State formation and empire building

JOHANN P. ARNASON

This chapter focuses on states and empires, and more precisely on imperial trends and turns in processes of state formation, political developments that lend support to the idea that the mid-first to mid-second millennium was a distinct period.

That period coincides roughly with the conventional definition of the Middle Ages, a periodization that is often dismissed as Eurocentric, unduly levelling, and laden with value judgments that now seem unfair to the long epoch in question. This periodization also has a history of coming back, however, reinvigorated by new evidence from outside its original context, although those who want to avoid ‘medieval' connotations may prefer to speak of the Middle Millennium or of an ‘intermediate age'.^{^[618]} Views on the events, episodes and restructurings that mark the beginning and the end of this period depend on thematic choices; the approximate dates to be preferred will therefore vary from one regional setting to another. As will be seen, the main patterns at work in imperial turns in state formation intertwine with religious cultures and institutions.

The political landmarks defining the Middle Millennium

To begin with Western Christendom, the inaugural chain of events links the fifth and the sixth centuries ce. The decomposition of the Roman Empire was followed by the rise of two Germanic kingdoms aspiring to great power status: the Ostrogoths in Italy and the Franks in Gaul. The destruction of the Ostrogoths by Justinian's counter-offensive did not lead to imperial reunification, and the definitive failure of that project, together with the survival and strengthening of the papacy, set the scene for further geopolitical developments. As for the final episode, the grand but failed strategic project pursued by Charles V is the most salient landmark.

This was an attempt to put the wealth and prestige derived from trans-oceanic conquest at the service of older aspirations to imperial supremacy in Europe; its failure and the consequent change to geopolitical conditions make the mid-sixteenth century an obvious turning point.

A slightly different chronology seems required for the Byzantine world. Here we can begin with the last decades of the sixth century. The critical state of the empire after Justinian's bid to reconquer the West, aggravated by plague and culminating in a struggle for survival against threats from north and east, may be seen as the beginning of the transition from an East Roman to a Byzantine Empire; this process then took a more decisive turn when Islamic expansion changed the regional balance of power. The Byzantine trajectory came to an end with two major but contrasting geopolitical shifts. The post-imperial state system that had - with active involvement of Western invaders and rivals from within the Byzantine world - taken shape in southeastern Europe after 1204 was absorbed by the Ottoman Empire, whose background reflects the last great wave of Inner Eurasian expansion. The Russian periphery, long subject to an Inner Eurasian power, gained the upper hand against its former overlords and embarked on a multi-secular expansion across northern and Inner Eurasia.

With regard to the Islamic world, there can be no doubt about the point of departure. But the emergence of Islam is not to be understood as an abrupt leap into a new religious universe. The sixth-century geopolitical upheaval in the Arabian peninsula, in which Sasanian Persia participated directly and the East Roman Empire indirectly through its Axumite allies, is an important part of the story, and so are the troubles preceding the late seventh-century consolidation of an Islamic realm. At the other end of our period, the formation of the early modern Islamic empires (Ottoman, Safavid and Mughal) represents a major historical divide; after the ninth- and tenthcentury fragmentation of the ‘Abbasid Empire, this was the first stable reorganization of the Islamic world along imperial lines.

The foundation of the Mughal Empire in 1526 may therefore be seen as the beginning of a new epoch. This chronological framing also has an obvious bearing on Indian history. In the South Asian subcontinent, the sixth-century downfall of the Gupta dynasty is the most decisive mid-first millennium landmark. But soon afterwards, at the very beginning of the eighth century, Islamic expansion gained its first foothold in India with the conquest of Sindh. There are thus good grounds to regard Islamic expansion as a co-determinant of Indian history during most of the period in question, unfolding through successive stages and taking its most decisive turn with the foundation of the Mughal Empire.

China and the broader East Asian region pose specific problems. It is relatively easy to construct an opening phase. The late sixth-century reunification of the Chinese Empire, carried out by the Sui dynasty and consolidated by their Tang successors, was accompanied by the rise of unified states in Korea and Japan. The mid-second millennium is a more difficult case. Historians of China often work with the model of a late imperial period beginning in the mid-sixteenth century; the middle imperial stage is most frequently equated with the Tang dynasty and the ascendant phase of Song rule (before the loss of North China). The idea of a Song-Yuan-Ming transition, from the early twelfth to the late fifteenth century, has been used to bridge the gap between the two periods. From the present point of view, focusing on changing imperial regimes, there is much to be said for a periodization stressing the Manchu takeover in 1644 and the establishment of the Qing dynasty.

As for other parts of the East Asian regional core, the rise of the Qing roughly coincided with changes that are now generally seen as the beginning of an early modern epoch in Japanese history: unification under the Tokugawa regime (1603-1868) and withdrawal from interstate relations. There are both contrasts and parallels between the Tokugawa settlement and the seventh- to early eighth-century developments that brought core regions of the archipelago under the control of a reformed central state.

In both cases, intensified reception of Chinese cultural models went hand in hand with strategic withdrawal; the main contrast is that the state created in the wake of the mid-seventh-century Taika reform drew on Chinese institutional patterns, whereas the Tokugawa regime grew out of more indigenous trends. Korean history calls for a different approach; here the main landmark of the second millennium was a late fourteenth-century combination of dynastic change, political restructuring and ideological reorientation, often described as a Neo-Confucian transformation. The chronological discrepancy is, however, less important than the lasting results. In East Asia, the period in question ends with a regional bifurcation that was to have a major impact on the roads to modernity. On the one hand, a new wave of Chinese cultural influence on the outlying countries (the Neo-Confucian turn) coincided with a definitive parting of political ways. Japan abandoned all political contact with China and all political ambitions on the mainland, and after the Mongol interlude, no Chinese dynasty exercised direct control over Korea. On the other hand, the Qing conquests brought China into political union with an Inner Eurasian periphery where its cultural reach was much more limited.

It is more difficult to draw chronological boundaries for Southeast Asian history; but it seems reasonably clear that the sixth century saw the decline and disintegration of the first known powerful political centre in the region (the state known from Chinese sources as Fu-nan), and that both its demise and its legacy were important for later developments. A good case for a closure and a new beginning in the mid-second millennium can be made on the basis of Anthony Reid's work.^{^[619]} As he sees it, the modern phase of Southeast Asian history begins with the fifteenth-century resurgence of Islamic expansion and the durable division of the region between a predominantly Islamic archipelago (including the Malay peninsula) and a predominantly Theravada Buddhist mainland.

Finally, the Middle Millennium history of Inner Eurasia is very clearly demarcated from earlier and later stages. The sixth-century rise of the Turkic Empire (khaqanate) brought the eastern and western parts of this region into closer contact than before. The Turkic imperial formation was the first to extend across Inner Eurasia and leave a cluster of successor states - from the Avars in Europe and the Khazars on the South Russian steppe to the Tibetan and Uighur empires on the eastern frontier. The era of large empires and lesser successors came to an end with the very transformations already noted in the East Asian and Russian context.

