Understanding the past

Each generation has a fresh opportunity to understand the past, by perceiving it from a unique standpoint: that of the present. Since this standpoint constantly shifts, so also do views of the past. This common-sense assertion does not imply a retreat into relativism - that there can be no common description of phenomena, only an infinite choice of attitudes - because it is true of all our cognitive processes: out of the uniqueness of personal experience, common perceptions emerge. It does, however, help to make sense of why certain views of the world are espoused at certain times, and why our perceptions differ from those of our predecessors (and, not unusually, many of our contemporaries). What contemporary conditions do is foreground certain properties of the world around us, because of their current relevance. They direct our attention to comparable conditions in the past. To a previous generation these particular properties were unimportant or even invisible, because they had no pressing relevance. They were not within the visual field of the backward-looking gaze. It is only by accumulating the insights of successive generations, and observations made from a diversity of standpoints, that a fully three-dimensional view of the past can be obtained. That is why intellectual history is one of the most fundamental disciplines of the social sciences: it explains why opinions differ, and why our colleagues can be blind to what seem to be self-evident truths.

The current, minority, interest in world-systems' perspectives falls squarely within this formulation. The institutional structure of academic life contains a multiplicity of barriers which insulate disciplines and shield them from the implications of a global view. World-systems approaches diner, in this respect, from other contemporary shifts in perception such as those achieved by feminism. When gender roles are in flux, it is easier to see them as culturally constructed rather than god-given or determined by nature. Although globalization is a phenomenon which affects us all, it does so in ways which are less direct, and are interpretable in a multiplicity of different paradigms. World-systems viewpoints are also easily mis-represented as the revival of nineteenth century conceptions when 'globalization' was essentially colonialism. Since many 'subjects' in the humanities are set within subject areas whose definition implies a cultural genealogy ('classics,' 'oriental studies'), models which imply that these entities are

[115]

misleading are likely to be stoutly resisted by established interests who are happy to rule within their own (shrinking) domains.

Nevertheless, the time is propitious for a global academic view, and the experimental formulations now surfacing are a foretaste of what is likely 'to become the common coin of discussion in the social and historical sciences - until such obvious truths, hardened into unimaginative dogma, are themselves attacked and re-formed. Symptomatic of these changes is the way in which certain phrases and metaphors have an instinctive appeal. One of these is the idea of the network. It is hardly coincidental that many of the contributors to this volume have communicated with each other via the Internet. It may be hoped that the immanence of this concept will lead to a more ready acceptance of Eric Wolfs classic appeal (1982: 1-7) for an emphasis on connections rather than on entities, since an entity-by-entity approach renders invisible many of the most significant features of human history. Bounded structures are secondary phenomena arising from the properties of networks: only the crude convenience of national history and national accounting have given these transitory entities a disproportionate role in the social sciences (including anthropology, where illusory equivalents of nation-states have been sought in the mysterious entities called 'tribes'). Networks have no logical boundaries, save at the global scale; all social systems are 'world systems.'

One other current re-envisioning of the world is of relevance to the world-systems challenge: consumerism. In a world of shopping malls and competing brand images, it is not surprising that there has been a rash of books with tides such as 'Consuming culture.' It is easy to dismiss such interests; but I argue that they reflect a more fundamental rediscovery of the importance of culture, and more particularly the simultaneous meaningfulness and materiality of material culture. This is not to resurrect a reductive materialism, but to assert me power of ideologico-material phenomena, whether these be recent features of the world scene, such as 'coca-colanization, ogearlier phenomena such as 'farming,' which has traditionally been treated as a set of subsistence arrangements but must now be seen as a socio-subsistence system embodying powerful ideological elements. This approach would see the world as being made up, not of cellular units of culture, but of growing arenas for competing regimes of value. The propensity of certain ways of life (such as 'farming' or 'civilization') to spread at the expense of others is fundamental to the study of world systems, while scarcely apprehensible within the conventional disciplines.

What is required to take advantage of these intellectual opportunities is a new kind of discipline. Neither the conventional comparativism of me social sciences, nor the inherent insulation of the classical humanistic disciplines, is equipped to deal with the connectivity of the real world. What is necessary is a larger, structural approach. Aspects of this approach can be found scattered through subjects which have already obtained intellectual respectability. One is anthropology, which has resolutely opposed the rigid forms of categorization (for instance into

[116]

economic, religious or social aspects) which form me basis of the conventional academic division of labor; but its concentration on small-scale, face to face interaction - which gave it this comprehensiveness of oudook in me first place -has directed disciplinary attention more to on-the-ground manifestations than toward larger, structural properties.² Another is human geography, which — had its recent history been different — might constitute just the discipline mat is required; but its lack of a time dimension (for historical geography rarely penetrates beyond the medieval period in Europe), and successive concerns with urban planning and cognitive topography, have largely inhibited the more comprehensive formulation mat is now needed.³ History, of course, should be the answer: but the longitudinal view of me development of past societies over a timescale of millennia is interrupted by a differently constituted domain of 'ancient history,' which is subject to me pressures of its own peer community in ways which inhibit its contribution.

