February 16, 2004

 

 

 

 

 

 WORLD CITIES, EMPIRES, AND WORLD SYSTEM EVOLUTION

 

 

 

George Modelski

 

 

 

 

 

 

 

                        The world system (that is, the species-wide social organization, and institutions, of humans) was not built in a day, not even in the rapidly globalizing last decades of the 20th century.   Nor, for that matter, has it now attained its “final’ form – whatever that might be.   It is the product of a long process whose past and future shape needs to be better understood.

 

If we stipulate for the purpose of this paper that the widely accepted division of world history into three eras: ancient, classical,

and modern – also represents distinct phases in the trajectory of the world system since about -3000.  then the question arises about the characterization of these phases: are they in fact be phases of an evolutionary learning process of the world system [1]?

 

In this paper we test these questions against systematic data on major urbanization over the long period of five millennia brought

together in the recently issued World Cities:-3000 to 2000  (Modelski 2003a) [2].     The results of such testing point to an affirmative answer to our question.   Yes, the world system at its most general level does indeed give evidence of passing through three major and qualitatively distinct phases, each one of which maximizes one particular set of critical social innovations.   Of course, that is not the entire story of world system evolution – that also comprises, in a spectacular cascade, essential social, political, and economic processes,  but it does provide an outline of the envelope within which other elements of that cascade that make up the story of the making of the social organization of mankind find their place.

 

In the main, then, this is an attempt to answer a qualitative question about the character of world system development on the basis of

a quantitative survey of world cities.   But we shall also give attention to a related question about the role of empires as focus of world organization, and our finding suggest that their importance while notable, might have been exaggerated; they do not shape or dominate any of the three phases.      . 

           

World system evolution

 

                        The discussion of ‘phasing’ with which we began this paper should not obscure the commonsense observation that the trajectory of humankind is a manifest case of evolution.  To-day’s world is obviously much different from what it was one thousand years ago, or what we can imagine it to have been five thousand years ago.    What is more, any even cursory look reveals a picture that can make sense only if we approach it via a form of evolutionary theory.   For what body of knowledge if not evolution can explain humanity moving, in a space of five thousand years, quantitatively from less than ten million to a size now approaching  tem billion members, and qualitatively, from a condition of small dispersed communities to one of a constellation of great metropoles?  And so, what does ‘phased’ urbanization reveal about world system evolution?

 

World system evolution, as all of the ‘cascade”, is a macro-learning process that is Darwinian in character [1].  The term Darwinian in this context is not intended to suggest a biological (e.g. genetic) explanation.  It is designed as an instance of “universal Darwinism”, the argument (advanced i.a. by Richard Dawkins and Henry Plotkin) that evolution everywhere, including in this particular case, changes in the social organization of the human species, occurs in accordance with certain generalized Darwinian principles.   Plotkin (1997:84) uses the term “heuristic” – a pattern that leads to discovery and invention – for the sequence of consecutive and continuous phases in the overall evolutionary learning process [3]. 

 

At this point we stipulate that the trajectory of the world system shows evidence of major, 2000-year long, phases.   The evidence for this will not be reviewed at this point, and the reader is referred to evidence on the shape of the growth of world population, and also on the long trend in major urbanization (taken as a proxy for world urbanization,  Modelski 2003a,b).

 

At this stage of our study we intend to test the ‘qualitative’ proposition that world system evolution is a sequence of four such macro-phases : generation of variants (the nature of which remains unspecified), cooperative mobilization (or agglutination),  selection of variants (by a mechanism to be specified), and regeneration and consolidation.    We propose that the phases that we have recognized throughout, the three eras of urbanization:  ancient, classical and modern, are in fact the first three of the four phases of the macro-learning process that we call world system evolution.  Let us spell out these points in some more detail. 

 

 

A test of the learning thesis

 

                        Viewed synchronically, all parts of the world system, the social organization of the human species, constitute a structure.   That structure results from the fact that at a certain point in world time, a point that coincided with the onset of urbanization, humans began to relate to each in other ways that raised the possibility, and increasingly the actuality, of species-wide organization.   Hence, viewed diachronically, across world time, world system might be viewed as a process, in fact a learning process that we also call world system evolution.   World system did not emerge, fully-fledged, and all at once but rather as a process extending over long periods.   That process continues and projects well beyond the future that is foreseeable by to-day’s social science methodologies.

 

                        By speaking about this “world system process” we imply that we adhere to a “unitary” rather than plural, or “multi-civilizational” conception of world system evolution.   We do not scan the world for multiple world systems or civilizations each with their own developmental trajectory, or review the historical record in search of convincing ways of slicing it into several singular units of analysis.   Such a  unitary conception of the world system may be contrasted with a multi-civilizational approach originally propagated by Oswald Spengler and Arnold Toynbee, and that more recently came to be associated with the concept of a “clash of civilizations”.    

 

                   In the present approach, we take human species as one, and ask:  what has been its social evolutionary trajectory?   That is, by what path have we reached to-day’s social organization, and where might such a path lead us in the future?

 

                        To answer that question,  conceive of social organization on a world scale as a millennial learning project:   the human species learning to live with themselves, and with increasing numbers of their own kind, by evolving new forms of organization.   Such a project, like all (social evolutionary) learning processes, might be expected to proceed in long stages, that we take to be some 2,000 years in duration and about  equal to historical eras, each era lending distinct emphasis to a different evolutionary problem [4].

 

                        We propose that world system evolution can be analyzed, over the past 5,000 years, and well into the future, as a sequence of four phases of a major social learning process.   Its first, cultural, phase  might be understood as the laying down of the learning infrastructure of the entire system, creating the conditions for what we can call the emergence of civilization, conditions in the absence of which there could be no extended (potentially world-wide) social learning, and in the absence of which variety could not emerge on a sustained basis.  

 

                        The second, ‘social’, phase, that can also be described as cooperative and community-building, highlights the learning of more extensive yet also more inclusive, - and potentially universal - solidarities, on the basis of which new forms of cooperation could take root and without which social organization on a world scale could not arise.  

