Introduction
Sri Lanka has tended to remain isolated from projects of comparative history.Footnote 1 If it is physically separated from the subcontinent only by the Palk Strait, the religion of its Sinhalese inhabitants and its recent history outside British India have displaced it from models of South Asian history.Footnote 2 On the other hand, while it has obvious cultural affinities with Southeast Asia, it is literally displaced by its physical remoteness from attempts to define and theorise that region. It is only the conspicuous absence of such comparative reflections that can justify an exploratory and provisional paper such as this.Footnote 3 It is intended as an initial advertisement of Lankan history for theoreticians—or a pre-digestion of a broader theory for the consumption of Lankanists—and is presented in the hope that other scholars will use their expertise to revise, challenge and improve upon what is essayed here. The bulk of this paper concerns the pre-1600 period, and is rather more vague on developments after the European interruption. It focuses on the Sinhalese kingdoms and leaves to one side the history of the Tamil kingdom in the far north of the island.
The notion of an Asian ‘early modernity’ is rapidly gaining acceptance, but the particular theory of early modernity that I shall be focussing on here explodes the boundaries of Asia for a Eurasian framework and restlessly exceeds the broadest limits of early modernity itself, peeling back for a vision of 800–1830: this is the vision presented by Victor Lieberman. There are other ways to approach early modernity, of course. One immediately thinks of Anthony Reid's well-known characterisation of Southeast Asia 1450–1680 as an Age of Commerce, or of the ‘connected histories’ pursued in Asia by Sanjay Subrahmanyam for example.Footnote 4 The latter sets out to illuminate movements of ideas, people, technologies, goods and political practices that could bring the same conjunctural realities to apparently separate parts of the world. We shall touch on these movements from time to time in this essay, particularly when they intrude upon us rudely in the shape of the European interruption. But much more could be done in this vein.Footnote 5 One could consider how South Indian personnel influenced Rajasinha I's (1581–1593) vision of kingship, or how consciousness of Southeast Asian courts shaped that of Rajasinha II (1635–1687)—not to mention whether Portuguese and Dutch administrative systems influenced the internal development of the Kandyan state.Footnote 6 Other developments of great interest today, such as the enhanced salience of religious and ethnic identities, cannot be seen in isolation from the habits of thought introduced by the Portuguese as well as their violent impact.Footnote 7 The pursuit of connectedness does not, however, amount to a theory in the sense of an explanatory model; indeed, it may particularly suit those with an aversion to the generalisations and abstraction such modelling entails. It is rather an angling of the light onto chains of cause-and-effect that have been cast into shadow by the projections of national, imperial, civilizational and area studies historiographies; it is a predisposition to see any given location of study as a cross-roads rather than a town square.Footnote 8
Lieberman's work, on the other hand, is comparative in the strict sense, and presents a model that we can sit Lankan history alongside or within.Footnote 9 The theory is principally concerned with what Lieberman refers to as the ‘protected rimlands’, namely northwestern and northeastern Europe, Japan and mainland Southeast Asia, which he sees as sharing the same historical path from a milieu of warring little kingdoms to increasingly large, solid and well-defined states. In Braudellian fashion geography is the root here: states in other Eurasian areas, such as India, which were comparatively unprotected from external influence, may sometimes appear to be travelling along the same path, but their route is uncertain and wavering, and can double back on itself or disappear into the bush for long stretches of time.Footnote 10 By contrast the route of protected rimlands is rather linear, more apparently progressive, even if it only proceeds in convulsive bursts of energy (or ‘administrative cycles’), punctuated by periods of chaos and desperate improvisation.
In the first volume outlining this theory, Lieberman concentrates on the states of mainland Southeast Asia. At some point late in the first millennium AD or just after it, these ‘rimlands’ (by virtue of the fact that they lay on the periphery of older civilizations) played host to ‘secondary state formation’, domesticating world religions and constructing newly grand architectural complexes or public works.Footnote 11 They formed, in fact, charter states, insofar as their traditions were regarded by successors as normative and legitimating. In the thirteenth and fourteenth centuries they succumbed to disintegration. To describe the fortunes of the post-Charter regimes that sprang out of their ruins, Lieberman presents us with patterns for the development of territorial consolidation, administrative centralisation and cultural integration which are different but clearly related. With regard to the former, the post-Charter states rose and fell in turn according to three phases: c.1350–1570, 1600–1752, 1760–1830/40.Footnote 12 These last two phases can be seen as making up an extended ‘early modern period’.Footnote 13 With each successive phase, smallish kingdoms tended to be amalgamated into larger polities.
