1
Interlocking Threads:
Examining the Relationships between Cord-Marked
Pottery, Gender, and Communities of Practice in
Western North Carolina
Helen Albea
Senior Thesis
Warren Wilson College
Spring 2012
SOC 410 Directed Research
Advisors
Dr. Christey Carwile
Dr. David Moore
2
Table of Contents
LIST OF TABLES AND FIGURES……………………………………………….……………3
ABSTRACT…………………………………………………………………………….….........4
INTRODUCTION………………………….………..…………………………………….........5
THESIS STATEMENT, RESEARCH QUESTIONS, AND PURPOSE OF STUDY……........9
SIGNIFICANCE AND BENFITS OF STUDY……………………………………….…........10
RESEARCH FRAMEWORK: OVERVIEW OF THE ARCHAEOLOGICAL SITES………..........11
LITURATURE REVIEW……………………………………………………………….……..13
GLEANING INFORMATION FROM CORD-MARKED POTTERY.…………….….……..18
METHODS……………………………………………..…………………….…………….....19
DATA CODING……………………...……………………………………………….……....24
EXPIREMENTAL METHODS…………………………………………………………........26
FINDINGS AND DISSCUSSION OF LARGE VESSEL…………..…..……..……….…....27
SPSS ANALYSIS…………………………………………...…………………………..…....28
CONCLUSION……………………………………………………….…………………..…..40
LIMITATIONS AND DELIMITATIONS…………………………………….…………...…41
QUESTIONS FOR FUTURE RESEARCH…………………………………………..….......42
AKNOWLEDGEMENTS……………………..………………………………………….…..42
BIBLOGRAPHY…………………………………………………...………………………....43
APPENDIX A: CODING SPREADSHEET…………………..……………………………...49
APPENDIX B: ORIGINAL CODING KEY.………………………………………….…......50
APPENDIX C: REFERENCE DIAGRAM FOR TWIST ANGLE AND DIRECTION….....52
APPENDIX D: INSTRUMENT FORM…………………………………………………......53
APPENDIX E: WIDTH OF CORD/NUMBER OF TWISTS PER CENTIMETER KEY…..54
3
APPENDIX F: SPSS KEY...………………………………………………………….…..…..55
List of Tables and Figures
Table 1: Hurley’s Cordage Tension Diagram (Hurley 1979:7)……………………….…..….21
Table 2: Frequency of Pottery Types ……………………………………………….………..29
Table 3: Frequency of Final Twist Directions.……………………………………….……….30
Table 4: Frequency of Final Angle of Twists……………………………………………..….30
Table 5: Frequency of Cordage Hardness………………………...………………….……....30
Table 6: Frequency of Cordage Ply………………………………………………….…..…...30
Table 7: Frequency of Replied Cords……………………………………………………...…31
Table 8: Frequency of Cordage Widths……………………………………....…………...….31
Table 9: Crosstabulation of Final Twist Direction and Pottery Type by Site…………..……...32
Table 10: Statistical Significance of Table 9…………………………………………………33
Table 11: Crosstabulation of Archaeological Site and Cordage Width………………..….....36
Table 12: Statistical Significance of Table 11………………………………….…………….36
Table 13: Crosstabulation of Pottery Types and Cordage Width………………………….…37
Table 14: Statistical Significance of Table 13…………………………………………….....37
Table 15: Crosstabulation of Final Angle of Twist and Cordage Hardness…………………39
Table 16: Statistical Significance of Table 15…………………………………………...…..39
Figure 1: Positive Cast of Cordage Impressions………………………………………...…..21
Figure 2: Measuring Twists Per Centimeter………………………………………….…...…21
Figure 3: Single Ply Cord and Positive and Negative Impressions ……………….……...…22
Figure 4: 2-ply Cordage and Positive and Negative Impressions…………………….…...…22
Figure 5: 3-Ply Cord and Positive and Negative Impressions ……………………….………23
4
Figure 6: Replied Positive and Negative Impressions…………………………………………23
Figure 7: One Side of Partially Reconstructed Dan River Cord-marked Vessel………………27
Figure 8: Pottery Type’s and Final Twist Directions on the Berry Site…………………….…34
Figure 9: Pottery Types and Final Twist Directions on the Warren Wilson Site………….…..35
Figure 10: Pottery Types and Final Twist Directions on the Biltmore Mound Site………..….35
Abstract
Native Americans produced a wide variety of textiles in the Southeast; unfortunately
these artifacts are not well preserved. In Western North Carolina the only evidence of Native
textiles is cord- and fabric-impressed pottery. This research examines prehistoric textile
production on three archaeological sites through the description and analysis of cord-marked
pottery using quantitative data coding, SPSS statistics software, and experimental archaeological
methods. The theoretical foundation for this research is feminist archaeology and situated
learning theory. These theoretical perspectives provide the means to understand communities of
craftspeople and the belief that Native American women were the primary practitioners of fiber
arts. This study thereby opens a discussion on the lives and social and economic contributions of
prehistoric women in this area. This study tracks variations in cordage construction temporally
and geographically by statistically analyzing S- and Z-final twist directions of cordage in relation
to pottery typologies and site locations. Labor intensity of cordage construction is also addressed
via statistical analysis. A non-statistically significance tendency toward S-twist exists both
temporally and geographically suggesting a possible community of practice in Western North
Carolina. The labor intensity necessary for cordage construction is moderate and employed a
practical use of both time and labor.
5
Introduction
Who were they? What was their daily life like? What would my role have been if I was
born among them? These are the questions I always ask in the back of my mind whenever I first
hear of an ancient civilization or culture. I often try to relate to the past by seeing myself there.
What is more relatable than daily life? I am sure my fascination with the mundane seems strange
from the outside but the truth is these are the moments that fill up most of a human life and are
something that everyone through time has in common. Even if the routines and actions are
different, the needs they sustain are the same. So naturally, when I first visited the Berry and
Warren Wilson sites —or heard about in the case of the Biltmore Mound site—a question
appeared in my mind, just above the horizon of my subconscious, “If I grew up and lived in this
Native American town what would my life be like?”
Archaeology is the best way to answer questions like these for it studies the objects left
behind by past societies. But it takes more than just a general study of the artifacts on the Berry
or Warren Wilson site to fully answer my question. I not only want to know what life was like, I
want to know what it was like for female-bodied individuals. This requires research on preColumbian Native American culture with a feminist archeological perspective. What follows are
the pieces of the puzzle I have been able to put together so far.
Women were key economic contributors to their societies throughout the Southeast. Most
of the cultures were matrilineal so women’s importance was both material and social (Braund
1990, Rifkin 2005). They were involved in every aspect of food preparation, they both
harvested/hunted wild plants and animals and tended gardens and/or large fields and were often
responsible for cooking, tanning hides and other meat processing activities (Adovasio et al
2007:268, Bridges 1989, Koehler 1997:223-224, Claassen 1997:69-70). We know this from
6
archaeological evidence such as Patricia Bridges’ study on stresses on long bones from
Mississippian agricultural and archaic hunter-gatherer societies (Bridges 1989). She found that
women’s arm and leg strength increased in the agricultural population, which indicates use of
mortar and pestle and other shifts in labor related to agriculture (Bridges 1989, Claassen
1997:69-70). They were also skilled craftspeople that made pottery and all manner of textiles
(any fiber-based item of material culture) from baskets to clothing (Adovasio et al 2007:268,
Koehler 1997). Their handiwork left its subtle mark on the archaeological record along with the
food they prepared and the crops they tended. While there is variation from population to
population as to who was primarily responsible for creating and maintaining different objects of
material culture, it is generally agreed that Native American women produced textiles and
pottery (Braund 1990, Hurcombe 2000 Koehler 1997: 224). While pre-historic pottery studies
are fascinating and equally valid in terms of women’s arts, this research will not be focusing on
pottery production. This study concerns cord-marked pottery—the intersection between pottery
and fiber arts—with an emphasis on cordage (i.e. the individual elements of twisted or spun
fibers that make up cloth and textiles).
One of the reasons spinning and weaving in particular is so commonly done by women is
its compatibility with multi-tasking. None of the materials involved are fragile or dangerous,
cordage can easily be picked up or left behind without any damage to the project, and it requires
minimal attention once you get the hang of it. Producing the textiles for themselves and their
families and/or communities was therefore intermixed with other responsibilities that made up
daily life. These tasks included planting, tending, and harvesting domesticated crops, gathering
wild crops, hunting small game, making pottery, processing meat and hides, cooking, and caring
7
for children (Adovasio et al 2007, Braund 1990, Bridges 1989, Hurcombe 2000, Koehler 1997
Claassen 1997).
This means that women on the Berry, Biltmore Mound, and Warren Wilson sites were
probably spinners, weavers, potters, farmers, and gatherers, some were specialists but probably
most were masters of all these tasks in addition to cooking and childcare. They were also
included in hunting and war parties and capable of flint knapping (Claassen 1997, 68-71,
Koehler 1997: 224). Despite this information it is very difficult to find evidence of the women as
individuals. The plants they harvested and the children they raised are long gone and the pottery
and stone tools can only tell us so much about individual moments in time. So how can we reach
them? This is where the magic of cordage and fabric (the objects made from either spun or
unspun fibers such as nets or woven cloth) comes in. Cordage and cloth display within their
construction every decision the artist made during the entire process from when to splice in new
fibers in individual threads, to the over-all pattern of the fabric, how to tie off the ends and how
to repair holes (Eriksen et al 2000, 69). For example, nets can be made in a variety of ways from
looping, weaving, or knotting. In other words, cordage and cloth are records of thought
processes, which are in turn interrelated to cultural traditions and communities of craftspeople.
