If the question is: what explains world system change over the past five thousand years of its trajectory? Our answer is: such change is the product of an evolutionary process, or better still, of an array of evolutionary processes.'

We therefore propose that world system history be examined in the first place as an incident in the evolution of the human species, for world system is one form of social organization humankind can assume, and our task is to flesh out the characteristics of that concept and its real world references. Secondly, we intend to examine the structural components of the world system, such as economic or political, as well as global and national, each of which might be expected to be subject to evolutionary process, of a distinct character but subject to the same mechanism. We assume finally that evolutionary processes are not randomly scattered but rather exhibit synchronization of a coevolutionary character.

Such a social science-based conception yields a number of conjectures about the shape and timing of world system change, hence about some crucial developments in world history. It is a conception that is especially fruitful for contemplating the big picture of the fate of the human species; it yields a periodization of world history viewed as a phased evolution of the world system; it serves as a way of ordering the growth of the world economy, and of world politics, and accretion of a world community; and for the last millennium of our experience it affords a finer-grained depiction of structural change at the global level, with an explicit role for agents of change, leading sectors and world powers. As though miraculously, all these separate processes appear interdependent, and well-synchronized.

All of this adds up to an extended theoretical conjecture. At its heart, it consists of the spinning out of a few basic rules of evolutionary processes. But unlike nineteenth-century schemes of evolutionism, it also purports to be testable in accord with contemporary methods of the social sciences and with full attention to systematic historical data, and has in parts already been tested. Correspondences between predictions derived from these conjectures and standard accounts of world history will also be discussed. But our principal purpose here is to lay out the basic logic of this argument and to demonstrate how all of it forms an interrelated, symmetric, and mutually supportive structure that derives its strength from being grounded in evolutionary theory.

World system evolution is the story of humans learning to be human: learning

[25]

to live with each other, and doing so on a large scale and in global settings. It is the story not of a movement guided either solely by instinct or directed toward a preordained end or design, but rather that of a continual and continuing search and selection among variety generated by evolutionary processes.

Let us begin by establishing a definition of our basic term, world system, and show how it accommodates a 'process' conception. As already noted, we define world system as the social organization of the human species. That means that world system is an attribute of the human species viewed as one population. That also implies that such a population might at any time be in either one of at least two states: either unorganized, or more or less organized; and having in common basic institutions such as cities or writing, states or state systems, technologies or trading networks.

World system therefore is, in the first place, a form of species organization. A species is a population of individuals that interbreed, and that share common attributes and a common fate. Such a population forms a complex system and represents a collection of organisms that is subject to evolutionary process. Biologists inform us that the number of animal species on this planet is in the order of one million (Lumsden and Wilson 1981:4) and about each of these the questions might be asked: is that species organized, and how well is it organized? For by the condition of interbreeding alone, they are intimately related, and share important symbioses that include forms of social organization. They signal in order to establish mutual recognition as members of the same species; they cooperate if only in the process of reproduction; they experience competition and conflict as selective mechanisms; and they depend on certain resources in order to survive and continue breeding.²

The species concept thus implies interdependence among its members. That interdependence is initially a matter of quite close common origin. It is now thought that all modern humans might derive from a small population of maybe 2,000 individuals 400,000 years ago whose descendants then populated the earth, spreading to various parts of the globe over routes that have remained significant to this day, such as those of Central Asia north of the Himalayas, and the maritime ones of South and Southeast Asia. Common origin, as well as mobility (search for independence), seem to be the characteristic features of human behavior from early on.

Thus in their distribution, modern humans have settled worldwide, probably as a result of superior communication skills, expanding from Africa to every continent by, at the latest, 15,000 years ago (Asia, 60,000 years ago, Europe 35,000, Australia 40,000, the Americas 35-15,000 (Cavalli-Sforza et al. 1994: 112,154-7). But this was clearly not a coordinated movement; the numbers were quite small, and long-range communication let alone organization was lacking, and did not

[26]

satisfy the minimum conditions of world system: common identity, solidarity, collective action, and a resource base. Yet in our view the first steps toward such a world system seem to have been undertaken not too long after the spread of-modern humans worldwide, some 5-6,000 years ago.

If interdependence is a characteristic of the human species, then various kinds of linkages are pan of its experience, and basically unsurprising, even though unevenly distributed. That is why we do not make the existence of networks of interdependence as the primary research question facing students of world systems. Networks of interdependence are important parts of a world system but pointing to, for example, certain trade or migration routes of modern humans is not enough to show that a world system exists. To demonstrate the existence of a world system we need to show institutions that potentially or actually are of species-wide impact and significance.

The first basic real-world referent for the concept of world system is the population of humans. Table 2.1 is a summary of the salient figures of world population over the time span relevant to this analysis.

If we ask: has the human species, as depicted in the long perspective of Table 2.1, formed a stationary system, a condition in which social processes merely reproduce themselves, without any changes? Or has it been a dynamic one, undergoing substantial, and even dramatic change? The answer is obvious: the human species, as indexed by world population, has been, throughout historical experience, not just reproducing itself but expanding at a rapid rate. It is therefore to be presumed that its social organization, too, must have been changing with comparable rapidity.

We might wonder about the direction of the causal arrow between population growth and social organization it, the rise in population that caused changes in the social organization and the emergence of world system potential? Or is it that new forms of social organization, epochal innovations such as the rise of cities, or industrial revolutions, made possible the rise of populations? We would tend to argue that both types of causation are in fact relevant, at different points of human social experience.

Even at this preliminary stage it is easy to see that world system history is not about a stationary world system that keeps reproducing itself but about human social organization undergoing a sustained process of long-term change. That is why the appropriate stance for such an undertaking is not a reduction of the

Source: After McEvedy and Jones 1978:344, 349.

[27]

system to a number of static entities such as societies or civilizations but a process conception that allows us to capture variability of structures and movement in time. We do not hold the world system to have emerged fully fledged and perfectly connected some five millennia ago. Rather we are attempting to model the rise of the world system from a condition of potential, over some preliminary stages, toward a more fully developed status, and a future condition yet to be determined.

All this presupposes what we might call the 'singular' conception of the world system. Such a conception maintains that the human species has exhibited, and will continue to exhibit, important regularities and common behavioral patterns not only at the level of individuals but also at the collective level of institutions, and organizations for humankind. It stands for the idea that behavioral uniformities of humanity viewed as an interacting whole - uniformities such as urbanization, or war - are worth knowing about, and that the changes in those characteristics are a most important subject of study.

There are some interesting parallels here with interdisciplinary debates about the origin of modern humans. On the plural (or polycentric) view, homo sapiens emerged separately but by parallel evolution to form distinct races in the world's major regions. The 'singular' view, supported by contemporary genetic research, sees to-day's world population as the product of a single expansion of 'anatomically modern humans' from Africa over the past one hundred thousand years (Cavalli-Sforza el al. 1994:62).