The focus of this chapter is the Eurasian macro-region (including the southern shores of the Mediterranean), but a brief comment on other parts of the world is in order. The history of state formation in sub-Saharan Africa is a long and variegated one, and includes cases of rapid rise and fall as well as more sustained development. This vast regional field and its internal divergences are beyond the scope of our discussion. But when it comes to the more specific question of imperial trends and traditions in Africa, connections to the broader Eurasian context are more important, and the main ones have to do with the dynamics of Islamic expansion. In the region later known as the Horn of Africa, the Axumite kingdom had by the sixth century grown into an imperial centre that ‘united under its rule... all the civilizations to the south of the Roman Empire, between the Sahara in the west and the Rub' el-Khali desert in the east'.^{^[620]} Among the three imperial formations surrounding the birthplace of Islam, it was most directly involved in the peninsular upheaval that preceded Muhammad's mission, and the Islamic impact on its history differed from the two other cases. Sasanian Persia was conquered; Byzantium survived as a diminished but genuine empire; Axum was marginalized without direct confrontation, but a shrunken imperial tradition - Christianized roughly at the same time as the Roman Empire - seems to have survived and left its mark on the later trajectory of the Ethiopian state under successive dynasties.

At the other end of the period, the failed Islamic assault on Ethiopia in the first half of the sixteenth century sealed geopolitical and geo-religious divisions for a long time to come. In West Africa, the ramifications of Islamic expansion were very different, but important to the course of state formation. Here the main example of regional power centres with imperial reach is the sequence of three formations with core areas gradually shifting from west to east: Ghana, Mali and Songhay. The beginnings of this story are obscure, and no convincing reconstruction has gone back to a mid-first millennium date. The terminating event is closer to our time frame: the late sixteenth-century Moroccan conquest of Songhay (a more visibly imperial power than its predecessors) changed the geopolitical profile of West Africa.

As for the Americas, the sixteenth-century destruction of Mesoamerican and Andean states and civilizations by Spanish conquerors was part of a broader geopolitical upheaval at the end of the Middle Millennium. It is more difficult to find parallels to the beginnings of our period. The most significant historical divides before the conquest have to do with collapses of political and cultural centres, not always markedly imperial in character: Teotihuacan and the classical Maya city-states in Mesoamerica during the second half of the first millennium, Wari and Tiahuanaco in the Andean zone around and after 1000.

Interactions between Inner Eurasia and adjacent civilizations

Interaction between inner and outer parts of Eurasia was of particular importance for the history of states and empires during the Middle Millennium. It is now generally agreed that the interactive process was too complex and too formative on both sides to be reduced to being understood as outbreaks of barbarian aggression. Specific aspects of the relationship between the two historical complexes are more disputed. This problem begins with varying labels for historical and geographical regions; and here I make a distinction between Inner Eurasia, broadly defined, and a more narrowly demarcated Central Asian part of that region. The Inner Eurasian region is to be identified with the whole area between the Siberian Arctic and the shifting borders of settled civilizations in the European, Iranian, Indian and Chinese worlds; in terms of west-to-east, it reaches from the Russian steppe to Manchuria. Central Asia proper is a zone of deserts, oases, mountains and valleys. The typical political pattern in the larger part of this region was the oasis-state with long-distance trade connections. There was, however, one case of a larger Central Asian state flourishing from the early seventh to the middle of the ninth century: the Tibetan Empire. It was based in the agricultural south, but drew on the historical experience of empires with a nomadic pedigree.^{^[621]} The conversion to Buddhism may be seen in that light: as a strategic and limited change of religious orientations, comparable to earlier cases of conquest dynasties in Northern China. But if the initial conversion was part of an empire building strategy, it also paved the way for a post-imperial re-appropriation of Buddhism more original than anything developed elsewhere in Central Asia, and capable of spreading further afield. This indirect legacy of the Tibetan Empire has overshadowed its geopolitical record.

However, the main impetus to empire building across the great civilizational divide came from further north. Popular notions of ‘nomad conquerors' may suggest a simple equation of nomadism and Inner Eurasian dynamics. In fact, the interaction of nomadism with agricultural and urban modes of settled social life is a precondition, not just a result of imperial expansion. These interconnections developed through several levels. In the first place, the Inner Eurasian world is best understood as an ‘entire spectrum from intensive cultivation to strict steppe pastoralism',^{^[622]} and the continuum between these extremes allowed for varying balances. Not all empirebuilders were primarily nomads. At more advanced stages of state formation,

the incorporation of urban-based societies in Central Asia was crucial to fUrther growth of nomad-ruled empires; the role of the Uighurs in the Mongol Empire is a classic case. Finally, the economic dependence of nomads on the more productive economies of sedentary neighbours did not necessarily lead the former to embark on conquest; but when it did, and when the conquerors maintained their rule without assimilating to domestic elites, the resulting combination could become a strategic basis for new imperial projects. It was the Mongol takeover of such innovations in northern China, as well as the very different symbiosis between nomadic and settled cultures in eastern and western Turkestan, that provided foundations and incentives for the building of an unprecedentedly cross-Eurasian empire.

There are nevertheless good reasons to stress a strategic role of nomadism within a more complex configuration. Peter B. Golden argues that ‘given their tribal organization, continual training for war and the executive talents needed to move herds and people some distance, the state was latent in most Eurasian nomadic societies'.^{^[623]} This generalized disposition to state formation could be activated by a variety of factors. But a more unitary view is suggested by Nicola Di Cosmo: the nomadic state is typically a ‘social response to a state of crisis' and to the accompanying ‘endemic, low-level violence'.^{^[624]} The particular crisis-proneness of nomadic societies is due to ecological and geographical conditions: the vulnerability to environmental change, even on a small scale, and the high mobility of ethnic groupings, often forced to compete for scarce resources, led to periodic destabilization. Not that the emergence of state structures was an automatic response - the history of steppe societies contains episodes of long-drawn-out disintegration as well as cases of durable organization halfway between tribe and state. It is a basic fact about Inner Eurasian history that states were more likely to form in the eastern than in the western part of the steppe, and that invaders or refugees from the east often set the course in the west. But state formation is frequent enough for recurrent basic mechanisms to be identified. Two internal transformations were decisive. Social stratification gave rise to a hierarchy of tribal groupings and identities. This process was combined with - and counterbalanced by - the decimal military organization, in which armies were arranged into inter-ethnic units of ten and commanded to be loyal to all, regardless of tribe or ethnic group. The decimal military system seems to have been invented at an early stage and transmitted with remarkable continuity to later nomad states. But since these changes imposed further demands on a fragile resource basis, they only prefigured a more ambitious solution to the problem.

