<%BANNER%>

Locating the Gift

Permanent Link: http://ufdc.ufl.edu/UFE0024242/00001

Material Information

Title: Locating the Gift Swift Creek Exchange along the Atlantic Coast (A.D. 200 to 800)
Physical Description: 1 online resource (361 p.)
Language: english
Creator: Wallis, Neill
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: anthropology, archaeology, ceramics, exchange, gift, mortuary, personhood, pottery, southeast
Anthropology -- Dissertations, Academic -- UF
Genre: Anthropology thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: LOCATING THE GIFT: SWIFT CREEK EXCHANGE ON THE ATLANTIC COAST (A.D. 200 to 800) This dissertation evaluates anthropological theories of exchange by detailing the history of social interactions along the Atlantic coast of Georgia and northern Florida during the Middle and Late Woodland period (ca. AD 200 to 800). Rare or finely crafted objects are often interpreted by archaeologists as exchanged valuables through which persons and social groups entered into relationships of reciprocal obligation and competition. In contrast, ostensibly mundane artifacts like utilitarian pottery are rarely considered to have held any exchange value. However, a truly unique dataset for studying past social interactions comes from Swift Creek Complicated Stamped pottery that linked sites throughout much of the Eastern Woodlands but was primarily distributed across the lower Southeast. Archaeologists have demonstrated that Swift Creek vessels and the carved wooden paddles used to decorate them were carried long distances across the landscape. The considerable social diversity evident among social groups necessitates studying these interactions through a genealogy of past material practices at multiple intersecting scales. Toward this end, I employ three complimentary methods in the analysis of numerous pottery assemblages from both mortuary mounds and village middens along the Atlantic coast. Instrumental Neutron Activation Analysis (INAA) and petrographic analysis of clay samples and pottery are used to differentiate between local and foreign-made vessels at sites. Technofunctional analysis of pottery documents site-specific trends in the manufacture, use, and deposition of vessels. Together, these data indicate that forms and functions of vessels deposited at mortuary mounds were often different from the limited forms of village wares, but domestic cooking vessels from distant villages were sometimes carried long distances to be finally deposited at mortuary mounds. Rather than the de facto refuse of moving people, the prevalence of foreign-made cooking vessels at mortuary mounds denotes the intentional emplacement of gifts, likely distributed in the context of deaths that obligated repayment of debts in which relationships between descent groups were reworked. Based on the content and distribution of designs on vessels, I argue that exchanged vessels were parts of distributed persons that were extended in social space and time, transforming cooking pots into powerful tools of commemoration, affiliation, and ownership.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Neill Wallis.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Sassaman, Kenneth E.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2011-05-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024242:00001

Permanent Link: http://ufdc.ufl.edu/UFE0024242/00001

Material Information

Title: Locating the Gift Swift Creek Exchange along the Atlantic Coast (A.D. 200 to 800)
Physical Description: 1 online resource (361 p.)
Language: english
Creator: Wallis, Neill
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: anthropology, archaeology, ceramics, exchange, gift, mortuary, personhood, pottery, southeast
Anthropology -- Dissertations, Academic -- UF
Genre: Anthropology thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: LOCATING THE GIFT: SWIFT CREEK EXCHANGE ON THE ATLANTIC COAST (A.D. 200 to 800) This dissertation evaluates anthropological theories of exchange by detailing the history of social interactions along the Atlantic coast of Georgia and northern Florida during the Middle and Late Woodland period (ca. AD 200 to 800). Rare or finely crafted objects are often interpreted by archaeologists as exchanged valuables through which persons and social groups entered into relationships of reciprocal obligation and competition. In contrast, ostensibly mundane artifacts like utilitarian pottery are rarely considered to have held any exchange value. However, a truly unique dataset for studying past social interactions comes from Swift Creek Complicated Stamped pottery that linked sites throughout much of the Eastern Woodlands but was primarily distributed across the lower Southeast. Archaeologists have demonstrated that Swift Creek vessels and the carved wooden paddles used to decorate them were carried long distances across the landscape. The considerable social diversity evident among social groups necessitates studying these interactions through a genealogy of past material practices at multiple intersecting scales. Toward this end, I employ three complimentary methods in the analysis of numerous pottery assemblages from both mortuary mounds and village middens along the Atlantic coast. Instrumental Neutron Activation Analysis (INAA) and petrographic analysis of clay samples and pottery are used to differentiate between local and foreign-made vessels at sites. Technofunctional analysis of pottery documents site-specific trends in the manufacture, use, and deposition of vessels. Together, these data indicate that forms and functions of vessels deposited at mortuary mounds were often different from the limited forms of village wares, but domestic cooking vessels from distant villages were sometimes carried long distances to be finally deposited at mortuary mounds. Rather than the de facto refuse of moving people, the prevalence of foreign-made cooking vessels at mortuary mounds denotes the intentional emplacement of gifts, likely distributed in the context of deaths that obligated repayment of debts in which relationships between descent groups were reworked. Based on the content and distribution of designs on vessels, I argue that exchanged vessels were parts of distributed persons that were extended in social space and time, transforming cooking pots into powerful tools of commemoration, affiliation, and ownership.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Neill Wallis.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Sassaman, Kenneth E.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2011-05-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024242:00001


This item has the following downloads:


Full Text

PAGE 1

LOCATING THE GIFT: SWIFT CREEK EXCHANGE ON THE ATLANTIC COAST (A.D. 200-800) By NEILL J. WALLIS A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2009 1

PAGE 2

2009 Neill J. W allis 2

PAGE 3

To Michelle 3

PAGE 4

ACKNOWL EDGMENTS I have benefited tremendously from many people in the course of this research. I must first thank my dissertation committee. Ken Sa ssaman has been a superb committee chair and advisor, always a stanch advocat e and supporter but at the same time encouraging me to freely pursue my own research. Although I took my own direction in my dissert ation research, it has been heavily shaped by Kens thoughtful guidance a nd intellectual influence in more ways than I can enumerate. I am deeply indebted to him in bringing this research to fruition and in my professional development generally. Susan Gilles pie has also been a significant force in my graduate career. From the outset, she always inspired me to think broadly and theoretically, and helped me to develop my research questions and interpret the wider significance of the data. I owe her a great debt of gratitude for inspiration and for her careful and cr itical reading of my work. Jane Southworth, from the Geography department, was alwa ys helpful in considering the spatial aspects of data, and also shared invaluable advice about life in academia. John Krigbaum was a late, though welcomed, addition to the committee, and brought his characteristic enthusiasm in discussing this research. I must also thank many of my fellow gradua te students and profe ssional colleagues who helped make this research possible. I am foreve r indebted to Keith Ashl ey, who I attribute with initially sending me down the path toward Swift Cr eek research in northeastern Florida. When I met Keith, I had a simple interest in the Swift Creek archaeological culture. In his typical fashion, Keith generously divulged all of his knowledge about Swift Creek and, through our collaboration, brought me up to speed in a way that lay the groundwork for my dissertation. I consider him an honorary member of my committee. Ann Cordell has always been willing to share her expert advice and for this I have been very fortunate. In my first years as a student, Ann dedicated many hours to teaching me about po ttery, eventually employed me periodically at 4

PAGE 5

the FLMNH, and finally, for m y dissertation, an alyzed many petrographic thin sections on the cheap. This work could not have been completed without her. I also gratefully acknowledge my friends and fellow graduate students for many helpful discussions along the way, giving me inspiration or commiseration when I needed it, especially Jamie Waggoner, Debby Mullins, Asa Randall. This research was heavily dependent on extant collections and I si mply could not have carried it out without the generosity and helpfuln ess of all the people and institutions that made loans to me. In addition, destru ctive analysis is often difficult to get approved, and I appreciate the willingness of my colleagues to give me that permission. Many, many of the pottery assemblages from Georgia sites came from Fr ankie Snows voluminous collection at South Georgia College, and, as has often been the ca se, to him I owe my deepest gratitude. These assemblages also came with consent and help fr om Fred Cook, Dwight Ki rkland, and Bill Stead. I also benefited from the use of collections from the University of Georgia with the help of Mark Williams and the University of West Georgia Waring Laboratory with assistance from Ray Crook and Susan Fishman-Armstrong. During my trav els to these institutions, I stayed in the homes of Jim and Marjorie Waggoner and my uncle and aunt, Rick and Mary Jane Taylor. I thank them for their warm hospitality. Over the years, I benefited tremendously from my relationship with Environmental Services, Inc. They granted me full access to th eir collections and, from time to time, provided gainful employment. Brent Handley, Greg Hendr yx, and Greg Smith always kept me informed of any projects that included Swif t Creek, and I am grateful that I was able to excavate at several of those sites and incorporate them into my dissertation. Florida Archeological Services provided an important collection from work on the Greenfield peninsula. I also benefited from 5

PAGE 6

discussions with Bob Johnson a bout archaeological work in the Jacksonville area. The Jacksonville Museum of Science and History pr ovided full access to the Dent Mound collection and Christy Leonard was incredibly helpful in this regard. The Florida Museum of Natural History, with the help of Donna Ruhl and Ann Cordell, granted access to the Mayport Mound collection. Finally, with fundi ng from a John Griffin Grant fr om the Florida Archaeological Council, I was able to go to the National Museum of the American Indian (NMAI) to analyze collections from C.B. Moores excavations on the Lower St. Johns River. I extend my thanks to the extremely helpful staff at the NMAI and also thank my cousin Tom Fry and his wife Michelle (and of course their da ughters Trisha and Ann) for thei r hospitality during my stay in the D.C. area. The use of raw clays was an important part of this research, and I was fortunate to have access to many samples that had already been lo cated and collected. Vicki Rolland sent me a whopping eight different clay samples from various parts of NE Florida and I was overwhelmed by her willingness to share them. Carolyn Rock provided two clay samples from recent Camden county excavations. Brian Floyd sent me clay sa mples from the lower Ocmulgee area of Georgia and also was helpful in identifying some potential paddle matches within the Jacksonville area. Keith Ashley and Buzz Thunen loaned clay from the Grant Mound. Finally, Fred Cook and Bill Stead led me on an expedition along the Altamaha River to obtain clay samples from sources with which they were familiar. Ken Sassaman graciously offered his boat for this adventure. I thank all of these colleagues for th eir assistance in obtaining clay samp les. I was very lucky to have been awarded a University of Missouri Research Reactor (MURR) doctoral internship (s upported by NSF grant #0504015), which opened the door to my use of Instrumental Neutron Activation Analysis for a very large sample. Mike Glascock was 6

PAGE 7

very helpful at every step of m y preparat ion, analysis, and interpretation and I greatly appreciated his enthusiasm and candor. Jeff Fe rguson spent many hours with me as I gained an understanding of GAUSS and other statistical packages. Upon my arrival in Columbia, Matt Boulanger immediately became my tour guide at MURR and the University of Missouri. Matt and I carpooled to MURR, but more importan tly, I benefited tremendously from our many conversations about INAA and other chemical me thods and their use in archaeology. Corinne Rosania showed me the technical side of sample preparation, including supervising me for hours while I (slowly!) learned how to seal vials with the torch. Corinne was also directly responsible for getting all of my samples irradiated on schedu le. Mark Beary and Jacob Masters sealed my long count vials for me before I was certified on the torch. Finally, Michelle Amor gave me an affordable place to stay for my three-month residence in Columbia, without a lease. Besides the cats always wanting to jump in my lap while I was working, I could not have hoped for a better place to live. Michelle was a gr eat vegetarian chef and a fast friend in my months away from home. I also thank John and Tanya (Peres) Le mons for putting me up for a night in Tennessee on my drive up to Missouri. Funding to subsid ize INAA and petrography came from a National Science Foundation Dissertation Improvement Grant (#0744235). From the very beginning my family has b een exceptionally supportive of my goals and achievements. My parents were staunch supporters, always keenly interested in what I was up to and proudly bragging about their son the way all parents should. Although she reneged on her promise to buy me dinner once a week if I came to th e University of Florida, it was really fun to live near my sister Kate for my entire graduate career. I must also thank my soon-to-be affinal kin as well, Jennifer, Sheron, and Doug, for thei r encouragement, interest, and books! I dont 7

PAGE 8

think I could have com pleted this work without Michelles unwavering support, love, and faith in me. She has always been able to put things in to perspective when I n eed it the most. 8

PAGE 9

TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ........12 LIST OF FIGURES.......................................................................................................................14 ABSTRACT...................................................................................................................................17 CHAPTER 1 INTRODUCTION AND OVERVIEW..................................................................................19 Swift Creek Complicated Stamped Po ttery and Social Interaction........................................23 Organization and Outline....................................................................................................... .27 2 WHAT IS A GIFT?.......................................................................................................... ......33 The Legacy of Mauss............................................................................................................ ..34 The Social Life of Thi ngs: Object Biographies......................................................................38 Object Extension and Personhood..........................................................................................40 Genealogies of Objects, Genealogies of Practice...................................................................44 The Contexts of Exchange......................................................................................................48 3 THE SWIFT CREEK CULTURES........................................................................................55 Swift Creek Technology: Wood and Earthenware.................................................................58 Chronology.............................................................................................................................61 The Diverse Social Landscape................................................................................................62 Ceremonial Centers.........................................................................................................67 Swift Creek Villages........................................................................................................69 Swift Creek Interaction........................................................................................................ ...71 Representational Meaning of Swift Creek Designs................................................................80 Summary and Conclusions.....................................................................................................86 Notes.......................................................................................................................................87 4 CULTURAL HISTORY AND ARCHAEOLOGICAL OVERVIEW OF THE WOODLAND PERIOD ON THE ATLANT IC COASTS OF GEORGIA AND NORTHEAST FLORIDA......................................................................................................90 The Ecological Setting......................................................................................................... ...90 Pre-Swift Creek Culture History............................................................................................93 Swift Creek Radiocarbon Chronol ogy, Typology, and Culture History................................99 Early Swift Creek (ca. A.D. 200 to 550).......................................................................101 Late Swift Creek (ca. A.D. 550 to 850).........................................................................103 9

PAGE 10

Swift Creek Site Types ......................................................................................................... 104 Village Middens............................................................................................................105 Northeastern Florida...............................................................................................106 Southeastern Georgia.............................................................................................110 Mortuary Mounds..........................................................................................................113 Summary of the Cultural Landscape.............................................................................117 Paddle Matches.....................................................................................................................118 Summary...............................................................................................................................121 Notes.....................................................................................................................................122 5 INSTRUMENTAL NEUTRON ACTI VATION ANALYSIS: PATTERNS OF SWIFT CREEK INTERACTION, PART 1......................................................................................134 The Sample...........................................................................................................................134 Instrumental Neutron Activa tion Analysis (INAA): Methods.............................................136 INAA Results........................................................................................................................140 Composition Groups......................................................................................................141 Discussion..................................................................................................................... .147 Summary...............................................................................................................................150 Notes.....................................................................................................................................151 6 PETROGRAPHIC ANALYSIS: PATTERNS OF SWIFT CREEK INTERACTION, PART 2.................................................................................................................................167 The Sample...........................................................................................................................167 Methods................................................................................................................................168 Results...................................................................................................................................170 Complementing the Ch emical Evidence..............................................................................174 Summary...............................................................................................................................179 Notes.....................................................................................................................................180 7 THE FORM, TECHNOLOGY, AND FU NCTION OF SWIFT CREEK POTTERY.........190 Technofunction.....................................................................................................................191 Considering Style..................................................................................................................194 Methods................................................................................................................................199 Vessel Forms........................................................................................................................200 Open Bowl.....................................................................................................................2 01 Restricted Bowls............................................................................................................202 Restricted Pots...............................................................................................................2 03 Open Pots.......................................................................................................................204 Flattened-Globular Bowls.............................................................................................205 Collared Jar................................................................................................................... .207 Small Cups and Bowls...................................................................................................208 Sm all Jars.......................................................................................................................209 Boat-Shaped Bowl.........................................................................................................211 Double Bowl..................................................................................................................212 10

PAGE 11

Multi-Com partment Tray..............................................................................................213 Beakers..........................................................................................................................214 Bottle......................................................................................................................... ....214 Shallow Bowl................................................................................................................214 Double-Globed Jar.........................................................................................................215 Vessel Morphology Summary.......................................................................................215 Orifice Diameter, Soot, and Mend Holes.............................................................................218 Rim Thickness......................................................................................................................222 Paste Characteristics.......................................................................................................... ...225 Summary and Conclusions...................................................................................................228 Notes.....................................................................................................................................231 8 THE SWIFT CREEK GIFT..................................................................................................255 A Genealogy of Swift Creek Materi ality on the Atlantic Coast...........................................255 The Forms and Meanings of The Swift Creek Gift..............................................................262 Future Directions..................................................................................................................271 APPENDIX A INSTRUMENTAL NE UTRON ACTIVATION ANALYSIS............................................276 B PETROGRAPHIC ANALYSIS...........................................................................................297 C RIM AND BASE PROFILES..............................................................................................323 LIST OF REFERENCES.............................................................................................................332 BIOGRAPHICAL SKETCH.......................................................................................................361 11

PAGE 12

LIST OF TABLES Table page 4-1 Calibrated radiocarbon assays for Early and Late Swift Creek contexts in northeastern Florida.........................................................................................................123 4-2 Calibrated radiocarbon assays for Earl y and Late Swift Creek contexts in southern coastal Georgia.................................................................................................................124 5-1 Site and type dist ribution of INAA pottery samples........................................................153 5-2 Clay samples in the INAA study......................................................................................15 4 5-3 Mean and standard deviation of el emental concentrations in each composition group...155 5-4 Pottery chemical group assignments by site.....................................................................156 6-1 Site and type distribution of petrographic analysis pottery samples................................181 6-2 Clay samples select ed for petrographic analysis..............................................................181 6-3 Summary descriptions of vari ability in petrographic paste categories.............................182 6-4 Summary descriptions of va riability in gross temper categories......................................183 6-5 Mineralogical paste categories by county and INAA group............................................184 6-6 Paddle matching samples by INAA and petrographic groups..........................................185 6-7 Percent quartz among chemical groups............................................................................185 6-8 R-squared value for the linear regression model and correlation between elements or principle components and quartz proportion...................................................................186 7-1 Vessel form summary statistics........................................................................................ 232 7-2 Soot and mend hole frequencies in each vessel form.......................................................233 7-3 Frequency of vessel form by surface treatment and pottery type.....................................234 7-4 Mound assemblage orifice diameter and rim thickness summary statistics.....................235 7-5 Midden assemblage orifice diam eter and rim thickness summary statistics....................236 7-6 Soot frequency grouped by orifice diameter....................................................................237 7-7 Aplastic constituen ts of gross paste categories.................................................................238 12

PAGE 13

7-8 Frequency of gross paste groups by site. ..........................................................................239 7-9 Rim thickness summary statistics by gross paste groups.................................................239 A-1 Sample Provenience, Type, and Group Membership......................................................277 A-2 Mahalanobis distance-based pr obabilities of group membership for Group 1 members...........................................................................................................................288 A-3 Mahalanobis distance-based pr obabilities of group membership for Group 2 members...........................................................................................................................291 A-4 Mahalanobis distance-based projec tions of group membership probability for unassigned specimens......................................................................................................293 A-5 Mahalanobis distance-based projections of group membership probability for clay samples........................................................................................................................ .....295 A-6 Eigenvalues and variance fo r the first ten principle components....................................295 A-7 Eigenvectors for each element......................................................................................... 296 B-1 Raw point count data (1)..................................................................................................298 B-2 Raw point count data (2)..................................................................................................301 B-3 Raw point count data (3)..................................................................................................303 B-4 Percentage data (1)...........................................................................................................306 B-5 Percentage data (2)...........................................................................................................308 B-6 Estimated frequency data (1).......................................................................................... .311 B-7 Estimated frequency data (2).......................................................................................... .313 B-8 Sand size data...................................................................................................................316 B-9 Sand size data and indices................................................................................................318 B-10 Key to headings and abbr eviations for petrographic data................................................321 13

PAGE 14

LIST OF FI GURES Figure page 1-1 The Atlantic coast of Georgia and northeastern Florida.....................................................32 3-1 Primary distribution of Swift Cr eek Complicated Stamped pottery and notable occurrences at distant sites.................................................................................................88 3-2 Swift Creek design attributes characteristic of split representation...................................89 4-1 Distribution of recorded Deptford and St. Johns sites......................................................125 4-2 Sites with Swift Creek pottery along the Atlantic coast...................................................126 4-3 Sites with paddle ma tches mentioned in the text..............................................................127 4-4 Reconstructed desi gn 34and select paddle matches.........................................................128 4-5 Reconstructed desi gn 36 and select paddle matches........................................................129 4-6 Reconstructed desi gn 38 and select paddle matches........................................................130 4-7 Select paddle matc hes with unnumbered design..............................................................131 4-8 Reconstructed design 291 and paddle matches................................................................132 4-9 Reconstructed design 151 and sher ds with identical and similar designs........................133 5-1 Distribution of sites with assemblages used in the INAA study......................................157 5-2 Distribution of clay samples in the INAA study..............................................................158 5-3 Bivariate plot of Cr and Ca in assemblages from she ll-bearing and shell-devoid sites...159 5-4 Inverse Distance We ighted (IDW) interpolation of Ca concentrations in both natural and archaeological clay samples......................................................................................160 5-5 Biplot of the fi rst two principal components....................................................................161 5-6 Bivariate plot of PC 2 and PC 4 with tentative clay groups.............................................162 5-7 Bivariate plot of PC 2 and PC 4 with pottery groups.......................................................163 5-8 Bivariate plot of Cr and Co...............................................................................................164 5-9 IDW interpolation based on clay samples for Co and Cr.................................................164 5-10 Bivariate plot of Cr and Co with unassigned samples......................................................165 14

PAGE 15

5-11 Bivariate plot of Cr and Co with pa ddle matching samples .............................................166 5-12 Percentage of chemical group assi gnments from mound and midden sites on the Lower St. Johns River......................................................................................................166 6-1 Mineral constituents useful for distinguishing clay resource groups...............................187 6-2 Group assignments by quartz, Cr, and Co........................................................................188 6-3 Group assignments by quartz, PC1, and PC2...................................................................189 7-1. Open bowl profiles...........................................................................................................240 7-2 Restricted bowl profiles.............................................................................................. ......241 7-3 Restricted pot vessel profiles........................................................................................ ....242 7-4 Restricted pot rim profiles................................................................................................243 7-5 Open pot profiles..............................................................................................................244 7-6 Flattened-globular bowl profiles......................................................................................245 7-7 Collared jar profiles................................................................................................. .........246 7-8 Small cup and bowl profiles........................................................................................... ..247 7-9 Small jar profiles.................................................................................................... ..........248 7-10 Boat-shaped bowls............................................................................................................249 7-11 Double bowls....................................................................................................................250 7-12 Multi-compartment trays................................................................................................ ..251 7-13 Beakers............................................................................................................... .............252 7-14 Double-globed jar............................................................................................................253 7-15 Rim profiles of small cups, bow ls, and jars from midden contexts.................................254 7-16 Soot frequency grouped by orifice diameter...................................................................254 C-1 Rim and base profiles from the Tillie Fowler site (8DU17245)......................................323 C-2 Rim and base profiles fr om Greenfield site #8/9 (8DU5544/5)......................................324 C-3 Rim and base profiles from Greenfield site #7 (8DU5543).............................................325 C-4 Rim profiles from various sites....................................................................................... .326 15

PAGE 16

C-5 Rim and base profile s from the Dent mound (8DU68)....................................................327 C-6 Rim and base profile s from Cathead Creek (9MC360)...................................................328 C-7 Rim profiles from Evelyn (9GN6) shell midden.............................................................328 C-8 Rim profiles from Lewis Creek (9MC16).......................................................................329 C-9 Rim profiles from McArthur Estates (8NA32)................................................................330 C-10 Rim profiles from Sidon (9MC372)................................................................................331 16

PAGE 17

Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy LOCATING THE GIFT: SWIFT CREEK EXCHANGE ON THE ATLANTIC COAST (A.D. 200-800) By Neill J. Wallis May 2009 Chair: Kenneth E. Sassaman Major: Anthropology This dissertation evaluates anth ropological theories of exchan ge by detailing the history of social interactions along the A tlantic coast of Georgia and northern Florida during the Middle and Late Woodland period (ca. AD 200 to 800). Rare or finely craf ted objects are often interpreted by archaeologists as exchanged valuables through which persons and social groups entered into relationships of reciprocal oblig ation and competition. In contrast, ostensibly mundane artifacts like utilitarian pottery are rarely considered to have held any exchange value. However, a truly unique dataset for studying past social interactions comes from Swift Creek Complicated Stamped pottery that linked sites throughout much of the Eastern Woodlands but was primarily distributed across the lower Southeast. Archaeologists have demonstrated that Swift Creek vessels and the carved wooden paddles used to decorate them were carried long distances acro ss the landscape. The considerable social diversity evident among social groups necess itates studying these in teractions through a genealogy of past material practices at multiple intersecting scales. Toward this end, I employ three complimentary methods in the analysis of numerous pottery assemblages from both mortuary mounds and village middens along the Atla ntic coast. Instrumental Neutron Activation Analysis (INAA) and petrographic analysis of clay samples and pottery are used to differentiate 17

PAGE 18

18 between local and foreign-made vessels at sites. Technofunctional analysis of pottery documents site-specific trends in the manuf acture, use, and deposition of vessels. Together, these data indicate that forms and functions of vessels deposited at mortua ry mounds were often different from the limited forms of village wares, but dom estic cooking vessels from distant villages were sometimes carried long distances to be fina lly deposited at mortuary mounds. Rather than the de facto refu se of moving people, the prev alence of foreign-made cooking vessels at mortuary mounds denotes the intentional emplacement of gifts, likely distributed in the context of deaths that obligat ed repayment of debts in which relationships between descent groups were reworked. Based on the content and distri bution of designs on ve ssels, I argue that exchanged vessels were parts of distributed pers ons that were extended in social space and time, transforming cooking pots into powerful tools of commemoration, affiliation, and ownership.

PAGE 19

CHAP TER 1 INTRODUCTION AND OVERVIEW Bronislaw Malinowski (1922) was the first to suggest that no nwestern economies could be fundamentally different from western market s in terms of how exchange systems were organized, how value was determined, and what motivated people to exchange. In his ethnographic work among the societies of the Trobria nd Islands that he perc eived to be relatively uninfluenced by western markets, Malinowski (1922: 175) saw a system of giving for the sake of giving seemingly based on generosity rather th an self-interest. What Malinowski observed in the Trobriands, the Kula ring, came to be known as the classic example in anthropology of a non-market gift economy that conformed to the rules of reciprocity. Theories of the gift continue to be focused on the perplexing questions of why pe ople give gifts and what it is that obligates a return. The motivations for gift exchange have been variously explaine d as characteristic of primitive societies, as fundamental to the primeval nature of all humans in all societies, as a social phenomenon that created important bonds between people, and as a mechanism of both social solidarity and compet ition (Levi-Strauss 1969; Mali nowski 1922; Mauss 1925; Sahlins 1972). Whatever its origins and ma ny functions, gift exchange is concerned with creating, maintaining, or altering social re lationships as much as it is about economic principles. Gifts can create bonds of equality, constitute social difference, or be used to threaten and cajole. Indeed, there can be menacing and destru ctive gifts (Parry 1989; Raheja 1988). Thus, a gift is never truly free, and although it is given with concern for social relationships, the intentions behind it are often not magnanimous. The organization, logic, and meanings of gift exchange ultimately depend on the specifics of histor ical and cultural contexts. Yet anthropologists have often interpreted the gift as evidence of a cross-cultural principle of reciprocity that operates mostly in 19

PAGE 20

kin-based societies and confounds the logic of self-interest in neoclassical econom ic theory. Sahlins (1972) Stone Age Economics a manifesto on reciprocity, has had a lasting effect on anthropology, particularly in archeology. Sahlins (1972) discer ned in stone age economies two forms of reciprocity corresponding with the domestic econo my and political economy of small-scale, kin-based societies. Generalized reciprocity was understood to operate between kin within the domestic (immediate) economy of the household and was ostensibly altruistic. These gifts included everything from a mothers milk given to her child to th e daily sharing of food among coresidents and kin. Alternatively, balanced reciprocity involved pe ople of greater social distance in their attempt to forg e social connections. This form of reciprocity constituted the political economies of small-scale societies, charact erized by the give and take of gift exchange and attempts to engineer social relations. Following the influe nce of Malinowski (1922), these two economic domains can be seen as also c oupled with gendered divisions, with women typically confined to the internal workings of the domestic economy and men operating in the outreaching operations of the political econom y (Weiner 1976:11-19). These dichotomous ideas have been very infl uential in archaeological investigations of exchange in past societies without markets or st andardized tribute system s. According to these conventional understandings, objects that are pr oduced and consumed w ithin the household tend not to be subject to exchange. This means that utilitarian items such as earthenware vessels, basketry, netting, or stone tools are generally not expected to be items of exchange in a gift economy. In contrast, especially exquisite or ornate examples of utilitarian items, or objects that are rare, made of materials from far away, or produced with speciali zed ritual knowledge are particularly valuable items of exchange, and this is the material from wh ich prestige is acquired in the political context of gift exchange. These two categories of material culture, utilitarian and 20

PAGE 21

cerem onial/prestige (i.e. exchangeable items), are therefore imagined as distinct, separate, and unshifting. According to this reasoning, domestic products are not regular ly exchanged because they are ubiquitous, easily replaced, and are ofte n the products of women, who are assumed to not be major players in the public political domai n. These generalizations might be borne out in some social contexts in which highly valued and access-restricted objects were produced, but there is nothing inherent in an object that determines its value for exchange (contra Hayden 1995, 1998). Rather, objects are exchanged in the c ontext of meanings that are embedded in social practice. Because all material technologies are consti tuted through culturally informed practices that are embedded w ith symbolic understand ings, any category of object has some potential of becoming caught up in political st ruggles and systems of exchange, no matter how seemingly mundane and quotidian it seems to the archaeologis t (Dobres 2000:116-117). This dissertation is concerned w ith one such class of material culture, pottery, that in its utilitarian form is rarely consider ed to have held any symbolic value, much less exchange value. By convention, long-distance transport and exchange should generally be limited to luxury or ceremonial vessels (e.g., Drennan 1985; Harry 2003; Struever and Houart 1972), while utilitarian vessels might be ex changed between close kin alon g with other subsistence or maintenance goods (e.g., David and Hennig 1972; Duff 2002; Fie 2000, 2006; Graves 1991). Perhaps explaining the distribution of vessels in some contexts, these unyielding categories of utilitarian and luxury object s, with their inherent measures of value, result in a myopic view of the materiality of exchange. In reality, objects of material culture ofte n get recontextualized through different cultural milieus whereby va lue and significance shift in complex and unexpected ways (Meskell 2004; Mille r 1995b; Thomas 1991). This is especially true of objects that are exchanged because they continually move in and out of different cultural contexts. 21

PAGE 22

Archaeolog ists have often been slow to rec ognize the potential for multiple and alternative symbolic capacities for individual objects, with some notable ex ceptions. In the Southwestern United States, for example, widespread vessel exchange among Pueblo and ancestral Pueblo populations has been explained in terms of the biographies of vesse ls and with recognition of the multiple uses of individual vessels in both dom estic and public performative contexts (Crown 2007; Duff 2002; Mills 2004, 2007; Walker 1995). However, archaeologists working in the Southwest often continue to use categories such as undecorated and decorated, falling back on the Sahlins (1972) dichotomy to designate th e former as low-val ue utilitarian goods exchanged informally between close kin and the la tter as the material of formal interaction among members of different communities (Duff 2002:25-27). While these categories may very well explain Pueblo systems of exchange, by ap plying these conventions too broadly we risk masking the multiple ways that objects become mobilized through exchange. Eschewing the presumption that utilitarian pottery carried less symbolic or exchange value than other artifacts, this study begi ns with the premise that all thi ngs have the potential to create bonds between people through exchange This idea takes inspirati on from Marcel Mauss (1925) The Gift perhaps the first work to outline how objects come to be entangled in social relationships through inalienable qua lities that link them to pers ons and places in the context of exchange. It was Mauss (1925) who first demons trated that the excha nge value of objects is essentially contextual, often deriving from histories of associ ation that imbue objects with cultural significance. The inherently arbitrary nature of symbolic signification combined with continually developing social practices make for malleable values among objects, potentially exploding the ceremonial (politi cal economic) and quotidian (dom estic) analytical divide. Ultimately, even the material of daily practices, su ch as pottery associated with food preparation 22

PAGE 23

and eating, can be seen as linked to particular people, places, or ideas and thus provide the substance through which social relationships are constructed and personhood is enacted in salient socia l contexts. For archaeologists, then, th e challenge is to outline the everyday material practices of past populations in enou gh detail as to be able to identi fy contextual shifts in the use of objects that might indicate symbolic transformations that do not conform to preconceived categories. Based on the history of pottery pr oduction and use detailed in this study, along the Woodland period coast of present-day Georgia and Northeast Florida the ex change of utilitarian pottery appears to have been caught up in ce remonial contexts that indicate more than reciprocal exchanges of subsistence goods. Swift Creek Complicated Stamped Pottery and Social Interaction During the second half of the Woodland period (ca. AD 100 to AD 850), Swift Creek Complicated Stamped pottery gained widespread popularity across much of the lower Southeastern United States, eventually beco ming common in assemblages throughout presentday Georgia and major portions of adjacent states. As with many archaeological types of pottery that are ubiquitous across the landscape, the hi storical reasons for th e extensive adoption of complicated stamped pottery are poorly understood. However, unlike other types, complicated stamped pottery preserves tantalizing evidence of th e social interactions that may have fueled its growth in popularity. Today, complicated stampe d vessels and sherds yield definitive evidence of social interaction at an unparalleled level of detail, owing to distin ctive vestiges of the manufacturing process (Snow 1998). To make the characteristic potter y, wooden paddles were carved with various motifs, some representing animals, plants, and faces, and were subsequently impressed into earthenware vessels before fi ring. The indelible impressions of the wooden paddles preserved inimitable signatu res or fingerprints such as wooden cracks or asymmetrical design flaws. These unique signatures allow ar chaeologists to identify pa ddle matches, that is, 23

PAGE 24

vessels sometim es hundreds of miles apart that were stamped with the same paddle. Hundreds of these paddle matches have been id entified; therefore either pots or the paddles that were used to stamp them were frequently carried acr oss the landscape (Snow 1975, 1998; Snow and Stephenson 1998; Stephenson et al. 2002; Stoltman and Snow 1998). Swift Creek pottery might also be unequaled in giving archaeologists a glimpse of Middle and Late Woodland symbolic repres entation in contexts where little else but pottery is preserved (Snow 1998, 2007). Indeed, well over 400 unique designs have been recorded pertaining to a variety of themes, and this is sure to represent onl y a small fraction of the immense corpus of designs. Although preserving the impressions of exquisite woodcarvings and coming in a variety of vessel forms, most Swift Creek pottery can be considered utilitaria n. Made for cooking or storing food and, when broken, swept into village garbage middens, complicated stamped pottery is ubiquitous at Swift Creek sites. Some theref ore warn archaeologists to beware of mistakenly elevating the social importance of these everyday humdrum artifacts (Williams and Elliott 1998:10). Along these lines, evidence for the move ment of both paddles and pots have received cursory explanation mostly as the byproduct of soci al practices in which other objects were more important: marriage alliances in which mates were exchanged and economic transactions in which vessel contents were exchanged (Stephenson et al. 2002; Stoltman and Snow 1998). In contrast, I submit that archaeo logists must be attuned to the contextual details of pottery production, use, distribution, and de position, before discounting vessels as mere utilitarian tools with little symbolic value. Swift Creek Comp licated Stamped vessels were made and used by populations that manifested a huge array of so cial diversity across a wide expanse of the Southeastern landscape. In terms of cultural attr ibutes, what constitutes Swift Creek in one area is not the same in another locale (Ashley and Wallis 2006), therefore the ways in which 24

PAGE 25

com plicated stamped pottery was embedded in social life certainly varied as well. Indeed, the social significance of complicated stamped vesse ls and the ways that individual designs were disseminated via pots or paddles may have been as variable as the Swift Creek social landscape itself, making multiple scales of analysis indispen sable to studies of Swift Creek interaction. To understand Swift Creek culture on a global scale, we must ve nture to understand it on a local scale. This means detailing th e range of variation within and between many assemblages from different kinds of sites in a region. This dissertation focuses on the material pr actices of pottery pr oduction, distribution, and use during the proliferation of co mplicated stamping on the Atlantic coast of present-day Florida and Georgia (Figure 1-1). Swift Creek pottery came to the Atlantic seaboard through specific historical circumstances, adopted first along the Lower St. Johns River around AD 200 and along the Altamaha River and intervening areas three centuries later (Ashley and Wallis 2006; Ashley et al. 2007). Swift Creek influences, whether due to migration, assimilation, or other social interactions, derived from differe nt geographical areas in each river valley, from the Gulf Coast of Florida to the Lower St. Johns River and fr om central and south-cen tral Georgia to the Altamaha River and Georgia coast. Perhaps as a consequence, cultural differences are observed between the two regions in the spatial struct ure of the built landscape of burial mounds and villages, in the modes of burial, and the tec hnological style of pottery. After AD 500, when complicated stamped pottery was being produced all along the coast from the mouth of the St. Johns River to the Altamaha River, paddle matches between sites became more abundant, not just between proximate sites but between rive r drainages separated by well over 100 km. These contexts on the Atlantic coast therefore offer th e potential to understand Swift Creek interaction as a historical process at the intersection of multiple temporal and spatial scales. Based on 25

PAGE 26

paddle m atches, pottery was obviously involved in social interactions of some kind. In this research I attempt to determine more precisely how pottery was embedded in social practice and interaction by outlining a genealogy of the material practices of pottery production, use, distribution, and deposition along the Atlantic coast. By using the term genealogy, I am referring to a record of te mporal and spatia l variation in earthenware vessel attributes that can be used to trace historical connections and disjunctures in material forms (cf. Gosden 2005). The record is developed in this dissertation through empirical evidence of vessel provenance, function, use, and deposition that is situated within specific archaeological contexts that relate to distinctive social practices. Two different spaces, mortuary mounds and village sites, were inscribed onto the landscape and clearly reflect the materiality of particular actions and events. While mortuary mounds and village sites are widespread across the Eastern Woodlands, they take a variety of re gionally distinct spatial configurations. Along the Lower St. Johns River, a series of low sa nd burial mounds each constructed over the course of several centuries stand separated from cont emporaneous villages by hundreds of meters (Ashley and Wallis 2006; Wallis 2007). This mo rtuary landscape of continuous-use mounded cemeteries spatially segregated from habitation site s contrasts with the pattern of sites in coastal Georgia and along the Altamaha River where village-adjacent mortuary mounds are more common (Ashley et al. 2007). The separate cemetery and habitation spaces in both regions present the opportunity to discover how pottery production, use, and deposition varied with social context. In turn, these data can be used to reconstruct patte rns of interaction and exchange. I argue that the spatial distri bution, and technofunctional and styl istic attributes of nonlocal Swift Creek Complicated Stamped vessels are char acteristic of gifts that were exchanged as 26

PAGE 27

im portant material for the constitution of soci al relationships, the most fundamental of which were marriage alliances. This conclusion is buil t on multiple lines of empirical data. Nonlocal vessels are identified by a robust sa mple of pottery and clays subjected to Instrumental Neutron Activation Analysis (INAA) and petrographic an alysis. Chemical and mineralogical data indicate that nonlocal vessels were deposited al most exclusively at mortuary mounds and that they were predominantly complicated stamped. Technofunctional analysis reveals that village assemblages were nearly entirely comprised of domestic cooking vessels while mortuary mound assemblages included a diverse ar ray of special-use ceremonial vessels in addition to cooking vessels. In the context of these many ceremoni al vessel forms, a significant percentage of cooking vessels at mortuary mounds on the Lo wer St. Johns River were nonlocal, made somewhere along the Altamaha River. This dist ribution of nonlocal co oking vessels indicates that rather than the de facto refuse of m oving people, complicated stamped vessels became important ceremonial material that was intentio nally emplaced at mortuary sites, likely deriving symbolic density from the fingerprinting capabi lities of impressed designs. With faces and animals emblazoned on their surfaces, complicated stamped vessels were the quintessential distributed object (e.g., Gell 1998:2 21) that could be used to embody parts of a person linked across the landscape by matching designs. Conseque ntly, I argue that simple cooking vessels from distant villages wound up in Lower St. J ohns mortuary mounds as gifts used for honoring and renegotiating social relationshi ps that were threatened in the event of a death, particularly marriage alliances between important descent gr oups living on the St. Johns and Altamaha rivers. Organization and Outline This study begins by establishing a background in social theories of exchange and Swift Creek archaeological contexts, continues with presentation of the INAA, petrographic, and 27

PAGE 28

technofunctional data, and closes with a synthesis of Swift Cr eek exchange on the Atlantic coast. In Chapter 2, I review theories of gift excha nge and also draw more broadly on anthropological considerations of materialit y to consider how and why objects are exchanged. I begin by focusing on the limitations of normative and f unctionalist conceptions of reciprocity that unrealistically limit the objects of exchange and r easons for transactions. I then review the various ways that material culture becomes entangled in social life, particularly in the context of exchange. Objects take on social lives through recognition of thei r biographies and as extensions of social persons or as agentive persons themselves. Because the kinds of objects that become meaningful and that are exchanged are variable with context, I end by advocating a contextually-based genealogy of material practice that details the range of variation in how a class of objects was made and used in order to infer the significance of their mobilization for exchange. Chapter 3 is an overview of the Swift Creek archaeological culture that covered a vast expanse of the lower Southeast. I focus on the social and cultural di versity among populations who made and used complicated stamped pottery, suggesting that the mean ings of Swift Creek carved designs and their role in so cial life was probably just as vari able. Next I review available evidence and previous interpretations of Swift Creek social interaction at a variety of scales. While there is some evidence to suggest that both wooden paddles and earthenware vessels were moved considerable distances, previous explan ations for interaction have been mostly decontextualized. Moreover, with few exceptions the representational meanings of designs have been considered separately from their distributio n on artifacts. Alternatively, I argue that the movement of designs via pots or paddles was likely critical to their meanings. Toward this end, I present my identifications of split representati on in Swift Creek design execution that supports 28

PAGE 29

the contention that vessels a nd paddles were conferred degr ees of personhood, which helps explain their dissem ination. Chapter 4 synthesizes current evidence of Sw ift Creek manifestations on the Atlantic coast. I begin with an outline of the ecological se tting of the coastal sector, discussion of the preSwift Creek culture history of the Woodland period, and review of the Middle and Late Woodland culture chronology of the Atlantic coast. I then integrate Swift Creek site information along the coast to make important cultural di stinctions between northeastern Florida and southeastern Georgia. I draw sp ecial attention to the mortuary landscapes of the Lower St. Johns River that consisted of series of mortuary m ounds segregated from contemporaneous villages. These distinctive mortuary landsca pes provided important contexts for social interactions across considerable distances, as evidenced by paddle matches between sites. The chapter concludes with a review of known paddle ma tches along the A tlantic coast. With this necessary background in place, the remaining chapters convey the major empirical contributions of the study. Chapter 5 presents new data derived from Instrumental Neutron Activation Analysis (INAA) of clay and pottery samples fr om the Atlantic coast. By discovering patterns in the chemical composition of samples, I use these data to determine the prevalence and locations of forei gn-made vessels at each site. Wh ile a larger and more diverse sample would be beneficial, the current results are quite compelling. The frequency of nonlocal vessels within mortuary mound assemblages is significantly greater than among midden samples on the Lower St. Johns River. What these data indicate is that on the St. Johns River people were depositing foreign-made Swift Creek vessels at mounds but were not using or breaking foreign vessels very often at habitation sites. 29

PAGE 30

I discuss in Chapter 6 new m ineralogical data from petrography of thin sections from a subsample of pottery and clays used in the INAA study. These data complement the chemical data to allow for better resolu tion in resource groups that firmly establish the foreign manufacturing origins of some vessels. For instance, the combination of chemical and petrographic data show that some vessels with paddle matches were likely made with the same clay source while others were ma de with slightly different clay sources from the same region. The mineralogical data also help assess the effect of quartz sand temper on the chemical composition of vessels. Because size of quartz temper closely corresponds with chemical group assignment, point count data were compared with elemental con centrations, showing that while some chemical differences may be due to the di luting effects of temper, other differences are almost certainly linked to variation among clay sources. Chapter 7 presents the results of technofunctional anal ysis of pottery. I begin by reviewing how technofunctional data are important for understanding the practic al functions of earthenware vessels and, by extension, their social significance. In what follows, I detail vessel form, function, and paste charact eristics among assemblages from thir ty sites. Using mortuary mound assemblages of reconstructed vess els, I identify 15 dis tinct vessel forms that were used for a variety of functions. Based on rim profiles and estimates of orifice diameter among assemblages, I infer that midden vessels are conf ined almost exclusively to cooking forms while mound assemblages have a wide range of formal variation that includes many special-function vessels. In addition, differences in rim thickne ss and paste constituents among Late Swift Creek vessels along the coast conform to a geographic pattern that li kely corresponds with the longlived traditions of different social groups. Thes e differences lend further support to the idea that 30

PAGE 31

foreign vessels at Lower St. Johns River sites were m ade in Georgia by Georgia-local Swift Creek potters. All of these data are brought to bear on an interpretation of Swift Creek interaction in Chapter 8. The history of the adoption of Swift Creek Complicated Stamped pottery on the Atlantic coast is a story of cultu ral and historical dis tinction between the A ltamaha River and St. Johns River regions. I explain that out of the specific context of these cultural distinctions grew systems of earthenware vessel exchange that be came integral to the constitution of social relationships along the coast, namely marriage alliances among descent groups. Based on the available evidence, I suggest that domestic cooking pots with complicated stamping were exchanged as indexes or citations of social persons, thus providing the material to forge and rework social relationships transcendent of th e time and space of face-to-face interactions. Complicated stamped surface treatment was a t echnological innovation that imparted greater resonance to social citations during important events, and I suggest th at these capabilities allowed for transformations of mundane cooking vessels into ceremonial material. Through the indexical qualities of paddle impressions, comp licated stamped vessels became distributed objects (e.g., Gell 1998:221) that gave presence to im portant persons or social bodies, especially in ceremony. This dissertation therefore negates the longstanding bias of anthropologists that utilitarian and mundane items are symbolically unimportant by deta iling the specific ways that commonplace earthenware vessels became inextr icably bound up in social and political processes. I close with a discussion of this studys limitations and th e potential for further research. 31

PAGE 32

32 Figure 1-1. The Atlantic coast of Georgia and northeastern Florida.

PAGE 33

CHAP TER 2 WHAT IS A GIFT? The gift has been a seminal idea in anthropol ogy for the better part of a century, in its various guises informing considerations of economy, social solidarity, and human nature. Originally published in 1925, Marcel Mauss The Gift is at once abstruse and rich with meaning, qualities that invite new readings of the work by each generation of anthropologists. In this way interpretations of The Gift can be seen to follow, and in some ways shape, the contours of the discipline through time (Sykes 2005). The most recent spate of Maussian revisionism is foremost a critique of the anthropo logical considerations of gifts as the quintessential material of nonwestern primitive economies that provide ex ception to the logic and motives of western capitalism (Miller 2001; Sigaud 2002). Countering th is idea, other research has come to focus on the various ways that objects and people are enga ged in processes of mutual constitution that give special impact and meaning to gift exchanges, not as a particular cate gory of transaction that conforms to social classifications but rather as sp ecific moments that are critical to the process of cultural construction in a variety of settings. The arbitrary natu re of symbol-based behavior itself contradicts general models that attempt to predict what objects will ta ke the form of gifts, the motivations for their exchange, or the significance of these transactions. In short, a gift is not a type of thing but part of a process by which things, social relationships, and persons are created. Accordingly, I do not view gifts as being charac teristic of a partic ular type of economy (e.g., reciprocity) or type of society (e.g., smallscale) but as a proce ss by which objects of exchange become variously imbricated with social worlds in ways that ar e dependent on context. This chapter outlines the social process of gifting and important implications of theories of materiality in archaeological st udies of exchange. 33

PAGE 34

The Legacy of Mauss In The Gift Mauss (1925) undertakes an explanation for prestations that are seem ingly voluntary, spontaneous, and putativ ely munificent but which are in fact obligatory, premeditated, and carefully calculated. More specifically, Mauss attempts to explain why a gift obligates a return and what it is that struct ures the nature of the return gift One answer that Mauss seems to offer is that the giving and receiving of gifts ope rates within a kind of original morality that stimulates systems of reciprocity. Indeed, Maus s (1925) work has a distinctively evolutionary flavor that betrays a deep concern with orig ins (Parry 1986), in which reciprocity can be understood as the primordial and inherently mora l state of primitive economies. The idea of reciprocity in The Gift comes largely from Malinowskis (1922) work in the Trobriand Islands, from which Mauss drew heavily. Malinowski (1922:175) himself beli eved that the gift from its very general and fundamental natureis a unive rsal feature of all primitive societies. However, Malinowski (1922) presents an impor tant difference: Mau ss (1925) describes reciprocity as operating in all societies while Malinowski (1922) restri cts its operation to a particular type of primitive society (Gudeman 2001:84). It has been the latter view that has garnered the most influence in anthr opology until recent decades. Levi-Strauss (1969), Polanyi (1944), and Sahlins (1972) were each influential in establishing what amounts to a law of reciprocity that confor ms to evolutionary social typologies. Levi-Strauss (1969) made the clai m that exchange was the primary fundamental phenomenon of social life, making reciprocity the very foundation of society. Polanyi (1944) introduced a three part typology of economies that was implicitly arranged on an evolutionary scale: reciprocity was practiced in societie s dominated by kinship concerns, redistribution pertained to ancient societies where religious and political authority were established, and market exchange occurred in modern capitalist c ontexts. In many ways Sahlins (1972) combined 34

PAGE 35

aspects of the work of Mauss (1925), Levi-Strauss (1969), and Polanyi (1944) to devise a typology of stone age econom ies based on social distance, with reciprocity as its foundation. Sahlins (1972) described generali zed reciprocity as exchange between kin that was ostensibly altruistic, balanced reciprocit y as less personal and more economic gift exchange between nonkin, and negative reciprocity as bart er and theft. Each of these types of exchange was a version of reciprocity at its core but was defined by different levels of social distance, with negative reciprocity characterized by the greatest social distance and the least personal form of transaction. Largely through the influence of these works, Ma uss idea of the gift enjoyed somewhat of an invented legacy in anthropology as it came to represent part of a dichotomy between reciprocity and market exchange that is probably best codified in the work of Gregory (1982). Through perspectives that were deeply rooted in colonialist economics and po litics, critics argue, the gift ultimately came to be synonymous with reciprocity, understood as the natural, primitive, and inherently moral state of the economy of the ethnographic other that stood in opposition to western market economies (Har t 2007; Sigaud 2002; Weiner 1992; 1994). Yet instead of a dichotomy between disparate economi c systems, Mauss himself actually viewed the gift as a total social fact, a common reality that pervades all of society and all institutions: economic, political, religious, and aesthetic (Hart 2007). For Mauss the purely altruistic gift was indeed the polar opposite of pure se lf interest but gift exchange systems are no more altruistic than capitalist markets are base d purely on self interest. In fact, Mauss conceived of the archaic gift as a mixture of these two modes of exchange and rarely used the term reciprocity (Hart 2007). 35

PAGE 36

W hatever Mauss original intent, the idea of reciprocity as an inherent regulatory mechanism in primitive societies devoid of ownership, legal codes, or political hierarchy has confounded anthropological efforts to understand what motivates gift exchange. Other research has demonstrated a more calcula tive dimension among societies ot herwise described as solidarity writ small (Appadurai 1986; Bourdieu 1977; We iner 1992, 1994). For example, Weiner (1992, 1994) refocused the question of why a gift oblig ates a return by centering on the inalienable qualities of exchanged objects. Ma uss described the spirit of th e gift, most famously in the Maori hau, as a way that things create bonds between people by retaining inalienable qualities of the givers person in the context of exchange. Gifts essentiall y embody the nature, substance, and spiritual essence of a person, and thus beco me powerful, dangerous, alive, and personified (Mauss 1925:10). Weiner (1992:150) develops this idea further to argue that inalienability itself is the thrust of systems of exchange as people engage in a process of keeping while giving. Weiner (1992) outlines how, through exchange of other objects, a person or social group attempts to conserve their most valuable po ssessions that establis h differences between themselves and other persons or groups. These items, such as Samoan fine mats or Kwakiutl coppers, are ranked against one another and some rare ly circulate. In this way the exchange of fine mats or coppers is as much about the objects that are being offered as about the ones that are being kept out of exchange. These are the items that Mauss described as immeuble, immovable objects that are closel y linked to soil, clan, the family, and the person (Weiner 1992:46). Thus, according to Weiner (1992:150) it is the radiating power of keeping inalienable possessions out of exchange that fuels reciprocal exchange networks. While invaluable items may be literally kept according to Weiner (1992), the other related implications of inalienability are to do with the qualities of place and person that adhere 36

PAGE 37

to an object even through changes in possessi on and geography. In th is sense objects are inalienable not becaus e they cannot be given away but because they represent linkages that are inextricable from their material form. Recognition of the spirit of the gift in terms of the entanglement of things and persons opens opportuniti es to understand specific social contexts in which objects are mobilized as more than their mate rial parts. Indeed, so me things matter in the profound sense that they are critical to the pr ocess of self-constructi on (Miller 1995a). As theories of material culture have long demonstrated, social worlds are constituted by the object world, not just the other way around (Bourdieu 1977; Appadurai 1986; Miller 1987, 1995a, 2005). Rather than reducing mate rial culture to essentialized models of the social world, materiality can be seen as inte gral to the process of sociality. Things inevitably become entangled in peoples lives in ways that bind them to particular ideas, meanings, memories, places, and persons, and confer to them various degrees of agency (Thomas 1991:16). Therefore, objects mediate social agency in ways that are contextually specific and historically situated. Although this realization is deceptive ly simple and far from new, anthropological research invested in this perspective of materi ality, as the mutually constituting dialectic of people and things, arguably brings us closer to studying the social as a process in a continual state of becoming rather than focusing on analyti cal abstractions such as society and culture (e.g., DeMarrais et al. 2004; Meskell 2004, 2005; Miller 2005; Thomas 1999). No less important, a materiality perspective with a focus on the engagement of objects within social worlds enhances the relevance of archaeology, tr ansforming the discipline from an ineffectual attempt to understand past societies through the vestiges of long-gone th oughts and behaviors to resurrecting some of the very substance through which past socialities were constituted and transformed. 37

PAGE 38

The Social Life of Thin gs: Object Biographies As Appadurai (1986) describes, m a terial things have social live s in the sense that they are inextricably bound up in social process, perhaps mo st significantly in the intercalibration of the biographies of persons and things. The con cept of object biographies has become popular among anthropologists and particularly archaeologists, who often outline the life cycle of objects (e.g., Schiffer 1975, 1976; Schiffer and Skibo 1997). However, Appadurai (1986) was interested not just in the cultural biography, or individual life hist ories of objects, but also in the social history of things, the collective history of a particul ar class of object a nd how it was imbricated with social life. In an effort to understand the relationship of material things to human actors, such a project is necessarily deeply contextual. Because symbols are not inherently meaningf ul but depend on specific social practices (e.g., Leach 1976), there are different kinds of object biographies that pertain to particular social and cultural contexts. Gosden and Marshall (1 999) distinguish between objects that gather biographies to themselves and those that serv e mostly to contribute to the biography of a ceremony or body of knowledge. Some objects have the ability to accumulate histories, deriving significance from the people, places, and events to which they are connected. Most famously, Kula valuables in the Trobriands have this qu ality, maintaining links with named persons who possess and transact each object. The direct relationship between a person and an object is retained through each subsequent possession by another person, setting up a process of enchainment whereby social relations become defined by exchanged objects (Chapman 2000). In this context, the identities of people and objec ts are mutually creating as objects gain value through their associations with powerful peopl e and people build reputations through their possession of famous objects. Not only its attachment to people, but also where something is from (or perceived to be from) is a significant part of its biography. An object may be valued 38

PAGE 39

because of its procurem ent from an important pl ace or simply because it comes from far away (Helms 1988). As pieces of places (Bradley 20 00), objects can be seen to materialize distant places, events, and persons, and in their excha nge thereby constitute social connections and define social differen ces (Thomas 1999). Objects that accrue biographies ca n also be distinguished by the level of specificity in their recognized histories. Gosden and Marshall (1999) point to Kula valuables as examples of objects that have very specific hi stories of association with named persons. In contrast, tabua, which are whales teeth that are circulated (singl y) throughout Fiji, take on value through a rather generic understanding of their age. Over time, ta bua become darker in color due to the oil from peoples hands in contact with them. The darker the tabua, the older and mo re valuable it is, but this biographic quality is not linked to particular named owners or places, making its story generic compared to famed Kula valuables. In contrast to objects with accumulative biog raphies, some objects hold little inherent meaning outside of performative contexts. For example, among the Kwakwakawakw on the Northwest coast of North America, carved wooden masks were a means through which ceremonial privileges could be ma nifested in material form. However, possession of the mask itself was not significant because its meaning was tied to the context of performance. It was the act of showing the mask that was central to its power and meaning. Thus, in material form the masks were alienable and the Kwakwakawakw were not wary of selling them to outsiders. However, neighboring Nuxalk groups to the nort h understood and performed the relationships between masks and people much differently, ma king their sale as commodities much more problematic (Seip 2001). Hence, the biographic qualities of objects are historically and 39

PAGE 40

cultu rally contextual to the extent that the same class of artifact used in similar ways can have quite different biographic capabilities. Object Extension and Personhood One productive way to explore contextual variat ions in the social liv es of objects is in thinking about an artifacts extensi on in social space and time. This focus returns to the idea of a social history of things propos ed by Appadurai (1986), recognizi ng that objects have a social habitat, a correct place and tim e of use (Robb 2004). Material objects are part of culturally specific practices that are associated with particular social roles for practitioners, a body of knowledge including bodily knowledge and com portment, and symbolic significance (Robb 2004:134). A common way that object s function in the social wo rlds of Melanesia is as extensions in time and space of the social life of a person or lineage (Munn 1983, 1986, 1990). Kula valuables that are associated with a person s name are viewed as indexes of his physical presence, knowledge, age, and intelligence, and a Kula operator must carefully calculate how these parts of his self are distributed (Gell 1998:230-231). As objectific ations or indexes of personhood and of thoughts, intentions, and mental states, Gell (1998:231232) argues that the Kula system as a whole is a form of cognition which takes place outside the body, which is diffused in space and time, and which is carried on through the medium of physical indexes and transactions involving them. In this way the Kula necklaces and arm-shells transacted by a Kula operator are a distributed object, an object having many sp atially separated parts with different micro-histories, and through their di ssemination the person who set the objects in motion becomes a distributed person, extended through time and space (Gell 1998:221). The Kula system is, in effect, an objectified social world in which social relationships between persons can occur beyond the face-to-face intera ctions of biological individuals. 40

PAGE 41

The extension of persons is not a capability restricted to Kula valuables, but is also typical of m any of the objects employed in mortuary ce remony. For example, in the Trobriands, when someone died their possessions (earrings, armbands clan-associated feathers for men or skirts and earrings for women) or even bones from thei r body were carried for up to a year by members of the matrilineage of the spouse or father of the deceased. In an in terview with Myers and Kirshenblatt-Gimblett (2001), Annett e Weiner convincingly interpre ts these actions as extending the social life of a person afte r their death, specifically by maintaining relationships between lineages that had been forged through marriage. By carrying part of the material identity of the deceased, reciprocal obligations originally fo rged by marriage were continued and womens skirts and banana leaf bundles were subseque ntly given by the deceased persons lineage. The extension of the deceased pers on through exchanges maintained social relationships between lineages in anticipation of a new marriage betw een them in the future. The Sabarl axe in the southern Massim area of Papua New Guinea goes even further to extend personhood in time and space (Battaglia 1983, 1990). According to symbolic representations, Sabarl axes are conceived as bodies with part icular named body parts that are correlated with human body parts: th e blade represents the right hand or genitals, the shaft is an arm with a crook that represents an elbow, and so on. These axes are animated by the reproductive potential of a person, called hinona, which is embodied in the heat-generating greenstone blade that represents the genitals and the right hand that guides exchange. Significantly, the shape of the axe is also a material metaphor for exchange relations and the movement of gifts in mortuary exchanges as well as a representation of two different aspects of the person (clan member and individual) united in a single form. In the event of a death, five axes are presented by the paternal clan to the ma ternal clan of the deceased. The maternal clan 41

PAGE 42

then uses these axes, along with food, to constitute the corps e of the deceased. Once the mortuary cerem ony is over, the axe corpse is deconstructed by reprodu cing axes, in effect killing the dead person and seve ring links between clans while at the same time creating an ancestor. The axes serve not only as substitutes for a person in this case, but can be seen to act as persons because they participate in the same so cial actions as persons: they gather at mortuary ceremonies and have the power to reproduce ne w people, new objects, and new social ties through exchange. These examples of objects conceived as persons or as parts of persons are consistent with Stratherns (1988) descriptions of Melanesian personhood. Rather than individuals, Strathern (1988) argues that Melanesian persons are co nceived as dividuals that have partible components. Every person is multiply authored and created out of rela tions between others, starting with both parents. Therefore each person is a composite of the substances and actions of others, encompassing constituent things and relations received from other people. These relationships may be condensed into physical subs tances or objects and, through partibility, some parts can be extricated from a person. This par tibility can be important for a number of reasons, such as allowing a person to remove parts of their dividual self that are the same as others in the community for the purposes of marriage (Mos ko 1992). Through the ex trication allowed by partibility, the partible person becomes a partial version of that person, in which the extracted part is presented as the whole (Fowler 2004:25). For example, a pig in the Highlands of New Guinea is multiply authoredit is produced through the labor of a man and his wives. However, in offering a pig as a gift, a man temporarily presen ts it as a unitary version of himself, standing in for his family in the process. Consequent ly, it is the mans person who enters into the exchange relationship and it is the man who garner s prestige from the trans action. In this way a 42

PAGE 43

person can f luctuate between their dividual natu re constituted by various relations, and their partible nature in which parts of themselves, wh ich are also multiply authored, can be extricated and given away (Fowler 2004:31). The concepts of a fragmented and multiply-constituted person and the consequent personlike qualities of objects ha ve been employed by archaeologists studying patterns of exchange in many parts of the world. It is perhaps the l ack of individual or bi ographic specificity in Stratherns (1988) work, along w ith it being one of the most thorough negations of the bounded Western individual that has made these ideas so popular for prehistoric archaeologists (Meskell 2004:55). Indeed, the concepts of personhood developed through ethnographic work in the Trobriands and Papua New Guinea have become so pervasive among some researchers as to give the European Neolithic a Melanesian flavour (Jones 2005:195). Of course, the partibility of persons does not pertain to all times and pl aces because their definition and creation is contextual. As LiPuma (1998) argues, both individual and dividual modalities of personhood exist in all cultures but one or the other aspect s is masked according to different contexts. Generally speaking, in Western cultures the multicomponent, dividual parts of the person are masked while the individual is emphasized. The o pposite is true in Melane sian contexts. It is out of this tension between dividual and individual aspects of personhood that social persons emerge (LiPuma 1998:57). In essence, people are configured in historically and culturally specific ways and the criteria for defining exactly who or what may or may not be a person is contextually variable. Rather than simply populating the prehisto ric past of Europe or North America with Melanesian models of personhood, archaeologists must attempt to define the contextual world of objects and how they interrelate with human lives. Jones (2005:126) points to other possibilities derived from Mayan (Gil lespie 2001; Houston and Stuart 1998; Meskell and 43

PAGE 44

Joyce 2003; Joyce 1998, 2003) and Andean (Allen 1998) examples in which less em phasis is placed on object exchange as the most salient medium for the production of persons. Instead, persons emerge primarily through bodily practic es, ritual practice, a nd the inhabitation of architectural spaces. Yet in all events the person can be located in the sum of their relationships, which are archaeologically accessible through the relations between people and things, between people and architectural spaces, and more generally with their landscape and environment (Jones 2005:199). Deciphering the proce ss by which materiality helps constitute the social world, including personhood, is principally ba sed on a concern with context. Genealogies of Objects, Genealogies of Practice Gosdens (2005) focus on the genealogy of obj ects is relevant here. The genealogy of objects refers to a historical conc ern with descent lines of materi al forms and their modifications through time and space, paying part icular attention to how thi ngs of different origins and histories are put together in cohe rent ways. Gosden (2005) is fore most interested in how things shape people and argues that rather than focus on meanings, per se, that we should first focus on the effects of object worlds. The agency of objects defined as the social effect on people, is in the combination of their forms, historical trajectories, and perc eived sources. Gosdens (2005) compelling example comes from the British Isles incorporation into the Roman Empire, in which changes in form and attributions of source am ong pottery vessels and metal fibulae and brooches with demonstrable influences from the continent resulted ultimately in the construction of new and locally distinctive t ypes of people. These were not Romans, but locally unique persons who were constituted out of new combinations of material practices and their simultaneous transformation. There is a commonality in Pauketat and Alts (2005) approach to genealogies of practice that examines the socially cons tituting qualities of materiality. In their study of agency in a 44

PAGE 45

postm old, Pauketat and Alt (2005) argue that physical objects such as posts used in house construction can be the catalyst through which social change happens or con tinuity is secured. In pre-Mississippian times, house construction was in itiated with posts that were secured in individually dug holes that made necessary th e coordination and colla boration of labor among familial or communal groups. The necessity for community cooperation in the construction of each house was later obviated by a new kind of hous e introduced from the political center of Cahokia that included wall trenches and possibly prefabricated walls. These houses could be constructed by a single person. Even in the mi dst of adopting these new house construction techniques, however, some farmer s continued to dig individual posts beneath the trenches, presumably referencing the symbolic importance of the communal effort of building houses. Concurrently, the erection of large marker pos ts in the plazas of ceremonial centers can be seen as an inversion of the familial-communal basis of the post-setting practices that once were used for building houses. Given the frequent resetting and replacement of marker posts, it was the communal practice of setting posts rather than ju st the physical post itse lf that was important. Through the genealogy of the practice of post-setti ng in various contexts, we can begin to see how the materiality of posts came to shape the lives of people. Thus, objects can be agents in the sense that they have social effects. The quite influential idea of object agency has been explored in detail by Bruno Latour (1999) and Alfred Gell (1998). Latour (1999) emphasizes the fact that ju st like people, objects cause actions and their effects have real social conseque nces. This can be true of object s that are perceived by people to be animate and have agency as well as those that do not. Latour (1999) uses the term actant to describe all people and things with agency in a symmetrical and equal way, thereby entirely dissolving the theoretical oppositio n of agentive subjects and inan imate objects. Gell (1998) is 45

PAGE 46

som ewhat less willing to grant full agency to objects, instead creating a category of secondary agency that describes the effect of human agency through objects. In effect, Gell proposes that primary agents create secondary agents thr ough the process of objectification. The extreme position of Latour (1999), though perhaps philos ophically coherent, introduces the danger of asserting object agency to th e exclusion of keeping track of how objects were important in particular times and places. The agency of objects certainly does not always pertain and as Meskell (2004:5) argues, theories of object agency are most efficacious when they enhance our understandings of peoples intentions and practices (my emphasis). In archaeological terms, such an understandi ng of past social practice only comes from a deep concern with historical, so cial, and cultural contexts as described by Gosden (2005) and Pauketat and Alt (2005). In essence, a focus on the genealogy of objects or the genealogy of practice (as it is implicated in materiality) can provide the robust understanding of context through which to infer the social ramifications of changes in ma teriality. These genealogies provide the basis for the identification and in terpretation of recontextualizations whereby objects move through various social contex ts. Thomas (1991) uses the idea of recontextualization to describe the transformation of ethnographi c art objects through their movement into transnational commodity market s, museum exhibitions, and the holdings of private collectors. However, the term can be usefully applied to the movement of any object through various social contexts that redefine its referential meaning. What is more, when viewed at a larger scale, recontextualization can refer to the transformation of a corpus of material practices over time, as in the case of Roman pottery in Brit ain (Gosden 2005) or the changing use of posts in a variety of domestic and ritual settings (Pauketat and Alt 2005). 46

PAGE 47

In agreem ent with the procedur al agenda set forth by Pauketat and Alt (2005), I believe the most efficacious way to study past social practice is through an understanding of three related aspects of materiality: the gen ealogy of objects (the corpus of a class of material culture, including its range of variation in time and space), the genealogy of practice (how objects (including architecture) were made and used through time), and finally, the interrelations of practices as they were implicated across various cl asses of material culture. These avenues of study refer to the indexicality (G ell 1998) and citational (Jones 2005) capabilities of objects and social practice. Drawing on the f unction of citations in texts, whic h refer to other texts and in the process reiterate their importance, we can view social practice its elf as embedded in a network of citations (Butler 1993; Derrida 1982). For example, Butler (1993) and Joyce (1998, 2000a) discuss how actions are gendered with reference to prior gendered performances that thereby build and accentuate their social connotation. Similarly, the object world is enmeshed in referential fields of material citations (or ind exes) as each action th rough materiality makes reference to and acquires its meaning from past a nd present objectifications. Importantly, social practice through material referent ial or citational fields, as Jones (2001, 2005) calls them, is not necessarily conservative. Rather, the citation of past or spatially removed events and actions in new social contexts or in innovative combinations enables the constitution of new and different meanings (Jones 2005). When based on cor puses of material culture that are specific to particular contexts, the concept is therefore useful in understandi ng how material practices were used to create social worlds and how these practices were changed or transformed through time. The work of Gosden (2005) and Pauketat and Alt (2005) is above all a focus on the transformations of material practices, the most obvious of which are transformations in the form, 47

PAGE 48

tem porality, and space of material production, use, and deposition, but by inference these correlate also with changes in the constitution of persons and communities. The Contexts of Exchange The forms and reasons for exchange in the pa st are best understood by situating acts of exchange within the broad social contexts of materiality outlined above. While all material culture is a citation in one way or another, at the very least referencing the practices that engage objects of the same kind, individual exchanged obj ects are often particularly important citations that bridge multiple and significant distances in social space and time. Approaching the indexicality of exchanged objects as citations or extensions re quires the sort of contextual genealogies that allow us to be attuned to instances of transf ormation. This is an empirical enterprise that necessitates l eaving behind the received wisdom of evolutionary social typology and its corresponding economic cate gories. As Thomas (1991:39) argues, the transformations and contextual mutations of objects cannot be appreciated if it is presumed that gifts are invariably gifts and commodities invariably commodities. This contextual approach runs counter to many archaeological studies of exchange in the Eastern Woodlands of North America that have b een mired in categories of social organization and types of exchange, thereby often overlooking act ual past human practice. In general, modes of exchange have been differentiated accordi ng to a dichotomy that opposes group mechanisms of integration and the fulfillment of economic requirements with individuals tendencies to aggrandize power and wealth. Exchange among th e smallest and most mobile social groups has been most often interpreted as generalized reci procity, argued to be bot h economically practical (or necessary) and socially integrative (Sahlins 1972). These functionalist interpretations of social interaction and exchange foreground the economic and biological needs of small-scale societies, ranging from the biological imperative for mate exchange to insurance against periodic 48

