier 1961:1-7). Drier’s contribution through this effort includes two radiocarbon
dates from Pit 25 at the Minong Mine which very loosely date the feature between
the first and third millenium B.C. A privately printed collection of earlier works on
prehistoric copper mining was published in 1961, making available under one cover
a number of obscure articles (Drier and Du Temple 1961).

Subsequent to Quimby’s and Dustin’s involvement, intermittant survey and
excavation by the University of Michigan Museum of Anthropology continued, firmly
placing Isle Royale in the broader context of Upper Great Lakes archaeology. Survey
and testing by Spaulding in 1959 was followed by three years of work under the

leadership of Tyler Bastian. Bastian’s unpublished report of the survey (1963a) and

121

masters thesis (1963b) give the first comprehensive descriptions of prehistoric and
historic sites on the island and effectively provide the baseline for all subsequent
archaeological reconnaissance. An additional hallmark of Isle Royale archaeology is
Griffin’s (1961) edited volume on prehistoric copper and the Indians of the Lake
Superior basin, much of which is a compilation of older sources and informal
discussions on the subject.

Since the University of Michigan survey in the early 1960S, archaeological
research on Isle Royale has been limited to a small number of impact assessments
of development associated with the National Park’s visitor and staff facilities, largely
undertaken by park personel or by archaeologists from Michigan Technological
University in Houghton. In 1984 Maass produced the park’s Cultural Sites Inventory,
a major compilation of cultural resources on the island, largely taken from extant
secondary sources.

Between 1981 and 1986 the National Park Service’s Submerged Cultural
Resources Unit conducted a submerged cultural site inventory on Isle Royale
(Lenihan 1987). Shipwrecks and a number of selected historic sites with terrestrial
associations were examined and documented. It was this survey which unintentionally
discovered a nearly whole Terminal Woodland vessel off the Siskiwit Mine.

In the spring of 1986 Patrick Martin conducted an archaeological field school
on Isle Royale in which the location and mapping of historic features at Todd
Harbor, McCargoe Cove, and Daisy Farm was the basis of instruction in

archaeological field techniques (P. Martin 1988). This effort resulted in the discovery

122

of one of the first copper smelter in the Lake Superior basin, and generated a series
of detailed maps for management purposes.

In the fall of 1986 the area around the Rock Harbor channel was examined
in an attempt to establish a terrestrial context for the J untunen phase pot found off
the Siskiwit Mine by divers. The survey area extended from Daisy Farm on the
southwest to Threemile on the northeast on the main island. On the Lake Superior
side of the channel coverage was from the Bangsund Fishery to Mott Island,
including the Rock Harbor Lighthouse, West Canbou, East Caribou, Cemetery, and
Rabbit Islands. Among the results of the Rock Harbor survey included the discovery
of an occupation site at the Siskiwit Mine, confirmed the presence of subsurface
prehistoric deposits at the Rock Harbor Lighthouse and Cemetery Island, and
identified new a site at West Caribou Island (Clark 1987).

The most recent study of Isle Royale National Park by the Midwest
Archaeological Center of the National Park Service provides survey coverage of
existing development areas within the park, along with survey coverage of selected
sampling areas and limited testing of selected archaeological Sites. The initial phase
Of the project, the field survey, assessed the effects of recent fluctuations in the level
of Lake Superior on cultural resources and the impact of visitors and park facilities
such as trails and campgrounds on cultural sites. Survey work is focussed on
shorelines and campgrounds in order to locate new sites in critical areas as well as

monitor the ongoing condition of previously known sites.

123
OVERVIEW OF THE TERMINAL WOODLAND SUBSTAGE ON ISLE ROYALE

These data are based on the results of the 1986 Rock Harbor survey and the
1987-90 surveys, as well as on published and unpublished accounts of prehistoric sites
on Isle Royale. For those sites not visited or tested by the Midwest Archaeological
Center (MWAC) survey Bastian’s (1963a, 1963b) work is the primary source. (Table
1; Figures 12-14).

The Terminal Woodland substage has more complete documentation than the
earlier prehistory of Isle Royale. If relative numbers of sites are an accurate
reflection of activity, the Initial Woodland use of Isle Royale was only slightly greater
than in the preceding Archaic stage, while in the subsequent Terminal Woodland
substage there appears to have been a major increase in the use of the island’s
resources evinced by a dramatically increased number of Sites. Within the Terminal
Woodland ceramic evidence indicates that the increase in population on Isle Royale
peaked ca. A.D. 1250-1450, and continued at equivalent or somewhat diminished
levels as late as ea. A.D. 1550-1650.

A possrble source of error in the determination of temporal placement
involves those sites which lack ceramics and are believed to belong to the Woodland
stage on the basis of their lithic raw materials. At this level, the determination is
highly reliable. However, at the present we lack the ability to discriminate between
Initial and Terminal Woodland raw material preferences. For this reason no
aceramic sites are included in Table 1 or considered in any detail in the following

discussion.

124

 

Table 1. Summary of Terminal Woodland Sites at Isle Royale
National Park.

ITE# NAME COMPONENTS" SITE TYPE
201R] Chippewa Hbr. #1 TW/H occupation
201R14 Massee Rockshelter TW burial
201R17 Grace Island TW/H occupation
201R18 Grace Point TW/H occupation
2OIR27 Birch Is. TW occupation
2OIR28 Indian Point IW/TW/H occupation
201R29 Belle Isle #1 IW/TW/I-I occupation
2OIR30 Lookout A/TW mine
201R31 Grass Point TW/H occupation
2OIR41 Siskiwit Mine TW/H occ/mine
201R42 Cemetery Is. TW occupation
201R45 Daisy Farm IW/I'W/H occupation
201R46 Ransom TW(?) mine
201R52 Baker Point TW occupation
201R53 Chippewa Hbr. #2 TW occupation
201R56 Rock Harbor TW occupation
201R65 Washington Is. #1 IW/TW occupation
201R78 Merritt Lane IW/I'W occupation
201R80 Singer TW mine
201R81 Phelps TW mine/occ.
201R82 Boys Island TW occupation
201R107 West Canbou Is. TW occupation
201R111 Rock Hbr. Light. TW occupation
2OIR114 Lone Tree Cove TW occupation
201R116 Threemile #1 TW occupation
201R118 2 1/2 Mile TW mine
2OIR120 Chickenbone Lk.#1 TW occupation
201R124 Lake Ritchie #1 TW occupation
201R127 Chippewa Hbr. #3 TW pot break
201R128 Lane Cove TW/H occupation
20LR134 Mott Sauna Beach TW occupation
201R140 Duncan Bay #1 TW pot break
2OIR142 Belle Isle #2 TW occupation
201R143 Belle Isle #3 TW occupation
201R144 Pickerel Cove #1 TW/H occupation
201R147 Brady Cove #2 TW occupation
201R148 Brady Cove #3 TW pot break
201R149 McCargoe Cove #2 TW occupation
201R150 McCargoe Cove #3 TW occupation
201R151 McCargoe Cove #4 Prehistoric occupation
201R152 Cove #2 TW occupation

125
Table 1 (Cont’d)

201R153 Cove #3 TW occupation
201R154 Cove #1 TW occupation
201le Washington Is. #2 TW occupation
201R174 Malone Bay Camp. IW/I‘W occupation
201R175 Malone Island #2 TW occupation
201R179 Greenstone Beach TW occupation
201R180 Little Greenstone TW occupation
201R181 Wright’s Island TW/H occupation
201R183 Malone Bay #3 TW occ./mine

*Components: A=Archaic, IW=Initial Woodland, TW=Terminal Woodland,
H=Historic

With only a few exceptions, notably the Indian Point and Chippewa Harbor
#1 sites, the data utilized in this discussion are derived from limited sources. The
fact that we are looking through very small windows, surface collections or
excavations consisting of less than seven, and in many cases one or two units, places
obvious constraints on how far the data are to be trusted. Sample sizes are small
and, perhaps not representative in a quantitative sense for individual sites.

The absolute chronology for the Terminal Woodland substage is weakly
developed on Isle Royale: there are twelve dates germane to this period, representing
six sites (Table 2). This is a major handicap as concerns the internal partitioning of
the Terminal Woodland. It is expected that over 700-800 years significant differences
in group locations, movement, and interaction potential occured which may address
what is ostensibly an association of different ceramics. For example, the association
of Huron and Blackduck ceramics on the Indian Point site could be attributable to
chronological factors and be independent of cultural ones, the events responsrble for

the ultimate deposition of the ceramics taking place 200 or 300 years apart. This

126

problem cannot be addressed in this analysis: future work should utilize the potential
for radiocarbon dates taken from residues deposited on the vessels themselves in
order to refine both the absolute chronology and the stylistic seriation of the various

ceramics found in the Lake Superior region (cf. Hamilton 1988, Lovis 1990).

SUBSISTENCE AND SETTLEMENT

The amount of domestic refuse recovered at many occupation sites indicates
small groups with ancillary subsistence activities conducted from an occupation site
nearby. Typically, one to four pots were broken/discarded in the immediate vicininty
of hearths, around which chipped stone tools were manufactured and repaired. The
picture is one of a family or a few families engaged in brief episodes of mining
supported by fishing, collection of berries, the hunting or trapping of beaver, with the
occasional taking of caribou.

The larger sites at Chippewa Harbor (2OIR1) and Indian Point (201R28) do not
necessarily contradict the above characterization of small groups as the primary
feature of Terminal Woodland demography on Isle Royale. It is not certain if the
impression of these sites as large obtains from repeated occupation of these strategic
locations by the same small groups whose more discrete occupations are in evidence
elsewhere on the island. An argument in favor of these sites as representing larger
aggregations of groups drawn from other areas of the island can be made on the
basis of their strategic positioning at either end of the major corridor across the
island. In such a location it is expected that groups would anticipate contact with

other groups for purposes of information exchange in the context of visitation. The

.Em cam swam 2a 8 3% @383,» 35.53. .S 05mm

 

127

 

mOEmnam 3.2..

l
0/. Am.
0 ll 0
“ware. j .
as r a... x. a
m0 e. I
I I... ..
b.0211. . a W fl» nv
‘
I79- &

 

 

128

.Em a»? seam 2a 8 8% “.5803 35.58. .2 05mm

 

 

mOEmanm

3.

m¥<._

 

 

129

 

 

/'

Figure 14.

9
/ 14 277 14.,

5 1 km
433
\0“
90¢“
a!" O
0
47
154 \‘J
150
28 148

 

/

Terminal Woodland Sites on Isle Royale, McCargoe Cove area.

130

Table 2. Calrbrated Terminal Woodland substage radiocarbon dates. (Calrbrations
are calculated according to Stuiver and Becker (1986) and determined using the
CALIB Radiocarbon Calibration Program, Version 2, Quaternary Isotope Lab,

University of Washington.)

201R], CHIPPEWA HARBOR #1

Sample Age
AD. 890 (M-1272)
840 z 100 B.P.

Crane and Griffin 1965:127-128

AD. 1130 (BETA-23115)
820 t 80 B.P.

AD. 1360 (BETA-23114)
smzman

AD. 1410 (BETA-23113)
5wzwan

201R18, GRACE POINT
AD. 1250 (BETA-35054)
700 2 70 B.P.

20IR30, LOOKOUT MINE
AD. 1510 (M-1640)

440 t 100 B.P.

(Crane and Griffin 1966:26)

AD. 1540 (M-1276b)
410 t 100 B.P.
(Crane and Griffin 1965:128)

AD. 1625 (M-1276a)
325 1 100 B.P.
(Crane and Griffin 1965:128)

201R41, SISKLWIT MINE
AD. 1760 (UCR-2243a)
190 z 60 B.P.

(Clark 1987:11)

AD. 830 (BETA-23116)
1120 x 80 B.P.

201R45, DAISY FARM
AD. 1670 (BETA-35055)
280 z 50 B.P.

Calibrated Ages (Intercept : two sigma)
AD. 694 (985) 1206

AD. 1020 (1219) 1280

AD. 1264 (1328, 1350, 1391) 1440

AD. 1280 (1410) 1440

AD. 1210 (1280) 1406

AD. 1298 (1439) 1650

AD. 1320 (1446) 1660

AD. 1410(1523, 1581,1625) 1955

AD. 1527(1668, 1751, 1758, 1777, 1796,1947,
1947, 1953) 1950

AD. 687 (898, 920, 942) 1147

AD. 1480 (1642) 1955

131

two interpretations are not mutually exclusive and, in fact, are supported by similar
lines of evidence. The ceramic heterogeneity found at Chippewa Harbor and Indian
Point would be expected in the case of multiple reoccupation by small groups or by
the coalescence of these into larger, multiethnic groups.

The subsistence and settlement patterns of the region are characterized by
adaptations which emphasize local resources. In some cases these resources are thinly
dispersed across a southern boreal forest landscape while in others the local resources
are concentrated and abundant seasonally (Heffley 1981). Major native fisheries at
Sault Ste. Marie and the St. Louis river, or the wild rice district of northern
Wisconsin, Minnesota, and adjacent portions of Ontario and Manitoba are resource
concentrations which permitted large aggregations of people during the harvest
season. Along the north shore of Lake Superior river mouths provided locations
which served both subsistence and logistical needs: spawning runs could be met in
season and small hunting units could coalesce for the exchange of information which
would guide impending decisions regarding the movement and distribution of people
for the summer season (Moore 1981).

The subsistence data from archaeological sites on Isle Royale suggest that the
island was occupied primarily from spring to fall. The availability of suitable prey
species over the winter months was probably as good as could be found on the
mainland, but the irrevocable nature of a decision to remain on the island may have
motivated many persons to return to the north shore, leaving behind a few small
family groups in a winter hunting range. Spring may have been the time for some

groups to return to their favorite fishing sites and to make sugar; summer, the varied

132

strategies of fishing and hunting of beaver, caribou, and small game. In fall groups
decampcd for mainland hunting ranges where, as dispersed families or hunting
groups, they would pass the winter season. There is nothing to suggest that Isle
Royale was in any way an inappropriate place to spend a winter, although, in the
event of a failure of the resource base on the island winter residents would find
themselves committed to privation until the ice and lake conditions permitted a
return to the mainland. A network of reciprocal relations among winter groups
operating on the mainland would be effectively unavailable to island winterers.
Storable resources such as maize, rice, and dried meats acted as a buffer to seasonal
scarcity in many localities, and one would expect that on an island these would be of
particular value. To date, neither maize or rice has been recovered although it seems
highly likely that both, in some measure, were used by the prehistoric visitors and
residents of the island.

Faunal remains were not sought from archaeological sites on Isle Royale until
the UMMA survey (Bastian 1963a). At this time bone was collected by hand from
the ground or from the screens during excavation, a technique which systematically
favors larger pieces to the exclusion of entire classes of remains (e.g., fish scales).
This bias is expressed in the report on the faunal remains from the Indian Point site
(201R28) in which Cleland (1966) identified beaver, caribou, moose, and lake
sturgeon. The moose is questionable on contextual and historical grounds, and is
probably not associated with the prehistoric occupation of the site. Cleland (1966)

expressed some surprise that more fish was not represented in the collection.

133

The MWAC survey initiated the use of flotation separation of biological
remains from soil matricies from sites on the island, resulting in a much broader
image of prey species and subsistence practices (Appendix B). Martin and Masulis
(1989az44ff) have identified a wide range of fish species from both cold deep water
and warm shallow water habitats. At Chippewa Harbor #1 (201R1) evidence for the
harvest of spring sucker runs and fall spawners exists with a variety of terrestrial
resources. Short term seasonal occupations are masked at this and other large sites
by multiple episodes of reoccupation.

Predictably, more restrictive faunal assemblages are present at smaller sites
(e.g., 201R114, 201R120), but the extent to which these are attributable to any form
of specialization in subsistence practices is equivocal. In my opinion, the small sites
reflect the relationship between the season of occupation, the intended duration of
the occupation, the function or purpose of the occupation, and the surrounding site
catchment. Further, considering the shape of the island, its physiography, and the
transportation advantages of the canoe, it can be argued that at no location on Isle
Royale is one distant from any of the potential resources available there. There is
no clear pattern of site location and specialization in subsistence practices or in
seasonality of occupation, or as stated by Martin and Masulis (1989a), "the pattern
is one of variability."

The MWAC survey strategy of the Isle Royale project concentrated on modern
shorelines and developed areas, thus precluding any statistical treatment of site
location data. Based on the distribution of known sites the following characteristics

are assumed to be among the primary factors in site location: 1) physical

134

characteristics of the setting, 2) proximity to resources, and 3) relations to other
human groups.

Occupation sites are found in settings where there is a gravel beach above
which is a relatively flat (rarely level) surface of sand, soil, or rock. A combination
of physical features which repeatedly occurs on Isle Royale is a configuration of one
or two bedrock projections into the water forming a sheltered gravel beach with older
and higher beach deposits providing an appropriate living area above the beach.
Artifacts and midden deposits have been found in virtually all areas relative to the
beach or bedrock; sometimes reflecting the exposed nature of a point of land as it
projects into the water, or the sheltered aspects of a location as much as 40 m inland
from a beach on a fossil beach strand. Sandy points at or near the lake level are also
common site locations; in each case a highly exposed situation.

As stated earlier, nowhere on Isle Royale is one distant from any of the island’s
resources, provided one has a canoe. The shape of the island, in combination with
its physiographic features render it accessrble with major limitations being vegetative
cover along inland routes and weather conditions for water passage. The major
avenue of traversing Isle Royale along the transverse fault between Chippewa Harbor
and McCargoe Cove reflects the logistical advantage of this route by the presence of
the two largest (most frequently reoccupied) sites: Chippewa Harbor #1 (201R1) and
the Indian Point site (2OIR28). Aside from the logistical advantages of certain
localities, site location relative to food resources are seen as secondary to the physical
setting of the site location. Martin and Masulis (1989a, 1989b) have identified few

sites in which the faunal assemblage is sufficiently specialized to suggest that the site

135

was occupied for the express purpose of exploiting a particular target species or
habitat.

In many instances the close proximity between copper mines and occupation
sites has been noted. At the Siskiwit Mine (201R41), Phelps (201R81) and Singer
(201R80) the relation is immediate or separated by less than 60 m. The Threemile
#1 site (2OIR116) is removed from the nearest known copper mine by a distance of
350 m. By canoe, the relationship between the entire McCargoe Cove area and the
Minong mine is effectively immediate, involving only a nominal commute to enjoy the
benefits of some of the best copper mining on the island with an occupation setting
in easy distance from a wide range of habitats.

One expects that a hunting group newly arrived on the island would anticipate
encountering other groups at the most optimal of site locations. Assuming that social
interaction and information exchange were operating on the decision making
structure of the parties, we predict that these optimal sites, such as Chippewa Harbor
#1 , Indian Point, and Daisy Farm, would be the most heavily used sites on the
island. Most sites conform to a pattern that suggests the movement of very small
units, one or two families, around interior lakes and the Lake Superior shoreline to
harvest island resources. Where copper sources occur sites tend to be larger (more
freqently reoccupied) but continue to reflect the activities of a small group over a
short period of time. I

In certain areas of Isle Royale a "non-site" approach is appropriate (Thomas
1975). It is evident from a map showing the locations of sites in McCargoe Cove that

it is possible to interpret the distribution of sites both as a configuration of

136

penecontemporaneous activities or as a set of discrete and unrelated loci of artifact
concentrations. All occupation sites in this area had immediate access to the Minong
Mine copper resources, and likely incorporated it into a daily round of activities.
Likewise, there may have been an ancillary function of certain sites to others. For
example, McCargoe Cove #4 (201R151), a small occupation, may have served as a
temporary flintknapping site with a hearth used by individuals whose main habitation
was across the channel at Indian Point. Likewise, the rich midden concentration of
McCargoe Cove #2 could have functioned as a secondary site of domestic refuse
which originated at Indian Point. Although a formal non-site analysis is not
developed here, the concept should be borne in mind when considering the functional
relationships of closely spaced sites in a small resource-rich area.

A hypothetical vignette of site visitation/use hierarchy serves to illustrate: a
group consisting of a few families canoes from the north shore to Isle Royale, putting
in at the mouth of McCargoe Cove where they know that they are likely to encounter
others. They stay at Indian Point for a few weeks fishing and visiting with the
relatives/friends they find there, and from whom they learn the disposition of other
groups and resources around the island. The group divides into smaller units which
disperse for short periods to hunt caribou and beaver, and to fish. Over the next few
months they coalesce at popular sites like Chippewa Harbor to visit, after which they
depart to resume their subsistence activities and/or to extract some copper. At some
point the decision is made to stay or leave the island for the winter and, once again
they are found at Indian Point (or Washington Harbor) to prepare for the trip across

the lake to the mainland. Perhaps there was some advantage to making the open

137

water voyages with as large a group as possrble; in which case it would be
advantageous to meet at a Site like Indian Point. The dog and tobacco thrown into
the lake; speeches made; manitous propitiated; the entourage hove off for the
northwest, leaving behind the few families for whom Isle Royale served as a

traditional winter hunting range.

ARCHAEOLOGICAL CULTURES AND INTERACTION

The Terminal Woodland substage on Isle Royale is varied with respect to the
archaeological cultures represented in ceramic style. There are numerous examples
of Huron, Blackduck, Sandy Lake, Selkirk, and J untunen ceramics. But there are few
sherds on Isle Royale which are possrbly Oneota in derivation, in contrast to western
Upper Michigan where Oneota maintains a limited but regular presence in late
prehistory. Similarly, there are few sherds that are immediately comparable to
pottery styles of the Lakes phase of northern Wisconsin, or the Sand Point ceramic
series of the Keweenaw Peninsula. Assuming a correlation between the absolute
minimum number of vessels and the proportionate representation of an actual or
potential presence of a given archaeological culture, the data presented in Table 3
indicate this diversity of Terminal Woodland usage of the island.

Huron pottery was found at 13 Sites. Baker Point, Chippewa Harbor #3, and
Brady Cove #2 appear to be nominally single component Iroquoian sites, insofar as
no other ceramic styles occur in association with them. At Belle Isle #1 and #2 the

context suggests that the pottery was used in a single event not necessarily in

138

conjunction with other ceramic types. At all other sites Huron pottery is found in a
mixed assemblage.

Juntunen pottery is found at eight sites, often with Huron pottery. Although
the historic relations between the Ottawa/Chippewa and the Huron ranged from
mildly antagonistic to amicable (Trigger 1976), the association between these two
ceramic traditions comes as no surprise, representing trade, multiethnic association,
or an coincidence of deposition.

Blackduck and Selkirk cultures are represented at eight sites apiece on Isle
Royale. In most cases there is a nominal association with ceramics indicative of an
eastern orientation. Given the proximity of Isle Royale to the heartland of
Blackduck culture in Ontario, Manitoba, and northern Minnesota, a stronger showing
might be expected on the island.

Sandy Lake pottery was recovered from two sites. If the paucity is a real one,
it may reflect a contraction in the use of Isle Royale’s resources by groups with a
northwestern affiliation from late prehistory and into the early historic period. It is
during the same time that the Huron ceramics begin to make a showing on Isle
Royale sites.

The incidence of Lakes phase and Oneota ceramics in any of the Isle Royale
collections remains very low. This situation is in accord with expectations which hold

Lake Superior as a barrier to regular north-south travel, and which considers the

139

Table 3 Summary of the Terminal Woodland Archaeological Cultures at Isle
Royale National Park.

Archaeological Cultures" n(%)

BDK HUR JUN MAC ONE SAN SEL UNC MNV"
SITE#
201R1 4(16) 1(4) 3(12) 0 1(4) 0 15(60) 4(5) 25
201R5 0 0 0 0 0 0 0 3(100) 3
201R14 n/a
201R17 0 1(50) 0 0 0 0 0 1(50) 2
201R18 0 2(67) 1(33) 0 0 0 0 0 3
201R27 n/a
201st 1(3) 3(9) 0 1(3) 0 1(3) 10(30) 17(52) 33
201R29 1(12) 1(12) 0 0 0 0 3(38) 3(38) 8
201R31 0 0 0 0 0 0 0 1(100) 1
201R41 0 0 3(60) 0 0 0 0 4(40) 5
201R42 0 0 0 0 0 0 0 2(100) 2
201R45 1(33) 1(33) 1(33) 0 0 0 0 0 3
201R52 0 1(100) 0 0 0 0 0 0 1
201R53 0 0 0 0 0 0 0 1(100) 1
201R56 0 0 0 0 0 0 0 2(100) 2
201R65 0 0 0 0 0 0 1(100) 0 1
201R81 0 0 0 0 0 0 0 1(100) 1
201R107 0 0 0 0 0 0 0 1(100) 1
20IR111 2(40) 0 0 0 0 0 0 3(60) 5
201R114 0 1(33) 2(66) 0 0 0 0 0 3
201R116 0 1(50) 0 0 0 0 0 1(50) 2
201R120 0 0 0 0 0 0 0 1(100) 1
201R124 1(100) 0 0 0 0 0 0 0 1
201R127 0 0 1(100) 0 0 0 0 0 1
201R128 0 0 0 0 0 0 0 1(100) 1
201R134 0 1(100) 0 0 0 0 0 0 1
201R140 0 0 0 0 0 0 1(100) 0 1
201R142 0 2(33) 1(17) 0 0 0 3(50) 0 6
201R143 0 0 0 0 0 0 0 1(100) 1
201R147 0 1(100) 0 0 0 0 0 0 1
201R149 0 0 0 0 0 0 0 1(100) 1
201R150 0 0 0 0 0 1(100) 0 0 1
201R152 0 0 0 0 0 0 0 1(100) 1
201R153 0 0 0 0 0 0 0 1(100) 1
201R154 0 0 0 0 0 0 0 1(100) 1

140

Table 3 (cont’d)

201le 0 0 0 0 0 0 0 1(100) 1
201R174 4(71) 0 0 0 0 0 0 3(29) 7
20mm 0 0 0 0 0 0 0 1(100) 1
201R179 1(50) 1(50) 0 0 0 0 0 0 2
201le 0 0 0 0 0 0 0 9(100) 9
201R18] 0 1(50) 0 0 0 0 1(50) 0 2
201R183 0 0 0 0 0 0 1(100) 0 1

 

TOTAL 15(11) 17(12) 13(9) 1(<1)1(<1) 2(<1) 35(25) 54(39) 138

*BDK=Blackduck, HUR=Huron, J UN=J untunen, MAC= Mackinac, ON E= Oneota
SAN =Sandy Lake, SEL=Selkirk, UNC=Unclassified Terminal Woodland

"MNV=Minimum Number of Vessels used to make this determination.

Ontonagon-Keweenaw area as an alternative source of copper not entailing the risks
of travel to Isle Royale. Relations to the south are represented by one equivocal
example of Oneota pottery from the Chippewa Harbor #1 site (Bastian 1963a). A
few ceramics from Chippewa Harbor #1 have been classified as belonging to the
Lakes phase, the Sand Point series, and the Door Peninsula types, in addition to
those related to the Straits of Mackinac sequence (Mark Lynott, personal
communication 1987).

Within the time frame embraced by the Terminal Woodland substage, the Lake
Superior basin acts concurrently as a barrier and facility to mobility. Although canoe
trips out of sight of land are reported (see Little 1987), and Radisson in 1659 states
that trips were made from the Keweenaw Peninsula to Isle Royale when good

weather prevailed, voyages across the lake involved immense risks:

141
They say that from the isle of copper, which is a league in the

lake, when they are minded to thwart it in a faire and calme wether,

beginning from sun rising to sun sett, they come to a great island [Isle

Royale], from whence they come the next morning to firme lande att

the other side; so by reason of 20 leagues a day that lake should be

broad of 6 score and 10 leagues. The wildmen doe not much lesse

when the weather is faire. [Kellog 1926, quoted in Griffin 1961:38]

Among Griffin’s (1961:133) conclusions is a rejoinder to the tenent that "the
Indians would not have been able to canoe to Isle Royale." In this analysis, it is
assumed that it was possrble to travel to Isle Royale across Lake Superior from the
Keweenaw. Insofar as archaeology is concerned with patterned behavior, the
evidence of regular translake trips should be reflected in the archaeological presence
of north shore raw materials on the Keweenaw Peninsula. At the present, there is
but one documented example of jasper taconite from an archaeological Site on the
south shore of Lake Superior: a notched projectile point found on a beach near the
mouth of the Misery River in Ontonagon County, Michigan (Larry Sutter, personal
communication 1987).

The relations between copper mining sites and associated habitation Sites are
largely based on an argument of proximity. On the one hand, there are copper mines
and their associated artifacts, including hammerstones, hammerflakes and, in one
case, chert flakes. On the other hand, there are occupation sites with a relative
wealth of artifactual material, including copper and stone waste material, copper and
stone tools, ceramics, food remains and domestic features. It is not possrble in most

cases to demonstrate relations between specific sites other than by the logic of

association of copper source and copper tool/ornament fabrication. The relationship

142
between copper mines and specific occupation sites is discussed at length in the
following chapter.

The question of propriety of a group or family over a given mine is raised by
the apparent discreteness (i.e., the apparent single component nature) of the
archaeological remains at some occupation sites. Le J eune’s 1636 oft cited reference
to the "master of the trade line" may serve as an hypothetical construct for
understanding the "ownership" of a copper pit, the first to discover it or work it being
the practical owner. The observation was made concerning the Huron, but has been
applied as a heuristic device in the Great Lakes region:

And first, concerning commerce; several families have their own
private trades, and he is considered Master of the line of trade who was

the first to discover it. The children share the rights of the parents in

this respect, as do those who bear the same name; no one goes into it

without permission, which is given only in consideration of presents; he

associates with him as many or as few as he wishes. [Thwaites JR

10:223-225]

However, physical possession and occupation of a mine may have defined
"ownership" with the subsequent abandonment terminating any proprietary
considerations. Although it remains an open question, framing the relationships
between access to unique and localized resources and trade involving those resources
is an important step in understanding the systematics of the relation between the
ethnicity manifest in the material culture of a Site and that groups’ role in regional

trade and exchange. We would predict that more strict hegemony over copper

resources would obtain where territorial lines are fairly distinct, the corollary being

143

that, in cases of high group mobility and permeable or fluctuating band territories,
access to copper resources will be less restricted or unrestricted. A third alternative
is that Isle Royale’s copper resources, and the spiritual qualities embued therein
constituted an open area of free access to all groups.

Travel parallel to the lake shore, on the other hand, increased mobility and
made possrble the transportation of quantities of goods that would have been
otherwise quite impossrble overland. In agreement with Bastian’s (1963a:319)
conclusions, extant data suggest that goods and persons made frequent use of Lake
Superior’s east-west routes throughout prehistory, both on the north shore and on
the south, resulting in the opportunity for interaction among groups of Algonkian,
Iroquoian, and Siouan speaking peoples. In her study of long distance travel Little
(1987) catalogs historic documentation which descrrbes canoe trips in the Great
Lakes. Marquette travelled from St. Ignace to Green Bay via Lake Michigan in 21
days, Tonti from Green Bay to Ft. St. Joseph down the west side of Lake Michigan
in 30 days and Chicago to St.Ignace in 14 days, and Bonnecamps travelled up the St.
Lawrence river to the Niagara on Lake Ontario in only seven days (Little 1987:62).
Voyageurs and engages could make even better time when so motivated.

It is 35 km to Grand Portage, Minnesota, and 90 km to Copper Harbor on the
Keweenaw Peninsula of Upper Michigan. Nonetheless, there has been a strong
tendency to associate Isle Royale with the copper district of the Keweenaw and
Ontonagon, and to portray Isle Royale as it is today, a part of Michigan and the

United States. The idea that most travel to and from Isle Royale originated from the

144

Keweenaw Peninsula to the south is pervasive in many of the writings concerning the
prehistory of the island (e.g., Dustin 1957, Griffin 1961, Karamanski, Zeitlin, and
Derose 1988). The reasons for this conclusion emanate from two related sources,
one is an implicit belief in the unity of the copper district which links Isle Royale to
the Keweenaw and Ontonagon areas of the Upper Peninsula of Michigan, and
perceives the entire copper-producing area as an entrepot for points south which saw
elaboration of the copper industry in the Late Archaic "Old Copper" complex and in
the Middle Woodland substage. The second source is a continuing reluctance to see
Isle Royale as part of the north shore’s environment and cultural ecology, and to
ignore the presence of alternate sources of copper on Superior’s north shore. This
does not mitigate either the uniqueness or importance of Isle Royale as a major
source of copper. For the entire north shore of Lake Superior, Isle Royale served
as the single most important mining district, independent of the alternative sources
across the lake to the south.

Archaeological evidence for prehistoric use of Isle Royale, either as a copper
source or subsistence option, favors the north shore as a point of origin. Evidence
in the form of lithic raw materials unequivocally points to the north shore, the
Thunder Bay vicinity in particular, as the departure point for trips to Isle Royale.
Whatever the purpose for the trip, the need for stone tools was anticipated by the
importation from the mainland to the island of finished items as well as unfinished
cores and blanks. Lithic raw materials contain a high proportion of types whose
primary source is in the Thunder Bay area (taconite, Gunflint formation cherts and

carbonate Shales), although trace amounts of exotic cherts do occur. Quartz and

145

quartzite which dominate much of the lithics in the Upper Peninsula of Michigan are
rare. Unfortunately, there are no good primary sources of chert in western Upper
Michigan where Hudson Bay Lowland cherts supplied the necessary raw materials to
prehistoric knappers. Thus, it is impossrble to categorically state that raw materials
obtained in the Upper Peninsula do not occur on Isle Royale sites. On the other
hand, it is possrble to state that the common raw materials from the north shore do
not, with a single exception occur on sites in Michigan’s copper district. Isle Royale
was part of the north shore cultural sphere, supplying northwestern Ontario with
much of its raw and processed copper amid the interactions of a variety of
archaeological cultures representing the Straits of Mackinac to the east, the rice and
lakes district to the west, and horticultural Huronia on northeastern Lake Huron.
The motivation for someone to travel across Lake Superior from the south would
necessarily be related to something other than copper, since copper is available in the
Keweenaw and Ontonagon areas without entailing such extreme risks.

If ceramics are discounted as trade items, there is almost no evidence of the
movement of nonlocal goods to Isle Royale. As we have seen, lithic raw materials
were obtained from the Thunder Bay area, and copper, of course, was obtained from
the island itself. Exotic durables such as marine shell, and foodstuffs such as wild
rice and maize have not been found here. This apparent absence of evidence of
trade/exchange must not be construed as evidence of nonpeaceful relations, however.
As Gary Wright (1974) has noted, differential preservation probably accounts for
much of the one sidedness of the archaeological record. Overall, the interaction

among Terminal Woodland groups, as evidenced by the almost ubiquitous

146

co-occurence of Juntunen, Huron, Selkirk, and Blackduck ceramics on many sites,
must have been at least amicable.

The only possible case of violent interaction is found at the Massee
Rockshelter (201R14), the only known mortuary site on the island, and even here the
interpretation of violence is, at best, speculative. Sauer (1990) found evidence for
scalping on one skull. This, in combination with the recovery of two projectile points
amid the scatter of secondary interrnents, may indicate that some of the individuals
buried there met a violent death. However, there are no diagnostic artifacts in
association with the remains to indicate which archaeological culture(s) may be
represented by the skeletal population. The site remains an anomaly with only
uncertain bearing on the issue of interaction in the Terminal Woodland substage.

In summary, it appears that several archaeological cultures, Juntunen and
Huron from the east and Blackduck and Selkirk from the west, had equivalent access
to Isle Royale. But was the nature of access the same in each case, or were certain
resources the domain of specific groups? Was the island a neutral area for all who
took the trouble to get there? Or was there a substantive difference between the
types of resources, copper and subsistence resources for example, that limited the
activities of certain groups visiting the island? In the following chapter we examine
the role of copper in a regional setting to assess the potential of copper as an

indicator of differential interaction in the Terminal Woodland substage.

CHAPTER 6
COPPER AND COPPER MINING IN THE

TERMINAL WOODLAND SUBSTAGE.

INTRODUCTION

To what extent did the copper resources of Isle Royale serve as a draw on
the prehistoric groups of the surrounding area and to what extent did copper serve
as a medium of interaction among these groups? Of what value was copper in the
systemic context of the Terminal Woodland in the Lake Superior region?
Archaeologically, how should we use copper as an indicator of group interaction?
When we examine questions of value, whose value is it that is being measured? In
order to assess these questions it is necessary to make certain determinations
concerning copper and, where determinations are precluded by an absence of
substantive data, assumptions. In this chapter unique resources are discussed with
respect to how they condition and are conditioned by social realities, and how copper
as a resource of the Lake Superior basin is distributed archaeologically. These data
are then used to address the role of copper as a mediating factor affecting group
interaction in the Terminal Woodland substage.

Europeans first became aware of Great Lakes region copper in the early
sixteenth century when copper ornaments were found among the Indians along the
St. Lawrence River valley. In the winter of 1535-36 Cartier was entertained and
intrigued by tales of the "Kingdom of Saguenay" which lay in some ambiguous

direction to the north and west and from whence had come the copper objects in

147

148

possession of his hosts, the St. Lawrence Iroquois. In subsequent years there was a
flow of rumor and information concerning the source of copper. In 1603
Champlain’s Huron informants described a copper mine somewhere to the north
where they had obtained some bracelets. Seven years later on a trip upriver from
Quebec, Champlain encountered some Montagnais and a single Algonkian who
presented him with a piece of native copper said to have been obtained from the
bank of a river (Biggar 1925).

With the establishment of a permanent French presence in Huronia and the
burgeoning fur trade, more substantive information concerning the source of copper
was forthcoming. Penetration into the Superior basin by the traders Brule and
Grenole around 1623 brought back copper specimens and more tales of mining from
Indians living to the west of Huronia. No mines were actually visited, however, and
much of the information was still couched in vague and often mythic language. In
his review of references to copper in the Jesuit Relations, Whittlesey (1863)
concludes that all such references prior to 1847 are, at best, second hand in nature.
Boucher writing in the 1640s descrrbes the process of copper extraction by making
"fires on top of it, and then hew pieces out with their axes," a process he probably did
not witness. Dablon (Thwaites JR 1669-70) makes mention of the tradition of a
floating island of gold (the type of metal is likely an elaboration on the part of the
translator) 40 or 50 leagues north of the Sault and opposite Michipicoten Island. His
native informants, he tells us, lacked consensus on the matter; "the savages did not

agree as to the source of the copper. Some say it is where the [Ontonagon] river

149

begins, others say that it is close to the lake, and others at the forks and along the
eastern branch of the river," (Winchell 1881:613).

European interest in Lake Superior copper may be characterized as sporadic
and uncertain according to the scattered references in the Jesuit Relations. Some
prospecting was done by the French in 1739 at bidding of De la Ronde who
employed two German miners in the Ontonagon area. Alexander Henry, in 1771,
established a mine on the Ontonagon River which by the following year had failed.
Mining of copper and silver by Euro-Americans did not begin on Isle Royale,
however, until 1847 by which time any tangible link between the prehistoric miners
and their historic survivors had been largely obscured by the processes of cultural
change. Goods of European manufacture had since supplanted the exchange in
native copper although it continued in use locally and was revered as a manitou in

its natural state (Halsey 1983).

GEOLOGICAL DISTRIBUTION OF COPPER

Primary deposits of native copper are found at several places around the Lake
Superior basin and along the Keweenaw rift to the southwest (Figure 15). South
shore sources are found from the St. Croix River on the Minnesota-Wisconsin border
northeast to Ontonogan and the Keweenaw Peninsula (Huber 1975). The Snake and
Kettle rivers in Minnesota, tributaries of the St. Croix, are also mentioned as a
copper source (Irving 1883:241). Copper-bearing rocks are found on the north shore
from Duluth to Grand Portage. There is a gap in the distribution of copper sources

from the international border to Black Bay, Ontario. Irving’s (1883) survey of copper

150

 

I KEWEENAWAN VOLCANIC SEQUENCE

 

3% COPPER HARBOR CONGLOMERATE

Figure 15. Distribution of Geological Copper Sources.

151

bearing rocks of the Lake Superior basin specifically names the locations of copper
sources from Thunder Bay to the Sault: Edward Island, Point Porphory (Black Bay),
Battle Island, Flour Island, Simpson’s Island, St. Ignace, North Bay, Michipicoten
Island, Point Gargantua, and Point Mamainse (Irving 1883:78).

Secondary sources of copper are found to the south of the areas of primary
deposits; particularly in Minnesota, Wisconsin, and the Upper Peninsula of Michigan.
Glacial till deposits and sorted till and outwash occurring on river bottoms and
lakeshores in this area have produced large amounts of native copper, ranging from
small pieces to large masses weighing thousands of pounds (e.g., Winchell 1911:505).
Drift or float copper, as it is called, has the advantage of being free of surrounding
matrix, making it easier to work into large items. However, this advantage is
diminished by the difficulties in locating float copper. No data exist by which to
assess the relative frequency of the use of secondary copper as opposed to primary
sources. It is assumed that, in most instances, the procurement of float copper was
the result of accidental finds, although it is expected that a level of awareness of the
likelihood of encountering copper in certain areas must have operated. This was
certainly the case in the "Old Copper Culture," the distrrbution of which is largely

coterminus with the distribution of float copper south of Lake Superior.

DISTRIBUTION OF ABORIGINAL COPPER MINES
Given the large area of primary copper sources, the number and distribution
of aboriginal copper mines is surprisingly limited. Excluding Isle Royale, prehistoric

copper mines are reported from few locations: the eastern end of Lake Superior at

152

Point Mamainse where three pits are recorded (Griffin and Quimby 1961:77), the
Ontonagon and Mass City area, and on the Keweeenaw Peninsula at the Calumet
Ancient pit (Griffin 1966) (Figure 16). By contrast, there are 15 prehistoric copper
mines recorded on Isle Royale with anywhere from one pit or fissure to more than
one hundred. This reflects the archaeological knowledge of the region, rather than
the actual distribution and density of copper mining in prehistory, and explains the
historic tendency to over emphasize the role of Isle Royale as the source of
prehistoric copper.

