UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Ceramic variability of Shang society at Huanbei in Anyang, China Fong, Denise Catalina 2008

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata


ubc_2008_fall_fong_denise_catalina.pdf [ 1.31MB ]
JSON: 1.0058374.json
JSON-LD: 1.0058374+ld.json
RDF/XML (Pretty): 1.0058374.xml
RDF/JSON: 1.0058374+rdf.json
Turtle: 1.0058374+rdf-turtle.txt
N-Triples: 1.0058374+rdf-ntriples.txt
Original Record: 1.0058374 +original-record.json
Full Text

Full Text

CERAMIC VARIABILITY OF SHANG SOCIETY AT HUANBEIIN ANYANG, CHINAbyDENISE CATALINA FONGB.A., The University of British Columbia, 2003A THESIS SUBMITTED IN PARTIAL FULFILLMENTOFTHE REQUIREMENTS FOR THE DEGREEOFMASTER OF ARTSinTHE FACULTY OF GRADUATE STUDIES(Anthropology)THE UNIVERSITY OF BRITISH COLUMBIA(Vancouver)August2008© Denise Catalina Fong, 2008ABSTRACTThe study of ceramic variability in Chinesearchaeology is conventionally understoodinthe context of temporal and regional differences,where emphasis is placed on explainingvariability in terms of identifying regionalstyles and long-term changes. Inthis thesis, Iexamine ceramic variability of Shang pottery betweentwo contiguous daily-use contextsatHuanbei, a Middle Shang period (1400-1250BCE) site located in the CentralPlains of China.Based on the analysis of pottery sherds collected fromdaily-use contexts at Hanwangdu(HWD)and Huyuanzhuang (HYZ), I argue that ceramicscollected within a single-site contextcan behighly varied and distinct due to differencesin use-context. Assemblage differences andceramicvariation are evaluated according torim sherd attributes including vessel shape, rimand lipshape, dimensional properties, and surfacetreatment styles. Possible interpretivemodels forexplaining observed patterns of variability arepresented.Results of this study suggest that siginificantvariability in pottery vessel designcan beobserved in the samples examined from the Huanbeisite. Consumers from the HWD(a palacecontext) consumed a greater variety of pottery vesseltypes but with a more limited rangeofshapes and decorative styles. In contrast, consumersfrom HYZ (a non-palatial context)consumed a limited range of pottery vesseltypes but with a greater variability in the rangeofshapes and decoration. The observed patternsof variation reinforce current assumptionsregarding the contextual differences betweenHWD and HYZ, and also provide newinsight intothe differential pottery consumption patternsby different social classes at Huanbei.Results ofthis study indicate the potential valueof studying intra-site ceramic variation in Chinesearchaeology and its importance in creatingnew knowledge on the material consumptionbehaviorof different social classes.11TABLE OF CONTENTSAbstract.11Table of ContentsiiiList of TablesivList of FiguresvAcknowledgmentsivIntroduction1Ceramic Variability in Shang Society3Huanbei8Chronology9Hanwangdu and Huayuanzhuang12Fieldwork, Methodology, and Dataset17Data Analysis23Pottery Vessel Types at HWD and HYZ23Form Variability25Metric Variability32Decorative! Stylistic Variability34Discussion and Conclusion36Bibliography42Appendix A Rim Shape Variations46Appendix B Surface Treatment Variations49Appendix C Surface Treatment Combinationsby Vessel Type and Form 52111LIST OF TABLESTable 1 Vessel Types at HWD andHYZ24Table 2 Complete dataset24Table 3 Percentages of pottery subtypes for guan jarsand pen basins 26Table 4 Rim/lip combinations by vessel type andform 29Table 5 Metric values for rim sherds at HWD and HYZ based onvessel type 33ivLIST OF FIGURESFigure 1 The location of the HuanbeiShang city8Figure 2 Shang chronology based onC-14 dating12Figure 3 Map showing the proximity of HWD andHYZ at Huanbei 14Figure 4 Map indicating layout of features at HYZ14Figure 5 The distribution of HWD building remainsalong the section-cleaning ditch 15Figure 6 Diameter template21Figure 7 Vessel types identified at HWD and HYZ27VACKNOWLEDGEMENTSThere are many individuals who I am indebted to thoughout thisentire graduate studentexperience. First and foremost, I thank my thesis supervisor, Dr.Zhichun Jing, who gave me theopportunity to explore the archaeology of ancient China.I also thank my thesis committeemember Dr. Michael Blake, who always made learning archaeologyjust that much moreinteresting throughout all these years at UBC. I express my sinceregratitude to Dr. Jigen Tang,who generously accommodated me at the Anyang field station to conductmy research on theHuanbei materials. I also would like to thank many individuals whotaught me lessons aboutarchaeology and beyond. Dr. David Pokotylo, a verygood statistics teacher who helped me toget my foot in the door. Dr. Kenichi Takashima, who gaveme the entry ticket into the world oforacle bone studies. Dr. Brian Chisholm, who has always been just agreat friend and mentor.Professor Fang Hui from Shandong University, China, for sharingwith me his profoundknowledge of Chinese archaeology. Funding for my research waskindly supported by ProfessorJing’s Social Sciences and Humanities Research Councilgrant, the ACLS/Henry LuceFoundation, and the Chinese Government Scholarship Council. I alsothank Patricia Omerod,who was always willing to help out during needy times.Throughout the past few years I met numerous individualsfrom around the worldbecause of our common interest in archaeology. Be it a short chatin the hallway or livingtogether out in the field, I gained friendships, insightson life, and academic knowledge that havehelped me grow tremendously as an individual and a graduate student.I thank all of you whohave been a part of my graduate student experience. Iespecially thank the wonderful crew whohelped out during the summer of 2004 at the Anyang workstation — ZhangHua, Patty Telford,Christina Neele, Shauna Janz, and Meng Ying who offered me tremendoushelp in measuring,organizing, and photographing my samples. I also thankSue Formosa, my very wonderful labbuddy, who saw me throughout the years and was always there to givesupport and motivation.Finally, I send my deepest gratitude to my family and friendsfor supporting me in this unusualventure. My biggest thanks go out to my family who hassupported me in so many ways. To mydear friends who have stuck beside me throughout theyears, thanks for all your encouragement.And to Meng who was always there through all the ups and downs inthe past few years, mysincere gratitude to you.viINTRODUCTIONArchaeological and historical research since the1 920s have demonstrated that the Shang(1600-1040 B.C.) was a state-level society that exercised considerablecultural influence in theCentral Plains region during the early Bronze Age (Bagley1999; Chang 1980; Shaughnessy1989; Thorp 2006; Trigger 1999; Yates 1994). Pottery vessels areamong the most commonmaterial correlates of the Shang culture and have longbeen relied upon as an index for studyingregional interaction and long term social change. Bystudying patterns of inter-site ceramicvariability, archaeologists have gained much insight intothe regional distribution of ceramicstyles and their evolution over time (Yan 1997). Incomparison, little is known of ceramicvariability within a single site context during a single timeperiod. Recent studies have shownthat patterns of variability encode information regardingvarious aspects of social behavior suchas strategies of production and consumption/use behavior(Masson and Rosenswig 2005; Rice1987). The objective of this study is to examine attribute variability ofpottery vessels to addressquestions of variability and consumption during the Shangperiod.This study was conducted in the summer of 2004 duringmy visit to the Anyang fieldstation located in the city of Anyang in the Henan provinceof China. A large collection ofpottery from the Huanbei site b1L (1400-1250 BCE),by far the largest Shang-period urbansettlement discovered in the central plain of China (Jinget al. 2004:6), had been recentlyexcavated and was stored at the field station. Thisparticular pottery collection included potterycollected from the Hanwangdu locus (hereafterHWD) during the 2001 excavation and theHuayuanzhuang locus (hereafter HYZ) duringthe 1999 excavation. The two loci are locatedonekm. apart within the parameters of the Huanbei site, andoffer a unique context for discussiondue to their distinct use-functions: HWD was likely designatedas the palace sector while theneighboring HYZ locus was a non-palatial residentialsector for local elites. Comparisons drawnfrom pottery sherds excavated from these two distinctcontexts are meant to reveal unique1patterns of consumption associated with different socialclasses and activities. In this study,emphasis is placed on studying daily-use ceramics throughthe analysis of pottery rim sherdscollected from a series of refuse pits associated with severallarge building foundations in eachcontext. New information on the excavated pottery assemblagesfrom Huanbei is presented here,followed by qualitative and quantitative studies of attribute variationof specific pottery vesselforms from HWD and HYZ. The analysis of attribute variationcan provide insight into the kindsof pottery produced for and consumed at the two contexts. Thisstudy also aims to explore newapproaches to explain variability that can improve ourunderstanding of differences between theHWD and HYZ ceramics and provide new insight into Shang potteryconsumption andvariability.The focus of this paper is to build upon current understandingof Shang ceramics in threeways: 1) identify and compare major characteristics of the HWD andHYZ pottery assemblage, 2)identify and compare patterns of variability among serving, cooking,and storage pottery vesselsfrom HWD and HYZ, and 3) consider interpretive modelsto explain the observed patterns ofdifferences and variability. This thesis begins with a briefdiscussion on the utility of ceramicvariability analysis, and relevant approaches to examining ceramicvariability in archaeologicalcontexts. Focus is placed on the evaluation current approaches toexplaining ceramic variabilityin Chinese archaeology. Background information onthe Huanbei site is introduced, including itschronological and geographical significance. Next, researchmethods are presented, including adescription of fieldwork, introduction to the dataset,and specific quantitative and qualitativeanalysis methods used. The results of my analyses, discussedin terms of vessel shape variation,rim/lip variation, dimensional variation, and decorative/stylisticvariation are presented.Implications behind the observed patternsof ceramic variability in Shang society basedon theHuanbei case are discussed.2CERAMIC VARIABILITY IN SHANG SOCIETYDiscussions on ceramic variability in Shangarchaeology have long been underminedbythe assumption that ceramics from the Shang periodwould have been highly standardizedresulting from controlled and centralizedproduction. As discussed by Liu Li and ChenXingcan(2007), “ceramic utensils were produced at themajor centre as well as at each regionalcentre,and then distributed to the entire population, includingcommoners in the immediate surroundingareas.. .It can only be generally argued that the productionand distribution of ceramics utensilswas carried out at regional and subregionallevels” (Liu and Chen 2007:137). Underhill(2002:202) suggests that ceramic production was eithercontrolled by political elitesorindependent producers. At the local level, there wouldbe considerable division of labour forceramic production such that artisans may have specializedin different kinds of pastes and