Against the Grain is both a wide-ranging voyage of discovery and a regionally focused study of the trajectory of agriculture from its earliest appearance until historical times, coupled with discussion of the mechanisms that maintained early states. For Scott, the state is a fragile entity (pp. 21, 23, 118, 125) based on the production of grain, along with water transport, city walls, tax collection, specialized administrators, monumental centres, kings, social hierarchy, filth, epidemic disease and an insatiable demand for enslaved labour. With such a definition, there is a little hope that the societies of Eurasian pastoral nomads can be seen as anything other than ‘barbarians’ living outside the laws and hierarchies of agricultural states. It is these Eurasian nomadic pastoralists and their relations with the state that will form the focus of this commentary.

Eastern Eurasian nomadic pastoralists—known by ethnonyms such as Xiongnu, Türks, Uighurs, Mongols and Manchus among others—are the ‘stars’ of the barbarian world described by Scott (p. 179). The key characteristics of the barbarians of Against the Grain is that they lived in non-state communities, did not practise centrally administered farming on a large scale, lacked permanent settlement apart from a few small towns and were generally disease-free. They were organized in unadministered tribes and chiefdoms with weak sovereignty and were difficult to tax. Most importantly, they were violent raiders whose major economic engine was the plundering of sedentary states (p. 180).

Scott's presentation of Eurasian nomads is heavily influenced by the work of Barfield (Reference Barfield1989), but also invokes the ideas of the fourteenth-century social theorist Ibn Khaldun, in which nomads are seen as the antithesis of civilization. Drawing on both Nissen's (Reference Nissen and Barrelet1980) ancient Mesopotamian textual analysis and Lattimore's (Reference Lattimore1940) ethnohistory, Scott recognizes that the sharply contrasted worlds of farmer and herder have actually been a continuum (pp. 66, 172). Although the permeability of the boundary between these two worlds is fundamental to the book's argument with information, objects, materials, domesticates, skill-sets and people all passing from the realm of the state to that of the barbarians, the asymmetry of Scott's sources serves to highlight differences and provide power to his strongly dualistic argument. In short, the model of the shadow empire (Barfield Reference Barfield1981; Reference Barfield1989) is profoundly influential in Against the Grain. This model suggests that nomadic pastoralists are dependent on sedentary states for critical elements in their nutrition and technology, as well as for inspiration for their political complexity.

To sedentary sources the actions of pastoral nomads are often unfathomable or random. Detailed studies that engage with pastoral subsistence practices have helped to define pastoral nomadism with a specificity similar to that of farming in prehistory. Just as there are particular soils and work schedules associated with farming, there is a similar discipline associated with nomadic pastoralism. On the local scale, nomadic pastoralist groups regularly move most of their human and animal members to locations that are advantageous for the animals. This movement system is seasonally based, repetitious and responsive to environmental variation, and crises are dealt with through exceptional activity to protect herds. This mobile economy creates the potential for particular political processes. Honeychurch (Reference Honeychurch2015) describes the long history of polities in which multi-dimensional networks allowed for a particular ‘spatial politics’ where population mobility, the extensive scale of territory and a particular broadly participatory politics were made possible by the underlying pastoralist-based economy. Reflecting this, the complexity of the archaeological record in Mongolia raises questions for Scott's strongly dualistic thesis about the foundations and practices of complex societies.

It is in the study of the Xiongnu polity, the original shadow empire, that the dependency model has been most strongly critiqued. There is no evidence, for example, of the dramatic picture presented in Against the Grain of border markets where barbarian Xiongnu traded slaves for steel. Archaeology has provided the remains of a few small towns, central places, trade systems, monumental tombs (Brosseder & Miller Reference Brosseder and Miller2011), traces of farming (Wright et al. Reference Wright, Honeychurch and Chunag2009) and the manufacture of iron. This last is a key issue, because iron was central to the border economic policies of the Han Empire. Analysis of objects from elite tombs and quotidian habitation contexts (Park et al. Reference Park, Gelegdorj and Chimiddorj2010) and the discovery and excavation of industrial smelting and foundry facilities (Sasada & Chunag Reference Sasada and Chunag2014) show that within Xiongnu territory a range of iron manufacture was taking place and that iron objects were being made in the Han style, rather than imported from Han territory (Park et al. Reference Park, Gelegdorj and Chimiddorj2010). In both production and the performance of rulership, Xiongnu were actors within a larger east Asian world. The political complexity, wealth distribution and nature of administration of the Xiongnu has been extensively written about and debated (Barfield Reference Barfield1981; Batsaikhan Reference Batsaikhan, Brosseder and Miller2011; Honeychurch Reference Honeychurch2015; Kradin Reference Kradin2005), but recent research suggests that for the Xiongnu state the Han empire was not the wellspring from which it flowed, but a foreign policy issue to be managed through a variety of methods.