 

The third, political, phase that is our own, modern, experience,

revolves around the central Darwinian principle of selection,  that is on the timing and character of the (explicit or implicit) choice of evolutionarily stable conditions of world organization.   In the case of world system it concerns changes in the scope and nature of collective organization at the global level, the control of the means of violence and instruments of mass destruction, and the choice of conditions of world order.  

 

The fourth, yet to come, could be envisaged as ‘economic’, and residing

 in the stabilization and consolidation of the economic and material basis of that process,  and possible regeneration of its products.   But what about the economic factors in social action so far?   They cannot be ignored and are hardly forgotten.   The “urban revolution” first arose against the background of, and as the successor to, the invention and successful dissemination of agriculture in the previous millennia, and in particular in West Asia.   That is, cities rose up with the help of food surpluses generated by the “agricultural revolution”, and in an area where successful irrigation works greatly increased the availability, predictability, and distribution of such surpluses.  The even more immediate context was the network of trade links that antedated even cities and had already enmeshed much of that area.  All this suggests that the fourth phase could be laying the material foundations for a new period of world system evolution to take the human species onto yet undreamed of paths, to other planets or maybe even into other parts of the galaxy.  

 

It is obvious that this is an attempt to lend conceptual coherence to the

 large sweep of human experience, covering the ancient, the classical, the modern,(and the post-modern?) eras, and that is an important and possibly urgent task.  Now comes the question that we shall ask of our data:   do our data on world cities support the expectation  that the three phases of the world system show qualitative differences of emphasis that support the notion of a learning process at work, that each of the eras that we have reviewed and documented stands for a distinct civilizational experience that justifies it being recognized as one of the predicted phases of world system evolution?  

 

 

Major urbanization

 

Urbanization is how the world population comes to live in urban

 places.  Quite evidently, it is social change on a major scale.   Our own survey of ‘world cities” suggests strongly that major urbanization, just as world population growth, is a phased process.   On this occasion we take the “fact” of phasing as a given, having previously argued for it at some length (Modelski 2003a,b, Devezas and Modelski 2003).

 

                        The growth patterns of world citiesare, of course, just one portion of urbanization. The inventory of world cities that forms the data base for this test gives us information about what might be called “major urbanization”, that is about the formation and the life cycle of the world’s largest cities.   This is, as it were, the tip of the iceberg of urbanization..   We call it “major” – or maybe “high-end” – urbanization  because the cities we have observed are the greatest and the most prominent of their eras.   It is also “major” because it highlights those urban places that are best positioned to be observed, and also to serve as links in long-range and world-wide networks of every kind.  

 

The question now is: what does our record of “major urbanization” –

viewed as an index of social transformation -   tell us about world system evolution, and in particular about the make-up of the three major eras of world history (or phases of the world system)?   For each of the three eras  it constructs an inventory of “world cities”.[5]   The primary and sole relevant qualification for the designation of a town or settlement as “world” city is the presumed (estimated) size of its population in a given year (at one century intervals).   In other words, in this instance world cities are the most populous urban sites of their time.   That is, deliberately a simplification.   In recent years a considerable literature has arisen around the concept of “world city” as the focus of globalization at the end of the 20th century (see e.g. King and Taylor 1995).   That literature    That literature takes for granted population figures (the number of “millionaire”

cities is currently in the 300+range, and attention has shifted to mega-cities of 10-20 million +) and seeks to distinguish among them in particular those that host major clusters of economic activity, such as headquarters etc. of transnational corporations.

They can afford to take population figures for granted because they are now freely available, in contrast with the situation for most of the five millennia past.

 

                        As a criterion that must work across such a long time-span, the number of inhabitants seems the most useful.   More refined approaches will no doubt find that not all of our listings merit the designation of “world city”, and that some smaller urbanisms that fall below our thresholds will be found will be found that also deserve that title.  For ancient world cities (-3500 to -1000), our threshold criterion is an estimated population of 10,000 or more;  in the classical world  (-1000 to 1000) it is 100,000, and in modernity (since 1000)one million.   The surprise is that these simple criteria yield, for each era, a comparable number of entries:  in the first two cases, a maximum of 26, in the  first, and 27 in the second.   In the modern era we reach 16 cases by 1900, but then urbanization explodes, and we find some 300 cases in our inventory.   There are grounds for  believing that this is an instance of a hidden order in the world of cities (as in the world system).    But let us now turn to our main question:  what do our data reveal about the qualitative character of world system phases?

 

                                 

Cities, writing, and calendars in the ancient world

 

                        Was the ancient era the first critical phase of the world system process, a distinct learning experience of an infrastructural character?   What are the elements of a civilizational learning-infrastructure?

 

The distinguishing feature of that era resided in the launching of two

major social innovations, cities, and writing and calendars,   and these might jointly be interpreted as having laid down the learning-infra-structural basis for world system evolution.   Because we had cities and writing, the human species could undertake the task of building the elements of social organization on a world scale.   Without cities and writing the humans could not have continued with the project of the world system. .  

                       

Why do we regard cities, and writing and calendars, as the central

 elements of this stage of the civilizational process?    A system of cities provided as it were the hardware making possible collective learning on a continuous basis;   it was the platform for sustained and intense interaction with the expectation of repeat experience, independent of the fate of any one city.  The first cities arose around cult centers;   city-states were the political basis of that system, and the nodes of networks of economic cooperation.   World cities, and a fortiori, a system of cities (and not just one big cosmopolis of the classical thinkers)  became the basis for the sustained generation of variety from which sprang innovation.  Cities, too,  are the cradles of  writing (the organization and storage of information) and of calendars (the organization of time) which are the software of stable social interaction and learning.  They made possible collective memory (we know that lists and records are the first forms of writing), and they enabled communication across space and time, thus programming complex human enterprises.

