The Florida anthropologist

Material Information

The Florida anthropologist
Abbreviated Title:
Fla. anthropol.
Florida Anthropological Society
Place of Publication:
Florida Anthropological Society.
Publication Date:
Quarterly[<Mar. 1975- >]
Two no. a year[ FORMER 1948-]
Physical Description:
v. : ill. ; 24 cm.


Subjects / Keywords:
Indians of North America -- Antiquities -- Periodicals -- Florida ( lcsh )
Antiquities -- Periodicals -- Florida ( lcsh )
serial ( sobekcm )
periodical ( marcgt )


Contains papers of the Annual Conference on Historic Site Archeology.
Dates or Sequential Designation:
v. 1- May 1948-

Record Information

Source Institution:
University of Florida
Holding Location:
Department of Special Collections and Area Studies, George A. Smathers Libraries, University of Florida
Rights Management:
Copyright Florida Anthropologist Society, Inc. Permission granted to University of Florida to digitize and display this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
01569447 ( OCLC )
56028409 ( LCCN )
0015-3893 ( ISSN )

Full Text
Florida Anthropological Society
Publications No. 11
i The Florida Anthropologist, Vol. 38, No.l-2,Pt.2 July 1985

THE FLORIDA ANTHROPOLOGIST 1b published quarterly by tne Florida Anthropological Society, Inc.,
p.O. Box 1013, Tallanassee, Florida 32302* Subscription is by membership in tne Society for
individuals, families and institutions interested in tne aims of the Society. Annual dues are
$12 (Individual), $18 (Family), $15 (Institutional), $25 (Sustaining), $100 (Patron) and $150
(Life). Foreign subscriptions are an additional $5 U.S. currency to cover added postage costs
for individual, family or institutional membersnip categories. Requests for information on tne
Society and membersnip application forms, as well as notifications of changes of address,
should be addressed to the Membership Secretary. Donations snould be sent to the Treasurer.
Requests for copies of the Editorial Policy and Style Guide (re: FA 37(1)), orders for back
issues, submissions of manuscripts for publication and notices of non-receipt or damaged issues
snould be sent to tne Editor. Newsletter items snould be sent to the President. Address
cnanges snould be made AT LEAST 30 days prior to the mailing of tne next issue. The Post Office
will not forward bulk rate mail.
Joan Deming
1839 Pine Cone Circle
Clearwater, FL 33520
M. Katherine Jones
#28 Bureau of Archaeological
Dept, of State, 4th Floor,
R.A. Gray Building
Tallanassee, FL 32301-8020
Harold Caldwell
1343 Woodbine Street
Daytona Beach, FL 32014
Elizabetn Horvatn
7404 12tn Street N.
Tampa, FL 33604
John F. Scarry
P.0. Box 1013
Tallahassee, FL 32302
AGENT: Joan Haas
1602 Alabama Drive, Apt. 304
Winter Park, FL 32789
(Three Years):
Jeffrey Mitchell
Florida State Museum
University of Florida
Gainesville, FL 32611
(Two Years):
William Goza
Florida State Museum
University of Florida
Gainesville, FL 32611
(One Year):
Mitchell Hope
111 Sunset Drive
Sebring, FL 33870
Louis D. Tesar
Route 1 Box 209-F
Quincy, FL 32351
Katnv Poppell
P.0. Box 1013
Tallahassee, FL 32302
Gandy Printers, Inc.
1800 South Monroe Street
Tallahassee, FL 32301
Robert S. Carr
Historic Preservation
Office of Community and
Economic Development
Warner Place-Suite 101
111 SW Fifth Avenue
Miami, FL 33130
George M. Luer
3222 Old Oak Drive
Sarasota, FL 333579
John W. Griffin
Route 5, Box 19
James J. Miller
Division of Archives,
History 6 Records
Department of State
The Capitol
Tallahassee, FL 32301-8020
William H. Marquardt
129 Florida State Museum
University of Florida
Gainesville, FL 32611
John F. Scarry
Division of Archives,
History 4 Records
Department of State
The Capitol
Morgan R. Crook, Jr.
Department of Sociology
and Anthropology
West Georgia College
Carrollton, GA 30118
St. Augustine, FL 32084 Tallahassee, FL 32301-8020
COVER ILLUSTRATIONS: (FRONT) Field notebook, pot and point. Record keeping is an
important part of archaeological surveys. Illustration of pot after the "Little
Pot" found at site 9Se9o during a survey conducted by Nancy White. (Illustration
courtesy of Louis D. Tesar). (BACK) Archaeological site location wanted poster.
Archaeological sites are historically significant, non-renewable cultural resources.
It is important that site location information be recorded for use in preparing
settlement pattern models and historic preservation plans. Without such informa
tion many sites are lost to development, agricultural practices and erosion. You
can help protact these sites by reporting site location information to the Florida
Department of State, Division of Archives, History and Records Management in
Tallahassee or to your local FAS Chapter representative who can do so. (Poster
design and illustration by Louis D. Tesar).

Papers from a Symposium given at the 1984 Southeastern Archaeological
Conference in Pensacola, Florida.
Editor's Page 94
Guest Editor's Preface to Archaeology of Northwest Florida and
Adjacent Borderlands: Current Research Problems and
Approaches Nancy Marie White 95
Prehistoric Settlement and Subsistence Systems of the Apalachicola
National Forest, Florida Sandra Jo Forney 98
PIXE Analysis of Gulf Islands Ceramics Glen H. Doran 104
The Deptford to Santa Rosa/Swift Creek Transition in the Florida
Panhandle Prentice M. Thomas, Jr. and L. Janice Campbell 110
Archaeological Resource Management Plan in Pensacola, Florida, and
the Hawkshaw Project Judith A. Bense 120
Shellfish Utilization During Deptford and Mississippian Times in
Escambia Bay, Florida Cheryl Claassen 124
An Archaeological Study on the Conecuh Drainage Ned J. Jenkins
and Cyril B. Mann 136
The Pensacola Variant and the Bottle Creek Phase Noel R. Stowe 144
The Bear Point Phase of the Pensacola Variant: The Protohistoric
Period in Southwest Alabama Richard S. Fuller 150
"Willey-Nilly" or the Archaeology of Northwest Florida and Adjacent
Borderlands Revisited David S. Brose 156
Nomenclature and Interpretation in Borderland Chronology : A
Critical Overview of Northwest Florida Prehistory Na(ncv White. 163
Discussion and Comments Jerald T. Milanich 175
Comments on the Archaeology of Northwest Florida in 1984 Gordon
R. Willey 178
Comments on Archaeology of Northwest Florida and adjacent Borderlands:
Current Research Problems and Approaches Louis D. Tesar 184
References Cited 187
BOOK REVIEW: Final Report of the United States De Soto Commission
by John R. Swanton, Classics of Smithsonian Anthropology
Reprint, reviewed by Louis D. Tesar 194

Tnis issue is of special interest to me
for three reasons. First, it is a
pleasure to bring you the first Florida
Anthropological Society Publication
since 197 This is our irregular
series, and it depends on grants,
donations and back issues sales and for
its publication, as membership dues are
used to cover our quarterly issues of
The Florida Anthropologist and other
Society activities. Funding for this
issue was obtained by Nancy Marie
White, our Guest Editor, through a
faculty grant from the University of
South Florida President's Council.
Second, the subject area, "Archaeology
of Northwest Florida and Adjacent
Borderlands," has been one of continu
ing research interest to me (cf. Tesar
1965, 1973a, 1973b, 1974, 1976, 1977,
1979, 1980a, 1980b, 1980c, 1981). As
such, at Nancy's suggestion, I have
added my comments at the end of this
issue to those of Millanich and Willey.
Finally, as one of many professionals
working on efforts to update Florida's
Comprehensive Historic Preservation
Plan, and recognizing the need to
coordinate such efforts with those of
adjacent states to achieve concensus on
our efforts m the borderlands, I
welcome such issues as this one, as
well as the preceding sixteenth century
St. Augustine issue and the forthcoming
Ft. Walton ceramics issue, as which
will assist in these efforts.
I need not repeat the introduction to
the papers contained in this issue, as
they are presented by our Guest Editor
on the following pages. Please note
the references cited in the articles in
this issue have been combined to save
space and avoid duplication.
Putting The Florida Anthropologist
together has sometimes been considered
akin to magic. While certainly not
Vol. 38 Numbers 1-2 Part 2 THE FLORIDA
claiming to be magician, it is an
experience which is difficult to
describe to the uninitiated. Never
theless Frank Gilson's illustration
seem appropriate. Hope you enjoy this
Louis D. Tesar
June 1, 1985

This volume presents some of the papers
given at a symposium at the forty-first
annual meeting of the Southeastern
Archaeological Conference, on November
8, 1984 in Pensacola, Florida. The
symposium topic was current research
and approaches in the archaeology of
northwest Florida and adjacent border
lands, an area encompassing southwest
Georgia and south Alabama, in addition
to the entire Florida panhandle west
ward from the Aucilla River located
east of Tallahassee, (see map). This
area is very important in prehistory,
and not only because of certain techno
logically and aesthetically superior
artifacts, or great population densi
ties through time, or evolution here of
what were apparently truly complex
societies for the Eastern U.S.
More and more archaeological survey and
excavation are taking place lately in
this area, though often little gets
published outside the limited distribu
tion of cultural resource management
(CRM) reports or research papers given
at meetings. This collection runs the
gamut from pure description to extended
critique, and shows how far we have and
have not come in the forty years since
one of the first great regional synthe
ses in eastern U.S. archaeology was
done here by Willey (1949).
In so vast a region many large areas
have not yet been well examined.
Forney's paper deals with an interior
area, the Apalachicola National Forest,
away from both the Gulf and major
rivers. This type of area, according
to earlier models of prehistoric
settlement, was thought to be of
little use to most aboriginal cultures.
Her data suggest re-evaluation of that
model. Sites of many time periods now
known from the pine flatwoods indicate
differing intensities and several possi
ble functions of occupations through
Jenkins and Mann explore cultural chro
nology and settlement in the Conecuh
Valley of south Alabama, another little
known area. Their description of their
finds suggests a gap in the record,
during the Early Woodland.
Data from excavated sites continue to
clarify the particular picture for
specific time periods. The papers by
Stowe and Fuller define two phases from
the latest indigenous cultures in the
far western panhandle and south Alabam-
a, the late prehistoric and historic
Pensacola variants of Fort Walton
culture. Thomas and Campbell describe
a Deptford site in Fort Walton Beach
where they obtained materials charac
teristic of a transition into Santa
Rosa/Swift Creek, suggesting the
picture for Early Woodland to be less
simple and clear cut than previously
A Deptford occupation on Pensacola Bay,
with an adjacent historic settlement,
was the focus of Bense's investiga
tions. However, her report is signifi
cant primarily for its description of
a local effort to involve city govern
ment in cultural resources management.
Patterns of utilization of a subsis
tence resource through time are ex
plored by Claassen. Her study of
several shell middens around Escambia
Bay is important, and not only for
showing interesting differences in
shellfish exploitation manifested from
Deptford to Mississippian times. It
also suggests the potential of
Vol. 38 Numbers 1-2 Part 2 THE FLORIDA ANTHROPOLOGIST July, 1985


archaeology to investigate the conse
quences of human harvesting on a
specific natural resource.
Doran's study applies mid-eighties
technology to traditional archaeologi
cal problems of sorting ceramics, in
this case from sites near Pensacola.
The potential of such new methods to
help refine or even redefine old
categories is exciting.
Critiques are presented by both Brose
and White. Brose explores in detail
the original meanings of our archaeo
logical constructs known as Deptford,
Swift Creek, Weeden Island, and Fort
Walton, and what they have lately come
to mean. In light of the large body of
new evidence, not to mention new appro
aches, he suggests significant altera
tions in our chronological/explanatory
frameworks. My own paper is a summary
critique of general research trends in
the area today, and suggests some more
fruitful directions for investigation,
different research questions or strate
gies for our next foci of attention.
Already known for his recent overall
synthesis in Florida archaeology,
Milanich provides thoughtful commentary
on the papers and several more valuable
suggestions for the researchers.
Finally, we are honored to have in this
volume the worthy comments of Gordon R.
Willey. He has the perspective of one
who not only started it all, but also
has done more archaeology than any of
us, and has both explored and chroni
cled the evolution of our science
through the decades.
Many have assisted in the preparation
of this volume. Some papers from the
original symposium were not able to be
included here, but I thank all the
participants. Archaeology graduate
students Susan Henefield and Jill
Halchin of the University of South
Florida provided considerable assis
tance in organizing and aiding
preparation of the manuscripts submit
ted for final editing.
Whether or not one agrees with all the
authors' approaches and interpreta
tions, getting the data and analyses
into print is always a task of some
difficulty, but of paramount impor
tance. The University of South Florida
President's Council is deserving of
great thanks for supporting my endeavor
to compile the latest research and make
it available in this publication.
Nancy Marie White
March 15, 1985

Sandra Jo Forney
U.S. Forest Service, Tallahassee
Previous researchers have made some
general statements asserting the lack
of a continuous prehistoric occupation
in the pine flatwoods regions of
northwest Florida (Brose and Percy
1978; Larson 1980). When viewed in
comparison with other optimal local
environments, the pine flatwoods has
been described as an inhospitable place
for permanent habitation due to dense
vegetation and wetness of the ground
throughout much of the year (Brose and
Percy 1978:84). This conclusion may
have been guided by a somewhat ethno
centric bias based on historic western
concepts of what constitutes an ade
quate environment for human occupation.
Prehistoric populations certainly
selected those locations most endowed
with environmental diversity and with
the potential to yield the maximum
amount of usable resources. Research
in these areas, however, often leads to
explanations of cultural process and
the development of cultural reconstruc
tion based on ideal situations or on
locations central to cultural develop
ment. Although these centers of cul
tural development by all means must be
studied, smaller sites representing the
socio/economic maintenance of popu
lations must also be interpreted. To
accomplish this task, sites in areas
considered "unfit for human occupation
should be studied.
The pine flatwoods of the Apalachicola
National Forest is a low energy envi
ronment compared to coastal and river
ine locations. Archaeological surveys
in the pine flatwoods, have revealed
that this environment is sufficient to
maintain an extensive if not intensive
prehistoric population, as indicated by
the presence of Archaic through Late
Woodland and possibly early Fort Walton
A closer look at the micro-environments
of the region in the Apalachicola
National Forest provides a better in
sight into the distinct cultural pat
terns that exist in the pine flatwoods.
The Apalachicola National Forest,
encompassing 328,316 ha (588,871
acres), is located within four counties
of northwest Florida (Figure 1). Ten
percent of the Forest is contained
within the Karst Plain regions of the
Woodville and Lake Munson Hills physio
graphic units. These areas contain
predominately high pineland vegetation
such as longleaf pine and a variety of
scrub oaks and southern mixed hardwood
forests. Solution depressions or sink
holes and lakes are common features of
the landscape. Evidence of continuous
prehistoric occupation in this area of
the Forest has been previously demon
strated (Forney 1983).
Approximately 90% of the Forest is
located in the Apalachicola Delta
District (formerly included in the Gulf
Coastal Lowlands) a physiographic unit
characterized by slash and longleaf
pine flatwoods. These nearly level to
convex ridges are interspersed with
titi, cypress, blackgum and bay swamps
which drain into six major watersheds.
The pine flatwoods of the coastal
lowlands experience moderate fluctua
tions in the water table throughout the
year, suggesting the possibilities of
seasonal foraging and migration during
prehistory. As with agricultural acti
vities, the seasonal intensity of
July, 1985

5b 1 Lake Munson Hills 8c Delta Plain
5b2 Woodville Karst 8d Coastal Strip
8a3 Hosford Delta
Figure 1. The Apalachicola National Forest and its physio
graphic divisions.
Apalachicola National Forest
Norwood .5%
Swift Creek .5%
Deptford 1%
Ft. Walton
Figure 2. Prehistoric sites in the Apalachicola National Forest:
relative frequencies of different types.

hunting pursuits may have been
dependent upon these water levels.
According to a recently completed soil
survey of the Apalachicola National
Forest, 31% of the Forest contains soil
types considered conducive to small
scale horticulture. These nine soil
types include somewhat poorly drained,
moderately well drained, and well
drained soil classes (Hart 1980; Waite
1984).The flatwoods could supply small
semi-permanent villages with a variety
of natural resources. This environment
under natural conditions was constantly
burned by natural or wild fire, allow
ing the maximum amount of new under
story growth to sustain herds of deer
and flocks of turkey. In addition to
supplying browse and mast for deer,
fruit- and nut-producing species
located in the hardwood hammocks and
scrub oak ridges also yielded foods to
supplement a diet based on cultivated
plants. Saw palmetto fruits, the
second most important source of food
for deer in Florida during the fall,
are abundant in the pine flatwoods.
The surrounding bays and swamps provid
ed additional resources such as turtle,
bear and other faunal and floral
The Inventory
During the past seven years, approxi
mately 7% of the Apalachicola National
Forest has been inventoried for cultur
al resources prior to land disturbing
activities conducted by the U.S. Forest
Service. Of 267 cultural resources
recorded for the area, 72% are prehis
toric and 28% are historic. This
averages one site location per 160
Evaluation of previous archaeological
research in the north Florida pine
flatwoods was of little value in the
initial planning of the survey strate
gy. Prior to 1977, only 20 prehistoric
site locations were recorded in the
Apalachicola National Forest.
These sites include the Brickyard Creek
Weeden Island Mound, initially located
by Moore (1903) and later described by
Willey (1949), and 19 prehistoric
properties recorded during the Florida
Department of State, Division of
Archives, History and Records
Management survey conducted within the
proposed Sopchoppy River Wilderness
Area (Chance 1977). All surveyed areas
were located within riverine, non-
flatwoods environments. Most of the
more recent inventory work has been
conducted within the pine flatwoods
region of the Forest, because the
majority of the impact-related
mechanical site preparation activities
occur in that area. Certain methodo
logical and interpretative problems
occur when determining if a statisti
cally viable sample resulted from the
inventory. Site preparation areas are
arbitrarily bounded units that usually
do not correspond to cultural or
natural divisions. They are determined
merely on the basis of stand vegetative
age, and may be considered anything but
random from an archaeological stand
point .
On the other hand, investigations con
ducted prior to proposed road recon
struction and construction tend to
traverse upland ridges as well as swamp
creeks. This provides a stratified
transect sample of distinct natural
zones to supplement the bulk of the
units sampled from the reforestation
areas of the Forest's pine flatwoods.
Current emphasis of U.S. Forest Service
cultural resource programs is inventory
and preservation of located sites
considered potentially eligible for
nomination to the National Register of
Historic Places. This overall direc
tion results in a lack of data recovery
or site impact mitigation often associ
ated with archaeological research on
federal lands. Preliminary results of
the ongoing study in the Apalachicola
National Forest, therefore, do not
include excavation data such as site
(38, 1985)

size, the presence of features, faunal
or floral remains. Obviously, no
firmly established inter- and intra
site comparisons could be made on the
basis of these data.
Results and Interpretations
All prehistoric cultural manifestations
of the Northwest Gulf Coast Tradition
are represented in the Apalchicola
National Forest pine flatwoods.
Although no bona fide in situ Paleo-
Indian sites are recorded within the
flatwoods, a few points from this
period have been recognized in local
collections. Evidence of Early and
Middle Archaic period occupations is
also limited, indicating little beyond
the presence of such occupations in the
Forest. Although there is only one
single-component Late Archaic Norwood
site in the sample, all 12 of the
multi-component sites recorded contain
evidence of Norwood. Other components
at these sites include Swift Creek and
Weeden Island. All the multi-component
sites are located on arable soils,
perhaps suggesting a shift toward small
scale plant cultivation.
The pine flatwoods may not have been
attractive for intensive long term
habitation during the Early Woodland
Period. Only one site revealed evi
dence of a Deptford period occupation.
In addition to Swift Creek components
recorded for the multi-component sites,
only one single-component Swift Creek
site has been located in the flatwoods.
These interpretations may, however, be
unjustly founded on a biased sample.
Thirty-six percent of the known sites
in the Forest's pine flatwoods are
Weeden Island (Figure 2). In previous
Weeden Island research, emphasis has
been on coastal and riverine locations
(Willey 1949; Percy 1971; Percy and
Brose 1974; Milanich 1974). In addi
tion to emphasis on areas which may be
considered high energy locations,
research has often focused on larger
sites of a ceremonial nature (Kohler
1975, 1978). Little work has been
published concerning smaller mainte
nance sites. As a result, Weeden
Island has been considered a ceremonial
society based on horticulture.
Weeden Island sites in the Apalachicola
National Forest pine flatwoods are
characterized by assemblages containing
Wakulla Check-Stamped ceramics and
incised and punctated wares, such as
Carabelle Punctate and Incised and
Keith Incised, suggesting that these
sites date from Weeden Island III to
Weeden Island V (See Percy and Brose
Most of these sites are situated within
or adjacent to areas containing moder
ately well drained soils. In addition
to pine flatwoods, Weeden Island sites
are also recorded in a variety of
environmental settings particularly
those adjacent to lakes, ponds, rivers,
and swamps.
The settlement patterns during the Late
Woodland probably involved small
extended family groups practicing
shifting cultivation and supplementing
horticultural produce with extensive
use of limited faunal and floral
resources from flatwoods. Site loca
tions were, no doubt, selected for
access to the few small areas of
tillable soils in the individual
drainage systems throughout the Forest,
in addition to providing access to the
flatwoods resources. People may have
occupied these sites for several years
and then abandoned them, returning
after the recovery of soil fertility.
At this preliminary stage of inventory
in the Apalachicola National Forest,
the cultural adaptation appears to be
similar to those of the previously
defined Wakulla Weeden Island peoples,
who exploited the river valley hardwood
forests and the small streams draining
into the Apalachicola River (Milanich

and Fairbanks 1980:127).
As may be expected, lithic scatters,
accounting for 40% of the known sites,
are also prevalent in the Forest.
These sites contain predominately non
decortication flakes and an occasional
unifacial tool. More of these sites
were located on wetter, less arable
soils than were ceramic bearing sites.
Certain lithic scatters may represent
remains from pre-ceramic groups using
the Forest for intermittent hunting
expeditions. Then again, they may be
Late Woodland lithic debris from activ
ities performed away from the hearth,
such as butchering.
One major research question to consider
in the Apalachicola National Forest is
why and how did the horticultural,
hunting-and-gathering Weeden Island
peoples utilize the pine flatwoods?
Why were the pine flatwoods occupied if
other local ecosystems were more effi
Several alternative hypotheses may be
considered. (1) These potential farm
steads may represent sites utilized to
complement the riverine subsistence
system. Interpretations concerning the
existence of village habitations based
on limited informal shovel tests and
surface-collected data may be presump
tuous. (2) These sites could be farm
steads of the riverine Weeden Island
peoples, utilizing the flatwoods as an
overflow area during times of popula
tion increases. A more dispersed
settlement system was needed to support
increased population as well as the
increased dependency upon small scale
horticulture during this time. Earlier
occupants of the flatwoods, lacking
large populations and dependence upon
plant cultivation, probably did not
need to migrate beyond the choicest
locations. Changes in socio-political
motives during the Weeden Island phase
may have also created a need for such
an extensive distribution of sites.
(3) Although materially these sites
show Weeden Island traits, they may not
have actually participated in the
"Greater Weeden Island Culture Sphere."
Given time, cultural groups living in
small drainage systems, such as those
in the pine flatwoods, often develop
their own subtle variations.
A number of site locations, 9% of the
total sample, yielded grit-tempered
ceramics, both plain and incised.
Forty-five percent of these sites also
contained debitage and possible tools
in association with the ceramics. All
are located on soils considered to be
arable for small scale horticulture.
Initially defined by Willey and
Woodbury, grit-tempered pottery, more
specifically Lake Jackson Plain, is
considered a marker of the Fort Walton
culture (Milanich and Fairbanks 1980;
Willey and Woodbury 1942).
An additional question to be considered
in the Apalachicola National Forest
concerns the nature of these ceramic
types: Is the presence of grit tem
pered wares indicative of Early Fort
Walton groups developing in the flat-
woods as they were in the riverine
areas to the north?
It has been speculated that the earli
est manifestations of adaptations
toward Mississippian socio-economic
patterns would occur in those regions
of northwest Florida where late Weeden
Island populations practicing short-
fallow swidden horticulture expanded
into diverse but ecologically or
socially restricted areas. These
criteria are met in major river valleys
as well as the smaller tributaries of
the Apalachicola River (Brose and Percy
As in the Apalachicola River Valley,
Fort Walton settlement in the pine
flatwoods may be early. Sites in the
Forest yielding grit-tempered wares may
represent the activities of semi-
sedentary communities prior to the
inception of full scale permanent late
Fort Walton phase agriculture. Site
(38, 1985)

location favors varied resource pro
curement carried out during the early
Fort Walton period, but does not easily
support activities associated with Fort
Walton agricultural communities. These
later groups probably moved inland (to
the Tallahassee Red Hills area), as did
those descendants of the Wakulla Weeden
Island peoples in the upper
Apalachicola River Valley.
As horticulture became more important
to the total subsistence pattern of the
post-Weeden Island Mississippian
culture, it was no longer as necessary
to move seasonally to harvest nuts and
hunt deer. With little need to supple
ment diet with hunting and gathering,
there was a decreasing need to exploit
the pine flatwoods region which, as
mentioned earlier, was certainly not
capable of supporting full scale
Prehistoric settlement and subsistence
in the north Florida pine flatwoods
displays both temporal and functional
variability. The apparent paucity of
basically agricultural sites of the
late Fort Walton period and the pres
ence of all other northwest Florida
Gulf Coast cultural manifestations
strengthens the hypothesis that the
pine flatwoods region was intermittent
ly occupied by distinctly hunting and
gathering groups and at least semi
permanently occupied by those groups
which supplemented a horticultural
economy with hunting and gathering.
Sandra Jo Forney
USDA/Forest Service
Suite 4061
227 North Bronough Street
Tallahassee, Florida 32301

Glen H. Doran
Department of Anthropology
Florida State University
PIXE (proton induced x-ray emission) is
a technique of elemental analysis
employing a high energy beam of protons
directed at a target sample. The
proton beam strikes the sample and
excites the electrons in the constit
uent atoms of the target sample. When
excited, the electrons leave their
normal shell orbits and emit detectable
x-rays of specific spectral frequen
cies. Each element exhibits a charac
teristic x-ray spectrum. The x-ray
counts are statistically proportional
to the quantity of each element present
(Cullity 1978). Most PIXE facilities
monitor the emitted spectra with a
multichannel analyzer capable of simul
taneously monitoring a wide spectra of
x-rays. A computer monitors and sorts
the record of emitted spectra and then
fits the data to a set of calibrated
The procedure used here to analyze Gulf
Islands prehistoric ceramics provided
data on calcium, chlorine, nickel, sil
icon, sulphur, vanadium, sodium, magne
sium, aluminum, potassium, phosphorus,
titanium, copper, zinc, iron and manga
nese. Few specimens contained copper
or zinc, and these elements were
eliminated from statistical analysis.
Contrasting the cost of PIXE to other
elemental analysis procedures is
worthwhile. Neutron activation and
atomic absorption are undoubtedly more
accurate. Unfortunately, the cost for
these techniques can exceed $450.00 per
sample (General Activation Analysis
1984 Price Schedule). Additionally,
scans or identification of specific
elements must be requested prior to
submission. This is a dramatic con
trast to the highly cost-effective PIXE
procedures. The estimated PIXE cost at
Florida State University (FSU) is
between $10.00 and $20.00 per sample.
Simultaneous analysis of a wide spec
trum of elements is possible without
the necessity of prior elemental
selection. The cost/benefit ratio for
PIXE analysis is very favorable for
archaeological applications.
Dr. William Nelson and his staff have
been developing the FSU PIXE facility
for the last four years and have ana
lyzed over 25,000 samples. The bulk of
the lab's analysis has focused on
aerosol pollutants (Albert, Leslie and
Winchester 1978) and substances
introduced into the atmosphere by lo
calized disturbances such as factories
and missile launches from Cape
Canaveral. Prior to 1982 the lab's
involvement with archaeological samples
was very limited. Since 1982, however,
nearly 200 samples of prehistoric and
historic ceramics have been analyzed in
conjunction with the FSU Department of
Anthropology (Doran 1984a; Doran,
Houmere and Bauman 1985). This paper
summarizes portions of the research
pertaining to northwest Florida
Analysis Procedures
Several sample preparation procedures
have been used in PIXE studies of
archaeological materials. The most
common technique involves taking a
sample from a sherd, crushing it to a
uniform size and then pelletizing it
with a high pressure press (Carlsson
and Akselsson 1981). This approach
creates a uniformly dense pellet of
homogeneous composition. Arguably,
this procedure may be the best techni
que, but the time and increased cost of
sample preparation can strain
chronically tight archaeological
Vol. 38 Numbers 1-2 Part 2 THE FLORIDA ANTHROPOLOGIST July, 1985

budgets (approximately $50.00 a sample;
personal communication from William A.
Nelson, 1984). This is particularly
true where analysis of hundreds not
dozens of samples may be an analytical
A second approach submits the entire
unmodified artifact for analysis. Most
targets are placed in a vacuum chamber,
and this approach generally limits the
process to one sample analysis per
chamber "pump down." Again, this can
be time-consuming and costly. In some
cases samples can be analyzed external
to a vacuum chamber though the mounting
procedures would prove very time con
suming for larger numbers of samples
(Chen et al. 1980). The nondestructive
nature of these approaches makes them
very suitable for rare or exotic arti
facts which cannot be destroyed or have
samples removed from them.
A third approach, the one used here,
involves scoring a freshly exposed
sherd core section with a stainless
steel blade. The resultant powder
falls to the adhesive surface of a PIXE
sample ring. At the present time eight
sample rings can be placed in the
chamber. Chamber design is such that
each ring can be rotated into the
proton beam's path by a simple ratchet
drive. Thus for each vacuum evacuation
eight samples can be analyzed
The ultimate test of the effectiveness
of this preparation procedure is
whether or not sherds of different
types or from different locations can
be statistically separated. The re
sulting table of data is then submitted
for statistical evaluation. Since the
resulting data are proportional in
nature, percentages of elements can be
compared. PIXE studies often "stan
dardize" the data by expressing each
element proportion as a fraction of a
consistently occurring element. For
example, if all samples contained mea
surable quantities of aluminum, each
element is divided by the amount of
aluminum. Thus the ratio of element X
to aluminum can be compared between
samples. Silicon and aluminum were
consistently encountered in this study
and both ratios were employed in
statistical analysis.
Mississippian ceramics often have
appreciable quantities of shell
introduced as a temper element. To
avoid the undue influence of this
additive, specifically the elevated
calcium fractions, the elements were
expressed as a percentage with the
calcium fractions included and then
recalculated after excluding the
calcium percentage. All ratios were
calculated after excluding the calcium
Simple univariate analysis, as will be
shown, is of limited interpretative
use. Therefore, more complex multi
variate analysis was extensively emplo
yed. Discriminant function analysis
(DFA) was the statistical tool of
choice. DFA basically addresses the
question of whether two or more samples
(here sherd types or proveniences) can
statistically be separated. DFA essen
tially plots the respective elements
and then mathematically defines a line
which separates the maximum number of
specimens into the respective groups
(Nie et al. 1975).
The discriminant function is employed
in a manner similar to that of a re
gression equation. There is a constant
which is multiplied by the individual
sherd element value or element ratio
and the resulting value is used to
assign group membership. To test the
accuracy of assignment the DFA assign
ment is then compared to the known
assignment and the percentage of
accuracy is computed. The higher the
accuracy the more effective the DFA is
in its assignments. This process is
relatively straightforward with one
element. Multiple elements involve
plotting the values in a multi-

dimensional space, and the mathematics
becomes more complex though the prin
ciple is the same.
Sample Composition
To evaluate the PIXE approach with
archaeological materials from northwest
Florida, 47 sherds from three sites
near Pensacola, Florida were selected.
All specimens were from the FSU
Department of Anthropology's excava
tions in the Gulf Islands National
Seashore funded by the Southeastern
Archeological Center, National Park
Service, Tallahassee (Doran 1985a).
Twenty-four sherds from Butcherpen
Mound (8Sr29), a Weeden Island II site;
5 sherds from Plantation Hill West
(8Sr67), a Weeden Island II site; and,
18 sherds from Third Gulf Breeze
(8Sr8), a Santa Rosa-Swift Creek and
Pensacola site, were studied. The
sites are all within 5 km of each
other. The sample includes 7 Swift
Creek Complicated-Stamped sherds, 4
Alligator Bayou sherds, a Basin Bayou
sherd, a Franklin Plain sherd, a Santa
Rosa Punctated sherd, 5 Wakulla Check-
Stamped sherds, 2 Carabelle Incised
sherds, 2 Coles Creek-like sherds, 6
Weeden Island Plain rim sherds, a
Mississippian Plain sherd, 2 Moundville
Engraved sherds, and 15 unidentified
plain sherds.
The Mississippian Plain, Moundville
Engraved and the Coles Creek-like
sherds may be trade wares, but the rest
are presumably locally manufactured
wares. Analysis of local clay sources
has not been undertaken, but is planned
to provide more accurate data on the
distinction between trade wares and
locally manufactured materials.
Representative statistics for several
element percentages are presented in
the accompanying table to illustrate
the difficulty of relying on simple
percentages and, more particularly, the
difficulty of univariate statistics to
separate different sherd types or
sherds from different sites.
The aluminum mean values are fairly
consistent from sherd type to sherd
type, with only minor exceptions. The
mean of the Coles Creek-like wares is
low. Furthermore, this ware shows the
most restricted range, with only a 2%
difference between the minimum and
maximum values. The Santa Rosa-Swift
Creek Complicated-Stamped wares show
the greatest variation and produced
both the maximum high value and the
minimum low value reported. This is
possibly a reflection of the larger
sample size and not a reflection of
significant elemental differences. The
Weeden Island decorated rim and the
Moundville Engraved wares both exhibit
relatively restricted ranges and tend
to be on the low end of the total
aluminum range.
Chromium concentrations in general are
low and relatively uniform in the
sherds examined. The failure to
identify chromium in the Franklin Plain
sherd is probably also an anomaly of
the small sample size.
Concentrations of iron are character
istically in excess of 17%, though the
single samples of Franklin Plain and
Santa Rosa Punctate both fall below
this level. There is substantially
more variability in maximum values and
the Weeden Island decorated rims show a
maximum of 39.7% iron.
Manganese was not consistently encoun
tered in many of the sherds. In the
wares with larger sample sizes the
minimum and maximum values of the Swift
Creek Complicated-Stamped and the
Weeden Island decorated rim specimens
are almost identical. However, the
ranges of both of these groups do not
overlap the ranges of the unidentified
plain (UID) wares. It appears the UID
wares internally represent a cohesive
group. The Coles Creek-like sherd
(38, 1985)

