Detecting political stratification

Political stratification concerns the unequal distribution of political rewards and inequalities in access to political offices. How to define presence or absence of political stratification is a secondary methodological question here, rather than a primary substantive focus as in cultural evolutionary studies of the origins of ranking or (economic) stratification (Service 1971; Carneiro 1981; Earle 1978). For the Rosario polity, I begin with the working assumption that it had clear political stratification: a regime featuring political inequalities, with privileged access to political offices for some. A basic division into rulers and ruled, leaders and subjects, or elite and commoner groups characterized most ancient Mesoamerican polities, at least from the Late Formative Period onward. There is no reason to believe that the Rosario polity was exceptional. Beyond sweeping generalizations, three specific lines of evidence locally sustain this assumption. First, the clear presence in the settlement record of a political settlement-hierarchy (an uneven distribution of civic-ceremonial facilities) means that the residents associated with differently ranked centers have differing access to political offices and activities. Second, considerable variability exists in domestic architecture attributes indicating hierarchical inequalities in sociopolitical status. In simplest terms, there was probably a set of sumptuary status-related rules about the appearance of dwellings. Third, differential distribution of special luxury items well meets the expectations for a stratified sociopolitical system. These items include polychrome pottery, fine imported pottery, figurines, jade-greenstone celts, seashell pendants, and so forth, all viewed as sumptuary sociopolitical status markers. Given a defensible assumption that there was political stratification (and not egalitarianism) in the Rosario polity, the problem then becomes one of investigating its nature.

Pyramidal versus hierarchical regimes

The continuum between pyramidal and hierarchical political regimes deals with both the decision-making and decision-implementing aspects of a political system. The practical reason for treating these aspects jointly is that archaeological evidence in the Rosario Valley does not allow anything like a distinction between policymaking (politics) and policy-implementation (administration or bureaucracy) to be made. There are no executive buildings and artifacts as opposed to administrative buildings and artifacts. Taking the argument onto a more interpretive plane, Maya ethnohistoric sources suggest that in the Postclassic Period there were no clearly separable groups of people involved in policy-making as opposed to policyimplementation (with the exception of menial administrative “flunkies ” such as the tupiles mentioned in a few Yucatec sources). More precisely, this assertion is based on an ethnohistorical survey covering a variety of Contact Period Maya polities (the Yucatec Maya-de Montmollin 1980; the Guatemala Highland Quiche Mayade Montmollin 1982b; and the Chiapas Highland Maya-de Montmollin 1979c). Once again applying historical-evolutionary logic, one would not expect earlier periods to feature fully professional bureaucratic structures (after Weber, see Gerth and Mills eds. 1946: ch. 8). This logic is a form of substantivism applied to politics instead of economics. It resembles Giddens' discontinuist perspective on the development of the state, one which draws a sharp contrast in terms of bureaucratic structure and efficacy between traditional states and modern nation-states (Giddens 1985). Thus, the lack of evidence for pure administrators or bureaucrats in the Rosario Valley polity may be taken to reflect a genuine absence of such specialized personnel.

With a theoretical framework in hand, a range of methodological problems need to be tackled before analytical procedures can be developed and applied towards the goal of characterizing political structure and organization in the Rosario polity. But discussion of methodological problems and application of analytical techniques only make sense with fairly specific reference to the properties of a given archaeological settlement record. To introduce the Rosario settlement record, I now provide a brief sketch of the Rosario polity in its local and wider Maya context.

The Rosario polity occupies a small valley within the Upper Grijalva Tributaries of Chiapas, Mexico (Figures 1 and 2). The Upper Tributaries lie on the southwest edge of the tropical rainforest Usumacinta Lowlands, a core area of Classic Period (AD 300–950) Maya political and cultural development, with major centers such as Yaxchilan, Bonampak, and Piedras Negras (Figure 1). The surveyed part of the Rosario Valley covered just under 53 sq km, estimated to have been almost the entire extent of the Rosario polity's densely settled core. Within this, there was an estimated (maximum) population of 20,000 in the Late/Terminal Classic Period, AD 700–950 (Figures 3 and 4). Adding on (cursorily examined) peripheral area gives a total polity area of 100-150 sq km. Consisting primarily of rugged hills separating the Rosario polity from neighboring polities, most of the peripheral area was very sparsely settled. The Rosario polity core has several nested districts. Two sections correspond to upper and lower valley halves. Seven pockets consist of further divisions of the sections, corresponding to small sub-basins (Figure 5). A political settlement-hierarchy has four discrete levels of centers that include civicceremonial plazas (Figure 5, Table 2).