Tyler Bastian’s (1963b) thesis on the aboriginal copper mines of Isle Royale
is the most comprehensive treatment to date. Although inconclusive with respect to
the cultural-historical aspects of prehistoric mining and the systemic cognates of
copper extraction, fabrication, and redistribution, Bastian describes the geological
context of native copper deposits on Isle Royale, the aboriginal copper mines, and
associated artifacts. He defines two forms of aboriginal mines: pits and fissures. Pit
mines are, as the name implies, generally circular excavations in bedrock usually
associated with a lode deposit, while fissures are linear excavations that follow a
copper vein.

Seven copper mines on Isle Royale can be placed in the Terminal Woodland
substage, some by radiocarbon dates, some by direct association of Terminal
Woodland artifacts, some by their elevation relative to modern lake levels, and some
by an inference of proximity to documented Terminal Woodland occupation sites.
Only one other site in the entire region, the Calumet Ancient Pit on the Keweenaw

Peninsula with an uncorrected radiocarbon date of AD. 770 +/-140 (M-1776) has

153

.352 8&8 358205 me 83:95me .3 oSmE

 

><m m204<2 .. .

1‘.

2825.. \ W. ..
:2:me .v \ 0202.:

Ewan—“WM. u .m>oo 39.56:

._ \ ..
.5... .\ \\
\\\\ \ \

\\.\\,y‘
\

hDOXOOJ

osz.
$ .030255
.W/wfiuza
.\\ 5.525
02.53
OER SO
NU

m5: 02(40003 122—Sam... 0
m3: GEOhwimm—a O

 

r 3 3205.200 .
oflo
$225: ,0)
. 5.0.. 8.9
. v

0 356: \

mam.

 

 

 

 

154
been documented for the Terminal Woodland (Griffin 1966), although there are no

diagnostic artifacts associated with the site. The following detailed site descriptions
are used to deduce the social and technological requirements of extraction and
fabrication of Isle Royale copper in the Terminal Woodland substage.

2OIR30 (LOOKOUT): One of the first aboriginal mines identified by whites
on the island, this site was first discovered by Shaw in 1847 who excavated at least
a portion of one pit (Foster and Whitney 1850:162). The site was then forgotten
until its rediscovery by Park Superintendent R. Gibbs in 1950. Ten years later the
UMMA survey located five fissure mines and excavated two pits (Pits 4 and 56)
(Bastian 1963az49-53). In addition to the wide range in radiocarbon dates, six
hammerstones and five "scrapers or choppers" were found.

2OIR41 (SISKIWIT MINE): This site is a combination of mine and
occupation/fabrication site on the north side of the Rock Harbor channel opposite
Mott Island. The Siskiwit Mine site is the location of a prehistoric mine and
occupation, an American Fur Company fishing establishment, and a nineteenth
century copper mine (Bastian 1963a:55-56, Clark 1987).

The University of Michigan survey (Bastian 1963a) recovered 16
hammerstones from a fissure mine at this locality. Crane and Griffin (1965:128)
report an uncorrected radiocarbon date of 1420 BC. +/- 130 from charcoal collected
from the mine (erroneously reported as coming from 201R6, the "Siskiwit [River]
site"). In 1986 a sheet midden was located and samples recovered. The midden
contained beaver and fish bones, stone tools and debitage, copper waste, and very

small amounts of nondiagnostic pottery. A radiocarbon date from the midden places

155

the occupation in the Terminal Woodland and Historic time frames (Clark 1987). In
order to clarify the nature of the occupational deposit and its relationship to the
mining area above, the Siskiwit Mine site was selected for limited test excavation in
the spring of 1987.

Historic artifacts were limited to the first 27 cm BS. Prehistoric materials
were present in small numbers in this mixed upper zone. Feature 1 was initially
defined as a hearth feature on the basis of a shovel test (Clark 1987). The black soil
was found to be a sheet midden with a maximum thickness of 22 cm. It was
restricted to the westernmost excavation unit where it was encountered from 12 to
34 cm BS. Midden contents included copper in both waste and finished tool forms,
unifacial stone tools and debitage, abundant calcined bone, wood charcoal, and
several examples of Terminal Woodland pottery. No historic artifacts were
recovered from the midden.

Feature 2 was defined at 17 cm BS as an area roughly 80 cm by 80 cm of
closely spaced pieces of FCR and dark charcoal rich soil. Thirty centimeters in
thickness, the feature ended at 47 cm BS. In profile Feature 2 consists of five
discrete layers of alternating sterile soils and organic rich fill. The uppermost level
was the area of greatest FCR density, although F CR continued throughout the
deposit. At 30 cm BS in the center of the feature a bright red-orange lens of burned
clay was found. Radiocarbon dates are consistent with the artifacts and previous
date, placing occupation at the Siskiwit Mine site in the Terminal Woodland substage.

A total of 27 pieces of waste copper and two finished tools were found in the

1987 excavation units. Initial stages of fabrication are well represented by raw

156

copper with and without matrix, as well as examples of all flat copper waste types.
Two small butterknives were also found in the midden deposits. Both exhibit a
narrow tang and are made of relatively homogeneous pieces of copper although
laminated structures are present on both. They are quite short, even for this category
(43.21 mm and 35.58 mm).

Fragments of three Terminal Woodland vessels were recovered from the
excavation units at the Siskiwit Mine site. All have typological affinities with
Juntunen phase pottery of the Straits of Mackinac sequence.

Faunal remains at the Siskiwit Mine site consist of 1600 total elements, 1100
of which are identified to class (T. Martin and Masulis 1989a:21-26). Fish constitute
40%, including lake trout (91% of all fish) and lake Whitefish. Mammals account for
48% with examples of cow, pig, moose (?), caribou, beaver, and hare. Small amounts
of loon and turtle were also recovered.

201R4§ (ISLE ROYALE AND OHIO MINE): Located above the Daisy
Farm campground (see 201R45, Appendix A), the Isle Royale and Ohio Mine, also
known as the Ransom site, operated between 1846 and 1849. In 1986 a single
hammerstone was located here, identifying the site as an aboriginal mining site (Clark
1987:23). Based on its close proximity, this site may have served some of the
immediate needs of the Daisy Farm occupants for copper. It should be noted that
the Daisy Farm site has both Laurel and Terminal Woodland components, making
the temporal placement of the associated mine site uncertain.

201R80 (SINOER): The Singer site is located on the south side of

Washington Island on steep bedrock (Portage Lake flow volcanics) just to the east

157

of a level saddle which formerly was the site of a cabin. A site was recorded by the

UMMA survey and described by Tyler Bastian as a historic mining prospect:

There are at least seven veins in a distance of about 400 feet
which have been very superficially mined or prospected. Drill holes
are present in a few cases indicating that at least some of the mining
can be attributed to Euro-American miners, but there is no way to
demonstrate that any of the veins have also been worked by the
Indians. The only known historic records of mining on Washington
Island are from 1847 or earlier (Foster and Whitney 1850:91, Jackson
1849:426). The workings of the Ohio and Dead River Company on the
north side of Washington Island could not be located by the survey.
[1963a:70]

Dustin notes that early mining explorations on Washington Island, "have been
reported on the site of ancient works" (1957:11). The location indicated by Bastian
was relocated by the NPS-MWAC crew in 1989 which found that the site is an intact
and undisturbed aboriginal copper mine with associated copperworking and living
sites (see 201R160, Appendix A). No evidence of historic disturbance was found and
Bastian’s assessment remains a puzzle. There is a gap in site information between
2OIR79 and 201R82 in Bastian’s field notes, suggesting that this site and Phelps
(20IR81) were described on the basis of secondhand information.

Three mining pits (Features 1-3) were identified and documented and two
test units excavated on a level area immediately above the pits. Hammerstones,
hammerflakes, and worked copper were collected from the exposed surface, shovel

tests, and excavation units.

158

Feature 1, the largest of the mine pits and the one closest to the water, was
located by following a copper vein from the water’s edge upslope. At the time of
its discovery Feature 1 was water filled. The water was pumped and bailed to reveal
numerous large to small boulders, sand, and driftwood. All but one large boulder
were removed. The pit has dimensions of 3 m following the vein, 1.75 m
perpendicular to the vein, and is 1.25 m deep, measured from the water mark. The
vein of pure copper averages 5 cm in width and exhibits battering from wave action
on the cobbles and debris which had collected in the pit.

The second pit, just above Feature 1 and following the same vein, measures
2.2 x 1.0 x 0.3 m deep. Large rocks that had fallen from the bedrock face above this
pit have filled the upper end of the feature and were too heavy to remove. Metallic
copper is visrble at the southern end of the pit.

A third feature along the same vein was not excavated, although it has the
greatest potential for in situ remains from prehistoric mining activity as it is well
above the level of Lake Superior and is spared the impact of the waves. The pit is
filled with vegetation and is partly obscured. It is roughly rectangular in shape,
measuring 65 x 47 cm. Depth is at least 37 cm.

Two excavation units were located on a level area just above these pits to
investigate the possrbility of copperworking activities. One shovel test contained one
piece of flat, vesicular copper. The unit at this location, however, was otherwise
sterile. Owing to the saturated condition of the soil it was excavated to a depth of
15 cm BS. A second test unit 5 m to the southwest contained one piece of flat,

laminated copper and two basalt hammerflakes.

159

Hammerstones were collected from three areas of the site. A small basalt
fragment was located on the surface between Features 2 and 3; a quartzite hammer
fragment was found in an a shallow depression above Feature 1. In an actively
eroding area to the east of the primary vein, one complete and two fragmentary
basalt hammerstones were collected. Massive boulders in this area have thin sheets
of copper on some outer surfaces which could be easily detached, suggesting a low
energy alternative to copper extraction.

The activities reflected in this site indicate only copper extraction and
preliminary fabrication. The nearby occupation site (201R160) probably served as
the focus of related domestic and subsistence activities for the miners of the Singer
site.

PHELPS (201R81): Phelps Island was one of the early names for what later
became Washington Island. There is no known record of who Phelps was in the
history of the island, although this information might shed some light on the basis
upon which Bastian initially identified this site. The name Phelps was not found on
any of the available documents of persons holding mining claims or among the
employees of the Wendigo Mining Company which operated in this area in the late
nineteenth century. As was the case with the Singer site, the Phelps site was
apparently misidentified by Bastian:

The mines are near the middle of the south coast of Washington

Island about 2100 feet due east of the line between sections 9 and 10.

There are at least five mines in a distance of 300 feet, but only one

amounts to more than a superficial prospect. As in the case of the

mines at the Singer site, it was not determined if they had been worked
by the Indians. [1963a:70]

160

In an attempt to ground truth Bastian’s report, the NPS-MWAC survey team
found one well defined aboriginal pit with hammerstones, hammerflakes, worked
copper, and a small quantity of Terminal Woodland ceramics lrberally strewn around
the pit and across the bedrock surface of the site. No evidence of drilling or other
historic disturbance was noted. Although only one pit was identified, other areas of
exposed copper veins were recorded to the southwest of the site. These areas are
subject to heavy wave action from Lake Superior and, as a consequence, no surface
evidence of aboriginal mining was preserved.

The mining pit (Feature 1) is located astride a copperbearing vein. The pit
measures 3 m along the vein, 1.4 m perpendicular to the vein; its depth could not be
determined due to the fill in the pit. Test unit #1 was initiated at the northern edge
of the pit where a natural drainage had redeposited soils containing ceramics and
hammerflakes. Fragmentary hammerstones and hammerflakes were abundant around
the orifice of the pit.

Localized shovel testing in the area immediately above the pit feature located
an activity area associated with the mining. One shovel test contained black
anthrosols with five chert and two basalt hammerflakes. A second unit, TU #2, was
opened to explore this deposit. A thick duff layer extended to 13-20 cm BS where
it gave way to a dense deposit of cultural material in a matrix of black loam and pea
gravel. This midden deposit (Feature 2) persisted to a depth of 32 cm BS and
contained a sandstone anvil, hammerflakes, worked copper, and small amounts of

chert debitage, sherds, charcoal, and calcined bone.

161

The lithic artifacts from this site are almost exclusively associated with the
extraction and fabrication of copper. A total of twelve chert flakes include HBL (10),
J ST (1), and PLQ (1). By contrast, there were 125 hammerflakes, 35 of which were
from the mine pit. Basalt is the most common material. Ten hammerstone
fragments and two complete hammerstones were collected from the combined
provenience units. One is bilaterally notched for hafting. All specimens are basalt
except for one example of granite. The sandstone anvil from the midden exhibits a
central area of pitting on an otherwise smooth surface. The lateral edges of the anvil
are trimmed by flaking. No copper tools or ornaments were recovered although
limited testing produced a typical range of waste categories.

One neck, 13 body, and 26 indeterminate sherds were collected from Feature
1, seven of which are attributable to a pipe or miniature vessel. The midden feature
contained four indeterminate sherds from a single vessel. The ceramics are
significant as the first which are directly associated with a mining feature on Isle
Royale; unfortunately they are not diagnostic beyond a general similarity to Terminal
Woodland pottery.

The Phelps site, in combination with Singer and Washington Island #2,
constitutes a configuration of activities associated by proximity and necessity. Copper
extraction and initial processing took place very near a site of domestic activity where
pottery and tools used in tasks not related to mining were repaired and discarded.
The midden feature is likely a product of fires for annealing and warmth. While it

meets the requirements of artifact and feature content of our operational definition

162

of an occupation/mine site, the physical characteristics of the site suggest a high
intensity of copper working, but low intensity of domestic activities.

201R118 (2 12 MILE): This site was recorded as a spot find at the time of
its discovery, but during a reexamination of the site hammerstones, hammerflakes,
and chert flakes pointed to a much more complex deposit. The site is located about
400 m from the Threemile campground in a broad area of exposed bedrock and jack
pine within 20 m of the Rock Harbor channel.

Artifacts were collected from the main trail and its branches over a distance
of about 70 In. Major concentrations of hammers and hammerflakes occur
immediately above two vegetation-filled depressions designated mine pits #1 and #2.
Additional pits are located nearby: a total of five pits were mapped as probable
mines. A small number of chert flakes were collected from one area adjacent to pit
#3: an unusual association for a mining site. Hammerflakes were collected from a
bedrock exposure at the northeast end of the site (where the initial spot find was
found), some distance apart from the pits and major concentrations of artifacts,
suggesting an ancillary area of initial processing.

Seven hammerstones, 19 hammerflakes, and 12 chert flakes were collected
from the vicinity of the mine pits. The hammerstones are all waterworn basalt
cobbles, probably collected from the beach immediately to the west of the site. All
show well developed battering on one or more surfaces and several have massive
flake sears from the detachment of hammerflakes. One unusual hammer has, in

addition to the typical wear on both poles, an area of battering and flake detachment

163

on its midsection indicating that it was held by both ends while the middle of the
hammer served as the percussive surface.

The hammerflakes constitute an important means of identifying an aboriginal
mining area, even when the pits are filled with vegetation and difficult to recognize.
They exhibit the same morphological characteristics as other flakes, but are usually
larger and are made of igneous rock rather than chert. Platform characteristics are
a function of use and do not reflect modification for detachment found in reduction
trajectories for chert or quartz tools. Of the 12 chert flakes recovered, six are J ST
and six are unidentified chert. All are very small in size and represent on-site
maintenance of chert tools.

The 2 and 1/2 Mile site is probably related to the Threemile site (see 2OIR116,
Appendix A), the latter serving as the occupation and fabrication locus of a small
Terminal Woodland encampment (Clark 1989a).

291R183 (MALONE BAY #3): The Malone Bay #3 site is the only recorded
aboriginal copper mining site which occurs in the Copper Harbor conglomerate
formation. Located on the south shore of the island a short distance from the
Malone Bay Campground, this site was initially identified by the presence of
hammerstones, worked copper, and ceramics lying on the open surface of a flat
conglomerate shelf, sparsely covered with mosses and juniper. The site was mapped,
and careful removal of moss revealed the presence of additional artifacts, mostly
hammerflakes and hammerstones, and a pot break was found near the edge of a
fissure. Closer examination of the fissure revealed a seam of copper running fiom

the water of Lake Superior up to the level of the site about 5 m above the lake.

164

Here the single fissure split into two which continued in a northeasterly direction
defining the southern edge of the site. Three excavation units were dug to further
investigate this site. Artifacts include hammerstones, hammerflakes, worked copper,
chert and quartz debitage, and a Sandy Lake (grit tempered) pot break. No midden
or datable materials were recovered. This Terminal Woodland site probably
represents a single episode of copper extraction and fabrication conducted by a small
group with affinities to northern Minnesota.

The content and distribution of mining sites, associated fabrication loci at
occupation sites, and sites more peripheral to the copper industry on Isle Royale
establishes a pattern which can be applied to other areas within the Lake Superior
basin.

Mining sites must be located where there are primary sources of copper and
are characterized by the presence of hammerstones, hammerflakes, and pits and/or
fissures in bedrock matrix. In at least two instances (Malone Bay #3 and 2 1/2 Mile)
the presence of aboriginal copper mines was inferred prior to their actual discovery
based on the surface recovery of these artifacts in a quantity sufficient to indicate
that a mine was nearby.

Fabrication loci may be found at any distance from the actual source of
copper; either immediately proximate to the mines or well away at a location more
suitable for occupation and subsistence preferences. Fabrication is indicated by the
presence of worked waste copper and smaller hammerstones than those used for
mining. On Isle Royale, fabrication loci are most frequently associated with

occupation sites, and only at the Phelps site is the association an immediate one.

165

More typical is the spatial separation between mine and fabrication loci, as is the case
between the 2 1/2 Mile and Threemile sites, Singer and Washington Island #2, and
more generally between the Minong Ridge area and the McCargoe Cove sites,
respectively.

Sites peripheral to the extraction and fabrication of copper, but where copper
was used are more typical of the areas outside the copper range, reflecting the
movement of copper from source to ultimate place of loss or discard. Finished items,
typically beads, awls, small knives, or other utilitarian items are usually encountered
at these sites.

These patterns are used to evaluate the distribution of copper in the Lake
Superior basin and surrounding areas. These artifact patterns suggest, on a regional
level, the degree to which specific archaeological cultures participated in copper
mining or fabrication, or if their acquisition of copper was dependent on some agency

other than direct procurement.

DISTRIBUTION OF ARCHAEOLOGICAL COPPER IN THE TERMINAL
WOODLAND SUBSTAGE

The subsequent movement of copper opens the question of relations with
other groups and the nature of social interactions required to either preclude access
to copper or make it available to one’s neighbors. This inquiry begins with an
examination of the geographic distrrbution of archaeological copper around the Lake
Superior basin as means of determining whether or not certain archaeological

cultures seem to have had differential access to, or interest in, copper. Renfrew

166

(1977) offered a number of alternative models for reviewing the distribution of
unique resources across cultural landscapes, all of which require the reduction of
archaeological data to mathematical equivalents. Another requisite is knowing in
precise terms the source of the unique resource so that the relation of the
distribution of the products originating at that sources may be measured.
Unfortunately, we are not in a position to differentiate Isle Royale copper from
Keweenaw copper, or any of the other alternative copper sources within the Lake
Superior basin, rendering any mathematical approach absurd. Trace element
analysis is currently in progress that may make this possible in the future. Therefore,
the distribution of Terminal Woodland copper in the Lake Superior basin is
presented in a narrative form which considers in general terms the density, form, and
context of copper artifacts.

NORTHWESTERN ONTARIO AND NORTHERN MINNESOTA:
Published sources were consulted to investigate the distn'bution of archaeological
copper in the area viewed as the territory of the Blackduck and Selkirk
archaeological cultures. The area includes the western end of the Lake Superior
copper district (i.e., Isle Royale and north shore sources from Minnesota at least as
far east as the Sault). A survey of sites from the western end of Lake Superior to the
northwest, including parts of northwest Ontario, northern Minnesota, and
southeastern Manitoba suggest, informally, a sudden decline in density with increased
distance from the geological sources of copper. An inventory of copper tools,
ornaments, and waste copper from Isle Royale is presented for the Terminal

Woodland sites in Appendix A.

167

Cultural site surveys in the Superior National Forest of northern Minnesota
have contained only small amounts of copper. Out of 39 sites with Terminal
Woodland diagnostics (typically Blackduck pottery) four also yielded copper artifacts
(Peters et al. 1983, Peters 1984, 1986). Most sites found in the forest survey are lithic
scatters and probably include a number of aceramic Terminal Woodland components
which are not counted here. But even with the addition of these, the numbers of
sites which contain copper artifacts do not increase significantly. Excavation by the
Superior National Forest at the Big Rice Lake site revealed a large occupation with
Blackduck and Sandy Lake ceramics. Copper artifacts were limited to a single awl
and three pieces of waste (Peters 1984).

At the Martin-Bird site on Whitefish Lake (southwest of Thunder Bay) one
awl and 46 pieces of worked copper were found in what Dawson (1987) interprets
as an annealing hearth. Also on Whitefish Lake is the Mound Island site where
Dawson (1978) found a major occupation site with Blackduck, Mackinac (?),
Pickering Branch, and Peninsular Woodland ceramics: no copper was recovered
here. In a survey of sites at Voyageurs National Park on the international border,
including Rainy, Namakan, and Kabetogama Lakes, only three sites (21SL47, 141,
and 153) out of a total of 35 Terminal Woodland sites contained one or two pieces
of copper in association with Blackduck and/or Selkirk pottery (Lynott, Richner, and
Thompson 1986). Nearby at the Long Sault site on the Rainy River, which was
intensively excavated and produced evidence for a Blackduck and Selkirk habitation

and burial site of long duration, one copper awl was recovered (Arthurs 1986).

168

In Kenyon’s (1986) survey of burial mounds in western Ontario one finds only
sporadic use of copper associated with burials. At the Hungry Hall site on the Rainy
River only two copper beads were found with Blackduck burials in Mound 1,
although there was a wide assortment of other goods included (e.g., clay pots, marine
and freshwater shell and bone beads, scrapers, bone awls, and other worked bone,
shell gorgets, and a lump of hematite). At Mound 2 grave goods included clay pots,
marine and bird bone beads, Catlinite tubes, and a bone awl. Kenyon describes a
number of other Blackduck mounds which have sparse or no grave goods and lack
copper.

At the Ballynacree site in the Lake of the Woods area no copper was
recovered in the excavation of the Blackduck and Selkirk occupation (Rajnovich and
Reid 1987). Also near Kenora, the Ballysadare site produced no copper (Rajnovich
1980). The Spruce Point site at the north end of the Lake of the Woods also failed
to produce copper with its late prehistoric Blackduck and Selkirk occupations
(Rajnovich 1983). North of Kenora at the Forestry Point site one awl and two pieces
of waste copper were found with Blackduck and Selkirk ceramics (Pelleck 1983).

The West Patricia District of northwestern Ontario covers 223,500 square
kilometers (86,300 square miles) of rivers and lakes of the southern boreal forest.
The extensive survey of this area disclosed approximately 364 Terminal Woodland
sites of which eleven contained copper artifacts (Reid 1980, Reid and Ross 1981,
Ross 1982). In most cases the components are identified as Blackduck and/or
Selkirk, and in most instances where copper artifacts were found, only one item is

involved. Artifacts include butterknives, awls, bars, and socketed or conical points.

169
A survey in the North Canbou Lakes area 500 km north of Isle Royale

disclosed five Terminal Woodland sites, no copper artifacts were found (D. Gordon
1985). Similarly, a survey in the extreme southwest corner of northwestern Ontario
which found 14 Blackduck and Selkirk sites contained no copper (Halverson 1988).

NORTH CENTRAL ONTARIO: By contrast, Terminal Woodland sites
along the north shore of Lake Superior contain larger amounts of copper artifacts.
The attribution of this copper to a specific archaeological culture is, however,
rendered uncertain by the association of the widest range of ceramics. The problem
is further exacerbated by the presence of primary copper sources at a number of
points along the north shore which make any association with the Isle Royale sources
equivocal.

The Cobinosh Island site is located on the north shore of that island at the
eastern end of the Nipigon Bay archipelago in Lake Superior. According to David
Arthurs (personal communication, 1990), the site is stratified and includes copper
artifacts (mostly awls) with Heins Creek, Madison, Mackinac, Juntunen, Blackduck,
Selkirk(?), and Iroquoian ceramics. In one test excavation there is a direct association
of Blackduck ceramics and a copper awl.

The Michipicoten site at the mouth of the Michipicoten River on Lake
Superior contains stratified deposits ranging from AD. 1100 to AD. 1700 (J .V.
Wright 1968). One copper awl was found in Stratum VII (ca. A.D. 100-1400) with
Juntunen and Peninsular Woodland ceramics. Stratum III (AD. 1460) contained 9
awls, 1 knife, 1 punch, and 17 pieces of waste copper with Huron, Juntunen, and

Peninsular Woodland ceramics. Stratum 11 (ca. AD. 1700) contained 2 rings, 1 awl,

170

and 7 pieces of waste copper in association with Huron, Juntunen, and Peninsular
Woodland ceramics.

The Whitefish Island site at Sault Ste. Marie produced an assemblage which
duplicates the association of copper artifacts and ceramic wares seen at Michipicoten
and at several Isle Royale sites. Conway (1977) describes 10 awls, 4 bars, 2 earrings,
3 beads, 1 ring, and 15 pieces of waste copper in association with Huron, Mackinac,
Bois Blane, Juntunen, Blackduck, and Algoma style ceramics. ("Algoma" ceramics
represent Conway’s interpretation of a local Algonkian ceramic tradition. It has not
been formally published or described, however.) Based on what is known from Isle
Royale, this assemblage of worked and waste copper is what one would expect from
a site in close proximity to a primary copper source, in this case, probably the Point
Mamainse and Point Gargantua sources.

STRAITS OF MACKINAC: As both McPherron (1967) and Bastian (1963a
and 1963b) noted, the copper assemblage at the Juntunen site is one of the largest
east of Isle Royale. A total of 776 copper artifacts from 133 provenience units are
described by McPherron (1963:164-175), and in form and inferred technology
recapitulate the findings on Isle Royale. The large numbers of copper artifacts must,
in part, reflect the intensity of excavation of the Juntunen site as compared to the
more limited testing at most of the other sites considered here. However,
McPherron’s hypothesis of a resurgence of a local copper industry (the first since the
Archaic) based on the stratigraphic distribution of copper at the Juntunen site

appears to be borne out by the paucity of Mackinac phase ceramics elsewhere in the

171

Lake Superior copper district coupled with the widespread occurrence of Juntunen
ceramics in this area.

LOWER MICHIOAN AND SO OTHWESTERN ONTARIO: In the Lower
Peninsula of Michigan the problem of different sources of Lake Superior copper
becomes irrelevant, since most of it would have had to funnel through the Straits of
Mackinac regardless of geological source. Here, too, the context of copper takes on
a different value with increased distance where it shifts from secular utilitarian forms
and contexts into increasingly ritual contexts as burial goods or items of personal
adornment (usually beads). In northern Lower Michigan at the Pine River Channel
site on Lake Michigan two beads, one tool fragment, and four pieces of waste copper
(?) were found (Holman 1978). Nearby at the O’Neill site one rolled tubular bead,
one butterknife, and three pieces of waste copper were found with Mackinac,
Juntunen, Skegemog, and Traverse wares (Lovis 1973).

On the western side of the Lower Peninsula at the Dumaw Creek site
Quimby (1966) describes a late prehistoric (ca. A.D. 1605-1620) burial component.
Copper artifacts are numerous and include hair pipes (rolled tubular beads), beads,
tinkling cones, and a snake effigy. These in combination with the many other items
of personal adornment and utilitarian function included as grave furniture suggest an
ascribed "value," but one which ranks along side of stone projectile points, ceramic
vessels, and marine Shell beads as a reflection of social prestige of the interred
individual and not necessarily an intrinsic value of the material itself.

Far to the south at the Riviere Au Vase site in Macomb County, Michigan,

one awl and 34 beads were found with a burial (Fitting 1965). Copper was also a

172

common constituent in the Wayne Mortuary Complex of the Saginaw Valley and
southeastern Michigan (Halsey 1976), although the relations to Middle Woodland
cultures renders this marginal, both culturally and geographically to this consideration
of Late or Terminal Woodland copper.

Copper does not appear to have been an important component of sites in
Huronia in spite of evidence of long distance contact between this area and the Lake
Superior region. At the Nodwell site on Lake Huron awls were found in a
longhouse, midden, and pit contexts (J .V. Wright 1974). One awl was found at the
tenth century AD. Iroquoian Boys site, north of Lake Ontario (Reid 1975). And,
in general comments concerning the use of native copper in Glen Meyer sites in
southwestern Ontario, Noble (1975:48) states that, "A second non-indigenous
commodity, native copper, also appears in Glen Meyer territory, but again not in
marked volume. Wright (1966:39) reports one rolled bead from the Stafford site,
while Fox (1972:23) recovered two beads from DeWaele. Reputedly, a native copper
awl came from Van Beisen."

UPPER PENINSOLA OF MICHIOAN AND NORTHERN WISCONSIN:
In the area of the Upper Peninsula and northern Wisconsin, the problem of Isle
Royale versus other sources of copper is a moot point. In practical terms, one must
accept that virtually all of the archaeological copper encountered here is from sources
on the south side of Lake Superior. While there are a few well documented sites, the
area is, overall, poorly known.

In the protohistoric component at the Summer Island site in northern Lake

Michigan there are 3 rolled beads, 4 conical points, 3 awls, and one snake effigy (?)

173

(Brose 1970:211) in association with Lake Winnebago Trailed, Huron Incised, Bell
Type II, and Dumaw Creek ceramics.

On the Keweenaw Peninsula in the Upper Peninsula’s copper range and to the
southwest toward the Wisconsin border archaeological copper takes on expected
similarities of form and context comparable to that found on Isle Royale and north
shore sites. At the Sand Point site at the southern end of Keweenaw Bay the copper
assemblage includes waste, beads, crescents, awls, conical points, and fishhooks with
a total of 130 copper artifacts (Hoxie 1980). The site is dated between A.D.
1100-1400 and is associated with a ceramic complex related to the Lakes phase of
northern Wisconsin with stylistic ties to the Juntunen phase (Claire McHale Milner,
personal communication 1989). The site includes eleven burial mounds and the
remains of a minimum of 117 individuals was recovered from site testing (Cremin
1980). The context of the copper artifacts is nowhere specified: one gets the
impression that copper is found throughout the site with other domestic refuse.

At a site near the tip of the Keweenaw Peninsula, the Montreal River site,
copper waste, one awl (?), and one bar (?) fragment were surface collected in an
assemblage which is dominated by Mero phase grit tempered Oneota ceramics and
also includes Sand Point and J untunen ware. The artifacts from this site reside in the
Archaeology Laboratory, Michigan Technological University, Houghton, and in the
private collection of John Williams, Arcadia, Michigan, and were examined by the
author.

Buckmaster’s (1979) settlement analysis of the Menominee River watershed

on the Michigan-Wisconsin border also recapitulates the forms and context of

174

copper on the north shore of Lake Superior, although there are differences in some
of the artifact forms and, of course, in ceramic associations. Knives, awls, beads, and
waste copper predominate and are found with Heins Creek, Pt. Sauble, Madison,
Grand River, Carcajou, and Lakes phase ceramics.

In the interior of the Upper Peninsula of Michigan United States Forest
Service surveys are the primary source of archaeological site data. A cursory
examination of the 1985 and 1986 annual reports from the Hiawatha National Forest
cultural resource surveys produced no examples of prehistoric copper (Franzen 1987,
Gilbert/Commonwealth 1987). No data were obtained for the Ottawa National
Forest, although in the northern part of the Forest copper mines are known but not
formally reported.

Out of a total of eight Terminal Woodland sites in the St. Croix river drainage
on the Wisconsin-Minnesota border two sites have copper artifacts (Perry 1986a,
1986b). One includes six pieces of "raw and worked copper" nominally indicative of
fabrication activities. These are lower than expected densities for an area in which

native copper is known to occur in primary contexts.

EXAMPLES OF OTHER UNIQUE EXOTIC RAW MATERIALS

Copper was not the only material to capture the fancy of prehistoric
craftsmen, or the only substance to be carried long distances across cultural
boundaries. Unique raw materials other than copper which occur on archaeological
sites may offer some insight with respect to inferring the technology and organization

required to extract, fabricate, and distribute the products of that material, and the

175

nonmaterial cognates of its use. To the extent that it will explicate some of the more
general aspects of the role of unique raw materials, a brief discussion of some
analogous archaeological and ethnographic examples is in order.

Holmes (1919) descrrbes a number of quarries, among which is the Flint Ridge
chert quarry in Ohio, as well as sources of obsidian, mica, steatite, hematite, and
turquoise. To generalize, in all cases there is a basic pattern in the structure of
activities and activity areas relative to the quarry pits and mines. Workshops are
close to the primary source where extraction and initial reduction anticipatory to
fabrication take place. Intermediate and final stages of working tend to occur at or
near the living sites. Holmes, does not consider the social implications of these
activities.

Pipestone or catlinite was quarried from shallow surface pits at its source in
Minnesota, and has been of interest to explorers, missionaries, and archaeologists
since the early seventeenth century. It found mention in the writings of the Jesuits
in 1637, Groseilliers and Radisson in 1658-60, Marquette in 1673, Le Sueur in
1700-02, Lewis and Clark in 1804, and Catlin in 1837 (Beaubien 1957). Only Catlin
supplies an eye-witness account, however.

Holmes (1919:253-264) visited the quarry where he found shallow pits,
hammerstones (grooved and ungroovcd), with nearby occupation sites where
fabrication took place. The tradition of the quarry as a neutral ground where all
intergroup hostilities were suspended was already well established in the oral history
surrounding the pipestone quarry. In the mid eighteenth century Jonathan Carver,

who did not himself visit the quarry, left this account:

176
On the plains between the river St. Piere and Missouri is a large
mountain of red marble where all the neighboring nations resort for

stone to make pipe of. Even those who hold perpetual wars in all

other parts meet here in peace. The pipe being a symbol of peace, it

shews the prevalence of custom when people of so cruel and

implacable a disposition can so far bridle their impetuosity as to be

diverted from revenge by any means. [Parker 1976:138-139]

Beaubien (1957) questions the veracity of the notion of neutral ground. But,
unfortunately, there are no data of an archaeological or ethnohistorical nature to
confirm or refute it. Further, there are alternative sources for pipestone elsewhere
in the Midwest to complicate the situation. However, the physical layout of the
quarry, and the associated domestic and fabrication debris is Similar to the Isle
Royale copper mines on formal grounds.

Catlinite, unlike copper, cannot serve basic technological, subsistence related
needs: its domain is strictly socio-religious or, at the very least, ornamental. The
importance of the pipestone calumet is amply documented (e.g., Swanton 1911), but
in no instance is there an account of pipestone serving as a tool in the same sense
as copper or other harder stone. In this regard the analogy between it and copper
is imperfect.

Galena is known from prehistoric archaeological sites in the Midwest. Like
pipestone, its use in prehistory was for nonutilitarian purposes. Galena, a source of
white pigment or occasionally worked into effigy forms, occurs on both
ceremonial/mortuary and occupation sites (Walthall 1981:3ff). There is historic

documentation of mining which took place in the seventeenth through the early

nineteenth centuries when it was used as a source of lead for Shot and ornaments

177

(Walthall 1981:18-25). The historic extraction and use of galena are strongly
influenced by European technology. However, Schoolcraft’s (1821:345-346, quoted
in Walthall 1981:19) observations of a galena mine which he observed in 1820 are
instructive with respect to the use of labor and mining techniques:

The lead ore at these mines is now exclusively dug by the Fox

Indians, and, as is usual among savage tribes, the chief labour devolves

upon the women. The old and superannuated men also partake in

these labours, but the warriors and young men hold themselves above

it. They employ the hoe, shovel, pickaxe, and crow-bar, in taking up

the ore. These things are supplied by the traders but no shafts are

sunk, not even of the simplest kind....They run drifts into the hills so

far as they can conveniently go, without the use of gunpowder and if

a trench caves in, it is abandoned. They always dig down at such an

angle that they can walk in and out of the pits, and I descended into

one of these which had probably been carried down forty feet. All this,

is the work of the Indian women and old men, who discover a degree

of perseverance and industry, which is deserving of the highest

commendation.

Walthall’s analysis identifies several chemically discrete sources of galena in
the Midwest and middle South, and traces the exploitation and archaeological
distribution of galena through prehistory. The issue of differential access to the
resource is not addressed, but at a broad level his conclusions support the decline of
widespread interaction networks with increasing emphasis on local sources (1981:44).
His inferential linkage of galena and Great Lakes copper is stated in terms of
"abundance" of the latter resource, which he feels declined in late prehistory.

The question of social boundaries is raised in the study of the distribution of

greenstone axes in southeastern Australia (Gould 1980:206-212). The physical

characteristics of the quarries is essentially the same as the other examples cited

178

earlier, with shallow pits and concentrations of debris and blank production near the
quarry. But the interpretation of the archaeological distribution of the final products
across the landscape is of interest. Citing a study by McBryde, Gould (1980)
describes a nonuniform distribution of axes from the point of geological origin to as
far away as 500 km. The distribution lacked a simple fall-off profile (Renfrew 1977),
and in one case increased in frequency at the most distant point of the survey. The
distribution suggested that axes followed major avenues of travel along river valleys
while avoiding open plains. Ethnohistoric evidence showed that the distrrbution of
these axes corresponded to the historic distributions of two competing groups with
traditionally antagonistic relations. The lack of knowledge of competing sources for
other axes, and a real uncertainty of the chronological dimensions of the axe sample
used by McBryde limit the conclusions of her study: for example, it is not known to
what extent the archaeological distrrbution of greenstone axes can be attributed to
the ethnohistoric groups referred to in the study. Gould (1980) is quick to point out
that discontinuities in material distributions do not necessarily signal a cultural

boundary.

VALUE

The archaeological determination of value of an object, substance, or
commodity often presupposes an environment which includes a market economy, or
an established system of trade/exchange equivalencies that mediate the transfer of
goods among groups (e.g., Earle and Ericson 1977). Value is also inferred from the

presumed function of the object, substance, or commodity relative to a continuum

179

from utilitarian to nonutilitarian domain (Binford 1962). The distance a given item
or material has travelled is used as an indication of its value (e.g., Brose 1990). The
context of an item’s ultimate entry into the archaeological record, such as grave
furniture or general domestic refuse, also suggests the value invested in that item or
material (Binford 1962).

According to Sahlins (1972:277), "The diverse values put on things depend
specifically on barriers to their interchange, on the inconvertibility of goods from
different spheres; and as for the transactions (’conveyances’) within any one sphere,
no determinants of the rates have yet been specified." Within the Lake Superior
basin, the question of value of copper is translatable into a question of access to
copper sources, rather than in the area of exchange or trade. By extension, the
question of access is translatable into the question of territory and group boundaries
and, more fundamentally, the area of intergroup interaction. Traditional definitions
of value become largely irrelevant when one concedes that copper resources are
widespread (if nonuniform) throughout the basin, and only become important when
the resource is considered outside the region. In the context of direct acquisition
of copper or exchange between nearest neighbors there are no fixed values ascrrbed:
"a ’reciprocity’ that comprehends precise material rates is rarely encountered. The
characteristic fact of primitive exchange is indeterminacy of rates" (Sahlins 1972:278).
If, as Sahlins (1972:279) asserts, the rules of exchange are modelled after the social
and moral spheres of social organization, archaeological data indicative of boundary

permeability, or the absence of boundaries, must be interpreted as evidence of equal

180

access to copper resources by all groups proximate to the copper district of the Lake
Superior basin.

Similarly, mathematical models which describe the distribution of commodities
or materials across a cultural landscape are inapplicable within the area of the
source(s) of that material. Down-the-line exchange, directional trade, and
concentration effect models (Renfrew 1977) only make sense if one maps the
distribution of the material in a regional universe that precludes direct acquisition.
Monotonic decrement, the decline in frequency with increased distance from the
source, is applicable in the case of copper and lithic raw materials with discrete
spatial sources. However, the problem is complicated by the lack of ability to
discriminate among the different copper sources within the region, making suspect
any quantitative approach which, for example, measures the density and distrrbution
of copper from one end of the Lake Superior basin to the other.

In contrast to earlier cultural manifestations in Archaic or Middle Woodland
cultures where it occurs as a frequent inclusion as grave furniture or as an exotic in
ritual settings, the role of copper in the Terminal Woodland appears to have had less
value, status, or importance. This determination is based on the archaeological
context of copper which, in the Lake Superior region, is found along side of such
domestic refuse as ceramics, lithic waste, and food remains. The fact that the
regional setting of this analysis is coterminous with the range of native copper
deposits may explain a large part of the secular nature of copper during this period
of prehistory: the material was, in effect, on a par with chert quarries or good

sources of clay. Secondly, the objects from which copper is made fall into utilitarian

181

and ornamental categories, but do not rank any higher or lower than lithics or
ceramics as status articles when the archaeological context in which they are found
is considered. All of the utilitarian items, awls, gaffs, fishhooks, small knives, can be
duplicated on other materials. Perhaps only copper beads are unique products
unavailable in another medium: copper artifacts are unique only by virtue of their
being copper. Therefore, while copper served a number of technological functions
related to subsistence and fabrication needs, it was not essential or critical to
survival.

It is argued that the social organization required to extract, fabricate, and
ultimately redistribute copper items did not exceed that which was required to pursue
other forms of subsistence or technological necessities. Those mines which can be
associated with the Terminal Woodland substage, and which have evidence of related
occupation and fabrication activities, suggest that the effective group required for
extraction and fabrication of copper did not exceed one or two families or the
equivalent number of individuals. There are no data to indicate that, in the Terminal
Woodland substage, there were special groups whose sole purpose was to go to Isle
Royale to extract copper for a trade with neighboring or distant archaeological
cultures. Rather, the extraction and fabrication of copper seems to have been one
of many activities undertaken in concert with the pursuit of subsistence needs and
probably did not require any special organization or planning outside of that which
was required for other activities.