Aluminum Chromium Iron
SC Comp.-Stmp.
Alligator Bayou
Basin Bayou
Franklin Plain
Santa Rosa Punct.
Wakulla Check-Stmp.
Carabelle Incised
Coles Creek-like
W. I. dec. rim
Mndvlle. Engrv.
UID. Plain
SC Comp.-Stmp.
Alligator Bayou
Basin Bayou
Franklin Plain
Santa Rosa Punct.
Wakulla Chk.-Stmp.
Carabelle Incised
Coles Creek-like
W. I. dec. rim
Mndvlle. Engrv.
UID. Plain
Table I. Representative Elemental Concentrations. Percentages were calculated after
first excluding the calcium amounts. Sample sizes vary depending on the number of sherds
exhibiting the elements. Where only min. and max. values are reported only 2 sherds
contained identifiable amounts. Single values (reported in the min. column) indicate only
one sherd exhibited the element.

which exhibited manganese is more like
the UID group than any of the other
samples examined.
The phosphorus data show little coher
ent pattern in this small sample. It
may or may not be significant that the
Swift Creek Complicated-Stamped and the
Alligator Bayou material showed the
highest maximum values of any of the
materials studied.
The Basin Bayou sherd shows a high per
centage of sulphur. This is in marked
contrast to the Swift Creek Compli-
cated-Stamped sherds, which show a
consistently low sulphur percentage.
Larger samples are needed to confirm
these observations.
The analysis of individual wares breaks
the total sample into the smallest
analytical groups. It also illustrates
the difficulty of interpreting uni
variate analysis of PIXE data. All
that can be observed are occasional
tantalizing hints of differences.
Univariate analysis of larger sub
samples, such as by site, show similar
analytical difficulties. Univariate
analysis of sherds from different
regions has only been informative in a
few cases where sherds from certain
areas or certain types revealed the
consistent absence of specific ele
ments. In general this phenemenon is
not commonly observed. Mulitvariate
analysis has been necessary for effec
tive separation (Doran 1985b).
Discriminant Function Analysis of
Ceramic Types
Discriminant function analysis (DFA) is
more effective with larger samples than
presented here. The elements or ele
ment ratios which were most useful in
the greatest number of separations
include iron, silicon, chromium,
aluminum, vanadium, magnesium, man
ganese, and titanium. With respect to
the element ratios or percentages,
elements ratioed to aluminum tended to
be more useful than the silicon ratios
or simple percentages.
DFA of the seven Swift Creek Compli-
cated-Stamped sherds and the five
Wakulla Check-Stamped sherds using only
the iron/silicon ratio was 91.67%
accurate. Clearly there are distinct
enough differences between these types
to provide accurate separation. Out of
the total of 12 sherds only 1 Wakulla
Check-Stamped sherd was misclassified
as a Swift Creek Complicated-Stamped
sherd. As a test, a second variable
was included in this analysis, the
magnesium/phosphorus ratio, and the
accuracy rate rose to 100%. Using this
many element ratios in samples this
small is statistically unwise but
indicates the potential of PIXE
analysis. If such consistent differ
ences can be observed in plain wares
from the two periods, the Santa
Rose/Swift Creek and the Weeden Island
periods respectively, such an obser
vation might be of considerable value.
A DFA of Wakulla Check-Stamped and the
Alligator Bayou wares was also per
formed. Between these two groups the
chromium-aluminium ratio proved to be
77% effective. Out of the total of
nine sherds only two were incorrectly
assigned to the wrong groups one
error in each sherd category. When the
discriminant function developed in the
comparison of Wakulla Check-Stamped
and Alligator Bayou sherds was applied
to the Weeden Island decorated rim
sherds all six were assigned to the
Wakulla Check-Stamped category. These
two Weeden Island wares thus showed a
high internal similarity in contrast to
the Alligator Bayou sherds.
A comparison between the Wakulla Check-
Stamped wares and the Weeden Island
decorated rim materials was accurate
90.9% of the time. Only one sherd, a
Wakulla Check-Stamped sherd, was in
correctly assigned. The most effective
element ratio in this separation was
(38, 1985)

the vanadium/silicon ratio.
A three-way comparison of the Wakulla
Check-Stamped, Alligator Bayou, and
Weeden Island decorated rim sherds was
accurate in 80% of the sherd ascrip
tions. Of the 15 sherds in the analy
sis only 3 were misassigned. One
Alligator Bayou and one Weeden Island
decorated rim sherd were assigned to
the Wakulla Check-Stamped category and
one of the Wakulla Check-Stamped speci
mens was incorrectly assigned to the
Alligator Bayou category. For this
separation the chromium/silicon ratio
was employed.
Intersite Comparisons
After excluding the possible nonlocal
wares already mentioned, comparison
between 8Sr8 and 8Sr29 sherds was
performed. The sample sizes from each
site were, respectively, 15 and 13.
DFA analysis was 92.8% accurate in
separating the sherds from these two
sites. Two element ratios and one
percentage were used to achieve this
degree of accuracy: the chro
mium/phosphorus percentage, the
chlorine percentage, and the magnes
ium/aluminum ratio. Inclusion of all
wares, regardless of the probable
origin was less accurate as might be
anticipated; only 74% accuracy was
attained. Twelve of the 42 sherds in
the analysis were incorrectly assigned.
For this analysis a totally different
set of element ratios proved most
valuable and included the zinc/aluminum
ratio, the magnesium/silicon ratio and
the titanium percentage calculated
without the calcium contribution.
PIXE analysis of northwest Florida
ceramics provided several types of
information. The simplified and cost
effective sample preparation procedure,
the "scratch and bombard" approach,
appears capable of producing analytical
results of potential archaeological
interest. Univariate analysis of
simple percentages is difficult to
interpret. Multivariate analysis
appears more appropriate for PIXE data.
In spite of the small sample sizes in
this preliminary analysis a variety of
differences could be observed. Differ
ences between wares of the same time
period and of different time periods
could be identified with relatively
high accuracy rates. Furthermore,
differences in elemental composition of
presumed locally manufactured ceramics
between sites could be observed, though
with a reduced degree of accuracy. The
degree of accuracy dropped when pre
sumed nonlocal wares were included.
Explanations for the differences
observed at this time can not be
offered without additional analysis of
larger samples. Possible sources of
the ceramic variation could lie in
differences in clay sources, manu
facturing processes, temper elements,
and use. Undoubtedly analysis of
extant clay sources near the sites
would help resolve some of these
It seems clear that PIXE analysis of
ceramics is a potential tool of arch
aeological interest. PIXE analysis of
other archaeological materials such as
human bone, copper, and cherts may
prove equally effective and certainly
seems warranted. PIXE facilities
around the world have analyzed a broad
range of archaeological materials
(Baijot-Stroobants and Bodart 1977;
Fontes et al. 1984; Demortier and
Hackens 1982; and others) but to the
best of my knowledge this is the first
time this simplified preparation pro
cedure has been employed, and the first
time northwest Florida ceramics have
been studied with PIXE.
Analysis funded by 1982 Committee for
Faculty Research Award No. 195001281.
Glen H. Doran
Department of Anthropology
G-24 Bellamy
Florida State University
Tallahassee, Florida 32306

Prentice M. Thomas, Jr. and L. Janice Campbell
New World Research, Inc.
The Transition from Deptford to
Santa Rosa/Swift Creek
The nature of Deptford culture has been
widely discussed in the archaeological
literature, with attention paid not
only to characteristics related to
settlement, subsistence, and societal
composition, but also to when it began
and when it ended. Milanich (1973a)
suggested 500 B.C. as a beginning date
for the Deptford culture. However, the
Alligator Lake site, in Walton County,
yielded a radiocarbon date of 625 B.C.
(W. Lazarus 1965). That site, which
produced fiber tempered and Alexander
series sherds in addition to Deptford,
but in the absence of any Swift Creek
materials, clearly falls at the early
extreme of the Deptford temporal range.
Our concern in this paper is with the
end of Deptford and the nature of the
transition from Deptford to Santa
Rosa/Swift Creek.
Milanich (1973a) suggests a date of
between A.D. 200 and A.D. 300 for the
end of Deptford and the beginning of
Santa Rosa/Swift Creek. At the other
extreme, Knudsen (1979) places the
emergence of Santa Rosa/Swift Creek as
sometime between 300 B.C. and 200 B.C.,
a period well out of line with most
scholars' thinking. A middle ground of
around A.D. 0 has been suggested by
Percy (1974) and Tesar (1980a) for the
end of Deptford and the beginning of
Santa Rosa/Swift Creek. To our knowl
edge, the only absolute date available
for a Deptford site on the Florida
panhandle, prior to those presented in
this paper, is that for the Alligator
Lake site. It adds nothing to an
examination of when Deptford became
Santa Rosa/Swift Creek.
Another issue related to the end of
Deptford is how it developed into Santa
Rosa/Swift Creek. Because of continui
ty in the production of stamped ceram
ics, numerous writers such as Milanich
(1973a), Sears (1966), and Tesar
(1980a) have suggested that the
Deptford culture developed directly
into Santa Rosa/Swift Creek in the
Florida panhandle. In 1949, Willey
hinted that an actual tribal migration
from the Lower Mississippi Valley may
have occurred at the end of Deptford.
However, besides a passing comment and
the clear evidence of outside influence
in the material culture of northwest
Florida, Willey does not give serious
attention to the migration hypothesis.
Instead, he tends to concentrate on the
nature of the ceramic assemblages that
characterize Deptford and Santa
Rosa/Swift Creek.
The relationship between Deptford and
Santa Rosa/Swift Creek ceramic types
has strengthened the concept of
straight-line evolution. Sears
(1966:4-6) pointed to the Tucker site,
in Franklin County, Florida, as sup
port for a continuum between Deptford
and Swift Creek ceramics. He noted
that at Tucker, as well as at a second
site, Willey Brown, there were a
"number of odds and ends" associated
with the Deptford occupation that
suggested there was about to be a
change from Deptford to "something
else." For example, on rims there were
the beginnings of notching, better
associated with early Swift Creek
(Sears 1963).
Fairbanks and Keel (1966) seem to have
reached a similar conclusion about the
evolution from Deptford to Santa
Rosa/Swift Creek following the analysis
of their materials from the Elsom
mound. That site produced decorated
ceramics dominated by Deptford Bold
Check-Stamped with tetrapodal supports.
July, 1985

Another major type was Deptford Simple-
Stamped. There was also a small amount
of complicated-stamped wares that Keel
said were probably Swift Creek. The
ceramic assemblage led Fairbanks to
believe the site represented a late
Deptford population coming into contact
with Swift Creek influences originating
out of the northeast. In other words,
a culture in transition.
At the Mandeville site in south
Georgia, McMichael's preliminary
analysis identified both late Deptford
and early Swift Creek components.
However, Betty Smith's (1975) re
analysis of the Mandeville site
produced some rather sharp differences
in the site's interpretation. She
stated that there was little or no
difference between the complicated
stamped materials from the lower and
upper levels of the Middle Woodland
layers of Mound A. Smith found the
temper, paste and execution to be
identical for what McMichael called the
late Deptford and early Swift Creek
componenets; moreover, she observed
that the motifs remained generally
similar. As further support for a
continuum in ceramic tradition at
Mandeville, Smith pointed out that,
contrary to McMichael's analysis,
notched and scalloped rims, character
istic of early Swift Creek, represented
over 14% of the rims from the lower
level that had initially been
classified as Deptford.
Betty Smith concluded that the
Mandeville data clearly suggested an
uninterrupted continuum in the
occupations at the site. She also
noted that the addition of new ideas or
traits, though present, did not
materially affect the ceramic tradition
at the site. According to Smith (1975)
the problem came down to deciding
whether to call it late Deptford or
early Swift Creek. Because of the
dominance of early Swift Creek
materials, Smith opted for the latter
cultural designation and went on to
define an Early Swift Creek Complex
that included early Swift Creek Compli-
cated-Stamped, Crooked River Complicat-
ed-Stamped, Check-Stamped, Simple-
Stamped, Cord-Marked, and Plain. To
this list, she added the Santa Rosa
series ceramics found at sites in
northwest Florida.
We do not wish to reopen the question
of the interpretation of the Mandeville
site, but we certainly agree that
sherds of the three series do occur in
unmixed contexts, and for at least a
brief period of time, occur contempora
Finally, at the Fort Walton Temple
Mound, in Okaloosa County, Florida, the
sub-mound midden produced a mix of
ceramics that might be used as support
for a transition phase. Of the total
ceramics in the sub-mound midden,
Deptford types made up 57.7% (Y.
Lazarus 1970). Swift Creek ceramics
comprised only 2.7%, but Santa Rosa
decorated types represented 21.1% of
the collection. Polished plain wares,
which perhaps should be classed with
the Santa Rosa sherds, comprised 5.5%,
while other plain wares represented
The Pirates Bay Site
Understanding of the transition from
Deptford to Santa Rosa/Swift Creek is
complicated by the lack of good dates
on transition sites. We believe that
data to begin filling these gaps, at
least on the northwest Florida Gulf
Coast, have been derived from our
recent excavations at 80kl83, the
Pirate's Bay site.
The Pirate's Bay site is situated on
the north shore of Santa Rosa Sound in
Okaloosa County, Florida (Figure 1). A
surface collection made by the Fort
Walton Temple Mound Museum when the
site was first recorded included
materials diagnostic of Deptford, Santa
Rosa/Swift Creek, Weeden Island, and
Fort Walton periods; however, our



excavations have demonstrated the site
is dominated by Deptford and Santa
Rosa/Swift Creek remains. This is not
surprisingly in light of the site's
coastal setting adjacent to what was a
former stream and marsh, most of which
was filled in the early 1900s for
construction of several cottages. The
environmental characteristics of the
setting are identical to those reported
for coastal Deptford settlement (cf.
Milanich 1973a; Milanich and Fairbanks
Our excavations, conducted from Decem
ber 1983 to March 1984, were designed
to mitigate the adverse impact from
planned construction of a multi
structure condominium and marina
complex on the site (Figures 2 and 3).
The site contains an intact midden of
dark greasy sand with some shell and
numerous artifacts. The midden is
between 50 cm and 60 cm thick, but the
upper 25 cm have been disturbed by
plowing in several small areas and over
most of its area by the earlier housing
construction. The old development,
however, consisted of concrete slab
duplexes and only in a few areas on the
site have septic tanks or pipe trenches
extended below the 25 cm level (Figure
5). The extent of disturbance was
carefully delineated in the early
stages of excavation so that in full
data recovery we were able to separate
disturbed zones from intact deposits.
Below the midden, the excavations
uncovered 72 features (Figure 6), which
included postmolds, large refuse pits
with shell, bone and artifacts, small
pits, and shallow linear trenches
similar to those described by Milanich
(1973b) at a Deptford site on
Cumberland Island, Georgia. Only three
features produced any materials later
than Deptford or Santa Rosa/Swift
Creek. In these three features, six
Weeden Island sherds were found,
clearly an insignificant presence.
Small quantities of Weeden Island and
Fort Walton ceramics were recovered
from the excavated squares; however,
these are limited both horizontally and
vertically in their distribution. In
the western portion of the site, there
is an intact Weeden Island oyster shell
midden that overlies a Deptford-Santa
Rosa/Swift Creek dark sand midden, but
this late occupation is confined to
that portion of the site which was not
the focus of extensive data recovery
(Figure 4). In the eastern portion of
the site, where the block excavations
were located, no Weeden Island shell
midden occurs and materials assignable
to the Weeden Island or Fort Walton
periods all occur in the upper dis
turbed 25 cm of the midden.
It is in the analysis of materials from
undisturbed features and midden depos
its that the interpretation of the
Deptford to Santa Rosa/Swift Creek
transition, which we have designated
the Okaloosa Deptford phase, is based.
To simplify the presentation, we
combined individual sherd types into
the Deptford, Santa Rosa or Swift Creek
series. The Deptford series includes
Deptford Linear Check-Stamped, Deptford
Bold Check-Stamped, Deptford Simple-
Stamped, and Deptford Cross-Stamped.
Santa Rosa series types are Alligator
Bayou Stamped, Basin Bayou Incised,
Santa Rosa Stamped, Santa Rosa
Punctated, and a distinctive fine
paste, thin walled plain ware. These
plain sherds have been classed as Santa
Rosa because they have a paste
identical to that of some Santa Rosa
Stamped sherds at Pirate's Bay and are
easily sorted. The Swift Creek series
is comprised of Swift Creek
Complicated-Stamped, St. Andrews
Complicated-Stamped, New River
Complicated-Stamped, and West Florida
Cord-Marked. Although Crooked River
Complicated-Stamped is traditionally
included in the Swift Creek series, the
features from Pirate's Bay produced no
materials assignable to this type, so
it is omitted from consideration at
this time.
For the purposes of analysis, we
selected only those features which were

Site Map
Santa Rosa Sound
Santa Rosa Sound
(38, 1985)

Trench 4
East Profile
0 2 TP 6 TP 11




Main Excavation

— ——seees——j m

Trench 6 North Excavation

Trench 5


South Excavation



clearly undisturbed and yielded more
than 20 sherds. There were 20 such
features. All contained both Deptford
and Swift Creek ceramics, and all but
two produced sherds of the Santa Rosa
series. Not only is the co-occurrence
of these series unmistakable, but also
their relative frequencies are remarka-
bly consistent throughout both the
features and the midden at the site.In
the features, Deptford materials
comprise the majority, averaging 52.2%
of all ceramics. Swift Creek sherds,
predominantly Swift Creek Complicated-
Stamped or New River Complicated-
Stamped, are second in frequency,
representing and average of 13.64 of
the three series. Santa Rosa is third
in frequency, comprising an average of
7.6%, of the sherds. Plain sherds
represent an average of 9.0%, and the
remaining 17.6% unidentified obliter-
ated check- or other stamped sherds.

The consistency of series frequencies
across features was notable in itself,
but examination of materials from
undisturbed midden revealed similar
ceramic percentages in the test units.
Deptford series materials were best
represented, with an average of 39.1%
in the midden, followed by Swift Creek
at 8.9% and Santa Rosa at 7-02.
Although the series frequencies follow
the same relative trends as in the
features, the percentages for Deptford
and Swift Creek are lower because in
the midden the materials are more
weathered, so that there is a larger
number of obliterated check- and
complicated-stamped sherds. When the
obliterated check-stamped sherds are
counted with the Deptford materials,
the average frequency of that series
rises to 55.1% in the test units,
almost identical to the percentage of
Deptford in features. Similarly, when
the obliterated complicated-stamped
sherds are combined into Swift Creek,
that series accounts for an average of
16.3% of test unit ceramics, just
slightly higher than the percentage for
Swift Creek in the features. Santa
Rosa remains almost identical at 7.02

and plain sherds make up the remaining
16-1% of midden ceramics. Although the
percentage of plain ceramics is a
little higher in the midden than in the
features, the difference probably
relates to their presence in the upper
levels. This will doubtless factor out
when the percentages are recalculated
by strata for the final report on
Pirate's Bay.

The co-occurrence and relative frequen-
cies of these three ceramic series is
in line with what other researchers,
such as Sears, have seen as the
straight-line evolution from Deptford
to Santa Rosa/Swift Creek. Further
evidence for the transitional nature of
Pirate's Bay is found by looking more
closely at some of the features of the
ceramics. For example, the Swift Creek
Complicated-Stamped sherds definitely
do not fit Willey's description of
classic Early Swift Creek. At Pirate's
Bay, they are more like those described
by Sears (1955:5) for the Tucker site
where complicated-stamped sherds, “in
full Deptford association (are) heavy
and rather crude." He went on to say
"I am inclined to classify most of
the(se) sherds not found in Deptford
context as Late Swift Creek.”

The Swift Creek Complicated-Stamped
sherds at Pirate's Bay fit Sears'
description perfectly and, as at
Tucker, occur with Deptford ceramics.
These sherds in no way exhibit the
quality of execution and variety of
design that is described for this
region during Willey's early Swift
Creek. Nor do they have the scalloped
or “pie crust" rims that occur on Swift
Creek Complicated-Stamped and Franklin
Plain as diagnostics of Willey's early
Swift Creek. In fact, only three
notched rims are present in the Pi-
rate's Bay collection and two of these
were effects produced by continuing the
paddle design onto the rim rather than
classic notching. In contrast to the
scalloped rims, Swift Creek Complicat-
ed-Stamped rims at Pirate's Bay tend to
exhibit a gentle undulation, which, in


time, doubtless evolved into the
classic "pie crust" rim. This incipi
ent technique of treating the rims
seems reminiscent of the "odds and
ends" Sears noted for late Deptford at
the Tucker site.
Another suggestion of transition is the
complete absence in features of the
Swift Creek type Crooked River Compli-
cated-Stamped, which was included by
Smith in her Early Swift Creek Complex.
Also, the features produced only two
sherds of the Swift Creek type St.
Andrews Complicated-Stamped, which is
thought to have a longer life span in
Swift Creek than some of the other
complicated-stamped wares. In con
trast, 30 sherds of the type New River
Complicated-Stamped were recovered in
13 of the 20 features used in our
analysis. Willey (1949:386) noted that
New River Complicated-Stamped is
related to both check- and complicated-
stamped series and occurs "...strati-
graphically at the point of transition
from Deptford into Swift Creek. But
the New River Complicated Stamped types
at Pirate's Bay have an earlier "look"
about them than those described by
Willey. The check stamping on the New
River sherds from Pirate's Bay more
closely resembles Deptford Bold Check-
Stamped than Gulf Check-Stamped.
To place the Pirate's Bay site
accurately in time, four samples of
materials recovered from features were
submitted for C-14 analysis to Dicarb
Radioisotope Co. of Norman, Oklahoma.
Two dates were on samples taken from a
refuse pit (Feature 18) containing
Deptford and Swift Creek series sherds.
One of these, run on oyster shell, came
out to 1810 B.P.+50 or A.D. 140 (Dicarb
2997). The second, on a charcoal
sample from the same pit, was dated at
1830 B.P.50 or A.D. 120 (Dicarb 2995).
A third date was taken on a second
large refuse pit (Feature 10/11)
containing an almost complete Deptford
Linear Check-Stamped vessel, and other
sherds of the Deptford, Santa Rosa, and
Swift Creek Series. That sample of
oyster shell was dated to 1830 B.P.+50
or A.D. 120 (Dicarb 2998). The fourth
sample, again on charcoal, was taken
from a basin-shaped pit (Feature 17)
which also contained sherds of all
three series, but had a higher than
average percentage of Santa Rosa types
(12.7%). That sample was dated to 2000
B.P.55 or 50 B.C. (Dicarb 2996).
Thus, the C-14 dates reflect a range
from about 50 B.C. to A.D. 140.
We are defining this transitional
occupation as the Okaloosa Deptford
phase. Based on the dates from
Pirate's Bay, we suggest the phase
spans, at a minimum, the period between
50 B.C. and A.D. 140, and represents a
dynamic period in northwest Florida
prehistory. Influences from the west
are evident, since the Santa Rosa
series sherds are present in a village,
rather than ceremonial, context. At
the same time, ceramic ideas were
probably coming in from the north and
east, influencing the manufacture of
Swift Creek series ceramics. It was
also apparently a period of local
experimentation, as witnessed by the
beginnings of rim treatment that later
became the notched or pie crust rims of
Swift Creek, and the production of new
designs on Deptford paste.
Although we have defined the Okaloosa
Deptford phase on the basis of C-14
dates and ceramics, the excavations at
Pirate's Bay have produced a wealth of
data that will be elaborated in our
final report on that project. A large
number of bone and shell tools were
recovered from the excavations, as well
as a distinctive lithic assemblage that
includes a variety of small, unifacial
cutting tools manufactured on imported,
small milk quartz cobbles. At this
point, we are unsure of the geographic
limits of the Okaloosa Deptford phase.
However, Willey (1949) reported 18
sites that produced Deptford and Swift
Creek ceramics on the northwest Florida
(38, 1985)

Coast, and 14 of these also yielded
Santa Rosa sherds. More thorough
investigations may determine which of
these are multicomponent sites and
which, if any, are transition sites.
The transitional Okaloosa Deptford
phase seems to be represented at the
Carrabelle site in the middle levels of
Pit I and in the lower levels of Pit
II, but there are no Santa Rosa
materials. At the Gulf Breeze site
(Willey 1949) the transition may also
be present. The sub-mound midden at
the Fort Walton Temple Mound, located
less than a mile from Pirate's Bay,
seems to fit clearly within the
Okaloosa Deptford phase. There are
doubtless characteristics of the
Okaloosa Deptford phase, such as those
related to ecology, subsistence, and
material culture, that can be examined
in the final report on Pirate's Bay and
through the investigation of other,
similar sites in northwest Florida. We
hope, however, that definition of the
Okaloosa Deptford phase will provide a
framework for interpreting similar
components along the Florida panhandle
and in neighboring regions.
Prentice M. Thomas, Jr. and
L. Janice Campbell
New World Research
Post Office Box 4246
Fort Walton Beach, Florida 32549

