Apart from a few stray generalizations, I have confined myself to a single millennium of Near Eastern history. Yet a body of literature within comparative sociology generalizes about historic empires located over the whole globe and throughout the five millennia of recorded history. To do so requires that
B. C.
Key
• Power of central state
----Power of decentralized elites
Figure 5.1. Dialectics of Mesopotamian empires
Empires shared broad similarities that shine through the many and variegated differences of time and place. “Is it not amazing,” asks John Kautsky rhetorically,
That there should be substantial similarities between Assyrians, Almoravids and Aztecs, between the empires of the Macedonians, the Mongols and the Moguls, between Ostrogothic kings, Umayyad caliphs and Ottoman sultans, between the Sassanid, Son-ghay and Saudi empires, between Ptolemies, Teutonic Knights and the Tutsi, between the Vandals, the Visigoths and the Vikings? [1982: 15]
Kautsky observes acutely that the basic similarity enabled conquerors like the Romans or the Spanish conquistadores to exploit politically the weak spots of their apparently “alien” opponents - for they recognized their power structure.
1 do not dispute Kautsky’s essential argument. This type of comparative sociology has established points of similarity between such varied regimes. I present three of these before turning to the principal defects of the model - a neglect of history, an inability to produce a theory of social development, and a failure to recognize dialectical processes.
The first point of similarity between such regimes is that, as Kautsky labels them, they were “aristocratic empires.” They were dominated by a ruling class that monopolized landownership (sometimes in the sense of effective possession, rather than legal ownership) and so controlled the economic, military, and political power resources that land provided. And ideologically, their dominance was expressed through genealogical claims to moral and factual superiority - an aristocrat was superior because, through birth, he (or she) was connected to an endogenous kin group stretching back to an original ancestor group that founded the society, were descended from heroes or gods, or performed some other noble feats. With its hands firmly on all four sources of social power, the class was so entrenched that no ruler could dispense with its support. This is worth stating simply and forcibly because many of these regimes made a contrary ideological claim, namely, that all power flowed from it and it alone, and also because some writers have been taken in by the claim. Sargon’s grandson, Naram-Sin, claimed divinity. His Akkadian or Sumerian aristocrats only claimed genealogical connections to the divine. This became a standard pattern for the more pretentious empires of history up to the modem period. It justified the personal despotism of the mler, which in theory was exercised no less over the aristocrats than over anyone else. Some of the more credulous writers have believed that this could lead to actually “absolute” rule. This number included Wittfogel, whose theories I dismissed in Chapter 3, as well as a few other comparative sociologists (e. g.. Wesson 1967: esp. 139-202). In practice, however, such regimes were feeble.
It is helpful at this point to distinguish between two types of state power. I make the distinction more fully in Mann (1984). Despotic power refers to the range of actions that the mler and his staff are empowered to attempt to implement without routine, institutionalized negotiation with civil society groups. A supreme despot, say a monarch whose claim to divinity is generally accepted (as in Egypt or China throughout much of their imperial histories) can thus attempt virtually any action without “principled” opposition. Infrastructural power refers to the capacity to actually penetrate society and to implement logistically political decisions. What should be immediately obvious about the despots of historic empires is the weakness of their infrastructural powers and their dependence upon the class of aristocrats for such infrastructure as they possessed. For many purposes, and especially in the provinces, their infrastructure was the aristocracy. So, in practice, empires were “territorially federal,” as 1 expressed it - looser, more decentralized, more prone to fission, than the state’s own ideology usually claimed.
All these points stemming from the first regime similarity have been made often enough, using slightly different terminology, in recent comparative sociology (see, e. g., Bendix 1978, Kautsky 1982).
The second regime similarity leads to a rather different emphasis, however. In emphasizing the power of the aristocratic class, we should not lose sight of the fact that a state still exists with power resources of its own. States exist because they are functional for social life beyond a fairly simple level. It is more relevant to the present issue that they provide something that is useful to the aristocratic class. This is territorial centralization. A number of activities, such as judicial rule making and enforcement, military organization, and economic redistribution, were usually more efficiently performed at this level of historical development if centralized. This central place is the state. Thus any autonomous power that the state can acquire derives from its ability to exploit its centrality.
