The Changing Role of Cities in World-Systems


Christopher Chase- Dunn

Institute for Research on World-Systems

University of California, Riverside

(v. 8-2-04)



An earlier version was published   in Volker Bornschier and Peter Lengyel ( eds .) World Society Studies, Volume 2. Frankfurt and New York: Campus, 1991.


The purpose of this essay is to examine the relationships between human settlement systems, intersocietal networks (world-systems) and transformations of modes of production. What has been the role of cities with regard to changes in the basic logic of social reproduction? I will examine how the roles of cities and systems of cities have changed depending on the nature of the world - systems in which they are embedded.

After introducing several theoretical problems that motivate the comparative study of world -systems I will discuss:

·        a class of world-systems without cities ,

·        the rise of the first cities and settlement systems,

·        interstate systems composed of city-states,

·         the effects of empire-formation on cities,

·        the effects of capitalist city-states on tributary empires,

·        the role of cities in the rise of Europe and the emerging predominance of the capitalist mode of production,

·        the subordination of cities to national states and capitalist firms in the capitalist world -economy, and


I do not believe that urban sociology or the study of cities ought to constitute a theoretically separate topic from the general study of social change. The consideration of urban-rural relationships, larger societal structures, core/periphery relations, and whole world -systems are necessary to the understanding of the development of cities. But the study of cities does provide an interesting angle from which to analyze whole socio-economic systems and the historical processes by which the deep structural logics of systems become transformed. I will choose as my empirical scope for comparison, not the contemporary global political economy, nor even the expanding Europe-centered world-system over the last 500 years, but rather the set of sedentary small, medium-sized and large world -systems that have been on Earth over the last ten thousand years.

There are three justifications for this unusually broad scope of comparison. It is possible that the comparative study of whole intersocietal systems will produce a more powerful theory of social evolution than has emerged from the more traditional focus on societies. The broad temporal scope of comparison maximizes variation in the structural characteristics of world - systems. And it allows for empirical study of fundamental watersheds in the transformation of systemic logics: We can analyze how tributary modes of production[1] emerged out of kin-based modes, and the processes by which the capitalist mode became predominant over tributary modes. This scope of comparison may also be useful for helping us to understand contemporary and future processes by which systemic logic may fundamentally change.

I will be employing a set of theoretical distinctions about modes of production[2] developed by Amin (1980) and Wolf( 1982) , and supplemented to some extent by the approach to forms of social integration developed by Karl Polanyi(1977) . To summarize briefly, I distinguish between four classes of systemic logics that are designated as:

1. kin-based modes of production (in which social labor, distribution, and collective accumulation is mobilized by means of normative integration based on consensual definitions of value, obligations and rules of conduct --amoral order).

2. tributary modes (in which social order is produced by politically institutionalized coercion based on codified law and formally organized military power).

3. capitalist modes (in which land, labor, wealth and goods are commodified and strongly exposed to the forces of price-setting markets; and the logic of accumulation through the production of commodities using commodified labor is predominant) , and

4. socialist modes (in which major policy and investment decisions are controlled democratically by the people they affect according to a logic of collective rationality)


            The world-systems perspective has stimulated a new approach to the understanding of capitalism, one which emphasizes the necessity of peripheral forms of capitalism, the importance of the interstate system, and the various forms and degrees of the commodification of labor within the capitalist world - economy (Chase-Dunn 1989) .The application of the world -systems perspective to pre capitalist systems raises new questions and reopens old debates about other modes of production and their transformations.[3]

Since Karl Polanyi and his colleagues developed the substantivist approach to economic sociology in the 1950s a debate has waned and waxed among archaeologists, anthropologists, historians and sociologists over the "substantivist/formalist" question. The substantivists argue that exchange relations are embedded in social structures, and that markets are historically created institutions, not timeless logics expressing the truck-and-barter instincts of "economic man. " Polanyi distinguished between three qualitatively distinct forms of integration: reciprocity, redistribution and market exchange.[4] The formalists argue that economic rationality has similar properties in all human societies, and they emphasize the importance of rational choice approaches for nomadic foragers as well as contemporary consumers.[5]

 A related debate over the similarities and differences between modern, classical, and ancient societies has raged among "primitivists" and "modernists”[6] The modernists argue that economic development in the classical or ancient worlds already involved commodified relations and processes of economic development quite similar in their basic nature to modern societies ( e. g. Rostovtzeff, 1941).  The primitivists emphasize the importance of differences in the logic of competition, the rationality of accounting practices , the nature of taxation, forms of property, types of labor control, etc. between modern and classical societies (e .g. Finley, 1973) .

These old debates have received a new incarnation in the context of the effort to use the world-system perspective to understand precapitalist social change. The earliest and most influential formulation of the dynamics of ancient and classical world-systems argues that these share a common "capital- imperialist" mode of production with the modern world-system (Ekholm and Friedman, 1982). More recently Gills and Frank (1991) contend that accumulation occurred historically within the context of a continuous world -system over the last 5000 years, and they explicitly reject the idea that transitions have occurred between qualitatively different modes of production. Though these scholars focus explicitly on a new unit of analysis, whole world-systems, their theoretical contentions are similar in form to those of the formalists and modernists mentioned above. On the other side are those world-system Marxists and recalcitrant Polanyians who insist that qualitative transformations have occurred in the development of world-systems .The analysis of the growth of cities and systems of cities is germane to the many issues which these contending perspectives raise.


World-Systems Without Cities

Does it make sense to think of the interaction networks among stateless, classless, city-less societies as constituting world-systems ? Certainly these were very different from the modern world -system. Are they so different as to be non-comparable? I will argue that most of the very small intersocietal systems composed of unstratified societies fit the definition of a world -system based on the necessity of the whole for the parts. And I will contend that it is important to study this class of world -systems because they provide instances of historical transformation of one mode of production into another. We need to understand how kin -based systems were reproduced in order to analyze how classes, states and cities emerged and transformed the kin-based mode into a tributary mode.

Wallerstein (1984: Chapter 14) has classified stateless societies as "minisystems" --not world-systems --in which each cultural entity encompasses an autonomous system of the production of basic goods. It is implied that these are non-comparable with world-systems, which are defined as a division of labor of the production of basic goods (bulk foods and raw materials) which encompasses a set of separate cultures. In other words, cross-cultural trade of basic goods is the main criterion in Wallerstein's distinction between world -systems and minisystems. I contend that, though trade of basic goods is an important aspect of interaction for defining and bounding world -systems, it is not the only important kind of interaction that constitutes systemness. It can be demonstrated that many stateless intersocietal systems fit Wallerstein's definition of a world-system (based on cross-cultural bulk-goods exchanges) , and manyothers fit a somewhat broader definition proposed by Chase-Dunn and Hall (1991) which includes political/military competition and prestige goods exchange along with trade in basic goods as important intersocietal interactions which produce systemness .

There are several fundamental issues that arise when we extend world- system concepts to stateless societies. In many such systems it would appear that there is little in the way of core/periphery hierarchy. In order to understand why core/periphery hierarchies emerge and are reproduced it is desireable to compare systems that have such structural features with those that do not. To explore these conceptual and comparative problems further I have undertaken to study a very small and very different system, that of "precontact" Northern California (Chase-Dunn and Mann 1998). Though one case study cannot be the arbiter of such basic theoretical problems, I believe that this study will show that a broadly construed world -system perspective can be fruitful for understanding the reproduction and historical development of social structures in an intertribelet network of sedentary foragers. Other studies already confirm that intersocietal relations are important for reproduction and transformation in somewhat more complex stateless societies.[7] 

Though there are no cities in these systems, there are settlements and socially constructed space.[8]  Chase-Dunn and Hall (1997) argued that we should begin world-systems analysis with the emergence of sedentism. But nomads do not wander aimlessly across space and so it does make sense to study world-systems composed solely of nomads. The yearly migration cycles of nomads reflect both the seasonal location of resources and interactions with other groups of humans. The sequence in which larger yearly migration routes become gradually smaller, and the emergence of regional differences in Paleolithic toolkits shows that the same processes of growing population density that affect sedentary peoples are also operating on nomads.

Once some of the actors are territorial it is no great leap to study intergroup interactions to see if they are indeed systemic or not. In California, sedentary foragers built villages and defined communally- held territory with definite boundaries between adjacent groups. The regulation of land use was an extremely important aspect of intergroup relations because foraging is space-extensive and resources can easily be overexploited. My analysis of this little world-system will focus on whether or not, and how, intervillage and intertribelet relations are important for the reproduction and/ or transformation of local social structures and forms of resource use. My study will also search for evidence of unequal exchange and domination among tribelets and villages. If none is found I will try to understand how such an regionally-egalitarian world -system was reproduced .

I have also studied more hierarchical stateless systems, such as that of the “precontact” Hawaiians (Chase- Dunn and Ermolaeva 1994). Here was a system in which classes had emerged and hierarchical accumulation was occurring, but there was no state in the sense of a specialized organization exercising regional control that is substantially autonomous from kin-based obligations. Johnson and Earle (1987) classify the Hawaiian system as composed of  “complex chiefdoms” in which the polities are more hierarchical and larger than “big man “ or simple chiefdom systems. Class exploitation in Hawaii was based on the control of land and other resources by means of hierarchical forms of kinship. The work of Gailey (1987) examines the interaction of gender hierarchies and forms of marriage that are involved in the emergence of more hierarchical class societies. Friedman (1982) and Friedman and Rowlands (1977) add to this the consideration of relations among societies, especially intermarriage, warfare and prestige goods economies.