Social statics and dynamics

The tendency of polities to segment or cohere is an aspect of political organization (dynamics). At the segmenting end of a continuum, polities show a tendency to break apart into constituent districts, while at the non-segmenting end, polities are less prone to break apart, maintaining their cohesiveness over long(er) periods (Chapter 2). It is particularly difficult to determine where the Rosario polity fits along this continuum, and to answer the fifth research question concerning the tendency of Maya polities to segment. In working around the edges of the problem, I examine several themes (listed in decreasing order of generality): structure (social statics) versus organization (social dynamics) and how one moves from the former to the latter, inferring diachronic (possibly cyclical) developments from synchronic evidence, and the relation of the Rosario polity's breakdown to the Classic Maya Collapse.

To reiterate an earlier point (Chapter 2), I draw a distinction between organization and structure. Structure (social statics) consists of relatively enduring and abstract norms, principles, or institutions for arranging and regulating relations among societal actors. Organization (social dynamics) is a contingent situational set of activities and pragmatic behaviors on the part of societal actors (after Firth 1964). My prior analysis has concentrated on structure (Chapters 5–9), with a problem orientation tailored closely to the chronological possibilities and limitations of settlement data (Chapter 4). A climax-crash view of settlement development allowed interpretation of patterns detectable from single-period settlement maps, according to principles of political structure. To get at questions of organization with such material is exceedingly more difficult.

The study of political structure in ancient complex polities exercises a powerful attraction on archaeologists working in Mesoamerica and many other parts of the world. Intense interest and energy are invested in trying to describe and understand complex polities which are ancient or traditional in the sense that they often predate and always differ from modern European nation-states (Southall 1965; Giddens 1985). Before proceeding with a study of politics in one or more ancient complex polities, pausing to look at the general reasons for the attraction of such studies proves instructive since these reasons directly affect problem orientation, theory, and methodology. Reasons for studying ancient complex polities revolve around four broad themes: the evocation of great questions (origins of the state, complexity, civilization); the use of data and concepts from political anthropology; the problems encountered in constructing analogies (models); and the vigorous effort and reasoning required for constructing bridging arguments to link theoretical concepts and archaeological data. A number of approaches may be adopted with reference to these four broad themes. My own approach is one of bias in favor of anthropological archaeology which combines a comparative search for general principles with an (intellectual) respect for diversity in political structure and behavior. Such an approach lies towards the relatively more fruitful middle ground of a spectrum. At one end are the highly generalizing approaches in the archaeology of ancient complex societies which deal in political universals and sweeping conclusions about the human condition. At the other end of the spectrum are the highly particularistic approaches which refuse to deal with any comparative generalization and limit themselves to descriptions of single ancient complex polities or cultures.

Differentiation

Degrees of replication in settlement patterns among the Rosario polity's centers and districts provide insights into degrees of differentiation (along the second subcontinuum contained in the segmentary-unitary continuum. Table 1). Presence or absence of replication (for centers and districts) is judged with reference to values arrived at on the various centralization measures (forced settlement, TDI, and TLI – Figure 15). Other available measures (for districts) are: territorial size, settlement density (and environmental composition), and tributary base size (TBS – Figure 15). A high degree of (horizontal) replication in centralization measures among all hierarchically equivalent units indicates a segmentary polity. Vertical replication of centralization measures across levels of a hierarchy (the PH) indicates segmentary structure, as does increasing degrees of centralization down the PH. Decreasing degrees of centralization down the PH indicate unitary structure. To a very great extent, the differentiation subcontinuum has an analytically residual quality with reference to the centralization subcontinuum. Many of the values examined with reference to differentiation have been generated while investigating centralization (Chapter 5).

At the outset, it appeared that a search for replication in internal layout and composition among sites (and wards) might be a useful way of examining degrees of differentiation in the Rosario polity (de Montmollin 1982a), but it soon turned out that this kind of analysis was inappropriate. Sites are too numerous and richly variable in their internal layout and composition to be effectively compared. Besides practical difficulties of too much detail, there are theoretical objections to the site scale.