Pottery does not display the same kind of detail because the stages of its construction are usually
smoothed over leaving only the exterior decoration and clay type as cultural markers.
Unfortunately all of the cordage from the Berry, Biltmore Mound, and Warren Wilson
has long since decayed along with the plants these women worked so hard to gather. This
situation is not unique to those two sites, in fact very few textile artifacts associated with the
humid American Southeast survived to reach the hands of modern archaeologists because these
plant-based artifacts often decompose soon after disposal. By plant-plant based I mean that most
8
Native American cordage was made of wild plant fibers such as dogbane/milkweed, fibrous tree
barks, stinging nettle and other locally available fibrous plants. Organic matter such as this does
not take long to decompose if left to the elements so we have very little surviving fiber artifacts
outside of the occasional burial or rock shelter. The Spiro Mound in Oklahoma, the rock shelters
of the Ozark Mountains, and Hopewell culture burial sites in Ohio are some of the handful of
sites that contain intact fabric remains; unfortunately this study does not have access to any
physical cloth or cordage. All we have left of Native American textiles at the sites in question are
the impressions of cordage left on potsherds. Cord-marked and fabric-impressed pottery was
made throughout the American Southeast and East coast so information about Native fiber arts in
this area is often obtained from these types of pottery. Though this information is obviously less
detailed than that from actual fiber remains it is still useful and can provide archaeologists with a
great deal of information. One of the best examples of this is the cultural information represented
in the final twist direction of cordage. The final twist direction is the clockwise or
counterclockwise direction that a spinner twists her yarn to make a final product. This can
indicate communities of practice that in turn indicate ethnic boundaries which will be discussed
later on.
Fabric-impressed and cord-marked pottery is commonly found on most Southeastern
Woodland (roughly 600 BC to AD 1000) and Mississippian-period (circa AD 900 to AD 1600)
sites. The Biltmore site has a majority of Woodland artifacts. The Warren Wilson and Berry sites
contain majority Mississippian artifacts. The Berry site contains comparatively rare instances of
Woodland objects. The cord-marked pottery from these sites is therefore the basis for the
following study.
9
The above example represents the long twisting road an archaeologist must follow in
order to gain insight into the past. We find evidence of an individual’s handiwork, which is also
a representation of the larger cultures artistic traditions, hidden in an impression in a broken
piece of another handcrafted object. Using these humble objects I will attempt to answer at least
a small part of my ever-present questions. Who were they, what was their daily life like, and
what would my role have been if I were born among them?
Thesis Statement, Research Questions, and Purpose of Study
This study examines prehistoric Native American textile production in Western North
Carolina through the description and analysis of cord-marked pottery using situated learning
theory and feminist archaeological perspectives. The goal of this research is to answer the
following questions:
1)
What did pre-historic cordage in Western North Carolina look like i.e. what are
its most common attributes and did they change over time?
2)
Are there any culturally indicative patterns represented in the frequency of Sand Z-final twists in relation to temporal or geographic association that could
indicate a pre-historic Western North Carolinian cordage making community of
practice?
3)
How labor intensive was it to make cordage in prehistoric Western North
Carolina?
This research explores the design elements and methods of production used to create
Native American cordage, and through this knowledge gains insight into who made it, how, and
why. More specifically, this study looks for traces of culturally distinct methods used to craft
these cords and how that community of practice relates to others within the Southeast. I believe
10
this perspective is the most valuable theoretical approach for the Southeast because we have so
few physical fiber remains with which to conduct materially based studies. We must therefore
rely on design elements as the basis of our studies because these attributes remain in cordage
impressions long after the original fiber has rotted away.
As previously emphasized, another fundamental part of this study is its relation to the
lives of women in Western North Carolinian Native American society. Rebecca Kugel and Lucy
Murphy make the excellent point that “The field of women’s history can also benefit greatly
from more culturally accurate interpretations of Native women’s experiences.” (Kugel and
Murphy 2007:xv). This cultural accuracy is what I seek out by focusing on fiber arts, which
directly links back to Native American women’s lives. Cultural specificity is especially needed
because most of the work on women’s association with textiles is based on the European
experience. This study examines what women’s work in pre-historic America looked like by
studying cord-marked and fabric-impressed pottery from the Berry Site near Morganton, The
Biltmore Mound Site in Asheville, and the Warren Wilson site in the Swannanoa valley. This
research contributes to our knowledge of textiles and Native American women’s role in preColumbian Western North Carolinian society.
Significance and Benefits
This study is significant because it is the first study of cordage for the Berry,
Biltmore, and Warren Wilson sites or for any sites in Western North Carolina as a whole.
Cordage and other fiber-based artifacts are essential items of material culture of ancient Native
Americans (Adovasio et al. 2007: 242). Therefore I would argue that a study of cordage and/or
textiles should be included in every discussion of a Native American site or people group. For
this reason the current study is an important component of the overall body of archaeological
11
knowledge pertaining to this geographic region as well as the individual sites. It also contributes
to the general pool of data pertaining to the Warren Wilson, Biltmore, and Berry sites and could
serve as a useful starting point for a more intensive study on the subject.
Research Framework:
Overview of Archaeological Sites
The present study focuses on cord- and fabric-impressed pottery and builds on the
literature pertaining to the three sites in Western North Carolina. These sites are relevant for my
study because they represent a fairly diverse array of geographic and temporal components of
Western North Carolinian prehistory. The Biltmore Mound site is a woodland period sites
located in the Appalachian Mountains and are associated with Cherokee ancestors. The Warren
Wilson site is also located in the mountains and associated with Cherokee ancestors but contains
mostly Mississippian artifacts. The Berry site is located on the Piedmont and is associated with
Catawba ancestors. To more fully address the geographic and temporal aspects of my research
questions I must also associate my findings with indicators of regional cultural chronologies.
Ceramic typologies (along with radio carbon dates and other artifact typologies) are often used to
identify cultural chronologies on a given site in this area thereby better pinpointing who lived on
a given site and when. By identifying what pottery series and type within that series each of my
cord-marked sherds belongs to I was able to make rough inferences about when and what culture
groups were making the cordages I discuss in this paper. The cord-marked pottery series I
analyzed in this study were Burke, Cowans Ford, Uwharrie, Dan River, Connestee, and
Swannanoa. Burke and Cowans Ford are Mississippian Piedmont pottery dating from A.D. 14001600 and represented the Mississippian component of my study and were all found on the Berry
site (Moore 2002). Uwharrie and Dan River represented the Late Woodland component of this
12
study dating from A.D. 800-1200 and A.D. 1000-1450 respectively (Archaeology of North
Carolina: The Woodland and Mississippian Periods in North Carolina 2010:
http://rla.unc.edu/ArchaeoNC/time/wood_pied_L.htm). Both are found on the Piedmont and in
small amounts on the Berry site. The Connestee series is the middle Woodland component of
the study dating from A.D. 200-800. Connestee sherds from both the Warren Wilson and
Biltmore Mound site were used, and occurred on these sites in abundant amounts. Swannanoa
pottery is the early woodland pottery used in this study dating from 1000-300 B.C. and was
found on both the Warren Wilson and Biltmore sites. Each of these sites is discussed in more
detail below.
The Warren Wilson site is a village site on the banks of the Swannanoa River that was
inhabited by Cherokee ancestors (Dickens 1976). The Warren Wilson site is located on the
Warren Wilson College campus. The site was intermittently excavated as part of the Warren
Wilson College field school starting in 1964-1975, then again from 1978-1983, and finally from
1996 to 2000. Units on the site are still occasionally reopened for archaeology classes. The
Warren Wilson site contains majority Mississippian ceramics from the Pisgah phase. However,
the Woodland period Swannanoa phase pottery found on the Warren Wilson site is known for its
cord-marked and fabric-impressed surface treatment (Keel 1972:246). Keel also indicates that
Connestee series sherds were recovered on the Warren Wilson site (Keel 1972:237). Extensive
excavations of the site reviled overlapping palisade walls representing different habitation phases
primarily during the (early Mississippian) Pisgah phase (Dickens 1976). The majority of the
sherds for this study came from pit features and subsoil excavated during the 1996, 97, and 98
Warren Wilson College field seasons.
13
The Biltmore Mound site is a small village site with an adjacent short wide platform
mound. The mound was a ceremonial center built in three stages by the Connestee culture (North
Carolina’s Hopewellian contemporaries) over a 200-year time span from A.D. 400 to A.D. 600
(Kimball et. all 2010, 44). One of the defining characteristics of Connestee pottery is cordmarked surface treatment. Cord-marked sherds were the second most common Connestee
ceramic type found on the site (Kimball et. al. 2010, 49-50). The pottery I analyzed from the
Biltmore mound was from pit features and surface contexts collected during David Moore’s
survey of the site in 1984.
The Berry site is located in the North Carolina Piedmont on Upper Creek eight miles
north of Morganton in Burke County NC. The Berry site is a Mississippian period site inhabited
by Catawba Indian ancestors rather than Cherokee. It is also the village where Juan Pardo built
fort San Juan in 1567 in the native town of Joara (Moore et al. 2004). Burke series is the most
common pottery found on the Berry site (Moore and Beck 2002:9). Cord and fabric-impressed
Burke pottery is rare but did make up a fair portion of my sample. Uwharrie, Cowans Ford, and
other less common forms found on the Berry site were also used in my analysis. Over half of the
pottery from the Berry site analyzed in this study came from pit feature, subsoil, and structure
contexts excavated during the 2007-2011 Warren Wilson College field school seasons; the rest
were found in the plowzones. These plowzones resulted from decades of agricultural activity on
this privately owned site.
