CHAPTER 1 GENERAL MODELS OF WORLD-SYSTEMS Capital imperialism Let us now turn to a discussion of Ekholm and Friedman's (1982) article, "'Capital imperialism' and exploitation in ancient world-systems." The main thrust of this fascinating essay is that world-systems, at least since the emergence of Southern Mesopotamian civilization in the period following 3000 B.C., have exhibited structures and processes strikingly similar to those found in our modern world-system. Many of the ancient world-systems have had a political structure composed of several competing states, city-states, which interacted with one another in a fashion which resulted in something like the hegemonic cycle found in the contemporary global political economy. For long periods the Mesopotamian system remained a world-economy without turning into a world-empire. This early world-system, as well as the better known ancient world-empires, contained a core/periphery division of labor in which core areas exchanged manufactured goods for raw materials and semi-processed goods produced in peripheral areas. They discuss Mesopotamia and other ancient world-systems, as well as the modern one. Let me quote extensively from their article: Center/periphery relations are not necessarily defined in terms of their import-export pattern. Thus, it is unnecessary that a center be the sole locus of industrial manufacturing in the system, or that the periphery be the sole supplier of raw materials. A relation based on a technical division of labor does not correspond to either the mechanisms of development or functioning of global systems. Center/periphery relations refer, rather, to different structural positions with respect to total accumulation. The possession of extremely "valuable" commodities such as silver (Athens) makes it possible to accumulate a disproportionate part of the production of the larger system. If Athens only imported and exported goods, it would never have become a great center. Its accumulation of wealth was, in the first instance, a result of large-scale military tribute-taking and plunder, mercantile profit, and the export of silver. This primary accumulation laid the foundation for a formidable expansion of industrial production. Generally speaking, industrial production is not the means by which centers accumulate initially and accumulation of wealth from the larger system proceeds well in advance of home production. This initial and often continual "primitive accumulation" has always taken the forms of tribute, booty, and enormous mercantile profits. It is usually argued that capitalism can be reduced to the production of industrial capital to the exclusion of all other accumulative activities. Imperialism in such a system is a logically secondary phenomenon related to the needs of self-expanding industrial capital. This construct is opposed to the ancient economy where the struggle for prestige and political power predominated, thus where industrial growth, imperialism, and profit are marginal phenomena. Thus, it is assumed that capitalism is a self-igniting and self-accumulative process while the ancient economy was a more "embedded" system in which production for specific social uses determined the degree and form of growth. This distinction builds upon the subjectivity of the industrial capitalist in one case and on that of the classical Greek and Roman aristocrat in the other. In neither case is the structure of the total reproductive cycle taken into account in the definition. In the modern capitalist mode of production, for example, the accumulation of money capital is not a dependent function of production but rather operates parallel and in contradiction to production. The purpose of production here is the accumulation of money and it is certainly not the only means although it establishes the limit conditions of wealth accumulation. Large portions of the total liquid wealth of capitalist society are invested in "non-productive" and even non-commercial activities. Similarly, while it is clearly the case that the landed aristocracy and, later, the imperial bureaucracy may have been the dominant class faction in the ancient society, their power depended upon the enormous wealth and profit gained in commercial agriculture, and their direct involvement in urban and international commerce, as well as their access to imperialist tribute. This situation is not different in kind from the medieval Arab economy or early modern Europe. It is often overlooked that mercantile Europe operated very much like Rome in its expansion, that it accumulated and "squandered" great amounts of wealth, not primarily by producing, but by pillaging large parts of the globe, and that capitalist production only began within this larger imperialistic process upon which it was materially speaking, entirely dependent. That a capitalist mode of production became dominant in Europe is, of course, related to specific local structures. The emergence of wage-labor in one place and slavery in another is dependent on a difference in initial conditions, but the resultant social forms are worlds apart. Our argument is that the general properties of im- perialist-mercantilist expansion are common to ancient and modern worlds irrespective of specific local forms of accumulation. The growth of industry and commercial agriculture, by whatever form of exploitation, occurs within an already constituted imperialist structure and is not a local and closed process (emphasis in the original, 1982: 90-92). The most controversial aspect of the Ekholm-Friedman essay is their