PAGE 49

food production shortfalls (e.g., Anderson 1995; Anderson and Hanson 1988; Braun 1986; Braun and Plog 1982; Brose 1994; Clay 1998; Dye 1996; Jefferies 1996; W althall and Koldehoff 1998). This risk management perspective, ul timately supported by the work of behavioral ecologists with hunter-gatherers in marginal envi ronments, posits that the function of exchange and associated practice such as feasts among hunter-gatherers is to build alliance networks and redistribute wealth and subsis tence goods, thereby pooling econo mic risks (Brown 1985; Kelly 1995:168-201; Wiessner 1982). Thus, recognized pe riods of heightened exchange are assumed to correspond with amplified social and economic risk associated with changes in climate, population size, and subsistence systems (Braun and Plog 1982). In contrast, theories for exchange among la rger, more socially complex, and typically agricultural social group s have focused on its sociopolitical function, namely the development and maintenance of hierarchical power th rough unequal access to goods and information networks and the prestige goods or primitive valuables trafficked through them (e.g., Brown et al. 1990; Peregrine 19 91; Welch 1991). The concept of a prestige good, however, presents a confounding tautology within archaeo logical interpretation, as pres tige is somehow defined by what is gained through the posse ssion and use of prestige goods and prestige goods are those whose use gives one prestige (Robb 1999:6). Indee d, defining whether prestige or the prestige good came first is problematic, as some researchers argue that pow er and domination would have been necessary to orchestrate the widespread and sometimes high-volume exchange networks evident at large ceremonial cent ers or gateways yet others see power as derived from the products of these very networks (e.g., Gibson 1996; Smith 1986). The actions of elites are not always interpreted as the consequence of i ndividual political ambiti on alone but are also attributed to economic adaptations. Thus, a di chotomy between functional and exploitative, 49

PAGE 50

adaptation ist and political and ultimatel y, corporate and network strategies lend dynamism to elite behavior and their social c ontexts (e.g., Blanton et al. 1996; Brumfiel and Earle 1987; Gilman 1981; Saitta 1999). Even still, prestige g oods models arguably condense and limit the kinds and scales of social relationships created or maintained by the use, display, or exchange of different objects. These interpretations of non-capitalist exchange derive explicitly and implicitly from the economic typologies proposed by Sahlins (1972) a nd Polanyi (1944). In fact, various models used to explain the same archaeological cont ext can sometimes be seen to embody several typological categories. Hopewell exchange, characterized by the long distance movement of a variety of artifacts and raw materi als throughout much of North Amer ica, has been interpreted in a variety of ways that can be grouped into three categories. First, exchange of Hopewell objects has been construed as part of a system of centr alized redistribution that implies a chiefdom-like tribute system (Griffin 1965). Second, Hopewell objects are unde rstood as prestige goods that became significant markers of status, and were exchanged between high ranking individuals in a network with important elements of compe tition (Braun 1986; Brose 1979a, 1979b; Struever and Houart 1972). Finally, exchange is argued to be part of alliance bu ilding strategies that were not necessarily competitive and did not only involve elites (Braun and Plog 1982; Seeman 1995). Although archaeologists now recognize that Hopewell interaction was comprised of many distinct social proce sses (Carr 2006a; Seeman 1995), the diff erences in interpretation described here are drawn as much from typological assumpti ons as they are from archaeological data. In many of these models, the exchange of subsistence or utilitarian materials is viewed as ancillary to prominent formal exchanges. Thus, formal exchanges paved the way for beneficial exchanges 50

PAGE 51

of goods that either provided in surance against food shortages or fostered elite control of those goods to bolster political power. These are good examples of how long-distan ce exchange has been generally interpreted among small-scale and middle-range societies. The use of types of exchange that ostensibly correspond with social variation are problem atic because interpretations tend to be decontextualized and either uncritically accept so me sort of law of reciprocity or give suspiciously western economic and political read ings to gifts, reciproc ity, authority, and gender (Weiner 1994). Consequently, researchers using exchange models rooted in economic theory and social typology not only posit the value of object s a priori but also pr edetermine which items are significant for the purposes of exchange. For example, outsi de of standardized tribute or market exchange systems, pottery is nearly al ways considered to be locally made and not exchanged. The reasons for this assumption usually include the wide availability of materials, the consequent ubiquity of pottery production ac ross the landscape, and the tendency for vessels to be bulky and heavy, making them cumbersome to transport compared to most preciosities (Fie 2006). More importantly, earthenware vessels are presumed not to have been exchange items because they are understood to be a product of the domestic economy and particularly the immediate labor of women, who are assumed to not be major players in the public domain. However, if we take seriously the transformativ e capabilities of objects, all things have the potential to become important items of exchange depending on the social context. As Melanesian ethnography repeatedly shows, the produc ts of womens labor are often quite critical objects of exchange (e.g., MacKenzie 1991; Strath ern 1988; Weiner 1992). Ultimately, even the most common materials of daily pr actice can be seen as linked to particular people or qualities of people and thereby be constitutive of social rela tionships. MacKenzies (1991) description of 51

PAGE 52

bilums m ade by women in Papua New Guinea is a prom inent example of a ut ilitarian item with significant exchange value deriving, in part, from biographic qualities of the objects. Bilums are string netbags used by men, women, and children for the utilitarian tasks of storing or carrying a wide range of objects. Netbags are therefore functionally useful but they also have aesthetic value as an ornament that is worn on the body and a variety of styles serv e as markers of cultural and social identity. What is more, bilums are symbolically linked to the womb and reproduction, and they are central to much of social life, ma rking individual achieveme nts, clan membership, and transitions in the life-cycle, as well as main taining ritual and kinship ties, and serving as items of wealth (Tilley 1999:64). The netbags themselves are produced by women but men take some of them and add additional features such as feathers in the context of initiation ceremonies, in the process claiming the bilums as products of their own labor. However, this elaboration of the netbag does not erase the womans relationship to it. Rather, the bilum stands for detachable aspects of a womans identity that ar e still indelibly linked to her (Tilley 1999:66). Thus, a mans bilum is the product of multiple authorship, and as it passes through various stages of production and use it acquires a biography th at embodies relationships between men and women. As exchange items that accumulate biographies, bilums mediate relationships between men and women and different social groups, sometimes serving as payments of bridewealth. Thus, meaningful exchange objects emerge out of a corpus of material practices that help constitute the social world. In a generic sense, objects ac quire value because they embody memory and knowledge (Hendon 2000), and it is the task of the researcher to investigate the various functions served by a class of objects that reveal its contributions to the ongoing process of cultural construction. In archaeological research, this is best accomplished by tracing genealogies of material pract ice over time and space and analyzing potential contextual 52

PAGE 53

transform ations. Rather than be limited by func tionalist platitudes, such an endeavor can lend specificity in terms of actual so cial and cultural process to expl anations for observed variations in patterns of exchange. Indeed, archaeological evidence in the Americas has already indicated that pre-Columbian exchange was exceptionally va riable in terms of what, how often, and in what social context it took place (Earle 1994: 420). Moreover, most archaeologists also recognize that exchange is continge nt upon variable social and histor ical factors so that the forms and scales of exchange do not dependably correspond with structural aspects of social or political organization (Saitta 2000). The present study is concerned with two commo nplace and widespread classes of material culture that were each routinely moved acro ss the landscape among Woodland Period Swift Creek cultures: carved wooden paddles and earthenw are vessels. Removed from the specificities of context, a number of social practices might all potentially explain th e distribution of these objects, including migrations, seasonal rounds, post-marital residence patter ns, pilgrimages, and various kinds of exchange. Of course, deciphe ring among these alternatives is possible only through detailed contextual analyses that attempt to reconstruct histori cal, social, and cultural circumstances. With enough data, a genealogy of material practice can re veal how these objects were employed in daily life and how practices va ried over time and space, thereby presenting the opportunity to understand how and why the objects were moved. Identification of variation and transformations in material practice ultimately depend on definitions of scale, and the most thorough understandings derive from tacking back and forth between multiple scales of analysis (Nassaney and Sassaman 1995; Pauketat and Alt 2005) Our understandings of the Swift Creek archaeological culture are impoveri shed without situating the phenom enon within the context of the various local social groups that made and us ed complicated stamped pottery, as well as vice 53

PAGE 54

versa. Indeed, Swift Creek Com plicated Stampe d pottery appears as a veneer across the lower southeastern landscape and seems to have connect ed a variety of distinct social groups with unique histories (Ashley and Wallis 2006). The r easons for the adoption of Swift Creek pottery, the use of complicated stamped vessels, and ultim ately the movement of individual carved or stamped objects, grew out of unique (though connected ) historical and social circumstances that require analytical attention. In what follows in subsequent chapters, multiple lines of evidence indicate that earthenware vessels were exchange d as significant gifts th at constituted social relationships across considerable distances along the Atlantic coast. This conclusion emerges out of a reconstruction of object worlds and the speci ficities of past practi ce at local levels that can be situated within the globa l-scale phenomenon that is Swift Creek culture. As a point of departure, I now turn to the Swif t Creek archaeological culture. 54

PAGE 55

CHAP TER 3 THE SWIFT CREEK CULTURES The Swift Creek archaeological culture is defi ned almost exclusivel y by the production and use of complicated stamped pottery across th e Woodland period lower southeastern United States. Specifically, Swift Creek complicated stamped pottery became popular circa AD 100 and continued to be produced until around AD 850 in pr esent-day Georgia and portions of adjacent states (Figure 3-1; Stephenson et al. 2002:318). Although the term Swift Creek has often been used to designate both a type of pottery and an archaeological culture, an explanation for the fluorescence and persistence of complicated st amping must begin with recognition of the apparent social diversity that existed among the people who made a nd used this type of pottery (Anderson 1998:275; Ashley and Wallis 2006:5). Indeed, as Milanich (1999:704) indicates, likeits taxonomic siblings Swift Creek has proven to be a ha ndful. When it comes to its role as a ceramic assemblage, it is well behaved. But wh en Swift Creek is used to designate a single culture, it can be a schizoid problem. This ch apter summarizes previous research of Swift Creek contexts and highlights the archaeological di versity that the term encompasses with focus on technology, chronology, social an d cultural attributes, social in teraction, and the meaning of designs. At the end of the chapte r, I explore new interpretations of Swift Creek interaction based on the representative techniques evident in designs and their spatial distribution on vessels. The Middle Woodland Period (ca. 200 B.C. to A.D. 400) saw extensive long-distance exchange and interaction networks which culminated in the Hopewell interaction sphere that spanned most of the Midwest and surrounding regi ons (e.g., Caldwell 1964). With artifacts and raw materials of chert, copper, galena, marine shell, mica, and obsidian exchanged or carried long distances across the landscape, Hopewell interaction imparted a semblance of cultural and religious continuity among many otherwise dist inct Middle Woodland so cieties (Caldwell 1964; 55

PAGE 56

Struever 1964). The fluorescence of long-dist ance exchange and earth en m ound building in some areas, along with the preferen tial treatment of some individua ls in burial, arguably reflects the rise of more powerful descent groups that vied for increasing political influence (Anderson and Mainfort 2002:10; Smith 1986:48). While cultural connections were indeed far-reaching, many Middle Woodland societies were comparativel y insular and apparently did not participate in exchange networks or mani fest power disparities among kin groups to a significant degree (Anderson and Mainfort 2002:10). In fact, materi als derived from Hopewe ll interactions were overwhelmingly concentrated at major ceremoni al mound centers, especially the geometric earthworks and large conical burial mounds in southern Ohio and west-central Illinois. Likewise, connections further afield, such as in the lower Mississippi va lley, northern Alabama, and western Georgia, were also restricted to earthen mound center s, where mortuary regimes and exotic materials indicate H opewell influence (Cobb 1991:176). Hopewell materials were brought south to the Swift Creek cultures at mound centers in western areas of Georgia and Florida, and, moving in the oppos ite direction, Swift Creek potte ry was deposited at Hopewell sites toward the north (Brose 1979b; Kellar 1979; Smith 1979). Swift Creek cultures toward the east were somewhat less connected to Hopewell ex change networks (Seeman 1979). Like Hopewell, Swift Creek complicated stamping seems to have been implicated in social connections across various distinct societies rather than the trademark of a monolithic and insular culture. Some of the earliest complicated stamping is associated with Deptford contexts in northwestern Florida and southwestern Ge orgia (Knight and Mi stovich 1984:217-220; Stephenson et al. 2002:335). As the popularity of complicated stam ping proliferated across large portions of present-day Florida, Georgia, and Al abama, as well as the edges of South Carolina and Tennessee, broad regional c onnections became apparent. There is a strong association 56

PAGE 57

between W eeden Island cultures and Late Sw ift Creek complicated stamped pottery in southwestern Georgia and northwestern Fl orida (Willey 1949:396-409, Sears 1962; Milanich 1994:166). Fairbanks (1952) referr ed to the relationship betw een Swift Creek and Weeden Island as an intimate but separate presence. Going further, Sears (1956) equated pottery directly with people to suggest that Weeden Isla nd pottery indicated activ ity areas of an elite class in contrast to a subser vient Swift Creek pottery-using gr oup. This correlation is unlikely, but the Swift Creek series is known to have ge nerally preceded the production of Weeden Island pottery and then continued to be produced alongside Weeden Island wares in northwestern Florida and southweste rn Georgia. While the eastern panhandle Late Swift Creek populations between the Aucilla and Apalachicola rivers are associated with Weeden Island culture, Early Swift Creek west of the Apalachicola River shows similarities to the contemporaneous Marksville archaeological culture of the Lower Mississippi Valley. Western panha ndle Swift Creek manifest ations are different enough to receive a different appellation, Santa Ro sa-Swift Creek (or Flo rida Marksville). Santa Rosa-Swift Creek sites include both Swif t Creek Complicated Stamped and Santa Rosa Series pottery, the latter consisting of many types that are considered Marksville variants such as Alligator Bayou Stamped, Basin Bayou Stampe d, Santa Rosa Stamped, and Santa Rosa Punctated (Milanich 1994:152; Stephens on et al. 2002:334; Willey 1949:372-378). There are other affinities expressed on the various geographic periphe ries of the primary Swift Creek pottery distribution. Pickwick Complicated Stamped pottery in the Tennessee River Valley is recognized as a varian t of Swift Creek and tends to be a minority ware on multicomponent sites (Elliot 1998). In areas of northern and western Georgia Swift Creek complicated stamped pottery is found among Carter sville assemblages while Deptford sites in 57

PAGE 58

eastern Georgia and South Carolina som etimes contain complicated stamped sherds (Anderson 1998:277). On the lower St. Johns River, Flor ida, Swift Creek assemblages often include Weeden Island, St Johns, and Deptford series sherds Late Swift Creek sites in particular seem to include St. Johns Plain pottery as a minority ware, with the highest fr equencies at mortuary mounds (Wallis 2007). Swift Creek Complicated Stamped pottery ha s also been found at many important Middle and Late Woodland period sites far outside its primary distributi on. These include various sites in southern Florida and Hopewell sites in Oh io and Indiana (Milanich 1994:142; Ruby and Shriner 2006; Stephenson et al. 2002:349). Clearl y, complicated stamped pottery often served as the material for social connections in the Woodland period Southeast and the diversity of contexts in which it is found may reflect an intr icate history of various kinds of interaction between social groups. Swift Creek Technology: Wood and Earthenware Swift Creek pottery is defined by the stamped impressions of complicated designs on the exterior surfaces of vessels. These designs were achieved by impressing a paddle into the surface of still-wet vessels before drying and firing. These paddles followed a long-lived tradition of pottery manu facturing technology in the Eastern W oodlands, in which various simple stamped, check stamped, and fabric impressed surface treatments were used in some of the earliest Woodland period pottery traditions (Chase 1998:49). Al most always made of wood, paddles had lines carved into them or were wrap ped with fabric or cordage and stamped into vessels to achieve the desired e ffect. Complicated stamping appear s to be a departure from the technology of the previous millennium only in the c ontent of the designs carved into the paddles. Rather than being limited to pottery, the intr icate and sometimes orna te designs found on Swift 58

PAGE 59

Creek pottery m ay in fact be a mere glimpse of a prolific wood-carving trad ition that culminated in entire object worlds being engraved (e.g., Williams and Elliott 1998:10). Owing to the unlikelihood of preservation, no carved wooden paddles have ever been reported, although at least one earthenware complicated stamped paddle has been found (Milanich 1994:146). Impressions on vessels that show the grain of the wood and characteristic cracks indicate that the paddles were primarily wooden. Snow (1998:71) suggests that the long axis of the wooden paddles was oriented parallel to both the wood grain and the cracks that sometimes developed on the paddle face. Judging from stamped impressions, the faces of the paddles themselves appear to have been roughl y rectangular (Snow 1998:70) and generally about 10 to 15 cm in length and 10 cm or less in width. Both sides of the paddle head may have been carved, but interestingly only one example of a vessel stamped with two designs has ever been documented (Snow 1998:71). There are also rare examples of negative impressions of known designs, indicating that complicated stamped sherds were occasiona lly used as a substitute for carved paddles (Snow 1998:67). Snow (1998:67) calls these sher ds convenience paddles and documents paddle (positive) and sherd (negative) matche s at the same site in central Georgia. The technique for impressing designs into vessels appears to have been somewhat variable and may have been context specifi c (Wallis 2007:215). Snow (1998:72), who has analyzed more Swift Creek designs than any ot her researcher, argues th at the primary purpose for paddle stamping was to bind clay coils during manufacture. This assertion is based on a number of attributes common to Swift Creek co mplicated stamped pottery. First, the whole design, as it would have appeared on the wooden paddl e, is almost never registered on a vessel. Instead, the extremities of the paddl e, particularly the proximal end, seem to have rarely had contact with the wet clay surface. Second, design orientation is variable on vessels. 59

PAGE 60

Representative figures such as m asks or anim als, although they are al most always oriented with the head toward the distal end of the wooden paddle, are turned a number of directions on vessel surfaces. Third, overstamping is common, resulting from the frequent overlapping of paddle impressions. Finally, stamped impressions were often smoothed over prior to drying and firing. All of these characteristics seem to indi cate that clear, unadulterated design impressions were not intended, but that overl apping designs covering the vessel achieved the desired effect (Snow 1998:72). In contrast to Snows (1998) assertion, Br oyles (1968:54) argues th at the primary purpose of Swift Creek paddle stamping was to transf er designs onto vessels. Analyzing vessels primarily from Kolomoki, Fairch ilds Landing, Mandeville, and the Quartermaster site, Broyles observed a general lack of overstamping a nd the prevalence of spaces between stamping impressions. If the carved wooden paddles were used to shape the vessels, Broyles (1968:54) reasoned, then overstampingwould necessarily result from pounding the vessel into shape. I believe the divergent conclusions of Snow a nd Broyles are due, in part, to their analysis of different assemblages and the contextually variable role of carved paddles in the manufacturing process (Wallis 2007 :215). Like Broyles (1968), I find it unlikely that carved paddles were always used to form vessels into their final shapes, unless the un-carved sides of the paddles were used or the vessels were smoothe d before final design application. At the same time, the seemingly haphazard and overlapping placement of paddle impressions, and prevalence of smoothing on some vessel surfaces, demonstrates that crisp and complete designs were often not desired. In short, carved paddle stamping may have been a pplied after vessels had been formed into their desired shape but the resulting surface treatment was somewhat variable in its 60

PAGE 61

execution. The practice of stam ping a vessel with a carved paddle may have been more significant than the quality of the de sign registered on the final product. While there are some general trends in vesse l attributes that lend themselves to fine seriation (as discussed below) the overall morphology and func tion of Swift Creek vessels was variable. In fact, Swift Creek complicated stamping seems to have been used on vessels of many shapes, sizes, and tempers. Indeed, in many regions Swift Creek complicated stamping appears to have simply replaced the earlier Dept ford check stamping on a similar suite of vessel forms (Sears 1952:103). Chronology Swift Creek Complicated Stamped was formally defined as a pottery type by Jennings and Fairbanks (1939) one year after Kelly (1938) had published a prel iminary report of excavations at the Swift Creek type site ne ar Macon, Georgia. Based larg ely on Kellys (1938) stratigraphic excavations in Mound A, Early, Middle, and Late Swift Creek types were defined. Early Swift Creek pottery is characterized by weakly impresse d designs with narrow lands, the prevalence of rectilinear designs, tetrapodal ba sal supports, and very small rim folds or notched or scalloped rims. Middle Swift Creek pottery tends to have more curvilinear motifs, stamped impressions are better registered, and rim folds are more prom inent. Finally, Late Swift Creek pottery has carelessly applied designs, designs applied in re stricted zones (usually the top third of the vessel), and very large rim folds. This chronolo gy has been generally in terpreted as the rise, culmination, and decline of a pottery style (Caldwell 1958:37). Outside of central Georgia, the Middle Sw ift Creek designation has been rarely used. Based on stratigraphic analysis of Swift Creek on the Gulf Coast of Florida, Willey (1949:378) independently defined Early and Late varieties. Many of the attribut es used to differentiate Early and Late Swift Creek on the Gulf Coast were cons istent with Kellys (1938) findings, including 61

PAGE 62

the differences noted in rim form, design st yle, and stamping execution. However, Willey (1949) found no evidence of Middle Swift Creek in northwestern Florida, a disparity that may be due to the limited date range of the type site. In comparison to other assemblages from northwestern Florida and southern Georgia, Mound A at the Swift Creek site primarily dates to the Early and Middle Swift Creek phases with very few examples of Late Swift Creek pottery (Fairbanks 1952:288; Price 2003). Throughout the geographic distribution of Swift Creek pottery, chronological trends have been noted not only in the ex ecution of stamping but in the designs themselves, with Early Swift Creek de signs generally less complex in comparison to Late Swift Creek designs (Anderson 1998:277; Fair banks 1952:288). However, these labels are subjective and the trends may be locally specific, making rim forms more reliable time markers than stamping designs. Knight and Mistovich (1984) offered a refi nement of Kellys (1938) chronology based on work in the Chattahoochee River drainage in sout hwestern Georgia and southeastern Alabama. Their Mandeville, Kolomoki, and Quartermaster phases, based on pottery from each of these respective sites in the region, roughly correspond with Kellys (1938) Early, Middle, and Late chronology. A 50 year gap between Mandeville a nd Kolomoki phases was inferred by Knight and Mistovich (1984), but this temporal gap is filled by potte ry from the submound midden at Hartford in central Georgia (Snow and Stephe nson 1998; Stephenson et al 2002:342). However, outside of the Chattahoochee Rive r area, the Knight and Mistovich (1984) chronology is rarely used. In northwestern Florida, for example, Kolo moki phase pottery is simply called Late Swift Creek (Smith 1999). The Diverse Social Landscape The Middle and Late Woodland periods sa w major earthen mound-building projects, numerous multi-house circular villages, and wide spread long distance exch ange and interaction. 62

PAGE 63

However, there was also much social and cultura l diversity across the Southeast during this time and some populations were comparatively unconnected and apparently chose not to participate in these cultural trends. The Swift Creek archaeologi cal culture played a major role in the social landscape of the Woodland period bu t it also encompassed an impre ssive amount of diversity in monuments, settlement, mobility, and interaction. The broad range of this diversity necessitates an emphasis on particular contexts and highlight s the problems associat ed with treating Swift Creek as a coherent culture. As with many archaeological cultures of the Eastern Woodla nds, Swift Creek culture and culture history was constructed primarily th rough pottery typology and seriation, and more recently, absolute dating of carbon attached to or associated with pottery. This regional trend results from the ubiquity of pottery in the archaeological record, particularly at sites with poor preservation where little else is found. However, the failure to incorporate other archaeological data in the construction of arch aeological cultures has sometimes led to interpretive problems in considering social interaction a nd culture change, or worse, the uncritical conflation of ceramic types and actual cultural groups. Archaeological cultures are admittedly necessary abstractions that archaeologists must use to structure the he terogeneous archaeological record, yet uncritical use of these devices harbors the corresponding tendency to mask archaeological differences among assemblages, sites, and landscapes, especial ly when they are based on a single class of artifact. At the very least, variatio n in subsistence, technol ogy, and settlement patterns among regions provide the basis for differentiating subgroups within widespread archaeological cultures. For example, the Deptford archaeological culture, which stretches across much of the same expansive lower southeastern region as Swift Creek culture, has been recently 63

PAGE 64

reconstructed based on new archaeological evidence th at indicates considerab le cultural diversity corresponding with geographical an d ecological regions. Deptford was first defined as a coastal culture by excavations at the type site (9CH2) near the m outh of the Savannah River (Caldwell 1952; Waring and Holder 1968), and subsequent research by Milanich (1971) on Cumberland Island. A year-round coastal tradition that was based on maritime resources was indicated by the predominance of shellfish and fish in midde ns and the discovery of both winter and summer houses (Milanich 1971, 1994:124-215). Sites with Deptford series pottery on the interior coastal plain, much smaller than most coastal sites, we re assumed to indicate short-term occupation of inland forests, probably hunting a nd foraging expeditions that targ eted deer, hickory nuts, and acorns (Fradkin and Milanich 1977; Milani ch 1994:120-123; Tesar 1980:688-794). However, this model of coastal settlement with only peri odic forays inland is complicated by the discovery and excavation of substantial, y ear-round Deptford sites on the interior coastal plain (Anderson 1985; Brooks and Canouts 1984; Stephenson et al 2002). Although these sites contain checkstamped pottery characteristic of the Deptford se ries, the assemblages are qualitatively different, with a much higher frequency of linear check stamping decoration on pottery (Milanich 2004). In addition, so-called Deptford in western Ge orgia appears more closely related to other archaeological cultures with check-stamped pottery to the north and west, such as Cartersville (Smith 1975; Stephenson et al. 2002:331-2). Moreover, inland Deptford sites contain consistently greater numbers of stone and bone tools in combina tion with a subsistence regime focused on deer (Milanich 2004). Due to these differences, and because there is little convincing evidence that inland and coastal populations we re connected beyond the popularity of checkstamped pottery, separate archaeological comple xes and different developmental sequences are 64

PAGE 65

recognized, including Gulf, Atlant ic, and Interior/Riverine Dept ford (Milanich 2004; Stephenson et al. 2002). The Swift Creek archaeo logical culture is in n eed of similar revision in order to emphasize the temporal and cultural variati on that the term encompasses. As Milanich (2002:371) argues, [I]t does not make a great deal of sense to labe l as Swift Creek an A.D. 100 shell midden site on the panhandle coast of Florida, an A.D. 700 s ite near Macon, Georgia, and a site of unknown antiquity near Brunswick, Georgia. In fact, si milar Gulf, Atlantic, and Interior/Riverine Swift Creek subregions might be usefully recognized based on different subsistence and settlement patterns. Santa-Rosa Swift Creek sites are ov erwhelmingly located on the coastal strip, on or near the coast (Stephenson et al. 2002:342). Swift Creek sites along the Atlantic seaboard are also primarily within the coastal sector, and subsistence remains from sites in both regions indicate an almost exclusively maritime diet of fish and shellfish (Byrd 1997; deFrance 1993; Fradkin 1998; Reitz and Quitmeyer 1988). Among Santa-Rosa Swift Creek sites, at least three configurations of midden have been identified : ring middens, linear middens with long axes oriented along the coastline, and small dumps (Milanich 1994:144-5; St ephenson et al. 2002). The same configurations have also been identi fied along the Atlantic coast (Ashley and Wallis 2006; Saunders 1998). Riverine-focused Swift Creek groups occupied central and southern Georgia with an entirely different subsistence and settlement stra tegy. Subsistence data from the interior coastal plain are woefully underrepresented because of a persistent preservation bias. However, data from a few well-preserved contexts in the p ine barrens indicate that deer overwhelmingly made up the bulk of the Woodland pe riod diet (Carder et al. 2004). At the Hartford site (9PU1) near the Ocmulgee River, deer comprised 35% of the MNI and nearly 95% of the biomass in the 65

PAGE 66

vertebrate assem blage (Carder et al. 2004:31). Birds and othe r mammals made up most of the remainder of the assemblage, with almost neg ligible meat weights coming from fresh water fishes and turtles. Settlement data come primarily from the nearby Ocmulgee Big Bend region from Snows (1977) extensive pedestrian surveys. In this area the settlement pattern consisted mostly of small seasonal resource extraction sites with a few larger sites serving as central or base sites (Snow 1977; Stephenson et al. 2002:345) The implication is that communities were relatively mobile on a seasonal basi s as part of an ad aptive strategy that focused on deer hunting. This degree of mobility stands in contrast to th e more permanent settlements evident on the Gulf and Atlantic coasts. However, the Ocmulgee Big Bend data should not be used as generalizations for the entire interior of Georgi a since many large and more permanently settled sites (including Hartford) existed with in the interior as well. In contrast to the Deptford evidence, where cultural relationships between subregions are somewhat tenuous, design matches on Swift Creek pottery indicate that ar tifacts were carried between coastal and inland villages (Ashley et al. 2007; Kirkland 2003). Thus, diverse cultures across the lower Southeast were connected via ca rved paddles or stamped pots but often with little other evidence of cultural influence acro ss subregions. Swift Creek pottery-producing cultures appear to have embodied the entire rang e of cultural and social diversity that existed during the Woodland period, ranging from small groups of mobile hunter-gath erers to larger and more centralized populations that undertook majo r mound-building projects. Even with this range of diversity, however, the proliferation of ceremonial mound centers and circular villages are important topics of discussion for their apparent connections to Swift Creek potteryproducing cultures. 66

PAGE 67

Ceremonial Centers More than m erely a precocious pottery tradition of otherwis e simple societies (Caldwell 1958:44), Swift Creek is associated with larg e monuments that were once assigned to the Mississipian period. Platform m ounds were constructed for the first time on a large scale during the Middle Woodland period (Lindauer and Blitz 1997; Jefferies 1994; Mainfort 1988; Pluckhahn 2003; Thunen 1998). Many Woodland peri od platform mounds were associated with the Swift Creek archaeological culture, includ ing at the sites of Swift Creek, Kolomoki, Mandeville, Annewakee Creek, Cold Springs, and pr obably Evelyn in Georgia, Garden Creek in western North Carolina, and McKeithen and possibly Block-Sterns in Florida (Anderson 1998:290). These mounds generally lack evidence of buildings on their summits and probably functioned as ceremonial stages, access to which might have been restricted to privileged factions (Jefferies 1994; Lindauer and Blitz 1997). Many of the mounds, which vary in size and scale, were built in stages over time, indica ting that the functions and meanings of the monuments were continually ch anging through recurrent practice (Jefferies 1994; Lindauer and Blitz 1997). Anderson (1998) suggests that sites with platfo rm mounds were positio ned as gateways or trading centers to take advantage of the exchange of shell and other items in to the interior of the Eastern Woodlands. This idea is consiste nt with a common Middle Woodland theme of dispersed populations living in scattered households or small villages coming together periodically at regional cent ers (Anderson 1998:283). However, some platform mounds were also bordered by villages with resident populations. Kolomokis Mound A, the largest example of a Middle Woodland platform mound (ca. 17 m high), had a substantial village population (Pluckhahn 2003). The function of platform mounds may also have varied. At McKeithen, a Weeden Island village site w ith three mounds, the platform mound was apparently built to 67

PAGE 68

support the residence of a high ranki ng big m an (Milanich et al. 1997 ). It is also possible that mound complexes with associated villages we re ceremonial centers for surrounding social groups as well, but the association of ceremonial practice with a select re sident group may have been distinctly different from the practice of converging at vacant ceremonial centers. In northern Georgia, Williams and Harris (1998) noted an even spacing among Swift Creek mound sites, about 30 km apart. These sites appear to have been mortuary mounds without associated villages and ve ry little associated midden. Fu rthermore, these sites are often located on the tops of high ridges, away from the river floodplains that ar e comparatively richer in subsistence resources. Williams and Harris (1998) label these mound sites as shrines, substantiating the idea that Middle Woodland populations gathered at vacant ceremonial centers. Other sites that lack platform mounds are also likely to have been civic-ceremonial gathering centers for disparate popul ations, including sites such as Little River, Milamo, Fortson Mound, and Hartford (Stephenson et al. 2002). At th e Hartford site (9PU1), excavation within a submound midden revealed a large structure with interior hearth (Snow and Stephenson 1998). At 12.5 m across, this structure was larger than might be expected for Middle Woodland domiciles (cf. Anderson 1985; Pluckh ahn 2003). The site is at th e confluence of major historic Indian trails and was apparently a crossroads during the Woodla nd period as well. The site contained exotic marine fauna and other nonloc al materials such as quartz crystals, mica, hematite, graphite, galena, copper-bearing silver and chert from the Ridge and Valley province (Carder et al. 2004). Faunal analysis of the hear th contents also indicates the preferential selection of choice cuts of deer meat. All of this evidence le d researchers to conclude that the Hartford submound context represents a ceremoni al center where groups gathered periodically (Carder et al. 2004; St ephenson et al. 2002). 68

PAGE 69

Elsewhere, perm anently nucleated villages c onstructed their own mortuary and ceremonial spaces sometimes without obvious contributions from populations li ving beyond the immediate region, in some cases perhaps co rresponding with more insular, alienated people (Knight 2001; Milanich 1994; Milanich et al. 1984, 1997). Residential sites with adja cent mortuary mounds, for example, are not uncommon on the Florid a Gulf Coast (Willey 1949). However, even relatively modest mounds and the village ri ng middens themselves frequently contain Hopewellian prestige items that show connec tions to the Midwest (Seeman 1979; Stephenson et al. 2002:242-243). In fact, ba ked clay figurines are often found within Santa Rosa-Swift Creek and Swift Creek middens along the Gulf Coast and are presumed to be associated with the Hopewell Interaction Sphere (K eller and Carr 2006; Milanich 1994:148). In general, Hopewell Interaction Sphere items seem to be concentrated at Early Swift Creek and Santa Rosa-Swift Creek mortuary sites in western Florida and Georgia, with the la rgest sites containing the most numerous exotic items (Seeman 1979). Althoug h there is still much to understand about the social structure and practice that was manifested at mound centers, the variations in the size and orientations of locations on the built landscape may reflect major differences in both the mobility and the degree of connectivity of various so cial groups across the lower Southeast. Swift Creek Villages Many large Swift Creek sites a ppear to have been organize d according to a similar, basically circular community pl an, showing continuity with Ar chaic and Early Woodland site structure. Circular and horseshoe-shaped middens in Swift Creek contexts have been frequently interpreted as village sites corresponding to the arrangement of houses around a central plaza (Bense 1998; Bense and Watson 1979; Stephenson et al. 2002:346; Thomas and Campbell 1993; Wallis 2007). At some sites numerous postmolds and even substantial fragments of burned daub near the perimeter of these predominantly shel l-constituted rings supports the interpretation of 69

PAGE 70

these sites as villages, although no posthole patt erns have been recorded (Stephenson et al. 2002:346). Considerable m orphologi cal variability exists among the Swift Creek ring middens in Florida and Georgia. A continuous midden of dense shell occasionally over a meter high or a series of discrete middens arranged in an ar c are both common configur ations of the village perimeter. Both complete, closed circles and semi-circular or horseshoe shaped middens have been observed. The orientations of the openings of the semi-circular middens are variable and do not conform to any pattern (S tephenson et al. 2002:346). Based on seasonality studies and radiocarbon assays, many Swift Creek circular villages appear to have been occupied for long periods of time or were returned to repeatedly over many years. At the Horseshoe Bayou site near the Choctawatchee Bay, Florida, for example, the species of fish represented in the faunal a ssemblage suggest occupa tion during every season, while radiocarbon results indicate that the duration of Swift Creek activity may have been protracted over more than two centuries (Bense 1998:253-255). The size of Swift Creek circular villages s eems to have been at least inter-regionally variable. Bense (1998:257) reports an average diameter of approximately 100 m for Santa RosaSwift Creek villages but villages may be as sma ll as 50 m on the Atlantic coast of Florida (Smith and Handley 2002; Wallis 2007). Plazas are generally characterized as swept clean (Stephenson et al. 2002:342; Thomas and Ca mpbell 1993), however, a notable exception was discovered at the Bernath site in the western panhandle of Fl orida, where more than 300 individuals were estimated to have been interred within the cen tral space (Bense 1998). Another plaza cemetery, perhaps comprised of a very low mound in the center of a circular v illage, may have been discovered in northeastern Florida at Greenfi eld site no. 8/9 (Johnson 1998). Burials at ring middens may be more common than currently un derstood due to inadequate sampling within 70

PAGE 71

sterile plazas (Bense 1998:269), but adjacent or nearby m ounds seem to have served as mortuary repositories for most Swift Creek village populations (Stephenson et al. 2002:345). An alternative pattern is recognized in northeastern Florida, where village sites normally appear to have been situated some distance away from mortuary mounds (Ashley and Wallis 2006; Wallis 2008). The consistency with which Swift Creek villag es were organized according to a circular plan is indeed striking, showi ng continuity across thousands of miles and hundreds of years. Regardless of their consistency, however, circular villages are not a uniquely Swift Creek trait, having been popular among Archaic cultures (e .g., Stallings) and many contemporary and adjacent Woodland cultures (e.g., Deptford, Weeden Island, Marksville) as well (Milanich et al. 1984:53-54; Phillips 1970:265-353; Sa ssaman 2006). Regardless of the specific social and structural implications of circular village design, these site plan data are important for establishing that Swift Creek groups commonly lived in multi-household villages that were oftentimes permanent or semi-permanent settlements. Swift Creek Interaction Swift Creek cultures were important participan ts in the various far-reaching interactive networks of the Hopewell Interaction Sphere. The Hopewell Interaction Sphere is characterized by the long-distance movement of raw materials and finished artifacts, in the past interpreted as various unitary phenomena, including a trade network, mortuary cult, shared religion, and interaction of a peer polity (Bra un 1986; Caldwell 1964; Ca rr 2006b; Struever and Houart 1972). However, more recent research has recognized separate and geographically distinct patterns in the distribution of Hopewell artifacts that are likely to have corresponded to different and distinct practices (Carr 2006a; Seeman 1995). The H opewell Interaction Sphere is thus beginning to be deconstruc ted into its constituent parts in a way that beguiles the old 71

PAGE 72

unif ying synthetic concepts that celebrated a monolithic Hopewell society across the landscape. Carr (2006a:53) defines Interregi onal Hopewell as a composite of multiple, diverse kinds of practices, ideas, and symbols, which had their or igins in multiple, differing regional traditions and were shared or operated at multiple, different supraregional scales. Caught up in this complex web of material distribution and exch ange, Swift Creek cultures were on both giving and receiving ends of materi al exchanges and influence (K ellar 1979; Seeman 1979; Smith 1979). Copper panpipes and cymbal ornaments, stone plummets and gorgets, along with other ostensibly Hopewellian artifacts are found in variable frequenc ies at Santa Rosa-Swift Creek, Swift Creek, and other Middle Woodland sites in the Deep South and Gulf Coast (Kellar 1979; Sears 1962). Conversely, conch shells were proc ured from the Gulf coastal areas inhabited by Swift Creek populations and distri buted as far north as Michig an and New York (Carr 2006b). In a more limited way shell from the Atlantic coastal plain was also distributed north, as evidenced by Altamaha Spiny Mussel shells ( Elliptio spinosa) with Hopewell engravings recovered from a mound site in Indiana (Farnsworth and Atwell n.d.). Swift Creek Complicated Stamped pottery is also found at several Hopewe ll sites in Ohio and Indiana (Ruby and Shriner 2006; Stephenson et al. 2002). Swift Creek cultures were certainly partic ipants in various Hopewellian networks, however, the heyday of Hopewell interaction (ca. 200 BC to AD 400) predates many of the Swift Creek cultures that thrived until at least the ninth century. The temporal extent of Hopewell manifestations is well-defined in Sears (1962) definition of the Yent and Green Point ceremonial complexes along the Gulf Coast of Flor ida. Based on the site s of Crystal River, Pierce, and Yent, the Yent Complex dates to th e Deptford phase and exhibits a long list of Hopewellian ceremonial traits. The related but later Green Point Complex, defined by four 72

PAGE 73

Swif t Creek phase sites, demonstrates compar atively fewer Hopewellian traits. While the coherency of Sears (1962) complexes is questiona ble, they provide a desc ription of the extent and timescale of Hopewell influen ce along the Gulf Coast. Hopewell traits were on the wane during Early Swift Creek times and were complete ly absent by the Late Swift Creek phase. However, Swift Creek interaction, as broadly de fined below, mostly operated independently of Hopewell interactive networks a nd continued long after the demise of Interregional Hopewell. As with recent deconstructions of Hopewell interaction, Swift Creek interaction might be explained by various social practices that operated at differing geographical scales. Researchers have long noticed patterns in the distribution of Swift Creek complicated stamped designs. Betty Broyles (1968) pioneered the study of Swift Creek design distribution through her painstaking reconstructi on of designs that facilitated their identification at multiple sites. This work was taken up primarily through the efforts of Frankie Snow (1975, 1977, 1998, 2003), who has outlined many of the attributes necessary for identifying vessels that were stamped with the same wooden paddle. These attr ibutes include design flaws, or idiosyncratic and non-symmetrical elements, and cracks in th e wooden paddle that are evident in stamped impressions. Using these attributes as part of the fingerprint of each design, Snow has identified many hundreds of paddle matches that li nk sites across Georgia and adjacent states. Over the course of more than three decades of research, Snow (2003) has compiled an impressive database of Swift Creek designs and lists of sites where each has been found. An important limitation of the database, however, is that the bulk of the designs derive from pottery from south-central Georgi a. This has resulted in more paddle matches being linked to south-central Georgia sites than to sites in any other region. 73

PAGE 74

Major gaps in our knowledge of Swift Creek de signs include the Gulf and Atlantic coasts of Florida. Snow has reconstr ucted a lim ited number of designs from these regions. There are also some records of design elements (Moore 1894, 1895; Saunders 1986; Wallis 2007; Willey 1949), but these have somewhat limited utility co mpared to Snows (2003) robust database of south-central Georgia complete designs. An exception to the general dearth of design data from outside south-central Georgia is southwestern Georgia, where a fa ir number of complete or near ly-complete designs have been recorded. Both Broyles (1968) and Snow (2003) have reconstruc ted designs from some of the most important sites in the region, including Kolomoki, Mandeville, Quartermaster, and Fairchilds Landing. This more limited inventory of designs has allowed for the identification of several paddle matches, most notably with sites in south-central Georgia (e.g., Snow and Stephenson 1998). Paddle matches between Swift Creek sites provide direct evidence of some form of social interaction (Snow and Stephenson 1998). On a regional scale, the identification of paddle matches or merely Swift Creek pottery outside of its primary distribution can be combined with petrographic and neutron activati on analyses to help determine whether pots or paddles were carried across the landscape (Stoltman and Snow 1998; Mainfort et al. 1997; Ruby and Shriner 2006; Smith 1998). Each of these studies that attempt to explain the dissemination of Swift Creek designs deserves careful c onsideration as important background to the present research. Using a fairly robust sampling strategy, Stoltman and Snow (1998) analyzed 69 petrographic thin sections from 11 sites, including 22 thin sections from vessels that had paddle matching designs linking them to two or more sites. The samples came primarily from the nearpiedmont site of Hartford (n=19) and the coastal plain site of Milamo (n=15). Most of the 74

PAGE 75

rem aining samples (n=24) came from sites near Milamo in the coastal plain of south-central Georgia. Using a paddle match between the Hartford site and three sites in the coastal plain, the authors argue that the identification of metamorphi c rock temper in all ma tching vessels indicate that vessels were moved from the piedmont into the coastal plain. Further, another paddle match between Hartford and a small site on the Oc onee River showed enough paste similarity to indicate the movement of a vesse l from the piedmont. An alternative explanation is offered for five other designs that were identified as paddle matches linking Milamo to various sites, including Hartford, Kolomoki, and several sites in the coastal plain. The variability in paste cons tituents among these vessels led the authors to conclude that paddles were moved between sites and that vessels were made with local materials. However, the authors acknowledge that the rang es of variation in paste composition overlap between many of the sites and in most cases sa mple sizes are inadequate (Stoltman and Snow 1998:149-151). Therefore, the number of paddle matches attributed to paddle movement and local production may be inflated. Whatever the ca se, the petrographic analysis was successful in identifying both instances in which paddles we re carried across the landscape and examples when vessels were moved. In a separate study, Betty Smith (1998) used Neutron Activation Analysis to identify nonlocal vessels at Mandeville, Swift Creek, and Bl ock-Sterns using 75 samples from earthenware sherds and figurines. Conducted in 1975 at th e University of Georgia, this early NAA study unfortunately suffered from a number of limita tions. Only eight elements were recorded (compared to the standard 33 at the University of Missouri Research Reactor (MURR)) and some of the most powerful multivariate statistical methods were not employed to analyze the results. Combined with the incomplete publishing of results, these data are therefore incomparable with 75

PAGE 76

other NAA research. Based on a cluster analysis and com parisons of summary statistics, Smith (1998) concluded that at leas t one sherd from Swift Creek and one from Mandeville was nonlocal. In addition, three ceramic figurines were identified as probabl e nonlocal productions. Mainfort and colleagues (1997) conducted INAA at MURR on a sample of sherds from the Pinson mounds. Along with other presumably nonlocal types, there were a total of 74 Swift Creek Complicated Stamped sherds recovered during Mainforts excav ations at Pinson. Mainfort et al. (1997) irradiated 163 samples from Pinson, including 4 Swift Creek sherds. Based on the chemical data, all of the Swift Creek samples were assigned to the same statistical group, A, which was judged to be comprised of only locally-produced pottery. Thus, in this case a Swift Creek carved wooden paddle appears to have been carried to Pinson mounds or manufactured there. Later, Stoltman and Mainfo rt (2002) objected to this finding with reference to petrographic analysis of these Swift Creek sh erds. The authors argued that slight textural differences between the Swift Creek sherds and a limited number of presumably local specimens from Pinson (n=32) and other near by sites (n=16) provide d "suggestive but not conclusive evidence of nonlocal status" (S toltman and Mainfort 2002:16). Given the preponderance of the data from the INAA study, I vi ew these Swift Creek sh erds as likely local productions. Finally, Ruby and Shriner (2006) attempted to as sign a source to Swift Creek pottery at the Mann site. The Mann site, a large Middle Woodland habitation site in southwestern Indiana, is known to contain copious amounts of Swift Creek Complicated Stamped pottery. Rein (1974) analyzed a sample of 984 complicated stamped sh erds from Mann and suggested that they were very similar to Early Swift Creek designs from cen tral Georgia. Further, Rein argued that three of the designs found at Mann were identical to ce ntral Georgia examples (i.e. they may be paddle 76

PAGE 77

m atches). To help explain the unusually high fre quency of Swift Creek pottery at Mann and five other habitation sites within a 40 km radius, Ruby and Shriner (2006) combined petrography with a number of other analytical methods, including scanning electron microscopy, x-ray diffraction, and sherd refiring. Of the 80 petrographic thin sec tions included in the study, 11 were derived from Swift Creek complicated stam ped sherds. Based on the similarity of clays and paste constituents, Ruby and Shriner (2006) c oncluded that all of the Swift Creek samples were produced locally. In fact, of all the pottery analyzed for the project only five simple stamped sherds were judged to have been manufactured nonlocally in the Southeast. In sum, paddle matches can be explained by both paddles being carried across the landscape and vessels being moved. Alternatively, the occurrence of Swift Creek complicated stamped pottery at more distant sites like Pinson and Mann appears be the re sult of local pottery production. Several explanations pertain to the distribution of local and nonlocal Swift Creek pottery at sites. First, group residential mob ility might explain the movement of both paddles and pots across the landscape as groups carri ed them during seasonal rounds and other migrations. Snow (1977:22; Stoltman and Snow 1998:152), for example, has noted that many Swift Creek sites in the Ocmulgee Big Bend regi on of southern Georgia appear to be small campsites used intermittently from year to year with some found on the floodplain and others on oak-hickory sand ridges, perhaps indicating a seasonal round. Paddle matches between these sites, evident on both locally made and imported ve ssels, likely indicate th e curation of material culture among residential sites. Paddle matches between more distant sites and the occurrence of Swift Creek pottery far outside its primary distribution may call for othe r explanations. Long-distance paddle matches with local paste have been hypothesized to co rrespond with patterns of post-marital residence 77

PAGE 78

am ong presumably female potters who curated their wooden paddles (Stephenson and Snow 1998; Stoltman and Snow 1998). Alternatively, nonlocal paste among Swift Creek vessels may be the result of gift giving and feasting (Ashle y and Wallis 2006; Stoltman and Snow 1998). Of course, both of these explanations could pertain to related parts of a coherent social process; they are not mutually exclusive. If Swift Creek descent groups formed marriage alliances then we might expect to find both the movement of obj ects owned by women with their changes in residence and gifts given to foster these relationships and repay associated debts. Further, the marriage alliance and gift explanations have been posited with implicit assumptions about the social context of wooden paddle and ear thenware pot production and use. Most fundamentally, wooden paddles are assumed to have been owned and used exclusively by female potters, though this was not necessarily the case. Instead, paddles may have been produced and owned by wood-carving specialists or owned by de scent groups rather than individuals. Evidence for the movement of wooden paddles does not necessarily correspond with the movement of individuals, married women or otherw ise, since paddles may have been objects of exchange (Saunders 1998; Wallis 2006). Identifying what kind of social interaction re sulted in the dissemination of Swift Creek designs in general and paddle matches in particular is inseparable from an understanding of Swift Creek materiality; specifically, the way that padd les, vessels, and designs were embedded in social life. Identifying the locations and frequencies of individual paddle designs is a prudent start toward this goal. For example, Snow and Stephenson (1998) iden tify the place where a paddle resided based on the comparatively high fr equency of its design on vessels at a site. These sorts of design distribution studies can be profitably combin ed with painstaking analysis of the archaeological contexts of Swift Creek pa ddle matches. In particular, a more complete 78

PAGE 79

understanding of the social cont ext of paddle production, use, a nd m ovement can come from a detailed analysis of large assemblages in wh ich paddle matches are identified. With this objective in mind, in the chapters that follow I pr esent the results of my analysis of many whole assemblages. Intrasite Design Distributions. There have also been attempts to identify patterns in the distribution of designs within sites. Bettye Broyles (1968:5 0) and Betty Smith (1998:113) observed possible sacred and secu lar dichotomies in designs at the Swift Creek and Kolomoki sites, with some designs restricted to pottery in either buria l mounds or midden areas. More recently, Pluckhahn (2007) employed symmetry an alysis to compare the designs that are differentially distributed at Kolomoki. Pluckhahn (2007) identified a higher frequency of symmetry in mound assemblages compared to n earby village contexts, and interpreted this difference as an active strategy of social incorporation in ceremonial contexts. Alternatively, Saunders (1998) recorded patterns in design distributions at village sites that might represent kin group affiliations. At the Kings Bay site (9CM171a), Saunders (1998) recognized that particular designs were circumscribe d within either the nort hern or southern half of the arc-shaped village midden while other de signs were found throughout the site. Saunders (1986, 1998) employed a concept of design hierarchy to argue that all of these designs were representations of group affiliation, with those more restricted in space signifying lineages, clans, or moieties and those more widely distributed repr esenting a larger social group or polity. In her comparison of stamping designs between Kings Bay and the nearby Mallard Creek site (9CM185), Saunders (1998) concluded that the same population inhabited both sites, perhaps for different purposes. 79

PAGE 80

At the Swift Creek village area at G reen field site #8/9 (8DU5544/5), I identified a minimum of 22 individual carved paddle designs stamped into a minimum of 187 vessels (Wallis 2007). Of the seven paddle designs with conclusi ve intrasite paddle matches, vessels with the same design were distributed widely across the site in ostensibly separate household middens. Consequently, I argued that if designs represente d clans or lineages, then they were shared among these household groups. This spatial distri bution does not negate the possibility that Swift Creek designs were used as emblems or totems, because their circulation on vessels may have been a significant practice in constituting relationships between descent groups (Wallis 2006a). I approach the interpretation of the m eaning of Swift Creek designs by combining two independent sets of data, the designs themselves and their distribution. The former is discussed below and data pertaining to the latter is the subject of s ubsequent chapters. Representational Meaning of Swift Creek Designs Many Swift Creek designs seem to incorporat e variations on a limite d number of common elements: spirals, concentric circles, ladders, and chevrons, to name a few. In addition to these common elements, a robust database of recons tructed whole designs reveals common motifs, some seemingly abstract and unrecognizable and others suggestive of plant, animal, and human representations (Broyles 1968; Snow 1975, 1977, 1998, 2007). Designs generally appear to represent birds, animals, flowers, and various anthropomorphic and zoomorphic faces. Further, Snow (1998) speculates that some designs may have attributes suitable for making more specific inferences, tentatively identif ying rabbits, serpents, mosquito s, buzzards, woodpeckers, horned owls, roseate spoonbills, and mythological beings th at have characteristics of multiple animals. Among the designs that appear to be faces or m asks, Snow (1998) identifies a wolf, owl, bear, and long-nosed god. These are potentially importa nt observations for tr acing the historical trajectory of particular representations. For ex ample, the eye, circle-i n-cross, and long-nosed 80

PAGE 81

god m otifs are important Mississippian and historic al-period symbols in the Southeast that may be traceable back through the Woodland period in Swift Creek art (Snow 1998). Interpreting the representationa l meaning of Swift Creek designs is nearly impossible without an understanding of the rules of representation that se em to dominate Swift Creek art. There is an undeniable grammar to the struct ure of Swift Creek designs, and Frankie Snow (1998) has outlined many of its principles. Firs t, Snow (1998) has recogn ized general geometric similarities, such as the particular prevalence of a quadripartite symmetry within designs. Bilateral symmetry and particul ar geometric transformations are also common (Pluckhahn 2007a). In addition, there are of ten purposefully asymmetrical elements incorporated into otherwise perfectly symmetrical designs. A second observation of Snow (1998) is that particular elements appear to be interchangeable between designs. For example, the concentric circle (sun/fire) and eye elements appear in larg ely identical contexts and may therefore be interchangeable and have similar meanings. Third, Snow (1998) has identified particular elements that are normally associated with one an other. For example, spirals often are present in association with the supposedly mo ving parts of represented bodies, su ch as the legs of a bear or the wings of a bird. Drawing on some of Snows observations, I have advocated an i nductive and formal analytical framework for assessing the constitu ent parts of Swift Creek styles (Wallis 2006b). This framework could be profitably modeled on the work of Washburn (1995), who delineated a corpus of representational transformations for categorizing and interpre ting decorative styles within their cultural contexts. Her premise is that visual images are more often structural descriptions of relationships rather than iconic pictures, and that these representations rest on a culturally informed repertoire of geometri c and metric transformations (Washburn 1995). 81

PAGE 82

Structu ral transformations are a necessary t ool for artists who attempt to render threedimensional images onto two-dimensional surfaces, and a relatively conservative corpus of these transformations are assumed to make up an artistic tradition with limited degrees of freedom for individual artisans working w ithin it (Washburn 1995). The essential points are that transformational rules inform artis tic styles, that these rules ca n be systematically understood by archaeologists, and that rules of representation and transformation may reflect structural aspects of a society. Split Representation. An important trend in Swift Cr eek representation that I have observed primarily concerns designs that appear to depict faces (or masks using Snows (1998) terminology), but also animals and other beings (Wallis 2006b). Although these faces do indeed have characteristics that make them distinct a nd might differentiate them according to different animals or persons, an equally profitable focu s of inquiry concerns the strict rules of representation to which these designs conform. These imag es are unlike Western graphic representational art with conventions for repres enting three dimensions on a flat surface, where depth perspective is achieved by interposition, rela tive height and size, linear perspective, texture gradient, and shadow. Instead, these faces are flattened by split representation, as defined by Boas (1955) and Levi-Strauss (1963). Split representation is a convention whereby a th ree-dimensional object is imagined as cut into parts and spread onto a flat surface, often with only one end of the object cohering. In its simplest form, split representation corresponds to tw o profiles conjoined at a central point, thus showing both sides of a face or a body at once. Mo re complex split representation can be used as well, depending on the number and orientation of the cuts through a body. The desired effect in all split representation is to display at once multiple sides of a three-dimensional body in two82

PAGE 83

dim ensions. The use of split representation may reveal important structural relationships within a culture. For example, Levi-Strauss (1963) and Gell (1998) argue that this technique is exclusively employed when the artwork itself constitutes a person or animal rather than merely representing or referring to one In other words, to the culturally informed subject, split representation does not just re nder an image that looks like an animal or person, the image or object is actually conferred a de gree of personhood. The same has al so been argued in regards to the representation of faces (irrespective of split representation) in contradistinction to heads As Deleuze and Guattari (1987:167-191) argue, heads are normally conceived of as connected to a body and are coded by the body in the sense that th ey are part of a complete organism. In contrast, giving an object a face and an expressi on dispenses with these corporeal coordinates and arguably lends a subject to or behind the face (e.g., personhood). If correct, these interpretations would be essential for interpre ting the movement of Swift Creek designs across the landscape. There are at least three attribut es of Swift Creek designs that reveal them to be structured by the principles of split repres entation. First, a flattened face in split representation often appears wide because the profiles have been spre ad apart (Figure 3-2A). Indeed, like the Haida and Maori examples referenced by Boas ( 1955:242-247) and Levi-Strauss (1963: Plate VII), Swift Creek faces are often nearly as wide or wider than they are long. Second, the splitting of the face into two halves only joined in limited pl aces frequently results in the appearance of a depression between the brows, as the forehead is cut down the center a nd spread to each side (Figure 3-2B). Finally, an exceptionally wide m outh is typical of split representation, as it is spread apart in the process of split ting (Figure 3-2C). Each of these traits can occur either in 83

PAGE 84

com bination or in isolation within each Swift Cr eek design. Thus, the over all similarity of many Swift Creek faces to those that are rendered by split representation indicates that they may have been conceived by similar principles. In addition to faces, whole bodies of anim als might also be represented by split representation. A Swift Creek design depicting birds (Snow 1998:94) is very similar to a Haida design illustrated by Boas (1955:226), differing primar ily in the parts that are cohering, the head in the former and the tail feathers in the latter There also exist much more complex forms of dislocation and recombination of body parts with in traditions of split representation, and this could potentially explain some of the rotati on and complex symmetry in other Swift Creek designs. There are significant social implications for sp lit representation in Swift Creek art, more than simply denoting a recognizable non-western representative technique. As Levi-Strauss (1963:263) indicates, split representa tion is used when social fact ors prevent the dissociation of graphic two-dimensional and plas tic three-dimensional images. This is most commonly the case when the plastic component consists of the face or human body along with the painted or tattooed decoration that is applied. In other word s, when socially constitutive indelible forms are inscribed onto the body, neither the body nor the design can be graphically represented without the other. This inextricable link between the design and the human face or body, in turn, implies the strict conformity of persons to their social roles (Levi-Strau ss 1963:264). Split representation is especially associated with masking cultures in which the person and mask are permanently linked, often achieved by indelible means such as tattooing (Gell 1998:195). While masks, for example, frequently enable persons to tempor arily become sacred be ings during particular ceremonial occasions, the use of split representa tion implies that the mask identity is not 84

PAGE 85

tem porary. Thus, split representation is necessa ry when motifs are critica l to, and inextricable from the constitution of social persons, and ther efore often corresponds with prestige struggles and rivalry among ranked lineages as ancestral identity is understood as embodied in living persons. The foregoing interpretation may provide a usef ul framework for answ ering the question of why Swift Creek designs were carved into small wooden manufacturing tools. As others have surmised, paddle designs were probably part of a prolific woodcarving tradition that included other artifacts that may have carried more impor tance than paddles (Williams and Elliott 1998). However, we should not downplay the social im portance of carved paddle s that were easily transported, and more importantl y, that provided an efficient way to disseminate images on a broad scale. Given the prominence of exch ange in pottery and/or paddles that is archaeologically visible via matching stamp desi gns on different vessels, complicated stamping may have been an effective way to extend soci al relationships and fu rther the renown of the descent group or person that ultimately claimed ownership of the design. If the principles espoused by Levi-Strauss (1963) and Gell (1998) apply to Swift Creek designs, then objects bearing designs (including vessels) may have be en conferred degrees of personhood or may have acted as extensions of persons. Consequentl y, the ownership, moveme nt, and exchange of paddles and pots with Swift Creek designs might have been directed by their qualities of inalienability, perhaps as embodiments of social persons such as important ancestors claimed by particular descent groups. Exchanges of artifacts with these designs might reflect heightened concerns with social rank constructed and legi timated through myths, ritual, and pedigree that are typical of societies that use split representation (LeviStrauss 1963:264). Of course, determining the symbolic density of Swift Creek complicated stamped vessels depends not only 85

PAGE 86

on understanding the principles of design representa tion, but also the entanglem ents of vessels in social practice that can be evaluated with context-specific studies of provenance and technofunction. In subsequent ch apters, I present data that show nonlocal complicated stamped cooking vessels were restricted mainly to mortuary contexts, which I argue shows that designs conferred symbolic density to vessels in a wa y that made them appropriate material for negotiating relationships be tween kin groups. Summary and Conclusions Identified by the wide distribution of complicated stamped pottery across the lower Southeast, the Swift Creek archaeological cu lture encompassed a dive rse array of Woodland period cultures with unique traditions and histories. C onsequently, the social connections evinced by the identification of complicated stam ping at sites across the region were probably as variable as the social groups who made and used Swift Creek potter y. Swift Creek pottery gained and maintained popularity throughout the lower Southeast through interactive networks that doubtless connected groups in various ways, l eading some to share ot her cultural features beyond complicated stamped pottery that lent a semblance of pan-regional cultural equivalence (Ashley and Wallis 2006:5). What constitutes Swift Creek in one area, however, is not necessarily the same as Swift Creek in other places. In order to begin to explain the widespread distribution and prom ulgation of Swift Creek pottery, an exploration of the nuances of local archaeological contexts is essential. Keeping with the idea of materiality as a dialectic of people and things, I am most interested in understanding the wa y that the material world was bound up in social relationships, and specifically in the case of Swift Creek pottery, how objects became extended artifacts (Robb 2004:134). A general aspect of an artifacts extension is in the sense of its positioning within social space and timethe normal or correct social context for an objects production and consumption. With th is in mind, I explore Swift Creek object 86

PAGE 87

worlds with focus on the socially prescribed u ses of pottery in general and complicated stamped pottery in particular. It is to the material wo rlds of Swift Creek potte ry-makers on the Atlantic coast to which I now turn. Notes 1. Although four Swift Creek sherds were reported in articles by Mainfort et al. (1997) and Stoltman and Mainfort (2002), five Swift Creek sherds from the site appear in the MURR database. 87

PAGE 88

Figure 3-1. Primary distribution of Swift Cr eek Complicated Stamped pottery and notable occurrences at distant sites (modified from Williams and Elliott 1998:6). 88

PAGE 89

89 A B C Figure 3-2. Swift Creek design attr ibutes characteristic of split representation. A) wide faces B) depression between brows C) wide mouth [reprinted with permission of Frankie Snow (2007)].