Literature Review
Theory is the weft that binds together the loose threads of literature pertaining to this
study of pre-historic fiber arts. The theoretical foundations for this study are feminist
archaeology—built on Frances Dahlberg’s pioneering work Woman the Gatherer (1983)—and
14
situated learning theory (Lave and Wegner 1991). These two conceptual spheres intertwine in a
two-ply construction that describes what we know about prehistoric cordage, the women who
made it, its cultural implications, and how we can scientifically study all of this from our twentyfirst-century vantage point. Each of these theories is discussed in detail below.
Gendered interpretation of textile production is interconnected with feminist archaeology.
Fabric and gender are often inextricable because—aside from a few exceptions—ethnographic
studies of historic hunter/gatherer and pre-industrial agricultural communities indicate that
women were the primary producers of textiles all over the world (Adovasio et al. 2007:192,
Hurcombe 2010). This is due, in part, to fiber preparation’s compatibility with child rearing
(Barber 1994:29, 30). Spinning is especially compatible with multi-tasking, in any form, in that it
can be done with minimal attention and while walking or even riding horseback (Barber
1994:37). Fiber’s general association with women has provided an outlet for feminist
archaeological inquiry. Since its introduction in the 70s feminist archaeology has greatly
enhanced archaeological inquiry’s understanding of gender on sites world wide. In other words,
feminist archaeology was transformative for the discipline and. This is the sentiment theorists
Margaret Conkey and Joan Gero were putting forward in their paper on the subject where they
so boldly stated that “[…]Feminist inquiry in archaeology […] [is] committed to changing the
way archaeology is practiced, the way it is presented, and the nature of archaeological
interpretation” (Conkey and Gero 1997: 412).
Jean Lave and Etienne Wenger’s Situated Learning Theory also plays a vital role in this
inquiry. Situated Learning Theory states that all learning takes place through legitimate
peripheral participation in socially located communities of practice. A community of practice is
15
any situation where people are learning skills from one another. The producers of cordage in
ancient society are an example of a community of practice.
Archaeologists have begun to apply situated learning theory to artifact analysis by
identifying which attributes are the result of unconscious motor skills and which are stylistic
choices. Motor skills are indicators of communities of practice because they come from
ingrained taught skills rather than conscious aesthetic choices. These studies show that final twist
direction in cordage and twining direction in fabric are the primary indicators of geographic and
temporal cultural differences (Gilligan 2008, Hayden 2009, Maslowski 1996, Minar 1999,
Peterson 2000).
Through the application of Lave and Wegner’s situated learning theory, changes in
cordage construction can be tracked temporally and geographically by statistically analyzing
certain culturally indicative attributes of cordage. These attributes must be directly tied to craftspecific motor skills that tend to be taught and learned in the same way over time and are usually
not a conscious design choice but rather founded in muscle memory; a break from these motor
skills therefore represents a break from an individual culture’s community of practice. Most
studies of the significance of twist direction in cordage incorporate situated learning theory
(Minar 2000, Maslowski 1996, Peterson & Wolford 2000). The learning processes and related
culturally-unique motor skills are especially useful in regional studies in the United States that
focus on ethnic identity and are of special interest to many scholars who study the evolution of
culture groups. For example situated learning theory has been used to identify the origins of the
Iroquois, and study a long-standing community of practice in the Ozark Mountains (Gilligan
2005, Horton 2010, Peterson & Wolford 2000). These studies draw conclusions largely through
analysis of basic design elements in existing fabric remains and cord- and fabric-impressed
16
pottery. Annette Ericksen, Kathryn Jakes, and Virginia Wimberley open their paper with the bold
assertion that, “textiles embody the sum of individual and societal behaviors, combining
individual decisions and group-derived conventions with social patterns of a given time and
place” (Erickson et al. 2000:69). Though they do not use the term, these group-derived
conventions are an aspect of communities of practice, and individual decisions visible in the
weaving of the fabric are their manifestation. The thought process of the individual who made
the fabric is visible in it, from finite decisions about when to splice in new fibers, to the overall
design plan. These techniques are often passed down from teacher to student and remain
relatively consistent within a culture.
Though it is widely agreed that textiles can tell us a great deal about personal and group
identities and interaction, there is still some debate over which attributes are the most telling and
why (Drooker 2000:2). One such debate that applies to the current study of cordage involves the
culturally indicative attributes associated with direction of twist. C. Jill Minar studied the reasons
for and meanings behind the standardized final twist direction (the twist of the finished product
rather than the twist of the internal elements) of fibers in cordage present among cultural groups
(2000). Her study argues that spinning technique, fiber type, and handedness were not associated
with final twist directions, as was previously accepted. Instead she presents the following four
reasons as to why final twist direction persists temporally and geographically: the teaching and
learning process, automatization of motor skills, efficiency and practicality, and cultural beliefs
about directionality (Minar 2000:96-98). This study challenges assumptions about the
significance of textile remains by introducing new lenses through which to view meanings of
humble cordage. Categories of meaning attributed to twist direction, such as Minar’s provide
researchers like myself with a backdrop for statistical analysis that aids in making sense of clear
17
long-standing patterns found in cohesive groups of artifacts. In fact, her first three categories are
directly reinforced by my findings.
Theories such as Minar’s are especially useful for studies where no physical cordage
exists on a site because methodologically, fiber based studies vary depending upon the type of
artifacts available. When only fabric and cordage-impressed pottery can be found the spectrum
of possible research topics is more limited and lends itself best to descriptive analysis and
discussions of communities of practice. Fortunately most fabric and cordage remains come in the
form of impressions east of the Rocky Mountains in the US, so many studies incorporate this
type of artifact (Drooker 1992, Hurley 1979, Johnson 1996, Kuttruff and Kuttruff 1996,
Maslowski 1996, Peterson and Woldford 2000). As William Hurley (1979) and Robert
Maslowski (1996) demonstrate with their studies on cord and fabric-impressed pottery, specific
information on the ethnic distribution through time can be gleaned through the study of these
artifacts. Unfortunately the two authors do not totally agree in terms of interpretation.
Maslowski—who studied ethic distribution in Virginia using final twist direction found in
impressions—argues that Hurley approaches cordage as merely a decorative attribute of pottery
rather than an important and separate aspect of material culture deserving of study for its own
sake (Maslowski 1996:89). However Hurley still makes valuable contributions to
methodological and descriptive approaches to cord-marked pottery with his detailed cordage
identification manual based on cordage impressions found on sites in Michigan. Hurley provides
a field-guide of sorts that is useful for the current study because of his efficient style of notation
and detailed series of photographs of a diverse array of fabric impressed pottery and how to
identify them. I found Hurley’s book invaluable during my research.
18
Gleaning Information from Cord-Marked Pottery
Mississippian Village Textiles at Wickliffe by Penelope Drooker is one of the only booklength studies on southeastern textiles. In it Drooker describes how she conducted her study in
detail and discusses its implications at length. This study is especially useful for my research not
only because of its depth but also because she gathered her data using fabric-impressed pottery.
She offers a step-by-step description of her methods, which proved extremely valuable for the
data coding and methods section of my research proposal and served as a guide as I conducted
my research. Though Drooker’s focus is textiles rather than cordage, her basic methods of
pottery impression analysis still apply to my study in terms of appropriate size of sherd,
significant attributes of cordage and most importantly the cultural implications those attributes
convey.
In her introduction she points out many of the ideas that first inspired me to do this
research, such as the simultaneous rarity and importance of fiber based artifacts. She points out
that even though we have abundant evidence through impressions on pottery that the Native
Americans of the southeastern US had a complex textile industry, most site reports dismiss these
objects as footnotes and most images of pre-Columbian people still depict them wearing animal
skins. This is the degree to which we do not understand the scope and detail of the textile arts in
prehistoric native cultures of the southeast. Penelope Drooker seeks to fill that void by drawing
on every resource she can find within the literature on the few surviving textile objects from that
time period in addition to ethnographic research. She applies this knowledge to her study of
fabric-impressed potsherds found in abundance on a mound site in Kentucky. In this way her
bibliography is as useful to my own research as her methods and instrumentation sections.
19
She briefly mentions that final angle to twist is “one indicator of the amount of labor
involved in fabric production” because this shows how loose or tight the cordage is twisted
(Drooker 1992, 48). Hurley measures the tightness of cordage by comparing the width and
number of twists per centimeter. I use both of these approaches and compare the two in the hopes
that it would give me the clearest indication of labor intensity. I measure labor intensity based on
the assumption that all the cordage represented in my sample is handmade without the use of
tools. Mississippian spindle whirls are extremely rare but not unheard of according to Drooker
but there has been no evidence of such artifacts on any of the sites in my sample (Drooker 1992,
159).
While Drooker’s research outlines the wealth of information available in regards to labor
intensity, Peterson and Wolford’s 2000 study discuss how best to interpret cultural boundaries
through cord-marked pottery. For cord-marked pottery studies final twist direction is often the
most telling in terms of cultural boundaries because this aspect of cordage construction is more
often than any other unconsciously chosen based on how the individual was taught to make
cordage rather than intentional design decision (I was taught to make Z-twist cordage and find it
very difficult to make S-twist cordage for example). Peterson and Wolford focused on Woodland
Iroquois ancestor populations in the Northeast and found that costal groups made Z-twist
cordage and interior populations made S-twist (Peterson and Wolford 2000, 112). As a
subconscious motor-skill final twist direction is not considered a stylistic choice but rather a
result of the learning process.