The process that led to the formation of states and empires with nomads in key roles was no mere reaction under pressure; it had a cultural framework, continuity and vision of its own. In this regard, the Turkic Empire that stands at the beginning of our epoch was - despite its brief duration - a paradigmatic achievement. It entered the cultural memory of later states and shaped their self-understanding. The Turkic rulers had not attempted any large-scale conquest of sedentary regions, but their claim to supreme dignity, inherited by later contenders, appears to have been easily translatable into visions of universal domination, most effectively in the case of the Mongols. This imperial ideology was rooted in a religious imaginary well attested for the imperial Turks, and clearly of older origin; the core notion was a celestial pedigree of royal power, defined in terms that indicate a Mandate of Heaven.

The ideological continuity noted above was compatible with a wide variety of strategies applied in relations with sedentary societies and powers. Inner Eurasian states were, even more than the tribal groupings from which they emerged, dependent on resources provided by more complex economies in southern and eastern Eurasia. Recent scholarship distinguishes four solutions to the problem: raiding, trade, tribute, and territorial conquest followed by direct taxation. If the accumulation of military means and the extraction of resources are key factors in state formation, the Inner Eurasian pattern was thus marked by a problematic relationship between them: internal conditions enabled rapid and spiralling militarization (through ongoing incorporation of tribes), but the resources needed to sustain this process had to come from elsewhere. In favourable circumstances, the dynamic thus generated could lead to self-perpetuating expansion. Whether the history of states built on this basis reveals cumulative or evolutionary trends is a more complex question. A glance at the overall trajectory suggests not so much a cumulative trend as a renewable but, in the very long run, selfcancelling pattern. The Turkic Empire did not expand into sedentary zones, but its reach from east to west and its control over trade zones were unprecedented. Its successor states on the eastern and western sides of Inner Eurasia varied widely in terms of strength, structure and relations with neighbouring civilizations. Conversions to marginal or even persecuted religions from the sedentary world (Judaism among the Khazars, Manichaeism among the Uighurs) represent the highest point of a cultural demarcation strategy that was not adopted by later arrivals in the borderlands of Inner and Outer Eurasia. But although these political formations retained links to the Turkic imagery and nomenclature of power, none of them could aspire to imperial rule on a corresponding scale. The second wave of imperial expansion across Inner Eurasia (and this time beyond its borders on all sides) came with the Mongol conquests.

There is some justification for seeing the first stages in the rise of the Mongols as shifts in the balance of power within a China-centred constellation. The marginal society within which Chinggis Khan rose to power appears to have been a decomposing former tributary of the Jin dynasty that had ousted the Liao from power in North China in 1125 and inherited from them a distinctive mode of rule, known to historians as dual administration. It is worth noting that this innovative technique of expansion - maintaining the separate identity of the conquerors and their ultimate control over territories and resources, while making use of indigenous governmental traditions - was first developed by the Liao outside China proper, when in 926 they conquered Bohai, a Korean kingdom beyond the borders of the unified Korean peninsula. The whole process can thus be seen as a dynamic unfolding between a Chinese regional order and its Inner Eurasian neighbours, and Chinggis Khan's enterprise as a new beginning favoured by particularly unsettled conditions on a distinct but potentially strategic periphery. As the Mongol expansion gathered momentum, it resulted in conquests of sedentary regions far beyond all earlier limits. The thirteenth-century Mongol Empire is often described as the largest contiguous one in human history, but it should also be emphasized that this phase of Mongol power - an east-to-west Eurasian realm ruled from a single centre - was very short-lived. The conquest of Iran and adjacent regions almost coincided with an intra-dynastic conflict that marked the beginning of the end for imperial unity; the most enriching and empowering conquest (that of Song China, completed in 1279) was carried out by a North China branch of the empire, with only symbolic claims to universal sovereignty.

This is not to suggest that the Mongol Empire began to fall apart as a result of overstretch. It seems more likely that the extremely effective powermaximizing strategies of the Mongols could have sustained further growth. The centralized military machine, the eclectic but highly extractive taxing methods, and the systematic gathering and redistribution of ‘human talent and skill as a form of booty'^{^[625]} added up to an unequalled supra-regional dominance. The once widely accepted view that the Mongols imposed a universal law code (yasa) has been questioned; but the alternative model of context-dependent modifications of existing legal traditions underlines the adaptive strengths of the regime. The halting of the Mongol advance in various places at successive junctures is probably to be explained in specific terms for each case. But it can be argued that some distinctive features of Mongol empire building constituted both strengths and weaknesses, and that their mutually reinforcing impact also led to disintegrative pressures. Chinggis Khan's ascent to unchallenged power gave rise to an extreme version of dynastic charisma, strong enough to survive the empire as well as the first generation of successor states and leave its mark on the later history of the Inner Eurasian heartland; but this incentive to continuity proved inseparable from the regional tradition of intra-dynastic struggle as a test of legitimacy. On all fronts of expansion, the Mongols also drew on more specific methods and experiences of nomadic or nomad-dominated precursors. Seen as a whole, this pattern represents a synthesis of previously separate developments, rather than a culmination of general trends. But the combination also posed new problems. The adoption of diverse precedents led to adaptive processes that pulled the main regions of the empire further apart. Finally, the mixing of civilizational traditions, especially in administrative affairs, reflects the reach of the empire and the pragmatism of its rulers, but it also provoked resistances that fed into the decomposing process; Chinese objections to Muslim influence, and vice versa in West Asia, exemplify this point. Given all these background factors, it is not surprising that the division of the empire into separate states ‘was almost as bloody and as tortuous a business as was the expansion of the empire itself for those powers and peoples that were its victims'.^{^[626]} The ‘largest contiguous empire in history' was a fleeting formation, unduly magnified and sometimes romanticized by later historians.

The sequel to the Mongol Empire is best understood as a three- or four- generation series of successor states, with some regional variations. The first separate realms were domains (ulus) of imperial family members, hardly to be described as states, but gradually transformed into more clearly demarcated entities under dynastic rule and with power bases in pre-existing regions (China, Central Asia, Iran and the western steppe). A third phase saw the further partition of these imperial fragments into shifting mixtures of dynastic centres and tribal formations reverting to earlier ways. Timur's vast raids, perhaps best described as ‘plundering tourism',^{^[627]} ^{^[628]} were the most destructive episode of this intermediate stage. Their spectacular character has led to misconceptions of a culminating Inner Eurasian empire; in fact, they did not result in state-building on an imperial scale, and closer analysis of Timur's strategy¹¹ suggests a balancing act between tribal basis and despotic conquest. The only long-term upshot was a charismatic aura without institutional anchor, transferable to those who could claim a genealogical connection and embark on other ventures elsewhere, most notably in India.