evenfunctional types of vessels (Underhill 2002:256).Based on oracle bone and textual data, shealsopoints out the possibility that controlled specialized producedwas practiced in such a way that“potters who belonged to one descent group made domesticpottery vessels for the late Shangroyal household” (Thid:256).Ceramic variability has often been discussed interms of broad-scale questions related tothe culture-history of Bronze Age China.While considerable ceramic variability hasbeenobserved in Shang pottery vessels (Zhongguo1994), observed patterns of variabilityare oftende-emphasized in the interest of identifyingtypical ceramic styles for understandingculturalhistorical development (Yan 1997:56-57).The method of leixingxuebiaoxingxue,5Eçpottery seriation, grounded in principles of culture-historyand stratigraphy, relies on theidentification of typical vessels uniqueto specific temporal and spatial units to identifychangeover time (ibid:59). Specifically, this method assumes thatcertain vessel types or surfacetreatments that exhibit rapid change or existfor only a short period of time can be identifiedasrepresenting different temporal or regionalstyles (Ibid:59). A common practice for classifying3Shang pottery is to sort the sherds according to so-called“typical” and “non-typical” styles,which is frequently used by local archaeologists anddiscussed in excavation reports (see Anyang1994). The criteria used to differentiate between “typical”and “non-typical” styles aredeveloped based on the following attributes, vessel shape,rim shape, lip shape, and surfacetreatment combinations. These are commonly associatedwith the local region and time period.Vessel type and surface treatment variation is generally subsumedunder general formscategories to emphasize the most typical characteristics. Less is knownabout how to interpretpatterns of ceramic variability to examine ceramic productionand consumption in site-basedcontexts.Research from other areas has shown that ceramic variabilitygenerally encodesinformation regarding an array of social processes notsimply related to large-scale culturalhistory developments (Gosselain 1992, 1999; Lemonnier1986; 1993; Rice 1989; van der Leeuw2002). These include: 1) strategies of productionrelated to the process of variety generation;and2) consumption or use behavior involving the process ofvariety selection (Rice 1987:201). Thegeneral premise behind this inferential process is the ideathat variability is often reduced andregularized in different production modes, and thereforeits study can provide informationregarding production organization and use (ibid:20 1).In her essay on ceramic diversity,Prudence Rice (1989:113) argues that variability examinedthrough pottery attribute systems canreflect different aspects of production and consumption.Following Rice (1989:113), at least four pottery attributesystems can be identified tostudy variability: resources, manufacturing technology,form, and decoration! style. Variabilityinresources (ie. clay preparation) provides information aboutgeological variability of the localenvironment and the availabilityof clay and temper sources. At the same time, it isalsoindicative of a culturally-defined ‘ceramic environment’defined by potters’ knowledge of andpreference for the resources potentially available tothem. Variability in technological attributes4systems (forming and firing) are generally related to the level of technologicaldevelopment, andthe degree of standardization andlor mass production inthe manufacturing process. Variabilityin primary forms (vessel shapes such as jar, bowl, etc.) is significant interms of functions andactivities among the archaeological contexts being compared, whilevariability in secondaryforms (lip or base variations, vessel dimensions, appendages)is more likely to inform onproduction standardization and specialization. Lastly, variabilityin decorative or stylisticattribute systems generally provides information regarding contextof use of vessels or reflectscontact between local and non-local potters (ibid: 113).The chaIne opératoire approach discussed by OlivierGosselain (1992; 2000) has offeredsimilar insight into the source and meaning of variability. It emphasizesan understanding ofceramic attribute variation in relation to production choices, and arguesthat variation essentiallyreflects the dynamics between the control of productiontechnologies and its use in thenegotiation of social identities among different categoriesof social group membership(Gosselain 2000:208). Resource variability pertains togeological variability of the localenvironment, the availability of clay and temper sources,as well as the selection of resources bypotters. Based on the chaIne opératoire approach, resourceand technological attributes are botheasily transmissible through postlearning interactions among membersinvolved in the potterymaking process, and tends to reflect technological variationsover large span of time and space.Gosselain (2000) argues that each ofthe three major stages of ceramics’ manufacture isshaped by various social and technical factors. The firstcategory involves techniques that leavevisible evidence on the finished products, includingcertain processing techniques (e.g.,tempering or mixing clays to modify texture or color),“preforming”, “secondary forming”,decoration, certain firing techniques (e.g. smudging),and most postfiring treatments. Dueto thevisibility and technical malleability of these techniques,they tend to be “easily transmissiblethrough postlearning interactions and should displaya tendency to fluctuate through time andto5be transmitted widely across space to reflect more superficial, situational,and temporary facetsof identity” (ibid: 191).The second category involves clay selection, extraction, processing,and firing. Theseprocesses are also technically malleable, but because their effectsare not as visible as theprevious category of techniques potters’ choices are mostly influencedby others involved in thepottery making process. Both categories are largely influenced by thetransmission of ideas andtechniques among members involved in the pottery productionprocess. It is particularly usefulin reflecting variation over large regional areas or changesover long periods of times (ibid: 191).The third category involves primary forming, or “roughingout”. These stages are themost resistant to change due to the limited visibilityand limitation through motor habits(2000:191-192). There are countless ways of shaping ceramics whichare not dictated by externalrestrictions such as the availability of raw materials or functionalmorphology. Gosselain(1992:572) argues that the choice of shaping techniquesis largely cultural-based since limitedvisibility and motor habits limitations make primary forming largelyresistant to change. Amongthe potters interviewed by Gosselain, most were convincedthat there are no possible alternativesto their shaping technique; it was rather a technological tradition whosehistorical backgroundwas unclear but was consistently carried on as a tradition(Gosselain 1992:572). The shapingprocess appears to be influenced by the potter’s ethnicitydefined in terms of linguisticboundaries (Ibid:582). For this reason, he argues thatthe shaping process is a particularly reliableindex for understanding cultural diversity.In terms of decorative attributes, potters claim that “decorationusually follows personalfancy of the maker or is produced to the requirementsof the customers” (Gosselain 1992:574). Itis generally acknowledged that the finished surfaces of ceramicsvessels have a multitude offunctions beyond pure decoration (Rice 1987:244). A conventionalview highlight their technicalfunctions which include reduction of permeability, improvement ofgrip during transportation,6and modification of thermal properties (Ibid:232). Morerecently, the role of visible ceramicattributes style in communication and information transfer;research on ceramic productiontechnology have demonstrated that attributes with high visibilitybest reflect superficial,situational, and temporary facets of identity (Gosselain 2000;Stark 1999). Since decoration andfinishing can be easily adopted or modified by any experienced potter,they are more sensitive toinnovation since the techniques can be easily copiedand reproduced. Consequently, they areoften vague in terms of their fashion and diffusion but are particularlyeffective mediums used insocial identity expression (Gosselain 1992:582). In addition,other scholars have also suggestedthat decorative/stylistic variability is potent information regardingthe expression of socialidentities in social interaction and in defining socialboundaries (Dunnell 1978; Sackett 1977,1985, 1986a, b; Wiessner 1983, 1984, 1985; Wobst 1977).7HUANBEIFigure 1 has been removed due tocopyrightrestrictions. The information removedis a mapindicating the locationof Huanbei Shang city(from Anyang 2004b:Figure 1).The data examined in this study were collectedfrom the Huanbei site, situated in theHuan river valley in the northern part of Henan provincein North China (Figure 1). Locatedinthe northern suburb of modern Anyang city,Huanbei sits on flat terrain surroundedby low-reliefhills to the north, the foothills of the TaihangMountains to the west, and broadalluvialfloodplains to the east and south (Anyang 2004c:6).Over two dozen pounded earth foundationshave been located within the siteincluding a large palace foundation (Thorp2006:131). Theconstruction of the palace-temple compound,which is by far “the largest Shangor even BronzeAge palace-temple compound ever excavatedto date”, implies some membersof society hadgained enough power to organize anddirect the labour and resources of others(Jing, et al.2004:10). It is also located immediatelynortheast of many late Shang periodsites within an areadesignated as Yinxu’, which later becamethe site of the first official state capitalof the Shangperiod.The discovery of the Huanbei sitein 1999 was one of the major outcomesof theThe name Yinxu corresponds to the term“the ruins of Yin” which derives fromChinese historical documents(Thorp 2006:120). The Yin people describedin these texts refer to the Shang people.8interdisciplinary collaborative project betweenthe University of Minnesota and the InstituteofArchaeology of the Chinese Academy of SocialSciences (Jing, et al. 2002). Its discoveryissignificant for two reasons: 1) the site was occupiedduring the middle Shang (ca. 1400-1250B.C.), a period which foreshadowed the riseof Yinxu, “the only major settlementthat canreadily claim status as a state capital”to date (Thorp 2006:214); therefore, it providesus with abetter understanding of the transition processfrom pre-state to state-level societiesin this region(Liu 2005:13); and 2) stratigraphic informationindicates the Huanbei site was abandonedafter a50-year occupation, which offers a fine-scale temporal contextfor discussing socialdifferentiation within a single site.ChronologyThe recent discovery of Huanbeihas been an important factorin shaping currentunderstanding of the Shang cultural chronology.Traditional definition of the Shangculturalchronology was mostly based on earlier archeologicalresearch on major Shang type-sitessuch asZhengzhou (discovered in the early1 950s) and Anyang (discovered in the late1 920s) (Bagley1999:3 3). The traditional view suggests thatthe Shang cultural chronology is composedof threeperiods: predynastic Shang early Shangand late Shang periods (Zou1980:95.