Archaeologists' perception of the character of Anasazi society has changed dramatically in the last five years. Oddly, this is not so much in response to new data, although there are a great deal of those, as it is to new ways of looking at the data.
The 1960s and early 1970s in Anasazi archaeology were devoted to the development of local sequences. This was in large part a result of increasing knowledge about local variation. In addition, there was frustration over the inability to structure local developmental sequences into the Pecos Classification stages and the geographical divisions based on the major drainage systems, largely unchanged since Kidder's (1924, 1927) formulations.
Traditional temporal and spatial categories have been deemed inadequate for years (Martin and Plog 1973; F. Plog 1981), and pleas have been made for schemes that permit the use of continuous rather than categorical measurements. Unfortunately, in very few localities is either the chronological control or the understanding of behavioral nuances and environmental perturbations sufficiently detailed to structure a continuous dynamic model. The result of the construction of local sequences was a tendency to view the Anasazi area as made up of discrete, relatively autonomous social and economic entities that changed periodically at what are considered phase boundaries. Somewhat in reaction to this discrete temporal and spatial packaging have been recent attempts to demonstrate that the Anasazi engaged in large-scale social and economic interaction, which included economic interdependence and, in some cases, complex political organization and social ranking (Lightfoot and Feinman 1982; F. Plog 1983b; Upham, Lightfoot, and Feinman 1981; Upham 1982). While these scholars have addressed the excesses of past parochial views, we suggest they may be erring in the opposite direction. There was, as many chemical, physical, and stylistic analyses indicate, much greater interaction and information flow among localities than previously supposed; however, this was not necessarily accomplished through an elite class, a hierarchical social system, or even a formally organized network.
We contend that (l) the Anasazi were for the most part egalitarian, except for short periods in certain areas; (2) minute social, technological, and demographic adjustments were made partly in response to spatial and temporal changes in the natural environment; (3) these adjustments, sometimes over a distance of only a few kilometers, produced variations on the Anasazi pattern that are much more localized and distinctive than previously realized; (4) interaction networks expanded and contracted, and cooperation and competition between localities increased and decreased depending, in part, on demography and on spatial and temporal patterning of the environment; (5) cooperation between social entities and relative absence of environmental circumscription hindered the development of more complex socially hierarchical forms; and (6) the generally egalitarian nature of Anasazi societies was partially responsible for the persistence of the culture for some two millennia. If these propositions are correct, the Anasazi concept should be reevaluated.
The main thrust of this paper is that the basic behavioral mechanism of Anasazi adaptation is cooperation (Fig. 12). Cooperation is effected by the coordination of activities and the flow of information, sometimes restricted, sometimes faulty, which conditions decision making. This coordination can be accomplished through one of several levels of social organization: egalitarianism, ranking, or stratification, which may represent a continuum of variation rather than discrete categories. In addition, social interaction involving phenomena such as exchange, the development of cultural boundaries, and other kinds of information flow facilitates adaptation.
This scheme provides a framework for understanding the connectivity and interaction of horizontally organized social entities. Change in the system results from changing relationships among behavioral, demographic, and environmental variables (Dean 1988; Dean et al. 1985) . When change in any component of the system crosses a systemic boundary, it triggers major adaptive change in other components. These transformations make certain information obsolete and may rebuke changes in the intensity and form of cooperation and competition, in the nature of the social order, and in the rate, direction, and form of social interaction.
