Settlement and Community at San Estevan, Belize

Journal of Anthropological Archaeology 21, 120–141 (2002) doi:10.1006/jaar.2001.0394

An Institutional Perspective on Prehispanic Maya Residential Variation: Settlement and Community at San Estevan, Belize

Laura J. Levi

Department of Anthropology, The University of Texas at San Antonio, 6900 North Loop 1604 West, San Antonio, Texas 78249-0652

E-mail: [email protected]

Received December 4, 2000; revision received June 18, 2001; accepted November 8, 2001

This article explores implications of a pattern of residential settlement found at the prehispanic Maya community of San Estevan in northern Belize. Although an awareness of the decision-making behavior of domestic groups helped to isolate the pattern, a household approach cannot by itself account for the overall community structure that emerged from the differential distribution of San Estevan’s residential forms. Making sense of the organizational forces underlying this distribution requires a more comprehensive set of explanatory tools drawn from conceptualizations of institutions and their powers to shape the spaces of social practice. © 2002 Elsevier Science (USA) Key Words: archaeology; Maya; institutions; power.

INTRODUCTION decision making. On the surface, this would appear to be a likely discovery, An archaeological settlement pattern largely anticipated by the maxim that soci- study conducted at the lowland Maya site eties are systemically related wholes. As of San Estevan, Belize, has demonstrated Ortner [1990 (1984):390] observes, however, that the residential composition of site our longstanding assumptions about the areas varies with distance to monumental systemic properties of human interaction precincts. Although the research findings have tended to deflect attention away from may recall Sjoberg’s (1960) model of con- the question of “where ‘the system’ comes centric zonation, popularized in Maya ar- from” (see also Mann 1986:1; Trigger chaeology in the 1970s (Folan et al. 1979; 1989:27). In light of that question, the San Hammond 1975; Haviland 1970; Kurjack Estevan findings lose some of their trans- 1974; Marcus 1983), there is little basis for parent logic. Analysis of the site’s residen- comparison. In contrast to the predictions tial architecture was informed by an ap- of concentric zonation, San Estevan’s resi- proach that attributes decision-making dential distributions do not correlate with autonomy to individual domestic groups. archaeological indicators of “wealth” or Yet, because of its emphasis on household “elite” status (cf. Abrams 1987; Arnold and agency, this same approach posits only the Ford 1980; Carmean 1991). Instead, the di- most “tenuous” of connections between so- mensions of variation examined in the ciety’s domestic and political institutions study were expressly chosen to address di- (Netting 1993:19). The San Estevan study is versity in the organizational and produc- therefore vulnerable to criticism about tive strategies of prehispanic domestic where best to situate agency in social groups. At San Estevan the spatial associa- process. More significantly, the study’s tions between residential forms and findings suggest a counterintuitive rela- civic/ceremonial architecture indicate that tionship between agency and power. activities housed in monumental precincts Should it prove reasonable to understand had important consequences for household domestic groups as strategic decision mak- 120 0278-4165/02 $35.00 © 2002 Elsevier Science (USA) All rights reserved. PREHISPANIC MAYA RESIDENTIAL VARIATION 121 ers, then how do we also understand a set- munity spanned three limestone ridges, or tlement system composed of such groups uplands, ringing the perennial wetland, but seemingly structured by powers exter- Long Swamp (Fig. 2). Each of these ridges nal to any of them? supported a precinct of monumental archi- San Estevan presents the curious case of a tecture, with the most prominent housed at settlement system revealed inadvertently the center of San Estevan’s largest upland and on the basis of premises that are to zone. Each ridge also sustained myriad res- varying degrees compromised by the re- idential groupings of variable scale and search results. My purpose here is to recon- spatial configuration. My own archaeologi- cile this discrepancy between expectations cal fieldwork at the site documented this di- and outcomes and, in the process, to clarify versity of residential arrangements within a some of the organizational forces realized sample of 20 survey blocks measuring 250 through San Estevan’s spatial patterning. m on a side. Three additional residential The article begins with a description of how zones, each comparable in size to a survey the anthropological literature on house- block, also were investigated, but in a more holds was used first to discriminate among opportunistic fashion as they gradually be- residential arrangements and subsequently came cleared of bush or cane. All architec- to reveal spatially sensitive dimensions of tural remains in these 23 survey localities variation in the site’s assemblage of domes- were mapped and surface collected, and se- tic architecture. Discussion then moves to lected residential units were test excavated consideration of where the San Estevan set- in order to obtain information about their tlement system came from. The power to construction histories. structure that system cannot solely be attributed to either administrative or house- The Archaeology of Domestic Groups hold action. Rather, it must be located in the interplay of both institutional arenas. The Fieldwork at San Estevan laid the foun- remainder of the article considers how best dation for an examination of the distribu- to model agency and power as they relate tional parameters underlying diversity in to institutions, in general, and to house- the composition and layout of residential holds, in particular. units at the site. Empirical in intent, the research sought to identify and account for RESEARCH AT SAN ESTEVAN organizational differences among the community’s prehispanic domestic groups. San Estevan is located just east of the That variation in residential forms com- New River, at the far western margins of prised the principal vehicle used to ad- northern Belize’s flat, coastal plain (Fig. 1). dress these issues requires some justifica- Initial archaeological studies by William tion given increasing concerns over the Bullard (1965) and Norman Hammond interpretive potential of architectural data. (1973) helped to define major chronological Critics of architectural approaches to do- and architectural components of the site. mestic group behavior register two pri- Settled in the Middle Preclassic and pos- mary objections: first, that residences are sessing significant occupation by Early unresponsive to temporal dynamics span- Classic times, San Estevan reached its great- ning reductions in domestic personnel or est population levels and areal extent in the their wholesale replacement (e.g., Hirth Late Classic Period (Hammond 1973; Levi 1993); and, second, that structure plans and 1993). Bounded by the New River to the interiors do not always constitute sensitive west and by large seasonal wetlands to the or unambiguous indicators of domestic north, east, and south, the Late Classic com- activities (Allison 2000; Goldberg 2000). 122 LAURA J. LEVI

FIG. 1. Map of selected archaeological sites in northern Belize.

