LIBRARY
Michigan State
University

 

 

 

PLACE IN RETURN Box to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

Ht! 0 n Ami-:7
L L I r

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 cJCIRC/DatoDuo.p65-p.15

THE LAUREL WORLD: TIME-SPACE PATTERNS OF CERAMIC STYLES
AND THEIR IMPLICATIONS FOR CULTURE CHANGE
IN THE UPPER GREAT LAKES IN THE FIRST MILLENNIUM A.D.
BY

Margaret Grace Nell Rajnovich

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

DOCTOR OF PHILOSOPHY

Department of Anthropology

2003

ABSTRACT

THE LAUREL WORLD: TIME-SPACE PATTERNS OF CERAMIC STYLES AND
THEIR IMPLICATIONS FOR CULTURE CHANGE IN THE UPPER GREAT
LAKES IN THE FIRST MILLENNIUM A.D.

BY

Margaret Grace Nell Rajnovich

The Middle Woodland Laurel “culture" has been recognized as one of the most
geographically extensive cultural phenomena in Eastern North America
stretching from western Quebec to Saskatchewan and from the edge of the
Hudson Bay Lowlands to northern Michigan and Minnesota. Radiocarbon
dating has revealed that it also is extensive in time, ranging from about 100 BC.
to after AD. 1000. Yet it has been studied in only piecemeal fashion as each
research effort concentrated on organizing data for the subregions of Laurel.
Culture histories have been developed for the Boundary Waters of Ontario-
Minnesota and have been partially devised for northern Michigan. Major
questions remain concerning Laurel origins, its internal dynamics over such a
wide area and long time span and its eventual demise. This dissertation
addresses those issues on a regional scale using multi-variate cluster analysis
to discern ceramic patterning in both the Boundary Waters of western Laurel
and in northern Michigan and northeastern Ontario in the eastern Laurel area.
These are then compared to delineate time-space patterns across Laurel as a
whole.

The study concludes that Laurel had a number of origins including the
preceding Archaic of the western and eastern areas with influences from
adjacent groups. The western and eastern ceramic assemblages are most

similar in early Laurel, then diverge through time to eventually define western

and eastern style zones later occupied by distinctly different Late Woodland
cultural entities. The pattern fits John H. Moore's model of ethnogenesis
wherein ethnic groups initially converge to form a new cultural entity which
eventually diverges along new lines creating new cultural entities.

This dissertation also concludes that the study of the development of
geographical sub-regions, or style zones, in cultural entities like Middle
Woodland that precede the appearance of two or more Late Woodland style
zones in the same geographical spaces can produce a new and useful tool for
understanding cultural origins and continuities.

The style zone approach advocated in this dissertation lends support to
the hypothesis that there was continuity between Middle and Late Woodland
peoples of the Upper Great Lakes, previously glimpsed only by the few ceramic

similarities between those cultural periods.

ACKNOWLEDGMENTS

A regional approach to a subject such as this dissertation requires
masses of data. I am indebted to those who toiled in the trenches over many
years in Ontario and Michigan before I was able to analyze the ceramics they
excavated and “crunch” the information with cluster statistics. I thank C. S.
Paddy Reid, David Arthurs, William A. Ross, Thor Conway and their crews who
amassed Middle Woodland information through the 19705 and 19803, as I did,
for the Ontario Ministry of Citizenship and Culture (it had other names at various
times). In those days, we began with a minimal amount of understanding of
Laurel but we now know so much more. That is equally true of northern
Michigan. I was able to use data collected by Dr. William A. Lovis, Dr. Charles E.
Cleland and Christine Branstner. I also thank the University of Michigan,
Museum of Anthropology, for allowing me to analyze the large assemblage from
the Naomikong Point Site excavated by Dr. Donald Janzen. Barbara Mead of
the Office of the State Archaeologist of Michigan gave me access to material
from the Gyftakis Site excavated under the direction of Dr. James E. Fitting.

The Batchewana Band of Ojibway First Nations allowed access to the
Laurel ceramics of Whitefish Island in Sault Ste. Marie, Ontario, and Bill Ross of
the Ontario Ministry of Culture, Tourism and Recreation provided space to
analyze those ceramics now housed in Thunder Bay, Ontario. He also helped
locate photos of the Fisk and Ballynacree assemblages. Stacey Bruyere of The
Kay-Nah-Chi-Wah-Nung Historical Centre provided photos of the Long Sault
ceramics. The archaeology laboratory of the University of Wisconsin-Milwaukee
gave me access to ceramics from the Richter Site.

I am grateful to Dusan Markovich of Avoca, Wisconsin, for allowing me to

write my dissertation at his pretty farmhouse that serves as the Gottschall

iv

Rockshelter project headquarters. Dr. Robert J. Salzer created the Gottschall
Rockshelter project for which I work and he too gave me permission to use the
Avoca house. He also gave me access to ceramics from the Squirrel Darn Site
housed at Beloit College.

I especially appreciate the help from my brother, Dr. James J.
Rajnovich of Algoma University College, who wrote the cluster program for this
dissertation and answered my questions even while he was working on his own
dissertation in computer science at the University of Western Ontario.

My dissertation committee was helpful at all times and with all things. It
was chaired by my major professor, Dr. Charles E. Cleland, and included Dr.
William A Lovis, Dr. Patrick LeBeau, and Dr. Helen Pollard of Michigan State
University and Dr. Ronald J. Mason of Lawrence University. I’ll always
remember the late Dr. Jerome A. Voss who served on my committee until his
untimely death.

This dissertation was born in my desire to contribute to the work of the
Rainy River First Nations in their 30—year-long effort to preserve and understand
the magnificent Kay-Nah-Chi-Wah-Nung (aka. Manitou Mounds), one of our
most intriguing Canadian monuments to the endurance of the Indian people on
the continent. My efforts in these pages are puny compared to those of my Rainy

River colleagues.

TABLE OF CONTENTS

LIST OF TABLES ....................................................... v

LIST OF FIGURES ..................................................... vii

CHAPTER 1

INTRODUCTION ........................................................ 1

CHAPTER 2

PREVIOUS RESEARCH ................................................ 32

CHAPTER 3

METHODOLOGY ....................................................... 44

CHAPTER 4

THE SITES .......................................................... 52

CHAPTER 5

LAUREL AND ITS MIDDLE WOODLAND NEIGHBORS .................... 67
Laurel and the Northeastern Plains ............................... 68
Laurel and Central Minnesota .................................... 70
Laurel and North Central Wisconsin ............................... 72
Laurel and Northeastern Wisconsin ................................ 76
Laurel and Southern Michigan and Southern Ontario ................ 81

CHAPTER 6

LAUREL TEMPORAL AND SPATIAL DYNAMICS .......................... 85
Temporal Dynamics ............................................. 85
Spatial Dynamics ............................................... 104

CHAPTER 7

SUMMARY AND CONCLUSIONS ........................................ 113

BIBLIOGRAPHY ...................................................... 121

APPENDICES ....................................................... 137
A. Legend for the Ceramic Charts ............................... 138
B. An Example of the Ceramic Decorative Variables

Used in the Laurel Study ..................................... 139

C. Whitefish Island Site ......................................... 141
D. Naomikong Point Site ....................................... 153
E. Cloudman Site .............................................. 172
F. Gyftakis Site ................................................ 182
G. Portage Site ................................................ 196

vi

H. Fisk Site .................................................... 207

I. Long Sault Site .............................................. 216
J. Ballynacree Site ............................................. 232
K. Squirrel Dam Site ........................................... 254
L. Richter Site .................................................. 263
M. Vessels by Site for Each Super-Cluster in the Laurel Study ....... 273
N. Vessel Clusters for the Squirrel Dam and Richter Sites

Based on the Laurel Super-Clusters ........................... 276
O. The Cluster Tree for the Whitefish Island Site: A Sample of the

Cluster Analysis Used in the Study ............................. 277

vii

LIST OF TABLES

1. Frequencies of Vessels in Each Assemblage ........................... 54

2. Exterior Rim Decorative Technique

for Selected North Bay Components ................................. 79
3. Coefficients of Similarity for Selected North Bay Components ........... .80
4. Relationships Among Selected North Bay Assemblages ................ 80
5. Laurel Radiocarbon Dates (uncorrected) ............................ 88-89
6. Calibrated Radiocarbon Dates for Laurel Components ............... 90-92
7. Vessel Clusters for Laurel Sites
Described by the Primary Trait for Each Cluster ......................... 98
8. Coefficients of Similarity for the Eastern Laurel Sites ................... 100
9. Relationships among the Eastern Laurel Assemblages ................. 100
10. Coefficients of Similarity for all Laurel Sites in the study .............. 103
11. Relationships among all Laurel Assemblages in the study ............. 103
12. Clusters Present at each Laurel Site ................................ 106
13. An Example of the Ceramic Decorative Variables
Used in the Laurel Study ....................................... 139-140
14. Whitefish Island Vessels ....................................... 145-147
15. Naomikong Point Vessels ..................................... 157-161
16. Distribution of Vessels at Naomikong Point ........................... 162
17. Cloudman Vessels ............................................ 175-177
18. Gyftakis Vessels ............................................. 184-187
19. Portage Vessels ............................................. .198-199
20. Fisk Vessels ................................................. 209—210
21. Long Sault Vessels ......................................... 221-224

viii

22. Ballynacree Vessels ........................................... 236-241

23. Ballynacree Vessel Clusters by House .............................. 242
24. Squirrel Dam Vessels ......................................... 257-258
25. Richter Vessels ............................................... 265-267
26. Vessels by Site for each Super-Cluster in the Laurel Study ....... 273-275

27. Vessel Clusters for the Richter and Squirrel Dam Sites
Based on the Laurel Super-Clusters .................................. 276

ix

LIST OF FIGURES

1.Location of Northern Tier Middle Woodland cultures of the Great Lakes ..... 2
2. Location of Laurel sites discussed in the text ............................ 39
3. Legend for the ceramic charts ....................................... 138
4. Location of Whitefish Island at Sault Ste. Marie ...................... . 142
5. Excavation areas on Whitefish Island ................................. 142
6. Whitefish Island ceramic chart ................................... 148-150
7. Vessels from the Whitefish Island Site ................................ 151

8. Vessels from the Whitefish Island Site ................................ 152
9. Naomikong Point on the south shore of Lake Superior ................. 154
10. Excavation blocks at Naomikong Point ............................... 154
11. Naomikong Point ceramic chart. .............................. .164-170
12. Vessels from the Naomikong Point Site ............................. 171

13. The Cloudman Site on Drummond Island, Michigan .................. 173
14. Cloudman ceramic chart ....................................... 178-179
15. Vessel from the Cloudman Site ..................................... 180
16. Vessels from the Cloudman Site ..................................... 180
17. Vessel from the Cloudman Site .................................... 181
18. Vessel from the Cloudman Site ..................................... 181
19. Gyftakis ceramic chart .......................................... 189-191

20 Vessel from the Gyftakis Site ....................................... 192
21 . Vessels from the Gyftakis Site ..................................... 193
22. Vessels from the Gyftakis Site ...................................... 194

X

23. Vessel from the Gyftakis ........................................... 195 '

24. Portage ceramic chart ......................................... 201 -204
25. Vessels from Group 3 at the Portage ................................ 205
26. Vessels from Group 4 at the Portage Site ............................ 206
27. The Fisk Site on a tributary of the Winnipeg River

north of Kenora, Ont ............................................... 211
28. Excavation units at the Fisk Site .................................... 212
29. Features at the Fisk Site ........................................... 213
30. Fisk ceramic chart ................................................. 214
31. Vessels from the Fisk Site .......................................... 215
32. The Manitou Mounds Site (Kay-Nah-Chi-Wah-Nung)

in Northwestern Ontario ............................................ 219
33. The Long Sault Site on the Rainy River, northwestern Ontario .......... 220
34. Long Sault ceramic chart ....................................... 225-229
35. Vessels from the Long Sault Site .................................... 230
36. The Ballynacree Site on the Winnipeg River in Kenora, Ontario ......... 233
37. The excavation block at the Ballynacree Site ......................... 233
38. The three houses and related features at the Ballynacree Site .......... 234
39. Ballynacree ceramic chart ...................................... 244-248
40. Shoulder portion of a vessel from the Ballynacree Site ................ 249
41. Vessels from the Ballynacree Site . ................................. 249
42 Vessel from the Ballynacree Site .................................... 250
43 Vessels from the Ballynacree Site .................................... 251
44. Incised vessel from the Ballynacree Site ............................. 252
45. Zoned dentate stamped vessel from the Ballynacree Site .............. 253

X!

46. Squirrel Dam ceramic chart ................................... .259-260

47. Vessel from the Squirrel Dam Site. ................................ 261
48. Vessel from the Squirrel Dam Site ................................. 262
49. Richter ceramic chart ......................................... 268-269
50. Vessels from the Richter Site ....................................... 270
51. Vessels from the Richter Site ....................................... 271
52. Vessels from the Richter Site ....................................... 272

53. The cluster tree for the vessels from the Fisk Site:
A sample of the cluster analysis used in the study .................... 278

xii

1. INTRODUCTION

The Middle Woodland Laurel “culture” of the Upper Great Lakes has
been recognized since Lloyd Wilford’s pioneering explorations in the 19403 in
Minnesota as one of the most geographically extensive cultural phenomena in
Eastern North America, stretching from the Quebec-Ontario border to eastern
Saskatchewan, and from the edge of the Hudson Bay Lowlands to northern
Minnesota and Michigan. Laurel is an archaeological culture defined mainly by
its decorated ceramics, but also represented by medium sized projectile points
and scrapers, a fishing and hunting economy adapted to the Upper Great
Lakes and Boreal Forest environments, and the construction in some areas of
burial mounds. Radiocarbon dating has placed it in an extensive time range,
from about 100 BC. to after AD. 1000. Several researchers have suggested
Laurel people were the Middle Woodland precursors of the AIgonkian-speaking
people who still inhabit the region.

Because of its extensive geographical and temporal range, it has been
the subject of several major analyses that have indicated variations within
Laurel across time and space.Yet it has been studied in only piecemeal fashion
as researchers concentrated on their own subregion. By now, culture histories
have been devised for Saskatchewan, Manitoba and the Boundary Waters of
Ontario-Minnesota, and partially developed for northern Michigan (Figure 1).
No study has synthesized the data from the Laurel universe as a whole in order
to understand the dynamics of such a vast cultural landscape - the mechanisms
of interaction and change that operated at various times throughout the cultural
sequence both to knit together the subregions into a distinctively Laurel
phenomenon, and ultimately to divide them. A number of significant questions

remain, including the question of Laurel origins. Was Laurel a single entity that

.mmxfl 890 05 Co 85:30 9.6682, @625. Lot. Emctoz 9: Lo c2803 ”F 9:9...

 

(33> Fuzzum

   

‘\
‘
‘
‘
‘
‘
‘
‘
‘

 

‘C

-
"‘
.0

D‘.
||||||||||
I.

    

240.20.!

><m «.02

:Fi

I
.....

 

 

 

o \_
A
.. .. S
-. V)
S
N
a M
02320 3
M
V
N

 

 

 

diverged or several entities that converged. In other words, did Laurel have one
origin (Reid and Rajnovich 1991) or several origins (Mason 1991)? What was
the nature of Laurel dynamics over such a long period and across such a vast
northern world?

Investigations of social dynamics have recently taken the form of stylistic
analysis on the assumption that social groups express themselves through
style: as social networks change, stylistic features change. This dissertation will
examine Laurel ceramic style change across space and through time in an
effort to clarify the social dynamics of the Laurel world. A study of dynamics must
stress variability rather than normative descriptions, interaction rather than
isolated phase definitions and exotic characteristics as well as local ones. This
study will concentrate on these issues using multivariate analysis of three
ceramic assemblages from northwestern Ontario representing hypothesized
early, middle and late Laurel in that area and five ceramic assemblages from
northeastern Ontario-northern Michigan which were not previously analyzed
using a multivariate approach. Ideally, a study of Laurel interaction must include
assemblages external to the Laurel world, neighbors such as Malmo, Nokomis,
North Bay, Saugeen and Point Peninsula (Figure 1), an undertaking beyond the
resources of this dissertation; however, this study will use data from those areas
previously analyzed using other approaches and will include two assemblages,
one Nokomis and one North Bay, not previously reported elsewhere
(Appendices C and D).

Laurel is a manifestation of the Middle Woodland Period represented by
eastern North American components contemporaneous with the Hopewell
Interaction Sphere. Researchers have long recognized similarities among
several ceramic taxa in the Upper Great Lakes, and Ronald J. Mason has

labeled these the Northern Tier Middle Woodland to distinguish them from the

Havana-related complexes to the south (Mason 1967). James E. Fitting uses
the term Lake Forest Middle Woodland for the same regional group of Laurel-
related styles to emphasize not only ceramic decorative similarities but also
similar ecologically focused subsistence and settlement patterns as opposed to
the patterns adapted to the more southerly Carolinian zone (Fitting 1975). Along
with Laurel, these include the Malmo Focus of central Minnesota (Wilford 1955),
North Bay components of the Green Bay area (Mason 1967), later portions of
the Nokomis Phase in north central Wisconsin (Salzer 1969), Saugeen and
Point Peninsula phases in southern Ontario and New York (Wright 1967, Ritchie
and MacNeish 1949). They share traits with Laurel such as pseudo-scallop
shell and dentate stamp decoration. Formal comparisons among these wares
has not been made on a systematic basis, thus the nature and strengths of
interaction, such as might be expected to be physically manifested among them,
is not clear.

Laurel is characterized by a presence of burial mounds at some sites and
a hunting-fishing-gathering adaptation to the Lake Forest and Boreal zones of
the Laurentian Shield. The boundary between Laurel and Hopewellian
complexes is generally very sharp, corresponding closely with the Carolinian-
Canadian transition zone or the southern edge of the Lake Forest. Laurel’s most
characteristic feature is an assemblage of highly decorated ceramics to the
point of baroque omateness. The distinctive, plain surfaced ceramics have
conical bases and very varied decorative techniques and motifs, often multiple
horizontal bands encircling the upper third of the vessel. Laurel is unique in the
Great Lakes area in that it is the only ceramic complex of that region that has no
in situ Early Woodland antecedent equivalent to the thick series (Vinette,
Marion, Schultz) or the trailed series (Fox Lake, Dane, early Nokomis) which

precede Middle Woodland wares in other areas of the Great Lakes. Also

4

characteristic are medium-sized stemmed, side-notched and corner-notched
projectile points, net-fishing technology, and the use of toggle-head harpoons.
Copper is common especially on eastern Laurel sites. While there are some
links to Hopewell as indicated by the presence of burial mounds and a few
exotic items, it is not considered Hopewellian and extends in time much later
than that interaction sphere.

While Laurel shares some characteristics with its Middle Woodland
counterparts in the Upper Great Lakes, the distinctive, elaborately decorated
ceramics set it apart, making them the defining feature of Laurel. They have
well-fired, relatively thin paste with small temper particles (less than 2 mm in
diameter), plain surface treatment, and a design structure of three basic fields
on the upper rim, lower rim-neck, and shoulder. Decoration consists of lines of
stamps or incising, most often intricately executed with small tools repeatedly
pressed into the clay in closely spaced, multiple rows. Techniques include
dentate stamps, linear stamps, incising, pseudo-scallop shell stamps, and
occasionally cord-wrapped stick stamps. The pseudo-scallop shell stamp, a
hallmark of Laurel, is a small dentate stamp with offset notches that create a
wavy line. Laurel designs vary considerably because of small-scale choices by
the potters concerning number of rows or columns, spacing, use of
punctateslbosses, and structure of each row that is composed of either a simple
line of incising or stamps placed end-to-end or a complex line of closely spaced
stamps placed vertically or obliquely.

The small-scale variations create polythetic sets unsuitable for
typological classification; nevertheless, the large-scale similarities make Laurel
easily recognizable throughout its very wide distribution from Saskatchewan to
Michigan. As Laurel is distinguished by its ceramics on the large-scale, and

they display variability on the small-scale over space and time, they are

5

appropriate for a study of Middle Woodland dynamics in the Upper Great Lakes.

The Laurel manifestation has been termed a focus of the Midwestern
Taxonomic System (Wilford 1955), a tradition and a culture of the Willey and
Phillips System (Wright 1967, Stoltman 1973), and a configuration of the Syms
System (Reid and Rajnovich 1991). The confusion of taxonomic terms is due to
the fact that the complete internal structure of Laurel has not been investigated
and remains unclear. Due to the extremely wide spatial distribution of Laurel,
there has been, understandably, a good deal of provincialism among some
pioneering investigators. This study is designed to clarify the issue. In the
Boundary Waters area along the Ontario-Minnesota border, four phases of the
Willey-Phillips system have been established on the basis of extensive data for
the time period of about 100 BC. to some time after AD. 1000 (Stoltman 1973,
Lugenbeal 1977, Reid and Rajnovich 1991). Fitting has suggested two more
phases for the Straits of Mackinac of northern Michigan based on very limited
data (Fitting 1979). The broad patterns of Laurel have not been established. C.
S. Paddy Reid and l, following other researchers (Syms 1977, Stoltman 1973),
have noted the unsuitability of the Willey and Phillips taxonomy for Laurel
studies. We have therefore applied the Syms taxonomic system (to be
discussed below; see Syms 1977) in formulating four “complexes” for the
northwestern Ontario-Minnesota material comprising a Boundary Waters
Composite (Reid and Rajnovich 1991). We speculated that an eastern Laurel
composite also exists with its own internal dynamics perhaps different from from
the Boundary Waters composite, but we did not investigate further.

Following Wilford’s (1955) first identification of Laurel in northern
Minnesota, Richard S. MacNeish devised a culture history for southeastern
Manitoba that included Laurel foci (MacNeish 1958). Since then, James V.

Wright conducted archaeological surveys across central Canada and noted

6

Laurel’s range from Manitoba to the Quebec border (Wright 1967). K. C. A.
Dawson (1980, 1981) and several researchers for the Ontario government
located many previously unknown sites in northern Ontario (see Reid and
Rajnovich 1991) and work by David Meyer has located late Laurel sites in
Saskatchewan (Meyer 1983). Donald E. Janzen investigated a major Laurel
site on Lake Superior west of Sault Ste. Marie (Janzen 1968); James E. Fitting
began a culture history of Laurel sites at the Straits Mackinac of northern
Michigan (Fitting 1975, 1979), and others have noted Laurel and Laurel-like
ceramics from the western shores of Lake Michigan (Mason 1967, 1969, 1991;
Brose 1970). To date, hundreds of Laurel components have been located and
more than 50 radiocarbon assays are known (Reid and Rajnovich 1991). In
contrast to the previous Archaic period, the sites are large and numerous.
Indeed, some Laurel sites are impressively large even by Late Woodland
standards. Excavations of the Naomikong Point Site, to be discussed in this text,
stretched 190 feet along the southern shore of Lake Superior, recovering
133,000 sherds and disclosing no boundaries to the deposits within the
excavation areas. The site probably represents multiple occupations.

The above research has led to the development of culture histories
based on ceramics for southeastern Manitoba (MacNeish 1958), the Boundary
Waters (Stoltman 1973, Lugenbeal 1977, Reid and Rajnovich 1991) and
partially for the Straits of Mackinac area (Fitting 1979). No detailed attempt has
been made to compare ceramics across these subregional taxa.

Burial mounds have been located only in the Boundary Waters of
Ontario-Minnesota (Arthurs 1986, Lugenbeal 1977, Stoltman 1973, 1974,
Wright 1967, Kenyon 1986) and at Drummond Island near Sault Ste. Marie,
Michigan (Branstner 1992). Other burial modes have been discovered at

Arrowhead Drive on Bois Blanc Island in the Straits of Mackinac (Bettarel and

7

Harrison 1962), the Gyftakis Site in St. Ignace (Fitting 1979) and possibly Cave
95B at Burnt Bluff on the Garden Peninsula, northern Michigan (Fitting 1975).

One shrine related to pictographs, has been discovered at Burnt Bluff
(Cleland and Peske 1968) and it is possible that some of the hundreds of
pictographs located across the Canadian Shield are products of Laurel
(Rajnovich 1994).

Six Laurel sites have produced house outlines, all similar in size and
shape - oval structures roughly eight metres long by four metres wide with post
construction probably representing single- or extended-family wigwams (Reid
and Rajnovich 1986, 1991; Brose 1970; Peters et al 1983). The Ballynacree
Site produced three, apparently contemporaneous, houses. The house pattern
contrasts with the previous Archaic period when lodges were circular and only
about three to five metres in diameter (Pollock 1976, Wright 1972). The
demographic pattern at the onset of Laurel appears to be that of an increased
number of sites with increased numbers of people inhabiting each site.

Subsistence patterns changed as well. Laurel sites from Manitoba to
Michigan contain toggle-head harpoons and greater numbers of fish remains
(Mason 1981); Charles E. Cleland has also posited the technological
innovation of net fishing for the early Laurel period (Cleland 1982). Wild rice
appeared in the Upper Great Lakes as early as the last centuries BC. and has
been found associated with a Laurel component in Minnesota (Valpuu and
Rapp 2000). It should be noted that most Laurel components have a matrix of
acidic soils incompatible with faunal and floral preservation, so it may be that
increased fishing and wild ricing was more widespread than the archaeological
record indicates.

A number of researchers have speculated on the origins of Laurel.

MacNeish suggested that similarities of assemblages between Manitoba Laurel

8

and those of the Eastern Woodlands, and the almost complete absence of
shared traits west, north and south of the Laurel world in Manitoba, indicated
influence or migration to Manitoba from the Eastern Woodlands (MacNeish
1958: 78) . Wright, noting the similarities of Laurel ceramic decorative
techniques with those of Bronze Age sites in northwest Asia, concluded that
Laurel developed as a result of Asiatic influences diffusing southeastward,
although he recognized that evidence at the time of writing was “tantalizingly
suggestive but frustratingly limited and disjointed”. He hoped that future
research would fill in the gaps between the Kamchatka Peninsula and the
Laurentian Shield (Wright 1967: 132-133). Such evidence has not since come
to light. Stoltman suggested Laurel was a result of either migration or diffusion
from the south and was ultimately influenced by Havana. He based his
hypothesis on his ceramic seriation of Laurel components in Minnesota which
indicated the most southerly ones were also the oldest (1973: 122). Edward N.
Lugenbeal reanalyzed Minnesota Laurel components and found that the
earliest were not necessarily the most southerly and that early Laurel ceramics
were least like Havana Middle Woodland of Illinois (Lugenbeal 1977: 661). He
concluded that Laurel was an in situ, independent development. E. Leigh Syms
hypothesized a migration from the south, basing his theory on the possibility of
population growth in the Early Woodland Period to the south and a domino
effect causing territorial expansion northward (1977: 77). Mason has suggested
that Laurel had more than one point of origin (Mason 1991: 138). While the
Asiatic theory has generally been refuted, none of the others has been tested.
Reid and l plotted 25 components with radiocarbon dates and noted that
the first century BC. dates occurred in the Boundary Waters area of Ontario-
Minnesota; we therefore chose that area as the point of origin (Reid and

Rajnovich 1991: 222-223). The later dates indicated an expansion

northwestward into Manitoba in the first century AD. and a further expansion
eastward to northeastern Ontario and northern Michigan subsequent to that
time.Mason, however, has suggested that radiocarbon dates and ceramic
seriation indicate that Laurel in the northern Lake Michigan basin was probably
contemporaneous with Boundary Waters Laurel (Mason 1991). The Reid-
Rajnovich study indicates that, by the ninth century AD, Laurel had contracted
again to the Boundary Waters and areas northwest of it. We cautioned,
however, that problems in interpretation existed; for instance, one of the few
dates for eastern Laurel was the Naomikong Point assay of AD. 4301400
covering an 800-year time span (1991: 212) making the chronology of eastern
Laurel tentative at best. The most westerly Laurel picture is somewhat clearer.
David Meyer has studied Laurel in Saskatchewan and suggested that it entered
the province late in the sequence, possibly as late at AD. 800 (Meyer and
Hamilton 1994: 108). There are, so far, no early dates for Laurel there.

Laurel apparently developed into the Pine River series about AD. 600 in
northern Michigan, leading to the Mackinac Phase of Late Woodland about AD.
800 (Lovis and Holman 1976, Holman 1984, Cleland 1992) which
characterized the whole area previously occupied by Laurel in both Michigan
and northeastern Ontario. Laurel persisted into much later times in the west.
The Late Woodland Blackduck, Rainy River and Selkirk series of northern
Minnesota, northwestern Ontario, Manitoba and Saskatchewan, probably
descendants of Laurel and other cultures (Dawson 1976, Koezur and Wright
1976, Lugenbeal 1977, Rajnovich 1983, Meyer and Russell 1987), made their
appearance after about AD. 800, and they seem to have coexisted with some
Laurel occupations. Reid and l (1991: 198) obtained three controversial Laurel
assays from the Ballynacree Site in Kenora, Ontario, dating after AD. 1200,

about 200 years younger than most researchers would like. However, Meyer et

10

3! obtained other dates for late Laurel in Saskatchewan placing it at
A.D.1175¢155 and AD. 1305170 (Meyer et al 1988: 37).

Some researchers have hypothesized that Laurel represents Algonkian
speakers. Their arguments are based mainly on data that indicate an unbroken
ceramic transition from Middle to Late Woodland and Historic components in
the region (Stoltman 1973, Koezur and Wright 1976, Dawson 1976, Rajnovich
1983, Meyer and Russell 1987, Holman 1984, Lenius and Olinyk 1990, Cleland
1992). However Edward N. Lugenbeal and David Arthurs find the evidence
inconclusive (Lugenbeal 1977, Arthurs 1986), and Alan McPherron perceives a
gap between the ceramics of Middle and Late Woodland in the Straits area
(McPherron 1967). Peter Denny has suggested that Laurel peoples were
Algonkian, basing his argument on linguistic analysis that would place
Algonkian speakers in the Great Lakes area during the late Archaic Period
(Denny 1992). Today, the area is the home of the Algonkian-speaking
Anishinabe (Ojibway, Odawa and Cree) who claim the Woodland traditions as
their own and have taken control and preservation of two of the region’s most
important Laurel sites, Whitefish Island at Sault Ste. Marie in northeastern
Ontario, and Long Sault on the Rainy River in northwestern Ontario, both to be

analyzed in this dissertation.

ASSUMPTIONS

This study is based on a number of assumptions dealing with (1)
ethnogenesis, (2) taxonomy, (3) time-space dimensions of eastern Laurel, (4)
the relationship between style and social organization, and (5) the levels of

stylistic analysis.

11

(1) Ethnogenesis

A major question in archaeology centers around the issue of cultural
origins. John H. Moore has noted that the process has often been expressed in
a unilineal model requiring a single antecedent to the culture in question based
on cladistic theory that emphasizes a process by which daughter populations
are derived from a single parent population (Moore 1994: 925). He emphasizes,
however, that rarely do we find such phenomena among contemporary ethnic
groups, pointing, for example, to the formation of hybrid bands in California, the
Great Basin, Tierra Del Fuego, Australia and the northern plains of both the
United States and Canada (1994: 936). He proposes that a theory based on the
principle of uniformitarianism - what we observe in the recent record was also
likely in prehistoric social systems - would embrace what he terms
“ethnogenesis”, emphasizing the probability that each population has its roots
in several different antecedent groups. Rather than the cladistic approach ( from
the Greek clados, “branch”) which emphasizes development of several cultures
from a single source, he opts for a rhizotic approach (from the Greek rhizo,
“root”) (1994: 925) which emphasizes the emergence of a culture from several
roots. Using another simile, he likens the cultural evolutionary process to the
channels of a river which separate and recombine in complex ways resembling
braiding (1994: 930).

The ethnographic perspective supports a multi-regional approach,
arguing that human history has always been characterized by interaction across
ethnic and regional boundaries (1994: 937). Moore suggests that an example of
the theory operating in archaeology can be found in the Oneota case in the
Upper Midwest of the United States. He suggests that the record can show that
an assembly of diverse elements developed into the mature culture known as

Oneota which lasted for centuries, then underwent a process of diversification

12

resulting in historically identifiable cultures such as the Kansa, Omaha, Ponca
and others.

Martin Wobst also has been critical of models of bounded, isolated ethnic
units often described in ethnographic literature, but derived from the limitations
of the ethnographer’s study rather than the real situation (Wobst 1978).
Likewise, David L. Clarke has emphasized the rhizotic nature of “cultural
phylogeny" (Clarke 1968: 165) with its rhythmic pattern of repeated cycles of
cultural birth, growth, expansion, regional divergence, re-patteming and
realignments or "death” (1968: 404).

The pattern has been described in a number of studies of Indian groups
in central Canada. For instance, Robert A. Brightman studied the Cree of
northern Manitoba, finding “a complex history of amalgamations and
migrations” (Brightman 1993: 5). The genealogies of members of one band
show that they descended from families known to be in the area in the
eighteenth century combined with others who came into the area during the late
nineteenth century from three different locations. The contemporary group
descends from ancestors with at least two different dialects (1993: 7).

Janet E. Chute describes a situation in the mid-nineteenth century
among the Lake Superior Ojibway wherein tensions within and among bands
threatened to split known social groups. Furthermore, the dispute over lands
between the Ojibway bands and the United States government led
Shingwaukonse, a famous leader of a band of Ojibways at Sault Ste. Marie on
the Canadian side of the St. Mary’s River, to propose that several bands in
Minnesota, Wisconsin, and Michigan join his band in Ontario (Chute 1998: 152-
153). Had the plan succeeded (the Canadian government opposed it), the mass
migration and union at the Sault would have provided an example of the

cultural river channels joining.

13

David P.Braun and Stephen Plog discuss the evolution of “tribal” social
networks previously viewed by researchers as simply an integration of
autonomous local entities. They suggest, on the other hand, that local units
were not autonomous; rather, “tribal” units were the development of increasing
connectedness and formalization already existing in regional networks (Braun
and Plog 1982: 508-509). Likewise, we might expect the “tribal” networks to
eventually split along already developing regionalized sub-networks within the
“tribe”, be they lineages, sodalities, clans or other sub—groups.

In archaeological terms, then, the question of the origin of Laurel may be
misleading, as the answer may require more than one origin. Likewise,
questions of the origin of each Late Woodland culture post-dating Laurel in the
Upper Great Lakes are also misleading if we presuppose that only Laurel is the
antecedent. The evolution of several separate cultural entities evolving at least
partly out of Laurel becomes more understandable if we see mature Laurel,
figuratively speaking, as the river, and the various Late Woodland
developments as a final separation into new channels. Then we must ask if we
can identify regionalization within what appears at first glance to be
homogeneous Laurel leading to the Late Woodland threshold. In other words,
changes in pottery styles in early Late Woodland are not the gaginning of a new
cultural phenomenon but the Mme of previous developments in Laurel.

(2) Ethnogenesis and Taxonomy

E. Leigh Syms has grappled with the issue of the rhizotic model of
cultural evolution and archaeological taxonomy. He rejects the Willey and
Phillips system on the basis that it has been widely used to reflect an untrue
picture of cultural groups spatially limited to the scale of a locality or region
(Willey and Phillips 1958: 22) with a unilinear development (Syms 1977: 5). He

proposes instead the “Co-Influence Sphere Model” that takes into account

14

known ethnographic records of more than one ethnic group in a region and
interaction across biomes either by actual movements of bands within the yearly
cycle or by trade. He envisions a “Core Area" where an ethnic group spent most
of the year, ”Secondary Areas” to which the group went at specific times for
specific resources and/or deliberate interaction with other groups, and “Tertiary
Areas” that were marginal, being used briefly and intermittently (1977: 5-6). For
instance, Cleland describes core, or home-base, territories for the Ojibway
bands of the Upper Great Lakes and secondary regional sites at which these
groups gathered such as Mo-ning-wun-a—kawn-ing (Madeline Island) and
Bawating (Sault Ste. Marie) (Loew 2001: 54; Cleland 1992: 101). During the
Fur Trade Period, a tertiary site for many groups was Moniang (Montreal).

The Syms model suggests that archaeological taxonomy requires more taxa
than the simplified Willey and Phillips two-tier system and he proposes four
units. The “complex” is like the phase, encompassing assemblages left by the
same group over a sufficiently narrow time period, but it differs from the phase in
that it can incorporate vast areas across several biomes including core,
secondary and tertiary areas (Syms 1977: 71). The complexes are not isolated
but overlap: Syms emphasizes interaction.

The complexes are subsumed into a “composite”, a number of
complexes that share a set of traits sufficiently similar to indicate a common and
recent ancestry but different in that micro-evolutionary changes have taken
place (Syms 1977: 71). The concept is similar to \Mlley and Phillips' term
“culture”. Composites are included in a “configuration” that combines
composites sharing sufficient traits to indicate distant generic relationships or
cultural convergence (1977: 72). Syms drops the terms “phase” and “culture"
because of ambiguity of definition stemming from their differential use in

different taxonomies (1977: 70-71). Reid and l have suggested that the Syms

15

system is more appropriate when dealing with the complexities of the long and
widespread Laurel world (Reid and Rajnovich 1991). We suggested that the
“phases" defined for western Laurel be termed “complexes” and that they be
subsumed into a Boundary Waters Composite of the Laurel Configuration. The
system is broad enough to allow for eastern and western Laurel composites
each with their own internal evolution but also with interaction between them.

Recently Brian Lenius and David Olinyk have used the Syms system to
great effect in realigning the previously confused world of the Late Woodland in
northern Minnesota, northwestern Ontario and Manitoba (Lenius and Olinyk
1990). The emphasis in the system stresses overlaps in the archaeological
record that indicate interaction and ethnogenesis (Syms 1977: 72); the goal of
the system is to delineate real social groups as opposed to arbitrarily defined,
bounded archaeological units. The attempt is to define not only the complexes
but also their relationships to each other both temporally and spatially.

(3)Temporal Sequence

A study of the dynamics of Laurel social behavior necessitates a
temporal continuum of sites from each of the western and eastern areas. The
western complexes have already been developed. As a first step, this study
requires control of the time dimension of eastern Laurel. Radiocarbon dates for
the eastern sites are available but ambiguous, overlapping to some degree.
Therefore the formulation of a temporal continuum must rely on temporal
seriation. While seriation can be affected by both time and space, I have chosen
to deal with assemblages from a number of sites in the Sault Ste. Marie-to-
Mackinac Straits area on the assumption that seriation will show that
differences among these are time related; that is, they represent a number of
sequential complexes rather than spatially defined unitsl do this for both

ethnographic and archaeological reasons.

16

Ethnographic accounts describe the Sault Ste. Marie-northern Lake
Michigan water route as a common corridor for travel with no geographical
barriers. Rather, the Great Lakes of this area enhanced communication along a
route of more than 160 km. The earlith ethnographic accounts of the area, The
Jesuit Relations, indicate that the Noquet extended across the Upper Peninsula
of Michigan from the eastern end of Lake Superior to the north shore of Lake
Michigan. The Jesuit Relations also report that they frequented the Sault Ste.
Marie area and were in the process of uniting with the Saulteaux (Thwaites
1959); that is, the subregion encompassing the eastern Upper Peninsula and
the Sault Ste. Marie area of Ontario was the home of closely interacting social
groups. As the landscape and transportation routes of the area remained very
similar from early Laurel times to the historic period, there is good reason to
believe that the people of each period moved about their environment similariy.

Cleland has noted that similar ecological, economic, social and
ideological factors produce similar territory sizes (Cleland 1992).
Archaeological phases/complexes for the Boundary Waters area of western
Laurel, environmentally and economically similar to the eastern Laurel area,
have been inferred from ceramic style similarities and radiocarbon dating. Each
complex contains sites that are located more than 160 km (100 miles) apart. In
the eastern Laurel area, the complex sizes should be roughly equal to those of
the Boundary Waters. The Sault Ste. Marie-to-northem Lake Michigan region
matches the complex size.

(4) Style

This study using ceramics to define interrelationships among people is
founded on the assumption that there is a relationship between social
organization and style (Deetz 1965, Engelbrecht 1974, Sackett 1982, Carr and
Neitzel 1995).

17

For this dissertation, style is defined using the definition provided by
Whitney Davis: it is “(a) a description of a polythetic set of similar but varying
attributes in a group of artifacts, (b) the presence of which can only be explained
by the history of the artifacts, (c) namely common descent” (Davis 1991: 19). A
polythetic set is one in which each artifact possesses a large number of
attributes of the group; each attribute may be found in a large number of the
artifacts of the group but no single attribute need be in every artifact, that is , no
attribute is both necessary and sufficient (Davis 1991: 19).

James R. Sackett has suggested that stylistic significance can be
assigned to any attribute whether it is utilitarian such as vessel form or non-
utilitarian such as decoration (Sackett 1985: 157). Margaret Conkey defines
style as design form and technique and she adds the process of production
Ieamed at least in part through interaction (Conkey 1991: 13). Thus, style is
related to interaction and interaction among all Laurel potters is apparent in the
first stages of production, the use of coiled construction, conical shape, plain
surfaces and design structure of three fields. Variability lies in the other aspects
of style, technique and design, and these are the focus of this study. Given that
there is interaction, what is the nature of those interactions? Do we perceive
groups fissioning, amalgamating or forming regional alliances? A number of
researchers have suggested that we can use style to formulate specific tests to
explore those processes of interaction.

Archaeology has treated style from a number of perspectives. The culture
historical approach treats styles as normative reflections of ideas, or norms, and
culture as a shared set of concepts. Members of a culture express those
differentially and therefore artifact style distribution synchronically should
approximate a bell curve reflecting variation around a central norm (Caine-

Hohman 1983: 15). Diachronically, styles form a battleship curve. Accordingly,

18

culture historians have stressed a type-variety system of analysis emphasizing
central tendencies such as Stoltman’s Laurel typology (1973). The approach
reflects a core-periphery cultural paradigm inappropriate for Laurel, the diffuse
distribution of which suggests a nodal, or modular, model in which each
component forms a node of equal stature with the others. The model could be
termed a “rendezvous pattern” like the one described by David Meyer for the
Saskatchewan Cree of the 18th and 19th centuries in which there were a
number of aggregating centers no one of which dominated the others in the
manner of a core-periphery pattern (Meyer 1995).

Since the 1960s, archaeologists have used a processual or systemic
approach stressing culture as behavior with style reflective of those behaviors.
In this approach, style distribution should reflect the patterns of interaction.
Polythetic analysis has gained favor in that it can explore patterns that may be
the result of a core-periphery behavior or some other behavior such as the
“rendezvous pattern” and can deal with style in a manner other than the core
type-variety method.

The behaviors of paramount interest to the processualists have been
adaptive responses. Claire McHale Milner has noted that in small-scale
societies, social integration is often a risk-buffering mechanism: times of high
natural, cultural or social stress cause social fission, intensified localized
amalgamation or increased regionalized alliance building (Milner 1998: 18).
Style should reflect these.

Milner discusses two paradigms that have driven approaches to cultural
evolution, the descent theory and the alliance theory (1998: 15). The descent
theory stresses internal development of lineally bounded descent groups, a
notion resulting from biases in the anthropological literature which

underestimated regional social interaction, and overestimated stability in social

19

composition and geographical range. On the other hand, alliance theory
emphasizes external ties and broad social networks (1998:16) such as those of
the “rendezvous pattern”.

Debate has centered on the nature of the behaviors reflected in style.
James Sackett has stressed the “social interaction” theory which states that
styles are learned in a social system that provides “isochrestic” sets of choices,
that is, there are a limited number of choices all of which can serve the same
purpose, and style attributes are shared among social entities to the extent
directly proportional to the frequency of interaction (Sackett 1982, Braun and
Plog 1982). Accordingly, some archaeologists have attempted to use style to
discern the nature of the social entities. For instance, David Brose, working with
Laurel-related Upper Peninsula Ware from the Summer Island Site on northern
Lake Michigan, attempted to determine the family residence pattern of the
occupants of the four houses at the site. He assumed that 1) women made the
pots, 2) culturally conditioned behavioral patterns were responsible for artifact
style patterning, and 3) the distribution of design attributes on pots can be used
to infer distribution of women sharing a ceramic learning experience (Brose
1970: 53). He suggested that a matrilocal residence pattern would produce
similar assemblages among the structures and between hearths within each
house. He found that the assemblages vary, however, and concluded that the
residence pattern was bilocal or patrilocal, probably the latter (1970: 55).

None of his assumptions are compelling. We have no direct evidence
that women, and only women, made the pots. We also have no direct evidence
that only passive cultural conditioning determined the choices of styles. Martin
Wobst has introduced an “information exchange” model of style that takes into
account active interaction reflecting changing preferences and message

signaling via emblemic style (Wobst 1977) both within and across social

20

boundaries. Laurel is easily identified across a very large region of central
North America, a fact that underscores wide interaction (Brose 1970: 68). It is
likely that the high degree of mobility among hunter-gatherers produced styles
reflective not only of familial cultural conditioning but also of band-level and
extra-band-levels of information exchange involving learning from kin and non-
kin alike. A number of social systems also affected interactions, including
marriage and residence patterns, sodalities, and alliances. Considering the
various types and scales of social interaction probably represented in the
ceramic record, it is unlikely that family structure in isolation from all other social
structures can be discerned in ceramic style. Therefore, this study will focus on
social groups larger than the family.