 

                        When and where do world cities, and writing and calendars, appear on the world scene?    Jean-Jacques Glassner, among others, has situated the “origin of writing” in Sumer  in the century between -3400 and -3300.   The data assembled in World Cities:-3000 to 2000 demonstrate that a system of cities, the nucleus of the world city system, emerged in Sumer after about 3500.    That system was centered on Uruk as its largest unit.              We also know that the oldest extant examples of writing concern Uruk, and also have been found in that city.   The reason for that might well be that writing was invented there (though in a context that included wide-ranging contacts).   In one ancient Sumerian text entitled “Enmerkar and the Lord of Aratta”, the invention of writing is indeed affirmed to be a human (rather than a divine) creation, attributed to the Lord of Uruk-Kulaba, also described as a founder of the city.   The epics of Sumerian literature that took off about -2500 concentrate on the legendary rulers of that city, and in particular on Gilgamesh.  

 

                   The earliest, and the only substantial cuneiform finds of the earliest of the Uruk IV period (-3400 to 3100),  consisting of 600 inscribed tablets, are recorded for that city  (as compared to one for Kish).   That means that we have good reasons for believing that the origin of writing coincided quite closely with the origin of the world city system, probably in a synergistic fashion.   The cities of Sumer grew around ceremonial centers focussed on temples, and temples also soon became the centers of writing, store houses of knowledge, and the locus of teaching.   The principal such center was Nippur.    Evidence for specialized political, or economic activities comes later.   An incipient system of cities also appeared at about this time in (pre-unification) Upper Egypt, about Abydos, Nekhen, and Hierakonpolis;  but on a smaller scale, and here too incipient writing has now been observed.  Over time, cities and writing do appear also in Iran, in the Levant, and in the Indus Valley, in roughly similar circumstances.   At the close of the ancient era, these have dispersed throughout a wide area, and Sumer is no longer the center of the system.   By 1200, a new, alphabetic, form of writing is documented for the Levant coast.  

 

                        Parallel to the invention of writing has been the consolidation of concepts used for the management of time.   The earliest tablets (Uruk IV) contain signs for such measurement: days, months, and years.   That would seem to confirm that the lunar-solar calendar was probably invented in Sumer at about that time because  the seven-day week may have evolved out of moon phases.   The Egyptians (especially at Heliopolis) excelled in astronomical skills, and computed the length of the year at 365 ¼ days and it is their calendar that was the foundation of the Julian system introduced by Julius Caesar  on the advice of an Alexandrian astronomer (-46), and that in adapted (Gregorian) form is still with us.   in China, the lunar-solar calendar shows up in the oracle bones of the Shang period and in India, in texts at about -1000.

 

Our data on world cities therefore support the conjecture about the

(qualitative) character of the ancient era  namely that its distinguishing feature at the most general level resided in the creation of a learning infrastructure, including such institutions as schools, chancelleries, and archives, and of course the cities they were located in, as essential for the continued evolution of the world system.   That would lead us to expect that subsequent eras  will also likely exhibit their own distinctive features.

 

 

Religions in the classical era

 

                        Given the character of our data, and the focus of our present interest,  the next concern is with the prediction derived from the framework of “world system evolution”  that the characteristic features of the classical era (the second phase of the world system process, covering -1000 to 1000) center on the wider solidarities that are the products of religious evolution, in other words, on the world of religions. 

 

                        The observation that the classical era was associated with the development of major cultural and religious traditions is hardly news to historians who routinely comment on the rise of i.a. Buddhism, Christianity, or Islam, in this time frame.   Less frequent are the attempts to put these developments in a wider framework.   One prominent exploration on the canvas of world history is Karl Jasper’s concept, in  The Origins and Goal of History (1953) of the “axial period”  .   That is the period, from 800 to 200 BC, that on his view, constituted a major turning point in human affairs, and laid down the foundations from which all contemporary civilization derives.   Because that is when, he argues, simultaneously and across Eurasia, the beginnings of the world religions, by which human beings still live, were created.    That is the time when Confucius lived in China, Buddha was active in India;  Zoroaster taught in Persia,  the prophets preached in Israel, and the philosophers argued in Greece.  For Jaspers, the close simultaneity of these developments demonstrates the spiritual unity of mankind, but represents no “general law” but only “a specific historical, unique fact”.

 

                        We might wonder how simultaneous a process can be that extends over a period of 600 years; we might also point out that Christianity and Islam appear on the world scene, on Jaspers’ account,  as latecomers, well past the “axial period”.    But our own analysis does concur with his in recognizing the importance of religions for the classical era;  it differs from it, though,  in some important respects:   in its analysis that puts the emphasis less on ideas and more on social organization, one that is sequential rather than stressing simultaneity,  and  one that sees it in terms of the general processes of world system evolution, as previously described, and not as the unique product of a specific historical situation.   Our urban series could therefore help us to answer the two following questions:   do religions in the classical era show sequential development as part of world system evolution, and when and how did they take root at the center of the world system?

 

                        We propose that the classical process might have followed in Eurasia the following sequence:  first, Confucianism and Taoism; second, Buddhism; third, Christianity, and fourth, Islam.   This could be conjectured to be a learning sequence: Confucianism prioritizing education; an ideology of scholar-officials “learning to be human”;   Buddhism as the way of the monastic community of those “freed from fetters’ of traditional society;  Christianity notable for strong religious organization;  and Islam as the wide-flung and proselytizing community of the faithful.

 

                        Table 1:  “Classical World Cities:  Buddhist, Christian, Moslem, Others”

helps to test such propositions.   It classifies all those we designate as “world cities” according to whether they might be regarded as influenced (or just colored) by one of these three religions communities.   In other words, whether, in their general orientation, they might be described, at specific points in time, as likely to be noticeably (though not exclusively, or completely) Buddhist, Christian, or Moslem.   Table 1 displays developments from 200 BC onward (that is, past Jaspers’ axial period) because prior to that date such communities have demonstrably not yet registered an impact on cities in general.

 

                        Confucianism comes first in this sequence, and it does not really make an appearance in Table 1 because it emerges as an official ideology first in the Han Empire, after -2000, with Changan or Luoyang as possibly its cities.   It does become the public face of that empire, and of succeeding systems of Chinese governance, but has little transnational appeal as a more general system.