-182). Prior to the discovery of Huanbei, scholarshad challenged this chronologysystem byciting evidence from art historical researchby Max Loehr (1953) (discussionon Loehr’s researchto follow); namely, it was argued thatthe transition from early to late Shang wasintersected byan intermediary period (Bagley1999; Thorp 1985).Max Loehr’s seminal workon the periodization of Shang bronzedecorative styles hasbeen particularly influential in the discussionof Shang’s cultural chronologybecause his workdemonstrated a gap in the development fromZhengzhou to Anyang bronze decorativestyles (seeLoehr 1953). By rigorously studying unprovenancedShang bronze vessels from various9museum collections, Loehr developed a chronologycharting the development of bronzedecorative styles through the seriation of surface decorationson ritual bronze vessels. Thedevelopment of decorative styles from theErligang to the Yinxu period was markedby fivemajor styles that reflected a “sustained and intensely self-consciousartistic development”(Thid: 155). Each style was characterized by a combinationof changes in vessel shapes,production technique, decorative layout, image content(motifs), forms of relief’ image-grounddistinction and draftsmanship (Ibid: 147-154). Five stylesthat represented the development ofbronze technology during the Shang period were defined:“Style I” and “Style II” have beenassociated with ErlitouJXia and Zhengzhou/Erligangsites; “Style III” is an elaboration of “StyleII”, which is characterized by modulated lines, taotie motifs and dragonmotifs. Early versionsof “Style III” vessels were found at Zhengzhou while itsfully developed form was later found inseveral other parts of China. “Style IV” and “Style V”vessels have been found in Anyang.With the subsequent discovery of new archaeologicalsites, Loehr’s seriation of bronzedecorative styles linked the Zhengzhou and Anyang sitesin a dynamic pattern of growth (Thorp1985:7). However, vessels from Anyang appeared “exceedinglyrich, varied in style, and farmore advanced in every way—casting, typology, anddecoration—than vessels assigned to suchEarly Shang sites as Zhengzhou and Panlongcheng,Hubei” (Thorp 1985:9). This influencedscholars to re-consider the Shang chronology (Bagley1999; Tang 1999; Thorp 1985), and ledsome to hypothesize that there was “a so-called transition period interveningbetween theZhengzhou and Anyang sites” (Bagley 1999: 150) to accountfor the development from EarlyShang styles to the mature and complicated Late Shangstyles. With the discovery of the largeurban settlement at Huanbei as well as about 200 small-scaleMiddle Shang period sitesthroughout different regions of China, the presence of suchan intermediary period has beenaffirmed by archaeological evidence (Tang 1999, 2002). Due to differentviews regarding thedefmition of Shang sites, this period has been labeledin several ways: Robert Bagley (1999)10refers to it as the Transition Period, RobertThorp (1985) uses the term Transitional Period,whileChinese scholar Tang Jigen (1999; 2002) refersto it as the Middle Shang period.Preliminary dating of Huanbei was achievedby the identification of diagnostic ceramicsherds collected from the Huanbei city walland the modem Hanwangdu and Huayuanzhuangvillages (Jing et al. 2004:25). Absolute datingof Huanbei was conducted through AMS andconventional ‘4C dating techniques in the datingof human and animal bone samples as partofthe Xia Shang Zhou chronology project (XiaShangZhou2000:51,71). The Shang chronologyofthe Central Plains includes three periods: theEarly Shang period (1600-1400 BCE), theMiddleShang period (1400-1250 BCE), and finally the Late Shangperiod (1250-1040 BCE) (Jing,et al.2002) (Figure 2). Results of absolute dating indicatethat the occupation of Huanbei (1400-1250BCE) fits chronologically between the height of occupationat Zhengzhou during the EarlyShang period and the royal occupation at Anyang duringthe Late Shang Period (Thorp2006:131). Therefore, Huanbei potentially offersimportant evidence supporting earlierfindingsregarding an intermediary transition phasebetween Early and Late Shang.11Figure 2 has been removed due tocopyright restrictions. Theinformation removed is a diagram ofthe Shang chronology based on C-14dating (from Jing 2002: Figure 8).Hanwangdu and HuayuanzhuangThe selection of samples for diversity or standardization comparisons requiresthat thesamples or populations being compared be linked through temporal continuity orgeographicalcontiguity (Rice 1989:112). The reason for this standard is to control extraneousvariables and toestablish a ‘background’ or baseline range of variability against which diversitycan bemeaningfully interpreted (Ibid: 112). In this study, HWD and HYZ providetwo ideal populationsfor assessing variability because 1) the two loci are located in close proximitywithin theHuanbei site; and 2) both of the ceramic assemblages are associated with the 50-yearoccupationof the Huanbei site. Moreover, HWD and HYZ represent two functionally distinctcontextslikely associated with different demographic features — a large palace sectorat HWD and a nonpalatial sector at HYZ.Based on excavated and surveyed sections at HYZ and HWD, the two contexts represent12areas of distinct layout and scale (Figure 3). At HYZ,four pounded-earth platforms of buildingfoundations (F 1 -F4), wall foundation trenches, post holesand associated stone plinths, refusepits, wells, and trenches, sacrificial pits as wellas burials have been excavated (Anyang 2004c)(Figure 4). HWD is “characterized by a concentration ofplatforms of building foundations madeof hard pounded earth, including at least two large-sizedbuildings with courtyards” located inthe center of the Huanbei settlement (Jing, et al. 2004:14).In addition, 30 densely distributedbuildings were discovered along a 600m section-cleaningditch north of the two large buildingfoundations (Anyang 2004b: 12) (Figure 5).Significant differences in the scale of HWD and HYZare demonstrated by a comparisonof building sizes. At the HWD locus, the largest building enclosure(F 1) measures 174 m eastwest by 90 m north-south, totaling 1.6 hectares (Anyang 2004b:16;2004c:296). Situated 27 mnorth of it is F2, another rectangular enclosurewhich is about 90 m by 70m (approximately 6300m2) in size based on intensive coring detection. In contrast,the largest rectangular enclosure (F2)at the HYZ locus measures 30.2 m by 4.1 m (approximately124 m2); the second largest Flcompound measures 16.5 m by 6.25 m (approximately103 m2). To put this difference intoperspective, the largest structure at HWD is over 150times larger than that at HYZ.13Figure 3 has been removed due to copyright restrictions. The informationremoved is a map showing the proximity of HWD andHYZ at Huanbei.The rectangular area enclosed by the dotted line in zoneIV is HYZ, whilethe area across zone V and VIII is HWD (from Anyang2004b:Figure 2).Figure 4 has been removed due to copyright restrictions. The informationremoved is a map indicating layout of features at Huayuanzhuang(from Anyang 2004c:Figure 3).14Figure 5 has been removed due to copyright restrictions. The infonnationremoved is a map indication the distribution of HWD building remains alongthe section-cleaning ditch (from Anyang 2004b: Figure 9).15As described in the excavation reports, the differences betweenconstruction quality ofthe HWD and HYZ building foundations suggests thatmuch labour effort was devoted to thechoice of building materials and the actual constructionof the HWD structures (Anyang2004b: 12; 2004c). Building remains at HWD were madewith high quality rammed earthshowing extremely high density and uniform material composition (Jing,et al. 2002:25). In theFl foundation at HWD, two types of pounded earth canbe distinguished: very dark gray clayand silt clay in the foundation trench; and yellow silt orclay silt above the foundation trench(Anyang 2004a:27). The pounded layers were relatively thinand consistent, measuring about 8-9 cm for the former and 12-15 cm for the latter. In comparison, the HYZF 1 foundation wascomposed of reddish-brown clay and yellow earth with small amountsof calcium carbonateinclusions (Anyang 2004c: 301). This platform was poundedin some areas only and ranged inthickness from 0.2m to 1.1 m. A large portion of the F 1 foundation utilizedthe platform of theneighbouring F2 building foundation, and the quality of the pounded earthvaried within differentsections of the platform (Ibid:301).Based on the layout and size of building foundationsand other features within these threezones, excavators developed preliminary interpretations about their nature:the HWD locus mayhave been the palatial sector of Huanbei (Anyang 2003a:16), and the HYZ locus may have beenthe non-palatial residential compound (Anyang 2004c:35).For these reasons, Huanbei offers aunique context for discussing ceramic variability during the Shang period:HWD appears to bethe palatial sector within the Huanbei site, which suggests that the materialcorrelates associatedwith F{WD may have been linked to public governmental or religiousactivities, as well ashousehold activities of the noble classes at the palatial context(Jing et al. 2004:15). On thecontrary, the settlement patterns of HYZ suggest a small-scaleresidential context that was likelynot associated with the royal family or any large-scale public activities(Anyang 2004c:35).16FIELDWORK, METHODOLOGY AND DATASETFor this research, I conducted fieldwork at the Anyang field station fromJune to July2004 based on previously excavated ceramics collectedfrom Huanbei during the 1999 fieldseason at HYZ and 2001 field season at HWD. The total data set includes1042 rim sherds (905valid cases2). Rim sherds were selected for analysis to aid in classificationand quantification,because rim sherds “provide the most information for assessing the sizeand shape of a vessel”necessary in vessel identification and size estimation (Orton, et al. 1993:l68).I particularlyfocus on pottery excavated from refuse pits because they were closely associated withthebuilding remains both spatially and temporally4.To aid data classificationand organizationduring and after the field season, formal and surface attributesof the samples were recordedusing field-notes, computer database, photography, hand drawing,and digital illustration.Fieldnotes contained detailed descriptions of the measurements and attributesof all potterysherds examined in the field; this data was initially entered intoan Excel database and later setup as an SPSS database to improve database management and aid in basic statisticalanalyses.Photographs of the pottery sherds were taken using a digital camera for visualreference, some ofwhich were later used in revising the classification scheme after the field season.Hand drawingsof the rim profile for selected sherds were done in the field by myself aswell as a local fieldtechnician to provide a reliable reference of rim shape. These hand drawingswere later scannedand traced using Adode Illustrator software in order to placethe rims on the same plane oforientation to aid comparison. Selected photographs and rim profiledrawings are presented inAppendix 1 and 2 in reference to rim shape and surface treatment classifications2Based on sherds larger than 5% of estimated vessel equivalent (EVEs) in vessel diameter.Discussion of EVEs iscontained in the methods section.Body sherds, while usually being the most abundant in archaeological ceramic collections,are usually harder towork with because of the difficulties in determining vessel shape and size (Orton,et al. 1993:223). Although basesherds are