C. Chase-Dunn, Th. D. Hall

"Cross-World-System Comparisons: Similarities and Differences"

World-systems are intersocietal networks of regularized and systemically important competitive and cooperative interaction. It is often the case that such networks are nested and overlapping because different kinds of interaction have different geographical patterns. When we say that world-systems range from small to global we mean that populations are linked by networks that vary in size and with regard to the spatial extent of interactional consequences. These definitional and conceptual matters are described and defended in Chase-Dunn and Hall (1993). In this paper we will formulate hypotheses about similarities and differences among different types of world-systems based on the project of comparing world-systems. Concepts and hypotheses are still being debated, while case studies of single world-systems are being completed. What will eventually be needed is systematic study of a large number of world-systems. It is toward this eventual end that this paper is focused.

We need to have explicit hypotheses about similarities and differences among world-systems and we need to construct a comparative framework for these. We employ the typology advanced in Chase-Dunn and Hall (1993, pp. 866-71). True cross-world-system research would have access to comparable data on at least ten of each of our ten types of world-systems (see Table 4.1). As world-systems get larger they become fewer. Thus it is not possible to have ten whole capitalist world-economies because there has been only one. This "sample" of world-systems would number about sixty Hypotheses about similarities and differences and the "typical" parameters of systems can be systematically confronted with evidence once we have such a comparative data set. As it is. we must put forth our generalizations based on the patchy evidence of case studies and the comparison of those state-based systems on which we have some comparable evidence.

This paper suggests several hypotheses about similarities and differences across the different kinds of world-systems specified in our typology We set up a framework for comparisons and we specify some hypothetical parameters of network size and demographic size. This presumes our conceptual approach to world-system spatial boundaries, in which we distinguish between bulk goods, political/military, and prestige goods interaction networks (sec Chase-Dunn and Hall 1993, pp. 858-62).


Our structural approach to studying world-systems abstracts from population size and territorial scale to-compare the structural patterns of very small systems with very large ones. Nevertheless, it is worthwhile to consider the "typical" sizes of different kinds of world-systems and to hypothesize how different interaction networks vary in relative size in different sorts of systems.

It is likely to be the case that the sizes of different kinds of interaction networks vary in their ratios to one another across different kinds of world-systems. Table 4.1 shows rough estimates of "typical" population sizes of types of polities, bulk goods networks, intermarriage networks, and prestige goods networks across our world-system types ranging from small to large. By "typical" we mean to hypothesize a central tendency for each type of world-system while recognizing that there are probably wide variations among different real systems of each type. The numbers in Table 4.1 are hypothetical means for the universe of world-systems of each type. Within each type there is undoubtedly considerable variation, and this is likely to be systematically related to the developmental trajectories of particular systems. Additionally, all systems experience "pulses" of expansion and contraction of population and territorial size.

Furthermore, the relative sizes of different networks may differ depending upon the prior developmental history of each system. For example, those simple chiefdom systems that have never been through a phase of complex chieftainship may differ significantly in terms of the size of their trade networks from those simple chiefdoms that have devolved from former prestige goods-based chiefdoms (e.g., Friedman 1981). Such complications need to be subjected to careful cross-world-system research, but for now the following central tendencies are our best hypotheses.

The hypothesized polity sizes are for the largest polities in each system. As world-systems have gotten larger the different interaction networks have converged in terms of both population size and territorial extent, and simpler systems have been either eliminated or incorporated into larger, more complex, and more hierarchical systems. Table 4.1 is a heuristic suggesting research questions for comparative studies of interaction networks in different kinds of systems.

Population Sizes of Polities and Interaction Nets Across Different Types of World-Systems

Type of World-System


Bulk Goods


Military (PMN)

Prestige Goods

Nomadic foragers






Sedentary foragers






Big Man






Simple chiefdoms






Complex chiefdoms





1 million

Primary states




1 million

2 million

Primary empires




2 million

4 million

Secondary empires


1.6 million

1.6 million

3 million

6 million

Commercializing systems

50 million

75 million

20 million

100 million

200 million

Modem world-system

25 million

150 million

25 million

500 million

700 million

(AD 1750)













Another way of comparing world-systems is to map spatiotemporally the boundaries of different systems as they merged to become the global modem world-system. Here we will adopt the approach developed by David Wilkin-son (1987 [chapter 2, this volume], 1991). Wilkinson bounds his "world systems/civilizations" by means of regularized political/military interaction networks. All states and regions that are regularly engaged in either military conflict or alliance with one another are part of the same political/military network (PMN). Using this rule Wilkinson has produced a chronograph (chapter 2, this volume. Figure 2.1). This shows how thirteen political/military networks containing cities and states were engulfed by what we shall call the "Central PMN." The Central PMN was formed in the merger between the Mesopotamian and Egyptian PMNs in about 1500 BC. Wilkinson, coming from a comparative civilizations approach, excludes systems that do not have cities larger than 10,000 members. Thus, although all stateless and cityless world-systems also have PMNs, those without large cities are not included in Wilkinson's chronograph. If they were included, each of the tributary branches of the Central PMN would be composed of several smaller branches that came together at an earlier time.

Wilkinson's chronograph shows when the several separate PMNs became integrated into the Central PMN. Wilkinson (1992b, 1993) has also designated the spatial boundaries of larger "oikumenes," or trade networks, and the points in time at which they merged. These correspond to the prestige goods networks of our multicriteria approach to spatial boundaries.2 If we were to redraw Wilkinson's chronograph using the boundaries of'oikumenes" it would look quite similar except that the dates along the left margin would shift down because different regions were usually linked by trade earlier than they were linked into the same political/military network. For example, the linkage of the separate Egyptian and Mesopotamian political/military networks into the same oikumene trade network occurred before 2250 BC, whereas their merger into a single political/military interaction network did not occur until 1500 EC.