The cumulative effect of disciplinary divisions is a disabling fissivity which disguises me common aspects of their enterprises. An academic division of labor which partitions spatial aspects as 'geography' and temporal aspects as 'history' calls for the satirical skills of a latter-day Jonathan Swift in imagining how a university might set about measuring a cube, when this involved an inter-faculty committee to coordinate me departments needed for the different horizontal and vertical measurements. Academia is constituted in such a way as to obfuscate recognition of large-scale spatio-temporal structures, and to deny advancement to those mat do recognize them. That is why the field is studded with me names of intellectual mavericks whose works defy disciplinary attribution, and are me bane of librarians forced to work within the conventional categories.⁴

One field of inquiry ought to have the relevant credentials, and that is archaeology. With memods applicable irrespective of social complexity or cultural specificity, its perspective can encompass me broad sweep of prehistory and track the flow of items which moved around the world in historical times, whether or not the bills of lading have survived. In practice, however, archaeology offers a microcosm of the problems which beset the wider academic community. Its few practitioners are haphazardly deployed across a trivial fraction of the evidence, and subject to the same forces which beset their colleagues in other historical disciplines. Nevertheless archaeology commands an increasing body of observations, over vast expanses of time and space; and its relative freedom from the constraints of linguistic specialization, together with its comprehensive approach, make it a natural seed-bed for large-scale interpretations. It might even be argued that archaeology manifested this propensity too soon, in the set of doctrines known as diffusionism, and that the contemporary effort to re-introduce world-system ideas is made more difficult by the taint of association with these now unfashionable ideas. In this respect it parallels the cycle of interpretations in ancient history from the nineteenth century on, between primitivists and modernizers,⁵ which was both a debate about whether the economy of ancient Greece was simple or complex, and about the role of the Near East in relation to Europe. At me same time as the 'primitivist' (substantivist) interpretation of me ancient economy was being forcefully restated in the post-World War II period by Karl Polanyi (1958)

[117]

and Moses Finley (1973), prehistorians were at the furthest extent of their reaction to the difiusionism of Gordon Childe, so that an emphasis on local autonomy has recently prevailed both in prehistory and in the study of the ancient world.

These attitudes are currently in course of change; and it seems a propitious moment to attempt to develop a common conceptualization with those world-system theorists (e.g. Frank and Gills 1993) who have broken out of Weberian constraints and are attempting to extend an interactionist framework back from the sixteenth century to the beginnings of urban life in the fourth millennium BCE. This chapter is offered as an archaeologist's view of what a common description of global processes might look like, using insights from a variety of disciplines, but with a terminology taken directly from none. It is based on a simple premise: that time and space are the dimensions of the process to be studied, and are not, therefore, a useful basis for long-term academic specialization. History and geography must fuse.

If the continuity of temporal development (and thus the common properties of social systems through time) has been lost in the proliferation of specialist disciplines, it is the specificity of spatial processes which has been partly abandoned by geographers. Seduced by mathematical elegance, a generation of 1960s ('New') geographers gave up the inductive approach in favor of attempts to specify fundamental properties of spatial organization on an isotropic surface. But the world is not isotropic, and important developments have taken place in topologically complex and atypical parts of the Earth's surface. It might even be argued that change itself is closely related to the existence of such singularities. A unified spatio-temporal approach must balance comparability and specificity.

Informing such a methodology must be a sound grasp of the shape of the structures being sought. Two competing visions of the dominant processes of world development have alternated in the minds of theorists. One might be called the 'layer-cake,' the other the 'calyx.' The former is based on stratigraphic succession, and gives the impression of a set of common phases or stages through which populations progress. Development theory in economics (Rostow 1960), is a recent example, paralleled by the still-powerful school of regional autonomism in prehistoric archaeology or the candelabra model of human origins in biological anthropology (early dispersal of a common stock, followed by the parallel development of different regional populations towards modern humanity). These were all, in their different contexts, elaborated in rhetorical opposition to diffusionist models, such as colonialism. While contagious processes such as difiusionism largely lack such an explicit theoretical description, there is a clear kinship between the kinds of models which have recently tended to succeed 'layer-cake' models in a variety of disciplines: core-periphery models of economic develop-

[118]

ment, both in historical and prehistoric times; or the 'out of Africa' models which have recently become the new orthodoxy of human evolution. By comparison with the horizontal stratification through time which characterized the layer-cake, these latter attempts at visualization describe structures which are the shape of inverted cones, beginning in specific areas and growing upwards through time and outwards in space like a bush or flower, hence the image of a flower-calyx.

