university alabama press circular villages of the monongahela tradition aug 2007

Changes in village lay-outs re®ect lessons learned from earlier manifestations of a community andthe development of new or modi¤cation of extant social institutions to bettermanage incre

Circular Villages of the Monongahela Tradition

Circular Villages of the Monongahela Tradition

BER NAR D K M E ANS

THE UNIVERSITY OF ALABAMA PRESS

Tuscaloosa

Copyright © 2007

The University of Alabama Press

Tuscaloosa, Alabama 35487–0380

All rights reserved

Manufactured in the United States of America

Typeface: Minion

∞

The paper on which this book is printed meets the minimum requirements of American tional Standard for Information Sciences-Permanence of Paper for Printed Library Materials, ANSI Z39.48-1984.

Na-Library of Congress Cataloging-in-Publication Data

Means, Bernard K (Bernard Klaus), 1964–

Circular villages of the Monongahela tradition / Bernard K Means.

p cm.

Includes bibliographical references and index.

ISBN-13: 978-0-8173-1573-3 (cloth : alk paper)

E78.P4M35 2007

974.8′801—dc22

2006102622

List of Illustrations vii

Acknowledgments ix

1 Village Spatial Layouts and Social Organizations 1

2 A Review of the Late Prehistoric Monongahela Tradition and the NewChronology for Allegheny Mountains Villages 13

3 Villages, Communities, and Social Organizations 31

4 Building Models of Village Spatial and Social Organizations 40

5 Models and Hypotheses Related to Community Organization 69

6 Data Sources, Variables, and Analytical Approaches 86

7 Modeling Community Patterning from Select Village Components in theAllegheny Mountains Region 106

8 Comparative Analyses from Modeling Individual

1 Maximum extent of the Monongahela tradition 2

2 DeBry’s version of John White’s 1585 watercolor of the village

of Pomeioc, North Carolina 4

3 Villages sites in Somerset County, Pennsylvania 10

4 Select drainages in the Allegheny Mountains region 17

5 Reconstruction of a Monongahela village 20

6 Diametric model of a ring-shaped settlement 46

7 An Omaha tribal camping circle 47

8 Map of a Zulu homestead 49

9 Concentric model of a ring-shaped settlement 50

10 Village of Omarakana 51

11 Yellen’s model of a ring-shaped settlement 56

12 Portnoy’s model of a ring-shaped settlement 57

13 Map of Fort Ancient Mayo Site 58

14 Radial model of the Mayo Site 59

15 Map of Fort Ancient SunWatch site 60

16 Circumferential model of a ring-shaped settlement 62

17 Hub-and-spoke model of a ring-shaped settlement 66

18 Circular histogram of grave orientations at Gnagey 3 104

List of Illustrations

19 Circumferential graph of features at Gnagey 3-2 105

33 Site plans arranged by increasing size 147

34 Site plans arranged by age and by increasing size 148

TABLES

1 De¤nite Allegheny Mountains region village sites 25

2 Radiocarbon assays for Allegheny Mountains region village sites 27

3 Areal extent of major social spaces at Allegheny Mountains regionvillage sites 107

4 Descriptive statistics and population estimates for dwellings at AlleghenyMountains region villages 109

5 Dwelling clusters and population at select Allegheny Mountains

region villages 152

viii / Illustrations

A portion of the material presented in this work was supported by the NationalScience Foundation under Grant No (BCS-0226785) titled “Modeling Somer-set Monongahela Village Organization Within a Chronological Framework De-veloped through AMS Dating of Curated Organic Remains.” My research alsobene¤tted materially from a Pennsylvania Historical and Museum Commis-sion Scholars in Residence grant Without the generosity of artifact loans fromThe State Museum of Pennsylvania and the Carnegie Museum of Natural His-tory, this work would not have been possible The State Museum of Pennsyl-vania and the Pennsylvania State Archives made available all the original ¤eldrecords and photographs from the New Deal excavations in Somerset County,Pennsylvania, and these are relied on throughout this work Washington andLee University’s Glenn Grant Publication Fund aided publication of this book.

My wife, Laura J Galke of Washington and Lee University, provided a keyillustration in this work, helped to further clarify its contents, and patientlysupported the writing process Finally, I would like to thank the critical ap-praisal of John Hart and an anonymous reviwer of an earlier draft of thismanuscript This work is stronger for their comments, but all misinterpreta-tions or errors in this work are solely my responsibilty

Acknowledgments

Circular Villages of the Monongahela Tradition

Across time and space, a signi¤cant shift in how social groups con¤gured selves has repeatedly taken place: families abandoned their millennial-longpractice of living in small dispersed settlements to reside with other families

them-in aggregated village settlements Wills (1991:161) stressed the importance ofthis shift when he noted that “the organization of unranked social groups intovillage communities is a remarkably widespread phenomenon that bespeaks aprofound adaptive strength.” In northeastern North America (hereafter theNortheast), villages became a ubiquitous part of the social landscape duringthe Late Prehistoric period (ca a.d 900 to Contact), following the adoption ofmaize horticulture as the primary subsistence strategy (Church and Nass 2002;Hart and Means 2002; Smith 1992) Despite the widespread presence of villagesites in this region, archaeological studies at the community level remain attheir infancy

A serious limitation on our understanding of the past peoples that once habited the Late Prehistoric Northeast is that their af¤liations to historicallyknown tribes have been long lost or are ambiguous Pre-Contact inhabitants

in-of a region that encompasses large portions in-of Pennsylvania, Ohio, Indiana,and most of Kentucky and West Virginia include groups referred to as “PoorlyKnown Tribes of the Ohio Valley and Interior” (Trigger 1978:ix) Direct knowl-edge of native inhabitants in this area is limited primarily to what can be ex-tracted from the archaeological record

The upper segment of the Ohio Valley has an archaeological record knownbest from the remains of excavated village sites Village sites from this re-gion have been assigned to a taxonomic unit referred to as the “Monongahelaculture” (Figure 1) A robust understanding of village community organiza-tion has been hampered by widespread use of the overly broad Monongahelataxon, which was de¤ned within the cultural-historical paradigm in the 1930sand 1950s (Means 2003) The Monongahela taxon subsumes a considerable

1

Village Spatial Layouts and

Social Organizations

amount of variation in the material expressions of cultural practices (Hart1993; McHugh 1984; Raber et al 1989:39), which differed over time and space.The use of the Monongahela taxon has led to an overgeneralization of simi-larities and a suppression of differences within and between village sites Indi-vidual village sites are often characterized as having been created according to

a broadly de¤ned typical settlement form By ignoring or dismissing variation

between Monongahela taxon village sites, archaeologists have been unable tosuccessfully compare these sites within a developmental sequence Complex so-cial relationships too often have been reduced to a level of abstraction that haslittle explanatory value, if considered at all For reasons explored in Chapter 2,the phrase “Monongahela tradition” is preferred over Monongahela taxon orthe more commonly used “Monongahela culture.”

