Frontier process in eighteenth century colonial Georgia : an archeological approach


Material Information

Frontier process in eighteenth century colonial Georgia : an archeological approach
Physical Description:
xiii, 320 leaves : ill. ; 28 cm.
Honerkamp, Nicholas, 1950-
Publication Date:


Subjects / Keywords:
Antiquities   ( fast )
Fort Frederica National Monument (Ga.)   ( lcsh )
Antiquities -- Saint Simons Island (Ga.)   ( lcsh )
Georgia -- Fort Frederica National Monument   ( fast )
Georgia -- Saint Simons Island   ( fast )
Anthropology thesis Ph. D   ( lcsh )
Dissertations, Academic -- Anthropology -- UF   ( lcsh )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 1980.
Includes bibliographical references (leaves 296-308).
Additional Physical Form:
Also available online.
General Note:
General Note:
Statement of Responsibility:
by Nicholas Honerkamp.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 023443539
oclc - 07333734
lcc - F292.G5 H6 1980a
System ID:

Full Text








A great many people contributed to this study. Some

offered suggestions and encouragement; others, their labor

and expertise. In the case of the National Park Service,

funds were provided to carry out the research. Individuals

will be mentioned here with the understanding that their

contributions can never be acknowledged to the extent that

they deserve.

The members of my doctoral committee, Charles H.

Fairbanks, Kathleen A. Deagan, Prudence M. Rice, Jerald T.

Milanich, and John K. Mahon, have been extremely generous

with their advice and criticism during the course of the

project. Besides their impact on the present study, two

members in particular have had a profound influence on my

development as an archeologist. Ten years ago Dr. Fairbanks

stimulated my interest in anthropology through his lectures

on North American prehistory. After attending a Fairbanks-

led archeological field school, that interest changed into an

overwhelming ambition to persue archeology on a full-time

basis. Dr. Deagan helped me realize the importance of a

systematic, scientific approach to the study of human

behavior. I am indeed fortunate to have worked closely with

both of these scholars.

I was also fortunate to have been associated with the

members of the three field crews that participated in this

project. The field people are:

Wendy Bolles Charles Chambers Julie Emrich

Bruce Ferguson Kevin Galleger Jane Gray

Wayde Hanna Joan Hebb Patricia McKay

Roberta Owens Martha Pinello Tricia Sokol

Diane Sylvia Patricia Welsh Dan Yannette

Drew Yaros

Among the many volunteers that contributed to the project

were John Battle, Anita Fulton, Julia Furgeson, Lisa

Laudadia, Susan Loftin, Vincent Pinoso, Robin Smith, Mrs. R.

Welsh, and Clint Wills. Chad Braley assisted with production

of the artifact photographs used in Chapter V.

At the Florida State Museum, the following persons

provided invaluable instruction and assistance during the

faunal analysis: Elizabeth Wing, Elizabeth Reitz, Sylvia

Scudder, Erica Simons, Arlene Fradkin, and Tom Chase. In

addition, Dr. Reitz has contributed much to my knowledge of

colonial resource utilization and is primarily responsible

for stimulating my interest in this subject. Many of the

ideas presented in this study concerning colonial patterns of

faunal use stem from suggestions she has made in her own work

or in our collaborative efforts.

The National Park Service has been extremely supportive

of this project in every way. Deserving special praise for

their cooperation and assistance are Chief Richard Faust,


Archeologist George Fischer, and Laboratory Director James

Stoutamire of the Southeast Archeological Center in

Tallahassee. At Fort Frederica, the entire staff supported

and sometimes actively participated in the project. Former

Supervisor Janet Wolf and Assistant Supervisor George Berndt

were particularly helpful in seeing that the field work was

successfully carried out.

Finally, I would like to thank my wife Robin L. Smith

for her editorial contributions and especially for her

constant encouragement and criticisms. Her personal and

professional support was invaluable during all phases of the

project, and will continue to be essential to me in the




LIST OF TABLES . . . . .


ABSTRACT . . . . . ..


Location and Setting .
Previous Research . .
Project Background . .


Competing Paradigms in Historical
Archeology . . . . . .
Old Particularism . . . . .
New Particularism . . . . .
Pattern Recognition . . . .
Eclectic Approach . ....
Theoretical-Methodological Position


Concepts and Definitions


Historical Context: Geo
Historical Context: Fre


General Temporal-Demographic Parameters
Spatial Parameters * . . *.. .
The Hawkins-Davison Site . . . .
The Hird Site . . . . . ..
The Dobree Site . . . . ..


Excavation Procedures . . . . .
Horizontal Control . . . . ..
Stratigraphy . . . . . ..
Analytical Methods and Results . .
Faunal Analysis . . . . . ..

. . 018

. . 20

. . . 30
* * 31
* . 314^

* 3 314

* . . 41

# . 41
. 45
. . 51
. 55
* 57
. . 59

. . 60

S. 65

S . 66
. . 78
. . 79
. . 80
* . 142

. . vii


* *

* .

* *


Dobree Site Features . . . . . .. 158
Hird Site Features . . . . . . 210
Hawkins-Davison Site Features . . . . 214
Summary . . . . . . . . . 214


Cultural Affiliation . . . . . . 217
Temporal Parameters . . . . . . 220
Site Form and Function . . . . . 225
Refuse Disposal and Artifact Distribution . 262
Subsistence and Diet . . . . . 274


REFERENCES CITED . . . . . . . . . 296



DOBREE SITE . . . . . . . 314


BIOGRAPHICAL SKETCH . . . . . . . . 320


Table 4-1.

Table 5-1.
Table 5-2.

Table 5-3.

Table 5-4.

Table 5-5.

Table 5-6.

Table 5-7.
Table 5-8.

Table 5-9.

Table 5-10.

Table 5-11.

Table 5-12.

Table 6-1.

Table 6-2.

Table 7-1.

Table 7-2.
Table 7-3.
Table 7-4.
Table 7-5.

Table 7-6.

Table 7-7.

Table 7-8.

Table 7-9.

Summary of Street and Lot Dimensions,
Frederica . . . . . . . . . 56
Artifact Class Frequencies, Dobree Site . 88
Identifiable Ceramics from Three Sites,
Frederica, Georgia . . . . . . 90
Product-Moment Coefficients for Three
Ceramic Categories . . . . . . 101
Product-Moment Coefficients for Three Nail
Types, Dobree Site . . . . . . 101
Summary of Furniture Hardware Class
Artifacts, Dobree Site . . . . . 120
Summary of Artifacts in Gun Parts Class,
Dobree Site . . . . . . . . 122
Artifact Class Frequencies, Hird Site . .. 143
Allometric Constants Used in Biomass
Calculations . . . . . . . 148
Species List for the Dobree Site, Lot 31
South, Frederica . . . . . . . 151
Summary of Six Faunal Categories, Dobree
Site, Frederica . . . . . . . 155
Summary of Six Faunal Catergories, Hawkins-
Davison Site, Frederica . . . . . 156
Summary of Six Faunal Categories, Hird
Site, Frederica . . . . . . . 157
Artifact Class Frequencies, Features, Dobree
Site . . . . . . . . . . 162
Frequency Percentages for Four Artifact
Classes, Wells, Dobree Site . . . . 184
Comparison of Total Bone and Artifact Mater-
ials, Hird and Dobree Sites . . . . 232
Soil pH Values, Dobree Site . . . . 234
Emperical Artifact Profile, Dobree Site . 242
Emperical Artifact Profile, Hird Site . . 243
Summary of Three Group Categories, Frederica
And Carolina Sites . . . . . . 244
Product-Moment Coefficients For Six Artifact
Categories, Dobree Site . . . . . 251
Product-Moment Coefficients for Slag,
Clinkers, and Coal, Dobree Site . . . 253
Summary of Three Species From Three Sites,
Frederica, Georgia . . . . . . 282
Ages of Three Species by Element Fusion, Hird
Site, Frederica . . . . . . . 284



Figure 1-1.
Figure 1-2.

Figure 1-3.

Figure 1-4.
Figure 4-1.
Figure 5-1.
Figure 5-2.
Figure 5-3.
Figure 5-4.
Figure 5-5.
Figure 5-6.

Figure 5-7.

Figure 5-8.
Figure 5-9.

Figure 5-10.

Figure 5-11.
Figure 5-12.
Figure 5-13.
Figure 5-14.

Figure 5-15.
Figure 5-16.
Figure 5-17.
Figure 5-18.

Figure 5-19.
Figure 5-20.
Figure 6-1.
Figure 6-2.

Figure 6-3.

Figure 6-4.

Figure 6-5.

Figure 6-6.
Figure 6-7.

Figure 6-8.

Figure 6-9.
Figure 6-10.

The Barrier Islands of Coastal Georgia
Vicinity Map, St. Simons Island,
Georgia . . . . . . . .
Layout of The Fort And Town of Frederica,
circa 1740 . . * * * *. . .
Composite Excavation Map, Frederica . .
1796 Miller Map of Frederica . ....
Excavation Plan, Hawkins-Davison Site .
Excavation Plan, Hird Site . . . .
Excavation Plan, Dobree Site . . .
Actual Area Excavated, Dobree Site . .
Field Work In Progress, Dobree Site . .
Distribution of Postcreamware Ceramics,
Plowzone, Dobree Site . . . ...
Distribution of Identifiable Colonial
Ceramics, Plowzone, Dobree Site . . .
Ceramic Artifacts, Dobree Site . . .
Tableware Ceramic Vessels, Dobree
Site . . . . . . . . .
Cross Section Drawings of Ceramic
Vessels, Dobree Site . . . . .
Wine Bottles, Dobree Site . . . .
Glass Tableware Items, Dobree Site . .
Tableware Artifacts, Dobree Site . .
Architecture and Furniture Group
Artifacts, Dobree Site . . . .
Arms Artifacts, Dobree Site . . . .
Clothing Artifacts, Dobree Site . . .
Personal Group Artifacts, Dobree Site .
Activities Group Artifacts, Dobree
Site . . . . . . . . .
Snaffle Bit, Dobree Site . . . .
Military Artifacts, Dobree Site . . .
Composite Map of Dobree Site Features .
Top of Feature 1 Barrel Well, Dobree
Site . . . . . . . . .
Initial Excavation of Feature 1 Barrel
Well, Dobree Site . . . . . .
Profile of Feature 1 (Barrel Well),
Square 31, Dobree Site . . . . .
Profile of Feature 2 Well Pit, Dobree
S i t e . . . . . . . .* .
Feature 2 Pit and Shaft, Dobree Site .
Closeup of Feature 2 Pit and Shaft,
Dobree Site . . . . . .
Profile of Feature 3 Barrel Well,
Dobree Site . . . . . .* . .
Feature 3 Pit and Shaft, Dobree Site .
Frequency or Weight Percentages for Nine
Artifact Groups, Features, Dobree
Site . . . . . . . . .


* 5

. 7

* 11
* 13
* 55
* 68
. 70
. 72
. 74
. 77

* 93













Figure 6-11.

Figure 6-12.

Figure 6-13.

Figure 6-14.

Figure 6-15.

Figure 6-16.

Figure 6-17.

Figure 6-18.

Figure 6-19.
Figure 7-1.

Figure 7-2.

Figure 7-3.

Figure 7-4.

Figure 7-5.

Figure 7-6.

Figure 7-7.

Figure 7-8.

Profile of Feature 5 Trash Pit, Dobree
S i t e . . . . . . . .* .
Profile of Feature 10 Storage/Trash
Pit, Dobree Site . . . .. .
Composite Map of Dobree Site
Postholes . . . .. . .. .
Distribution of Construction Materials,
Plowzone, Dobree Site . . . ...
Tabby Plaster From Wattle and Daub
Structure, Dobree Site . . . . .
Edge of Feature 29 Pallisade Wall Trench,
Dobree Site . . . . . . . .
Profile of Feature 29 Pallisade Wall
Trench, Dobree Site . . . . . .
Profile of Feature 29 and Underlying
Postholes, Dobree Site . . . . .
Compostie Map of Hird Site Features . .
Distribution of Slag Waste, Plowzone,
Dobree Site . . . . . . .
Contours of Mean Ceramic Dates for
Plowzone Ceramics, Dobree Site . .
Distribution of Oriental Porcelain
Fragments, Plowzone, Dobree Site . .
Distribution of Wine Bottle Fragments,
Plowzone, Dobree Site . . . . .
Distribution of Faunal Materials,
Plowzone, Dobree Site . . . ...
Distribution of White Clay Pipe
Fragments, Plowzone, Dobree Site . .
Distribution of Wrought and Square
Nails, Plowzone, Dobree Site . . .
Distribution of Window Glass Fragments,
Plowzone, Dobree Site . . . . .

















Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy



Nicholas Honerkamp

December 1980

Chairman: Charles H. Fairbanks

Major Department: Anthropology

Frontier regions offer unique opportunities for

anthropologists to study the adaptive responses of intrusive

colonizing societies to new social and natural conditions.

The process of colonization is reflected archeologically

through the patterning of settlement structure and function.

The purpose of this study is to examine the archeological

correlates of a British colonial frontier adaptation in

Georgia and to interpret this evidence through reference to

documented 18th century sociocultural, political, and

demographic factors.

Analysis of archeological materials from three sites at

Fort Frederica, Georgia (1736-1750), and contemporaneous


British colonial sites in the southeastern United States is

used to address hypotheses concerning frontier adaptations.

Using Stanley South's "pattern recognition" methodological

approach to organize the data, interpretation of the evidence

from Frederica was accomplished through the application of a

frontier model proposed by Kenneth E. Lewis. Questions

concerning cultural affiliation, temporal parameters, and

site structure and function were examined at the town and lot

level. Additionally, the nearly complete excavation of one

of the sites allowed an empirical test of South's Brunswick

Pattern of Refuse Disposal to be made. Zooarcheological data

from Frederica were also presented for comparison with a

traditional model of resource utilization designated as the

British Barnyard Complex.

Results of the analysis demonstrated the applicability

of Lewis' model to the Frederica data. Due to the need for

defense against a competing state power, the settlement

pattern at Frederica more closely resembled the concentrated,

evenly-spaced row pattern of European market towns than the

dispersed, random arrangement found at frontier towns such as

Camden, South Carolina. At the lot level, clustering of

subsurface features, maximum utilization of lot elements,

differential bone deposition, and demarcation of lot

boundaries is believed to have resulted from the same

demographic and sociopolitical factors that shaped the

settlement pattern of the town as a whole. Quantitative and

qualitative evidence of site function was also found to

reflect frontier constraints and conditions. As with Camden,

the occurrence of domestic and craft activity by-products at

the Frederica sites indicates the multifunctional nature of

frontier occupations. Comparison of artifact profiles from

sites at Frederica, Camden, and Brunswick Town, South

Carolina, revealed patterning of frequency relationships that

were explicable in terms of intersite functional differences.

Graphic presentation of artifact distributions at one of

the sites at Frederica indicates a departure from the

Brunswick Pattern described by South. Instead of disposal of

refuse adjacent to a post-supported structure, there is

evidence of bimodall trash disposal behavior" by the site's

occupants. The first behavioral mode involves dispersed

surface disposal of trash, while the second is associated

with purposeful concentration of trash, particularly food

bone, in subsurface features. An orientation toward trash

disposal in areas adjacent to the house was not associated

with either mode. It is suggested that, instead of

systematic British colonial refuse disposal behavior cited by

South, the limited size of the samples used to define the

Brunswick Pattern may be responsible for its occurrence.

Little correspondence was seen between the traditional

model of British colonial resource utilization and the

zooarcheological evidence from Frederica. The British

Barnyard Complex, characterized by an overwhelming emphasis

on the use of swine, sheep, and less importantly, dairy

cattle, was not replicated at Frederica. Substantial


adjustments to New World conditions are indicated by an

increased reliance placed on wild terrestrial fauna and a

shift from swine to cattle, as measured through comparison of

biomass values derived from archeological bone. Similar

results from contemporaneous British and Spanish colonial

sites in St. Augustine, Florida, indicate that the

traditional colonial foodways model is in need of major




The British colonial settlement of Frederica was

established on St. Simons Island, Georgia, in 1736. Planned

as a major military-civilian fortified outpost, Frederica's

existence owed more to the bellicose machinations of Britain

and Spain than to anything else. Significantly, earthworks

for the walls of a fort were among the first things con-

structed by the 40 English families who founded the town.

Subsequent events during the settlement's short history serve

to underscore the interdependence between military and civil

segments of the community. When the regiment left in 1748,

Frederica's civilian population quickly relocated to other

settled communities or frontier areas. By the early 1750s

only a dozen or so families remained out of the more than

1000 inhabitants who had lived there a few years earlier.

Although it existed as a community for little more than

15 years, a great deal of historical and archeological

information dealing with various aspects of 18th century

British colonial life at Frederica is potentially available.

Since 1945, when it was established as a national monument,

Fort Frederica has been the focus of considerable archeo-

logical and documentary research. As part of the U.S.

National Park system, most of this research has been directed


toward recovering information useful for on-site reconstruc-

tion and interpretation of British colonial lifeways at the

fort and associated town. In recent years this traditional

National Park Service goal has been augmented by research

aimed at addressing problems of anthropological significance,

such as defining socio-economic indicators in the documentary

and archeological records, examining the extent of wild

versus domestic animal utilization, and revealing the

presence of craft activities and delimiting areas where they

were practiced. It should be emphasized that these questions

are complementary to those aimed at site reconstruction, but

they involve different methodological emphases and different

field and analysis techniques.

The archeological research reported here has been sup-

ported by a National Park Service grant for $19,381.00 to the

Department of Anthropology, University of Florida (Contract

No. CX500080754). This project consisted of excavation of a

domestic town lot at Frederica (believed to be Lot 31, South

Ward), analysis of the documentary and archeological data

pertaining to the site, and synthesis of these data and

information derived from previous excavations. Principal

Investigator for the project was University of Florida

Distinguished Service Professor in Anthropology Charles H.

Fairbanks. The author supervised field and preliminary

laboratory work and was responsible for analysis of artifacts

and faunal materials.

Field work was carried out at Lot 31 South in three 10

week periods in the spring and fall of 1978 and in the winter

of 1979. Excavation crews consisted of undergraduates from

the University of Florida and Florida State University, who

were enrolled in the University of Florida Archeological

Field School course, as well as volunteers. There were three

full-time excavators in the spring, five in the fall, and

eight during the winter session, for a total of 160 crew

weeks. Numerous part-time volunteers also contributed to the

project. The author was assisted part-time by two under-

graduate students in cataloging artifacts for a nine week

period following the field work.

Location and Setting

Located approximately 10 kilometers east of Brunswick,

Georgia, Fort Frederica National Monument is situated on a

bluff overlooking the Intracoastal Waterway on the western

edge of St. Simons Island (Figures 1-1 and 1-2). St. Simons

is part of a barrier island chain extending along the

southern Atlantic coast from Amelia Island, Florida, to Cape

Fear, North Carolina (Figure 1-1). As is true of most of the

barrier islands, St. Simons is characterized by diverse

habitats, notably the beach-dune configuration facing the

ocean, the maritime forests on the island interior, and the

tidal marsh lying between the island and the mainland. This

diversity of habitats contrasts sharply with the essentially

homogenous environment of the pine forests on the coastal

plain. Detailed descriptions of geology and ecology of the

Figure 1-1.

The Barrier Islands of Coastal Georgia.

25 km











Figure 1-2.

Vicinity Map, St. Simons Island, Georgia.


. _-- --- F red er i ca. _~ -:_ \."\

{C Marsh GA
1km -N- F-
I dla and Map Location
Iskylan ,____Land Map______ Location___
,o ,, ,:i ... -,

coastal region are provided by Johnson et al. (1974). The

low upland ridge upon which Frederica lies is part of an

ancient marine terrace known as the Princess Ann Formation.

Soils at Frederica conform to the somewhat excessively

drained, fine sand characteristics of the Cainhoy soil phase

that is associated with ridgetops of relict marine terraces

on the Georgia coast (Rigdon and Green 1980).

According to Robin Smith (1978:8-11), the geological

characteristics of areas similar to the one in which Fred-

erica is located have had important implications in human

settlement patterning. Smith has convincingly demonstrated a

pattern of prehistoric and historic settlement activities

which focus on the fine sand bluffs adjacent to the salt

marshes of the coast. The high, well-drained soils of these

areas comprised an important scarce resource supportive of

human life, and as a consequence a succession of indigenous

and immigrant populations has used this land in various ways.

The 18th century British occupation at Frederica is seen as

conforming to this general barrier island marsh-bluff set-

tlement orientation.

As can be seen in Figure 1-2, the fort and town are

located at a sharp bend of the Frederica River. The military

advantages accruing from this position in the context of 18th

century martial technologies have been fully discussed by

others and need no reiteration here (Cate 1943; Ivers 1974;

Manucy 1962). Ironically, some of the factors contributing

to the selection of Frederica's location as a site for a

frontier settlement have also been responsible for the

destruction of part of the archeological record. The depth

of the Frederica River at this bend allowed colonial period

ships of considerable draught to closely approach the fort

for offloading of people, material, and supplies. Over the

years the swift movement of the current at this deep water

point has eroded the east shore of the river, resulting in

the loss of a significant portion of the fort area, including

the entire battery spurwork shown in Figure 1-3. Addition-

ally, the soil characteristics of the upland bluff area that

successfully supported both aboriginal and colonial popula-

tions also attracted more modern occupants. Alterations to

the landscape associated with postcolonial occupations

include plow cultivation and construction of roads and

dwellings, both of which have adversely affected archeolo-

gical resources. The present preservation policy of the

National Park Service has substantially reduced further

degradation of these resources.

Previous Research

Although a good deal of dirt has been moved at Frederica

in the interest of archeology, most of the excavations have

been devoted to delineating the layout of the town and to

exposing military construction features and domestic archi-

tecture (Deagan 1975). In the town itself, all of the

domestic sites facing Broad Street and the northern section

of Cross Street have been excavated (Figure 1-4). Several

interpretive reports have emerged from this work, including

Figure 1-3.


Layout of the Fort and Town of Frederica,
circa 1740.

1 Hird Site, 12 North

2 Hawkins-Davison Site, 1 and 2 South

3 Dobree Site, 31 South

Marsh North Ward
/ I Barracks

(^ "'"^ ^ T______
]Ilfl IIH~ll

\D I I I lZI I

South Wa rd


Figure 1-4.

Composite Excavation Map, Frederica.

Excavated areas are shown in black.


^y v^^/ a e -r -I





Manucy's synthesis of the archeology and history of the

military and civilian portions of the settlement (1960,

1962), Fairbanks' study of the architecture and associated

artifacts from the Hawkins-Davison site (1952, 1956), and

Reese's popularization of Manucy's studies (Reese 1969).

Despite the considerable accumulation of artifactual and

housing data resulting from these projects, analysis of arti-

facts which would allow testing of research hypotheses of a

processual nature has been lacking. More recent work has

been oriented toward addressing problems of anthropological

significance. Deagan's (1972) report on the analysis of the

Hawkins-Davison material was concerned with such questions as

delineating the material correlates of social status, live-

lihood, and behavior. Her study represents the first fully

analyzed site report made at Frederica. The present author's

excavation and analysis of a site in the North Ward (Lot 12)

was primarily descriptive, although it did identify some

methodological questions of concern to archeologists working

at colonial sites and indicated hypotheses for future testing

at Frederica (1975). A popularized interpretive summary of

this project was also produced (Honerkamp 1977a).

In 1975 Kathleen Deagan reviewed and assessed the arche-

ological research carried out at Frederica in a report pre-

pared for the National Park Service. In this document, which

now functions as a master plan for archeology at Frederica,

Deagan has

1) outlined the results of all excavations carried out
from 1945 through 1975;

2) summarized the types of information obtained (and
ignored) through the archeological field work;

3) assessed the research potential of the extant
artifact collections and of the unexcavated sites;
4) identified problem areas and research emphases for
which Frederica is best suited.

Specific recommendations included sampling sites that did not

face the main streets of the town (especially back lot

features), delineation of dietary, occupational, and social

status elements from the archeological assemblages, and defi-

nition of the extent of military involvement in the domestic

sector of town life (Deagan 1975:25-26). The research prior-

ities defined in her report were explicitly incorporated in

the grant proposal for the 1978-1979 archeological excavation

of Lot 31 South (Fairbanks and Honerkamp 1978).

Project Background

Based on Deagan's recommendations, the Fairbanks and

Honerkamp proposal listed a number of rather ambitious

research objectives that were designed to meet the needs of

the Park Service as well as to address problems of particular

interest to the authors. These objectives were to be

achieved through the complete excavation of a single 90 by 60

feet (27.4 by 18.3 meters) colonial lot. Complete excavation

of the site was considered necessary in order to delineate

the site's boundaries, to identify back lot elements, and to

locate and define areas associated with craft and domestic

activities. In addition, total excavation would ensure the

retrieval of an unambiguous representative sample of a single

colonial occupation which could be used as a basis for

deriving hypotheses dealing with various aspects of 18th

century life at Frederica. Since the site was assumed to

have been occupied by known, historically documented

colonists, it was through the documentary records that

spatial, temporal, and social variables affecting the

archeological record could be accounted for, and it was

against this baseline of historical documentation that

hypotheses could be formulated and tested.

The original major research interest of the author was

to formulate hypotheses concerning specific socioeconomic

differences and similarities between the occupants of the

Dobree, Hird, and Hawkins-Davison sites which could be tested

through comparison of various classes of artifacts, including

faunal remains. A preliminary attempt along these lines was

made by the author using only faunal materials; documentary

evidence was used to control for temporal and social vari-

ables associated with the three sites (Honerkamp 1980).

However, archeological and documentary analysis carried out

after the Dobree Site field work revealed evidence neces-

sitating major revisions to the research goals. These are

discussed fully in subsequent chapters.


With the publication of Method and Theory in Historical

Archeology (South 1977a), historical archeology has reached a

watershed in the development of the methods upon which it is

based. Although the controversy associated with South's

"pattern recognition" approach as a viable research orienta-

tion is sometimes acrimonious, it has at the same time pro-

duced fertile ground for nurturing theoretical and method-

ological constructs, and it has been found to be essential

for stimulating the author's own consideration of an

important and often neglected question in archeology: why

are we interested in digging into the clutter of past

cultures? It is with the hope of somehow attempting to

answer this and other questions that the present study was


The purpose of this chapter is to clarify the theore-

tical biases inherited from the author's mentors, teachers,

and colleagues so that the reader can judge for him- or

herself the costs and benefits of such biases to the results

of the research. Following an overview of what is considered

to be the primary theoretical orientations in historical

archeology and the research implications inherent in them, my

own views on the strengths and weaknesses of each will be

presented. Finally, several broadly-stated problems derived

from the theoretical discussion will be identified as the

focus of the more specific research hypotheses presented in

Chapter VII.

Competing Paradigms in Historical Archeology

Although historical archeologists are fond of pointing

out the great strides that have been made in the theories,

methods, and techniques associated with their field, there is

nevertheless much disagreement among the practitioners of

this "science" concerning specific aspects of these advances.

Indeed, there is even disagreement on the categorization of

historical archeology as a "science" at all, with or without

the capital 's' (Binford 1972; Cleland and Fitting 1968;

Dollar 1968; Flannery 1973; Fontana 1968; South 1977a, 1977b;

Walker 1967, 1974). Historical archeology is still methodo-

logically and theoretically unstable, much as it has been

since its inception. However, in recent years there has been

a gradual emergence of two main paradigms that are usually

characterized as being diametrically opposed to one another.

Evidence that some methodological-theoretical battle lines

have in fact been drawn is seen in the tendency of many

archeologists to attach identifying (and slightly deni-

grating) labels to the orientations of their less enlightened

contemporaries in order to distinguish them from their own

sensible approaches. This is certainly not new in arche-

ology, but it seems to be more conspicuous now than it has

been in the past.

The first approach, discernible in most of Stanley

South's work, is referred to as the "pattern recognition"

orientation. It is based on the quantification and com-

parison of artifact types, groups, and classes among and

between historic sites to define inter- and intrasite

relationships of artifacts (i.e. "patterns") that are thought

to have temporal, functional and/or behavioral significance

(South 1977a, 1978). Implicitly South's methodology seems to

stem from an anthropological perspective, but the guiding

paradigm upon which it is based has never been clearly

stated. The other approach has been labelled "historical-

istic" by Schuyler (1978:1), or "particularistic" by South

(1977a:8). Whatever the name might be, it refers to a

philosophical approach that is usually associated with

methodologies that are quite distinct from that employed by

South and his adherents. South has also identified an

offshoot of the particularistic school that seems to be

proanthropology as well as interested in explanation: the

"world viewers" (South 1979). Since the present author views

South's orientation to be in part a reaction against the

initial particularistic approach, it is this earlier

theoretical orientation that will be examined first. As part

of my own small contri-bution to the archeological literature

devoted to the labelling of the theoretical approaches of

other archeologists, the two versions of the particularistic

school have been identified in this study as the "old" and

"new" particularism.

Old Particularism

Much of the early work in historical archeology centered

on elite-occupant sites that were usually impressively large

and ornate, historically famous or infamous (often as a

result of bellicose events occurring there), or possessing

some other "historically interesting" quality such as being

the oldest or only example of a style of architecture

(Fairbanks 1977). Not coincidentally, these types of sites

were attractive to the historically-oriented and -trained

researchers in the field, as well as to the agencies willing

to fund historic site research. The combination of a narrow

historicalist theoretical position among even the anthropo-

logically trained archeologists, along with the social and

economic context in which they worked, resulted in an

emphasis on a preservation-restoration approach which,

according to Schuyler (1978:1), "reduced archeology to a

supplemental technique in the service of architecture, narrow

specific historical questions, and the National Park

Service." Speaking as the herald for "pattern recognition,"

South characterizes particularistic archeology as emphasizing

...individualistic analysis and synthesis.
The paradigm (idea set) is idiographic (intensive
study of an individual case) and particularistic
(often characterized by an antinomothetic stance
against the search for general laws)...((particular-
istic archeologists)) are often implicitly scientific...
while at the same time disdaining the use of the
hypothetical-deductive method....The particularistic
approach has been accompanied by an antiscientific,
antianthropology phenomenon. (South 1977a:8)

For reasons discussed below, this orientation will be refer-

red to as "Old Particularism."

Inherent in this approach, as both South and Schuyler

point out, is an antipathy toward processual goals in

archeology and a rejection of anthropology as it is applied

to the study of historical archeology. The most prominent

proponents of this view have been Ivor Noel Hume (1969a) and

Ian Walker (1967, 1974), both British archeologists with

training in history and the humanities, and Clyde Dollar

(1968), a historian by training. Cleland and Fitting

(1968:124-126) and Schuyler (1978:201) believe that these

authors share a narrow view of what constitutes a science,

and by extension archeology, which certainly is not shared by

all historians or archeologists. Walker's assertion that

"...historical archeology by definition relies on historical

interpretation for explanation" (1974:168) is even less

likely to be supported by historical archeologists.

Anthropological questions, whether or not they dovetail with

the research concerns of historians, will continue to be of

central importance to many historic site researchers. Yet,

as Harrington has recently observed (1979:75-76), a great

many site reports--particularly contract reports--are devoid

of any hint of historical or anthropological interpretation.

Perhaps as a consequence of the early emphasis on site-

specific reconstruction and restoration, a number of archeo-

logists avoid asking any but the most facile questions about

their data, and as a consequence end up producing intensive

intrasite artifact-architecture studies which at best are

descriptive catalogs and at worst are excercises in

psuedoscience (South 1977a:326). Meanwhile, historians

conspicuously ignore the field of historical archeology

altogether (Harrington 1955; Schuyler 1977; Wilderson 1975),

despite the fact that the research concerns of some

historians and archeologists are becoming indistinguishable

(Bloch 1968; Carson 1978; White 1969).

At the risk of stating the obvious, systematic artifact

studies and architectural reconstructions have always been

and will continue to be important components of historical

archeology. Without the further refinement of our method-

ological tools that is afforded bv careful artifact studies,

not to mention the funding support found in reconstruction

oriented projects, advances in historic site research would

not be possible. While recognizing its achievements and

contributions, however, we should also realize that the

limited scope and aims of particularistic archeology preclude

it from achieving truly anthropological goals.

This approach has been labelled as "Old Particularism"

not only because it represents the initial theoretical

orientation embraced by early historical archeologists, but

also in order to differentiate it from the "world view"

orientation that is particularistic but stems from an anthro-

pological framework. This latter approach will be referred

to as the "New Particularism."

New Particularism

The "world view" orientation in historical archeology is

a relatively recent phenomenon, and if the papers presented

at the 1980 meetings of the Society for Historical Archeology

in Albuquerque, New Mexico, are any indication, it is a

popular one as well. The best known work in this area has

been that of James Deetz (1974, 1977) and Henry Glassie

(1968, 1975). Both of these researchers take an approach to

the study of material culture that emphasizes structural

analysis and mentalistic interpretation of data. In his book

In Small Things Forgotten Deetz is concerned with the

systematic relationships of various aspects of material

culture in the 17th through the 18th centuries in New England

(1977). He traces the evolution of the form and function of

ceramics, architecture, mortuary art, eating utensils,

butchering practices, and furniture, among other things, and

identifies major shifts in their use in England and the

American Colonies. He attributes the presence of the

cultural traditions and horizons that he defines to ideolo-

gical forces, particularly a shift from a "medieval" to a

"Georgian world view" that ultimately resulted from the

intellectual ferment of the Renaissance. This cognitive

model of material culture stems from Deetz's long-term

interest in mentalistic processes and their potential for

investigation by archeologists. For instance, the goal of

discovering "mental templates" of the people who made and

used artifacts was emphasized in his 1967 book, Invitation to

Archeology. Since that time Deetz apparently has recognized

the futility of this approach in prehistoric archeology and

has concentrated on adapting it to the field of historical

archeology, with its greatly expanded data base. In Small

Things Forgotten is a convincing demonstration of the

potential that the cognitive approach has for constructing

models of cultural change. It is explicitly anthropological

in outlook, it synthesizes and interprets written historical

and archeological data, and it incorporates sound

methodological principles from such sources as Willey and

Phillips (1958) and Binford (1962a) in the interpretation of

the data.

Similarly, Glassie has applied a structural approach to

the study of architecture in a temporally and spatially

circumscribed area in Virginia (1975). His imaginative and

exhaustive analysis of vernacular architectural forms was

designed to explain diachronic change and continuity in folk

housing as a function of the unconscious mental structures of

the people that built the houses. According to Glassie,

studying artifacts is the best way to discover how the minds

of people operated over long periods of time; conversely, he

thinks that this is an unnecessarily complicated and hence

futile way for discovering the principles governing human

behavior (1975:vii). It is precisely this attitude that

makes the world view approach descriptive rather than

explanatory in nature. Discovering how minds operated,

reconstructing mind sets, and elucidating aspects of a

Georgian world view does not explain the behavior resulting

from the mental structures, mind sets, and world views. This

concern with description and reconstruction is in my view the

principal element connecting the old and new particularism.

Whereas the old particularism was largely theoretical and

rejected anthropological questions as a basis for doing

historical archeology, the new version recognizes the

essential importance of theory in archeology, and it is to

anthropology that many of its practitioners turn for their

theoretical constructs. Both approaches attempt to

reconstruct something from the archeological and documentary

data. In the former, it is the site that is reconstructed,

while in the latter, it is the world view which created the

site that is sought. The underlying factors that might

account for the content, form, and function of the site (or

the mind set) are not proposed or investigated. It is this

failure to go beyond description that links both approaches

to a particularistic paradigm.

The most vocal critic of the new particularism has been

Stanley South. He is disturbed by the explanatory short-

comings that are associated with mind-set reconstructions and

by the practical problems that must be faced in recon-

structing world views from artifacts:

If cognitive models or world views are the goal
of the archeological research then a good way to
begin would be to study the modal personality of
living people and abstract their "world view" and
define their "mind set." Theoretically then, we
should be able to read in the different litera-
ture highly sophisticated models based on ethno-
graphic data for the Irish world view, the Black
world view, the Scottish mind set...on and on ad
infinitem. However, I see no mass of predictable
ethnographic or ethno-archeological literature on
such world views...To suggest that such models can
be abstracted from the archeological data base when
such has yet to be demonstrated from living human

populations, appears to me to be expecting a little
((too)) much from the material data base of archeology...
(South 1979:2)

South's criticisms seem to be relevant. If well trained

ethnographers, who are supposedly adept at recognizing emic

and etic distinctions, cannot discern the world views of

those they study, it would appear highly unlikely that

archeologists would be able to successfully glean mind sets

from studying the artifacts of extinct cultures.

South views the present popularity of the cognitive

approach with dismay since it distracts many archeologists

from the goal of understanding cultural process. However,

the appeal of the new particularism is not difficult to

understand. In addition to the attractive theoretical and

methodological characteristics that have already been

discussed, there are other factors involved. Both Deetz and

Glassie have presented eloquent and compelling reasons for

doing archeology that seem to have struck a responsive chord

in many of us. Echoing Ascher (1974), both strongly reject

the elite-only emphasis of the Old Particularism in favor of

what Charles Fairbanks has called the "archeology of the

proletariat" (1978). Not only does this shift from the elite

sites to the sites of the "common people" tremendously

increase the number of sites to be dug, it also provides

moral imperatives for digging them up. When Henry Glassie

tells us that the written record is "...superficial and

elitist--a tale of viciousness, a myth for the contemporary

power structure," it is clearly the duty of archeologists,

particularly cognitive archeologists, to provide a humanistic

counterpoint to the biased documentary data base (1977:29).

In a world of shrinking research dollars and increasing

skepticism toward the value of all social research, Glassie

and Deetz have given historical archeologists an almost holy

mission to accomplish and at the same time provided a

forceful argument against critics of the social sciences.

Small wonder that the New Particularism is attractive to so

many archeologists.

Pattern Recognition

South's approach, detailed in Method and Theory in

Historical Archeology (1977a) and operationalized in

different ways by a number of authors in Research Strategies

in Historical Archeology (South 1977c), is based on

quantitative analysis of artifacts, especially the derivation

of frequency variations of artifact types, groups, and

classes. Once this is accomplished, "patterns" can be

recognized for intrasite, intersite, and temporal contexts.

These patterns of artifact association (i.e., regular

frequency variations in artifact types-classes-groups) are

believed to have resulted from the patterned behavior of the

site occupants. The basic regularity of the replicated

patterns is then used as a reference by which variability in

the archeological record can be measured. Using this method,

South has defined a number of distinct patterns at British

colonial sites. These include the Brunswick Pattern of

Refuse Disposal, which measures the spatial distribution of

artifact classes and groups by comparing frequency

distributions around dwellings; the Carolina and Frontier

patterns, which are believed to monitor site function through

comparison of the frequency relationships between artifact

groups; and the Kitchen Pattern, which also monitors site

function as indicated through comparison of artifact classes

within the Kitchen artifact group (South 1977a, 1978). Using

artifact assemblages from several domestic, military, and

craft-oriented British colonial sites, the range of

variability that can be expected for the frequency

percentages of the artifact classes and groups are presented,

along with the hypothesized function of each site. By

establishing the normal variation that can be expected for

each type of site, South hopes eventually to be able to

determine the function of any British colonial site,

including those for which no documentary information is

available. The uniqueness of South's method lies in the

insistence on complete quantification of artifact assemblages

and a classification scheme that allows direct intersite

comparisons to be made. By contrast Deetz's methods seem

highly impressionistic (Schuyler 1977:113).

In the opinion of the present author, South's method is

believed to possess the potential to revolutionize the field

of historical archeology, similar to the way in which Ameri-

can prehistoric archeology was affected in the 1960s by Lewis

Binford. South works with historic site data in a singular

and original way. In so doing he gives primacy to the

archeological rather than the documentary record, and by

necessity this results in anthropological rather than narrow

historical interpretations. But, as is common with any

revolutionary paradigm, there are problems associated with

South's approach that must be resolved before it will gain

overwhelming acceptance. The most serious one is that South

has not provided a convincing explanation as to why the

patterns that are defined occur. This lack of an explanatory

component is probably why Schuyler refers to pattern

recognition as "a form of structural-functionalism"

(1980:200). Like Levi-Strauss, Glassie, and Deetz, South has

done an exemplary job of description without accounting for

the underlying causes of the phenomena described. Until this

drawback is overcome, the revolutionary potential of South's

methodology cannot be realized.

On a technical level, there are also questions

concerning the adequacy of the samples used by South in

delineating artifact patterns. Since most of his excavations

were oriented toward locating architectural features in order

to meet the requirements of the sponsoring agencies involved

in each project, almost all of his artifact samples were

recovered inside or directly adjacent to structural remains.

In view of the reliance placed by South on Michael B.

Schiffer's "behavioral archeology" concepts (Schiffer 1972,

1976, 1977), it is surprising that ruin-associated artifacts

are assumed to be representative of the entire site,

especially considering Fairbanks' (1977) and South's own

(1977a) observations concerning differential refuse disposal

practices at colonial sites. This assumption may be valid,

but this certainly needs to be demonstrated before it is

accepted. If it is not valid, then neither are the patterns.

Other less serious criticisms of South's pattern recognition

are mentioned in reviews of Method and Theory in Historical

Archeology (Chance 1977; Honerkamp 1977b).

Eclectic Approach

Another approach used by historical archeologists with

distressing tenacity is to avoid thinking about theory at all

and simply excavate sites as carefully as possible. The

adherents of this approach argue that commitment to specific

research strategies will invariably result in the collection

of limited types and amounts of data which have relevance

only to the specific research strategies employed. This

position, which is in itself a research strategy with its own

set of unspecified data limitations, is common in

anthropology and archeology and has been referred to by

Harris as "eclecticism" (1979:x,287-314). Archeologists

employing this supposedly "open-minded" approach are really

only excavation technicians, and their meticulous site

reports usually lack interpretation as well as usefulness to

other archeologists or historians. This in itself

illustrates the truthfulness of Glassie's observation that

"the scholar who believes that he works without theory, works

with bad theory" (1977:9).


Theoretical-Methodological Position

The manner in which archeological research is conducted

relates directly to the theoretical-methodological position

that the archeologist holds, including the archeologist who

rejects theory (Cleland and Fitting 1968; Glassie 1977;

Watson et al. 1971). Although the approaches outlined above

are by no means exhaustive of those held by historical

archeologists, they are thought to be representative of a

majority of the archeologists actively engaged in historic

site research. My own orientation is based on the foregoing

consideration of the strengths and weaknesses of each

approach, especially as they can be applied to the study of

the archeological record at Fort Frederica. This study

applies the structural-functionalist method to interpretation

of data within a positivist, materialist theoretical

framework. Neither an in-depth description of the material

remains recovered at Frederica, nor a historical-

archeological chronicle, nor an interpretation of the world

view of the site's occupants will be attempted, for reasons

discussed above. At the same time, the author is not unaware

of the needs of the sponsor concerning traditional site

interpretation goals. Accordingly, the material remains used

to derive archeological evidence of past activities and

behavior are described in more detail than is customary in

pattern recognition studies, and reconstructed or complete

artifacts are identified and illustrated. In this sense, the

present study is particularistic. No attempt will be made to


describe a nontestable mind set at Frederica because, even if

this could be accomplished, it would not explain the

archeological record. However, the level of analysis

reported in this paper should be sufficiently detailed to

allow a cognitive study to be made by some future New

Particularist. Although the pattern recognition approach is

seen to be similarly incomplete without reference to the

underlying causal factors that could account for the

patterns, it is felt to be a necessary methodological step in

attempting a coherent explanation of the British colonial

cultural system by allowing intersite comparisons of data to

be made.

Probably the best illustration of the potential that the

application of the pattern recognition method has in

historical archeology is found in Kenneth Lewis' study of the

frontier model in archeology (1976, 1977). Lewis

demonstrates the usefulness of South's method when it is

combined with processual archeology. He first outlines the

general characteristics of a frontier model on the basis of

past geographical studies. He then makes several predictive

statements (hypotheses) concerning the process of frontier

change and tests these against the data recovered from

eighteenth century Camden, a frontier community in South

Carolina. From this he is able to draw several conclusions

concerning the adaptations of colonizing societies to

frontier conditions. The utility of this approach for the

study of the frontier town of Frederica is obvious.

In summary, the theoretical position taken here is one

that derives in part from the pattern recognition approach of

Stanley South. Methodologically, the "structural-functional"

orientation with its reliance on artifact quantification will

be followed, but the shortcomings of this approach are

recognized. In relation to the study of archeological

evidence, a positivistic stance is taken which rejects the

ideational basis of the world view approach in favor of a

materialist explanation of culture (Harris 1977, 1979). The

systematic nature of the archeological record is assumed, as

is the ability of the archeologist to discern, through the

formulation and testing of appropriate research hypotheses,

the systematic cultural and natural processes that account

for the archeological record. Before this is attempted,

however, the data base must be defined. The data base at

Frederica is believed to reflect a British colonizing

adaptation to specific frontier conditions. The goal of the

present study will be to identify and explain components of

this frontier adaptation. The following chapters are devoted

to the definition of the historical-archeological data base.


Any study of the structure, function, and nature of the

frontier process must necessarily be regional in scope (Lewis

1976:157). The frontier system at Frederica will be investi-

gated through the use of documentary information pertaining

to the region and town during the 18th century, and through

combined documentary and archeological data from the three

sites excavated: the Hawkins-Davison Site (Cate 1956; Deagan

1972; Fairbanks 1956), the Hird Site (Honerkamp 1975, 1977a),

and the Dobree Site. As this last site is reported here for

the first time, it will be examined in more detail than

either of the other two.

Concepts and Definitions

Since the frontier phenomenon is identified as the focus

of investigation, it must necessarily be delimited. Using

the work of geographers such as Casagrande et al. (1964),

Kristof (1959), Weigert et al. (1957), and Hudson (1969),

Lewis defines the frontier as a region in which the dispersal

of settlement into a new territory takes place. It also

includes the zone separating settled and unsettled areas of a

territory which lies within effective control of the state

(Lewis 1977:145). The frontier appears with the first

permanent settlement and ceases to exist with the leveling

off of settlement growth and the stabilization of settlement

patterning. The frontier is directly dependent on the trade

and communications network that is established with the

already settled origin area. Frontier settlements

...reflect distribution of personnel and materials
in the most efficient way to permit the integra-
tion of activities in a sparsely settled area.
The limits of the exchange network at any given
time effectively mark the boundaries of the area
of colonization. (Lewis 1977:154)

The five conditions characterizing the frontier model are:

1) prolonged contact/continuity maintained between the
colonists and the parent society;

2) loss of complexity due to attenuation of networks
with the homeland;

3) settlement pattern more geographically dispersed
than the homeland unless temporarily impeded by
restrictive conditions;

4) dispersed frontier towns serving as the nucleus for
social/political/economic/religious activities and
as a terminus for the transportation network link-
ing the area of colonization with the homeland; and

5) temporal and spatial change, i.e., the pattern of
temporal growth is replicated spatially, with set-
tlements closest to the moving frontier represent-
ing the earliest stages of frontier development. As
the frontier expands, early settlements experience
marked changes in population density and settlement
pattern, and eventually become integrated at the
national level.

As with Lewis' study, two broad questions concerning the

cultural and temporal characteristics for the three sites at

Frederica will be investigated. Once these parameters are

established, it will be possible to test hypotheses

concerning the form and spatial extent of the occupations at

Frederica, the functions of the sites in question, the

distribution of archeological materials having behavioral

significance (especially with reference to South's recent

work), and resource utilization in an 18th century frontier

environment. Unlike Lewis' study, which was based on a large

stratified unaligned random sample from the entire town of

Camden, the sample from Frederica consists of three

intensively excavated sites. Since only a small proportion

of the total archeological variability at Frederica has been

sampled, inferences concerning the frontier process as an

intersite phenomenon must be made with caution. A brief

summary of the research questions addressed in this study is

given below. These questions will be presented in greater

detail in Chapter VII.

The cultural-ethnic affiliation of the sites can be

determined through reference to documentary and archeological

evidence. As will be seen in the next chapter, documentary

information indicates that the cultural affiliation of

Frederica as a whole, and the three sites in particular, was

primarily British. This assumption is tested in Chapter VII

with respect to the ceramic assemblages at each site.

Briefly stated, the economic ties between Frederica and

Britain, and by extension the cultural affiliation of the

town, should be reflected in the ceramic artifacts associated

with most sites in the town. The ceramic assemblages from

each site should therefore be characterized by (1) a

predominance of ceramics manufactured in Britain, (2) a

variety of British ceramic types that mirrors the diversity

of types found at other British sites, and (3) based on

findings from other British colonial sites, the occurence of

a predictable amount of re-exported foreign-made ceramic

types such as Oriental porcelain and Westerwald stoneware.

The temporal period of the town occupation is

established in Chapter IV as extending from 1736 to circa

1750 for most sites in the town. However, the chronological

position of the Hird, Hawkins-Davison, and Dobree sites

cannot be determined satisfactorily from the documentary

evidence alone. Temporally sensitive artifacts--ceramics and

white clay pipes--are used to estimate the occupation dates

for the sites. A lack of congruence between the mean

occupation dates derived from the documentary, ceramic, and

pipestem materials is used to interpret the site occupation

sequences as well as the behavioral characteristics of the


Site form and function are investigated with reference

to the frontier model. Settlement patterning at the town

level is expected to reflect specific demographic, political,

and economic factors that were present in the South Carolina

and Georgia frontier areas. This should be evinced by

distinctive settlement patterns for Camden and Frederica.

For instance, at Camden a combination of abundant land, low

population density, and the absence of a need for concerted

defense against a state-level power would result in a more

dispersed and uneven town structure than at Frederica. Both

documentary and archeological data from each town are used to

contrast the settlement patterns. Adaptive pressures

affecting the town structure should also be reflected in the

structure of the town's components, which in the case of

Frederica would be the freeholder lots. Some implications of

this hypothesis are (1) there should be evidence of cluster-

ing of features within the lots (depending on the intensity

of occupation at each site), (2) evidence indicating attempts

at maximizing the trash disposal function of certain areas of

the lot through reuse of subsurface features should be

present, (3) efficient use of subsurface features for the

disposal of faunal remains possessing objectionable odors

should be seen at the sites, and (4) there should be evidence

of an emphasis on demarcation of lot boundaries.

The question of site function is investigated by compar-

isons of domestic and nondomestic activity by-products at

each site. Quantification of the artifact assemblages in

terms of South's (1977a:92-102) type-ware-class-group

classification is carried out in order to define site

function as it is reflected by the empirical artifact

profiles associated with each site (see Chapter V). Fred-

erica, as a planned frontier community, should show archeo-

logical evidence of domestic activities as well as craft,

marketing, small-scale manufacturing, and tavern socializing

activities. Due to the emphasis placed on crafts and trades

by the organizers and sponsors of the colonization effort at

Frederica, nondomestic activities should be more in evidence

at Frederica than at Camden. Besides the comparison of

quantitative artifact profiles, it is also possible to use

the results of qualitative or presence-absence analysis to

determine site function. This is done for the artifact

assemblages from all three sites at Frederica, with

particular attention given to the by-products of a small-

scale manufacturing operation (a forge) that were recovered

from the Dobree Site.

The nearly complete excavation of the Dobree Site has

resulted in a data base that is well suited for testing

hypotheses concerning British colonial refuse disposal

behavior. South has defined a pattern of artifact

distributions at British colonial sites which he has

designated as the Brunswick Pattern of Refuse Disposal. This

pattern monitors secondary and primary deposition of refuse

as it occurs in and around colonial structures (South

1977a:48). However, portions of the sites used by South to

derive the Brunswick Pattern constitute a highly limited

sample in that they are all oriented around structural

foundations and their directly adjacent areas. Virtually no

testing of peripheral site areas is evident in South's work

(1977a:50-76). At the Dobree Site graphic representation of

the frequency distributions of several artifact classes and

groups is used to test the applicability of the Brunswick

Pattern at Frederica.

Finally, a subsistence model based on traditional

English foodways is defined in Chapter VII and compared with

the zooarcheological evidence from Frederica. A traditional

pattern of meat consumption in early Georgia that mirrors the

European pattern is expected to contrast sharply with a

pattern of resource utilization that reflects adaptive

responses to frontier conditions.

Of central importance in testing the hypotheses

mentioned above is the definition of the temporal and spatial

aspects of the particular sites investigated, as well as the

general historical context of the town and region. This is

undertaken in the following chapter through the combined use

of documentary and archeological resources.


In this chapter the historical context of Frederica as a

frontier settlement, and of the Hawkins-Davison, Hird, and

Dobree sites as components of such a settlement, will be

developed. The concern here will be with clarifying some of

the characteristics of the frontier adaptation in the region

and relating them to Frederica and the three sites

investigated. On a more specific level, an attempt will be

made to establish temporal and spatial control for the town

and sites. For regional and local background information of

a more general nature, any of the numerous publications

dealing with Georgia and Frederica history and archeology

should be consulted (Cate 1943, 1956; Coleman 1976; Davis

1976; Fairbanks 1956; Honerkamp 1975, 1977a; Ivers 1974;

Manucy 1962; Reese 1963, 1969; Saye 1943).

Historical Context: Georgia

Georgia was established in 1733 as the last British

proprietary colony in America and as the last to be founded

by settlers coming directly from Europe. By that time South

Carolina had developed sufficiently in economic importance

through its plantation exports and British-Indian trade that

expansion west and south from the settled coastal areas was

officially recognized as necessary for the consolidation and


security of the planting and trading enterprises (Brown

1963:2; Coleman 1972:169-170). In 1730 the royal Govenor of

South Carolina submitted a detailed plan for expanding the

Carolina frontier which was enthusiastically endorsed by the

Board of Trade. The Board added the stipulation that two

settlements be located on the Altamaha River on land that had

long been claimed by both Britain and Spain. By combining an

economic-military expansionist policy with a philanthropic

movement in England aimed at making productive colonists of

the mother country's many poor and insolvent subjects, the

founding of Georgia seemingly accomplished three objectives

at once. First, it protected the economically valuable

Carolina colony by providing a military buffer against

Spanish or French incursions and by reducing the threat of

slave rebellions on the Carolina plantations. Second, it

expanded the frontier trade and plantation networks that were

so essential to the mercantilistic economy envisioned for the

colonies by the Board of Trade. Third, it helped relieve the

mother country of a substantial domestic burden, the

unemployed and poor of London and other cities (Reese 1963:8-

9; Coleman 1976:9-13). Thus it was the "unfortunate poor,"

many of whom were supported by the Trust charity, who

accompanied James Oglethorpe to the new colony in 1733.

The frontier town of Savannah, located near the mouth of

the Savannah River, became the nexus of social, political,

and economic life in early Georgia. The settlement was

positioned on a major river linking the piedmont with the

coast. Within a few years after its founding, a number of

small military and domestic settlements extending along the

Savannah River and southern coast were established. The main

military settlement was Fort Frederica while the most

important trading town was Augusta. Savannah soon became the

center of a land and water transportation network that

developed into the hinterland, although until the 1760s

Charleston remained the principal entrepot from which English

merchandise was transshipped to Georgia (Davis 1976:52-54).

Initially the system of land distribution in Georgia was

explicitly designed to ensure a dispersed settlement pattern.

In the interests of military security, a man-land ratio of

one male to every fifty acres was established for land

grants. "Gentlemen" were allowed up to 500 acres on the

condition that they bring 10 male servants to the colony at

their own expense. Once land was granted, its inheritance

and sale were severely restricted. Large individual

landholdings were prohibited under this arrangement in the

belief that the colony would instead be populated by numerous

small-plot farmers who would make up a strong standing

militia; slaves were prohibited for the same reasons. As

Reese (1963), Coleman (1976), and others have emphasized,

this slave-free agrarian system of land division discouraged

the establishment of productive plantations with their high

capital and labor investments. The flexibility associated

with the private, large scale, economically oriented

companies that had successfully established colonial

settlements elsewhere in America was precluded under

Georgia's philanthropic charter. Until the land and labor

restrictions were removed by the royal administration in

1750, Georgia's frontier settlements remained unstable in

terms of population, economy, and settlement patterning.

Once a reliable economy based on rice, indigo, and marine

stores became established, frontier towns such as Savannah

began to stabilize and take on increasingly complex

communication, transportation, and commercial functions. By

1773, 25 ocean-going vessels were owned by Georgians (Wright

1873:175) and many were offloading at Savannah, which had

largely replaced Charleston as the major sea port linking the

colony with Britain.

From the documentary information reviewed thus far, it

can be seen that the conditions proposed by Lewis as being

necessary for the development of a frontier are fully

satisfied (1977:160-164). Briefly stated, these are:

1) Georgia, located on the periphery of a prev-
iously settled area, was physically occupied
by an intrusive colonizing society (Great Britain).

2) The intrusive European society possessed a highly
developed state level of organization. The
presence of a concomitant legitimizing force
necessary to maintain the logistical support of the
colonization effort is also apparent. Frederica,
as a fortified frontier settlement and as the staging
point for a major military campaign against Spain in
1740, exemplifies this state level legitimization of
force (Ivers 1974).

3) In the second quarter of the 18th century, external
sociocultural barriers to expansion consisted only
of a competing state level society in St. Augustine,
Florida. Decimated by internecine warfare, slave
raids, and European diseases, the coastal Guale
Indians had been removed to Spanish missions in

Florida by 1686. By the 1730s only sporadic con-
tact with remnant inland Creek groups occurred in
the Frederica area (Gannon 1965; Ivers 1974).
These groups posed little if any competition for

4) Georgia was amenable to subsistence and commercial
exploitation. This was conclusively demonstrated
with the removal of land and labor restrictions by
the royal administration. Natural barriers pre-
venting access to different parts of the frontier
were absent, as indicated by the maintenance of
trade and communication routes during the politi-
cally and economically unstable proprietary period.

Historical Context: Frederica

Shortly after Savannah was established, a string of

military outposts was built along the coast as far south as

Fort George Island, Florida. The center of this defensive

network was the fortified settlement of Frederica. Founded

in 1736, the town originally was occupied by 40 civilian

families, but by 1738 a military regiment was permanently

stationed there. Eighty-four house lots were laid out within

the walls of the town and a corresponding number of 50-acre

farming plots were located in the surrounding countryside;

every freeholder received one of each. The town was divided

into a North and South Ward by the main east-west street

(Broad Street) extending from the town entrance to the fort

(Figure 1-3 and Figure 1-4). Besides the 600-man Regiment,

Frederica's inhabitants included various craftsmen and

skilled workers, but as was true of the rest of Georgia's

population at the time, farmers, husbandmen, and laborers

were in short supply (Coleman 1976:20-22; Coulter and Saye


The military orientation of the town is clearly

discernible in the settlement structure. Unlike Camden,

which was an unplanned frontier town founded for and

supported by economic considerations, Frederica was designed

as a compact, defensible settlement containing a high density

population capable of bearing arms. The tightly regulated

settlement pattern, based on military necessity, accounts for

the circumscribed layout of the town as compared to Camden.

Frederica's row pattern of lots more closely resembles the

contiguous arrangement of structures in an English market

town than the uneven dispersal of structures at Camden (Lewis


The town's main economic activities are of importance in

establishing some of the major characteristics of the

frontier adaptation. This topic has been of interest to

historians as well as archeologists at Frederica. Although

the settlers were expected to become self-sufficient through

subsistence farming and by practicing their crafts and

trades, their initial material needs were supplied by the

Trust. Upon arriving at Frederica the head of each household

was supplied with a year's provision of food, clothing,

tools, and kitchenware. By 1738, when the charity period was

to end, most of the settlers were unable to support

themselves and had to depend on weekly "advances" of food

(Manucy 1962:54,100). This continued for a number of years,

for only sporadic success at farming was ever achieved by

Frederica's inhabitants and craft activities were apparently

not frequently practiced (see discussion below).

The Trustees gave specific instructions concerning the

construction of dwellings, which were to be built in the

Georgian style and to measure at least 16 by 20 feet (Candler

1904-37:XXXIV,288). At Frederica the documentary and

archeological evidence gives a fairly clear picture of the

types of houses built by the colonists. A contemporary

visitor to the town mentioned "...some houses built entirely

of Brick, some of Brick and Wood, some few of Tappy-Work, but

most of the meaner sort of Wood only" (Jones 1878:122).

Archeological excavations support this description. Most of

the inhabitants built "common freeholder" houses which Manucy

describes as timber framed clapboard huts (1960:20).

Architectural variability as a function of socioeconomic

status seems to be evident at Frederica. Timber frame

structures were the most economical and easiest to build, and

by far made up the majority of residences. Tabby and brick

houses, which were more secure but entailed much higher costs

in labor and materials than their wooden counterparts, were

selected by the wealthier colonists (Manucy 1960:20-23).

There is also some indication of a correlation between lot

location and type of house, with the brick and tabby examples

more commonly located near the fort or facing Broad Street.

Archeological evidence of wooden houses has yet to be

recorded and remains a high research priority.

Most authors emphasize the overall importance of the

Regiment in the town economy. The civilian sector quickly

became oriented toward and dependent on the military payroll,

either by providing goods or services through the full time

practice of crafts and trades, or by engaging in the tavern

trade and practicing crafts on a part time basis, if at all.

This "artificial prosperity," as Davis calls it (1976:61), is

indicated by contemporary descriptions of the civilians

engaged in "selling to the Camp" and "keeping public Houses"

(Reese 1973:8,74). Except for early attempts by a small

minority of the town's more industrious residents, farming

was never an important activity on either a subsistence or

cash crop level. A 1745 letter from John Terry to the

Trustees of the colony is informative concerning the role of

farming and other occupations in the community at that time.

In his letter he complains about the

...badness of the Land in Regard to the Expces.
attending the Clearing of it So, that the few &
very few inhabitants that are here do Not go on
planting And Neither Are they able to go Upon
Such an Expensive Undertakins, All they Do is to
build houses on their town Lott, wch. turns to a
very good Accot. to them By Letting or Selling of
them to the officers, And hiring of them for Stores
to Merchts. Believe me My Lords & Gentlemen, these
Are the Improuvemts. Made or makg.
(Candler 1915:401-402)

This letter echoes the sentiments expressed five years

earlier by George Whitfield when he remarked that "Frederica

is wholly kept up by the Soldiery & that too I fear by their

Intemperance. Very few as I could hear of intended planting

any Corn..." (Phillips 1947:104). Clearly the economic

endeavors of Frederica's inhabitants were limited to a narrow

range of activities that focused on servicing and supplying

the regiment. Thus a loss of complexity resulting not only

from attenuation of networks with the homeland but also from

a highly restricted economy is indicated by the documentary


General Temporal-Demographic Parameters

Through numerous documentary references a terminus post

quem of February 1736 has been established for the colonial

occupation at Frederica. Determination of an end date for

the settlement is more difficult due to the erratic

fluctuations in population that characterized Frederica's

history. Most authors emphasize a severe reduction in the

domestic population as a result of and immediately following

the departure of the Regiment in 1749 (Cate 1956:204; Jones

1878:124-125; Reese 1969:69-71). Indirect evidence of the

town's dissolution is seen by Reese in the closing of the

town storehouse in 1751 due to a lack of customers (1969:70).

In 1758 a fire "...wasted nearly all of the town which time

had spared," presumably because no one was around to put it

out (Stevens 1847:446). Contemporary accounts, including one

by Royal Governor John Reynolds in 1755, give the impression

of a town "in Ruins," and sporadic attempts to revive

Frederica for military or economic purposes during the last

half of the 18th century were unsuccessful (Jones 1878:126-

136). Small military detachments were stationed there

through the American Revolution and half-hearted efforts at

repairing the military structures were occasionally made.

William Bartram, viewing Frederica in 1774, mentioned that

what had once been "a very considerable Town" was "now almost

in ruins" (Bartram 1943:145), while in 1839 Fanny Kemble

portrayed the town as a collection of rose-covered ruins with

only two standing houses (1961:131). Small numbers of

families continued to live and farm in the Frederica area

into the middle of the twentieth century (Cate 1926:37,

1956:204; Lovell 1932:272-273).

The foregoing evidence indicates that the bulk of

Frederica's population was absent by 1750, scarcely 15 years

after its establishment. However, a review of other

documents suggests that the occupation span of many of the

nonmilitary residents was considerably less than 15 years.

The list of Georgia settlers compiled by the Earl of Egmont,

first President of the Trustees for Establishing the Colony

of Georgia in America, reveals that, of the 63 colonial

households or individuals who were assigned lots at

Frederica, 24 were either "Dead, Quitted, or Run Away" after

five years in the colony; for many, only two years of

frontier life was enough (Coulter and Saye 1949). This

figure is derived from an extremely limited sample, i.e.,

only those settlers for whom Egmont had some information, and

would undoubtedly be higher if a systematic study were made

of all of Frederica's documented inhabitants. Underscoring

the findings from Egmont's List are the frequent complaints

and sometimes dismal descriptions of frontier life that were

voiced by many of Georgia's inhabitants (see Candler 1904-

37:I,256-345, XXII,17, XX,183-184; Reese 1973:47,106,294-


The present author feels there is a need for increased

internal criticism of secondary documentary references

dealing with this and other aspects of Frederica's past.

Frederica is commonly characterized as a thriving community

of artisans and craftsmen that prospered by serving the

Regiment until 1749 (Cate 1956:204; Coleman 1976:141; Reese

1969:47). The town population supposedly reached its peak of

over 1000 soldiers and civilians in the early 1740s, a figure

that should be considered controversial in view of highly

contradictory contemporary estimates (Candler 1904-

37:VIII,488, XXIV,140-141; Reese 1973:111). Adverse

descriptions of Frederica's conditions cannot simply be

dismissed as the machinations of "Clamorous Malcontents." At

the same time that Frederica was said to be in its economic

heyday, the rest of Georgia was apparently experiencing hard

times, as summarized by Coleman:

From 1737 through 1747 there was decline in the
colony. This period saw the closing of the
Trustees store ((in Savannah)), the objections
of the malcontents, and the Spanish War. Immi-
gration slowed, people left, and population
actually declined. Charitable contributions to
the Trustees declined to almost nothing, and
the colony was kept going only by Parlimentary
grants totaling some 88,000 during the decade.
The year 1740 was undoubtedly the low point
economically in the colony's history. It was
during this decade that the inadequacy of many
of the charity colonists for a frontier area
became clear. (Coleman 1976:142)

What seems to be generally ignored or underestimated in

historical studies of Frederica is that the settlement was

characterized by an extremely high turnover rate for the

inhabitants. The emphasis that some historians have given to

the number and diversity of crafts and skills represented at

Frederica is the result of a synchronic interpretation of a

diachronic phenomenon of population replacement. In other

words, the historical interpretation was derived from a

listing of all craft specialists over a span of time, rather

than from those present at any one time. Reese's contention

that a list of the names of Frederica's settlers "could be

extended to include practically every trade that a pioneer,

frontier community needed" may be correct (1969:47), but the

accompanying assumption that all these essential craftsmen

were living in Frederica at the same time cannot be supported

by the documentary evidence.

The effects that this high rate of population

replacement might have on the archeological record vary.

Most sites at Frederica can be assumed to have been abandoned

prior to 1750. Some site assemblages can be expected to

reflect single residence occupations of short duration or

multicomponent occupations of various durations. These

should contrast in discernible ways with a small number of

sites with long duration (eight to ten years or more) single

components. Intrusive 19th and 20th century components

should also be present at a small number of the colonial


Spatial Parameters

Rarely has the utility of integrated archeological and

documentary research been better demonstrated than at

Frederica. Fairbanks' excavation of the Hawkins-Davison

house was based on documentary information provided by the

late Margaret Davis Cate, a coastal Georgia historian who

devoted much of her life to the historical research of

Frederica. The identification and discovery of the Hawkins-

Davison common wall on the lot line between Lot 1 and Lot 2

of the South Ward allowed the archeologists to relate the

colonial lot layout to Frederica's present topography (Cate

1956; Fairbanks 1956).

Of considerable interest to many researchers at

Frederica has been the spatial configuration of the town.

Beginning with Cate's and Fairbanks' early research, every

excavation undertaken or sponsored by the National Park

Service has incorporated into its research design some

emphasis on the spatial delineation of the town lot locations

and boundaries. Lot sizes have been reliably established

through the use of information obtained from 18th century

documents and maps (Auspourger 1736; Miller 1796; Moore

1840:82; Stacy 1784) combined with archeological verification

(Fairbanks 1956:225). Estimates of the size of the various

colonial streets and alleys, which are particularly important

for determining the locations of lots that do not face Broad

Street, are more problematic. The 1796 Miller map (Figure 4-

1) shows specific street-alley widths that contrast with











~" I *

aS f




'\ -

4 I

-6 "


I t. 0- 4M

.-re's contempor-ry account and the dimensions listed -n t-e

1736 Auspourger mrap. Fairb:nks fou-d no direct evidence of

'r-ad Street during his excavations and attempts by Shiner

(1958a) an4 Moore (1958) to determine the width of Cross

Street were also unsucc-ssful. Using the archeolo<{ical

-:vidence of foundations uncovered at the Hawkins-Davison and

Lot 1 North sites, Fairbanks concluded that Broad Street was

75 feet wile (22.9 mete-s), .s shown in t'-e Ausoourzer man.

However, the alley to the south of the Hawki-s-Davison site

was found to be slightly mor- than 14 feet wide ('.4 metes)

w'.I -h is in accordance with t'-e Mi ler mao. A su=marv of the

i.-formation on stee and lot dimensi -ns shows some

concordance, as indicated nda in Table 4-1.

Table 4-1.

Summary of Street

and Lot Di.ensions, Frederica.

'usoourgr (e 73r

road 75 feet

H ~ 1 ^' 0 0 feetC
To.. o6eX Oeel

S> t r qQ e teet



S o'nd U" t

%!/1i n-e oee' or-

I'J ^(i f p ^ ^ ~ ^1 ^^p -4

,, 4

Miller '1796

3 feet

?2 feet

1 4 17 f

*0 X 90 ? pt

q1 f~ Y~ !*? y~t *,'

vDr pOr

Fai banks'
Excavation (1Q56

75 feet

I14 feet-

70 t Q ,

rpp i r> ? 1

L' ) A

$ 0 (-

4- 4-

I~\ ;n
1 '3

~r 1r~r<~ ~r AT -, ~N i~ r 4- V

-1 M; e nu m(=> P V t 3 i'

other or with the field calculations of street dimensions,

locational information on lot and street placement remains a

high priority for any excavation at Frederica. Lots not

adjacent to Broad Street are of particular concern since

possible errors made in measuring their locations may be

compounded with increasing distance from the Hawkins-Davison

base line. Variations induced by the surveying techniques

used in 1736 could also contribute to inaccurate determin-

ations. Despite these drawbacks it is assumed that the

approximate location of any lot can be found using the

documentary-archeological evidence outlined above.

The Hawkins-Davison Site

The documentary record pertaining to this site has been

presented fully by Cate (1956), Deagan (1972), and Fairbanks

(1956). A brief synopsis of their work will be given in

order to indicate some of the temporal, spatial, and social

parameters associated with the site. Thomas Hawkins has been

positively identified as the owner of Lot 1 South. He, his

wife, and two servants arrived at Frederica in 1736 and

apparently left the settlement in 1743, although an ambiguous

reference given by John Terry in 1745 suggests a later date

of departure (Candler 1904-37:1,462, XXIV,402). Despite

coming to Georgia "at the Trustees' expense," Hawkins seems

to have held a relatively high socio-economic status among

the town's inhabitants. He was employed as the Regiment

surgeon, town apothecary-doctor, and Bailiff for Frederica.

He lived in one of the most impressive residences in the

town, a three story home built of bricks. Hawkins owned a

large "plantation" on a nearby island and possessed a number

of cattle (Candler 1904-37:V,500). A petition for his

property in 1767 and 1768 suggests that his house was not

reoccupied after his departure but this is unlikely given its

expensive construction.

Like Thomas Hawkins, Samuel Davison came to Frederica in

1736 on the charity of the Trust. The family, consisting of

Davison, his wife and their three children resided at Lot 2

South in an expensive brick and tabby house which shared a

wall (on the lot line) with the Hawkins house. Trained as a

chairman, Davison held positions as Constable, Searcher of

Ships, and Overseer of the Trustee's Servants, although his

main livelihood was in running a tavern in his home. He made

a vigorous attempt at farming, at least during his first few

years in the colony. By virtue of his expensive housing and

contemporary comments on his industrious nature, Davison can

also be assumed to have held a high socioeconomic status

relative to other civilians in the town. In a 1741 letter he

was listed as having "20 head of Cattle, servants, 2 or 3

carts, 8 horses..." (Candler 1904-37:V,501), but the family

left for South Carolina in that year. Documentary evidence

of a later occupation on his property has not been found, but

it is unlikely that one of the best houses in Frederica would

have remained unused for long in the 1740s.

The Hird Site

Documentary background on the Hird Site has been

presented by Honerkamp (1975, 1977a). Thomas Hird, his wife,

and three children came to Georgia in 1736 at Trust expense.

According to the list compiled by the Earl of Egmont, Hird

was a dyer who resided at Lot 12 North. As shown by the

Miller Map, this would indicate a lot position in the

northeast section of the town (see Figure 4-I). According to

the documents, Hird may have engaged in dyeing, brewing, and

livestock raising, but there is definite evidence that he

served as town Constable, lay preacher, and farmer exemplar.

He also made frequent business trips to Savannah and

Charleston. His socioeconomic position in the town,

especially with respect to Hawkins and Davison, is difficult

to determine. Although Honerkamp has interpreted the

documentary-archeological information as indicating a

slightly higher socioeconomic level for Hird (1980), this

interpretation is open to question. Contemporary accounts

concerning Hird's industrious nature and material

improvements, when combined with evidence of his considerable

occupational endeavors, support the assumption of a

relatively high status position in the town. His long term

occupation at Frederica (12 years) can also be interpreted as

supporting this contention, especially in view of the high

turnover of other inhabitants who were unable to make a

living in the town. Hird occupied his lot until his death in

1748, and the rest of the family departed soon after that


The Dobree Site

Symptomatic of the difficulties encountered in using a

direct interpretation level of research is the documentary

background on the Dobree Site. Using the lot numbers shown

on the Miller Map, (Figure 4-1) and the information on lot

ownership available in Coulter and Saye (1949:40), this site

was originally thought to have been occupied by Robert

Patterson, a accountant in the town store. However, after

excavation of the site was nearly half finished, another

reference found in Egmont's List indicated that Lot 21 South

was occupied by Elisha Dobree, an accountant who worked as a

clerk in the town store (Coulter and Saye 1949:71).

Underscoring the confusion, a review of Margaret Davis Cate's

research notes revealed contradictory listings for the lot

number. In some cases Cate originally had "31-S" listed for

Patterson's lot, but she later crossed this number out and

replaced it with "21-S" on several reference cards (Cate

1961). Using Cate's notes, Reese later attributed a

description made by Dobree of his own house and lot to

Patterson (1969:back cover). Finally, the Assistant

Superintendent at Fort Frederica, George Berndt, informed the

archeologists that the lot numbering for the rows containing

Lots 21 through 31 in both Wards was reversed on the Miller

Map (Figure 4-1). He cited as evidence the out-of-synch

numbering of these two rows, and more conclusively, a grant

made in 1763 to Pickering Robinson in which town lot 25 in

Frederica is described as being on a corner lot on the north

side of Broad Street (Bryant 1972:30). Lot 25 in either Ward

can only be on a corner if the lot numbers are reversed

according to Berndt's interpretation. Since the Miller Map

was drawn nearly one half century after the town was

abandoned, a numbering error would not be surprising. The

easternmost lot on the third row of the south ward is

therefore considered to be Lot 31.

Unfortunately the new documentary evidence concerning

the lot designation has not cleared up the question of the

lot ownership and occupancy. Since both Patterson and Dobree

could not have owned the same lot at the same time, one of

Egmont's entries must be in error. Dobree is known to have

been in Frederica from at least 1736 to 1738 and to have sold

his lot to David Provost sometime prior to 1743 (Candler

1904-37:I,424, XXI,283). In that year John Provost, David's

heir, assigned a town lot in Frederica to Captain George

Dunbar of Oglethorpe's Regiment. The lot was identified as

Dobree's former property, but the lot number was not

mentioned (Candler 1904-37:V,705). Dunbar died in Jamaica in

1763 while still a Captain in the Regiment; the extent of his

involvement on his Frederica lot is unknown. In 1759 William

Mackintosh requested a grant for Lot 21 South in Frederica

(Candler 1904-37:VIII,143). Presumably, if Dunbar was still

in possession of Dobree's lot at this date, Mackintosh would

not have requested it, so he must have been trying to obtain

the "real" Lot 21 belonging to Robert Patterson. This

contention is supported by a similar request made in 1759 by

Thomas Goldsmith for Patterson's Frederica lot, which was

specified by name and not by number (Candler 1904-37:VII,32-

33). By 1763 Goldsmith owned this lot and Patterson was said

to be deceased (1904-37:IX,53). Patterson, who arrived with

a wife and three sons in 1736, is known to have been in

Frederica as late as 1741. By this time his entire family

had died and he was supposedly running a "bawdy house" (Cate

1961; Coulter and Saye 1949:40,100). An oblique reference to

Patterson made by John Terry in 1745 does not definitely

indicate the presence of Patterson at Frederica (Candler


On the basis of the admittedly incomplete documentary

information reviewed above, it is suggested that the Lot 31

South site belonged to the Dobree-Provost-Dunbar series of

owners from 1736 to sometime prior to 1763. It is believed

that the lot designation by the Earl of Egmont for the Dobree

Lot was an error resulting from the inadvertent replacement

of a "3" with a "2" or from incorrect information recorded by

Egmont. Dobree is the only known occupant of the site--

Provost or Dunbar may have resided there, but this cannot be

definitely established from the documents.

Elisha Dobree was a controversial figure in early

Georgia. He came to Savannah in 1734, where he was assigned

a lot in the town (Coulter and Saye 1949:71). His arrival in

Georgia was immediately precipitated by his flight from South

Carolina in order to escape his creditors (Candler 1904-

37:XX,72). Although trained as a merchant-bookkeeper, he

wrote a number of letters to the Trustees describing his

efforts at clearing land, farming, and preparing naval stores

for export to the mother country. Notable among his many

letters written in 1735 is one in which he mentioned the

possibility of bringing slaves to Georgia (Candler 1904-

37:XXI,612-613). This is one of the earliest references of

dissatisfaction with the Trustees policies among the

"Malcontents" (Wood 1979). In reply he received a strongly

worded letter from the Trustees which emphasized the reasons

for the laws prohibiting slavery and the consequences of

disobedience to local authority. This letter also

encouraged Dobree's farming attempts and almost in the same

sentence reprimanded him for "...hiring so many Lots ((in

Savannah)). Because it destroys poor Men, unites Lots, and

drives away Inhabitants..." (Candler 1904-37:XXIV,133). Two

months later he was discharged from his job as clerk in the

town store in Savannah and he wrote a despairing letter

concerning his "greatly reduced circumstances" (Candler 1904-

37:XXIV,377-380; McPherson 1962:97). In 1736 he gave a

description of his house and lot at Frederica where he was

again a clerk in the town store: "...a small house with a

brick chimney, built on his town lot which is fenced and has

palisades and clapboards, well dunged and now every way fit

for propogation of fine plants..." (Candler 1904-37:XXI,345).

Apparently Dobree possessed one of the small clapboard huts

so common to the nonaffluent segment of the colonial

population at Frederica (Manucy 1960:20). A single servant

may have resided with him in his new home (there are at least

five servants listed for Dobree by Egmont). His wife,

however, refused his repeated requests that she join him,

saying that he was a "whimsical man, and not able to

maintain" her and their three children (Egmont 1923:377).

Dobree started complaining about the harshness of his new

home almost immediately and probably left Frederica sometime

after 1738 when the last reference to his employment at the

store was made (Candler 1904-37:V,70).

Although not extensive, the documentary records contain

evidence that Dobree had a difficult time as a colonist. His

short, inauspicious stays in Carolina, Savannah, and

Frederica indirectly indicate a lack of success in exploiting

the natural and social environments in each location.

Certainly his wife's opinion of his abilities as a family

provider and his own admission of economic hardship are not

inconsistent with an inference of low socio-economic status

at Frederica, at least in comparison to the occupants of the

Hawkins-Davison and Hird sites. Questions relating to the

identification of the Dobree occupation and how it contrasts

with other colonial occupations are developed in Chapter VII.


This chapter reviews the methods and results of

excavation and analysis for the three sites investigated. An

attempt is made to describe the research frame used in this

study in terms of (1) the ways in which the research frame

was investigated in the field and during analysis, and

(2) the archeological data obtained from application of the

excavation and analysis methods and techniques. A

distinction is maintained throughout this chapter in the

discussion of faunal and nonfaunal methods and materials.

This contrast arises from the specialized analytical

procedures used in zooarcheological studies and the nature of

the fauanal material itself, which is recognized as

representing specialized by-products of human behavior which

is distinct from the behaviors accounting for the presence of

ceramic, glass, and metal artifacts in the archeological

record (South 1977a:97; Wing and Brown 1979:1-10).

Excavation and analysis procedures employed at the Hird,

Hawkins-Davison, and Dobree sites were fairly consistent

despite the extended intervening periods between projects.

One reason for the continuity achieved in the excavations is

the close professional association among the researchers

involved in the archeology of Frederica. Also contributing


to the possibility of making valid intersite comparisons is

the quality of the field work performed by Fairbanks in 1952,

which resulted in the systematic recovery of both architec-

tural and nonarchitectural data at the Hawkins-Davison site.

His approach was in many respects ahead of its time in

comparison to contemporary historic site projects, and his

work served as a standard that was rarely equalled in later

excavations at Frederica.

Excavation Procedures

The procedures used in excavating the three sites

reported here reflect the specific research goals of the

archeologists. The orientation at the Hawkins-Davison Site

was toward discovering and excavating architectural features

that could be used as a locational datum and as evidence of

British colonial housing. In addition, anthropological

questions were investigated. In order to determine the

extent to which differences in the crafts performed by the

residents of the site would be reflected in the archeological

assemblage, Fairbanks sifted all excavated dirt using one

half inch screen, retained all artifacts found, and recorded

provenience information while excavating. Ten foot square

units were excavated in and around the house; lot lines were

determined by trenching. An area of 209 square meters (2249

square feet) was uncovered at the site, including the west

trench shown in Figure 5-1. Horizontal and vertical control

was attained through reference to a permanent datum station

that was established at the site.





< <



I I- .
0 3: -0


wi- -

0 jr




At the Hird Site the research focus was on the

definition of activity loci in the back lot or "toft" area of

the site, and on the recovery of artifacts and features

indicative of craft activities, food storage and preparation

practices, and trash disposal behavior. Rather than attempt

to locate and expose foundations, the back lot area was

sampled through intensive excavation of eight ten by five

foot units and eight ten by ten units (Figure 5-2). A total

area of 111.5 square meters (1200 square feet) was excavated.

The field strategy at the Dobree Site was to excavate

the entire 90 by 60 foot lot using three by three meter

squares. Complete excavation of the hypothesized lot was

expected to reveal archeological correlates of street and lot

edges, house construction, craft activity loci, and lot

structure and function. The large, detailed body of

archeological data generated from this approach was expected

to be useful for testing hypotheses concerning the patterning

of artifact associations on intrasite, local intersite, and

regional intersite levels. Sixty-three excavation units were

dug at the site, although not all of them were within the lot

lines. Due to the presence of numerous features on the north

end of the site and to the lack of a clearly defined fence

line marking the actual lot boundary, several excavation

units were placed to the north and west of the lot, as

indicated in Figure 5-3. The actual area uncovered during

the excavation, which totaled 465 square meters (5005 square

feet), is illustrated in Figure 5-4.

r -----------------------------------------------------


I4 5


8 9 10 II

13 14

15 16
I -I

13 14!I



3 meters
3 meters

Excavation Plan, Hird Site.

Figure 5-2.

Figure 5-3.

Excavation Plan, Dobree Site.



-- - - -- - - -












* 1

A permanent datum was established at the Hird and Dobree

sites to facilitate use of a transit, chain, and stadia rod

for horizontal and vertical control. Excavated dirt was

screened through one-fourth inch square mesh fitted to

gasoline-powered shaker screens (Figure 5-5), with the

exception of Square 13 at the Hird Site where three-eighths

by one inch diamond mesh was used. Water screening was

employed whenever necessary, for instance when excavating

features containing wet soil or high bone densities. One-

eighth inch screen was used for excavating part of a barrel

well at the Dobree Site, but this technique resulted in

greatly increased excavation time and labor without a

corresponding increase in artifact recovery (most of the

small bone recovered with this screen size could not be

identified). Other features at the site were therefore

screened with the one-fourth inch mesh.

Besides the extent of the sample frames, the most

important difference between the field strategies employed at

the Hird and Dobree sites consisted of the way in which some

artifact classes were collected and recorded. Brick, tabby,

and shell were noted only on a presence-absence basis at the

Hird Site, while at the Dobree Site these artifact categories

were quantified for nearly half of the excavation units in an

attempt to define the location of a domestic structure. In

view of the extreme limitations that presence-absence

approach to artifact collection imposes on the interpretation

of the archeological record (Taylor 1948), and the





b) I



..H bO






* A

;I **~~ *~

^ Y I / .. ',.' ,: ,'
Ii J i
,* ~ ~ ., .. ' ,; / *,'
** ", .er 1 '': ** v i
"I, I i

S ** '
I 4 K




I o

1. I.,-

l ". I-

demonstrable research utility of quantification of

construction-related artifacts at historic sites (Lewis 1977,

Kaplan and Coe 1976), brick and tabby quantification was felt

to be justified despite the time and effort devoted to it.

The quantification technique used at the Dobree site

consisted of measuring the volume of all construction

materials recovered while screening and recording the total

volume recovered for each provenience in the field notes.

Volume measurements to the nearest liter were taken by

depositing the brick and tabby .in plastic buckets with liters

marked on the sides.

Horizontal Control

The discovery in 1952 of the Hawkins-Davison common wall

on the boundary line between Lots 1 and 2 South has provided

a convenient and accurate datum for subsequent archeology at

Frederica. The hypothesized outlines of the Hird and Dobree

sites were determined by measuring the appropriate distances

from the common wall, using colonially reckoned cardinal

directions as determined from the layout of Broad Street

(this street extends in a direction that is 99 degrees east

of magnetic north). At the point where a straight line from

the Hawkins-Davison wall forms a right-angle intersection

with the center of Broad Street, the distance measured to the

southwest corner of the Hird Lot was 450 feet east (137.1

meters) and 142 feet north (43.2 meters). This latter

distance was erroneously reported as 190.75 feet north in the

1975 Hird Site study (Honerkamp 1975:70). Widths of 75 feet

for Broad Street, 30 feet for Cross Street, and 14.5 feet for

the east-west alley are assumed, as are lot dimensions of 60

by 90 feet (see Table 4-1). The northwest corner of the

Dobree Site was found using the same assumptions for street

widths. In addition, the east-west street lying between the

lot and Broad Street is assumed to be 22 feet wide. From the

Hawkins-Davison wall-Broad Street intersection to the

northwest corner of the lot the distances used were 510 feet

east (155.4 meters) and 245 feet south (74.6 meters).


As is probably true of most sites at Frederica, both the

Hird and Dobree sites were found to have been extensively

disturbed by 19th and 20th century plowing. At both sites

the dark gray A soil horizon was equivalent to the plowzone;

this zone was excavated in two arbitrarily defined levels.

Zone 1-A was 0.15 to 0.20 meter thick as measured from the

ground surface after the removal of sod in each square. Zone

1-B included all soil down to the light brown to tan sterile

sand which made up the B soil horizon. This lower level of

the plowzone was usually 0.10 to 0.15 meter in thickness.

Man-made features, which were discernible only when they

extended below the plow zone into sterile sand, were

excavated separately. The Hawkins-Davison Site apparently

was not plowed during recent times. Fairbanks found a zone

of sandy humus covering the colonial remains that was 0.20 to

0.30 meter thick (0.7-1.0 foot).

Natural disturbance processes were also present at the

site. Animal burrows and root stains were commonly

encountered, as were numerous burrowing insects, especially

beetle larvae. At the Dobree Site a count was made of the

number of grubs encountered in a 15 centimeter level of the

plowzone; 24 were seen in one three meter square. Since the

life cycles of most of the grubs identified in the sample are

less than one year in the larval stage, more than 200 years

of concentrated "grub activity" would be sufficient to

obliterate most of the evidence of stratification of cultural

or natural zones. As Wood and Johnson point out, one of the

most common consequences of insect action is the blurring of

natural or cultural boundaries in the soil (1978:322). The

lack of stratigraphic information noted by Fairbanks at the

Hawkins-Davison Site, and the difficulty that the present

author experienced in defining the outlines of features at

the Hird and Dobree sites are certainly characteristic of

faunalturbation activities, and it is suspected that similar

processes will have affected all archeological sites at


Analytical Methods and Results

As indicated in Chapters II and III, the methodological

approach taken in this study is based on South's pattern

recognition method. This has had far greater influence on

the ways in which the artifacts have been analyzed than on

the ways in which they were excavated. The earlier work done

by Deagan (1972) and the author (1975) was not based on

South's approach, and as a consequence there are differences

in the analytical techniques used and the results obtained

from the application of these distinct techniques. For

instance, Deagan did not completely quantify all artifact

classes from the Hawkins-Davison Site, so that only the

ceramic and glass from this site are useful for testing some

of the hypotheses presented in the next chapter.

Artifact Analysis

The first step in the analysis of the archeological

materials was artifact identification. Ivor Noel Hume's A

Guide to Artifacts of Colonial America (1974) was the primary

reference used by Deagan and the author in the identification

of ceramic types. The work by Miller and Stone (1970) was

also frequently used. The primary sources consulted in the

identification of glass artifacts were Brown (1971) and Noel

Hume (1969b, 1974). References for metal and flint artifacts

include Stone (1974), Hanson and Hsu (1975), and Hamilton

(1976). The archeological type collections at the Department

of Anthropology, University of Florida also facilitated

identification of ceramic and nonceramic material. Floral

materials were identified by Dr. David W. Hall of the

Vascular Plant Herbarium, Botany Department, University of


Data Management

How to handle the sheer quantity of artifacts identified

from the Dobree Site was a major problem during the analysis.

It was quickly realized that the hand-tabulated quantifica-

tion techniques used by the author for the Hird Site material

was inadequate for organizing and analyzing the data base

resulting from the nearly complete excavation of the Dobree

Lot. It became obvious that the most efficient means of

working with the extensive Dobree data base would be through

the application of computer data handling capabilities. Of

the several computer packages available for this purpose from

the Northeast Regional Data Center (NERDC) at the University

of Florida, the Statistical Analysis System (SAS) was

considered to be the most suitable package for meeting the

particular organizational and analytical-needs of the author.

This choice was based on the highly versatile and powerful

capabilities SAS has for information storage and retrieval,

data modification, file handling, and statistical analysis.

In addition the system is relatively easy to learn and comes

with an understandable though terse user's manual (Helwig and

Council 1979). The package is limited to IBM hardware, which

at NERDC consists of an IBM-360/370 computer.

The input data used to create a SAS data set for the

Dobree ceramics consisted of 343 cases, with each case

containing information for three identifying variables and

frequency values for 20 ceramic types and six ceramic wares.

The identifying variables used were the original field

specimen number, the excavation unit number, and the

provenience designation. Nonceramic artifacts were also

coded into SAS data sets using the same identifying

variables. Through reference to one or a combination of the

identifying variables it was possible to combine, divide, or

alter the ceramic and nonceramic SAS data sets in any way

that was useful for analytical purposes. Although the

conversion from analysis forms to SAS data sets was time

consuming, it ultimately spared the author many frustrating

hours of hand tabulation. Conversion to SAS data sets also

allowed application of statistical techniques that would have

been extremely difficult, if not impossible, to perform by

hand, such as computing the correlation coefficients for

several ceramic types based on frequency of occurrence for all

63 squares at the Dobree Site. The SAS data sets created for

the artifacts and faunal materials from the plowzone and

features are presented in summary form in Appendix A through

Appendix D. Original analysis forms are on file at the

National Park Service Southeast Archeological Center,

Tallahassee (artifactual) and the Florida State Museum,

Gainesville (faunal).

Of obvious utility for establishing temporal parameters

for almost any historic site is the mean ceramic date formula

(South 1972). Mean ceramic dates for the excavation units,

features, and the sites themselves were found for the Dobree

and Hird sites. The Hawkins-Davison ceramic analysis did not

allow application of the formula. The dates derived for the

Dobree ceramic data for both the plowzone and features were

calculated through the use of a SAS subroutine devised by the


Another computer program used in the analysis of the

Dobree Site artifacts was the Synagraphic Mapping System, or

SYMAP. This computer graphics program produces maps on which

spatial data are graphically displayed through variable

darkness and texture. It is especially useful for providing

a method of interpolating data values for locations in a

mapped area based on the values of the closest known data

points in that area. The construction of artifact density

isopleth maps composed of contour lines connecting all

locations having the same density values is one example of

SYMAP's interpolation capabilities. Although the user's

manual for this package is obscurely written (Dougenik and

Sheehan 1975), the program was easy to use once the basic

mechanics for obtaining output were finally mastered. SYMAP

contour maps delineating relative densities of several

classes of artifacts from the Dobree Site were used to

address questions of site structure and function.

Classification Format: Dobree Site

The specific analytical tools used in this study were

chosen for their applicability in addressing the research

questions of interest to the author. The type-ware-class-

group classification scheme, which is an integral part of the

pattern recognition approach (South 1977a:92-102), has been

followed. The artifacts were divided into the same classes

and groups that South used to construct empirical artifact

profiles for deriving the Carolina-Architecture artifact

patterns. As already mentioned, analysis of the

Hawkins-Davison nonceramic material was on a presence-absence

basis which did not allow direct group-class comparisons with

the totally quantified Hird and Dobree assemblages.

The artifact classification for the Dobree and Hird

sites conformed closely to the format proposed by South in

order to allow intersite comparisons to be made of the

empirical artifact profiles that were constructed for each

site. In doing so, however, a number of decisions concerning

the structuring of the data base associated with the sites

had to be made, especially with respect to the inclusion of

various categories of artifacts in the pattern recognition

format. Since this study is concerned with aspects of the

British colonial occupation at Frederica, artifacts that were

known to have been deposited later were excluded from the

analysis. For instance, wire nails and ironstone pottery

were certainly not discarded at the Dobree or Hird sites

during colonial times.

The reasoning behind the decision to exclude artifacts

of the "wrong" temporal period from the analysis is explained

as follows. South's approach attempts to delineate patterns

in artifact relationships and distributions as they apply to

the British colonial system at historic sites. It is

reasonable to assume that inclusion of noncolonial artifacts

in the analysis would tend to obscure evidence of colonial

activities or behavior. It is therefore desirable to

eliminate from consideration those aspects of the data base

which are associated with a later noncolonial system. It is

also possible to further delimit the data base by excluding

colonial materials which must have been deposited after the

primary occupation at Frederica (1736-early 1750s). This

approach has been followed in the present analysis.

Unfortunately South does not delineate the data base

used in his studies in a manner that would allow other

researchers to replicate his classification format, and it is

this lack of explicit data definition that is considered to

be a major drawback in the pattern recognition approach.

Although it is true that studies of historic artifacts over

the last 20 years have rendered further detailed analysis of

some classes superfluous (i.e., ceramics, wine bottles,

pipes), it is equally true that other classes have been

neglected and need to be described so that other

archeologists will know what is or is not being included in


A brief review of the type-ware-class-group classifi-

cation procedure as it was applied to the Dobree and Hird

artifact assemblages will serve to point out the adjustments

made to South's original format as a result of unique aspects

of the data base at Frederica.

Since the Dobree site material is reported here for the

first time, descriptions of some of the artifact types used

in the class categories are more detailed than is usual for

most pattern recognition studies. This serves a dual purpose

by: 1) allowing researchers with differing orientations to

incorporate the data reported here, and 2) making explicit