Cooperation and Competition
Our theoretical stance is that in the heterogeneous but relatively uncircumscribed environment of the Colorado Plateaus there is a distinct adaptive advantage to behavior that (l) initiates and reciprocates cooperation; (2) provokes easily and quickly retaliates against aggressive behavior; and (3) forgives aggressive behavior after retaliation. This proposition runs counter to the common anthropological belief that conflict, warfare, and inequality are in large part the cause of, or co-occur with, social ranking and stratified societies (R. Adams 1975; Carneiro 1970, 1978; Haas 1982). It does appear that stratified, aggressive societies tend to coincide with the development of chiefdoms and the pristine state; therefore, ranking, stratification, and aggressive behavior are seen as attributes of these evolutionary stages. While these stages are conceived as advances in evolutionary form, however, individual societies within them are usually unstable and fall to competing entities. As Wright (1977:385) has observed, chiefdoms and simpler states exist in networks ever-changing in scale, intensity, and direction regulated by competition and alliance. It cannot be assumed, however, that in the absence of systemic circumscription, nonegalitarian societies are in the long run more evolutionarily fit, or that chiefdoms or states based on political centralization, warfare, and competition for resources necessarily have adaptive characteristics superior to those of more egalitarian or tribal entities. In fact, just the opposite may be the case. Adaptation based on cooperation as opposed to competition has been studied using game theory and computer simulation based on the iterated Prisoner's Dilemma game, which models ecological adaptation (Axelrod and Hamilton 1981; Axelrod 1983). A strategy that combined cooperation with other entities, retribution when acted against, followed by a rapid return to cooperative behavior proved most successful against competing strategies when run for a great many generations of encounters. Interestingly, this cooperation-retribution-cooperation (CRC) strategy did not defeat a single rival in the many generations of encounters that have been run; the best result it produced was a tie, and often it lost to competition by a slight margin. The CRC strategy is successful for the entire system because it elicits behavior that permits all strategies to do well. Aggressive strategies that attempt to exploit more cooperative programs often did well for many generations of encounters but, in closed system environments, proved to be self-defeating in the long run.
Cooperation (Mechanism)
Coordination
(Information Flow and
Decision Making) (Process)
Environmental
and Demographic
Matrix
Social Social
Order Interaction (Structure)
(Egalitarian/ (Exchange/
Asymmetrical Boundaries)
Fig. 12. A model for the mechanism, process, and structure by which the Anasazi adapted to environmental and demographic variation.
Lessons from the iterated Prisoner's Dilemma game seem to be especially useful in understanding the long range adaptive stance of the Anasazi, the generally conservative nature of Anasazi culture, and the occasional and short-lived emergence of possibly stratified societies exemplified by complex entities such as the "Chaco Phenomenon."
In order for social entities to cooperate, coordination is necessary. Social systems are in part arenas of interaction that coordinate and dissemination information, usually on a selective basis. Articulation of the constituent parts of the social system depends on information flow about human-human and human-environment interactions that can vary in direction, frequency, and intensity (Johnson 1978; Root 1983). In egalitarian societies, comparatively unrestricted information flow promotes relatively equal access to critical resources with the result that no major economic or political advantage occurs through control of those resources (Fried 1967). Information flow integrates and coordinates components of the social system, and institutionalized social relationships are the channels through which the information flows (Van der Leeuw 1981; Root 1983).
In a heterogeneous environment, such as that of the Western Anasazi area, the unregulated flow of information through institutionalized social relationships would mitigate potential differential access to resources caused by spatial and temporal variability in those resources.
Anthropologists have sometimes assumed that decisions based on received information are made under conditions of omniscient rationality by individuals striving for economic optimization. There has been a widespread failure to recognize that decisions are made in social contexts that may or may not involve economic optimization. As Moore (1983:183) has noted, "In reality, decisions are based on a mix of information, ignorance, error,and lies."
In sum, archaeologists should not expect a one-to-one correspondence between environmental change and human optimizing behavior, even if there is equal access to information and critical resources. However, archaeologists have to assume that coordination and information flow existed on some level, or the society would cease to function.
As stated in the introductory section, cooperation is hypothesized to be the basic behavioral mechanism for adaptation in the Western Anasazi area. It ensured the relatively unrestricted flow of information, and facilitated decisions necessary to mediate the unequal distribution of natural resources in the heterogeneous environment. This is not to say that at some times and places competition over important resources did not restrict the flow of critical information. In addition, interaction, in terms of exchange and other forms of social interchange, shifted in direction, duration, and intensity in relation to demographic, environmental, and behavioral variables.
Recent provocative publications have suggested that at certain times and places asymmetrical, nonegalitarian societies developed in the Southwest (Lightfoot and Feinman 1982; Upham et al. 1981, Upham 1982).