Specific associations between copper mines and the archaeological cultures

engaged in mining activities are generally wanting. Argument by geographic

182

proximity in which the distribution of an archaeological culture and primary sources
of copper are contrasted indicates that all groups had access to copper in varying
degrees. The people of the Juntunen phase had access to virtually all of the Lake
Superior sources from Isle Royale to Point Mamainse, and probably some access to
the Keweenaw sources as well. The extent of Blackduck and Selkirk pottery suggests
that their producers had access to Isle Royale and north shore sources at least as far
as the Michipicoten River. The Wanikan culture, while only marginally represented
at Isle Royale probably had access to some north shore copper sources, those in the
St. Croix drainage, and may have utilized drift sources in Minnesota and Wisconsin
as well. Oneota and Lakes phase groups clearly did not use Isle Royale or north
shore copper sources: their needs were more than met by south shore (Keweenaw
and Ontonagon) and St. Croix copper deposits.

Among those groups with the highest frequency of occurrence on Isle Royale,
Blackduck, Selkirk from the west, and Juntunen and Huron from the east, we may
conclude that all had some access to primary sources of copper. The apparent
coincidence of the distribution of the J untunen phase and most of the primary copper
sources may inflate the seemingly greater incidence of copper on Juntunen sites, or
it may reflect a differential interest in copper by these people. Assuming roughly
equivalent geographical access to copper among these groups, the regional
distribution of copper suggests, on the contrary, that differences exist which may
indicate differential access to or interest in copper which may have been conditioned
by factors of intergroup interaction. This conclusion will be further explored in

Chapter 8.

CHAPTER 7
THE MOVEMENT OF POTS AND PO'ITERS IN THE

TERMINAL WOODLAND SUBSTAGE: TRACE ELEMENT ANALYSIS.

INTRODUCTION

This chapter presents the results of the neutron activation analysis of 95
archaeological and five geological clay samples from sites within and near the Lake
Superior basin. It was suggested in the previous chapter that, although most
archaeological cultures were in a geographical position to exploit copper resources
on Isle Royale, the distribution of copper on a regional basis indicate that
sociocultural factors may be responsrble for the differential access to and use of
copper. Explanation of this requires some confirmation of the distribution of
archaeological cultures, the extent to which the presence of their diagnostic ceramics
on Isle Royale reflects sustained use of the island or merely sporadic visits.
Hypotheses pertaining to the potential interaction of groups manufacturing
Blackduck, Juntunen, Huron, Oneota, and Lakes phase style ceramics are addressed.
Results identify geographic zones of ceramic production which may correspond to the
distribution of Terminal Woodland archaeological cultures.

This study has created a body of reference data for the use of neutron
activation analysis (NAA) as a means of enhancing the use of ceramic data in
addressing fundamental definitions of group membership and mobility in the Great
Lakes region. While it is not anticipated that the results of NAA will lead directly

to conclusions about ethnicity manifest in ceramic style, information concerning the

183

184

relationships among clay sources used in the production of a variety of types which,
in turn, are attributable to discrete archaeological cultures, contribute to the
understanding of interaction in the context of the Terminal Woodland substage.

In the Lake Superior region we are dealing not with craft specialists in
relatively well circumscribed cultural and geographical units. Instead, the Great
Lakes potters, and particularly those in the nonhorticultural Algonkian-speaking
groups, are characterized as persons who, through a variety of agencies, moved across
permeable boundaries in a seasonal pattern of subsistence and/or socially motivated
transhumance. The affects of mobility and boundary permeability on the distribution
of ceramic vessels, both in terms of style and clay source, can be tested using NAA
but will likely result in a different and more diffuse patterning than is seen in

residentially stable societies.

ASSUMPTIONS

Trace element analysis is not unknown in the Great Lakes and surrounding
areas where it has been applied by geologists and archaeologists on clays, lithics, and
copper (e.g., Brizinski and Buchannan 1977, Brumbach 1975, Brumbach and Bender
1986, Julig, Pavlish, and Hancock 1988, Kuhn 1986, Luedtke 1976, Nussman 1965,
Ramsden 1988, Rapp 1984, Trigger et al. 1984). The fundamental underlying
assumption of trace element studies is that the data are structured and may be
explained by geological and/or cultural factors. This type of analysis has as its
weakest point the explanation of the patterns which ensue from the statistical

manipulation of the data. Sample sizes are rarely large enough or lack contextual

185

and chronological controls sufficient for definitive conclusions. Furthermore, the
various applications of trace element analysis done within a region have not been
pooled in order to maximize the potential for a more broad consideration of these
types of data on a panregional level. Nonetheless, trace element studies have made
possible the evaluation of the epistemology of ceramic types (Syms 1977) and the
interpretation of style with respect to the composition of prehistoric archaeological
cultures.

Beyond the assumption of underlying structure are other assumptions
regarding the systemic context of ceramic production and use in the Upper Great
Lakes region. Pots can be made at any time of the year, but the period from spring
through fall is most likely, Since the acquisition and processing of clay is easier during
the warm months (Syms 1977:63). Pots as curated items are more likely to be made
at intervals of greater residential duration as a function of practical limits of mobility,
however, pots as expedient items could be produced under almost any conditions.
The greatest residential stability occurs at spring and early summer fishing sites, early
spring sugaring camps, or fall ricing camps. It is under these conditions that factors
operate on the use of style during ceramic decoration, and have the highest potential
for transfer of vessels among groups. It is assumed that style encodes some level of
corporate membership, and that the comingling or discreteness of style will reflect a
level of interaction or integration among and within groups (see Chapter 3). Further,
it is assumed that most pottery was made by women and that the conceptual and

procedural steps involved in ceramic production were transmitted through the female

186
line (Syms 1977:59). Finally, it is assumed, following Syms (1977:43), that for the

most part, ceramics were not articles of trade in this region.

CONDITIONS AND IMPLICATIONS OF COVARIATION
The conditions and implications of covariation anticipated in the application
of NAA to the sample include the following:
1) Conditions: Pots from area A are consistently made of geologically
related clays from the same area.

Implications: There is a degree of sedentism or, at least, a correlation
between where a pot is made and where it is used and
discarded.

Archaeological Expectations: A local ceramic tradition, interpreted as a
resident population, may be defined on the basis of a
localized distribution of pots on related clays.

@mment: Within-style-group homogeneity in trace elements suggests that the
same (geologically related) source was used by potters from a related archaeological
culture and implies spatial integrity of an archaeological culture with minimal
production, use, or discard of ceramics outside of the area.

2) @nditions: A pot from style group A made of clay associated with

area A is found in area B.
Implications: The finished pot has been physically transported from

area A to area B. Alternatively, raw clay may have been

187

transported from area A to area B where it was formed,
fired, and discarded.

Arphaeologjcal E_xpectations: Ceramics identified as exotic, foreign, or
nonlocal are present on a site or in a region.

@mment: Between-style-group heterogeneity in trace elements suggests that
different (geologically unrelated) clay sources were used by potters from unrelated
archaeological cultures. As a corollary of the first stated condition, the movement
of finished pots from the site of manufacture to the site of discard is implied.

3) @nditions: A pot of style group A is found in area B and is made

of clay associated with area B.

Implications: Three interpretations are possrble. The potter has
moved from area A to area B where the pot was
manufactured on local clays; a potter from area B has
copied a style associated with area A; and style and
source areas for clays overlap used by two groups.

Arghaeological apectations: Different recipes for ceramic production
may be used: e.g., the pot that is stylistically exotic may
be thicker and more heavily tempered than stylistically
equivalent ceramics actually from the area of presumed
cultural origin.

Cgmmpnt: Within-style-group heterogeneity in trace elements may correspond
to technical or stylistic variables, such as the thick-walled, heavily tempered vessel

made on local clays or the thin-walled temper-poor vessel made on nonlocal clays,

188

or may relate to the mobility of the potter from a familiar clay source to one which

is less familiar.

4)

Opnditions:

Implications:

Pot styles A and B are manufactured on the same
clay(s) and are found in both areas A and B.

Potters from both groups share access to the same or
geologically similar clay sources, and there is extensive
movement of pots and potters between areas, and/or a
single group is producing two stylistically distinct

products.

Archaeological Emtations: Ceramics from both style groups will be

frequently associated on sites. If two groups are
represented there should be some proportionate
difference between the two groups in certain areas of
their range, depending on their seasonal pattern of
movements. Alternatively, the composition of the group
responsrble for the two ceramic styles includes members
from two distinct ceramic traditions and is independent

of factors such as seasonal movement.

@mment: Between-style-group homogeneity in trace elements suggests that

the same (geologically related) clay sources were used by potters from unrelated

archaeological cultures, or that potters manifesting discrete stylistic traditions were

members of a common group.

189

5) Conditions: Clays are selected from a wide variety of sources which
are geologically unrelated; ceramic production takes
place under conditions of mobility.

Implications: There is insufficient homogeneity in the clay sources
selected for ceramic production for patterning of trace
elements.

Archaeological Expectations: Differences at this level may not be
discemable without trace element analysis. Under these
conditions, trace elements from a given style group or
region will fail to cluster.

@mments: Within style group heterogeneity in trace elements suggests that
several (geologically unrelated) clay sources were used by potters from related
archaeological cultures, or that similar styles were employed by unrelated
archaeological cultures using different clay sources. The mobility of potters and the

production of vessels at several loci is implied.

THE SAMPLE

NAA is applied to ceramic artifacts and geological clay samples to determine
the statistical relationships of composite materials among and within stylistically
coherent groups. Geological clay samples were obtained from five localities: three
on Isle Royale, one from the north shore of Lake Superior at Grand Portage
National Monument, and one from the Apostle Islands in northern Wisconsin (Figure

17). The archaeological samples were selected on the basis of stylistic, geographical

190

SEES <<z 86828262 28 .8380 E 8283 .2 2:5

 

 

           

man. E<m ..<0.001_Ow<10m< O

mga2<m ._<0.00._0m0 O

mmtmuamww

 

a...
5.80 - .
$3.3 9.

9530 a
£23 ‘

 

 

191

and temporal qualities. A minimum number of ten vessels per ware group was
initially sought and, in most cases, attained. However, problems in provenience,
identification, or with the potential destruction of small sherds account for the actual
sample. The sample was drawn to address specific questions of relationships

described in the hypotheses below. A catalog of NAA samples is provided in

Appendix D.

CULTURAL AND CHRONOLOGICAL DIMENSIONS OF THE SAMPLE
Terminal Woodland substage sites dating ca. A.D. 700-1600 in the Upper
Great Lakes and southern boreal forest typically contain a number of stylistically
diverse ceramic ware groups. Common among them are Blackduck, Sandy Lake, and
Selkirk from northwest Ontario and northern Minnesota, the Straits of Mackinac
sequence, the Lakes phase from northern Wisconsin, and Ontario Iroquois Tradition
ceramics from the Lake Huron basin. Ware groups utilized in this analysis include
Blackduck, Mackinac, Juntunen, Huron, Sand Point, and Mero phase Oneota. While
no less important, samples of Selkirk and Sandy Lake were not included.
BLAOKDOOK: Blackduck has both a broad geographical and temporal
distribution in the Lake Superior basin and in the lakes district to the northwest
where they become common after AD. 800 and persist until contact. Blackduck
samples are drawn from three areas: eight from Isle Royale National Park, twelve
from an assortment of sites in Ontario on Lake Superior’s north shore, and ten from

Voyageurs National Park in northern Minnesota.

192
STRAITS OF MACKINAO: The Straits of Mackinac sequence (McPherron

1967) includes the Mackinac phase (A.D. 700-1000), Bois Blane phase (A.D.
1000-1250), and Juntunen phase (A.D. 1250-1450). The Juntunen site has been an
important datum for late prehistoric ceramic analysis in the Upper Great Lakes, and
it is important to include it in a comparison with western Lake Superior basin
ceramics. The J untunen site (20MQl) provided ten specimens of Mackinac ware and
ten of J untunen ware. Isle Royale contributed ten samples of J untunen style pottery,
while sites in northwestern Ontario and the Montreal River site on the Keweenaw
Peninsula in northern Michigan contributed two each.

HORON: Huron ceramics are believed to be a late (post AD. 1450) arrival
in the Superior basin, however, ceramics with stylistic affinities to the Ontario
Iroquois Tradition are widespread throughout the Terminal Woodland substage. Ten
samples of Huron style ceramics were submitted from sites on Isle Royale.

SAND POINTZflS PHASE: The Sand Point phase (Dorothy 1980) was
defined for ceramics from one of the few excavated sites on the Keweenaw Peninsula
in Upper Michigan and is believed to belong to the latter part (ca. A.D. 1100-1400)
of the Lakes phase defined for northern Wisconsin by Salzer (1974). A sample of
ten vessels was obtained from the Sand Point site, as well as a single specimen from
the Montreal River site.

ONEOTA: Grit tempered Mero phase Oneota pottery from the Upper
Peninsula of Michigan was included to provide a contrastive data set. The Sand
Point and Montreal River sites from which the samples were obtained are located in

Michigan’s mainland copper district and are included here to evaluate the possibility

193

of regular travel across Lake Superior between the Keweenaw and Isle Royale
suggested by Raddison in 1659 and subsequently reiterated by a number of
archaeologists (e.g., Dustin 1957, Griffin 1961). Oneota is at best only nominally
represented on archaeological sites on the northern side of the Superior basin and
is virtually unknown from sites on Isle Royale. The Montreal River site is near the
tip of the Keweenaw Peninsula and also in the mainland copper district of Upper

Michigan, provided ten samples of Oneota pottery.

GEOGRAPHICAL DIMENSIONS OF THE SAMPLE

The sources of the archaeological samples used in this analysis was largely
directed by the availability of materials as much as by the integrity of the sample. In
the case of the samples from Voyageurs National Park, the Straits of Mackinac, and
the Keweenaw Peninsula, samples were drawn from only a few or single sites.
Artifacts from Isle Royale were obtained from the entire island, representing sixteen
different sites. The Ontario samples are from twelve sites from across the north
shore of Lake Superior, some from submerged contexts (Figure 17). Specific data
for each sample submitted for NAA are provided in Appendix C. A general

breakdown of the sample is presented in Table 4.

194
Table 4. NAA Sample Locations.

8 Blackduck ware, Isle Royale National Park
12 Blackduck ware, Northwest Ontario
10 Blackduck ware, Voyageurs National Park
10 Mackinac ware, Straits of Mackinac
10 Juntunen ware, Straits of Mackinac
10 Juntunen ware, Isle Royale National Park
2 Juntunen ware, Northwest Ontario
2 Juntunen ware, Montreal River site, Keweenaw Peninsula
10 Huron, Isle Royale National Park
10 Sand Point ware, Sand Point site, Keweenaw Peninsula
1 Sand Point ware, Montreal River site, Keweenaw Peninsula
10 Oneota, Montreal River site, Keweenaw Peninsula
3 clay samples, Isle Royale National Park
1 clay sample, Grand Portage National Monument
1 clay sample, Apostle Islands National Lakeshore

RESULTS

The processing of the samples and statistical manipulation of the resulting data
were performed by Michael Glascock and Hector Neff of the University of Missouri
Research Reactor Facility, Columbia, Missouri (Appendix C; Glascock and Neff
1990). Cluster analysis utilizing the elements LA, LU, SM, YB, CE, EU, FE, RB,
SB, SC, TB, TH, BA, DY, MN, TI, and V, suggested divisions along the east-west
axis. Principal component analysis further refined these groups, the first four of
which account for 78% of the total variance. Bivariate plots of principal components
(Figures 18-19, Appendix C) and selected elemental (SC/FE and EU/LA)
concentrations (Figures 20-21, Appendix C) illustrate the results.

Five reference groups resulted which are broadly associated with zones of clay

procurement. These are identified by geographical designations: Isle Royale,

Thunder BayNoyageurs, Straits of Mackinac, Sand Point and, Montreal River. The

195
Isle Royale, Thunder BayNoyageurs, and Straits of Mackinac groups are strongly

correlated, indicating their common membership in a similar geochemical regime
extending from northern Minnesota to northern Lake Huron across the north shore
of Lake Superior. Straits of Mackinac samples were distinguished from the Thunder
Bay/Voyageurs group, but the latter could not be further broken down, indicating

either a closely related geochemical system or widespread movement of pots between
these areas. Although mutually distinct at certain levels of analysis, the Keweenaw
Peninsula samples from the Montreal River and Sand Point sites constitute a zone(s)

of clay procurement unrelated to any of the other samples.

HYPOTHESES

Specific hypotheses pertaining to the relations of trace elements in the ceramic
sample are derived from the cultural-historical data and the implied or hypothesized
relationships among the sites/regions and archaeological cultures considered. These

are presented with the interpretations of the NAA:

__1: Mackinac ware is distinct from J untunen ware from the Straits of
Mackinac. The sociocultural changes hypothesized by MacPherron (1967)
regarding a fundamental shift in interaction orientation from the
Algonkian groups to the west (Wisconsin) to the northern Iroquoian in
the east and southeast involved a change in the use of clay sources.

NAA results suggest that clay sources used by Mackinac and Juntunen phase

potters remained the same in spite of possible shifts in cultural interaction spheres.

Ceramics were produced from clays obtained from within an area habitually utilized

196

by Mackinac and Juntunen phase potters. Selection of clays used for pots discarded
at the Juntunen site in both phases remained the same in spite of changes in stylistic

elements of the industry.

H2: Blackduck samples from Voyageurs National Park will form a discrete
cluster apart from other samples of Blackduck and from other styles of
pottery. This sample exhibits homogeneity in paste qualities but variety
in decorative elements. All but one sample are drawn from a single site,
also increasing the potential for homogeneity in source derivation.

NAA identified a cluster of Voyageurs and Thunder Bay samples with

additional geochemical linkage with clays from the Straits of Mackinac making a clear
chemical distinction between them impossible. A cultural explanation involves the
widespread movement of ceramics between the Voyageurs and north shore areas; a

geological explanation points to related geochemical sources.

H3: Ontario Iroquois Tradition ceramics will form a discrete cluster. Huron
style ceramics are included in the analysis to test the proposition that
what is stylistically Huron in Lake Superior is derived from Algonkian
groups who either copied the style or obtained them through trade or
exogamy.

In their analysis Glascock and Neff (1990:6-7) note that the Huron sample

from Isle Royale, while belonging to a related ceramic source zone, exhibit
geochemical differences explainable by the use of contrasting manufacturing

techniques which call for greater amounts of temper in the clay body. It has been

197

suggested that potters unfamiliar with a clay source will increase the amount of
aplastic to diminish the possibility of firing failure (David Arthurs, personal
communication 1988). If this is the case, it demonstrates a technological distinction
suggestive of the actual presence of Huron potters on Isle Royale instead of the
mimicry of Huron styles by Algonkian potters. Not all of the Huron style ceramics
fall into the Isle Royale cluster and may represent ceramics transported from areas

outside the north shore-Straits of Mackinac geochemical area (i.e., Huronia).

HA: Juntunen ware from Isle Royale will form a cluster distinct from
Juntunen ware from the Straits of Mackinac.

The NAA results suggest that the J untunen ware samples are the best evidence
of regional mobility (in terms of movement of vessels) with <1% probability of
membership in other groups sampled (Glascock and Neff 1990:7). Among the
Juntunen ware vessels from Isle Royale two belong to the Straits of Mackinac
geochemical cluster. Long distance movement of Juntunen pots is also evident in the
inclusion of one J untunen specimen from Ontario and three from the Keweenaw
Peninsula in the Straits of Mackinac geochemical cluster. One Juntunen sherd from
the Keweenaw Peninsula is grouped with Isle Royale, giving limited support for
travel across the lake. In concert with the evidence for widespread movement of
Juntunen style pottery is the strong indication that it was locally produced on Isle
Royale and at north shore sites in Ontario. In sum, the distribution of Juntunen

ceramics and the combined evidence for long distance movement of pots and local

198

manufacture are taken as an indication of the geographical extent of the Juntunen

phase encompassing all but the western end of the Lake Superior basin.

OS: The peoples of the Lakes phase and Mero phase did not travel across
Lake Superior to Isle Royale. Therefore, it is expected that Sand Point
and Oneota ceramics will form a discrete mutually inclusive cluster.
The NAA identifies a loose association between Sand Point and Oneota
ceramics, suggestive of a geochemically related source on the Keweenaw Peninsula
and one which is distinct from the both the Isle Royale and Thunder Bay/Voyageurs
reference groups. Supported by the lack of evidence of the movement of Oneota or
Sand Point ceramics from south to north, or of the movement of lithic raw materials
from north to south, the notion of regular traffic across Lake Superior is rejected.
It is apparent, and by no means surprising, that the conclusions indicate a
combination of geographical discreteness and intraregional movement of ceramics
and style. Returning to the conditions and implications of covariation of trace
elements, ceramic style, and location of discard of ceramics, one finds examples of
each possibility outlined there. Through NAA it has been determined that the local
production of ceramics, the movement of finished vessels (or raw clays) across the
region, and the use of more than one decorative style occurring on chemically related
clays are all operative factors contributing to the ostensrbly chaotic condition of the
archaeological record of the region.
If one accepts the premise that a seasonally resident population is, in part,

defined by the manufacture of stylistically discrete ceramics on locally obtained clays,

199

the range of the Juntunen phase must include the northern shore of Lake Superior
as far west as Thunder Bay and including Isle Royale. Here, the Juntunen phase
potters and their families interacted with the makers of Blackduck ceramics, although
the specific content of that interaction is unknown. Huron potters may have also
traveled with Juntunen groups where they mined copper, fished, and hunted with
Blackduck and Selkirk people from the west, bringing with them a distinctive set of
techniques for pottery manufacture.

The social setting of ceramic production, use, and discard in the Lake Superior
region places very high demands on the sample size and stylistic integrity of the
ceramics used in trace element analysis. The small number of samples used here
cannot provide more than an introduction to the potential of NAA for understanding
the cultural dynamics of interaction vis-a-vis ceramics in the Terminal Woodland
substage. Glascock and Neff (1990:8) recommend for further research a minimum
of 20-30 geological clay samples and 50-100 additional samples of Juntunen ware
ceramics to explore the question of movement of pots and potters. In addition to
their suggestions, other specific samples tailored to the hypotheses discussed above
could be stipulated. The question of Huron ceramics, for example, could be further
explicated by merging extant trace element data from previous studies (e.g., Brizinski
and Buchanan 1977, Ramsden 1988, Trigger et al. 1984) could be merged once
interlaboratory variation is controlled (see Bishop et al. 1990). This notwithstanding,
the initial application of NAA to this sample has resulted in the identification of
zones of clay procurement which have been linked to other data (lithic and ceramic

style) to suggest that there is some correspondence between them and the distribution

200

of prehistoric archaeological cultures. The implications of these findings for the

broader issue of intergroup interaction is discussed in the final chapter.

CHAPTER 8

CONCLUSIONS

In this analysis we have moved from an examination of interaction between
ethnographically and ethnohistorically documented groups to the sphere of
archaeological cultures. In the transformation between the two, we have noted that,
in the case of the former, the levels of specificity in context wherein interaction
occurs is a structurally relative one in which the operative parties have available to
them a wide variety of interaction alternatives which are conditioned by several
factors, and which may or may not independently reflect the general patterns of
interaction between two groups. In the archaeological case, interaction among
archaeological cultures must, by necessity of the more general nature of the data
base, be seen as a correspondingly generalized phenomenon which is not expected
to reflect the subtle variations of interaction between households and individuals, but
instead represents broad patterns of interaction among archaeological cultures on a
regional level. This analysis has not rendered archaeological cultures obsolete, but
recommends that their application to the area of interaction and group composition
may color the outcome. Boundaries are not of one type: a situation of passive
territoriality may have been inclusive with respect to the harvesting of subsistence
resources, but proprietary restrictions may have operated simultaneously to limit the
access of certain groups to copper. The comingling of a wide variety of ceramics in
the Upper Great Lakes is a fair representation of the ethnic composition of the

groups responsrble for leaving the archaeological record; what Cleland (1977 :93) has

201

202
referred to as "the cosmopolitan quality" of the Late Woodland sites throughout the
region. Schortman (1989) also advocates the need to depart from a spatially bounded
and environmentally based view of culture. As an alternative he suggests:
Salient social identities, ethnic or class based, are self-ascribed
cultural categories whose members share common assumptions, values,

and Standards for evaluating proper behavior. In order to recognize

salient affiliations archaeologically, therefore, emphasis must be placed

on the specification of consistent associations of materials that

functioned in those behavioral spheres which are most likely to reflect

the operation of these assumptions, values, and standards. [1989:57]

More than two decades earlier, J .V. Wright anticipated the difficulty
associated with discerning boundaries between groups who shared fundamentally
similar life-ways:

The broad mosaic of politically independent bands, loosely

related at the specific level through clan and/or marriage and, at a

more general level, through language and way of life, limits, in part,

the reality of discrete tnbal designations to taxonomic units of

anthropological convenience. [J .V. Wright 1965:90]

Clearly, archaeologists are faced with a dialectical dilemma; that is, the
juxtaposition between archaeological cultures which are by definition exclusive and
homogeneous, and the real actors on the prehistoric landscape which were inclusive
and heterogeneous. The search for Mason’s (1976) "site-unit ethnicity" or sites which

manifest the remains of a single historically identifiable group may, in fact be a

hollow victory, given the overall pattern of cultural diversity throughout the region.

203

The Terminal Woodland substage offers an incredrble diversity of ceramics,
interpreted here as the product of several archaeological cultures who utilized Isle
Royale’s resources independently and/or in concert with one another. Blackduck,
Selkirk, Juntunen, and Huron are the most prevalent manifestations occurring on the
island with additional but numerically minor examples of Sandy Lake. Lithic raw
materials suitable for tool production are not available on the island and were
imported from north shore sources where embarkation to the island likely originated.
Evidence for trans-lake travel is limited to isolated examples of lithics and ceramics
which, in aggregate, do not suggest that it formed a regular pattern of cultural
behavior. Copper cannot be used in this determination since it is widely available on
both sides of Lake Superior, and has not yet been chemically differentiated among
its various sources.

There is no evidence for trade or exchange with neighboring areas if one
discounts ceramics as a possible commodity. Even in the event that ceramics were
traded, the evidence is largely restricted to the J untunen phase. In this case, mobility
of Juntunen potters between the Straits of Mackinac and Isle Royale across the north
shore seems a more parsimonious explanation. The extent to which Huron potters
travelled with the Juntunen people remains uncertain. The association between the
two ceramics may reflect cooperation between them, or may be accidental, although
given McPherron’s (1967) comments regarding the basic elements of shared or

borrowed style and other dimensions of culture, cooperation is more likely.

204

That Blackduck and Selkirk occur together reflects an ongoing use of the
island by people from the west of Lake Superior, whether or not one ascribes to the
theory that they are coeval or that one supplants the other through time. Isle
Royale, while clearly within the range of Blackduck and Selkirk, is in an area where
both begin to diminish in occurrence as one moves east. Isle Royale was a peripheral
area in terms of Blackduck and Selkirk territory, periodically used for subsistence but
not apparently held in the same manner by them as by the Juntunen phase people.

Interaction among all groups represented on Isle Royale appears to have been
at least ambivalent, and at best, amicable. There is no substantive evidence for
violent interaction, or of exclusive boundaries which precluded access to the island
except in the case of south shore Oneota and Lakes phase groups. South shore
groups had no material need to cross the lake and would have done so only for social
or religious purposes: there is no evidence that they did. Thus, the strongest
exclusive boundary, and that a barrier, seen in the region is the physical presence of
Lake Superior.

The definition of archaeological cultures employed here utilizes an isomorphic
view of the relationship between style and raw materials, and a corresponding
archaeological culture. But as we have seen from the Isle Royale data,
archaeological sites usually include a wide variety of ceramic styles indicative of more
than one archaeological culture. Traditional explanations of this phenomenon
include exogamous marriage practices, raiding and warfare in which female potters

are captured, stylistic mimicry, trade, and exchange. All can be documented to

205
varying degrees for the Upper Great Lakes region, but none fully or independently

explain the archaeological association of so many different varieties of pottery on so
many sites in so large a region, nor do they exhaust the scope of potential
explanations of this phenomenon.

If heterogeneity in ceramic style is the rule and not the exception, several
questions must be addressed with respect to regional culture-history and more
broadly applicable principles.

1. IS the apparent association of diverse ceramics a function of a lack of
chronological control wherein these diverse ceramics represent unrelated episodes of
deposition?

Chronology, both on Isle Royale and in the region as a whole, is poorly
controlled with respect to the fine-tuning of ceramic traditions. Seriation of
Blackduck (Lugenbeal 1978), Selkirk (Rajnovich 1988), and Juntunen ceramics
(McHale Milner and O’Shea 1990) has begun to refine the formal evolution of
ceramic styles, but the fundamental problem of relating these parts to a panregional
whole remain. Clearly, the very general approach to chronology used in this analysis,
in which the entire Terminal Woodland substage is treated as a single entity greatly
oversimplifies the actual range of interactions which occurred in prehistory.
However, the requirements of the data to treat the historical realities in specific
terms cannot be met by the archaeological record now extant. A large portion of the
chronological uncertainty stems from the spotty documentation of archaeology along

the north shore of Lake Superior. Conway’s "Algoma" pottery (1977) and perhaps

206

other undocumented archaeological cultures between Lake Nipigon and the Sault
may have influenced the direction and scope of interaction in the region in both
spatial and temporal terms.

2. Is the situation of heterogeneity as it has been interpreted caused by an
inappropriate definition of archaeological cultures which effectively precludes other
possible permutations of group composition?

The isomorphic view of material culture’s relation to a sociocultural
counterpart, defined here as an archaeological culture, is a technique to simplify data
so that they can be used to address questions posed on the regional level. Hamilton
(1988) has pointed out that ceramic typologies which are formulated to define
archaeological cultures remain specific to a limited range of issues and are not
appropriate for all research goals. His interest in intermediate styles reflecting the
diffusion, evolution, or other mixing of design and its implications for social
organization and interaction looks for the interstices between traditionally defined
archaeological cultures. In the absence of evidence for stylistically or technologically
intermediate forms, the mixing of archaeological cultures is further indicated by the
consistent association of diverse ceramics on sites. Discounting for the moment all
of the alternative explanations mentioned earlier (exogamy, raiding, warfare, mimicry,
trade, and exchange) the potential of mixed group composition requires further
consideration.

The association of J untunen and Huron ceramics, or Blackduck and Selkirk

ceramics does not require any of these vehicles to account for their co-occurrence on

207

individual sites or across the region as a whole, since they represent closely related
cultural phenomena associated in space, time, and by the sharing of fundamentally
similar adaptive strategies and cultural values. Therefore, the conceptualization of
a group of people responsrble for the archaeological record Should assume some
degree of diversity in their composition, rather than resorting to a de facto
assumption of homogeneity in which the isomorphic view prevails. This is not to be
confused with a redefinition of an archaeological culture: these will likely continue
to be defined as unitary constructs necessary for the organization of archaeological
data for the reconstruction of cultural-historical issues. The operative unit
responsible for depositing the archaeological remains, however, may be composed of
a fluid membership, potentially representing elements of any or all members of the
archaeological cultures in a region. The degree of inclusiveness/exclusiveness in
ceramics in a regional pattern of archaeological sites should constitute the index of
potential if not actual group composition.

3. Is the identification of boundaries among archaeological cultures an
impractical or impossrble goal, given the apparent mixing of ceramic styles throughout
the region?

The feasibility of this goal is measured by the scope of the available data base.
Looking as single sites, or even at relatively large areas such as Isle Royale, may be
insuficient to disclose patterns of association among artifacts which reflect cultural
boundaries. All the usual caveats of sampling bias, differential preservation,

collapsed stratigraphy, etc., apply here. However, if one begins with the assumption

208

that the heterogeneity of ceramic styles in the Lake Superior region is chaotic, the
outcome of an investigation of boundaries is likely to be strongly predetermined, if
not doomed. As an alternative, we can view the consistent association of diverse
styles as the very pattern of interaction among these prehistoric groups which is not
chaotic at all, but reflects a consistent and stable level of interaction lacking
boundaries in the usual sense.

4. Is there more than one type of boundary or territoriality manifest in the
archaeological record which is masked by the ceramic mixing but is expressed in some
other dimension of material culture?

If we accept the assumption that the heterogeneity of ceramics reflects a
situation of diverse group composition, and not simply a function of trade or
exchange, or any of the other "explanations," it is still possible to discern differences
in cultural behavior indicative of boundaries at a different level. The regional
distribution of archaeological copper relative to geological sources strongly suggests
that something is operating on cultural behavior which results in a marked difference
in its value, use, and availability in the Terminal Woodland substage. The difference
in the regional distribution of copper between the Blackduck and Selkirk cultures on
the one hand, and the Juntunen on the other may be interpreted as one of
differential access to copper in which the J untunen people exercised some proprietary
restriction over the primary sources on Isle Royale and the north shore. Hamilton’s
(1988:53) use of "passive territoriality" explains differential access to certain

subsistence resources in which groups from outside a territory may harvest only

209

certain items without evoking feelings of transgression on the part of the host group.
Such may have been the case with copper. Hypothetically, the Blackduck and Selkirk
people may have been welcome on Isle Royale to fish and hunt along side of the
Juntunen people, so long as access to copper remained in control of the latter.
Conversely, the difference in the distribution of archaeological copper may reflect a
lack of interest in copper on the part of the Blackduck and Selkirk groups, in which
case a cultural difference in the value of copper, but not necessarily a spatial
boundary, is indicated.

5. Assuming that certain types of boundaries can be identified among the
archaeological cultures of the region, can any conclusions regarding territory and
resident versus transient populations be reached?

Exercising hegemony over certain resources, such as copper, reflects a
substantive territory. If the Juntunen phase people felt that they had proprietary
rights to the copper resources of Isle Royale and the north shore, they would have
included it in their own definition of territory. In addition, consistent patterns of
movement between areas on a seasonal basis, in which the composite parts of a range
formed an integrated whole of subsistence and social related behaviors, might also
suggest that the various parts constituted a whole territory for a given archaeological
culture.

The NAA of ceramics indicates that pottery was both imported and locally
manufactured on Isle Royale, and that the best evidence for long distance movement

of pots is found with the Juntunen phase. In the Lake Superior region there is

210

virtually no such thing as a year-round resident population, but the pattern of
seasonal residency for the J untunen phase is quite strong, indicating that Isle Royale
was an integral part of the seasonal cycle which was tied both to the north shore and
to the Straits of Mackinac. Given that J untunen phase pottery occurs in numbers
equivalent to that of other contemporaneous archaeological cultures, we cannot
exclude Isle Royale from the seasonal ranges of these other groups who obviously
visited the island for subsistence needs. However, the differences with respect to
copper may be taken as a substantive indication of how the island was perceived by
Juntunen in contrast to Blackduck and Selkirk peoples.

While it enters into the shady area of direct historic inference, the historical
accounts describing the spiritual aspects of copper among the Ojibwa may relate to
this difference in the distribution of copper seen archaeologically. Although the
copper mining technology and the technological applications for copper tools had
almost disappeared by the time observations concerning the aboriginal use of and
attitudes regarding copper were made, there is a vestigial remnant of a tradition
suggesting that it was an important substance to the Ojrbwa who continued to revere
it in an unmodified state after it had ceased to serve any technological needs (Halsey
1983, Kohl 1985).

This analysis has grappled with the problems of chronology and site formation
processes as they relate to the association of diverse ceramic styles on sites. That the
issue requires more dates and more data for clarification is obvious. However, in the

context of searching for a workable definition of an archaeological culture which

211

accurately reflects the composition of those groups responsrble for the archaeological
record anticipates an archaeological situation already manifestly obvious: that sites
are not simple reflections of activities nor are artifacts simple reflections of the
groups whom they served. A cognate of the finding of complex group composition
is the absence of boundaries in the traditional sense, that is, boundaries which are
defined by abrupt discontinuities of artifact distrrbutions on the landscape. This
analysis has indicated that another type of boundary potentially discemable from
archaeological data is one in which access to resources is not an either/or proposition,
but is instead contingent upon the type of resources under consideration and the
perceived distance between interacting parties.

Finally, the determination of proprietary rights to certain resources as it is
inferred from archaeological data is very complex. The correlation of the distribution
of primary sources of a specific resource, in this study copper, and that of the
potentially competing groups, or archaeological cultures, who made use of that
resource is combined with the archaeological distribution of raw and finished items
on a regionwide basis. From this it is inferred that certain groups exercised
differential access to the resource without compromising the other dimensions of

their interactions with their neighboring groups.

APPENDICES

APPENDIX A. TERMINAL WOODLAND SITES ON ISLE ROYALE.

Sites on Isle Royale which have Terminal Woodland components are briefly
described below. Artifact content is summarized in tabular form for each site.
Detailed artifact data are not presented here but may be found in Clark (n.d.).
Artifacts belonging to components other than Terminal Woodland are not included
in the tables.

201R1 CHIPPEWA HARBOR #1: Chippewa Harbor #1 is a major
Terminal Woodland occupation site on the south end of the transverse fault that
meets McCargoe Cove on the north side of the island. Archaeological investigations
began as early as 1928 (West 1929) and have continued to the present (Guthe 1930,
Dustin 1957). The site was most recently tested in 1988 by the NPS-MWAC survey
which found intact cultural features and artifacts. As can be seen from the tabular
data for this site (Table 5) the Terminal Woodland component is complex. In terms
of artifact density Chippewa Harbor #1 is second only to the Indian Point site
(201R28), although this is in part a reflection of the amount of excavation that has
taken place here. Although some Laurel ceramics are found on the site, most of the
material and all of the radiocarbon dates obtain from the Terminal Woodland
substage. The intensity of the occupational deposits is believed to be the result of
multiple reoccupations of this site, owing to its attraction with respect to logistics,
proximity to a wide variety of resources, and its aesthetic appeal. Faunal remains
indicate no resources specialization or seasonality of occupation. Virtually all
habitats available on the island are represented here: deep water, seasonal spawners,
waterfowl, migratory passenger pigeon, caribou, beaver, and riparian amphibian
species (Martin and Masulis 1989a).

Table 5. Artifact and raw material summary, 2OIR1.

CERAMICS: BLACKDUCK 4, JUNTUNEN 3, SELKIRK 15, HURON 1,
ONEOTA (?) 1, HEINS CREEK CORDED (?) 1.

LITHIC TOOLS: JST 3, GFS 2, HBL 8, QTZ 2
LITHIC DEBITAGE: J ST 5, GFS 2, HBL 29

OOPPER: WASTE 86, BARS 7, TOOLS 10, ORNAMENTS 1

212

213

201R5 FINN POINT: The Firm Point site is located on the north side of Hay
Bay opposite the Hay Bay Campground dock. Bastian (1963a) tested the site and
found a thin and diffuse scatter of materials mixed with twentieth century fishery
deposits on a level shelf of land near the water. The NPS-MWAC survey shovel
tested the site and found no prehistoric materials and no further work was done here.
Of particular note, however, is the notched projectile point made of Onondaga chert
from the Niagara Escarpment of Ontario and New York. This is the only example
of this material on Isle Royale.

Table 6. Artifact and raw material summary, 201R5.
OERAMIQ: UNCLASSIFIED TERMINAL WOODLAND 3
LITHIO TOOLS: HBL 1, ONONDAGA 1

LITHIO DEBITAGE: RSPT 1, HBL 8

COPPER: WASTE 7

201R14 MASSEE ROCKSHELTER: Isle Royale’s only known aboriginal
burial site, the Massee Rockshelter (201R14), near Point Houghton on the south

shore of Isle Royale. The name of "rockshelter" is misleading. The site is best
described as a small and narrow cleft in a sandstone ridge which is far too small to
have ever served as a rockshelter in the sense of its more common usage as a type
of occupation site in a shallow cave or overhang.

The Massee Rockshelter was first documented in 1928 when the
McDonald-Massee Expedition visited the site. The existence of the site was
apparently known to the Isle Royale fishermen and their families, and was first
discovered by fisherman E.T. Seglem’s children in 1908. According to a
communication between Seglem and West who accompanied the McDonald-Massee
Expedition one or more skulls were removed as curiosities by local fishermen (West
1929:25). The McDonald-Massee Expedition photographed the removal of skeletal
material from the site but did not produce a plan view of the in situ materials. One
artifact, a biface, was found in association with the burial feature.

In 1960 the University of Michigan Museum of Anthropology survey of Isle
Royale relocated the site and removed the remainder of the skeletal material
(Bastian 1963). Bastian (1963a:57) observed that the human bone was in a good
state of preservation and concluded that, on this basis, the burial feature likely dated

214

to the Late Woodland or Historic stage. In addition to human remains, a small
quantity of animal bone and a second bifacial tool were recovered.

In 1989 the Massee Rockshelter was relocated and examined by
archaeologists from the National Park Service’s Midwest Archaeological Center. No
ground disturbing activities were undertaken in the immediate area of the deposit but
it appears that no skeletal materials remain after initial removal in 1928 and final
excavation by the University of Michigan. Shovel tests were made in flat areas with
soil outside the cleft and along the point of land to the east in an effort to identify
associated activity areas but no materials were found.

Analysis of the osteological materials obtained by the McDonald-Massee
Expedition was performed by W. C. McKern who estimated that a minimum number
of 12 individuals was present at this site (West 1929:38-40). West (1929:24-25) notes
that, "as a large number of the small bones were missing it is possrble that the
remains were gathered up and deposited, after first having been suspended in the
trees for some time, as was the custom of the Chippewa and the Sioux." Given the
amount of bone recovered by the subsequent UMMA excavation, West’s observation
concerning the "small bones" is questionable. However, it is evident from the small
size of the cleft in the sandstone that the burials were indeed secondary since there
is inadequate space for twelve primary intemments to have been deposited at one
time.