Both images are 'true.' Each metaphor describes one aspect of the temporal process, and the two images have continuously alternated in popularity in biological and social theory — Darwin and Spencer being notable nineteenth century exemplars of the genealogical bush and the stadial sequence respectively. The more organic metaphor appeals to the Romantic imagination, the regular succession accords better with the Enlightenment sense of order. It is not accidental that one image is based on an individual life, the other on impersonal sedimentary accumulation. Such metaphors have alternately risen to prominence in both science and literature as they have reflected dominant aspects of their contemporary economic, social and political environments (Sherratt 1992; 1997).⁶ Each has a measure of validity, and it would be wrong to emphasize their mutual exclusive-ness or to claim that one is more 'scientific' than the other. After all, the growth of an organism is as scientifically describable a process as is sedimentary deposition, but there are undoubtedly times when the usefulness of one image needs to be asserted against the dominance of its alternate. This would seem to be the case today: organic growth needs to be seen as an important property of human social systems, as opposed to neoevolutionary stadial descriptions. Long-lasting structures which expand through time (the calyx image) seem to be a better representation of important phenomena as they can be reconstructed over long periods than does the more passive image of successive layers.

Such growing and expanding structures are not unique to human society. They are a fundamental feature of biological evolution also, describing how new species emerge. Improved palace-environmental reconstructions not only provide a more realistic background to biological evolution, but also form a significant mechanism of change itself. This has led to a new conception of 'punctuated' change, produced by episodes of rapid environmental forcing. What has been less emphasized, however, is that this temporal punctuation has a spatial dimension in that the specific conditions which give rise to change are limited to particular places or regions, from which successful innovations then spread (Sherratt 1997). Nowhere is this better exemplified than in the rift valleys of East Africa where human populations had their ancestry. Successive generations of human ancestors emerged in the environmentally diverse conditions of this geologically unusual setting; and from here successive waves of more advanced hominids spread out to replace or overtake earlier generations who either survived unchanged, or (like the Neanderthals) underwent physical specialization as they adapted to the cold northern margin. It is only the speed with which successful innovations spread over large areas that gives the impression of a common, stadial or layer-like sequence.

The calyx structure of spatial expansion, and the center/edge processes that

[119]

have given rise to it, are seen as fundamental even to biological evolution, and not peculiar to human societies and phenomena such as the emergence of capitalism. Such structures underlie the entirety of human history. What is striking about the ' last 10,000 years is the multicentric pattern of such structures: several regions stand out as the foci of fast-spreading contagious processes. Whereas biological change had been uni-centric, the cultural changes made possible by the emergence of modern behavioral complexity have been notably multicentric, including foci in both the eastern and western parts of the Old World, and the central and Andean parts of the New World. During the last 10,000 years, in the warm conditions of the Holocene interglacial period in which we are still living, successive developments have consistently arisen within the same focal areas of change -notably the geographically unusual conditions at plate boundaries in sub-tropical environments where narrow montane belts are surrounded by sea and desert. These foci or 'nuclear areas' have fundamentally affected their surrounding regions, in large-scale center/edge processes; it was these macro-structures which gave rise to the idea of 'diffusionism.'⁷ This formulation has been unpopular during the last half-century, and the term is no longer used in a respectable sense, but it is vital that the idea should be revived since it correctly expresses essential properties of the process. Whereas biological changes such as the spread of successive human species involved the emergence of a new genotype and its. spread by migration and replacement, Holocene changes came about as much by the adoption of new practices by existing populations as by migratory spread. 'Nuclear' areas are a property of behaviorally modern humans, capable of forming networks of trade and exchange, transmitting complex messages and learning new practices by imitation.

The appearance of consistent foci of change was closely related to the domestication of a set of high-yielding but labor-intensive cereal grasses, together with certain legumes. These annual plants, adapted to strong seasonal stress, formed parts of vegetational communities which expanded in certain areas in inter-glacials. Although modern humans had probably experimented with plant propagation and cultivation during the last glacial period, it was only with warmer and wetter conditions that cereals came to be present in sufficient numbers to justify special attention. It is likely that they were only taken into cultivation as a result of climatic changes which forced the use of all available resources, however labor intensive. Initial cultivation systems were essentially horticultural, and involved little forest clearance. The principal form of labor input was the back-breaking process of grinding the seeds, undertaken by women. Since cereals and beans could supply protein as well as carbohydrate, their cultivation led to population growth; since cultivation could be extended to new areas, population growth could be sustained. This was expressed both in increasing local densities and in a propensity for farming to spread. The spread was so rapid that they overtook more slowly developing forms of cultivation (e.g. tending root-crops); though local