sometimes more accurate than rims in predicting the total number of vesselsin an assemblage, they maynot always provide information about the size and shape of the vessel if the sherd size istoo small (Ibid:223).The site was re-occupied during much later periods, therefore I only selected those pitsthat were associated withthe Shang period occupation.17At HWD, 371 individual rim sherds were collected from the refuse pits 112, H6,H7, H8,and 1110. A substantial amount of ceramics and artifacts were excavated fromthese refuse pits,which include ceramics dated to Phase II of the middle Shang period (Jing, etal. 2004:17). Withthe exception of pit H2, which is located near the foundations of Fl and F2, allother refuse pitsare located along the profile of section cleaning close to the north building cluster.The refusepits (including 112-H 10) at HWD were superimposed by level 4, a 0.2 - 0.4 m layer ofbrown oryellowish brown silt clay on the same strata as the platform of the HWD building foundations,which suggests that they were contemporaneous to the buildings (Anyang 2004b: 17; Jing,et al.2004:18).At HYZ, 605 individual rim sherds were collected from the refuse pits H24, J25,and J4.The pits are closely associated with the HYZ Fl and F2 buildings on the west sideof theexcavation unit. 1124 is cut into by the F2 west building foundation. J4 openson the same levelas the F 1 building foundation, and J2 cuts into the platform of the F 1 building foundation(Ibid:300). Both the buildings and the refuse pits are located below level 3, a Shang culturallayer;therefore, all three pits are closely associated with the HYZ building foundations.To evaluate the differences between the pottery assemblages from HWD and HYZ,Iclassify the sherd samples according to their respective vessel forms and calculate the ratiosofvessel types based on the number of individual sherds. Classification is guided by the Huanbeisite survey and excavation reports (see Anyang 2003a, b; Anyang 2004c) with the assistanceofProfessor Tang Jigen, director of the Anyang field station. The assistance of ProfessorTang inpottery rim classification was particularly crucial in this study because I had no prior experiencein ceramic analysis. To gain familiarity with the characteristics of Middle Shang pottery,Istudied the illustrations and descriptions of pottery excavated fromHWD and HYZ as describedin the excavation reports. In the field, Professor Tang assisted me in identifying variousceramicPits labeled as ‘H’ and ‘J’ both represent refuse pits in general, but ‘J’ specifically refers to waterwells.18attributes by sorting the pottery sherds based on type with me and describing to me the typicalpottery attributes for each vessel type. These attributes included general paste types, inclusiontypes, overall vessel size, rim shape, lip shape, surface color, and strategies of differentiatingamong various vessel types. Complete and refitted vessels were also studied for generalreference to help me understand the overall shape and possible functions for various vessel type.After sorting the sherds according to general vessel type, Professor Tang guided me in thedetailed classification of various styles for each type for pottery sample from HWD H2 and HYZH24. These two samples were selected by Professor Tang and Professor Jing to be mostrepresentative of the pottery assemblages from the HWD and HYZ contexts. Detailedclassification was conducted on these two samples based on Professor Tang’s assessment ofsignificant attribute differences in rim shape and vessel size, lip shape features, as well asvariations in surface color and inclusions, which is used to guide the following analysis.In addition to individual sherd count, the EVE (estimated vessel equivalent) measure isapplied as a supplementary method of sherd quantification, because sherds which are toofragmented may create a large sample compared to a small collection of large sherds (Gosselain2000:191-192). The term estimated-vessel equivalent (EVE) was originally coined by CliveOrton in 1975, based on the assumption that “all sherds in an assemblage that come from thesame vessel type a certain proportion of that vessel, that is they are equivalent to a certainfraction of it” (Orton 1993:179). The EVE method is determined by measuring a specific part ofa vessel such as the rim that is calculable as a fraction of a complete vessel (Ibid: 173).Specifically, EVEs are calculated “from the lengths (or percentages) of rims of a particular typeof pot, which then may be divided by the mean rim diameter” (Orton 1993:173). The fraction ofthe orifice represented by the sherd is then read on the radius of a rim diameter template whichcomes closest to the opposite end of the rim sherd (Ibid:352). The EVEs method produces anestimate of the number of complete vessels based on the curvature and completeness of the sherd19as a percentage of the total orifice circumferenceas well as an estimate of the rim diameterof thecomplete vessel (Egloff 1973:352). TheEVE method will aid in the evaluation ofthecomparability of the two datasets.Form variability is explored in moredetail through the identification of vesselform, andsecondary form characteristics, which includerim shape, lip shape, and vessel dimensions(Rice1989:113). As discussed earlier, variability in terms ofresources, technology, form,anddecorative/stylistic attributes can all be identified toinvestigate the structure of variability inpottery (Rice 1989:113). Variability in form attributesis measured by studying the primaryforms (shapes) of pottery vessels (Ibid: 113).This requires a basic sorting of pottery sherdsaccording to vessel form using classification terminologyspecifically relevant to Shang potteryfrom the local region (see Anyang 2004c).Form variability is further explored in termsofrim/lip combinations, which involvesproducing a list of rim shapes and lip shapecombinationsobserved for each pottery vessel type evaluated(Masson and Rosenswig 2005:369). FollowingMarilyn Masson and Robert Rosenswig (2005:369), variouscombinations of rim shape and lipshape attributes are identified and are thenlabeled according to terminology used intheexcavation report of the HYZ locus(Anyang 2004c). All possible rim and lipshapescombinations observed in the two contextsare identified and tabulated. Variability is evaluatedbased on the absolute and relative numbersof combinations for different subtypes ofpen basins,guan jars, and ii tripods from HWD H2and HYZ H24 refuse pits, which containlarge samplesrepresentative of each context.20Figure 6 has been removed due to copyright restrictions.Theinformation removed is a diagram ofa diameter template(from Egloff 1973:Figure 1).Metric variability is evaluated based on rimdiameter and rim thickness measurements(Figure 6). Rim diameter is measured using a diameter chartby measuring a vessel’s orificeagainst a graded series of concentric arcs separated at5% intervals (Rice 1987:292). The sherd isaligned on the arc of best fit with one edge of the rim at the0% radius and measurements weretaken along the interior edge of the rim sherd(Ibid:352). Two rim charts with 1cm. intervalswere used in this study, one that had a maximum diameterof 25 cm applicable to most sherdsand another with a maximum diameter of 40cmfor sherds with larger diameters. Sherds thathadless than 5% of the circumference intact hadto be ignored in the final tabulation as well sinceitwas too difficult to obtain an accurate size estimate from them. Sherdsthat were warped ordamaged were also omitted from the measurements butwere noted in the field records. Rimthickness is measured by using electronic calipers,and is consistently measured at the widestpart of the rim for each rim sherd.Variability is evaluated using coefficients of variation (hereafterCV), which can provideadditional insight regarding standardization (Massonand Rosenswig 2005:373). CV is definedasthe ratio of the standard deviation to the mean, and is reportedas a percentage by multiplying theabove ratio by 100 (Ibid:357). Calculating coefficientsof variation is a useful measure of21standardization by comparing metric variability between twosets of data; higher coefficientsimply less standardization, likely due to the involvement of greater numbersof specialists inpottery making, while lower coefficients suggest less dimensional variabilityand imply greaterstandardization (Masson and Rosenswig 2005:377).To assess decorative/stylistic variation, surface treatment attributes are observed andrecorded for comparison. Variability in surface treatment attributes is evaluated bycomparingsurface treatment combinations for different subtypes ofpen basins, guan jars,and ii tripods.Surface treatment attributes include decoration found on the body, neck, rim,lip, as well as anyattributes observed on the inner surface of the sherd. The description of attributes followsterminology used in the excavation report of the HYZ locus and includes moredetailedexplanation of the decorative attributes observed in the study sample (Anyang2004c).22DATA ANALYSISPottery Vessel Types at HWD and HYZThe following types of pottery vessels are observed at HWD and HYZ:dou-stemmeddish, gui-pedestal bowl,jia-tripod, pen-basin, guan-jar, weng-jar, zun-vase,li-tripod andjiangjunkui-vat (Table 1). Pen basins, ii tripods, guan jar, zun vases, gut bowls, doustemmeddishes are found in both contexts; thejiangjunkui vat only appearsat HWD, and thejia tripodonly appears at HYZ. Pen basins, ii tripods, guan jars are the most common vesselforms in bothcontexts according to both the number of rim sherds and the EVE calculation. Percentagesofvessel types per context calculated by individual sherds and EVE both show similarpatterns ofdistribution.Based on individual sherd counts, the HWD assemblage (H2, H6, H7, H8,HlO) consistsof 36% pen basins, 20% ii tripods and 18% guan jars. The remaining 26% of the assemblageconsists of five vessel forms, including: dou-stemmed dishes, gui-pedestalbowls, weng-jars, zunvases, andjiang/unkui-vats. The HYZ assemblage (H24, J2, 34) is characterizedby li tripods at37%, guan jars at 29%, andpen basins at 32%, which account for98% of the total assemblage.The remaining 2% of the assemblage consists of five vesselforms, including low percentages(l%) of dou-stemmed dishes, gui-pedestal bowls, jia tripods, weng-jars, andzun-vases.23Table 1. Vessel types at HWD and HYZTYPE HWD HYZN % EVE % N % EVE %Dou cup 3 0.9% 0.28 0.6% 3 0.5% 0.54 0.9%Guan jar 62 18.4% 9.94 22.2% 163 28.7% 18.8 31.6%Gui bowl 7 2.1% 0.95 2.1% 2 0.4% 0.21 0.4%Jia tripod 0 0.0% 0 0.0% 1 0.2% 0.06 0.1%Jiangjunkui vat 17 5.0% 2.1 4.7% 0 0.0%0 0.0%Li tripod 68 20.2% 10.7 23.9% 210 37.0% 24.3 41.0%Penbasin 120 35.6% 14.9 33.3% 179 31.5% 14.123.8%Weng jar 8 2.4% 0.97 2.2% 4 0.7% 0.85 1.4%Zunvase 52 15.4% 4.91 11.0% 6 1.1% 0.44 0.7%Total 337 100.0% 44.8 100.0% 568 100.0% 59.3 100.0%Table 2. Complete datasetTYPE N TOTAL EVETOTALHWD H2 H6 H7 H8 H10 N % H2 H6 H7H8 H10 EVE %Doucup 1 - 1 - 1 30.9% 0.09 - 0.08 - 0.11 0.28 0.6%Guanjar 29 5 16 1 11 62 18.4% 4.01 0.97 2.990.1 1.87 9.94 22.2%Gui bowl 3 1 2 - 1 7 2.1% 0.39 0.17 0.34 -0.05 0.95 2.1%Jia tripod - - - - - 0 0.0% - - - -- 0 0.0%Jiangjunkui vat 17 - - - - 17 5.0% 2.1 - - -- 2.1 4.7%Li tripod 51 4 7 2 4 68 20.2% 8.4 0.521.23 0.19 0.35 10.7 23.9%Pen basin 64 5 35 7 9 120 35.6% 7.2 0.78 5.22 10.72 14.9 33.3%Weng jar 5 3 - - - 8 2.4% 0.62 0.35 - -- 0.97 2.2%Zunvase 51 - 1 - - 52 15.4% 4.83 - 0.08 -- 4.91 11.0%TOTAL 221 18 62 10 26 337 100.0% 27.6 2.799.94 1.29 3.1 44.8 100.0%HYZ H24 J2 J4 N % H24 J2 J4 