Comparing World-Systems:

Empire Upsweeps and Non-core marcher states

 Since the Bronze Age

Alexis Álvarez, E.N. Anderson, Elisse Basmajian, Hiroko Inoue, Christian Jaworski, Alina Khan, Kirk Lawrence, Andrew Owen, Anthony Roberts,

Panu Suppatkul and Christopher Chase-Dunn
Institute for Research on World-Systems
University of California-Riverside
National Science Foundation Grant #:  NSF-HSD SES-0527720

To be submitted to Social Forces  draft v. 8/7/13, 6316 words.

This paper is available as IROWS Working Paper #56


Abstract: This is an examination of one of the implications of the hypothesis of semiperipheral development: that major increases in the sizes of polities have been accomplished mainly by the conquests carried out by semiperipheral marcher states. We use the comparative evolutionary world-systems perspective to frame our study of twenty-two upsweeps of the largest polities in four regional world-systems and in the expanding Central System since the Bronze Age. We seek to determine whether or not each of these twenty-two upsweeps were or were not instances in which a semiperipheral marcher state produced the polity size upsweep by means of conquest. The hypothesis of semiperipheral development holds that polities that are in between the core and periphery (semiperipheral polities) have been, and continue to be, unusually fertile locations for the implementation of organizational and technological innovations that transform the scale and the developmental logic of world-systems. This is because semiperipheral polities have less invested in older institutional structures and than do core polities and they also have greater incentives to take risks on new technologies and organizational forms. One important manifestation of semiperipheral development is the marcher state phenomenon: a recently founded sedentary polity out on the edge of an older core region conquers the older core polities and puts together a core-wide empire that is significantly larger than earlier polities have been. This phenomenon has occurred repeatedly, but it is not the only way in which large empires have emerged.  We find that over half of the twenty-two identified empire upsweeps were likely to have been produced by marcher states from the semiperiphery (10) or from the periphery (3).


            This paper is a part of a larger project that is studying the growth/decline phases and upward sweeps of settlement and polity sizes in order to test explanations of long-term patterns of human socio-cultural evolution. We use the comparative and evolutionary world-systems perspective first outlined by Chase-Dunn and Hall (1993,1997) as our orienting theoretical approach. The focus is on interpolity systems (world-systems) rather than on single polities.[1] We study how sociocultural evolution occurred in systems of interacting polities. We propose a somewhat revised definition and typology of semiperipherality based on our study of empire upsweeps. The main focus of this article is on the evolutionary significance of semiperipherality since the period in the early Bronze Age in which states emerged.

Regions, Political-Military Networks and Upward Sweeps

          There has been a long-term upward trend in which human polities have grown in population and territorial size while the total number of sovereign polities has decreased (Carneiro 1978).[2] Human polities have evolved from bands to tribes to chiefdoms to states and then to empires. This long-term trend has occurred in a series of events that we call upward sweeps (upsweeps).  Upsweeps are defined as those instances in which a polity emerges that is at least 1/3 larger in territorial size that the average of the three previous peaks of polity size (Inoue et al 2012).

            All hierarchical world-systems have experienced a cycle of centralization and decentralization in which a large polity in an interpolity system emerged and then declined. This sequence of rise and fall is seen in interpolity systems composed of chiefdoms (Anderson 1994), states, empires and modern hegemons. In such cycles most of the upward phases result in a polity that is nearly the same size as the one that existed at the previous peak of polity size. This we call a “normal rise.” An upsweep involves a significant increase in the size of the largest polity relative to the previous peak. These upweeps are much less frequent than are normal rises.  But they are the events that instantiate the long-term trend toward larger polities and so they are very important for the evolution of sociocultural complexity.

             The Polities and Settlements Research Working Group at the Institute of Research on World-Systems[3] has quantitatively identified twenty-two such polity upsweeps in five world regions since the early Bronze Age (Inoue et al 2012). In order to identify these polity upsweeps we have mainly used Rein Taagepera’s (1978a,1978b,1979, 1997) estimates of the territorial sizes of the largest states and empires in four world regions and in the interpolity system that David Wilkinson (1987) has called “Central Civilization.”[4]

Figure 1:  Largest polities in the Mesopotamian PMN, 2800 BCE-1500 BCE

            Figure 1 shows the territorial sizes of the largest polities in the Bronze Age Mesopotamian political-military network (PMN) between 2800 BCE and 1500 BCE. A political-military network is a set of fighting and allying polities. This is equivalent to what International Relations Political Scientists call “the international system” except that they rarely study such systems before the treaty of Westphalia.  Figure 1 illustrates the difference between empire upsweeps and normal rises. Lagash carried out an upsweep because it became significantly larger than the largest earlier polities had been. The Akkadian Empire was a gigantic upsurge. The Babylon/Mitanni upsweep was not as big as the Akkadian had been, but it was more than 1/3 larger than the average of the three previous peaks and so qualifies as an upsweep. Around 1500 BCE the Mesopotamian and Egyptian PMNs merged to become what we call the Central PMN (or the Central System).

