Introduction
Walter Scheidel

   THE “History of the Later Han Dynasty” reports the customs of Da Qin, or “Greater China,” a distant realm near the western ends of the earth. Its inhabitants were tall and shaved their heads, wore embroidered clothes, and planted silkworm mulberry trees. Their ruler occupied five palaces whose columns were made of crystal glass. Wary of natural disasters that would require him to step down and be replaced by someone else, he was known to honor this convention without complaint. That these features bear no discernible resemblance to the Roman Empire as we know it may well have something to do with the fact that access to this remote place was inconveniently blocked by “many lions and ferocious tigers which intercept and harm travelers: if the party does not include over a hundred men furnished with arms, they are invariably devoured.”1 Roman observers faced a similar predicament: for them, the easternmost reaches of Asia were “not easy of access; few men come from there, and seldom.” This made it difficult to visit the Seres or “Silk–People,” atheists who lived for more than two hundred years, occupied themselves with scraping silk from trees, were fierce and warlike as well as gentle and peaceful, sported blue eyes and flaxen hair, and never talked to strangers.2

This was unfortunate: had communication been less arduous, contemporary observers could hardly have failed to notice numerous similarities between their two mighty empires. Indeed, conditions in the Han state would have seemed less remarkable to a Roman observer than the sheer scale and order of Kublai Khan’s China were bound to appear to visitors from medieval Europe such as Marco Polo and their audience.3 Under even more auspicious circumstances, centuries of sustained contact might have allowed historians and political philosophers on both sides to track convergent trends over time: shifts from city–states to territorial polities and from military mass mobilization for interstate warfare to professional armies for border control; the growth of a protobureaucratic civil service accompanied by functional differentiation of power; formal dichotomies in provincial organization eclipsed by centralization of governmental control; the settlement and military use of peripheral groups in frontier zones; massive expansion of the money supply through standardized state–controlled minting; state intervention in manufacturing and trade; census registration and formal status ranking of the general population; codification of law; the growth of markets in land and the gradual concentration of wealth among elites; the transformation of smallholders into tenants, coupled with the growing strength of private patronage ties encroaching on state authority; unsuccessful attempts at land reform and eventual rural unrest; ideological unification through monumental construction, religious rituals, and elite education; the creation of a homogeneous elite culture and of corpora of classics; the emergence of court–centered historiography; ideologies of normative empire sustained by transcendent powers; and, later on, religious change leading to the formation of autonomous church systems and a philosophical and religious shift in emphasis from community values to ethical conduct and individual salvation. They might also have pondered the significance of conspicuous differences, such as the Republican background of the Roman state; the relative weight of local landowners and salaried officials in imperial government; the scale and functions of slave labor; the degree of autonomy of military power; the other party’s lack of a close equivalent to Rome’s civil law tradition or its emperor cult and to China’s dynastic stability or its Confucian–Legalist philosophy that underpinned state authority and legitimacy.

But distances were too great to permit these kinds of comparisons: the overland route from Chang’an to the Mediterranean coast wound its way across 4,500 miles (7,000 km) of steppe and mountains while even the most direct sea route from Egypt to northern Vietnam measures almost 6,500 nautical miles (12,000km). For each side, empirical knowledge of its counterpart remained confined to the goods that had been hauled across this forbidding expanse by intrepid intermediaries: silk, jade, and iron objects from China, linen, glass, and gypsum from the Mediterranean.4

Modern students of ancient history have no such excuse. While linguistic requirements and academic conventions continue to impede cross–cultural research, a vast amount of readily accessible information invites a comparative approach.5 Yet even today, scholarly interest in contact and exchange, in the objects and mechanisms of the transcontinental luxury trade, and in the concurrent transmission of supernatural beliefs and technical skills dramatically outweighs the amount of attention paid to the potential benefits of comparative analysis. The growing popularity of “Silk Road Studies” is emblematic of this imbalance, which, for all its persistence, has always been hard to justify.

