1.1 Resource Management and the Commons

Since we are examining issues of collective action in the archaeological and historical record, we thought it appropriate to start off with a utilitarian artifact. Figure 1 depicts a cowbell. It is not the sort of relic from the past that gets romanticized by popular media, but this cowbell tells a story that connects with bigger issues of how societies organize themselves. It was discovered in Boston Common during a 1987 archaeological survey and likely dates between 1770 and 1830 (Reference BagleyBagley 2016: 113–114, 131–133). The later date provides what archaeologists call a terminus post quem, or limit after which the cowbell could reasonably date to, because 1830 is when Boston decreed the Common to be a public park, prohibiting residents from grazing their animals on it. For almost two centuries prior, starting in 1634, Boston Common had been set aside for communal pastureland, as well as less bucolic uses such as storing munitions in the town’s powder house and the occasional public execution, including of women accused of witchcraft. In the early years of Boston Common, its use for grazing was restricted to adult white males and overuse during its first decade led to a town ordinance on how many animals could graze, limiting an owner to one cow or four sheep at a time.

Figure 1 Photo of a cowbell recovered from an excavation at Boston Common and dating to ca. 1770–1830, when the common could be used for grazing animals.

Used with permission from Reference BagleyBagley (2016).

Commons of this sort were, well, common in England too and had been transposed to Boston and other North American settlements by Anglo colonists. Soon after Bostonians outlawed grazing on Boston Common, the English author Reference LloydWilliam Forster Lloyd (1833) published two lectures using a parable of the commons titled Two Lectures on the Checks to Population. Lloyd noted the tensions between individual and group interests inherent to managing shared resources, as it is in any single person’s interest to use a resource as much as possible, but it is in the group’s interest that the resource not be overexploited – or overgrazed, in this particular case. Lloyd’s parable was revived and popularized more than a century later as the “tragedy of the commons” by the American ecologist Reference HardinGarrett Hardin (1968), who reflected three decades later that he should have been more specific in branding it a “tragedy of unmanaged commons” (Reference HardinHardin 1998). Although Hardin’s popular coinage for communal resource management has been very influential, his proposed solutions were immoral as he espoused white nationalist and anti-immigrant ideas as the means of managing the commons (Reference MildenbergerMildenberger 2019). Hardin’s notions of who among a population actually gets to use the commons were not much different than those of the settler colonists of three centuries earlier.

What we consider abstractly as “the commons” is therefore subject to societal dynamics of governance, social inequality, definitions of in-groups versus out-groups, and related factors. This has been particularly well illustrated by the recent work of Reference HeatherHeather McGhee (2021), who examines how the intersection of racism and classism in the United States has historically undermined collective action in managing common-pool resources, including, most evocatively and literally, in access to public spaces such as community swimming pools. These cultural, historical, and societal factors defining inclusion and exclusion can be among the most determinative in how individuals define themselves as groups and whether they do or do not cooperate with one another.

Since its elaboration as a coherent body of theory in the 1960s by scholars such as Reference OlsonMancur Olson (1965), collective action theory has offered a productive, interdisciplinary framework for explaining how and why people act out of selfish interests or as contingent cooperators and how that affects who has access to particular resource systems. We are partial to the work of Reference OstromElinor Ostrom (e.g., 1990, Reference Ostrom1998, Reference Ostrom2000, Reference Ostrom2005, Reference Ostrom, Boix and Stokes2007), the first woman to be awarded the Nobel Prize in economics, as she skillfully bridged multiscalar levels of analyses, from elegant yet abstract mathematical models derived from game theory to the thick description of how irrigation and other resource management systems work on the ground, drawing from actual ethnographic and historical cases. Reference Ostrom, Gardner and WalkerOstrom and colleagues (1994) developed an organizational framework for classifying types of goods, resource systems, and the cooperation dilemmas that may arise from misalignments between individual and group interests or from asymmetries in power (see also Reference OstromOstrom 2005: 24–26). In Figure 2, we elaborate on the simple graphic schema they developed to illustrate these points with a sampling of examples that are often considered in archaeological discussions of managing land and other resources (e.g., Reference Bayman and SullivanBayman and Sullivan 2008; Reference Blanton and FargherBlanton and Fargher 2008; Reference Carballo and FeinmanCarballo and Feinman 2016; Reference EerkensEerkens 1999; Reference KohlerKohler 1992; Reference OosthuizenOosthuizen 2013).

Figure 2 Text box schema presenting a classification of goods and resource systems, based on Reference Ostrom, Gardner and WalkerOstrom and colleagues (1994), with examples of infrastructure and management strategies frequently discussed by archaeologists.

The two major axes for classification are the subtractability of a resource, meaning how much people using it depletes the resource, and its potential for excluding others from using it.

The representation is only schematic, and the arrows in Figure 2 are intended to indicate how any type of good or resource could vary depending on factors beyond ecology and economy, especially cultural, social, or historical ones. In general, we might consider a spectrum from fully public to fully private resources that reflects how easy it is for an individual or small group of people to exclude others from that resource – the exclusion axis on the left. But the framework also considers the scale of the resource – the subtractability axis on top – and the degree to which any use of it impacts others doing so as well. Resources with high subtractability are more finite than resources with low subtractability, such as energy from fossil fuels versus solar power. Their higher susceptibility to overexploitation is why so much attention has been given to investigating common-pool resources (CPRs, in this literature). Management of common-pool resources among ancestral populations who primarily hunted, fished, or collected foods (i.e., “hunter-gatherers” or “fisher-foragers”) includes examples from the Southeastern United States, where rich ecosystems permitted seasonal or permanent sedentism on the part of relatively large populations (Reference ThompsonThompson 2022). Since these subsistence resources were not privatized and were subtractable, they present common-pool resource problems analogous to contemporary management of the same – such as preventing the overexploitation of game species, fisheries, and forests (Reference AchesonAcheson 2011, Reference Acheson2015). Although global populations have vastly increased, and our exploitation of such resources has been mechanized and industrialized, the basic logic of collective action remains the same: in any given social-ecological system, how resources are classed and managed derives from interactions between different nested scales of resources, users, and governing structures (Reference Andersson and OstromAndersson and Ostrom 2008; Reference OstromOstrom 2009).

1.2 Governance and Fiscal Systems

To provide other examples of the sorts of resource issues outlined in Section 1.1 and their intersection with governing structures, let us consider the hypothetical example of an agricultural society where successful farming is based on receiving adequate rainfall. In a fully rainfed system, rainwater could be classed as a public good since it falls equally on everyone’s field and does not exclude anyone or subtract from the possibility of it going somewhere else. Yet if this same society devises an irrigation network to channelize rainwater towards particular fields, that network could now be classed as a common-pool resource and governing structures relating to how the irrigation canals and associated fields are managed (i.e., centrally vs. locally, collectively vs. privately) become the key variable in understanding how resource management was organized. Governing systems can impact the full spectrum of resource systems, not simply common-pool resources. Within public goods we list examples mostly drawn from the built environment and urban infrastructure, assuming high access among a population. Yet an aqueduct designed to bring water to a palace rather than a wider public would, in contrast, be classed at the other end of a spectrum as a private good. Alternatively, if the use of a public good were somehow governed by expectations that users participated in its construction or maintenance, there could be social mechanisms to exclude those who did not contribute – termed “free riders” in this literature – from enjoying the benefit. Sanctions for such transgression could range from light and informal measures, like gossip or shaming, to formal ostracism from the group or even death (Reference BoehmBoehm 2012: 84–87; Reference Carballo and CarballoCarballo 2013a: 11–13; Reference OstromOstrom 1990: 94–100).

Spot resources, listed roughly in the middle of the chart, are easier to exclude others from obtaining than are those with wide distribution and could even be monopolized by a few individuals or small, powerful groups, classing them as private goods or toll/club goods. On their own, the management of spot resources offers dilemmas that social scientists usually find less interesting to probe, but not when they are part of the fiscal systems of governing institutions. These relationships were examined in comparative, historical perspective by the political scientist Reference LeviMargaret Levi (1988) and were elaborated from historical premodern states by the archaeologists Reference Blanton and FargherRichard Blanton and Lane Fargher (2008) in considering the relationships between fiscal streams and systems of governance more broadly. Their detailed analyses of comparative cases demonstrate that a key factor determining variability in the organization of large-scale societies is whether revenues are largely internal (more communal and widely available resources) or external (such as spot resources, often coming from long distances, but also war booty, tenant rent, and toll taxes on long-distance trade). This axis of political-economic variability is also classified as a spectrum between inclusive and extractive political institutions (Reference Acemoglu and RobinsonAcemoglu and Robinson 2012). It is a relatively simple distinction that has major consequences for the levels of accountability that leaders face from the populace at large and the likelihood that governance and power will be shared versus individualized (Reference Blanton and FargherBlanton and Fargher 2008: 114–132, Reference Blanton and Fargher2016: 106–114). In addition to greater accountability for political elites, higher levels of internal financing and communal resources tend to be positively correlated with higher dissemination of public goods and bureaucratization of civic offices, together creating systems that may grow larger and be more resilient to social perturbations than more autocratic or despotic formations (Reference Blanton and FargherBlanton and Fargher 2016; Reference Feinman and CarballoFeinman and Carballo 2018).

An important point to emphasize here is that collective action operates at multiple social scales – ranging from cooperating households to the political alliances of large states (Reference BirchBirch 2022) – and while the particular forms of collective action characteristic of any particular society shape the broad contours of societal organization, they are not immutable (Reference Arroyo Abad and MaurerArroyo Abad and Maurer 2021). In fact, they change frequently as individuals and groups strategize their actions in association with shifting socioeconomic variables and historical contingencies. Human decisions to shift their social norms and levels of trust in institutions, as well as in the variable arrangements for financing governance, have profound cumulative effects for broader societal organization. We can draw from two canonical works of anthropology to illustrate the types of mechanisms by which such variability and change can occur. In the classic work of political anthropology Political Systems of Highland Burma, Reference LeachEdmund Leach (1954) developed a model for Kachin social organization that involved oscillations between a more egalitarian and pluralistic structure (gumlao) and a more hierarchical and exclusionary structure (gumsa).

Leach’s work was important in showing the possibility for political variability and change among people with strongly shared cultural norms. He illustrated how relatively minor changes in ritual practices could be a major driver of this variability. For instance, within gumsa groups a chief was entitled to conduct major sacrifices, promote supernatural entities associated with his lineage over those of others, and pass his ritual powers off to sons, whereas such initiatives were more distributed, community responsibilities in gumlao groups, who prioritized loyalty to place over lineage. Since Leach’s study, the reasons for variability and change in Kachin groups along the gumlao/gumsa spectrum have been debated by scholars, with some arguing that more egalitarian organization was the result of Kachin chiefs losing the external fiscal streams (in opium, enslaved peoples, and taxes extracted on trade caravans and as a protection racket) that underwrote their power and others arguing that egalitarianism was a conscious choice made by particular groups in strategic opposition to the hierarchical states (the Shan and the English) within their broader sphere of interaction (Reference NugentNugent 1982; Reference Robinne and SadanRobinne and Sadan 2007; Reference ScottScott 2009).

This second explanation, of definition in purposeful opposition to another, recalls the yet earlier canonical work in anthropology by Gregory Bateson, who in Naven (Reference Bateson1936; see also Reference BatesonBateson 1935) elaborated the concept of “schismogenesis” as the process by which individuals in a society might progressively differentiate themselves, thereby creating schisms with others. Like with the gumlao/gumsa oscillations, schismogenesis provides a heuristic tool for thinking about purposeful and agentive change, rather than envisioning particular social “types” to be fixed or inevitable due to external ecological factors or power dynamics. The concept has also been applied to entire cultures in contact, with examples from classical antiquity such as democratic Athenians versus oligarchic Spartans or, more broadly, pluralistic Greeks versus autocratic Persians (Reference Graeber and WengrowGraeber and Wengrow 2021: 56–58, 180). If considered as a process of divergent ethnogenesis, schismogenesis has value in considering ideational mechanisms that can lead to the definition of in- and out-groups structuring conflicts over more material concerns, such as resource management. For instance, it has been used to describe progressive differentiation leading to factional conflicts over common-pool resources, including salmon fisheries among Native populations of the Pacific Northwest (Reference Harrison and LoringHarrison and Loring 2014), connecting back to issues of resource management (see also Reference BlantonBlanton 2015; Reference FeinmanFeinman 2022).

The interactions between people, resources, and governing structures therefore have great explanatory value in considering variability and change in societal organization in the deep past. Archaeology offers a material lens on these issues and a diachronic, deep-time perspective on the resources people used and their identification as part of polities, settlements, ethnic or religious groups, gender or age divisions, households, and even sometimes as individuals. A challenge is how to identify governance archaeologically, including what Reference Holland-Lulewicz, Thompson, Birch and GrierHolland-Lulewicz and colleagues (2022) term the “keystone institutions” that promote and sustain collective action and more pluralistic governance.

1.3 Institutions, Infrastructure, and the Archaeological Record

Because archaeology engages primarily with a material record of the past, the sorts of undertakings that people may have worked more or less collectively towards are interpreted largely through elements of the built environment such as architecture and infrastructure, the relative distributions of economic goods, and imagery that may convey past ideas about human relations among one another and with supernatural entities (Reference Blanton and FargherBlanton and Fargher 2016; Reference CarballoCarballo 2013c; Reference StanishStanish 2017). These material remains are widespread across the globe, but determining the cultural institutions that promoted and organized them can be more challenging as, “[b]roadly defined, institutions are the prescriptions that humans use to organize all forms of repetitive and structured interactions including those within families, neighborhoods, markets, firms, sports leagues, churches, private associations, and governments at all scales” Reference OstromOstrom (2005: 3). Although the term often implies formality or rigidity, as in their imposition by powerful groups, we adopt a broad definition of institutions in this Element because anthropologists and other social scientists have demonstrated how they can form through more organic, bottom-up processes of repeated interactions by individuals adhering to widely agreed upon cultural norms and rules (Reference GlowackiGlowacki 2020). Being cultural templates or prescriptions for behavior, understanding the institutions of past societies usually requires textual evidence from the period, historical connections drawing on oral histories and contemporary practices of descendant communities, or judicious analogy with cases that appear similar in related institutions. Interpretations based on the first two lines of evidence tend to be more compelling than those based on the third since they are grounded in the cultural logic of a given society.