Judith A. Bense
Department of Sociology and Anthropology
University of West Florida
Archaeological Resources Management
In Pensacola
During Che last fifteen years, there
have been significant advances in the
protection of cultural resources.
Federal and state laws concern primari
ly public lands, and affect private
property only when licenses, permits or
government aid is involved. Many
archaeologists have been involved in
the management of archaeological
resources in compliance projects
involving governmental regulations.
(Historic structures are also addressed
by federal and state laws, but they are
not the focus of this paper, ed.).
In spite of the protection afforded
sites by state and federal laws, there
is a significant area which has not
been dealt with effectively: public
land in the county and city domains.
These governmental bodies are not bound
by any of the protective regulations
unless there is state or federal
involvement, and then only on a pro-
ject-by-project basis. Even when this
occurs, city and county officials are
often ignorant of the compliance
process. This has led to widespread
and unnecessary destruction of archaeo
logical resources.
Established in 1698, Pensacola is the
second oldest continuously occupied
European founded city in the continen
tal United States (an earlier Spanish
colony established in Pensacola on
August 14, 1559 by Tristan de Luna y
Arellano was struck by a hurricane and
finally abandoned in 1561, thus passing
to St. Augustine, wnich as founded in
1565, the right to claim oldest contin
uously occupied city status, ed. note).
Its history has been recognized as
significant by the State of Florida
with the establishment in 1969 of the
Pensacola Historic Preservation Board.
Since then, two additional historic
preservation districts have been
established. Local residents recognize
the importance of this long history to
Pensacola's character. The Pensacola
Historical Society membership numbers
over 1000, and there are other quite
active historical groups in the city.
Several museums are supported by local
and state groups, and many historic
standing structures have been pre
Unfortunately, there has been almost
total disregard for Pensacola's archae
ological resources. With almost 400
years of history and 10,000 years of
human prehistory, there are great
numbers of archaeological deposits
continually being exposed during
construction activities. Furthermore,
an active market for old bottles and
other artifacts has developed and many
people benefit financially from the
looting of sites.
During my residence at the University
of West Florida in Pensacola, I have
frequently been told of the exposure
and vandalism of historic deposits in
the city. I have even performed
several contract archaeology projects
in the Historic District. However, I
was generally busy with other archaeo
logical work and, basically, ignored
the situation in the city.
Last winter, two of our students
reached me on a weekend to relate that,
in the excavation of the foundation for
a new city hall, many historic pits had
been exposed and bottle collectors were
descending in force. Actual "claims"
were being made by collectors on the
pits, which were being jealously
guarded. Soon the area began to look
like a mine field. Our students
"claimed" two pits of their own and
began to excavate with proper controls.
Later they called me for help with the
excavation of. an unusual wooden fea
ture, which turned out to be a rare
type of well encased in boards and
July, 1985

containing a hogshead barrel. The wood
below the water table had been perfect
ly preserved. Though this area,
formerly a residential neighborhood,
had been previously disturbed by the
construction of a parking lot, surpris
ingly, significant archaeological
deposits were still present.
During this endeavor, I became uncom
fortably aware of the fact that the
remaining features would soon be
destroyed, that this was public proper
ty, and that tne project was years in
the planning. I realized then that the
situation which had developed was, in
many ways, my responsibility. I and my
colleagues had the education, training,
and experience to prevent this from
happening. This lead to a meeting of
three archaeologists (two from the
university and one from the Gulf
Islands National Seashore) and four
historians who had worked with us on
different projects. We agreed that we
had to get involved and find a way to
protect archaeological resources in
Pensacola. We formed a small action
committee to come up with a plan. A
non-confrontational approach to the
city was adopted. We resolved to
interact as professionals presenting a
well-organized program to city adminis
trators and city council. Our tasks
included: 1. Background research; 2.
Basic design of a compliance program;
3. Structuring a city research design;
and, 4. Design of an ordinance.
With these basic concepts established,
we called a general meeting of the
community, presented our findings, and
asked for suggestions. The resulting
new idea which developed was that of
tax incentive for private parties to
support archaeological work. We were
urged to meet with city administrators
to get their reactions and suggestions
as to how to deal with the situation.
We promptly drafted an ordinance
modeled on existing city ordinances and
scheduled a meeting with city manage
ment and engineering departments. The
reaction was cautious at first but
became encouraging when they were made
aware of possible federal and state
grant funding for survey and planning
activities. We were asked to return
with a detailed presentation and a
revision of the ordinance with more
The proposed ordinance and Archaeologi
cal Resources Management Plan contains
several features. First, the basis of
the plan is an ordinance which will
require all projects which disturb the
surface of the ground in the city to be
reviewed at the pre-application stage
by an Archaeological Review Board.
This is similar to the existing Archi
tectural Review Board which evaluates
all building projects in the special
Second, a master plan and research
design for the archaeological deposits
is being designed for the city. Tms
will be both a planning document and a
quick reference for city administra
A third aspect of the Archaeological
Resources Management Plan is the
identification of appropriate monetary
support for the archaeological investi
gations. Funding is available through
compliance processes, grants, and
donations from local sources. Unused
city facilities may be modified for use
in archaeological fieldwork. Equipment
can be supplied from the university on
a lend-lease type of arrangement.
Expertise can be found in the profes
sional community here in Pensacola or
hired from outside. Facilities for
curation will have to be developed
through cooperative programs with the
university and the community.
Another aspect of funding involved the
private sector. Where there is no
compliance process and accompanying
funding, we have developed a plan to
provide tax incentives to fund the
necessary work. While obtaining
professional legal and tax advise is
recommended, the following approaches
to tax benefits have worked in the
past: Non-profit organizations exist

in both the historic preservation
system and the university system.
Individuals and companies can donate
funds earmarked for archaeological
purposes and receive one-to-one tax
credits. In addition, the appraised
prices of any collection donated to the
university or state can be a complete
tax credit; although, it is more common
to value the artifacts and associated
data as the coast of the data recovery
project. This program is extremely
beneficial for small and medium-sized
businesses and individual landowners.
Corporations also have the added
incentive of the "good press" which
accompanies archaeological investiga
tions .
So, at this time, tne Archaeological
Resources Management Plan for the City
of Pensacola is moving along. The
consideration of archaeological re
sources in the special districts is
written Into the Land Development Code
being adopted by the city.
Tbe Hawkshaw Project
One of the first archaeological pro
jects conducted in Pensacola under this
program was undertaken with the cooper
ation and support of Gulf Power Compa
ny. In the fall of 1983, it was
brought to the attention of the small
archaeological action committee that
the company had purchased over eleven
acres in downtown Pensacola on which to
build a corporate headquarters. We
checked with the State Historic Preser
vation Office in Tallahassee and
determined that there were no historic
preservation compliance procedures
applicable to this project, as there
was no federal involvement and the
project had been determined not to be a
Development of Regional Impact under
state laws. However, we developed a
proposal and met with company represen
tatives to explain the potential of the
cultural resources on the property.
Also, an article came out in the local
newspaper describing the history of the
planned construction area. We proposed
to evaluate the cultural resources in a
testing program which included research
in history, folk history and archaeolo
gy. The company agreed to this endeav
or and the Hawkshaw Project was begun.
The project has been and will continue
to be a joint effort between the
company, the local community, the
university, and historic preservation
Testing a site this large in an urban
neighborhood had to be approached from
an historical perspective. The site is
located in a densely populated, low
income, working class area which has
been occupied continuously since at
least the Civil War. There also was
the possibility of a colonial residence
on the property. A sawmill, railroad
and wharf, and fishing provided the
primary employment for the historic
residents. The Hawkshaw community
began ethnically mixed and changed to
primarily black in the last thirty
Prehistoric occupation had been indi
cated by the collection of a few Early
Woodland pot sherds from the area. The
environment was conductive to prehis
toric occupation. A small stream
emptied into the bay adjacent to tne
project area, which Is hlgn and well-
Archaeological testing began with a
controlled surface collection aligned
to the current plat system. This work
was performed by the Gulf Power employ
ees on "archaeology day". The items
were washed, classified, recorded and
boxed on-site in one day. The students
and professionals from the area greatly
aided in this procedure. A posthole
testing program was then initiated in
the sensitive areas of the site. This
was followed by a series of test units,
which were placed in specific and
representative areas of the site. Road
grader test strips were also used to
expose subsurface features.
Historical "testing" included a review
and evaluation of the records and docu-
(38, 1985)

merits available on tne Hawkshaw commun
ity. The folk historian identified
and evaluated possible informants who
clearly remembered che unrecorded
recent past of the neighborhood and its
way of life during the late nineteenth
and early twentieth century.
Archaeological testing revealed that
deposits from both prehistoric and
nistorlc occupations were intact. The
prehistoric occupation was confined to
a 50 by 100m area along the bay bluff
edge. It is an early Deptford occupa
tion, radiocarbon dated at 1950 + 50
B.P. (A.D. 0; Beta 9410) and 1790 + 60
B.P. (A.D. 160; Beta 9411). Two
vessels from a feature were recon
structed. Several refuse pits were
present, and those with shell midden
had excellent bone preservation. Tne
site is significant, as it is on the
western edge of the Deptford cultural
area. Minority wares of complicated-
stamped, shell-impressed, incised, and
punctated surface decoration are
associated, and western indicated
contact along the Gulf coastal plain.
The historic remains include three
intact components: a Territorial
period midden, a Victorian period
midden and features, and recenc depos
its. Features are extremely abundant
(approximately 9,000 are projected in
the 11 acres). The historical records
for the area are of mixed usefulness.
With this wealth of information
demonstrated and a series of favorable
stories in the press, we presented
recommendations for a data recovery
effort to the company. The basis of
the recommendations was that both Gulf
Power and the community would benefit
from this new information about
Pensacola's past. What the company
would get from this non-compliance
endeavor is the recognition that it
contributed to the discovery and
interpretation of a significant part of
Pensacola's cultural heritage. This
will be communicated to the public
through the production of pamphlets,
videotapes, and a documencary movie
which will be placed in the public
school media centers and other public
areas. Student work sheets and in
struction booklets will accompany the
material. In addition, the design of
the new building lobby will emphasize
historic and prehistoric themes in
murals, exhibits, and push-button
The public archaeology aspect of this
project is quite exciting and meaning
ful. We have had a design consultant
in all aspects of the work who inter
acts with and interprets for the
public. This kind of dispersal of
information, we believe, will raise the
awareness of the community in under
standing archaeology and its benefits
to our present and future.
This project is setting an example in
the community, showing that archaeology
can be managed when adequate time is
allowed. The action committee members
are pleased to see that their efforts
are fruitful. The project would have
never been imagined had we not gotten
involved. Planners of several other
construction projects have contacted us
concerning the possibility of doing an
evaluation of their property. They now
see tnat there are real benefits, and
that archaeology does not impair
construction. The use of local profes
sionals, students, and facilities
brings respect from the local community
that would have been impossible before.
With a working Archaeological Resource
Management Plan, we hope to make
Pensacola an example that other cities
can follow. We are well on our way,
and we have tried to put our money
where our mouth is.
Judith A. Bense
Sociology/Anthropology Department
University of West Florida
Pensacola, Florida 32514

Cheryl Claassen
Appalachian State University
The Escambia Bay Drainage Archaeologi
cal Research Project, under the direc
tion of Judith A. Bense, for several
years has been studying the prehistoric
adaptation of human populations in this
region. During the summer of 1984,
field schools of the University of West
Florida and Appalachian State Universi
ty jointly conducted research in the
drainage. As part of this research, I
directed testing at five small middens
(Figure 1).
Preliminary testing in 1983 at the
Thompson's Landing site (8Es950) on a
ridge toe terminating in Thompson Bayou
on the northwest side of the Bay
encountered shell in the midden and in
features. In 1984, two samples of
shell were recovered from two
Mississippian features. Feature 1,
approximately 40 cm deep, was excavated
in 10 cm levels for the purposes of the
shell study Table 1 lists its shell
species ratios. Oyster shape indicated
collection from a cove. In addition to
shellfish, Feature 1 contained numerous
Mississippian sherds, perforated
alligator teeth and scutes, a projec
tile point, a bone awl and a large
piece of a steatite bowl.
Feature 2, some 12 meters southwest of
Feature 1, contained a thick lense of
wedge Rangia (Rangia cuneata) (86%) and
Carolina marsh clam (Polymesoda
caroliniana) (14%) and additional
arcifacts. Oyster was absent.
West of the Thompson's Landing site,
and higher up the ridge toe shell
midden was again evident (8Es947).
Unit 2, one m 2, was opened in the
center of the exposed midden (Table 2).
A grab sample of shell from an area 30
x 13 x 6 cm was made one meter to the
south of this unit, and labeled Unit 1.
A late prehistoric or protohistoric
Pensacola Red sherd was found in this
shell, as were two bone fragments.
Located in a low-lying field on the
east side of the bay on Pa^e Mill Creek
is 8Srl69. A single 50 ctn unit was
opened here in an area containing
surficial deposits of R. cuneata. Two
distinct shell middens zones were
encountered. They are separated by 20
cm of sand. Mississippian sherds were
found on the surface and in level 6, in
the sand (Table 3).
Moccasin Mound (8Sr85) was more inten
sively tested. The mound, marked by a
clump of cedar trees in a dense stand
of saw- and needle-grass, measures 30 m
x 15 m and consists of a shell midden
1.2 m thick. Three 50 cm square units
were excavated into the midden in 10 cm
levels, producing a 0.2% sample of the
shell in the site. Deptford fabric-
marked sherds were found on both the
surface and in Levels 9 and 13 of Unit
1. Numerous other sherds from both the
surface and the midden indicated a
subsequent Mississippian occupation.
Three large "potholes" had been dug by
a single curious individual in the
summer of 1982. The walls of two of
these holes were cleaned for the
removal of the shell units. The third
shell unit sample area was placed at
the southern edge of the island.
Drumfish, catfish, gar, alligator,
snake, turtle, deer, mussel (C_.
demissa) periwinkle, crab, numerous
unidentified fish species, sherds, and
charcoal were mixed with the shells
throughout the upper 11 levels of the
units. Very little other than shell
July, 1985

Figure. The Escambia Bay drainage area in extreme northwest Florida,

20-30 cm
30-40 cm
40-50 cm
Wedge Rangia
(Rangia cuneata)
Carolina Marsh Clam
(Polymesoda caroliniana)
Common Oyster
(Crassostrea virginica)
Mean length
4.04 cm
Table 1. Thompson's Landing
Shellfish Species
(8Es950), Feature 1 Percentage (gms) of
Wedge Rangia
(R. cuneata')
Carolina Marsh Clam
(P. caroliniana)
Common Oyster
(C. virginica)
Table 2. 8Es947 Unit
2, Percentage
of Shellfish Species
Species cmbs
40-50 50-60
cmbs cmbs
Wedge Rangia
(R. cuneata) 92

Carolina Marsh
(P. caroliniana) 7
Common Oyster
(C. virginica) 2

Table 3. 8SR169 Unit 1, Percentage (gms) of Shellfish Species
(38, 1985)

was encountered in lower levels. All
material removed was screened through a
0-635 cm (&") mesh. Soil samples were
selectively taken and water-screened.
The water table was intersected in two
of the units. Cave-in of the walls
terminated excavation prematurely in
Units 1 and 3, but the sub-midden
stratum was found at a depth of 150 cm
in Unit 2. The species ratios of each
provenience are given in Table 4.

The final site investigated in 1984 was
Gilligan's Island (8Sr143), a
Mississippian shell midden very similar
in setting to Moccasin Mound. Located
one km east of Moccasin Mound and near
the eastern shore of Escambia Bay, this
longer cedar covered island is today in
an estuary setting surrounded by moving
water. Prehistoric peoples discarded
large quantities of shell in the center
of the island over an area approximate-
ly as big as Moccasin Mound. Six 50 cm
square units were excavated in 5 cm
levels, resulting in an estimated 0.43%
sample of the shell at the site. The
loci of units were arbitrary. Several
smaller shovel tests were made the
previous summer to determine the site
area. Although the shell recovered
during the 1983 season is useful for
confirming several observations to be
reported later, the subdivision of the
midden in these tests into 20 cm levels
precluded its use in this analysis.
Table 5 presents the species ratios
from this fifth Mississippian midden.
Notable is the appearance of the
freshwater bivalve Elliptio spp.

The Shellfish Species

Five bivalve species were recovered
from the five Escambia Bay sites. A
brief description of each is necessary
background for succeeding portions of
this paper.

Wedge Rangia (R. cuneata) is the most

numerous species in these sites, and
Many others in the bays of the Gulf


Coast. This species has an experi-
mentally determined salinity tolerance
range for adults of 0-30% (Castagna and
Chanley 1973:82)- In natural settings
it is found in salinities of 2.5%
(Godwin 1967) to 17% (Parker 1965). R.
cuneata is typically found landward of
oyster populations and below the
optimal salinity range for that species
(Fairbanks 1963:4). Prior to the 1960s,
humans would have encountered R.
cuneata only along the northern Gulf
Coast. Today it has colonized areas of
the Atlantic northward to Chesapeake
Baye The appearance of this species at
Cahokia and at the Arrowhead Farms site
in Kentucky has yet to be explained
(Hill 1983).

R. cuneata thrives on almost any
organism small enough to be ingested
(Fairbanks 1963:38)- It spawns almost
continuously, beginning some time in
late summer or fall and continuing
through the winter and spring into
early summer. They are found clumped
across the bottom, burrowing only
slightly, and are easily collected by
hand. Other predators are red and
black drum, catfish, Atlantic croaker,
blue crab, and white shrimp. This is
the only brackish water shellfish that
can osmoregulate, and it is always the
dominant shellfish species in an

Archaeologist frequently confuse (P.
caroliniana) with R. cuneata. The
latter is distinquished by a large hole
located in the hinge area, long lateral
hinge teeth, and two muscle scares, one
on either side of the arch formed by
the shell. P. caroliniana lacks all
three features, and has three small
hinge teeth. Although this species has
not been previously reported in middens
of northwest Florida, the author has
found it at Fort Walton, in numerous
other middens along Choctawhatchee Bay,
and in all five middens investigated in
Escambia Bay.

P. caroliniana is a characteristic


species of Spartina marshes on the
northern shore of the Gulf of Mexico.
It is frequently found interspersed in
R. cuneata populations, although always
in lesser numbers. During collecting
by the author in Escambia Bay, a ratio
of 12 R. cuneata to 1 P^. caroliniana
was frequently encountered. P.
caroliniana is usually found in water
whose salinity lies in the range 0-19%
(Castagna and Chanley 1973:68) (Adults
can tolerate salinities as high as
26.3% for short periods). It is most
commonly found in salinities less than
10% (Castagna and Chanley 1973:68). In
a tidal marsh in Mississippi the
largest populations of P_. caroliniana
were found in regularly flooded areas
and in the upper reaches of a tidal
creek. In Virginia it is found regu
larly in muddy areas.
The common oyster, Crassostrea
virginica, is ubiguitous in middens of
the Atlantic coast and southwestern
Florida. It Is rarer environmentally
and archaeologically along the Gulf
Coast. Oyster is extemely sensitive to
silt and grows best on sand substrates.
According to the data summarized in
Castagna and Chanley (1973:66), the
range of salinity tolerable by adult
oysters is 5%-30%, while that optimum
for growth is 12.5%-25%. Oyster reef
formation occurs only between salini
ties of 10%-30%. Predators include the
drums, starfish, and whelks.
Geukensis demissus (Modiolus demissus),
the ribbed mussel, is plentiful in salt
marshes and brackish water. Andrews
(1953) reports it at all salinities
above 8% 10%. This species was
present in nearly every level of the
Escambia Bay sites, but it was always
less than 1% of the shell by weight.
Ellipto spp. (probably dilatata) is a
freshwater mussel found in a range of
substrates from silt to cobble, in
water 1 cm to 30 cm deep, standing to
swiftly flowing (Buchanan 1980:38).
Its appearance in the midden at 8Srl43
is surprising since populations of this
species would live north of the salt
water/freshwater interface. These
specimens must have been transported to
the island from the north.
The Deptford and Mississippian
The archaeologist must consider the
specie ratios of R. Cuneata, P.
caroliniana and C. virginica in shell
middens of Perdido, Pensacola,
Escambia, Choctawhatchee Bay and
Apalachicola Bays. A relatively high
proportion (by weight) of P^.
caroliniana and low proportion of £.
virginica means either very low salini
ty (5%) or high siltation. The two
species utilize nearly mutually exclu
sive habitats, certainly so for repro
ducing populations. This exclusiveness
is apparent in the respective propor
tions of these two species in the
Deptford levels of 8Sr85 (Unit 1 Levels
9-15?, Unit 2 Levels 9-15?), where
oyster is frequently absent altogether.
Moccasin Mound (8Sr85) today is sur
rounded by saw- and needle-grass.
Flowing or even stagnant water is some
distance removed from the island in
most seasons. That this location was
an island when utilized by prehistoric
peoples is evident today by its loca
tion on the delta, and through subsur
face shovel testing into the grass.
The local environment in 1984, reflec
tive of much silting in the bay, is
dramatically different from what it
must have been In either Mississippian
or Deptford times. Presumably, during
the period of prehistoric occupation,
the island was surrounded by brackish
water supportive of the Rangia and
Polymesoda populations discarded on
shore. All the species identified from
the Deptford levels would have been
available in the water surrounding the
The species ratios in the Mississippian
levels of 8Sr85 suggest a stable, low
salinity regime that fostered the
(38, 1985)

Unit 1
Unit 2
Unit 2
t = trace total g sorted 540,007.5g
Rc = Rangia cuneata
Pc = Polymesoda caroliniana
Cv = Crassostrea virglnica
Table 4 8SR85 "Moccasin Mound" Shellfish. Percentages by weight, 10 cm

Unit 1 Unit 2 Unit 3 Unit 4 Unit 5 Unit 6

t trace total g sorted 221,293-12g
Rc Rangia cuneata
Pc Polymesoda carolinian
Cv Crassostrea virginica
E = Elliptlo spp.
Table 5 8SR143 "Gilligans Island" Shellfish. Percentage of weight,
5 cm levels.
Size Class
Presence Period Comments
Highest densities of this
class in March.
January, February Mo6t abundant size class.
Hign densities occur in
January, April, May.
Largest class in September.
Sumer only
47 of September 15 collec
late sumner,
Mean lengtn of population shows decline
young individuals are recruited.
in late spring, early summer as
Table 6 Size Classes and Seasonality in R. cuneata.
(Taken from Fairbridge 1963:20-27).

Unit 2
late summer,
late summer ,
late summer,
late summer,
late summer,
late summer,
late -summer,
late sunnier,
(1, 2)
late summer,
late summer,
late summer,
spring or fall (4)
late summer,
late summer,
sunnier (5)
summer (5)
late sumuer,
late summer,
summer (5),
fall (4)
summer (5)
(1. 2)
late summer,
Key: (1) .5-4% of size class 14.25-23.75 min
(2) shells o size class 26.75-30.75
(3) >4% of size class 14.25-23.75
(4) thin sectioning fast growth >33%
(5) 1.75-22.75 mm
(6) highest % class 27.0-30.0 mm
Table 7. Seasonality of Proveniences at 8SR85 Moccasin Mound.

Unit 1 Unit 2 Unit 3 Unit 4 Unit 6

late suomer (3)

late summer,
fall (1)
late sumner,
fall (2)
summer (3)
spring, fall (5)
laCe summer,
fall (2)
late summer,
fall (1)
late summer (3)
late summer,
fall (1)
summer (4)
summer (3)
late summer,
fall (2)
late summer,
fall (1)
late summer,
fall (2)
summer (4)
late summer,
fall (2)
late summer,
fall (l)
late summer,
fall (2)

late summer,
fall (1)
late summer,
fall (2)
late summer,
fall (2)
(1) 0.5-4.02: of size class 14.25-23.75 mm
(2) shells of size class 26.75-30-75 mm
(3) shells of size class 24 mm-26.7 mm (late summer)
(4) 1.75-22.75 mm
(5) thin sectioning fast growth >33%
Table 8. Seasonality of Proveniences at 8SR143 Gilligan's Island.

expansion of Rangia, which is typically
the dominant shellfish species when it
occurs. The high percentage of oyster
in level 4 of Unit 1 would require a
dramatic, rapid increase in salinity
and rapid decline in subsequent levels
to have grown in the shallow water
surrounding the island. Such an
increase in salinity would have
required the island's occupants to
travel shoreward to gather Rangia and
particularly Polymesoda. More likely,
the oysters were collected in a more
southern area of Escambia Bay, and then
consumed and discarded at 8Sr85 along
with the shells of locally collected R.
cuneata and P^. caroliniana. Today
oyster lives approximately four miles
south of either 8Sr85 or 8Srl43.
The initial Mississippian levels at
8Es950, 8Srl69, 8Sr85 and 8Srl43 Units
4 and 5 lack oyster altogether, while
the inital levels at 8Es947 and 8Srl43
Units 1, 2 and 6 are notably high in
oysters, although it quickly declines
in percentage. If no oyster in this
section of the delta is the rule,
oyster in the founding levels of these
two sites must also have been trans
ported north from a more southerly part
of the bay. The relatively high amount
of P. caroliniana further suggests that
one or the other species was transport
ed to the site from elsewhere. While
oyster could stimulate special collect
ing trips, P. caroliniana most likely
would not. An hypothesis arising from
these observations that may deserve
attention in the future is that
Deptford populations were more local
ized in their exploitations of marine
resources than were Mississippian
peoples. Not only was oyster apparent
ly secured to the south of these sites,
but also the freshwater genus Elliptio
was transported from north of the delta
by Mississippians.
Seasonality of Shellfishing
Numerous characteristics of individuals
and populations of shellfish have
calendrical import. Distances between
the surficial grooves on archaeological
R. cuneata have been measured and
compared to those of control collec
tions of recently killed animals to
infer month of shellfishing (Aten 1981,
others). Repeated attempts by the
author to correlate these surficial
grooves with internal annual growth
increments have failed.
Several other approaches to seasonality
have been explored. The most simple of
these is suggested by the observations
of Fairbanks (1963:20) who reported
several size classes present only
during specific periods of the year.
These lengths and their observed
seasonality are presented in Table 6.
Collections by members of the Escambia
Bay Drainage Archaeological Research
Project in the spring, summer and fall
of 1984 revealed juveniles in the size
class 276-300 mm present only in the
19 September collection, agreeing with
Fairbanks' conclusion.
Measuring only the whole valves of one
side from each provenience in 8Sr85,
(N=5500) and 8Srl43 (N=3038) suggests
several summer-late summer collections
based on Fairbanks' data. These data
summarized in Tables 7 and 8.
If one to four percent of a collection
fell between 14.25 and 23.75 mm, a late
summer assignment was made. Other
criteria are specified on the two
tables. The evidence from shell
lengths points overwhelmingly to late
summer-fall shellfish collecting at
these two sites. The Deptford collect
ing pattern is indistinctive from the
Mississippian pattern at 8Sr85.
A living population of Rangia cuneata
in Mulatto Bayou on the northeastern
side of Escambia Bay was sampled six
times from January 1984 to September
1984 in order to determine what, if
any, growth characteristics registered
in the shell could be used to
distinguish specific months of the

year. A fast growth-slow growth annual
profile similar to that found in a
contemporary North Carolina Mercenaria
population (Claassen 1982, 1983) was
discovered. Animals killed in fast
growth were 33% 16 May (N=24), 24% 1
June (N=37), 13% 18 June (N-34) and 8%
19 September (N=38) (The January
collection netted only 2 individuals
and has not been included).
After a thin-section was made of one
valve of each individual, reflected
light was used to determine whether the
final growth period was fast (white
shell at margin) or slow (grey shell at
margin). The resulting graph indicates
that the majority of animals are
undergoing fast growth in the period
after September 19 and before May 16.
Analysis of two archaeological levels,
8Sr85 Unit 1 Levels 7 and 13 and 8Srl43
Unit 6 Level 3, indicates that the
shells were collected sometime during
the September to May period.
Intensity of Exploitation
The intensity of human harvesting can
be measured in several ways (see
particularly Claassen 1982). Relevant
to the data gathered in this study are
changes in the size classes and mean
length of R. cuneata over the time
period represented in the midden. Such
a study requires that the shells
considered together are indeed a single
death assemblage and not subsets of
numerous temporally distinct collec
tions whose random spatial association
has been intersected by the archaeo
logists' unit. To increase the proba
bility of this assumption being true,
each shell collection discussed was
removed from an area 30 x 30 x 5 cm
thick (8Srl43) or 30 x 30 x 10 cm thick
If the deepest or initial cultural
level of each unit at each site repre
sents a previously unexploited
molluscan population, then size class
profiles and mean lengths of shells in
that and subsequent levels can serve as
standards for gauging the impact of
human harvesting on the Rangia popula
tions. The presence of fewer size
classes in later levels, a shift toward
a smaller dominant size class, or a
steady decline in a mean length are
indicative of either an increasingly
unsuitable natural environment or
intensive human harvesting. If the
decreases are bimodal or multimodal,
then either human exploitation or
season changes are implied. At this
time the test implications are inade
quate for distinguishing between
There are several gradually declining
sequences of mean lengths in the upper
levels of the shells from Gilligan's
Island that do indeed suggest very
intensive harvesting of the Rangia
population: Unit 2 levels 5-1, Unit 3
levels 9-2, Unit 4 levels 6-3, and Unit
6 levels 6-3. Whether or not this
exploitation was severe enough to
explain the cessation of the prehistor
ic use of the island may never be
Summary and Conclusion
Human utilization during Deptford times
in the Escambia Bay drainage area is
represented by only one site in this
study, Moccasin Mound apparently a
mealtime camp on an island located in
what is today the delta of the Escambia
River. All species found in the column
samples are typical of low salinity
estuaries and were obtainable in the
water surrounding the island. The
dominant shellfish species in both
number of valves and weight was R.
cuneata, followed by P. caroliniana.
Oyster was absent in most of the
relevant levels and, given the nearly
mutually exclusive requirements of
oyster and P. caroliniana, was appar
ently unavailable locally. Deptford
ceramics were found only in Unit 1
levels 9 and 13, but the virtual
absence of oyster and the relatively
high percentages of jP. caroliniana
suggest that Deptford levels are
(38, 1985)

Unit 1 levels 9-15, Unit 2 levels 9-15
and Unit 3 levels 9-11. Shellfishing
was apparently a late summer-fall
activity, as suggested by seasonal
occurrences of certain sizes of shell
in modern settings.
Mississippian populations utilized the
Rangia and Polymesoda populations in
much the same way as did Deptford
people, but added features can be
detected. If this interpretation is
correct and not a sampling problem,
Mississippian peoples exploited the
delta not only through mealtime camps
on tiny islands such as 8Sr85, where
little other than shell was found, but
also from village sites both on islands
(8Srl43) where the sherd density is
quite high and the mainland (8Es947,
8Es950, and 8Srl69) to which estuarine
foods were transported. It has been
argued that oyster was collected in a
more southerly section of the bay and
transported to 8Sr85 and 8Srl43, and
the freshwater mollusc Elliptio was
transported from up river south to
8Srl43 on the bay.
Like the Deptford population,
Mississippian peoples utilized the
shellfish resources of the delta almost
exclusively in late summer and fall.
Incompletely analyzed data suggest at
this point in the research that shell
fish harvesting in the water surround
ing 8Srl43 may have been intensive
enough to have had an adverse effect on
the Rangia population's growth profile.
The appearance of Elliptio in two
distinct vertical halves of the midden
suggests that this exploitation spanned
several seasons rather than being one
intensive short-lived clambake.
The conclusion of late summer-fall
shellfishing is at odds with the April-
May harvesting posited for Rangia
middens on the Texas coast (Carlson
1983). Numerous technical
difficulties have been encountered by
the author and others (e.g., Irv
Quitmyer, personal communication, 1984)
when the customary seasonality
determination technique (Aten 1981) is
employed. Future research will concen
trate on refining the seasonal correla
tions with size class data in Rangia
populations and exploring the problem
of subdividing shell middens into
distinctive shellfishing seasons.
Cheryl Claassen
Department of Anthropology
Appalachian State University
Boone, North Carolina 28606