claim that some of the core states of the original Mesopotamian world-economy had social systems in which oligarchies somewhat autonomous of the state accumulated capital wealth based on private property. This contention is based on recent research which challenges the "Temple economy" thesis first developed by Anton Deimel and employed in Polanyi's (1957a) considera- tion of Mesopotamia. According to Polanyi, the Sumerian economy was a non-market redistributive system in which surpluses of food and prestige goods were collected, stored and distributed by a centralized religious bureaucracy. According to Ekholm and Friedman, this kind of economy did characterize the earliest states of the Early Dynastic period, but later differentiated into a system in which authority was shared with a secular palace specializing in political-administrative, trade and warfare functions, and an aristocracy, partially autonomous from both temple and palace, which accumulated capital wealth based on private ownership of land and slaves. Production became divided into private and public sectors, and a merchant class emerged to monopolize inter-regional circulation, accumulating a substantial portion of the total wealth through mercantile profits.2 The forms of labor exploitation in this differentiated system included "helots" directly dependent on the state, "private and state slaves, and a "free" population exploited by taxation. There is also evidence, perhaps in the free sector, of contract or wage-labor connected with skilled or more specialized tasks" (Ekholm and Friedman, 1982: 99). Ekholm and Friedman are careful not to argue that all world-systems are the same. We do not deny that there are important differences between industrial capitalism and the ancient systems. It is clear that the modern system in which industrial capitals compete for survival by direct investment in the productive forces implies a kind of dynamic unknown in the past. The accumulation of capital as a form of abstract wealth, however, is a truly ancient phenomenon. To say that this ancient "capital" played a fundamental economic role is not to say that it functioned directly in the production process, but that its accumulation and control were dominant features of those economies (1982: 88). And yet theirs' is a generalized model which is asserted to describe the overall structural parameters of all world-systems. Before describing some other general models let us describe several interesting hypotheses mentioned by Ekholm and Friedman. Based on their comparison of the Southern Mesopotamian city-states interacting within a world-economy and Ancient Egypt, an empire in which a single state encompassed the economic network, Ekholm and Friedman suggest that: 1) a less dense economic network is easier to centralize, and conversely a more dense network which is not easily isolated from outside contacts is harder to centralize; and 2) a less centralized and more competitive world-system is more likely to rapidly develop technology, especially military technology. They compare Egypt's slow innovation rate with the rapid development of warfare techniques in Mesopotamia. This point has been made also with respect to other kinds of innovation by other authors. Anderson (1974b) argues that feudal Europe's "parcellization of sovereignty" allowed capitalist cities, especially in Northern Europe to develop polities under the control of merchants and guilds, a feature which was propitious for the development of productive technology. Wallerstein (1974) argues that it was Europe's decentralized interstate system which allowed the Portuguese expansion around Africa to the Indies, whereas China's more centralized empire, concerned with maintaining an imperial state and the boundaries of a large empire, quashed such expansive tendencies when they developed during the Ming dynasty. Of course it is evident that decentralization is not always an advantage, and so we must try to determine under what conditions it leads to greater innovation. Ekholm and Friedman also contend, based on their study of Southern Mesopotamian city-states, that differentiation of authority which allows resources to escape the control of a state can lead to disintegration and conquest by an alien power. This contention is supported by those (Hawley, 1981) who understand Rome's "time of troubles" as a breakdown of central control over regional outposts in conjunction with rising conflicts over the control of the central state. And another hypothesis of Ekholm and Friedman is that successful emergence as a core area producing manufactures for export to peripheries requires prior primary accumulation of resources through conquest, pillage and plunder. Schneider (1979) points out that gold and silver, rather than being irrelevant luxuries, were important means for the mobilization of armies and the waging of wars in the precapitalist world. Successful core areas were ones that could develop manufactures (especially textiles) for export to draw precious metals toward the center, and the transition from periphery to core required the development of mercantilist policies which would prevent the importation of textiles from older cores as well as the export of precious metals. These hypotheses will be discussed again below, but here I would like to describe some other general models of evolution of civilizations and the dynamics of the rise and fall of empires. Other general models At the most general level are the evolutionary ideas of Talcott Parsons (1966, 1971) who described the development of social