State, empire and religion in Outer Eurasia

A conception of state formation as a long-term process, indeed a set of such processes, is needed to make sense of connections between early medieval backgrounds and early modern outcomes. The most seminal argument of that kind is to be found in Norbert Elias' work on the civilizing process as a long-term dynamic with durable trends and cumulative consequences, and constellations that impose their logic on historical actors and channel their projects in specific directions. This explanatory framework highlights the twin monopolies of violence and taxation as links between the strategic and structural levels in state formation.^{^[629]} To summarize a complex discussion, three main themes have emerged as necessary correctives to Elias' approach. The twin monopolies have been subsumed under the more general and interconnected imperatives of control and mobilization, applying to human and non-human resources. More attention must be paid to contextual factors, affecting the long-term processes in significant ways. To mention only European examples, the Eliasian model does not do justice to the role of the Catholic Church, nor to the reactivation of Roman law. Finally, the cultural frameworks of state formation call for closer analysis.

With these perspectives in mind, a few basic points about the imperial versions of state formation should be added before discussing particular cases. Given the enormous variety of imperial regimes, and the unsatisfactory state of comparative research, we cannot begin with a general definition of empire as a category. It seems more useful to treat imperial turns as specific episodes within processes of state formation, and to allow for a range of patterns and developments that tend to accompany such turns in widely varying degrees and combinations. Imperial regimes are thus linked through ‘family resemblances', rather than constant defining features. The starting point for comparative analysis would then be the very transition to imperial rule, i.e. the attempt to unite separate processes of state formation - or their outcomes - under a single centre. A recurrent but unequally developed trend leads to the transfiguration of imperial power into a higher order; in the most pronounced cases, this culminates in visions of world domination. Higher- order interpretations of power give rise to institutionally and ideologically eminent centres and correspondingly elevated notions of rulership. How the imaginary meaning of ultimate sovereignty translates into practical terms is a more complex issue. Imperial power favours autocratic rule, but other factors may turn this form into a facade for more oligarchic regimes.

With the above comments on states and empires in mind, the main paths of outer Eurasian history during our period should now be compared in broad outlines, with due allowance for the fact that here the interrelations of religious and political dynamics were more significant than on the Inner Eurasian side. If we begin with the Western Christian world, it is certainly not being suggested that this region was central to the whole scene. The key point is, rather, the long-drawn-out transition, from a fragmented and marginal condition to the early modern field of consolidated states in quest of empire, that opened a new chapter of history. Also, as analyses and theories of state formation have mainly drawn on the western European experience, there are valid reasons to deal first with that part of the world.

Western Christendom

For much of our period, the western Eurasian periphery - in civilizational terms, Western Christendom - was geopolitically marginal, highly fragmented, and limited to expansion on separate and narrowly circumscribed frontiers. But around the middle of the second millennium, a small group of states emerged from the long-fragmented European field and embarked on overseas conquest. They included the unified Spanish state, the economic and cultural powerhouse of the Netherlands, as well as the English and French monarchies. But a historical interpretation of early modern European ascendancy must take note of preconditions emerging at much earlier stages.

Such aspects include political factors, and recent scholarship tends to put the state back into times and places where earlier approaches had tended to dissolve it into networks of ritualized interpersonal relations.^{^[630]}

A glance at the history of medieval Western Christendom will highlight not only political fragmentation, but also the qualitative diversity of political centres. This civilization was not durably dominated by an imperial centre, yet a ruler and a realm with claims to imperial dignity were, for most of the time, present as key components of the civilizational complex. The emperor as an institution was a model of supreme rulership, intermittently translated into aspirations to more effective and extensive power, but also imitated by lower-ranking rulers who wanted to enhance their legitimacy; the empire as a geopolitical entity brought together more diverse communities than other states of the times, but this was a double-edged asset that could both serve to elevate the centre and - in the long run more decisively - to enable the creation of autonomous power structures at lower levels. Within Western Christendom, the position of the empire was defined by its relationship to the papacy. The latter was the civilizational institution par excellence, a potential rival to the empire - on the basis of claims to ultimate authority over Christian rulers - and an active player in political conflicts and alliances on multiple levels. But in the very long run, the most sustainable setting for state formation was a changing number of kingdoms with shifting territorial boundaries, typically defined with some reference to ethnic identity and stabilized - in varying degrees - by dynastic continuity. Finally, the urban transformation that gathered momentum from the eleventh century onwards gave rise to autonomous legislation, self-government and a republican political tradition. Whether the quest for autonomy resulted in a revolutionary challenge to authorities, and whether self-government came to involve a claim to sovereignty, are more specific issues that should not be conflated with the general question of the city as a political innovator. These urban communities represented an alternative line of state formation; they flourished within a geopolitical setting that blocked the road from city-state to empire, with one limited exception: the expansion of Venice.

These various forms of political power were also different versions of the religio-political nexus. The conflict between empire and papacy was not simply a dash of secular and sacred authorities. A mixture of the two was invoked by both sides, and the relative weights changed over time: eleventh-century emperors made more emphatic claims to sacral rulership than their successors, and the twelfth-century successes of the papacy markedly increased its influence on the level of power politics. As for the urban communities, the religious background to their history manifested itself in multiple ways. Urban centres shrank during the transition from antiquity to medieval times, but without a basic continuity the whole subsequent pattern of growth would have been unthinkable. The Church, and particularly the position of bishops as holders of social power, gave institutional support to this enabling legacy. The High Medieval breakthrough to urban autonomy was, in significant measure, achieved against local Church powers and related to new currents in religious culture. The eleventh-century rise of the northern Italian communes - the most radical version of a civilization-wide, though not uniform, change - is instructive in both regards. The devolution of Carolingian power in Italy had strengthened episcopal authority in the cities; the communes established themselves in conflict with this offshoot of the weakening alliance between Church and Empire, and more generally as a rejection of ‘aristocratic-oligarchic reality', to which they opposed the ‘idea of brotherly unity'. In short, ‘the foundation of the commune is, in the last instance, a religious idea'.^{^[631]}

The geopolitical patterns evolving on this complex basis underwent major changes. If we regard the post-Roman centuries of fragmentation and fluctuating hegemonies as a transitional stage, the formative turn was the rise of Carolingian power and its imperial expansion. The conquests and the close relationship with the papacy made it possible to revive the idea of imperial rulership and claim equality with the surviving Byzantine Empire. Empire and Church were intertwined on ideological, institutional and administrative levels; efforts to develop cultural centres and resources were attuned to this power structure, and the intellectual revival, although limited to a narrow circle, was important enough to justify the notion of a Carolingian Renaissance as the first of such transformative turns in the medieval West. For all these reasons, the Carolingian Empire deserves to be described as a civilizational matrix, even though it did not incorporate the whole area of Western Christendom. In the shorter run, however, it gave way to a more diverse and fragmented pattern. The Ottonian revival and relocation of empire in 962 was a landmark. The imperial institution was transferred to a frontier of the Carolingian world and superimposed on pre-existing power centres and ethnic identities in formation, over which it had only limited and precarious control. The efforts to strengthen the Roman - and more generally Italian - connection were meant to compensate for these weaknesses. But if the Holy Roman Empire was to a high degree an imaginary institution, this does not mean that its historical impact was minor or illusory. Visions of statehood and/or empire enter into the pursuit of power and the resultant long-term processes; in this particular case, the presence of the empire influenced various paths of state formation. The imperial institution as such was capable of reforms that consolidated state structures, and the need to accommodate multiple centres gave a constitutional twist to this trend; such projects became more ambitious towards the end of our epoch. Territorial units within the empire moved towards statehood, both in connection and in rivalry with the centre. Distinctive political developments on the periphery of the empire led to lasting results (for example in the territories that later became Switzerland and the Netherlands).