Further, it has been suggested there were periods and areas when populations coalesced into formal political and economic "alliances" in contrast to a simpler "resilient" pattern, characterized by small, relatively autonomous farmsteads or semi-nomadic groups (Upham 1982; F. Plog 1983b; Cordell and Plog 1979; Stuart and Gauthier 1981). These studies have provided an important awareness of spatial and temporal variability in social organization and interaction. They have forced archaeologists to consider information flow, social order, and interaction in a non-normative way. There are, however, several alterative explanations for the patterning observed in the archaeological record. Some studies (Lightfoot and Feinman 1982; Upham et al. 1981; Upham 1982) do not view egalitarian and asymmetrical societies as social entities along a continuum, but rather, like classical evolutionary categorizations, which viewed culture change in broad stages. This is in spite of the faa that some of the authors themselves caution against the use of nonscalar concepts (Lightfoot and Feinman 1982). The authors often appear to view any evidence of social status difference as proof of a nonegalitarian society. No human society is completely egalitarian if that term is taken to mean the absence of status differences. Even simply organized societies like those of the Shoshoneans, which most anthropologists agree were egalitarian, had "talkers" and "rabbit chiefs" who acted as headmen at certain times in order to accumulate, process, and transmit information about subsistence resources and to organize collection and hunting activities (Steward 1938).
Attempts have been made to rationalize models of a hierarchical social order in prehistory by emphasizing the inequities that exist in present Western Pueblo society (Upham 1982; Upham et al. 1981). Indeed, inequities do exist, even in those aspects of society that control access to vital resources.
Such inequities, however, are based less on the accumulation of material wealth than on ritual knowledge (Smith 1983). Commonly, ritual knowledge is widely apportioned throughout the society and is, therefore, diffused rather than accumulated. Smith ( 1983:42) goes on to state that "the distinction, then, between egalitarian, ranked, and stratified societies may, at best have only narrow analytical utility, particularly if it rests on such ethnocentrically materialistic criteria as access to real goods."
Nevertheless, a scale of asymmetry is useful for understanding degrees of social complexity and can be used effectively to describe and understand culture change. Furthermore, it matters little for our purpose whether or not access to ritual power results in the accumulation of material wealth or whether the two are concomitant or independent. For our purposes, an appropriate definition of egalitarian society is one whose members have essentially equal access to critical resources and to a relatively unrestricted flow of information about those resources. This definition allows for the development of leadership through achievement, personal abilities, and even hereditary status, so long as that status does not convey the right to restrict the access of others to critical resources. To confine the term egalitarian to less than this definition excludes most, if not all, human societies. The ideal is, of course, to consider degrees of egalitarianism and asymmetryÑdifficult to do in an archaeological situation.
Within the Western Anasazi area there is no archaeological evidence for the existence of an asymmetrical social order, at least until the late fourteenth century; there is no evidence of surplus labor, large capital investment, or great specialization of production. Furthermore, there is no indication that the need to process information about resources was so enormous that it required asymmetry in the social order, structuring the flow of information to reduce production or procurement costs.
There is no need to invoke the nuances of environmental or demographic variability to explain the egalitarian nature of Western Anasazi society prior to the fourteenth century. In spite of the vagaries of environmental and demographic pressure or release, the most effective strategy for dealing with most situations would have been one of cooperation and access to information and resources necessary for survival. During the late fourteenth century, however, population density and capital investment at large communities may have increased tendencies toward a conscriptive society.
Prior to the fourteenth century evidence for "central places" is minimal. Klesert (1982b) proposes that Standing Fall House, a cliff structure at the extreme northeastern edge of Black Mesa dating just prior to abandonment of the area about A.D. 1150, served as a central storage and redistribution center, although the evidence for this is minimal.
The Plaza site in the Hopi Buttes was almost certainly a "central place" (Gumerman l975, 1986).The Plaza site,in contrast to the surrounding small (three to eight room) habitation sites, is distinguished by a high ratio of public to nonpublic space. There is no evidence, however, to suggest that the Plaza site restricted access to critical information or resources as there are no qualitative or quantitative differences in artifact inventory from the contemporaneous small habitation sites in the immediate vicinity. On the contrary, the Plaza site's main function was probably to enhance accumulation and dissemination of information and facilitate the distribution of critical resources, thus mediating the effects of a heterogeneous and high risk environment.
A dispersed population, dwelling in numerous small sites, was economically most efficient for food production and procurement in the Hopi Buttes area. Nevertheless, with an expanding population in a heterogeneous environment, a mechanism to ensure an unrestricted information flow had to be developed to monitor the environmental situation for effective cooperation. The Plaza site provided that function.