systems from small undifferentiated societies in which kinship institutions carried out all societal functions to modern large-scale complex social systems in which societal functions are differentiated into specialized institutions. The process of development is said to procede through stages of differentiation followed by strains among the differentiated subunits which are eventually overcome by the emergence of new specialized integrating institutions, especially the state, but also the generalized or "universalistic" value systems embodied in world religions. While Parsons' description of this evolutionary sequence is pacific, a more historical and conflictive version with elements of class conflict and devolution is presented in Arnold Toynbee's works (Tomlin, 1978). Toynbee contrasts the successes of grand organization in the material world of power with the more sublime development of universal religious ideas, which often emerge and spread during periods when secular power is disintegrating. Sorokin (1941) also distinguishes between "sensate" periods of civilization and "ideational" or "idealistic" periods in which spiritual values come to the fore. While it may be au courant to denigrate these "idealist" analysts of evolutionary civilization it should be pointed out that they do deal with social change on a scale which at least makes possible the understanding of world-system transformation. Parsons' theory, while it may ignore much which we consider historically central about how change occurs, does paint with broad strokes the main changes which have occurred in human society over the past nine millenia. And Toynbee's work, much more attuned to the problematic and conflictive nature of growth and decline, emphasizes the role which religious value systems have played in the integration of world-systems, a matter which has been neglected by most, but not all, historical materialists. Perry Anderson's (1974a) poignant metaphor, that the Christian church was the vessel which carried the cultural heritage of the ancient world across the sea of the dark ages, suggests that we ought not ignore Toynbee, Parsons, or the functioning of ideological products in the transformation of world-systems. Toynbee claims to observe a pattern in which value-generalizing world religions emerge during the disintegrating phase of a civilization and become institutionalized in the civilization which emerges next. Gerhard Lenski's ecological-evolutionary scheme (Lenski and Lenski, 1982) focusses on the importance of productive technology for the development of complex societies which are able to better adapt to the natural and social environment. Lenski begins with a comparison between biological and social evolution,3 in which the information storage capacity of human culture is compared to the much less flexible storage capacity in the genetic structures of non-human life forms. Social insects are hard-wired, operating on genetically programmed instincts, while homo sapiens evolved a great capacity for individualized learning in the uncommitted cortex of the human brain. The invention of written language extended the possibility of collective memory and facilitated the specialization of individuals in particular kinds of knowledge. Our libraries are our collective information gene pools, and for Lenski it is the techniques of material production which are the most important information stores. Lenski's typology of societies focusses on productive technology. Human societies are described as having modes of production based on hunting/gathering, simple and complex horticulture (agriculture without the plow, irrigation or animal energy), agrarian production (plow, metals, irrigation, and animal energy), and industrial production (machine production, non-human, non-animal energy sources). His theory is not unilinear in the sense that individual societies are thought to develop through a series of stages. It allows for branching specialized types (fishing, pastoral, maritime) and views evolution as a competitive process of borrowing, innovation and conflict in a world system arena which has been characterized by interaction among societies for several millenia. Elman Service (1975) has analyzed the emergence of states and civilization in the six locations where primary states developed: Mesopotamia, Egypt, China, the Indus River valley, the valley of Mexico, and Peru. He disputes a number of prevalent notions about how these instances of primary state formation came about. He notes for example that "hydraulic civilization" seems more a consequence than a cause of primary state formation. He critiques Engels' (1972) contention that the state emerged to protect the interests of merchants by pointing out what Polanyi and most other scholars have noted: in the primary states it was at first the state itself that was the organizer of the economy and long distance trade, and thus there were no pre-existing merchants who needed state protection. Service's own theory is a version of the functional theory of stratification. He argues that chiefdoms emerged to organize collection and storage of a surplus to tide the population through hard times. Some of these chiefdoms later evolved into states engaging in redistributive centralization and mobilization of resources for religious and military purposes. Service (1971) propounds a general theory of the rise and fall of core states which he applies to ancient empires as well as modern nation-states. His theory utilizes the notions of adaptation and adaptivity. He notes the pattern in