Ancient Maya Pottery Classifi

8/16/2019 Ancient Maya Pottery Classifi

1/311


2/311

Ancient Maya Pottery

Florida A&M University, Tallahassee

Florida Atlantic University, Boca Raton

Florida Gulf Coast University, Ft. Myers

Florida International University, Miami

Florida State University, Tallahassee

New College of Florida, Sarasota

University of Central Florida, OrlandoUniversity of Florida, Gainesville

University of North Florida, Jacksonville

University of South Florida, Tampa

University of West Florida, Pensacola


3/311

This page intentionally left blank


4/311

Ancient Maya Pottery

Classification, Analysis, and Interpretation

Edited by James John Aimers

Foreword by Diane Z. Chase and Arlen F. Chase

University Press of Florida

Gainesville/Tallahassee/Tampa/Boca Raton

Pensacola/Orlando/Miami/Jacksonville/Ft. Myers/Sarasota


5/311

is publication is made possible in part with support frome Research Foundation of the State University of New York.

Copyright by James John AimersAll rights reservedPrinted in the United States of America. is book is printed on Glatfelter Natures Book,a paper certied under the standards of the Forestry Stewardship Council (FSC). It is arecycled stock that contains percent post-consumer waste and is acid-free.

is book may be available in an electronic edition.

A record of cataloging-in-publication data is available from the Library of Congress.ISBN - - - -

e University Press of Florida is the scholarly publishing agency for the State UniversitySystem of Florida, comprising Florida A&M University, Florida Atlantic University,Florida Gulf Coast University, Florida International University, Florida State University,New College of Florida, University of Central Florida, University of Florida, University

of North Florida, University of South Florida, and University of West Florida. University Press of Florida

Northwest th StreetGainesville, FL -http://www.upf.com

http://www.upf.com/http://www.upf.com/


6/311

Contents

List of Figures viiList of Tables xForeword xiPreface xiii

. Introduction James John Aimers

. Type-Variety: What Works and What Doesn’t

Prudence M. Rice . Types and Traditions, Spheres and Systems: A Consideration of Analytic

Constructs and Concepts in the Classication and Interpretation ofMaya Ceramics Cassandra R. Bill

. Interpreting Form and Context: Ceramic Subcomplexes at Caracol, Nohmul, and Santa Rita Corozal, Belize

Arlen F. Chase and Diane Z. Chase . Ceramic Resemblances, Trade, and Emulation: Changing Utilitarian

Pottery Traditions in the Maya Lowlands Robert E. Fry

. Type-Variety on Trial: Experiments in Classication and Meaning Using Ceramic Assemblages from Lamanai, Belize James John Aimers and Elizabeth Graham

. Establishing the Cunil Ceramic Complex at Cahal Pech, Belize Lauren A. Sullivan and Jaime J. Awe

. Technological Style and Terminal Preclassic Orange Ceramics in theHolmul Region, Guatemala Michael G. Callaghan, Francisco Estrada-Belli, and Nina Neivens de Estrada


7/311

. Acanmul, Becán, and the Xcocom Phenomenon through a Type-Variety Looking Glass: Resolving Historical Enigmas through Hands-OnTypological Assessments

Joseph Ball and Jennifer Taschek. Looking for Times: How Type-Variety Analysis Helps Us “See” the Early Postclassic in Northwestern Honduras Patricia A. Urban, Edward M. Schortman, and Marne T. Ausec

. Slips, Styles, and Trading Patterns: A Postclassic Perspective from Central Petén, Guatemala Leslie G. Cecil

. Mayapán’s Chen Mul Modeled Effigy Censers: Iconography and Archaeological Context Susan Milbrath and Carlos Peraza Lope

. Problems and Prospects in Maya Ceramic Classication, Analysis, and Interpretation James John Aimers

References Contributors Index


8/311

Figures

. . Sovedeso Negative-painted vessels and sherds from Copán

. . Polished black/brown pottery from Copán

. . Painted and incised jars from Copán

. . Plan of Nohmul Structure showing location of recovered vessels

. . Vessels associated with Nohmul Structure

. . Plan of Santa Rita Corozal Structure showing location of recovered vessels

. . Vessels associated with Santa Rita Corozal Structure

. . Plan of Caracol Structure A

. . Vessels associated with Caracol Structure A

. . Sample of Belize Red vessels recovered from Caracol

. . Pottery classied as “censer ware” from Santa Rita Corozal

. . Stacked set of contemporary Mexican bowls

. . Map of sites in southern Quintana Roo studied by the Uaymil SurveyProject

. . Two-dimensional MDS solution of stylistic resemblances of LateClassic slipped bowls from Quintana Roo

. . Two-dimensional MDS solution of technological resemblances of LateClassic slipped bowls from Quintana Roo

. . Two-dimensional MDS solution of stylistic resemblances of LateClassic unslipped jars from Quintana Roo

. . Two-dimensional MDS solution of technological resemblances of Late

Classic unslipped jars from Quintana Roo . . Local red slate ware annular-base basin from Chau Hiix . . Pedestal-based jar and dish (“chalice”) from Lamanai . . Pozo Unslipped system jar from Lamanai and Pozo Unslipped system

jar rim from Tipu . . Cahal Pech site map


9/311

. . Cahal Pech Plaza B

. . North Wall prole, Column /

. . Baki Red-incised: Baki Variety

. . Zotz Zoned-incised: Zotz Variety . . Zotz Zoned-incised: Zotz Variety

. . Kitam Incised: Kitam Variety

. . Sikiya Unslipped: Sikiya Variety

. . Location of the Holmul region and other Terminal Preclassic sites

. . Sierra Red composite bowl form

. . Sherds of Ixcanrio Orange-polychrome composite bowl forms

. . Aguila Orange composite bowl form

. . Boleto Black-on-orange composite bowl form

. . Actuncan Orange-polychrome composite bowl form

. . Bivariate plot of Sr and Ca-corrected Cr base- logged concentrationsshowing compositional groups

. . Bivariate plot of uncorrected Ca and Sr base- logged concentrationsshowing compositional groups

. . Areal map showing sites mentioned in text and other key centers

. . Plan view of Acanmul site core . . Comparison of original and revised Becán, Acanmul, and EdznáClassic-Postclassic ceramic phase sequences

. . Diagnostic “early Xcocom” or “Puuc” lance points

. . Select Acanmul (Pa’xil)–Becán (Xcocoma) ceramic identities

. . Select Acanmul (Pa’xil)–Becán (Xcocoma) ceramic homologies . . Map of Northwestern Honduras showing areas and sites mentioned in

the text

. . Sample of lip form coding sheet

. . Sample of jar neck form coding sheet

. . Completed readout sheet

. . Graph comparing trends in paste composition, middle Chamelecon-Cacaulapa rural area and El Coyote

. . Central Petén archaeological sites

. . Trapeche ceramic group sherd thin section demonstrating two slips andSEM image of a double slip

. . Early Postclassic period pottery

. . Bivariate plots of ( ) Cr and and ( ) Sc and Nd base- loggedconcentrations showing, in turn, separation of Trapeche and Fulanoand separation of Paxcamán ceramic group slips

Figuresviii


10/311

. . Bivariate plots of Early Postclassic period Trapeche and Fulano ceramicgroup pastes and slips

. . Bivariate plots of Late Postclassic period Paxcamán ceramic group

pastes and slips . . Mayapán Chen Mul Modeled censer representing God N . . Mayapán Hoal Modeled effigy censer . . Step-eyed Maize God from Structure Q , Mayapán . . Skeletal deity from Structure Q , Mayapán . . Dresden Codex [ ], Opossum or Mam wearing an oyohualli on a

braided breastplate . . Madrid Codex , Step-eyed Maize Gods . . Location of effigy censers and other gures in Cache , Lot C in

Structure Q

Figures ix


11/311

Tables

. . Schematic of classicatory principles and characteristics of resultantunits

. . Parallel distributions of decorative types and varieties within PostclassicPetén paste wares

. . Sherds and vessels recovered from latest use of Santa Rita CorozalStructure

. . Attributes used in the analysis of pottery collected by the UaymilProject from sites in east-central Quintana Roo

. . Wares, groups, types, and varieties of the Cunil complex . . All paste variants of composite bowls

. . Diversity data for all paste variants of composite bowls

. . Major stereomicroscopic and corresponding petrographic paste variants

. . Petrographic paste variant and groundmass type

. . INAA groups and type-forms . . Numbers of paste wares, groups, types, and varieties in Middle Ulua–