 

                        The first to show up well in Table 1 is Buddhism, with the city of Pataliputra at -200, seat of the 3rd Buddhist Council  convoked by Ashoka c.-250.   Over time, the several tendencies of Buddhism rose gradually in their salience among the world cities, and in time, too, also reaching out across the major regions of Asia, generally following the Silk Road.  Overland, the Central Asian route took Mahayana Buddhism to China, with one of the earliest  foundations opening near Luoyang, then the Han capital,  as early as 68, and assuming major importance in the Tang era and by 800 notable also in Korea and Japan;  the other major movement occurred via the maritime routes, first to Ceylon, and then to Burma and several other parts of Southeast Asia, also reaching South China.   Cities prominent in transnational and/or commercial linkages such as Nanjing, Guangzhou, or Kyoto were particularly likely to assume a Buddhist coloration.   The origin and the initial impact of Buddhism was in India but it is its subsequent and far from secondary influence, as facilitating communications across Asia that arouses particular interest, and shows up well in our table.

 

The third phase is that of Christianity, appearing initially in the form of

small communities  in Antioch, Alexandria, and Rome  but after 330 becoming the official cult of the Roman empire  centered on Constantinople,  and taking hold of up to one half of the then world cities.   From about 330 to 632 the Mediterranean zone of the world system was saliently Christian, even though it was also rent by local divisions, heresies and persistent schisms.

 

The fourth phase is Islamic and Table 1 shows how rapidly

the Christian cities of eastern Rome were taken over, and how quickly, after 700, the center of the world system thus turned into that direction.   The Moslem world at that moment was becoming strongly urban, and trade oriented, and by 1000 more than four out ten (10/25)  of the world cities were situated there.   Constantinople was the only substantial urban center holding on in non-Moslem Europe, with Buddhist-influenced areas in the East completing the picture.                 

 

Table 1 therefore suggests that the rise and diffusion of certain religions in

 the classical era was more nearly a sequential process, rather than a simultaneous experience.  The process was one characteristic of the world system, and not just of its component parts.   The table shows most of the classical era as significantly molded by that preeminent tendency.

 

 

Table 1:   Classical World Cities:   Buddhist, Christian, Moslem, Others,

-200 to 1000

 

  -200

1

200

400

600

800

1000

Buddhist

 

 

 

 

 

 

PATALIPUTRA

VAISHALI

VAISHALI

TAXILA

MADURAI

KANAUJI

KANAUJI

UJJAIN

PATALIPUTRA

PATALIPUTRA

LUOYANG

KANAUJI

ANURADHP.

THANJAVUR

TAXILA

TAXILA

TAXILA

CHANGAN

ANURADHP.

SRI KSETRA

CHENGDU

VAISHALI

PURUSAPURA.

PURASAPURA

NANJING

SRI KSETRA

LUOYANG

KAIFENG

 

PRATHISTARA

PRATHISTARA

YE

ISANAPURA

CHANGAN

KYOTO

 

ANURADHP.

ANURADHP.

DATONG

LUOYANG

GUANGZHOU

NARA

 

 

LUOYANG

 

CHANGAN

CHENGDU

ANGKOR

 

 

 

 

WUCHANG

KAIFENG

DALI

 

 

 

 

 

KYOTO

 

 

 

 

 

 

NARA

 

 

 

 

 

 

LHASA

 

 

 

 

 

 

KYONGJU

 

 

Christian

 

 

 

 

 

 

 

ALEXANDRIA

ALEXANDRIA

ALEXANDRIA

CONSTANT.

CONSTANT.

 

 

ANTIOCH

ANTIOCH

ANTIOCH

 

THESSAL.

 

 

EPHESUS

ROME

EPHESUS

 

ANI

 

 

PERGAMUM

CARTHAGE

JERUSALEM

 

VENICE

 

 

CAESAR.MAR..

CONSTANT.

CAESAR.MAR.

 

 

 

 

CAES.MAZ.

EMERITA

CONSTANT.

 

 

 

 

 

CAPUA

 

 

 

 

 

 

CAESAR.MAR.

 

 

 

 

 

 

 

Moslem

 

 

 

 

 

 

 

BAGHDAD

FUST/CAIRO

 

 

 

 

 

CORDOVA

TINNIS

 

 

 

 

 

BASRA

BAGHDAD

 

 

 

 

 

ALEXANDRIA

NISHAPUR

 

 

 

 

 

FUSTAT

ISPHAHAN

 

 

 

 

 

RAYY

CORDOVA

 

 

 

 

 

MERV

SEVILLE

 

 

 

 

 

DAMASCUS

RAYY

 

 

 

 

 

KUFA

SAMARKAND

 

 

 

 

 

MECCA

AL AHSA

Other  Cities

 

 

 

 

 

 

 

LUOYANG

LUOYANG, LU

LINZI

CTESIPHON

SUZHOU

SUZHOU

SUZHOU

 

LINZI, LU

LINZI

ROME

MERV

CTESIPHON

XINJANG

LINHUANG

 

XIANYANG

CHANGAN

CAPUA

 

.

YOUSHOU

ANHILWARA

 

CHANGAN

MAOLING

SMYRNA

 

 

 

 

 

CARTHAGE

ZHANGLING

CAESAR.IOL

TEOTIHUA.

TEOTIHUA

 

TOLLAN

 

ROME

YANGLING

CARTHAGE

CARACOL

CARACOL

CARACOL

 

 

ALEXANDRIA

CHENGDU

APAMEA

TICAL

 

 

 

 

ANTIOCH

WAN

 

 

 

TIHUANACO

 

 

EPHESUS

ANTIOCH

 

 

 

 

 

 

SELEUCIA/T

ROME

 

 

 

 

 

 

JERUSALEM

EPHESUS

 

 

 

 

 

 

PERGAMUM

PERGAMUM

 

 

 

 

 

 

 

JERUSALEM

 

 

 

 

 

 

 

APAMEA

 

 

 

 

 

 

 

CYBIRA

 

 

 

 

 

 

 

SELEUCIA/T

 

 

 

 

 

 

 

ALEXANDRIA

 

 

 

 

 

 

16

25

20

19

18

27

26

 

 

 

 

 

Note, too, that the spread, or diffusion, (as distinct from the origin and

 inception) of these religions is principally the product of regional developments of the second part of the classical era, after about -200.   The first part saw the rapid growth of populations, growth of cities, and expansion of cultivated areas, especially in East and South Asia.    But as this rapid growth, for several reasons, slowed down, religious diffusion assumed prominence.   It served to promote the equalization of intra- and extra-societal differentials, and to smooth long-range communications.     That process was also promoted by migrations, and hinterland incursions.