These authors have modified Leach's (1954) work on the alternation between more and less hierarchical forms of social organization to explain the emergence of class- based accumulation and intersocietal hierarchies. The processes of the rise and fall of chiefdoms[9] in interchief systems are roughly analogous to sequences of political centralization/ decentralization which occur in larger-scale world-systems --state formation, empire formation and the rise and fall of hegemonic core powers in interstate systems --but the differences are undoubtedly more important than the similarities.

The systemness of world-systems composed of chiefdoms is not in doubt, but the nature of this systemness needs closer analysis. What kind of core/periphery relations existed in these systems and what role did these play in reproduction and transformation? Friedman and Rowlands(1977) argue that interpolity hierarchy in such systems relied heavily on prestige goods economies and that these hierarchies were unstable because it was relatively easy for peripheral groups to redefine the hierarchical ideologies and to find local substitutes for the symbolic goods which the core societies were trying to monopolize.[10]

The way in which such societies organize space is important, and a matter for which archaeological surveys are directly relevant. It is to be expected that complex chiefdoms produced the first settlement systems in which villages exhibited a two-tiered size distribution --larger villages were systematically surrounded by smaller villages (Nissen,1988:41; Lightfoot and Feinman,1982) . Of interest is whether or not three-tiered settlement systems sometimes emerged in complex chiefdoms, or whether this more hierarchical spatial system was only found in state-based systems.[11]

Though many of the important questions about the transformation from kin-based logics to state-based logics of accumulation have not been directly addressed from a world -systems perspective, I can make a few observations here. First, this transformation did not begin with the emergence of states, but rather with the emergence of classes. Complex chiefdoms use hierarchically politicized kinship metaphors (e .g. conical clans, ranked lineages) to extract surplus product from direct producers. The way in which kin -based relations are made more hierarchical is through the restructuring of gender hierarchy (Gailey, 1987) and the reorganization of intermarriage-based intersocietal alliances (Friedman, 1982).

These kinds of hierarchies are reinforced by prestige-goods economies in which symbolic goods are used to distinguish between nobles and commoners, and to reward subalterns. These hierarchies can be further stabilized, however, by the creation of specialized and more autonomous organizations which control important societal resources directly and which use politically organized coercion in order to enforce these controls.


The First Cities

Erech, now usually called Uruk, --the city of Gilgamesh --is considered to have been the first true city on Earth. Uruk was built on the Mesopotamian plain in about 3200 BC. The distinction between a city and a town can be drawn in many ways but, however this is done, there was a qualitative difference in population size and spatial extent of the built-up area between Uruk and any earlier settlements. True, a three-tiered settlement system had emerged earlier on the Susiana plain (adjacent to the Mesopotamian plain) but the Uruk settlement system had four tiers and the size of Uruk was much greater than any prior settlement (Nissen, 1988:65-72).[12]

The state-based system which emerged in Lower Mesopotamia has been studied using world -system concepts by a number of archaeologists and anthropologists. Adams's (1966) comparison of early urbanization and state formation in Mesopotamia and Mesoamerica emphasized the importance of settlement systems, regional analysis and the transformation of kinship relations. This was undoubtedly an important stimulus to Friedman and Rowlands(1977) , who have formulated the most comparative and theoretically polished approach to the emergence of early states and  “civilization “ from a world -systems perspective. Important empirical work which has addressed questions about the nature of core/periphery relations in the Mesopotamian system has been done by Lamberg-Karlovsky(1975), Kohl(1987a,1987b) and Algaze (1988,1989).


The “pristine”[13] state in Mesopotamia was a city-state and there quickly emerged a rather long-lasting interstate system composed of city-states. This state system exhibited many of the characteristics of the modern interstate system. The shifting alliances among different city-states exhibited features of a balance of power mechanism which prevented, for a long while, takeover by any single state (Cooper, 1983) .[14] Before the Akkadian conquest created an empire composed of formerly-autonomous core city-states, there were earlier and nearly successful efforts at empire-formation which resembled in some respects the process of the rise and fall of hegemonic core powers in the modern world -system.

Theories of the emergence of the ancient cities are numerous and varied.  The general connections between increases in social hierarchy, intensification of production and the formation of states are agreed upon by everyone, but very different explanations of these transformations are offered. Many approaches take account of regional and inter-regional aspects of the process of pristine state-formation and the emergence of  “civilizations.”[15] Older emphases on the development of productive technology (e.g. Lenski and Lenski 1987) have given way to theories which emphasize the importance of increasing population pressure within a region (e.g. Johnson and Earle,1987) .Renfrew(1975) has developed the notion of the Early State Module (ESM) , an approach which emphasizes co-evolutionary interaction among a set of states within a region (see also Renfrew and Cherry ,1986) .

Carneiro's(1970) circumscription theory is perhaps now the most widely accepted explanation of pristine state formation. Michael Mann (1986) has interpreted and applied Carniero's approach in a way which combines population pressure, resistance to hierarchy, intensified production and the historical development of "technologies of power. " Mann (1986) and Johnson and Earle (1987) apply this general approach to both chiefdom-formation and state- formation. The basic idea is that intensified production allows higher population density (people per unit of land), but it involves harder work compared to foraging. Foragers enjoyed "the original affluent society" (Sahlins  1972) . Methods of planting were known long before they became intensively used. People preferred to forage as long as this was possible. Population growth leads to "hiving off" and emigration as long as there are new places to go. Increasing general population density, however, leads to the adoption of more intensified types of production that allows population density to go higher within a region. As population density increases this exacerbates competition for land and other resources, which increases the prevalence of warfare. People will emigrate if there are ecologically appropriate less populated areas to which to go. But if available options for emigration are circumscribed, hierarchical organizational forms may emerge to regulate group interaction. People resist this, as the cyclical rise and fall of chiefdoms shows. More stable hierarchies depend on the development of new technologies of control -the state.

This general formulation is admirable because it integrates several elements in a convincing way. It also is much less functionalist than earlier explanations that focused only on the socially desirable consequences of hierarchies --conflict management, social investment and savings, and facilitation of trade (e .g. Service, 1975). But there still remain problems of functionalist emphasis in some applications of the circumscription approach (e.g. Johnson and Earle 1987:256-69). The alternative to this should not be a rigid "class struggle model" which argues that all hierarchies are equally exploitative, but rather a more historical approach which studies how particular institutional changes actually occurred.

Mann's (1986) emphasis on the historical nature of social development is admirable in many ways, but his analysis falters when he examines stateless, classless societies. The word "prehistoric" has traditionally been used to refer to non-literate peoples who have not written down their own histories. Mann claims that stateless peoples are prehistoric in a different sense. He argues that general and deterministic evolutionary models adequately describe human social change up until the emergence of the state and "civilization" because the constraints on action produced generally similar processes that occurred everywhere. He points to the small number of places where pristine state formation occurred as evidence for the idea that history in the sense of less- constrained, more conjunctural and intentional action had begun. I follow Eric Wolf (1982) in arguing that the idea of history, in the sense of historical action that is to some extent open -ended, should also be applied to stateless, classless societies. Indeed the Neolithic adoption of horticulture, like state- formation, occurred "pristinely" in only a very few places on Earth, and all of these were pre-literate, stateless, and classless, but not "prehistoric" in Mann's sense.

Some authors stress the predatory or parasitical nature of ancient cities compared to modern industrial cities that, it is alleged, produce something in exchange for the resources they extract from the countryside ( e .g. Sjoberg , 1955) .While this broad comparison may be of some value, a closer look reveals that cities within tributary modes of production varied greatly in the extent to which they were predatory. Algaze's (1988,1989) work argues that the Uruk expansion extracted resources from the periphery by building major urban trading centers in the periphery at key transportation sites and smaller trading posts closer to the origins of valuable raw materials. Algaze argues that these Sumerian traders located in the periphery were able to carry out core/periphery exchange on terms advantageous to the core for a time, but that this exchange led to changes in peripheral societies that worsened the terms of trade for the core. Trade not backed up by military force was not a stable way to extract resources, and the core Sumerian cities were unable to organize effective military control of the periphery.

Urbanization spread within the core region creating competing core city- states and leap-frogged to distant new areas, creating competing core regions (Kohl, 1987a, 1987b). Even those regions that did not develop their own cities and states were often able to resist unequal exchange. It is probably the case that chiefdoms and early states were less exploitative than later states and empires, and that there are important connections between the forms of hierarchy within societies and the nature of core/periphery hierarchies (see Chase- Dunn and Hall 1991) .

In the early state-based world-systems the tributary mode took forms which were partial and imposed over the top of kin-based structures (Zagarell 1986) .The "temple economy" existed beside and above a kin-based system, and the emergent autonomy of military leaders in the context of rising inter-city-state military competition added a greater degree of coercion to class relations .

Diakonoff (1973) has argued that early states in Mesopotamia relied largely on the exploitation of their own local direct producers while later "despotic states" relied more on exploitation of peripheral regions. In these systems "backwash effects" were weak and “spread effects” diffused core characteristics to new regions, creating new cores[16] and overcoming the “development of underdevelopment.”  Kohl (1987b 1988) argues that key military and productive technologies could easily be adopted by peripheral regions and that the "techniques of power" (Mann,1986) for concentrating resources and cumulating inequalities were not well developed in these early state-based world-systems.


