The SetPol Frameworks for

comparing World-Systems

Galactic or mandala polities in Southeast Asia

Christopher Chase-Dunn,

Marilyn Grell-Brisk and Hiroko Inoue

Institute for Research on World-Systems

University of California-Riverside

V 9/16/2020 8929 words

Forthcoming in Global Perspectives Section on Global epistemologies, concepts, methodologies, and data systems. This is IROWS Working Paper # 138


Abstract: The article explains two comparative research designs for studying the sociocultural evolution of world-systems and their expansion to become the single global system of today. An anthropological framework of comparison allows the study of similarities and differences between small stateless, medium-size regional state-based and modern global whole systemic human interaction networks. Case studies of single systems and comparative quantitative studies using settlement sizes and the territorial sizes of polities have used geopolitical military networks to bound separate whole systems, but because these networks change their spatial boundaries over long periods of time, we also use a set of spatially constant world regions selected with regard to their settlement size histories to study changes in settlement and polity sizes and to compare world regions. We also present our study of the spatio-temporal boundaries of Japan in the East Asian world-system as an example.

Structural globalization involves the geographical expansion and intensification of large-scale interaction networks relative to local interaction networks. As such, a single, complex, hierarchical, and systemically integrated global human interaction network emerged in the nineteenth century CE. This creates a special challenge for comparative world-systems analysis. With only one “case” the quantitative study of changes in the whole system must use changes over time in variable characteristics or examine subsystems that were systemically linked with one another. But whole autonomous systems can be studied comparatively if we go farther back in time, because there were smaller regional autonomous systemic interaction networks.

Chase-Dunn and Hall (1997) defined world-systems as systemic interaction networks that linked settlements[1] and polities[2] in reciprocal interaction networks that conditioned the reproduction and change of local social structures. The important insight is that all human polities have systemic interactions with their neighbors, so it does not make sense to study them one at time.

The word “world” here refers to the world of systemic interactions (exchange, warfare, diplomacy, communication, intermarriage, etc.) that reproduce the social structures and institutions of human groups. In this sense “worlds” were small when transportation and communication technologies imposed the tyranny of distance that constrained the consequences of interaction to extend relatively short distances. These were the small social worlds in which people lived (Chase-Dunn and Lerro 2013; Chase-Dunn and Mann 1998).

This systemic network conceptualization of a world-system makes it possible to study the sociocultural evolution[3] of these systems over very long time periods, and to observe the attributes that engendered social complexity[4] and hierarchy within and among interacting polities. We can compare the structures and processes of egalitarian world-systems (in which hierarchies were minimal within and between polities) and to study how and where hierarchical relations evolved as world-systems got larger and became global.[5]

 The goal of this article is to explain and justify the research designs and operationalizations that we propose for comparatively studying the structures of small, medium-sized and large world-systems and the causes of the evolution of structural complexity and hierarchy within and between polities. We explain the frameworks that have been used by the Settlement and Polities (SetPol) research working group[6] for the specifying the spatio-temporal boundaries of whole interpolity systemic networks and give examples of studies that have been done using these cases and another framework that uses constant world regions.

The Comparative Method

One of the main tenets of the logic of the comparative method requires independent instances of a process as cases for inferring causal relations among variables (Zelditch 1971). Valid inferences can be drawn from comparing cases only when those cases are independent instances of the process under study.  One cannot increase the number of cases meaningfully by including the same case twice because this is not an independent instance.  This problem was first highlighted in social science by Sir Francis Galton's critique of the compara­tive method that had been proposed by Edward Tylor.  Tylor had proposed that the causes of social struc­tural and cultural characteristics could be studied by comparing correlations of attributes across cultures.  Galton pointed out that some of the resulting correlations would stem from cultural diffusion rather than from the functional interdependence of institutions or other endogenous social characteristics.

  Raoul Naroll (1968) suggested that this problem could be solved by only comparing cases that were distant enough from one another that cultural diffusion could not have occurred.  This is one way of "controlling out" the effects of variables that are the result of interactions or connections among cases, in this case diffusion. Another approach measures connectedness or interaction and includes these sources of non‑independence as measured variables in the causal model, thus statistically controlling for them. This solution has been used extensively in cross-national comparative research to study the effects of dependence and interdependence among interacting cases (national polities) within a single larger interaction network of national societies and it could also be used in comparisons of interpolity networks that are becoming linked with one another if measures of the links could be operationalized. Another approach is to use Granger causality that estimates causal effects among characteristics that vary in a single case over time (time-series) based on the rule antecedence.[7] This can be used when you only have one case, as with the contemporary global world-system. But the simplest and cleanest way to test causal models of the rise of complexity and hierarchy within world-systems is to develop a set of largely independent cases interpolity interaction networks) to use as comparative cases in which the processes under study cannot be due to connectedness or interaction.

The comparative method is an adaptation of the general logic of the experimental method for the study of instances that cannot be feasibly subjected to experiments. But some situations are closer to experiments than others, especially when confounding processes such as diffusion or long-distance interactions have been controlled out by relative isolation, as in islands that are far from continents or other large islands. The sociogeography of islands has been a fruitful approach used by archaeologists to study human sociocultural evolution (e.g. Patton 1996).  The frameworks presented here have been developed to allow us to use the comparative method to study processes of development by comparing independent systems to one another.


Theoretical Underpinnings

We are institutional materialists[8] who are inspired by theoretical perspectives from anthropology, archaeology, sociogeography, world history, ecology, complexity theory, sociology, global studies, political science, and economics. Of course, we are standing on the shoulders of the builders of the world-system perspective developed by Immanuel Wallerstein, Terence Hopkins, Samir Amin, Andre Gunder Frank, and Giovanni Arrighi. This article focusses on research design, units of analysis, and the comparative method. It is not about theories but about how macrosociological hypotheses about the causes of social change need to be studied. We are working on a synthetic model of the causal processes that have produced the expansion and evolution of human world-systems (Inoue and Chase-Dunn 2020), but that is not the main topic we are addressing in this paper.[9]