The two bifaces associated with the remains are not stylistically diagnostic.
The artifact found in 1928 is lanceolate, approximately 57 mm long and 15 mm wide
(West 1929: Plate XVIII:3). The second biface is stemmed with a serrate blade
margin (UMMA #62111). The length is approximately 38 mm, the width 16 mm.
The raw material is possrbly GF S which may indicate a Woodland time/cultural frame
for the burials. Since there was no grave furniture, per se, with the burial feature,
the relationship between the bifaces and the skeletal material is open to a variety of
speculation. Accidental association is considered unlikely, given the context of the
deposit in a rock cleft. It is possible that the two artifacts were incidental grave
offerings or were in the personal affects with the remains at the time of interment.
Or, it may be that the bifaces are in some way related to the manner of death of
some of the individuals buried there.

The material associations are insufficient to make a sure determination of
cultural or chronological assignation. On the basis of the site’s location near the
modern water level of Lake Superior it is likely that the site postdates the Archaic
stage which, on Isle Royale, conforms to elevated beach features between 40 and 60
feet above modern lake levels.

215

Sauer’s (1990) complete reanalysis of the skeletal material increased the
minimum number of individuals from McKem’s (1929) estimate of 12 to 15 and
identified a number of postmortem modifications of the remains, including cut marks
and perforations on long bones, and evidence of scalping on one skull. As the only
known burial site on the island, and the only possrble instance of violent interaction,
it is unfortunate that no linkage can be

made to any of the archaeological cultures of the area using either ceramics, raw
materials, or epigenetic traits.

Table 7. Artifact and raw material summary, 201R14.

LITHIC TOOLS: HBL 2

201R17 GRACE ISLAND: This is an occupation site on the eastern tip of
Grace Island. Dustin (1957:19) made some tests but found nothing. The UMMA
survey located Terminal Woodland artifacts but no significant features or deposits.
Shovel testing by 1989 MWAC survey identified a Terminal Woodland occupation
mixed with late nineteenth and twentieth century deposits.

Table 8. Artifact and raw material summary, 201R17.
CERAMIOS: HURON 1, UNCLASSIFIED TERMINAL WOODLAND 1
LITHIO TOOLS: PLS 1
LITHIO DEBITAGE: GFS 1, HBL 3
OOPPER: WASTE 16

201R18 ORACE POINT: This site was known only as a possible location of
an American Fur Company fishery in the mid nineteenth century until the UMMA
survey team found two flakes in their excavations (Bastian 1963a). In 1989 MWAC

survey shovel tested and excavated test units here, finding intact Terminal Woodland
deposits but little evidence of the American Fur Company fishery.

216

Table 9. Artifact and raw material summary, 201R18.
OERAMIOS: JUNTUNEN 1, HURON 2

LITHIC DEBITAGE: JST 19, GFS 12, HBL 13, AGATE 67, PLS 23, QTZT 2

201R27 BIRCH ISLAND (LINKLATER): The Birch Island site, a small

island immediately south of Indian Point at the entrance to Brady Cove, was first
reported by the MacDonald-Massee Expedition in 1928 (West 1929:31) and
subsequently tested by Guthe in 1930, Spaulding in 1957, Bastian in 1961 (Bastian
1963a), and in 1983 by NPS personnel. While Bastian (1963az28) states that both
Initial and Terminal Woodland components are present on Birch Island, the UMMA
survey tests did not recover any ceramics, and the basis of Bastian’s statement was
the small collection of pottery from the MacDonald-Massee Expedition. To further
exacerbate the problem, it is evident from Bastian’s analysis of the ceramics that the
provenience of sherds from Chippewa Harbor and Birch Island became mixed at
some point. We are, therefore, left with uncertain evidence for the nature of the
Woodland components at this site, although it seems valid to assume that both Initial
and Terminal Woodland stage occupations are represented, given the proximity to
Indian Point where both are amply documented.

291R28 INDIAN POINT: In many respects the Indian Point site is the north
Shore’s counterpart of the Chippewa Harbor #1 site, although a more heavily used
location than the latter. Like Chippewa Harbor, Indian Point represents an intensive
prehistoric occupation site which has received a commensurate amount of attention
by archaeologists. Its logistically advantageous location near the head of McCargoe
Cove places it close to the Minong Ridge area of intensive copper mining and the
north end of the major natural route across Isle Royale.

The site was first discovered by Fred Dustin in 1929 and first tested by Guthe
the following year, although it was earlier identified by Ives (1847) as an historic
Indian camp. Guthe and Dustin were rewarded by the recovery of a "gallon can full
of sherds, a number of animal bones, two flint chips, one probably a scraper, two
hammer stones, and three other stones" (Guthe 1930). A circular depression was also
examined with uncertain results. The UMMA survey conducted more extensive
excavations here in 1960 and 1962, including "many small test pits" and 16 five foot
square units near the west side of the point. The Indian Point site has Initial and
Terminal Woodland components, and a strong 18th and early 19th century aboriginal
component. The 1988 MWAC survey interest in the Indian Point site was limited to
an assessment of the impact of shoreline erosion, and did not include any testing.

217

Bastian (1963a:31-32) recognized the multicomponent nature of the Indian
Point site and sought to define stratigraphic separation between them in his
excavations. While his discussion is unclear as to the vertical relations between the
various components, it seems safe to conclude that there is no stratigraphic
separation and that all cultural materials are found near the surface. No attempt at
discerning the horizontal relations among components is discussed in his report.

A rough estimation of the artifacts found at the Indian Point site indicates
that both the Initial and Terminal Woodland components are quite substantial.
There is a minimum of 25 Initial Woodland vessels, most of which are clearly Laurel.
The Terminal Woodland component has a minimum population of 33 vessels,
exclusive of eight miniature vessels. The Terminal Woodland ceramics are
characteristically variable, with examples of push-pull, stamped, CWS impressed,
punctate, plain, and notched rim motifs. There is at least one example of Huron
Incised pottery.

Lithic and copper artifacts are not so readily differentiated between Initial
and Terminal Woodland. There is a minimum of nine triangular projectile points
which are attributable to the Terminal Woodland, and examples of large notched ( 1)
and small notched (3) points. Endscrapers are the most common formal tool on the
site. Copper artifacts include an array of the forms of finished items and waste
products that are typical of the copper found on later prehistoric sites. Finished
products include a hook (I), tanged knives (3), awls (11), discoidal beads (13),
tubular beads (7), spiral bead (1), small bars (24), and large bars (13).

Table 10. Artifact and raw material summary, 20IR28.

SERAMIQS BLACKDUCK 1, MACKINAC 1, SANDY LAKE 1, SELKIRK 10,
HURON 3, UNCLASSIFIED TERMINAL WOODLAND 17

LITHIO TOOL_.S: Total of 138

E I A E: Total of 2049
OOPPER: WASTE 242, BARS 62, TOOLS 15, ORNAMENTS 34

218

201R29 BELLE ISLE #1: In spite of the long period of historic activity on
Belle Isle no records of prehistoric material from this locale exist prior to the
UMMA survey in 1961. Roy Drier established a base camp here for his work on the
Minong Mine but did not comment on any collecting activity while at Belle Isle in
1953 and 1954 (Drier 1961:l). Neither did the McDonald-Massee Expedition remark
on any field work at Belle Isle during their operation in 1928. It seems unlikely that
such a site, with extensive and abundant artifacts of all categories would go unnoticed
while undergoing development as a resort with a golf course and other facilities
located on the prehistoric component. As it stands, however, the first work for which
documentation is available is that of the University of Michigan.

Bastian (1963a:34) characterizes the results of the excavation of "several small
test pits and one trench" as "relatively unproductive." A total of ten pieces of
debitage, 156 sherds, and one piece of worked copper were recovered at this time.
Bastian’s interpretation of this small assemblage placed it in the Terminal Woodland
stage on the basis of his "rectangular impressed" and "thickened lip" ceramic
categories being present.

Seven test excavation units were made in 1988 by the NPS-MWAC survey
team. The site contains an Initial Woodland Laurel component defined on the basis
of its characteristic pseudo-scallop shell stamped ceramics. There is also a Terminal
Woodland component with a wide variety of ceramics, including Huron, Blackduck,
and Selkirk. The site has a broad area of occupation midden with good preservation
of faunal material. Copper-working is documented by the abundant waste and
finished copper artifacts and large hammerstones.

Table 11. Artifact and raw material summary, 201R29.

OERAMIQ: BLACKDUCK (Nett Lake Cord Imp.) 1, SELKIRK 3, HURON
(Huron Incised) 1, UNCLASSIFIED TERMINAL WOODLAND 3

LITHIO TOOLS: JST 1, RSPT 2, HBL 5

LITHIC DEBITAOE: JST 23, GFS 6, KAK 1, RSPT 54, HBL 104, PLS 1, QTZ
5

OOPPER: WASTE 40, BARS 7, TOOLS 6, ORNAMENTS 4

HAMMERSTONESZE ES: 7/0

219

2 I 1 S POINT DUNCAN NARROWS) Now known as Duncan
Narrows, the Grass Point site is located on a level point of land on the south side of
the entrance to Duncan Bay. Fred Dustin, acting on the report of the finding of a
stone pipe, visited the Grass Point site in 1929-30 (1957). His excavations produced
equivocal results pertaining to the prehistoric occupation he believed was here. More
convincing evidence in the form of Terminal Woodland ceramics (variety unspecified)
was found by a NPS employee in 1960, and the site was tested by the UMMA survey
team in 1962. Historic materials relating to the 19th century American Fur Company
fishery were abundant in the test units. However, one unit contained 34 pieces of
stone debitage and one piece of worked copper. The 1988 MWAC survey team
made surface collections, excavated two test units, and salvaged a Terminal Woodland
feature eroding at the shoreline.

A total of ten flakes were found at the Grass Point site of which seven are
diagnostic. All are J ST and, with one exception, were recovered from the eroding
feature. One episode of flintknapping is indicated by the homogeneity of the JST
which is derived from the same parent source. Production of a single tool from a
bifacial preform is possrbly represented by the debitage from this feature. The feature
also yielded a unifacial side scraper made from the same parent raw material as the
debitage. A diabase hammerstone was found on the surface near the eroding feature.
Given the proximity to documented copper mines above Grass Point on Mount
Franklin and Point Lookout, the absence of copper here is likely a function of
sampling this low density site.

Table 12. Artifact and raw material summary, 201R31.
OERAMIQ: UNCLASSIFIED TERMINAL WOODLAND 1

LITHIO TOOLS: JST 1
LITHIO DEBITAGE: JST 43, HBL 1

HAMMERSTONESZE MS: 1/0

220

SISKIWIT MINE (201R41): The Siskiwit Mine site was the location of a
prehistoric mine and occupation, an American Fur Company fishing establishment,
and a nineteenth century copper mine. The University of Michigan survey (Bastian
1963a) recovered 16 hammerstones from a fissure mine at this locality. In 1986 a
sheet midden was located and samples recovered. The midden contained beaver and
fish bones, stone tools and debitage, copper waste, and very small amounts of
nondiagnostic pottery. In order to clarify the nature of the occupational deposit and
its relationship to the mining area above, the Siskiwit Mine site was selected for
limited test excavation in the spring of 1987.

A total of 27 pieces of waste copper (Table 13) and two finished tools were
found in the 1987 excavation units. Initial stages of fabrication are well represented
by raw copper with and without matrix, as well as examples of all flat copper waste
types. Two small butterknives were also found in the midden deposits. Fragments of
three Terminal Woodland vessels were recovered from the excavation units at the
Siskiwit Mine site. All have typological affinities with the Straits of Mackinac
sequence.

Table 13. Artifact and raw material summary, 201R41.

@RAMICS: JUNTUNEN PUSH-PULL 3, UNCLASSIFIED TERMINAL
WOODLAND 2

LITHIO TOOLS: HBL 3

LITHIO DEBITAOE: JST 1, GFS 19, HBL 5, PLQ 1

OOPPER: WASTE 27, TOOLS 2

20IR42 OEMETERY ISLAND: This small island in the Rock Harbor channel
is known and named for the cluster of historic graves on the island’s summit. A
prehistoric Terminal Woodland occupation was identified here in 1985 when an NPS
employee found a rim sherd on the surface near the nineteenth century cemetery.
An additional example of Terminal Woodland pottery representing a pot break was
found in shovel testing another area of this small island in 1986 (Clark 1987).

221

Table 14. Artifact and raw material summary, 201R42.

ERAMIQS: UNCLASSIFIED TERMINAL WOODLAND 2

291R4§ DAISY FARM: The Daisy Farm site is situated at the mouth of
Benson Creek This is as desirable a location as can be had anywhere on Isle Royale,
and its complex occupational history is reflected in the surface and subsurface
deposits distributed over the large area between the shore and first bedrock ridges
behind the site. The prehistoric component at the Daisy Farm campground was
initially discovered in 1957 by Albert Spaulding of the University of Michigan. In
1987, the Daisy Farm was test excavated by the 1988 survey. Two 1 m units were
established in the vicinity of positive shovel test finds made in 1987 by MWAC on the
east side of the mouth of Benson Creek. This effort was rewarded by the discovery
of Initial Woodland (Laurel) ceramics in one unit, and in the other by a dense
Terminal Woodland midden containing ceramics, copper, lithics, bone, and botanical
remarns.

The variety of lithic raw materials showing a high proportion of HBL, is
typical for a Terminal Woodland site. There is one flake which is possrbly Bois Blane
or Fossil Hill/Collingwood chert from the eastern Lake Superior or northern Lake
Huron basin.

Twenty nine pieces of waste copper were recovered from excavation unit #2
at Daisy Farm. The association is strictly Terminal Woodland and all stages of
fabrication are present with a proportionate emphasis on later stages. The
co-occurrence of copper waste, ceramics, lithic tools, and abundant food remains
argues for a lack of spatial separation between activities at this site. The nearest
known prehistoric copper mine is 300 m to the north at the Ransom mine site
(201R43) and it is likely that most of the initial processing took place at or near the
site of extraction prior to transporting the raw copper to the occupation site at Daisy
Farm.

All ceramic artifacts from the Daisy Farm site have been found in subsurface
contexts. Test unit #1 contained the sherds of two Initial Woodland vessels while
TU #2 had the remains of at least four distinct Terminal Woodland vessels. Seven
bodysherds and 41 indeterminate sherds were not assigned.

222
Table 15. Artifact and raw material summary, 201R45.

@RAMICS: BLACKDUCK 1, JUNTUNEN PUSH-PULL 1, HURON (Ontario
Oblique) 1

LITHIO TOOLS: HBL 2
LITHIO DEBITAGE: JST 1, GFS 1, HBL 55, QTZ 5, AGATE 1, BBL 1

COPPER: WASTE 30

201R52 BAKER PT.: The UMMA survey discovered this site in 1960
(Bastian 1963a:38). It is located on a prominent point at the entrance to Moskey
Basin. A small prehistoric component was identified by a thin scatter of lithic flakes,
a copper awl, and a substantial portion of a Lalonde High Collar (Huron) vessel,
dating ca. AD 1450.

Table 16. Artifact and raw material summary, 201R52.
CERAMICS: HURON (Lalonde High Collar) 1
LITHIO DEBITAGE: GFS 1, HBL 2

COPPER: TOOLS 1

@1353 @IPPEWA HARBOR #2: This site is located opposite Chippewa
Harbor #1 on a small point on the west side of the entrance to the harbor. Bastian

(1963az22) reports a shallow deposit of prehistoric materials mixed with historic
artifacts and tentatively assigns it to the Terminal Woodland on the basis of its
proximity to Chippewa Harbor #1. No prehistoric artifacts were found here in the
1987 survey.

223
Table 17. Artifact and raw material summary, 201R53.

CERAMIQ: UNCLASSIFIED TERMINAL WOODLAND 1

LITHIO TOOQ: JST 1

LIIflIO DEBITAOE: HBL 7
OOPPER: WASTE 1, BARS 1

201R56 ROCK HARBOR: The Rock Harbor site was originally found by
Spaulding who collected a sherd from the beach here in 1957. The site is located on
a small sandy point about one half mile southwest of Daisy Farm. Subsequent tests
by the UMMA survey were negative with the exception of "one large unit" which
contained Terminal Woodland (Bastian’s large punctate and linear punctate
categories) and recent historic artifacts (Bastian 1963a:37-38). In May 1987 the site
was relocated by the presence of lithic and ceramic artifacts on the beach and in the
shallow water off the sandy point upon which the site is situated. The site was again
checked in September 1987 and additional collections of ceramics, lithic tools, and
debitage were made.

Table 18. Artifact and raw material summary, 201R56.
OERAMIOS: UNCLASSIFIED TERMINAL WOODLAND 2
LITHIC DEBITAGE: JST, GFS 1, KAK, RSPT 1, HBL, QTZ 1, BBL 1

OOPPER: WASTE 2

ZOIRQQ BflNOM ISLAND: The island is located north of Washington
Island which protects it from the open waters of Lake Superior to the south and
southwest. The UMMA survey excavated several test trenches in which one fiagment
of worked copper and 15 flakes were found (Bastian 1963a:61). As part of the 1989
NPS-MWAC survey Barnum Island was shovel tested from end to end at 10 m
intervals. This effort confirmed the location of a prehistoric site on the northeastern
end in the same area occupied by the Johns Hotel and extended the site a short
distance to the west. The site was mapped and one excavation unit dug in an attempt
to define the prehistoric component. Although intact cultural deposits were
encountered, no diagnostic materials were recovered from the excavation.

224

The site is considered Terminal Woodland on the basis of the lithic raw
materials. A total of 221 waste flakes were collected from the shovel test and
excavation unit in front of the hotel. One is HBL chert and the rest are J ST
representing the reduction of only a few parent pieces of material.

Table 19. Artifact and raw material summary, 201R64.
LITHIO DEBITAGE: J ST 223, HBL 4, QTZ 1

COPPER: WASTE 2

201R65 WASHINGTON ISLAND #1: This site is located on the north side
of the extreme northeastern end of the island and combines an Initial Woodland and
late nineteenth/early twentieth century fishery in its archAeological remains. The
UMMA survey tested the site in 1960 and 1961, finding "rather evenly distributed
but sparse occupational debris for about 100 feet along the shore and for 20 to 30
feet back from the beach," (Bastian 1963a:24). Artifacts included 22 sherds with
Initial Woodland decorative techniques and coil breaks: pseudo scallop shell and
complex push pull. The NPS-MWAC survey confirmed and expanded upon the
earlier findings. The site was shovel tested in 5 m intervals and one unit was
excavated. Ceramic artifacts indicate both Initial and Terminal Woodland
occupations. A total of seven sherds represent two vessels; one Initial Woodland and
one Terminal Woodland. An additional two Laurel vessels were identified by the
UMMA excavations.

Table 20. Artifact and raw material summary, 201R65.
CERAMIQ: SELKIRK 1
LITHIO DEBITAGE: JST 1, HBL 10

OOPPER: WASTE 26, TOOLS 2

HAMMERSTONEZE MS: 0/1

225

EIRB MQOAROOE COVE: The prehistoric component of the McCargoe
Cove site was discovered by the UMMA survey in 1960 when a surface collection was
made in the area of the exposed rock surfaces just above the NPS dock. Tests of
"sodded areas" were made in 1961, resulting in the recovery of nondiagnostic lithic
debitage (Bastian 1963a:27). Additional surface collections were made in 1962 by the
UMMA survey crew, in 1986 and 1987 by MTU, and in 1988 by NPS-MWAC. The
prehistoric site occupies a small rock point which extends into the southern end of
McCargoe Cove on the northwest side, just north of where the outlet stream of
Chickenbone Lake empties into the cove. It is the nearest known occupation site to
the extensive mining complex on the Minong Ridge, but has yielded only scanty
evidence of prehistoric activities. No pottery has been found here: artifacts consist
of lithic debitage and worked copper. The latter category is somewhat problematic,
given the intensive historic mining operations here. Some of the worked copper
found in surface collection may, in fact, obtain from the historic period, thus leaving
only the waste flakes to attest to the prehistoric occupation.

Table 21. Artifact and raw material summary, 201R73.

LITHIO DEBITAGE: JST 12, GFS 2, RSPT 1, HBL 20, RHY 1

OOPPER: WASTE 15

HAMMERSTONESZE MS: 0/1

A)IR7§ MERRITT LANE: The site’s location is "a small shelf along the
rocky northwest shore of Merritt Lane north of the channel separating Merritt and
Boys islands" (Bastian 1963a:36). Test excavations by the UMMA survey contained
Initial and Terminal Woodland ceramics and a large projectile point "of probable
Middle Woodland affinity." A grooved axe was found by NPS personnel prior to the
UMMA survey (Bastian 1963a:377, Plate 50a). Surface collection of the site in 1987
added 19 pieces of lithic debitage.

Table 22. Artifact and raw material summary, 201R78.
LITHIO DEBITAOE: JST 1, HBL l6, AGATE 1

OOPPER: WASTE 7

226
201R29 MERRITT ISLAND: Another site discovered by the UMMA survey,
the Merritt Island site is a late prehistoric/early historic occupation, documented by
a small projectile point, a blade gunflint, and four tubular copper beads. The site is
situated near the center and on the northwest side of the island, just opposite the
Merritt Lane site, and may be island #5 mentioned by Ives (1847) where he noted

"an old campground of the Indians." In the 1987 survey one flake was found in a
shovel test, and a netsinker of PLS was found on the surface of the site.

Table 23. Artifact and raw material summary, 201R79.

CERAMICS: UNCLASSIFIED TERMINAL WOODLAND (?) many small sherds

LITHIC TOOLS: HBL 1, PLS 1
LITHIO DEBITAOE: HBL 43, QTZ 1
OOPPER: WASTE 7

201RIQ7 WEST CARIBOU ISLAND: A prehistoric occupation was
discovered in the campground on West Caribou Island in 1986 (Clark 1987). The

small ceramic sample from West Caribou suggests, but does not confirm, a placement
in the Terminal Woodland substage. One is a neck sherd with portions of two
punched through punctuations.

Table 24. Artifact and raw material summary, 201R107.

(LRAMIQ: UNCLASSIFIED TERMINAL WOODLAND 1

LITHIC DEBITAOE: JST 8, GFS 2, HBL 4

OOPPER: WASTE 5, AWL 1

227
201R111 ROCK HARBOR LIGHTHOUSE: The Rock Harbor Lighthouse

site is located at one of the major entrances into Rock Harbor. The prehistoric
occupation of the site reflects intermittent, short term use during the Terminal
Woodland substage. The presence of prehistoric artifacts at Rock Harbor Lighthouse
was initially noted by Dustin (1931:9) with the recovery of a flake west of the
lighthouse. Surface examination and shove] testing in 1986 (Clark 1987) provided
evidence that prehistoric materials were present in subsurface context. The 1987 test
excavations at the Rock Harbor Lighthouse produced a total of 85 waste flakes and
shatter, ten chipped stone tools and 30 sherds of Terminal Woodland ceramics.
Prehistoric materials were distributed from immediately below surface to about 40 cm
below surface. Native copper occurs in relatively large quantities at the site. The
copper artifacts occur in forms that reflect the intermediate and final stages of tool
manufacture, rather than extractive processes.

Table 25. Artifact and raw material summary, 201R111.

CERAMIQ: BLACKDUCK 2

LITHIC TOOLS: HBL 1
LITHIO DEBITAOE: JST 2, GFS 2, HBL 17

OOPPER: WASTE 17, BARS 4, ORNAMENTS 1

2OIR114 LONE TREE COVE: The Lone Tree Cove site was discovered in
1988 by Shovel testing an elevated beach ridge above the Rock Harbor shoreline
(Clark 1990). The Lone Tree Cove site was selected for limited testing; two 1 m by
1 m excavation units were placed close to the location of a positive shovel test. One
test unit was devoid of cultural material while the other, only five meters distant,
contained a sheet midden, ash feature, a dog burial, and numerous artifacts. The
midden contained an abundance of pottery and calcined bone with small quantities
of lithic debitage and one copper artifact. In addition to the dog burial, faunal
remains at the Lone Tree Cove site (T. Martin and Masulis 1989a:26-42) include a
minimum of eight beaver and unidentified large mammal in the form of small
calcined bone fragments. No plant food remains were recovered from this site.

228

Table 26. Artifact and raw material summary, 201R114.
OERAMIQ: JUNTUNEN PUSH-PULL 2, HURON 1
LITI-IIO TOOLS: KAK 1, RSPT 1

LITHIC DEBITAGE: JST 12, GFS 2, HBL 15, PLQ 1

OOPPER: TOOLS 1

20IR116 THREEMILE #1: The site sits above and to the northeast of a
small cove with a good beach, somewhat analogous to the Lone Tree Cove site
although smaller in scope. The lithic artifacts from the Threemile #1 site include
46 waste flakes, one bipolar core/wedge, one projectile point, a pebble core, and two
small distal biface fragments. The projectile point is very small; probably a true
arrow point of the Terminal Woodland period, and is made of HBL chert. Both
biface fragments are the working ends of tools; one a graver or awl, the other likely
a projectile point or knife. Both are made of HBL chert. Thirteen sherds, all
probably from one Terminal Woodland vessel, were found at the Threemile #1 site.
Copper waste and bars are abundant and it is believed that this site is associated with
a nearby mine (201R118).

Table 27. Artifact and raw material summary, 201R116.

OERAMIQ: HURON 1, UNCLASSIFIED TERMINAL WOODLAND 1

LITHIO TOOLS: HBL 3
LITHIO DEBITAOE: J ST 2, HBL 44, QTZ 4

OOPPER: WASTE 80, BARS 6, TOOLS 2, ORNAMENT S 1

QIRIZO CHIOQNBONE LK.#1: The presence of prehistoric sites on
Chickenbone Lake is alluded to in general terms by West (1929) in the

McDonald-Massee Expedition report. However, this site found in 1987 is the first one
recorded on Chickenbone Lake. It is located on the only bedrock exposure on the
north shore of the east arm of the lake. Debitage, calcined bone, pottery, and burned
clay was found on the surface in an area roughly 4 m by 4 m. One shovel test
contained a large amount of small calcined bone fragments; all other tests proved
negative. Thirteen very small sherds were found on the surface. One has a corded

229

exterior and several exhrbit smoothed interior surfaces. Two fired clay lumps lack
temper and have the impressions of spruce needles. Martin and Masulis (1989az42)
have identified the bone samples, finding examples of caribou, beaver, and white
sucker.

Table 28. Artifact and raw material summary, 201R120.
OERAMIQ: UNCLASSIFIED TERMINAL WOODLAND 1

LITHIO TOOLS: HBL 1, PLS 1

LITHIO DEBITAOE: HBL 3

201R124 LAKE RITCHIE #1: In 1987 the Lake Ritchie #1 site was found
on a bedrock point facing Hastings Island, a little more than half way between the
Lake Ritchie/Chippewa portage and the Lake Ritchie/Lake LaSage portage. A
sunbumed chert scraper was found lying on top of the bedrock 25 m west of the end
of the point. Debitage was collected from the surface of a 2 m by 2 m area a short
distance to the south of this find. The site was extended 90 m to the west when
prehistoric ceramics were found in a shovel test about 10 m from the shoreline.
Plans to relocate the Lake Ritchie canoe campsite to the location of a known
prehistoric site necessitated the testing of this site in 1988. Three test units were
excavated and two contained substantial portions of a single Blackduck vessel. In
addition, there was a thin sheet midden containing a burned uniface, and a large
piece of unworked copper. A second large piece of unworked copper was found
during tentpad construction within 5 m of the excavation. Nineteen pieces of bone
include two beaver elements and 17 other small calcined fragments (Martin and
Masulis 1989b:2—3).

Table 29. Artifact and raw material summary, 201R124.
OERAMIOS: BLACKDUCK 1
LITHIO TOOQ: GFS 1, HBL 1

LITHIO DEBITAGE: GFS 2, HBL 8

COPPER: UNMODIFIED MASS 2

230

201R127 CHIPPEWA HARBOR #S: A ceramic scatter of a single broken
pot was found in the shallow water at the back of a small bay at the east end of the
west arm of Chippewa Harbor. All sixteen sherds recovered from the Chippewa
Harbor #3 site belong to a single vessel. In general the sherds are in a poor state
of preservation except where covered by a thick deposit of cooking residue.
Although the rim is badly eroded, it is similar to Huron Incised.

Table 30. Artifact and raw material summary, 201R127.

OERAMIQ: HURON (Huron Incised) 1

20IR128 LANE COVE: The Lane Cove site is located within and around the
NPS campground on the east end of a sheltered cove. It was discovered through
shovel testing by NPS personnel in 1982 when the development of the campground
was initiated (Maass 1984). The 1988 survey team surface collected all exposed areas
and excavated three units. Substantial portions of a Terminal Woodland vessel were
recovered. The rim is straight and thickened in profile with a pinched "pie crust" lip
that is smoothed. This style of pottery spans the late Terminal Woodland into the
historic period. Similar types are found at the Bell site in Wisconsin (cf, Bell site
Type II, Wittry 1963). Mason (1986) has attributed Bell site Type II pottery to the
historic Potawatomis. It is a common style at the Michipicoten site (Wright 1966)
where it is referred to it as Peninsular Woodland.

Table 31. Artifact and raw material summary, 201R128.
CERAMIQ: UNCLASSIFIED TERMINAL WOODLAND 1

LITHIO DEBITAOE: JST 1, RSPT 1, HBL 6

EIR134 MOTT SAONA BEACH: In 1987 a water rolled copper awl was
found on a gravel beach locally referred to as the "sauna beach" to the east of the

Park Service dorms and sauna on Mott Island. Its presence on this beach suggested
that an undiscovered prehistoric component was present along the eroding shoreline
and developed area immediately adjacent to the beach. A compliance related survey
in October 1989 succeeded in detecting such a site. A pot break intersected by a
shovel test contained 212 sherds from a single Juntunen drag-jab vessel. Eight
additional sherds from the trail to the beach are heavily encrusted with cooking
residue under which a smoothed cordmarked surface is discemable. They are not
diagnostic but are likely Terminal Woodland as well.

231

The site as delineated by the survey represents a Juntunen phase (ca. A.D.
1200-1450) occupation. Although copper artifacts were recovered, the absence of
copper waste materials indicates that the occupation was functionally marginal to the
copperworking activities found elsewhere on the island.

Table 32. Artifact and raw material summary, 201R134.
OERAMIQ: JUNTUNEN 1 (Juntunen Push Pull)

OOPPER: BARS 2, TOOLS 2

201R140 DUNCAN BAY #1: Two waterworn sherds from one Terminal
Woodland vessel were found on the beach of the second point 550 m southwest of
Grass Point, on the south shore of Duncan Bay. One rim and one bodysherd
represent a Terminal Woodland Selkirk pot.

Table 33. Artifact and raw material summary, 201R140.

OERAMICS: SELKIRK 1

2OIR142 BELLE ISLE #2: The Belle Isle #2 site is situated in a setting
similar to Belle Isle #1: a broad cove with a long gravel beach almost 100 m long lies
at the base of a series of beach terraces that extend 60 m back (west) from the water.
Shovel tests were made on the first six beach terraces resulting in the recovery of
prehistoric artifacts on the first, second, third, and fifth terraces. A pot break and
copper artifacts were found on the open gravel surface on the southern side of the
site. Consequently, a grid system was established and eight excavation units were dug
at the Belle Isle #2 site. Artifact density was spotty. Shovel testing in an area over
65 m from the beach resulted in the discovery of an additional occupation area. In
this part of the site the land is level but drops off steeply into a marsh, or what would
be a very small bay or backwater if higher lake levels prevailed. Judged by current
conditions, this location is sheltered, if not secreted. Two contiguous excavation units
were placed in the area of the positive shovel test. Although no midden soils or
features were identified the density of ceramics is high with a minimum of five
Terminal Woodland stage vessels. Lithics and copper, by comparison, were sparse.
The only bone identified out of a total of 51 pieces is a caribou unciform. The
remainder of the faunal material is largely calcined and attrrbutable to medium or
large sized mammals (Martin and Masuilis 1989b:3).

232

Six discrete ceramic vessels and over 300 small sherds which could not be
confidently assigned to a specific vessel were recovered from surface and excavated
contexts. All are considered representative of the Terminal Woodland substage.

Table 34. Artifact and raw material summary, 2OIR142.

OERAMIQ: JUNTUNEN PUSH PULL 1, SELKIRK 3, HURON 2

LITHIO TOOQ: HBL 2

LITHIO DEBITAOE: JST 6, GFS 1, HBL 47, AGATE 2, QTZ 3, QTZT 1

OOPPER: WASTE 6, BARS 3, TOOLS 1

HAMMERSTONESZE LigS: 1/1

ZQIR14S BELLE ISLE #3: This site represents a small activity area,
probably related to the Belle Isle #2 site. Its location is on a very small area (scarcely
5 by 10 m) of a level shallow pocket of soil, in a rock cleft above the water. Shovel
testing by members of the 1988 survey party found a diffuse shallow hearth
containing worked copper. One test unit, incorporated into the Belle Isle #2 grid
system, was subsequently opened to collect matrix samples for flotation. Screened
matrix contained a relatively large amount of small pieces of copper sheets, fish bone,
Terminal Woodland pottery, and a one HBL waste flake. The feature is likely an
annealing hearth and conforms to other possible annealing hearths descrrbed by
Bastian (1963a:30-31) at Indian Point.

Fourteen very small nondiagnostic body and indeterminate sherds were
excavated from the shallow hearth in association with the copper and faunal remains.
Exterior surfaces are smoothed and without decoration. A nonspecific Terminal
Woodland stage placement is suggested. The faunal assemblage at the Belle Isle #3
site includes lake trout, whitefish or cisco, beaver, and unidentified mammal (Martin
and Masulis 1989b:3).

233
Table 35. Artifact and raw material summary, 201R143.

OERAMIQ: UNCLASSIFIED TERMINAL WOODLAND 1

L DEBITA E: HBLI

OOPPER: WASTE 50

201R144 PIOKEREL OOVE #1: The Pickerel Cove #1 site occupies a
natural portage on a narrow neck of land which separates Pickerel Cove from the
waters of Herring Bay on Lake Superior. Its small size would have accommodated
only a limited number of people at one time. Surface survey of the tent site and fire
ring at the Pickerel Cove portage disclosed the presence of a late prehistoric and/or
early historic aboriginal site. Surface collections and the excavation of one test unit
produced a sample of glass trade beads, lithic waste, copper, and faunal materials, as
well as some later historic materials. A total of 71 waste flakes were found at the
Pickerel Cove #1 site. Fourteen pieces of copper waste were found. No finished
copper items were found and the procurement of copper and copperworking do not
appear to have been any more important than the minimal flintknapping that was
undertaken here. Martin and Masulis (1989b:3-4) have identified three large
unidentified calcined bones from the surface. The test unit contained two unburned
humerus shaft fragments from a common loon (unburned), 30 unburned pieces of
unidentified large mammal, and 13 calcined pieces of unidentified mammal.

Table 36. Artifact and raw material summary, 201R144.

LITHIC DEBITAGE: JST 5, GFS 43, RSPT 1, HBL 18, PLS 1

OOPPER: WASTE 14

HAMMERSTONESZE ES: 0/4

A)IR147 BRADY OOVE #2: Brady Cove #2 is located on the southeast
side of a small terrace at the base of a steeply sided point on the south side of the
cove. A single J ST flake was found in a shovel test and subsequent investigation of
the shoreline was rewarded with the discovery of fragments of a Huron pot and a
flake of HBL chert.

234
Table 37. Artifact and raw material summary, 201R147.

OERAMIQ: HURON 1

LITHIO DEBITAGE: JST 1, HBL 1

201R148 BRADY COVE #3: The second site in Brady Cove was found by
shovel testing a shallow soil pocket on a steep bedrock bank at the northeast end of
the cove. A single pot break is represented by the six sherds found in one shovel
test. A nonspecific Terminal Woodland placement is suggested.

Table 38. Artifact and raw material summary, 201R148.

OERAMICS: UNCLASSIFIED TERMINAL WOODLAND 1

201R149 McCARGOE COVE #2: This site is situated on a small point
across McCargoe Cove from Birch Island. Shovel tests in an area of soil within a
bedrock point revealed a dark midden: one test unit was excavated. Midden contents
included a large amount of calcined bone, an aboriginal clay pipe bowl fragment, one
body sherd, a small hammerstone, and ten waste flakes. The pipe fragment is
untcmpered clay with smoothed undecorated surfaces. The bodysherd is likewise
undecorated. A Terminal Woodland placement is tentatively suggested.

Faunal material fi'om the site are summarized by Martin and Masulis
(1989b:4): "Although six unidentified fish bones were recovered from the
fine-screened sample, the fauna] assemblage is dominated by mammals. A minimum
of one individual beaver an one individual caribou are represented."

Table 39. Artifact and raw material summary, 201R149.
OERAMIOS: UNCLASSIFIED TERMINAL WOODLAND 1

LITHIO DEBITAGE: HBL 8, QTZ 2

235

ZQIRISQ McCARGOE COVE #3: A small occupation site was identified by
the 1988 survey on the first point of land on the west side of the entrance to
McCargoe Cove. It was discovered by shovel testing a small flat area at the end of
the point. A pot break was encountered and a test unit initiated to further define the
deposit. All cultural material, mostly sherds from a single Terminal Woodland vessel,
were in the upper portion just below the duff layer and interspersed among the large
cobbles which constitute the soils here. Three flakes and a small copper bar were
also recovered from this unit.

Table 40. Artifact and raw material summary, 201R150.
CERAMICS: SANDY LAKE 1
LITHIC DEBITAGE: JST 1, HBL 2

OOPPER: BARS 1

@IRISZ OOVE #2: In order to redefine the location of the Cove site
(201R72) intensive shovel testing was undertaken along all parts of the first
(unnamed) cove on the east side of the entrance to McCargoe Cove. Three discrete
areas were defined on this basis. Time did not permit the formal testing of these
locales, but the indications of the shovel test suggest a high potential for significant
deposits. A piece of vesicular copper, one HBL bipolar core, and two sherds of
Terminal Woodland ceramics were recovered in shovel tests in a small level area
between the northem-most end of the cove and a small beach to the north open to
Lake Superior. This site is to the west of a narrow rocky isthmus which separates
Cove #2 from Cove #3.

Table 41. Artifact and raw material summary, 201R152.
OERAMIOS: UNCLASSIFIED TERMINAL WOODLAND 1

LITHIO DEBITAGE: HBL 1
OOPPER: WASTE 1

236

201R153 OOVE #3: Midden soils containing animal bone, worked copper,
three sherds of nondiagnostic Terminal Woodland ceramics, and lithics were found
in shovel tests on the east side of the isthmus. The site is situated on an older
elevated beach ridge with easy access to water on both sides. Shovel tests contained
35 pieces of mammal bone, all of which were calcined with the exception of one
caribou and three beaver elements. One caribou bone exhibits cut marks (Martin
and Masulis 1989b:4).

Table 42. Artifact and raw material summary, 201R153.
CERAMICS: UNCLASSIFIED TERMINAL WOODLAND 1
LITHIC DEBITAGE: HBL 3

OOPPER: WASTE 3

201R154 COVE #1: Cove #1 is located on the point of land at the western
entrance to the cove opposite the McCargoe Cove #3 site and extends for about 60
meters to the east along the raised shoreline of the cove. The site contains well
developed but discontinuous midden with abundant artifactual and biological remains.
Six waste flakes and a long diabase hammerflake were found in shovel tests. A
bifacial drill or perforating tool made of HBL, and a basalt celt blank were also
recovered from the shovel tests. A total of one rim and eleven body and
indeterminate sherds were recovered from shovel tests. Faunal remains consist of
eight pieces of clacined medium and/or large sized mammal (Martin and Masulis
1989b:4).

Table 43. Artifact and raw material summary, 201R154.
W: UNCLASSIFIED TERMINAL WOODLAND 1*
LITHIO TOOLS: HBL 1, BASALT 1

LITHIO DEBITAGE: HBL 6
HAMMERSTONESZEflS: 0/1

‘Approximately 4 Blackduck vessels were recorded for 201R72, the precise location
of which is uncertain, although it is known to have been some where in the Cove
area.

237

201le WASHINGTON ISLAND #2: A short distance from the Singer site
(201R80) is a previously unrecorded Terminal Woodland occupation site. The site
is located along the actively eroding shoreline in front of a former cabin site and
clearing on the south side of Washington Island. Chert debitage and worked copper
were collected from the actively eroding shoreline. Shovel tests revealed intact
occupation deposits in a wooded area to the west of the clearing. The site was
mapped and two excavation units positioned to further test the cultural deposits.

Five indeterminate sherds were collected from the eroding shoreline. TU
#2 contained 23 sherds representing a minimum of three vessels. One sherd bears
the impression of a spruce needle. One shoulder, three body, and 15 indeterminate
sherds are clearly attrrbutable to the Terminal Woodland. The presence of a possrble
coil break reflects a tentative Initial Woodland component.

Table 44. Artifact and raw material summary, 201R160.

CERAMICS: UNCLASSIFIED TERMINAL WOODLAND 1

LITHIC TOOLS: JST 1
LITHIO DEBITAGE: JST 4, HBL 15

OOPPER: WASTE 14

201R174 MALONE BAY CAMPGROUND: A prehistoric site was located

in a narrow area of soils between the actively eroding beach and bedrock below the
NPS campground at Malone Bay. Midden deposits containing ceramics, lithics,
copper, and bone are actively eroding into Lake Superior. The remaining area
measures approximately 30 m long and is 4 m wide. The site was mapped, a grid
established, and four 1 m square units excavated. Midden deposits were found in
each of the units which tended to contain similar kinds of artifacts. These consist of
Terminal Woodland ceramics, mostly Blackduck, debitage, worked copper, calcined
bone, and a few stone tools.

One notable exception was found in one unit (83-4, W26-27). Artifacts found
from 0 to 20 cm BS conformed to the types of contents described above. Between
20 and 30 cm BS (bedrock) exclusively Laurel ceramics were found with a small
number of flakes and three pieces of worked copper. Although the midden deposit
is continuous from the surface to bedrock, this is evidently a culturally stratified
deposit, unique on Isle Royale.