Debate among adherents of social interaction and information exchange
theories has been whether the source of style lies in a passive culturally
conditioned choice or in active exchange of emblemic messages (Sackett 1985,
Wressner 1985). While the ma of style is debatable, the functians of style are
clear: a particular style can be habitual and emblemic at different times
(Wiessner 1985, Conkey 1991: 13). Wiessner's study of Kalihari San groups
shows that they regarded their projectile point styles as habitual choices while
interacting within the multi-band language group and as emblemic mechanisms
across language groups. They functioned as indicators of formal group
membership, providing cohesion within the muIti-band social network that
provided the risk-buffering mechanism. They also functioned as active
boundary maintenance mechanisms across the different language groups
(Wiessner 1983: 272-3).

Christopher Carr has pointed out that adherence to either the “social
interaction” model or the “information exchange” model has been based more

on the particular sets of attributes chosen for analysis. The two models are, in

21

fact, complementary (Carr 1995: 156). Jerome Voss and Robert L. Young also
suggest that style is multidimensional, reflecting both active signaling and
passive enculturation, and the two theories predict similar distributional patterns
(Voss and Young 1995: 93-94). The analytical approach to style in this
dissertation is not designed to clarify whether styles functioned as passive
interaction or active signaling in any particular case.

One spatial pattern predicted by the“social interaction” model is the effect
of distance-decay: style results from a passive learning, or imitative process,
and the frequency of exchanged aspects of style decrease with distance/time. It
predicts unfettered style drift (Sackett 1982: 74).

However, researchers also point out that there are constraints affecting
style drift: 1) friction affect of distance (imperfect communication), 2) social
structure, 3) geographical conduits and barriers, 4) values of objects, 5)
population density, 6) cultural factors such as social distance (Clarke 1968,
Wobst 1977, Hodder 1979, Braun and Plog 1982). Robert D. Sack has termed
constraints such as these “realist geography”, a geographical theory that takes
into account perceptions of space based on active, subjective perspective and it
counters a passive distance decay model (Sack 1980).

Ian Hodder suggested that there are widespread similarities in style
among diffuse hunter-gatherers where there is considerable free movement
from camp to camp (Hodder 1979), while Guy Gibbon has noted that bands in
relative isolation produce different styles (Gibbon 1972). Hodder described the
formation of socially bounded groups with restricted interaction networks with
increasingly restricted styles. That is, there are both spatial and social factors at
work.

The notion that style is reflective of social interaction is significant for

Laurel studies. Ethnographic data depict the Lake Superior area as an

22

interaction zone not restricting ability to communicate; only social or linguistic
barriers would restrict communication. Therefore, we could expect an overall
similarity in styles with a chaining of attributes indicative of style drift among
freely interacting bands throughout the Laurel area. On the other hand, an
increasing restriction of style zones would be indicative of social change, new
boundary maintenance with changing social patterns as newly defined bands
increasingly restrict their social interaction or mark new ethnic boundaries. If
style zones emerge, we may see similarities between two style zones some
distance apart that indicate social interaction across a great expanse, perhaps
indicating the origin of one group out of the other, or a convergence of styles
indicating a newly formed social cohesion. For instance, William Engelbrecht
shows that early Seneca and Niagara Frontier (non-League people) had similar
styles while these same groups later possessed quite different styles, an
indication of a new Seneca bond with eastern Iroquois groups and a near
complete split from the western, Niagara groups as social interaction changed
(Engelbrecht 1978).

(5) Levels of Stylistic Analysis

A number of researchers have noted that stylistic studies must take into
account two levels of analysis, one dealing with the primary attributes of overall
design, the other with the secondary attributes of details that produce particular
motifs (e.g. Hardin 1970, Braun and Plog 1982, DeBoer 1991). These do not co-
vary and Laurel ceramic analyses would require two separate stages. Margaret
Hardin and Warren DeBoer noted that pottery production groups, be they kin-
based or community based, develop styles that share not only overall design
concepts but also the intricacies of the individual designs - the secondary
attributes - whereas spatially or socially distanced groups share only the

primary design concepts or none at all across social barriers. Using Wobst’s

23

study of style and social distancing along with Hardin’s and DeBoer’s studies of
production groups, we should expect that decorations among a local group will
be intricately similar, but meant to appeal to a group at a medium social
distance (e.g. feasters). Two groups at medium social distance will display only
primary and a few secondary attributes while a number of groups at long social
distance will display a minor number of primary attributes. Those across social

barriers will display no likenesses.

HYPOTHESES

The broad question addressed in this dissertation is about Laurel social
dynamics. Both the social interaction and the information exchange theories of
style predict that social groups express themselves through style; as social
networks change, stylistic features change. Thus fissions, amalgamations and
alliances will be observed in stylistic divergences and convergences. Analysis
of Laurel ceramic styles can be used to explore three significant questions: (1)
origins of Laurel, 2) the processes of interaction externally with its neighbors
and (3) the dynamics of change within Laurel.

Hypotheses can be addressed using two sets of data. One study
involves a comparison of Laurel to its neighbors in the Upper Great Lakes such
as Malmo, North Bay and Nokomis, and to the southeast such as Saugeen and
Point Peninsula. Perhaps the data would indicate greater social ties between
Laurel and a neighbor than hitherto recognized. The other involves the internal
trends within the Laurel ceramic data themselves; a comparison of
assemblages from a number of Laurel components should establish directions
of change.

A number of specific, testable hypotheses can be generated concerning

origins:

24

1) Spatial analysis of ceramic assemblages will reveal that the

earliest Laurel assemblage/complex is most similar to one or more

of its external neighbors. The test for strength of similarity through

style analysis will show that similarity is strong enough for the

conclusion that one or more neighbors had a major influence on

Laurel, or vice versa.

2) Temporal analysis will reveal that the neighbor(s) is earlier than

the earliest Laurel assemblage/complex and is therefore a good

candidate for an origin of Laurel style.

3) If there is no candidate for Hypotheses 1) and 2), then internal

Laurel dynamics will explain the development of Laurel.

A number of hypotheses can be generated concerning the internal
origins of Laurel:

4) One Laurel assemblage/complex is earlier than others, thus

representing an expansion model. a spread from one source within

the Laurel world.

5) All early assemblages/complexes are apparently

contemporaneous, thus representing a convergence model with

two or more origins of Laurel.

Hypotheses concerning Laurel origins and the question of Laurel
dynamics have long been interrelated in the minds of analysts. They have seen
the former question answered by the latter, as is apparent in the following
hypotheses that invoke trends within the Laurel data.

(a) Laurel originated to the west or northwest of the Laurel world (Wright
1967). This hypothesis was based on similarities between Laurel ceramics and
those of the Bering Straits area. However the proposition has been refuted, and

is no longer considered viable. Syms (1977), Mason (1969) and Meyer and

25

Hamilton (1994) have pointed out the lack of specific similarities in ceramics
between those two regions and the paucity of close interaction between Laurel
and Plains groups. The strongest refutation lies in the internal patterning of
Laurel that indicates a late entry of Laurel into Saskatchewan (Meyer and
Hamilton 1994).

(b) Laurel originated to the south of the Laurel world. Stoltman used
internal data to suggest that the earliest Laurel sites are at the southern extreme
of the Laurel world, and he suggested an origin for the ceramic style ultimately
in Illinois Hopewell (Stoltman 1973).

(0) Laurel originated in situ (Lugenbeal 1977). Lugenbeal noted that the
earliest Laurel sites do not cluster at the southern boundaries, therefore the
data do not indicate a south-north trend. He suggested this lack of trending
indicates in-place development. Mason has argued for a number of origins
within the Laurel universe (Mason 1991).

(d) Laurel originated to the southeast of the Laurel world. This argument
has not received strong support and has been dismissed by Wright (1967) and
Mason (1969) on the basis of ceramic evidence showing that the presence of
attributes such as pseudo-scallop shell stamping is strongest within Laurel and
weaker to the southeast, perhaps indicating a diffusion from Laurel to the
southeast. However, Cleland has supported the hypothesis on the basis of the
appearance of net sinkers in early Laurel, a technological innovation derived
from an earlier use of them in the Archaic of the Atlantic seaboard and the
Northeast (Cleland 19922769) and on the basis of Anishinabe traditions of a
migration from the eastern seacoast (Cleland, personal communication). The
tradition relates that the Anishinabe once lived on the shore of the Atlantic
Ocean where they first received the rites of the Midewiwin, or Grand Medicine

Society. They moved to the Montreal area setting up a new Mide lodge there,

26

then proceeded to the shore of Lake Huron, then to Sault Ste. Marie, then to La
Pointe and westward, each time establishing the Midewiwin in those stopping
places (Warren 1970, Dewdney 1975, Kohl 1985, Loew 2001).

The migration story from oral traditions must be given critical attention as
an hypothesis similar to the historian’s own interpretation of the past (Vansina
1985: 196). It necessitates independent verification in the form of a study of the
processes of interaction (diffusion and migration) with Laurel’s neighbors and
especially within the Laurel World itself from Lake Huron westward. Diffusion
involves down-the-Iine transmission of ideas by contact, including local
migration of small groups in seasonal rounds, trading, and via social
gatherings. Groups of one core area will exhibit more similarity with neighbors
in their secondary area than in their tertiary area. Migration involves movements
of people from a core area to a tertiary area.

Debate has arisen over the effects of migration on the archaeological
record. Migration no doubt occurred in prehistory but is difficult to distinguish
from diffusion. A few archaeologists have attempted to develop models of the
processes of migration that can be tested with archaeological data. Colin
Renfrew champions A.J. Ammennan and L. L. Cavalli-Sforza’s (1973) “wave of
advance” model for the spread of lndo-European-speaking farmers into western
Europe (Renfrew 1984). On the assumptions that they were segmentary
societies, or non-hierarchical, and that they experienced little natural or social
resistance in movement, those populations would have migrated outwards from
a point of origin in a modest, clinal fashion as populations increased. No long-
distance moves would have been necessary. He suggests that a new mode of
production, in this case farming, made possible an initial increase in population,
then waves of population expansion set in as populations continued to grow

and required more territory. A steady state would have been reached with the

27

carrying capacity of the the mode of production or at environmental barriers
(1984: 188-189).

Irving Rouse has attempted to demonstrate such movements in the
archaeological record (Rouse 1986). He has developed tests for migration
involving both the relationship among cultural complexes created by a single,
unimpeded expanding group and the relationship of the complex of an
expanding group encroaching upon another's territory.

First, he suggests that population movements are not seen in the
expansion of single cultural complexes because a number of factors can be
present that encourage modification of the complex: 1) as people migrate, they
encounter natural and cultural conditions and will modify their complex
accordingly; 2) replacement of on entire population takes time and complexes
are likely to experience cultural drift over that period, and 3) migrants can be
atypical of the parent population producing a “founder effect” in which only a
part of the entire parent population’s complex is reproduced by the migrating
group. As a result, the steps to test migration include recognition of a series of
related complexes with clear ties of development from one another. Plotted on a
map, these will reveal a geographical shift of the series from which we can infer
migration (Rouse 1986: 10-11).

We may also decide between migration and diffusion by comparing the
complex of the original inhabitants of a particular area with that of the new
complex in the same area. If the change is gradual and traits of the new
complex are integrated into the old, we may conclude that diffusion is present. If
the change is abrupt and the previous complex is replaced by the new complex,
we may conclude migration is present (1986: 12).

The question arises of how to treat hunter-gatherer systems in models of

migration. Ethnographic literature is replete with descriptions of the extreme

28

mobility of hunter-gatherers. For instance, the early-19th century Euro-American
captive John Tanner provided an account of his Odawa mother taking him west
from their home in northern Michigan to Lake of the Woods in northwestern
Ontario where they resettled. He later returned east to Sault Ste. Marie to
resettle there (James 1956). The ethnographic literature also gives accounts of
whole groups of people in possible migration situations. In 1852, Ojibway
leaders in Minnesota and Wisconsin proposed a resettlement-of whole bands
from those areas to Sault Ste. Marie (Chute 1998).

Taking ethnographically documented long-distance shifts into account,
David Anthony has posited a long-distance model of “Ieapfrogging” in which
migrations are directed to specific far-flung places, ignoring those between, as a
result not only of “push” factors such as population increase in the parent group
but also “pull” factors such as kin affiliation and opportunities for subsistence in
distant locations (Anthony 1990). In this model, similar assemblages need not
be in close proximity.

Both models face the problem of ambiguity in the data concerning levels
of similarity necessary to declare migration present as opposed to diffusion.
Researchers agree with the notion that the founder effect should create
assemblages in migrating populations somewhat dissimilar to the parent
assemblage because they would be carrying a narrower pool of variability than
the whole parent assemblage (Rouse 1986: 10; Anthony 1990: 903). However,
they establish no minimum levels of significance to test similarity. The data from
the Ballynacree Site in Kenora, Ontario, to be presented in this dissertation,
may provide such parameters for testing migration. The site contains three
houses with clearly defined assemblages for each. If we assume that the Laurel
assemblage from Ballynacree represents the parent population, then the three

house assemblages form samples drawn from that, or “founder effect” samples

29

that would be similar to a split-off, migrating sub-group of the parent. With
these, one can produce minimum levels of similarity to act as a test for migration
in the data. Two hypotheses can be stated:
6) Diffusion or short-distance migration will be supported if
similarity co-varies with distance; assemblages exhibiting greatest
similarity will be closest in space.
7) Long-distance migration and not diffusion will be supported if
similarity does not co-vary with distance; two assemblages from
western and eastern Laurel areas will exhibit similarity.
The internal development of Laurel after its origin is established can be
explored further with two hypotheses concerning culture change:
8) All early assemblages/complexes are similar with increasing
variability among later assemblages, representing a aivarganca
mo l.
9) All early assemblages/complexes are variable with increasing

similarity among later assemblages, representing the continuation

of the canvargence magal discussed in Hypothesis 5.

Laurel internal dynamics may be effective in a demonstration of the
formation of specific social groups. Analysts have postulated the development
of Late Woodland complexes, represented by the ceramics of Selkirk
(Rajnovich 1983, Meyer and Russell 1987), Blackduck (Wright 1967, Lugenbeal
1977), and Pine River (Holman 1984), as phenomena with a Laurel base and
the addition of Late Woodland traits. However, the evidence has been
ambiguous. We can hypothesize that Laurel stylistic dynamics will demonstrate
a formation of subregional social groups within the Laurel world which later

develop into the distinctive style zones of the Late Woodland period. The

30

subregions within Laurel will demonstrate divergence, convergence or parallel
developments leading toward the emergence of those Late Woodland zones.
The demonstration of a development of subregional groups within late Laurel
that match those of the early Late Woodland would strengthen arguments for an

in situ Laurel-Late Woodland transition.

31

2. PREVIOUS RESEARCH

For the last 60 years, researchers have been attempting to classify the
components of Laurel using mainly ceramic analysis of local areas or sub-
regions, but no classification is compelling and none helps to elucidate the
nature of interactions among the sub-regions. A discussion of research to date
can be divided into two general topics: internal relationships of Laurel
components, or the dynamics of Laurel itself, and the external relationships of
Laurel to its neighbors in the Northern Tier Middle Woodland, including its
origins, interactions and developments into the Late Woodland Period.

Archaeological interest in the area began in the late 19th century mainly
as antiquarian fascination with burial mounds and associated ancient relics and
stirred by debates about the identity of the mysterious “Mound Builders”, an
hypothesized race of people thought to have inhabited eastern North America
before the present Indian nations (MacNeish 1958: 13; Lugenbeal 1976: 40;
Arthurs 1986: 23). Cyrus Thomas debunked the hypothesis in his famous report
to the Bureau of American Ethnology in 1894, but antiquarian interest in the
mounds continued into the early 20th century (Branstner n.d.: 19). The sole
contribution to Laurel research from these endeavors was a series of reports on
the locations of some of the most extensive sites in the area.

Laurel research moved into a descriptive stage in the 1940s and 19503
with the publication of Wilford’s classifications of prehistoric cultures in
Minnesota These included Laurel in a classification based on the Midwestern
Taxonomic System. Basing his conclusions on stratigraphy and ceramic
similarities or dissimilarities with Havana-like wares, Wilford considered Laurel
to be a focus of the Rainy River Aspect of the Lake Michigan Phase, a taxon of
the Middle Woodland Period distinct from the Hopewellian Phase in Minnesota

(Wilford 1955: 131). He also defined the Malmo focus as Laurel's closest
32

relative in Minnesota (1955: 135). The focus was located in the Mille Lacs
region of central Minnesota and subsequent investigations there have
unearthed very little evidence of Malmo compared to the thousands of Laurel
ceramics, and Gibbon (1975: 19; 2000: personal communication). has
questioned the concept of Malmo. Later researchers have also noted the
discrepancies of Wilford’s taxonomy in light of radiocarbon dating.

In. 1958, MacNeish extended the culture-historical investigation to
Manitoba where his short survey and excavations led to a classification of
Laurel into the Anderson and Nutimik foci, one earlier than the other. Later
research has shown that these may both be late in the Laurel sequence, and
Nutimik may in fact contain some Late Woodland material.

From the 19603 to the 19803, many surveys of Northern Ontario,
Manitoba and Saskatchewan located hundreds of Laurel components across
the Canadian Shield from the southern edge of the Hudson Bay Lowlands
(Pilon 1987: 180) to the international boundary (e.g. Wright 1967; Dawson
1980, 1981; Wiersum and Tisdale 1977, Meyer 1983; see also Reid and
Rajnovich 1991). While serving to flesh out the culture history of Laurel with
large amounts of data, some investigations refined Laurel chronology with
radiocarbon dates, finding that sites range in age from 1501;165 B.C. (Reid and
Rajnovich 1991: 202) in the Lake of the Woods area to as late as A.D.1305¢70
in Saskatchewan (Meyer etal 1988: 37).

During the same time period, several investigators undertook major
analyses of ceramics to define local sequences. By the 19603, researchers
began studies of variability within the Laurel sequence, emphasizing questions
of classification including both Laurel taxonomy and ceramic typology, two
closely intertwined subjects. Each study was confined to the western or eastern

Laurel areas.

33

WESTERN LAUREL

In 1973, James B. Stoltman undertook a reanalysis of Wilford’s
collections from the Smith, McKinstry and Pike Bay mounds housed at the
University of Minnesota and he added an assemblage from the Pearson village
site. He used the Willey and Phillips taxonomic system to define three “phases”
within a Laurel “culture” based on his ceramic analysis. Stoltman used the chi-
square statistic to discover correlated modes (Spaulding 1953) in order to
define types. He then used the Brainerd-Robinson technique (Robinson 1951)
and Double-Link Close Proximity analysis (Renfrew and Sterud 1969) to seriate
his assemblages into what he presumed were temporal phases - Pike Bay,
McKinstry and Smith (Stoltman 1973: 4-5). In 1974 he refined his analysis by
publishing results of his excavation of a portion of the McKinstry Mound #1
where he was successful in obtaining data from an undisturbed stratigraphic
sequence relating to his three phases. Reliable radiocarbon dates also
confirmed his hypothesis that the differences in ceramic content of the three
phases were age related rather than spatially significant. Dates from the lowest
strata containing Pike Bay Phase material were 30 8.0.145 (Stoltman 1974: 80)
to AD. 120155 (1974: 86). The McKinstry Phase date was A.D.560i55
(1974:83). He obtained no material of the Smith Phase but estimated its date to
be after A.D.500 and before about A.D.900 (1974: 88).

One encounters a number of problems in attempting to adapt Stoltman’s
data to a new study of Laurel dynamics. First, he used rim sherds as the units of
analysis rather than vessels. He recognized the unsuitability of that procedure
but he had limited time to commit to the study and used that procedure as a
time-saving device (1973: 47). His results, then, are not comparable to those

using vessel counts. Second, he used a limited list of attributes (“modes” in his

34

study), mainly exterior rim decorative technique, and some attributes such as
cord-wrapped stick and trailing were rejected on the assumption that they are
intrusive or idiosyncratic (1973: 62). Third, he found correlation between the
push-pull technique and absence of bosses/punctates so he devised a Laurel
Oblique type regardless of tool technique. Researchers have found that type to
be unworkable in that it does not specify tool type. Thus, in Stoltman’s data,
there is little discussion of the variation within Laurel Oblique, no recognition of
a Laurel Cord-wrapped Stick type, and no discussion of the presence of
minority types that would be helpful in intra- or inter-regional comparisons.

Stoltman’s work was not intended to provide a formal study of inter-
regional comparisons but he did recognize trends. For instance, he noted that
Naomikong Point of eastern Laurel would be intermediary between the Pike
Bay and McKinstry phases of Minnesota Laurel “although, depending upon how
many of Janzen’s Pseudo-scallop Shell rims are classifiable into our Laurel
Oblique type, the site may be much more Pike Bay-like than McKinstry-like”
(1973: 119).

As for the Willey and Phillips system he proposed for Laurel, he
recognized difficulties with that two-tier approach. If one accepts his phases
subsumed within a Laurel “Culture”, he asked, what then do we call Mason’s
“Northern Tier Middle Woodland”? He suggested this supraculture level of
taxonomy be termed a “Culture Group” or similar inclusive phrase (1973: 113).
In any case, he neither assigned a name to it nor formally analyzed it.

Edward N. Lugenbeal continued Stoltman’s approach to western Laurel.
In his 1976 dissertation at the University of Wisconsin-Madison, he presented
analysis of his excavations at the Smith Site (aka Grand Mound Site) where he
had hoped to open large blocks of units to study economic and social patterns

of the Laurel habitation area However, demonstrations by a militant Indian

35

group precluded that endeavor and necessitated a shorter study of smaller units
with a new plan to address only a refinement of Stoltman's culture history for the
area (Lugenbeal 1976: 98). He combined his results from the Smith Site village
with his reanalysis of assemblages from nine other sites in the region including
southeastern Manitoba, the Rainy River area of Ontario, and northern
Minnesota, comprising 356 vessels as the units of analysis (1976: 424). He
used the same methodology as Stoltman (Lugenbeal 1976: 420-422) but
recognized the limitations in Stoltman’s list of attributes and included a slightly
larger array of his own. He concluded that Stoltman’s types were essentially
supported with the addition of a Laurel Cord-wrapped Stick type and a splitting
of Laurel Oblique into Laurel Dragged Oblique and Laurel Undragged Oblique
(1976: 460-462).

He emphasized the association of the definitions of ceramic types with
the definition of phases - one being only at a higher taxonomic level than the
other (Lugenbeal 1976: 572) - and, on the basis of his expanded typology, he
recognized four rather than Stoltman’s three phases within Laurel. His work at
the Smith Site coincided with Stoltman’s definition of the Smith Phase and
added radiocarbon dates of AD. 4801-60 to AD. 760155 for the phase. His
expanded reanalysis of regional data led him to introduce a later phase, Hungry
Hall, based on the presence of cord-wrapped stick decoration and decrease in
other techniques at the type site. Lugenbeal’s dissertation represents an
extensive and comprehensive study of western Laurel, the results of which have
since been supported by independent evidence (see below). However
Lugenbeal did not undertake inter-regional analyses of Laurel.

In the 19803, K. C. A. Dawson of Lakehead University excavated two
sites in the western Lake Superior area, the MacGillivray Site on Whitefish Lake

southwest of Thunder Bay, Ontario, and the Wabinosh River Site on the west

36

shore of Lake Nipigon, north of Lake Superior (Dawson 1980, 1981). The
former site contained a burial mound with central pit and log-crib construction
dating to AD. 201200, but containing no ceramics, and a village area where
sherds of 59 Laurel vessels were excavated. The mound did not overlie the
village area and Dawson concluded the habitation could be later than the
burial (1980: 6). The Wabinosh River Site contained two Laurel habitations, an
early one in the low terrace of the site and a later one in a higher terrace dated
to AD. 855:180 and AD. 1240:175. He seriated these sites with four other
Canadian Laurel sites using exterior rim motif as the unit of analysis (Dawson
1981: 36) and produced a seriation that is consistent with Stoltman’s and
Lugenbeal’s suggested sequence of Laurel decorative dynamics. The sites
included Tailrace Bay in Manitoba, Long Sault and Hungry Hall on the Rainy
River, and Heron Bay on Lake Superior more than 150 km southeast of Lake
Nipigon. Dawson recognized the possibility of spatial as well as temporal
dimensions in his seriation (1981: 28). He did not devise phases on the basis of
his limited study but confirmed the general pattern of Laurel temporal variability.

David Arthurs excavated and analyzed a portion of the Long Sault Site
(aka Manitou Mounds) for his Masters thesis at the University of Manitoba
(Arthurs 1986). He also conducted a seriation of Laurel components based on
some of Stoltman’s and Lugenbeal’s sites and using their typology. His analysis
concluded that Long Sault falls into the McKinstry Phase of Stoltman’s
taxonomy (1986: 118). Reid and Rajnovich of the Ontario Ministry of Culture and
Communications conducted an expanded seriation using 16 sites from both
Canada and the United States but with reduced criteria The study was
independent of the Stoltman-Lugenbeal approach using types; rather exterior
rim decorative technique was the unit ofianalysis. The study confirmed the

Minnesota sequence (Reid and Rajnovich 1991: 193), but it had problems of its

37

own. It was based on limited data concerning exterior rim decoration and it
placed Long Sault at the earliest end of the seriation, a placement disputed by
Arthurs (personal communication). This dissertation includes a reconsideration
of the Long Sault Laurel assemblage as well as two others from northwestern
Ontario - Fisk considered early Laurel, and Ballynacree considered late Laurel
(Figure 2). These sites were excavated by Reid and Rajnovich for the Ontario
government. The Reid-Rajnovich study also addressed the need for a new
taxonomy for Laurel, one suggested by E. Leigh Syms (1977) that would
include more taxa than the Willey and Phillips system that Stoltman had
previously found wanting in light of Laurel studies. We suggested that Laurel be
termed a configuration comprising a number of composites, including a
Boundary Waters Composite containing the Stoltman-Lugenbeal phases, now
termed complexes (Reid and Rajnovich 1991: 224-228). We suggested that a
number of composites would define Laurel, including possibly an eastern one,
but our study confined itself to the western Laurel area and did not attempt a

discussion of the relationships among the composites.

EASTERN LAUREL

From the 19603 to the 19803, researchers have been stymied by the lack
of local systematics for eastern Laurel, mostly due to paucity of material on
which to build taxa.

In Ontario, two major survey projects were undertaken. James V. Wright
of the National Museums of Canada conducted a survey of the Canadian
Shield in the early 19603, reporting 16 Laurel components, eight of them near
the north shore of Lake Superior in northeastern Ontario. He confined his
analysis to a discussion of Laurel as a whole, as a “tradition” with long temporal

and wide spatial continuities, and refrained from attempting a discussion of the

38

3x9 35 E 88:86 8% .053 .o 5333 ”N 05ml

2

 

         

llllllllllll
II-
1n.-
o

  

 

 

 

..l 3

. t...

coeduczimlrwi.

iaA/r .
c .

 

VMEIHDJszys

N

 

\

Lamp—.232

 

it?

 

39

phenomenon with more precision concerning internal dynamics because of the
paucity of evidence at the time (Wright 1967: 2-3). He even dismissed
MacNeish’s Anderson and Nutimik fool for Manitoba because they lacked
sufficient data for confirmation and both appeared to represent late Laurel
occupations (1967: 107). Moreover, he confined his study of ceramics to
attribute analysis rather than formulation of typological definitions because of
the small number of sherds recovered. The Heron Bay Site, the largest
assemblage, consisted of only 123 rim sherds; the Sand River site, the smallest,
contained only one rim sherd. In the 19703 and 19803, Thor Conway conducted
surveys of northeastern Ontario for the Ontario government, recording many
small Laurel components and one large one, the Whitefish Island Site at Sault
Ste. Marie (Figure 2). He failed to publish data on these, and the Whitefish
Island material, now housed at the Ontario culture ministry’s office in Thunder
Bay is, presented comprehensively for the first time in this dissertation.

In Michigan, more than 70 Middle Woodland components have been
reported for the Upper Peninsula and the northern section of the Lower
Peninsula, but research on Laurel components has been confined to only a few
major sites in the area, a region that has received far less intensive
investigation than the western Laurel region. Moreover, there has been little
attempt to devise a taxonomic scheme for the eastern Laurel data. In the 19603,
a number of researchers located and collected data on Laurel occupation of the
northern Michigan area Donald E. Janzen, working on a dissertation at the
University of Michigan, excavated the Naomikong Point Site (Figure 2), a large
Laurel component on the south shore of Lake Superior, and recovered 133,000
sherds (Janzen 1968: 35), far more than Wright’s Ontario assemblages
combined. David S. Brose, also working on a dissertation at the University of

Michigan, investigated the Summer Island Site at the mouth of Green Bay in the

40

Upper Peninsula, a Laurel-related site, recovering 4,451 sherds (Brose 1970:
47). Both researchers had few other assemblages in the eastern Laurel area
with which to compare their samples (Janzen 1968: 100; Brose 1970: 94) and
they were forced into broad areal comparisons of assemblages from across the
Great Lakes. They also lacked chronological control, producing what Wright
termed, and Brose reiterated, as reconstructions “speculative to the point of
fantasy” (Wright 1967:126; Brose 1970: 94). Ceramics from Naomikong Point
housed at the University of Michigan have been reanalyzed for this dissertation.

In 1973, Jeffrey Richner reported material from the Winter site on the
Garden Peninsula. His Masters thesis for the University of Western Michigan
does not provide data on the ceramics and my attempt to relocate the collection
failed. In 1974 and 1975, \Mlliam A. Lovis of Michigan State University
excavated the Portage Site near Cheboygan, Michigan (Figure 2); the ceramics
from there are housed at Michigan State University and they have been recently
reported by Lovis, Rajnovich and Bartley (1998). Those data have been
incorporated into this dissertation.

By the mid-19703 investigators began to turn their attention to preliminary
statements concerning Laurel dynamics in Michigan. James E. Fitting, on behalf
of the State of Michigan History Division, reported on excavations at the Gyftakis
(Figure 2) and McGreggor Sites in St. Ignace at the straits of Mackinac. He
concluded that the sites represent two phases of Laurel (n.d. 182: 1978: 112).
His findings are reported comprehensively in an unpublished manuscript a
copy of which is at Michigan State University. There are two problems: first, my
reanalysis of the ceramics, now housed at the State Archaeologist’s laboratory,
Michigan Historical Center, does not agree with the figures presented by Fitting
for his phases and, second, I could not relocate the McGregor Site ceramics.

The analysable Laurel rims from the Gyftakis Site have been included in this

41

dissertation

In 1984, Margaret B. Holman of Michigan State University presented data
on Pine River Ware from the Pine River Channel Site in Charlevoix County
which she demonstrated represented a phase transitional between Laurel and
Late Woodland Mackinac Ware, dating after about AD. 600 (Holman 1984).

In the early 19903, Charles E. Cleland and Christine Branstner of
Michigan State University investigated the Cloudman Site on Drummond Island
(Figure 2), a Laurel habitation and mound site, and Branstner provided
preliminary data in an unpublished report in 1995. The collections, housed at
Michigan State University, are analyzed in this dissertation.

In summary, the assemblages chosen for analysis in this study are from
the Ballynacree Site in Kenora, Ontario, the Long Sault Site on the Canadian
side of the Rainy River, and the Fisk Site north of Kenora, all assemblages from
the western Laurel area. Also included are the Naomikong Point Site on the
Michigan shore of Lake Superior, the Whitefish Island Site in Sault Ste. Marie,
Ontario, the Cloudman Site on Drummond Island, Michigan, the Gyftakis Site in
St. Ignace, Michigan, and the Portage Site near Petoskey, Michigan, all
assemblages from the eastern Laurel area. All are comparable in that they meet
the size requirement for the study (greater than 20 vessels) and all appear to
represent habitations. The Long Sault and Cloudman sites are situated near
burial mounds that may be Laurel, however, the ceramic assemblages are from
habitation areas of those sites. The Gyftakis Site assemblage contains ceramics
that are from both a burial pit and habitation area nearby, but mainly from burial
context. The study may suffer from distortion on that account, as well as by the
fact that seasonality and duration of the sites chosen are not known. However,
without these assemblages, the study cannot proceed for lack of available data.

Other Laurel or Laurel-like assemblages have been discovered at sites

42

such as the Goodwin-Gresham Site in losco County, Michigan (Fitting et al
1969, Fitting 1975), Arrowhead Drive on Bois Blanc Island, Michigan (Bettarel
and Harrison 1962), and the EckdahI-Goodreau Site in Schoolcraft County,
Michigan (Fitting 1975), parts of which are housed at the University of Michigan
and Michigan State University. The assemblages from these sites are too small
for inclusion in this analysis. In addition, Fitting has suggested that the
Goodwin-Gresham data may represent a mixing of two or more components

(Fitting 1975: 142).

43

3. METHODOLOGY

Researchers have identified Laurel ceramics from Saskatchewan to
Michigan; while vessel styles vary from place to place and time to time, they all
share distinctive Laurel attributes, and we assume that they are somehow
related. This study uses vessels as the units of analysis and assumes that the
relationships among them will take the form of polythetic sets in which each
vessel shares a majority of its attributes with some others in the set but no
attribute is necessary in order to define the set This is opposed to the
assumption of monothetic sets in which vessels share a group of explicitly
necessary attributes. This dissertation will use the Laurel assemblages from
each of the eastern and western areas to conduct a ceramic stylistic analysis
based on presence or absence (attribute states) of more than 90 variables
(attributes) in each assemblage.

The units of analysis are vessel counts, not rim sherd counts. The vessels
are depicted in ceramic charts for each assemblage in the appendices (see
Appendix A for the legend). The chart, originally devised by Ritchie and
MacNeish (1949) and modified by Reid for Laurel to include more detail (see
Rajnovich, Reid and Shay 1982), depicts vessel profiles, exteriors to the left, at
the top of the chart. The side columns present the exterior, lip and interior
decoration for each vessel or group of vessels that are identical. The interior
cells provide frequencies of vessels with the indicated profile and decoration.

By observation of the attributes depicted on the ceramic charts, I
developed a spreadsheet to code data for each vessel (an example of the data
sheet for each assemblage is shown in Appendix B). The variables are listed by
field: interior, lip, exterior lip-rim juncture, exterior upper rim, exterior lower rim.

Variables include surface treatment, decorative technique, and decorative motif

44

for each field. For example, “exterior upper rim linear stamps” is one attribute.
Presence (listed on the spreadsheet as “1”) or absence (listed as “0”) is the
attribute state. Exterior shoulder technique and motif were also coded but not
used in the analysis because of low frequencies of extant examples.

The necessary first step in intersite comparisons is the temporal ordering
of the eastern Laurel components in order to isolate representative
assemblages of early-to-late Laurel ceramics of the subregion. This facilitates
the second step involving a comparison of eastern and western assemblages of
comparable time periods. The major task is to examine both the primary and
secondary decorative characteristics, the sensitive indicators of interaction as
predicted by the style theory discussed above.

Debate over the past four decades has focused on the evaluation of
methodologies for the discovery of sets, involving advocates of monothetic vs.
polythetic analysis. Robert Whallon Jr. (1972) proposed a monothetic
approach, basing his decision on ethnographic accounts purporting to discover
monothetic sets in the cognition of native informants. He cited Duane G.
Metzger and Gerald E. Williams’ (1966) study of Maya classification of firewood
in Chiapas, Mexico. However, a number of cultural anthropologists (Black 1968,
Kaplan and Levine 1981, Kempton 1981) have warned that field methods may
inadvertently “lead” informants, producing unreliable results. In the Chiapas
study, the authors began by providing a set of frames, or carefully constructed
questions and possible responses “which led to the establishment of one
contrast set“ (Metzger and Williams 1966: 391), enforcing a monothetic
construct assumed for the data without evaluating its ‘emic’ thth.

Kempton (1981), Kaplan and Levine (1981), and others opted for
multivariate analysis allowing for both contrastive and integrative sets using

measures of similarity. Kempton found that his Mexican study group lumped

45

pottery into polythetic sets with each category displaying a ”family of features”.
“Like physical resemblances among members of a family, some members share
some features, other members share other features, but no features are shared
by all members” (Kempton 1981: 15). He demonstrated further that the potters
allowed for more variation within sets than non-potters (1981: 132). Kaplan and
Levine’s analysis of folk taxonomy of Mexican pottery used a multivariate
approach and discovered sets that have cultural significance in terms of folk
belief. They made an interesting observation that has implications for
archaeological research: informants did not conceptualize the whole picture of
the native system of classification; each person understood only portions of the
whole (Kaplan and Levine 1981: 875). The notion is similar to Kehoe and
Kehoe’s (1973) suggestion that members of each generation participate in the
cultural “cognitive map” which is, in fact, a collection of slightly differing
individual “maps”. A polythetic analysis has the advantage over monothetic
approaches of discovering such overlapping sets and also has the ability to
isolate true monothetic sets at the same time.

Kaplan and Levine advocated cluster analysis as one powerful tool to
assist in the determination of cognitive structures in an assemblage of data H.
Charles Romesburg (1984: 39) has also championed cluster analysis as a tool
for discovering assemblage patterning. This study is intended to clarify the
ceramic stylistic patterning of each assemblage. From those, the intersite
comparisons allow for tests of the hypotheses concerning social dynamics.

There have been questions concerning the most efficient statistical tool
for multivariate analyses, the choices usually consisting of multidimensional
scaling that produces scatter diagrams, and cluster analysis producing
hierarchical trees. R.G. Matson and D. L True (1974) tested the two approaches

on the same sets of data and found both to be useful. This study will use

46

clustering on the assumption that the trees produced are better suited to
analysis of large numbers of items; the dendrogram clarifies the pattern among
hundreds of objects better than a scatter diagram does.

Numerical models aid in understanding complex Laurel data because
they help to simplify and synthesize them, allowing for productive interpretation
(Johnson 1972). Cluster analysis, in particular, has been considered an
appropriate archaeological tool because it can handle a multitude of objects
and displays resemblance coefficients visually - as easily understood
dendrograms (Clarke 1968, Hodson 1969, Engelbrecht 1974). It allows for a
grasp of the structure of similarity among many objects; for instance, chaining
(total seriation without clustering), clustering, or chaining within clusters may all
occur. These indicate the internal structure of the clusters as well as, the
structure of the whole assemblage (Romesburg 1984). Clustering is appropriate
for a Laurel ceramic study for three reasons.

1) It provides scaling necessary for a stylistic study based on theory that
assumes scaling. It provides higher- and lower-order linkages that should
reflect primary and secondary decorative characteristics. The higher-order
clusters, as suggested by style theory, will be. shared by several groups, while
the lower-order clusters should contain secondary attributes shared by social
units in close contact.

2) Clustering can provide a polythetic analysis in which all attributes are
initially assigned equal weight. In this stylistic analysis, we will not know at the
outset which attributes are primary or secondary.

3) Clustering can be undertaken using either the Q-mode or R-mode
technique. Hodson has advocated Q—mode analysis that clusters objects rather
than R-mode analysis that clusters attributes. He demonstrates that attribute

analysis obscures some categories; attributes may appear to be “randomly

47

associated” when classes do indeed exist, and low frequencies of some
attribute associations lead to the idea that they are atypical or transitional.
Previous Laurel research has either defined types arbitrarily (MacNeish 1958,
Janzen 1968) or used attribute analysis using bivariate analysis (Stoltman
1973, Lugenbeal 1976, Fitting n.d.). The attribute studies follow Spaulding’s
(1953) approach using the chi-square statistic to measure significance of
association between pairs of attributes. However, Lugenbeal found some
problems with the approach, suggesting the pairings are somewhat arbitrary
(1976: 448) and that reformulation of the chi-square tests using various pairings
would change the results (1976: 457). He suggested that multivariate statistics
would correct the problems of bivariate analysis. This study will follow Hodson’s
and Lugenbeal’s suggestion and will use multivariate clustering of objects
(vessels) as opposed to attributes (decorative techniques or motifs).

Preliminary observation of the data suggests that Laurel ceramics have
fewer attributes in the early stages and increasing diversity later. It has been
suggested that this will produce high coefficients of similarity among the earliest
ceramics and lower ones later as diversity increases. With a larger array of
choice, two polythetic sets drawn from the same population of choices can be
different. A debate between Michelle Hegmon (1986) and David P. Braun and
Stephen Plog (1982) is pertinent here. The latter researchers hypothesized that
increasing interaction often produces increasing homogeneity across
assemblages, while the former analyst suggested that increased interaction can
produce increasing homogeneity and diversity of attributes; that is,
assemblages are alike and display a large array of attributes. Whatever the
case, both sides in the debate agree that increasing similarity results. For
instance, Hegmon found both increasing similarity and increasing attribute

diversity among the ceramic assemblages of Black Mesa, Arizona, as

48

populations increased from about A.D. 850-1050. She inferred increased
interaction. The ceramics of the late Laurel Hungry Hall phase are
characterized by diversity (Reid and Rajnovich 1991: Figures 11-15), yet
assemblages from the Ballynacree Site at the north end of Lake of the Woods
and the Hungry Hall Site, more than 100 miles away at the south end of the
Lake, are about 80% similar, a situation like Hegmon’s. in which diversity does
not decrease similarity.

Concerning technique, almost all Laurel analysts have commented that
some techniques are sometimes difficult to distinguish from others; the
technique of dragged stamping, also referred to as push-pull or stab-and-drag,
is sometimes indistinguishable from undragged stamping, and sometimes both
are used at the same time. Similarly, the technique of pseudo-scallop shell
stamping can be transformed into simple dentate stamping. I have opted for the
most frequent technique on each vessel - if some decorative elements are not
dragged but most are, the technique is considered dragged.

Concerning motif, one pattern has been referred to as banked stamped,
a group of closely spaced horizontal lines encircling the vessel in vertical
banks. I have distinguished banked stamping from unbanked stamping by the
spacing; banked stamped in this study refers to groups of closely spaced lines
that are literally banked together, touching or almost touching each other, while
unbanked stamping refers to lines with spacing between them.

In this study, a distinction is made between exterior rim-lip decoration,
often called lip ticking, from exterior upper rim decoration on the basis of their
differences in technique, size or direction of application.

The analysis involves several steps:

1) Sorting of rim sherds into minimum number of vessels; the unit of

analysis in this study is the vessel. Each vessel is then depicted on a ceramic

49

chart for the assemblage.

2) Coding of attributes in a spreadsheet to be used in the clustering
procedure. This step is accomplished by observation of the attributes depicted
on the ceramic chart.

The spreadsheet is then used in the cluster statistics program, written by
James J. Rajnovich of the Computer Science Department, Algoma University
College. It performs the following steps:

3) Formation of a chart of coefficients of similarity. The measure of
similarity used in this study is the Jaccard Coefficient which ignores 0-0
matches and counts only 1-1, 1-0 and 0-1 (Romesburg 1984: 143). In other
words, it ignores the cases of matches between vessels on the basis of absence
of an attribute on both vessels.

4) A clustering technique based on those coefficients. The clustering
technique is the Unweighted Pair-Group Method Using Arithmetic Averages
(UPGMA) in which the average of the distance measures of objects in the group
is used for comparison to the next objects to be clustered. The results are
shown in a dendrogram for each site (an example is in Appendix O) and the
clusters are described in a table for each site.

5) A correlation coefficient to test the accuracy of the clustering process.
Each cluster tree has a Cophenetic Correlation Coefficient, or Pearson
Coefficient, which compares the strength of the resemblance coefficients of step
1 with the computed values on the dendrogram at step 2 which are slightly
altered through the process of averaging the values in the group. A perfect
concordance is 1.0; no concordance is 0.0. A generally accepted value is .8
(Romesburg 1984: 27). All the trees in this study meet this requirement.

The level at which significant clusters are produced along the height of

the tree has been a matter of debate among statisticians. Mark S. Aldenderfor

50

and Roger K. Blashfield (1984) have said that the issue of the cut-off points in
the dendrogram is among the unsolved problems of cluster analysis
(Aldenderfor and Blashfield 1984: 53). Both they and Romesburg have
suggested that the problem of defining the number of clusters has not been
important in analyses with a goal to explore general patterns of relationship
among entities (Aldenderfor and Blashfield 1984: 54, Romesburg 1984: 213).
Aldenderfor and Blashfield present some formal statistical tests for addressing
the cut-off points, however they caution that few are widely accepted (1984:
58).Rather, they and Romesburg suggest that the cut-off can be done by
inspection using two methods: 1) by choosing the number of clusters at a level
above which no new significant information is portrayed on the dendrogram
(Romesburg 1984: 213-215) , or 2) by choosing the number of clusters at a
level displaying the greatest width of range from that below it; that is, discover a
significant “jump” in the value of the coefficient where two relatively dissimilar
clusters merge and choose the cut-off point below that (Aldenderfor and
Blashfield 1984: 57).

In this study, two levels are considered significant in that they may
display similarity at a cluster level and a sub-cluster level possibly indicating not
only general similarity but also more close-range similarity. Therefore, I chose to
cut the primary group at a level of about .5 similarity and subgroups at about .35
or less. Outliers are single vessels that cut at the group level and could form
another group at a different site. If a single vessel splits at about .35 it is called a

subgroup as it is not different enough from the group to be termed an outlier.