Bounding World-Systems Across Space and Time

As stated above, we contend that whole world-systems of different sizes existed prior to the advent of the now-global (Earth-wide) single world-system of today. This perspective on the human past enables us study and to test theories that explain social change by comparing whole systems that are independent from one another.  But how do we bound world-systems spatially and temporally to compare them? We begin with the understanding that all world-systems are defined by interaction networks that are systemic, meaning that the reproduction and change in local social structures are heavily dependent on the interactions. We do this by employing a method suggested by archaeologist Colin Renfrew (1977). Because all human polities interact with neighboring polities there has been a single global human interaction network since modern humans came out of Africa if we count all indirect connections. But this would be to ignore what geographers call the “tyranny of distance.” The real consequences of human actions did not extend very far when communications and transportation technologies were mainly about carrying things on ones’ back (or head) or moving in small boats. The neighboring polities with which a home place exchanged, cooperated, and competed, and perhaps the neighbors of those adjacent neighbors were affected by what happened within the home polity, but beyond that the effects were not systemic. So, in order, to study small world-systems we need to employ the idea of fall-off, that the consequences of actions decrease as one moves away in space. And to bound a system this way we need to pick a focal locale from which to start. The method of bounding is place centric. That place can be a settlement or a polity or a region. Then the question becomes what is the nature and strength of the human interaction networks that are systemically(strongly) connected with the focal locale?

Interactions in systemic networks are on-going and two-way. They involve the production and distribution of the necessities of life as well as the provision of security from attacks and opportunities for alliances and conquest. The networks of production and distribution of food and everyday raw materials (bulk goods) are always systemic, and the political/military interactions among polities are always systemic. There two kinds of network often have different spatial scales. The bulk goods network is often smaller (more local) than is the political-military network of warfare and alliances among polities. And in some systems trade in prestige goods are also systemic. Prestige goods usually

travel farther than either bulk goods or political military interactions (see Figure 1). Information flows for both local and long-distance interaction but are rarely systemic in the absence of trade. Ideas and technologies may diffuse very long-distances with systemically linking their points of origin with those that adopt them. These exogenous factors may have large effects without creating a systemic link.  Whether they are systemic or not depends on the importance the reproduction of local social structures. Prestige goods can be important in different ways. In some systems elites monopolize access to prestige goods that are necessary requirements for marriage. Anthropologists call these prestige goods systems (Blanton, Kowalewski, and Feinman 1992; Peregrine 1991, 1996. In this case the spatial boundary of the world-system may include the places from which the prestige goods come if interruption of the supply would have major consequences for the focal locale.

A close up of a logo

Description automatically generated















Figure 1: Nested Interaction Networks:BGN= Bulk Goods Network; PMN= Political/Military Network; PGN= Prestige Goods Network; IN= Information Network


A Very Small World-System

The Wintu, indigenous sedentary hunter-gatherers late pre-historic Northern California, have been studies as the center of a very small world-system (Chase-Dunn and Mann 1998). Their

interaction networks linked households and tribelets (small-scale polities) across major linguistic divisions. These interaction networks were established through warfare, trade, and intermarriage ties. A prestige goods network connected Northern and Central California based on the use of clam-shall disk beads – a storable form of wealth that was used to trade for both food and other valuables. This allowed villages heads to use disc-beads to trade for needed food when food was scarce and so this exchange network reduced raiding to get food. This was not a “prestige goods system” of the kind describe above, but the exchange of valuables and long-distance down-the-line trade networks[10] were systemic for reproducing a rather high-density system of sedentary foragers in Northern California.

 Chase-Dunn and Hall (1997) made a distinction between core/periphery differentiation and core/periphery hierarchy. Core/periphery differentiation exists when polities with different degrees of population density are systemically interacting with one another. Core/periphery hierarchy involves interactions among different polities that is asymmetric because one or more of the polities (the core) is

A close up of a map

Description automatically generatedFigure 2: Nested Interaction Networks in Northern California: BGN= Bulk Goods Network; PMN= Political/Military Network; PGN= Prestige Goods Network (Source: Chase-Dunn and Mann 1998: 74).


exploiting and/or dominating the non-core polities. This is imperialism. In indigenous precontact Northern California there was some core/periphery differentiation – the valley-living Wintu had bigger villages than the hill-living Yana, but there was little in the way of core/periphery hierarchy. Wintu headmen were somewhat more likely to marry non-Wintu women that were Yana headmen. This was partly due to a difference in the availability of food. So, this was a small world-system that did not have much in the way of interpolity hierarchy.  The Wintu had been slowly moving into Yana territory over a period of several centuries. These sedentary hunter-gatherer polities were not internally very stratified. There were no social classes. The headman was typically a good hunter and could deliver effective speeches at Big Times and Trade Feasts. But he did not have much power over the other families within the polity. So, a situation of relative internal equality corresponded with a situation of relatively equal relations between polities.

The prestige goods network of the Wintu and their neighbors was organized as gift-giving among village leaders who competed with each other to establish and maintain reputations of generosity.[11] This also served as an alternative to raiding during times of scarcity helping to reduce the level of interpolity conflict (warfare). A reduction in warfare facilitated the conditions for increasing population density, which is often associated with greater political stability and growing economic wealth.

All world-systems, large and small, exhibit some cyclical changes. Once hierarchies have emerged in chiefdom and state-based systems there are cycles of the rise and fall of dynasties within polities and the rise and fall of largest polities within interpolity systems. But in small world-systems with little in the way of within-polity or among-polity hierarchies there are still cycles of the expansion and decline of trade networks—periods of “globalization” in which long-distance trade expanded and became more dense, and periods of “deglobalization” in which long-distance trade decline. These cycles can be seen in the archaeological evidence as operating in the small world-system of Northern California. Chase-Dunn and Mann (1996:36, 140-141) discuss “pulsating” trade networks and describe archaeological evidence for the rise and fall of interpolity trade networks based on different kinds of shell beads that emerged to link the small-scale polities of Northern California with peoples in the Great Basin and in Central California. The first wave that linked the coast of Northern California with the Great Basin emerged from about 2000 BCE to 200 BCE, then contracted from 200 BCE to 700 CE, and then expanded again from 700 Ce to 1500 CE.[12] Beginning in the 16th century CE there was a major expansion within what became California based on a different kind of shell (clam disk beads), that linked Northern California with bead producers near Clear Lake in Central California.[13]


Comparative Studies of State-Based World-Systems:

The Settlements and Polities (SetPol) Research Working Group

            The SetPol Research Working Group uses the same anthropological framework of comparison for studying world-systems small, medium and large world-systems since the Paleolithic, but for purposes of quantitative studies it has had to focus on systems for which estimates of key indicators are available at regular intervals over relatively long period of time.