During the later half of the Middle Millennium, new political centres proliferated through civilizational and military expansion. But this was also the age of pioneering state-strengthening measures in key centres of the older civilizational core, notably the English and French kingdoms, and of emerging city-state cultures in rapidly developing regions, especially northern Italy and the northwestern corner of the continent. The pluralism of political centres thus reached its highest level; so did the divergence of regional trajectories, and that aspect became more salient during the last centuries of our period. The ‘new monarchies' of the Atlantic seaboard became exemplary cases for analyses of state formation, but they also gave rise to a somewhat misleading perspective. England, France and the Spanish kingdom created by the dynastic union of Castile and Aragon represented new levels of control and mobilization, and their experience came to be seen as a decisive turn towards the modern conversion of monarchies into nationstates. This supposedly paradigmatic path of state formation obscured the variety of historical trajectories, and in particular the continuing presence of empires. The type of statehood represented by the new monarchies was, on the larger European and incipiently global scene, intertwined with imperial trends in two different ways. On the one hand, the seaboard kingdoms became pioneers of trans-oceanic empire building. The united Spanish state - the most traditional and, in the longer run, least transformative of the new monarchies - was the first to acquire major overseas possessions. On the other hand, composite states were a key element oflate medieval geopolitics, and this trend sometimes took imperial turns. The composite state may be seen as a more stabilized version of purely dynastic unions, and was in some cases accompanied by dynastic change. It is common in late medieval Europe (Portugal is the main exception), but more dominant in some regions than elsewhere. Even the new monarchies had composite backgrounds. France is an outstanding case of a centralizing monarchy gradually prevailing over the component principalities; the fourteenth- and fifteenth-century career of the Burgundian state exemplifies the chances then still open to the principalities. England was in part a composite state, but the late medieval and early modern push to strengthen the centre kept this pluralistic aspect in check. Spain was the most durably composite of the key states, but this was for some time counterbalanced by a succession of strong rulers and by gains from conquest. Elsewhere the composite character of state formation was more pronounced. The Kalmar Union (1389-1523) that brought the Nordic kingdoms under one ruler was an exemplary composite state, and even after its downfall the Danish kingdom retained this character. Another prominent case is the Polish-Lithuanian commonwealth, whose ruling dynasty briefly came to power in other central European states, but without personal union; here it seems justified to speak of an uncompleted imperial upgrading. Finally, the Holy Roman Empire restructured itself as a more formally institutionalized composite state, with the specific feature of a centre endowed with higher dignity but confined to more limited political options than the rulers of the rising monarchies.

East Asia

Comparisons with the Western Christian constellation are easier to envisage in some regions than in others, and are particularly revealing in the Chinese case. Chinese institutions, cultural orientations and geopolitics were - to a higher degree than those of any other civilization during the Middle Millennium - embedded in older patterns. In the first place, state formation in China had a head start on other regions. Rival centres competing for hegemony raised their capacities of control and mobilization; the result was a unified imperial regime, durably identified with the Han dynasty (206 BCE-220 ce). The core institution was a very strong version of sacral rulership, operating through a close integration of ritual, political and administrative practices. The staying power of this sacral-imperial centre prevented developments comparable to Western Christian institutional and territorial division. But the model that endured was also a source of variations within its framework. The imperial centre, relying on its own sacral legitimacy, could emphasize and privilege different parts of a composite tradition which also came to include Buddhism. On the other hand, the sovereignty of the emperor, enhanced by his autonomously sacral status, could serve to impose autocratic rule.

Alongside the cultural legacy and the imperial model, a third aspect of the Chinese pattern also emerged during the formative phase that preceded our period. As noted above, the beginnings of Inner Eurasian state formation on imperial scale go back to the early Han period, and smaller conquest states with a more sedentary social background came to the fore in the aftermath of Han decline. Both kinds of frontier formations became permanent presences in Chinese history, but in the first phase of the Middle Millennium their role was more marginal than later. The Turkic origins of the dynasties that united and consolidated the empire, Sui (581-618) and Tang (618-907), did not translate into any institutional fusion or combination of traditions. The Tang dynasty built on Sui foundations and stayed in power long enough to develop more variations within the received model. Religious pluralism and regional influence have often led historians to describe the Tang Empire as the most cosmopolitan in Chinese history, but there are good reasons to emphasize another side. The application of Chinese models to autonomous state-building in Korea and Japan during the early Tang era completed the formation of an East Asian region, made up of three main political centres within a cultural koine. On the other hand, Tang expansion did not result in an enlarged empire. Korea resisted conquest, and Japan was never within reach; on the continental side, the Tibetan Empire became a serious threat, and warfare in the Turkic zone ended in a stalemate conducive to fragmentation on both sides.

The three dynastic regimes following the tenth-century fragmentation after the Tang collapse - Song, Yuan and Ming - cover more than the second half of our period; they are crucial for China's place in world history, not least in relation to Inner Eurasian empires and their trans-regional dynamic. This was an intermediate phase between the Tang empire, widely regarded by later generations as a paradigm of cultural and political achievement, and the late imperial stage that began with the Manchu conquest in 1644. From one point of view the whole period appears as a long-drawn-out withdrawal (not without backlashes) of the imperial state from direct involvement in social life, accompanied by urban growth and a progressive empowerment of local elites. On this view, the eleventh- and early twelfth-century contest between statist reformers and advocates of a more decentralized regime was a turning point. The centralizing drive led by Wang Anshi (1021-86) was closely linked to plans for a military offensive against northern neighbours seen as usurpers of imperial territory and threats to Song power. The outcome appears as an irreversible defeat on both fronts: Inner China was divided between Song emperors and conquest dynasties from the north, later followed by Mongol rule over the whole realm (and thus by a reconstitution of the imperial domain that had previously been divided between a Chinese core and peripheral states with expansionist ambitions), and on the internal side, social power shifted from the imperial centre to a broadly based and regionally diversified gentry that combined access to offices with landed property and local authority. The focus is thus on connections between a changing geopolitical situation in continental East Asia, a resumption of state formation on an imperial scale, and a redistribution of social power. This is a useful reminder of broader contexts. But in the long run and the overall picture, the resilience of the imperial state seems more important than the restructuring of its social basis. The ascendancy of the gentry under the Song can also be seen from this angle, and in contrast to aristocratic power during the Tang era: from the beginning of the second millennium to the nineteenth century, ‘there were no challenges to the emperor by elites with independent bases of authority’.^{^[632]} Access to office was crucial to the position of the gentry, and this institutional link to the imperial centre was reinforced by ideological changes. Neo-Confucianism, gradually established as a dominant school of thought, was not simply an ideology of gentry power; it also served to integrate the latter into the imperial order, and in that capacity, it may be seen as an adaptation of the religio-political nexus. Moreover, the whole regime of education and indoctrination, centred on Neo-Confucianism, represents ‘efforts by a state to influence belief and behaviour patterns of the general population well before such activities were imagined, let alone pursued, in Europe’.^{^[633]} This highly developed cultural framework could compensate for shortcomings of state structures, but it was also conducive to underestimation and neglect of structural problems that became acute when the empire faced external challenges. The Ming dynasty that ousted the Mongols in 1368 inherited their reunified realm, but its founder also attempted to restore a comprehensive Chinese order with strongly nativist overtones, a shifting preference for particular traditions within the Chinese civilizational complex, and an extremely autocratic model of rulership. The restoration seems to have been marked by a lasting dissonance between cultural-political patterns and economic development, and this flaw helps to explain the seventeenth-century collapse. The mobilizing capacity of the imperial centre was tried and found wanting.