Santa Barbara area, Naco Valley, and Middle Chamelecon–Caculapazone


12/311

Foreword

A decade ago, when we decided to start the Maya Studies series with the Uni- versity Press of Florida, we did so with two specic targeted books in mind. erst of these was a volume that synthesized all of the results of the various ar-chaeological projects that were then taking place in central Belize; that volumewas published by the press in ase Ancient Maya of the Belize Valley:Half a Century of Archaeological Research, edited by James Garber. e secondbook that we solicited was one that dealt with new approaches to and issues inMaya ceramic analysis. While this second book took a bit longer to come to

fruition, it was worth the wait. Ancient Maya Pottery: Classication, Analysis,and Interpretation , edited by James John Aimers, is an excellent contribution tounderstanding Maya pottery and how it is analyzed and interpreted. For a Maya archaeologist, ceramics are crucial to interpreting the archaeo-logical record. Large numbers of pottery sherds are found at all Maya sites. Tosome extent, they are both a bane and a blessing for archaeologists. ey are ablessing because they can be dated and used to interpret past societies—pro- vided that their stratigraphic and contextual situations are understood; theyare a bane because their quantity and o en poor preservation make them diffi-cult and time-consuming to analyze. Pottery is plastic and malleable, changingto reect cultural mores and preferences. In a society that did not use metalobjects until very late in their history, this meant that most containers weremade of pottery and that these containers morphed into different forms andstyles as time passed. us, ceramics are a keystone in Maya archaeology forrelative dating, for interpreting social differences within past groups at any one

point in time, and for determining contextual functions. Spectacular examples of Classic period Maya polychrome ceramics (A.D.–A.D. ) are rightly prized by the countries in which they have been

found, and many examples are professionally displayed in the world’s muse-ums. However, the vast majority of Maya pottery does not meet this artisticstandard, and most of the materials with which archaeologists work wouldnever be placed on exhibit. Yet all pottery—whether beautiful or mundane,


13/311

partial or whole—is of great signicance to archaeological interpretation, andmethodology relating to its classication and interpretation is important for anunderstanding of past civilizations.

Ceramics rst appeared sometime a er B.C. in the Maya area; exactlywhat constitutes the earliest Maya pottery is still under discussion. e earliestceramic complexes (Swasey, Cunil, Eb, Xe, Ox, and Ek), distributed through-out the Maya region, are among the least well known because of how difficultthey are to locate archaeologically. Cunil pottery, described in detail in this vol-ume, is one of the earliest known ceramic complexes from the Maya area, butwhether it is in fact Maya is contested. Transitional ceramic complexes, occur-ring at times of change in the Maya past, are similarly difficult to identify, de-ne, and interpret—whether they are the orange wares and polychromes thatappeared between the Preclassic and Classic periods or the modeled wares, newares, and red wares relating to the transition between the Classic and Post-classic periods. Attempts at better dening and interpreting these enigmaticmaterials—as is undertaken by almost half the chapters in this volume—provide great benet to all students of Maya archaeology. Pottery will always be of critical importance to the archaeological inter-

pretation of ancient Maya society. is book provides a needed backgroundto the serious issues involved in the classication and interpretation of Mayaceramics and also addresses the signicance of ceramic analysis for deningtransitions that occur within the archaeological record. e chapters in this volume will have lasting value for researchers engaged in both Maya studiesand ceramic analysis in other parts of the world.

Diane Z. Chase and Arlen F. ChaseSeries Editors

Forewordxii


14/311

Preface

Most of the chapters in this volume originated as papers presented at two elec-tronic symposia held at the Society for American Archaeology (SAA) meetingsin and , and all deal in one way or another with type: variety-modemethod and theory. I organized the symposium because I was strugglingwith the classication of Postclassic period pottery from Lamanai in north-ern Belize. Having worked previously in the upper Belize River valley, I wasaccustomed to using Gifford’s ( ) Barton Ramie study as the basis of mywork, which mainly amounted to the identication of wares, groups, types,

and varieties established by Gifford and the other contributors to the vol-ume. ere are problems withPrehistoric Pottery Analysis and the Ceramics ofBarton Ramie in the Belize Valley (Gifford ), some of which I address in theconclusion of this volume, but that classic work is still the backbone of moststudies in the Belize Valley. When I got to Lamanai in northern Belize, I was faced with a large collec-tion of well-preserved pottery that had not been previously classied using thetype-variety system. Because the Lamanai collection is, in my opinion, one ofthe most important ceramic collections in the Maya lowlands, and becauseits original excavator, David Pendergast, has been a critic of the type-varietyapproach (see, for example, Pendergast : ), I was hesitant to assign type- variety names immediately, based on my more-or-less intuitive recognition ofstylistic groupings. As a rst step I set out to identify the standard methods oftype-variety. is brought me back to an issue I had noticed in the Belize Val-ley, which has been discussed by Rice ( ): problems with the ware concept.

Even if one ignores the original concept, which includes both paste and sur-face characteristics (as Willey and colleagues [ ] did for Copán), in a type- variety classication one must decide where to place paste variation. Doesit belong at the top of the hierarchy to dene wares or some similar category(such as Rice’s paste wares), at the bottom to dene varieties, or as a modalquality associated with any of the hierarchical levels of type-variety?


15/311

When I walked into the symposium in April this was the main issueon my mind, but there was much more in store. In an electronic symposium,papers are posted in advance on a Web site and the two hours of the meeting

are devoted to discussion of them, so I was expecting a fairly placid conversa-tion about archaeological method. What I experienced was one of the mostimpassioned discussions I have seen at the SAA meetings since I began attend-ing in . ere seemed to be disagreements at every turn, not just about theware concept but about what type-variety actually is, what it can (or should)do, and especially about how to use it. By the time I le the room two hourslater it was abundantly clear to me that there are many versions of type-varietyand very little agreement about its goals and methods. is prompted a secondsymposium in where the discussion was more subdued but no less useful,and thus this volume came into being. Compiling and editing this volume has taught me a great deal about Mayapottery generally and type: variety-mode classication specically, but I amnow only more aware of the complexity of many of the issues and how muchI have to learn. It is surreal to edit a volume with scholars who have been soinuential in the study of Maya pottery. As I put together these chapters I

continually drew upon their expertise, as well as that of others not directlyinvolved in the volume, such as Carol Gifford, and I am grateful to them all fortheir patience and intellectual generosity. I hope that this volume is useful toothers who, like me, continue to grapple with the issues raised by ancient Mayapottery classication, analysis, and interpretation.

Notes

. e term type-variety was coined soon a er the concepts of types and varieties wereestablished (as in Smith, Willey, and Gifford’s classic American Antiquity article).Later, the term was expanded totype: variety-mode (e.g., Gifford : – ) to indicatethat modal attributes were not consistently presented in type and variety descriptions andmore explicit ways of dealing with modes were desirable (see, e.g., the various tables inSabloff : ch. , and Forsyth’s use of paste and other modes). Independent modalclassication has only inconsistently been part of published type-variety reports, however,so I do not think it is accurate to use the termtype: variety-mode in all cases. ere is also

variation in the terminology used by different writers. Compare, for example, Culbertand Rands’s ( ) general use of the termtype-variety in an article that mentions type: variety-mode classication (but hyphenates the three terms asType-Variety-Mode) andAdams’s ( ) consistent use of the termType: Variety-Mode. ere is probably a goodargument to be made to usetype: variety and type: variety-mode, because the colon be-tween the rst two terms reects the intimate connection between types and their variet-ies and is also used in notation (e.g., San Felipe Brown: San Felipe Variety), whereas the

Prefacexiv


16/311

hyphen nicely indicates the less consistent and potentially separable inclusion of modaldata in type variety. In this volume I have chosen to use the general term that seems mostappropriate (uncapitalized) for the study or topic at hand (sometimestype: variety-mode but more commonlytype-variety ) and to allow the individual authors to decide whichterm to use.

. e terms pottery and ceramics are used interchangeably in this volume, as theynormally are in Maya archaeology, but the termceramics is sometimes used to refer onlyto high-red or vitried cooking and serving utensils and art objects (for a discussion,see Rice c: – ).

. I use the term paste to refer to the composition of red pottery excluding surfacetreatment, but the term fabric for the red product is more accurate.Paste should refer tothe clay or clay mixture used to produce the pottery before ring (see denitions in Rice

c). is distinction is more consistently made in Old World archaeology than in NewWorld studies (see comments in Howie ).