 

                        Note, in addition, that of the religious communities we have followed, each started in the competitive context of systems of independent city states:   the Spring and Autumn, and the Warring states in China,  the newly rising urbanized polities of the Ganges valley, the Hellenistic world of the Mediterranean, and the merchant cities of Arabia.   Only after the movements formative of these communities were first launched were their universalistic tendencies taken up by great imperial formations as the themes that might lend them greater legitimacy:   Confucianism by the Chinese centralized state;  Buddhism by the Mauryan kingdom and later in China and Japan,  Christianity by Rome, and Islam founding its own Umayyad, and then Abbasid caliphates.   But each of these empires soon faded out.   It was the religions communities that outlasted them as the glue that held together large portions of civilization, and shaped the regions wherein they flourished.

 

 

How important were empires in world system evolution?

 

                        Historians, political scientists, and sociologists pay much attention to the role of empires” as the most interesting, and as though ever-present, social structures of the pre-modern world system.    Immanuel Wallerstein regards empires as a “constant feature of the world scene for 5000 years”.   What does the inventory of world cities reveal about those quite obviously prominent structures, and why do we need to under-emphasize , in an evolutionary account, the role of empires?

                       

                        The importance of empires - those multi-level political organizations, seeking power monopoly mostly imposed by force upon diverse populations, yet also claiming universal authority at regional or world scales - is sometimes assessed by the extent of the territory they controlled.   That measure is of questionable validity unless qualified  by some index of value.   The world city data makes it possible, for the ancient and classical eras, to gauge the importance, or lack of it,  of those grand political constructions in another way, by determining whether they mattered in global proportions,  by showing what share of world urban resources (hence also skilled populations, trade and industry, and communication networks) they commanded.    We analyze those empires that offer lessons or experience in regional or world organization – not just large land powers.

 

                        The urban record reveals empires to have been not as important as historical memory would have it, if we measure that importance by the proportion of world cities controlled by them, at points of time captured in our record.  In the ancient world, the two prominent multi-level political organization with a claim to regional standing were those of Akkad (at -2200), and of the Third Dynasty of Ur (at -2100).    At these dates, each controlled a significant portion of the world cities, that is about one half of the respective totals (Akkad, 13/21;  Ur III 10/19), which means they both controlled most of Mesopotamia and some neighbouring areas, but not Egypt, Central Asia, or the Indus Valley.    Both were relatively short-lived:   Akkad lasted about one century and a half, and Ur III just about one century.   Neither was notably innovative because the great achievements of Sumer, cities and writing, antedated their activities and had little to do with them.   Their overall impact was less positive than negative,  despite bombastic claims from Akkad whose second king, Naram-Sin, fond of writing, famously claimed in his inscriptions, not just divinity but also rule “of the four quarters” of the earth, (probably the first recorded, but hardly last, claim to universal dominion)  and the rationalizing administrative practices of Ur III.   

 

Both the wars of Akkad, and the collapse of Ur III rule under

peripheral pressures paved the way  toward the disappearance,  at the hands of Babylon,   of the civilization of Sumer in the following 2-3 centuries,    But Hammurabi’s short-lived Amorite empire does not even show up on our screen;   it was substantially gone by -1700.   Nor does Egypt’s New Kingdom have much claim to universal significance, despite important inroads in Nubia and the Levant.   In East Asia, the rule of the Xia or the Shang may have laid out the basis for wider political system in North China, but we do not know enough about it (one scholar – Mark Edward Lewis - described these ancient Chinese empires as “an imaginary realm created” in the Han era “within texts”).  In short, in the ancient era,  empires (if that is the right word for the cases cited here) were, at best, costly and short-lived experiments in political construction above the scale of city-states.  Far from being continuously present, they were at best attention-getting incidents in a general picture of autonomous units.    It was the systems of independent city states that served as the loci of innovations that shaped this ancient world, and the emerging world system.

 

In the classical world, imperial regional organizations make a somewhat

stronger impression, but not a truly positive contribution.   An overview of the relevant data appears in Table 2, below; it gives a rough measure of the weight of classical empires at five data points  and demonstrates that the classical world showed minimal imperial structures in the early part of the era (at -1000 and -500), and maximum control at about mid-point, in year 1, with some three-quarters (18 out of 25) of  world cities forming part of an imperial structure, that share declining sharply over the remaining period, such that by 1000, only about 1/5th of our world cities can be shown to have had imperial color..  

                     

       Table 2:  How important were the classical empires?

                                 (as share of all world cities, by major region)  

Imperial / NON-IMPERIAL

world citiies IN

 

at

-1000

 -500

1

500

1000

East Asia

0         1

0        8

9        0

0        5

3       4 

South Asia

0

0        1

0        6

0        1

0       4

Mediterranean

0         2

1        0

7        1

6        1

2       2

West Asia & Moslem world

0         1

2        0

2        0

2       0

0      10

The Americas

         

 

 

0        3

0        1

Totals

0         4

3       13

18      7   

8       10

5      21

 

 

 

 

 

 

% distribution (rounded)

 

        100

20     80

75      25

30      70

20    80

.                      