To conclude, a comparison of Western Christendom and China will not simply contrast a failure of empire in the former case with its continuous presence in the latter. Imperial ventures and aspirations were also involved on the Western side; but the staying power, the cultural reach and the exclusive status of the Chinese imperial centre had no Western parallel. There was no Chinese analogy to the Western split between empire and papacy, but a core conflictual relationship of another kind affected the course of events: the mutually formative rivalry of Chinese and Inner Eurasian claimants to empire. A further feature of the East Asian scene is the transfer of the Chinese imperial model to outlying countries (Korea, Japan, Vietnam) where it became a framework for state formation on a significant scale but without any impact on the geopolitics of the Chinese heartland and its northern frontier. No such separation of centre and periphery occurred in the West.

India and Southeast Asia

Certain aspects of the Indian experience suggest affinities with China. In both cases, Inner Eurasian invasions were a defining feature of our period, whereas their impact on Western Christendom was minimal. But the Indian conquest dynasties differed from the Chinese ones in that they were converts to a universal religion and heirs to its empire building tradition, and their impact on the subcontinent was correspondingly more radical. Further contrasts and parallels have to do with the spread of Chinese and Indian models in East and Southeast Asia. A common feature of the two regional configurations was the transfer of state-building strategies, traditions and ideologies from centres to peripheries, without military conquest, and sometimes - as in seventh- and eighth-century Japan vis-a-vis China - with a view to mimetic rivalry. But there were also major differences. Korean and Japanese adaptations of Chinese models resulted in a remarkably stable geopolitical constellation of one major and two minor centres, modified only by intermittent waves of Inner Eurasian conquests. Southeast Asia saw the emergence, flowering and fall of numerous states, often loosely demarcated and sometimes barely present in the historical record. Despite these divergences, the two regional examples of derivative statehood share a contrast with European trends. State formation along the lines of borrowed models was certainly not unknown in medieval Western Christendom, but in the upshot, military conquest played a much greater role in the medieval making of Europe than in the corresponding East and Southeast Asian processes.

A closer analysis should link developments on the Indian subcontinent to the record of insular and continental Southeast Asia. This expanded cultural orbit is best approached from a vantage-point often taken to mark the end of classical Indian civilization, but now more plausibly seen as the onset of a new phase. In the subcontinent, the beginning of the Middle Millennium was an imperial aftermath; this resembles the situation of Western Christendom, but differs from imperial renewal in China and imperial succession in the Islamic and Byzantine worlds. The Gupta dynasty, collapsing in the early sixth century under the pressure of internal tensions and invaders from the northwest, was the last indigenous contender for pan-Indian empire. But it also developed a pattern of statehood that could be adopted by later rulers, albeit on a smaller scale and in mutual contest. Variously described as concentric, segmentary or galactic, this type of regime combined a domain of central power with less effective control over local rulers and subaltern allies. Control over trade routes was essential, and so was the ritual integration of the rulers' domains achieved through cooperation with priests and temples; the historical alliance of rulers and Brahmins was perfected in that context. But this form of the religio-political nexus also allowed for mutual interference and rivalry. The deification of kings combined with the refiguration of gods as kings or sovereigns of the realm in question. Finally, this new paradigm of the post-Gupta state stresses a process that was repeated and replicated in successive phases and shifting regional settings (one of the characteristics of this period was the cultural and geopolitical integration of the whole subcontinent). Beginning with local lordships in pursuit of power and moving on to kingdoms with more expansive aims, the trend culminated in large regional states with imperial ambitions. None of them established the kind of dominance that the Gupta rulers at their strongest had enjoyed, but a sequence of pre-eminent powers can be reconstructed. While historians disagree on some cases, they seem to agree that the last one before the eleventh-century turning point (marked by a new wave of Islamic invasions) was the Chola kingdom based at the southern end of the subcontinent. The Cholas sent military expeditions to the Northern Indian plains as well as Southeast Asia, and entered into diplomatic relations with China.^{^[634]}

The Southeast Asian political formations that took shape in the course of a multi-secular process varied widely in time and space, and in individual cases, their strength and size seem to have fluctuated. On the whole, however, it seems clear that Southeast Asian patterns of rulership never shifted to the unconditional Brahmin primacy posited in the most ideological constructions of the Indian model, and where Buddhist religious culture predominated, it remained - during the first half of our period - more syncretic than the elite versions of Indian Buddhism. Both these historical features seem related to indigenous notions of sacral power and their continuing active role in shaping the course of Indianizing developments.

Given the overall context, it is not surprising that Southeast Asian trajectories of state formation have proved difficult to trace, and boundaries between minor or middle-range states more elusive than in most other places. During the first half of our period, two very different and unequally known cases stand out against a blurred background; both have been labelled as empires, although it is hard to pinpoint a transition to that level. In the western part of the Malay-Indonesian archipelagic zone, the kingdom of Srivijaya became a power centre with far-flung contacts and control over trade routes, but no continuous history of it can be written. Much better known is the Khmer kingdom on the mainland, whose beginnings can with reasonable certainty be dated to the early ninth century; this was, at least in its culminating twelfth-century phase, a genuinely imperial formation, and its basis was control over a key agricultural region. But in contrast to the East Asian region, no stable geopolitical pattern crystallized during the first half of our period. It was only during the second half that the outlines of such a configuration began to emerge, and this was also the time of the most significant extension of East Asian models to Southeast Asia: the rise of an expansionist Vietnamese state that continued to develop its own versions of Chinese institutions. Burmese and Tai states based in western and central reaches of the area emerged during the same period.¹⁸ Towards the end of the Middle Millennium, their trajectories culminated in large unified states on both sides; more specifically, the sixteenth-century expansion of the Burmese kingdom made it - for a short time - one of the most markedly imperial formations in Southeast Asian history. Burmese and Tai directions

Brajadulal Chattopadhyaya, ‘Political Processes and the Structure of Polity in Early Medieval India', in Kulke (ed.), State in India, 195-232; and Andre Wink, Al-Hind: The Making of the Indo-Islamic World (Leiden: Brill, 2003), vols. ι-ιιι.