Preface xv


17/311

This page intentionally left blank


18/311

1Introduction

e ancient Maya produced a broad range of pottery, which has attracted con-certed scholarly attention for nearly a century. e authors of the chapters inthis volume address a range of issues around the classication, interpretation,and analysis of ancient Maya pottery linked in one way or another to the type: variety-mode approach. Given the many years of research devoted to Mayapottery, one might expect that there would be agreement over the goals and

methods of the type: variety-mode approach, but as I note in the preface, this isnot the case. In fact, it seems that every Maya archaeologist has an opinion ontype-variety, including (and perhaps especially) those who know little about it.

e authors of the chapters here also do not agree on a number of issues, andno doubt readers will have plenty to say. In this chapter, I present a brief reviewof each of the chapters. In the concluding chapter, I address some of the issuesthat have been raised about type: variety-mode more generally. In the chapter that follows this one, Prudence Rice provides a concise his-torical and theoretical introduction to type-variety and discusses importantaspects of classication in general. She highlights the difference between a pos-itivist view of classication, in which types are dened in any number of waysdepending on the questions one is asking, and what Hill and Evans ( ) havecharacterized as an empirical view in which types are seen to be inherentlymeaningful (for example, as expressions of individual and group values andnorms, as in the work of Gifford and others). She links the empirical view to

the “revisionist history” of Gifford’s ( : ) claim that type-variety is concep-tually based on whole vessels and argues that type-variety was in fact intendedto help manage large amounts of sherds, a task to which it is well suited. Rice also returns to the issues she raised in her important article on theware concept (Rice ). She reiterates and further explains her argument thatpaste and surface should be treated separately and the value of the category of


19/311

James John Aimers

“paste ware” for paste alone. She follows this point with an example from herresearch in the Petén Lakes region (in this volume, see also chapters by Ceciland by Urban and colleagues). Rice also discusses ceramic systems, a some-

what neglected concept that I nd very useful (see Henderson and Agurcia; Urban and Schortman ; Wheat et al. ; in this volume, see alsochapters by Aimers and Graham and by Bill). Systems lump stylistically analo-gous types (that is, ceramics that look basically the same and are from the sameperiod but have been given different type names for various reasons). Rice isa leader in the study of archaeological ceramics, and her contribution to this volume provides an excellent starting point for the chapters that follow. In her contribution, Cassandra Bill addresses some of the integrative catego-ries of type-variety, including ones that are commonly used (types and spheres)and others that are rarely used (such as horizons and systems). Appropriately(or perhaps ironically), Bill creates a classication of systems and spheres. Billidenties two types of systems (which we might call regular and long-lived)and three types of spheres. Because systems as originally dened lump together types, which are typi-cally associated with a particular time period (phase), they are not normally

conceived as long-lived. Following observations by archaeologists working inHonduras, where the systems concept has been used most extensively, Bill re-minds us that styles o en continue through time (that is, across phases identi-ed as ceramic complexes) and that the concept of system can thus be used tolump analogous types that continue through time (that is, stylistically similarceramics that span multiple phases/complexes). She notes that these long-livedsystems are rather like ceramic traditions without the spatial restriction char-acteristic of traditions, and because systems are ideally relatively circumscribed

in time she suggests calling these long-lived systems “macro-traditions.” Bill looks closely at the concept of spheres, which group together entire ce-ramic complexes when they share a majority of their most common types (SeeBall for a good discussion). Bill makes distinctions between at least threekinds of spheres. First, there are spheres in which pottery vessels themselvesare produced and exchanged across a region, indicating a relatively high levelof economic integration in that area. I think of these as “economic” or “produc-tion and distribution” spheres. Types in these kinds of spheres would be identi-cal down to the varietal level. Second, there are spheres in which potterystyles are produced and exchanged across a region, indicating shared ceramic pro-duction practices and social interaction of some sort but not necessarily a highlevel of economic integration in that area. I think of these as “social” or perhaps“stylistic” spheres in which types are shared but are expressed locally through


20/311

Introduction

varieties designated (as varieties normally are) based on differences in techno-logical attributes, minor formal attributes, or both. ird, there are spheres inwhich the sameconventions are found at different sites but are manifest in ce-

ramics that are different enough to be considered different types. Nevertheless,the similarities of the types in such a sphere would be close enough so that theycould be lumped into a ceramic system, so I think of these as “systems spheres.”

is third type of sphere is intriguing because it resolves a problem raised bythe fact that some archaeologists use existing type names when analyzing anew sample, whereas others prefer to assign new type names. As Bill notes, us-ing the same type names may suggest closer similarities than there are betweensamples; assigning new types does the opposite. Henderson and Agurcia ( :

) have also suggested creating sphere-like entities based on systems, andI agree since it resolves this methodological bind and allows us to efficientlyuse ceramics to address questions of interaction (Aimers a, b). Bill’sdifferent types of spheres are concepts through which we may rene our ap-proaches to different kinds of interaction (for example, vessel exchange versusstylistic emulation). e complexity of Bill’s ideas here will hopefully stimulate reection on our

varied and o en uncritical (or worse, simplistic) use of some of the categoriesshe discusses, and perhaps further renement of them. One of the gratifyingelements of her chapter is that it squarely addresses a common criticism oftype-variety: the idea that type-variety classications tell us little about ancientMaya ideas and practices (or Maya “culture” more generally). As Bill showswith a number of specic examples, each of the concepts she discusses can beused as a starting point for inferences about ancient Maya life, especially therelationships among groups of potters (or potentially even individual potters)

and the communities of which they were a part. Type-variety need not be usedas anything other than a common language—that alone is of great benet—butthe idea that type-varietycannot usefully be used as a basis for cultural infer-ences is discredited by the array of observations Bill derives from just a few ofthe concepts of type-variety. Arlen Chase and Diane Chase’s chapter serves as a counterpoint to Bill’sbecause they highlight the inadequacies of type-variety. ey argue for con-text-based approaches to Maya pottery with an emphasis on the associationsof whole vessels and retting of sherds because full or reconstructable vesselsfrom good primary deposits are the ideal material from which to derive infor-mation about ancient Maya behaviors and beliefs. Unfortunately, this is notalways possible, for any number of reasons (for example, when excavationsare limited due to funding or as part of early investigations at a site). I suspect


21/311

James John Aimers

that fewer and fewer projects will have the resources to engage in the sort ofintensive sampling the Chases quite reasonably advocate. e reality is that weare o en forced to deal with poor samples of sherds (sometimes eroded ones)

from secondary or poorly understood contexts. If the resources of the twomajor projects at Tikal could not solve some of these issues, one wonders howsmall projects could hope to do so. Despite beginning the chapter with a provocative quotation from omp-son ( ), the Chases end on a more conciliatory note with comments similarto those made by Culbert and Rands ( ): we can derive the most informa-tion from Maya pottery when we use multiple approaches. is is a themethat runs through this volume as a whole, and I return to it in the concludingchapter. In his contribution, Robert Fry interrogates assumptions about pottery pro-duction and exchange with case studies from two regions using monochromeslipped bowls and unslipped jars—the more ordinary pottery of the Maya,which has not been given as much attention in regard to these issues. Becauseof issues of sample size, Fry (unlike the authors of other chapters in the vol-ume) does not use type-variety, but this would be easy to do with signicant

samples. Instead, Fry uses multidimensional scaling (MDS) to conduct what isessentially a complex modal analysis. Importantly, due to the relatively homo-genous geology of the region, stylistic variables were more revealing of regionalpatterning than were technological attributes. In my contribution with Elizabeth Graham, we describe how ceramics havebeen approached at Lamanai, Belize. Several people are conducting multipleindependent analyses of the ceramics using contextual associations of vessels,type: variety-mode classication, material science approaches, and icono-

graphic analysis. We emphasize type-variety in the chapter, particularly theusefulness of ceramic system assignments as a preliminary stage of analysis orfor sherds that lack important features diagnostic of specic types, as well aspoorly preserved sherds. We also briey indicate some of the results of com-bining various approaches to pottery at Lamanai. At this point the focus of the volume shi s slightly from chapters orientedto method and theory to chapters that deal more directly with specic sites andsamples, although each of these also has much to say about pottery classica-tion and analysis generally. Lauren Sullivan and Jaime Awe discuss the pottery of the earliest-knownceramic complex in the Maya region, Cunil, which they date from to B.C. As the authors note, because these dates are so early, the Cunil ceramicshave been assessed very cautiously. For those outside the debates around the


22/311

Introduction

Preclassic period and its pottery, this chapter may appear straightforward, butit represents nearly two decades of work on the denition of the Cunil com-plex, especially its dating, and the result of years of discussion among archae-

ologists. Sullivan and Awe’s chapter makes a tting starting point to this section ofthe volume not only because the Cunil complex is the earliest currently knownpottery complex in the Maya lowlands but also for the attention they draw toone of the most basic functions of type-variety analysis: stylistic comparisonfor chronological assessment. e capabilities of radiocarbon dating still donot match the precision of date estimates that can be gained by close stylisticanalysis of Maya pottery (especially for later and better-known periods likethe Late Classic, where archaeologists now discuss chronology in hundred- oreven y-year segments with some condence). Overcondence inC datescan lead to serious misunderstandings, as in the confusion and controversycaused by the misdating of the Swasey complex at Cuello based onC dates(see, for example, Andrews and Hammond ; Hammond et al. ; Ko-sakowski ). Sullivan and Awe note that it was “gratifying” to see the levelof agreement among different archaeologists when samples of Cunil pottery

were laid out stratigraphically at the Belize Ceramic Workshop in andcompared directly with those of Swasey and other complexes by archaeologistsintimately familiar with these materials. is reinforces Ball and Taschek’s ob-servation in this volume that hands-on visual inspection of pottery is not justdesirable but essential for reliable comparisons, a fundamental point that is tooo en overlooked. Sullivan and Awe present here their preliminary typology of the wares,groups, types, and varieties of the Cunil complex. ese provide a succinct lan-

guage for further discussion of Cunil across the Maya lowlands; as they note,this discussion has already begun. But, as they caution, types are always opento further revision. Although I have only been on the margins of the protracteddiscussions about Cunil, I suspect this revision will continue, as it should forthis early and very important pottery and, in fact, any Maya pottery. In their chapter, Michael Callaghan, Francisco Estrada-Belli, and NinaNeivens investigate the idea that the glossy orange-slipped polychrome potteryproduced in the late Terminal Prelassic to early Early Classic period in the Hol-mul region was a form of “social currency” linked to the emergence of formalelite power signied in various ways (for example, the appropriation of sacredspace for elite burials). ey note, however, that these fancy serving vesselsare found not only in elite contexts but also in ritual contexts of various sorts,includingchultuns and caves. Reporting on some of the ndings in Callaghan’s