Empire Formation: City-states and imperial capitals


The first imperial capital in Mesopotamia was Akkad (Agade) , an new city built to be the capital of the Akkadian empire when its king, Sargon, successfully conquered the entire Sumerian core region and beyond. The nature of this conquest is related to both core/periphery relations and the development of class and ethnic relations in the core city-states. Both Diakonoff (1989) and Mann (1986) argue that part of Sargon's advantage was that he combined core and peripheral military techniques in away that gave him an advantage over the older Sumerian core states. Both authors also focus on the important fact that Akkad’s ruling class displayed greater solidarity than did the ruling classes of the older city-states. In addition Diakonoff adds conflict between classes to his explanation. The older core city-states had developed differentiated ruling classes with competing and contradictory interests among priesthoods, military men and holders of private wealth. They had also become very stratified, with wide differences between rich and poor that corresponded with ethnic differences (Yoffee 1988). The Akkadian conquerors were Semitic-speakers who had greater solidarity with their soldiers and were able to exploit the class and ethnic divisions in the core city-states that they conquered.

Chase-Dunn and Hall (1997) interpreted the Akkadian Empre as an early instance of what they call "semiperipheral development".  But the evidence that Akkadian speakers had long been present within the Sumerian core societies challenges this interpretation. Semiperipheral marchers are usually recent settlers on the edge of an old core (such as the Aztecs). The Akkadian conquest may have been more akin to an ethnic rebellion than a true semiperipheral marcher state conquest.  Older core cities lose dominance to a semiperipheral city. The rising centrality of the former semiperiphery is at least partly due to its ability to combine elements of core forms of organization with peripheral forms to create new techniques that produce a comparative advantage. We do know that the pattern of "semiperipheral marcher states" is repeated over and over again in the spiraling rise, fall and territorial expansion of empires that occurs in the Near East, India, China, Mesoamerica and Peru. To mention a few well-known cases of semiperipheral marcher states: the Assyrians, the Persians, the Macedonians , the Romans, and the Islamic Caliphates.

The first state-based world-systems had cities only in the core, but soon world-systems developed in which semiperipheral and, eventually, peripheral regions also contained cities. While some of these systems approximated the definition of a "world-empire" in which a single overarching state encompassed the territorial network of the exchange of fundamental goods, many others did not. There were many systems in which, though political power in the core was centralized in a single empire, basic goods exchange occurred within a much wider area outside of the direct control of the empire-state. In many other systems empires interacted with one another in the context of a larger division of labor, and non-contiguous core regions interacted with each other indirectly through intervening peripheral zones.

In the early state-based world-systems specialized kinds of cities emerged. Already within the Sumerian system there was Nippur, a pan- Sumerian religious city. Nippur was accorded special religious status by the competing city-states despite the fact that intercity rivalry was represented in the Sumerian religious pantheon as competition for ascendancy among city deities. With empire formation cities also soon diverged with respect to there specialization in trade versus as political centers.


Ports of Trade and Semiperipheral Capitalist City-States


The "port of trade" concept developed by Polanyi and his colleagues (Polanyi, Arensberg and Pearson, 1957) is of relevance to the issue of modes of production and transformation. Polanyi argued that trading in the Babylonian-centered world-system was "marketless" redistribution. He saw the extensive local and regional trade networks as "state-administered" trade. He characterized the specialized traders involved with long-distance trade (which many other scholars have called merchants) as political functionaries who carried out state-administered exchange, not market trade.

Polanyi and his colleagues noticed a number of instances in which politically autonomous cities located in the interstices between empires specialized in trade, and the "ports of trade" category was developed to deal with these. A "port of trade" is understood as neutral territory on which political exchange can be carried out between two redistributive empires.

Unfortunately the particular traders that Polanyi (1957a) chose to illustrate his notion of state trade functionaries have been shown by more recently discovered evidence to have been trading on their own account and paying great attention to prices (Curtin 1984:70). The Old Assyrian City-State (Assur) and its colonies left records that showed that merchants were trading on their own account, not as agents of the king. Convincing evidence of the existence of commodified forms of wealth, land and goods has been found for the Mesopotamian world-system such that no one now claims that these societies were completely marketless. On the other hand, it is obvious that there have been marketless societies --smaller and less hierarchical ones --, and so the general Polanyian point that markets are an institutional invention, not a timeless historical logic, is much easier to defend if we are willing to consider stateless world -systems within our scope of comparison. The existence of some commodified relations in early state-based systems does not tell us how important these were for social reproduction. Rather our attention ought to focus on how, where and why the processes of commodification emerged and became more important. Once we acknowledge that the importance of commodified forms varies across different world-systems (or in the same world- system over time) the analysis of transformation can proceed. I will argue that autonomous trading cities very early took up the role of promoting commodification in the interstices between empires and in peripheral regions. This does not require the conclusion that true ports of trade never existed. It may well be that Dilmun was a port of trade in the Polanyian sense. In each case it is important to look for evidence of market exchange and to try to determine its importance relative to state-administered trade.

Most autonomous city-states in world -systems dominated by larger tributary states and empires maintained their autonomy by a combination of serving the interests of more powerful neighbors and the exercise of their own coercive power. We know that the early Phoenician city-states (Byblos , Tyre , Sidon, Arados) acted as trade intermediaries among empires, naval allies of empires, and agents of the expansion of trade networks to smaller states and stateless regions in the Mediterranean (Frankenstein, 1979) .Were these cities only passive ports of trade? I will argue instead that they were active agents of commodification that ought to be designated as the first capitalist states.

The Phoenician city-states certainly engaged in merchant capitalism (buying cheap and selling dear) and in that way they expanded and intensified the territorial division of labor in the Mediterranean littoral and beyond. Their activities encouraged the production of surplus for sale and the partial economic integration of formerly separate societies. True, they did this in a context in which tributary modes of production were dominant in many regions and kin- based modes in others. I am not saying that capitalism was predominant in the Near Eastern world-system, but that the Phoenicians were agents of commodification who spread the logic of markets and the production of commodities.

So, as Marx (1973) affirms[17], they were merchant capitalists. A more complex question is whether or not they were also production capitalists. We know that they not only bought and sold, but they also produced commodities for sale. The dye for imperial purple came from a sea snail that they harvested. They made cheap mass produced copies of Egyptian glass, Greek pottery and statuary that they sold throughout the Mediterranean. They were the Taiwanese of the ancient world. But did they produce these commodities with wage labor or other semi-commodified labor? Elayi (1982) finds evidence that there may have been a Greek quarter within the Phoenician city of Arados, and thus it may have been Greek artisans who produced the vases and statues that the merchants of Arados sold throughout the Mediterranean. But later Phoenician city-states apparently produced manufactures for export without the help of the Greeks (Tsirkin, 1979) .I argue that the Phoenicians engaged in production capitalism, though the nature of the labor process was, perhaps, less commodified than in modern production capitalism.

I also contend that semiperipheral capitalist city-states played an important role in the processes of growing commercialization that occurred in the tributary world-systems. Few would deny this of the later Italian city- states. Chaudhuri(1985) , Curtin(1984) and Braudel(1984) have examined the role of Malacca in the Indian Ocean in this light .

Cox's (1959:26-7) typology of capitalist cities is a helpful beginning for understanding the roles that cities played in the spread of market integration and the development of capitalist institutional forms inside world-systems in which the tributary modes of production were still predominant. The comparative world -system approach to understanding the development of capitalism within the tributary modes still faces many basic problems despite the great strides that have been made by Braudel(1984) , Lane(1973) , Chaudhuri(1985) and Abu-Lughod(1989). Weber's (1978) insistence on a radical disjuncture between East and West continues to find resonance ( e .g . Chirot, 1986) despite important efforts to qualify or displace it [ e .g . Chaudhuri(1985); Abu-Lughod(1989)]. [18]

World-system theory's interpretation of the role of state action in successful capitalist accumulation in the modern world -system ( e .g. Chase- Dunn, 1989: Part 2) allows us to look at earlier city-states in a new light. Lane's (1979) analysis of protection rent and "violence-controlling enterprises" is relevant for both modern national states (e .g. Bornschier, 1988) and ancient capitalist city-states. The issue of the "capitalist state" involves two related aspects: who controls the state? , and the purposes for which state power is used. Clearly the Phoenician cities used their naval power in pursuit of a strategy of accumulation through commodity trade and commodity production. Of course their enemies, Greeks and Romans, characterized them as pirates rather

than traders. Violence-controlling organizations can be used either to provide protection rents to merchants and capitalists or for direct appropriation of surplus product.[19] In this second usage they are instruments of tribute- gathering, not capitalism. These uses are usually mixed, but the balance in the Phoenician cities was typically closer to Cox's (1959) "Venetian model. "

I should mention that other than maritime and naval cities have been commercially important and may have served as agents of commodification as well. Caravan cities and fairs certainly have played some role. Cox's (1959:26- 7) typology of capitalist cities is organized around the distinction between those that were politically autonomous or sovereign and those were which dependent on larger polities. Maritime cities were often in a better position to prevent conquest by empires because they could employ naval power to protect themselves. The Phoenicians built their cities on promontories or on islands that could be defended from land armies by naval power --like Venice. Political autonomy made maritime city-states more important agents of commodification and they also had advantages in terms of transport costs, but the roles of other kinds of trading cities should not be neglected. Moseley (1983 ) makes this point about caravan cities in her study of the precolonial  development of West Africa. Abu-Lughod's(1989: Chapter 2) discussion of Champagne fair towns demonstrates the political and geographical constraints on non-maritime trading cities. The Silk Road caravan cities need to be

investigated with regard to their role in the development of commodification .