          We contend that interaction networks, rather than homogenous cultural or ecological regions, are the best way to bound evolving human systems for the purposes of studying the causes of socio-cultural evolution (Chase-Dunn and Jorgenson 2003)[5].  But one problem with using interaction networks is that they expand (and contract) over time, which can make the results of comparisons dependent on the decisions one has made about the timing of changes in the spatial boundaries of networks. Our project uses four world regional political-military networks (PMNs) and one expanding PMN – what we call, following David Wilkinson (1987), the Central System (see Figure 2).

Figure 2: Chronograph the expansion of the Central PMN and its engulfment of regional PMNs since the Bronze Age. [revised from Wilkinson (1987);  the PMNs studied here are underlined]


The four regional PMNs are Mesopotamia, Egypt, East Asia and South Asia. The Central System, a network of allying and warring states and empires, is bounded following Wilkinson.[6] It begins around 1500 BCE when the formerly separate Egyptian and Mesopotamian interpolity systems merged and it then expanded to eventually become the contemporary global international system. The East Asian region was linked by long-distance trade in prestige goods with the Central System since the time of the Roman and Han empires, but East Asia had a substantially separate interpolity system (PMN) until China was surrounded and penetrated by the European powers in the 19th century.[7] 

Core/Periphery Relations

            The notion of core/periphery relations has been a central concept in both the modern world-system perspective (Wallerstein 1974) and in the comparative evolutionary world-systems perspective (Chase-Dunn and Hall 1993, 1997). World-systems are systems of interacting polities and they often (but not always) are organized as interpolity hierarchies in which some polities exploit and dominate other polities. Chase-Dunn and Hall redefined and expanded the core/periphery concept to make it more useful for comparing the modern world-system with earlier regional world-systems. The core/periphery distinction is a relational concept. In other words, what coreness, peripherality and semiperipherality  are depends on the larger context in which they occur – the nature of the polities that are interacting with one another and the nature of their interactions.

            When we use the idea of core/periphery relations for comparing very different kinds of world-systems we need to broaden the concept and to make an important distinction (see below).  But the most important point is that we should not assume that all world-systems have core/periphery hierarchies just because the modern system does. It should be an empirical question in each case as to whether core/periphery relations exist.  Not assuming that world-systems have core/periphery structures allows us to compare very different kinds of systems and to study how core/periphery hierarchies themselves have emerged and evolved.

            For comparing different kind of systems it is also helpful to distinguish between core/periphery differentiation and core/periphery hierarchy.  “Core/periphery differentiation” means that polities with different population densities different amounts of internal hierarchy are interacting with one another. As soon as we find village dwellers in sustained interaction with nomadic neighbors we have core/periphery differentiation.  “Core/periphery hierarchy” refers to the nature of the relationships between polities.  Interpolity hierarchy exists when some polities are exploiting and/or dominating the people in other polities. Well-known examples of interpolity domination and exploitation include the British colonization and deindustrialization of India, or the conquest and subjugation of Mexico by the Spaniards. But core/periphery hierarchy is not unique to the modern Europe-centered world-system of recent centuries. Roman and Aztec imperialism are also famous.[8]

Distinguishing between core/periphery differentiation and core/periphery hierarchy allows us to deal with situations in which larger and more population dense polities are interacting with smaller ones, but are not exploiting them. It also allows us to examine cases in which smaller, less dense polities were exploiting larger and more dense polities such as occurred in the long and consequential interaction between the nomadic horse pastoralists of Central Asia and the agrarian states and empires of China and Western Asia. The most famous case was that of the Mongol Empire of Chingis Khan, but confederations of Central Asian steppe nomads had managed to extract tribute from agrarian states long before the rise of Mongols (Barfield 1989). 