From our point of view both prestige goods trade and political/military networks are important for understanding the operation of whole world-systems, so we do not choose to call one or the other the "real" world-system. Rather; the whole system from the perspective of any location is the nested set of networks which impinge systemically upon that place. What Wilkinson calls "civilizations" we will call "PMNs". We shall refer to his trade oikumenes as "prestige goods networks" (PGNs). We also want to consider the importance of bulk goods networks—trade networks that include the flows of basic goods such as foods and everyday raw materials. We shall call Wilkinson's Central Civilization the Central PMN, and his "Old Oikumene" we shall call the Central prestige goods network (PGN).

Here we will summarize Wilkinson's (1992b, 1993) boundaries of trade oikumenes in order to paint a picture of the expansion of the Central PGN. As we have already mentioned, the Egyptian and Mesopotamian PMNs were already joined into the single Central prestige goods network in 2250 BC. This merger of prestige goods nets probably occurred as early as 3000 BC (Marfoe 1987). By 2000 BC the Central PGN had grown to include more of northeast Africa, western Asia, and the Aegean (Wilkinson 1992b, p. 59). By 1800 BC the Central PGN included also the PMN centered in the Indus River valley, which Wilkinson (1992b, p. 60) calls "Indie civilization." In 1600 BC the Central PGN included also the cities of the Aegean PMN on Crete (Wilkinson 1992b, pp. 61-62), but the cities of the Indus had disappeared and Wilkinson indicates that this region was no longer in the Central PGN. In 1360 BC the Mesopotamian and Egyptian PMNs had joined to create the Central PMN, and the Central PGN had expanded to include more of the Mediterranean littoral and had moved back into western India despite the fact that there were no large cities there (Wilkinson 1992b,p. 63) A separate citified Far Eastern PMN had arisen in China. In 1200 BC the Central PGN had shrunk once again in both the east and the west. India was out, as was the western Mediterranean (Wilkinson 1992b, p. 66). This is an instance of the phenomenon of "pulsation" in which world-systems experience waves of territorial expansion followed by either slower expansion or contraction. In 1200 BC there were separate PMNs and PGNs in India and China.

In 1000 BC the Central PGN had only a single PMN because Central PMN had incorporated the Aegean PMN. The boundaries of the Central PGN had moved south to link with the city of Saba at the southern tip of the Arabian peninsula, but the eastern and western boundaries were about the same as they had been in 1200 BC. By 800 BC not much had changed regarding the boundaries of Central PGN according to Wilkinson. By 650 BC the city ofNapata in Nubia had emerged, extending the system south and Wilkinson designates Miletus as being in a separate Aegean PMN that was engulfed by the Central PMN shortly thereafter (Wilkinson 1992b, pp. 69-70). By 430 BC the Central PGN and the Central PMN had begun again to spread east and west. Carthage was part of the Central PGN. Though Wilkinson shows some movement east, the Indie PMN was still characterized as having its own separate oikumene. In Mesoamerica we have the emergence of a new and separate system indicated by the city of Cuicuilco (Wilkinson 1992b, p. 71). Wilkinson notes that "this table is the first to have a notable presence of cities which, though surely or most likely capitals of states, had such small states and/or hinterlands to extract from that they must have flourished (demographically) and lived largely by trade: most Greek and Phoenician cities are of this character" (Wilkinson 1992b, pp. 70-71).

By 200 BC the Central PGN trade net incorporated the Indie region while the PMNs of Central and Indie remained separate. Far Eastern was still outside of the Central PGN according to Wilkinson. This was also the situation in AD 100 except that the Central PMN and the Central PGN had moved northwest a bit to incorporate England. Regarding the relationship between the Far Eastern and the Central PGN, Wilkinson (1992b, pp. 75-76) says:

The Later Han dynasty is weakening at the center, but has recently extended its power in Turkestan and opened an important silk trade with Rome. Accordingly, it is possible... that Far Eastern civilization has now been linked into die Old Oikumene via Rayy, Merv and Balkh. But I have delayed acknowledging that link until AD 622, when the growth of connector cities in Central Asia makes the case more persuasive.

We suggest that this is another case of pulsation, similar in form to the expansion and contraction of the Central PGN mentioned earlier. The Silk Road trade has been shown to have had important consequences for both Han China and Rome (Teggart 1939). The fact that this trade declined with the decline of the major terminus empires is not surprising and the Eurasian PGN once again became two disconnected systems, not to be firmly attached again for another 400 years. By AD 500 Wilkinson contends that the Central PGN had contracted slightly on its western frontier. The British isles had dropped out with the collapse of the Roman Empire.

In AD 622 Wilkinson shows a single Afroeurasian PGN that included North Africa, Latinized Europe, India, Central Asia, India and Ceylon, China, Japan, the northern Philippines, Sinified Southeast Asia, but not the Malay peninsula or Indonesia. Korea was included in the Far Eastern PMN and in the Central PGN. By AD 800 the Central PGN had fallen back a bit in the northwest, but it had moved southwest into savannah Africa, where it joined with the still separate West African PMN. The Central PGN had also expanded southeast to include by trade the still separate Indonesian PMN. On the eastern edge, within the trade net, was a separate Japanese PMN. In AD 900 the Central PGN had about the same boundaries, but the West African PMN had been incorporated into the Central PMN.