Table 3:   presents the data in greater detail, naming names, quite a few of which the reader might recollect as forming part of our

 historical memory.   It also offers visual support to the trends depicted in Table 2.   The basic distinction is that between world cities (of 100,000 inhabitants+) that are part of empires:  “imperial cities”, and other such cities that we call: “non-imperial”.   The distinction is not always as neat as we should like it to be, and it also raises important definitional questions about what was (or is) an empire (a political structure with a monopoly of regional reach).    The important cases are clear:   the Persian empire in -500 (its rulers were ‘kings’); and Rome, and Han China, in the year 1, but even these had powerful opponents on their frontiers (i.a. Rome had to content with the Parthians, and then the Sasanids for over half a millennium) and were never entirely at peace.   Other cases are even less clear:  the Abbasids overthrew the Umayyads c.750, and seized control over the Caliphate, the political-religious institution of the Moslem world.   But the Caliphs in Baghdad soon came to be controlled by their Turkish professional soldiers (the Buyyids, and the Seljuks) and the Moslem domain split apart into a number of independent entities, with i.a. the Umayyads ruling in Spain, and the Fatimids controlling Egypt.   At 1000, the Uighurs were gnawing at the northeastern frontiers.   Was this still an empire?   The basic generalization remains that the “growth” period of the classical era was basically one animated until the 1st century BC by autonomous units in the Mediterranean, the Ganges valley, and in China, that were then brought into imperial structures that soon set on a path of decay such that by the end of the era empires were in retreat.       

 

Three out of the four major regions that we have reviewed show one

strong pattern:  vigorous urban growth in the first half of the era, followed by imperial rule of uncertain duration and stability and combined with cessation of growth in the second half [such cross-regional synchrony in  respect of changes in the size of empires and the largest cities was earlier demonstrated by Christopher Chase-Dunn).  In other words, empires seemed to have had little to do with the creative parts of the process; at best they may have had some effect of consolidation (with a much overrated Pax Romana), but also vulnerability to hinterland intrusions.   They never offered a sustainable model of regional cooperation, let alone of world organization.

  

The path to empire was somewhat different in West Asia, actually the

original seat of the urban revolution.   Three prominent conquest states made an early appearance there in succession:  the Neo-Assyrian, the Neo-Babylonian, and the Persian empires, with Babylon as their largest city.   As their names suggest, the first two were as if ghosts of the ancient world (Assyrian rulers harked back to Sargon of Akkad, and the Neo-Babylonians, to Nebuchadrezzar I, etc).   The king of Persia loomed large c.-500, holding sway from Egypt to the Indus valley, but his rule collapsed c.330 under Macedonian-Greek onslaught that remade the region.  One millennium later, the Arab empire of the Umayyads held sway for about a century, but the Moslem world soon devolved into a system of basically independent units.

 

The Americas of the classical era, and in particular the Mayan world, do not

seem to have had an imperial structure, but rather one of autonomous city-states.

 

 

Table 3:  Classical world cities:  imperial and non-imperial

                       (est.100,000 inhabitants and over)

 

 

Year / totals

Imperial world cities

Non-imperial world cities

-1000

 

                4

 

HAOQING   capital of Western Zhou

MEMPHIS 21st Dynasty city, Egypt

THEBES    21st Dynasty city, Egypt

BABYLON   

-500

 

               16

BABYLON   Persian empire

ECBATANA    Persian empire

MEMPHIS   Persian empire

LUOYANG   Eastern Zhou

LINZI   capital of Qi

XIATU   lesser capital of Yan

XINTIAN  capital of Jin

HANDAN  capital of Zhao

LU   capital of Lu

ANYI   capital of Wei

SHANGQIU   capital of Song

RAJAGRIHA   capital of Maghada

CARTHAGE

AGRAGAS,  SYRACUSE  

ROME   Republic

1

 

               25

CHANGAN   capital of Han

LUOYANG  lesser capital of Han

LINZI ,   LU     Han

MAOLING. ZHANGLING Han

YANLING  Han

CHENGDU, WAN  Han

CTESIPHON/SELEUCIA/T   Parthia

MERV   Parthian empire

ROME   capital of Roman empire

ALEXANDRIA   Roman empire

ANTIOCH   Roman empire

EHESUS   Roman empire

PERGAMUM   Roman empire

CIBYRA   Roman empire

APAMEA   Roman empire

TAXILA   Saka Kingdom

PURUSAPURA  Saka Kingdom

VAISHALI  city of the Licchavis

PATALIPUTRA

PRATHISTARA Satavakhanas

JERUSALEM   Kingdom of Judea

ANURADHPURA Sinhalese capital

 

500

 

              18

CONSTANTINOPLE  Capital of Eastern Rome

ALEXANDRIA   Eastern Rome

ANTIOCH   Eastern Rome

EPHESUS    Eastern Rome

JERUSALEM   Eastern Rome

CAES.MARITIMA   Eastern Rome

CTESIPHON   Sasanid capital                                                                                                                                                                                                                                                                                                                                                                                                    

MERV   Sasanid empire

ROME   Ostrogothic Kingdom

MADURAI   capital of the Pandyas

LUOYANG   capital of the Northern Wei

CHANGAN   Northern Wei

DATONG   Northern Wei

YE   Northern Wei

 NANJING  capital of the Ch’i

TEOTIHUACAN 

TIKAL,  Mayan

CARACOL Mayan

1000

 

               26

CONSTANTINOPLE   capital of Byzantium

THESSALONIKI   Byzantine Empire

KAIFENG  capital of Song China

CHENGDU

SUZHOU

 

BAGHDAD  capital of Abbasids (Buyids)

RAYY,   ISPHAHAN

NISHAPUR

SAMARKAND Uighurs

AL AHSA  capital of Qarmatians

CORDOVA   capital of Umayyads

SEVILLE

FUSTAT/CAIRO   capital of Fatimids

TINNIS

LINHUANG  capital of Liao

KYOTO,   NARA,  Japan

ANI  capital of Armenia

VENICE   Republic

KANAUJI   capital of Pratiharas

ANHILWARA   main city of Gujerat

THANJAVUR   capital of Chola Kingdom

ANGKOR   capital of Khmer kingdom

DALI capital of Nanzhao

TOLLAN capital of the Toltecs

.                      