Methods
In this research I describe cordage, and attempt to identify patterns in cordage
construction found on the Berry, Biltmore and Warren Wilson sites using purposive sampling of
20
cord- and fabric-impressed pottery. The sampling frame is based on sherd size (sherds 1 x 1
centimeter or larger). I used a quantitative coding spreadsheet to recode cordage attributes, and
experimental archaeological techniques designed as a reference and visual aid for identifying
different attributes of cordage construction. The coding and methods used in this study identify
the construction of cords and fabric by breaking them down into individual aspects. The
terminology used to describe these aspects is best understood in the context of how the cord is
constructed. Each cord is made of fibers twisted together or spun either to the right or left, which
is the initial spin direction. These single-ply yarns are then spun together, usually in the opposite
direction (to prevent unraveling) of the initial twists, to make cordage; this is the secondary or
sometimes final twist. More cords can be added to this in a variety of ways, such as twisting two
or three or even four larger cords together to make rope or overlaying a smaller cord over a
larger one to re-enforce the later. I will discuss each of the cordage and potsherd attributes I
analyzed in detail below.
Because twist direction is best studied “in conjunction with other diagnostic artifacts
whenever possible” (Maslowski 1996:96), and because pottery is an artifact associated with
regional cultural chronologies and culture groups I first recoded the pottery type for each sherd
during my analysis. I then began to identify the following cordage attributes final twist direction
(S- or Z-twist), width of cord in millimeters, number of twists per centimeter, cordage spin
(single yarn or combined cords)1, ply (single, 2, 3, 4, etc.), re-ply (2:2, 2:3, 3:2, etc), final angle
of twist, and hardness of cords. When I discuss cordage width, tightness, angle of final twist,
final twist, and ply number to in my analysis the final cord is what I am referring rather than the
1
I then identified the spin of the cordage which I later realized was redundant and is encompassed by cordage ply so
I did not include any of my single/combined cordage tables in the findings section of this paper.
21
internal elements. Data was collected by first taking a positive cast of each sherd with modeling
clay (See figure 1).
Figure 1: Positive Cast of Cordage Impressions
Figure 2: Measuring Twists Per Centimeter
The number of twists per centimeter was taken from the positive cast using a centimeter
scale printed on a transparency so I could lay it directly onto the positive impression (See Figure
2). The twists per centimeter were used in conjunction with the width of the cord millimeters—
measured with a ruler—to find the cordage tension/hardness as per Hurley’s diagram (see table
1).
Tensions of Cord
Diameter of Cord in
Millimeters
Soft Cords
0.5
1
Medium Cords
0.5
1
2.5
Hard Cords
0.5
1
2.5
4
Very Hard Cords
0.5
1
2.5
4
Table 1: Hurley’s Cordage Tension Diagram (Hurley 1979:7).
Twists Per Centimeter
7.2
5.6
9.6
7.2
1.6
12
9.6
2
2
16
12
3.2
2.4
22
The diagram is lacking many of the possible cord width and/number of twist per
centimeters combinations so I had to construct a table that filled in some of these gaps. (See
Appendix E). Because the relation between hardness and twists per cord width/ is based on a
range of possible combinations and follows no clear formula I constructed my table in as logical
a manner as possible by making sure my calculations remained consistent with those of the
Hurley’s table (for example, a 3 mm cord is hard if it has 2 twists per cm because a 2.5 mm cord
requires 3.2 twists and a 4 mm requires 2).
The next group of variables identified is ply types. Single-ply cords are identified as
those not combined with other cords (see figure 3). Plied cordage means any cord that is made up
of two or more single cords twisted together, usually in the opposite direction of the initial twist
to prevent unraveling. So two Z-twist single cords would make a S-twist 2-ply cord (see figures
4 and 5). Re-Plied cords are those that are made of twisted together plied cords, such as 2 twoplied cords twisted together to make a single cord or rope. This type of cord is called a 2:2 cord
(see figure 6).
Figure 3: Single Ply Cord and
Positive and Negative Impressions
Figure 4: 2-ply Cordage and Positive
and Negative Impressions
23
Figure 5: 3-Ply Cord and Positive
and Negative Impressions
Figure 6: Replied Positive and
Negative Impressions
Final angle of twist was measured using Penelope Drooker’s diagram (See appendix C)
(Drooker 1992:45). I printed the diagram onto a transparency and put it on top of the positive
cast to get as accurate an assessment of the twist angle as possible. The diagram is designed for
S-twist cords but I had no trouble ascertaining the angle of Z-twist cords when carefully
comparing their positive casts beneath the transparency.
To select sherds for the coding process I use purposive sampling to select the cordmarked potsherds that were at least one by one centimeters in diameter. Some sherds did not
have clear impressions of cords, i.e. they lacked discernible yarns/components and twist
directions. These sherds were a part of my sample but I was unable to code any of their cordage
attributes.
To give the analysis process more structure I prioritized pottery recovered from postholes
and features such as hearths and trash pits on the sites and analyzed them first, and then looked
through potsherds found in the sub-soil and plowzones. The cordage assemblages on the
24
particular sites have more significance than their specific locations on those sites. Cord and
fabric-impressed pottery—more so with the latter—are unusual enough on these sites that I
analyzed all the sherds I found that fit the size and clarity of impression criteria. To examine
each sherd, I took a positive cast of the impressions with modeling clay to fully identify the
clarity and desirability of the sherd for my study and used the positive cast rather than the
impressions as my primary source of measurements.
Throughout my research and the instrument construction process I reference Hurley’s
descriptive analysis of cord-marked pottery, as outlined in Prehistoric Cordage, Drooker’s
Mississippian Village Textiles at Wickliffe, and Emery’s The Primary Structures of Fabric
(Hurley 1979, Drooker 1992, Emery 1966). I use the photo references and analysis discussions in
these texts the way that a naturalist uses a field guide. I use my coding spreadsheet to create
descriptive data of cordage represented in my sample of ceramic artifacts (See Appendix A). I
use Hurley’s style of notation to record the few instances of complex cordage (i.e. replied cords)
(Hurley 1979:7). When I encounter the more rare net- and fabric-impressed pottery, I only
analyze the individual cords using the above-listed methods. There was not sufficient time
allotted for this study to conduct any sort of analysis of fabric structure.
Data Coding
My coding forms for this research is based primarily on Kuttruff’s fabric complexity
index, specifically her yarn structure and fabric patterning sections and David Moore’s pottery
analysis spreadsheet (Kuttruff 2000, Moore 2002). Constructing the coding forms was one of the
most complicated aspects of my research. They went through many incarnations beginning with
an almost unaltered version of Kuttruff’s fabric complexity index that reads like a questionnaire
with a side column for coding. Dr. Moore helped me to see the inefficiency in this setup and
25
suggested I collect my data using a spreadsheet format. That was the first of many overhauls of
my coding sheet. Once the spreadsheet was set up choosing what variables best operationalize
my research questions was the next challenge. I chose the variables outlined by Kurttruf and
began creating a coding key that encompassed all the possible attributes for cordage in this area
resulting in the list of attributes discussed above. Creating the key was the first step that took
much longer than I anticipated because I was unfamiliar with the pottery types from this area. I
therefore had to do a good deal of research very quickly to make sure all the possible pottery
series and types for the three sites were represented. I then had to become familiar with the
attributes of all these series and types (21 in total). I ended up only identifying six series with
certainty (Burke, Cowans Ford, Uwharrie, Dan River, Connestee, and Swannanoa) and the rest I
listed as either Woodland or Mississippian indeterminate.
It took collecting a page worth of data to see the flaws in my spreadsheet. These flaws
resulted in the removal of the number of beads per cm and the unspun columns from the
instrument because these attributes were too difficult to discern in the impressions and turned out
to be unnecessary. I first removed the number of components/beads column from my spreadsheet
because I initially did not find any replied cords and when I did come across this cordage type
the number of beads were difficult to see and rare enough to be unnecessary information. This
was also true of unspun cords. I found I could not differentiate between an unspun cord and a
weathered surface on the potsherd that had lost its clear impressions of twist. Due to this
uncertainty I rejected all sherds that did not represent clear twisting elements and removed the
unspun column from my spreadsheet. Lastly I had to add in a notes column because I found I
was writing all over the margins. I also added a hardness column after I collected all my data so
that I could calculate the hardness based on the twists per centimeter in conjunction with the
26
width of the cord. After I had recorded all my data I realized that single/combined cords were
encompassed by the ply column of my spreadsheet. Therefore, even though I did enter spin data I
did not use it in my final analysis and instead relied on the ply number to convey this
information.
Once all was coded I used a more refined system that left out all undiscovered attributes
such as 4-ply cords in preparation for entering my data into SPSS. I also documented some noncoded information on an extra form in but found this extra form cumbersome and therefore did
not use it for many sherds (see Appendix D). In addition to recording data I took photographs of
especially interesting and diagnostic potsherds and the positive impressions of the cordage or
fabric.
Once my data was compiled with the final version of my spreadsheet I began entering
data into IMB’s SPSS (Statistical Package for the Social Sciences). This was a very simple and
straightforward process though I did have to create a new key (See Appendix F) catered to the
codes I used in SPSS where they differed from my original key. This is partly because I did not
find all the pottery types or replied varieties that I included on my first key so it was pointless to
code those into SPSS. Another example of why I had to change the coding when I entered my
data into SPSS is that had to code all indeterminate variables as 1 to make the data tables
uniform. There were nine variables in total (See Appendix F) I then ran frequency and crosstab
analyses to find answers to my research questions.