51

4. THE SITES

The assemblages chosen for analysis in this study are from both the
eastern and western regions of the Laurel world. The eastern assemblages are
from the Naomikong Point Site on the Michigan shore of Lake Superior, the
Whitefish Island Site in Sault Ste. Marie, Ontario, the Cloudman Site on
Drummond Island, Michigan, the Gyftakis Site in St. Ignace, Michigan, and the
Portage Site near Petoskey, Michigan. As stated previously, they were chosen
because they are all comparable in that they meet the size requirement for the
study (greater than 20 vessels); they all appear to represent, at least in part,
habitations: and the assemblages have been kept intact for study by curators at
the Office of the State Archaeologist of Michigan, the University of Michigan,
and Michigan State University.

The western group is composed of three previously analyzed ceramic
assemblages known to represent a series from early to late Laurel in the
Boundary Waters area. They, too, meet the size requirement for the study and
are from habitation areas. The Fisk Site, reported by Rajnovich, Reid and Shay
(1982), the Ballynacree Site, reported by Reid and Rajnovich (1991), and the
Long Sault Site, reported by Arthurs (1986), were chosen over other Laurel
sites of the western area for two reasons: 1) they were fully reported with the
use of master ceramic charts so that the data from each could be incorporated
into this study; and 2) they were chosen to represent early, middle and late
assemblages of the western area that would provide a comparison to an
eastern seriation produced by this study. Reid and I conducted a seriation of 11
western Laurel assemblages using only exterior rim decorative technique, and
found the Fisk assemblage to be among the earliest and the Ballynacree

assemblage to be the latest (Reid and Rajnovich 1991: 225). In addition, the

52

Fisk Laurel component was radiocarbon dated to AD. 50:115 from charcoal
from a hearth in a possible house outline, a date among the earliest for Laurel
in the western area (Rajnovich, Reid and Shay 1982: 97). Ballynacree
produced three dates of 1240145 from charred wood on a Laurel house floor,
1240:65 from charcoal from the central hearth of the house, and 1270:55 from
charcoal of the floor of the house (Reid and Rajnovich 1991: 198). The dates
are very late for Laurel, considerecd too late by some researchers, but the
contexts of the samples were clear; both the hearth and house floor contained
Laurel ceramics. In the Reid and Rajnovich study, the Long Sault Site seriated
early, but Arthurs, who produced a Master’s thesis on the site, disagreed and
suggested a fuller analysis would reveal Long Sault to be a middle Laurel
assemblage (Arthurs: personal communication). The present study is designed
to test that hypothesis using a larger data set than the Reid and Rajnovich study.
Table 1 provides frequencies for each assemblage used in the study.
The Richter Site in Door County, Wisconsin, and the Squirrel Dam Site in
Oneida County, Wisconsin, have also been included to provide some
comparisons of Laurel with its neighbors in the eastern area. Both sites are near
the Upper Peninsula of Michigan. Richter is a North Bay Phase component and
Squirrel Dam is a Nokomis Phase assemblage.
A summary of each site is given in this chapter. A fuller discussion is
presented in the appendices, including the ceramic charts, cluster descriptions,

maps, and ceramic illustrations.

53

Table 1: Frequencies of Vessels in Each Assemblage

 

 

 

§IIE NW
Eastern Laurel

Whitefish Island 80
Naomikong Point 162
Cloudman 29
Gyftakis 61
Ponage 39
Western Laurel

Fisk 23
Long Sault 79
Ballynacree 112
Richter 38
Squirrel Dam 33
IQIAL- fl

 

Whitefish Island (CdIc-2) (See Appendix C) - The site extends across a
large island in the rapids of the St. Mary’s River between the twin cities of Sault
Ste. Marie, Ontario and Michigan. Thor Conway of the Ontario Ministry of
Culture and Recreation conducted test excavations from 1976 to 1979 as part of
a Sault-area survey. Conway produced only preliminary statements of the
results, commenting that the Laurel component was confined to one area of the
site (Conway n.d.: l). Site maps were lost during a move of the ministry office.
The artifacts are now housed in the archaeology laboratory of the Ontario
Ministry of Culture, Tourism and Recreation in Thunder Bay.

John W. Pollock, an archaeological consultant, provided the ministry with
a map of the excavation areas as part of a resource evaluation study in 1983. It
reveals that the site is horizontally stratified, with Laurel in one area in the
center of the island at its highest point. The island is composed mainly of
cobbles with no apparent vertical stratigraphy (Conway n.d.: 3). No radiocarbon
dates were obtained.

The 80 vessels of sufficient size for this analysis reveal a surprising
homogeneity in motifs. Statistical analysis produced seven clusters, five of

54

which are banked stamped. Some of the vessels have identical or near-
identical decoration. For instance, five vessels all have right-leaning banked
linear stamps, tick marks at the lip-rim juncture, and interior vertical linear
stamps. Another six pots have exactly the same exterior decoration but lack
interior decoration. Banked stamping is an early trait of Laurel Ware in its
western distribution in Minnesota and northwestern Ontario, so it is possible that
Whitefish Island belongs at the early end of the seriation of eastern sites.

The Whitefish Island assemblage is most unusual in that the majority of
vessels are in a single cluster: 52.5 per cent of the vessels are in Group 1 alone,
banked oblique linear stamped designs, with lip ticking on 18 of the 42 vessels.
Another 13.8 per cent make up Group 2, banked vertical linear stamped
vessels. Group 3 vessels (7.5 per cent) are banked linear stamped with the
addition of punctates. In all, banked linear stamping makes up nearly three-
quarters of the assemblage. Group 4 is a banked pseudo-scallop shell stamped
set; Group 5 is banked dentate stamped, and Groups 6 and 7 represent
unbanked pseudo-scallop shell stamped and unbanked dentate stamped
vessels. Only one pot has punctates as the sole exterior decoration. Three
vessels are outliers of all groups. Vessel 7 has horizontal single cord
impressions reminiscent of Mackinac Ware but with a classic Laurel profile and
paste. Vessel 30 has a dentate stamped exterior common to Laurel, but with a
cord-marked interior, a trait of Early Woodland ceramics elsewhere. Vessel 77
has interior punctates that do not form exterior bosses, an idiosyncratic feature
for Laurel ceramics which commonly have exterior bossing.

Whitefish Island is so homogeneous that it must be considered a single
occupation where the same potters often repeated their template-like designs. It
also contains one vessel with an Early Woodland trait and, as discussed above,

that feature points to the probability that seriation will place Whitefish Island at

55

the early end of the eastern sequence.

Naomikong Paint Sita (20CH2) (See Appendix D) - The Naomikong

Point Site is a multi-component site at the east end of Tahquamenon Bay on the
south shore of Lake Superior, about 30 miles west of Sault Ste. Marie in the
Upper Peninsula of Michigan. The sandy beach had been the focus of surface
collections since at least the 19303 (Janzen 1968: 1). In the 19603, James E.
Fitting and Donald E. Janzen of the University of Michigan set about a multi-
year excavation of the site to address the issues of community patterns. Fitting’s
and Janzen’s excavations totaled 2,750 square feet and stretched along 190
feet of the Lake Superior shoreline. Janzen’s surface surveys and test trenches
were designed to find the limits of the site, but he found none. He concluded
that the Naomikong Point Site represented an enormous occupation area, and
he estimated that his work and Fitting’s together covered only 1% of the site
(Janzen 1968: 91). Its main component was a rich Laurel deposit eventually
yielding sherds of more than 200 Laurel vessels.

Fitting’s 1965 excavation and Janzen’s 1967 excavation resulted in two
large block units with a 5x60-foot trench connecting the two major areas. The
units were excavated in two levels (Janzen 1968: 19). The top level contained
historic and Late Woodland material. Janzen found that the Laurel component,
a shallow deposit no more than 0.7 of afoot, consisted of sherd concentrations
extending from top to bottom, so he excavated the Laurel deposit as a single
level. The results of the three years of work were produced as Janzen’s doctoral
dissertation in 1968. He interpreted the Middle Woodland spatial data as
indicating two separate Laurel occupations in the two large block excavations,
one later than the other. Janzen obtained a radiocarbon date of AD. 43035400

from scrapings on a pseudo-scallop shell stamped vessel (1968292).

56

l undertook a cluster analysis to test Janzen’s hypothesis of two
occupations, and seriation analysis to assess their relative temporal placement.
The study included 162 Naomikong Point vessels represented by sherds large
enough for inspection from rim to full neck.

Analysis revealed 12 clusters: Groups 1 to 4 are pseudo-scallop shell
stamped vessels; Groups 5 to 8 are linear stamped; Group 9 is a category
based on vertical motifs; Group 10 represents plain vessels; Groups 11 and 12
are incised. There are 11 outliers.

The outliers are trailed, more akin to wares southwest of Naomikong
Point such as Lake Nokomis Trailed of northern Wisconsin (Salzer 1974), and
Black Sand, Dane, and Prairie Incised (sic) of southern Wisconsin (Famsworth
and Emerson 1986). Dane Incised also occurs on North Bay sites on the Door
Peninsula of northeastern Wisconsin (Mason 1967).

Three miniature vessels all have a pendant triangle motif that also occurs
on Dane, Prairie, Black Sand, and Nokomis vessels. The motif may also occur
on Point Peninsula Wars of southern Ontario and New York (Ritchie and
MacNeish 1949) and Hopewell Ware of southern Michigan (Fitting 1975). The
motif is uncommon wherever it is found and indicates a widespread exchange
of ideas.

A Coefficient of Similarity for the two blocks of 135.3 indicates that they
are nearly 68% similar, however there are significant differences to support
Janzen’s hypothesis of two separate occupations. Whereas Janzen
emphasized the predominance of pseudo-scallop shell stamped vessels in the
east block (1968 80-82), the present analysis suggests that the significance lies
in a number of factors. 1) There are more pseudo-scallop shell stamped vessels
in the east than west. 2) There are more banked linear stamped vessels in the

west than east. 3) There is a large percentage of vessels in the west block that

57

are plain. 4) There is a large percentage of vessels in the west block that are
decorated on the upper rim only. This group is reminiscent of the process of
simplification of rim decoration seen on later, transitional Pine river Ware of
northern Michigan (Holman 1984: 35). Seriation of the two blocks at Naomikong
Point with other Laurel assemblages (see Chapter 6) should indicate that the

east block assemblage constitutes an earlier point in the series than the west.

nggdman Site (ZOCHQ) (See Appendix E) - The site is at the mouth of
the Potagannissing River on the west side of Drummond Island in the Upper
Peninsula of Michigan It has one burial mound that may be Laurel (Branstner
n.d.). The site has been known since amateur archaeologists excavated the
mound in the early 20th century. They reportedly found skeletons with their feet
toward the center and a number of artifacts including ceramic sherds with
geometric designs (n.d.: 19). Charles E. Cleland and Christine Branstner of
Michigan State University conducted excavations at the site in 1991, 1992, and
1994 to determine site function and settlement-subsistence practices (Branstner
n.d.: 2). Little more than test pits were excavated in the extremely large and
intensively occupied Laurel area of this multi-component site. The ceramic
assemblage includes 29 analyzable Laurel vessels used in this study. .

Statistical analysis revealed seven clusters: Groups 1, 2, and 7 are
pseudo-scallop shell stamped vessels differing in banked/unbanked and
presence/absence of bosses and lip ticking. Group 3 is incised; Group 4 is
dentate stamped; Group 5 is plain, and Group 6 is linear stamped. The largest
cluster, the group of banked linear stamped vessels, represents 24.1 per cent of
the assemblage, however the three groups of pseudo-scallop shell stamped
vessels comprise 32 per cent of the assemblage. Four outliers all have traits
that preclude them from Laurel. Vessels 28 and 29, both with cord-marked
exteriors, may be transitional forms between Laurel and Late Woodland. Vessel

58

26 is Saugeen Ware and Vessel 23 is trailed, perhaps related to Nokomis
Phase vessels of north central \Msconsin or to Dane Incised of the Green Bay

area.

Qyflakis Sita (29MK51) (See Appendix F) - The site in St. Ignace,
Michigan, was located by John Franzen and Mike Manfredo during a survey

conducted for the Mackinac State Park Commission and the Michilimackinac
History Society. James E. Fitting and Timothy Smith conducted excavations
there in 1973 for the Michigan History Division, Michigan Department of State
(Fitting n. d.: 8).

Fitting, who undertook the excavations as part of a salvage operation
during urban construction in the village, uncovered a large burial pit with
associated Laurel Ware (n.d.: 18). In his report to the History Division, Fitting
mentioned that the site had been leveled previous to his work (n.d.: 11); what he
may have discovered was a mound similar to that at the Cloudman Site on
Drummond Island. An unpublished report of Fitting’s work is on file in the
Department of Anthropology, Michigan State University.

Most of the Laurel ceramics were concentrated in and around a large pit,
Feature 15, covering most of two 10x10-foot squares. The pit contained
secondary burials of at least seven people and grave goods including a
smashed pot, two complete pots, one inside the other, a number of bird bills, red
ochre nodules, and a large sandstone “lump” (n.d.: 18-21). The fill of the pit
contained other Laurel sherds. Beside the pit, and associated with it, was
Feature 22, a large fireplace with many fire-cracked rocks and a crushed pot
similar to those in Feature 15. A post mold extending nine inches into the living
floor was associated with the fireplace. A radiocarbon date of AD. 170:80 was

obtained from Feature 22 (n.d.: 23).

59

All ceramics were Middle Woodland except for three intrusive shell-
tempered sherds probably associated with the nearby historic Tionontate
village (n.d.: 25-26). Fitting conducted chi-square analysis in search of
significant attribute groups but he discovered no clustering (n.d.: 53). He also
concluded that attribute distribution was random stratigraphically and spatially.

I reanalyzed the assemblage to test the hypothesis of randomness.
Statistical analysis revealed 10 clusters: Groups 1 and 2 form a super-cluster of
banked and dragged stamped vessels, with dentate stamped in Group 1 and
linear stamp in Group 2. Group 3 is composed of pseudo-scallop shell stamped
vessels. Group 4 is a cluster of banked, undragged designs with punctates. A
super-cluster of Groups 5 to 10 are unbanked, and each group is small,
comprising only a few vessels each. The clustering partially supports Fitting’s
hypothesis in that the 61 vessels analyzed at Gyftakis form 10 groups: Gyftakis
is one of the most varied assemblages in the study. Contrast this to Whitefish
Island where 80 vessels comprised only seven groups, and half of the vessels
were in one group.

Concerning distribution, as Fitting noted, there is no stratigraphic
clustering; clusters include vessels from various levels. Also, the 19 vessels
from Feature 15 are in every cluster except Groups 5, 6, 9 and 10 all of which
are small, subject to sample size problems. The assemblage does not split into
two occupations as at Naomikong Point.

Of the outliers, all are variations of Laurel designs except Vessel 7, a
combed pot with no clear association with any other ware, and Vessel 63 , a pot
with unusual dragged incising and trailed pendant triangles similar to vessels at

Naomikong Point and Squirrel Dam (see Appendices E and M).

Pgrtage Site (2QEM22) (See Appendix G) - The site lies on dunes

60

between an interior lake chain and Lake Michigan near Petoskey in
northwestern Lower Michigan. Vlfilliam Lovis of Michigan State University set
out to examine the site in 1974 and 1975 with the expectation that it would
provide transitional Middle-to-Late Woodland material. The ceramic analysis
was completed by Lovis, Rajnovich and Bartley (1998)

Lovis uncovered two components separated by sterile aeolian sand.
Lovis, Rajnovich, and Bartley conducted a cluster analysis of the ceramic
assemblage to define the characteristics of the upper and lower strata, and
examine the relationships between the two components (Lovis, Rajnovich and
Bartley 1998). While the upper and lower strata represent mainly Late and
Middle Woodland respectively, our cluster analysis found that the ceramic traits
of the two strata overlap, indicating at least a partial continuum from Middle to
Late Woodland in the area. The vast majority of ceramics from the lower
component are Middle Woodland, mostly Laurel. Two overlapping dates of AD.
1205;120 and AD. 330:150 from charcoal from a small pit or post mold, and
wood or pitch from a hearth, both from the lower stratum, were accepted as
reasonable dates for the assemblage (Lovis, Rajnovich and Bartley 1998).

The present analysis concerns only those vessels from the lower levels,
named Groups 3 and 4 in the previous publication of ceramics from the site.
There are significant subgroups in the assemblage. Some are recognizably
Laurel types previously defined. Subgroup 3A is Laurel Incised; 3B is Laurel
Plain var. Punctated; 4A is Laurel Oblique var. Dentate Stamped, and 4B is
Laurel Oblique var. Linear Stamped. However Subgroup 48, contains stamps
that are unusual in a Laurel assemblage, linear stamps done with a cord-
wrapped stick (Vessels 65, 66 and 68). These may be related to Upper
Peninsula Looped-end Cord stamped at Summer Island in the Upper Peninsula
(Brose 1970) and North Bay Corded Stamped dated to AD. 160 i100 at Porte

61

des Morts in Wisconsin (Mason 1967). In addition, Group 3C contains banked
pseudo-scallop shell stamped vessels common in Laurel assemblages, but it
also contains single-cord impressed vessels perhaps related to vessels at
Summer Island (Brose 1970). The impressions there are vertically oriented
whereas they are horizontal at the Portage site. Subgroup 3D has cord-
wrapped stick designs perhaps more common on North Bay Ware than Laurel,
and Vessel 39 has added trailing, a trait more at home on Dane Incised or Lake
Nokomis Trailed of northern Wisconsin.

In summary, while the previous study indicated a connection between
the Portage Site’s Middle and Late Woodland assemblages, this study notes a
connection also between the Middle Woodland vessels of the Portage Site and
vessels not only of Laurel sites to the north but also of Middle Woodland sites

across Lake Michigan to the west.

Fiek Site (DIKe-I) (See Appendix H) - The Fisk Site, on a portage of a
tributary of the Winnipeg River just north of Kenora, Ontario, was the focus of an
early survey of northwestern Ontario by the provincial Ministry of Culture and
Recreation. In 1975 and 1976, C. S. “Paddy” Reid excavated a block of 25 one-
metre units to expose spatial patterning as well as stratigraphy of this multi-
component site. It contained material from Late Woodland complexes over, and
mixed with, Middle Woodland Laurel material. The main feature consisted of a
pattern of post molds and rocks, apparently the outline of an oval house.
Charcoal from the central hearth produced a date of AD. 501115, one of the
earliest assays for Laurel (Rajnovich, Reid and Shay 1982).

The Laurel assemblage is small, only 23 vessels. It consists of a simple
set of four clusters with one outlier. Group 1 is banked dragged pseudo-scallop
shell stamped; Group 2 is plain with punctates/bosses; Group 3 is unbanked

62

pseudo-scallop shell stamped, and Group 4 is unbanked linear
stamped/punctated. The outlier is the only dentate stamped vessel. The
complexity of designs seen, for example, at the Gyftakis and Portage sites, is

missing in this early western Laurel assemblage.

Long Sault Site (Dde-1) (See Appendix I) - The Long Sault site in
northwestern Ontario is part of a group of sites on the Rainy River known in the
archaeological literature as Manitou Mounds, the largest existing group of burial
mounds and related habitation areas in Canada. The Rainy River First Nations
developed Manitou Mounds as a national historic park and opened the park
and interpretive center to the public in 1999.

In 1973, Manitou Rapids Reserve of the Rainy River First Nations
requested a survey and test excavation in response to destructive gravel
operations in the area. David Arthurs of the Ontario Ministry of Natural
Resources conducted both survey and excavations between 1973 and 1975.
Included in his investigation was the Long Sault Site. He produced his results
as his master’s thesis at the University of Manitoba and as a report for the
Ontario Ministry of Citizenship and Culture (Arthurs 1986).

The site is deeply stratified with Archaic to historic materials. He
excavated 14 one-metre units in 3-cm levels (Arthurs 1986: 28), and foUnd the
Laurel stratum to be the richest. He estimated that three Laurel occupations
were present but so close in time as to be treated as one, and he estimated its
date to be middle Laurel McKinstry phase (1986: 133).

I reanalyzed the assemblage from Long Sault to test Arthurs’ estimate of
a mid-Laurel date. The 79 vessels form only six clusters with plain vessels,
Groups 1 and 2, representing nearly 57 per cent. Another 32 per cent are

vessels with dentate and pseudo-scallop shell stamping in unbanked designs,

63

and only 7.6 per cent have banked stamping. The unbanked motifs are
reminiscent of later ones on Blackduck vessels, an indication perhaps that Long
Sault Laurel is mid-Laurel as Arthurs suggested, or even later. Of the three
outliers, Vessel 46, with exterior lip-rim tick marks, is significant because lip

ticking is common on eastern Laurel pots and uncommon in the west.

Ballyaacree Site (Dka-8) (See Appendix J) - Beginning in 1983, the
Regional Archaeologist’s office of the Ontario Ministry of Citizenship and
Culture began a four-year excavation of the Ballynacree Site on the Winnipeg
River in Kenora, Ontario, to obtain data concerning the community pattern of an
entire Laurel habitation site (Reid and Rajnovich 1991).

The excavations were in a large block of units covering an entire flat,
grassy terrace about 30 x 20 metres. The area was dug in 3-cm arbitrary levels
exposing a stratified, but somewhat mixed, series from Historic to Laurel. The
Middle Woodland component consisted of the outlines of three complete
lodges, none of which overlapped, an indication of probable contemporaneity.
Preliminary seriation of the ceramics from each house showed close similarity
among the houses, another indication of probable contemporaneity (Reid and
Rajnovich 1991).

Three radiocarbon assays were obtained from charcoal from the floor
and central hearth of House #1 dating to AD. 1240:45, A.D. 1240165 and AD.
1270:55 (Reid and Rajnovich 1991: 198). The dates are among the latest
obtained for Laurel and have been questioned by several researchers,
however, preliminary seriation of the ceramics from Ballynacree and other
Laurel assemblages, based on external decorative techniques, indicated that
Ballynacree is the latest Laurel site in the western Laurel area, regardless of the

questionable validity of the absolute dates (Reid and Rajnovich 1991: Figure

64

28).

The Ballynacree ceramics are included in this study to reexamine their
position in the seriation of Laurel sites, with the expectation of providing a very
late assemblage representing the Hungry Hall Complex. The assemblage is
characterized by a small percentage of banked stamped motifs (10.7 per cent)
and the presence of Late Woodland characteristics such as cord-wrapped stick
stamps, single-cord impressions, and oblique-over-horizontal motifs. It also
contains a unique incised design on Vessel 88, and a large number of dentate-
stamped vessels (>40 per cent) in comparison to the tiny number of pseudo-
scallop shell stamped ones (4.5 per cent).

The statistical analysis for this dissertation produced 12 major clusters.
Groups 1 and 2 are 18 vessels that are plain, with or without punctates/bosses.
Vessel 88 of Group 1 has an isolated, incised, triangular devise on the
shoulder. Group 3 contains vessels with cord-wrapped stick decoration usually
considered late Laurel. Group 4 is the largest group and it, along with Group 5,
represent unbanked dentate stamped vessels. Group 6 is a pseudo-scallop
shell stamped cluster, and Group 7 represents very rare single-cord impressed
designs. Group 8 is linear stamped, and Group 9 is incised. The last three
groups represent motifs rather than techniques: Group 10 represents vessels
with predominantly horizontal motifs; Group 11 has vessels with upper rim
decoration as the predominant feature; and Group 12 has banked and dragged
stamped designs.

Of the 12 clusters, House #1 has all of them, House #2 has nine, and
House #3 has seven. While all of the seven groups of House #3 are shared with
House #1, and five of its seven with House #2, it lacks five groups represented
in House #1 and four in House #2.

If we assume that the houses are contemporary, they may provide us with

65

some indication of the similarity coefficient that should be expected in the event
of a migration. We should expect a Founder Effect to be present in assemblages
representing the emigrant group splitting from a parent population; that is, the
assemblage of the emigrant group would represent only a sample of that of the
parent population because the emigrants themselves are a sample of the
parent group, selecting styles representative of a portion of the full style
repertoire of the homeland (Sackett 1982: 72-73). But how similar would the
emigrant and parent assemblages be? Consider House #1, with all 12 clusters
represented, to be the parent population from which samples for Houses #2 and
#3 are drawn. If people from either of these emigrated, the assemblages would
provide us with a minimum coefficient of similarity between the home and
emigrant groups. The similarity coefficients between the parent group of House
#1 and samples from the other houses is 131 (65 per cent) for House #2, and
128 (64 per cent) for House #3. Thus, if residents of either houses emigrated,
they would take a minimum of 64 per cent (a similarity coefficient of 128) of the

decoration of the parent sample.

66

5. LAUREL AND ITS MIDDLE WOODLAND NEIGHBORS

Three hypotheses were generated concerning Laurel origins. 1) spatial
analysis will reveal that the earliest Laurel assemblage/complex is most similar
to one or more of its extemal neighbors, and tests for strength of similarity
through style analysis will show that similarity is close enough to conclude that
one or more neighbors had a major influence on Laurel. 2) Temporal analysis
will reveal that the external neighbor(s) is earlier than the earliest Laurel
assemblage/complex and is therefore a good candidate for an origin of Laurel
style. 3) If there is no candidate for Hypotheses 1) and 2), then we can conclude
that an external source is not responsible for Laurel origins: internal Laurel
dynamics will explain the development of Laurel.

Neighboring ceramic traditions stretch from Saskatchewan in the west
through Manitoba and North Dakota, across the northern and central portions of
Minnesota and Wisconsin, to southern Michigan and southern Ontario. Little
work has been completed in most of these areas relative to Laurel studies,
making detailed comparisons impossible. It is striking how much we know about
Laurel as opposed to its neighbors: compare, for instance, the works of
MacNeish (1958), Wright (1967), Stoltman (1973, 1974), Fitting (1979),
Lugenbeal (1977), Dawson (1980, 1981), Meyer (1983), and Reid and
Rajnovich (1991), just to name a few publications providing Laurel overviews, to
publications concerning two adjacent ceramic traditions in Wisconsin, Nokomis
and North Bay, that have been defined by Robert J. Salzer (1969) and Ronald J.
Mason (1967, 1969, 1991) respectively but investigated by few other
researchers (but see C. Mason 1981, Dirst 1995, Moffat 1999). While a detailed
comparison of all of these Northern Tier Middle Woodland neighbors was an

aim of this dissertation as a means of obtaining a more complete understanding

67

of the interactions of the Laurel World, that lofty goal cannot be accomplished at

this date. Nevertheless, a more modest analysis follows.

LAUREL AND THE NORTHEASTERN PLAINS

The earliest ceramics of the area have been termed Crawford Ware
(Benn 1990197) and are related to the cord-marked and trailed family of the
Early Woodland period known as Fox Lake Trailed in Minnesota, Black Sand
Incised in Illinois, Prairie lncisedfT railed, Dane Incised and Lake Nokomis
Trailed in Wisconsin, and Shiawassee Ware in Michigan (Famsworth and
Emerson 1986, Fischer 1972), and possibly ceramics of the Couture Complex
in southwestern Ontario (Spence and Fox 1986). Crawford Ware has
radiocarbon dates for the fifth century BC. (Gregg 1990: 31) and for several
centuries later in North Dakota (Benn 1990: 129). If the earliest dates are
supported by future evidence, Crawford Ware would be a candidate for the
earliest pottery in proximity to western Laurel. However, a development of
Laurel out of Crawford is highly unlikely as the two are vastly different.
Crawford was constructed using the paddle-and-anvil method, has cord marked
surfaces and rounded lips, and has trailed designs (Benn 1990: 97), whereas
early western Laurel was constructed by coiling, has plain surfaces with flat lips,
and has a complex of design techniques that excludes trailing. The only
similarity is that both wares are sometimes bossed. The pattern is of little
interaction among the potters of these wares - indeed, nearly total avoidance.

The Northeastern Plains Middle Woodland series of wares comprising
Besant, Sonota, Avonlea, and Brainerd surround the western Laurel area, from
eastern Saskatchewan through southwestern Manitoba and North Dakota to
north central Minnesota, spanning a time period of about A.D. 200-900 (Davis
1988, Gregg 1994, Meyer and Hamilton 1994). Crawford Ware might be

68

considered a candidate for the precursor of these wares that exhibit surfaces
treated by various cord markings, grooved paddles, and net impressions, with
minimal decoration. If so, the potters of these wares acted little differently than
their predecessors. Meyer and Hamilton have noted a few cases in which
Laurel and Avonlea, or Laurel and Brainerd, are found in the same contexts,
indicating some interaction; however, they also note that these occurrences are
present only along the borders of the Laurel region that coincide generally with
the Plains-Woodland interface, and that the interaction seems neither to have
been intensive nor to have had an extensive cultural impact (Meyer and
Hamilton 1994: 111). Meyer and Hamilton conclude that there is a pattern of
avoidance between Laurel and Besant, and of only weak interaction between
Laurel and Avonlea/Brainerd, suggesting distinct cultural groups. They note the
fact that Laurel does occur in the park land and grassland edge in southeastern
Manitoba and adjacent Minnesota, an indication that Laurel people were
entirely capable of adapting to that environment. However, Laurel is absent
from the parkland and grasslands of southwestern Manitoba and
Saskatchewan; sociopolitical relations must be considered a factor in restricting
middle Laurel occupation in these bison ranges (1994: 112).

Meyer and Russell paint a different picture for late Laurel. Reacting to an
hypothesis put forth by me (Rajnovich 1983) that Selkirk Composite ceramics of
the early Late Woodland in northern Manitoba were derived from a Laurel base,
they suggest rather that ceramic traits were derived from the interaction of two or
more regional cultural groups including Laurel, Avonlea, and Blackduck (Meyer
and Russell 1987: 22). The textile-impressed body surfaces and relative paucity
of decoration that characterize early Selkirk material could readily be derived
from Avonlea net impressed vessels, while vessel shape and uses of punctates

on early Selkirk vessels could be derived from Laurel.

69

There is strong evidence to support an ethnic identification of Selkirk as
Cree although Meyer and Russell, in accepting the evidence, caution that not all
Selkirk material, especially some on the fringes of the territory, need have been
Cree (1987: 27). One could add that not all people who interacted to create
Selkirk Ware in the early Late Woodland need have been Cree. The long, weak
interaction between Avonlea and Laurel may have been a factor of language
differences, and the hypothesized coalition of some Avonlea with Blackduck
and Laurel to form Selkirk would then have possibly involved the adoption of a
new language by some participants.

However, the weak interaction between early Laurel and Plains potters

precludes the Plains wares as candidates for the origin of Laurel.

LAUREL AND CENTRAL MINNESOTA

Malmo Ware is the earliest ceramic manifestation found to date in central
Minnesota. Wilford defined the Malmo Focus of the Mill Lacs Aspect on the
basis of sites with predominantly plain surfaced sherds. He described its
ceramic characteristics as intermediate between Howard Lake Hopewellian
and Laurel wares (Wilford 1955: 135). A small number of mounds have been
assigned to the focus as have a few habitations, one of which, the Brower Site,
was excavated by Wilford in 1949 and by Guy Gibbon in 1972. Gibbon has
described the ceramics as plain-surfaced vessels with Havana-derived traits of
incised, punctated, bossed. cord-wrapped-stick stamped, and dentate stamped,
and the Laurel-derived trait of dragged stamped. The Havana traits occur in
greater numbers than Laurel traits (Gibbon 1975: 18). Malmo has been
radiocarbon dated at 690 BC. to AD. 150 (Gibbon 1975: 18-19), dating early
enough to be a possible forerunner of Laurel.

There are problems, however, with the concept of Malmo and the

70

associated ceramics. Gibbon questions the reliability of Malmo as representing
an integral unit (personal communication 2000). He has noted that the definition
is hampered by the small number of sites assigned to the unit, small sample
size, early research that concentrated on burial mounds which had few or no
associated diagnostics, and unreported excavations, with the Brower Site as
the only single component habitation site that has been excavated (Gibbon
1975: 11). He has also questioned the early radiocarbon date of 690 BC. from
the Morrison Mound that had no diagnostics, and he reassigned dates for
Malmo to about 200 BC. to AD. 200 on the basis of the presence of Havana
and Laurel ceramic characteristics (1975: 19).

Christy Caine-Hohman, who contributed the description of Malmo
ceramics for the Handbook at Minnesota Ceramies (Anfinson 1979), included a
comparison of Malmo ceramics with other Minnesota wares in her dissertation,
but did so “reluctantly” because they are “so ill-defined that formal comparison
may not be very productive” (Caine 1979, Cains-Hohman 1983: 137).

Sample size is a problem - there are “over 200 rim sherds” defined as
Malmo in Minnesota (Caine 1979: 137). The two major sites excavated and
reported, Brower (Gibbon 1975) and Gull Lake (Johnson 1971), produced only
72 rims representing perhaps 40 vessels. Gibbon concluded that the integrity of
Malmo remains a research question.

Given that future research may support Malmo as an archaeological
entity, the possibility that Laurel emanated from Malmo seems remote for a
number of reasons. 1) The high frequencies of pseudo-scallop shell stamping in
Laurel is not observed in Malmo; the Brower Site has none. 2) The shared traits
of dentate stamping, punctates, bosses, and cord-wrapped stick are
characteristic of late Laurel. 3) The small number of Malmo vessels compared to

Laurel with perhaps more than 1200 vessels suggests an influence to Malmo

71

from Laurel. 4) The diffuse nature of Malrno, if it exists as an entity at all, in a
restricted locality is in contrast to the strong cultural homogeneity and extent of
Laurel.

The wisest conclusion is that Laurel influenced Malmo, as Gibbon stated.
Scott Anfinson has suggested that some sherds recovered in central Minnesota
are Laurel-like and that the range of Laurel should be considered further south
than northern Minnesota(Anfinson 1979: 126).

Malrno could still have contributed to the northern development, not
directly and not in early Middle Woodland, but later on. Cains-Hohman
hypothesized that St. Croix ceramics of the middle to late Middle Woodland in
Central Minnesota were derived from Malmo and, basing her conclusions on
ceramic stylistic analysis, she posits the idea that they may very well have
contributed to the development of Blackduck ceramics that characterize the
early Late Woodland of northern Minnesota, northwestern Ontario, and
Manitoba. In her scenario, Blackduck has at least two origins, St. Croix and
Laurel (Caine-Hohman 1982: 252-253). A number of researchers have
provided evidence for mixtures of Laurel and Blackduck traits on some vessels
(Koezur and Wright 1976, Stoltman 1973, Lugenbeal 1977, Dawson 1981), and
Lugenbeal has put forth an argument that the ceramic evidence can reasonably
be interpreted as an in situ development. If these represent ethnic entities as
well as ceramic styles, the picture is similar to the development of Selkirk

discussed previously - ethnogenesis from a number of streams.

LAUREL AND NORTH CENTRAL WISCONSIN
North central Wisconsin is also under-represented in the literature.
Robert J. Salzer conducted extensive surveys of the North Lakes district of

Wisconsin in the mid-19603, but the area has received little attention since then.

72

Salzer divided the Woodland period into two phases, Nokomis and Lakes. He
suggested that the Nokomis Phase straddled Early and Middle Woodland
periods and was characterized by thick, cord-marked ceramics with wide,
finger-trailed lines named Lake Nokomis Trailed, but also common were
vessels with dentate stamps, cord-wrapped stick stamps and incising indicating
interaction with potters from the east (North Bay Phase) and south (Waukesha
Phase) (Salzer 1969, 1974). Carol l. Mason added the pottery type Little Eau
Pleine Punctated after excavation of two mounds in Marathon County. The type
is similar to Lake Nokomis Trailed in construction and surface treatment but
lacks the trailed lines and adds lower rim punctates (C. Mason 1981). Ceramics
from the Squirrel Dam Site in Oneida County, excavated by Salzer from 1965 to
1968, comprise one of the major assemblages used by Salzer to define the
Nokomis Phase, but they have not been completely reported. The Middle
Woodland vessels from Salzer’s excavations of Squirrel Darn are presented in
Appendix K for comparison to Laurel.

Salzer noted two major features in the phase. There is an impressive
number of copper tools and detritus from tool-making, with copper wastage
found throughout Nokomis Phase deposits, and there was an apparent reliance
on trade, as exotic lithic material is more common than local lithics (Salzer
1974: 49). On the basis of exotic lithics and ceramic characteristics, he
concluded that the trade was with the south and east but he suggested there
was “limited trade” with Laurel of Minnesota, based on the presence of a few
pseudo-scallop shell stamped vessels (1974: 49).

Salzer could provide no dates but posited a range from sometime before
A.D.1 to about AD. 200. Recently, Norman M. Meinholtz and Steven R. Kuehn
reported a series of dates from the Deadman Slough Site ranging from 360 BC.
to AD. 670 (Meinholtz and Kuehn 1996: Table 11.3). Dane Incised and Lake

73

Nokomis Trailed ceramics were found in association with dates from 190 BC. to
AD. 530. Charles R. Moffat has added dates of 280 :50 80., AD. 4355 and
AD. 80: 140 from Mark Bruhy’s excavations of Nokomis Phase material at
Luna Lake (Moffat 1999: 36). Moffat recently re-excavated a portion of the
Squirrel Dam Site and obtained dates of 560:70 B.C. from a pit fill containing
Lake Nokomis Trailed and Dane Incised sherds and AD. 58:90 and AD.
160:80 from two superimposed pits, the uppermost containing Dane Incised
ceramics (Moffat 1999: 36). He noted the “alarming amount of heterogeneity in
the ceramic assemblage” of Nokomis Phase, and suggested it may be a mixture
of Early and Middle Woodland materials (1999: 93). Salzer has also suggested
that collapsed stratigraphy so common on Upper Great Lakes sites may account
for the tremendous variety, and has expected future work to clarify the sequence
and redefine the “Nokomis Phase” (personal communication 2000). Like the
“Malmo” problem, the “Nokomis” problem presents difficulties for comparative
studies.

The burial pattern of Nokomis is equally murky. C. Mason reported two
mounds on the Little Eau Pleine River that have Nokomis Phase ceramics in the
fill but none associated with the burials (C. Mason 1981), and Nokomis-like
ceramics appear in illustrations of vessels from the fill between burials at the
Riverside Site in Menominee, Michigan, a large cemetery radiocarbon dated
from 510:140 B. C. to AD. 1:130 (Hruska 1967). No ceramics were directly
associated with the burials.

There are few similarities between Nokomis and Laurel. As noted in
appendix K, only five vessels from the Squirrel Dam Site share any traits with
Laurel. Vessel 28 (Figure 46) has pseudo-scallop shell stamping but it is used
in Havana-like zoned decoration unlike Laurel. Vessel 15 (Figure 46) has a

classic Laurel profile but not classic Laurel decoration. Vessel 25 (Figure 46)

74

has dentate stamped columns like Laurel but the profile is different. Vessel 21
(Figure 46) has banked stamping like Laurel but the motif is unlike Laurel. All of
these vessels are outliers of the clusters at Squirrel Dam. Vessel 17 (Figure 46),
in a cluster of trailed/incised vessels, shares the verticals-over-horizontals motif
with Laurel.

All of these matches can be found at the Naomikong Point Site which
also has some Nokomis-like traits. Vessels 23, 25, 43, and 160 from Naomikong
Point (Figure 11) have trailing that is uncommon in Laurel and common in
Nokomis. Vessels 59, 142, and 147 from Naomikong Point (Figure 11) have
pendant triangle motifs uncommon in Laurel but somewhat similar to Squirrel
Dam Vessels 1 and 20 and possibly Vessel 21 (Figure 46). Vessels 42 and 125
of Naomikong Point (Figure 11) have a motif of isolated vertical columns similar
to Vessel 22 of Squirrel Dam (Figure 46).

The pattern is of occasionally shared traits such as the use of pseudo-
scallop shell stamping in place of Havana-like dentate stamps. While Nokomis
Phase ceramics are early enough to be a progenitor of Laurel, there are so few
shared traits that the likelihood is small. Nevertheless, the evidence points to
contact between Nokomis and Naomikong Point, not western Laurel as Salzer
suggested. The rare and intriguing pendant triangle motif seen in both Squirrel
Dam and Naomikong Point assemblages must convey some message. The
three examples from Naomikong Point are miniature vessels, that is, special
purpose pots, a fact that strengthens the notion that these are special purpose
messages.

The nature of contact between the Laurel and Nokomis people is
unclear. Salzer has established contact between Nokomis and North Bay
potters (Salzer 1974: 49), so down-the-line contact with Laurel as a third party is

possible. However, I have identified at least one Laurel vessel from the

75

Robinson Site. another Nokomis Phase assemblage (Salzer 1969). and,
considering Nokomis’ proximity to the Upper Peninsula of Michigan where the
Naomikong Point Site is located, there is a higher probability that contact was
direct. However, contact was apparently weak: Nokomis and Laurel contributed

very little to each others’ repertoire.

LAUREL AND NORTHEASTERN WISCONSIN

Ronald J. Mason defined North Bay Ware as the major ceramic complex
of the Middle Woodland period of the Door Peninsula of northeastern Wisconsin
(R. Mason 1966). He has described it as demonstrating a “common kinship”
with Laurel Ware (1966: 123), and suggested that the two share partial ancestry
with considerable diffusion between them (1967: 333). The two wares are often
found together on northeastern Wisconsin sites, with Laurel the predominant
type at the Rock Island Il Site at the mouth of Green Bay (Mason 1991 : Table 1).
The two wares are distinguished by paste and temper: North Bay is coarser,
more friable, and with bigger temper than Laurel in addition to being thicker.

North Bay has been recently dated as early as the second century BC. at
the Shanty Bay Site in Door County (Dirst 1995). It endured into the fourth
century AD. when a transition to Late Woodland Heins Creek Ware took place.
The North Bay Phase has been dated as early as 59035270 80. at the Rock
Island ll Site and 520165 BC. at the Richter Site on Washington Island at the
mouth of Green Bay, but Mason rejects both as too early for North Bay on the
basis of several inconsistencies with other dates collected from the same
components (Mason 1991: 126). He accepts a date of 651120 SC. for the
earliest date at Rock Island Il (1991: 126). The Richter Site was excavated in
1968 and 1973 by Richard Peske and Gordon Peters, the latter under the

direction of Guy Gibbon, but a report was never produced. Ceramics from the

76

Richter Site are presented in Appendix L for comparison to Laurel and for
seriation with North Bay components to test the early radiocarbon date.

It should be noted that the early dates from Rock Island and Shanty Bay
were run on charred food incrustations on ceramic vessels (Mason 1991: 123;
Dirst 1995: 54), a method recently brought into question. Rose Kluth has
presented a series of radiocarbon dates derived from food residues on Brainerd
Ware from northern Minnesota (Kluth 1995). The dates ranged in the first and
second millennia B.C., far too early for the accepted range of dates for Brainerd,
commonly thought to be from the first millennium A.D., a conclusion based on
Brainerd’s stratigraphic position and co-occurrence with dated ceramics
(Johnson 1971: 53). Researchers have also noted Brainerd’s striking similarity
to Avonlea Ware of the Plains (Meyer and Hamilton 1994: 110; Gibbon 1994:
142) dated to the first millennium A.D. (Morlan 1988). It is apparent that there
are difficulties with unconditionally accepting residue radiocarbon dates from
the Upper Great Lakes area. While some food-residue radiocarbon dates from
Door County, for instance, Oneota dates, are in line with conventional dates
from elsewhere (R. Mason, personal communication), the early North Bay dates
have not been reproduced elsewhere as yet.

Nevertheless, it is possible that North Bay could have antedated the
beginnings of Laurel. Early North Bay vessels such as the thick, linear stamped
vessel in the lower North Bay stratum at Shanty Bay (Dirst 1995: 33) could have
been the developmental stage between generalized Early Woodland linear
stamped vessels ( see those reported in Wisconsin by C. Mason [1964] and in
southern Ontario by Lawrence Jackson [1968]) and later North Bay and Laurel.

The question arises, then, about the nature of early North Bay and its
relationship to Laurel. I undertook an analysis of vessels from the reportedly

early Richter Site and combined it with Ronald J. Mason’s reports of the

77

ceramics from Rock Island II (1991), Mero (1966), and Porte des Morts (1967),
and Dirst’s from Shanty Bay (1995) to ascertain a seriation of North Bay
components. The exterior decorative techniques of each vessel are shown in
Table 2 and the results of the seriation in Tables 3 and 4, the coefficients of
similarity and the relationships of the North Bay assemblages.

As Table 4 shows, the North Bay assemblages do not form a neat
seriation, an observation previously made by R. Mason (1991: 130); however,
Shanty Bay falls at one end of the seriation and is probably the earliest
assemblage, in agreement with the date of 195:50 30. for Shanty Bay (Dirst
1995: 54). However, the small number of vessels comprising the Shanty Bay
assemblage precludes any confident conclusions based on seriation. It is
closely related to Rock Island II and they are only weakly linked to a cluster
formed by Mero 1, Porte des Morts, and Richter. Marc 2 is an outlier only weakly
linked to Mero 1. Porte des Morts has two radiocarbon dates of AD. 1601100
and AD. 120:80 (R. Mason 1991: 126). The date of 520:65 80 from the
Richter Site is not supported.

