Theories of ethnicity and the dynamics of ethnic change in multiethnic societies

Abstract

I modify Fredrik Barth’s approach, which sees ethnic group building as a signaling system, to place it within a framework that draws from collective action and costly signaling theories. From these perspectives, ethnic signaling, although representing a costly penalty to group members, is one effective form of communication that facilitates collective management of resources. I then identify three contexts in which the benefits of ethnic group building are likely to outweigh its signaling costs: in politically chaotic refuge and periphery zones; in the context of long-distance specialist trading groups; and within the territorial scope of failed states. I point to selected data from the Mughal and Aztec polities to illustrate how a combination of effective public goods management, in highly collective states, and the growth of highly integrated commercial economies will render ethnic group building superfluous.

Early in the 20th century, anthropologists turned to a focus on culture as a challenge to the biologically reductionist race thinking of 19th century evolutionists. This strong cultural program, promoted by Franz Boas and his followers, was influenced by a German tradition tracing its origins to Romanticists, such as Johann Gottfried Herder, and their “Aufklärer” followers of the 19th century (1). This group opposed ideas of British Enlightenment authors, such as John Locke, who argued that a society, as a commonwealth, ideally is the product of socially purposed institution building. In the Romantic reaction, society is understood less in terms of rational social action and institution building and more in terms of a people’s shared origins and history, and an emotional attachment to their culture, language, and local territory (2). Boasians drew inspiration from Romanticism to make the argument that even though there might some diffusion of culture traits across societal boundaries, still, each cultural unit or “tribe” (in the case of smaller scale societies) was understood to develop a distinct social, linguistic, and cultural configuration shared by its members. What I call a strong cultural program proved to be a source of disciplinary unity for decades but eventually proved to be problematic in the way it ignored the role of human agency in social group building.

The Strong Cultural Program in Archaeology and in Theories of State-Building

We can see the impact of the strong cultural program on 20th century archaeological practice, for example, when archaeological cultures or regions are understood to reflect the distribution of a people and thus are ethnically labeled, for example, as “Sumerian” and “Akkadian” in early Mesopotamia, or “Zapotec” or “Mixtec” in the pre-Hispanic Valley of Oaxaca, despite evidence for immigration and ethnic pluralism [early Mesopotamian city-states were demonstrably pluralistic according to Yoffee (ref. 3, p. 49); Blanton et al. (ref. 4, p. 41) and Flannery and Marcus (5) point to the possibility for immigration and pluralism in the Valley of Oaxaca]. In this regard, it is also worth mentioning how episodes of social and cultural change have been understood as the result of large-scale invasions of new dominant cultural groups. For example, we see this device in the Mesopotamian historical and archaeological literatures when change episodes are attributed to invasions by various groups, including Amorites, Canaanites, and Hyksos, an approach critiqued by Kamp and Yoffee (ref. 6, p. 97). And Cohen (ref. 7, p. 381) finds, in other branches of anthropology, a tendency toward overly simplistic ethnic labeling that ignores ethnic plurality.

The strong cultural program is on full view in the theory of state dynamics developed by Clifford Geertz (8). Here, he addresses the role played by local tribal cultures in the dynamics of postcolonial states in Africa and in South and Southeast Asia. As he argues, a tension between cultural heterogeneity, counted as regionalism, religion, language, or tribe, on the one hand, and the need for a civil order, on the other, is “one of the central driving forces in the national evolution of the new states; as it is, at the same time, one of the greatest obstacles to such evolution” (ref. 8, p.108). The key problem, Geertz argues, is found in the fact that, within the boundaries of the new states, there are groups in which attachment to one’s culture can be understood to constitute a “natural” or “primordial” state of human experience. Local attachments persist despite nation-building projects, he argues, because such attachments are more natural than national-scale attachments owing to the “great extent to which a peoples’ sense of self remains bound up in the gross actualities of blood, race, language, locality, religion, or tradition… . To subordinate these specific and familiar identifications in favor of a generalized commitment to an overarching and somewhat alien civil order is to risk a loss of definition as an autonomous person, either through absorption into a culturally undifferentiated mass or, what is even worse, through domination by some other rival ethnic, racial, or linguistic community that is able to imbue that order with the temper of its own personality” (ref. 8, pp. 108–109).