Analytical tools

Several analytical tools help us to arrive at the archaeological measures needed for examining political structure in the Rosario polity. Narrowing the earlier focus on general methodological difficulties (Chapter 4), my methodologically oriented discussion of analytical tools is geared even more specifically to the Rosario settlement record's qualities and possibilities. Nevertheless, the choices faced and the logic used for constructing analytical tools are common to many settlement studies of politics in ancient complex polities. The needed analytical tools consist of: a territorial subdivision of the valley, functional classifications of buildings, a site classification of dwellings, a hierarchical political classification of civicceremonial plazas, and a size classification of sites. Particular emphasis is placed on the widely relevant analytical importance of the relationship between political and size classifications of settlements.

Territorial subdivisions

The survey area is readily divisible into smaller districts defined topographically by ranges of low hills or constrictions in valley-floor width – i.e., one section in each valley half, and within each section a set of sub-basins termed pockets (Zorrillo, Nuestra Señora, Chihuahua, Momón, Rosario, Santa Ines North, Santa Ines South). Another district, the Midvalley Range, has a different character, covering part of the range of hills that bisects the valley. Because its settlement pattern is so different (de Montmollin n.d.a: ch. 5), the Midvalley Range is left out of most comparisons. A consistency in the number of political hierarchy levels within equivalent topographically defined districts (four levels in each section, three levels in five of seven pockets – Figure 5) suggests a correspondence of topographic and political boundaries and reinforces the district's analytical validity.

Bridging arguments for ancient complex polities

Archaeological studies of ancient complex polities often rest on a base of weakly developed bridging arguments for linking theoretical concepts to data in the archaeological record (Chapter 1). Bridging arguments are indeed difficult to develop for such complex subject matter. Furthermore, devoting a lot of attention to them detracts from the time available for pondering what seem to be more fascinating great questions (Chapter 1) and substantive details associated with civilizations. However, there can be no doubt that building bridging arguments has a fascination of its own, requiring intricate problem solving and vigorous imagination on the part of the archaeologist (well exemplified in Binford's Palaeolithic studies- 1983a, 1983b).

In looking at bridging arguments and their uses it makes sense to distinguish “between theory treated methodologically as a means of investigating another theory and theory treated substantively as the theory to be investigated” (Bailey 1983: 177). From this perspective, constructing bridging arguments is a methodological means to an end. For example, alternate notions about politics (behavior) are related through bridging arguments to distinguishable settlement patterns (material culture). From a different perspective, the distinction between substantive and methodological theories begins to blur. What look like bridging arguments take on a more intrinsic theoretical interest. For example, an argument that links settlement-density patterns (seen in material culture) and sociopolitical patterns (behavior) becomes the central focus of theoretical interest (Fletcher 1981). This seems to be the trend in certain forms of ethnoarchaeology where material culture is the main archaeological subject matter (Hodder 1982) and the call is for theoretical work to develop generalizations about the role of material culture in creating political structure and organization.

In this study of settlement and politics in a Classic Maya polity I have set out to discuss and exemplify some elements of an anthropological approach to the archaeology of ancient complex polities. To begin, I want to sketch out briefly and rather informally the basic motivations for the work, saving the more detailed formal rationales for later. I see an anthropological approach to archaeology as something more than a simple mining of ethnographic lore in order to fill out archaeological interpretation. I also see it as something other than the uncritical adoption of a single paradigm from social or cultural anthropology (e.g., cultural ecology, structuralism, or post-structuralism). Rather, to my mind, an effective anthropological approach to the archaeology of complex polities requires the adoption of some of the more appealing intellectual traits that characterize the social and cultural anthropology of complex societies. Very broadly, these traits include sustained scepticism about received conceptual tools, respect for the diversity of behaviors and institutions in the record, a sensitivity to issues of social scale, and a sensitivity to the tension between approaches that use abstract formal models as opposed to approaches that use more particularistic substantivist models (the contrast between general comparative and particularizing approaches).

From my chosen anthropological-archaeological perspective, it seems that over the last few years many archaeologists have reached a kind of impasse in their studies of complex polities. Conclusions have been and continue to be drawn about such great issues as state formation or the rise of civilization. But in a more sceptical (and post-heroic) intellectual climate, it seems increasingly difficult to continue discussion exclusively along this track without falling into conceptual routinization.