18 To avoid confusion with the modern Thai nation, the term “Tai” refers to a cluster of peoples associated with earlier processes of state formation. of state formation were characterized by a new and more firmly structured relationship between religion and politics. The shift to Theravada Buddhism, now seen as a gradual process rather than an outright conversion, led to closer connections between the ruler, idealized in more distinctively Buddhist terms, and the monastic community; in a more general sense, it strengthened control and discipline.

The Islamic world

After the eleventh century, Islamic conquest dynasties became the dominant factor in Indian history, and this situation must be seen against the background of Islamic expansion as a long-term process. In the context of the Middle Millennium, Islam stands apart from other civilizations in several ways. As argued above, its emergence marks the beginning of our period; it represents the closest ever union of religious and imperial expansion; but it also exemplifies a uniquely thoroughgoing fusion of religion and politics, and reinterpreted memories of this origin could later be invoked as a corrective against the drifting apart of the two component forces. Yet no retrospective idealization could suppress a problem built into Islamic approaches to power and authority. The founder who embodied the perfect union of religious and political leadership was canonized as the last of the prophets, and hence irreplaceable; on the other hand, some kind of succession was needed to keep the community intact. The response to this dilemma was the institution of the caliphate, initially conceived as a strong version of sacral rulership. Whether the definition of the caliph as God's deputy was fully established is a matter of debate, but it could certainly be activated in times of crisis in response to internal conflicts and a later slowdown of expansion. On the other hand, it was vulnerable to erosion. The ruler could neither inherit nor re-establish the status of prophet, and this flaw prefigured further retreat. But the process was never linear: periodic efforts to reclaim the religious authority of the caliphate left their mark on Islamic history. At the same time, the succession problem was complicated by another divisive factor. Claims could be made on the basis of a kinship-based connection to the Prophet, or from a position of strength and authority within the community of his followers. That kind of dispute was the original trigger of the split between Sunnis and Shi'ites; but in a situation where the relationship between religion and politics had to be restructured, further constructions developed on both sides. When a strong and temporarily successful bid for dynastic rule by descendants of the Prophet was made (by the Isma'ilis and their Fatimid leaders in the tenth century), it was combined with a highly elaborate paradigm of spiritual authority (the imamate), complementary to prophecy rather than a substitute for it.

The Islamic trajectory thus began with a very distinctive form of the religio-political nexus. But this was a disputed field of issues and interpretations, not a consistent model, and it had to do with ways of dignifying and sacralizing power, rather than organizing it on the level of state structures or imperial domination. Further development, and more specifically the path from peninsular unity to all-round expansion, was shaped by the encounter with two imperial monarchies, Byzantium and Sasanian Persia, but the circumstances of these two cases differed, and so did the effects upon Islamic modes of rule. The Sasanian kingdom was obliterated by rapid conquest, but this initiated a long process of adapting Persian statecraft and experience to the successive phases of Islamic empire building and post-imperial state formation. The Persian connection was also crucial in another regard: this was where the problem of integrating non-Arab converts became most acute, and its repercussions unfolded on two levels. It was a major factor in the mideighth-century change of caliphal dynasty, often described as the ‘Abbasid revolution, and the ensuing restructuring of the empire; but in a more diffuse fashion, the interaction between Islam and local communities surviving the conquest gave rise to sectarian rebellions on the border between heterodoxy and secession; although they did not lead to political transformations, they seem to have paved the way for the ninth-century erosion of caliphal power. By contrast, the Byzantine Empire survived, and even in its massively weakened state it became an enduring challenge to Islamic power as well as an inducement to maintain the claim to imperial universality.

In short, early Islamic expansion confronted two separate worlds of monarchy, empire and civilization, with unequal success and different consequences. But the very fact of conquest on both sides reinforced the universal and unifying ambitions of the emerging power. After the troubled transition from peninsular state to Near Eastern empire, the first clearly defined Islamic polity (consolidated under ‘Abd al-Malik, 685-705) resumed expansion on all fronts and relied on a combination of invented and adapted institutions to support it. It was the coincidence of stalled expansion in several regions and pressures for a more inclusive imperial order that triggered a regime crisis and a change of dynasty in 750. The ‘Abbasids became the main state-builders of early Islamic history. A new capital became the most metropolitan city in the history of the Near East and the centre of a complex bureaucracy organized for taxation on an imperial scale; but the bureaucratic structures were intertwined with local elites and clientelistic ties, and to sum up, ‘the degree of control varied from highly centralized administration to loosely held suzerainty’.^{^[635]} The ascendant phase of ‘Abbasid power also saw a major effort to restore the religious authority of the caliph (under al-Ma’mun, 813-33). This attempt to upgrade the religio- political nexus took place at a stage when the differentiation of schools and trends had progressed beyond the initial pattern, and ideological power was to a decisive degree held by religious and legal scholars; the controversy provoked by caliphal policies was correspondingly sharp, and their failure bound to have broader effects.

Seen in the long-term context of Islamic history, two aspects of ‘Abbasid decline - beginning in the ninth century - seem especially noteworthy. On the one hand, the loss of central control and geopolitical weight set in very soon after the retreat from religious interventionism. The record thus highlights the interdependence of the various aspects of caliphal power. On the other hand, the caliphate as an institution proved very resilient. Its migrations and metamorphoses included rival versions set up in opposition to or at least defiance of the ‘Abbasid one (the tenth- and eleventh-century Fatimid caliphate in Egypt and the more remote one in the Iberian peninsula), as well as the latter’s shadow in exile and out of power, surviving as a paradigm of political order (in the final phase, it was appropriated by the Ottoman Empire and used to back up a claim to pre-eminence). The enduring civilizational significance of the caliphate - as a vision, not as a renewable project - was also reflected in limits to the alternatives that emerged in the course of its decline. New ethnic communities were integrated into the Islamic world, but contrary to trends in the post-Roman West, there was no turn to ethnic definitions of statehood. The main contenders during the middle centuries of our period, separated by a geopolitical divide, were the Fatimid counter-caliphate in the western and a changing constellation of militarized local or regional regimes in the eastern lands of Islam. The Fatimid state, of imperial dimensions at its strongest and backed up by a schismatic vision of Islamic history as well as (at a more esoteric level) an ambitious reinterpretation of the Quranic message, failed to develop a political model capable of refounding the caliphate. The military elites that gained the upper hand in provinces escaping central control paved the way for a power structure commonly seen as a new type of state: ‘not a direct expression of Islam, but a secular institution whose duty it was to uphold Islam'.^{^[636]} But this ambiguous status also meant that the militarized mode of government remained in the shadow of the caliphate. Even when bereft of power, the latter could still function as a source of legitimacy. On the other hand, the very construction of the military regimes, and a fortiori their expansion, depended on two factors from outside the Islamic orbit: tribal groupings from Inner Eurasia, migrating into the Near Eastern heartland and converting more or less at the same time, and borrowings from Persian statecraft. Both these sources could serve to upgrade the image of rulership and claim some kind of divine sanction; the doctrine of two unequal powers, prophecy and kingship, could in turn be invoked to contain such trends.