 Carthage was a prototypical example of a semiperipheral capitalist city-state, but the Carthaginian settlements in Spain took the other road -- development of a land-based empire. Hannibal's attack on the Romans was more similar to the marcher state approach, and the delay and weakness of support from Carthage (which was fateful in his defeat) was partly the result of ambivalence over the adoption of the tributary strategy. Braudel's(1984) discussion of the tradeoffs for maritime city-states of leaving agriculture to others is of relevance here. While such states are almost completely dependent on their trade networks for obtaining necessities, they need not invest in the protection of their own agricultural and raw material hinterland, and may thus devote all their resources to trade. The capitalists also avoid having to make compromises with landed aristocrats. This works well in a world in which there are many alternative sources for obtaining necessities, but growing competition for trade opportunities or politically-centralized control of such opportunities decreases the viability of the pure Venetian model.

The Greek city-states were in some ways intermediate between the Phoenicians and the larger states and empires. While they allowed more latitude for markets and commodity production than the empires did, they were not as specialized in trade and commodity production as were the Phoenicians. Greek agriculture combined subsistence production with production for the urban market, a model which allowed the Greeks to colonize widely in the Mediterranean and to develop a system in which local and long-distance market interactions were relatively balanced and integrated ( Rostovtzeff  1941). The result was a less purely capitalist state than that of the Phoenicians, but a society that nevertheless was at an advanced level of commercialization.

Braudel (1984:27) contended that early world-economies had a single city at their centers. If we broaden our definition of world-system connections to include prestige goods exchange and interconnections produced by political/military rivalry, [20] the capitalist city-states were never the political centers of any world -system. They were, of course, the center of their own trading networks, but these were always parts of larger systems of production and exchange. After the rise of empires world-systems have never been centered on a single city-state. Braudel(1984:91) himself implies as much when he says of the capitalist city-states ,

Nor is it paradoxical to think that the cities needed the space

around them, the markets, the protected circulation zones --in

short that they required larger states to batten on: they were

obliged to prey on others to survive. Venice would have been

unthinkable without first the Byzantine and later the Turkish Empire.

Most world-systems are multicentric and some even contain non- contiguous core regions. The idea of multicentric systems, however, does not mean that we should pay no attention to the extent to which world -systems are politically centralized or decentralized. Wallerstein's distinction between world- empires and interstate systems is valuable precisely because it points to this important dimension. Most empire- based world -systems alternate between periods in which there is a single dominant empire controlling most of the core versus periods in which two or three competing empires of unequal size and smaller states are interacting. A statistic like the Ray-Singer (1973) power concentration index could be used to measure the degree of concentration or dispersion of power among core states and empires. If this measure was computed across systems I believe the following would be found: though concentration varies with the rise and fall of "universal states" in the tributary systems and with the rise and fall of hegemonic core powers in the modern world -system, the modern world -system has been, on the average, less politically centralized than pre capitalist systems.[21] This is really only a more nuanced way of understanding the point which Wallerstein makes by contrasting world -empires with world-economies .

I would not argue that autonomous capitalist city-states were the only agents of commodification in the world -systems in which the logic of tribute was predominant. Obviously the tributary empires themselves became increasingly commercialized in the sense that commodity forms developed and expanded. The Persian rulers learned to grant a certain amount of autonomy to cities such as Babylon in order to thereby be able to tax the profits of successful merchants (Cook, 1983) .More direct efforts to monopolize trade often stifled the golden goose. China, India and Rome were rather commercialized systems with highly developed forms of money, credit, and markets. Polanyi  (1957b) granted that classical Athens was a rather highly commodified society because money in small denominations was in wide use and it was possible to purchase a prepared meal. Markets, money, rather complex institutions of credit, the substantial commodification of land, and the commodification of labor in the form of chattel slavery and even wage labor --all these things played an important part in the commercializing world-systems, and yet capitalism was not yet predominant. In China market forces and capitalist accumulation threatened the tributary empire powerfully (especially during the Sung dynasty) , but these were always brought back under the control of the mandarins. In Rome, Byzantium, the Islamic empires and the Ottoman Empire market forces were strong, but the logic of tribute gathering was never overcome.

In my judgment the problem of the development of capitalism and the rise of the West needs further work from the point of view of a comparative Marxist world -system perspective. In response to the overemphasis on the uniquenesses of Europe it is natural to emphasize the similarities and the connections between Europe and the larger Afroeurasian world-system. But a more sophisticated comparative world-system approach will allow us to reexamine which of Europe's special characteristics were indeed relevant for its rise, while taking account of its important interactions with older core regions.

Capitalism does not become predominant until: 1. capitalists control important core states, not just semiperipheral city-states, and 2. market forces and the logic of capitalist accumulation becomes predominant over accumulation by direct political means. Capitalism grew strong in the interstices of tributary modes, and especially in the extremely decentralized tributary mode that was European feudalism. The medieval European cities in which capitalists took state power were not unique in kind, but their dense concentration within a small region amplified their effects on the logic of accumulation in that region. Earlier capitalist city-states had been dispersed in the interstices of empires.

The period from 235 AD to the twelfth century was a period of devolution of political power and of economic networks in Europe and the geographical shift of dominant state power "back East" to the older core region. According to Pirenne (1980) Western political decentralization was not accompanied by full economic decline until the rise of Islam turned the Mediterranean into a "Moslem Lake," cutting off the long distance trade of Western Europe with the East.

Political and economic involution in Europe led to the "manorial economy, " and feudalism, a very decentralized version of the tributary mode of production. This was based on a synthesis of Roman and Teutonic cultural institutions (Anderson, 1974), and was fertile soil for the growth of capitalist trade and production when the long distance trade started up again. The very decentralization of the polity and the "parcellization of sovereignty" (Anderson, 1974) created great latitude for market forces and frequent interstices within which merchants and commodity producers could find political protection. Many of the medieval cities of Europe were similar to the capitalist city-states of old, except that the tributary mode of production within which they operated was itself fragmented and unable to repress the growth of merchant wealth or to prevent the escape of serfs to the cities. The social inventions which made capitalist commodity production possible --money, credit, commodified labor, price-setting markets, contract law, etc. --were present within the semi-commercialized tributary empires, especially the Roman Empire, but they had not dominated the logic of accumulation as they were to do once imported and planted in the fertile and unconstrained soil of medieval Europe.

There had been feudalism in other places at other times without the florescence of capitalism. Mann (1986) has claimed that the cultural integration of elites by the ideology and organization of Christendom provided a "normative pacification" which facilitated the development of market forces and agricultural technology. But earlier interstate systems and feudal structures had shared religious ideologies, and though this may have facilitated communication , cooperation and alliances, these effects did not lead to exceptionally strong market forces in the absence of opportunities for long distance trade or the other institutional elements necessary for a market economy. It was the semiperipheral position of certain European regions within the larger multicentric Afroeurasian super-world-system (in which commercialized relations and the institutional and cultural artifacts of capitalism had already been developed [Abu-Lughod 1989]), which made it possible for capitalism to become a dominant regional mode of production for the first time. European feudalism, unlike earlier decentralized polities, was in the right place at the right time.

It was the growth of capitalism which provided the resources which enabled kings to centralize power sufficiently to create the European states, and thus the European interstate system. The first half of the "longevity of the interstate system" in the European sub-system since the fall of Rome was thus a period in which parts of Europe became semiperipheral relative to the core states of Byzantium, the Arab Caliphates, and the Islamic Empires. This was followed by a period that saw the upward mobility of parts of Europe based on the development of market forces, the strengthening of the European states, and the creation of a regional interstate system with several core national states at its center. The rise of capitalist national states (as opposed to city-states) was a European first, and the first of these was the Dutch Republic of the seventeenth century.


Cities in the Modern World-System

Many scholars have compared the hegemonies of Venice and Amsterdam, and claimed that the Dutch hegemony was more similar to that of a city-state than a national state (e.g. Barbour 1963) .True, the federal structure of the Dutch state allowed Amsterdam a great deal of autonomy and facilitated its leading influence within the state. But the Dutch hegemony also revealed the growing importance of a protected home market for the most successful players in a world -system in which capitalist accumulation had become more important than tributary accumulation. If we draw a line in time from Venice to the Netherlands, through the United Kingdom to the United States we see that the size of the hegemon's home market has grown with the size of the world-system.

Charles Tilly's (1989) important study of European cities and states from AD 900 to the present illustrates how countries pursued different strategies of state formation. The optimum territorial scale of state power was changing with the increased institutionalization of capitalism in Europe. The earlier capitalist cities that resisted the formation of large states in their territories were superceded by larger states that were able to exploit the growing scale of production, transportation and regional integration. Tilly's study, while it is far superior to both those "state-centric" political scientists who completely ignore capitalism (e .g. Wilkinson, 1987) those students of cities who assume that market competition is the only relevant form of interaction ( e .g. Jacobs , 1984) , nevertheless employs an ahistorical definition of capitalism[22] and oversimplifies the relationship between capitalism and cities. Though differentiation among states is an important part his story, Tilly ignores the emergence of a new core/periphery hierarchy in which some regions of Europe come to dominate other areas. He does not analyze the rise and fall of hegemons and neither does he try to explain why Europe is so resistant to empire formation.