Societal typologies

An effective study of political structure and organization in ancient complex polities cannot go very far without a good initial conceptualization of such polities. Societal typologies or bundled continua of variation are two principal approaches to meeting this requirement. Along with other Americanist archaeologists, students of Maya politics have usually opted for the first approach. There are numerous attempts to classify Maya sociopolitical systems as belonging to one or another societal type. Some have tried to decide whether the Maya had theocratic as opposed to secular-militaristic government (Wolf 1959; Webster 1976a). Contrasts are drawn between theocratic (lowland) Maya and secular-militaristic (highland) Central Mexican cultures or between the theocratic Classic Maya and the secularmilitaristic Postclassic Maya. Others have adopted a cultural-evolutionary typology (Service 1971), discussing whether the Maya had a chiefdom or a state level of development, and trying to determine when and why the shift from one to the other might have occurred (Sanders and Price 1968; Marcus 1976). Whether the Maya had an urban or non-urban form of civilization has also intrigued Mayanists (Haviland 1966a; Kurjack 1974) as has the question of whether the Maya had egalitarian, ranked, or stratified social structure (Rathje 1973; Haviland 1966b; Vogt 1983). An energy-capture view of cultural evolution has been used to categorize the Classic Maya as a Low Energy Society, contrastable with High Energy Societies in Central Mexico (Sidrys 1978). Other studies have placed the Maya in a Durkheimian typological framework, contrasting the Maya's mechanically solidary socioeconomic system with organically solidary socioeconomic systems and contrasting societas with civitas (Coe 1961, 1965).

Conceptualizing an ancient polity

I selected bundled continua of variation in place of societal typologies in order to conceptualize the Rosario polity as an example of an ancient complex polity. Why was this? By so doing, some difficulties associated with societal typologies could be avoided. The full range of these difficulties concerned: resistance of whole societies (or polities) to typological analysis; difficulties of choosing scale and locating boundaries for societal types; undue reification of society; undue reification of societal types; inappropriately categorical thinking (for the study of continuous variation); inability to account for change from one type to another; a priori technoenvironmental determinism (entailed by using a typological approach for societies but not for their physical environments); real rather than nominal definitions (Service 1985) for types, with assumed co-variation of several attribute levels and elimination of worthwhile research problems; and, finally, dubious extrapolation from documented attribute levels to undocumented attribute levels (Chapter 2).

The bundled continua approach derives from work done in political anthropology. The central premise is that it is “useful to place phenomena on a continuum, with the expectation that to do so will make it possible to locate cluster points” along several of the aligned continua (Easton 1959: 239). Reasons for the recurring clusters may then be sought. With such an approach, awkwardly multivariate societal types are broken down into more easily studied constituent variables. The existence of continuous as well as discontinuous variation is allowed for by avoiding a priori polar-categorical thinking. The object of study is given a nominal rather than a real definition (Service 1985), thus avoiding an essentialist approach which seeks to identify and discuss the true aspects of phenomena.

The mechanical-organic solidarity distinction finds distant inspiration in Durkheim's studies on the division of labor in society (Chapter 2; Durkheim 1933). Mechanical arrangements feature economically autarchic constituent units, limited exchange, and independence of parts. Contrarily, organic arrangements have a great deal of economic specialization among constituent units, extensive exchange, and interdependence of parts. It then follows (or does it?) that mechanical arrangements tend to be less solidary and less cohesive than organic ones, especially as scale increases. Aspects of the Rosario settlement record may be used to address the fourth set of research questions. What was the degree of mechanical versus organic economic solidarity that characterized Classic Maya society (4a)? If there was markedly mechanical economic solidarity, how closely was this associated with segmentary political structure (4b)?

My analysis of the mechanical–organic continuum in the Rosario polity must be quite a summary one, in keeping with a primary focus on politics (rather than economics), and to remain within the limitations imposed by very sketchy archaeological evidence concerning the presence or absence of economic specialization. However, at least some preliminary conclusions can be reached about the position of the Rosario polity and its districts along the mechanical to organic solidarity continuum. However imperfect, these conclusions are worthwhile in order to give at least some economic underpinning to the predominantly social and political factors investigated so far.

Certain patterns, which are potentially recoverable through archaeological survey methods, allow a distinction to be made between mechanical and organic solidarity. These patterns shed light on the question of relative community (site)-scale involvement in basic agricultural production activities.