EVE%Dou cup 3 - - 3 0.5% 0.54 -- 0.54 0.9%Guanjar 125 14 24 163 28.7% 14.4 1.75 2.6418.8 31.6%Gui bowl 2 - - 2 0.4% 0.21 -- 0.21 0.4%Jia tripod - - 1 1 0.2%- - 0.06 0.06 0.1%Jiangjunkui vat - - - 0 0.0%- - - 0 0.0%Li tripod 155 21 34 210 37.0% 18.2 2.693.4 24.3 41.0%Penbasin 130 6 43 179 31.5% 10.3 0.553.3 14.1 23.8%Weng jar 2 - 2 40.7% 0.35 - 0.5 0.85 1.4%Zun vase 6 - - 6 1.1% 0.44 -- 0.44 0.7%TOTAL 423 41 104 568 100.0% 44.4 4.999.9 59.3 100.0%Calculations based on EVE estimation reveal similar patterns of distribution,suggestingthat the effects of sherd breakage did not have dramatic impact onthe relative percentages ofvessel types (Table 2). The HWD assemblage (H2, H6, H7, H8, H1O)is characterized by 34%pen basins, 23% ii tripods and 21% guan jars, which altogether accountfor 79% of the entireassemblage. The remaining 21% of the assemblage consists of five vesselforms, including: dou24stemmed dishes, gui pedestal bowls, weng jars, zun vases,andjiangjunkui vats. The HYZassemblage (1124, J2, J4) is characterized by ii tripods at 40%, guanjars at 31%, and pen basinsat 26%, which account for 96% of the total assemblage. The remaining4% of the assemblageconsists of five vessel forms, including: dou stemmed dishes,gui pedestal bowls andjia tripods,weng jars, and zun vases.Based on the relative percentages of vessel types from HWD and HYZ,the former ischaracterized by a more evenly distributed range of different pottery vesseltypes relative to thelatter. The major difference between the two contexts is reflected in highpercentages ofjiangjunkui vat and zun vase at HWD relative to the typicalpen basin, guan jar, and ii tripodcombination. At HWD, there are relatively low percentagesof other vessel types relative to penbasins, guan jars, and ii tripods. Second, at HWD, pen basins are the mostpopular, followed byii tripods and guan jars. At HYZ, ii tripods are the mostpopular, followed by guan jars andpenbasins. Differences in the relative percentages of these three most typicalvessel types suggesttwo different sets of consumption requirements were demanded by the consumersat HWD andHYZ. Both patterns reflect distinct differences in the consumption demandsand preferences ofthe consumers from HWD and HYZ. The significance of theseobserved patterns will be furtherelaborated in the discussion section.Form VariabilityAs often described in excavation reports (see example in Anyang 2004c),conventionalceramic typology in China addresses pottery vessel types ingross functional terms; labels suchas guan jars are often used when referring to groups ofpottery vessels interpreted as havingsimilar functions though distinct in shape. Appendix 1 shows illustrationsof typical vesselshapes per vessel type. In this study, variations in vessel shape are observedfor two vessel typesin the HWD and HYZ samples: pen basin and guan jar rimsherds (Table 2). Among them, guan25jars exhibit the most variability in the number of vessel shapes (n=3).The same number of vessel shape variations in guan jars is observed in both HWDandHYZ rim sherd samples. Three forms ofguan jars, include: 1) large jar with incurvedsides,vertical neck, and a flat base; 2) different forms of a small jar with recurved sides,outcurvedneck, and a flat base; and 3) small jar with outcurved neck, globular shape,and round base.While the number of varieties is the same, the percentages of these subtypesare different.Globular guan jars are the most popular at HWD, while large-sized guan jarsare the mostpopular at HYZ. In terms of guan jars, the distribution of each of the threeforms is fairly equalin terms of their relative percentages at both contexts. At HWD, the globular formpresents aslightly higher percentage compared to the other types, while the large-size formsare mostpopular at HYZ.Table 3. Percentages of pottery subtypes for guan jars and pen basins.TYPE SUBTYPES HWD H2 HYZ H24n % n %Guan jar Large-size (diameter 14-40 cm) 9 31.0% 49 39.2%Small-size (diameter 12-28 cm) 9 31.0% 39 31.2%Globular (diameter 10-18 cm) 11 37.9% 37 29,6%Total 29 100.0% 125 100.0%Pen basin Flat-base (diameter 20-28 cm) 44 68.8% 108 83.1%Round-base (diameter 26-42 cm) 20 3 1.3% 22 16.9%Total 64 100.0% 130 100.0%Two general forms ofpen basin are observed in both contexts, including: 1) basinwithincurved sides, outfiared neck and flat base; and 2) basin with flared sides and slightlyroundedbase. Both varieties appear in the HWD and HYZ pottery sherd samples, and inboth cases flat-based pen basins are more popular than round-based pen basins. In fact, based on the relativepercentages ofpen basins at HWD, the ratio of flat-based pen basins is significantlyhigher thanround-based basins. The same pattern is observed in the HYZ, where a largemajority of samplesare flat-based basin sherds.261) 2)3)’\J4)C6) 7) 8) 9)11) t 12)0I713)Figure 7. Vessel types identified at HWD and HYZl)Pen (Deep belly) (Underhill 2002:291-297,308); 2) Pen(Shallow belly) (Anyang 2004c:Figure18); 3) Guan (Anyang 2004c:Figure 18); 4) Guan (Anyang2004c:Figure 19); 5) Guan (Globular)(Anyang 2004c:Figure 19); 6) Li (Tang 1999:Figure 3); 7) Li(from Anyang 2004c:Figure 14); 8)Weng (from Anyang 2004c:Figure 14); 9) Gui (from Tang 1999:Figure 10);10)Dou (fromAnyang 2004c:Figure 17); 11) Zun (from Anyang 2004c:Figure20); 12) large-mouthed Zun(Tang 1999:Figure 2); 13) Jiangjunkui Vat (Anyang 2003a:Figure12).Variation is also observed in zun vases at HWD and HYZ. Two forms ofzunvase,including: 1) large-mouthed vase with wide outcurving rim,deep body, and round base; and 2)small vase with outcurving rim, recurved sides, and a round base. However, thelarge-mouthedstyle is observed only in the HWD sample, while the small vase is observed onlyin the HYZ27sample.Form variability is also examined in terms of the number of rim/lip combinationsfor penbasin, guan jar, and ii tripod sherd samples from HWD H2and HYZ H24 refuse pits (Table 4).At HWD, there are four rim/lip combinations for flat-basedpen basins and three for round-basedpen basins. At HYZ, there are five rim/lip combinations for flat-based pen basinsand four forround-based pen basins. In both contexts, flat-based pen basins display more variabilitythanround-base basins.Flat-based pen basins in both contexts are associated with doubleoutflaring rim, which ischaracterized by the outfiaring rim and wide rim edge which gives the rimthe appearance of twoeversion points. Four rim/lip combinations are shared by HWD and HYZ, the mostpopularcombination being the double outfiaring rim and thick square lip combination.Based on thenumber of rim/lip combinations observed at HWD (n4) and HYZ (n5), flat-basedbasins showrelatively greater variation at HYZ. One combination thatis found exclusively at HYZ context isthe double outfiaring rim and round lip style.Relative to flat-base pen basins, less variability is observed in round-base penbasins inboth contexts. At HWD, there are three rim/lip combinations, and at HYZthere are four rim/lipcombinations, which suggests greater variability in rim/lip combinations is observedin the HYZcontext. The most popular style at HWD 112 is the direct rim and square lip withgroovecombination, while at HYZ H24 the direct rim/square lip combinationis most popular. The wideoutflanng rim/square lip with groove combination is unique toHYZ H24.28Table 4. Rim/lip combinations by vessel type and formHWD HYZGuan (Large) Outfiaring rim Round lip0Outfiaring rim Thick squared lip 20Straight rim Squared lip2 5Straight rim Thick squared lip5 43Total N=3N=3Guan (Small) Outcurving rim Pointed lip0 2Outcurving rim Pointed lip wI exterior thickened0 9Outcurving rim Round lip0 12Outcurving rim Square lip0 7Outcurving rim Square lip wI groove0Outcurving rim Thick round lip0 3Outfiaring rim Square lip6 4Outfiaring rim Square lip wI groove0Straight rim Square lip3 0Total N=2N=8Guan (Globular) Outcurving rim Pointed lip6 2Outcurving rim Round lip 028Outcurving rim Square lip5 9TotalN=2 N=3Pen (Flat-base) Double outfiaring rim Pointed lip5 4Double outfiaring rim Round lip0 8Double outfiaring rim Square lip6 6Double outfiaring rim Square lip wI groove 1213Double outfiaring rim Thick square lip21 77Total N=4N=5Pen (Round-base) Direct rim Pointed lip1 1Direct rim Square lip wI groove14 2Direct rim Squared lip5 16Wide outfiaring rim Square lip w/ groove0 10Total N=3N=4LiOutfiaring rim Pointed lip0Outfiaring rim Round lip2 3Outfiaring rim Square lip3 30Outfiaring rim Square up wI double groove0Outflaring rim Square lip w/ groove9 26Outflaring rim Thick square lip3 6Outflaring rim Thick square lip wI groove0 6Outfiaring rim Thin square lip7 4Outfiaring rim Thin square lip wI groove0 9Pan-shaped rim Square lip 198Pan-shaped rim Square lip w/ groove8 61TotalN=7 N=1 129Large guan jar rim/lip combinations are equal in number at both HWD and HYZ. Thereare three rim/lip combinations for the large guan jar sherds in both contexts, whichreflect thesame degree of variability. Two combinations are common in both contexts: the straightrim andsquared lip, and straight rim and thick squared lip combinations. The most popularcombinationis the short straight rim with thick squared lip, followed by the straight rim/squaredlipcombination, where there is no thickening of the lip. Two combinations of outfiaring rimareunique to each context, including 1) outfiaring rim shape, round lip found only in HWD,or 2)outfiaring rim, thick squared lip found only in HYZ.Nine rim/lip combinations are observed for small-size guan jars, which makes it themostvaried among the three guan jar forms identified here. Two rim/lip combinationsare observed atHWD while eight combinations are observed in HYZ; therefore, small-sized guan jarsaresiginificantly more varied at HYZ than at HWD. Only one combination (outfiaring rim,squarelip) is shared by both contexts, while the other seven combinations are exclusiveto HYZ. Theseinclude various outcurving rim combinations associated with a variety of lip shape;nooutcurving rim combinations are observed in HWD. The most popular rim/lip combinationfound at HWD is the outfiaring rim and square lip, which appears to be a smaller versionof thelarge-size guan variety. The most popular rim/lip combination found at HYZis the outcurvingrim and round lip combination. The outfiaring rim and square lip with groovecombination isonly found in HYZ.Globular guan jars have the lowest number of rim/lip combinationsamong the three guanjar subtypes. At HWD, two combinations are observed, while at HYZ three combinationsareobserved. Globular guan jars are by far the only form ofguan that show more variationat HWDthan at HYZ. Rim shape is highly consistent for this vesseltype as globular guan jars are onlyassociated with the outcurving rim. Variations in the lip shape are observed,and based on thesamples examined the outcurving rim and pointed lip combinationis most popular at HWD30while the outcurving rim and round lip combination ismost popular at HYZ. Two combinations(outcurving rim and pointed lip, and outcurving rim andsquare lip) are common in both contexts.The outcurving rim and round lip combination is found exclusivelyat HYZ H24.At HWD, large size guan jars show more variability indicatedby the highest number ofrim/lip combinations among all three guan jar types in that context (n=3).Both small size andglobular guan jars have the same pattern of variability (n=2). At HYZ,small size guan jarsdisplay the most variability in terms of rim/lip combinations (n8), whilelarge size and globularjars both display relatively less variability (n=3). High variability isobserved for large guan