As capitalism became predominant in the European core, the modern interstate system --composed primarily of national states --came into existence and eventually spread to the whole globe. Polanyi(1944) and Braudel(1984) discuss the importance of a protected and integrated home market for competition in a capitalist world-economy. With the rise of capitalist national states sovereign capitalist city-states could no longer compete with larger players who took up the accumulation of resources through commodity production and trade. A few survived as ex-colonial enclaves or trade entrepots, but no longer were city-states the main bearers of commodification. Most cities had long been the hostages of empires, while a small number of sovereign capitalist city-states operated in the semiperipheral interstices. Now national states and capitalist firms became the main players in a world -system in which the logic of capitalist accumulation had become predominant over other logics of accumulation.

There has been a lot of excellent research on urban development, national city-systems and the role of world cities in the cyclical processes and secular trends that characterize the capitalist world-economy. Some of the best of this is contained in Timberlake (1985). More recent studies have examined the connections between core financial and global service cities and peripheral and semiperipheral countries (e.g. Meyer,1986; Sassen,1988) and the development of single cities in a world-system context (e.g. Hill and Feagin,1987; Feagin, 1988) .These excellent studies add immensely to our understanding of how cities have changed in the context of the restructuring of the international division of labor. My only criticism of this work is its often somewhat limited time horizon that portrays very recent changes as completely new departures. If you think a global economy first emerged in the 1960s your analysis of world cities will be likely to neglect comparisons with earlier periods that are entirely relevant for understanding recent changes. Now that we have a fairly well specified description of long-run world-system cycles and trends it becomes possible to find out what is really new about recent changes.

The most recent restructuring of the global division of labor has exported manufacturing and heavy industry from core states to the semiperiphery and concentrated global services in world cities such as New York and Los Angeles. Sassen(1988) convincingly argues that this reorganization is connected to new flows of immigrants into these world cities from peripheral countries which have received heavy doses of recent foreign investment. The expansion of the informal sector and the polarization of income distributions in these world cities is partly due to global restructuring and new immigration.

This recent set of developments bears comparison with earlier world - system periods in which a hegemon was declining and the international division of labor was being reorganized. London experienced a pattern of development in the 1870s similar in some respects to that of contemporary New York. The nineteenth century English city system had been moving toward a less hierarchical city-size distribution because of the rapid growth of the industrial cities (Manchester, Birmingham, etc.), but with the beginning of the British industrial loss of leadership London's rate of growth increased relative to the industrial cities. London became an increasingly important center of world financial services and markets as the British position in industry was eroding. The inflow of Irish immigrants into the East End provided a supply of cheap labor to the burgeoning "informal sector" in the world city (Jones 1971) .There is some evidence that Amsterdam became similarly more important as a world city during the Dutch hegemonic decline .


The World City System


The study of the world city system as such was first advocated by John Walton(1976) .The structure of exchange connections among cities in the world- system has long been a nested hierarchy with regional and national city systems woven into a global network of cities. Core cities are more interconnected with one another than peripheral cities are, just as is the case with core and peripheral countries. If we had a global map of trade, communications and other interactions among all the cities on Earth it would resemble an airline flight map. In order to get from one part of the periphery to another you have to go through the core. This hierarchical network feature undoubtedly varies to some extent over time. Unfortunately, comprehensive data are not available to study connections among cities for the whole world, but it is likely that network studies in which national states rather than cities are the nodes of the analysis give us a fairly close approximation of how the hierarchical structure changes over time (e. g. Smith and White, 1989). Unfortunately such studies have only been done for very recent time periods.

Using the population sizes of the largest cities within the Europe- centered system I have shown that the world city size distribution varies approximately with the rise and fall of hegemonic core powers (Chase- Dunn 1985b ).[23]  The rise of the semiperipheral megacities has flattened the world city-size distribution in recent decades. In 1975 Mexico City was the fourth largest urban agglomeration in the world and Sao Paolo was seventh (Chase-Dunn 1985b:Table 12.1) .My analysis of changes in the world city-size distribution shows that the trend toward a flatter global city system began in 1875 with the downturn of the British hegemony (Chase-Dunn,1985b:Figure 12.2) .The former correspondence between the hegemonic sequence and the steepness/flatness of the world city-size distribution failed with the rise of the United States hegemony. It is possible that this has been partly due to certain ceiling-effects that slow the growth of the very largest megacities and allow smaller ones to catch up.

Is this flattening of the world city size hierarchy and movement of semiperipheral cities in to the club of the very largest world cities in anyway associated with a reduction in core/periphery inequalities? Certainly the industrialization of the semiperipheral "NICs" has reorganized the international division of labor. Formerly core specialties like heavy industry have migrated to the semiperiphery while new lead industries have emerged in the core. There is considerable evidence that all this reorganization has not, however, reduced the magnitude of relative inequalities between core and periphery. Table 1 shows that, though world relative energy usage has grown in the semiperipheral countries as we would expect, relative world income has not become more equally distributed. Other data on recent changes in the distribution of world resources is reviewed in Chase-Dunn (1989: Chapter 12) and supports the conclusion that core/periphery inequalities have not diminished.


Table 1: Concentration of resources among countries in the world -system: 1960-1980


1960  1970  1980

%        %       %

The proportion of world energy consumed by:


the countries highest on per capita energy consumption

with 20% of world population                                                  83.3  80.3  69.9



the middle countries on per capita energy consumption

with 60% of world population                                                 16.1  18.8  29.1



the countries lowest on per capita energy consumption

 with 20% of world population                                                   .6  .9  1.0



The proportion of world GNP going to:


the countries highest on GNP per capita

with 20% of world population                                                    79.2  79.8  81.1



the middle countries on GNP per capita

with 60% of world population                                                   19.4  18.9  18.1



the countries lowest on GNP per capita

with 20% of world population                                                    1.4  1.3  .8



Percentages are based in sets of countries for which data are available for all time points. Thus comparisons over time are not confounded by missing data. N = Number of countries . Sources: GNP is from World Bank's Economic Analysis and Projections Department. Energy consumption is from World Tables, 1983, Volume II.


National City Systems


The general connection between urbanization (the rise in the proportion of the population of national societies who live in cities) and industrialization in the core societies is well-known. Comparative research has shown, however, that the causes and consequences of urbanization are quite different for core and peripheral countries (Bairoch 1975; Kentor,1981; Timberlake,1985). Colonialism and the processes of peripheralization heavily influenced the city systems of most peripheral countries. In many of these countries the largest cities functioned as administrative centers of core colonialism and as ports for exporting peripheral products and importing core goods. These observations were used by several analysts of peripheral urbanization to explain why peripheral city-systems displayed greater "urban primacy than core city- systems.” Urban primacy is a situation in which the largest city is many times larger in population than the second largest city, and the size distribution of cities is generally steep. A country in which all the cities were of the same size would have a perfectly flat city-size distribution. The relative steepness of the national city-size distribution is a characteristic of urban systems that can be measured for many countries over a relatively long time period by gathering data on the population sizes of cities. A research project at Johns Hopkins University gathered data on the population size of the ten largest cities in each country from 1800 to 1980 (Chase-Dunn,1982,1983; Walters,1985) .

Analyses of these data show that significant differences in the level of urban primacy between core and peripheral countries did not emerge until  the 1930s and 1940s (Chase-Dunn,1984a,1985a,). It is well known that some core national city-systems have long been primate (e. g. England, France) and that some peripheral countries had relatively flat city-size distributions (e .g . Colombia) .Our research has shown that peripheral city systems had indeed become, on the average, much more primate than core ones, but that this did not happen during colonization or original peripheralization but rather during the 1930's and '40s.

Another finding was that the El-Shakhs(1972) hypothesis that national city systems go through a sequence from flat to primate to "log-normal" is not supported by the data (Chase-Dunn,1984b) .A more recent and thorough multivariate crossnational analysis of these data (Lyman,1989) shows that the level of economic development ( GNP per capita) decreases urban primacy in the 1960's and 1970's. The causes of the shift of the periphery toward urban primacy in the middle decades of this century are still not well understood.

Portes(1989) has shown that the trend toward urban primacy in Latin America has stemmed and even reversed in the 1980's. Though these countries now seem to be moving toward what appears on the surface to be a more core- like distribution of population across cities, this is unlikely to be the result of processes associated with core development such as industrialization, the growth of middle-sized industrial cities, the creation of a more integrated and balance national economy, etc. More likely people have stopped flocking to the "megacities" because life in them has become unsustainable in the period of debt crisis and stagnating economic growth that has been the lot of most peripheral and semiperipheral countries in the 1980's. The causes of this reversal need to be studied comparatively.


Cities and Socialism

Boulding's (1978) discussion of cities in the international system portrays cities as civil hostages to war-making national states. From the long-run perspective I have used in this paper it is obvious that we should not adopt a rigid general notion of pacific trading cities and war-like states. Following my discussion above, the role of cities has changed with the rise and fall of different logics of accumulation. Early city-states were protagonists of the tributary mode of production. They developed the techniques of power that made empire formation by conquest possible. Later, in the context of already- formed tributary empires, a few city-states in the semiperiphery became protagonists of the capitalist mode of production. When the capitalist mode became predominant, national states and firms pushed cities out of the transformational role. Is it possible for cities to become important protagonists of socialism?