Experimental Methods
Though I did not conduct an experimental study I did use some experimental methods to
help me visualize and contextualize my data. Without replicating the cordage for this study I
would have failed to confidently recognize three-plied cordage and how to differentiate replied
27
from plied cordage impressions. I initially wanted to collect local wild Indian hemp/dogbane but
was unable to locate a wild stand in the allotted time. Instead I ordered a pound of raw flax with
which to conduct my experimentation. I replicated 2-ply, 3-ply 3:2, and 2:2 re-plied cordage and
pressed these cords into modeling clay to verify my observations of pre-historic potsherds. I also
use cordage replication as a tool to more accurately understand cordage hardness as described by
Hurley. I did this by constructing a 2mm 2-ply cord that varied from soft to hard. This also
reinforces my later finding that hard cords are more labor intensive because it takes much more
concentration and causes a great deal more hand cramping to produce a hard cord than a medium
or soft cord. The replied cords also take a lot more mental and physical effort and dexterity to
create than the 2- or 3-plied cords constructed from single-plied elements.
Findings and Discussion of Large Vessel
Figure 7: One Side of Partially Reconstructed Dan River Cord-marked Vessel
28
I am very fortunate to have access to a nearly complete reconstructed cord-marked vessel
for this study. The over 55 sherds that make up the Dan River vessel were found together in
feature 84 in 2007. This vessel provides me with the most complete representation of a cohesive
cordage assemblage. By comparing the cordage clearly impressed on 27 sherds that made up the
vessel I am able to describe three distinct cordage types. The first is a 2-ply S-Twist cord that
varies from .75-3 millimeters in width averaging 1.6 mm. It also varies from a soft to a medium
cord in strength and tightness. This is represented both by the number of twists per centimeter
when compared to the average width that varies from 2-3 twists per cm, and the angle of the
twist which varies from <10 to 25 degrees. This cord is the most common cord represented and is
most likely intentionally stamped on the vessel as a surface treatment. The other two cordage
types impressed on the vessel are less common and only occurring on a few of the sherds on one
side of the vessel. On three of the twenty-seven sherds analyzed a re-plied S-twist cord made up
of two smaller Z-twist cords appears intermixed with a heavily over-stamped section of the twoply S-Twist cord. This cord is 1.5-3 millimeters wide with a twist angle of 10 to 25 degrees. This
is still probably a rather inconsistent cord varying from soft to hard depending on the width. The
most interesting thing about this cord is it appeared to be unraveling on sherds 20, 21, and 22
providing me with a clear impression of the Z-twist cords it is made of. The Z-twist cord is 1 mm
wide with a hardness ranging from soft to medium and a twist angle of 25-45 degrees.
SPSS Analysis
The sample includes 346 sherds, but because some sherds had indiscernible impressions
the sample of cordage impressions analyzed is smaller (N=206); thirty-one percent of distinct
cordage impressions analyzed came from the Berry site (31% n=64), forty percent from the
Warren Wilson site (40.8% n=84), and twenty-eight percent from the Biltmore site (28.2%
29
n=58). I ran frequency analyses using IBM’s SPSS software on all the variables. What follows
is the discussion of each variable.
The overall most common pottery types, quantified by the number of distinct cordage
impressions they displayed, were Connestee cord-marked (47.1% n=97), Swannanoa cordmarked (20.4% n=42), and Burke cord-marked (16.5% n=34) (see table 2). The high frequencies
of these pottery types were expected because they are very common on the sites they are
associated with. The less common pottery series and types are Mississippian Indeterminate
(6.8% n=14), Woodland Indeterminate (2.9% n=6), Burke Fabric-impressed (1.5% n=3),
Cowans Ford Cord-marked (1.5% n=3), Uwharrie Net-impressed (1% n=2), Uwharrie Cordmarked (.5% n=1), and Cowans Ford Fabric-impressed (.5% n=1). There were only three distinct
cordage impressions on the Dan River vessel discussed above making those only account for
1.5% of cordage analyzed. All of the minority pottery types except for Woodland Indeterminate
were found on the Berry site and are associated with Piedmont pottery traditions.
Valid
Mississippian Indeterminate
Woodland Indeterminate
Swannanoa Cord-Marked
Connestee Cord-Marked
Uwharrie Cord-Marked
Uwharrie Net-impressed
Dan River Cord-Marked
Burke Cord-Marked
Burke Fabric-Impressed
Cowans Ford Cord-Marked
Cowans Ford Fabric-Impressed
Total
Frequency
14
6
42
97
1
2
3
34
3
3
1
206
Percent
6.8
2.9
20.4
47.1
.5
1.0
1.5
16.5
1.5
1.5
.5
100.0
Cumulative
Percent
6.8
9.7
30.1
77.2
77.7
78.6
80.1
96.6
98.1
99.5
100.0
Table 2: Frequency of Pottery Types
The most common cordage attributes in the sample were also easily interpreted from the
frequency tables (see tables 3-8). Sixty-six percent of the cords were S-twist (66% n=136), fiftyeight percent had a twist angle of 10-25 degrees (58.7% n=121), ninety percent of cords were 2-
30
ply (90.8% n=187), sixty percent of cords were soft (60.7% n=125), and fifty-eight percent of
cords were between 1 and 1.9 millimeters in width (58.7% n=121). The rarity of replied (2.4%
n=5) and single spun cords (5.3% n=11) is also notable and seems to indicate a tendency toward
neither overly simple nor overly complex cords (see tables). This data indicates that the most
common attributes of cordage in Western North Carolina were soft, S-twist, 2-ply, 1-1.9
millimeter cords, with a twist angle of 10-25 degrees.
Frequency
Valid
Indeterminate
S-Twist
Z-Twist
Total
3
136
67
206
Percent
1.5
66.0
32.5
100.0
Cumulative
Percent
1.5
67.5
100.0
Table 3: Frequency of Final Twist Directions
Frequency
Valid
Indeterminate
<10 degrees
10-25 degrees
26-45 degrees
>45 degrees
Total
8
15
121
59
3
206
Percent
3.9
7.3
58.7
28.6
1.5
100.0
Cumulative
Percent
3.9
11.2
69.9
98.5
100.0
Table 4: Frequency of Final Angle of Twists
Valid
Indeterminate
Soft
Medium
Hard
Very Hard
Total
Frequency
23
125
37
14
7
206
Percent
11.2
60.7
18.0
6.8
3.4
100.0
Cumulative
Percent
11.2
71.8
89.8
96.6
100.0
Table 5: Frequency of Cordage Hardness
Valid
No Ply
2 Ply
3 Ply
Total
Frequency
11
187
8
206
Percent
5.3
90.8
3.9
100.0
Table 6: Frequency of Cordage Ply
Cumulative
Percent
5.3
96.1
100.0
31
Valid
Not Applicable
2:2
3:2
Total
Frequency
201
4
1
206
Percent
97.6
1.9
.5
100.0
Cumulative
Percent
97.6
99.5
100.0
Table 7: Frequency of Replied Cords
Valid
< 1mm
1-1.9 mm
2-2.9 mm
3-4 mm
Total
Frequency
59
121
21
5
206
Percent
28.6
58.7
10.2
2.4
100.0
Cumulative
Percent
28.6
87.4
97.6
100.0
Table 8: Frequency of Cordage Widths
From this information alone, patterns already begin to emerge that could indicate a
common community of practice in Western North Carolina. Of special interest is the high
quantity of final S-twists and the high number of Woodland phase Connestee and Swannanoa
pottery. From this frequency data I began to suspect that S-twist cordage was a tendency of
Woodland Cherokee ancestors in the Appalachians.
To confirm these initial trends indicated by the frequency data I ran Pearson chi-square
crosstabulation statistical tests to see if there was a relationship between the variables and pottery
types on each archaeological site. The most important of these crosstabulations that best
addresses my research questions is the test of association between final twist direction—a known
community of practice indicator—pottery type and site (see tables 9 and 10).
S-twist was the most common twist form on all three sites making up fifty-seven percent
of the cordage impressions from the Berry site (57.8% n=37), fifty-nine percent from the Warren
Wilson Site (59.5% n=50), and eighty-four percent from the Biltmore Site (84.5% n=49). S-twist
32
is also the most common twist direction in most of the pottery types making up seventy-one
percent of Connestee Cord-marked impressions (71.1% n=69), sixty-nine percent of Swannanoa
Cord-marked impressions (69% n=29), sixty-seven percent of Burke Cord-marked impressions
(67.6% n=23), one hundred percent Uwharrie Net-impressed (100% n=2), sixty-six percent of
Dan River Cord-marked (66.7 n=2), one hundred percent of Burke Fabric-impressed (100%
n=2), and sixty-six percent of Cowans Ford Cord-marked impressions (66.7% n=2). The few
pottery types that show majority Z-twist impressions are Mississippian and Woodland
Indeterminates (64.3% n=9 and 66.7% n=4), Uwharrie Cord-Marked (100% n=1), and Cowans
Ford Fabric Impressed (100% n=1).
Despite these clear trends, the data for each site is not technically statistically significant
but the overall tendency toward S-twist is. The Pearson chi square test still indicates that there is
an over .05 chance that this is a random occurrence on the individual sites as table 10 illustrates.
This does not mean that the data is altogether irrelevant; the majority S-twist is still a potential
indicator of a temporally and geographically stable community of practice. The creation of Stwist cordage was probably a common motor skill in Western North Carolina both in the
Woodland and Mississippian periods and in the mountains and on the piedmont. With a larger
sample size these trends could be tested further. This is very interesting because it could
potentially indicate a Swannanoa and Connestee influence on Burke Cordage production.