This has been explored by Eisenstadt (1963). Following Weber’s lead, he argued that the imperial state claims universalism and that this claim has some actual grounding in fact. A state cannot be merely aristocratic. Genealogical claims are inherently particularistic; they are the anthithesis of centrality and of the state. Societies that develop permanent states have already proceeded farther than particularism. They have rationalized the symbolic sphere and begun to conceptualize the cosmos as subject to general forces with universal impact. The state, not the aristocracy, expresses this rational divinity. Materially, further argues Eisenstadt, the state’s interests lie in fostering “free-floating resources,” resources that are autonomous of any particularistic power actor. Eisenstadt instances many of these, and I will return repeatedly to them in the course of my historical narrative. The most striking (especially to the person concerned!) is the use of eunuchs by the state. As I have emphasized, any of the state’s agents may “disappear” into civil society, escaping the control of the ruler. One way of stopping an agent from disappearing into the aristocracy is to prevent genealogical issue by castration.
Of the universalizing techniques of early states glimpsed in this and the last chapter, let me pick out three. First, in the realm of ideology, comes the attempt by the Akkadian conquerors to rationalize and systematize the pantheon and the creation myths of the Sumerian cities. Under the Akkadian Empire a “religion” is written down, codified, and given hierarchy and centrality. Second, in the realm of material infrastmcture is the attempt (or at least the claim) by Sargon and his imperial successors to have improved and coordinated as a whole the communications structure of the empire. These are not just power-enhancing actions: They attempt to universalize power, and, consciously or not, their force is to reduce the power of local, particularistic elites. Third, and perhaps the best example because it combines ideology with infrastructure, is the “decimal” administrative structure imposed by the Inca conquerors on the Andean peoples (referred to in Chapter 4). In practice, of course, the Inca could only rule conquered provinces through local, indigenous elites. They might impose alongside them an Inca governor, import some loyal settler-soldiers and build roads, storehouses, and relay-posts - indeed, no conquerors were more ingenious in such respects. But they could not overcome those brute logistical problems of rule that I have outlined in this chapter. Hence the significance of the decimal rationalization. Its ideological function and perhaps, to a degree, its actual effect (though the con-quistadores exposed its weaknesses), was to say to the local elites: “Yes, you may continue to rule your people. But remember that your rule is part of a wider cosmos that subordinates tribal and regional particularisms to a rational Inca order, centered on the Lord Inca himself.” It reflects great credit on Eisenstadt to say that if the Lord Inca or Sargon or the Chinese or Roman emperor were to return and read his book, they would recognize his characterization of their policies and know what was meant by universalism, free-floating resources, the rationalization of the symbolic sphere, and the other pieces of Eisenstadt jargon.
I have drawn two insights from comparative sociology: on the one hand, a socially useful, despotic, universal state; on the other, a decentralized, particularistic aristocracy in actual possession of much of the power infrastructure of the society. The contrast between the two means that comparative sociology has also provided a third insight, a clear exposition of the contradictions, and sometimes of a part of the dynamics, of such regimes. For there was a continuous struggle between the two, mitigated only (but most substantially) by their mutual interdependence in order to preserve their exploitation of the mass of the population. The most famous discussion of the struggle was presented by Weber in his analysis of patrimonialism in Economy and Society (1968: III, 1006-69).
Weber distinguished patrimonialism and feudalism as the predominant types of political regime in preindustrial civilized societies. Patrimonialism adapts an earlier, simpler form of patriarchal authority within the household to the conditions of larger empires. Under it, government offices originate in the ruler’s own household. This continues to provide the model even where the official function has little connection with the household. For example, the cavalry commander is often given a title, like “marshal,” which originally denoted supervision of the ruler’s stables. Similarly the patrimonial ruler shows a preference for appointing members of his own household, kinsmen or dependents, as government officers. The ensuing rule is autocratic: The ruler’s authoritative commands assign rights and duties to other persons and households. Sometimes associations of persons and households are designated by the ruler as collectively responsible for rights and duties. By contrast, feudalism expresses a contract between near equals. Independent, aristocratic warriors Ireely agree to exchange rights and duties. The contract assigns one of the parties overall political rule, but he is restrained by the terms of the contract and he is no autocrat. Weber distinguishes these two forms of rule as ideal types and then proceeds in his characteristic fashion to elaborate the logical consequences and subdivisions of each. But he also notes that in reality ideal types become blurred and transform one into the other. In particular he acknowledges the logistical impossibility in preindustrial conditions of a “pure” patrimonialism. The extension of patrimonial rule necessarily decentralizes it and sets in motion a continuous struggle between the ruler and his agents, now become local notables with an autonomous power base. It is exactly the kind of struggle I have described in Mesopotamia. Weber details examples from ancient Egypt and Rome; ancient and modem China; and medieval Europe, Islam, and Japan. His analysis has so influenced historical understanding that it looms large in modem scholarship on all these cases and more.