The process of administrative centralisation Lieberman sees as being rather more linear. Naturally, the various states of the mainland did not march neatly in step towards governmental reform, so the categories overlap chronologically. The charter states, existing between 900–1450, are typed ‘Pattern A’.Footnote 14 From 1450–1840, one can find Pattern B states (Decentralised Indic Administrations), which make do with a circumscribed administrative function and little direct political control beyond the nuclear zones, but in which the autonomous role of monastic establishments is also reduced. From 1600–1840, however, the new regimes in Siam and Burma reveal enough centralising will to be characterised as Pattern C (Centralised Indic Administrations), extending the reach of the court beyond the nuclear zone, extracting tax and services with greater efficiency and so on.Footnote 15 The characteristics of these different types will emerge in greater detail below. And over this epoch political centralisation was matched by cultural integration. Theravadin and Confucian ideologies were promoted and controlled by increasingly powerful kings, and as more and more people identified with the ethnicity of the central elite, a deepening sense of politicised ethnic awareness was born. Lieberman also weaves into his analysis such matters as economics, commerce, demography and climate change, which are obviously important but cannot be addressed here.
Now, Lieberman does not present Sri Lanka as a fellow traveller of these sporadically-advancing mainland Southeast Asian states. In the second volume of his project due to be published by 2009, Lieberman intends to extend his comparative reflections to the global level, promising to show us how ‘these trajectories differed from patterns in South Asia, island Southeast Asia and other more exposed areas of Eurasia’.Footnote 16 These exposed zones ‘lacked effective barriers to nomadic or seaborne invaders’, and so remained vulnerable to domination by elites who were substantially foreign to their subject populations.Footnote 17 So where should we place Sri Lanka?
Sri Lanka is clearly a ‘rimland’, lying as it does on the periphery of the original heartland of Hindu-Buddhist civilization in India even if it took inspiration from that heartland remarkably quickly. In a broad sense, before 1500 it was also comparatively ‘protected’ from military projects—if not cut off from commercial movements. It is perhaps this mixture of exposure to benign movements of men, goods and ideas and relative insulation from malign projections of power, particularly early in the first millennium, that gave rise to the Anuradhapura civilization. This enjoyed an extraordinarily long lease of life from roughly the second century BC to the late tenth century AD, of which the period circa 67–429 AD was apparently remarkably stable. Incidentally, this chronology seems analogous to that of the Roman Empire in some ways.Footnote 18 To be sure, its close proximity to South India meant that from the fifth century Sri Lanka was exposed to incursions from across the straits, and from 900–1500 AD, the course of Lankan history was heavily influenced by its vulnerability to subcontinental powers such as the Colas. Nevertheless, from the sort of very large-scale comparative perspective that Lieberman asks of us, the resilience of indigenous rulership is impressive.
Indeed, it was only after 1500 that the sea became more of a conduit than an obstacle and the island was rendered magnificently unprotected, standing as it did athwart the sea-lanes connecting the eastern and western halves of the Indian Ocean. Moreover, by this time the centre of political gravity had moved to the least topographically-protected and most commercially-interesting part of the island: the city of Kotte was located on the southwest coast surrounded by spice-growing hinterlands. One could say, in other words, that the dynamic movements of the early modern period were felt as a more violent perturbation by unprotected lands. In Sri Lanka's case these were present in the successive European colonial projects that it was largely unable to resist. The implication here is that such interruptions ensured that the island would not properly be able to engender the shift from the fissiparous Pattern B states to the hardier types of Pattern C.Footnote 19
Long after this present essay was written—but shortly before it was published!—I saw an extract from Lieberman's work-in-progress, from which it became clear that in broad outline he traces a similar trajectory for the whole of island Southeast Asia in his overarching scheme: it too was ‘protected’ by the sea until the sixteenth and seventeenth centuries at which point it became subject to the vicissitudes of the exposed zone.Footnote 20 However, the great majority of this essay was written with only Lieberman's first volume to hand, and thus it takes mainland Southeast Asia as its main point of comparison. I considered Sri Lanka as a particularly revealing test-case because of its close kinship with the states of this region, which meant that many of the variables remain constant while its geographic situation is left as the outstanding difference. In short, Sri Lanka looks analogous to the mainland world in socio-political terms, and analogous to the island world in geopolitical terms. The result is that from c. 900–1500 Lankan political history displays some real affinities with chronological patterns of mainland Southeast Asia, and from c. 1500–1800 with those of island Southeast Asia. What would happen then when the political animals of the Southeast Asian ‘galactic’ species were subjected to the harsh ecology of European imperialism?