[120]

crops might be integrated into an expanding complex, and even take over the role of principal cultigen as farming spread to new climatic zones (as with temperate millets in North China or tropical millets in sub-Saharan Africa). The date at which such explosive processes began, and the area which they affected, depended on local vegetation and climate. The western Old World focus, with winter rainfall crops, began right at the onset of the Holocene. New World developments, in tropical (summer rainfall) settings, began much later (Sherratt 1997). The seasonality of crops affected the area to which their cultivation could spread: western Old World crops spread rapidly into a temperate (winter-rainfall) hinterland, while New World crops spread more widely in adjacent tropical areas and penetrated only slowly into temperate North America. The availability of potentially domesticable ungulates also affected the character of farming. The New World provided fewer candidates than the Old, and only in the Andes was animal raising more important than hunting. For these reasons, Old World societies evolved faster than New World ones. The following discussion deals principally with the western Old World.

Farming systems were characterized by a tendency to increasing sociocultural complexity and a tendency to spread. Once farming was in existence, the process of cultural change was reinforced by features which accompanied it: larger communities, increased sedentism and the associated arts of architecture, pottery, textile production, and stone-working. This horticultural complex spread westward over Mediterranean and temperate Europe, and eastward through Iran to western India. Within the new village communities, patterns of private consumption grew up by which certain families and villages came to control a larger portion of consumable and durable commodities than others; ostentatious display became an increasingly prominent part of village life. The more complex forms of material production associated with this way of life were important in the transmission of farming and associated practices to indigenous foraging groups. Elaboration of material culture was further promoted by technological innovations. Copper became a widely desirable commodity and medium of exchange, promoting liquidity. In temperate Europe, this accompanied an enhanced role for cattle-raising. In the ecologically more diverse environments of the nuclear region it permitted a degree of specialization in the production of products such as cheese, wool and fruit. These gave rise to more specialized forms of husbandry, such as pastoral livestock rearing or arboriculture. They were matched by more intensive techniques of cereal production: irrigation, the use of animal traction, and the development of a simple plough. These innovations remained largely restricted to the nuclear area where farming began. In temperate Europe these more advanced forms of agriculture and pastoralism did not appear until urbanism had begun in the Near East, perhaps as a consequence of increasing long-distance links.

The processes by which farming spread can be summarized as a nuclear/ margin system. Innovations, including primary farming ('founder crops' and meat-animals) and secondary farming (commodity crops including animal products and tree fruits), occurred primarily within the nuclear area. New forms of

[121]

food-getting were accompanied by technological innovations (pottery, metal) and social changes. Some of these spread ('diffused,' whether by migration or adoption) out of the nuclear region, but this extension had little effect upon the nuclear • area itself: the process of spread did not create an outer sustaining area, and the term 'margin' reflects this lack of two-way linkage. What the existence of a margin did was to provide an external arena of farming and simple metal-working societies, which might be tapped by long-distance routes for particularly desirable commodities. This, however, was a feature of the next phase.

The process by which more differentiated structures arose may again be followed in the Old World: but once more it stands for three or four independent regions. The model proposed here is based on an interpretation of urban centers as locations of manufacturing activity: more precisely, of added-value production. This definition does not exclude administrative or religious functions; indeed, it is formulated so as to encompass them. Urban centers import raw materials and add value, exporting some of their products: the difference pays for the materials and other consumables. The value created in the manufacturing process is ideological as much as material. It is the desirability of the products which keeps the cycle in existence (cf. Hornborg, this volume). The outward growth of an urban system can be described as the expansion of a hegemonic regime of value, in which new consumption practices are promoted.⁸ Its typical commodities are metalwork, textiles and psychotropic consumables such as alcoholic dnnks, packaged within a system of beliefs and practices within which they have a coherent ideological role. These practices are (he prerogative of a minority, and societies characterized by these forms of specialized elite consumption have traditionally been termed 'civilizations.' These form a special category of high-consumption societies, whose needs have to be supplied through the transport of large volumes of goods, far beyond the requirements of subsistence. Ideological and material aspects are thus intimately related: missionaries tell the natives they arc naked, and traders sell them clothes! Flows of ideas are as important as the flows of products. Nevertheless the process has a definable economic base, which is the asymmetrical division between suppliers of raw materials and providers of manufactures. The emergence of elites is therefore not a local process, but always takes place within a larger structural setting (Sherratt 1995). An important element in the continuing expansion of such a system is import substitution: peripheral areas acquire the skills to undertake their own manufacturing processes, becoming pans of an expanding core region, and generate their own external supply area which becomes the new periphery. Some innovations spread further, as easily transferable skills spread among surrounding populations, though without entailing a continuing economic articulation. This outer area, culturally affected by the existence of urban communities but not actively participating in their economy, forms a margin in the same sense as the spread of farming created a penumbra of farming societies around the original nuclear region.