238
Table 45. Artifact and raw material summary, 20IR174.

@RAMIQ: BLACKDUCK 5, UNCLASSIFIED TERMINAL WOODLAND 2
LITHIO TOOIé: JST 1, RSPT 1, HBL 10
LITHIO DEBITAGE: JST 11, RSPT 2, HBL 133, QTZ 7, PLO 14, RHY 2

OOPPER: WASTE 35, BARS 1, TOOLS 2

HAMMERSTONESZE fiS: 0/5

EIRIZS MALONE ISLAND #2: This site consists of one waterworn
Terminal Woodland rim sherd and one bipolar HBL core on a gravel point on the
north side of the island. An examination of eroding surfaces and shovel tests on
intact surfaces above the point failed to disclose additional remains.

Table 46. Artifact and raw material summary, 20IR175.

CERAMICS: UNCLASSIFIED TERMINAL WOODLAND 1

LITHIO TOOLS: HBL 1

2()IR179 OREENSTONE BEAOH: Shovel tests and an examination of
actively eroding surfaces by the 1990 MWAC survey crew discovered Terminal
Woodland deposits with intact midden. The site is located in a small bay between
Malone Bay and Chippewa Harbor on Isle Royale’s south shore. Dimensions of the
site determined by shovel testing are approximately 40 m N-S by 10 m E-W.

Seven units were excavated along the length of the site, one of which was
sterile. Virtually all cultural material is limited to the top 20 cm of the deposits. At
the eastern end and center of the site area soils are characterized by dense beach
deposited cobbles with little soil. At the western end of the site where TU #6 and
#7 were located soils are coarse sands and gravels, also beach deposited. Artifact
content is typical for a Terminal Woodland site on the island: units contain debitage,
worked copper, ceramics, and bone in a discontinuous sheet midden.

239

Table 47. Artifact and raw material summary, 201R179.

OERAMIOS: BLACKDUCK 1, HURONI

LITHIO TOOLS: JST 1, HBL 4
LILLIIO DEBITAOE: JST 9, GFS 1, HBL 63, QTZ 5, PLO 13

OOPPER: WASTE 53, BARS 1, ORNAMENT S 1

HAMMERSTONESZE LA_KES: 5/3

201R180 LITTLE GREENSTONE BEAOH: In an interview with Tyler

Bastian, Chippewa Harbor resident Violet Johnson Miller reported collecting
prehistoric sherds from this small cove in 1937 but the area was not visited by
UMMA survey (Bastian 1963az80). The MWAC survey crew visited Little
Greenstone in 1990 and found actively eroding cultural deposits with intact midden.
The site is located amid the dense beach deposited cobbles approximately 10 m above
the modern level of Lake Superior.

Like Greenstone Beach, Little Greenstone is open to the wave action of the
lake which accounts for the attrition of the older beach deposits. Three 1m square
units were excavated. Most of the cultural material was contained in the top 20 cm.
Between 20 and 30 cm BS the sparse amount of soil among the cobbles gave way to
pure cobbles: that is, the soil ended and the interstices between the cobbles were
filled with air. Some small artifacts (flakes) were lost during the excavation of the
third level in one unit by falling into the cobble matrix.

Nine discrete ceramic vessels are identified by rimsherds from the three
excavation units. In aggregate, this large number of vessels represented by very small
rimsherds is unlike any other assemblage on the island. The group lacks the more
typical elements of Juntunen, Huron, or Blackduck ceramics found elsewhere and
does not readily conform to the other north shore ware groups (i.e., Selkirk or Sandy
Lake). The use of trailing as a decorative technique on one sherd suggests a
southern (Wisconsin) relationship.

240

Table 48. Artifact and raw material summary, 201R180.

OERAMIQ: UNCLASSIFIED TERMINAL WOODLAND 9

FIRED OLAY LOMPS: 12
LITHIC TOOLS: HBL 3

LITHIC DEBITAGE: JST 3, GFS 1, HBL 39, QTZ 3, RHY 1
OOPPER: WASTE 180, BARS 1,
HAMMERSTONES/FLAKES: 0/12

201R181 WRIOHT’S ISLAND: The MWAC survey team found evidence
of a late prehistoric mid nineteenth century occupation on Wright’s Island in Malone
Bay. Shovel testing and the excavation of four units produced a wide array of data
showing the mixing of prehistoric and historic materials. Only at the western end of
the site did the Terminal Woodland materials seem to be in a relatively isolated
context. One Huron potbreak was found in a shovel test where it was associated with
cut nails and burned white ware. Another interesting find was that of 17 stone
netsinkers, the context of which is uncertain. These could potentially date from both
the Terminal Woodland and Historic components.
Table 49. Artifact and raw material summary, 201R181.
OERAMIQ: HURON 1, SELKIRK 1
LITHIO TOOLS: RSPT 1, HBL 1
LITHIO DEBITAGE: RSPT 4, HBL 7, PLQ 1
OOPPER: WASTE 1

NET INKER : 17

HAMMERSTONESZE ES: 0/1

241

20IR18§ MALONE BAY #3: This site is the first and only aboriginal
copperminc recorded located on a copper deposit in conglomerate matrix. The site
is situated a short distance west of the Malone Bay Campground and immediately
west of a small gravel beach. Artifacts, primarily hammerstones and hammerflakes
are scattered across the top of a flat conglomerate shelf about 4 m above Lake
Superior. A copper-bearing fissure is located immediately to the south and west of
the site. Three Shallow excavation units and an intensive piece-plotting of surface
finds indicate that the site was a single occupation, primarily concerned with
exploitation of the copper deposits with some initial fabrication taking place on site.
A broken Sandy Lake pot was found near the fissure, along with a small amount of
unidentified bone, hammerflakes, worked copper, and FCR.

Table 50. Artifact and raw material summary, 201R183.
OERAMIQ: SANDY LAKE 1
LITHIO DEBITAOE: HBL 2, QTZ 1

OOPPER: WASTE 24

HAMMERSTONESlE LA_KES: /22

APPENDIX B: FAUNAL REMAINS FROM TERMINAL WOODLAND
SITES ON ISLE ROYALE.

The data presented in this appendix are derived from Martin and Masulis
(1989), Martin (1989), and Martin (1990). For larger collections additional data,
including weight, biomass, and ubiquity index is provided. For smaller collections,
only the number of identified specimens (NISP) and minimum number of
individuals (MN I) is provided.

242

Table 51. 201R], Chippewa Harbor #1

243

 

 

Ubiquity
Taxon NISP MNI WT(g) Biomass(kg) Index
MAMMALS .92
Snowshoe Hare 12 3 2.8 .066 .19
Red Squirrel 2 1 .4 .012 .04
Beaver 100 3 62.1 1.081 .69
cf. Deer Mouse 5 1 .2 .006 .04
Muskrat 7 1 1.3 .033 .19
Canid sp. 9 1 5.2 .116 .15
Caribou 14 1 33.9 .650 .31
Large Mammal 11 - 20.9 .406 .12
Med/Lg Mammal 2416 - 473.0 6.720 .77
Small Mammal 21 2.0 .049 .35
BIRDS .42
Common Loon 91 4 47.3 .683 .35
Duck sp. 4 1 .8 .017 .08
Passenger Pigeon 1 1 .2 .005 .04
Songbird 1 1 .1 .003 .04
Unident. Bird 44 - 3.6 .065 .23
REPTILES .23
Painted Turtle 2 1 .6 .022 .08
Unident. Turtle 8 - 2.0 .050 .19
FISH .62
Lake Sturgeon 5 1 1.0 .030 .15
Lake Trout 90 4 8.4 .162 .35
Lake Whitefish 2 1 .2 .008 .04
Whitefish/Cicso 2 1 .2 .008 .04
White Sucker 8 (2) 2.7 .066 .15
Longnose Sucker 3 (2) .3 .012 .08
White/Longnose Sucker141 7 10.7 .196 .42
Redhorse(?) 1 (1) .1 .005 .04
Pike/Muskellunge 48 5 8.4 .162 .31
Burbot 2 1 .2 .008 .08
Walleye 3 3 .4 .013 .12
Unident. Fish 1441 - 42.8 .619 .42
Scales 836 - 3.4 - .31
Unident. Vertebrate 1949 - 59.8 - .50
Totals 7279 43 795.0 11.275
Total Identified 533 187.5 3.366
Percent Identified 7.6 23.6 29.9

244
Table 52. 201R111, Rock Harbor Lighthouse.

 

 

Ubiquity
Taxon NISP MNI WT(g) Biomass(kg) Index
MAMMALS
Beaver l 1 1.0 .026 .17
Large Mammal 2 1 14.5 .292 .33
Med/Large Mammal 44 - 11.9 .259 1.00
FISH
Lake Sturgeon 7 1 2.4 .060 .17
REPTILES
Turtle 1 1 .2 .011 .17
GASTROPODS
Unident. 2 1 .1 - .17
Totals 57 5 30.1 .648
Total Identified 8 2 3 4 086

Percent Identified 14.0 40.0 11:3 13:3

245
Table 53. 201R41, Siskiwit Mine.

 

 

 

 

Ubiquity
Taxon NISP MNI WT(g) Biomass(kg) Index
MAMMALS
Snowshoe Hare 1 1 .3 .009 .05
Beaver 38 2 20.3 .415 .21
Pig 2 1 .7 .019 .05
cf. Moose 1 1 3.2 .075 .05
Caribou 1 1 1.2 .031 .05
Moose/Caribou 2 - .9 .024 .05
Cattle 2 1 1.2 .031 .05
Moose/Cattle 3 - 15.7 .320 .11
Large Mammal 58 - 26.4 .841 .26
Med/Lg Mammal 858 - 206.5 3.724 .68
BIRDS
Common Loon 4 1 .5 .011 .05
Unident. Bird 1 - .1 .003 .05
FISH
Lake Trout 198 10 15.5 .570 .37
Lake Whitefish 16 1 1.4 .046 .16
Trout/Whitefish 2 - .1 .002 .05
Unident. Fish 439 - 20.6 .349 .21
Unident. Vertebrate 23 - .9 .21
Totals 1649 19 314.3 6.470
Total Identified 270 19 59.8 1.553
Percent Identified 16.4 - 19.0 24.0
Table 54. 201R114, Lone Tree Cove.

Ubiquity
Taxon NISP MNI WT(g) Biomass(kg) Index
MAMMALS
Beaver I63 8 130.4 2.107 .88
Dog 79 1 123.3 2.004 .38
Large Mammal 13 (1) 19.4 .379 .25
Med/Lg Mammal 220 - 16.4 .326 .63
Medium Mammal 60 - 15.5 .310 .50
Totals 535 10 305.0 5.126
Total Identified 242 9 253.7 4.111
Percent Identified 45.2 90 83.2 80.2

Table 55. 201R29, Belle Isle #1.

246

 

 

 

 

Ubiquity
Taxon N ISP MNI WT(g) Biomass(kg) Index
MAMMALS
Med/Lg Mammal 79 - 10.3 .087 .83
FISH
Lake Sturgeon 1 1 .1 .005 .17
Unident. Vertebrate 389 - 6.0 - .67
Totals 469 1 16.4 .092
Total Identified 1 - .1 .092
Percent Identified <.1 - <.7 100.0
Table 56. 20IR31, Grass Point.

Ubiquity
Taxon NISP MNI WT(g) Biomass(kg) Index
MAMMALS
Beaver 3 1 7.6 .163 .67
Med/Lg Mammal 59 - 11.3 .223 1.00
Unident. Vertebrate 168 - 4.5 - .33
Totals 230 1 23.4 .396
Total Identified 3 - 7.6 .163
Percent Identified 17.6 - 54.7 41.2

247

Table 57. 201R45, Daisy Farm.

 

 

 

Ubiquity
Taxon NISP MN I WT(g) Biomass(kg) Index
MAMMALS
Red Squirrel 4 1 .4 .012 .11
Beaver 1 16 3 80.4 1.364 .78
Muskrat 1 1 .1 .003 .11
Caribou 12 2 222.5 3.409 .22
Caribou/Moose 1 - 18.3 - .11
Large Mammal 44 - 82.8 1.337 .33
Med/Lg Mammal 191 - 43.8 .789 .44
Medium Mammal 2 - .6 .017 .11
FISH
Whitefish/Cicso 2 1 .1 .005 .11
Unident. Fish 2 1 .1 .005 .11
Unident. Vertebrate 2146 - 33.5 - .44
Totals 2521 9 482.6 7.214
Total Identified 137 344.3 3.350
Percent Identified 40.8 86.2 74.2
Table 58. 201R149, McCargoe Cove #2

Ubiquity
Taxon NISP MNI WT(g) Biomass(kg) Index
MAMMALS
Beaver 3 1 2.6 .062 .29
Caribou 11 2 40.8 .741 .43
Large Mammal 49 - 31.1 .580 .57
Med/Lg Mammal 1254 - 381.9 5.543 .86
FISH
Unident. Fish 6 1 .3 .003.14
Unident. Vertebratec14700 - 223.2 - .43
Totals 16025 4 679.9 6.929
Total Identified 5 31.8 803

Percent Identified 0.7 13.0 11:6

248

Table 59. 201R144, Pickerel Cove #1 (NISP/MNI).

Moose 1/1
Large Mammal 33l-
Med/Lg Mammal 7/-
Common Loon 2/1
Unident. Vertebrate 8/-

Table 60. 201R124, Lake Ritchie #1 (NISP/MNI).
Beaver 1/1
Med/Large Mammal 17/-

Table 61. 201R142, Belle Isle #2 (NISP/MNI).
Canbou 1/1
Med/Large Mammal 50/-

Table 62. 201R143, Belle Isle #3 (NISP/MNI).

Beaver 1/1
Med/Large Mammal 6/-
Lake Trout 1/1
Whitefish/Cicso 17/1
Whitefish family 1/1
Unident. Fish 41/1
Unident. Vertebrate 76/-

Table 63. 201R153, Cove #3 (NISP/MNI).

Beaver 9/1
Caribou 1/1
Med/Large Mammal 25/-

Table 64. 201R17, Grace Island (NISP/MNI).

Snowshoe Hare 1/1
Beaver 7/1
Pig 1/1

Med/Large Mammal 104/-

249

Table 65. 20IR18, Grace Point (NISP/MNI).

Beaver 7/1
Med/Large Mammal 89/-
Common Loon 11/1
?Red-Necked Grebe 1/1
Unident. Large Bird 25/-
Trout family 1/1
Unident Fish 1/-
Unident Vertebrate 49/-

Table 66. 201R64, Barnum Island (NISP/MNI).
Med/Large Mammal 4/1

Table 67. 201R65, Washington Island #1 (NISP/MNI).

cf. Snowshoe Hare 1/1
Med/Large Mammal 125/-
Small Mammal 2/-
Unident Vertebrate 711/-

Table 68. 20IR81, Phelps (NISP/MNI).
Med/Large Mammal

Table 69. 2OIR116, Threemile #1 (NISP/MNI).
Canid (wolf?) 3/1

Table 70. 20IR120, Chickenbone Lake #1 (NISP/MNI).

Beaver 4/1
Caribou 1/1
Large Mammal 188/-
White Sucker 1/1

Table 71. 201le, Washington Island #2 (NISP/MNI).
Unident Vertebrate 3/-

APPENDIX C: NEUTRON ACTIVATION ANALYSIS OF PREHISTORIC
CERAMICS FROM ISLE ROYALE AND THE UPPER
GREAT LAKES. Michael D. Glascock and Hector Neff

250
INTRODUCTION

The following report discusses sample preparation, analytical results and
statistical interpretations for a set of 100 ceramics specimens from Isle Royale and
the Upper Great Lakes which were submitted to the Missouri University Research
Reactor (MURR) for analysis by Instrumental Neutron Activation Analysis (INAA).
The analytical procedures used are briefly described below. For more information
about the sample preparation and analysis procedures, the reader is referred to
Glascock and Elam (1988) and Elam and Glascock (1988).

SAMPLE PREPARATION

Upon receipt of the specimens and a shipping log at MU RR, an inventory was
conducted to insure that the samples had been marked with the same ID names
included on the shipping log. These were found to be in agreement. Because the
sample IDs provided by the principal investigators were not acceptable to our system,
we modified their IDS by replacing each prefix 89- with the letters IRP (e.g., 89-108
became IRP108). Our IDs were used throughout the entire analysis.

Samples were delivered as powders which had been drilled from sherds using
a carbide drill bit. Prior to sample preparation, the powders were dried in an oven
at 100 degrees Centigrade for 24 hours to reduce possible water content. After
removal from the oven, they were placed in a vacuum desiccator for cooling prior to
weighing. Samples were prepared by weighing about 200 mg of each into small
polyvials used for short irradiation. After completion of the short irradiations, the
samples were transferred into high-purity quartz vials used for long irradiations and
reweighed. In addition to these unknown samples, a number of reference standards
of SRM-1633a Fly Ash, SRM-688 Basalt Rock and SRM-278 Obsidian Rock were
similarly prepared.

IRRADIATION AND ANALYSIS

The first measurements conducted on each sample were by use of the
pneumatic tube irradiation system at MURR. In this measurement, samples were
sequentially irradiated for five (5) seconds at a neutron flux of 8 x 10" n/cmz/s.
Following irradiation, each sample was allowed to decay for 25 minutes before being
counted on a high-resolution germanium detector for 720 seconds. The short-lived
elements which were determined by this procedure were A], Ba, Ca, Dy, K, Mn, Na,
Ti and V.

The samples were allowed to decay for two weeks and then all specimens were
transferred into quartz vials as mentioned above. The entire batch of samples was

251

subjected to a 24-hour irradiation at a neutron flux of 5 x 1013 n/cmZ/s. After the long
irradiation, the samples decayed for seven days before a first measurement of 2,000
seconds was conducted on each sample using a germanium detector coupled to an
automatic sample changer. This measurement resulted in the determination of
several medium halflife elements including: As, La, Lu, Nd, Sm, U and Yb. The
samples were then allowed to decay for an additional four (4) weeks before a final
measurement of 10,000 seconds each. The latter measurement yielded the following
long-lived elements: Ce, Co, Cr, Cs, Eu, Fe, Hf, Ni, Rb, Sb, Sc, Sr, Ta, Tb, Th, Zn
and Zr.

Following collection of all element concentration data, a tabulation of the data
was assembled using the LOTUS 1-2-3 spreadsheet program. Table 72 provides a
complete listing of concentrations determined for all 100 specimens. The data were
then transferred into a DBASE file where the descriptive information was attached
to the specimens.

STATISTICAL ANALYSIS

In broad terms, the goal of compositional analysis is to identify subgroups that
are meaningful from an archaeological viewpoint. Various strategies for identifying
patterning in compositional data are discussed by Harbottle (1976), Sayre (n.d.), and
Bishop and Neff (1989). Assumptions underlying both the compositional analysis and
the subsequent search for subgroups in the data are ( 1) that chemical variation within
sources of ceramic raw materials (clay and temper) is less than variation between
sources and (2) that patterns of variation in raw materials carry over into the pottery
made from the raw materials, in effect producing a "fingerprint." Assumption #1
carries the qualification that "source" can be conceived on several levels of
geographical inclusiveness (Arnold, Neff, and Bishop 1990). Assumption #2 carries
the qualification that paste preparation techniques (mixing clays, mixing temper with
clays, etc.) affect the composition of the pottery, producing a composite fingerprint
that may not match any single raw material source (Arnold, Neff and Bishop 1990;
Neff, Bishop and Arnold 1988; Neff, Bishop and Sayre 1988, 1989).

The first step in the analysis of compositional data is to identify elements with
too many missing data to be used in group formation and to identify outliers that
might have resulted from weighing error, contamination, or other kinds of analytical
error. In the case of the northern Great Lakes data, consideration of missing data
and precision of the determinations led us to exclude the elements As, U, Co, Cr, Ta,
Hf, Ni, Sr, Zn, Zr, Al, Ca, and K. Drill-bit contamination is suspected in the case
of IRP025 and possrbly some other specimens with high Co, Cr, Ta (W was not
determined, so drill bit contamination could not be identified with certainty).
Anomalous high values of rare earth elements in the case of IRP105 are thought to
reflect inhomogeneity in the paste.

252

Following preliminary screening, average linkage cluster analysis based on
mean euclidean distances derived from log base 10 data was carried out. This
analysis suggested divisions along an east-west axis. Cluster analyses of a reduced list
of elements and of principal component scores (1-7) showed a similar pattern.
However, the tendency for cluster analysis to force data into hyperspherical groups,
coupled with the known tendency for pottery and clay compositional groups to be
elongated (due to interelemental correlation) demand skepticism regarding any
cluster solution.

The results of the cluster analysis combined with provenience and typological
information provided a basis for defining preliminary groups that were refined in
subsequent analyses. Compositional groups can be thought of as centers of mass in
the hyperspace defined by elemental concentrations (Harbottle 1976). The groups
represent "sources" at some level, whether each source is a single mine or a large
region. Such groups are defined by the centroid (multivariate mean) and variance-
covariance structure derived from the elemental concentrations (19 in the present
case). The distinctiveness of groups and probability of membership of individual
specimens can be evaluated by multivariate techniques that are analogous to t-tests
in that both the centroid and the dispersion (variance-covariance structure) of data
points are considered. In the present case, the small number of specimens in all the
proto-reference groups precluded group evaluation and refinement with multivariate
statistics based on the full set of 19 elemental concentrations.

To reduce the dimensionality in the data without arbitrarily excluding elements
that might be important in group definition, a principal components analysis of the
entire 100-specimen data set was carried out on log base 10 data. Transformation
to log concentrations partially counteracts the weighting inherent in the fact that
trace, minor, and major elements are used, the concentrations ranging from parts per
million up to percent. We conducted a parallel analysis on data standardized across
the entire data set, and noted no significant differences in the resulting groups. For
both the standardized and logged data, group refinement was based on the first 4-6
principal components, which subsume 78 and 87 percent of the variance in the data
set, respectively, as shown in Table 73. Besides effecting a reduction in
dimensionality, principal components analysis provides a new set of reference axes
on which the data may be plotted, in effect providing a new window through which
to explore the elemental concentration space.

Decisions on what specimens to shift between groups and what specimens to
exclude from all groups were based on a number of considerations, including relative
probabilities of group membership of individual specimens and the potential for a
specimen to broaden the group, making it less distinct from other groups.
Examination of bivariatc plots, both of elemental concentrations data and of principal
components scores, were used extensively in this process. The accompanying plots
(Figures 18-21) are examples of these bivariatc plots.

253
REFERENCE GROUPS

Table 74 contains the final group assignments along with typological and
provenience information and scores for each specimen on the first four principal
components. The group names reflect the most common proveniences of member
specimens. In the case of Isle Royale, two of the raw clays collected on the island
are group members (IRPOOI and IRP002) providing yet a stronger tie to geographic
space.

Table 75 shows probabilities of group membership obtained by calculating
Mahalanobis distances from each specimen to the centroid of each of four reference
groups. The first four principal components, which subsume about 78% of the
variance in the 19-dimensional elemental concentration data set, formed the basis for
these calculations. Probabilities for specimens within a group were "jackknifed" by
removing the specimen from the group before calculating probability of membership.
This practice counteracts the tendency for inclusion of a specimen in a small group
to inflate artificially its own probability of membership, a phenomenon referred to
as "stretchability" by Harbottle (1976). Some overlap between the four groups is
evident in the near misassignments. This is undoubtedly due to (l) the fact that the
groups are rather small and (2) the fact that each group probably represents a fairly
extensive source region rather than a localized source.

Another set of probabilities was obtained using only the first 3 principal
components, which subsume about 72% of the variance in the 19 dimensional
elemental concentration space (Table 76). Further dimensionality reduction made
it possible to include a fifth group, centered on the Keweenaw Peninsula. A few
misassignments and more near misassignments appear in the Table 76, but that is an
artifact of the loss of information with reduction in dimensionality. High probabilities
for membership in the Keweenaw group shown by some Thunder Bay/Voyageurs and
Juntunen specimens are due mainly to the small size of the Keweenaw group.

The basis for separation of the five reference groups is depicted in Figures 18
and 19, on which the various groups are plotted relative to the first four principal
components of the data. All the groups are separable on principal components 1 and
2 (Figure 18) with the exception of Thunder Bay/Voyageurs and J untunen which are
clearly distinct on components 20 and 21 (Figure 19).

As is common in compositional studies of pottery, a significant proportion of
specimens (37) were not included in any of the reference groups. The ungrouped
specimens were compared to each of the reference groups by calculating Mahalanobis
distances, as discussed above. Probabilities of membership of these outliers in all
groups except Keweenaw are given in Table 77. The probabilities for outliers should
be regarded skeptically and in light of other information, particularly provenience and
typology of the samples. The small sizes of the core groups and the likelihood that

254

each represents a fairly general source zone rather than a localized source means that
some non-members may show erroneously high probabilities of membership. To
reiterate, decisions as to which specimens to include and which to exclude were based
not only on the specimen’s probability of membership but also on the effect of
including the specimen on the group’s distinctiveness from other groups. Further
analyses might clarify the compositional affiliations of some of the outlier specimens.

MEANING OF THE REFERENCE GROUPS

The reference groups probably represent source zones rather than localized
sources. Source zones consist of the total ceramic environment perceived and utilized
by prehistoric potters living in a particular region during the time periods represented
in an analyzed sample. Multiple clay and temper sources as well as multiple
production locations may be represented in such groups. Thus, for example, the Isle
Royale group contains, in addition to pottery from a number of sites on Isle Royale,
raw clays from three different collection locations, two separate locations on Isle
Royale itself and one on the adjacent mainland. Further subdivision of the identified
reference groups and more precise identification of production locations would
require a much more intensive program of sampling and analysis.

Above the source zone level (represented by the five compositional reference
groups), there is a fundamental dichotomy in the data set between Sand Point and
Keweenaw groups on the one hand and the Juntunen, Isle Royale, and Thunder
Bay/Voyageurs groups on the other hand. Isle Royale, Thunder Bay/Voyageurs, and
Juntunen probably belong to a similar geochemical system in which variations in
elemental concentrations tends to occur along parallel hyperplanes of correlation.
For example, the Fe-Sc (Figure 20) plot shows the Juntunen and Isle Royale groups
lying on basically the same correlation line, though Juntunen specimens tend to be
lower in both elements. On La and Eu (Figure 21), Juntunen and Thunder
Bay/Voyaguers are completely overlapping. On both the Fe-Sc plot and the La-Eu
plot, the Sand Point and Keweenaw groups lie on correlation lines with different
slopes, suggesting derivation from a distinct geochemical environment. Principal
components #1 and #2 (Figure 18) combine the effect of distinct rare earth and Fe-
Sc correlations: through this different "window," the Thunder Bay/Voyageurs, Isle
Royale, and Juntunen groups are elongated along a line with neutral or positive
slope, while the Keweenaw and Juntunen groups are elongated along lines with
negative slope.

As reflected in the probability calculations presented in Tables 75-76, both the
Sand Point and Keweenaw groups are "loose" in that specimens in other groups
sometimes show high probabilities of membership. This looseness arises from the
fact that there is substantial variation in elemental concentrations in the small sample
of analyzed specimens, 3 result not unexpected for groups inferred to represent

255

"zones" rather than localized sources. Identifying Keweenaw and Sand Point as
distinct groups depends on analysis of correlation in the data, not just on proximity
in Euclidean space. This example provides an elegant demonstration of the fallacy
of removing correlation from a data set before attempting to define groups.

Subgroups in the data below the source zone level are suggested by substantial
variation in some of the reference groups, particularly the J untunen group. Principal
components analysis of the J untunen group alone (Table 78) indicates that the largest
dimension of variance, which subsumes about 48% of the total variance, is a
dimension on which higher scores indicate enrichment of the rare earth (or
lanthanide) elements combined with slight dilution of Rb, Ba, and Na. The rare
earth elements tend to stay together during the weathering process, making them
particularly appropriate for differentiating the end results of weathering, i.e., clays.
Rb, Ba, and Na, on the other hand, are associated with rock forming minerals: Na
occurs as a major constituent of certain feldspars, whereas Rb and Ba substitute for
K and Na in the crystal lattice of feldspars; Rb, Na, and Ba go into solution easily,
and tend to be carried away in ground water during weathering. In sum, the major
dimension of variance in the J untunen group seems to reflect differing paste textures,
with coarser pastes (containing more rock fragments) scoring lower on the PCI and
finer pastes scoring higher on PCI.

Texture may also explain some of the variation in the Isle Royale group. The
group is too small to run a principal components analysis without excluding some of
the elements. However, Table 73 indicates that the first principal component of the
whole data set in part expresses enrichment of rare earths combined with dilution of
the elements carried in rock-forming minerals. Relatively low values for the Huron
specimens from Isle Royale (IRP047 - IRP053) on PC] (see Table 73) may indicate
that the makers of Huron pottery, although exploiting basically the same ceramic
environment as other potters who worked on Isle Royale, followed a paste recipe
calling for slightly greater amounts of crushed rock or sand. With a larger sample
of Huron specimens, it might be possrble to define a group that represents not just
resource procurement in the Isle Royale source zone, but a particular set of paste
preparation practices as well.

The Thunder Bay/Voyageurs group is problematic in that it includes specimens
from two widely separated regions. Efforts to form two subgroups, one representing
Thunder Bay and one representing Voyageurs, were unsuccessful. Whether this
overlap indicates widespread geochemically similar ceramic resources or movement
of vessels in prehistory is an open question. Specimens in this group also tend to
Show high probabilities (10 - 20%) of membership in the Juntunen group, a finding
related to chemical heterogeneity of the Juntunen group, discussed above. This
results suggests a need for further refinement of the Juntunen group.

256
POSSIBLE ARCHAEOLOGICAL LNTERPRETATIONS

Although each of the compositional groups is dominated by specimens from
a particular area (giving rise to their interpretation as source zone groups), several
groups contain significant proportions of specimens found archaeologically outside
the inferred source zones. If the source zone groups were larger and better defined,
these anomalies could be taken as evidence of trade or movement of people during
prehistory. However, some skepticism regarding such an interpretation is warranted
because of the partial overlap between the groups; some of the anomalies may
represent chance variation in ceramic pastes derived from chemically heterogeneous
source materials.

The Juntunen group provides perhaps the best evidence for movement of
vessels, either through trade or movement of people. With only a few exceptions, the
members of this group all show below 1% probability of membership in any of the
other groups (Table 75). But, as shown in Table 74, J untunen specimens come not
only from Bois Blane but from Isle Royale (2), Thunder Bay (1), and Keweenaw (3).
One of the Juntunen specimens inferred to have been transported (IRP106, from the
Keweenaw Peninsula) shows a high probability of belonging to the Isle Royale group,
but none of the others shows above 5% probability of belonging to any group other
than Juntunen. All specimens inferred to have been transported show above 30%
probability of membership in the Juntunen group.

While some Juntunen pottery made in the vicinity of Bois Blane appears to
have been transported to Isle Royale and farther west, typologically similar specimens
(designated Juntunen) also appear to have been made locally on Isle Royale (4) and
in the Thunder Bay/Voyageurs region (3). Such a pattern could result from
movement of people, perhaps through intermarriage. All the Isle Royale specimens
Show above 50% probability of membership in the Isle Royale group and below 5%
probability of membership in the three other groups. Two Juntunen specimens
included in the Thunder Bay/Voyageurs compositional group (IRP032 and IRP064)
were excavated on Isle Royale, while another (IRP014) was excavated in Thunder
Bay. However, as mentioned previously, Thunder Bay/Voyageurs specimens tend to
show high probabilities of membership in the Juntunen group; the three specimens
typed as Juntunen are no exception, all showing above 10% probability of
membership in the Juntunen group.

SUGGESTIONS FOR FURTHER RESEARCH

As stated several times in the foregoing discussion, sample sizes constituted
an obstacle to data analysis. The core groups ranged in size from 5 to 22 members,
and these core groups themselves represent source zones rather than localized

257

sources or production centers. If further compositional research is undertaken,
augmenting the size of the core groups should be a priority.

Because Juntunen pottery seems to present the best evidence for movement
of people and movement of pots, it may be the best focus for the next stage of
compositional research in the northern Great Lakes. One thrust of the research
should be to improve the characterization of compositional patterning in the Bois
Blane region, where the Juntunen compositional group seems to be centered.
Further analyses should include 50 to 100 specimens of Juntunen, Mackinac, and
other pottery types, and perhaps 20-30 clay samples from the region. In addition, 50-
100 specimens of Juntunen pottery from Isle Royale and other parts of the western
Lake Superior region should be analyzed in order to provide a further test of the
hypothesis that Juntunen pottery was transported over long distances.

Several approaches to clarifying suggested patterning in the western Lake
Superior region could be taken. For one thing, the nature of the Thunder
Bay/Voyageurs compositional group should be clarified through further sampling of
pottery from the two regions combined with sampling of clays. To provide further
evidence bearing on the suggestion of movement of vessels between Thunder Bay and
Isle Royale, 3 more secure characterization of the Isle Royale pottery and clays is
also desirable. The possibility that Huron pottery falls into a distinct compositional
subgroup of the Isle Royale group should also be explored.

A sampling strategy directed at broad regional coverage (as in this study) is
a reasonable first step in compositional investigations, but one should anticipate only
tentative conclusions from such a study. More secure conclusions will result from
subsequent analyses of larger samples from particular locations and particular ceramic
types. This is the general approach suggested for further compositional investigations
in the Northern Great Lakes region.

ACKNOWLEDGEMENTS

We gratefully acknowledge funding from the National Science Foundation
through the Archaeometry program grant (BNS-8801707) and the Department of
Energy, through its Reactor Sharing Grant to MURR. We also appreciate the
assistance of J. Michael Elam (Research Assistant) who performed the sample
preparation. Of course, we are responsrble for any errors of fact or interpretation.

258

~96 nNo no.3 on.$ can awn; honn cnnv Shad ocdn Snfi nnomyfl
oona4 o.vo who“ oodo Rog $34 2:. 0 844m 856 oodv vwwé anal»:
Kfin w.n: 50.5 3.3 cwfim mafia n34. n ooAv vacnd Sdn on: anon”:
NNQn ode.4 No.0 onfin Nnn.~ NNbN n3. v whom oNHnd wvnm ozn onommm
at .v cow bnn H no. 5 Had End anon no.5. cnnmd Hmnn Nbvd omen”:
89v oficfi 8am ens: oEN woo; oons vane nan vodn neon wacky:
coon n.nc hwofi ovfic SYN coo; bwmn 86v nonnd ocdn CVNN Sony:
an >63 oh: bwnv one.“ c3.~ Hmnn oo.o~ oFHd 8.8 mnbé 895:
3:. We; cN.NOn cono 3h; 98.: 3nd oo.nn Snad 3.9. Good 893:
omen NS 3.: 2&5 Rod vnfin Son cnén oand ncén oon.n enemy:
waé one wnnfi wfimh noné oonA n36 no.3 nocmd nHOn noon n85”:
Gown QR enofi no.3 wwmfi wNnnH SHE SAn mowed vooN vowNN Noon”:
name n63 no.3 3.2. wnoé cwod cnfin dev 336 monn “Sn wan—y:
nwon wnofi n~.nn NEon NwNA ohwd «an 3.3 N336 wn.£ oSn Sonny:
con.”e Wm: wonN 8.8 mood {SN mwbé Nnnn Send nhdn vwfin can”:
coon odw cog: code SYN wocn own;4 cném mannd fionm new: mam”:
nncé fin: no.8 ~n.vo ens; cooA nwod hnnn Sonmd «N. 3. vwwd van—y:
nbwn «no 8.2 «Wan non; woo; mean 55.3 onoH d oodm mon.n n Sm”:
n 5.0 wdo Neda vmnb nood vans “8.0 nNdn «Hand ncén ann Nam”:
Emmi w.no No.3 Sén con; bnmn nwnn S .oH noomd nfimm enmd SCAD:
n36 9:: 3.3 made nan; nown own;4 8.5m 3 36 mesh cnnn 38y:
nvn.v cdo Nndfi N13 awe; can cvné no.3 n 3N6 coda Rode ooomamu
ownn Nae own bnno nvné nond ~91. mnna HnNNd coAn ocHn 895:
“new boo Emma wmdn wnNA nnnd ooN.n Eda owwfic ~n.v~ Hoes Sonny:
mcHn QR: oo.wH on. 5 cenn oan CANE conn wvmnd vwdn 806 Sony:
wood fins no.3 mono H84 and Hood Rufin nmwmd Hnén ennd 8cm”:
won.“4 ofio oNdN 36o REA vonn NNNC hnnn 835 3.; owed deem”:
3&4 one wvdn $.34 wocN nnwd and vaN Send endm vwfim nos”:
nnn.~ n.nn~ vodn when ennd novd ncon ovnm cHNnd on.w~ motn 893:
anon Woo l.nw om cnén ean wan anion NcAn onNnd chwm nnvd “any:

A8936 A836 A1388 A3860 AEAEVew can—VD Easam Aaaaez A353 A353 A332 .862

gram 3:232 038.: 03 Sch 8383 03380 5 Eouubcooeoo Eon—2m— .Nh 2:3.

259

vnon nAn nn.n~ bnnn nnn.n “En nnn.n nnén nand nn.vn n52: nnommm
ovnN fion oho nn.mn nnoé nvnA n36 nn.on oonmd nn.oN “own one:—
ono.n o.on nN.nH no.3 anN oKA nno.n mg .3. onond 3.3. 536 nnomyfl
enhn n.nn. 3.2 none onNN nnnA nnon nodv hound Nnnv H n: nncmmm
own. n ado“ onna nbdh N3: SON ohoé nNSN Send bonn nnn.~ vnommu
oNNd o.no no.o~ Eunn vnoé nova ncoé 3 .nm Ebnd an. H n Nnnn nnom~=
S fin finvn 5.9. 3.2. So.“ on A vmnn nan nvnmd nn.on SEN anon”:
vncn o. fl n in“ NH .no omnd nnH A nnn.n nnnn nnnnd nn.on hnnA Sonny:
woo; n.nn Eda nn.on ooQN och; hon.»V Sém cofind nnéa bomn omen”:
nnn.nv v.No cn.n~ No.2. nnn.n noo.N nnfin ~n.o~ finond onnn nocn nnommu
Zvé finn 3.3 onéb ovoé oSN nnn.n anv hnvmd cmnn nNon Rom”:
nnn.n nob 3.3 nvdo nob; onbn ~on.n vnnv onnmd nvnv nnn.n nanny:
nmon n.nn n2: ocon finnA b2...“ NNNn nvnN n~n~d oh: ~nn.n mam”:
mg; n: nod Snn cooA onné nnn.n $.VN noomd nn.n~ nmon Romy:
mod n. fl n nc.o nNév nnn.n nnné nnvn 9.. 3 nmnwd nv.o~ Good n non—N:
nnn.n ens on.n~ 8.3. nmod oond ocvn S .oN no: 6 3.8 8nd Nam”:
ono.N v.nv nnnfi bndb nnn.n Hob; onn.n ~n.nv named nn.on nno.n Romy:
Novn n.vn 3:: vnén vnnA nnoA and ncAn bnnmd NnnN hob; enemy:
Nnoé nfio own“ nn.on nan; nond 39¢ ofinN «nnn.n nnnN omnn og—
nfinN n.nn on.nH nn.on Sn; bnné nHo.“V nn.on mnHNd ohdn Nana. n3:—
Non.N Nov andfi Nnnn nnnA no.“ nnné 8.0m noHNd nvnm nnn.n recap:
nvmn N.nn 8.3 ~o.nn 3n; 30H nnn.n cnnH ovvwd ofivm coed nan—M:
Sun.“ n.nZ hwnfi Rob go; oocn ~36 nvév nnvmd oNnn nnfin nag—M:
n51. n.nS ofimm nn.nn :NN nbnN noné No.5 nofind nfinm Hood. $.93:
8n.n ndo no.nN nfiNn onn.~ vonN nonn no.9. nnbnd nonm nNo.n nan—y:
Sen néo v0: coda nnfim nnod homo cn.nn nooNd boon Hovn Nam”:
vonn v.no KEN 3.nn nnnd 8N.N nnvn an. 3. §nd Sow nvné finch":
nan n.nn S.vm moén oon.~ vnQN nnn.n ncnn nannd no.n~ :Nn nnomyz
nH fin fino hoAN ann van nonA nfioé Nonn vooNd nvém “nod. Romy:
nnn.n v. fl n co. Hm Na . Ho Ego; 8nd nnn.n vvdn bvnmd andn gnn enemy:

3&va ABA—$6 €958 €935 A895“; 3&3: Aeaaem AEaacz 3533 A353 A282 .803

Anus—€83 {do 3:232 298% 02 Bob 8.953. 2883 E anon—«.5588 8082m— fi. 2an

260

N36 fig Nw.w «Nda NcN.m 5va.w a5N.NH ens moon: ww.$. wch macaw:
8N6 H60 3.: om.m5 H5m.m N03.” N35 36m 5%.: mm.wN O9 .9 58AM—
35. m w.5m 532 v08 meN Need vmmfi mm.wN omovd vméN a; 08mm“
58. N flow 3.3 2.3 #34 omefi 89v ovéN 5oaNd mmdN $56 macaw:
NNw. 5 H.5w mv.cN Nvéw wad wwod 93.5 3.0m 33.6 mon 35.2 3cm”:
wo5.w w.w5 onfiN 3.8 SwN vewé wand mvéN omovd eN.mN 5002 many:
m3.5 v.05 3.2 3.5m ame 356 wvod NN.5N .2de MNNm 5mm.NH Noam”:
wmwé N.m5 m2: 8N5 NmmN 30m 5V5.m wmém c5mmd amfim 55... 395:
w5m.w .35 mmNH $.55 wmvd nnn.n 58.3 HQNw 83.: w5de w35 cocky:
wNod m.w5 V5.3 5m.mm 05H; Sod Nmo.m 8.9 33.: mcdfi owed axon“—
m~56 méw nn.oN ma . Hm cawN 50c.”V mead “Nam :93 Nde wNo.w £95:
awmé v. 3 3.3 53% 956 Sod VNNd w5.mm commd owdm 392 5%—
aNmN v.v5 HN.mm 8.3. VNwA ~53V wwfim 5de vmoNd vwda 636 £93:
Nmo.3 v.50 5wéN W565 omvN god m5w.v SAN w~mmd waNN 39.9 $9”:
m5N.v 0.3 55.3 59.3 owed SYN 805 59mm «Nmnd 532. 20w v8.5:
03. m m.~5 8.: 3.8 omaé vmmé Nwmé SéN onNd 55.mN owed $93:
own. m 5.3 m5.£ 3.8 wmoN mch acmd cm.wm 5Namd mva mwod Nwommu
one. m v.3. 8.3 mm.3 85.N Nch 315 3.9. Sand 5.9. 89w $95:
58. v ado VNdm 3.8 3.5.6 mmm.m 58.5 3.; 35.6 8.5V NHw.5H 893:
m8. N Ndw 3.9 565 ENN mmaé 586 05.3 mammd 09mm MNvd @595:
58. v c.5m no.9 342 mmaN aw5.N NMN.5 w~.5v 33.6 3.8 mood mam”:
Smd wa 8.3 5965 ome wONA Nvmw woém mammd H5.Nm wvaN 559$:
Nmm.”q ado wwma 2.ch wcmN Rod Sad $5.3 Nvad mmdm 5:2: @593:
SNd WNW 3w: VN.m5 SQN mvoN $66 8.5N m5~md 8.3” 30.3 W58”:
3N6 9N5 8.: wv.m5 mocN OMNA nnn.n mH . H m oHaNd 8.3” 5m5.N V595:
ch.v wNm 3.2 owde 25A mowfi n36 :.wN 5V5Nd oNAm V5N.N M595:
chN v.9. M5.w wméo awe; Hmfim nnn.n 3.3 wMMNd N5.MN 55c. m N58”:
53.N wdv mmd 2.3 v8.N SQN wfimw 9.6m wand 8.3. mg. H #599:
me. n mém 5m. 3 86w SYN MNNN mm56 c5dm m5mmd 3.? 95. N @595:
wNo. m WNW N. 5 365 cNmN 5Nc.N v5w.v mNéN wNvmd omdN m8. N $95:

€836 €336 A355 @3500 3&3» ABA—e: Aeafiam A8932 3:53 A335 €5.83 .283

.Auoacuaoov in.” 3:282 2.98m 03 Bob 8388 0.2528 5 8355:0800 Beau—m .N5 035.5

3.3
waH
wmdfi
8.2
mfiefi
3.3
3.2
NNHH
3.3
nnma
3.2
8.3
8.9

3.6
end
ER.

waned
embed
mvwmd
mmwmd
acmbd
mwhmd
Nowmd
Need
mocvd
NmaNd
wmmcd
owwmd
wand
Nde
836
H H 86

weomd
Maud
cmmmd
NmbNd
vwmvd
mvmmd
Vmwmd
386
Rde
cammd
wcomd
mmcvd
85nd
vuvmd
momvd
393

n.nn
own
Ndm
Ham
wdm
ode

c6

mAv
NR
59.
bdm
Owe
NNV
fivw
ode
WE

mi: can”:
vmud mam”:
c3; 395:
$56 mam”:
Had Nam”:
356 fi Shy:
newd can”:
emwd gem”:
3&6 macaw:
86¢ 89?:
VNNA 8cm”:
6%.: 393:
amfia 395:
gm." mock”:
30H N853:
an“; 593:

A332 A335

261

33ku A586 Aaaaouo Ascend abuser

wad
:56.
90mm
mm?»
ahwfi
H86
304.
mama
womé
mmz.

