What, then, were the root causes of the onset of modernity? Two circumstances, one exceedingly remote and the other far less so, were critical for European fragmentation and polycentrism. From the late Cretaceous onward, the collision between the African and the Eurasian tectonic plates caused the Alpine orogeny, a process that formed the Carpathians and Alps and raised them ever higher. Without the former, Transylvania would not have appeared and the great Eurasian steppe would extend to Vienna; and without the Alps, it might stretch even farther west. Tens of millions of years later, mounted warriors might have pushed state formation in a different direction, toward serial empire.46
In actual history, at roughly the time when such influences would have made themselves felt, Germanic warriors in the Iberian peninsula, Gaul, Italy, and Germany gained control over their means of subsistence and lordly autonomy undermined central state power. Had taxation and centralized governance been sustained, the odds of imperial restoration would have been improved, and a rerun of the Roman experience might no longer have been impossible.
A third factor arguably contributed as well, albeit in a more ambiguous way: the perseverance of the followers of an obscure Jewish prophet in building up a far-flung network that evolved into a hierarchical and fairly cohesive transnational organization. Its presence simultaneously contributed to and helped offset post-Roman fragmentation.
Plate tectonics, warriors, and—perhaps—preachers interacted in very specific ways to create the environment that gave birth to modernity. But where does all this leave the Romans? What did they ever do for us before their empire fell apart? What seems to me the most honest answer is also disappointingly vague: that they quite possibly may have done something very important for us—if, but only if, their empire, by turning to Christianity, laid some crucial foundations for much later development—but that they just as likely may not have contributed anything essential at all to this eventual outcome and thus failed to shape the general appearance, if not some of the finer points, of the world we live in today.
In the end, competitive fracture may well have mattered more—or rather, even more—than residual cultural unity: the unanswerable question is whether the former, on its own, would have been enough. Rome’s unreversed demise was an indispensable precondition of modernity. But when it comes to explaining this breakthrough, does it really matter that its empire ever existed at all?
GLOSSARY
Some of the key terms employed in this book are bound to mean different things to different people. Explaining my own usage of the most common ones should help avoid misunderstandings.
CHINA Depending on context, China refers to the (fluid) geographical area that was controlled by the main dynasties from the Qin onward, as well as to the territory of the current People’s Republic of China (PRC). Where applicable, I differentiate China from “core China” or “China proper,” shorthand for the PRC exclusive of Manchuria, Inner Mongolia, Xinjiang, Tibet, and Taiwan. For the sake of convenience, if anachronistically, I loosely apply the terms “China” and “Chinese” to these core areas going back to the second millennium BCE. I do not use China and East Asia interchangeably; the latter includes Korea and Japan. My use of any of these terms is not intended as commentary on ethnic identity or current affairs.
EMPIRE A generally large, composite, and often multiethnic state organized around a dominant center that controls subordinate and sometimes distant territories and populations (or “peripheries”). Even this conservative definition is far from precise: imperial formations did not have to be particularly large and varied greatly in terms of heterogeneity and intensity of demarcation between dominants and subordinates. In practice, I apply this term only to physically and demographically substantial polities that covered from several hundred thousand to tens of millions of square kilometers and claimed between millions and hundreds of millions of subjects. Where applicable, I employ characterizations such as “agrarian,” “traditional,” or “tributary” to highlight salient attributes. “Hegemonic” empires were states that effectively dominated territory and population on a subcontinental scale by a combination of direct rule and often highly asymmetric relations with neighboring political entities, in periods when their existence and overall territorial integrity were not systematically at risk from challengers. Whenever I use the term “empire” without further specification, I refer to contiguous land empires (based in agrarian areas, the steppe, or both). I distinguish as “colonial empires” the overseas territories European powers acquired from the late fifteenth century onward. I employ the term “imperiogenesis” to describe state formation on a large scale.
EUROPE I adopt the conventional definition of Europe as comprising the lands west of the Urals, the Ural River, the Caspian Sea, the Caucasus, and the Black Sea. I also single out specific parts thereof. “Latin Europe” refers to Europe west of the areas traditionally dominated by the Orthodox Church (currently west of Russia, Belarus, Ukraine, Romania, Serbia, and Greece, a line that has not substantively shifted since the Schism of 1054). “Western Europe” is a slightly smaller portion from the Atlantic to the Germanic- and Italian-language areas in the east. “Northwestern Europe” encompasses the British Isles, the Scandinavian Peninsula, the Low Countries, northern France, and northern Germany. More narrowly but somewhat poorly defined, the “North Sea region” centers on Britain and the Low Countries. I have sought consistency in usage but may use “European” as shorthand if the context is clear. I generally avoid the labels “West” and “Western”: when they are used, they refer in the first instance to Europe (and especially Latin Europe) plus its main colonial settler societies in North America and the Antipodes.