Not shown in Table 2 are two secondary North Bay decorative forms,
rocker stamps and annular punctates that can be assignable to influence from
Havana ceramics where they are much more prevalent. Vessels with rocker
stamping were recovered from Rock Island ll (f=1), Mero 1 (f=1), and Porte des
Morts (f=4) (R. Mason 1966, 1967, 1991). Three vessels with annular punctates
were found at Porte des Morts (Mason 1967). It is noteworthy that they do not
occur in the earliest assemblage but are more prevalent in later ones, an
indication that North Bay did not descend from Havana but gained its Havanoid
attributes by diffusion in middle Havana times. These attributes never entered
the Laurel repertoire.

North Bay has the spatial and temporal criteria to be acceptable as a

78

79

 

 

 

 

388 m 3.8: 8 3.8503 3. 6: z. 3.8: m: 338 5 293
3&8 k 3.8 m 3.08 mm 338 2 -- 3.8: m .3 2950
- 3.3 c v 3.8 R 38: v 3.: P - mic
- 3.8 m 3.8 v 3.8 P 3.8 m - 686:.
- 3.8: m 3.0: R 38.: o 38 N - 8.26%..
3.5 F 33: s 3.8 9 3.8 m 3.8 v 3.8 F 83.6.2 98
- 3.8: m 3.8 mm 3.8 F 3.8 m - wan.
- 3.8 P 3.8 m 3.8 P 3.8 E 3.28 m .85.. .5
3.3 c F A38 m 3.8 m 3.08 m 3.8 mm 3.8 3 ca...
- 33.8 m 3.8 R. 3.08 m 338 S. 3.98 m cm .85
m 26.2 622m «to: was much P ems. __ Ema. xoom 3m :55

 

«Eocene—co >3 :tez nouns—om .2 3:233... 252309 5:... 3:05 "a 03¢...

 

.QcoE N 0.0.2
\ / .
8.. R w 5

\ / _
wees. won 36.. ii weir 965. i S. i__ 65...... xoom inns Ilsa... 3:65

 

000a_nE000< >3 stoz 33.0w acoE< 02:23:20». 3 03a... 0
8

 

- , 8. em 5 mm on m 6.6.2
8. - 2: RF 8 8 .96...
«m 8. - no. «k 3 9.0.2 «.8 6.6..
5 RF 8? - For mos 396.2
mm 8 S. B. - a? 6. xoom
as 8 3 no? «9 - 3m 355

 

N ems. .253 arcs. 6% mean. 36.2 ._ .26.... .66.. 3m 3.55
0233500 >3 ztez 33.0w .2 3:22.53 .0 02.205000 "n 030...

logical candidate for Hypotheses 1 and 2. The Rock Island ll assemblage is
closely related to the earliest, Pike Bay, complex of western Laurel (R. Mason
1991: 134) with a coefficient of similarity for Fisk-Rock Island II at 137. However,
it is most closely related to Whitefish Island, apparently the earliest assemblage
of eastern Laurel, with a coefficient for Whitefish Island-Rock Island II of 146
based on high incidences of linear stamping. The earliest North Bay appears to
be slightly earlier than Laurel. The small assemblage from the lower North Bay
stratum at Shanty Bay dated to the second century BC. includes decorated
sherds found in undisturbed context from a single thick, linear stamped vessel.
Sherds from another linear stamped vessel and a plain pot lack good
provenience but all are thick and “far more crude” than those associated with
the upper North Bay stratum (Dirst 1995: 33). These could be the precursors of
both later north Bay and early Laurel.

However, there is more to early Laurel than linear stamped and plain
vessels. The distinctive dentate and pseudo-scallop shell stamping of Laurel is
not present in the lower Shanty Bay stratum. These traits developed either
elsewhere or later in time, possibly by Laurel potters themselves. Even if the
Shanty Bay material is ancestral to Laurel, the conclusion would have to be that
Laurel had “multiple origins” (R. Mason 1991: 138); Laurel did not obtain all of

its attributes from one place.

LAUREL AND SOUTHERN MICHIGAN AND SOUTHERN ONTARIO

Middle Woodland components of southern Michigan are composed of
ceramic assemblages similar to Havana-Hopewell of Illinois, and Fitting has
noted evidence for trade between Laurel and southern Michigan sites (1975:
99) in the form of northem-derived copper in the south and minor amounts of

Bayport chert from the Saginaw area in northern Laurel contexts (1975: 99), but

81

he also concluded that there was little actual cultural interaction between the
two areas. He suggested that Laurel people regarded copper “goodies” in the
same manner that modern Mackinac Island inhabitants regard the fudge, made
for consumption by outsiders (Fitting 1979: 140). In his study of Middle
Woodland sites of the Mackinac Straits, Fitting pointed to the presence of
Havana or Hopewell influence on Laurel ceramics as indicated by dentate
stamping, fingernail impressing, and pseudo-scallop shell stamping on Laurel
sherds (1979: 111). He equates the latter decorative technique with rocker
stamping, a questionable conclusion. Pseudo-scallop shell stamping does not
occur in Havana-Hopewell, and dentate stamping and fingernail impressions
are rare in early Laurel. It is apparent that Laurel people of this study were
unimpressed by, or only vaguely aware of, Havana-Hopewell ceramics.

Most researchers in the southern Michigan-southern Ontario area look to
Saugeen and Point Peninsula series as most akin to Laurel (e.g. Fitting 1975:
129; Janzen 1968: 105-108; Wright 1967: 125). Point Peninsula ceramics,
defined by William A. Ritchie and Richard S. MacNeish (1949), occur in upper
New York state, southern Quebec, and southeastern and southern Ontario.
Point Peninsula traits blend with Laurel in eastern Ontario (Wright 1967: 110). It
was originally separated into three phases but is presently under reanalysis by
Robert Pihl for a dissertation at the University of Toronto (Pihl, personal
communication). Saugeen, a related Middle Woodland ceramic series
recovered in the western portion of southern Ontario was defined by Wright and
Anderson (1963) and is presently under reanalysis by James Wilson for a
dissertation at McMaster University (Pihl, personal communication). Saugeen
ceramics were recovered from the Cloudman Site (see Vessel 26 in Figure 14),
and Laurel sherds were found in direct association with Saugeen refuse at the

Donaldson Site in southern Ontario (Wright 1967: 118).

82

Wright suggested that the origins of Point Peninsula and Saugeen lay
within an Archaic/Early Woodland base, with diffusion from Laurel in the north
and Hopewell in the south (Wright 1967: 126). Although dates as early as the
sixth century 30. are known, several researchers have dismissed those (R.
Mason 1991, Spence, Pihl and Murphy 1990). A beginning date “sometime in
the third century BC.” has recently been accepted as the most reasonable
estimate (Pihl 1995), with a transition into Late Woodland wares at about AD.
700 (Spence, Pihl and Murphy 1990). Thus, the earliest components of Point
Peninsula/Saugeen could have been ancestral to Laurel, contrary to Wright’s
conclusion.

Research on both ceramic series has been hampered by
multicomponent and mixed sites with radiocarbon dates in questionable
contexts. The Ault Park Site provided some of the earliest dates for Point
Peninsula, but the ceramics from the site contain what may be a mixture of early
and late vessels (Pihl, personal communication).

Wright undertook an extensive comparison of decorative techniques for
ceramics from 13 Point Peninsula, Saugeen and Laurel components (Wright
1967). Data from his Table 36 (translated into the terms used for decorative
techniques in this dissertation) can be used for comparison of Laurel to the
early Point Peninsula assemblages from the Ault Park Site in Ontario and the
Vinette Site in New York, plus the presumably early Saugeen assemblage from
the Donaldson Site in Ontario. The frequency of pseudo-scallop shell stamping
at the Vinette Site (aboUt 3.5%) is far less than any of the Ontario sites - Ault
Park at 39%, Donaldson at about 38%, Fisk at 39% and Whitefish Island at
16.3%. Two other characteristically Laurel traits, banked stamping and dragged
stamping, are present in only minor quantities in Saugeen and Point Peninsula:

banked stamping - Vinette (about 5%), Ault Park (not present), Donaldson (2%),

83

Fisk (39%) and Whitefish Island (81%); dragged stamping - Vinette (5%), Ault
Park (not present), Donaldson (2%), Fisk (35%) and Whitefish Island (30%).
Banking and dragging increase with time on Point Peninsula and Saugeen
assemblages. At least two of the three sets of figures suggest a diffusion from
Laurel to Point Peninsula, as Wright suggested (1967: 109). Linear stamping is
absent from early Saugeen and Point Peninsula components, while it is a major
technique in Laurel assemblages. It is possible that pseudo-scallop shell
stamping was an early southern Ontario trait that diffused northward, but the
majority of the traits distinguishing Laurel decorative techniques and motifs did
not.

While Laurel is related to Saugeen and Point Peninsula, it does not
represent an extension from them, or a daughter population on a cladistic tree. It
has an internal development of its own with possible borrowings from
neighbors, perhaps linear stamps from North Bay, incising from the Dane series
found in North Bay or Nokomis, and pseudo-scallop shell stamps from Saugeen
or Point Peninsula. It appears that Laurel introduced banking and dragging and
passed those traits to the others.

While researchers have long noted some similarities between Laurel and
some of its neighbors, no neighbor is a good candidate for Hypotheses 1 and 2:
they were that 1) spatial analysis of ceramic assemblages will reveal that the
earliest Laurel assemblage/complex is most similar to one or more of its
external neighbors; and 2) temporal analysis will reveal that that neighbor is
earlier than the earliest Laurel assemblage/complex and is therefore a good
candidate for an origin of Laurel style. Thus, Hypothesis 3 is supported: Laurel
had its own dynamics, and a study of the internal spatial and temporal

dimensions of Laurel is necessary to explain it with satisfaction.

84

6. LAUREL TEMPORAL AND SPATIAL DYNAMICS

Two hypotheses concerning the internal development of Laurel were
generated in Chapter 1: 4) one Laurel assemblage/complex is earlier than
others, and indicates expansion from one source within the Laurel world, and 5)
all early assemblages/complexes are apparently contemporaneous,
representing a convergence of two or more sources of Laurel. Two hypotheses
concerning the processes of interaction were stated: 6) diffusion or short-
distance migration will be supported it similarity co-varies with distance;
assemblages exhibiting greatest similarity will be closest in space; and 7)
migration, not diffusion, will be supported if similarity does not co-vary with
distance; two assemblages from western and eastern Laurel areas will exhibit
similarity. The internal development of Laurel after its origin was addressed with
two hypotheses concerning culture change: 8) All early assemblages are
similar with increasing variability among later assemblages, representing a
givergenee mmel; and 9) all early assemblages are variable with increasing
similarity among later assemblages, representing a egnvergence megel.

A regional study of Laurel requires the clarification of temporal and
spatial parameters; the temporal order of assemblages from each of the western
and eastern Laurel areas must be understood before spatial comparisons

between those areas can be undertaken.

TEMPORAL DYNAMICS

The temporal order from the two areas can be drawn from two studies,
radiocarbon dating and statistical seriation. Researchers in the western Laurel
area have had the opportunity to obtain numerous radiocarbon dates that

clarify the ceramic sequence of the sub-region and the long time span for

85

western Laurel (Tables 4 and 5). Eastern Laurel researchers have not been so
fortunate; few dates have been obtained and those that are available are
ambiguous. Western Laurel assemblages have been subjected to repeated
seriation tests all with simin results (Stoltman 1973, Lugenbeal 1976, Dawson
1980, Reid and Rajnovich 1991), supporting Stoltman’s 1973 conclusions of
three Laurel phases or complexes - Pike Bay (circa 100 BC. to circa AD. 300),
McKinstry (circa A.D. 300-600), and Smith (circa AD. 600 and later) and a
fourth added by Lugenbeal, Hungry Hall, arising possibly after AD. 800 or AD.
900 (Lugenbeal 1977). The stylistic seriations fit well with the series of
radiocarbon dates available. Eastern assemblages have not been seriated.

This dissertation attempts to order the eastern assemblages.

L r I R i ' n t

Table 5 lists uncorrected Laurel radiocarbon dates, along with locations
and original sources. The eastern and western assays are placed side-by-side
for easy comparison. The table indicates that the western dates extend much
later than the eastern ones, and that the western dates also extend earlier than
the eastern ones except for one assay from the Portage Site that has been
deemed anomalous due to questionable context (Lovis, Rajnovich and Bartley
1998: 91).

Table 6 provides further details of the radiocarbon dates including
laboratory numbers, material (when given in the original source), and corrected
dates. The calibrations are from Stuiver and Reirner (1993, updated in 2000).
They demonstrate that assays of about AD. 200 and later (uncorrected), with
standard deviations of about 100 years or less, indicate a range (corrected) with
the earliest date about the same as the mean uncorrected assay. For instance,

the date for The Pas in Manitoba of AD. 240380 (uncorrected) has a range of

86

AD. 240 to AD. 422 (corrected). The date from Sinclair Cove in northeastern
Ontario of AD. 360:60 (uncorrected) has a range of AD. 411 to AD. 540
(corrected).

On the calibrated series, the eastern Laurel dates all fall before about
AD. 900 at the latest except for one from Heron Bay that has been dismissed on
the grounds of questionable context (Wright 1967: 95). There are numerous
dates after AD. 900 in the western area. In fact, the calibrations convert the
latest dates into even later ones.

Among the early dates on the calibrated series, those before AD. 200
uncorrected may represent age ranges that extend earlier than the mean
uncorrected dates. For instance, an assay from Fisk in northwestern Ontario of
AD. 150:115 (uncorrected) has a corrected range of 36 BC. to AD. 236. In all,
ten western assays may represent B.C.dates, while three eastern dates may
range into years 80..

In summary, while the calibrations in Table 6 change the absolute
numbers, the relative placement of dated components, as shown in Table 5,
does not change. A detailed discussion of Laurel radiocarbon dates and
problems arising from them follows.

The earliest date for‘western Laurel of 150 BC. : 165 from a small
component on the Winnipeg River near Kenora, Ontario, is enigmatic in the face
of the large standard deviation, however there are other B.C. dates for western
Laurel including one of a series of important assays collected by Stoltman at
McKinstry Mound 1 on the Rainy River. The date of 30 BC. : 45 is from a
habitation layer underlying the earliest stage of mound building at the location,
and Stoltman concludes that it is an acceptable date for the earliest Laurel of
Minnesota (1974: 80). Dates from subsequent stages of the mound above the
basal layer are AD. 250 : 55 from Mound A which Stoltman

87

 

.E8 .28 .86.". .< 52...; so .32 .22 882.2 2.3.8. .o.<
E: 6.83 2a 5:222. 5.2 .z 8n. 65 88:8. .o.<
E: 0.28: .65 5:222. .532 .z 8n. a: 8.18... .o.<
22 c8260 so .02 3.2832... 83.63 .o.<
:2 5.5.63 2.22 .z 5 .22 $8.26: $3.83 .o.<
48. 2mm .5 .32 28.2 3 2+8... .o.<
.58 .88 .22.... 828 .222 .z 8.... 3 9.18... .o.<
22 83.20 .52 .z 9a.. 32%; 8:68 .o.<
E: 2.8.8 2... 2.6.. .222 .z 8.. 2: 8.18m .o.<
32 22.. so .32 28.2 2728.. .o.<
E! .3282... 2.22 .2 5.5.0 8.2.9. .o.<
48. 22. so .32 2652.23 8718.. .o.<
v8. 22. so .32 8.8.. .63 8.13 .o.<
:2 5.5.63 .522 .z o. .22 b.8222 $.18... .o.<
E: .6852... 2.2.2 .z 5.5 8.738 .o.<
8.: 222m .5 .02 8.5 682. 8....on .o.<
52 :85... so .32 5586.0 333% .o.<
£2 89.20 .322 .z 3. ...w 87:88 .o.<
8.: 222m .5 .02 8.5 28>, 8.28: .o.<
8.2 83.3 5.2 .2 mu :2: 87:8. .o.<
E: .6859... .522 .2 5.3.0 3.183 .o.<
2.2 89.20 .222 .2 ea... .88 8.3:. .o.<
:3. 838 2.0 .o.z .62... 268.33 8233 .o.<
48. 22. 2.0 25.2 28.2 87:83 .o.<
82 82:3. so .32 22 9528.2 8738 .o.<
on: :8on .522 .z mm :2: 87182 .o.<
82 a a 5.6.2 .23 .z 56.0 $135. .o.<
48. 2cm .5 .32 8.8m 32 3.18.... .o.<
.8. 28:3 .5 .o.z .9... 282925 3.183 .o.<
32 2232.3 25 26m .20 25.2 8.83.3 9.332 .0...
.8. 22.6%.. 2... 22. so .32 8.62.2.8 $7.8m. .o.<
.8. 2262...... 28 2cm .5 .32 3622.8 3.2.5 .o.<
82 a a .222 .23 .2 5.5.0 2338. .o.<
8:228 22.83 6.5 sun.
.8?

 

3030:033 0030 3900030.... .053 3 038.

88

 

 

 

8... .228 228262 .28. 22.2 .2 86.62 81.0.2 8.
«8. 262.6... 26 6628 22.2.2 20 82.265. 8718 .0.<
6.... 68.2 22.2.2 .2 .2258 8.16. .0.<
8... .228 22.862 .28.. 22.2 .2 68.62 8.18. .0.<
.8. 262...; .20 2.2 .8 .662 8.18. .0...
6.... 892 22.2.2 .2 6.22.8 8718. .0...
.22 8.2.2 22.2 .2 2.6.20 816.. .0.<
6.... 892 22.2.2 .2 .2258 8.18. .0.<
.8. .282. .20 2.2 622 2.8 8.18... .0...
8... .228 2282.62 .28.. 22.2.2 68.62 8.186 .0...
2.8 .28 2.62.80 62. .20 2.2 260 .6622 8.18.. .0.<
.8. .282. .20 .22 .8 .662 8.16... .0.<
.8. 826.. 22.2.2 .2 262.282 87181.0...
.8. .282. .20 .22 .8 .66.. 8.16... .0.<
.8. 22.2. .20 .22 .8 .66.. 8.18. .0...
.8. 22.2. .20 22 .8 .66: 8718. .0...
8280.0... 20:200.. 85 220
.662
68. 22.6262 .60 2.2 66862.8 8.1.0.2 8.
88 262 26 =86> 22.2 .2 8.2 22 8.1.0.2 8
68m 862 26 .628 22.2 .2 8.2 22 81.0.2 6.
2.... 28260 .60 .02 828.022 81.0.2 8
2.... 8.2.68 22.2 .2 . .22 .8226: 81.0.2 8
.8. 22.2 8.2 .z 82 2.. 6.1.0.2 o.
8.... 8.2.68 22.2 .z <. .22 2.62262 8.16. .0...
68. 28260 .20 .02 8.22082 68.18 .0.<
6.6. 82.20 26.2 .z 8 22: 84.8 .0...
28 282 26 =86> 22.2 .z 8.2 22 8.18 .0.<
N8. .68 .262 228.62 .20 2.2 222 m. .18 .0...
6.... .6220 .62 .z 8 ._.m 8.18 .0...
8.... 8.5.22 22.2 .2 m. .22 262.222 8.18. .0.<
.8. 2.96.. 26 26.. .82 .z 82 2: 8.18. .0.<

 

22.50. m .3...

89

 

 

dww Ewe EN dd 3.10% dd - - 82 9m

mmo 3 s; v.2 dd map-18m dd - mmd-m 9.3 38;

23 $3 m:- dd 8.108 dd - - man. 9:

m5 89.. d3 .83 gm dd on 718w dd .8220 coo; 812d xooz

So ace 0% dd 8.10? dd 83 Sosa-.2, 5:5

m3 86 d; .88 mmv dd om 718d dd .0855 80: :m-ua 96522.8

o8 8X: dam dd 8-1on dd .8220 coo; m E 705 833. >23

mm 6mg 0% dd 3.10% dd .8220 .803 odd-m5, 8 .2. 925.0:

.8 58 0% dd 8-13m dd 82. mmméz 5:5

mom 38 Sm dd 8-1on dd - m1 Tod 2% d3;

08 A2831... dd mop-1mg dd .8220 Bo; 9% 7m 5:586

mg 82 .85 d2 don .58 So dd om 7108 dd - - 3. am
can 8E so dd 8.10; dd - 03.9.. 9.5 a3;

:3 33 .1» .m 6 .02 .th 2o dd on 7103 dd - - mm «2:
03 Ed 68 dd» .2: .35 mt dd 3.18: dd 88. and-mi 5:5
:3 8mm .va .38 NE dd $-1m 5 dd - - E8 .88

3: 83V dam dd 02.1%» dd .8220 woo; ode-m .22 5.85%;

Na: 85: 3d dd map-18¢ dd .8820 So; $5-2“. v.82

8: 3.8: mg dd cop-15mm dd .8820 coo; .53-. .22 oceans-d

SN; 8m: .8: .3: .82 .23: 33 dd 8.189 dd - 3.1m mm «2:
mm: 38: $2 dd $713: dd - mmmN-m 566

8S $8: EN. dd 3-182 dd .8055 n83 mom-2d «2.5. 5d

:1 :3: No: dd 8.7133 dd 58.20 coo; Pom-m .95. £858;

83 EN: EN. dd 3-19.2 dd .8038 So: 33-05 8822.8

82 £8: £3 dd 3-133 dd 3856 39s $3.05 3623.8

82 83: 52 dd 3.123 dd .8036 89s mod-ores 362:3

mam. Sam-1:2 :8: 3.2 dd 2.13? dd - RmNd 5.65
3.5a... Fauna;

vuao vvuootou .0.<\.U.m vuao .Ntvuaz 509532 on; vfiw

 

AwEEm P .< 35»: dmé m_._<u 9.33 mucvcanou .053 .8 mouse con-.3235. ago-.930 no 03a...

90

moo ENE m¢~ dd
0% 3va :w dd
dmd 83 Ed dd
93 $3 E dd
ddd ads 3% dd
2d Git d8 dd
5 SF 8de 8d dd

Nu dd 8 2 d3 dd: 2m dd
on dd :- d d3 3— dd

mu dd 2.2.5.8 d8 mm: dd
NS dd E d d ddv 5 dd

mu dd 5 dd .3 ddv 9 dd
d2 dd 8m dm .5 ddv :. dd
8. 88 N dd

dmm dd 5 ddv 02 dd

2N dd an di 8 dd

dmm dd ad d.$ mm dd

Nvm dd ANN. do, dd di um dd
mm" 8.2 dd :8 mm dd

EN 35 d2 dd: 8 dd

dam Adm-ad dN dd

0% ads of dd

«mm :5 don dds 5 dd

-v in :5 do.“ .38 ox... dd
8d 83 .on $8 ~v~ dd

d3 SR 65 is E dd

mmm 58 SN dd

Om 71.0mm .o.<

ow-\+oom .o.<
oar-1+0; .o.<
oov-\+om¢ .o.<
GNP-15$ .o.<

ow-\+oo~ .o.<
0m —-\+omn .o.<

map-Stud Omp
cop-\fud om
om..\+.u.m on
om-\+.o.m ov
mv-\+.u.m om
OWE-dd O—
ow-\+op .o.<
OON-\+ON .o.<
mm-\+0m .o.<
COTE-CV .o.<
m P T\+Om .o.<
om-\+om .o.<
mm-\+ONP .o.<
Om T\+omp .o.<
mn-\+owp .o.<
oo~-\+omp .o.<
ow-\+o¢N .o.<
om-\+OmN .o.<
mm-\+OmN .o.<
m p T\+omN .o.<

Woo-find too;

.3935 too;
.3220 too;
.3230 too:

Woo-ace too;

.8220 too;

mean
3239.8
.3955 too;

mmoéa
mvvémo
mmOm-Z
mow-0mm
K Tm
m¢¢¢m0

mum-0.0

NmN Tx<u
www-m_>>
va _.-<
www-mS)
va Tv_<o
mmné

anéa

@919;

VNV 7<
mmm T<

Vow-m
{Tm—Z
anéa

wanton.

o>ou -._a_oc_m
xwm :90:

.K deEomz
xam :20:
>3 :90:
A5 :90...

 

_0-=awn_ Cuflumwm

23333

8.1 9m

82 9m
xmxmBUoz

_ dz $25.02
«an on...

S .22 bumcioz
dam-.3508:

om m2:

002 9m

me

Vm .=w

m: .3. 925.0:
and 2F

6: 90me
v.3 2F

mad 2:.
umxmawzaufid

5 dz .9230:
x32

 

.Pucou @ 033.

91

 

0m dd 8v ddv 05 dd 0d-1dd 00F Fwd-0.0 canton.

mvw dd 83 d2 dd ddv mm dd 09.18 dd ~37: .00 82205
F? dd SE .030 2: dd 03.12 dd 38.2 .2 3558

3m 35 ddpdd: vm dd 02-18. dd 30-0.0 wanton.

now 880 S dd 02-101 dd mod-0mm >8 83:

mow 850 E dd 02-102 dd {om-z 6.. 3228

2m 53 Pm. dd 8-102 dd 82.2 usage

8m 3% .03 .38 md dd 08-103 dd 8d 7: .2 3528

03 2% EN dd 007102 dd - .25. ufim

 

.350 d can»

92

considered unacceptable in the context of the other dates in the series (1974:
81), AD. 10 _-l; 60 also from Mound A (1974: 82), AD. 120 i 55 from Mound B,
and AD. 560 i 55 from Mound C, the uppermost mound stratum (1974: 82-83).
Both the series and the strata of the mound supported Stoltman’s earlier
formulation of the Pike Bay and McKinstry phases, represented in the mound by
Mounds A and B for the early phase and Mound C for the later one (Stoltman
1974: 88). Lugenbeal obtained three dates for the Smith habitation site on the
Rainy River. The earliest at AD. 480 i 60 may relate to the McKinstry phase
although the paucity of sherds in the stratum made assignment inconclusive
(Lugenbeal 1976: 118). The later dates of AD. 565 i 60 and 760 i 55 are
acceptable for the Smith phase (1976: 582), and Lugenbeal added another late
but undated phase named after the Hungry Hall Site in Ontario (1976: 578).
Since Lugenbeal’s study, several late dates for Laurel have been
determined but most have been questioned on some grounds, some on
contextual evidence and others on the grounds of the considerable overlap with
Late Woodland dates from the same area. some researchers have pointed to a
total absence of Laurel in the eastern area by AD. 800 (uncorrected) at the
latest. Walter Kenyon produced a date of AD. 940 i 100 from the Armstrong
Mound at the Manitou Mounds Site on the Rainy River in Ontario (Wilmeth
1971), however a recent assay of the bone produced a date of AD. 160 j: 75
(William A, Ross, personal communication), a date more consistent with both
the nature of the ceramics which appear to be early Laurel, and with the
construction of the mound that has a Hopewellian-like log crib tomb. Two
Hopewell stone pipes were found in the fill of the mound. Site UNR 23 in
northern Manitoba produced an apparently acceptable date of AD. 1030 .4: 150
(Dickson 1976), however sediments are typically very thin in that area, and both

mixture and contamination are possible. The Ash Rapids Site on Lake of the

93

Woods in Ontario produced an assay of AD. 1230 i 55 (Reid 1984) from a
hearth at the interface between Middle and Late Woodland strata. The feature
contained both Laurel and Late Woodland ceramics. Again it is unclear whether
some mixture may be involved. Dawson (1981) received a date of AD. 1240 .4:
175 from the Wabinosh River Site at Lake Nipigon in north central Ontario,
however he questions the date and accepts a date of AD. 855 i 180 for the
same stratum. Reid and I reported three dates from House #1 at the
Ballynacree Site in Kenora, Ontario, of AD. 1240 i 45, AD. 1240 i 65, and AD.
1270 :t 55 (Reid and Rajnovich 1991). All three were from the floor or posts of
the house and appeared in good contexts, but they have been questioned as
far too late for Laurel. However, Meyer et al reported two dates for the Crown
Site in northeastern Saskatchewan of AD. 1175 i 155 and AD. 1305 i 70 from
a stratum containing both Laurel and Avonlea material (Meyer ef al 1988). It
may very well be that Laurel extends into the Late Woodland period in the
western area. Several researchers have commented on the possibility of a
temporal overlap between Laurel and Late Woodland in the area (Stoltman
1974, Lugenbeal 1976, Syms 1977), a concept in line with multilineal
ethnogenesis: not every one in a culture changes his/her style all at once.

Ceramics for each western Laurel complex are characterized by relative
frequencies of decorative techniques rather than presence or absence factors.
Lugenbeal noted, using types defined by Stoltman (1973), that the Pike Bay
complex has a high proportion of Laurel Oblique and Laurel Plain. The
McKinstry complex has a large amount of Laurel Pseudo-scallop Shell and
Laurel Bossed, and the Smith complex is characterized by a high proportion of
Laurel Dentate and Laurel Punctate. Laurel Cord-wrapped Stick is introduced
in the Hungry Hall complex (Lugenbeal 1976: 571).

The series of radiocarbon dates from eastern components is ambiguous

94

for the following reasons. Dawson (1981) has discussed a number of early
dates in the area including two from the mouth of the Michipicoten River on the
northeast shore of Lake Superior of 1165 BC. i 425 from the Michipicoten
Harbour Site and 535 B.C. : 250 from the Wawa Site (Brizinski and Buchanan
n.d.). While both sites contained ceramics, the manufacturing technique is
paddle-and-anvil, not characteristic of Laurel pottery which is coiled. The lithics
have affinities with Archaic and Early Woodland types rather than Laurel
(Dawson 1981: 37). Dawson mentions another early date of 490 BC. i 140
from the eastern Lake Superior area, reported by Thor Conway, but he notes
that contextual details have not been published. He rejects another date of 180
30. i 280 from the Montreal River Site on Lake Timiskaming because the
sample is from pit refuse associated with Early Woodland Vinette 1 Ware, not
from the Laurel component at the site (Dawson 1981: 37). The Montreal River
date, along with a date of 80 BC. i 200 from the Killamey Site on the north
shore of Lake Huron that has Adena or Havana Plain sherds (Crane and Griffin
1959), may provide terminal dates for Early Woodland in the region but not very
informative ones; we are dealing with large standard deviations.

The same is true for the series of dates specifically associated with
Laurel and Laurel-like ceramics in the eastern area. Dates from Summer Island
at the mouth of Green Bay of AD. 70 i 280, AD. 160 i 130, and AD. 250 i 200,
from a component with Laurel-like material, range over 600 years (uncorrected)
and span nearly 800 years when calibrated. (Brose 1970). The date from
Naomikong Point of AD. 430 i 400 is nearly meaningless (Janzen 1968).

Other eastern dates have been questioned for a number of reasons:
Wright rejected the series of dates from Heron Bay on the north shore of Lake
Superior on the basis of a lack of clear association with Laurel material (Wright

1967: 95). Conway received a date of AD. 360 i 60 from Sinclair Cove on the

95

northeastern shore of Lake Superior, but he has not published a report on the
site. The Sand River date of AD. 320 i 100 from a Laurel component on the
northeastern shore of Lake Superior contained one rim sherd (Wright 1967:
69). Of the three dates from the Portage Site, the earliest has been rejected on
the grounds that its location in the upper zone of the site is out of order with the
dates from the lower zone (Lovis, Rajnovich and Bartley 1998: 91).

Reid and I plotted the distribution of known radiocarbon dates and found
that the earliest, based on the mean uncorrected assays, were in the Boundary
Waters of Minnesota and Ontario. Two expansions apparently occurred, the first
northwestward to Manitoba and a succeeding one in the second century AD.
eastward to the Upper Peninsula of Michigan. The plots also suggest a
retraction at about the ninth century AD. to the Boundary Waters (Reid and
Rajnovich 1991: 222-223). The Portage dates have been published since that
study and the plot remains the same, but so does the problem: the radiocarbon

dates are too ambiguous. As a result, this study will rely on seriation.

Stylistic Seriation
Hypotheses 4 and 5 can be tested using style analysis. They state that 4)
one Laurel assemblage/complex is earlier than others, and indicates expansion
from one source within the Laurel world, and 5) all early
assemblages/complexes are apparently contemporaneous, representing a
convergence of two or more sources of Laurel. Some testable corollaries can
be inferred:
4a) One western assemblage will seriate earlier than the
earliest eastern one.
4b) One eastern assemblage will seriate earlier than the

earliest western one.

96

5a) Some eastern assemblages are as early as western
assemblages.

To seriate the western and eastern assemblages, the clusters for each
site (see appendices) must be integrated into super-clusters drawn from all the
assemblages studied in order to produce a common set of categories for
comparison. Each cluster procedure produced groups that were duplicated for
the most part in the others. For instance, all assemblages have a
plain/punctated cluster; many have an unbanked pseudo-scallop shell stamped
group, and a few have a dragged, banked pseudo-scallop shell stamped
cluster. The super-clusters could be observed simply by inspection, an
analytical procedure to define second-order patterning of data (Romesburg
1984: 280). The individual clusters are the first-order patterns of similarity, or
intra-site vessel-to-vessel similarity; the super-clusters form second-order
patterns of similarity derived from comparison of the site clusters, or inter-site
assemblage-to-assemblage similarity. Table 7 describes the clusters obtained
and Appendix M lists the vessels in each super-cluster for each assemblage.
Each super-cluster is given a letter designation and is described by its major
attribute; all have exterior rim technique as the major attribute except Clusters A,
B and C which are mostly undecorated and are defined by surface treatment,
and Cluster P which is defined by having decoration on the upper rim only,
eschewing decoration on the lower rim. Frequencies and percentages are
given for each assemblage, and the Naomikong Point assemblage has been
split into the hypothesized east and west occupations. Of the 162 Naomikong
vessels, only 147 had extant provenience identification. Cluster B, the group
with cord-wrapped paddle impressed surface treatment, is included in the list for
the purpose of comparisons to material at an inter-regional level to be

conducted later in this study.

97

03090 n .5 ”09.03 H dd Edam £8900:de u mmm ”0.0% 800936.00 u 950

 

 

Sofia 2393 3.09 Fd 20d9d0F 3.099 8.0058 80: NFF 60: 2 8.298 39238
- - 6.9 d E9 2 - - 3.9 d - ad: d 2:0 55 580.0
Add: ANdcd 6.0; EFFVNF 8.va A0.09F.N AdFVN fiNvN 6.0de 58:85.0
8.de FF 3.9 m a9 d E9 2 ANN: d 8.90 dd :9 d - - .80: dd .2
- 8.9 d 3.9 d 8.9 F 1.9 F - - 6.9 N :9 N .85.. .2
8.0: F. 6.9 N 8st NF 6.9 d 8.9 N 3.9 N 2.9 d 8.9 d 8.9 F 99:8 dd .._
- 8.9 F 6.9 F. 6. F0 N Add: 1 6.9 d 3N9 dd 8.9 d - 29:8 .0.

- 8.9 F 6.9 0 EN: dF - - - - A59 d ddn. dd .5 .F.
:9 N d. F F9 d a9 d Add: 9 ANN: d 8.9 v - - - ddn. dd ._
- Add; 8.200 Addch 8.2? 3:00 39d 21.09% cat: dda .:
5.0 d Add: 0 6. F0 F 6.9 d 8.0: d 8. F0 F :9 d 8. F9 F - 03.8. .0
- - 8. F0 F 6.9 d 3.9 F - - - - 02d: d
8st a - - a: N - 8.: F :9 d - - 28 20:5 .d
6.9 F - - - - - 6.9 v - - 950 .0
- - 6F: - - - - - - Bawad .o

- - - . - - - - . 00200 .d
8.de 9 3.09 2 Sid aid 8.9N 8.5 8.30? FNd9 Ne EFNVd 9805955 d

 

wadtoa >> 39:88: 8.3.30 m ocovmlEOdz :dEgeo d. 5E5, omaodcédm gnaw 9.0.. 0.90
523.0 zoom .3. =9; BEE...“— 05 3 32.830 «25 .953 do. 222:0 .3oo> E 2nd...

.285

98

The first step of analysis was a temporal ordering of the eastern Laurel
assemblages. By comparing frequencies of vessels in each super-cluster from
site to site, the Brainerd coefficient of similarity was developed for all pairs of
assemblages (Brainerd 1951): 200 denotes identical assemblages, and 0
denotes no similarity between assemblages. A flow chart depicting the best
seriation was produced from the highest coefficients. The second step involved
both spatial and temporal patterns: it compared the eastern seriation to western
temporal seriation, thereby adding the spatial dimensions for early, middle and

late Laurel.

Eastern Laurel

The coefficients of similarity for the six eastern assemblages are shown
in Table 8 in the best-fit order, and they are charted in Table 9. The seriation
generally works except for the western component of the Naomikong Point Site,
and it could be argued that it is out of order. The ordering indicates temporal
rather than geographical order; for instance, Whitefish Island is spatially
between Cloudman and Naomikong Point but it appears at one end of the
seriation. Whitefish Island is probably at the earliest end of the eastern Laurel
seriation and the Portage Site at the latest. Whitefish Island has a vessel with
interior cord marking, a generally Early Woodland trait, while the Portage Site
introduces a significant amount of cord-impressed decoration, a generally Late
Woodland trait.

As Table 7 demonstrates, Whitefish Island is characterized by a nearly
uniform series of banked linear stamped vessels - Clusters N and O comprising
banked and dragged banked linear stamped vessels make up nearly three-
quarters of the assemblage. Whitefish Island is weakly related to the Cloudman

Site assemblage, about a quarter of which is banked linear stamped.

99

 

000000 SS .2, 00055002 0.00 020.56 0.02 .0 00005002 F.00F 0050006 0.00 50932,

 

aoua.nE¢on< .953 533m 05 96:5 09:05:20: "0 030... 0

10

 

- 39 0.00 2:. SF 30 000000
33 - 0.00 0.0: 3.: 0.00 .2, 003.5002
0.00 0.00 - 0.00F 0dNF 0.00 00.0036
SK 0.0: 0.02 - N00F 0.0 .0 00055002
BF 0.3 F 0dNF ~00F - 0.00 00280.0
:0 0.00 0.00 90 0.00 - £000.23

wagon. .>> 0000:8002 didfio .w 0000.00.82 000.0020 @233
3am .230.— Eoudum 0.: .3. 3.3.3.5 .0 3:22:30 "0 030...

Cloudman also contains an increased number of Clusters H and l, pseudo-
scallop shell stamped groups, and Cluster K, dentate stamped, and a decrease
in the linear stamped varieties. Cloudman is followed by the eastern occupation
of the Naomikong Point Site with an increase of Clusters H, I and J, the pseudo-
scallop shell groups. The Gyftakis Site follows with a marked increase in
Clusters K and L, the dentate stamped varieties. Next comes the western
occupation of the Naomikong Point site with a notable increase in Cluster A,
plain vessels, and then the Portage Site, also with plain vessels but with a
renewed emphasis on banked linear stamped pots and an introduction of a
significant number of single-cord impressed vessels.

The coefficients of similarity in Table 9 indicate that at least three, and
perhaps four, complexes may be present. Some complexes apparent in this
study are represented by only one assemblage and should be tested with more
data. Whitefish Island is only weakly related to the Cloudman Site at 98.5 and
could form a separate Whitefish Island Complex representing early Laurel in the
eastern composite. Cloudman, Naomikong East and Gyftakis are strongly
related through the Cloudman-Naomikong East coefficient of 138.7 and the
Naomikong East-Gyftakis coefficient of 133.5. Gyftakis is also more strongly
linked to Cloudman at 124.6 than to Naomikong West at 99.9, and possibly
should be included in a Cloudman Complex. However, the high incidence of
banked dentate stamped designs at Gyftakis compared to any other site in the
study (nearly 30%) would beg for a separate Gyftakis Complex as Fitting
suggested for a time period of about AD. 300 to AD. 500 (1979:112). It should
be noted that the Gyftakis assemblage contains vessels from burial context, a
fact that may skew the results. Another weak link is Gyftakis and Naomikong
West at 99.9: Naomikong West appears to form a separate Naomikong

Complex. Likewise, the Portage Site is only weakly related to Naomikong West

101

at 100.1 and forms a separate Portage Complex (Lovis, Rajnovich and Bartley
1998: 109). I could not relocate the assemblage from the McGreggor Site that
Fitting used as his definition of the McGreggor Phase, estimated to date after
about AD. 500 (Fitting n.d.: 182). It may be that his McGreggor Phase has been
separated in this study into the Naomikong and Portage Complexes, however
the frequencies Fitting presented for the McGreggor Site ceramics are
significantly different than either Naomikong West or Portage (Fitting n.d.: 146
and 149).

The Whitefish Island Complex is characterized by the use of extremely
homogeneous banked linear stamped designs. The Cloudman Complex
continues the use of banked linear stamped decoration, and increases the use
of pseudo-scallop shell stamps. The Gyftakis Complex emphasizes dentate
stamping while the Naomikong Complex emphasizes plain pottery. The final,
Portage Complex presents a diverse set of vessels that are dominated by plain,
banked linear stamped and single-cord-impressed decoration. As stated
previously, the complexes must be seen as tentative in light of their definition

using single assemblages in some cases.

Western Laurel

The vessel clusters for the three selected Laurel assemblages from the
Fisk, Long Sault and Ballynacree sites are shown in Table 7, and the
coefficients of similarity are given in Table 10. None of the coefficients are large:
the Fisk-Long Sault coefficient is 52.5, Fisk-Ballynacree is 30.3, and Long
Sault-Ballynacree is 64.0. This is an expected result given the long time span
involved between the early Fisk Site radiocarbon dated at AD. 50 i 115 and
the Ballynacree Site dated after AD. 1200. Nevertheless, the coefficients place

the assemblages in the proper order according to the radiocarbon assays, with

102

 

ccbmcémm

_
0.00
_

gnaw 0:3

_
0.3
_

wa

md:

000000 I402 1.; 0805002 -000 lmxmfiolmgp rm 085500210003100.008.010.00 I 000232,

 

>35 05 5 30035002 .953 =0 9.95 09:05:22“. 0:. 030...

 

 

- v0 00 0.00 3.00 0.00 00 «.00 0.0... 008002.00
00 - 00 «.00 0.00 0.00 0.30 300 0.3. 0.000 0:3
00 00 - 0.00 0. 3 0.00 0.0: 30 W00 02“.
0400 «.00 0.00 - 3.8. 0.00 0.00 0.: 3.0 000:8
3.00 30 0: 3.8. - 0.00 0.0: 0.3. 0.00 .2, 002582
0.00 0.00 0.00 0.00 0.00 . 0.000 0.00. 0.00 02000.0
00 0.30 04.0.: 0.00 0.0: a - H.003. 0.30 .w 009.382
0.00 30 F0 02 0.3 3 0.00. H004 . 01.00 00508.0
0.03. 0.3. 3.00 3.0 0.00 0.00 0.30 0.00 . 0000002,
0080.503 035 95.. 0.00.0 canton .>> 99.05002 0% .m 0:0th002 5.5520 5..me

 

 

 

0020 .5 s 8:0 .20.... =0 a. 3:23.70 3 35.0580 "2 2...:

103

Fisk at one end and Ballynacree at the other. The Fisk assemblage is
characterized by a high number of Cluster P, upper rim decoration only, and
large amounts of Clusters H and J, pseudo-scallop shell stamps varieties. The
Long Sault assemblage continues the use of pseudo-scallop shell designs but
introduces plain vessels. Ballynacree continues use of plain vessels and adds
an increase_of dentate stamping. In addition, Ballynacree shows an introduction
of cord-wrapped stick and single cord impressing. The percentages for the
clusters place Fisk in the Pike Bay Complex, Long Sault in the late Pike Bay or
early McKinstry Complex and Ballynacree in the Hungry Hall Complex
(Stoltman 1974: 87, Lugenbeal 1976: 540, Reid and Rajnovich 1991: 221).

SPATIAL DYNAMICS

Laurel dynamics involve both temporal and spatial dimensions. With the
temporal order of both eastern and western assemblages established, the
ceramics could be compared spatially across the Laurel world, in conjunction
with time.

Coefficients of similarity for all Laurel sites in the study are listed in Table
10, and Table 11 presents the relationships between the eastern and western
Laurel sites of the study based on their coefficients of similarity. The Fisk
assemblage is most closely related to Naomikong East while Long Sault is most
closely connected to Naomikong West and Ballynacree is most Closely
_ associated with Portage. The coefficients connecting the western ceramics to
the eastern components are all small, and they decrease with time, suggesting
a decreasing similarity as time goes on. While Ballynacree aligns with Portage
in Table 11, it undoubtedly represents a ceramic style considerably later than
the Michigan sites, and should ideally be charted to the far right of the Portage
Site in Table 11.

104

The Fisk-Naomikong East alignment is unexpected in that one would
predict from the Reid-Rajnovich hypothesis that the early western site would be
earlier than the earliest eastern one, but Fisk is not related to Whitefish Island or
Cloudman.