PAGE 90

CHAP TER 4 CULTURAL HISTORY AND ARCHAEOLOGICAL OVERVIEW OF THE WOODLAND PERIOD ON THE ATLANTIC COASTS OF GEORGIA AND NORTHEAST FLORIDA This chapter presents an overview of Woodl and period archaeology on the Atlantic coasts of Georgia and Northeast Florida, focusing on local manifestations of the Swift Creek archaeological culture. In many ways, the coasta l sectors of Georgia and Northeast Florida were different from one another in environment and culture throughout the Woodland period but these distinctions converged around the area of the Lower St. Johns River. With its long history as a boundary or transitional zone (Russo 1992: 107), the region that includes northeastern Florida and southeastern Georgia seems to have been a consistent location of inter-societal interaction. In terms of archaeo logical visibility, and ostensibly past practice as well, these interactions culminated with Middle and Late Woodland period Swift Creek cultures. This chapter begins by outlining the eco logical setting of the Atlantic coast and reviewing the preSwift Creek culture history of the Woodland pe riod. Building on this background, a review of the Middle and Late Woodland culture chronology of the Atlantic coast follows, focusing on the distinctions between north eastern Florida and southeastern Geor gia. I draw special attention to the mortuary landscapes of the Lower St. Johns River that consisted of series of mortuary mounds segregated from contemporaneous villages These distinctive mortuary landscapes were important inscriptions of descent-group ident ity and loci of social interactions across considerable distances, as evidenced by paddle matches between sites. The chapter concludes with a synthesis of important Swift Creek sites a nd a review of the evidence for social interaction in the form of paddle matches. The Ecological Setting The Georgia and northern Florida coasts occupy a central portion of the Georgia Bight, or Georgia Embayment, the long, westward-curving shoreline that extends from Cape Hatteras, 90

PAGE 91

North Carolina to Cape Canaveral, F lorida. Al ong portions of this stre tch of shoreline are a series of barrier islands that front the open ocean waters of the Atlantic and protect a network of tidal estuaries and wetlands that thrive between the mainland and leeward side of the islands. The islands themselves consist of a series of Pleistocene and Holocene barrier islands that formed through marine deposition and erosion duri ng fluctuations in sea levels, with some of them now combined to form large composite isla nds (Pilkey and Fraser 2002:245). Variation in the shape of these coastal plain islands, which in large part determine the character of the nearby estuaries, are also defined by differences in th e forces of tide and wave s along the length of the Georgia Bight. The contours of the coastline cau se tidewaters to accumulate toward the center of the Georgia Bight, giving the shoreline in Geor gia extreme tidal fluctuations, more than 3 meters, and trending toward more minimal tides of less than a meter toward each of the northern (Hatteras) and southern (Canav eral) capes (Davis 1997:158). In addition, the Georgia coast abuts a very wide and shallow c ontinental shelf that diminishes the size of waves through friction with the seafloor, resulting in the lowest wave s on the eastern coast of North America (Pilkey and Fraser 2002:248). This distinctive tide/wave en ergy balance results in a series of sea islands in Georgia (as well as one each in South Carolina and Florida) backed by wide and deep tidal inlets and extensive tid al flats and marshes many miles wide (Davis and Hayes 1984). Based on sea level data, essentially m odern coastal conditions began to be established by around 3000 years ago, when rapid sea level rise gave way to much more modest rates of rise (Davis 1997:157-158). The estuarine environments of the Georgia and northeastern Florida coasts sustain a rich and diverse maritime ecosystem. The salt marshes, comprised of a thick mat of salt-tolerant grasses on mudflats that are repe atedly submerged in salt water, provide nursery ar eas in the life 91

PAGE 92

cycles of m any marine vertebrate s and invertebrates. These marshes also supply nutrients to the surrounding estuarine and open waters as they drain with each e bbing tide (Pilkey and Fraser 2003:72-73). The marshes are punctuated by small is lands, many of them relict barrier islands covered by maritime oak hammocks, and an intricat e network of tidal creeks. Maritime forest comprised mainly of live oak also dominates the mainland and barrier island shorelines that face the marsh, while further inland the high sandhills are covered by pines and turkey oaks (Wharton 1978). This landscape is periodically interrupted by several large rive rs with freshwater sources. From north to south these rivers include the Savannah, Ogeechee, Altamaha, Satilla, St. Marys, Nassau, and St. Johns. The first three of these rivers ultimately drain the piedmont while the latter four all originate in the coastal plain. The low gradient of all of th ese rivers subjects their flow to heavy tidal influence at their lower reach es that typically results in a flooded estuary. Tidal influence and estuarine habitats are thus able to reach further inland at these rivers compared to the salt marshes limited to areas betw een the mainland and barrier islands. The St. Johns River has an exceptionally low gradient so th at tidal effects may reach as far south as Lake George, over 100 miles upstream, but true estuarin e habitats do not occu r much beyond 15 miles of the coast (Anderson and Gools by 1973:1-2; Ashley 2003:62). The coastal network of estuar ies, salt marshes, and maritime hammocks support extensive wildlife populations. Zooarch aeological studies indicate that Woodland period populations particularly targeted coastal area s for fish and shellfish, along with a diverse array of terrestrial and marine vertebrates, includi ng turtles, snakes, alligators, deer, opossum, rabbit, raccoon, and numerous species of birds (Fradkin 1998; Reitz and Quitmyer 1988; Reitz 1988). Paleobotanical research is conspicuously absent, but wild plants are assumed to have made up an important part of the diet as well. Coastal sector ecology is a ffected by fluctuations in temperature, rainfall, 92

PAGE 93

salinity, wind, and tide, which would have m ade res ource availability somewhat variable at least on a seasonal basis (Hackney et al 1976:273-276). The subtropical climate of the Georgia and northern Florida coasts is comprised of warm, humid summers and short mild winters, with water temperatures in the estuaries ranging fr om below 10 C to above 30 C (Reitz and Quitmyer 1988:96). Consequently, these fluctuating conditions result in pa tterns in the presence and distribution of particular spec ies of fish within the estuary. Based on the relative frequencies of various species of fish, along with seasonality data collected from Mercenaria clams, the Woodland period coastal sector seems to have been occupied during every season, and perhaps permanently year-round (Reitz and Quitmyer 1988: 105-106). This subsistence regime, focused on year-round exploitation of estuarine habitats is likely to have continued throughout the Woodland period and beyond, through various cultural changes. Pre-Swift Creek Culture History The first half of the Woodla nd period in eastern Georgi a and Northeast Florida was dominated by two archaeological cultures (Figur e 4-1). From roughly 800 BC to as late as AD 500, Deptford sites pervaded along the coast as fa r south as the Lower St. Johns River (Milanich 1971, 1994:112-115; Stephenson et al. 2002). The St. J ohns I culture thrived toward the south of this Deptford distribution, occupying most of the eastern half of the Florida peninsula from at least 500 BC to AD 750 (Milanic h 1994:244). These archaeologica l cultures are each defined primarily by differences in potter y. Deptford series pottery along the coast is tempered with sand or grit with various surface treatments of check stamping, linear check stamping, simple stamping, cordmarking, and at late sites, complicated stamping. In contrast St. Johns I pottery is mostly plain and occasionally incised, with a distinctive sponge spicule temper that gives vessel surfaces a chalky feel. In addition to these differences in pottery assemblages, the two 93

PAGE 94

archaeological cultures m ay be usefully disti nguished by variation in mortuary ceremonialism and subsistence regime. Atlantic Deptford populations are presumed to have built low sand burial mounds less than a meter high, although the paucity of diagnostic artif acts at mortuary sites of this time period make assignments of cultural affiliation so mewhat tenuous (Stephenson et al. 2002:327). Thomas and Larsen (1979) excavated nine mounds on St. Catherines Isla nd, Georgia, most of which are presumed to be Deptford. Some of these mounds contained up to a few hundred Refuge and Deptford series sherds while others produced virtually no pottery at all and few other artifacts. Nonetheless, a series of radiocar bon assays from these artif act-poor mounds confirms that most were constructed during a Deptford time period (Caldwell 19 70; cited in Milanich 1973:55; Thomas and Larsen 1979:138-143). Thomas and Larsens (1979) extensive fieldwork seems to demonstrate that Deptford mortuary regimes were consistently limited in terms of burial accoutrements. These material restrictions appear particularly austere when juxtaposed with other Middle Woodland burial treatments across the Eastern Woodlands that incorporated the largess of the Adena and Hopewell traditions Both primary extended and secondary bundled burials are present in Deptford mounds, usually with fewer than 25 individuals in each mound. The preponderance of female burials in the m ounds led Thomas and Larsen (1979) to propose a matrilineal society among the Atlantic Deptford. To the south, St. Johns I populations bu ilt low sand burial mounds around a meter high although a few are more than 3 meters high (Mil anich 1994:260). Most information about these mounds comes from C.B. Moores (1894, 1895) early excavations summarized by Goggin (1952). The preparation of bodies for interment may have been largely identical to Deptford practices, with both primary and secondary burials present. Mo st mounds contain less than 25 94

PAGE 95

individuals although a f ew mounds contain as many as one hundred. The distinguishing feature of the St. Johns burial regime compared to Dept ford is in the quantity and types of material culture found at mounds. While Deptford mound s can be characterized by the paucity of material culture, St. Johns I mounds contain arti facts in abundance. Local ly produced vessels of St. Johns Plain and Dunns Creek Red, a red pain ted ware most common in ceremonial contexts, are ubiquitous at mounds. Mounds also contain Deptford, Swift Creek, and Weeden Island vessels of foreign manufacture or locally-made copi es, and earthenware plan t and animal effigies (Milanich 1994:247). Extralocal influence is further evident in nonlocal mineral, metal, and stone (Seeman 1979), as well as occasional loglined tombs reminiscent of Hopewellian burial treatments (Bullen et al. 1967; LaFond 1983). In sum, St. Johns mortuary ritual was rich in material accoutrements and demonstrates pervasive connections with many contemporary Woodland period cultures. Differences between Deptford and St. Johns I cultures are also evident in subsistence regime, with the two cultures having been primarily adapted to distinct ecological zones. In fact, the extent of estuarine and freshwater ecosy stems along the Atlantic coast may have been influential in the formation of cultural boundaries There seems to be a strong correspondence between the distribution of Deptfo rd and St. Johns I sites and part icular ecological zones (Figure 4-1). Atlantic Deptford populati ons were adapted to the especial ly wide and extensive estuaries that back the sea islands of the southern South Ca rolina, Georgia, and northeastern Florida coast. Notably, these estuaries become much narrower a nd eventually intermittent further toward the south. Only the two northernmost islands in Flor ida (Amelia and Fort George) are true barrier islands with natural inlets and an extensive sa ltwater estuary behind them (Pilkey and Fraser 2002:70). With a more wave dominated environm ent toward the south, ba rriers are long, inlets 95

PAGE 96

are unstable, and tidal inf luence is limited, lead ing to lagoons behind barrie rs in eastern Florida that are essentially freshwater (Davis 1997:159). Indeed, before the dredging and blasting of channels to the ocean during the last centur y, freshwater ecosystems near the central and southern Florida coast were much more common. Perhaps as a partial consequence, Deptford sites are uncommon south of the St. Johns River mouth, where estuarine environments are more limited. Atlantic Deptford populations seem to have been overwhelmingly focused on the exploitation of oyster and saltwater fish while making only periodic special use camps for the procurement of upland resources (Espenshade et al. 1994). Alternatively, the St. Johns I culture was cente red on the exploitation of aquatic freshwater resources with the Upper and Middle St. Johns River as its heartland. Freshwater banded mystery snail (Viviparus georgianus), fish, and turtles make up the bulk of middens along rivers and lakes (Wing and McKean 1987). While there ar e also St. Johns I sites that clearly show extensive maritime adaptation (Russo et al. 1989), these estuarine enviro nments south of the mouth of the St. Johns River were smaller and more isolated than their co unterparts to the north in Georgia and northeastern Florida. As suc h, freshwater ecosystems were a subsistence mainstay for many St. Johns I populations. As am ong Deptford sites, no large St. Johns I sites have been located any distance away from major rivers, lakes, or coasta l estuaries. Smaller extractive camps exist in the piney flatlands, bottomland swamps, and cypress domes between the St. Johns River and Atlantic Ocean (Hardi n and Russo 1987; Hardin et al. 1988; SieglerEisenberg et al. 1985). The extensive estuarine habitats of the north and widespread freshw ater environments of the south meet dramatically at the mouth of the Lower St. Johns River in northeastern Florida. Here, the 310 mile river empties into the ocean to become part of a huge flooded estuary. This is 96

PAGE 97

also the area where W oodland period typological distinctions begin to break down, creating difficulties for archaeologists attempting to construct culture histories. Based on the published work of C.B. Moore (1894), John Goggin (1949, 1952:36) included the Lower St. Johns River in his definitions of St. Johns I and II cultures of the St. Johns River valley, largely ignoring the typological complexity of the northern fringes of the region. In his estimation, St. Johns I was a predominantly plain chalky ware period, wi th subperiods defined by minority wares of extralocal origin (Goggin 1952:47). However, S ears (1957) later stratig raphic excavations along the southern bank of the Lower St. Johns River failed to yield any Woodl and period sites with predominantly St. Johns Plain pottery. Instea d, Sears (1957) work produced assemblages dominated by sand-tempered plainwares with only minor occurrences of St. Johns Plain. Interpreting the Lower St. Johns as a boundary between the Georgia coast and Northern St. Johns culture areas, Sears (1957 :2) defined a locally-specific Woodland period chronology that began with Deptford, followed by a long sand-tempered plain p eriod that included minority occurrences of Swift Creek Complicated Stampe d wares, and ended with the sherd-tempered series known as Colorinda (Ashley 2003:67). Despite Sears (1957) attempt at reformul ating a chronology tailore d to the Lower St. Johns River area and the taxonomic difficulties en countered in other excavations (Bullen and Griffin 1952; Wilson 1965), Goggin s (1952) culture areas and ch ronology were subsequently repeated by Milanich and Fairbanks (1980) in their synt hesis of Florida arch aeology. As Ashley (2003:66-74) argues, Milanich a nd Fairbanks (1980) gloss of the Lower St. Johns River region was the chronological framework used for many cont ract archaeology projects that ensued in the following decades. Consequently, the Florida Master Site File is filled with sites from the Lower St. Johns River that are labeled St. Johns I even when only a few, or even just one, St. Johns 97

PAGE 98

Plain sherd was recovered am ong a large assembla ge of sand-tempered plain pottery (Ashley 2003:69). To this day only one site along the Lower St. Johns River, the Wood-Hopkins Midden (8DU9185), has produced a preponderance of St. Johns Plain pottery (Johnson 1994:52-69). This small freshwater snail (Viviparus georgianus) midden located just 15 km west of the Atlantic Ocean may represent a rare foray of Middle St. Johns groups into the area (Ashley 2003:72). The site is anomalous in terms of both pottery assemblage (all St. Johns Plain) and shell species (mostly Viviparus with some oyster), in an area wher e St. Johns Plain pottery is rare and middens are dominated by oyster. Michael Russo (1992) was the first after Sears (1957) to attempt to re fine the problematic chronology by proposing a separate culture area for the Lower St. Johns River locale: the St. Marys region. As Russo (1992) envisioned it, th e St. Marys region included the area along the Atlantic coast between the Lower St. Johns River a nd the Satilla River in s outheastern Georgia. Drawing primarily on extensive survey work from the Timucuan Historic and Ecological Preserve near the mouth of the St. Johns Rive r, Russo (1992; Russo et al. 1993) provided a new view on the complex chronology of the local Woodland period. Like Sears (1957), Russo emphasized that sand-tempered plain pottery was often dominant on Woodland period sites, with Deptford, Swift Creek, St. Johns, and Colorinda al so occurring in varyin g amounts. Due to the multicomponent nature of the sites he excav ated, Russo (1992) was unable to connect the ubiquitous sand-tempered plainwares to a specif ic archaeological culture or develop a Woodland period chronology with the few temporally di agnostic pottery types. In the decade that followed Russos (1992) work, a healthy database of Woodland period radiocarbon dates was developed. Ashley (2003 :81) compiled these radiocarbon assays and found that they largely substa ntiated Sears (1957) original attempt at a chronology through 98

PAGE 99

seriation of m idden pottery. As Ashley (2003:74) observes, Dept ford series pottery began the Woodland period chronology and was produced from roughly 500 BC to AD 1 (Kirkland and Johnson 2000). Following Deptford, a 500 year pe riod ensued in which sand-tempered plain pottery dominated assemblages al ong with minor occurrences of check stamped and complicated stamped (Early Swift Creek) surf ace treatments. St. Johns Plai n pottery was made locally in limited amounts or brought in from the Middle St Johns River. Charcoal-tempered plain and complicated stamped pottery was produced at this time as well, from roughly AD 300 to 500. After AD 500 to roughly AD 850, sand-tempered potte ry continued to dominate assemblages but Late Swift Creek Complicated Stamped became more prevalent than any previous minority ware. Finally, Colorinda pottery concluded the Woodland period chr onology, produced for a short time between AD 850 and 900 (Ashley 2003:74-75). This most recent chronology provides the basis for a more detailed discussion of typology and chronology of Swift Creek contexts on the Atlantic coast. Swift Creek Radiocarbon Chronology, Typology, and Culture History While Swift Creek Complicated Stamped pottery was produced for almost a millennium across the southeastern United States, the specif ic temporal range of its production varied considerably across the region (Stephenson et al. 2002; Williams and Elliott 1998). This is true along the Atlantic coast itself, where complicated stamping appear s on the Lower St. Johns River several centuries earlier than along the estuaries of southeastern Georgia. As with many areas of the Southeast, Swift Creek pottery in northeastern Florida can be partitione d into Early and Late varieties based on attributes such as rim and lip form as well as quality of design execution and application (e.g., Kelly 1938; Kelly and Smith 1975; Phelps 1969; Willey 1949). In southeastern Georgia, Deptford pottery persisted through Early Swift Creek occupations of northeastern Florida, and only the Late variet y of Swift Creek pottery is fo und at coastal Georgia sites. 99

PAGE 100

Com plicated stamping continued well into the 9th century in both sout heastern Georgia and northeastern Florida. Along the Atlantic coast, a to tal of 43 radiocarbon assays have been recorded from 13 sites (Table 4-1, 4-2). This series of dates confirms that Early Swift Creek pottery is restricted to northeastern Florida while Late Swift Creek pottery is contemporaneous along much of the Florida and Georgia coasts. Early Swift Creek contexts in northeaster n Florida are represented by six radiocarbon dates from three sites (Table 4-1). Four of th ese are AMS assays derived from soot samples on Early Swift Creek charcoal-tempe red plain and complicated stamped vessels from the Dent and Mayport mounds. Another comes from soot on a complicated stamped sponge-spicule-tempered vessel from the Tilley Fowler (817245) villag e site (Hendryx and Wallis 2007). Finally, one date comes from a charcoal sample from a fire pit 60 cm below the ground surface at the Mayport mound (Wilson 1965:31). Calibration of the entire series of assays indicates a conservative time span of circa AD 30 to 600 for Early Swift Creek pottery. However, both the earliest AMS assay from the Dent mound and the ra diocarbon date from ch arcoal at the Mayport mound should be viewed with caution. As an anomalous outlier among AMS dates from soot on charcoal-tempered pottery, the Dent assay may be a result of the old-wood problem, in which the wood used as fuel was much older than the cook ing activity that caused s oot to adhere to the vessel surface (Ashley and Wallis 2006:8). Altern atively, the cultural affiliation of the Mayport charcoal sample is uncertain because Wilson (1965) did not provide a record of material culture from this feature, although th is assay may well relate to the earliest stages of mound construction. Although more assays from Early Swift Creek contexts are needed, the current suite supports a conservative calibra ted date range of AD 200 to 600. 100

PAGE 101

Presently, 37 Late Swift Creek radiometric da tes have been recorded from 13 sites along the Atlantic coast between the m outh of the Altamaha River and the mouth of the St Johns River (Table 4-1, 4-2). This series of dates indicates that Late Swift Creek contexts have a time span of roughly AD 500 to 850, with the 7th and 8th centuries particularly we ll represented. Nearly half of the assays come from the Kings Bay s ite (9CM171) in southern Camden County that itself spans the entire temporal range of Late Swift Creek occupation along the coast (Adams 1985:358-359). The 6th century was a time of transition in pottery production, indicated by the overlap in the temporal ranges of Early and Late Swift Creek radiometri c dates in northeastern Florida. Along the Lower St. Johns River, ch arcoal tempering and early notched rim forms appear to have gradually diminished to give way to the ubiquity of sand tempering and late folded rims. In southeastern Georgia, Late Swift Creek pottery production began during the 6th century but enjoyed its greatest popularity in the following two cen turies. While Early and Late Swift Creek manifestations clearly overlap acco rding to radiometric assays, for convenience I have divided the transitional 6th century in half for my outline of Swift Creek culture history along the Atlantic coast. By circa AD 850, a shift in pottery produc tion occurred and was accompanied by other cultural changes. These wa ning Late Swift Creek ma nifestations, labeled the Kelvin culture by Cook (1979), are represen ted by two radiometric dates from extreme northeastern Florida and possibly one more from the Evelyn site on the Altamaha River (Ashley et al. 2007; Hendryx 2004). Early Swift Creek (ca. A.D. 200 to 550) In northeastern Florida, Early Swift Creek assemblages consist of locally produced plain and complicated stamped wares, with rim treatme nts that include simple rounded and flattened lips as well as hallmark forms such as notch ed, nicked, scalloped, a nd crenulated (Ashley 1992:130-131, 1998:204; Sears 1959:155). Although sand tempering seems to predominate at 101

PAGE 102

m any Early Swift Creek sites, early vessels are often charcoal-tempered. Charcoal-tempered pottery contains quartz sand and ch arcoal inclusions that range from very common to sparse. While the production of charcoal-tempered pottery seems to have been mainly restricted to sites near the mouth of Lower St. Johns River, it ha s been recorded in lim ited quantities as far upstream as the River Point site (8SJ4790) on the St. Johns, as far south along the coast as the Shannon Road Midden site (8SJ3169), and as far nor th as the Sadlers Land ing site (9CM233) in Camden County, Georgia, a straight line distance of about 90 km (Ashley and Wallis 2006:8). Charcoal tempering seems to have been a loca l, spatially circumscri bed tradition among groups making Early Swift Creek pottery. At both mort uary mounds and trash middens plain charcoaltempered sherds outnumber complicated stamped versions, but sand-tempered plain represents the dominant type in most Early Swift Cr eek assemblages (Ashley 1998:200; Russo 1992:115; Sears 1957:29). The technological tradition of charcoal tempering has been interpreted through the results of petrographic analysis by Ann Cordell and Lee Newsom (Wallis et al. 2005). In general, most extant charcoal particles in the paste tend to exhibit little shrinkage from vessel firing and cooking, indicating that predominantly charcoal, rather than wood, was added as temper. On both the interior and exterior su rfaces of vessels, charcoal particles often appear to have been burned to ash, leaving holes or voids (Ashley 19 98:202; Russo et al. 1993:35-36). These are the characteristic holes that S ears (1957, 1959) and Wilson (1965) identified to designate holetempered and limestone-tempered pottery, re spectively. Petrographic analysis reveals the prevalence of charcoal, crushed bone, and grog inclusions in the paste of many vessels, supporting the possibility that thes e materials were hearth contents added to the cl ay during paste preparation. In addition, the consis tent size of temper particles, ranging from medium to coarse 102

PAGE 103

on the W entworth scale, suggest that charcoal was consistently pounded and sieved before being added to the raw clay. Designs on Early Swift Creek pottery tend to be sloppy in execution, differentiating them from Late Swift Creek varieties (Ashley and Wallis 2006:8). Paddle impressions typically cover the entire exterior surface of vessels, which are predominantly sub-conical pots and bowls. The apparent prevalence of overstamping and smoothing has generally hindered attempts to reconstruct Early Swift Creek desi gns, although they are presumed to be similar to those of the Gulf Coast of Florida (Ashley and Wallis 2006:13). One exception is the diamond and raised dot design characteristic of the type Sun City Comp licated Stamped, which has been identified at Swift Creek sites in the Florida panhandle as well as on charcoal-tempered vessels in northeastern Florida (Hendryx and Wallis 2007; Thomas and Campbell 1993:571; Willey 1949:437). Late Swift Creek (ca. A.D. 550 to 850) Late Swift Creek pottery has a much broade r distribution than th e early variety, being prevalent at sites along the Atlantic seaboard from just south of the St. Johns River to the north side of the Altamaha River (Figure 4-2) (Ash ley et al. 2007; Ashley and Wallis 2006:9). Rim treatments include hallmark thick rim folds as we ll as simple rounded or flattened lips. Temper among Late Swift Creek assemblages is typically sand, with vessels from the Georgia coast and northern-most Florida coastal is lands (e.g., Amelia Island, Martin Island) having coarser (i.e. larger) quartz temper than Lower St. Johns Rive r vessels (Ashley and Wallis 2006:9). Two other temper variants have also been noted with some frequency. First, bone tempering, which typically consists of bone crushed into fine particles, has been observed in limited amounts among classic sand-tempered Late Swift Creek a ssemblages. Second, grog tempering, almost always crushed into very fine particles, occurs w ithin some Late Swift Creek vessels as well. In 103

PAGE 104

fact, grog tempering is characte ristic of the waning Late Swif t Creek tradition that Cook (1979) labels Kelvin (Ashley an d Wallis 2006:9). Thus grog tempering may be a temporal marker in the latest Swift Creek sherds, its use signifying a precursor to the grog particles of the terminal Late Woodland cultures of Colorinda and Kelvin (Ashley and Wallis 2006:9). Sand-tempered plain continues to dominate assemblages throug hout Late Swift Creek and waning Late Swift Creek contexts. Designs on Late Swift Creek complicated stamped vessels appear more carefully executed than on Early Swift Creek vessels. Co mplete stamping all over the vessel continued but zoned stamping also became popular. Zoned stamping is also coincident with new vessel forms, particularly collared jars and bowl s that are not present among Early Swift Creek assemblages. At the same time, sub-conical pots and bowls continue to be common. Paddle matches and design similarities on Late Swift Creek pottery along the Atlan tic coast demonstrate cultural connections among populations of north eastern Florida and southeastern Georgia (Ashley and Wallis:12-14). By the end of 9th century carved paddle designs became more simplistic and paddle impressions were more s loppy in execution. Nested chevron designs, often typed Crooked River Complicated Stampe d, became a persistent surface decoration on vessels, dominating some assemblages. Rim folds became uncommon and were replaced by an incised line or false fold below the lip. Thes e characteristics define the waning Late Swift Creek or Kelvin pottery assemblages that seem to have continued into the 10th century. Swift Creek Site Types Swift Creek pottery is common in small amounts at many sites on the Atlantic coast (Ashley 1992; Ashley et al. 2007; Russo 1992; Russo et al. 1993). The widespread occurrence of a few complicated stamped sherds in assemblages dominated by other types is perhaps testimony to the breadth of social interactions mediated through Swift Creek material culture. 104

PAGE 105

However, Swift Creek sites, defined by the predom inance of complicated stamping in assemblages (usually following sand-tempered plai n), essentially consist of three types: small artifact scatters, large shell middens, and low sa nd mortuary mounds. Th e distribution of these types of archaeological sites a nd their relationship to one an other is variable across the landscape. Along the Lower St. Johns River in northeastern Florida, a series of mounded cemeteries stand segregated from the multi-house hold villages represented by extensive shell middens (Wallis 2008). In cont rast, mortuary mounds are sc arce along the Georgia coast, perhaps as the result of both site destructi on and different burial pr actices (Ashley et al. 2007:22). At sites where mounds are present or likely to have been present in coastal Georgia, villages were placed nearby. Therefore, even wi th the sampling bias that may pertain, the built landscapes among local Swift Creek populations were different al ong the Lower St. Johns River and the Georgia coast. Village Middens It is difficult to infer the structure of Sw ift Creek settlement from complicated stamped pottery found in hundreds of sm all artifact scatters and many dense multi-component sites (Ashley 1998; Ashley et al. 2007). However, large single-component si tes seem to correspond with multi-household villages and many reveal pa tterns in community structure. A conspicuous feature of many shell middens along the coast is their circular or semi-circular arrangement, presumably corresponding with the circular shap e of villages, although other configurations are also noted. The failure to di scover clear architectural featur es at arcuate arrangements of middens may indicate that houses were located adjacent to middens, perhaps inside the shell ring. These areas, which tend to be mostly devo id of artifacts, have received very limited excavation coverage. At the same time, the lack of shell in these areas makes preservation of cultural features very unlikely. The shallow nature of many of these shell middens suggests 105

PAGE 106

seasonal rather than extended year-round site oc cupancy. T hus, Swift Creek populations on the coast were probably at least seasonally mobile. Subsistence data along the coast indicates a diet overwhelmingly dominated by maritime resources, es pecially fish and she llfish. Further inland, beyond the saltwater habitats of oysters and clams, poor faunal preservation has resulted in a paucity of subsistence data. Below is a brief re view of the largest Swif t Creek middens along the coast that are interprete d as village sites. Northeastern Florida Work in the last two decades has grea tly expanded our knowledge of Swift Creek habitation in northeastern Fl orida. Presently, there are at least five sites that represent substantial villages from the east bank of the St. Johns Ri ver near downtown Jacks onville to Amelia Island near the Georgia border. These sites represent a range of ecological se ttings, adaptations, and village structure. On the eastern bank of the St. Johns Rive r across from Jacksonville, numerous Swift Creek deposits have been identified (Ash ley and Hendryx 2008; Hendryx and Wallis 2007). Small amounts of Swift Creek Complicated Stampe d pottery are widely scattered at sites along the waters edge and a hundred or more meters inland on the sand ridges that flank the eastern and southern banks of the river (Ashley 1998:215). These sites are commonly quite small, but a significant Swift Creek component wa s recently excavated at the T illie Fowler site (Hendryx and Wallis 2007). The northern portion of the Tillie Fowler site (8DU17245) was comprised of two distinct loci of artifact concentration primarily dating to the Early Swift Creek phase. Along a portion of the river too brackish for freshwater snails and too fresh fo r saltwater oysters and clams, the site contained almost no shellfish, whic h resulted in very poor preservation of artifacts and features. Even with the paucity of non-ceram ic artifacts and cultural f eatures, the density of the pottery deposits, and their spat ial distribution covering nearly one hectare is very likely to 106

PAGE 107

correspond with a village. As elsew here along th e river, Swift Creek pot tery has also been recovered from contexts closer to the waters edge such as at the nearby Dolphin Reef Site (8DU276), where more than 200 Swift Creek and many presumably contemporaneous sandtempered plain sherds were documented (Ashley and Hendryx 2008). Moving east from Jacksonville along the St. Johns River toward the Atlantic Ocean, there are many sites that indicate intensive Swift Cr eek occupation. At least four multi-component sites (8DU62, 8DU66, 8DU5602, 8DU5611) with appr eciable quantities of Swift Creek pottery have been tested in the vicinity of Mill C ove (Ashley 1998; Johnson 1988) and near St. Johns bluff (Sears 1957, 1959). The limited nature of tes ting at these sites frus trates attempts to interpret their Woodland period components (A shley 1998:216). However, three single component sites further to the east, on the Greenfield peninsula, provide a compelling view of Swift Creek occupation. The Greenfield Peninsula is bordered by salt ma rshes that extend from the St. Johns River to the north, Greenfield Creek to the west, and San Pablo Creek (Intracoastal waterway) to the east. Situated between 50 and 200 m away from th e present western shoreline, Greenfield site 7 (8DU5543) consisted of three hectares of sparse midden debris interspersed with several small areas of dense shell midden (Ashley 1998; Florid a Archaeological Services 1994). Of the five areas of dense shell identified, three were intens ively investigated and subsequently interpreted as household refuse middens (Florida Arch aeological Services 1994: 121). These middens contained solidly packed, homogeneous oyster sh ell matrix averaging 10 to 15 cm thick and with abundant vertebrate faunal remains and pot tery. Two adjacent middens, measuring 2 x 7 m and 6 x 12 m in plan view, contained exclusivel y Early Swift Creek Plain and Early Swift Creek Complicated Stamped pottery. The third and sm allest (3 x 5 m) midden investigated, roughly 107

PAGE 108

120 m southeast of the others, contained only La te Swift Creek Plain and Late Swift Creek Complicated Stamped pottery. Combined, th ese three middens produced approximately 1000 sherds, which were equally divided between complicated stamped and plain varieties. Notably, Early Swift Creek and Late Swift Creek contexts were discrete at the site, with no mixing of diagnostic wares in any of the excavated features or shell middens. Although only five shell midden concentrations were identified, and only three of these were intensively investigated, the site clearly represents two chr onologically distinct occupations that were both substantial. The form of these occupations may have been either solitary houses or multi-household villages, but is more likely th e latter. As the authors note, it is highly probable that undemonstrated refuse heaps ex ist within the site boundaries because these features were smaller than the shovel testing in terval used across much of the site (Florida Archeological Services 1994:121). The spatial orientations of th ese villages, however, cannot be determined from available evidence. A clearly circular Late Swift Creek village was located nearby. Roughly in the middle of the Greenfield peninsula, the Swift Creek Middens Area at Greenfield site #8/9 (8DU5544/5545) was a non-mounded horseshoe-sha ped shell midden approximately 50 m in diameter with two interior household middens ri fe with fauna and sherds (Smith and Handley 2002:49; Wallis 2007). Although fe w posts and no definitive evidence of structures were identified, the spatial configuration of midden deposits are consistent with many other Swift Creek ring middens interpreted as circular villages (Stephenson et al. 2002:346). Three radiocarbon assays put occupation of the site solidly in the 7th century and possibly the 8th. The pottery assemblage consisted almost exclusively of sand-te mpered plain and complicated stamped sherds, however, microscopic analysis id entified charcoal tempering in a few sherds 108

PAGE 109

(W allis 2007). Because the charcoal-tempered sh erds were recovered from various depths throughout the dense shell midden they are likely to be contemporaneous with the Late Swift Creek occupation of the site and represent the last waning years of th e tradition of charcoal tempering. The Swift Creek Middens Area represents a discrete Late Swift Creek village, perhaps the best example of one in northeastern Florid a. Another Late Swift Creek ring midden was recorded on the same site (8DU5544/5545) just 175m to the east (Johnson 1998). This roughly arcuate midden (Midden A) is notable not only for its large size, at 120 m across its long axis, but also for its probable plaza cemete ry. Five test units were excavated in the center of the arcshaped site with human remain s and Swift Creek pottery encountered each time. This area was labeled Mound C, described as a low, barely perceptible 30 cm rise with an oval shape cradled by Midden A (Johnson 1998:71). However, the recorded relative elevations of the surface of Midden A and Mound C reveal that the apex of the mound stood only 10 cm above the highest portions of the midden (Johnson 1998:22, 64). Thus, Mound C may not have been a mound, per se, but rather a plaza cemetery akin to the Bernath Place Swift Creek village on the Gulf Coast, where dozens of individuals were recorded (Bense 1998). Whether mounded or not, the cemetery-village complex at Greenfield #8/9 is unique in a region where solitary mortuary mounds were more common during Swift Cr eek times (Ashley and Wallis 2006). To the north of the Greenfield peninsula, multicomponent sites at Cedar Point (8DU81), Crane Island (8NA709), and Amelia Island (McArt hur Estates, 8NA32) indicate substantial Swift Creek occupations but little information about village structure ca n be discerned (Ashley 2003; Dickinson and Wayne 1999; Handley et al. 2004). At McArt hur Estates (8NA32) numerous postholes and widespr ead pit features filled with domestic refuse likely correspond 109

PAGE 110

with a Late Swift Creek village, yet no spatial patterns were appa rent. Radiocarbon assays from Swift Creek pit features date the site to the 7th and 8th centuries. Another Amelia Island site s hows more clearly an approximate ly circular village layout. Ocean Reach (8NA782) was a Late Swift Creek s ite consisting of 22 discrete shell middens (Johnson et al. 1997). These middens, averagi ng between 3 m and 4 m across and uniformly between 10 and 20 cm in thickness, were interp reted as individual househ old refuse deposits. Three clusters of individual middens were appa rent at the site, with each forming roughly the corner of a triangle about 80 m on a side. An im portant distinction at Ocean Reach was that the shell middens were comprised primarily of quahog clam ( Mercenaria spp. ), as opposed to oyster ( Crassostrea virginica ) that dominates most Swift Creek shell middens. Southeastern Georgia Large Swift Creek components in coastal Georgi a are mainly confined to southern areas, between the St. Marys River and just north of the Altamaha River. While there are more than 60 sites that contain Swift Creek pot tery along the coast (Ashley et al 2007:7), at present seven sites that have been investigated are likel y to represent village sites. In Camden County, Georgia, three Swift Creek sites correspo nd with substantial habitation sites. Two large village complexes at the Kings Bay locality were excavated by the University of Florida during the 1970s and 1980s (Adams 1985; Smith 1978, 1986; Smith et al. 1981; Smith et al. 1985). The first was the Kings Bay site (9CM171), which included a series of discrete shell middens arranged in an arc on a bluff overlooking th e salt marsh. While this area of the site was interpreted as the location of a village with individual household middens, evidence of a structure and hearth s/trash pits at two adjacent local es were interpre ted as special processing areas (Saunders 1986:17-18). Season ality data indicated occupations during all seasons of the year, perhaps denoting year -round residency (Quitmyer 1985:89; Reitz and 110

PAGE 111

Quitm yer 1987). The Mallard Creek site (9CM 185), located 1.5 km inland from Kings Bay (9CM171), exhibited a settlement structure entire ly different than the typical arcuate-shaped arrangement of middens. Instead, the site consiste d of a single centralized refuse disposal area with domestic activities apparently confined to its periphery (Sm ith et al. 1985:140). An analysis of quahog clam growth rings points to a spri ng season of occupation (Saunders 1998:163). The Sadlers Landing site (9CM233), located 8 km north of the Kings Bay site, included numerous discrete oyster shell middens but thei r spatial arrangement was not clearly arcuate (Kirkland 2003:122). In one area of the site partial, multiple, and primary burials were also encountered along with red ochre, stone celts, de er antler fragments, a bear tooth, and mica sheets. While no evidence of a mound was encountered during excavat ion, there is a strong possibility that one existed at the site but had been razed by plowing (Kirkland 2003:152). The dense shell middens recorded nearby may be eviden ce of a village directly adjacent to the site of the former mound. Based on the field map, discrete clusters of shell midden also were recorded between approximately 150 m and 600 m from the human remains (Kirkland 2003:122). Two AMS assays date the site between AD 600 and 800, firmly within the Late Swift Creek phase. To the north, significant Sw ift Creek sites exist at Cumberland Harbour Shell Midden (9CM249) on the northeast end of the Point Peter Peninsula as well as Pi kes Bluff Shell Midden (9GN200) on the west side of St. Simons Island but reports of excavations at these sites have yet to be released (Ashley et al. 2007:24). Further to the north along the Altamaha River and its tributaries, numerous important sites have been investigate d, most notably Evelyn (9GN6), Cathead Creek (9MC360), Sidon (9MC372), and Lewis Creek (9MC16). The Evelyn site includes seven mounds across about 200 acres overlooking the tidal marsh on the southern side of the Altamaha Ri ver. A Swift Creek shell midden covering more 111

PAGE 112

than an acre is situated just north of mounds B, a platform mound, and C, a conical burial mound, both of which have been demonstrated to be Swift Creek (Ashley et al. 2007:14-15; Chance 1974; Waring and Holder 1968:140). Surface collecting and very limited subsurface testing have revealed that the midden is Late Swift Creek with a probabl e Kelvin component, and Ashley and colleagues (2007:14) su ggest that further investiga tion may reveal an arc-shape habitation area. With further work, the Evelyn site may reveal a substantial Swift Creek village directly adjacent to both a conical burial m ound and platform mound. Roughly across the river from Evelyn at the confluence of Cathead Creek and the Darien River is the Cathead Creek site (9MC360). This site is known solely through mitigation efforts that were focused on narrow utility corridors. This spatial restriction and the multi-component nature of the site made delineations of site structure difficult, yet concentrations of Swift Creek pottery suggested the po ssibility of an arc-lik e settlement pattern (Wa yne 1987:59). The Swift Creek components of the site included both shallo w and deep pits as well as shell middens. One calibrated radiocarbon age dates Swift Cr eek occupation of th e site to the 6th or 7th century. A face-down burial was also discovered at the site but due to its radiocarbon age (10th century) and burial treatment it is likely to be associated w ith Kelvin (e.g., terminal or waning Swift Creek; Ashley et al. 2007:11). Zooarch aeological data from the Swift Creek component indicate that the site may have been occupied on a year-round basis (Reitz and Quitmyer 1987). Located 3 km to the northwest of Cathead Creek, the Sidon Plantation site (9MC372) was excavated by Fred Cook (1995) during construc tion of a shopping mall. His salvage work documented a series of Swift Creek pit features, including a shallow Swift Creek phase house pit that contained a flexed human intermen t (Cook 1995:8-10). Another burial was located nearby, consisting of disarticulated and cremat ed remains from several individuals. These 112

PAGE 113

interm ents that seem to be associated with houses rather than mounds ma y be the best evidence for an entirely different mortuary tradition in Swift Creek contexts. One AMS date from human bone indicates that occupation of the site was in the 7th century. Located less than 7 km west of Sidon Planta tion along a tributary of the Altamaha River, the Lewis Creek site (9MC16) c onsists of a shell midden along th e waters edge and five burial mounds further inland. Only one of the mounds ha s been investigated an d dates to the Savannah II phase, however, the cultural affiliation of th e other four mounds remains unclear (Cook 1966). The shell midden itself has never been excavated, but has been extensively surface collected over the years as Swift Creek pottery continually erode d onto the bluff facing the creek (Ashley et al. 2007:11). Little has been determined about the spa tial structure of the site but a recent visit to the site in 2008 confirmed that shell midden deposits are extensive. Mortuary Mounds There are two or more Swift Creek mortuary traditions represented on the Atlantic coast. Along the Lower St. Johns River, low mounds mostly less than a meter in height were raised over a period of centuries as human remains, artifact s, and earth were repeatedly added. At least fifteen of these mortuary mounds have been re corded, and all but one appear to have been located some distance away from village sites (Ashley and Wallis 2006). Alternatively, Swift Creek mounds are rare to the north of the St. Johns River along the Georgia coast, undoubtedly in part due to site destruction in the absen ce of recorded investig ations. As mentioned previously, a mound probably once existed at Sa dlers Landing (9CM233), and other Swift Creek mounds have been noted along Turtle River in Glynn County and on St. Simons Island (Ashley et al. 2007:24). While Swift Creek mortuary mounds certainly existed in southeastern Georgia, another mode of burial, such as underneath houses like at Sidon (9MC372), may also have been more common than current evidence suggests. 113

PAGE 114

The mounds at Evelyn (9GN6) near the southern bank of the Altamaha River may represent still another form of mortuary ceremoni alism. The Evelyn site itself is anomalous for several reasons. First, it is the only site along the coast that contains multiple Swift Creek mounds, with at least two and possibly more pres umably dating to the Late Swift Creek phase (Ashley et al. 2007:13). Second, th e site contains the only possi ble Swift Creek platform mound (Mound B) along the coast, consisti ng of several distinct layers of cultural strata with Swift Creek sherds and intermittent burials (Waring a nd Holder 1968). If Mound B does indeed have a Swift Creek affiliation, its construction and use may have been similar to other Woodland period platform mounds in the Southeast (Ashley et al 2007:13; Jefferies 1994). Third, the Evelyn site provides convincing evidence of a substantial vi llage site adjacent to the two Swift Creek mounds, a close proximity between mound and vill age that is not common to the south on the Lower St. Johns River. Fourth, Mound C, at 3 m high in recent records of elevation (Ashley et al. 2007), is considerably taller than the major ity of Swift Creek mounds on the Lower St. Johns River. Stratigraphy in the mound revealed that it was built in four or five stages (Chance 1974; Waring and Holder 1968:140), which stands in contrast to the lo w mounds of the St. Johns that mostly lack discrete internal stratigraphy that would be form ed by major earthmoving events. Finally, Mound C contained artif acts such as quartz crystal and small bar gorgets that are uncommon on the coast but have been recovere d at Swift Creek mounds much further inland (Ashley et al. 2007:14). In fact, Anderson (1998) suggests that site s such as Evelyn were located along important exchange routes to the interior and were places of public consumption and competitive display of wealth between individu als and lineages. In all, the ceremonialism evidenced at Evelyn may have been more similar to that of central Georgia populations than to populations to the south along the St. Johns River. 114

PAGE 115

Fifteen Woodland period mounds along the Lo wer St. Johns River have been attributed to the Swift Creek archaeological culture (Ash ley and Wallis 2006:11). Twelve of these are known primarily through Moores (1894, 1895, 1896) early work, with the remaining three excavated in later years ha ving somehow escaped Moores detection (Ashley 1995; Johnson 1998; Wilson 1965). Most mounds along the river tend ed to be sited on blu ff edges not directly on the water, although they very well may have comm anded a view of the wate r. In general, the mounds yielded mainly complicated stamped and undecorated vessels, along with fewer St. Johns Plain, Dunns Creek Red, and Weeden Island series wares. Each of these minority wares appears to have been more common in mounds than in coeval middens. Other common items in the mounds included shell beads, modified and unmodified marine shell, pebble hammerstones, unmodified pebbles, hafted bifaces and chert flakes, celts, hematite (ocher), sandstone, mica sheets, and various copper artif acts (Ashley 1998:212). While some of these objects are similar to those exchanged throughout Hopewell intera ction networks of the Eastern Woodlands (Seeman 1979), the majority of extralocal artifacts at Swift Creek mounds were acquired after the heyday of Hopewell interactions (Ashley and Wallis 2006:12). Based on the osteological populations and the context of mound artifacts, often placed throughout the mound rather than with particular individuals, Thunen and Ashley ( 1995) inferred an essentially egalitarian social structure. Certainly the oste ntatious burial treatments record ed in other Hopewell-related contexts in the Southeast and Midwest are mos tly lacking on the Lower St. Johns (e.g., Brose and Greber 1979; Mainfort 1988). Judging from Moores (1894, 1895, 1896) descripti ons, there appears to have been some variation in burial regimes and quantities of objects placed within mounds. While there is disparity in the height of m ounds, ranging from .6m to 2.1m (Ashley 1998:209), the most notable 115

PAGE 116

disparity is in the quantities of hum an re mains and objects (mostly pottery) that mounds contained. In some mounds, bur ials were abundant and objects we re deposited by themselves or in caches, often with ochre. It is estimated that the Dent Mound contained more than 100 human interments, with males, females, and child ren represented (Ashley 1995). Excavation of approximately 30% to 40% of the Mayport Mound yi elded more than 50 individuals, both adults and children (Wilson 1965:12-13). At these sites, pottery was ubiquitous, including many scores of whole vessels, intentionally br oken or basally-perforated vesse ls, and sherds numbering in the thousands. Moore (1894, 1895) seems to describe similar evidence at Arlington (8DU33), Point LaVista C (8DU40), Monroe (8DU13), Low Gr ant A (8DU15), and Low Grant E (8DU19) mounds. At most of these mounds Moore notes that human remains were encountered at points, presumably indicating at least 25 interr ed individuals. These mounds also contained appreciable amounts of earthenwa re vessels and sherds. Moor e (1895:490) describes Grant Mound E as literally filled w ith earthenwarewhole vessels, fr agmentary ones, and sherds. Grant Mound E actually contained few human burials but this discrepancy may have been due to extremely poor preservation, which Moore (1895: 489) himself acknowledges. All of the mounds rich in artifacts and human remains appear to be the continuous-use type of mound (cf. Sears 1958:277) in which the mound grew by accretion as artifacts and human remains were added over a period of centuries (Ashley 1998:213-214; Thunen and Ashley 1995:5). I propose that among Moores descriptions ther e may have been another type of mound that was constructed in fewer stages and over a mu ch shorter period of time. At mounds such as South Jacksonville B (8DU36), Point la Vi sta A (8DU38) and B (8DU39), and Johnson (8DU10), human remains and pottery were comp aratively limited. While human remains at these sites may have escaped detection due to poor preservation, the disp arity in the abundance 116

PAGE 117

of pottery is striking. At thes e sites it appears th ere were no opportunitie s to incorporate new offerings over tim e, resulting in far fewer fragments of [vessels that were ] many feet apart, as though strewn upon the mound in course of c onstruction than at other mounds (Moore 1894:199). Instead, these mounds may have been erected quickly and material was not continually added in subsequent years. Among the clearly accretionary mo unds there is some variati on as well. At the Mayport mound, extended burials predominated while at the Dent mound large aggregations of disarticulated remains were more common (A shley 1995; Wilson 1965). These discrepancies may indicate very different patterns of group mobility or simply di fferent mortuary regimes. As Milanich and Fairbanks (1980: 160) suggest, masses of bones of many individuals interred together may indicate that remains were processed in a charnel facility before interment. Both mounds contained Early and Late Swift Creek di agnostic pottery and a ppear to have been continuously used for centuries. In addition, no v illage site has been located in proximity to either of these mounds. Summary of the Cultural Landscape The cultural landscape of the Atlantic coast ca n be divided into somewhat of a northern and southern dichotomy. Along the Lower St. Johns River, many accretionary mounds were initiated during the Early Swift Creek phase and c ontinued to be used for centuries until the end of the Late Swift Creek phase. A few others, as noted, may have had shorter use lives. In all but one solitary case (Greenfield #8/9 ), villages on the Lower St. J ohns River were situated some distance away from mounds. While Late Swift Creek populations continued to use many of the same mortuary mounds initiated by Early Swift Creek groups, La te Swift Creek village sites rarely overlap with Early Swift Creek occupations. Early Swift Creek interactions seem to have 117

PAGE 118

been f ocused toward the Gulf Coast of Florida, whereas Late Swift Creek interactions were directed toward the Georgia coast as well. On the Georgia coast there is little ev idence of Swift Creek mounds, with Evelyn a confirmed exception, and possibili ties at the unexcavated mounds at Lewis Creek, and the mortuary components at Cathead Creek and Sadlers Landing that ma y have been razed prior to excavation (Ashley et al. 2007:22) While site destruction ma y partially explain regional differences, other modes of burial are also appare nt, as demonstrated at the Sidon site (9MC372). At Evelyn (9GN6), the comparatively large conical burial mound and broad platform mound may show more cultural similariti es to Swift Creek sites in cent ral Georgia than Lower St. Johns mounds. In addition, the ostensible village midden at Evelyn is directly ad jacent to the mortuary mound(s), a close proximity not recorded amon g the mounds and villages in northeastern Florida. Paddle Matches Even in the midst of these apparent differen ces during the Late Swif t Creek phase, there is definitive evidence of social interaction in the form of paddle matches identified from complicated stamped designs. Paddle matches confirmed by sherd to sherd comparison link sites in coastal Georgia to sites in south-central Ge orgia and northeastern Fl orida. Paddle matches link sites within these regions as well. There are presently nine paddle matches linking 13 sites in northeastern Florida and southeastern Georgi a (Figure 4-3) (Ashley et al. 2007; Ashley and Wallis 2006). This tally does not include several designs that link solita ry coastal sites with inland sites (Ashley et al. 2007). Some of these paddle matches come from complete vessels or large portions of vessels and confirm variation in both the shape of matching vessels and orientation of paddle impressions. 118

PAGE 119

There are three designs linking nor theastern Florida and southeastern Georgia sites. Using Frankie Snows num erical designations, Desi gn 34 connects the Dent (8DU68) and Grant (8DU14) mounds on the Lower St. Johns River with the Sidon site (9MC372) near the Altamaha River, about 115 km away. Based on illustrations, this design is likely present at Kings Bay (9CM171) as well. Interestingly, at least two di stinct vessel forms are represented, an open pot from Sidon and a collared jar from Grant (Figure 4-4). The orientation of the paddle stamping is identical, however. Design 36 links the Dent Mound (8DU68) with three sites near the Altamaha (9MC16, 9MC360, 9MC372) as well as Kings Bay (9CM171) and Crane Island (8NA709) just west of Amelia Island (Ashley and Wallis 2006). Vessel forms are very similar, consisting of restricted pots or bowls, but the size of the vessels vary dramatically with orifices ranging from 17 cm (9MC16) to 30 cm (8DU68). What is more, pa ddle orientation is unique on the Dent mound vessel compared to the specimens from sites near the Altamaha River. While paddle impressions on most vessels were made with the long axis of the paddle perpendicular to the rim, the Dent Mound vessel has impressions nearly para llel to the rim (Figure 4-5). Design 38 connects the Mayport Mound (8DU96) and the same three sites along the Altamaha River (9MC16, 9MC360, 9MC372), as we ll as the Shadmans Field (9GN271) site on St. Simons Island (Ashley and Wallis 2006). Tw o of these vessels, one from Mayport and the other from Lewis Creek, were partially reconstructed. These vessels are strikingly similar in terms of vessel form and rim treatment: both ar e sub-conoidal pots with thick folded rims. However, their sizes are quite distinct, with orifice diameters of 26 cm (Mayport) and 36 cm (Lewis Creek). Also, paddle orie ntation is variable, oriented n early perpendicular to the rim (Mayport) or nearly parallel to the rim (Lewis Creek) (Figure 4-6) Notably, the Mayport vessel 119

PAGE 120

exhibits crisp, clear im pressions while the Le wis Creek vessel shows a crack in the wooden paddle, indicating that the Lewis Creek vessel was made sometime after the vessel from Mayport (Wallis 2004). There are three designs that link sites within nor theastern Florida. First, there is a match between the Dent (8DU68), Alicia B (8DU31), a nd Beauclerc mounds (8DU43). This is the sole Early Swift Creek paddle match in northeastern Florida, identified by charcoal temper and an AMS radiometric range for the Dent vessel between the 4th and 5th century (Ashley 1995). Notably, the Beauclerc specimen does not contain ch arcoal temper (Figure 4-7). Rim treatments among the Dent and Alicia B vessels, the latter illustrated by C.B. Moore (1895:plate 80), are identical. Second, I recently identified anothe r match between the Dent Mound and a Greenfield peninsula (8DU5544/5) village site (Figure 4-8). AMS dating of soot from the rim of the village vessel yielded a calibrated 2-sigma range of AD 660 to 790 (Stephenson et al. 2002). A third match links two potentially quotidian contexts, comprised of one sherd that was included in the fill of the Browne Mound (8DU62) that may postd ate Swift Creek occupation and one from the Schmidt site (8SJ52) in St. Johns Co. to the south. A few sherds from the latter site show design wear and well-developed paddle cr acks, indicating that this vessel was made later than other matching vessels at the site as well as the vessel from the Browne Mound (Ashley and Wallis 2006:12) There are three designs that link sites in Camd en Co. in southeastern Georgia. Designs 188 and 238 both link Kings Bay (9CM171) to Sadlers Landing (9CM233) along with several south-central Georgia sites. Design 142 links Sadlers Landi ng (9CM233) with Mallard Creek (9CM185) and several sites much further inland (Ashley et al. 2007:18). These Camden County 120

PAGE 121

sites are linked by eight designs w ith more than a dozen sites toward the interior coastal plain, mostly along the lower Ocmulgee River (Ashley et al. 2007:18-19). To summarize, of the six paddle matches with sites on the Lower St. Johns River, all but one includes definitive mortuary mound assemblages. In contrast, none of the six matches with southeastern Georgia sites includ es unambiguous mortuary contexts This difference may reflect both sampling bias and different modes of burial along the coast. Among paddle matches, the diversity of vessel forms and orientation of paddle impressions among matching vessels separated by many kilometers may invalidate the idea that wo oden paddles were personal property used by one individual (Wallis 2006). Georgia coastal assemblages include many paddle matches with the interior coastal plain, yet northeastern Florida assemblages contain no such matches. However, designs found on vessels from the Beauclerc Mounds (8DU43) on the Lower St. Johns River and the R.L. Smith Field site (9JD8) near the Oc mulgee River, though not a match, are similar enough to deserve mention (Figure 4-9). The size a nd number of elements in the de signs are not identical, but they clearly represent the same fundamental theme. Ra rely are two paddle designs so similar, except perhaps in the case of sun theme designs, charact erized by a concentric circle element in the design center (Frankie Snow, personal communica tion, June 2008). While the similarities in these designs do not necessarily reflect direct so cial contact between so uth-central Georgia and northeastern Florida, they indicate some degree of mutual influence in specific forms of artistic representation. Summary The Swift Creek archaeological culture was manifested on the Atlantic coast in regionally specific ways as populations with distinctive hi stories engaged larger-scale social and cultural developments. Variation in the rate at whic h complicated stamping was adopted, the size and 121

PAGE 122

122 number of mounds, type of burial practices, an d the layout of villages and mounds across the landscape show a spatial structure th at mirrors deep historical differe nces of at least the previous millennium. It is probably no coincidence that the southern extent of site s with large amounts of Swift Creek pottery essentially mirrors the distribution of Deptford sites because the two archaeological cultures were clearly related (compa re Figures 4-1 and 4-2). However, the closer proximity of the Lower St. Johns River to the St. Johns I cultures and Swift Creek cultures of the Gulf Coast led to a different cultural trajectory for each region. On the Atlantic coast, Swift Creek Complicat ed Stamped pottery was first adopted on the Lower St. Johns River, presumably reflecting soci al interactions with pop ulations on the Florida Gulf Coast. After about A.D. 500, populations living along the Atla ntic coast from the Altamaha River to south of the mouth of the St. Johns River were making complicated stamped pottery. With the ubiquity of complicated stamping also came evidence of interaction along the coast in the form of paddle matches, which seem to ha ve become more prevalent and conspicuous. While there are no Swift Creek paddle matches that link the Lower St. Johns and the Gulf Coast, social interaction between populations continue d after A.D. 500 as evidenced by Weeden Island series vessels at Lower St. Johns sites. By deta iling the contexts of pottery production, use, and deposition, the remaining chapters are devoted to deciphering the soci al interactions among Swift Creek populations on the Atlantic coast that have been only vaguely understood. Notes 1. This map represents sites that are presumed to have significant Deptford and St. Johns components, defined here as comprising one of the first three mo st prominent archaeological cultures for eac h site as listed in the site file. Although I believe that this map approximates the distribution of Deptford and St. Johns I, the St. Johns designation is particularly problematic and probably overestimates the number of St. Johns sites along the Lower St. Johns and north into Georgia. The Georgia Site File data code does not differentiate between St. Johns I and St. Johns II and, as Ashley (2003:69) has outlined, many Lowe r St. Johns sites are misrepresented in the Florida Site File as St. Johns I when only a few St. Johns sherds were recovered among many sand-tempered plain sherds. 2. Although this vessel was labeled Beauclerc Mounds at the National Museum of the American Indian, other artifacts labeled Alicia, Beauclerc Mounds may indicate that some objects attributed to the Beauclerc mounds may actually derive from the Alicia mounds.

PAGE 123

Table 4-1. Calibrated radiocarbon assa ys for Early and Late Swift Creek contexts in northeastern Florida. Site Beta # Material Measured C14 age (BP) C13/C12 ratio (o/oo) Conventional C14 age (BP) Calibrated 1 Sigma (AD) with intercept Calibrated 2 Sigma (AD) Reference 8DU68 182333 soot 1930 + 40 -24.2 1940 + 40 30 (65) 95 30 BC 135 Stephenson 2002 8DU96 GX315* charcoal 1865 + 95 -25.0 1865 + 95 55 (130) 250 50 BC 395 Wilson 1965 8DU17245 217829 soot 1830 + 40 -26.7 1800 + 40 150 (230) 250 120 340 Hendryx and Wallis 2007 8DU68 182332 soot 1690 + 40 -24.7 1690 + 40 330 (385) 410 250 430 Stephenson 2002 8DU96 168177 soot 1560 + 40 -24.5 1570 + 40 430 (460, 480, 520) 540 410 580 Stephenson 2002 Early Swift Creek 8DU96 190255 soot 1510 + 40 -25.3 1510 + 40 530 (560) 610 440 640 Wallis 2004 8DU96 169421 soot 1450 + 40 -22.8 1490 + 40 540 (580) 620 460 480 and 520 Stephenson 2002 8DU68 169420 soot 1330 + 40 -18.4 1440 + 40 580 (600) 650 550 660 Stephenson 2002 8DU81 181303 oyster 1460 + 70 -3.6 1810 + 70 540 (610) 670 440 710 Ashley 2003 8DU5545 163598 oyster 1390 + 60 -3.2 1390 + 60 600 (660) 690 540 740 Smith and Handley 2002 8DU5545 163597 oyster 1350 + 60 -2.3 1350 + 60 620 (670) 700 560 770 Smith and Handley 2002 8NA32 190666 oyster 1340 + 60 -1.6 1340 + 60 640 (680) 720 580 -780 Handley et al. 2004 8DU5545 168176 soot 1290 + 4 0 -24.5 1300 + 40 670 (690) 770 660 790 Stephenson 2002 8NA32 190665 oyster 1310 + 60 -1.8 1310 + 60 660 (700) 770 590 850 Handley et al. 2004 8DU68 182334 soot 1270 + 40 -25.0 1270 + 40 685 (720, 745, 760) 780 670 -870 Stephenson 2002 8DU81 182335 soot 1250 + 40 -25.4 1240 + 40 705-815 (775) 840-855 680 -885 Stephenson 2002 8DU68 54645 oyster 1250 + 70 -2.7 1610 + 70 695 (775) 865 655 -955 Ashley 1995 8NA709 126313 oyster 1180 + 60 0.0 1590 + 60 730 (800) 875 685 -945 Dickinson and Wayne 1999 8NA910 159965 tagelus 1120 + 60 0.0 1540 + 60 780 (870) 920 720 -1000 Hendryx and Smith 2001 Late Swift Creek 8NA910 159964 soot 1150 + 40 -23.4 1180 + 40 790 (880) 900 770 -970 Hendryx and Smith 2001 123 *Geochron Laboratories, Cambridge Massachusetts All others Beta Analytic, Miami, FL.

PAGE 124

Table 4-2. Calibrated radiocarbon assays for Early and Late Swift Creek contexts in southern coastal Georgia. Site Beta # Material Measured C14 age (BP) C13/C12 ratio (o/oo) Conventional C14 age (BP) Calibrated 1 Sigma (AD) with intercept Calibrated 2 Sigma (AD) Reference 9CM171 2122 oyster 1490 + 90 0.0 1891 + 90 415 (515) 615 295-680 Adams (ed.) 1985 9CM171 4005 oyster 1470 + 80 0.0 1871 + 80 430 (545) 632 330-690 Adams (ed.) 1985 9CM171 4421 oyster 1410 + 70 0.0 1811 + 70 515 (605) 670 420-720 Adams (ed.) 1985 9CM171 4418 charcoal 1470 + 100 -25.0 1470 + 100 450 (605) 665 385-765 Adams (ed.) 1985 9MC360 nr charcoal 1450 + 50 -25.0 1450 + 50 560 (620) 650 450-680 Wayne 1987 9CM171 4420 charcoal 1440 + 80 -25.0 1440 + 80 550 (625) 670 420-760 Adams (ed.) 1985 9CM171 3994 charcoal 1420 + 80 -25.0 1420 + 80 555 (640) 675 435-775 Adams (ed.) 1985 9CM171 3996 charcoal 1410 + 140 -25.0 1410 + 140 530 (645) 760 360-955 Adams (ed.) 1985 9MC372 82086 bone 1280 + 60 -17.1 1400 + 60 620 (650) 675 560-760 Cook 1995 9MC233 157590 soot 1400 + 40 -25.2 1400 + 40 630 (650) 660 600-680 Kirkland 2003 9CM171 3989 charcoal 1360 + 80 -25.0 1360 + 80 620 (665) 755 540-875 Adams (ed.) 1985 9CM171 4010 oyster 1330 + 60 0.0 1730 + 60 610 (670) 715 530-800 Adams (ed.) 1985 9CM171 3988 charcoal 1340 + 80 -25.0 1340 + 80 630 (675) 770 550-885 Adams (ed.) 1985 9CM171 4417 charcoal 1330 + 70 -25.0 1330 + 70 645 (675) 770 580-880 Adams (ed.) 1985 9CM171 4000 charcoal 1320 + 80 -25.0 1320 + 80 645 (680) 780 565-895 Adams (ed.) 1985 9CM171 3993 charcoal 1320 + 60 -25.0 1320 + 60 655 (680) 770 605-880 Adams (ed.) 1985 9CM233 157591 soot 1300 + 40 -26.4 1280 + 40 680 (710) 780 660-860 Kirkland 2003 9CM171 4429 quahog 1250 + 60 0.0 1650 + 60 670 (720) 805 610-900 Adams (ed.) 1985 9CM171 3995 charcoal 1260 + 130 -25.0 1260 + 130 655 (760) 950 545-1020 Adams (ed.) 1985 9CM171 4004 oyster 1220 + 90 0.0 1620 + 90 675 (760) 880 590-995 Adams (ed.) 1985 9CM171 4012 oyster 1190 + 60 0.0 1590 + 60 705 (790) 885 665-975 Adams (ed.) 1985 9CM171 4015 oyster 1180 + 60 0.0 1580 + 60 715 (800) 895 670-985 Adams (ed.) 1985 9GN6 225771 soot 1100 + 40 -25.2 1100 + 40 890 (970) 990 880-1020 Ashley et al. 2007 124

PAGE 125

Figure 4-1. Distribution of reco rded Deptford and St. Johns s ites at their intersection on the Atlantic coast based on Florida and Georgia state site file data. 125

PAGE 126

Figure 4-2. Sites with Swift Creek po ttery along the Atla ntic coast. 126

PAGE 127

Figure 4-3. Sites with paddle ma tches mentioned in the text. 127

PAGE 128

Figure 4-4. Reconstructed desi gn 34 [reprinted by permission of Frankie Snow (2007)] and select paddle matches. Courtesy of the National Museum of the American Indian, South Georgia College, and the Jacksonville Museum of Science and History. The Grant vessel is 18.5 cm tall. 128

PAGE 129

Figure 4-5. Reconstructed desi gn 36 [reprinted by permission of Frankie Snow (2007)] and select paddle matches. Courtesy of the Jacksonville Museum of Science and History, South Georgia College, and the Antonio J. Waring, Jr. Archaeological Laboratory, University of West Georgia. 129

PAGE 130

Figure 4-6. Reconstructed desi gn 38 [reprinted by permission of Frankie Snow (2007)] and select paddle matches. Courtesy of the Florida Museum of Natural History, South Georgia College, and the Antonio J. Waring, Jr. Archaeological Laboratory, University of West Georgia. 130

PAGE 131

Figure 4-7. Select paddle matches with unnumbered design. Courtesy of the Jacksonville Museum of Science and History and the National Museum of the American Indian. The Dent vessel is 20 cm wide. 131

PAGE 132

Figure 4-8. Reconstructed desi gn 291 [reprinted by permission of Frankie Snow (2007)] and paddle matches. Courtesy of the Jacksonville Museum of Science and History and Environmental Services, Inc.. 132

PAGE 133

133 Figure 4-9. Reconstructed design 151 [reprinted by permission of Frankie Snow (2007)], sherds bearing this design from 9JD8, and a near ly identical design from 8DU43. Courtesy of the National Museum of the American Indian.

PAGE 134

CHAP TER 5 INSTRUMENTAL NEUTRON ACTIVATION ANALYSIS: PATTERNS OF SWIFT CREEK INTERACTION, PART 1 The complicated stamped designs on Swift Creek pottery often provide compelling evidence for social interaction across the landscape, but the potential of th ese serendipitous data is only beginning to come to fruition as they are combined with detailed consider ations of context. Depositional context is routinely record ed by archaeologistsit is the identification of contexts of manufacture that holds new poten tial for understanding in teraction. Identifying where vessels were made compared to where they were deposited is a significant step toward outlining object biographies and crucial for unders tanding patterns of inte raction in the past. Instrumental Neutron Activation Analysis (I NAA) is an exceptionally powerful analytical technique for determining provenance that was not often used in the southeastern United States until recent decades (but for discussion of earlier work see Rice 1980 and Smith 1998). However, as the INAA chemical database for th e region grows, so too does the potential to decipher the provenance of vessels and thereby construct more robust models of interaction. Particularly compelling is the combination of INAA chemical data and the social contacts evidenced by Swift Creek paddle matches. This chapter presents the results of Instrumental Neutron Activation Analysis (INAA) of a sample of earthenware vessels and raw clays from Swift Creek contexts on the Atlantic coast. These data indicate that nonlocal vessels, particularly complicated stamped ones, were deposited almost exclusively at mortuary mounds on the Lower St. Johns River. This pattern requires new explanations for paddle matches and forms of interaction that seem to have been b ound up in mortuary ceremony. The Sample The sampling strategy for the INAA portion of this study was informed by two related goals. First, the project was designed to esti mate the range of compositional variation in 134

PAGE 135

presum ably local pottery for the various locations sampled. Based on achievement of this first goal, the second objective was to ev aluate the frequency of nonlocal pottery at sites. The spatial scale by which local and nonlocal pottery can be defined, of course, ultimately depends on the amount of chemical variation in the constituent materials distributed across the landscape. Thus, a large number of samples from each site was desirable for understanding intrasite variation while a large number of sampling locations of both raw materials and sherds from throughout the region would be advantageous fo r understanding chemical variation across the landscape. With a limited time and funding budget for INAA, these parameters were combined with a delicate balance. Samples were taken from 313 vessels from 16 sites across north eastern Florida and southeastern Georgia (Table 5-1; Figure 5-1). These samples were divided fairly equitably by region: 173 derived from eight si tes in northeastern Florida and 140 came from nine sites in southeastern Georgia. There were three criter ia that were employed in choosing samples from each site. When assemblages were large and came from discrete contexts from well-documented sites, samples were chosen at random, resulti ng in a diversity of sa mpled pottery types. Alternatively, samples from surf ace collections or poorly documented excavations presented a unique challenge. In these cases, Swift Creek Complicated Stamped sherds were chosen in order to insure contemporaneity and archaeological culture affiliation among samples. Finally, in some cases samples were limited to the preferences of curators who bristled at the suggestion of taking samples from reconstructed or whole vessels. Other than this last restriction, which may have introduced bias that I discus s in more detail with the analysis results, the sherds chosen from each site represent a random sample. 135

PAGE 136

Twenty clay sam ples from around the region we re also subjected to NAA, with data from an additional three samples appropriated from the MURR database from Keith Ashleys (2003) research (Figure 5-2; Table 5-2). These clay sa mples were drawn mostly from areas near the sherd sample sites, although a few from further south and west were tested in an attempt at evaluating larger spatial trends in chemical composition. Instrumental Neutron Activat ion Analysis (INAA): Methods Instrumental Neutron Activation Analysis (I NAA) is an analytical technique used to characterize the chemical consti tuents of a material. The analysis is based on the physical properties of the atomic nucleus, whereby gamma ra ys are released from a sample that has been bombarded with neutrons (i.e. irradiated) within a nuclear reactor (Glascock 1992:11-12). Specifically, the bombardment of a sample with neut rons results in unstable radioactive isotopes that decay with characteristic half-lives and em it gamma radiation characte ristic of each of the elements. As these isotopes decay, the gamma ra ys can be measured to determine the quantities of various elements within a sample. As an especially sensitive and increasingly standardized chemical characterization techni que, INAA continues to be a pow erful analytical tool for the study of pottery in the context of exchange and interaction (Gla scock 2002; Neff 1992, 2000; Steponaitis et al. 1996, etc.). INAA of pottery results in the characteriza tion of the bulk chemical composition of a sample. Because pottery samples are powdered and homogenized before irradiation, no attempt is made to differentiate the chemical profiles of clays and aplastic materials that make up a ceramic paste. Therefore, the aboriginal practice of removing materials from natural clay bodies or adding temper in the process of vessel manuf acture can certainly affect the bulk chemical composition of earthenwares. However, experi mental and statistical studies confirm that inordinate amounts of temper would usually need to be added to the paste of vessels in order to 136

PAGE 137

com pletely obscure the chemical profiles of different clays (N eff et al. 1988, 1989). There are several techniques that have been successfully used for evaluating the effects of temper in archaeological samples. One powerful technique employs a ceramic ecology approach: sampling a range of possible raw clay sources and tempering agents to tie chemical variation in pottery to different paste recipes (e.g., Ne ff and Bove 1999). Also, statis tical procedures have been developed to remove the chemical effects of some tempers such as shell (Cogswell et al. 1998:64; Steponaitis et al. 1988). Shell temperi ng adds large amounts of calcium and strontium so that other elements are diluted. However, the diluting effects of shell can be mathematically corrected using the recorded pr oportions of calcium and strontium measured for each sample. Pottery samples were prepared by the author according to standard procedures at MURR and under the direction of MURR staff (Glascoc k 1992). A small fragment was removed from each sample and abraded using a silicon carbide burr. This burring procedure removes any adhering soil, glazes, slips, paints, or other po tential contaminates on sample surfaces. The samples were then washed in deionized water and allowed to dry. Once completely dry, the samples were ground into a fine powder with an agate mortar and pestle, thus homogenizing each sample. The powder samples were then placed in an oven at 100 C for 24 hours to remove excess moisture. Two analytical samples were prepared from the resulting powders. Approximately 150 mg of powder was weighed in to high-density polyethylene vials used for short irradiations while 200 mg of each sample was weighed into high-purity quartz vials used for long irradiations. Each sample weight wa s recorded to the nearest 0.01 mg using an analytical balance and the vials were sealed prior to irradiation. As standard at MURR, INAA consisted of tw o irradiations and a total of three gamma counts on high-purity germanium de tectors for each sample (G lascock 1992; Neff 1992, 2000). 137

PAGE 138

The short irradiation was perform ed through a pn eumatic tube irradiation system, in which samples were each irradiated for 5 seconds in the reactor and then s ubjected to a 720 second gamma count. The second irradiation consisted of 24 hours within the nuclear reactor, with two counts subsequently performed, one after seven days of decay for 1800 seconds (i.e. the middle count) and the other af ter three weeks of decay fo r 8500 seconds (i.e. the long count). Combined, the two irradiations and thre e gamma counts allow detection of 33 elements. Based on the gamma counts, element concentration data were tabulated in parts per million. The goal of the quantitative analysis was to identify groups among the samples that shared a consistent chemical composition presumed to correspond with geographically restricted sources or source areas (e.g., Weigand et al. 1977). The locations of sources can be deduced by comparing specimens with unknown provenience (e.g., earthenware vessels) to substances with known provenience (e.g., clay samples). Provenien ce can also be inferred by indirect methods such as the criterion of abunda nce (Bishop et al. 1982) whereby th e chemical group(s) with the most samples at a site is presumed to be local. As detailed by Glascock (1992:16), compositional groups can be viewed as centers of mass in the hyperspace defined by the measured elemental data. Each group is ch aracterized by the locati on of its centroid in compositional hyperspace and unique correlations between the elements. Group assignment for each specimen is determined by the overall probab ility that the measured concentrations of a sample could have been obtaine d from a particular group. Standard quantitative proce dures at MURR were used to analyze the elemental data (Glascock 1992). The raw parts per million values were converted into base-10 logarithms in order to offset differences in magnitude between major and trace elements and to produce a more normal distribution for many of the trace elements Because the elemental datasets produced by 138

PAGE 139

neutron activation analysis are very large in num ber and often include highly correlated variables, data manipulation and interpretation can be difficult. Principle Component Analysis (PCA) is a data transformation technique that can be used to reduce the original variables into a smaller set of uncorrelated variables. Each Principle Component (PC) is a linear combination of measured variables based on eigenvector methods used to determine the dire ction and magnitude of maximum variance in the data (Glascock 1992:17). The first PC is orient ed in the direction of maximum variance while the second PC lies in the direction of maximum re maining variance oriented perpendicular to the first PC. Each subsequent PC lies in the directio n of maximum remaining variance and is orthogonal to the previous PCs until the number of PCs is equal to the number of original dimensions (Glascock 1992:18). PCA is effective for both finding partitions in a dataset and to test provisional groups developed on the basis of other criteria. One important strength of PCA is that it can be applied as a simultaneous Rand Q-mode techni que, with both variables (i.e. the elements) and objects (i.e. individual analyzed samples) displayed on the same set of PC reference axes (Baxter 1992; Baxter and Buck 2000; Neff 199 4, 2002). The inter-relationships between variables that are inferred from biplots can be verified by comparing bivariate elemental concentration plots. While PCA allows visual evaluation of group differentiation in two dimensions, Mahalanobis distance (or generali zed distance) is a metric that makes possible the measurement of separation between groups and/or individual samples in multiple dimensions (Bieber et al. 1976; Bishop and Neff 1989). Mahalanobis di stance takes into account the variances and covariances in a multivariate group, which can be converted into probabilities of group membership for each sample specimen. When group sizes are small, probabilities based on 139

PAGE 140

Mahalanobis distance can vacillate depending on the assum ption of whether or not each specimen is a member of the group to which it is being compared (Bishop and Neff 1989; Harbottle 1976). To circumvent this problem, a conservative approach can be used whereby each specimen is removed from its presumed gro up before calculating probability of membership (Baxter 1994; Leese and Main 1994). Erring on the side of caution, this approach may sometimes exclude true group members. Anothe r difficulty of small sample sizes is that membership groups must include more samples than variables (elements) to make Mahalanobis distance calculations possible. As an alternative to calculations based on the full elemental concentration space, the scores on PCs can be us ed instead as long as enough components are used to subsume 90% of the tota l variance in the data. A final important benefit of Mahalanobis distance calculations is that they can be used for substituting missing values, thus limiting the number of specimens that must be removed from group calculations. INAA Results The standard procedures at MURR resulted in the measurement of 33 elements. However, three of these were ultimately re moved from the quantitative data analysis. As is often the case with pottery, nickel (Ni) was belo w detection levels in the majority of samples and was therefore omitted during the analysis. After careful review of the distribution of calcium (Ca) and strontium (Sr) in the samples, these too were ul timately removed. These elements were present in measurable amounts in most samples, but upo n viewing their spatial distribution, I became suspicious that calcium and strontium levels reflected the chemistry of the samples depositional environment rather than the original constituent materials of pottery. Indeed, calcium and strontium levels were most elevated among a ssemblages from shell middens and typically diminished in samples from non-shell contexts For example, samples from sandy sites like Tillie Fowler (8DU17245) and Sidon (9MC372) had consistently low levels of calcium while 140

PAGE 141

sam ples from shell midden sites such as Greenfield #8/9 (8DU5544/5) and Cathead Creek (9MC360) had dependably high calcium levels (Figur e 5-3). Thus, there is some support for the possibility that calcium and strontium, as highly water soluble elements, could have been deposited in the pores of the permeable, low fire d pottery, especially in wet environments. The calcium data from clay samples provided little consolation. Calcium levels are generally highest among coastal cl ay deposits, however some clay samples from very near the coast are conspicuously low in cal cium (Figure 5-4). Furthermor e, many of the clay samples with the highest calcium levels are from archaeol ogical contexts in she ll midden and thus were presumably subjected to the same processes that might have caused calcium and strontium enrichment in sherds. Given the isomorphic re lationship between shell middens and calcium and strontium levels in samples, the possibility of diagenesis in shell midden contexts, or alternatively, leaching (chemical loss) in non-shel l contexts, may be the source of variation. Insitu alteration of calcium levels was a potential ly serious problem because variation in calcium explained a large percentage of variance with in the sample in the first two PCs during preliminary quantitative analysis. As a pr ecaution, a correction procedure was used to mathematically remove calcium and strontium fr om the samples and compensate for the diluting effect on the remaining 30 elements (Cogswe ll et al. 1998:64; Steponaitis et al. 1988).4 Composition Groups Two distinct composition groups are recognized within the chemical data from pottery samples. There are three additional provisional groups among the clay samples that are apparent in bivariate plots but small sample sizes currently preclude statistical support for thes e partitions. Among pottery samples, Group 1 is comprised of 129 samples and is dominated by assemblages from sites near the Lower St. Johns River. Gr oup 2 is made up of 98 samples and contains a 141

PAGE 142

m ajority of specimens from sites near the Lo wer Altamaha River. There were 86 pottery samples left unassigned to any group defi ned in the analysis. PC plots do not reveal convincing separati on of the groups, but were instructive in identifying the elements that most contribute d to compositional variation among samples. The first 5 PCs account for 77.5% of the variance in the dataset. PC1 has many equally contributing elements, primarily potassium (K), sodium (Na), rubidium (Rb), cobalt (Co), manganese (Mn), and many of the rare earth elements (terbium (T b), europium (Eu), cerium (Ce), samarium (Sm), neodymium (Nd), lanthanum (La)). PC2 has a very strong cont ribution from arsenic (As). Contributions to PC3 are widely distributed, coming mostly from cesium (Cs), barium (Ba), potassium (K), and many rare earth elements. PC4 shows a strong contribution from sodium and somewhat less from manganese, while PC5 is dominated by cobalt. The bivariate plots of the PCs s how considerable overlap in th e two groups. In general the PCs demonstrate that Group 1 members have higher levels of each element on average while Group 2 members are comparatively deficient in most elements (Figure 5-5). Furthermore, Group 1 has more compositional variability among samples compared to Group 2, which has far less variation. These conclusions are borne out in the arithmetic means and standard deviations for each element among groups (Table 5-3). The only exception to lesser amounts of each element in Group 2 is cobalt, which is more abundant in Group 2 than Group 1. Thus, while Group 1 can be described as having comparativ ely high levels of every element and Group 2 contains comparatively low levels, cobalt demonstr ates the opposite distri bution among groups. The overlapping bivariate plots of principal components show that the large degree of compositional variation in Group 1 generally subsumes the smaller range of variation in Group 142