Approximations to the ideal-type regimes, plus mixed cases, have dominated much of the globe. The stmggle between centralized, patrimonial empires and decentralized, loosely feudal, aristocratic monarchies constitutes much of the history recorded by contemporaries. But if this were all of our history, even all of our upper-class history, it would be essentially cyclical, lacking long-term social development. In this chapter I have tried to add something else: an understanding of how such a stmggle continuously revolutionizes the means of power and so constitutes a dialectic of development.
Perhaps it is open to misinterpretation to accuse Max Weber of lacking interest in historical development, since he concerned himself with this more than has any other major sociologist of the twentieth century. But his use of these ideal types was at times static. He contrasted East and West, arguing that massive social development occurred in Europe, rather than the East, because it was dominated by a contractual, decentralized feudalism, which (in contrast to Eastern patrimonialism) fostered a relatively rational spirit of acquisition and an activist orientation to the conquest of nature. In his view a relatively feudal, or at least decentralized, stmcture must be in place before dynamism can occur. This is incorrect. As we shall see repeatedly, it is the dialectic between the centralizing and decentralizing that provides a considerable part of social development, and this has been especially pronounced in the history of the Near Eastem/Mediterranean/Westem world.
The subsequent development of neo-Weberian comparative sociology has become more static. Whatever the insights provided by Bendix, Eisenstadt, Kautsky, and others, they neglect development. To concentrate, as Kautsky does, upon the similarities of regimes such as the Inca Empire and the kingdom of Spain (both “aristocratic empires”) is to forget what happened when 180 Spaniards entered an Inca Empire of millions. The Spaniards possessed power resources undreamed of by the Inca. Those resources - body armor, horses, gunpowder; the military discipline, tactics, and cohesion to use these weapons; a Salvationist, literate religion; a monarch and a church able to enforce commands over six thousand kilometers; a religious/national solidarity able to overcome differences of class and lineage; even their diseases and personal immunities - were products of several millennia of world-historical development denied to the American continent. We will see the resources emerge gradually, unsteadily, but undeniably cumulatively over the next six or so chapters. Comparative sociology must be restrained by an appreciation of world-historical time.
Thus, when neo-Weberian analyses come to explain social development, they look outside of their theoretical model. Kautsky regards “commercialization” as the main dynamic process. It emerges, he says, through towns and traders who are largely outside the structure of “Aristocratic Empires,” and whose emergence he cannot therefore explain. Bendix, whose aim is to explain the transition from monarchy to democracy, also turns to extraneous factors. In his case they are a number of unexplained, independent variables like population growth; technological changes; and the growth of towns, communications infrastructures, education systems, and literacy (1978; esp. 251-65). Eisenstadt has a more adequate model for understanding social development. In a few pages (1963: 349-59) he outlines how a few empires were transformed into more “modem” polities and societies. For him the decisive factor was the ability of various decentralized elites, supported by rational, Salvationist religion, to appropriate the universalism and free-floating resources hitherto monopolized by the state. As we shall see in later chapters, this is indeed an important part of the answer. But after 350 pages delineating a static or cyclical model of empire, he can hardly go far in this direction in 10 pages. All these works (as with most comparative sociology) combine promiscuously material gathered from different phases in the development of social-power resources. This is their greatest weakness, for it is often the very thing they are supposedly trying to explain.
My criticism of the methodology of the comparative sociology of ancient empires is not the “typical historian’s” objection that every case is unique. Though this is tme, it does not preclude comparison and generalization. It is rather that comparative analysis should also be historical. Each case develops temporally, and this dynamic must itself be part of our explanation of its stmcture. In the present case, the dynamics of “imperial” (or “patrimonial”) and “feudal” regimes have constituted a dialectic of development, ignored by comparative sociology.
World History