[122]

In many respects, therefore, the formation of urban systems can be seen as a continuation of the structures and processes created by farming. Instead of a nuclear/margin system there was now a core/periphery/margin system, with the nucleus functionally differentiated into core and periphery. Whereas the relationship between nucleus and margin involved only the transmission of innovations and not any continuing interdependence, that between core and periphery involved a series of real-time interactions, where changes in one partner actively affected the fortunes of the other. This assymetrical relationship was the first regional division of labor. Such zonally specialized systems emerged within each of the areas which had previously been central to the origins of farming, and core regions with urban economies grew up at specific locations within them: typically in alluvial river valleys which were major transport arteries (the Euphrates, Nile or Indus). These were associated with improved transport systems including boats with sails and pack-animals. Such alluvial environments also provided habitats where the productivity of farming could be intensified by irrigation, and this correlation has in the past suggested explanations of urban origins in terms of the achievement of an agricultural surplus. While this provided a means of sustaining larger communities, it can no longer be seen as simply causal, since it is the new consumption patterns which demand explanation, not their calorific base.

Agricultural surpluses were not exported directly, but in the form of labor-time in the production of desirable manufactured commodities: improved farming supported sheep and people, and textiles were exported as part of a total ideological package. The process behind the emergence of a core was in fact industrialization. This is reflected in the appearance both of techniques of mass-production (like wheelmade pottery) and of advanced craftsmanship (like sheet-metal vessels or elaborate jewelery). Innovations in technology were often directed toward import substitution, as in the invention of blue glass as a substitute for lapis lazuli. The Industrial Revolution of the eighteenth century can be seen as a logical extension of this system, on a new, global scale (and with a fossil fuel subsidy), rather than a fundamental rupture with 'traditional' economies.

Although the system began within a strongly religious and theocratic framework, the proliferation of competing centers within the advanced ideological/ manufacturing core created tensions which gave rise to armed conflict, and precipitated forms of social organization in which the explicit use of force was institutionalized: the state, and soon also a system of temporarily dominant regional powers which were experimental forms of empire. Such structures typically attempted to control flows of high-value products and their nodal points rather than territory as such. As the system grew in area, it altered in scale and configuration. Increases in scale were reflected both in centralization and in the emergence of competing foci of power. As new external supplies were tapped, major shifts in arterial flows could occur, which brought new centers to prominence and isolated old ones. New transport technologies could also alter the balance of advantage between land and water transport, and since sailing vessels could cope with increasingly bulky cargoes, maritime transport privileged certain coastal areas. All these processes were capable of producing rapid alterations in

[123]

the wealth of particular regions, and such shifts in the topology of the system were probably more important than cyclical phenomena in explaining ups and downs in prosperity.

The spread of urban trading networks, and their extension along the Persian Gulf and eastern Mediterranean, created a complex molecular structure of regional foci so that as well as the zonation of core and periphery (originally created around Mesopotamia) there was a series of interacting civilizations: Mesopotamia, Egypt, the Indus valley; then also Syria, central Anatolia (Hittites) and the Aegean (Minoans and Mycenaeans). Beyond this was a margin which included not only temperate areas such as Europe, but the dry steppe corridor of central Asia. This was truly a world system, even if it occupied only a restricted portion of the western Old World. Whilst each civilization emphasized its ideological autonomy, all were identifiably part of a 'common-world' of interacting components. In the eastern Old World, China had only indirect contacts with the complex of western Old World civilizations through their common margin, which now extended across the forest and forest-steppe belt of Eurasia. Nevertheless, important technological innovations passed between East and West in the Old World during the Bronze Age.

Aspects of bronze-casting metallurgy reached Europe, and chariotry reached China, across the steppes. Societies on the margin were also increasingly penetrated by long-distance trade. In Europe, the Danube formed a major axis of contacts from the Black Sea to the center of the continent, succeeded (as trade spread west along the Mediterranean) by contacts across the Alps as far north as Scandinavia. These carried precious materials over long distances, such as the small quantities of amber which reached Troy and Mycenae, and were even buried in the tomb of Tutankhamun. In mirror-image of these northern routes, tropical products such as incense and spices came to Egypt from Nubia and Somalia.