PAGE 143

2. However, the groups are confirm ed by Mahalanobi s distance to be statistically distinct. For Group 1, no specimen has greater than 0.25% ch ance of being in group 2. By comparison, Group 2 members show an increased probability of membership in Group 1 because the more heterogeneous Group 1 subsumes much of the va riation in the smaller group and causes more probability of inclusion. Still, no sample in Group 2 has greater than 4.8% probability of being in Group 1. The chemical composition of clay samples is instructive in defining the compositional differences between pottery groups. While the nu mber of clay samples is too small to create statistically meaningful groups that can be tested by Mahalanobis distance, there are recognizable geographical trends in the chemical data that conform to differences observed among pottery samples. Clays taken from three general regions tend to cluster together into three distinct chemical profiles: the Lower St. Johns River, Lower Altamaha River, and lower Ocmulgee/Upper Altamaha River (Figure 5-6). Fi ve clays from near the Lower St. Johns River (Group 3) are characterized by comparatively high levels of arsenic and chromium, and low levels of cobalt, just like the pottery presumed to be locally made. In comparison, two lower Altamaha clays (Group 4) have low levels of ar senic and chromium, and high levels of cobalt. A third clay (NJW333) from near the lower Alta maha River appears to have an anomalous chemical profile. Taken from a dirt road along the high uplands several h undred meters south of the river in Wayne County, this sample may suffe r from contamination or, more likely, simply reflect a different chemical com position compared to clays formed along the river banks. Four Upper Altamaha/Lower Ocmulgee clays (Group 5) are characterized by high levels of all elements except arsenic, which is consistently low. In esse nce, along the Ocmulgee/Altamaha drainage the clay deposits closest to the piedmont and mountains contain the highest 143

PAGE 144

concentrations of m ost elements (except arsenic) and these gra dually diminish along the course of the river toward the coast, as is expected in comparisons of mountain, piedmont, and coastal plain regions of Georgia (Crocker 1999). Relatively high levels of cobalt and low levels of arsenic are attributes that are re tained in the Lower Altamaha clay s and consequently enable their differentiation from Lower St. Johns clays. A lthough there is much overl ap, the orientation of the two pottery groups in bivari ate plots of PCs generally conf orm to the differences observed among their presumed parent clays (Figure 5-7). Clay samples taken from areas along the coast between the St. Johns and Altamaha Rivers are quite variable in chemical composition. However, among six samples some generally shared characteristics include high levels of arsenic, cobalt, and chromium. With characteristic levels of arsenic and chromium similar to Lower St. Johns clays and levels of cobalt similar to Lower Altamaha samples, there is not a clear geographic trend in chemical composition from north to south. More clay samples are needed to outline the nuances of local chemical variation across the clay deposits of this region. The partition in the pottery sample data is most easily viewed in a bivariate plot of chromium and cobalt because the distribution of these elements shows little overlap between groups (Figure 5-8). While Mahalanobis distan ce was used as the primary test for group membership, chromium and cobalt levels prov ided an extremely reliable comparison of group distinctions, with high levels of chromium and low levels of cobalt for Group 1, and low levels of chromium and high levels of cobalt for Group 2. The differences in coba lt levels are likely to be the result of differences in the origins of se dimentary materials along the two rivers, with the Altamaha River ultimately draining the piedmont and mountains where cobalt levels are higher, and the St. Johns River running through the coasta l plain where cobalt leve ls are low (Figure 5144

PAGE 145

9a). The chrom ium levels among groups also co rrespond with differences in clay sources, with Lower Altamaha clays exhibiting slightly less chromium than their Lower St. Johns River counterparts (Figure 5-9b). The pottery samples left unassigned to any group fall into three categories (Figure 5-10). Some of the samples (n=27) are likely members to one of the two defined groups but each varies enough to preclude official inclusion based on statistical probabil ity. The remaining unassigned samples (n=58) are either repres entative of various local manuf acturing materials whose range of variation is poorly understood or nonlocal pottery from source lo cations poorly represented in the study (e.g., Middle St. Johns, Gulf Coast, central Georgia, etc.). Sites near the Lower St. Johns River are dominated by Group 1 samples (Table 5-4). Diagnostic Early Swift Creek samp les (charcoal-tempered) have es pecially high frequencies of Group 1 membership, reflecting th e unwavering prevalence of loca l production before A.D. 500. Among charcoal-tempered samples, 89% (n=42) are Group 1 members and the remainder (n=5) are unassigned. Limited Deptford (n=1), Mayport Dentate Stamped (n=2), and grog-tempered plain (n=1) samples are all included in Group 1. As should be expected, only samples from Late Swift Creek contexts show any evidence of pot ential coastal Georgia manufacture. Among sandtempered plain and Late Swift Creek samples, 65 are Group 1 members, 10 are confirmed Group 2 members, and one more is a very likely me mber of Group 2 but contains too great of a probability of Group 1 membership to be officially included. However, this last sample has greater than seven times the proba bility of belonging to Group 2 than Group 1. Several unassigned samples from Lower St. J ohns River sites are statistical outliers that may derive from nonlocal sources undefined in th is study. All St. Johns Plain (n=5) and Dunns Creek Red (n=1) samples remain unassigned. These samples tend to have diminished levels of 145

PAGE 146

all elem ents which may be the result of a past e recipe characterized by high levels of sponge spicule temper. Addition of this silica-based te mper would be expected to have a diluting effect on the chemical constituents of spiculate pastes similar to the effect of quartz sand temper (Michael Glascock, personal communications, Ma rch 2008). Thus, the chemical differences in St. Johns Plain specimens may merely result from a different paste recipe, perhaps simply more temper by volume compared to sand-tempered samp les. However, comparative pottery and clay specimens from the Middle St. Johns River area are needed to evaluate the possibility that these vessels were made using materi als from much further upriver (south). Carrabelle Punctated (n=1) and Weeden Island Red (n=1) samples from the Lower St. Johns River area and a Weeden Island Incised (n=1) sample from a site on Am elia Island (8NA32) are also unassigned and may have come from somewhere along the Gulf Coast of Florida where these pottery types are more common. Comparative chemical data from speci mens along the Gulf Coast and elsewhere are required to convincingly link the ma nufacture of these vessels to a particular nonlocal region. Samples derived from sites in the areas be tween the mouth of the St. Johns River and the Altamaha River show a nearly even split be tween Group 1 (25%, n=15), Group 2 (34%, n=21), and unassigned (30%, n=18) members. The remai nder of the sample is comprised of probable Group 1 members (11%, n=7). Th is broad distribution of sa mples among groups may result from the presumably variable chemical profiles of local clay sources in the coastal plain between the major drainages. Perhaps some clay deposits in this region are chemically similar to Lower St. Johns River clays while others are more si milar to Lower Altamaha River clays. This possibility is supported by the high degree of ch emical variation among six clay samples tested in the present study. However, gi ven their proximity to both the St. Johns and Altamaha rivers, sites along this central region of the coast may also simply c ontain nonlocal pottery that was 146

PAGE 147

m anufactured in areas to the north and south. Indeed, there is a co rrelation between group assignment and the size of quartz temper in a sa mple. The majority (n=11) of Group 1 members from these central sites are tempered with fine sand, much more common at sites on the Lower St. Johns River. Alternativel y, all 21 Group 2 members are temp ered with coarse sand, the nearly exclusive tempering material of the Lower Altamaha River region. Of the many unassigned samples, more than two-thirds contain medium and coarse sand temper and the remaining third contain fine sand temper. With traditional tempers from their respective regions, the Group 1 and Group 2 members may be vessels derived from the St. Johns and Altamaha River regions, respectively. Alternatively, pe rhaps the unassigned samples represent locally manufactured vessels from constituent material s poorly represented in the chemical study. Samples from Altamaha River sites are main ly Group 2 members (67%, n=68) or probable Group 2 members (9%, n=9). A mere 3% (n=3) are Group 1 members and 7% (n=7) are likely Group 1 members. The remaining 14 samples are unassigned to any group defined in the analysis. Some of these samples, with comparatively elevated levels of most elements, may derive from the Ocmulgee River area, where clay sa mples exhibit these chemical characteristics. In bivariate plots of some elements, these Lowe r Altamaha River samples show similarities to two samples from the Hartford site (9PU1) n ear the Ocmulgee River (Mainfort et al. 1997). However, many more comparative samples are ne eded to link manufacturing origins of these vessels to the Ocmulgee River region. Discussion The potential variation in the chemical composition of clays in the areas between the St. Johns and Altamaha rivers impedes the identifi cation of foreign vessels at sites in Nassau, Camden, and Charlton counties. However, this is not a problem among samples drawn from sites along the Lower St. Johns and Lower Alta maha Rivers, where clay samples establish 147

PAGE 148

consistent chem ical profiles. I therefore restricted the identification of foreign vessels to sites along the two major drainages. From the Lower St. Johns River sites there are 10 samples that are Group 2 members and one other that is a very likely Group 2 member. Including th is likely member, nine of the Group 2 vessels are Swift Creek Complicated Stamped a nd two are sand-tempered plain. Three of these Group 2 members have paddle matches with Altama ha river sites (Figur e 5-11). First, both vessels that make up the Design 34 match betw een the Dent Mound and Sidon site belong to Group 2. Second, the Dent Mound vessel with Design 36 belongs to Group 2 along with matching vessels from Cathead Creek and Lewis Creek. A fourth paddle matching vessel with Design 36 from Sidon is an unassigned outlier that is more likely to be a Group 1 member. Finally, the Mayport Mound, Cathead Creek, and Le wis Creek samples that share Design 38 are all Group 2 members. In sum, the three vessel s from the lower St. Johns that bear paddle matches that link them to Altamaha sites were all apparently made on the Altamaha river based on the chemical evidence. Thus, in all of thes e cases the vessels themselves, not the carved wooden paddles, were moved considerable distances. In comparison, three vessels that share a pa ddle match (Design 291) between two sites on the lower St. Johns were all placed in Group 1, indicating that they were all made locally, defined broadly as the Lower St. Johns region. In the case of paddle matches in such close proximity that share the same chemical group, th e data cannot be used to identify whether the earthenware vessels or the wooden paddle was moved. The distribution of foreign-made vessels am ong lower St. Johns sites is significant. Including the probable Group 2 me mber, 9 of 48 vessels from mounds are nonlocal, comprising 19% of the mound samples. Two of 96 vessels from middens were nonlocal, making up 2% of 148

PAGE 149

these assemblages. If we rem ove from the cal culations the charcoal-tempered samples, which we know come from early contexts that pre-date Late Swift Creek inte ractions with coastal Georgia, the difference is more dramatic. At mounds, nine of 34 vessels (nearly 27%) were made on the Altamaha, compared to only 2 of 69, or 3% of vessels from middens (Figure 5-12). Notably, the Group 2 members found on lower St. Johns sites are all temper ed with coarse or medium grit, the predominant temper size in coasta l Georgia. In fact, grit temper in a lower St. Johns vessel appears to be a fairly reliable indicator that it was made in the coastal sector of Georgia, a point that will be return ed to in the next chapter. The identification of nonlocal vessels in nor theastern Florida is limited to those that, based on chemical composition, were probably ma de somewhere near the Altamaha River. Unfortunately, this limitation may result from a severe sampling bias introduced through institutional restrictions that constrained the types of vessels that could be analyzed by destructive analysis. Specifically, the avoidance of many whole vessels from mounds resulted in no INAA samples of types that indi cate Florida Gulf Coast affiliati ons, especially Weeden Island Incised, Weeden Island Red, and Crystal River In cised. Based on the limited INAA samples of these types from midden contexts, which were all unassigned outliers, I believe that many of these vessels from mounds derived from the Gulf Coast of Florida. If foreign-made, the prevalence of these Gulf Coast vessels at mort uary mounds compared to habitation sites on the Lower St. Johns River conforms to the same pattern as the rest of the assemblage: foreign pottery was overwhelmingly deposited in mounds. The assemblages from sites along the Altamaha River are comprised of comparably fewer foreign-made vessels. Among 102 vessels from six sites, only two vessels from the Evelyn site and one from Paradi se Park are Group 1 members and may have been made along 149

PAGE 150

the Lower St. Johns River. Two of these vesse ls are Swift Creek Com plicated Stamped and the third is sand-tempered plain. There does not a ppear to be any correlation between temper and group membership, with fine sand, grit, and gr og tempers all represented. The paucity of foreign-made vessels at Altamaha River sites co mpared to Lower St. Johns sites may be due to sampling bias, specifically a lack of analyzed mort uary assemblages. In fact, the proportion of vessels identified in the middens of each region as being made in the other river valley is nearly identical, at roughly 3%. Futu re INAA of samples from mortuary mound contexts, such as Evelyn Mound C, might very we ll identify a much higher propor tion of pottery made on the Lower St. Johns River.5 Whatever the case of mound pottery in Georgia, the chemical data indicate that vessels made nearly one hundred kilometers distant were not a major part of assemblages that were used for routine domestic tasks and that ultimately became part of domestic midden assemblages. Therefore, Swift Creek interactions, at least at the scale of contact between populations on the Lower St. John s and Altamaha rivers, did not often involve the exchange of pots and/or their contents that were intended for normal use at villages. Summary To summarize the basic findings of the INAA study, evidence for Early Swift Creek influence from the Gulf Coast remains fairly intractable in terms of foreign-made vessels. As we might expect, almost all charcoal-tempered vess els were made locally. Some of the outlier samples that could not be assigned to either of the two chemical groups are likely from the Gulf Coast, especially the Carrabelle Punctated, Weeden Island Incised, and Weeden Island Red vessels. Alternatively, St. Johns Plain and Dunns Creek Red vessels may be from the Middle St. Johns River area or simply have a different paste recipe. In all cases, comparative chemical data will be necessary to confidently assign a regional provenience. 150

PAGE 151

W e can much more definitively see some form of exchange in Late Swift Creek contexts along the Atlantic coast. According to the chem ical data, Late Swift Creek interaction on the Lower St. Johns River seems to have been pr imarily centered on mortuary ceremony. The two analyzed mound assemblages from the Lower St. J ohns River have a proportionally much greater number of vessels identified as made on the A ltamaha River compared to village sites. The limited number of vessels at Altamaha River village sites that were identified as having been made somewhere near the Lower St. Johns Rive r indicate that the sa me pattern may have pertained here as well, with mounds being the pr imary locus for the deposit of foreign pottery. What seems fairly certain with these data is that the foreign vessel assemblages found in each region are probably dominated not by the de fact o refuse of moving people (from either marriage alliances or migration), or the containers for exchanged subsis tence goods, but rather by gifts intentionally emplaced at locations of heightened ritual importance. The final offering of these gifts seems to have been in mortuary contexts perhaps given in the context of deaths that obligated recognition of social connections and repayment of de bts. Before exploring the implications of this conclusion further, in the next chapter I outline mineralogical data derived from petrographic analysis that complements the chemical groups to give a more nuanced view of production origins and patterns of exchange. Notes 1. As Neff and Glascock explain (2001:4): The two dimensional plot of element coordinates on the first two principal components is the best possible two-dimensional representation of the correlation or variance-covariance structure in the data: Small angles between vectors from the origin to variable coordinates indicate strong positive correlation; angles close to 90 indicate no correlation; and angles close to 180 indicate negative correlation. Likewise the plot of object coordinates is the best two-dimensional representation of Euclidean relations among the objects in log-concentration space (i f the PCA was based on variance-covariance matrix) or standardized logconcentration space (if the PCA was based on the correlatio n matrix). Displaying objects and variables on the same plots [i.e., biplots] make it possible to observe the contributions of specific elements to group separation and to the distinctive shapes of the various groups. Such a plot is ca lled a biplot in reference to the simultaneous plotting of objects and variables. 2. The Mahalanobis distance of a specimen from a group centroid (Bieber et al. 1976, Bishop and Neff 1989) is defined by: 151

PAGE 152

152 ][][2 ,XyIXyDx t Xy where y is the 1 x m array of logged elemental c oncentrations for the specimen of interest, X is the n x m data matrix of logged concentrations for the group to which the point is being compared with X being its 1 x m centroid, and is the inverse of the m x m vari ancecovariance matrix of group X. Because Mahalanobis distance takes into account variances and covariances in the multivariate group it is analogous to expressing distance from a univariate mean in standard deviation units. Like standard deviation units, Mahalanobis distances can be converted into probabilities of group membership for individual specimens. For relatively small sample sizes, it is appropriate to base probabilities on Hotellings which is the multivariate extension of the univariate Students t (G lascock 1992). xI2T3. As Glascock (1992:19) explains: When analyzing many hundreds of specimens for a large number of elements, it is almost certain that a few concentrations will be missed for some specimens. This occurs more frequently when the concentration for an element is near its detection lim it in a group of specimens. Rather than eliminate such specimens from consideration, it is possible to substitute a missing value by choosing a value that minimizes the Mahalanobis distance for the specimen from the group centroid. Thus, those few specimens which are missing a concentration can be included in all group calculations. 4. The following mathematical correction was used, as defined by Cogswell et al. (1998:64) and Steponaitis et al. (1988): c e e 5.210 10 '6 6 where e is the corrected concentratio n of a given element in ppm, e is the measured concentration of that element in ppm, and c is the concentration of elemental calcium in ppm. After the calcium correction, statistical analysis was subsequently carried out on base-10 logarithms of concentrations on the remaining 31 elements. 5. Although collections from Evelyn are rumored to be at the Smithsonian, my inquiries there and at the National Museum of the American Indian turned up nothing.

PAGE 153

Table 5-1. Site and type distribution of INAA pottery samples. CCS CP SWCRCS STP DEPT STJP DCR MDS GROG WII WIR CAR Incised NI Total Northeastern Florida Dent Mound (8DU68) 3 7 10 8 28 Mayport Mound (8DU96) 2 2 4 6 1 2 2 1 20 McArthur Estates (8NA32) 2 19 4 1 26 Greenfield #7 (8DU5543) 9 3 9 5 26 Tillie Fowler (8DU17245) 7 9 4 4 1 1 1 1 28 JU Temp Sites 5 12 1 18 Greenfield #8/9 (8DU5544/5) 1 1 11 10 1 24 Grant (8DU14) 3 3 Southeastern Georgia Sidon (9MC372) 11 10 1 22 Cathead Creek (9MC360) 17 6 23 Evelyn (9GN6) 20 3 23 Lewis Creek (9MC16) 22 1 23 Kings Bay (9CM1 7 1) 4 4 Kings Lake 18 1 19 Paradise Park (9WY8) 6 2 8 Oak Landing 2 1 3 Hallows Field (9CM 15325) 15 15 Total Pottery Samples 313 CCScharcoal-tempered Swift Creek Complicated Stamped MDSMayport Dentate Stamped CPcharcoal-tempered plain GROGgrog-tempered plain SWCRCSSwift Creek Complicated Stamped WIIWeeden Island Incised STPsand-tempered plain WIRWeeden Island Red DEPTDeptford Check Stamped CARCarrabelle Punctated STJPSt. Johns Plain Incisedinci sed (non-diagnostic) DCRDunns Creek Red NInet impressed (sand-tempered)

PAGE 154

Table 5-2. Clay sam ples in the INAA study. Sample # ANID Provenience 1 NJW-315 New Smyrna, South Riverside Drive 2 NJW-316 Green Spring, Creek Bank 3 NJW-317 Grant Mound, Feature 1 4 NJW-318 Oxeye Island, NE Jacksonville 5 NJW-319 Grand Shell Ring, TU-4, L-13, Area U 6 NJW-320 Amelia Island Airport, Fernandina Beach 7 NJW-321 Little Talbot Isla nd, Black Rock, Nassau Sound 8 NJW-322 White Oak Plantation, St. Mary's River Bank 9 NJW-323 Osceola Forest, Comp. 2, St 13W, Tr. 2, St.2 10 NJW-324 Cabin Bluff Shell Ring, TU-201, 95 cmbs 11 NJW-325 Cabin Bluff Shell Ring TU-201, Level 8 12 NJW-326 China Hill, Telfair County, Georgia 13 NJW-327 Coffee Bluff, Ocmulgee River 14 NJW-328 "Hog Wallow" near Coffee Bluff, Ocmulgee River 15 NJW-329 Jekyll Island (south) 16 NJW-330 Jekyll Island (north) 17 NJW-331 "Clay-hole Island", Altamaha River bank 18 NJW-332 Lower Sansavilla, Altamaha River bank 19 NJW-333 Lower Sansavilla, upland road cut A KHA089 38CH42 B KHA033 Deen's Landing, upper Altamaha C KHA088 8LE151 Note: Samples 3, 5, 10, and 11 are derived from archaeological contexts. All others are natural clay deposits. Samples A, B, and C are from Ashley (2003). 154

PAGE 155

155 Table 5-3. Mean and standard deviation of el emental concentrations in each composition group. Group 1 Group 2 Element Mean (ppm) Standard Deviation Mean (ppm) Standard Deviation As 7.17 7.39 3.24 1.74 La 35.11 9.50 26.18 3.96 Lu 0.41 0.13 0.32 0.05 Nd 29.99 11.00 21.26 3.51 Sm 5.87 2.18 4.18 0.62 U 3.87 1.38 2.60 0.51 Yb 2.78 1.03 2.07 0.29 Ce 72.17 21.91 53.15 7.79 Co 4.65 1.76 6.59 1.10 Cr 75.19 12.78 51.87 5.19 Cs 3.40 1.10 2.92 0.73 Eu 1.02 0.51 0.68 0.12 Fe 31143.27 7489.33 26861.47 3873.80 Hf 13.20 3.03 9.72 2.14 Rb 33.36 10.83 27.50 6.92 Sb 0.32 0.11 0.22 0.04 Sc 12.29 2.25 11.00 1.38 Ta 1.10 0.20 1.15 0.14 Tb 0.74 0.33 0.53 0.10 Th 12.41 1.90 10.76 1.34 Zn 41.53 10.95 36.81 7.12 Zr 334.07 84.43 232.64 54.39 Al 71150.66 11536.09 68840.96 8603.35 Ba 272.02 100.23 319.92 150.46 Dy 4.46 1.85 3.38 0.59 K 6506.06 2552.45 3884.72 1192.57 Mn 126.56 56.75 118.54 58.84 Na 1870.48 859.34 1337.29 425.15 Ti 5008.90 690.66 4805.10 488.35 V 92.30 18.46 76.53 9.50

PAGE 156

Table 5-4. Pottery chemi cal group assignments by site. Group 1Group 2Probable 1Probable 2 Unassigned Total Lower St. Johns Dent Mound (8DU68) 20 5 3 28 Mayport Mound (8DU96) 11 3 2 1 4 21 Greenfield #7 (8DU5543) 24 2 26 Tillie Fowler (8DU17245) 23 5 28 JU Temp Sites 10 7 17 Greenfield #8/9 (8DU5544/5) 20 2 2 24 Grant (8DU14) 2 1 3 Total 110 10 2 1 24 147 Nassau, Camden, Charlton Co. McArthur Estates (8NA32) 7 13 1 5 26 Kings Bay (9CM171) 4 4 Hallows Field (9CM25) 5 4 3 3 12 Kings Lake 4 3 12 19 Total 16 21 7 20 64 Lower Altamaha Sidon (9MC372) 12 3 4 3 22 Cathead Creek (9MC360) 17 3 3 23 Evelyn (9GN6) 2 15 1 2 3 23 Lewis Creek (9MC16) 19 2 1 1 23 Paradise Park (9WY8) 1 3 1 3 8 Oak Landing 1 1 1 3 Total 3 67 8 10 14 102 Total Pottery Samples 129 98 17 11 58 313 156

PAGE 157

Figure 5-1. Distribution of sites with assemblages used in the INAA study. 157

PAGE 158

Figure 5-2. Distribution of cl ay samples in the INAA study. 158

PAGE 159

Figure 5-3. Bivariate plot of Cr and Ca in assemblages fr om shell-bearing (8DU5544/5 and 9MC360) and shell-devoid (8DU17245 and 9MC372) sites. 159

PAGE 160

Figure 5-4. Inverse Distance Weighted (IDW) interp olation of Ca concentrations in both natural and archaeological clay samples. The highest concentrations come from archaeological deposits in shell midden. 160

PAGE 161

Figure 5-5. Biplot of the first two principal components along with the relative influence of each of the elemental variables. Ellipses represent 90% confidence level for group membership. 161

PAGE 162

Figure 5-6. Bivariate plot of PC 2 and PC 4 for the data wi th only clay samples plotted. Tentative chemical groups are outlined. 162

PAGE 163

Figure 5-7. Same bivariate plot of PC 2 and PC 4 as figure 5-6 but with pottery group members and Group 3, 4, and 5 clay samples plotted. Ellipses represent 90% confidence of group membership. 163

PAGE 164

Figure 5-8. Bivariate plot of Cr and Co showing separation of pottery groups and tentative clay groups. A B Figure 5-9. IDW interpolation based on clay samples for A) cobalt and B) chromium. 164

PAGE 165

Figure 5-10. Bivariate plot of Cr and Co with unassigned samples represented by crosses. Ellipses represent 90% confidence of group membership. 165

PAGE 166

166 Figure 5-11. Bivariate plot of Cr and Co with paddle matching samples plotted. Solid symbols represent samples from St. Johns River site s. Ellipses represen t 90% confidence of group membership. Figure 5-12. Percentage of chemical group a ssignments, excluding charcoal-tempered vessels, from mound (n=34) and midden (n=69) sites on the Lower St. Johns River.

PAGE 167

CHAP TER 6 PETROGRAPHIC ANALYSIS: PATTERNS OF SWIFT CREEK INTERACTION, PART 2 The effectiveness of INAA in differentia ting local and nonlocal pottery has been questioned (Shrarer et al. 2006; St oltman et al. 2005; Stoltman a nd Mainfort 2002). Specifically, Stoltman et al. (2005:11214) argue that petrographic poi nt count data are superior to chemical data because the identification of the mineral constituents in ceramic pastes allows for nuanced understandings of both added tempers and the parent rock of constituent clay s. However, this is not always the case, as sometimes the chemical va riation in constituents (especially clays) is not attributable to microscopically visible mineralogical variation (Stoner et al. 2008). I follow a more moderate approach that recognizes th e complementary nature of chemical and mineralogical data (e.g., Bishop et al. 1982; Neff et al. 2006; Stoner et al. 2008). Indeed, the most problematic aspect of attempting to source clays using bulk chemical composition analyses of pottery is that vari ation in composition between samples ma y reflect a suite of materials added to clay. As repeatedly cauti oned by researchers using INAA, ce ramics are composite materials that require careful considerations of life histories that may have contributed to distinct chemical compositions through processes like the addition of temper and diagenesis (Glascock 2002:3; Neff et al. 2006). Given these potentially complica ting factors, studies of pottery distribution benefit from multiple sourcing methods that can be used to test th e results derived from different data. With this goal in mind, a sample of th e INAA assemblage was selected for petrographic analysis. The Sample Petrographic analysis was performed on thin se ctions from a total of 69 vessels from 14 sites and 10 unique clay samples (T able 6-1; Table 6-2). Of these, 57 of the pottery samples and all of the clays were analyzed expressly for the current study while additional data from 12 167

PAGE 168

pottery samples was incorporated from previous research (e.g., Wallis 2004; Wallis et al. 2005). All samples were selected with the goals of representing the range of variation in aplastic constituents, approximating the relative freque ncy of each paste recipe within the total assemblage, and being proportionate with th e INAA samples from each site. Using characterizations of the paste from analysis with a binocular microscope as well as typological designations, 43 vessels from site s in northeastern Florida and 26 from sites in southeastern Georgia were selected (Table 6-1). These pe trographic samples repres ent 25% and 19% of the INAA samples from each region, respectively. A comparatively greater proportion of the INAA sample from northeastern Florida was selected for petrographic analysis in order to accommodate the greater range of variation in apla stic constituents and typological diversity in this region compared to southeastern Georgia. Petrographic analysis of clays was limited to samples from the Atlantic coast near the archaeological sites repr esented in this study (Table 62). Methods The methods of petrographic analysis empl oyed in this study are explained by Cordell (2008). In brief, the analysis was conducted to evaluate compositional and textural variability in the samples and to document potential matches betw een pottery samples and clays. Point counts were made for quantifying relativ e abundance of inclusions. This procedure involved using a petrographic microscope with a mechanical st age and generally followed recommendations by Stoltman (1989, 1991, 2000). A counting interval of 1 mm by .5 mm was used in most cases. A 1 mm by 1 mm counting interval was used in the 12 cases that had been analyzed for a previous study (Wallis 2004; Wallis et al. 2005). Each point or stop of the stage was assigned to one of the following categories: clay matrix, void, si lt particles, charcoal temper, grog temper, 168

PAGE 169

bone tem per, biogenic silica (sponge spicules, phyt oliths, diatoms), and very fine through very coarse quartz and other aplastics of varying compositions. For cases in which fewer than 200 points were counted (n=8), the thin sections were rotated 180o on the mechanical stage and counted a second time (after Stoltman 2000:306). Mo st of the point counts were made using the 10X objective, but the 25X objective (with plan e-polarized light) was used to search for occurrence of siliceous microfossils such as sp onge spicules, phytoliths, and diatoms. Size of aplastics was estimated with an eyepiece microm eter with reference to the Wentworth Scale (Rice 1987:38). A comparison chart of percen t particle abundan ce (Rice 1987:349 [Figure 12.2]) was also used for estimating relative abundance of constituents occurring in low frequency All analyses were carried out by Ann S. Cordell in the Florida Museum of Natural History Ceramic Technology La boratory (FLMNH-CTL). In addition to the petrographic analyses, pottery samples of sufficient size were refired to standardize color comparisons between samples, to assess relative iron content of the clays represented by the samples, and for comparison to the clay samples. The lapidary saw was used to control the desired size of fragments for refiri ng but not all sherds in the sample were large enough to spare removal of pieces for refiring. Sher ds were refired in an electric furnace at a temperature of 800oC for 30 minutes, conditions that probably exceeded those of the original firings. Raw clay samples were also fired under comparable c onditions. The kiln temperature was initially set at 275oC and held for 10 minutes with the kiln door slightly open to allow for the escape of water vapor. The kiln door was then shut completely and the temperature was raised to 800oC. After about 15 minutes, the 800oC temperature was achieved and maintained for 30 minutes. The total firing time was approximately 75 minutes. Prior to refiring, original core color/degree of coring was recorded for each sher d (from a fresh break) with reference to five 169

PAGE 170

nom inal categories ranging from "no coring" to "heavy dark cori ng." A fresh break was made after refiring to note color cha nges. Four nominal refired colo r categories were distinguished on the basis of gross visual differe nces and correspond with relative iron contents ranging from very low to high. Results Mica, sponge spicules, phytoliths, diatoms, silt grains, and very fine sand were considered significant for defining clay re source groupings among the 69 pottery samples. The first five constituents are considered naturally occurring in some clays. Quartz aplastics falling into silt and very fine Wentworth particle sizes are us ually also considered as naturally occurring constituents of the clay source (Rice 1987:411; also see Stoltman 1989:149-150, 1991:109-111). Differences in fine through very coarse quartz particle sizes and other constituents are attributed to tempering practices, although so me fine sand may be naturally present in some cases based on variability in some of the clay samples. Base d on these six naturally occurring constituents, six petrographic paste groupings we re specified for pottery samples (Table 6-3). Each group represents a hypothetical resource group made up of one or more cl ay resources that are similar in terms of these six criteria. Based on the same criteria, some of the clays were assigned to one or more of the six pottery paste categories, while others formed their own categories. Five predominant temper categories were obser ved in the sample: charcoal temper, quartz sand, quartz grit (p article size > .5mm; includes some quartzite), quartz sand and grit, and grog temper (Table 6-4). Bone temper was observed in some samples, but was never the predominant constituent. Other constituents include mica (m uscovite and biotite or some other slightly pleochroic mica), feldspars (mainly microcline and plagioclase), granitic rock fragments (rarely), ferric concretions or nodules, birefringent gr ains (epidote, amphibole, UID minerals), and siliceous microfossils (sponge spicules, phytoliths, diatoms). Most of these other constituents 170

PAGE 171

that are probably naturally occurring in the pottin g clays were variously ob served in the sample, especially mica, ferric concretions, and the siliceous microfossils. Feldspars and other birefringent minerals may be naturally pres ent or introduced along with sand temper. The siliceous microfossils in the present sample are only detectable in thin section with magnifications ranging from 250x to 400x. Most of the sponge spicules and diatoms are fragmentary and are presumed to be natural constituents of the clay resources used for vessel manufacture. The six petrographic groups can be descri bed according to the re lative abundance of aplastic constituents considered to be natural in clusions in the exploited clays. Of these, variation in the occurrence of mica, sponge spicules, phytoliths, and diatoms are most significant (Figure 6-1). Group A is comprised of the most samples (n=31), and is characterized by rare to occasional mica, absent or rare sponge spicul es, and most conspicuously, absent or rare phytoliths. Based particularly on the absence or ra rity of phytoliths, four clays are also assigned to Group A. However, the potential in each clay for variability in some constituents makes other group designations possible as well. Group B (n=16) is similar to Group A, but is differentiated by the occasional to frequent occurrence of phyt oliths. Group C (n=4) contains variable frequencies of phytoliths and mainly differs from Groups A and B in the occasional to frequent occurrence of sponge spicules. Group D (n=13) is defined primarily by occasional to frequent mica. Group E (n=4) is characterized by hi gh frequencies of mica like Group D, but with occasional to frequent sponge spicules. Group F is comprised of a single sample and is similar to Group D but with occasional diatoms. Group F/G clays (n=3) are a potential match for the Group F sherd because of matching species of diatoms. However, these clays differ from Group 171

PAGE 172

F in having only rare sp onge spicules and rare to occasional mica. The single clay that constitutes Group H, from a prepared clay stoc kpile from an archaeolo gical context at the Grant site (8DU14), is defined by sponge spicules as the predominant aplastic inclusion and does not match any of the pottery samples in this st udy (Figure 6-1C). Finally, Group I clays (n=2) are defined by very high mica content, occasional to frequent sponge spicules, and rare diatoms. The five gross temper groups defined in the an alysis are not isomorphic with the six clay resource groups (Table 6-4). For example, Group A and B specimens are found in each of the five temper categories. Alternatively, the smaller Groups C, E, and D demonstrate some important correlations. Group C samples (n=4) are comprised entirely of grit-tempered sherds. Group E (n=4) samples are made up of grit-tempered or grit-and-sand-tempered sherds. Group D (n=13), consists primarily of sand-tempered sher ds but also some charcoal-tempered samples. These correlations reveal an intersection between mineral ogically distinct clays and geographically circumscribed tempering traditions To review, grit temper predominates in Swift Creek assemblages along the Altamaha River and as far south as Amelia Island while fine sand temper (often with charcoal before A.D. 500) dominates Lower St. Johns River assemblages (Ashley and Wallis 2006). The petrographically defined resource groups correspond with geographical areas that can be usefully summarized by county (Table 6-5). Group A samples come mostly from the southernmost counties, Duval (58%) and Nassau (19%). Likewise, Group D samples are mostly from Lower St. Johns sites in Duval Co. (77%). Group E is made up exclusively of samples from Bryan, Glynn, and McIntosh counties, all in Geor gia. Group B and C samples are more evenly divided by county but many of the Duval county specimens may be foreign imports based on INAA data. Specifically, the INAA results indica te that two of the Group B specimens from 172

PAGE 173

Duval Co. are foreign imports from the Alta m aha (e.g., chemical Group 2) while one from Nassau Co. is unassigned to either chemical gr oup. This leaves only two (13%) of the Group B specimens as locally produced in northeastern Florida based on the INAA data. Similarly, both Duval Co. specimens in petrographic Group C are likely imports based on chemical data, one a chemical Group 2 member and the other unassig ned to any chemical group. In sum, based on the geographical correlation with mineralogical groups the petrographic analysis identified two resource groups presumed to be local to the Lower St. Johns River area and three resource groups local to the Altamaha River area. The two Lower St. Johns groups are Group A and Group D, which differ from each other mostly in terms of mica content. The three Altamaha groups are B, C, and E, the latter two groups sharing occasional to frequent sponge spicules. Without more samples, the single sherd containing diatom s and sponge spicules (Group F) cannot be confidently assigned a geographic origin. In general, the mineralogical groups defined by the petrographic analysis are corroborated by the INAA chemical groups (Table 6-5). In mineralogical Group A, more than twice as many samples are chemical Group 1 members (local to the Lower St. Johns River) compared to chemical Group 2 members (local to the Altamaha River). Group B contains more than three times as many chemical Group 2 members as ch emical Group 1 members. Group C includes only chemical Group 2 or unassigned sample s while Group D contai ns only Group 1 or unassigned specimens. Group E is the most variable in terms of chemical composition but also suffers from small sample size, with only four members. The data from petrographic analysis of limite d clay samples help clarify some of the discrepancies between minera logical group members, chemic al group members, and their 173

PAGE 174

geographic distribution. Group A clays com e from sites throughout the study region, from Glynn, Nassau, and Duval counties. Therefore, Group A clay res ources, and by extension Group A pottery, are unlikely to be restri cted exclusively to the Lower St. Johns River. In other words, the wide distribution of Group A clays sets up an expectation for multiple origins among Group A pottery members. There are no natural mineralogical differences between Group A clay resources distributed throughout th e project area, but INAA was ab le to identify geographically significant chemical differences between them. However, based on the criterion of abundance (Bishop 1982) among sherds, Group A mineralogical characteristics may at least be more prevalent, though not exclusive, in cl ay resources toward the south. Group F/G clays are also widely distribute d, derived from Glynn, Camden, and Nassau counties, although none come from Duval County. For the purposes of this study, the spatial distribution of these Group F/G cl ays has little bearing on the sourcing of sherds because only one sherd potentially matches this group. Group I seems to be the only clay group with a significant spatial correlation as the two clays comprising th is group both come from the Ocmulgee/Altamaha river drainage. Group I clay is the only mineralogical group that contains moderate amounts of both sponge spicules and mica, firmly tying pottery Groups C and E to this drainage area. In fact, occasional to frequent naturally occurring sponge spicules only occur in these two Georgia clay samples and pottery samp les in these two Georgia pottery groups. The lone member of Group F also contains occasional to frequent sponge spicules but this vessel is tempered with spiculate paste grog that may have introduced them to the prepared paste. Complementing the Chemical Evidence Through the similarity of mineral constituents in some clays across the region, the mineralogical Group A cross-cuts the two chem ical groups determined by INAA. The other mineralogical groups mostly conform to the two chemical groups but also parse them further into 174

PAGE 175

subdivisions based on m ineralogical differences. This relationship is evident in comparisons of the mineralogical and chemical categories assi gned to vessels with matching paddle designs (Table 6-6). With the exception of one unassi gned sample, all vessels with paddle matching designs 34, 36, and 38 share the same chemical Group 2 but are split among two different mineralogical groups, A and B. These vessels were all probably made near the Altamaha River based on the chemical evidence, but with two or more mineralogically different clay sources. However, paddle matching vessels belonging to the same chemical and mineralogical groups are likely to have been made from the same clay source. This is the case among three of the paddle matches. Vessels with design 36 from the near by sites of Cathead Creek (9MC360) and Lewis Creek (9MC16) are assigned to chemical Group 2 and mineralogical Group B. Vessels with design 38 from Lewis Creek and the quite di stant Mayport Mound (8DU96) are members of chemical Group 2 and mineralogica l Group A. Finally, vessels sh aring design 291 from the Dent Mound (8DU68) and Greenfield #8/9 (8DU5544/ 5) belong to chemical Group 1 and mineralogical Group A. Alternatively, two paddle matching vessels from the same site (8DU5544/5) have different mineralogical designations, Group A and Group D, distinct groups both local to the Lower St. J ohns River area. Thus, the da ta from INAA and petrographic analysis complement one another, each providing data for further distinctions where the other indicates homogeneity. The petrographic analysis was also useful in evaluating the potential sources of chemical variation in the INAA data. One potentially obfus cating factor in identifying chemical variation in the clay resources used to manufacture vessels was variation in the amount of temper used. As reviewed in the previous chapter, there is a strong corre spondence between chemical group assignment and type of temper. Group 1 member s are overwhelmingly tempered with fine sand 175

PAGE 176

and/or charcoal while Group 2 m embers are pred ominantly grit-tempered (medium or coarse sand). Quartz tempers are known to dilute the chem ical constituents in bu lk chemical profiles, because, except for traces of hafnium and zirconium the elements that comprise quartz itself are not detected by INAA (Neff et al. 1989:66; Steponaitis et al 1996:559). Because Group 2 contained, on average, lower con centrations of each measured el ement and larger quartz particles compared to Group 1, a record of the proportion of quartz tempers based on petrographic point counts was useful in unders tanding the chemical effect s of this temper. The point count data revealed that chemical Group 2 contains slightly more quartz than Group 1 and would therefore be presumed to have a diluting effect on the chemical composition of members of this group (Table 6-7). However, the comparatively low quartz percentage in Group 1 also corresponds with charcoal temper in many samples, a siliceous material that might also have a diluting effect. When charcoal temper percentages are added to the quartz percentages, the mean for Group 1 is 31% (with a standard deviation of .052), closer to the Group 2 mean of 35%. Scatter plots and correlation coefficients s how correlation between many elements and the amount of temper in a sample (Figure 6-2; Ta ble 6-8). Among 31 measured elements, all but two show a negative correlation between element concentrati ons and proportion of quartz. However, the strength of this correlation varies dramatically among elements. For example, the elements scandium (Sc), aluminum (Al), chromium (Cr), and uranium (U) show fairly strong negative correlation with quartz while cobalt (C o), titanium (Ti), calcium (Ca), and manganese (Mn) have comparatively weak negative corre lations. The majority of elements have correlations with quartz somewhere between thes e two extremes. Some of these correlations help explain the chemical differences between INAA groups. For example, chromium, which 176

PAGE 177

was a fairly reliable indicator of group m embership in the ch emical analysis, shows a strong negative correlation with quartz (r= -.59), indicating th at chromium deficiency in many of the Group 2 samples is likely due to the diluting e ffect of quartz temper in high proportions. However, there is also a noticeable paucity of chromium in two of the three lower Altamaha clays, indicating that low chro mium values in the clays used for Group 2 pottery may have contributed to this deficiency as well. Among the elements that show weak corre lation with quartz, some correspond with particularly small chemical differences between INAA groups. For instance, there is a 6% difference in manganese between Group 1 and Gr oup 2 and an equally low correlation between this element and percent quartz (r=-.04). A lternatively, the distribu tion of cobalt between chemical groups does not correspond with the propor tion of quartz. Cobalt shows relatively little correlation with quartz (r=-.22), yet all chemical Group 2 members contain comparatively high levels of cobalt, demonstrating that this chem ical difference stems from differences in the constituent clays rather than the diluting effect s of temper (Figure 6-2). Thus, although quartz temper seems to have had a dilu ting effect on cobalt levels, the di fferences in cobalt between the constituent clays used for each chemical group are significant enough to remain useful in partitioning the data. Hafnium (Hf) and zirconium (Zr) are the onl y elements that demonstrate the opposite correlation with proportion of quartz, with each el ement increasing slightly in direct proportion to the amount of quartz. This relationship is expected because hafnium and zirconium are common trace elements in quartz. These elem ents were not fundamental to determining chemical group affiliations, but it is worth notin g that Group 1 includes an average of roughly 30% more hafnium and zirconium than Group 2, yet Group 2 contains more quartz than Group 1. 177

PAGE 178

Therefore, although the hafnium an d zirconium levels of pottery samples were inflated in relation to the amount of quart z temper added by ancient potters, Group 1 clay resources naturally contained more of these elements befo re temper was added. Scatter plots and correlation coefficients of quartz and principle components, which can be conceptualized as linear combinations of the el ements, show variable co rrelation that correspond with the various contributions of each element (F igure 6-3; Table 6-8). PC1 and PC3 show the strongest correlations with quartz, reflecting strong contributions fr om elements subjected to the heaviest effects of dilution: pot assium (K), rubidium (Rb), cesium (Cs), and many of the rare earth elements. In contrast, PC2, PC4, and PC5 s how the weakest correlations due to significant contributions from elements apparently less su sceptible to dilution th rough the addition of quartz, especially barium (Ba), manganese (Mn), cobalt (Co), hafnium (Hf), and zirconium (Zr). Consequently, bivariate plots of these principl e components structured the partitions in the chemical data much better than PC1 and PC3, which may have been more heavily influenced by the diluting effects of quartz. Thus, as many researchers have rec ognized in their analyses of a variety of temper materials (Bishop et al. 1982 ; Neff et al. 1988, 1989), temper can affect the levels of each element differently, diluting some and augmenting others. In the case of this particular sample, quartz temper appears to have diluted some elements while having little effect on others. As might be expected for quartz, which is almost pure SiO2, there is no evidence that the temper added measurable chemical constituents to the paste of vessels except small amounts of hafnium and zirconium. Notably, the same relationships between chemi cal concentrations and proportion of quartz pertain to clay samples, most of them raw clay samples from natural deposits (i.e. not archaeological). The elements chromium (Cr), antimony (Sb), and zinc (Zn) are especially 178

PAGE 179

diluted in direct proportion to quartz, while other elem ents appear to be only weakly correlated. There is a very wide variation in the proportion of quartz in the clay samples, ranging from 8% to 57% in the natural deposits, and these propor tions do not correspond with specific geographic areas but are distributed broadly across the landscape. Yet these disparities in quartz and their presumed diluting effect on some chemical cons tituents do not obfuscate the chemical partitions in the data that are geographically significant. In element and principle component plots, clays from the Lower St. Johns, Lower Altamaha and Upper Altamaha/Lower Ocmulgee each generally cluster together in their respective geog raphic groups, regardless of the wide variation in quartz proportions. The same is true of pottery samples, in which some elements, particularly cobalt, were not diluted by quartz temper enough to conceal the chem ical differences attributable to regionally distinct clays. Summary The petrographic analysis re sulted in six distinctive mineralogical groups among the pottery samples. Based on the geographic correlation of these po ttery groups and comparisons with clay samples, three of the groups are comprised of Altamaha River area resources (B, C, E), one group is local to the Lower St. Johns River (D), and the remaining groups are either not geographically distinctive (A), or sample size is too small to designate a geographic affiliation (F). These mineralogical groups complement the chemical data obtained from INAA. Partitions in the chemical data usefully distinguish the regional origins of ubiquitous mineralogical Group A samples, while mineralogical differences were used to identify distinctive clay resources within the broadly defined chemical groups. The combination of methods therefore provides an exceptionally nuanced view of pottery production and distribution. When combined wi th paddle match data, the mineralogical group assignments show that some matching vessels were very likely to have been made from the same 179

PAGE 180

clay deposits, while others were m ade in the sa me region but with mineralogically distinct clays (perhaps, though not necessarily, at different villages). These conc lusions bring us much closer to understanding the production orig ins of some vessels, especially those with paddle matches. Petrographic point counts were also useful in evaluating the chemical differences between groups. Although the correlation between quartz a nd the chemical data indicates that most elements were diluted by the addition of quart z temper, not all elements were diminished equally. Fortunately, geographic differences in the concentration of some elements were extensive enough to enable assignment of a re gional provenance to most pottery samples regardless of dilution through temper. Notes 1. The results of refiring are included in descriptions of the mineralogical groups in tables 6-1 and 6-2. However, comparison of refired color designations and the more precise quantification of iron by INAA reveals only a very general correlation between color and Fe content. Therefore, the refired color of sherds was not considered to be a significant variable in distinguishing mineralogical groups. 180

PAGE 181

181 Table 6-1. Site and type distribution of petrographic analysis pottery samples. CCS CP SWCRCS STP GROG WII WIR Total Northeastern Florida Dent Mound (8DU68) 1 2 5 1 9 Mayport Mound (8DU96) 1 3 1 1 6 McArthur Estates (8NA32) 2 5 1 8 Greenfield #7 (8DU5543) 3 2 1 6 Tillie Fowler (8DU17245) 1 1 1 1 1 5 JU Temp Sites 2 2 4 Greenfield #8/9 (8DU5544/5) 1 2 2 5 Total 6 9 18 7 1 1 1 43 Southeastern Georgia Sidon (9MC372) 5 1 6 Cathead Creek (9MC360) 5 5 Evelyn (9GN6) 3 2 5 Lewis Creek (9MC16) 4 4 Kings Lake 3 3 Paradise Park (9WY8) 1 1 Hallows Field (9CM25) 2 2 Total 23 3 26 Total Pottery Samples 6 9 41 10 1 1 1 69 CCScharcoal-tempered Swift Creek Complicated Stamped CPcharcoal-tempered plain SWCRCS Swift Creek Complicated Stamped STPsand-tempered plain GROGgrog-tempered plain (non-diagnostic) WII Weeden Island Incised WIRWeeden Island Red Table 6-2. Clay samples select ed for petrographic analysis. Sample # ANID Petrographic ID Provenience 3 NJW-317 C03-58c Grant Mound, Feature 1 (Duval Co.) 4 NJW-318 C04-59c Oxeye Island, NE Jacksonville (Duval Co.) 5 NJW-319 C05-60c Grand Shell Ring, TU-4, L-13, Area U (Duval Co.) 6 NJW-320 C06-61c Amelia Island Airport, Fernandina Beach (Nassau Co.) 7 NJW-321 C07-62c Little Talbot Island, Black Rock, Nassau Sound (Nassau Co.) 10 NJW-324 C10-63c Cabin Bluff Shell Ring, TU-201, 95 cmbs (Camden Co.) 13 NJW-327 C13-64c Coffee Bluff, Ocmulgee River (Telfair Co.) 15 NJW-329 C15-65c Jekyll Island (south) (Glynn Co.) 17 NJW-331 C17-66c "Clay-hole Island", Altamaha River bank (Glynn Co.) 18 NJW-332 C18-67c Lower Sansavilla, Altamaha River bank (Wayne Co.)

PAGE 182

Table 6-3. Summary descriptions of variability in petrographic paste categories. petrographic paste group sample size estimated silt very fine quartz mica sponge spicules phytoliths diatoms refired color temper group naa group comments relationship to clay samples A 31 3-5% 6% rare to occasional absent or rare absent or rare moderatehigh iron charcoal = 10 sand = 7 grog = 1 grit&sand = 5 grit = 8 1 (most charcoal, grog, sand) and 2 (most grit, grit&sand) silt ranges from 3% to 35%, a difference which is of debatable significance (grit samples with slightly higher silt); differs from B in absence or rarity of phytoliths possible affinity to clays 4, 5, 6 and 17 A-clay 4 1-5% 5-10% rare to none, but might vary none to rare none to rare moderate to high iron clay 4, 5, 6, 17 3, 4, UO potential for variability in silt, sand content and perhaps other constituents most similar to paste A; possible affinity to pastes B and C if deposits vary in frequency of phytoliths and sponge spicules B 16 3-5% 5% rare to occasional in most rare to occasional occasional to frequent low to moderatehigh iron charcoal = 2 sand = 1 grog = 1 grit&sand = 3 grit = 9 mostly 2 silt ranges from 3% to 35%, a difference which is of debatable significance; differ from A in terms of presence of noticeable phytoliths possible affinity to clays 5 and 6 if deposits vary in frequency of phytoliths C 4 3-5% 2% rare to occasional in most occasional to frequent variable low to moderate iron grit = 4 2 heterogeneous group in terms of phytoliths; differs from A and B in presence of noticeable sponge spicules and lower frequency of very fine sand possible affinity to clays 5 and 6 if deposits vary in frequency of sponge spicules D 13 3-10% variable 11% occasional to frequent absent to rare rare to absent low to moderatehigh iron charcoal = 3 sand = 9 grog = 1 1 relatively micaceous compared to groups A-C; variable silt content possible affinity to A clays 4, 6, 17 if deposits vary in mica frequency E 4 3-10% variable 5% occasional to frequent occasional to frequent variable 2 n=1 grog = 1 grit&sand = 2 grit = 1 variable similar to group D, but with higher sponge spicule content possible affinity with I clays 13, 18 or A clays if deposits vary in mica, spc and phytoliths F 1 3% 19% occasional to frequent occasional to frequent rare to o ccasional occasional grog = 1 1 similar to D, but with more very fine sand and occasional diatoms possible affinity to clay 18 in mica, spc, and phytolith frequency, but diatom species differernce precludes a match; might be more similar to F/G clays F/G clay 3 1-3% variable rare to occasional rare absent to rare occasional to frequent moderate to high iron clay 7, 10, 15 3, UO potential for variation in silt, very fine sand, mica and perhaps other constituents possible affinity to paste F in diatom species H clay 1 3-5% 9% rare 18% not observed moderate iron clay 3 3 spiculate clay, stockpiled no matches with any of the pottery samples I clay 2 variable variable frequent occasional to frequent variable extremely rare low to moderate clay 13, 18 4,5 mica frequency is higher than sherd with F paste similar to paste F, but diatom species difference precludes a match 182

PAGE 183

183 Table 6-4. Summary descrip tions of variability in gross temper categories. gross temper sample size Tempers matrix aplastics sand SSI.5/1 silt e/p vf quartz fine quartz medium quartz coarse quartz vc quartz Mica, sponge spicules, phytoliths Petrographic paste other constituents charcoaltemper 15 9% charcoal temper (grog present in 7 cases; bone temper in 3 cases) 60% 40% 26% 0.96/1.10 4% /3% 6% 16% 2% <1% mica rare to occasional in most cases; sponge spicules and phytoliths present/ rare in most cases A = 67% B = 13% D = 20% 2% polyxQ 1% feldspars <1% heavies grogtemper 5 2-4% crushed sherds (bone temper in one case) 55% 45% 36% 1.05/1.21 5% /3% 9% 17% 5% 1% variable mica, sponge spicules, and phytoliths A,B,D,E,F 2% polyxQ 2% feldspars 2% heavies sandtemper 17 quartz sand (grog temper rare in 3 cases; charcoal temper rare in one case) 58% 42% 36% 0.94/1.10 4% /4% 10% 20% 2% <1% <1% variable mica; sponge spicules and phytoliths absent or rare in most cases A = 41% B = 6% D = 53% 2% polyxQ 2% feldspars 1-2% heavies grit and sand 10 quartz sand and grit (grog temper rare in two cases) 56% 44% 39% 1.52/1.62 4% /3% 6% 14% 10% 4% <1% mica, sponge spicules rare to occasional in most cases; variable phytoliths A = 50% B = 30% E = 20% 2% polyxQ 1% feldspars 1-2% heavies grittemper 22 grit-sized quartz and quartzite (charcoal present in one case) 60% 40% 38% 1.97/2.02 4% /2% 3% 8% 11% 8% 2% mica rare to occasional in most cases; sponge spicules present/ rare in all but 6 cases; phytoliths occasional-frequent in half the cases A = 36% B = 41% C = 18% E = 5% 4% polyxQ 1% feldspars 1% heavies SSI.5/1 sand size index (with very fine grains counting as .5)/ (with very fine counting as 1) Silt e/p estimated percentage of silt grains/ actual percentage calculated from point counts vf very fine vc very coarse polyQ polycrystalline quartz or quartzite heavies sum of amphibole, epidote and UID minerals

PAGE 184

Table 6-5. Mineralogical paste categories by county and IN AA group. Paste Category County INAA G1 INAA G2 INAA unas Total A Duval 13 3 3 18 Nassau 3 1 2 6 Camden 1 1 Bryan 1 1 Glynn 1 1 McIntosh 3 3 Total 17 8 6 31 B Duval 2 2 4 Nassau 1 1 2 Camden 1 1 Glynn 2 2 McIntosh 5 2 7 Total 2 11 3 16 C Duval 1 1 2 Bryan 1 1 Wayne 1 1 Total 3 1 4 D Duval 9 1 10 Glynn 1 1 McIntosh Total 10 2 3 2 13 E Bryan 1 1 Glynn 1 1 McIntosh Total 1 1 1 1 2 2 4 F Duval 1 1 unasunassigned to any group in the INAA study G1 Group 1 G2Group 2 184

PAGE 185

Table 6-6. Paddle m atching sample s by INAA and petrographic groups. petid INAAid site temper petpaste INAA paddle # 2008-26 NJW-174 9MC372 GRIT A 2 34 2008-32 NJW-027 8DU68 GRIT B 2 34 2008-04 NJW-010 8DU68 GSAND A 2 36 2008-27 NJW-177 9MC372 GSAND B unas 36 2008-39 NJW-195 9MC360 GSAND B 2 36 2008-49 NJW-241 9MC16 GSAND B 2 36 2004-01 NJW-242 9MC16 GRIT A 2 38 2004-18 NJW-038 8DU96 GRIT A 2 38 2008-40 NJW-196 9MC360 GRIT B 2 38 2008-06 NJW-021 8DU68 SAND A 1 291 2008-21 NJW-152 8DU5544/5 SAND D 1 291 2008-25 NJW-169 8DU5544/5 SAND A 1 291 petid petrographic analysis indentification INAAid instrumental neutron activation analysis identification Petpaste petrographic (mineralogical) paste category INAA instrumental neutron activ ation analysis chemical group unas unassigned to any chemical group defined in the analysis Table 6-7. Percent quartz among chemical groups. Group 1 Group 2 Number 29 25 Mean 27% 35% Standard Deviation .078 .054 Minimum 9% 21% Maximum 38% 42% 185

PAGE 186

Table 6-8. R-squared value for th e linear regression m odel and co rrelation between elements or principle components and quartz proportion. Element or PC R-squared Correlation (r) As 0.127 -0.357 La 0.159 -0.398 Lu 0.1551 -0.3939 Nd 0.1278 -0.3575 Sm 0.1445 -0.3802 U 0.2903 -0.5388 Yb 0.1339 -0.3659 Ce 0.1435 -0.3788 Co 0.0477 -0.2184 Cr 0.34 -0.587 Cs 0.1497 -0.3869 Eu 0.2185 -0.4674 Fe 0.1761 -0.4196 Hf 0.0305 0.1746 Ni n/a* n/a* Rb 0.2384 -0.4883 Sb 0.217 -0.4658 Sc 0.4627 -0.6802 Sr n/a* n/a* Ta 0.0631 -0.2513 Tb 0.1574 -0.3967 Th 0.0837 -0.2894 Zn 0.2227 -0.4719 Zr 0.0075 0.0867 Al 0.3588 -0.599 Ba 0.063 -0.251 Ca 0.0164 -0.128 Dy 0.1827 -0.4274 K 0.2449 -0.4949 Mn 0.0014 -0.038 Na 0.114 -0.3376 Ti 0.0212 -0.1456 V 0.0929 -0.3048 PC1 0.1241 -0.3523 PC2 0.0358 -0.1892 PC3 0.1365 -0.3695 PC4 0.004 -0.0636 PC5 0.0026 0.0511 *too many missing values in chemical data 186

PAGE 187

A B C D Figure 6-1. Mineral constituents useful for dis tinguishing clay resource groups: A) phytolith, sample 2008-63 plane-polarized light 25X B) diatom, sample 2008-62, C) sponge spicules, sample 2008-58, D) absence of th ese natural constituents (with fine sand temper), sample 2005-24. All images plane-polarized light, 25X magnification, except D (cross-polarized li ght, 10X magnification) 187

PAGE 188

A B Figure 6-2. Group assign ments by quartz and A) chromium, B) cobalt. 188

PAGE 189

189 A B Figure 6-3. Group assignments by quartz and A) PC1, B) PC2.

PAGE 190

CHAP TER 7 THE FORM, TECHNOLOGY, AND FUNCTION OF SWIFT CREEK POTTERY Swift Creek pottery has already proven to be an excellent time marker for seriation and, through the reconstruction of designs a useful database of social interaction. However, these data tell us little about the ut ilitarian functions of vessels, wh ich are critical to understanding their roles in the social lives of people. Building on a respectable understanding of the temporality of the Swift Creek contexts in this study as well as the transport of vessels across the landscape, there were several qu estions that the analysis of the technological and functional aspects of pottery was designed to address: (1 ) How were vessels made? (2) What kinds of vessels (in a functional sense) were made? and (3) Where and fo r what purpose were the various kinds of vessels used? These ar e basic questions about how potte ry was employed in daily life, and as Skibo (1992a:4) relates, I view them as foundational to the ar chaeological pursuit of culture reconstruction. Having knowledge of the transport of vessels, the formal, technological, and functional variation of earthenware vessel s are clues to decipheri ng patterns of social interaction in Swift Creek contexts. This chapter begins with a review of two c oncepts that are critical to comprehending the diversity of Swift Creek pottery assemblages. First, the id ea of technofunction can be used to highlight the properties of material s and patterns of use, which in th e case of pottery can be used to define fairly specific functional parameters for each individual vessel form defined from archaeological assemblages. Sec ond, discussion of technofunction ne cessitates a review of style, defined simply as alternative ways of doing. Drawing on the idea of technological style, variations in the technological prope rties of pottery can reflect soci al identities in various ways and thereby be useful in delim iting patterns of interaction. 190

PAGE 191

Technofunctional data are presented in four sec tions. First, vessel form s are defined based on the sizes and shapes of whole or substantiall y reconstructed vessels. The function of these vessels is inferred through thei r morphological suitabili ty for particular tasks and corresponding frequencies of use alteration. These data help define functio nal categories for various vessel forms. Second, vessel size and use alteration fre quencies are discussed for all assemblages in order to identify vessel forms a nd potential functions even among small sherds. Third, rim thickness is compared across all assemblages, draw ing particular attention to the temporal and geographical trends of th is attribute. Fourth, the general paste characteristics of vessels are categorized and their distribu tion described. The chapter c oncludes with a synthesis and discussion of all technofunctional data and implica tions for understanding the social contexts of pottery production and use. Ultimately, the tech nofunctional data reflect conspicuous functional differences between mortuary mound and villag e midden pottery assemblages and stylistic differences between northern and southern site collections. In the midst of ceremonial and special-use vessels at Lower St. Johns mort uary mounds were buried nonlocal Swift Creek Complicated Stamped vessels with the technological style of Altamaha village wares. I argue that their use and deposition at mortuary mound s indicates a recontextu alization of nonlocal cooking vessels in which indexical qualitie s could be enacted in ceremony. Technofunction The analysis of pottery described in this ch apter follows in the tradition of what Rice (1987:310) calls technologi cal ceramic analysis or what othe rs have called the analysis of technofunction (e.g., Skibo 1992, term taken fr om Binford 1962, 1965). Recent years have seen a dramatic increase in the number of anal yses of ceramic function and use (Rice 1996:138), moving beyond the traditional culturehistorical utility of potsherds as time markers. Instead, pots can be considered as tools (Braun 1983), an d analysis is focused on the physical properties 191

PAGE 192

of materials and patterns of use. In this type of technological analysis an important analytical distinction must be made between data that can be used to infer vessel function and data that can be taken as direct evidence of use (Linton 1944; Rice 1996; Skibo 1992). To paraphrase Lintons (1944) oft-cited statement, we may not always be able to determine with certainty what a ceramic vessel was used for but we may often identify the functions for which it would have been well suited. Vessel attributes such as mo rphology (size and shape), paste composition, wall thickness, and surface treatment can all be manipul ated during manufacture so that the vessel is better suited for a particular func tion, but these data are not eviden ce of vessel use (Skibo 1992). The size and shape of vessels, particularly the openness of vessel profile, rim diameter, and volume, have proven to be general predicto rs of patterns of use in ethnographic studies (Smith 1988). A common drawback to these useful data in archaeological research is that large portions of vessels are necessar y, a situation not common in all archaeological contexts. This limitation can be somewhat ameliorated by using some of the many established techniques for estimating vessel shape and capac ity from sherds (Rice 1987:222-224). In contrast to vessel morphology, the analysis of paste composition is not limited by the small portions of vessels commonly found in archaeological contexts. Paste constituents affect performance characteristics during manufacture and use, includ ing clay workability, paste shrinkage, thermal shock resistance, impact and abrasion resistance heating effectiveness, and evaporative cooling effectiveness (Rice 1987:54-110, 226-232; Skibo 1992:36-37). A major focus of experimental studies has been the function of cooking pots as th ey relate to the physical properties of surface treatments and paste constituents (Schiffer 1990, Schiffer and Skibo 1987; Schiffer et al. 1994; Skibo et al. 1989). While knowledge of the physical properties of materials can help to infer the best function for vessels, two caveats should be noted. First, alterations in the physical 192

PAGE 193

properties of vessels can at once affect multiple performance characteristics and the researcher cannot always be sure which (if any) were th e impetus for adoption. For example, organic temper can be advantageous in the manufacturi ng process but results in a vessel that is less durable and less efficient in direct heat cooking than a sa nd-tempered alternative (Skibo 1992:37). Yet the disadvantages of organic temper as a conductor of heat cannot be used to predict that vessels were not used to cook over fi re, as demonstrated by use alteration data that indicate that some fibe r-tempered vessels were used for di rect-heat cooking in the Southeast (Sassaman 1993:144-148; Waggoner 2006). Second, alt hough the physical properties of vessels make them better suited for some tasks more th an others, ethnographic re search indicates that many vessels are multifunctional (Deal and Hagstrum 1994; DeBoer and Lathrap 1979; Skibo 1992). Even if an intended function can be inferred from the physical properties of vessels, impromptu or expedient uses cannot be de termined by these methods. In the strictest sense, evidence of use can be derived only from use alteration (Arthur 2002; Hally 1983, 1986; Skibo 1992). Use alterati on comes in two basic forms: deposits on vessels such as soot or food re sidue and vessel attrition such as abrasion or pitting (Skibo 1992). Sooting on vessel surfaces is commonly used to show that a vessel was used for cooking and the orientation of soot can indicate the position of a pot on the fire (Hally 1983; Linton 1944; Mills 1986; Skibo 1992). The residue of vessel contents can be identified using a variety of chemical techniques (Rice 1996:144-147). A limitation of using deposits on vessels as evidence of use is that not all uses of a vessel result in accretions. Alternatively, Skibo (1992a:40) argues that surface attrition has the potential to provide evidence for any type of pottery use. The most commonly used category of attrition is abrasion, which results from actions such as stirring, scraping, cutting, or beating during food preparation, consumption, or cleaning of vessels (Hally 193

PAGE 194

1983; Skibo 1992:40-41; Skibo et al. 1997). Another comm on fo cus of attrition studies is pitting, which can result from thermal shock, chemical corrosi on, or physical abrasion (Arthur 2002; Hally 1983; Jones 1989). I view as major inspiration for this analys is Hallys (1983, 1986) inferences about the function of Barnett phase Lamar pottery forms. In brief, Hally (1983, 1986) summarized the mechanical performance characteristics of vessels and compared these data to evidence of use in the form of sooting, oxidation discolora tion, and surface pitting. By comparing with ethnohistorical accounts of vessel use, Hally (19 86) was able to convincingly show how each of 13 Barnett phase vessel forms was used for various purposes. I have taken a similar approach to the analysis of Woodland period pottery, albeit with less reliance on ethnohistorical data. Attributes pertaining to vessel form were employed to infer tr ends in vessel function and evidence of use was defined by the occurrence of soot adhering to vessel surfaces and presence of drilled holes for mending and/or suspension. Considering Style The data recorded in this study are not onl y useful in delineati ng the technological and functional parameters of Swift Creek earthenware vessels but also in making inferences about the history and culture of pot produc tion and use. The consideration of style, in a broad sense, is germane to this endeavor. There have been myriad and competing expositions on the definitions and functions of style in material culture, especially pottery (reviews by Hegmon 1992; Rice 1996). Proponents of the ceramic sociology of the 1960s and 1970s, which inspired a surge of interest in style, held that sim ilarity in material culture was directly related to levels of social interaction and shared learning contexts (e.g., H ill 1970; Longacre 1970). This passive view of style was attacked through more active theories of style, most notably by proponents of the information-exchange theory of Wobst (1977). Wobst (1977) argued that style functions within 194

PAGE 195

cultures as a for m of communicatio n. Although a perspective of style as part of social strategies of identity and communication are now commonplace, several of W obsts (1977) original tenets that rested on functionalist and systems-theory approaches have been repeatedly criticized (Hegmon 1992:520). Among Wobsts ideas that have been disputed are his claim that stylistic messages will operate within the limits of optimal efficiency, that unambiguous, simple, and recurrent messages are most commonly communicat ed with style, and that stylistic messages will be found primarily in the most visible contexts. The most parsimonious perspectives on style acknowledge that learning and interaction (ceramic sociology) and information-exchange are not mutually exclusive theoretical perspe ctives. Indeed, information exchange only emphasizes one dimension of style, namely the mobilization of particular symbolic aspects of objects in some social contexts (Hegmon 1992). The learning/interaction and information-exchan ge debate in many ways coalesced in the disputes between Sackett (1985) and Wiessner (1 983). Sackett (1985) argues that style is the result of choices made from functionally equivale nt alternatives. This he calls isochrestic variation, which he suggests is learned or otherwise socially transmitted because options (e.g., styles) are dictated by the cr aft traditions within social groups. Sackett (1985) contrasts isochrestic style with iconological style, with Wiessners (1983) description of style as an example. Drawing largely on Wobsts (1977) idea s, Wiessner (1983) classifies style into two categories that relate to the conveyance of different kinds of information: emblemic style communicates information about group identity and social boundaries while assertive style relates to individual expression and identity. In essence, the arguments proffered by Sackett (1985) and Wiessner (1983) are debates about th e passive versus active nature of style in material culture. 195

PAGE 196

In an overview of the subject, Rice (2005a: 152-153) lam ents that many debates on style are merely reworking half-century old debate s about typology, whether types are heuristic abstractions created by the resear cher or inherent to the artifacts (e.g., Ford 1954; Spaulding 1954). Yet the more recent debates about styl e, as clarified by Wiessner (1985, 1989) and Sackett (1986, 1990), are not over wh ether all style is best define d as emblemic, assertive, or isochrestic variation, but rather which of these coexisting kinds of style pertain to specific contexts. It is the specificities of local context and particular histories that should ultimately guide studies of style (Carr and Nietzel 1995:4; Shanks and Tilley (1987:148-149). A theoretical concept that posit s the importance of social cont ext in both technological and stylistic choices, is technological style (L echtman 1977; Lemmonier 1986, 1993; Stark 1998). According to proponents of technological style, technology itself has a style because the activities which produc e the artifacts are stylistic (Lechtm an and Merrill 1977:5). Thus, it is differences in social practice that reproduce differences in material culture that can be recognized as style. A useful way to examine culturally-embedded systems of production is through analysis of the ope rational sequence or chaine operatoire (Lemmonier 1993; Leroi-Gourhan 1943, 1945). Technological style is essentially alternative ways of doing (cf. Sacketts (1985) isochrestic variation), and each step of the ope rational sequence of production can be examined in an attempt to determine or infer reasons for particular choices. Alt hough the chaine operatoire is now often used in archaeology as simply a method to detail past technological strategies, sequences, and practices, a more holistic perspect ive comes from the heritage of Mauss (1935) enchainement organique that links the simultane ous becoming of artifacts and technical agents (Dobres 2000:155-156). In his Techniques du corps, Mauss (1935) descri bes how technological practice is socially embedded to the extent that making and using material culture is a process 196

PAGE 197

that con stitutes agents as meaningful members of society, especially th rough everyday acts of bodily comportment that are conditioned through social tradition. Thus, the structural features of technological systems are reflections of, and indeed reproduce, the st ructure of social relations. Through the legacy of Mauss (1935), the con cept technological style and the chaine operatoire method approach the idea of habitus, especially in terms of bodily praxis and the reproduction of social structur e (e.g., Bourdieu 1977). Style is a sort of technological performance of social (re)produc tion that need not carry obser vable or communicable meaning, although it certainly can serve this purpose (Lechtman 1977; Lemonnier 1993). In the concept of technological style, analysis is focused on how production fits into broader social and symbolic systems and cognitive structure is often privileged: different technological styles may represent different mental processes that underlay and direct [persons] actions on the material world [and] are embedded in a broader, symbolic system (Lemonnier 1993:3). The cognitive focus of some applications of th e technological style concep t in itself fails to explain all aspects of material variation but I find it to be a useful heuristic in examining differences within and among pottery assemb lages. The fundamental contribution of technological style is in the rec ognition that all aspects of technologies are at least in part cultural constructions and thus social a nd historical context and continge ncy are critical parameters in defining style. Importantly, each step of the manufacturing sequence, or chaine operatoire, is potentially situated differently within social st ructure so that similar geographical or temporal patterns in two attributes of a cl ass of material culture may refl ect very different phenomena. For example, some steps in the production sequence of pottery ar e perhaps more likely than others to be the reflection of distinct learning environments and fields of interaction. Those parts of the manufacturing sequenc e that depend on ingrained moto r habits are more likely to 197