Aaaefi age.

mom
H40
meow
ca:
QG
9.3
“am
fine
Wmv
v.8

chafing.

N2:
cw; H
v2:
mw.m~
cod

mud

3.3
wm.:
bed“
mm. ol

3&5 A535 995% Egan

01%
228
mm.mm

869

3.3.
mmdh
mafia
mwdw
cwfia
3.6m

wmmé
mmcé
034
03.6
vwwd
EQN
RNA
99.6
mg.“
Velma

EON
8mg
mwcé

£th

ensue
8m;
mmmé
mad
owed
~86

mwmé
N34.

ond
28.8

wood

82:

wfimfi
39w
Hmmfi

08.3

ofimm
3.5
swam
mmda
mvém
wm.mm
84m
$9.
wvdm
wnxnm

Q 86
Smmd
wand
9386
Named
embed
8:6
Rowe

cmdw
3.9m
8.3
03%
awn
avwm
vwda
Ngsm
26m
no.3

€53: A335 Anaeuz A333

.Qoaawaoov $2.538 fun 15wa 01.3% 02 Bet 3389. 3888 E maouabaooaoo Beau—m .N5. 2%...

A8963

«SN 85%
83 SEE
85 8:5
32 85:
83 3:5:
$3 83.3
$2. 85:
2% SEE
«Sm SEE
8mm 82%
Em? €03

262

NN.N NH.NH..N NNH.N.N H.NNN NN.N H.NNNN N.NN N.N NNN.N NNN.N NHN.N NNN.H”:
HN.N HNHN.N NNNNN N.NNN H.H..N NHHHN H.. N N. NH. NH.N.N H.NN. N HH.N.N NNN.H”:
NH..N NNNN.N NNNNN H..NNH NN.NH NNNNN N. NHH N. N NNN.N NH.N. N H.NN.H HNNnHmH
NN.NH NNNH.N HNNNN N.NHH NN.HH NHH.N.N N. NNH N. NN NNN.H. NNN. N HNN.N NNNnHmH
NN.N NHNN.N HNNN.N N.NNN NNH.H HNNN.N H.. NN N.NN NNN.H. NNH. H. 82 NH.NNHNHH
NN.N NNNH.N NNHN.N N.NNN NN.HH NNNN.N N. HN N.HN NH H.H. NNN. H. NH.N. H NNNHHNHH
NN.N NNNH.N NNNNN N.NNH NH.N NNNH.N N.NN N.NN H.HN.N HH.N. N 9.3 NH.NAEH
HN.N NNNH.N NNNN.N N. HNN NN.NH NNNNN H .H.NH H .NH. NNN.N NH.N. H. NNN. N NH.NbHH
NN.NH H.NNN.N NNNN.N N. NNN NN.H.H NNNN.N N NH.H N. N HNN.H. NNN. H. NNN.H NH.NNmH
NN.N NNNN.N NNNNN N.NN HH..NH NNNNN N. N N. NN NNN.H. NNN. N NNN.H 3.9::
NH..N NNNN.N NNNN.N N.NN NN.NH H.H.NN.N N. NN N. NN NNN. H. HHN. N NNH. H Nab:
NN.NH H.NH.N.N HNNNN N.HHN NN.NH 898 N. HNH N. N NNN. H. NNN.N NNN. H NH.N.HNHH
NN. N NHNN.N NNH.N.N N.H.H.H HN.NH NNHNN N. NN N.NNH NHN.N NNN.N NNN. H NNN.H”:
NN. N NNNN.N NNNN.N N.NNN NN.NH NH.NH..N N.NN N.HN NNN.N NNN.N NNN.H NNNHHNHH
NN.N NNNNN NH.HN.N N.NHN NNH.H NNNNN H..NN N. NN NNN.N HNN.H. NNN.H NNNHHE
HN.NH NNNNN H.NNN.N N.NNH HN.HH NHNN.N N.NNH N.N HNN.H. NNN.N NNN.H H.NNnHmH
NN.NH NHNN.N NNNNN N.NNH NN.N NNNN.N N.NHH N.N NNN.H. NNN.N HNN.H NNNNNHH
H.N. H.H NNNN.N NHNN.N N.NNN NN.NH NNNN.N H..NHH N.NNH NNN.H. HNN.H. NNH.H NNNNNHH
NN.N NHNN.N NNNNN H..HNN NN.NH NNHNN H.NN N. NNH NNN.H. NNN.N NNN.H HNEmH
NN.NH NNNNN NNNNN N.NH. NN.N NNNH.N H..NN N.N NNN.N NNH.N NNN.N NNNHHNHH
NN. HH NNNNN NNNNN N.H.NN HN.NH NNNN.N H..NN H.. NN NHN.H. NNN.H. H.NN. H NNNHHNHH
9.. NH NHNN.N NNNH.H N.NNN NH.H.H NNNN.N N.NHH H..HN NH.N.H. HNH.N HNN. H NNEmH
NN.N NHNN.N NH.NN.H N.NNH NH.NH NH.NH..N H..H.NH N.HN NH.N.N NNH.H. NNH. H NNN.H”:
NN.N NNNNN HNNNN N.NNN NN.NH NH.H.N.N N.NN N. NN NHN.N NHN.H. NNN. N NNN.H”:
NN.NH NNNNN NNNNN N.NHN NH.NN H.NNNN N.HNH N. NN NNN.N HNN.H. NH.NH NNNAHNHH
NH.NH NNNNN NH.NNN N.HNH NN.NH NNNNN N.NN N.HN NNN.N HHN.N NNH.H H.NNHHNHH
NN.NH NNNH.N NNNNH N.HNH NN.NH NNNNN N.NNH H..NN NNN.H. NNH..H. NNN.H NNN.H”:
NN.N H.NNNN NNH.N.N N.NNH NN.NH NHNN.N N.NN N.HN NHN.N NHN.H. NNN.H NNNnHmH
NH.NH NNNH.N H.NNN. H N.NNH NN.: NNNNN N.NHH N.NN NNN.N NH.NH. NNN.H NHNHHHHH
NN.N NHNN.N NH.NNN H..NNH NH.NH NNNH.N N.NN N.NN NNN.N H.HH.N NNN.N NHN.H.“:

HENNNHHN 33.: HaNeNN HENNNNN HENNVoN ENNNNN Hagen €932 €935 ANNE HENNNNNH N.NNH

.Goaaucoov “tam 3:232 2.98m NH.NH ENE NN.—9:8 £888 5 823.5588 308on— Nh 035.

263

med“ Ngod Shed Yam Sdm pend fie: dd Nbdé eavfi ewod ewe”:
mod 33d deA Name 3.2 bird mdu edv newd the «veg vwdmmm
mmdfi maevd vwmed odS med wmeNd @de dd addd endm vwwd mwdmmd
N.N.mH Ndfiwd vwmdd ode“ bed“ bdavd ddN. fire wand 35m dam.“ $de—
mmdfi Maud vowwd afimfi and mvvmd adv vdm beds mafia eNvA Edna—
mfimfi Heewd eHeNA ed: 3.2 wemdé Nde edm wand e36 mama dwdmmfl
5d Naed hwmhd N63 mad“ vamd odd dd dfiwé Zed vmmé add”:
e02 mvmwd efimod mdmfi >03 mehed ode wde mwmd Newd new; wbdmyfl
whdfi Hmeed oeved u. add 8.2 amend e. HQ dd ebNd Hewé SNA Edam:
3 .9 Nedbd mmeNA Ndw eedH thbd mdw ddm was ddmd 92 A e593:
N~d~ wwwvd dwawd efiwm Sud :med wdb dfiv weed. awhm dfidé mg”:
mudfi emwmd unwed edmfi :6 dmde addH dd eNHd wSxm 9 HA Edam":
emda ovevd Mdmwd mdea mvdfi dwwmd deH mdv need need mad; Mbdmmu
Eda waded Emed wdbfi NN.N. deemd 0de dd Edd SYN Nwed N8:—
wa: waved thwd ed: eNd hwdvd H.Ndfi NdH fined edaé wdod Eda—M:
deH amend vuewd N. _ mm 3.6 Hwevd Nde mdv :Bd NNN.N Eh; dbdmmu
mafia abded 2.84 vdmfi and sewed odd dd eevd daHN mead dedmwfl
who mmomd mbvwd admfi N.N.m emded mdN. edm need. and mod; wedmyfl
who Vdmed Shed v.3; mds mead vddH mdN add dHNN dead neona—
fiwdfi vmdad mmmwd vddm mvda ebNed hdH d vdv :mé edmgv van; eedmmm
3.3 oeved mmmwd d.md~ deH wHeNd HddN v.3 mead emed beNA medal»:
N03 aged bdddd e.e- weNH eedmd wdm m mdm add eddé mddA vedmm—
2 .a rmmmd mNVhd mdeN ee.NH mumvd ER. adv add vw H .v. Ned.“ medmmu
Neg. edmed Edd.“ Ndvm N.NNN omNNd 5:. d.~e~ eNdd has daeA Nedmmu
eh: added Kamd eAwH 3.: med «.34 den NHvd Nddé d3; Edam”:
3 .5 eewed bdebd mddN dedfi N§d v.3. m4 m Nwmd damé mod.“ ands—M:
N.NH.: dZmd dmwed add NN.: mused vddfi ddm dwmé dmvé ebdé wmdmmu
evdfi devd 3“de Vida med“ waNd fiddH dd H H H4. Nwhé wvwé bmdmwfl
ovdm e5 ed wmmhd Ndev we; H Nmmmd efid d.me aeH .m dbvd omvé emdmmu
dmd mfiemd muaed WNWN Ned amid Md: dd deed 3:. eeNA $de—
Nwd vmevd whmvd firm and $.de de. dd eewé wmaé wwmd vedmmu
Nae: ENE ENE 35.... Egg €35 Nae; HEN: ENE NNN.N flags; .5:

.Goscuaoov 38m Ecoeaz 2.98m 03 Bob NN.—mg £828 5 Ncouabaoocoo Evan—m .3. 035.”.

264

8: 88 a: 88 -~.~ :8 m8 : -m :8 v.2: 8.8 38%
8: $8 8 a: 8: 82 33. R8 8- 3.8 ~ 8: 88 88m:
88 «.38 83: -R E: 88 8%...“ 8R mm: 8: 8% ~88.
8: 3:8 82 83. mm: 28 :2 o.~..~ 88 8: ~88 :88:
33> A8? 9:2 Heads: 3:: €85 H88 €858 GEE 8&5 @8ch Bus
8.8 $88 .98 8a :8 E8 9:: man 83. 9.8 28 web:
2.: 58 ~38 :8 -8 :88 8.2 8 $8 88 88 SEE
:8 ~38 888 o.-~ :8 ~88 8: 3m 83. R8 888 85:
$3 m§~ ~23 8.8 3..- 58 38 mg 38 88. 9.8 85:
~84 ~38 :8 «an 8: ~88 n8 8 £8 §~ 3.3 3:3:
3.: E8 E8 83 :8: 38 as 8 88 82 33 SEE
3.: 88:. ~88 8% -.: ~88 2: 8~ 3.8 $3. 88 ~83”:
~3~ $8.: 58 9:: 3: ~28 8: 8 22 $2. 33 SEE
3.: -~8 ~38 88 88: :8 8: 8 8:8 38 88 8:5:
3.: 2:: 8:8 3.: 8.: ~88 8.8: w- 28 :8 c~:.~ 88E
2.: 22.: 58 8: 3.: 8:8 new 8.9. 88 £8 o:.~ 88.2:
3.: S8 38 3:. ~02 88 3.2 8 22 a? $2 8%:
3.: 58 ~38 ~.S~ :5 E8 8: 8 88 $3 E: 88m.
:8 E8 228 £- 88: E8 v.8 8.: 88+ 88:. $3 8.2:
mm: 38.: a-8 8.: c~...~ :38 82 8» ma... :8 83 8mm:
~98 ~§8 338 8 8.: :88 8: 8 88 ~93 9.3 89::
8: 38 3.2.: 82 3.2 888 8: 8 8% 38 32 ~25:
3.: ~28 .88 fig 3.: £98 8: 8 38 :8. 88 89::
~3~ ~83 388 ~.% 8.: 89.8 82 8.2. ~88 23.. $8 88%
VS $~8 3.8 88 -+~ 888 8m 8: 88 ~88 88 88.2.
8.: om-.: 8m~8 8m m8~ ~38 8.8 8m 33. 88 $3 £85
8.: 83.: 3:8 9% ~3~ :88 an: 8 28 88 83 8%:
8.: 8:: ~88 8 SR «82 E... 8: 23 88.2 22 88mm:
€99: HERE 3&5 €935 €838 €2.55 Hagen €52 @8qu @E Aegean: .252

.Heoscecoov ism 3:232 298m 23 Bob 8388 2888 5 8085588 Eon—2m .N5 033.

265

odmfi
0N3
v63
fimm
0mm
Qmofi
fimmfi
wdm
QR:
Qma
Na
Ems
emu
3:.
mm:
céc
9::
9me
w.m-
Qmw
N.N:
«.3
0.63
mm:
063
Ndfi
odw
as:
oAE
m. a:

mmbmd
«god
35mg.
2wde
wgd
Swmd
med
NHde
weomd
“mmmd
33.:
856
chmd
vmad
omawd
3.3.0
mmwmd
Smmd
83.:
2.95
ammmd
HNmNd
vmvmd
amend
mmwmd
Smmd
mmmmd
mcmmd
:36
mfivmd

AER—mg $5.

wNvA
39A
who;
woe;
mmcé
«‘34
c3;
3%.“
Sum.“
chA
Sm;
3m.m
hum."
856
854
ma;
non.“
m3:
NNvA
bmmd
moo.“
8v;
8N;
NmmA
OR:
N2;
Nmmé
vmmA
v86
mad

$22

063
adww
edge
v.85
Nd:
a; a
w.vcm
mémm
vane
kow
v.33
N6?
:63
wdbom
admv
H.003
Qmev
“dz:
@va
9.3m
flaw
Wmmc
~63
fiflh
mdcm
H.ch
hfiwm
9on
Gama
fimav

Guava—z

mmm.~
0:;
3a.“
cww.N
wmnm
mhwm
wmcé
vow;
SYN
EQN
mSN
ohm;
mafia
gem
8:,”
mg;
End
amp.“
wa.N
owe.“
mch
vafim
3mm
wvmm
owfim
mmm.~
3mm
wcm.N
mwfim
mmm.N

@Ea

mwmm
c2:
8mm
SEN
Ehm
and
34m
“gm
5mm
KEN
3cm
we;
3mm
cad
mend
aamfi
§.N
amaé
3cm
mmcm
mgd
vmeé
swmd
33.
Sad
wwwé
Bmd
can;
3am
a3;

Exam
Se;
coed
084
mvmé
Sm;
owN.N
:vd
ammd
H3;
$5.0
Sm;
mcwd
Goad
Ewd
mbwd
wild
mmeA
can;
oomd
mama
memd
NN.:
Sad
35.:
cmaA
3&6
mwwd
mch
am. e

g8

fiHmN
wdnm
odoN
«.mmm
mm?
fivwo
bfiwm
v.mmm
wdfi.
aéwv
odaw
fihmm
@ch
«dew
N.Nmm
aémm
odmm
“menu
cécm
GEN
0.30
wécv
m.mNm
odes
9QO
cémb
wAE
v.30
1ch
04mm.

3:3an

$5.5
$0.0
wwcw
mmcfi
3N6
mmwfi
mwc.w
Sad
owes
3mm
mmmfi
Haw
mmms
meH
mvwfi
O86
:‘ms
mad
E‘ww
meme
mEk‘
Ems
Hmvw
macs
39w
32%
mm»;
Sod
89.x
weed

GEE

A335

m.mm~ 3. R Romy:
:63 8. c omen”:
v.03 cod mmommu
v.9; cod «8%:
hNE 3.2. mmom~=
neg 3.2: «mag
mm: 2.3 Hmong—
“..ch wwdw emana—
twfi av. 52 mg—
onH 8. o 395:
cémfi $.va >85:
mg 3.3 each”:
QE wade mach”:
wdfl owdmfi v85:
mm: cad macaw:
068 3.8 mag”:
ma: 8.: mam”:
ado umfib Say:
3: hmmw Sona—
fifim cod mamm—
vwfi 34% 38.»:
5..me 5.8 many:
wdfi cod Nam”:
v.8 No.5 Sony:
«4.: wmdw can”:
mm: «3% coon”:
mdmfi 86m mock”:
5.: mméw boom”:
>53 $62 08.5:
and: 86 neona—
AEQBHN AEqucN 4:03

.3338an fig Raouaz 298m 03 Bob 3388 £838 5 8395588 Eon—2m .N5 038.

266

93
five
New
{mu
0%
9w:
mdw
wfim
0.3
52:
m.m-
mm:
~63

mmcmd
wvmmd
aamd
vmbmd
Novmd
SmNd
Hmmmd
vaaad
83.:
vad
camwd
Novmd
ammvd
8:3
8mmd
aamd
ochd
Namd
83d
35d
cecmd
NSmd
ammmd
Sofia
NSNd
mammd
aid
hvwmd
mNde
Send

9va

NNN.N
momm
3:“
RS.
vmmm
3 Hm
hmmé
comm
fimvm
3mm
ommm
fimam
m~m.m
mvwm
mafia
mhvfi
comm
8am
ommN
mva
3mm
EHN
onN
wam
comN
mid
3am
mEhm
“Rum
mwcm

omvé
NS.“
own;
m2:
owwm
QNNA
mama
och.“
on“:
2.54
8N6
9AA
wow;
cam.“
054
m3;
356
hwmd
3m;
vofia
036
mm;
3.3
Nmmé
“he.“
man;
#84
:9:
var.“
mmbd

063
méwc
cdmv
Q80
wéwm
omfi m
Omen

was

N.me
2.5.
mag
wdmm
03m
9on
vdmc
N.Nch

odfimfi

wmbm
wdma
home
”Now
Qmwm
qgm
wfimm
wdmh

wdm:

m.~mm
adcv
N63.
035

Bad
mfivfi
bmvd
>38
cam.»
Sod
39w
hmofi
82.
39.0
N506
fimms
890
8th
and
3mg.
30.»
gem
mmvd
mzfi
was
Sad
mmom
m8.w
owns
>32.
8mm
mafiw
hams
mwcd

mm:
06:
ad:
@th
H6:
Yum“
ad:
flog
9mm

fivmfi
ad:
Qua
b. NmH
N. mg
m. mmH
w.wm~
ad:
N.Nmfi
hcmm
«mow

m.¢m~
ho:
mdNH
Néw

fivw

mm:
EN:
Wm:
vAmH
Némm

AEAEV>

@E

3.55

3&6

@958

$5 355 €35

€03

.Goscucoov fan 3:232 298% 03 Boa 8388 £8.28 5 Eon—350:8 82.8—m .N5 2an

267

0000.0
0000.0
0000.0.
0000.0
0000.0
0000.0.
0000.0
0000.0
0000.0.
0000.0
0000.0
0000.0.
0000.0
0000.0
0000.0.
0000.0.
0000.0-
0000.0
0000.0
0000

0000.0.
0000.0.
0000.0.
0500.0.
0000.0.
0000.0-
0000.0
0000.0
0000.0
0000.0
0000.0
0000.0.
0000.0
0000.0.
0000.0
0000.0
0000.0.
0500.0-
0000.0
0000

0500.0-
0500.0
0500.0
0000.0
0000.0
0000.0
0000.0-
0000.0
0000.0
0000.0-
0000.0.
0000.0-
0500.0-
0000.0
0500.0.
0000.0
0000.0-
5000.0
0005.0
5000

0000.0
5000.0
0000.0
0000.0
0000.0-
0000.0-
0500.0
0000.0
0000.0-
0000.0
5000.0
0000.0
0000.0
0000.0-
0000.0-
5000.0.
0050.0
0050.0.
0000.0
0000

5000.0.
0000.0
0000.0.
0000.0
5000.0-
0000.0.
5000.0
0000.0.
0000.0
0000.0
5000.0
0000.0-
0500.0
0000.0.
0000.0-
5000.0.
5000.0
0000.0
0000.0
0000

0500.0.
0000.0.
0050.0-
0000.0.
0005.0.
0000.0
0050.0
0000.0
0000.0
0000.0.
0000.0-
5050.0
5000.0
0000.0.
0000.0-
0000.0
0000.0
5000.0
0000.0-
0000

0000.0
0050.0-
0000.0.
0000.0-
5050.0.
0000.0-
0000.0-
0000.0.
0000.0-
0000.0.
0000.0-
0000.0
0000.0
0000.0
0000.0
0050.0
0000.0-
0000.0
0000.0
0000

0000.0
0000.0
0000.0-
0000.0-
0000.0-
5500.0-
0050.0-
0000.0
0000.0-
0000.0-
0000.0
0050.0.
0000.0.
0000.0-
0000.0
0000.0
0000.0
0050.0
0000.0
0005

0000.0
0005.0.
0000.0.
0000.0-
0000.0
0500.0.
0000.0-
0000.0
0000.0
0000.0.
0000.0
0000.0
0500.0
0500.0-
0000.0-
0500.0
0000.0
0500.0
0000.0.
0000

5500.0
0500.0-
5000.0-
0000.0
0000.0.
5000.0
0000.0-
0000.0
0000.0
0000.0-
0050.0-
0000.0.
0000.0
0000.0
0000.0.
0000.0-
0050.0
0500.0.
5000.0.
0000

5500.0
0000.0
0000.0-
0000.0
0000.0-
0000.0
0000.0
0000.0.
0000.0-
0000.0-
0000.0.
0050.0
0000.0
0000.0-
5000.0
0000.0.
0000.0-
0000.0.
5000.0
0000

0000.0
0500.0.
0000.0
0000.0
0000.0
0000.0
0000.0
5050.0.
0000.0-
5000.0
0500.0.
0000.0.
0000.0-
0000.0.
0000.0-
0000.0
5000.0.
0500.0
0000.0.
0000

0000.0
0000.0
0000.0.
5550.0.
0000.0
0000.0
0000.0-
0000.0
0050.0
0000.0.
0000.0
0000.0
5000.0
0000.0
0000.0-
0000.0.
0500.0-
0000.0.
0000.0-
0000

0000.0-
5000.0.
0000.0.
0500.0
5000.0.
0000.0
0000.0.
0500.0
0500.0-
0000.0-
0000.0
0000.0-
0000.0.
0000.0.
5000.0
0000.0.
0000.0-
0500.0
0000.0-
0000

2:02.... 3 200...:

0000.0 0000.0 5000.0 >
0000.0. 5000.0 5000.0 05
0500.0 0000.0- 0000.0. <2
0500.0 5050.0 0000.0 :1
0500.0. 5000.0. 0000.0 >0
0000.0 0000.0. 0000.0- (0
0000.0 0600.0. 0000.0 :0
0500.0. 5000.0- 0000.0 00
0000.0 0000.0 0000.0 00
0000.0- 0500.0 5000.0 00
0000.0 0500.0. 5000.0. ma
0000.0 0000.0 0000.0 00
5000.0 0000.0. 0500.0 00
0050.0 0000.0- 0000.0 00
0000.0 5000.0. 0000.0 00
0000.0- 0000.0. 0000.0 0»
0000.0 0500.0. 0000.0 :0
0000.0. 0000.0. 0000.0 04
0000.0 0000.0. 0500.0 (0
0000 0000 000;
00000000000u neoconfiou 0000uc0ua
00000.0 0000ooo.0
00000.0 0000ooo.o
0000.0 0005ooo.o
0000.0 000000.0
0500.0 5000oo.o
0000.0 000000.0
0000.0 0000oo.0
5005.0 005000.0
0000.0 0000oo.o
0000.0 050000.0
000.0 0005oo.0
000.0 00000.0
000.0 00000.0
005.0 00000.0
050.0 00000.0
000.5 00000.0
00.00 00050.0
00.00 0000.0
00.00 0000.0

62.236 35...; 0.. «03:33.. 05 32:50;

5.58. cue-0:30.350»: .3 0300 «3220

03:0: 30: 05 cc 0300 2.0022 35.338 0022...;

020100000 000 000 so (000 000000 «On

000>0<z< 0020200200 000002000

"05 00000

268

 

 

0050.0 0000.0 0000.0 0000.0- 00<>0¢ 0000 o>o0 .050000 003000000 500000 00
0000.0 0000.0 0000.0- 0000.0- w0<>0¢ 0000 .L0: 03000000 .00000 :oc3uc30 000000 00
0000.0 0000.0- 0000.0 0000.0- 0<>0¢ 0000 00<>0¢ 0000 0:.oa 50000 .000000 .00008 :00: ouco000 000000 00
0000.0- 0050.0- 0000.0- 0000.0- 00<»0¢ 0000 0:0000 :0030 .00a< 000500 0000 000000 50
0000.0 0000.0- 0000.0- 0000.0- 0<>0¢ 0000 m0<>0¢ 0000 0:0000 ououo .500000 :oL3: 000:00 00
0050.0 0000.0 0000.0 0000.0- 00<>0¢ 0000 00 0000 00000 .000000 :oL3: 000000 00
0000.0 0000.0- 0000.0- 0000.0- 00<>0¢ 0000 00 0000 00000 .0000000 :oL3: 000:00 00
0000.0- 0000.0- 0050.0 0055.0- 0<>0¢ 0000 m0<>om 0000 5L0: 00.00 .000000 :ou3: 000000 00
0000.0 0000.0 0005.0 0005.0- 0<>0¢ 0000 00<>om 0000 00 0000 00000 .0000000 :oL3: 000000 00
0000.0 0000.0- 0000.0 0000.0- 0<>00 0000 00<>00 0000 00 o>o0 000:0 .5000000 :oL3: 000000 00
0000.0 0005.0- 0000.0- 0000.0- 00<>0¢ 0000 00 .La: 03000.00 .5000000 :oL3: 500000 00
0000.0. 0000.0 0000.0 0000.0- 00<>00 0000 00 .50: oxuoa050 .00000 003000000 000000 00
0000.0- 0000.0 0000.0- 0000.0- 0<>0>520:0 00<>0¢ 0000 Egon 00000 .000000 003000000 000000 00
0000.0- 0000.0- 0000.0 0000.0 0<>0¢ 0000 00<>0¢ 0000 00 0000 00000 .000000 003000000 000000 50
0000.0 0000.0 0000.0 0000.0 0<>0¢ 0000 00<>0¢ 0000 00 00:00.: 0:00 .0000000 xu3uxuo00 000:00 00
0005.0- 0000.0 0000.0. 0000.0 00<>oa 0000 0030:00000 .La: 30o: .0000000 003000000 000000 00
0050.0 0000.0 0000.0 0050.0 0<>00 0000 (0000220: :ozuu0xx00o: 000:0 L00: .0000 000500 0000 500:00 00
0000.0 0000.0 0000.0 0000.0 0<>0¢ 0000 m0<>0¢ 0000 0:0o: 0:00 .000000 :0:30:30 000000 00
0000.0 0000.0 0000.0 0000.0- 0<>0¢ 0000 m0<>0¢ 0000 :Eou 00000 .000000 :o:3u:30 000000 00
0000.0- 0000.0 0000.0- 0000.0- 0-20200200 00<>0¢ 0000 o>o0 00:» oco0 .0000000 :o:3u:30 000100 00
0000.0 0000.0 0000.0- 0000.0- 0-20200200 00<>0¢ 0000 o>00 00;» ucog .0000000 :uc3uc30 000000 00
0000.0- 0000.0 0000.0. 0000.0- 0<>0>52010 00<>0¢ 0000 20000 0:300 000: .0000000 co:3u:30 000:10 00
0000.0 0000.0 0000.0 0000.0 w0<>0¢ 0000 00 0000 00000 .0000000 :o:30:30 000000 00
0000.0- 0000.0- 0000.0- 0000.0- 0-20230200 ><0 «0020:» 0-00 u0 :u:3u:30 000100 50
0000.0 0000.0 0000.0- 0000.0 >00 «0020:» 0-30 00 003030000 000000 00
0000.0 0000.0 0000.0- 0000.0 0<>0>\z::0 ><0 «0020:» 0-30 00 003000000 000000 00
0000.0 0000.0 0000.0 0000.0- ><0 000:0:0 0-00 00 003000000 500000 00
0000.0. 0000.0 0000.0 0000.0- ><0 00020:5 0-x0 no 003000000 000000 00
0000.0 0000.0 0000.0 0000.0 ><0 0002::5 0-50 00 303000000 000000 00
0000.0 0000.0 0000.0 0000.0 ><0 «0020:» 0-00 uo 003000000 000000 00
0000.0- 0000.0 0000.0- 0000.0- 0<>0>xzaxh ><0 00020:» 0.00 00 003000000 000:00 00
0000.0 0000.0 0000.0 0000.0 0<>o¢ 0000 ><0 0002015 :o0uuo_0o0 :ouo3cx 003000000 000000 00
0000.0- 0005.0 0000.0- 0000.0- 0<>0>\z::5 >00 «0020:» 0-00 00 003000000 000000 00
0000.0- 0050.0 0000.0 0000.0- ><0 «002015 0-00 00 003000000 500:00 50
0000.0- 0000.0 0000.0 0000.0- ><0 «0020:» 00-00 00 003080000 000000 00
0005.0- 0000.0- 0050.0 0000.0 ><0 «0023:» 0-00 00 003000000 000000 00
0000.0 0000.0 0000.0- 0000.0- 0<>0>52010 ><0 «0020:» 0-50 00 :o:3u:30 000100 00
0000.0- 0000.0 0000.0 0000.0- 00:00<>0> 000000 003000000 000000 00
0000.0- 0000.0 0000.0- 0000.0 0<>0>\z::5 magmu<>0> 000000 003000000 000:00 00
0000.0- 0000.0 0000.0- 0050.0- 0<>0>xzazp 00:00<>0> 000000 003000000 000000 00
0005.0- 0000.0 0000.0 0000.0- 00:00<»0> 000000 003000000 000000 00
0000.0- 0000.0 0000.0- 0005.0- 0<>0>\20:5 00000<>0> 000000 003000000 000000 0
0000.0- 0000.0- 0000.0- 0000.0- mazwu<>o> 000000 003030000 000000 0
0000.0- 0000.0 0000.0 0000.0- 00000<>0> 000000 003000000 500000 5
0000.0- 0000.0 0000.0 0050.0 masuu<>0> 000000 003030000 000000 0
0000.0- 0000.0 0000.0 0000.0- 00000<00> 000000 003000000 000000 0
0000.0- 0050.0 0000.0- 0000.0- 0<>0>\:::0 mxawo<>0> 000000 003000000 000000 0
0000.0 0000.0- 0000.0 0050.0- 00<>0¢ 0000 0:0000 Lo>0om 000500 0000 000000 0
0000.0 0000.0 0005.0 0000.0 0<>0¢ 0000 w0<>0¢ 0000 “rogue: auxno: 000600 0000 000000 0
0000.0 0000.0- 0000.0 0000.0- 0<>0¢ 0000 00<>oa 0000 gonna: 03000030 000500 0000 000000 0
00m 00: 000 00: 00000 .0.0
000000 020200200 0(0002000 00000¢<oz<00 0(00xw:0 0:00 0000 00>5 .0<z< 0vsouo:

.000 <5<0 <0¢< 00<>0¢ 0000 000020 00: (000 «0:00 0:0 0020120000< 00000 00:00

"05 0000»

269

 

0000.0 0000.0 0000.0- 0000.0- 0-00000000 00000200 000000 00000 0000 000000 000
0000.0- 0000.0- 0000.0- 0000.0- 0-00000000 00200300 000000 00000000 000000 00
0000.0- 0000.0- 0000.0- 0000.0- 0-00000000 00000000 000000 00000000 000000 00
0000.0 0000.0- 0000.0- 0000.0 00000000 000000 000000 000000 00
0000.0 0000.0 0000.0- 0000.0 00000000 00000300 000000 000000 000000 00
0000.0 0000.0- 0000.0- 0000.0 00000000 00000300 000000 000000 000000 00
0000.0- 0000.0 0000.0- 0000.0- 00000300 000000 000000 000000 00
0000.0 0000.0 0000.0- 0000.0 30000000 00200300 000000 000000 000000 00
0000.0- 0000.0 0000.0 0000.0 00000300 000000 000000 000000 00
0000.0 0000.0- 0000.0. 0000.0 00000000 00200000 000000 000000 000000 00
0000.0 0000.0- 0000.0- 0000.0 00200300 00200000 000000 000000 000000 00
0000.0- 0000.0- 0000.0- 0000.0 00:00:00 000000 000000 000000 00
0000.0- 0000.0 0000.0- 0000.0 00000000 000000 000000 000000. 00
0000.0 0000.0- 0000.0 0000.0- 00000 0000 000000 0000 000000 00000000 000000 00
0000.0- 0000.0 0000.0 0000.0 00000 0000 00000 0000 000000 00000 0000 000000 00
0000.0- 0000.0 0000.0 0000.0 00000 0000 00000 0000 000000 00000 0000 000000 00
0000.0- 0000.0 0000.0 0000.0 00000 0000 000000 00.00 0000 000000 00
0000.0- 0000.0- 0000.0- 0000.0 00000 0000 000000 00.00 0000 000000 00
0000.0 0000.0- 0000.0. 0000.0 00000 0000 00000 0000 000000 00.00 0000 000000 00
0000.0- 0000.0- 0000.0 0000.0 00000 0000 00000 0000 000000 00000 0000 000000 00
0000.0- 0000.0 0000.0 0000.0 00000 0000 00000 0000 000000 00.00 0000 000000 00
0000.0- 0000.0 0000.0 0000.0 00000 0000 00000 0000 000000 00.00 0000 000000 00
0000.0 0000.0- 0000.0 0000.0 00000 0000 00000 0000 000000 00000 0000 000000 00
0000.0- 0000.0 0000.0 0000.0 00000 0000 00000 0000 000000 00000 0000 000000 00
0000.0 0000.0- 0000.0 0000.0 00000 0000 00000 0000 00000 000.0000 000000 00
0000.0- 0000.0- 0000.0- 0000.0- 0-00000000 00000 0000 00000 00000000 000000 00
0000.0 0000.0- 0000.0- 0000.0 0-00000000 00000 0000 00000 00000000 000000 00
0000.0 0000.0- 0000.0- 0000.0 0-00000000 00000 0000 00000 000.0000 000000 00
0000.0- 0000.0- 0000.0- 0000.0 00000 0000 00000 00000000 000000 00
0000.0 0000.0- 0000.0- 0000.0- 0-00000000 00000 0000 00:00 00000000 000000 00
0000.0 0000.0- 0000.0- 0000.0 0-00000000 00000 0000 00000 00000000 000000 00
0000.0 0000.0 0000.0- 0000.0 0-00000000 02000 0000 00000 00000000 000000 00
0000.0- 0000.0- 0000.0- 0000.0 0-00000000 00000 0000 00000 00000000 000000 00
0000.0- 0000.0- 0000.0- 0000.0- 0-00000000 00000 0000 00000 00000000 000000 00
0000.0 0000.0- 0000.0. 0000.0- 0-00000000 00000 0000 00:00 00000000 000000 00
0000.0- 0000.0 0000.0- 0000.0- 0-00000000 00000 0000 00000 00000000 000000 00
0000.0- 0000.0- 0000.0- 0000.0- 0-00000000 00000 0000 00000 00000000 000000 00
0000.0- 0000.0- 0000.0- 0000.0- 00000 0000 00:00 00000000 000000 00
0000.0 0000.0- 0000.0- 0000.0 0-00000000 00000 0000 00:00 00000000 000000 00
0000.0- 0000.0- 0000.0- 0000.0- 0-00000000 00000 0000 00000 00000000 000000 00
0000.0- 0000.0- 0000.0- 0000.0- 0-00000000 00000 0000 00:00 00000000 000000 00
0000.0 0000.0- 0000.0- 0000.0- 0-00200000 00000 0000 00:00 00000000 000000 00
0000.0 0000.0- 0000.0- 0000.0 0-00000000 00000 0000 00000 00000000 000000 00
0000.0- 0000.0 0000.0. 0000.0- 00000 0000 00000 00000000 000000 00
0000.0- 0000.0 0000.0- 0000.0- 0000000000 000000 0000 00000 00000000 000000 00
0000.0 0000.0- 0000.0 0000.0. 00000 0000 000000 0000 00000 00000000 000000 00
0000.0 0000.0 0000.0 0000.0 000000 0000 .000000 000000000 000000 00
0000.0 0000.0 0000.0- 0000.0- 000000 0000 .000000 00000 000000 00
0000.0 0000.0- 0000.0 0000.0- 00000 0000 000000 0000 .000 00000000 .00000 00000000 000000 00
0000.0- 0000.0 0000.0 0000.0 000000 0000 .000000 000000000 000000 00
00000 .0.0
000 000 000 000 00000000 0000 0000 0000 .0000 0000000

270

MAHALANOBIS DISTANCE CALCULATION AND POSTERIOR CLASSIFICATION
FOR REFERENCE GROUPS, BASED ON FIRST FOUR PRINCIPAL COMPONENTS.

TABLE 75:

Groups are:

1.000 PCTV (Thunder Bay/Voyageurs)
2.000 PCISR (Isle Royale)

3.000 PCSAND (Sand Point)

4.000 PCJUNT (Juntunen)

Variables used are:

pc01

pc02

Probabilities are jackknifed for specimens included in each group.