The Uighur Empire (ad 744–840), which figures in Against the Grain’s study of barbarians as successful nomadic raiders preying on the Tang Dynasty, is known archaeologically primarily through the remains of its capital Kharbulgas and also through a growing number of smaller discoveries throughout Mongolia. Tang visitors to the Uighur heartland (Mackerras Reference Mackerras1973) described fields, orchards and irrigation systems, and recent archaeological investigations have begun to reveal the extent of the urban and suburban structure of the Uighur capital (Hüttel & Erdenebat Reference Hüttel and Erdenebat2010) and its hinterland (e.g. Ahrens Reference Ahrens2016). This work reinforces the historical accounts and shows a sprawling city surrounded by a network of enclosed fields, mortuary zones, smaller towns and suburban compounds or estates. This urban core region of the Uighur polity included smaller cities up to 120 km away (Kradin et al. Reference Kradin, Ivliev, Ochir, Vasiutin, Satantseva, Kovychev and Erdenebold2014). This all suggests that Uighur conflict with Tang was much more of a contest between states than an extraction strategy of mobile raiders. This, along with the many generations of people living in urban and agricultural contexts, makes it possible to argue that the Uighur state was a state like any other in the early medieval world and its culture and aesthetics, tombs and urban forms, as well as its politics, would not have been out of place in Tang China or Silla Korea.

Eurasian states were not dependent on alluvium to prosper. Instead, administrators developed systems that managed, taxed and maintained diverse subsistence economies. An example of this hybridity of subsistence and administration can be seen in the Kitan-Liao empire (ad 907–1125) (e.g. Kradin & Ivliev Reference Kradin and Ivliev2014). Often overshadowed by their vastly more successful successors the Mongol Empire (ad 1206–1368), much of the infrastructural and administrative successes of that later world empire appears to have been rooted in the early Liao state. The founders of the Liao were drawn from an ethnic group known as the Kitan, who celebrated their nomadic pastoralist heritage through art, recreation and elite spatial politics, but also ruled a territory with one of the longest histories of sedentary farming in the world (Shelach Reference Shelach2000). Within a few decades of its foundation the Liao had incorporated both large swathes of what is today northern China, including Beijing, and the old Mongolian Turkic heartland in the Orkhon (Kradin et al. Reference Kradin, Ivliev, Ochir, Vasiutin, Satantseva, Kovychev and Erdenebold2014). As rulers who managed both mobile pastoralist, sedentary farmers and urban populations the Kitan were acutely aware of the different modes of living of their subjects, regulating dress, law codes and land use both to highlight and to obscure differences among the population. Government officials managed populations with different skill sets, collected taxes in grain from farming communities and managed animal stocks and grazing systems. Subsistence skills were a commodity managed by the state. Populations of pastoralists and city dwellers were moved throughout the empire (Kradin & Ivliev Reference Kradin and Ivliev2008; Wittfogel & Feng Reference Wittfogel and Feng1949), bringing their special and complementary skills to different regions. Moreover, it should also be stressed that the Liao (as well as the Turkic Khaganates, the Uighur empire and the Mongol empire) developed a writing system for their own use. In all but the case of the Khaganates, they drew on different existing systems, but shaped them to their own needs. The widespread, though not dense, distribution of inscriptions and painted texts across Mongolia demonstrates that literacy was found in more places than the royal courts of Eurasia.

Examples could be multiplied, but the overall point here is that the detailed examination of the most successful barbarians identified in Against the Grain shows them to have been complex, state-level societies in their own rights, visiting on their citizens at least some of the ills of civilization and interacting with alluvial states as peers. Each of these states endured for one or two centuries, a brittleness on a par with most early states. Taken together, however, the Turkic and Mongol polities form a continuous sequence of at least 600 years of urban and partially agrarian living communities well outside the major alluvial plains of east Asia.

What are we to make of the thesis of Against the Grain if the star barbarians are not so barbaric after all? From the advent of large polities such as the Xiongnu into historic times we see a series of states built up on an alternative model to that presented in Against the Grain. Though there are commonalities, such as hierarchies, powerful rulers, monumental centres, the management of land resources and taxes, there are also alternative practices: spatial politics (Honeychurch Reference Honeychurch2015), headless states in which it is not the apex leaders who hold the polity together over the long term (Sneath Reference Sneath2007), and hybrid governments that successfully rule populations of farmers and pastoralists.

In the end, Scott's thesis is limited as a potential universal statement about all agriculturally based states and their more dynamic neighbours, but it is an accessible and readable presentation of the state of the art in the Levant and Mesopotamia. It might be that the true golden age of the barbarians is to be found in one of the lacunae of the book. Instead of an age of raiding nomads, it is the Neolithic period that might have been the real golden age of barbarians. During those several millennia, a diversity, creativity, mobility at different scales, and skilful mastery of materials and biology, along with exuberant human explorations of the world, were at their height and all without the overarching oppressive structure of a state's administrative hierarchies.