Much recent research in lowland Meso- is still pending on whether activity area re- america attempts to compensate for the lat- search captures temporal ﬂuctuations of a ter problem. To great effect, emphasis has different order or on a ﬁner scale than shifted from architecture to artifact and architectural studies (but see Alexander from structural space to extramural space 2000:92–93). Furthermore, activity studies (e.g., Alexander 2000; Johnston and Gonlin do not necessarily provide comparable in- 1998; Killion 1990, 1992; Killion et al. 1989; formation on domestic group behavior. As McAnany 1992a). Nevertheless, the verdict research is ever more closely trained upon PREHISPANIC MAYA RESIDENTIAL VARIATION 123

FIG. 2. San Estevan Project area showing locations of monumental precincts, survey blocks, and other mapped settlement localities. human action, behavior increasingly behavior in its organizational aspect (cf. Flan- comes identiﬁed with task—the work that nagan 1989:248), and on the whole, analy- people do. The connections forged among ses of the “built environment” provide the people doing all that work often appear some of our most critical insights into orga- incidental to the tasks performed. Yet, in- nization as it is materially represented terconnections among people speak to be- (Lawrence and Low 1990). 124 LAURA J. LEVI

Not surprisingly, most investigations of In anthropology, the agrarian ecological residences in Maya archaeology have studies of Robert Netting have lent the stressed social affiliations. I would argue greatest support to this latter perspective. that dissatisfaction with this kind of re- Netting’s early work among Nigeria’s Kof- search arises more from the organizational yar agriculturalists highlighted a striking perspectives brought to bear on domestic flexibility of domestic forms and activities architecture than from any interpretive pit- in relation to the variables of demography falls intrinsic to the data set. The lineage and land availability (Netting 1968, 1993). models employed for many years by At issue was how to account for the exis- Mayanists furnish a case in point. Although tence of such variation within the context of extremely sophisticated both methodologi- a single cultural group subsumed by a sin- cally and theoretically (see especially Hen- gle overarching political structure. Circum- don 1991; McAnany 1992b, 1995; Sanders venting a facile demographic determinism, 1989; Sheehy 1992), lineage approaches in- Netting directed attention to the tasks per- variably produced typological assessments formed by Kofyar households and the ways that invoked a set of normative and uni- in which their labor was deployed. The formly experienced beliefs to seamlessly household, he emphasized, was a corporate merge prehispanic Maya domestic and po- entity whose membership shared responsi- litical institutions. Domestic organization bility for production, consumption, and re- was asserted to be fully determined by uni- production. Discharging those responsibili- lineal descent principles that comprised, as ties was contingent upon the household’s well, the pervasive idiom for power rela- ongoing mediation of external forces and tions throughout Maya society. Distance to constraints. Variation in domestic group or- apical ancestors was argued to have gov- ganization should therefore be understood erned access to political aegis and produc- to follow from strategic differences in tive lands. Accordingly, variations in the household decision making. size of residential units, their construction Appropriating the term “household” to histories, and expressions of inequality signal a shift away from an a priori holism were interpreted to reflect the divergent that empirical studies could not support, fortunes of lineages and their constituent Netting and his students challenged both families. the structural-functionalist causality that While a few detractors have challenged gave primacy to societal norms as well as a the specific relevance of unilineal descent persistent, if tacit, unilineal evolutionism groups to the prehispanic Maya (e.g., Gille- that informed arguments about the corre- spie 2000; Wilk 1988), the vast majority con- spondence between domestic and political tend more generally that “families” and do- institutions (Netting et al. 1984:xvii–xix; mestic groups are rarely coterminous (e.g., Wilk 1991:18–26). A concept of household Hendon 1996; Johnston and Gonlin 1998; agency informed both critiques. The au- Ashmore and Wilk 1988). Introducing a thority attributed to household decision fundamental question of causality, both making suggested the overall expedience of concerns invite debate over the organiza- tradition (Wilk and Netting 1984). At the tional determinants of domestic life. Typo- same time, it undermined conventional no- logical perspectives pointed to a set of uni- tions of a determinant stream of causality form and stable forces located somewhere flowing from society’s most powerful outside the mundane activities in which members to its least. The conduct of politi- domestic group members engaged. The al- cal life was geographically diffuse and pro- ternative, of course, is to situate causality ceeded at multiple scales, while the house- directly within the domestic realm, itself. hold was a locus where land and labor PREHISPANIC MAYA RESIDENTIAL VARIATION 125 converged in sets of situated activities coercion like servitude and slavery. Not (Wilk 1991:31–33). In consequence, house- merely contingent, therefore, households hold agency compelled investigations of or- are also negotiated constructs, pursuing (or ganizational dynamics to be thoroughly being compelled to accept) some forms of grounded in “local environment” (Netting recruitment and not others. And for the 1993:21). household to work as such, it requires organizational strategies of incorporation (cf. Finding Households in Residence at San Arnould 1984; Bachnik 1983; Blanton 1995). Estevan The close connection between household production and recruitment, and the orga- San Estevan’s prehispanic community nizational strategies necessary to effect this had been dispersed over more than 30 connection, suggested a way to identify San km2 of an ecologically diverse area. Its Estevan’s prehispanic households while at residential units ranged from isolates the same time gauging differences among and paired structures to several different them. One of the most impressive facets of kinds of multistructure, plaza-focused San Estevan’s residential assemblage was units. A household approach, grounded the range of architectural mechanisms used in “local environment,” contained the to incorporate individual structures into implication that the formal and distribu- larger residential groupings. Plazas, struc- tional properties of residence should be ture abutments, shared substructural plat- examined together, a necessary contextu- forms, and broad basal platforms support- alization that typological perspectives ing whole residential groups were some of often neglected. Moreover, by situating the more common mechanisms that helped sources of organizational diversity in to confer spatial integrity to residential household decision-making processes, units at the site. Many of these devices were the approach offered a means to inspect differentially distributed across the site, residential variation without recourse to and no one residential unit possessed the assumptions about how the domestic entire array. San Estevan’s domestic archi- realm was tied to broader socioeconomic tecture seemed highly strategic as a result. and political forces. Finally, household Although an inferential leap, I began to research contained the germ of an idea view this array of mechanisms as a reper- about connections between people and toire of architectural elements available to the places they inhabit, pointing to the domestic groups for the purpose of orga- central role recruitment plays in the func- nizing their membership. Differences in the tioning of any domestic group. elements selected should speak to differ- Households are highly contingent social ences in the connections that had been forms, poised at the precipice of failure forged among domestic group members should relationships between economic ori- through their efforts to balance productive entations and household membership be options and labor requirements. Prehis- disrupted (e.g., Cain 1988; McDonald 1991). panic households at San Estevan could be There are many ways for households to ad- simultaneously discerned and differenti- just the balance sheets—tallying produc- ated by the kinds of structure incorporation tion and consumption and juggling among mechanisms they had employed. alternative productive and recruitment op- On the basis of a combined consideration tions. In addition to childbearing, recruit- of structure incorporation mechanisms and ment strategies may include various mar- structure number, five principal classes of riage practices, adoption, extensions of real residential arrangements were identified at or fictive kinship, and forms of sanctioned the site (Fig. 3). Isolates and Paired Plat- 126 LAURA J. LEVI