In AD 1000 the boundaries of the Central PGN were nearly the same, but the Central PMN had expanded a bit closer to India. In AD 1100 the boundaries of the Central PGN were about the same, but a new citified political/military network, the Mississippian PMN, centered at Cahokia (East St. Louis), had emerged in North America. Wilkinson (1992b, p. 86) is unsure about whether the Mississippian had its own PGN or was a northern extension of the Mexican PGN. In AD 1200 there were no longer sizable cities in West Africa so Wilkinson does not indicate a citified PMN in that region. By AD 1300 the Central PGN had expanded its boundaries further into Africa and the West African PMN had reappeared. Now the Central PMN was even more separated from the Indie PMN than it had been because of the decline of intervening and linking cities (Wilkinson 1993, p. 43). The citified Indonesian PMN had reappeared.

By AD 1400 the boundaries of the Central PGN remained about the same, but Central PMN had moved again back toward India with the emergence of Samarkand. Large cities had disappeared from the East Indies so Wilkinson does not show an Indonesian PMN. A new citified PMN appeared in Peru, and Wilkinson is unsure whether or not it was part of a larger New World PGN or had its own separate trade net (Wilkinson 1993, p. 5). In AD 1500 the Central PGN was still about the same. This was the last point at which the Indie PMN could be seen as separate from the Central PMN. There were still no large cities in the East Indies.

By AD 1600 the Central PGN and the Central PMN had expanded to include the New World Mexican, Peruvian, and Chibchan PMNs. Chibchan was a separate citified PMN that Wilkinson designates as having arisen in Colombia. The Central PGN now encompassed all the large cities of the world. The Central PMN had incorporated the Indie and West African PMNs. By AD 1700 things were much the same except that "Tokugawa isolationism has taken Japan out of the Old Oikumene and into an oikumene of its own, where it remains . . . until 1900" (Wilkinson 1993, p. 49). Of course Wilkinson's focus on large cities obscures the rapid incorporation of many new areas in which there were no cities. By 1800 things were much the same regarding the cities, but the Old Oikumene and the Central PMN had continued to expand into nonurbanized regions. European explorers and traders incorporated much of Oceania, North America, and central and southern Africa. By 1900 the Japanese PMN was included in the Central PGN. Wilkinson still shows separate Japanese and Ear Eastern PMNs as of 1900, but mentions that these were both incorporated into the Central PMN by the time of World War I. He says that by 1900 we have "the (approximate) end of all oikumenes but the Old Oikumene and (approximate) end of all civilizations but Central civilization" (Wilkinson 1993, p. 51). There may still have been a few isolated stateless and cityless regions outside of the modem world-system, but Wilkinson does not consider these. In summary Wilkinson (1993, p. 56) says:

A world economy, lacking a coextensive world polity, but containing world polities of smaller area than its own, existed from (at least) the 4th millennium BC (when it linked the world polities of Egyptian and Mesopotamian civilizations) to the 19th century AD (when a world polity became global, and coextensive with the world economy that had theretofore contained it.) Other such "oikumenes," trade-linked but not polirico-militarily bonded, probably connected Chibchan with Peruvian civilization, and may have linked Mexican with Mississippian and/or Mexican with Peruvian civilization. But it is particularly noteworthy that Central civilization, from c. 1500 BC to c. AD 1900, formed a politically coherent social system smaller than, nested within, expanding in pace with and into the space pioneered by, an economically coherent but politically unlinked oikumene.

This is the big picture regarding the spreading engulfment of the earth by the Central prestige goods network and the Central political/military network. The details need finer-grained research done on a comparative basis. And the whole chronograph needs to consider the incorporation of stateless and cityless regions as well as the links between those regions in which there were cities. These details can be important for understanding the historical developments that occurred in particular regions, but for comprehending the big picture Wilkinson's monumental work is quite sufficient.


What are the developmental patterns that can be observed in all world-systems? Remember that our universe of world-systems includes those very small-scale systems composed of sedentary foragers as well as the contemporary global political economy. All world-systems pulsate in the sense that the spatial scale of integration, especially by trade, gets larger and then smaller again. During the enlarging phase trade networks grow in territorial size and become more dense in terms of the frequency of transactions. During the declining phase, trade falls off, local areas become less connected, and they reorganize around self-sufficiency Local identities and the differentiation of identities between local groups and outsiders are emphasized. We saw several examples of pulsation in the overview of the expansion of the Central PGN above.. The point here is thataW world-systems experience these sequences of expansion and contraction, even very small and egalitarian ones. In hierarchical world-systems these horizontal sequences of linkage and delinkage become entwined with processes of the rise and fall of hierarchies. In these, the processes of universal-local identity formation become linked with the rise and fall of hegemonies (Friedman 1994).

In the next section we shall discuss a pattern which is widespread but not universal: the rise and fall of hierarchy and of the sizes of individual polities. That pattern does not occur in very egalitarian world-systems because there are no hierarchies to rise and fall. Of course there are some inequalities even within and between societies of sedentary foragers. But ethnographers have never observed cycles of increasing and decreasing inequality in such very egalitarian societies. The rise and fall patterns begin in earnest in chiefdoms. But pulsation occurs in all systems.

In systems of sedentary foragers trade networks are institutionalized as down-the-line exchanges of gifts among local leaders. Goods move long distances but there arc no long-distance traders. In California we saw archaeological evidence for pulsation cycles as the rise of one trade network, its decline, and the rise of a new, larger trade network (Chase-Dunn 1993). Thomas L. Jackson (1992) links the declines to periods of "localization" in prehistoric California. Similar sequences have been noted in Mesolithic Europe (Price 1991).