Today we often think of infrastructure in relation to cities or the transportation networks that link urban areas together. Yet a recent emphasis on “green infrastructure” in cities, usually meaning green spaces like parks or environmental mitigation strategies in response to climate change, articulates with a much longer history of managing ecologies for subsistence through a human built environment – and we are not the only species to do so, as beaver dams, termite mounds, and other nonhuman built environments attest to (Reference BaruaBarua 2021). A broader archaeological definition of infrastructure could include its materialization of some sort of consensus undertaking between people through “the manifestation of multi-user physical networks” (Reference SmithSmith 2016: 165). Though these arrangements certainly ramped up with more “users” living near one another on a permanent basis (Reference Feinman and NeitzelFeinman and Neitzel 2023), they can also be seen among mobile foraging populations. In fact, some of the earliest subsistence infrastructure anywhere in the world could be said to be the desert kites of Pre-Pottery Neolithic B (PPNB) in southwest Asia – large linear stone formations that PPNB hunters began constructing and depicting in engraved-stone monuments some 9,000 years ago as communal hunting strategies to funnel herds of gazelles, aurochs, and other species to specific kill zones or pit traps (Reference Crassard, Abu‑Azizeh and BargeCrassard et al. 2022, Reference Crassard, Abu-Azizeh and Barge2023). Zooarchaeological remains and human-animal imagery in the sculpture from Göbekli Tepe and other sites in southeastern Turkey, representing some of the earliest documented large communal structures in the world, connect this type of collective construction to public architecture and monumental sculpture of early religious traditions (Reference Pöllath, Dietrich and NotroffPöllath et al. 2018). Similar features of stone lanes for driving large migratory animals such as caribou were constructed by groups of cooperating hunters in North America at least 9,000 years ago (Reference O’Shea, Lemke, Sonnenburg, Reynolds and AbbottO’Shea et al. 2014). Large fish weirs are in a sense the aquatic parallel to these sorts of terrestrial features and allow for the corralling of fish for easy spearing and netting. Their construction among fisher-forager populations such as Florida’s Calusa represents a related form of collective subsistence infrastructure (Reference Thompson, Marquardt, Walker, Roberts Thompson and NewsomThompson et al. 2018).

Cases of communal hunting and fisheries management such as these have been productive lines of analysis for collective action relating to common-pool resources and how its organization articulated with cultural norms and institutions. Among agricultural populations, investment in subsistence infrastructure continued or intensified, with the construction of networks of terraces, irrigation canals, reservoirs, wetland raised fields, and others that would fit in the schematic of Figure 2 representing forms of landesque capital or sunk costs in the built environment whose construction and maintenance generate long-term economic dividends if cooperative arrangements can be sustained (Reference Feinman, Carballo, Nicholas and KowalewskiFeinman et al. 2023; Reference Widgren, Hornborg, McNeill and Martínez-AlierWidgren 2007).

And how were cooperative arrangements sustained? Through the archaeological record we can see a few pathways that operate at different social scales and implicate different forms of archaeological data: one set relating more to space (i.e., architecture and the built environment), from supra-household groups like corporate houses or neighborhoods to public arenas for gathering, ritual, and performance; and one set relating more to things (i.e., artifact and ecofact distributions, household assemblages, or mortuary deposits) bearing on levels of social inequality and the fiscal underpinning of power relations. We will examine these two broad sets of data in turn, along with the interpretations archaeologists have drawn from them regarding collective action.

The first, more spatial dimension to sustaining cooperation dovetails with a large literature in the social and behavioral sciences on mutual monitoring – the shared understanding generated when people are able to keep track of one another and evaluate their level of participation in and commitment to collective undertakings, leading them to trust that a given system is “fair” (e.g., Reference Bowles and GintisBowles and Gintis 2002; Reference ChweChwe 2001; Reference Ostrom, Gardner and WalkerOstrom et al. 2003).

One means of reducing the costs of monitoring is to have multiple families live together in exceptionally large houses featuring continual interaction and shared domestic tasks capable of creating scalar economies within the confines of shared residential walls (Reference Banning and CouplandBanning and Coupland 1996; Reference HirthHirth 2020: 22–23, 36–38). These will naturally have limits on their total size, but some “big house” traditions featured residences that could accommodate several dozens to low hundreds of people. At the larger scales of neighborhood or entire settlements, the costs of monitoring could be achieved through integrative public ritual and the ritualization of collective work (Reference StanishStanish 2017). Here, elements of the built environment such as open plazas or courtyards could provide venues to organize and incentivize participation in collective endeavors through work feasts or public recognition. The archaeological record is full of plazas and communal architecture that could be interpreted as integrative, but it is important to consider other contextual information because these settings can also be used for public performance of inequality and displays of hierarchy (Reference Dungan and PeeplesDungan and Peebles 2018; Reference Feinman and MurphyFeinman 2016; Reference FisherFisher 2009; Reference InomataInomata 2006). Open plazas could just as easily be used for pluralistic purposes of public assembly and lively political debate (the classical Athenian Agora) as for displays of autocratic rule and squashing dissent (Beijing’s Tiananmen Square in the late twentieth century). Likewise, all pyramidal type monumental architecture was not created for the same reasons, and those that were built as mortuary complexes for divine rulers or to extoll the virtues of a dynasty clearly projected more exclusive and unequal symbolism than those built in honor of widely venerated deities or broad cosmological principles. Levels of access, related symbolism, and other cues to how such public spaces were actually used are therefore critical to their analysis and interpretation.

Another important subset of the built environment are spaces created or set aside for purposeful deliberation in a more pluralistic, rather than autocratic, system of governance. These can also vary greatly by culture and through time so require other lines of information other than spatial ones to determine their uses. For instance, because of a rich textual record and well-preserved architectural remains, we can see tiers of democratic decision-making in the urban plans of classical Greek poleis centers, such as dedicated council houses (bouleuterion) adjacent to the open and multifunctional agora, but with restricted access and seating for a council that would deliberate on items to present at a larger popular assembly (ekklesia) that may need to meet at the town’s theater (Reference Paga and NevettPaga 2017; Reference SmallSmall 2009). The openness of the agora and semicircular seating of the council and theater buildings were conducive to a spectrum of political discourse among those defined as citizens. Other spaces for pluralistic assembly might be hard to detect in the absence of texts because of their minimal modification or architectural elaboration, such as the Germanic thing and later progeny in the Icelandic Althing (Reference ThurstonThurston 2022). The construction of more elaborate, centrally located architecture with high accessibility may serve as a spatial cue for deliberative spaces in context that lack supporting texts, as has been proposed for colonnaded structures in cities of the Indus Valley civilization (Reference GreenGreen 2022). For cases lacking textual documentation, it is also difficult to discern who were able to participate in political discourse within deliberative spaces and how a society defined in-groups and out-groups. Comparative studies of Greek and Roman political history highlight another axis of variability in the inclusion/exclusion of people defined as citizens, with the first having defined this in-group more restrictively but empowered those select few with decision-making abilities in a more pluralistic system of governance, while the second defined citizens more inclusively, particularly with imperial growth, but as part of a less pluralistic system of governance (Reference Rhodes and BalotRhodes 2009).

In some cases, early institutions of more pluralistic governance can be gleaned from triangulating between the known institutions of later sources and the archaeological record. Such is the case for council houses in the Southeastern United States, where Native peoples such as the Muskogee have centuries of documented use of these deliberative spaces and can be identified archaeologically dating back some 1,500 years at Ancestral Muskogean sites (Reference Holland-Lulewicz, Thompson, Birch and GrierHolland-Lulewictz et al. 2022; Reference Thompson, Holland-Lulewicz and ButlerThompson et al. 2022). The largest council houses measure more than 35 m in diameter and could have held hundreds of people, not unlike the bouleuterion of classical Greek democratic instructions. A smaller but particularly well-preserved example was discovered nearly a century ago at Ocmulgee in central Georgia (also known as Macon Mounds), which was identified both as an “earth lodge” and as a “council chamber” during the national collective labor projects of New Deal archaeology (Reference FairbanksFairbanks 1946; Reference KellyKelly 1938). The circular structure was well preserved, with burnt timbers from the roof covering a packed clay floor some 13 m in diameter with a bench running along the circumference that becomes progressively lower towards the entrance from a platform depicting a raptorial bird with a forked-eye motif – a trait characteristic of peregrine falcons, admired as the animal capable of the fastest speeds in the world (Reference ByersByers 1962) (Figure 3). Featuring forty-seven seats along the bench and spaces for another three on the falcon platform, the structure was designed to seat fifty people in a manner that simultaneously expressed some social hierarchy in spatial proximity to the platform versus the entrance but also in a spatial arrangement designed to ensure mutual visibility and allow for group deliberation. The structure is therefore a wonderful material encapsulation of tensions between hierarchy and heterarchy, as well as the deep-time history of pluralistic institutions within the Southeastern United States.

(based on Reference FairbanksFairbanks [1946: fig. 21])

Figure 3 Illustrations of a circular council chamber from the Ocmulgee site in Georgia, showing (a) reconstruction of falcon platform, adjacent seats, and central hearth (based on Reference Dickens and McKinleyDickens and McKinley [2003]) and (b) plan view illustration from excavations.

Another relevant line of research on the built environment, based mostly in sociology, deals with public gathering spaces as “social infrastructure” that fosters more equitable and resilient communities by providing opportunities for neighbors to have face-to-face contact in recreational settings, such as libraries, parks, playing fields, and town squares (Reference KlinenbergKlinenberg 2018). Scholars researching contemporary social infrastructure have found strong positive correlations between its equitable distribution and greater social capital in the form of more trusting and cooperative interpersonal relations within both urban and rural settings (Reference FlorFlora 1998; Reference Latham and LaytonLatham and Layton 2019). In some cases, social infrastructure is a form of public good distributed by more collective civic or governmental agencies, while in others it can be instigated through self-organizing processes by more localized, community-driven configurations, such as by coordinated families or neighborhoods. In either case, its presence as an arena for convivial social interaction enables virtuous cycles of prosocial behavior (Reference Blanton and FargherBlanton and Fargher 2011). As an example, US post offices in the nineteenth century could be considered somewhere between hard and social infrastructure, as they were gathering places that kept communities connected and information flowing between them. A historical survey of where they were first established and how many existed per county demonstrates strong correlations between the establishment of early post offices and higher incidences of social capital today, measured in indices such as having more per capita community newspapers and civic and social organizations, and fewer social ills, such as lower arrests and mortality rates (Reference Jensen Jeffrey and RameyJensen and Ramey 2020). Historical analyses of this type indicate that early investments in community infrastructure have long-term ramifications for social capital and community resilience (see also Reference Putnam, Leonardi and NanettiPutnam et al. 1993; Reference Yue, Antonietti and AlirezaeiYue et al. 2022).

Issues of equity and social capital can also be evaluated through smaller-scale material remains in the archaeological record, such as distributions of artifacts and ecofacts, health indices reflected in skeletal remains, and the relative size and elaboration of houses and burials. These are all long-standing archaeological concerns, but recent approaches interested in understanding variability in past sociopolitical organization have found novel ways to evaluate them more empirically through calculating Gini indices using various archaeological proxies of wealth (e.g., Reference Kohler and SmithKohler and Smith 2018) or through comparative analyses of the fiscal financing of governance and institutional economies (Reference Blanton and FargherBlanton and Fargher 2008: 112–132; Reference HirthHirth 2020: 154–193). Much of the work on Gini calculations for gauging relative levels of inequality use domestic living space as a proxy, since it is widely generalizable to different cultural contexts unlike determining what commodities were valued in any setting or accounting for cultural variability in mortuary practices and furnishings. The most exclusive and inequitable architectural complexes of the premodern world were of course palaces, and any society centered on them or showing great variance between the sizes of elite and non-elite houses is clearly less collective than one in which living space was more equitably allocated. Nevertheless, other variables such as location, quality of construction materials, and elaborateness of decoration also figure prominently in the social differentiation materialized in houses (Reference BlantonBlanton 1994).

Other archaeological signatures of fiscal systems and inequality include whether revenue streams for political economies were based more on staple or wealth goods (e.g., Reference D’Altroy and EarleD’Altroy and Earle 1985) and the possibility for more exclusionary control along the lines illustrated in Figure 2. A few key axes of variability in how we interpret the collectivity of social organization as seen in the archaeological record are outlined in Table 1. Variables are generalized so that they could be applied to settings throughout the globe and also may be manifested in different types of material remains, necessitating integrated and multiscale analyses of space, things, and imagery. Most societies feature mixtures of attributes along these axes, and we need to envision possibilities of change over time, as gumlao/gumsa cycles or other well-documented societal oscillations highlight. We agree with Reference Blanton and FargherBlanton and Fargher (2008: 293) that “Collective action process may be expressed in diverse cultural settings and time periods, and is especially sensitive to local conditions that have an impact on revenue streams, while cultural code may be a less influential causal force, or may be more of a dependent variable.”

1.4 Summary

To this point, we have outlined a framework, drawn from a range of disciplines, that is focused on the contingent factors that foster or impede human cooperation and underpin diverse forms of collective action among social groups. Human agency, hence the potential for selfishness, is universal (Reference Melis and SemmannMelis and Semmann 2010; Reference SökefeldSökefeld 1999), and yet as a species we are unrivaled in our abilities to cooperate in large aggregations and affiliations with nonkin. For us, the key assignment is to outline the factors that constrain and foster such collaborative formations. Whereas scholars studying resource management, such as Reference OstromOstrom (1990, Reference Ostrom2005), tend to engage more with the allocation and organization of public goods among smaller groups of “conditional cooperators,” those drawing on comparative historical cases, such as Reference LeviLevi (1988, Reference Levi1996), take a more macro-approach to political and economic systems, and especially the role of revenue streams and fiscal financing. It is important to note that we do not see these approaches as at odds with one another; rather, they provide insights for dealing with different scales of human interaction.