  The modern world was initially characterized by the strength of

urbanization in West and East Asia.   That was when the Moslem domains in effect became the centers of the world system, and Song China recovered from internal and external troubles.   The spectacular successes of the Mongol cavalry army dominated the 13th century and this putative world empire reached a high point in about 1250-1280.   Table 4 shows for 1300 all of the world cities to be under Mongol rule.   The dream of world empire faded quite promptly, but the mere attempt  ended in internal collapse and catastrophic failure for all concerned, including the world cities, and delayed the growth spurt that might have otherwise been expected.   The three prominent Asian empires that lasted until well into the 19th century:  Manchu in China, Moghul in India, and Ottoman in West Asia, can best be described as post-Mongol formations:  fierce on the outside but hollow inside, large but devoid of innovation, and of which only China displayed one or two world cities. 

 

The impetus for change came from Western Europe that was managing

a shift of focus from the Mediterranean to the oceanic world.   When that transition took effect, about 1500, substantial cities began to appear after 1700 and grew with amazing speed in the 19th century.   Having stepped into the shoes of the Moguls in India after 1859, Britain, at about 1900, briefly entertained a dream of empire, also holding much sway in the Ottoman and the Manchu worlds.   But that dream did not last long and was soon over. 

 

By 1900, the network of world cities (1m +) holds especial interest.   Of

the sixteen “millionaire” cities of that year, a majority could have been described as “imperial”:   Beijing, clearly, Tokyo (Empire of Japan), St. Petersburg, Moscow (Russian empire), Vienna (Austro-Hungarian empire), Berlin (German Reich), Paris (French Republic, with large colonial holdings), and London (with Manchester, Birmingham, Glasgow, with Calcutta, of the short-lived Indian Empire -1876-1947).   Only the United States (with New York, Boston, Philadelphia and Chicago) might have might have shunned that label.   What is striking is that all these self-described empires were more so in name than in substance.   Though they could be recognized as “great powers”, none dominated the world – only

       

         Table 4:   Modern world cities: imperial and non-imperial, 1000 to 1500

                                        (est. one million inhabitants and over)

Year

Imperial cities

Non-imperial cities

1000

 

BAGHDAD  Buyids (Abbasid Caliphate)

1100

KAIFENG   No.Song capital

BAGHDAD   Seljuks (Abbasids)

SONGDO   Koryo

1200

 

BAGHDAD   (Caliph an-Nasir)

KAIFENG   Ruzhen Jin

HANGZHOU   So.Song capital

1300*

BEIJING   Mongol capital

HANGZHOU   Mongol China

 

1400

NANJING   Ming capital

 

 

1500

BEIJING   Ming capital

 

 

*   In 1300, Baghdad, Kaifeng, and Songdo, while much diminished, were also under

 Mongol rule.

                       

 

Britain seemed to approach that condition, and did indeed control South Asia (the domain of the Moghuls), and was strong in East Asia and West Asia.   But in hindsight all were in different stages of democratization, and when that took hold their time was past.   Democracy corrodes empire.   

 

In other words, as evidenced by the record of world cities, empires

can hardly be said to have been “ever present”, and even less, creative, in the past five millennia.   They can take little credit for urban growth or economic development though they did attempt to harvest the fruits of that growth when they ripened,  for empires are a primitive means  of  domination all round, and not just instruments of economic domination.    Their much touted products, peace and order, failed to hold   The Roman, and the Chinese empires may have served lessons for state builders in the modern era but it is the example of democracy in the city states of classical Greece that captures real interest, and claims most relevance.   But even the Athenians failed to translate their democracy into a prescription for fruitful and more than passing co-existence with their fellow city states, let alone for the Mediterranean world of their time.   Empires had a place in one phase of the evolution of the world political system, but they were not one of the essential innovations, and in the modern world increasingly became an anachronism.

 

 

Selection in the modern world

 

                        Selection refers, of course, to the third phase of evolutionary learning.

 In evolutionary biology, this is natural selection, that is selection by the forces of nature, such as climate. environmental change, or other species.  In social arrangements, there is social selection, either by collective agencies or individual choice.   Decision is selection among alternatives.   At the most general level, the modern world system might be undergoing a phase characterized by “selection” in two respects:  as human species in a changing environment, and as a species internally selecting viable forms of world-wide species organization.  Our concern is primarily with the latter.

 

                        World system evolution is about the arrangements that humans devise for living together on this planet in conditions of rising inter-connectedness.   Theoretical analysis (Devezas and Modelski 2003) shows it to be a project that is about 80 percent complete.   That situation did not obtain 5000 years ago but world system organization is now inescapable, and probably irreversible.   We also know that over the experience of the modern world system, world empire has already been selected against, twice in a major way.   In the 14th century the spectacular failure of the Mongol project was one strong signal.  In the centuries since 1500, a European imperial solution was four times selected against, albeit at heavy expense, the preferred system being global leadership, a transitional solution mid-way between empire and modern world organization.   The time has now come to move the world system toward a more stable solution.   That which is being selected is a form of governance that will make possible human survival into the “long” future.   These will be multi-level forms of social organization sturdy enough to cope with such priority issues on the world agenda as the role of war and nuclear weapons, center-hinterland interactions, wealth disparities, and environmental stress.

 

                        The study of major urbanization over the past several millennia cannot tell us precisely what forms of world organization will be, or need to be, selected.  That topic needs a more comprehensive treatment.   What it does make clear though is this:  for long stretches of the modern era the need for change hardly appeared urgent or convincing.   Currents of change were already underway but they remained mostly subterranean, hidden from view.   For a short time, and until the 20th century, it seemed as though only ‘the West’ was changing.

 

                        What the study of urbanization does bring into full view is the abruptness of recent social change.   We now know that over what seems only a moment, the last century between 1900 and 2000, the world’s social make-up has undergone sudden, and unprecedented change: the number of people on earth has risen dramatically, and qualitatively the world’s makeup has moved, from chiefly rural to majority urban.   As noted, the number of “millionaire” cities shot up from 16 to some 300 in just the last one hundred years.   (Table 5)    What also follows is not just that urbanization cannot possibly continue at the speed of the past century because soon we would all be living in cities (or maybe even in one world city but that the social organization of the species is on the verge of a qualitatively new form.