FIRST GREAT DIVERGENCE As explained in chapter 7, I use the term “First Great Divergence” to describe the secular divergence in patterns of state formation between Latin Europe on the one hand and the Middle East, South Asia, and especially East Asia on the other, which commenced and solidified during the second half of the first millennium CE. Unlike the Second Great Divergence, this process was political in nature: it witnessed the emergence of a durable polycentric state system in Latin Europe as opposed to the serial constitution of very large empire in other parts of the Old World.
POLYCENTRISM A polycentric system is characterized by multiple centers of power, authority, and control. Polycentric arrangements exist between polities within state systems that are not dominated by a single hegemon as well as within any given polity in which different constituencies autonomously exercise power that balances that of the central political leadership. With regard to both types of arrangements, I use polycentrism and fragmentation as effectively interchangeable terms. They serve as an ideal-typical antithesis to the (equally ideal-typical) monopolistic exercise of power within a given polity or the hegemonic status of an imperial state within a very large catchment area. Monopoly, hegemony, and polycentrism are inevitably a matter of degree, and I spend much time in this book exploring their relative preponderance in different historical settings.
SECOND GREAT DIVERGENCE I employ this term as the equivalent of what is generally called the “Great Divergence,” the process whereby the most advanced economies of Western Europe pulled ahead of other parts of the world in terms of economic productivity. Definitions of this divergence vary: while some apply it broadly to relatively subtle differences in output and welfare that can be traced back to the Middle Ages, others use it in a more focused manner to contrast the onset of industrialization and sustainable economic growth in eighteenth-century northwestern Europe with the persistence of more traditional modes of economic activity and technological development elsewhere, most notably in Asia. I generally call this process the “(Second) Great Divergence,” a compromise that seeks to reconcile standard usage (which recognizes only a single “Great Divergence”) with my introduction of the term “First Great Divergence.” When mentioned without qualifiers, “Great Divergence” refers to the “Great
Divergence” as conventionally defined.
STATE For the purposes of a wide-ranging comparative study that reaches far back in time, I define the state very loosely as a somewhat durable and hierarchical political organization that claims and at least to some degree exercises authority over territory, people, and resources, and does so by asserting priority over certain claims of other groups within or outside that territory and through its ability to establish this priority by coercive means, at the very least on an occasional punitive basis. This minimalist definition omits features commonly associated with more mature states, such as legitimacy, sovereignty, or a monopoly of legitimate violence in the enforcement of rules and the legal order. Historically, such political organizations greatly varied in their level of centralization and effectively shared many claims, prerogatives, and functions with other constituencies. Unless we apply formally Eurocentric and modernizing concepts of statehood, these common characteristics should not deter us, however, from labeling such polities “states.” I use the more neutral term “polity” interchangeably with the term “state” and also to describe political entities that might not have met even the basic requirements set out here, such as chiefdoms, tribal confederations, or particularly decentralized or ephemeral political organizations. I do not employ the term “nation” as a generic equivalent of “state” but occasionally switch between “interstate” and “international” to avoid repetitiveness.
STATE FORMATION I use this term in a very general sense to refer to any developments that create and modify institutions of governance over people and territory beyond the level of small-scale communities. I envision state formation as an ongoing process of structural change that may (but need not) lead to increased centralization or state capacity: historically, state power within given polities waxed and waned over time. Very occasionally, I refer to “state (de)formation” to emphasize the decentralizing and regressive character of a particular process.
STATE SYSTEM A cluster of several adjoining independent states that systematically interact over the long term in the political, military, economic, and cultural domains. I use this concept as the stylized opposite of “hegemonic empire” as defined above. The “Warring States” of fifth- to third-century BCE China and the post-Carolingian states of Latin Europe are the most straightforward historical examples and are frequently referenced in this book. Others that are beyond the scope of my study, such as various city-state cultures from Sumerians to Maya, also qualify.
STEPPE FRONTIER I apply this term to the shifting borderlands and intermediate zones between grasslands that primarily sustained herding and hunting on one side and settled agricultural areas on the other. These frontiers could be wide or narrow depending on local circumstances, and their extent was sensitive to ongoing changes in climate and land use. They served as interaction zones between nomadic or seminomadic pastoralists and sedentary agriculturalists. The most extensive steppe frontier in history was that between the great Central Eurasian grasslands and agrarian regions to the south.
TECHNICAL NOTE TO CHAPTER 1
IN CHAPTER 1, I seek to determine the share of the largest power in the population of a given region at a certain point in time. This approach requires information about the territorial extent of the principal polities and the probable size of their population. The former can readily be obtained from standard historical atlases. Although these may differ in points of detail, such differences are fairly trivial compared to the much greater uncertainty surrounding population number. Specialized historical scholarship, inasmuch as it exists at all, is of highly uneven quality, and so is the evidentiary base (such as it is). For present purposes, the single most important objective is to ensure consistency across time and space in order to facilitate systematic comparison.