The empirical goal of the comparative quantitative SetPol research is to identify and quantify the specific time points when the scale of sociocultural systems and interaction networks changed significantly. Since the Bronze Age, polities and interpolity systems have gone through iterations of centralization through conquest (as with the semiperipheral marcher states described by Chase-Dunn and Hall (1997)) and incorporation (see for example, Hall (1998) or Wallerstein (1973)) and decentralization as a result of declining centralized power. We focus on quantitative estimates of the territorial sizes of the largest polities in each system and the population sizes of the largest settlements.[14] These are proxies of scale for which comparable estimates are available for several world regions at regular intervals over long periods of time.

For estimating the spatio-temporal boundaries of autonomous state-based world-systems we rely on David Wilkinson’s (1987) chronograph of the changing boundaries of “state systems” – interaction networks in which states are fighting and allying with one another for purposes of military competition. Contrary to what many functionalist sociologists who study social systems believe, Wilkinson sees conflict as an important form of sociation.  Citing sociologists Georg Simmel and Lewis Coser, Wilkinson writes:

Conflict always integrates in a mildly significant way, in that the transaction of conflicting always creates a new social entity, the conflict itself.  But durable conflict also integrates more significantly, by creating a new social entity that contains the conflict but is not reducible to it, within which the conflict must be seen as occurring, which is often of a larger scale and longer-lived than the conflict that  constituted it.  It is therefore legitimate, and it is indeed necessary, to posit the existence of a social system, a single social whole, even where we can find no evidence of that whole existence other than the protracted, recurrent or habitual fighting of a pair of belligerents.  Such continuing relations, however hostile, between groups however different, necessarily indicates that both are (were or have become) parts of some larger group or system.  (Wilkinson, 1987: 34; emphasis in the original.)

Wilkinson is talking about the Political/Military Network that is shown in Figures 1 and 2 above. A slightly revised version of Wilkinson’s chronograph is presented as Figure 3.

A screenshot of a cell phone

Description automatically generated

Figure 3: Spatio-temporal Chronograph of the incorporation of regional PMNs into the Central PMN: Modified from Wilkinson (1987)

            The Chronograph shown in Figure 3 depicts where and when fourteen “urban” state-based PMNs that contained a city with at least twenty thousand residents emerged and when these separate PMNs became linked with one another with regard to durable military and alliance interactions among states. By durable, Wilkinson means the interaction link produced by a military incursion that is distant from the conquered area that only results in a temporary conquest does not produce a sufficient degree of systemic integration to count. Thus, for example Alexander’s army conquered large tracts in South Asia in the 4th century BCE and established Hellenistic state there, but it did not last. Wilkinson also does not (now) count the incursion of Mahmud of Ghazni into South Asia in 1008 CE as having produced a durable connection. So, Figure 3 shows South Asia as a separate system until it was colonized by the European states in the sixteenth century CE.  This is a high bar criterion for connectedness that does not err on the side of premature inclusion.).[15] Figure 3 also depicts the expansion of what we, following Wilkinson, call the Central PMN as it became the global (Earth-wide) system of today.[16]

            SetPol project focusses on changes in the scale of polities and cities because these are proxies of quantitative changes in the scale of complexity and hierarchy for which estimates are available over long periods of time and in spatially separate systems. We have also assembled data sets with other quantitative and categorical variables. We measure the frequency of interpolity conflicts (warfare), the frequency of rebellions, David Wilkinson’s “power configuration” variable (coding of the distribution of relative military power among the states in a PMN; see Wilkinson 2004) and a categorization of polities regarding their position in a core/periphery hierarchy (see Inoue et al 2016).

All hierarchical world-systems exhibit cyclical processes of change –through the rise and fall of large polities and oscillations in the spatial extent and intensity of interaction networks. We have studied five separate PMNs– Mesopotamia, Egypt, South Asia, East Asia and the Central PMN using Wilkinson’s spatio-temporal bounding as shown in Figure 3.[17] [18] Our study of scale changes in the population sizes of largest cities and territorial sizes of largest polities uses the following typology to designate what we mean by upswings, downswings, upsweeps, downsweeps, surges and short term and long-term collapses (see Figure 4).

Figure 4: Types of medium-term scale changes

                Upswings and downswings are normal oscillations. Upsweeps and downsweeps are unusually large-scale changes that are at least 1/3 greater than the average of the three previous peak or troughs (see Inoue et al 2012).  With these distinctions we can count the instances of different kinds of scale changes to compare the systems we are studying. The identification of upsweeps shows the timing and location of the events that constituted the medium-term instances of the long-term upward trends in the sizes of human settlements and the sizes of polities, which are important components of increases in complexity and hierarchy – sociocultural evolution. We can also see how frequently short and long-term collapses have been in whole system s(Inoue et al. 2012, 2015). And we can test theories of the causes of upsweeps by comparing the historical events that we have identified. For example, we have tested the idea that upsweeps of the territorial sizes of empires have been caused by conquests by peripheral or semiperipheral (non-core) marcher states. Inoue et al 2016 found that ten of the twenty-one instances of territorial upsweeps were produced by conquests by semiperipheral marcher states, and three were due to conquests by peripheral marcher states.  So more than half of the examined instances of territorial upsweeps were caused by conquests by noncore marcher states and the other eight were not. This means that the hypothesis of noncore development explains a lot of the upsweeps but does not explain about half of them. The phenomenon of noncore development cannot be ignored in any explanation of the long-term rise of polity sizes, but it is not the only explanation. Therefore, we are constructing a multilevel model that includes both world-system level variables and within-polity processes (Inoue and Chase-Dunn 2020).

We also found that nine of the eighteen urban upsweeps we identified in these same PMNs were produced by noncore marcher state conquests and eight directly followed, and were caused by, upsweeps in the territorial sizes of polities (Inoue et al 2016).  Whereas about half of the upsweep events were caused by one or another form of non-core development, there were a significant number of upsweep events in which the causes seem to be substantially internal.        