There are instances in which belonging is couched in the language of blood and kin (e.g., ref. 9, p. 27). However, does a primordial sensibility preclude rational social action? In the case of Geertz’s sense of primordialism, the answer seems to be yes, but his view is problematic in identifying distinct categories of persons I roughly label as “rational” and “tribal.” On the rational side, the goal of social actors is to construct civil order based on notions of democratic modernity. The tribal peoples, by contrast, mired in primordial emotional attachments to “the gross actualities of blood, race, language, locality religion, or tradition” seem to be lacking in rational social agency. I also suggest that Geertz’s scheme errs in viewing postcolonial state building in terms of a singular notion of political modernity, when, in reality, state building, whether modern or premodern, exhibits a wide range of variation in form, function, political goals, and the ability to enact goals (10). I return to that point below in Ethnicity Within the Fabric of a Complex Society.

Fredrik Barth and the Critique of the Strong Cultural Program

The key turning point in anthropological thinking about culture and ethnicity came from Fredrik Barth’s brief but very useful introduction to “Ethnic Groups and Boundaries” (11). Barth proposed that we rethink culture from the vantage point of how social action drives the formation of ethnic groups that are aimed to “organize interaction” for a social purpose (ref. 11, p. 10). Typically, in ethnic groups, a sense of belonging is created through the symbolism of blood relations and shared history. However, to Barth, symbolic meaning is less important than the fact that ethnically specific behaviors constitute a system of signals with the dual purpose to establish intergroup boundaries and, intragroup, to confirm group members’ commitment to the “basic value orientations” of the collectivity (ref. 11, p. 14).

Barth and Collective Action Theory

Barth’s ideas have been influential (12), but the bulk of research using his instrumentalist approach has emphasized signaling that denotes ethnic boundaries and thus manifests difference (e.g., ref. 13, p. 128), but here I take a different direction, with the goal to bring Barth’s ideas into the light of collective action theory. To do that, I emphasize Barth’s suggestion that intragroup signaling allows a group’s members to demonstrate they are, in Barth’s terms “playing the same game” (ref. 11, p. 15).

In collective action theory, as in Barth’s theory, the human subject is regarded as a rational social actor. As a result, with only minor modification, Barth’s formulation can be productively incorporated into a collective action perspective by recasting the rational self as a “conditional cooperator.” This notion of the rational self identifies cooperation as a rational course of action, when the goal is to realize both collective and self-benefit, but behavior that illustrates a selfish disregard for collective benefit is also rational. As Mancur Olson (14) pointed out, humans have difficulty cooperating because rational individuals may not act in the common interest (15) and, as a result, to construct a socially purposed cooperative group entails problem-solving strategies of institution building with the goal to foster cooperative behavior.

From this perspective, ethnic construction can be considered as one possible institutional strategy suited to the establishment of a socially purposed cooperative group. According to collective action theory, organizing for common defense and for the cooperative management of resources will be the most pressing problems requiring institution building. Resource management might include the goal to maintain control over and sustainably manage finite “common pool” resources, as in the case of the nomadic pastoralist Pathan tribes Barth studied (16), or to manage a “public goods” system where the cooperative goal is to gain mutual benefit from jointly produced resources. In both cases, because mutual benefits are gained from the devoted efforts of other group members in tandem with one’s own efforts, a group’s members will make mutual claims of accountability with respect to each other. However, how is it possible to understand that the intentions and actions of others will be consistent with collective benefit, given that the conditionally cooperative human has the capacity to find individual gain as a rational but egotistical agent?

A group’s system of visual signaling provides one clue as to the manner in which trust is generated among members of a collectively organized group. For example, as the collective action theorist Michael Hechter put it, to produce cooperative groups, “…individuals must be highly visible to one another in order to reduce the severity of the free-rider and assurance problems” (ref. 17, p. 21; cf. ref. 18). Although intervisibility is one solution to the trust problem, especially in very small groups, ethnicity provides another form of trust-building signaling. To communicate reliable ethnic signals mandates that one will make an overt commitment to a whole way of life, including the consumption of ethnically specified material culture [what Wobst (ref. 19, p. 12) calls “stylistic messaging”] and the display of proper public etiquette and language competency, and often will include other commitments, such as participation in public ritual or other public behaviors; as Cohen (ref. 20, p. xiii) observed, to join an ethnic group, one must “pay the price of membership” (cf. ref. 11, p. 23). Thus, ethnic signaling is productively understood as one form of “costly signaling” (e.g., ref. 21) or “reputational signaling,” notions that provide us with one way to understand how signals may be evaluated in terms of their probable validity and thus enhance trust. This idea, first proposed by Zahavi (22), points out that signals that are costly to the signaler (i.e., that are a handicap) will be perceived as having more reliability and thus are more likely to enhance trust than lower cost signals. Ethnic signaling also is an ideal form of intragroup communication and trust building because ethnically based signals, especially when considering such factors as language competency, are difficult to fake. A person who obviously has made the kind of deep commitment to a way of living is likely to be understood as a person who also embraces those moral values of a group that undergird cooperation.