FIG. 3. Residential classes at San Estevan. (a) Large Composite Group, (b) Basal Platform Group, (c) Focus Group, (d and e) Paired Platform Groups, and (f and g) Isolates. form Groups, consisting of 1 and 2 struc- classes. Members of the Focus Group class tures, respectively, showed little or no in- possessed 3 or 4 structures positioned to vestment in architectural mechanisms of define a central plaza area (Fig. 3c). In the structure incorporation, and both lacked Basal Platform Group class the primary in- clear vestiges of plaza foci (Figs. 3d–3g). The tegrating feature was a broad basal plat- rest of San Estevan’s residential units were form supporting from 2 to 5 structures plaza-focused arrangements, and the vast arranged along the platform’s perimeters majority of these fell into three well-defined (Fig. 3b). Large Composite Groups, on the PREHISPANIC MAYA RESIDENTIAL VARIATION 127 other hand, possessed from 6 to 13 struc- tinuously modified into the Terminal Clas- tures and exhibited a profusion of structure sic, long after building activity had ceased incorporation mechanisms, including one in other kinds of residential units. Finally, or more plazas, numerous structure abut- the residential units incorporating the most ments, linear basal platforms supporting 2 structures at San Estevan were also among or 3 contiguously aligned structures, and the shortest lived groups at the site. Despite small alters or shrines (Fig. 3a). their sprawling size and spatial complexity, The residential classification used at San Large Composite Group coalesced and de- Estevan partitioned the architectural as- clined during the Late Classic. semblage according to pronounced differ- In summary, the residential classification ences in the composition and layout of the offered little support for the idea that a site’s residential units. These differences, single, enduring set of normative values however, could not be correlated with obvi- governed the behavior of San Estevan’s ous dimensions of inequality. Significant prehispanic domestic groups. Typological disparities existed in the labor and re- approaches that invoke the uniform opera- sources used to construct residential units, tion of unilineal descent principles to but variation within classes was as great as account for residential diversity gain cre- the variations between them (Levi dence only if the residential units possess- 1993:229). On the other hand, the classifica- ing the largest numbers of structures at a tion did help to systematize observations site also exhibit the greatest temporal on temporal dynamics. Although excava- depth. This simply did not hold true at tion and surface collection data indicated San Estevan, the site’s Large Composite that over 80% of San Estevan’s domestic ar- Groups providing a case in point. Taken as chitecture possessed Late Classic occupa- a whole, the construction histories charac- tion (Levi 1993:113), construction sequences terizing San Estevan’s suite of residential often differed dramatically among San Es- classes suggested that households had tevan’s residential classes. Isolate and been variously organized in response to a Paired Platform classes showed the most far more nuanced “local environment.” In internally variable patterns. Of the Focus order to make the connection between Groups tested at the site, half showed households and their immediate environs, lengthy occupations extending from the it was necessary to model how local condi- Early Classic through the Late Classic, tions, the context of residence, had varied while the remainder originated in the Late as well. Classic period. Both early and later variants consolidated into Focus Group layouts Household and Community at San Estevan within a single ceramic period. This fairly rapid process was apparently offset by a Each of San Estevan’s domestic groups low ceiling upon group expansion, how- had existed within a context that could be ever, and Focus Groups never possessed framed at two scales. First, any particular more than four structures, regardless of group had occupied a particular locality de- temporal duration. San Estevan’s Basal fined by the presence of other groups in its Platform Groups, averaging four structures immediate vicinity. The immediate context per unit, evinced similar constraints upon of a group therefore could be described in group expansion but consistently showed terms of a given tract of land and the array great temporal depth. Comprising the com- of domestic forms that land supported. San munity’s most enduring residential loci, Estevan’s survey localities provided the these groups were founded no later than principal means to assess context at this the end of the Late Preclassic and were con- scale, and there were 17 in the sample that 128 LAURA J. LEVI possessed visible remains of residential set- cies for residential classes could be deter- tlement. Assessments of the residential as- mined by multiplying actual frequencies by semblages characterizing each of these lo- numerical ranks. It was then possible to calities began with tabulations of the characterize the context of any given unit frequencies of occurrence of specific resi- according to the spatial preponderance of dential classes. Raw frequencies, however, residential classes in its immediate envi- could not convey how the residential units rons. A particular residential class was con- typical of some classes clearly occupied sidered to dominate a locality if its more three-dimensional space than those of weighted frequency was at least twice as others. Although these larger units might large as the sum of the weighted frequen- occur in relatively low frequencies within a cies of other classes (Table 1). given area, they nevertheless managed to Obviously, many residential units at San dominate that area spatially. In order to bet- Estevan possessed comparable contexts by ter capture this spatial dimension, residen- virtue of their presence in the same survey tial classes were next ranked according to blocks. Units located in widely disparate their average architectural volume. On av- areas of the site, however, also could share erage, Large Composite Groups were volu- important quantitative and qualitative sim- metrically larger than other kinds of resi- ilarities in the residential makeup of their dential units at San Estevan, and their rank respective localities. For example, some lo- was arbitrarily set at 100. This figure al- calities evinced a similarly broad range of lowed the ranks of San Estevan’s remaining residential classes, with no one class residential classes to be computed as a per- achieving dominance (Dominant Class ϭ centage of the architectural volume of NONE; Fig. 4). Other areas, while support- Large Composite Groups. For any particu- ing a diverse residential assemblage, nev- lar site area, therefore, weighted frequen- ertheless showed a preponderance of either

TABLE 1 Raw and Weighted Frequencies (f and wf) of Residential Classes by Survey Localities

BPG FG PP I LCOMP Surveyed Dominant localities E fwfE fwfE fwfE fwfE fwfclass

SB1 2 4 344 2 4 68 1 6 36 BPG SB2 1 1 17 1 5 75 1 7 42 PP/I SB3 1 1 86 3 3 51 3 45 5 30 NONE SB6 1 15 5 30 2 4 400 LCOMP SB9 1 15 2 12 PP/I SB10 1 86 3 51 2 30 1 6 NONE SB11 1 1 15 5 30 1 1 100 LCOMP SB12 1 1 17 3 45 PP/I SB13 1 1 6 PP/I SB15 1 86 2 30 4 24 NONE SB16 1 3 45 2 4 24 PP/I SB17 2 34 1 15 FG SB18 1 1 15 1 1 6 PP/I SB19 1 17 3 45 1 6 PP/I Central zone 3 258 1 2 30 2 12 BPG Martinez zone 2 2 200 LCOMP Chowacol zone 4 344 1 17 1 15 3 18 BPG Column totals 3 14 7 16 5 29 6 47 5 7