Interaction Networks and World Regions

          Most of the SetPol studies compare whole interaction networks (PMNs) with each other. But we realize that this framework necessitates controversial decisions about when and where interaction networks got larger. While we are confident in our decisions about these issues based on our understanding of the work of David Wilkinson and with his help with complicated decisions (see Chase-Dunn and Inoue 2021), we also recognize that many world historians and social scientists will be uncomfortable with the decisions that have been made.  For this reason, we also pursue a related, but different, framework strategy that designates temporally constant spatial boundaries of ten world regions. These regions have been chosen based on what we know about the timing of the emergence of large settlements, some being places where large settlements emerged early, and some in which they emerged late.

            But we have also considered the social science literature that has hypothesized comparisons and connections among regions in our designation of regions. We are aware of the issue of Eurocentrism in social science and the obvious point that “Europe” is not a continent but is rather a promontory of Eurasia (Lewis and Wigen 1997). Social science itself has been constructed around comparisons between East and West and so an important way to scientifically address the issues of comparison and connections is to use some of the categories that have been constructed in the past to see whether alleged differences (or similarities) are supported or contradicted  by quantitative data.

Using world regions designated in this way allows us to address the important issues raised by world historians and civilizationists who compare regions (e.g. Pomeranz 2000; Scheidel 2009, Wong 1997; Morris 2010, Frank 1998). The SetPol project is also be able to compare the use of these spatially constant regions with what we find when we use expanding networks (e.g. Chase-Dunn et al 2015b).

The SetPol project has used the world regions shown in Figure 5 to do comparisons using the quantitative estimates that we have assembled on the population sizes of largest settlements and the territorial sizes of largest polities in order to compare the results with what we have found using temporally changing interaction networks. We also use world regions to examine the idea that cycles and upsweeps may have been synchronous in different world regions.


Figure 5: Ten world regions for studying the emergence of large cities and polities

          The boundaries of the ten world regions shown in Figure 5 are specified in Chase-Dunn et al (2015a) which is a study of changes in the sizes of largest territorial states and empires.[19] The data appendix for that paper compared Europe as a world region to the expanding Central PMN to see what difference it makes when we compare an expanding network with a temporally constant world region. The SetPol project has also used world regions to examine changes in the sizes of largest cities (Chase-Dunn et al 2015b)


Bounding a World-System: Japan and the East Asian Interpolity System (PMN)

            Here we provide a discussion of the spatio-temporal bounding of a state system – the East Asian system that emerged in the Huang He (Yellow) River valley in what is now China as an example of our approach to bounding a whole world-system.[20] According to Wilkinson, the East Asian political-military network (which he calls the Far Eastern Civilization) developed separately from the Central PMN from about 1500 BCE, when a city of 20,000 residents and an early state emerged, until 1850 CE when it became engulfed by the Central PMN (Wilkinson 1987:31). Wilkinson contends that after the Battle of Baekgang (or the Battle of Hakusukinoe) around 663CE[21], Japan established a separate PMN – what he calls Japanese Civilization. Twelve centuries later, this separate Japanese PMN became absorbed into the Central PMN (around 1850 CE).  Wilkinson contends that this Japan-centered PMN was substantially autonomous from the East Asian PMN. It’s island location and its relatively homogenous ethnic composition are mentioned as evidence that it was an autonomous system as opposed to being part of the East Asian civilization (Wilkinson 1987, 1992b:68)

            We contend that Wilkinson’s logic for arguing for the Battle of Baekgang as a turning point for Japan in the formation of its own interpolity system is supported by the internal and external political-military history of Japan. Still, while Wilkinson treats Japan as an autonomous system that had formed its own political-military network from the 7th century on, for us the structural influence of the East Asian system—not only the political-military aspects but also trade and information flows—on Japan is clear and is crucial for bounding the East Asian world-system.

            Using our place-centric approach, we reinterpret Wilkinson’s Far Eastern Civilization by focusing on Japan.[22] Japanese state formation started in the 3rd century, during the Kofun era, and was completed in the 7th century with the state’s bureaucratization (Ritsuryo system) and centralization programs (Batten 2003). The emperor became the head of the state, institutionalizing and incorporating the local elites’ roles into the bureaucratic system.  Japan was renamed as “Nihon” –a name that appears in the Kojiki and Nihon Shoki, mythologizing Japan’s common ancestry amongst its people and solidifying its ethnonational identity.

            Furthermore, we maintain that despite establishing its own interpolity system (i.e. the Japanese Civilization as per Wilkinson), Japan was also part of the larger East Asian political-military network, expanding its interaction network before and after the 7th century. Local elites had historically engaged in interactions with their neighbors across the sea and although the state centralization process somewhat curtailed these types of interactions, the Yamato polity of Japan, as part of the central government, led more state-to-state interactions with neighboring polities during the 7th century period. Japanese missions were sent to Sui China during this period and their missions to Tang China continued through to the 9th century CE. 

            Prior to the 7th century, the Japanese islands were politically multipolar. Although there was a central imperial government, powerful local elites had relative political independence. There were contending local powers including Katsuragi (present Nara), Kibi (present Wakayama), and Tsukushi (Present Fukuoka), and they interacted with Kudara, a Korean state (Mori 2010). In fact, from the mid-3rd to the 6th and the 9th to the 10th century CE, there was considerable interaction between Japan and Korea, and this enhanced the power and legitimacy of both local and state elites. In Japan, Kofun (mound tombs) represent the power of elites and this tomb-building style peaked in the 5th century. Evidence of intensified interactions of Japanese and Korean polities is represented by the Japanese-style burial mounds, built between the 5th and 6th century, found in the area around the Yeongsan River in Southwest Korea. 

            Some scholars have argued that the chiefs of Yeongsan River area adopted Japanese-style tombs as a signifier of their relationship with Japan, which they believed would help stave off the competing power, Baekje (Habuta 2008; Mori 2010; Tanaka 2002). Other scholars contend that the tombs belonged to Japanese immigrants allied with Baekje and that the tombs served to restrain the chief in the Yeongsan River area (Nishitani 2001; Pak 2007; Yamamoto 2008). Either way, these burial monuments imply a significant connection between Japan (Wa Kingdom) and the Baekje Kingdom on the Korean peninsula (Pak 2010:7). Additionally, local elites in Japan took advantage of the symbolic power of tombs to enhance their political influence.  Local powerful families in Northern Kyushu (southern-most of the Japanese islands) benefited from their relationship with Baekje, which provided these local families with greater influence in Japan (Pak 2010).