PAGE 198

preserv e the routines of a potters initial training during her or his formative years. In particular, the shaping stage of vessel manuf acture has proven to be particul arly conservative and resistant to change in a way that corresponds to social identities in many et hnographic studies (Arnold 1998:358; Reina and Hill 1978:230; Ri ce 1984; van der Leeuw et al. 1992). Alternatively, other steps in the manufacturing sequence such as clay processing, firing, and applying surface treatments (i.e. decoration) can be more delib erately acquired by potters for various social reasons and are therefore more sensitive to ch ange (Gosselain 2000:191-193). In sum, through changes in residence in the course of marriage alliance or migration, the vessel shaping process is most likely to preserve the signatures of a potters initial learning environment while other attributes of pottery might be easily changed to accommodate new social contexts. Thus, there may be both habitual and purposef ul (i.e. conscious) aspects of ceramic technological style that can l eave archaeological residues. So me manufacturing steps reflect learning environments (as proposed by ceramic sociologists) while other attributes might be used in active messaging or become identity markers in certain contexts. The parameters that influence and inhibit a potters technical choi ces are important consid erations in studying interactive contexts in which al ternative views of the correct socially-accepted way to make and use pottery might collide. For example, among Swift Creek village assemblages, the mobility of potters through marriage alliances might be reflected in particular attributes of pottery that correspond with processes relying he avily on ingrained motor skills such as vessel forming techniques. Indeed, the work of outsiders might be identifi ed by looking for subtle anomalies in how clay coils were formed and bonded or how vessels were scraped and trimmed to achieve their final shape. 198

PAGE 199

Methods Inferences about vessel form were a central part of the pottery analysis. Consequently, the vessel, as opposed to the sherd, was the unit of an alysis chosen for this study. A vessel unit of analysis has the advantage of not only better approximating th e number of vessels in an assemblage (e.g., minimum number of vesselsMNV) but also inferring the relative frequencies of vessel forms. Determining a minimum number of vessels (MNV) is analogous to zooarchaeological techniques for determining mi nimum number of individuals (MNI) for faunal specimens (Rice 1987:292). In the present st udy, an attempt to identify a MNV for each assemblage began by isolating all rim sherds. Th ese were combined or se parated into numbered vessel lots representing individual vessels accord ing to similarities or differences in lip form, rim form, surface treatment, and paste composition. Using surface treatment and paste composition only, each body and basal sherd was in ferred to be either part of an existing numbered vessel lot or a new vesse l lot. Every attempt was made to mend individual sherds that appeared similar based on the above criteria. In this way, vessel numbers were eventually assigned to individual sherds or groups of sherds that were in ferred to be representative of individual vessels. The technique described above usually undere stimates the number of vessels in an assemblage (Rice 1987:292), probably among some types of sherds more than others. Body sherds, having fewer recorded attributes than rim sherds (lip form, lip fold depth, orifice diameter, rim form, etc), are more susceptible to being incorrectly gr ouped into too few vessel lots. However, the uniqueness of each complicated stamped design on Swift Creek sherds allowed for better than normal precision in differentiating body sherds from separate vessels. Sand-tempered plain body sherds, with only pa ste composition and sometimes surface smoothing to differentiate them, were the most difficult to assign to individual vessel lots and were therefore 199

PAGE 200

underrepresented according to MNV to a greater degree than other types of sherds. In all, 1222 vessels from 30 sites were identified and analyzed. Recorded vessel form data included orifice diameter, rim thickness, vessel height, lip form, rim form, lip fold depth, and basal form. In addition, rim profiles were drawn for all rims greater than 3 cm in length. For the majority of the vessel lots with rims, orifice diameter was calculated using a rim diameter template. Vessel s exhibiting less than 5% of the total orifice were generally considered to give unreliable es timates of total diameter. For comparability, rim thickness was recorded at the same point on each specimen, exactly 3 cm below the lip of the vessel. Accurate vessel capacity measurements we re impractical due to the fragmentary nature of most vessels. Among the rec onstructed vessels, volume was estimated using a version of the summed cylinders method, in which one cm hi gh cylinders were stacked inside a vessel profile and then the volume (V= rh) of each was added together (Rice 1987:222). Surface treatments (impressed, incised, painted, punctated, scraped, slipped, smoothed, stamped) were recorded for both exterior and in terior portions of each vessel. Recorded use alteration data included both the relative amount and location of soot adhering to vessel surfaces and the number of perforations through vessel walls, both susp ension and mend holes. Using fresh breaks at sherd edges and a binocular ( 70x) microscope, the paste of each vessel was characterized according to the rela tive frequencies, size (Wentworth scale), and shape of aplastic inclusions. Vessel Forms Willey (1949:496-506) provides an overview of ve ssel forms from the Florida Gulf Coast, many from Weeden Island a nd Swift Creek contexts. With some variations discussed below, the vessel forms outlined by Willey (1949) are largel y similar to the contemporaneous forms found 200

PAGE 201

on the Atlan tic coast. I have used these Gulf Coast descriptions as a comparative guide in differentiating vessel forms. Determining the range of vessel forms is di fficult without large por tions of vessels and large sample sizes. Fortunately, mortuary mound contexts that contain whole and reconstructable vessels can pr ovide the framework through whic h to infer vessel dimensions from collections of smaller sherds. Based on the 102 substantially w hole or reconstructed vessels from mortuary mounds on the lower St. Jo hns River, at least 15 vessel forms can be differentiated. In addition, seven vessels from village middens can be definitively assigned to one of these forms. Although the remaining ve ssels from both mortuary mounds and middens cannot be distinguished as specific forms, ri m profiles reveal that standard cooking forms predominate, including open pots and bowls and restricted pots and bowls. Outlined in this section and summarized in Table 7-1, Table 72, and Table 7-3 are the morphological and functional attributes of each of the vessel forms identified from the analyzed mortuary assemblages and the few definitive specimens from middens. Open Bowl The open bowl form is not defined by Willey (1949), but only differs from his simple bowl form in generally having greater vessel height The widest diameter of the vessel, always at the mouth in this form, is greater than the ve ssels height. The tallest of these specimens are morphologically similar to the open pot form. Base s are rounded or, in th e shallowest versions, flattened. There are nine complete examples of this vessel form from mounds on the Lower St. Johns River. The form is somewhat difficult to identify among collections of small sherds because a steep sided open bowl is indistinguishabl e from an open pot without more than half of the vessel wall present. However, the dram atic outward slope of vessel walls among the shallowest examples can be identified usi ng smaller rim sherds. Among collections from 201

PAGE 202

m iddens, five vessels were identified as havi ng this form. Due to the difficulties of identification, the actual frequency of open bowls among the vessels deposited in middens is assumed to be much higher. Open bowls were made in a range of sizes, fr om medium to very large (Figure 7-1). The orifice diameter of open bowls ranges from 17 cm to 40.5 cm with a mean of 26 cm. Among the nine substantially complete vessels, height ranges from 13 cm to 27.5 cm with a mean of 18.75 cm. This vessel form is not correlated with a particular temporal range or culture historical type. The total 14 vessels of this form include charcoal-tempered plain (n=3), Early Swift Creek Complicated Stamped (charcoal-tempered) (n=1), Dunns Creek Red (n=1), sand-tempered plain (n=7), St. Johns Plain (n=1), and Late Sw ift Creek Complicated Stamped (n=1). Nearly half (n=6) of open bowls have soot adhering to exte rior surfaces, thus demonstrating that this vessel form was commo nly used for cooking over fire. Four of the vessels have mend holes while another vessel has suspension holes drilled into the rim. Open bowls provided an alternative cooking container th at was generally stable without extra supports (particularly the flatter and shallower examples) and in which contents were easily accessed and manipulated but consequently also easily spil led (cf. Hally 1986:283). Given these features, these vessels may have been used to cook food that needed frequent turning or stirring and that could be shared straight out of the vesse l after preparation. Restricted Bowls The restricted bowl form is la rgely similar to the open bowl fo rm but differs in exhibiting its greatest width below the lip. This vessel fo rm is not defined by Willey (1949) but is a variation of his simple bowl. As with all bowls, the widest di ameter of the vessel is greater than the vessels height. The tallest of these sp ecimens approaches the form of restricted pots. There are only four complete specimens from mo unds on the Lower St. Johns River. Restricted 202

PAGE 203

bowls could be definitively id en tified only once within the mi dden assemblages owing to the lack of large portions of vessels. Indeed, the upper portions of restricted bowls near the rim are identical to restricted pots and thus small sherds prevent their differentiation into specific vessel forms. The limited sample of restricted bowls shows a large array of sizes with orifice diameters ranging from 12 cm to 30 cm with a mean of 20 cm (Figure 7-2). Four of the vessels are simply smoothed on exterior surfaces, three sand-temp ered and one charcoal-tempered. The fifth specimen is a painted Weeden Island Red vessel. The two largest vessels of this form were used for cooking over fire and have soot adhering to thei r rim exteriors. The three smaller versions of this vessel form have no evidence of soot. Thus there are at least two sizes of restricted bowl that may each correspond with different functions : the larger was for cooking while the smaller served some non-cooking function. The orifices of these vessels are not conducive to sealing, making a storage function unlikely. While the slightly restricted orifice of these bowls served to reduce heat loss and spillage of contents (c f. Hally 1986:288-289), whether these vessels were employed differently than open bowls is unclear. Restricted Pots Willey (1949:498-502) makes a distinction between simple jars and pots. This distinction pertains to the degr ee of restriction at the orifice and the location of the maximum width of the vessel. However, Willey (1949) define s no measurable attributes to differentiate the two forms. Given the rather dubious definitiona l parameters, the uncertain functional differences between jars and pots in this co ntext, and the small number of ostensible jar forms in the study collections, all restricted vessels th at are taller than they are wide are included as restricted pots here. 203

PAGE 204

There are a total of 16 restri cted pots from mound assemb lages on the Lower St. Johns River (Figure 7-3; Figure 7-4). This vessel form also appears to be common in the sherd assemblages from middens but cannot be confid ently differentiated from restricted bowls without the majority of a vessel wall present. The restricted pots from mounds range from 12.25 cm to 28 cm in orifice diameter with a mean of 20.7 cm. They include charcoal-tempered plain (n=4), sand-tempered plain (n=7), St. Johns Pl ain (n=1), and Swift Creek Complicated Stamped (n=2), two of which are definitively Late Swift Creek. Restricted pots were primarily used as cooking vessels, with 11 (69%) from the mound samp le having soot adhering to exterior rims or vessel walls. Only two (13%) have mend holes demonstrating that these vessels were not commonly repaired. These slightly restricted vessels and larg ely similar open pots (discussed below) were apparently very common cooking pot forms dur ing the Woodland period in general and among many Swift Creek groups in par ticular (Sears 1962; Willey 1949:379). The smooth transition between base, wall, and rim provide an excel lent conductive surface for cooking food and lend a high thermal shock resistance for cooking over fire. The restricted orifice varies only slightly from the morphology of open pots and in fact the l east restricted versions grade into open pot forms. A slightly restricted orifice may have reduced spillage and h eat loss but only to a negligible degree. Restricted pots may merely be a morphological variation of open pots that served the same cooking functions. Open Pots The open pot form is similar to the restricted po t but with an open, unrestricted orifice (Figure 7-5). There are 12 ve ssels represented in the mound assemblages although open pots are assumed to be very numerous in both mound and midden assemblages. As with restricted pots and bowls, open pots cannot be differentiated fr om open bowls without major portions of the 204

PAGE 205

vessel wall. The m ound sample of open pots is sim ilar to restricted pots in orifice diameter size, ranging from 12 cm to 26 cm with a mean of 16.9 cm. These vessels include many pottery types, including Early Swift Creek Complicated Stampe d (charcoal-tempered; n=4), charcoal-tempered plain (n=2), sand-tempered plain (n=1), Dept ford Simple Stamped (n=1), Mayport Dentate Stamped (n=2), Swift Creek Complicated Stam ped (n=1), and Weeden Island Red (n=1). The open pot is quite possibly the quintessential cooking form among Swift Creek populations. All but one (92%) of these is so oted in the mound sample. The only unsooted vessel is of the type Weeden Island Red, which is nearly exclusive to mortuary contexts in northeastern Florida and appears to have been rarely used in cooking. Only one (8%) open pot in the sample contains mend holes, showing that th ese mostly utilitarian pots were not commonly repaired. Flattened-Globular Bowls Willey (1949:496-498) defines this vessel form as a medium-deep to deep bowl with maximum diameter at about midpoint of [the] ve ssel and with inturned sides and constricted orifice. Flattened-globular bowls are somewhat variable in their appearance owing to a variety of heights that range from half to nearly equal the maximum diameter of the vessel (Willey 1949:498). However, the defining ch aracteristic of this vessel form is the sharply incurvate vessel wall near the rim. The tallest of thes e specimens grade into Willeys (1949:498) simple jar form. On the Lower St. Johns River, flattened-globular bowls were made in a variety of sizes with orifice diameters ranging from 7 cm to 17.5 cm (mean 11.7 cm). However, the mean vessel size in this case closely mirrors the mode: more than half (n=5 ) of the nine specimens have orifice diameters between 11 cm and 12 cm. This vessel form and particularly this modal size appears to have been used for cooking over fire at least some of the time as evidenced by soot 205

PAGE 206

adhering to the exterior s urfaces of three of the nine (33%) ve ssels. Interestingly, the three sooted vessels are all of the moderate modal size (vessels LG27, D7, D12; Figure 7-6). The smaller vessels that are not sooted were pres umably not used for cooking. Flattened-globular bowls of all sizes appear not to have been extensively repair ed; only one of the nine (11%) vessels contains mend holes. Flattened-globular bowls are str ongly correlated with Weeden Is land series pottery. Nearly half of the sampled vessels are Weeden Island Red (n=2) or Weeden Island Incised (n=2) and another is Crystal River Incised. The remain ing vessels are sand-tempered plain (n=3) and charcoal-tempered plain (n=1). Thus, the idea of fla ttened-globular bowls, and/or the vessels themselves, are likely to have come from the west along with other items of early Weeden Island and Crystal River material culture. Indeed, th e flattened-globular bowl is a form quite common to Weeden Island contexts but rare in cont emporaneous cultures throughout the Eastern Woodlands (Willey 1945:249). These vessel forms are notably absent from non-mortuary contexts in northeastern Florida regardless of their comparatively facile identification with small sherds. Only one of the nine vessels comes fr om a non-mortuary context and it too is Weeden Island Incised. Moreover, there may have been a functional difference between the diagnostic Weeden Island types and wares that were not painte d or incised. All three of the sooted vessels of this form are plain in surface treatment. Thus, flattened globular bowls we re ceremonial wares that seem to have been reserved for use and deposition in mortuary contexts. Th is is one of several vessel forms found in northeastern Florida that substantiates a sort of sacred and s ecular dichotomy akin to Sears (1973) description of Weeden Is land contexts in western Florid a. Although neither Sears (1973) nor Milanich et al. (1997:127) found globular bowls to be rest ricted to mortuary contexts 206

PAGE 207

within and among W eeden Island sites, populations in northeastern Florida limited their deposition, and presumably their use, mostly to mortuary mounds. Collared Jar Willey (1949:498-502) divides collared jars into two categories: long collared jars in which the length of the vessel neck is between 1/3 and 2/3 of the total vessel height, and short collared jars in which the vessel neck comprises between 1/4 and 1/5 of the vessel height. The Lower St. Johns River specimens (n=7) do not conform to these parameters, with some necks much longer than 2/3 of the total height and others definitively between the length ranges specified by Willey (1949). Therefore, all vessels with collars are included in the same category here (Figure 7-7). Further, a single doubleglobed jar (e.g., Willeys terminology 1949:503) with a distinct collar is also in cluded in this category due to its overall morphological similarity. The morphology and overall size of collare d jars is variable on the Lower St. Johns River but orifice diameters are consistently simila r across vessels, with si x of the seven vessels ranging between 9 cm and 11.5 cm. A single short co llared vessel with an or ifice diameter of 18 cm is an outlier in this limited sample. While the collared jars grouped together here show a wide range of morphological vari ability, they are also unified by common surface treatments. Five of the vessels are Late Swift Creek Complic ated Stamped and the remaining two are Crystal River Incised. All are from mortuary mound contexts. Collared jars appear to have been mortuary-s pecific wares that were an important part of the Swift Creek ceremonial complex that manifest ed in northeastern Fl orida. These vessels appear uniformly to have been expertly construc ted and carefully used an d repaired. Five (71%) of the vessels contain mend holes. Some collared jars were used over fire as evidenced by soot adhering to the shoulders of two (29%) of the ve ssels. The long collared jars with restricted midsection and flaring orifice would have been well-suited for quickly heating liquids and 207

PAGE 208

transporting and pouring these cont ents without spilling. Given the m ortuary-specific recovery context for all of these vessels, their function is li kely to have been in preparing and distributing ceremonial beverages. Several ve ssels have a thick angular rim fold suitable for tying covers across vessel openings for secure transport (Figure 7-7b). There is a single example of a collared bowl form that may be differentiated from the collared jars by having an overall width that is greater than maximum height. The vessel approximates the shape of shorter co llared jar. This vessel is also from a mortuary context and is likely to have served functions similar to collared jars. Small Cups and Bowls Small cups and bowls take the forms of larger pots and bowls, respectfully, but differ in their diminutive size. Willey (1949:506) defines no such category but does mention miniature vessels which he describes as very small vessels most of which are rath er carelessly made. Willey (1949:506) also estimates that these miniature vessels mostly conform to the morphological categories of larger vessels. While there are miniature ve ssels in the Lower St. Johns mound assemblages that might fit Willey s definition (see below), the cups and bowls described here do not conform to this descript ion, being neither very small nor carelessly made. Rather than being miniature versions of large functional vessels, small cups and bowls are likely to have served a pa rticular function, namely to consume individual portions of food and drink and as containers for mortuary offerings. There are 15 whole or nearly whole small cups and bowls among the mound assemblages from the Lower St. Johns River (Figure 7-8). As with other vessel forms, a considerable portion of the vessel wall must be present in order to di fferentiate between vessel forms, with small rim sherds potentially conflati ng identifications of small cups, bowls, and jars with beakers, bottles, and multi-compartment trays. Consequently, sma ll cups and bowls are not specifically defined 208

PAGE 209

for any m idden assemblages, but the extremely low relative frequency of small orifice diameters among midden assemblages indicates their rarity. The orifice diameters of small cups and bowls range mostly from 5.2 cm to 13 cm (mean 10.1 cm) with one outlier shallow bowl with an orifice diameter of 16.5 cm. Several of these vessels ha ve slightly irregular (i .e. non-circular) openings but do not approach the truly el ongated shape of the boat shaped bowl described below. The small cup and bowl assemblage includes a wide range of pottery types: Early Swift Creek Complicated Stamped (charcoal-tempered; n=1), charcoal-tempered plain (n=3), Dunns Creek Red (n=2), non-diagnostic punctate d and incised (n=1), sand-tempered plain (n=4), St. Johns Plain (n=1), Swift Creek Complicated Stamped (n=2), and Weeden Island Incised (n=1). Therefore, these vessels include both cer emonial pottery types (e.g., Dunns Creek Red, Weeden Island series, Crystal River series) that seem to be mostly restricted to mortuary sites as well as common types found in village middens. Notably, only one (7%) vessel exhibits a small am ount of soot adhering to the exterior of the rim, demonstrating that small cups and bowls usually did not se rve a cooking function. These vessels were also only infrequently repa ired, with mend holes present on only one (7%) vessel. Small Jars Small jars are diminutive vessels with restrict ed orifices and greater height than width (Figure 7-9). The larger of these vessels are si milar in proportion to the small cups and bowls and may have served similar functions. In cont rast, the smallest jars often appear to be diminutive versions of larger vessels in othe r morphological categories, what Willey (1949:506) refers to as miniature vessels. For instance, an incised vessel from the Low Grant Mound E, only 4 cm tall and less than 2 cm wide, takes the shape of the larger collared jar form described previously and seems to mimic either the zone d decoration of Swift Creek Complicated Stamped 209

PAGE 210

or the incising of Crystal River or Santa Rosa-Swi ft Creek series vess els (Figure 7-8). Likewise, a Crystal River Incised jar, just over 10 cm tall, is a miniature version of Willeys (1949:498) jar with lobes from an early Weeden Island mound in Bay County (8BY14) (Moore 1902:145; Willey 1949:239-240). There are eleven small jars from Lower St. Johns River mound assemblages, the largest three of which resemble cups and the remaining eight which can be considered miniature vessels. The orifice diameter of small jars ra nges between 1.8 cm and 13 cm with a mean of 5.9 cm. The vessels are overwhelmingly decorated, with only three (27%) having entirely smoothed exterior surfaces. According to typology, the sm all jar assemblage includes Basin Bayou Incised (n=1), Carrabelle Punctated (n=1), charcoal-tempe red incised (n=1), grog-tempered plain (n=1), Crystal River Incised (n=2), sand-tempered plai n (n=2), Late Swift Creek Complicated Stamped (n=2), and sand-tempered incised (nondiagnostic; n=1). Small jars were very rarely used for c ooking; only one (9%) vessel contains traces of soot. While the three larger jars are well-suited for individual servings and offerings like similar small cups and bowls, the remaining miniature vessels may have had an altogether different function. There are no mend holes in any of the sm all jars, but suspension holes near the rim are present in four (36%) of the miniature jars. Thus, these miniature jars were commonly suspended, perhaps with cordage, for easy carr ying and display. Many of these vessels were small enough to have been attached to the body as ornamentation. As such, they may have contained materials important for ceremonial even ts. These vessels can be easily identified in midden assemblages by their very small orifice diam eters but are very rare or absent in these contexts. None are definitively identified in the midden sa mples of this study. 210

PAGE 211

Boat-Shaped Bo wl Willey (1949:498) defines boat-shaped bowls as medium-deep or shallow with oval or ovate-rectangular form. There are five of these in the Lower St. Johns River mound assemblages (Figure 7-10). These were made in at least two distinct sizes. The larger of the two (n=3) has orifice diameters between 13.2 cm and 18 cm on the long axis and between 6.5 cm and 11 cm on the short axis. The smaller of the tw o vessel sizes (n=2) is comprised of diminutive versions of the larger, with orifice diameters ra nging from 7 cm to 8 cm on the long axis and 4 cm to 4.5 cm on the short axis. Three of the vessels, including the tw o miniature vessels, are Dunns Creek Red while the remain ing two are charcoal-tempered pl ain. Only the largest vessel exhibits soot on the exterior rim. None of the specimens contain mend holes but the smallest vessel has two suspension holes. Boat-shaped bowls served mostly non-cooking functions and may have been used in ways similar to small cups and bowls. Likewise, the sma llest of these vessels could have been used as ornaments worn on the body or otherw ise displayed. The use of this vessel form appears to have been limited to mortuary contexts, with none id entified from non-mortuary sites and a strong association with the Dunns Creek Red type that is limited mostly to mortuary mounds. Willey (1949:558) describes boat-shaped bowls as a distinctive feature of the St. Johns River region, where this vessel form enjoye d the greatest popularity. However, Willey (1949:558) also suggests that the idea for this ve ssel form may have come from the Gulf Coast, where a few such vessels have been recovered from Deptford and Early Swift Creek sites. Regardless of whether eastern Florida was mostly on the receiving end in the exchange of ideas and material culture (Willey 1949:562), all five of the vessels in the present study are presumed to have been produced either on the Lower St. J ohns (charcoal-tempered) or the Middle St. Johns (Dunns Creek Red) based on thei r typological designations. 211

PAGE 212

Double Bo wl The double bowl vessel form, as depicted by Willey (1949:499), is made up of two conjoined bowls separated by thic k compartment walls (Figure 7-11). There are three of these vessels from mound assemblages and they are unifo rmly small, having orifice diameters between 4.4 cm and 5.7 cm for each bowl. The asse mblage of double bowls includes one Weeden Island Red and two sand-tempered plain vessels. None are sooted or have mend holes. The three double bowls in the sample include a variety of unique attributes. One rather amorphous and poorly smoothed vessel has small flattened lobes described as handles by Moore (1894) that protrude from the base of th e vessel on either side. Another vessel has four pointed protrusions symmetrically placed on the exteri or rim. The breakage pattern of this vessel is indicative of the manufacturi ng technique, with the bowls simply coming apart in the area they were conjoined. The two bowls appear to have b een made separately and later joined together while still wet and plastic. Willey (1949:410-411) describes the double bowl as a Weeden Island series vessel form, comprised of Weeden Island Plain wares on the Gulf Coast. However, this temporal and cultural designation is apparently based on C.B. Moor es work on 56 pure Weeden Island sites in which only two of 248 vessels (less than 1%) we re double bowls. In the present sample, one Weeden Island Red vessel supports a W eeden Island series affiliation. Double bowls have not been found at mi dden sites on the Atla ntic coast and are presumably specialized mortuary wares serving a function similar to multi-compartment trays. However, these vessels are notably different in execution, being less uniform in shape and exhibiting less smoothing and polishing than multi-compartment vessels. 212

PAGE 213

Multi-Compartment T ray The multi-compartment tray comes in a variety of shapes and sizes (Figure 7-12). Some vessels are similar in form to double bowls but w ith three, rather than two, bowls conjoined. Another form with three compartments is much more reduced in height and is best described as a tray. A third variety, which Willey (1949:502) describes as common, consists of two or more low compartments and a single larger bowl in a raised position above the rest. The five multi-compartment vessels in the sample exhibit a variety of sizes, with individual compartments ranging from 4.5 cm to 12 cm (7.6 cm mean) in diam eter. No soot or mend holes are observed in the sample. As with double bowls, Willey (1949:410-411) presum es that the multi-compartment tray is a Weeden Island vessel form. Based on the same work by C.B. Moore at 56 pure Weeden Island sites, Willey (1945) lists 11 specimens (4.4 %) taking this form. One Weeden Island Red vessel in the Lower St. Johns study collection corroborates this susp ected Weeden Island affiliation. The remainder of the assemblage is comprised of one charcoal-tempered plain and three sand-tempered plain vessels. Multi-compartment vessels were specialized cere monial containers apparently designed to hold and separate important substances. At the Dent Mound, a partial multi-compartment tray was discovered with different materials in each of the two extant compartments. Found in situ, smeared on the interior of the compartments and in whole pieces within, was red ochre (hematite) in one compartment and yellow ochre (limonite) in another. Presumably a third substance would have been contained within th e missing compartment. These substances may have been important ritual items, perhaps for pa inting the bodies of the living and the dead in mortuary ceremony. 213

PAGE 214

Beakers The defining characteristics of beakers are st raight walls and flat bases with a definite angle between the walls and base (Willey 1949:500). The three examples from Lower St. Johns River mounds conform specifically to Willeys (1949:500) cylindrical beaker form (Figure 713). Two of the beakers are near ly identical in size, 18 cm tall and between 8.2 cm and 9 cm in orifice diameter. The third vessel is smaller, 15 cm tall with an orifice diameter of 6.75 cm. Two vessels are sand-tempered plain and one is Early Swift Creek Complicated Stamped (charcoal-tempered). None were id entified in the midden assemblages. There is neither soot nor mend holes in any of the specimens. Beakers were probably not used for cooking nor were they suitable for it, w ith sharp profile angles that would have been particularly susceptible to thermal shock and br eakage. Because they were not practical cooking containers and only fluid contents could be easily removed, beakers are likely to have served as drinking vessels. Bottle There is one example of a bottle vessel form. This St. Johns Plain vessel has an elongated neck with an orifice diameter (4.5 cm) that is markedly smaller than the maximum diameter of the vessel (18 cm). There is no s oot or mend holes, but two suspension holes were drilled into the rim. With a small orifice that is easily closed, this vess el would have been wellsuited for carrying liquids. Shallow Bowl There is one shallow bowl from the mound a ssemblages, with a diameter of 17 cm and a height of 4 cm. The specimen is sand-tempered plain. No other information can be garnered because a thick lacquer applied to all surfaces has unfortunately ob literated the evidence. This vessel form was popular among Weeden Island culture s to the west. For example, shallow bowls 214

PAGE 215

and plates are fairly com mon within the village and the mounds at the Mc Keithen site (Milanich et al. 1997). This vessel form would have been suited for holding only solid foods or other nonliquid items. Double-Globed Jar There is one example of a double-globed jar, which Willey (1949:413) defines as a ceremonial form found at Weeden Island burial m ounds along the Gulf Coast. However, in the case of the Lower St. Johns River specimen, th e surface treatments more closely resemble Crystal River Incised (Figure 7-14). The vessel is 30 cm in height, 18 cm in maximum width, and has a highly restricted orifice of 8 cm. The exceptional workmanship reflected in this vessel caused Moore (1895:491) to describe it as by far the finest specimen of earthenware recovered by us from any Florida mound. Whatever its speci fic ceremonial function, this vessel was used directly over fire as demonstrated by soot a dhering to exterior portions of the rim and vessel wall. Vessel Morphology Summary The fifteen vessel forms outlined in this chapter can be grouped into three categories: cooking vessels found in both mound and domestic contexts, vessels used occasionally for cooking and present only at mounds, and non-cooking vessel forms restricted almost entirely to mounds. Cooking vessel forms common to both ceremoni al and domestic contexts include open pots and bowls and restricted pots and bowls. Co mbined, 64% of these vessels (30 of 47) show definitive evidence of coming into contact with fire. These vessels are moderate to large in size and were most likely used for boiling meat and vegetable foods and in some contexts, perhaps, beverages such as black dri nk. The morphological variations among these cooking vessel forms may correspond to practical considerations of content accessibility during cooking, serving, or 215

PAGE 216

eating and stability on the ground without external supports. Indeed, the use of pots versus bowls m ay have varied: soot is more frequent among pots (ca. 80%) than bowls (ca. 40%) while conversely, bowls (ca. 30%) have a higher freq uency of mend holes than pots (ca. 10%). Whether these attributes pertain to the nature of the food or dr ink being prepared or to the conventions of use in particular cultural contexts is a questio n addressed below with closer comparison of mound and village assemblages. The remaining vessel forms appear to be mainly restricted to mound assemblages. Of these, the flattened-globular bowl, collared jar, and double-globed jar assemblages show at least occasional cooking functions. Combined, just over one -third (6 of 17) of th ese vessels have soot adhering to exterior rims or walls. The morphology of all of these vessels would have limited the type of cuisine that could have been prepared. In particular, a highly restricted orifice would have prevented access to conten ts during cooking tasks such as stirring that might prevent burning meats and vegetables during boiling. Altern atively, teas can be bo iled without risk of charring and a restricted orifice makes heating fast er and more efficient. Consequently, some of these vessels were probably used for preparation of ceremonial beverages such as black drink. In fact, a vessel form resembling the flattened globu lar jar was still used to prepare black drink during early colonial times in St. Augustine according to Father Francisco Ximnezs 1615 description (Sturtevant 1979: 150-151). During the Woodland period, collared jars may have been used for black drink as well, and because of their reduced ability to withstand thermal shock due to somewhat angular profiles, they were especially susceptibl e to cracks. The high frequency of mend holes among this vessel form (70 %) not only indicates th eir fragility but also the cultural importance of keeping them in use. Also notable is the correlation between the flattened-globular bowl, collare d jar, and double-globed jar forms and the Crystal River Incised 216

PAGE 217

type (n=4 ). Willey (1949:389) lists these vesse l forms, along with cylindrical beakers and composite-silhouette jars, as the most common ve ssel forms for Crystal River Incised along the Gulf Coast. Either garnered from the Gulf Coas t or made locally as copies, these vessels likely carried considerable ceremonial importance on the Lower St. Johns River. Several vessel forms restricted to mound assemblages had almost entirely non-cooking functions. These include small cups and bowls small jars, boat-shaped bowls, double bowls, multi-compartment trays, beakers, bottles, and possibly, shallow bowls. Combined, only 7% (n=3) contain evidence of soot. In addition, only one vessel shows eviden ce of repair with mend holes, corresponding with the possibility that many of these vessels were not heavily used. Indeed, many of these vessels, particularly the multi-compartment trays, double bowls, and various miniature vessels, are clearly ceremonial forms intended for only specific ritual tasks. The comparably high frequency of incised and painted pottery types among these vessel forms, rarely found in midden contexts on the Atlantic coast, corroborates their special function. In sum, contained within mortuary mounds were numerous vessel forms ostensibly designed for various specific functions and part icular cultural contexts Based on many rim profiles from Swift Creek middens along the Atla ntic coast, the vast majority of domestic cooking vessels were identical in shape and size to the sooted pots and bowls placed in mortuary mounds. In contrast, many vessels with specia l cooking functions and non-cooking functions in mound assemblages are apparently absent in village and midden contexts. As noted, the limitations of small sherds, as are typical at mi ddens, impede the conclu sive identification of many vessel forms. However, the disparity between vessel forms among mound and midden assemblages can be understood through comparisons of vessel orifice diameter. 217

PAGE 218

Orifice Diameter, Soot, and Mend Holes Am ong assemblages from mortuary mounds, wh ole or nearly whole vessels were not uncommon, and overall morphology and vessel cap acity could be accurately estimated. However, the majority of vessels in this study were each represented by only small sherds and vessel dimensions were much more difficult to infer. In the ab sence of whole vessels, orifice diameter estimates from rim sherds were used as a general proxy for vessel size. Orifice diameter corresponds most directly with overall vessel size when all vessels being compared have similar morphology. As noted, the midden assemblage consists overwhelmingly of open and slightly restricted pots and bowls, simple vessel forms w hose orifice diameters correspond more or less directly with overall size. Othe r vessel forms identified in the mound assemblages have variable vessel morphologies that make for poor predictions of vessel size based on orifice diameters. For example, bottles, collared jars and bowls, and beakers have greater capacities than their orifice diameters might suggest because of highly restricted orif ices or great height, while the opposite is true of shallow bowls. Thus, differences in orifice diameters between assemblages can at once reflect bo th distinctions in the overall volume of vessels and differences in the frequency of particular vessel forms. Out of 1222 total vessels, 517 orifice diameters were recorded, including 141 from mortuary mounds and 376 from midden contexts Taken as a whole, the combined mound assemblages have a much smaller mean orifice diameter (15.7 cm) and a much larger standard deviation (8.2) than the combined midden a ssemblages (Table 7-4; Table 7-5). These differences reflect the higher frequency at mound s of small vessels such as small cups, bowls, and jars, as well as vessel forms with small restri cted orifices such as flattened globular bowls and collared jars. In compar ison, the orifice diameters record ed from midden assemblages are distributed almost entirely within the range of variation among th e reconstructed cooking vessels 218

PAGE 219

from mounds. At middens, no orifice diameter sm aller than 8 cm was recorded, revealing that miniature vessels, double bowls, and multi-compartment trays rarely, if ever, were broken or deposited in domestic contexts. Furthermore, only 4% of vessels (n=15) from middens have orifices smaller than 12 cm compared to 37% (n=52) in the combined mound assemblages. Because cooking vessel forms in the mound assemb lages were determined to be universally larger than 12 cm in diameter, rims with esti mated orifice diameters less than 12 cm are very likely to be from vessels other than the basi c cooking forms. The rim profiles among the few small vessels within the midden assemblages indicate that they take the form of small cups, bowls, and jars, although beakers and collared jars cannot be comp letely ruled out for some of these small rim sherds (Figure 7-15). Regardless of whether rim sherds with small estimated orifice diameters represent small vessels like cu ps or bowls or larger vessels with highly restricted orifices, the relative frequency of both is demonstrably limited in domestic contexts, where simple cooking vessel forms predominate. A comparison of the frequency of soot ed vessels within both midden and mound assemblages corroborates the hypothesized functions of various vessel forms (Table 7-6). In both midden and mound assemblages the relative frequency of soot by approximate orifice diameter conforms to a bell-shaped curve in which frequency is greatest among vessels with orifice diameters between 20 and 24 cm and freque ncies diminish toward either end of the size distribution. Thus, the relative frequency of soot among very sma ll and very large vessels is low among both midden and mound assemblages. The c onsistent discrepancy of soot frequency between midden and mound assemblages of all sizes is very likely due to sampling biases. Specifically, the smaller portions of vessels repres ented from village assemblages leads to lower frequencies of soot identification than within mound assemblages in which the entire rim of a 219

PAGE 220

vessel is often available. This sam pling bias results in a sort of logarithmic effect in the midden assemblage data, decreasing the order of ma gnitude by which size classes differ in the frequencies of recorded soot (Fig ure 7-16). Many small sherds that lack soot are likely to have come from vessels that contained soot on other portions of the rim or wall. The frequency by which sooted vessels were misclassified using sher ds that lacked soot appears to have increased as the actual occurrence of soot increased in the whole vessel sample Yet this bias was not great enough to conceal the relative tre nd of soot frequency among small sherds that mirrors the whole vessel assemblage. In sum, vessels with small orifice diam eters from midden assemblages appear to correspond with the small cups, bowls, and jars fr om mound assemblages that were very rarely used in cooking. On the other end of the size distribution, very large vessels in both mound and midden assemblages have a lower frequency of soot that may also indicate a non-cooking function. Notably, among whole or reconstructed vessels, only open bowls have recorded orifice diameters greater than 28 cm and therefore sh erds with the largest orifice diameters from middens are likely to have come from open bow ls as well. Given the comparatively low frequency of soot on these vessels, many open bowls clearly served non-cooking functions. Mend holes were recorded for their potential in discriminating th e breakage rates and relative cultural importance of different kinds of vessels (DeBoer a nd Lathrap 1979; Senior 1994). Most important, the number of mend holes in a vessel is likely to re flect its longevity, an important aspect of an objects biography that could re veal some of its symbolic density. Mend hole data have proven useful for distinguishing some vessel forms, such as the collared jars that appear to have been repaired with some frequency, however, the combined number of vessels with mend holes from all assemblages (n=39) provides too small a sample to reveal significant 220

PAGE 221

differences in frequencies between size classes or pottery types. There is a general correlation between m end holes and larger vessel sizes but th e sample size makes this trend statistically insignificant. In addition, the same sampling biases that pertain to soot data are far more detrimental with such a small sample and inhibi t useful comparisons between the frequencies of mend holes in mound (n=17, 11%) and village (n=22, 2%) assemblages. While I have previously argued, based on just two assembla ges, that mound assemblages have a higher frequency of mend holes than village assemblage s (Wallis 2007), this statistic disregards the formidable sampling bias inherent in comparing small sherds to whole vessels. Whatever the degree of mathematical misrepresentation in co mparisons of entire assemblages, there are two vessels from mound assemblages that deserve specia l mention. One St. Johns Plain vessel, from the Mayport Mound, contained over two dozen mend holes along its bottom half that seem to have been an attempt to repair coil breaks (Wallis 2007:226). A Dunns Creek Red vessel from the Dent Mound contained twelve me nd holes that were oriented to repair a large portion of the vessel rim. Both of these vessels are large open pots, nearly 40 cm in orifice diameter. The large number of mend holes in these specimens indicates both the fragility of spiculate paste wares and the cultural concern for extending the lives of these particular vessels. For the Mayport Mound vessel especially, this concern may have surpas sed consideration of th e objects functional capabilities in order to keep the he irloom alive. To summarize, assemblages from mortuary mounds and middens are clearly comprised of different vessel forms, with mounds contai ning many small vessel forms that are very uncommon or absent at villages Frequencies of soot on vess el surfaces corroborate the noncooking function of smaller-sized vessels at a ll sites and further bolst er the much higher proportion of non-cooking vessel form s at mounds. Small sample size limits the comparability 221

PAGE 222

of m end holes among assemblages, although particular vessel forms such as collared jars appear to have been repaired more often than other vessels and two large open bowls from mounds are notable for exceptional numbers of holes. Combine d, these data delineate a suite of special-use vessels nearly exclusively deposited in ceremoni al contexts as well as cooking pots and bowls that are common to both mound and midden contexts. Importantly, this pattern is evident by the Early Swift Creek phase, as clearly indicated by the range of mound-specific vessel forms that c ontain temporally diagnostic charcoal-tempered paste. Charcoal tempering is found in flattened globular bowls, small cups, bowls, and jars, boat-shaped bowls, multi-compartment trays, and beakers, all forms restricted mostly or exclusively to mounds. Altern atively, there are no examples collared jars with charcoal tempering, although Crystal River Incised examples are presumed to be contemporaneous with Early Swift Creek. In comparison to mound assemb lages, midden vessels with charcoal temper are restricted almost exclusively to standard dom estic cooking forms. Th us, a sort of dichotomy between mortuary and domestic forms was well established by the Early Swift Creek phase and many of the unique mound vessels were made in the local paste tradition. Rim Thickness Rim thickness was measured 3 cm below the lip on a total of 443 ve ssels, including 138 vessels from mounds and 305 from middens (Table 7-4; Table 7-5). Ther e is a wide range of variation in rim thickness across the combined assemblages, ranging from 3 mm to 11.1 mm. While there are small differences in the rim thicknesses of mound and midden assemblages on the Lower St. Johns River, the most notable differences appear betw een midden assemblages themselves and correspond with both geography and temporality. There is a significant difference between Lower St. Johns River midde n rims, which are generally thin (mean=6.39 mm), and Lower Altamaha River rims, which ar e comparably much thicker (mean=7.83 mm). 222

PAGE 223

This geographic difference is especi ally clear between villag e sites in each area that are likely to have been contemporaneous based on radiocarbo n assays. Each dating to the seventh century, Greenfield #8/9 (8DU5544/5) on the Lower St. Johns River and Sidon (9MC372) on the Lower Altamaha River have mean rim thicknesses of 6.9 mm and 8.1 mm, respec tively. Sample sizes are regrettably small for assemblages from site s in the areas between Greenfield #8/9 and Sidon, but rim thicknesses appear to vary between the thin and thick extremes of the southern and northern sites. However, analysis of larger assemblages from Camden and Nassau counties may demonstrate a graduated continuum in the distri bution of rim thickness. Saunders (1986), for example, recorded mean thicknesses of 7.2 mm (n=20) and 7.6 mm (n=18) for the two most common vessel forms at the Late Swift Creek si te of Kings Bay (9CM171). Because thickness was not measured according to the same paramete rs in Saunders (1986) work (just below the lip) as the current one (3 cm below the lip), th ese data are not directly comparable. However, the profiles of the vast majority of vessels from Swift Creek site s on the Atlantic coast indicate that thickness increases with distance from the lip. Therefore, the Kings Bay data are likely to underestimate the rim thickness at 3 cm below the lip. Variation in rim thickness is also highly co rrelated with temporal change. Assemblages from predominantly Early Swift Creek sites, su ch as Greenfield #7 (8DU5543) and Tillie Fowler (8DU17245), are comprised of mostly thin vessel s, with rim thickness means of 5.9 mm (n=45) and 6.2 mm (n=60), respectively. In comparison, Late Swift Creek assemblages from sites nearby have significantly thicker rims, such as at Greenfield #8/9 (mean=6.9 mm). The correlation between temporality and rim thickne ss becomes more pronounced when comparison between assemblages is limited to vessels that ar e irrefutably Early Swif t Creek and Late Swift Creek. Charcoal-tempered vessels, which clearl y date to the Early Swift Creek phase on the 223

PAGE 224

Lower St. Johns River, have a m ean thickness of 5.8 mm (n=118). In comparison, Late Swift Creek Complicated Stamped vessels with hallmark folded rims from Lower St. Johns River sites have a mean rim thickness of 6.9 mm (n=34). Thus, on the Lower St. Johns River rims were thin during Early Swift Creek times and became thicker with the emergence of Late Sw ift Creek pottery. Not coincidentally, this increase in rim thickness corresponds with Late Swift Creek paddle matches that link sites on the Lower St. Johns River to sites as far north as the Altamaha River, Georgia, where rims are very thick. A logical conclusion is that increasi ngly thicker rims in Lower St. Johns River assemblages are a consequence of social interactions with populations to the north where thick rims were more common, combined with the ac tual import of thick-rimmed vessels made in Georgia. Indeed, all of the measurable vessels on Lower St. Johns sites that were identified by INAA and petrographic analysis as originati ng from near the Altamaha River have rim thicknesses greater than average, ranging from 7.0 mm to 9.8 mm (avera ge 8.15 mm, n=6). At the same time, however, thicker rims are partly a consequence of slightly increasing vessel size over time. There is some positive correlation between rim thickness and orifice diameter. In a general sense, as rim thickness increases so doe s orifice diameter, but this correlation does not explain a large portion of va riation in rim thickness ( =.303). In fact, among midden assemblages rim thickness is more strongly corr elated with the latitude where a vessel was recovered ( =.551) than it is with orifice diameter Comparisons of all midden assemblages from the Atlantic coast reveal that most have a mean orifice diameter near 20 cm, while rim thickness averages vary widely. Unless there we re differences in height that could not be measured in this sample, coastal Swift Creek populations from north to south were producing roughly the same size subconoidal cooking pot regardless of rim thickness. 224

PAGE 225

Although thicker walls on vessels would have been less efficient in conducting heat, soot frequencies indicate that cooking pots in a variety of thicknesses were all used in direct heat cooking over fire. The differences in thickness are not likely to have been related to differences in diet or cuisine, at least based on available zooa rchaeo logical data that indicate a consistent diet dominated by the same species of fish and sh ellfish all along the Geor gia Bight (deFrance 1993; Fradkin 1998; Reitz 1988; Reitz and Quitmyer 1988). The majority of variation in rim thickness must be explained by other, presumably cultural, factors. More specifi cally, the thickness of vessels is likely to correspond w ith the habitual, routinized ve ssel forming routines that are acquired in a potters original learning environment (Arnol d 1998:358; Gosselain 2000; Reina and Hill 1978:230; Rice 1984; van der Leeuw et al 1992). Vessel thickness may be the residue of ingrained bodily practice, with generations of Swift Creek populations along the Georgia coast having learned to make thick vessels while thei r counterparts in coastal Florida inherited the technique of making thin vesse ls. Thus, rim thickness may provide critical data for differentiating social groups with in the Swift Creek archaeologi cal culture and investigating interaction between them. Paste Characteristics Using a binocular microscope (70x), the paste of vessels was analyzed in order to identify the frequency and size of a variety of aplastics. As reviewed in Chapter 6, aplastics in the paste of vessels may represent intentiona lly added temper, natura l inclusions in the clay, or both (Rice 1987:409-411). In general, charcoal, grog, and bone, are almost certainly in tentional additions to the paste because these constituents do not occu r with any frequency in natural clays (Rice 1987:410). Depending on their frequency in the past e, sponge spicules may be either natural constituents of the clays used for pottery (Cordell and Koski 2003; Espenshade 1983; see Chapter 6), or added temper (Rolland and B ond 2003). I concur with Rolland and Bond (2003) 225

PAGE 226

that the density of spicules within the paste of m ost St. Johns vessels denotes an intentionally added temper in Northeast Florida. Alternativ ely, mica, calcite, and fi ne quartz particles are naturally abundant in some clays and might have b een incorporated into th e paste of vessels with selection of particular clays. The present di scussion concerns the resu lts of analysis that characterized vessel paste const ituents by relative abundance and size. Abundance was recorded on an ordinal scale (absent, rare, moderate common, abundant) and size was grouped by the Wentworth scale for sand (very fine, fine, me dium, coarse, very coarse). Variation in the aplastic paste constituents in the sample can be usefully separated into five temper categories: spiculate, ch arcoal, fine sand, medium sand, a nd coarse sand-tempered pastes (Table 7-7). Bone, limestone, and especially grog were also observed in a few samples but never considered the dominant temper and therefore were not separated by category. Spiculate pastes are characterized by common to abundant spon ge spicules and, on occasion, angular or subangular very fine or fine sand. Charcoal-tempered pa stes contain occasional to common charcoal fragments ranging in size from very fine to very coarse, often within the same sample. Bone fragments occur on occasion as well as grog to a more limited extent. Very fine and fine sand that is angular or subangular is quite common. Fine sand pastes mostly contain abundant very fine and fine sand that is angular to subangular in shape. Mica is normally absent but in rare instances is common, usually correspondi ng with uncommon pottery types like Weeden Island Red. Medium sand and coarse sand pastes contain a majority of medium or coarse sized sand, respectively, with these large grains tending to have a subrounded or rounded shape. Very fine and fine sands are common in these pastes as well and tend to be more angular than the larger grains. 226

PAGE 227

Based on the ubiquity of angular and subangular quartz grains in all paste categories, these m ay have been natural inclusions in the clays ra ther than added temper. Spiculate, charcoal, medium sand, and coarse sand pastes commonly cont ain abundant very fine and fine sand. This hypothesis is also supported by the angular shap e of most small sand particles. Rounded and water-worn grains are not common in clay bodies, particularly in Florid a (Ann Cordell, personal communication, 2006; Shepard 1968). Alternatively, the medium and coarse pastes with more rounded grains may have been added to clay bod ies that already contai ned smaller and more angular quartz grains. The results of petrographic analysis of a limited number of samples were used to corroborate the temper categories assign ed to the entire assemblage. The petrography confirmed the gross temper categor ies in nearly every case. An important pattern that emerged from this analysis was in the distribution of the major paste constituents of vessels among sites. As expected, charcoal-tempered paste occurs only in northeastern Florida and comprises the majority of assemblages from Early Swift Creek sites like 8DU5543 and 8DU17245 (Table 7-8). Spiculate pa stes are a consistent minority at sites in northeastern Florida and seem to have been more frequent at mounds than middens. Finally, sand temper or inclusions are ubiquitous within most assemblages, but the size of the grains show consistent geographic distributions. Fine sand temper predominates on sites in northeastern Florida while medium and coarse sand temper is most common on sites in southeastern Georgia. Interes tingly, sites near the Florida-Georgia border in Nassau and Camden counties may yield a more equal distribution of grain size. The McArthur site (8NA32) contained roughly equal proportions of fine and coarse quartz te mper. An increasing frequency of coarse grain temper is apparent just to th e north: the Kings Bay site (9CM171) assemblage is reported to have been predominantly grit-tempered [1.0-3.0 mm], with sand and shell being 227

PAGE 228

m inority tempers (Espenshade 1985:304). Thus moving north along the coast from the St. Johns River there is a general trend toward increasing grit temper within Swift Creek assemblages. This geographic trend may be the result of bot h geological and cultural factors. Overall, sand grain sizes are likely to be smaller on the St. Johns River compared to the Altamaha River due to differences in gradient and consequent sediment load. At the same time, however, individual sand bars and different parts of a river channel might have widely varying grain sizes (Orrin Pilkey, personal communication, 2008). In fact, Espenshade (1985:302) notes sand and grit in a wide range of sizes found together on sand bars of the Crooked River near the Kings Bay site. Thus, sand grain size in earthenware pa stes may be related, in part, to the regional availability of sand sizes but is also likel y to correspond with cultural prescriptions for appropriate tempering agents. There is a notable correlati on between medium and coarse sand pastes and thick rims (Table 7-9). While this pattern mostly corr esponds simply with the parallel geographical distribution of grain size and ri m thickness, vessels from northeas tern Florida sites demonstrate this relationship as well. This correlation wi thin northeastern Florida samples denotes nonlocal vessels made in coastal Georgia or the work of potters from coastal Georgia accustomed to coarse tempers and thick vessel walls. Based on the provenience studies outlined in Chapters 5 and 6, most of the grit-tempered, thick-rimmed vessels on the Lower St. Johns probably were made in Georgia near the Altamaha Rive r and were ultimately exchanged. Summary and Conclusions The combined data from technofunctional anal ysis lead to two compelling discoveries. First, specialized forms of po ttery were made for specific tasks and were deposited almost exclusively at mortuary mounds while ostensibly domestic cooking vessels were placed at both 228

PAGE 229

mounds and habitation s ites. Second, resilient manufacturing traditions that correspond with populations in certain areas can be identified. To summarize, the fifteen vessel forms defined from the assemblage of whole or nearly w hole vessels were grouped into three functional categories based on their physical suitability to ce rtain tasks and the presence or absence of soot and mend or suspension holes. These categories include vessels used primarily for cooking, those used occasionally for (ceremonial) cooki ng and non-cooking tasks, and those used for entirely non-cooking functions. Based on profil es, estimated orifice diameters, and soot frequency among rim sherds, midden assemblages are dominated by moderately sized cooking vessels. Mounds contain many of these cooking vessels too, but they comprise only half of mound assemblages. The remainder of mound asse mblages are made up of vessels intended for specific ritual tasks, both c ooking and non-cooking. Many of these non-cooking vessels were small cups and bowls that may have had serving functions while other forms had other functions as containers of potent ritual substances. Based on the prevalence of rare pottery types among these special-function vessels (Crystal River, Weeden Island, Dunns Creek Red, etc.), many may have been made nonlocally. Unfortunately, chemical and mineralogical data were not generated from any of these presumably foreign vessels. Finally, geographic patt erns in average rim thickness and the size of qua rtz temper in vessels are likely to reflect the unconscious, ingrained aspects of technological style that correspond with particular lear ning environments in at least two different pottery traditions, one located on th e Lower St. Johns River and the other centered around the Altamaha River. Thus, the story that is emerging from the Middle and Late Woodland pottery on the Atlantic coast is reminiscent of the sacred a nd secular dichotomy that Sears (1973) used to rather simplistically describe Weeden Island pottery assemblages. Yet exactly how these objects 229

PAGE 230

operated in social life m ay have been quite different from the cl assic Weeden Island contexts of the Gulf Coast area. Sears (1973) clearly viewed the ornate and fantas tic earthenware vessels from Weeden Island mounds as ceremonial rather than purely mortuary forms, arguing that the vessels were used for some period of time in various non-secular contexts and periodically buried all at once in caches. In comp arison, the continuous use type of mound was predominant on the Lower St. Johns River and there were not east-side caches of vessels or other artifacts (Sears 1967). Instead, at many of the mounds, vessels were deposited continuously over time, often over a period of several centuries. Many of these vessels we re clearly not produced for immediate burialthey were used repeatedly as indicated by soot and mend holes in many specimens. While there is presently no way to determine whether individual special-use vessels eventually buried at mounds were used exclus ively (and repeatedly) at mounds or whether their roles in social life extended to events at ot her locations, the periodic deposition of individual vessels, compared to the caching of entire a ssemblages, may relate to a more obvious heterogeneity in the biography of objects. Rather than an assemblage of vessels that were used repeatedly for the same periodic events, it seems more likely that some of the special-use vessels deposited at mounds along the Lower St. Johns River were used for a variety of rituals (perhaps involving healing, divination, various initiations, etc.) at multiple locations. For some of these vessels it may have been only its last use that took place at mounds, thus invoking and mobilizing object biographies in particularly salient social events. Notably, the vessels identified as foreign-made from the Georgia coast are nearly all domestic cooking pot forms. They bear th e chemical, mineralogical, and morphological trademarks of cooking vessels made by potters local to the Altamaha River for use in domestic cooking. It appears, then, that Swift Creek vessels represent an unusual kind of gift, one that was 230

PAGE 231

drawn out of dom estic and mundane contexts to make manifest its latent symbolic potential. These were a special kind of gift that achie ved symbolic transformations through dramatic contextual shiftsthe significance and meaning of which are discu ssed in the final chapter. Notes 1. The maximum and minimum diameters of each non-unifo rm orifice was averaged fo r the purposes of summary statistics. 2. Exceptions include the following: (a ) one instance of charcoal temper id entified under low magnification but not observed in petrographic thin section; (b) two instances in which coarse sand was identified under low magnification but only fine sand was identified in petrographic thin section. 231

PAGE 232

Table 7-1. Vessel form summary statistics. Orifice Diameter (cm) Height (cm) Rim Thickness (mm) Volume (l) Open Bowls (n=14) Range 17 40.5 13 27.5 5.1 7.35 1.5 15 Mean (Standard Deviation) 25.7 (8.11) 18.75 (4.99) 6.39 (.8) Restricted Bowls (n=5) Range 12.0 30.0 11.0 28.0 5.32 9.10 1.5 18 Mean (Standard Deviation) 19.7 (8.15) 21.25 (7.27) 6.68 (1.54) Restricted Pots (n=16) Range 12.25 28.0 19.0 41. 0 4.8 9.8 2.5 13.5 Mean (Standard Deviation) 20.67 (5.0) 25.42 (6.31) 6.48 (1.38) Open Pots (n=12) Range 12.0 26.0 13.75 27. 0 4.45 7.0 1.0 7.0 Mean (Standard Deviation) 16.86 (4.31) 19.8 (4.01) 5.88 (.75) Flattened-Globular Bowls (n=9) Range 7.0 17.5 10.0 19.0 4.7 8.5 .5 4 Mean (Standard Deviation) 11.72 (3.07) 13.19 (3.25) 6.83 (1.25) Collared Jars (n=7) Range 9.0 18.0 16.0 27.0 4.6 9.7 .8 2.5 Mean (Standard Deviation) 11.21 (3.12) 21.6 (4.35) 6.59 (1.83) Small Cups and Bowls (n=15) Range 5.2 16.5 3.2 13. 0 4.0 9.0 0.1 0.75 Mean (Standard Deviation) 10.13 (2.93) 8.12 (3.05) 5.54 (1.27) Small Jars (n=11) Range 1.8 13.0 3.0 19.0 4.0 8.1 .02 1 Mean (Standard Deviation) 5.92 (3.7) 9.47 (4.67) 6.08 (1.64) Boat-shaped bowls (n=5) long /short axis Range 6.85-18.0 / 4.0-11.0 2.15 9.0 4.5-6.0 .02 1 Mean (Standard Deviation) 11.91(4.3) /7.13(3.08) 6.08 (2.96) 5.12 (.58) Double Bowls (n=3) Range 4.38 5.7 3.0 8. 0 3.0 7.0 0.15 0.2 Mean (Standard Deviation) 5.03 (.66) 5.83 (2.57) 4.5 (2.18) Multi-Compartment Trays (n=5) Range 4.5 12.0 6.0 10. 35 2.0 11.50 .15 1.5 Mean (Standard Deviation) 7.63 (2.96) 8.09 (1.78) 5.74 (3.68) Beakers (n=3) Range 6.75 9.2 15.0 18. 0 4.6 5.15 0.4 0.9 Mean (Standard Deviation) 8.05 (1.23) 17.0 (1.73) 4.88 (.28) Bottle (n=1) 4.5 4.9 19.2 1.1 Shallow Bowl (n=1) 17.0 4.5 4.0 0.7 Double-Globed Jar (n=1) 8.0 6.0 30.0 4.3 232

PAGE 233

233 Table 7-2. Soot and mend hole freq uencies in each vessel form. Vessel Type Soot Mend Holes Open Bowls (n=14) Frequency 6 4 Percent 43 29 Restricted Bowls (n=5) Frequency 2 2 Percent 40 40 Restricted Pots (n=16) Frequency 11 2 Percent 69 13 Open Pots (n=12) Frequency 11 1 Percent 92 8 Flattened-Globular Bowls (n=9) Frequency 3 1 Percent 33 11 Collared Jars (n=7) Frequency 2 5 Percent 29 71 Small Cups and Bowls (n=15) Frequency 1 1 Percent 7 7 Small Jars (n=11) Frequency 1 0 Percent 9 0 Boat-shaped bowls (n=5) Frequency 1 0 Percent 20 0 Double Bowls (n=3) Frequency 0 0 Percent 0 0 Multi-Compartment Trays (n=5) Frequency 0 0 Percent 0 0 Beakers (n=3) Frequency 0 Percent 00 Bottle (n=1) 00 Shallow Bowl (n=1) no data 0 Double-Globed Jar (n=1) 1 0

PAGE 234

Table 7-3. Frequency of vessel form by surface treatment and pottery type. 234 Plain Stamped Incised/Punctated Painted Vessel Form CP STPSTJGRGESCLSCSCOtherS WII CRI OtherIPWIRDCRTotal Open Bowls 3 7 1 1 1 1 14 Restricted Bowls 1 3 1 5 Restricted Pots 4 7 1 2 2 16 Open Pots 2 1 4 1 3 1 12 Flattened-Globular Bowls 1 3 2 1 2 9 Collared Jars* 6 2 8 Small Cups and Bowls 3 4 1 1 2 1 1 2 15 Small Jars 2 1 2 2 4 11 Boat-shaped bowls 2 3 5 Double Bowls 2 1 3 Multi-Compartment Trays 1 3 1 5 Beakers 2 1 3 Bottle 1 1 Shallow Bowl 1 1 Double-Globed Jar 1 1 TOTAL 17 35 4 1 7 13 3 3 3 6 5 6 6 109 *includes one collared bowl CPcharcoal-tempered plain OtherS other stamped (simple stamped, dent ate stamped) STP sand-tempered plain WII Weeden Island Incised STJ St. Johns Plain CRI Crystal River Incised GRG grog-tempered plain OtherIP non-diagnostic incised and/or pun ctated ESC Early Swift Creek (charcoal-tempered) WIR Weeden Island Red LSC Late Swift Creek DCR Dunns Creek Red SC Swift Creek Complicated Stamped (cannot be defined as Early or Late)

PAGE 235

Table 7-4. Mound assemblage orifice diam et er and rim thickness summary statistics. Orifice Diameter (cm) Rim Thickness (mm) n 60 53 Dent (8DU68) mean 18.28 5.93 sd 7.29 1.08 min 8.00 3.80 max 40.50 8.70 n 34 35 Mayport (8DU96) mean 16.40 6.26 sd 9.32 2.62 min 3.50 4.10 max 39.00 10.35 Other LSJ Mounds n 47 50 mean 11.76 6.87 sd 7.46 1.65 min 1.80 3.00 max 30.00 10.30 TOTAL n 141 138 mean 15.68 6.22 sd 8.18 1.37 min 1.80 3.00 max 40.50 10.35 235

PAGE 236

Table 7-5. Midden assemblage orifice diam eter and rim thickness summary statistics. Orifice Diameter (cm) Rim Thickness (mm) Tillie Fowler (8DU17245) n 78 60 mean 19.86 6.23 sd 5.40 1.21 min 8.00 3.90 max 31.00 9.2 Greenfield #7 (8DU5543) n 79 45 mean 20.05 5.85 sd 4.67 0.98 min 10.00 4.15 max 29.00 8.3 Greenfield #8/9 (8DU5544/5) n 64 78 mean 23.5 6.86 sd 5.59 1.00 min 10 5.1 max 34.00 9.8 McArthur (8DU32) n 14 12 mean 19.93 6.80 sd 5.50 1.63 min 10 3.6 max 30.00 9.8 Cathead Creek (9MC360) n 30 22 mean 20.4 7.60 sd 5.25 1.39 min 9 5.15 max 30.00 10.0 Evelyn (9GN6) n 24 17 mean 19.8 7.95 sd 4.93 1.22 min 10.00 5.50 max 28.00 9.51 Hallows Field (9CM25) n 7 8 mean 22.86 7.39 sd 6.89 1.53 min 13.00 5.45 max 32.00 9.25 Kings Lake n 5 6 mean 21.00 6.42 sd 5.83 1.13 min 14.00 5.35 max 28.00 7.3 Lewis Creek (9MC16) n 10 11 mean 19.89 8.50 sd 7.98 1.56 min 10.00 5.8 max 36.00 10.9 236

PAGE 237

Table 7-5. Continued. Orifice Diameter (cm Rim Thickness (mm) Sidon (9MC372) n 51 34 mean 20.51 8.09 sd 5.29 1.23 min 8.00 5.18 max 34.00 11.10 Florida Residual n 8 6 mean 19.25 5.90 sd 5.42 1.07 min 14.00 4.55 max 28.00 7.5 Georgia Residual n 6 6 mean 22.83 7.71 sd 5.31 0.62 min 18.00 6.7 max 32.00 8.4 TOTAL n 376 305 mean 20.68 6.91 sd 5.43 1.39 min 8 3.9 max 36 11.1 Table 7-6. Soot frequency grouped by orifice diam eter. <12 12.0-16 16.1-20 20.1-24 24.1-28 28.1-32 >32 TOTAL Midden # sooted 1 7 15 14 9 1 0 47 total vessels 15 66 117 84 64 25 5 376 % 7 10.6 12.8 16.7 14.1 4 0 12.5 Mound # sooted 5 9 10 11 13 2 0 50 total vessels 52 29 19 13 20 5 3 141 % 9.6 31 52.6 84.6 65 40 0 35.5 237

PAGE 238

Table 7-7. Aplastic constituents of gross paste categories. Paste category vessel count percent description spiculate 38 3.3 common to abundant sponge spicules none to common subangular very fine to fine sand rare ferruginous lumps no mica one sample with grog charcoal 238 20.7 occasional to common charco al fragments: very fine to very coarse rare to occasional bone fragments rare grog common to abundant angular to subangular very fine sand occasional to common angu lar to subangular fine sand rare instances of medium and coarse sand none to occasional mica rare to occasional ferruginous lumps fine sand 543 47.1 common to abundant angular to subangular very fine sand occasional to abundant subangul ar fine sand; rare rounded fine sand none to occasional mica; in rare instances common none to occasional ferrugino us lumps; in rare instances common medium sand 127 11.1 none to common angular to subangular very fine sand occasional to common subangular to subrounded fine sand occasional to common subangular to rounded medium sand rare coarse and very coarse sand no mica none to rare ferruginous lumps one sample with grog coarse sand 204 17.8 none to common angular to subangular very fine and fine sand none to common subangular to rounded medium sand occasional to common subangular to rounded coarse sand none to occasional subang ular to rounded very coarse sand none to rare mica none to rare ferruginous lumps; in rare instances common TOTAL* 1150 100.0 *when fresh breaks on vessels were not permitted, no paste category was assigned. 238

PAGE 239

Table 7-8. Frequency of gross paste groups by site. Charcoal Fine Sand Medium Sand Coarse Sand Spiculate Other* Total # % # % # % # % # % # % # % 8DU5543 85 52.8 62 38.5 6 3.7 6 3.7 2 1.2 161 100.0 8DU17245 93 66.0 37 26.2 5 3.5 1 0.7 5 3.5 141 100.0 8DU5544/5 26 6.1 316 74.2 38 8.9 20 4.7 23 5.4 3.0 0.7 426 100.0 8NA32 3 7.9 14 36.8 2 5.3 18 47.4 1.0 2.6 38 100.0 8DU14683/14686 4 20.0 15 75.0 1 5.0 20 100.0 8DU68 14 38.9 12 33.3 4 11.1 6 16.7 36 100.0 8DU96 6 14.6 21 51.2 3 7.3 5 12.2 4 9.8 2 4.9 41 100.0 Other LSJ Mounds 6 11.8 26 51.0 7 13.7 9 17.6 3 5.9 51 100.0 Other Florida Middens 1 16.7 4 66.7 1 16.7 6 100.0 Northeastern Florida 238 25.9 507 55.1 66 7.2 65 7.1 38 4.1 6 0.7 920 100.0 9MC360 5 10.4 13 27.1 30 62.5 48 100.0 9GN6 3 13.0 3 13.0 17 73.9 23 100.0 9MC25 4 26.7 4 26.7 7 46.7 15 100.0 9MC16 3 13.0 4 17.4 16 69.6 23 100.0 9MC372 19 19.8 32 33.3 45 46.9 96 100.0 Kings Lake 1 5.3 2 10.5 16 84.2 19 100.0 Other Georgia Middens 1 7.7 3 23.1 8 61.5 1 7.7 13 100.0 Southeastern Georgia 0 0.0 36 15.2 61 25.7 139 58.6 0 0.0 1 0.4 237 100.0 *includes grog, sponge spicule and grog, sand and bone, and limestone Table 7-9. Rim thickness summary st atistics by gross paste groups. Fine Sand Medium Sand Coarse Sand All Pottery n=162 n=47 n=56 Mean Rim Thickness 6.65 7.75 7.66 Standard Deviation 1.23 1.41 1.45 Min 3.00 5.20 4.70 Max 9.90 11.10 10.90 Florida Pottery n=148 n=18 n=23 Mean Rim Thickness 6.5 7.3 7.4 Standard Deviation 1.2 1.5 1.4 Min 3.0 5.2 4.7 Max 9.8 10.4 9.8 239

PAGE 240

A B Figure 7-1. Open bowl profiles. A) Vessels (wall thickness not to scale, D= Dent, M = Mayport), B) Rims with estimated orifice diameters. 240

PAGE 241

A B Figure 7-2. Restricted bowl prof iles. A) Vessels (wall thickness not to scale, D= Dent, M = Mayport), B) Rim. 241

PAGE 242

Figure 7-3. Restricted pot vesse l profiles (wall thickness not to scale, D= Dent, M = Mayport). 242

PAGE 243

Figure 7-4. Restricted pot rim profile s and estimated orifice diameters. 243

PAGE 244

Figure 7-5. Open pot vessel profil es (wall thickness not to scale, D= Dent, M = Mayport, G = Grant, GE = Grant Mound E). 244

PAGE 245

A B Figure 7-6. Flattenedglobular bowl profiles. A) Vessels (wall thickness not to scale, AL=Alicia, AR=Arlington, D=Dent, DN=De nton, LG=Low Grant ), B) Rims (AL=Alicia, M=McArthur). 245

PAGE 246

A B C Figure 7-7. Collared jar profiles. A) Vessels (wall thickness not to scale, A=Alicia, D=Dent, M=Mayport, GE=Grant E, R=Reddie Point), B) Sooted vessel (M5) with prominent rim fold. Courtesy Florida Museum of Natural History. C) Collared bowl. 246

PAGE 247

Figure 7-8. Small cup and bowl vessel profiles (wall thickness not to scale, A=Alicia, B=Beauclerc, D=Dent, LG=Low Grant, M=Mayport, MO=Monroe). 247

PAGE 248

A B Figure 7-9. Small jar profiles. A) Forms (wall thickness not to scale, B=Beauclerc, LGA=Low Grant A, LGE=Low Grant E, GI=Grant with specific mound not determined, M=Mayport, P=Point La Vista), B) Incise d designs. Courtesy of the National Museum of the American Indian. 248

PAGE 249

Figure 7-10. Boat-shaped bowls (LG=Low Grant, D=Dent (top view), F=Floral Bluff, G=Grant, M=Mayport (top view)). Courtesy of the Na tional Museum of the American Indian, the Jacksonville Museum of Science and Hi story, and the Florida Museum of Natural History. 249

PAGE 250

Figure 7-11. Double bowls (B=Beauclerc, P=Point La Vista, LG=Low Grant). Courtesy of the National Museum of the American Indian. 250

PAGE 251

Figure 7-12. Multi-compartment trays (B=B eauclerc, D=Dent, GE=Grant E, M=Mayport, ME=Monroe). Courtesy of the National Museum of the American Indian, the National Park Service, and the Jacksonvill e Museum of Science and History. 251

PAGE 252

A B Figure 7-13. Beakers (L G=Low Grant, M=Mayport) Courtesy of the Na tional Museum of the American Indian and the National Park Service. A) profile view, B) oblique view. 252

PAGE 253

Figure 7-14. Double-globed jar from Grant Mound E. Courtesy of the National Museum of the American Indian. 253

PAGE 254

254 Figure 7-15. Rim profiles of small cups, bowls, and jars from midden contexts. Figure 7-16. Soot frequenc y grouped by orifice diameter.