Although the cultural af¤liations of Monongahela tradition populations arenot de¤nitively known, the manner in which they commonly con¤gured theirvillage sites is quite evident Ongoing excavations beginning in the 1930s dem-onstrated that Monongahela tradition village sites frequently consisted of acircular or oval occupation zone that formed a band around a central openspace—a formally de¤ned plaza—devoid of most cultural features (George1974; Hart 1993; Johnson 2001; Johnson et al 1989; Mayer-Oakes 1955) That is,

their village sites were ring-shaped.

1 Maximum geographic extent of the Monongahela tradition.

The so-called typical ring-shaped Monongahela tradition village site asdescribed by regional scholars is a widespread, almost archetypal settlementform Ring-shaped settlements once existed throughout much of the EasternWoodlands, from New England and New York into Virginia and the Carolinasand throughout the Middle and Upper Ohio Valleys (Bushnell 1919; Drooker1997:48; Grif¤n 1978:559; Hart 1993; Johnson et al 1989; Mayer-Oakes 1955;Ward and Davis 1993) This basic settlement form has also been recorded else-where in North America and throughout the world, including the campingcircles once formed by some native Plains Indian groups during annual buf-falo hunts (Bushnell 1922:129; Dorsey 1884:215, 1894:523, 1897; Fletcher and LaFlesche 1911; Fraser 1968:20–21; Guidoni 1975:31–36; Lévi-Strauss 1953:528) and

in settlements located in New Guinea (Fraser 1968:31; Lévi-Strauss 1963a:136),Central Brazil (Bennett 1949:13; Fabian 1992:37; Gross 1979:329; James 1949;Lévi-Strauss 1953:528, 1963a:137; Lowie 1946a:383, 1946b:420, 1946c:482; Wüstand Barreto 1999), Puerto Rico (Siegel 1997:109), southern Africa (Kuper 1993;Yellen 1977), and in early agricultural villages in Mesoamerica and the NearEast (Flannery 1972:30–38, 2002:422)

A ring-shaped village is the subject of John White’s iconic illustration of thevillage of Pomeioc, located in what is now North Carolina (Figure 2) Pomeiocand some of its inhabitants were painted by White following his visit to thevillage in July 1585 as part of Sir Walter Ralegh’s failed attempt to establish acolony in the New World (Hulton 1984:10; Quinn 1985:68) White’s watercolor

of Pomeioc depicts villagers actively interacting within a palisaded nity where a ring of houses surrounds an open plaza Several individuals areshown performing a ceremony around a rather large ¤re located in the cen-ter of the plaza Monongahela tradition villages likely would have resembledPomeioc, except that their dwellings had circular ®oor plans The watercolors

commu-of Pomeioc and its inhabitants are commu-of unparalleled value; they breathe life into

an otherwise static past that often left few material traces (Quinn 1985:182).Impermanent aspects of material culture captured by White include architec-tural details and modes of personal dress, body adornment, and even hair styles(Hulton 1984:27–28)

Ring-shaped sites in the Eastern Woodlands are not limited to villages habited by maize agriculturalists of the recent past Middle Archaic foragers inLouisiana, for example, initiated construction of the Watson Brake moundcomplex around 3350 b.c (Saunders et al 2005:631) Two curved earthworksform an oval around an open area devoid of cultural remains The two earth-works consist of eleven mounds connected by arti¤cial ridges (Saunders et al.2005:632) The central open area is thought to have served as ritual space andthe encircling ridges had a more domestic function (Saunders et al 2005:631,665) As explored in Chapter 4, Watson Brake is structured in a similar fashion

in-to more recent ring-shaped villages of the Eastern Woodlands, where a tic zone encircled a central ritual or ceremonial space.

domes-The ring-shaped settlement form arose repeatedly and independently indifferent times and places—and among cultures of varying complexity Thephrase “ring shaped” was adopted in this work from Wüst and Barreto’s (1999)characterization of Central Brazilian village settlements, the layouts of which

2 DeBry image of John White’s 1585 watercolor of the village of Pomeioc, North

Carolina (from Lorant, The New World, 1946).

closely correspond to those of many village sites in the Eastern Woodlands Thering-shaped village has been viewed as a fairly modular settlement form (Wills1991:162), which readily enables its inhabitants to arrange material elements—notably dwellings—to accommodate a wide variety of social organizationspresent at varying scales The layout of a ring-shaped settlement is intention-ally manipulated to reinforce the local social order and often is explicitly per-ceived as a microcosm of the universe, rather than passively re®ecting its con-stituent social groups (Maybury-Lewis 1989a:11; Pearson and Richards 1994:12).The act of linking their settlement’s layout to their model of reality en-hances the stability of the local group (Fletcher 1977:64) In other words, theinhabitants of ring-shaped settlements are cognizant on some level of the ac-tive role that material elements and their arrangements play in maintaining,perpetuating, and even generating social interactions (Fletcher and La Flesche1911:138; Gross 1979:329, 337; James 1949:48; Lowie 1946a, 1946b; Means 1999a:35,2000a:44, 2001, 2002a; Pearson and Richards 1994:12) Changes in village lay-outs re®ect lessons learned from earlier manifestations of a community andthe development of new or modi¤cation of extant social institutions to bettermanage increasingly large groups of people working and living alongside oneanother (Carneiro 1967:239; Eggan 1955:495; Gumerman 1994:9).

Ring-shaped settlements attracted the attention of ethnographers as early

as the late nineteenth century, with some noting that these settlements are signed by their inhabitants according to geometric models (Dorsey 1884:215,1894:523, 1897) Especially beginning with the work of Lévi-Strauss in the 1950s,ethnographers have argued that geometric models had social and ideologicaldimensions Community-wide planning principles generated by these geo-metric models in®uence to some degree the spatial con¤guration of socialgroups—and their associated architectural elements, especially dwellings As

de-is evident from thde-is work’s cross-cultural sampling of societies, presented inChapter 4, discrete social groups of varying sizes and levels of political, so-cial, or economic integration can arrange themselves within their ring-shapedsettlements according to one of a variety of geometric models

Ethnoarchaeological and archaeological studies in the latter half of thetwentieth century argued that geometric models in®uence the distribution ofactivities within a ring-shaped settlement, usually with respect to the con¤gu-ration of architectural remains (Dunnell 1983; Yellen 1977) Depending on theirnature, activities are arranged spatially with respect either to individual dwell-ings or to the entire band of dwellings that encircle a central communal space.That geometric models in®uence the distribution of activities, architec-ture, and social spaces associated with social groups at ring-shaped settlementsshould not come as a surprise My review of these settlements indicates thatgeometric models in®uence not only the locations of social groups but also thelocations of the material aspects of a settlement’s layout, resulting in a non-

random distribution of all or some material remnants At known ring-shapedvillages, social groups are evident physically in the distributions and con¤gu-rations of their dwellings The placement of these dwellings structures inter-actions within the village, including the nature and location of activities withrespect to individual dwellings and to all dwellings, which form an occupationzone around the centrally located plaza One can expect village social organi-zations to be re®ected in the spatial arrangement of some or all material re-mains at their village sites.