            Evidence of Japan’s networked interactions with its neighbors is also revealed by the fact that while Japan did not possess the technology of producing iron in this period, there is documentation of their usage of iron weapons.  Both state and local-family powers on the Japanese islands were dependent on the Southern area of Korea for their supply of iron tools and weapons. To establish military power and authority on the Japanese islands, where many small states competed with one another, many of these sought iron from Gaya (or Mimana, a confederacy of territorial polities on the Nakdong River in southern Korea) during the period from, 42 CE to 532CE (Mori 2010).  Iron weapons along with other artifacts have been found in the Kofun tombs in Japan.

            We must also understand that war-making with Korea and China, particularly the Battle Baekgang, forced the Japanese to reorganize their political institutions and military capabilities.  Defense facilities were built along the shores of Kyushu (southernmost part of the island) and in the Seto inland sea shore (Batten 2001), and centralization became essential for the coordination of military protection. The emergent Japanese central state promoted and institutionalized the Ritsuryo bureaucratic system which it had imported from China. Local elites continued to retain political importance, with some locations in Kyushu, such as Dazaifu and Hakata serving important roles in the central government. However, they were essentially under the Yamato government’s control.

            For the early states in Northeastern Asia—Japan, Korea, and China–sustaining alliances with one another were more significant than territorial conquests for establishing stability and strengthening internal authority. A place-centric approach allows us to understand the formation and expansion of the interpolity system and how the boundaries of these systems changed with time.

            According to Wilkinson (1987), ethnicity was an important factor in Japan’s evolution  and the formation of its own civilization in the 7th century.  However, we would argue that Japan’s ethnonational identity was subject to geographical “fall-off.” The emergence of Japanese identity was impacted by structural and geographical conditions. Batten (2003, 2006) contends that “a sense of common identity as “Japanese” was not present until around 700 CE.[23] He maintains that the diffusion of ethnic identity came from the top down as elites expanded and incorporated lower classes to seeing themselves as part of a Japanese society. According to Batten this only began in the 7th century during after the crystalization and centralization of the Yamato state. Initially the identity of “Japanese” was limited to the elite classes and those in the lower classes had little consciousness of being “ethnically Japanese.” Furthermore, it was only at the end of the nineteenth century that Japanese ethnic identity was adopted by the lower classes (Batten 2003:91).

            Batten (2001) also argues that the diffusion of Japanese ethnic identity followed the “loss of strength gradient” (to borrow military parlance), as it expanded the national borders of the state toward the “ends of Japan”. To clarify, in the 7th century CE, the whole archipelago was not fully integrated, but the national borders expanded as the central polity incorporated peripheral areas, mainly the Tohoku region (the North East portion of Honshu). Other areas were incorporated even later: Hokkaido (the northernmost of the Japanese islands) at the end of the sixteenth century; Okinawa (the southernmost islands of what is now Japan) at the beginning of the nineteenth century (Batten 2001:28). The ethnic identity and culture of the core Japanese polity diffused from the center to the geographically distant areas as the political border of the central state expanded.  Formally sovereign territories were incorporated into state control and Japanese cultural ethnicity expanded to the lower classes and geographically, although this process was uneven and only partial on Hokkaido and Okinawa where elements of older local cultures are still present.


A close up of a logo

Description automatically generated

Figure 6. Diffusion and fall-off effects of Japanese state control and ethnonational formation on the Japanese islands

                   Figure 6 illustrates how integration was achieved by a top-down process in which the elite classes incorporated the lower classes into the Japanese national identity and into control by the central Japanese state, and that these processes also expanded territorially to include the “ends of Japan” -- Hokkaido and Okinawa.[24] The ruling classes in the Japanese core were bound up with a dense sense of the ethnonational identity, but as the social and geographic locations became more distant, this sense of “Japanese” identity became uneven and sparse. The long length of time that a unified Japanese ethnic identity took to expand suggests that the single ethnic identity began from the center and that the Japanese islands were significantly multicultural for centuries and that remnants of the former non-Japanese cultures still survive. It is of course, indicative of the fall-off rule for bounding the system. Further, this suggests that until the 7th century, this heterogeneity of national identity and the lack of the sense of unity with core Japan allowed peripheral powers within Japan to seek their independent interactions with the overseas polities. We conclude that what David Wilkinson called Japanese Civilization was never an autonomous PMN but was part of the larger East Asian system since the emergence of a state on the Japanese islands. Prior to that there were autonomous small-scale stateless world-systems on the islands that became Japan that were only intermittently connected with the Korean peninsula by migrations and diffusions that were exogenous to the small world-systems of hunter-gatherers, horticulturalists and fishing peoples of the islands.

A Comparative Framework for Studying Human Social Change Using World-Systems as a Unit of Analysis

We have outlined and explained the comparative frameworks that have been developed by the SetPol Research Working Group for studying human sociocultural evolution using systemic networks and world regions. We do not claim that these are the only valid ways to test theories of social change. The SetPol project is collaborating with the SESHAT Project[25] that is developing a Global History Databank focused on thirty Natural Geographic Areas spread around the world and targeted to represent low, medium, and high levels of complexity (see Turchin et al 2015a and 2015b). Other world-system scholars claim that there has been a single global system since the emergence and dispersal of modern humans from Africa (see the discussion of these in Chase-Dunn, Inoue, Neal, and Heimlich 2015) and some researchers have done quantitative studies of urbanization using the whole Earth as the unit of analysis (e.g. Harper 2017). The SetPol data sets have been collected in ways that make different units of analysis easy to employ and we think that the way forward is to compare different frameworks of comparison with one another.  But we also contend that the causal structures of human interaction systems were local and regional, not global, until the nineteenth century CE, and so using the framework that actually captures the processes of development can best be done by getting the spatio-temporal boundaries of real interaction systems right. Claiming that there was a single global system in the distant past throws apples, oranges, and other things (figs) into the bowl and occludes what may be important differences in different world-systems.



Batten, Bruce. 2001. Kokkyo No Tanjo: Dazaifu Kara Mita Nihon No Genkei. (The Birth of National Border: Japanese Prototype by Looked from Dazaifu). Tokyo: NHK Books.

Batten, Bruce. 2003. To the Ends of Japan: Premodern Frontiers, Boundaries, and Interactions. Honolulu: University of Hawaii Press.

Batten, Bruce. 2006. Gateway to Japan: Hakata in War and Peace, 500-1300. Honolulu: University of Hawaii Press.