As the marketization of socialist states proceeds, those of us who still want to create a democratic and collectively rational world need to think of new organization forms that can act as the bearers of socialist logic. The taking of state power by socialist movements in many ways reproduced the logic of the interstate system and worked against the logic of socialism at the world -system level. Unlike capitalism, which thrives on competition and conflict in the interstices of larger systems, socialism is a holistic logic of cooperation that is undercut by external threats. Transnational and trans organizational forms of organization will be important for constructing socialism at the level of the world-system. Alliances of cities might be a contender for the role of key transformational actor. Many would agree that it has now been conclusively shown that a socialist mode of production in one country cannot be sustained in a larger world -system in which capitalism remains a dynamic and expanding force.[24]  If this is true of even large countries such as the Soviet Union and China, it would be absurd to assert that a self -sustaining socialist mode of production could be constructed and sustained in a single city. Most cities in the modern world -system are dependent on national states and capitalist firms, and are pitted against each other in a struggle for state support and private capital investment. Of course, states differ in the extent to which cities are delegated autonomous political powers, but no cities, even those few politically autonomous city-states which still exist, have sufficient resources to do what large semiperipheral socialist states have not been able to do. This does not mean that socialist politics in the municipal arena is meaningless, but rather that such efforts must be construed as part of a larger struggle.[25]

David Harvey (1985: 275) has suggested that "federated structures of interurban cooperation " might be organized to help progressive movements within countries. A similar approach might be tried internationally. Transnational alliances of cities might be able to mobilize movements that are transformative. Creating alternative exchanges and political structures that would undermine the logic of the interstate system and the world market could do this. A socialist international of cities could (in principle) organize mutual aid and collectively bargain with large firms and states. It could refuse to participate in the madness of nuclear weaponry and pollution by organizing and coordinating municipal nuclear-free zones. Chadwick Alger (1990) reviews the international municipal movements that have confronted global issues of war prevention, disarmament, poverty, and human rights.

The above proposal for a socialist international of cities certainly seems utopian in the present climate. On the other hand, the spiraling of interaction between capitalist expansion and anti-systemic movements will undoubtedly continue. We need to think creatively about new organizational forms for resistance and possibilities for the creation of humane institutions. [26] The explosion of awareness about globalism and increasing concern about the structural conditions that threaten our species and our planet provide fertile ground for new social movements. Municipal networks are one form of organization that such movements should utilize.




Abu-Lughod, Janet L. 1989. Before European Hegemony: The World-System A.D.

1250-1350. New York: Oxford University Press.

Adams, Robert McCormick. 1966. The Evolution of Urban Society: Early

Mesopotamia and Prehispanic Mexico. Chicago: Aldine .

Algaze, Guillermo. 1988. "Mesopotamian expansion and its consequences:

'momentum towards empire' in the second half of the fourth millenium BC"

 Ph.D Dissertation, Universityof Chicago.

-_.1989. "The Uruk expansion: cross-cultural exchange as a factor in early

Mesopotamian civilization. " Current Anthropology 30, 5 ( December) .

Alger, Chadwick F. 1990. "The world relations of cities: closing the gap between

 social science paradigms and everyday human experience." International

 Studies Quarterly 34,4:493-518 (December).

Amin, Samir. 1980. Class and Nation. Historicallyand in the Current Crisis. New

York:Monthly Review Press.

Anderson, Perry .1974. Passages From AntiQuity to Feudalism. London: New Left


Bairoch, Paul. 1975. The Economic Development of The Third World Since 1900.

Berkeley: University of California Press .

Barbour, Violet. 1963 .Capitalism in Amsterdam in the 17th Century .Ann

Arbor:University of Michigan Press.

Barfield, Thomas. 1989 .The Perilous Frontier: Nomadic Empires and China.

Cambridge, MA. : Basil Blackwell.

Blanton, Richard E. and Gary Feinman. 1984. "The Mesoamerican world system."

American Anthropologist 86, 3: 673-92.

Bornschier, Volker 1988. :W:estliche Gesellschaft im Wandel. Frankfurt and New

York: Campus.

Boulding, Kenneth. 1978. "The city as an element in the international system. "Pp .

150-8 in L. S. Bourne and J. W. Simmons(eds. ) Systems of Cities. New York:

 Oxford University Press .

Braudel, Fernand. 1975. Capitalism and Material Life, 1400-1800. New York: Harper

and Row.

-1984. The Perspective of the World Volume 3 of Civilization and Capitalism,

15th-18th Centuries. New York: Harper and Row.

Carneiro, Robert L. 1970.  " A theory of the origin of the state. " Science 169: 733-


Chase-Dunn, Christopher. 1982a. "World division of labor and the development of

city systems, " Comparative Research 9,3 : 3 -9 ( Winter) .

__________1982b. (ed. ) Socialist States in the World-System. Beverly Hills: Sage.

__________1983. "Urbanization in the world-system: new directions for research," in

 Comparative Urban Research 9,2 : 41-46. Reprinted in Michael P. Smith ( ed. )

 Cities in Transformation: Class, Capital and the State, Beverly Hills, CA :


__________1984a. "Levels of urban primacy in zones of the world-system since 1800:

 a preliminary report, " Presented at the Annual Meetings of the American

 Sociological Association, San Antonio, August 27-31.

__________1984b. "Patterns of change in national city systems : 1800-1970" Presented to the Program on Interdependent Political Economy, University of Chicago , November 15,1984.

___________1985a. "The coming of urban primacy in Latin America, " Comparative Urban Research 11,1-2:14-31. Also published as "El fenomeno de primacia de una ciudad an los sistemas urbanos latinoamericanos: su surgimiento, " Pp. 27 - 46 in J.E. Hardoy and A. Fortes (eds. )~_g~des and Sistemas Urbanos: Economia Informal y Desorden Espacial, Buenos Aires: CLACSO.

___________1985b. "The system of world cities:A.D. 800-1975," pp.269-292 in Michael "

Timberlake (ed. ) Urbanization in the World Economy. New York:Academic Press.

__________1986. "World division of labor and the development of city systems, " Final report to the National Science Foundation. Grant #SES 8318793. May.

__________1988. " Comparing world -systems: toward a theory of semiperipheral develop- ment." Comparative Civilizations Review 19:39-66, Fall.

___________1989 .Global Formation: Structures of the World- Economy. Cambridge, MA :

Basil Blackwell.

__________1990a. "World state formation." Political Geography Quarterly 9,1(April).

__________1990b. "Resistance to imperialism: semiperipheral actors. " .R~view 13,1 : 1- .32 (Winter) .

Chase-Dunn Christopher and Thomas D. Hall 1991. "Conceptualizing core I periphery hierarchies

for comparative study. " Pp. 5 -44 in Chase- Dunn and Hall ( eds .) Core I Periphery Relations in Precapitalist Worlds. Boulder , CO. : Westview Press .

Chase-Dunn, Christopher and Elena Ermolaeva, "The ancient Hawaiian world-system: research in progress," A paper presented at the annual meeting of the International Studies Association, March 29, 1994, Washington, DC.

Chase-Dunn, Christopher and Kelly M. Mann 1998 The Wintu and Their Neighbors: A Very Small World-System in Northern California. Tucson: University of Arizona Press.

Chaudhuri, K.N. 1985. Trade and Civilization in the Indian Ocean: An Economic

History from the Rise of Islam to 1750. Cambridge: Cambridge University Press.

Chirot, Daniel 1986. Social Change in the Modern Era. New York: Harcourt,

Brace, Jovanovich.

Cook, J.M. 1973. The Persian Empire. New York: Shocken.

Cooper, Jerrold S. 1983 "Reconstructing history from ancient inscriptions: the

Lagash-Umma border conflict." Sources from the_~ncient Near Eas! (SANE) 2, Malibu: Undena Publications .

Cox, Oliver. 1959. The Foundations of Capitalism. New York: Philosophical


Curtin, Phillip. 1984 .Cross -cultural Trade in World History .Cambridge :

 Cambridge University Press .

Diakonoff, Igor M. 1973. "The rise of the despotic state in ancient Mesopotamia ."

Pp. 173-203 in I. M. Diakonoff (ed. ) Ancient Mesopotamia. G. M. Sergheyev (trans. ) Walluf bei Weisbaden: Dr. Martin Sandig.

1989. "Early despotisms in Mesopotamia. " Chapter 3 in History of the Ancient -World:

Early Antiquity, Trans. and ed. by Philip Kohl. Chicago: University of Chicago Press.

Dodgshon, Robert A. 1987. The European Past: Social Evolution and Spatial Order. London:

Macmillan .

Ekholm, Kasja and Jonathan Friedman 1982. "'Capital' imperialism and exploitation in the ancient

world-systems. " Review 6, 1 :87-110.

Elayi , Josette 1982. " Studies in Phoenician geography during the Persian period. " Journal of

Near Eastern Studies 41,2: 83-110.

El-Shakhs, Salah. 1972. "Development, primacyand systems of cities," ~nal of Developing Areas

7:11-36 (October).

Feagin, Joe R. 1988. Free Enterprise City: Houston in politico-Economic Perspective.

New Brunswick,NJ:Rutgers University Press.

Fiala, Robert and David Kamens. 1986. "Urban growth and the world polity in the nineteenth

and twentieth centuries: a research agenda," Studies in Comparative International Development 23-35 (Spring) .

Finley, Moses. 1973 .The Ancient Economy. Berkeley: University of California


Frankenstein, Susan. 1979. "The Phoenicians in the Far West: A function of Neo- Assyrian imperialism.

II Pp. 263-94 in Mogens T. Larsen ( ed. ) Power and ~ro~aganda: A Symposium on Ancient Empires. Copenhagen: Akademisk Forlag.

Friedman, Jonathan. 1982. "Catastrophe and continuity in social evolution." Pp. 175-196 in Colin

Renfrew, Michael J. Rowlands and Barbara Abbott Segraves ( eds .) Theoryand Explanation in Archeolog-y: The Southampton Conference . New York: Academic Press.