PAGE 255

CHAP TER 8 THE SWIFT CREEK GIFT The exchange of Swift Creek Complicated Stamped vessels was an important social practice that, on the Lower St. Johns River at least, became inextricable from mortuary ceremony. The specific contexts of production a nd deposition and the ap parent functions of these vessels reveal the signatures of gifts that were transformed into significant citations from the seemingly mundane material of everyday pr actice. This conclusion emerges out of a genealogy of the material practices of pottery production, use, and deposition that has been developed in preceding chapters. This fi nal chapter synthesizes the sourcing and technofunctional data from pottery assemblages in order to construct an outline of material practice and interpret eviden t patterns of exchange. A Genealogy of Swift Creek Materiality on the Atlantic Coast At first glance, Swift Creek Complicated Stam ped pottery seems to have been initially adopted on the Atlantic Coast as simply a di fferent stamping technique that was added to existing pottery technology. Indeed, where co mplicated stamping first occurred on the Lower St. Johns River, simple, check, and a variety of other stamping forms had undergone periods of fluorescence amidst a persistent sand-tempered plain pottery tradition that spanned several centuries (Ashley and Wallis 2006). Judging by limited data from earlier contexts, cooking vessels in the form of sub-conical pots and open bowls were widespread by at least the Deptford phase and persisted through the adoption of co mplicated stamping (DePratter 1979; Kirkland 2000). Cooking vessel forms did not change, yet the adoption of complicated stamping was coincident with a number of new material prac tices. First was the development of a locally distinctive temper that consiste d of pounded charcoal fragments. Charcoal-tempered pottery is found only within a circumscribed area around the Lower St. Johns River and seems to have 255

PAGE 256

corresponded with production by soci al groups that shared close cult ural ties with one another. As the incorporation of hearth contents from hou seholds or villages into vessels that were moved across the Lower St. Johns landscape, charcoal tempering may signify a growing concern with descent groups and their dist ribution. Second, the adoption of complicated stamping was coincident with the initiation of a series of mortuary mounds along the Lower St. Johns River, some of which continued to serve as mortuary re positories for centuries. Third and finally, with complicated stamping, charcoal tempering, and mortuary mounds, came many new vessel forms that were deposited only in mound contexts. Along with human interments, shells, exotic objects of copper, mica, and stone, and many plain a nd complicated stamped cooking pots that were identical to local village forms, were fl attened globular bowls, boat-shaped bowls, multicompartment trays, beakers, collared jars, and small cups, bowls, and jars. Many of these new vessel forms were made with charcoal-tempe red paste and were therefore produced locally. However, a few vessels had Crystal River series surface treatments and may have derived from the Gulf Coast, although these could not be sampled by INAA or petrography. Therefore, upon closer examination, the adoption of Swift Creek pottery brought a revolution of material practice, particularly in hearth-tempered pottery, the construction of a mortuary landscape, and in the ra nge of vessel forms that were buried exclusively at mortuary mounds. The initiation of each of these material practices arguably entrenched peoples concerns with descent groups and their standing vis--vis one another, a point to which I will return later. The distincti on between mound and village pottery forms was likely a significant departure from earlier Deptford phase ceramic tr aditions on the Atlantic coast (e.g., DePratter 1979) but followed in the traditi on of the various sacred-secular divisions of the Gulf Coast Yent and Green Point complexes that culminated in later Weeden Isla nd contexts (Sears 1962, 256

PAGE 257

1973). Even as the m ortuary tradition and cerem onial life that included specific vessel forms seemed obviously to derive much inspiration fr om populations to the we st, most Early Swift Creek vessels in both mounds and middens were locally produced along the Lower St. Johns River. Instrumental Neutr on Activation Analysis and petrography confirm local production origins for nearly all vessels that are demonstr ably Early Swift Creek. However, this sample consists primarily of charcoal-tempered specimens and includes no Crystal River series examples. Swift Creek Complicated Stamped pottery on th e Georgia coast also developed out of the Deptford tradition that lingered there until relativ ely late, with apparent influences from the interior of central and southern Georgia. Presently unclear is whether there was a wide-scale migration of Swift Creek pottery-making populations into the Georgia coas tal zone from further inland (Wayne 1987). Regardless of population movement, Swift Creek groups were permanent residents of the Georgia coast and developed distinctive elements in their carved designs that are distinguished from the corpus of interior Georgia paddle designs (A shley et al. 2007). Nevertheless, social ties between populations on the coast and the interior are evident in several paddle matches between the two areas as well as mound-building practic es at Evelyn that resemble interior traditions. Th ere is no evidence to suggest th at the ceremonial attributes found at mounds like Evelyn were autochthonous developm ents. Rather, these practices seem clearly to represent a departure from De ptford mortuary traditions (Thomas and Larsen 1979). Thus, complicated stamping was adopted along the Atlantic coast in two separate historical processes. Swift Creek Compli cated Stamped pottery came early (ca. AD 200) to the Lower St. Johns River from the Florida Gulf coast, along with a whole host of new material practices. By all accounts, Swift Creek on the Lower Altamaha River also came with other cultural changes, 257

PAGE 258

but these new ideas cam e from central and southern Georgia several centuri es later (ca. AD 500). These distinct historical traj ectories are important because be tween AD 200 and 500 there appear to be no material connections between Lowe r St. Johns River and Lower Altamaha River populations. Instead, Early Swift Creek Lower St. Johns River populations are characterized by the stalwart maintenance of mate rial practices that were distin ct but in some aspects borrowed from populations living along the Gulf Coast of Florida. These separate historical traj ectories along the Atlantic coas t are evident in the village pottery assemblages of the subsequent Late Sw ift Creek phase. Two attributes distinguish between Lower St. Johns River and Altamaha River midden assemblages. First, there are differences in temper, with grit temper more prevalent along the Georgia coast and fine sand temper predominant in the Lower St. Johns Rive r area. Second, Lower Altamaha River vessels are on average significantly thicker than Lower St Johns River vessels, even in the context of identical vessel capacities and similar diet and cu isine. Temper and vessel thickness both seem to correspond with longstanding di vergent cultural traditions that were inculcated for several centuries. While type of temper may have been consciously chos en, the thickness of vessel walls may very well have corresponded with habitual ways of making pots that were ingrained in somatic memory. The advent of Late Swift Creek rim form s and designs were the result of a global phenomenon in the sense that style changes a ffected the entire Swif t Creek pottery-making world. Indeed, thick folded rims are diagnosti c of Late Swift Creek pottery across the lower Southeast. Concomitant with folded rims on the Atlantic coast are paddle designs that seem to be more systematically carved and more carefully applied on vessel surfaces compared to sloppy Early Swift Creek examples (Ashley and Wallis 2006:8). This difference between 258

PAGE 259

Early and L ate Swift Creek designs may be due, in the former, to lack of concern for clearly registering designs on vessels as well as the e ffect of charcoal tempering that left designobliterating holes in vessel surf aces. The more systematic carvi ng and careful execution of Late Swift Creek designs make the id entification of paddle matches much easier for archaeologists, and it seems that this is what they were intended to do in the past as well. The distinctive and recognizable qualities of designs were ultimately mobilized by people to act as powerful material citations on the Atlantic coast. In the context of this widespread stylistic shift, Swift Creek pottery was introduced to the Altamaha River and southward down the Georgi a coast, charcoal tempering was abandoned on the Lower St. Johns River, and vessels made at Altamaha River sites were brought to the Lower St. Johns River. In this way social connections and a semblan ce of cultural continuity spread across Southeast Georgia and northeastern Florida for the first time in at least several centuries. There is no clear dividing line between the pottery traditions that characterize the Altamaha and St. Johns River regions during the Late Swift Creek phase. Instead, assemblages from sites between the two rivers appear to contain a mixt ure of the two styles. However, the Altamahabased tradition, with thick rims and grit temper had a notably stronger influence toward the south than did Lower St. Johns styles toward th e north. For instance, many Swift Creek sites on Amelia Island, just 15 km north of the mouth of the St. Johns River, are characterized by assemblages made up of more than 50% vessels w ith grit temper, some with very thick rims. There are also paddle matches between Georgia inte rior sites and sites in extreme southeastern Georgia that indicate continued so cial interactions. The southern most distribution of Late Swift Creek pottery characteristic of the Georgia coas t and paddle matches with the Georgia interior does not overlap with the extent of Early Swif t Creek manifestations on the Lower St. Johns 259

PAGE 260

River. This correspondence is no t coincidental. Rather, the dist ribution is likely to reflect the continued social cohesion of groups living along the Lower St. Johns River for centuries. Even with the b road changes in Swift Creek pottery st yles, there was significant continuity in the material practices of Late Swift Creek populatio ns on the Lower St. Johns River, continuing to use the same mortuary mounds initiated during the Early Swift Creek phase and continuing to produce similar ceremonial vessel forms. But there were new social interactions that took place along the Atla ntic coast with the spread of Late Swift Creek materiality. In strumental Neutron Activation Analysis and petrography together indicate that at least nine vessels from two Lower St. Johns River mounds, comprising 19% of the mound sample, were likely made somewhere near the Altamaha River. Three of these foreign vessels correspond with pa ddle matches that link them to specific village sites on the Altamaha River. In comparison, tw o vessels from midden contexts on the Lower St. Johns were made at Altamaha River sites, co mprising 2% of the sample. The data from Altamaha River midden contexts are basically identical: three vessels, comprising 3% of the sample, may have been made near the Lower St. Johns River. Unfortunately, no mortuary mound assemblages were sampled from the Altamaha River for this study, which might have manifested a similar pattern to St. Johns mound assemblages. Due to the multi-phase nature of mound asse mblages on the Lower St. Johns River, the frequency calculations are sure to underestim ate the proportion of fore ign pottery at mounds. The mound assemblages used in the analysis, from Dent and Mayport, span several centuries of both the Early and Late Swift Creek phases, fr om circa AD 300 to 800. Therefore, roughly half of these assemblages may date to the Early Swif t Creek phase, which pred ates interaction with populations in Southeast Georgia. If we rem ove from the calculations the charcoal-tempered 260

PAGE 261

sam ples, which we know come from Early Swift Cr eek contexts, nearly 27% of vessels at Lower St. Johns River mounds were made near the Alta maha River. However, this is still too conservative an estimate, because at least half of Early Swift Creek assemblages were tempered with fine sand rather than charcoal. This m eans that the relative frequency of Georgia-made vessels in Lower St. Johns mounds may be much higher, up to half of the Late Swift Creek Complicated Stamped and sand-tempered plain assemblages. After AD 500, foreign vessels derived from other regions also seem to have become more common in mounds. These include Weeden Island series vessels possibly from the Gulf Coast or North Central Florida, and St. Johns series vessels probably from the Middle St. Johns Rive r area, although none of these received adequate testing in the INAA or petrogr aphic studies. While some of these vessels may be local renditions of foreign wares, macroscopi c differences in the paste constituents of many may indicate distant production orig ins. Moreover, vessels of some of these nonlocal types, such as Weeden Island Red, are demonstrated by petrographic analysis of north -central Florida and Gulf Coast assemblages to have been carried considerable distances (Cordell 2006; Rice 1980) Thus, in the purview of ceremonial mate rial practices, Late Swift Creek phase populations on the Lower St. Johns River increasi ngly emphasized the unique and the foreign in ways that mirrored Weeden Is land cultures to the west (Sears 1973). Yet unlike some Weeden Island populations, for these Swift Creek pottery-making peoples t he foreign was manifested not just in ceremonial forms but also in domestic cooking vessels that were superficially similar to local village pots. From the available data, each of the vessels demonstrated to have been carried from the Altamaha River area and de posited at mounds on the Lower St. Johns River were domestic cooking forms. Based on the oc currence of coarse sand temper, some other special-use vessel forms may have been brought to mounds as well, including small cups and 261

PAGE 262

bowls, beakers, and collared jars. H owever, the foreign cooking vessels ar e especially important as recontextualized materials that became a new kind of citation in ceremonial contexts. The Forms and Meanings of The Swift Creek Gift On the Lower St. Johns River, domestic c ooking vessels were deposited in mounds for centuries. During the Early Swift Creek phase these were nearly all locally produced within the immediate region. However, these locally-mad e cooking vessels deposited at mounds are not likely to have all simply come from the nearest vi llage site. As dramatically demonstrated by an Early Swift Creek paddle match design shared am ong vessels deposited at the Dent, Alicia B, and Beauclerc mounds, transporting vessels so me distance to mounds may not have been unusual. In the heightened context of cerem ony in which other vessels were important for ritually specific ta sks, the salience of cooking vessels seems to have derived mostly from where they were made, who made them, and other reco gnizable aspects of biography. There may have been important biographical qual ities of plain cooking vessels th at were eventually buried at mounds, but complicated stamping enabled biographic specificity to become indelibly attached through the differentiation of design. Indeed, am ong Late Swift Creek phase assemblages, the overwhelming majority (82%) of foreign-made vesse ls identified in the sourcing analyses were complicated stamped. Thus, the citational capa bilities of complicated stamped vessels were employed during the Early Swift Creek phase, but the distance and frequency of these citations appears to have increased drama tically during the Late Swift Creek phase. This increase in the spatial extent of references corresponded with the culmination of distinctively clear and unadulterated paddle impressions. In short, the deliberate use of complicated stamped vessels as citations was related to the increased visu al impact of designs. Complicated stamping was, and continues to be, an exceptionally effective technology of citation that conferred upon si mple cooking pots new capabilities of enchainment (Chapman 262

PAGE 263

2000; Jones 2001). As a wet clay vessel was stam ped with a carved wooden paddle, it becam e connected to the paddle by beco ming a copy of its image and preserving evidence of contact (e.g., Frazer 1922; Mauss 1972). Through the impr ession of many vessels during manufacture, a carved wooden paddle was an effective tool for the creation of a recognizable chain of associations in objects that coul d be distributed across the lands cape. In this way, the biography of a stamped vessel could become linked not only to a carved wooden paddle but also to every other vessel that was stamped with the same paddle. The fingerpr inting capabilities of complicated stamping, making vessels that preserve d an exact copy of an image and evidence of direct contact with it, arguably transferred the unmodified essence of whatever was depicted on the wooden paddle. Indeed, part of the power of mimesis is in the inabil ity to distinguish the copy from the original (Taussig 1993:53). Gi ven the distribution of nonlocal vessels, found almost exclusively at mortuary mounds on the Lo wer St. Johns River, th ese were capabilities recognized by people who made and used Swift Creek vessels. Through complicated stamping, cooking vessels became material c itations or indexes of persons who made and used the design. These qualities made complicated stamped vessels parts of a quintessentia l distributed object, an image with many spatially separated manifesta tions across the landscape, each with a unique micro-history (Gell 1998:221). As distributed ob jects, complicated stamped vessels were not merely representations that sy mbolically stood for the wood en paddle that bore the same design. Rather, each image was manifested in a corpus of vessels that were moved across the landscape and distributed independen tly of one another. The power of these distributed objects culminated in their functions as indexes of bodi ly presence of an image and a person (e.g., Gell 1998:231; Munn 1990). 263

PAGE 264

Distributed objects are often funda mental to the constitution of distributed persons. S o it is, for example, that an arm-shell or necklace exchanged through Kula networks does not just represent a renowned person who owns it, but rather the object itself is the person of renown in the sense that it is conceived as embodying the age, influence, and wisdom of that person (Gell 1998:231). Thus, distributed objects incorporate the substance of a human subject in a way that allows them not only to represent a person but also to embody personhood in a corporeal way. The interpretation of exchanged Swift Creek Comp licated Stamped vessels as distributed objects, and hence vehicles of distributed personhood, pr ovides a cogent explanation for the data presented in previous chapters: designs depic ting faces and animals and their rendering by split representation, the high frequency of nonlocal ve ssels at mortuary mounds, and the apparent symbolic importance of nonlocal domestic cooki ng vessels amidst many special-use vessel forms in mortuary mound assemblages. Carved paddle designs that were impressed in to vessels often clearl y depicted faces and animals, and as I argued in Chapter 3, these im ages were rendered by split representation. This representational technique is significant because it tends to indicate an inex tricability of an image and what it represents (e.g., signifier and signified, in semiotic terms). Split representation renders images that are faces or animals as subjects, rather than as objects symbolically standing for the real life versions of them. Split repr esentation is common among tattooing cultures in which images are viewed not only as indelible on the skin but also constitutive of the skin of social persons (Gell 1998:195). Because images literally are what they represent, to be represented in two dimensions, a three-dimensional body must be c onceived as cut into sections and spread apart. The result is an image that em bodies and is part of the thing it represents. Faces, animals, and other designs rendered by spl it representation impresse d into an earthenware 264

PAGE 265

vessel could have conferred a degree of personhood to a vessel, m aking cooking pots into persons, and objects into subjects. This explan ation for the symbolic efficacy of complicated stamped designs resonates with the qualities of ten ascribed to long-di stance exchanged objects such as Kula necklaces and arm-shells. With a carved design embodying a social person, Swift Creek Complicated Stamped vesse ls were vehicles of distri buted personhood, disseminating the image, and therefore the person, across the landscape and through time. The fact that nonlocal complicated stamped vessels, as vehi cles of distributed personhood, would be most often deposited at mortuary m ounds is not surprising because this is where distanciations of personhood are most common. Indeed, as Alfred Gell (1998:223) conveys: The idea of personhood being spread around in ti me and space is a component of innumerable cultural institutions and practices. Ancestral shrines, tombs, memori als, ossuaries, sacred sites, etc. all have to do with the extension of personhood beyond the c onfines of biological life via indexes distributed in the milieu. If Swift Cr eek Complicated Stamped vessels were conceived as persons, what sorts of persons were bei ng distributed via the pot s and what does their deposition at mortuary mounds signify? The answer I believe, lies in marriage alliances and the social obligations established between descen t groups, the wife-givers and wife-takers. Levi-Strauss (1969:264) argues that split repr esentation corresponds cross-culturally with societies that are obsessed with competition over geneal ogical credentials that link men to gods because of the strict conformity of the actor to his role and of social rank to myths, ritual, and pedigree. The persons dist ributed via complicated stampe d pots are likely to have had genealogical significance as important ancestors whose possession by a lineage or clan legitimated the descent groups identity and rank. Swift Creek societies al ong the Atlantic coast were almost certainly not hierarchical to the degree that Levi-Strauss ( 1969:264) associates with 265

PAGE 266

split rep resentation. However, given the age dist ributions of burial populations, with very few subadults interred in mounds, not everyone ha d access to mound burial. While there were probably limited degrees of achieved status for individuals (Thunen and Ashley 1995), the premier concern in mortuary ritual was likely me mbership in and ranking of descent groups, with burial in a mortuary mound restricted to member s of high ranking lineages or clans. The ranking of descent groups was reproduced by clai ming important ancestors and coordinating advantageous marriages and exchange networks with other ranked des cent groups. Thus, not only the split representation of the designs themselves, but also the placement of Swift Creek Complicated Stamped vessels within and on top of mortuary mounds re veals their symbolic density as genealogical represen tations that may have been exch anged in the context of marriage alliances or in the event of a death that recalled the obligations of affinal kin. Without data from mortuary mound contexts al ong the Altamaha River and in the absence of other (especially perishable) forms of material culture besi des pottery, determining whether these were asymmetric affinal alliances and which descent groups were wife-givers and which were wife-takers is quite difficult. If descent groups located on the Lower St. Johns River were wife-givers to Altamaha River descent groups, th en vessels may have been exchanged in a reciprocal way as distributed pe rsons exchanged for biological pe rsons (wives). When a descent group on the Altamaha River accepted a wife, th e gifting of a complicated stamped cooking vessel to a descent group on the Lower St. Johns River was probably one constituent of creating affinal bonds. Thus, as a woman moved north through patrilocal resi dence, a person was extended south through the exchange of a vessel th at embodied the apical ancestor or some other important person of the wife-taki ng descent group. The vessels ma y have been exchanged in the event of a marriage and kept as the physical presence (through a totem) of the descent group 266

PAGE 267

whom had taken a woman as a wife. If this were the case, the gifted vessel was carefully guarded and not often used in village contexts, as very few nonlocal vessels were identified in village assemblages in the analysis. Alternativel y, vessels may have been given in the event of a death, in which the death of affi nal kin recalled debts accrued through histories of exchange and provisioning that are typical of marriage alliances (e.g., Battaglia 1983, 1990; Khan 1989; Weiner 1992). In either case, the end result was the same: the extension of the social person embodied in the vessel was ritua lly killed and placed in (or on top of) a mortuary mound. The ritual death of the vessel did not signify the dissolu tion of the relationships that had been developed between descent groups. More likely, placing a piece of the affinal descent group into a cemetery was a way to ensure the retention of past alliances and protention of future ones (e.g., Munns (1990) use of Husserls (1962, 1964) te rminology). By placing a piece of an allied descent group within a mortuary mound, ancestors were commingled and a permanent point of reference to these social connections was in scribed onto the landscape (e.g., Joyce 2000b). As the embodiment of important social pers ons claimed by descent groups, the designs and the carved paddles used to distri bute them on vessels were inalie nable (e.g., Weiner 1992). They could not be given away. On the Atlantic coas t there is no evidence to suggest that carved paddles were exchanged or used out side of local villages. Instead, the paddles seem to have been jealously guarded even as stamped vessels were broadly distributed. In fact, by overstamping and smoothing stamped impressions on vessels, a prohibition against una uthorized reproduction of the image was enforced. For the most pa rt, enough of the design was registered on an earthenware vessel for a clan or lineage ancestor to be recogni zed, but not enough of the design was present on a vessel to make an exact replica from sherds and thereby steal the capability of reproducing it. The apparent ra rity of negative impressions made by sherds in vessel production 267

PAGE 268

reveals the p ower of these prohibitions. Swift Creek vessels with designs were likely understood as embodying social persons but the full extent of the design remained partially concealed through overstamping, thus protecting the distributed person from co-option by other descent groups. With its bewildering effect on those trying to precisely interpret designs, overstamping may also have imparted an apotropaic function to vessels, that is, as prot ective devices (i.e., Gell 1998:83-90). As Gell (1998:84) summarizes, apotropa ic patterns are dem on-traps that act by compelling evil spirits to be fascinated by intric acy or sheer multiplicity. Spirits become so engaged in attempting to comprehend a design, such as a maze, that they are distracted from their malicious plans. The complexity of overstamped Swift Creek designs on vessels exemplifies these mystifying qualities, as anyone attempting to reconstruct designs knows well. What is more, the multiplicity of each design on Swift Cr eek vessels is metaphorically linked to the protection of the vessel in functional terms: paddles were used to bond clay coils during manufacture, creating a solid earthenware body, and thus protecting the vessel from death (i.e. breakage). It seems highly likely that these de signs adorned peoples bodies for the same reason, as a way to ward off dangerous forces. By exchanging Swift Creek vessels, the protective devices that were rendered potent on the skin could be loaned to allied descent groups without giving up control of them. Given the prominen ce of exchanged vessels at mortuary mounds, these designs may have been especially important in death, as persons negotiated transitions into the ancestral and spirit world. In light of this interpretation of exchange d vessels as distributed persons, it is worth mentioning some other person-like qualities of vessels. Copying is often synonymous with reproduction (Taussig 1993:112) and in this sens e both wooden paddles and earthenware vessels 268

PAGE 269

m ight be viewed as having procreative power. Along these lines, a particularly compelling way to interpret the (re)production of complicated stamped vessels is as multiply-authored by two genders. Domestic pottery production is likely to have been the domain of women (Vincentelli 2000, 2004). However, carving wooden paddles may ha ve been the responsibility of men. In fact, wood carving seems to have been widely as sociated with men in North America according to ethnohistoric accounts (Drive r and Massey 1957:371). If thes e gendered divisions of labor pertained to Swift Creek phase populations, ther e may have been metaphorical and metonymical conceptions of these gendered technologies th at paralleled ethnographi c Southeastern Indian understandings, such as among the Creek Indians. For Creeks, women are associated with uncontrolled, generative force (water and corn) while men are identified with structuring forces (fire, wood, and bone) (Bell 1990). In the produ ction of Swift Creek complicated stamped pottery, the generative capacity of women in molding wet clay into a vessel, and the structuring capacity of men in finishing the vessel with stamped designs, may have been an act of reproduction that paralleled human biological concep tion. The logical conclusion of this idea is that once the vessel was conceived through padd le stamping (by men or by women using an object made by men), it went through a gesta tion period in the drying and firing process, whereby it emerged as a functional vessel. This vessel was essentially a completed person that represented the agency and the substance of at le ast two persons. This in terpretation of pottery production paralleling the constitution of persons appears more plausible in the context of paddle stamps that took the form of faces and in the likely commonality of both people and pots becoming cultural subjects through the renderi ng of designs on their surfaces. Conceived in a way that parallels biol ogical reproduction, and often given faces, complicated stamped vessels might also have been seen to act with person-like qualities. First, 269

PAGE 270

through the production process that included stamping m any vessels with a distinctive image, vessels could be seen to have procreative power s whereby they reproduced themselves. Indeed, many vessels in a village appear to have shared the same paddle stamp impression (Wallis 2007). The fecundity of vessels was also demonstrated by their function in cooking, repeatedly producing nourishing cooked food out of an orifi ce. Second, vessels had mobility and traveled across the landscape with people through residential m oves and, most importantly, to important places for ceremony. Because of the unique and recognizable impressions of each design, the movements of complicated stamped vessels could be more easily traced than among other types of vessels. Finally, vessels di ed, either during use at villages or as they were deliberately dispatched at mound ceremonies. The spirit of the object may have been released through the death of its image in salient acts of destruction (Taussig 1993:135). In light of their importance in mortuary ceremony, where webs of social relations between both living and deceased persons were (re)construc ted and differences between persons in terms of kinship and status were given currency (e.g., Joyce 2001), exchange d complicated stamped vessels represent a recontextua lization and inversion of the si gnificance of ordinary cooking pots. At mounds, among a corpus of special-use vessels th at were not used in daily life, were the very same cooking pots that were strewn across the garbage heaps at distant village middens. Exchanged vessels were taken out of the corpus of stamped vessels bearing the same (clan?) design within a village, and in this way a part was extricated from the social body of the descent group. The everyday ubiquity of a design on cooki ng vessels within the vi llages of a particular descent group was likely part of its biographic po wer in manifesting a distributed person, making reference not just to an ethereal ancestor but to an organic, living, working descent group. As 270

PAGE 271

parts of distributed persons, exchanged vessels coul d quite literally give bodily presence to social persons who were not otherwise pr esent within mortuary m ounds. The use of complicated stamped vessels as enlivened descent emblems seems to have begun with the initial adoption of Swift Creek pottery on the Lower St. Johns River, but the spatial scale and intensity of their exchange, al ong with the clarity of designs, increased during the Late Swift Creek phase. In the context of incipient social complexity, this escalation is presumably linked to increasing competition among descent groups that struggled over claims to ancestry and desirable alliances in marriage. Thus, marriage patterns and economic alliances are likely to have contributed to patterns of paddl e matches between distant sites (e.g., Stoltman and Snow 1998), but I contend that vessels carried across the landscape do not re present the de facto refuse of these practices. Rather, I find that u tilitarian cooking wares seem to have been caught up in political processes as cita tions of social relati onships during momentous mortuary events. This transformation of value for cooking pots was contextual and historica l: at a time and place where material references to places, peopl e, and genealogies became most important, complicated stamping on domestic cooking pots offere d material that was especially suitable for these purposes. This co-opti on and transformation of ostensib ly mundane material culture reinforces the arbitrariness of cultural attributions of value that anthropologists have long recognized. Just as the choice of seashells, banana leaf bundles, and cloth as important items of exchange in Melanesia are embedded in cultur al practice (Myers a nd Kirshenblatt-Gimblett 2001), so too are the reasons why cooking vessels became important items of exchange on the Lower St. Johns River. Future Directions I hope that this research has built a foundation fo r future studies of Swift Creek cultures as well as archaeological considerations of exchange more generally. To begin with the foregoing 271

PAGE 272

case study, several practical lim itations could be su ccessfully addressed with future work. In the nascent stages of the project, I set out to analyze pottery asse mblages from mortuary mound and village midden sites along the Atlantic coast from the St. Johns River to the Altamaha River. However, these goals were compromised when I was unable to locate mortuary mound assemblages from Waring and Holders (1968) ex cavations at the Evelyn site (9GN6) or any assemblages from Swift Creek mortuary sites alo ng the Georgia coast. This deficiency in the sample prevents comparison between mortuary assemblages and potential identification of exchanges of complicated stamped vessels in reci procation of vessels brought to the Lower St. Johns River. While many mounds have been dest royed and their co llections lost, a few mortuary sites remain for future excavations, including unexcavated portions of Mound B and Mound C at Evelyn and one or more potential Swift Cr eek mounds at Lewis Creek (Cook 1966). Another sampling limitation was introduced by the institutional restrictions placed on many collections. Because whole or especially ornate vessels were mostly off limits to destructive sampling by INAA and petrographic thin sections, specialized mortuary mound vessels were not represented in the chemical or mineralogical data. As a consequence, the samples from mounds were limited mostly to sa nd-tempered plain and Swift Creek Complicated Stamped cooking vessels and excluded the incise d and painted wares, as well as diminutive vessels that had been recovere d whole or had been completely reconstructed. The provenance portion of this research theref ore only discovered significant exch ange relationships between the Lower St. Johns River and Altamaha River. However, many of the unsampled specialized vessels may have been made along the Gulf Coas t (Crystal River and Weeden Island types) or Middle St. Johns River (St. Johns types) wh ere these pottery types are more common. 272

PAGE 273

Discovering provenance for these ostensibly non local vessels will require some destructive analysis as well as a large comparative sample of vesse ls and clays from suspected source areas. Even in the areas along the coast where patterns in chemical and mineralogical data were fairly conclusive, more clay samples would be beneficial for outlining the range of compositional variability within the region. Clays from along th e coast in the areas be tween the St. Johns and Altamaha rivers appear to be particularly variable in chemical composition, necessitating a much larger sample to begin to outline spatial distri butions of elements. Future work might also benefit from a greater focus on ceramic ecol ogy, sampling not only potential clay sources but also potential temper sources, particularly sa nd. Although quartz sands are likely to make only minimal contributions to the elemental com position of pottery (except in hafnium and zirconium), determining the geogr aphic availability of variou s grain sizes in itself might contribute to our understanding of tempering traditions along the coast. There is tremendous potential for future rese arch of Swift Creek in teraction and exchange in other areas of the lower Southeast and beyond. Future development of a searchable digital database of paddle matches is necessary for realiz ing this potential on a larg er scale. There have been some attempts to create a digital databa se of designs, but toda y Frankie Snows (2007) unpublished manuscript that compiles reconstructe d designs and paddle matches is the primary resource for design studies. The application of technology, such as fingerprinting software for design recognition, will be critical for making paddl e match identification more efficient and for standardizing design reconstructions. Ultimately, design data should be combined in a database with other vessel attributes such as vesse l morphology, and chemical and mineralogical composition. 273

PAGE 274

Finally, m ore excavations at Swift Creek village sites would be bene ficial for exploring intrasite patterns in the use of designs, and ultimate ly, to test inferences a bout the social structure and reproduction of descent groups that seem to have been the driving forces of vessel exchange, at least on the Atlantic co ast. Aside from work at Kings Bay (Saunders 1986, 1998), opportunities to identify and excavat e multiple discrete households have been limited. Restricted budgets often restrict the scope of work on thre atened sites and ethical considerations limit excavations at protected sites; nonetheless, thorough excavation of a multi-household Swift Creek village holds great potential to build on our understandings of social structure. For future studies of the Swif t Creek archaeological culture, I advocate analyses of large assemblages across many sites in order to reconstruct the contextual details of material practice. I have suggested that the data pr esented in this study indicate the exchange of used village cooking vessels as indexes of persons and de scent groups that operated as material for constituting social relationships on momentous occasions at mort uary mounds. However, these material practices were not necessarily shar ed among other Swift Creek populations across the lower Southeast. I have emphasized that comp licated stamping was embraced on the Lower St. Johns in specific ways that made cooking vesse ls appropriate as mean ingful citations for reworking social relationships. Given the lack of data from mortuary mounds along the Altamaha River, the region where exchanged Swift Creek vessels on the Lower St. Johns derived, I have been reticent to make conjectures about the ways material culture at Altamaha River mounds may have been mobilized in recipro cal ways. It may be even less likely in other areas of the Eastern Woodlands that Swift Creek pottery was di stributed by similar social practices with analogous logics of value. Swift Creek Complicated Stamped pottery was a global-scale phenomenon that was incorporated differently across space and through time. To 274

PAGE 275

275 assert that Swift Creek pottery was moved by the same practices or had similar meanings across thousands of kilometers ignores the possibili ties of recontextualization, whereby even mass produced objects in a globalized market take on local connotations of identity (e.g., Miller 1995b). In the particular case of populations on th e Lower St. Johns River, I have argued that the indexical capabiliti es of cooking vessels as vehicl es of distributed personhood were appropriated for constructing webs of social connections between descent groups. With continued empirical research, I believe we w ill find that Swift Creek Complicated Stamped vessels were frequently exchanged by other populati ons as well, but that th e specific ways that the items were mobilized and the manifestation of their citational potential were quite variable with context.

PAGE 276

APPENDIX A INSTRUME NTAL NEUTRON ACTIVATION ANALYSIS 276

PAGE 277

Table A-1. Sample Provenience, Type, and Group Membership. anid PID PMD field_id Vessel # site cer_type Paste Group NJW-001 1 A3-9 9 8DU68 CCS charcoal 1 NJW-002 A3-13 13 8DU68 CCS charcoal 1 NJW-003 A3-21 21 8DU68 CP charcoal 1 NJW-004 A3-27 26 8DU68 STP fine sand 1 NJW-005 A3-27 27 8DU68 CP charcoal 1 NJW-006 A3-30 30 8DU68 CP charcoal 1 NJW-007 2 88.8.607 33 8DU68 SWCRCS coarse sand 2 NJW-008 3 88.8.613 34 8DU68 SWCRCS coarse sand unasO NJW-009 88.8.595 35 8DU68 SWCRCS medium sand 2 NJW-010 4 36 88.8.396 37 8DU68 SWCRCS medium sand 2 NJW-011 88.8.204 40 8DU68 STP find sand and light charcoal 1 NJW-012 88.8.411 44 8DU68 SWCRCS (CROOKED RIVER) coarse sand 2 NJW-013 88.8.631 48 8DU68 CP fine sand and moderate charcoal unasO NJW-014 88.8.200 49 8DU68 STP find sand and light charcoal 1 NJW-015 88.8.203 50 8DU68 SWCRCS fine sand and moderate charcoal 1 NJW-016 88.8.203 52 8DU68 SWCRCS medium sand 1 NJW-017 5 88.8.632 54 8DU68 STP fine sand 1 NJW-018 88.8.406 55 8DU68 CCS fine sand and heavy charcoal 1 NJW-019 88.8.627 58 8DU68 STP fine sand 1 NJW-020 88.8.543 59 8DU68 STP find sand and light charcoal 1 NJW-021 6 291 88.8.565 60 8DU68 SWCRCS fine sand 1 NJW-022 88.8.471 62 8DU68 STP none 1 NJW-023 88.8.336 64 8DU68 SWCRCS fine sand 1 NJW-024 n/a 65 8DU68 STP fine sand unasO NJW-025 2005 #25 n/a 66 8DU68 CP fine sand and heavy charcoal 1 NJW-026 2005 #24 n/a 68 8DU68 CP charcoal 1 NJW-027 32 34 88.8.232 67 8DU68 SWCRCS coarse sand 2 NJW-314 n/a 66 8DU68 CP fine sand and heavy charcoal 1 NJW-028 103001 1 8DU96 SWCRCS fine sand unasO NJW-029 2004 x 103002 3 8DU96 GROG-TEMPERED PLAIN fine sand and medium grog 1 NJW-030 103003 4 8DU96 STP coarse sand unas1 NJW-031 103005 7 8DU96 STP coarse sand unasO NJW-032 2005 #1 103006 8 8DU96 CP fine sand and moderate charcoal 1 277

PAGE 278

Table A-1. Continued anid PID PMD field_id Vessel # site cer_type Paste Group NJW-033 103007 9 8DU96 STP fine sand 1 NJW-034 103009 11 8DU96 STP coarse sand 2 NJW-035 103011 13 8DU96 MDS fine sand 1 NJW-036 103014 16 8DU96 STP coarse sand 2 NJW-037 103015 17 8DU96 ST. JOHNS PLAIN spiculate unasO NJW-038 2004 x 38 103016 18 8DU96 SWCRCS medium sand 2 NJW-039 103017 20 8DU96 STP coarse sand 1 NJW-040 103020 24 8DU96 ST. JOHNS PLAIN spiculate unasO NJW-041 103021 25 8DU96 SWCRCS fine sand unas2 NJW-042 103022 26 8DU96 SWCRCS fine sand 1 NJW-043 2005 #2 105196 29 8DU96 CCS fine sand and moderate charcoal 1 NJW-044 105196 30 8DU96 CCS fine sand and moderate charcoal 1 NJW-045 103025 33 8DU96 DEPTFORD CHECK STAMPED fine sand 1 NJW-046 103024 38 8DU96 MDS fine sand 1 NJW-047 2005 #26 103025 37 8DU96 CP fine sand and light charcoal 1 NJW-313 2005 #3 103025 41 8DU96 CP fine sand and moderate charcoal unas1 NJW-048 140 3 8NA32 SWCRCS ( CROOKED RIVER) coarse sand 2 NJW-049 140 4 8NA32 STP coarse sand 2 NJW-050 33 44 6 8NA32 SWCRCS coarse sand 2 NJW-051 44 8 8NA32 SWCRCS (CROOKED RIVE R) coarse sand 2 NJW-052 44 9 8NA32 STP fine sand 1 NJW-053 34 surface 10 8NA32 SWCRCS (NEW RIVER) fine sand unas1 NJW-054 surface 11 8NA32 SWCRCS coarse sand 2 NJW-055 35 49 13 8NA32 WI fine sand unasO NJW-056 surface 15 8NA32 SWCRCS fine sand 1 NJW-057 143 16 8NA32 SWCRCS coarse sand 2 NJW-058 118 17 8NA32 STP fine sand unasO NJW-059 36 148 19 8NA32 SWCRCS coarse sand 2 NJW-060 148 20 8NA32 SWCRCS coarse sand 2 NJW-061 130 21 8NA32 SWCRCS fine sand and light bone 1 NJW-062 130 22 8NA32 SWCRCS fine sand 1 NJW-063 37 130 23 8NA32 SWCRCS coarse sand 2 278

PAGE 279

Table A-1. Continued anid PID PMD field_id Vessel # site cer_type Paste Group NJW-064 38 118 27 8NA32 SWCRCS coarse sand 2 NJW-066 138 31 8NA32 SWCRCS coarse sand 1 NJW-067 138 32 8NA32 STP fine sand and light charcoal 1 NJW-068 143 33 8NA32 SWCRCS coarse sand 2 NJW-069 142 34 8NA32 SWCRCS fine sand unasO NJW-070 139 35 8NA32 SWCRCS coarse sand 2 NJW-071 144 36 8NA32 SWCRCS fine sand unasO NJW-072 2005 #8 140 37 8NA32 CP fine sand and moderate charcoal unasO NJW-073 2005 #9 126 38 8NA32 CP fine sand/charcoal/ light bone 1 NJW-074 77 2 8DU5543 CP fine sand and heavy charcoal 1 NJW-075 203 14 8DU5543 CCS fine sand and moderate charcoal 1 NJW-076 188 31 8DU5543 CCS fine sand and moderate charcoal 1 NJW-077 152 44 8DU5543 SWCRCS fine sand 1 NJW-078 20b 73 43 8DU5543 CCS fine sand and moderate charcoal 1 NJW-079 16 224 70 8DU5543 SWCRCS coarse sand unasO NJW-080 216 71 8DU5543 CCS fine sand and moderate charcoal 1 NJW-081 17 185 72 8DU5543 CCS fine sand 1 NJW-082 67 79 8DU5543 STP fine sand 1 NJW-083 67a 80 8DU5543 STP medium sand 1 NJW-084 67a 93 8DU5543 SWCRCS fine sand 1 NJW-085 18 4 99 8DU5543 SWCRCS fine sand 1 NJW-086 67a 97 8DU5543 SWCRCS fine sand 1 NJW-087 67 101 8DU5543 SWCRCS fine sand 1 NJW-088 67a 105 8DU5543 STP fine sand 1 NJW-089 228 119 8DU5543 CP fine sand and moderate charcoal 1 NJW-090 65 139 8DU5543 SWCRCS fine sand 1 NJW-091 20 65 143 8DU5543 STP fine sand 1 NJW-092 190 162 8DU5543 CP fine sand and heavy charcoal 1 NJW-093 19 228 166 8DU5543 CCS fine sand and heavy charcoal unasO NJW-094 83 170 8DU5543 CCS fine sand and heavy charcoal 1 NJW-095 10 A 8DU5543 SWCRCS fine sand 1 279

PAGE 280

Table A-1. Continued anid PID PMD field_id Vessel # site cer_type Paste Group NJW-096 83 B 8DU5543 STP fine sand 1 NJW-098 2005 #23 33 D 8DU5543 CCS fine sand and moderate charcoal 1 NJW-099 212 103 8DU5543 SWCRCS fine sand 1 NJW-100 surface 2 8DU17245 STSS fine sand 1 NJW-101 surface 4 8DU17245 STP fine sand 1 NJW-102 7 78 22 8DU17245 WIR coarse sand unasO NJW-103 78 26 8DU17245 CP fine sand and heavy charcoal 1 NJW-104 78 27 8DU17245 STP fine sand 1 NJW-105 86 29 8DU17245 CCS fine sand and moderate charcoal 1 NJW-106 86 32 8DU17245 CP fine sand and moderate charcoal 1 NJW-107 surface 34 8DU17245 CARABELLE fine sand unasO NJW-108 surface 36 8DU17245 ST JOHNS spiculate unasO NJW-109 92 40 8DU17245 CP fine sand and heavy charcoal 1 NJW-110 190 43 8DU17245 STP fine sand 1 NJW-111 8 191 45 8DU17245 SWCRCS medium sand 1 NJW-112 9 211 54 8DU17245 CP fine sand and moderate charcoal 1 NJW-113 231 63 8DU17245 CCS fine sand and heavy charcoal 1 NJW-114 234 64 8DU17245 CCS fine sand and heavy charcoal 1 NJW-115 234 65 8DU17245 CCS fine sand and moderate charcoal 1 NJW-116 240 67 8DU17245 CP fine sand and moderate charcoal 1 NJW-117 269 76 8DU17245 SWCRCS fine sand 1 NJW-118 269 77 8DU17245 DCR spiculate unasO NJW-119 10 278 79 8DU17245 STP fine sand 1 NJW-120 305 92 8DU17245 CCS fine sand and heavy charcoal 1 NJW-121 196 100 8DU17245 CCS fine sand and moderate charcoal 1 NJW-122 11 197 102 8DU17245 CCS fine sand and heavy charcoal 1 NJW-123 317 112 8DU17245 SWCRCS fine sand 1 NJW-124 345 123 8DU17245 CP fine sand and moderate charcoal 1 NJW-125 357 131 8DU17245 CP fine sand and moderate charcoal unasO NJW-126 357 132 8DU17245 CP fine sand and heavy charcoal 1 NJW-127 388 140 8DU17245 CP fine sand and heavy charcoal 1 280

PAGE 281

Table A-1. Continued anid PID PMD field_id Vessel # site cer_type Paste Group NJW-128 46 1 8DU14683 STP fine sand unasO NJW-130 46 3 8DU14683 STP fine sand 1 NJW-131 45 4 8DU14683 ST JOHNS PLAIN spiculate unasO NJW-132 45 5 8DU14683 STP fine sand 1 NJW-133 30 1 8DU14686 STP fine sand unasO NJW-134 32 2 8DU14686 STP fine sand unasO NJW-135 34 3 8DU14686 STP fine sand unasO NJW-136 12 34 4 8DU14686 SWCRCS medium sand/ moderate charcoal 1 NJW-137 36 5 8DU14686 STP fine sand and moderate charcoal 1 NJW-138 13 39 6 8DU14686 STP fine sand 1 NJW-139 14 42 8 8DU14686 STP fine sand 1 NJW-140 43 10 8DU14686 STP fine sand 1 NJW-141 43 11 8DU14686 SWCRCS fine sand unasO NJW-142 43 12 8DU14686 STP fine sand unasO NJW-143 44 13 8DU14686 STP fine sand unasO NJW-144 44 14 8DU14686 SWCRCS fine sand and light charcoal 1 NJW-145 15 44 15 8DU14686 SWCRCS fine sand 1 NJW-146 2 3 8DU5544/5 SWCRCS fine sand 1 NJW-147 2 10 8DU5544/5 SWCRCS coarse sand 2 NJW-148 4 22 8DU5544/5 STP fine sand 1 NJW-149 10 56 8DU5544/5 SWCRCS fine sand 1 NJW-150 11 65 8DU5544/5 STP fine sand 1 NJW-151 22 13 71 8DU5544/5 CP fine sand and moderate charcoal 1 NJW-152 21 291 19 96 8DU5544/5 SWCRCS fine sand 1 NJW-153 13 99 8DU5544/5 SWCRCS fine sand 1 NJW-154 22 100 8DU5544/5 SWCRCS fine sand 1 NJW-155 22 103 8DU5544/5 STP fine sand 1 NJW-156 22 151 8DU5544/5 ST JOHNS PLAIN spiculate unasO NJW-157 25 109 8DU5544/5 STP fine sand 1 NJW-158 23 26 117 8DU5544/5 STP coarse sand 2 NJW-159 26 118 8DU5544/5 SWCRCS coarse sand 1 NJW-160 38 187 8DU5544/5 SWCRCS fine sand 1 281

PAGE 282

Table A-1. Continued anid PID PMD field_id Vessel # site cer_type Paste Group NJW-161 41 200 8DU5544/5 SWCRCS fine sand 1 NJW-163 49 231 8DU5544/5 STP medium sand unasO NJW-164 24 53 241 8DU5544/5 STP fine sand 1 NJW-165 67 262 8DU5544/5 STP fine sand 1 NJW-166 93 330 8DU5544/5 STP medium sand 1 NJW-167 118 368 8DU5544/5 SWCRCS fine sand 1 NJW-168 118 372 8DU5544/5 CCS-SUN CITY fine sand and light charcoal 1 NJW-169 25 291 124 400 8DU5544/5 SWCRCS fine sand 1 NJW-170 1 8DU14 SWCRCS fine sand 1 NJW-171 2 8DU14 SWCRCS fine sand unasO NJW-172 3 8DU14 SWCRCS fine sand 1 NJW-173 5 9MC372 SWCRCS medium sand unas1 NJW-174 26 34 7 9MC372 SWCRCS medium sand 2 NJW-175 10 9MC372 STP coarse sand 2 NJW-176 12 9MC372 SWCRCS fine sand unas2 NJW-177 27 36 16 9MC372 SWCRCS fine sand unas1 NJW-178 21 9MC372 INCISED medium sand 2 NJW-179 23 9MC372 STP fine sand unas1 NJW-180 27 9MC372 STP medium sand 2 NJW-181 40 9MC372 SWCRCS medium sand 2 NJW-182 43 9MC372 STP medium sand unas2 NJW-183 46 9MC372 STP/UID coarse sand 2 NJW-184 51 9MC372 SWCRCS fine sand unasO NJW-185 54 9MC372 STP medium sand unas2 NJW-186 55 9MC372 STP coarse sand 2 NJW-187 57 9MC372 SWCRCS medium sand 2 NJW-188 58 9MC372 STP fine sand unasO NJW-189 72 9MC372 SWCRCS fine sand 2 NJW-190 28 74 9MC372 SWCRCS medium sand 2 NJW-191 29 85 9MC372 STP coarse sand unas2 NJW-192 30 86 9MC372 SWCRCS fine sand unasO NJW-193 31 89 9MC372 SWCRCS coarse sand 2 282

PAGE 283

Table A-1. Continued anid PID PMD field_id Vessel # site cer_type Paste Group NJW-194 94 9MC372 STP coarse sand 2 NJW-195 39 36 950 A 9MC360 SWCRCS medium sand 2 NJW-196 40 38 913 B 9MC360 SWCRCS coarse sand 2 NJW-197 41 974 C 9MC360 SWCRCS coarse sand 2 NJW-198 42 893 D 9MC360 SWCRCS coarse sand 2 NJW-199 43 766 E 9MC360 SWCRCS coarse sand unasO NJW-200 950 F 9MC360 SWCRCS coarse sand unas2 NJW-201 950 G 9MC360 SWCRCS coarse sand 2 NJW-202 766 H 9MC360 SWCRCS fine sand unasO NJW-203 766 I 9MC360 SWCRCS coarse sand 2 NJW-204 766 J 9MC360 SWCRCS coarse sand 2 NJW-205 1005 K 9MC360 SWCRCS coarse sand 2 NJW-206 1005 L 9MC360 SWCRCS medium sand 2 NJW-207 1005 M 9MC360 STP medium sand 2 NJW-208 1005 N 9MC360 STP coarse sand unas2 NJW-209 887 O 9MC360 SWCRCS coarse sand 2 NJW-210 887 P 9MC360 STP coarse sand unasO NJW-211 893 Q 9MC360 SWCRCS coarse sand 2 NJW-212 893 R 9MC360 SWCRCS medium sand 2 NJW-213 974 S 9MC360 SWCRCS coarse sand 2 NJW-214 974 T 9MC360 SWCRCS coarse sand 2 NJW-215 974 U 9MC360 STP coarse sand 2 NJW-216 893 V 9MC360 STP coarse sand unas2 NJW-217 893 W 9MC360 STP coarse sand 2 NJW-218 44 A 9GN6 STP fine sand 1 NJW-219 45 B 9GN6 SWCRCS coarse sand 2 NJW-220 46 C 9GN6 SWCRCS medium sand and moderate grog 1 NJW-221 47 D 9GN6 SWCRCS coarse sand 2 NJW-222 48 E 9GN6 STP fine sand 2 NJW-223 F 9GN6 SWCRCS coarse sand 2 NJW-224 G 9GN6 SWCRCS coarse sand 2 NJW-225 H 9GN6 SWCRCS coarse sand 2 NJW-226 I 9GN6 SWCRCS coarse sand 2 283

PAGE 284

Table A-1. Continued anid PID PMD field_id Vessel # site cer_type Paste Group NJW-227 J 9GN6 SWCRCS coarse sand 2 NJW-228 K 9GN6 SWCRCS coarse sand unas2 NJW-229 L 9GN6 SWCRCS coarse sand 2 NJW-230 M 9GN6 SWCRCS coarse sand 2 NJW-231 N 9GN6 SWCRCS coarse sand unas2 NJW-232 O 9GN6 SWCRCS fine sand unasO NJW-233 P 9GN6 SWCRCS coarse sand 2 NJW-234 Q 9GN6 SWCRCS coarse sand unas1 NJW-235 R 9GN6 SWCRCS medium sand 2 NJW-236 S 9GN6 SWCRCS coarse sand 2 NJW-237 T 9GN6 SWCRCS coarse sand 2 NJW-238 U 9GN6 SWCRCS coarse sand unasO NJW-239 V 9GN6 SWCRCS coarse sand 2 NJW-240 W 9GN6 STP medium sand unasO NJW-241 49 36 2 9MC16 SWCRCS coarse sand 2 NJW-242 2004 x 38 1 9MC16 SWCRCS medium sand 2 NJW-243 3 9MC16 SWCRCS fine sand 2 NJW-244 4 9MC16 SWCRCS coarse sand 2 NJW-245 5 9MC16 SWCRCS coarse sand 2 NJW-246 6 9MC16 SWCRCS coarse sand 2 NJW-247 7 9MC16 STP medium sand 2 NJW-248 8 9MC16 SWCRCS coarse sand 2 NJW-249 9 9MC16 SWCRCS fine sand 2 NJW-250 10 9MC16 SWCRCS coarse sand 2 NJW-251 11 9MC16 SWCRCS coarse sand 2 NJW-252 12 9MC16 SWCRCS coarse sand 2 NJW-253 14 9MC16 SWCRCS fine sand and moderate bone 2 NJW-254 15 9MC16 SWCRCS coarse sand 2 NJW-255 16 9MC16 SWCRCS coarse sand 2 NJW-256 17 9MC16 SWCRCS coarse sand 2 NJW-257 50 18 9MC16 SWCRCS coarse sand 2 NJW-258 21 9MC16 SWCRCS coarse sand 2 NJW-259 22 9MC16 SWCRCS fine sand unas1 284

PAGE 285

Table A-1. Continued anid PID PMD field_id Vessel # site cer_type Paste Group NJW-260 51 23 9MC16 SWCRCS coarse sand unas1 NJW-261 25 9MC16 SWCRCS coarse sand 2 NJW-262 26 9MC16 SWCRCS coarse sand unas2 NJW-263 36 9MC16 SWCRCS medium sand unasO NJW-264 1 9CM171 SWCRCS fine sand and moderate bone 1 NJW-265 2 9CM171 SWCRCS coarse sand 1 NJW-266 4 9CM171 SWCRCS fine sand 1 NJW-267 5 9CM171 SWCRCS medium sand 1 NJW-268 1 Kings Lake SWCRCS coarse sand unasO NJW-269 2 Kings Lake SWCRCS coarse sand unasO NJW-270 3 Kings Lake SWCRCS coarse sand unasO NJW-271 4 Kings Lake SWCRCS coarse sand unasO NJW-272 5 Kings La ke SWCRCS coarse sand 2 NJW-273 6 Kings Lake SWCRCS coarse sand unasO NJW-274 7 Kings La ke SWCRCS coarse sand 2 NJW-275 8 Kings Lake SWCRCS fine sand unasO NJW-276 10 Kings Lake SWCRCS coarse sand unasO NJW-277 52 11 Kings Lake SWCRCS coarse sand unasO NJW-278 12 Kings Lake SWCRCS coarse sand unas1 NJW-279 53 13 Kings Lake SWCRCS coarse sand unasO NJW-280 14 Kings Lake SWCRCS coarse sand unasO NJW-281 54 15 Kings La ke SWCRCS coarse sand 2 NJW-282 16 Kings Lake SWCRCS medium sand unas1 NJW-283 17 Kings La ke SWCRCS coarse sand 2 NJW-284 18 Kings Lake SWCRCS medium sand unas1 NJW-285 19 Kings Lake SWCRCS coarse sand unasO NJW-286 20 Kings Lake NET IMPRESSED coarse sand unasO NJW-287 1 9WY8 STP medium sand 1 NJW-288 2 9WY8 STP medium sand unasO NJW-289 3 9WY8 SWCRCS coarse sand unasO NJW-290 4 9WY8 SWCRCS coarse sand 2 NJW-291 5 9WY8 SWCRCS coarse sand unas1 NJW-292 6 9WY8 SWCRCS coarse sand 2 285

PAGE 286

Table A-1. Continued anid PID PMD field_id Vessel # site cer_type Paste Group NJW-293 55 7 9WY8 SWCRCS coarse sand 2 NJW-294 8 9WY8 SWCRCS coarse sand unasO NJW-295 1 Oak Landing SWCRCS coarse sand unas1 NJW-296 2 Oak Landi ng SWCRCS coarse sand 2 NJW-297 3 Oak Landing STP coarse sand unasO NJW-298 1 9CM25 SWCRCS fine sand 1 NJW-299 2 9CM25 SWCRCS medium sand 1 NJW-300 3 9CM25 SWCRCS medium sand unas1 NJW-301 4 9CM25 SWCRCS/ WI PUNCTATED coarse sand 2 NJW-302 5 9CM25 SWCRCS coarse sand 2 NJW-303 6 9CM25 SWCRCS fine sand unas1 NJW-304 56 7 9CM25 SWCRCS fine sand 1 NJW-305 8 9CM25 SWCRCS coarse sand 2 NJW-306 9 9CM25 SWCRCS fine sand 1 NJW-307 57 10 9CM25 SWCRCS coarse sand 2 NJW-308 11 9CM25 SWCRCS coarse sand unasO NJW-309 12 9CM25 SWCRCS coarse sand unasO NJW-310 13 9CM25 SWCRCS coarse sand unas1 NJW-311 14 9CM25 SWCRCS medium sand 1 NJW-312 15 9CM25 SWCRCS medium sand unasO NJW-315 1 New Smyrna CLAY n/a unasO NJW-316 2 Green Spring CLAY n/a unasO NJW-317 58 3 8DU14 Feature 1 CLAY n/a 3* NJW-318 59 4 Oxeye Island CLAY n/a 3* NJW-319 60 5 8DU1 CLAY n/a 3* NJW-320 61 6 Amelia Is. CLAY n/a unasO NJW-321 62 7 Nassau Sound CLAY n/a 3* NJW-322 8 St. Mary's River CLAY n/a unasO NJW-323 9 Osceola Forest CLAY n/a unasO NJW-324 63 10 9CM157 CLAY n/a unasO NJW-325 11 9CM157 CLAY n/a unasO NJW-326 12 China Hill CLAY n/a 5* NJW-327 64 13 9TF115 CLAY n/a 5* 286

PAGE 287

287Table A-1. Continued *provisional clay chemical group anid PID PMD field_id Vessel # site cer_type Paste Group NJW-328 14 Near 9TF115 CLAY n/a 5* NJW-329 65 15 Jekyll Is. (south) CLAY n/a unasO NJW-330 16 Jekyll Is. (north) CLAY n/a unasO NJW-331 66 17 Altamaha River CLAY n/a 4* NJW-332 67 18 Lower Sansavilla CLAY n/a 4* NJW-333 19 Lower Sansavilla CLAY n/a unasO KHA089 A 38CH42 CLAY n/a unasO KHA033 B Deen's Landing CLAY n/a 5* KHA088 C 8LE151 CLAY n/a unasO KHA091 D Amelia Island CLAY n/a 3* anid INAA analytical identification nu mber CCS Early Swift Creek Complicated Stamped (charco al-tempered) PID petrographic identification number STP sand-tempered plain PMD paddle match design CP charcoal-tempered plain (Early Swift C reek) field id field or laboratory sp ecimen number MDS Mayport Dentate Stamped vessel # -vessel number within site assemblage CARABELLE Carrabelle Puncatated cer type ceramic typological designation SWCRCS Swift Creek Complicated Stamped paste temper characterization WI Weeden Island Incised group INAA chemical group WIR Weeden Island Red

PAGE 288

Table A-2. Mahalanobis dist ance-based probabilities of group m embership for Group 1 members. Anid G1 G2 NJW001 94.272 0.000 NJW002 99.762 0.000 NJW003 94.268 0.000 NJW004 27.236 0.000 NJW005 95.970 0.000 NJW006 86.618 0.000 NJW011 52.544 0.000 NJW014 64.397 0.001 NJW015 99.240 0.001 NJW016 32.561 0.000 NJW017 62.337 0.000 NJW018 78.268 0.000 NJW019 77.171 0.000 NJW020 44.461 0.002 NJW021 39.239 0.000 NJW022 49.458 0.000 NJW023 15.259 0.000 NJW025 87.625 0.000 NJW026 5.044 0.000 NJW029 27.756 0.000 NJW032 3.129 0.000 NJW033 91.034 0.000 NJW035 18.153 0.000 NJW039 2.839 0.000 NJW042 97.946 0.000 NJW043 90.915 0.000 NJW044 99.691 0.001 NJW045 10.863 0.000 NJW046 46.285 0.001 NJW047 14.801 0.000 NJW052 79.438 0.000 NJW056 82.380 0.000 NJW061 79.283 0.000 NJW062 93.793 0.000 NJW066 51.435 0.004 NJW067 43.479 0.000 NJW073 79.292 0.000 NJW074 84.976 0.000 NJW075 75.758 0.000 NJW076 46.845 0.000 NJW077 94.278 0.005 NJW078 97.742 0.000 NJW080 87.719 0.020 NJW081 23.341 0.000 NJW082 3.926 0.000 NJW083 78.123 0.000 NJW084 81.806 0.000 NJW085 5.665 0.000 NJW086 40.531 0.001 NJW087 19.728 0.000 NJW088 43.097 0.000 288

PAGE 289

Table A-2. Continued Anid G1 G2 NJW089 44.536 0.000 NJW090 72.432 0.000 NJW091 90.893 0.000 NJW092 27.002 0.000 NJW094 75.708 0.002 NJW095 86.137 0.000 NJW096 11.240 0.000 NJW097 20.812 0.000 NJW098 94.732 0.000 NJW099 73.547 0.000 NJW100 12.607 0.000 NJW101 9.085 0.000 NJW103 47.588 0.000 NJW104 1.408 0.000 NJW105 19.889 0.000 NJW106 34.654 0.000 NJW109 9.638 0.000 NJW110 88.163 0.001 NJW111 11.965 0.000 NJW112 14.690 0.000 NJW113 88.650 0.000 NJW114 57.543 0.000 NJW115 23.963 0.000 NJW116 32.685 0.000 NJW117 97.577 0.000 NJW119 57.973 0.000 NJW120 19.681 0.000 NJW121 65.782 0.000 NJW122 99.675 0.000 NJW123 21.082 0.000 NJW124 83.548 0.000 NJW126 81.370 0.000 NJW127 2.640 0.000 NJW129 47.083 0.000 NJW130 15.003 0.000 NJW132 11.012 0.000 NJW136 9.777 0.000 NJW137 4.196 0.000 NJW138 58.569 0.000 NJW139 47.130 0.000 NJW140 33.524 0.000 NJW144 51.383 0.000 NJW145 5.548 0.000 NJW146 55.143 0.000 NJW148 38.591 0.000 NJW149 31.178 0.000 NJW150 26.905 0.000 NJW151 86.215 0.205 NJW152 79.460 0.000 NJW153 95.127 0.000 NJW154 31.060 0.000 NJW155 66.154 0.000 289

PAGE 290

Table A-2. Continued Anid G1 G2 NJW157 81.181 0.000 NJW159 2.983 0.000 NJW160 24.466 0.000 NJW161 69.969 0.000 NJW162 12.412 0.000 NJW164 4.388 0.000 NJW165 31.658 0.001 NJW166 99.537 0.000 NJW167 47.952 0.000 NJW168 11.907 0.000 NJW169 98.276 0.000 NJW170 35.645 0.000 NJW172 2.451 0.000 NJW218 75.946 0.000 NJW220 14.428 0.068 NJW264 97.460 0.006 NJW265 2.281 0.000 NJW266 3.353 0.000 NJW267 11.925 0.000 NJW287 25.489 0.001 NJW298 7.510 0.000 NJW299 51.360 0.000 NJW304 34.393 0.000 NJW306 27.933 0.000 NJW311 62.021 0.000 NJW314 84.150 0.000 290

PAGE 291

Table A-3. Mahalanobis dist ance-based probabilities of group m embership for Group 2 members. Anid G1 G2 NJW007 1.473 2.755 NJW009 4.760 96.627 NJW010 4.613 57.584 NJW012 0.037 13.105 NJW027 0.000 14.289 NJW034 0.000 13.595 NJW036 0.030 8.890 NJW038 0.633 46.987 NJW048 0.557 48.477 NJW049 2.111 26.768 NJW050 0.719 49.109 NJW051 0.002 2.433 NJW054 0.021 9.757 NJW057 0.101 54.325 NJW059 0.000 1.401 NJW060 0.175 95.743 NJW063 0.929 60.461 NJW064 3.647 19.648 NJW065 0.206 97.040 NJW068 0.071 22.493 NJW070 0.319 64.721 NJW147 0.001 72.909 NJW158 4.749 93.725 NJW174 0.278 98.503 NJW175 0.000 1.146 NJW178 0.061 69.048 NJW180 0.000 7.299 NJW181 0.025 34.819 NJW183 0.029 97.297 NJW186 0.000 83.582 NJW187 0.003 98.291 NJW189 0.713 78.323 NJW190 0.775 70.017 NJW193 0.009 96.410 NJW194 1.674 67.834 NJW195 3.247 60.579 NJW196 0.242 99.718 NJW197 0.003 0.745 NJW198 0.000 5.332 NJW201 0.044 91.776 NJW203 0.102 37.488 NJW204 0.080 97.173 NJW205 0.773 92.469 NJW206 0.001 64.938 NJW207 0.253 78.920 NJW209 0.649 88.646 NJW211 0.000 81.456 NJW212 0.013 19.174 NJW213 0.000 16.873 NJW214 0.000 9.287 291

PAGE 292

Table A-3. Continued Anid G1 G2 NJW215 0.077 87.604 NJW217 0.002 24.629 NJW219 0.024 95.783 NJW221 0.004 56.178 NJW222 2.772 42.294 NJW223 0.000 62.012 NJW224 0.025 14.533 NJW225 0.000 77.808 NJW226 3.275 66.073 NJW227 0.001 83.203 NJW229 0.017 77.534 NJW230 0.038 14.951 NJW233 0.025 2.705 NJW235 0.368 53.801 NJW236 0.015 84.480 NJW237 0.002 73.528 NJW239 0.001 13.273 NJW241 0.000 57.629 NJW242 2.402 44.020 NJW243 0.000 98.365 NJW244 0.000 3.349 NJW245 0.029 80.294 NJW246 2.519 44.032 NJW247 0.000 3.662 NJW248 0.009 35.024 NJW249 0.000 56.740 NJW250 0.000 40.935 NJW251 0.000 43.172 NJW252 0.000 4.735 NJW253 0.000 9.223 NJW254 0.000 83.641 NJW255 0.138 94.466 NJW256 0.617 39.922 NJW257 0.001 51.794 NJW258 0.000 78.933 NJW261 0.264 18.195 NJW272 0.000 5.238 NJW274 0.003 9.903 NJW281 0.000 17.437 NJW283 0.736 68.639 NJW290 0.111 91.651 NJW292 0.018 61.207 NJW293 0.000 2.200 NJW296 0.008 13.456 NJW301 0.343 32.881 NJW302 0.082 54.529 NJW305 0.198 72.107 NJW307 0.236 95.747 292

PAGE 293

Table A-4. Mahalanobis dist ance-based projections of gr oup m embership probability for unassigned specimens. Anid G1 G2 BEST GROUP NJW008 0.000 0.000 1 NJW013 0.648 0.000 1 NJW024 0.000 0.000 1 NJW028 0.213 0.000 1 NJW030 21.441 1.205 1 NJW031 0.000 0.000 1 NJW037 0.000 0.000 1 NJW040 0.000 0.000 1 NJW041 10.520 73.683 2 NJW053 33.021 10.055 1 NJW055 0.000 0.000 1 NJW058 0.316 0.000 1 NJW069 0.057 0.000 1 NJW071 0.061 0.000 1 NJW072 0.311 0.000 1 NJW079 0.007 0.063 2 NJW093 0.229 0.000 1 NJW102 0.030 0.000 1 NJW107 0.000 0.000 1 NJW108 0.000 0.000 1 NJW118 0.000 0.000 1 NJW125 0.891 0.000 1 NJW128 0.000 0.000 1 NJW131 0.000 0.000 1 NJW133 0.000 0.000 1 NJW134 0.000 0.000 1 NJW135 0.247 0.000 1 NJW141 0.000 0.000 1 NJW142 0.605 0.000 1 NJW143 0.001 0.000 1 NJW156 0.000 0.000 1 NJW163 0.005 0.000 1 NJW171 0.001 0.003 2 NJW173 41.463 9.191 1 NJW176 24.529 79.132 2 NJW177 15.677 3.444 1 NJW179 57.552 12.922 1 NJW182 14.316 62.401 2 NJW184 0.000 0.001 2 NJW185 8.625 81.015 2 NJW188 0.003 0.000 1 NJW191 12.790 57.904 2 NJW192 0.005 0.000 1 NJW199 0.000 0.000 1 NJW200 6.320 19.998 2 NJW202 0.435 0.000 1 NJW208 5.712 42.015 2 NJW210 0.000 0.000 1 NJW216 7.689 99.936 2 NJW228 2.908 5.309 2 293

PAGE 294

Table A-4. Continued Anid G1 G2 BEST GROUP NJW231 14.698 98.222 2 NJW232 0.007 0.001 1 NJW234 44.434 59.185 2 NJW238 0.000 0.000 2 NJW240 0.820 0.006 1 NJW259 28.795 12.224 1 NJW260 19.003 0.072 1 NJW262 3.527 47.077 2 NJW263 0.781 1.274 2 NJW268 0.666 0.045 1 NJW269 0.000 0.000 2 NJW270 0.156 0.000 1 NJW271 0.000 0.000 1 NJW273 0.017 0.008 1 NJW275 0.475 0.016 1 NJW276 0.000 0.126 2 NJW277 0.007 0.000 1 NJW278 8.301 0.313 1 NJW279 0.000 0.000 1 NJW280 0.161 0.000 1 NJW282 4.611 1.285 1 NJW284 3.126 0.000 1 NJW285 0.224 0.000 1 NJW286 0.000 0.000 1 NJW288 0.039 0.000 1 NJW289 0.000 0.000 1 NJW291 12.803 1.256 1 NJW294 0.001 0.002 2 NJW295 2.699 2.152 1 NJW297 0.001 0.000 1 NJW300 64.501 11.854 1 NJW303 55.847 8.360 1 NJW308 0.059 0.773 2 NJW309 25.106 21.580 1 NJW310 3.125 1.207 1 NJW312 0.643 0.000 1 NJW313 0.541 0.000 1 294

PAGE 295

Table A-5. Mahalanobis distan ce-based projections of group m e mbership probability for clay samples. Anid G1 G2 BEST GROUP KHA033 0.000 0.000 1 KHA088 0.000 0.000 1 KHA089 0.000 0.000 1 NJW315 0.000 0.000 1 NJW316 0.000 0.000 1 NJW317 0.069 0.000 1 NJW318 0.000 0.000 1 NJW319 0.067 0.000 1 NJW320 0.000 0.000 1 NJW321 0.002 0.000 1 NJW322 0.000 0.000 1 NJW323 0.000 0.000 2 NJW324 0.000 0.000 2 NJW325 0.000 0.000 2 NJW326 0.000 0.000 2 NJW327 0.001 0.000 1 NJW328 0.025 0.000 1 NJW329 0.000 0.000 1 NJW330 0.000 0.000 1 NJW331 0.066 0.000 1 NJW332 0.000 0.000 1 NJW333 0.000 0.000 1 Table A-6. Eigenvalues and variance for the first ten principle components with simultaneous R-Q Factor Analysis based on variance-covariance matrix Eigenvalue %Variance Cum. %Var. 1 0.3201 39.1965 39.1965 2 0.1278 15.6542 54.8507 3 0.0772 9.4522 64.3029 4 0.0564 6.9079 71.2108 5 0.0514 6.2947 77.5055 6 0.0329 4.0236 81.5291 7 0.0278 3.4080 84.9371 8 0.0228 2.7922 87.7293 9 0.0201 2.4618 90.1911 10 0.0148 1.8065 91.9977 295