To study the relationships further, the presence/absence factors of each
ceramic super-cluster were plotted for each assemblage. They are listed in
Table 12. Concerning the three western Laurel assemblages, of the seven
original clusters at Fisk, six are retained throughout the sequence. Only Cluster
J, dragged banked pseudo-scallop shell stamped, drops out. Of the seven
clusters at Long Sault, six are retained in the Ballynacree assemblage.
Ballynacree has the addition of Cluster N, banked linear stamped, replacing
Cluster M, unbanked linear stamped. It also introduces two new clusters, D and
E, both characteristic of Late Woodland ceramic traits - cord-wrapped stick and
single-cord impressed. The pattern is of a very strong tradition in western Laurel
that persists throughout the sequence, and adds Late Woodland traits at the
end of the sequence. Only Clusters J, dragged banked pseudo-scallop shell
stamped, and M, unbanked linear stamped, do not survive the sequence.

Concerning the eastern assemblages, six of the nine clusters at Whitefish
Island persist throughout the sequence. Clusters C, brushed, F, trailed, J,
dragged banked pseudo-scallop shell stamped, M, unbanked linear stamped,
and P, decoration on the upper rim only, are introduced in the middle of the
sequence and then are dropped before the end. The pattern is of clusters
increasing then decreasing in frequency. The eastern group displays a much
less stable tradition than observed in the western assemblages. While 78 °/o of
the clusters in the west are retained by the latest assemblage (Ballynacree),
only 46% of the clusters in the eastern group are retained in the latest

assemblage (Portage).

105

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

- - n. n. - - n. - a
L O O O - O O O O
z z z z z z z - -

. _2 .2 .2 .2 - - 2 s.

._ ._ ._ ._ 4 4 ._ l. ..

- x x x x v. v. x -

- a a a - - - - a

_ _ _ _ _ _ - - -

- I I I I I I I I

w 0 O 0 0 0 0 0 -

- - u n. u. - - - -

m. - - m - m w - -
o - - - - - o - -

I I 0 I I I I I I

< < < < < k < < <
wanton .B 955E002 8.830 .m 99:80.02 5:08:20 .0_ 50:0: cobmcgwm gnaw 0.5.. it

._.w<m hwlmB’

 

35 .953 zoom no «conga 9.32:0 "Np oEah

106

The pattern of a conservative tradition in the western group is reinforced by the
radiocarbon dates that indicate a lengthy span for Laurel in the west, Fisk at one
end and Ballynacree at the other. The Fisk assemblage is characterized by a
high number of Cluster P, upper rim decoration only, and extending later than
AD. 1000, compared to the east. An hypothesis for a less conservative tradition
in the east is also reinforced by radiocarbon dates indicating a transition to Late
Woodland about AD. 700 (Lovis and Holman 1976, Fitting 1979, Holman
1984), and by a greater interaction with people having non-Laurel ceramic
traditions as seen in the use, for example, of trailing.

The less conservative tradition is also indicated by the frequencies of
vessels in the east that share decorative designs. The early assemblage at
Whitefish Island is unique in its homogeneity - for instance, Vessels 1 to 5 have
the same decoration, as do Vessels 8 to 13, Vessels 15 to 17, Vessels 18 to 22
and Vessels 23 to 26 (Figure 6); only the profiles vary in these groups. In
contrast, the latest sites are characterized by extreme variability with very few
redundant decorative designs (see Figures 19 and 24, the Gyftakis and Portage
Sites). Whitefish Island has a rate of redundancy at 34% (27 of 79 vessels share
design), while the rate drops precipitously to 21% at Naomikong, 15% at
Cloudman, 11% at Gyftakis and 0% at Portage. Variability in decorative designs
in the west is not so extreme throughout the sequence - 38% at Fisk, 27% at
Long Sault and 17% at Ballynacree, indicating a more conservative tradition.

While the east and west groups share a number of clusters, the data do
not clearly indicate the direction of influence. The coefficients of similarity
indicate that the western Fisk assemblage has an eastern connection with
Naomikong Point East. Three conclusions are possible:

Corollary 4a states that a western Laurel assemblage is earlier than the

earliest eastern one. In this hypothesis, Fisk is earliest and contributes Clusters

107

J, dragged banked pseudo-scallop shell stamped, and P, decoration on the
upper rim only, to Naomikong Point. Cluster M, unbanked linear stamped, could
also have been contributed by Fisk to the east. The justification for this
hypothesis is that the frequencies for those are higher at Fisk. However, the
data indicate that Whitefish Island and Cloudman could be earlier than, or
contemporaneous with, Fisk. They contributed Cluster E, single-cord impressed,
Cluster F, trailed, Cluster l, banked pseudo-scallop shell stamped, and Cluster
N, banked linear stamped, to Naomikong Point. In this scenario, Naomikong
Point represents an amalgamation of two distinctly different style zones in the
east and west.

Corollary 4b states that one or more eastern Laurel assemblages are
earlier than the earliest western one. In this corollary, Naomikong Point East is
earlier than Fisk which represents a selection process involving a few traits
chosen from the larger pool of decorations at Naomikong Point. All of Fisk’s
clusters could have been derived from Naomikong Point. The east could also
have contributed Cluster G, incised, and Cluster K, unbanked dentate stamped,
to the west at any time later than Fisk. This hypothesis would also make
Whitefish Island and Cloudman earlier than Fisk. This scenario is an expansion
model, but expansion only of ceramic traits, not populations, as will be shown
below.

Hypothesis 6 stated that diffusion will be supported if similarities co-vary
with distance, while Hypothesis 7 stated that migration will be supported if two
assemblages in the western and eastern Laurel areas exhibit similarity. The
minimum level of similarity necessary for migration, derived from the
Ballynacree houses, is 128. The Fisk-Naomikong Point similarity coefficient of
119.5 does not reach the critical number. Rather, similarity co-varies with

distance, supporting Hypothesis 6, diffusion.

108

Furthermore, it would be difficult to support a migration hypothesis in this
case because the projectile points from the two sites differ extensively. The most
common point type at Naomikong Point, a side-notched form with flaring convex
base (Janzen 1968: 62), was not recovered at Fisk. The closest affinities of the
Fisk points are to Late Archaic-Early Middle Woodland types to the west such as
Pelican Lake and Larter Tanged from Manitoba (MacNeish 1958), not to
eastern Laurel (Wright 1967; Rajnovich, Reid and Shay 1982: 103). Fisk has
groundstone, an Archaic trait, as well as Archaic-like points, and thus appears
to represent a grafting of Laurel ceramics and other characteristics onto a Late
Archaic base. The influence, then, in this scenario would be diffusion rather
than migration.

Traits represented in clusters such as P, decoration on the upper rim
only, and M, unbanked linear stamped, are common on wares adjacent to the
eastern Laurel area such as North Bay Ware, and could have been introduced
to the western area through eastern Laurel. The same could hold for the
subsequent introduction to the west of Cluster G, incised, and K, unbanked
dentate stamped, also traits common on wares adjacent to eastern Laurel, for
example, Dane, Nokomis, and North Bay.

In an effort to determine the kind of diffusion occurring, a study of
secondary traits was undertaken on the assumption discussed in Chapter 1 that
groups in close contact will share not only primary traits but also some
secondary ones as well. An examination of the presence of interior decoration,
lip decoration, and rim-lip juncture ticking, indicates that neither of the first two
are confined to certain clusters nor to particular time periods in the east and
west groups. However, the third trait, lip ticking, is highly significant. That trait is
common on all sites in the east, but is confined to Long Sault in the west. This

minute decoration, visible only on close inspection, is a “smoking gun”

109

indicating close contact between the eastern area and Long Sault potters, a
conclusion consistent with the fact that Long Sault is part of a community of very
special ceremonial sites comprising the main mound-building district of the
whole Laurel world. The inescapable conclusion that people were directly
interacting, not participating only in down-the—line diffusion of ideas, is
supported by evidence that sites reported to date between the east and west
study areas do not exhibit lip ticking (Wright 1967, Dawson 1980, 1981). Wright
does not report lip ticking from his survey of sites along the north shore of Lake
Superior except on rims from the Heron Bay Site and the Eaka Site (Wright
1967: 10 and 80) but Heron Bay is in the eastern area and Eaka is near Lac
Seul in the western area. Dawson does not report lip ticking for either the
MacGillivray Site near Thunder Bay, Ontario, or the Wabinosh River site on
Lake Nipigon (Dawson 1980, 1981). Lip ticking is a common trait of the
eastern assemblages, occurring throughout the sequence, and is uncommon
on western ceramics. The logical conclusion is that the influence is from east to
west at the time of the Long Sault Site.

Most other secondary traits such as the use of punctates or bosses with
specific decorative techniques or designs, or the use of specific designs
coupled with design tool type, show no significant temporal-spatial pattern.
However, the use of both punctates and bosses in multiple rows is a common
western trait not transferred to the east. Trailing, whatever the motif, is an
eastern decorative technique not transferred westward. Unbanked pseudo-
scallop shell stamping combined with bosses is common at Long Sault and
found in lesser numbers at Whitefish Island and Naomikong, suggesting
perhaps a transfer eastward of that combination.

The general pattern is that of east-west sharing among potters in close

contact but neither the primary nor the secondary traits clarifies questions of

110

origin of those traits except in the case of lip ticking. While the direction of
influence remains somewhat obscure, it is clear that at least two separate
populations in the east and west were involved, selecting from a retinue of
shared decorative choices. R. Mason has suggested that there may never have
been an origin for Laurel but rather multiple origins (1991: 138). This study
supports that hypothesis, a view in concert with Moore’s (1994) and Clarke’s
(1968) notions of ethnogenesis. Populations did not spread throughout the
Laurel world, but ideas did, creating a recognizable cultural entity out of
disparate groups.

In later Laurel, a different model emerges; east and west assemblages
become increasingly different with low coefficients of similarity; while Fisk
seriates with Naomikong Point at nearly 120, the later assemblage coefficients
fall below 100, supporting the divergence model of Hypothesis 8. The model
can be demonstrated most clearly in the fact that Laurel continued after about
AD. 700 (uncorrected) in the west but underwent extreme changes in the east
to emerge as early Late Woodland.

In summary, the evidence indicates that at least two groups of people in
the eastern and western zones of the study area interacted to form Laurel. The
direction of influence remains unclear, although at least one set of data
suggests at least one direction of information exchange was east-to-west: the
presence of lip ticking on many eastern vessels and only a few western ones
may imply a development of that trait in the east and its transfer to the west.
Migration was not a major factor in the expansion of Laurel: the coefficients of
similarity between the zones are small, indicating relatively weak interaction not
consistent with an expectation of migration. The early western Laurel artifact
assemblages include some Archaic traits common to the western zone, but not

the eastern one, an indication that the western Laurel people retained some

111

local cultural traits. The early development Laurel as a whole must be
understood as a convergence. The later evolution of Laurel can be expressed
as a divergence: western Laurel ceramics remained stylistically conservative
while the eastern assemblages demonstrate probable influences from various

neighboring groups.

112

7. SUMMARY AND CONCLUSIONS

Moore has put forth a theory of “ethnogenesis” to explain the formation of
social groups, suggesting that the process can be viewed as the convergence
of several streams which then flow through time on a somewhat homogeneous
course, and eventually diverge into several new streams (Moore 1994:930).
Archaeologists can never study the formation of social groups directly, but
theories of style as constituted by patterned human behavior open the door to
the probability that all the pots studied in this dissertation have something to say
about the people who made them. They indicate that Moore’s model is
appropriate for the Laurel data. The ceramic evidence suggests Laurel potters
interacted with most groups adjacent to them, but there were no wholesale
borrowings of style. Early Laurel lithic assemblages contain indigenous Archaic
features, and early Laurel pottery has traits probably derived from neighbors,
but it also has a distinctive character of its own. It is likely that the Laurel
population was a convergence of people from several of these sources. Laurel
most likely did not have one origin; Laurel people “were not half-baked
Hopewellians”, as Janzen put it (1968: 108), nor were they half-baked Point
Peninsulans or Malmoans. Laurel developed a widespread but variable culture
with important traits, or adaptations to the Lake Forest and Boreal Forest, such
as fishing nets, toggle-head harpoons, lithic technology based on the bipolar
core technique, and use of copper as well as ornately decorated ceramics that
can be described as baroque when compared to many other Woodland wares.

Analysis of Laurel ceramics can be used to explore three significant
questions: (1) origins of Laurel, 2) the processes of interaction externally with its
neighbors and (3) the dynamics of change within Laurel.

A number of hypotheses were generated concerning origins:

113

1) Spatial analysis will reveal that the earliest Laurel
assemblage/complex is most similar to one or more of its external
neighbors. The test for strength of similarity through style analysis

will show that similarity is close enough to conclude that one or

more neighbors had a major influence on Laurel, or vice versa.

The study has shown that Laurel and Crawford Ware of the northeastern
Plains exhibit a pattern of avoidance, but Laurel shares some traits with its other
neighbors, Malmo, Nokomis, North Bay, Saugeen, and Point Peninsula and all
are in close proximity to Laurel sites, even appearing with Laurel on some sites.
The percentage of shared traits is low and the direction of influence is two-way.

2) Temporal analysis will reveal that the external neighbor(s) is

earlier than the earliest Laurel assemblage/complex and is

therefore a good candidate for an origin of Laurel style.

While the radiocarbon dates for Malmo have been questioned, Gibbon
has suggested that it is no earlier than Laurel. Nokomis, North Bay, Saugeen,
and Point Peninsula have been dated earlier than the earliest Laurel
assemblages, however the major traits of early Laurel as revealed by seriation
statistics consist of pseudo-scallop shell stamped, banked stamped and
dragged stamped that do not appear on the early ceramics of Nokomis, North
Bay, Saugeen, or Point Peninsula. Therefore Hypothesis 3 is supported:

3) If there is no candidate for Hypotheses 1) and 2), then internal

Laurel dynamics will explain the development of Laurel.

Two hypotheses were stated concerning Laurel origins.

4) One Laurel assemblage/complex is earlier than others

representing an expansion model, a spread from one source within

the Laurel world.

5) All early assemblages/complexes are apparently

114

contemporaneous representing a convergence model with two or

more origins of Laurel.

The radiocarbon dates from the eastern Laurel area are ambiguous,
either because of their questionable contexts or their large standard deviations.
Therefore, seriation statistics were used to secure their relative temporal
positions. The direction of influence between eastern and western areas is
unclear in the statistics. Those reveal that the Fisk Site, the earliest assemblage
in the western Laurel area, seriates most closely in the east with Naomikong
Point East, not the earliest eastern assemblage; Whitefish Island and Cloudman
precede it. If the direction of influence was Fisk-to-Naomikong Point, then two
sources, or populations, were involved in the development of Laurel, those at
Fisk in the west and those at Whitefish Island and Cloudman in the east. If the
direction of influence was Naomikong Point-to-Fisk, then it is possible that there
was one source for the ceramics, the early assemblage in the east. However
there must have been two sources of populations, east and west, because
Laurel is more than ceramics and its western lithic assemblage reveals a local
Archaic origin. While Hypothesis 4 could be supported in terms of ceramics, it
cannot be supported in terms of actual populations.

Two hypotheses were stated concerning the processes of interaction,
migration and diffusion:

6) Diffusion will be supported if similarity co-varies with distance;

assemblages exhibiting greatest similarity will be closest in space.

7) Migration and not diffusion will be supported if similarity does

not co-vary with distance; two assemblages from western and

eastern Laurel areas will exhibit similarity.

The minimum level of similarity necessary for an inference of migration

was obtained from the assemblages from each of the three houses at

115

Ballynacree. The similarity coefficient for Fisk-Naomikong Point is 120, below
the critical number of 128. Therefore, Hypotheses 7 cannot be supported. The
major process of interaction was diffusion.

Two hypotheses concerning Laurel dynamics were presented:

8) All early assemblages/complexes are similar with increasing

variability among later assemblages, representing a divergence

Mil-

9) All early assemblages/complexes are variable with increasing

similarity among later assemblages, representing the continuation

of the convergence model.

Seriation coefficients decrease from early to late Laurel, culminating in a
complete split about AD. 700 when eastern Laurel ceramics began to develop
Late Woodland traits while western Laurel continued on its traditional course.
Thus, while convergence characterized early Laurel, divergence marked all
ensuing developments.

Brown and Plog have suggested that the pattern of the formation
of “tribal” social networks is one of increasing connectedness in existing
regional networks. A corollary from Moore's theory is that “tribal” networks
eventually split along devel0ping sub-networks in the social unit. The ceramics
studied in this dissertation conform to that pattern. Sub-regional networks that
formed during middle Laurel times occupied the same geographical space as
later complexes in the Late Woodland period. Eastern and western Laurel
gradually took on different forms that appear at the point of the Late Woodland
emergence as separate entities, Mackinac in the east and Blackduck/Selkirk in
the west.

The theoretical importance lies in the fact that the beginning of a new cultural

phenomenon can very well be the outcome of developments in the old one, not

116

an abrupt break in development, however different the styles appear to be: style
change can be drastic as a result of the influx of different influences affecting
each of the sub-regions produced. The study of the development of
geographical sub-regions, or style zones, in cultural entities like Middle
Woodland that precede the appearance of two or more Late Wobdland style
zones in the same geographical spaces can produce a new and useful tool for
understanding cultural origins and continuities.

The style zone approach advocated in this dissertation lends a great deal
of support to the idea that there was continuity between Middle and Late
Woodland peoples of the Upper Great Lakes, previously glimpsed only by the
few ceramic similarities between those cultural periods.

The style changes at the onset of the Late Woodland period could have
been the result of new alliances, sodalities, residence patterns, or the formation
of new lineages or clans. All of the above may have played a role. The ceramic
analysis of this dissertation cannot address these issues specifically. Perhaps a
larger data base that would include a large number of assemblages from each
region could approach the issues of specific social processes.

A major problem with working with Laurel ceramics is that they are
extremely variable, as previous researchers who have attempted to devise
typologies have understood. Ceramic analysis has a built-in contradiction: it
necessarily reduces Laurel and, in doing so, we lose the brilliance of it. The
method of cluster analysis for the formation of polythetic sets is an attempt to
address the problem, but it is cumbersome because the multiplicity of data are
not easily managed. Nevertheless, there is evidence that a polythetic approach
like the cluster method is appropriate. Mary Black, in her dissertation on the
ontology of the Ojibway people who reside today in the Laurel area, some of

whom are most likely descendants of Laurel people, speaks of their world view

117

as composed of polythetic sets (Black 1967, 1968). For instance, the culture
hero Nanabojou can be included in both the categories of Anishinabe (person)
and manitou (spirit). Anishinabe itself can be put into different categories;
depending on the context, it can mean person, Indian or specifically Ojibway
Indian. It seems appropriate, then, to produce polythetic ceramic sets that
include, for instance, dentate stamping in two categories in Table 6, K. Dentate
Stamped and P. Upper Rim Only. A Laurel study should not be confined to
normative types upon which researchers have never agreed, but be open to
free floating sets dependent for their formation on the question at hand.

The substantive contribution of this dissertation has been the production
of a body of data and a seriation for eastern Laurel assemblages. It has been
shown that Whitefish Island is earliest followed by the Cloudman Site, the
eastern occupation at Naomikong Point, the Gyftakis Site, the western
occupation of Naomikong Point, and the Portage Site. The strongest
relationship to selected Western Laurel sites is early in the sequence, with the
Fisk Site most closely related to the eastern portion of Naomikong Point. It could
be that the eastern assemblages each represent a complex of the Syms (1977)
taxonomic system but more data are needed. A complex should be represented
by more than one assemblage because the taxon should represent not only
time but also spatial distribution that reflects primary and secondary utilization of
resources. As it stands, the seriation indicates a north-to-south temporal
gradient similar to the pattern discussed by Lovis and Holman (1976), but data
beyond the selected assemblages in this dissertation will be needed to support
their hypothesis. Whatever the case, the eastern sites form a composite of the
Syms system, what Reid and l have termed the Superior Composite (Reid and
Rajnovich 1991: 226-227).

There are clues that the eastern and western Laurel people maintained

118

regional interaction that was face-to-face, not down the line. The occurrence of
the secondary, minor attribute of lip ticking at the Long Sault Site is intriguing.
Lip ticking is an eastern trait appearing on vessels found at one of the major
ceremonial sites in the Laurel World. Could it be that the Rainy River which has
at least six mound sites, the largest of which is Kay-Nah-Chi-Wah-Nung that
includes Long Sault, attracted people from long distances to its sacred shores?
It would be an example of Syms’ secondary or tertiary resource zone for eastern
Laurel people, but this resource is food for the soul. The idea of such long
distance travel should not be surprising given the wealth of ethnographic data
describing regular journeys between such places as Michilimackinac and Lake
of the Woods.

Evidence for interregional interaction is confined mainly to the eastern
sites where Laurel grades into Point Peninsula, and has been found on
Saugeen and North Bay sites. There is also some indication that Naomikong
Point people interacted in a small way with Nokomis Phase populations. By the
middle of the first millennium, the Laurel World was changing in the east with
the introduction of Late Woodland traits on Pine River Ware, while western
Laurel became more isolated and continued on a conservative, traditional
course for some centuries.

It is tempting to view these processes as contributing to the development
of Cree and Ojibway, the languages known to have been spoken at Lake of the
Woods and at Sault Ste. Marie respectively in the early Historic Period
(Greenberg and Morrison 1982). Support for such a hypothesis requires
extensive studies of the Late Woodland world covering the large gap between
Laurel and the Historic era. Whatever the case, this study has shown that the
Laurel World was complex; it arose out of a number of ethnic streams and

culminated in the genesis of a number of new ones flowing into the Late

119

Woodland. We can discern strong ties between these periods, indicating a

continuous ethnogenesis.

120

BIBLIOGRAPHY

121

BIBLIOGRAPHY

Aldenderfor, Mark S. and Roger K. Blashfield
1984 Cluster Analysis. Sage, London.

Ammerman, A. J. and L. L. Cavalli-Sforza
1973 A Population Model for the Diffusion of Early Farming in Europe. In
The Explanation of Culture Change: Models in Prehistory, edited
by Colin Renfrew, pp. 343-358. Duckworth, London.

Anfinson, Scott E, ed.
1979 A Handbook of Minnesota Ceramics. Minnesota Historical Society,
Occasional Publications in Minnesota Anthropology 5.

Anthony, David W.
1990 Migration in Archaeology: The Baby and the Bathwater.
American Anthropologist 92(4): 895-914.

Arthurs, David
1986 Archaeological Excavations at the Long Sault Site (Manitou
Mounds). Ontario Ministry of Citizenship and Culture,
Conservation Archaeology Reports, Northwestern Region
7.

Benn, David W. N
1990 Woodland Ceramics. In Woodland Cultures on the Western
Prairies: The Rainbow Site Investigations, edited by David W.
Benn, pp. 96-144. University of Iowa, Office of the State
Archaeologist, Report 18.

Bettarel, Robert and Sidney Harrison
1962 An Early Ossuary in Michigan. Michigan Archaeologist 8:37-42.

Black, Mary
1967 An Ethnoscience In vestigation of Ojibwa Ontology and World
View. Ph.D. dissertation, Sandford University.

1968 Eliciting Folk Taxonomy in Ojibway. In Cognitive Anthropology,
edited by Stephen A. Tyler, pp. 165-189. Holt, Rinehart and
Winston, New York.

Brainerd, G. W.

1951 The Place of Chronological Ordering in Archaeological Analysis.
American Antiquity 16(4): 301-313.

122

Branstner, Christine N.
n.d. Archaeological Investigations at the Cloudman Site (20CH6): A
Multicomponent Native American Occupation on Drummond
Island, Michigan - 1992 and 1993 Excavations. Paper prepared for
the Consortium for Archaeological Research, Michigan State
University.

Braun, David P. and Stephen Plog
1982 The Evolution of Tribal Social Networks: Theory and Prehistoric
North American Evidence. American Antiquity 47(3): 504-525.

Brightman, Robert A.
1993 Grateful Prey: Rock Cree Human-A nimal Relationships. University
of California Press, Berkeley.

Brizinski, Morris G. and Kenneth T. Buchanan
n.d. Ceramic, Chert and Culture: An Analysis of Three Prehistoric Sites
Located in the Michipicoten Area. Manuscript on file, Ontario
Heritage Foundation.

Brose, David S.
1970 The Archaeology of Summer Island: Changing Settlement
Systems in Northern Lake Michigan. University of Michigan,
Museum of Anthropology, Anthropological Papers 41.

Bruyere, Stacey
1998 Kay-Nah-Chi-Wah-Nung: “Place of the Long Rapids” - A Gathering
Place. Bear Country, In-Flight Magazine of Bearskin Airlines: 12-
20.

Caine-Hohman, Christy Ann
1979 MalmolKern. In A Handbook of Minnesota Ceramics, edited by
Scott F. Anfinson, pp. 137-141. Minnesota Historical Society,
Occasional Publications in Minnesota Anthropology 5.

1983 Normative Typological and Systemic Stylistic Approaches to the
Analysis of North Central Minnesota Ceramics. Ph.D. dissertation,
University of Minnesota.

Carr, Christopher
1995 Building a Unified Middle-Range Theory of Artifact Design:
Historical Perspectives and Tactics. In Style, Society and Person:
Archaeological and Ethnological Perspectives, edited by
Christopher Carr and Jill E. Neitzel, pp. 151-170. Plenum Press,
New York.

123

Carr, Christopher and Jill E. Neitzel, eds.
1995 Style, Society and Person: Archaeological and Ethnological
Perspectives. Plenum Press, New York.

Chute, Janet E.
1998 The Legacy of Shingwaukonse,: A Century of Native Leadership.
University of Toronto Press, Toronto.

Clarke, David L.
1968 Analytical Archaeology. Methuen, London.

Cleland, Charles E.
1982 The Inland Shore Fishery of the Northern Great Lakes: Its
Development and Importance in Prehistory. American Antiquity
47(4): 761-784.

1992 From Ethnohistory to Archaeology: Ottawa and Ojibway Band
Territories of the Northern Great Lakes. In Text-Aided

Archaeology, edited by Barbara J. Little, pp. 87-102. CRC Press,
Ann Arbor.

Cleland, Charles E. and Richard Peske
1968 The Spider Cave. University of Michigan, Museum of
Anthropology, Anthropological Papers 34: 20-60.

Conkey, Margaret
1991 Experimenting with Style in Archaeology: Some Historical and
Theoretical Issues. In The Uses of Style in Archaeology, edited by
Margaret Conkey and Christine Hastorff, pp. 3-17. Cambridge
University Press, Cambridge.

Conway, Thor A.
n.d. Whitefish Island - A Remarkable Archaeological Site at Sault Ste.
Marie, Ontario. Manuscript on file, Ontario Ministry of Culture,
Tourism and Recreation, Thunder Bay.

Crane, H. R. and James B. Griffin
1959 University of Michigan Radiocarbon Dates IV. American Journal of
Science, Radiocarbon supplement 1.

Davis, Leslie B., ed.

1988 A van/ea Yesterday and Today: Archaeology and Prehistory.
Saskatchewan Archaeological Society, Saskatoon.

124

Davis, Whitney
1991 Style and History in Art History. In The Uses of Style in
Archaeology, edited by Margaret Conkey and Christine Hastorff,
pp. 18-31. Cambridge University Press, Cambridge.

Dawson, K. C. A.
1974 The McCIuskey Site. National Museum of Man, Mercury Series
25.

1976 Algonkians of Lake Nipigon: An Archaeological Survey. National
Museum of Man, Mercury Series 48.

1980 The MacGillivray Site: A Laurel Tradition Site in Northwestern
Ontario. Ontario Archaeology 34: 3-16.

1981 The Wabinosh River Site and the Laurel Tradition in Northwestern
Ontario. Ontario Archaeology 36: 3-36.

DeBoer, Warren
1991 Interaction, Imitation and Communication as Expressed in Style:
The Ucayali Experience. In The Uses of Style in Archaeology,
edited by Margaret W. Conkey and Christine A. Hasdorff, pp. 32-
43. Cambridge University Press, Cambridge.

Deetz, James

1965 The Dynamics of Stylistic Change in Arikara Ceramics. University
of Illinois Press, Urbana.

Denny, J. Peter
1992 The Entry of the Algonkian Language into the Boreal Forest. Paper
presented at the annual conference of the Canadian
Archaeological Association, London, Ontario.

Densmore, Frances

1910 Chippewa Music. Smithsonian Institution, Bureau of American
Ethnology, Bulletin 45.

Dickson, Gary A.
1976 Recent Radiocarbon Dates from Northern Manitoba. Manitoba
Historic Resources Branch, Papers in Manitoba Archaeology,
Miscellaneous Paper 3.

Dirst, Victoria
1995 Shanty Bay: a Great Place to Camp in Door County Vlfisconsin
Department of Natural Resources, Bureau of Property
Management.

125

Engelbrecht, William
1974 The Iroquois: Archaeological Patterning on the Tribal Level. World
Archaeology 6(1): 52-65.

1978 Ceramic Patterning Between New York Iroquois Sites. In The
Spatial Organization of Culture, edited by Ian Hodder, pp. 141-
152. University of Pittsburgh Press, Pittsburgh.

Farnsworth, Kenneth B. and Thomas E. Emerson, eds.
1986 Early Woodland Archaeology. Center for American Archaeology.
Kampsville.

Fischer, Fred W.

1972 Schultz Site Ceramics, In The Schultz Site at Green Point: A
Stratified Occupation Area in the Saginaw Valley of Michigan,
edited by James E. Fitting, pp. 137-190. University of Michigan,
Museum of Anthropology, Memoir 4.

Fitting, James E.
1975 The Archaeology of Michigan: A Guide to me Prehistory of the
Great Lakes Region. Cranbrook Institute of Science, Bloomfield
Hills.

1979 Middle Woodland Cultural Development in the Straits of Mackinac
Region: Beyond the Hopewell Frontier. In Hopewell Archaeology:
The Chillicothe Conference, edited by David S. Brose and N’omi
Greber, pp. 109-112. Kent state University Press, Kent.

n.d. The Gyftakis and McGreggor Sites: Middle Woodland Occupations
in St. Ignace, Michigan. Paper prepared for Michigan History
Division of the Michigan Department of State.

Fitting, James. E., David S. Brose, Henry T. Wright and James Dinerstein
1969 The Goodwin-Gresham Site, 20 IS 8, losco County, Michigan. The
Wisconsin Archeologist 50(3): 125-183.

Gibbon, Guy E.
1972 Cultural Dynamics and Development of the Oneota Life-way in
Wisconsin. American Antiquity 37(2): 166-185.

1975 The Brower Site: A Middle Woodland Mound and Camp Site on
Lake Onamia. The Minnesota Archaeologist 34(1-2): 1-42.

1994 Cultures of the Upper Mississippi River Valley and Adjacent
Prairies in Iowa and Minnesota. In Plains Indians A.D. 500-1500:
The Archaeological Past of Historic Groups, edited by Karl H.
Schlesier, pp. 128-148. University of Oklahoma Press, Norman.

126

Greenberg, A. M. and J. Morrison
1982 Group Identities in the Boreal Forest: The Origin of the Northern
Ojibway. Ethnohistory 29(2): 75-102.

Gregg, Michael L.
1990 An Early Plains Woodland Structure in the Northeastern Plains.
Plains Anthropologist 35-127: 29-44.

1994 Archaeological Complexes of the Northeastern Plains and Prairie-
Woodland Border, A.D. 500-1500. In Plains Indians AD. 500-
1500: The Archaeological Past of Historic Groups, edited by Karl
H. Schlesier, pp. 71 -95. University of Oklahoma Press, Norman.

Hardin, Margaret Friedrich
1970 Design Structure and Social Interaction: Archaeological
Implications of an Ethnographic Analysis. American Antiquity
35(3): 332-343.

Hegmon, Michelle
1986 Information Exchange and Integration on Black Mesa, Arizona,
AD. 931 -1 1 50. In Spatial Organization and Exchange:
Archaeological Survey on Northern Black Mesa, edited by
Stephen Plog, pp. 256-281. Southern Illinois University Press,
Carbondale.

Hoddenlan
1979 Economic and Social Stress and Material Culture Patterning.
American Antiquity 44(3): 446-454.

Hodson, F. R.
1969 Searching for Structure Within Multivariate Archaeological Data.
World Archaeology 1(1): 90-105.

Holman, Margaret R.
1984 Pine River Ware: Evidence for In Situ Development of the Late
Woodland in the Straits of Mackinac Region. The Wisconsin
Archeologist 65(1): 33-48.

Hruska, Robert
1967 The Riverside Site: A Late Archaic Manifestation in Michigan. The
Wisconsin Archeologist 48(3): 145-260.

Jackson, Lawrence

1968 New Evidence for Early Woodland Seasonal Adaptation from
Southern Ontario, Canada. American Antiquity 51: 389-401.

127

James, Edwin, ed.

1956 A Narrative of the Captivity and Adventures of John Tanner. Ross
and Haines, Minneapolis. (First published 1830)

Janzen, Donald E.
1968 The Naomikong Point Site and the Dimensions of Laurel in the
Lake Superior Region. University of Michigan Museum of
Anthropology, Anthropological Papers 36.

Johnson, Elden

1971 Excavations at the Gull Lake Dam (21CA27). The Minnesota
Archaeologist 31(2): 44-69.

Johnson, Leroy Jr.
1972 Introduction to Imaginary Models for Archaeological Scaling and
Clustering. In Models in Archaeology, edited by David L. Clarke,
pp. 309-381. Methuen, London.

Kaplan, Flora S. and David M. Levine
1981 Cognitive Mapping of a Folk Taxonomy of Mexican Pottery: A
Multivariate Approach. American Anthropologist 83(4): 868-884.

Kehoe, Alice B. and Thomas F. Kehoe
1973 Cognitive Models for Archaeological Interpretation. American
Antiquity 38(2): 150-153.

Kempton, Willett

1981 The Folk Classification of Ceramics: A Study of Cognitive
Prototypes. Academic Press, New York.

Kenyon, Walter A.
1970 The Armstrong Mound on Rainy River, Ontario. National Historic
Sites Service, Occasional Papers in Archaeology and History 3:
65-83.

1986 Mounds of Sacred Earth. Royal Ontario Museum, Toronto.

Kluth, Rose
1995 The Cass Lake I site: Direct Evidence for Prehistoric Wild Rice in
the Mississippi Headwaters Region. Midwest Archaeological
Conference, 40th annual meeting, South Beloit.

Koezur, Polly and James V. Wright

1976 The Potato Island Site: District of Kenora, Ontario. National
Museum of Man, Mercury Series 51.

128

Kohl, Johann Georg
1985 Kitchi-Gami: Life Among the Lake Superior Ojibway. Minnesota
Historical Society, St. Paul.

Lenius, Brian and David Olinyk
1990 The Rainy River Composite: Revisions to Late Woodland
Taxonomy. In The Woodland Tradition in the Western Great Lakes:
Papers Presented to Elden Johnson, edited by Guy E. gibbon, pp.
77-112. University of Minnesota, Publications in Anthropology 4.

Loew, Patti
2001 Indian Nations of Wisconsin: Histories of Endurance and Renewal.
Wisconsin Historical Society Press, Madison.

Long, A. and M. Tamplin
1977 University of Arizona Radiocarbon Dates from Archaeological
Sites in Manitoba. Manitoba Historic Resources Branch, Papers in
Manitoba Archaeology, Miscellaneous Paper 4.

Lovis, William and Margaret B. Holman
1976 Subsistence Strategies and Population: A Hypothetical Model for
the Development of Late Woodland in the Mackinac Straits-Sault
Ste. Marie Area. Michigan Academician 8(3): 267-276.

Lovis, William, Grace Rajnovich and Aryn Bartley
1998 Exploratory Cluster analysis, Temporal Change and the Woodland
Ceramics of the Portage Site at L’Arbre Croche. The Wisconsin
Archeologist 79(1): 89-1 12.

Lugenbeal, Edward N.
1977 The Archeology of the Smith Site: A Study of the Ceramics and
Culture History of Minnesota Laurel and Blackduck. Ph.D.
dissertation, University of Wisconsin-Madison.

MacNeish, Richard S.
1958 An Introduction to the Archaeology of Southeast Manitoba.
National Museums of Canada, Bulletin 157.

Mason, Carol I.
1964 An Early Woodland Vessel from Wisconsin. The Wisconsin
Archeologist 45(4): 158-160.

1981 Excavations on the Little Eau Pleine: Site 6. The Wisconsin
Archaeologist 62(1): 84-108.

129

Mason, Ronald J.
1966 Two Stratified Sites on the Door Peninsula of Wisconsin.
University of Michigan, Museum of anthropology, Anthropological
Paper 26.

1967 The North Bay Component at the Porte Des Morts Site, Door
County, Wisconsin. The Wisconsin Archeologist 48(4): 267-342.

1969 Laurel and North Bay: Diffusional Networks in the Upper Great
Lakes. American Antiquity 34(3): 295-302.

1981 Great Lakes Archaeology. Academic Press, New York.

1991 Rock Island and the Laurel Cultural Frontier in Northern Lake
Michigan. Midcontinental Journal of Archaeology 16(1): 119-155.

Matson, R. G. and D. L. True
1974 Site Relationships at Quebrada Tarapaca, Chile: A Comparison of
Clustering and Scaling Techniques. American Antiquity 39(1): 51-
74.

McPherron, Alan
1967 The Juntunen Site and the Late Woodland Prehistory of the Upper
Great Lakes Area. University of Michigan, Museum of
Anthropology, Anthropological Papers 30.

Meinholz, Norman M. and Steven R. Kuehn
1996 The Deadman Slough Site. State Historical Society of Wisconsin,
Archaeology Program, Archaeological Research Series 4.

Metzger, Duane G. and Gerald E. \Mlliams
1966 Some Procedures and Results in the Study of Native Categories:
Tzeltal Firewood. American Anthropologist 68(2-1) 389-407.

Meyer, David
1983 Saskatchewan Laurel: An Overview. Saskatchewan Archaeologist
4: 3-24.

1995 Saskatchewan River Rendezvous Centers and Trading Posts:
Continuity in a Cree Social Geography. Ethnohistory 32(3): 403-
444.

Meyer, David and Scott Hamilton
1994 Neighbors to the North: Peoples of the Boreal Forest. In Plains
Indians A. D. 500- 1500: The Archaeological Past of Historic
Groups, edited by Karl H. Schlesier, pp. 96-127. University of
Oklahoma Press, Norman.

130

Meyer, David and Dale Russell

1987 The Selkirk Composite of Central Canada: A Reconsideration.
Arctic Anthropology 24(2): 1-37.

Meyer, David, Olga Klimko and James Finnegan
1988 Northernmost Avonlea in Saskatchewan. In Avonlea Yesterday
and Today: Archaeology and Prehistory, edited by Leslie B. Davis,
pp. 33-42. Saskatchewan Archaeological Society, Saskatoon.

Milner, Claire McHale
1998 Ceramic Style, Social Differentiation, and Resource Uncertainty in
the Late Prehistoric Upper Great Lakes. Ph.D. dissertation,
University of Michigan.

Moffat, Charles R.
1999 Archaeological Investigations at Several Wisconsin River
Headwaters Reservoirs: Results of the 1997 and 1998 Field
Seasons. University of Wisconsin-La Crosse, Mississippi Valley
Archaeology Center, Reports of Investigations 315.

Moore, John H.
1994 Putting Anthropology Back Together Again: The Ethnogenetic
Critique of Cladistic Theory. American Anthropologist 96(4): 925-
948.

Morlan, Richard E.
1988 Avonlea and Radiocarbon Dating. In Avonlea Yesterday and
Today: Archaeology and Prehistory, edited by Leslie B. Davis, pp.
291-310. Saskatchewan Archaeological Society, Saskatoon.

Peters, G. R., J. O. Hunn, K. A. Motivans and W. A. Okstad
1983 Cultural Resource Management of the Lake Superior National
Forest: 1982 Report. USDA Forest Service, Superior National
Forest, Cultural Resource Report 1.

Pihl, Robert H.
1995 Ceramics at the Ault Park Site, Ontario: Wnette 2 Ware and Point
Peninsula Chronology. Paper presented at the 40th annual
Midwest Archaeological Conference, South Beloit.

Pilon, Jean-Luc
1987 Ecological and Cultural Adaptation along the Severn River in the
Hudson Bay Lowlands of Ontario. Ontario Ministry of Citizenship
and Culture, Conservation Archaeology Reports, North western
Region 10.

131

Pollock, John W,

1976 The Culture History of Kirkland Lake District, Northeastern Ontario.
National Museum of Man, Mercury Series 54.

n.d. Archaeological Resource Evaluation of Whitefish Island
Prehistoric Site, Sault Canal. Manuscript on file, Ontario Ministry of
Culture, Tourism and Recreation, Thunder Bay.

Rajnovich, Grace

1980 Ballysadare (Dka-10): a Laurel-Blackduck Site at the Source of
the Winnipeg River. Ontario Archaeology 33: 37-58.

1983 The Spruce Point Site: a Comparative Study of Selkirk
Components in the Boreal Forest. Ontario Ministry of Citizenship
and Culture, Conservation Archaeology Reports, North western
Region 1.

1994 Reading Rock Art: Interpreting the Indian Rock Paintings of the
Canadian Shield. Natural Heritage, Toronto.

Rajnovich, Grace, C, S. “Paddy” Reid and C. Thomas Shay
1982 Rescue Excavations at the Fisk Site (Dle-1) in Northwestem
Ontario. Ontario Ministry of Citizenship and Culture,
Archaeological Research Report 18.

Reid, C. S. “Paddy”
1984 Some Woodland Radiocarbon Dates from Northwestern Ontario
with Comments on Significance. Ontario Archaeology 41: 37-44.

Reid, C. S. “Paddy” and Grace Rajnovich
1986 Laurel Architecture: Five Case Studies. The Minnesota
Archaeologist 44(2): 5-30.

1991 Laurel: A Re-evaluation of Spatial, Social and Temporal
Paradigms. Canadian Journal of Archaeology 15: 193-234.

Renfrew, Colin

1984 Approaches to Social Archaeology. Harvard University Press,
Cambridge.

Renfrew, Colin and Gene Sterud

1969 Close-Proximity Analysis: A Rapid Method for Ordering of
Archaeological Materials. American Antiquity 34: 265-277.

132

Riddle, David K.
1980 Archaeological Survey of the Upper Albany River. In Studies in
West Patricia Archaeology No. 1: 1978-1979, edited by C. S.
Reid, pp. 153-225. Ontario Ministry of Culture and Recreation,
Archaeology Research Report 15.

Ritchie, William A. and Richard S. MacNeish
1949 The Pre-Iroquoian Pottery of New York State. American Antiquity
2: 97-124.

Robinson, W. S.
1951 A Method for Chronologically Ordering Archaeological Deposits.
American Antiquity 16: 293-301.

Romesburg, H. Charles
1984 Cluster analysis for Researchers. Lifetime Learning, Belmont.

Rouse, Irving
1986 Migrations in Prehistory: Inferring Population Movement from
Cultural Remains. Yale University Press, New Haven.

Sack, Robert D.
1980 Conceptions of Space in Social Thought: A Geographic
Perspective. University of Minnesota Press, Minneapolis.

Sackett, James R.
1982 Approaches to Style in Lithic Archaeology. Journal of
Anthropological Archaeology 1 (1 ): 59-1 12.

1985 Style and Ethnicity in the Kalahari: A Reply to Wiessner. American
Antiquity 50(1 ): 154-159.

Salzer, Robert J.

1969 An Introduction to the Archaeology of Northern Wisconsin. Ph.D.
dissertation, Southern Illinois University.

1974 The Wisconsin North Lakes Project: A Preliminary Report. In
Aspects of Upper Great Lakes Anthropology: Papers in Honor of
Lloyd A. Wilford, edited by Elden Johnson, pp. 40-54. Minnesota
Historical Society, St. Paul.

Spaulding, Albert C.

1953 Statistical Techniques for the Discovery of Artifact Types.
American Antiquity 18: 305-313.

133

Spence, Michael W. and William A, Fox
1986 Early Woodland Occupations of Southern Ontario. In Early
Woodland Archaeology, edited by Kenneth B. Farnsworth and

Thomas E. Emerson, pp. 4-46. Center for American Archaeology,
Kampsville.

Spence, Michael W., Robert H. Pihl and Carl R. Murphy
1990 Cultural Complexes of the Early and Middle Woodland Periods. In
The Archaeology of Southern Ontario to AD. 1650, edited by
Chris J. Ellis and Neal Ferris, pp. 125-169. Ontario Archaeological
Society, London Chapter.

Stuiver, M. and P. J. Reimer

1993 Extended 140 Data Base and Revised CALIB 3.0 140 Calibration
Program. Radiocarbon 35: 215-230.

Stoltman, James B.
1962 A Proposed Method of Systematizing the Modal Analysis of
Pottery and its Application to the Laurel Focus. Master’s thesis,
University of Minnesota, Minneapolis.

1973 The Laurel Culture in Minnesota. Minnesota Historical Society,
St. Paul.

1974 An Examination of Within-Laurel Cultural Variability in Northern
Minnesota. In Aspects of Upper Great Lakes Anthropology: Papers
in Honor of Lloyd A. Wilford, edited by Elden Johnson, pp. 74-89.
Minnesota Historical Society, St. Paul.

1997 The Archaic Tradition. The Wisconsin Archeologist 78(1/2): 112-
139.

Syms, E. Leigh
1977 Culture Ecology and Ecological Dynamics of the Ceramic Period
in Southwestern Manitoba. Plains Anthropologist 22(76-2),
Memoir 12.