The comparatively large number of completely excavated village sites fromthe Allegheny Mountains region of Pennsylvania are well suited to address-ing three broad goals central to this work First, geometric models responsiblefor the con¤guration of all or some archaeologically recovered elements ofMonongahela tradition village sites should be determinable from each site’scommunity pattern Second, the kinds of social groups once present withinthese village sites should be identi¤able at least to some degree from the na-ture and con¤guration of archaeologically recovered elements Third, objec-tively identifying variation—if any—in village community patterns and corre-sponding social organizations should enable a consideration of whether thesechanged systematically over time The third goal is directed toward potentialsocial transformations that may have occurred in village communities afterthey emerged in the Upper Ohio Valley This work is not concerned directlywith the question of why villages ¤rst emerged in this region

THEORETICAL PERSPECTIVE

In recent American archaeology, a continued focus on the household as themajor unit of study has led some researchers to lose sight of a basic fact: house-holds in village settlements intentionally chose to be part of a larger socialentity—the community Village sites such as those inhabited in the AlleghenyMountains region represented a particular type of community—the nucleatedcommunity—and differed from communities made of dispersed households.Yaeger and Canuto (2000) argued that the community as an entity in and ofitself has largely been neglected in many household archaeology studies, and

in larger-scale settlement pattern studies as well To understand communityorganization, one must examine patterning evident at the level of the entirevillage site and not simply from one segment extrapolated to characterize anentire village site All too often, there is at least an implicit assumption that anyone segment of a village site—such as one or two households—is interchange-able with other broadly similar segments The concept of the community is theappropriate scale for examining social relationships at and above the householdlevel (Adler 2002:25) I consider a theoretical framework that relies on anthro-

pological and ethnoarchaeological studies of community organization to bebest suited to interpreting the archaeologically recovered aspects of village lay-outs in terms of village social organizations.

Studies of built environments and the social and behavioral use of spaceindicate that a settlement’s layout, “society’s largest and most complex artefact”(Parkington and Mills 1991:365, quoted in Widlok 1999:399), can shape and beshaped by the various social groups present, as well as the activities that thesegroups conduct within the settlement The ring-shaped settlement is a specialcase of a settlement form that is clearly built according to models that attempt

to impose and maintain a geometric order on the layout of the settlement as itexists in a horizontal plane As touched on earlier, a cross-cultural and cross-temporal review of ring-shaped settlements and models developed to explaintheir form and shape suggests that these geometric models used to spatiallyorder major social groups within a settlement re®ect an ideal framework—aframework that frequently has a cosmological basis

Because geometric models can spatially order both social groups and ties within a ring-shaped settlement, they generate activity structures that aregeometric in their patterning Considering more broadly both cognitive andbehavioral aspects of these geometric models, one can argue that, for ring-shaped villages, community-wide planning principles exist that in®uence theorganization of, and place certain constraints on, the intravillage distribution

activi-of the architectural remains and social spaces that were created and tained by various social groups In other words, a community’s activity struc-ture is generated partly by the interaction between village social groups withthe geometric models they used to plan a site’s layout For these reasons, ring-shaped settlements should be ideal for considering how social organizationscan appear on at least some level within the spatial structure of a settlement’slayout Although writing of village settlements in general, Gillespie (2000a:7)underscored a similar point when she noted that the village is “a meaningfullyconstituted layout around which people organize behaviors.”

main-More than one geometric model can operate at any one time or at differentstages in the occupational history of a settlement, depending on how social,behavioral, and ideological factors affect the relationships that generate thelayouts of a given ring-shaped settlement For example, one ideal geometricmodel may have been responsible for the distribution of village social groupsand certain material elements, especially dwellings or dwelling clusters associ-ated with the social groups Other geometric models can then come into playthat act to in®uence the locations of various activities Disposal of a commu-nity’s more noxious refuse, for example, can take place in a ring outside of themain occupation zone, regardless of how social groups and their associateddwellings are distributed within this zone Dwellings are expected to provide

one of the strongest indicators of a settlement’s geometric models, either withrespect to their own arrangement within a settlement or in terms of the dis-tribution of other material elements relative to dwellings.

Thus, the basic premise of this study is that community organization can bemodeled from the remains of village sites; the layout of a village settlement canre®ect past social organizations There is a wide range of possible social groupsthat could be represented spatially within a ring-shaped village Social groupscould range from spatially discrete and functionally redundant households tohousehold clusters representing more formal corporate entities, such as line-ages, clans, or Lévi-Straussian “societies of houses” (Brooks and Yellen 1987;Dancey 1988; Hart 1993; Hull 1987; Lea 1995; Lévi-Strauss 1982; O’Connell 1987;Pollack and Henderson 1992) The recently formulated Lévi-Straussian “socie-ties of houses” concept describes a form of social organization where socialgroups form to maintain “property” from generation to generation, includingtheir “ownership” of part of a village circle (Gillespie 2000a)

STATEMENT OF THE PROBLEM, MODELS, AND HYPOTHESES

The social transformations that occurred after Monongahela tradition lations—and other groups in the Northeast—shifted to seasonal or year-roundco-residence in village settlements are not well understood For the Mononga-hela tradition, this failure to fully understand social transformations associatedwith living in villages stems partly from a lack of understanding of the orga-nizing principles that generated the layouts of their ring-shaped settlements.Not knowing until recently the age of occupation for many Monongahela tra-dition village sites also curtailed efforts to determine whether there was a tem-poral shift in the nature of the organizing principles that operated at thesesites To address these issues, two sets of models with archaeological correlatesand one set of hypotheses were developed for this work

popu-The ¤rst set of models considers selected Monongahela tradition villages

in terms of geometric models that potentially characterize all or some aspects

of a village’s community pattern and were formulated following a review ofethnohistorical, ethnographic, and archaeological studies of ring-shaped settle-ments The ¤rst model is concerned with whether the material remnants of avillage site form two discrete clusters, potentially indicating the former pres-ence of a dual social organization Assessment of the next four models helpsascertain the nature of radial patterning present at a ring-shaped village site,ranging from a generalized zone of features, artifacts, and dwellings around acentral plaza to functionally distinct radial bands The sixth geometric modelconsiders whether the radial band or bands of material elements present at aring-shaped village are divided into discrete segments akin to pie wedges, i.e.,

representing circumferential patterning A seventh model argues that radialand circumferential patterning in the distribution of material elements at aring-shaped village site should be considered simultaneously.