The singular conception of world system is exemplified also in the work of A. Gunder Frank and Barry Gills (1993, also this volume) and might be contrasted with the 'plural' one, that proceeds from the existence of a number of world systems (as in Chase-Dunn and Hall 1997, also in this volume). These are called world systems not because they extended potentially or actually to the human species as a whole or because they covered the entire planet, but because they are deemed to have been self-contained, or regarded by their members as 'worlds-in-themselves.' In the plural conception, the major questions become those of system identity and differences among systems, of relations among world systems, and of mergers of world systems into 'super' world systems.³

Our own preference would be to use terms such as 'regional' or 'local' for systems that may seem to have been self-sufficient over some periods, or whose distinctive traits appear to confer a special identity, and reserve the term 'world-system' to that one whose operations, potentially or actually are coextensive with humanity. The distinction is not unimportant. It is not merely an idiosyncratic choice between 'lumpers' who see things writ large, and 'splitters' who see reality, in the first place, only in the microcosm. It is also a choice between the sort of question one deems important: a broad-gauged inquiry into the behavior of the human species, or the detailed accounting of the fates of some individual societies as compared with others, each interesting in their own right, and that, in the aggregate, might add up to the story of humanity. The difference is not a new one,

[28]

with merit on both sides, and it reflects contemporary debates in the study of world history.

For Condorcet, in his pioneering account of human progress, or for Kant, in his concept of 'universal history,' the story of the world was, in the first place, the story of humanity. For them, there was only one (evolving) world system, and one world civilization. Contrast this with influential conceptions of 'multiple' (or plural) civilizations viewed, in Toynbee's (1934:51) words as 'the intelligible field' of history conceived as the 'comparative study of civilizations.' For the British historian, as for Oswald Spengler, 'civilizations' were the societies that had a 'greater extension, both in Space and in Time, than national states or city states' but none of which embraced 'the whole of Mankind' (ibid., p.45). The story of mankind was for them an account of the life cycle of these distinct civilizations, of which more than twenty were identified by name.⁴

This 'plural' perspective, of a number of separate civilizations pursuing essentially independent careers may be contrasted with William McNeill's (1963/ 1991) position that the cultures of mankind have experienced a significant degree of interaction with other cultures at every stage of their history, and never more so than when great transformations were underway in the world system. For McNeill, the present state of world organization is the consummation of a single continuous process that he recently (1991:xxii) described as 'ecumenical.'

A process conception of the world system naturally focuses upon sequences of change in the social organization of humanity. But which, specifically; are those changes that need to be explained?

It is the strength of this approach that it allows for asking questions both about structure, and about agency. World System Evolution, in its broad sweep, is about structural change of planetary scope, and it proposes questions at two principal levels of structure: about major institutional change, such as the rise of the market economy, and second, about organizational change, such as the emergence of global organizations in tandem with the nation state. But it does not ignore either the role of agents. For social change is driven, at the grass-roots level, by innovation, and it is the innovators who are the agents of change. The long cycles of global politics that drive global political evolution, and Kondratieff (K-) waves that motivate their economic counterparts, are propelled by important innovations, and often well-known innovators.

What changes need to be well explained in the world system?

1 Eras of world system history, which reflect the conventional periodization of the story of the human community. We shall conceive of them as a sequence of major changes in world institutional arrangements.

[29]

2 World social, political, and economic change. This too can be represented as a sequence of major institutional changes of a more specialized kind.

3 Global change in the modern era that might be seen as organizational change, in turn driven by the world powers of the long cycle, and by leading industrial sectors of the global economy.

4 What is the broad context of world system change? What is human evolution and what constitutes civilization?

These are, at different levels of analysis, four sets of questions about key aspects of the clearly expanding world system of the past five millennia. The first looks upon the world system as a whole. The second distinguishes between the several components that make it up, while the third highlights prominent vertical structures of the modern era, the global system, and its product, globalization. The last concerns the evolution of modern humans. All four pose large questions and demand solid explanations. Together, they compose a scaffolding for the exploration of world system evolution.

To answer these large questions about the fate and vicissitudes of the social organization of the human species we employ an evolutionary explanation. What are the elements of an evolutionary explanation applicable to world system processes (see also Modelski 1991, 1994, 1996, 1999a, Andersen 1994:14ff)?

Let us propose that an evolutionary explanation of a 'social fact' requires reference to previous such facts as well as a causal link that includes the following four mechanisms: those of(l) variety-creation; (2) cooperation (and segregation);

(3) selection; and (4) preservation and transmission. In turn, these four mechanisms, taken together in that sequence, constitute a social learning algorithm.

We also propose that world system processes might be seen, at their several levels of analysis, at the institutional and organizational levels, and in respect of agency, as propensities for major social learning processes. At a very rough approximation, we might say that the first of the mechanisms is cultural, the second primarily social, the third political, and the fourth, economic. Since such a synthesis has to be an ordered one, all world system processes have a time-structure that allows for successive optimizations of these mechanisms in a formal-logical learning sequence, in the order in which the mechanisms were presented in the previous paragraph. World system processes can therefore be seen as possessing the make-up of four-phase temporal learning experiments.

The underlying premise is that evolution is proportional to time. That is the hypothesis of a 'social evolutionary' clock, not unlike that of a 'molecular clock' timing genetic mutations over long time periods, and helping to chart the time elapsed in the separation of two species. Cavalli-Sforza et al. (1994:33) use that same postulate to study the history and geography of human genes. They define the rate of evolution as the amount of evolutionary change - measured as the genetic distance between an ancestor and a descendant - divided by the time in

[30]

which it occurred, and they report to have found that the hypothesis yields good results when used for comparing major population changes. A directly analogous procedure, known as 'glottochronology' (Swadesh 1971:271), has also been, employed in linguistic analysis and is used to date the origin of languages by assuming a 'relatively constant rhythm of substitution' in a basic vocabulary, with an average retention rate of some 86 per cent per one thousand years. Results from linguistic work tend to reinforce genetic studies even though the rate of genetic change is much slower than that of languages. This leads to the following postulates:

1 World system processes, both at the level of major institutions, and of major organizations, as well as agents, are evolutionary in make-up. They are self-similar in that the same explanatory logic applies at each level, implying that the system has a fractal structure. Self-similarity is 'symmetry across scale' a 'repetition of structure at finer and finer scales' (Gleick 1987:103,100).

2 World system processes each undergo change at their own rate that is proportional to time, and they each have a time-structure that integrates the four evolutionary mechanisms of variety creation, cooperation, selection, and reinforcement. Each period of a world system process consists of four such learning phases.

3 World system processes flourish in conditions of high evolutionary potential conducive to innovation.

4 World system processes are nested and synchronized (they coevolve). Nesting means that large-scale processes enfold, and are in turn animated by, smaller scale processes of determinate proportion in conditions of synchronization.

In other words, we do not search for a distinct logic for each era or structure of the world system because 'one system' requires 'one process.' We propose one common logic (or algorithm), an evolutionary learning one, to explain each world system process we identify.

The classical source of nineteenth century philosophies of history is Immanuel Kant's Idea for a Universal History (1784/1991). Kant himself was not an evolutionist but his basic ideas might be thought of as foundational for evolutionary explanations. He raised the question whether it is possible to discover, among free-willed human actions considered on a large scale, a 'regular progression,' 'in accordance with natural laws,' in the 'history of the entire species,' such that it might be recognized as a 'steadily advancing but slow development of man's original capacities.' He also advanced nine propositions as guidelines for such a history, the fourth of which identifies 'antagonisms in society' as the long-run source of a 'law-governed social order' and comes close to portraying them as mechanisms of selection. We cannot but take heart in the ninth proposition, that 'a philosophical attempt to work out a universal history of the world in accord-

[31]

ance with a plan of nature ... must be regarded as possible,' and read it as a prescription of a search for a better theory of world system change, though we also need to debate in what sense and to what degree such a 'plan of nature' might be thought to be 'aimed at a perfect civil union of mankind.'