In the modem world-system we have had the crisis of the 17th century, and periodic waves in international economic integration followed by declines of trade and emphasis on national autarchy This system also continues to experience spatial expansion and contraction despite the fact that it encompasses the whole globe. Pressures for the exploration and exploitation of the sea beds, the moon, and outer space obviously vary conjointly with economic cycles. We may ask if there are similar relationships between pulsation and cycles of the rise and fall of hierarchy in different kinds of world-systems? We shall return to that question after we have considered the matter of rise and fall.

Rise and Fall

All hierarchical intersocietal systems go through sequences ofcentralizarior and decentralization of economic, political, and social power. Like states chiefdoms emerge in sets in which chiefly polities interact and compete witr one another, and these "interchiefdom systems" exhibit a pattern of rise anc fall in which the territorial and population size of the largest polities rise; and then declines (Sahlins 1972, pp. 144-48; Mann 1986. ch. 2; Friedmar and Rowlands 1978). The dynamics of this sequence in systems composed of chiefdoms, in which power is organized around hierarchical kinshij relations, diner in important ways from the dynamics of rise and fall, politics centralization and decentralization, that operate in systems composed o true states.3 Some chiefdoms develop techniques of power which supple ment and go beyond the metaphors of hierarchical kinship. Thus singl polities become able to exercise control over larger areas and the interactiol networks composed of these larger states grow larger and more densel interconnected.

The cycle of the rise and fall of states occurs in all known interstate systems. In some the competition among a set of states within a single coi region of a world-system takes the form of the rise and fall of hegemon core powers, a process which we know well in the modern world-systen In others, and more frequently, the cycle of political centralizarion/decentralization takes the form of an alternation between interstate systems (i which there are a number of competing stares within a core area, calle "states systems" by Wilkinson [1991, pp. 116-21]) and world-empires i which a single state succeeds in unifying an entire core area by means ( conquest (called universal empires or world states by Wilkinson). Wilkinson (1992b, p. 54) provides us with periodizations of states systems and world states for eleven state-based PMNs.

In addition to the cycles of the rise and fall of polities, there is a long-n trend toward the increasing size of polities and the decreasing number autonomous polities on earth (Carneiro 1978). Rein Taagepera's (1978 1978b, 1979) studies of changes in the territorial size of the largest empires on earth over the past four thousand years demonstrates the cycles political centralization and decentralization previously discussed. The combination of the long-term trend of increasing size of polities with t medium-term process of political centralization/decentralization is illustrated in Figure 4.1. Taagepera's studies show that the size of the larger empire on earth oscillated for long periods and then jumped up in rapid rises corresponding to the wide conquests of semiperipheral marcher states, which created empires across whole core regions. Well-known examples are the Akkadian Empire, the Assyrian Empire, the Alexandrian conquests, and the Roman Empire. Figure 4.1 is a simplified and idealized model based on Taagepera's studies of the territorial size of the largest states and empires on earth and Carneiro's (1978) discussion of the long-term evolutionary trend from many small polities to few large ones.

Of course Figure 4.1 should not be read to imply that each system experiences such a pattern. This is the pattern we get when we plot the largest systems on earth over time on the same graph. Some systems never rise, for example the sedentary foragers of California. Others rise and then decline and never rise again. Friedman's (1981) model of Melanesian systems understands them as former prestige goods chiefdoms that have devolved when trade networks became so dense that hierarchies based on monopolizing imports could no longer be sustained. Even though some individual systems do not exhibit the rise and fall pattern, many experience it repeatedly and also undergo the occasional qualitative leaps in polity size indicated in the graph. These are of interest not only because they change the spatial scale of states and empires, but also because they are associated with the development of new "techniques of power" (Mann 1986) that allow rulers to extract resources over a much broader territory.

It is also the case that the strategies of players at the different size levels in Figure 4.1 arc qualitatively different. Thus the strategies appropriate to the creation of a complex chiefdom are quite different from those that will work to create an empire out of separate states or to consolidate hegemony over other core states in the modem world-system. It is also undoubtedly the case that there are qualitatively different strategies that work under different circumstances in the same world-system types. Thus Friedman notes that prestige goods systems allow for the creation of complex chiefdoms when long-distance trade is feasible, but that a very different strategy is necessary in regions in which regularized long-distance trade between quite different polities is more costly, and therefore irregular or lacking. So prestige goods chiefdoms emerged in Melanesia and western Polynesia where inter archipelago trade was more feasible, whereas in eastern Polynesia and Hawaii, where long-distance trade was much more difficult, large and hierarchical polities had to rely on the ability of the chiefly class to control access to land and other resources. Blanton, Feinman, Kowalewski, and Peregrine (forthcoming) have analyzed two distinct strategies that interacted in the development of the Mesoamerican world-system. Certainly there are different strategies that are appropriate to the rise of hegemons in the modem world-system. It will not do to simply copy what was done before. But the differences among strategies within similar world-systems (i.e., chiefdom-based systems) are probably smaller than the differences between strategies in completely different types of world-systems.

Andre Gunder Frank and Barry Gills (1993; Gills and Frank 1991) build on the work of Kajsa Ekholm and Jonathan Friedman (1982) to conceptualize and study the sequences of political centralization and decentralization that have occurred in the Central world system over the last 5,000 years. This sequence is well-known to students of political history as the rise and fall of empires. The sequence of political centralization/decentralization is a prime example of a continuity between the modem world-system of the last 500 years and the earlier Central system. Indeed, even chiefdom-based world-systems exhibit a somewhat similar pattern. But these processes also exhibit important differences in different kinds of systems. Both chiefdom systems and state-based systems become centralized through military conquest, but the polities erected by chiefly conquerors must rely on kinship alliances in order to implement regional control, while states make use of specialized nonkin control institutions. This is why state-based empires usually were able to incorporate larger territories and populations than chiefdom-based polities did.