23/311

James John Aimers

( b) dissertation, the authors combine manufacture-related evidence withdistributional and stylistic data to suggest that the change from glossy red-slipped pottery to glossy orange-slipped vessels reects more general changes

in worldview that may be linked to changes in social structure. ey discusstechnological attributes (including paste recipes, ring technology, and surfacenish), possible number of production groups, and distribution within theregion and beyond. Overall, it seems that the number of production groupsexpanded through time, or at least that manufacturing technologies becamemore varied. Specically, Preclassic period monochrome Sierra Red vesselsshowed less technological variation than later orange-slipped types (bothmonochrome and polychrome). e usefulness of type-variety here as a descriptive language is clear, butthe complexity indicated by the technological analyses is much greater thanthat conveyed by stylistic or formal classication alone. Notably, paste recipescrosscut stylistic types, suggesting that a single production group made mul-tiple stylistic types with essentially the same technology (for similar examplesfrom Lamanai, Belize, see Howie ). Another important nding in terms ofthe themes of this volume is that orange-slipped ceramics were not introduced

to the Holmul region from elsewhere in the Terminal Preclassic but derivedfrom local production traditions (see, for example, Brady et al. ). e authors conclude with some speculative but provocative suggestionsabout possible links between changes in technological style (as dened byLechtman ) and social structure and worldview, including the idea thatcrushed sherd (grog) temper in new vessels represents an attempt to connectwith the past. Milbrath and Peraza (this volume) note a similar idea in refer-ence to ethnohistorical evidence that temper made from ground Chen Mul

Modeled censers was used in the manufacture of new ones as a means to trans-fer the power of the deity image (Chuchiak ). Callaghan and colleaguessuggest, conversely, that the abandonment of grog temper may reect an at-tempt to break with the past. Overall, this chapter is well suited to this volumesince it demonstrates how productive the use of multiple analytical techniquescan be in the study of ancient Maya pottery. e combination of stylistic analy-sis and technological analysis provides a window into ancient Maya behavior(and, as they suggest, perhaps worldview as well) that neither could providealone. In a chapter rich with culture-historical interpretations as well as method-ological commentary, Joseph Ball and Jennifer Taschek begin with the usefulreminder (via Robert Sonin) that type-variety classication is not analysis butmerely a step toward analysis (see also quotes from Dunnell in Rice’s chapter).Too o en in Maya archaeology there is a sense that type-variety classication is


24/311

Introduction

standard operating procedure for pottery and if one simply “does” type-variety,then any number of interpretations on a variety of topics can then be extractedfrom the classication. I believe this is a part of a usually unexamined assump-

tion that all type-variety classications are essentially equivalent and easilycomparable. On the contrary, the questions one is posing will shape how theclassication proceeds (as Rice notes at the beginning of her chapter, the corol-lary is that classications also limit the research questions one can address), soresearch questions should be denedbefore the classication begins. Once thequestions and methods of classication are dened, classication occurs andthe analysis ows from there (see also Aimers ). Similarly, Ball and Taschek conclude their chapter by forcefully emphasizingthe need for hands-on examination of sherds and vessels in type-variety iden-tications/classications and the related inadequacy of publications for thistask. eir distinctions betweenidentities (ceramics that are stylistically andtechnologically indistinguishable) andhomologies (ceramics that share stylisticelements but are technologically distinguishable) are essential to their inter-pretations yet usually impossible based on photographs or illustrations alone.

e need for rsthand observation may seem obvious in a discipline as empiri-

cal as archaeology, yet it is astonishing—even scandalous—how o en peoplemake assessments based on written descriptions and/or black-and-white il-lustrations alone, and the problems this procedure causes are disconcerting.Erroneous equivalencies between ceramics at different sites and the opposite (aproliferation of identical types with different names) are just two of the prob-lems this procedure causes. is chapter also contains examples of how ceramic phases may overlap intime and space. Ceramic phases are not always sequential: they can overlap

or be separated by a hiatus, or they can be partially overlapping and partiallysequential (see also comments by Urban and colleagues, this volume). iscomplexity is not easy to diagram or describe but ts better with the reality ofstylistic change. Even today where stylistic change is rapid in many areas (forexample, in fashion or interior design), new styles do not immediately and to-tally replace old ones, yet in archaeological ceramic chronology the battleshipcurves that were the basis of the earliest stylistic seriations are o en ignored inthe presentation of data. Foias ( ) and Forsyth ( ) have discussed this;Gifford ( : ) explains the logic behind it. us, even the slanted and dottedlines and the nested Xcocoma phase in Ball and Taschek’s chronological chartmake important methodological points. In substance, Ball and Taschek’s chapter is a reevaluation of their olderideas about the signicance of the Xcocom phase ceramics at Becán using thestrengths of type-variety: intersite comparison and chronology. Ball and Tas-


25/311

James John Aimers

chek both redate and reinterpret the meaning of these deposits, arguing thatXcocom comprises two subcomplexes (Xcocoma and Xcocomb), represent-ing a Gulf Coast/Puuc-associated occupation and a Caribbean-associated one,

respectively. is highlights another o en-overlooked characteristic of type- variety: classications are not meant to fossilize a er publication but shouldalways remain open to reassessment and renement as data accumulate. efact that many archaeologists acknowledge this (for example, Gifford : ;Forsyth : ) has not prevented fossilization from occurring. Continual revision is, of course, the nature of science as well. I suspect, as Ball and Taschek indicate, that their specic culture-historicalinterpretations (for example, about Itza invasion at Acanmul) may be ques-tioned and, as usual, more excavation would be useful, as they acknowledge.Still, their chapter is unusually clear in revealing the processes through whichceramic data (including contexts and associations), type-variety methods, andculture-historical interpretation are linked. Patricia Urban, Edward Schortman, and Marne Ausec examine their useof type-variety classication to identify the Early Postclassic in three areas ofHonduras. Like Sullivan and Awe, these authors stress the exibility of type:

variety-mode, which they continue to modify as new types of data are rec-ognized and new questions are asked. While recognizing some shortcomings(particularly that type-variety alone tells us little about vessel function), theyargue that type: variety-mode has been useful for the identication of spatialand temporal variation and has allowed them to deal efficiently with sherdsamples that sometimes number in the millions (a prospect that makes meshudder) in a relatively small area with striking pottery diversity. ey make these points with discussions of the three areas. In the middle

Ulua–Santa Barbara area, the identication of the Early Postclassic was rela-tively straightforward due to the emergence of new types—including a newtype within an existing group—and some reliable nonlocal markers such asLas Vegas Polychrome. ey caution, however, against reliance on nonlocalmarkers, because such markers were not present at all loci known to have EarlyPostclassic occupation based on shi s in the presence or absence of local types. For the Naco Valley area, they note that in the absence of ceramic mark-ers of the Early Postclassic they relied on negative evidence (the absence ofcommon Early Postclassic types) and shi s in the frequencies of types. In theMiddle Chamelecon-Cacaulapa zone, they note rst the great variety of pasteformulas in use with the same surface treatments (for a similar case see Rice,this volume) and the fact that increasing coarseness in paste composition hasbeen a useful temporal diagnostic. us, they have used paste as an important


26/311

Introduction

distinguishing feature in their classications, as do Rice and others. Shape-based modes are also important to their work, but new vessel forms do notappear and disappear neatly—their relative frequency changes.