Nineteenth century evolutionary thought assumed two main forms. The founders of sociology, Auguste Comte and Herbert Spencer, put forward bold new conceptions of the development of humanity. While aspects of their ideas, of a phased advance toward industrial society, remain influential to this day, they arc on the whole regarded as dated. Marxist thought developed along similar lines. This broad 'evolutionist' strand of thought might be contrasted with the 'selectionist' approach of Charles Darwin, whose innovative contribution was the identification of natural selection and variation as the mechanisms of species development. Darwin also launched the grand idea of a 'tree of life,' mapping the common origins of life on earth, thus in effect laying the foundations for a macro-history of biology; but he avoided any large claims to explaining patterns of human social evolution.

Social evolutionary thought experienced a revival in the mid-twentieth century, just as Darwinian biology was reinvigorated through the 'modern synthesis' founded on genetics. In economics, the work of Joseph Schumpeter, or Friedrich von Hayek, is now regarded as evolutionary, and viewed as offering an important alternative to neoclassical analysis. In sociology, Talcott Parsons was, in his later work, a significant contributor to social evolutionary theory. Karl Popper and Donald Campbell advanced evolutionary epistemology as a methodological foundation for the natural and social sciences.³

More recent social thought has been less hospitable to 'big picture' theorizing. Anthony Giddens (1984:243,236-9) for example, compounds evolutionary theory with historical materialism, and argues that it is necessary to deconstruct them both. 'In explaining social change, no single or sovereign mechanism can be specified. There are no keys that will unlock the mysteries of human social development.' 'Human history does not have an evolutionary shape' because 'history is not a "world-growth story'"; 'the relatively short period since the emergence of civilization in Mesopotamia is not marked by the continuing ascent of civilization; it conforms more to Toynbee's picture of the rise and fall of civilizations.'

We disagree with that position. The task of social scientists is to unlock the mysteries of social development, and to discern its shape, evolutionary or not. We need to attempt a synthesis of 'evolutionism' of the Big Picture, with the rich detail of Darwinian mechanisms of selectionism. We aim to explain large-scale social change, but seek to do it by carefully tracing the processes that propel such change, and the mechanisms, selection being crucial but not the only one among them.

In social theory, we need to contrast the evolutionary explanation with both rational choice, and functionalism. Rational choice (Elster 1989), whose paradigm is neoclassical economic analysis, is an elaboration of what Alfred Weber called the ends-means schema. That approach takes opportunities and resources (and constraints) as well as preferences (or interests) as given, and generates from

[32]

them a stream of intended outcomes. By contrast, selectionist and evolutionary models do not depend upon intentions; they focus on actual outcomes, and they make changes in constraints and interests as part of the model. Design, the ideal, product of rational choice, is not part of selection, that plays upon trial-and-error. Speed of response is the glory of rational choice; evolutionary change is the tortoise that moves slowly but surely over the long haul.

Selectionist models therefore appear preferable for studying evolutionary processes and macro-level phenomena such as world system change. They are not to be confused with functionalism. At the heart of functionalism is the sound idea that structures have consequences for wider systems. But classical functionalism asked: what maintains social systems? And answered, practices that respond to (postulated) social need. Critics (such as Litde 1991:101-20) were right in arguing that this did not get us very far, and that it implied an excessively optimistic, 'Panglossian' meta-theory, that 'societies will produce practices that satisfy their long-term needs.' Evolutionary explanations, by contrast, do not seek to explain persistence, but rather social change and the processes that transform social structures.

One way into the macro-analysis of world systems is through the established field of periodization of world history. McNeill's (1963) divisions of world time (shown in Table 2.2, first column) might be regarded as an example of standard specification of epochal divisions over the past several millennia (subject to his comments in 1991:xvii-xix). That periodization conforms quite comfortably under the familiar headings of ancient, classical, and modern. It is also based on the additional criterion of 'dominance,' in that it relies, in that sense, on changes in macro-political organization. In doing so, it illuminates just one facet of the world system process, but there is no reason to suppose that facet to be unimportant or unrelated to other principal components. Indeed, there are grounds for thinking of political periodicity as a kind of 'clock' timing the entire social system.

On McNeill's account, the political organization of the world system has now passed through three stages. In the first, starting well before 3000 BC, the Middle East was the center of world development. In the second, no one region occupied a similarly striking position, and the situation was one of 'cultural balance' in which 'each of the four major civilizations developed more or less freely along its own lines' (1963:253). The third stages might be thought of as a return to 'dominance,' and this time, 'Western dominance.' The long-term vision of the future, hence also possibly the fourth stage, is indicated by a reference, in the book's closing pages, to the 'establishment of a world-wide cosmopolitanism' (a Kantian term) that 'would enjoy vastly greater stability' (ibid.:806-7).

McNeill's shifts in the pattern of dominance suggest a world system tendency toward long-term alternation between equal and unequal structures, but imply a

[33]

Table 2.2 Eras of the world system

long-term tendency toward equalization, in the Kantian spirit. We might also note that me overall phase structure, of ancient, classical, modern, dovetails nicely with the general trend portrayed in the summary of population growth present earlier in Table 2.1. We observe in that earlier tabulation not only the striking pattern of growth but also its patterning in line with conventional periods of world history. With population measured roughly at mid-point, to each of the well-known stages corresponds a new order of demographic magnitude. Throughout this 'history,' the expanding population also maintains a fairly even distribution in space, with the major regions of Asia invariably accounting for more than one half of me world total. We have here a vast expansion in die numbers of the human species as a whole, hence a systemic process, but one that is also clearly patterned, roughly in line with historical periodization.

What we have is a process with a strong phase structure, both political and social (demographic), that is also suggestively cumulative and in a sense progressive. We propose an evolutionary explanation. We argue that the process is one of the emergence and consolidation of the world system. Or else we say that the process is one of launching the world system as a major institutional complex for the human species, hence an epochal innovation. The standard periodization reflects the political phasing of that process, and the cumulative demography indicates successively more effective conditions of world organization. Each phase of that process, broadly corresponding to die conventional eras is also marked by the optimization of one of the four evolutionary mechanisms.

Our model therefore proposes that me world system has now nearly passed through three (out of four) evolutionary phases: the learning-infrastructural phase (laying down the cultural base for the entire process); the community-building phase (foundational for enterprises of large-scale cooperation); and collective organization (selection of forms of worldwide organization); me fourth phase being that of consolidation (reinforcement and replication). These would represent

[34]

successive optimizations of evolutionary mechanisms of variety-generation, cooperation, selection, and reinforcement.⁶

If structural change, that is social learning viewed as problem-solving, is to occur these phases must be passed in the sequence just presented. The completion of one phase depends on the conditions created by the preceding phase, and becomes a necessary condition for the phase that follows. Each of these phases represents the principal theme of the social universe in the major epochs of its functioning; it reflects major social priorities but does not imply neglect of other domains of social organization.