Tsegi phase communities of the Klethla, Laguna Creek, and Long House Valleys also appear to have been internally differentiated, with the central pueblos functioning as "central public places." Although data pertaining to this issue are scant, central pueblos probably served as loci of information exchange among inhabitants of the villages clustered around them. The mutual intervisibility of central pueblos in Long House Valley suggests that intercommunity information exchange, perhaps regulated in some fashion, also existed.
Northeastern Black Mesa had no need for a "central place" to monitor information because the physiography (unlike the Hopi Buttes) is dominated by a dendritic drainage system where deep alluvium and water are arranged in linear fashion. Relatively small sites could maintain the majority of social, economic, and religious functions. Because most sites had equal access to all critical resources, it was not necessary to establish a central place for the monitoring of environmental information over a large area. In contrast, the dispersion of population throughout a heterogeneous environment stimulated the sharing of information and coordination of effort over the entire area. The discontinuous distribution of usable farmland under the conditions of alluvial degradation that prevailed during the Tsegi phase may account in part for the development of "central places" in the eastern Kayenta area.
There is evidence in the study area for greater social complexity after the Hopi Buttes and Kayenta areas were abandoned. Upham (1982) disputes the egalitarian nature of the Western Pueblo during Pueblo IV (ca. A.D. 1300-1450), suggesting that "hereditary oligarchies" controlled ritual and other knowledge resulting in a much more asymmetrical social order in the Upper and Middle Little Colorado River than previously supposed; certainly more than existed at the time of European contact. Upham's evidence consists largely of the "central place" spacing of contemporaneous sites, which suggests a network of what could be called power systems, and the differential distribution of certain artifacts, which may reflect the differential access to commodities attendant on status differences.
Several arguments can be marshalled against the supposition that a ranked, hereditary elite exercised control over the system of communities. First, the existence of interaction or regulated exchange among large communities does not necessarily specify a nonegalitarian society. On the contrary, the evidence for interaction among smaller, earlier sites indicates that such relationships should be expected for the later, larger sitesÑeven if the distance between individual sites is much greater. If there are fewer sites, interaction will be maintained or even augmented in spite of increased distances between the sites. Second, and more important, there is no necessary correlation between the differential distribution of artifacts and the denial of access to those artifacts on status grounds.
As Plog and Upham themselves state, archaeologists probably cannot distinguish situations where access to resources is based on preference or opportunity rather than differential power (Plog and Upham 1983:208-209). It is our contention that in spite of deteriorating environmental conditions, demographic pressure, and capital investment, the advantages of cooperation in a process of unrestricted information flow and equal access to resources outweighed the advantages of hereditary asymmetrical social organization. That may be why asymmetrical social order along the Little Colorado River was so ephemeral, if it existed at all.
There can be little question that there was a Feat diversity in the manner in which Anasazi societies were organized, and some were certainly more complex than Western Pueblo societies or those that existed at the time of Spanish contact. Some of this complexity, evidenced for example by Chacoan road systems and great public architecture, must have required a ranked or even stratified society, based on competition and on control of information flow, as well as preferential access to resources. We do not intend to discuss the origins of these manifestations, except to note that characteristically they are short lived. The short-lived phenomena of an aggressively exploitative strategy in a social environment of cooperation is precisely what is expected in the iterated Prisoner's Dilemma game. One of the fascinating results of the game is that a group of cooperating entities can successfully resist aggressively exploitative strategies if the cooperating entities continue to interact with one another. This demonstrates how an egalitarian cooperation strategy can successfully adapt in an environment of social exploitation, whether or not that exploitation derived from external social systems, such as Mesoamerica, or was internally derived.
In sum, the Western Anasazi area provides no concrete evidence for the development of an asymmetrical social system, with the possible exception of late in the prehistoric period when, for a short time, environmental circumscription, high population, and the capital expenditure of labor may have encouraged social ranking along the Little Colorado River. Outside the Western Anasazi area, evidence for short-lived ranked or stratified societies exists, but cooperation and egalitarianism were the primary adaptive modes. Considerably more archaeological evidence exists for the degree, direction, and intensity of social interaction than for any determination of the social order (Dean et al. 1985). Social interaction in an archaeological sense includes such traditional spheres as exchange and "cultural boundaries" as well as recently articulated concepts of provinces and alliances.