The question of core/periphery status also should be considered with regard to different spatial scales of interaction. Chase-Dunn and Hall (1993,1997) point out that regional world-systems often were composed of important interaction networks that had different spatial scales. They use a “place-centric” approach to spatially bounding systemic interaction networks that begins from a focal settlement or polity. The network is spatially bounded by considering how many indirect links are needed to include all the actions have an important impact on the reproduction or transformation of institutions at the focal locale. Bulk goods networks (in which everyday foods and raw materials were exchanged) were usually smaller in extent relative to political-military networks (PMNs) of allying and fighting polities. Prestige goods networks (in which high value per weight goods were exchanged) usually were larger than PMNs and systems vary in terms of how important the exchange of prestige goods are for the reproduction of social institutions.  The issue of core/periphery status always needs to be asked for both the bulk goods and political-military networks.

Semiperipheral Development

            The semiperiphery concept was also originally developed for the study the modern world-system (Wallerstein 1976). But it too has been expanded for the purposes of comparing world-systems. For Wallerstein the semiperiphery is a middle stratum in the global hierarchy that helps to reduce the strains that emerge from polarization. Chase-Dunn and Hall (1997: Chapter 5) contend that semiperipheral polities are often the “seedbeds of change” because some of them implement new technological and organizational features that allow them to successfully compete with core polities. This is thought to account for the phenomenon of uneven development and the movement of core regions from their original locations.

Hub theories of innovation have been popular among world historians (McNeill and McNeill 2003; Christian 2004) and human ecologists (Hawley 1950). The hub theory holds that new ideas and institutions tend to emerge in large and central settlements where information cross-roads bring different ideas into interaction with one another. The hub theory is undoubtedly partly correct, but it cannot explain some of the long-term patterns of human sociocultural evolution, because if an information cross-road was able to out-compete all contenders then the original information hub would still be the center of the world. But that is not the case. We know that cities and states first emerged in Bronze Age Mesopotamia. Mesopotamia is now Iraq. It had 100% of the world’s largest cities and the most powerful polities in the Early Bronze Age (Morris 2010, 2012). Now it has neither the largest cities nor the most powerful polities. All of the regional world-systems have undergone a process of uneven development in which the old centers were eventually replaced by new centers out on the edge.

            Chase-Dunn and Hall contend that it is often polities out on the edge that transform the institutional structures and accomplish the upsweeps. This hypothesis is part of a larger claim that semiperipheral polities often play transformative roles that cause the emergence of complexity and hierarchy within polities and in world-systems. This is the most important justification of the claim that a world-systems rather than a single polity (or society), is a necessary unit of analysis for explaining human socio-cultural evolution. [9]

The node theory does not well account for the spatially uneven nature of evolutionary change. The cutting edge of evolution moves. Old centers have often been transcended by polities out on the edge that were able to rewire network nodes in a way that expanded the spatial scale of networks.

            The phenomenon of  semiperipheral development has taken various forms: semiperipheral marcher chiefdoms, semiperipheral marcher states, semiperipheral capitalist city-states, the semiperipheral position of Europe in the larger Afroeurasian world-system, modern semiperipheral nation-states that have risen to hegemony, and contemporary semiperipheral societies that are engaging in novel and potentially transformative economic and political activities that may change the nature of the contemporary global system.

            There are several possible processes that might account for the phenomenon of semiperipheral development. Randall Collins (1981) has argued that the phenomenon of marcher states conquering other states to make larger empires is due to the geopolitical “march land advantage.” Being out on the edge of a core region of competing states allows more maneuverability because it is not necessary to defend the rear. This geopolitical advantage allows military resources to be concentrated on vulnerable neighbors. Peter Turchin (2003) argued that the relevant process is one in which group solidarity is enhanced by being on a “metaethnic frontier” in which the clash of contending cultures produces strong cohesion and cooperation within a frontier society, allowing it to perform great feats. Carroll Quigley (1961) distilled a somewhat similar theory from the works of Arnold Toynbee. Another factor affecting within-group solidarity is the different degrees of internal stratification usually found in premodern systems between the core and the semiperiphery. Core societies often have old, crusty and bloated elites who rely on mercenaries and “foreigners” as subalterns, while semiperipheral leaders are more often charismatic individuals who attract strong loyalty from their soldiers. Less stratification can mean greater group solidarity. And this may be an important part of the semiperipheral advantage in systems in which within-polity inequality is greater in the core than in the non–core.

            But Arnold Toynbee (1946) also suggested another way in which the peoples of semiperipheral regions might be motivated to take risks with new ideas, technologies and strategies. Semiperipheral polities are often located in ecologically marginal regions that have poor soil and little water or other natural disadvantages. Patrick Kirch relies on this idea of ecological marginality in his depiction of the process by which semiperipheral marcher chiefs were most often the conquerors that created island-wide paramount chiefdoms in the Pacific (Kirch 1984). It is quite possible that all these features combine to produce what Alexander Gershenkron (1962) called “the advantages of backwardness” that allow some semiperipheral societies to transform and to dominate regional world-systems.