Determination of the dominant residential class in each locality has been made on the basis of weighted class frequencies. “E” indicates the number of groups in which structures were test excavated. This ﬁgure does not include groups where excavations tested only plazas or other extramural areas. Nor does “E” indicate instances where construction sequences were obtained through examination of looters trenches. PREHISPANIC MAYA RESIDENTIAL VARIATION 129

FIG. 4. Example of a settlement locality where no single residential class dominates the assemblage (Dominant Class ϭ “NONE”).

Basal Platform Groups or Large Composite attributed to the relatively small size of the Groups (Dominant Class ϭ BPG or survey sample. LCOMP; Figs. 5 and 6). In contrast to the On the whole, therefore, the majority of diversity manifest by these ﬁrst 3 kinds of San Estevan’s residential localities ap- localities, several site areas exhibited a peared to share a rather limited range of remarkably narrow range of residential residential patterns. Those patterns, in turn, classes, with Paired Platform Groups and suggested that certain areas of the settle- Isolates dominating the assemblages ment, by virtue of commonalities in resi- (Dominant Class ϭ PP/I; Fig. 7). Finally, of dential composition, may have afforded a the 17 localities examined, only 1 pos- comparable range of economic options to sessed a preponderance of Focus Groups San Estevan’s prehispanic domestic groups. and thus failed to show a residential as- In order to investigate this possibility, it semblage comparable to any other site was necessary to determine whether areas area. For the present, this anomaly must be that overtly shared similar residential as- 130 LAURA J. LEVI

FIG. 5. Example of a settlement locality where Basal Platform Groups dominate the residential assemblage (Dominant Class ϭ “BPG”). semblages also exhibited similar spatial re- practices (Levi 1996). Ultimately, however, lationships to spheres of economic opportu- both the diversity and kinds of residential nity within the context of the wider com- classes found at any given locality speciﬁ- munity. At this second scale, the residential cally correlated with proximity to nodes of composition of settlement zones could be monumental architecture (Fig. 8). Areas broadly linked to settlement ecology, with dominated by Basal Platform Groups never clear implications for prehispanic agrarian appeared further than 0.50 km from one of PREHISPANIC MAYA RESIDENTIAL VARIATION 131

FIG. 6. Example of a settlement locality where Large Composite Groups dominate the residential assemblage (Dominant Class ϭ “LCOMP”).

San Estevan’s three nodes of monumental matic shift, and Large Composite Groups architecture. Areas characterized by highly dominated the landscape. diverse residential assemblages and the absence of a dominant residential class only THE HOUSEHOLD PARADOX occurred within a 0.7-km radius of monumental precincts. Beyond this point, and ex- While a notion of household agency mo- tending for roughly 1/2 km, residential as- tivated much of the San Estevan research, semblages were marked both by a loss of this idea bestows what Gillian Hart (1992) diversity and by the preponderance of has called an “imagined unity” upon some- Paired Platform Groups and Isolates. Mov- thing that is actually an aggregate of indi- ing even further away from monumental viduals. Hart and others have argued quite precincts to the margins of San Estevan’s persuasively that viewing the household in upland zones, there was yet another dra- terms of the separate and often conﬂicting 132 LAURA J. LEVI

FIG. 7. Example of a settlement locality where Paired Platforms and Isolates dominate the residential assemblage (Dominant Class ϭ “PP/I”). interests of its constituents is critical to un- making is embedded in broader institu- derstanding larger social processes (e.g., tional settings (de Montmollin 1987; Folbre 1988; Hendon 1996). Yet a rationale Halperin 1985). A second contradiction for the variation evidenced by San Este- arises from this critique. Only by granting van’s residential architecture emerged only agency to San Estevan’s prehispanic house- from a model of households as decision- holds could clear limits to their decision- making entities, negotiating strategies of re- making autonomy then be discerned. Quite cruitment and production. Still other schol- simply, certain organizational strategies ars have questioned the validity of proved likely, unlikely, or impossible for attributing agency to either households or San Estevan households, depending on the individuals, pointing out that such decision settlement localities they occupied. One PREHISPANIC MAYA RESIDENTIAL VARIATION 133

FIG. 8. Distribution of survey localities in relation to nodes of monumental architecture. Localities have been grouped according to the dominant residential class characterizing each.