PAGE 296

296 Table A-7. Eigenvectors for each element (largest to smallest). As 0.1349 0.8489 0.0542 0.2027 0.2845 -0.0445 0.2762 -0.0477 0.0974 -0.0714 La 0.2066 -0.0297 0.1561 -0.0457 -0.0121 -0.0264 0.0157 0.0072 0.1607 0.2344 Lu 0.1789 -0.0489 0.1375 -0.0174 -0.0523 -0.0607 0.0779 0.0177 -0.1133 -0.2087 Nd 0.2403 -0.071 0.233 -0.0285 0.0242 0.0035 0.0166 -0.0393 0.1606 0.0729 Sm 0.2363 -0.0747 0.217 -0.0244 0.0315 0.0081 0.0274 -0.0531 0.0638 0.0015 U 0.0142 0.0991 0.0792 -0.1277 -0.3261 0.0534 -0.1641 -0.1454 -0.2691 -0.3412 Yb 0.1994 -0.0739 0.1761 -0.0257 -0.002 -0.0303 0.0743 0.0241 -0.0717 -0.134 Ce 0.2263 -0.0549 0.1856 -0.0488 0.0191 -0.0209 0.0242 -0.0045 0.135 0.1786 Co 0.2003 -0.1497 -0.2614 -0.2641 0.4911 -0.158 0.0852 0.4393 -0.0918 -0.314 Cr 0.1129 0.1303 0.0218 -0.1261 -0.1192 0.0545 -0.1869 -0.1713 -0.0356 0.3128 Cs 0.1399 0.1098 -0.3196 -0.2579 -0.1185 -0.0274 -0.1272 -0.0265 0.5901 -0.021 Eu 0.2973 -0.0961 0.2079 -0.1109 0.1407 0.1348 0.0301 -0.218 0.0849 -0.0395 Fe 0.1032 0.1392 -0.055 -0.1138 0.1411 -0.0414 -0.0376 0.1577 -0.1864 0.3795 Hf 0.0613 0.0106 0.1417 0.1995 -0.2666 -0.3205 0.0436 0.3781 0.0499 0.0656 Rb 0.2422 0.0359 -0.3662 0.0045 -0.2194 0.0185 0.0587 0.0585 0.2661 -0.2519 Sb 0.0724 0.3033 0.1065 -0.1749 -0.1559 -0.015 -0.2219 -0.0319 -0.3478 -0.2429 Sc 0.1176 0.0509 -0.1076 -0.2127 0.0108 0.0538 -0.0651 0.0408 -0.0712 0.2784 Ta 0.0574 0.0108 -0.0319 -0.1576 -0.0544 -0.1341 0.0021 0.1536 -0.0738 0.0016 Tb 0.2546 -0.1073 0.2152 -0.0322 0.1038 0.0241 0.1035 -0.1003 0.0607 -0.095 Th 0.1013 0.0418 0.0412 -0.0415 -0.0902 -0.1014 -0.0184 0.1659 -0.0232 0.2223 Zn 0.1826 0.0462 -0.173 -0.1633 -0.0315 -0.06 -0.212 -0.0198 -0.2153 -0.0206 Zr 0.0821 0.0385 0.1803 0.2103 -0.2968 -0.3443 -0.0174 0.3542 0.0253 0.0192 Al 0.0901 0.0335 -0.1185 -0.2203 -0.0015 0.0161 -0.0566 0.0304 -0.0944 0.2123 Ba 0.1604 -0.1078 -0.3051 0.0178 -0.1786 -0.0539 0.7116 -0.198 -0.3372 0.1916 Dy 0.2236 -0.0972 0.1963 -0.0594 0.0808 0.0054 0.0666 -0.0272 -0.021 -0.0778 K 0.3082 -0.0147 -0.241 0.2495 -0.3372 0.2136 0.0109 -0.0487 0.0398 -0.0527 Mn 0.2037 -0.1087 -0.2317 0.3557 0.2353 -0.6068 -0.334 -0.4301 -0.0852 0.0443 V 0.2819 -0.0493 -0.1275 0.5044 0.1718 0.5088 -0.2511 0.2931 -0.1645 0.0893 Ti 0.0642 0.0229 -0.0303 -0.1013 -0.0461 0.0329 0.0039 0.0514 -0.027 0.1226 Na 0.0837 0.1303 -0.0183 -0.1958 -0.0711 -0.0817 -0.137 0.1514 -0.1028 0.0579

PAGE 297

APPENDIX B PETROGR APHIC ANALYSIS 297

PAGE 298

Table B-1. Raw point c ount data (1). sample NJWID interval V o I d s Twit h V total M a t r I x a p l a s t i c s # s i l t vf Q fine Q M e d Q C r s Q vc Q poly xQ F e l s i c R K s p a r p l a g U I D f e l d h e a v y a m p h i b e p i d o t e m i c a f e r r i c F e s a n d G r o g C l a y l u m p C h a r c o a l B o n e S p c p h y t 2004-01 NJW-242 1x1x2 28 359 331 180 151 6 18 43 41 23 5 7 1 2 . 5. . 2004-03 NJW-029 1x1x2 12 323 311 173 138 8 59 47 1 2 22 1 1 22 52. 2 2 . 2004-18 NJW-038 1x1 21 267 246 143 103 2 12 16 27 34 5 6 1 . . . 2005-01 NJW-032 1x1x2 21 328 307 188 119 8 31 47 4 1 1 1 13. 28 1 2005-02 NJW-043 1x1x2 32 310 278 161 117 5 31 32 9 2 1 2 12. 2 30 2005-03 NJW-313 1x1x2 48 394 346 156 190 4 32 122 7 1 . . . 24 2005-08 NJW-072 1x1 40 357 317 178 139 1 12 77 16 4 5 5 1 1 7 10 2005-09 NJW-073 1x1 36 410 374 217 157 6 16 75 20 2 6 2 . 1 2 8 17 2 2005-23 NJW-098 1x1 26 293 267 150 117 14 6 15 3 . 2 1 1 75 2005-24 NJW-026 1x1x2 40 343 303 156 147 10 12 27 2 1 1 1 . 9 13 70 1 2005-25 NJW-025 1x1x2 31 356 325 237 88 13 18 31 4 . 1 . 1 2 18 2005-26 NJW-047 1x1x2 48 404 356 250 106 4 20 44 8 6 . . . 24 2008-01 NJW-001 1x.5 30 356 326 216 110 14 24 55 5 1 1 3 . 1 5 1 . 2008-02 NJW-007 1x.5 30 366 336 234 102 2 1 20 39 19 7 9 2 2 1 . . . 2008-03 NJW-008 1x.5 27 296 269 159 110 4 17 25 13 13 4(1g) 21 4 1 7 . 1 . 2008-04 NJW-010 1x.5 32 387 355 193 162 7 23 46 37 22 4 12 1 4 . 1 5 . 2008-05 NJW-017 1x.5 60 399 339 181 158 21 39 81 4 4 4 1 1 1 1 1 . 2008-06 NJW-021 1x.5 25 369 344 224 120 10 25 74 4 1 1 3 . 2 . 2008-07 NJW-102 1x.5 18 232 214 116 98 15 38 26 3 1 4 3 1 2 1 4 . 2008-08 NJW-111 1x.5 49 468 419 257 162 14 15 92 13 3 1 6 1 8 2 2 4 1 . 2008-09 NJW-112 1x.5 22 385 363 253 110 15 24 33 9 1 7 1 1 19 2008-10 NJW-119 1x.5 42 531 489 332 157 15 16 96 11 5 1 1 2 1 1 5 3 . 2008-11 NJW-122 1x.5 10 286 276 185 91 21 8 22 3 1 . . 1 2 33 2008-12 NJW-136 1x.5 34 357 323 159 164 21 22 86 12 1 9 1 8 1 1 2 . 2008-13 NJW-138 1x.5 19 244 225 129 96 7 14 53 9 2 3 4 . 2 2 2008-14 NJW-139 1x.5 65 405 340 179 161 8 24 96 10 12 1 3 . 3 1 3 . 2008-15 NJW-145 1x. 5 19 343 324 212 112 15 37 41 2 4 2 1 7 1 2 . 2008-16 NJW-079 1x.5 24 365 341 216 125 8 4 24 30 17 4 24 2 1 5 . 1 4 1 2008-17 NJW-081 1x.5 17 318 301 190 111 14 10 55 5 6 5 1 1 12 298

PAGE 299

Table B-1. Continued sample NJWID interval V o I d s Twit h V total M a t r I x a p l a s t i c s # s i l t vf Q fine Q M e d Q C r s Q vc Q poly xQ F e l s i c R K s p a r p l a g U I D f e l d h e a v y a m p h i b e p i d o t e m i c a f e r r i c F e s a n d G r o g C l a y l u m p C h a r c o a l B o n e S p c p h y t 2008-18 NJW-085 1x.5 67 521 454 260 194 10 41 101 15 1 4 11 5 1 2. 12 . 2008-19 NJW-093 1x.5 37 270 233 131 102 4 14 45 1 5 2. 1 . 12 1 26 2008-20 NJW-091 1x.5 35 392 357 224 133 9 19 85 13 1 1 4 . 1 . 2008-21 NJW-152 1x.5 28 242 214 121 93 18 32 27 4 5 221 1 1 2008-22 NJW-151 1x.5 48 400 352 197 155 29 47 45 4 5 13 12 1. 7 1 . 2008-23 NJW-158 1x.5 47 373 326 188 138 6 6 33 44 27 6 9 1 1 1 . 4 2008-24 NJW-164 1x.5 46 441 395 222 173 8 40 97 8 7 1 5 1 1 5 2008-25 NJW-169 1x.5 53 516 463 279 184 18 30 114 8 1 5 1 2 1 1 2 1 2008-26 NJW-174 1x.5 56 545 489 279 210 10 18 53 48 36 5 30 2 4 1 2 1 . 2008-27 NJW-177 1x.5 66 358 292 144 148 10 19 61 31 9 6 1 1 1 3 3 3 . 2008-28 NJW-190 1x.5 83 581 498 263 235 10 15 83 55 30 6 26 2 1 3 1 1 2 . 2008-29 NJW-191 1x.5 38 395 357 179 178 18 6 45 50 29 7 16 1 1 1 1 2 . 1 2008-30 NJW-192 1x.5 71 501 430 194 236 23 71 107 3 10 3 7 7 3 1 1 2008-31 NJW-193 1x.5 57 501 444 240 204 13 21 54 49 37 6 17 1 2 1 . 1 2 2008-32 NJW-027 1x.5 70 605 535 292 243 9 15 61 74 46 12 21 1 2 . 1 1 2008-33 NJW-050 1x.5 31 419 388 228 160 6 12 30 52 39 7 10 2 . . 2 . 2008-34 NJW-053 1x.5 47 447 400 218 182 13 25 76 37 11 5 4 . 6 5 . 2008-35 NJW-055 1x.5 31 256 225 124 101 11 33 32 13 6 3 1 1 . . 1 2008-36 NJW-059 1x.5 50 485 435 273 162 5 11 44 40 25 7 23 3 1 . 1 2 . 2008-37 NJW-063 1x.5 50 382 332 166 166 7 14 37 38 42 7 15 4 1 . . 1 2008-38 NJW-064 1x.5 53 589 536 315 221 10 24 77 58 20 1 14 1 1 8 . 2 1 2 2 2008-39 NJW-195 1x.5 69 435 366 188 178 6 28 55 46 25 1 8 1 4 1 2 1 . 2008-40 NJW-196 1x.5 47 435 388 220 168 8 9 33 52 38 7 13 1 5 . 1 1 . 2008-41 NJW-197 1x.5 44 369 325 194 131 9 7 29 38 25 2 10 1 1 2 2 . 5 . 2008-42 NJW-198 1x.5 48 425 377 211 166 10 12 22 34 46 10 25 2 1 1 3 . . . 2008-43 NJW-199 1x.5 31 383 352 220 132 13 47 38 2 8 3 7 1 1 1 6 1 1 3 . 2008-44 NJW-218 1x. 5 49 477 428 267 161 17 31 87 7 7 2 2 2 1 2 3 . 299

PAGE 300

Table B-1. Continued sample NJWID interval V o I d s Twit h V total M a t r I x a p l a s t i c s # s i l t vf Q fine Q M e d Q C r s Q vc Q poly xQ F e l s i c R K s p a r p l a g U I D f e l d h e a v y a m p h i b e p i d o t e m i c a f e r r i c F e s a n d G r o g C l a y l u m p C h a r c o a l B o n e S p c p h y t 2008-45 NJW-219 1x.5 82 532 450 236 214 15 18 40 52 45 15 19 1 21 1 . 14 . 2008-46 NJW-220 1x.5 40 412 372 210 162 13 15 46 35 3 9 3 5 1 2 11 15 8 4 1 2008-47 NJW-221 1x.5 59 475 416 243 173 7 24 34 46 42 4 10 3 . 21 . 2008-48 NJW-222 1x.5 59 469 410 204 206 9 36 62 41 6 1 17 13 4 1 12. 14 5 2 1 2008-49 NJW-241 1x.5 41 375 334 199 135 7 25 42 39 7 2 7 2 1 3. . 2008-50 NJW-257 1x.5 77 547 470 238 232 14 26 79 52 31 1 15 2 2 2 4 3 1 . 2008-51 NJW-260 1x.5 18 353 335 252 83 23 16 8 3 9 5 3 1 1 1 2 2 1 . 8 2008-52 NJW-277 1x.5 23 262 239 173 66 5 17 7 15 11 3 6 1 1 . . 2008-53 NJW-279 1x.5 15 217 202 120 82 5 18 20 17 6 3 5 1 6 . 1 . 2008-54 NJW-281 1x.5 33 295 262 176 86 5 5 17 31 15 1 8 1 1 1 . . 1 2008-55 NJW-293 1x.5 28 396 368 273 95 1 9 12 25 25 5 11 1 2 1 2 . . 1 2008-56 NJW-304 1x.5 25 225 200 114 86 10 12 40 2 2 1 2 1 1 1 1 9 4 . 2008-57 NJW-307 1x.5 27 360 333 216 117 10 6 20 36 18 7 10 3 1 . 3 2 . 1 C03-58c NJW-317 1x.5 59 430 371 193 178 11 32 52 5 2 2 4 1 2 1 . 66 C04-59c NJW-318 1x.5 87 428 341 278 63 2 16 23 8 1 5 1 2 1 1 . 1 2 C05-60c NJW-319 1x.5 106 466 360 179 181 16 36 85 33 2 4 1 2 . 2 . C06-61c NJW-320 1x.5 58 341 283 197 86 12 23 37 4 1 1 1 2 1 1 1 2 . C07-62c NJW-321 1x.5 35 408 373 297 76 3 14 46 2 2 2 1 5 1 . . C10-63c NJW-324 1x1 41 262 221 84 137 6 19 78 23 5 1 3 . 1 1 . C13-64c NJW-327 1x1 6 296 290 191 99 38 25 9 2 3 1 4 1 2 3 4 2 1 1 3 C15-65c NJW-329 1x.5 52 395 343 309 34 4 5 16 7 1 . 1 . . C17-66c NJW-331 1x1 31 318 287 201 86 10 25 25 10 4 5 1 2 1 1 2 . C18-67c NJW-332 1x1 60 382 322 174 148 11 54 51 6 3 7 1 1 3 6 . 4 1 . 300 1x.5=counting interval 1mm by .5mm m=medium Q=quartz plag=plagioclase 1x1=counting interval 1mm by 1mm crs=coarse polyxQ=polycrystalline quartz or quartzite UIDfeld=UID felspars 1x1x2=counting interval 1mm by 1mm, counted twice vc=very coarse felsicR=felsic or grantitic rock fragm ent heavy=UID minerals vf=very fine phyto=phytoliths g=granule Kspar=microclin e or potassium feldspar amphib=amphibole ferric=ferric concretions or nodules charred=charcoal temper FE sand = ferric with imbe dded quartz grains spc=sponge spicules

PAGE 301

Table B-2. Raw point count data (2). sample NJWID polyX Q felsicR Kspar plag UID feld heavy amphib epidote mica ferric 2004-01 NJW-242 7(2vf2f3c) 1vf 2(vff) 2004-03 NJW-029 2f 2(vff) 2(vff) 1vf 1vf 2vf 2(vff) 2004-18 NJW-038 6(1f3m1c1vc) 1vf 2005-01 NJW-032 1vf 1vf 1vf 3(2m1vf) 2005-02 NJW-043 2f 1vf 2(vff) 1vf 2vf 2005-03 NJW-313 1f . 2005-08 NJW-072 5f 5(4vf1f) 1f 1vf 2005-09 NJW-073 6f 2vf 1vf 2005-23 NJW-098 2vf 1vf 1vc 2005-24 NJW-026 1vf 1vf 9(2c4m2f1vf) 2005-25 NJW-025 1vf 1m 2005-26 NJW-047 6(2vf3f1m) . 2008-01 NJW-001 1(f) 1(vf) 3(vf) . 1(vf) 2008-02 NJW-007 9(1f1m6c1g) 2(m) 2(1m1c) 1(m) . 2008-03 NJW-008 21(3f9m6c2vc1g) 4(2c2vc) 1(f) 7(3vf2f2m) . 2008-04 NJW-010 12(5f3m3c1vc) 1(f) 4(vf) . 1(f) 2008-05 NJW-017 4(1vf3f) 4(2vf2f) 1(f) 1(vf) 1(m) 2008-06 NJW-021 1(f) 1(f) 3(1vf1f1m) . 2008-07 NJW-102 4(1f3m) 3(2vf1f) 1(vf) 2(vf) 1(vf) 4(2vf2f) 2008-08 NJW-111 6(1vf3f1m1vc) 1(f) 8(4vf3f1m) 2(1vf1f) 2(vf) 2008-09 NJW-112 1(f) 7(6vf1f) 1(vf) 1(m) 2008-10 NJW-119 5(3f2m) 1(vf) 1(f) 2(1vf1f) 1(f) 1(m) 5(1vf3f1m) 2008-11 NJW-122 1(f) . . 1(vf) 2008-12 NJW-136 9(2vf5f1m1c) 1(vf) 8(1vf7f) 1(f) brn ple 1(f or opq) 2008-13 NJW-138 3(f) 4(2vf2f) . 2(1vf1f) 2008-14 NJW-139 12(3vf8f1m) 1(f) 3(2vf1f) . 3(2vf1f) 2008-15 NJW-145 4(f) 2(1vf1f) 1(f) 7(4vf3f) 1(vf) 2(vf) 2008-16 NJW-079 24(9m10c4vc1g) 2(f) 1(vf) 5(2vf2f1m) . 2008-17 NJW-081 8(1vf7f) 5(1vf3f1m) 1(vf) 2008-18 NJW-085 4(1vf1f2m) 1(vf) 1(f) 5(vf) 1(vf) 2(1vf1f) 2008-19 NJW-093 5(f) 2(1vf1f) 1(f) . 1(f) 2008-20 NJW-091 1(f) 1(f) 4(2vf2f) . 2008-21 NJW-152 4(1vf3f) 5(vf) 2(1vf1f) 2(1vf1f) 2008-22 NJW-151 5(2vf3f) 1(vf) 3(vf) 12(9vf3f) 1(vf) 301

PAGE 302

Table B-2. Continued sample NJWID polyX Q felsicR Kspar plag UID feld heavy amphib epidote mica ferric 2008-23 NJW-158 9(5f4c) 1(f) 1(vf) 1(f) uid . 2008-24 NJW-164 7(1vf5f1m) 1(vf) 5(2vf3f) 1(vf) 2008-25 NJW-169 5(1vf4f) 1(vf) 2(1vf1f) 1(vf) opq 1(vf) 2008-26 NJW-174 30(1vf9f10m9c1vc) 2(f) 4(1vf2f1c) 1(vf) 3o 2(1vf1f) 2008-27 NJW-177 6(3f2m1c) 1(f) 1(f) 1(vf) Trm 3(2vf1f) 2008-28 NJW-190 26(1vf11f8m5c1vc) 2(1f1c) 1(f) 3(vf) 1(vf) 2o 1(f) 2(vf) 2008-29 NJW-191 16(1vf7f4m4c) 1(f) 1(f) 1(vf) 1(m) Z 2(1vf1f) f opq? 2008-30 NJW-192 10(1vf8f1m) 3(1vf2f) 7(3vf4f) 7(5vf2f) 3(1vf1f1m)1vf B 1(vf) 2008-31 NJW-193 17(3f7m6c1vc) 1(f) 2(f) 1(vf) lowo . 2008-32 NJW-027 21(1vf2f8m6c2vc2g) 1(vf) 2(vf) . 2008-33 NJW-050 10(2f3m5c) 2(f) . 2008-34 NJW-053 5(2f3m) 4(2vf2f) . 6(1vf2f3m) 2008-35 NJW-055 3(f) 1(vf) 1(vf) . 2008-36 NJW-059 23(3f7m11c2vc) 3(1vf1f1m) 1(vf) . 1(vf) 2008-37 NJW-063 15(7m7c1vc) 4(f) 1(m) . 2008-38 NJW-064 14(3vf3f4m2c2vc) 1(m) 1(f) 8(4vf3f1m) . 2(vf) 2008-39 NJW-195 8(1vf5m1c1vc) 1(vf) 4(3vf1f) 1(vf) lowo ple 2(1vf1m) 2008-40 NJW-196 13(3f2m6c1vc1p) 1(f) 5(4vf1f) . 1(vf) 2008-41 NJW-197 10(2f3m2c3vc) 1(vc) 1(vf) 2(1f1m) 2(1vf1f) . 5(3vf1f1m) 2008-42 NJW-198 25(1vf2f4m14c4vc)1c chert 2(1f1c) 1(c) 1(f) 3(2vf1f) . 2008-43 NJW-199 7(2vf4f1vc) 1(vc) 1(f) 1(vf) 6(3vf1m1c1vc) 1(vf) Ky or Px 1(vf) 3(1vf2f) 2008-44 NJW-218 7(3vf4f) 2(1vf1f) 2(1vf1f) 2(f) 1(f) brn ple 2(m) 3(1vf2f) 2008-45 NJW-219 19(2f5m7c5vc) 1(vc) 2(1f1m) 1(f) 1(f) . 1(vf) 2008-46 NJW-220 9(1vf4f4m) 3(1vf1f1m) 5(4vf1f) 1(vf) opq or Fe 2(1vf1f) 1(vf) 2008-47 NJW-221 10(1f6m3c) 3(vf) . 2(1vf1f) 2008-48 NJW-222 17(6f8m3c) 1(f) 3(1vf2f) 4(3vf1f) 1(f) 1(f) 2(1vf1m) vf opq? 2008-49 NJW-241 7(1vf1f3m2c) 2(vf) 1(vf) 3(2vf1f) NJW-257 15(3f7m4c1vc) 2(1vf1m) 2(1vf1m) 2(m T vf opq) 4(2vf2f) 2008-50 2008-51 NJW-260 3(vc) 1(vf) 1(vf) 1(vf) 2(vf) 2(1vf1f) 302

PAGE 303

Table B-2. Continued sample NJWID polyX Q felsicR Kspar plag UID feld heavy amphib epidote mica ferric 2008-52 NJW-277 6(2f1m3c) 1(vf) 1(vf) 2008-53 NJW-279 5(1vf1f2m1c) 1(m) 6(3vf1f1m1c) . 1(vf) 2008-54 NJW-281 8(1f3m2c2vc) 1(vf) 1(vf) 1(f) . 2008-55 NJW-293 11(1f2m5c2vc1g) 1(m) 2(c) 1(c) 2(m) . 2008-56 NJW-304 1(vf) 2(1vf1f) 1(vf) opq 1(vf) 1(f) 2008-57 NJW-307 10(2f5m3c) 3(1vf1f1m) 1(vf) . 3(2f1m) C03-58c NJW-317 2(f) 2(f) 4(vf) 1(vf) or T 2(1vf1f) C04-59c NJW-318 5(2f2m1c) 1(f) 2(vf) 1(f) opq 1(vf) or T C05-60c NJW-319 4(3f1m) 1(vf) 2(vf) . 2(vf) C06-61c NJW-320 1(vf) 1(f) 1(f) 2(vf) 1(vf) 2o blue 1(f) 1(vf) 2(1vf1f) C07-62c NJW-321 2(f) 1(f) 5(2vf3f) 1(vf) C10-63c NJW-324 1(f) 3(1vf2f) . 1(vf) C13-64c NJW-327 4(1vf2m1vc) 1(f) 2(1vf1f) 3(2vf1m) 4(vf) 1 B 2(vf) C15-65c NJW-329 1(f) . 1(m) C17-66c NJW-331 5(2f2m1c) 1(m) 2(vf) 1(f) 1(f) 2(1vf1f) C18-67c NJW-332 7(1vf2f2m2c) 1(vc) 1(f) 3(1vf2f) 6(5vf1f) . 4(2vf1f1m) 303 Table B-3. Raw point count data (3). sample NJWID Fe sand grog temper clay lumps charcoal temper bone temper 2004-01 NJW-242 . 2004-03 NJW-029 2(f,m) 2(vc) 2004-18 NJW-038 . 2005-01 NJW-032 28(5vf7f6m8c2vc) 2005-02 NJW-043 2(vc,gr) 30(2vf10f10m8c) 2005-03 NJW-313 24(6c18vc) 2005-08 NJW-072 7(4c3m) 10(1vf1f2m5c1vc) 2005-09 NJW-073 2(m,c) 8(1vf3f2m3c) 17(4vf5f3m5c) 2005-23 NJW-098 75(18vf6f22m27c2vc) 2005-24 NJW-026 13(1vc5c3m3f1vf) 70(14vf12f20m20c4vc) 1c 2005-25 NJW-025 2(gr,f) 18(5vf3f3m7c)

PAGE 304

Table B-3. Continued sample NJWID Fe sand grog temper clay lumps charcoal temper bone temper 2005-26 NJW-047 24(1m6c17vc) 2008-01 NJW-001 5(3m1c1vc) 1(vc) 2008-02 NJW-007 . 2008-03 NJW-008 1(c) 2008-04 NJW-010 5(1m2c2vc) . 2008-05 NJW-017 1(m) 1(vc) 2008-06 NJW-021 2(1f1c) . 2008-07 NJW-102 . 2008-08 NJW-111 4(vc) or cl 1(m) 2008-09 NJW-112 19(4f8m7c) 2008-10 NJW-119 3(2c1vc) . 2008-11 NJW-122 2(m) 33(7vf6f7m10c3vc) 2008-12 NJW-136 2(c) or org . 2008-13 NJW-138 2(1f1m) 2008-14 NJW-139 1(f) 3(2m1c) or Fe sand 2008-15 NJW-145 . 2008-16 NJW-079 1(m) or cl 4(3vf1f) or Fe 2008-17 NJW-081 1(vc) or Fe sand 12(2vf5f5m) 2008-18 NJW-085 12(1f4m2c5vc) 2008-19 NJW-093 2(m) 1(vc) or Fe sand 26(6vf13f5m2c) 2008-20 NJW -091 1(c) 2008-21 NJW-152 1(m) 1(vc) 1(m) 2008-22 NJW-151 7(1m3c3vc) 1(c) 2008-23 NJW-158 4(1f2c1vc) 1c org? 2008-24 NJW-164 1(c) 5(1vf1m1c1vc1g) 2008-25 NJW-169 2(1m1c) 1(m) 2008-26 NJW-174 1(m) . 2008-27 NJW-177 3(2m1c) 3(2c1vc) or Fe sand 2008-28 NJW-190 . 2008-29 NJW-191 . 2008-30 NJW-192 1(f) org 2008-31 NJW-193 1(f) 2(1vf1f) org 2008-32 NJW-027 1(m) 1(m) org? 2008-33 NJW-050 2(c) 304

PAGE 305

Table B-3. Continued sample NJWID Fe sand grog temper clay lumps charcoal temper bone temper 2008-34 NJW-053 5(1m2c2vc) . 2008-35 NJW-055 . 2008-36 NJW-059 2(vc) 2008-37 NJW-063 1(vf) org sand 2008-38 NJW-064 1(c) 2(1c1vc) or Fe sand 2(1vc1g) 2008-39 NJW-195 1(c) . 2008-40 NJW-196 1(m) . 2008-41 NJW-197 . 2008-42 NJW-198 . 2008-43 NJW-199 . 2008-44 NJW-218 . 2008-45 NJW-219 4(1f3c) . 2008-46 NJW-220 1(f) 15(1vf2f3m5c3vc1g) 8(1vf2f1m3c1vc) 4(1f2m1vc) 2008-47 NJW-221 1(m) . 2008-48 NJW-222 14(2f4m5c2vc1p) 5(2f3m) 2(1f1m) 1m org 2008-49 NJW-241 . 2008-50 NJW-257 3(1m1c1vc) 1(vc) 2008-51 NJW-260 1(c) . 2008-52 NJW-277 . 2008-53 NJW-279 . 2008-54 NJW -281 . 2008-55 NJW-293 . 2008-56 NJW-304 1(m) 9(1m3c3vc2g) 4(2f2m) or Fe 2008-57 NJW-307 2(1c1vc) . C03-58c NJW-317 1(c) . C04-59c NJW-318 1(m) 2(1f1m) org C05-60c NJW-319 . C06-61c NJW-320 . C07-62c NJW-321 . C10-63c NJW-324 1(gr-pb) C13-64c NJW-327 1(c) or Fe C15-65c NJW-329 . C17-66c NJW-331 . C18-67c NJW-332 1(vc) 305

PAGE 306

Table B-4. Percentage data (1). sample NJWID temper total %voids %matrix %aplast %sand %nonsand %qtz %Tqpq %feld %heavy %silt %vfq %fineQ %medQ %crsQ 2004-01 NJW-242 GROG 331 8% 54% 46% 42% 3% 39% 41% 1% 2% 5% 13% 12% 7% 2004-03 NJW-029 GRIT 311 4% 56% 44% 38% 6% 34% 35% 2% 2% 3% 19% 15% <1% 2004-18 NJW-038 GRIT 246 8% 58% 42% 41% 1% 38% 41% <1% 1% 5% 6% 11% 14% 2005-01 NJW-032 CHAR 307 6% 60% 40% 28% 11% 27% 27% <1% 3% 10% 15% 1% <1% 2005-02 NJW-043 CHAR 278 10% 58% 42% 27% 15% 26% 27% <1% 1% 2% 11% 12% 3% 2005-03 NJW-313 CHAR 346 12% 45% 55% 47% 8% 46% 47% 1% 9% 35% 2% 2005-08 NJW-072 CHAR 317 11% 56% 44% 38% 6% 34% 36% 2% <1% <1% 4% 24% 5% 1% 2005-09 NJW-073 CHAR 374 9% 58% 42% 32% 10% 30% 32% <1% 2% 4% 20% 5% <1% 2005-23 NJW -098 CHAR 267 9% 56% 44% 10% 34% 9% 9% 1% 5% 2% 6% 1% 2005-24 NJW-026 CHAR 303 12% 52% 48% 14% 34% 14% 14% <1% 3% 4% 9% 1% <1% 2005-25 NJW-025 CHAR 325 9% 73% 27% 17% 10% 16% 16% <1% 4% 6% 10% 1% 2005-26 NJW-047 CHAR 356 12% 70% 30% 22% 8% 20% 22% 1% 6% 12% 2% 2008-01 NJW-001 CHAR 326 8% 66% 34% 27% 2% 26% 26% 1% 4% 7% 17% 2% 2008-02 NJW-007 GRIT 336 8% 70% 30% 30% 26% 28% 1% 1% <1% 6% 12% 6% 2008-03 NJW-008 GRIT 269 9% 59% 41% 39% <1% 27% 35% 3% 2% 6% 9% 5% 5% 2008-04 NJW-010 GSAND 355 8% 54% 46% 42% 2% 37% 41% 1% 2% 6% 13% 10% 6% 2008-05 NJW-017 SAND 339 15% 53% 47% 40% 1% 37% 38% 1% 1% 6% 12% 24% 1% 2008-06 NJW-021 SAND 344 7% 65% 35% 31% 1% 30% 30% 1% 3% 7% 22% 1% 2008-07 NJW-102 SAND 214 8% 54% 46% 36% 2% 32% 34% 3% <1% 7% 18% 12% 1% <1% 2008-08 NJW-111 SAND 419 10% 61% 39% 34% 2% 30% 31% 2% <1% 3% 4% 22% 3% 1% 2008-09 NJW-112 CHAR 363 6% 70% 30% 21% 6% 18% 18% 2% <1% 4% 7% 9% 2% 2008-10 NJW-119 SAND 489 8% 68% 32% 27% 2% 25% 26% 1% <1% 3% 3% 20% 2% 2008-11 NJW-122 CHAR 276 3% 67% 33% 12% 13% 12% 12% 8% 3% 8% 1% 2008-12 NJW-136 CHAR 323 10% 49% 51% 43% 1% 38% 40% 3% <1% 6% 7% 27% 4% <1% 2008-13 NJW-138 SAND 225 8% 57% 43% 38% 2% 35% 36% 2% 3% 6% 24% 4% 1% 2008-14 NJW-139 SAND 340 16% 53% 47% 43% 2% 38% 42% 1% 2% 7% 28% 3% 2008-15 NJW-145 SAND 324 6% 65% 35% 29% 1% 25% 26% 3% <1% 5% 11% 13% 1% 2008-16 NJW-079 GRIT 341 7% 63% 37% 33% 2% 23% 30% 2% 2% 1% 7% 9% 5% 2008-17 NJW-081 CHAR 301 5% 63% 37% 28% 5% 23% 26% 2% <1% 5% 3% 18% 2% 2008-18 NJW-085 GROG 454 13% 57% 43% 37% 3% 35% 36% 2% <1% 2% 9% 22% 3% 2008-19 NJW-093 CHAR 233 14% 56% 44% 29% 13% 26% 28% 1% 2% 6% 19% <1% 2008-20 NJW-091 SAND 357 9% 63% 37% 34% <1% 33% 33% 1% 2% 5% 24% 4% 2008-21 NJW-152 SAND 214 12% 56% 44% 32% 3% 28% 29% 2% 1% 8% 15% 13% 2008-22 NJW-151 SAND 352 12% 56% 44% 33% 3% 27% 29% 4% <1% 8% 13% 13% 1% 306

PAGE 307

Table B-4. Continued sample NJWID temper total %voids %matrix %aplast %sand %nonsand %qtz %Tqpq %feld %heavy %silt %vfq %fineQ %medQ %crsQ 2008-23 NJW-158 GRIT 326 13% 58% 42% 39% 1% 36% 38% 1% <1% 2% 2% 10% 14% 8% 2008-24 NJW-164 SAND 395 10% 56% 44% 40% 2% 37% 38% 2% <1% 2% 10% 25% 2% 2008-25 NJW-169 SAND 463 10% 60% 40% 35% 1% 33% 34% 1% <1% 4% 6% 25% 2% <1% 2008-26 NJW-174 GRIT 489 10% 57% 43% 40% 1% 33% 39% 1% <1% 2% 4% 11% 10% 7% 2008-27 NJW-177 GSAND 292 18% 49% 51% 44% 3% 41% 43% 1% <1% 3% 6% 21% 11% 3% 2008-28 NJW-190 GSAND 498 14% 53% 47% 45% <1% 38% 43% 1% <1% 2% 3% 17% 11% 6% 2008-29 NJW-191 GRIT 357 10% 50% 50% 44% 1% 38% 43% 1% <1% 5% 2% 13% 14% 8% 2008-30 NJW-192 SAND 430 14% 45% 55% 48% 1% 42% 44% 4% 1% 5% 16% 25% 1% 2008-31 NJW-193 GRIT 444 11% 54% 46% 42% 1% 38% 41% 1% <1% 3% 5% 12% 11% 8% 2008-32 NJW-027 GRIT 535 12% 55% 45% 43% <1% 39% 43% 1% 2% 3% 11% 14% 9% 2008-33 NJW-050 GRIT 388 7% 59% 41% 39% <1% 36% 39% <1% 2% 3% 8% 13% 10% 2008-34 NJW-053 GSAND 400 10% 54% 46% 40% 3% 37% 38% 1% 3% 6% 19% 9% 3% 2008-35 NJW-055 SAND 225 12% 55% 45% 40% <1% 37% 39% 1% 5% 15% 14% 6% 3% 2008-36 NJW-059 GRIT 435 10% 63% 37% 35% 1% 29% 34% 1% 1% 2% 10% 9% 6% 2008-37 NJW-063 GRIT 332 13% 50% 50% 48% <1% 42% 46% 2% 2% 4% 11% 11% 13% 2008-38 NJW-064 GSAND 536 9% 59% 41% 38% 1% 34% 36% 2% 2% 4% 14% 11% 4% 2008-39 NJW-195 GSAND 366 16% 51% 49% 46% 1% 42% 44% 1% <1% 2% 8% 15% 13% 7% 2008-40 NJW-196 GRIT 388 11% 57% 43% 41% <1% 36% 39% 2% 2% 2% 8% 13% 10% 2008-41 NJW-197 GRIT 325 12% 60% 40% 36% 2% 31% 34% 2% 3% 2% 9% 12% 8% 2008-42 NJW-198 GRIT 377 11% 56% 44% 41% 33% 40% 1% 3% 3% 6% 9% 12% 2008-43 NJW-199 SAND 352 8% 62% 38% 33% 1% 28% 30% 2% 1% 4% 13% 11% 1% 2% 2008-44 NJW-218 SAND 428 10% 62% 38% 32% 1% 29% 31% 1% 1% 4% 7% 20% 2% 2008-45 NJW-219 GRIT 450 15% 52% 48% 43% 1% 38% 42% 1% 3% 4% 9% 12% 10% 2008-46 NJW-220 GROG 372 10% 56% 44% 32% 8% 27% 29% 2% 1% 4% 4% 12% 9% 1% 2008-47 NJW-221 GRIT 416 12% 58% 42% 39% 1% 36% 38% 1% 2% 6% 8% 11% 10% 2008-48 NJW-222 GROG 410 13% 50% 50% 42% 6% 36% 40% 2% <1% 2% 9% 15% 10% 2% 2008-49 NJW-241 GSAND 334 11% 60% 40% 37% 1% 34% 36% 1% <1% 2% 8% 13% 12% 2% 2008-50 NJW-257 GSAND 470 14% 51% 49% 45% 2% 40% 43% 1% <1% 3% 6% 17% 11% 7% 2008-51 NJW-260 GSAND 335 5% 75% 25% 14% 4% 12% 13% 1% 1% 7% 5% 2% 1% 3% 2008-52 NJW-277 GRIT 239 9% 72% 28% 25% <1% 22% 25% <1% <1% 2% 7% 3% 6% 5% 2008-53 NJW-279 GSAND 202 7% 59% 41% 38% <1% 32% 34% 4% 2% 9% 10% 8% 3% 2008-54 NJW-281 GRIT 262 11% 67% 33% 30% <1% 26% 29% 1% 2% 2% 6% 12% 6% 2008-55 NJW-293 GRIT 368 7% 74% 26% 25% <1% 21% 24% 1% <1% 2% 3% 7% 7% 307

PAGE 308

Table B-4. Continued sample NJWID temper total %voids %matrix %aplast %s and %nonsand %qtz %Tqpq %feld %heavy %s ilt %vfq %fineQ %medQ %crsQ 2008-56 NJW-304 GROG 200 11% 57% 43% 30% 8% 28% 28% 1% 1% 5% 6% 20% 1% 1% 2008-57 NJW-307 GRIT 333 8% 65% 35% 30% 2% 26% 29% 1% 3% 2% 6% 11% 5% C03-58c NJW-317 CLAY 371 14% 52% 48% 26% 19% 24% 24% 2% <1% 3% 9% 14% 1% C04-59c NJW-318 CLAY 341 20% 82% 18% 17% 1% 14% 16% 1% 1% 1% 5% 7% 2% <1% C05-60c NJW-319 CLAY 360 23% 50% 50% 45% 1% 43% 44% 1% 4% 10% 24% 9% 1% C06-61c NJW-320 CLAY 283 17% 70% 30% 25% 1% 23% 23% 1% 1% 4% 8% 13% 1% C07-62c NJW-321 CLAY 373 9% 80% 20% 20% 17% 18% 2% <1% 1% 4% 12% <1% <1% C10-63c NJW-324 CLAY 221 16% 38% 62% 58% 1% 57% 57% 1% 3% 9% 35% 10% 2% C13-64c NJW-327 CLAY 290 2% 66% 34% 17% 4% 14% 15% 2% 1% 13% 9% 3% 1% 1% C15-65c NJW-329 CLAY 343 13% 90% 10% 8% <1% 8% 8% <1% 1% 2% 5% 2% C17-66c NJW-331 CLAY 287 10% 70% 30% 25% 1% 22% 24% 1% 1% 4% 9% 9% 4% 1% C18-67c NJW-332 CLAY 322 16% 54% 46% 41% 2% 35% 38% 3% 3% 17% 16% 2% 1% Table B-5. Percentage data (2). sample NJWID %vcQ %polyxQ %felsicR %ferric %FeSand %grog %clylmp %char %bone %spc %phyt 2004-01 NJW-242 2% 2% <1% 2% . 2004-03 NJW-029 1% 1% 1% 1% . 2004-18 NJW-038 2% 2% . . 2005-01 NJW-032 <1% 1% 9% <1% 2005-02 NJW-043 1% 1% 1% 11% 2005-03 NJW-313 <1% . 7% 2005-08 NJW-072 2% 2% 3% 2005-09 NJW-073 2% <1% <1% 2% 4% <1% 2005-23 NJW-098 <1% 28% 2005-24 NJW-026 <1% 3% 4% 23% 1% 2005-25 NJW-025 <1% 1% 6% 2005-26 NJW-047 2% . 7% 2008-01 NJW-001 <1% <1% 2% <1% . 2008-02 NJW-007 2% 3% 1% . . 2008-03 NJW-008 2% 8% 2% <1% . 2008-04 NJW-010 1% 3% <1% 1% . 2008-05 NJW-017 1% <1% <1% . 308

PAGE 309

Table B-5. Continued sample NJWID %vcQ %polyxQ %felsicR %ferric %FeSand %grog %clylmp %char %bone %spc %phyt 2008-06 NJW-021 <1% 1% . 2008-07 NJW-102 2% 2% . 2008-08 NJW-111 <1% 1% <1% 1% <1% . 2008-09 NJW-112 <1% <1% 5% 2008-10 NJW-119 1% 1% 1% . 2008-11 NJW-122 <1% <1% 1% 12% 2008-12 NJW-136 3% <1% 1% . 2008-13 NJW-138 1% 1% 1% 2008-14 NJW-139 4% <1% 1% <1% 1% . 2008-15 NJW-145 1% 1% . 2008-16 NJW-079 1% 7% <1% 1% <1% 2008-17 NJW-081 2% . <1% 4% 2008-18 NJW-085 <1% 1% <1% 3% . 2008-19 NJW-093 2% <1% 1% <1% 11% 2008-20 NJW-091 <1% . <1% . 2008-21 NJW-152 2% 1% <1% <1% <1% 2008-22 NJW-151 1% 2% <1% . 2008-23 NJW-158 2% 3% . 1% 2008-24 NJW-164 2% <1% 1% 2008-25 NJW-169 1% . <1% <1% 2008-26 NJW-174 1% 6% <1% <1% . 2008-27 NJW-177 2% 1% 1% 1% . 2008-28 NJW-190 1% 5% <1% . 2008-29 NJW-191 2% 4% 1% . <1% 2008-30 NJW-192 2% <1% <1% 2008-31 NJW-193 1% 4% <1% <1% 2008-32 NJW-027 2% 4% <1% <1%o 2008-33 NJW-050 2% 3% . <1% . 2008-34 NJW-053 1% 2% 1% . 2008-35 NJW-055 1% . . <1% 2008-36 NJW-059 2% 5% <1% <1% . 2008-37 NJW-063 2% 4% . <1%o 2008-38 NJW-064 <1% 3% <1% <1% <1% <1% 2008-39 NJW-195 <1% 2% 1% <1% . 309

PAGE 310

Table B-5. Continued sample NJWID %vcQ %polyxQ %felsicR %ferric %FeSand %grog %clylmp %char %bone %spc %phyt 2008-40 NJW-196 2% 3% <1% <1% . 2008-41 NJW-197 1% 3% <1% 2% . 2008-42 NJW-198 3% 7% <1% . . 2008-43 NJW-199 1% 2% <1% 1% . 2008-44 NJW-218 2% 1% . 2008-45 NJW-219 3% 4% <1% <1% 1% . 2008-46 NJW-220 2% <1% <1% 4% 2% 1% <1% 2008-47 NJW-221 1% 2% <1% <1% . 2008-48 NJW-222 <1% 4% <1% 3% 1% <1% <1% 2008-49 NJW-241 1% 2% 1% . 2008-50 NJW-257 <1% 3% 1% 1% <1% . 2008-51 NJW-260 2% 1% 1% <1% . 2% 2008-52 NJW-277 1% 2% . . 2008-53 NJW-279 2% 2% <1% . 2008-54 NJW-281 <1% 3% . . <1% 2008-55 NJW-293 1% 3% <1% . <1% 2008-56 NJW-304 <1% <1% <1% 4% 2% . 2008-57 NJW-307 2% 3% 1% 1% . <1% C03-58c NJW-317 <1% <1% <1% . 18% C04-59c NJW-318 2% . <1% 1% C05-60c NJW-319 1% 1% . C06-61c NJW-320 <1% 1% . C07-62c NJW-321 <1% . . C10-63c NJW-324 <1% <1% <1% . C13-64c NJW-327 <1% 1% 1% <1% <1% 1% C15-65c NJW-329 <1% . . C17-66c NJW-331 2% 1% . C18-67c NJW-332 2% <1% 1% <1% . 310

PAGE 311

Table B-6. Estimated frequency data (1). sample NJWID temper silt coarse Q vc Q polyX Q felsicR Kspar plag UID feld EPheavy amphib epidote mica1 mica2 2004-01 NJW-242 GROG 3-5% 3% 1% 1% P P P P 1% L P 2004-03 NJW-029 GRIT 3% P 1-3% 1-3% 1% 2% 1% 1-3% M 3%b 2004-18 NJW-038 GRIT 3-5% 3-5% 1-3% 1-3% 1% P P 1% P P L P 2005-01 NJW-032 CHAR 3-5% P 1% P P P 1% L 1%b 2005-02 NJW-043 CHAR 5% 1% P 1-3% P P 1% P 1% M 3%b 2005-03 NJW-313 CHAR 3% P 1% P P P 1% L 1%b 2005-08 NJW-072 CHAR 3% 1% P 1% P P P P 1% L P 2005-09 NJW-073 CHAR 3% 1-3% P P 1% P P 1% P L P 2005-23 NJW-098 CHAR 5% P P P P P L 1%b 2005-24 NJW-026 CHAR 3-5% P P P L 1%b 2005-25 NJW-025 CHAR 3-5% P P P P P L 1%b 2005-26 NJW-047 CHAR 3% P P P P P P 1% N 2008-01 NJW-001 CHAR 3-5% 1% P P P P P P L 1%b 2008-02 NJW-007 GRIT 3% 1-3% 1% 1% 1% P P 1% P P L Pb 2008-03 NJW-008 GRIT 3-5% 3% 1% 1-3% P P 1% 1-3% 1% P P L 1%b 2008-04 NJW-010 GSAND 3-5% 3% 1% 1% 1% P P 2% P P L Pb 2008-05 NJW-017 SAND 5-10% P 1-3% P 1% 2% P P LM 1-3%b 2008-06 NJW-021 SAND 3% P P 1% P 1% P P P L Pb 2008-07 NJW-102 SAND 3-5% 1% P 1-3% 1% 1% 2% 1-3% M 3-5% 2008-08 NJW-111 SAND 3-5% 1% P P 1% P P 2% P P L Pb 2008-09 NJW-112 CHAR 3-5% P P 1-3% 1% 1% 1% 1% M 3%b 2008-10 NJW-119 SAND 3% P P P P P P P L Pb 2008-11 NJW-122 CHAR 5-10% P P P P P P L 1%b 2008-12 NJW-136 CHAR 5-10% P P P 1% P 1% 2% P 1% LM 1-3%b 2008-13 NJW-138 SAND 3-5% P P P P P P LM 1-3%b 2008-14 NJW-139 SAND 3-5% P 1-3% P P 1% P 1% L Pb 2008-15 NJW-145 SAND 3-5% P 1% P P P P L Pb 2008-16 NJW-079 GRIT 3% 3% 1% 1-3% P P P 1% P N 2008-17 NJW-081 CHAR 3-5% 1% 1% P 1% P P L 1%b 2008-18 NJW-085 GROG 5-10% P P P 1-3% 1% 1% 2% P 1% L 2008-19 NJW-093 CHAR 3% P P 1-3% 1% 1% P P P L P 2008-20 NJW-091 SAND 5-10% P P 1-3% P 1-3% 2% P 1% LM 1-3%b 2008-21 NJW-152 SAND 3-5% P 1-3% P P 1% 1% LM 1-3%b 311

PAGE 312

Table B-6. Continued sample NJWID temper silt coarse Q vc Q polyX Q felsicR Kspar plag UID feld EPheavy amphib epidote mica1 mica2 2008-22 NJW-151 SAND 5-10% P 1% 1% 1% 2% 1% 1% LM 1-3%b 2008-23 NJW-158 GRIT 3-5% 3% 1% 1-3% 1% 1% 1% 1% P L P 2008-24 NJW-164 SAND 3% P P P P P 1% P P L Pb 2008-25 NJW-169 SAND 3-5% P P 1% P P P P P L 1%b 2008-26 NJW-174 GRIT 3-5% 3-5% 1% 1-3% P P 1% P P L Pb 2008-27 NJW-177 GSAND 3% 1-3% P P P P P P L Pb 2008-28 NJW-190 GSAND 3-5% 3-5% 1% 1-3% 1% P P P P P L P 2008-29 NJW-191 GRIT 3-5% 3-5% 1% 1% P P P 1% P P L Pb 2008-30 NJW-192 SAND 3% P 1-3% P P 1% P 1% M 3%b 2008-31 NJW-193 GRIT 3-5% 3-5% 1% 1% P P 1% 1% P L 1%b 2008-32 NJW-027 GRIT 3% 3-5% 13% 1-3% P P 1% 1% P L Pb 2008-33 NJW-050 GRIT 3-5% 5% 13% 1% P P P P P P N 2008-34 NJW-053 GSAND 3% 1% P 1% 1-3% 1% 1% 1% P 1% L Pb 2008-35 NJW-055 SAND 3-5% 1% P P 1% 1% P P P P N 2008-36 NJW-059 GRIT 3-5% 3-5% 13% 1-3% P P P P P P L Pb 2008-37 NJW-063 GRIT 3-5% 3-5% 1% 1-3% 1% P P P P P L Pb 2008-38 NJW-064 GSAND 3% 1-3% P 1% P P P 1% P L Pb 2008-39 NJW-195 GSAND 3% 3% P 1-3% 1% P 1% P 1% L P 2008-40 NJW-196 GRIT 3-5% 3-5% 1% 1% 1% P P P P L Pb 2008-41 NJW-197 GRIT 3% 3-5% P 1% P P P P P N 2008-42 NJW-198 GRIT 3-5% 3-5% 1% 1-3% 1-3% P P 1% P P LM 1-3%b 2008-43 NJW-199 SAND 5% P 1% 1% 1-3% 1% 1% 2% P 1-3% M 5%b 2008-44 NJW-218 SAND 5% P P 1% P P 1% P P LM 1-3%b 2008-45 NJW-219 GRIT 3% 3-5% 1% 1-3% 1% P P P L Pb 2008-46 NJW-220 GROG 5-10% 1% 1% P P P 2% P P LM 1-3%b 2008-47 NJW-221 GRIT 3-5% 3% P 1-3% P P P P P N 2008-48 NJW-222 GROG 3% 1-3% P 1% 1% P P P P 1% L Pb 2008-49 NJW-241 GSAND 3% 3% P P P P P 1% P N 2008-50 NJW-257 GSAND 3-5% 3% P 1% P P P 2% P P N 2008-51 NJW-260 GSAND 5-10% P 1% 1% 1% P P 2% P P M 3%b 2008-52 NJW-277 GRIT 3-5% 1-3% P 1% P P P P N 312

PAGE 313

Table B-6. Continued sample NJWID temper silt coarse Q vc Q polyX Q felsicR Kspar plag UID feld EPheavy amphib epidote mica1 mica2 2008-53 NJW-279 GSAND 3-5% 1-3% P 1% 1% P P 1% P P LM 1-3%b 2008-54 NJW-281 GRIT 3-5% 3% 1% 1-3% P P P 1% P L Pb 2008-55 NJW-293 GRIT 3% 3% 1% 1% P P P P P P L Pb 2008-56 NJW-304 GROG 3% P P 1% P P 1% P L Pb 2008-57 NJW-307 GRIT 3% 3% 1-3% 1% P P P P P P N C03-58c NJW-317 CLAY 3-5% P P P P P P P P L P C04-59c NJW-318 CLAY 1-3% P P P P 1% P P L Pb C05-60c NJW-319 CLAY 3-5% P P P P P P P N C06-61c NJW-320 CLAY 3% . 1% P P 1% P P L 1%b C07-62c NJW-321 CLAY 3% P P 1% 1% P P 1% P L 1% C10-63c NJW-324 CLAY 1-3% 1% P P P P 1% L P C13-64c NJW-327 CLAY 10% 1% P P 1% P P 1% P P M 5%b C15-65c NJW-329 CLAY 3% P 1% P P P P L 1%b C17-66c NJW-331 CLAY 3% 1% P 1% 1-3% P P 1% P P L Pb C18-67c NJW-332 CLAY 3-5% 1% P P P 1-3% 1% P 2% P 1% M 3%b 313 Table B-7. Estimated frequency data (2). sample NJWID ferric Fe sand grog clay lumps charcoal organic bone spc1 spc2 phyt1 phyt2 diatoms 2004-01 NJW-242 3-5% P . L P L P 2004-03 NJW-029 1-3% P 1-3% 1% P M 1-3% L P P 2004-18 NJW-038 Pdk P . L P L P 2005-01 NJW-032 1% P P 9% P L P L P 2005-02 NJW-043 1-3% P 1-3% 11% P N N 2005-03 NJW-313 Pdk 7% P N N 2005-08 NJW-072 Pdk P 1% 3% P N N 2005-09 NJW-073 Pdk 1-3% 2% P 4% L P L P 2005-23 NJW-098 1% P P 28% P L P LM 1% 2005-24 NJW-026 3% 1% 1-3% 23% P 1% L P L P 2005-25 NJW-025 1-3% P P 6% P 1% L P L P 2005-26 NJW-047 Pdk 7% L P L P 2008-01 NJW-001 3-5% 1-3% P P L P L P

PAGE 314

Table B-7. Continued sample NJWID ferric Fe sand grog clay lumps charcoal organic bone spc1 spc2 phyt1 phyt2 diatoms 2008-02 NJW-007 Pdk . P M 3% N 2008-03 NJW-008 Pdk 1-3%Fe N N 2008-04 NJW-010 3% 1% P L P L P 2008-05 NJW-017 Pdk P P N N 2008-06 NJW-021 3% P . N N 2008-07 NJW-102 3-5% P P N N 2008-08 NJW-111 1-3% 1% 1% P L P N 2008-09 NJW-112 1-3% P 5% L P(g) L P(g) 2008-10 NJW-119 3-5% 1% . N N 2008-11 NJW-122 1-3% 1% 12% 3-5% L P LM 1% 2008-12 NJW-136 3% P P P N N 2008-13 NJW-138 3% . P N N 2008-14 NJW-139 3% P 1% or 1% N N 2008-15 NJW-145 3% P L P L P 2008-16 NJW-079 3-5% 3% M 3% L P 2008-17 NJW-081 1% P P 4% N N 2008-18 NJW-085 3% 1% 3% 1% N N 2008-19 NJW-093 1% 1% P 11% 3-5% N N 2008-20 NJW-091 1-3% . P N N 2008-21 NJW-152 3% P P N L P 2008-22 NJW-151 3% P 1-3% . N N 2008-23 NJW-158 Pdk P 1% L P LM 1% 2008-24 NJW-164 1-3% P 1% P N N 2008-25 NJW-169 1-3% P N N 2008-26 NJW-174 Pdk P . N L P 2008-27 NJW-177 3% 1% P L P LM 1% 2008-28 NJW-190 3% P P L P L P 2008-29 NJW-191 1% P L P LM 1% 2008-30 NJW-192 3% P . N L P 2008-31 NJW-193 3% P . L P LM 1% 2008-32 NJW-027 Pdk P P P L P LM 1% 2008-33 NJW-050 Pdk P P L P L P 2008-34 NJW-053 3% P 1% . N L P 2008-35 NJW-055 Pdk 1% . L P M 1-3% 314

PAGE 315

Table B-7. Continued sample NJWID ferric Fe sand grog clay lumps charcoal organic bone spc1 spc2 phyt1 phyt2 diatoms 2008-36 NJW-059 Pdk P P P P L P L P 2008-37 NJW-063 1-3% . L P M 1-3% 2008-38 NJW-064 3% 1% P P N N 2008-39 NJW-195 3% 1% . L P LM 1% 2008-40 NJW-196 3% P . L P LM 1% 2008-41 NJW-197 3-5% P P L P M 3% 2008-42 NJW-198 Pdk . M 3% L P 2008-43 NJW-199 3-5% P N N 2008-44 NJW-218 3% P P N L P 2008-45 NJW-219 3-5% 1% . L P LM 1% 2008-46 NJW-220 1-3% 3-5% 1% M 3% M 3% 2008-47 NJW-221 3-5% P . L P L P 2008-48 NJW-222 1-3% P 3-5% P P L P M 3% 2008-49 NJW-241 3% . L P LM 1% 2008-50 NJW-257 3% 1% 1% . N N 2008-51 NJW-260 3% P P M 3-5% LM 1% 2008-52 NJW-277 3% P P L P L P 2008-53 NJW-279 Pdk P P M 1-3% L P 2008-54 NJW-281 Pdk . P M 1-3% M 1-3% 2008-55 NJW-293 Pdk P M 3% LM 1% 2008-56 NJW-304 3-5% P 3-5% 3%Fe P N N 2008-57 NJW-307 3-5% P P P L P LM 1% C03-58c NJW-317 1-3% P 3% H 18% N C04-59c NJW-318 Pdk . N N C05-60c NJW-319 3-5% P . P L P L P C06-61c NJW-320 3% . 3-5% L P L P C07-62c NJW-321 1-3% . L P N 3% C10-63c NJW-324 3% P P L P N 3% C13-64c NJW-327 1-3% . P M 3% M 1-3% P C15-65c NJW-329 Pdk . P L P L P P? C17-66c NJW-331 3-5% P P L P C18-67c NJW-332 3-5% P P M 1-3% L P P 315

PAGE 316

Table B-8. Sand size data. sample NJWID temper vf Q fine Q med Q crs Q vc Q T QTZ vf other f other m other c other vc other g-p other Tother TpxQ TQPQ 2004-01 NJW-242 GRIT 18 43 41 23 5 130 4 3 3 10 7 137 2004-03 NJW-029 GROG 59 47 1 107 7 5 . 12 2 109 2004-18 NJW-038 GRIT 12 16 27 34 5 94 1 1 3 1 1 7 6 100 2005-01 NJW-032 CHAR 31 47 4 1 83 2 . 2 1 84 2005-02 NJW-043 CHAR 31 32 9 72 2 3 . 5 2 74 2005-03 NJW-313 CHAR 32 122 7 161 1 . 1 1 162 2005-08 NJW-072 CHAR 12 77 16 4 109 4 7 . 11 5 114 2005-09 NJW-073 CHAR 16 75 20 2 113 2 6 . 8 6 119 2005-23 NJW-098 CHAR 6 15 3 24 2 . 2 24 2005-24 NJW-026 CHAR 12 27 2 1 42 2 . 2 1 43 2005-25 NJW-025 CHAR 18 31 4 53 1 . 1 53 2005-26 NJW-047 CHAR 20 44 8 72 2 3 4 9 6 78 2008-01 NJW-001 CHAR 24 55 5 84 4 1 . 5 1 85 2008-02 NJW-007 GRIT 1 20 39 19 7 86 1 5 7 1 14 9 95 2008-03 NJW-008 GRIT 17 25 13 13 4(1g) 72 3 6 11 8 4 1 33 21 93 2008-04 NJW-010 GSAND 23 46 37 22 4 132 4 6 3 3 1 17 12 144 2008-05 NJW-017 SAND 39 81 4 124 4 6 . 10 4 128 2008-06 NJW-021 SAND 25 74 4 103 1 3 1 5 1 104 2008-07 NJW-102 SAND 38 26 3 1 68 5 2 3 10 4 72 2008-08 NJW-111 SAND 15 92 13 3 1 124 6 8 2 1 17 6 130 2008-09 NJW-112 CHAR 24 33 9 66 7 2 . 9 66 2008-10 NJW-119 SAND 16 96 11 123 2 6 2 10 5 128 2008-11 NJW-122 CHAR 8 22 3 33 1 . 1 1 34 2008-12 NJW-136 CHAR 22 86 12 1 121 4 13 1 1 19 9 130 2008-13 NJW-138 SAND 14 53 9 2 78 2 5 . 7 3 81 2008-14 NJW-139 SAND 24 96 10 130 5 10 1 16 12 142 2008-15 NJW-145 SAND 37 41 2 80 6 9 . 15 4 84 2008-16 NJW-079 GRIT 4 24 30 17 4 79 3 4 10 10 4 1 32 24 103 2008-17 NJW-081 CHAR 10 55 5 70 2 10 1 13 6 78 2008-18 NJW-085 GROG 41 101 15 1 158 8 2 2 12 4 162 2008-19 NJW-093 CHAR 14 45 1 60 1 7 . 8 5 65 2008-20 NJW-091 SAND 19 85 13 117 2 4 . 6 1 118 2008-21 NJW-152 SAND 32 27 59 6 3 . 9 4 63 316

PAGE 317

Table B-8. Continued sample NJWID temper vf Q fine Q med Q crs Q vc Q T QTZ vf other f other m other c other vc other g-p other Tother TpxQ TQPQ 2008-22 NJW-151 SAND 47 45 4 96 15 6 . 21 5 101 2008-23 NJW-158 GRIT 6 33 44 27 6 116 1 7 4 12 9 125 2008-24 NJW-164 SAND 40 97 8 145 5 8 1 14 7 152 2008-25 NJW-169 SAND 30 114 8 1 153 4 5 . 9 5 158 2008-26 NJW-174 GRIT 18 53 48 36 5 160 3 13 10 10 1 37 30 190 2008-27 NJW-177 GSAND 19 61 31 9 120 1 5 2 1 9 6 126 2008-28 NJW-190 GSAND 15 83 55 30 6 189 5 14 8 6 1 34 26 215 2008-29 NJW-191 GRIT 6 45 50 29 7 137 2 10 5 0 21 16 153 2008-30 NJW-192 SAND 71 107 3 181 10 16 1 27 10 191 2008-31 NJW-193 GRIT 21 54 49 37 6 167 1 6 7 6 1 21 17 184 2008-32 NJW-027 GRIT 15 61 74 46 12 208 4 2 8 6 2 2 24 21 229 2008-33 NJW-050 GRIT 12 30 52 39 7 140 4 3 5 12 10 150 2008-34 NJW-053 GSAND 25 76 37 11 149 2 4 3 9 5 154 2008-35 NJW-055 SAND 33 32 13 6 84 2 3 . 5 3 87 2008-36 NJW-059 GRIT 11 44 40 25 7 127 2 4 8 11 2 27 23 150 2008-37 NJW-063 GRIT 14 37 38 42 7 138 4 8 7 1 20 15 153 2008-38 NJW-064 GSAND 24 77 58 20 1 180 7 7 6 2 2 24 14 194 2008-39 NJW-195 GSAND 28 55 46 25 1 155 6 1 5 1 1 14 8 163 2008-40 NJW-196 GRIT 9 33 52 38 7 139 4 5 2 6 1 1p 19 13 152 2008-41 NJW-197 GRIT 7 29 38 25 2 101 2 4 4 2 4 16 10 111 2008-42 NJW-198 GRIT 12 22 34 46 10 124 3 5 4 16 4 32 25 149 2008-43 NJW-199 SAND 47 38 2 8 3 98 8 5 1 1 3 18 7 105 2008-44 NJW-218 SAND 31 87 7 125 5 9 . 14 7 132 2008-45 NJW-219 GRIT 18 40 52 45 15 170 5 6 7 6 24 19 189 2008-46 NJW-220 GROG 15 46 35 3 99 8 7 5 20 9 108 2008-47 NJW-221 GRIT 24 34 46 42 4 150 3 1 6 3 13 10 160 2008-48 NJW-222 GROG 36 62 41 6 1 146 5 11 8 3 27 17 163 2008-49 NJW-241 GSAND 25 42 39 7 2 115 4 1 3 2 10 7 122 2008-50 NJW-257 GSAND 26 79 52 31 1 189 2 3 11 4 1 21 15 204 2008-51 NJW-260 GSAND 16 8 3 9 5 41 3 3 6 3 44 2008-52 NJW-277 GRIT 17 7 15 11 3 53 1 2 1 3 7 6 59 2008-53 NJW-279 GSAND 18 20 17 6 3 64 4 2 4 2 12 5 69 2008-54 NJW-281 GRIT 5 17 31 15 1 69 2 2 3 2 2 11 8 77 2008-55 NJW-293 GRIT 9 12 25 25 5 76 1 5 8 2 1 17 11 87 317

PAGE 318

Table B-8. Continued sample NJWID temper vf Q fine Q med Q crs Q vc Q T QTZ vf other f other m other c other vc other g-p other Tother TpxQ TQPQ 2008-56 NJW-304 GROG 12 40 2 2 56 3 1 . 4 1 57 2008-57 NJW-307 GRIT 6 20 36 18 7 87 2 3 6 3 14 10 97 C03-58c NJW-317 CLAY 32 52 5 89 5 4 . 9 2 91 C04-59c NJW-318 CLAY 16 23 8 1 48 3 4 2 1 10 5 53 C05-60c NJW-319 CLAY 36 85 33 2 156 3 3 1 7 4 160 C06-61c NJW-320 CLAY 23 37 4 64 4 3 . 7 1 65 C07-62c NJW-321 CLAY 14 46 2 2 64 3 6 . 9 2 66 C10-63c NJW-324 CLAY 19 78 23 5 125 1 3 . 4 1 126 C13-64c NJW-327 CLAY 25 9 2 3 1 40 4 2 3 1 10 4 44 C15-65c NJW-329 CLAY 5 16 7 28 1 . 1 1 29 C17-66c NJW-331 CLAY 25 25 10 4 64 2 3 3 1 9 5 69 C18-67c NJW-332 CLAY 54 51 6 3 114 7 6 2 2 1 18 7 121 Table B-9. Sand size data and indices. sample NJWID Tfeld Theavy Tvf T fine T med T c T vc T g-p Tsand Tnon sum.5 sum1 SSI.5 SSI1 2004-01 NJW-242 2 22 46 41 26 5 140 11 237 248 1.69 1.77 2004-03 NJW-029 5 5 66 52 1 119 19 87 120 0.73 1.01 2004-18 NJW-038 1 13 17 30 35 6 101 2 212.5 219 2.1 2.17 2005-01 NJW-032 1 33 47 4 1 85 34 0.88 1.07 2005-02 NJW-043 1 2 32 34 9 75 42 0.91 1.12 2005-03 NJW-313 32 123 7 162 28 0.94 1.04 2005-08 NJW-072 5 1 16 83 16 4 119 20 1.13 1.20 2005-09 NJW-073 2 18 81 20 2 121 36 1.12 1.20 2005-23 NJW-098 2 8 15 3 26 91 0.96 1.12 2005-24 NJW-026 1 14 27 2 1 44 103 0.93 1.09 2005-25 NJW-025 1 19 31 4 54 34 0.90 1.07 2005-26 NJW-047 22 47 9 78 28 0.97 1.12 2008-01 NJW-001 4 28 56 5 89 7 80 94 0.90 1.06 2008-02 NJW-007 3 1 21 44 26 7 1 100 0 220.5 221 2.20 2.21 2008-03 NJW-008 8 20 31 24 21 7 2 105 1 190 200 1.81 1.90 2008-04 NJW-010 5 27 52 40 25 5 149 6 240.5 254 1.61 1.70 2008-05 NJW-017 4 3 43 87 4 134 3 116.5 138 0.87 1.03 2008-06 NJW-021 4 26 77 5 108 2 100 113 0.93 1.05 318

PAGE 319

Table B-9. Continued sample NJWID Tfeld Theavy Tvf T fine T med T c T vc T g-p Tsand Tnon sum.5 sum1 SSI.5 SSI1 2008-07 NJW-102 6 1 43 28 6 1 78 5 64.5 86 0.83 1.10 2008-08 NJW-111 9 2 21 100 15 3 2 141 7 157.5 168 1.12 1.19 2008-09 NJW-112 7 1 31 35 9 75 20 68.5 84 0.91 1.12 2008-10 NJW-119 4 2 18 102 13 133 9 137 146 1.03 1.10 2008-11 NJW-122 8 23 3 34 36 33 37 0.97 1.09 2008-12 NJW-136 9 1 26 99 13 2 140 3 144 157 1.03 1.12 2008-13 NJW-138 4 16 58 9 2 85 4 90 98 1.06 1.15 2008-14 NJW-139 3 29 106 11 146 7 142.5 157 0.98 1.08 2008-15 NJW-145 10 1 43 50 2 95 2 75.5 97 0.80 1.02 2008-16 NJW-079 8 7 28 40 27 8 1 111 6 229.5 233 2.07 2.10 2008-17 NJW-081 5 1 12 65 6 83 14 83 89 1.00 1.07 2008-18 NJW-085 7 1 49 103 17 1 170 14 165.5 190 0.97 1.12 2008-19 NJW-093 3 15 52 1 68 30 61.5 69 0.90 1.02 2008-20 NJW-091 5 21 89 13 123 1 125.5 136 1.02 1.11 2008-21 NJW-152 5 2 38 30 . 68 7 49 68 0.72 1.00 2008-22 NJW-151 16 1 62 51 4 117 9 90 121 0.77 1.03 2008-23 NJW-158 2 1 7 40 44 31 5 1 128 4 249.5 253 1.95 1.98 2008-24 NJW-164 6 1 45 105 9 159 6 145.5 168 0.92 1.06 2008-25 NJW-169 3 2 34 119 8 1 162 4 155 172 0.96 1.06 2008-26 NJW-174 6 1 21 66 58 46 5 1 197 3 355.5 366 1.80 1.86 2008-27 NJW-177 2 1 20 66 33 10 129 9 172 182 1.33 1.41 2008-28 NJW-190 6 2 20 97 63 36 7 223 2 369 379 1.66 1.70 2008-29 NJW-191 3 1 8 54 55 33 6 1 157 3 296 300 1.89 1.91 2008-30 NJW-192 17 3 81 123 4 208 5 171.5 212 0.82 1.02 2008-31 NJW-193 3 1 22 60 56 43 7 188 3 340 351 1.81 1.87 2008-32 NJW-027 3 19 63 82 52 13 3 232 2 459.5 469 1.98 2.02 2008-33 NJW-050 2 12 34 55 44 7 152 2 310 316 2.04 2.08 2008-34 NJW-053 4 27 80 40 11 158 11 206.5 220 1.31 1.39 2008-35 NJW-055 2 35 35 13 6 89 1 96.5 114 1.08 1.28 2008-36 NJW-059 4 13 48 48 36 8 1 154 3 295.5 302 1.92 1.96 2008-37 NJW-063 5 14 41 46 49 8 158 1 319 326 2.02 2.06 2008-38 NJW-064 10 31 84 64 22 3 204 7 305.5 321 1.50 1.57 2008-39 NJW-195 5 1 34 56 51 26 2 169 3 261 278 1.54 1.64 2008-40 NJW-196 6 13 38 54 44 8 1p 158 2 322.5 329 2.04 2.08 319

PAGE 320

320Table B-9. Continued sample NJWID Tfeld Theavy Tvf T fine T med T c T vc T g-p Tsand Tnon sum.5 sum1 SSI.5 SSI1 2008-41 NJW-197 5 9 33 42 27 6 117 5 226.5 231 1.94 1.97 2008-42 NJW-198 5 15 27 38 62 13 1 156 0 353.5 361 2.27 2.31 2008-43 NJW-199 8 2 55 43 3 9 6 116 3 127.5 155 1.10 1.34 2008-44 NJW-218 6 3 36 96 7 139 5 128 146 0.92 1.05 2008-45 NJW-219 4 18 45 58 52 21 194 5 410 419 2.11 2.16 2008-46 NJW-220 8 3 23 53 40 3 119 30 153.5 165 1.29 1.39 2008-47 NJW-221 3 27 35 52 45 4 163 3 303.5 317 1.86 1.94 2008-48 NJW-222 8 2 40 73 49 9 1 172 25 222 242 1.29 1.41 2008-49 NJW-241 2 1 29 43 42 9 2 125 3 176.5 191 1.41 1.53 2008-50 NJW-257 4 2 28 82 63 35 2 210 8 335 349 1.60 1.66 2008-51 NJW-260 2 3 19 8 3 9 8 47 13 82.5 92 1.76 1.96 2008-52 NJW-277 1 1 18 9 16 14 3 60 1 104 113 1.73 1.88 2008-53 NJW-279 7 22 22 21 8 3 76 1 111 122 1.46 1.60 2008-54 NJW-281 3 7 19 34 17 3 80 1 153.5 157 1.92 1.96 2008-55 NJW-293 5 9 13 30 33 7 1 93 1 209.5 214 2.25 2.30 2008-56 NJW-304 2 2 15 41 2 2 60 16 58.5 66 0.98 1.10 2008-57 NJW-307 4 8 23 42 21 7 101 6 202 206 2.00 2.04 C03-58c NJW-317 6 1 37 56 5 98 69 84.5 103 0.86 1.05 C04-59c NJW-318 3 2 19 27 10 2 58 3 62.5 72 1.08 1.24 C05-60c NJW-319 3 39 88 34 2 163 2 181.5 201 1.11 1.23 C06-61c NJW-320 4 3 27 40 4 71 3 61.5 75 0.87 1.06 C07-62c NJW-321 6 1 17 52 2 2 73 0 70.5 79 0.97 1.08 C10-63c NJW-324 3 20 81 23 5 129 2 152 162 1.18 1.26 C13-64c NJW-327 6 4 29 11 5 3 2 50 11 52.5 67 1.05 1.34 C15-65c NJW-329 0 1 5 17 7 29 1 33.5 36 1.16 1.24 C17-66c NJW-331 3 2 27 28 13 5 73 3 82.5 96 1.13 1.32 C18-67c NJW-332 10 61 57 8 5 1 132 5 122.5 153 0.93 1.16