            Those new technologies and organizational forms that transform the logic of development and allow world-systems to get larger, more complex and more hierarchical, are often implemented by semiperipheral polities. Some semiperipheral polities are unusually fertile locations for the implementation of these new adaptive strategies. Semiperipheral polities are often involved in processes of rapid internal class formation and state formation and they do not have large investments in, and commitments to, doing things the way they have been done in older core polities. They do not have institutional or infrastructural sunk costs. So they are freer to implement new institutions and to experiment with new technologies.

 There are several different ways to be semiperipheral (see below) and semiperipheral polities not only sometimes transform systems but they also sometimes take over and become the new predominant core polities.  We have already mentioned semiperipheral marcher chiefdoms. The chiefdoms that conquered and unified a number of smaller chiefdoms into larger paramount chiefdoms were usually from semiperipheral locations.  Peripheral peoples did not usually have the institutional and material resources that would allow them to implement new adaptive strategies or to take over older core regions.  It was in the semiperiphery that core and peripheral social characteristics could be recombined in new ways.  Sometimes this meant that more adaptive and competitive techniques of power were strongly implemented in semiperipheral polities. 

Much better known than semiperipheral marcher chiefdoms is the phenomenon of semiperipheral marcher states. Many of the largest empires in all world regions were assembled by conquerors who came from semiperipheral polities.  The most famous examples are the Achaemenid Persians, the Macedonians led by Alexander, the Romans, the Islamic Caliphates, the Ottomans, the Manchus and the Aztecs.

But some semiperipheries transform institutions, but do not take over the interpolity system of which they are a part.  The semiperipheral capitalist city-states operated on the edges of the tributary empires where they bought and sold goods in widely separate locations, encouraging farmers and craftsmen to produce a surplus for trade.  The Phoenician cities (e.g. Tyre, Sidon, Biblos, Carthage, etc.), as well as Malacca, Venice, Genoa and the cities of the Hanseatic League, spread commodification and expanded markets by producing manufactured goods and trading them across great regions.[10]  In this way the semiperipheral capitalist city-states helped to commercialize the world of the tributary empires without themselves becoming core powers (see Inoue et al 2013).

The modern world-system has experienced a sequence of the rise and fall of hegemonic core states. The Dutch, the British and the U.S. were countries that had formerly been in semiperipheral positions relative to the modern core/periphery hierarchy.  And indeed the rise of Europe within the larger Afroeurasian world-system was also a case of semiperipheral development, one in which a formerly peripheral and then semiperipheral region eventually rose to become a new core and to bring all the regions into a now-global interpolity system (Chase-Dunn and Lerro 2013).[11]

Indicators of Semiperipherality

            The main purpose of this article is to determine which and how many of the twenty-two quantitatively identified empire upsweeps were brought about by semiperipheral marcher states. In order to do this we need to specify what we mean by semiperipherality. This is not a simple task because, as we have mentioned above, world-system position (core, periphery, semiperiphery) are relational concepts. In other words, what semiperipherality is depends on the larger context in which it occurs – the nature of the polities that are interacting with one another and the nature of their interactions. The most general definition of the semiperiphery is: an intermediate location in an interpolity core/periphery structure. The minimal definition of core/periphery relations, as mentioned above, is that polities with different degrees of population density and internal hierarchy and complexity are interacting with one another. This is what Chase-Dunn and Hall 1997) have called “core/periphery differentiation.” The idea of “core/periphery hierarchy” is more stringent. It requires interpolity domination and exploitation. In this study we will be looking for evidence that a polity that conquered other polities and was responsible for an upward sweep in territorial size was semiperipheral relative to the other polities it was interacting with before it started on the road to conquest. We will use four main empirical indicators to make such determinations:

·                     the geographical location of the society relative to other societies that have greater or lesser amounts of population density. Is it out on the edge of a region of core polities?;

·                     the relative level of development: population density, which is usually indicated by the sizes of settlements, the relative degree of complexity and hierarchy, the mode of production: e.g foraging, pastoralism, nomadism vs. sedentism, horticulture vs. agriculture, the size of irrigation systems, etc. Foragers (hunter-gatherers) and pastoralists are usually peripheral to more sedentary agriculturalists and those who dwell in large settlements;

·                     the recency of the adoption of sedentism, agriculture, class formation and state formation; and

·                     relative ecological marginality.