jarsat HWD and small guan jars at HYZ. Consistent low variability in globularguan jars in bothcontexts suggests relatively greater standardization for this particular vessel subtypecompared toother guan jar forms.Li tripods exhibit significant variability in terms of the number of rim/lip combinations.A total of eleven combinations were present in both contexts combined,seven are present inHWD and eleven in HYZ. At HWD, the most popular combinationis the pan-shaped rim!square lip. The pan-shaped rim is a style unique to ii tripod and is characterizedby an outfiaringrim with slightly incurving edges. At HWD, the total numberofpan-shape rim sherds (n=27) isgreater than the number of outfiaring rim sherds (n=24), which suggeststhe pan-shaped rimshape is more popular at HWD. The number ofpan-shaped rim/lip combinations(n=2) is lowerthan outfiaring rim/lip combinations (n=5), which indicates highervariability in outfiaring rims.At HYZ, the opposite is true; the number ofpan-shaped rim sherds (n69)is lesser thanoutfiaring rim sherds (n86), which suggests that ii tripodswith outfiaring rims are generallymore popular. The number ofpan-shaped rim combinations (n=2)is significantly lower thanoutfiaring rim combinations (n=9), which indicates that outfiaringrim styles are more varied interms of the number of rim/lip combinations. However, the singlemost popular rim/lipcombination at HYZ is the pan-shaped rim! square lip with groove (n=61).31In examining patterns of variability in pen, guan, and ii pottery sherdsfrom HWD andHYZ, guan jars exhibit the most variability in terms of the numberof rim/lip combinations.Among the three types of vessels examined, pen basins arethe least varied in terms of thenumber of rim/lip combinations. These patterns suggestsignificant differences in the degrees ofvariability in rim/shape combinations between the twocontexts. Though the reported patternsmay be affected by sampling vagaries, the observed variabilitydemonstrates that variability ismore significant than expected both withinthe guan jar, pen basin, and ii tripod vesselsubtypes.Metric VariabilityMetric variability is assessed based onrim diameter and rim thickness measurements forpen basins (flat-based and round-based), guan jars (large,small, and globular), and ii jars toevaluate the relative degree of metric variability withinsingle vessel types6.Table 6 show themetric values calculated for these forms based on sherdsamples from HWD H2 and HYZ H24.Coefficients of variation (CV) values acrossall three categories were generally high,whichsuggests that ceramics at HWD and HYZ generally displayhigh variability in both contexts.Rim diameter measurements for flat-base penbasins display higher CV values relativetoround-base pen basins in both the HWD andHYZ samples, suggesting greater rim diametervariability in the former vessel type. Roundbase basins provide a particular useful measureofrelative variability due to relatively equalsample sizes in both contexts. CV valuesfor rimdiameter suggest that round base pen basinsare more standardized in HWD than in the HYZsample. Rim thickness measurements show similar patterns,such that flat-base pen basinspossess higher CV values in relation to round basepen basins in both samples. Again, rimthickness measurements indicate that roundbase pen basins from HWD are more standardizedLow sample size made it difficult to measure standardizationin other vessel types. For example, althoughthere isa sufficient sample ofzun vases at HWD, the samplefrom HYZ is too smaLl to produce a reliable indexforcomparison.32than those from HYZ.The table of metric values for rim diameter shows that small sizeguan jars from theHWD 112 sample have the lowest CV value among all typesofguan in both contexts. Theseresults may be influenced by the classification procedure,such that this particular vessel subtypeis defined by guan jars possessing small vessel size. Results contrastwith the high degree ofvariability observed in rim/lip combinations and surface treatmentcombinations observed forsmall guan jars in both contexts. In the HWD sample, globular guanjars have the second lowestCV value for rim diameter measurements, followedby large size guan jars. This suggests thatthe rim diameter of large size guan jars is most varied inthe HWD sample. In the HYZ sample,large size guan jars have the second lowest CV value,followed by globular size guan jars. Thisindicates that rim diameter measurements of HYZglobular guan jars display the highestvariability.Table 5. Metric values for rim sherds at HWD andHYZ based on vessel typeCONTEXT TYPE SUBTYPES Rim Diameter (cm)Rim Thickness (mm)N Mean SD CV N Mean SD CVHWD 112 Guan jar Large-sized9 29.5 5.2 17.6 9 8.4 0.7 8.5Small-sized 9 15.0 1.4 9.3 9 8.3 1.720.8Globular 11 15.0 2.1 14.0 118.4 0.8 9.5Lijar 51 17.2 4.7 27.3 516.3 1.4 22.2Pen basin Flat-based 44 33.0 7.2 21.8 448.4 2.0 23.8Round-based 20 37.0 5.0 13.5 209.6 1.5 15.6CONTEXT TYPE SUBTYPES Rim Diameter (cm)Rim Thickness (mm)N Mean SD CV N Mean SD CVHYZ H24 Guan jar Large-sized 49 23.7 4.719.8 49 8.7 2.1 24.1Small-sized 39 14.4 1.5 10.439 7.2 1.3 18.4Globular 37 16.5 3.3 20.0 377.2 1.1 15.3Li jar 155 18.9 5.7 30.2 1556.2 1.3 21.0Pen basin Flat-based 108 33.67.6 22.6 108 8.3 1.6 19.3Round-based 22 34.0 5.9 17.4 227.8 1.4 17.9In terms of rim thickness measurements, CV values for HWDguan jars show low CVvalues for large-size and globular jars relative to small-sizeguan jars. This pattern indicates thatrim thickness for large size guan is the least varied, whileglobular guan jars are the most varied33in the HWD sample. In the HYZ H24 sample, rim thickness measurementsof globular guan jarhave the lowest CV value, followed by small sizeguan jars and finally large size guan jars. Theobserved pattern contrasts with the HWD values in that globular guan jarsare the least varied interm of rim thickness while large guan jars are most varied. Smallsize guan jars have a higherCV value for rim thickness measurements even when the HWD samplesize is significantlysmaller than the HYZ sample, which suggests that the HYZ sample isrelatively standardizedcompared to HWD.Based on rim diameter measurements of ii tripods, high variabilityis observed in bothHWD and HYZ contexts. Rim thickness measurements produce similar patternsof variability assuggested by high CV values. Interestingly, while the ii jar sample size fromeach context isvery different, CV values for both rim diameter and rim thicknessare similar between HWD andHYZ. In fact, rim thickness measurements indicate a slightly higherCV value (greatervariability) for HWD ii tripods despite having only a third of the HYZsample size. This suggeststhat the rim thickness of li tripods from HYZ is more standardized thanthose from HWD.Decorative! Stylistic VariabilityVariation observed in decorative characteristics is evaluated for penbasins, guan jars,and ii tripods. Appendix 2 shows photographs of general surfacetreatment types according tovessel type. This includes an evaluation of surface treatment attributesobserved on the body,neck, rim, lip, and the sherd inner surface. Combinations are defined differentlydepending onsurface treatment styles, but are generally determined by the styleand placement of surfacetreatment on the outer body, and surface treatment alongthe rim or inner surface of the sherd(Appendix 3).Surface treatments associated for flat base basins inthe samples from HWD H2 and HYZH24 samples include cordmark, line, line and appliqué,line and appliqué and cordmark, and34plain. A total of twenty surface treatment combinations are observed in the two samples.Sixcombinations are shared in both contexts, three are exclusive to HWD H2, and elevenareexclusive to HYZ H24. The most popular combination in both contexts is the incisedline andcordmark with one pair of incised lines on the vessel body, 1 single incised linealong the neckcircumference, and a wide groove along the inner rim surface. Commonsurface treatments forround base pen basins observed at HWD and HYZ include various cordmark,line and plainsurface treatment, represented by a total of eight surface treatment combinations.Threecombinations are exclusive to HWD, while five combinations are exclusive to HYZ.At HWD,the most popular combination (n= 11) is the cordmark surface treatmentthat beginsapproximately 5 cm. below the rim. At HYZ, plain surface treatment with a widegroove alongthe rim inner surface (n 15). There are no common surface treatment styles betweenthe twocontexts, suggesting that the consumers from HWD and HYZ preferred completelydistinct typesof round base pen basins. This is only observed in this particular vessel subtype.More variability is observed in the decorative attributes of flat-basedpen basins (n=9)from HWD relative to round base pen basins (n3). At HYZ, greater variability isobserved inflat-basedpen basin (n= 17) compared to round-base pen basins (n=5). Surfacetreatmentvariation for pen basins is most visible in terms of the numbers! combinationsof incised lines onthe outer surface as well as the presence or absence of groovesalong the inside of the rim. Thisis demonstrated by the large number of surface treatment combinationsobserved for incised lineand cordmark style (n= 14), which is the largest number of combinations fora single surfacetreatment style observed in both HWD and HYZ.35DISCUSSION AND CONCLUSIONThis study of Shang pottery from HWD and HYZ providesinformation on ceramicvariability of Shang society based on ceramic evidencecollected recently from Huanbei. Newinformation regarding the general character of the Huanbei potteryassemblage is presented,particularly since no detailed reports have yet been publishedfor pottery collected from theHWD context. Three primary objectives of this study include: 1) identifyingand comparingmajor characteristics of the HWD and HYZ pottery assemblages,2) identifying and comparingpatterns of variability among serving, cooking, and storage potteryvessels from HWD and HYZ,and 3) considering interpretive models to explainthe observed patterns of differences andvariability. Findings of this study demonstrate that pottery sherdscollected from FIWD andHYZ demonstrate distinct differences in terms of formand stylistic attributes. Moreover,significant attribute variability is observed inthe pottery sherds examined from both contexts.Results of my study suggest that significant ceramicvariability is evident in the pottery producedfor and consumed by the settlers at Huanbei such thata model of controlled and centralizedproduction may not offer adequate explanationfor the observed patterns. Here I present somepossible interpretations for the observed patterns of ceramicvariability.As suggested by Prudence Rice (1989:113), vessel form can reflectfunctions andactivities among the archaeological contexts being compared.I identified significant differencesbetween the pottery assemblages from HWDand HYZ based on the ratios of pottery vesseltypes.The high ratios ofpen basins, ii tripods and guan jarssuggest that these vessels were importantin both HWD and HYZ contexts. In particular, globularguan jars are the most common amongall three guan jars subtypes at HWD, while large size guanjars are most common at HYZ. Ifguan jars are generally associated with storage functions,this suggests that the most preferredtype of storage vessel in the two contexts is different. The exclusivepresence of large-mouthzun vases andjiangjunkui vats at HWD indicates that theiruse may have been associated36exclusively with palatial activities. Anne Underhill (2002:206)has suggested