The following specimens are in the file PCTV

Probabilities:
ID. NO. PCTV PCISR PCSAND PCJUNT From: Into:
IRP004 99.636 0.004 0.092 14.354 1 1
IRP009 22.249 0.013 0.041 4.969 1 1
IRP011 17.607 0.932 0.043 6.327 1 1
IRP012 14.681 0.048 0.220 7.796 1 1
IRP014 59.565 0.001 0.064 11.719 1 1
IRP018 91.220 0.070 0.053 18.108 1 1
IRP023 32.134 0.017 0.032 19.652 1 1
IRP028 25.587 0.000 0.150 10.242 1 1
IRP032 69.089 0.003 0.125 12.703 1 1
IRP045 25.999 0.002 0.112 5.934 1 1
IRP064 71.041 0.063 0.043 10.317 1 1
The following specimens are in the file PCISR

Probabilities:
ID. NO. PCTV PCISR PCSAND PCJUNT From: Into:
IRP001 0.001 14.697 0.141 1.851 2 2
IRP002 0.000 7.294 0.394 0.001 2 2
IRP022 0.002 59.072 0.245 2.572 2 2
IRP035 0.001 54.352 0.107 0.649 2 2
IRP036 0.000 64.624 0.633 0.066 2 2
IRP037 0.000 56.992 0.826 0.075 2 2
IRP043 0.001 51.705 1.139 0.754 2 2
IRP044 0.005 41.579 0.388 4.782 2 2
IRP048 0.013 75.442 0.026 15.221 2 2
IRPO49 0.001 58.580 0.038 0.435 2 2
IRP050 0.011 30.399 0.046 11.027 2 2
IRP053 0.007 10.481 0.009 5.243 2 2
IRPOSS 0.010 37.121 0.011 13.782 2 2
IRP059 0.000 64.955 0.109 0.329 2 2
IRP063 0.003 78.919 0.094 4.506 2 2
IRP095 0.001 69.659 0.042 0.220 2 2
The following specimens are in the file PCSAND

Probabilities:
ID. NO. PCTV PCISR PCSAND PCJUNT From: Into:
IRP084 0.000 0.192 91.064 0.186 3 3
IRP085 0.000 0.012 65.846 0.001 3 3
IRP086 0.000 0.000 37.121 0.000 3 3
IRP087 0.013 0.001 25.628 0.233 3 3
IRP088 0.001 0.012 77.958 0.060 3 3
IRP089 0.000 0.000 19.531 0.000 3 3
IRP090 0.024 0.000 16.018 0.133 3 3
IRP093 0.000 0.055 73.666 0.001 3 3
IRP094 0.012 0.000 40.413 0.310 3 3

271

The following specimens are in the file PCJUNT
Probabilities:

ID . NO . PCTV PCISR PCSAND PCJUNT From : Into
IRP033 1.050 0.038 0.043 59.803 4
IRP034 1.605 1.272 0.102 50.869 4
IRP072 0.030 5.128 0.013 26.878 4
IRPO73 4.490 0.642 0.040 54.446 4
IRP077 0.164 0.125 0.219 40.360 4
IRP083 0.030 2.889 0.030 85.772 4
IRP106 0.067 33.340 0.014 62.267 4
IRP107 0.431 1.001 0.036 86.173 4
IRP108 0.151 3.460 0.013 54.146 4
IRP067 0.004 0.348 0.025 37.804 4
IRP079 0.020 0.265 0.109 39.408 4
IRP066 0.224 0.000 0.348 14.182 4
IRP070 0.004 0.004 0.108 10.321 4
IRP076 0.032 0.022 1.330 27.176 4
IRP078 0.027 0.001 1.672 48.387 4
IRP081 0.018 0.001 1.128 42.093 4
IRP082 0.025 0.034 1.286 63.866 4
IRP069 0.003 0.005 0.092 17.207 4
IRP074 0.023 0.034 0.036 59.812 4
IRP030 0.005 0.319 0.046 48.034 4
IRP068 0.011 1.339 0.217 35.088 4
IRP075 0.022 0.590 0.122 51.361 4

hhhfibbhhhhfihbbéhb-fib-Ahh“

272

MAHALANOBIS DISTANCE CALCULATION AND POSTERIOR CLASSIFICATION
FOR 5 REFERENCE GROUPS, BASED ON FIRST THREE PRINCIPAL COMPONENTS.

TABLE 76:

Groups are:

1.000 PCTV (Thunder Bay/Voyageurs)
2.000 PCISR (Isle Royale)

3.000 PCKEE (Keweenaw)

4.000 pcsann (Sand Point)

5.000 PCJUNT (Juntunen)

Variables used are: pc01 pc02 pc03

Probabilities are jackknifed for specimens included in each group.

The following specimens are in the file PCTV

Probabilities:
ID. NO. PCTV PCISR PCKEW PCSAND PCJUNT From: Into:
IRP004 98.005 0.001 38.401 0.075 16.189 1 1
IRP009 51.460 0.002 27.515 0.025 13.545 1 1
IRP011 8.695 0.630 20.555 0.018 46.507 1 5
IRP012 7.700 0.015 22.633 0.119 44.710 1 5
IRP014 40.763 0.000 35.596 0.067 7.451 1 1
IRP018 92.850 0.018 30.976 0.031 33.677 1 1
IRP023 38.144 0.004 23.351 0.020 19.678 1 1
IRP028 36.846 0.000 39.340 0.238 4.820 1 3
IRP032 49.130 0.001 28.185 0.118 13.082 1 1
IRP045 60.584 0.001 45.125 0.082 12.476 1 1
IRP064 51.433 0.015 26.999 0.022 30.877 1 1
The following specimens are in the file PCISR

Probabilities:
ID. NO. PCTV PCISR PCKEW PCSAND PCJUNT From: Into:
IRP001 0.001 8.326 1.536 0.149 1.351 2 2
IRP002 0.000 4.344 0.537 4.189 0.001 2 2
IRP022 0.001 42.992 1.607 0.346 1.049 2 2
IRP035 0.000 61.151 1.580 0.145 0.256 2 2
IRP036 0.000 69.519 0.788 2.523 0.033 2 2
IRP037 0.000 71.143 0.760 2.103 0.021 2 2
IRP043 0.000 34.252 0.993 2.133 0.251 2 2
IRP044 0.001 44.116 1.661 0.330 2.429 2 2
IRP048 0.106 62.077 6.110 0.019 13.403 2 2
IRP049 0.001 40.475 2.230 0.048 0.422 2 2
IRP050 0.012 89.263 3.549 0.029 5.164 2 2
IRP053 2.530 23.507 11.711 0.006 20.059 2 2
IRP055 0.622 30.236 9.114 0.006 13.229 2 2
IRP059 0.000 46.076 1.319 0.151 0.272 2 2
IRP063 0.003 85.541 2.211 0.101 3.433 2 2
IRP095 0.001 56.438 1.858 0.053 0.440 2 2
The following specimens are in the PCKEW

Probabilities:
ID. NO. PCTV PCISR PCKEW PCSAND PCJUNT From: Into:
IRP098 0.001 0.000 14.073 0.483 0.328 3 3
IRP099 0.000 0.000 33.850 2.244 0.039 3 3
IRP101 0.017 0.000 15.434 2.316 0.232 3 3
IRP103 0.015 0.000 87.764 0.434 2.672 3 3
IRP104 0.069 0.000 84.704 0.304 4.868 3 3

The following specimens are

ID.
IRP084
IRPOBS
IRP086
IRP087
IRP088
IRP089
IRPO9O
IRP093
IRPO94

The following specimens are

ID.
IRP033
IRP034
IRPO72
IRPO73
IRPO77
IRP083
IRP106
IRP107
IRP108
IRP067
IRPO79
IRP066
IRPO7O
IRPO76
IRPO78
IRP081
IRP082
IRP069
IRPO74
IRPOBO
IRP068
IRPO7S

NO.

NO.

Probabilities:
PCTV PCISR
0.000 0.103
0.000 0.008
0.000 0.000
0.002 0.000
0.000 0.004
0.000 0.000
0.004 0.000
0.000 0.099
0.002 0.000

Probabilities:
PCTV PCISR

59.555 0.058
4.414 0.672
0.033 2.157

28.965 0.284
0.356 0.275
0.338 1.263
2.249 20.880
3.857 0.801
6.683 2.126
0.165 0.286
0.597 2.647
0.200 0.002
0.302 0.095
0.042 0.011
0.130 0.009
0.048 0.003
0.053 0.077
0.008 0.002
0.112 0.196
0.150 0.092
0.028 0.446
0.112 0.171

273

in the file PCSAND
PCKEW PCSAND PCJUNT
0.487 78.279 0.054
0.894 44.656 0.052
0.115 30.883 0.000
1.156 50.066 0.585
0.775 62.461 0.060
0.091 10.613 0.000
12.233 6.243 0.043
0.711 84.429 0.038
1.465 84.435 0.384
in the file PCJUNT
PCKEW PCSAND PCJUNT
30.474 0.034 45.006
10.299 0.068 52.098
7.479 0.004 25.211
26.203 0.021 61.267
5.141 0.251 25.535
14.703 0.011 75.141
12.729 0.006 44.651
23.784 0.017 79.868
13.577 0.007 41.085
12.141 0.010 56.040
5.651 0.101 70.625
43.189 0.235 7.998
9.563 0.078 91.735
2.913 0.455 47.229
3.933 1.232 36.117
5.547 0.495 26.357
2.637 0.828 45.938
10.924 0.024 10.368
21.987 0.018 52.738
5.899 0.017 34.854
2.705 0.100 23.084
4.284 0.044 49.893

"Q

bhhbbhhbhfl
.8

From:

mmmmmmmmmmmmmmmmmmmmmm

H

huhtdhhbbb$5

I

mmmmwmmmmmwmmmmmmmwmmwa

r?
O

ff
0

274

m0~0.0
00H0.0n
”~00.0a
m0~0.0
mcv~.0-
~0v0.0u
~000.0
00m0.0n
~0-.0
5000.0
0~0~.0
00-.0o
0-v.0
ven~.0:
0050.0.
0000.0
~H0~.01
00~0.0u
n000.0
0Hua

~0~0.0
00~0.0
~m00.0
nono.0-
0fi00.0
0000.0
~0~0.0u
“~00.0u
v-0.0
00~0.0
0000.0:
~0~0.0
0~00.0o
000~.0u
0m~0.0u
0-0.0
"000.0
c0~m.0u
0000.0
mfiua

0050.0
m~0¢.0u
0000.0-
vvco.0
-0~.0
~000.0
"000.0-
00~m.0
00~m.0
~000.0
m~0~.00
~0m~.0-
0000.0-
n000.0
mvm~.o
0~0~.0u
0~no.0:
0c~v.0u
n0~v.0u
naua

~00~.0
s~0n.0
0m00.0o
0m00.0u
"000.0-
Kan—.0
00m~.0a
0000.0
~v~m.0n
000~.0u
-00.0u
~000.0
~m~0.0u
v00~.0
000n.0
00n0.0u
0~00.0
50"“.0o
-00.0u
0~ua

0~00.0u
~0m~.0a
0-0.0u
~0c~.0
0050.0-
000~.0
0~0~.0u
~m~0.0
0000.0
~00~.0
0000.0
0050.0-
v~0~.0a
nn-.0:
«~00.0n
-00.ou
0000.0-
c~0~.0
0~n~.0
muua

0¢0~.0
~0H0.0
vv~0.0n
n~m0.0
00am.0
m~00.0n
0000.0:
00-.0
00-.0c
0000.0
~¢0~.0
00~0.0u
«00n.0n
0~0o.0u
nmoo.0
00n~.00
00n¢.0o
00-.0s
0000.0
vfiua

0a0~.0u
OM00.0o
c~0~.0u
00v~.0.
~00M.0o
0000.0
-m~.0
-~n.0
000fl.0
msmfi.0u
mns~.0u
000~.0
0000.0
“050.0
00H0.0n
gnu—.0-
0~00.0
0~n~.0u
0~0~.0
0H0;

0000.0
nn0~.0u
0~00.0n
0m0~.0u
0m—~.0u
wfivm.0u
«0m~.0
v0-.0
~0~N.0n
00~0.0u
000~.0
000~.0u
000~.0
m00~.0n
-0~.0u
38.0..
«mam.0
n~m0.0
0000.0
«Hun

n0n~.0n
~mv~.0
~0-.0
0~00.0
mvflo.0
00m0.0-
nm~0.0c
oscn.0
Nvmo.0
~00~.0
0000.0
«000.0
0-0.0
-n0.0u
n~nn.0u
00~n.0u
m~m0.0o
0000.0-
0-~.0u
"~00

~00~.0
0000.0-
m~m0.0
vo~0.0u
000H.0-
0000.0-
~0~0.0
00-.0
000H.0o
00“".0-
~500.0u
mon~.0
~m0~.0u
0«00.0
~000.0
-me.0
000~.0n
0m~m.0
n0~0.0u
onua

000n.0
m0-.0
0m0~.0
somn.0
00-.0
00-.0
0~00.0
0m-.0
~0~_.0a
0~0~.0
0000.0:
m00m.0u
Hvoa.0u
000~.0
~¢0c.0u
00-.0
~00H.0:
000~.0
000n.0n
0000

000~.0a
~00~.0-
cam—.0-
0~00.0
0000.0
~00~.0
n«00.0-
n0n~.0o
00~m.0u
0-m.0
50mm.0
mac—.0
nmmfl.o
00-.0
000~.0n
00~0.0c
00~—.0
0m00.0n
H00o.0n
000a

0~0e.0
0000.0-
00H0.0c
000~.0
0~0~.0-
~0m~.0
«000.0
0~00.0c
~0-.0a
'0-.0
0000.0:
n0v~.0
0-0.0n
mow—.0:
00n0.0u
00-.0a
5000.0-
000~.0u
n0-.0
noun

605390 35...; 0o 003595.. 2.- 332.505

5.55. consigouéucatg so 938 n.nn—22

£52 3:... 05 co van-o “3322 3559-8 3205....

.00000 002000000 20200200 «cu m~m>4<z< mpzwzoazou 4<m~uz~¢a

000~.0
~000.0
~vv0.o-
0~0~.0a
~00~.0
000~.0-
-00.0n
0000.0-
0~0~.0
cmmm.0o
000m.0u
0m00.0u
0~No.0
~c-.0n
-v~.0o
00-.0
~¢¢0.01
«~00.0
-~0.0u
0000

c00~.0u
~00~.0n
~00v.0
-0~.0u
vm0_.0
~00v.0
5000.0:
00¢".0
~000.0-
0m00.0
~ov~.0s
00~H.0u
0m-.0
n~c~.0c
0~m~.0
-n0.0
n~n~.0
0000.0:
vfiwn.0
000a

0~00.0u
0m~0.0-
0050.0
~000.0
0~00.0-
000~.0n
n-0.0o
~000.0
-0n.0u
~000.0o
000~.0u
000~.0u
00~0.0u
00~0.0o
«000.0-
0000.0
0wm0.0
0000.0-
0~00.0
000a

0000.0
~000.0u
0NHN.0
m0~0.0n
ns0—.0o
00~v.0u
m0-.0o
-00.0o
00¢~.0
v00~.0
0~00.0
v~c~.0
0—00.0
0—sn.0
«~00.0o
-v0.0n
~000.0
0000.0:
mm~0.0
noun

000~.o
00~0.0
n¢0~.0
0000.0
0000.0
00-.0
00~0.0n
n0~0.0
~v-.0
"mmc.oo
00¢~.0
~00m.0.
000".0
0000.0
0000.0
0000.0:
00¢".0
0000.0-
0m0~.0
Noun

~00~.0
0~N~.0
0m0~.0u
00-.0
~0n~.0
0~u~.0u
00~0.0
n~0~.0
'00—.0
0m~0.0
0~0~.0o
m00~.0
00-.0
n-~.o
0~v~.0
n~0~.0
00-.0
¢0v~.0
0~0~.0
"com

>
N»
(z
2:
>0
<0
:»
m»
00
mm
ma
um
:0
no
00
a»
:0
:0
<4

"33:13 3 «30.33 3:28.30 3305....

-00.0 0000.0
0000.0 0000.0
m0~0.0 ”000.0
0000.0 "000.0
s~00.0 ~000.0
050~.0 0000.0
0_v~.0 0000.0
0000.0 -00.0
~0~0.0 0~00.0
n0oo.0 -00.0
n00m.~ vn00.0
0000." ”000.0
m0mo.N 0000.0
0¢~0.0 ~000.0
0500.0 -fi0.0
0000.0 ¢0~0.0
vvvo.- 00~0.0
mm~0.0~ mono.0
-~H.00 ~0-.0

“sh wam<p

275

TABLE 78: HAHALANOBIS DISTANCE CALCULATION AND POSTERIOR CLASSIFICATION FOR THO OR MORE
GROUPS.
Reference groups and masters of specimens:
1 PCTV 11 (Thunder Bay/Voyageurs)
2 PCISR 16 (Isle Royale)
3 PCSAND 9 (Sand Point)
4 PCJUNT 22 (Juntunen)

Variables used:
pc01 pc02 pc03 pc04

The following specimens are in the file PCOUTS (Thirty-seven outliers)

Probabilities:
10. IO. PCTV PCISI PCSAND PCJUNT BEST GP.
1RP003 0.000 0.035 0.064 0.000 3
IRPOOS 0.228 0.035 0.272 0.127 3
1RP006 0.999 0.003 0.488 5.378 4
1RP007 0.711 0.034 0.037 0.282 1
IRPOOB 0.599 10.481 0.026 77.030 4
1RP010 0.761 0.230 0.104 0.732 1
1RP013 0.338 0.124 0.024 1.726 4
1RP015 0.001 0.000 0.897 0.334 3
1RP016 0.322 0.018 0.145 2.774 4
1RP017 0.000 0.000 0.058 0.000 3
IRP024 0.006 6.143 0.310 10.335 4
1RP025 0.479 0.002 0.208 8.154 4
IRP026 0.002 0.002 0.011 0.097 4
1RP027 0.106 0.782 0.130 36.674 4
1RP029 2.377 0.064 0.155 44.573 4
1RP031 0.002 1.484 0.209 0.615 2
IRP042 0.257 0.000 0.466 0.304 3
IRP046 0.207 6.898 0.065 20.341 4
1RP047 0.004 3.875 0.009 25.432 4
1RP051 0.010 0.065 0.060 12.764 4
IRPOSZ 0.046 0.138 0.004 2.638 4
IRP054 0.000 0.000 0.015 0.001 3
IRP056 0.534 0.002 0.026 21.053 4
IRP057 0.595 0.037 0.063 15.146 4
1RP058 0.237 4.727 0.291 23.220 4
1RP061 0.065 0.000 0.046 7.045 4
IRP062 0.000 0.288 0.216 0.052 2
IRP065 0.001 0.000 0.064 0.118 4
1RP071 0.001 0.001 0.043 1.574 4
1RPO80 0.016 0.007 35.558 6.105 3
1RP091 0.080 0.018 0.179 10.868 4
IRP092 0.208 0.021 3.983 0.520 3
IRP096 3.158 0.014 1.374 15.526 4
IRP097 0.150 0.003 31.779 1.039 3
IRPIOO 0.660 0.002 0.279 1.652 4
1RP102 16.321 0.013 0.016 12.190 1
IRP105 0.001 0.000 11.792 0.003 3

Summary of Probabilities for Specimens in the file PCOUTS

Probability Cutoff Values:

Group.

------- 0.01000 0.10000 1.00000 5.00000 10.00000 20.00000 100.00000
PCTV 12 5 17 2 0 1 0
PCISR 15 11 5 3 2 1 0
PCSAND 2 15 15 2 0 1 2
PCJUNT 4 2 8 6 4 6 7

Sumary of Best Classification of Projected Specimens:
Classified Into Group:
PCTV PCISR PCSAND PCJUNT Total

From Group:
PCOUTS 3 2 10 22 37
Total 3 2 10 22 37

276

.0. 05 E. 3:22:50 3.10:5 ”SE 2806

20 0:039:00 6305...;

m;

.vé F
_ _ _ F

0.0

 

 

 

5:30:30

\mcomoxo>\>om Longs...

 

 

4

 

Eon Bum

   
 

r. N0
1 To
1 0.0
l 0.0

T N;
r. 0;
l 0.?
T w;
T N

I N.N
l .v.N
T 0d
1 0d

 

 

.2 2:5

z# iusuodwoo lodpupd

277

N.N

.30.. one 03. 3:280:80 30655 ”Ho—m 0350mm

0* “cocanoo 690500

N. 0.0 0.0

_ _ _ _ _

.v.0
_

N0

.2 2st0

 

 

meaomo>o>\>om Longs...

0.98m 23

5:30:30

   
    

l 0;

i h;

l 0.—

 

 

04

mi! zuauodwog lodpuyd

278

ON?

00—
_

mocomnofi Log 3.63 :9:

00

P

00

h

.8: Ba 855% ”8: 2.52m

.8 20E

 

 

 

T0

10

lo—
INF
1.:
10F
19
TON
INN
T¢N
T0N
[mm
Ton

 

 

an

(wdd) wngpuoog

279

0.\.

00

8230053 28 8:30.26 ”SE 33.820

AEQQV 83:05:04
on 9

_ _ _ L _

.R 20::

 

 

mcammo>o>
\xom 80:3:

  
     

5:30:30

accomaox

T N.N
T ¢.N
I 0.N
l 0.N

 

(wdd) wngdomg

280
APPENDIX D: CATALOG OF NAA SAMPLES.

Samples are numbered consecutively with 89 prefix (eg, 89-1).

Abbreviations for source of samples:
APIS = Apostle Islands National Lakeshore

GRPO = Grand Portage National Monument

ISRO = Isle Royale National Park

MCC = Ministry of Culture and Communications, Thunder Bay
MTU = Michigan Technological University

MWAC = Midwest Archeological Center, National Park Service
UMMA = University of Michigan, Museum of Anthropology
VOYA = Voyageurs National Park

WMU = Western Michigan University

NAA# Source, Site, Ceramic Type, Catalog #, Other

89-1 MWAC, Chippewa Harbor, Clay sample

89-2 MWAC, Moskey Narrows, Clay sample

89-3 MWAC, Beaver Island, Clay sample

89-4 VOYA, 218L213, Blackduck, MWAC# 605-002
89-5 VOYA, 218L183, Blackduck, MWAC# 492-001
89-6 VOYA, 218L183, Blackduck, MWAC# 467-001
89-7 VOYA, 218L183, Blackduck, MWAC# 534-001
89-8 VOYA, 218L183, Blackduck, MWAC# 601-001
89-9 VOYA, 218L183, Blackduck, MWAC# 526-001
89-10 VOYA, 218L183, Blackduck, MWAC# 603-001
89-11 VOYA, 218L183, Blackduck, MWAC# 505—001
89-12 VOYA, 218L183, Blackduck, MWAC# 499-001
89-13 VOYA, 218L183, Blackduck, MWAC# 001-000
89-14 MCC, Dd Ir-l, Juntunen, MCC# 01

89-15 MCC, De Ip-3, Blackduck, MCC# 05 [alternate]
89-16 MCC, D1 Iq-15, Blackduck, MCC# 14

89-17 MCC, Di Jf-4, Blackduck, MCC# 16

89-18 MCC, De Ip-3, Blackduck, MCC# 06

89-19" MCC, De Iu-3, Selkirk, MCC# 08

89-20" MCC, Mason Collection, Selkirk, MCC# 19

89-21 "‘
89-22
89-23
89—24
89-25
89-26
89-27
89-28
89-29
89-30

89-31
89-32
89-33
89-34
89-35
89-36
89-37
89-38”
89-39"
89-40"

89-41 *
89—42
89-43
89-44
89-45
89-46
89-47
89-48
89-49
89-50

89-51
89-52
89-53
89-54
89-55
89-56
89-57
89-58
89-59
89-60

281

MCC, Di Jf-5, Selkirk, MCC# 17

MCC, Knudson Collection, Blackduck, MCC# 18
MCC, De Ip-3, Blackduck, MCC# 04

MCC, De Ik-3, Blackduck, MCC# 07

MCC, Dk Im-8, Blackduck, MCC# 13

MCC, Db Jx-l, Blackduck, MCC# 15

MCC, De Ip-2, Blackduck, MCC# 03

MCC, De Iu-3, Blackduck, MCC# 10

MCC, De Iu-3, Blackduck, MCC# 12

MCC, De Ik-l, Juntunen, MCC# 02

MWAC, 201R142, Belle Isle #2, Juntunen, MWAC 885-001
MWAC, 201R134, Mott Sauna Beach, Juntunen, MWAC 890-005
MWAC, 201R114, Lone Tree Cove, Juntunen, MWAC 871-001/003
MWAC, 201R114, Lone Tree Cove, Juntunen, MWAC 870—001
MWAC, 201R45, Daisy Farm, Juntunen, MWAC 882-002

MWAC, 201R18, Card Point, Juntunen, MWAC 891-001

MWAC, GRPO, clay sample, vicinity of Great Hall/Kitchen
MWAC, 201le, Duncan Bay #1, Selkirk, MWAC 880—001
MWAC, 201R128, Lane Cove, Selkirk, MWAC 883—001

MWAC, 201R29, Belle Isle #1, Selkirk, MWAC 880-006

MWAC, 201R29, Belle Isle #1, Selkirk, MWAC 880-005
ISRO, 201R111, Rock Hbr. Lighthouse, Blackduck, 102
MWAC, 201R124, Lake Ritchie #1, Blackduck, MWAC 871-1
MWAC, 201R29, Belle Isle #1, Blackduck, MWAC 886-001
MWAC, 201R45, Daisy Farm, Blackduck (?), MWAC 882-003
MWAC, 201R1, Chippewa Hbr. #1, Blackduck, MWAC 872-2
MWAC, 201R127, Chippewa Hbr. #3, Huron, MWAC 870-0
MWAC, 201R147, Brady Cove #1, Huron, MWAC 880-001
MWAC, 201R142, Belle Isle #2, Huron, MWAC 885-001
MWAC, 201R45, Daisy Farm, Huron, MWAC 882-003

MWAC, 201R142, Belle Isle #2, Huron, MWAC 885-001

MWAC, 201R29, Belle Isle #1, Huron, MWAC 887-001

MWAC, 201R17, Grace Island, Huron, UMMA 62136

MWAC, APIS, Outer Island, Clay sample

UMMA, 201R52, Baker Point, Lalonde High Collar, UMMA 62237
UMMA, 201R1, Chip. Hbr., Juntunen, UMMA 52477 (20a)
UMMA, 201R72, Cove, Blackduck, UMMA 62363 (16b)

UMMA, 201R72, Cove, Blackduck, UMMA 62363 (not illust)
UMMA, 201R1, Chip. Hbr, Juntunen, UMMA 5329(130)

UMMA, 201R], Chip. Hbr., Juntunen/Ont.0blique UMMA 5329
(26b)

89-61
89-62
89-63
89-64
89-65
89-66
89-67
89-68
89-69
89-70

89-71
89-72
89-73
89-74
89-75
89-76
89-77
89-78
89-79
89-80

89-81
89-82
89-83
89-84
89-85
89-86
89-87
89-88
89-89
89-90

89-91
89-92
89-93
89-94
89-95
89-96
89-97
89-98
89-99

89-100

282

UMMA, 201R28, Indian Pt., Huron, UMMA 5325 (29b)
UMMA, 201R72, Cove, Blackduck, UMMA 39588/89 (17a)
UMMA, 201R1,Juntunen, UMMA 52477 (180)

UMMA, 201R1, Huron, UMMA 5329 (30a)

UMMA, 20MK1, Juntunen, 62947 (McPherron 1967zPlate XXIb)
UMMA, 20MK1, Juntunen, 41754 (XXd)

UMMA, 20MK1, Juntunen, 40687 (not illust)

UMMA, 20MK1, Juntunen, ?????tape #21 (not illust)

UMMA, 20MK1, Juntunen, 41594 (XXIa)

UMMA, 20MK1, Juntunen, 41022 (not illust?)

UMMA, 20MK1, Juntunen, ?????

UMMA, 20MK1, Juntunen, 41501 (122) (not illust?)
UMMA, 20MK1, Juntunen, 41531 (not illust?)

UMMA, 20MK1, Juntunen, 40962 [Fea. 1] McP Vessel 12)
UMMA, 20MK1, Mackinac, 41580

UMMA, 20MK1, Mackinac, 41590

UMMA, 20MK1, Mackinac, 41670 (XIIIa)

UMMA, 20MK1, Mackinac, 62978

UMMA, 20MK1, Mackinac, 41071 (XXIg)

UMMA, 20MK1, Mackinac, 62834

UMMA, 20MK1, Mackinac, 41588 (XIIa)
UMMA, 20MK1, Mackinac, ????? (painted #61)
UMMA, 20MK1, Mackinac, 41795? (XIVa?)
UMMA, 20MK1, Mackinac, 41582 (XIVb)
WMU, 20BGl4, Sand Point, SP-1254 (VIII:1)
WMU, 2OBGl4, Sand Point, SP-2014

WMU, 20BGl4, Sand Point, SP15-754

WMU, 20BGl4, Sand Point, SP1-1749

WMU, 208614, Sand Point, SP1-1004

WMU, 208614, Sand Point, no number

WMU, 20BGl4, Sand Point, SP1-2015

WMU, 20BGl4, Sand Point, SP1-682/1960/1705(VI:1)
WMU, 20BGl4, Sand Point, SP15-817/109

WMU, ZOBGl4, Sand Point, SP1-1329 (VI:5)

ISRO, 201R41, Juntunen (submerged)

MTU, 20KE15, Oneota, Williams Collection

MTU, 20KE15, Oneota, Williams Collection

MTU, 20KE15, Oneota, Williams Collection

MTU, 20KE15, Oneota, MTU 80-22-1

MTU, 20KE15, Oneota, MTU 81-1-1

89—101
89-102
89-103
89-104
89-105
89-106
89-107
89-108
89-109
89-110

283

MTU, 20KE15, Oneota, MTU 81-1-1

MTU, 20KE15, Oneota, MTU 81-1-1

MTU, 20KE15, Oneota, MTU 81-1-1

MTU, 20KE15, Oneota, Williams Collection
MTU, 20KE15, Oneota, Williams Collection
MTU, 20KE15, Juntunen, MTU 80-22-2

MTU, 20KE15, Juntunen, MTU 80-22-7

MTU, 20KE15, Sand Point, Williams Collection

'=sample taken but not submitted

NAA Sample, Listing by Ceramic Type:

BLACKDUCK WARE
Voyageurs National Park, MN

89-4
89-5
89—6
89-7
89-8
89-9
89-10
89-1 1
89-12
89-13

VOYA, 21$L23, Blackduck, MWAC# 605-002
VOYA, 218L23, Blackduck, MWAC# 492—001
VOYA, 21$L23, Blackduck, MWAC# 467-001
VOYA, 218L23, Blackduck, MWAC# 534-001
VOYA, 218L23, Blackduck, MWAC# 601-001
VOYA, 2181.23, Blackduck, MWAC# 526-001
VOYA, 218L23, Blackduck, MWAC# 603-001
VOYA, 21$L23, Blackduck, MWAC# 505-001
VOYA, 218L23, Blackduck, MWAC# 499-001
VOYA, 218L23, Blackduck, MWAC# 001-000

North Shore Lake Superior and BWCA, ON

89-15
89-16
89-17
89-18
89-22
89-23
89-24
89-25
89-26
89-27
89-28
89-29

MCC, De Ip-3, Blackduck, MCC# 05 [alternate]
MCC, Dl Iq-15, Blackduck, MCC# 14

MCC, Di J M, Blackduck, MCC# 16

MCC, De Ip—3, Blackduck, MCC# 06

MCC, Knudson Collection, Blackduck, MCC# 18
MCC, De Ip-3, Blackduck, MCC# 04

MCC, De Ik-3, Blackduck, MCC# 07

MCC, Dk Im-8, Blackduck, MCC# 13

MCC, Db Jx-l, Blackduck, MCC# 15

MCC, De Ip-2, Blackduck, MCC# 03

MCC, De Iu-3, Blackduck, MCC# 10

MCC, De Iu-3, Blackduck, MCC# 12

284

Isle Royale National Park, MI

89-42
89-43
89-44
89-45
89-46
89-57
89-58
89-62

ISRO, 201R111, Rock Hbr. Lighthouse, Blackduck, 102
MWAC, 201R124, Lake Ritchie #1, Blackduck, MWAC 871-1
MWAC, 201R29, Belle Isle #1, Blackduck, MWAC 886-001
MWAC, 201R45, Daisy Farm, Blackduck (?), MWAC 882-003
MWAC, 201R1, Chippewa Hbr. #1, Blackduck, MWAC 872-2
UMMA, 201R72, Cove, Blackduck, UMMA 62363 (16b)
UMMA, 201R72, Cove, Blackduck, UMMA 62363 (not illust)
UMMA, 201R72, Cove, Blackduck, UMMA 39588/89 (17a)

ONTARIO IROQUOIS TRADITION WARES
Isle Royale National Park, MI

89-47
89-48
89-49
89-50
89-51
89-52
89-53
89-55
89-61
89-64

MWAC, 201R127, Chippewa Hbr. #3, Huron, MWAC 870-0
MWAC, 201R147, Brady Cove #1, Huron, MWAC 880-001
MWAC, 2OIR142, Belle Isle #2, Huron, MWAC 885-001

MWAC, 201R45, Daisy Farm, Huron, MWAC 882—003

MWAC, 201R142, Belle Isle #2, Huron, MWAC 885-001

MWAC, 201R29, Belle Isle #1, Huron, MWAC 887-001

MWAC, 201R17, Grace Island, Huron, UMMA 62136

UMMA, 201R52, Baker Point, Lalonde High Collar, UMMA 62237
UMMA, 201R28, Indian Pt., Huron, UMMA 5325 (29b)

UMMA, 201R1, Huron, UMMA 5329 (30a)

J UNTUNEN WARE
North Shore Lake Superior, ON

89-14
89-30

MCC, Dd Ir-l, Juntunen, MCC# 01
MCC, De Ik-l, Juntunen, MCC# 02

Isle Royale National Park, MI

89-31
89-32
89-33
89-34
89-35
89-36
89-56
89-59
89-60
89-95

MWAC, 201R142, Belle Isle #2, Juntunen, MWAC 885—001
MWAC, 201R134, Mott Sauna Beach, Juntunen, MWAC 890-005
MWAC, 201R114, Lone Tree Cove, Juntunen, MWAC 871-001/003
MWAC, 201R114, Lone Tree Cove, Juntunen, MWAC 870—001
MWAC, 201R45, Daisy Farm, Juntunen, MWAC 882-002

MWAC, 201R18, Card Point, Juntunen, MWAC 891—001

UMMA, 201R1, Juntunen, UMMA 52477 (20a)

UMMA, 201R1, Juntunen, UMMA 5329(130)

UMMA, 201R], Juntunen/Ont.0blique UMMA 5329 (26b)

ISRO, 201R41, Juntunen (submerged)

285

Straits of Mackinac, MI

89-65 UMMA, 20MK1, Juntunen, 62947 (McPherron 1967:Plate XXIb)
89-66 UMMA, 20MK1, Juntunen, 41754 (XXd)

89-67 UMMA, 20MK1, Juntunen, 40687 (not illust)

89-68 UMMA, 20MK1, Juntunen, ?????tape #21 (not illust)
89-69 UMMA, 20MK1, Juntunen, 41594 (XXIa)

89-70 UMMA, 20MK1, Juntunen, 41022 (not illust?)
89-71 UMMA, 20MK1, Juntunen, ?????

89-72 UMMA, 20MK1, Juntunen, 41501 (122) (not illust?)
89-73 UMMA, 20MK1, Juntunen, 41531 (not illust?)
89-74 UMMA, 20MK1, Juntunen, 40962 [Fea. 1] McP Vessel 12)
Keweenaw Peninsula, MI

89-106 MTU, 20KE15, Juntunen, MTU 80-22-2

89-107 MTU, 20KE15, Juntunen, MTU 80-22-7
MACKINAC WARE

Straits of Mackinac, MI

89-75 UMMA, 20MK1, Mackinac, 41580

89-76 UMMA, 20MK1, Mackinac, 41590

89-77 UMMA, 20MK1, Mackinac, 41670 (XIIIa)

89-78 UMMA, 20MK1, Mackinac, 62978

89-79 UMMA, 20MK1, Mackinac, 41071 (XXIg)

89-80 UMMA, 20MK1, Mackinac, 62834

89-81 UMMA, 20MK1, Mackinac, 41588 (XIIa)

89-82 UMMA, 20MK1, Mackinac, ????? (painted #61)
89-83 UMMA, 20MK1, Mackinac, 41795? (XIVa?)

89-84 UMMA, 20MK1, Mackinac, 41582 (XIVb)
ONEOTA

Keweenaw Peninsula, MI

89-96 MTU, 20KE15, Oneota, Williams Collection

89-97 MTU, 20KE15, Oneota, Williams Collection

89-98 MTU, 20KE15, Oneota, Williams Collection

89-99 MTU, 20KE15, Oneota, MTU 80-22-1

89-100 MTU, 20KE15, Oneota, MTU 81-1-1

89-101 MTU, 20KE15, Oneota, MTU 81-1-1

89-102 MTU, 20KE15, Oneota, MTU 81-1-1

89-103 MTU, 20KE15, Oneota, MTU 81-1-1

89-104 MTU, 20KE15, Oneota, Williams Collection

89-105 MTU, 20KE15, Oneota, Williams Collection

286

SAND POINT WARE
Keweenaw Peninsula, MI

89-85 WMU, 208614, Sand Point, SP-1254 (VIII:1)
89-86 WMU, 208614, Sand Point, SP-2014

89-87 WMU, 208614, Sand Point, SP15-754

89-88 WMU, 208614, Sand Point, SP1-1749

89-89 WMU, 208614, Sand Point, SP1-1004

89-90 WMU, 208614, Sand Point, no number

89-91 WMU, 208614, Sand Point, SP1-2015

89-92 WMU, 208614, Sand Point, SP1-682/1960/1705(VI:1)
89-93 WMU, 208614, Sand Point, SP15-817/109
89-94 WMU, 208614, Sand Point, SP1-1329 (VI:5)
89-108 MTU, 20KEIS, Sand Point, Williams Collection

SELKIRK (no samples of Selkerk were submitted for NAA)
North Shore Lake Superior, ON

89-19 MCC, De Iu-3, Selkirk, MCC# 08

89-20 MCC, Mason Collection, Selkirk, MCC# 19

89-21 MCC, Di Jf-5, Selkirk, MCC# 17

Isle Royale National Park, MI

89-38 MWAC, 201R140, Duncan Bay #1, Selkirk, MWAC 880-001
89-39 MWAC, 201R128, Lane Cove, Selkirk, MWAC 883-001
89-40 MWAC, 201R29, Belle Isle #1, Selkirk, MWAC 880-006
89-41 MWAC, 201R29, Belle Isle #1, Selkirk, MWAC 880-005
CLAY SAMPLES

89-1 Chippewa Harbor, ISRO87

89-2 Moskey Narrows, ISRO88

89-3 Beaver Island, ISR089

89-37 Depot/kitchen area, 6RPO89

89-54 Outer Island APIS88

REFERENCES

287

REFERENCES

Afinson, Scott F. (editor)
1979 A Handbook of Minnesota Prehistoric Ceramics. O c c a s i o n a 1
Publications in Minnesota Anthropology, no. 5. St. Paul.

Arima, Eugene Y.
1984 Caribou Eskimo. In Arctic, edited by David Damas, pp. 447-462.
Handbook of North American Indians, vol. 5, William 6. Sturtevant,
general editor. Smithsonian Institution, Washington, DC.

Arnold, D.E., H. Neff, and R. L. Bishop
1990 Compositional Analysis and "Sources" of Pottery: An
Ethnoarchaeological Approach. American Anthropologist (in press).

Aronson, Dan R.
1976 Ethnicity as a Cultural System: an Introductory Essay. In, Ethnicig in
the Amerigg, edited by Frances Henry, pp. 19, Mouton & Company,
The Hague.

Arthurs, David
1978 Sandy Lake Ware in Northwestern Ontario: A Distributional Study.
Archae-Facts 5(2-3):57-64.

1986 Arghaeologigfl Investigations at the Long Sault Site, a Stratified

Habitation and Burial Station on the RM River in Northwestern
Ontario. Conservation Archaeology Report 7, Northwestern Region,

Ministry of Citizenship and Culture.

Barth, Fredrik, editor

1969 Ethnic Groups and Bogdaiies: the Smial Qrggization of 930mm!
Differenggs. Little, Brown, and Company, Boston.

Bastian, Tyler J.
19633 Archaeological Sruvey of Isle Royale National Park, Michigan, 1960-
1962. Manuscript on file, University of Michigan Museum of
Anthropology and Midwest Archeological Center, Lincoln.

1963b Prehistoric Copper Mining in Isle Royale National Park, Michigan.
Unpublished MA thesis, Department of Anthropology, University of
Utah, Salt Lake City.

288

Baxter, P.T.W.
1977 Boran Age-sets and Warfare. In Warfare Among East African Herders,
edited by Katsuyoshi Fukui and David Turton, pp. 69-95. Senri
Ethnological Studies 3, Osaka.

Beaubien, Paul L.
1957 Notes on the Archaeology of Pipestone National Monument. lh_e

Minnesota Archaeologist 21(3):1-22.

Biggar, H. P. (ed.)
1925 The Publications of the Champlain Socieg, Volume 2, The Works of

Samuel De Champlain. Toronto.

Binford, Lewis R.
1962 Archaeology as Anthropology. American Antiguig 28:217-225.

Birk, Douglas
1977 The Norway Lake Site: A Multicomponent Woodland Complex in North
Central Minnesota. The Minnesota Archaeologist 36:16-45.

Birket-Smith, Kaj
1929 The Caribou Eskimos: Material and Social Life and Their Cultural

Position. Repogt of the Fifth Thule E_xpedition, 1921-1924, Volume 5,
Copenhagen.

Bishop, Ronald L., Veletta Canouts, Patricia Crown, and
Suzanne P. De Atley
1990 Sensitivity, Precision, and Accuracy: Their Roles in Ceramic
Compositional Data Bases. American Antigpity 55(3):537-546.

Bishop, Ronald L., and Hector Neff
1989 Compositional Data Analysis in Archaeology. In Archaeological
Qhemist_ry IV, edited by RC. Allen, pp. 576-586. Advances in
Chemistry Series 220, American Chemical Society, Washington, DC

Bray, Martha C., editor

1970 The Journals of Joseph N. Nicollet, A Scientist on the Mississippi
Headwaters With Notes on Indian Life. Minnesota Historical Society,

St. Paul.

Brizinski, Morris J. and Kenneth T. Buchanan

1977 Ceramig, Chegt Qd @lture; An Anelyeis of Three Prehistoric Sites
meted in the Michipicoten Area. Laurentian University, Sudbury.

289

Brose, David S.
1970 The Archaeology of Summer Island: Changing Settlement Systems in
Noghern fie Michigan. Anthropological Papers No. 41. Museum of
Anthropology, University of Michigan, Ann Arbor.