Leslie Cecil’s chapter, like the contribution by Callaghan and colleagues,demonstrates the advantages of combining information gleaned from techno-logical and stylistic analysis. In this case, the focus is predominantly on slips,although research on pastes is included, and Cecil is able to incorporate an-other valuable type of information: ethnohistorical accounts of the Petén Lakesgroups, especially the Kowoj and Itza. By marshalling these disparate sources,Cecil is able to make suggestions about links between style, technology, andsociopolitical identity and even trading patterns that would remain speculativeusing archaeological data alone. One of the exciting aspects of Cecil’s research is her ability to link patternsin the pottery data with these ethnohistorically and historically known groups.

us, her work contributes to a substantial literature in ceramic studies onthe contested links between style (technological and otherwise) and identity,particularly the problematic concept of “ethnic” identity (for an entire volumeon the concept of ethnicity in Maya studies, see Sachse ). Debates over

the presence or denition of ethnicity among the ancient Maya aside, Cecilconcludes that the multiple known sociopolitical groups in the Petén Lakesregion shared some of their manufacturing traditions and that trade amongthese groups was relatively unrestricted. Again, the productivity of bringingmultiple sources of data and interpretive frameworks to bear on Maya ceram-ics is evident. In their chapter, Susan Milbrath and Carlos Peraza discuss the famous ChenMul Modeled full-gure effigy censers of Mayapán. Variants of these are found

across the Maya area in the Late Postclassic; in an article we wrote with LyndaFolan (Milbrath et al. ), we argued that archaeologists should be morecautious in using the Chen Mul Modeled type name for effigy censers withsignicant stylistic variation from the Mayapán censers. We suggested thatmany people are using the Chen Mul Modeled type name like a system name,to identify broad stylistic similarity that nevertheless encompasses signicantlocal variation that may be of use in mapping intersite and interregional inter-action. Milbrath and Peraza provide the most comprehensive summary to date ofthe possible origins of Chen Mul and conclude that the origins of this ceramicsystem in all likelihood lie to the west of the Maya heartland in the Late andTerminal Classic, possibly as part of the famous Mixteca-Puebla tradition atsites like Cacaxtla and Cholula, or perhaps on the Gulf Coast. In any case, the


27/311

James John Aimers

most current data suggest that the style is rst found at Mayapán near the be-ginning of the thirteenth century. e authors also provide a concise and well-illustrated stylistic description

of the Chen Mul Modeled censers from Mayapán, an analysis that brings to-gether formal, iconographic, ethnohistorical, and even linguistic evidence andwill be of great value to Maya scholars given the widespread distribution of theChen Mul Modeled system. Near the end of the chapter, they point out that thecentral Mexican deities depicted on Chen Mul Modeled censers at Mayapánsuggest trade contacts between the site and the central highlands (for example,the Aztec empire) at the end of the fourteenth century and into the eenth.

ey note that Mayapán’s most famous export, Maya blue pigment, was one ofthe Maya elements that went west; this may help explain why Mayapán showssuch strong iconographic connections to the central highlands of Mexico. Along with Petén polychromes, the Mayapán Chen Mul Modeled censersare icons of Maya pottery in general because of the visual richness of theirstyle and iconography. e widespread distribution of the Chen Mul Mod-eled ceramic system makes it an excellent resource for understanding ceramicexchange and stylistic interaction. Milbrath and Peraza’s chapter provides a

wealth of insights of value to those of us trying to reconstruct Postclassic pat-terns of interaction across the Yucatán Peninsula and beyond. As a whole, the chapters in this volume raise a host of issues about type: variety-mode specically and the study of Maya pottery more generally. Givenwhat I saw at the symposia from which most of the chapters are derived, therewill no doubt be disagreement among archaeologists about some of the opin-ions expressed and the various conclusions reached. I address some of thebroader issues in the concluding chapter to the volume.

Note

. In response, I am tempted to create a type called Abracadabra Black, although place-names are more typically used in type names.


28/311

2Type-Variety

What Works and What Doesn’t

.

Because classications organize and structure data, they inevitably also orga-nize and structure the formulation of research problems. However, classica-tions are only tools: they are a means to an end, not ends themselves. usit is necessary to ask continually if the goals and the procedures of ceramicresearch are existing in optimal relationships, or if classicatory systems,which are always conservative and resistant to innovation, may be impeding

rather than enhancing ceramic research.Rice :

e type-variety system has been used for pottery classication in the Mayaarea for nearly half a century, surviving—or “doomed to success” (Adams

)—despite recurrent attacks. e type-variety approach works well atwhat it was originally intended to do—structure descriptions of archaeologi-cal pottery for spatiotemporal comparisons—but its detractors complain whenit fails to accomplish things far beyond this goal. Different units of the type- variety system can have different kinds and degrees of utility, depending on thegoals of any particular analysis, but type-variety is, rst and foremost, simply ahierarchical framework for organizing descriptive data about pottery.

Pottery Systematics: An Overview of Classication

Study of a collection of archaeological pottery generally proceeds in threestages or levels: classication, analysis, and interpretation or explanation (Hilland Evans : ; Rice : – ). My focus in the following discussionis on the rst and lowest level, classication and its constituent procedures. Analysis (Greek, “break up, loosen, separate”) involves collapsing a collection(of pottery) into some set of constituent attributes to determine their nature,


29/311

Prudence M. Rice

proportions, and/or relations; thus, pottery analysis may be stylistic, compo-sitional, modal, morphological, and so forth. e highest level of study isin-terpretive or explanatory, in which the pottery of interest is situated into some

broader context (theoretical, spatial, chronological, social, economic) to con-tribute to the resolution of a theoretical issue, answer a question, or solve aproblem. Classication (also known as taxonomy or systematics) refers to the mostgeneral scientic process of ordering objects, be they potsherds, plants, oranimals. e process involves organizing the objects into “similarity groups”based on one or more characteristics of interest. Members of similarity groupsare more similar to each other than they are to members of other groups. is“high within-group homogeneity” is the goal of all scientic systematics andinformatics operations (Rice c: – ). Classicatory activities may bepursued in two ways, through categorization or identication. Categorization is the creation of groups of previously unclassied or un- grouped materials. An example would be a ceramic assemblage from a rela-tively unknown site in a relatively unexplored area or time period. Few suchassemblages emerge from the heavily explored Maya lowlands in the twenty-

rst century, but my work at Spanish-colonial wineries in Moquegua, south-ern Peru, provides a case in point. Excavations produced mixtures of locallymade, hand-built, pre-Hispanic pottery, plus Spanish-inuenced wheel-made,tin-enameled, lead-glazed, and unglazed pottery. Initial categorization of thismaterial was based on two variables: surface treatment (presence/absence ofenamel/glazing) and technique of manufacture (wheel-thrown versus hand-built). Categorization o en results in the creation of new, provisional ceramictypes: for Maya pottery, see, for example, the procedures of early classica-tion of the pottery of Altar de Sacricios (Adams : – ) and Barton Ramie(Smith, Willey, and Gifford : ; Gifford , ). Procedures of ceramic taxonomy can be contrasted with those of other archaeological materi-als: for example, classication of human, deer, or chicken bones does not—in-deed, cannot—involve the process of categorization or naming of new types,because all such bones are already known and named. Identication is the process of assigning individual objects to previously estab-lished classes. In the case of human bone, classication is always a process ofidentication: a specimen can only be identied as one of the long-knownand long-named bones in the human body, the variability of which is furtherconstrained by le -side/right-side distinctions. With pottery, identicationconsists of evaluating objects with respect to already-established categories thatset out the parameters of variability that we can expect in our collections. In the


30/311

Type-Variety: What Works and What Doesn’t

case of the Moquegua material, I compared my hand-built, unglazed sherdsto numerous published reports on pre-Hispanic southern Peruvian potteryto identify them as examples of those materials. However, very little system-

atic description of early colonial Peruvian wheel-made and/or tin-enameled/glazed pottery existed. us, for these colonial materials I began with simplecategorization and used type-variety structuring principles to create a typology(Rice ). e distinction between categorization and identication seems obscu-rantist, but it is important. It is based on—and also illuminates the historyof—what is already known or has been accomplished in Maya pottery analy-ses, as compared to what has not. It is important to remember that catego-rization and identication are empirical processes: both refer to the assign-ment ofactual, physical objects, such as the Moquegua sherds, to organizingunits. is distinction between categorization and identication has been ex-panded and effectively diagrammed by Robert Dunnell ( b: ). He distin-guishes between the phenomenological or empirical (objectively “real”) versusthe ideational (“conceptual”) realms in the general activity of ordering objects,

such as pottery. Dunnell also makes the distinction between “grouping” proce-dures and procedures of “classication proper” or paradigmatic classicationsensu stricto (table . ).

Grouping occurs in the phenomenological (that is, materially “real”) realm.It refers to the physical sorting (or categorization) ofactual objects, such assherds excavated from the Moquegua Valley, into similarity groups created

Table 2.1. Schematic of classicatory principles and characteristics of resultant units

REALM

PHENOMENOLOGICAL IDEATIONAL

PROCEDURE GROUPING CLASSIFICATION

(categorization) (identication)

Creates: Groups of objects Conceptual classes

Based on: Objects, single attributes Attribute clusters, ideas

Interpretation: Actual, historical, nite Potential, historical, inniteUNIT RELATIONS

EQUIVALENT (unordered) Statistical clustering, most modalanalysis

Some modal analysis

HIERARCHICAL Numerical taxonomy, clusteranalysis

Typology, classication(e.g., Linnean; type-variety)

Source: After Dunnell 1971a: g. 9.