Ned J. Jenkins and Cyril B. Mann
During the Spring of 1984 Auburn
University at Montgomery conducted an
archaeological reconnaisance of the
Conecuh Drainage Area for the Alabama
Historical Commission. Three months
were spent locating sites in portions
of seven counties in South Central
Alabama (Figure 1). The primary ob
jectives of this project were to locate
sites eligible for nomination to the
National Register of Historic Places,
determine the destructive agencies
impacting these sites, develop models
of sites location, and develop a
culture history for the region. This
paper deals primarily with the post-
Archaic culture history of the Conecuh
Drainage Area and how it articulated
with that of surrounding regions.
Very little previous archaeological
research had been conducted in the
Conecuh Drainage Area. Brief forays by
Sears (1962), Fornaro (1974), Chase
(1982) and Bense (1983), however
provided some good insights on what
kinds of sites and materials to expect.
One hundred sixty-nine sites were
located and evaluated by this study.
An analysis of the artifacts revealed
that the Conecuh Region was sparsely
occupied during the Paleo-Indian
Stages. Activity increased during the
Archaic Stage, and was focused on the
procurement of Tallahatta quartizite.
Several large Tallahatta quartzite
quarries and reduction stations were
recorded within the Tallahatta Forma
tion, situated on the southern edge of
the Southern Red Hills Physiographic
District. One large site covered over
1.6 km During the Middle Woodland
Period the Conecuh Region was virtually
vacant. Later, during the Late Wood
land Period, a populous Weeden Island
Culture dominated the region. During
the Mississippian Stage, the Upper
Conecuh was controlled by a relatively
small Rood Creek-Lamar population,
while the Lower Conecuh was dominated
by a pure Pensacola Culture with little
admixture of Fort Walton. During the
Proto-Historic and Historic Periods,
the area was virtually abandoned.
Establishing a regional chronology is a
basic step which must be accomplished
prior to any meaningful studies of
prehistoric lifeways or cultural
process. This procedure involves the
recognition of regional artifact
typologies and the arrangement of these
types into a coherent temporal se
quence, i.e., outlining the morphologi
cal change of artifacts through time.
Recognition of artifact types and
temporal positions in a previously
unstudied area is often obtained
through analogy and cross-dating with
artifacts from adjacent areas. A given
study area is understood within the
broader contexts of neighboring re
gions. The present chronology has been
formulated in this manner (Figure 2),
since there are no excavated stratified
sites or radiocarbon dates from the
Conecuh Region.
The two classes of artifacts that work
best for chronology building in the
eastern United States are lithic tools
and ceramics. These artifacts are more
durable, numerous, and regionally
diagnostic than others. The chronology
outlined here is based on the morpho
logical/temporal ordering of ceramic
and lithic artifact types recovered
from surface contexts within the study
Throughout the post-Archaic prehistory
Vol. 38 Numbers 1-2 Part 2 THE FLORIDA ANTHROPOLOGIST July, 1985

C onecub
11 p p er Ala
River Valley
Lower Talla poosa
River Val Iry
Lower Ala
River Valley
AAobi l e.
J> 1 la
Lower Tombig bee Northwest
River- Valletj Florida
AD. 1700
Historic Creek
Historic Creek
Historic Creek
Historic C
/ lohome
Histone Apalarbee
'XPen sacla
Fort N\
Historic Cree.V
(Lawson r .elrl )
FUar kmon
Aberc rornhle
A D 1500
Ala River PKase
(Burial lira)
r Phase
(Rear TFird)
(Dr Lake)
Lnrnar^ /
Moundvili e
Sbine H
Shine I
(BollIf Creek)
(f3ottle Creek)
Bull Creek
Roods Creek
yf ^Pensacola
? ?
? ?
A D 1000
Union ^
9 9
A D 800
Ten?, aw
Weeden Island
Wf eden
Hopa Hull
A D feOO
D^ad River
Earlg Weeden
ude 5v*if| Creek j
Eadg WeeHrn
1 si and
C a 1 loway
F)r pt ford
(Cobbs Swamp)
A D 200
De,pt ford
(Cobbs Swamp)
Pe pi ford/
Millfrl mrin
For te.r
F?1 \ ^osa
r*/ \
n / \.
/ \
fMandeville n j
bar lii
Swift Creek
(Mnndevil)e I)
Late D*ptfprd
200 B C.
('.Ivy Knol 1 )
Ale xander
^Tvy Knoll)
Farlg Dept ford
000 BC
? ?
00 B C
800 B.C.
1000 B C
Potle ry
Mill brook
Plain Fiber
Tempered R>Hnj
Flam Fiber
Tempered Fittenj
Plain Fiber
Tempered Rdtenj
Fiber ^
Tempered Rtlrrg
Tempered Rflenj
Fiber Tempered
Fiber Tempered
FMtery ?
Late"- > ^
Stalling' IsUni^.
Figure 2 Culture Chronology of Fh>
Conecuh Drainage and Surrounding Regions

of the Conecuh Region, little innova
tive local development seems to have
occurred. Rather, the Conecuh seems to
have been a reflection of the environ
mentally richer and more culturally
dynamic northwest Florida Gulf Coast.
If not for large deposits of Tallahatta
quartizite, the population density
would probably have been minimal during
the Archaic and Gulf Formational
stages. After the end of the Gulf
Formational stage (ca. 100 B.C.), use
and export of Tallahatta quartzite
diminished considerably. So did the
population density of the Conecuh,
which had peaked during the Late Weeden
Island Period (A.D. 400 1000).
During this period the Conecuh seems to
have served primarily as a population
overflow area for the northwest coastal
region of Florida.
Gulf Formational Stage
(ca. 2500 ca. 100 B.C.)
As elsewhere in the earliest ceramics
in the Conecuh region were fiber-
tempered. All fiber-tempered pottery
recovered by this study is plain
surfaced. Several large sherds indi
cate that the vessel shape was the
hemispherical bowl. It is virtually
impossible to assign this pottery to
any of the previously defined fiber-
tempered ceramic series, i.e.,
Stallings Island, Orange, Wheeler, or
Norwood. It might, however, belong to
the Norwood series, considering its
lack of decoration and its proximity to
the northern coast of Florida where it
was defined (Phelps 1965).
The projectile point type most fre
quently associated with fiber-tempered
pottery in the Conecuh region is the
Elora point, usually made of Tallahatta
quartzite. Other point types which may
also be associated are the Pickwick,
Savannah River, Kays, MeIntire, Lime
stone, and Wade Points. Chase (1966,
1978) has also noted an association of
plain fiber-tempered pottery and Elora
points in the Terminal Archaic
Millbrook Phase for the Central Alabama
Dating the period of manufacture of
fiber-tempered pottery in the Conecuh
Region is not easy task. It could date
as early as 2500 B.C., considering the
early dates for Stallings Island
pottery along the Savannah River
(Stoltman 1972). In that area and in
north Florida the manufacture of fiber-
tempered pottery ends around 1000 B.C.,
(Stoltman 1972; Bullen 1972), while to
the west, on the Tombigbee River, it
lasted till ca. 500 B.C. (Jenkins 1975,
1982). A reasonable guess date for the
duration of fiber-tempered pottery in
the Conecuh Region might range from
approximately 1500 B.C., or slightly
Woodland Stage
(100 B.C. A.D. 1000)
After the disappearance of fiber-
tempered pottery, the Conecuh was
virtually abandoned until the appear
ance of the early Weeden Island Culture
at approximately A.D. 400.
The reasons behind this abandonment
probably lie in the realm of environ
mental change and associated settlement
pattern redistribution. Research along
the Escambia and lower Conecuh River by
Judy Bense (1983) hinted that by 1000
B.C., there was evidence of a gradual
shift in settlements from the interior
headwaters of the Escambia, the
Conecuh, and the Patsalagi, to the
lower Escambia and Pensacola Bay area.
By 1000 B.C. the small interior rivers
which drain into the Escambia were
The Late Archaic Period coincided with
the end of the Altithermal climatic
episode, during which extreme drying
conditons prevailed. As xeric condi
tions progressed, groups no doubt moved
out of the more sensitive dry upland
(38, 1985)

tributary areas and into the larger
floodplain of the lower Escambia River.
Associated with this drying condition
was the closing of the mouth of
Pensacola Bay by Santa Rosa Island and
Perido Key, resulting in a reduction
of water velocity in the lower bays.
This, along with the previous 10,000
years of sedimentation caused the
development of the rich marsh systems.
With the gradual increase of the
productivity and size of the marsh
systems, the interior floodplains were
gradually abandoned. This phenomenon
has been documented as occurring
between 4000 and 6000 years ago in many
areas along the eastern United States
coast (Kraft and Margules 1971, Bense
1983). However, the Conecuh basin is
the only region of the lower south
where it is believed to have caused the
full scale abandonment of an area for
several hundred years.
By approximately A.D. 400 the Conecuh
region was again populated, by people
of the Weeden Island culture. Early
Weeden Island can be defined cera
mically in this region by an assemblage
of approximately 80% Weeden Island
Plain, 0-10% Wakulla Check-Stamped, and
10-12% decorated pottery, i.e., Weeden
Island Incised and Punctated, Carabelle
Incised and Punctated, Keith Incised,
Ruskin Dentate-Stamped, and Weeden
Island Red. The most numerous decorat
ed types are Carabelle Incised and
Punctated. Swift Creek Complicated-
Stamped, of eastern origin, and Furrs
Cord-Marked, of western origin, togeth
er comprise less than one percent of
the assemblage. The Late Weeden Island
complex is defined by 50-60% Weeden
Island Plain, as much as 40% Wakulla
Check-Stamped, and less than 5% deco
rated types. Woodland ceramics for the
area are shown in Figures 3-6. The
early Weeden Island complex is virtual
ly identical to the early Weeden Island
complex found between Pensacola Bay and
Mobile Bay in that it lacks any appre
ciable amount of Swift Creek Complicat-
ed-Stamped pottery.
At least three point types appear to be
associated with Weeden Island ceramics:
Bakers Creek, an expanded-stemmed or
weak sided-notched point similar to the
Florida Columbia point, and the small
triangular Madison point. Occurring
infrequently is a small contracting
stemmed point defined in this study as
the Andalusia Stemmed type. The most
frequently used raw materials were
Tallahatta quartzite on the lower
Conecuh and Ocala chert on the upper
During the Late Weeden Island period
there was a very large population in
the Conecuh Drainage. Most sites
recorded by this study contained some
sort of Weeden Island component. Two
basic types of components were recog
nized, base camps and specialized
camps. Specialized camps were defined
as very small, sparse artifact scatters
reflecting temporary seasonal use.
Base camps were defined as dense
artifact scatters covering a large
area. These sites were probably
occupied on a semi-permanent basis.
One mound center, near Adalusia, is the
Mitchell site, previously located by
Sears (1966). This very large site
contains two mounds as well as spatial
ly distinct early and late Weeden
Island habitation areas. It is located
on a ridgetop 1.6 km east of the
Conecuh River. The water sources for
this site were spring heads that emerge
from the sides of the ridges. A number
of large base camps are located near
the site, on the upper terrace of the
Conecuh, where streams or creeks enter
the river.
Considering the apparent absence of any
Middle Woodland complex in the Conecuh
Region, it is reasonably clear that the
appearance of Weeden Island in this
region is not a product of local
cultural development. Rather, the
initial appearance of Weeden Island in
this region is probably a product of
site-unit intrusion from the north
Florida coast, specifically the western

Figure 3. Weeden Island Plain rim sherds from the Conecuh drainage.
Scale in centimeters.
Figure 4.
Carrabelle Punctate and Tucker Ridge-Pinched sherds from the
Conecuh drainage. Scale in centimeters.

Figure 5. Carrabelle Incised sherds and one Weeden Island Incised sherd
(lower left) from the Conecuh drainage. Scale in centimeters.
Figure 6. Weeden Island Red sherds (upper left and right, lower right),
Indian Pass Incised sherd (lower left) and unidentified stamped
sherd from the Conecuh drainage. Scale in centimeters.

Pensacola Bay area. By approximately
A.D. 400 the Pensacola Bay area may
have been approaching its maximum
carrying capacity at a hunting and
gathering level of technology supple
mented by small scale horticulture.
The addition of horticulture to the
economy could have initiated a popula
tion increase. A similar phenomenon
seems to have occurred along the
central Tombigbee River between A.D.
400 and A.D. 500 (Jenkins 1982). The
addition of a small amount of horticul
ture initiated population increase,
which stressed the natural environment
al resources. When this process was
repeated by succeeding generations,
environmental stress increased until a
cultural and settlement reorganization
occurred which intensified horti
cultural production. This process
signalled the beginning of Mississ-
ippian in the Southeast.
The intrusion of Weeden Island Culture
into the Conecuh drainage had a dramat
ic effect upon the subsequent ceramic
evolution of the upper Alabama and
Tallapoosa River Valleys. Interaction
between the Central Alabama and Conecuh
Regions was no doubt facilitated by the
overlapping of headwater creeks of the
Tallapoosa and Alabama with those of
the Conecuh. Just prior to A.D. 400
two spatially and morphologically
distinct complexes existed within the
upper Alabama and lower Tallapoosa
River Valleys. The Calloway Complex
occupied the lower Tallapoosa between
A.D. 300 and A.D. 500. Ceramically
this phase was defined by a contorted
plain ware tempered with sand and
intentionally added mica. Interaction
with Weeden Island groups ca. A.D. 400
caused the Calloway Complex to evolve
into a morphologically distinctive
ceramic complex, the Dead River Com
plex, which adopted several modes of
Weeden Island ceramic technology. Two
ceramic types comprise this complex:
Dead River Red Filmed is clearly a copy
of Weeden Island Red, while Kilby Plain
has a burnished surface similar to
Weeden Island Plain (Chase 1968, 1969).
The Cobbs Swamp Phase existed within
the upper Alabama River Valley between
approximately 100 B.C. and A.D. 400.
This is a Deptford-related complex
comprised primarily of check-stamped
pottery. As a result of Weeden Island
influence, shortly after A.D. 400 this
complex evolved into the Henderson
Complex, which is dominated by check-
stamped pottery, with a minority of
punctated designs that are clearly
crude copies of Weeden Island Punc
tated. The Henderson Complex also
contains a burnished plain ware with
rim forms similar to Weeden Island
Plain (Dickens 1971).
Mississippian Stage
(A.D. 1000 1540)
Mississippian occupation in the Conecuh
Region appears to have been relatively
sparse in contrast to the earlier more
numerous Weeden Island components. At
least two cultures of Mississippian
were recognized by this study. South
and west of Andalusia the Pensacola
Culture occupied the lower Conecuh and
its western headwaters. These compo
nents were indicated by the shell-
tempered Pensacola ceramics. Sherds of
the sand-tempered Fort Walton ceramics
were virtually absent. Pensacola sites
were relatively small, probably repre
senting farmsteads or hunting camps.
North of Andalusia, along the upper
Conecuh, the Mississippian occupation
was significantly different. It
appears to have been of the Roods-Lamar
continuum, although few diagnostic
artifacts were recovered. Mound and
village pairs were identified: one 3
km southwest of Luverne, another 24 km
southwest of Luverne, and another 24 km
southeast of Luverne.
The Conecuh Region was never a center
of innovation and change. Few cultural
(38, 1985)

developments seem to have occurred here
with a lasting impact on surrounding
regions. During the Archaic, the
conecuh Region was an important source
of Tallahatta quartzite. This is
attested to by its numerous large
quarries and reduction stations. From
around 500 B.C. until A.D. 400 (almost
1000 years) the Conecuh was uninhabit
ed. The reasons for this abandonment
are unclear, perhaps there was climatic
change and associated settlement
redistribution. Between A.D. 400 and
A. D. 1000 a large Weeden Island popula
tion inhabited the area. During this
period the large ceremonial center at
the Mitchell Site was occupied. The
Weeden Island occupation of the Conecuh
was significant in that this culture
had a substantial influence on the
cultural development within the adja
cent upper Alabama and lower Tallapoosa
River Valleys. Mississippian occupa
tion was small compared to that of the
Weeden Island period. The south and
western portion of the Conecuh Drainage
Basin was occupied by the Pensacola
culture, while the northern part was
occupied by the Roods and Lamar cul
tures. Proto- historic and historic
Indian occupation of the Conecuh Region
was very sparse. No definite Protohis-
toric or historic sites were encoun
tered by this study.
Ned Jenkins
Fort Toulouse/Jackson State Park
Route 6 Box 6
Wetumpka, Alabama 36092

Noel R. Stowe
University of South Alabama
During the last 284 years archaeo
logical data of varying quality have
been collected from hundreds of sites
on the north-central Gulf Coast. Some
250 of these sites are associated with
the Mississippi and Protohistoric
periods. In this area the most widely
used reference for these periods is
Gordon Willeys "Archeology of the
Florida Gulf Coast (1949). In that
monumental work he described the
Pensacola (shell-tempered) ceramic
series of the Fort Walton Culture
(Willey 1949:452). Subsequently these
assemblages have been referred to as
"Fort Walton-Pensacola" (Wimberly
1960:179); "Moundville, Fort Walton"
(Trickey and Holmes 1971:127); "Western
coastal variant" (Brose and Percy
1978:100), Pensacola phase (Stowe
1981:231), and "Pensacola Fort Walton"
(Tesar 1980a:143). With the exception
of Willey's "Pensacola series" and
Tesar's "Pensacola Fort Walton" none of
these constructs have been defined.
Some of this confusion in terminology
can be traced back to W.H. Holmes, who
in 1903 referred to the shell tempered
pottery found between Mobile and
Choctawhatchee Bay as "Mobile-Pensacola
Ware" (1903:104-114). Holmes' summar
ized the problem at the turn of the
century by stating: "Speculation as to
the peoples to whom these wares should
be attributed will for the present be
practically unavailing" (1903:105).
Holmes' uncertainty regarding the cul
tural placement of the ceramics un
earthed by his contemporaries is
certainly understandable. Unfortun
ately, recent researchers who have
incorrectly imposed or appended (some
times "in toto") cultural descriptions,
chronologies, or artifact taxonomies
and typologies developed for adjacent
sub-areas on this region are not as
easily understood.
This paper briefly describes the
Pensacola variant and the Bottle Creek
phase. My use of the term phase
follows the definition of Willey and
Phillips (1958:22). Use of the term
"variant" conforms to those definitions
produced by Lehmer (1971:32), Krause
(1977:10), and Jenkins (1982:10).
Lehmer defined variant as "a unique and
reasonably uniform expression of a
cultural tradition which has greater
order of magnitude than a phase, and
which is distinguised from other
variants of the same tradition by its
geographic distribution, age and/or
cultural content" (1971:31). In other
words, Kraus called it "a mid-range
taxon which has less content, greater
time span and greater spatial spread
than a horizon" (1977:10). Jenkins
successfully utilized the variant con
cept to replace "culture" in his
definition of the Moundville variant,
a regional manifestation of the
Mississippian tradition in the
Tombigbee and Black Warrior River
valleys (1982:10). Jenkins' Moundville
variant is "characterized as a series
of phases related by similar content,
and by spatial and temporal continuity"
(1982:10). The use of variant in
association with Pensacola seems espe
cially appropriate since, among other
things, it is culturally, spatially,
and temporally closely related to
Moundville. However, at the outset I
was somewhat reluctant to use the term
"Pensacola" for several reasons. As
noted, considerable confusion exists
concerning the use of the term.
Vol. 38 Numbers 1-2 Part 2 THE FLORIDA ANTHROPOLOGIST July, 1985

Willey initially used it as a ceramic
series and appended it to Fort Walton.
Subsequently, numerous authors have
noted the problem of including the
shell-tempered pottery assemblages with
Fort Walton (Willey 1949:463-464, Sears
1959, Lazarus 1971:40, Fuller and Stowe
1982). The term Pensacola implies
affiliation with a historic tribe of
the same name. While this may be true
in some cases it is certainly not in
others. The town of Pensacola,
Florida, is geographically located near
the south-eastern edge of the distri
bution of this variant. Finally,
sufficient data exist to divide the
assemblages heretofore called Pensacola
culturally chronologically and spa
tially into at least two and possibly
four phases. However, the long
standing usage of the term overrode
these objections, and Pensacola it is.
As it now stands Pensacola contains two
phases, Bottle Creek (circa. A.D. 1200-
1450, Middle Mississippi Period) and
Bear Point (circa. A.D. 1450-1700, Late
Mississippi and Protohistoric). It may
be desirable, in the future, to include
the coastal Louisiana Bayou Petre phase
(Kniffen 1936; Gagliano et al. 1979;
Davis 1981) in this variant.
Geographically, the Pensacola variant
extends across the north-central Gulf
Coast from Choctawhatchee Bay in
northwest Florida westward to the
Mississippi Gulf Coast, and northward
to the Tombigbee at least to Jackson,
Alabama, and up the Alabama perhaps to
Camden (Figure 1). Temporally it
begins during the Middle Mississippi
period (about A.D. 1200) and lasts
through the late Mississippi and
Protohistoric periods until A.D. 1700.
The 98 traits identified for the Bottle
Creek phase and 87 traits for the Bear
Point phase define the Pensacola vari
ant (See Table ). As one would expect
given the continuity from Bottle Creek
to Bear Point there is an overlap of
some traits. These traits include 22
pottery types and varieties for Bottle
Creek and 14 types and varieties for
Bear Point. Descriptions of these
ceramics and their relationships to
Moundville, Fort Walton, Bayou Petre,
and the Natchez phase are published
elsewhere (Fuller and Stowe 1982). The
majority of the types are Pensacola,
D'Olive, and Moundville series cera
mics. Perhaps the most diagnostic
ceramic attributes are the Gasque and
D'Olive Rim modes for Bottle Creek and
the Bear Point and Douglas rim modes
for Bear Point.
Environmental Setting
The area's climate is mild, with a mean
annual temperature of 68 F (20 C.)
and a growing season of approximately
270 days. The average annual rainfall
is 60 inches (155 cm). Seasons with
the greatest rainfall are winter and
summer, and the driest time of the year
is the fall. The Alabama and Tombigbee
Rivers and the Mobile-Tensaw Delta
often flood during late winter and
early spring.
Bottle Creek and Bear Point phase
components are found on the shores of
the barrier islands, bays, rivers,
bayous, creeks, swamps, and marshes of
the north-central Gulf Coast. Cer
tainly more important than the Gulf
Coast biome to the Mississippian and
Protohistoric peoples who inhabited
this area were the micro-environments
which include oyster and Rangia clam
beds, the salt springs of southern
Clarke County, and the fertile, sandy
loam flood plains of the lower Alabama
and Tombigbee River valleys.
Mississippian components have not been
found in the upland pine barrens, even
though these areas may have been
visited for hunting and collecting.
Pensacola does not appear in south
western Alabama until around A.D. 1200.
We have no good evidence for the early

Figure 1. Geographic distribution of the Pensacola variant. Figure 2. Geographic distribution of the Bottle Creek phase.

Mississippian West Jefferson or
Moundville I phases in the area.
Moundville engraved pottery types are
rare, and the markers of the West
Jefferson phase are absent (Jenkins and
Neilsen 1974; Stephonaitis 1980:271).
A few sherds of Kimmswick Fabric-
Impressed, var. Langston (Jenkins
1981:70), a variety linked with the
Moundville variant to the north, are
found in southern Alabama associated
with Late Woodland McLeod and Tensaw
Lake types. These associations at the
salt works in Clarke County and the
Boggy Gut site in the northern Mobile
Delta are probably the earliest
evidence for Mississippian in the area.
The first Mississippian people to enter
the region may have done so to obtain
salt, an important dietary supplement
for a population engaged in agriculture
(Fuller, Fuller and Stowe 1984). Other
later Moundville III types, Moundville
Incised, var. Snows Bend and and
Carthage Incised are also present at
the salines.
Five geologically and culturally stra
tified sites with Pensacola components
have been tested or excavated:
Bryant's Landing (lBal76, Trickey and
Holmes 1971:116); Hubbards Landing
(lBal81, Stowe, et al. 1981:173);
Middle River (lMbl07, Stowe, et al.
1981:217); the Douglas Mound (lCk217,
Fuller, Fuller and Stowe 1984:65); and
Vessel Point (lBa37). Without excep
tion, Mobile Cord-Marked, Furrs Cord-
Marked, or McLeod pottery types,
associated with the terminal Late
Woodland Tensaw Lake or McLeod phases,
are found stratigraphically immediately
below or mixed in the lower levels of
these sites with Bottle Creek pottery
types. A similar situation occurs with
late Weeden Island Wakulla Check-
Stamped occurring below Fort Walton in
northwest Florida.
Some 25 C-14 dates exist for the Mobile
Bay-Delta-Lower Tombigbee region. The
earliest date we have for Pensacola is
A.D. 1295 (SI-5348) for lCk210 (Peavys
Landing, Brose, Jenkins and Weisman
1983:326), a Bottle Creek Phase site on
the lower Tombigbee. Four additional
dates, A. D. 1435 (1-7658), A.D. 1400
(1-7657), A.D. 1430 (1-7142) and A.D.
1590 have been obtained from the
Dauphin Island Shell Midden (lMb72,
Bland 1981). These dates are probably
associated with the Bear Point
component at that site. We also have a
date of A.D. 1345 on a Mississippian-
style dugout canoe from Peavy's Landing
(Stowe 1974:199). The Bear Point Phase
has been coin-dated at A.D. 1560-1570
(Fuller, Fuller and Stowe 1984; Stowe
Bottle Creek Phase (Ca. A.D. 1200-1450)
This phase extends from extreme north
western Florida (it does not go
eastward of Choctawhatchee Bay) west
ward along the Alabama and Mississippi
Gulf Coasts perhaps as far west as the
Mississippi River Delta, northwestward
up the Tombigbee as far as Peavy's
Landing above Jackson, and north
eastward up the Alabama perhaps to
Selma (Figure 2).
Site types associated with this phase
include coastal oyster and fish bone
middens; brackish water Rangia, fish
bone, and mammal bone middens; mussel,
mammal, and fish bone middens on the
lower Alabama and Tombigbee; camps;
salines in Clarke and Washington
Counties; and single and multiple mound
This phase is named for the Bottle
Creek site, a large ceremonial center
and village located in the center of
the Mobile-Tensaw Delta. The site,
which covers about 50 ha, contains 15
mounds. The largest, constructed of
sand and clay, is rectangular, flat-
topped, 14 m high and 97.2 m by 76.3 m
at the base. Also present at Bottle
Creek are borrow pits, canals, a
stockade, and possible burial mounds.

The site has attracted sporadic
attention since the early 1700's when
Bienville, Governor of Louisiana,
visited it and made a collection. In
the early 1850's Andrew Bigelow des
cribed the site, made a sketch map, and
published his observations (1853:188-
189). The first and only controlled
excavations at Bottle Creek were
carried out in the 1930s by David L.
DeJarnette. An analysis of the field
notes and artifacts from those first
excavations was published 30 years
later by Nicholas Holmes (1963:23).
Middle Mississippi Horizon markers
occurring on Bottle Creek phase sites
include Kimmswick Fabric-Impressed,
var. Langston salt pan ware; Mound
Place Incised varieties; pottery rim
effigies (Fuller and Fuller 1984); and,
Southern Ceremonial Complex Markers
(Stowe 1984). The pottery types
associated with the Bottle Creek phase
include varieties of Mississippi Plain,
Bell Plain, Pensacola Incised, D'Olive
Incised, Mound Place Incised, Kimmswick
Fabric-Impressed, Salt Creek Cane
Impressed, Moundville Incised, and
rarely Moundville Engraved, Lake
Jackson Plain, and Fort Walton Incised.
Other traits include specific pottery
modes, shapes, and appendages; stone,
bone, pottery, shell, and copper
artifacts; a double-ended platform-
style dugout canoe and specific burial
types, structures, and site types.
Subsistence Activities and Settlement
Subsistence activities during the
Mississippian and Protohistoric periods
in the Mobile area have been described
elsewhere (Curren 1976; Knight 1984).
Curren (1976) based his study on
previous archaeological work (primarily
at Bottle Creek and the D'Olive Creek
sites) and ethnohistorical records,
including the Lincecum manuscript.
From these data he developed a model
that described a seasonal round that
alternated between villages and col
lecting stations "with large numbers of
people converging at centers like
Bottle Creek at appropriate times of
the cycle, such as the communal Green
Corn Ceremony" (1976:77-78). In my
opinion his use of the Lincecum
material and extending the green corn
ceremony described for the historic
period back into the Middle Mississippi
Period is questionable. It is clear,
however, that the river and delta
bottoms are flooded and inhabitable
during portions of the year. But there
is some evidence for a winter-spring
occupation of the coastal oyster
middens, such as the presence of drum
and sheepshead fish bones and menhaden
scales in those sites. These are fish
that come in to spawn during the winter
months. Also, most oyster shells in
these middens have very thin edges,
which is indicative of growth that
occurs during fresh water intrusions on
the oyster reefs along the coast in
late winter and early spring.
There is little evidence for agri
culture in either the Bottle Creek or
Bear Point phases. A handfull of
carbonized cobs, possibly associated
with Bear Point, from the D'Olive Creek
site and a few cob-marked shell-
tempered sherds from the upper Delta
have been recovered. However, the
absence of evidence for agriculture is
probably due to the lack of excavation
at Pensacola camps and villages on the
lower Alabama and Torabigbee Rivers.
At this time a model proposing a north-
south seasonal round from lower rivers
and delta to the coast in the winter
has some validity.
The Bear Point phase of the Pensacola
variant is discussed in the following
paper by Fuller.
Noel R. Stowe
Department of Sociology and
University of South Alabama
Mobile, Alabama 36688
(38, 1985)

Pottery Type
Mississippi Plain, var. Warrior (Steponaitis
1978:15; 1980:94, 436; Puller and Stowe
Bell Plain, var. Hale (Steponaitis 1978:15;
1980:95, 427-429; Fuller and Stowe 1982:50)
Pensacola Incised (Willey 1949:464)
Pensacola Incised, var.
Stowe 1982:77)
Pensacola Incised, var.
and Stowe 1982:77).
Pensacola Incised, var.
and Stowe 1982:77)
Pensacola Incised, var.
and Stowe 1982:80)
Pensacola Incised, var.
Stowe 1982:75)
Gasque (Fuller and
Jessamine (Fuller
Pensacola? (Fuller
Perdido Bay? (Fuller
Holmes (Fuller and
D'Olive Incised, var. D' Olive (Fuller and
Stowe 1982:56)
D'Olive Incised, var. Mary Ann (Fuller and
Fuller 1982:60)
D'Olive Incised, var. Dominic (Fuller and
Stowe 1982:58)
Vessel Appendages
Human Head Adornos (Bun and Forelock, Rattles)
Animal Head Adornos ("cookie cutter" Style Birds
Crested Birds, Owls, Fish, Feline, Canine/Bear
Animal Tail Adornos
Rim Nodes
Rim Flanges
Docorative Rim Nodes
Loop Handles
Strap Handles ?
Other Artifacts
Pottery Discoidals
Stone Discoidals
Shell Pins
Shell Beads
Busycon Shells
Bone Awl6
Bone Needles with Eyes
Stone Abraders
"Recycled" Projectile Points
Gar Scale Projectile Points
Greenstone Artifacts
Cooking Stone6
Antler "Points"?
Cooper Artifacts?
Ear Spools
Platform Dugout Canoes
Mound Place Incised, var. Waltons Camp (Fuller
and Stowe 1982:66; Willey 1949)
Salt Creek Cane Impressed, var. Salt Creek
(Fuller and Stowe 1982:83)
Langston Fabric Marked (Heimlich 1952:26;
Wimberly 1960:185-188)
Moundville Engraved (Steponaitis 1980:98, 437;
Jenkins 1976:228; Willey 1949:466; Wimberly
1960; McKenzie 1964; 1965; Fuller and Stowe
Moundville Engraved, var. Wiggins (Steponaitis
1978:19; 1980:445)
Moundville Incised, var. Snows Bend (Steponaitis
1978:16; 1980:104, 448)
Vessel Inverted over Skull
Extended Burials
Bundle Burials
Flexed Burials
Secondary Burials
Turtles and Amphibians
Other Mammals
Moundville Incised, var. Moundville (Steponaitis
1978:16; 1980:104, 448; McKenzie 1966)
Moundville Incised, var. Bottle Creek (Fuller
and Stowe 1982:63)
Moundville Incised, var. Carrolton (Steponaitis
1978:16; 1980:104, 447)
Lake Jackson Plain (Willey 1949:458) (Rare)
Fort Walton Incised (Willey 1949:460) (Rare)
Rectangular Post Mold Pattern
Daub With Cane Impressions
Site Locations
Gulf Coast Strip
Mobile Bay
Mobile-Tensaw Delta
Alabama River
lower Tombigbee River
Upland Sites (Rare)
Pottery Traits
"D'Olive Rim Mode (Fuller and Stowe 1982:84)
"Gasque" Rim Mode (Fuller and Stowe 1982:82)
"Black Filming" (Fuller and Stowe 1982:86)
Vessel Shapes
Globular Bowl
Shallow Bowl
Salt Pan
Beaker (Rare)
Structural Effigy (Rare)
Site Types
Oyster Middens
Rangia Middens
Mussel Middens
Rectangular Temple Mounds
Oval Temple Mounds
Burial Mounds
Hunting and Collecting Stations
Salt Collection Stations
Southern Ceremonial "Cult" Complex
Multiple and Single Mound Centers
Sites Seasonally Occupied
Ranked Social Organization


Richard S. Fuller

Mobile, Alabama

For a long time after Gordon Willey
defined the Pensacola series it was
treated much like an unwanted step-
child: either abused or ignored.
Although Willey and, subsequently,
others recognized the distinctive
origins and content of Pensacola, it
was often subsumed under the rubric of
Fort Walton or Moundville. This has
resulted in deformation of the latter
concepts to the detriment of what may
now be viewed as a distinct variant of
Mississippian culture with a traceable
developmental history and definable
geographic limits.