The Atlas of World Population History compiled by Colin McEvedy and Richard Jones in 1978 is the only resource that provides the required level of internal coherence and consistency. McEvedy and Jones give estimates of the population of modern countries (or clusters of countries), a few historical entities (such as the Ottoman and Russian empires), and entire continents and the whole world. Their time series commences in 400 BCE and ends in 1950, with projections offered for 2000. The level of resolution is 200-year intervals between 400 BCE and 1000 CE, 100-year intervals from 1000 to 1500, and 50-year intervals thereafter.1
In arriving at their figures, they rely on an eclectic mix of earlier scholarship and they frequently extrapolate backward when there are no data at all. Their extrapolations are undergirded by key assumptions about the overall direction and relative scale of long-term change, which are open to debate. Based on my own research experience, I can say with confidence that their numbers systematically underestimate population density in the ancient world. For example, a tally of 60 million–70 million inhabitants for the Roman empire at its peak in the mid-second century CE is much more realistic than their minimalist estimate of 45 million. The ancient Greek world was also more heavily populated than they allow for.2
Similar problems undermine their results for East Asia. They assume a population of 50 million in China (within its current borders plus Taiwan) in 1 CE, although the Han census for 2 CE refers to 59.5 million, even as not all of China was actually under Han rule and not everyone who was a subject can be expected to have been counted. Their claim of dramatic demographic growth after 1000 stems at least in part from their tendency to underestimate earlier Chinese population size, which then requires an unrealistically rapid catching-up to match documented census totals from the Song period. Their figures for South Asia are inevitably even more conjectural and may likewise underestimate earlier population densities.3
All of this is intentional, given that McEvedy and Jones operate on the assumption of a strong global net increase in population number during the first millennium CE, whereas other scholars favor a boom-and-bust scenario with little if any net increase overall. More recent research lends greater support to the latter view. As McEvedy and Jones correctly observe, their estimates for the second millennium CE are for the most part less controversial and the margins of uncertainty keep shrinking as we approach the present.4
What matters most for us is that insofar as their premodern population figures aim too low, they do so in a consistent fashion. This would be a concern if we were interested in actual population size, but it is of no particular relevance to the computation of population shares as undertaken in chapter 1. Thus, if a given polity covered an area whose population had in fact been 25 percent or even 50 percent larger than McEvedy and Jones believe it was, this undercount would not affect the ratio of imperial to macro-regional population size. I have therefore generally refrained from adjusting their figures so as not to introduce imbalances into their internally coherent edifice of guesstimates.
A more serious problem arises from the need to break down national population totals into smaller elements that can be assigned to imperial states that held only portions of the territory of particular modern countries. In Europe and the MENA region this is a minor issue, given that modern countries there tend to be of fairly moderate size. This makes it relatively easy to come up with plausible fractional estimates, and modest margins of error do not greatly affect overall outcomes for polities that extended across a number of modern states. South and East Asia, by contrast, present us with formidable challenges, dominated as they are by just two massive modern states that form the basic units for McEvedy and Jones’s estimates.5
In those two cases I had to employ (even) cruder ways of apportioning historical population to parts of modern states. For China, I drew on the distribution of Han census results for 140 CE to establish rough proportions for northern and southern China and Sichuan. I also used the present distribution of population among China’s various provinces to extrapolate a gradual shift over time from the ancient to the current pattern. Broadly speaking, northern China, which was demographically heavily dominant in antiquity,
gradually lost out to other areas, above all to the South, which witnessed substantial development especially from the Song period onward. Although it might be possible to devise a somewhat finer-grained adjustment method by taking account of historical census data from the second millennium CE, this is made redundant by the fact that China was for the most part either united or split in two. This pattern greatly facilitates estimation of the population share of the largest polity in East Asia.6
My estimates for South Asia remain much more problematic, both because of the dubious nature of any early population guesstimates and because of our ignorance regarding the spatial distribution of the population in the more distant past. In the most general terms, early India—not unlike China—would have been characterized by greater demographic concentration in the north, in the Indus and Ganges basins, than it is today. In the absence of information comparable to the ancient Chinese census records, there is no obvious way of quantifying this shift even in a crude manner. I have therefore relied on the current distribution of the population of India and Pakistan in order to estimate population shares in the past, an uncomfortably simplifying approach that produces a good fit for the recent past but inevitably becomes increasingly shaky the farther we move back in time.7
This is significant because the largest South Asian empires commonly originated in the north, either indigenously in the Ganges basin (Maurya and Gupta) or as the result of incursions from the northwestern frontier (Saka, Kushan, and a series of Muslim polities). Only some of them arose in central India (Satavahara, Chalukya, and Rashtrakuta), and Vijayanagara and Chola remained the only major empires in the far south. My population shares likely underestimate the relative demographic dominance of northern empires and somewhat inflate the weight of more southerly ones, a problem I duly note in chapter 1. This problem is limited to amplitude and does not affect measurement of the duration of the dominant imperial formations. Thus, although my reconstruction of the South Asian profile ought to be approached with greater caution than that for the other macro-regions, it is nevertheless defensible in its general outlines.
Escape From Rome Page 57