The nature and types of village social groups are illuminated through amination of village sites in terms of the second set of models Some kinds ofsocial groups that might have been present at Monongahela tradition villagesites include lineages, clans, sodalities, moieties, and Lévi-Straussian societies

ex-of houses The second set ex-of models also generates a more thorough standing of the social, behavioral, or ideological factors that acted to in®uencespeci¤c Monongahela tradition village layouts These models assist a determi-nation of whether households within a village were autonomous or more for-mally linked, whether a dual organization or sodalities were present at a site,and whether status distinctions in®uenced village spatial organizations.Finally, intercommunity comparisons are considered through implementa-tion of a set of hypotheses These hypotheses examine whether villages in-creased in size steadily through time—as some have argued—or whether therange of village sizes increased through time—as others contend Particular at-tention is directed toward a consideration of whether observed variation be-tween community patterns from individual sites re®ected directional culturalchange or individual responses to unique historical and local circumstances.Examination of the ¤nal hypothesis should enable a determination of whetherlarger village components were more recent and more structurally complexthan smaller village components

under-NATURE OF THE DATA

The majority of excavated Monongahela tradition village sites—and manyother village sites in the Eastern Woodlands for that matter—are not wellsuited to an examination of patterning at the community level because only asmall portion of their site plans were exposed horizontally These village lay-outs were usually revealed through relatively small and discontinuous excava-tion blocks Many uncertainties are generated in the modeling of communitypatterns when village layouts are extrapolated from the excavated portions ofsuch sites To facilitate this study, village components that were selected hadtheir community patterns completely, or nearly completely, exposed

Most completely exposed Monongahela tradition village components wereexcavated as part of federal relief excavations in Somerset County, Pennsylva-nia, in the 1930s and 1940 (Figure 3) (Augustine 1938a, 1938b, 1938c, 1938d, 1940;Butler 1939; Cresson 1942; Means 1998a) However, these village componentsonce lacked secure chronological data No matter how objectively the differ-ent village components can be compared, the amount of variation in village

organization that existed at any one point in time—and the degree to whichthis variation was related to changes in intra- and intercommunity relations—cannot be determined in the absence of reasonably secure chronological data;nor can comparisons be made to general trends in the Northeast during theLate Prehistoric period without these data A project funded by the NationalScience Foundation (NSF) enabled the ages of many of these village compo-nents to be determined through accelerator mass spectrometry (AMS) dating

of organic remains and ceramics with organic residues curated in museum lections (see Chapter 2 for an overview; see also Means 2005a, 2006a; Meansand Galke 2004)

col-Whether excavated in the 1930s or more recently, Monongahela traditionvillage sites investigated in this work are examined in terms of ¤ve majorgroups of variables Variable Group 1 is related to the overall nature of eachvillage component, including its areal extent and whether its layout was in®u-

3 Village sites investigated as part of federal relief excavations in Somerset County, sylvania, between 1934 and 1940.

Penn-enced by topographic and environmental factors Variable Group 2 pertains toarchitectural elements and includes dwellings and post-enclosed features usedfor storage The size of individual dwellings and their distribution within eachvillage component offer one of the strongest lines of evidence for inferring thenature of past village social groups.

Nonarchitectural features—such as hearths, graves, and pits of unknownprimary function—characterize Variable Group 3 These features provide a di-rect indication of some activities and where they occurred, and to what extentthe activity structure of Monongahela tradition village communities was in®u-enced by geometric models Variable Group 4 consists of spaces de¤ned by ar-chitectural elements, including the plaza The phrase “ring-shaped village” im-plies the presence of a plaza, which was usually the largest space within aMonongahela tradition village and was one of the more important elementsin®uencing village layouts Finally, Variable Group 5 considers material culture,with a particular emphasis on how the distributions of artifact classes are use-ful for examining whether geometric models in®uenced the distribution of dis-crete activity areas The spatial distribution of stylistic artifact classes and in-dividual stylistic attributes also potentially reveals whether certain types ofvillage social organizations, such as dual organizations, were present

CONCLUSIONS

Monongahela tradition village sites in this work are considered against ric models that, at their root, are ideal representations of reality These villagesites represent the remnants of past communities Villagers lived and died here,welcomed new members through birth or emigration, and saw others leave,perhaps to join another village or to found new settlements of their own Howwell community patterns correlate to one or more of a suite of geometricmodels is in®uenced—like any other archaeological site—by the once dynamicnature of these communities and, of course, by postdepositional processes(Ascher 1968; Schiffer 1987) It is expected that some individual Monongahelatradition community patterns deviate considerably from the ideal geometricmodels that in®uenced their layouts at their initial establishment

geomet-Nonetheless, simply describing and quantifying spatial patterning withineach village’s layout in terms of models with a geometric basis—in and ofitself—produces an enhanced understanding of community organization Spa-tial patterning is examined here to infer the presence of and interrelationshipsamong various social groups, potentially allowing villages to be distinguishedalong a rough continuum: from sites with households that appear to havebeen fairly autonomous, to those with households apparently linked infor-mally (e.g., informal economic pooling networks), to villages where house-holds appear to have more formal, institutional kinds of social groups (e.g.,

lineages, clans, moieties, or Lévi-Straussian societies of houses) This studysuggests that social changes were generated in part by attempts to impose andmaintain a geometric order on the arrangement of various elements in a vil-lage’s layout, beginning at its initial establishment These geometric modelswere balanced against the reality of living in a dynamic community, whereindividuals—and the social groups of which they were a part—sometimesmade decisions that put them at odds with one another.

This work shows that analysis of village spatial layouts can lead to a greateranthropological understanding of the past peoples who inhabited the UpperOhio Valley than the still prevalent cultural-historical emphasis on de¤ning a

“Monongahela culture.” Geometric models clearly were a major factor encing the initial spatial layouts of Monongahela village components, espe-cially dwelling clusters that indicated social divisions within a village For-mal economic, social, or political links between households are evident at allMonongahela tradition village components that had layouts uncompromised

in®u-by postdepositional or other factors These formal links appeared as ing clusters that occupied discrete segments of dwelling rings at ring-shapedMonongahela tradition villages At some villages, these corporate groups weresmall—approximately 15 to 20 residents—and probably represented extended ormultifamily groups Other village sites had multidwelling corporate groups—ranging from approximately 30 to 100 residents—that were minimally lineagesegments, clans, or perhaps a form of social organization akin to the Lévi-Straussian society of houses concept Evidence for other types of social groupsthat one might have expected at Monongahela tradition village sites was lim-ited and ambiguous Further, contrary to expectations based on a reading ofthe literature, ring-shaped Allegheny Mountains region villages that were morerecent in the local developmental sequence did not necessarily have better de-signed layouts than older village components

dwell-This focus on examining ring-shaped village spatial layouts in terms ofgeometric models can be extended to sites located throughout the EasternWoodlands—and beyond—regardless of “cultural” af¤liation Although thiswork uses Monongahela tradition village sites as the primary case studies, theorganizational principles underlying their settlements are not unique to the Up-per Ohio Valley Rather than comparing one “culture” to another, archaeolo-gists can use the tools and techniques presented in this work to objectively ana-lyze and directly compare the spatial structure of ring-shaped settlements andalso to elucidate how potential social organizations were distributed withinthese settlements according to geometric models Such an effort will success-fully challenge our ideas about social organizational change in the EasternWoodlands during the last few centuries before sustained European contact

A basic tenet of this work is that—because they were generally ring shaped—Monongahela tradition village sites are best studied through application ofgeometric models derived from a cross-cultural and cross-temporal survey ofother ring-shaped village communities The reader does not necessarily need

a detailed knowledge of the Monongahela tradition to follow the analysis andinterpretation of village community patterns and village social organizationspresented in this work To enable readers to situate this work’s ¤ndings in alarger regional context, a general culture history of the Upper Ohio Valley dur-ing the Late Prehistoric period is presented here, with additional emphasis onthe Allegheny Mountains region

All Monongahela tradition village sites analyzed in this work are from theAllegheny Mountains region These village sites were selected for three princi-pal reasons First, most had their layouts completely exposed and therefore arewell suited for evaluating the nature of geometric models underlying and in-