            Core polities usually hold to best locations in terms of soil and water. The semiperipheral marcher chiefdoms of the Pacific Islands were typically from the dry side of the island where land was steeper and soil was thinner (Kirch 1984). Of course, which land is better depends on the kind of resources that are being used and the technologies available for appropriating resources. But ecological marginality is often an important aspect of semiperipherality. Polities in ecologically marginal regions have a powerful incentive for taking the risks of conquest.

The Aztecs are a proto-typical example of a semiperipheral marcher state. They were nomadic hunter-gatherers who migrated into the valley of Mexico and settled on an uninhabited island in a lake. There had already been large states and empires in the valley of Mexico for centuries. The Aztecs hired themselves out as mercenary soldiers, developed a class distinction between nobles and commoners and claimed to have been descended from the Toltecs, an earlier empire. Then they began conquering the older core states of the valley of Mexico, strategically picking first on weak and unpopular states, until they had gathered enough resources to “roll up the system.” The Aztec story has all of the elements specified above that we will use in examining our upsweep cases.

            One issue that complicates the determination of world-system position is: semiperipheral to what? A polity may have different relationships with different other polities in the same interpolity network. For example, Macedonia had one kind relationship with the other Greek states, and a different kind of relationship with the Persian Empire. Semiperipherality is relative to the system as a whole, but may also be affected by important differences between other states in a system and by the existence of different kinds of relations with those other states.

            The logical alternatives to semiperipherality are coreness and peripheralness. Core states are older, more stratified, have bigger settlements, and they have had the accoutrements of civilization, such as writing, longer. Peripheral polities are often nomadic hunter-gatherers or pastoralists, hill people, or desert people. If they are sedentary, their villages are small relative to the settlements of those with whom they are interacting.  We also note that some conquest empires were formed by peripheral marcher states or by old core states that made a comeback. David Wilkinson’s (1991) survey of the core, peripheral and semiperipheral zones of thirteen interpolity systems, is helpful in suggesting criteria for designating these zones, but Wilkinson did not address the question we are asking here: were the polities that produced empire upsweeps semiperipheral before they did this?

            We should also note that some large empires have been formed by internal revolt in which a subordinate ethnic group or caste revolted and took power in an existing state and then carried out an expansion by conquest. The slave-soldiers of the Mamluk Sultanate are an obvious example, and Norman Yoffee (1991) has contended that the Akkadian empire was the result of an ethnic revolt (but see below). The point here is that this is a possible alternative to the semiperipheral marcher state route to an empire upsweep.

So we may find four types of upsweeps:

·                     semiperipheral marcher state, a polity that is in a semiperipheral position within a regional system conquers a large area and produces an upsweep;

·                     core state conquest or restoration, conquest by a state that has been in the core of a regional system for centuries, or a restoration of domination by an older core civilization that had been conquered by a semiperipheral or peripheral polity (e.g. the Third Dynasty of Ur, a Sumerian restoration in Mesopotamia or the Ming Dynasty in China in which the Han Chinese threw out the Mongol Yuan rulers;

·                     internal revolt,  a state formed by an internal rebellion, such as what Yoffee argues for the Akkadian Empire, or the Mamluk Empire; and

·                     peripheral marcher states, in which a polity that is peripheral in a regional system conquers the core, (e.g. the Mongol Empire).

Our review of the upsweeps also suggests that there have been different kinds of semiperipheral marcher states that have different combinations of the features discussed above. The summary of our efforts to find evidence for each case is contained in the Appendix for this article.[12]  The effort to locate evidence in favor or against the semiperipheral origins of the upsweeps has been more difficult for the early cases because evidence for them is circumstantial or lacking. Thus the categorization of the origins of upsweeps must be qualified in terms of the degree of certainty.  

            Table 1 shows the twenty-two cases identified as empires upsweeps in our quantitative study, which was based mainly on Rein Taagepera’s estimates of the territorial sizes of large empires (Inoue et al 2012).  It also shows the results of our effort to categorize each of these upsweeps into the four types outlined above.

The Mongol empire is shown twice in Table 1 because it was important for both the East Asian and the Central PMNs, but it is not counted twice in Table 2 below. The first three upsweeps in Table 1 are from the Mesopotamian regional world-system (see also Figure 1 above).

Mesopotamia 2800 BCE to 1500 BCE



Polity name

Semiperipheral Marcher State? (SMS)




Probably not




combined SMS and ethnic revolt





Egypt     2850 BCE to 1500 BCE



Polity name

Semiperipheral Marcher State?