In the last few years a stronger case
has been made for the concept of
Pensacola as a discrete culture-histor-
ical unit by Sears, Jenkins, Knight,
and others, and, more recently, by
Stowe, and myself. Read Stowe (this
issue) presented our current perception
of Pensacola as a regional
Mississippian variant and defined its
Middle to Late Mississippi manifesta-
tion, the Bottle Creek phase. Its
protohistoric manifestiation is the
Bear Point phase.

Prior to 1981 our working concept of
Pensacola as an archaeological entity
amounted to little more than Willey's
Pensacola ceramic series. Based
largely on his 1949 typology, which had
been modified somewhat by Wimberly in
1960, the series covered a temporal
span, roughly A.D. 1100-1700, that was
too broad for investigating culture
change, population dynamics, settlement
patterns, or the various other funda-
mental aspects of an archaeological
culture. Progress toward sub-dividing
Pensacola into more discrete and
workable units, ideally complexes or
phases, was given an unintentional
boost three years ago by the misguided
shovel of a relic hunter.

Vol. 38 Numbers 1-2 Part 2


This shovel, a double-edged spade if
you will, found its way into the Pine
Log Creek site (1Ba462), a proto-
historic, Pensacola-affiliated cemetery
located in the Alabama-Tombigbee
confluence basin. Through the concert-—
ed efforts of the land-owner, the State
of Alabama, and concerned professionals
and lay people, the looted artifacts
were recovered for analysis and cura-
tion before their very imminent and
irretrievable dispersal into the
collectors market. Thus, intercepted
enroute from obscurity in the ground to
obscurity in the colector's case, this
important collection of aboriginal and
sixteenth century European artifacts
provided a much needed datum for
investigating the Pensacola variant.
The aboriginal ceramic fraction,
comprising a nearly pure late Pensacola
assemblage, formed a nucleus around
which the definition of the Bear Point
complex and, now, the Bear Point phase
has coalesced (Figure 1).

Site 1Ba462 consists of a series of
sandy knolls or small mounds resting
upon a natural levee bordering Pine Log
Creek, a tributary of the extreme lower
Alabama River. Ecologically, the site
lies just within the northern limits of
the Mobile-Tensaw Delta, a complex
system of streams, ponds, lakes, levees
and swamps with a large and diverse
suite of readily available subsistence
resources. Archaeological research has
shown that the Mobile-Tensaw Delta
supported a viable and relatively
stable aboriginal population from Late
Archaic to Historic times.

Qur examination of the site indicated
that most or all of the looted materi-
als came from the rise referred to as
mound no. 1. Without evidence derived
from controlled excavation we are
uncertain if we are dealing with true

July, 1985



Figure 1. Geographic distribution of the Bear Point phase.
Ft. Stoddard
Doctor Lake
Late Lithic
Bear Point
^ Island )
Burial Urn
Bottle Creek
Moundvi 11 e ?
Boggy Gut?
Salt Pan
Figure 2. Late prehistoric-historic aboriginal sequence in
the Alabama-Tombigbee Confluence Basin.

burial mounds or slightly enhanced
natural knolls used as burial sites;
thus, the term mound is used advisedly.
Although we lack strict contextual
data, informant interview and the
collection itself suggested multiple
burials were present in the mound.
Secondary interment, including burial
in pottery vessels (urn burial) was
common, and at least one extended
burial also occurred. The ceramic
assemblage, including some 30 vessels
and over 4000 sherds, could largely be
classified according to the existing
Mississippian types: Bell Plain,
Mississippi Plain, Pensacola Incised,
D'Olive Incised, and Moundville In
A key to deciphering the cultural
context of this assemblage was the
recognition of a number of vessel
shapes and design motifs which repli
cated those in early collections from
Bear Point, (IBal), a cemetery located
on the Alabama Gulf Coast on Perdido
Bay. The Bear Point site produced a
series of pots, illustrated by W.H.
Holmes and C.B. Moore, that played a
major role in Willey's original defini
tion of the Pensacola series. Bear
Point also yielded European artifacts,
including a sixteenth century Spanish
coin, which placed at least part of the
mortuary activity in the Protohistoric
period. Significantly, the salvaged
collection from Pine Log Creek also
contained European artifacts, a number
of which can be dated to the sixteenth
or early seventeenth century.
Also crucial to dating the Pine Log
Creek ceramics was the observation that
there were certain stylistic relation
ships to Moundville IV, Alabama River
phase, and late Fort Walton pottery,
all of which date to the Protohistoric
or Early Historic period. It was
further noted that there were several
decorative elements and motifs with
analogs in protohistoric Natchezan
pottery types in the Lower
Mississippian Valley, specifically
Leland Incised, Fatherland Incised and
Maddox Engraved.
All this direct and indirect evidence
of a protohistoric date for the Pine
Log Creek assemblage prompted the
definition of a late Pensacola ceramic
complex which could best be expressed
as varieties and modes within the
existing Pensacola series. Thus we
established a number of provisional
varieties, naming the complex "Bear
Point after the better known site.
Additional surveys and further reviews
of the archaeological literature
provided a somewhat firmer base for
most of the varieties and modes and
yielded further information on distri
bution and dating. The Bear Point
complex has therefore been promoted to
phase status with the understanding
that much "fleshing-out" is still
needed. The phase is believed to
correspond roughly to the Protohistoric
period, or about A.D. 1450 to 1700.
Currently, the Bear Point phase is
still defined primarily by ceramics
(see Table) including five types, nine
varieties, and four vessel shape modes.
All of the pottery is shell-tempered,
indicating its membership in the
Mississippian ceramic tradition. This
membership is reinforced by the occur
rence of the Mississippi Plain/Bell
Plain, or coarse ware/fine ware duad
and by the persistence of certain
vessel shapes, particularly the globu
lar utilitarian jar. These elements
were undoubtedly inherited from the
earlier Bottle Creek phase which, in
turn, received them from the Moundville
variant in central Alabama. Part of
this heritage also includes Moundville
Incised arch motif and free-standing
Southern Cememonial Complex design
elements, specifically hands and skulls
which, by the Bear Point phase, had
undergone noticeable stylistic evolu
tion. Many traits, however, appear to
have been introduced from sources other
than Moundville. These include:
interior decorated plates and the
parallel-line scroll with triangular
filler motif, both exhibiting ties to
the Natchez Bluffs segment of the Lower
Mississippi Valley; and, the "pseudo-
casuek;" and flattened globular bowl
forms, reflecting influences from
(38, 1985)

the Alabama and Coosa River Valleys.
In addition to pottery, various non
ceramic traits, though less common,
have been recognized in the Bear Point
phase. These includes: plain circular
shell gorgets and variety of shell bead
types; polished stone discoidals and
celts, including the spatulate form;
and copper ornaments, including
arrowhead" style emblems. Together
these socio-technic artifacts are
further evidence of the Mississippian
heritage of the Bear Point phase.
New elements which entered the
Pensacola variant during the Proto-
historic period include European
materials, the best documented of which
are from Pine Log Creek. Among these
are faceted seven-layer chevron beads
which are excellent sixteenth century
markers. An equally good marker is a
brass, capstan-style candlestick which,
along with a small brass pail, may have
been part of a portable altar set.
Other items are dark blue glass beads
and a bead or earspool cut from Colum
bia Plain glazed earthenware. Also
recovered was an assortment of iron
atifacts, including sword fragments, a
gun barrel, an axe, a bridle cheek
plate, a pike head, a knife, a sickle,
a mule-or horseshoe, spikes, chisels,
and several re-worked items that may
have originally been chain links.
The quantity of European artifacts,
some 37 in all, indicates either
coastal shipwreck salvage by the
Indians or possibly prolonged contact
with a sixteenth century Spanish
expedition. A possible source is the
abortive colonizing effort of Tristan
De Luna. Some 1500 would-be colonists
spent nearly two years, from 1559 to
1561, in and around Mobile Bay,
Pensacola Bay and the Lower Alabama
River Valley. Toward the end of this
period the expedition was literally
starving and disheartened, and readily
parted with their possessions in
exchange for food. Significantly, four
silver coins, minted in Mexico during
the period of 1554 to 1570, have been
found in extreme southwestern Alabama.
Three of these came from sites with
strong Bear Point phase components.
We have recently completed a prelimi
nary survey in the Alabama-Tombigbee
Confluence Basin that focused upon the
location of Late Mississippian, Proto-
historic and Early Historic sites in an
effort to place Pine Log Creek and the
Bear Point phase in a firmer context.
The most important result of this
survey was the recognition of a cluster
of small Bear Point phase burial mounds
and habitation sites situated on
natural levees of the riverine flood-
plain. Urn burial appears to be common
in these mounds and the overall mortu
ary pattern is suggestive of Choctaw
burial practices during the Historic
period. Also, the distribution of
habitation sites seems to fit a dis
persed riverine settlement pattern
deduced from French acccounts of the
Choctaw-speaking Mobilians and Tomehs
who occupied the Forks region in the
early eighteenth century.
Typical of the burial mounds is
Ginhouse Island, (lWn86), located
within the historic domain of the
Tomehs. Test excavations at Ginhouse
Island have yielded urn burials and
Bear Point phase ceramics. This is
just one of a scattered group of at
least 10 such burial mounds in the
Forks region. We are provisionally
referring to this aggregate as the
Ginhouse Mortuary Complex which,
although part of the Bear Point phase,
appears to be somewhat more restricted
geographically. Burial in pottery
receptacles, which has not been report
ed from coastal Bear Point phase
sites, reflects participation in a more
wide-spread Burial Urn horizon. Often
referred to as the Burial Urn culture
or variant, it seems to have been
manifested in a number of distinct
phases and variants, including
Savannah, Lamar, Alabama River phase,
Moundville IV and the Bear Point phase,
all on a proto-historic time-line, thus

prompting our use of the term "hori
zon. It is interesting to note that
some of the earliest contacts between
European and Muskoghean groups occurred
in the areas represented by these
various archaeological constructs. It
is tempting to speculate tht the trait
of urn burial may have been somehow
stimualted by such contact.
At present, the known distribution of
Bear Point phase ceramics appears to be
centered in extreme southwest Alabama,
southeast Mississippi and northwest
Florida. It extends along the Gulf
Coastal strip from about Biloxi Bay on
the west to Choctawhatchee Bay on the
east (Figure 2). In the latter area it
overlaps with the Choctawhatchee
variant of the Fort Walton variant.
The Bear Point and Choctawhatchee
materials share some design and vessel
shape attributes in common but are
separable based on distinctive ware
characteristics: specifically, shell
tempering in the former versus sand
tempering in the latter. This distinc
tion may ultimately prove to be impor
tant not only archaeologically, but
ethnically since Choctawhatchee Bay may
have served as an interface zone
between proto-Choctawan grops and
Proto-Apalachees. In Alabama the Bear
point phase extends northward from
Mobile Bay, through the Mobile-Tensaw
Delta, into the Lower Alabama and Lower
Tombigbee Valleys. This distribution
may include all or part of the province
of Tascalusa of the de Soto accounts.
Thus, it is not inconceivable that when
the site of the embattled town of
Mauvilla is finally located it will
exhibit a Bear Point phase component.
As it now stands, the best candidates
for the Historic period heirs to the
Pensacola variant are the Mobilians,
Tomehs, Naniabas, Pensacolas and,
perhaps, the Pascagoulas. Certainly,
early French descriptions of the
locations of these "Eastern Choctaws"
correspond rather neatly with the known
distribution of the Bear Point phase.
So far, Bear Point ceramics have not
been found in association with French
artifacts or with any site known to
post-date the seventeenth century.
However, several ceramic complexes with
apparent early eighteenth century
associations have been provisionally
defined. Significantly, two of these,
the Doctor Lake complex and the Port
Dauphin complex, exhibit a limited
number of traits which could easily
have been inherited from the Bear Point
phase. The coarse ware/fine ware
dualism and the parallel-line scroll
motif found in both these and the Bear
Point phase seems to have survived even
later in the region, as represented by
an assemblage of Choctaw pottery from a
ca. 1800 fort site located on the
Mobile River. Further refinement of
the Bear Point phase and the Pensacola
variant will forge a much needed link
between prehistory and history.
Ultimately it will be a key to investi
gating aboriginal cultural and demo
graphic change in the face of European
exploration and colonization on the
North-Central Gulf Coast.
Richard S. Fuller
105 DuRhu Drive, Apt. 109
Mobile, Alabama 36608
(38, 1985)

Pottery Types
Mississippi Plain, var. Pine Log (Fuller and
Stowe 1982:54).
Mississippi Plain, var. Warrior (Steponaitis
1978:15; 1980:94; 436; Fuller and Stowe
Bell Plain, var. Hale (Steponaiti6 1978:15;
427-429; Fuller and Stowe 1982:50).
DOlive Incised, var. Arnica (Fuller and Stowe
D'Olive Incised, var. D'Olive (Fuller and Stowe
Pensacola Incised, var. Moore (Fuller and Stowe
Pensacola Incised, var. Pensacola (Fuller and
Stowe 1982:69).
Pensacola Incised, var. Perdido Bay (Fuller and
Stowe 1982:80).
Pensacola Incised, var. Bear Point (Fuller and
Stowe 1982:72).
Pensacola Incised, var. Gasque? (Fuller and
Stowe 1982:74)(Rare).
Pensacola Brusned ? (Lazarus 1961:56; Stacy
Mound Place Incised var. Waltons Camp ?
(Pnillips 1970; Fuller and Stowe 1982:65;
Phillips, Ford and Griffin 1951; Steponaitis
Moundville Incised, var. Douglas (Fuller and
Stowe 1982; Cottier 1970; Sheldon 1974).
Alabama River Incised (Fuller and Stowe 1982;
Cottier 1970; Sheldon 1974.
Other Pottery Types
"Bear Point" Rim Mode (Fuller and Stowe
"Douglas" Rim Mode (Fuller and Stowe 1982:84).
"Black Filming"
"Red Filming"?
Vessel Shapes
Collared, Flattened, Subglobular Bowls
Carinated Bowls
Casuela Bowls
Hemispherical Bowls (common)
Shallow Bowls (common)
Plates (common)
Cylindrical Bowls or Beakers (rare)
Bottles (rare)
Structural Effigies (rare)
Vessel Appendages
Loop Handles
Strap Handles
Decorative Rim Modes
Node Lugs
Flange Lugs
Small, Simple, Solid Rim Effigies
Other Artifacts
Pottery Discoidals
Stone Discoidals
Busycon Shells
Shell Beads
Shell Pins
Shell Ear Spools
Projectile Points
Spatulate Ax
"Recycled" Projectile Point6
Copper Artifacts
Copper "Symbol Badge" Arrowhead Headdress
Small, Battered Poll, "Greenstone" Celts
European Goods (16th/17th Century)
Spansish-Mexican Coins (Circa, mid-to-late
Glass Beads (Faceted Chevron)
Brass Items
Spanish Candlestick (Circa. 16th Century)
Iron Spikes/tfails
Sword Fragments
Gun Part6
Horse Furniture
European Ceramic Ear Spool
Iron Tools
Urn Burials
Wooden Chest Burial
Vessel Over skull
Extended Burials
Flexed Burials
Bundle Burials
Secondary Burials
Skull Burials
Other Mammal
Site locations
Coast Strip
Lower Tombigbee River
lower Alabama River
Mobile-Tensaw Delta
Mobile Bay
Site Types
Rangia Middens
Oyster Middens
Sand Temple Mound ?
Sand Burial Mound
Ranked Social Organization (Warrior Class)
Sites Seasonally Occupied

David S. Brose
The Cleveland Museum of Natural History
It is now almost a half century since a
graduate student named Gordon R. Willey
first traveled along the Florida Gulf
Coast. Willey came to see for himself
those prehistoric sites which had
attracted Moore; sites which yielded
ceramics of such interest to Holmes.
As graduate student projects go,
Willey's should be considered a suc
cess. He came. He saw. He conquered.
Indeed, the monument to his conquest,
the whole of volume 113 of the
Smithsonian Miscellaneous Collections,
has been regarded for most of the 35
years since its publication, as if
Willey had entitled it The Archeology
of the Florida Gulf Coast.
This is a valuable publication. Of its
first 200 or so pages nearly 160 are
detailed site excavation descriptions.
The last 400 pages are avowedly syn
thetic discussions of the eight culture
area/periods, each organized under nine
topics (see Willey 1949:IX). It is
worth remembering that with the limited
data with which Willey dealt, most of
his discussion for five of the nine
topics was necessarily based upon
rather intuitive extrapolations. These
were in part derived from accepted
historical or anthropological ana
logues, refined in light of the then
limited archaeological record from the
southeast. From word one Willey
(1949:1) appreciated that his effort
was a descriptive step in the direction
of fuller understanding, rather than
some statement of arrival. Zen rather
than Satori.
As he tried to make clear, his method
and his goal was to relate the prehis
toric pottery of west Florida
with chat known in Louisiana and
Georgia to create a chronological
reconstruction. The ceramics in
stratigraphic sequence, were grouped
into sub-areas, and "In the northwest,
each of the culture periods in that
region was first established upon the
presence of certain pottery types and
upon the percentage configuration of
these types in a particular strati
graphic context...It will be seen from
this that the cultural periods, as they
were first defined, were essentially
ceramic periods" (1949:xix, 1, 4-5).
After describing what he meant by
"type" (based upon decoration) and
"series" (paste and temper), Willey
described the "ceramic complex" as an
inclusive group of types or series
which occur together in the same
general area at the same time. He
remarked that while the actual situa
tion for any type or series was chrono
logically, spatially and culturally
dynamic, so that they may have over
lapped sequential complexes, "...they
have, for purposes of description and
analyses, been made static and classi
fied as components of one complex or
the other" (Willey 1949:6, emphasis
This is not an attack on Willey's work.
However, I do not feel it unfair to
look critically at some of Willey's
cultural periods in the northwest
Florida with the advantage which two
generations of other archaeologists
have obtained. Indeed, this symposium
suggests reevaluation is long overdue.
I wish to focus not on what Willey made
of his ceramic culture, but on what we
have done (or rather, should have done)
with these temporary and artificial
constructions (cf. Gardin 1980) such as
Deptford, Santa Rosa/Swift Creek,
Vol. 38 Numbers 1-2 Part 2 THE FLORIDA ANTHROPOLOGIST July, 1985

Weeden Island I, and Fort Walton.
Continued use of the super term
"Deptford" has maintained an illusion
that there were really peoples living
along the northwest Florida coast with
a culture resembling that of the south
Atlantic coast and islands where
Deptford was born. (I call Deptford a
super type as its application seems to
have been faster than comparative
analysis; more powerful than a dated
component; and able to leap state
borders in a single bound).
Since its appearance in that area,
Deptford has referred to a ceramic
series, a set of chronological phases,
and a cultural tradition. We can now
recognize that these three differing
archaeological constructs are of
varying duration and distribution,
based upon data unavailable in 1949.
But for Willey (1949:353), there was
less ambiguity:
The Deptford period is defined
by the appearance and numer
ical predominance of the pottery
types Deptford Linear Check Stamped
and Deptford Bold Check Stamped.
These types are often accompanied
by Deptford Simple Stamped.
Deptford is the earliest established
period for the northwest coast
region of the Gulf Coast area and in
Florida is known only for this
If we are satisfied to use the term
Deptford to mean little more than a
millennial "Southeast Coastal Woodland"
tradition (Sears 1963, 1959; Milanich
1973b; Milanich and Fairbanks 1980), we
will probably not do much harm, but
probably not accomplish much good
During the past decade, I have devel
oped misgivings about the existence of
even an early Deptford Ceramic Complex
in northwest Florida to which other
cultural behavior could be tied. I do
not mean to suggest that Deptford
pottery is the result of spontaneous
generation, but rather, that the
Deptford ceramic series by which Willey
identified a Deptford period in fact
represents several periods.
Willey was not unaware of this possi
bility when he stated that (other than
several sorts of unclassified or plain
sherds) the most common type, Deptford
Bold Check-Stamped, was difficult to
distinguish from other check-stamped
types and thus a poor period marker.
He also stated that the chronological
position of Deptford Simple-Stamped was
not clear in Florida, as it seemed to
coexist with Deptford Linear Check, was
quite consistently found with Santa
Rosa/Swift Creek, and occasionally
occurred in Weeden Island Period
contexts (1949:355-358).
There were also a few Deptford Series
ceramic types Willey reported which
seem to have been either associated
with, or which actually were Alexander
Incised and St. Simons fiber-tempered
ceramics at the lower levels of those
very few sites where he postulated
evidence for this Deptford Period.
Sears (1963) working with similar
stratigraphically mixed midden deposits
at the Tucker site, was forced to deal
with them in a similar methodological
Perhaps we should really take the time
to look at those Deptford period sites
which Willey discussed, whether he
found them in (or on) the ground, or in
earlier collections (or reports of
collections). A detailed review of the
data is not reassuring. Of Willey's
ten sites, only Carrabelle provided
anything like stratigraphic control.
Of sites subsequently assigned to this
Deptford Period, Sears' (1963) excava
tions in the Tucker midden were all coo
similar to those at Carrabelle. At
both sites Deptford ceramics (however
arbitrarily defined) were found mixed
throughout the deposits along with what
assumed to be later types of Swift
Creek, Santa Rosa and Weeden Island
pottery (dated as late as A.D. 345).

At Carrabelle, Tucker (Phelps 1966),
and the Alligator Lake Site (Lazarus
1965), similar Deptford ceramics were
apparently associated with Poverty
Point objects and perforators, and
with fiber-tempered Orange or St.
Simons, Norwood, and/or unmistakably
Alexander and Bayou la Batre ceramics,
respectively dated (with occasionally
loose proveniences) between 1200 B.C.
and 200 B.C. locally.
Now, either the deposits were to some
degree mixed, and these dates are all
to be ignored, or we must recognize
that we have a Deptford period whose
few diagnostic ceramic criteria extend
over a least half the time during which
ceramics were made here. I am afraid
we are dealing with the latter situa
tion. Despite the often stated need to
create phases within this long lived
Deptford tradition, (Sears 1963, 1977;
Milanich 1971, 1973; Milanich and
Fairbanks 1980), and the vague acknowl
edgement that early Deptford is not
late Deptford, or coastal Deptford is
not interior Deptford, or Gulf Coast
Deptford is not Atlantic Coast
Deptford, we appear to have accepted
what Willey explicitly wished to avoid.
Our Deptford is less a chronological
episode than an archaeological process.
It is certainly not a cultural period,
and we generate little but confusion
when we simplistically match archaeo
logical questions from one Deptford to
archaeological answers from a different
Deptford. The alternative approach
(Tesar 1980a (1):67-108), with multiple
Deptfords forced into variable interac
tions with every aboriginal culture
from Poverty Point to late Ohio
Hopewell, becomes needlessly confusing
and in fact tends to obscure the
dynamic regional relationships which
existed through time. We may even end
up by postulating evolutionary sequenc
es between contemporaneous events.
If ever the archaeological record of
northwest Florida presented us with a
need to resist the temptation to change
what Willey meant a period to be, it is
here with Deptford. If it is not to be
a cultural period, it must not be
treated as such.
Santa Rosa/Swift Creek and
Weeden Island I
Since Willey's 1949 formulation of the
Santa Rosa/Swift Creek period as an
organizing structure, it has become
increasingly clear that its juxtaposed
components were in serious structural
disequilibrium. The Santa Rosa and the
Swift Creek ceramic series are now
known to display significantly differ
ing distributions, both in time and in
space. Their variable and peripheral
overlap in northwest Florida has been
of unquestioned value for aligning
chronological sequences in regions as
disparate as southwestern Georgia and
southwestern Alabama. Nevertheless,
after half a century, interest has
shifted to understanding the internal
and external behaviors of societies
which existed in this region of the
Gulf Coast. It was this point which
motivated Sears (1954, 1962) to super
impose the Yent and Green Point ceremo
nial complexes atop Willey's Deptford
and Santa Rosa/Swift Creek ceramic
Willey of course had specific chrono
logical limits for Deptford, far more
restricted (and earlier) than these of
Sears. As I have noted (Brose and
Percy 1974:3, 8) and more recently
probably overemphasized (Brose 1979),
Sears' construction was significantly
flawed not so much by his (1954) paleo-
sociological interpretations (which I
rejected though I confess to no more
evidence than Sears had), but by his
excision of Kolomoki from the transi
tion into earliest Weeden Island. In
logical consequence Sears was led to
deny the existence of Crystal River as
a complex, leaving it uncomfortably
astride Weeden Island: half early,
half late, and wholly confusing.
Crystal River and Kolomoki now show
radio-carbon equivalence. Their
(38, 1985)

socioceremonial integrity (and that of
sites of coeval duration such as Bird
Hammock, Aspalaga Landing, Pierce Mound
A, early to middle McKeithen, et al.,
must be seen as the apogee (not culmi
nation) of a cultural complex which
from perhaps A.D. 100 to about A.D. 400
displays far more internal cohesion and
similarity in external connections than
would be suggested by the borrowed
ceramic terms in whicn it has been
The erratic and nearly disjointed
references to various Ohio and Ohio
Valley Hopewellian phenomena, to the
Crystal River Complex, to Porter, to
early to Late Marksville, and to
various presumptive phases of early,
middle and late Swift Creek, to say
nothing of the confusing culture area
maps which bedeck various modern
syntheses (e.g. Sears 1977; Milanich
and Fairbanks 1980: 89-93, 114-129;
Tesar 1980a: 72-108), bear witness to
the fact that the Santa Rosa/Swift
Creek ceramic period of the 1940s is
simply irreconcilable with archaeologi
cal data of the 1980s- What an indict
ment of all of our efforts it would be
if this were not true.
Willey noted that there were both pure
Santa Rosa/Swift Creek mounds, and
Swift Creek/Santa Rosa and Weeden
Island I possibly accretional mounds
(1949:370). He also discussed a few
(very few) pure middens, which I have
reviewed (Brose and Percy 1974, Brose
1979). Indeed, Willey's recognition of
the lack of clear ceramic distinctions
between the Santa Rosa/Swift Creek and
Weeden Island I periods is best seen in
his own interpretation of the Mound at
Eleven Mile Point (Willey 1949:276),
where the ceramics spanned the heuris
tic break between these cultures,
cultures which he had explicitly
separated for preliminary analyses.
This was the same interpretation Willey
Offered for the Crystal River Mound
assemblage (1949:320-322), adding that
he felt "that there is little doubt
that Crystal River, Yent, and Pierce
Mound A are all part of the Santa
Rosa/Swift Creek Period (1948;
1949:323). Without (I hope) getting
too sidetracked, it seems worth looking
first at the long Georgia Swift Creek
tradition, to determine just which part
supposedly co-occurs with Santa Rosa,
and then at Santa Rosa ceramics them
selves and tneir neighbors in Mobile
Bay, to discover whether they might be
more precisely ordered.
First, it must be recognized that the
distinction of "Early" from "Late"
Swift Creek is based on different
things for different people. For
Caldwell and Sears the stamped motifs
were the key. For Kelley and Smith,
the rim and basal modes were the most
diagnostic criteria. For Willey the
decorative field was the basis for the
first cut, the rim and basal modes
next, and the stamp motif the last
Smith's statement that the Swift Creek
Mound A was early to middle Swift Creek
seems to have been based on the absence
of either Napier or Lamar materials
within the mound (Smith and Kelley
1976:31-34). Thus, her "Late Swift
Creek" seems later than what in north
west Florida terminology (Percy and
Brose 1974) is Weeden Island 3 or 4.
The Early Swift Creek at Mandeville,
securely dated to between A.D. 245 and
A.D. 420 (Smith 1975, 1979), she seems
to define by the complicated stamped
motifs. Thus when Smith stated
(1975:125-131) that "...most of the
material from these sites (in southeast
and south central Georgia) appears to
be Late Swift Creek she follows
immediately with a placement of A.D.
500-700 for Middle Swift Creek.
Indeed, every site she reported for
that misty region seems so late that it
cannot relate to the Santa Rosa/Swift
Creek of northwest Florida (1976:197-
224). In addition, Snow (1977) has
proposed that these sites and others at
which he has worked be assigned to a
lingering Deptford III sub-phase, which
would be temporally equivalent to