®uencing the layouts of these ring-shaped settlements Second, these villagesites are located in relatively close geographic proximity and have broadly simi-lar environmental and topographic settings that played a minimal role in in-

®uencing village layouts Third, all village sites in the region were thought tohave been con¤ned to the same temporal span All were previously assigned tothe Somerset Phase of the Early Monongahela period based on apparent cul-tural similarities and a perception that all were occupied between a.d 900 anda.d 1250 (George 1983a; Johnson et al 1989) Thus, the perception was that vil-lagers participated in a social network that had at least some major culturaltraditions in common Following from this, it was expected that examination

of their ring-shaped sites could show whether geometric models used to designtheir settlements drew primarily from general geometric principles or from ashared set of organizing principles endemic to a local cultural tradition.This chapter begins with a consideration of maize agriculture and villages,

because the adoption of maize agriculture is often viewed as the major catalystleading to the rise of village settlements in the Northeast and elsewhere Aheavy reliance on maize agriculture characterized the subsistence economy ofmany village communities in the Northeast during the Late Prehistoric period.The issue of how effective culture-historic taxa are for some analyses of LatePrehistoric groups is then broached, particularly relative to how this issue per-tains to the Monongahela tradition The next part of this chapter turns to thephysiography of the Allegheny Mountains region The chapter then concen-trates on the Monongahela tradition itself, including a consideration of ma-jor trends in settlement-subsistence data and possible relationships to histori-cally known native groups Finally, a revised temporal framework is developedthrough AMS dating for the Allegheny Mountains region.

MAIZE, AGRICULTURE, AND VILLAGES IN THE NORTHEAST

A few general comments on maize-oriented agricultural economies, their pact on social groups, and the potential connection to the rise of village settle-ments are warranted before further consideration of the Monongahela tradi-tion Maize agriculture ¤gures prominently in some explanations of the riseand development of village settlements in the Northeast and beyond Like else-where, archaeologists in the Northeast are exploring the multiple pathwaysthat were followed in the domestication of maize and the rise of nucleated vil-lages (Hart and Means 2002) There is a tenuous understanding throughout theNortheast regarding the timing of the ¤rst appearance of nucleated villages inlocal developmental sequences and what material and social structural changesthese villages exhibited through time (Hart and Means 2002) New techniquesand a reassessment of earlier ¤ndings have led to a reevaluation of local devel-opmental sequences and a rewriting of chronological frameworks for some re-gions Analyses of older collections, some dating to the 1930s, have been cen-tral to this effort (Hart 1999, 2000; Hart and Means 2002; Means 2001, 2002a,2002b, 2002d, 2005a, 2005b, 2006a; Means and Galke 2004)

im-During the Late Prehistoric period, maize-oriented agricultural economieswere established over a broad area of the Eastern Woodlands (Smith 1992:292)

In the archaeological record of most regions of the New World, nucleated lage settlements appeared either as maize was locally domesticated or a fewcenturies after its initial appearance Hart (2001) pointed out that the relation-ship between the domestication of maize and human populations was neces-sarily one of mutual dependence The close relationship between maize agri-culture and nucleated village sites in many areas suggested that inhabitants

vil-of those regions chose to reinforce their social and economic networks by residence

co-Adoption of maize agriculture did not always inspire or support the

aggre-gation of local populations into village settlements Maize was adopted by avariety of cultures at different levels of complexity throughout the EasternWoodlands These included some groups who either never inhabited villagesettlements or did so long after they began cultivating maize (Hart and Means2002) Bender (1978:204) asserted that commitment to agriculture is not pri-marily a question of technology but rather one of changing social relations.Therefore, it is not surprising that tribal social networks became more wide-spread in the Northeast by the beginning of the Late Prehistoric period (Carrand Maslowski 1995; Dragoo 1976; Ford 1974) Tribal social networks likely de-veloped to even out environmental unpredictability by distributing goods fromdifferent spatial loci and information about environmental conditions at thesesame loci (Braun and Plog 1982).

Studies of modern slash-and-burn agriculturalists in Mesoamerica indicatethat individuals rely on other members of their community—who are not al-ways kin—to tend their ¤elds while they construct or repair their own dwell-ings, or for help with other tasks including clearing the area for their dwellings.Individuals who draw on the labor of other community members in turn pro-vide their own labor to others when they need assistance (Sandstrom 2000:59–61) Group cooperation would be necessary to clear an area for a village settle-ment (Lowie 1949:336; Myers 1973:244) Concentrating a social group into asingle village settlement—at least seasonally—was one major strategy for en-suring that suf¤cient people were present in the right places at the proper times

to perform labor-intensive tasks associated with maize horticulture, such asclearing areas for planting (Moeller 2001:2)

The appearance of nucleated settlements is usually associated in the east with a shift to maize horticulture as the primary subsistence strategy.These settlements eased the scheduling and mobilization of labor required on

North-a locNorth-al community level to successfully implement this mNorth-aize-focused tence strategy (Church and Nass 2002; Hart and Means 2002) However, themechanisms and developmental sequences that led to the nucleation of dis-persed hamlets into villages remain unclear for many parts of the EasternWoodlands (Hart and Means 2002) Also unclear are the social transformationsthat occurred after populations shifted to at least seasonal co-residence in vil-lage settlements

subsis-CULTURE-HISTORIC TAXA AND THE LATE PREHISTORIC

PERIOD IN THE NORTHEAST

The archaeological record of large parts of the Northeast during the Late historic period is organized around a series of culture-historic taxa, many de-veloped by the 1950s These taxa include not only the Monongahela culturetaxon of the Upper Ohio Valley but also the Fort Ancient culture taxon of the

Pre-Middle Ohio Valley, the Clemson’s Island culture taxon located along the main,north, and west branches of the Susquehanna River, and the MontgomeryComplex along the Middle Potomac River Some archaeologists have attackedthe appropriateness and utility of these culture-historic taxa (Dent et al 2002;Graybill 1984; Hart 1999; Hart and Brumbach 2003; Hart et al 2005; Means2003) A critique of the Owasco cultural taxon includes a number of salientpoints that can be extended to other taxa in the Northeast Hart and Brumbach(2003:738) noted that culture-historic taxa as units of analysis “are formally,temporally, and spatially bounded units that are extensionally de¤ned so as tominimize internal variation.” The use of culture-historic taxa is consistentwith an Americanist archaeological tradition dating to the 1940s and 1950s(Hart 1999) The efforts of these earlier archaeologists were intended to providechronological and spatial control for taxa that were used for subsequent pro-cessual and evolutionary analyses (Hart 1999:20; Hart and Brumbach 2003:738).Hart and Brumbach (2003:749) emphasize that, despite what many archaeolo-gists consider today, the Owasco and other taxa are not equivalent to an eth-nographic culture or to an ethnic group The misuse of the Owasco taxon hasled archaeologists to ignore variation in the material record of the past in favor

of broad generalities with limited explanatory merit (Hart and Brumbach2003) Hart and Brumbach (2003:749–750) emphasize the dangers in equatingculture-historic taxa with direct correlates of ethnic groups, language groups,

or cultures at any time in the past The use and abuse of culture-historic taxaalso can lead archaeologists to interpret change as progressive and occurringonly at major temporal divisions within the taxa (Hart and Brumbach 2003).Similar arguments can be extended to the Monongahela, Fort Ancient, andMontgomery Complex culture taxons as well Hart (1999) and Hart and Brum-bach (2003:750) suggested abandoning culture-historic taxa and emphasizedthat cultural changes are best viewed from the perspective of a local develop-mental sequence, where differences in the variables under consideration—inthis case, village organization—are more likely to be caused by social interac-tions within a local population Some archaeologists in the Northeast now sub-stitute the word “tradition” for “culture” in an effort to avoid implying thatthese traditions necessarily equate to a single cultural entity such as a tribe (see,for example, Drooker and Cowan 2001 and Hart et al 2005) It is for similar rea-sons that the phrase “Monongahela tradition” is favored over the more value-laden “Monongahela culture” or “Monongahela taxon.”