which ancient empires rise and then are taken over by "marcher states." He argues that the original cultural adaptation which allowed the empire to rise becomes rigidified in the institutional structure of the empire, and this reduces the adaptivity of the empire. Newer innovations are adopted by competing centers which are then able to conquer the original empire. This general notion is applied by Service (1971) to modern nation states in which economic competitive advantage is thought to create rigid structures in core states, which allow other competing states to adopt newer technologies, leading to the hegemonic cycle. This is a generalized understanding of uneven development similar to Gerschenkron's (1962) discussion of "the advantages of backwardness." Rein Taagepera (1981: 12-14) posits a "deterioration" model to account for the patterns he discovers in the size and duration of empires. He assumes that "any self-sustaining system produces deterioration within its own structure. The gradual accumulation of deterioration eventually brings about a slow decline or a sudden collapse. During the growth period, new healthy components are being created, and the system continuously outgrows its defects. When growth stops accumulation of deterioration soon takes over... The nature of deterioration may be varied and complex. The very factors which enable growth to occur in the first place may be destroyed by successful growth. Thus the remarkable performance of the Roman army of citizen- farmers led to the demise of the small farms and a shift to mercenary armies. Secondly, a factor contributing to success at small size may itself become dysfunctional at a larger size. Strong personal and non-delegatable leadership might be an example. Thirdly, the very fruits of a successful expansion may be poisonous in a figurative or even a literal sense. The enriched Roman elite could afford the luxury of leaden tableware which gradually poisoned them" (1978: 13-14). The debate between the "formalists" who analyze historical economies in terms of supply and demand, and the "substantivists" who focus on the institutional and social embeddedness of economic processes in precapitalist societies has waxed and waned and waxed since Karl Polanyi (1944, 1977) and his associates (Polanyi et al. 1957) first argued the substantivist position and supported it with comparative research. I will comment on this ongoing debate as we analyze various types of world-systems. Here I will summarize the theoretical positions. Polanyi's work propounds a number of types of exchange which are present in different degrees in all societies. He argues that these types of exchange are forms of social integration whose importance differs across different types of societies. The first of his modes of integration is reciprocity, based on normatively consensual rules of symmetrical exchange, which is the dominant form of integration in kinship-based simple societies. Redistribution is the most important form of integration in the ancient empires, in which central state institutions extract a designated surplus product from direct producers and determine the uses of the surplus. The last form of integration is the price-setting market in which the inter- action of a large number of competing buyers and sellers, trying to maximize their individual gains, determines the relative rates (prices) at which different commodities exchange. Polanyi's most important effort was to distinguish between these very different modes of exchange and to argue against the error of projecting our modern "market mentality" back upon pre-capitalist societies. As George Dalton (1975) has pointed out, Polanyi's work does not contain a theory of how these types of integration evolve or change. Polanyi's approach and his interpretation of particular societies have been criticized on many points. Forms of exchange which appear to be responding to supply and demand have been discovered in many precapitalist societies, and a debate has raged over the importance of these for system logic and system change. Philip Curtin's (1984) historical study of trade diasporas, while criticizing the substantivist tradition at many points, does not formulate its own theory of development. Yet, we can impute a theory which is a modified version of formalism. Trade diasporas are ethnic or culturally solidary groups which specialize in trade across cultural boundaries. Beginning with his own area of specialization, Africa, Curtin analyzes how various ethnic groups establish communities across a wide area and specialize in trade which crosses the boundaries of other cultures. He then examines trade diasporas in the ancient world, the non-European medieval world, and the overseas trading empires of the European states. Curtin's main point is that mixtures of reciprocity, redistribution and market exchange characterize all social systems, although he does not examine the question of system logic or transformation. He argues that trade diasporas are necessary to exchange across cultural boundaries before an "ecumenical" culture of trade has emerged. He disputes the status of Polanyi's categories as an evolutionary sequence. The Marxian tradition of historical materialism has so many sub-tendencies that it is difficult to summarize. Marx's own analysis focussed primarily on the unique nature of capitalism as a mode of production and its emergent dominance. His comments on precapitalist modes of production were few, although some subsequent Marxists have converted these into a general theory of historical