Although divided into 'civilizations' and a multiplicity of polities, the growth of urban systems is a clear example of contagious expansion (corresponding to the 'calyx' model rather than the 'layer-cake'). The concentric structure of core, periphery and margin expanded simultaneously: the nuclear area (core and periphery) because of the dynamic of competition and import substitution, and the margin because innovations which spread beyond them created new cultural possibilities (as with wheeled vehicles on the steppes, or iron-working in Africa). The system was very sensitive to transport costs, so the urban core and its peripheral supply area expanded differentially along maritime sea-lanes and tributary rivers. The network of urban links expanded first along the length of the Mediterranean and Black Seas, opening peripheral hinterlands in temperate Europe and the steppes and semi-desert regions of eastern Europe and central Asia. By the end of the first millennium BCE, both the western and eastern Old World systems had grown to such an extent that their peripheries overlapped, and trade between the cores themselves became a significant factor, increasingly in amounts which gave the advantage to transport by sea.

[124]

While maritime expansion in the Mediterranean had begun in the Bronze Age, it was interrupted by a period of system-wide crisis at the end of the second millennium BCE, which coincided with the discovery and spread of ironworking in the east Mediterranean (Sherratt 1994). Fundamental technological and social changes produced a dissolution of old forms of political organization and the emergence of a new pattern with a block of territorial empires bordered by maritime city-states, which spread all round the Mediterranean. A new order of metropolitan cities (with more than 100,000 inhabitants) appeared at nodal points within this enlarging network, and such primate cities no longer had to sit in the middle of their breadbaskets. Maritime supply allowed them to be provisioned by sea. Slaves also became a commodity to be shifted over long distances to sustain industrial and agrarian production. New external connections began to exercise an important influence. The ancient road to the east, along which lapis lazuli had flowed from Afghanistan, took on a new importance in the second half of the first millennium as it extended over the Hindu Kush to the upper Indus, and continued to the Bay of Bengal. A new generation of Iron Age civilizations developed in India, centered in the Ganges valley. This axis developed first by extending further into central Asia, and second by being to some extent superseded by the maritime monsoon route between the Red Sea or Persian Gulf and India. As the overland trade route into central Asia from the west encountered the similar long-distance route supplying China from the east (the 'horse route'), the effect was to produce the first real-time contacts between the eastern and western Old Worlds: the Silk Route (Franck and Brownstone 1986). As the name implies, it transmitted oriental luxury goods from this tropical civilization to the west, paid for largely by tlie export of silver. The increase of economic activity where these routes entered western Asia (i.e. in Persia) balanced the expansion of trading along the Mediterranean, and created a latitudinal corridor of civilizations within which hegemonic control could shift suddenly between cast and west: from Persia to Hellenistic Greece (which linked the east Mediterranean and Silk-Route terminus) and then to Rome (which unified the Mediterranean).

Political control of the principal arena of maritime actory allowed Rome to develop its extensive European hinterland, principally along coastal and river routes; but the relative underdevelopment and dispersed pattern of resources within this vast area necessitated unparalleled provision of military forces along its margins. This relatively primitive imperial structure nevertheless lasted for many centuries, but it was undermined by a combination of forces: the economic expansion of the Indian Ocean directed attention eastwards, while the growth of 'barbarian' political units increased pressure to the north. The combination of these factors led to incursions, as Germanic-speaking groups attempted to gain access to trading activity which was increasingly oriented to western Asia, symbolized by the shift of capital from Rome to Byzantium. For the next millennium, it was the Indian Ocean that was to be the principal arena of maritime activity, paralleled by (and to some extent alternating with) contacts along the Silk Route. A gravity model of interaction between the western and eastern Old World core/

[125]

periphery/margin systems, and their effective fusion into a single common-world, correctly predicts the internal shifts which each of them experienced, together with the shifting relative advantages of the two transport systems which linked them, by land and sea respectively. Andrew Bosworth's elegant description in this volume details the economic and political implications of this evolving topology of east-west links.

These routes gave renewed importance to precisely those areas where urban civilization had begun: Mesopotamia and Egypt. Baghdad and Cairo became the largest cities of their times. The western Old World nuclear region remained the heartland of an expanded system, now with the added advantage of being situated on the narrow isthmus separating the Mediterranean and Indian Oceans, Europe and Asia (Lombard 1975). Religious ideologies, and especially Islam, gave a cultural unity to areas of interaction too vast to remain under the control of a single territorial power. Agrarian intensification (Watson 1983) and technological advance had their locus in the central area of the expanded system, in the Islamic world of south-western Eurasia and in the coastal southern as opposed to northern China. The growth of Indian Ocean traffic (Chaudhuri 1985), in the three overlapping cycles of Southeast Asia, the Bay of Bengal and the Arabian Sea, provided the motor of hemispheric growth. Yet overland contacts remained important, and just as the Roman Empire gave rise to a barbarian hinterland in temperate Europe, so the activity in southern and central Asia stimulated political development on the steppe belt and adjacent montane and desert areas, to produce a succession of steppe empires, culminating in the Mongols (Abu-Lughod 1989). Ironically, it was the very integration of this corridor and surrounding areas (and imposition of the 'Pax Mongolica,' allowing individuals to traverse the entire route for the first tiine) which permitted the spread of the Black Death, Eurasia's first pandemic (McNetll 1976). The overland corridor never regained its prosperity.