Beaujard, Philippe. 2005. “The Indian Ocean in Eurasian and African World-Systems Before the Sixteenth Century.” Journal of World History 16(4):411–65.

Beaujard, Philippe. 2009. Les Mondes de L’Océan Indien Tome I: De La Formation de l’état Au Premier Système-Monde Afro-Eurasien (4e Millénaire Av J.-C. -6e Siècle Apr. J.-C.). Paris: Armand Colin.

Beaujard, Philippe. 2012. Les Mondes de L’Océan Indien Tome II: L’Océan Indien, Au Coeur Des Globalisations de l’ancien Monde Du 7e Siècle. Paris: Armand Colin.

Blanton, Richard, Stephen A. Kowalewski, and Gary Feinman. 1992. “The Mesoamerican World-System.” Review (Fernand Braudel Center) 15(3):418–26.

Chase-Dunn, Christopher and Marilyn Grell-Brisk. 2016. “Uneven and Combined Development in the Sociocultural Evolution of World Systems.” Pp. 205–18 in Historical Sociology and World-History: Uneven and Combined Development over the Longue Durée, edited by K. Anievas and Martin Alexander

Chase-Dunn, Christopher and Marilyn Grell-Brisk 2019 "World-System Theory" Oxford Bibliography of International Relations.:// https www.oxfordbibliographies.com/view/document/obo-9780199743292/obo-9780199743292-0272.xml?rskey=NKUFuv&result=271

Chase-Dunn, Christopher and Thomas D. Hall. 1997. Rise and Demise: Comparing World-Systems. Boulder: Westview Press.

Chase-Dunn, Christopher, Hiroko Inoue, Teresa Neal, and Evan Heimlich 2015 “The Development of World-Systems.” Sociology of Development 1(1):149–72.

Chase-Dunn, Christopher and Bruce Lerro. 2013. Social Change: Globalization from the Stone Age to the Present. New York: Routledge. an auxiliary chapter of this book entitled “Indigenous North American World-Systems Before the Rise of Chiefs” available at http://s3-euw1-ap-pe-ws4-cws-documents.ri-prod.s3.amazonaws.com/9781612053288/9781612053288_online.pdf . The Data Appendix for this book is at https://irows.ucr.edu/cd/appendices/socchange/socchangeapp.htm

Chase-Dunn, Christopher and Kelly M. Mann. 1998. The Wintu and Their Neighbors. Tuscon: University of Arizona Press.

Chase-Dunn,C.  Hiroko Inoue, Alexis Alvarez, Rebecca Alvarez, E. N. Anderson and Teresa Neal 2015a

"Uneven Political Development:: Largest Empires in Ten world Regions and the Central International

                System since the Late Bronze Age."  This is IROWS Working Paper #85

available at  https://irows.ucr.edu/papers/irows85/irows85.htm Data appendix for this paper is

at https://irows.ucr.edu/cd/appendices/worregs/worregsapp.htm

Chase-Dunn, C.  Hiroko Inoue, Alexis Alvarez, Rebecca Alvarez, E. N. Anderson and Teresa Neal

2015b "Uneven Urban Development: Largest Cities since the Late Bronze Age"  IROWS

Working Paper # 98 available at https://irows.ucr.edu/papers/irows98/irows98.htm Data

appendix for this paper is  at https://irows.ucr.edu/cd/appendices/worregs/worregsapp.htm

Chase-Dunn, C., Hiroko Inoue, Levin Welch and Manjing Gao 2019 "Upsweeps and Waves of

Conflict in the East Asian World-System, 1800 BCE-1830 CE" IROWS Working

Paper #135 available at http://irows.ucr.edu/papers/irows135/irows135.htm   

The data Appendix to this paper is at http://irows.ucr.edu/cd/appendices/eaconf/eaconfapp.htm

Chase-Dunn C. and Hiroko Inoue (eds.) 2021 Systemic Boundaries: Time Mapping Globalization since the

Bronze Age. New York: Springer

Habuta, Yoshiyuki. 2008. “Zenpou Kouen Fun o Meguru Kan to Wa.” In Kodai Nihon No Ibunka Kouryu. Tokyo: Bensei Publishing Company. 土生田純之 「前方後円墳をめぐる韓と倭」。『古代日本の異文化交流』勉誠出版、. Tokyo: Tsutomu Publishing.

Hall, Thomas D. 1998. “The Effects of Incorporation into World-Systems on Ethnic Processes: Lessons from the Ancient World for the Contemporary World.” International Political Science Review/Revue Internationale de Science Politique 19(3):251–67.

Harper, Antony 2017 "The Punctuated Equilibrium Macropattern of World System Urbanization and the Factors that Give Rise to that Macropattern" Social Evolution & History, Vol. 16 No. 1, :86–127

Inoue, Hiroko, Alexis Alvarez, Eugene N. Anderson, Kirk Lawrence, Teresa Neal, Dmytro Khutkyy, Sandor Nagy, Walter DeWinter, and Christopher Chase-Dunn 2016 "Comparing World-Systems: Empire Upsweeps and Non-Core Marcher States Since the Bronze Age" IROWS Working Paper 56. https://irows.ucr.edu/papers/irows56/irows56.htm

Inoue, Hiroko, Alexis Alvarez, Eugene N. Anderson, Andrew Owen, Rebecca Alvarez, Kirk Lawrence, and Christopher Chase-Dunn. 2015. “Urban Scale Shifts Since the Bronze Age: Upsweeps, Collapses and Semiperipheral Development.” Social Science History 39(2).

Inoue, Hiroko, Alexis Alvarez, 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).

Inoue, Hiroko and C. Chase-Dunn 2020 "A Multilevel Spiral model of Sociocultural Evolution:

            Polities and Interpolity Systems" IROWS Working Paper

126  http://irows.ucr.edu/papers/irows126/irows126.htm

La Lone, Darrell E. 1993 “World-Systems and Transformations of Production in the Inca Empire.”  Pp. 17-41 in The Economic Anthropology of the State, edited by Elizabeth Brumfiel. Monographs in Economic Anthropology, No. 11. Lanham:  University Press of America.

Mauss, Marcel. 1973. “Essai Sur Le Don: Forme et Raison de l’échange Dans Les Societés Archaiques.” in Sociolgie et Anthropologie. Paris: Presses universitaires de France (PUF).