In subsequent sections, we focus our conceptual lens on Mesoamerica, one of the parts of the world where a diversity of keystone institutions, urban settlements, large-scale polities, and extensive socioeconomic networks arose independently. Yet those interpersonal affiliations took a diversity of forms, across space and social scale, and they also varied markedly in cultural regions and through time. Although we recognize that our current interpretations and hypotheses are certainly not final, we hope that the new framing offered here provides a firmer foundation to understand the prehispanic Mesoamerican world and to integrate more comprehensively the lessons and Indigenous knowledge of that world into global comparisons and constructs.

2.1 Cultural Institutions and Social Theories

During the early twentieth century, anthropology and archaeology emphasized culture-history as researchers sought to define culture areas and create chronological sequences of their developments and associated material culture. The German-Mexican anthropologist Reference KirchhoffPaul Kirchhoff (1943) first defined Mesoamerica along a series of traits that he compared with those of other peoples of the Americas. Subsequent research has shown how some of these traits are problematic while others are of uneven utility for studying the archaeological record, due mostly to differential preservation of organic remains and nonmaterial forms of culture (Reference CreamerCreamer 1987). Among those traits or practices that Kirchhoff proposed as distinctive to Mesoamerica were a dual-calendar system combining the 365-day solar year with a 260-day ritual year; the processing of maize in an alkaline solution (nixtamalization), allowing for the formation of tortillas as a more nutritious staple food; specialized commercial markets; lakeshore wetland agricultural fields (chinampas); stepped temples covered in lime stucco; and ballcourts featuring rings for scoring. Of these, probably only the first three were present in most or all of Mesoamerica, but they are worth emphasizing as broadly shared means of coordinating people by parsing time and subsistence/economic activities potentially amenable to collective action. Team sports such as Mesoamerican ballgames provide a nice encapsulation of human dynamics of cooperation with an in-group in order to compete with an out-group, and whether or not temples represent a public good depends on how they are used and what symbolic messaging they convey, as we saw in Section 1. In a next tier of traits present in Mesoamerica but also in other parts of the Americas, Kirchhoff listed agricultural terraces, specialized craft economies, and the importance of intermediate-scale corporate-kin groups such as the Aztec calpolli (“big house”) – all also relevant to considerations of how people affiliated with nonkin and worked either cooperatively or antagonistically.

More recently, Mesoamerican scholars have considered the culture area as an assemblage of institutions, rather than checkbox-type trait lists as were compiled by cultural historians, allowing for more comparative assessment by subregions or through time. Pervasive keystone institutions at the time of the Spanish invasion included in domains we could categorize as politics (states); settlement tiers (cities, neighborhoods/barrios, rural communities, households); economics (marketplaces and irrigation societies); and ritual/religion (temples) (Reference Kowalewski, Heredia Espinoza, Bondarenko, Kowalewski and SmallKowalewski and Heredia 2020). A reframing of the Mesoamerican past to emphasize institutions and individual practices within them (see Reference Joyce, Hendon, Overholtzer and JoyceJoyce 2021) allows us to query issues such as the relationship between social scale or population with the number of institutions, their complexity, or their degrees of collectivity and trust, and to identify which institutions contributed to sustaining or worked to undermine collective action.

Relationships between the scale of societies and the nature of their institutions turn out to be not as tightly correlated or predictable as early models of social evolution imagined. For instance, rather than observing clear correlations between social scale and the likelihood that resource systems will be successfully self-organized, comparative analyses by Reference OstromOstrom (2005: 251–252) and others highlight the greater importance of actors making strategic tradeoffs between the costs of providing a public good or managing a common-pool resource (for instance, in transaction costs or in sanctioning for noncompliance) and overall value of the resource system (for instance, the subtractability of the particular resource). Operating at the highest level of concentrated power and largest scale of polities, Reference LeviLevi (1996: 10–16) emphasizes the critical role of trust in state institutions to societal stability and the historical resilience of governing systems that (1) make credible commitments to its citizens, (2) are seen by them as upholding fair procedures, and (3) foster reciprocal sentiments of trust. She also identified key factors affecting perceptions of fairness as involving a combination of (a) coercion or sanctioning of those who neglect their responsibilities, (b) cultivation of a universalistic, merit-based ethos, (c) creation of impartial institutions (judicial, electoral), and (d) more pluralistic decision-making allowing for broad participation in the institutions of governance (see also Reference LeviLevi 2022). Several of Levi’s points overlap with efforts by archaeologists to define corporate cognitive codes that have historically promoted more pluralistic systems of governance (e.g., Reference Fargher, Fargher and EspinozaFargher 2016). It is worth considering both sets of relationships in the light of Mesoamerican institutions for managing resources and governing.

A reframing of Mesoamerican archaeology around institutions, including some of the ethnographic traits used in cultural-historical definitions like Kirchhoff’s, enables more multiscalar, comparative analyses and moves us beyond top-down or unilinear evolutionary models of societal development. Top-down perspectives of Mesoamerican sociopolitical organization derive primarily from two historical currents of the twentieth century: an early sample bias in archaeological research towards seats of power – political capitals, their temples and palaces – and elite material culture such as tombs and courtly art; and the uncritical use of ideas stemming from Marxist historical materialism, corresponding to modes of production as stages of evolution to group early cultures, such as “Oriental Despotism” (Reference Blanton and FargherBlanton and Fargher 2016: 99–106; Reference Gándara, Nichols and PoolGándara 2012; Reference OffnerOffner 1981). While the mid-twentieth-century shift to consider settlements across the entire scalar spectrum and the houses of non-elites helped to start to rectify the top-down biases of archaeologists excavating only pyramids and palaces, conclusions of the research were also often framed using unilinear models of societal development that did not allow for variable forms of organization and institutions. Our goal in this section is to illustrate precisely that variability.

2.2 Indigenous Mesoamerican Institutions of Working Together

A rotary tax delivered in labor, sometimes referred to using the French term corveé and also as labor tribute, was a key institution of Aztec society and is documented among other Mesoamerican groups as well (Reference Carballo and CarballoCarballo 2013b; Reference Rojas Rabiela, Frost, Meyer, Vázquez and DíazRojas Rabiela 1977, Reference Rojas Rabiela, Medina, Austín and Serra1986). The above passage was written by Nahua scribes working in a trilingual (Nahuatl, Spanish, and Latin) documentation project overseen by the Spanish friar Bernardino de Sahagún in the decades following the establishment of colonial New Spain. It speaks to a type of civic service termed coatequitl – literally “snake/twin work” but implying a sense of reciprocity – performed by youths on behalf of their ethnic state (the altepetl) as part of their compulsory system of education open to all youths (the telpochcalli) and covering subjects we might gloss as civics, history, calendrics, martial arts, religion, and ritual. The rich historical corpus from sixteenth-century central Mexico, compiled by both Spanish and Indigenous authors, makes such institutions of the greater Aztec world the best known for Mesoamerica. In Table 2, we provide some examples of Nahuatl terms relating to key resources and the social and political institutions involved in managing them. Some terms overlap and represent regional variability, but they help us move from abstract notions of collective action by providing concrete examples of Indigenous institutions and potential axes of variability within them. We also consider parallel cases among other Mesoamerican peoples.

Forms of collective labor that mirror the Nahuatl terms tequitl – a suprahousehold scale yet smaller than town or polity – and coatequitl are still practiced by descendant Indigenous and mestizo (mixed, Spanish-speaking) communities in Mesoamerica today. They are well-documented ethnographically through the twentieth century, including in Nahua communities of central Mexico (Reference GoodGood 2005; Reference LewisLewis 1963: 108–111), Maya communities of the Yucatan (Reference Redfield and Villa RojasRedfield and Villa Rojas 1967: 77–80), and Mixtec (Reference MonaghanMonaghan 1990) and Zapotec (Reference CohenCohen 1999) communities of Oaxaca. Labor is typically organized through social units intermediate in scale between households and towns or municipalities. Depending on the setting, the intermediating units might be Spanish synonyms for neighborhoods, such as barrios or colonias, or may be termed demarcaciones or secciones, usually implying something larger along the lines of a ward or district. Work teams tackle public infrastructure projects involving water systems, roads, churches and municipal buildings, schools, and social infrastructure such as recreational fields. Some examples of contemporary, small-scale water acquisition and distribution infrastructure from the semiarid highlands of Oaxaca are illustrated in Figure 5 and include wells for tapping groundwater and check dams and irrigation channels for capturing rainfall and distributing it to fields.

Figure 5 Photos of water management infrastructure in the contemporary Valley of Oaxaca, including (a) a small well, (b) check dams, and (c) a small irrigation channel.

In communal work of this type, individuals who neglect their labor duties might be sanctioned through fines or even jail time, while participants in the system are rewarded through collective meals and the positive reputation that comes with good citizenship – a form of ritual economy that Reference MonaghanMonaghan (1996) has alliteratively termed “fiesta finance.” Group work responsibilities often represent one of the defining principles for Mesoamerican community affiliation (Reference CohenCohen 1999: 9; Reference MonaghanMonaghan 1990: 768). In fact, in the central Mexican state of Tlaxcala during the mid-twentieth century public works were referred to as comunidad (community) and were an expected weekly contribution on the part of adult males. One project within a seasonal wetland environment (humedal) was documented by Reference WilkenWilken (1968: 228) as involving more than 100 community members who participated in maintaining the common-pool resource of a drained-field system by clearing and widening canals over two days, overseen and celebrated by municipal officials.

The example of drained-field maintenance in the wetlands of southern Tlaxcala evokes the most intensive agricultural systems of prehispanic Mesoamerica: the chinampas (chinamitl) constructed along the lakeshores of the Basin of Mexico (Reference CordovaCordova 2022: 196, 206–214; Reference Frederick, Thurston and FisherFrederick 2007). As a form of landesque capital that involved sunk costs of labor to create and continually maintain, with the payoff of exceptionally productive agricultural yields, chinampas involved the interplay of institutions that could be classified as top-down and bottom-up (Reference Morehart, Nichols and Rodríguez-AlegríaMorehart 2017, Reference Morehart and Holt2018). In the most productive setting of the southern freshwater lakes, Aztec chinampas had a long and narrow configuration, measuring approximately 50 m by less than 4 m, and as much as half of the crops grown on them were ceded to imperial tax officials as a tax in kind. Relic Aztec canals could still be seen in aerial photos of the 1940s (Reference Luna GolyaLuna Golya 2014), while fields that remained active at that time had been modified into shorter and wider parcels (Figure 6). Certain pre-Aztec chinampas of the saline northern lakes were, in contrast, abandoned following imperial expansion in the region. Like other resource systems, chinampa management therefore involved political dynamics between users and governing structures who, in this case, oversaw intensification efforts in order to provide a consistent, internal tax base in staple goods.

Figure 6 Base image shows (left) relic Aztec chinampas and (right) active chinampas from a 1941 aerial photo modified from the Compañía Mexicana Aerofoto, while inset image shows the similar configuration of the chinampas from the sixteenth-century Plano Parcial de la Ciudad de México (Mediateca INAH). https://mediateca.inah.gob.mx/islandora_74/islandora/object/codice%3A635

Chinampas and other agricultural fields existed on land that was managed as a resource along a spectrum from private to common-pool, as indicated in Figure 2. The Nahuatl land terms in Table 2 provide an indication of land that was more privately owned, typically by nobles or other social elites, and land that was owned and managed by suprahousehold corporate groups. In the case of the latter, use was allocated to individual families, making them semi-privatized, but under norms of land tenure that fields would be tended to or risk being absorbed and reallocated by the calpolli or another corporate group (Reference Harvey, Harvey and PremHarvey 1984; Reference HirthHirth 2016: 36–41; Reference LockhartLockhart 1992: 141–163). The base of Aztec and other Mesoamerican subsistence economies thereby rested on land and the labor required to construct and maintain infrastructure of agricultural intensification, such as irrigation canals and terraces (Reference Rojas Rabiela, Ruiz and Murillo LiceaRojas Rabiela et al. 2009). These networks of subsistence infrastructure were generally managed as common-pool resources because, although individual fields may be allocated to particular families, any dereliction in maintaining one’s individual segments of the system adversely impacted others. Communal lands such as those managed by a calpolli (calpotlalli) also included forest lands that included various common-pool resources that were in the group’s interest not to overexploit, such as stocks of lumber and game (Reference Harvey, Harvey and PremHarvey 1984: 91).

Critical to the organization of Aztec labor obligations was the coordination provided by the ceremonial calendar and its associated sequence of monthly (twenty-day) rites (Reference GillespieGillespie 1989: 210–215; Reference van Zantwijkvan Zantwijk 1985: 24–25). The ceremonial calendar “mapped social relations in space and time” (Reference GillespieGillespie 1989: 211). This calendrical scheduling of communal labor is emphasized by the Nahua author of the quote at the beginning of this section, discussing coatequitl and then noting how the work of the youths and maidens involved singing and dancing while holding hands. As a result, the keeping of the ceremonial calendar by priests or others trained in calendrical and glyphic notation in elite religious schools (calmecac) constituted an important bureaucratic role for the organization of labor in Aztec society and likely also in other Mesoamerican societies that kept calendrical records through hieroglyphic notation or oral histories.