1978 Late Prehistory of the Upper Great Lakes Area. In Northeast, edited by
Bruce Trigger, pp. 569-582. Handbook of North American Indians, vol.
15. William G., Sturtevant, general editor. Smithsonian Institution,
Washington, D. C.

1990 Toward a Model of Exchange Values for the Eastern Woodlands.
Midcontinental Journal of Archaeology 15(1):]00-136.

Brumbach, Hetty J o
1975 "Iroquoian" Ceramics in "Algonkian" Territory. Mg In the Nertheast
10:17-28.

Brumbach, Hetty Jo, and Susan J. Bender
1986 Winney’s Rift: A Late Woodland Village Site in the Upper Hudson

River Valley. The Bpiletin and Journal of Archaeology for New York

State 92:18.
Buchner, A. P.
1979 The Shield Archaic: A Review. Mgitobe Archaeological Quarterly
3(2):2-11.

Buckmaster, Marla
1979 Wooleand and Oneota Settlement and Subsistence Systems in the
Menominee River Watershed. Unpublished Ph.D. dissertation,
Department of Anthropology, Michigan State University, East Lansing.
University Microfilms International, Ann Arbor.

 

Buckmaster, Marla and James Paquette
1988 The Gorto Site: Preliminary Report on a Late Paleo-Indian Site in
Marquette County, Michigan. The Wisconein Archeolegist 65(2):]59-168.

Burch, Ernest S.
1972 The Caribou/Reindeer as a Human Resource. American Antiguiiy
37(3):339-368.

1978 Caribou Eskimo Origins: An Old Problem Reconsidered. Arctic
Anthropology 15(1):1-35.

290

Casagrande, Joseph, Stephen Thompson, and Phillip Young

1964

Colonization as a Research Frontier. In, Prgy'ess epd Pettern in ngpre:

Essaye in Honor of Jplian H. Steward, edited by Robert Manners, pp.
281-325. Aldine Press, Chicago.

Christian, Sarah

1932

Winter on Isle Royale, A Narrative of Life on Isle Reyale During the
Years of 1874 and 1875. University Microfilms International, Ann Arbor.

Clark, Caven P.

1987

1989a

1989b

1990

n.d.

Archaeological Survey of ihe Rggk Harbor Aree, Isle Royale Natienei
Park, 1986. Cultural Resource Management Report 14. Michigan

Technological University, Houghton.

Analysis of Copper Waste and North Shore Ties in the Terminal
Woodland Substage: The 1988 Isle Royale Archaeological Survey.

Journal ef me Steward Anthrepological Socieiy 18(1-2):151-164.

Plano Tradition Lithics from the Upper Peninsula of Michigan. The
Michigan Archaeologist 35(2):88-112.

A Dog Burial from Isle Royale, Lake Superior: An Example of
Household Ritual Sacrifice in the Terminal Woodland Period.

Mideontinenial Journal of Archaeology 15(2):265-278.

The Isle Royale Archaeological Survey: 1986-1990. Unpublished
manuscript on file at Isle Royale National Park, Houghton and the
Midwest Archeological Center, Lincoln.

Cleland, Charles E.

1966

1971

1985

Analysis of the Fauna of the Indian Point Site on Isle Royale National
Park in Lake Superior. The Miehiggi Arehaeelegisi 14(3-4):143-146.

The Lasnen Site: An Historie Burial Localng in Mackinac Coung,
Michigan. Anthropological Series 1, The Museum, Michigan State

University, East Lansing.
Naub-Cow-Zo-Win Disks and Some Observations on the Origin and

Development of Ojibwa Iconography. Ar t_ie Anthropelogy
22(2):131-140.

Cooper, L. and E. Johnson

1964

Sandy Lake Ware and its Distribution. Amerim Antiguiy
29(4):474-479.

291

Conway, Thor
1977 SlQ-Whitefish Island- A Remarkable Archaeological Site at Sault Ste,

Marie Ontario. Research Manuscript Series, Ministry of Culture and
Communications.

 

Crane, H. R. and James B. Griffin
1965 University of Michigan Radiocarbon Dates X. Radiocarben 7.

Cremin, William C.
1980 Sand Point (208614): A Lakes Phase Site on the Keweenaw Peninsula,
Baraga County, Michigan. In The Sand Point Site (208614), edited by
Margaret Holman and Terrance Martin. The Michigan Archaeologist
26(3-4):7-16.

Dawson, K. C. A.
1974 MeCluskey Site. Mercury Series, Archaeological Survey of Canada,
Paper 25, National Museum of Man, Ottawa.

1978 The Mound Island Site: A Multi-component Woodland Period
Habitation Site in Northwestern Ontario. Ontario My 30.47-66.

 

1980 The MacGillivray Site: A Laurel Tradition Site in Northwestern Ontario.
Qnime Arehaeology 34:45-68.

1983 Cummins Site: A Late Palaeo-Indian (Plano) Site at Thunder Bay,
Ontario. Ontario Archaeology 39:3-31.

1987 The Martin-Bird Site. Qntarie Areheeelegy 47:33-57.

Despres, Leo A.
1975 Toward a Theory of Ethnic Phenomena. In, Ethnicig and Resogce

@mpep’tion in Plprei Sgieties, edited by Leo Despres, pp. 187- 207.
Mouton & Company, The Hague.

Dorothy, Lawrence 6.

1980 The Ceramics of the Sand Point Site (208614), Baraga County,
Michigan: a Preliminary Description. In, The Sand Point Site (208614),
edited by Margaret Holman and Terrance Martin. The Miehigg
Arehaeolegist 26(3-4):39-86.

Drier, Roy W.
1961 The Michigan College of Mining and Technology Isle Royale
Excavations, 1953-1954. In Lake Sumrior Copper ad the Indians,

Miscellaneous Studies of Great Lakes Prehistogz, edited by James B.
Griffin, pp. 1-7. Anthropological Papers 17. Museum of Anthropology,

University of Michigan, Ann Arbor.

292

Drier, Roy W. and Octave DuTemple

1961 Prehieierie @pper Mining in the Lake Superior Region: A Compilation
ef Referenee Articles. Privately published by the authors, Calumet.

Dustin, Fred
1929 Narrative Report of an Archaeological Reconnnaissance of Isle Royale
During July and August, 1929. Unpublished manuscript on file,
University of Michigan, Museum of Anthropology.

1957 An Archaeological Reconnaissance of Isle Royale. Michigan Histeg
41(1):1-34.

Earle, Timothy K., and Jonathon E. Ericson
1977 Exehmige Systems in Prehi§tog. Academic Press, New York.

Elam, J .M. and MD. Glascock
1988 A Re-examination of the Provenance of Eastern Sigillata A. In

Prggeedings of the 26th Internetional Areheeometg Symposium, edited
by RM. Farquar, R.G.V. Hancock, and LA. Pavlish, pp. 179-183.

University of Toronto.

Evans, 6. Edward
1961 ramic An sis of Blackduck Ware end Its General Cultural
Relationships. Proceedings of the Minnesota Academy of Science,
29:33-54. St. Paul.

Evans-Pritchard, BE.
1940 The Nper. Clarendon Press, Oxford.

Farrand, W. R.
1969 The Quaternary History of Lake Superior. International Association of
Great Lakes Research, 12th Conference, Great Lakes Research 1969
Proceedings, pp. 181-197.

Fitting, James E.

1965 Late Woodland @kures of Southeastern Michigan. Anthropological
Papers No.24, Museum of Anthropology, University of Michigan, Ann

Arbor.

1975 The Archaeology of Miehiggp. Second Edition. Cranbrook Institute of
Science, Bloomfield Hills.

Fitzhugh, William
1972 The Eastern Archaic: Commentary and Northern Perspective.

Pennglvania Archaeologist 42(4):1-19,

"L

 

293

Flannery, Kent V.
1971 Archaeological Systems Theory and Early Mesoamerica. In Man’s
Imprint from the Past, edited by James Deetz. Little Brown, Boston.

 

Flannery, Kent V. (editor)
1976 The Early Mesoamerican Village. Academic Press, New York.

Ford, Richard I.

1972 Barter, Gift, or Violence: An Analysis of Tewa Intertribal Exchange. In,
Social Exchange and Interaction, edited by Edwin Wilmsen, pp. 21-45.
Anthr0pological Papers 46, The University of Michigan, Museum of
Anthropology, Ann Arbor.

Foster, Brian L. and Geoffrey M. White
1982 Ethnic Identity and Perceived Distance Between Ethnic Categories.
Hpman Qrgapizatien 41(2):121-130.
Foster, J. W., and J. D. Whitney
1850 Remrt on the Geology and Topography ef a Portion ef the Lake

Sumrior Land District in the State ef Michigan; @pper Lande. House
Executive Documents No 69, Part 1. Washington, DC.

Franzen, John G.
1987 T st Ex v ' n d tional An sis f Prehist ri Sites in e
Hiewatha National Forest, Michigan. Cultural Resource Management
Report No. 5, Hiawatha National Forest, Escanaba.

Fried, Morton
1966 On the Concepts of "Tribe" and "Tribal Society." Transaetions ef the

New York Agdemy ef $ieng 28:527-540.

Fukui, Katsuyoshi
1977 Cattle Colour Symbolism and Intertribal Homicide Among the Bodi. In
Warfare Among East African Herders, edited by Katsuyoshi Fukui and
David Turton, pp. 147-177, Senri Ethnological Ethnological Studies 3,
Osaka.

Fukui, Katsuyoshi and David Turton
1977 Introduction. In Warfare Among East African Herders, edited by
Katsuyoshi Fukui and David Turton, pp. 1-13, Senri Ethnological
Ethnological Studies 3, Osaka.

Gibbon, Guy E. and Christy A. H. Caine
1980 The Middle to Late Woodland Transition in Eastern Minnesota.

Midmntinental Jepmel pf Archaeolegy 5(1):57-72.

294

Gilbert/Commonwealth, Inc. 1987
1986 Cultural Resource Survey, Hiawatha National Forest. Submitted to the
USDA, Forest Service, Milwaukee, Contract No. 53-5380—5-01018.

Glascock, Michael and J .M. Elam
1988 Differentiation of Obsidian Sources in Mesoamerica. In Preeeedinge ef

the 26th Intemation Ar haeom osium edited by RM.
Farquar, R.G.V. Hancock, and LA. Pavlish, pp. 245-251. University of
Toronto.

Gordon, Bryan HQ

1975 Of Men Ed Hergg in Barrenlgpds Prehistory. Paper 28, Mercury Series,
Archaeological Survey of Canada, National Museum of Man, Ottawa.

Gordon, Diana Lynn

1985 North @‘bou Lake Arehaeology, Conservation Archaeology Report,
Northwestern Region, Report 9.

Gould, RA.
1980 Living Archaeology. Cambridge University Press, Cambridge.

Griffin, James 8., editor
1961 . eSuooerir Aurr nd h Indi Mi 1] ‘0 rdi of r.
les Prehistog. Anthropological Papers 17, Museum of Anthropology,
University of Michigan, Ann Arbor.

Griffin, James B.
1961 Early Historical Accounts of the Lake Superior Copper District. In Lake

Sumrior Copper and the Indians, Miscellgpeops Sipdies of Greg; Qiges
Prehistogy, edited by James B. Griffin, pp. 32-45. Anthropological Papers

17, Museum of Anthropology, University of Michigan, Ann Arbor.

Griffin, James B. and George I. Quimby
1961 Prehistoric Copper Pits on the Eastern Side of Lake Superior. In Lake

Superior Copper and the Indians, Miscellaneeps Studies of Great Lakes
Prehistogz, edited by James B. Griffin, pp. 77-82. Anthropological Papers

17, Museum of Anthropology, University of Michigan, Ann Arbor.

Griffin, James B.
1966 The Calumet Ancient Pit. The Michigan Archaeologist 12(3):]30-133.

Guthe, Carl
1930 Unpublished field notes on file at the University of Michigan, Museum
of Anthropology.

 

295

Halsey, John R.

1976 The Bassinger Site: A Malp'compenent Site in Lhe Saginaw Vailey of

Michigan With a Review of Early Late Woodland Mortum Complexea
in the Northea_stern Woodlands. Unpublished PhD dissertation,

University of Chapel Hill. University Microfilms International, Ann
Arbor.

1983 Miswabik-Red Metal: Lake Superior Copper and the Indians of Eastern
North America. Michigan Histog 67(5):32-41.

Halverson, Colleen
1988 An Archaeological Survey ef Stephen, Cameron, Fling, and Kakgi Lakes
in Northwestern Ontap'q. Conservation Archaeology Report 12, Minstry
of Citizenship and Culture, Northwestern Region.

Hamilton, Scott
1981 The Archaeology of Wenesaga Rapids. Archaeology Research Report
17, Archaeology and Heritage Planning Branch, Ministry of Culture and
Recreation, Toronto.

1988 The Concept of Cultural Edge in the Boreal Forest of Manitoba and
Northwestern Ontario. In r For d ub-Ar ' Ar h
edited by CS. Reid, pp. 37-71. Occasional Publications of the London
Chapter, Number 6, Ontario Archaeological Society.

Harbottle, Gar
1976 Activation Analysis in Archaeology. In Radimhemism 3:33-72. The
Chemical Society, London.

Hearne, Samuel

1958 A Jearney frem Prince of Wale’s Fee; in Hudson’s Bay te ihe Northern
chap in (he Years 1769, 1770, 1771, and 1772. Edited by Richard

Glover, MacMillan and Company, Toronto.

Helm, June
1968 The Nature of Dogn'b Socioterritorial Groups. In Man the Hunter
edited by Richard B. Lee and Irven DeVore, pp. 118-125, Aldine Press,
New York.

 

Hickerson, Harold
1960 The Feast of the Dead Among the Seventeenth Century Algonkians of
the Upper Great Lakes. American Anthropologist 62:81-107.

1970 The Chippewa Qd Their Neighbors: A SEQ in Emnehiaieg. Studies
in Anthropological Method, Holt, Rinehart, and Winston, Inc., New

York.

296

Hlady, Walter M.

1970 Ten Thopsand Years; Archaeology in Mgitoba, Commemorating
Mgiteba’s (fingennial. 1870-1970. Manitoba Archaeological Society,

Altona.

1971 An Introduction to the Archaeology of the Woodland Area of Northern
Manitoba. Manitoba Archaeologig Newsleiier 8(2-3).

Hodder, Ian
1979 Economic and Social Stress and Material Cultural Patterning. Amerigip
Anp'guig 44(3):446-454.

Holman, Margaret
1978 The Settlemen]; System of the Maekinac Phase. Unpublished PhD
Dissertation, Michigan State University. University Microfilms
International, Ann Arbor.

Holmes, William H.
1901 Aboriginal Copper Mines of Isle Royale, Lake Superior. American

Anihropologist 3:684-696.

1919 Handbeok ef Aboriginal American Anp'gaities, Part I, Introductog, The
Lithic Indusgies. Bulletin 60, Bureau of American Ethnology,

Smithsonian Institution. US. Government Printing Office, Washington,
DC.

Hoxie, R. David
1980 An Analysis of the Late Woodland Copper Assemblage from the Sand
Point Site, Baraga County, Michigan. In The Sand Point Site (208614),
edited by Margaret Holman and Terrance Martin. The Miehigan
Arehaeologisi 26(3-4):25-38.

Hruska, Robert
1967 The Riverside Site: A Late Archaic Manifestation in Michigan. _T_h_e

W'seensin Archeologist 48(3):145-260.

Huber, N. King

1975 The Geolegic Stegy pf Isle Reyale Natienai Paik. Bulletin No. 1309, U.
8. Geological Survey, US. Government Printing Office, Washington,

DC.

Innis, Harold A.
1970 The Far Trade in Capada. University of Toronto Press, Toronto.

 

297

Irimoto, Takashi

1981 Qm'pemap Ecology; Group Strucgre and Cariboa Hpnting Sgetem.
Senri Ethnological Studies 8, Osaka.

Irving, Roland D.
1883 The Cepper—Bearing Reeks ef le Superior. Monographs of the

United States Geological Survey, Volume V, US. Government Printing
Office, Washington, DC

Jacobs, Alan
1977 Maasai Inter-Tribal Relations: Belligerent Herdsmen or Peaceable

Pastoralists. In Warfare Ameng East Afrieap Herders, edited by
Katsuyoshi Fukui and David Turton, pp. 33-51, Senri Ethnological

Ethnological Studies 3.

J anes, Robert R.
1973 Indian and Eskimo Contact in Southern Keewatin: An Ethnohistoric

Approach. Ethnohistoiy 20(1):39-54.

Janzen, Donald E.

1968 The Naemikeng Pain; Siie apd ihe Dimensions of Laurel in the Lake
Saperior Basin. Anthropological Papers No. 36, Museum of

Anthropology, The University of Michigan, Ann Arbor.

J ochim, Michael A.
1981 Strategies for Survivfl: Culturai Behavior in an Ecological Context.

Academic Press, Inc., New York.

Julig, Patrick, L.A. Pavlish, and R.G.V. Hancock
1988 INNA of Lithic Materials from the Western Lake Superior Region,
North America: Determination of Raw Material Sources. In,

Prgfieedings of the 26th International Smposium ef Arehaeomem.

Toronto.

n.d. Report on INNA Analysis of Lithic Materials from Isle Royale, Lake
Superior. Unpublished manuscript in possession of authors.

Justeson, John and Steven Hampson
1985 Closed models of open systems: boundary considerations. In, fie

Archaeelegy pf Frenp'efi apd Beppdmes, edited by Stanton W. Green
and Stephen M. Perlman, pp. 15-30. Academic Press, New York.

Karamanski, T., R. Zeitlin, and J. Derose
1988 Narrative Histoiy of Isle Royale National Park. Mid-American Research
Center, University of Chicago, Loyola.

298

Kellog, Louise Phelps
1926 The French Regime in Wisconsin apd ihe Northwest. State Historical

Society of Wisconsin, Madison.

Kenyon, Walter A.

1986 Mounds of Saered Earth; Barial Mounds of Ontario. Archaeology
Monograph 9, Royal Ontario Museum, Toronto.

Koelz, Walter
1929 A survey of the lakes of Isle Royale with an account of the fishes
occuring in them. Typed ms., Institute of Fisheries Research, University
of Michigan, Ann Arbor.

Kohl, Johann Georg

1985 Kimhi-Qami: Life Among ihe Lge Superier Ojibwa. Translated by
Lacelles Wraxall. Minnesota Historical Society Press, St. Paul.

Kuhn, Robert D.
1986 Interaction Patterns in Eastern New York: A Trace Element Analysis of
Iroquoian and Algonkian Ceramics. The Bulletin and Journal of

Arehaeology for New York State 92:9-21.

Lake Superior Basin Workshop
1988 Desparately Seeking Siouans: The Distrrbution of Sandy Lake Ware.
Wanikap 88-1.

Leach, Edmund R.
1954 Political Systems of Highlgd Burma. Beacon Press, Boston.

Leacock, Eleanor

1954 The Moniagpais "Hunting Territegz" and ihe Far Trade. American
Anthropological Association, Memoir 78.

Lenihan, Dania] J ., editor

1987 bmer ed C ltural Resources tu : Isle R0 Na 'on P k.
Professional Papers No. 8, Southwest Cultural Resources Center, Santa
Fe.

Linton, Ralph (editor)
1940 Aegilturation in Seven Ameriean India Tribes. New York.

Little, Elizabeth A.
1987 Inland Waterways in the Northeast. Midginp'nengal Jeurnal of

Arehaeolegy 12(1):55-76.

299

Longacre, William A.
1968 Some Aspects of Prehistoric Society in East-Central Arizona. In New

Perspeetives in Archeology, edited by Sally R. Binford and Lewis R.
Binford, pp. 89-102. Aldine Press, New York.

Lovis, Wm. A.
1973 Qte Weedland giltural Daamics in the Northern Lower Peninsula of
Miehigan. Unpublished PhD Dissertation, Michigan State University.
University Microfilms International, Ann Arbor.

1990 Accelerator Dating of the Ceramic Assemblage from the Fletcher Site:
Implications of a Pilot Study for Interpretation of the Wayne Period.

Midcontinental Joarnfl of Archaeology 15(1):37-50.

Luedtke, Barbara
1976 Limic Material Distribatiens ad Interaeiion Patterns During the Laie
Weodland Period in Michigan. Unpublished Ph.D. dissertation,
Department of Anthroplogy, University of Michigan. University
Microfilms International. Ann Arbor.

Lugenbeal, Edward N.
1978 The Blackduck Ceramics from the Smith Site (21KC3) and Their
Implications for the History of Blackduck Ceramics and Culture in

Northern Minnesota. Midantinental Jearnal of Arehaeelegy 3(1):45-68.

Lynott, Mark J ., Patrick E. Martin, and Caven P. Clark
1987 Isle Royale Archeology: 1987. Paper presented at the Midwest
Archaeological Conference, Milwaukee.

Lynott, Mark J ., Jeffrey J. Richner, and Mona Thompson
1986 Archeological Investigations at Voyageurs National Park: 1979 and 1980.
Occasional Studies in Anthropology No. 16, National Park Service,
Midwest Archeological Center, Lincoln.

Maass, Carol
1984 Interim Cultural Sites Inventory. Unpublished manuscript on file at Isle
Royale National Park, Houghton, and Midwest Archeological Center,
Lincoln.
MacNeish, R. S.
1952 Iregaeis Potteg fiypes. Bulletin 124, National Museums of Canada,
Ottawa.

1958 An Ingedpetien to the Archaeolon pf Soathwesi Maa'toba. Bulletin
157, Anthropological Series 44, National Museums of Canada, Ottawa.

300

Martin, Christian J.

1988 An Ecologia Histegt of Shap—tailed Groase in Isle Reyale National
Park. Unpublished MS thesis, University of Minnesota, Duluth.

Martin, Patrick E.

1988 Teehnigyl Report on Archaeologiefl Survey and Evaluation Isle Royale

National Park Daring 1986. Ms. on file, Archaeology Laboratory,
Department of Social Sciences, Michigan Technological University,

Houghton.

Martin, Terrance and Mary Masulis

1989a Animal Remains from Archaeelegiefl Sites at Isle Royale National Park,
Michigan: 1987 Season. Quaternary Studies Program Technical Report

89-389-8, Illinois State Museum Society, Springfield.

1989b Animal Remains from Archaeological Sites at Isle Rmale Natioal Pg
Miehiga: 1988 Season. Quaternary Studies Program Technical Report

89-512-30, Illinois State Museum Society, Springfield.

Mason, Ronald J.

1966 Twp Saatified Sites en me Deer Peninsala of Wiaensin.
Anthropological Papers 26, Museum of Anthropology, University of

Michigan, Ann Arbor.

1967 The North Bay Component at Porte Des Morts Site, Door County,
Wisconsin. The Wiscensin Areheologist 48(4):267-345.

1969 Laurel and North Bay: Diffusional Networks in the Upper Great Lakes.
Ameriea Antiquity 34(2):295-302

1976 Ethnicity and Archaeology in the Upper Great Lakes. In, @ltural
an e and ntin i : Essa s in Honor of ames Bennett Griffin
edited by Charles E. Cleland, pp. 349-361. Academic Press, New York.

1981 Great Lakes Archaeelogy. Academic Press, New York.

1986 Rmk Island: Historical Indian Arehaeelogy in the Nerthem Lake
Miehigan Basin. Special Paper 6, Midcontinental Journal of

Archaeology.

Mathiassen, Therkel
1930 The Question of the Origin of Eskimo Culture. American

Anthrepelogist 32(4):591-607.

301

McHale Milner, Claire, and John O’Shea
1990 Life After the Juntunen Site? Late Prehistoric Occupation of the Upper
Great Lakes. Paper presented at the Midwest Archaeological
Conference, October 5-6, Evanston.

McLeod, Mike
1978 The Archaeology of Deg Lake, Thander Bay: 9000 Years ef Prehistosy.
Report Submitted to the Ontario Heritage Foundation, Lakehead
University.

McPherron, Alan
1967 The Junmnen Site and the Late Woodland Prehistog of the Upper
Great Lakes Area. Anthropological Papers 30, Museum of
Anthropology, University of Michigan, Ann Arbor.

Meech, L. David
1966 The Wolves of Isle Royae. Fauna of the United States National Parks,
Fauna Series 7. US. Government Printing Office, Washington, DC.

Michlovic, Michael
1981 Patterns of Social Interaction and the Dynamics of Artifact Style. In,
Current Directions in Midwestern Archaeology: Selected Papers from
the Mankato Conference. edited by Scott Anfinson, pp. 41-48.
Minnesota Historical Society, Occasional Publications in Minnesota
Anthropology 9, St. Paul.

Milner, George, Virginia Smith, and Eve Anderson
1988 Conflict, Mortality, and Community Health in an Illinois Oneota
Population. Paper presented at the conference entitled: Between Bands
and States: Sedentism, Subsistence, and Interaction in Small-Scale
Societies. Center for Archaeological Investigations, Southern Illinois
University, Carbondale.

Moore, James A.
1981 The Effects of Information Networks in Hunter-Gatherer Societies. In,

Hunter-Gatherer Foraging Strategies, edited by Bruce Winterhalder and
Eric Smith, pp. 194-217. The University of Chicago Press, Chicago.

Munsell Color
1975 Munsell Soil Color Gharts. Munsell Color, MacBeth Division of
Kollmorgen Corporation, Baltimore.

Neff, Hector
1988 Reconstructing Ceramic Production from Ceramics Compositional Data:

A Guatemalan Example. Jeurnal ef Field Archaeolegy 15:339-348.

302

Neff, H., R.L. Bishop, and EN. Sayre
1988 A Simulation Approach to the Problem of Tempering in Compositional

Studies of Archaeological Cerrnaics. The Journal of Arehaeologieal
aienee 15:159-172.

1989 More Observations on the Problem of Tempering in Compositional

Studies of Archaeological Ceramics. The Joumg of Archaeological
Scienee 16:57-69.

Noble, William C.
1975 Van Beisen (AfHd-2): A Study in Glen Meyer Development. Ontario

Archaeology 24.

Packard, R. L.

1893 Pre-Gelambia Copar Mining in Nosth America. Annual Report of the
Board of Regents of the Smithsonian Institution, July 1892. US.

Government Printing Office, Washington, DC.

Parker, John (editor)

1976 The 103,;sz ef Jonaman Carver and Related Dauments 1766-1770.
Minnesota Historical Society Press, St. Paul.

Parker, Kathryn E.
1988 Plant Remains from 1987 Archaeological Testing on Isle Royale. Ms. on
file, Midwest Archeological Center, Lincoln.

1989 Plant Remains from 1988 Archaeological Testing on Isle Royale. Ms. on
file, Midwest Archeological Center, Lincoln.

1990 Plant Remains from 1989 Archaeological Testing on Isle Royale. Ms. on
file, Midwest Archeological Center, Lincoln.

Pelleck, John A.
1983 Fores Point E -1:A Woodland site on Red Lake No western
Gntarie. Report 3, Northwest Region, Ministry of Culture and
Communications.

Penman, John T.
1977 The Old C0pper Culture: An Analysis of Old Copper Artifacts. The

Wisconsin Archeolegist 58(1):3-23.

Perry, Leslie A.
1986a Archeology Along the St, Groix Riverway 1981. Midwest Archeological

Center, National Park Service, Lincoln.

303

1986b Ar h l ' ie Eval tion t. roix Nati n ni Riverw 19 2.
Midwest Archeological Center, National Park Service, Lincoln.

Peters, Gordon, John Hunn, Karene Motivans, and Walter Okstad

1983 giltaral Resource Maagement on the Superior National Forest, 1982
Annpai Repeat. Cultural Resource Report No. 1. USDA. Forest

Service, Superior National Forest, Duluth.

Peters, Gordon (editor)

1984 Galtural Resoarce Management on the Superior National Forest, 19S;
Annaai Repeit. Cultural Resource Report No. 2. USDA. Forest

Service, Superior National Forest, Duluth.

Peters, Gordon (editor)

1986 giltural Resouree Maagement en the Superior Natienal Ferest, 1985
Annaal Repog. Cultural Resource Report No. 4. USDA. Forest

Service, Superior National Forest, Duluth.

Peterson, Lynell
1986 An Attnbate Analysis of Sandy Lake Ware from Norman Coung ad
N930 Qngal Minnesota. Unpublished Master’s thesis, University of
Nebraska, Lincoln.

Peterson, Rolf O.
1977 Welf Ecology ad Prey Relationships on Isle Reyale. National Park
Service Scientific Monograph Series 11. US. Government Printing
Office, Washington, DC.

Prescott, J. R. V.
1965 The Geegiaphy of Frentiefi ad Bepndaries. Aldine Press, Chicago.

Pye, E. G.
1969 010 and n North hor f Lake Su rior. Geological Guide
Book 2, Ontario Department of Mines, Toronto.

Quimby, George I.
1939 Aboriginal Camp Sites on Isle Royale, Michigan. Ameriea Antiguig
4(3):215-223.

1966a The Dumaw Creek Site: A Seventeenth Century Indian Village and
Cemetery in Oceana County, Michigan. FieldianazAnthropology
56(1):1-91.

304

1966b Indian Culture and European Trade Goods: The Archaeology of the
Historic Period in me Western Great Lakes Region. The University of

Wisconsin Press, Madison.

Rajnovich, Grace
1980 Balysadare (Dka-lO): A Laurel-Blackduck Site at the Source of the
Winnepeg River. Ontario Archaeology 33:37-58

1983 Th S ce Point ite' A m ative Stu of Selkirk Com nen in
the Boreal Forest. Conservational Archaeology Report No. 1,
Northwestern Region, Heritage Branch, Ministry of Citizenship and
Culture. Kenora.

Rajnovich, Grace and CS. Reid
1987 Rescuing Rat River Por_tage Prehisteiy; Preliminag Repert on me 1986

Exeavations of the Ballaacree Site in Kenora. Conservation
Archaeology Report 11, Northwestern Region, Ministry of Citizenship
and Culture.

Ramsden, Peter G.

1977 Refinement of Some Aspects of Huron Geramie Anflysis. Mercury

Series 63, National Museum of Man, Ottawa.

1988 The Huron-Algenkia Boandaty; The View from Hahbprton ad the
Kawartha fies. Paper presented at the Annual meeting of the

Canadian Archaeological Association, Whistler, BC.

Rapp, George, James Allert, and Eiler Henrickson

1984 Traa Element Discrimination of Discrete Saws of Native Copper. In
Archaeological Chemistry III, edited by J .8. Lambert, pp. 273-293.

American Chemical Society.

Ray, Arthur J.

1974 Indias in the Far Trade: Their Role as Huntefi, Trapars and

Middlemen in the Lands Southwest of Hudson Bay, 1660-1870.
University of Toronto Press, Toronto.

Reid, CS.

1975 The Boys site and the Early Ontario Iroquois Tradition. Mercury Series,
Archaeological Survey of Canada, Paper 42, National Museum of Man,

Ottawa.

1988 Some Ideas Concerning the Formulation of Research Designs and
Excavation Methodologies on Boreal Forest Habitation Sites.

Midcontinental Journal of Archaeology l3(2):187-221.

305

Reid, C.S.(editor)

1980 Studies in West Patricia Archaeology No. 1:1978-1979. Archaeological
Research Report 15, Historical Planning and Research Branch, Ministry

of Culture and Recreation.

Reid, CS. and WA. Ross, (editors)
1981 Studies in West Patrieia Archaeology No. L1979-1980. West Patricia
Heritage Resource Report 2, Archaeological Research Report 16,
Historical Planning and Research Branch, Ontario Ministry of Culture
and Recreation.

Renfrew, Colin
1977 Alternative Models for Exchange and Spatial Distribution. In, Exchange
tems in Prehisto edited by Timothy Earle and Jonathon Ericson,
pp. 71-90. Academic Press, New York.

Richards, Bruhy, and Lynne Goldstein
1987 Culture Contact in the Late Prehistoric and Historic Midwest: French
and Native American Interactions as a Model for Late
Woodland-Mississippian Dynamics. Paper presented at the annual
meeting of the Midwest Archaeological Conference, Milwaukee.

Rogers, Edward S.
1967 The Material Gulture of the Mistassini. Bulletin 218, National Museum
of Canada, Ottawa.

Ross, W.A. (editor)

1982 Smdies in West Patricia Arehaeology Number Threell980-1981,
Archaeology Research Report No. 2. Ministry of Citizenship and

Culture, Toronto.

Sackett James R.

1982 Approaches to Style in Lithic Archaeology. ,lournfl ef Anthrepelogia
Arehaeelogy 1(1):59-112.

Sahlins, Marshall
1961 The Segmentary Lineage: An Organization of Predatory Expansionism.

Amerifl Anthroalogist 63:322-345.
1972 Stone Age Economies. Aldine Publishing Company, New York.

Salzer, Robert J.
1974 The Wisconsin North Lakes Project: A Preliminary Report. In Aspects
fU r retLak Anthr 010 P e ianoroleod
Wilford edited by Elden Johnson, pp. 40-54. Minnesota Prehistoric
Archaeology Series 11, Minnesota Historical Society, St. Paul.

 

306

1986 Other Late Woodland Developments. The Wisconsin Areheologist
67(3-4):302-313.

Sampson, C. Garth

1988 Sadistic Boundaries Among Mobile Hunter-Foragefi. Smithsonian

Institution Press, Washington, DC.

Sauer, Norman J.
1990 An Analysis of the Human Skeletal Remains from the Massee
Rockshelter, Isle Royale National Park. Ms. on file, National Park
Service, Midwest Archeological Center, Lincoln, NE.

Sayre, E.V.
n.d. Brookhaven Procedures for Statistical Analyses jof Multivariate
Archaeometric Data. Brookhaven National lab, Report BNL—21693.

Schoolcraft, Henry R.

1821 Travels to the Seuras of the Mississippi River. E.M. Hosfor, Albany,
New York.

Schortman, Edward M.
1989 Interregional Interaction in Prehistory: The Need for a New Perspective.

American Antigpiht 54(1):52-65.

Service, Elman R.
1962 Primitive Social Grganization. Random House, Inc., New York.

1971 Cultural Evolutienism: Theog in Prah'ge. Holt, Rinehart, and Winston,
New York.

Sharp, Henry
1977 The Caribou-Eater Chipewyan: Bilaterality, Strategies of Caribou

Hunting, and the Fur Trade. Aretie Anthropolegy 14(2):35-40.

Sharp, Robert W. and Nord
1960 A Fishery Survey of Some of the Lakes of the Isle Royale National
Park. Bureau of Sport Fisheries and Wildlife, Minneapolis. Ms. on file,
Isle Royale National Park, Houghton.

Sharrock, Susan R.
1974 Crees, Cree-Assinrboines, and Assiru'boins: Interethinc Social
Organization on the Far Northern Plains. Ethnohistoty 21(2):95-122.

307

Shay, C. Thomas

1971 The Itam Bison Kill Site: An Ealegieai Analysis. Minnesota Prehistoric
Archaeology Series No 6, St. Paul.

Slavick, Allison D. and Robert A. J anke
1987 The Vascular Flora of Isle Royale National Park. The Michigan Boflist
26(3).

Smith, Beverly
1987 Dog Burials of Late Prehistoric Algonquian Sites in Northeastern
Ontario. Paper presented at the 52nd Meeting of the Society for
American Archaeology.

Smith, James GE.
1978 Economic Uncertainty in an "Original Affluent Societyz" Canbou and
Caribou-Eater Chipewyan Adaptive Strategies. Arctic Anthropology
15(1):68-88.

Smith, James 6.8. and Ernest Burch
1979 Chipewyan and Inuit in the Central Canadian Subarctic, 1613-1977.

Aretie Anthropology 16(2):76-101.

Spicer, Edward H.
1961 Types of Contact and Processes of Change. In, Perspectives in Ameriean

India giltare Change, edited by Edward H. Spicer, pp. 517-544.
University of Chicago Press, Chicago

Stanley, George M.
1932 Abandoned Strands of Isle Royale and Northeastern Q0 Superior.

Unpublished Ph.D. dissertation, University of Michigan.

Steinbring, Jack
1974 The Preceramic Archaeology of Northern Minnesota. In Asacts of
r r a kes Anthro, l 'P rsin Hon r le d Wilford
edited by Elden Johnson, pp. 64-73. Minnesota PrehistoricArchaeology
Series 11, Minnesota Historical Society, St. Paul.

Stoltrnan, James B.
1973 The Laurel Culture in Minnesota Minnesota Prehistoric Archaeological
Series No. 8, Minnesota Historical Society, St. Paul.

1986 The Archaic Tradition. The Wisconsin Archeologist 67(3-4):207-242.

308

Stone, Lyle and Donald Chaput
1978 History of the Upper Great Lakes Area. In Northeast, edited by Bruce
G. Trigger, pp. 602-609. Handbook of North American Indians, vol. 15,
William G. Sturtevant, general editor. Smithsonian Institution,
Washington, DC

Stuiver, Minz, and 8. Becker
1986 High Precision Decada] Calibration of the Radiocarbon Time Scale,
A.D. 1950-2500 BC. Radiocarbon 28:863-910.

Swanton, John R.
1911 India Tribes of me Lower Mississippi Vailey ad Adjacent Coast ef the

Gulf of Meg'ee. Bulletin 43, Bureau of American Ethnology,
Smithsonian Institution. US. Government Printing Office, Washington,
DC.

Syms, E. Leigh
1977 Orltural Ecology and Ecological Dynamics of the Ceramic Period in
Southwestern Manitoba. Plains Anthropologist 12(2).

Thomas, David Hurst
1975 Non-Site Sampling in Archeology: Up the Creek Without a Site. In
Sampling in Archeology, edited by James W. Mueller, pp. 61-81.
University of Arizona Press, Tucson.

Thwaites, Rueben Gold
1899 The Jesuit Relations and Allied Dapments: Travels
-1900 ad ixplorations of the Jesuits and Missionaries in New France

1619-1791. 73 Volumes. Barrows Brothers Campay, Cleveland.

Tornay, Serge
1977 Armed Conflicts in the Lower Omo Valley, 1979-1976: An Analysis
From Within Nyangatom Society. In Warfae Ameng East Afrim
Herders edited by Katsuyoshi Fukui and David Turton, pp. 97-117,
Senri Ethnological Ethnological Studies 3, Osaka.

 

 

Trigger, Bruce G.

1976 The @‘ldren of Aataentsie: A Histeiy ef the Huron People to 1660.
McGill-Queen’s University Press, Montreal and London.

1978 Time ad Traditions, Essays in Archaeological Intergpretations. Columbia
University Press, New York.

309

1984 The Road to Affluence: A Reassessment of Early Huron Responses to

European Contact. In Affluence and Cpltural Survival: 1981 P_rec_§e_dr_ng'

ef the American Ethnolegical Sociehz, edited by Richard F. Salisbury
and Elisabeth Tooker, pp. 12-25. American Ethnological Society,

Washington, DC.

Vernon, W. W.
1985 New Perspectives on the Archaeometalurgy of the Old Copper Industry.

MCA Joarnal Supplement 3(5):]54-163.

Walthall, John A.

1981 Galena and Aborigm’ al Trade in Eastern North America. Scientific
Papers, Volume 17, Illinois State Museum, Springfield.

West, George A.

1929 @pper: Its Mining and Lise ml the Aborigihes of the Lake Superior

Region, Report of the McDonald-Massee Isle Royale E_xpedition, 1928.
Bulletin of the Public Museum of the City of Milwaukee 10(1):1-182.

Whallon, Robert
1968 Investigations of Late Prehistoric Social Organization in New York
State. In New Pegpectives in Archeology, edited by Sally R. Binford and
Lewis R. Binford, pp. 223-244. Aldine Press, New York.

Whittlesey, Charles

1863 Aneient Mining en the Shares ef Lake Saperier. Smithsonian
Contributions to Knowledge 13(155). Washington, DC.

Wiessner, Polly
1984 Reconsidering the Behavioral Basis of Style. Journal pf Anthropologiga
Archaeology 3(2):190-234.

1985 Style or Isochrestic Variation? A Reply to Sackett. American Antiguity
50(1):160-166.

Wilford, Lloyd A.
1941 A Tentative Classification of the Prehistoric Cultures of Minnesota.

Amerim Anh'gaity 6(3):231-249.

1955 A Revised Classification of the Prehistoric Culture in Minnesota.
American Antiguig 21:130-142.

310

Winchell, N. H.'

1881 Ancient Copper-Mines of Isle Royale. Popular Science Monthly
19(5):601-620.
1911 The Aborigines of Minnesota. Minnesota Historical Society, The Poineer
Company, St. Paul.
Wittry, Warren
1957 A Preliminary Study of the Old Copper Complex. The Wisconsin
Archeologist 38(4):204-221.
1963 The Bell Site, Wn9, An Early Historic Fox Village. The Wiseonsin
Areheologist 44(1):1-57.
Wright, Gary A.
1974 Arehaeology and Trade. Addison-Wesley Module in Anthropology, No.
49. Reading.
Wright, J. V.
1965 A Regional Examination of Ojibwa Culture History. Anthropologica
7:189-227.
1966 The Ontario Iroquois Tradition. Bulletin 210, National Museums of
Canada, Ottawa.
1967 The Laurel Tradition and the Middle Woodland Period. Bulletin 217,
National Museums of Canada, Ottawa.
1968 The Michipicoten Site. Bulletin 224:1-85, National Museums of Canada,
Ottawa.
1972a Gntario Prehistog. National Museum of Man, National Museums of
Canada, Ottawa.
1972b The Shield Arehaic. National Museums of Canada, Publications in
Archaeology 3. Ottawa.
1974 The Nodwell Site. Mercury Series, Archaeological Survey of Canada,
Paper 22, National Museums of Canada, Ottawa.
1981 Prehistory of the Canadian Shield. In Subarctic edited by June Helm,

 

pp. 86-96. Handbook of North American Indians vol. 6. William G.
Sturtevant, general editor. Smithsonian Institution, Washington, DC.