Mori, Kimiaki. 2010. Mimana Nihon Fu Towa Nanika: 森公章 「「任那日本府」とは何か」。「"任那日本府"の謎」『日本と朝鮮半島2000 . Tokyo: NHK「日本と朝鮮半島2000年」プロジェクト)』日本放送出版協会、2010年。.

Morris, Ian. 2013. The Measure of Civilization. Princeton: Princeton University Press.

Naroll, Raoul 1968:  "Some thoughts on comparative method in cultural anthropology."  Pp. 236-

            277 in Hubert M. Blalock, Jr. and Ann B. Blalock (eds), Methodology in Social Research, New

            York:  McGraw-Hill

Nishitani, Tadashi. 2001. “Kankoku No Zenpou Kouen Fun o Meguru Shomondai.” In Zenpou Kouen Fun to Kodai Niccho Kankei. Chosen Gakkai. 西谷正 「韓国の前方後円墳をめぐる諸問題」。『前方後円墳と古代日朝関係』朝鮮学会、. Tokyou: 前掲:『朝鮮学報 179輯』 朝鮮学会、.

Ortman, Scott G. Andrew H. F. Cabaniss, Jennie O. Sturm and Luís M. A. Bettencourt. 2015. "Settlement scaling and increasing returns in an ancient society" Science Advances, 1 (2) http://advances.sciencemag.org/content/1/1/e1400066.full

Pak, Chon Su. 2007. Kaya to Wa—Kan Hantou to Nihon Rettou No Koukogaku. Tokyo: Koudan Sha. 朴天秀『加耶と倭 -韓半島と日本列島の考古学-』講談社、. Tokyo.

Pak, Chon Su. 2010. “‘Kanhanto Nanbu Ni Wajin Ga Tsukutta Zenpou Koen Fun—Kodai Kyusyu No Kokusai Koryu.’ Kokusai Kankei Gaku Ron Syu. 5. 朴天秀 2010「韓半島南部に倭人が造った前方後円墳 -古代九州との国際交流-.” in 『国際関係学論集』 5巻第12合併号、九州国際大学、. Fukuoka.

Patton, Mark 1996 Islands in Time: Island Sociogeography and Mediterranean Prehistory. London: Routledge.

Peregrine, Peter. 1991. “Prehistoric Chiefdoms on the American Mid-Continent: A World-System Based on Prestige Goods.” in Core/Periphery Relations in Precapitalist Worlds, edited by C. Chase-Dunn and T. D. Hall. Boulder: Westview Press.

Renfrew, Colin. 1977. “Alternative Models for Exchange and Spatial Distribution.” Pp. 71–90 in Exchange Systems in Prehistory, edited by T. K. Earle and J. E. Ericson. San Francisco: Academic Press.

SetPol Research Working Group, Institute for Research on World-Systems, University of California-Riverside (https://irows.ucr.edu/research/citemp/citemp.html)

SESHAT Global History Databank (https://evolution-institute.org/projects/seshat/)

Smith, Charlotte Ann. 2002. “Concordant change and core-periphery dynamics: a synthesis of highland Mesoamerican archaeological data” University of Georgia, PhD dissertation.

Smith, Michael E. 2005. “City size in late postclassic Mesoamerica” Journal of Urban History 31, 4: 403-434.

Tanaka, Toshiaki. 2002. “Kankoku No Zenpou Kouen Fun No Hisousha—Zoubo Syudan Ni Taisuru Shaken.” in Zenpou Kouen Fun to Kodai Niccho Kankei. Cyousen Gakkai. 田中俊明 「韓国の前方後円形古墳の被葬者・造墓集団に対する私見」。『前方後円墳と古代日朝関係』朝鮮学会. Tokyo.

Tilly, Charles. 1995. “Globalization Threatens Labor’s Rights.” International Labor and Working-Class History 47(Spring):1–23.

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

Turchin, Peter, Rob Brennan, Currie Thomas, Kevin Feeney, Pieter Francois, Daniel Hoyer, Joseph Manning, Arkadius Marciniak, Daniel Mullins, Alessio Palmisano, Peter Peregrine, Edward A. L. Turner, and Harvey Whitehouse. 2015a. “SESHAT: The Global History Databank.” Seshat Website.

Turchin, Peter, Rob Brennan, Currie Thomas, Kevin Feeney, Pieter Francois, Daniel Hoyer, Joseph Manning, Arkadius Marciniak, Daniel Mullins, Alessio Palmisano, Peter Peregrine, Edward A. L. Turner, and Harvey Whitehouse. 2015b. “The Global History Databank.” Cliodynamics 6(1).

Turchin, Peter and Sergey Nefedov. 2009. Secular Cycles. Princeton: Princeton University Press.

Wallerstein, Immanuel. 1973. “Africa in a Capitalist World.” Issue: A Journal of Opinion 3(3):1–11.

Wikipedia n.d. "Archaic Globalization" https://en.wikipedia.org/wiki/Archaic_globalization. accessed September 11, 2020.

Wilkinson, David. 1987. “Central Civilization.” Comparative Civilization Review 17:31–59.

Wilkinson, David. 2004. “The Power Configuration Sequence of the Central World-System, 1500-700 BC.” Journal of World-Systems Research 10(3):655–720.

 Zelditch, Morris, Jr.1971 "Intelligible Comparisons" Pp. 267-308 in Ivan Vallier (ed.), Comparative

            Methods in Sociology, Berkeley:  University of California Press


[1] The term “settlement” includes camps, hamlets, villages, towns and cities. Settlements are spatially bounded for comparative purposes as the contiguous built-up area.

[2] We use the term “polity” to generally denote a spatially-bounded autonomous realm of sovereign authority such as a band, tribe, chiefdom, state or empire.” We use this term instead of “societies” because autonomous realms of authority are usually easier to bound spatially and are societies, as persuasively argued by Charles Tilly (1984) and Michael Mann (1986). Tilly (1984) pointed out that societies (defined as communities that share a common language and culture) are messy entities when we consider interaction networks.

[3] Use of the word “evolution” still requires explanation. We mean long-term patterned change in social structures, especially the development of complex divisions of labor and hierarchy. We do not mean biological evolution, which is a very different topic, and neither do we mean “progress.”

[4] Here we mean complexity regarding the division of labor—the number and specialization of roles in the mobilization of social labor. Even small-scale human polities have complex languages and kin-ship systems.