Pottery Type * Final Twist Direction * Archaeological Site Crosstabulation
Count
Archaeological Site
Berry Site (31BK22)
Pottery Type
Mississippian Indeterminate
Woodland Indeterminate
Uwharrie Cord-Marked
Uwharrie Net-impressed
Dan River Cord-Marked
Burke Cord-Marked
Burke Fabric-Impressed
Cowans Ford Cord-Marked
Final Twist Direction
Indeterminate
S-Twist
Z-Twist
0
5
9
0
1
2
0
0
1
0
2
0
0
2
1
1
23
10
1
2
0
0
2
1
Total
14
3
1
2
3
34
3
3
33
Cowans Ford Fabric-Impressed
Warren Wilson Site (31BN29)
Total
Pottery Type
Biltmore Mound (31BN174)
Total
Pottery Type
0
2
0
0
1
1
Swannanoa Cord-Marked
0
37
1
22
27
50
7
1
25
2
10
21
33
3
1
64
3
32
49
84
10
Connestee Cord-Marked
42
6
48
49
9
58
5
2
29
69
0
2
2
23
2
2
0
136
9
4
13
27
1
0
1
10
0
1
1
67
14
6
42
97
1
2
3
34
3
3
1
206
Woodland Indeterminate
Swannanoa Cord-Marked
Connestee Cord-Marked
Total
Total
Pottery Type
Mississippian Indeterminate
Woodland Indeterminate
Swannanoa Cord-Marked
Connestee Cord-Marked
Uwharrie Cord-Marked
Uwharrie Net-impressed
Dan River Cord-Marked
Burke Cord-Marked
Burke Fabric-Impressed
Cowans Ford Cord-Marked
Cowans Ford Fabric-Impressed
0
0
0
1
0
0
0
1
1
0
0
3
Total
Table 9: Crosstabulation of Final Twist Direction and Pottery Type by Archaeological Site
Chi-Square Tests
Archaeological Site
Berry Site (31BK22)
Pearson Chi-Square
Likelihood Ratio
Linear-by-Linear Association
N of Valid Cases
Warren Wilson Site (31BN29) Pearson Chi-Square
Likelihood Ratio
Linear-by-Linear Association
N of Valid Cases
Biltmore Mound (31BN174)
Pearson Chi-Square
Continuity Correctiond
Likelihood Ratio
Fisher's Exact Test
Linear-by-Linear Association
N of Valid Cases
Total
Pearson Chi-Square
Likelihood Ratio
Linear-by-Linear Association
N of Valid Cases
Value
21.145a
18.379
6.134
64
2.936b
3.284
.032
84
1.933c
.829
1.676
4
4
1
.569
.512
.858
1
1
1
.164
.363
.195
1
.168
20
20
1
.006
.210
.057
Exact Sig. (2sided)
.177
1.900
58
39.371e
24.785
3.630
206
a. 23 cells (85.2%) have expected count less than 5. The minimum expected count is .03.
b. 5 cells (55.6%) have expected count less than 5. The minimum expected count is .04.
c. 1 cells (25.0%) have expected count less than 5. The minimum expected count is 1.55.
d. Computed only for a 2x2 table
e. 26 cells (78.8%) have expected count less than 5. The minimum expected count is .01.
Table 10: Statistical Significance of Table 9
16
16
1
Asymp. Sig. (2sided)
.173
.302
.013
df
Exact Sig. (1sided)
.177
34
The overall tendency toward S-twist is made abundantly clear on in figures 1-3. These
graphs show the strong majority of S-twist on Burke cord-marked pottery (Figure 1), the less
remarkable but still general tendency toward S-twist on the Warren Wilson Site (Figure 2), and
the clear preference for S-twist on the Biltmore Site (Figure 3). The comparison of the Warren
Wilson to the Biltmore site is especially interesting in light of the fact that these two pottery
assemblages represent contemporary and culturally similar groups yet they sharply contrast in
twist preferences.
35
Figure 8: Pottery Type’s and Final Twist Directions on the Berry Site
Figure 9: Pottery Types and Final Twist Directions on the Warren Wilson Site
Table 10: Pottery Types and Final Twist Directions on the Biltmore Mound Site
36
I also ran crosstabulations on all the other variables by site and pottery type separately to
test for significant correlations between other cordage attributes and time and geography. Most
of these relationships were not statistically significant. However one variable in particular shows
some interesting statistical significance. Cordage width exhibits geographic and temporal trends
(see tables 14-17). This analysis showed that the Berry site was the only site with a slight
majority of fine cords 1 millimeter wide or narrower (46.9% n=30). The Warren Wilson and
Biltmore sites showed a clear majority of slightly wider cords in the 1-1.9 category (65.5% n=55
and 67.2% n=39). This may also be a temporal relationship despite the fact that the chi-square
test came out with exactly .05 making it just barely insignificant. I still believe this could—with
further research—be a significant trend because Burke pottery shows a clear tendency toward
narrow <1 mm cordage impressions (58.8% n=20) whereas Connestee and Swannanoa sherds
tend toward the slightly wider 1-1.9 mm category (67% n=65 and 61.9% n=26). This could
mean that Mississippian period Piedmont cordage was finer than Woodland Appalachian
cordage and might represent an increase in spinning technique over time.
Crosstab
Count
Archaeological Site
Berry Site (31BK22)
Warren Wilson Site (31BN29)
Biltmore Mound (31BN174)
Total
< 1mm
30
10
19
59
Width of Cord
1-1.9 mm
2-2.9 mm
27
6
55
15
39
0
121
21
Table 10: Crosstabulation of Archaeological Site and Cordage Width
Chi-Square Tests
Pearson Chi-Square
Likelihood Ratio
Linear-by-Linear Association
N of Valid Cases
a.
Value
34.621a
42.049
.083
206
df
Asymp. Sig. (2sided)
6
.000
6
.000
1
.773
3 cells (25.0%) have expected count less than 5. The minimum expected count is
1.41.
Table 11: Statistical Significance of Table 10
3-4 mm
Total
1
4
0
5
64
84
58
206
37
Crosstab
Count
< 1mm
Pottery Type
Mississippian Indeterminate
Woodland Indeterminate
Swannanoa Cord-Marked
Connestee Cord-Marked
Uwharrie Cord-Marked
Uwharrie Net-impressed
Dan River Cord-Marked
Burke Cord-Marked
Burke Fabric-Impressed
Cowans Ford Cord-Marked
Cowans Ford Fabric-Impressed
6
0
9
20
0
0
0
20
3
0
1
59
Total
Width of Cord
1-1.9 mm
2-2.9 mm
8
0
5
1
26
5
65
10
1
0
2
0
2
1
10
3
0
0
2
1
0
0
121
21
3-4 mm
Total
0
0
2
2
0
0
0
1
0
0
0
5
14
6
42
97
1
2
3
34
3
3
1
206
Table 12: Crosstabulation of Pottery Types and Cordage Width
Chi-Square Tests
Pearson Chi-Square
Likelihood Ratio
Linear-by-Linear Association
N of Valid Cases
a.
Value
43.798a
47.608
4.336
206
df
30
30
1
Asymp. Sig. (2sided)
.050
.022
.037
36 cells (81.8%) have expected count less than 5. The minimum expected
count is .02.
Table 13: Statistical Significance of Table 12
I ran crosstabulation analyses comparing angle of twist with hardness to find the labor
intensity of cordage production (see tables 18 and 19). These crosstabs found a large
concentration of soft cords with a final angle of 10-25 degrees on each site with other types of
cordage falling well behind these in percentages. For example this type of cord makes up forty
percent of the total cordage impressions from the Berry site (40.6% n=26) while the next most
common soft 26-45 degree cordage only makes up fifteen percent (15.6% n=10). Similar trends
exist on the Warren Wilson site with soft 10-25 degree cordage making up twenty-seven percent
of impressions there (27.4 n=23) while the second two most common cordage impressions, soft
38
and medium 26-45 degree cordage, only make up fourteen percent of the whole analyzed
assemblage from that site (14.9% n=12). This trend appears again on the Biltmore site with soft
10-25 degree cordage making up exactly fifty percent of cordage impressions analyzed (50%
n=29) and soft 26-45 degree cordage only making up twenty percent of impressions (20.7 n=12).
Statistical analysis indicated that all these trends were statistically significant with the exception
of the Berry. However I believe the majority is strong enough to still show a clear trend toward
soft 10-25 degree cords on the Berry site. The rarity of single and replied cords discussed above
also illustrates the point that the cordage was neither unnecessarily elaborate and strong nor
overly simple and fragile. The cordage and the work put in to producing it seems to be well
suited for practical use of both time and resources. Because cordage was probably all hand made
this preference for practicality over complexity or extreme strength makes logical sense. 26-45
degree cords are the stronger and more difficult to produce of the two because maintaining the
necessary tension for that level of tightness requires a strong grip and more concentration to
maintain an even cord. This is still not nearly as difficult to make as the hard and very hard cords
and those with greater than 45-degree twist angles. These all indicated that cordage was simple
to moderately intensive to create and took just enough effort to produce functional cords of
ordinary strength.