The first phase (in Table 2.2, column three) labeled 'learning-infrastructural' might be understood as generating the variety for building the world system, and does so by drawing upon the resources developed in the preceding era (in this case, during the agricultural revolution, that preceded it by up to 8,000 years). That variety arises by means of city-building and gradually spreads, in a system of interconnected - and partly preexisting - networks of intercontinental proportions: cities being by definition cultural constructions that are oriented to, and closely connected with, other cities, coming to form the center of an emerging world system. Cities are the hardware, the invention of writing supplies the software of the infrastructure of world system learning. Writing records and stores information, and it organizes social life both to the past and to the future; it lends continuity to social organization and makes systematic structural changes possible. It leads to the emergence of professional classes, such as scribes or teachers; centered on temples it helps, as do cities, in differentiating culture from the social system. Writing means the start of systematic learning, and science (astronomy, calendars), makes possible intensive agriculture, and is essential in disseminating the elements of bureaucracy.

Proceeding from this learning-infrastructural foundation thus understood, the next major phase of the world system goes on to community-building on a scale going beyond tribe and city. This is the time that innovates the structures of wider cooperation implicit in the world system, because in principle such cooperation must extend to all humanity. We propose that in this second phase such cooperative potential is actuated via universal religions. Religion, in turn, forms the basis for solidarity and cooperation, enhances education and communication, large-scale political organization, and long-distance trade. Each in their own way, the great religions are in turn forms of differentiating populations and building larger regional ensembles.

Given a set of major communities, the stage is set for organizational selection. Competitive pressures of several kinds (economic competition, political conflict, ideological confrontation, scholarly debates) select the organizational forms best attuned to the emerging complexities of the world system. The collective management of human affairs becomes the operative problem, both at the new national, and new global levels. Such collective organization finally gives way to adaptation, a stabilization comprising an adaptive adjustment to the environment, preparing the stage for yet other evolutionary developments.

[35]

How does such a model square with the conventional understanding of the unfolding of the human story? Let us look once again at Table 2.2, and note some differences between the conventional, and the analytic account. Both start before 3000 BC, in Sumer. But the first column consists of periods of unequal length that nevertheless average out to a little over 2,000 years; the right-hand column shows world system phases as of roughly uniform length, and only in the first two cases is that evidently so, at just over 2,000 years; the duration of the modern and postmodern phases remains to be determined. The postulated four-phased world system process would thus extend over a total of 8,000 years; the possibility of verifying that extends to no more than five millennia.

Can the era of 'Middle Eastern dominance' be convincingly labeled as pre-eminently cultural, or 'learning-infrastructural,' as predicted by the model? We observe that the Middle East was not then dominant in the sense that it might have controlled say China, or Europe, but only in so far as it was the location of the period's major institutional innovations. For this was where cities emerged, at first mainly in Sumer, and cities were the first great human artifacts that manifested cultural achievement and laid the foundations for an entirely new form of social organization. This is, too, where writing first appeared, another major breakthrough toward social organization. Indeed, all evidence suggests that these two developments first occurred in one Sumer city, in Uruk, that around 3000 BC was the largest city of the world system (Modelski 1999b). Cities, and writing, in turn stimulated new departures in society, sparked technological and scientific innovation, as in transportation, civil engineering or astronomy, creating a set of conditions in the absence of which a world system could not emerge, and on the basis of which the great religions rose in the next era. The course of the period saw both cities and writing spread to other world regions, including China and Europe.

At the start of the classical era, that of Eurasian cultural balance, 'there existed four distinct regions of high culture in Eurasia,' and 'two thousand years later the physiognomy of Eurasia was recognizably the same' (McNeill 1963:249) and, we might add, persists to this day. As balance replaced dominance, conditions were established not just of relative autonomy but also, more importantly, in which each major world region contributed its share to world development, and shared in its evolution. What was the nature of that contribution?

Our model proposes that in this 'classical' era the major priority, and critical innovations, lay in 'community-building.' This was the era of institutional innovation of universal religions. The Eastern Chou (Zhou) era of Chinese history brought forth Confucius, whose ideas 'catalyzed the institutional and intellectual definition of Chinese civilization' (McNeill 1963:232), and in due course began to mould the cadres of its scholar-officials. The teachings of Buddha first came to be institutionalized in the Indian subcontinent, but then spread far and wide along the Silk Roads, powerfully influencing not just West and Southeast Asia but also China, Korea, and Japan. Buddhism was organized around the practice of

[36]

monastic communities. Christianity formed in the Hellenistic culture of the Mediterranean and then under Roman rule, but also outlived it. Islam first arose among the city states of Arabia but then shaped the Ummayad and Abbasid Caliphates as its framework, which resolved the issue as to how 'the community of the faithful should be led and by what principles it should be governed' (ibid.: 429). Accounting for 'much of the institutional pattern' that gave Islam its strength was 'the strategic position of the mercantile class' (Hodgson 1993:107).

We observe that all four of these cases have in common the creation of a new basis for an extended, overarching community, even potentially world system-wide. None fully succeeded, but each served to lay the foundations for a more inclusive world and each in turn contributed some important strains to the enterprise of community-building. We observe, though, a contrast with Karl Jaspers' concept of the 'axial age' of Confucius, Buddha, and Socrates (800 BC to 200 BC), that he regarded as having witnessed the creation of all fundamental cultural constructs simultaneously all across Eurasia. Our own 'learning' conception is resolutely sequential, as can be demonstrated in a study of the world cities of the classical age (Modelski 1999c) that can be shown to have become first importantly Buddhist, then Christian, and between 700 the majority Moslem, with a strong continuing Buddhist participation.

If the classical era was, in its basic thrust, Eurasian, then the modern age might more appropriately be labeled 'global,' and we might wish to suspend judgment on the question whether this should be called the era of 'Western dominance.' If, as McNeill (1982) has more recently argued, we see the modern world system to take off in Sung (Song) China, just before the turn of the first millennium, then its complexion must be a mixed one, with the Chinese Renaissance being followed by the Italian, and Atlantic Europe only taking a lead after 1500. Maybe it is too early to give a final definition to that era whose likely reach might extend for some more centuries but also possibly into space.

But we might be entitled to assert that, propelled by a powerful 'organizational revolution,' those dominating the social landscape in the modern age are the prominent forms of collective (species-wide) action, not old-fashioned empires, but nation-states, armies and navies, corporate business entities, universities and other non-governmental and governmental organizations, not forgetting either die growing network of global institutions. That is how we would define the most striking social innovations of the modern era, and it is these new powerful organizational creations that give substance to the term 'world system.'

We have now traced these through five millennia but might also observe that only in die last millennium has die world system begun to assume concrete shape through organizations and institutions of worldwide impact: in die Mongol's bid for world empire around 1250; in Portugal's global network of fleets and bases after 1515; and most recently, in die growing array of global organizations since die middle of die nineteenth century. Does that mean that no world system existed prior to die modern era, and are we indeed justified in labeling those earlier ages as world system evolution?

[37]

The answer is: world system is not built in a day, or even a millennium. It is an epochal learning project for die entire species that starts slowly at first, then gamers up nuclei of cooperation, and only in good time, after some false starts, reaches die possibility of crystallization (or punctuation, as in punctuated equilibrium), at die selection phase of that process. For die world system, that would be die stage of collective (species-wide) organization that has been underway for the past millennium, which is about die time when die fact of world organization first started to become a reality.

What have we learnt so far about world system evolution? That a gross ordering scheme cast in evolutionary terms is at least conceivable, and that a rough characterization of the eras of world history in terms of modalities of a learning process, and die innovations punctuating that process, is not inherently implausible. Such a model cannot be rejected as glaringly at variance with the conventional scheme of things, and is in fact better than might be expected from an exceedingly macro-model. It is equally clear we are still far from a full explanation, and must make this ordering more convincing by reducing the extended time-span between cause and effect, and by introducing additional mechanisms of less-grandiose temporal dimensions.