However, I shall suggest here that it is not simply the case that Sri Lanka was blown off course by the sixteenth-century conjuncture, but that her preceding history was idiosyncratic by virtue of the vexed question of continuity from much earlier epochs. This is particularly important with regard to religion and religio-political ideas, because Sri Lanka was the incubator, over many centuries, of the ‘Mahaviharan revolution’ that would come to be so definitive of early-modern dynamics in mainland Southeast Asia. Moreover, the problem of apparent continuity also arises when we come to locate the charter-state period in Sri Lanka. If we are moved by synchronicity, then the Polonnaruva period (c. 993–1250 AD) is the obvious candidate; yet if we make attributes alone decisive then the Buddhist polity of Anuradhapura that arose early in the first millennium, with its massive irrigational and public works, bears many of the appropriate hallmarks—including the critical sense in which it was seen as normative by its successors.
This is by no means to deny that the Polonnaruva period was one of profound transformation. On the contrary, I shall suggest here that if one is concerned to isolate a pivotal epoch for the island's history (an intrinsically elusive ambition), together with its aftermath in the thirteenth century, may well form the best candidate.Footnote 21 That is to observe that many of the grand themes that most concern our model-builders—the political, the ethnic, the literary and the religious—seem to find their most striking advancement during the Polonnaruva and early Dambadeniya eras when the Sinhala kings were often engaged in intense rivalry with the Colas for dominion over the island.Footnote 22 The problem is that the ‘early modern period’ is thereby drained of specificity from the other end, so to speak, as Polonnaruva begins to look more like a template for the second millennium than a contrastive prologue.
We have already waded into troubled waters here, since even pondering the more ‘static’ or slow-moving features of Sri Lankan history is to stand against the tide of recent Asian historiography, which has been concerned to recover diachrony for pasts smoothed over by the characterisations of Orientalist scholarship. Of course, it also brings us uncomfortably close to the agendas of vamsas, and the story told by some with nationalist preoccupations of the long decline from the glories of Anuradhapura, in which even the Polonnaruva period can be presented as a reunification of the island and a reform of Buddhism. Anuradhapura is not only placed at the beginning of an unbroken history of the Sinhala Buddhist nation stretching to the present day, it is also represented as a purer, more real manifestation of that primordial identity, the flowering of its greatest glory.Footnote 23 Historicist-minded scholars have set out to deconstruct such stories and their work is by no means done.Footnote 24 However, it ought to be remembered that where our reflections do turn to the continuous it is in the spirit of a comparative exercise which uses other Asian regions as its foil and is designed to capture the varying pulse of change rather than to stop it dead. Moreover we ought not to lose sight of a respectable ‘traditionalist’ perspective that would draw our attention towards what is particular or idiosyncratic about Sri Lankan history.
The State Before 1600
The Nature of Government
The Polonnaruva kingdom seems to settle comfortably into the clothes Lieberman has laid out for his charter-polities of Pattern A. It is thus one of the ‘classical’ temple-building civilizations that drew their wealth from the maintenance of irrigation systems and which flourished early in the second millennium AD. So the celebrated Polonnaruva of Parakramabahu I (1153–1186), the centre of a unified Lanka, is easily compared to Anawratha's Pagan (1044–1070) or Suryavarman's Angkor (1113–1150). And, to step into ‘connectedness’ briefly, there were real currents of influence that may have helped propel the common tide of history. One such flowed out from the island: Parakramabahu saw Southeast Asia as a legitimate field for the demonstration of his power, lording it over parts of the Malay peninsula, sending an invasion force into Burma, setting up marriage alliances with Angkor.Footnote 25 His kingdom would then also be subject to the same conjuncture of fragmentation, collapsing as it did into a number of unstable lesser polities. It was only in the fifteenth and sixteenth centuries that newly authoritative political centres arose, often now based in coastal regions and more interested in the exploitation of maritime trade.