5th Dynasty




12th Dynasty





SMS and ethnic revolt

Central PMN -1500 BCE to 1991AD



Polity name(s)

Semiperipheral Marcher State?











18th Dynasty








Achaemenid Persia








Islamic Empires





Peripheral Marcher State






East Asia 1300 BCE to 1830 AD



Polity name(s)

Semiperipheral Marcher State?



Shang/Western Zhou (Chou)

Probably SMS



Qin/Western Han/Xiongnu *

Qin=SMS; Western Han=no; Xiongnu=Peripheral Marcher State



Western Han




Eastern (Later) Han




E. Turks

Peripheral Marcher State







Mongol-Yuan (same as above in Central System. Counted only once)

Peripheral Marcher State





South Asia -420 BCE to 1008 AD



Polity name

Semiperipheral Marcher State?




Probably SMS

*The Qin/Western Zhou/Xiongnu was a composite upsweep produced by three polities with three different original world-system positions. Each of these is counted as 1/3 in Table 2 below.

Table 1: World-system position origin of those polities that produced territorial upsweeps

            Table 1 shows the results of our efforts to determine whether or not each of the twenty-two polity size upsweeps were the result of the rise of a semiperipheral marcher state. In the case of Lagash we have not been able to find any hard evidence regarding the world-system position origins of this polity upsweep, but our best guess is that it was not a semiperipheral marcher state (see the Appendix for the review of evidence).  For the rest of the cases we have been able to find enough evidence to determine with a fair degree of certainty whether or not they were semiperpheral marcher states.


SMS and Ethnic Revolt

Probably a SMS (Shang, Mauryan)

Peripheral Marcher State


Not enough evidence to tell (Lagash)

Not a SMS

6 1/3*



2 1/3*


7 1/3 *

Table 2: Count of the 22 upsweep cases in Table 1 with regard to status as semiperipheral marcher state prior to upsweep. 

SMS = Semiperipheral Marcher State

*See note to Table 1.



We studied the twenty-two upsweeps in the territorial sizes of states and empires that account for the long-term evolutionary trend toward increasing size of polities. Of the twenty-two upsweeps we found, thirteen were due to the actions of either semiperipheral marcher states (10) or peripheral marcher states (3). The three peripheral marcher states (Xiongnu, Eastern Turks and Mongols) were Central Asian steppe pastoralist confederations that conquered large territories.  

            Despite that it is difficult to tell with available evidence whether or not three of the upsweeps were due to the actions of semiperipheral marcher states or not, there are enough fairly certain cases to demonstrate support for the hypothesis that over half of the polity upsweeps were the result of conquest by non-core polities. This is strong support for the idea that uneven develop is important and important cause of sociocultural evolution.  The nine cases that were not due to the rise of a non-core power were coups in which a new dynasty came to power within an existing core state and led a successful military expansion.

            Ten of the twenty-two upsweeps were due to the actions of a semiperipheral marcher state.  This means that the theory of semiperipheral development does not explain everything about the events in which polity sizes significantly increased in geographical scale, but also that the semiperipheral development hypothesis cannot be ignored in any explanation of the long-term trend in the rise of polity sizes.

Appendix: Classification of Empire Upsweeps with regard to status as produced by semiperipheral marcher states



Allen, Mitchell. 1997  Contested Peripheries:  Philistia in the Neo-Assyrian World-System.  Unpublished Ph.D.

dissertation, Interdepartmental archaeology program, UCLA.

_____.  2005. “Power is in the Details: Administrative Technology and the Growth             of Ancient Near

Eastern Cores.” Pp. 75-91 in The Historical Evolution of World-Systems, edited by Christopher Chase-Dunn

and E. N. Anderson. New York and London: Palgrave.

Anderson, David G. 1994 The Savannah River Chiefdoms. Tuscaloosa: University of Alabama Press.

Arrighi, Giovanni.  1985  Semiperipheral Development: The Politics of Southern Europe in the Twentieth Century.

            Beverly Hills: Sage.

            Barfield, Thomas J. 1989 The Perilous Frontier: Nomadic Empires and China. Cambridge, MA.: Blackwell

            Carneiro, Robert L.  1978.  "Political Expansion as an Expression of the Principle of Competitive Exclusion."

             Pp. 205-23 in Origins of the State: the Anthropology of Political Evolution, edited by Ronald Cohen and Elman R. Service.

            Philadelphia:  Institute for the Study of Human Issues.

Chase-Dunn, Christopher.  1988  "Comparing World Systems: Toward a Theory of Semiperipheral             Development."