31/311

Prudence M. Rice

for convenience. Groups are thus historical or contingent by virtue of beingbounded by that nite set of objects being grouped. Classication proper (or paradigmatic classication) is based in the ide-

ational realm and is aconceptual structure for orderingattributes or features (not actual objects!) intoclasses that are proposed on the basis of somecri-terion or theory . e features or variables of pottery being ordered may besize, form, color, iconography, technology, composition, and so on. Criteria fortheir selection may be overarching social, political, economic, or other theo-retical issues, concepts, or problems encompassing power, status, production,trade, feasting, ideology, mortuary ritual, and so on. In the case of the colonialMoquegua pottery, subsequent sorting criteria included glazing versus enamel,and color of the enamel. e outcome of classication proper is the creation of atypology : “a theo-retically oriented classication that is directed toward the solution of someproblem or problems” (Gifford : , quoting Kluckhohn) and broadlyuseful comparatively. e units in a typology are based onideas about attri-butes or features—method of manufacture, kind of surface treatment, color,form—and their relations as criteria for membership. us typologies are, in

principle, ahistorical and innite in their spatiotemporal applicability, ratherthan being limited to a nite set of excavated objects on the sorting table. ese ideational classes also organize the phenomenological (“grouping”)

realm for later ordering operations—which permits the process of identica-tion. e advantage of paradigmatic classication systems (as opposed to group-ings) lies in their replicability and testability. Groupings (statistical clusters andnumerical taxonomy) are created on the basis of aspecic data set, and their

utility cannot be tested on the same set or even, technically, on a different one.An important contribution of groupings is to generate hypotheses about theformation of classes and then, subsequently, to test them with other data sets(Spaulding : ). In summary, classication is a general term that refers to several proceduresfor organizing objects into units having high within-group similarity.

Grouping refers to procedures of physically sorting actual objects to create

groups of similar things, a process that may be accomplished in two ways:. Categorization: creation ofnew groups of objects. Identication: assigning objects toexisting groups or categories

Classication proper , or paradigmatic classication, refers to proceduresbased on, and formalized into, a typology: a conceptual structure of at-tributes relating to theoretical needs.


32/311


Dunnell ( b) further distinguishes between attributes that are equivalentand of equal weight in the sorting versus those that are nonequivalent, ordered,or weighted. An example of the former is classes based on form: bowl, jar, plate,

and cup are equivalent, unordered dimensions of the attribute “form,” and weneed not have these categories physically present to conceive of their existence.Type-variety, by contrast, is a scheme of units ordered by quasi-theoreticalconcepts (or objectives) in which the various pottery attributes of interest arenonequivalent and ordered in different levels. us, the resulting classicatoryunits or levels are nested: varieties are organized into types, types into ceramicgroups, and ceramic groups into wares. Most simply stated, the type-varietysystem is a hierarchically organized typology. e important point is that the theoretical basis of a typology is manifestin the selection of, and relations among, the attributes that underlie its ide-ational structure and operations. roughout most of the twentieth century,the research problem(s) addressed by classications of archaeological ceramicswere those of comparative chronology and dating of sites. In the twenty-rstcentury, other issues are of greater interest and may or may not be addressedby the type-variety system. Regardless, it is these underlying theoretical goals

that move us beyond simple systematics and into the realms of analysis andinterpretation of pottery.

Historical Development of the Type-Variety System

Methods of classication in general and type-variety in particular are auniquely Americanist concern of the past two centuries (see Willey and Sabl-off ), particularly with respect to determining the origins—ethno-cultur-

ally and temporally—of the indigenous inhabitants of North America and theearthen mounds in the eastern United States. Absent reliable dating methods,the study of archaeological remains was largely conned to description anddevelopment of chronologies. Advances occurred primarily in the southwest-ern United States in the s, where vigorous and systematic efforts to createpottery types bearing chronological or historical signicance, rather than sim-ply descriptive clarity, were led by a handful of archaeologists (Gladwin andGladwin ; Hargrave and Colton ; Colton and Hargrave ). ese efforts resulted in a hierarchical classicatory system built on theearlier binomial nomenclature adopted at the Pecos Convention (Kid-der ; Willey and Sabloff : , n ). For pottery, this was a varianof biological taxonomy in which the ceramic equivalent of a “genus” namerefers to the surface treatment and the “species” name is taken from local ge-ography. Descriptions of types were to be presented in a set format of name,


33/311

Prudence M. Rice

shape, decoration, “type site,” cultural affiliations, and temporal data (Willeyand Sabloff : ). is system was adopted in the southeastern UnitedStates (for example, Phillips ). At the same time, a crosscutting analytical

or modal system of descriptive organization (technically, phenomenological“grouping”), based on modes/attributes rather than on sherds, was introducedby Irving Rouse ( ). e apparent agreement on the conceptual structuring of pottery classica-tions in the American Southwest and Southeast was followed by two decadesof disputes about practice and interpretation (see Hill and Evans ). edebate is typically framed in terms of two polarized positions. On the one sidewas James Ford ( ) and his adherents (see, for example, Brew ; Dun-nell a, b), who believed types were arbitrary constructs or conceptsimposed over a collection of artifacts by the archaeologist. On the other wasAlfred C. Spaulding ( ), who saw types as meaningful, statistically veriableassociations of attributes discovered by archaeologists; like-minded archaeolo-gists (see, for example, Gifford ; Rouse ) claimed that pottery typesare inherent in a collection and represent the ideas (“mental templates”) of theproducers and users.

e contentious atmosphere was inuenced by the publication of Wal-ter W. Taylor’s Harvard University dissertation, “A Study of Archaeology.”In it, Taylor inveighed against archaeologists’ practice of mere “comparativechronicle” and the vacuity of Maya pottery studies as practiced to date, quip-ping famously that “the road to Hell and the eld of Maya archaeology arepaved with good intentions” (Taylor : ). Taylor’s dissertation was inu-enced by an earlier diatribe against Middle American archaeology penned byhis advisor, Clyde Kluckhohn ( [ ]). Kluckhohn, an ethnologist of the

culture and personality school, promoted an emic approach: ethnographersmust seek to understand the minds of their subjects, their attitudes, values,ideas, and feelings. Both publications excoriated the Carnegie Institution ofWashington (CIW) and particularly Alfred Vincent Kidder for focusing onelite architecture instead of trying to understand “the culture,” and both de-rided Carnegie’s lack of rigor and theory and its emphasis on (allegedly badlywritten) culture history. Both advocated a more robust, “scientic” approachto archaeology, which Taylor termed “conjunctive.”

Type-Variety in the Maya Lowlands

In , the classicatory procedures of North American archaeology enteredthe realm of Maya lowland pottery studies. e main proponents were Har-


34/311


vard University archaeologists and students Richard E. W. Adams, Jeremy A.Sabloff, Gordon R. Willey, and particularly James C. Gifford, who had cut histeeth in Southwest archaeology and pottery classication (Gifford ). Ear-

lier Maya pottery studies had followed procedures of modal or attribute analy-sis and description, typically segregating pottery into generic “wares” on thebasis of slip color and form as at Uaxactún (Smith ) and San José, Belize( ompson ). In an important article in American Antiquity , Robert E. Smith (of theCarnegie Institution), Willey, and Gifford ( : ; herea er cited as SWG

) announced the merger of the rened variety concept (from Wheat, Gif-ford, and Wasley ; herea er abbreviated as WGW ), with the type,as modied for application in the Southeast (Phillips ), into what theyreferred to as the “type-variety concept.” In the same issue, Gifford ( )published his ideas about this method “as an indicator of cultural phenom-ena.” e newly named type-variety “system” was adopted into the Maya areabecause an “explosion” of archaeological eldwork had produced massiveamounts of pottery fragments that introduced “complications . . . into theculture-historical picture,” requiring a new classicatory structure to facili-

tate data sharing (Willey, Culbert, and Adams : ; herea er abbrevi-ated as WCA ). Smith, Willey, and Gifford ( : – ) begin their exposition of type- variety with what could be read as a tacit apologia for the Carnegie Institution’sprocedures, so heavily criticized by Taylor. ese authors claimed that theCarnegie’s inadequacies resulted because “no two people followed the samedenition or concept of ware and type” and because, except for some of AnnaShepard’s work, compositional analyses were “largely neglected.” eir solu-

tion was the type-variety concept, and they redened various concepts, dis-cussed methods of sorting and rules of nomenclature, and advocated integrat-ing modal studies. ey identied the variety as the “basic unit” of analysis(following WGW : – ), suggesting that classication should begin with varieties; types and larger, more inclusive units should be distinguished later inthe process (SWG : ; contra Sabloff and Smith : ). e Mayanists argued that types and varieties were “realities within the cul-tural conguration of their origin and it is our job as analysts to recognize”them; they were “mental templates,” not merely analytical constructs imposedon a collection (SWG : ; Gifford : ). Gifford ( : ,

– ) expanded the interpretation of these “realities” beyond the spatiotem-poral to include societal values and “innate human tendencies.” Within tra-ditional societies, he writes, there is an acceptable range of variation in ideas,