It is hard to justify because only comparisons with other civilizations make it possible to distinguish common features from culturally specific or unique characteristics and developments, help us identify variables that were critical to particular historical outcomes, and allow us to assess the nature of any given ancient state or society within the wider context of premodern world history. Comparative history can take many forms. For instance, social scientists have distinguished between “analytical comparisons” between equivalent units for the purpose of identifying independent variables that help explain common or contrasting patterns or occurrences, and “illustrative comparisons” between equivalent units and a theory or concept that evaluate evidence in relation to predictive theory rather than particular units in relation to one another. Others think in terms of “parallel demonstration of theory” (equivalent to “illustrative comparison”) that aims for the empirical verification of theory, “contrast of contexts” that shows how the unique features of particular cases affect the unfolding of common social processes, and “macrocausal analysis” that employs comparisons in order to draw causal inferences about macrohistorical processes and structures and, ideally, to generate new theory. Others still advocate “universalizing,” “encompassing,” and “variation–finding” techniques.6 Most actual work in this area has followed a “case–oriented” rather than a “variable–oriented” approach that views historical cases as configurations of characteristics that are to be related to particular outcomes.7 Comparative history, by its very nature, is not about “laws” but about the search for what has been called “robust processes,” defined as combinations of characteristic initial conditions that produce a particular outcome. The main questions are which factors were crucial rather than incidental to observed developments and how different contexts could produce similar outcomes, or vice versa. In other words, comparative history uses case–based comparisons to investigate historical variation, to offer causal explanations of particular outcomes by identifying critical differences between similar situations and/or by identifying robust processes that occur in different settings.8 These are the main goals of the following chapters.

Comparative research is necessarily sensitive to sample size. In principle, consideration of a large number of cases makes it easier to identify significant variables or conjunctures and to support generalizations. In the present case, however, the absolute scarcity of what world–systems theorists would call “corewide empires” imposes severe constraints on the range of comparative analysis.9 Our focus on two very large and durable states that were created through the absorption of all or almost all state–level polities in their respective ecologically bounded macroregions ensures direct comparability in terms of observation as well as analysis: both proceed at the (very) macrosocial level of the emerging or mature near–monopolistic super–state.10

As I have already noted, this approach has seldom been adopted in modern scholarship. Moreover, although some explicitly comparative work on China and the Mediterranean in antiquity has appeared in recent years, it is very heavily weighted in favor of intellectual history. These studies tend to focus on the nature of ethical, historical, and scientific thought in ancient Greece and early China. The most prominent and prolific proponent of this line of inquiry has been Geoffrey Lloyd, who has published no fewer than six books on science, medicine, and ways of understanding the world in these two environments.11 A small number of other scholars have produced comparative work in related areas.12 Fritz–Heiner Mutschler has shown particular interest in the historiographical traditions of imperial Rome and Han China.13

At the same time, comparative studies of political, military, social, economic, or legal institutions have remained extremely rare.14 The comparativist interests of Max Weber, Karl Wittfogel, Shmuel Eisenstadt, and Samuel Finer have had little impact on the research agenda of specialist historians in either field.15 Recent historico–sociological studies of imperialism and social power that deal with ancient Greece and Rome comparatively and within a broader context do not normally pay much attention to conditions in China.16 Concrete case studies by professional historians are almost impossible to find: Hsing I–Tien’s unpublished dissertation on the political role of the Roman and Han military is the only booklength study that comes to mind. More general but shorter comparative surveys by Günther Lorenz, Christian Gizewski, and Samuel Adshead have recently been joined by a similar contribution by Maria Dettenhofer but have thus far generally failed to generate further debate.17 It is emblematic of the ideational focus of existing research that the most ambitious project to date, a substantial collection of papers prepared for an international conference on “Conceiving the ‘Empire’: Ancient China and Rome—An Intercultural Comparison in Dialogue” held in Germany in 2005, deals exclusively with textual and artistic representations and reflections of large–scale state formation.18 Victoria Tin–bor Hui’s recent political science analysis of balancing mechanisms in Warring States China and early modern Europe provides an attractive model for the comparative study of Rome and China but eschews a synchronic approach.19

The present volume is the first in a series of works to engage in the comparative institutional study of ancient Rome and early China. A few years ago, I launched an international collaborative research initiative called the “Stanford Ancient Chinese and Mediterranean Empires Comparative History Project.” It has three principal objectives:

To contribute to our understanding of state formation in the ancient Mediterranean (with particular emphasis on the Roman empire) and in China (with particular emphasis on the Warring States and Qin–Han periods). Two conferences have been devoted to this goal.20

To study the character and causes of the long–term divergence between periodic imperial reunification in China and the absence of core–wide empire from western Eurasia following the fall of the Han and Roman empires. A separate workshop has focused on this phenomenon.21

To ask whether ancient and/or early medieval state patterns of formation and associated developments in eastern and western Eurasia were instrumental in determining the nature of what has been called the “Great Divergence” of the last two centuries that witnessed a dramatic acceleration of technological progress and increases in consumption and wellbeing in the modern “West.” This is the subject of a fourth meeting.22

The first of these issues calls for case studies of different aspects of state-society relations in these two historical environments. The chapters in this volume cannot be more than a modest first step in this direction. The opening chapter proposes a preliminary interpretative framework for more detailed work by highlighting the scale and limits of convergent trends in ancient imperial state formation. Nathan Rosenstein’s contribution elaborates on a key element of this perspective in his comparative analysis of the relationship between interstate conflict and the development of state institutions. His focus on systemic forces is complemented by Karen Turner’s study of internal coercion as embodied by penal law. Maria Dettenhofer considers the location of women and eunuchs in the emerging power centers of the Han and Roman imperial courts. The remaining chapters deal with the social and political contexts of economic issues. Peter Bang offers a wide–ranging survey of imperial styles of surplus extraction and consumption. Mark Lewis deals more specifically with traditions of euergetism and seeks to account for the different mechanisms of welfare provision in Han and Roman society. My own final chapter surveys the divergent evolution of coinage in these two systems and explores its underlying causes and economic consequences.

In their own ways, all of these contributions share the goal of identifying and explaining specific features with reference to particular contextual variables. Their comparative perspective heightens our appreciation of similarities as well as differences between the Roman and Chinese experiences that is crucial to this endeavor: without it, causal analysis lacks vital controls. At the same time, these contributions show how much work remains to be done. The sequel to this volume will return to some of these issues while introducing additional themes: the character and functioning of different levels of imperial rule from the monarchs to state officials and local elites; the accommodation and instrumentalization of religious beliefs by the state; the political and economic dimension of urbanism; and the relations between self–proclaimed “universal empires” and their peripheries.23 In a forthcoming monograph on state–society relations in ancient Rome and China and their immediate successor states, I will develop a more synthetic account of many of these features by assessing the changing configurations of political, military, economic, and ideological power.24 In addition, my colleague Ian Morris and I are planning to publish the results of two separate meetings that deal with divergent trends in eastern and western Eurasian state formation since late antiquity and their long–term consequences.25 Early periods of history also occupy a prominent position in Morris’s comparative study of social development in western Eurasia and China.26

Together with the present volume, these forthcoming studies are meant to contribute to the creation of a broader framework for the study of particular regions, periods, and processes that transcends the historical specifics of those regions, periods, and processes. They are also meant to link up to the efforts of other cross–culturally oriented collaborative initiatives such as the European research network “Tributary Empires Compared,” which juxtaposes developments in the Roman, Mughal, and Ottoman empires, and the United Kingdom–based “Network on Ancient and Modern Imperialisms.”27 All these efforts are necessary for creating a basis for multicase comparisons: larger samples make it easier to design and test causal hypotheses and, in time, may even allow us to complement caseoriented comparisons with variable–oriented analysis of premodern historical societies. A generation ago, Moses Finley mused that “[i]deally, we should create a third discipline”—in addition to anthropology and sociology—“the comparative study of literate, … pre–industrial, historical societies,” and suggested “preMaoist China, pre–colonial India, medieval Europe, pre–revolutionary Russia, medieval Islam” as appropriate comparanda for students of the Greco–Roman world.28 But what the (academic) world needs is not yet another discipline that would inevitably end up policing its very own boundaries: what we need instead is the specialists’ willingness to overcome existing compartmentalization and contribute their expertise to collaborative efforts that address bigger questions. The study of ancient civilizations, all too often weighed down by the need to accumulate recondite yet indispensable technical knowledge, has much to gain and nothing to lose from broader perspectives.