An approach that emphasizes signaling and, especially, the cost of signaling has two advantages for understanding ethnic group building. First, there is a material dimension to signaling (Wobst’s stylistic messaging), which is suited to archaeological methodology (ref. 23, pp. 173–195). Second, to evaluate cost raises an important question for consideration: Under what social conditions is it likely that the benefits of ethnic signaling will outweigh its costs? In the following, I identify several social contexts in which ethnic signaling likely will be a solution to the problem of building and sustaining collective action groups: in zones of weak periphery incorporation along the boundary zones or frontiers of polities or world systems; in the context of intercultural trade; and in contexts internal to poorly functioning or failed states.

Periphery Incorporation and Ethnogenesis

Hall (24) provides a useful scheme of varying degrees and forms of periphery incorporation, referring to situations in which autonomous boundary or frontier zones are impacted by forces of economic and political change emanating from multicultural economic systems (world systems) or expanding polities. Across a continuum from weak to strong incorporation, Hall finds that ethnogenesis is expressed most strongly in areas of weaker incorporation, his “refuge” and “contact” zones, where there is little direct political control from a polity or a world-system core zone. In refuge and contact zones, collective action at the local level emerges as a strategy for mutual defense and control of resources in socially chaotic environments plagued by demographic collapse, forced migration, social disruption, pressure from slavers, and competition for resources. Examples of ethnogenesis in refuge zones are discussed comparatively by Kowalewski (ref. 25; cf. ref. 26). He found that refugee groups from diverse cultural backgrounds overcome heterogeneity to build coalescent social formations through the creation of new modes of community integration. Ethnic group building in these cases emphasized, among other features, “universalizing ideologies and cults” that transcend cultural variability and the implementation of forms of collective leadership that downplay the role of centralized hierarchical authority (ref. 25, p. 117).

Populations in contact periphery zones of world systems may inhabit a chaotic environment while, at the same time, they face cooperation problems exacerbated by economic change as they are incorporated into an expanding economy as suppliers of raw materials or labor destined for core zone consumption. In these situations, ethnogenesis may be a strategy to manage an intensifying competition for valuable resources: for example, as is clear in the case of the North American plains region, during the 18th and 19th centuries, when fur became an important trade good (summarized in ref. 27, pp. 176–182). This growing periphery economy spawned a phase of ethnogenesis as local groups faced the dilemma that cooperation was sorely needed for mutual defense and to maintain control over important hunting territories at the same time that social cohesion was threatened when wealth and glory came to those who were individually successful in hunting, trading, and warfare. The Cheyenne, described by Hoebel (28), exemplify how the development of a new sociocultural program and ethnic identity were designed to enhance intragroup cohesion while overcoming the growing force of individualism. An important dimension of their strategy is evident in how the tribal leadership, the “Council of Forty-Four,” confirmed their devotion to Cheyenne value orientations through a costly signaling strategy. Upon assuming office, council members were obligated to cease participation in warfare, normally an important source of male status, and they were obligated to display a pattern of generosity that was likely to diminish a chief’s wealth over the mandated 10-y term of office (ref. 28, pp. 43 and 51).

“Alien Traders,” Ethnic Signaling, Trust, and Economic Monopoly

The alien-trader phenomenon is one in which an existing system of ethnic signaling is manipulated as a way to enhance the functionality of special-purposed long-distance trader groups. Ethnic signaling is used in the context of this kind of intercultural long-distance trading for two reasons. First, ethnic boundary marking defines sets of specific actors who are able to maintain monopoly control over the profit-making potential inherent in long-distance trading. And, given that intercultural commerce presents many cooperation problems, intragroup signaling is a source of trust building between traders. Cooperation problems in long-distance trading are difficult to overcome owing to the possibility for opportunistic self-interest–seeking behavior when marketplace transactions consist of fleeting encounters with strangers, and when transactors from different cultural or social backgrounds fail to mutually agree on the relative value of goods and services. In addition, market transactors may not share the same moral concepts and so may underestimate or overestimate the likelihood of market cooperation by others. Long-distance trading over great distances and across cultural boundaries also is challenging because it requires special expertise in social instruments for arranging credit and contracting at a distance.