Willey's Weeden Island II.
If looking north only complicates the
placement of Florida Middle Woodland,
perhaps looking west will prove more to
the point.
Jenkins recently suggested that,
although they have been called Porter
Hopewell and Porter Marksville, the
ceramics of the Porter complex are in
fact a local manifestation of the Santa
Rosa series (Brose, Jenkins and Weisman
1983:129). But, based upon Jenkins'
excellent analyses of ceramic data
compiled for that Black Warri-
or/Tombigbee study, it makes far better
sense to consider the Santa Rosa
ceramics as a local manifestation of a
late Bayou la Batre-Porter continuum.
In northwest Florida the earliest
decorated Santa Rosa series ceramics
seem to have been Santa Rosa Stamped,
Basin Bayou Incised, and Alligator
Bayou Stamped, possibly in that order
of appearance (Willey 1949:75-95; 372-
378). I believe Willey was indeed
correct in his stratigraphic intuition.
The design configurations represented
by these ceramics predate the appear
ance of early Swift Creek Complicated
Stamped motifs on the Florida Gulf
Coast, therefore they must be seen as
an introduced stylistic concept,
without known local precedents.
The Santa Rosa series must, therefore,
have originated as part of some other
developmental sequence, certainly one
in which paddle stamped decoration was
of little importance. This is, of
course, the Gulf Tradition (Walthall
and Jenkins 1976) generally. Specifi
cally, I offer as prototypes those
ceramics from late Bayou la Batre
and/or early Porter levels of sites in
the Lower Tombigbee river valley and
the Mobile Delta and Bay area. These
levels date consistently to the period
between 200 B.C. and A.D. 150 (Brose et
al. 1983).
The conclusions from such regional
review data question the validity of
Willey's trial construct. Santa Rosa
series ceramics develop from late Bayou
la Batre in southwest Alabama prior to
A.D. 1. Only after interaction with
Marksvillian ceramic technology do they
spread east into northwest Florida.
When they appear, later in our region,
they are NOT in mound contexts, and
they carry decorative motifs and
designs unambiguously related to those
on ceramics which are late in Early
Marksville, about A.D. 100 (Toth 1974,
1979; Belovich et al. 1983:412). I
believe this is prior to even the
earliest Swift Creek ceramics on the
Florida coast where, as in Georgia,
they are nearly overwhelmed by their
association with plain and check-
stamped ceramics. It is not until well
after A.D. 100 that northwest Florida
shows strong evidence for an ex
change/mortuary complex. When it
occurs, the exchange/mortuary complex
has almost none of the Santa Rosa
series ceramics. Indeed, it has both
Early and Late Swift Creek ceramics
(identified by their "Weeden Island
Rims). Those assemblages also have
what Willey (1945) called the Crystal
River Series, which is consistently
associated with Weeden Island plain and
decorated types.
Santa Rosa, while a useful ceramic
series, is merely an ephemeral influ
ence from Alabama Porter, not a local
cultural period. Swift Creek is a
Georgia ceramic tradition, hardly a
cultural period even there. In north
west Florida the juxtaposition of
Porter and Swift Creek results in a
change to both. From A.D. 100 to A.D.
300 Porter ceramics are adapted into
Crystal River Series, and somewhat
later into the Weeden Island decorated
types. These Gulf Formational ceramic
concepts are accepted into the Woodland
tradition, resulting in curvilinear
stamped motifs (Early Swift Creek), and
ultimately in the overall adoption of
the Weeden Island rim morphology (late
Swift Creek). Porter and Weeden Island
concepts of design field were used to
limit application of Swift Creek
(38, 1985)

decorative techniques (Late Swift
Santa Rosa is rather late when seen
from Mobile or from Mississippi (see
also Brose 1984a). In west Florida it
occurs with both Late Deptford and
Early Swift Creek, and it basically
predates the Midwestern Middle Woodland
interaction. It is difficult to decide
how much we may legitimately use
ceramic criteria to separate Late Swift
Creek from Early Weeden Island as
cultural periods. Indeed, Willey's
original 1949 ceramic periods ought to
be superceded by incorporating what we
have since learned into a better
configuration, one reflecting more
cultural behavior than seen in the
presumptive promiscuity of potsherds.
Fort Walton
The final prehistoric ceramic/cultural
period Willey constructed for the
northwest coast was unfortunately
called Fort Walton, unfortunately
because the type site was at best only
partly Fort Walton. It was the very
example used in my earlier conjecture,
of which I remain increasingly suppor
tive (Brose and Percy 1978:100-103),
that in northwest Florida Pensacola is
a Moundville-derived ceramic veneer,
overlapping the late, most westerly
manifestations of Fort Walton. Yet
Fuller and Stowe have shown that even
Pensacola is not wholly protohistoric.
In 1941 Griffin explained the 1520-50
coin Moore recovered at Bear Point as
an artifact curated for some time
before burial. He thought that the
expansion of Middle Mississippian,
accompanied by the rapid spread of the
"Buzzard Cult" occurred in late proto
historic or even historic times (Grif
fin 1946:79-80). Willey's conflation
of Fort Walton and Pensacola was an
almost inevitable consequence of this
chronological perspective of the 1940s.
Mississippian cultural developments
were compressed into a period which
began late in the fifteenth century and
lasted for 200 years, at most
Willey recognized the multiple occupa
tions at Bear Point mound (1949:198-
200) and the intrusiveness of sixteenth
century burials into levels with
earlier pottery (which Moore (1901)
recorded but, as usual, did not col
lect). The entire complex of Pensacola
(and Moundville) ceramics, along with
the very few Fort Walton sherds, were
used to assign mound and midden to the
Fort Walton Period.
The association at Bear Point of
Southern Cult ceramics and Spanish
coins was unfortunate. Based upon the
archaeological maxim that any archaeo
logical deposit can only be as early as
its latest contents, the entire Fort
Walton period became sicklied over with
the pale cast of Spanish history at the
western end the Panhandle. If it were
post-1550 in Alabama (where it obvious
ly had originated from some downriver
Moundvillian project) could it be
present in Tallahassee before the
Spanish Mission period except that it
were some rapid, and violent intrusion?
Could it be other than an invasion?
This was exactly the type of
Mississippian-related horizon marker
that the Cult was supposed to have been
according to Griffin (1946), and which
Willey still considered it to represent
in 1958.
Of course we no longer accept the
timing of this purported blitzkrieg; we
have dates which Willey, Philips and
Griffin did not have. Nor can many of
us agree with the contemporaneity of
"cult" manifestations everywhere or
their supposed socio-ceremonial equiva
lence (Brose 1984c).
We can, I hope, all now agree that
early Fort Walton phases may be coeval
with early Moundville and Roods, and
the Pensacola phase of west Florida may

be coeval with the Alabama River and
Bull Creek phases (Brose and Percy
1978; Schnell, Knight, and Schnell
1979; Brose et al. 1983; Brose 1984a;
Knight and Mistovich 1984). That is,
Fort Walton is both early and mature
Mississippian, while in Florida,
Pensacola is late Mississippian. After
two generations of reliance on Willey's
1949 definition and descriptions for
Fort Walton (even as a ceramic period),
we still cannot agree upon "rules of
evidence for the kinds of data needed
to determine temporal limits, internal
sequential phases, or changing geo
graphical distributions. Indeed, we
have not yet been able to decide even
how ceramics (the archaeologist's best
friend) might distinguish between Late
Weeden Island and Early Fort Walton, or
Early Fort Walton and Roods, or Fort
Walton and Safety Harbor. It is about
time these problems were critically
I do not think we have understood what
Willey actually did on the Florida Gulf
Coast. In one masterful volume he
collated the extant archaeological
data; he carefully evaluated and
organized these data with the most
sophisticated methods available; and he
interpreted his integrated results in
terms of what was then the most theo
retically dynamic framework. Since
1949 we have masses of new data. We
have at our disposal numerous new
methods and techniques, techniques
undreamed of in Willey's day. Thus, it
is difficult to understand why archae
ologists in the 1980s continue to
interpret the evidence of prehistoric
societies in the northwest Florida area
within the archaeological frameworks of
the 1940s.
David S. Brose
Archaeology Department
Cleveland Museum of Natural History
Wade Oval, University Circle
Cleveland, Ohio 44106

Nancy Marie White
University of South Florida
This paper reviews current archaeologi
cal work and explanation in the region
encompassing northwest Florida, south
Alabama, and southwest Georgia.
The term "borderland" area is used for
obvious reasons. Science is always a
product of its time and place. Modern
political boundaries do more than cut
across coherent ecosystems such as
river basins; they also enclose differ
ent groups of archaeologists working
within their own traditional approach
es, assumptions, and opportunities for
conducting fieldwork, analysis and
An example is the heavy dependence on
various ceramic chronologies developed
from the individual vantage points of
each state and the earliest sites
excavated there. No really large-scale
attempts have been made since
Caldwell's (1958) to correlate typol
ogies of one state, region, or environ
ment with the next. Depending on the
training of the individual archaeo
logist, for example, the same pot or
sherd can be classified into any of
several types (Figure 1). The task is
far worse when one is confronted with
slightly less common types. Whether
one is a lumper or a splitter, there is
bound to be controversy in applying
terminology developed for more interior
portions of the state to the border
lands. The same can be said for
projectile points which from state to
state may have different names for
otherwise indistinguishable points
recovered from similar chronological
The variability in fieldwork opportuni
ties also structures our archaeological
explanation. In the 1970s and 1980s
most studies have been the result of
cultural resources management projects.
These tend to be limited to types of
environment where government instal
lations occur. They are not randomly
distributed, and seldom allow complete
investigation of any one ecological
The inherent nature of CRM surveys -
budgets, deadlines, and logistics
constraints, not to mention differing,
often inappropriate sampling tech
niques leads to survey biases that
are often not even discussed, let alone
corrected for, although the reporting
guidelines in Florida request a discus
sion of field methodology. The problem
may be even worse in non-CRM projects,
as they are not subject to ever improv
ing scopes of work which increasingly
demand that at least lip service be
paid to research design construction.
Client-oriented archaeology and compe
tition for contracts generally leads to
expanding extensiveness of survey
coverage but decreasing intensities.
A recent overview of Southwestern
archaeology describes a comparison of
results of a transect sample with those
of a complete survey of the same
project area. Site frequencies esti
mated from the transect sample results
departed from those obtained in the
complete survey by 167% (Cordell
1984:305). The difference was from
different levels of intensity of
coverage, measured in worker-days per
unit sampled (which of course, had
depended upon major differences in the
research goals of the two surveys).
In our region some of the same incon
sistencies may exist. For projects
aiming to find as many sites as possi
ble, surface survey as practiced 40
years ago may still be valid, along
eroding coastal areas or in plowed
fields, although such techniques are
deemed inadequate or incomplete for CRM
projects under most professional and
other guidelines. Site clusters on
dirt roads may reflect either lack of
Vol. 38 Numbers 1-2 Part 2 THE FLORIDA ANTHROPOLOGIST July, 1985

Figure 1. Vessel from the Little Pot site, 9Se98, on
the Chattahoochee River bank at the Florida-
Georgia-Alabama border. Archaeologists in
all three states could not agree on what
type to place it in (White 1981:548-549).
Figure 2. Fieldworker Susan Henefield of USF doing
archaeological survey on the bank of the
middle Apalachicola River in midsummer, 1984.
Long sleeves and snake leggings permit survey
and shovel testing of the thick forest.

time or an unwillingness to plunge into
the thick forest and dig (Figure 2).
Those who do shovel testing disagree
widely over the adequacy of 30 cm
square, 30 cm deep tests compared with
50 cm square, 130 cm deep tests, or 20
m test intervals compared with 150 m
intervals. It is now clear that wide,
deep tests are essential because of the
hign probability of buried sites in
northwest Florida.
Tne "judgemental survey," as the old-
fashioned "look-for-the-hot-spots"
technique is now called, is at least
recognized as something less than
rigorously "scientific," but it does
indeed produce the most sites per
worker-day. Statistically-derived
sampling designs, intended to yield
better controlled data, are commonly
abandoned when they prove too consump
tive of time or labor, or else they are
manipulated to provide more time in
"higher probability zones." This at
least leaves until final report writing
time any worries about quantitative
generalizations derived from such data.
Management concerns dictate the neces
sity for these generalizations. We are
forced to synthesize for summary and
explanation chapters in our reports
(even in our contract proposals).
Willeys great regional systhesis
(1949), as a product of the "cultural-
historical" phase of our discipline,
emphasized three things: definition of
"cultures" used to construct local
chronologies, reconstruction of life-
ways of these cultures, and derivation
of explanations of culture change from
external factors such as diffusion and
migration (Willey and Sabloff 1980:83-
180). We now stress new approaches and
explanatory frameworks such as systems
theory, cultural ecology, processual
explanation, social archaeology. Self
consciously, we now- even produce
introspective syntheses of our own work
(Dunnell 1984:489-90). While striving
to advance anthropological theory,
however, many of us in the Southeast
must continue to be preoccupied with
chronology (see Williams 1983:76-77).
This criticism may be unfair. Recent
data from well excavated, dated con
texts are hard to obtain amid the
hundreds of pages listing surface
collections from surveys. But I find
two particular flaws in recent synthet
ic attempts (e.g., Belovich et al.
1982; Bense 1983; Braley and Mitchelson
1984, Knight and Mistovich 1984,, Scarry
1980, Tesar 1980a), including my own
(White 1981b, 1982, 1984a, 1984b).
First, we are increasingly desperate
for greater organization, for clarity
of classification in the midst of the
daily information overload of life in
the 1980s. So some of us set up new,
more detailed classification schemes;
but many are based by necessity on
poorly controlled data. The plethora
of new phase names is bewildering.
Many are formulated from materials at
only one site. Others are based on too
few or poorly dated sites, or surface
collections. The problem is confounded
when the next researcher simply accepts
the newly proposed phase as a valid
entity to fit into or build upon,
instead of a hypothesis to be tested.
The second flaw is more fundamental, if
expectable (and I cannot yet suggest
any way to remedy it): We continue to
speak of archaeological "cultures" or
cultural periods (often the two are
used interchangeably, which is another
problem), as classificatory units
comprising our chronologies. What we
usually mean of course, are certain
artifact (usually ceramic) assemblages.
But we breathe life into them, see them
"migrating or "intruding," rising or
declining or being replaced by each
other. There are conceptual problems
with these constructs: they are
defined by trait lists, ignoring
adaptation and social interaction. It
is always harder to look at archaeo
logical manifestations from the
viewpoint of, say, social structure.
But besides general objections to the

concept of an archaeological culture,
there are specific problems in applying
such a construct, designed for analyz
ing sedentary agricultural societies,
to the analysis of open hunting-gather
ing societies or complex political
structures of early civilizations
(Trigger 1984:283).
For the kinds of fast summaries needed
in CRM reports, cultural chronology
sections of theses or research papers,
even study units for syllabi of survey
courses, archaeological cultures in the
traditional sense are presently essen
tial. They are less useful in answer
ing our latest research questions,
however. This becomes clear when we
try to understand culture process and
see what happens "when the sherds hit
the fan" (original quote from J.R.
Williams). Because the real problem is
that "the sherds don't ever hit the
fan" (original quote from D.S. Brose).
In other words, few real research
hypotheses of late have explored the
territory beyond the current frame
works .
If current project constraints preclude
development of middle range theory, not
to mention Big Science (in the sense of
theoretical contributions, not fund
ing), at least we can still collect
excellent data. These can serve us
well when all of the Gulf Coast is
under concrete and condos, and archaeo
logical research means work on dusty
collections. Meanwhile, instead on
continuing in too negative vein, I wish
to discuss worthwhile avenues of
pursuit in both analysis and explana
Method and Methodology
Presumably easier to collect than
direct social or religious data,
information on subsistence and settle
ment now increasingly gathered in
survey projects offers a way out of the
total dependence on pottery. Middens
can easily be sampled in shovel tests.
Analyses of lithic artifact assemblag
es, from raw material sources to tem
porally diagnostic forms to manufac
turing trajectories, is recognized as
being sorely needed in northwest
Ceramic sorting requirements are
inescapable, but few recent reports
agree on their type definitions. It
would be wonderful to hold a three-
states ceramic conference and all
compromise on a single clear, usable
typology system for everyone. Lacking
this, if types, or better, specific
sorting criteria are at least opera
tionally defined before being applied,
data classification can still be
amenable to comparison from one report
to the next. Major problems in chro
nology building and settlement modeling
can be traced to this issue. Sorting
criteria are seldom clear or, more
important, mutually exclusive.
Willey's (1949) original classifica
tions have served us well, but contain
many overlapping types when viewed at
the sherd level. For example, a Weeden
Island Incised rim sherd from the pot
illustrated in Willey's Figure 38a
could be identical to one from his Fort
Walton Incised pot in Figure 57c.
For another example, various overlap
ping descriptions of check-stamped
types have continued to make chis
surface treatment the subject of much
debate. I have measured several
thousand check sizes with sharp caliper
points and ever-dulling vision over the
last few years. The issue loses import
when one is unable to find any signifi
cant association between check size or
linearity, temper, or temporal place
ment (White 1981a:626-631, 652-59;
1982:139-57, 215-222, Appendix F). So
far most evidence indicates that Early
Woodland folk in northwest Florida
began making check-stamped pots about
2000 years ago and, except in a few
localizd areas or short time periods
that remain to be well determined, they
never quit making it. It is erroneous
to call this pottery "decorated."
Check stamping was a surface treatment,
undoubtedly utilitarian. So was cord
(38, 1985)

marking, which appears to be distribut
ed complementarily, increasing enor
mously where check-stamping becomes
less common, to the east and north of
the central Florida panhandle.
We claim the ability to sort varieties
of check-stamped types, especially
Deptford from Wakulla or Weeden Island.
One scheme flatly states that Deptford
contains mica while Wakulla does not
(Knight and Mistovich 1984:58); another
states that Wakulla is distinguished by
fine sand paste (Brose, in a review of
the aforementioned report, following
Willey's original description). These
observations apparently hold true only
for individual sites or localized
areas, however. At the Curlee site on
the upper Apalachicola, just below the
confluence of the Flint and
Chattahoochee Rivers, most of the
clearly late Weeden Island-early Fort
Walton Wakulla sherds were usually
packed full of both mica flakes and
large coarse grit.
Such identify problems complicate
interpretation in a region where one
ubiquitous site type is the one produc
ing only plain and check-stamped
sherds. At first I preferred to label
these finds Wakulla Check-Stamped, in
the tradition of earlier workers, whose
materials of this type nearly always
did date either chronometrically or
stratigraphically to the late Weeden
Island (e.g., Bullen 1950, 1958;
Caldwell 1978; Kelly 1950a). But we
are fooled by similar pots, down to the
folded rims, that appear on Deptford
sites (W. Lazarus 1965; my own unpub
lished work at the Trestle Bridge site,
8Jal86). It is now considered better
to use associated material for classi
fication: accompanied by Carrabelle
types, Keith Incised, etc., such sherds
are Wakulla; found with large fabric-
marked tetrapods or simple-stamped
sherds, they are Deptford (other
criteria are needed to pick out the
rest of the vast multitude of check-
stamped types, such as Mercier, Gulf,
Leon, etc.). By themselves, in my
opinion, check-stamped sherds can only
characterize a site as "indeterminate
Discourse at length on this issue at
least points out how we then go on to
have problems at the model-building
stage, as we design our survey or
interpret our overall results. Further
difficulties arise from a lack of
familiarity with the latest work in the
area. One recent survey, citing only
older reports, designed a model for
surveying coastal flatwoods based upon
work done in the only truly non-flat
land sector of the whole region, the
steep Torreya Ravines on the east side
of the Apalachicola, which further
differed by containing rare and unusual
floral and faunal species.
Many recent studies are only available,
however, in that "gray" literature of
archaeology: cultural resource manage
ment reports, dissertations, departmen
tal papers, all of limited distribu
tion. Worse still, many accepted type
definitions are generally agreed upon
by long-time workers in the area but
not ever really written down anywhere.
I found a case of this with Willey's
original extremely brief mention of the
type Marsh Island Incised (1949:466),
understood differently by earlier and
later researchers (White 1982:91). Use
of types generally considered non
standard these days is another abuse by
archaeologists newer in the region,
unfamiliar with what types were long
ago questioned or thrown out (e.g.,
Bullen 1968).
Taking all these concerns under advise
ment, we can concentrate on the specif
ic research questions that are most
significant and also most able to be
tested. Wtiat follows is a brief
summary of the kinds of avenues we
should pursue in our investigations,
both survey and excavation.
General Research Goals
We must move beyond ceramic typologies

to analyze style, technology, and raw
materials. Subsistence data are
essential. Rigorously controlled
survey and quantitative treatment of
resulting data may help us revise our
models, make them more testable, and
avoid the craze for unverifiable causal
explanations (Smith 1982:77). Perhaps
then we may ask some of the really
exciting questions that promote the
rise of anthropological theory.
We also need more valuable and (dare
one hope) practical kinds of knowledge
about human behavior in the past. An
example is the adaptation to coastal
versus interior environments through
time. (Davis 1984 begins discussion of
this issue). Location in the ecolog
ically richer interior is clearly
correlated with differences in material
culture, doubtless of significance in
understanding not only human subsis
tence and settlement, but also the
evolution of varying kinds of socio
political interaction systems as well.
Use and even abuse of subsistence
resources, explored by a few new
studies (see Claassen's paper, this
volume), is another extremely pertinent
subject for exploration today.
Some Specific Questions
We are stuck, for the moment, with our
archaeological cultures and temporal
stages. Figure 3 shows the current
framework. Particular research ques
tions pertinent to chronology, as well
as culture process, are suggested
below, including some testable hypothe
sis. Paleo-Indian and Archaic: The
extreme rarity may mean sites of this
age are few. All we have are surface-
collected Suwanee and Tallahassee
points from localized areas. But
perhaps other tool kits were important,
either in exploiting Pleistocene
species whose remains are abundant, or
in other adaptations. A.R. Kelly's
(1950b) "early tool industry" in
southwest Georgia included crude, heavy
choppers and large, plano-convex
scrapers of heavily weathered cert. He
associated these with the Paleo-Indian
time period, but also with Bolen
Beveled points of the Early Archaic.
Both crude chopping tools and beveled
points are ubiquitous in northwest
Florida, but their ages are virtually
Early sites may also be scarce because
they are now underwater due to post
glacial rise in sea levels, or under
several meters of alluvium. On the
lower Chattahoochee and upper
Apalachicola we often saw lines or
clusters of heavy chert artifacts and
debitage exposed in the riverbanks 3 to
5 m below the surface, clearly unreach
able in typical shovel tests (Figure
4). Obtaining some excavated, dated
materials may permit recognition of
more diagnostic lithic types, and help
us not to lump all sites without
potsherds into the Archaic.
Archaic points are often grossly lumped
into the type "Florida Archaic
Stemmed; further refinement of this
situation is crucial, as these are also
ubiquitous. Furthermore, they are
commonly found at clearly later sites.
Determining if they represent an early
component or simply recycled tools
picked up by prehistoric surface
collectors is difficult, unless there
are clear signs of "retooling" (Figure
5). An ambiguous type is the
Tallahassee, an isoceles triangle with
serrated edges, perhaps dating as early
as late Paleo-Indian, but also increas
ingly associated with late Weeden
Island; and possibly ancestral to the
triangular Pinellas point of the later
Fort Walton period. Chert raw
materials are little known, but some
trends should be examined. The whitish
"Gulf Coast chert" is the most common,
but there is also a lot of agatized
coral. Quarry sites are known for
both. There is also limited use of
Talahatta quartzite all around south
Alabama and western northwest Florida.
As for Archaic settlement patterns,
should we accept the view that sites
concentrate in uplands in the interior?
There must be hundreds deeply buried
(38, 1985)

Figure 3.
Civil War
Aboriginal &
Late Archaic
Lower Creek/
based camp-
trade goods
Majolica, Lamar
hunting and
Fort Walton
temple mounds
Late Weeden
Island, Wakulla
Late Woodland
Swift Creek-
Weeden Island
incised, pune-
tated ceramics,
copper, galena,
hunting &
Early Swift
Creek compli-
ceramics, some
Deptford Simple-
Stamped, Fabric-
Marked, Check-
small, semi-
ground stone
many stemmed
and notched
points, other
Bolen Beveled
points, heavily
weathered (?)
Tallahassee (?)

Figure 4. Example of a deeply buried site, 8Ja412, on the middle Apalachicola.
Exposed in the bank 2-3 m below the pale alluvium (dark bank across
middle of photo), these dark soils contained only lithic remains,
though nothing yet recovered is clearly diagnostic of the Archaic.
Figure 5. Archaic stemmed point of silicified from site 9Sell7 (White 1981a:580).
Point reworked by later Woodland peoples to reveal the lighter colored
stone underlying the darker patination, portions of which remain on
the lower surface of the blade and most of the stem.