PHYSIOGRAPHY OF THE ALLEGHENY MOUNTAINS REGION

Because all case studies considered in this work are from the Allegheny tains region, only the physiography of this region is considered here (Figure 4)

Moun-This region falls primarily within the Appalachian Plateau Province, where it

is otherwise referred to as the Allegheny Mountain Section The AppalachianPlateau Province is also separated into the Glaciated Section, Allegheny HighPlateau, and the Pittsburgh Section (Raber 1985:164) Most archaeological in-vestigations in the Allegheny Mountains region have occurred in an area coter-minous with Somerset County, Pennsylvania, especially in the vicinity of theBorough of Meyersdale (Means 2002c) The Allegheny Mountains region isdrained by the Youghiogheny and Casselman rivers, which form a portion ofthe headwaters of the Ohio River Drainage (Augustine 1938a:6; Flint 1965:14;Wall 1981:3) The Raystown Branch of the Juniata River—part of the Susque-hanna River Drainage—originates in Somerset County, as do several head-waters of the Potomac River (Flint 1965:14) Thus, this region’s geographicplacement indicates that it potentially served as a crossroads between cultural

4 Select major drainages in the Allegheny Mountains region and the locations of the Con®uence and Meyersdale areas within Somerset County, Pennsylvania.

developments that occurred in different parts of the Northeast and Middle lantic regions.

At-The Casselman River originates in western Maryland as a headwater age Few sections of developed ®oodplain are present along the western Mary-land segment of the river Shortly after the Casselman River enters Pennsylva-nia, it develops major zones of ®at, extensive ®oodplain These occur at bends

drain-in the river and at major con®uences (Wall 1981:78) The majority of villagesites discussed in this work occur on or adjacent to the extensive areas of ®ood-plain near the Boroughs of Meyersdale and Con®uence In the vicinity ofMeyersdale, landforms that contained cultural remains included ®oodplains ofthe Casselman River and tributary streams; the riverine terrace system alongthe Casselman River; small rises on the ®oodplain or riverine terraces; lowerslopes; rockshelters on middle slopes; hillside benches; and ridge tops Eleva-tions of these landforms range from 584 meters at the banks of the CasselmanRiver to 695 meters in the uplands that rise quickly from the river’s banks(Means 2006a)

MONONGAHELA TRADITION OVERVIEW

Major archaeological investigations in the Upper Ohio Valley were initiated

at the height of the Great Depression partly to learn more about the poorlyknown aboriginal inhabitants of southwestern Pennsylvania (Augustine 1938a,1938b, 1938c, 1938d, 1940; Butler 1939; Means 1998a) Some leading scholars atthe time disputed whether there had ever been a substantial aboriginal occu-pation of the region From 1934 to 1940, the Somerset County, Pennsylvania,Relief Excavations (SCRE) undertook large-scale excavations of village sites toaddress this issue, as well as to provide suf¤cient work to keep a crew of mencontinuously employed (Means 1998a, 2000b, 2002b) Butler (1937, 1939:9, 71–73) developed the ¤rst formal de¤nition of the Monongahela tradition as a pre-historic group based largely on SCRE-excavated village sites (Means 1998a:39–

44, 2002b, 2003)

Hart (1993:88) pointed out that the SCRE investigations are important to acontinuing understanding of the Monongahela tradition, because they exposedmore habitation area of more village sites than has been done since For thisreason alone, data on Monongahela tradition village sites at the center of thiswork’s analyses and interpretations were largely derived from SCRE investiga-tions The usefulness of SCRE-investigated village sites was further enhanced

by this author’s successful attempt to obtain AMS assays from several villagecomponents excavated in the 1930s, as detailed later in this chapter (see alsoMeans 2005a, 2005b, 2005c, 2006a) Chapter 6 includes an assessment of thequality of the decades-old SCRE data

Overall Monongahela Tradition Chronological Framework

Drawing on Mayer-Oakes’s (1955) pottery seriation and initial chronologybuilding efforts (assessed in Means 2003), Johnson et al (1989:24) divided theMonongahela tradition developmental sequence into a tripartite scheme: Early(a.d 1050/1100 to a.d 1250), Middle (a.d 1250 to a.d 1580/1590), and Late

or Protohistoric (a.d 1580/1590 to a.d 1630) Some researchers (Raber et al.1989) have questioned the validity of this tripartite chronological scheme andits further subdivision into phases that usually have a geographic dimension.Most Monongahela tradition periods and phases are constructed based on traitlists, particularly those traits related to ceramic attributes Raber et al (1989:39)argued that a focus on trait lists obscures substantial regional and local diver-sity Thus, where possible in this work, radiocarbon assays and not culture-historic temporal constructs are used to assign ages to individual village com-ponents

Site Structure and Community Patterns

An understanding of Monongahela tradition village sites is central to this work,

so this section considers how these sites have been described by other ers Researchers from outside the Upper Ohio Valley can see that the basic or-ganization and constituent elements of Monongahela tradition village sites aresimilar to many settlements located throughout the Eastern Woodlands (Bush-nell 1919) These sites were generally circular or oval, bounded by small posts,and contained an open central plaza until the end of the Monongahela tradi-tion sequence (Johnson et al 1989:9–12) A typical Monongahela tradition com-munity pattern is usually said to consist of an occupation zone commonly en-circled by a palisade and containing activity areas, dwellings, pits, graves, andother features surrounding an open plaza (George 1974:2; Johnson et al 1989:9).Residential dwellings may have been wigwam-style, dome-shaped structuresdocumented ethnohistorically throughout the Eastern Woodland (Nabokovand Easton 1989), although not all scholars agree with this conclusion (Fig-ure 5) (Means 2006b)

research-Monongahela tradition sites are also noted for post-enclosed features thatwere either freestanding or attached to dwellings and were presumably usedfor storage These features have long been considered unique to the Mononga-hela tradition (Cresson 1942:66) and are seen by some as an important cul-tural indicator (George 1974:2,19; Hart 1995:42; Herbstritt 1981:14) They prob-ably represented structures that extended well above the ground surface (Hart1995:42) If these post-enclosed features were storage facilities, they would haveserved as a buffering mechanism for coping with both predictable and unpre-dictable short- and long-term spatial and temporal variation in resource avail-

ability (Hart 1995:42) Hart (1995:50) interpreted most attached post-enclosedfeatures as representing greater domestic control over surplus than was possiblewith freestanding post-enclosed features By the Late Monongahela period, thepetal-shaped structure present at a few Monongahela tradition village sites mayhave indicated increased communal control over some surplus (Hart 1995:51).The open plaza area of a village was generally devoid of substantial culturalremains during most of the Monongahela tradition sequence (Johnson et al.1989) It was probably swept clean on a regular basis (Johnson 1981) However,evidence for cultural activities within the open plaza area has been recovered

at some village sites For example, George (1974:5) found evidence for frequentand intensive ¤ring within the central plaza of the Ryan site, especially at thegeographic center of the village He did not provide an explanation for thispattern, although a ceremonial function is likely, given ethnographic data dis-cussed in Chapter 4 Other sites, including some in the Allegheny Mountainsregion, had evidence for centrally located ¤res or posts (Means 2001)