 

Table 5:   Number of modern world cities:  1000 to 2000

(est. one million inhabitants and over)

 

1000

1100

1200

1300

1400

1500

1600

1700

1800

1900

2000

1

3

3

2

1

1

1

1

4

16

299

 

What we have learnt from the phasing of urbanization tells us that at

mid-points of both the ancient and the classical eras, a change of pace occurred that signaled new dangers and new opportunities.  Has the world system reached a comparable mid-point in the modern era?   Our analysis implies that much but should not be taken too literally because it is anchored in a “rounded” conception of the duration of the modern (and earlier) eras, as extending for 2000 years, from 1000 to 3000.   A finer analysis, of global political and economic processes, might suggest an earlier mid-point, as early as 1850.  (Modelski 2003b).   The onset of an “Age of Redistribution” (Modelski 2003b) makes a decision process even more likely.   Current population projections indeed indicate a growth slowdown for this century and next and that, too, implies that the world is near, or at, the mid-point it had reached in two previous eras.

                         

                        Developments other than urbanization also signal a time for decision.  By 2000, a majority of the world’s inhabitants have come to live in democracies, an advance that clearly owes much to urbanization.   Politically, nuclear weapons are now in the hands of governments of about one half of the world’s population;   the weapons revolution makes  the destruction of urban life, and even mass extinction of humans a theoretical possibility, hence poses another sort of limit. The information revolution now supplies the knowledge requirements of a world system available to much of a highly inter-connected population.  That is, decision might now be seriously in order but it is also worth remembering that this still leaves several centuries for that phase to be completed.

 

                        What might be the mechanisms of selection?  The first lesson of our analysis might be that this will not be empire, or imperial rule.   The record of such solutions is worse than uneven and does not justify their reputation.  The most damaging thing to be said about them is that they are incompatible with democracy.   Imperial dominance is monopoly of power and absence of accountability, and these are inimical to democratic arrangements, as well as to evolutionary stability.

 

         The best general mechanism of selection in fact appears to be that implied

in democratization.   Recall that at its base, world system is about the social organization of human species, and that means that the selecting mechanism will be actuated by the world’s population.   This might take the form of each individual acting on their own  as for instance in the decision to form, and have or not have, families, to move to a city or to leave it, or to participate in, or withdraw from, markets, by buying or selling.    Or else it might find expression in the decisions of collective agencies and organizations such as governments, at several levels, that are accountable, and responsive to, their respective constituencies.   Being self-correcting, and self-legitimating, democracy is the system most compatible with the requirements of evolutionary stability and change.     

                       

 

In summary

 

                        This has been a test of propositions maintaining that the human experience of the past 5000 years can be understood in a unitary perspective as one period of a continuous process of (prospectively) four major phases but one that can be portrayed and narrated in one uninterrupted sequence (except for the Americas in the classical age where such continuity may be in question).   This continuity assumption (also confirmed by David Wilkinson for Central Civilization) is supported by the evidence of major urbanization, combined with data on the world population.

 

                        But the test goes beyond the continuity thesis.   It showed that viewed though the prism of urbanization world system evolution is neither linear nor random, but  phased because at fairly regular intervals and at increasing levels of complexity it exhibits distinct phases.  That makes it possible to argue that those phases are successive installments of a Darwinian learning process:  the process of world system formation.

                       

                        Grounds for this conclusion lie in the recognition that the course of world system evolution has been marked by a qualitative succession of major innovations theoretically predicted for these phases, and it is these innovations that best characterize the entire process.  The first of these phases (or eras) laid down the groundwork for further advance, the very infrastructure of the learning process.   It launched cities, invented, and broadcast writing, and organized time, assuring conditions without which further learning would have been inconceivable.  In the second phase, the narrow basis of human solidarity was expanded, and the formation of larger regional communities enabled, by the dissemination of religions. 

 

As indicated i.a. by city data for the third, selection, phase of this macro-

learning process,  conditions are now changing at an extremely fast rate and will be seen as helping to justify calls for action toward viable forms of world organization.   The quality of that selection and the innovations linked to that process will have much to do with the survival capacity of the human species.   We do not know what forms of world organization will, or will not be, actually selected and innovated.   That set of decisions will be up to all the members of the human family.   What it does graphically show is that the time for decision is nigh.

                       

 

 

 

Notes

 

 

*    Paper presented to the “Globalization and World System Dynamics” panel,   annual conference of the International Studies Association,

                                Montreal, March 18, 2004

 

1.   For a fuller discussion see in particular Modelski (2000) and Devezas and Modelski (2003b).

 

2.  This paper draws substantially from Part III of World Cities:-3000 to 2000,  and most of its data from that volume.  

 

3.     Henry Plotkin enumerates three phases of the “g-t-r” (Campbell-Lewontin) heuristic.

 We add a fourth, cooperation, that too can be traced to Darwin’s writings, but that has also been elaborated over the years by such writers as Peter Kropotkin, Robert Wright. or Matt Ridley.

 

4.   In this way we go beyond postulating a ‘heuristic’, to proposing that each element of that heuristic is optimized in a definite and significantly long period, and that each such phase of that process has roughly the same length, namely 2000 years.   Why should phases be of equal length?  They facilitate coordination with other evolutionary processes.

 

5.   The network of world cities maps the center of the world system, and its mapping yields information, in the first place, about the center. But the condition of the center also reflects the state of center-hinterland relations, hence also the state of the whole system

(see also Modelski 2003b).

 

 

 

 

References

 

Devezas, Tessaleno and George Modelski   (2003)   “Power Law Behavior and

World System Evolution”   Tech. Forecasting and Soc Change  November, 819-860.

King, Paul L. and Peter Taylor eds.  (1995)  World Cities in a World-System,  

Cambridge :  Cambridge University Press.

Modelski,   George   (2003a)   World Cities: -3000 to 2000,   Washington DC: Faros2000.

-----          (2003b)   “Ages of Redistribution”   Conference on “World System History

and Global Environmental Change”, Lund.

-----          (2000)   “World System Evolution”   in R. Denemark et al eds.   World

 System History,   New York:  Routledge.

Plotkin,  Henry   (1997)   Darwin Machines, and the Nature of Knowledge, 

Cambridge:  Harvard University Press..