The direction, degree, and intensity of interaction changed dramatically several times in the Western Anasazi area. These changes are evidenced by either shared or distinctive styles in ceramics and architecture, and in the results of chemical and physical material analyses (Deutchman 1980; Green 1984; S. Plog 1980; Upham 1982). The stage names proposed in the first chapter of this volume are largely indicative of variations in interactive behavior.
The interpretation of changing interaction as periods of alliances and breakdowns of alliances by F. Plog (1983b, 1984) and Upham and Plog ( 1983) appears to be too narrow an interpretation of the data. Alliances are usually considered to be formal, political, highly organized arrangements among elite segments of large social entities. Alliance behavior, however, is only one extreme in a continuum of interactional orders that can produce shared styles or materials. Interactional forms range from episodic contacts between individuals to highly structured, continuing, formal relationships between large social and political entities. Interaction between individuals or small groups of individuals is more consistent with our views of the Western Anasazi. However, it is also evident that the extent and formality of relationships varied through time. For example, at times there is a sense of great parochiality when architecture and, to some extent, ceramic style were consistent over only a few kilometers on Black Mesa and in the Hopi Buttes. At other times, especially early and late in the sequence, there is evidence of intensive and continuing pan-Western Anasazi interaction.
The Western Anasazi archaeological record suggests regional interactive behavior far less complex than that implied by the concept of formal political alliances. The alteration of periods of widespread interaction with intervals of more localized interchange may be explained as reflections of cooperational-retribution-cooperation ( CRC ) behavior rather than of participation in formal political alliances. The same sort of environmental variability that is thought to have triggered alliance-forming behavior in other areas may have stimulated CRC behavior among the Western Anasazi. Thus, high spatial variability in climate may have enhanced inter-areal information exchange (i.e., increased cooperation), while periods of uniform climatic conditions reduced the need for information exchange and cooperation, exacerbated intersocietal differences, and stimulated competition and retributive behavior that persisted until environmental conditions once again favored cooperation. The CRC response to ecological diversity suggested by the iterated Prisoner's Dilemma game provides an alterative to the political alliance explanation of archaeological evidence for intermittent increases in interactive behavior. While the CRC model seems to better account for the nature of this evidence in the Western Anasazi prior to A.D. 1300, the alliance model may be more applicable to other areas or other time periods. Evaluation of both models in particular instances of interactive behavior undoubtedly will increase resolution and the explanatory power of each, and thereby enhance understanding of southwestern prehistory.
The recent dialogues over the nature of the social order and the degree and form of interaction have provided a more realistic understanding of Western Anasazi social behavior. In spite of differing points of view, all scholars paint a much more variegated portrait of Anasazi society than had previously been observed. Of especially great value is that the discussion centers not on whether the Anasazi were hierarchically organized or the exact forum of social interaction, but on what these social conditions tell us about changing Anasazi behavior.
The Western Anasazi archaeological record suggests regional interactive behavior far less complex than that implied by the concept of formal political alliances. The alteration of periods of widespread interaction with intervals of more localized interchange may be explained as reflections of cooperational-retribution-cooperation ( CRC ) behavior rather than of participation in formal political alliances. The same sort of environmental variability that is thought to have triggered alliance-forming behavior in other areas may have stimulated CRC behavior among the Western Anasazi. Thus, high spatial variability in climate may have enhanced inter-areal information exchange (i.e., increased cooperation), while periods of uniform climatic conditions reduced the need for information exchange and cooperation, exacerbated intersocietal differences, and stimulated competition and retributive behavior that persisted until environmental conditions once again favored cooperation. The CRC response to ecological diversity suggested by the iterated Prisoner's Dilemma game provides an alterative to the political alliance explanation of archaeological evidence for intermittent increases in interactive behavior. While the CRC model seems to better account for the nature of this evidence in the Western Anasazi prior to A.D. 1300, the alliance model may be more applicable to other areas or other time periods. Evaluation of both models in particular instances of interactive behavior undoubtedly will increase resolution and the explanatory power of each, and thereby enhance understanding of southwestern prehistory.