35/311

Prudence M. Rice

standards, and preferences about the quality of any good, such as pottery, andproducers and consumers conform to these norms. us, varieties “closely ap-proximate actual material ceramic manifestations of individual and small so-

cial group variation in a society” (Gifford : ). It follows, then, that typesare summations of individual and small social group variation and representthe “crystallization of conscious or unconscious ceramic esthetic images con-ditioned by values” (Gifford : ). I nd it difficult to believe that this line of thought is anything other thana response, whether direct or indirect, to Kluckhohn’s and Taylor’s erce cri-tiques that Maya archaeology was oriented toward accumulating facts ratherthan driven by (cultural anthropological) theory and synthesis. Gifford ( :

) thanks Kluckhohn in his acknowledgments and cites two of his papers onthe study of values (Kluckhohn , ). e introduction of type-variety to Mayanists, particularly the endlessly dis-puted esoterica of nomenclature (Gifford ), did not sail smoothly, prompt-ing the organization of the “Conference on the Prehistoric Ceramics of theMaya Lowlands” in Guatemala City in . As reported in American Antiquity (WCA ), this conference of only ten participants, all North American, in-

cluded discussions of theoretical concepts and units, particularly spheres, andhands-on comparisons of actual sherds. In hindsight, the conference summary,while chronicling participants’ disagreements, may convey more solidarity andacceptance than actually existed, because arguments about type-variety con-tinued to rage throughout the s and s. In , Sabloff and Smith updated and critiqued current practices, includ-ing a comparison of procedures used in the Mayapán and Seibal reports. eyreiterated the earlier (for example, SWG ) call to combine typological and

analytical/modal analysis, if only by listing the principal identifying modes,and detailed the criteria for and signicance of ceramic groups and wares. Ina sharp counterpoint, Robert Dunnell ( a, b) assailed the type-varietysystem from the point of view of general systematics and positivism: “If clas-sications of any kind are to be devices useful in constructing explanations,. . . they must behypotheses about the ordering of data for a specic problem. . . .Only with specically dened problems is it possible to evaluate the utility, parsi-mony, elegance, and sufficiency of a given classication” (Dunnell b: – ,emphasis added; also Hill and Evans ). Opposed to a basic premise oftype-variety, Dunnell ( b: ) concludes that “[c]lassications are logicalconstructs whose justication lies in their utility. ey are not inherent, nor dothey explain. ey are imposed constructs that function to order data so thatexplanation is possible.”


36/311


On Wares and Pastes and Paste Wares

I am particularly interested in the “ware” unit of type-variety classication. Asdened and used by Southwestern archaeologists, a ware is “a large groupingof pottery types which has little temporal or spatial implication but consistsof stylistically varied types that are similar technologically and in method ofmanufacture” (WGW : – , citing Hargrave and Colton : – ;Colton and Hargrave : – ). In the introduction of type-variety to Ma-yanists, wares were characterized as displaying “consistency in such techno-logical attributes as paste or surface nish” (SWG : ). e report of the

Maya ceramic conference described a ware as “a number of ceramic types

sharing a cluster of technological attributes” (WCA : ). Later, this waselaborated such that “a dened ware is a ceramic assemblage in which all at-tributes of paste composition (with the possible exception of temper) and ofsurface nish remain constant” (Sabloff and Smith : ). Paste composi-tion variables were those of texture, temper material, hardness, porosity, andcolor, whereas surface nish attributes included presence/absence of a slip,smoothness, luster/matte nish, and color. e Maya ceramic conference reported some disagreement aboutwhere ware belonged in the hierarchical arrangement of taxonomic units:Smith felt it was a higher-order unit, “making the separation of wares the rstprocedural step in ceramic analysis,” whereas others thought wares were moreintegrative and “abstracted from completed type denitions” (WCA : ).

us, wares organized the Mayapán and Barton Ramie type-variety classica-tions but not the Copán analysis (Willey et al. ). is ambivalence bespeaks a continuing fundamental problem with theware unit. Note that as originally presented (SWG : ), the attributes ofware were pasteor surface nish, echoing the Southwesternists’ concern withtechnology. Later, wares were dened by pasteand surface nish (Sabloff andSmith : ). More than thirty years ago, I published a commentary ontype-variety, laying out my views of what “ware” should be and mean (Rice

). I believed then, and still believe, that identifying wares by both pastecomposition and surface treatment is illogical and confusing. In my view, forMaya pottery, at least, wares should be reconceived specically as “paste wares”

that refer specically to paste characteristics alone, rather than a combinationof paste and surface. One reason is that I always have been suspicious of a too-heavy relianceon slip characteristics, particularly color variations. Slip colors can vary enor-mously on individual vessels red using bonre-type techniques, as well asby conditions of use and depositional environment (which may completely


37/311

Prudence M. Rice

remove slips), so that it is difficult to know whether different colors were in-tentionally achieved or accidental. Paste, however, is a characteristic of all sherds and pots, slipped and un-

slipped, no matter their size or degree of preservation. Paste composition alsorepresents choices of clay and temper resources, primarily locally available,that allow insights into areas of manufacture and patterns of distribution. Ad-ams ( : ) seems to turn this into a detriment in noting that surfacetreatment is “most culturally dened and inuenced,” whereas paste (ostensi-bly by negative contrast) is “greatly dependent on the clay and trace elementsavailable locally and therefore is partially a product of natural elements thato en were unknown or controlled in prehistory.” But that is precisely the point,albeit garbled: clay and temper resourcesare locally available, and theycan becharacterized in the recovered pottery, thereby providing a basis for insightsinto manufacturing areas as well as potters’ behavior. Adams ( : ) goes on to complain that “paste analysis is also timeconsuming and therefore the sample sizes are miniscule.” To the contrary,the attributes used for identifying paste wares in a eld lab are those visibleto the naked eye or with a X hand lens, including color; texture; fracture

characteristics; general kinds, quantities, sizes, and shapes of inclusions; andtactile qualities (for example, the sandpapery feel of volcanic ash-temperedwares and the silty/clayey texture of Postclassic Snail-Inclusion Paste Ware).

ese rough eld sorts can later be tested and rened by mineralogical andthen chemical analyses, which permit insight into potters’ decisions (Cecil

b, , ). Potters make a cascade of decisions, or “technologicalchoices,” in selecting and manipulating resources on the basis of the desir-ability of their properties in forming, drying, decorating, ring, and using

the nal products (see Rice c: – ). e decisions may be the resultof experimentation or dictated by custom, but they can be revealed by rigor-ous research designs. For example, Leslie Cecil ( , a, b,

, ; Cecil and Neff ) used a combination of basic classicatoryand stylistic observations plus petrography, x-ray diffraction, SEM, ED-XRS,INAA, and LA-ICP-MS to characterize the pastes and pigments of Postclassicslipped and decorated pottery from sites in the Petén Lakes region (see alsoCecil, this volume). True, these latter studies are time-consuming and expensive and thereforeare performed on only small samples, which must be judiciously selected toanswer specic questions about the meaning of detected variability. ey areproperly carried out as the nal step of analysis: basic eld lab identicationsof paste wares are not only useful in and of themselves but also are essentialpreliminaries to interpretively sound compositional studies.


38/311


Case Study: A Petén Postclassic Slipped Pottery “System”

Postclassic slipped and decorated pottery at sites in the Petén Lakes region

was manufactured of three distinctive paste wares: coarse red-orange carbon-ate (Vitzil Orange-red Ware; A. Chase ), silty gray-to-brown with snailinclusions (Snail-Inclusion Paste Ware; Cowgill ), and marly “white” orcream (Clemencia Cream Paste Ware; Rice ). ese indicate minimallythree production nodes of Petén Postclassic pottery, at least one for eachpaste. We do not know exactly where those production nodes were, althoughit is fairly certain that Clemencia Cream Paste Ware pottery was manufac-tured from clays in the vicinity of the Topoxté Islands and Lake Yaxhá (Cecil

). Snail-Inclusion Paste Ware pottery incorporates a variety of freshwateraquatic snails and hematite lumps, apparently naturally present in these lacus-trine clays (Rice b: ), and it exhibits sufficient variability in both pastesand pigments that there were probably multiple centers of manufacture in thelakes region (Cecil a, b; Cecil and Neff ). ese paste wares wereused to manufacture the three most common red-slipped ceramic groups ofthe Postclassic: Augustine (Vitzil), Paxcamán (Snail-Inclusion), and Topoxté(Clemencia Cream). Besides these common red-slipped groups, we recoveredsmall quantities of pottery with other slip colors—“pink,” black, brown, andpurplish red. ere is a striking coherence in the “design structure” (Friedrich ) ofthis pottery, evident in its layout: whether painted or incised, decoration ap-peared in circumferential bands on the interior walls of tripod dishes (Rice

, , ). e bands are dened above and below by multiple linesand divided into two or four panels by vertical lines or placement of simple

motifs, which also may appear inside the panels (see Rice and Cecil ).is decorative coherence structures the classicatory coherence of types and

varieties in the slipped ceramic groups across the three paste wares: mono-chrome, black-painted, red-painted, and incised (table . ). In other words,potters across the Petén Lakes area had a fairly uniform and widely shared setof ideas about what constituted proper pottery decoration regardless of theirdifferent clay and temper resources.