Famously lacking in the subsistence economies of Mesoamerica were large domesticated animals, an important point of contrast with other nuclear areas of early agriculture and urbanism in Afro-Eurasia and South America. This deficit had at least two important implications regarding social organization and resource management. The first is that large domesticated animals often drove increased inequality in early societies since, depending on the setting, they could represent “wealth on hoof” in and of themselves; they could be used for plowing fields and creating more extensive tracts of arable land; or they could be used in warfare or raiding (Reference BoixBoix 2015: 140–159; Reference GlowackiGlowacki 2020; Reference Kohler, Smith and BogaardKohler et al. 2017). In all cases, large domesticates had the potential to create and exacerbate disparities in wealth and power that were not available pathways to these ends for aspiring Mesoamerican elites. The second implication is that the management of hunted game species, fisheries, and concentrations of collected resources remained more important among Mesoamerican societies than was the case for many other early agriculturalists. Along these lines, Reference ParsonsJeffrey Parsons (2008), who spent decades studying the collecting of aquatic resources in the Basin of Mexico and the use of drought-tolerant plants, like cacti and succulents, proposed that these activities provided functional and nutritional equivalents of pastoral niches in Mesoamerica. The cultivation and processing of maguey and other agaves found in more arid regions fostered suprahousehold management and gender complementarity in labor (Reference Evans and BlandonEvans 2005). Fishing, hunting waterfowl, salt production, and the collection of lacustrine resources such as insects and algae did not require large groups, but they created communities with shared identities, networks, and patron deities and offered “opportunities for cooperation, possibilities for conflict, ecological change, and other unintended consequences” (Reference Millhauser, Nichols and Rodríguez-AlegríaMillhauser 2017: 311). In other words, these functional equivalents of pastoralism in Mesoamerica were structured by different dynamics of resource management than true pastoralism, creating more common-pool resource issues than are typical for privately owned herds of animals.

In addition to studies of subsistence economies, recent research on Aztec markets and exchange systems helps to distinguish what fiscal streams were more internal to a particular polity, what were more external or mixed strategies, and how these revenue streams were allocated to supporting a bureaucratic infrastructure and dissemination of public goods to varying degrees (Reference Blanton, Fargher, Babones and Chase-DunnBlanton and Fargher 2012; Reference HirthHirth 2016; Reference Nichols, Berdan and SmithNichols et al. 2017). Central Mexican economies were commercialized prior to the Aztec empire, and the strategies of Tenochtitlan and Texcoco in particular focused generally on encouraging preexisting market systems and trade networks, or incentivizing the creation of new ones, though they also intervened militarily in punitive ways in cases of political dissent from provincial centers. Exotic items originating outside of central Mexico and luxury items requiring skilled labor in their manufacture made their way disproportionately to Aztec elites through systems of imperial tribute and palace-based coatequitl, and certain forms of dress and adornment were restricted to nobility based on sumptuary laws, but a range of jewelry and other luxury items were also available in marketplaces for individuals who had the purchasing power to obtain them (Reference HirthHirth 2016: 52–57, table 3.1). This could include a “middle class” in Aztec society of non-nobles with achieved status through service in the military or in the political bureaucracy, skilled artisanship or long-distance trade, and as priests or other ritual specialists (Reference HicksHicks 1999). More widely available were “bulk luxuries” such as highly decorative polychrome pottery, made within central Mexico but requiring skilled manufacture, and valued commodities from outside of central Mexico, including cacao and cotton. These became more widely available to non-elites in the Late Postclassic than they had been during earlier periods of Mesoamerican history (Reference Blanton, Fargher, Heredia Espinoza and BlantonBlanton et al. 2005) and moved through complex commodity chains that saw fluid and overlapping relationships between intermediate or “mesoscale” social groups amendable to analysis using collective action frameworks (Reference Millhauser and OverholtzerMillhauser and Overhaltzer 2020).

Aztec tax systems in goods and labor, and questions relating to the financing of states and other political institutions, bring us to the issue of governance and its variable organization documented among cases from the Aztec world and other parts of Postclassic Mesoamerica.

2.3 Indigenous Institutions of Mesoamerican Governance

We are able to glean prehispanic Mesoamerican perspectives on what constituted good governance from passages such as the one above, which is a compilation of huehuehtlahtolli (“words of the ancients”) recorded by the friar Andrés de Olmos working with Nahuatl-speaking scribes in the 1530s, a little over a decade removed from the Spanish-Aztec War. When people use the term Aztec colloquially, they are often referring to the Mexica ethnic group and, more specifically, the Tenochca-Mexica who occupied the imperial capital of Tenochtitlan. This problematic shorthand muddles understanding because a broader definition of Aztec – for instance, the peoples of Postclassic central Mexico with a shared narrative of origins in Aztlan – includes tremendous diversity in forms of governance, equal to or greater than what existed in Europe at the time (Reference CarballoCarballo 2022) (Figure 7). Reference Zorita and KeenAlonso de Zorita (1963: 86), the sixteenth-century Spanish chronicler who recorded a wealth of information on land, labor, and governance in prehispanic Mexico noted, “it is impossible to state a general rule as concerns any part of Indian Government and customs, for there are great differences in almost every province.” The micro-patriotism and divergent views of governance within the Aztec world are nicely illustrated by a passage in book 12 of the Florentine Codex, where a scribe from the Mexica sister-city of Tlatelolco, occupying the same island as Tenochtitlan, evinces his low regard for Moctezuma II (or Xocoyotzin), the tlatoani of Tenochtitlan, compared with Itzquauhtzin, the tlatoani of Tlatelolco. In funeral services for both rulers during the Spanish-Aztec War, we are told that people cursed Moctezuma’s name, as he had been feared, and when his funeral pyre was lit the ruler’s body gave off a foul stench; Itzquauhtzin, in contrast, was grieved, wept for, and cremated with great splendor (Reference LockhartLockhart 1993: 150–152).

Figure 7 Map of western Mesoamerica, featuring selected sites and areas of political hegemony by Aztec (Triple Alliance) empire (shaded) and rival states (dotted lines). https://mediateca.inah.gob.mx/islandora_74/islandora/object/codice%3A635

Also emphasized by Nahua scribes in recording the oral histories of the elders is the centrality of the altepetl to political organization and a moral philosophy of rulership legitimated on the basis of historical precedent set by past governors said to have ruled justly and prudently. Although the term altepetl is often glossed in English as city-state, the metaphoric couplet of waters and mountains evokes a territorial realm or polity with the urban component of secondary importance. Historical records document a diverse array of altepetl organization, ranging from polities with a single ruler (Tenochtitlan) to others with as many as seven corulers (Cuauhtinchan) (Reference LockhartLockhart 1992: 14–58; Reference PaulinyiPaulinyi 1981). Confederated polities, such as Chalco and Tlaxcallan, were also a common arrangement and could have four or more corulers plus large and powerful governing councils featuring more pluralistic institutions of decision-making (Reference Gutiérrez Mendoza, Daneels and MendozaGutiérrez Mendoza 2012).

Even Tenochtitlan, the most powerful city in prehispanic Mesoamerican history, was part of a confederation between three altepemeh. Its paramount ruler, the “great speaker” (huey tlatoni), ruled in consultation with the rulers of those polities along with another powerful Tenochca (the cihuacoatl or “woman snake”) and a council of four dignitaries (Reference Pastrana FloresPastrana Flores 2020: 123–126). Later Tenochca rulers appear to have purposefully suppressed an earlier history of quadripartite corule beginning with the empire builder Itzcoatl (r. 1427–1440), who ordered earlier histories be destroyed to recast rulership in more autocratic terms, a centralization of power that culminated under Moctezuma II (Reference HassigHassig 2016: 56; Reference LockhartLockhart 1992: 25; Reference López AustinLópez Austin 1961: 20–52). Other documents reveal that for some fifty years following its founding Tenochtitlan had no single ruler and, as a subject polity to Colhuacan and then Azcapotzalco, was governed internally by the tlatoani of its barrios or wards (Moyotlan, Teopan, Atzacualco, and Cuepopan), who retained autonomous authority to organize calpollis to build and maintain infrastructure through the city’s transition to an imperial capital (Figure 8) (Reference Florescano, Nichols and Rodríguez-AlegríaFlorescano 2017: 100; Reference HassigHassig 2016: 26). As a multiethnic capital, Tenochtitlan’s political organization was deliberately divided into factions with different, complementary roles in the system:

Figure 8 Map reconstructing Tenochtitlan-Tlatelolco in the sixteenth century, with depictions of Tenochtitlan’s four wards or barrios, ceremonial precincts, chinampa zones, and other features of urban infrastructure.

Direct succession of rulership from parent to child, whether following primogeniture or not, was the exception rather than the rule in the Aztec world, with the imperial “second city” of Texcoco representing one of the notable exceptions (Reference Carrasco, Harvey and HannsCarrasco 1984; Reference OffnerOffner 2010). Succession was more typically collateral, meaning it was determined through election by ruling councils from a slate of candidates from certain noble families, and represented a more fluid system than was practiced by the Postclassic period Mixtec and Classic period Maya (Reference Carrasco, Harvey and HannsCarrasco 1984: 46; Reference HassigHassig 2016: 30–59; Reference MarcusMarcus 1992: 306–320; Reference van Zantwijkvan Zantwijk 1985: 25, 178–179). During the Postclassic period, succession within Mixtec kingdoms most closely adhered to rules of primogeniture and a concern with maintaining royal bloodlines, which were legitimated materially through art emphasizing royal genealogy (Figure 9), the focal role of palaces within urban centers, and what were perhaps the most elaborate tombs and personal ornamentation from prehispanic Mesoamerica (Reference Caso and PaddockCaso 1966; Reference Spres and BalkanskySpores and Balkansky 2013: 89–103). Mixtec codices foreground royal genealogies and the exploits of certain kings and queens yet also indicate the presence of small councils that served political and religious functions (Reference Jansen and Aurora Pérez JiménezJansen and Pérez Jiménez 2011: 357–461; Reference PohlPohl 1994). Aztec and other central Mexican codices, in turn, more frequently depict cosmological themes relating to the ritual calendar (critical to coordinating communal labor), migration narratives of ethnogenesis, and tax records, though individual rulers were depicted in postconquest codices and in a few instances of prehispanic sculpture, particularly under the more consolidated rule of Moctezuma II (Reference HajovskyHajovsky 2015).

Figure 9 Scene from prehispanic Mixtec codex (Codex Vindobonensis) depicting noble genealogy and founding narrative of emergence from sacred tree.

Austrian National Library, Vienna.

Although the Mixtec kingdoms were in close contact with the peoples of Tlaxcallan, current Tlaxcala, these major rivals to the Mexica-Aztec developed a very different system of governance, likely in a process analogous to schismogenesis as a form of resistance to the imperial ambitions of the Triple Alliance. Tlaxcallan’s governance structure was highly pluralistic, with Reference Cortés and PagdenCortés (1986: 68) equating it to the northern Italian republics he knew from Europe. Tlaxcaltec political organization featured a distribution of executive power through rotating officeholders, a governing council that probably numbered in the hundreds of members and included some level of participation on the part of non-elites, and no clear cases of residences that could have served as central palaces (Reference Fargher, Blanton and Heredia EspinozaFargher et al. 2010, Reference Fargher, Blanton and Heredia Espinoza2011, Reference Fargher, Blanton and Heredia Espinoza2022; Reference López Corral, Stanton, Taube, Coltman and Marengo CamachoLópez Corral 2023). Aspirants to high office, the teuhctli, underwent elaborate ceremonies of investiture involving acts of penance, self-sacrifice, and fasting that could last up to a year (Reference ReyesReyes 2020). During the early colonial period, the Tlaxcaltecs presented themselves to the Spanish Crown as a confederation of four nearby city-states: Tizatlan, Ocotolulco, Tepeticpac, and Quiahuiztlan (Figure 10). Although conquistadors singled out Xicotencatl the elder and Maxixcatzin, representing the first two of these divisions, leading figures of other important factions in the governance structure are also named and the four-part confederation seems a colonial period reconfiguration of what were previously wards or districts of a single urban agglomeration (Reference LockhartLockhart 1992: 20–23). The city was organized as a polycentric yet interconnected network of residential terraces and equitably distributed social infrastructure of plazas, small shrines, and altars, with a disembedded ritual precinct not connected to any one faction (Reference Fargher, Blanton and Heredia EspinozaFargher et al. 2011, Reference Fargher, Blanton and Heredia Espinoza2022). Governance in Tlaxcallan can be said to have featured democratic institutions and was also inclusive in terms of the definition of citizenship, with ethnic groups other than the dominant Nahuas, such as the Otomi and Pinome, afforded that status (Reference Fargher, Blanton and Heredia EspinozaFargher et al. 2022: 124).

Figure 10 Scene from colonial period Lienzo de Tlaxcala, depicting ruling figures and houses represented in the council governance of (a) Tizatlan, (b) Ocotolulco, (c) Tepeticpac, and (d) Quiahuiztlan.

To the south of Tlaxcallan, Cholula (former Cholollan), was one of the most prominent market and pilgrimage centers of Postclassic Mesoamerica. Its inhabitants turned from being allies of the Tlaxcaltecs to becoming largely foes once Cholula was incorporated into the Triple Alliance empire. Documentary evidence indicates that Cholula’s leadership structure was somewhere between Tlaxcallan and Tenochtitlan in terms of its pluralism as it featured two high priests who coruled the polity and were charged with external affairs, the Tlachiach and Aquiach, another ruler charged with internal affairs, the Chichimecatl teuctli, and a noble council depicted in the map of the city in the Historia Tolteca Chichimeca (Figure 11) as the “turquoise house” (Reference Lind, Daneels and MendozaLind 2012; Reference López Corral, Stanton, Taube, Coltman and Marengo CamachoLópez Corral 2023; Reference Plunket and UruñuelaPlunket and Uruñuela 2018). The six councillors of the turquoise house acted as legislators and judges and were responsible for electing the three rulers in consultation with a larger governing council. Cholula’s more oligarchic governance, relative to Tlaxcallan, is apparent in the fact that the Tlachiach and Aquiach could only be selected from among the nobility of a single, highly ranked calpolli (Reference Carrasco, Ekholm and BernalCarrasco 1971: 372).

Figure 11 Depiction of the city of Cholula from the Historia Tolteca Chichimeca with temple to Quetzalcoatl at center left, calmecac at lower right, calpoltin in rectangles along border, and rulership structure with the corulers (a) Tlachiach and (b) Aquiach, ruler of internal affairs the (c) Chichimecatl teuctli, and (d) noble council in the “turquoise house.”

Base images from Wikimedia (https://commons.wikimedia.org/) and Bibliothèque nationale de France.