We have treated the conventional periodization as applying across die board to the world system process. But we noticed too that the basis of such periodizations is in the first place expressed along one dimension, die political, or die geopolitical one, because it invites us to view world system development through the lens of such concepts as Middle Eastern 'dominance,' or Eurasian 'balance.' Clearly political evolution constitutes one aspect of world system process, meaningfully related to other processes. How might political evolution be explained, and how might the political process be related to the main currents of economic or social change in die world system?

The world system, at least in its more developed form, exhibits along its horizontal dimension, not only political but also economic, social, and cultural (or learning) structures.⁷ It has states and intergovernmental organizations, financial markets, humanitarian projects, and the worlds of media and scholarship. These are, by now, conventional distinctions but we do note that they broadly match die array of evolutionary mechanisms discussed earlier on. If the world system has 'structural potential' then we should also be able to trace the evolution of these structures over time.

What is more, we assume these four structures to be self-similar. That is, we view each one of diem as subject to an evolutionary process of die same logical make-up, but die process proceeds at different time scales in each case. The economy is expected to change at a faster rate than political structures, and so on. Because die time scales of these processes differ, some of them might be thought of as nesting within others of longer periods, and in that sense economic change might be thought of as proceeding with the political framework. Nesting, in turn,

[38]

calls for synchronization, and that might be the reason why the time scales of the four evolutionary processes here discussed must stand in a determinate relationship to each other. The set of these forms a spectrum of interlocking periodicities, We hypothesize a set of relationships in Table 2.3.

In other words, we propose that the constituent processes of the world system interlock in a determinate manner. While all four of them undergo change at a rate that is constant in respect of a particular process, the periods of these differ in a determinate ratio. Four periods of the world economy process are equal to two periods of the (political) active zone process, and one period of the (social) center-hinterland process. The relationship, at the world level, is thus regulated by the relative scale of these event sequences, in a manner that reflects a 'cybernetic hierarchy' according to which 'the longer the time perspective, and the broader the system involved, the greater is the relative importance of higher, rather than lower factors in the control hierarchy' (Parsons 1966:9,24,113). In such terms, the overall process of world system evolution that we reviewed in the previous section might, for some purposes, be analyzed (or refracted) into four distinct evolutionary processes whose temporal dimensions stand to each other in a relationship of 1:2:4:8, each in turn composed of four phases.

World system process We hypothesize that this is the process that gives overall shape to the world system, programs it, orders its priorities, and times the major phases we discussed. It takes its cue from the world information networks and, giving a distinct texture to major eras, is at bottom cultural.

World socialisation The evolution of the human community and the growth of human solidarity is not a process of linear expansion but one of persistent tension between-the pressures for innovation that are the consequences of evolutionary processes, and a necesiary outcome of learning, and the demands for equality that is the operative condition of every community. Innovations produce concentrations of metropolitan power, and peaks of prestige, often centered on opulent cities and brilliant empires. Forming in opposition to them are the hinterlands, or the margins of civilized society, that from time to time organize themselves to effect a system leveling (or dependency reversal). It is hypothesized that major phases of concentration, a millennium in length, alternate with equally significant intervals of hinterland assertiveness and that this alternation constitutes the process of world socialization.

Table 2.3 Interlocking periodicities

[39]

We borrow this pair of terms 'center/hinterland' from Frank's studies in the 1990s. It parallels the conceptual pair of 'core-periphery' but needs to be distinguished from it because the latter proceeds from a differential division of labor while the former refers to a social relationship that might also be grounded in a differential access to political power, social prestige, or claims to innovation.

Active zone process Let us define an active zone as the spatial locus of innovation in the world system. Social and cultural evolution proceeds by means of innovation and its diffusion, and it flourishes in conditions that favor the generation of variety and, more generally, of high evolutionary potential. The political seedbeds of such variety are not powerful empires that tend to attract the attention of historians but zones of autonomous entities, such as state systems, and intermediate political networks, and more broadly, regimes and domains in which individuals, and communities enjoy openness, freedom and autonomy that foster creativity.

Thus conceived, the active zone becomes the center of the world system but only as long as it generates the innovations that respond to world problems. The active zone process is the political process that in each period of 2,000 years focuses upon a broad geopolitical zone (such as the Middle East), but in each of its phases of about 500 years moves along spatially to a new region. Standard eras of world history might be seen as periods of the active zone process.

Production/commerce The hypothesized world economy process defines changes in the major modes of organization of production and exchange, in agriculture, mining, and industry. Periods of productive development, and surges of new technologies, such as bronze, or iron, alternate with others that expand networks of interchange, pioneer new trade routes, and generally disperse innovations.

Table 2.4 presents this model for visual inspection, to show how the processes coevolve. It adds detail, and adduces some other information that puts flesh on the bare bones of this structural argument. The second column shows the three eras of the world system, reviewed earlier. In addition to bearing a conventional designation (ancient, etc.), each of them also defines a major theme of cultural development.

The next column of Table 2.4 is a schematic outline of world socialisation. We know that the center-building that produced the flowering of civilization in Sumer, and then in the Nile and Indus Valleys, did not continue in a straight line upward. If the growth of major cities is the right indicator then we observe a rise in their numbers until about 2300 BC, after which the growth abates. The number of hinterland migrations and incursions rises, and Sumer in particular experiences a drastic decline. By 1200 BC only Egypt remains secure in the line of 'barbarian' onslaughts but loses its drive and falls behind. While the Indus Valley civilization collapses, the qualities that made Mesopotamia and Egypt special now diffuse more widely in Eurasia, to Europe, China, and later India (Modelski and Thompson 1996a; Modelski 1999b).

[40]

[41]

The dispersion affected by this process served to consolidate the central portions of four civilizations, East Asian, South Asian, Mid-Eastern, and European that came to constitute the 'balance of Eurasia' in the classical era. Within that balance, each of these made its own contribution to the major process of community-building previously reviewed. Each of these civilizational regions served as the basis of a major religion. But within that process, the pulse of center-hinterland interaction was also palpable. After about the year 100, the new social structures in the several regions came under pressures of new invaders, and urban growth abated once again. The movements of Germanic tribes that broke up the Roman empire were part of the same great migration as the barbarian occupation of North China, and the Ephthalite invasions of India. The eruptions of the tribes of Arabia upon the Mediterranean world after 632 completed what appeared to be another sustained process of systemic leveling (Modelski and Thompson 1999; Modelski 1999c).

In the early modern era it appeared for a moment as though, under Mongol rule, China might become the center of the world system. But attempts at world empire collapsed, and the active zone moved first to the Mediterranean. We are familiar with the thought that after 1500 the center shifted to Atlantic Europe, in a manner such that various parts of the world became increasingly dependent upon it, in colonial and semi-colonial situations. Our analysis leads to the prediction that, once again, this extensive period of systemic concentration that culminated with the Industrial Revolution might be due for a reversal: for the world system might have entered, after 1850, onto a movement of systemic leveling, though hopefully at a higher level of organization. This leveling is taking the form of democratization, in which an initially small nucleus of democratic societies (about 10 per cent of world population at the end of the nineteenth century) is gradually bringing into the fold of a future democratic community an increasingly larger portion of the world's peoples (about 50 per cent of a much larger population at the turn to the twenty-first century).