Figure 6. Fabric-marked tetrapodal sherd from 9Jal86, a Deptford site
on the upper Apalachicola, over 100 river miles inland.
Figure 7- Stemmed point of clear quartz, a highly unusual raw material,
from the Middle Woodland site 8Ja204, a few kilometers west
of the Chattahoochee River in northwest Florida.

along the major river bottomlands. It
has been suggested by Tesar (1980a) and
others that such sites are concentrated
along the edge of river drainages,
around wet prairies, and around sink
holes and lakes.
Late Archaic: The debate continues
over nomenclature for fiber-tempered
ceramics, variously termed St. Simons
Plain, Orange, and Norwood (Willey
1949, Bullen 1958, Phelps 1965). The
latest synthesis for Florida refers to
either Norwood or Orange (Milanich and
Fairbanks 1980), while in Alabama
Orange is seen to be replaced by
Norwood within a new temporal period
called the "Gulf Formational Stage"
(Walthall 1980), and in southwest
Georgia these earliest ceramics are
considered to be in the Stallings
Island series (DePratter 1975). Type
names are actually a minor problem;
there is not yet enough evidence of a
distinctive cultural entity such as
Northwood was intended to represent.
More well-dated materials are needed.
Deep excavation is required to find
them, as Bullen found on the
Chattahoochee (1958), where stratified
deposits containing fiber-tempered
sherds dated to 1200 B.C. Later work
(e.g., Milanich 1974, Scarry 1975,
Henefield 1985), suggests that if one
keeps going deeper below Woodland
sites, the labor will pay off when the
typical thick, crude, plain fiber-
tempered sherds appear.
Deptford: Ceramic chronology here
includes the question of overlap at
both ends of the Early Woodland tempo
ral span (Also discussed in this issue
by Thomas and Campbell, and by Brose).
Bullen's work recovered sand-tempered
pottery in association with the fiber-
tempered sherds. Other investigators
see the unmistakeable types (fabric-
marked, simple-stamped) coming up
together with complicated-stamped
sherds originally thought to be cultur
ally different. Sites yielding unmixed
materials have been located (e.g.,
White 1981:636-640, Belovich et al.
1982:411-417) (the problem with check-
stamped has alreay been discussed).
Types once unexpected have been recov
ered. I found a large fabric-marked
tetrapod at 8Jal86, 100 river miles
inland on the Apalachicola (Figure 6).
Settlement is another question; tradi
tional descriptions have emphasized
coastal patterns (Milanich and
Fairbanks 1980), with only occasional
utilization of inland locales for wild
resources collection. Good subsistence
materials could easily demonstrate
seasonality or a temporary nature of
sites. Tesar (1980a) has suggested
both permanent coastal and inland
settlement, with associated special use
seasonal activities.
Early Woodland burial mounds are almost
unknown, but all of us probably suspect
our favorite mound sites of later
cultural affiliation to be covering
Swift Creek: For later Swift Creek
times there is accumulating excellent
evidence for resource utilization,
ceremony, and society. However, an
"early" or "pure" horizon unmixed with
later Weeden Island materials, or with
earlier Deptford ceramics, remains less
well-defined, or may not even exist
except due to sampling error. Further
more, there may be serious difficulties
with assuming that pure and early Swift
Creek are the same phenomenon.
Specific complicated-stamped designs
may be generally considered early or
late, but there is yet no comprehensive
assessment of their distribution over
space and time.
Late Swift Creek-Early Weeden Island:
The archaeologically recognized flores
cence of Middle Woodland and Gulf Coast
cultures include a great number of
unusual items, such as exotic cherts
(Figure 7) and other artifacts that
should be more closely examined in this
region. The Baker's Creek and Savannah
River point types may prove to be more
narrowly diagnostic of the time period
than is currently thought. We need to
(38, 1985)

clarify ceramic type relative frequen
cies permitting attribution of an
assemblage to one part of the Woodland
or the next. (One possible method is
Doran's PIXE analysis, this issue).
It may be during this time period that
the dichotomy between coastal and
inland adaptations intensifies. Large,
often ring-shaped coastal shell midens
suggest greater population densities
relative to the interior. But are
interior sites really not often on
riverbanks (Milanich et al. 1984:26)?
Weeden Island was originally estab
lished by Willey as two time periods, I
and II, even then with the realization
that the earlier one overlapped late
Swift Creek. The now classic refine
ment of this model (Percy and Brose
1974, Brose and Percy 1974), dividing
the period into 5 segments based on
different ceramic type frequencies in
middens in the Apalachicola River
drainage, remains to be really rigor
ously tested (see critique in White
1981a:647-649). We need to understand
the conjoining of two very different
ceramic traditions, complicated stamp
ing and incision/punctation. Possibly
more significant, botanical data can
help evaluate the timing and rates of
the shift from gathering to horticul
ture. Social correlates of ranked
society may be present in burial mound
data; might there be incipient social
stratification, unlike the situation
in northeast Florida, where there is no
Swift Creek (Milanich et al. 1984).
Late Weeden Island: This may be the
postponed "decline" from Middle Wood
land splendor seen earlier elsewhere in
the eastern U.S. It is uncertain why
ceramic types considered more "sacred"
disappeared in late Weeden Island.
Midden sites of this time period far
outnumber those of any other. Subsis
tence materials demonstrate continued
utilization of wild resources, but also
suggest the beginning of maize horti
culture (Milanich 1974) in the inter
ior .
There is enormous environmental varia
bility and extensiveness of distribu
tion of late Weeden Island sites.
Great socioeconomic instability and
enormous population growth are postu
lated at this time, possibly the result
of increasing need for arable land.
Evidence for these processes is cru
cial, but still lacking. Settlement
data should be compared and the pres
ence of corn among macrobotanical
specimens from middens should be
sought. What is the explanation for
the tremendous increase in use of
freshwater molluscs apparently taking
place at this time (White 1982:199-
122)? Is there a state of the art
zooarchaeological study of the contents
of even a small shell pile on a river
ine site? There are so many different
forms of freshwater shellfish deposi
tion, from tiny piles to entire strata,
but what it all means in terms of human
behavior (except the obvious, full-time
clambakers or quick picnics), has not
been addressed.
Finally, if development of the succeed
ing more complex sociopolitical systems
in situ out of this adaptation did take
place, as most recent interpretations
agree, we need to establish the materi
al correlates expected to exist and
then go look for them.
Fort Walton: The evolution of these
systems of temple mound building,
intensive maize agriculture, and
relatively complex, probably stratified
society is generally seen as having
taken place earliest in the lower
Chattahoochee-Apalachicola Valley, but
no one can say exactly why. There is
no agreement on phase divisions in Fort
Walton (White 1982:231-241). Work on
possible spheres of political domina
tion by individual mound centers
(Scarry and Payne 1982) is exciting,
but suffers from lack of confirming
hard evidence. There is a clear and
radical settlement shift in the interi
or, to riverine locales. This has been
vaguely attributed to growing agricul
tural dependence. Riverine mollusc
procurement drops markedly for as yet
unknown reasons; people are certainly
still collecting bountiful wild re-

sources, such as large amounts of deer
and acorn.
A fascinating question concerns the
great lack of chert materials at Fort
Walton sites, especially as compared
with earlier and later sites. Perhaps
this is related to subsistence practic
Coastal adaptations remain enormously
different from those in the interior,
with a great emphasis on shellfish
collecting, not maize agriculture,
presumably for ecological reasons. The
roles played by small short-term camps
as opposed to hypothesized permanent
villages must be investigated through
both settlement archaeology and ethno
graphic records research. If settle
ment hierarchies can actually be
demonstrated, perhaps the material
evidence for social stratification can
be as well. Might interior and
coastal sites be seasonal manifesta
tions of the same culture? Might
ethnicity account for Pensacola ceramic
series differentiation?
Later on in Fort Walton some specific
phenomena are in desperate need of
exploration. The first is the role of
the Southeastern Ceremonial Complex or
Southern Cult (whatever that entity may
actually be in itself). No centers of
cult objects have been uncovered so far
in the area between Lake Jackson
(Tallahassee) and southwest Alabama,
but the motifs are seen on ceramics
(Lazarus and Hawkins 1976; White 1982:
Plate 18y). Is it merely participation
in religious ceremonialism or some
thing more? The nature of Lamar, a
ceramic complex perhaps associated with
proto-Creek, also remains unclear in
northwest Florida, where it rather
suddenly appears in late prehistoric
times. If it is only a new sort of
pottery traded down the larger rivers,
it may not be connected with any actual
movement of peoples at all, but why
does it appear, and exactly when? Is
there a relationship here with unusual
developments in the Mississippi Valley
heartland area in protohistoric times
(Williams 1983:77-78)?
Historic Native American: Clearly, in
northwest Florida the Spanish first met
with groups possessing a Fort Walton
material culture, since we find their
trade goods at some sites. The ethnic
identity of these peoples is unknown,
however. West of the Apalachee, whose
extent apparently did not reach west
ward beyond Tallahassee, there are no
good records. Other names known in the
Panhandle, such as the Sawokli or
Sabacola, the Chatot, the Apalachicola,
and the Pensacola are apparently later.
Ceramic associations are difficult. On
the coast but also at some inland
locations heavily shell-tempered pots
accompany cemetery burials. But also
inland is found a proliferation of
Lamar, Leon-Jefferson and Ocmulgee
Fields ceramic types, suggesting a
newer Lower Creek identity.
Why did later historic natives, unques
tionably Lower Creeks (somehow rapidly
changing into Seminles) quickly switch
to making rather dull brushed pottery?
Just the kinds of responses to the
invading Euro-Americans reflected in
the selective use of non-aboriginal
artifacts is a fascinating area still
waiting for proper study. One could
conceivably continue aboriginal materi
al culture studies to the present day
among the isolated enclaves of people
who were able to avoid removal to
Indian territory.
At the risk of sounding overly pedantic
petty, and horribly critical, I have
presented what amounts to my own
personal summary of some directions we
should take in the future, and some we
should avoid. It is my aim to jump
right into the investigation of many of
these avenues of research with our next
field season, if we can only keep the
survey boat from breaking down!
Nancy Marie White
Department of Anthropology
University of South Florida
Tampa, Florida 36620
(38, 1985)

Jerald T. Milanich
Florida State Museum
I think we would all agree that the
major goal of archaeology is to learn
about the pasthow people lived, their
relationships to other peoples and to
their environments. We also wish to
understand the processes which affected
these peoples through time. Not all
research must be formulated to produce
new culturological theorems; there is
nothing wrong with doing good old
cultural history (what is now often
referred to as past lifeway studies).
In the eastern United States there are
three major research questions regard
ing past lifeways which lie at the
basis of almost all the archaeology
that we do; most culture history
research questions which we ask are
corollaries of these major questions.
First, of course, is the question of
the peopling of the New World. Who
came, when, and why? What was the
effect of those early populations on
the plant and animal populations of the
New World? What was the effect of the
New World environment (which previously
did not contain a single human) on the
early aboriginal cultures and their
subsequent development? These are
intriguing questions, questions which
are of great interest to us today.
Unfortunately, northwest Florida and
its borderlands have not generally
served as a laboratory for studies of
Paleoindian and early Archaic peoples.
This may be the fault of the archaeolo
gists more than of northwest Florida;
such early sites are indeed present,
although they may not be readily
apparent (e.g., White 1981a:606-618;
Webb et al. 1984).
The second question revolves around the
origins of plant domestication (both
native and introduced species) and the
concommitant development of social
complexity. A great deal of recent
research has centered on the appearance
in che archaeological record of cucur
bits during Archaic times and even more
is being focused on the other end of
the time scale, the emergence of
Mississippian societies and economic
systems out of more egalitarian Wood
land systems. Northwest Florida and
its borderlands have in the past
provided a great deal of information on
Woodland societies (e.g., Weeden
Island) and the evolution of
Mississippian period groups (e.g., Fort
Walton). Many archaeologists in the
Southeast are jealous of the corpus of
sites (and the published data) dating
from ca. 500 to 1539 found in northwest
Florida. This region has the potential
to provide excellent understanding of
the development of complex societies,
and it is doing so (e.g., Brose 1984b;
Scarry 1984). There is no reason to be
apologetic for northwest Florida or the
research carried out there in the past.
Best of all, symposia like this one are
focusing attention on the types of
problems which can readily be ap
proached in the region. Building on
the early work of C.B. Moore and Gordon
R. Willey and the recent work of White
(1982) and Scarry (1984) and others,
the future looks even brighter.
The third and final general research
problem is determining the nature of
aboriginal populations at the time of
contact with Europeans and the effects
of that contact on native societies and
their aboriginal environment. What was
the magnitude of disease-caused popula
tion reduction? What was the magni
tude of culture change which resulted
from demographic collapse? Here again,
northwest Florida can provide answers.
The Apalachee Indians have been the
focus of much documentary research.
Interest in Spanish mission sites and
in the Fort Walton culture has also
generated investigations of Apalacnee
archaeology. Work presently being
undertaken by the Florida Department of
State, Division of Archives, History
and Records Management (DAHRM) and by
Rochelle Marrinan of Florida State
University, building on the work
carried out in the 1940s through 1970s,
Vol. 38 Numbers 1-2 Part 2 THE FLORIDA ANTHROPOLOGIST July, 1985

has tiie potential to provide unprece
dented understanding of the demise of a
Mississippian lifestyle and subsequent
cultural changes which occurred as a
result of historic contact.
The amount of available data on north
west Florida's aboriginal cultures from
500 B.C. to 1700 is considerable. It
may be one of the most studied areas in
the eastern United States. The work of
Moore, Willey and Woodbury, Willey,
William Lazarus, and others, along with
a host of Florida State University
archaeologists and students has left a
real legacy. The banner has been taken
up by the DAHRM under whose auspices
many projects have been carried out.
From the papers included in this volume
we can see that the U.S. Forest Serv
ice, CRM firms, and other universities
and museums in Florida, Alabama, North
Carolina and Ohio can be added to the
As a result of all of this research our
collective knowledge is indeed great.
What is sorely needed is for someone to
devote several years to reviewing all
of the publications and contract
reports and in-house circulars and
writing a book length synthesis (I can
almost promise that the Bullen series
would publish it). Out of synthesis
comes problem orientation. This giant
step must be undertaken, perhaps as a
dissertation by a devoted graduate
student. I feel that without such a
synthesis we will remain securely
bogged down in arguments and counter
arguments regarding taxonomic trivia
such as what is Deptford and what is
We can no longer define cultures almost
solely on the basis of ceramic assem
blages; instead we must adopt a taxo
nomic system that takes into account
other criteria as well, such as envi
ronmental adjustments in the form of
settlement patterns, subsistence
activities, and artifact assemblages
(e.g., Milanich 1973:51; Smith
1978:480-488). As many of the papers
point out, it seems ludicrous to speak
of a Deptford culture that ranges from
the Gulf coast through Mandeville to
northern Georgia and down to the
Deptford type-sice near Savannah at
various times during the period 500
B.C. to A.D. 700.
The papers in this collection are a
mixed bag and they reflect the range of
current research orientations. The
papers by Forney, Bense, and Thomas
and Campbell are an outgrowth of
cultural resource management projects.
Bense's deals largely with attempts by
archaeologists and the citizens of
Pensacola to preserve and conserve
their archaeological resources. As
municipalities learn that archaeology
is good for education and tourism and
that it excites the interest of the
public (not to mention archaeologists)
city fathers and mothers are going to
be more willing to undertake cultural
resource management. All they need is
a resident archaeologist to push things
through; perhaps we could clone Judy
and send her to different cities. The
other two CRM papers both provide data
that is grist for the mill, information
that supports other data already
published and helps to establish
patterns. We need such studies,
especially if they are oriented cowards
a specific goal which will increase our
understanding of major research ques
tions (or corollaries of those ques
tions) .
The contributions by Jenkins and Mann,
Fuller, and Stowe focus on westernmost
northwest Florida, including what we
Floridians call the borderland to the
northwest (I'll bet those Alabamans see
it differently; also, I have always
wondered if aborigines in southeastern
Alabama worried about being taxonom-
ically excised from conferences, or if
they preferred being called McLeod or
Wakulla or perhaps something else).
Ned has begun to investigate a little
known area, the Conecuh, where Bill
Sears found the Mitchell site, a Weeden
Island site which had previously been
added to that list of unexcavated sites
which have figured strongly in our
(38, 1985)

interpretations. It will be interest
ing to see if the region was indeed
inhabited by early Weeden Island people
and if the Mitchell site was a lineage
center like the McKeithen site well to
the east in North Florida (Milanich et
al. 1984).
Richard Fuller and Reed Stowe have
begun to unravel further the mystery of
the Pensacola culture, an undertaking
which is going to increase our knowl
edge of the impact of the Spanish
explorations in the sixteenth century
on northwest Florida, not to mention
our understanding of Mississippian
period coastal populations. New
taxonomy was needed and they have not
hesitated to organize their data
Glen Doran's and Cheryl Claassen's
respective PIXE ceramic analysis (are
we going to have Twinklebell next?) and
shellfish analysis are examples of the
specialized types of laboratory work
which are now available and which will
add a great deal to our knowledge of
prehistoric lifeways. I am glad we
don't have to eat Rangia today, but can
stick with those Apalachicola oysters
(I suspect that Apalachicola has become
a generic term for the northwest
Florida oyster, much like Deptford has
evolved into a term for much Southeast
check-stamped pottery).
Personally I have no problem with using
the term Deptford to refer to a ceramic
assemblage associated with a specific
coastal plain culture and using
Deptford period to refer to the time
period of that culture. Deptford does
have meaning and has been defined.
When it is a problem is when people
apply the same term to assemblages,
cultures, and periods well inland at
Mandeville or even worse, in the
Georgia piedmont. However, I do agree
with Brose's contention that we must
not remain impaled on our own heuristic
devices, our taxonomy, but we should be
willing to revise our definitive
nomenclature as needed. I do not agree
that it is not being done (e.g., see
the papers here by Fuller and Stowe;
also see Schnell, Knight, and Schnell
1981, for an example from the northern
My last comments are for Nancy White,
who almost overnight has gone from
being a bright-eyed new Ph.D. to a
veteran. Organizing a symposium began
the aging process; getting these edited
papers out of the contributors finished
it. She has now learned about life in
the real world and is entitled to
criticize all of those who have gone
before her. However, as she will
shortly find out, it is all too soon
that one's own students repeat the
process. She deserves a great deal of
credit for undertaking this task.
New generations of scholars bring new
interpretations, new techniques, new
conclusions. Our knowledge continues
to evolve, changing more rapidly all
the time. When you are just climbing
down from the trees to join the other
scavengers on the savannah, it may be
difficult to see how far you have come
and how much your evolution has depend
ed on all your ancestors who slept in
those trees. Northwest Florida archae
ology has also come a long way from the
work of Holmes and C.B. Moore. Without
them, Gordon Willey may never have
published his synthesis; without
Willey, John Scarry and Nancy White
would not have written their disser
tations, and without them someone down
the line will not be able to write an
overview of northwest Florida. From my
perspective over here in the
borderlands, things don't look bad at
Jerald T. Milanich
Department of Anthropology
Florida State Museum
Museum Road
University of Florida
Gainesville, Florida 32611

Gordon R. Willey

Nancy Marie White's gracious invitation
to me to offer a few comments on the
papers of the recent symposium of
Northwest Florida archaeology was
impossible to resist, invoking, as it
did, nostalgia for my own past as well
as stirring my curiosity about what has
been happening in recent decades in the
pursuit of prehistory in this region.
Her suggestion that I might “offer
words of wisdom" is flattering but wide
of the mark; I have been away from this
particular field for too long for me to
do this. Rather, what little I have to
say will consist of some reminiscences
that may be of didactic interest and
some questions about substantive
matters in Florida archaeology. The
answers to some of the latter may be
well known to Florida specialists, and
my asking them will betray my ignorance
of current literature; on the other
hand, I am sure that some of my ques-—
tions relate to continuing problems for
those of you working in the Florida
Gulf Coast and general Southeastern

First as to the reminiscences, and
relating to some of Brose's introducto-
ry remarks, it is true that I began my
Florida research in 1939-40, but
Archeology of the Florida Gulf Coast
(Willey 1949) was written largely in
the summer and fall of 1948, when I was
no longer a graduate student but a 35
year old professional with an estab-
lished position in the Bureau of
American Ethnology and several seasons
of archaeological field experience
behind me in the Southeast, Peru, and
Panama. The point is of significance
in suggesting that I should have
produced a more sophisticated piece of
work than I did, especially as at that
time I considered myself a bit of an
expert in “archaeology theory." While
I do not intend to make this an apolo-
gia - anymore than a paean of self-
congratulation — let me tell you

Vol. 38 Numbers 1-2 Part 2


about the writing of Archeology of the
Florida Gulf Coast.

When I began my Florida Gulf research,
I was in a mood to move rapidly on
problems of chronology and culture
history and evolution. [I felt that the
Midwestern Taxonomic System, then
widely in use in the northern part of
the Eastern North American Woodlands,
and even to some lesser extent in the
south, moved too slowly toward such
problems; its avowed “timelessness”
disturbed me, and the connotations of
its terminology and its hierarchical
structuring seemed “northern” in their
application and not well-adapted to
what we were finding in Georgia,
Louisiana, and Florida. So in Florida
I opted for what seemed to me to be a
system of extreme simplicity and one
geared to chronology - especially to
the chronology of pottery styles and
types- Iwas also looking for a system
that would do the least violence to the
available pre-existing literature on
the region as well as one that - or at
least I felt so then - would create the
fewest problems for those who followed
me. To a degree, I think I was suc-
cessful in this. One of the strengths
of Archeology of the Florida Gulf Coast
is its extraordinary simplicity. For
the most part, it has turned out to be
a surprisingly accurate chronicle of
ceramic traditions and even of specific
pottery types. Its classificatory
structure, the “period-cultures” of the
chronology, is spare and has not
encumbered you who have followed me
with an excess of deadwood to clear
awaye- At the same time, the unfortu-
nate linkages of such terms and con-
cepts as “period” and “culture”, and
the interchangeable usages of these
terms and concepts, were archaeological
mistakes and have saddled you with
something that complicates present
archaeological research in the area.
Clearly, linkages of “period” and

July, 1985


"culture," "culture" and "pottery
type," or "pottery type" and "period,
if made on a one-to-one basis are
inhibiting in the pursuit of modern
archaeological problems and prejudicial
of results.
As an example, to which several in the
symposium have alluded, it was a
mistake for me to use the name
Deptford, after a site and pottery
style on the Georgia Coast, for a
culture and major period in my Gulf
Coast Florida scheme. I should have
called the Northwest Florida complex,
or culture, with the Deptford or
Deptford-like pottery, something else:
Carrabelle, perhaps, after my best
local type site. The time period
should have been designated quite
separately as, say, Period III, leaving
room above and below on the time-scale.
Irving Rouse (1939) had already done
something like this in Haiti, and, as
an admirer of his work, and familiar
with it, I should have installed such
machinery or such a framework in
Florida. Or so logic seems to dictate
to me now. But who knows for sure?
Maybe in so doing I would have created
what some would believe to be a more
cumbersome structure than what I did
In any event, it is clear that the
chronological-classificatory structure
of Archeology of the Florida Gulf Coast
is adequate only in a preliminary and
general way for the definition of the
kinds of problems you are interested in
now: culture-environmental adapta
tions, the content and organization of
functioning communities, the trajecto
ries of cultural and social change, and
the various meanings in cultural
variation. Certainly one of the things
that has emerged in the archaeology of
recent decades is that, as we proceed
with the defining and asking of ques
tions of this nature basically
processual questions we are
frequently brought up against
inadequacies in the space-time and
culture unit frameworks as
these have been laid out in the exist
ing literature. But, as I read these
papers of the Pensacola symposium, as
well as other writings in Florida
archaeology that have appeared in
recent years, it seems to me that you
are setting about with modifications
and corrections and the building of new
syntheses through trial-and-error
procedures. The Milanich-Fairbanks
(1980) book is one such synthesis, and
I would presume there are others in the
making. As you can see, at age 72, I
remain optimistic perhaps sometimes
bemused and confused but on the whole
optimistic about the future of archae
ology in Florida as well as elsewhere.
So much for these reminiscences that
are perhaps more interesting to the
writer than to the reader. Let us turn
to the questions. How does the archae
ology of Florida now stand on various
important substantive matters? Appar
ently, there is little more that has
been found in the northwest part of the
state that pertains to the Paleo-Indian
era, beyond what was known in the
1940s, although its presence, in some
degree or other, is presumed. As I
read what you have to say, the Archaic
still seems skimpy in the Northwest.
Did the Gulf Coast bays and estuaries
fail to provide the same kind of volume
of resources provided along the St.
Johns River where the enormously deep
shell middens showed preceramic Archaic
levels as well as those with fiber-
tempered pottery? I found scatterings
of fiber-tempered sherds at a few
sites, with hints of their early
chronological position; Lazarus (1965)
describes similar findings at the
Alligator Lake site in Walton County
where fiber-tempered wares probably
antedate the 625 B.C. radiocarbon date
for early Deptford pottery. Where are
the closest cultural affiliations of
this Northwest Florida fiber-tempered
pottery? Eastward, to the St. Johns
and the Georgia Coast? Or
northwestward to Alabama, as hinted at
by Jenkins and Mann's Conecuh finds? I
am inclined to think that some big

shell middens although not as big as
those of the St. Johns may have been
destroyed for commercial or road
building purposes in Northwest Florida.
Woodbury and I saw signs of this in
1940. The West Goose Creek site in
Wakulla is an example (Willey,
1949:292-293) where the lower levels of
a deep midden were probably of Archaic
age. I also wonder about the big,
lunate-shaped middens at East Point, in
Franklin County, opposite the city of
Apalachicola. I speculated that they
had been Archaic sites (Willey
1949:284). Have they ever been
explored by an archaeologist? Or have
they been bull-dozed away in the march
of commercial progress? If they
haven't they should be given some
Another question of general cultural
historical interest is when did burial
mound ceremonialism first appear in
Northwest Florida? While it was my
conclusion that this probably did not
happen until the Santa Rosa-Swift Creek
Period, there were some hints that it
might have had an inception on the
earlier Deptford time level. Moore
(1918) described some cremated burials
at Carrabelle. Were these from an
artificial sand mound? His account is
somewhat equivocal about this. The
ceramic associations were Deptford,
including a little tetrapodal jar of
the Deptford Simple-Stamped type and an
Alexander Incised sherd (Willey
1949:267-268). Burial mound ceremon
ialism is known from Early Woodland
contexts in other places in the Eastern
United States. Adena and Adena-related
cultures so date, as does Poverty Point
and Tchefuncte in Louisiana. More
specific dating also relates to this
question. An early Deptford date, well
back into the first millennium B.C.,
might be too early for the artificial
mound trait in northwest Florida; a
date of 50 B.C. to A. D. 140, as given
for the Okaloosa Deptford phase, or
Deptford to Swift Creek transition,
might not be too early.
Several of the papers have indicated
the definite chronological overlap
between Deptford series pottery types
and those of the complicated-stamped
Swift Creek genre, and this, of course,
is consistent with my earlier northwest
Florida stratigraphic results. It
seems likely, from all that we know of
Swift Creek Complicated-Stamped, and
related complicated-stamped pottery
distributions, that the complicated-
stamped tradition did develop in the
general south Georgia-north Florida
territory and from the observations of
Thomas and Campbell, Brose, and others
this development took place in the
early centuries of the Christian era.
As Brose emphasizes, the Santa Rosa-
Swift Creek Period, as I defined it, is
an obvious ceramic mixture of two
traditions. I think that my hyphenated
conflation of the names so expresses
this. The Santa Rosa elements are
western; the Swift Creek are eastern.
Brose relates the Santa Rosa series
pottery types to the Bayou La Batre-
Porter continuum of ceramic development
of the Lower Tombigbee and Mobile Bay
in Alabama and suggests a date of 200
B.C. A. D. 150. If I understand him
correctly, he argues that this is
earlier than the inception of the
complicated-stamping tradition. I
should think we need more evidence on
this point. One would expect Santa
Rosa ceramics to be earlier than
complicated stamped types in the west,
but I wonder if Swift Creek doesn't
have equally early origins farther
Again, the question of burial mound
ceremonialism may be brought up. Is it
a trait linked closely with Santa Rosa
series pottery, and both relating to
Marksville still farther to the west?
This seems a reasonable possibility.
On the other hand, James Griffin
(personal communication 1984) offers
the opinion that there are Hopewellian
traits spreading down the western side
of the Appalachians, from the Ohio
Valley to Florida, that stand apart
from Marksville and Santa Rosa-Swift
Creek and probably have a different
history? Some of the Crystal River
(38, 1985)

elements for instance, the copper
tubes would appear to belong to this
diffusion. What of Crystal River? As
is quite evident to anyone who has read
Archeology of the Florida Gulf Coast,
as well as some earlier papers of mine
(Willey 1948a, 1948b; Willey and
Phillips 1944), I had a great deal of
difficulty in deciding just what to do
with this famous site. I think that
the Santa Rosa-Swift Creek time period
is where it, or most of it, belongs
chronologically; but from a typological
point of view it does not rest easily
there. Brose, judging by what he says
in his symposium paper, has less
trouble with it than I did, and I note
with interest his statement that
Crystal River and Kolomoki "now show
radiocarbon equivalence." I always
felt, in contradiction to Sears' (1954)
interpretations, that the Kolomoki
burial mound was somewhere transitional
between Santa Rosa-Swift Creek and
Weeden Island I, and this certainly
seems to be the chronologically conven
ient slot for Crystal River. Before
moving away from Kolomoki and Crystal
River, what is the consensus now about
the dating of the big flat-topped
pyramid mounds at these two sites? Are
they later additions on a Mississippian
time level? Or do they belong back in
the early centuries A.D., on the same
time level with the burial mounds? I
have often felt that worrying about
this led Sears (1956) to insist upon a
"mature Mississippian dating for the
Kolomoki burial mound. For the model
carried in the minds of most of the
archaeologists working in the Eastern
United States in the early 1950's was
that temple mounds and truly elaborate
pottery were hallmarks of the
Mississippian tradition and horizon.
The Santa Rosa-Swift Creek and, espe
cially, the early Weeden Island wares
of the far south were just too fancy
looking to fit the Woodland taxonomic
model. I don't think there can be
doubt any longer that Weeden Island
pottery is quite definitely distinct
from and earlier than anything that is
usually thought of as Mississippian;
and I think the same thing can be said
for platform mound ceremonialism in
south Georgia and Florida although,
admittedly, I plead ignorance about
recent research on this question.
Almost 20 years ago, however, the late
Ripley P. Bullen (1966), in discussing
one of the temple mounds at the Crystal
River site, provided some radiocarbon
dates, from a putative "stela found in
front of the mound as well as from
mound fill refuse, of ca. A.D. 500 or a
bit earlier. This can scarcely be
considered firm dating, but has
anything more been done to explore and
date the Crystal River temple mounds?
Or che big temple mound at Kolomoki?
Weeden Island as a period in Northwest
Florida seems to have lasted a long
time, from A.D. 400 until about 1000 -
longer and earlier than my 1949 guess-
date of A.D. 1000-1500. (In those pre
radiocarbon days why were we always so
short on chronological estimates? We
thought of it then as commendable
conservatism.) Weeden Island is
certainly the "Classic" culture of
Northwest Florida or, indeed, Gulf
Coast Florida, with its numerous burial
mounds and midden sites and its grace
fully incised, punctated, and modelled
pottery. A word may be in order about
the spelling of the name. Milanich
raised the point in his symposium oral
presentation (although not in his later
written commencs). The first litera
ture referecne to the Weeden Island
site, in the Tampa Bay region, is that
of J.W. Fewkes (1924) in his little
excavation monograph on the site. He
refers to an early settler and landown
er in these parts, a "Dr. Weeden."
Stirling (1936) followed Fewkes in
spelling the name this way. In our
preliminary report on the 1940 North
west Coast survey, I was responsible
for spelling it two ways, "Weeden" in
the chart on page 235 and "Weedon in
the chart on page 237 (Willey and
Woodbury, 1942). I cannot remember how
this came about. Anyway, a short time
later, in an article devoted speci
fically to the Weeden Island culture

(Willey 1945), I settled on the "en"
spelling. Following Fewkes, it seemed
to be the appropriate one, and I stuck
with it in the 1949 monograph. Perhaps
more important historically, than the
spelling of the name, was its selection
as the culture and period name. In
this, I followed Stirling (1936) who
had used Weeden Island as a general
name for a pottery style of the Florida
Gulf Coast, one which preceded the
later Safety Harbor style. I might add
that, in beginning work in the
northwest, I thought of all Weeden
Island pottery as fine, sand-tempered
ware, and it was not until I began to
examine some of the collections from
farther south, including that from the
Weeden Island site, that I realized
that quite characteristic Weeden Island
incised and punctated decoration was
often found on vessels of a soft paste,
or "temperless, quality. These were
then set apart as the Papys Bayou
I am much pleased by the chronological
subdivisions which present workers have
come up with for Weeden Island. It was
obvious, even to me in the 1940's that
such was necessary, and I am gratified
that my beginning with Weeden Island I
(complicated-stamped associated) and
Weeden Island II (Wakulla Check-Stamped
associated) has proved useful in this
regard. I also infer that you are
making headway with regional as well as
chronological subdivisions of this
culture and period. Certainly, Jenkins
and Mann's Weeden Island sites, from
the far west end of the geographical
range, show a regional difference in
their lack or scarcity of complicated-
stamped types.
What is the general opinion about
Weeden Island and agriculture now? It
was my long ago guess that it was of
some consequence in the economy of the
Weeden Island sites, even the coastal
plain ones. Presumably, farming would
have been of more importance further
inland and along major streams. Was it
becoming of greater importance as time
passed? Forney's and White's referenc
es to an increase in the number of
Weeden Island II sites may reflect
this. This time period the centuries
we think of as the Middle Woodland
decline in the Ohio and Central Missis
sippi Valleys seems to have seen
something of a florescence in the far
south. It is all a part of the larger
and complex problem of the changeover
or transition between the general
Hopewellian and Mississippian climaxes
in the Eastern United States.
What of Fort Walton? I wil admit to
having been somewhat less interested in
it than in Weeden Island or Santa Rosa-
Swift Creek. After all, it was more
grubbily represented along the Gulf
shores than these earlier cultures, and
I sensed the catchall quality of the
construct which embraced such disparate
places as the temple mound at Fort
Walton beach (honored as the type site
because it was known in the literature)
and Lake Jackson, that impressive
temple mound establishment near
Tallahassee (up to then virturally
unknown). What I was doing was brack
eting off my last time period, a period
in which Northwest Florida was reacting
to its contacts with Mississippian
cultures in the north, and in which
these reactions produced a number of
quite distinct developments. You are
beginning to get a handle on it now.
Stowe and Fuller, at the western end of
the geographical range, have blocked
out a Bottle Creek phase (A.D. 1200-
1450) and a Bear Point phase (A. D.
1450-1700) as representing middle and
late divisions of a general
Mississippian Period. Brose seems to
suggest an early division of this
period for Lake Jackson. I take it
this would be ca. A.D. 1000-1200 and on
a chronological level with Macon
Plateau in central Georgia? Farther
south, of course, on the Gulf Coast, is
Safety Harbor. Does it date to the
full range of A. D. 1000-1700? Or was I
correct in sliding an Englewood Period
and culture between Weeden Island and
Safety Harbor? This moves beyond the
(38, 1985)

purview of your symposium but is
certainly a closely related matter.
From the papers I get the idea that
there is a consensus that Mississippian
type cultures, at Lake Jackson and at
other inland places, were representa
tive of agricultural towns and villages
but that the immediately coastal sites
were probably fishing and shellfishing
stations utilized seasonally.
Claassen, particularly, contributes to
such an interpretation, which seems a
likely and reasonable one.
There is, throughout most of the
symposium papers, a healthy new archae
ological emphasis on in situ develop
ment of the Northwest Florida cultures.
What I swept up together as Fort Walton
and Safety Harbor is certainly the
outcome of subtler, more complex
processes of contact. But to under
stand just what went on, it will
continue to be necessary to explore the
network of relationshops between Gulf
Coast cultures and sites and those
farther north for these relationships
are as much a part of process as
cultural-natural environmental interac
tions. And in making this plea for
continued interest in "diffusion,"
whatever manifold meanings this term
may have, I hope you will not altogeth
er forget the possibility of more
remote and exotic contacts. It seems
to me highly likely that the Middle
Woodland (Santa Rosa-Swift Creek and
Crystal River) and immediately post-
Middle Woodland (Weeden Island)ceramic
flambuoyance of the Florida Gulf Coast
must have something to do with its
relative proximity to Mesoamerica (see,
again, Bullen, 1966).
Harvard University
March 1985