In the modern world-system the pattern of political centralization/decentralization takes the form of the rise and fall of hegemonic core powers. This is analytically similar to the rise and fall of empires, but the differences are important. In the process of empire-formation a "rogue power" - most often a semiperipheral marcher state - conquers the other core states to form what Wilkinson (1987) terms a "universal empire." Well-known examples are the Roman Empire and the Han Empire. This phenomenon corresponds roughly to what Immanuel Wallerstein calls a "world-empire." Wallerstein claims that the modern world-system is politically structured as an interstate system of competing states, while earlier world-systems frequently took the form of world-empires in which the economic division of labor came to be encompassed by a single state.

Rarely did the "universal states" encompass entire world-economics, but they often did conquer an entire adjacent core region. This is the peak of political centralization in such systems. It is convenient to conceptualize centralization and decentralization as two ends of a continuum. Thus, though there have not been true "world-empires" in the sense that a single state comes to encompass an entire world-system, this idea may be understood to point to a relatively high concentration of control over a relatively great proportion of a world-system. The term we prefer, because it is more precise, is "core-wide empire."4 State-based world-systems prior to the modem one oscillated between core-wide empires and interstate systems. In some regions the decentralization trend went so far as to break up into mini-states. Thus feudalism may be understood as a very decentralized form of a state-based system and a hegemonic core state. This may be understood as simply a difference in the degree of the concentration of political/military power in a single state. It is in this sense that Wallerstein's distinction between world-empires and world-economies points to an important structural difference between the modem system and earlier state-based systems. But this is not only a systematic difference in the degree of peak political concentration. The whole nature of the process of rise and fall is different in the modem world-system. The rise and fall of hegemonies has occurred in a very different way from the rise and fall of empires. Empire formation was a matter of conquering adjacent core states. The rise of modem hegemons did not occur in this way Modem hegemons did not conquer adjacent core states to extract taxes and tribute. Rather they sought to control international trade, especially oceanic trade, that linked cores with peripheries.

This is why the modern world-system is resistant to empire formation. The most powerful state in the system acts to block empire formation and to preserve the interstate system. Thus the cycle of political centralization/decentralization takes the form of the rise and fall of hegemonic core powers. This important difference is due primarily to the relatively great importance that capitalist accumulation has for the modern system.

Settlement Systems

All world-systems have settlement systems in which people locate themselves in space in a certain way and regulate the use of land and natural resources. Our broadly comparative perspective on world-systems suggests that we examine the structural similarities and differences of the settlement systems found in each of our world-system types. As with other differences in size and scale, we move from the village system of sedentary foragers to the urbanized world city system of today5 Sedentary foragers lived in rather permanent winter villages, but they traveled to temporary camps during the summer. The settlement systems of sedentary foragers differ in terms of their degree of size hierarchy Some formed clear two-tiered size hierarchies in which larger villages were regularly separated by smaller hamlets, while others had no regular distribution of larger and smaller settlements. In general, polity size and social hierarchy are reflected in the size hierarchy of the settlement system (e.g., Lightfoot and Feinman 1982; Nissen 1988, p. 41). But this is not always the case. Some rather densely populated and hierarchical societies do not concentrate people together in villages (e.g., ancient Hawaii). Thus it is not always possible to assume a one-to-one correlation between settlement size hierarchies and political hierarchies.

Settlement Size Hierarchies

What is a "city-size hierarchy?" All human settlements interact with other settlements. The size of individual settlements can be studied as they grow or decrease, and the relative size of settlements can also be studied. This means looking at the distribution of settlement sizes within a region. Some regions contain settlement systems that are very hierarchical in the sense that there is a single very large settlement that is surrounded by very small settlements. Such settlement systems are called "primate" because there is a single center that is much larger than any other settlement. Geographers have developed theories suggesting that a "normal" settlement size hierarchy will correspond to the rank-size rule in which the second largest settlement is half the size of the largest, the third largest is one-third the size of the largest, the fourth largest is one-fourth the size of the largest, and so on. The rank-size rule is also called the "log normal" rule because the distribution of settlement sizes approximates a straight line when the settlement sizes are logarithmically transformed and plotted.

Some settlement systems are "flat" in the sense that the towns or cities or villages of which they arc composed arc all about the same size. So we can discuss different settlement systems as primate, rank-size, or flat depending upon the relative size of the settlements of which they are composed. The size hierarchy aspect allows us to compare very different kinds of settlement systems to one another because we arc looking at the relative, rather than the absolute, sizes of settlements. Thus a system composed of villages can be just as hierarchical as a system composed of great cities if one village is much larger than the other villages.

In order to make such relative comparisons, a statistic called the Standardized Primacy Index (SPI) was developed by Pamela Walters (1985). The SPI takes a value of zero when a settlement size distribution corresponds to the rank-size rule. It takes on negative values when the distribution is less hierarchical (flatter) than the rank-size rule and positive values when the distribution is more hierarchical than the rank-size rule (primacy). Though the measures of urban populations arc subject to error, and to greater error as we go further back in time, the SPI is less sensitive to absolute errors because it examines the differences in the population sizes of settlements.

Here we will report the results and conclusions of a research effort that examined the relationship between processes of political centralization/decentralization and changes in the relative population sizes of cities located within several separate political/military networks (Chase-Dunn and Willard 1993). The simplest hypothesis is that city systems will become more hierarchical - that is, the largest cities will be much larger than other cities in the same network—when political/military power is more centralized.