that the large-mouth zun vase was used in alcohol fermentation, whileothers have suggested a role in alcoholserving (Anyang 2004a). The exclusive presence of largemouthed zun vases at HWD has thefollowing implications: 1) alcohol consumption was moreimportant at HWD than atHYZ; and 2)the large percentage of large sized zun vases at HWDsuggests that they demanded fermentationvessels with large capacity possibly to support large groupconsumption.By way of contextual association, the exclusive presenceof thejiangjunkui vat at HWDsuggests that its function is closely related to activitiesthat would take place in a palace context.Current interpretations of its use include alcohol fermentationvessel, bronze smelting, and saltproduction (boiling sea water to extract salt) (Underhill2002; Zhongguo 1987). Future researchusing chemical analyses methods may allow us to identifyits specific use and therefore identifythe association of certain activities with a palace context.Gui pedestal bowls and dou stemmedcups have appeared in more labor-intensive styles duringthe late Shang period: white kaolin claydou stemmed cups and jade gui pedestal bowls appearedin Anyang associated with elite burials(Orton, et al. 1993:169-171, 330). The generally lowquantities of these vessels at both HWDand HYZ suggest they were not popular inboth contexts. However, the relatively higherratio ofboth vessels at HWD suggest that these vessels were morepopular but not exclusively used inthe palace context.Form variation offers clues to the degree of standardizationin the manufacturing process(Rice 1989:113). Examination of forming techniquesused to produce pen basin, guan jar,and iitripod pottery vessels display low variability.The vessel shape forms of flat-base pen basins,round-base pen basins, large guan jars, globularguan jars, and ii tripods show a highlevel ofconsistency in terms of overall form.The only vessel type that displays high variabilityin shapeis the small guan jar. It appears that the productionlconsumptionofguan jars may have beengoverned by a different set of standards. One possibleexplanation is that due to the small-sizeof37these guan jars, the exchange of these vessels among different groups of people wouldhave beenpopular due to their transportability. Because forming is most resistant to change due to limitedvisibility and motor habits, variations in vessel shape tend to reflect distinct cultural traditionsgoverned by linguistic or other social boundaries (Gosselain 2000:191-192; 1992:572).SinceHuanbei was one of the largest settlements in the regions at that time, it is likelythat foreign-styled pottery would have been imported and exchanged in the local marketsas travelers visitedthe area.Metric variation provides specific information regarding the volume of food preparedandserved, group size of the consumers, and the variety of food processing tasks practiced(Blitz1993:85), but can also establish a measure of relative standardization. In a similar casestudy, ithas been suggested that elite/religious/public contexts have a more restricted vessel size rangeand disproportionately larger vessels to support the limited set of large-group food consumptionand storage needs (Blitz 1993:87). Greater size variability is associated with village/householdcontexts to reflect the greater variety in domestic activities practiced (Blitz 1993:90).In thisstudy, the observed patterns suggest that vessels used in the palace context werenot always morestandardized than vessels from residential context. In fact, it is observed that small size-guanjars and flat-base pen basins from HWD were more varied in rim thickness measurementsaccording to the calculated CV values. This reflects that these vessels may havebeen associatedwith domestic activities, and that domestic activities practiced atHWD were relatively diverse.Mean rim diameter comparisons for pen basin, guan jar, and ii tripod vessels do not indicatesignificant differences in the average vessel size across similar categories.However, asmentioned earlier, there are additional large sized vessels that were associated exclusivelywiththe HWD context which may have been dedicated to activities practiced exclusively inthe palacesector.Comparisons of secondary form variation (rim form,lip variation, vessel size) best38represent the observed ceramic attribute variability inthe samples from Huanbei. Like vesselshape, this aspect of form variation would have beenlargely influenced by cultural traditionsfollowing Gosselain’s argument (1992:572). Unlikevessel shape attributes, rimllipcombinations indicate that pottery from both HWDand HYZ display high variability. Based onthe number of rimllip combinations, round-base basinsdisplay greater standardization than flat-base basins in both contexts. Also, globular guan jarsdisplay greater standardization than largesize and small size guan jars. It appears that these twosubtypes were more standardizedcompared to other subtypes ofpen basins and guan jars. A possibleexplanation for this patternis that the producers of these two vessel subtypes arepotters shared similar cultural backgroundsand may have been employed as specialized producersof these pottery vessels. In comparison,high variability in small size guan jars and ii tripods suggest these vesselswere obtained from amore diversified set of sources, the sources of whichmay have involved a multitude of pottersfrom diverse cultural backgrounds and pottery traditions.Evidence from decorative/stylistic attributes corroboratewith these findings. Decorative!stylistic atthbutes are generally useful in providing information aboutthe context of use oroutside contacts of the makers (Rice 1989:113), and is particularly potentwith informationregarding the expression of social identities in socialinteraction and social boundaries. Resultsof this comparison are particularly telling of the differencesbetween the HWD and HYZ round-base pen basins. Based on the types of surface treatment combinationsobserved, the round-basepen basins used in the two contexts are completelydistinct in style. Vessels from HWDgenerally exhibit some variability, except for globularguan jars where only one surfacetreatment combination is observed. In comparison, vesselsfrom HYZ display high variabilityfor all pen basin, guan jar, and ii tripod vessels. Dueto the lack of information regardingresource variability, it is difficult to determine whetherthe observed variations are the results ofactive cultural interaction between foreign and local potters,or if foreign pottery vessels were39frequently imported and consumed by the local people. The observedpatterns, however,strongly suggest that the pottery vessels produced for and consumedat Huanbei do notdemonstrate characteristics of a highly standardized potteryproduction system. Some vesselsexhibit more standardized attributes, which indicate thatthese particular types may have beenmore involved in socially visible contexts. Variabilityin forming and decorative attributessuggest 1) consumers from the HWD context demanded a greater varietyof pottery vessel typesfor palatial activities but often a more limited range of shapesand decorative styles; and 2)consumers from the HYZ used a limited range of pottery vessel types likelydedicated todomestic activities but greater variability in the range of shapes anddecoration.This study has attempted to demonstrate the potential value of investigatingintra-siteceramic variation. Traditional ceramic research in Chinaoften attributes ceramic variation toculture-historic differences resulting from cultural, regional andtemporal differences (Zhongguo2003:253). Results of my analysis suggest form and surface treatmentvariability are particularlyeffective in illustrating patterns of variability for pen, guan, and ii vesselsfrom Huanbei. Metricvariability in terms of rim diameter and rim thickness has also revealedinteresting patterns.Future research may benefit from the consideration of additional attributesin the study ofcomplete or refitted vessels. To better understand the expressionsof variability in Shang pottery,methods in experimental archaeology may offer new insightson sources of production variability.Resource variability and technological variability should be exploredto better understand theorganization of pottery production, particularly since firing technologyhas revealed significantvariability. It will also provide more information for assessing whetherthe source of variabilityobserved in this study is due to variation in local productionor high volume import of foreignpottery vessels.An additional perspective yet to be considered in future researchis the role of bronzevessels in daily contexts. Cooking, storage, and serving vessels fabricatedin bronze were a40common part of the Shang elite’s lifestyle as suggestedby mortuary evidence (Underhill2002:238-239). Since HWD was a palatial compoundwhere governmental/religious activitiestook place and the royal family resided, many more bronzevessels would have been consumedcompared to HYZ. To what degree would this haveaffected the observed ceramic patterning atHuanbei? Were bronze vessels frequently consumedin daily contexts, and if so, what rolewould they have played in different activities? Futureresearch should consider the significanceof pottery vessel use and variability in relation to othercraft industries.41BIBLIOGRAPHYAnyang2003a Zhongguo Shehui Kexueyuan Kaogu YanjiusuoAnyang Gongzuodui,H111E. Henan Anyangshi Huanbei Shang cheng de kancha yushijue (Surveyand Test Excavation of the HuanbeiShang city in Anyang). Kaogu -428(5):3-16.2003b Zhongguo Shehui Kexueyuan Kaogu Yanjiusuo Anyang Gongzuodui,‘1 l1±Henan Anyang shi Huanbei Shang cheng gongdianquyihao jizhi fajue jianbao(A Brief Reporton the Excavation of Palatial Compound Fl at the Huanbei Shangcity in Anyang,Henan). Kaogu -428(5):17-23.2004a Anyang Work Team of the Institute of Archaeology,CASS. A Brief Report onthe Excavation of Palatial Compound Fl at the HuanbeiShang City in Anyang, Henan.Chinese Archaeology 4:21-28.2004b Anyang Work Team of the Institute of Archaeology,CASS. Survey and TestExcavation of the Huanbei Shang City in Anyang. Chinese Archaeology4:1-20.2004c Zhongguo Shehui Kexueyuan Kaogu Yanjiusuo AnyangGongzuodui,1998-1999 Anyang Huanbei Shang chengHuayuanzhuang dong di fajue baogao 1998-1999(Excavation Report on the Locus Eastof Huayuanzhuang of the Shang Period City atHuanbei in Anyang in 1998-1999). Kaoguxue JikanfIJ15:296-358.Bagley, R.1999 Shang Archaeology. In Cambridge History ofAncientChina: From the Origins ofCivilization to 221 B.C., edited by M. Loewe and E. L. Shaughnessy,pp.124-231.Cambridge University Press, Cambridge.Blitz, J. H.1993 Big Pots for Big Shots: Feasting and Storage ina Mississippian Community.American Antiquity 58(1):80-96.Chang, K. C.1980 Shang Civilization. Yale University Press,New Haven.Dunnell, R. C.1978 Style and Function - Fundamental Dichotomy.American Antiquity 43(2): 192-202.Egloff, B. J.1973 A Method for Counting Ceramic Rim Sherds. AmericanAntiquity 38(3):351-353.Gosselain, 0. P.1992 Technology and Style - Potters and Pottery amongBafia of Cameroon. Man27(3):559-586.421999 In Pots We Trust The Processing of Clay and Symbols in Sub-Saharan Africa.Journal ofMaterial Culture 4(2):205-230.2000 Materializing Identities: An African Perspective. Journal ofArchaeologicalMethod and Theory 7(3):187-217.Jing, Z., G. R. Rapp, J. Tang and J. Stoitman2004 Coevolution ofHuman Societies and Lanscapes in the Core Territoiy ofLateShang State - An Interdisciplinary