39
Final Angle of Twists * Tension/Hardness of Cord * Archaeological Site Crosstabulation
Count
Archaeological Site
Berry Site
(31BK22)
Final Angle of
Twists
Warren Wilson Site
(31BN29)
Total
Final Angle of
Twists
Biltmore Mound
(31BN174)
Total
Final Angle of
Twists
Total
Total
Final Angle of
Twists
Indetermi
nate
<10 degrees
10-25 degrees
26-45 degrees
>45 degrees
Tension/Hardness of Cord
Sof
Med
t
ium
Hard
0
0
0
Indetermin
ate
3
Indeterminate
<10 degrees
10-25 degrees
26-45 degrees
>45 degrees
Indeterminate
<10 degrees
10-25 degrees
26-45 degrees
Indeterminate
<10 degrees
10-25 degrees
26-45 degrees
>45 degrees
Total
0
6
3
0
12
2
1
2
0
0
5
3
1
1
1
6
8
2
9
4
0
23
3
26
10
1
40
0
6
23
12
1
42
0
2
29
12
43
0
11
78
34
2
125
0
5
2
0
7
0
2
12
6
1
21
0
0
4
5
9
0
2
21
13
1
37
Very
Hard
0
4
1
0
5
0
0
6
3
0
9
0
0
3
4
0
7
0
0
10
4
0
14
0
0
3
4
0
7
Figure 14: Crosstabulation of Final Angle of Twist and Cordage Hardness
Chi-Square Tests
Asymp. Sig. (2Archaeological Site
Berry Site (31BK22)
Value
16.119a
12
.186
Likelihood Ratio
14.539
12
.268
1.918
1
.166
39.199b
16
.001
21.985
16
.144
6.816
1
.009
N of Valid Cases
Pearson Chi-Square
Likelihood Ratio
Linear-by-Linear Association
N of Valid Cases
Biltmore Mound (31BN174)
sided)
Pearson Chi-Square
Linear-by-Linear Association
Warren Wilson Site (31BN29)
df
64
84
Pearson Chi-Square
32.980c
6
.000
Likelihood Ratio
20.958
6
.002
Linear-by-Linear Association
13.483
1
.000
N of Valid Cases
58
Tota
l
3
3
41
16
1
64
2
9
46
25
2
84
3
3
34
18
58
8
15
121
59
3
206
40
Total
73.010d
16
.000
Likelihood Ratio
46.399
16
.000
Linear-by-Linear Association
15.298
1
.000
Pearson Chi-Square
N of Valid Cases
206
a. 17 cells (85.0%) have expected count less than 5. The minimum expected count is .08.
b. 21 cells (84.0%) have expected count less than 5. The minimum expected count is .12.
c. 9 cells (75.0%) have expected count less than 5. The minimum expected count is .31.
d. 17 cells (68.0%) have expected count less than 5. The minimum expected count is .10.
Table 15: Statistical Significance of Table 14
Conclusion
Establishing a community of practice for Western North Carolina is the first step toward
better understanding prehistoric Western North Carolinian cordage. The discovery that both
Cherokee and Catawba ancestors exhibit a preference for S-twist cordage could suggest cultural
interaction between the two but needs more research from a wider variety of sites in the
mountains and the Piedmont to confirm. This is especially interesting because it suggests
possible interaction among Native American women and outlines their handiwork in detail. The
change in cordage thickness over time and across space may also suggest a refinement of
techniques or technological advancement over time or differing techniques and technologies in
the mountains than in the Piedmont. This research shows that the craftspeople that created the
cordage in this area were pragmatic and knew just how to make cords quickly and simply while
still maintaining its strength. The community of practice and the labor intensity come together in
the most common cordage type, the soft S-twist 2-ply cords between 1-1.9 millimeters in width,
and the majority of the cordage displayed on the large vessel. Both are easily made cords that are
stronger than single ply but not rigid or complex like hard or replied cords. This tells us that the
women who walked the earth on our campus roughly 3-2,00 years ago and the generations that
followed across North Carolina learned and taught each other how to make cordage in basically
41
the same way for thousands of years and also passed on the ethic of practicality and efficiency
that comes with a largely self-sufficient society. Because cordage was one of the most useful and
versatile tools used in Native American society, ranging in use from making up the nets used for
catching small animals and fish to lashing together the roof beams of houses. These tools
represent women’s technological ingenuity and contribution to their society. The twist direction
and tightness of the cords impressed in the pottery I analyzed are a record of the thoughts and
actions of women of the past. I am grateful I was given this opportunity to learn to read the
messages they unwittingly left for me all that time ago.
Limitations and Delimitations
This study does not create generalizations about all Native American cordage due to the
small sample size and limited time available to conduct the study. The replication aspect of the
project is not able to address the specific plant species used in cordage production because no
physical remains exist. This study it is not an exhaustive analysis of the cord-marked material
from the Biltmore Mound, Warren Wilson, and Berry sites; there are still many unanalyzed cordmarked sherds in the artifact collections from those sites. This is partially due to the length of
time it takes just to locate cord-marked potsherds amidst the thousands of potsherds in these
collections. This study’s primary focus is description with the overall theoretical themes of
situated learning in communities of practice and feminist archaeology. A true situated learning
study of cordage in Western North Carolina is beyond the scope of this research because it
requires a much larger sample from a broad spectrum of contemporary sites. The intention of this
project is simply to build the groundwork for such a study by observing initial trends in cordage
assemblages in this area and to open a discussion on the roles and lives of women on the
42
Biltmore Mound, Warren Wilson, Berry sites. Studies such as these must first have solid
descriptive data on which to stand and I provide that to the best of my ability.
Questions for Future Research
A larger sample size is needed to confirm the trends discovered in my research. The
construction of a more conclusive understanding of the region’s prehistoric textiles requires the
analysis of more potsherds from a wider variety of sites in Western North Carolina. Studies
primarily focused on fabric- and net-impressed pottery and the plethora of additional information
available through that lens would greatly flesh out our understanding of fiber artifacts in this area
as well. Experimental studies based around time, effort and possible tools and techniques used to
make cordage and textiles would also make a profound impact on this body of research. More indepth studies on settlement patterns in this area and how cordage twist preferences relate to those
patterns would also be an interesting way to build on this research. This type of analysis would
need to heavily rely on cordage twists trends in conjunction with pottery series and lythic
analyses to fully outline trade routes, culturally specific craft-specialization and design
preferences, and other archaeological evidence that denotes cultural interaction. Cordage twist
analyses apply to spirituality-based studies; especially in the southeast where directionality is
know to be a key part of regions ceremonial complex. Access to fiber or cord-marked pots from
burials and mounds with known ceremonial significance would be necessary for such a study.
Acknowledgements
I would first like to thank my advisors David Moore and Christey Carwile for their
continual guidance and support. Laura Vance deserves equal thanks for her relentless guidance
through the proposal writing process prior to the culmination of this research and for offering
help with SPSS and interpreting my findings; without her I would not have been able to
43
statistically analyze and understand my data at all. I would also like to thank Linda Hall for
allowing me access to the Western Office of North Carolina State Archaeology and its Biltmore
Mound pottery collection. Special thanks go to my fellow students both in my Directed Research
class and all those who provided me with encouragement and support and allowed me to discuss
my ideas with them. Of these wonderfully helpful peers I would like to especially call attention
to, Elizabeth Creech at the writing center who helped me with the first draft of the paper, my
roommate Valorie Nichols who is also writing her senior thesis in Anthropology (I am so lucky
to have a roommate who is right there with me pulling all nighters to finish our research drafts),
Kat Kipfer and Maureen Vaughan, my fellow archaeologists, and everyone else on the
Archaeology crew. I am proud and grateful to be situated in the Warren Wilson College
community of learning, without the unique cultural interaction that occurs here I would not be
who I am today and I will carry the teachings of this work college community of practice with
me always.
Bibliography
Adovasio, J. M, Page, J, and Soffer, Olga
2007 The Invisible Sex: Uncovering the True Roles of Women in Prehistory New York:
Smithsonian Books: Collins.
Albers, Patricia, and Medicine, Beatrice.
1983 The Hidden Half: Studies of Plains Indian Women. Boston: University Press of America.
Barber, Elisabeth.
1995. Women’s Work: The First 20,000 Years, Women, Cloth, and Society in Early Times.
New York: Norton.
Binford, Lewis R.
44
1972 Archaeological Perspective. Academic Press Inc.
Braund, K.E.H.
1990 Guardians of Tradition and Handmaidens to Change: Women’s Roles in Creek
Economic and Social Life During the Eighteenth Century. American Indian Quarterly
14(3):239.
Bridges, Patricia S.
1989 Changes in Activities with the Shift to Agriculture in the Southeastern United States
Current Anthropology 30(3):385–394.
Brumbach, Hetty Jo, Robert Jarvenpa
1997 Women the Hunter: Ethnoarchaeological Lessons from Chipewyan Life-Cycle
Dynamics. In Women in Prehistory: North America and Mesoamerica Pp. 17-32.
Philadelphia: University of Philadelphia Press.
Claassen, Cheryl
1997 Changing Venue: Women’s Lives in Prehistoric North America. In Women in
Prehistory: North America and Mesoamerica Pp. 65-87. Philadelphia: University of
Philadelphia Press.
Claassen, Cheryl, Rosemary A. Joyce
1997 Women in Prehistory: North America and Mesoamerica. Philadelphia: University of
Philadelphia Press.
Conkey, Margaret W., and Joan M. Gero.
1997. “Programme to Practice: Gender and Feminism in Archaeology.” Annual Review of
Anthropology 26:411–437.
Dahlberg, Frances.
45
1983 Woman the Gatherer. Yale University Press.
Drooker, Penelope B
1992 Mississippian Village Textiles at Wickliffe. Tuscaloosa: University of Alabama Press
Drooker, Penelope B.
2000 Beyond Cloth and Cordage: Archaeological Textile Research in the Americas. Salt Lake
City: University of Utah Press.
Emery, Irene
1966 The Primary Structures of Fabrics. Washington, Textile Museum
Erickson, Annette; Kathryn Jakes; Virginia Wimberley
2000 Prehistoric Textiles: Production, Function, Semiotics. In Beyond Cloth and Cordage:
Archaeological Textile Research in the Americas Pp. 69-83. Salt Lake City: University of
Utah Press
Gilligan, Shane Patrick
2008 Identifying Communities of Practice through Cord Impressed Ceramics at the Early Late
Prehistoric Sites of Thomas/Luckey and Chenango Point in Central New York Anthropology
Masters Thesis Binghamton University New York.
http://search.proquest.com/docview/304330640?accountid=14472.