This is based on the idea that political power is an important component of the ability of large cities to gather the resources necessary to sustain large populations. On the other hand, when political power is relatively less centralized within a region, cities should be more equal in size and so the size distribution of cities should be less hierarchical.

The phenomenon of urban primacy - the concentration of population in a very large central city with only much smaller cities in the same region - has been extensively studied in national societies in the modern world-system (Chase-Dunn 1985a, 1985b). It is well known that France has a very primate city-size distribution, as do most peripheral and semiperipheral countries in the modern world-system. Urban primacy is seen as a problem by many contemporary urban and regional planners and they have constructed and tried to implement policies for encouraging the growth of small and middle-sized cities (rather than further increasing the size of the largest city in a country).

The Chase-Dunn and Willard study used data on the populations of cities and the relative population sizes of cities within regions defined as networks of political/military interaction. They studied the relationship between changes in the distribution of city sizes, the growth and decline of the largest cities, the rise and decline of centralized polities, and hypothesized long cycles of political/economic expansion and contraction. Gills and Frank (1991) and Frank (1993) have periodized the history of their 5,000-year Eurasian world system into phases of expansion and contraction. Wilkinson (1992a) has examined the relationship between the periodizations proposed by Gills and Frank and changes in the rate of city population growth. He used data from Chandler's compilation of city populations, Four Thousand Tears of Urban Growth (1987), to study the sizes and number of the largest cities in both political/military interaction networks and larger prestige goods trade networks. Wilkinson found some support for the expansion/contraction periodization proposed by Gills and Frank. The Chase-Dunn and Willard study used the city population data from Chandler in a somewhat different way to examine the Gills-Frank periodizations. And they examined changes in city-size hierarchies of several PMNs over time with Wilkinson's periodization of these into "states systems" and "universal states." This last examines the relationship between settlement size hierarchies and the cycle of political centralization/decentralization previously discussed.

Power and the City-Size Distribution

Why should a city system have a steeper city-size distribution when there is a greater degree of concentration of power? The simple answer is that large settlements and especially large cities require greater concentrations of resources to support their large populations. This is why population size has itself been suggested as an indicator of power (Taagepera 1978a, p. 111). But these resources may be obtainable locally, and the settlement-size hierarchy may simply correspond to the distribution of ecologically determined resources. For example, in a desert environment populations cluster near oases. It is not the political or economic power of the central settlement over surrounding areas that produces a centralized settlement system, but rather the ecological distribution of necessary or desirable resources. In many systems, however, we have reason to believe that relations of power, domination, and exploitation do affect the distribution of human populations in space. Many large cities are as large as they are because they are able to draw upon far-flung regions for food and raw materials. If a city is able to use political/military power or economic power to acquire resources from surrounding cities it will be able to support a larger population than the dominated cities can, and this should produce a hierarchical city-size distribution.

Of course the effect can also go the other way Some cities can dominate others because they have larger populations. Great population size makes possible the assembly of large armies or navies, and this may be an important factor creating or reinforcing steep city-size distributions.

The Chase-Dunn and Willard research examined the extent to which changes in the degree of population-size hierarchies in city systems correspond or do not correspond with changes in the degree of political centralization. They also examined the hypothesized expansion/stagnation phases proposed by Frank (1993) by observing changes in the size of the urban population. For information about political centralization, Chase-Dunn and Willard used several sources dealing with historical changes in the relationship among states in each PMN.

As stated above, the primary unit of analysis in the Chase-Dunn and Willard study is the political/military interaction network (PMN) as constructed by Wilkinson. Is the PMN the best unit of analysis for studying city systems? We will argue that it is a good unit of analysis, but that it would also be desirable to use PGNs and bulk goods networks and to compare the results with those based on PMNs. For the present, the boundaries ofPMNs are well enough specified that we can easily use them for the comparative study of ancient city systems.

The Chase-Dunn and Willard study used Chandler's (1987) data on city population6 to construct a series of SPI indices showing the steepness of city-size distributions for eight PMNs: Egyptian in 1600 BC, Mesopotamian in 2000 and 1600 EC; Aegean in 1360 BC; Mesoamerican in 200 EC, AD 500, and AD 800; West African in AD 1350, AD 1500. and AD 1550; Indie from 430 BC to AD 1550; Far Eastern from 650 BC to AD 1825; Japanese from AD 1575 to 1850 and the Central PMN from 1360 EC to AD 1988 (see Table 4.2). The Central PMN incorporated the other PMNs after the dates at which we cease to calculate their SPIs. Chase-Dunn and Willard produced SPIs for time points at which we have estimated populations for at least the three largest cities. They used at least three (but a maximum of five) cities when populations for these were available. The SPI is standardized to correct for changes in the number of cities across time points.

Observe in Table 4.2 that the SPI for the Mesopotamian PMN was -0.1 in 2000 BC. The negative sign means that the Mesopotamian city-size hierarchy was flatter than the rank-size rule. In 1600 BC the Mesopotamian SPI was -0.19 and so the city-size hierarchy had become a bit flatter than it had been in 2000. At the bottom of the table are the average SPIs for each of the PMNs. The grand mean for all the SPIs in the table is -0.820. Thus the average city-size distribution for all the PMNs for which we have data tends to be flatter than the rank-size rule. An earlier study of city-size distributions in modern national societies showed that these SPIs are, on the average, steeper than the rank-size rule (Chase-Dunn 1985b). But modem national societies are parts of a larger intersocietal system, not whole systems unto themselves. An earlier study of the Europe-centered modem world-system as bounded by the Braudel Center scholars (Chase-Dunn 1985a) showed that this larger intersocietal system composed of multiple states has city-size distributions which are usually flatter than the rank-size rule over the period from AD 800 to 1975. Intersocietal systems composed of multiple states are much more likely to have relatively flatter city-size distributions than are single states precisely because political power is more decentralized relative to the relations among cities in these larger systems than it is in single states.