Friedman, Jonathan and Michael Rowlands. 1977. "Notes towards and epigenetic model of the evolution

of 'civilization' .II Pp. 201-278 in J. Friedman and M . J. Rowlands ( eds .) The Evolution of Social Systems. London: Duckworth .

Gailey, Christine Ward. 1987 .!.$i~s~p tQ Ki!!e:ship: Gender Hierarchy and State Formation in the

Tongan Islands. Austin: University of Texas Press .

Gills, Barry K. and Andre Gunder Frank 1991. "5000 years of world system history: the

cumulation of accumulation." Pp. 67-112 in Chase-Dunn and Hall.

Gledhill, John and Mogens Larsen. 1982. "The Polanyi paradigm and adynamic

analysis of archaic states. II Pp. 197 -230 in Colin Renfrew ~ Theory and ~xplanation in Archeoloe:v New York: Academic Press .

Hall, Thomas D. 1991. "The role of nomads in core/periphery relations," Pp. 212- 239 in Chase-

Dunn and Hall.

Harvey, David 1985 Consciousness and the Urban  Experience: Studies in

the History and Theory of Capitalist Urbanization. Baltimore: Johns Hopkins University Press.

Hill, Richard C. and Joe R. Feagin. 1987. "Detroit and Houston: two cities in

global perspective. II Pp. 155-77 in Michael P. Smith and Joe R. Feagin( eds .) The Capitalist City. London: Basil Blackwell.

Hopkins, Keith. 1978a. "Economic growth and towns in classical antiquity." Pp.35-

78 in Philip Abrams andE.A. Wrigley (eds. ) Towns in Societies. Cambridge: Cambridge University Press.

-1978b Conquerors and Slaves. Cambridge: Cambridge University Press. Hopkins,

Terence K. 1957. "Sociology and the substantive viewof the economy."

Pp. 270-306 in Polanyi, Arensberg and Pearson.

Jacobs, Jane 1984 Cities and the Wealth of Nations. New York:Random House.

Johnson, Allen W. and Timothy Earle. 1987. Th~ Evolution of Human Societies:

From Foraging Group to Agrarian State Stanford: Stanford University Press.

Jones, Gareth Stedman. 1971. Outcaste London. Oxford: Clarendon Press.

Kelly, Raymond C. 1985. The Nuer Conquest: 1:he Structure and Development o an Expansionist

System. Ann Arbor: University of Michigan Press.

Kentor, Jeffrey. 1981. II Structural determinants of peripheral urbanization: the effects of

international dependence, II American Socioloe:ical Review 46: 201- 11 (April)

Kohl, Phillip. 1987a. "The use and abuse of world systems theory: the case of the 'pristine'

West Asian state." In Archeoloe:ical Advances in Method and Theory II. New York: Academic Press, pp. 1-35.

-1987b.”The ancient economy, transferable technologies and the Bronze Age world-system: a view

from the northeastern frontier of the Ancient Near East." Pp. 13-24 in Rowlands, Larsen and Kristiansen.

Kristiansen, Kristian. 1987.  “Centre and periphery in Bronze Age Scandinavia.”

Pp. 74-86 in Rowlands, Larsen and Kristiansen.

Lamberg- Karlovsky , C. C. 1975. "Third millennium modes of exchange and modes

of production." Pp. 341-368 in J. A. Sabloff and C. C. Lamberg-Karlovsky

(eds.) Ancient Civilization and Trade. Albuquerque: University of New

Mexico Press.

Lane, Frederic. 1973. Venice: A Maritime Republic. Baltimore: Johns Hopkins


-1979 .Profits From Power: Readings in Protection Rent and Violence-

Controlling Enterprises. Albany: SUNY Press .

Leach, Edmund. 1954. The Political Systems of Highland Burma. London: Athlone


Lenski, Gerhard and Jean Lenski. 1987. Human Societies. New York:McGraw-Hill.

Lightfoot, Kent G. and Gary M. Feinman 1982. "Social differentiation

and leadership development in early pithouse villages in the Mogollon region of

the American Southwest." American Antiquity 47,1:64-86.

Lyman, H. Bradford. 1989. "Urban Primacy: A Crossnational Study," Ph.D.

Dissertation, Department of Sociology , University of Maryland.

Mann, Michael. 1986 .The Sources of Social Power: A History of Power from the

Beginning to A.D. 1760. Cambridge: Cambridge University Press.

Marx, Karl. 1973[1857-8] .The Grundrisse. New York: Vintage.

Meyer, David. 1986. "The world system of cities: relations between international

financial metropolises and South American cities. " Social Forces 64: 553-81 (March).

Moseley, Katherine P. 1989. " Caravel and caravan: West Africa and the world –

economies ca. 900-1900 A.D." Presented at the Annual Meetings of the

International Studies Association, Washington, DC, April 11.

Myrdal, Gunnar. 1971. Economic Theory and Underdeveloped Regions. New York:

Harper and Row

Nissen, Hans J. 1988 .The Early History of the Ancient Near East,9000-2000 B. C,.

Chicago: University of Chicago Press .

Pearson, Harry W. 1957. "The secular debate on economic primitivism. " Pp. 3-11

in Polany, Arensberg and Pearson.

Peregrine, Peter 1991. "Prehistoric chiefdoms on the American mid-contentinent:

a world-system based on prestige goods." Pp. 193-211 in Chase-Dunn and Hall.

Pirenne, Henri. 1980[1939] .Muhammad and Charlemagne. Totowa, N.J.: Barnes

and Noble Books

Polanyi, Karl. 1957a. "Marketless trading in Hammurabl's time. " Pp. 12-26 in Polanyi,

Arensberg and Pearson.

-1957b"Aristotle discovers the economy." Pp. 64-96 in Polanyi,Arensberg

and Pearson.

-1977. The Livelihood of Man. Harry W. Pearson (ed.). New York:

Academic Press .

Polanyi, Karl, Conrad M. Arensberg and Harry W. Pearson (eds.) .1957. Trade

and Market in the Early Empires. Chicago: Regnery .

Portes, Alejandro .1989. "Latin American urbanization in the years of the crisis, "

Latin American Research Review ( December) .

Portes, Alejandro, Manuel Castells and Lauren A. Benton (eds.). 1989. The

Informal Economy: Studies in Advanced and Less Developed Countrie Baltimore: Johns Hopkins University Press.

Ray, James Lee and J. David Singer 1973. "Measuring the concentration of power

in the international system, " Sociological Methods and Research, 1, 4: 403- 437.
Renfrew, Colin R. 1975. "Trade as action at a distance: questions of integration

and communication." Pp. 3-59 in J.A. Sabloff and C.C. Lamborg-

Karlovsky(eds. ) Ancient Civilization and Trade. Albuquerque,N.M. : University of New Mexico Press .

Renfrew, Colin and John F. Cherry (eds.). 1986. Peer Polity Interaction

and Socio-political Change. Cambridge: Cambridge University Press .

Rostovtzeff, Michael. 1941. The Social and Economic History of the

Hellenistic World. Oxford: Clarendon Press.

Rowlands, Michael, Mogens Larsen and Kristian Kristiansen ( eds .) .1987. Centre

and Periphery in the Ancient World. Cambridge: Cambridge UniVerBrtY Press.

Sahlins, Marshall. 1972. Stone Age Economics. Chicago: Aldine.

Sassen, Saskia. 1988. The Mobility of Labor and Capital: A Study in International

Investment and Labor Flow. Cambridge: Cambridge University Press .

Schneider, Jane. 1977. "Was there a pre-capitalist world -system'?" Peasant Studies

6,1:20-29. Reprinted as Chapter 2 of Chase-Dunn and Hall, 1991.

Service, Elman R. 1975. The Origins of the State and Civilization. New York:


Sjoberg, Gideon. 1955. "The preindustrial city. " American Journal of Sociology

60:438-45 (March).

Smith, David A. and Douglas R. White. 1989. "Centrality in international trade

flows: implications for world-system hegemony and mobility." A paper presented at the joint meeting of the American and British International Studies Associations, London, March 24.

Taylor, Peter. 1990. " Britain I s changing role in the world -economy, " Review

13,1: 33-48(Winter) .

Tilly, Charles 1986. "Since Gilgamesh." Social Research 53,3:391-410 (Autumn).

-1990. Coercion. Capital. and European States. AD 990-1990. Cambrdige,

MA. : Basil Blackwell.

Timberlake, Michael. 1985. "The world-system perspective and urbanization." Pp.

3-24 in M. Timberlake(ed. ) Urbanization in the World-Economy. New York: Academic Press .

________1985. Urbanization in the World-Economy. New York: Academic Press.

Tsirkin, Yu. B. 1979. "Economyof the Phoenician settlements in Spain." Pp.547-

564 in Edward Lipinski ( ed. ) State and Temple Economy in the Ancient Near East Volume 2. Departement Orientalistiek, University of Leuven.

Wallerstein, Immanuel. 1974. The Modern World-System, V. 1. New York:

Academic Press .

_________1979. The Capitalist World- Economy. Cambridge: Cambridge University Press.

_________1984. The Politics of the World-Economy. Cambridge: Cambridge University


_________1989. The Modern World-System, III. New York: Acadenric


WaIters, Pamela B. 1985. "Systems of cities and urban primacy: problems of

definition and measurement." Pp. 63-120 in TimberIake, 1985.

Walton, John. 1976. "The political economy of world urban systems: directions

for comparative research. " In J. Walton and L. H. Masotti( eds .) .The City

in Comparative Perspective: Crossnational Research and New Directions in

Theory. New York: Halstead Press.

