Southeast Asia is a paradox to Western scholars. Few are familiar with its history, yet Southeast Asia has been a veritable intellectual resource extraction zone for twentieth- and twenty-first-century social thought: imagined communities, galactic polities, agricultural involution and the moral economy of peasants all emanate from work done in Southeast Asia. The region's archaeological record is equally paradoxical: late Pleistocene ‘Hobbit’ hominins disrupt models of human origins, the world's largest Buddhist monument of Borobudur now sits in a wholly Muslim land mass in central Java, and the world's largest premodern city of Angkor is located in Cambodia, a country that remains resolutely rural. So we should not be surprised that Scott's Against the Grain: A deep history of the earliest states draws from a career in Southeast Asian studies to study human history (the entire Anthropocene). This essay concentrates on how Scott believes early Mesopotamian states became legible.

Scott's engaging book will please the educated non-archaeological public and rile archaeologists who dislike treatments of our field by those outside it. Certainly some of Scott's key points are conventional wisdom in our field, like the reciprocal nature of plant and animal domestication (e.g. Rindos Reference Rindos1984) and the short lifespans of ancient states (e.g. Baines & Yoffee Reference Baines, Yoffee, Feinman and Marcus1998; Yoffee & Cowgill Reference Yoffee and Cowgill1988). And few working archaeologists are environmental determinists vis-à-vis collapse: most agree that human agency shaped a world of climatic ‘predictable unpredictability’ (Miller Reference Miller, Miller, Moore and Ryan2011). Yet much of Scott's model is generalizable to premodern Southeast Asian states. One contribution from his work concerns literacy as a ‘technology of communication’ (p. 148; see also Scott Reference Scott1999; Reference Scott2010). Southeast Asian states emerged soon after the embrace of Sanskrit-derived writing systems. The role of tax versus tribute is currently under discussion, and libraries that Chinese annals describe have not been found, perhaps because they held records written on bamboo. Remaining stone inscriptions, however, fit Scott's model: writing made Southeast Asian states legible and elites into kings.

A second contribution lies in Scott's vision of ancient states as brittle structures that were prone to shatter. He is at least as interested in continuities in rural agrarian life through time as he is in particular ‘collapse’ events (a term he dismisses as ‘histrionic’, p. 201) of the small elite and urban sector of the population. Elites formed delicate superstructures that rested precariously atop foundations of ‘smaller and more stable components’ (pp. 187, 253), and this elite superstructure was vulnerable to climatic disasters, epidemiological events and human-induced ecological degradation. Scott's ancient states had long arms that extended into rural agrarian communities to wring resources from farmers and create ‘normal, grinding forms of oppression’ (pers. comm., December 2017). Counterposing rural resilience against urban fragility makes sense in a Southeast Asian context, where alluvial lowlands hosted sequential states for a millennium before the Early Modern period (e.g. Aung-Thwin Reference Aung-Thwin, Aung-Thwin and Hall2011; Stark Reference Stark, Schwartz and Nichols2006b).

Scott's application of modern peasant political ecology to ancient states, however, may be analogical over-reach for premodern Southeast Asia. State power (or ‘iron control’) in his Chayanovian model (p. 152; see also Scott Reference Scott1985; Reference Scott2010) drove populations to flee to upland refugia to escape the mantle of state oppression. That this happened in the early colonial Philippines is suggested by recent archaeological research (Acabado Reference Acabado2018), but such developments arrived late, with world mercantile capitalism, and then intensified under colonialism. Few premodern Southeast Asian states exercised effective control much beyond a 30 km radius of the capital: the approximate distance an army or civilians could travel in a day (Hendrickson Reference Hendrickson2011, 448; Lustig Reference Lustig2011; also see e.g. Deloche Reference Deloche and Walker1993, 245–51, 284). Warfare and direct management were expensive. Maintaining webs of patron-client relationships was more effective, but required continuous nurture: such linkages were subject to fracture.