Certainly the way in which we bound the regions within which we study city systems is important for what we find. In principle we want to study cities that are engaging in important interactions with one another. If we include within the same "region" cities that arc not connected to one another by such human interactions as trade, warfare, and communications, we arc lumping together in our analysis things which were not importantly connected in reality Thus we could consider the entire earth as a single region since 2000 BC and analyze all the known cities as parts of a single system since then (as seems to be suggested by the approach developed by Gills and Frank [1991]). But this would be a mistake because we know, for example, that Mesoamerican cities were not linked to Old World interaction networks in any important way until the 16th century AD. Lumping them in with the Central world-system before then would be a mistake.

Comparing the Mean SPIs of Different PMNs

Notice at the bottom of Table 4.2 that the mean SPIs for the different PMNs arc quite different from one another. Though the data points are few, it is interesting to note that the Mesopotamian PMN has a flatter city-size distribution than does the Egyptian PMN. This is consistent with what we know about differences between the Egyptian and Mesopotamian political systems. In Egypt a centralized empire formed very early and was relatively stable for a long time, producing a more hierarchical city system. In Mesopotamia a number of city-states emerged and interacted within a larger intercity-state system for a long while before the unification of the Mesopotamian core by the Akkadian conquest. And that early empire was not long-lived. The Mesopotamian city-size hierarchy is flatter than the Egyptian one, and we are tempted to say that this is due to the differences in political structure between the two regions.

The Indie PMN shows an average SPI of-0.981 based on 19 time points. This is flatter than the average for the Central PMN. The Far Eastern PMN has an average SPI of -0.184 based on 27 time points. This is more hierarchical than is the average for the Central PMN. In this case a city system which is more hierarchical corresponds to what many have perceived as a relatively more centralized political system. John Fitzpatrick (1992) has recently emphasized that Chinese development before the Mongol conquest occurred in the context of a competitive interstate system in which geopolitical competition among states often played an important part in shaping outcomes. Fitzpatrick contends that in this respect Chinese development was more similar to that in the European regional system than most of the many contrasters of Chinese empire and European state system would admit. This approach to the Far Eastern PMN is further strengthened by Thomas Barfield's (1989) work on the systemic relations between Chinese and steppe nomad states. While we may agree with Fitzpatrick's overall point, the difference in the average degree of city-size hierarchy between China and the Central PMN indicates that political centralization was great on the average, in China.

The Japanese PMN was even more hierarchical on the average than was the Far Eastern, with a mean SPI of-0.100 based on eight time points. This is quite close to the rank-size rule. It might be thought that the relatively hierarchical shape of the Japanese city system is due to the fact that Japan is rather small and small regions arc more likely to have hierarchical city-size distributions. Studies of city-size distributions in modern national societies show a negative correlation between the territorial size of a country and the degree of hierarchy of the city-size distribution. Obviously Hong Kong and Singapore, city-states, are extreme cases of urban primacy, but the negative correlation between regional size and urban primacy holds even when such cases are omitted. Does small size account for the relatively more hierarchical average SPI in the Japanese PMN'? It might account for some of it, but it cannot account for all of it. Though four of the eight individual SPIs in Japan are more hierarchical than the rank-size rule, the SPI for AD 1650 is -0.733, which is flatter than the average of all the SPIs in Table 4.2. Since the territorial size of the Japanese PMN did not greatly change between 1600 and 1700, the less hierarchical city-size distribution cannot be due to size of the region. If urban hierarchy can be shown to be a reflection of political centralization, the city-size data for Japan will contradict the notion that Japan was a decentralized feudal system between 1575 and 1850.

Though these differences between average SPIs across different PMNs suggest some support for the notion that city-size hierarchies are correlated with regionally centralized political power, the main analysis of the relationship between city-size hierarchies and periods of political centralization/decentralization requires the comparison of SPIs over time within PMNs. Inspection of the standard deviations of the distributions of SPIs for each PMN in the bottom row of Table 4.2 shows that there are big differences over time in all the PMNs for which there are several time points. Chase-Dunn and Willard plotted the changes in SPIs for each PMN for which there were several time points and they compared the changes in the SPIs over time with information about the rise and fall of states and empires.

Changes in City-Size Hierarchies and Power Concentration

Does the steepness or flatness of city-size distributions vary systematically with changes in the concentration of political and/or economic power? Unfortunately we do not have continuous and comparable measures of the distributions of economic and political power over long periods of time. This is one reason why it would be useful to know whether or not city-size distributions can serve as a proxy for direct measures of the distributions of power. In order to study changes in the distributions of political and economic power Chase-Dunn and Willard relied on historical accounts based on textual evidence. These arc imperfect sources that are undoubtedly selective and error-prone in ways that may have led to false inferences. Closer studies of PMNs are needed, especially to improve upon the estimation of changes in the distribution of political power.