final contradiction more fully conjures the In short, where do we locate the systemic specter of systemic connections between properties of the San Estevan community? domestic and political life. Although the Answers to these questions require the analysis was able to show that the organiza- combined input of an eclectic array of tional strategies pursued by San Estevan’s scholars whose works converge upon the households had varied in relation to their related subjects of institutions, power, and productive activities, it also underscored agency. the influences exerted by the community’s most prominent political arenas. Institutions and Power The San Estevan study thus reveals a critical paradox delineated by three questions It is not my intention to provide a com- to be addressed below. First, why may prehensive history of the term “institution” households behave as if they were individ- nor even a thorough critique of definitions. uals, when clearly they are not? Second, Rather, I limit the discussion to a few theo- why must households be credited with retical developments relevant to considera- agency before limits to their autonomy can tions of institutional power. By and large, be perceived? And, third, if agency is not theoretical exploration of institutions has solely the prerogative of individuals—if been conducted outside of anthropology, there are other “active entities” in society primarily in sociology, but also in history, (Johnson 1989:208)—then what is it in this economics, and psychology. Until the prac- complicated interplay of individuals, tice theorists, 20th-century sociological tra- households, and suprahousehold entities ditions were dominated by the tendency to that contributes to the organization and erect clear conceptual boundaries between inter-relatedness of the entire social field? people and institutions. This tendency is 134 LAURA J. LEVI best exemplified in the writings of Talcott power, Giddens adopted an “epoché” ap- Parsons (1951, 1960), who maintained that proach á la Braudel (1973) that collapses in- people form collectivities while institutions stitutions into “rules and resources,” effec- consist of formal rules that delimit the roles tively neutralizing both history and people people assume and the actions they take (Giddens 1984:28–34). (Parsons 1951:39–40). Historical depth was By contrast, there have been very few implied: Institutions endure, people do not. generalizing treatments of institutional People, as a result, became the ahistorical phenomena by anthropologists who, in- subjects of network analysis (Coleman et al. stead, have opted to focus on how specific 1957) and interaction theory (Goffman institutions work. In rare, early exceptions, 1963, 1967), whereas the study of institu- anthropological discussions show a clear tions fell under the purview of an histori- indebtedness to sociology’s Parsonian tra- cally grounded macrosociology with strong dition (e.g., Lowie 1948:3). Later commen- Weberian roots. Whether practice theorists tary, however, points to an altogether con- have managed to successfully breach the trary assessment best illustrated in the divide between people and institutions work of Mary Douglas. According to Dou- through poststructuralist bridging is cer- glas (1986:46), institutions are “legitimate tainly open to debate. With the notion of social groups.” The drawbacks to this defi- habitus, for example, Pierre Bourdieu re- nition are obvious. Outwardly, at least, it of- constituted people as historical subjects fers a synchronic perspective and reflects but, in doing so, preserved their opposition anthropology’s traditional ambivalence to- to institutions [Bourdieu 1990 (1980):56–58]. ward history. It is also inherently reduction- Habitus is historically and experientially ist, the emphasis on “legitimacy” revealing derived knowledge embodied by an indi- anthropology’s deference to—and unwill- vidual. The content of habitus, although se- ingness to unpack—the “emic” point of lectively incorporating a history that may view. If the emphasis shifts to “social extend far into the past, nevertheless is in- groups,” however, the definition allows for formally configured during the course of an the possibility that institutions are situated individual’s life span and remains in large within groups of people rather than the re- part unexpressed or unrecognized. Al- verse. In consequence, institutions have the though institutions are also historically de- potential to be understood as problems not rived knowledge, they perpetuate aware- only of enduring structure but of emergent ness of their pasts in the present through organization, as well (Kowalewski 1994:1). recognized symbolic schemes that people The productivity of an organizational ap- may consciously elucidate and manipulate proach to institutions has been amply illus- (Bourdieu 1991:105–106). Ultimately, there- trated by historian Michael Mann in his fore, institutions stand external to individu- treatise on social power. In Mann’s view, als, shaping the determinacy of habitus to the linkages forged between people always the extent that people can be considered to involve power relationships, although the “inhabit” institutions [Bourdieu 1990 sources, effects, and expressions of power (1980):57]. Anthony Giddens (1979, 1984) may vary (Mann 1986:1–33). Gathering the also seemed to have some difficulty defin- insights of Parsonians, Weberians, histori- ing institutions in terms of people. Arguing cal materialists, and practice theorists, from his standpoint of the fundamental Mann described organizational power in “duality of structure,” he asserted the for- its collective, distributive, extensive, inten- mer to be those “practices which have the sive, authoritative, and diffuse aspects. greatest space–time extension” (Giddens Collective and distributive powers refer to 1984:17). However, to explore institutional sources—power in as opposed to power PREHISPANIC MAYA RESIDENTIAL VARIATION 135 over. Extensive and intensive refer to effec- otherwise unavailable should its members tive strength as measured, respectively, by act alone. Because this power is contingent size and areal extent of the relevant popu- upon the existence of the group, and dissi- lation and by degree of “mobilization” or pates when the group breaks apart, it is not “commitment” (Mann 1986:7). Finally, the erroneous to grant the household a kind of last two terms implicate the manner in autonomy, rooted in but distinct from the which power relationships are expressed separate interests of its members. Although and experienced, either authoritatively not an individual, the household is never- through “definite commands and con- theless an actor. scious obedience” or diffusely through If the household can act by virtue of its shared “social practices . . . not explicitly collective powers, then the actions of its commanded” (Mann 1986:8). It should be members must achieve some outward sem- noted that although Mann defined his six blance of coherence and coordination. Over terms according to three oppositional pairs, the years, there have been many attempts to no one pairing truly delineates a specific model this phenomenon, from Durkheim’s dimension of contrast along which organi- collective conscience and the practice theo- zations will vary. For example, all organiza- rists’ habitus to economists’ formulations of tional forms necessarily will rest upon col- the household’s joint utility function. lective power while the degree to which Mann’s intensive, authoritative, and diffuse they manifest distributive powers may dif- powers also constitute a response to the fer. Similarly, some organizations may ex- problem of collective action and the indi- hibit great degrees both of intensive and vidual compliance it requires. It is Mary extensive powers, while others will show Douglas, however, who has demonstrated an inverse relationship between the two. how these aspects of organizational power Mann’s contribution is twofold. First, an stem from a common cultural fund that is institution gains its social presence or both epistemic and conventional. Compar- nodality as much from power relationships ing diverse institutions, from small descent generated by people as from knowledge groups to international scientific communi- structured for people. Second, the social ties, Douglas (1986) explored how they force of an institution may be gauged by the form and reform around rationales that ap- organizational power it summons in rela- pear “natural,” morally justified, and his- tion to other institutions. Together, these torically legitimated. The individuals com- two points help to resolve some of the ap- prising an institution can and do think and parent contradictions in the San Estevan act independently, but communication and findings. the achievement of goals are enormously expedited when expressed in terms of these Agency and Institutional Power established logics. An institution gains at least some of its power through this appeal Why households can be profitably stud- of efficiency, providing a ready set of con- ied “as if” they were individuals rests prin- ventions that its constituents may use to or- cipally on an understanding of collective ganize their experiences and actions. power. Collective power is an emergent As institutions, then, San Estevan’s property of all institutions in that individu- households held certain organizational als by means of “cooperation . . . enhance powers to define and justify goals of pro- their joint power over third parties or over duction, consumption, and reproduction. nature” (Mann 1986:6). Through the com- Although the site’s architectural record di- bined efforts of its members, therefore, the rectly reflects neither the goals nor the pre- household wields power that would be cise strategies households adopted to 136 LAURA J. LEVI achieve their ends, it does provide clues to of fixity, modeling those relationships as how those strategies were variously ex- historically and situationally specific. Soci- pressed. San Estevan’s mechanisms of ety’s institutions stand overlapping and structure incorporation and the residential “counterpoised” (Crumley 1995:3). They variants they produced must be considered are comprised of diverse and intersecting different sets of conventions that at one time memberships with divergent and intersect- framed different kinds of household experi- ing goals. More powerful institutions, ence and action. However, as the San Este- therefore, do not always and necessarily van findings suggest (and as critics of subsume less powerful ones. Very often household decision making duly note), the they function to influence and constrain power exercised by households through the and can be influenced and constrained in organization of their membership is not es- turn. This heterarchical understanding of pecially formidable. There are other institu- organizational power helps to resolve the tions in society that command far more second contradiction of the San Estevan power. We sense this greater power because analysis. Household agency may not be ob- it embraces larger numbers of individuals viated but shaped by other, more powerful and a broader areal expanse. In Mann’s institutions; and it is only by seeking the (1986:3) terminology, these “sociospatial” limits to household agency that the greater consequences reveal power in its distribu- organizational powers of these other insti- tive and extensive aspects. At San Estevan, tutions become known. there were institutions of governance, phys- ically anchored in monumental precincts, DISCUSSION: SPACE AND “THE that had far greater power than any individ- SYSTEM” ual or household in the community. Their greater power was made manifest through In this article, I have stressed institutions the ability to limit the organizational op- and their organizational entailments in tions that households could pursue at any order to make the point that agency and particular settlement locality. Yet, it would power are emergent capacities of organiza- be a mistake to argue that these larger, more tion rather than intrinsic characteristics of dominant institutions completely overpow- persons (see Dobres and Robb 2000 and ered San Estevan’s households. They influ- Sweely 1999 for summaries of the range of enced only the range of options available. debate on loci of agency and power). For ar- Individual households were left to work chaeologists, there are important theoretical through those options, with variable results and methodological consequences to this and differential success. position. Theoretically, our ability to model The San Estevan findings highlight that broad social processes is not automatically society’s institutions are not consistently or directly enhanced by breaking house- arranged in a kind of nested hierarchy, with holds apart into the individual actions of the most powerful fully determining the their constituents. Gillian Hart recom- nature and powers of lower institutional or- mended this tactic as an expedient to con- ders. A more apt rendering comes from structing politicized views of the house- Carol Crumley’s (1979, 1995) model of het- hold. According to Hart (1992:122), erarchy, which imbues organization with a household institutions are best understood dynamism lacking in conceptions of society to be “dense bundlings of rules, rights, and that are more rigidly hierarchical. Although obligations,” all of which can be “subject to not discounting the existence of hierarchi- contestation” (Hart 1992:122). A politicized cal relationships of dominance and subor- household, therefore, is one in which the dination, heterarchy anticipates an absence sources of power for any given member PREHISPANIC MAYA RESIDENTIAL VARIATION 137 may at times differ from and compete with might be overlooked should we lose sight those of other members. An essentialist per- of the system. Social systems exhibit pow- spective, on the other hand, would construe ers well in excess of those attributable to in- institutions as neatly bounded, inwardly stitutions. They possess a persistent, almost coherent, and outwardly cohesive. Curi- dogged, durability that resists all efforts to ously, it was Robert Netting’s (1993) essen- politicize their constituent institutions, to tialized view of political formations (and situate those institutions within groups of not households) that justified his efforts to people, and to imbue those people with or- distance sources of household variation ganizational powers. Practice theorists from political influence. Institutions of gov- have suggested that the persistent, structur- ernance, conceived to possess unitary goals ing components of any system reside in and invariant impacts, could in no way be regimes of knowledge (i.e., institutions) made to account for the variety of domestic that produce real material inequities among forms over which they presided. Clearly, an people and groups. I would argue that “the essentialist perspective is unrealistic. But system” has a far greater material presence. given the organizational powers that accrue Taken individually, institutions achieve to their memberships, institutions wield a substance over and above the visible eco- social force that would be incalculable nomic disparities they might produce. should we focus too narrowly on the ac- Through collective power, institutional actions of individuals. I suggest, in contrast to tion summons conventional modes of rep- Hart, that we consider institutions to coa- resentation, and archaeologists should be lesce through “dense bundlings” of interac- alert to a potentially wide array of material tion among people who possess diverse or- conventions through which organizational ganizational affiliations. It then becomes affiliations are expressed and experienced. possible to politicize all institutions—to see But, additionally, the organizational pow- the tensions and contradictions of each— ers of institutions have decided spatial con- without losing sight of the ways in which sequences (Mann 1986:3). At San Estevan, organizational power motivates institu- for example, architectural conventions of tional action. What makes a politicized structure incorporation signaled group af- view of institutions and, especially, a politi- filiation while simultaneously carving out cized household both necessary and useful the spaces of domestic life. As organiza- is the realization that society’s most power- tions of people, therefore, institutions craft ful institutions are exposed through the ac- their own spatial realities (cf. Kus 1983). tions taken by its least powerful ones. Taken together, however, San Estevan’s in- Methodologically, therefore, institutions stitutions and their respective powers were comprise some of the most salient units of realized not simply as discretely bounded analysis accessible to archaeologists. landmarks on the terrain, but through the The San Estevan findings underscore the way the entire community was spatially need for a more flexible, heterarchically constituted. based understanding of the institutional Accordingly, the San Estevan analysis of- media of action in society. Not surprisingly, fers another vantage on the locus and deter- with an awareness of the cross-cutting or- minative dynamic of “the system.” In a ganizational powers of institutions comes very concrete sense, the systemness of a so- the risk of overplaying agency while under- ciety resides in the space forged jointly by conceptualizing the system as a whole. particular institutions possessed of variable Given the growing emphasis on agency in goals and characterized by variable degrees anthropological thinking, it seems appro- and kinds of organizational power. This priate, if unfashionable, to consider what spatial presence, perhaps more than knowl- 138 LAURA J. LEVI edge, is what makes societies so resilient in Arnold, Jeanne, and Anabel Ford the face of efforts to change them. Knowl- 1980 A statistical examination of settlement pat- edge may be readily contested, negotiated, terns at Tikal, Guatemala. American Antiquity and reformulated. Historically, space has 45:713–726. been a far more invidious enemy of change, Arnould, Eric requiring tangible (and often vast) inputs of 1984 Marketing and social reproduction in Zinder, labor to alter flows of people and their Niger Republic. In Households: Comparative and historical studies of the domestic group, edited by knowledge and resources (Harvey 1990). R. Netting, R. Wilk, and E. Arnould, pp. 130– Space, therefore, has a duality all its own, 162. Univ. of California Press, Berkeley. part enduring structure and part organiza- Ashmore, Wendy, and Richard Wilk tional medium (Soja 1985). It is in this latter 1988 Household and community in the Meso-amer- capacity that we can refer to space as a crit- ican past. In Household and community in the ical component of “organizational process,” Mesoamerican past, edited by R. Wilk and W. a phrase used by Eric Wolf (1990:591) to Ashmore, pp. 1–27. University of New Mexico communicate the power of the system as a Press, Albuquerque. whole. Bachnik, Jane 1983 Recruitment strategies for household succes- sion: Rethinking Japanese household organi- ACKNOWLEDGMENTS sation. Man 18:160–182.