How might this be so? On the present analysis, world socialization is now in its second period (of 4,000 years each). The first, labeled here 'Rough World' was, in a sense unsurprising, for in the course of it the splendor of civilization was regularly, albeit at long intervals, balanced by the excesses of those who attacked it in order to share in it, and sometimes to destroy it. While this was indeed a rough world, maybe there was also in it an element of rough justice.

Can the rough edges be taken off the world system, while increasing its civilizational quotient? Axelrod (1984) argues that even in a 'nasty' world, 'nice' strategies can arise as mutations of established operating procedures, and in certain situations, namely those of clustering, might not just survive but also prosper. Let us therefore entertain the proposition that the second period of center/hinterland interaction might be experiencing a move toward a 'nicer world.'

At first, we would once again expect a stage of reconcentration, powered by prodigious innovation of various kinds, giving rise to much dependency. But the same process would also be responsible for setting off mutant forms, say in Sung

[42]

China, or in Renaissance Italy, launching experiments, reformist, republican and liberal that build up potential for social and political development. It is from those mutant forms that cooperative arrangements could arise and lift the raw conflict among center and hinterland to a higher level of performance. The diffusion of democracy since the nineteenth century to i.e. Japan, India, East Europe, and parts of South America is beginning to create conditions in which demands for greater equality might find expression in forms that are 'nicer.'

The sequence of 'active zones' (column 4 of Table 2.4) has been neither arbitrary nor unsystematic. The progression described by that sequence has been along zones of spatial contiguity, and has successively lent priority to cultural, social, political and economic factors. Thus for instance in Confucianism, emphasis rests on scholarship and learning; in Buddhism, on the creation of monastic communities; in Christianity, on individuals and on church organizational factors; and in Islam, on the evolution of long-range communications based on pilgrimages and trade routes facilitated by that community.

The active zone process can be documented with die help of data on urban population growth, cities in the active zone being seedbeds of innovations. Andrew Bosworth (1995a:198ff) has shown that the process appears to capture 'those regions whose population growth outpaced those of die rest of the world, with each active zone building upon the foundations of its forerunner.' For each of the zones identified in column four after 1000 BC, that is for all those for which such data were available, its share of the population of die world's top twenty-five cities exceeded that of every other region in the relevant time period.

As expected, the active zone is shown to have been, in most cases, an area occupied by autonomous political systems, from the city states of Mesopotamia, to the interstate systems of the Middle East, the Eastern Chou, and India at the time of the Buddha. The Roman Empire splintered not long after its emperor embraced Christianity, and the Islamic Caliphates were noted for the flexibility of their political organization. The city states of classical Greece might appear as an omission from that list but it was in their shaping of the Hellenized civilization of the Mediterranean in which Christianity arose that Greek culture found its enduring place.

At about 1000 it appeared as though Sung China, then the most conspicuous country of East Asia, might take the lead in the world system, and for a while the most salient feature of that system appeared to be Chinese predominance. But the Sung first lost the North, and then fell, in the South, before the onslaught of the Mongols. It is the Mongols' design for a world empire at the center of the Eurasian landmass and the dominance of their cavalry armies that defined the age as one of 'Eurasian transition,' a transition that moved the wellspring of innovation of the system from East Asia toward Europe. The Mongols' project extended from China over most of the Eurasian landmass, but it never materialized completely, nor did that of Timur a century later, and imposing though both were, they soon collapsed under their own weight just as the last European attempts at medieval empire (those of Charles d'Anjou) crumbled before even getting properly underway. The republican regimes of Genoa, and Venice, were

[43]

the initial beneficiaries of that transition, took up the challenge and became the springboards from which the power of Atlantic Europe was projected upon the world after 1500.

The last column in Table 2.4 suggests the outlines of the evolution of the world economy (see also Modelski and Thompson 1996a, ch.8). The world economy process begins with bronze as a basic new technology and source of productive organization, with implications for tool-making, construction, and weaponry in particular. That technology encouraged urbanization, helped to raise levels of production, and increased trade. The age of bronze is rounded off with the broadening of what initially was a Sumerian center area to the entire Fertile Crescent becoming the basis of economic organization, extending from the Persian Gulf to the Mediterranean.

Between 1200 BC and 1000 BC in the Middle East and Europe, (though in China not until 600 BC), iron begins to assume general importance, gradually replacing bronze as the more cost-effective primary metal for tools and weapons. The 'closure of the Eurasian ecumene' in turn provides the opportunity for an expansion of long-distance trade, chiefly by land but also by sea, with the Silk Roads (including the maritime Spice Roads) assuming a key role in that process in the fourth period in particular.

The overall movement in the world economy over this long period has been from command to market economy. That is how the modern period is at first notable for the emergence of more flexible and productive market economies at the national and regional level, in the sense of an increasing differentiation between economy and polity, the consolidation of the economy as a potentially and substantively a self-organizing system, and the increasing role of autonomous, and increasingly corporate, business organizations. The groundwork for that development was laid over the past millennium, and more recently, say after 1850, conditions began to ripen for the emergence of world markets as the framework of exchange in a now significantly productive economy. But the world market still has a long way to go.

So far so good. World system evolution has now assumed a fuller-bodied complexion and presents a more rounded picture. The resolution of its analysis, represented by the shortest phase interval, is now reduced from 2,000 to about 250 years, but it does remain quite a wholesale view of the human experience, and especially so for the modern times, and the role of human agency in innovation remains essentially dim.

What is it precisely that we are trying to explain? We are trying to explain the finer texture of evolution, no longer as a sequence of major institutional changes, but this time in the grainier detail of organizational developments principally at the global level, and most importantly, by tracing the agents of innovation who propel these movements.

Let us define evolution at the global level as globalization. It consists of the

[44]

emergence of organizations that actually or potentially operate in planetary scope. Globalization thus comprises a set of processes: global economic evolution (of trading systems and world markets); global political evolution (of nation-state systems, world power competitions, and international organization); democratization (forming a potential democratic community); and the creation of a world public opinion (via media, and learning).

And who are the agents of globalization? They are individuals and organizations advancing and sponsoring innovation that results in strengthening the global layer of interactions. Among these we would name business firms that work transnationally, world financial markets, nation-states in positions of global leadership, non-governmental organizations of an humanitarian character or individuals who fund such projects, leaders of social movements, Nobel prize

Agency is most clearly apparent when we study processes that actually drive globalization: the rise and fall of leading industrial sectors, in the rise and decline of world powers, the course of the democratic lineage, and the long movements of world opinion. That is where agents of innovation can be seen at their most active.

Let us make more fine-grained the analysis for the modern era. Some treatments of this subject set the start of modernity, or of the modern world system, at about 1500, give or take some decades, and it is admittedly clear that a 'birth' of sorts did indeed occur at that time. But students of evolutionary processes would tend to go beyond mere birth, to the actual sources of such an event and ask: when and in what conditions did 'inception, (or conception?) take place? Our model suggests that it happened earlier, as early as maybe 930 (if we focus on the Chinese context), and accordingly we shall assign the onset of global processes, and of globalization, to that date.

Second, our model takes account of the increased organizational complexity of modernity because its main institutional emphasis is one of 'collective (species-wide) organization.' We know that the premodern world was, organizationally-speaking, fairly simple, basically a two-tier arrangement, one that combined the world of the 'great tradition' based on imperial courts, cities and temples, with a multitude of little traditions of the village peasantry (Modelski 1987:24—6). We hypothesize that the modern era produced, along the vertical dimension, a division of this two-tier set-up into a potentially four-fold structure ideally comprised of local, national, regional, and global layers of organization. The inception of that process of vertical differentiation coincides with the modern era, and its unfolding has produced, and continues to produce, two processes of major consequence: the rise of nation-states and the nation-state system, and the formation of organizations of global scope. We focus our attention on the second, the global system process.

Globalization, or global system process, might be decomposed (or refracted)

[45]

into four nested sequences of cultural, social, political, and economic elements, each subject, just as the institutional movement was, to self-similarity, and bringing forth new organizational structures. These processes are labeled in Table 2.5.

Globalization might again be defined as the formation of a planetary organizational framework and might best be viewed as a spectrum of four processes. We observe that the period of each of these structure-building processes at the global level equals one phase (that is one-quarter) of the overarching world system periods previously discussed, on the ground that its scope is that much more limited. Once again, we postulate that the four processes stand in the 8:4:2:1 relationship we showed for the larger process. In the paragraphs that follow we briefly review this model, and then support it with some data.

'Global system process' might be expected to program the global system, much in the same way that we have observed the world system process to do earlier. Its working depends on the more detailed specification of the world system priority of 'collective (species-wide) organization,' and its phases represent the steps by which such organization framework might be thought of as emerging at the global level, via Preconditions, Global Nucleus, Global Organization, Consolidation.

'Democratic community process' traces the evolution of community at the global level, and embodies the premise that such community can only be built upon democratic foundations. It docs so by tracing the antecedents, and then the members of the 'democratic lineage,' defined as the 'line or succession of societies that have shaped world democratization' (Modelski 1999d:154). Two periods of that process will be distinguished: that of 'experiments,' and of 'democracy.' The terms are inspired by Robert Axelrod's (1984) analysis of the evolution of cooperation, previously mentioned, that shows that even in a 'nastv' world, mutant cooperative forms might arise in an experimental exploration of the potential for cooperative (that is, higher-yielding) undertakings. We propose that reform movements in Sung China (ca. 1100), and republican experiments in the city states of northern Italy (ca. 1300) constituted two sets of such trials, but that it was the religious turmoil of the European Reformation, centering upon the Dutch Republic that laid the foundations for the cumulative growth of a nucleus of a global system, embodied in particular in the 'liberal-maritime alliance,' with England. That is how some of these experiments succeeded in conditions of clustering, and we propose further that conditions particularly favorable to such clustering have prevailed since the mid-nineteenth century, laying the groundwork

Table 2.5 Global system processes

[46]

for a future democratic community, in a process first anticipated by Immanuel Kant, and then by Alexis de Tocqueville who in 1835 postulated the ultimate success of democratization on the basis of his American experience. Why not call the successive periods of democratic community formation D-waves (for Democracy)?

'Global Political Evolution' has a period of some 500 years that corresponds to one phase of the active zone process. Each such period therefore tends to center upon one region of the world system, and we name these periods provisionally after the periods of the active zone process. The first is 'Eurasian transition,' followed by 'Atlantic-Europe,' and 'Atlantic-Pacific,' the names tending to suggest the shift in the geopolitical center of the global system. The phases of each period are activated by successive instances of the rise and decline of world powers, in their learning cycles, this being the mechanism whereby some nation-states have been selected to the role of leadership at the global level, and have been able by that means to shape global organization.

'Global Economic Evolution' describes structural change in worldwide commercial and industrial arrangements, and has a period of some 225 to 250 years. Within each of these periods nest four K-waves. These changes reflect movements in the world economy process previously discussed, from 'National' to 'World' market, and spell out the finer structure of that transformation, as a sequence that leads in the now familiar path from Sung China, through Italy, to an oceanic trade system that generates an 'industrial take-off.' In other words, the famed industrial revolution of the eighteenth and nineteenth centuries is but one, albeit dramatic, incident along a long path of growth for the global economy, that is now being followed by the 'Information Economy.'

Table 2.6 presents a model of global evolutionary processes derived from the considerations just discussed. It also adduces some data that lend it greater validity. We note that the resolution of these processes is now down to 50-60 years, the average length of a K-wave, in the last column. (For reasons of clarity of presentation, only the last three K-waves are shown in the last column; for a complete list see Modeiski and Thompson 1996a.)⁸ But we do need to emphasize that the K-wave, the surge of a leading industrial sector (currently the information industries) and its effect upon the global economy is another instance where agency conspicuously enters the evolutionary process.

There is no reason to suppose that the breakthrough to a global system could have occurred at only one place or in one time interval. Table 2.6 presents a concept of an 'Eurasian inception' of the global system, namely that such a breakthrough occurred, in Eurasia, at about 1000, both in China, and soon afterwards in Mediterranean Europe. The developments in China looked, for a while, more substantial and promising, and it was there that the technologies first developed that animated that breakthrough. There was a surge of sea power and a notable expansion of water-borne trade. Society moved in a 'bourgeois' direction.

[47]

Table 2.6 Processes of globalization (930-2080)

[48]

But Sung China faltered and failed to carry through at the global level. The Sung saw themselves as the bearers of China's classical tradition (recently reinforced through the printing of Confucian classics), but they had to contend ' with aggressive Sinicized dynasties of Inner Asian origin that came to rule much of North China. In 1004 they reached a stand-off with the Khitans (Chinese dynasty name of Liao). But the Liao dynasty was destroyed with Sung help, in 1125, by the Jurchen who took the dynastic name of Chin (Jin). By 1224 Chin rule in turn succumbed to the Mongols under Genghiz Khan who then took on the southern Sung and, under Kublai Khan, completed the conquest of China in 1279 (dynastic name of Yuan) whereby the Mongol world empire reached its greatest extension. The Sung paid much attention to seafaring and maritime affairs, especially in their southern phase, but the nomadic threat from the North never allowed full energy to be devoted to activities of truly global scope.

Structural politics in China 930-1420 could be presented as a sequence of four dynastic cycles, each of about 100 to 120 years in length, and each composed of the four phases of a political learning process (see Table 2.7). China's structural evolution was timed by four macrodecisions: the war with Liao, the war with Chin that forced a shift of the capital from Kaifeng to Hangchow and launched the Southern Sung; the Sung's conquest by the Mongols, who established themselves at Peking; and the Ming rebellion that expelled the Mongols and coincided with a system-wide collapse of Mongol power, also under the attack of Timur (1405). Ming rule was first founded at Nanking, with a potentially maritime orientation, but despite the great expeditions to the Indian Ocean, China soon turned inward, and the move to Peking in 1421 put a seal on that fateful shift.

The rule of the Mongols extender not only to China, but also through major parts of the Eurasian mass, in an imposing structure that embodied a bid for world empire. But that bid soon crashed, in just the way similar imperial bids, though on a more vast scale, also collapsed in the Far West at the same time. The same model we just used to oudine the evolution of East Asia could also be employed, with an identical time grid, to depict the salient developments in Mediterranean Europe, whose original center was then Byzantium. The first three cycles had the same imperial bent as those of the Chinese system: an attempt at Byzantine recovery that collapsed after the disaster at Manzikert (1054); the Holy Roman Empire and the Hohenstauffen bid to rule the Mediterranean; and the French (Anjou) and Papal try for universal monarchy, launched at the peak of Mongol power. All three were carried through to a macrodecision. But they all failed to make good on execution.

Running parallel to these imperial gestures was a series of bids for commercial supremacy by prominent Italian city states (Braudel 1984:106-11), beginning with Amalfi, a dependency of Byzantium, followed by Pisa, in league with the Holy Roman Emperors, and then Genoa, working with the Popes and the French and indirectly coordinate, via Black Sea trade and alliances with the Mongol Empire. Each city was in turn routed by its successor, synchronous with the macrodecision of the imperial bids: Pisa sacked Amalfi; Genoa crushed Pisa. And

[49]

Venice defeated Genoa after much bitter fighting and for a time profited on its own from die victories of Timur.

But it was only Venice mat stood on its own, dominated its golden quattrocento, and served as the regional prototype of a global power and a bridge to the next phase of global politics. These are the circumstances that inform the earlier parts of the Matrix of Evolutionary World Politics shown in Table 2.7.

In global political evolution, depicted in Table 2.7 as a series of long cycles (LC1-11) the Chinese antecedents, and the Italian prototypes, were inchoate, proto-global political sequences, laying at the regional level the foundations of future global enterprises. With Portugal, we move beyond laying the foundations, to building, in Atlantic Europe, the nucleus of die global system, and then adding to it with each successive cycle, with the successive world powers being die principal agents of leadership that they did or might not exercise. We note, too, that me four-phase evolutionary structure, observable both at the level of organizational structure (A,B,C), and at the agency level of long cycles (LC), remains the same but in contrast to its fumbling beginnings is, in the later cycles, more steady and, from the Dutch case onward, noticeably cumulative. In die 'Atlantic-Pacific' period we see the beginnings of international organization of a complex kind.

That makes it clear that die long cycle which depicts the sequence of world powers, is in fact a mechanism of global political evolution, wherein powerful nation-states, pursuing their own goals and interests, also forwarded the tasks of political construction. What Table 2.7 also shows is that each world power passed through its own learning and selection process as it was reaching a position of global leadership. What is more, the leaders of the world powers, interacting with the challengers they confronted, were the principal agents of international change. We note that the resolution of the analysis is down to 25-30 years (one phase of the long cycle, such as coalition-building - consult Table 2.7), that is to what is usually considered to be die generational period (the time it takes a generation to replace itself). This period of one generation is the basic time unit of this analysis.

The key elements of the political process are now well understood, and explored. In particular die way in which global leadership was presented, in their own words, by those who exercised it, and how leadership was transferred among them (Modelski and Modelski 1988), and how sea power distributions can be measured over five centuries to determine empirically who wielded global power (Modelski and Thompson 1988). Its basis in world economic evolution is explored in Modelski and Thompson 1996a. The recognition of the special role of the United States, Britain, and die Netherlands, in global arrangements, is now quite widely shared (Thompson 1988); the present study (and others) extend the reach of that process to die onset of modernity.

We have now shown world system evolution to be an array of major and lesser processes, all of which though do 'hang together.' But we also know that the world

[50-51]

Table 2.7 Matrix of modern evolutionary politics

Note: LC long cycle of global politics (numbered)

[52]

system is not a free-standing or autonomously self-propagating process. It is likely to be, in a nesting fashion, and extending the postulate of self-similarity, part of some larger-scale arrangement of the learning variety. That larger process, too, is likely to be of the logical form of the evolutionary processes we have just studied. What might that be?

These are complex but intriguing questions, possibly insoluble in the light of present knowledge. Is there a larger process of which world system evolution (with a period of 8,000 years) is a part? How might world system evolution place in that larger scheme of the evolution of modern humans? How do we explain the origins, or the inception of the world system, and what is the larger framework in which that system must be situated?

'Good questions.' Could it be that, at first, human evolution wrought the differentiation of humanity from nature? Could that prehistorical period, extending for over 30,000 years, be thought of as having comprised four major phases of some 8,000 years each? It might have begun with the adoption of a fully efficient human language, that included the use of syntax. Such a cultural breakthrough, we might suppose, would in turn promote, in the next phase, kinship linkages and extended family groups. Then, at the peak of the last ice age, humans might have been driven into caves and other permanent settlements. Finally, some 10,000 years ago, agriculture might have begun, slowly at first in today's Middle East, and then elsewhere. This celebrated 'Agricultural Revolution,' and the networks of exchange that it might have promoted, would lay down the preconditions for the next step in human evolution, the rise of civilization on the basis of sociality.

What is not in doubt is that, the centuries between 4000 and 3000 BC mark a strategic turning point, towards sociality, and the onset of world system evolution, and of all that we have reviewed as making up our major social institutions. They also signaled the beginnings of civilization, a deeper change that extended not only to society but also to culture, human nature, and to the humans' relation to their environment. We might define (singular) civilization (a word whose root is the same as that for city) as a condition of humanity characterized by a culture of urbane living, contrasting with barbarism in a condition of nature. The city, made possible by agricultural surpluses, is itself a large-scale cultural artifact and makes possible the differentiation of culture from society, and its more autonomous development. Human evolution entered civilization (and history proper) with cities and writing, and that is the stage that we are still living through, and are likely to continue well into the future.

This has been an exposition of the logical structure and some indication of the range of supporting evidence that suggest the initial plausibility of an evolutionary world system analysis. Species-wide evolutionary processes have been identified at three major levels of analysis: institutional, organizational, and agency;

and a four-phased learning process could then be observed. Surprisingly, the logic is the same at each of the levels, while the period of the learning is in each case

[53]

proportional to the scope of the process, in a manner that preserves symmetry, nestedness, and synchronization in coevolution. Even more encouraging is the thought that important processes including the rise and decline of world powers, as well as that of the rise and decline of leading sectors, might in each case be seen to be incidents in an unfolding panorama of world system history.

1 This is a reworking of the paper 'World System Evolution: A Learning Model' first presented at the thirty-second annual convention of the International Studies Association, Vancouver BC, March 1991.

2 But Lumsden and Wilson (ibid.:3-7) also argue that the human species is unique in the magnitude of its enculturation process. 'Mankind has attained' the 'complete' or 'true' cultural state because its repertoire includes not only simple learning and imitation, but also complex learning: teaching linked to socialization of the young, and the employment of symbols by human agents. This creates the potential for the creation of more advanced forms of organization via species-wide learning processes. But the focus of the present analysis is social organization and social structure, not culture as a complex of meanings and the codes that govern it, changes in the make-up of personalities, or the adaptations in the biological organism that mediate between its physical existence and human action.

3 For full comparative treatment of the principal approaches see contribution by William Thompson to this volume.

4 Another, recent, example is Samuel Huntington's The Clash of Civilizations (1996).

5 Jane Azevedo's Mapping Reality: Evolutionary Realist Epistemologv for the Natural and Social Sciences (1997) is an excellent recent exposition of the current state of that field.

6 They could also be seen as phases of a learning process that Parsons and his associates described in 1950 as the LIGA sequence (Modelski 1987:104ff).

7 Along the vertical dimension, we distinguish global from regional, national, and local structures.

8 The outline of global economic evolution comes close to what Fernand Braudel (1984:

76-88) identified as the secular trends in the European (world) economy, each linked to four K-waves. We place less emphasis than Braudel on the ups and downs of the economy, and on its price trends, than on ongoing transformations in economic structure at the global level via lead sectors.