Thomas, Cyrus
1894 Report on the Mound Explorations of the Bureau of Ethnology.
Smithsonian Institution, Bureau of American Ethnology, 12th
Annual Report.

Thwaites, Reuben Gold, ed.

1959 The Jesuit Relations and Allied Documents, 1670. Pageant Press,
New York.

134

Valpuu, Seppo H. and George (Rip) Rapp
2000 Paleoethnobotanical Context and Dating of the Laurel Use of Wild
Rice: The Big Rice Site. The Minnesota Archaeologist 59:81-87.

Vansina, Jan
1985 Oral Tradition as History.University of Wisconsin Press, Madison.

Voss, Jerome and Robert L. Young
1995 Style and the Self. In Style, Society and Person: Archaeological
and Ethnological Perspectives, edited by Christopher Carr and Jill
Neitzel, pp. 77-99. Plenum Press, New York.

Warren, William W.
1970 History of the Ojibway Nation. Ross and Haines, Minneapolis.

Whallon, Robert Jr.
1972 A New Approach to Pottery Typology. American Antiquity 37:13-
33.

Wiessner, Polly
1983 Style and Social Information in Kalahari San Projectile Points.
American Antiquity 48(2): 253-276.

1985 Style or lsochrestic Variation? A Reply to Sackett. American
Antiquity 50(1): 160-166.

Wiersum, W.E. and M. A. Tisdale
1977 Excavations at UNR23, the Notigi Lake Site. Manitoba Historic
Resources Branch, Publications in Archaeology.

VViIford, Lloyd A.
1955 A Revised Classification of the Prehistoric Cultures of Minnesota.
American Antiquity 21(2): 130-142.

\Nilley, Gordon R. and Philip Phillips
1958 Method and Theory in American Archaeology. University of
Chicago Press, Chicago.

\Nilmeth, Roscoe
1971 Canadian Radiocarbon Dates. National Museum of Man, Bulletin
232.

Wobst, H. Martin

1974 Boundary Conditions for Paleolithic Social Systems: A Simulation
Approach. American Antiquity 39(2): 147-178.

135

1977 Stylistic Behavior and Information Exchange. In For the Director:
Papers in Honor of James B. Griffin, edited by Charles E. Cleland,
pp. 317-344. University of Michigan, Museum of Anthropology,
Anthropological Papers 61.

1978 The Archaeo-Ethnology of Hunter-Gatherers or the Tyranny of the
Ethnographic Record in Archaeology. American Antiquity 43(2):
303-309.

Wright, James V.
1967 The Laurel Tradition and the Middle Woodland Period. National
Museums of Canada, Bulletin 217.

1972 The Shield Archaic. National Museums of Canada, Publications in
Archaeology 3.

Wright, James V. and J. E. Anderson
1963 The Donaldson Site. National Museums of Canada.,BulIetin 184.

136

APPENDICES

137

APPENDIX A

Heunesmecenomnmecemmcmms
Profits-mutton...) ) I ’
Manhunt“
I

LowerRiimledt [M
Shoulder -
fibrfipfluior
c Cord-umppedpaddiemessadi—ahumws)
S Smoothed-ovarcord-wrwpadpaddeitvessad
B arusiiedrcombed
m UNI-MM
£532 Dentatestariped
1w 13%me
«It 089906le
///l W
I] Trailed
:1- Museum
anu- Single-COMIMM
fifi Notimpimed
.e Exteriorpunctatemmss
0 Hole
0 Annulapunctde
a Exterioraoas
n Thumbna'lincised
w Rockerstampcd

 

 

 

 

 

 

 

 

 

 

138

APPENDIX B

TABLE 13: AN EXAMPLE OF THE CERAMIC DECORATIVE VARIABLES USED
IN THE LAUREL STUDY

Table 13 lists the decorative variables used for the Whitefish Island Site
assemblage. The left column delineates the surface finish, decorative technique
and decorative motif for each of the fields. They are composed of interior, lip,
rim-lip juncture, exterior upper rim, exterior lower rim/neck, and shoulder. The
second column describes the attributes present in the assemblage, and the
third column lists the attribute state (1 or 0) for each vessel. The data sheet for
each assemblage varies slighfly according to the attributes present at each site.

For example, the criss-crossed motif is present in some assemblages but not in
others.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

FIELD {ATTRIBUTE PRESENCE!
.............. ABSENCE
VESSEL NUMBER W1

PROFILE §-4s Degrees Everted 1
................................... EStraight : 0
. ........ Inverted 0
INTERIOR RIM SURFACE Plain 1
.......... Corded 0
INTERIOR RIM TECHNIQUE _______________ Incised 0_
. ____ Linear Stamped _1
Dentate Stamped 0.
_________ Pseudo-Scallop Shell 0,
, ______ gPunctated o
INTERIOR RIM MOTIF Diagonals 0
Verticals , 1
LIP SURFACE .............................................. Plain 1
LIP TECHNIQUE Pseudo-Scallop Shell 0
..................... Incised 0
_. ________ §Linear Stamped 0.,
.UP MOTIF ......................................... EDiaaonals ....................................................... Q.
iTransverse 0

 

 

 

139

Table 13 (cont'd).

 

EXTERIOR RIM-UP TECHNIQUE

LInearStamped

 

Pseudo-Scallop Shell

 

EXTERIOR RIM-LIP MOTIF

Verticals

 

 

Diagonals

 

EXTERIOR UPPER RIM SURFACE

Plain

 

EXTERIOR UPPER RIM TECHNIQUE

Pseudo-Scallop Shell

 

 

gLInear Stamped

 

: Dentate Stamped

 

 

L Punctated

 

' Incised

 

 

Single Cord Impressed

 

EXTERIOR UPPER RIM MOTIF

Crlss-Cross

 

Verticals on Horlz.

 

Diagonals on Horlz.

 

 

5 Banked

 

:BankedDragged

 

Horlzontals

 

Horizontals on Vert.

 

Diagonals on Dlag.

 

LOWER RIMINECK SURFACE

Plain

 

LOWER RIM/NECK TECHNIQUE

Punctated

 

Bossed

 

 

Incised

 

 

UnearStemped

 

Pseudo-Scallop Shell

 

Dentate stamped

 

 

 

LOWER RIMINECK MOTIF

Single Cord Impressed

Crlss-crossed

 

Horlzontals

 

Verticals on Horlz.

 

 

 

Diagonals on Horlz.

 

 

 

 

 

SHOULDER SURFACE

 

SHOULDER TECHNIQUE

Pseudo-Scallop Shell

 

Incised

 

 

UnearStamped

 

Dentate Stamped

 

SHOULDER MOTIF

Horlzontals

 

Verticals on Horlz.

 

Cries-crossed

 

Rocker Stamped

 

 

 

i Banked

 

OOOOOOOOOJOOJJOOOOOOJCOO-‘O:OO‘O-‘JOOO

 

140

 

APPENDIX C

WHITEFISH ISLAND (CdlC-2)

The rapids of the St. Mary’s River between the twin cities of Sault Ste.
Marie, Ontario and Michigan, has long been known as a special place of the
Ojibway people. Its traditional importance as a center of the Midewiwin is
emphasized in a Mide’s song collected by Frances Densmore at the turn of the
twentieth century: “At the long rapids, I am called to go in, my Mide Brethren”
(Densmore 1910: 63). Whitefish Island in the rapids was chosen in 1990 as the
site for a rare gathering of the Three Fires nations (Ojibway, Odawa and
Potawatomi) and it has been the subject of a hotly disputed land claim for much
of the century. The island was under the control of the Canadian federal
government for several decades, but has recently been returned to the
possession of the Batchewawa Band of the Ojibway First Nations.

The island was the focus of test excavations from 1976 to 1979. Thor
Conway of the Ontario Ministry of Culture and Recreation located the site and
conducted the investigation (Figure 4) as part of a Sault-area survey intended to
develop a culture history. Conway produced only preliminary statements of the
results of the survey, prior to completion of the excavation of the Laurel
component. He commented that Laurel material was confined to one area of the
site (Conway n.d.: i). Site maps were lost during a move of the ministry offices
(letter from Conway to Andrew Hinshelwood 1991) and notes are also missing.
Nevertheless, the artifacts have been preserved for study in the archaeology
laboratory of the Ontario Ministry of Culture, Tourism and Recreation in Thunder
Bay.

John W. Pollock, an archaeological consultant, provided the ministry with

141

 

 

 

 

 

Figure 4: The location of Whitefish Island at Sault Ste. Marie.

 

 

’Np .

St. Mary's Island

   
    
 
 

Whitefish Island

St. Mary's River Rapids

100m

—=l

 

Figure 5: Excavation areas on Whitefish Island. Area L is Laurel.

142

 

a map of the excavation areas as part of a resource evaluation study in 1983. It
reveals that the site is horizontally stratified (see Figure 5, a modification of
Pollock’s map) with Laurel at Area L, in the center of the island. It is the highest
point of the island, between 597 and 600 feet above sea level, the same
elevation as the Cloudman Site down river on Drummond Island (Branstner
n.d.: 6). An Archaic component was found at Area M; Mackinac Ware was
recovered at Area C, and Bois Blanc at Area BB. The eastern areas labeled A
contained historic material (Pollock n.d.). The island is composed mainly of
cobbles with no vertical stratigraphy apparent among the rocks (Conway n.d.:
3). No radiocarbon dates were obtained.

The 80 vessels of sufficient size for this analysis reveal a surprising
homogeneity in decoration. The vast majority are banked stamped, and some
vessels are identical. For instance, Vessels 1 to 5 (see Figure 6, the ceramic
chart, and Figure 7) all have right-leaning banked linear stamps, tick marks at
the lip-rim juncture, and interior vertical linear stamps. Vessels 8 to 13 have
exactly the same exterior decoration but lack interior decoration. Banked
stamping is an early trait of Laurel Ware in its western distribution in Minnesota
and northwestern Ontario.

The cluster tree for the 80 Whitefish Island vessels produced the clusters
described in Table 13. The Whitefish Island assemblage (Figures 7 and 8) is
most unusual in that the majority of vessels are in a single cluster: 52.5 per cent
of the vessels are in Group 1 alone, banked oblique linear stamped designs,
with lip ticking on 18 of the 42 vessels. Another 13.8 per cent make up Group 2,
banked vertical linear stamped vessels. Group 3 vessels (7.5 per cent) are
banked linear stamped with the addition of punctates. In all, banked linear
stamping makes up nearly three-quarters of the assemblage. Group 4 is a

banked pseudo-scallop shell stamped set; Group 5 is banked dentate stamped,

143

and Groups 6 and 7 represent unbanked pseudo-scallop shell stamped and
unbanked dentate stamped vessels. Only one pot, Vessel 34, an outlier of
Group 5, has punctates as the sole exterior decoration. Three vessels are
outliers of all groups. Vessel 7 has horizontal single cord impressions
reminiscent of Mackinac Ware but with Laurel profile and paste. Single-cord
impressing occurs also at the Portage Site and Summer Island. Vessel 30 has a
dentate stamped exterior and a cord-marked interior, a trait common to Early
Woodland ceramics elsewhere. Vessel 77 features interior punctates that do not
form exterior bosses, an idiosyncratic feature for Laurel ceramics which
commonly have exterior bossing.

Whitefish Island is so homogeneous that it must be considered a single
occupation where the same potters often repeated their template-like designs. It
also contains one vessel with a trait characteristic of Early Woodland and
seriation (see Chapter 5) places Whitefish Island at the early end of the eastern

sequence.

144

Table 14: Whitefish Island Vessels (f=80)

Group 1: (f=42: %=52.5) Banked, oblique linear stamped vessels.

8 r 1A i=2 '%= 1
Profile - Everted less than 45 degrees and straight
Interior - Plain or vertical linear stamps
Up-Pmm
Exterior Lip-Rim Juncture - Linear stamp tick marks on 14 vessels
Exterior Upper Rim - Banked oblique linear stamps
Exterior Lower Rim/Neck - Banked oblique linear stamps
Vessel # - 1 to 5, 8 to 13, 15to 17, 18to 22, 35, 38, 39, 44, 67

Outlier (f=1; °/o=1.3)
Lower Rim/Neck - Banked linear stamp with punctates and criss-
cross incising
Vessel # - 46

Su rou 1B f=15' %=1 .8

Exterior Lip-Rim Juncture - Linear stamp tick marks on 4 vessels
Exterior Upper Rim - Banked, dragged oblique linear stamps
Exterior Lower Rim/Neck - Banked, dragged oblique linear stamps
Vessel # - 6, 23, 24, 25, 26, 36, 40, 45, 48, 52, 54, 57, 63, 64, 79

Millslfi/Ew
Interior - Vertical dentate stamps
Exterior Lower Rim/Neck - Banked, oblique linear stamps and
punctates
Vessel # - 65

Group 2: (f=11; %=13.8) Vessels with banked vertical linear stamps.

Subgroup 2A ”=5; °/o=§.3)
Profile - Straight or everted less than 45 degrees
Interior - Plain
Lip - Plain
Exterior Upper Rim - Banked, undragged vertical linear stamps
Exterior Lower Rim/Neck - Banked, undragged vertical linear
stamps
Vessel # - 32, 66, 68, 69, 7O

Subgroup 23 jf=§1 °/o=7.§)
Lip - Plain (Vessel 53 has oblique linear stamps)
Exterior Upper Rim - Vessel 80 has vertical P88

145

Table 14 (cont’d).

Exterior Lower Rim/Neck - Banked, dragged oblique linear stamps
Vessel # - 27, 47, 53, 55, 59, 80

Outliers: (f=2; °/o=2.5) Two vessels are outliers of groups 1 and 2. Vessel 60
has upper rim punctates and linear stamp lip ticking, and lower rim banked
linear stamp. Vessel 71 has linear stamp lip ticking over oblique banked
stamping in oblique lines.

Group 3: (i=6; °/e=7.5) Vessels with banked linear stamps in patterns similar to
Groups 1 and 2 with the addition of lower rim punctates.
Vessel # - 28, 29, 58, 61, 62

Outlier:(F=1;°/o=1.3)
Exterior Upper Rim - Criss-cross incising

Group 4: ( :6; %=7.5) Banked pseudo-scallop stamped vessels.

Subgroup 4A “=4; °/o=§.Ql
Profile - Everted less than 45 degrees and inverted

Interior - Plain

Lip - Plain

Exterior Upper Rim - Banked vertical or oblique PSS (Vessel 14
also has punctates)

Exterior Lower Rim/Neck - Banked vertical or oblique PSS and
punctates

Vessel # - 14, 31, 33, 41

W
Exterior Rim-Lip Juncture - PSS tick marks
Exterior Upper Rim - Unbanked oblique PSS
Exterior Lower Rim/Neck - Banked oblique PSS
Vessel # - 50, 51

Group 5: (f=3; %=3.8) Banked dentate stamped vessels.

Profile - everted less than 45 degrees

Interior - Plain

Lip - Plain

Exterior Upper Rim - Banked, oblique dentate stamps

Exterior Lower Rim/Neck - Banked, oblique dentate stamps and
punctates

Vessel # - 72, 78

146

Table 14 (cont’d).

Outlier (f=1;°/o=1@
Exterior Upper Rim - Plain
Exterior Lower Rim - Punctates on a plain surface
Vessel # - 34

Group 6: (f=6; %=7.5) Vessels with vertical-over-horizontal pseudo-scallop
shell stamping.

Subgroup 6A (f=3; °/o=3.8)
Profile - Everted less than 45 degrees

Interior - Plain

Lip - Plain or transverse PSS

Exterior Upper Rim - Unbanked vertical PSS
Exterior Lower Rim/Neck - Unbanked horizontal PSS
Vessel # - 73, 74, 75

Outlier (f=1; °/o=1.3l
Exterior Lower rim - Vessel 42 has punctates and SCI horizontals

W5)
Exterior Upper Rim - Unbanked oblique PSS
Exterior Lower Rim/Neck - Horizontal PSS or dentate and bosses
Vessel # - 43, 49

Group 7: (f=2: %=2.5) Unbanked dentate stamped vessels.

Profile - Everted less than 45 degrees

Interior - Plain

Lip - Plain

Exterior Upper Rim - Unbanked vertical or oblique dentate stamps
Exterior Lower Rim - Unbanked horizontal dentate stamps

Vessel # - 37, 76

Outliers: (f=3; °/o=3.8) Three vessel are outliers of all groups.

Vessel 7 has exterior horizontal single-cord impressions.

Vessel 30 has dentate stamps over a plain surface exterior and a cord-marked
interior.

Vessel 77 has vertical linear stamps over PSS on the exterior and interior
punctates that do not form exterior bosses.

 

147

 

W... I I l I I l

 

 

 

m
B

a... I-sl l 2|

 

N
19

 

 

 

 

 

 

 

 

 

l

 

 

gig
HE]

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ll l 3° a
'quu I F
M 7 I e e o o
l 3|
[:3 IIIII
mum ‘I3 2 l
Vllll’; a "u ,
32
.0 U 0000 '4 | G
"If"! e o
"H"
rmru 33
mum ‘
I5
-'7 2 e o e 0 ‘ O
. V 34 _T A
rm l8 , ”II I
G -22 2 2 l 32:1: (1
In”
um 35
I?”
ll 53., l l l G
00:0
. “I I
l 27 I 252': ‘
' , z ..'.-'..:.2. . .,
37
”1777'
31:35: 28 raft:
' " 32‘s}.
38 film

 

 

Figure 6: Whitefish Island ceramic chart.

148

 

WHTEFISH ISLAND SITE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

, 1 ii, 39 ..
A l 49
ll .. fill U '.".'1
‘ 50 ‘\‘\‘I§l
L”...- he‘s
‘ 5' O.
f' l, U 42 u
.W, e, I
ll "4. j a
(l ‘” 5
.ml
55 ‘m~ W.
. T46 I am
* 56
U 47 ' ll
" u 48 " (I

 

 

 

 

Figure 6 (cont’d): Whitefish Island ceramic chart.

149

 

 

I I II II

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

111159 I '
I . ., ll
:3: l
2311': (I ...f‘éa ' ' . U
nu w l A—
I ' ' I
I
I ' ' l
I vs “ '
' ll I I}; U
l 75 -3"
ll U 6" ' (l
II I .
r; 67 '
..r G ... I 78.0 ...
, es | 5.“.-
.ll ' - I 79% ll
99,, I I Tr""l
IICI , .fifl

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 6 (cont’d): Whitefish Island ceramic chart.

150

 

Figure 7: Vessels from the whitefish Island Site. Top and middle rows, banked,
oblique linear stamped, constituting Group 1; Bottom row, banked vertical linear

stamped, constituting Group 2. The vessel at middle left also has incising.

151

‘f'
'50

Figure 8: Vessels from the Whitefish Island Site. Top row, banked pseudo-

 

- -. 2-1

scallop shell stamped from Group 4; bottom row, unbanked pseudo-scallop

shell stamped from Group 6. The vessel at lower left also has dentate stamps.

152

APPENDIX D

NAOMIKONG POINT SITE (20CH2)

In the 19603, little was known about Laurel settlement distribution across
the Great Lakes area, and even less was understood about Laurel community
patterns, the spatial plans of a single site. James E. Fitting and Donald E.
Janzen of the University of Michigan set about a multi-year excavation of the
Naomikong Point Site to address the second question (Janzen 1968). The
researchers wanted to conduct more than a culture history; they were interested
in spatial and cultural dimensions of Laurel with an emphasis on the cultural
dimension including group size, subsistence practices, and habitation
patterning.

The Naomikong Point Site (Figure 9) is a multi-component site at the east
end of Tahquamenon Bay on the south shore of Lake Superior, about 30 miles
west of Sault Ste. Marie in the Upper Peninsula of Michigan. The sandy beach
had been the focus of surface collections since at least the 19303 (Janzen
1968: 1).

Fitting and Janzen chose the site to elucidate issues of community
patterns for two reasons. First, surface collections indicated it was large: Fittings
and Janzen’s excavations totaled 2,750 square feet and stretched along 190
feet of the south Superior shore. Janzen’s surface surveys and test trenches
were designed to find the limits of the site. He found none and concluded that
the Naomikong Point Site represented an enormous occupation area; he
estimated that his work and Fitting’s together covered only 1% of the site
(Janzen 1968: 91). Second, its main component was a rich Laurel deposit

eventually yielding sherds of more than 200 Laurel vessels.

153

wmmsul
POINTI

-— I 4..

3“

 

 

 

 

WHITEFISH BAY l

‘

in“, I NAOMIKONG
“ °_ POINT!
SAULT STE.

MARIE

 

 

Figure 9: Naomikong Point on the south shore of Lake Superior (from Janzen

  

 

1968).
//1 _ fizz—x

 

 

 

 

/

3
II“ ‘
I
O

 

 

 

 

 

 

 

 

 

 

 

 

\... . ' ks

\,0. %
\/;~> ~/

/ K” °"‘\

V/\ \

Figure 10: Excavation blocks at Naomikong Point. The darkened portion
represents a 1965 test trench. The light squares are the 1966 and 1967
excavations (from Janzen 1968).

154

In 1965, Fitting excavated a 5x20-foot test trench and added three block
units, A, B, and C in 1966 (Figure 10). In 1967, Janzen added blocks D and E
and a 5x60-foot trench connecting the two major areas. The intent was to
expose large areas of the site to reveal spatial patterning reflecting social
activities. The units were excavated in two levels, surface to the top of the
Middle Woodland component and the Laurel component itself (Janzen 1968:
19). The top level contained historic and Late Woodland material. Janzen
attempted to excavate the Laurel component, a shallow deposit no more than
0.7 of a foot, in two levels but he found that concentrations of sherds extended
from top to bottom. Accordingly, he excavated the Laurel deposit as a single
level. The results of the three years of work were produced as Janzen’s doctoral
dissertation in 1968. He interpreted the Middle Woodland spatial data as
indicating two separate Laurel occupations, one later than the other. Janzen
obtained an unhelpful radiocarbon date of AD. 4301400 from scrapings on a
pseudo-scallop shell stamped vessel (1968292). The assay spans the entire
Laurel period in the eastern Lake Superior area.

He noted a discontinuity of artifact debris near the east end of the 5x60-
foot trench connecting the two blocks (Janzen 1968: 21). He concluded that
features and artifact concentrations occur in in two separate areas, the east and
west blocks (1968227). He interpreted the concentrations as houses or activity
areas and estimated the population of each at a minimum of 15 people (1968:
92). As at most northern sites, there were few faunal remains, but Janzen
concluded that the site was a fishing station; he recovered 296 notched stones
interpreted as net sinkers. The site was historically known as the best place in
the bay to catch whitefish (1968: 90).

In his ceramic analysis, Janzen considered decorative tool type as the

principal attribute superseding all others (1968: 39). On that basis, he separated

155

the assemblage by inspection into five ceramic types, Laurel Incised (f=1
vessel), Laurel Dentate (i=3 vessels), Laurel Plain (f=25 vessels), Laurel Linear
Stamped (f=102 vessels), Laurel Pseudo-scallop Shell Stamped (f=93 vessels),
and Laurel Pseudo-scallop Shell Stamped var. Naomikong Point (62 vessels)
(Janzen 1968: 47-55). The last category was distinguished on the presence of a
greater proportion of dragged stamps and the use of a smaller tool than the
main type. He compared the distribution of the three largest categories and
found that the linear stamped vessels conformed to the expected distribution of
uniformity, if it was assumed that the two excavation blocks represent a single
occupation. The expected distribution would be 54 vessels in the larger east
block and 35 in the west block; the actual distribution was 57 and 32. However,
the two pseudo-scallop shell stamped categories showed entirely different
distributions: 93% were recovered from the east block. Janzen concluded that
the east block was earlier than the west area. He noted that Laurel Linear
Stamped and Late Woodland Mackinac Ware share the modes of undecorated
body and oblique rim decoration, and that the distribution at Naomikong Point of
linear stamped vessels approximated the distribution of Mackinac vessels.

l undertook a cluster analysis to test Janzen’s hypothesis of two
occupations, and seriation analysis to assess their relative temporal placement.
The study included 162 Naomikong Point vessels represented by sherds large
enough for inspection from rim to full neck. The analysis did not weight
attributes or assume types by inspection as in Janzen’s analysis. The vessels
are depicted in Figure 11, the master ceramic chart and Table 15 presents a
description of each cluster. Selected sherds from Janzen (1968) are shown at
Figure 12.

Groups 1 to 4 are pseudo-scallop shell stamped vessels; Groups 5 to 8

are linear stamped; Group 9 is a category based on vertical motifs; Group 10

156

Table 15: Naomikong Point Vessels (f=162)

Group 1: (f=20; %=12.3) Unbanked Pseudo-scallop shell stamped vessels in
patterns combining upper verticals or obliques over horizontal lines.

Profile - Everted less than 45 degrees, straight or inverted
Interior - Plain

Lip - Plain, transverse PSS or oblique linear stamped

Exterior Upper Rim - Vertical or oblique PSS

Exterior Lower Rim/Neck - Horizontal PSS with or without
punctates or bosses (Vessels 10 and 15 have vertical PSS)
Vessel # - 4, 10, 15, 20, 33, 34,46, 48, 58, 60, 69, 75, 85, 90, 91.
94, 99, 100, 101, 102

Group 2: (i=9, %=5.6) Banked and dragged pseudo-scallop shell stamped
vessels.

Profile - Straight or inverted

Interior - Plain

Lip - Plain (Vessel 1 has transverse linear stamps)

Exterior Upper Rim - Dragged or undragged vertical or oblique
PSS (four vessels also have punctates)

Exterior Lower Rim/Neck - Banked and dragged, oblique/vertical
PSS (Vessel 1 has oblique columns of oblique PSS)
VesseI#-1,2,3,61,79,80, 133,134,163

Group 3: (f=25, %=15.4) Banked, undragged pseudo-scallop shell stamped
vessels.

Profile - Everted less than 45 degrees or straight

Interior - Plain or PSS or linear stamps (Vessel 62 has a combed
interior surface)

Lip - Plain or transverse or oblique PSS

Exterior Upper Rim - Banked oblique PSS (several are unbanked
and Vessels 9 and 31 have SCI rather than PSS)

Exterior Lower Rim/Neck - Banked oblique PSS with or without
punctates (Vessels 9 and 31 are SCI with punctates)

Vessel #- 9, 14, 16 19, 31, 53, 54, 55, 56, 57, 62, 70, 71, 72, 73.
76, 77, 81, 82, 83, 84, 98, 103, 108

Outlier IL=1. °/o=0.6)

Vessel 50 includes interior linear stamp and exterior lip-rim PSS tick
marks.

157

Table 15 (cont’d).

Group 4: (f=6, %=3.7) Vessels like Group 2 pseudo—scallop shell stamped
designs but with lip decoration.
Profile - Everted less than 45 degrees or inverted
Interior - Plain or oblique PSS
Lip - Oblique or transverse PSS
Exterior Upper Rim - Banked, dragged oblique PSS
Exterior Lower Rim/Neck - Banked, dragged oblique PSS (Vessel
141 has an incised line and undragged PSS and Vessel 45 has
horizontal undragged PSS))
Vessel # - 45, 63, 64, 65, 67, 141

Outlier: (f=1; %=0.6) Vessel 152 has unbanked oblique dentate stamps over
banked, undragged PSS.

Group 5: (f=14; %=11.7) A strong cluster of banked and dragged linear
stamped vessels.
Profile - Everted less than 45 degrees, straight or inverted
Interior - Plain or vertical linear stamped
Lip - Plain (Vessel 89 has transverse linear stamps)
Exterior Upper Rim - Banked, dragged oblique linear stamps
(Vessel 159 has undragged vertical linear stamps and Vessel
150 has both PSS and linear stamps)
Exterior Lower Rim/Neck - Banked, dragged oblique or vertical
linear stamps with or without punctates
Vessel # - 5, 8, 28, 38, 74, 89, 137, 138, 145, 150, 159

Outliers: (f—g; °1e=1.9) Vessels 44 and 122 have vertical columns of
banked, dragged linear stamps instead of horizontal rows. Vessel 153
has the horizontal rows plus banked, dragged linear stamps in arcs.

Group 6: (f=8; °/0=4.9) A weak group of vessels with linear stamp designs with
different upper and lower motifs.

Profile - Everted less than 45 degrees

Interior - Plain

Lip - Plain or linear stamps

Exterior Upper Rim - Unbanked vertical linear stamps, PSS,
dentate or plain

Exterior Lower Rim/Neck - Banked, dragged vertical or oblique
linear stamps

Vessel # - 24,40, 51, 68, 155, 157, 164, 165

158

Table 15 (cont’d).

Group 7: (i=9; %=5.6) Vessels with undragged, banked linear stamps.

Profile - Everted less than 45 degrees

Interior - Plain

Lip - Plain

Exterior Upper Rim - Banked, undragged oblique linear stamps
(Vessels 41 and 96 have dentate stamps and Vessel 162 also
has punctates)

Exterior Lower Rim/Neck - Banked, undragged oblique linear
stamps with or without punctates

Vessel #- 21, 22,41, 47, 96, 97, 130, 151,162

Group 8: (f=7; %=4.3) A weak group of vessels with banked linear stamps in
oblique designs.

Profile - Everted less than 45 degrees, straight or inverted
Interior - Plain

Lip - Plain

Exterior Upper Rim - Unbanked vertical or oblique linear stamps
with or without punctates

Exterior Lower Rim/Neck - banked, undragged linear stamps in
oblique columns

Vessel #- 13, 49, 139, 131, 143

Outliers: (f=2; %=1.2) Two vessels are outliers, Vessel 132 has criss-
cross incising over banked vertical and oblique linear stamps. Vessel 86
has vertical dentate stamps over banked vertical linear stamps.

Group 9: (f=13; %=8.0) Vessels with vertical designs in dentate stamps.

r A f: ' °/o= .7
Profile - Everted less than 45 degrees or inverted
Interior - Plain
Up-Pmm
Exterior Upper Rim - Vertical dentates, dragged or undragged,
banked or unbanked (Vessel 126 also has bosses)
Exterior Lower Rim/Neck - Vertical dentates, dragged or
undragged, banked, with or without punctates or bosses (Vessel
66 has horizontal dentates)
Vessel # - 32, 66, 78, 88, 92, 126

159

Table 15 (cont’d).

I'OU B 1:3' °/o=‘l.9
Exterior Upper Rim - Vessel 154 has vertical incised lines
Exterior Lower Rim/Neck - Vessels 135 and 154 have dragged,
banked oblique linear stamps and Vessel 52 has banked, oblique
PSS
Vessel # - 52, 135, 154

Sumroup QC (f=4; %=2.5)
Exterior Upper Rim - Banked vertical linear stamps (Vessel 12 has

vertical PSS)

Exterior Lower Rim - Banked, vertical linear stamps or plain with
punctates (they are like North Bay Ware. Vessel 12 has an oblique
column of linear stamps)

Vessel #- 12, 18, 121, 144

Group 10: (f-=27; %=16.6) Vessels that are plain or decorated on the upper
rim only.

Sumrgup 1QA (f=19; %=11.7)
Profile - Everted less than 45 degrees or straight

Interior - Plain

Lip - Plain

Exterior Lip-Rim - Linear stamp tick marks on Vessels 29 and 39
Exterior Upper Rim - Plain or vertical incising or bosses

Exterior Lower Rim/Neck - Plain

Vessel # - 29,37,39,104 -107,109 - 120

Subgroup 19B (f=6, %=3.7)
Exterior Upper Rim - Oblique PSS or linear stamps

Exterior Lower Rim/Neck - Vessel 23 has two oblique trailed lines
Vessel #- 6, 7, 11, 23, 93, 149

Quilters: (f=2; °/o=1.2) Vessel 30 has exterior lip tick marks over upper rim
linear stamps and Vessel 127 has exterior lip tick marks over lower rim linear

stamps.
Group 11: (f=4; %=2.5) Vessels with banked incised designs.

Profile - Everted less than 45 degrees

Interior - Plain

Lip - Plain or incised

Exterior Upper Rim - Banked vertical or oblique incising (Vessel 26
is not banked)

160

Table 15 (cont’d).

Exterior Lower Rim/Neck - Banked oblique incising with or without
punctates
Vessel #- 26, 136, 146, 148

Group 12: (i=9; %=5.6) Vessels with unbanked incising.

Profile - Everted less than 45 degrees or straight

Interior - Plain

Lip - Plain (Vessel 128 has criss-cross incising)

Exterior Upper Rim - Unbanked diagonal or vertical incised lines
Exterior Lower Rim/Neck - Unbanked horizontal, diagonal, criss-
cross or vertical incising with or without punctates

Vessel # - 27, 35, 36, 123, 128, 129, 140, 156, 161

Outliers: (f=10; °/e=6.2) Ten vessels outlying all groups at the site.

Vessel 17 has lip ticking on the upper rim underlined by a horizontal PSS line
above oblique PSS. The zoning is possibly related to Nokomis Phase.

Vessel 42 has various zones of linear stamps, and Vessel 125 has zones of
both linear stamps and incising.

Vessel 158 has vertical dentates on the upper rim only.

Vessel 160 has linear stamps on the upper rim over horizontal trailed lines and
Vessel 43 has Vertical trailing over punctates. Both may be Lake Nokomis
Trailed vessels.

Vessels 124 and 87 have oblique dentate over vertical columns of oblique
dentate with punctates.

Vessel 95 has vertical oblique dentate over punctates and horizontal incising.
Vessel 25 has oblique trailing over punctates and may be Lake Nokomis
Trailed.

Not Clustered: Vessels 59, 142 and 147 are much alike, all with variations of
a pendant triangle design. They clearly form a cluster but were not included in
the cluster analysis because they are miniature vessels with functions probably
dissimilar to the larger pots.

 

161

represents plain vessels; Groups 11 and 12 are incised, and there are 11
outliers. Thee outliers, Vessels 25, 43 and 160 are trailed, more akin to wares
southwest of Naomikong Point, such as Lake Nokomis Trailed of northern
WIsconsin (Salzer 1974), and Black Sand, Dane, and Prairie Incised (sic) of
southern WIsconsin (Farnsworth and Emerson 1986). Dane Incised also occurs
on North Bay sites on the Door Peninsula of northeastern Wisconsin (Mason
1967). The three miniature vessels, 59, 142 and 147 all have a pendant triangle
motif that also occurs on Dane, Prairie and Black Sand vessels, and on
Nokomis Phase vessels (see Vessels 1 and 20 from Squirrel Dam in this
dissertation). The motif may also occur on Point Peninsula Ware of southern
Ontario and New York (Ritchie and MacNeish 1949) and Hopewell Ware of
southern Michigan (Fitting 1975). The motif is uncommon wherever it is found
and indicates a widespread exchange of ideas.

Table 16 lists the distribution of the 144 vessels of known provenience at
Naomikong Point.

Table 16: Distribution of Vessels at Naomikong Point

Group East Block West Block
1. 13 (12.6%) 3 (6.8%)
2. 8 (7.8%) 1(2.3%)
3. 18 (17.5%) 5 (11.4%)
4. 6 (5.8%) 0

5. 7 (6.8%) 6 (13.6%)
6. 4 (3.9%) 1 (2.3%)
7. 6 (5.8%) 0

8. 4 (3.9%) 3 (6.8%)
9. 9 (8.7%) 4 (9.1%)
10. 13 (12.6%) 13 (29.5%)
11. 1 (1,0%) 3 (6.8%)
12. 6 (5.8%) 2 (4.5%)
Outliers 8 (7.8%) 3 (6.8%)
Total 103 (100.0%) 44 (99.9%)

The Coefficient of Similarity for the two blocks is 135.3 which means they are
nearly 68% similar, however there are significant differences to support
Janzen’s hypothesis of two separate occupations. Whereas Janzen
emphasized the predominance of pseudo-scallop shell stamped vessels in the
east block (1968 80-82), the present, more detailed, analysis indicates that
significance lies in a number of factors: 1) the presence of more pseudo-scallop
shell stamped vessels in the east than west (Groups 1 to 4 in Table 15); 2) more
banked linear stamped vessels in the west than east (Group 5 of Table 15); 3)
the large percentage of vessels in the west block that are plain; and 4) the large
percentage of vessels in the west block that are decorated on the upper rim only
(Group 10 in Table 15). This group is reminiscent of the process of simplification
of rim decoration seen on later, transitional Pine river Ware of northern
Michigan (Holman 1984: 35). Seriation of the two blocks at Naomikong Point
with other Laurel assemblages (see Chapter 6) indicates that the east block
assemblage constitutes an earlier point in the series than the west.

Janzen’s major hypotheses are supported by this reanalysis. 1)There are
two occupation areas, one earlier than the other; 2) these may represent small
groups of people per occupation or at least there is no evidence to suggest that
Naomikong Point represents a single enormous occupation, and 3) some
specific designs especially on the miniature vessels suggest a widespread

exchange of cultural ideas.

163

 

NAOMIKONGPOINTSITE ’ ‘

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

g v“ I I4
,5 IIII
4 l [IV/I l
A 4: mm
I6 ”77”
w-
”..5-7 I l
III
I?
...- 8 |
NH .0. can
I I8 "H
"HT
000

 

 

 

 

I0' I

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ZI

 

III
m“
I \\\\\w
h\\\\

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mmflmmmmmmm
ID

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 11: Naomikong Point ceramic chart.
164

 

NAMKONG POINT SITE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

24 III ..
35 ll”
7m
25 mm
“is .. II
III. a .. I” I
fill/II 37
(l 2’ ll]
28 w"
39
" ll "'9 "ll
0.0.-'0'. U 30 .. u
a: 3' .
..n‘ ['0'
42 g
32 I"
43 “‘
33
u -34 fl

 

I;

 

Figure 11 (cont'd): Naomikong Point ceramic chart.

 

NAMWGPOINTSITE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

45
58 AA
“I "
46 \\~:
59
am 47
\.§=
38:. L-F/ "
4e _'
g: .
SI “5...;
00.0 49 B
T” I"
62 52:12:.
'7 7' so >
‘50 ' ’00.
.0 ‘
5|
"‘3’":
'5:
sz "‘35:
51.5331 - o 00
m: A
"/ s3
. ' '0 -56 ,
' 57
O. O
l

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 11 (cont’d): Naomikong Point ceramic chart.

166

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

if? . . 69
7o

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

7|
”,4 co 0 ’77-

“ I M-J
IW/l/ '" 73 -
W
g 74
‘ 000: 00.0 o o
”I

75
O C

 

 

 

fi

[:3 76
~77
o a

 

 

 

 

 

 

 

 

 

 

 

78

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mmmflmsmmmm-

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

‘ 79
9 3
(Ill/i 99'2

 

czzczmczziczzmczzczzczzmm

 

 

 

 

 

Figure 11 (cont’d): Naomikong Point ceramic chart.

167

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

WI 35 .
O O 0
122
[zissi 96
$3» - . .
$2 123 ”1”
oval: U 97 {5.3-};{1‘ "
” I 0000 b 2....
'35-'55 {£13
“a; 124 31:33::
/f
1111/ 93 W3
' z
'25 N ...-a-
99- W
I02 0 o
126 '
mm H 103 -
{C’C’f’ ,
....' . O . 00.0.
'~”'J ‘27 no“
IO4
0 ° ' ‘ 407
U m)
l28 v"
[tr/m U 108
my;
' ~ ' ° m)
\\ I .29 U
'09
a “2° :3,
II
I30 w
353's" '2]
'3‘ § 0 o o

 

 

 

 

 

Figure 11 (cont’d): Naomikong Point ceramic chart.

168

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Wmniil )HH
. . '1 7 6 0M
9‘. '32 I V ”PHI 0 c
"\j I I44 can
“I“? U .000 334 I I a U ....
”...“
I . 145 [m1
I35 I 111;)
l I46
I36 I I .-
____,J l I47 ..
1375 I I ”
To... l 148/Ilia
o 0 I39 I
m | I49
I40 | "'
... 0".
W l '50
WM I4I | 3'3: 0
{Ill/f4 g M I IS! 545;; ° .
I42 1 hi
I I52 51;???
I43 “ti-‘1
u u I I 153 SE? a .1

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 11 (cont’d): Naomikong Point ceramic chart.

169

 

 

 

 

 

154

 

 

 

 

 

 

 

WW“ 1 I H

 

ISS

 

 

 

'65
0'1

 

 

 

 

 

 

 

 

156 I

 

 

 

 

 

 

 

 

 

 

 

 

I57 |

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I59!

 

 

 

 

 

 

 

 

 

 

 

 

I60 I

 

 

 

 

 

 

 

 

 

 

 

 

 

161 I

 

 

 

 

 

 

 

 

 

 

 

 

l62 |

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

CZCZCZZCZZCZZCZZCZZCZCZC:

 

 

 

 

 

Figure 11 (cont’d): Naomikong Point ceramic chart.

170

 

Figure 12: Vessels from the Naomikong Point Site (from Janzen 1968): a)
unbanked incised constitute Group 12; b) banked incised constitute Group 11;
c) banked, linear stamped with lip decoration forming Group 4, d-h) banked,
undragged linear stamped forming group 7; i-k) decorated on the upper rim only
forming Group 10. (Photo courtesy of University of Michigan Museum of

Anthropology)

171

APPENDIX E

CLOUDMAN SITE (20CH6)

While burial mounds are common in the western Laurel area, occurring
mainly along the Boundary Waters between Minnesota and Ontario, they are
rare in the eastern area. The Cloudman Site on the west side of Drummond
Island in the Upper Peninsula of Michigan (figure 13) contains one mound that
may be Laurel (Branstner n.d.). The site at the mouth of the Potagannissing
River has been known since amateur archaeologists excavated the mound in
the early 20th century. They reportedly found skeletons with their feet toward the
center and a number of artifacts including ceramic sherds with geometric
designs (n.d.: 19). Later surface collecting and shovel testing at the site
recovered Middle and Late Woodland material along with some Historic
artifacts, and the mound was measured as 20 m in diameter and 1.7 m high
(n.d.: 21). Its association with Laurel is not confirmed (n.d.: 21).

Charles E. Cleland and Christine Branstner of Michigan State University
conducted excavations at the site in 1991, 1992, and 1994 to determine site
function and settlement-subsistence practices (Branstner n.d.: 2). The site is
rare in the eastern Upper Peninsula not only for the presence of a mound but
also for its multi-component nature that can be used to address the transition
between Middle and Late Woodland. At least 85 per cent of the excavated area
was in the Late Woodland part of the site. Little more than test pits were
excavated in the extremely large and intensively occupied Laurel area. The
ceramic assemblage includes 29 analyzable Laurel vessels used in this study.

The researchers excavated 102 square metres in 5-cm levels and found

that the deposits were 35 to 40 cm deep. Branstner notes that the site is

172

 

 

 

 

 

 

 

 

 

 

 

 

Figure 13: The Cloudman Site on Drummond Island, Michigan (from Branstner

n.d.)

173

horizontally stratified on three terraces, the upper terrace with Laurel, the middle
with Late Woodland, and the lowest with protohistoric material (Branstner n.d.:
23). She reports very little overlap among the three cultural zones. No
radiocarbon dates are available but Branstner (n.d.: 9) estimates a date of AD.
100-300 for the Laurel component based on general comparison of ceramics to
other dated sites.

The vessels are depicted in Figure 14, the master ceramic chart, and a
description of the clusters obtained in analysis is outlined in Table 17.

Groups 1, 2, and 7 (Figures 15 and 16) are pseudo-scallop shell
stamped vessels differing in banked/unbanked and presence/absence of
bosses and lip ticking. Group 3 is incised; Group 4 (Figure 17) is dentate
stamped; Group 5 is plain, and Group 6 (Figure 18) is linear stamped. The
largest cluster, the group of banked linear stamped vessels, represents 24.1 per
cent of the assemblage, however the three groups of pseudo-scallop shell
stamped vessels comprise 32 per cent of the assemblage. Four outliers all have
traits that preclude them from Laurel. Vessels 28 and 29, both with cord-marked
exteriors, may be transitional forms between Laurel and Late Woodland. Vessel
26 is Saugeen Ware, and Vessel 23 (Figure 16) is trailed, perhaps related to
Nokomis Phase vessels of north central Wisconsin or to Dane Incised of the

Green Bay area.

174

Table 17: Cloudman Vessels (f=29)

Group 1: (f=6; %=20.7) Vessels decorated with unbanked pseudo-scallop
shell stamping.

Subgroup 1A (f=2; %=6.9)

Profile - Everted

Interior - Plain

Lip - Transverse PSS

Exterior Upper Rim - Unbanked vertical PSS

Exterior Lower Rim/Neck - Unbanked vertical PSS and punctates
Vessel # - 12, 13

SUQgrQUQ 18 (1:4; 70:13.5)

Lip - Plain

Exterior Upper Film — Unbanked vertical or oblique PSS

Exterior Lower Rim/Neck - Unbanked horizontal, or horizontal and
oblique PSS

Vessel #- 15, 16, 17, 18

Group 2: (1:1, %=3.4) One vessel with banked vertical PSS, bosses and lip

ticking.

Profile - Everted less than 45 degrees

Interior - Plain

Lip - Plain

Exterior Rim-Lip Juncture - Vertical linear stamped tick marks
Exterior Upper Rim - Banked vertical PSS and bosses
Exterior Lower Rim/Neck - Banked vertical PSS

Vessel # - 10

Group 3: (f=2; %=6.9) Incised vessels at the cluster level of subgroup.