Griffin, James Bennett
1952 Archeology of Eastern United States. University of Chicago Press.
Hayden, Anna
2009 Ceramic Production in Middle Woodland Communities of Practice: A Cordage Twist
Analysis in Tidewater Virginia. Honors Senior Thesis for Bachelors of Arts in Anthropology,
College of William and Mary, Virginia.
46
Horton, Elizabeth Temple
2010 The Ties that Bind: Fabric Traditions and Fiber use in the Ozark Plateau. Ph. D.
dissertation, Washington University in St. Louis.
"http://search.proquest.com/docview/757657396?accountid=14472" \n
_blank__http://search.proquest.com/docview/757657396?accountid=14472_.
Hurcombe, Linda
2000 Time, Skill, and Craft Specialization as Gender Relations. In Gender and Material
Culture in Archaeological Perspective. Palgrave Macmillian Press. New York.
Hurley, W.
1979. Prehsitoric Cordage: Identification of Impressions on Pottery. Washington, DC:
Taraxacum.
Jensen, Joan M.
1977. Native American Women and Agriculture: A Seneca Case Study. Sex Roles 3: 423-441.
doi:10.1007/BF00287408.
Keel, Bennie C
1972 Woodland of the Appalachian Summit Area. P.H.D dissertation Washington State
University. University of Microfilms International An Arbor Michigan.
Keel, Bennie C
1976 Cherokee Archaeology: A Study of the Appalachian Summit. The University of
Tennessee Press. Knoxville.
Kimball, Larry R., Thomas R. Whyte, and Gary D. Crites.
2010 The Biltmore Mound and Hopewellian Mound use in the Southern Appalachians.
Southeastern Archaeology 29(1):44–58.
47
Koehler, Lyle
1997 Earth Mothers, Warriors, Horticulturists, Artists, and Chiefs: Women Among the
Mississippian and Mississippian-Oneota Peoples, A.D. 1000 to 1750. In Women Prehistory:
North America and Mesoamerica Pp 211-226. Philadelphia: University of Philadelphia Press.
Kugel, Rebecca and Murphy, Lucy Eldersveld
2007 Native Women’s History in Eastern North America before 1900: A Guide to Research
and Writing. Lincoln: University of Nebraska Press.
Kuttruf, Jenna T and Kuttruff Carl
1996 Mississippian Textile Evidence of Fabric-Impressed Ceramics from Mound Bottom,
Tennessee. In A Most Indispensable Art: Native Fiber Industries from Eastern North
America Pp. 160-173. Knoxville: University of Tennessee Press
Lave, Jean and Wenger, Etienne.
1991. Situated Learning: Legitimate Peripheral Participation. Cambridge University Press.
Maslowski, Robert
1996 Cordage Twist and Ethnicity. In A Most Indispensable Art: Native Fiber Industries from
Eastern North America. Pp. 88-99. Knoxville: University of Tennessee Press
Minar, C. Jill
1999 Impression Analysis of Cord-Marked Pottery, Learning Theory, and the Origins of the
Alachua. Ph. D. Dissertation. University of California, Riverside
Minar, C. Jill
2000. Spinning and Plying: Anthropological Directions. In Beyond Cloth and Cordage:
Archaeological Textile Research in the Americas. Pp. 85-100. Salt Lake City: University of
Utah Press.
48
Moore, David G.
2002 Catawba Valley Mississippian: Ceramics, Chronology, and Catawba Indians.
Tuscaloosa: University of Alabama Press.
Moore, David; Rob Beck
2002 The Burke Phase: A Mississippian Frontier in the North Carolina Foothills.
Southeastern Archaeology. 21;2:192-205. http://www.warrenwilson.edu/~arch/Mississippian_Frontiers.pdf
Moore, David, Robin A. Beck, Jr., Christopher B. Rodning.
2004 Joara and Fort San Juan: Culture Contact at the Edge of the World. Antiquity. 78:299
http://www.antiquity.ac.uk/projgall/moore299/
Newton, Dolores
1974 The Timbira Hammock as a Cultural Indicator of Social Boundaries. In The Human
Mirror: Material and Spatial Images of Man. pp 231-251Louisiana State University Press.
Baton Rouge.
Peterson, James B and Jack Wolford
2000 Spin and Twist as Cultural Markers: A New England Perspective on Native Fiber
Industries. In Beyond Cloth and Cordage: Archaeological Textile Research in the Americas
pp 101-117 Salt Lake City: University of Utah Press
Rifkin, Mark
2005 Representing the Cherokee Nation: Subaltern Studies and Native American Sovereignty
Boundary 2 32(3):47–80.
Univeristy of North Carolina at Chapel Hill College of Arts and Sciences Research Laboratories
of Archaeology.
49
Archaeology of North Carolina. The Archaeology of North Carolina: The Woodland and
Mississippian Periods in North Carolina. Last modified 2010. Accessed April 29, 2012.
http://rla.unc.edu/ArchaeoNC/time/wood_pied_L.htm. Wilson, Gilbert
1917 Buffalo Bird Woman’s Garden: Agriculture of the Hidatsa Indians. Minnesota Historical
Society Press.
Appendix A
#
#
#
#
#
#
#
#
ent
of
Com
pone
nt s
Arra
Not e
s
ngem
Num
b
Twis er of
ts Pe
r cm
Har d
n ess
Aver
age
Diam Yarn
eter
in
mm
ry Ty
pe
Pot te
Prov
Site
AC C
#
Sher
d Co
de #
eni e
nce
Coding Spreadsheet
Spun
Plied
Replied
Final
Twist
Direction
Final
Angle of
Twist
50
Appendix B
Original Coding Key
Cordage Structure key from Kuttruff’s Fabric Complexity Index
Site
14) Dan River Net-Impressed
1) 31BK22 (The Berry Site)
15) Qualla Cord-Marked
2) 31BN29 (Warren Wilson Site)
16) Burke Cord-Marked
3) Other
17) Burke Fabric-Impressed
Pottery Type
18) Cowans Ford
1) Badin Cord-Marked
19) Pee Dee
2) Badin Fabric-Marked
20) Pigeon
3) Yadkin Cord-Marked
21) Pisgah
4) Yadkin Fabric-Marked
22) Other
5) Swannanoa Cord-Marked
6) New Hope Cord-Marked
Number of components/beads 1-10 or
7) New Hope Fabric-Impressed
indeterminate
8) New Hope Net-Impressed
9) Connestee Cord-Marked
Number of twists per centimeter 1-20 or
10) Uwharrie Cord-Marked
indeterminate
11) Uwharrie Net-Impressed
Arrangement of components:
12) Haw River Net-Impressed
13) Dan River Cord-Marked
Unspun
1) Single
51
2) Combined
5) Indeterminate
3) N/A
6) N/A
Spun
Final twist direction
1) Single
1) No Twist
2) Combined
2) S
4) Complex (specify)____
3) Z
5) Other (specify)____
4) Indeterminate
6) Indeterminate
Plied (1-6 or indeterminate)
Final angle of twist in degrees
1) No twist
Replied
2) <10
1) 2:2
3) 10-25
2) 2:3
4) 26-45
3) 3:2
5) > 45
4) Other____
6) Indeterminate
52
Appendix C
Reference Diagram for Twist Angle and Direction
53
Appendix D
Instrument Form
Photo _________________________________
Date __________________________________
Artifact Number_________________________
Provenience____________________________
______________________________________
Potsherd Dimensions_____________________
Yarn Diameter in mm____________________
Measurements taken from:
a) Impression
b) Positive Cast
c) Both__________________________
Twist direction of components using Hurley’s notation style.
References in Literature:
Photo references:
Description:
Comments
54
Appendix E
Width of Cord/Number of Twists per Centimeter Key
Tensions of Cord
Soft Cords
Medium Cords
Hard Cords
Very Hard Cords
Diameter of Cord in Millimeters
Twists Per Centimeter
0.4
0.5
0.6
0.7
0.8
0.9
1
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
2
0.5
1
1.5
1.6
1.8
1.9
2
2.5
4
0.5
1
1.6
1.8
2
2.1
2.5
3
4
0.5
1
2
2.2
2.3
2.5
3
3.5
4
7.2
7.2
7
7
6
5.6
5.6
4
4
4
4
3
2
2
2
1
9.6
7.2
4
4
4
2
2
1.6
1
12
9.6
5
5
3
2.5
2
2
2
16
12
4
3.2
3.2
3.2
3
3
2.4
55
Appendix F
SPSS Key
Pottery Types
1) Mississippian Indeterminate
2) Woodland Indeterminate
3) Swannanoa Cord-Marked
4) Connestee Cord-Marked
5) Uwharrie Cord-Marked
6) Uwharrie Net-impressed
7) Dan River Cord-Marked
8) Burke Cord-Marked
9) Burke Fabric-Impressed
10) Cowans Ford Cord-Marked
11) Cowans Ford Fabric-Impressed
Final Twist Direction
1) Indeterminate
2) S-twist
3) Z-twist
Final Angle of Twist
1) Indeterminate
2) <10
3) 10-25
4) 26-45
5) >45
6)
Site
1) Berry Site (31BK22)
2) Warren Wilson Site (31BK29
3) Biltmore Mound Site (31BK174)
Hardness
1) Indeterminate
2) Soft
3) Medium
4) Hard
5) Very Hard
Spin
1) Single
2) Combined
Ply
1)
2)
3)
4)
Indeterminate
No Ply
2 Ply
3 Ply
Replied
1) Not Applicable
2) 2:2
3) 3:2