By the Late Postclassic period, some signicant variations in this schemeare noted: the basic structure continued at the ware and group level, butinnovations occurred at the type and variety level. For example, instead ofred or black painted decoration, there began to be red-and -black painteddecoration. In addition, elaborate decoration began to be placed over theentire interior of tripod dishes, rather than restricted to encircling bands. Bythe very Late Postclassic period, still more decorative freedom can be noted


39/311

Prudence M. Rice

Table 2.2. Parallel distributions of decorative types and varieties within Postclassic Peténpaste wares

Snail-inclusion Cream Red Uapake

Group:Paxcamán

(Red)Fulano(Black)

Trapeche(Pink)

Macchiato(Brown)

Topoxté(Red)

Augustine(Red) Chipotle

Types

Monochrome X X X X X X (2) X

Polychrome

Black paint X X X X X X

Red paint

Banded X X X XCurvilinear X X

Geometric X X

Red + black X (2) X X X

Incised X X X X X

Fine X X X X X

Groove X X X (2) X X

Incised + paint X

Censer X X X

as greater variation and combinations of possible choices came into use: forexample, some vessels exhibit both paintingand incising, or interiorand exterior decoration, and so on. is pattern conforms well to what was called in Southwestern pottery anal-ysis a “ceramic system” (WGW : – ; see also Colton’s [ ] “principleof analogous pottery types”) and was introduced to lowland Maya type-vari-ety practice by Gifford ( ) in discussing the classication of pottery fromBarton Ramie. Gifford described it as having relatively little time depth: “anessentially horizontal or very shallow diagonal arrangement of roughly con-temporaneous pottery types that range over a wide area and that are related toone another in particular from the standpoint of decorative treatment, design

styles, and surface manipulation” and that crosscut wares ( : ). Gifford’sexamples included a possible Usulután system (Gifford : ) and an Agua-cate system (Gifford : , ), but he did not actually apply the concept tothe Barton Ramie collection. e ceramic system concept has been little usedin Petén, although it was incorporated into Honduran classication schemes(Henderson and Agurcia ; see also Bill, this volume).


40/311


What does a ceramic system “mean”? Most obviously, it means that Post-classic Petén potters recognized and adhered to a decorative canon—a set ofrules—for design structure, layout, and colors. Vessel forms and proportions

varied from ware to ware and settlement to settlement, but the principlesthat structured what kinds of decoration were to be applied, and where andhow, were strictly adhered to in at least three production resource-groups(minimally dened by paste compositions) throughout the lakes area (ap-proximately seventy-ve kilometers east–west) over three hundred years.In addition, we were eventually able to successfully predict the decorativetypes we would nd (which would otherwise be a process of categorization)in each ceramic group. is is the difference between a classication struc-tured by conceptual categories, as opposed to phenomenological groupingsor modal analysis. is Petén Postclassic ceramic system—based on the traditional rules forslipped pottery decoration—appears to have broken down in the very LatePostclassic period, with a surge in decorative variability, and a new paste wareappeared that was used not just for slipped pottery but also for unslipped ves-sels and censers. Signicantly, the Late Postclassic in the lowlands was a time

of population movements, in-migration of new groups, and general culturaldisruptions. Some of the pottery appears to have declined in certain techno-logical variables (quality of nishing, and incomplete oxidation), as is com-monly noted in times of societal stress. e increase in decorative variety andfreedom, plus some new forms, in Petén Postclassic pottery also coincidedwith an increase in hostilities among the Petén populations and vis-à-vis theimmigrants, accompanied by what we believe was the assertion of distinctethno-linguistic or social identities, Kowoj versus Itza, in this regional context

(Rice and Cecil ).

Type-Variety: What Works and What Doesn’t Work?

e type-variety system establishes a series of concepts and principles for or-dering attributes and attribute clusters, and this means it can be universallyapplicable for purposes of organizing descriptions. Proponents of type-variety,

of which I am one, highlight its advantages for quickly and efficiently handlinglarge quantities of pottery fragments (Adams : ) and its usefulness in in-tersite comparisons, especially with the growth of regional and settlement pat-tern studies since the s (Sabloff and Smith : – ). Many users havenoted the exibility and versatility of its hierarchical structure, which allowscomparative analyses at different levels of variability. Sabloff ( : ) noted its


41/311

Prudence M. Rice

strengths in both categorization and identication: “[I]t is easy to work with inanalyzing ceramics from a previously unstudied site.” Mayanists’ dissatisfactions with the type-variety system have abounded

(for example, Ball b, b; Culbert and Rands ; Hammond ;Smith ), and several criticisms recur: it leaves a lot of variability unac-counted for, it should be accompanied by modal analysis, and it has limiteduse with whole vessels. For my part, I have long been concerned about thelegacy of Gifford’s mapping mentalistic meanings and values onto what aresimply descriptive classes. For example, he (Gifford : , emphasis in origi-nal) interprets

the ceramic variety as a reection ofovert individual and small groupbehavior (Gifford ), the pottery type as reecting the interplay ofboth covert individualness andcovert culturalness, and the ceramic groupas reectingovert culture . . . [such] that the ceramic group is telling us ofthe everydayceramic activities of a culture; that the pottery type is tellingus of the subconscious ceramic value orientations of both theculture andthe individual ; and that the ceramic variety is an expression through pot-tery of individual or small social group ceramic preferences.

Before addressing these critiques, let us remember that the type-variety sys-tem of pottery classication used by Mayanists today had its ultimate origins inthe southwestern United States in the s to meet the needs of the emergingdiscipline of Americanist archaeology for improvements in description, no-menclature, chronology, regional integration, and culture change. Its conceptsand procedures were adopted into the southeastern United States and fromthere into Maya archaeology in the s. ese beginnings are important be-

cause they help explain the how and why of type-variety practice and some ofthe problems perceived by Mayanists. For example, in both the Southwest and the Southeast, the early emphasiswas on surface nish. Similarly, in the Maya area, variables of surface nish anddecoration were deemed most important and therefore prescribed as the basisfor the rst sorting, because surface nish was held to be the “most susceptibleof ceramic attributes to culturally determined change” (Adams : , :

). Regardless of the veracity of this assertion, mitigating preservation issuesmust be remembered: Southwestern pottery enjoys excellent preservation ofits complex polychrome decoration because of the region’s dry environment,and in the humid Southeast the surface nish is usually plastic manipulation,which also preserves well. However pottery from the tropical forests of thesouthern Maya lowlands o en has eroded surfaces, making it difficult to sortreliably by surface treatment.


42/311


In addition, type-variety was developed in the Southwest to sort, classify,and organize descriptions of large volumes of pottery fragments. Many Mayaarchaeologists who work with whole vessels, especially complex Classic picto-

rial polychromes, from burials or other contexts nd the type-variety systemunworkable and therefore dismiss its usefulness (Ball ; D. Chase and A.Chase : – ; Pendergast : , ). But in its original formulationtype-variety analysis was not intended for whole pots. Only with the “newarchaeology” and other developments did the logical inconsistencies betweenstudying sherds as proxies for pots come to be addressed. As Willey and Sabl-off ( : ) commented about the Southwest, potsherds were a convenient“statistical unit highly adapted for counting and manipulation,” but classica-tions had “grown steadily farther away from the whole pot.” Concerns about“the Indian behind the artifact,” as it was sometimes crudely put, promptedGifford ( : ) to assert that “in conceptual substance ‘type-variety’ is basedon whole vessels as opposed to broken pieces (sherds). eoretical implica-tions adduced from varieties and types are founded on the realization that theproduction of whole vessels was the intent of the prehistoric potters.” But thisis revisionist history. Today, except for comparative exercises, there is generally

little need for or value added by classifying whole vessels through type-variety’sformulaic organizational principles designed to create condensed descriptions. e most recent critique of the type-variety system has come from Culbertand Robert L. Rands, whose focal complaint is that the resultant classica-tions do not take into consideration many variables, resulting in “the loss ofimportant information” ( : ). Butall classications result in loss of in-formation: it is theoretically and practically impossible for any classicationto take into consideration all possible attribute states of an entity (see Dunnell

b: , ). e fundamental issue is to specify initially what informationis “important,” that importance dictated by the underlying goal or purpose ofthe classicatory exercise. According to the basic precepts of systematics, classes of whatever ob- ject of interest are supposed to represent something “inherently signicant”in the objects as indicated by the goal of high within-group similarity. It isgenerally agreed that classications structure inquiry by furnishing a sys-tem for describing and naming the objects of study, fostering communica-tion through shared t

Documents

Ancient Maya Pottery Classifi