The complex altepetl of Chalco was detailed by the prolific Nahua chronicler Chimalpahin and its political organization had several points of overlap with Tlaxcallan’s, though its four altepemeh were more geographically dispersed and the polity’s rules of succession and governing council were more restrictive. Chimalpahin’s chronicles list some thirteen royal titles for the four altepemeh, each of which oversaw four to six suprahousehold groups, designated as calpolli, chinamitl, tlaxilacalli, and other terms listed in Table 2 and discussed in subsequent sections (Reference SchroederSchroeder 1991: 109–113). In succession patterns, Chalco’s governance structure might have been more dynastic or lineage based, akin to Mixtec polities, yet the “Chalca reveal considerably more flexibility, allowing the succession of commoners, the exchange of rulerships, and atypical marriage arrangements” (Reference SchroederSchroeder 1991: 220). Interestingly, both the Chalca and the Mixtec featured more queens as paramount rulers than did more elective or pluralistic systems since maintenance of lineage ties was of greater concern and male-dominated councils tended to elect other men as successors to high office, a tendency seen in the kingdoms of early modern Iberia as well (Reference CarballoCarballo 2020: 116).

Through this selected set of cases from the Postclassic period, we can appreciate some of the temporal changes in political organization and governance, such as is documented for Tenochtitlan, and the regional variability in the same observed between more dynastically oriented Mixtec kingdoms and more pluralistic Tlaxcallan, or elsewhere in Mesoamerica as well. A notable example from the Maya world is the operation of corule or council rule (multepal) at Mayapan and possibly other Postclassic capitals of the northern Yucatan (Reference Masson and LopeMasson and Peraza Lope 2014). The diversity of sociopolitical arrangements underscores the need to consider the same fluidity in earlier periods, which we will do in Section 3. Nevertheless, much of the literature on prehispanic Mesoamerican politics fixates on rulers and carries implicit assumptions of top-down leadership structures, while our aim is to consider political relations and governance more broadly so as to include multiple levels of interaction on the part of various social groups, bureaucratic roles, and the public goods and expectations placed on leaders in different polities. Some terms in Nahuatl that provide actual Mesoamerican examples appear in the infrastructure and governance sections of Table 2, while Figure 12 provides a schematic model for considering governance and collective action within Aztec polities. The model proposes that societies with more internal resource bases as fiscal streams are positively correlated with greater bureaucratization and dissemination of public goods (Reference Blanton and FargherBlanton and Fargher 2008: 250–254) as members of state bureaucracies assume roles both in mobilizing internal resources via tax regimes and in organizing public labor. These variables also are positively correlated with checks on governing authorities, who relinquish elements of their decision-making in a bargain with subjects to, following Reference LeviLevi (1996: 10–16), (1) make credible commitments, implicating public goods; (2) uphold generally agreed upon fair procedures, through bureaucratic roles and checks on rulers; and (3) foster reciprocal sentiments of trust that sustain the general compliance of taxpayers who finance the system. The text adjacent to these key variables to collective action provides examples from the Aztec world, many of which have analogs for other culture regions and time periods in Mesoamerica.

Figure 12 Schematic text-and-arrow model of collective action with examples of public goods, internal revenues financing government, and types of accountability of leaders provided from Aztec states drawing on Reference Blanton and FargherBlanton and Fargher (2008: 251–254) and sources cited in Table 2.

Arrow width reflects correlation strength between the variables in the authors’ sample of thirty societies including Aztec but is calibrated for all cases.

In Reference Blanton and FargherBlanton and Fargher’s (2008) comparative study of thirty premodern states, the Aztec empire scored as tied with the Mughal empire as the sixth most collective and as in the top quarter of cases in terms of its dissemination of public goods, degree of bureaucratization, and checks on authorities. Yet, as we saw in the comparative cases discussed earlier, the Tenochca-Mexica were by no means the most pluralistic governance system of Postclassic Mesoamerica and its rulers had increasingly consolidated power during the imperial period, particularly under Moctezuma II. Based on textual accounts of the governance structure, Tlaxcallan was much more pluralistic, but it is useful to engage archaeological evidence pertaining to the built environment and relative levels of access to resources on the part of a broader populace. For instance, a comparison of Gini coefficients calculated based on the amount of residential living area scores Tenochtitlan (0.30) higher than Tlaxcallan (0.23) as we would expect, though both are considerably lower than the contemporary United States (0.49) and many premodern Eurasian state capitals (Reference Fargher, Blanton and Heredia EspinozaFargher et al. 2022; Reference Kohler, Smith and BogaardKohler et al. 2017). How much a differential economic base, lacking in large domesticated animals, utilitarian metals, and with fewer opportunities for elites to monopolize spot resources, relates to these values of more equitable living space is an important question that requires further study.

Although later Tenochca-Mexica rulers lived in large and lavish palaces, depicted themselves in state art on equal footing with gods, and conveyed their hegemony through ceremonies involving human sacrifice, they also had limits on their power and were expected to distribute public goods in the form of providing food stores in times of famine, constructing hydraulic projects such as aqueducts to bring fresh water in and dikes to keep flooding at bay, and expand the city’s agricultural base through major canal systems that could take several years to construct (Reference BerdanBerdan 2014: 76–81; Reference Rojas Rabiela, Ruiz and Murillo LiceaRojas Rabiela et al. 2009). A nice encapsulation of this comes from Chapultepec, located on the mainland to the west of Tenochtitlan. Here, Mexica sovereigns had their pleasure palaces and gardens but also oversaw projects to channel the hill’s freshwater springs into the city via aqueducts – a pattern continuing into the viceroyalty of New Spain and representing a semipublic colonial park almost a century earlier than Boston Common (Reference Cervantes de SalazarCervantes de Salazar 1554/2014: 131). By bureaucratizing economic roles associated with markets and trade networks, the Mexica political economy also fostered broad sentiments of fairness in exchange by establishing principles of judicial neutrality in marketplaces such as Tlatelolco and in the entrepreneurial agency permitted to the pochteca and other merchant and artisan groups (Reference Blanton and FargherBlanton and Fargher 2016: 96). State judicial and military oversight of fair and free trade were therefore public goods but paid for by taxes in kind and on markets and labor, which included military conscription. Irrespective of how Aztec rulers were promoting themselves, then, Aztec governance falls on the more collective or pluralistic end of the spectrum when we consider broader relationships between economic systems, bureaucracies, and public goods. Governance that we identify as more autocratic, in contrast, tends to be transactional to revolve around exclusionary networks of political or social elites.

A key axis of variability well documented in Aztec institutions underlay the tensions between forms of social organization based on hierarchical class relations, between nobles and commoners or patrons and clients, and among corporate groups with greater rank or status parity internally and heterarchical relations externally. In Nahuatl terminology, the tension was foremost between hierarchical patrimonial noble estates (the teccalli) and heterarchical corporate groups connected by kin and affinal ties (the calpolli and tlaxilacalli) or by economic bonds arising from shared occupation and market interests (the pochteca and certain high-status artisan groups). These institutions can likely be generalized to other Mesoamerican states, though we must be cautions as terms were used differently even within Postclassic central Mexico, with notable distinctions in terms between eastern and western Nahuas. For instance, subdivisions of the altepetl in eastern Nahua polities such as Tlaxcallan and Cuauhtinchan were often termed teccalli, while in western Nahua polities such as Chalco they were often termed tlayacatl, though particularly in the case of Tlaxcallan the teccalli subdivision was much less hierarchical than how the term was used to the west and allowed for achieved noble status (Reference Fargher, Blanton and Heredia EspinozaFargher et al. 2022; Reference SchroederSchroeder 1991: 215–216). Further, in some cases the standard of living for a non-elite may have been higher as a subject to an affluent noble in a teccalli structure than as a more autonomous member of a calpolli that possessed little land or social status. Still, Reference LockhartLockhart (1992: 98) rightfully emphasized the teccalli–calpolli tension as critical for understanding variability in precolonial Nahua sociopolitical organization, noting: “Teccalli and calpolli competed, threatening in principle to swallow each other up … Where the noble and patrimonial teccalli won out, the majority of the population might be special dependents … where the ethnic calpolli was predominant, the majority would be ordinary macehualtin [commoners].” A generalized notion of these two institutions from the Aztec world, including their fuzzy boundaries and tensions as hierarchical versus heterarchical bonds of affiliation, provides Indigenous templates grounded in historical texts for evaluating variability in the archaeological record of periods lacking such textual documentation.

2.4 Summary

By reframing the study of Mesoamerica based on key institutions and their articulation with one another, we allow for comparative analyses that consider multiple scales of human action, not simply views from the top of elite culture and governance structures. Scales intermediate between individual households and polities are particularly important for generating more nuanced social histories of the Mesoamerican past. The tension between the Nahua institutions of patron–client-type noble houses, such as the teccalli, and corporate-kin groups such as the calpolli is not simply a point of semantic debate, it illustrates a socio-structural conflict between more hierarchical and more heterarchical cultural institutions that affected the management of land and other key resources underlying the economies of Mesoamerican societies. It therefore connects the collective action frameworks relating to more local-level labor and resource issues covered by researchers such as Ostrom with the frameworks relating to polity-level fiscal and governance systems covered by researchers such as Levi. Somewhat like with the tensions between gumsa and gumlao within Kachin society, teccalli-like and calpolli-like social structures provide us with Indigenous institutions from Mesoamerica for considering variability in collective action in resource management and governance. With these better-documented institutions in mind, including their variability, we will now turn to earlier periods in Mesoamerican history, which have far fewer texts but offer millennia-long archaeological and iconographic records for evaluating known institutions through material remains.

3.1 Chronological Trends in Resource Management and Economies

Mesoamerica’s transformation from being occupied mostly by mobile or seasonally mobile foraging populations to being occupied mostly by sedentary farmers residing in permanent villages unfolded over millennia of the Archaic period (ca. 8000–2000 BCE), extending into the Early Formative/Preclassic (ca. 2000–900/800 BCE) in several regions (Figure 14). The tempo was irregular and change was not unidirectional. The more ephemeral nature of traces of occupation from this period and a long-standing research bias to more monumental centers dating from later periods make this early record spotty, but we can glean ways that people came together and shared in resource exploitation and labor from several sites. For instance, in the semiarid Valley of Oaxaca during the Middle Archaic period, sites such as Gheo-Shih attest to macroband gatherings of a few dozen people who camped together during the rainy season while collecting plant resources, perhaps sharing food, and demarcating “plaza” or “dance platform” space on the order of 140 m² bounded by parallel lines of boulders (Reference Joyce and FlanneryMarcus and Flannery 1996: 52–59; see also Reference Marcus, Alarcón, Lohse, Borejsza and JoyceWinter and Alarcón 2021: 306–308). In Archaic coastal settings featuring abundant marine resources, human intentionality in the formation of large shell mounds and the degree to which they were residentially occupied are debated points, with Reference Clark, Hodgson, Lohse, Borejsza and JoyceClark and Hodgson (2021) proposing a more sedentary and intentional occupation for those located on the Soconusco coast. If accurate, these would represent pre-agricultural villages, some measuring up to 3.3 ha (33,000 m²), with hypothesized houses along the edges of mounds that clustered around open community plazas. This proposal is admittedly speculative and in need of more investigation but, if one day confirmed, could represent cases of community cooperation in exploitation of a common-pool resource while simultaneously signaling to other groups, through the mass of mounded shell, their territorial claim to particularly productive mollusk beds.

Figure 14 Map with sites used in comparative analyses with others mentioned in the text.

With the development of farming based on maize and other crops, intentionality in communal labor becomes more archaeologically visible since it resulted in the landesque capital of hydraulic works, field systems, terrace networks, and other anthropogenic landscapes of subsistence infrastructure. Mesoamerica is an area of high climatological and microclimatic variability, but to generalize, farming in the semiarid highlands typically benefits from moving water to fields and terracing slopes to retain soil while farming in the humid lowlands typically benefits from moving water away from fields and houses during periods of inundation and diverting it to reservoirs. Reservoirs were also critical to communities in the southern Maya lowlands as sources of potable water during the dry season, since its karstic landscape means that groundwater is deeply buried. Wells were of limited utility except in select places with less porous substrates (Reference AshmoreAshmore 1984; Reference Lentz, Dunning and ScarboroughLentz et al. 2015). The northern lowlands of the Yucatan Peninsula diverge from Mesoamerica’s general highland/lowland ecologies in being semiarid and lacking rivers, so settlements clustered around subterranean water sources in the form of sinkholes (cenotes) and depressions (rejolladas). Groundwater levels are generally higher in highland settings, allowing potable water to be tapped through wells (Reference Sanders, Parsons and SantleySanders et al. 1979: 221–293).

Understanding the socioeconomic relations that underlay any given case of subsistence infrastructure typically requires careful geoarchaeological research to establish their formation processes and chronology as well as accurate mapping of how people were distributed across the landscape and how households of varying statuses may have had disproportionate access to the resources. In areas that were riskier for crop failure, people were generally more incentivized to work together and buffer against risk, though relatively lush areas also offered potential gains to households that worked together in intensifying agricultural practices and it is clear that risk alone does not explain the distribution of subsistence infrastructure. As noted in Section 2, the lack of large domesticated animals in Mesoamerica meant that there were no true pastoral economies or plow-agriculture extensification of fields as drivers of differential wealth/land accumulation. Hunting and collecting activities remained relatively more important to Mesoamerican subsistence compared to early urban societies in Afro-Eurasia, providing other types of common-pool resources that were not privatized.