In what Curtin (29) terms “trade diasporas,” the trust basis for market transactions in these challenging situations often is to be found in some combination of shared ethnic affiliation (“ethnically homogeneous middleman groups”) (30) and network capital (a personal social network), such as among Chinese diaspora groups in Southeast Asia, but many other examples can be cited, including Jews and Italians in Medieval Europe, Parsees in India, and Sikhs, Orma, Hausa, and Julas in Africa, as well as the Putun and Itzá merchant groups of late pre-Hispanic Mesoamerica.

I propose that trade diasporas are one reflection of periphery incorporation and a source of core-periphery inequality because the commercial success of trading groups reflects the differential development of institutional capital for market cooperation in core and periphery zones (31). In this scenario, trading groups are able to make use of social instruments for trading at a distance that have a long history of institutional development in their home region. Social instruments, in conjunction with their ethnic signaling and network capital, allow them to realize their goal to monopolize long-distance trading with periphery groups where there has been less social development along these lines. For example, as Granovetter (ref. 32, p. 32) points out, expatriate Chinese trading groups in Southeast Asia are successful because they form small close-knit communities based on shared clan membership or places of geographical origin. Within the close-knit networks, “credit is extended, capital pooled, and authority delegated without fear of default or deceit.” Jean Ensminger’s (33) study of sub-Saharan African trading, during the 11th to 19th centuries CE sub-Saharan Africa, points to religious conversion as a form of costly signaling. Islam brought a shared language of trade and “a monetary system, an accounting system, and a legal code to adjudicate financial contracts and disputes…making outsiders, insiders” (p. 7) and thus restricting access to the profits of the long-distance trade to a successful and privileged few. At the same time, to communicate reliable signals of conversion entailed personal costs, including giving up alcohol, devotion to fasting, costly pilgrimages, and the building of mosques.

Ethnicity Within the Fabric of a Complex Society

Although ethnic group construction often provides positive benefits for group builders in chaotic zones or as a strategy to facilitate long-distance trading of minority “alien” traders who connect cores with peripheries, it has a more complex and varied role within the fabric of complex societies. For this reason, Barth’s instrumentalist theory of in-group and between-group ethnic signaling applies only in certain situations. For example, McKay (ref. 34, pp. 401–402) pointed out that some expressions of ethnicity may be primarily ideational or affective (his “pseudo ethnics” and “symbolic ethnics”): for example, when persons gain personal satisfaction from carrying forward traditional elements of a threatened culture or when ethnicity is used in a person’s search for self-realization of identity. In these contexts, however, there may be little interest in group building to achieve a social purpose in the Barthian sense.

I suggest that, to explain variation in forms of ethnic expression, we consider the costs and benefits of ethnic signaling in the context of state-building processes, but also in the context of economic processes, the latter to understand the costs and benefits of ethnic construction within the framework of a market economy.

Comaroff and Comaroff (35) have pointed to how social inequality within contemporary states may prompt ethnogenesis among disenfranchised subaltern populations, who then use their organizational capacity to oppose state power or to make claims on the state (see also ref. 36, p. 343). Brumfiel (ref. 37, p. 94) identifies a similar process in premodern states, when ethnic group formation was a strategy to maintain local autonomy in the face of pressure from an exploitative elite. I propose a more general approach that sees ethnic group building as a way to realize local organizational capacity in the context of state failure. Failure may accompany phases when state authority is weakened or when a state collapses, or in the context of poorly functioning and exploitative states. All of these situations create conditions in which ethnicity is used to mobilize local-scale organization, either in defensive reaction to chaotic conditions, or as a strategy to gain access to a state’s public goods (e.g., ref. 38, p. 24). By poorly functioning or exploitative, I refer to situations in which a state is either lacking the will to provide, or is lacking the institutional capacity to provide, collective benefit: for example, the more autocratic or segmentary states identified by Blanton and Fargher (10) from their comparative study of premodern states.