The most fundamental change in the topology of inter-continental connections came about from the sixteenth century onwards, marking the decisive shift to long-distance maritime routes. The exploration of the Atlantic is conventionally ascribed to the blocking of Near Eastern transisthmian routes by the Ottoman Empire; but this 'push' factor must be balanced by the 'pull' factor of the growing maritime competence of the European Atlantic community. Fertilized by the maritime experience of the Mediterranean and the Baltic, and stimulated by riverine feeder routes which linked the Baltic to the northern Silk Route, the North Sea trading communities of the early medieval period found themselves in a modestly advantageous position. Although links to the Mediterranean were primarily overland to northern Italy, the Alpine obstacle limited the quantities of goods that could be carried in this way. There was every incentive to develop a coastal trade linking Europe's two inland seas along the Atlantic coast, and this maritime enterprise was promoted by the development of North Atlantic fishing and by the incentive to outflank land routes across the western Sahara by sailing down the west coast of Africa. The improved ocean-going vessels that resulted led first to the discovery of the Atlantic islands (Canaries, Madeiras and Azores), and

[126]

then to the great clockwise wind systems of the northern and southern Atlantic (Crosby 1986). West Europeans erupted simultaneously into the Americas and into South and East Asia; completely altering the topology of interregional connections. During the crisis years of the seventeenth century, the European center of gravity shifted from Venice to northern Italy to Amsterdam and the North Sea Basin, which now became Europe's point of articulation with the outside world (de Vries 1984). From being on the edge of a Eurasian system, western Europe suddenly occupied its center. Areas now bypassed diminished in importance, until the Ottoman Empire became the 'sick man of Europe.'

It is impossible (though some economic historians manage to do so) to divorce the economic history of modern Europe - the 'genesis of capitalism' and the Industrial Revolution - from this enlarged setting of economic activity (Blaut 1993) and the local zonation of productive activities to which it gave rise within Europe (Wallerstein 1974a; Nitz 1993). In the perspective of millennial growth adopted here, the concentration of added-value production and its associated technologies has always been most marked at the nodal points of interregional connections and the center of the system. The incorporation of all three core/ periphery/margin structures into a single global system gave a unique advantage to the European maritime nations at the center of the new pattern, who reorganized the world to their advantage. The European 'Middle Ages' thus separated two phases during which the European continent rose to prominence because of its centrality: the 'Ancient,' Mediterranean-centered world, and the 'Modern,' Atlantic-centered world - down to the new middle ages which face the Atlantic world as centrality shifts to a new and larger arena of maritime activity, the Pacific.

This chapter has described the development of human societies in terms of evolving, long-term structures, undergoing successive transformations and acquiring new properties. These transformations have taken the form of contagious processes. They may be characterized in hierarchical terms (Figure 5.1) as processes of increasing specificity, from center/edge through nuclear/margin to core/periphery/margin systems. Successive phenomena are clearly related, and each generation of dispersals - farming, urbanism, industrialization, each 5,000 years apart - was nested within its predecessor. Farming and urbanism began within the same set of three or four nuclear regions, though industrialization began in a new central region (Wallerstein's core) which emerged rapidly. In this sense, the 'Modern' world does have a different regional focus and new properties of scale from its predecessors, though it is not fundamentally different in its motivations.

Each subsequent advance was predicated on an earlier set of changes: the dispersal of modern humans in the case of farming; the existence of agrarian populations in the case of urbanization; the existence of a city network in the case of industrialization. These nested structures were manifested geographically as a

[127]

set of concentric zones. The existence of a margin was a precondition for an extension of the periphery and thus of the core itself: communities which were already mobilizing useful commodities could be 'tapped' by centrally placed societies with the advantage of superior technologies and concentrations of capital. Because of this set of spatial and temporal dependencies, the analysis of 'world systems' cannot begin at any arbitrary point in the sequence, but is always faced with the problem of infinite regress. That is why an archaeological perspective is essential.

The genesis of urban systems introduced new properties into the process.

[128]

Specific routes became more important than the spread of innovations across broad fronts. Trading contacts became highly directional. Growth in the quantity of goods traded caused a general shift from land to water transport. Rivers were always important, especially in the creation of urbanism, but sea routes frequently succeeded overland routes and caused recession in the bypassed areas. Major phases of prosperity and population growth have occurred during periods of interaction around the successively larger seas of the Mediterranean, the Indian Ocean and the Atlantic, and perhaps in the future the Pacific. Indeed, the global demographic cycles identified by McEvedy and Jones (1978:343-51), ('primary,' 'medieval' and 'modernization'), can be seen as reflections of these basic geo-economic structures. At this level of abstraction, world history appears highly deterministic, in a way summarized in Figure 5.2. Cyclical upturns and downturns within this pattern (Figure 5.3) may also be related to further discontinuities encountered by expanding spatial processes. This vision of a world with a highly constrained set of macro-structures but great freedom at the level of micro-structures is one which accords well with descriptions of the world offered by complexity theory.