Take the case of ninth–fifteenth-century ad Angkor (northwest Cambodia), where more than a century of research has generated a massive literature in which archaeological research remains subordinate to historical study. An inland agrarian state, Angkor was a ‘grain-population complex’ (p. 204). The precise geographic reach of Angkor at different points in its history remains hazy, despite important recent geospatial work (e.g. Evans et al. Reference Evans, Pottier, Fletcher, Hensley, Tapley, Milne and Barbetti2007; Hendrickson Reference Hendrickson2010). The large urban complex that we call Greater Angkor absorbed resources from across the empire, in areas that lay beyond the 48 km reach that Scott (p. 205) suggests. In fact, one of Angkor's two biggest agrarian areas lay nearly 300 km southeast of the capital, in the Mekong Delta.

What explains the Angkorian civilization's resilience over nearly 600 years? Not territorial circumscription from population pressure, which underwrites ‘coercion theory’ approaches to archaeological state formation in Vietnam (e.g. Kim et al. Reference Kim, Kusimba and Keeley2015). Current Angkorian research suggests its capital was characterized by low-density urbanism (Fletcher et al. Reference Fletcher, Barbetti and Evans2003). Only Java, across all of Premodern Southeast Asia, has well-documented demographic packing (Christie Reference Christie1991), although northern Vietnam surely housed large populations. Truly dense urbanism emerged later, in the fifteenth–sixteenth centuries, as port cities grew to take advantage of new European trade opportunities (Reid Reference Reid1980).

We can fit Scott's Mesopotamian model (viz. ‘late-Neolithic resettlement camps’) to mainland Southeast Asia's archaeological trajectory by moving the clock forward. Mainland Southeast Asians did not aggregate into large settlements until the mid first millennium bc (known variously as the ‘Iron Age’ or protohistoric period). When they did, they chose areas of rich alluvia with annual flooding that brought high returns from rice agriculture. Some of these early settlements are the basal layers of historic capitals associated with the late first millennium ad (Ng Reference Ng, Smith and Watson1979). Indic ideas and ritual practices infiltrated the region's coasts and river valleys in the early centuries ad (Stark Reference Stark2006a), and populations adopted these new ideas to sacralize their landscapes (Stark Reference Stark and Emberling2015a). Newly established shrines, temples and hermitages (practices, ritual paraphernalia, sounds and smells) attracted people to these centres and organized their lives. An eclectic mix of Brahmanic, animist and Buddhist ideas and practitioners co-existed in a social pluralism that integrated rural communities with an ever-expanding state and thrived for nearly a millennium.

Epigraphic studies increasingly suggest that establishing patronage networks was key to growing the Angkorian state: it ensured a population influx to the capital, provided raw materials needed to craft wealth finance, and bound provincial elite families, whose support was needed to convert rural agrarian production into staple finance for the capital. Like contemporary states throughout the region, patron-client relationships formed the warp of the Angkorian polity; public ritual practice, replete with processions and pageantry at the capital (Stark Reference Stark and Yoffee2015b), was the weft.

In what ways has James Scott produced a general model that Southeast Asian archaeologists can use? Scott's masterly development of central themes has a conceptual elegance that occasionally also falters on factual grounds (e.g. Brookfield Reference Brookfield2011; Lieberman Reference Lieberman2010). A self-described anarchist, he has experience with and sympathy for rural Southeast Asian peasants: the subjects of colonial and post-colonial grand state experiments that produced deep collateral damage. Where he sees states as repressive, archaeologists also understand their magnetism and the ‘pull’ factors that attracted premodern Southeast Asians to growing cities. Their capitals were sites of monumental temples, ritual and political performance, and elites whose personal and familial success hinged on a delicate balance of charisma, political acumen and good climate.

Reading James Scott's Against the Grain: A deep history of the earl iest states is an unmitigated pleasure and stimulus for archaeological theory building. Its value for Southeast Asian archaeologists lies primarily in his contention regarding the intrinsic brittleness of ancient states that contrasted with resilience in rural agrarian communities from which they grew. Answering James Scott's fundamental, if provocative question—namely, why did ancient states work at all?—is a very good archaeological starting point.