 Comparative Civilizations Review  19(Fall):29-66.

Chase-Dunn, C. and Andrew K. Jorgenson, “Regions and Interaction Networks: an institutional materialist

            perspective,” 2003 International Journal of Comparative Sociology 44,1:433-450.

Chase-Dunn, C. and Thomas D. Hall 1993 "Comparing world-systems: concepts and working hypotheses"

Social Forces 71,4 (June).

_____________________________.  1997 Rise and Demise: Comparing World-Systems.           Boulder:

            Westview Press.

Chase-Dunn, Christopher, Daniel Pasciuti, Alexis Alvarez and Thomas D. Hall  2006 “Waves of

            globalization and semiperipheral development in the ancient Mesopotamian and Egyptian

            world-systems. Pp. 114-138 in Globalization and Global History edited by Barry K. Gills and

            William R. Thompson. London:Routledge.

Chase-Dunn, C. and Bruce Lerro 2013 Social Change: Globalization From the Stone Age to the Present.

            Boulder, CO: Paradigm.

Christian, David 2004 Maps of Time: An Introduction to Big History. Berkeley: University of California


Collins, Randall.  1981 “Long term social change and the territorial power of states,”           Pp. 71-106 in

            R. Collins (ed.)  Sociology Since Midcentury. New York: Academic Press.

Ferguson, R. Brian, and Neil L. Whitehead, eds.  1992  War in the Tribal Zone:          Expanding States and

            Indigenous Warfare.  Santa Fe, NM: School of American Research Press.

Hall, Thomas D.  2000 "Frontiers, Ethnogenesis, and World-Systems: Rethinking the Theories." 

            Pp. 237-270 in A World-Systems Reader: New Perspectives on Gender, Urbanism, Cultures, Indigenous

            Peoples, and Ecology edited by Thomas D. Hall. Lanham, MD: Rowman & Littlefield.

Chase-Dunn, Christopher and Thomas D. Hall 2006 “Global social change in the long run” Chapter

            3 in C. Chase-Dunn and Salvatore Babones (eds.) Global Social Change. Baltimore: Johns

            Hopkins University Press.

            Gershenkron, Alexander 1962 Economic Backwardness in Historical Perspective. Cambridge, MA: Harvard
            University Press.

Hawley, Amos. 1950. Human Ecology: A Theory of Community Structure. New York: Ronald Press Co.

Hopkins, Terence K. and Immanuel Wallerstein. 1987 "Capitalism and the Incorporation of New

            Zones Into the World-economy." Review 10:4(Fall):763-779.

Inoue, Hiroko, Alexis Álvarez, Kirk Lawrence, Anthony Roberts, Eugene N Anderson and

            Christopher Chase-Dunn 2012 “Polity scale shifts in world-systems since the Bronze Age: A

   comparative inventory of upsweeps and collapses” International Journal of Comparative Sociology 53,3:    210-229.

Inoue, Hiroko, Alexis Álvarez, Eugene N. Anderson, Andrew Owen, Jared Ahmed and Christopher

   Chase-Dunn 2013“Settlement Scale Shifts Since the Bronze Age: Upsweeps, Collapses and 
   Semiperipheral Development” Submitted for publication

Kirch, Patrick 1984 The Evolution of Polynesian Chiefdoms. Cambridge: Cambridge University Press

Kradin, Nikolay N..  2002.  “Nomadism, Evolution and World-Systems: Pastoral Societies in         Theories

of Historical Development.”  Journal of World-Systems Research, 8,3.

Lattimore, Owen.  1940.  Inner Asian Frontiers of ChinaNew York:  American Geographical Society,

            republished 1951, 2nd ed.  Boston: Beacon Press.

La Lone, Darrell.  2000  "Rise, Fall, and Semiperipheral Development in the Andean World-System"

              Journal of World-Systems Research 6:1(Spring):68-99    

Lawrence, Kirk. 2009 “Toward a democratic and collectively rational global commonwealth:

            semiperipheral             transformation in a post-peak world-system” in Phoebe Moore and Owen

            Worth, Globalization and the Semiperiphery: New York: Palgrave MacMillan.

Mann, Michael 1986. The Sources of Social Power: A History of Power from the Beginning to A.D. 1760.

            Cambridge: Cambridge University Press.

            McNeill, John R. and William H. McNeill 2003 The Human Web. New York: Norton.

Morris, Ian 2010 Why The West Rules – For Now. New York: Farrer, Straus and Giroux

_____. 2012 The Measure of Civilization. Princeton: Princeton University Press

Quigley, Carroll 1961 The Evolution of  Civilizations. Indianapolis: Liberty Press.