Starting in the lowlands, the expression of precocious art and early urbanism termed Olmec had a core in the humid southern Gulf Coast along seasonally inundated rivers such as the Coatzacoalcos, home to the monumental Early Preclassic center of San Lorenzo. To avoid flooding, the inhabitants of San Lorenzo created a patchwork of forty-seven elevated (1–5 m) residential mounds termed islotes that would have required concerted effort to build (Reference Cyphers, Di Castro, Fash and LujánCyphers and Di Castro 2009). Mound construction was also directed at public spaces in San Lorenzo and subsequent La Venta, and both places exhibit a discernable sociopolitical hierarchy that emphasized individualized rule and elite differentiation through access to spot resources such as jade and iron ore, both procured through exchange from hundreds of kilometers away. Basaltic stone, used for sculptures expressing religious concepts and human power relations, was absent at these Olmec capitals and likewise needed to be imported. At San Lorenzo, more than 30 metric tons of stone was also sculpted to construct an elaborate covered drain or aqueduct running 170 m east–west (Reference Coe and DiehlCoe and Diehl 1980: 118–127, fig. 83). The relatively short but elaborately built nature of the drain, its association with sculptures interpreted as water deities and aquatic fauna, and its association with small pools in central San Lorenzo are consistent with a ceremonial rather than a subsistence function, or as a means of providing potable water, and one that likely bolstered the power of San Lorenzo’s ruling elite (Reference Coe and DiehlCoe and Diehl 1980: 393). Contemporaneous with these Olmec sites, Middle Preclassic Maya communities began building earthen architecture on an equally monumental scale, with sites such as Aguada Fénix requiring an estimated 10–13 million person-days to construct (Reference Inomata, Triadan and Vázquez LópezInomata et al. 2020). Interestingly, Aguada Fénix and associated sites featuring colossal platforms and reservoir complexes were built by people who may have still retained patterns of seasonal mobility and did not exhibit notions of individualized rule in their art or as marked sociopolitical distinctions as did Gulf Olmec centers.

The Maya region is large and home to a range of strategies for subsistence infrastructure, including terraces in hilly regions like the Maya Mountains of Belize, low-lying bajos in Guatemala’s Peten district, irrigation in certain areas often as a means of creating reservoirs, and raised or drained fields in swampy areas (Reference Fedick, Morell-Hart and DussolFedick et al. 2023; Reference Lentz, Dunning, Sheets, Larmon, Lucero and ValdezLentz et al. 2022; Reference PalkaPalka 2023; Reference Scarborough, Dunning and TankersleyScarborough et al. 2012). The latter have been studied in detail in Pulltrouser Swamp (Figure 15), where raised-field construction was underway by the Late Preclassic and increased through the Late/Terminal Classic, eventually covering between 331 ha and 668 ha and likely allowing year-round cultivation (Reference Turner, Harrison, Turner and HarrisonTurner and Harrison 2000: 247). By excavating canals on the mainland adjacent to the swamp, the Maya created channelized fields in depressed areas. Using the low estimate for total area, Reference Turner, Harrison, Turner and HarrisonTurner and Harrison (2000: table 13-1) still extrapolate construction rates in the low thousands of days per hectare of the system, meaning a process that would have taken many years even if hundreds of workers participated in the undertaking. Household archaeology from the site of K’axob, adjacent to the raised fields at Pulltrouser Swamp, indicates that larger, more corporate households had longer occupations and more diversified economies, but bone-isotope analyses do not support the notion that local elites managed staple crop production and instead suggest this was managed by corporate-kin groups (Reference HendersonHenderson 2003). Other Maya raised-field and reservoir systems have been mapped using LiDAR (light detection and ranging) (Reference Beach, Luzzadder-Beach and KrauseBeach et al. 2019; Reference Hansen, Morales-Aguilar and ThompsonHansen et al. 2022; Reference Hutson, Dunning and CookHutson et al. 2021) and are amenable to future research on how hydraulic infrastructure was organized, whether as common-pool resources or as more privatized, toll-goods systems benefiting an elite class. In cases of the large Classic period Maya cities, the reservoir systems of centers such as Tikal may have been elite-managed and incorporated into rituals of royal aggrandizement, while more decentralized systems, such as with terraces at Caracol and other sites, afforded opportunities for neighborhood-level collective action and more bottom-up management processes (Reference Lentz, Dunning, Sheets, Larmon, Lucero and ValdezLentz et al. 2022).

Figure 15 Map of raised-field system in Pulltrouser Swamp, Belize, with associated Maya settlements.

Map courtesy of Marcelo Canuto.

In addition to diverting water from inundated areas to create dry land, bringing water to crops, and storing water for drinking, lowland Maya waterscapes also included the purposeful cultivation of freshwater fisheries containing species of fish, turtles, crustaceans, and mollusks (Reference PalkaPalka 2023). If managed sustainably, such species would have provided dependable protein sources, as well as field fertilizer, in a similar fashion to the lacustrine environments of highland Mesoamerica and may have similarly been managed as common-pool resources by social groups intermediate between households and states. The maintenance of aquatic landscapes, both natural and anthropogenic, was therefore a key collective action issue for lowland Mesoamerican populations. In some cases, their contamination may have been observable by ancestral Maya populations and adapted to, as appears to have been the case with the creation of sand filtration systems in certain reservoirs (Reference Lentz, Dunning and ScarboroughLentz et al. 2015, Reference Lentz, Dunning, Sheets, Larmon, Lucero and Valdez2022; Reference Scarborough, Dunning and TankersleyScarborough et al. 2012). In others, such as with sedimentation processes resulting from deforestation and urban-waste runoff near lowland lakes, the centuries-long unfolding of water contamination may have been difficult to notice or adapt to, leading to population decline (Reference Birkett, Obrist-Farner and RiceBirkett et al. 2023).

Management of water and soil resources in the Mesoamerican highlands differed from that of the lowlands in several ways, including the construction of more terrace networks to flatten land on hillier terrain, more canal irrigation systems to deal with semiarid precipitation regimes, fewer reservoirs for potable water since groundwater could typically be tapped through wells, and less mounding of earth into dry land since inundation is less of a concern. These nevertheless reflect general trends and it is clear that highlanders and lowlanders were both familiar with this range of subsistence infrastructure that they could adapt to a particular ecological setting when appropriate. In the highlands, the construction of some of these forms of subsistence infrastructure began during the Archaic period, but they became more pervasive during the Middle Formative (900–500 BCE) associated with agricultural intensification and the rise of nascent urban centers.

A form of land-making analogous to what was practiced in the lowlands can be seen in the tlateles constructed along lake margins of the Basin of Mexico (Reference CordovaCordova 2022; Reference Cordova, Morett-Alatorre, Frederick and Gámez-EternodCordova et al. 2022). Unlike with the longer and more linear chinampas, tlateles functioned as residential platforms and were therefore constructed with more proportionally sized square or rectangular sides, though the two types of land-making shared similarities in using palisaded enclosures of wooden stakes and lattice to retain sediments (Reference CordovaCordova 2022: 195–214). Tlateles ranged in area from tens of square meters – suitable for a dwelling or two and necessitating only household labor – to hundreds of square meters, similar in scale to the islotes of Gulf Olmec settlements and necessitating suprahousehold labor. In both cases, however, a primary driver for tlatel construction was lake aquaculture, including the harvesting of common-pool resources such as fish, algae, salt, aquatic birds, and other lake fauna.

Irrigation networks began with Archaic maize agriculture in the highlands but proliferated during the Middle Formative (Reference DoolittleDoolittle 1990; Reference Nichols, Frederick, Morett Alatorre, Sánchez Martínez, Lucero and FashNichols et al. 2006). These could be fed by springs that provided water throughout the year in select locations or by branching off canals from rivers and seasonal streams near lakeshores or seasonal wetlands termed humedales, as was the case in the ethnographic example from southern Tlaxcala documented by Reference WilkenWilken (1968). In this same area millennia earlier, drained-field systems were constructed by the inhabitants of the large regional center of Xochitecatl at the confluence of the two major rivers of the Puebla-Tlaxcala Valley (Serra Reference Puche, Carmen, Carlos Lazcano Arce, de la Torre Mendoza and ManzanaresPuche et al. 2004). An example of a modest, community-sized system was documented at Amalucan, today on the outskirts of the contemporary city of Puebla (Reference FowlerFowler 1987). This Middle to Late Formative network of canals was fed by springs and seasonal streams (Figure 16). The settlement grew to be a large town of the period, covering some 10 km², but began more modestly, with an irrigation system providing water to an estimated 1,000 ha of fields and later temple construction covering some of these earlier canals – a stratigraphic indication that the estimated 80,000 m³ of earth moved for excavating the canals was accomplished prior to the development of a more hierarchical political-religious system. The inhabitants of the larger cities of Cuicuilco and Teotihuacan created more extensive irrigation and reservoir networks drawing on nearby springs (Figure 17) and, particularly in the case of Teotihuacan, situated in the more arid northern Basin of Mexico, dug canals on the slopes of nearby mountains to capture rain water and humidify fields prior to planting once rains became more consistent (Reference Evans, Nichols, Gonlin and FrenchEvans and Nichols 2015; Reference Sanders, Parsons and SantleySanders et al. 1979: 230–260).

Figure 16 The Formative town of Amalucan, with its irrigation network, ceremonial core of mounds, and isolated mound on Cerro Amalucan. Map based on Reference FowlerFowler (1987) overlay on 2001 satellite image from Google Earth. Today the site is almost completely covered by Puebla’s urban sprawl.

Figure 17 Map of Teotihuacan superimposed over a false-color composite LANDSAT satellite image that highlights currently vegetated areas southwest of the city at the confluence of the San Juan and San Lorenzo rivers and where springs provide water for permanent irrigation.

These Formative and Classic period hydraulic systems have their origins in strategies developed earlier in the more arid Tehuacán Valley of southern Puebla, where irrigation networks were created in tandem with maize cultivation during the Archaic period. Reference Neely, Aiuvalasit and WinsboroughNeely and colleagues (2022) draw on the work of Ostrom (e.g., Reference Ostrom1990) in documenting five prehispanic irrigation networks in Tehuacán that were all substantially larger than Amalucan’s, ranging from 825 ha to 3,852 ha of estimated irrigation farmland, and all constructed by smaller-scale communities that do not exhibit signs of any coercive authority to organize the labor. Yet the most impressive hydraulic project from Tehuacán was the construction of the Purrón Dam Complex during the Early and Middle Formative period as a means of creating a reservoir to feed the irrigation systems. At its largest extent during the Middle Formative, the complex comprised some 390,000 m³ of construction volume, creating an estimated reservoir volume of nearly a million cubic meters of water (Reference Neely, Aiuvalasit and ClauseNeely et al. 2015) (Figure 18). It is therefore centuries earlier and larger than the dam and reservoir systems at Maya cities such as El Mirador, extrapolated thus far based on LiDAR mapping (Reference Hansen, Morales-Aguilar and ThompsonHansen et al. 2022), and Tikal, which have been documented through excavations and geoarchaeological investigations as well (Reference Scarborough, Dunning and TankersleyScarborough et al. 2012). The construction volume of the Purrón Dam Complex is also significantly larger than the volume of Teotihuacan’s Moon Pyramid or any of the largest pyramidal temple platforms of Formative period central Mexico (Reference CarballoCarballo 2016; Reference MurakamiMurakami 2015). Hydraulic projects from the Tehuacán Valley should therefore serve as an example to global archaeology of the possibilities for large-scale collective action on the part of small-scale societies lacking the trappings of significant sociopolitical hierarchy such as elaborate palaces or elite tombs. We hope it serves as a nail in the coffin to facile notions of monumental constructions requiring state or other hierarchical oversight.

Figure 18 To-scale renderings of the Purrón Dam Complex, Moon Pyramid at Teotihuacan, and El Mirador (La Jarilla) reservoir. Redrawn from Reference Neely, Aiuvalasit and ClauseNeely et al. (2015: fig. 4), Reference Sugiyama and López LujánSugiyama and López Luján (2007: fig. 1), and Reference Hansen, Morales-Aguilar and ThompsonHansen et al (2022: fig. 21).

Unfilled lines for Purrón and Mirador indicate the water capacity of reservoirs. Note that the largest construction of the Purrón Dam stretches some 400 m wide, not visible in cross-section, whereas the base of the Moon Pyramid is square.

Irrigation systems were also constructed farther south in the southern highlands of Oaxaca, but here we will instead emphasize the region’s terrace networks, which were built on slopes to flatten living surfaces, to retain soil and moisture for plant cultivation, or some combination of the two. In Oaxaca’s smaller valley systems, Zapotec and Mixtec communities developed a characteristic mode of urbanism of terraced hilltowns, which can be defined as a coherent “type” of Mesoamerican settlement that was especially enduring in the region (Reference Kowalewski, Feinman, Nichols, Heredia, Durrenberger and MartíKowalewski et al. 2006). Frontal terrace walls were generally shared by multiple residential units and drains often were situated in the narrow spaces between individual terraces, requiring neighboring families to coordinate labor for their construction and maintenance. At the base of residential sectors, networks of interconnected terraces known as lama-bordo are cross-channel irrigation systems and provide a clear example of a common-pool resource that require sustained cooperation at the suprahousehold level, as dereliction in maintenance on the part of one family would have negative consequences on others, particularly associated with downslope erosion (Reference Pérez RodríguezPérez Rodríguez 2016).

Serving as the Zapotec capital city for some 1,250 years, Monte Albán was the most prominent and longest lived Oaxacan hilltown, eventually being enmeshed in a network of more than 2,000 terraces along with fortification walls that protected certain slopes (Figure 19) (Reference Blanton, Feinman, Kowalewski and NicholasBlanton et al. 2022; Reference Joyce and FlanneryMarcus and Flannery 1996: 139–154; Nicholas and Reference FeinmanFeinman 2022). During the initial mapping of the city, a range of civic hydraulic features were documented that include drains, dams, canals, and pools (Reference O’Brien, Lewarch, Mason and NeelyO’Brien et al. 1980). More recent investigations have documented the presence of a spring located at the bottom of the northeast slope of Monte Albán, approximately 1 km from the urban epicenter and emanating from the hill itself. Water was channelized into a series of canals and pools, and the flow of water from the hilltown may have given Monte Albán a symbolic identification with a water-mountain (Reference Martínez GracidaMartínez Gracida 2017), like many other Mesoamerican cities and urban monuments.

Figure 19 Aerial photo of the long-lived hilltown and Zapotec capital city of Monte Albán.

In addition to the world of subsistence resources, visible archaeologically through various types of infrastructure or landesque capital, was the Mesoamerican world of goods that circulated through various networks of patronage and commercialized exchange and is generally visible through the record of portable artifacts. Reference Blanton, Fargher, Heredia Espinoza and BlantonBlanton, Fargher, and Heredia (2005; Reference Blanton, Fargher, Babones and Chase-DunnBlanton and Fargher 2012) have outlined a useful classification for major types of goods within Mesoamerica and other premodern economies: (1) regional goods, the backbone of any system of economic exchange, tending to be widely available to non-elites and more “utilitarian” in nature; (2) prestige goods, tending to be exclusive to elites, more “symbolic” in nature, and made from rare materials and/or with labor-added value; (3) bulk luxury goods, existing somewhere between regional and prestige goods in being rarer or more elaborate but also available to non-elites through greater commercialization of economies and more robust market circulation. In very broad terms, the three millennia of precolonial Mesoamerican history, from its first villages to the Spanish invasion, can be characterized by three major episodes in the relative import of these types of goods. The Early–Middle Formative Olmec interaction sphere saw a prioritization of prestige goods as status markers during processes of developing sociopolitical hierarchies. The Late Formative through Classic periods was marked by an intensification of regional goods production and circulation, as urban centers flourished in most parts of Mesoamerica, creating more scalar economies, tax/tribute structures, and early markets. In some cases, like with Thin Orange pottery at Teotihuacan, forms of bulk luxuries circulated widely, but this intensified during the Postclassic period with the intensification of markets, professional merchant classes, and political confederations like the Aztec Triple Alliance promoting relatively open exchange across political boundaries (Reference HirthHirth 2016: 188–236).

Major transformations and axes of variation in Mesoamerican economies can then be connected with collective action frameworks and known cultural institutions by considering what types of production and circulation activities would have been more exclusionary, concentrated disproportionately in the hands of social elites, versus those that would have been more inclusive, with greater accessibility to finished products on the part of non-elites and potential profits to be made on the part of intermediate-scale social groups such as cooperating households or guild-like corporate economic groups. The intensification of regional goods economies during the Late Formative to Classic, such as the wider circulation of mass-produced pottery and obsidian blades, and of bulk luxuries during the Postclassic, such as the wide availability of polychrome pottery, cacao, and cotton textiles, engendered new socioeconomic relations and networks among Mesoamericans (Reference Blanton, Fargher, Heredia Espinoza and BlantonBlanton et al. 2005; Reference Feinman, Nicholas, Golitko, Feinman and RiebeFeinman et al. 2022). Much of the resource management of these craft production systems and the distribution of goods through corporate trading groups who operated semi-autonomously from polities (e.g., the pochteca) would be classed as toll goods in schema like Figure 2, generally of lower analytical interest since their management dynamics are more privatized and therefore relatively straightforward. However, the political maintenance of the market systems and open trade routes that circulated these goods could be viewed as a public good to consumers, with a social contract of urban/state taxation to keep the flow of goods moving, like the schema in Figure 12. Together with the subsistence economy and its associated infrastructure, the variable emphases in types of goods formed the foundations of political economies and the fiscal streams of governance. These foundations had both material and ideological components.

3.2 Governance and Ideologies of Inclusion and Exclusion

The interactions between different nested scales of subsistence and economic resources reviewed in previous sections, the definition of users permitted to harness and consume them, and the cultural institutions and governing structures involved in their management created a diversity of urban and political arrangements in early Mesoamerica. We can draw from the textually documented institutions from later periods, covered in Section 2, to build interpretive frameworks that allow for consideration of a similar range of diversity both regionally and temporally. Our dataset here includes variables such as how individuals arranged themselves residentially, coordinated group action at intermediate scales between households and entire settlements, worked together to create shared infrastructure and other undertakings, organized their urban centers, and expressed difference in wealth, power, and other aspects of social identity. These choices and actions were enmeshed within cognitive systems of religious belief and political ideology, which shared some core elements across Mesoamerica but also varied over space and time.

Returning to Reference KirchhoffKirchoff’s (1943) early formulation of Mesoamerica as a culture area, calendrical systems provided broadly shared means of parsing time and synchronizing group activities, including ritual cycles and organizing communal labor. The Calendar Round pairing of a 365-day solar year and 260-day ritual year is a culturally distinctive attribute of Mesoamerican cultures that appears to have had its origins at the same time as the coordinating of hydraulic projects in the Early–Middle Formative period reviewed in Section 3.1. Hieroglyphic dates clearly establish the use of the calendar later in the Formative, but recent analysis of solar alignments from Aguada Fénix and other Middle Formative platform complexes of the Usumacinta region show an earlier interest in charting 260-day intervals within the 365-day solar year (Reference Šprajc, Inomata and AveniŠprajc et al. 2023). It includes one site of interest, La Carmelita, with alignments to sunset on April 30 and August 13, a splitting of the solar year into 260/105-day intervals that became canonical in central Mexico during the Classic period as major alignments for Teotihuacan and Cholula but was present in the region during the later Formative at sites such as Totimehuacan (Reference Carballo and AveniCarballo and Aveni 2012). In the case of central Mexico, the division might correlate with the agricultural year, parsing much of the rainy season from the dry season. Of the Middle Formative Usumacinta platforms, Reference Šprajc, Inomata and AveniŠprajc and colleagues (2023: 9) observe: “These constructions possibly symbolized a sense of attachment to fixed localities and provided concrete images of communal collaboration that could be shared among the growing populations.”

Although the Mesoamerican calendar was a strongly shared means of coordinating time, it was employed differently regionally and temporally and could be used more inclusively or more exclusively. In other words, this cultural institution of reconciling solar and human cycles, the Calendar Round, could intersect differently with other institutions, such as political ones (i.e., to legitimate rule) or economic ones (i.e., to time markets or redistributive festivals). Drawing on ethnographic examples from Maya communities in the highlands of Guatemala and ethnohistoric examples from sixteenth-century central Mexico, Reference BrumfielBrumfiel (2011) distinguishes between more exclusive (elite-focused) and inclusive (non-elite-focused) uses of the calendar in Aztec pottery, sculpture, and codices. Whereas relatively more exclusionary or politicized uses of the calendar included state-sponsored sacrificial rituals, tax and tribute from conquered polities, and the training of a predominantly elite class through the calmecac, ethnohistoric texts and calendrical motifs on the bulk-luxury of Aztec polychrome vessels evince widespread knowledge of the functioning of the calendar by non-elites and their differential uses for suprahousehold alliance building through consumption events relating to lifecycle rituals.

Teotihuacan’s writing system was a clear precursor to Aztec writing, dating more than a millennium earlier and possibly representing an earlier Nahua or at least Uto-Aztecan language, but the corpus of hieroglyphs is much smaller, currently numbering fewer than 400 documented examples (Reference Helmke and NielsenHelmke and Nielsen 2021; Reference WhittakerWhittaker 2021). Nevertheless, more than half of these come from ceramic vessels and a better cataloging of provenience, when available, could elucidate whether these circulated through more inclusive or more exclusive networks, akin to Brumfiel’s study of Aztec ceramics. What is clear, however, is that Teotihuacanos employed Calendar Round dates and emphasized naming places (toponyms) or buildings, social roles of people (i.e., offices) over individuals, and cosmological narratives that paired such text labels with imagery. The importance of the Calendar Round at Teotihuacan is apparent not only through its hieroglyphs and urban planning but also through a characteristic symbol the pecked cross, which often feature 260 dots and divisions thereof carved into stone and the floors of buildings as a highly legible representation of the ritual calendar. Of more than seventy examples documented in Mesoamerica, most have been found at Teotihuacan and the city’s hinterland (Reference AveniAveni 2005, Reference Aveni, Morley, Renfrew, Morley and Renfrew2010). The symbol’s attributes “leave little doubt that practical elements related to the seasonal calendar and its relationship to the sacred count are extant in the petroglyphs” (Reference AveniAveni 2005: 43). We agree with Aveni and others that this Teotihuacano emphasis on broad dissemination of the Calendar Round was a means of sharing a time–space cosmovision with other parts of Mesoamerica linked to the city’s intensification of exchange networks (Figure 20), many emphasizing regional goods, and promotion of periodic markets (Reference CarballoFeinman and Carballo 2022; Reference Feinman, Nicholas, Hirth, Carballo and ArroyoFeinman and Nicholas 2020a).

Figure 20 Map with distribution of Teotihuacan-style pecked crosses in Mesoamerica.

Classic Maya writing evolved to be the most sophisticated written communication of the precolonial Americas and was capable of visibly recording any element of speech. Nevertheless, the writing system was applied disproportionately as a means of recording dynastic history and of aggrandizing elite individuals, especially the k’uhl ajaw or “holy blood lords” who emerged as powerful monarchs during the Late Classic period (Reference MarcusMarcus 1992; Reference MartinMartin 2020). The individualizing emphasis of Classic Maya writing included use of the Long Count calendar as a means of more precisely recording time, largely to the end of situating the actions of powerful people within it, as well as statements of authorship of texts and ownership of the materials labeled by texts. These emphases are seen in the writing of many of the most powerful Classic Maya capitals, but it is also important to note the variability in the use of writing across the territorially extensive Maya world, with some meaningful axes between the northern and southern lowlands and between more urban and rural settlements (Reference HutsonHutson 2016; Reference Hutson, Dunning and CookHutson et al. 2021; Reference RobinRobin 2013). Nonetheless, in the aggregate the emphases of writing in capital cities differ significantly from how Teotihuacanos wrote in several respects, including thematically, spatially, and in terms of broad legibility (Table 3) (Reference CarballoFeinman and Carballo 2022).

Significant differences in the iconography of Teotihuacan and Classic Maya capitals are also apparent and mirror differences observed in writing systems in conveying cosmological narratives that share elements of Mesoamerican belief but apply them in relatively more inclusive or exclusive ways. Teotihuacan famously has no unambiguous depictions of paramount rulers, while these are abundant in Classic Maya capitals. Teotihuacano art emphasizes themes of collective concern, such as agricultural abundance, cosmic renewal, and military might, and when humans are depicted they are typically done so as social roles and subordinated only to deities, such as in the Tepantitla mural where priests are subordinated to a deity and non-elites frolic in a paradise of butterflies, flowering trees, and a water-mountain feeding irrigated fields (Figure 21(a)). Flowery paradises associated with the afterlife and ancestors were widespread in art and religious narratives of Mesoamerica and adjacent culture areas, but we again see important differences among Classic Maya capitals, where it was primarily associated with rulers and royal lineages, and Teotihuacan, where it was primarily associated with militarism and a glorious death associated with valor in the battlefield (Reference Turner, Mathiowetz, Mathiowetz and TurnerTurner and Mathiowetz 2021). That such a flowery paradise was accessible to non-elite warriors at Teotihuacan is supported by the fact that its related iconography was broadly disseminated in mass-produced portable art, such as composite censers, and used as mural decoration in neighborhood centers on the urban periphery of the city among its lower socioeconomic stratum (Figure 21(c)) (Reference Carballo, Barba and OrtízCarballo et al. 2021). Some sort of pathway to quasi-elite status through warfare at Teotihuacan, analogous to the Eagle Nobles of Aztec society (see Table 2), is certainly possible, but we can also draw on collective military ideologies among the Greek poleis, Roman Republic, and other cases of more pluralistic polities outside Mesoamerica.

Figure 21 Photos of murals from Teotihuacan depicting types of paradises: (a) portion of a watery paradise from Tepantitla, with non-elites frolicking in a water mountain, tending to irrigated fields, and dancing and playing among butterflies, with the Storm God floating above; portions of a flowery paradise with bird, butterfly, and flower motifs from (b) the center of the city (Palace of the Sun) and (c) the urban periphery (Tlajinga).

Importantly, as with long-lived Eurasian polities like Rome, governing structures and ideologies of inclusion/exclusion changed through time in Mesoamerican polities. The more individualizing foci of Maya writing and visual communication emerged and declined with the Classic period, especially the Late Classic. For the most part, political messaging of the Preclassic and Postclassic periods appears very different and likely indicates broad shifts in models of governance of the sort reviewed ethnohistorically in Section 2.3 (see Reference Eberl, Gronemeyer and Marie Vela GonzálezEberl et al. 2023). In some cases, due to the richness of textual and iconographic evidence, we can perceive such shifts between more exclusive and more inclusive governance at the timescale of years rather than centuries. Such is the case from the ornately sculpted city of Copán, which suffered a devastating blow to the polity when the long and prosperous reign of its thirteenth ruler, Waxaklajun Ub’ah K’awil, ended by this dynast being captured and killed in 738 CE by a former vasal city-state Quirigua. When Copán’s fourteenth ruler took the throne, his strategy appears to have been to more broadly share power among noble families, depicting their leaders on the façade of a council house – the Popol Na or “Mat House,” with the mat representing a symbol of rulership (Figure 22) (Reference FashFash 2001: 130–134). This shift to a more pluralistic or oligarchic system of governance at Copán, associated with a period of political turmoil and the erection of a purposefully deliberative building, was relatively short-lived as the monarchy reestablished itself and next commissioned the longest hieroglyphic text known from the precolonial Americas in commemoration of its dynastic line of succession.

Figure 22 Photo of the council House (Popol Na), with sculpted mat motifs and hieroglyphs, from the Maya city of Copán, Honduras.

In her study of Classic Maya political dynamics, Reference FoiasFoias (2013: 192–197) considers the revenue streams of various polities, along the spectrum presented by authors such as Reference LeviLevi (1988) and Reference Blanton and FargherBlanton and Fargher (2008). She proposes that externally derived, prestige resources were critical to palace economies but that tax/tribute systems drew more on internal resources, which is logical given the difficulties in moving bulk goods long distances in prehispanic Mesoamerica. Foias proposes that council houses such as Copán’s could serve as important buffers to the power of dynasts during the Late Classic period and proliferated during the Terminal Classic, leading to the more pluralistic systems seen in places such as Chichén Itzá and elsewhere. Maya council houses were “another power bloc that would have competed with the other political factions of the k’uhul ajaw (divine lords), non-royal aristocrats, priests, and warriors (although there may have been some overlap between these groups)” (Reference FoiasFoias 2013: 197). Such a multiscalar perspective that considers the motivations and relative agency of various factions – different noble lineages, suprahousehold corporate groups, and varied subalterns – moves us beyond simplistic debates in Mesoamerican archaeology that consider governance primarily or exclusively in terms of the structure of paramount rulers. The rich array of formations from later periods provide us with culturally meaningful parameters for interpreting earlier cases in the archaeological record.

Efforts to systematically compare Mesoamerican cities and settlements have been revived over the last decade or two following earlier attempts associated with the boom in regional studies of the 1960s–1970s and a subsequent decline during the 1980s–1990s. Some of the revival relates to new geospatial technologies applied to archaeology, such as geographic information systems (GIS) and remote-sensing, and some comes from cross-disciplinary borrowing from fields such as political science, economics, sociology, cultural demography, and urban studies. Examples include attempts to quantify socioeconomic inequality using Gini indices (e.g., Reference Kohler and SmithKohler and Smith 2018), to understand relationships between scales of settlement and various societal phenomena (e.g., Reference Ortman, Cabaniss, Sturm and BettencourtOrtman et al. 2014), and to create global databases of cultural-historical variables for comparative analyses (e.g., Reference Turchin, Currie, Turner and GavriletsTurchin et al. 2013). In all cases, judgment calls must be made on the part of regional specialists and collaborators in order to fit various complex and dynamic phenomena into much simpler and more static categories. They do not deal well with micro-historical shifts like we just discussed at Copán or saw earlier among Tenochtitlan’s rulers. Nevertheless, if the data are recorded and reported transparently they can be debated, modified, and improved for new iterations of such analyses.

It is in this spirit of comparative analyses as works in progress that we have collaborated with other colleagues on studies of urban centers in Mesoamerica focusing on issues of systems of governance, communication, economies, inequality, infrastructure, and the intersection of these in relation to the longevity of settlements as central places (Reference Carballo, Feinman and López CorralCarballo et al. 2022; Reference Feinman and CarballoFeinman and Carballo 2018, Reference Feinman and Carballo2022; Reference Feinman, Carballo, Nicholas and KowalewskiFeinman et al. 2023). Durable infrastructure of the sort we have been referencing throughout this Element – irrigation systems, terraces, roads, neighborhood public spaces, and the like – can be viewed as materialized consensus among the households that built, maintained, and benefited from such projects (Reference SmithSmith 2016). In many cases, who benefited and what sorts of consensus was reached can be difficult to discern in the absence of text, with land tenure and subcommunity-level irrigation systems being two prominent examples. Yet when people shared physical walls between residences, terraces, access-ways, or social infrastructure within neighborhoods, some bonds of interdependence would have been necessary.

To evaluate the relationship between shared infrastructure and broader societal dynamics we coded a sample of Mesoamerican urban centers in the publications reviewed thus far based on variables of housing and shared/social infrastructure. For housing, we created a scale of 1–4 in degrees of shared space from isolated houses, patio-groups, agglutinated house compounds, and larger apartment compounds, of which only Teotihuacan had as the dominant housing type. For shared infrastructure, we created a binary scale (0 or 1) of walls, terraces, roads, plazas, temples, and other presumed civic buildings present or absent at the scale of neighborhoods, giving a total range of 0–6 and a total ordinal scale of 0–10, adding residences. Figure 23 illustrates the statistically significant negative correlation between the shared infrastructure of a sample of nine Mesoamerican centers with published Gini coefficients and detailed site maps amendable to assessing types of urban infrastructure.

Figure 23 Graph depicting negative correlation between calculated Gini score, measured by living area, and ordinal scale of shared residential and neighborhood space for a sample of nine Mesoamerican cities, indicating that those with more shared space tend also to be more equal. For sources, see Reference Carballo, Feinman and López CorralCarballo et al. (2022).

Although the sample is admittedly small, and the results should be taken as preliminary, it suggests that urban centers with more shared infrastructure were more equitable than those with less shared infrastructure. For this analysis, we wanted to include shared residential space because that variable is the primary determinant of calculating Gini coefficients in comparative literature, but we acknowledge that this variable alone poses problems. When it comes to housing value, as any realtor will tell you, location matters, and as any contractor will tell you, the quality of construction materials matters. A more accurate evaluation of inequality in household archaeology would therefore include consideration of location and labor invested in construction, but this complicates comparative analysis across world regions at larger scales. We have therefore also used other indices of inequality such as household goods (Reference Feinman, Faulseit, Nicholas, Kohler and SmithFeinman et al. 2018). Likewise, infrastructure could better be quantified as a continuous rather than an ordinal variable – for instance, linear meters of terracing per person in a settlement – but this also adds to the time cost of comparative analyses. Nevertheless, the correlation highlights some of the variables of urban living that have been documented as unique or relatively unusual, such as apartment living organized around neighborhood centers with the properties of institutions like the later calpolli or tlaxilacalli at Teotihuacan (Reference Cabrera Castro, Gómez Chávez, Mastache, Cobean, Cook and HirthCabrera Castro and Gómez Chávez 2008; Reference Manzanilla and ManzanillaManzanilla 2017).

For a second analysis, we examined the relationship between shared infrastructure and the longevity of a larger sample of Mesoamerican urban centers. We define longevity here as the duration in years that a given center was at approximately half or greater its estimated maximal population and we omitted any Postclassic centers whose trajectories were truncated by the Spanish invasion. For settlements that had two discrete periods of occupation with population estimates and maps with architecture, such as Cerro Jazmín, we separated them as two urban centers, whereas those with multiple occupation periods but featuring population estimates and/or architectural reconstructions for just one of them we used only a single apogee of occupation score. For this analysis, we only consider shared neighborhood infrastructure, not residence type, which we wanted for the analysis of Gini values based on living area. Figure 24 illustrates the statistically significant positive correlation between apogee length and shared infrastructure in a sample of twenty urban centers. The two analyses together are consistent with a broader literature in urban studies and related fields on positive-feedback loops between iterated interactions among neighbors and that initial civic investments in shared infrastructure can promote sustained social capital needed for larger-scale collective action in more equitable and resilient social formations (e.g., Reference Jensen Jeffrey and RameyJensen and Ramey 2020; Reference KlinenbergKlinenberg 2018; Reference HeatherMcGhee 2021).

Figure 24 Graph depicting a positive correlation between apogee length and an ordinal scale of shared neighborhood infrastructure for a sample of twenty Mesoamerican cities. Cerro Jazmín’s two separated occupations are coded as Formative and Postclassic. For data sources, see Reference Carballo, Feinman and López CorralCarballo et al. (2022).

We also queried the relationships between other variables that were not correlated in our sample of urban centers. For instance, the relationship between total population of an urban center and its longevity is close to random (r = 0.02, p = 0.91 for 27 cases). A positive correlation does exist between population density and apogee length, but it is not statistically significant (r = 0.37, p = 0.57 for 27 cases). However, when we classified urban centers along an ordinal scale of relative degrees of collectivity following the axes summarized in Table 1 we discovered stronger correlations. For this scale, we coded inputs along three domains of political economy (internal vs. external fiscal finance, levels of socioeconomic inequality), governance (more pluralistic vs. more autocratic), and architecture (levels of accessibility to public space and infrastructure, centrality of palaces) to rank levels of collectivity in sociopolitical organization (see Reference Carballo, Feinman and López CorralCarballo et al. 2022; Reference Feinman and CarballoFeinman and Carballo 2018). Using this scale, we see a stronger correlation between total population and collectivity (r = 0.29, p = 0.14 for 27 cases), though still not statistically significant, yet we register a highly significant (p = 0.01) positive correlation between collectivity and apogee length (Figure 25). Taken together, these analyses indicate that sociopolitical organization, rather than population measures, was the key variable in the resilience and longevity of urban centers in early Mesoamerica.

Figure 25 Graph depicting positive correlation between apogee length and an ordinal scale of collectivity in internal sociopolitical organization for a sample of twenty-seven Mesoamerican cities. Cerro Jazmín’s two separated occupations are coded as Formative and Postclassic. For data sources, see Reference Carballo, Feinman and López CorralCarballo et al. (2022) and Reference Feinman and CarballoFeinman and Carballo (2018).

While these analyses of macroscale patterns in urban centers are illuminating, all the cases listed in Figures 23–25 eventually suffered declines, and prehispanic Mesoamerica also offers contemporary lessons relating to different forms of societal collapse and reconfiguration. Climatic stressors, particularly relating to extended drought and unpredictability in annual precipitation, are frequently proposed as primary drivers of collapse without much consideration of historical variables and counterfactual cases of periods of climate stress that did not result in collapse. When paleoclimate data are combined with detailed analyses of cultural and historical variables in more interdisciplinary studies, the conclusions are much more nuanced and compelling (see especially Reference Hoggarth, Restall, Wood and KennettHoggarth et al. 2017). Ultimately, we see crises in social trust to be the key variable determining whether polities endure through perturbations stemming from environmental forces, factionalism, or macro-socioeconomic processes, whether their governance structure lies at the more pluralistic or more autocratic end of the spectrum (e.g., Reference Golden and SchererGolden and Sherer 2013; Reference ManzanillaManzanilla 2015). Trust is the social glue that keeps collective action and large societal formations together (Reference LeviLevi 1996; Reference Ostrom, Gardner and WalkerOstrom and Walker 2003). Collective labor as implemented through shared infrastructure is a powerful means of fostering trust and interpersonal ties both with coresidents and neighbors. When implemented effectively, it can generate virtuous cycles and sustained positive outcomes. Yet declines in trust undermine the social contracts that underpin contributions to the institutional systems responsible for public goods dissemination. As is expected from the relationships outlined in the schematic rendering of Figure 12, such breakdowns are capable of undermining the interpersonal and institutional networks necessary to keep the system going, resulting in enough people rejecting it and so leading to collapse.

In the central and southern highlands of Mesoamerica, the collapses of Teotihuacan and Monte Albán appear to be related to the compromising of more collective political arrangements by patron–client systems centered on noble houses (Reference Feinman and NicholasFeinman and Nicholas 2016; Reference ManzanillaManzanilla 2015). The polities that arose in the wake of their decline – such as Cacaxtla, Xochicalco, and Zaachila – exhibit more exclusive and individualizing art and architecture, including the greater centrality of palaces and noble genealogies. In central Mexico in particular, Teotihuacan remained a powerful touchstone for claims to political legitimacy through the Aztec period (Reference Carrasco, Jones and SessionsCarrasco et al. 2000). Teotihuacan’s collapse may have also affected developments in Oaxaca, leading to state balkanization into patron–client systems that distributed the more centralized authority of Monte Albán. Yet the city continued in the historical memory of Zapotec and Mixtec peoples as a powerful place, serving as a necropolis and referenced in Postclassic codices (Reference Feinman and NicholasFeinman and Nicholas 2016; Reference Jansen and Aurora Pérez JiménezJansen and Pérez Jiménez 2011: 317–327). Historical memory of these powerful places is likely part of the moral rhetoric of good governance recorded in oral histories like the huehuehtlahtolli passage from Section 2.3. It likely contributed to developing the Toltec and later Aztec “syntheses” in central Mexico and their related pan-Mesoamerican interaction spheres known as the “international style” or Postclassic world-system, associated with the institutions we reviewed from historically documented periods.

3.3 Summary

In this necessarily cursory review of millennia of Mesoamerican history, we hope to have highlighted several ways in which interpretive frameworks grounded in collective action theory and drawing on established cultural institutions from the ethnographic and ethnohistoric record can recast investigations on major issues relating to variability in societal organization over time and place. The triangulation between the archaeological record, documented cultural intuitions, and broader social theory, often called the direct-historical approach, has a long history of debate in Americanist archaeology (Reference Lyman and O’BrienLyman and O’Brien 2001; Reference WylieWylie 1985). When applied with too heavy a hand, it can serve to stifle evidence of change over time, something that is clearly observable in the archaeological record. We see the benefits of the approach taken here to include the setting of reasonable parameters for the institutions and strategies that past peoples were likely to have pursued at varied scales of interaction in a way capable of considering human agency and change through time. Frameworks for collective action and governance informed by a diversity of plausible cultural institutions shed new light on how elements of Mesoamerican societies considered canonical to defining the culture region from early definitions such as Reference KirchhoffKirchhoff’s (1943) – subsistence strategies, parsing time, regional economies, urban life, governance – varied based on differing strategies for human–environment interactions across the spectrum of socioeconomic status, political power, cultural identities, and other axes of human variability.

Forms of collective action can be detected from some of the earliest chapters of Mesoamerican history, documented archaeologically in cooperative subsistence activities and gatherings as part of forager fusion–fission cycles. With transitions to agricultural lifeways, building projects directed at collective subsistence infrastructure such as hydraulic projects and terraces created more elaborate anthropogenic landscapes conducive to analysis of resource management and group labor. Importantly, the development of agricultural lifeways in Mesoamerica did not include the domestication of many animal species, and no large ones, with significant implications for available pathways that aspiring elites could pursue in attempts to exert disproportionate control over resources. It also had the effect of creating more common-pool resource strategies in hunted, fished, or collected species. Infrastructure projects such as the Purrón Dam Complex, built by small-scale communities, exceed the scale of others from much larger, state-level societies and should serve as correctives to facile models of labor organization. We can no longer presume that collective production and labor projects require top-down, hierarchical management and must recognize more heterarchical contributions to creating and maintaining public goods. Analyses of monumental constructions must also distinguish, based on architectural and iconographic attributes, whether they were built for inclusive or exclusive purposes.

In a similar vein, models of governance in the deep past should draw from known cases rather than abstract social theory. They should recognize the differing scales of interaction and the fact that more collective intermediate social formations can exist in tension with more exclusionary ones based on patron–client relations; in some places and times the first were relatively more powerful and could push back against elite organization of the second and in others they were weaker and more constrained by unequal social relations. We have queried these issues by exploring shared infrastructure and how it correlates with levels of social inequality and the longevity of urban centers. Finally, when things fell apart and Mesoamerican societies underwent political or demographic declines and societal transformations, we should consider erosion in trust in institutions at least as much as ecological and social stressors, as the two often function relationally. These cases from early Mesoamerica therefore offer varied examples for us to draw on today in managing resources, living in cities, participating in governance, and generally getting along. We return briefly to those in the concluding section.