As we have seen, two major inter-civilizational encounters - with Byzantium and Persia - affected the course of early Islamic history. By contrast, the contact with Western Christendom - seen from the Islamic side - did not develop into a significant encounter during the Middle Millennium (that stage only began with the sixteenth-century Mediterranean expansion of the Ottoman Empire). But the final centuries of the Middle Millennium were marked by two other encounters: with Inner Eurasia through the Mongol conquests and their aftermath, and with the Indian world through Islamic advances into the subcontinent. In both cases, significant long-term effects resulted, but along different lines, not least in regard to the destinies of states and empires. The Mongol impact began with massive destruction inflicted by unbelievers and was then muted by conversion, but this did not mean that the whole episode was absorbed into an ongoing Islamic trajectory. Apart from lasting changes to the interrelations of nomadic and settled populations in the Islamic heartland, converted elites of Mongol and Turkic origin transmitted key elements of political culture to the Islamic world. This complex of operative notions included visions of empire, revitalized by the Mongol example; an exceptionally strong dynastic charisma attached to Chinggis Khan's descendants; and, to use Marshall Hodgson's term, a more systematic version of the military patronage state already adumbrated by the Turks.^{^[637]} The Indian encounter did not give rise to that kind of generalizable pattern, but it produced an expanding state based on a more stable and regulated coexistence with another civilization than any comparable power structure in the earlier history of Islam. Conversion seems to have been most effective in areas marginal to Hindu society, not at the centre of Islamic domination; there was, admittedly, no appropriation of Indian thought that could be compared to earlier engagement with Hellenic traditions, but in institutional and practical regards, the imperial power that grew out of Islamic expansion into India rested on an inter-civilizational basis.

Byzantium

Among the regional and civilizational formations discussed here, the Byzantine trajectory stands out as most nearly coextensive with the Middle Millennium. It was also the one whose destinies were most decisively affected by forces and processes generated elsewhere (more precisely within Western Christendom and Islam). The end of the Byzantine state was one of the most salient terminating events, although aspects of the Byzantine legacy survived within the Islamic empire of the Ottomans and as elements of the composite imperial tradition that took shape in Russia. The beginning of Byzantine history is more controversial, but as we have seen, it can be equated with a concatenation of major conflicts that also involved other civilizations and led to massive territorial losses. Internal transformations of the seventh and eighth centuries - affecting Byzantine state structures, society, worldview and identity - were in large measure due to the geopolitical downsizing and re-centring of the erstwhile East Roman empire.

The formation that emerged from these upheavals is perhaps best described as an ‘amalgamation of faith-zone, imperial idea and state apparatus'.^{^[638]} To call it an amalgam is to underline the less-than-complete unity of the component parts, and that is doubly important. The relationship between the three factors was flexible enough to allow for significant changes; it also enabled the diffusion of cultural and institutional models beyond imperial borders, with some variations in the relative weight and visibility of separate aspects. Historians have widely used the term ‘Byzantine Commonwealth' for the cluster of neighbouring states - from Bulgaria to Kiev Rus' - that developed autonomous versions of Byzantine patterns on all three abovementioned levels.

The imperial idea, inseparable from the affirmation of continuity with the Roman past, was an enduring and defining feature. More precisely, the core imaginary institution of Byzantium linked the imperial status of the ruler to the imperial sovereignty of the state and the imperial eminence of its urban centre (the last was more important for Byzantine civilization than for any other). There was, for much of the time, an imperial substance behind the idea, but this is no longer sustainable for the period from 1204 to 1453. During this final phase, the Byzantine world came to resemble a state system with a symbolic but gradually disempowered centre. One consequence of this was a loosening of the links between empire and faith. But even for the much longer, genuinely imperial period, it is now widely accepted that the long-held notion of Byzantine caesaro-papism is untenable. This is one of the most fundamental reappraisals in Byzantine studies, and it has led to better understanding of the religio-political nexus in question. There is no doubt that the Byzantine imperial institution had more power over the Church than its Western counterpart could ever claim. But the emperor was the protector of the Church, not its head; the power balance between the two summits of authority, however asymmetric in principle, could vary in practice, and disagreements could develop. Some notable imperial interventions in religious matters ended in failure; there was, in Dagron's terms, no successful fusion of imperial and priestly roles.^{^[639]}

Another reappraisal, not unrelated to the first, has to do with the state apparatus underpinning the empire. Contrary to earlier views, the changes triggered by Islamic expansion now appear as multiple long-drawn-out processes, rather than a grand design coordinating the infrastructures of state power; and the complex interrelations between sociopolitical forces, variously affected by successes and setbacks on external fronts, are central to present debates on Byzantine history.^{^[640]} For one thing, it seems well established that the Komnenoi dynasty, in power from 1081, was not an embodiment of feudal decline. Its state-building capacities were far from negligible.

Concluding remarks

The main conclusion from our survey is that imperial patterns - as distinct from more basic or common forms of statehood - are present in all Eurasian civilizations, but on very different levels. Imperial formations may be intermittent episodes in a longer civilizational trajectory, as in the Inner Eurasian case, or become durable components of a civilizational complex, as in the Chinese and Byzantine worlds (less successfully in the latter). In the particular context of Western Christendom, where the absence of a unifying empire was important for long-term developments, the imperial dimension of power was embodied in two centres: the Holy Roman Empire and the papacy. A minimal version of the imperial factor, evident in the ambitions and brief overlordships of regional dynasties, seems to have marked Indian history at least for the first half of our period. The simultaneous expansion of empire and religion was an Islamic invention, but the union was not perpetuated on a civilizational scale; later imperial ventures were of more modest size.

Several new imperial constellations emerged in the transition from the Middle Millennium to the early modern era. The Islamic world was more durably divided between empires than ever before; the Qing dynasty established the most balanced and lasting synthesis of Chinese and Inner Eurasian imperial traditions; Russian expansion across Siberia brought much of Inner Eurasia under the control of a power based northwest of the steppe; and the states of the western European seaboard acquired the first overseas empires. This last-named innovation was by no means a direct path to world domination, but it foreshadowed more radical changes to world affairs than the others.

State formation and empire building

More on the topic State formation and empire building: