MSU

LIBRARIES
___—..

 

 

RETURNING MATERIALS:
Place in book drop to
remove this checkout from
your record. FINES will
be charged if book is
returned after the date
stamped be1ow.

 

 

 

 

 

TII ANALYSIS 0’ OIAIGIIG PATTIIIS
OP PIPUSI DISPOSAL AID SITS UTILIZATION
OI A "ID-19TH TO SARLY 20TH C. BOUSISITS
II MISSISSIPPI

By

Leah Denise Rogers

A THESIS

Submitted to
Michigan State University
in partial fulfill-cut of the require-cuts
for the degree of

MASTER OF ARTS

Depart-ant of Anthropology

1985

.yl
4

Copyright by
LEAH DENISE ROGERS
1985

ABSTRACT
TEE ANALYSIS or annSIua PATTSRNS
or Rsrnss DISPOSAL Ann SITE UTILIZATION
on A HID—19TH To EARLY 20TH c. HOUSESITE
IN MISSISSIPPI
3!

Lash Dsniss Rogers

The Cedar Oaks housesite, located in northeastern
Mississippi on the Tonbigbee River, contains the last
standing dwelling fro: the extinct town of Barton
(ca. 1848-1870).

This thesis is concerned with the determination of
patterning in site utilization and disposal behavior as
related to the change from an in-town residence (ca.
1848-70) to a rural far-stead (ca. 1870-1940s) after
Barton’s de-ise.

By analyzing archaeological, oral, archival,
architectural, and artifactual data, a pattern was
discerned wherein the Barton-period occupation saw
deposition primarily in the backyard service area and
along the yard perineters, while the frontyard was swept
clean.

The post-Barton period saw refuse concentrations
closer to the house, in specific dusping areas, and along

the perineters. The backyard/frontyard dichotomy

continued, although the latter accumulated a shallow sheet
midden.

Comparison with analogous Barton housesites showed
the same shift in patterning after Barton’s demise,
related primarily to a change in the conception of spatial

utilization.

ACKNOWLEDGMENTS

There are many individuals without whose guidance,
assistance, and support this study would never have been
completed. To Dr. Charles Cleland I extend a special note
of thanks for his guidance throughout my career at
Michigan State University, and for giving me the
confidence to see this thesis through to completion.

Special thanks also to Dr. William Lovis for his
untiring assistance in the statistical analysis and for
being tolerant of my endless questions.

I must also extend deep appreciation to Kim and Steve
McBride for allowing me to utilize their ongoing research,
as well as for their friendship and support. The study of
Cedar Oaks would never have reached this stage without the
monumental effort put forth by Kim McBride in the
completion of the Phase III project report which formed
the basis of this thesis.

Thanks also to Michael Hambacher for his neverending
assistance and, above all, for his friendship. This study
would have been a far more arduous undertaking without his
input and moral support.

Last, but certainly not least, I wish to thank

G. Clark Rogers for his patience and loving support.

iii

His encouragement gave me the boost I needed to continue
this study, and without his assistance in the analysis
and preparation, it would never have been finished on

schedule.

iv

TAILS 0? CONTENTS

Page
List of Tables .................................. vi
List of Figures ......... ........................ viii
Chapter 1 INTRODUCTION ...... ........ .......... 1
Scope of Research ... ......... ..... 10
The Problem ....................... 12
Chapter 2 METHODS ............................. 22
Field methods ......... ....... ..... 22
Analytical methods .......... ...... 25
Chapter 3 THE DATA ............................ 46
Archival . ......................... 46
Oral history ............ ........ .. 51
Architectural ......... ..... ....... 56
Archaeological —-
architectural integrity .... ..... 61
Archaeological --
site utilization ........... ..... 67
Archaeological --
refuse disposal patterns ..... ... 88
Chapter 4 INTERPRETATIONS ..... ..... . ...... .... 114
Chapter 6 CONCLUSIONS ......................... 143
Bibliography .................................... 147

..‘H‘V

Table

10

11

List of Tables

Date ranges and references ...... ......... 26

General date range breakdowns for
features by percentages .................. 32

General date range breakdowns for
non-feature artifacts by percentage ...... 34

Feature date ranges by average density
pernzOOOO....0............OOOOOOOOOOO..O 40

Date ranges for non-feature assemblages
by average density per m2 ................ 41

Functional breakdown of artifacts for
selected features by percentages ......... 65

Functional breakdown of non-feature
artifacts by percentages ................. 77

Functional breakdown of non-feature

artifacts by percentages ..... ............ 78
Contingency Tables -- Average density

per m2

Backyard ................................. 102
Frontyard . ....... . ..... ............ ...... 102
Contingency Tables -— Average density

per m2

Mid-19th century ........ ................ . 105
Late 19th century .. ....... ..... ......... . 105
20th century ..... . ....... . ............... 105
Contingency Tables -- Average density

per m2

Mid-19th century .. ................. ...... 107
Late 19th century ........................ 107
20th century ............................. 107

vi

12 --

Contingency Tables -- Average density
per m2

Non-feature artifacts minus
architectural items ........

...........

Ceramic tableware -- non-feature .........

vii

110
110

10

ll

12

13

LIST OF FIGURES

Page
Tombigbee Historic Townsite and
Vicinity .... ....... ...... ................ 2
Hypothetical plat of Barton,
Mississippi ....... ....................... 3
Cedar Oaks photographs
House looking east ....................... 5
Shed looking northeast ...... .... ......... 5
Cedar Oaks floor plan -- present and
original, after Brito 1983 ............... 6
Cedar Oaks housesite map ... .............. 8
Cedar Oaks housesite -- soil map ......... 24
Cedar Oaks, map of features .............. 53
Feature 1 (brick foundation) ............. 55
Mid-19th century artifact distribution
map ........ . ............ ...... ........... 95
Late 19th century artifact distribution
.ap OOOOOOOOOQOOO ........... OOOOOOOOOOOOOO 96

20th century artifact distribution map ... 97

Non-feature artifacts minus architectural
items

Backyard ................................. 109
Frontyard .... ............................ 109
Non-feature ceramic tableware

Backyard ........ ..... ....... ............. 111
Frontyard ................................ 111

viii

l4 —— Griswold housesite map ................... 132

15 -- McGowan housesite map .......... .. ........ 133

ix

Chapter 1

INTRODUCTION

Project Description

In 1979, Michigan State University contracted with
Interagency Archaeological Services-Atlanta to investigate
historic period sites within the proposed Barton Ferry
Recreation Area of the Tennessee-Tombigbee Waterway. The
project area is generally located on the west side of the
Tombigbee River in Clay County, Mississippi (Figure 1).

Three nineteenth century townsites were studied as
part of the Tombigbee Historic Townsites Project and, in
order of chronology, these were as follows: Colbert
(ca.1834-1847), Barton (ca.1848-1870), and Vinton
(ca.1850-1920). It is known that Barton was founded, on
higher ground, after Colbert was destroyed by a flood in
1847 (Minnerly 1983: 2, 111).

The Cedar Oaks (2201809) housesite is situated in the
northeast portion of the Barton townsite, on a high knoll
approximately 175 m from a blufftop ridge on the south
side of the Tombigbee River (Figure 2). This housesite
functioned as a residence and farmstead and was occupied

almost continuously from the 1840s up to the 1940s.

é '. "if‘11

143.11 "P'C'U'X'Hi 'I‘VI

1:01.1qi1fiafju .1 «413) 1 8'

 

 

 

 

 

 

 

 

 

 

 

 

 

 

     
   

  

 

 

 

 

 

 

 

 

 

II 15
cows or ENGINEERS
souquav
vm'rou
sun-:4,
35 N
T. IS 3. 4—
11:7 3.
COLBERT /
SITE
2 i-
TWIN"
LOCATIONS
cuv couwT v Caefi“ U
0
In.
DRAWN: VI.D.LOLLAR e/Ie/So [ SCALE: 1:24.000
hi Fig. i Tombigbee Historic Townsites and Vicinity

 

 

Figure 1 . Tanbigbee Historic 'Ibmsite and Vicinity

 

«mmwnndnnaz .couumm mo usam Hmofluonuommm .N $5..va

 

wee. 2...... <
c 3

:16. o .51.!

u; 3 algal
9: , aw :91. a5:

:1. E 3....“

><Js mash-Xbogx 1

 

32-2-2: into. I user

 

>3: nah—0‘0» 9g... Bali‘s.—

 

 

 

 

 

    
    

_ 2:5

 

wussilo»

 

 

4

Standing structures on this site at the time of investi-
gation included a house, shed, privy, and two small barns
(Cleland and McBride 1983: 318-20).

This housesite was investigated in detail through
three phases of archaeological, oral, archival, and
architectural study. Attention was focused on Cedar Oaks
because the house represents the last remaining standing
structure from the town of Barton. It is a 48 x 36 ft
vernacular Greek Revival frame building with its long
dimension on a north—south axis (Howard 1978) (Figure 3).
It is one-story, with four rooms symmetrically placed on
opposite sides of a central, closed—in hallway.

The two front rooms are larger than the back, and
each room originally had separate fireplaces located on
the south and north ends, respectively (Figure 4). Only
one chimney remains standing and this was built in the
1940s to replace the southwest chimney.

The west facade serves as the front entranceway with
an open gallery spanning its full length. It has a
central doorway which is flanked by single windows in the
north and south front rooms. The northern facade shows
three windows with two in the front room and one in the
back room while the southern side only shows two windows
flanking the chimney in the front room. The back facade
(east side) shows two doors, one opening in the northeast

rear room and the other a folding door that opens from the

 

a. House looking east

;

 

b. Shed looking northeast

Figure 3. Cedar Oaks photographs

i'"'i
1-t£:

'1'.
I
I
I
I
I
I
I
I
I

.-.L_-.
d

mmmfl conga “mums .Hmchmfluo new summons I- amamuooam mxmo nmeoo .v masons
swam agmfino "Ills .

 

O§ — nu Cm -> Au :0: 2“ mnunl'lfl :MHAM “amen N "Huh-"III-"Hu
-G """"""""" ham—ell I 4 "Islsuubr II I' Ilnur ||||||||||| LU-
_

-

f

- 4

_

_

 

 

‘lmqq
I

I

I

I

I

Ii

.11

l-

3 5
3
---l

 

 

 

central hallway. No steps exist at present for either
door; whereas, wooden risers now serve as the front (west
side) porch steps. The house itself is raised approxi-
mately 30 inches on brick piers, and the roof is covered
with sheet metal which replaced a storm-damaged roof in
the 1970s (Howard 1978 and Brito 1983).

The existing shed is a 3 x 4 m frame structure with
the long axis running north-south (Figure 3). It is
situated 7 m east of the northeast corner of the house
with its only doorway facing west. This shed is a
twentieth century structure constructed of material
salvaged from an earlier building (Ibid.). The only other
standing outbuilding located in the immediate yard area of
this housesite is a twentieth century privy located in the
northeast corner of the backyard, 27 m off the northeast
corner of the house (Figure 5). An open brick-lined well,
approximately 16 m deep, is located between the shed and
privy and near a large southern red oak (Figure 5).

The house, shed, privy, and well are situated on a
relatively flat knoll in a cleared area approximately 70 x
40 m in dimension. The land slopes gently away from this
flat area, with the steepest slopes on the southern and
western sides. A dirt road extends down the western slope
away from the front of the house and connects with a more-

or-less north-south dirt road which runs from the present

El

 

 

IE]

E]

 

Bum c858 Hus “:5 .3 .

 

§u*

 

5800.3”

 

one 8388: mono 380 .m 853m

 

 

 

9
day main entrance into the Barton site up to the bluff
above the Tombigbee River (Cleland and McBride 1983: 308).

Presently, the vegetation of the site shows
purposeful placement and cultivation. The front of the
house is framed by two large cedar trees placed on either
side of an opening in the front fenceline. Two more large
cedar trees are located on the northwestern portion of the
fenceline with one actually serving as the corner
fencepost. As mentioned above, a large oak tree dominates
the backyard area and is situated approximately 16 m east
of the house. Fruit trees, predominantly peach, were
planted along the southern edge of the site. A flower
garden recently occupied the extreme northwest corner of
the frontyard (Figure 5). The surrounding land now shows
secondary hardwood forest dominating land once cleared for
cultivation but generally abandoned in the 1940s.

A portion of the fenceline remains standing and this
runs along the western edge of the frontyard and part of
the northern and southern edges (Figure 5). It is a wire
fence with an opening between two cedar trees that serves
as an entranceway into the site. A photograph taken in
front of Cedar Oaks ca.1909 shows a front fence in
approximately the same location but of a wooden picket
construction (Minnerly 1983: 30).

The remaining extant outbuildings are located 130 m

northwest of the house. These are two small barns, one of

10
log construction and the other plank (Minnerly 1983: 32).
They appear to be associated only with the twentieth-
century occupation of Cedar Oaks (Cleland and McBride
1983: 320).

A 70 x 40 m arbitrary boundary was delineated on the
Barton archaeological survey grid which encompassed the
majority of the cleared area around the house. This 2800
.2 area served as the focus for intensive archaeological
excavations during Phase I and a portion of Phase II
(i.e., November 1979 through July 1980) of the Tombigbee
Historic Townsites Project. These excavations yielded
106,452 artifacts from 99 excavation units or 364 m3.
Oral history interviews, archival research, artifact
analysis, architectural studies, and the archaeological

results provided the data for the following research.

Scope of Research

Initially, the primary focus of the research strategy
was to investigate the integrity of the house in its
present location. Excavation was geared to collect data
on the possible existence of earlier structures on this
site, to determine if the house had been moved or
structurally modified, and to elucidate the evolutionary
architectural history of this structure.

A second research problem concerned defining the role
of this site in relation to the town of Barton. In order

to achieve this goal the entire site was investigated as

t
. .
.
, I
‘1 .
.
‘.
'I
I O
\ (I
a

I

OIH‘I

.
. 1 .
' C
.
D I
‘ I
' I
I
i v .

‘0 But")?

I ' x I
o
i l
. I
I ' . I ' I '

11

to its utilization through time. This included the
definition of activity areas, the placement and function
of outbuildings, and ascertaining the changing nature of
its occupational history.

The third research problem involved the determination
of refuse disposal patterns and how they changed through
time. It was felt that such information could help
explain changing patterns in the material culture and
human behavior in this area during the nineteenth and
twentieth centuries.

The Cedar Oaks housesite is appropriate to this
research for a number of reasons. These are as follow:

1. The standing structures served as focal points
for the placement of excavation units facilitating
investigation of all three problem areas, as well as
serving as artifacts themselves.

2. The high density and wide variety of artifacts
provided an assemblage well suited as a collection.
Unfortunately due to budgetary limitations, the project’s
computer program never reached the stage where reliable
data and statistical manipulations could be obtained.
Therefore, analysis was limited to basic quantification in
the form of relative frequencies, percentages,
distributions, and basic statistical calculations. The
large and diverse assemblage did, however, provide

valuable chronological and functional information.

J.

-I o I .
I
I .7.
.‘I I-
I: Q I
i.
I II
In a I
. .
I I...
It
1 . .
It . .0.
I .I
II. I .
o
I I
V
) ’ I
lei o
u
C?
'1 .
II I
II
I. I-
I
I
. I I
s. . n
.. .u .i
I
I
I I
L
I
J
l

... a
I.
II.
I Q
I .l
:0.
n
I l
.
I
I I:
U
I. .b
..I
I
.I 0
I!
1
‘I I
I
II t.
.
.
l
a
d
I
I e
Q
t
o
l
d.

12

3. The horizontal and vertical extent of the
artifact distributions and feature concentrations, coupled
with data retrieved from the Barton (22C1807) townsite
surrounding Cedar Oaks, provided spatial and behavioral
information whereby activity areas could be defined, as
well as refuse disposal practices through time.

4. The recent nature of this site (i.e., mid-
nineteenth to early twentieth century) provided not only a
written record which could be examined for corroborating
and unique information, but also an oral record whereby
data unobtainable through other sources could be added to
the history of this site. Both sources proved to be of
great value in the interpretation of data from this
investigation.

5. The association of this housesite with a
specific town provided the opportunity to examine the site
as part of a social and cultural system; that of a
residence in a small nineteenth century Southern river
town, and later, as a survivor of a failed system when it

became a farmstead after Barton ceased to exist.

The Problem

The analysis of refuse disposal behavior and site
utilization has increasingly become a focus for historic
sites research. Studies centering on the determination of

behavioral patterns as reflected in the archaeological

l3
record have provided insights into status, ethnic, and
economic associations, to mention a few (see South 1977,
Drucker 1981, Otto 1977, and Moir 1982). Some studies
have approached such problems with indepth analyses of
site layout and activity areas as reflected in structural
and artifactual distributions (see Eeeler 1978, and Lewis
1977). Others have focused on refuse disposal patterns as
evidenced in the archaeological record and the
configuration of its deposition (see South 1977, and Moir
1982).

The underlying theory supporting such research,
postulates that human activity and behavior affect the
formation of the archaeological record in definable,
systematic ways. This is not to say that cultural
processes are the only factors in site formation, as
gggcultural processes also play a role in the final
configuration of the archaeological record. Erosion,
animal burrowing and rooting, as well as post-depositional
cultural processes such as plowing and salvaging, all have
an influence on artifact distribution and stratigraphic
integrity. It is the elucidation of these formation
processes that must be undertaken if the researcher is to
understand the archaeological record and the information
contained therein (Schiffer 1976:11 and 1983: 675-6, South
1977: 31-43, Binford 1981: 199, 205, Hayden and Cannon
1983: 117, and Murray 1980: 490).

A;

14

The attempt will be made in this study to define some
of the formation processes which are manifested in the
archaeological record found at Cedar Oaks. Noncultural
and post-depositional cultural factors will be considered,
however, the specific concern will be with the cultural
processes involved in the utilization of this site and how
this changed throughout its occupation as evidenced by
structural locations and their functional associations.

Refuse disposal patterns will be explored in
conjunction with this structural evidence as disposal is
also an aspect of site utilization. As Priscilla Murray
(1980: 490-1) has noted ”material elements come to rest
where they are found at an archaeological habitation site
basically as the result of two purposeful human behaviors
(besides burial behavior)--discard behavior and
abandonment behavior."

Two general types of discard behavior were noted at
Cedar Oaks, specific dumping and an overall refuse-laden
sheet midden. The former consists of secondary refuse
(i.e., discarded away from its location of use) dumped in
specific definable locations, and the latter consists of
an accumulation of primary (i.e., discarded at its
location of use), secondary, and de-facto refuse (i.e.,
lost or abandoned items) located primarily in the backyard
(Schiffer 1972: 161-163, South 1977: 296-299, and Smith et
a1. 1982: 224).

15

Often sheet deposits on historic sites are dismissed
as being of little use in determining behavioral patterns.
Randall Moir (1982: 139), however, in studying historic
sites in the Richland Creek area of Texas, has found that
”sheet refuse deposits... have considerable behavioral
integrity," and can provide fairly accurate information
concerning the chronology of a site’s occupation,
patterning of yard usage, and associated socioeconomic
information.

By integrating archaeological, artifactual, archival,
oral, and architectural data sets from the investigation
of Cedar Oaks and the Barton townsite, this study will
test the following general hypothesis: The changing
refuse disposal patterns at Cedar Oaks were a result of
the functional change from an in-town residence (ca. 1848-
1870) to a rural farmstead (ca. 1870-1940s).

Specifically, the change in site function and utilization
affected a change in the disposal behavior of its
occupants.

In order to examine this general hypothesis the
following, more specific hypotheses will be tested.

1. A change in site function from a town residence
to a farmstead involved changes in the spatial utilization
of the houselot, including the function, location, and I
necessity of certain outbuildings; the location of activ-

ity areas; and the patterning of refuse disposal behavior.

If?

.1
#w‘
I
I
9-.
(‘I
fit!
A!’
I 4
f
.I
(I

AU

.III.

"C iii

”1"

.U.

I
‘1
I
‘s
I
'I
I'
A
-I
0
{ll

' I

.. a; ‘:H .
a; Iii 'I w

J: I {133‘}
'1" 621

i’ i '
MI 'I ' '
,. H
:' nu
{:11 Ii :

?

l I ’ "Ir;
I I i I
in"!!! :1"
::. ,:

, .
)I.
it" ","
.' -..; ,
c 1.,1‘

, I i
xiii“. ‘ . ‘
II I . .
I ‘_ I 3:". I
i‘ In; I

.-"'I. :III'),,

.L'Iii'diii 3..)

:.~II:'II:i I. m

 

 

16

One test implication of this hypothesis suggests that
since a far-stead would have to be lore self-sufficient
than an in-town residence, there would be greater need for
kitchen gardens, small animal pens, and food processing
and storage buildings (e.g., a snokehouse). It is
therefore implied that such features would be sore
prevalent during the farmstead occupation than during the
town period.

A second test implication involves the idea that a
house generally serves as the centralized point from which
the majority of domestic activity originates and radiates.
It is therefore implied that the distance certain activity
areas, outbuildings, and disposal areas are located from
the house has significant meaning relative to the changing
function of this site.

Specifically, it is expected that during the town
period there would be few associated service structures,
and those that were present, would be in close proximity
to the house for easy access. A detached kitchen is
likely the only service structure needed during this
period as produce and other goods were readily available
in the local stores. Likewise, activity areas and their
associated refuse deposition, would be less evident owing
to the primary function of this housesite as a residence
during the town period. What evidence there might be is

expected to be within close proximity to the service

17
structure(s) and/or house where doaestic activity would
have taken place.

Further-ore, it is expected that refuse dumps
containing decaying organic aaterial and sharp, broken
items such as glass, would be located away from well-
traveled pathways and generally out-of—sight and snell.

It is entirely possible that a community dulp was
available during the town period.

On the other hand, a far-stead would require more
service structures including: a saokehouse to provide a
meat supply; sheds to store tools and machinery; and barns
to shelter aniaals and their food supply. It is expected
that a saokehouse would be close to the kitchen facilities
to allow easy access, yet far enough away to eliminate the
undesirable pollution created by neat processing. Sheds
and barns would be located at greater distances from the
house, in closer proximity to the animals and farmland
that required their presence.

Furthermore, refuse deposition night have become more
prevalent closer to the house and its surrounding lot
during this period, as doaestic activity would have
increased due to the need to be aore self-sufficient in an
increasingly isolated rural setting. .

A final test iaplication of this hypothesis involves
the conception of space and its utilization, and how this

was affected by the change in site function.

18
Specifically, it is iaplied that there was a distinction
made between the front and backyards as far as the
location of service structures, refuse dunps, and activity
areas is concerned.

During the town period it is proposed that the
frontyard would have served as a aaintained presentation
area with only moderate use for doaestic activities.
Therefore, it is expected that there would be little, if
any, sheet refuse accumulation, no outbuildings would be
present, and activity areas would be related only to
maintenance and ornamental activities (e.g., yard
sweeping, flower gardens, walkways).

It is further expected that this presentation area
would be more forsalised during the town period, breaking
down somewhat during the far-stead period. In particular
during the later occupations, there might be less evidence
for careful yard maintenance (i.e., a sheet midden might
begin to accumulate in the frontyard) owing to the effect
that the isolation of a rural environment would have as
opposed to the day-to-day scrutiny of a town setting.

On the other hand, it is proposed that the backyard
would have served as the focal point for the aajority of
domestic activities and outbuildings, and likely received
sore refuse deposition (intentional and otherwise). It is
postulated that during the town period the backyard served

as the locus for domestic activity related only to the

: .o;iisi 0 can w 91? i is z :130: a: :i .; .53;35790

on an ugt a: c an -30: n;u 3301' an? 1 swans -vnm

03 P40 ooxur 3 U? u .09 03011 9 svtvuo :o

1
(J

8:) "I:

(.1
D
V

.30 109u00 e: DZWB

(u

.2 9L #1 u .200 1w0' an: i .1”:

0:13 noza11 nanzn min: 3 as b91190 9 is n10

L l

D
I.)
‘—

9 .Lviwon miss on: not on 935:9060

LJ

1..
..2
O
(.1
I l
C:
..J
3'.

a

LL
l'v
$
d
O
6..
5-1
3
D
t .‘J

:1 .3113: ,Lurw 0 on: Jan: oeio-qxs u;

u: . ~

0351335: in 'nflc n;nn: 30 a. 3i .0: u 3:-

t
(3-
I

so: a 13 or 7nrnq 280 t 5 J on SUV‘B swan iUnw

:zviaos v: .I; on .820: .101330 one eoiirvz :o o:issoo:

. ‘ ~.:80190 03

M
In,
w
a

013 3303 09: 00:1 0 xx: 0:1 q':uu: 3nd 0510.

333 03 9:0 noisier V‘l 13 o 0:32 eon ~01 on o: 9d3 3:

19
site’s function as a town residence. Therefore, it is
expected that a ainiaal nuaber of outbuildings would be
present and located close to the house. Similarly,
activity areas would be in close proxiaity to these
structures. A sheet midden would probably accuaulate only
in the localized service area unless yard sweeping was a
strict practice in the backyard, as well as in the front.

Any intentional refuse duaps would be expected to be
located away from the house, where possible out-of—sight
and smell, for the benefit of visitors and occupants
alike. It is doubtful that the change fro. a town setting
to a rural farmstead would drastically change the general
conception of spatial utilization and the pride one would
take in an esthetic, sanitary environment. However, the
different needs and activities of a far-stead likely
served to lessen strict adherence to town standards.

It is proposed that the backyard, during the
farmstead occupations, would continue to function as the
service area; however, it is expected that sore
outbuildings would be present with a widening sphere in
their distance from the house relative to their respective
functions, with the emphasis now on self-sufficiency. It
would follow that there would be an intensification of
activity in the backyard with a sheet aidden accuaulating,
as a result, in the area of heaviest utilization. Kitchen

gardens and saall anisal pens could also be expected in

‘. .
.‘ ”‘HI’Z ‘
3
do? i .
’V 4P.
1‘ ~!
4 '-p‘,.\]
i: 1,: v
'V‘ .' EH 1! ’9
if?! {"31
'l' ' ' a .
'j-‘I‘, . f
.4' 1,. 'l'
{j ‘ 9- ‘ul'a.
,1 s , v
in l V '

“ T a')l '.
h" . "

, g . . ¢ ‘ .
' ' {I 19 I}:
l" -‘;:' .- . hr
. . .: "(H

I l .
‘ V , I '
v A l I . ' ‘ l
, l 1 J

"7 'F ‘H' m!
. 'I .I I ,'
'3' I}. ‘t I
,r [I ‘ O‘ '
‘1 I z ‘
. it A!
' L? ! \
;‘.‘ 1“! ‘i .
' "' .t :1;
It .1 '1' ,

3 h"! ‘v- ..' JUI It"! ' ) ‘n \ y'
'I , 1: «W o' .- a
I! I"' ‘ (gat' ‘.' I i. '
' "‘MH ‘0 H11 1 ., ”'
! .'" ~.-, 'Jtia ., 11. ,1 _ ‘. .3
4013.1”.05 ? 1‘. “1‘
" --., l. "t .14 '. ~
.0 1 v nu 1. no?! .5. . "' ~I
.-‘ 2 ‘1 ‘ 'H l N. ’ (

‘ I".
. I
rrHti

all!!!

 

 

20
the backyard near the structures where their products
would be used.

Finally, it is conceivable that refuse disposal
practices would also change as more of the surrounding
land was pressed into service as far-land, and less
eaphasis placed on esthetics. It sight be expected that
dumps would be located nearer to the house in the
backyard, where the activities generating the refuse were
focused.

Any patterning discerned at Cedar Oaks has little
significance when viewed in isolation. Therefore, it lust
be detersined whether this patterning is the result of
universal factors as far as cultural behavior relating to
changes in site function during this tine period and
region are concerned, or the result of site-specific,
idiosyncratic behavior. Comparative analysis is the best
approach to sake such a determination; therefore, the
following, final hypothesis will be examined:

2. Similar housesites in the Barton townsite
exhibit a changing trend in their pattern of site
utilization and refuse disposal, parallel to that
discerned at Cedar Oaks, and coinciding with the change in
site function.

A test implication of this hypothesis is that
deviations from the Cedar Oaks pattern would indicate that

other factors besides the change in site function affected

C!L.‘-'3fl.'. 9.7? fl‘:'- " "' i053 DB .TZ'H.‘ I Ti 75. [ifL'T’li'Wi’Wt

.7)U-',.C‘L'-’ ff‘i T "."H’IU ’3'}: 00'! 3 5"!" '}B " D'.‘ ”3.16 0'1."

7.7. ,. )0 DIUU)

L-..
'7
7A
in-
v
Oh

[man a:n n-';w' : n7 ,_osli=ufluu a a. ,v

:: v;11 c -ni : n 01 CH ennén‘ 3:.; a: on -‘ 3;ewq

.l 7: .BW .snzoa n~ nwnal prionqu

a.
v
w
—‘
i
v
‘u
-
o~
_
p.
a.-
p.
.4.
e.-
\
(W
i.
I
p...
‘
Us
3
s
I.
_
U
’C
HO
a
V
\
V
:._

'i (l 5. .' ,\'

S‘iU‘-'1 3'1’ H‘ .1" "11.”. 8.523;...23'33 5311! 2‘91“ 0’] '.‘.'.’.(‘.‘l

‘

.09 ll'!-:

O
'0‘

D
D
’4
'\
d
4
i
w
.2
l
[‘1
13
.4
‘-Ov
H
A
i w
'J‘
U
I-q
\
W
(O
.4
h

s; .‘T rum ' ..- (1 5 .1: I) :‘13 NHL? '1 : 0T 4

 

 

 

21
the cultural behavior which for-ed these sites. For
example, a factor such as differing socioeconomic levels

between the sites might be found to have had a greater

effect.

Chapter 2

HITBODS

Field Methods

The fieldwork on Cedar Oaks was geared specifically
to answer questions concerning site utilization and refuse
disposal; although the majority of the initial (i.e.,
Phase I) investigation was concentrated around the house
.in order to deter-inc its integrity and significance
(Minnerly 1983: 38).

As mentioned previously, the intensive excavations
esere confined to an arbitrary 70 x 40 a boundary. Of the
539 units which were excavated, 12 were 1 x 2 m and the
r‘emainder 2 x 2 m in diaension (Figure 5). A judgmental
t'esearch strategy was iaplemented in the placenent of
ulrits. Levels were excavated according to natural
stratigraphy, and soil was both dry-screened and water-
screened through 1/4 inch mesh (see Minnerly 1982: 178-
196, Minnerly 1983: 36-109, and Cleland and McBride 1983:
318-366 for detailed descriptions of the Phase I-III
investigations of Cedar Oaks).

Through the use of a systematic soil probe survey
over the entire site grid, a distinguishable stratigraphic

pattern was defined whereby feature locations and

22

f‘

ltssiiij'!"

f'fwt’

 

 

23
midden depth could be plotted and areas of potential
archaeological interest pinpointed (Figure 6).

Excavation units, and blocks of units, were
Judgaentally located in areas where architectural resains
and refuse middens were indicated. Features were
excavated primarily in block configurations (i.e.,
contiguous units) in order to fully expose their reaains,
as well as associated features and deposits. In all,
seven blocks of units and ten separate units were
excavated to sterile subsoil. A total of 66 features and
:seven subfeatures (i.e., small features such as postholes
:found within larger features) were discovered; 45 of these
users postholes.

The stratigraphy present on this site can be
ggeneralized as follows:

Level 1: Very dark grayish brown (lOYR3/2, Munsell
soil color code) to brown (lOYR5/3), depending upon its
«levelopment. This is a aidden resulting from the lengthy
(naltural occupation of this site and is not a natural
\uadisturbed Al or Ap soil horizon development (Murphee and
Miller 1976: 20-22). It is loany in texture and slightly
sandy in some areas, and the break between this aidden and
lover levels is abrupt and very distinct. It was thickest
in the backyard area, where it was 10 to 30 cm. ‘

Level 2: In some areas this is the A2 soil horizon,

which is yellowish brown (lOYR5/6) to pale brown (lOYR6/3)

~‘

'$.l'
:0
all
I, I

 

 

24

use :on ...... anaconda: 330 wsoou .o 9.3on

 

 

 

 

 

 

 

 

 

 

 

\‘II II I I, AmQUU 30 OH .eunOHQQflv
’Aagnou “nofiwae’ nkus can undo mo cowwanuo>o VPI/l confide gown
_ a l I /, x
_ y _ /.
/ — Econ 30 o 453 /
/ ’ sound: gown xumc _
x
/ ’ 3000 _ ~
/ 80 owned cocoa... gun xwoc _ H
/ z
/ / ...... .
/ _ N
/ / .
9530.?» an 00.3.39: /
5009.2: gown 8369: / / _
/ 326 .5 ~78 5632 .583 .
z /
x
n m . v z
E . /

25

sandy loan. This horizon shows a clear, snooth boundary
with the B horizon (Murphee and Miller 1976: 20-22). Over
much of the site, this level is either greatly disturbed,
leaving shallow discontinuous remnants, or is entirely
gone. The outer periaeters of the site do, however,
generally show a continuous, well-developed A2 horizon.

Level 3: This is the natural B horizon, the upper
part of which is strong brown (7.5YR5/8 to 7.5YR4/6) silty
to sandy clay loam or clay (Murphee and Miller 1976: 20-
22). A number of features were found to extend into this

level.

Analytical Methods

The analysis of the artifactual data went beyond a
mmere cataloging of iteas and consisted of detailed
[physical descriptions, vessel reconstruction, and research
into date ranges of nanufacture and popularity. A chrono-
ilogical classification was devised based on specific dates
and ranges assigned to the broader categories of 1800-1830
bearly nineteenth century), 1830-1860 (mid-nineteenth
century), 1860-1900 (late nineteenth century), and 1900-
1940s (twentieth century) (Table 1).

These date ranges were selected because they appear
in) best reflect the changing episodes in Barton’s history.
Specifically, 1800-1830 is pre-Barton; 1830-1860 covers
the heyday of Barton as a town; 1860-1900 encompasses

Barton’s decline and demise (i.e., 1860-1870) and the

E lnv‘oJ

Ef’JUd'1HM lswiJYIP-nlx

26

Table 1. Date ranges and references.

 

 

 

 

General
Qa£a_£aaga Artifact Data Reference
Iarly 19th Barly aachine cut nails 1815-30 Nelson 1968:6
century Handwrought nails pre 1830
(1800-30)
Mid-19th or c :
century Transfer print whiteware 1830-70 Lofstroa 1976:34
(1830-60) Price 1979:31
Flow blue whiteware 1840-70 Lofstroa 1976:9
Price 1979:31
Annular whiteware 1830-70s Lofstrom 1976:10
Price 1979:31
Shell-edged whiteware:
"Arrow” ca.1850-60s Miller n.d.
"Standard," scalloped 1st 1/2 18thc.
Sponged/spatter whiteware 1840s-60s Price 1979:31
Lofstroa 1976:9
Glass:
2 pc mold post 1840s Lorraine 1988:43
Blowpipe pontil mark pre 1857 Jones 1971:68-70
Sand-tipped pontil mark pre 1857 Jones 1971:68-70
Applied lip (hand finished) pre 1860 Lorraine 1968:43
Press-aolded (”Lecy" pattern) 1827-50 Lorraine 1968:43
Pro-chilled iron aold pre 1870 Lorraine 1968:43
Ganeral:
Modern machine cut nails post 1830s Nelson 1968:8
Late 19th Ceramics:
century Plain whiteware l840s-early South 1977:211

(1860-1800)

Shell-edged whiteware:
Painted-not impressed
"Standard,“ unscalloped

Sponge staaped whiteware

20thc.

ca.1850-80s
2nd 1/2 18thc.

1850-70s

Lofstroa 1978:10
Price 1979:22

Miller n.d.

Price 1979:31

27

 

 

 

Table 1. (continued)
General
2112.322127 Artifact Data Reference
glasa gottles
Applied lip (lipping tool
finished) 1850-1903 Lorraine 1968:43
Snapcase post 1857 Lorraine 1988:44
Mason Jar w/zinc cap post 1858 Lorraine 1968:44
”Mason” Jar distributed
by Dell Bros. post 1880 Reher 1977:237
Panel bottles w/eabossed
lettering post 1867 Lorraine 1968:44
Solarised (manganese)
bottle glass 1880-1915 Hunsey 1970:55
glasa: other
Press-molded, fire polished post 1850 Lorraine 1968:43
Kerosene lamps appear 1860s Lorraine 1968:44
Scoville 1948:47
Jar lid liner for screw-on
lids post 1869 Munsey 1970:146
Light bulb patented 1882 Lorraine 1968:44
Lightning fastener patented 1882 Lief 1965:13
General
Wire nails post 1850 Nelson 1968:6
Barbed wire patented 1867 McCallua and
McCallus 1965:244
20th Ceraaica:
century Decalcoaania (”Decal Ware”) late Hegars & Carley
(1900-40s) whiteware l9th-20thc. 1982:6-8

Glass:

Machine-sade bottles

"Coca-Cola," bobble-skirt
design

Applied color labeling

"Federal Law Forbids sale
or reuse of this bottle”
embossing

Glass bus candy container
eabossed w/"Victory
Lines/Special”

post 1903
1915

post 1930s
1933-64

1940-45

Lorraine 1968:44
Munsey 1970:105-6
Hunsey 1970:52
Munsey 1970:126

Muneey 1970:189

28

Table 1. (continued)

 

 

General

Date Ranga Artifact Date Reference
General:
Crown caps post 1892 Lorraine 1968:44

Lief 1965:17
Tin can w/locked
double seaa post 1897 Busch 1981:103

Beverage can pull tab post 1962 Busch 1981:103
Autosobile parts post 1908 Saith 1957:32
Plastic 20th c. Kaufsan 1963:74

 

29

changeover to a farming community; and 1900-1940s
represents primarily its ownership and occupation by the
Uithoven family (Cleland and McBride 1983: 331).

It is noted that some of the artifacts used in Table
1, such as modern machine cut nails, wire nails, and plain
whiteware, have date ranges which actually overlap other
categories. For example, plain whiteware spans the 1840s
through the early twentieth century (South 1977: 211,
Lofstrom 1976: 10, Price 1979: 22). It was placed in the
late nineteenth century category (Table 1) because its
median date falls within this period (South 1977: 211),
and most of the identified maker’s marks recovered at
Cedar Oaks date from the late nineteenth century. Similar
criteria was utilized in the placement of the other long-
range artifact types into specific date range categories.

Artifacts were further analyzed as to function, and
to accomplish this, nine categories were used. These are
as follows:

1. Ceramic and glass tableware--this included all
whitewares, refined stonewares (i.e., Ironstone), and
porcelain dishes (i.e., cups, plates, saucers, bowls, and
pitchers), as well as glass table items such as tumblers,
bowls, goblets, plates, etc. (many of these items were

press-molded).

. .
.
.
“I . I 1 ‘
1 . '0 .I .u
‘ I is. I ‘
'.' 4" " .

' r
e
'4
I '1 i "H
‘4‘ | .| I I 0.

' ull

"
m.

b...

I
v
I
r
1

30

2. Utilitarian ceramics--consisted of coarse
earthenwares and stonewares in the fore of crooks, jugs,
mixing bowls, and churns.

3. Bottle glass--all glass fragments and whole
vessels that could be identified as bottle pieces of any
function (i.e., alcoholic beverage, perfume, ink,
medicinal, etc.).

4. Foodstuff related material--this consisted of
all items, regardless of materials, that were positively
related to food storage, preparation, and consumption
(e.g., Mason Jars, Jar lid liners, milk bottles,
identifiable tin cans, spoons, forks, knives, butcher-cut
bones, seeds, and pits).

5. Clothing and personal items--includes all
clothing attachments and personal artifacts such as
buttons, snaps, rivets, eyelets, shoes, fabric, combs,
watches, eyeglasses, Jewelry, buckles, thimbles, smoking
pipes, mirrors, and scissors.

6. Architectural--includes nails, window glass,
brick and mortar samples, shingles, tarpaper, and
concrete.

7. Metal hardware-~consists of general non-
architectural hardware, tools, machinery parts, and
vehicle parts.

8. Faunal--includes all non-modified bone and

shell.

.u

31

9. Miscellaneous glass, ceramic, metal, and other
material--consists primarily of those items that were
unidentifiable as to function.

Any artifact could only be assigned to one functional
category, although it is noted that ceramic and glass
tableware, utilitarian ceramics, and some glass bottles
also functioned in foodstuff related activities.

The date range tabulations were first used on a
proportional scale for comparing levels within features,
as well as levels within the general non-feature
stratigraphy in the blocks of units (Tables 2 and 3).
These percentages included only datable artifacts, and
this data was used primarily to determine feature and
spatial function aiding in the identification of activity
areas and the changing pattern of site utilization through
time.

As proportional data, by itself, results in a
generalized impression of the actual data, further
analysis utilizing ratio and interval scaled data was
undertaken in order to test the proposed hypotheses (see
Chapter 1). As an initial step, the actual artifact
frequencies by function and date range were plotted on
graphs; wherein, the y axis = frequency and the x axis =
distance from the house in meters. In the latter case,
the house was designated as the (0,0) coordinate, and

excavation units were designated by their nearest distance

32

 

 

 

 

HN an e an em av n Nessa
ov ov a «N on u v floss“
an N I «H mm H m moved
3.49.43
an no cN on cm I N nosed
ma be nN Nv an - I H Ho>o~
m ousaseu
HN oc~ a N@ m I e ousaeou
me ha I c mm NH N gees”
N mm b we bN HN N ~o>o~
OH essence
be we I m Hm I N move“
an ee I I mm N N Ho>ou
mlflmmmmmm
ea m I I co“ I m ~e>e~
HN cN N ~N we I N nosed
me was «a non wNw I a doped
filmuwweww
ouasusv sIsHoo ousanloaae assuoeo husuooo assasoo hasaooo .oo Hooonx
aaoa>ous .uuoe sass» soon sums one; sumznesz gums sauna son-5a changes
we a no w

 

.aewsuoeouea he aehsaeeu so» assovxseun amass ease asheseo .N saber

33

 

 

 

 

 

 

 

 

Nm Hm mg as ma I N Hope“
He mm m me ha av H Hepem
be easasom
me m I NN am I m ~o>oa
we mN N we on I N ~o>e~
NN mm 6H an vN Hv H Moped
no essaoON
on can I mN Mb I we ossusON
av NV ca hm an I N Ho>ou
Ne hm NH an we N H HoboH
mv oususoh
an be m on vN I N Ho>e~
MN MN ha we mm I H depea
we a on
INN ensueoh
sunsuev clause ousfinlonas .hususeo hssuooo hususoo husvooo .oo Ho>on\
asow>oaa .usom Haven auoN numu eusa Acadia“: some Susan son-so essusON
ho a me &.

 

Avosouaooov .N o~noe

34

HsHueusl ensues» hHHsssHsa aHeooH sosoH a

 

 

 

«a 8 a so 3 I N :33
NH am e we Hm I H Ho>eH
«e zooHn
HN 5N N 5N om H N Ho>oH
mN 5N e we we H N He>eH
oN wN v we mN H H HoooH
auHos osoausoe
433
NN NH Hv mm «H I N Ho>oH
He He N we HN I N HepeH
be 5N HH Nb NH Hv H HoooH
euHe: soonusos
IN sooHN
m N I no 5N NH N HoooH
mN wN N ov mm N N HoooH
Ne Nb e we be H H Ho>oH
N xooHn
eN N N NN Oh I N HobeH
NN NH H NH NN H N He>oH
be new wN aNN new wH H Ho>oH
H zooHN
oHAeusv slsHoo Hsuou Assuseo husuooo hususeo ansuooo .oo He>eH\3
esoH>osa uHo: so xooHA nuoN numH ousH auNHIvH: nuNH hHusN sea-so
ya u me a uHss no xooHn

 

.aewsuneouea an auosMHuus easusemlsoe sou essovmsesa emssu ease

Hsuesee .N eHass

35

nfldhfluflfl OHSOUON

NHHuulHum oHo§oH uoonu

 

 

 

 

 

 

He 6N N NN NN I N Ho>oH
NN NH NN I vb I H Ho>oH
5N uHaa
Hv 0N I NN on I N Ho>oH
NN 0N N am NN I H Ho>oH
NN anD
NN No Hv NH HN I N Ho:oH
NV Nv N me we I H Ho>oH
NN van:
Hv NN H Nv pm I N Ho>oH
NN 5N oN NN vH I H Ho>oH
up sooHN
NN No N vN Nm I N Ho>oH
NN 5v vH Nv bN I H Ho>oH
N xuoHN
NH ON N NN vN I N Ho>oH
NH Ne N 5N vN Hv N Ho>oH
HN NN vH NN NN I H Ho>oH
n zooHN
oHaounv alsHoo Haaru Nunanoo husuaoo Ahaaaoo husuaoo .on Ho>oH\3
usoH>Ohn «Has no xooHa AuoN AaNH ouag AuNHIvH: AaNH hHuaN hon-:9
no a No a awn: uo aooHN

 

Avonnfiunoov

.N oHaua

HaHuoual ouaunom NHNNGIHNA nHo»oH LosoHu

 

36

 

 

NN ooN mN on 0N I N No>oN
NNIMNNN

mN Nc I on ON I N No»oN

NN an N Np «N I N No>oN
wmeNmm

. «N NN NN no NN I N No,oN

NN NN N NN an I N No>oN
mmImwmm

an m I NN Np I N No>oN

«N am NV Ne on N N No»oN

NN v mv e on I N No»oN
mm «Nap

«N on NN NN em I N No»oN

NN NN N on NN I N No>oN
NNIMNmm

mm mm NV NN N» I N No,oN

mN m I NN No I N NopoN
NNIMNMN

NN N I ocN I I N No»oN

NN NN I NN no I N No»oN

NN on N Na NN NV N No>oN
.NmImmmm
Odflludv BISHOO AIUOH hh’ufloo fihN—uflflo Mann—“BOO hhq—an-OO .Ofl ~O§OH\3

aaoNpoua “Na: no NuoNN NNNN NNNN ouuN NNNNINN: NNNN NNuaN hon-:2
No a No a ”Na: Na NuoNn

 

H602: anoov . N CHAIR

37

to the house. For example, a 2 x 2 I unit which began at
2 I frOI the house and extended up to 4 I away was
designated by the 2 I interval instead of the 4 I.

flhen the functional data was plotted it was found
that several categories appeared to be non-randOI in their
distribution in relation to the house. Of these, ceraIic
tableware exhibited the most obvious positive relation to
the house.

Further-ore, when the backyard units were c0Ipared to
the frontyard units, a pattern eIerged where a high
frequency was found near the house in the backyard with
frequency drapping off as distance increased; while in the
frontyard there was a low frequency near the house with an
increase in frequency as distance increased. This
dichotOIy between back and frontyard holds true for
virtually all of the plotted data even though other
functional categories, such as utilitarian ceraIics and
glass bottles, appeared to be sore randOIly distributed
overall.

A possible explanation for the difference in
distribution in ceraIic tablewares and utilitarian wares
lies in the fact that the forIer are ”generally used and
highly curated inside dwellings,” while the latter are
”used in a Iuch larger sphere that included porches, yard
areas, and outbuildings, and were curated to a such lesser

degree” (Hair 1982: 147).

I. .
' ?
I
I
o .I
t

0',

38

On the other hand, owing to technological
inroveIents in the glass industry, bottles becaIe Iore
available and lower in cost resulting in less reuse and
greater discard in the late nineteenth and twentieth
centuries (Hair 1982: 148). Therefore, higher frequencies
and perhaps a wider distribution would be expected during
these tiIe periods.

FrOI a visual inspection of this data it was
concluded that the artifact asserlage could be ”cleaned"
of certain variables that appeared to be randosly
distributed and therefore of little value to this
analysis. These variables included architectural iteIs
(i.e., nails, window glass, brick, etc.) which are better
indicators of structural Iodifications than of refuse
patterning, and bottle glass which was found in high
frequencies over Iuch of the site but in no definable
distance-related pattern. Although in general, the latter
did repeat the frontyard/backyard distinction described
above.

The second step in this portion of the analysis
involved the coaputation of artifact densities per I2 for
the functional and date range categories to be considered.
To accOIplish this the actual nuIber of artifacts under
consideration fros a given excavated area was divided by
the square Ieter areal extent of said area. Average

densities were utilized in order to reduce variability,

I ... I.
x L
J 0’
. I
I ' I .1 ‘1
’i |
V
I I I...
I .
s
V id
.
I . ’ C. .
- s I
.I I] ..
I
I t I . ..I . .I
(I
' s
I II
. .I .
II - ‘4‘
I v I
III.
I.
V t I
\I
.s I , I
I
I1 -
v
at

\I

39

saIple size, and the bias inosed by the distribution of
the excavation sample.

Tables 4 and 5 present the artifact densities for
feature and non-feature asserlages by date ranges. The
data set utilized for these tables did g2; include Iodern
Iachine cut nails and wire nails as these iteIs appeared
to be skewing the date range inforIation. It is felt that
the resulting data is a better reflection of the actual
date range of occupation. It is further noted that the
early nineteenth century category, which is priIarily
early Iachine cut and handwrought nails, is best
considered part of the lid-nineteenth century occupation.

In order to investigate the date range distribution
on this site, for the purpose of testing the stated
hypotheses (see Chapter 1), the datable asserlage (by
artifact densities), Iinus nails and bottle glass, was
applied to contingency tables; wherein, the date ranges of
aid-nineteenth, late nineteenth, and twentieth centuries
c0Iprised the row of attributes, and distance intervals in
Ieters froa the house were the coluInar attributes.

Initially, contingency tables were set up c0Iparing
date range to distance in the backyard as one table, with
the frontyard as a separate table. Tables were then
forIulated with the backyard and frontyard coaprising the
row of attributes and c0Iparable distance intervals as the

coluInar attributes. Separate tables were calculated by

4o

auosuuahs NNN.oH n AdhOh

.suasoo uosuHahs Hssaus suoosa Hay
.asoHuslmxounns she suns-shsassl usual shssom we
.aHHsu suHs vos sHHss use somaosl ossvol svsHosN uos ov senses sue: a

 

 

Name NNNNN NoNN NNV
oN.NNo.H «IN NNH.N NNN om.b¢ ON.NN No.0 cm.° hm
Ncmvv Nanny Nev Nev
NN.HNN «INH NNN.N bNN.H NN.NN oN.NN NN.o NN.o Nm
Nov
oo.NN NIH NN N oo.N Ne
Nva va ANN NNV
NN.NN use meN 5N Np.e om.H NN.o NN.N Ne
NmNN NNNV
HN.VN Nab va NN NN.N NN.H mv.cNIbN
NNm NmeNv .NNN Nee
5N.ONN NsNH NNm.e oNN N.> ON.NH n>.o NN.o HN
ANNV ANHV
NN.HN NINH mNN 5N NN.H NN.o N
NoNV va
co.NmN NIH NNN NH NH N e
ANN NeNV Nev NNV
H>.HN «an NNN He NN.o Nv.N H vH.H oH d N
Nev NNNN NeNv
cm.vm NINN NNH.H eoH NN.o bh.N «N.o H
«I\>uNsoov «I Hsuou coast .0 .onuNH .oAuNH .onuNH e
oussosc .ussu Hsuop nuoN ease INN: hHssN onsusom
was. was buHsasv sushsss ha nsswsss susv ensues: .v ansh

41

aussoo ausuHuus Hssuus souosov an

b zooHN use up .eb .bN .NN .eN sows: ovsHunN no: ssov «a
sHNss saw: was use seasons ssovol ovsHuoN aos soot a

 

 

NNNN NNNV NNV
«Ne mN ON.N ON.N oN.o I NaoN e NuoNN
saws:
NNNNN NceeNv NNNV NNNN unmeaso.
«NN.NN mmN.N sm.¢N hm.mN om.N Nm.o «new I N NooNN
sumo:
NNNNN NNNNN NNNN NNV atoastoa
NNN.NN om¢.N No.mN NN.NN NN.° Nc.o NamN IN NuoNN
NcNNV Nmmmv Nome NoNN
oNN.m one N¢.m NN.NN om.N Ne.o N-NN N NooNN
Nova NNNNNN NNNV Nae
Noo.NN mmN.N Nm.N NN.NN No.N oN.o «INN N NuoNN
.NoaNNNtu uusuv .u .oNNNN .UNNNN .oNNmN N. .02 NNga
.cz NaNoN . Nauoa NNoN «NaN INN: NNtum \NooNN

 

«I use hamassv sususps ha nssusHAIssas assassmlaos sou «sensed sue: .m sHash

42

sussoo «osNHuus Hssuus asuooov Nov

5 sooHN can up .vb .bN .NN

.vN saws: ovsHoaN nos soov «a

usss suHs was use osflnosl shovel ovsHosm uos soov s

 

 

 

NoNN NNNN NNN
New as om.s us.m I mN.o use an NNaa
AoHv Hva Nev
men No om.N mN.NN oo.N I «me so NNna
AHNV ANN
«NN mN I oN.m mN.N I «IN mm NNna
ANN Ava HHV
Nmo.N on om.N as.oN mN.o I use mm NNaa
ANNV ANNV
Now we oN.N an.N I I «soN m NooNN
NNNNN NoNNV NNV NNV
opm.m new No.m ms.N oN.o No.o «INN m NuoNN
.NoaNNuta sauce .0 .oNNmN .osNNN .oaamN «a .oz NNaa
.oz NaNoN Nouns NuoN oNaN IvN: sNuan \NuoNN
NuosaNNaooN .m oNnaa

season nosNHahs Hs=uos souoaov Adv

h zooHN was as .eb .NN .NN .VN saws: ovsHosH uos soov an
usss sum: was 950 seasons shovel ovsHosH cos ssov N

 

43

 

 

moN.NN NNN.N u NNNNoN
NNN NNN
NN N om.o oo.N I I N-N NN NNaa
NeNN NNNV
Nee No cc.» oo.eN I I «IN NN «Nap
NNNN Nose Noe
NNN.N NON oo.s mN.NN om.N I use mm NNaa
NNN NNNN NNNV NNN
NNN NNN oN.N NN.NN NN.N om.o «so am NNaa
Nssv Nsev NNN
New NNN NN.NN NN.NN ms.o I Nae NN NNaa
NNN Nave Nee
Noe mm om.o NN.NN oo.N I use om NNaN
.uoaNNsna coast .0 .oNNNN .oNNNN .oNNNN Nu .oz NNaa
.oz NuNoe_ Noses ANNN ouuN INN: NNuuN \NooNN
NeoaaNuaouv .o oNs-a

44

lid-nineteenth, late nineteenth, and twentieth centuries.
Such variables were selected with the idea that the
changing function of this housesite (i.e., town residence
to farIstead) resulted in changing patterns of disposal
specifically related to distance fros the house.

In order to test the significance of these
hypothesized relationships, a chi-square (x2) test was
calculated for each contingency table. This is a non-
paraIetric, statistical significance test which is

calculated by the following forIula: x2 = {0-823,
E

where 0 ' observed frequencies (in this case, average

densities per I3), and E = expected frequencies (Doran and
Hodson 1975: 54—55). It is noted that this statistic
indicates only whether or not the proposed variables are
independent in association (i.e., knowing the value of one
is no indication of the value of the other). It does 22;
indicate the strength of an association (Ibid.).
Therefore, CraIer’s V2, a Ieasure of association, was
applied to those tables which exhibited dependent
variables according to the chi-square test. This forIula

is as follows: V2 = x2 ,
Nmin(r-1,c-1)

where the chi-square is divided by the nuIber of units
tiIes the sIaller value of (rows - l) g; (coluIns - l)
(Doran and Hodson 1975: 147). This forIula is utilized

here with a cautionary note in that "the value of this

45

Ieasure is doubtful: it adIits no clear interpretation for
the values between 0 and l” (Doran and Hodson 1975: 148).
It is used in this analysis Ierely to show the possible
strength or weakness of an association.

To suppleIent this analysis, distribution Iaps of
actual artifact counts by date range were cOIpiled in
order to clarify, by visual presentation, the possible
relationships and patterns discerned by other Iethods.
These Iaps serve to illustrate the saIple distribution,

size, and liIitations.

Chapter 3

TB! DATA

Archival

Data frOI the archival research concerning Cedar Oaks
revealed that this "residence was probably built for Dr.
JaIes E. Curtis, a physician in the earlier town of
Colbert who loved to Barton shortly after the great flood
of Decerer 1847 inundated Colbert" (Minnerly 1983: 21).
As Curtis only paid $61.00 for this property, it is
extreIely doubtful that a structure was present on this
land at the tile of purchase around 1848 or 1849 (Ibid.).

However by 1851, when Curtis sold this property to
Miles Johnson, the price (i.e., $1,039) was "high enough
to suggest that the house was already built by that date"
(Minnerly 1983: 21). FurtherIore, "the orientations of
the house and yard axes,... generally correspond to the
grid of Elliott’s hypothetical plat [Figure 2], suggesting
that Cedar Oaks was not built before the platting of
Barton in 1848” (Ibid.).

The original owner, JaIes Curtis, along with his
brother John, entered into a Iercantile business in Barton
in 1848 or 1849 but the venture quickly failed. This
proIpted JaIes to sell Cedar Oaks in 1851 and Iove to

ColuIbus where he opened an inn (Minnerly 1983: 21).

46

.‘. 'I I'IIJHIIJ

Pl!“ 3??

Ievldnwh

t
7 I ' 4 l
I
I
I
I'I
I ~ \
1 II ‘ n
I l . ‘ ' I ‘
I f' . I
‘ 1. V V
i ‘ 'I I
.
\ ‘ I
‘ l ...: . I. ... 4
I Q‘
I ‘ I
.‘ ‘ l I N 6 s‘
I t I " ll 1 '

47

Miles Johnson was also a merchant in Barton; however,
he held title to Cedar Oaks for only one Ionth before he
sold it to James M. Collins, making only a $15.00 profit
in the process. FroI all indications, Johnson never lived
at Cedar Oaks; rather, he resided in the Barton Hotel
(Ibid.).

JaIes Collins was the Iain resident of Cedar Oaks
during Barton’s existence as a town. He and his faIily
occupied the house frOI 1852 to 1859, during which time be
operated the "principal Iercantile operation in the Barton
and Vinton area” (Minnerly 1983: 21).

When Collins opened his Barton store in 1852 he was
acting as a "branch officer for the proIinent ColuIbus
firI of Cozart, Billups and Hquhries” (Ibid.). His store
carried a large stock of general merchandise and was
”doing the best business in the area ($8,000-$10,000 per
annuI), excepting his principal c0Ipetitor at Vinton,
WilliaI E. Trotter" (Minnerly 1983: 21-22). In addition,
he also owned an interest in a warehouse and cotton shed
in Barton (Cleland and McBride 1983:27-8).

JaIes Collins was a man of considerable Ieans.

Before his arrival in Barton he presided over a large

plantation on Town Creek, about four Iiles northwest

of Barton. According to the 1850 census, he owned 20

slaves (Cleland and McBride 1983:37).

Due to the failure of navigation on the Tombigbee River in

1855-56, the per annuI sales from Collins’ store dropped

to about $7,000 indicating that "the cotton trade

48

constituted a significant portion of his business"
(Ibid.).

The Barton economy was based priIarily on the cotton
trade. SteaIboat landings at Barton and Vinton served as
focal points for the shipIent of the region’s cotton
downriver to Mobile. However, the upper Tonbigbee River
was only reliably navigable during high water,
approinately five Ionths out of the year (Doster and
Weaver 1981: 69). Warehouses near these landings, such as
that owned by Collins and his partners, protected the
cotton while it awaited shipIent (Doster and Weaver
1981:78).

With the extension of the Mobile and Ohio Railroad
into this area, the seasonality of the river trade gave
way to the ”all-season,” "all-weather" transport provided
by the railroad (Doster and Weaver 1981: 98, 102). In
fact, it was the extension of this railroad to West Point
in 1857, bypassing the Barton-Vinton area, that
effectively destroyed the viability of both towns.

By 1860, the overland feeder routes into the

cOIIunity [i.e., Barton-Vinton] had been diverted to

the railhead at West Point. Of the five principal

Iercantile establishIents operating in Barton

throughout the Iid-1850s, only one, possibly two,

reIained after 1860... By 1865, the drastic econonic
constriction of the Civil War finished the process '
begun in 1857 with the arrival of the railroad.

Occupied by only a few scattered inhabitants, Barton

was nearly abandoned with nothing but its ferry to

Justify its continued existence (Cleland and McBride
1983: 27).

49

In 1858, Collins saved his business to West Point
where it prospered until the Civil War disrupted the
econOIy (Cleland and McBride 1983: 37). He did not sell
Cedar Oaks until 1867 when the title was transferred to R.
J. Conner for $640.00 (Cleland and McBride 1983: 86). It
is not clearly known whether the house was occupied
between 1859 and 1867--it say have stood vacant or it
could have been leased at tiIes. Its history is obscure
during this period due to the rapid decline of Barton
after 1860 when, as a result, deed records changed froI
town blocks and lots to general land descriptions. The
ensuing civil and econoaic confusion resulting froI the
Civil War did not help to clarify land transactions in
this area (Minnerly 1983: 23).

Although H. J. Conner and his wife possibly resided
at Cedar Oaks for a tiIe, it appears that Sarah and
Bardine Richardson owned this property in the early 1870s,
along with such of the former Barton land. However, it is
not known whether they actually occupied Cedar Oaks.
Bardine Richardson had owned the Barton Perry in the
1850s. According to the 1870 census, he was by that tiIe
engaged in farsing, reflecting "Barton’s shift froI a
c0IIercia1 center to a rural agricultural coIIunity”
(Cleland and McBride 1983: 74).

Between 1875 and 1879, Mary B. Coltrane bought Iost

of the land surrounding, and including, Cedar Oaks

50
(Minnerly 1983: 23). By the 1880s, the Coltranes (i.e.,
Mary, husband WilliaI, and faIily) had acquired Iost of
the property that had once been Colbert and Barton, as
well as the southern portion of Vinton (Cleland and
McBride 1983: 87).

The probleIs the Coltranes faced in keeping their

property during this period of econoIic stagnation

and readjustsent were typical for Iany residents...

a large percentage of the land transactions during

the last quarter of the nineteenth century were

Iortgages. The Coltranes reIained well-established

and respected residents... In 1886, [they] were taxed

for seven cattle, three Iules, two carriages, one
piano, and one watch, holdings exceeding those of

Iany residents in the area (Ibid.).

WilliaI Coltrane was listed as a farIer in the 1880
and 1900 censuses, and he taught school in Vinton between
1874 and 1896. It is known that the Coltranes resided at
Cedar Oaks for Iany years, although the exact dates are
uncertain (Minnerly 1983: 23). Oral inforIants recognize
a Coltrane-Cedar Oaks association; however, the faIily is
Iore strongly associated in later years with a residence
on the bluff north of the Barton Ferry (Minnerly 1983:
28).

Oral sources further indicate that the WilliaI Sidney
Foote faIily say have rented Cedar Oaks froI the Coltranes
during the 1890s and/or early 1900s (Ibid.). In fact, the
1909 photograph Ientioned previously shows the Foote
faIily posing in front of the house (Minnerly 1983: 30).
Informants state that WilliaI Foote Iade his living farI-

ing the land to the west of the house (Minnerly 1983: 29).

51

Dr. Jan Uithoven purchased the property containing
Cedar Oaks from the Coltranes in 1913 but he Iay not have
lived in the house until 1919 or 1920 (Minnerly 1983: 23).
Initially, the Uithovens lived closer to the Barton Ferry
which Jan operated ca. 1917-1920 (Minnerly 1983: 29,31).
While residing at Cedar Oaks, Jan Uithoven made his living
by farming, raising sheep and goats, selling illicit
whiskey, and practicing medicine (Minnerly 1983: 31).
After his death in the late 19209, his daughter Frances
and her husband Charlie Rhea occupied the house and farmed
the land surrounding Cedar Oaks until the late 1930s
(Ibid.).

An oral informant recalled visiting the house in the
1940s and found it abandoned and in need of repair
(Minnerly 1983: 31). The house remained in the Uithoven
faIily’s possession until its recent purchase by the Corps
of Engineers for the construction of the Barton Ferry
Recreation Area. It appears that the house has not been

occupied since the Rhea’s vacated in the 1940s (Ibid.).

Oral History

One application of the oral historical data obtained
during this investigation was to fill in sOIe of the gaps,
encountered in the archival record as noted above. Oral
sources also provided valuable inforaation concerning the

location and function of for-er outbuildings, the material

52
culture and customs of the former occupants, and general
insights into the lifeways of the late nineteenth and
early twentieth centuries. As previously Ientioned, oral
informants held IeIories of Cedar Oaks that dated back to
the Coltrane family.

Oral sources further provided important information
concerning changes in the architecture of the house. Most
remerered an ell on the northeast corner of the house
which functioned as a kitchen, although none could recall
when it was built. Earliest recollections state that
there was no inside access to this ell, although a
connecting door was added later on which still survives
(Minnerly 1983: 31).

InforIants recalled fireplaces in each of the four
main rooIs with the all kitchen having a flue for a wood-
stove. The chimneys were described as handmade brick with
lime and sand mortar. The last original chimney fell off
the house in the 1940s and is evidenced archaeologically
by Feature 6 (Figure 7). This chiIney fall was not
salvaged unlike the chiIneys on the northern side of the
house (Minnerly 1983: 31, 68, 94, 95).

Two porches, one running the entire length of the
house front and one along the southern side of the kitchen
ell, were also recalled by informants. The latter porch
was reIoved in the 1940s when the ell was reIodeled

(Minnerly 1983: 31). This porch is evidenced

S3

sowsusom no use .nxso woven

 

 

8s.
4!

IP

‘81}.

 

 

 

 

 

 

I.
III!"

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Ill-ll. o s

8.3: Ills-lull!
ll Ills-II...
__ : .o . u Ills-ails:
so.

 

 

54
archaeologically by Features 9, 11, and 12, which are
postholes located in the central portion of Block 3
(Minnerly 1983: 71-73) (Figure 7).

A smokehouse, which was located archaeologically as
Feature 1 (Figure 8), was also recalled by some inforIants
(Minnerly 1983: 29-32 and 61-65). It was described as a
log structure which functioned as a smokehouse and food
storage shed. It was dismantled or fell down sometime in
the early twentieth century and the majority of the
construction material was removed or salvaged (Minnerly
1983: 63, McClurken and Anderson 1981: 561, 617, 619, 877,
882).

Other oral information includes the use of an
artesian well, located to the west of the house, as the
primary water source during the twentieth century
occupations. According to inforIants the brick-lined well
was not used during this time period (McClurken and
Anderson 1981: 562, 619, 883-884). It was therefore
concluded that this well represents a late nineteenth
century water source. After it was no longer used it was
merely abandoned and, like other late nineteenth century
brick-lined wells on the Barton townsite, was not used as
a trash receptacle (Cleland and McBride 1983: 373-374).

Further details were provided by oral sources and
these will be discussed later on in conjunction with

corroborating archaeological data. Some oral information

55

 

 

soc-assay

 

 

 

0".“ I l

Duo-amt.
mass-III».

Hun-«ssh.

 

ssssmcfi

”0-...”—
0

 

closes-n MIT!

 

 

 

 

 

Figure 8. Feature 1 (brick foundation)

56

was very detailed as to the types of dishes used, the
objects stored in the smokehouse, as well as refuse
disposal practices. It should be noted that all
informants remembered the house in its present four-room,
central hallway configuration.

In summary, the archival and oral information for
Cedar Oaks suggests a construction date in the late 1840s
with the occupants involved in Barton mercantile
operations through the 1850s. Occupancy became more
sporadic in the 18603 as Barton declined, with the house
possibly vacant for periods of time. After the 1870s, it
was occupied off and on by families engaged primarily in
farming with the house abandoned as a residence sOIetime

during the 19403.

Architectural

Three separate architectural studies were undertaken
at Cedar Oaks, and although all agree on certain points
there are also inortant differences. The first study was
completed in 1978 for the Historic American Buildings
Survey (Howard 1978). In this report a construction date
was postulated as sometime during the 1840s to 1850s
(Howard 1978: l).

Howard’s (1978) analysis of the structure of the
house led to speculation that the original floor plan may

have consisted of only the two northern roons. It was

57
felt that the northwestern Iantel piece, a board and
batten door in the same room, and the fact that the front
door is slightly off center gave credence to this original
configuration, especially as the mantel piece and door are
older in appearance than their counterparts in the
southern portion of the house (Howard 1978: 3).

Noted alterations and additions included the fact
that this house, in its present forI, once had four end
chimneys and an ell addition on the northeast corner,
allegedly removed around 1920 and the material used to
build the standing shed. It was further noted by Howard
(1978: 4) that the roof, most of the siding on the east
side, the flooring of the front porch, and two of the
front porch pillars are all recent replacements.

The inforIation concerning the all came from an
interview Howard (1978: 3) had with Felix Uithoven, son of
Dr. Jan Uithoven, and conflicts somewhat with other
informants which have the ell being remodeled in the 1940s
(Minnerly 1983: 31). Unfortunately, the Tombigbee
Historic Townsites Project was unable to interview Felix
Uithoven as he was, at that time, engaged in litigation
with the Corps of Engineers over the sale of Cedar Oaks
and was unwilling to speak with our investigators.

The second architectural study was made by Dr. M. B.'
Newton, Jr. in 1980 while excavations at Cedar Oaks were

underway. He concluded that this house was built sometime

58
between 1835 and 1850; however, because of the "earliness
of the feel of sale of the elements (mantel, joinery,
chair rail),” and allowing for ”the navigability of the
Tombigbee, plus the dynamisI of frontier regions," he
emphasized the earlier date (Newton 1980: 7).

Further-ore, Newton (1980: 1) concluded that the
house was built "where it now stands in the form that it
now has.” He felt that the house was built as a single,
integral unit owing to the principal sill supports which
he saw as ”fashioned in such a way that [its] lap joints
fall at points that do not Iatch any of the room
divisions” (Newton 1980: 2). Therefore, he claimed that
any renovation would have ”required disassembly of the
entire house so as to enable reassembly of the Iortise-
and-tenon joints of the studs where they join the sill”
(Newton 1980: 2-3).

Newton (1980: 6) also made suggestions which proved
very helpful in the interpretation of the archaeological
data. In particular, he suspected that the central back
door would not have opened to the exterior, but instead
onto a rear gallery or less-than-full length porch.
Features 14, 16, 17, and 23 are all postholes that are in
a north-south line parallel to the house, suggesting a
rear full-length gallery. Feature 21, a refuse deposit on
the southeast portion of the house (Figure 7), represents

refuse swept and/or dqued underneath such a structure.

59

The most recent architectural study was undertaken by
an historical architect, Heriberto J. Brito, in 1983. He
takes exception to both of the previous studies’ conclu-
sions concerning the original configuration of the house
and argues that the present floor plan of Cedar Oaks
evolved froI a simple two-room plan that consisted of two
equal size rooms, one north and one south, that say have
had porches on the eastern and western sides (Brito 1983:
4, Figure 4).

Brito (1983: 4) reached this conclusion because of a
cutline he detected in the floor of the southwest room, 12
inches fro- the south wall. This, coupled with
differences in the placement of the chair rail throughout
the house, the fact that the floorbeams are spliced near
the cutline, and the differences in placement of the brick
piers on the southern versus northern part of the house,
led hiI to argue strongly for such an evolution.

Unfortunately, no definite archaeological evidence
was found to corroborate this theory. Those units which
extended underneath the house on the east side revealed
only recent disturbances and features related to the house
in its present state. This includes: two driplines (one
corresponding to the present Ietal roof and another 10 cmv
in corresponding to the previous shingle roof), the older
brick pier bases (Features 22, 24, and 25) which have more

recently constructed brick piers on top of them, and

II .
‘ ‘
r
I I
.n I.
I I I
|C I 1. O
. .. .
. I . o
I
. a. .
- A A .'
I I. I
I I .I
I I
.I
I‘ U.) I
I I .
1
II .
’ I U
. x .
If
‘ -
I .I
I I l
. I
I I. I
.
. \
.
I I
A
K II
I I \
I
.

60
Feature 47, a wooden pier found next to, and in line with,
the second brick pier off the southeast corner of the
house (Figure 7). It is highly probable that evidence of
an earlier house for. was destroyed in this area by later
additions and isprovesents to the present house.

Not having had the benefit of Brito’s (1983) study
during the archaeological investigation, and being
influenced by Newton’s (1980) conclusions; the
plausibility of Brito’s version of the original floor plan
was not tested archaeologically. The fact that two out of
three studies concluded that the house evolved fros a two-
roon plan, albeit not agreeing on its orientation, would
indicate a higher probability of this house not being in
its original forn.

In sunsary, these three studies placed the
construction of Cedar Oaks between 1835 and 1850, with
Newton (1980) leaning towards the earlier date and the
others to the later one (Howard 1978, Brito 1983). while
two of the studies (Ibid.) concluded that the present
house had evolved fro- a sispler two-roos plan, the other
study (Newton 1980) argued that the house was in its sore-
or-less original state. All agreed that the house dated
fro. the Barton period and was historically significant as
the last remaining standing structure fros this town.
Brito (1983) especially reconnended that the building

possessed architectural and historic significance and

61
should be preserved. At present, the house is still

standing but its fate remains uncertain.

Archaeological--srchitectursl integrity

The results from the archaeological investigation
provided further details on the architectural evolution
and integrity of this house. Since no evidence of the
postulated original two-room floor plan was uncovered, the
following discussion will pertain only to the present
four-room configuration.

The front of the house (west side) has changed only
superficially, with the gallery having been refurbished
recently with new floorboards and two replacement pillars
which flank the front door. Archaeological and oral
infornation show that the present wooden steps are recent
replacesents (Howard 1978: 4, McClurken and Anderson 1981:
554-637). Postholes (Features 40, 50, and 59) for
previous steps were uncovered underneath the present ones
(Figure 7).

Excavations in the frontyard (Block 5) revealed a
brick walkway (Feature 27) extending westward from the
front steps towards the opening in the front fenceline
(Figure 7). Parallel to this walkway is a for-a1 garden
enclosure (Features 28-30, and 45, Figure 7) which
consisted of two parallel one-course high brick walls

around two circular brick flower beds.

62

The sanner and saterials of construction, as well as
associated artifacts, show that the walkway and garden
enclosure were contesporaneous and in existence in the
late nineteenth and early twentieth centuries.
Proportional data froa these features show that the
sajority of datable artifacts range fro- 1860 to 1900
(Table 2) (Cleland and McBride 1983: 331-332).
Further-ore, one oral inforsant recalls this walkway in
use during the early years of the twentieth century
(McClurken and Anderson 1981: 882), and when Iodern
aachine cut and wire nails are reaoved fro. the datable
asselblage (Table 4) this date range is reinforced.

The sides of the house revealed structural evidence
of the original chisneys. Features 6, 48, and 49 were the
southeast, southwest, and northwest chiIneys, respectively
(Figure 7). The southern chisneys were evidenced by large
fall and rubble reaains; however, the northwestern chisney
was represented only by a rectangular stain where its base
had been set. No clear indication of the northeastern
chisney was uncovered, although its existence is attested
to by architectural evidence within the house (Howard
1978: 3, Newton 1980: 9).

Since no evidence of previous chiIneys or structures_
was discovered in these areas, Features 6, 48, and 49 are
concluded to be the original chiIneys. The bricks and

sortar of Features 6 and 48 are identical, and brick

63
rubble recovered fro. a refuse du.p (Feature 53) on the
northern slope is also identical in co.position. This
rubble is believed to be the bricks fro. the northern
chi.neys (Cleland and McBride 1983: 334).

Excavations at the back of the house (Block 3)
revealed nu.erous features indicating greater
architectural changes in this area (Figure 7). All,
however, appear to be evolutionary changes related to this
particular structure and not indications of an earlier
house.

Perhaps .ost significant to the integrity of this
structure is Feature 47, the wooden pier discussed
previously (Figure 7). This pier suggests that the house
was once supported (in part, or wholly) by large wooden
piers which, upon decay, were replaced by brick piers.

The use of a variety of different pier .aterials on the
sa.e structure is not an unheard-of—practice. For
exalple, a 1913 double pen house located in the Bay
springs area of northeastern Mississippi ”sat upon various
types of piers including log stu.ps, brick and concrete”
(S.ith et a1. 1982: 66).

This particular pier was replaced so.eti.e in the
.id-to-late nineteenth century but no later, as the refuse
of Feature 21 entirely covered this pier re.nant. Table 2
shows that the .ajority of datable .aterial fro. this

feature, on a proportional scale, co.es fro. the

64
.id-nineteenth century with the next highest proportion
fro. the late nineteenth century. It is noted that this
is skewed by the inclusion of .achine cut and wire nails
in the datable assemblage. He.ova1 of these items shows
greater densities in the late nineteenth and twentieth
century date ranges (Table 4). Moreover, the .ixture of
dates fro. early nineteenth to the twentieth century
indicates a great deal of disturbance in this area
(Tables 2 and 3) and, as the .ajority of artifacts are
architectural in function (Table 6), there is a strong
indication of re.odeling having occurred.

Three structural additions to the back of this house
were evidenced by Features 14, 16, 17, and 23 which were
all in a north-south line parallel to, and approxi.ately 2
. away fro., the house and appear to indicate a rear full-
length gallery (Minnerly 1983: 50).

Features 8, l5, and possibly 14 and 16 .ay have
supported the ell that was known to have been attached to
the northeast corner of the house during the very late
nineteenth or early twentieth century (McClurken and
Anderson 1981: 561-562, 877, 880). Features 9, 11, and 12
represent a porch running east-west that would have been
attached to this ell extending out fro. the central
folding door (McClurken and Anderson 1981: 562, 880).
Archaeological and oral data indicate that the ell and

east-west porch postdate the rear gallery.

65

esoHusstosans ass sass-sasasel nous. cusses we
auosmHvus sususouase svsHosH no: sees a

 

NQN.¢H undue“
«Hm canyon NNHN NN N e NN H NH NH H N «IN hm
alsv seamen wNNNm HH N N NN H NH NN H NH «INH Nm
hoanao N@ m I I Nb I I NH I N «IH me
aoslHno mwN NH H N hm e N NH H N «ac we
suseoHoso va eH m N cm H H h H N «lb we
nevus. Henson .ONIbN
his. seamen meow NH N N mm H v NH H e «INH HN
NoslHao mNN NH H N N¢ H e mN e N «INH m
@000. nouom NmN NH H N hm H I HH H m «IH v
z : NHe e v N Nb H N N H v : oH
v0£a\:050an omH N N H up N I b I m «lb N
oasoaoxolm NNHH er xcH xH «NV I I wNN RN as «INN H
3.4 9TH N .d H V 3 an on .43 .....V d
m... 3.. m. m m x mm m m 3 E J m
as ....v. 3 u D. UI SW 3. WT. Tm m? 3.
3.... .....l a p. M I. o s T. t. a Z n
....c. e T. I e 1 UT. 3 a 1.3 At. a 1
CT. on T. 1 3 D.“ n 39. no ¥X a
E. I W a n I... u w ...... .... A...
I. 8m a n ....n. e e A N
8 O 1 .4 1 S u 9 9 m.
D. 1 n 9 8 a S T. 1v
3 S T. m T... D. 9 3
n O S p. S D. 1
1 3 1. s
. m

 

.sswsuseoaes ha soususeu veuoeHes you sausuHuus no stovxseun HssoHuoss.

.m sHAsh

66

Feature 4 (Figure 7) is evidence of a brick stoop in
front of the central back door. This was also confir.ed
by oral infor.ation (Howard 1978: 3) and was probably in
existence at the turn of the century after the rear
gallery had been re.oved and before the east-west porch
was built.

The only driplines uncovered were related to the
present house for.. Various other postholes found in the
vicinity of the house, front and back (i.e., Features 18,
33, 35, 41, 42, 46, 54, 55, and subfeature 1 of Feature
21) (Figure 7), were too saall to have supported a
structure of any great size. It is likely that they
represent fence enclosures associated with the front
garden, and porch or addition supports off the back and
north sides of the house. However, no distinct patterning
to their locations could be discerned (Cleland and McBride
1983: 331, Minnerly 1983: 53-54).

The integration of oral, archival, architectural, and
archaeological infor.ation strongly suggests that this
house is original to this site. Even though it .ay have
undergone a transfor.ation fro. a two-roo. to a four-roo.
plan, it retains its historic significance and integrity
as the last renaining standing structure fro. Barton.

The artifacts do range fro. the lid-nineteenth

century (Tables 2 and 3) when Barton was first platted,

67
and a construction date around this sa.e ti.e period is
strongly indicated. The low a.ount of early nineteenth
century .aterials (i.e., prisarily nails) would argue
against an earlier construction date, as would archival
data and so.e architectural opinions. The presence of
handwrought and early .achine-headed cut nails (1815-30,
Nelson 1968) is, perhaps, .ore a result of a lag in the
diffusion of new innovations into a rural area (especially
one located on a river which was navigable only on a
seasonal basis), than of an early nineteenth century
construction date and occupation. It is also conceivable
that their presence is indicative of the salvaging of

construction .aterials fro. the earlier town of Colbert.

Archaeological-~site utilisation

The data collected concerning the utilisation of this
site through ti.e was gathered prilarily fro. oral and
archaeological sources. Two for-er outbuildings were
located during the archaeological excavations and these
were a a.okehouse (Feature 1) and a detached kitchen/shed
(Features 3 and 10) (Figure 7). These buildings, along
with the existing shed, privy, and well, constitute the
.ajor structural features associated with this housesite,
and all are located in the backyard.

The standing shed (Figure 7) was constructed in the
twentieth century out of .aterials salvaged fro. the rear

e11 (Howard 1978). The privy was also of twentieth

{V‘JIJbXI1llH

QM?

*II': t DWIOSIHNZVIA

68

century construction, and no previous privies were located
during this investigation. It is possible that the
nineteenth century occupants did not have specially dug
facilities. Oral history in the study of the Bay Springs
farIsteads suggests general use of the woods for toilet
facilities, resulting in a paucity of privy features
(S.ith et a1. 1982: 57, 222). This is further supported
as a general practice in Barton, by the lack of privy
features at the other housesites associated with this town
(Cleland and McBride 1983:375).

The open brick—lined well, as .entioned previously,
was nineteenth century in construction, with an artesian
well west of the house serving as the twentieth century
water source (McClurken and Anderson 1981: 619, 883—884).
The brick-lined well was not used as a trash receptacle,
although so.e unlined wells found on the Barton townsite
were filled with refuse (Cleland and McBride 1983: 179-
181, 235-237).

The s.okehouse (Feature 1) was evidenced by a
rectangular solid brick foundation oriented on the sa.e
axis as the house and located approxi.ately 24 . east and
slightly south of the house (Figure 7). Oral data
indicates this structure functioned as a s.okehouse and
food storage shed, and further, that it was of log

construction (McClurken and Anderson 1981: 877, 882, 560—

562, 619).

I
I
‘I‘ILI
‘ I
t 7
3

II
'1
3a
I

l

I) '7
>1

69

Archaeological data indicates that the construction
.aterial was a1.ost entirely salvaged when the structure
was destroyed in the early twentieth century (Minnerly
1983: 61-65) (Figure 8). The artifactual data suggests a
.id- to late nineteenth century date range with the lower
levels exhibiting a predolinance of mid-nineteenth century
.aterial (Tables 2 and 4). The .ixture of .aterials in
the upper levels can be related to disturbances fro.
plowing in the southern half of Block 1 and ani.al
rooting.

Functional analysis shows architectural ite.s in the
highest proportion (Table 6), which is expected since this
structure was torn or fell down. When the architectural
ite.s are re.oved fro. the datable asseablage the eaphasis
shifts to the late nineteenth century date range (Table
4). The next highest functional percentage consists of
bottle glass (23 percent) and this too is expected as the
s.okehouse also functioned as a storage shed.

Perhaps the .ost telling percentage is that of faunal
re.ains, especially when co.pared to faunal re.ains found
Ielsewhere on this site. If this feature were a s.okehouse
and fbod storage shed, then higher proportions of bone and
shell re.ains could be expected. Terrance J. Martin
(1983: 286-306) conducted a detailed faunal analysis of
the Cedar Oaks asselblage. He noted that Block 1

(containing Feature 1) ”yielded 40.1 percent of all ani.a1

A

70

remains from the site," and "exhibited the third highest
concentration of bone and shell at 39.2 g per .2" (Martin
1983: 291-294). It also yielded ”the greatest nu.ber of
pig ele.ents (45 percent), oyster shells (70 percent), and
fish remains (38 percent), and the second highest nu.ber
of freshwater .ussel shells (22 percent)” (Martin 1983:
294).

The presence of a s.all proportion of tablewares and
utilitarian wares (Table 6) can also be related to this
feature having functioned as a s.okehouse/storage shed.
Utilitarian wares, such as crocks and Jugs, were co..only
used as food storage containers. The presence of
tablewares .ay be related either to dishes being
transported out to the s.okehouse to bring in food items
for consu.ption (sole of which .ay have broken in the
process), or ite.s deposited after the structure no longer
existed and which were .ixed with the feature asse.b1age
through plowing and ani.al disturbances.

The total absence of ite.s in the foodstuff related
category has .ore to do with the high proportion of
unidentifiable ite.s (14 percent) than with the total lack
of this .aterial in Feature 1. Further.ore, no clothing
or personal ite.s were recovered adding to the conclusion

that this feature functioned as a s.okehouse/food storage

shed.

71

Even though no firebox or s.udge pit was uncovered
inside the foundation, this does not negate its function
as a s.okehouse. There were ash and charcoal
concentrations within the feature level fill (Minnerly
1983: 65) indicating either that the fire was set upon the
ground surface with no specially prepared hole or brick
hearth, or that the firebox was subsequently re.oved when
the structure ceased to function as a s.okehouse. A
s.okehouse located on a si.ilar housesite in Barton did
have a brick-filled depression that appears to have served
as the s.udge pit (Cleland and McBride 1983: 234).

The area i..ediate1y surrounding the brick foundation
showed patterns of postholes indicating an addition to the
s.okehouse and/or fenced areas on its north and south
sides (Figure 8). Oral accounts do indicate that s.a11
ani.als were penned near this structure during the early
years of this century (Minnerly 1983: 32).

The other outbuilding discovered during the
archaeological investigation was evidenced by Features 3
and 10 (collapsed brick piers) and an associated .idden
(Figure 7). Table 2 shows that the .ajority of datable
ite.s is mid-nineteenth century with a notable proportion
of earlier .aterial (i.e., early .achine-headed cut nails)
in Feature 10. Few twentieth century ite.s were found in
the feature fill and these consisted of decalco.ania-

decorated ceranic tableware frag.ents in the upper level

72
of Feature 10. These were likely deposited after the
structure no longer existed and were subsequently .ixed in
with the feature fill by cultural and ani.al disturbances.
The re.oval of cut and wire nails fro. the datable
asse.b1age (Table 4) places greater e.phasis on the late
nineteenth century date range.

This data suggests that this structure was in
existence fro. the .id- to late nineteenth century and was
probably built when the house was first constructed. It
was destroyed in the late nineteenth century, and none of
the oral infor.ants have any recollection of a structure
in this area other than the present shed (Minnerly 1983:
28-32).

The presence of early nineteenth century nails .ost
likely indicates either that this structure was
constructed fro. materials salvaged fro. an earlier
building, or that there was access to early .achine cut
nails even though a .ore .odern type was being produced by
the ti.e this structure was being built (Nelson 1968).
Newton (1980: 5) does note that the cut nails used to
construct the house are also of this earlier type further
supporting the theory that both of these structures were
built si.ultaneously.

Functionally, this structure .ay have been a detached
kitchen. The presence of such structures is fairly co..on

in warner c1i.es due in part to the lessened need for the

73

extra war.th a kitchen inside a house would provide. It
would have been far .ore co.fortable for a kitchen to be
away fro. the house yet close enough for easy access.

Detached kitchens have been identified at higher
status ho.es (see Lees 1980: 119, Lewis 1977: 64), as well
as at lower status ho.es (see Beagan 1983: 111).

Ken Lewis (1977: 64) noted that ”co.parative
archaeological evidence indicates that separate kitchen
structures were generally situated Just to the rear of
larger dwellings" and that .any of the. are located
”nearest the left rear corner of the house." This holds
true for the location of this structure at Cedar Oaks in
accordance with the present configuration of the house
(Cleland and McBride 1983: 343).

Table 6 shows that once again the largest proportion
of artifacts is architectural in function. This too is in
keeping with the fact that this structure was destroyed.
The next highest proportions consist of cera.ic and glass
tablewares (9 percent and 4 percent, respectively for
Features 3 and 10) and bottle glass (7 percent and 8
percent, respectively). The higher percentages of
tablewares and bottles lends support to the theory that
this structure functioned as a kitchen, as both would have
been co..on ite.s related to kitchen activities (i.e.,

food preparation, cooking, consu.ption, and storage).

74

Martin’s (1983: 286-306) faunal analysis concerning
this structure revealed that

Although only 5.9 percent of the total faunal

asse.b1age was obtained fro. (Block 2 and Features 3

and 10), the density of 42.4 g of shell and bone per

square .eter of excavated area was the second highest
at the site. Oyster shell was the predominant aninal
re.ain in this area of the site and occurred along
with pig teeth, .ussel shells, a turtle ele.ent, one
chicken bone, one cattle bone and several
unidentified .a..al bones.

Added to this is the presence of charcoal and ash
concentrations in the feature fill. These deposits are
likely the result of cooking activity and either dropped
through the floor or were purposefully thrown out around
and underneath the kitchen. As the artifacts and brick
pier re.nants showed little evidence of burning, it is
concluded that the charcoal and ash deposits represent
cooking activity by-products rather than evidence that the
structure was destroyed by fire.

During the late nineteenth and early twentieth
century the ell attached to the rear of the house
functioned as the kitchen (McClurken and Anderson 1981:
561-562, 877, 880). This structure .ay be evidenced
archaeologically by Features 8 and 14-16, although 14 and
16 appear to be .ore closely associated with 17 and 23 as
evidence of a rear full-length gallery (Figure 7). The
all is best evidenced however, by the relatively shallow

.idden (average 10.7 c. deep) found in the northern half

75
of Block 3 as co.pared to the deeper .idden (average
14.2 c.) found in the southern half of this block, as well
as that found elsewhere in the backyard which is up to 30
c. deep in certain areas (Figure 6).

Table 3 shows that, for the northern half of Block 3,
the .ajority of datable artifacts are fro. the late nine-
teenth century. This appears to indicate that the rear
ell was not built until the very late nineteenth or, .ore
likely, the early twentieth century. Nhen .odern .achine
cut and wire nails are re.oved fro. the datable asse.b1age
(Table 5) the e.phasis is only slightly greater for the
twentieth century range, indicating that deposition after
the all had been re.oved was co.parable to that which
occurred in the late nineteenth century prior to its
existence.

The s.aller proportions and .arked decrease in
density of .id-nineteenth century artifacts (Tables 3 and
5) indicates that this area was kept relatively clean
during this ti.e period. The sweeping or hoeing of yards
to keep the. clear of refuse was a co..on practice in the
nineteenth and early twentieth centuries. Oral accounts
of this practice surrounding a dwelling have been noted by
Ada.s (1980: 216, 225) and S.ith et a1. (1982: 53, 57,
217). Such clearing was done around a house as this area
was often "the scene of .any outdoor activities” (Ada.s

1980: 225).

76

Besides the evidence of outbuildings as loci for
certain activities, other data obtained during this
investigation can be related to other aspects of site
utilisation. An artifact density scale was devised for

this analysis and it is as follows:

Sterile = O artifacts/per unit excavated
Sparse = 1-400 artifacts/per unit
Moderate = 400-l,500 artifacts/per unit
Dense = 1,500-4,600+ artifacts/per unit

This scale does not include feature asse.b1ages.

In general, the backyard area appears to have been
the .ain locus of activity for the entire occupation span
of Cedar Oaks. The deepest sheet .idden (10-30 c. in
depth) is found in this area (Figure 6), as well as the
.aJor outbuildings (i.e., shed, kitchen, privy, and
s.okehouse).

In all of the excavated areas of this portion of the
site (i.e., Blocks 1-3, and Units 85, 87, and 95) the
functional breakdowns (Tables 7 and 8) show architectural
ite.s and bottle glass in the largest proportions. This
is, however, true of the entire site, and is due to the
nu.ber of destroyed outbuildings and structural changes to
the house which encouraged the deposition of ite.s such as
nails and window glass, and the disposable nature of glass
bottles over tablewares, which would have been .ore
carefully curated. Certain functional differences can,

however, be noted in this backyard area.

'77

 

 

 

NNN.NN u AduON
Nv.vm NHo.H cH H H 0N H N NN N mH «INH N
o>.oo New NH Hv e mm H H NH Hv N «IoH N
ON.VNH opN.N NN H e an H H N Hv N «INN m
NN.NN «No NN N N NN N N HN Hv N «INN e
auHss
osonusoe
N>.NNN vNN.NN NH N N on H H NH H m «Iem N
euHs:
sausages
«N.NNN NNN.NH NH N N Hm H N NH H n «INN N
No.NVN oHN.m N H m mm H v NH H HH «IeN N
mm.mo« Hoo.m~ «NH «m an ace «H «n «e« «« as «n«m H
xooHN
d v o s w a H v 3 a a o n 1.5 1.x 8
3 In a. m n x mm m w. 3. m; um m.
1 99 3 D. U. 3.4 3 T. T. mu. 0
mp. IT. a e M t. on. s .... WT. am Z x.
86 .4 I T. p. 3 u.... 3 a 3. A a
a ....u T. I a an n 09 OD x u
3m. 9 a o 16 I. b s: 1 a m.
D. D. u 1. 1: .1 .... 33 p.
a o a n 1.3 e n. A
u 1. O 1 .4 1 S u 9 Q a
S 81 n a. m a S T. 1. 1
T? S I I D. a
a I e s p
A .4 s
a
D.
.eswsuseouea ha suosuHuus shsuseuIIos no ssovasesa HsnoHuussh .N sHAsa

 

78

NHN.Q u H<Hou

coda «H m m m s« m «« mH I HH «n« ma HHaa
oo.n«« eve «H « m ov H HH mH e e «n« ma sHaa
m«.mm« ««N.H eH H m be H Hv m« H m use we HHna
m«.«e« mum «H « H em Hv H «H m m use mm HHaa
oo.«mH mom pH Hv m m« «« mH m Hv m use Hm HHns
ma.ooH new a H « om Hv Hv MN m HH use om uHaa
me.mmH mum mH H m me H v m« Hv « «me mm «Hg:
oc.emH men mH Hv « mm Hv H mm m «H «ac em aHaa
co.es emH .H « m we H I mH m «H «av mm HHup

mb.NbN Hm°.H «NH NH am New RH va me va Ru «flu mm uHsD

 

dV 01 N d H w. .3 J 8 an .40 T. 0
3A 1.0 T. e p. 1 MT. 0 o 33 “.8 UN u
13 3 s m 1 3 10 o .4 1t. 1 a ...
1 99 a D. u. 33 D. 3 WT. Tm up. 1.
me 1T. 8 v. M .... ou. s ... T. a Z
7.5 1 T. T. p. 3 U... 3 a ....1. MT. 8 u
8 In T. 1 a an n 09 23 x
I. .e a a TED I. 5 81 1 a
D. p. u 1. I. T. T. 35 p.
a a a n ....8 e p. A a
u 1.3 o 1 1. 1 S u 9 p. 1
s 51 n e m a S T. 1.
1w 5 D. S D.
.A 1. s
8
D.

 

.aowsvs00hoa ha auosuHuus OIIHIOMInoI mo crevasshn HsIoHuossu .m eHasa

79

The units which comprise Block 1 (encompassing the
s.okehouse feature) exhibit, according to the above scale,
a .oderate artifact density. The average density per .2
is 206.53 artifacts. The functional breakdown, in co.-
parison with the other blocks of units (Table 7), shows
that Block 1 has the second highest percentage of bottle
glass (24 percent), the third highest percentage of
ceramic and glass tableware (7 percent), and relatively
high proportions of utilitarian cera.ics and faunal
re.ains (2 percent and 3 percent, respectively), all of
which are in keeping with this area as a locus for food
preparation and storage activities related to the
s.okehouse.

The sheet .idden in this block is an average of 15.8
c. in depth and all stratigraphic levels show a higher
proportion of .id-nineteenth century artifacts (Table 3).
When .odern .achine cut and wire nails are re.oved fro.
the datable assemblage (Table 5), the late nineteenth
century category shows a .uch higher density. As the
s.okehouse is known by oral accounts to have existed into
the late nineteenth or early twentieth century, a
predo.inance of late nineteenth century artifacts would
suggest that it was constructed and pri.arily in use
during this time span. Unit 87, which is located 4 . east

of Block 1 (Figure 5), exhibits si.ilar date range and

80

functional proportions and average artifact densities
(Tables 3, 5 and 8).

A garden area was noted running east-west across the
southern portion of Block 1. This was evidenced by plow
scars and concentrations of onion bulbs, and it postdates
the destruction of the s.okehouse indicating a twentieth
century date.

The units of Block 2 (containing the detached kitchen
feature) exhibit a .oderate artifact density with an
average density per .2 of 242.08 artifacts. The
functional proportions co.pared with the other blocks of
units, show relatively high percentages of foodstuff
related ite.s and bottle glass (4 percent and 15 percent,
respectively), as well as the second highest percentage of
cera.ic and glass tableware (11 percent). These can be
related to this area having been utilized for kitchen
activities.

The sheet .idden in Block 2 is an average of 18.2 c.
in depth and has high proportions of .id- to late
nineteenth century .aterial in all stratigraphic levels
(Table 3). Level 3 has a notable proportion of early
nineteenth century ite.s (18 percent) indicating the
presence of early-type nails. When .odern .achine cut and
wire nails are re.oved from consideration, the highest
density is found in the late nineteenth century range

(Table 5), although Block 2 exhibits the highest

81

aid-nineteenth century density per .2 (2.50/a2) among all
the blocks of units. This would be in keeping with the
supposition that the detached kitchen was constructed in
the .id-nineteenth century (ca. 1850).

A fenceline (Feature 13) along the southern wall of
this block indicates the presence of an aniaal pen or
separated garden area in the vicinity. As this fence was
found lying just beneath the sod layer it is concluded to
be twentieth century in date.

The northern units of Block 3, as discussed pre-
viously, show a .oderate artifact density and an average
density per .2 of 365.34 artifacts. The functional
breakdowns (Table 7) show a high proportion of bottle
glass (16 percent) and the fourth highest percentage of
tableware (5 percent). The sheet .idden is an average of
10.7 c. deep and predo.inantly late nineteenth and
twentieth century in date (Tables 3 and 5). This area was
the locus for kitchen activities in the twentieth century
and appears to have been swept clean before that.

The southern units of Block 3, on the other hand,
show a dense amount of artifacts with an average density
per .2 of 598.78 artifacts, the highest of the entire site
excluding features (Table 7). The sheet .idden found here
is an average of 14.2 c. in depth.

Functional breakdowns for the southern half of Block

3 are virtually identical to those found in the northern

82

half (Table 7) indicating the occurrence of the sa.e
general activities, with the southern half collecting
greater nu.bers of artifacts. This can be related to the
influence on deposition by Feature 21 (Figure 7) located
in the southwest portion of this block.

Proportionally, a .id- to late nineteenth century
date range is indicated (Table 3); however, the average
density (.inus nails) shifts the emphasis to the late
nineteenth and twentieth centuries (Table 5). A garden
area, running east-west, was evidenced in the extre.e
southern portion of Block 3 by plow scars present beneath
the deep .idden. This garden .ay, therefore, be .id- to
late nineteenth century in date.

Units 85 and 95 (Figure 5), which are also located in
the area of greatest backyard activity, exhibit relatively
deep sheet .iddens (10 to 26 c.), .oderate artifact
densities, average densities per .2 of 272.75 and 242.25
artifacts, respectively, and pri.arily a .id- to late
nineteenth century date range e.phasis (Tables 3 and 5).
Functionally, no specific activities are of note (Table
8).

The peri.eters of the backyard area revealed a less
developed .idden (8 to 12 c. deep and lighter in color)
which has been .odified by plowing along the eastern edge

of the site area. Artifact densities range fro. .oderate

83
to sparse with average artifact densities per .2 ranging
fro. 9 artifacts to 162 artifacts (Table 8).

Proportional date ranges in this area indicate a .id-
to late nineteenth century range for Units 86 and 90 and a
.id-nineteenth to twentieth century range for Unit 91
(Table 3). Date range average densities (.inus nails)
reinforce these ranges (Table 5). Unit 91 shows higher
proportions and densities of twentieth century .aterial as
this area was the locus of a twentieth century refuse du.p
along the eastern slope.

Functional breakdowns show unexpectedly high
proportions of tableware in Units 86 and 90 (Table 8).
Their distance fro. the house and kitchen area would lead
one to expect lower proportions in this category. Unit
91, on the other hand, exhibits high proportions of
clothing and personal ite.s and foodstuff related
.aterials (22 percent and 16 percent, respectively)
because of the large nu.bers of rubber boots, leather
shoes, and canning jars discarded in the twentieth century
du.p.

The plow zone in the extre.e northeast and eastern
portions of this site indicate that the fields, at least
in the twentieth century, reached up to the houselot’s '
edges. A posthole (Feature 56, Figure 7) found in Unit 86
is likely evidence of a fenceline (direction unknown)

separating the yard fro. the fields.

84

Along the southern edge of the backyard a layer of
clay was found overlying the .idden (Figure 6). This cap
of clay appears to be a twentieth century atte.pt to level
the slope in this area of the site to prevent or retard
erosion.

The sides of the house, Blocks 4 and 6, exhibit
shallow sheet .iddens (approxisately 6 c.), sparse
artifact densities, and average densities per .2 of 28.82
and 60.70 artifacts, respectively (Table 7). Block 4
shows higher proportions of .id- to late nineteenth
century ite.s, while Block 6 has a notable percentage of
twentieth century .aterial (Table 7). Date range
densities (.inus nails) shift the e.phasis for both blocks
to the late nineteenth and twentieth centuries (Table 5).
The construction and destruction of the four chi.neys in
these areas accounts for the .ixing of date ranges
throughout all levels (Table 3).

Block 5, located directly in front of the house,
contains the front walkway and for.al garden features
(Features 27-30 and 45, Figure 7). The sheet .idden here
is extre.ely shallow (average 5.9 c.) and the units
exhibit a sparse to .oderate density. The entire block
has an average density per .2 of 134.30 artifacts
Table 7).

The date range for all levels is pri.arily late

nineteenth to twentieth century with a s.aller proportion

85

of mid-nineteenth century material (Table 3). When nails
are removed from consideration the emphasis remains on the
late nineteenth to twentieth century range (Table 5).

Functional breakdowns for Block 5 show low
proportions in all categories except architecture and
.iscellaneous items (Table 7). This appears to indicate
an area consciously kept clean throughout the occupation
of this site, perhaps as an esthetic presentation of the
house to visitors.

The extreme northwest corner of the frontyard was
used as a flower and shrubbery garden, as well as a refuse
disposal area in the late nineteenth and twentieth
centuries. The latter is a trash pit (Feature 57) found
in Unit 89 (Figure 7).

Units 96 and 98 were situated along the southern and
western fencelines, respectively, and both revealed refuse
areas of .oderate artifact densities. Average densities
per .2 are 333.25 artifacts for Unit 96 and 223 artifacts
for Unit 98 (Table 8).

A .idden layer was present in both units which was
lighter in color than that found in the backyard but of
comparable depth (approxi.ately 19 cm). Date ranges show
higher proportions of mid- to late nineteenth century
.aterial, with Unit 96 having notable percentages of

twentieth century items (Table 3). Date range average

86
densities (.inus nails) shift the emphasis in both units
to the late nineteenth and twentieth centuries (Table 5).

Block 7 is the location of a refuse disposal area
(Feature 53) over the northern slope which was used
pri.arily during the late nineteenth and twentieth
centuries (Tables 2 and 4). Post.olds found intruding
into this refuse deposit (i.e., subfeatures 1-6, Figure 7)
appear to be for.er fenceline posts of varying ages
(Minnerly 1983: 98-99).

In su..ary, the utilization of space on this
housesite through ti.e consisted of outbuildings serving
as the loci for various activities, designated areas for
gardens and ani.al pens, and a conscious effort on the
part of the occupants to keep certain areas cleaner than
other areas.

Outbuildings in the backyard consisted of two
kitchens, a shed, s.okehouse, privy, and well, with the
detached kitchen, s.okehouse, and well dating generally
fro. the .id- to late nineteenth century occupations. The
kitchen may, in fact, date back to the construction of the
house. The smokehouse survived into the early twentieth
century when it was known to have had an ani.a1 pen next
to it. The shed, privy, and all kitchen, on the other
hand, date from the twentieth century occupation.

Another fenced area dating fro. the twentieth century

was located near the standing shed. Two gardens were

87
discovered, one to the southeast of the house dating
possibly fro. the nineteenth century occupation and the
other, in the southeast portion of the site, dating from
the twentieth century.

Throughout the occupation of this site it appears
that outdoor activities were focused on the area bounded
by the shed and detached kitchen on the north, the house
to the west, and the smokehouse to the east. The enclosed
area exhibits the deepest sheet .idden, as well as the
highest non-feature artifact densities.

The area off the northeast corner of the house, as
well as the surrounding sides and front area, show
indications of having been kept clean throughout the
site’s history. Only shallow middens were found in these
areas. The eastern peri.eter shows evidence of twentieth
century plowing and refuse disposal, while the southern
edge exhibits a twentieth century leveling activity. Oral
accounts indicate that the southern slope was planted with
a fruit orchard during the twentieth century occupation
(Minnerly 1983: 32).

The frontyard of the house was used in the late
nineteenth and twentieth centuries as a for.al
presentation and orna.enta1 area. The peri.eters of the
northern, western, and southern portions of the site,
however, were used to varying degrees as disposal areas

primarily during the later occupation of this site.

88

Archaeology--Hefuse Disposal Patterns

Two types of disposal behavior were discerned at
Cedar Oaks, specific dumping areas and an overall refuse-
laden sheet midden. The for.er consists of Features 21,
53, and 57 and areas indicated by Units 91, 96, and 98.
The midden is concentrated pri.ari1y in the backyard
extending out fro. the house approxi.ately 30 m to the
east; although a shallower .idden is present over .uch of
the re.ainder of the site including the frontyard
(Figure 6).

Feature 21 has been noted previously as a refuse dump
underneath the for.er rear gallery. It is concentrated on
the southeast corner of the house and extends eastward
fro. it approxi.ately 4 m (Figure 7). Proportional data
(Table 2) indicates a .id- to late nineteenth century date
range; however, when nails are re.oved from the datable
asse.b1age, the e.phasis shifts to the late nineteenth
century with the second highest average density per .2 in
this feature dating fro. the twentieth century (i.e., 7.6)
(Table 4). Overall, it contains the second highest
frequency of artifacts (4,568) and the third highest
average density per m2 (380.67).

Functionally, the artifacts fro. Feature 21 cover the
full range, with high percentages of architectural ite.s,
bottle glass, .iscellaneous items, and tableware (Table

6). Its proxi.ity to the house would .ake it a convenient

89

receptacle for all .anner of domestic refuse. The obvious
structural remodeling which took place in this area (see
previous section) accounts for 59 percent of the
asse.b1age being architectural in function. In general,
however, its functional .akeup is not unique, and other
areas exhibit a si.ilar co.position (Table 7).

Feature 53 (Figure 7) is a refuse dump on the
northern slope approximately 9 . off the northeast corner
of the house. In addition to the .aterial tabulated in
Table 6, it also contained a large concentration of brick
and mortar that .ay have co.e fro. the northern chi.neys.
Although only 12 square .eters were excavated, the entire
du.p is actually 18 x 23 . in visible extent.

Proportionally, the artifacts fro. this feature date
fro. the .id- to late nineteenth century with a sizable
percentage fro. the twentieth century (Table 2). However,
the average densities/.2 (nails excluded) exhibit a
pri.ary e.phasis on the late nineteenth and twentieth
centuries (63.20 and 38.30 artifacts/.2, respectively,
Table 4). This feature contains the highest frequency of
artifacts (6,382) but only the second highest average
density per .2 (531.83 artifacts/.2).

The functional breakdown for Feature 53 (Table 6)
shows the .ajority as architectural ite.s (32 percent),
with bottle glass (25 percent), tableware (13 percent),

and foodstuff related ite.s (12 percent) in notable

90

proportions. Despite the fact that less than 5 percent of
the site’s faunal re.ains were recovered from this
feature, it was found that the concentration of bone and
shell was the site’s highest at 47.7 g per .2 (Martin
1983: 295). The presence of such high proportions of
kitchen-related and domestic refuse, when co.pared to
other features and areas (Tables 6 and 7), can be
explained by the proxi.ity of this du.p to the two
kitchens and the house itself. It would have been a
close, sanitary, and out-of—sight du.ping area relative to
these structures.

Feature 57 is a circular trash pit that was found in
the northwest corner of the frontyard (Figure 7). This
was an excavated pit that was used as a trash receptacle
primarily during the late nineteenth and twentieth
centuries (Tables 2 and 4). During the twentieth century
it was used for burning refuse as the top layer was
co.prised of charcoal and burned artifacts (Minnerly 1983:
101). This pit was also used during the mid-nineteenth
century occupation although to a lesser extent than in the
later periods. In fact, it contains the highest average
density per .2 of .id-nineteenth century artifacts
co.pared to other features (i.e., 5.00/.2, Table 4). In'
addition, it contains the highest average density per m2
(l,068.5/.2, Table 4) when co.pared to other .ajor

features, but only the third highest frequency.

91

Functionally, the artifacts fro. Feature 57 cover the
full range of categories with an e.phasis on architectural
(32 percent), .iscellaneous ite.s (27 percent), bottle
glass (17 percent), and foodstuff related ite.s (13
percent) (Table 6). This too appears to have been a handy
receptacle for domestic refuse, although its placement in
the frontyard in full view of all visitors is so.ewhat
perplexing. Especially when great care was taken during
the sa.e time period to construct esthetic features such
as a for.el garden enclosure and brick walkway.

Unit 91 (Figure 5) was placed in the vicinity of a
twentieth century refuse du.ping area along the eastern
slope of the site. This area is located over 40 . to the
northeast of the house and was the receptacle for
discarded boots, shoes, canning jars, paint cans, and
other ite.s. The footwear accounts for the high
proportion of clothing and personal ite.s (22 percent,
Table 8), and the canning jars for the foodstuff related
ite.s (16 percent). This refuse area was utilized
primarily during the late nineteenth and twentieth
centuries (Table 3), with a higher average density per m2
(excluding nails) during the twentieth century occupation
period (i.e., 19.25 artifacts/.3, Table 5).

Units 96 and 98 (Figure 5) were located along the
southwestern and western fencelines, respectively. Both

revealed that the areas to the "outside” of these fences

92

were also used as refuse disposal areas although to a
lesser extent than the northern and eastern slopes. It is
likely that .uch of the refuse found along these
fencelines represents sweepings fro. the frontyard area.

Functionally, the areas investigated by Units 96 and
98 exhibit si.ilar compositions, although .ore items found
in Unit 98 could be related to a foodstuff function than
those found in 96 (i.e., 11 percent and less than 1
percent, respectively, Table 8). Proportionally, these
areas were utilized pri.arily during the .id- to late
nineteenth century (Table 3); however, the emphasis shifts
to the late nineteenth and twentieth centuries when nails
are re.oved from the datable asse.b1age (Table 5).

Besides these specific refuse du.ping areas there was
a sheet .idden which covered .uch of this site, to varying
degrees. This represents both the by-products of activity
areas (i.e., primary refuse) and those ite.s lost or
casually discarded anywhere they .ight fall on the site
area.

This .idden is deepest (10-30 c.) and contains the
highest average non-feature artifact density per .2
(334.71 artifacts/.2 or, 73,636 artifacts divided by
220.2) in the area directly behind the house and extending
approximately 6 m beyond Block 1 (Figure 6). Therefore,
the densest portion of the sheet .idden is bounded by the

house on the west side, the two kitchens (i.e., detached

93

and ell) to the north, the smokehouse to the east, and the
slope edge to the south. It is logical to assu.e that

this .idden was for.ed by the heavy day-to-day traffic and
concentration of activities which took place in this area.

The midden thins out in both depth and artifact
density to the sides and front of the house, indicating
areas little used and/or kept clear of refuse (Figure 6).
The peri.eters of the site exhibit a lighter .idden which
is 8-11 c. in depth, indicating areas little used for
specific activities and receiving less refuse overall.

It is likely that this peri.eter .idden, especially in the
frontyard, represents the final deposition of yard
sweepings.

The artifacts recovered fro. this sheet midden cover
the full range of function (Tables 7 and 8) with
architectural ite.s and bottle glass predo.inating over
the entire site. As for dates, the dense .idden in the
backyard is .id-nineteenth to twentieth century in range
with e.phasis upon the late nineteenth century (Tables 3
and 5). On the other hand, the front and side yards, as
well as the site peri.eters, date primarily fro. the late
nineteenth and twentieth centuries. Very little early to
.id-nineteenth century .aterial was recovered from these'
areas (Table 3).

Maps showing the distribution of datable non-feature

artifacts (excluding nails and bottle glass, see

94

Chapter 2) are presented in Figures 9, 10, and 11. The
distribution of .id-nineteenth century .aterial (Figure 9)
shows concentrations directly off the east-central and
southeast portion of the house (southern units of Block
3), the area of the detached kitchen (Block 2), and in the
vicinity of the s.okehouse (Block 1). The frontyard, side
yards, and perimeter areas appear to have received only a
sparse a.ount of .aterial during this period of
occupation. It should be noted that Feature 53 (Block 7)
and Feature 57 (Unit 89) also received a co.paratively
s.all a.ount of mid-nineteenth century .aterial (Table 4).

When the .id-nineteenth century distribution (Figure
9) is compared to that of the late nineteenth century
.aterial (Figure 10) it is apparent that density has
increased over all areas of the site. As glass bottles
and nails have been removed fro. the datable asse.b1age,
this increase in density is not influenced as greatly by
the .ass production and greater availability and
disposability of glass containers or by the structural
re.odeling and destruction which occurred on this site in
the late nineteenth and early twentieth centuries.
Rather, it appears that a change in disposal behavior has
occurred.

Specifically, the area im.ediately surrounding the
house (i.e., front, sides, and back within 0-6 I) received

a greater artifact density during the late nineteenth

l3.

 

95

 

 

 

mmE coflsflupmg #8ng SagosfimHIon . a snag

39:8 5 $06305 #0: 3.3:" ..nqomu-..‘
. . I

a ....._......_.. . .. H . E...

 

 

EH32 83...ng
.330 £38 .33. «835 8: 80H. u ..

 

96

mm: :UHuanHumHU uOMMHuHm Nusucwo InumH mung .OH ousmHm

 

 

L
mHmEmm CH covsHocH uoc.mu

. H" m¥<fiv¢<dwu Z Mmmfl
m «summiumum w m m V

 

 

>3...

Haaumumfi mnsummmqum

.mmmHm mHuyon .mHHm: m©5H0:H uoc meow u «

 

. O 0 an“ O O O
KINDS-JESSE

 

 

 

97

mma_gowusnflnumflc pooMHuum ansucmo ecuom .HH whomHm

 

 

 

 

 

 

 

 

 

 

«H96... a... @839: no: Bus..- 93 «(So . E
... mu . ,. ..
D . «$62.20 ~MU0 £ r at...
I. BE
HIH .u um=-. Mm
pewwmmww
E I EH...
1 I.
ma BEE wflfi
an "mg...
n_ ..J “mum...
D E ......3. flan.
FL DH.
nH... a

HMHHBmE egg
.mmMHm pruon .mHHmHH @9505 won moon I I

98

century than this sa.e area received during the previous
ti.e period. In fact, all areas of the site show
relatively dense deposition during the late nineteenth
century. This is reinforced when Features 53, 57, and 21
are considered, as they all have a predo.inance of late
nineteenth century material (Table 4).

The twentieth century distribution .ap (Figure 11)
shows yet another shift in deposition. During this period
artifacts appear to concentrate pri.arily in the 0-6 .
area surrounding the house on all sides. Blocks 1, 2, and
the perimeter units show lesser a.ounts than was found
during the late nineteenth century (Figure 10). Once
again, it is noted that Features 53, 57, and 21 did
receive a sizable a.ount of refuse during this ti.e period
(Table 4).

Unfortunately, the stratigraphic levels excavated on
this site exhibited .ixing of date ranges throughout (see
Table 3), preventing the designation of certain levels as
being fro. one definite time period. This served to
reduce the sa.ple asse.b1age to only those items which
could be positively identified as to date range. However,
care was taken to reduce the bias of this sa.ple, for a
portion of this analysis, by re.oving nails (which were
deposited in later periods unrelated to their original

context) and bottle glass (which artificially inflated

99
date range frequencies by the sheer volu.e of this
increasingly disposable item).

In order to investigate the refuse patterning in
greater degree, a series of contingency tables were
for.ulated to test certain hypotheses. As the general
thesis hypothesis proposed that the change in site
function through ti.e resulted in a changing pattern of
refuse disposal, contingency tables were devised to test
the relationship between distance fro. the house and date
ranges. The first variable was chosen as the house served
as the .ajor focal point for division of space (i.e.,
frontyard vs. backyard), as well as the originating point
for .uch of the refuse that was eventually deposited in
the surrounding site area as already proposed in related
hypotheses (see Chapter 1). It was felt that distance
from the house .ight have been a factor associated with
disposal behavior through time on this site. Such an
association between distance from the dwelling and refuse
patterning within a sheet .idden has been noted by Randall
Moir (1982: 147) for nineteenth and twentieth century
housesites in Texas.

In order to reduce so.e of the bias inherent in the
distribution of the excavated sa.ple, intervals of .
distance, (i.e., 0-6 ., 6-18 m, 18-30 ., and 30-50 .) were
chosen so units which otherwise were not connected could

be combined into one interval. Therefore, in the backyard

100
area, the 0-6 . interval consisted of only Block 3 (92
.2), 6-18 . included Block 2 and Unit 95 (28 .2), 18-30 m
included Block 1 and Unit 85 (96 m2), and 30-50 .
consisted of Units 86, 87, 90, 91, and 99 (18 .2).
Features were not included in this initial analysis,
therefore, Block 7 and Units 74 and 75 were re.oved from
the sample.

The frontyard area consisted of only two distance
intervals: 0-6 m (Blocks 4, 5, and 6 or 92 m2, excluding
Feature 6--Units 24, 26, and 27) and 6-18 . (Units 89, 96,
and 98 or 10 .2). The excavated areas of the side yard
were considered to be part of the ”frontyard” division of
space on this site.

Average densities per .2 of datable material from the
lid-nineteenth, late nineteenth, and twentieth centuries
were utilized in order to further reduce the sample bias,
as well as to produce a sa.ple size applicable to chi-
square analysis. The early nineteenth century category
was entirely removed fro. consideration as it represented
pri.arily architectural .aterial better related to the
mid-nineteenth century occupation.

Table 9a and b presents the initial contingency table
calculations testing the possible association between
distance intervals and date range. The hypotheses being

tested are as follows:

101

Ho (null hypothesis): there is no significant
difference between distance fro. the house and
distribution of .aterial through ti.e.

H1 : There is a significant difference between
distance fro. the house and distribution of material
through ti.e.

For both Table 9 a and b the null hypothesis would be
retained, as the resulting chi-squares (i.e., x2 = 8.37
with 6 degrees of freedo. and x2 = 2.61 with 2 degrees of
freedo., respectively) do not exceed, or even come close
to, a 0.05 level of confidence. In fact, values as large
as both of these would be observed in more than 20 percent
of si.ilar tests.

Even when datable feature artifacts and bottle glass
are added into this contingency table sample the resulting
chi-squares still do not allow rejection of the null
hypothesis. The contingencies, with bottle glass added
in, resulted in a x2 of 10.502 with degrees of freedo.
(df) = 6 for the backyard and a x2 of 1.59 with df = 2 for
the frontyard. When feature artifacts and bottle glass
are added, the resulting chi-squares are x2 = 7.33,
df = 6 for the backyard and x2 = 1.82, df = 2 for the
frontyard.

These results indicate that the tested association is

not a statistically significant one, and further, that one

102

 

 

 

 

Table 9. Contingency Tables-~Average density per .2
a. BACKYARD
distance
fro. MID-19th LATE-19th 20th
house centuryt centuryt century:
0-6 . 0.81 21.49 12.39 34.69

(1.95) (24.58) (8.16)
6-18m 2.39 20.78 2.61 25.78

(1.45) (18.26) (6.07)

18-30. 0.79 10.64 1.02 12.45

(0.70) (8.82) (2.93)

30-50. 0.89 8.67 4.44 14.00

(0.79) (9.92) (3.29)

4.88 61.58 20.46 86.92
x2 = 8.37, = 6, ‘*= 0.05, (n) = expected frequency
mammal;
distance

from MID-19th LATE-19th 20th
house centuryt centuryt centuryt
0-6 . 0.08 2.94 3.75 6.77

(0.20) (4.44) (2.12)
6-18m 0.60 11.80 3.30 15.70

(0.47) (10.30) (4.92)

0.68 14.74 7.05 22.47
x2 = 2.61, = 2, °‘= 0.05, (n) = expected frequency

1 = Date ranges .inus nails,

bottle glass, and features

103

variable is not dependent upon the other. It was then
hypothesized that perhaps the only .eaningful association
found at Cedar Oaks concerning date range and distance has
to do with the dichoto.y between the backyard and the
frontyard. Therefore, the following hypotheses were
for.ulated:

Ho = There is no significant difference between
distance from the house and the backyard/frontyard
division of space on this site through ti.e.

H1 = There is a significant difference between
distance from the house and the backyard/frontyard
division of space on this site through ti.e.

It should be noted that only the 0-6 m and 6-18 .
intervals could be tested as the frontyard area was not
investigated further than the latter distance.

Table 10 a presents the results of the chi-square
calculations testing these hypotheses for the mid-
nineteenth century .aterial (.inus features, nails, and
bottle glass). A x2 of 0.06 was obtained with df = l,
which would be observed in over 80 percent of si.ilar
tests. Therefore, for the .id-nineteenth century, the
null hypothesis would be retained.

Table 10 b shows the results fro. the testing of the'
late nineteenth century .aterial (.inus features, nails,
and bottle glass). A x2 = 4.27 with df = l was obtained

and, at a 0.05 level of confidence, this allows for

104

rejection of the null hypothesis. A Cra.er’s V2 (a
.easure of association) was calculated resulting in a
value of 0.07 suggesting a weak association.

Table 10 c shows the results concerning the twentieth
century .aterial (minus features, nails, and bottle
glass). A x2 = 2.10 with df = l was obtained which does
not allow for rejection of the null hypothesis. A value
this size would be found in more than 10 percent of
si.ilar tests.

When datable feature artifacts and bottle glass are
added to the above contingency table sa.ple, a x2 = 0.52,
df = l was obtained for the .id-nineteenth century
(Table lla). According to the critical values of chi-
square, the null hypothesis would be retained for this
date range. For this test, and the above .id-nineteenth
century test, it is indicated that there is no
statistically significant association between distance and
division of space through time. This, of course, only
applies to the maximum distance of 18 m away from the
house. One bias affecting this test could be the small
sample size of the mid-nineteenth century material;
however, when actual counts instead of average densities
are used the null hypothesis is still retained.

Table 11 b shows, for the late nineteenth century, a

x2 = 10.58, df = l which allows for rejection of the null

105

Table 10. Contingency Tables--Average density per .a
a. MID-19th c.t

 

 

 

 

distance

frogjhouge backyard frontyard

0-6 m 0.81 0.08 0.89
(0.73) (0.16)

6-18. 2.39 0.60 2.99
(2.47) (0.52)
3.20 0.68 3.88

x2 = 0.06, df = l, °<= 0.05, (n) = expected frequency

1:. Mt

 

 

 

 

 

distance

from house backyard frontyard

0-6 m 21.49 2.94 24.43
(18.11) (6.32)

6-18. 20.78 11.80 32.58
(24.16) (8.42)
42.27 14.74 57.01

x2 = 4.27, df = l, °‘= 0.05, V2 = 0.07,

(n) = expected frequency

c. 20;; c,t

distance

from house backyard frontyard

0-6 m 12.39 3.75 16.14
(10.98) (5.16)

6-18. 2.61 3.30 5.91
(4.02) (1.89)
15.00 7.05 22.05‘

 

x2 = 2.10, df = l, °‘= 0.05, (n) = expected frequency

* = Date ranges .inus nails, bottle glass, and features

106

hypothesis at a 0.01 level of confidence. A Craaer’s V2
0.11, however, suggests a weak association.
Table llc for the twentieth century, shows a x2 =

5.49, df = l which also allows for rejection of the null

hypothesis at a 0.02 level of confidence. A Cra.er’s V2
0.10 also suggests a weak association.

These results indicate that within 18 . of the house
there is a statistically significant difference between
distribution of material in the backyard versus the
frontyard in the late nineteenth and twentieth centuries.
Specifically, during the late nineteenth century the
backyard was receiving higher densities overall fro. 0-18
m than the frontyard (Table 10b). When looking at the
frontyard, it can be noted that the 0—6 range was
receiving less refuse than the 6-18 . interval (Table 10b
and Table 11b).

When features and bottle glass are added (Table 11b)
the sa.e general trend is evidenced; however, the 6-18 m
interval in the backyard received a greater average
density than the 0-6 . interval owing to the presence of
Features 3, 10, and 53. The 6—18 . interval in the
frontyard also shows a .arked increase over the 0-6 m
range as a result pri.arily of Feature 57.

During the twentieth century, the backyard in general

received a greater average density between 0—18 . than the

107

Table 11. Contingency Tables--Average density per .3

 

 

distance

fro! house backyard frontyard

0-6 . 1.12 0.09 1.21
(0.79) (0.42)

6-18. 2.10 1.60 3.70
(2.43) (1.27)
3.22 1.69 4.91

x2 = 0.52, df = l, °‘= 0.05, (n) = expected frequency

b. ELI! 19th c.t

 

 

 

 

distance

froa‘hoaae backyard frontyard

0-6 . 25.17 3.31 28.48
(18.18) (10.30)

6-18. 35.92 31.30 67.22
(42.91) (24.31)
61.09 34.61 95.70

x2 = 10.58, df = 1, -‘= 0.05, V2 = 0.11,

(n) = expected frequency

c. 20th cg:

distance

fro! hoaae backyard frontyard

0-6 . 17.73 4.75 22.48
(13.57) (8.91)

6-18. 14.95 16.70 31.65
(19.11) (12.54) ‘
32.68 21.45 54.13

x2 = 5.49, df = l, °‘= 0.05, V2 = 0.10, (n) = expected
frequency
* = Date ranges .inus nails, bottle glass, and features

108

frontyard; however, the 6-18 . interval in the frontyard
shows a higher density than the 6-18 m distance in the
backyard (Table 10c and Table 11c). This holds true for
both cases where features are excluded and included.

Selected functional types were also explored in this
manner in order to better understand possible associations
in refuse patterning. Initially, the .ean frequency
(i.e., the nu.ber of artifacts under consideration divided
by the nu.ber of units involved) of all non-feature
artifacts .inus architectural items was plotted on a graph
where the y axis = .ean frequency and the x axis =
distance fro. the house in meters. Figure 12a and b
present the results for the backyard and frontyard,
respectively, for this category. In general, it appears
that in the backyard the mean frequency decreases as
distance fro. the house increases; while in the frontyard,
frequency increases as distance increases. However, there
is .ore variation fro. this trend in the backyard,
especially in the 9 to 20 . range.

A chi-square test of this relationship between
backyard and frontyard at 0-6 m and 6-18 . was atte.pted
(Table 12a): however, the sample size (even when reduced
by using average density per .2) was too large to produce'
meaningful results (i.e., a x2 = 84.96 with df = l was

obtained which exceeds the critical value of chi-square,

109

     

1| 2| 3| 4| m
o. W distance from house

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

|24||l| 2| 3| 4| In
W!)

Figure 12 . Non-feature artifacts minus architectural iters

110

Table 12. Contingency Tables--Average density per m2

 

 

 

a. o - eat t ct c c a to

distance

from hoaae backyard frontyard

0-6 m 232.30 49.59 281.89
(181.96) (99.93)

6-18. 106.18 136.30 242.48
(156.52) (85.96)
338.48 185.89 524.37

x2 = 84.96, df

l, °‘= 0.01, (n) = expected frequency

 

 

 

be . -- '-

distance

fr on house backyard frontJLard

0-6 . 19.34 2.10 21.44
(17.25) (4.19)

6-18. 22.68 8.10 30.78
(24.77) (6.01)
42.02 10.20 52.22

82 = 2.20, df = l, °‘= 0.05, (n) = expected frequency

lll

 

distance from house

 

Figure 13. Non-feature ceramic tableware

112
at a 0.001 level of confidence, by approximately eight
ti.es). In general, the average density in the backyard
within 0-18 . was almost twice as high as that found in
the frontyard (i.e., 338.48/m2 to 185.89/m2,
respectively).

Ceralic tableware was also exa.ined since a
scattergra. of actual frequency and distance from the
house appeared to show the same 'backyard decrease in
frequency as distance increases versus a frontyard
increase in frequency as distance increases’ relationship,
as described above. Figure 13a and b present graphs of
the .ean frequency of cera.ic tableware by distance from
the house for the back and front yards, respectively. In
general, the trend appears to hold true, although in the
backyard the 0-20 . interval shows high .ean frequencies
overall with the highest values at 10 and 11 .eters. This
is likely influenced by the presence of the for.er
detached kitchen in this vicinity even though the feature
asse.b1age was not included in this sample.

A chi-square test utilizing the average density per
m2 of cera.ic tableware for 0-6 . and 6-18 . in the
backyard and frontyard produced a x2 = 2.20, df = 1 (Table
12b). A null hypothesis, asserting that there is no
significant difference between distance fro. the house and
the backyard/frontyard division of space, would be

retained. In general, however, the average density per m2

113
in the backyard is approxisately four times greater than
that found in the frontyard within 0-18 m (i.e., 42.02/m2

compared to 10.20/.2, respectively).

Chapter 4

INTERPRETATIONS

This study of the Cedar Oaks housesite has focused on
three aspects--the architectural history of the house,
utilization of space, and refuse disposal patterns. The
first aspect has already been discussed in detail through
the integration of archival, oral, architectural, and
archaeological sources. Therefore, the e.phasis of this
chapter will be on the interpretations of the latter two
aspects of this investigation as these were the pri.ary
factors used to exa.ine the central hypothesis of this
thesis (see Chapter 1).

The utilization of space, on this housesite, through
ti.e was studied in order to better understand the
changing function of this site (i.e., fro. a town
residence to a farmstead). Robert Heeler (1978: 10) has
noted that "the homelot is i.portant because it was a
center of human do.estic activity,” and further that

People create functional divisions of space. Areas

are bounded conceptually and physically in an effort

to specify spaces for particular activities (Keeler

1978: 14).

This aspect of human behavior can be .anifested

archaeologically by structural re.ains such as walls,

fences, and outbuildings, which can serve as spatial

114

115

dividers, and by less tangible re.ains such as the
presence of a midden in one area and not in another
(Keeler 1978: 7). There were fenced areas at Cedar Oaks,
with the major fenceline defining the open houselot
boundaries on the northern, western, and southern sides
(Figure 5). The present wire fence is twentieth century
in origin, and there is photographic evidence of an
earlier picket fence located in this sa.e position ca.
1909 which was likely late nineteenth century in origin
(Minnerly 1983: 30).

The opening in this fence serves as an entranceway to
the site, funneling traffic towards the front door. This
was reinforced during the late nineteenth and early
twentieth centuries by a brick walkway which extended out
fro. the front door towards the opening in this fence
(Cleland and McBride 1983: 353).

Other known fenced areas on this housesite functioned
as pens for ani.als and/or perhaps protected gardens from
ani.als. These were pri.arily located in the backyard
area during the late nineteenth and early twentieth
centuries (Cleland and McBride 1983: 354). There is some
evidence of a possible fenceline along the southeast
portion of the site which .ay have served to divide the
houselot from the adjacent fields (Cleland and McBride

1983: 348).

116

The sheet midden found on this site has also been
shown to be an indicator of utilization of space.
Specifically, the backyard contains the deepest and best
developed .idden concentrated primarily in an area bounded
by the house, the two kitchens and the smokehouse (Figure
6). This appears to have served as the focal area for the
majority of activities on this site. Furthermore, the
sheet .idden thins out in depth and artifact density in
the front and side yards, indicating areas kept relatively
clean throughout the occupation span of this house.

The major distinction in spatial utilization
evidenced on this housesite is the dichoto.y between the
frontyard and backyard. This is paralleled in Robert
Keeler’s (1978: 45, 72, 135) study of the St. John’s
housesite in St. Mary’s City, Maryland, and even though
this particular site dates fro. the seventeenth century
there are comparisons which can be drawn with later sites
such as Cedar Oaks.

In particular, Heeler (1978: 45) notes that "the
division of the yard into front and back is partly a
matter of convenience, but also seems to have been a
culturally .eaningful distinction," and further that "this
two yard division of space was part of English cultural
tradition in the seventeenth century and is still apparent
today" (Keeler 1978: 135). The frontyard would serve as

a "forecourt" kept relatively clear of refuse and often

117
formalized in plan; whereas, the backyard functioned as a
"service area" containing the majority of outbuildings,
activity areas, and sheet .iddens (Keeler 1978: 49, 72,
135).

At St. John’s the frontyard was found to be enclosed
by fences and kept relatively clean. It was clearly
distinct fro. the "cluttered” backyard which contained
nu.erous outbuildings, fenced areas, trash pits, a privy
and a sheet .idden (Keeler 1978: 49). A comparable
Tidewater region housesite exhibited a pattern wherein the
outbuildings were grouped around the backside of the house
creating a somewhat enclosed service area (Keeler 1978:
134-135).

At Cedar Oaks this frontyard/backyard distinction was
defined by various types of evidence. The only
outbuildings discovered within the confines of this
housesite were located to the back of the house up to a
.aximum of 30 . away (Figure 7). Further.ore, a well-
developed sheet .idden was "enclosed" by these
outbuildings grouped at the back of the house. This area
appears to have been the pri.ary locus for do.estic
activity.

The frontyard, on the other hand, exhibited a shallow
sheet midden, low artifact density, and no evidence of
outbuildings or specific do.estic activity areas. It is

indicated that this was an area of moderate usage which

118
was likely hoed or swept clean throughout much of the
site’s history. The peri.eters of the frontyard show a
.idden, lighter in color than that in the backyard, which
appears to represent the deposition of frontyard
sweepings.

Further.ore, the major features found in the
frontyard area are related to a garden enclosure and
walkway which served to define and formalize space, as
well as having functioned as an esthetic presentation to
visitors. A trash pit in the northwest corner of the
frontyard is the only detraction fro. this dichoto.y;
however, it was used primarily in the late nineteenth and
twentieth century occupations and .ay have been .asked by
orna.ental vegetation. The remains of a shrubbery and
flower garden still exist in this corner of the yard
(Cleland and McBride 1983: 318).

Chi-square analysis indicates that this dichotomy was
statistically significant primarily in the late nineteenth
century (Table lOa-c and Table lla-c). However,
distribution maps (Figures 9, 10, and 11) and the
comparison of average artifact densities per m2 (Table 5
and Table 9a-b) indicate the sa.e general trend throughout
the site’s history (see Chapter 3).

As the function of this site changed fro. a town
residence (ca. 1848-1870) to a far.stead (ca. 1870-1940),

it can be hypothesized that this resulted in a change in

119

the utilization of space relative to the occupants’
changing needs. In order to examine the changing layout
of this site comparisons were made to Kenneth Lewis’
(1977) three model classification which he used to deter-
.ine the function of the Kershaw house in Camden, South
Carolina. This was a late eighteenth to early nineteenth
century housesite, and it was investigated according to
criteria defined for plantation, far., and town residence
.odels. Lewis (1977: 40) proposed that the function of a
site would be reflected in the nature and arrangement of
structures and activity areas associated with the site.

Of these three models, only the far. and town resi-
dence models could be comparable to Cedar Oaks. The
defining criteria for the farm model includes a compact,
square arrangement of outbuildings to the rear of the
house, with the outbuildings facing inward and the house
facing away fro. this hollow square. The area within this
square might be subdivided into smaller parts, and the
house is likely to be adjacent to and facing a .ajor road
(Lewis 1977: 52).

The town residence, on the other hand, should exhibit
a simpler layout with the house located in front of all
outbuildings and facing away from them. The outbuildings
will be to the rear or side of the dwelling and may be
arranged in a contiguous row. The house should lie along

a through road and the borders of the property deaarcated

120
by fences or walls. Furthermore, evidence of a formal
garden may be present behind the house (Lewis 1977: 52).

Cedar Oaks fits neither .odel for all defining
criteria, although it most closely resembles the farm
.odel. The layout of this housesite through ti.e is as
follows. During the .id-nineteenth century it consisted
of the house, a detached kitchen to the rear of the house,
and possibly a s.okehouse to the southeast over 20 . away.
No other structures are known to have been associated with
this site during this ti.e period.

The s.okehouse is known to have existed in the late
nineteenth and early twentieth centuries and it is only a
possibility that it dates fro. the earlier occupation
period. Lewis (1977: 67) has noted that agricultural
processing structures (such as s.okehouses) are not
normally associated with a town residence. Further.ore,
"oral testi.ony strongly indicates that [s.okehouses] were
constant features of southern hill country far.steads, no
doubt because salting and smoking were the only plausible
.eans of preserving .eat in the southern cli.ate" (Cleland
and McBride 1983: 375).

In the late nineteenth to early twentieth centuries
the detached kitchen was destroyed and an ell, built on
the northeast corner of the house, functioned as the
kitchen during the twentieth century. After this e11 was

removed, a shed was constructed approximately 6 . east of

121
the northeast corner of the house. This small structure
is still standing.

The smokehouse was in use during the late nineteenth
century, but it was torn down so.eti.e during the early
years of the twentieth century. A bricklined well likely
served as the primary water source during the late
nineteenth century, while an artesian well, located to the
west of the house and down the slope, was the twentieth
century water source.

Two small barns located 130 . northwest of the house
served as the only known far.-re1ated outbuildings during
the twentieth century. The existing privy is also
twentieth century in origin and it is located in the
northeast corner of the backyard.

In comparison with Lewis’ (1977) far. .odel, Cedar
Oaks never contained the co.plex of outbuildings that
.ight be expected according to this model. It is likely
however, that the smokehouse and detached kitchen did
enclose a so.ewhat rectangular service area, and both .ay
have faced in towards the house, with the latter facing
away from them. Furthermore, the house did, and still
does, face a through road.

As for the town residence model, the housesite,
during the Barton period, did exhibit a simple layout
wherein the outbuildings were to the rear of the house

which, in turn, faced a through road. The borders of the

122
property .ay have been demarcated by a fence, however,
this is only positively known for the later occupations.
A for.el garden was present but this also was a late
nineteenth to early twentieth century feature and,
further, was located in front of the house. Kitchen
gardens were present in the backyard throughout the site’s
occupation.

Plausible reasons why Cedar Oaks is not well defined
by either .odel may be suggested. It has been concluded
through archival data that this house functioned as a town
residence during the mid-nineteenth century.

Specifically, the various owners were involved in
mercantile operations in Barton (see Chapter 3). However,
since this town never succeeded in establishing itself as
a full-fledged town in a for.el, permanent sense, it would
follow that residences likewise never developed into
"typical” town residence configurations.

As Barton declined in the 1860s-1870s, the function
of Cedar Oaks began to change to that of a s.all
far.stead. During much of this ti.e it was possibly
leased out and transient tenants may not have needed, or
have had the resources, to construct per.anent
outbuildings. After the Civil War the local econo.y was
severely disrupted and the ensuing crop-lien and tenant
farm syste. served to restrict far. size, diversity, and

profitability (Cleland and McBride 1983: 87, 97-110,

123
Doster and Weaver 1981: 117, 122-123). That Cedar Oaks
functioned as a far.stead during the late nineteenth
and twentieth centuries there is little doubt; however,
its s.all size and scope would have lessened the need for
a relatively co.plex layout such as that described by
Lewis’ .odel (1977: 41-42).

Perhaps a closer parallel can be drawn to the study
of the late nineteenth and twentieth century farmsteads in
the Bay Springs area of Mississippi. Investigators noted
a general pattern of far. layout wherein the outbuildings
for.ed both an inner and outer circle in relation to the
house. The outbuildings were arranged according to
function with the inner circle including a well,
s.okehouse, and chicken house, and the outer circle having
barns, vehicle sheds, and ani.al pens (Smith et a1. 1982:
240).

It was further noted that

The outer circle of outbuildings... was oriented

towards the production and storage of income related

activities like cash crops and animal husbandry

(barns, ani.al pens, cotton houses, corn cribs) with

lesser a.ounts going to the household. The inner

circle of outbuildings was mainly oriented toward the
production and storage of subsistence products

(s.okehouse, chicken house, garden, stor. cellars,

orchards, well) for household consumption (S.ith et

a1. 1982: 240-241).

The inner circle tended to range only 35-40 m out fro. the
house, while the outer circle covered a .uch wider area

(S.ith et a1. 1982: 241-242).

At Cedar Oaks the inner circle of outbuildings and

124

(activity areas has been evidenced. It includes the two
ikitchens, s.okehouse, well, gardens, and s.all animal
pens, and all are within 30 m of the house. The mid- to
late nineteenth century occupations’ inner circle was
comprised of the detached kitchen, s.okehouse, well, and a
garden off the southeast corner of the house. The outer
circle during this time period is not known.

The late nineteenth and early twentieth centuries saw
an inner circle comprised of a well, privy, shed, e11
kitchen, s.okehouse (for a time), a garden in the
southeastern portion of the backyard, an orchard along the
southern slope, and at least one small animal pen. The
outer circle consisted of two small barns approxi.ate1y
130 . to the northwest.

Further investigation of site utilization through
ti.e involved the exa.ination of the for.etion processes
which created the archaeological record. Cultural
processes included discard, loss, and abandon.ent as
evidenced by specific refuse du.ps, a sheet .idden, and
architectural re.ains. Post-depositional .odification of
this record included limited plowing, architectural
salvaging and re.odeling, ani.al rooting and foot traffic
(especially in the backyard), and yard sweeping.

In order to better understand the effect that the
change in site function had on the formation of this

record, attention was focused on the elucidation of

125

patterning in disposal behavior. By integrating all of
the data presented in Chapter 3, the following general
conclusions can be drawn concerning this patterning.

During the .id-nineteenth century, occupants disposed
of refuse away fro. the house and pri.arily in the
backyard area. Feature 21 is an exception, as this area
did receive sa.e refuse during this period (Table 4).

Higher frequencies were found in the area which
served as the focal point for the .ajority of site
activities bounded by the back of the house, the detached
kitchen to the north, and the s.okehouse to the east
(Figure 6). The .aterial found here .ost likely
represents the by-products of activities rather than
purposeful disposal.

The lack of a dense a.ount of .id-nineteenth century
.aterial, as well as the lack of specific du.ping areas
during this ti.e period, .ay be the result of re.oval of
refuse to a specific co..unity du.p while Barton was a
functioning town. Such a practice has been noted in oral
accounts fro. other nineteenth and twentieth century sites
(Ada.s 1980: 187, S.ith et a1. 1982: 32), and it has been
theorized that gullies away fro. housesites and wells were
used as pri.ary trash receptacles during Barton’s
existence (Cleland and McBride 1983: 380). Archaeological
evidence at several Barton sites shows that unlined wells

were used in this .anner during the 1850s and 1860s (Ibid.).

126

Another plausible explanation is that the yard area
was generally kept clear of refuse by sweeping the ground
clean. The depositing of .id-nineteenth century .aterial
along the yard peri.eters, as well as underneath
structures such as the rear gallery, likely represents the
sweeping of refuse out of .ajor traffic pathways to less-
used areas (Figure 9). The front and side yards, in
particular, were kept clear of refuse during the .id-
nineteenth century occupation.

Oral history fro. this area indicates that keeping a
"clean” yard was a source of pride to occupants (McClurken
and Anderson 1981: 128, 345, 507, 842, 1074, 1181).
Sililar testi.ony concerning Bay Springs far.steads in
northeastern Mississippi, indicate that yards were often
scraped clean with a hoe or swept (S.ith et a1. 1982: 53,
57).

During the late nineteenth century occupations of
Cedar Oaks, a shift began to occur in refuse disposal. In
fact, it is during this ti.e period that the difference
between the frontyard and backyard division of space, and
distribution of .aterial, beco.es statistically
significant (Table 11a and b). In general, overall
artifact density increases, and the spatial distribution
intensifies and widens to include all investigated areas
of the site. The area of greatest deposition re.ains the

focal area in the backyard (Figure 10), and the sheet

127

midden here has its greatest development during this time
period.

Refuse is now discarded in greater amounts up against
the back of the house, in the vicinity of the detached
kitchen and smokehouse, as well as in specific dumping
areas (i.e., Features 53 and 57, Figure 7). The frontyard
begins to accumulate more refuse than during the .id-
nineteenth century, although it is still .uch "cleaner"
than the backyard (Figure 10). This indicates that while
effort was still put forth to .aintain an esthetic
appearance, it was not to the same degree that it had been
during the previous occupation period.

This intensification of deposition is due, in part,
to the changes in technology, production, and .arketing of
containers (glass in particular), which resulted in
greater availability, less reuse, and increased
disposability of items (Moir 1982: 148). It is noted,
however, that the de.ise of Barton, coupled with the Civil
War, served to disrupt and depress the local economy. It
was not until the last quarter of the nineteenth century
that .ass produced ite.s, such as bottles and canned
foods, beca.e widely available in this area (Cleland and
McBride 1983: 379).

Despite these factors, it can be noted that there is

less effort overall to keep the yard, especially in the

128
back and up against the house, clear of refuse accumu-
lation than during the previous occupation period.
Furthermore, specific dumping areas within 10 m of the
house (i.e., Features 53 and 57) were now being heavily
utilized. The bricklined well, which may have been late
nineteenth century in origin, was not used as a receptacle
for trash.

The twentieth century occupation saw greater emphasis
placed on disposal in the peri.eter areas (i.e., Units 91
and 96), specific dumps (Features 53, 57, and 21), and the
area i..ediately surrounding the house (Figure 11). Once
again, the backyard received a higher density than the
frontyard, although the latter still received greater
a.ounts than it did during the .id-nineteenth century
occupation.

The greatest difference between the twentieth and
late nineteenth centuries’ depositional patterns is the
.arked decrease in density in the 6-30 m range during the
twentieth century (Table 9a). This is likely a result of
the disappearance of two major outbuildings (i.e., the
detached kitchen and the s.okehouse) in this area during
the late nineteenth century or very early twentieth
century. The kitchen activities shifted to the northeast
corner of the house, and a garden was created where the
smokehouse once stood. Therefore, the backyard service

area appears to have been used pri.arily for gardens and

129
s.all animal pens during this time period, resulting in
disturbance and .ixing of the .idden deposit and less
activity-related discard.

As Cedar Oaks was not an isolated housesite it is
expected that analogous housesites elsewhere in Barton
would exhibit si.ilar patterning. The Griswold housesite
(22C1807, Site 5448, see Cleland and McBride 1983: 224-
241) was located approxi.ately 300 . to the southeast of
Cedar Oaks (Figure 2). It was constructed in the 1850s by
Janes Griswold who, like the early owners of Cedar Oaks,
was a Barton .erchant. Other Barton period occupants
included the hotel owner, a .inister, and the Barton Ferry
owner (Cleland and McBride 1983: 225).

After Barton’s de.ise, this house was occupied by
various ferry owners and operators on an erratic basis, as
well as by tenant far.ers (McBride 1984: 4). Oral history
indicates that the house was in disrepair and used only
during ti.es of highwater by the 1920s. It survived into
the 19308-1940s when it burned down (Cleland and McBride
1983: 229, 234).

This site was also investigated by intensive
archaeological excavation during Phases II and III of the
To.bigbee Historic Townsites Project. A total of 180 m2
was excavated, and over 38,000 artifacts were recovered I
(Cleland and McBride 1983: 232, 237). This site is

comparable to Cedar Oaks for the following reasons:

130

1. It spans relatively the same occupation periods
(i.e., 1850-1940s), although its later occupations were
so.ewhat more erratic than at Cedar Oaks.

2. It was originally the residence of a Barton
merchant and was later used as a small far.stead and a
ferry operator’s residence during highwater.

3. Excavations uncovered structural remains of the
house, as well as an outbuilding which appears to have
been a s.okehouse (Cleland and McBride 1983: 234).

4. A sheet .idden is present on this site although
it has been modified by plowing to a greater extent than
that at Cedar Oaks (Cleland and McBride 1983: 232).

This housesite has been analyzed by Stephen McBride
(1984: l) in ter.s of the hypothesis that the change fro.
a town residence to a rural homestead resulted in a change
in refuse disposal practices. By focusing on the deposi-
tion and date ranges of bottle glass and refined ceramics,
it was found that the mid—nineteenth century artifacts,
although sparse in density, were concentrated near the
eastern slope (i.e., in the vicinity of the outbuilding)
and on the far northern edge of the site (McBride 1984: 9).

The late nineteenth to twentieth century artifacts,
on the other hand, were concentrated near the house with
twentieth century items in the greatest density. It was
noted that later ite.s were also broadcast over most of

the investigated area (McBride 1984: 9). An unlined well,

131

6 . north of the house, was used as a trash receptacle
during this period (Figure 14, Cleland and McBride 1983:
240).

McBride (1984: 21) concluded that this data indicated
a shift in refuse disposal patterns during the late nine-
teenth century shortly after Barton’s demise.
Specifically, during the town period,

a definite effort was made to deposit kitchen refuse

on the peri.eters of the yard, particularly the

backyard slope. Then, after the town’s demise, a

shift occurred, with less care taken to keep the

central yard clear (McBride 1984: 21).

Analysis of the McGowan housesite (2201807, Site
5442), another long-occupation site dating fro. the Barton
period and located approxi.ately 300 m southwest of Cedar
Oaks (Figure 2), indicated a similar refuse pattern.
Specifically, the .id-nineteenth century .aterial was
found pri.arily in the backyard area (Figure 15) while the
late nineteenth to twentieth century items concentrated
nearer to the house (McBride 1984: 19).

The only structural features found on the McGowan
site were related to the dwelling. An unlined well,
filled with refuse during the .id- to late nineteenth
century occupations, was located 8 m west of the house
(Figure 15, Cleland and McBride 1983: 153, 157). A small

refuse pit was discovered approximately 22 m northwest and

downslope from the house and was also utilized during the

132

 

 

 

Figure 14. Griswold housesite map

133

an... 95.0.50: cages: .mH ousuHm

 

 

l

 

3.832.
Z<>>OOO<<

...:
?

 

 

.5...qu

 

 

 

 

 

 

 

 

 

 

 

134
mid- to late nineteenth century occupations (Cleland and
McBride 1983: 154, 162).

This comparative data indicates that the changing
refuse disposal pattern discerned at Cedar Oaks is not
unique to this site. Further.ore, as the other housesites
demonstrate the sa.e general disposal trend and
occupational history, there appears to be validity in the
supposition that the shift fro. a town setting to a rural
farm setting was a critical factor affecting this change
in disposal behavior.