PAGE 321

Table B-10. Key to headings and ab breviations for petrographic data. Abbreviation Description sample Ann Cordell's sample id numbers (Appendix A1-A6) NJWID Neill Wallis's sample id numbers (Appendix A1-A6) interval counting interval (Appendix A1) petpaste petrographic paste/resource category (Appendix A6) match paddle match id number (Appendix A6) corec core color/dregree of co ring (see below) (Appendix A6) refcor refired core color (see below) (Appendix A6) aplast aplastics (Appendix A3, A6) sand sum of quartz, polyx quartz, feldspars, and heavies (Appendix A3, A5, A6) nonsand, non sum of other aplastics (mica, ferric, spc, etc.) (Appendix A3, A5) Q, QTZ quartz (Appendix A1, A3-A6) Tqtz total quartz (Appendix A5, A6) Tqpq sum of quartz and polyx quartz (Appendix A3, A5) polyxQ, pxQ polycrystalline quartz or quartzite (Appendix A1-A6) felsicR felsic or granitic ro ck fragment (Appendix A1-A4) kspar microcline or potassium feldspar (Appendix A1, A2, A4, A6) plag plagioclase (Appe ndix A1, A2, A4, A6) UID feld UID feldspar (Appendix A1, A2, A4, A6) Feld sum of feldspars (Appendix A3, A5, A6) amphib amphibole (Appendix A1, A2, A4, A6) heavy UID minerals (Appendix A1-A3) heavy sum of amphibole, epidote and UID minerals (Appendix A4-A6) other sum of feldpsars, polyxQ and heavies (Appendix A5) ferric ferric concretions or nodules (Appendix A1-A4, A6) Fe sand ferric with imbedded quartz (Appendix A1-A4, A6) clylmp clay lumps (Appendix A3) charcoal charcoal temper (Appendix A1, A2, A4, A6) char charcoal temper (Appendix A3) spc sponge spicules (Appendix A1, A3, A4, A6) phyt phytoliths (Appendix A1, A3, A4, A6) SSI.5 sand size index (with very fine grains counting as .5) (Appendix A5, A6) SSI1 sand size index (with very fine counting as1) (Appendix A5, A6) mica1 grouped relative frequency of mica (see below) (Appendix A4, A6) mica2 relative frequency of mica (P=present, rare; b=biotite or pleochroic mica) (Appx A4, A6) spc1 grouped relative frequency of sponge spicules (see below) (Appendix A4, A6) spc2 estimated and/or computed percent of sponge spicules (P=present, rare) (Appx A4, A6) phyt1 grouped relative frequency of phytoliths (see below) (Appendix A4, A6) phyt2 estimated and/or computed percent of phytoliths (P=present, rare) (Appendix A4, A6) Esilt estimated percent silt (Appendix A6) EPsilt average of estimated percent silt (Appendix A6) EPfeld estimated percent combined feldspars (Appendix A6) EPheavy estimated percent heavies (uid heavy minerals plus amphibole and epidote) (Appx A6) EPKspar estimated percent microcline or potassium feldspar (Appendix A6) EPplag estimated percent plagioclase (Appendix A6) EPUIDfeld estimated percent uid feldspars (Appendix A6) EPamphib estimated percent amphibole (Appendix A6) EPepidote estimated percent epidote (Appendix A6) EPferric estimated percent ferric concretions/nodules (Appendix A6) EPFesand estimated percent ferric with imbedded quartz (Appendix A6) 321

PAGE 322

Table B-10. Continued Abbreviation Description 1x.5 counting interval 1mm by .5mm, counted once 1x1 counting interval 1mm by 1mm, counted once 1x1x2 counting interval 1mm by 1mm, counted twice CHAR charcoal temper GROG grog temper SAND sand temper GSAND grit and sand temper GRIT grit temper CLAY clay sample vf very fine f fine m, med medium c, crs coarse vc very coarse g, gr granule pb pebble P present, rare, <1% not observed N not observed L low (present, rare to occasional, up to 1%) LM low to moderate (occasional to frequent, 1-3% M moderate (frequent, > 3%) L present, rare (for spc1 and phyt1) LM low to moderate (occasional to frequent, 1% (for spc1 and phyt1) M moderate (occasional to frequent, > 1-3%) (for spc1 and phyt1) H high (frequent to common) (for spc1 and phyt1) 322

PAGE 323

APPENDIX C RIM AND BASE P ROFILES Figure C-1. Rim and base profiles fro m the Tillie Fowler site (8DU17245). 323

PAGE 324

Figure C-2. Rim and base profiles from Greenfield site #8/9 (8DU5544/5). 324

PAGE 325

Figure C-3. Rim and base profile s from Greenfield site #7 (8DU5543). 325

PAGE 326

Figure C-4. Rim profiles from various sites. 326

PAGE 327

Figure C-5. Rim and base prof iles from the Dent mound (8DU68). 327

PAGE 328

Figure C-6. Rim and base prof iles from Cathead Creek (9MC360). Figure C-7. Rim profiles from Evelyn (9GN6) shell midden. 328

PAGE 329

Figure C-8. Rim profiles fr om Lewis Creek (9MC16). 329

PAGE 330

Figure C-9. Rim profiles from McArthur Estates (8NA32). 330

PAGE 331

331 Figure C-10. Rim profiles from Sidon (9MC372).

PAGE 332

LIST OF REFERE NCES Adams, William Hampton (editor) 1985 Aboriginal Subsistence a nd Settlement Archaeology of the Kings Bay Locality, Vol. 1. University of Florida, Department of Anthropology Reports of Investigations No.1, Gainesville. Allen, Catherine J. 1998 When Utensils Revolt: Mind, Matter, and Models of Being in the Pre-Colombian Andes. RES: Anthropology and Aesthetics 33:19-27. Anderson, David G. 1985 Middle Woodland Societies on the Lower South Atlantic Slope: A View from Georgia and South Carolina. Early Georgia 13:29. 1995 Paleoindian Interaction Ne tworks in the Eastern Woodlands. In Native American Interactions edited by Michael Nassaney and Kenneth Sassaman, pp. 3-26. University of Tennessee Press, Knoxville. 1998 Swift Creek in a Regional Perspective. In A World Engraved: Archaeology of the Swift Creek Culture, edited by M. Williams and D. Elliott, pp. 274-300. University of Alabama Press, Tuscaloosa. Anderson, David G., and Glen T. Hanson 1988 Early Archaic Settlement in the Sout heastern United States: A Case Study from the Savannah River Valley. American Antiquity 53:262-286. Anderson, Warren, and D.A. Goolsby 1973 Flow and Chemical Characteristics of th e St. Johns River at Jack sonville, Florida. State of Florida, Department of Natural Resources, Information Circular No. 82, Tallahassee. Appadurai, Arjun 1986 Introduction: commodities and the politics of value. In The Social Life of Things: Commodities in a Cultural Perspective edited by Arjun Appadurai, pp. 3-63. Cambridge University Press, Cambridge. Arnold, Dean E. 1998 Ancient Andean Ceramic Technol ogy: An Ethnoarchaeological Perspective. In Andean Ceramics: Technology, Organization, and Approaches ed. by Izumi Shimada, pp. 353367. MASCA Research Papers in Science and Archeology Supplement to Volume 15, 1998. Museum Applied Science Center for Archaeology, University of Pennsylvania of Archaeology and Anthropology, Philadelphia, PA. Arthur, John W. 2002 Pottery Use-Alteration as an Indicator of Socioeconomic Status: An Ethnoarchaeol ogical Study of the Gamo of Ethiopia. Journal of Archaeological Method and Theory 9(4):331-355. 332

PAGE 333

Ashley, Keith H. 1992 Swift Creek Manifestati ons along the Lower St. Johns River. The Florida Anthropologist 45:127-138. 1995 The Dent Mound: A Coastal W oodland Period Burial Mound Near the Mouth of the St. Johns River. The Florida Anthropologist 48:13-35. 1998 Swift Creek Traits in Northeaste rn Florida: Ceramics, Mounds, and Middens. In A World Engraved: Archaeology of the Swift Creek Culture edited by M. Williams and D.T. Elliott, pp. 197-221. University of Alabama Press, Tuscaloosa. 2003 Interaction, Population Movement and Political Economy: The Changing Social Landscape of Northeastern Florida. Ph.D. dissertation, Department of Anthropology, University of Florida, Gainesville. Ashley, Keith H., and Greg S. Hendryx 2008 Archaeological Site Testing and Data Recovery and Mitigation at the Dolphin Reef Site (8DU276), Duval County, Florida. Report on file, Division of Historical Resources, Tallahassee. Ashley, Keith H., and Neill Wallis 2006 Northeastern Florida Swift Creek: An Overview an d Future Research Directions. The Florida Anthropologist 59:5-18. Ashley, Keith, Keith Stephenson, and Frankie Snow 2007 Teardrops, Ladders, and Bulls Eyes: Swift Creek on the Georgia Coast. Early Georgia 35(1):3-28. Battaglia, Debbora 1983 Projecting personhood in Melanesia: Man 18:289-304. 1990 On the Bones of the Serpent: Person, Memory, and Mortality in Sabarl Island Society University of Chicago Press, Chicago. Baxter, Michael J. 1992 Archaeological uses of th e biplota neglected technique? In Computer Applications and Quantitative Methods in Archaeology, 1991 edited by G. Lock and J. Moffett, pp. 141-148. BAR International Se ries (S577) Tempvs Reparatvm, Archaeological and Historical Associates, Oxford. 1994 Exploratory Multivariate Analysis in Archaeology Edinburgh University Press, Edinburgh. Baxter, Michael J. and Caitlin E. Buck 2000 Data Handling and Sta tistical Analysis. In Modern Analytical Methods in Art and Archaeology edited by E. Ciliberto and G. Spoto, pp. 681-746. John Wiley and Sons, Inc., New York. Bell, Amelia R. 1990 Separate People: Speaking of Creek Men and Women. American Anthropologist 92:332-45. 333

PAGE 334

Bense, Judith A. 1998 Santa Rosa-Swift Cr eek in Northwest F lorida. In A World Engraved: Archaeology of the Swift Creek Culture edited by J.M. Williams and D.T. Elliott, pp. 247-73. University of Alabama Press, Tuscaloosa. Bense, Judith A., and Thomas C. Watson 1979 A Swift Creek and Weeden Island R ing Midden in the St. Andrews Bay Drainage System on the Northwest Florida Gulf Coast. Journal of Alabama Archaeology 25:85137. Bieber, Alan M., Jr., Dorothea W. Brooks, Ga rman Harbottle, and Edward V. Sayre 1976 Application of multiv ariate techniques to analytic al data on Aegean ceramics. Archaeometry 18:59. Binford, Lewis R. 1962 Archaeology as Anthropology. American Antiquity 28:217-25. 1965 Archaeological Systematics and the Study of Culture Process. American Antiquity 31:203-10. Bishop, Ronald L. and Hector Neff 1989 Compositional data analysis in archaeology. In Archaeological Chemistry IV edited by R. O. Allen, pp. 576. Advances in Chemistry Series 220, American Chemical Society, Washington, D.C. Bishop, Ronald L., Robert L. Rands, and George R. Holley 1982 Ceramic compositional analys is in archaeological perspective. In Advances in Archaeological Method and Theory, vol. 5, pp. 275. Academic Press, New York. Blanton, Richard E., Gary M. Feinman, St ephen A Kowaleswki, and Peter Peregrine 1995 A Dual-Processual Theory for th e Evolution of Mesoamerican Civilization. Current Anthropology 37:1-14. Boas, Franz 1955 Primitive Art Dover Publications, Inc., New York. Bourdieu, Pierre 1977 Outline of a Theory of Practice Cambridge University Press, Cambridge. Bradley, Richard 2000 An Archaeology of Natural Places Routledge, London. Braun, David P. 1983 Pots as Tools. In Archaeological Hammers and Theories edited by J. A. Moore and A. S. Keene, pp. 108-134. Academic Press, New York. 334

PAGE 335

1986 Midwestern H opewellian Exchange and Supralocal Interaction. In Peer Polity Interaction and Socio-Pol itical Change, edited by C. Re nfrew and C.F. Cherry, pp. 117-126. Cambridge University Press, Cambridge. Braun, David P. and Stephen Plog 1982 Evolution of Tribal Social Ne tworks: Theory and Prehis toric North American Evidence. American Antiquity 47:504-525. Brooks, Mark J., and Veletta Canouts 1984 Modeling Subsistence Change in the Late Pr ehistoric Period in the Interior Lower Coastal Plain of South Carolina Anthropological Studies 6, Occasional Papers of the South Carolina Institut e of Archaeology and Anthropology, University of South Carolina, Columbia. Brose, David 1979a A Speculative Model of the Role of Exchange in the Preh istory of the Eastern Woodlands. In Hopewell Archaeology: The Chillicothe Conference edited by David Brose and N. Greb er, pp. 3-8. Kent State Univer sity Press, Kent, Ohio. 1979b An Interpretation of the Hopewellian Traits in Florida. In Hopewell Archaeology: The Chillicothe Conference edited by David Brose and N. Greber, pp. 141-149. Kent State University Press, Kent, Ohio. 1994 Trade and Exchange in the Midwestern U.S. In Prehistoric Exchange Systems in North America, vol. 2, ed. T.G. Baugh and J.E. Ericson, pp. 215-40. Plenum Press, New York. Brose, David S. and Nomi Greber (editors) 1979 Hopewell Archaeology: The Chillicothe Conference. The Kent State University Press, Kent, Ohio. Brown, James A. 1985 Long-Term Trends to Sedentism and the Emergence of Complexity in the American Midwest. In Prehistoric Hunter-Gatherers: The Em ergence of Cultural Complexity, edited by T. D. price and J.A. Brown, pp. 201-231. Academic Press, New York. Brown, James A., R. A. Kerb er, and Howard D. Winters 1990 Trade and the Evolution of Exchange Relations at the Beginning of the Mississippian Period. In The Mississippian Emergence edited by B.D. Smith, pp. 251-63. Smithsonian Institution Press, Washington D.C. Broyles, Bettye J. 1968 Reconstructed Designs from Swift Creek Complicated Stamped Sherds. Southeastern Archaeological Conference Bulletin 8:49-74. Morgantown, West Virginia. 335

PAGE 336

Brum fiel, Elizabeth, and Timothy Earle 1987 Specialization, Exchange, and Complex Societies: An Introduction. In Specialization, Exchange, and Complex Societies, edited by Elizabeth Brumfiel and Timothy Earle, pp. 1-9. Cambridge University Press, Cambridge. Bullen, Ripley P., and John W. Griffin 1952 An Archaeological Survey of Amelia Island, Florida. The Florida Anthropologist 5:3764. Bullen, Ripley P., Adelain K. Bullen, and W.J. Bryant 1967 Archaeological Investigations at the Ross Hammock Site, Florida. William L. Bryant Foundation, American Studies Report 7. Butler, Judith 1993 Bodies that Matter: On the Discursive Limits of Sex Routledge, London. Byrd, John E. 1997 The Analysis of Diversity in Arch aeological Faunal Assemblages: Complexity and Subsistence Strategies in the Southeast during th e Middle Woodland Period. Journal of Anthropological Archaeology 16:49-72. Caldwell, Joseph R. 1952 The Archaeology of Eastern Georgia to South Carolina. In Archaeology of Eastern United States, edited by James B. Griffin, pp. 312-21. University of Chicago Press, Chicago. 1958 Trend and Tradition in the Prehistory of the Eastern United States Memior No. 88, American Anthropological Asso ciation, Scientific Papers, Volume X, Illinois State Museum, Springfield. 1964 Interaction Spheres in Prehistory. In Hopewellian Studies, edited by Joseph Caldwell and Robert Hall, pp. 133-43. Scientific Papers 12, Illinoi s State Museum, Springfield. 1970 Chronology of the Georgia Coast. S outheastern Archaeological Conference Bulletin 13:88-92. Carder, Nanny, Elizabeth J. Re itz, and J. Matthew Compton 2004 Animal Use in the Georgia Pine Barrens : An Example from the Hartford Site (9PU1). Southeastern Archaeology 23:25-40. Carr, Christopher 2006a Historical Insight into the Di rections and Limitations of Recent Research on Hopewell. In Gathering Hopwell: Society, Ritual, and Ritual Interaction edited by C. Carr and D.T. Case, pp. 51-70. Springer, New York. 2006b Rethinking Interregional Hopewellian Interaction. In Gathering Hopwell: Society, Ritual, and Ritual Interaction edited by C. Carr and D.T. Case, pp. 575623. Springer, New York. 336

PAGE 337

Carr, Christopher and Jill E. Nietzel 1995 Integrating Approaches to Ma terial Style in Theory and Philosophy. In Style, Society and Person edited by C. Carr and J. Nietzel, pp. 3-20. Springer, New York. Chance, Marsha A. 1974 The W .P.A. Glynn County Project: A Ceramic Analysis. Unpublished M.A. thesis, Department of Anthropology, Florida State University. Chapman, John 2000 Tension at Funerals: Social Practi ces and the Subversion of Community Structure in Later Hungarian Prehistory. In Agency in Archaeology edited by Marcia-Anne Dobres and John Robb, pp. 169-195. Routledge, New York. Chase, David W. 1998 Swift Creek: Lineage and Diffusion. In A World Engraved: Archaeology of the Swift Creek Culture, edited by M. Williams and D. Elliott, pp. 48-60. University of Alabama Press, Tuscaloosa. Clay, R. Berle 1998 Essential Features of Adena Ritual and their Implications. Southeastern Archaeology 17:1-21. Cobb, Charles R. 1991 Social Reproduction and the Longue Dure in the Prehistory of the Midcontinental United States. In Processual and Postprocessual Arc haeologies: Multiple Ways of Knowing the Past edited by R.W. Preucel, pp. 168-182. Occasional Paper No. 10. Center for Archaeological I nvestigations, Southern Illi nois University, Carbondale. Cogswell, James, Hector Neff, and Michael D. Glascock 1998 Analysis of Shell-Tempered Pottery Replicates: Implications for Provenance Studies. American Antiquity 63:63-72. Cook, Fred C. 1966 The 1966 Excavation at The Lewis Cr eek Site. Report on file, Department of Anthropology, University of Georgia, Athens. 1979 Kelvin: A Late Woodland Phase on the Southern Georgia Coast. Early Georgia 7(2):6586. 1995 An Archaeological Survey of Huma n Remains Found at Sidon Plantation. Manuscript on file, South Georgia College, Douglas. Cordell, Ann S. 2006 Finish What You Start: Revisiting the McKeithen Site Pottery Analysis. Paper presented at the 63rd Annual Meeting of the Southeas tern Archaeological Conference Little Rock, Arkansas. 2008 Explanation of Methods: Petrographi c Analysis of 67 NAA Pottery/Clay Samples. Report on f ile, Florida Museum of Natural History 337

PAGE 338

Cordell, Ann S., and Steven H. Koski 2003 Analysis of a S piculate Clay from Lake Monroe, Volusia County, Florida. The Florida Anthropologist 56:113-125. Crocker, Mark D. 1999 Geochemical Mapping in Georgia, US A: A Tool for Environmental Studies, Geologic Mapping and Mineral Exploration. Journal of Geochemical Exploration 67:345-360. Crown, Patricia L. 2007 Life Histories of Pots and Potters: Situating the Individual in Archaeology. American Antiquity 72:677-690. Davis, Richard A., Jr. 1997 Geology of the Florida Coast. In The Geology of Florida edited by A.F. Randazzo and D.S. Jones, pp. 155168. University Press of Florida, Gainesville. Davis, Richard A., and Miles O. Hayes 1984 What is a wave-dominated coast? Marine Geology 60:313-29. Deal, Michael, and Melissa B. Hagstrum 1994 Ceramic Reuse Behavior among the Maya and Wanka: Implications for Archaeology. In Expanding Archaeology edited by J. M. Skibo, W. H. Walker, and A. E. Neilsen, pp. 111-125. Univer sity of Utah Press, Salt Lake City. DeBoer, Warren R., and Donald Lathrap 1979 The Making and Breaking of Shipibo-Conibo Ceramics. In Ethnoarchaeology: Implications of Ethnography for Archaeology edited by C. Kramer, pp. 102-138. Columbia University Press, New York. deFrance, Susan D. 1993 Faunal Material From the Greenfiel d Site (8DU5543), Duval County Florida. Report on File, Florida Museum of Natural History, Environmental Archaeology Laboratory. Deleuze, Gilles, and Flix Guattari 1987 A Thousand Plateaus : Capitalism and Schizophrenia translated by Brian Massumi. University of Minnesota Press, Minneapolis. DeMarrais, Elizabeth, Chris Gosden, and Colin Renfrew 2004 Introduction. In Rethinking Materiality: The Engageme nt of Mind with the Material World edited by Elizabeth DeMarrais, Chris Gosden, and Colin Renfrew, pp. 1-7. McDonald Instit ute for Archaeological Research, Cambridge. DePratter, Chester 1979 Ceramics. In The Anthropology of St. Catherines Island: The Refuge-Deptford Mortuary Complex edited by David H. Thomas and Clark S. Larsen, pp. 109-32. Anthropological Papers of th e American Museum of Natural History 56(1), New York. 338

PAGE 339

Derrida, Jacques 1982 Signature, Event, Context. In The Margins of Philosophy edited by J. Derrida, pp. 307. Harvester P ress, Brighton. Dickinson, Martin F., and Lucy B. Wayne 1999 Island in the Marsh: An Archaeolo gical Investigation of 8NA59 and 8NA709, The Crane Island Sites, Nassau County, Florida. Report on file Division of Historical Resources, Tallahassee. Dobres, Marcia-Anne 2000 Technology and Social Agency: Outlini ng a Practice Framework for Archaeology Blackwell, Oxford. Driver, Harold E., and William C. Massey 1957 Comparative Studies of North American Indians, Volume 47, Number 2. Transactions of the American Philosophical Society, New Series, Philadelphia. Duff, Andrew I. 2002 Western Pueblo Identities : Regional Interaction, Migr ation, and Transformation. University of Arizona Press, Tuscon. Dye, David H. 1996 Riverine Adap tation in the Midsouth. In Of Caves and Shell Mounds, edited by K.C. Carstens and P. J. Wats on, pp. 140-158. University of Alabama Press, Tuscaloosa. Earle, Timothy A. 1994 Positioning Exchange in the Evolution of Human Socieity. In Prehistoric Exchange Systems in North America edited by T. Baugh and J. Ericson, pp. 419-437. Plenum, New York. Elliott, Daniel T. 1998 The Northern and Eastern Expressi on of Swift Creek Culture: Settlement in the Tennessee and Savannah River Valleys. In A World Engraved: Archaeology of the Swift Creek Culture edited by M. Williams and D. E lliott, pp. 19-35. University of Alabama Press, Tuscaloosa. Espenshade, Christopher T. 1983 Ceramic Ecology and Aboriginal Household Pottery Production at the Gauthier Site, Florida. Unpublished M.A. Thesis, University of Fl orida, Gainesville. 1985 Ceramic Technology of the Kings Bay Locality. In Aboriginal Subsistence and Settlement Archaeology of the Kings Bay Locality. Volume I: The Kings Bay and Devils Walkingstick Sites edited by W.H. Adams, pp. 295-336. University of Florida, Department of Anthropology, Reports of Investigations 1. 339

PAGE 340

Espenshade, Christopher T., Linda Kennedy, and Bobby G. Southerlin 1994 What is a Shell Midden? Data Recovery Excavations of Thoms Creek and Deptford Shell M iddens, 38BU2, Spring Island, South Carolina. Submitted to Spring Island Plantation by Brockington and Asso ciates, Atlanta. Fairbanks, Charles H. 1952 Creek and Pre-Creek. In Archaeology of Eastern United States edited by J.B. Griffin, pp. 285-300. University of Chicago Press, Chicago. Farnsworth, Kenneth B. and Karen A. Atwell n.d. The Blue Island and Naples-Russe ll Mounds and the Origin s of Hopewell in the Illinois River Valley Manuscript in production fo r the Illinois Transportation Archaeological Research Program. Fie, Shannon M. 2000 An Integrative Study of Ceramic Exchange During the Illinois Va lley Middle Woodland Period. Unpublished doctoral dissertation, Stat e University of New York, Buffalo. 2006 Visiting in the Inte raction Sphere: Ceramic Exchange and Interaction in the Lower Illinois Valley. In Recreating Hopewell edited by D.K. Charles and J.E. Buikstra, pp. 427-445. Univ ersity Press of Florida, Gainesville. Florida Archeological Services, Inc. (FAS) 1994 Phase II Archeological Investigations at Sites 8DU5541, 8DU5542, and 8DU 5543 at the Queens Harbour Yacht and C ountry Club, Duval County, Florida. Report on file, Division of Historic Resources, Tallahassee. Ford, James A. 1954 Spaulding's Review of Ford. American Anthropologist 56:109-112. Fowler, Chris 2004 The Archaeology of Personhood: an Anthropological Approach. Routledge, London. Fradkin, Arlene 1998 Appendix B: Quantification of 8DU5545 Faunal Assemblage. In A Phase II Archeological Investigatio n of Florida Inland Naviga tion District Tract DU-7, Greenfield Peninsula, Duval County, Florida by R.E. Johnson, B.A. Basinet, D.N. Dickel, and A. Fradkin, pp. 138156. Report on file, Division of Historical Resources, Tallahassee. Fradkin, Arlene and Jerald T. Milanich 1977 Salvage Excavations at the Law School Mound, Alachua County, Florida. Florida Anthropologist 30:166-178. Frazer, James George 1922 The Golden Bough Macmillan, New York. 340

PAGE 341

Gell, Alfred 1998 Art and Agency: An Anthropological Theory. Clarendon Press, Oxford. Gibson, Jon L. 1996 Poverty Point and Greater Southeastern Prehistory: The Culture That Did Not Fit. In Archaeology of the Mid-H olocene Southest, edited by K.E. Sassaman and D. G. Anderson, pp. 288-305. University Press of Florida, Gainesville. Gillespie, Susan D. 2001 Personhood, agency, and mortuary ri tual: A case study from the ancient Maya. Journal of Anthropological Archaeology 20:73-112. Gilman, Antonio 1981 The Development of Social Stratification in Bronze Age Europe. Current Anthropology 22:1-23. Glascock, Michael D. 1992 Characterization of Archaeological Ceramics at MU RR by Neutron Acti vation Analysis and Multivariate Statistics. In Chemical Characterization of Ceramic Pastes in Archaeology edited by H. Neff, pp. 11. Preh istory Press, Madison, WI. Glascock, Michael D. (editor) 2002 Geochemical Evidence for Long-Distan ce Exchange. Bergin and Garvey, Westport, Connecticut. Goggin, John W. 1949 Cultural Traditions in Florida Prehistory. In The Florida Indian and His Neighbors edited by J. W. Griffin, 13-44. Rollins College, Winter Park, FL. 1952 Space and Time Perspective in Northern St. Johns Archaeology, Florida Yale University Publications in Anthropology 47, New Haven. Gosden, Chris 2005 What Do Objects Want? Journal of Archaeologi cal Method and Theory 12:193-211. Gosden, Chris and Yvonne Marshall 1999 The Cultural Biography of Objects. World Archaeology 31:169-178. Gosselain, Olivier P. 2000 Materializing identi ties: an African perspective. Journal of Archaeological Method and Theory 7: 187-217. Gregory, Chris 1982 Gifts and Commodities. Academic Press, London. Griffin, James B. 1965 Hopewell and the Dark Black Glass. Michigan Archaeologist 11:115-155. 341

PAGE 342

Gude man, Stephen 2001 The Anthropology of Economy Blackwell, Malden, Massachusetts. Hackney, C. T., W. D. Burbanck, and O. P. Hackney 1976 Biological and Physical Dynamics of a Georgia Tidal Creek. Chesapeake Science 17:271-280. Hally, David J. 1983 Use Alteration of Pottery Surfaces : An Important Source of Evidence for the Identification of Vessel Function. North American Archaeologist 4:3-26. 1986 The Identification of Vessel Func tion: A Case Study from Northwest Georgia. American Antiquity 51:267-295. Handley, Brent M., Steve Ferrell, and Neill Wallis 2004 Addendum To: Data Recovery at 8NA32, The McArthur Estates Tract, Nassau County, Florida. Report on file, Division of Historical Resources, Tallahassee. Harbottle, Garman 1976 Activation analysis in archaeology. Radiochemistry 3:33. The Chemical Society, London. Hardin, Kenneth, and Michael Russo 1987 Phase II Test Excavations at the Piney Point site, 8NA31, Amelia Island, Florida. Report on file, Di vision of Historical Resources, Tallahassee. Hardin, Kenneth, John Ballo, and Michael Russo 1988 Cultural Reseource Assessment Survey of the Crossings at Fleming Island, DRI, Clay County, Florida. Report on file, Division of Historic al Resources, Tallahassee. Hart, Keith 2007 Marcel Mauss: In Pursuit of the Whole. Comparative Studies of Society and History 49:1-13. Hayden, Brian 1995 The Emergence of Prestige Technologies and Pottery. In The Emergence of Pottery Production: Technology and Innovation in Ancient Societies edited by W. Barnett and J. Hoopes, pp. 257-65. Smithsonian Institution Press, Washington, D.C. 1998 Practical and Prestige Technol ogies: The Evolution of Material Systems. Journal of Archaeological Method and Theory 5:1-55. Hegmon, Michelle 1992 Archaeological Research on Style. Annual Review of Anthropology 21: 517-536. Helms, Mary W. 1988 Ulysses Sail: An Ethnographic Odyssey of Power, Knowledge, and Geographical Distance. Princton University Press, Princeton, N.J. 342

PAGE 343

Hendon, Julia A. 2000 Having and Holding: S torage, Memory, Knowledge, and Social Relations. American Anthropologist 102(1): 42-53. Hendryx, Greg S. 2004 The Honey Dripper Site (8NA910): A Late Swift Creek Encampment in Northeastern Florida. The Florida Anthropologist 57:299-310. Hendryx, Greg S. and Neill J. Wallis 2007 The Woodland Period in Northeastern Fl orida: A View from the Tillie Fowler Site. The Florida Anthropologist 60:179-200. Hill, James N. 1970 Broken K Pueblo: Prehistoric Social Organization in the American Southwest. University of Arizona Anthropologogical Pape rs 18. University of Arizona Press, Tucson. Houston, Stephen D. and David Stuart 1998 The Ancient Maya Self: Personhood and Portraiture in th e Classic period. RES 33: 73. Husserl, Edmund 1962 Ideas: General Introduction to Pure Phenomenology Collier Books, New York. 1964 The Phenomenology of Inter nal Time and Consciousness. Indiana University Press, Bloomington. Jefferies, Richard W. 1994 The Swift Creek Site and Woodland Platform Mounds in the Southeastern United States. In Ocmulgee Archaeology 1936-1986 edited by David J. Hally, pp. 71-83. University of Georgia Press, Athens. 1996 The Emergence of Long-Distance Excha nge Networks in the Southeastern United States. In Archaeology of the Mi d-Holocene Southeast, edited by Kenneth Sassaman and David Anderson, pp. 222-34. University Press of Florida, Gainesville. Jennings, Jesse D., and Charles H. Fairbanks 1939 Pottery Type Description for Swift Creek Complicated Stamped. Southeastern Archaeological Conference Newsletter 1(2). Johnson, Robert E. 1988 An Archeological Reconnaissance Survey of the St. Johns Bluff Area of Duval County, Florida. Report on file Division of Historical Resources, Tallahassee. 1994 An Archeological and Historical Su rvey of the Wood-Hopkins Tract on Dames Point, Duval County, Florida. Report on file, Division of Historic Resources, Tallahassee. 1998 A Phase II Archeological Investigatio n of Florida Inland Navi gation District Tract DU7, Greenfield Peninsula, Duva l County, Florida. Report on file, Division of Historic Resources, Tallahassee. 343

PAGE 344

Johnson, Robert E., Myles C. P. Bland, B. Alan Basinet, and Bob Richter 1997 An Archeological Inv estigation of the Ocean Reach Site (8NA782), Nassau County, Florida. Report on f ile, Division of Historical Resources, Tallahassee. Jones, Andrew 2001 Drawn from Memory: The Aestheti cs of Archaeology and the Archaeology of Aesthetics in the Earlier Bronze Age and the Present. World Archaeology 33:334-356. 2005 Lives in Fragments? Personhood and the European Neolithic. Journal of Social Archaeology 5:193-224. Jones, Bruce A. 1989 Use-Wear Anlaysis of White Mount ain Redwares at Grasshopper Pueblo, Arizona. The Kiva 54(4):353-361. Joyce, Rosemary 1998 Performing the Body in Prehispanic Central America. RES: Anthropology and Aesthetics 33: 147. 2000a Girling the Girl and Boying th e Boy: The Production of Adulthood in Ancient Mesoamerica. World Archaeology 31: 473. 2000b Heirlooms and Houses: Mate riality and Social Memory. In Beyond Kinship: Social and Material Reproduction in House Societies, edited by Rosemary Joyce and Susan Gillespie, pp. 189-212. University of Pennsylvania Press, Philadelphia. 2001 Burying the Dead at Tlatilco: Social Memory and Social Identities. In Social Memory, Identity, and Death: Anthropological Perspecti ves on Mortuary Rituals, edited by M.S. Chesson pp. 12-26. Arch aeological Papers of the Am erican Anthropological Association 10, Arlington, VA. 2003 Concrete Memories: Fragments of the Past in the Classic Maya Present (500 1000AD). In Archaeologies of Memory edited by R.M. van Dyke and S.E. Alcock, pp. 104. Oxford: Blackwell. Kan, Sergei 1989 Symbolic Immortality: The Tlingit Potlatch of the Nineteenth Century Smithsonian Institution Press, Washington. Kellar, James H. 1979 The Mann Site and Hopewell in the Lower Wabash-Ohio Valley. In Hopewell Archaeology: The Chillicothe Conference edited by David Brose and N. Greber, pp. 100-107. Kent State University Press, Kent, Ohio. Keller, Cynthia, and Christopher Carr 2006 Gender, Role, Prestige, and Ritual Interaction across the Ohio, Mann, and Havana Hopewellian Regions, as Evidenced by Ceramic Figurines. In Gathering Hopewell: Society, Ritual, and Ritual Interaction edited by C. Carr and D.T. Case, pp. 428-460. Springer, New York. 344

PAGE 345

Kelly, Arthur 1938 A Preliminary Report on A rchaeological Explorations at Macon Georgia. Smithsonian Institution, Bureau of American Ethnology Bulletin 119. Washington D.C. Kelly, Arthur R., and Betty A. Smith 1975 The Swift Creek Site, 8-Bi-3, Macon, Georgia. Report submitted to the Southeast Archeological Center, National Park Service, Atlanta. Kelly, Richard L. 1995 The Foraging Spectrum: Diversity in Hunter-Gatherer Lifeways. Smithsonian Institution Press, Washington D.C. Kirkland, S. Dwight. 2003 Human Prehistory at the Sadl ers Landing Site, Camden County, Georgia. Early Georgia 31:107-192. Kirkland, S. Dwight and Robert E. Johnson 2000 Archeological Data Recovery at Gr eenfield Site No. 5, 8DU5541. Report on file, Division of Historical Resources, Tallahassee. Knight, Vernon James, Jr. 2001 Feasting and the Emergence of Pl atform Mound Ceremonialism I Eastern North America. In Feasts: Archaeological and Ethnographic Perspectives on Food, Politics, and Power, edited by M. Dietler and B. Hayden, pp. 311-333. Smithsonian Institution Press, Washington. Knight, Vernon J. and Timothy S. Mistovich 1984 Walter F. George Lake: Archaeologi cal Survey of Fee Owned Lands, Alabama and Georgia University of Alabama Office of Archaeological Research, Report of Investigations 42. Tuscaloosa. LaFond, Arthur A. 1983 The Queen Mound, Jacksonville, Florid a. Manuscript on file, Florida Museum of Natural History, Gainesville. Latour, Bruno 1999 Pandoras Hope: Essays on th e Reality of Science Studies Harvard University Press, Cambridge. Leach, Edmond 1976 Culture and Communication: The Logi c by which Symbols are Connected. Cambridge University Press, Cambridge. Lechtman, Heather 1977 Style in Technology: Some Early Thoughts. In Material Culture: Styles, Organization, and Dynamics of Technology edited by H. Lechtman and R.S. Merrill, pp. 3-20. American Ethnological Society, St. Paul. 345

PAGE 346

Lechtm an, Heather and Robert S. Merrill 1977 Material Culture, Styles, Organizatio n, and Dynamics of Technology. 1975 Proceedings of the American Ethnological Society. West, St. Paul, Minnesota. Leese, Morven N. and Peter L. Main 1994 The Efficient Computation of Unbiased Mahalanobis Distances a nd Their Interpretation in Archaeometry. Archaeometry 36:307. Lemonnier, Pierre 1986 The Study of Material Culture Today: Toward an Anthropology of Technical Systems. Journal of Anthropological Archaeology 5:147-186. 1992 Elements of an Anthropology of Technology Anthropological Papers of the Museum of Anthropology, vol. 88. Univ ersity of Michigan, Ann Arbor. Leroi-Gourhan, Andr 1943 LHomme et la Matire Albin Michelle, Paris. 1945 Milieu et Technique Albin Michelle, Paris. Levi-Strauss, Claude 1963 Structural Anthropology translated by C. Jacobson and B.C. Schoepf. Basic Books, New York. 1969 The Elementary Structures of Kinship, revised edition, translat ed by James Harle Bell, John Richard von Stur mer, and Rodney Needham. Beacon Press, Boston. Lindauer, Owen, and John H. Blitz 1997 Higher Ground: The Archaeology of North American Platform Mounds. Journal of Archaeological Research 5:169-207. Linton, Ralph 1944 North American Cooking Pots. American Antiquity 9:369-380. LiPuma, Edward 1998 Modernity and Forms of Personhood in Melanesia In Bodies and persons: Comparative perspectives from Africa and Melanesia edited by M. Lambek and A. Strathern, pp. 53-79. Cambridge University Press, Cambridge. Longacre, William A. 1970 Archaeology as Anthropology: A Case Study University of Arizona Anthropological Papers 17. University of Arizona Press, Tucson. MacKenzie, Maure 1991 Androgynous Objects: Strong Bags and Gender in Central New Guinea Harwood Academic Press, Melbourne. 346

PAGE 347

Mainfort, R obert C., Jr. 1988 Middle Woodland Ceremonialism at Pinson Mounds, Tennessee. American Antiquity 53:158-173. Mainfort, Robert C., Jr., James W. Cogswe ll, Michael J. OBrien, Hector Neff, and Michael D. Glascock 1997 Neutron Activation Analysis of Pottery from Pinson Mounds and Nearby Sites in Western Tennessee: Local Production vs. LongDistance Importation. Midcontinental Journal of Archaeology 22:43-68. Malinowski, Bronislaw 1922 Argonauts of the Western Pacific Routledge, London. Mauss, Marcel 1925 The Gift. W.W. Norton, New York. 1935 Les techniques du corps. Journal de Psychologie 32:271-93. 1972 A General Theory of Magic translated by R. Brain. Routledge, London. Meskell, Lynn 2004 Object Worlds in Ancient Egypt: Material Biographies Past and Present Berg, New York. Meskell, Lynn (editor) 2005 Archaeologies of Materiality Blackwell, Oxford. Meskell, Lynn and Rosemary Joyce 2003 Embodied Lives: Figuring Ancien t Maya and Egyptian Experience London: Routledge. Milanich, Jerald T. 1971 The Deptford Phase: An Archeological Reconstruction. Ph.D. di ssertation, Department of Anthropol ogy, University of Florida, Gainesville. 1973 The Southeastern Deptford Culture: A Preliminary Definition. Florida Bureau of Historic Sites and Properties, Bulletin 3:51-63. 1994 Archaeology of Precolumbian Florida University of Florida Press, Gainesville. 1999 Review of A World Engraved: Archaeology of the Swift Creek Culture edited by M. Williams and D. Elliott. American Antiquity 64:704-705. 2002 Weeden Island Cultures. In The Woodland Southeast, edited by D. G. Anderson and R. C. Mainfort, Jr., pp. 352-72. University of Alabama Press, Tuscaloosa. 2004 Archaeology of the Lower Atlantic Coastal Region, ca. 600 B.C. to Contact. In H andbook of the North American Indians, Volume 14: Southeast edited by Raymond D. Fogelson, pp. 219. Milanich, Jerald T., Ann S. Cordell, Vernon J. Knight, Jr., Timothy A. Kohler, and Brenda J. Sigler-Lavelle 1984 McKeithen Weeden Island: The Culture of Northern Florida. A.D. 200-900. Academic Press, Orlando. 347

PAGE 348

1997 Archaeology of Northern F lorida, A.D. 200-900: The McKeithen Weeden Island Culture. University Press of Florida, Gainesville. Milanich, Jerald T., and Charles H. Fairbanks 1980 Florida Archaeology. Academic Press, New York. Miller, Daniel 1987 Material Culture and Mass Consumption Blackwell, Oxford. 1995a Why Some Thi ngs Matter. In Material Cultures: Why Some Things Matter edited by Daniel Miller, pp. 3-21. The University of Chicago Press, Chicago. 1995b Coca-Cola: A Black Swwet Drink From Trinidad. In Material Cultures: Why Some Things Matter edited by D. Miller, pp. 169-188. 2001 Alienable Gifts and Inalienable Commodities. In The Empire of Things edited by Fred R. Myers, pp. 91-118. 2005 Materiality: An Introduction. In Materiality edited by Daniel Miller, pp. 1-50. Duke University Press, Durham, NC. Mills, Barbara J. 1986 "North American Cooking Pots" Reconsider ed: Some Behavioral Correlates of Variation in Cooking Pot Morphology. Paper presented at the Annual Meeting of the Society for American Archaeology, Denver. Moore, Clarence B. 1894 Certain Sand Mounds of the St Johns River, Florida, Parts I and II. Journal of the Academy of Natural Sciences of Ph iladelphia, Second Series 10:129-246. 1895 Certain Sand Mounds of Duval County, Florida. Journal of the Academy of Natural Sciences of Philadelphia, Second Series 10:449-502. 1896 Certain Florida Coast Mounds North of the St. Johns River. In Additional Mounds of Duval and Clay Counties, Florida pp. 22-30. Privately Printed. 1902 Certain Aboriginal Remains of the Northwest Florida Coast, Part II. Journal of the Academy of Natural Sciences of Philadelphia 12:127-358. Mosko, Mark 1992 Motherless Sons: Div ine Kings and Partible Persons in Melanesia and Polynesia. Man 27:697-717. Munn, Nancy D. 1983 Gawan Kula: Spatiotemporal Cont rol and the Symbolism of Influence. In The Kula: New Perspectives on Massim Exchange edited by E. Leach and J. Leach, pp. 277-308. Cambridge University Press, Cambridge. 1986 The Fame of Gawa: A Symbolic Study of Value Transformation in a Massim (PNG) Society. Cambridge University Press, Cambridge. 1990 Constructing Regional Worlds in Expe rience, Kula Exchange, Witchcraft and Gawan Local Events. Man 25:1-17. Myers, Fred R., and Barbara Kirshenblatt-Gimblett 348

PAGE 349

2001 Art and Material Culture: A Conversation with Annette W einer. In The Empire of Things edited by Fred R. Myers, pp. 269-314. School of American Research Press, Sante Fe. Nassaney, Michael S., and Kenneth E. Sassaman 1995 Introduction: Understanding Native American Interactions. In Native American Interact ions: Multiscalar Analyses and Interpretations in the Eastern Woodlands, edited by Michael Nassaney and Kenneth Sassaman, pp. xix-xxxviii. University of Tennessee Press, Knoxville. Neff, Hector 1992 Introduction. In Chemical Characterization of Ceramic Pastes in Archaeology edited by H. Neff, pp. 1. Prehistory Press, Madison, WI. 1994 RQ-mode Principal Components Analysis of Ceramic Compositional Data. Archaeometry 36:115. 2000 Neutron Activation Analysis for Provenance Determination in Archaeology. In Modern Analytical Methods in Art and Archaeology edited by E. Ciliberto and G. Spoto, pp. 81. John Wiley and Sons, Inc., New York. 2002 Quantitative Techniques for Analyzing Ceramic Compositional Data. In Ceramic Source Determination in the Greater Southwest edited by D. M. Glowacki and H. Neff. Monograph 44, Cots en Institute of Archaeo logy, UCLA, Los Angeles. Neff, Hector, Ronald L. Bishop, and Edward V. Sayre 1988 A Simulation Approach to the Problem of Tempering in Compositional Studies of Archaeological Ceramics. Journal of Archaeological Science 15:159-172. 1989 More Observations on the Problem of Tempering in Compositional Studies in Compositiona l Studies of Archaeological Ceramics. Journal of Archaeological Science 16:57-69. Neff, Hector, Jeffrey Blomst er, Michael D. Glascock, Ronald L. Bishop, M. James Blackman, Michael D. Coe, George L. Cowgill, Ann Cyphers, Richard A. Diehl, Stephen Houston, Arthur A. Joyce, Carl P. Lipo, and Marcus Winter 2006 Smokescreens in the Provenance Inve stigation of Early Formative Mesoamerican Ceramics. Latin American Antiquity 17:104-118. Neff, Hector, and Frederick J. Bove 1999 Mapping Ceramic Compositional Varia tion and Prehistoric Interaction in Pacific Coastal Guatemala. Journal of Archaeological Science 26:1037-1051. Neff, Hector and Michael D. Glascock 2001 Instrumental Neutron Activation Analysis of Prehistoric Pottery From Northern Florida and Southern Georgia, Part 1. Report on file, Research Reac tor Center, University of Missouri, Columbia. Parry, Jonathan 1986 The Gift, the Indian Gift and the Indian Gift. Man 21:453-73. 349

PAGE 350

Pauketat, T imothy R. and Susan M. Alt 2005 Agency in a Postmold? Physicality and the Archaeology of Culture-Making. Journal of Archaeological Method and Theory 12(3):213-236. Peregrine, Peter 1991 Prehistoric Chiefdoms on the American Midcontinent: A World-System Based on Prestige Goods. In Core/Periphery Relations in Precapitalist Worlds, edited by C. Chase-Dunn and T. Hall, pp. 193-211. Westview, Boulder. Phelps, David S. 1969 Swift Creek and Santa Rosa in Northwest Florida. Notebook 1(6-9):14-24. South Carolina Institute of Arch aeology and Anthropology, Univer sity of South Carolina, Columbia. Phillips, Philip 1970 Archaeological Survey in the Lower Yazoo Basin, Mississippi, 1949-1955 Papers of the Peabody Museum of Archaeology and Ethnology. Vol. 60, Pt. 1, Harvard University, Cambridge. Pilkey, Orrin H. and Mary Edna Fraser 2002 A Celebration of the Worlds Barrier Islands. Columbia University Press, Irvington, New York. Pluckhahn, Thomas J. 2003 Kolomoki: Settlement, Ceremony, and Status in the Deep South, A.D. 350 to 750 University of Alabama Press, Tuscaloosa. 2007 Reflections of Paddle Stamped Potte ry: Symmetry Analysis of Swift Creek Paddle Designs. Southeastern Archaeology 26:1-11. Polanyi, Karl 1944 The Great Transformation Rinehart, New York. Price, George 2003 Rethinking the Lithic Assemblage from the Swift Creek Type Site (9BI3). The Profile 119:8-9. Quitmyer, Irvy R. 1985 Aboriginal Subsistence Acti vities in the Kings Bay Locality. In Aboriginal Subsistence and Settlement Archaeology of the Kings Bay Locality, Volume 2 University of Florida, Department of An thropology Reports of Investigations No. 2, Gainesville. Rein, Judith S. 1974 The Complicated Stamped Pottery of the Mann Site, Posey County, Indiana. Unpublished M.A. thesis, Indiana University. 350

PAGE 351

Reina, Rubin E., and Robert M. Hill, II. 1978 The Traditio nal Pottery of Guatemala University of Texas Press, Austin. Reitz, Elizabeth J. 1988 Evidence for Coastal Adapta tions in Georgia and South Carolina. Archaeology of Eastern United States 16:137-158. Reitz, Elizabeth J., and Irvy R. Quitmyer 1988 Faunal Remains from Two Coastal Georgia Swift Creek Sites. Southeastern Archaeology 7:95-108. Rice, Prudence M. 1980 Trace Elemental Characterization of Weeden Island Pottery : Implications for Specialized Production. Southeastern Conference Bulletin 22:29-35. 1984 Some Reflections on Change in Pottery Producing Systems. In The Many Dimensions of Pottery: Ceramics in Archaeology and Anthropology edited by S.E. van der Leeuw and A.C. Pritchard, pp. 23193. CINGULA 7. Institute for Preand Proto-history, University of Amsterdam, Amsterdam. 1987 Pottery Analysis: A Sourcebook. University of Chicago Press, Chicago. 1996 Recent Ceramic Analysis : 1. Function, Style, and Origins. Journal of Archaeological Research 4:133-163. Robb, John 2004 Extended Artefact and the Monumental Economy: a Methodology for Material Agency. In Rethinking Materiality: the Engagement of Mind with the Material World, edited by E. DeMarrais, C. Gosden, and C. Renfrew, pp. 131-40. McDonald Institute for Archaeological Research, Cambridge. Rolland, Vicki L., and Paulette Bond 2003 The Search for Spiculate Clays near Aboriginal Sites in the Lower St. Johns River Region, Florida. The Florida Anthropologist 56:91-112. Ruby, Bret J. and Christine M. Shriner 2006 Ceramic Vessel Compositions and Styles as Evidence of the Local and Nonlocal Social Affiliations of Ritu al Participants at the Mann Site, Indiana. In Gathering Hopwell: Society, Ritual, and Ritual Interaction edited by C. Carr and D.T. Case, pp. 553-574. Springer, New York. Russo, Michael 1992 Chronologies and Cultures of the St. Marys Region of Northeast Florida and Southeast Georgia. The Florida Anthropologist 45:107-126. Russo, Michael, Ann Cordell, Lee Newsom, and Robert Austin 1989 Phase III Archaeological Excavations at Edgewater Landing, Volusia County, Florida. Report on file, Division of Historical Resources, Tallahassee 351

PAGE 352

Russo, Michael, Ann S. Cordell, and Donna L. Ruhl 1993 The Timucuan Ecological and Historical Preserve P hase III Final Report National Park Service, Southeast Ar cheological Center, National Park Service, Tallahassee. Sackett, James R. 1985 Style and Ethnicity in the Kalahari: A Reply to Wiessner. American Antiquity 50:154159. 1986 Isochrestism and Style: A Clarifi cation. Journal of Anthropological Archaeology 5:266277. 1990 Style and Ethnicity in Arch aeology: The Case for Isochrestism. In The Uses of Style in Archaeology edited by M. Conkey and C. Hastorf, pp. 32-43. Cambridge University Press, Cambridge. Sahlins, Marshall 1972 Stone Age Economics Aldine, Chicago. Saitta, Dean J. 1999 Prestige, Agency, and Change in Middle-Range Societies. In Material Symbols: Culture and Economy in Prehistory edited by John Robb, pp. 135-149. Occasional Paper No. 26. Center for Archaeological Investigations, Sout hern Illinois University, Carbondale. Sassaman, Kenneth E. 1993 Early Pottery in the Southeast: Traditio n and Innovation in Cooking Technology University of Alabama Press, Tuscaloosa. 2006 People of the Shoals: Stallings Culture of the Savannah River Valley University Press of Florida, Gainesville. Saunders, Rebecca 1986 Attribute Variability in Late Swift Creek Phase Ceramics from Kings Bay, Georgia. Unpublished M.A. Thesis, Department of Anthropology, University of Florida, Gainesville. 1998 Swift Creek Phase Design Assembla ges from Two Sites on the Georgia Coast. In A World Engraved: Archaeology of the Swift Creek Culture, edited by Mark Williams and Daniel Elliott, pp. 154-180. University of Alabama Press, Tuscaloosa. Schiffer, Michael B. 1975 Archaeology as a Behavioral Science. American Anthropologist 77:836-48. 1976 Behavioral Archaeology Academic Press, New York. 1990 The Influence of Surface Treatment on Heating Effectiveness of Ceramic Vessels. Journal of Archaeological Science 17:373-381. Schiffer, Michael B., and James M. Skibo 1987 Theory and Experiment in the Study of Technological Change. Current Anthropology 28:595-622. 1997 The Explanatio n of Artifact Variability. American Antiquity 62:25-50. 352

PAGE 353

Schiffer, Michael Brian, Jam es M. Skibo, Tamara C. Boelke, Mark A. Neupert, and Meredith Aronson 1994 New Perspectives on Experimental Archaeology: Surface Treatments and Thermal Response of the Clay Cooking Pot. American Antiquity 59:197-217. Sears, William H. 1952 Ceramic Development in the South Appalachian Province. American Antiquity 18:10110. 1957 Excavations on Lower St. Johns River, Florida. Contributions of the Florida State Museum 2, Gainesville. 1958 Burial Mounds on the Gulf Coastal Plain. American Antiquity 23:274-83. 1959 Two Weeden Island Period Burial Mounds, Florida. Contributions of the Florida State Museum 5, Gainesville. 1962 The Hopewellian Affiliations of Ce rtain Sites on the Gulf Coast of Florida. American Antiquity 28:5-18. 1967 Review of Excavations at the Mayport Mound, Florida by Rex K. Wilson. American Antiquity 32:120-121. 1973 The Sacred and the Secular in Prehistoric Ceramics. In Variations in Anthropology: Essays in Honor of John McGregor, edited by D. Lathrap and J. Douglas, pp. 31-42. Illinois Archaeological Survey, Urbana. Seeman, Mark F. 1979 The Hopewell Interaction Sphere: The Evidence for Interregional Trade and Structural Complexity Prehistory Research Series 5. Indiana State Historical Society, Indianapolis. 1995 When Words are Not Enough: Hopewe ll Interregionalism and the Use of Material Symbols at the GE Mound. In Native American Interactions: Multiscalar Analyses and Interpretations in the Eastern Woodlands edited by M. S. Nassaney and K. E. Sassaman, pp. 122-143. University of Tennessee Press, Knoxville. Seip, Lisa P. 1999 Transformations of Meaning: The Life History of A Nuxalk Mask. World Archaeology 31:272-287. Senior, Louise M. 1994 The Estimation of Prehistoric Values: Cracked Pot Ideas in Archaeology. In Expanding Archaeology, edited by J. M. Skibo, W. H. Walker, and A. E. Neilsen, pp. 92-110. University of Utah Press, Salt Lake City. Shanks, Michael, and Christopher Tilley 1987 Re-Constructing Archaeology: Theory and Practice. Cambridge University Practice, Cambridge. 353

PAGE 354

Shepard, Anna O. 1968 Ceramics for the Archaeologist Six th Printing (original 1954). Carnegie Institution of Washington, Washington D.C. Shrarer, Robert J., Andrew K., Balkansky, James H. Buron, Gary M. Feinman, Kent V. Flannery, David G. Grove, Joyce Marcus, Robert G. Moyle, T. Douglas Price, Elsa M. Redmond, Robert G. Reynolds, Prudence M. Rice, Charles S. Spencer, James B. Stoltman, and Jason Yaeger. 2006 On the Logic of Archaeological Inference: Early Formative Pott ery and the Evolution of Mesoamerican Societies. Latin American Antiquity 17:90-103. Sigler-Eisenberg, Brenda, Ann Cordell, Rich ard Estabrook, Elizabeth Horvath, Lee Newsom, and Michael Russo 1985 Archaeological Site Types, Distributi on, and Preservation within the Upper St. Johns River Basin. Miscellaneou s Project and Report Series Number 27, Department of Anthropology, Florida State Museum, Gainesville. Sigaud, Lygia 2002 The Vicissitudes of The Gift Social Anthropology 10:335-58. Skibo, James M. 1992 Pottery Function: A Use-Alteration Perspective. Plenum, New York. Skibo, James M., Tamara C. Butts, and Michael Brian Schiffer 1997 Ceramic Surface Treatment and Abra sion Resistance: An Experimental Study. Journal of Archaeological Science 24:311-317. Skibo, James M., Michael B. Schiffer, and Kenneth C. Reid 1989 Organic-Tempered Pottery: An Experimental Study. American Antiquity 54:122-146. Smith, Betty A. 1975 The Relationship between Deptford and Swift Creek Ceramics as Evidenced at the Mandeville Site, 9Cla1. Southeastern Archaeological Conference Bulletin 18:195-200. 1979 The Hopewell Connecti on in Southwest Georgia. In Hopewell Archaeology: The Chillicothe Conference edited by D.S. Brose and N. Greber, pp. 181-187. Kent State University Press, Kent, Ohio. 1998 Neutron Activation Analysis of Ceramics from Madeville and Swift Creek. In A World Engraved: Archaeology of the Swift Creek Culture, edited by M. Williams and D.T. Elliott, pp. 112-29. University of Alabama Press, Tuscaloosa. Smith, Bruce D. 1986 The Archaeology of the Southeastern United States: From Dalton to de Soto. In Advances in World Archaeology (vol. 5), edited by F. Wendor f and A. Close, pp. 1-92. Academic Press, Orlando. 354

PAGE 355

Sm ith, Greg C., Brent M. Handley 2002 Addendum To: Archaeological Data Recovery and Mitigation at 8DU5544/5545, Queens Harbour Yacht an d Country Club, Duval County, Florida. Report on file, Division of Historical Resources, Tallahassee. Smith, Karen 1999 From Pots to Potters: Ceramic Manufacture at Fairchilds Landing, Seminole County, Georgia. M.A. thesis, Department of Anth ropology, University of Alabama, Tuscaloosa. Smith, Marion F., Jr. 1988 Function from Whole Vessel Shape: a Method and Application to Anasazi Black Mesa, Arizona. American Anthropologist 90:921-922. Smith, Robin L. 1978 An Archaeological Survey of Kings Bay, Camden County, Georgia. Manuscript on file, Department of Anthropology, University of Florida, Gainesville. Smith, Robin L., Chad O. Braley, Nina T. Borremans, and Elizabeth Rietz 1981 Coastal Adaptations in Southeast Georgia: Ten Archaeological Sites at Kings Bay. Report submitted to the U.S. Department of the Navy, Kings Bay, Georgia. Smith, Robin L., Bruce Council, and Rebecca Saunders 1985 Three Sites on Sandy Run: Phase II Evaluation of Sites 9CM183, 184, 185 at Kings Bay, Georgia. Report subm itted to the U.S. Department of the Navy, Kings Bay, Georgia. Snow, Frankie 1975 Swift Creek Designs and Distributions: A S outh Georgia Study. Early Georgia 3:38-59. 1977 An Archaeological Survey of the Ocmulgee Big Bend Region: A Preliminary Report. Occasional Papers from South Georgia, Number 3. South Georgia College, Douglas. 1998 Swift Creek Design Inve stigations: The Hartford Case. In A World Engraved: Archaeology of the Swift Creek Culture edited by M. Williams and D. T. Elliott, pp. 61-98. University of Alabama Press, Tuscaloosa. 2007 Swift Creek Design Catalog. Re port on file, South Georgia College, Douglas. Snow, Frankie, and Keith Stephenson 1998 Swift Creek Designs: A Tool for Monitoring Interaction. In A World Engraved: Archaeology of the Swift Creek Culture, edited by M. Williams and D. T. Elliott, pp. 99-111. University of Alabama Press, Tuscaloosa. Spaulding, Albert C. 1954 Reply (to Ford). American Anthropologist 56:112-14. 355

PAGE 356

Stark, Miriam T. 1998 Technical Choices and Social Bounda ries in Material Culture Patterning: A n Introduction. In The Archaeology of Social Boundaries edited by M.T. Stark, pp. 111. Smithsonian Institution Press, Washington, D.C. Stephenson, Keith, Judith A. Bense, and Frankie Snow 2002 Aspects of Deptford and Swift Creek on the South Atlantic and Gulf Coastal Plains. In The Woodland Southeast, edited by D. G. Anderson and R. C. Mainfort, Jr., pp. 318351. University of Alabama Press, Tuscaloosa. Steponaitis, Vincas, M. Jame s Blackman, and Hector Neff 1996 Large-scale Compositional Patterns in the Chemical Composition of Mississippian Pottery. American Antiquity 61:555. Steponaitis, Vincas, M. Jame s Blackman, and Russell Weisman 1988 Chemical and Mineralogical Characterization of Mississippian Pottery. Paper presented at the 53rd Annual Meeting of the Societ y for American Archaeology, Phoenix, Arizona. Stoltman, B. James 1989 A Quantitative Approach to the Petrographic Analysis of Ceramic Thin Sections. American Antiquity 54(1):147-160. 1991 Ceramic Petrography as a Technique for Documenting Cultural Interaction: An Example from the Upper Mississippi Valley. American Antiquity 56 (1):103-120. 2000 The Role of Petrography in the Study of Archaeological Ceramics. In Earth Sciences and Archaeology, edited by Paul Goldberg, Vance T. Holiday, and C. Reid Ferring, pp.297-326. Kluwer Academic/Plenum Publishers. Stoltman, James B., and Robert C. Mainfort, Jr. 2002 Minerals and Elements: Using Petr opgraphy to Reconsider the findings of Neutron Activation in the Com positional Analysis of Ceramics from Pinson Mounds, Tennessee. Midcontinental Journal of Archaeology 27:1-33. Stoltman, James B., Joyce Marcus, Kent V. Flannery, James H. Burton, and Robert G. Moyle 2005 Petrographic Evidence Shows that Pottery Exchange Between the Olmec and Their Neighbors Was Two Way. Proceedings of the National Academy of Sciences 102:11213-11218. Stoltman, James B., and Snow, Frankie 1998 Cultural Interacti on Within Swift Creek Society: People, Pots, and Paddles. In A World Engraved: Archaeology of the Swift Creek Culture edited by M. Williams and D. T. Elliott, pp. 130-153. University of Alabama Press, Tuscaloosa. Stoner, Wesley D., Christopher A. Pool, Hector Neff, and Michael D. Glascock 2008 Exchange of Coarse Orange Pottery in the Middle Classic Tuxt la Mountains, Southern Veracruz, Mexico. Journal of Archaeological Science 35:1412-1426. 356

PAGE 357

Strathern, Marilyn 1988 The Gender of the Gift: Problems with Women and Problems with Society in Melanesia. Be rkeley: University of California Press. Struever, Stuart 1964 The Hopewell Interaction Sphere in Ri verine-Western Great Lake s Culture History. In Hopewellian Studies, edited by Joseph Caldwell and Robe rt Hall, pp. 86-106. Scientific Papers 12, Illinois State Museum, Springfield. Struever, Stuart, and Gail L. Houart 1972 An Analysis of the Hopewell Interaction Sphere. In Social Exchange and Interaction edited by E.N. Wilmsen, pp. 47-79. Anthropological Papers of the Museum of Anthropology, Un iversity of Michigan, No. 46. Ann Arbor. Sturtevant, William C. 1979 Black Drink and Other Caffeinecontaining Beverages among Non-Indians. In Black Drink: A Native American Tea edited by Charles M. Hudson, pp. 150-165. University of Georgia Press, Athens. Sykes, Karen 2005 Arguing with Anthropology: An Introducti on to Critical Theories of The Gift Routledge, London. Taussig, Michael T. 1993 Mimesis and Alterity: A Particular History of the Senses. Routledge, London. Tesar, Louis D. 1980 The Leon County Bicentennial Survey Report: An Archaeological Survey of Selected Portions of Leon County, Florida. Miscellaneous Project Re port Series 49, Bureau of Historic Sites and Properties, Florida Division of Archives, History and Records Management, Florida Department of State, Tallahassee. Thomas, Nicholas 1991 Entangled Objects: Exchange, Material Cultu re and Colonialism in the Pacific. Harvard University Press Cambridge. Thomas, Julian 1993 The Politics of Vision and the Archaeologies of Landscape. In Landscape Politics and Perspectives, edited by B. Bender, pp. 19-48. Berg, Providence and Oxford. 1999 Economy of Substances in Earlier Neolithic Britain. In Material Symbols: Culture and Economy in Prehistory edited by John E. Robb. Occasional Paper No. 26. Center for Archaeological Investigations, Southern Illinois University, Carbondale. 357

PAGE 358

Thom as, Prentice M., Jr., and L. Janice Campbell (editors) 1993 Technical Synthesis of Cultural Resour ces Investigations at Eglin Santa Rosa, Okaloosa and Walton Counties. Report of Investigations 192, New World Research, Inc., Fort Walton Beach, Florida. Thomas, David H., and C.S. Larsen (editors) 1979 The Anthropology of St. Catherines Island 2: The Refuge-Deptford Mortuary Complex Anthropological Papers of the American Museum of Natural History 56, New York, N.Y. Thunen, Robert L. 1998 Defining Space: An Overview of the Pinson Mounds Enclosure. In Ancient Earthen Enclosures of the Eastern Woodlands edited by R. C.Mainfort, Jr and L. P. Sullivan, pp. 57-67. Univ ersity Press of Florida, Gainesville. Thunen, Robert L., and Keith H. Ashley 1995 Mortuary Behavior along the Lower St. Johns: An Overview. The Florida Anthropologist 48(1):3-12. Tilley, Christopher 1999 Metaphor and Material Culture. Blackwell, Oxford. van der Leeuw, Sander E., Dick A. Papousek, and Anick Coudart 1992 Technological Traditions and Unque stioned Assumptions: The Case of Pottery in Michoacan. Techniques et Culture 16-17:145-173. Vincentelli, Moira 2000 Women and Ceramics, Gendered Vessels Manchester University Press, Manchester. 2004 Women Potters: Transforming Traditions Rutgers University Press, New Brunswick. Waggoner, James C. 2006 A Techno-Functional Analysis of Fibe r-Tempered Pottery from the Squeaking Tree Site (9Tf5), Telfair County, Georgia. Early Georgia 34:3-16. Wallis, Neill J. 2004 Perpetuating Tradition on the Lower St Johns: Pottery Technology and Function at the Mayport Mound (8DU96). The Florida Anthropologist 57(4):271-298. 2006a The Case for Swift Creek Paddles as Totemic Symbols: Some Anthropological Considerations. The Florida Anthropologist 59:55-61. 2006b The Production of Meaning in Swift Creek Iconography. Paper presented at the 71st Annual Meeting of the Society for American Archaeology San Juan, Puerto Rico. 2007 Defining Swift Creek Interaction: Earthenwar e Variability at Ring Middens and Burial Mounds. Southeastern Archaeology 26(2):212-231. 358

PAGE 359

2008 Networks of History and Mem ory: Creating a Nexus of Social Identity in Woodland Period Mounds on the Lower St. Johns River, Florida. Journal of Social Archaeology 8(2):236-271. Wallis, Neill J., Ann S. Cordell, and Lee A. Newsom 2005 Petrographic Analysis of Charco al-Tempered Pottery from Northeastern Florida. Poster presented at the 70th annual meeting of the Society for American Archaeology Salt Lake City, Utah Walthall, John A. and Brad Koldehoff 1998 Hunter-Gatherer Inte raction and Alliance Formation: Dalton and the Cult of the Long Blade. Plains Anthropologist 43:257-273. Waring, Antonio J., Jr., and Preston Holder 1968 The Deptford Ceramic Complex. In The Waring Papers: The Collected Works of Antonio J. Waring, Jr., edited by S. Williams, pp. 135-151. Peabody Museum of Archaeology and Ethnology. Harvard University, Cambridge. Washburn, Dorothy 1995 Perception, Geometry and style. In Style, Society and Person edited by C. C. a. J. Neitzel, pp. 101-122. Plenum Press, New York. Wayne, Lucy B. 1987 Swift Creek Occupation in the Altamaha Delta. Early Georgia 15:46-65. Weigand, Phil C., Garman Harbottle, and Edward V. Sayre 1977 Turquoise Sources and Source Analysis : Mesoamerica and the Southwestern U.S.A. In Exchange Systems in Prehistory edited by T. K. Earle and J. E. Ericson, pp. 15. Academic Press, New York. Weiner, Annette B. 1976 Women of Value, Men of Renown: Ne w Perspectives in Trobriand Exchange University of Texas Press, Austin. 1992 Inalienable Possessions: The Pa radox of Keeping-While-Giving. University of California Press, Berkeley. 1994 Cultural Difference and the Density of Objects. American Ethnologist 21:391403. Welch, Paul D. 1991 Moundvilles Economy. University of Alabama Press, Tuscaloosa. Wharton, Charles 1978 The Natural Environments of Georgia Georgia Department of Natural Resources, Atlanta. 359

PAGE 360

360 Wiessner, Polly 1982 Beyond Willow Smoke and Dogs Tails: A Comment on Binfords Analysis of HunterGatherer Settlement Systems. American Antiquity 47:171-178. 1983 Styles and Boundaries in Kalahari San Projectile Points. American Antiquity 48:253276. 1985 Style or Isochrestic Variation? A Reply to Sackett. American Antiquity 50:160-166. 1989 Style and Changing Relations Between the Individual and Society. In The Meaning of Things: Material Culture and Symbolic Expression edited by Ian Hodder, pp. 56-63. Unwin Hyman, London. Willey, Gordon R. 1949 Archaeology of the Florida Gulf Coast. Smithsonian Miscellaneous Collections 113. Smith sonian Institution Washington D.C. Williams, Mark, and Daniel T. Elliott (editors) 1998 A World Engraved: Archaeology of the Swift Creek Culture University of Alabama Press, Tuscaloosa. Williams, Mark, and Jennifer Freer Harris 1998 Shrines of the Preh istoric South: Patter ning in Middle Woodla nd Mound Distribution. In A World Engraved: Archaeology of the Swift Creek Culture, edited by Mark Williams and Daniel Elliott, pp. 36-47. University of Alabama Press, Tuscaloosa. Wilson, Rex L. 1965 Excavations at the Mayport Mound, Florid a. Contributions of the Florida State Museum 13, Gainesville. Wing, Elizabeth, and L. McKean 1987 Preliminary Study of the Animal Remains Excavated from the Hontoon Island Site. The Florida Anthropologist 40:40-46. Wobst, H. Martin 1977 Stylistic Behavior and Information Exchange. In For the Director: Research Essays in Honor of James B. Griffin edited by Charles E. Clela nd, pp. 317-42. Anthropological Papers No. 61, Museum of Anthropology, University of Michigan, Ann Arbor.

PAGE 361

BIOGR APHICAL SKETCH Neill Jansen Wallis was born and raised in Atlantic Beach, Florida. He attended Duke University, receiving a B.A. in English in 2000. Rethinking his prospects as an unemployed English major after his college rock band broke up, Neill found work as a field archaeologist in 2001 and was quickly inspired to pursue graduate school at the University of Florida to learn more about what he was finding. At U.F., he received an M.A. in anthropology in 2004 and a Ph.D. in anthropology in 2009. In May, 2009, Neill married Michelle LeFebvre. They currently live in Gainesville, Florida, with their dog, Kiko. 361