[5] The Wikipedia (n.d.) article on globalization as interconnectedness recognizes what is called “archaic globalization” as a process of network expansion that allegedly began with the rise of the early civilizations. This is an improvement over presentism, but it still occludes the study of small world-systems that did not have states or cities.

[6] The Settlements and Polities (SetPol) research working group is a project of the Institute for Research on World-Systems at the University of California-Riverside. It is affiliated with the Political Economy of the World-System Interuniversity Consortium (PEWS-IC) https://irows.ucr.edu/pewsic/pewsic.htm

[7] Granger causality infers causal relationships between time series variable by using the assumption that causality runs from the past forward (antecedence).

[8]Institutional materialism is an interdisciplinary approach that combines focusing on the historical emergence and development of humanly constructed institutions (language, kinship, production technology, states, money, markets, etc.) and the changing ways that humans interact with their biological and physical environment.


[9] A summary of the roots of the world-system perspective that emerged to comprehend the modern Europe-centered system is in Chase-Dunn and Grell-Brisk (2019).

[10] A down-the-line trade network is one in which the same goods pass from one group to adjacent neighbors and then move to from them to their neighbors and so goods can pass over long distances in this way. In indigenous Northern California bow staves made from Yew trees that only grow near the McCloud River moved by down-the-line trade to the San Francisco Bay area some 200 miles to the south. There were no long-distance traders because strangers from unknown places were usually killed.

[11] For extended reading on the role of gift and gift-giving in society, consider the work of Marcel Mauss (1973).

[12] The SetPol project uses Common Era (CE) and Before Common Era (BCE) to indicate calendar years.

[13] An overview of stateless world-systems, including the rise of systems that contain chiefdoms is provided in Chase-Dunn and Lerro 2017: Chapters 6 and 7 and in an auxiliary chapter of this book entitled “Indigenous North American World-Systems Before the Rise of Chiefs” available at http://s3-euw1-ap-pe-ws4-cws-documents.ri-prod.s3.amazonaws.com/9781612053288/9781612053288_online.pdf . The Data Appendix for this book is at https://irows.ucr.edu/cd/appendices/socchange/socchangeapp.htm

[14] Discussion of the use of the population size of the largest settlement as an indicator of complexity is in Inoue et al (2015). Discussion of the use of the territorial size of the largest polity is in Chase-Dunn, Inoue, Alvarez, Alvarez, Anderson, and Neal (2015). Bronze Age chronological issues are discussed in Chase-Dunn et al (2015). We are indebted to those prodigious coders who made quantitative comparative studies of settlements and polities possible: Tertius Chandler, Rein Taagepera and George Modelski.


[15] Other approaches to spatio-temporal bounding of systemic interactions are discussed in Chase-Dunn and Inoue (2021).

[16] Wilkinson’s term is “Central Civilization” but he is very clear that what he means is a network of military interactions among states, which is rather different from what the term “civilization” usually means.

[17] A good collection of information about these five cases is at https://irows.ucr.edu/cd/appendices/semipmarchers/semipmarchersapp.htm


[18] There are excellent world-system studies of Mesoamerica (Blanton, Kowalewski and Feinman (1992) and of the Andes (LaLone 1993). We have been unable to include cases from the Americas in our quantitative comparisons because estimates of city or polity sizes are not available with enough temporal resolution. Michael E. Smith (2005) provides city size estimates for Late Postclassic Mesoamerica (1200-1520 CE) but it is not possible to count cycles and sweeps because changes in city sizes over this period are not known. Charlotte Ann Smith (2002) has estimates over time for largest Mesoamerican cities, but the temporal resolution is not fine enough to see cycles and sweeps. The Ortman et al 2014 study of settlement sizes in the valley of Mexico also has temporal resolution based on archaeological phases that are too widely spaced for the study of cycles and sweeps.

[19] Admittedly some of the bounding decisions in our specification of world regions are somewhat arbitrary. We included the Caribbean with South America rather than with North and Central America because migrants from South America mainly peopled it. We made a great effort to have only ten world regions rather than some larger number of regions to keep our data gathering structure from becoming too complicated. We geocode settlements and polity capitals so that other researchers can redo the geo-temporal boundaries if they do not want to use ours.


[20] This is an elaboration of our bounding of the East Asian PMN as specified in Chase-Dunn et al (2019). That paper also discusses stateless world-systems in East Asia.

[21] One could argue for an earlier military intervention by Japan (Wa) against the state of Goguryeo on the Korean peninsula in the 5th Century CE.  The evidence is a somewhat mythical Japanese history written in the Kojiki and Nihon Shoki. The attack on Goguryeo by Wa in 404CE is also mentioned in the record about Goguryeo’s conflicts in the Gwanggaeto stele found in the city of Ji'an in China, which was the capital of Goguryeo. The stele celebrates King Gwanggaeto of Goguryeo’s success in defeating Wa and expanding his territory. This early incursion would not rise to the level of what Wilkinson means by a durable connection.

[22] Modern humans occupied the Japanese archipelago about 35,000 BCE when paleolithic hunter-gatherers began arriving from different locations in Asia. These peoples were some of the earliest on Earth to produce stone-ground tools. During the Pleistocene ocean levels were lower and land bridges intermittently connected the Japanese islands with the Asian continent. Between 14,000 and 10,000 BCE sedentary hunter-gatherers that made clay pots, and later clay figurines, arrived – the Jōmon culture. The village-living Jōmon peoples may have formed small world-systems similar in some ways to the small systems found in Northern California. Around 1000 BCE rice farmers from Korea arrived representing what is called the Yayoi culture. The rice farmers displace most of the earlier inhabitants and are the ancestors of most the later occupants of the archipelago.


[23] We use the terms ethnicity and nationality as two forms of collective identity and solidarity. This link is signified by other world-system scholars who discuss ethnonationalism.

[24] All core/periphery hierarchies in state-based world-systems are nested stratification structures, with core states containing some peripheral regions, and non-core polities including some core-like enclaves. So the Japanese islands were long in the non-core of the East Asian system, but included an internal core/periphery system. Japan eventually became a core state in the East Asian world-system in the nineteenth century CE at the same time that the East Asian PMN was being engulfed by the Central PMN.


[25] The SESHAT project is organized by the Evolution Institute (https://evolution-institute.org/projects/seshat/)