Specifically, McBride (1984: 21) has proposed that
these changes are ”related to a change in the conception
and use of the yard" where during the town period there
was a fornalized division and use of space between
frontyard and backyard, with the for.er kept clean and the
latter allowed to accu.u1ate refuse. However, in the
post-Barton period this formalization broke down and
refuse was allowed to accu.u1ate in all areas of the yard
(McBride 1984: 22).

Moreover, he notes that

to the Upland Southerner a town meant aggag, while

farmstead and dispersed settleIent were "what .any

would term disorderly and uncouth" (McBride 1984: 22,

after Newton 1974: 150-151).

However, McBride (1984: 22) points out that

the general declining socioecono.ic level at this

site may also have been a factor. Divergent

lifestyles and values along class lines should be

reflected in yard maintenance and use.... However, it
should be remembered that at the Barton sites none of

135

the residents were above a moderate socioeconomic
level, and the change frog the town to rural setting
gas likely more significant (emphasis mine).

 

This is to say that the general socioeconomic levels of
the Barton housesites did not differ greatly even while
Barton was a functioning town.

This area was relatively depressed economically
throughout its entire history and general economic levels
and social classes were never greatly divergent. For
exa.ple, it can be noted at Cedar Oaks that while the
principal resident during the Barton period (i.e., Ja.es
Collins) was one of the wealthier residents, the principal
residents of the post-Barton period (i.e., the Coltranes
and Uithovens) also enjoyed relative prosperity (see
Chapter 3). There was not a great divergence in
socioecono.ic level fro. one period to the next.

Therefore, what socioecono.ic differences there were
in the Barton area, were not as significant as factors
affecting behavior, as they would have been in an area
where a full range of econo.ic stratification was present
(i.e., upper to lower classes). The obvious change in
patterning evidenced on the three Barton housesites
compared in this analysis, coinciding as it does with the
failure of the town and the changeover to a far.ing
community, strongly suggests the i.portance of this change
in function as the .ajor factor affecting the shift in

site utilization and refuse disposal.

136

Comparison with other housesites in this region
further de.onstrates the universality of certain aspects
of this disposal pattern, as well as aspects which are
divergent. The Bay springs farmsteads are late nineteenth
to twentieth century in date range (Smith et al. 1982:
201) and are therefore only applicable to the post-Barton
pattern.

Sheet .iddens were evidenced on Bay Springs sites but
did not, in general, extend beyond 5 . out fro. the house.
It was also noted that ”these .iddens did not contiguously
surround the house but tended to concentrate in backyard
areas near the kitchen” and, further, that ”these areas
were still evident despite the sweeping that was done by
the occupants" (S.ith et a1. 1982: 217). Frontyards
received fewer artifacts and little .idden accu.u1ation
(Ibid.).

Other aspects of disposal on these far.steads
included du.ps and abandoned ite.s. The .ajority of the
dumping areas ranged fro. 15 to 59 m fro. the house, and
refuse burning often occurred in these areas (S.ith et a1.
1982: 225). On one housesite a slope, 15 to 20 m from the
house, was used as the pri.ary refuse du.ping area
(Ibid.).

In contrast to Cedar Oaks, the sheet .iddens on the
Bay Springs sites were not very thick (i.e., less than 5

c.) although one housesite did exhibit a fairly

137

well-developed and preserved .idden which was 10 to 20 cm
thick. These shallower, and less extensive sheet .iddens
are a result, in part, of shorter occupation spans, as
well as possibly lower econo.ic levels which li.ited
access to .aterial goods and promoted recycling and
curation over discard, .ore so than at Cedar Oaks.

One interesting phenomenon was noticed during the Bay
Springs investigation concerning abandoned ite.s. Many of
the sites had areas which contained large a.ounts of scrap
metal, abandoned vehicles, and glass containers. Oral
history indicated that such ite.s were not discarded but
rather being stored (S.ith et a1. 1982: 226). It was
concluded that

so.e of the accumulations of cans, jars, and scrap

.etal which [the investigators] called trash dumps

might in fact have been convenient storage areas

instead.... When the sites were abandoned, these
storage areas became de-facto refuse deposits
occurring on the surface, with or without an

underlying sheet .idden (S.ith et a1. 1982: 226).

Such an area was present at Cedar Oaks consisting of
four 1940s-vintage cars which were stored in the backyard
along the eastern site edge. Felix Uithoven (personal
co..unication, January 2, 1980) indicated that his brother
put the. there with the intention of restoring the.,
however, they were allowed to deteriorate beyond repair

and were ”abandoned.” It was noted that the area

underneath and around these vehicles had accumulated

138
numerous metal fragments and auto parts (Cleland and
McBride 1983: 363).

It is conceivable that the scattering of paint cans
and canning jars along this slope .ay also represent items
intended for recycling but which were subsequently
abandoned rather than purposely discarded.

Co.parisons can also be made with late nineteenth to
twentieth century tenant far.ing sites in the Richland
Creek area of eastern Texas. Sheet .iddens on these sites
were investigated indepth and found to possess
”considerable behavioral integrity” (Moir 1982: 139)
providing infor.ation about site chronology, spatial
variability, and socioecono.ic trends. In general,
.iddens concerntrated pri.arily in the back or side yards
of a house covering an area that was frequently greater
than 1,500 .2 containing a range of 20,000 to over 150,000
artifacts (Moir 1982: 147).

Spatial analysis indicated a general trend where

Artifact densities were often moderate in .agnitude

fro. i..ediately adjacent to a dwelling to distances

between 6 and 8 .eters away. After that point,
densities frequently increased by three to five-fold

and then dropped back down reaching zero about 15 .

away fro. the house. This pattern was most often

observed in the back or side yards of a dwelling
(Moir 1982: 147).

For certain artifact types and date range .aterial at
Cedar Oaks there is a peak in average density per .2 in
the 6 to 18 . range (Table 12b and Table 9a), however,

this is biased by the distribution of excavation units

139

towards structural remains, and the presence of the
pri.ary service area within this range. The major
divergence from the Texas pattern occurs in the backyard
up against the house which received the greatest average
density of artifacts overall (Block 3, Table 7), excluding
the dump features (i.e., Features 53 and 57), pri.arily
during the late nineteenth and twentieth century
occupations.

Therefore, it can be seen that the later refuse
patterning at the Barton housesites contains aspects which
are found on other comparable farmsteads for the late
nineteenth and twentieth centuries. This appears to be
affected to an extent by socioecono.ic level. Moir (1982:
151) has noted that in the Richland Creek area it appears
that ”the absence of intense sheet refuse deposits
indicates higher relative socioecono.ic status" and,
further, "that landowners .ay be better understood from
discrete features (i.e., trash pits) rather than from
their sheet refuse records.” As the Barton housesites
discussed in this chapter, were, in general, all at the
same .oderate econo.ic level, there should be greater
contrasts with higher econo.ic level rural sites than with
each other.

While Barton was a functioning town, Cedar Oaks was
one of the larger faaily dwellings in this town and was

owned by the ”wealthier" segment of that society.

140

The latter can also be said of the Griswold and McGowan
sites. It is possible that more conscious effort and
.oney, in the for. of slaves and hired help, might have
been expended on the upkeep of these houselots. Barton,
as a co..unity, .ay have had designated dumping areas,
such as gullies and/or the river, and refuse would have
been carried farther fro. the houselot before being
deposited.

Further.ore, it has already been noted that living in
a town setting carried a different conception of spatial
utilization, in terms of activities and discard, than a
rural far. setting. The for.er placed an e.phasis on
"order" and a more for.alized division of space (McBride
1984: 21-22, Newton 1974: 150-151), while for the latter
this for.alization broke down becoming .ore ”disorderly”
(Ibid.). Moreover, with the change to a far.stead, the
emphasis shifted to self-sufficiency resulting in an
increase in the activities conducted on the houselot.

After the de.ise of Barton, and the econo.ic and
social disruption of the Civil War, the occupations of
these housesites beca.e .ore sporadic and their function
changed fro. town residences to rural ho.esteads and
far.steads. The declining economy devalued property
status and the later occupants were, in general, from
lower economic levels although there was not a great

divergence fro. the previous period. It is likely that

141

they had less money and time to spend on upkeep of the
houselot, and the need to put .ore land to economic use
(i.e., far.ing and herding) .ay also have decreased the
areas available for non-organic refuse, necessitating
dumping closer to the house.

Another factor .ay lie in the fact that so.e of the
post-Barton occupations at these sites were of a short-
term, tenant nature. Such occupance may have entailed
less attachment to the property and less care for where
refuse was deposited (Cleland and McBride 1983: 365).

Finally, it must be considered that in the late
nineteenth and twentieth centuries there was greater
availability of mass produced items which lessened the
need for recycling and curation of containers, in
particular, than in previous times (Moir 1982: 148). This
resulted in greater discard of bottles and cans, and such
an increase has been noted at the Barton sites (see
Chapter 3, and McBride 1984). However, for a portion of
the Cedar Oaks analysis, bottle glass was removed from
consideration, and the patterning discerned was not
remarkably different fro. that found when this category
was included (see Chapter 3). Rather, this patterning was
sinply intensified.

Therefore, it appears that the important factor in
the changing refuse patterns at the Barton housesites is

not, in actuality, how .uch was being discarded, rather it

142

is gaaga it was being deposited. It has been shown in
this study that a shift in disposal patterns did occur at
Cedar Oaks, and si.ilar Barton housesites, relative to the
utilization and conception of space within these
houselots, with the .ajor shift occurring after the demise

of Barton.

Chapter 5

CONCLUSIONS

The house at Cedar Oaks is the last re.aining
standing structure from the extinct town of Barton. From
all known data it appears to have been constructed ca.
1848, when Barton was platted as a town, and was occupied,
off and on, up until the 19403.

Initially, it functioned as a town residence (ca.
1848-1870) occupied by Barton .erchants. After the town’s
de.ise (1860s-1870s), it functioned pri.arily as a
farmstead. This study focused on the possible effects
this change in function may have had on site utilization
and, in particular, refuse disposal behavior.

It was found that throughout the occupation span of
Cedar Oaks, there was a distinction .ade between the front
and backyards wherein the frontyard exhibited little use
and served as an ornamental presentation area, while the
backyard functioned as the .ain service area of the site.
The latter contained the deepest midden, the .ajority of
artifacts, the .ain refuse disposal areas, and most of the
outbuildings.

There was a slight shift noted in this dichotomy
after Barton ceased to be a town, where the frontyard

began to accumulate a sheet midden (albeit a shallow one)

143

144

and had a trash pit in its northwest corner. However, it
did continue to function as a presentation area, with a
brick walkway and garden enclosure accenting the frontyard
in the late nineteenth and early twentieth centuries.

Two for.er outbuildings, a s.okehouse and a detached
kitchen, were uncovered in the backyard. The area
"enclosed” by these structures and the house, functioned
as the focal point for domestic activities on this site.
It exhibited the deepest sheet midden and the greatest
density of artifacts.

Changing refuse disposal patterns were brought to
light where the earlier deposits were concentrated in the
backyard service area while the frontyard was kept clean.
A shift occurred after Barton’s de.ise, where refuse began
to accu.u1ate over the entire site with concentrations
nearer to the house and in specific du.ping areas. The
latter were located up against the back of the house, over
the northern and eastern slopes (10 to 50 . away fro. the
house), and in a pit in the northwest corner of the
frontyard.

Relating this changing pattern in disposal behavior
to the change in site function, is reinforced when
co.parisons are made with si.ilar housesites in the Barton
townsite. This sa.e general trend in patterning was
evidenced at the Griswold and McGowan housesites which,

like Cedar Oaks, were characterized by long occupation

145

spans, and a change from in-town dwellings to rural
ho.esteads and farmsteads (McBride 1984).

Comparisons with late nineteenth and twentieth
century far.steads within this geographic region,
de.onstrate general similarities with the later-period
refuse patterning and site utilization at Cedar Oaks,
primarily concerning sheet midden deposits. This
indicates some universality to the cultural factors
affecting disposal behavior and site formation during
these ti.e periods in the rural South. These factors
include: socioecono.ic differentiation, where higher
econo.ic level sites appear to exhibit different disposal
patterning than lower level sites (Moir 1982); rural
conceptions of spatial utilization appear to differ fro.
town conceptions (McBride 1984) resulting in diacernable
patterning; and .ass production of goods in the late
nineteenth and twentieth centuries which created greater
availability and disposability of ite.s over earlier
periods.

When these factors were considered in terms of Cedar
Oaks and the other Barton housesites, it was concluded
that the shift fro. a town setting to a rural farm setting
was .ore significant to the shift in disposal behavior
than socioecono.ic and .ass production factors. However,
it was noted that the latter two did play a role in the

formation of these sites.

146

Further analysis of the Barton housesites, using
these factors as test controls, should provide a clearer
understanding of the behavioral patterns which created
these sites. It is hoped that this thesis will serve as
the i.petus for future studies, thereby contributing to a

better overall understanding of past lifeways in the

South.

BIBLIOGRAPHY

BIBLIOGRAPHY

Adams, William H., editor
1980 Waverly Plantation: Ethnoarchaaolagy of a Tenant
Farming Comnunity. Report sub.itted to Heritage
Conservation and Recreation Service under Contract
#C—55026(79). Resource Analysts, Inc.,
Bloo.ington, Indiana.

Binford, Lewis R.
1981 Behavioral Archaeology and the "Pompeii Premise."
Journal of Anthropological Research 37: 195-208.

Brito, Heriberto J.
1983 Narrative Report on the Inspection of a Historic
Structure. Report sub.itted to the National Park
Service, H2217 (SER-PA).

Busch, Jane
1981 An Introduction of the Tin Can. Historical
Archagology 15: 95-104.

Cleland, Charles E. and Ki. A. McBride, editors
1983 anl Hiatorical, Docaaentary, and Archaeological
Investigations of Barton and Vinton, Mississippi:
An Interim Report on Phase III of the Tombigbee
Historic Townsites Project. Report submitted to
the National Park Service, Mid-Atlantic Region.
Michigan State University, Museu., East Lansing.

 

Deagan, Kathleen
1983 Spanish St. Augustine, The Archaeology of a
Colonial Creole Community. Academic Press, New
York.

Doran, J. E. and F. R. Hodson
1975 Mathematics and Co.putera in Archaeology. Harvard
University Press, Calbridge, Mass.

Doster, James F. and David C. Weaver ‘
1981 Historic:§ettleaent in the Upper To.bigbee lalley.
Report submitted to the Heritage Conservation and
Recreation Service, Albequerque under Contract #0-
5714(78). The Center for the Study of Southern
History and Culture, University of Alabama.

147

148

Drucker, Lesley M.
1981 Socioeconomic Patterning at an Undocumented Late
18th Century Low-country Site: Spiers Landing,
South Carolina. Historical Archaeology 15: 58-68.

Hayden, Brian and Aubrey Cannon
1983 Where the Garbage Goes: Refuse Disposal in the
Mayan Highlands. Journal of Anthropolagical
Archaeology 2: 117-163.

 

Howard, Ja.es Murry
1978 Tennessee-Tombigbee Waterway Project. Preli.inary
Sub.itta1, Part 111, Cedar Oaks. Historic American
Buildings Survey.

Jones, Olive
1971 Glass Bottle Push-ups and Pontil Marks. Historical
Archaeology 5: 62-73.

 

Kauf.an, M.
1963 The First Centary of Plastics: Celluloid and its
Sequel. Iliffe Books Ltd., London.

Keeler, Robert
1978 The Ho.elot on the l7th-Century Chesapeake
Tidewater Frontier. Ph.D. dissertation, University
of Oregon. University Microfil.s, Ann Arbor.

Lees, Willia. B.

1980 Li.erick, Old and In the Way: Archaeological
Investigations at Li.erick Plantation.
Anthropological Studies 5, Institute of Archaeology
and Anthropology, University of South Carolina,
Colu.bia.

Lewis, Kenneth E.
1977 A Functional Study of the Kershaw House Site in
Ca.den, South Carolina. Research Maaascript Series
Nu.ber 110, Institute of Archaeology and
Anthropology, University of South Carolina.

Lief, Alfred
1965 A Clqaeup of Closares: Hiatory and Progress.
Glass Containers Manufacturers Institute, New York.

Lofstro., Edward V.
1976 A Seriation of Historic Ceraaics in the Midwest,
1780-1870. Paper presented at the Joint Plains-
Midwest Anthropology Conference.

149

Lorraine, Dessamae
1968 An Archaeologists’ Guide to Nineteenth Century
A.erican Glass. Historical Archaeology 2: 35—44.

 

Martin, Terrance J.

1983 Animal Remains fro. the Cedar Oaks Housesite
(2201809), Clay County, Mississippi. In Oral
Historical, Docu.entary, and Archaeological
Investigations of Barton and Vinton, Missiasippi:
An Interim Report on Phase III of the Tpabigbee
Hiatoric Townaitea:Project. Charles E. Cleland and
Kim A. McBride, editors. Report sub.itted to the
N.P.S., Mid-Atlantic Region. Michigan State
University, Museum, East Lansing, pp. 286-306.

 

McBride, W. Stephen
1984 Changing Refuse Disposal Patterns at the Griswold
Housesite, Barton, Mississippi (1850-1940). Paper
presented at the Annual Meeting of the Society for
Historical Archaeology, Williamsburg, Virginia.

McCallu., Henry D. and Frances T. McCallu.
1965 The Wire that Fenced the West. University of
Oklahoma Press, Nornan.

McClurken, Ja.es M. and Peggy Uland Anderson, compilers
1981 Oral History Interview Transcripts, Tombigbee
Historic Townsites Project. Sub.itted to the
National Park Service, Southwest Region, under
Contract #C-07026. Michigan State University,
Museu., East Lansing.

 

 

Miller, George
n.d. Chronology of Edge Decorated Wares. Manuscript on
file at Michigan State University, Museu., East
Lansing.

Minnerly, W. Lee, editor

1982 Oral Hiatorical, Docaaentary,,andaArchaeological
Investigationagof Colbert, Barton, and Vinton,
Mississippi: An Interim Report on Phase I of the
Tombigaae Historic Townsites ProjectL Volume I.
Report sub.itted to the National Park Service,
Southwest Region under Contract #C-07026. Michigan
State University, Museum, East Lansing.

 

 

1983 Oral Hiatorical, Docuaentary, and Archaeolagical
Inveatigatioaa_of Barton and Vinton, Miasissippi:
An InteriagReport on Phaae II of the Tombigbee

150

Historic Townsites Project. Report submitted to
the National Park Service, Mid-Atlantic Region
under Contract #CX4000-3-0005. Michigan State
University, Museum, East Lansing.

Moir, Randall W.

1982 Sheet Refuse: An Indicator of Past Lifeways. In
Settleaent of the Prairie Margin: Archaeologyaof
the Richland Creek Reservoir, Navarro and Freestone
gaanties, Texaa,1980—1981, L. Mark Raab, Principal
Investigator. Archaeology Research Program,
Department of Anthropology, Southern Methodist
University.

Munsey, Cecil
1970 The Illastrated Gaide to Collecting_Bottles.
Hawthorn Books, New York.

Murphee, L. C. and K. H. Miller
1976 Soil Survey of Clay Coanty, Missiasippi. U.S.D.A.,
Soil Conservation Service in cooperation with

Mississippi Agricultural and Forestry Experiment
Station.

Murray, Priscilla
1980 Discard Location: The Ethnographic Data. American
Antiqgity 37: 490-502.

Nelson, Lee H.
1968 Nail Chronology as an Aid to Dating Old Buildings.
American Association for State and Local History
Technical leaflet 48, giatoryaNewa 24 (11).

 

Newton, Milton B.
1974 Cultural Preadaptation and the Upland South.
Geoscience and Man 5: 143-154.

1980 Cedar Oaks: an Integ;al,a§arly 19th Century
Architectaral Comaosition. Report submitted to
Interagency Archaeological Services-Atlanta under
Purchase Order #A-54l36(80).

Otto, John S.

1977 Artifacts and Status Differences - A Comparison of
Ceramics from Planter, Overseer, and Slave Sites on
an Antebellum Plantation. In Research Strategies
in Hiatorical Archaeolggy, Stanley South, editor.
Academic Press, New York, pp. 91-118.

 

151

Price, Cynthia R.

1979 Nineteenth Century Ceramics in the Eastern Ozark
Border Region. Center for Archaeological Research,
Monqgraph Series 1, Southwest Missouri State
University, Springfield.

Reher, Charles A., editor
1977 Settlement aad Subsistence Along the Lower Chaaa
River: The CGP Sagvey. University of New Mexico
Press.

Schiffer, Michael B.
1972 Archaeological Context and Systemic Context.
American Antiquity 37: 156-165.

1976 Behavioral Archaeology. Academic Press, New York.

1983 Toward the Identification of Formation Processes.
American Antiqaity 48: 675-706.

Scoville, Warren C.
1948 Revolation in Glassmaking. Harvard University
Press, Cambridge, Mass.

Smith, Oliver H.
1957 Automobiles and Motorcycles in the U.S. National
Museum. In United Statea_National MuseuaaBulletin
#213, Smithsonian Institution, Washington, D.C.

Smith, Steven D., David F. Barton, Timothy Riordan
1982 Ethnoarchaeology of the Bay Springs Farasteads: a
Study of Rural American Settleaent. Report
submitted to the National Park Service, Southwest
Region under Contract #C-54059(81). Resource
Analysts, Inc.

South, Stanley
1977 Method and Theory in Historical Archaeology.
Academic Press, New York.

Wegars, Priscilla, and Caroline Carley
1982 "The Very Latest Rage": Design Trends in Twentieth
Century Ceramics. Paper presented at the 15th
Annual Meeting of the Society for Historical
Archaeology, Philadelphia, Pennsylvania.