Settlements in the Allegheny Mountains Region

Monongahela tradition habitation sites are known along all major drainages inthe Allegheny Mountains region, including the Casselman, Youghiogheny, andPotomac rivers In addition, Late Prehistoric nonvillage components have beendocumented at a wide variety of topographic settings along these same drain-ages These components include hamlets, limited activity procurement camps,rockshelters, and at least one chert quarry site (Means 1998b, 1998c, 2000d,2002c, 2006a; Means et al 1998; Wall 1981)

5 Artist’s reconstruction of a Monongahela village site, shown in cross-section (courtesy

of Laura J Galke).

Major Trends Evident in Monongahela Tradition Subsistence-Settlement Data

New excavations and reviews of previously excavated village and other sitesenable an ongoing, dynamic reinterpretation of the Monongahela tradition de-velopmental sequence (Hart 1993; Hart et al 2005; Johnson 2001; Means 2006a;Nass and Hart 2000; Richardson et al 2002) The introduction of maize intothe native horticultural system is often viewed as the main catalyst for the coa-lescing of scattered Late Woodland hamlets (a.d 500 to 900) into Late Prehis-toric village sites (Fuller 1980, 1981a, 1981b; Wymer 1993) In light of new dataand recent theoretical developments (Hart 2001; Hart et al 2003), it is likely

that one should more properly consider the adoption of intensive maize

agri-culture as a potential catalyst for the appearance of village settlements, ratherthan the mere introduction of maize into existing subsistence economies The

development of village sites was not simply the restructuring of dispersed

ham-lets into composite, palisaded village sites (Hart 1993) Hamlet sites continued

to exist throughout the Late Prehistoric period in the lower Upper Ohio Valleyand played an integral role in local and regional subsistence-settlement systems(Hart 1993; Means 2002c)

Major changes in settlement distribution supposedly occurred around a.d

1250, with a brief climatic decline said by some to have led Monongahela tion peoples to abandon the more environmentally marginal Allegheny Moun-tains region and parts of Ohio and West Virginia As a result, Monongahelatradition settlements were once thought to have consolidated into the “Monon-gahela heartland” along the Allegheny and Monongahela rivers (Johnson 2001).The argument that the Monongahela tradition territory as a whole was envi-ronmentally marginal during this time period was based on the short growingseason for maize within this area The Allegheny Mountains region was con-sidered by Johnson et al (1989) to have been particularly adversely impacted

tradi-by this climatic decline This region contains the most rugged and steep raphy in Pennsylvania and, consequently, an even shorter growing season thanother Monongahela tradition areas Johnson et al (1989) considered maize ag-riculture to have been a fairly risky proposition in this region even prior to theapparent climatic decline

topog-Challenges to the presumed impact of the a.d 1250 climatic decline includerecent work at the Saddle site in West Virginia A component dating to arounda.d 1400 indicated that nucleated village sites persisted in the area longer thanhad been thought (Church 1995:70) Hart’s (1993) divided-risk model suggeststhat, even if a severe climatic decline did occur around a.d 1250, the link be-tween village sites and hamlets and a continued reliance on nonmaize resourcescould have ameliorated uneven resource distributions The interpretation that

the Allegheny Mountains region was abandoned was based in part on the weakchronological framework that once existed for this area New AMS dates ob-tained by the author—and considered below—unequivocally demonstrate thatthis abandonment scenario is untenable (Means 2005a, 2006a).

Again following from Johnson et al (1989), the Late Monongahela periodwas said to have witnessed a further reduction in the geographic distribution

of Monongahela tradition village sites from the earlier a.d 1250 contraction.This second reduction supposedly resulted primarily from con®ict with north-ern Iroquoian neighbors over the fur trade Monongahela tradition peoples dis-appeared as a recognizable cultural entity shortly after a.d 1635, within ¤veyears or less of the onset of the English fur trade in the region (Johnson et al.1989:28–30) The Seneca may have dispersed Monongahela tradition popula-tions (Johnson 2001) Recent research by Richardson et al (2002:86–90) sug-gested that drought conditions may have characterized the Late Monongahelasequence Along with increased aggregation of populations into larger villagesites and further intensi¤cation of agricultural production, drought conditionsmade Monongahela tradition populations particularly susceptible to hostileincursions by Iroquoian groups

Generally, Richardson et al (2002:90) concluded that climatic factors at eral critical junctures, particularly climate deterioration, were responsible forchanges in the distribution of settlements throughout the region assigned tothe Monongahela tradition They implied a correlation between climatic shiftsand increases in settlement size, the degree of forti¤cation for village sites, andthe intensi¤cation of maize agriculture Nass and Hart (2000:124) acknowl-edged the role of environmental factors but stressed heterogeneity, not homo-geneity, in subsistence-settlement data They further emphasized that pattern-ing in these data resulted from adaptive responses to both local and regionalenvironmental and social variables Nass and Hart (2000) concluded that cli-matic deterioration models and allied warfare models do not account for spa-tial and temporal variation within and between regions These models gener-ally fail to consider social and technological buffering mechanisms, such asstorage facilities, community-wide distribution of resources, and village siteslinked to hamlet and camp sites distributed in diverse ecological settings (Nassand Hart 2000:127; see also Hart 1993, 1995, and Hart et al 2005)

sev-Relationship to Historically Known Native Groups

Despite considerable speculation, no historically known tribal groups havebeen convincingly associated with the Monongahela tradition (George 1980,1994; Johnson 2001; Johnson et al 1989:31) This situation arises in part becauselittle research has been conducted on Contact period sites in western Pennsyl-

vania or on sites associated with Native American groups displaced from gions farther east during the Historic period (Cowin 1985:193; but see Johnson2001) Even if archaeologists succeed in assigning all or part of the Mononga-hela tradition to a speci¤c linguistic or tribal group known dimly from vaguehistorical documents, simply possessing the historic identity of the Mononga-hela tradition would tell us little about how they organized themselves socially,politically, or economically earlier To accomplish this feat, archaeologists mustturn to the material remnants of the Monongahela tradition, notably their vil-lage sites They need to consider the types of social organizations that likelywere present in village sites and more precisely to de¤ne exactly what is meant

re-by the word “village”—a term that is bandied about cavalierly re-by researchers.REVISING THE CHRONOLOGY OF THE