Louis D. Tesar
Florida Bureau of Historic Preservation
Tne papers in tnis issue range from
reaffirmations of already understood
data to efforts to recapitulate what is
know into more usefull syntheses. They
all should be viewed as another step in
a continuing process, and not as final
statements. Archaeological researcn
and theories are dynamic, or should be.
When they become static and we believe
that we have the final answers. Then
something is wrong. This is not to say
that we will not get closer and closer
to answering the questions which we ask
of our data. Rather, it is to say that
as our data base increases and our
methodology improves, we should shift
from broad generalized questions to
increasingly more specific questions.
It is gratifying to have Gordon R.
Willey's comments on the Archaeology of
Northwest Florida in this issue. Dr.
Willey's reminiscences place his (and
Dr. Woodbury's) work in 1939-40, and
his subsequent report, Archeology of
the Florida Gulf Coast (1949), in its
proper context. His comments on the
papers in this issue continue that
process and offer suggestions for the
Jerald T. Milanich's discussion and
comments are quite incisive. I will
not repeat his summaries of the papers
in this issue. Rather I will comment
on problem areas which are apparent in
some of them. These comments are
offered as an alternative viewpoint to
provide our readers with additional
information upon which to draw their
own conclusions. As Brose (this issue)
has noted there are apparently defects
in my own (Tesar 1980a:21-274) synthe
sis of the subject area's prehistory
and history. Thus, the "real" answer
may lie somewhere between, and not with
any of the viewpoints provided here.
It is up to future researchers, per
haps, to provide more complete answers
to the problems raised.
David S. Brose in his paper has provid
ed a wide ranging review of the Archae
ology of Northwest Florida and Adjacent
Borderlands during the time since
Willey presented his synthesis some 35
years ago. Brose acknowledges that
Willey did a remarkably good job with
the data available, but concludes that
it is time to reevaluate his tempo
ral/cultural constructs. He seems to
imply that Willey's constructs continue
to be accepted uncritically. This is
unfortunate since most of us have been
under the impression, and Brose's
review provides documentation for this
assumption, that Willey's work has been
continuously reevaluated. If he means,
however, that it is time to prepare a
new synthesis of the Archaeology of
Northwest Florida in a single reference
volume, then I and most other profes
sionals conducting Florida Gulf Coast
archaeological research agree.
Nancy Marie White entered the field of
Florida and adjacent borderlands
archaeology in the Chattahoochee-Flint-
Apalachicola drainage system, the only
truely Woodlands environment in Flori
da. She has been (and continues to be)
an enthusiastic and conscientious
researcher. Her efforts in organizing
the symposium which led to the papers
in this issue, and her assistance as
the Guest Editor are appreciated.
While I agree with much of Nancy's
summary and am encouraged by the
research questions which she is asking,
as she knows, I disagree on some
points. For instance, both Nancy and
Dave Brose point to difficulties in
sorting check-stamped ceramics and seem
to conclude that the problems are such
as to make the effort useless. While I
agree that sorting check-stamped
ceramics can be difficult (and occas-
sionally impossible) with small sam
ples and isolated sherds, I disagree
that it cannot be done consistently and
be indepently replicated with sufficent
Vol. 38 Numbers 1-2 Part 2 THE FLORIDA ANTHROPOLOGIST July, 1985

accuarcy to sort into distinct types.
It can and has been done for over forty
years. Sorting check-stamped ceramics,
however, involves consideration of
several variables, including paste,
method of manufacture, hardness,
surface treatment, form, and so forth.
By way of brief example, since nearly
all of the papers in this issue refer
ence Willey's Archeology of the Florida
Gulf Coast (1949), referring to the
check-stamped ceramics in that base
work seems appropriate. Willey listed
the definitions for Deptford Linear
Check-Stamped (1949:354-356), Deptford
Bold Check-Stamped (1949:357), Gulf
Check-Stamped (1949:387-388), Wakulla
Check-Stamped (1949:437-439), Biscayne
Check-Stamped (1949:445-446) and Leon
Check-Stamped (1949:491-492). Biscyane
Check-Stamped with its distinct chalky
paste will be omitted from this discus
sion. Deptford Linear Check-Stamped
with its distinct linear stamping will
also be omitted. There is little
distinction in hardness or thickness
between the four remaining types,
although Gulf Check-Stamped is general
ly thinner than the others. All are
manufactured using a coiling technique,
and while Willey does not discuss it I
have noticed that Deptford Bold Check-
Stamped sherds consistently show coil
fractures while the other three types
are better annealed and generally
fracture randomly. While Leon Check-
Stamped has a fine sand and coarse-grit
(or grog) temper and Gulf Check-Stamped
generally has a fine sand and mica
temper, there is considerable overlap
in the temper of the Deptford Bold and
Wakulla Check-Stamped types. However,
the surface texture of each type
varies. Deptford Bold Check Stamped
vessel interiors are generally smoothed
(showing striations from smoothing) and
occassionally have a low polish. While
somewhat subjective I have found
Deptford sherds generally texture akin
to modern-day cement. Gulf check-
stamped and Wakulla check-stamped both
have generally well smoothed, with some
polish interior surfaces, except for
weathered sherds, while Leon check-
stamped is generally smoothed and grog
is generally visible in the vessel
surface. Stamping in all four types
generally covers most, if not all of
the vessel exterior, and except for
Leon check-stamped there is wide
variation in the size of the checks.
However, while Deptford Bold Check-
Stamped and Gulf Check-Stamped
generally nave one set of parallel
lands dominant over the other, Wakulla
check-stamped and Leon check-stamped
generally have stamps of equal lands.
While Gulf check-stamped rims may be
scalloped or notched, and Wakulla
check-stamped rims may be folded or
have an appliqued band (false fold),
and Leon check-stamped may have a near
rim pinching or ticking, all four types
also have simple unmodified rims.
All of the preceeding is to say that
while it is not possible to sort check-
stamped ceramics into their proper
types all of the time, by taking a
multiple-variable approach it is
possible to confidently sort the
majority of such sherds with or without
the presence of other marker types to
reinforce the sorting identification.
The above is not to say that one of us
is right and the other wrong, but
rather to point out that a problem
exists, and that it is resolvable.
Perhaps studies such as Doran's PIXE
analysis will help in its resolution.
It is urged that someone make an effort
to restudy check-stamped ceramics, and
prepare a comparative typology paper.
Since Milanich has (almost) offered to
publish a new synthesis on Florida Gulf
Coast Archaeology, I will make a
similar offer for the publication in
The Florida Anthropologist of a concise
check-stamped ceramic typology.
Stowe and Fuller's research on the
Bottle Creek and Bear Point Phases of
the Pensacola Variant of Fort Walton,
particularly their efforts to go beyond
the ceramics to address other culture
traits should be well received. They
have in their work, perhaps, defined a
study unit or historical context for

Alabamas comprehensive historic
preservation plan, as well as providing
information of use to Florida. Full
er's chronology, however, seems a bit
late for the beginnings of Early
Mississippi which I had thought started
around 900-1000, rather than 1100.
Jenkins and Mann's summary of their
Conecuh Drainage project provides a
summary of the prehistory and early
history of a previously little known
area. It supplements the Pensacola
variant research of Fuller and Stowe,
and provides comparative data for
researchers in northwest Florida.
Thomas and Campbell's paper on the
Deptford to Santa Rosa/Swift Creek
transition reinforces and expands past
research. Their efforts to define the
attributes which characterize this
transition provide the elements for
future research to test. Their paper
also demonstrates that CRM archaeology
can and does contribute significantly
to our understanding of an areas
history and prehistory.
Forney's paper points out the misinter
pretations, or misevaluations, which
our past ethnocentric biases have lead
to in assessing "marginal" environ
ments. However, in spite of her
excellent efforts to correct this
problem in the coastal lowlands of the
Apalachicola National Forest by point
ing out the diversity of sites, she
seems to fall into a parallel bias of
failing to adjust her model to account
for the Paleo-environment of more
altithermal, lower sea level times.
This is a very common problem among
researchers in Florida and elsewhere.
Claassen's shellfish utilization paper
is important as it shows that there are
a lot of data which can be gotten from
food remains. While we need to refine
our analysis of artifacts, it is
apparent that the midden remains
themselves can contribute significantly
to our understanding of not only
subsistence activities but the past
environments in which those activities
occurred. How people responded to and
affected their environment and its
changes is an important aspect of
archaeological research.
Finally, Judy Bense's Hawkshaw Project
and its relationship to archaeological
resource management in Pensacola has
importance beyond the project. It
serves as an example to emphasize the
fact that if archaeological research
and site preservation are to continue
and be accepted as important aspects of
an area's planning and development,
then we who care about and are familiar
with such cultural resoruces are going
to have to work with the public and
public officials. It is the responsi
bility of both amateur and professional
archaeologists to attend public meet
ings and work with local officials to
educate and assist them in historic
preservation concerns, and to assure
that such concerns are included in
local ordinances and as part of local
government comprehensive plans. We
should do this by focusing on the value
of historic preservation to the commu
nity, both educationally and economi
cally. We should also work with
regional and state agencies to assure
that historic preservation concerns are
adequately met in regional and state
wide comprehensive plans. Since these
plans will set policies and priorities
of governmental decision making for
years to come, if we fail to take the
time and effort to make our contribu
tion, then we have only ourselves to
blame if archaeological resources are
not adequately addressed. In address
ing historic preservation concerns we
should also support the preservation of
historic structures, and make sure that
the historic archaeological features
associated with those properties are
considered in the planning process.
Louis D. Tesar
Bureau of Historic Preservation
Division of Archives, History and
Records Management
The Capitol
Tallahassee, FL 32301-8020
(38, 1985)

Ahlbert, M.S., A.C.D. Leslie and J.W. Winchester
1978 The Chemical State of Particulate Sulfur
in Ambient Aerosols Determined by PXXE
Analysis. Nuclear Instruments and
Methods 149:451-455.
Andrews, J.D.
1953 Fouling Organisms of Chesapeake Bay.
Chesapeake Bay Institute Inshore Survey
Program. Ref. 53-3. Interim Report 17.
Aten, lawrence E.
1981 Determining Seasonality of Rangia Cuneata
from Gulf Coast Shell Middens, Texas
Archaeological Society Bulletin 52:179-
BaiJot-Stroobants and F. Bodart
1977 Ancient Pottery Analysis by Proton
Bombardment and Mossbauer Spectroscopy.
Nuclear Instruments and Methods
Belovich, Stephanie, David S. Brose, and Russell
Weisman, with Nancy M. White
1983 Archaeological Survey at George W.
Andrews lake, Alabama and Georgia .
Report to the D.S. Army Corps of Engineers
Mobile District. Cleveland Museum of
Natural History Archaeological Research
Report No. 37.
Bense, Judith A.
1983 Settlement Pattern, Climate and Marine
Ecosystem Evolution Correlations in the
Escambia Bay Drainage System, Northwest
Florida. Paper presented at the
Southeastern Archaeological Conference,
Columbia, S.C.
Bigelow, Andrew
1853 Observations of Some Mounds on the Tensaw
River, American Journal of Science, 65,
Article 21.
Bradley, Cnad 0. and Ronald L. Mitchelson
1984 A Cultural Resource Survey of Fort Rucker,
Alabama. Report to Archaeological
Services Branch, National Park Service,
Atlanta. Southeastern Wildlife Services,
Inc., Athens Georgia.
Brose, David S.
1979 An Interpretation of the Hopewellian
Traits in Florida. Hopewell Archaeology:
The Chillicothe Conference. D. Brose and
G. Greber, eds. Midcontinental Journal of
Archaeology Special Papers 3:141-149.
1984a From tne Southeatern Ceremonial Complex to
the Southern Cult: You Can't Tell the
Players Without A Program. Paper
presented to the Cottonlandia Museum
Conference on the Southern Ceremonial
Complex, Greenwood, Misslsslppian,
September 1984.
1984b Mlssissippian Period Cultures in North
western Florida. In Gulf Coast Pre
history, David Davis, ed. Ripley P.
Bullen Monographs in Anthropology and
History from the Florida State Museum.
1984c Seeing the Mid-South and Lower Mississippi
Valley from the Southeast: Second
Century Stasis and Status. Paper pre
sented at the Fifth Annual Mid-South
Archaeologist Conference Pinson,
Tennessee, June 1984.
Brose, David S., Ned J. Jenkins and Russell Weisman
1983 Cultural Resources Reconnaissance Study
of the Black-Warrior-Tombigbee System
Corridor, Alabama: Vol. 1, Archaeology.
Report to the D.S. Army Corps of Engineers
Mobile District.
Brose, David S. and George W. Percy
1974 An Outline of Weeden Island Ceremonial
Activity in Northwest Florida. Paper
presented at the Society for American
Archaeology meeting, Washington D.C.
1978 Fort Walton Settlement Patterns. In
Misslsslppian Settlement Patterns, Bruce
D. Smith, ed. Academic Press, New York.
Buchanan, Alan C.
1980Mussels (Naiades) of the Meramec River
Basin. Missouri Department of Conser
vation Aquatic Series, No. 17.
Bullen, Ripley P.
1950 An Archaeological Survey of the
Chattahoochee River Valley in Florida.
Journal of the Washington Academy of
Sciences 40:101-125.
1958 Six Sites Near the Chattahoochee River
in the Jim Woodruff Reservoir Area,
Florida. River Basin Surveys Papers, No.
14. Bureau of American Ethnology Bulletin
169:316-358. SmithBonian Institution,
Washington, D.C.
1966 Stelae at the Crystal River Site, Florida.
American Antiquity 31(6):861-865.
1968 Area Group Discussions and Pottery Types:
Report of the Florida-South Georgia Group.
Southeastern Archaeological Conference
Bulletin No. 8.
1972 The Orange Period of Peninsular Florida.
In Fiber-tempered Pottery in Southeastern
United States and Northern Columbia: Its
Origin, Context, and Significance. Ripley
P. Bullen and James Stoltman, eds. pp.
9-33. Florida Anthropological Society
Publications No. 6, Fort Lauderdale.
Caldwell, Joseph R.
1958 Trend and Tradition in the Prehistory of
the Eastern United States. American
Anthropological Association Memoir No. 88.
1978 Report of excavations at Fairchild's
Landing and Hare's landing, Seminole
County, Georgia. Betty A. Smith, ed.
Report submitted to the National Park
Service, Tallahassee.
Carlson, David L.
1983 Rangia cuneata as a Seasonal Indicator for
Coastal Archaeological Sites in Texas.
Paper read at the Annual Meeting of the
Society for American Archaeology.
Pittsburg, Pennsylvania.
(38, 1985)

Carlsson, Iars-Eric and K. Roland Akelsson
1981 Rapid Determination of Major and Trace
Elements in Geological Material with
Proton-Induced X-Ray and Ganma-Ray
Emmission. Nuclear Instruments and
Methods 181:531-537.
Castagna, M. and P. Chanley
1973 Salinity Tolerance of Some Marine Bivalves
From Inshore and Estuarine Environments in
Virginia Waters on the Western Mid-
Atlantic Coast. Malacologla 12:47-96.
Chance, Marsha
1977 An Archaeological and Historical Survey of
Miscellaneous Areas in the Ocala, Osceola,
and Apalachicola National Forests in
Florida. Division of Archives, History
and Records Management Miscellaneous
Project Report Series, No. 44.
Chase, David W.
1966 A Stratified Archaic Site in Lowndes
County, Alabama. Florida Anthropologist
1-9 (2-3): 91-114.
1968 Pottery Typology Conmittee for Central
Alabama. Southeastern Archaeological
Conference Bulletin 8:11-22.
1969 New Pottery Types from Central Alabama.
Southeastern Archaeological Conference
Bulletin 10:17-20.
1978 An Outline of Central Alabama Prehistory
Paper presented at the winter meeting of
the Alabama Archaeological Society,
Auburn, Alabama.
1982 National Forests in Alabama. Cultural
Resources Overview. Department of
Agriculture, Montgomery, Alabama.
Chen, K., H. U, C. Ren, G. Tang, X. Wang, F. Yang
and H. Yao
1980PIXE Research with an External Beam.
Nuclear Instruments and Methods 168:437-
Claassen Cheryl P.
1982 Shellfishing Patterns: An Analytical
Study of Prehistoric Shell from North
Carolina Coastal Middens. Ph.D Disser
tation, Harvard University.
1983 Prehistoric Shellfishing Patterns in
North Carolina. British Archaeological
Reports, International Series 183:211-223
Cordell, Linda S.
1984 Southwestern Archaeology. Annual Review
of Anthropology 13:301-332.
Cottier, John W.
1970 The Alabama River Phase. Appendix in
Archaeological Salvage Investigations in
the Miller's Ferry Lock and Dam Reservoir
Report to the National Park Service.
Mound State Monument, Moundville, Alabama.
Cullity, B.D.
1978 Elements of X-Ray Diffraction. 2nd edi
tion. Addison-Wesley, Reading, Mass.
Curren, Caleb B., Jr.
1976 Prehistoric and Early Historic Occupation
of the Mobile Bay and Mobile Delta Area of
Alabama. Journal of Alabama Archaeology.
1978 The Zooarchaeology of the D'Olive Creek
Site (lBal96) Baldwin County, Alabama.
Journal of Alabama Archaeology 24(1).
Davis, David D.
1981 Ceramic Classification and Temporal
Discrimination: A Consideration of
Later Prehistoric Stylistic Change in
the Mississippi River Delta. Mid-
continential Journal of Archaeology
Davis, David D., editor
1984 Perspectives on Gulf Coast Prehistory.
University Presses of Florida,
Demortler, C. and T. Hackens
1982 Milliprobe and Microprobe Analysis of
Gold Itmes of Ancient Jewelry. Nuclear
Instruments and Methods 197:223-236.
DePratter, Chester B.
'1975 The Archaic in Georgia. Early Georgia
Dickens, Roy S., Jr.
1971Archaeology in the Jones Bluff Reservoir
of Central Alabama. Journal of Alabama
Archaeology 17(1).
Doran, Glen H.
1984a Proton Induced X-Ray Emission Analysis of
Prehistoric Ceramics. Florida Anthro
pologist^ 37(3):115-119.
1985a Report on the Florida State University
Archaeological Excavations at Gulf
Islands National Seashore. Report sub
mitted to the Southeastern Archaeological
Center, National Park Service, Tallahassee
1985b PIXE Analysis of Texas Ceramics. Draft
Manuscript on file, Department of
Anthropology, Florida State University,
Doran, Glen H., Pamela Houmere and Sene Bauman
1985 PIXE Analysis of Florida, Alabama, and
Georgia Ceramics. Draft Manuscript on
file, Department of Anthropology, Florida
State University, Tallahassee.
Dunnell, Robert C.
1984 The Americanist Literature for 1983:
A Year of Contrasts and Challenges.
American Journal of Archaeology 88:489-
Fairbanks, Charles H. and B. Keel
1966 North Florida Deptford. Proceedings of
the 17th Southeastern Archaeological
Conference, pp. 8-9.
Fairbanks, Lawrence D.
1963 Biodemographic Studies of the Clam Rangia
cuneata Gray. Tulane Studies in Zoology
Fewkes, J.W.
1924 Preliminary Archaeological Explorations
at Weeden Island, Florida. Smithsonian
Miscellaneous collections Vol. 76, No. 13.
'(38, 1985)

Fontes, P., I. Brissaud, G. Lagarde, J. Leblanc,
A. Person, J.F. Saliege and C. Heltz
1984 PIXE Analysis and Provenance Study of
Archaeological Potsherds from West Africa.
Nuclear Instruments and Methods B3:404-407
Fornaro, Robert J.
1974 A Conecuh River Site (Cv30), Alabama.
Florida Anthropologist 27.
Forney, Sandra Jo
1983 Archaeological Investigation of the River
Sink Tract, Apalachicola National Forest,
Wakulla County, Florida. Paper presented
at the 35th Annual Meeting of the Florida
Anthropological Society, Tallahassee.
Fuller, Richard S. and Noel R. Stowe
1982 A Proposed Typology for late Shell-
Tempered Ceramics in the Mobile Bay/Mobile
Tensaw Delta Region. In Archaeology in
Southwestern Alabama. A Collection of
Papers. Cailup Curren, ed. pp. 45-93.
Alabama Tombigbee Regional Commission,
Camden, Alabama.
Fuller, Richard S., and Diane E. Fuller
1984 Ceramic Rim Effigies in Southwest Alabama.
Submitted to the Journal of Alabama
Archaeology and on file at the University
of Alabama, Mobile.
Fuller, Richard S., Diane Fuller and Noel R. Stowe
1984 The Forks Project: An Investigation of
the late Prehistoric-Early Historic
Tradition in the Alabama-Tombigbee
Confluence Basin. Report on file at the
Alabama Historic Commission.
Gagliano, Sherwood M., et al.
1979 Cultural Resources Survey of the
Baratarla, Segnette, and Rigaud
Waterways, Jefferson Parrish, Louisiana,
vol. I and XI. CoaBtal Environments,
Inc., Baton Rouge, La.
Gardin, Jean-Claude
1980 Archaeological Constructs: An Aspect of
Theoretical Archaeology. Cambridge
University Press, Cambridge.
Godwin, W.F.
1967 Preliminary Survey of a Potential Hard
Clam Fishery. Georgia Fish and Game
Commission, Marine Fisheries Division,
Series 1.
Griffin, James B.
1946 Cultural Change and Continuity in Eastern
United States Archaeology. In Man in
Northeastern North America, F. Johnson,
ed. Papers of the Robert S. Peabody
Foundation for Archaeology 3:37-95.
Hart, James
1980 Soils and Vegetative Types on the
Apalachicola National Forest. Manuscript
on file, U.S. Forest Service, Tallahassee.
Heimlich, Marion Dunlevy
1952 Guntersville Basin Pottery. Geological
Survey of Alabama Museum Paper 32.
University of Alabama.
Henefield, Susan M.
1985 Woodland Settlements and River Channels in
the Lower Chattahoocnee-Upper Apalachicola
Valley. PaDer Presented at the Florida
Anthropological Society meeting, Daytona
Beach, 20 April 1985. On file at the
University of South Florida Department of
Anthropology, Tampa.
Hill, Frederick C.
1983 Unexplained Occurrence of the Mactrid
Bivalve, Rangia cuneata from the
Arrowhead Farms Indian Site Near
Louisville, Kentucky. The Nautilus
97 (2):79-81.
Holmes, Nicholas H., Jr.
1963 The Site on Bottle Creek. Journal of
Alabama Archaeology 9(1).
Holmes, William H.
1903 Aboriginal Pottery of the Eastern United
States. Bureau of American Ethnology,
Twentieth Annual Report, Washington D.C.
Jenkins, Ned J.
1975 Archaeological Investigations in the
Gainesville Lock and Dam Reservoir: 1974.
Report on file at the University of
Alabama, Department of Anthroplogy.
1981 Gainesville Lake Area Ceramic Descriptions
and Chronology. Archaeological Investi
gations in the Gainesville Lake Area of
the Tennessee-Tombigbee Waterway. Office
of Archaeological Research,
University of Alabama, Report of
Investigations. No. 12
1982 Archaeology of Gainesville Lake Area:
Synthesis. Archaeological Investigations
in the Gainesville Lake Area of the
Tennessee-Tombigbee Waterway. Office of
Archaeological Research, University of
Alabama, Report of Investigations. No. 23.
1983 Ceramic Chronology. In Cultural
Resources Reconnaisance Black Warrlor-
Tombigbee System Corridor, Alabama:
Archaeology. By D. Brose, N. Jenkins
and R. Weisman, University of South
South Alabama, Mobile.
Jenkins, Ned J. and Jerry J. Nielson
1974 Archaeological Salvage Investigations
at the West Jefferson Steam Plant Site,
Jefferson County, Alabama. Report on
file at Mound State Monument, Alabama.
Kelly, A.R.
1950a An Early Flint Industry in Southwest
Georgia. University of Georgia
Laboratory of Archaeology Manuscript
No. 53.
1950b Survey of the Lower Flint and
Chattahoochee Rivers. Early Georgia
Kniffen, Fred B.
1936 A Preliminary Report on the Indian Mounds
and Middens of Plaquemines and St.
Bernard Parishes. Lower Mississippi
River Delta: Reports on the Geology of
Plaquemines and St. Bernard Parishes,
Geological Bulletin No. 8. Department
of Conservation, Louisiana Geological
Survey, New Orleans.
(38, 1985)

Kraft, ,
Vernon James, Jr., and Tim S. Mistovich
Walter F. George lake: Archaeological
Survey of Fee Owned Lands, Alabama
Georgia. Report to the U.S. Army
Corps of Engineers, Mobile District.
Office of Archaeological Research,
University of Alabama, Report of
Investigations No. 42.
Gary D.
Partial Cultural Resources Inventory of
Tyndall Air Force Base, Florida. Final
report submitted to National Park
Service, Interagency Archaeological
Services, Atlanta. Southeast
Conservation Archaeology Center, Florida
State University, Archaeological Research
Reports No. 7.
C., and G. Margules
Sediment Patterns, Physiographic
Characteristics of the Water Mass and
Foraminiferida Distribution in Indian
River Bay Coastal Delaware. South
eastern Geology 12:223-250.
Richard A.
Taxonomic Practice and Middle Missouri
Prehistory: A Perspective on Donald
J. Lehmers Contributions. Plains
Anthropologist Memoir 13.
William C.
Ten Middens on the Navy Live Oak
Reservation. Florida Anthropologist
Alligator lake, a Ceramic Horizon Site
on the Northwest Florida Coast. Florida
Anthropologist 18:83-124.
Yulee W.
Salvage Archaeology at Fort Walton Beach,
Florida. Florida Anthropologist
The Fort Walton Culture West of the
Apalachicola River. Newsletter of the
Southeastern Archaeological Conference
Yulee and Carolyn B. Hawkins
Pottery of the Fort Walton Period.
Temple Mound Museum, Fort Walton Beach,
Donald J.
Introduction to Middle Missouri
Archaeology. National Park Service
Anthropology Papers, No. 1. National
Park Service, Washington D.C.
, Douglas
The Moundvllle Phase and Its Position
in Southeastern Prehistory. Ph.d.
Dissertation. Harvard University,
Department of Anthropology. Cambridge,
Pottery Types of the Moundvllle Phase.
In Southeastern Archaeological Conference
Bulletin 2. Cambridge, Massachusetts.
A Summary of the Moundville Phase.
Journal of Alabama Archaeology 12(1).
Moundville, Alabama.
Milanich, Jerald T.
1971 The Deptford Phase: An Archeaological
Reconstruction. Unpublished Ph.D.
dissertation, University of Florida,
Department of Anthropology, Gainesville.
1973a A Deptford Phase House Structure at
Cumberland Island, Georgia. Florida
Anthropologist 26(3):105-118.
1973b The Southeastern Deptford Culture: A
Preliminary Definition. Bureau of
Historic Sites and Properties,
Division of Archives, History
and Records Management, Bulletin
3:51-63. Tallahassee.
1974 life in a Ninth Century Indian Household
a Weeden Island Fall-Winter Site on the
Upper Apalachicola River, Florida. Bureau
of Historic Sites and Properties
Division of Archives, History and Records
Management Bulletin 4:1-44. Tallahassee.
Milanich, Jerald T., Ann S. Cordell, Vernon J.
Knight, Jr., Timothy A. Konler and Brenda Slgler-
1984 McKelthen Weeden Island: The Culture of
Northern Florida, A. D., 200-900.
Academic Press, New York
Milanich, Jerald T., and Charles Fairbanks
1980 Florida Archaeology. Acedemic Press, New
Moore, Clarence B.
1901 Certain Aboriginal Remains of the North
west Florida Coast, Part I. Journal of
the Academy of Natural Sciences,
Philadelphia, 11:450-514.
1902 Certain Aboriginal Remains of the
Northwest Florida Coast, Part II. Journal
of the Academy of Natural Sciences,
Philadelpnla, 12:127-355.
1903 Certain Aboriginal Mounds of the
Apalachicola River. Journal of tne
Academy of Natural Sciences.
Pniladelpnia, 12:440-490.
1918 The Northwestern Florida Coast Revisited.
Journal of the Academy of Natural Sciences
Nie, N., C. Hull, J. Jenkins, K. Steinbrenner and
D. Brent
1975 Statistical Package for the Social
Sciences. McGraw-Hill, Heightstown,
Parker, J.C.
1965 Bottom Fauna StudyDistribution and
Relative Abundance of Rangia cuneata. In
Annual Report of Bureau of Commerical
Fisnerles Biological Lab, Galveston,
Texas, Fiscal Year 1965. Fish and Wild
life Service. BCF Circular 246.
Percy, George W.
1971 Preliminary Report to the Divison of
Recreation and Parks, Department of
Natural Resources, State of Florida, on
Archaeological Work in the Torreya State
Park During the Year of 1971, by the
Department of Anthropology at Florida
State University. Manuscript on file,
Florida Department of Natural Resources.
(38, 1985)

Full Text