There is a clear zonality to the processes of contagious spread associated with urbanization, which have been crudely characterized here in terms of core, periphery and margin.⁹ These must always be relative terms, for the new properties

[129]

FIGURE 5.3

[130]

which have appeared have altered the nature of development in each, while preserving the tripartite zonal character. Crude estimates of the relative sizes of each of these zones in successive periods (Figure 5.4) suggest that they grew more or less in ratio, though the margin reached finite limits and was increasingly eaten up by the other two. Nevertheless large parts of sub-Saharan Africa remained marginal (in this terminology) down to the nineteenth century, and the term remains a useful one until the final phase of colonialism. It is especially useful as a reminder that most of the societies described in the ethnographic record (and often, therefore, used to construct neoevolutionary, stadial narratives of change), were in large part creations of the same global processes as those which produced the literate societies whose past is recognized as history (Wolf 1982). The way of life of arctic Inuit or East African pastoralists was alike made possible by innovations that stemmed from urban cores. Similarly, the barbarians who have from time to time erupted into history must likewise be situated within the interactions of core and periphery, even though the process has been chronicled only by one side. Such incursions have often been caused by local recession and lateral shifts in hegemony within the core, often as a result of new long-distance supply routes. The histories of all human societies are more intertwined, and at a more fundamental level, than conventional accounts allow.

This contribution has stressed the importance both of antecedent conditions and of geographical realities in understanding the character of social change. It is not intended to supplant other, richer and more subtle genres of history and anthropology, dealing with events and structures on a human scale: indeed, it specifically acknowledges their priority, in insisting on the culturally constructed nature of consumption and the definition of value. In many ways it is the most

[131]

tedious and reductive form of narrative, important only because it is largely unrecognized as a coherent scale of analysis; but it is one which may nevertheless illuminate and articulate more vivid accounts of the human past. In a world in. which histories are often partial and identities largely in Hux, this level of description and analysis may offer some element of common orientation for different cultures and traditions.

1 The hyphen indicates an adjectival usage, not a sectarian affiliation.

2 Notable exceptions include Wolf (1982), Ekholm and Friedman (e.g. 1982).

3 Important exceptions include Vance (1970) and Blaut (1993).

4 Jacobs (1969/1984) is an excellent example.

5 Bucher versus Meyer; Weber and Hasebroek versus Rosiovtzeff.

6 e.g. state-centralization, Keynesian, interventionist, welfare-state, public ownership, community responsibility, comprehensive education versus de-centralized, (Milton) Friedmanite, privatizing, individualist, Reagan-Thatcherite, etc.

7 The Near East (western Asia) has successively given rise to: behaviorally modern humans (Upper Palaeolithic), farming (Neolithic revolution), copper metallurgy, animal traction (secondary products revolution), the first cities (Urban revolution), the first empires, the first maritime states, iron metallurgy, the first metropolitan cities, and three of the major world religions. Yet in spite of this, difiusionism has for the last half-century been vilified as an interpretative framework! This argues for a very strong ideological influence in determining acceptable topics of discourse.

8 The key to this concept is the definition of added value. Value in general is a socially imposed desirability, and as such is culturally specific. There is no absolute, external standard. Rational Choice Theory cannot explain Calvin Klein underpants. Certain common features arise from the common behavioral propensities of modern humans: a delight in bright reflective materials, in methods of adorning the body, in pleasure-giving (and often psychotropic), consumables. Since humans are intensely competitive, these things must also be in short supply, so that their possession and consumption may be limited to a minority, and act as tokens of success. Besides the inherent properties of certain materials, there are also the properties conferred on materials by skilled labor. The combination of these properties has traditionally yielded the highest value, as in a finely wrought gold cup or a royal robe. Both the inherent and the added value are, however, arbitrary, in that they embody conferred meanmes. With the growth of technological sophistication, there has been a long-term shift from 'primary' to 'added' value as the principal determinant of 'wealth.'

9 Other systems of nomenclature, such as those describing economic or power structures within urban networks (e.g. Wallerstein's original 'core' and 'periphery'), have described a further level of differentiation, and it may be confusing to use the same terms in different senses. The zonality described here is ihe most fundamental, and it may be preferable to elaborate other terms (e.g. hegemony for the complex and rapidly changing patterns of global dominance.