            Sanderson, Stephen K. 1990 Social Evolutionism. Cambridge, MA: Blackwell.
            Taagepera, Rein 1978a "Size and duration of empires: systematics of size" Social Science Research
                        7:108-  27.

_____. 1978b "Size and duration of empires: growth-decline curves, 3000 to 600 B.C." Social Science

                                    Research, 7 :180-96.

_____.1979 "Size and duration of empires: growth-decline curves, 600 B.C. to 600 A.D." Social

                                    Science History 3,3-4:115-38.

_____.1997 “Expansion and contraction patterns of large polities: context for Russia.” International

                        Studies Quarterly 41,3:475-504.

Toynbee, Arnold. 1946. A Study of History (Somervell abridgement), Oxford: Oxford University Press.

Turchin, Peter.  2003. Historical Dynamics. Princeton, NJ:  Princeton University Press.

            Turchin, Peter, Jonathan M. Adams, and Thomas D. 2006. “East-West Orientation of Historical Empires

            and Modern States.” Journal of World-Systems Research.12:2(December):218-229.

            Wallerstein, Immanuel. 1974. The Modern World-System:  Capitalist Agriculture and the Origins of European

            World-  Economy in the Sixteenth Century.  New York:  Academic Press.

_____. 1976 “Semiperipheral countries and the contemporary world crisis” Theory and Society 3,4:461-         483.

Wilkinson, David. 1987 "Central civilization" Comparative Civilizations Review 17:31-59 (Fall).

_____. 1991  Cores, peripheries and civilizations”  Pp.113-166 in C. Chase-Dunn and T.D. Hall

            (eds.) Core/Periphery Relations in Precapitalist Worlds. Boulder, CO: Westview. 


Yoffee, Norman 1991 “The collapse of ancient Mesopotamian states and civilization.” Pp. 44-68 in

            Norman Yoffee and George Cowgill (eds.) The Collapse of Ancient States and Civilizations.

            Tucson: University of Arizona Press.



[1] We use the term “polity” to generally denote a spatially-bounded realm of sovereign authority such as a band, tribe, chiefdom, state or empire. World-systems are defined as human interpolity interaction networks that link trading, fighting and allying polities.


                [2] The long-term trend is masked in recent centuries by the waves of decolonization of the great colonial empires

                that had been the predominant form of core/periphery relations in the Europe-centered world-system. The number of          sovereign polities increased greatly as the Global South transitioned from colonies to formally sovereign members of the        international system of states. But this was also part of the emergence of a more uniform system of global governance      that is a continuation of the long-term trend noted by Carneiro (1978).

[4] An interpolity system is a set of interacting polities that make alliances and war with one another.  In other contexts we have called this a “political-military network” to distinguish it from other interactions that typically have smaller or larger spatial scales – bulk goods networks and prestige goods networks.

[5] This is because interpolity interaction often causes crosspolity differentiation, not homogeneity. Using “culture areas” obscures such cases of co-evolution.

[6] We want to include the Mesoamerican and the Andean systems as well as others, but quantitative estimates of the territorial sizes of polities are not currently available over enough time and with sufficient temporal resolution for the study of cycles and upward sweeps in these. Documentary evidence is required for the estimation of the territorial sizes of polities. Archaeological evidence by itself cannot be used to determine the territorial sizes of polities with enough temporal resolution for the identification of polity size upsweeps.

[7] Conquest is not the only road to state formation, as is illustrated by the partial success of the process of European unification since World War II. This is important for thinking about the potential for future global state formation.


            [8] Less well-known are instances in which core chiefdoms succeeded in extracting resources from distant polities such as        existed in the indigenous Chesapeake and the Pacific Northwest (Chase-Dunn and Lerro 2013).


[9] The intellectual history of the hypothesis of semiperipheral development is reviewed in Chase-Dunn and Hall (1997: Chapter 5).

[10] Dilmun and the Old Assyrian City-State (Assur) functioned similarly in the Bronze Age. Malakka was such in Southeast Asia, and Osaka, though it was never had autonomous sovereignty, played a similar role in Japan. These were the first capitalist states in which state power was mainly used to facilitate profit-making rather than the extraction of taxes and tribute.


[11] There may be an analogous phenomenon to interpolity semiperipheral development that occurs within polities. Organizations such as firms that are competing with each other may also exhibit aspects of the “advantages of backwardness.”

[12] Appendix: Classification of Empire Upsweeps