Weber, Max. 1975. Economy and Society. Volume 2. Berkeley: University of California


Wilkinson, David. 1987. "Central Civilization."


_______1990. " Cities and their civilizations. " A paper presented at the 12th World "

Congress of Sociology, Madrid, July 9-13.

Wolf , Eric R. 1982 .Europe and the People Without History.  Berkely: University

of California Press.

World Tables (I.B.R.D. ) 1983. Baltimore: Johns Hopkins University Press.

Yoffee, Norman 1988 “The collapse of ancient Mesopotamian states and civilization” Pp. 44-101 in Norman Yoffee and George C. Cowgill (eds.) The Collapse of Ancient States and Civilizations.  Tucson: University of Arizona Press.

Zagarell , Allen. 1986. "Trade, women, class and society in Ancient Western Asia. "

Current Anthropology 27,5:415-30 (December).



[2] The term "mode of production" is used here in the general sense of the deep

structural logic of production, distribution, exchange and accumulation

it does not restrict the focus of analysis to the point of production.


[3] Because the world -systems perspective was developed mainly to understand

the Europe-centered capitalist world-economy some of the key concepts need

to be redefined in order to make them appropriate for the task of comparing

very different world-systems. The definition of world-systems,

core I periphery hierarchies, and methods of spatially bounding world -systems

all require revision. These problems are confronted in Chase- Dunn and Hall



[4] Important statements of the substantivist position by Polanyi and his

colleagues are contained in Polanyi, Arensberg and Pearson, eds. (1957) .

Marxist variants have been argued by Sahlins(1972) and Wolf( 1982) .


[5] The formalist perspective has been defended in a recent study by Blanton ,

Kowalewski, Feinman and Appel(1981) .Curtin's(1984) studyof cross cultural

trade modllies ( and improves) the formalist position by adding some new

concepts: the trade diaspora and trade ecumenes. Gledhill and Larsen(1982)

present a review and critique of the application of Polanyi's approach to the

study of societies in the ancient Near East.


[6] A helpful review of the older literature on this debate is that by



[7] Sahlins (1972) demonstrates the necessity of the division of labor among

different horticulturalist and fishing societies in the Huon Gulf .Kelly( 1985 )

analyses a fascinating case of "tribal imperialism" based on systematic

raiding in the Nuer-Dinka relationship. Friedman and Rowlands(1977) and

Galley( 1987) show how intersocietal and gender relations interact in

emergence of complex chiefdoms and early states.    


[8] Dodgshon's(1987) fascinating study of social evolution and spatial order is

entirely relevant to the consideration of the causes and consequences of

cities and city systems, and Dodgshon also sheds important light on the way

in which nomadic hunter-gathers approach space, but he misses the

interesting category of sedentary foragers .



[9] Sahlins(1972:144-8) describes the rise and fall of chiefdoms in Hawaii and

Mann(1986:Chapter 2) discusses this topic as resistance to power among

"prehistoric" peoples based on his reading of archaeological studies in

Northern Europe.


[10] This argument is also made by Kristiansen(1987) for bronze age Scandinavia. Peregrine (1991) applies the prestige goods economy paradigm to Cahokia-centered Mississippian

world-system of middle North America.


[11] It is hazardous to infer social hierarchy from spatial hierarchy as many

archaeologists do, but settlement systems are an important feature of social

systems even if there is no direct equivalence between space and power .


[12] A helpful compilation of city population sizes for the Near Eastern-centered

world-system from 2250 B.C. to 1478 A.D. is presented in

Wilkinson(1987:Tables 2 & 3) .Wilkinson (1990) also has compiled lists of all

the major cities in each of fourteen civilizations/world-systems.


[13] The distinction between pristine (or primary) states and secondary states

focusses on the differences between the emergence of a state in a context

in which there are no other such organizations vs. the emergence of states

in interaction with other states. The same kind of distinction can be applied

to other organizational forms ( .e .g. chiefdoms, etc. )


[14] In the Egyptian world-system empire formation occurred much more quickly,

leading to a complicated discussion about the similarities and differences

between the Mesopotamian and Egyptian cases (Friedman and Rowlands,1977;

Mann,1986) .It is generally agreed that the emergence of "civilization" in

Egypt was, despite some similarities, substantially independent from the

somewhat earlier developments in Mesopotamia, thus constituting a truly

"pristine" case of state formation. The other cases of pristine or primary

state formation upon which there is substantial agreement are the lndus

River valley, China, Mesoamerica and Peru.


[15] Those who pay attention to core/periphery hierarchies are chary about the

application of the term "civilization" because it is often part of the

ideological justification for the exploitation and domination of "barbarian " or

"savage" peoples. Contrary to the notion that simpler, less

hierarchical human societies were "precultural" is the ethnographic evidence that these

societies were integrated primarily on the basis of normatively enforced

rights and obligations based on consensually held definitions of reality and

the good. In this sense they were more completely " cultural. " than larger

civilizations which need coercive organizations and the "crass cash nexus"

to function .



[16] I am employing terms used by Myrdal(1971) --spread effects and backwash

effects --in his analysis of regional patterns of uneven development. The

relative strength and basis of such affects differ across world-systems and

part of the motivation for comparing different kinds of core/periphery

relations is to enable us to understand why "co-evolution" occurs in some

systems while the development of underdevelopment occurs in others .


[17] In the famous discussion of the method of political economy in the Grundrisse

 Marx (1973:108)says,"The purity {abstract specificity) in which trading

peoples --Phoenicians , Carthaginians --appear in the old world is

determined precisely by the predominance of the agricultural peoples.

Capital, as trading-capital or as money-capital, appears in this abstraction

precisely where capital is not yet the predominant element of societies.


[18] I dare to join those who question whether or not the rise of the West was

purely occidental.


[19] Lane's(1979) concept of protection rent refers to returns to merchants and

production capitalists who benefit from the provision of protection for their

operations by a state that operates effectively and efficiently and provides

these services at cost. Mann(1986) uses the term protection rent in a

completely different sense to refer to the taxes and tributes which tributary

states place on merchants. Lane's concept provides the basis for

understanding capitalist states, while Mann's usage describes the operation

of tributary states.


[20] The respecified definition of world -system connections proposed by Chase-

Dunn and Ha1l (1991) for the comparative study of world-systems includes

basic goods exchanges (food and raw materials) , prestige good exchanges

(Wallerstein's "preciosities") and interconnections formed by on-going

political/military interactions . This last form of interconnectedness is

proposed and empirically applied by Wilkinson(1987) in his study of how

Near Eastern civilizations merged and expanded to become the now-global

"Central Civilization."


[21] An argument for the connection between the resistance of the modern

interstate system to empire formation and the emergent predominance of

capitalism is made in Chase-Dunn(1990) .


[22] Tilly (1989:17) defines capitalism as "any tangible mobile resources, and enforceable claims on such resources." Does this mean the cattle among the Nuer are capital? This definition ignores the importance of price-setting markets and commodified wealth, as well as production utilizing commodified labor.



[23] I used the boundaries of the European world-system proposed by the

Braudel Center scholars for my analysis of changes in the world city-size

distribution. The redefinition of world-system boundaries for the purposes

of comparing different world -systems proposed in Chase- Dunn and Hall

(1991) would change my results to some extent. For example Constantinople

was excluded from the European world -system in my analysis because of

Immanuel Wallerstein's argument that the Ottoman Empire was a separate

world-system. Wallerstein's contention is based on his usage of mode of

production as a criterion for spatially bounding world -systems, a practice

I have criticized and rejected ( Chase- Dunn ,1989 : Chapter 1) .By the method

of bounding proposed in Chase-Dunn and Hall the European cities and

Constantinople were part of the same system because of extensive trade in

basic goods, prestige goods and political military interaction .

The very large size of Constantinople would undoubtedly alter the outcome of my

analysis to some extent. A new study of long-run changes in the world city

system is needed.


[24] World -system scholars have been analyzing the reintegration of the socialist

states into the world market for some time (e.g. Chase-Dunn,1982b) .The

recent changes may be understood, as Wallerstein(1989) says of the French

Revolution, as the political and ideological superstructure catching up with

the economic base.



[25] One reason I am thinking about the possibility of cities as agents of the

transformation to socialism is because in the United States there have been only

two major recent accomplishments in progressive politics: the Jackson campaigns

of 1982 and 1988, and the electoral victories of progressives in Berkeley, Madison ,

Santa Monica and Burlington. Unfortunately these municipal victories have

illustrated the limitations of local approaches as much as their possibilities for

transformation. Municipal socialists in the U. S., as in Bologna, end up competing

with other players to efficiently provide urban services in a context in which

capital and its state hold most of the cards.

Certainly the “objective conditions” of declining industrial cities in the U. S .

ought to provide fertile ground for socialist politics. As the documentary

film , Roger and Me, illustrates tourism and "flexible accumulation" have little hope

of success in such places as Flint, Michigan. But the film also shows the many

ways in which victims blame themselves and adopt strategies of survival rather than struggle.

Peter Taylor's (1990) study of Britain shows how a declining hegemon can cool people out by revitalizing the imperial ideology .Reagan and Bush have shown how it is possible to get the losers to identify with the winners .  But this kind of smoke and mirrors cannot succeed for ever.



[26] Elsewhere (Chase-Dunn,1989,1990c) I have argued that the world-system

perspective implies that the semiperiphery is the weak link and that new

challenges to capitalism are likely to again emerge from that zone.