These are the conclusions of the study by Chase-Dunn and Willard: Both political and economic power distributions are reflected in changes in city-size distributions. In almost every case where a city system significantly changed its degree of flatness or hierarchy Chase-Dunn and Willard found indications of major and corresponding shifts in political power. They found only two cases in which the city-size distribution changed dramatically without an obvious corresponding shift in political power. In the first such case, the Far Eastern city system became much more hierarchical in 430 EC due to the growth ofYenhsiatu in the north. This was the Warring States period well before the rise of the first empire. There was a lot of economic development and urban growth in this period. If Yenhsiatu had been at a hub of transport nets linking the cities of northern China this might have accounted for its rapid rise. But it was not. This is a mystery

The second case is less mysterious. Wilkinson designates the period between 525 and 316 BC in the Central PMN as one in which a world state was constituted by the Persian and then the Macedonian empires. But the city-size distribution in 430 BC was moderately flat. The Persian Empire was still massive in 430 BC, but important challenging states outside of its domain had developed large cities. Thus the Persian Empire was not a core-wide state in this period. The case of Yenhsiatu may be an exception to the rule or simply one in which some important piece of information is missing. In any case, only two exceptions to the correspondence between the SPI and the concentration of political power would seem to be strong evidence in favor of the above hypothesis.

Another test of the relationship between power and the city-size distributions is provided by using Taagepera's (1978a, 1978b, 1979) data on the territorial size of empires organized by PMNs. Taagepera used maps in atlases to estimate the territorial size of the largest empires on earth from 3000 BC to AD 1975. The first thing to do is to examine the relationship between empire size and city size. If the above hypothesis is correct, large empires should generally have large capital cities, though there may be exceptions. Figure 4.3 plots the population sizes of the largest cities in the Central PMN along with the territorial size of the largest empires. The data on empire sizes are much more complete than the data on city sizes. Figure 4.3 demonstrates that there is no simple correspondence between the two variables.7 It is possible that this is due to missing data on city populations.

But there is another possibility It may be that the rise and peaking of the territorial size of empires occurs first and then city size rises after a considerable time lag. This appears to occur in three instances. Empire size rises after 850 BC with the Assyrian conquests and it peaks in 500 BC with the Achaemenids. Then there is another wave due to the Alexandrian conquests. The data on city sizes are sparse in this period, but it looks as if the largest cities came during the trough between the Alexandrian empires and the rise of Rome. The Roman rise is more of a hump than a peak, and Rome was always part of a PMN that contained other large empires. Thus Rome was not a world-empire in the Wallersteinian sense. The city data from Chandler are sparse during the Roman Empire. Then the Muslims took empire size to new heights. But it was only after the decline of the large Muslim empires that cities attained their greatest size. The Mongol Empire had little to do with city growth, as is well understood by the Mongol attitude toward cities. So the evidence from the Central PMN, though spotty, indicates that cities and empires do not always grow together. It also suggests that empires grow first, then decline, and then cities grow.

Figure 4.4 examines the same variables for the case of the Far Eastern PMN. Here there is a much more positive relationship between city sizes and empire size, though again the city data are quite spotty early on.8 The rise and fall of the Han Empire corresponds to a rise and decline in city sizes. From AD 600 to 800 there are a number of large empires, and these correspond to a rise of city sizes. As in the Central PMN, the Mongol Empire corresponds to only a small rise in city' sizes, and then both empire size and city size rise rapidly after AD 1600. The evidence from the Far Eastern PMN is much more supportive of a direct relationship between city sizes and empire sizes than are the results for the Central PMN. The hypothesized lag between empires and cities in the Central PMN is not found in the Far Eastern PMN.

A more direct test of the hypothesis of power concentration and city-size hierarchy is the relationship between empire size and the SPI.9 Correlational analysis finds no relationship between these two indicators for either the Central or the Far Eastern PMN. We have not examined this relationship for possible time lags. It would also be desirable to examine an analogous distributional measure of empire-size hierarchy using data on the second-and third-largest empires in each PMN.


Further work is needed before we can conclude that the hypothesized relationship between political power concentration and city-size hierarchy is false. One problem with the analysis presented here is that the measures of city populations for earlier time periods are too spread out over time. A data gathering project that supplements the pioneering work of Tertius Chandler (1987) on city populations could provide a much more certain ability to evaluate our hypothesis. There are also holes in Rein Taagepera's (1978a, 1978b, 1979) data set on empire sizes that need to be filled. Taagepera's data concern only the largest empires on earth, and so some empires that were important within regional PMNs are left out. Supplementing these two data sets would not only provide the opportunity' of an improved evaluation of the power concentration/city hierarchy hypothesis, but would be a helpful basic step toward the construction of a world history geographical information system.


1. For a review, see Hall and Chase-Dunn (1994).

2. Wilkinson's (19920, pp. 55-56) definition of oikumene is as follows: "An 'oikumeme' is here defined as a trading area, a domain internally knit by a network of trade routes, in which there is enough internal trade so that the whole trading area evolves to a significant degree as a system, while trade outside the area, though perhaps important both to the oikumene and to other oikumenes with which it trades, is not sufficiently dense and significant to cause system-level development to encompass these external systems."

3. We follow Johnson and Earle (1987) in defining true states as polities that are larger than chiefdoms and that have specialized institutions of regional administration and control.

4. Wilkinson refers to the decentralization phase as constituting a "states system."

5. An overview of the role of cities in world-system development is presented in Chase-Dunn (1992).

6. See Wilkinson (19920) for a comparison between Chandler's city lists and those derived from other sources. The business of estimating the population sizes of cities is fraught with danger, especially when we are limited to archaeological evidence. The estimates get more accurate as we approach the present.

7. The Pearson r correlation coefficient between Central city and empire sizes is only 0.36 based on 25 time points. This is statistically significant only at the 0.07 level.

8. The correlation coefficient for Far Eastern city and empire sizes is 0.46 based on 28 time points. This is statistically significant at the 0.014 level.

9. The Central correlation between the SPI and empire size is 0.15 based on 25 time points. The Far Eastern correlation is -0.02 based on 24 time points.

Rambler's Top100
Hosted by uCoz