In the comparative study, we demonstrated that a major cause of poor state functioning is revenue regimes in which governing principals maintained direct control over spot resources, such as export economies, or when they controlled other kinds of revenue sources, such as private estates or state-owned land. These resources could be mobilized to maintain the controlling position of a governing faction while making it possible to resist pressure to increase accountability and improve governing capacity. However, there were costs incurred by states that brought few collective benefits across social sectors. By comparison with the better organized states in our sample, these segmentary polities display a statistically significantly greater tendency to exhibit episodes of internal conflict, wars between ethnic groups, and various forms of oppositional movements (ref. 39, p. 48), perhaps analogous to the problems faced by the postcolonial states Geertz referred to.

I found little evidence for comparable problems in the context of those states possessing the institutional capacity to realize collective benefits (39). These polities implemented equitable tax-collection policies, provided public goods such as public security and effective judicial systems across the realm, accommodated taxpayer voice, and maintained effective institutional controls over the agency of governing principals and administrative cadre. In Blanton and Fargher (ref. 10, pp. 280–289), we demonstrate that collective action was achieved, in part, by extending administrative capacity deep into the social fabric to reorganize basal social units. However, this organizational process did not have a notable influence on multiethnicity; the great majority of polities in the sample (24 of 30), irrespective of degree of collective action, were coded as exhibiting multiethnicity. The question is, however, to what degree was ethnicity mobilized for group building within the territorial domain of a state? I predict that collective action would render ethnic group building within the state’s arena of control superfluous, for three reasons. (i) To the degree that a state is able to extend public goods such as military defense, effective judiciary, and the maintenance of public order across its realm, the functional purposes of ethnic group formations would be seen to duplicate state-provided services. As an example, I refer to how the provisioning of public goods influenced local-level ethnic groups in Mughal cities. Here, a tradition of highly closed and self-governed neighborhood units (mahallah), a residue of pre-Mughal periods, represented local-level adaptations to urban challenges, including the maintenance of public order, and often they were organized on the basis of shared caste affiliation or religion. The Mughal system emphasized urban administrative organization and public goods, so that, as Chaudhuri (ref. 40, p. 84) reports, increasingly during the Mughal period neighborhood-scale social formations were weakened. As a result, people increasingly purchased house lots where they could get them, and Muslims began to mix with Hindus and rich with poor [more examples of neighborhood decline under conditions of collective action are provided in Blanton and Fargher (41)]. (ii) In the more collective states, the costly signaling aspect of ethnic group formation may be seen to be burdensome because, in the fiscal system of more collectively organized states, revenues typically are derived from the broad population of taxpayers (ref. 10, pp. 253–256). In fact, this kind of “internal” revenue system, as we called it, is a basic element of collective action process because governing principals who depend on taxpayers will be more inclined to provide public goods and other services as a means to enhance confidence in government and, in turn, taxpayer compliance. Ethnic group members will thus face the growing costs of participating in the state’s economy, as they are drawn into it as taxpayers, at the same time they face the costly entailments of ethnic group signaling. (iii) Although the anthropologist Geertz may view a tribal “other” as a person mired in “the gross actualities of blood, race, language, locality religion, or tradition,” from my comparative work I note that, in the more collective polities, theories of the human mind were formulated that highlight the potential for rational thought that is not differentiated in terms of class position, religion, ethnicity, or rural/urban (in the less collective policies, a subaltern class is often viewed by the elite as being cognitively and morally challenged). This uniformitarian sense of the mind was not an attempt to deny ethnic heterogeneity to promote an ideology of cultural homogeneity in the sense of Anderson (42) because, in the cases I studied, ethnic difference was not ignored. To illustrate a pattern of ethnic recognition, I refer again to the interesting Mughal polity, which was ethnically pluralistic but where a cosmopolitan religious ecumenism was promoted by the state termed the “universal peace” (sulh-I kull). The goals here were to make possible open recruitment into positions of governing authority (which depends on a uniformitarian theory of mind) and also to valorize public reasoning to enhance intermingling and cooperation among ethnically distinct sectors. According to Amartya Sen (ref. 43, p. 16), these policies supported “dialogues between adherents of different faiths…[and argued that]…‘the pursuit of reason’ rather than ‘reliance on tradition’ is the way to address difficult problems of social harmony.”

Although uniformitarian policies acknowledged the reality of ethnic pluralism, I suggest that they did establish conditions that would tend to diminish the importance of local collective-action groups by effectively incorporating persons, as citizens, into a larger civic unit that is the source of public goods. And, collective state building is associated with uniformitarian notions of the human that are key to building judicial institutions that offer legal rights across social sectors. Such policies are also the conceptual foundation for policies of open recruitment of persons of different ethnic backgrounds that afforded meaningful participation in official capacities, so the state itself has the capacity to become ethnically pluralistic.