development. Marx, of course, focussed on class relations as the key to a society's logic. The institutions which, in class societies, allow a surplus product to be extracted from direct producers (relations of production), are thought to be the most important determinants of societal logic. Thus class societies can be categorized in terms of slavery, serfdom or wage-labor. His notion of a stagnant Asiatic mode of production, in which self-subsistent villages were largely unaffected by their payments of taxation or tribute to an overarching state, has been critiqued and abandoned by many Marxists. Marx's adoption of Hegel's dialectical model has been applied by many historical materialists to the logic of social change in precapitalist as well as capitalist societies. This model posits that a social system contains internal dialectical contradictions which eventually create a crisis, leading to the emergence of a new qualitatively distinct kind of system. These internal contradictions are thought to be the most important determinants of system logic and the potential of a system to transform. External social and environmental conditions are said to affect only the timing of the realization of a system's potential, not the nature of change. Mao's (1967) specification of the dialectical principle of contradictions and synthesis has been one of the most influential interpretations. Samir Amin (1980) is the only Marxist to combine an explicit analysis of modes of production with world-system concepts such as center, periphery and uneven development, and to apply this analysis beyond the confines of the rise of the West. Eric Wolf (1982) uses a similar approach to modes of production, but he does not explicitly use world-system concepts. Wolf's analysis of the interaction between precapitalist modes of production and European capitalism since 1400 A.D. is a valuable addition to our under- standing of the articulation of modes of production, but Wolf does not analyze the history and evolution of world-systems before 1400 A.D. We shall discuss at various points below how Amin and Wolf differ from others in terms of their definitions of modes of production. Here I want to mention that Amin applies the notion of uneven development to an important instance of the transformation of a mode of production--the emergence of dominant capitalism in Europe. According to Amin, feudal Europe is a peripheral area of a larger Eurasian world-system. European feudalism is an "incomplete" formation which is transition~~al between communal (or kinship-based) modes and the fully developed tributary mode of production. This relatively decentralized mode of production proves to be fertile ground for the development of capialism, according to Amin. Weber's (1976) comparative historical work was not formulated as general theory. Rather he compared the institutions of property and politics in different societies for the purpose of understanding the development of rationality which he saw as central to capitalism. His focus on political organization is reflected in recent interpretations of evolution, especially Service's (1975) focus on the development of authority. William McNeill's (1982) work on military technology and organization analyzes the major transformations of empires and European dominance in terms of breakthroughs in the art and science of warfare, supplemented by demographic changes resulting from the application of public health measures and increases in food production. A similar approach is taken by Taagepera (1978b) who attributes the rapid enlargement of empires which occurred around 600 B.C. to the development of techniques for delegating authority to regional representatives of the central empires. According to Taagepera this break-through occurred first in the Median- Persian empire, and then was adopted by the Han Chinese, Roman and Parthian-Sassinid empires. A similar phase of empire enlargement occurred in recent centuries as a consequence of European development of more rapid communications and transport, according to Taagepera. Most of the above theories were developed to account for the expansion and transformation of empires and societies, but with a little imagination they can be applied to Wallersteinian world-systems. Before formulating my own theory of world-system transformation I will employ a taxonomy of world-system types to investigate the processes which seem to operate in different kinds of world-systems. This exercise is an attempt to say what is particular about different kinds of world-systems. It utilizes the archeological, ethnographic and historical literatures about societies, intersocietal exchange and political systems, even though most of this literature was not written with world-systems as the unit of analysis. A taxonomy of world-systems We can synthesize a taxonomy of world-systems by focussing on the social structural characteristics of societies and the description of inter-societal relations in the existing literature. Without pretending to take account of all the cultural elements of societies, or even of all the variation in social structural elements, it is possible to suggest broad categories which will be useful in discussing the generality or particularity of different kinds of world-systems. I propose to describe the functioning of the following types of world- systems: 1. stateless, classless world-economies 2. primary state-based world-economies 3. primary world-empires 4. complex secondary world-systems 5. commercializing world-systems 6. the capitalist European subsystem