Grants from the National Science Foundation Blanton, Richard (BNS8910970) and the Wenner Gren Foundation for 1995 The cultural foundations of inequality in Anthropological Research funded the investigations at households. In Foundations of social Inequality, San Estevan that are discussed in this article. Many edited by T. Price and G. Feinman, pp. 105– thanks to Bruce Moses for producing computerized 127. Plenum, New York. images of the San Estevan survey maps. Once again, I Bourdieu, Pierre extend my gratitude to the many individuals associ- 1990 (1980) The logic of practice. Stanford Univ. Press, ated with the San Estevan Project: Dionicio Pech, Stanford, CA. Alfredo Pech, Noel Pech, Argelia Martínez, Ovel Martínez, Vildo González, Amir González, Omar 1991 Language and symbolic power. Harvard Univ. Vásquez, Armando Ya, Nellie Donaldson, Julian Pat, Press, Cambridge, MA. Daniel Pat, Jeffrey Jones, Daniela Triadan, Stacey Ruet- Braudel, Fernand ter, Jeffrey Baker, John Rose, Kevin Kuykendall, Mari- 1973 The Mediterranean and the Mediterranean world anna Wetters, Cynthia Robin, Colleen Gleason, Nor- in the age of Philip II. Fontana, London. man Hammond, T. Patrick Culbert, and Eric and Madelaine Levi. Bullard, William 1965 Stratigraphic excavations at San Estevan, northern British Honduras. Royal Ontario Museum, Art REFERENCES CITED and Archaeology Occasional Paper 9, Univ. of Toronto, Toronto. Abrams, Elliot Cain, Mead 1987 Economic specialization and construction personnel in Classic Period Copan, Honduras. 1988 The material consequences of reproductive American Antiquity 52:485–499. failure in rural South Asia. In A home divided: Women and income in the third world, edited by Allison, Penelope D. Dwyer and J. Bruce, pp. 20–38. Stanford 2000 Introduction. In The archaeology of household ac- Univ. Press, Stanford, CA. tivities, edited by P. Allison, pp. 1–18. Rout- ledge, London. Carmean, Kelly Alexander, Rani 1991 Architectural labor investment and social stratiﬁcation at Sayil, Yucatan, Mexico. Latin 2000 Mesoamerican house lots and archaeological American Antiquity 2:151–165. site structure: Problems of inference in Yax- caba, Yucatan, Mexico, 1750–1847. In The ar- Coleman, James, E. Katz, and H. Menzel chaeology of household activities, edited by 1957 The diffusion of an innovation among physi- P. Allison, pp. 78–100. Routledge, London. cians. Sociometry 20:253–270. PREHISPANIC MAYA RESIDENTIAL VARIATION 139