ALLEGHENY MOUNTAINS REGION

Until recently, one signi¤cant drawback to analyzing Allegheny Mountains gion sites has been that chronological data were generally weak, suspect, ornonexistent for most of these sites In particular, the most serious limitation toincorporating into this work the village sites investigated by the SCRE was thepoor understanding of their temporal positions within local or regional chro-nologies All were excavated long before the development of radiocarbon dat-ing and its regular application to archaeological research Therefore, until re-cently, none of the SCRE-investigated village sites had associated radiocarbondates Fortunately, technological advancements now enable AMS dating of or-ganic remains and ceramics with organic residues (Carr and Haas 1996; Lovis1990) Appropriate datable archaeological remains from SCRE sites were docu-mented in museum collections (Means 1999a) An NSF project designed to ob-tain AMS dates from SCRE-investigated sites was quite successful, as detailedbelow (see also Means 2005a, 2005b, 2005c)

re-Prior to these recently obtained AMS dates, only three de¤nite village siteshad associated radiocarbon dates: Quemahoning/Alwine, Gnagey 3, and Peten-brink 1 (George 1983a, 1983b; Means 2002c) Except for two dates obtained byHart and Scarry (1999), Gnagey 3’s extant suite of dates (George 1983a) is con-sidered questionable on a number of grounds, including problematic contextsand suspected errors with a laboratory that produced the dates (Means 2005a).The assays from Quemahoning/Alwine and Petenbrink 1 exhibited few obvi-ous problems Discounting those that were problematic, extant radiocarbon as-says from Quemahoning/Alwine, Gnagey 3, and Petenbrink 1 were insuf¤cientfor examining directional variation—if any—in village community organiza-tion in the Allegheny Mountains region A solution to strengthening the weak

regional chronological framework was clear at the onset of this study: carbon dating of SCRE-excavated village components using organic remainspresent in museum collections.

radio-A review of curated archaeological collections associated with the SCRE covered the presence of potentially datable organic remains from these sites(Means 1999b) Collections at The State Museum of Pennsylvania (TSMP) andthe Carnegie Museum of Natural History (CMNH) had curated organic re-mains suitable for radiocarbon dating from several Allegheny Mountains re-gion sites Organic remains selected for dating from these repositories includedsmall amounts of carbonized plant material (such as wood charcoal, charredmaize cobs and kernels, and charred beans); organic food residue on the inte-rior surfaces of ceramic vessel fragments; unworked shell objects in the form ofcrushed shell used as temper for ceramic vessels; and a worked animal bone awl

un-A number of factors in®uenced the amount and type of curated organicremains selected and submitted for AMS dating Several village sites consisted

of two or more discrete, and sometimes overlapping, components Becausethis analysis considered Monongahela tradition villages from a developmentalperspective, efforts were expended to date each discrete village component.Samples were submitted from features that were linked as closely as possiblethrough a study of intrasite patterning to a speci¤c component (Shott 1992:206) Features that fell into an area of overlap between two components werenot selected as sample sources to avoid confusion

It was intended to submit a minimum of three samples for each discretecomponent to ensure that the dates obtained were not anomalous (Thomas1986:244–250) The SCRE uncovered 16 or 17 village components at the 12 vil-lage sites they investigated (Table 1) Prior to submission of the NSF proposal,

it was determined that four sites—representing six components—did not havepreserved organic remains at TSMP or CMNH, although some organic re-mains could be preserved in undocumented private collections Because ideallythree samples were to be submitted for each village component, a total of 33samples were planned to be submitted for AMS dating of the 10 or 11 SCRE-investigated village components from eight sites then thought to have curatedorganic remains If some sites were determined not to have suf¤cient suitablematerials for three samples, the remaining samples were to be redistributedamong the other sites to increase their total sample number Subsequent de-tailed examination of the artifact collections at TSMP and CMNH determinedthat nine or ten components from seven SCRE-investigated village sites hadmaterial clearly suitable for AMS dating Of these, four sites had insuf¤cientcurated organic remains for the proposed number of AMS samples; their pro-posed samples were redistributed to the remaining three sites In case any ofthese 37 samples from the 12 or 13 Allegheny Mountains region components

proved problematic, an additional seven samples were budgeted for AMS ing to ensure that suf¤cient dates were obtained.

dat-Submission of an additional four samples for AMS dating from the Gnagey

3 village site was planned as well Radiocarbon assays obtained from the site inthe 1970s are problematic, as discussed earlier The two assays recently obtainedfrom Gnagey 3 (Hart and Scarry 1999) were insuf¤cient to verify with certaintythe chronological placement of the two components at this village site

A basic accomplishment of the AMS dating project was that, for the ¤rsttime, nine previously undated village components from seven New Deal exca-vated village sites were directly dated (Table 2) In addition, the chronologicalplacement of Gnagey 3’s two components was clari¤ed The more recently ex-cavated village components from Gnagey 3 were shown to have been occupiedaround the middle of the fourteenth century a.d rather than in the tenth andtwelfth centuries a.d as proposed by George (1983a) The larger village com-ponent at Gnagey 3 was built shortly after the earlier and smaller village com-ponent there (Means 2001, 2003).

The AMS dating project accomplished more than ascertaining the ages ofnine previously undated village components excavated during the New Dealand clarifying the ages of two more recently excavated components from asingle site This dating project successfully altered extant perceptions of theoccupational history of the Allegheny Mountains region These newly datedcomponents and a reconsideration of other radiocarbon assays enabled threelongstanding assumptions about the chronology of the region to be success-fully challenged

First, contrary to Mayer-Oakes’s (1955) ceramic typology, most village sitesdid not fall into the ¤rst part of the overall Monongahela tradition chronologi-cal framework Clearly, sites are not necessarily early in the temporal sequence

if their ceramic assemblages are dominated by limestone-tempered rather thanshell-tempered pottery Thus, the basic assumption of Mayer-Oakes’s ceramictypology that limestone-tempered ceramics were replaced by shell-temperedwares throughout Monongahela tradition territories at the end of the EarlyMonongahela period ca a.d 1250 is no longer tenable

Second, the revised dating of Gnagey 3 showed that the ¤rst de¤nite lage settlements did not become archaeologically visible much earlier in theAllegheny Mountains region—ca a.d 900—than was seen in other parts of theNortheast Rather, the ¤rst de¤nite village settlement appeared at least twocenturies later Petenbrink 1’s ¤rst component dated to ca a.d 1100 Based oncurrent data, it appears that American Indians in the region adopted a villagelifestyle two centuries more recently than was originally proposed by George(1983a) Of course, it is conceivable that future archaeological investigationscould uncover traces of earlier village sites

vil-Third, it is now evident that the Allegheny Mountains region was not doned after a.d 1250 owing to some perceived climatic shift In fact, most ofthe dated village sites were inhabited during the ¤rst two centuries followinga.d 1250 Further, the latest known village in the region—the larger occupation

aban-at Peck 2—was inhabited approximaban-ately three centuries after the supposed riod of abandonment at ca a.d 1530

pe-The principal purpose of the AMS dating project was not simply to

chal-C