Ethnicity and Commercial Development

Lane Fargher and I discovered a strong positive statistical correlation between our measures of collective action in state building and our measure of degree of commercial development (44). I turn to a consideration of commercialization because, I suggest, in tandem with collective state building, it is also a process that may have relevance for understanding the role of ethnicity in the dynamics of states. Specifically, I suggest that commercialization is a process that will operate side-by-side with political collective action to diminish the importance of ethnic group formation and maintenance. To investigate this possibility, I refer to the Late Postclassic period of Central Mexico (1350–1521 CE), which is an interesting case in which ethnic plurality is well represented, as we know from ethnohistoric documents (van Zantwijk, ref. 45), but sharp ethnic group boundary making is not evident, and, instead, diversity took the form of what Stark (ref. 46, p. 44) refers to as an “ethnic mosaic” in which ethnicity was a cultural fact but migration rendered particular localities ethnically pluralistic.

The Late Postclassic mosaic pattern is consistent with the meager evidence for ethnic signaling, at least insofar as it can be inferred from the distribution of the material culture recovered from archaeological investigations. Elizabeth Brumfiel et al. (47) tentatively identified lip plugs as an ethnic signal of Otomí speakers, at the site of Xaltocan, but, apart from this single example, the material culture of the period (known mostly from the study of ceramic vessels) shows little evidence for stylistic ethnic signaling. Beginning perhaps as early as the Early or Middle Postclassic periods (950–1350 CE), and fully in evidence by the Late Postclassic (1350–1521 CE), regional ceramics were highly stylistically uniform and what ceramic variability can be identified is understood to represent the commercial distribution systems of the major centers of Tenochtitlan and Texcoco rather than ethnic affiliation (48, 49). I suggest that this evidence of decline in ethnic groups can be related to commercialization in two ways. (i) By the Late Postclassic period, the Basin of Mexico and adjacent areas of the Central Highlands were economically integrated by a vast and interlocking periodic market system (50). This system linked together subregional specialist producers while strengthening the economic linkages between rural communities and the populations of the growing urban centers. With commercial growth, even commoners participated in a vastly expanded economic system that transcended local economies and markets. Marketers are known to have attended distant marketplaces outside their own polity and in the process crossing what were understood as traditional ethnic boundaries (e.g., ref. 51, p. 35). (ii) The vast interlocking market system of the Postclassic period and its associated market participation beyond local communities may have played a role in the decline of ethnic group boundaries. Ethnic affiliation and signaling also lose their meaning in a context of a social milieu in which market participants must be able to trust that, in market transactions, including the resolution of disputes, they will be judged only as individuals, not by the usual indicators of identity, including social standing, sex, and ethnicity (51). We can see how neutrality was expressed in the marketplaces, where a socially purposed organization, the Pochteca, served as a largely autonomous paragovernmental market-management system within the larger authority structure of the Late Postclassic Aztec empire (ref. 52, chap. 7). The purpose of the Pochteca organization was to manage Central Mexican marketplaces and to maintain monopoly control over long-distance trading; however, they did that without any reference to a particular ethnic identity. This nonethnic strategy, I suggest, was crucial to establishing trust among marketers that, in their commercial transactions and judicial actions, the Pochteca will maintain a high level of judicial neutrality in the face of ethnic plurality and variable social standing (ref. 52, p. 171).

Conclusion

Collective action theory allows me to expand on Fredrick Barth’s social and cultural constructivist theory of ethnicity. From this perspective, I am able to identify those conditions most conducive to varying forms of ethnic construction. One is a local-scale self-organizing strategy intended to provide mutual defense and to maintain control over resources when the larger political environment is chaotic, especially in zones of refuge and world-system incorporation. Ethnicity is also seen as a productive path to enhancing cooperation in the challenging contexts of long-distance trading between core and periphery zones of a world system. Chaotic conditions also provide a fertile ground for local-scale ethnic group building within the boundaries of failed states that are unable or unwilling to provide adequate public goods, and whose dominating and inefficient control of economy leads to the impoverishment of subaltern classes. From my comparative work, I found that, where collective action was a framework for state building, a goal was to enhance consensus and intermingling of the population in the face of ethnic or other sources of heterogeneity. In these cases, public goods disseminated by the state and the reorganizing of basal social units, alongside a growing commercial economy, very likely diminished the importance of local self-organization, including ethnic group building with its costly signaling strategy.