Profile - everted less than 45 degrees

Interior - Plain or oblique incising

Up-Pmm

Exterior Upper Film - Vertical or oblique incising

Exterior Lower Film/Neck - Vertical or vertical and criss-crossed
incising

Vessel # - 8, 27

Group 4: (f=5; °/o=17.2) Unbanked dentate and linear stamped vessels.

ro 4A 1:2' °/o=6.

Profile - Inverted and everted less than 45 degrees

175

Table 17 (Cont’d)

Interior - Vertical dentate stamps

Lip - Transverse dentate stamps

Exterior Upper Rim - Oblique dentate stamps (vessel 20 is banked)
Exterior Lower Rim/Neck - Punctates (Vessel 20 also has banked
oblique dentate stamps)

Vessel # - 19, 20

Subgrouo 4B (f=1; °/o=3.4)
Exterior Upper Rim - Unbanked horizontal dentate stamps
Exterior Lower Rim/Neck - Unbanked horizontal dentate stamps
and punctates
Vessel # - 22

Subgroup 4C (f=2; %=6.9)
Interior - Plain
Lip - Plain
Exterior Upper Rim - Unbanked oblique dentate or linear stamps
Exterior Lower Rim/Neck - punctates (Vessel 21 also has banked
oblique dentate stamps
Vessel # - 9, 21

Group 5: (f=2; %=6.9) Plain vessels.

Profile - Everted less than 45 degrees
Interior - Plain

Lip - Plain

Exterior Upper Rim - Plain

Exterior Lower Rim/Neck - Plain
Vessel # - 24, 25

Group 6: (f=7; %=24.1) Banked linear stamped vessels.

Profile - Everted less than 45 degrees or inverted

Interior - Plain (Vessel 6 has vertical linear stamps)

Lip - Plain or Linear stamped

Exterior Upper Rim - Banked vertical/oblique linear stamps,
dragged or undragged

Exterior Lower Rim/Neck - Banked vertical/oblique linear stamps,
dragged or undragged (Vessel 7 also has bosses)

Vessel#- 1, 2, 3, 4, 5, 6, 7

176

Table 17 (Cont’d)

Group 7: (i=2, %=6.9) Banked pseudo-scallop shell stamped vessels.

Profile - Inverted or Everted less than 45 degrees
Interior - Plain

Lip - Plain or transverse PSS

Exterior Upper Rim - Banked oblique PSS
Exterior Lower Rim/Neck - Banked oblique PSS
Vessel # - 11, 14

Outliers: (f=4; %=13.8) Four vessels not considered to be Laurel were
included in the analysis. All proved to be outliers of the Laurel assemblage.
Vessels 28 and 29 have cord-marked surfaces and decoration not characteristic
of Laurel - Vessel 28 has punctate pairs and long criss-crossed incising and
Vessel 29 has rocker stamping. Vessel 23 has trailing and Vessel 26 is
Saugeen Ware with large criss-cross dentate stamping.

 

177

 

W... Hill

 

 

I”:

 

 

 

an

 

II II

nun
”film
(Ill/Ir
hum

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

W

 

 

 

 

 

 

 

 

 

 

 

..IO

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Zl

 

 

 

 

 

 

 

a .
U I
3 I a
l
.222 U a I u
'IIOI‘ O 0
2.2:;
u 5 I u o e a
g m 6 I U o o o
o o 7 I
I 0
[MB l\\\\ 9 55733551 W
I/ IS -
a 9 x'w. a H:
15.1.?
LII}?

 

 

 

Figure 14: Cloudman ceramic chart.

178

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

=33? 5 22 |
:7. E
lllil U 23 | G
‘g

24
jag a 26 l U
m u 27 I (1
0.00.9.
a ...28 l U
“:3. as .
h‘ \“ \‘\“\“
¢~3 ~‘ *1
Exijfl‘

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 14 (cont’d): Cloudman ceramic chart.

179

 

 

 

 

 

 

 

 

 

 

 

 

Figure 15: Vessel from the Cloudman Site: Group 1, unbanked pseudo-scallop

shell stamped. (from Branstner n.d.).

 

Figure 16: Vessels from the Cloudman Site: upper and lower left, Group 7,
pseudo-scallop shell stamped; upper right, Group 2, pseudo-scallop shell

stamped; lower right, a trailed outlier. (from Branstner n.d.).

180

 

TH .91 interior

Figure 17: Vessel from the Cloudman Site: Group 4, dentate stamped (from

Branstner n.d.).

 

   

gum .1/73

Figure 18: Vessel from the Cloudman Site: Group 6, banked linear stamped

(from Branstner n.d.).

181

APPENDIX F

GYFTAKIS SITE (20MK51)

The Straits of Mackinac separating the Upper and lower Peninsulas of
Michigan was a nodal point in the Historic Period where both French and British
troops located military posts, but it was important long before then. The
Juntunen Site, the type site for the Late Woodland phase of that name, is
located on Bois Blanc Island in the Straits. But even long before the Anishinabe
occupied that site, Laurel people occupied sites on both the south and north
shores of the Straits. The Gyftakis Site is a Middle Woodland Cemetery in the
modern village of St. Ignace (Figure 2) located by John Franzen and Mike
Manfredo during a survey conducted for the Mackinac State Park Commission
and the Michilimackinac History Society. James E. Fitting and Timothy Smith
conducted excavations there in 1973 for the Michigan History Division,
Michigan Department of State (Fitting n. d.: 8).

Fitting undertook the excavations as part of a salvage operation during
urban construction in the village. He wanted to recover settlement pattern data
of the Middle Woodland component similar to that discovered by David Brose at
Summer Island (Fitting n.d.:10). He especially wanted house patterns but what
he got was a surprise. The major feature at Gyltakis was a large bun'al pit with
associated Laurel Ware (n.d.: 18). In his report to the History Division, Fitting
mentioned that the site had been leveled previous to his work (n.d.: 11); what he
may have discovered was a mound similar to that at the Cloudman Site on
Drummond Island. There was at least one other mound in the in the Straits
area, at the Late Woodland Juntunen Site (McPherron 1967: 201). Another

mound has also been reported in Sault Ste. Marie, Michigan (Cleland, personal

182

communication). An unpublished report of Fitting’s work is on file in the
Department of Anthropology, Michigan state University.

In 1973, Fitting and Smith added to the 5x10—foot unit excavated by
Franzen and Manfedo in 1972 by excavating a large block of units around it to
bring the total area to 1,600 square feet. He used 3-inch arbitrary levels in the
apparently mixed, loose sand matrix. A few fur trade items were located, most
likely from a large site adjacent to Gyftakis, the Marquette Mission, one of the
largest historic sites in the Great Lakes area (n.d.: 10). Most of the Laurel
ceramics were concentrated in and around a large pit, Feature 15, covering
most of two 10x10-foot squares. The pit outline was clearly delineated by
charred organic material and charcoal, and Fitting concluded it had been lined
with grasses, leaves, and twigs and burned as a ceremonial act (n.d.: 21). He
found secondary burials of at least seven people and grave goods including a
smashed pot, two complete pots, one inside the other, a number of bird bills
(genus unknown), red ochre nodules, and a large sandstone “lump” (n.d.: 18-
21). The fill of the pit contained other Laurel sherds. Beside the pit, and
associated with it, was Feature 22, a large fireplace with many fire-cracked
rocks and a crushed pot similar to those in Feature 15. A post mold extending
nine inches into the living floor was associated with the fireplace. A radiocarbon
date of AD. 170180 was obtained from Feature 22 (n.d.: 23).

All ceramics were Middle Woodland except for three intrusive shell-
tempered sherds probably associated with the nearby historic Tionontate
village (n.d.: 25-26). Fitting conducted chi-square analysis in search of
significant attribute groups but he discovered no clustering (n.d.: 53). He also
concluded that attribute distribution was random stratigraphically and spatially.

l reanalyzed the assemblage to test the hypothesis of randomness. The

vessels are depicted in Figure 19, the master ceramic chart for the 61

183

Table 18: Gyftakis Vessels (f=61)

Group 1: (1:12; %=19.7) Vessels with banked and dragged dentate stamps.
This is a strong group that could be a subgroup related to Group 2.

Profile - All types

Interior - Plain

Lip - Plain or transverse or oblique dentate or incised

Exterior Lip-Rim Juncture - 3 vessels have vertical linear stamp tick
marks

Exterior Upper Rim - Banked, dragged or undragged oblique
dentate stamps (3 vessels are not banked)

Exterior Lower Rim/Neck - Oblique or vertical dragged dentate
stamps (4 vessels also have punctates)

Vessel # -1, 2,3, 5,8, 9, 10, 17, 25,35, 36,37

Group 2: (f=6; %=9.8) Vessels with banked and dragged linear stamps.
§UDQFOUQ 2A (f=3; °/o=4.2))

Profile - Inverted

Interior - Plain

Lip - Plain or incised

Exterior Upper Rim - Vertical unbanked, undragged PSS, or
incised lines in criss—cross patterns or oblique lines

Exterior Lower Rim/Neck - Banked, dragged or undragged linear
stamps

Vessel # - 40, 41, 49

S f‘ 28 f: ' °/o=4.
Profile - More than 45 degrees everted or straight
Interior - Plain
Lip - Plain or tick marks in incising or linear stamps
Exterior Upper Lip-Rim Juncture - Vessel 46 has linear stamp tick
marks
Exterior Upper Rim - Banked and dragged linear stamps
Exterior Lower Rim/Neck - Banked linear stamps and punctates
(Vessel 53 has oblique lines of oblique linear stamps)
Vessel # - 46, 53, 54

Group 3: (f=10; %=16.4) Pseudo-scallop shell stamped vessels.
rou A f=2' %- 3
Profile - Everted less than 45 degrees
Interior - Linear stamped
Lip - Plain

184

Table 18 (cont’d).

Exterior Upper Rim - Banked, dragged PSS
Exterior Lower Rim/Neck - Banked, dragged PSS
Vessel # - 13, 14

 

Subgroup 3B (f=4; %=66)
Profile - Straight
Interior - Plain
Lip - Plain -
Exterior Upper Rim - Unbanked oblique PSS
Exterior Lower Rim/Neck - Unbanked diagonal or horizontal PSS

Vessel # - 16, 19, 39, 58

W529.)
Profile - Everted
Exterior Lower Rim/Neck - Unbanked horizontal PSS (Vessel 48
has oblique PSS
Vessel # - 4, 32, 48

S i‘OU D f=‘I‘ %=1.
Interior - Oblique PSS
Lip - Plain
Exterior Upper Rim - Vertical PSS over a horizontal incised line
Exterior Lower Rim/Neck - Oblique PSS
Vessel # - 38

Group 4: (1:11; %=18.0) Vessels with banked, undragged designs with
punctates.

@9932» 4A (f=6; 70:98)
Profile - Everted (Vessel 47 is more than 45 degrees)
Interior - Plain
Up-Pmm
Exterior Lip-Rim Juncture - Vessel 24 has linear stamp tick marks
Exterior Upper Rim - Undragged, banked, oblique linear, dentate,
PSS stamps
Exterior Lower Rim/Neck - Undragged, banked, oblique stamps
same as upper rim, with punctates
Vessel # -24, 47, 55, 56, 57, 6O

§QQQFOQQ 4B (f=3; 0/o=4.9)
Profile - Straight or inverted
Exterior Upper Rim - Undragged, banked oblique dentate stamps
(Vessel 27 also has punctates)
Exterior Lower Rim/Neck - Undragged, banked oblique dentate

185

Table 18 (Cont’d).
"""""" QERB§EEJBGE&SIE§_""“""""'""“""""
Vessel # - 27, 44, 62

Outliers (f=2; %=3.3): Vessel 43 has banked linear stamps and no
punctates. Vessel 59 has PSS rim-lip ticking and vertical columns of
oblique PSS stamps.

Group 5: (f=5; %=8.2) A very varied group of vessels with simple, unbanked
vertical stamps.

Profile - Everted less than 45 degrees

Interior - Plain

Lip - Plain

Exterior Upper Rim - Unbanked vertical rows in linear, PSS and
dentate stamps (Vessel 29 is plain)

Exterior Lower Rim/Neck - Same as upper rim (Vessel 29 has only
punctates)

Vessel # - 22, 29, 30, 33, 34

Group 6: (i=3; °/o=4.9) Vessels with unbanked oblique dentate stamps
combined with either punctates or pseudo-scallop shell stamps.

Profile - Straight or everted less than 45 degrees

Interior: Plain

Lip - Plain

Exterior Upper Rim - Unbanked oblique dentate stamps
Exterior Lower Rim/Neck - 2 vessels have only punctates and
Vessel 18 has horizontal PSS

Vessel#-11, 12, 18

Group 7: (f=2; °/o=3.3) Plain vessels

Profile - Inverted and everted less than 45 degrees

Interior - Plain

Lip - Plain

Exterior Rim-Lip Juncture - Vessel 26 has linear stamp tick marks
Exterior Upper Rim - Plain

Exterior Lower Rim/Neck - Plain

Vessel # — 26, 61

Group 8: (1:3; °/o=4.9) Vessels with unbanked linear stamps.

Profile - Straight
Interior - Plain

186

Table 18 (cont’d).

Up-Pmm

Exterior Upper Rim - Vertical linear stamps

Exterior Lower Rim/Neck - Vertical linear stamps (Vessel 42 also
has punctates)

Vessel # - 28, 42, 45

Group 9: (f=2; %=3.3) Dentate stamped and punctated vessels.

Profile - Straight

Interior - Plain

Lip - Plain

Exterior Upper Rim - Vertical or oblique, unbanked dentate stamps
Exterior Lower Rim/Neck - horizontal dentate stamps and
punctates.

Vessel # - 23, 31

Group 10: (1:2; %=3.3) Vessels with oblique dentate stamps on the upper rim

only.

Profile - Straight

Interior - Plain or vertical dentate stamps

Lip - Plain

Exterior Upper Rim - One row of oblique dentate stamps
Exterior Lower Rim/Neck - Plain

Vessel # - 15, 21

Outllers: (f=5; °/o=8.2) Five vessels not related to any cluster.

Vessel 6 is zoned with an area of oblique columns of linear stamps and an area
of right-angle dentate stamps.

Vessel 7 has vertical exterior combing but no decoration.

Vessel 50 has oblique columns of oblique dentate stamps.

Vessel 63 has horizontal dragged lines of incising on the upper rim and trailed
pendant triangles on the lower rim.

Vessel 64 has vertical columns of horizontal linear stamps.

 

187

analyzable vessels. The clustered vessels are described in Table 18. I did not
include Vessels 20, 51, and 52 in the clustering on the assumption that, as they
are miniature vessels, their function may have differed from the rest of the pots.

Groups 1 and 2 form a super-cluster of banked and dragged stamped
vessels, with dentate stamped in Group 1 and linear stamped in Group 2
(Figures 20 and 210). Group 3 is composed of pseudo-scallop shell stamped
vessels (Figure 22b). Group 4 has banked, undragged designs with punctates
(Figures 21a, 22c and 23). A super-cluster of Groups 5 to 10 is unbanked, and
each group is small, comprising only a few vessels each (Figures 21d and 22a).
The clustering partially supports Fitting’s hypothesis in that the 61 vessels
analyzed at Gyftakis form 10 groups: Gyftakis is one of the most varied
assemblages in the study. Contrast this to Whitefish Island where 80 vessels
comprised only 7 groups, and half of the vessels were in one group.

Concerning distribution, as Fitting noted, there is no stratigraphic
clustering; clusters include vessels from various levels. Also, the 19 vessels
from Feature 15 (see Figures 20, 21a, 22c and 23) are in every cluster except
Groups 5, 6, 9 and 10 all of which are small, subject to sample size problems.
The assemblage does not split into two occupations as at Naomikong Point.

Of the outliers, all are variations of Laurel designs except Vessel 7, a
combed pot with no clear association with any other ware, and Vessel 63, with
unusual dragged incising and trailed pendant triangles similar to vessels at

Naomikong Point and Squirrel Dam.

188

 

 

swam ‘ . l I l I

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 19: Gyftakis ceramic chart.

189

 

 

 

 

 

 

 

, 22 l
0' LL

 

 

 

 

 

 

 

 

23 l

 

I" ,

 

 

 

 

 

 

 

 

 

 

 

 

 

24 I

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

26 I

 

 

 

 

 

 

 

 

 

 

 

 

090 27

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

IE

 

 

| 4|

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

30 I III

 

 

 

 

 

 

 

 

' Ig
9:2: 2:: a: a: [:22 _M 5:: ::

 

 

 

 

 

llllll

 

 

 

 

 

 

 

 

 

 

 

24 5.1
czicziczzzczczztzzczczczczz

 

 

 

 

 

 

 

 

 

 

 

 

Figure 19 (cont’d): Gyftakis ceramic chart.

190

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

GYFTAKIS SITE
7000 45 $2;
a o o . gig”
O O O I 55
”7” 46 i”
2.2". :{fl 4
56
47 3:3: :I
II!!!
a o o 9 [LI 0 o
' , 57
Will! I ......
um 58
‘9 CI
59 I
:12 g 50 z“
r. 3"}. 31 933222
2,4. 0 - -- u
l 60
5|
6|
52 £22333}.
0 O O 0 O 0 1:90.37:
“AL ....
--- 53 W I ”:I
63
54 (I

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

64

 

 

 

 

 

 

 

 

 

Figure 19 (cont’d): Gyftakis ceramic chart.

191

 

 

Figure 20: Vessel from the Gyftakis Site: banked and dragged linear stamped

vessel from Group 2. (Photo courtesy of the Michigan Historical Center)

192

 

Figure 21: Vessels from the Gyftakis Site: a) banked, undragged linear stamped
from Group 4; b) linear stamped outlier; c) criss-cross incised over dragged
linear stamped from Group 2; d) dentate vertical columns from Group 9. (Photo

courtesy of the Michigan Historical Center)

193

III-

Figure 22: Vessels from the Gyftakis Site: a) dentate stamps on the upper rim
only from Group 10; b) unbanked pseudo-scallop shell stamped from Group 3;
c) vertical columns of oblique pseudo-scallop shell stamps from Group 4. (Photo

courtesy of the Michigan Historical Center)

194

 

Figure 23: Vessel from the Gyftakis Site: banked, undragged pseudo-scallop
shell stamped from Group 4. (Photo courtesy of the Michigan Historical Center)

195

APPENDIX G

PORTAGE SITE (20EM22)

While researchers have often noted that the spatial distribution of Middle
Woodland materials in the northern Lake Michigan area is consistent with that
of Late Woodland materials, suggesting in situ development from one period to
the other, few sites have actually produced ceramics with transitional
characteristics defining a Middle-to-Late Woodland sequence. William Lovis of
Michigan State University set out in 1974 and 1975 to examine the Portage Site
near Petoskey, Michigan (Figure 2), with the expectation that this site would
provide the necessary ties. Results of the ceramic analysis were completed by
Lovis, Rajnovich and Bartley (1998).

Lovis uncovered two components separated by sterile aeolian sand. The
31 excavated units encompassed 775 square feet, about 150 square feet of
which were stratified. Lovis, Rajnovich and Bartley conducted cluster analysis of
the ceramic assemblage to define the characteristics of the upper and lower
strata, and examine the relationships between the two components (Lovis,
Rajnovich and Bartley 1998). We discovered four main clusters, two mainly from
the upper stratum and two mainly in the lower. The upper stratum contains Late
Woodland Pine River Ware and a few Mackinac-related vessels, plus a group of
vessels that are transitional, displaying Late Woodland cord-marked exteriors
but with some Middle Woodland traits. A radiocarbon date of 100 B.C.4_-80 from
the upper stratum was rejected by Lovis because of its relationship to the Late
Woodland ceramics and its position in relation to two later radiocarbon dates
from the lower stratum (Lovis, Rajnovich and Bartley 1998). The vast majority of

ceramics from the lower component are Middle Woodland, mostly Laurel. Two

196

overlapping dates of AD. 120:120 and AD. 3301150 from charcoal from a
small pit or post mold, and wood or pitch from a hearth were accepted as
reasonable dates for the assemblage (Lovis, Rajnovich and Bartley 1998).
While the upper and lower strata represent mainly Late and Middle Woodland
respectively, our cluster analysis found that the ceramics traits of the upper and
lower strata form overlapping polythetic sets, indicating at least a partial
continuum from Middle to Late Woodland in the area.

The present analysis concerns only those vessels from the lower levels,
named Groups 3 and 4 in the previous publication of ceramics from the site.
They are shown on the master ceramic chart at Figure 24. Examples of vessels
are shown in Figures 25 and 26. Table 19 provides a description of Groups 3
and 4.

Vessels 70 to 74 are miniature vessels and are not included in the
cluster analysis although they do have counterparts among the vessels used in
the analysis. Vessels 38 and 40-45 are outliers in the analysis.

Some of the subclusters are recognizably Laurel types previously
defined. Subgroup 3A is Laurel Incised; 3B is Laurel Plain var. Punctated; 4A is
Laurel Oblique var. Dentate Stamped, and 4B is Laurel Oblique var. Linear
Stamped.

However, Subgroup 43 contains stamps that are unusual in a Laurel
assemblage, linear stamps done with a cord-wrapped stick (Vessels 65, 66 and
68). These may be related to Upper Peninsula Looped-end Cord stamped at
Summer Island in the Upper Peninsula (Brose 1970) and North Bay Corded
Stamped dated to AD. 160 i100 at Porte des Morts in Wisconsin (Mason
1967). In addition, Group 30 contains banked pseudo-scallop shell stamped
vessels common in Laurel, assemblages but it also contains single-cord

impressed vessels perhaps related to vessels at Summer Island (Brose 1970).

197

Table 19: Portage Vessels (1:39)

Group 3: (f=21; 96:52.9)

1' 3A 1:1' °/o=2.9

Profile - Inverted

Interior - Plain

Lip - Plain

Exterior Upper Rim - Vertical incised lines
Exterior Lower Rim/Neck - Punctates

Vessel # - 37 (closely related to Subgroup 3B)

Wm
Profile - Inverted or everted less than 45 degrees
Interior - Plain (Vessel 34 has vertical cord-wrapped stick stamps)
Lip - Plain (Vessel 33 has transverse punctates)
Exterior Lip-Rim Juncture - Vessels 30 and 35 have tick marks
Exterior Upper Rim - Plain
Exterior Lower Rim/Neck - Punctates
Vessel # - 29 (2 vessels), 30, 31, 32 (3 vessels), 33, 34, 35.

WW
Profile - Everted more or less than 45 degrees
Interior - Plain
Lip - Plain (Vessel 56 has linear stamps)
Exterior Upper Rim - vertical or oblique designs using various
tools
Exterior Lower Rim/Neck - horizontal single-cord impressed or
banked oblique pseudo-scallop shell stamped withylwithout
punctates
Vessel # - 51, 53-59

Subgroup 3D (f=2; °/o=5.2)
Profile - Everted less than 45 degrees

Interior - Vessel 52 has vertical cord-wrapped stick stamps

Lip - Plain

Exterior Lip/Rim Juncture - Vessel 39 has tick marks

Exterior Upper Rim - Vertical cord-wrapped stick stamps
Exterior Lower Rim/Neck - Punctates with single-cord impressed
or cord-wrapped stick stamps

Vessel # - 39, 52

198

Table 19 cont’d

Group 4: (f=16; 96:47.0)

u r u 4A 1:4' %:11.
Profile - Straight or everted less than 45 degrees

Interior - Plain

Lip - Plain

Exterior Upper Rim - Banked vertical or oblique dentate stamps
Exterior Lower Rim/Neck - banked vertical or oblique dentate
stamps with punctates

Vessel # - 4649

Su r 43 f=12° °/o= .
Profile - Straight or everted less than 45 degrees

Interior - Plain (Vessel 60 has vertical rows of linear stamps)

Lip - Plain (Vessel 62 is incised)

Exterior Upper Rim - Banked stamped with linear or cord-wrapped
stick stamps (Vessel 50 has pseudo-scallop shell stamps)

Exterior Lower Rim/Neck - Banked stamped with linear or cord-
wrapped stick stamps (Vessels 36, 60 and 68 have push-pull
banks

Vessel # - 36, 50, 60-69

M
Vessels 38 and 42 have incised exterior decoration, plain lips and
plain interiors.

199

The impressions there are vertically oriented whereas they are horizontal at the
Portage Site. Subgroup 3D has cord-wrapped stick designs perhaps more
common on North Bay Ware than Laurel, and Vessel 39 has added trailing, a
trait more common on Dane Incised or Lake Nokomis Trailed of northern
Wisconsin.

In summary, while the previous study indicated a connection between
the Portage Site's Middle and Late Woodland assemblages, this study notes a
connection between the Middle Woodland vessels of the Portage Site and
vessels not only of Laurel sites to the north but also of Middle Woodland sites

across Lake Michigan to the west.

200

 

PORIAGE SITE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2 | t . W
F HI I I
I l2
1‘ ‘1 i a:
/ I: I I3
4 I r
f: 00 0

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

film
I?

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I IS
F 7 I
III c
I i7
r II!” 3 I
IIII C

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I.“

 

 

 

 

 

 

 

 

 

::IIO I C

 

 

 

 

 

 

 

 

 

 

 

mmmmmmmmmm
Ln
a
mi: : "“
\n e
mammmmmmmm

 

 

 

 

 

 

 

 

 

Figure 24: Portage ceramic chart.

201

PORTAGE SITE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

21
st. 3'
E if}?! 22
' l 32
W [:123
33
i 5: SI 24
III a...
C 34
m I"! 25 '7
- 35

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

i
1:?!
S q” 27
A 37

0)!)
[:1 w
I 39 my

D” .

 

 

~ Ci

 

 

 

 

 

 

9%
1R
czzczzmczczczzmmmm
'5’»

30

 

 

 

J

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

czzczzczzmczzaczczzflczzczz

 

Figure 24 (cont’d): Portage ceramic chart.

202

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

PORTAGE SITE

 

 

I, "

 

 

 

 

 

"

I
I
ii:

 

l

 

SI

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

In
III/II

 

 

 

53 m"“{‘"

m

 

 

 

 

 

 

 

 

 

 

 

54 B

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mud
0001]; i

 

“3

 

 

”III/I
'0 ‘l I I

_ :53...
59 "'

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mQCZSDfl,::c:CZC2C2

iEmT—I

 

Figure 24 (cont’d): Portage ceramic chart.

203

 

 

PORTAGE SITE \ )

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

E

 

 

 

 

 

 

 

 

 

 

 

 

7|

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

rrrrr

 

 

 

 

 

 

 

 

 

 

 

 

 

 

\
§_.
flaming:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 24 (cont’d): Portage ceramic chart.

204

 

.Ammmw >oEmm new :o_>oc_mm .254 E95 mos—2:209
5:8 5:5 commoaé 28 20% B mcozmcfieoo 9m 29 Eaton new :3: 9 5:50 39 288 683 9a 26. 0:88
or: B 8:: :9 05 can 39 a9 9: yo mm: 9: £865 .39 a2. .95 08:01 9: 8 m 955 E0: m_mmmm> ”mm 059;

 

205

381298 95 6305mm .m_>o._ 59:
.8953 59:. .959 Eouon 0cm 063E 68:56 29:3 .39 8% gm omwtoa m5 8 v 9.90 So: m_omwm> ”mm 939”.

 

206

APPENDIX H

FISK SITE (Dle-i)

The Fisk Site, on a portage of a tributary of the Winnipeg River just north
of Kenora, Ontario (Figure 27), was the focus of an early survey of northwestern
Ontario by the fledgling regional archaeology system of the Ontario Ministry of
Culture and Recreation. What started as a simple rescue excavation of a minor
site suffering disturbance by extensive portage traffic ended in the gathering of
significant data on early Laurel.

In 1975 and 1976, Regional Archaeologist C. S. “Paddy" Reid excavated
a block of 25 one-metre units (Figure 28) in 2-cm and 3-cm levels in efforts to
expose spatial patterning as well as stratigraphy of this multi-component site. It
contained material from Late Woodland Selkirk and Blackduck complexes over,
and mixed with, Middle Woodland Laurel material. The main feature consisted
of spatial patterning of post molds and rocks, apparently the outlines Of an oval
house (Figure 29) approximately 7 x 4.5 m with a central hearth. Charcoal from
the hearth produced a date of AD. 501115, one of the earliest assays for Laurel
(Rajnovich, Reid and Shay 1982).

While the Laurel assemblage is small, only 23 vessels, it has great value
as an example of early Laurel pottery. The vessels are depicted in Figure 30,
the master ceramic chart. Examples of some are shown in Figure 31 and others
are depicted in Rajnovich, Reid and Shay (1982). The clusters are described in
Table 20.

The assemblage consists of a simple set of four groups with one outlier.
Group 1 is banked dragged pseudo-scallop shell stamped; Group 2 is plain with

punctates/bosses; Group 3 is unbanked pseudo-scallop shell stamped, and

207

Group 4 is unbanked linear stamped/punctated. The outlier is the only dentate
stamped vessel. The complexity of designs seen, for example, at the Gyftakis

and Portage sites, is missing in this early western Laurel assemblage.

208

Table 20: Fisk Vessels (f=2:!)

 

Group 1: (f=8; °/0=34.8) Banked, dragged pseudo-scallop shell or linear
stamped vessels.

S r 1A f=5' °/o=21.7
Profile - Everted less than 45 degrees
Interior - Plain
Lip - Plain
Exterior Upper Rim - Banked, dragged, oblique PSS
Exterior Lower Rim/Neck - Banked, dragged, oblique PSS
Vessel # - 5 to 9

Su ro 18 f: '%=1 .
Lip - Transverse decoration (Vessels 13 and 14 have lip notches;
Vessel 12 has incised lines)
Exterior Upper Rim - Banked, dragged, oblique linear stamps
Exterior Lower Rim/Neck - Bank, dragged, oblique linear stamps
Vessel # -12, 13, 14

Group 2: (f=5; %=21.7) Plain vessels with punctates/bosses on the exterior
upper rim.

Subgroup 2A (f=4; %=174)
Profile - Everted less than 45 degrees
Interior - Plain
Lip - Plain
Exterior Upper Rim - One row of punctates/bosses
Exterior Lower Rim/Neck - Plain
Vessel #- 16, 17, 21, 22

Sumrgup 28 (f=1; °/o=4.§I

Exterior Upper Rim - Two rows of punctates/bosses
Vessel # - 20

Group 3: (f=4; %=17.4) Vessels with unbanked vertical or oblique pseudo-
scallop shell stamps over punctates and horizontal pseudo-scallop shell
stamps. Within this group is only one cluster that is as strong as a sub-

subgroup.

Profile - Everted less than 45 degrees

Interior - Plain

Lip - Plain

Exterior Upper Rim - Unbanked oblique or vertical P88

209

Table 20 (Cont’d).

 

Exterior Lower Rim/Neck - Unbanked horizontal PSS with

punctates/bosses.
Vessel #- 1, 2, 3, 4

Group 4: (f=5; %=21.7) A group of vessels with unbanked oblique linear
stamps or punctates on the exterior upper rim.

Subgroup 4A (f=3; 70:13.9)
Profile - Straight

Interior - Plain

Lip - Plain

Exterior Upper Rim - Oblique punctates (Vessel 15 not oblique)
Exterior Lower Rim/Neck - Plain

Vessel #- 15, 18, 19

SQQQI‘QUQ 4B (f=2; °/o=§.7)
Exterior Upper Rim - Unbanked oblique linear stamps

Exterior Lower Rim/Neck - Punctates (Vessel 10 also has a row of
oblique linear stamps)

Vessel#-10, 11

Outller: (f=1, %=4.3) Vessel 23 has banked, undragged dentate stamps on the
exterior upper and lower rims, with punctates on the lower rim.

 

210

 

 

 

 

 

._-.: .. .51.: 3 E".
i} Lake

 

 

 

 

..
.90

 

 

 

 

 

 

Dallas '

 

The -.-‘ (N

 

 

Figure 27: The Fisk Site on a tributary of the Winnipeg River north of Kenora,
Ontario (from Rajnovich, Reid and Shay 1982).

211

 

 

 

 

 

 

 

 

 

 

 

 

    

 

Figure
THE FISK SITE, Dth-1

CI excavation unit

 

~ '43 9 9o ‘
63 rock with permanent datum 0:35;} D 1'” {:13 ,- ;.'
Gt. _, .' '
shovel test .; ‘ H
O ' § /
o r 2 a 4 s / 0"
metres E Riven ‘r/ , i
l
as metre contour interval / / ;
I3, 2. 1! WI 3' ’ é

 

Figure 28: Excavation units at the F-"Isk Site (from Rajnovich, Reid and Shay
1982)

212

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 29: Features at the Fisk Site (from Rajnovich, Reid and Shay 1982).
Horizontally lined features are pits filled with hard-packed black loam.
Diagonally hatched features are pits with black loam, ash, burnt bone and red
ochre flecks. Open circles are rocks and X-filled circles are fire cracked rocks.

Black dots arepoSt molds.

213

 

.. n

 

 

 

‘ | I lIll/”’l
:3 000
' I _|4 ll/yl/I

 

 

 

 

 

 

 

 

 

[III I I .5 EU

 

 

 

 

 

 

 

 

 

 

gap] ’ 2 Snug”

 

 

 

 

 

 

 

 

 

 

UB4 ' ' 2 slag“

 

 

 

 

lam“ ' . 20:05

 

 

 

 

 

 

2M I0. I eerE
"U + 2 -

 

 

 

 

 

 

IUI'l . '

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

IlEl‘i ' - - -

 

 

Figure 30: Fisk ceramic chart.

214

 

 

 

Figure 31: Vesselsfrom the Fisk Site: a—d) plain with punctates/bosses from

Group 2; e-g) unbanked pseudo-scallop shell stamped from Group 3.

215

APPENDIX I

LONG SAULT SITE (Dde-i)

The essence of spirituality for the Rainy River First Nations of the Ojibway
begins at Kay-Nah-Chi-Wah-Nung, the place of the long rapids (Bruyere 1998).
It is a group of sites known in the archaeological literature as Manitou Mounds,
the largest existing group of burial mounds and related habitation areas in
Canada. It stretches nearly 4 km along the banks of the Rainy River, and
contains 17 mounds and 30 habitations areas from all periods beginning with
Archaic (Figure 32). It is also one of the few sites producing intrusive Hopewell-
related artifacts in Laurel contexts. In 1970, the Canadian government declared
it to be of national historic significance. Since that date, the Rainy River First
Nations have worked to develop the site as a national historic park and, in
1999, opened the park and interpretive center to the public. A major component
in the development of the site has been archaeological investigation of
habitation areas, and training of Ojibway students in archaeological methods.

The Rainy River has several famous Laurel mound sites including
Mclfinstry Mounds and the Smith Site, or Grand Mound, on the American side
of the river (Stoltman 1973), and Hungry Hall and Pithers Point on the
Canadian side (Kenyon 1986). For the Laurel people, the Rainy River was
clearly a significant link in the chain of lakes and rivers forming the Boundary
Waters between Lake Superior and the central interior of the continent.

In 1973, Manitou Rapids Reserve of the Rainy River First Nations
requested a survey and test excavation of Manitou Mounds in the face of
destructive gravel operations in the area. David Arthurs of the Ontario Ministry of

Natural Resources conducted both survey and excavations between 1973 and

216

1975. As part of his investigation, Arthurs excavated the portion of Manitou
Mounds known as the Long Sault Site (Figure 33). He produced his results as
his master’s thesis at the University of Manitoba and as a report for the Ontario
Ministry of Citizenship and Culture (Arthurs 1986).

Previous excavations at Manitou Mounds focused on the Armstrong
Mound (Kenyon 1970, 1986), a Laurel mound containing a log crib structure
and bundle burials. Fill of the mound included early Laurel ceramics and two
Ohio pipestone pipes, one a frog effigy, the other a classic Hopewell platform
pipe. The mound has been radiocarbon dated to AD. 160175 (William A. Ross,
personal communication).

In choosing the Long Sault portion of Manitou Mounds, Arthurs focused
on a habitation area where stratigraphy was unknown (Arthurs 1986). His
project produced a basic culture history of the site necessary for the
development of interpretive displays planned by the First Nations. It was also
the first training project designed to develop archaeological skills of the First
Nations students slated to oversee future archaeology at Manitou Mounds.

The Long Sault Site is deeply stratified with Archaic to historic materials.
Arthurs excavated 14 one-metre units in 3-cm levels (Arthurs 1986: 28), and
found the Laurel stratum to be the richest. He estimated that three Laurel
occupations were present but so close in time as to be treated as one, and he
estimated its date to be middle Laurel McKinstry phase (1986: 133).

l reanalyzed the assemblage from the Long Sault Site to test Arthurs'
estimate of a mid-Laurel date. The ceramics are shown in Figure 34, Arthurs'
master ceramic chart (Arthurs 1986). The clusters are described in Table 21,
and samples of rims are depicted in Figure 35, an illustration from Arthurs
(1982).

The 79 vessels form only six clusters, with plain vessels, Groups 1 and 2,

217

representing nearly 57 per cent. Another 32 per cent are vessels with dentate
and pseudo-scallop shell stamping in unbanked designs, and only 7.6 per cent
have banked stamping. The unbanked designs are reminiscent of later motifs
on Blackduck vessels, an indication that Long Sault Laurel is mid-Laurel as
Arthurs suggested, or even later. Of the three outliers, Vessel 46, with exterior
lip-rim tick marks, is significant because lip ticking is common on eastern Laurel

pots and uncommon in the west.

218

928903. 0cm 9.252. .9230 .6 $952 25:5 9: B $2.58 amEV Eu: 9: «a 339 05
am: 9 coan gnaw 95.. of. 699:0 9933590: 5 aczz-cm>>-_co-cmz->mv: 95 3.522 39:85. 9:. ”mm 939”.

 

 

 

 

 

 

 

 

 

 

219

 

220

JJ

IIIIW [10"

F

 

 

 

 

 

Figure 33: The Long Sault Site on the Rainy River in northwestern Ontario. Arthurs’ excavation block is at the upper right

(from Arthurs 1986).

Table 21: Long Sault Vessels (f=79)

Group 1: (f=30; %=37.9) Plain vessels with punctates/bosses on the lower
rim/neck.

S rou 1A 1: '%=11.4
Profile - Everted less than 45 degrees.
Interior - Plain
Lip - Plain
Exterior Upper Rim - Plain (Vessel 77 has unbanked vertical PSS)
Exterior Lower Rim/Neck - Bosses and punctates ( Vessel 5 has
only punctates)
Vessel #- 5, 7, 8, 9, 11, 59, 61, 77, 93

SUDQI‘OUQ 18 (1:11; 70:13.91
Profile - Straight

Interior - Plain

Lip - Plain (Vessel 88 has oblique dentate stamps)

Exterior Upper Rim - Plain

Exterior Lower Rim/Neck - Punctates/bosses (Vessel 22 has
unbanked oblique PSS)

Vessel # - 2, 10, 12, 22, 55, 56, 58, 62, 72, 73, 88

Profile - Straight and inverted

Interior - Plain

Lip - Plain

Exterior Upper Rim - Plain

Exterior Lower Rim/Neck - Punctates and bosses over unbanked
horizontal PSS

Vessel # - 20, 21

Subgroup 1D (f=L°/o=8.8)
Profile - Inverted
Interior - Plain
Lip - Plain
Exterior Upper Rim - Plain
Exterior Lower Rim/Neck — Plain or punctates lbosses
Vessel # - 4, 6, 51, 52, 60, 64, 75

Outlier (f=1; °/o=1.3)
Profile - Straight
Interior - Plain
Lip - Plain
Exterior Upper Rim - Unbanked oblique P88

221

Table 21 (Cont’d).

——————-————————_——————_——————_—-—-——————————————_~___——_—

Exterior Lower Rim/Neck - Punctates and bosses
Vessel # - 78

Group 2: (1:15; %=19.0) Plain vessels.

Subgrbub 2A (f=7; %=89)
Profile - Everted less than 45 degrees

Interior - Plain

Lip - Plain

Exterior Upper Rim - Plain
Exterior Lower Rim/Neck - Plain
Vessel # - 1, 53, 68, 70, 74, 76, 95

U rou 28 1:6' °/o=7.
Profile - Straight

Exterior Lower Rim/Neck - Vessel 3 has a boss
Vessel # - 3, 50, 66, 67, 69, 71

Su rou 2 f=2’ %=2.
Lip - Vessel 47 has transverse linear stamped tick marks
Exterior Rim-Lip Juncture - Linear stamped tick marks

Vessel # - 13, 47

Group 3: (1:15, %=19.0) Vessels with vertical or oblique, unbanked,
undragged pseudo-scallop shell stamps on the upper rim, and horizontal PSS
with punctates/bosses on the lower rim/neck.

SUQQTOUQ 3A (f=7; °/o=8.§)
Profile - Everted less than 45 degrees or straight

Interior - Plain

Lip - Plain

Exterior Upper Rim - Unbanked oblique or vertical PSS (Vessel 25
is decorated partially with dentate stamps)

Exterior Lower Rim/Neck - Unbanked horizontal PSS with
punctates/bosses. (Vessel 25 has dentate as well as PSS)

Vessel # - 23, 24, 25, 27, 31, 32, 33

SU I' B T=3' °/o= .
Profile -Everted less than 45 degrees and inverted
Lip - Oblique PSS
Exterior Upper Rim - Unbanked oblique PSS (oblique linear
stamps on Vessel 36).

222

Table 21 (Cont’d).

Vessel # - 29, 30, 36

SU rou 1:2“ °/o=2.5
Lip - Transverse linear stamps
Exterior Upper Rim - Vessel 28 has dragged PSS
Vessel # - 28, 44

§UQQTOUQ SD (f=2; %=2.5)
Exterior Lower Rim/Neck - Unbanked oblique PSS with
punctates/bosses.
Vessel # - 35, 42

Outlier (1:1; 70:13)
Profile - Straight
Interior - Plain
Lip - Oblique punctates
Exterior Upper Rim - Unbanked oblique dentate stamps
Exterior Lower Rim/Neck - Unbanked horizontal PSS with bosses
Vessel # 34

Group 4: (f=6; %=7.6) Vessels with unbanked exterior diagonal dentate or
linear stamps.

Sumrbub 4A (f=4; °/o=§.Q)
Profile - Everted less than 45 degrees

Interior ~ Plain

Lip - Plain

Exterior Upper Rim - Unbanked oblique dentate stamping
Exterior Lower Rim/Neck - Punctates/bosses and oblique dentate
stamping (Vessel 26 has horizontal dentate)

Vessel # - 26, 40, 81, 84

S r U 48 1:2' °/o=2.6
Exterior Upper Rim - Unbanked oblique linear stamping

Exterior Lower Rim/Neck - Linear stamps
Vessel # - 41, 43

Group 5: (1:4; °/0=5.1) Vessels with bosses on the exterior upper rim and
unbanked horizontal pseudo-scallop shell stamping on the lower rim.

Subgrbup 5A (f=3; %=38)
Profile - Everted lass than 45 degrees, and inverted

Interior - Plain
Lip - Plain

223

Table 21 (Cont’d).

Group 6:

decoration.

Exterior Upper Rim - Bosses
Exterior Lower Rim/Neck - Unbanked horizontal PSS
Vessel #- 16, 17, 18

U r U B f=1'°/o=1.
Exterior Upper Rim - Punctates and bosses
Vessel # - 19

(f=6; %=7.6) A weak group of vessels with banked stamped

§U|bgrgup §A (f=2; %=2.5)

Profile - Everted less than 45 degrees and straight
Interior - Plain

Lip - Plain

Exterior Upper Rim - Banked and dragged linear stamps
Exterior Lower Rim/Neck - Vessel 39 has a punctate.
Vessel # - 38, 39

Subgroup 68 (f=4, %=5.0)

Exterior Upper Rim - Banked, undragged linear or dentate stamps.
Exterior Lower Rim/Neck - Banked, undragged linear or dentate
stamps.

Vessel # - 14, 15, 37, 82

Outliers: (f=3; °/o=3.8) Three vessel are outliers of all groups. Vessel 45 is
plain except for three random incised lines on the lower rim/neck. Vessel 46 has
linear stamps on the upper interior and lip, and linear stamp exterior lip/rim
ticking. Vessel 89 has linear stamps on the upper interior rim and lip, but no
exterior decoration.

 

224

.mcE:_oo 23650 9:
E 950% 6%? some So 9260 E: 9 .99 98E? :8 on» .88? 9:52 E9: :20 2E9¢o 22mm 0:3 new 939“.

....._._~

mCOaOZZJX

 

225

r
m
w
P
w
r
P
F
r

.920 2E98 gum 9.3 3.388 «m 9:9“.

 

226

taco 2E98 gnaw 0:3 ”25:09 vm 9:9“.

.9

I... .222— .—

mmOaOzZJXa.

 

227

.tmco 2E98 gnaw 9.3 ”25:09 vm 9:9“.