Crumley, Carol household activities, edited by P. Allison, pp. 1979 Three locational models: An epistemological 142–161. Routledge, London. assessment for anthropology and Archaeol- Halperin, Rhoda ogy. In Advances in archaeological method and 1985 Households as units of analysis in history and theory, edited by M. B. Schiffer, Vol. 2, pp. 141– anthropology. Reviews in Anthropology 12:346– 173. Academic Press, New York. 352. 1995 Heterarchy and the analysis of complex soci- Hammond, Norman eties. In Heterarchy and the analysis of complex societies, edited by R. Ehrenreich, C. Crumley, 1975 Lubaantun, a classic Maya realm. Monographs of and J. Levy, pp. 1–6. Archaeological Papers of the Peabody Museum, No. 2. Harvard Univ., the American Anthropological Association. Cambridge, MA. No. 6, American Anthropological Association, Hammond, Norman (Editor) Washington, DC. 1973 British Museum—Cambridge University Coro-zal de Montmollin, Olivier Project, 1973 interim report. Cambridge Univ. 1987 Forced settlement and political centralization Centre of Latin American Studies, Cambridge, in a classic Maya polity. Journal of Anthropologi- UK. cal Archaeology 6:220–262. Hart, Gillian Dobres, Marcia-Anne, and John Robb 1992 Imagined unities: Constructions of “the house- 2000 Agency in archaeology: Paradigm or plati- hold” in economic theory. In Understanding tude? In Agency in archaeology, edited by M. economic process, edited by S. Ortiz and S. Lees, Dobres and J. Robb, pp. 3–17. Routledge, Lon- pp. 111–129. Monographs in Economic An- don. thropology, No. 10, Univ. Press of America, Lanham. Douglas, Mary Harvey, David 1986 How institutions think. Syracuse Univ. Press, Syracuse, NY. 1990 The condition of postmodernity. Blackwell, Lan- Folan, William, L. Fletcher, and E. Kintz ham. Haviland, William 1979 Fruit, ﬁber, bark, and resin: The social organization of a Mayan urban center. Science 1970 Tikal, Guatemala and Mesoamerican urban- 204:697–701. ism. World Archaeology 2:186–197. Folbre, Nancy Hendon, Julia 1988 The black four of hearts: Toward a new para- 1991 Status and power in classic Maya society: An digm of household economics. In A home di- archaeological study. American Anthropologist vided: Women and income in the third world, 93:894–918. edited by D. Dwyer and J. Bruce, pp. 248–262. 1996 Archaeological approaches to the organization Stanford Univ. Press, Stanford, CA. of domestic labor: Household practice and do- Giddens, Anthony mestic relations. Annual Review of Anthropology 1979 Central problems in social theory. Univ. of Cali- 25:45–61. fornia Press, Berkeley and Los Angeles. Hirth, Kenneth 1984 The constitution of society. Univ. of California 1993 The household as an analytical unit: Problems Press, Berkeley and Los Angeles. in method and theory. In Prehispanic domestic Gillespie, Susan units in Western Mesoamerica: Studies of the household, compound, and residence, edited by R. 2000 Rethinking ancient Maya social organization: Santley and K. Hirth, pp. 21–36. CRC Press, Replacing “lineage” with “house.” American Boca Raton. Anthropologist 102(3):467–484. Johnson, Michael Goffman, Erving 1989 Conceptions of agency in archaeological inter- 1963 Behavior in public places: Notes on the social orga- pretations. Journal of Anthropological Archaelogy nization of gatherings. The Free Press, New 8:189–211. York. Johnston, Kevin, and Nancy Gonlin 1967 Interaction ritual: Essays on face to face behavior. 1998 What do houses mean?: Approaches to the Anchor Books, Garden City. analysis of Classic Maya commoner resi- Goldberg, Marilyn dences. In Function and meaning in Classic Maya 2000 Spatial and behavioral negotiation in Classical architecture: A symposium at Dumbarton Oaks, Athenian city houses. In The archaeology of 7th and 8th October, 1994, edited by S. Houston, 140 LAURA J. LEVI