Regional Archaeological Investigation in Anyang,China, 2001-2004. Institute of Archaeology, Chinese Academy of Social Sciences, andUniversity of Minnesota.Jing, Z., J. Tang, G. R. Rapp and J. Stoitman2002 Co-Evolution ofHuman Societies and Landscapes in the Core Territory ofLateShang State - An Interdisciplinary Regional Archaeological Investigation in Anyang,China, 2000-2001. Institute of Archaeology, Chinese Acadamy of Social Sciences, andUniversity of Minnesota.Lemonnier, P.1986 The study of material culture today: towards an anthropology of technical systems.Journal ofAnthropological Archaeology 5(2): 147-186.1993 Introduction. In Technological Choices - Transformation in Material Culturessince the Neolithic, edited by P. Lemonnier,pp.1-35. Routledge, London.Liu, L. and X. Chen2007 State Formation in Early China Duckworth Publishers, London.Loehr, M.1953 The Bronze Styles of the Anyang Period (1300-1028 B.C.). Archives oftheChinese Art Society ofAmerica 7:42-53.Masson, M. A. and R. M. Rosenswig2005 Production Characteristics of Postclassic Maya Pottery from Caye Coco, NorthernBelize. Latin American Antiquity 16(4):355-384.Orton, C.1993 How many pots make five? — an historical review of pottery quantification.Archaeometry 3 5(2): 169-184.Orton, C., P. Tyers and A. Vince1993 Pottery in Archaeology. Cambridge University Press, Cambridge.Rice, P. M.1987 Pottery Analysis - A Sourcebook. University of Chicago Press, Chicago.1989 Ceramic Diversity, Production, and Use. In QuantfjingDiversity in Archaeology,edited by R. Leonard and G. Jones,pp.109-117. Cambridge University Press, Cambridge.43Sackett, J. R.1977 Meaning of Style in Archaeology - General Model. AmericanAntiquity42(3):369-380.1985 Style and Ethnicity in the Kalahari - A Reply toWiessner. American Antiquity50(l):154-159.1 986a Isochrestism and Style - A Clarification. Journal ofAnthropologicalArchaeology5(3):266-277.1 986b Style, Function, and Assemblage Variability - A Replyto Binford. AmericanAntiquity 5 1(3):628-634.Shaughnessy, E. L.1989 Historical Geography and the Extent of the Earliest ChineseKingdoms. AsiaMajor 2(2): 1-22.Stark, M. T.1999 Social Dimensions of Technical Choice in KalingaCeramic Traditions. InMaterial Meanings - Critical Approaches to the InterpretationofMaterial Culture,edited by E. Chilton,pp.24-43. University of Utah Press, Salt LakeCity.Tang, J.1999 Zhong Shang Wenhua Yanjiu (AStudy of the Mid Shang Culture).Kaogu Xuebao 4:393-420.2002 Shang wang chao kao gu xue bian nian dejianliZhong yuan wen wu 3t’6:50-59.Thorp, R.1985 The Growth of Early Civilization, New Datafrom Ritual Vessels. HarvardJournal ofAsiatic Studies 45(1):5-75.2006 China in the Early Bronze Age - Shang Civilization.University of PennsylvaniaPress, Philadelphia.Trigger, B.1999 Shang Political Organization: A Comparative Approach.Journal ofEast AsianArchaeology 1(l):43-62.Underhill, A. P.2002 Craft Production and Social Change in Northern China.Kluwer Academic!Plenum Publishers, New York.van der Leeuw, S.2002 Giving the Potter a Choice - Conceptual aspectsof pottery techniques. InTechnological Choices - Transformation in Material Culturessince the Neolithic, editedby P. Lemonnier,pp.238-288. Routledge, London.44Wiessner, P.1983 Style and Social Information in Kalahari-San Projectile Points. AmericanAntiquity 48(2):253-276.1984 Reconsidering the Behavioral Basis for Style - A Case Study amongthe KalahariSan. Journal ofAnthropological Archaeology 3(3):190-234.1985 Style or Isochrestic Variation - A Reply to Sackett. American Antiquity50(1):l60-166.Wobst, H. M.1977 Stylistic Behavior and Information Exchange. In For the Director:ResearchEssays in Honor ofJames B. Grffln, edited by C. E. Cleland,pp.317-342. vol. 61.Museum of Anthropology, University of Michigan, Ann Arbor, Michigan.XiaShangZhou2000 Xia Shang Zhou duan dai gong cheng zhuanjia zu. Xia Shang Zhouduan daigong cheng 1996-2000 nianjie duan cheng guo bao gao:jian ben1 996-2000l* (Xia Shang Zhou Chronology Project). Shi jie tu shuchu ban gong si, Beijing.Yan, W.1997 Zou xiangshUi de kooguxue Sanqin Chubanshe, Xi’an.Yates, R.1994 The-City State in Ancient China. In The ArchaeologyofCity-States. Cross-cultural Approaches, edited by D. L. N. a. T. H. Chariton. vol. 7 1-90. SmithsonianInstitution Press, Washington, D. C.Zhongguo1987 Zhongguo Shehui Kexueyuan Kaogu Yanjiusuo,Yinxu Fajue Baogao (Excavation of Yinxu 1958-1961). ZhongguoShehuiKexueyuan Kaogu Yanjiusuot±FEF.1994 Zhongguo Shehui Kexueyuan Kaogu Yanjiusuo,Yinxu de Faxian Yu Yanjiu (Archaeology Excavationand Researchesin the Yin Ruins). Zhongguo Shehui Kexueyuan Kaogu Yanjiusuo2003 Zhongguo Shehui Kexueyuan Kaogu Yanjiusuo,Zhongguo Kaoguxue Xia Shang Juan (ChineseArchaeology Xia andShang). Zhongguo Shehui Kexueyuan Kaogu YanjiusuoZou, H.1980 Shi Lun Xia WenhuaI4-tL. In Xia Shang Zhou Kaoguxue LunwenjiK/Ypp.95-182. Wenwu ChubansheLH±, Beijing.45APPENDICESAppendix A. Rim Shape VariationsPottery SubtypeRim Shape Variations(Flat-based)H21H24))1H24H2401 110cm46H24” \OI-IlOcmGuan (Large-sized)3,4)Y‘.‘)01 10cmAppendix A. Rim Shape Variations (Cont’d)Pottery SubtypeRim Shape VariationsPen(Round-based)H2\:c..47Appendix A. Rim Shape Variations (Cont’d)Pottery SubtypeGuan (Small-sized)Guan (Globular)H2Rim Shape Variations?H24I-H2ui)1)(4)H201 10cm11)7H2401 110cm48Appendix A. Rim Shape Variations(Cont’d)PotteryRim Shape VariationsSubtypeLiH)1>11H1)7)flH2)Y1?H2) ‘)JFVOj10cm49Appendix B. Surface Treatment VariationsPotterySubtypePen(Flat-based)a. “Line & Cordmark”b. “Line & Applique”c. “Line”Pen(Round-based)a.”Cordmark”b.”Plain”c. “Line”Surface Treatment Variations50Appendix B. Surface TreatmentVariations (Cont’d)Pottery SubtypeGuan (Large-sized)“Line & Cordmark”onlyGuan (Small-sized)a. “Line”b. “Line & Cordmark”c.”Burnished”d.”Cordmark”Guan (Globular)“Outcurving”a. “Fine square stamp”b. “Medium squarestamp”c. “Coarse squarestamp”Surface Tr’”51Appendix B. Surface Treatment Variations (Cont’d)Pottery SubtypeLia. “Line &Cordmark”b. “FineCordmark”c. “MediumCordmark”d. “CoarseCordmark”e. “Cordmark &Applique”Surf Variations52AppendixC.SurfaceTreatmentCombinationsbyVesselTypeandFormGuan(Large)Appliqué Line&CordmarkLine&CordmarkLine&CordmarkLine&CordmarkTotalAppliquéband1cmbelowneckCordmarkbegins3cmbelowneckCordmarkbegins5cmbelowneckCordmarkbeginsatneckCordmarkbeginsatneckPlainrimsurfacePlainrimsurfaceRaisedlinealongneckoutersurfacePlainrimsurfaceRaisedlinealongneckoutersurfaceHWDHYZ1012122H3N=59N=4Burnished Burnished Burnished Burnished Line Line Line Line Cordmark Cordmark TotalFine,smoothsurfaceFine,unevensurfacewI1pairofIncisedlinew/1pairofIncisedline1pair&1singleincisedlineonbodyDoubleincisedlineonbodySingleincisedlineonbodyTripleincisedlineonbodyCordmarkbegins3cmbelowneckCordmarkbeginsatneckPlainrimsurfacePlainrimsurface1setofDoubleincisedlinealongneckoutersurface2setofDoubleincisedlinealongneckoutersurfacePlainrimsurfaceIncisedlinealongneckoutersurfaceIncisedlinealongneckoutersurfacePlainrimsurfacePlainrimsurfacePlainrimsurfaceGuan(Small)Guan(Globular)tJo4o o2o 10o o2o 1011N=3N=8SquareStampCoarseCordmarkbeginsatneck02SquareStampFineCordmarkbeginsatneck03SquareStampMediumCordmarkbegins3cmbelowneck02SquareStampMediumCordmarkbeginsatneck126TotalN=1N=4AppendixC.SurfaceTreatmentCombinationsbyVesselTypeandForm(Cont’d)CordmarkFineCordmarkLine Line Line Line&Appliqué&CordmarkLine&CordmarkLine&CordmarkLine&CordmarkLine&CordmarkLine&CordmarkLine&CordmarkLine&CordmarkLine&CordmarkLine&CordmarkLine&CordmarkLine&CordmarkLine&CordmarkLine&CordmarkLine&CordmarkPlainPlainrimsurfaceFinelineFinelineRaisedlinealongneckoutersurfaceRaisedlinealongneckoutersurfacePlainrimsurfacePlainrimsurfaceWidegroovealongriminnersurfacePlainrimsurface2groovealongriminnersurfaceWidegroovealongriminnersurfaceWidegroovealongriminnersurfaceWidegroovealongriminnersurfacePlainrimsurfaceWidegroovealongriminnersurface2groovealongriminnersurface2incisedlinealongriminnersurfacePlainrimsurfaceWidegroovealongriminnersurfaceWidegroovealongriminnersurfaceo2 013o416o710o122o1o6Pen(Round-base)Total Cordmark Cordmark Cordmark Line Plain Plain Plain PlainCordmarkbegins4cmbeloworificeCordmarkbegins4cmbeloworificeCordmarkbegins5cmbeloworifice2pairofIncisedlineonbodyUndecoratedbodyUndecoratedbodyUndecoratedbodyUndecoratedbody2groovealongriminnersurfaceWidegroovealongriminnersurfacePlainrimsurfacePlainrimsurface2groovealongriminnersurface1FinegroovealongriminnersurfacePlainrimsurface1WidegroovealongriminnersurfaceN=9N=174060110o o o3o10015Pen(Flat-base)1pairofIncisedlineonneck1singleincisedlineonneck2pairofIncisedlineonbody&1singleIncisedlineonneck1pairofIncisedlineonbody&IsingleIncisedlineonneckIpairofIncisedlineonbody1pairofIncisedlineonbody&1singleIncisedlineonneckIpairofIncisedlineonbody&1singleIncisedlineonneckIpairofIncisedlineonbody&3singleIncisedlineonneck1singleIncisedlineonbodyIsingleIncisedlineonbody&1pairIncisedlineonneck1singleIncisedlineonbody&1singleIncisedlineonneck2pairsofIncisedlineonbody2pairsofIncisedlineonbody&1singleIncisedlineonneck2pairsofIncisedlineonbody&IsingleIncisedlineonneck2singleIncisedlineonbody2singleIncisedlineonbody2singleIncisedlineonbody3singleIncisedlineonbody&3singleIncisedlineonneckUndecoratedbodyHWDHYZ051014010122027802TotalN=3N=5Table4.SurfaceTreatmentCombinationsbyVesselTypeandForm(Cont’d)CoarseCordmarkCoarseCordmarkCoarseCordmarkCordmark&AppliquéCordmark&DoubleIncisedlineCordmark&IncisedlineCordmark&IncisedlineCordmark&IncisedlineCordmark&IncisedlineCordmark&IncisedlineCordmark&TripleIncisedlineFineCordmarkFineCordmarkFineCordmarkFineCordmarkMediumCordmarkMediumCordmarkMediumCordmarkMediumCordmarkMediumCordmarkMediumCordmarkMediumCordmarkMediumCordmarkMediumCordmarkMediumCordmarkMediumCordmarkCordmarkbegins1cmbelowneckCordmarkbegins1cmbelowneckCordmarkbeginsatneckFineCordmarkwithappliquéCordmarkbegins3cmbelowneckCordmarkbegins1cmbelowneckCordmarkbegins3cmbelowneckCordmarkbegins3cmbelowneckCordmarkbegins3cmbelowneckCordmarkbeginsatneckCordmarkbegins3cmbelowneckCordmarkbegins1cmbelowneckCordmarkbegins3cmbelowneckCordmarkbegins3cmbelowneckCordmarkbeginsatneckCordmarkbegins1cmbelowneckCordmarkbegins1cmbelowneckCordmarkbegins1cmbelowneckCordmarkbegins1cmbelowneckCordmarkbegins3cmbelowneckCordmarkbegins3cmbelowneckCordmarkbegins3cmbelowneckCordmarkbegins3cmbelowneckCordmarkbeginsatneckCordmarkbeginsatneckCordmarkbeginsatneckPlainrimsurfaceWidegroovealongriminnersurfaceWidegroovealongriminnersurfaceFinegroovealongriminnersurfaceFinegroovealongriminnersurfaceWidegroovealongriminnersurfaceExtra-WidegroovealongriminnersurfaceFinegroovealongriminnersurfaceWidegroovealongriminnersurfacePlainrimsurfacePlainrimsurfaceWidegroovealongriminnersurfaceFinegroovealongriminnersurfacePlainrimsurfacePlainrimsurfaceExtra-WidegroovealongriminnersurfaceFinegroovealongriminnersurfacePlainrimsurfaceWidegroovealongriminnersurfaceExtra-WidegroovealongriminnersurfaceFinegroovealongriminnersurfacePlainrimsurfaceWidegroovealongriminnersurfaceExtra-WidegroovealongriminnersurfacePlainrimsurfaceWidegroovealongriminnersurfaceLiHWD0 2 3 0 0 6 0 3 0 0 0 0 2 0 0 4 04 0 2HYZ2 0 I17 7 0 0 I 0 8 4189 0 1 0 2 2 0TotalN=14N=19


Citation Scheme:


Usage Statistics

Country Views Downloads
China 20 20
United States 14 1
Canada 5 0
Russia 3 0
France 2 0
Australia 1 0
United Kingdom 1 0
City Views Downloads
Beijing 15 19
Unknown 12 20
Ashburn 5 0
Shenzhen 5 1
Millarville 2 0
San Francisco 2 0
Vancouver 1 0
University Park 1 0
Chard 1 0
Ann Arbor 1 0
Hornsby 1 0

{[{ mDataHeader[type] }]} {[{ month[type] }]} {[{ tData[type] }]}
Download Stats



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items