. D

p

.:::._::::.:

m¢0a025423.20mm000<

 

228

taco o_E9mo :33 0:3 “63:09 vm 930E

romance mo

.....—_.~....._¢.__

m¢0u0224x§_xouw0

 

229

9% :23 95.. 9: E9__ m_ommo> ”mm 939“.

abs-0:53:00

 

 

.
04.1.1 ,
444444 Ii
.i...4 .{h
.- 14...“...
. ..D

i ...ua

.9 a

bill.-

230

90009000 0:0 E0::0: 05:00 :0 >:0_:_s_ 0:020 00: :0 000000 .000: 0552 E0: 90:00

.0 0:90 E0: 000E000 :30 02.08.8300 03:00:: .0 0:90 E0: 000E000
:30 0200000000 03:00:: .0 0:90 E0: 000000 0:; 000E0:0 __0:0 02.08.0308 03:00:: ”39 E0000

.0 0:90 E0: 000E0:0 __0:0-00__000-00:000 00390:: .0 0:90 E0: 000E0:0 :30
02080088 03:00:: .0 0:90 E0: 0009000 00:50 03:00:: .0 0:90 E0: 000E000 0:00:00 03:00 ”39 0:00.

.0 0:90 E0: 000E000 0:00:00
03:00:: .0 0:90 E0: 000E000 03:00 .0 0:90 E0: 500 .0 0:90 E0: 000E000 0:00:00 03:00 ”39 0:0000

.F 0:90 E0: 000000 0:; E00 .0 0:90 E0: 500 .F 0:90 E0: 02000::0 0:; 50.0 .0 0:90 E0: :00 ”39 00 ._.

0:0 :.:00 0:00 0.: Eo: 0.0000> “V0 050i

231

APPENDIX J

BALLYNACR EE SITE (Dka-8)

In 1983, the Regional Archaeologist’s office of the Ontario Ministry of
Citizenship and Culture began a four-year excavation of the Ballynacree Site in
Kenora, Ontario (Figure 36), to obtain data concerning the community pattern of
an entire Laurel habitation site (Reid and Rajnovich 1991). A complete Laurel
village had never before been exposed.

The site on an island in the Winnipeg River had been located during
Kenora-area surveys conducted in the 19703, and it was known to contain a
substantial Middle Woodland component. It offered an excellent chance to
expose an entire Laurel encampment because it was bounded on all sides by
the \Mnnipeg River and a high rock outcrop of the Laurentian Shield (Figure
37).

The extensive excavations were in a large block of units covering the
entire flat, grassy terrace about 30 x 20 metres. The area was dug in 3—cm
arbitrary levels exposing a stratified but somewhat mixed series from Historic to
Laurel. The Middle Woodland component consisted of the outlines of three
complete lodges (Figure 38), none of which overlapped, an indication of
probable contemporaneity. Preliminary seriation of the ceramics from each
house showed close similarity among the houses, another indication of
contemporaneity. We used ethnographic analogy to formulate population size
and concluded that the Ballynacree Site consisted of three extended families of
about 10 or 11 people each (Reid and Rajnovich 1991: 220). Wobst has noted
that minimum band size for a number of hunter-gatherer populations is about 25

(Wobst 1974: 170). Cleland estimated that 30 people comprised the smallest

232

 

I ‘.
9 £3 E
o

   

 

 

a
tMEOF 339$
gm: wooos

 
 

 

 

 

 

Figure 36' The Ballynacree Site (Dka-8) on the Winnipeg River in Kenora

Ontario (from Reid and Rajnovich 1991).

 

     
    

WINNIPEG
RIVER

. ' ;- - -: ham-ll |
- WW, I

I., '. ‘ “‘ 2‘.

.- j,- . ' EXCAVATIONS \ . ~

3." . < \

9 ‘4'... .: 2.. ‘ '

A" . -_ " “mug. ‘.~‘\.

0’ I '3 l “ ,
M M" - ‘_ \ g ‘

m... «M

 

 

 

Figure 37 The excavation block at the Ballynacree Site (from Reid and

Rajnovich 1991).
233

 

 

 

 

 

 

 

 

  

 

 

 

—ucanno«
(00¢

 

 

 

Figure 38: The three houses and related features at the Ballynacree Site (from

Reid and Rajnovich 1991).

234

 

summer band of the historic Ojibway (Cleland 1992: 101). The Ballynacree Site,
then, would represent a minimum band or part of a band.

Three radiocarbon assays were obtained from charcoal from the floor
and central hearth of House #1, dating to AD. 1240:45, A.D. 1240:65 and AD.
1270:55 (Reid and Rajnovich 1991: 198). The dates are among the latest
obtained for Laurel and have been questioned by several researchers,
however, preliminary seriation of the ceramics from Ballynacree and other
Laurel components, based on external decorative techniques, indicated that
Ballynacree is the latest Laurel site in the western Laurel area, regardless of the
questionable validity of the absolute dates (Reid and Rajnovich 1991: Figure
28).

The Ballynacree ceramics are included in this study to reexamine their
position in the seriation of Laurel sites with the expectation of providing a very
late assemblage representing the Hungry Hall Complex. They are shown in
Figure 39, the ceramic chart. Table 22 describes the clusters. Representative
ceramics are in Figures 40 to 45.

Vessels 28, 29 and 105 were not clustered because they lack information
on lower rim/neck areas. The assemblage is characterized by a small
percentage of banked stamped motifs (10.7 per cent) and the presence of Late
Woodland traits such as cord-wrapped stick stamps, single-cord impressions,
and oblique-over-horizontal motifs. It also contains a unique incised design on
Vessel 88, and a large number of dentate-stamped vessels (>40 per cent) in
comparison to the tiny number of pseudo-scallop shell stamped ones (4.5 per
cent).

The cluster analysis for this dissertation produced 12 major groups as
shown in Table 22. Groups 1 and 2 are 18 vessels that are plain, with or without

punctates/bosses. Vessel 88 of Group 1 has an isolated incised triangular

235

Table 22: Ballynacree Vessels (f=112)

Group 1: (f=16; °/o=14.3) Plain vessels with punctates/bosses on the lower
rim/neck.

Sumroup 1A (f=12; %=1Q.7)
Profile - Everted less than 45 degrees, straight and inverted

Interior - Plain

Lip - Plain

Exterior Upper Rim - Plain

Exterior Lower Rim/Neck - Punctates and punctates-and-bosses.
(Vessel 97 also has horizontal CWS below the punctates.
Vessel # - 2, 3, 4, 5, 6, 7, 81, 82, 84, 86, 87, 97

 

Subgroup 1mf=3;%=2.7)
Profile - Inverted and less than 45 degrees everted I
Exterior Lower Rim - bosses (Vessel 88 has punctates and bosses
over a unique incised design)

Vessel # - 54, 85, 88

S f’ U 1 I=‘I' °/o= .
Exterior Upper Rim - Vertical single-cord impressed (SCI)
Exterior Lower Rim - Bosses above diagonal and vertical SCI
Vessel # - 115

Group. 2: (f=2; %=1.8) A group containing only one subgroup of plain vessels.

Profile - Everted less than 45 degrees and inverted
Interior - Plain

Lip - Plain

Exterior Upper Rim - Plain

Exterior Lower Rim - Plain

Vessel # - 1, 53

Group 3: (f=3, %=2.7) Vessels with horizontal motifs on the upper rim, done
with cord-wrapped stick (CWS) or dentate stamps and punctates.

Sumroup 3A (f=2; %=1,8)
Profile - Less than 45 degrees everted

Interior - Oblique CWS

Lip - Plain

Exterior Upper Rim - Horizontal CWS or oblique CWS
superimposed on horizontal CWS

Exterior Lower Rim - Punctates or punctates over oblique CWS
superimposed on horizontal CWS

236

Table 22 (Cont’d).

Vessel # - 52, 114

Subgroup SB (f=1, %=ug)
Interior - Oblique dentate stamping
Exterior Upper rim - Horizontal unbanked dentate stamping
Exterior Lower Rim - punctates.
Vessel # - 106. It is the same as Vessel 52 but decorated with
dentate rather than CWS

Group 4: (f=43; %=38.4) A large group of vessels with unbanked dentate
stamped decoration.

WWW
Profile - Mostly less than 45 degrees everted, but Vessels 16 and
26 are straight, Vessel 25 is inverted and Vessel 31 is more than
45 degrees everted.
Interior - Plain (Vessels 20 and 99 have oblique dentate)
Lip - Plain or various dentate designs ( Vessels 23, 24 and 25
have oblique linear stamps)
Exterior Upper Rim - Unbanked vertical/oblique dentate stamping
Exterior Lower Rim - Unbanked vertical/horizontal dentate
stamping.
Vessel # - 13 to 16, 20, 21, 22, 23, 24, 25, 26, 30, 31, 32, 57, 58.
59, 60, 61, 70, 99.

l‘OU 48 1:1 ' °/o=14.
Exterior Upper Rim - Dentate stamping, plain or horizontal rows of
linear stamps
Exterior Lower Rim - Various designs using dentate stamps
including verticals, diagonals, horizontals and chevrons.
Vessel # - 19, 62, 65, 66, 68, 69, 77, 89, 90, 98, 100, 101, 102,
104, 107, 108

UtIiel‘S 1:6' °/o= .4
Vessels with various combinations of unbanked dentate stamped
elements - verticals or obliques over horizontals, obliques or
chevrons (Vessels 17, 18, 63, 64, and 93) and one vessel (110)
with combined linear stamping and dentate.

23 7

..Jl ”—1 -w
' -.I—

 

I?

Table 22 (Cont’d).

Group 5: (1:5; %=4.5) Vessels with unbanked dentate stamping in
predominantly vertical motifs.

r u A i=4' %=3.
Profile - Everted less than 45 degrees
Interior - Plain
Lip - Plain
Exterior Upper Rim - Unbanked vertical dentate stamps and
punctates
Exterior Lower Rim/Neck - Unbanked vertical/oblique dentate
stamps (Vessel 71 has banked and dragged dentate stamps)
Vessel # - 67, 71, 91, 92

U l‘ B I=1 ' °/o= .
Exterior Upper Rim - Criss-cross and vertical dentate stamps in
zones, and punctates
Exterior Lower Rim/Neck - Criss—cross and vertical dentate stamps
in zones
Vessel # - 34

Group 6: (f=5; %=4.5) Vessels with unbanked pseudo-scallop shell stamps in
motifs that are verticals/obliques over horizontals. This cluster is at the level of a
subgroup.

Profile - Everted less than 45 degrees or straight

Interior - Plain (vessel 109 has oblique PSS)

Lip - Plain or transverse PSS

Exterior Upper Rim - Unbanked vertical or oblique PSS
Exterior Lower Rim/Neck - Unbanked horizontal PSS and
punctates or bosses

Vessel # - 33, 72, 95, 96, 109

Outller: (f=1. %=0.9) One vessel with with motif like those of Group 6 but done
with linear stamps and dentate stamping.

Profile - Everted less than 45 °/o

Interior - Plain

Lip - Transverse linear stamps

Exterior Upper Rim - Unbanked vertical linear stamps

Exterior Lower Rim/Neck - Unbanked horizontal dentate stamps
with punctates

Vessel # - 39

238

Table 22 (Cont’d).

Group 7: (f=6; %=5.3) Vessels with single-cord impressed decoration
clustered at the level of a subgroup.

Profile - Everted less than 45 degrees or inverted

Interior - Plain (Vessel 50 has oblique CWS)

Lip - Plain or decorated with SCI, CWS or incising

Exterior Upper Rim - Vertical, horizontal or oblique SCI
Exterior Lower Rim/Neck - Various motifs including vertical,
horizontal, oblique and chevron SCI with punctates
Vessel #-49, 50, 51, 83, 112, 113

Group 8: (f=8; %=7.1) Vessels with banked linear stamped designs.

Subgroup 8A (f=6; °/o=5.3)
Profile - Everted less than 45 degrees

Interior - Plain

Lip - Plain

Exterior Upper Rim - Banked vertical or oblique linear stamps
(Vessels 40 and 76 also have punctates)

Exterior Lower Rim/Neck - Banked vertical or oblique linear
stamps with punctates

Vessel #- 40, 42, 43, 76, 103, 111

Su rou 88 f=2' %=1.

Lip - Transverse or oblique linear stamps
Vessel # - 41, 78

Group 9: (f=4; °/o=3.6) Vessels with incised decoration.

Subgrpup 9A (f=3; °/o=2.7)
Profile - Straight or everted less than 45 degrees

Interior - Plain

Lip - Incised or plain

Exterior Upper Rim - Oblique incising (Vessel 48 also has
punctates)

Exterior Lower Rim/Neck - Horizontal or oblique incising with
punctates

Vessel # - 44, 46, 48

U lo 93 f=1'°/o=0.

Interior — Oblique incising and linear stamps
Vessel # - 73

239

 

Table 22 (Cont’d).

Outlier: (f=1; %=0.9) Vessel 45 has oblique over horizontal motif like Group 9
but done in cord-wrapped stick.

Group 10: (f=6; %=5.4) Vessels with predominantly horizontal designs in
dentate stamps, cord-wrapped stick and single-cord impressed.

U 1‘ 10A I: ' °/o=2.7
Profile - Everted less than 45 degrees
Interior - Plain
Lip - Plain or oblique incised !
Exterior Upper Rim - Horizontal dentate stamping and punctates
Exterior Lower Rim/Neck - Horizontal dentate stamping
Vessel # - 12, 55, 56

 

u ro 1 B 1:2' %=1.
Exterior Upper Rim - Horizontal CWS or SCI with punctates
Exterior Lower Rim/Neck - Horizontal CWS or SCI
Vessel # - 79,80

Outlier 0:130:09)

Vessel 27 has dentate stamps decoration like Group 10 but has added
zoning of oblique dentate stamps on the upper and lower rim. It also has
a profile everted more than 45 degrees.

Group 11: (f=7; °/o=6.3) A group of plain or incised vessels with punctates on
the upper rim.

Su ro 11A f:5' %=4.
Profile - everted less than 45 degrees or straight
Interior - Plain
Lip - Plain
Exterior Upper Rim - Punctates/ bosses
Exterior Lower Rim/Neck - Plain (Vessels 8 and 74 have oblique
incised lines)
VesseI#-8,9,10, 11,74

Su rou 118 f:2' %=1.
Exterior Upper Rim - Unbanked vertical incised lines and
punctates
Exterior Lower Rim/Neck - Unbanked vertical incised lines
Vessel # - 47, 75

240

Table 22 (Cont’d).

_—_——————_—_—————_——_—-———e——u_—————~“——_——_*-fl_———_——_*

Group 12: (f=4; °/0=3.6) Vessels with banked stamped, dragged designs.

Sumroup 12A (f=2; °/o=1.8)
Profile - Everted less than 45 degrees

Interior - Plain

Lip - Plain

Exterior Upper Rim - Banked, dragged dentate stamps
Exterior Lower Rim/Neck - Banked, dragged dentate stamps
(Vessel 38 also has punctates)

Vessel # - 35, 38

S f‘OU 128 1:2“ %=1.

Profile - everted or straight

Exterior Upper Rim - Banked, dragged linear stamps
Exterior Lower Rim/Neck - Banked, dragged linear stamps
Vessel # - 36, 37

Outlier: (f=1, %=O.9) Vessel 94 has vertical dentate stamps only on the upper
rim and lip. It is unrelated to all groups.

 

241

 

"I" 'l’. o '

devise on the shoulder (Figure 40). Group 3 contains vessels with cord-
wrapped stick decoration usually considered late Laurel (Figure 41, bottom).
Group 4 is the largest group and it, along with Group 5, represents unbanked
dentate stamped vessels (Figures 42 and 43, top). Group 6 is a pseudo-scallop
shell stamped cluster, and Group 7 represents very rare single-cord impressed
designs. Group 8 is linear stamped (Figure 43, middle and bottom), and Group
9 is incised (Figure 44). The last three groups represent motifs rather than
techniques: Group 10 represents vessels with predominantly horizontal motifs;
Group 11 has vessels with upper rim decoration as the predominant feature;
and Group 12 has banked and dragged stamped designs (Figure 41, top and
middle).

Table 23 delineates the vessel groups found in each house. Of the 12
groups, House #1 (vessels 1-52) has all of them, House #2 (Vessels 53-80) has
nine (75 per cent) and House #3 (Vessels 81-115) has seven (58 per cent).
While all of the seven groups of House #3 are shared with House #1, and five of
its seven with House #2, it lacks five groups represented in House #1 and four
in House #2.

Table 23: B II n

reV llrbHou

Cluster House #1 House # 2 House #3 Totals

 

 

 

1 6 (12%) 1 (3.6%) 9 (26.5%) 16 (14.3%)
2 1 (2%) 1(3.6%) - 2 (1.8%)
3 1 (2%) - 2 (5.9%) 3 (2.7%)
4 1 7 (34%) 14 (50%) 1 2 (35.3%) 43 (38.4%)
5 1 (2%) 2 (7.1%) 2 (5.9%) 5 (4.5%)
6 1 (2%) 1 (3.6%) 3 (8.8%) 5 (4.5%)
7 3 (6%) - 3 (8.8%) 6 (5.3%)
8 4 (8%) 2 (7.1%) 2 (5.9%) 8 (7.1%)
9 3 (6%) 1 (3.6%) - 4 (3.6%)
10 2 (4%) 4 ( 14.3%) - 6 (5.4%)
11 5 (10%) 2 (7.1%) - 7 (6.3%)
12 4 (8% ) - - 4 (3.6%)
Outliers 2 (4%) - 1(2.9) 3(2.5%)
T 1 IS 1 0/o 2 1 °/o 4 1000/0 112 100 0/o

242

 

If we assume that the houses are contemporary, they may provide us with
some indication of the similarity coefficient that should be expected in the event
of a migration. We should expect a Founder Effect to be present in assemblages
representing the emigrant group splitting from a parent population; that is, the
assemblage of the emigrant group would represent only a sample of that of the
parent population because the emigrants themselves are a sample of the
parent group selecting styles representing a portion of the full style repertoire of
the homeland (Sackett 1982: 72-73). But how similar would the emigrant and
parent assemblages be? Consider House #1, with all 12 clusters represented,
to be the parent population from which samples for Houses #2 and #3 are
drawn. If people from either of these emigrated, the assemblages would provide
us with a minimum coefficient of similarity between the home and emigrant
groups. The similarity coefficients between the parent group of House #1 and
samples from the other houses is 131 (65.per cent) for House #2 and 128 (64
per cent) for House #3. Thus, if residents of either houses emigrated, they would

take a minimum of 64 per cent of the decoration of the parent sample.

243

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

BALLYNAGREE SITE
House“
I
2 j .
0.. U 000 ‘5 I o o
6. l
I" U 000 7 ° '
I
as U 00 3 I
1% °
I
00 O U a o e 3 . 0.
9.0.0. G 0.. Q” :'::--
l
"332 .. '2 l
33:-3
53.1: I I'
IS
II] _,,, i l
9 I
I
'{eifevei leeeoo '3'8 ' | 000
a»: ° ° '9 ‘
€3.55

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 39: Ballynacree ceramic chart.

244

IT- _ _:.£‘.E_A'Zi_“ 11...".R.,L'|. a

BALLYNACREE SITE

IIouse

#1

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

34
oi oi '
3S - ’
§§7 I ' 000
33 o
OCDO
I
39 |
000
l L
0000. U 40 :.-j:‘:
o o '
Joooo
I
J ‘ ‘5 4|
w 42 | uunz.‘ : 3"
0:30 o o
52 I
43 |
44

 

Figure 39 (cont’d): Ballynacree ceramic chart.

245

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

:0000
L_r
oo
o o
000
r'r— ss /
#393: .0000
t::;::* I
...... 1’
:‘fi‘ loooo‘ so l 0..
L
I 'oo‘
. Lu,
000
:::'54 0000
I ‘55 000'
50000
L_ l
O O
0.0 66
L——J I

 

Figure 39 (cont’d): Ballynacree ceramic chart.

246

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

*0 °°
.

Figure 39 (cont’d): Ballynacree ceramic chart.

 

 

 

 

 

 

 

247

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

IOI,
000
I02
0 0 r
I03
00
I04 |
0 O I 000
I05 fie?“
o a flap;
:3 I
0.0. 000

 

 

 

 

 

I06
I07

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 39 (cont’d): Ballynacree ceramic chart.

248

 

 

 

 

 

Figure 40: Shoulder portion of a vessel from the Ballynacree Site with an
isolated triangular devise on the shoulder. The vessel, background, is plain with

bosses and punctates on the rim, from Group 1.

.w
I
A, >
‘1‘. ‘1
I .
\.
, ‘ ‘ “I
t
‘ h.
" i

. ...,
R :o-r ,

 

 

l'” , ‘ . ‘ “I: ‘ ~._ ‘.
.' 1‘ I
I" ‘ ’ I
1 1~ -’ ‘ ». -

Figure 41: Vessels from the Ballynacree Site: top and middle, banked dentate

 

stamped from Group 12; bottom, cord-wrapped stick stamped from group 3.

249

 

 

 

Figure 42: Vessel from the Ballynacree Site; unbanked dentate stamped from

Group 4.

250

 

Figure 43: Vessels from the Ballynacree Site: top, vertical dentate stamped from

Group 4; middle and lower, banked linear stamped from Group 8.

251

 

 

 

252

 

  

I;

Figure 45: Zoned dentate stamped vessel from the Ballynacree Site, an outlier

 

 

of Group 10.

253

APPENDIX K

SQUIRREL DAM SITE (470N21)

In the mid-19603, Robert J. Salzer of Beloit College undertook a number
of excavations in north-central WIsoonsin in an effort to devise a basic culture
history for a region that had been ignored in favor of extensive studies of
Hopewell and Mississippian sites to the south. Middle and Late Woodland
studies, including Laurel, were well underway in areas to the west, north and
east of the area (Stoltman 1962, Wright 1967, Janzen 1968).

The Squirrel Dam Site, at the mouth of the Squirrel River at the outlet of
Squirrel Lake in Oneida County, became a major contributor to the definition of
the Nokomis Phase (Salzer 1969). Salzer excavated 15 five-foot squares using
natural stratigraphy. There were five strata containing Archaic to Late Woodland
material. In Stratum 8, he found Early Woodland-like ceramics in similar
horizontal and vertical distribution to Middle Woodland-like sherds, and initially
concluded that he was dealing with a single assemblage (Salzer 1969: 152).
Accordingly, he defined the Nokomis Phase as containing both Early and
Middle Woodland characteristics, sometimes on the same vessel (1969: 153).
He also noted the presence of exotic traits on Nokomis Phase vessels including
Laurel’s characteristic pseudo-scallop shell stamping (1969: 407). Salzer did
not recover bone or charcoal samples for radiocarbon dating. Recently, Charles
R. Moffat obtained a radiocarbon sample from a pit at Squirrel Dam related to
Lake Nokomis Trailed and thick, Dane Incised, dating to 560170 BC.
uncorrected (calibrated to a range of 793 BC. to 427 BC. with one Sigma).
Moffat suggested that the material used to define the Nokomis Phase included a
mixture of Early and Middle Woodland material (Moffat 1999: 93).

254

 

 

Vessels from Squirrel Dam are shown in Figure 46, the master ceramic
chart. The clusters are described in Table 24. Samples are depicted in Figures
47 and 48.

Most vessels from Squirrel Dam have profiles that are everted less than
45 degrees and a few have straight rims. None are inverted or greatly everted
like some Laurel vessels. Also unlike Laurel, the vessels in this study produced
weaker clusters that indicate more variety within the clusters. Instead of Groups
splitting at about JC=.50, the Squirrel Dam Groups split at JC=.60 or greater.
Salzer originally noted the extreme variety of decoration in Nokomis Phase
(1974: 47), and the cluster analysis bears out the observation.

Groups 1, 2, 3, 4 and 5 are in a super cluster of Early and Middle
Woodland vessels. Most are Lake Nokomis Trailed and Dane Incised. Vessel
32 which forms Subgroup 1B is thin and is probably Late Woodland. Groups 6
and 7 are Late Woodland Lakes Phase.

Vessel 28 is especially interesting: it has zoned decoration like Havana
Ware but the technique is pseudo-scallop shell stamped, a Laurel trait. Vessel
21 is the only vessel in this assemblage with banked stamping similar to Laurel
vessels. Vessel 15 is the only vessel with a classic Laurel thinned-lip profile.
Vessel 25 has column of dentate stamps similar to Laurel. All of these vessels
are outliers. The only Laurel-like vessel that fits in a cluster is Vessel 17 with
vertical over horizontal incising and bosses.

The analysis suggests some temporal continuity among the ceramics.
While the cluster analysis separated the Early-Middle Woodland from Late
Woodland ceramics, there is some overlap; Vessel 32, a probable Late
Woodland vessel, is in Group 1 with Middle Woodland material. Vessel 18, a
Middle Woodland vessel, is in Group 7 with Late Woodland ceramics. The

analysis also did not isolate an Early Woodland from a Middle Woodland

255

cluster.

Early Woodland style is apparent on Vessel 19 with interior cord marking.
Vessel 34, not included in the cluster analysis because it lacks a rim, has a flat
base, an Early Woodland vessel form. Therefore, the radiocarbon date of about
500 BC. has to be seriously considered for some of the material. If the early
date receives further support, the hypothesis that Early Woodland is time
transgressive, occurring earlier in the south than the north of the Great Lakes

area (Stoltman 1997: 137), will have to be reconsidered.

256

 

 

Table 24: Squirrel Dam Vessels (f=33)

 

 

 

Group 1: (f=5; %=15.2) Cord-marked vessels with no exterior
decoration. These could be related to Early Woodland/Lake Nokomis Trailed.

Interior - Plain

Lip - Plain or oblique linear stamps or incising
Exterior Upper Rim - Plain

Exterior Lower Rim/Neck - Plain

Vessel # - 2,3,4, 5,7

Group 2: (f=3; %=9.0) Cord-marked vessels with horizontal incising. These
could be related to Dane Incised/Black Sand Incised etc.

Subgroup 2A (f=2; %=§.Q)

Interior - Plain

Lip - Transverse incising

Exterior Upper Rim - Horizontal incised lines (On Vessel 20, these
lines are enclosed in opposed triangles

Exterior Lower Rim/Neck - Horizontal incised lines

Vessel # - 16, 2O

Subgroup 28 lf=11 °/o=3.0)
Exterior Upper Rim - Horizontal incised lines beside diagonal
single-cord impressed lines
Exterior Lower Rim/Neck - Horizontal incised lines beside
diagonal SCI lines
Vessel # - 32. This vessel is very thin, probably Late Woodland

Group 3: (f=3; °/o=9.0) Cord-marked vessels with trailed designs. These are
probably Lake Nokomis Trailed.

Interior - Plain or trailed

Lip - Plain or punctated

Exterior Upper Rim - Horizontal trailed lines

Exterior Lower Rim/Neck - Oblique trailed or incised lines
Vessel #- 1, 23, 11

Group 4: (f=5; %=15.2) A cluster with corded or smoothed exteriors and
dominated by horizontal trailing or incising.

Interior - Plain

Lip - Plain or oblique or transverse CWS

Exterior Upper Rim - Undecorated or incisedftrailed lines. Vessel
17 has verticals over horizontals.

Exterior Lower Rim/Neck - Punctates and horizontal lines

257

 

 

Table 24 (Cont’d).

Vessels#-10, 12, 13, 17,24

Outliers (f=2; °/o=6.0) Vessels 9 and 19 are outliers weakly related to
Group 4. They have horizontal lines on the lower rim/neck with punctates
like Group 4 but they have oblique motifs done with trailing or linear
stamps on the upper rim. Vessel 19 also has interior horizontal cord
marking.

Group 5: (f=2; %=6.0) A weak cluster of vessels 8 and 30 distinguishable by
their exterior surface treatment. Both have only lower rim punctates
superimposed on either a smoothed surface or a net-impressed surface.

Group 6: (f=2; %=6.0) Single-cord impressed vessels with plain surfaces.
These thin vessels are probably Late woodland.

Interior - Plain

Lip — Plain or SCI

Exterior Upper Rim - Vertical or horizontal SCI
Exterior Lower rim/Neck - Horizontal SCI
Vessel # - 31, 33

Outlier: (F=L'%=3.0) Vessel 25 has horizontal dentate decoration on the
upper and lower rim.

Group 7: (f=3; %=9.0) Cord-wrapped stick stamped vessels with plain
surfaces. Vessels 27 and 29 are thin and probably Late Woodland.

Interior - Oblique CWS or plain

Lip - Transverse CWS or plain

Exterior Upper Rim - plain or horizontal CWS (Vessel 19 has
chevrons)

Exterior Lower Rim/Neck - Horizontal CWS (Vessel 18 has
bosses)

Vessel # - 18, 27, 29

Outliers: (f=7, %=21.2) Five vessels only weakly related to Groups 6 and 7.
Vessels 6, 15, 22, 26 and 28 have very varied decoration. Vessels 6 and 15 are
a weak cluster with plain exterior surfaces and punctates. Vessel 28 has
pseudo-scallop shell zoning on a plain surface. Vessel 22 and 26 form a very
weak cluster; they are related only by their oblique upper rim decoration.

Vessel 14 has trailed lines on the upper rim and a smoothed lip. It
probably is related to group 4 but information is lacking on the lower rim/ neck.

Vessel 21 has a design of incised and banked linear stamping only
weakly related to Groups 1, 2 and 3.

258

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0U.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

@§

 

 

 

 

 

 

 

 

 

259

 

 

 

 

 

 

IIII'IIIII

 

 

 

 

 

 

 

 

 

 

SQUIRREL DAM SITE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 46: Squirrel Dam ceramic chart.

 

SQURREL DAM SITE

 

I BASE

 

 

"

 

2|

 

 

 

 

 

 

 

3|

IIIS

 

 

 

 

 

 

 

O O O O

 

 

 

 

 

 

22

 

 

 

E

 

 

 

 

 

 

 

 

 

 

mmczz

 

 

 

 

é

 

 

III:

 

24

 

 

:L—aI—x
.%II

 

/___IU’

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

...]
../'
--/': ZS
-../
-../f:
23;:
26

 

 

 

 

 

 

 

 

 

 

33‘; 27

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I - EI
mmmmmmmmmm
fl ‘

 

...lll IIIII
\\\\\ 2“
£3? 29 | l
30

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

magma—.22:

 

 

 

 

 

 

 

Figure 46 (cont’d): Squirrel Dam ceramic chart.

260

 

 

Figure 47: Vessel from the Squirrel Darn Site: dentate stamped and incised.

261

 

Figure 48: Vessel from the Squirrel Dam Site: pseudo-scallop shell stamped.

262

APPENDIX L
THE RICHTER SITE (47Dr80)

The Richter Site produced one of the earliest dates for the North Bay
complex, therefore it has been of considerable interest to researchers (R.
Mason 1991: 126). However, details of the excavations have not been
published. The site is in a protected bay on the south end of Washington Island
at the mouth of Green Bay. It was excavated in 1968 and 1973 by Richard
Peske and by Gordon Peters (under the direction of Guy Gibbon) of the
University of Wisconsin-Milwaukee who unearthed several large features that
were apparently “pit houses” (R. Mason 1991: 126). Wood charcoal from House
ll produced a date of 520165 BC. uncorrected.

The ceramics from the Richter Site along with field notes are housed at
the University of Wisconsin-Milwaukee; however there are no site maps and no
directors’ notes with the assemblage. The ceramic chart is at Figure 49 and the
clusters are described in Table 25. Selected rims are shown in Figures 50 to 52.

The assemblage consists of a very varied group, including both
undecorated plain and cord marked vessels, cord-wrapped stick stamped,
incised, pseudo-scallop shell stamped, and linear stamped vessels and one
dentate stamped pot. Noteworthy among them is the high number of vessels
with linear stamped or punctated decoration only on the upper rim reminiscent
of Steuben Punctate of the Havana series except that the North Bay vessels
have punctates that differ from the hemiconical Steuben punctates. Decoration
only on the upper rim occurs in Laurel but it constitutes a minor Laurel style (see
Cluster P in Chapter 6). A Laurel tie to Havana through the North Bay series is

possible. Vessel 12 from the Richter Site has punctates formed by an annular

263

punch like Sister Creeks Punctate of Havana. The cord-wrapped stick stamping
may also be a Havana influence, but the pseudo-scallop shell stamps are a
northern characteristic. The assemblage is a good example of the “cultural
tension zone effect”, as Mason has aptly described North Bay (R. Mason 1967:
341).

Seriation of the Richter assemblage with other North Bay ceramics
places the site in the later portion of the series contrary to the radiocarbon date

produced by charcoal from the site (see Chapter 5).

264

Table 25: Richter Vessels (f=38)

The Richter site clusters are weaker than most Laurel assemblages, indicating
greater variation within each cluster as well as between clusters (Groups split at
about .60). groups 1 to 4 are plain surfaced vessels and Group 5 is cord-
marked.

Group 1: (f=16; %=42.1) Relatively plain vessels, the majority with upper rim
decoration done with linear stamps or shallow punctates. These vessels are
reminiscent of the motif of Steuben Punctate of Havana Ware but without the
characteristic hemiconical Steuben punctates.

Sumroup 1A (f=7; %=18.4)
Profile - Everted less than 45 degrees
Interior - Plain
Lip - Plain
Exterior Upper Rim - Single or multiple rows of vertical linear
stamps or shallow punctates (Vessel 34 is plain)
Exterior Lower Rim/Neck - Plain
Vessel # - 7, 13, 21, 22, 23, 25, 34

Subgroup 18 (1:2; %=53)

Lip - Plain or transverse linear stamps
Exterior Upper Rim - Plain

Exterior Lower Rim/Neck - Plain or punctated
Vessel # - 24, 25

W (f=4; °/o=10.5)
Profile - Everted less than 45 degrees or straight
Exterior Upper Rim - Vertical linear stamps
Exterior Lower Rim/Neck - Vertical linear stamps
Vessel # - 15, 16, 18, 20 (Vessel 15 is cord-marked)

wad; %=7.9)
Lip - Plain or oblique linear stamps
Exterior Upper Rim - Oblique linear stamps
Exterior Lower Rim/Neck - Plain or oblique linear stamps (Vessel 4
also has horizontal PSS)
Vessel #-4, 14, 17

Group 2: (f=3; °/o=7.9) Incised vessels.

Profile - Everted less than 45 degrees or straight
Interior - Plain

265

Table 25 (cont’d).
Lip - Plain
Exterior Upper Rim - Oblique dentate stamps or incising (vessel 37
has criss-cross incising)
Exterior Lower Rim/Neck - Horizontal incising (Vessel 37 has criss-
cross incising)

 

Vessel # - 8, 37, 38
Group 3: (f=2; %=5.3) Cord-wrapped stick stamped vessels.

Profile - Everted less than 45 degrees

Interior - Oblique CWS

Lip - Plain

Exterior Upper Rim - Plain

Exterior Lower Rim/Neck - Oblique or horizontal CWS
Vessel # - 9, 11

Group 4: (f=2; %=5.3) - Pseudo-scallop shell stamped vessels.

Profile - Straight or inverted

Interior - Plain or combed

Lip - Plain

Exterior Upper Rim - Unbanked oblique PSS
Exterior Lower Rim/Neck - Unbanked oblique PSS
Vessel # - 2, 5

Group 5: (f=8; %=21.1) Vessels with cord-marked surfaces and minimum
decoration.

Profile - Everted less than 45 degrees or straight

Interior - Plain

Lip - Plain, oblique CWS or oblique or transverse linear stamp

Exterior Upper Rim - Cord-marked (Vessel 19 also has oblique
linear stamps)

Exterior Lower Rim/Neck - Cord-marked

Vessel # - 19, 27, 28, 29, 30, 32, 33

Outlier (f=1: °/o=2.6)
Lip - Oblique incising
Exterior Upper Rim - Oblique incising

Outliers: (f=7; %=18.4) Several vessels are outliers of all groups.
Vessel 1 has exterior banked PSS.

Vessel 3 has unbanked PSS on both the interior and exterior.
Vessel 6 has unbanked oblique dentate over vertical PSS.
Vessel 10 has vertical CWS on the exterior upper rim only.

266

Table 25 (cont’d).

Vessel 12 has annular punctates similar to Sister Creeks Punctated.
Vessel 26 has a row of punctates on both the upper and lower exterior rim.
Vessel 31 is a smoothed vessel with no decoration.

 

267

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Si

 

 

 

‘
93

 

a“! :‘

 

 

 

 

 

 

 

III!

 

 

 

 

/U3 my”:
was ml!

gum .fifib

qu ..‘ (13
EU ‘ flung
"I q
a I
I

 

9&0”

 

 

 

 

 

 

 

 

 

 

 

 

 

 

FILM

 

 

 

Figure 49: Richter ceramic chart.

268

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

”I
x-

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

IIIIIIIIII

 

 

 

 

 

 

 

 

 

 

 

 

RICHTER SITE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

chad.

amic

Figure 49 (cont’d): Richter cer

269

 

Figure 50: Vessels from the Richter Site: upper rim linear stamped from Group
1.

270

 

Figure 51 : Vessels from the Richter Site: top left, dentate stamped from Group 2;
top right, linear stamped from Group 1; bottom, cord-wrapped stick stamped
from Group 3.

271

   

Figure 52: Vessels from the Richter Site: left, cord-marked from Group 5; right, a
smoothed outlier.

272

APPENDIX M

TABLE 26: VESSELS BY SITE FOR EACH SUPER-CLUSTER IN THE
LAUREL STUDY

Cluster A (Plain/Punctatelfissgj)

Fisk - 16, 17, 20, 21, 22

Long Sault - 1, 2, 3, 4, , 6, 7, 8, 9, 10, 11, 12,13, 46, 47, 50, 51, 52, 53, 55, 56,
58, 59, 60, 61, 62, 64, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 88, 89, 93 95
Ballynacree - 1, 2, 3, , 5, 6, 7, 53, 54, 81, 82, 97, 84, 85, 86, 87

Whitefish Island - 34

Cloudman - 24, 25

Naomikong East - 29, 39, 118, 119, 120

Gyftakis - 22, 26, 61

Naomikong West -104,105,106,107,109,110,111,112,113,114,115,116.
117

Portage - 29a-b, 30, 31, 32a-c, 33, 34, 35

luster mb
Gyftakis - 7
Clu t r D or -wra tick t m
Ballynacree - 45, 52, 80, 114
Portage - 39

Cluster E (Single-cord lmpressfl)
Ballynacree - 49, 50, 51, 79, 83, 112, 113, 115
Whitefish Island - 7

Naomikong East - 9, 31

Portage - 51, 52, 53, 54, 55, 56, 57

Cluster F (Trailed)
Cloudman - 23

Naomikong East - 25, 43, 160
Gyftakis - 63

Cluster G (Incised)

Long Sault. - 45

Ballynacree - 8, 44, 46, 47, 48, 73, 75, 88

Whitefish Island - 56

Cloudman - 8, 27, 28

Naomikong East - 35, 95, 123, 128, 132, 140, 146, 161
Gyftakis - 41

273

 

Naomikong West - 26, 27, 125, 129, 136, 148
Portage - 37, 38, 42

CltrHUn nk P d- ll hll tam

Fisk- 1, 2, 3, 4

Long Sault - 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 27, 28, 29, 30, 31, 32, 33, 34,
35, 36, 42, 44, 77, 78

Ballynacree - 33, 72, 95, 96, 109

Whitefish Island - 42, 43, 49, 50, 51, 73, 74, 75, 77

Cloudman - 15, 16, 17, 18

Naomikong East - 4, 10, 12, 15, 17, 20, 48, 58, 60, 75, 85, 90, 91, 94, 100, 141
Gyftakis - 4, 16, 18, 19, 29, 32, 39, 48, 58

Naomikong West - 33, 69, 99

Cluster l (Bankg: Pseudo-Scallop Shell Stampfi)
Whitefish Island - 14, 31, 33, 41

Cloudman -10,11,12,13, 14

Naomikong East - 14, 19, 50, 53, 55, 57, 62, 70, 71, 72, 73, 76, 83, 98, 103, 108
Gyftakis - 24, 38, 47, 55, 59

Naomikong West - 16, 54, 81, 82, 84

Portage - 58, 59

Clust r J Dr ed nk P o-Sc llo Sh II t m

Fisk - 5, 6, 7, 8, 9

Naomikong East - 1, 2, 3, 45, 63, 64, 65, 67, 79, 80, 133, 134, 163
Gyftakis-13, 14, 17

Naomikong West - 61

Cluster K (Unbanked Dentate Stamped)

Long Sault - 26, 40, 81, 82, 84

Ballynacree - 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 30, 31, 32,
39, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65,-66, 68, 69, 70, 77, 89, 90, 93, 98,
99, 100, 101,102, 104,106, 107, 108,110

Whitefish Island - 30, 37, 76

Cloudman - 19, 22, 26, 29

Naomikong East - 66, 152

Gyftakis -11, 12, 23,31

Naomikong West - 87

Cluster L (Banked Dentate Stampfi)

Fisk - 23

Long Sault - 14, 15, 37

Ballynacree - 27, 34, 35, 38, 67, 71, 91, 92

Whitefish Island - 72, 78

Cloudman - 20, 21

Naomikong East - 32, 78, 88, 124, 126

Gyftakis - 1, 2, 3, 5, 8, 9, 10, 25, 27, 30, 35, 36, 37, 44, 50, 56, 62

274

 

Naomikong West - 52, 92
Portage - 46, 47, 48, 49

Cluster M (Unbanked Linear stamped)
Fisk - 10, 11

Long Sault - 41, 43

Cloudman - 9

Naomikong East - 18

Gyftakis - 6, 28, 45

Naomikong West - 42, 127, 144

Cluster N (Banked Linear stamped)

Ballynacree - 40, 41, 42, 43, 76, 78, 103, 111

Whitefish Island - 1, 2, 3, 4, 5, 8, 9, 10,11,12, 13.15.1617, 18, 19, 20, 21, 22,
32, 35, 38, 39, 44, 46, 58, 60, 61, 62, 65, 66, 67, 68, 69, 7O

Cloudman - 1, 2, 4, 6, 7

Naomikong East - 13, 21, 22, 47, 51, 86, 96, 97, 139, 162

Gyftakis - 33, 34, 43, 57, 60

Naomikong West - 49, 143

Portage - 50, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69

Cluster Q (Dragged Benked Linear Stempfi)

Fisk -12, 13, 14

Long Sault - 38, 39

Ballynacree - 36, 37

Whitefish Island - 6, 23, 24, 25, 26, 27, 28, 29, 36, 40, 45, 47, 48, 52, 53, 54, 55,
57, 59, 63, 64, 71, 79, 8O

Cloudman - 3, 5

Naomikong East - 5, 8, 24, 89, 122, 131, 135, 137, 145, 150, 164, 165
Gyftakis - 42, 46, 49, 53, 54, 64

Naomikong West - 28, 38, 40, 44, 74, 138, 153, 154

Portage - 36

Cluster P (Decorated on the Upper Rim Only)
Fisk-15, 18, 19

Ballynacree - 9, 10, 11, 74, 94
Naomikong East - 6, 7, 11, 23, 36, 37, 93, 121, 149, 158
Gyftakis - 15, 21, 40

275

APPENDIX N

TABLE 27: VESSEL CLUSTERS FOR THE RICHTER AND SQUIRREL DAM
SITES BASED ON THE LAUREL SUPER-CLUSTERS

 

 

Richter Squirrel Dam
A. Plain 4 (10.5) 1 (3.0)
B. Corded 7 (18.4) 5 (15.2)
C. Brushed 1 (2.6) 2 (6.1)
D. CWS. 4 (10.5) 4 (12.1)
E. Single Cord - 2 (6.1)
F. Trailed - 9 (27.3)
G. Incised 4 (10.5) 7 (21.2)
H. PSS 3 (7.9) -
l. Ba. PSS 1 (2.6) 1 (3.0)
J. Dr. Ba. PSS - -
K. Dentate 1 (2.6) 1 (3.0)
L. Ba. Dentate - -
M. Linear 7 (18.4) -
N. Ba. Linear - 1 (3.0)
0. Dr. Ba. Linear - -

P.Upper Rim Only 6 (15.8)

 

 

Totals 38 (99.8) 33 (100)

 

CWS = cord-wrapped stick; PSS = pseudo-scallop shell; Ba. = banked;
Dr. = dragged.

276

APPENDIX 0
THE CLUSTER TREE FOR THE WHITEFISH ISLAND SITE:
A SAMPLE OF THE CLUSTER ANALYSIS USED IN THE STUDY

277

 

 

 

 

 

 

no
.5
3H
n:
n;
EN
3..
n5
.8
3
E
3
a

cluster (F12 to F14) are banked, dragged linear stamped vessels. the next
cluster (F 16, F17, F21, F22) are plain vessels with one row of exterior
punctates/bosses. Vessel F20 has two rows of punctates and bosses. The
cluster of F1 to F4 is a group of unbanked vertical/oblique pseudo-scallop shell
stamps over horizontal pseudo—scallop shell stamps. The cluster of F15, F18

Figure 53: The cluster tree for the vessels from the Fisk Site. The left cluster
(Vessels F5 to F9) are banked, dragged pseudo-scallop shell vessels. The next

F10 and F11 has unbanked linear stamps on the exterior upper rim. The outlier

and F19 has unbanked punctates on the exterior upper rim and the cluster of
(F 23) is dentate stamped.

278