pp. 141–185. Trustees for Harvard Univ., Mann, Michael Washington, DC. 1986 The sources of social power, Vol. 1. Cambridge Killion Thomas Univ. Press, Cambridge, UK. 1990 Cultivation intensity and residential site McAnany, Patricia structure: An ethnographic examination of 1992a Agriculture tasks and tools: Patterns of stone peasant agriculture in the Sierra de los Tuxt- tool discard near prehistoric Maya Resi- las, Veracruz, Mexico. Latin American Antiq- dences bordering Pulltrouser Swamp, Belize. uity 1:191–215. In Gardens of prehistory: The archaeology of set- 1992 Residential ethnoarchaeology and ancient site tlement agriculture in greater Mesoamerica, structure: Contemporary farming and prehis- edited by T. Killion, pp. 184–213. The Univ. of toric settlement agriculture at Matacapan, Ver- Alabama Press, Tuscaloosa, AL. acruz, Mexico. In Gardens of prehistory: The ar- 1992b A theoretical perspective on elites and the chaeology of settlement agriculture in greater economic transformation of classic period Mesoamerica, edited by T. Killion, pp. 119–149. Maya households. In Understanding economic The Univ. of Alabama Press, Tuscaloosa. process, edited by S. Ortiz and S. Lees, pp. Killion, Thomas, Jeremy Sabloff, Gair Tourtellot, and 85–103. Monographs in Economic Anthro- Nicholas P. Dunning pology, No. 10, University Press of America, 1989 Intensive surface collection of residential clus- Lanham. ters at terminal classic Sayil, Yucatan, Mexico. 1995 Living with the ancestors: Kinship and kingship in Journal of Field Archaeology 16:273–294. ancient Maya society. Univ. of Texas Press, Kowalewski, Stephen Austin. 1994 Local power and external power in Oaxaca: The Netting, Robert long view. Paper presented at the Annual Meet- 1968 Hill farmers of Nigeria: Cultural ecology of the Ko- ing of the American Anthropological Associa- fyar of the Jos Plateau. Univ. of Washington tion, Atlanta, December 1. Press, Seattle. Kurjack, Edward 1993 Smallholders, householders: Farm families and the 1974 Prehistoric lowland Maya community and social ecology of intensive, sustainable agriculture. Stan- organization: A case study at Dzibilchaltun, Yu- ford Univ. Press, Stanford, CA. catan, Mexico. Middle American Research Insti- Netting, Robert, Richard Wilk, and Eric Arnould tute, Publication 38, Tulane Univ., New Or- leans, LA. 1984 Introduction. In Households: Comparative and historical studies of the domestic group, edited by Kus, Susan R. Netting, R. Wilk, and E. Arnould, pp. xiii– 1983 The social representation of space: Dimension- xxxviii. Univ. of California Press, Berkeley. ing the cosmological and quotidian. In Archae- Ortner, Sherry ological hammers and theories, edited by J. Moore and A. Keene, pp. 278–304. Academic Press, 1994 (1984) Theory in anthropology since the six- New York. ties. In Culture/power/history: A reader in contem- Lawrence, Denise, and Setha Low porary social theory, edited by N. Dirks, G. Eley, and S. Ortner, pp. 372–411. Princeton Univ. 1990 The built environment. Annual Review of An- Press, Princeton, NJ. thropology 19:453–505. Parsons, Talcott Levi, Laura 1951 The social system. The Free Press, New York. 1993 Prehispanic residence and community at San Este- van, Belize. Ph.D. dissertation, Department of 1960 Structure and process in modern societies. The Anthropology, University of Arizona. Free Press, New York. 1996 Sustainable production and residential varia- Sanders, William tion: A historical perspective on lowland Maya 1989 Household, lineage, and state at eighth- domestic economy. In The managed mosaic: An- century Copan, Honduras. In House of the ba- cient Maya agriculture and resource use, edited cabs, Copan, Honduras, edited by D. Webster, by S. Fedick, pp. 92–106. Univ. of Utah Press, pp. 89–105. Studies in Precolumbian Art and Salt Lake City. Archaeology, No. 29, Dumbarton Oaks Re- Lowie, Robert search Library, Washington, DC. 1948 Social organization. Rinehart and Co., Inc., New Sheehy, James York. 1991 Structure and change in a late classic Maya do- PREHISPANIC MAYA RESIDENTIAL VARIATION 141

mestic group at Copan, Honduras. Ancient Wilk, Richard Mesoamerica 2:1–19. 1988 Maya household organization: Evidence and Sjoberg, Gideon analogies. In Household and community in the 1960 The pre-industrial city: Past and present. The Free Mesoamerican past, edited by R. Wilk and W. Press, Glencoe, IL. Ashmore, pp. 135–151. Univ. of New Mexico Soja, Edward Press, Albuquerque. 1985 The spatiality of social life: Towards a trans- 1991 Household ecology: Economic change and domestic formative retheorisation. In Social relations and life among the Kekchi Maya in Belize. The Univ. spatial structures, edited by D. Gregory and of Arizona Press, Tucson. J. Urry, pp. 90–127. St. Martin’s Press, New Wilk, Richard, and Robert Netting York. Sweely, Tracy 1984 Households: Changing forms and functions. In Households: Comparative and historical studies 1999 Introduction. In Manifesting power: Gender and of the domestic group, edited by R. Netting, R. the interpretation of power in archaeology, edited Wilk, and E. Arnould, pp. 1–28. Univ. of Cali- by T. Sweely, pp. 1–14. Routledge, London. fornia Press, Berkeley. Trigger, Bruce Wolf, Eric 1989 History and contemporary American archaeology. In Archaeological thought in America, 1990 Distinguished lecture: Facing power—Old in- edited by C. Lamberg-Karlovsky, pp. 19–34. sights, new questions. American Anthropologist Cambridge Univ. Press, Cambridge, UK. 92:586–596.