We shall not proceed too far here into the underlying reasons as to why this basic pattern should have been reproduced across many parts of Eurasia. One of the most plausible in a priori terms is the effect of climate. Lieberman points to the warming of the northern hemisphere during the Medieval Climate Anomaly of c. 850/900–1250/1300, which may have enhanced agrarian production in the protected rimlands and so laid the basis for charter florescence. During the subsequent climatic downturn there was a marked shift of economic activity from the charter core to peripheral regions, which naturally had a destabilising effect. Across the rimlands and China too, one sees severe political disruption and dislocation from c. 1250–1450.Footnote 26 This chronology captures almost exactly the rise of Polonnaruva, its disintegration in the early thirteenth century and the subsequent period of petty princelings and centrifugalism before the rise of Parakramabahu VI (1411–1465).Footnote 27 What remains unclear of course, is whether climate can be identified as the culprit in Sri Lanka.Footnote 28 On the face of it, the ‘drift to the southwest’ away from the heartland of dry-zone civilization seems simply a result of its vulnerability to military projections from South India. The marauding rule of one such warlord Magha (1215–1255) may have proved that the northern lowlands were now simply too ‘unprotected’ a location for a political centre. But, as C. R. De Silva points out, there had been wars and disorder before, after which the irrigation systems had been restored.Footnote 29
Whatever the reasons, Sri Lanka fits the pattern, which is also discernible in the nature of the regime that now arose in the new centre of Kotte, a city located less than 10 km from the sea. The rapid extension of Kotte's overlordship over the whole island during the reign of Parakramabahu VI can be compared to the rise of Pegu under the Toungoo kings.Footnote 30 However remarkable their rise, both dynasties failed to establish lasting institutions that might keep at bay powerful centrifugal forces; they remained firmly within the Pattern B mode.Footnote 31
Indeed, we are now in the age of the ‘galactic’ or ‘solar’ polity, of ‘tributary overlordship’ or ‘nested sovereignty’.Footnote 32 Rule beyond the core territories operated through a form of vassalage that allowed sub-kings to retain much of the symbolism of royalty, and depended on a continual cultivation and renegotiation of personal loyalties to the high king. This system of overlordship was not simply a practical accommodation to the business of rule but a desired end in itself: to have under-kings was a key signal of cakravartin status. The process of royal succession was typically fraught; its principles rarely beyond controversy; disintegration often won the day.Footnote 33 Given the institutional weakness of the state and its ephemeral boundaries, the control of manpower had absolute priority: ‘whenever the dominance of the capital was questioned it was reflected in the steady seepage of manpower away from royal control’.Footnote 34 And the more was lost the more difficult it became to retain the redistributive powers with which to secure loyalty. The exploitation of whatever symbolic capital that remained was crucial for the reassertion of authority. The magnificent grandeur of the royal image, particularly at the rites of tribute-presentation, was critical. Royal blood never lost its significance, and so it was possible to bind sub-rulers to the centre through marriage, or replace them with closer family members after pacification.
The above paragraph is drawn from surveys of Southeast Asian history, but it would work equally well as a snapshot of the fundamental political structures that underlay the turmoil of Sinhalese history in the later fifteenth and sixteenth centuries (and probably well before too).Footnote 35 In this context, for example, the great swiftness with which Mayadunne was able to undermine Bhuvanekabahu VII's (1521–1551) authority in the 1540s, as he sucked more and more local chiefs into his orbit and redirected the flow of revenue towards Sitavaka, loses some of its mystery. Throughout the sixteenth century, Kotte and Kandy sought to counter this threat through marital alliance.Footnote 36 The rapid rise and fall of Sitavaka itself, completed within the two generations of its only kings, Mayadunne (1521–1581) and Rajasinha I (1581–1593, but de facto rule from early 1560s), is of course also firmly within the Pattern B mould. Indeed in its rampant militarism and rapid self-combustion, Sitavaka seems an even better analogy for the First Toungoo empire than Parakramabahu VI's Kotte. Rajasinha's drive towards constant expansion while neglecting lasting administrative reform and his descent into unpopularity through overweening rule and religious oppression has interesting analogies with the dark reign of Nandabayin (1581–99).Footnote 37
The timing here indicates that the most mysterious, perhaps the most under-explained aspect of Lieberman's theory, the synchronicity of the administrative cycles, can be applied to Sri Lanka too, for the indigenous polities of the island undergo a late-sixteenth century crisis just as they do on the Southeast Asian mainland: Kotte subject to inexorable decline as an independent power with the accession of Dharmapala (1551–1597), who finally bequeathed his crown to Portugal in 1580; Kandy's flickering half-life smothered by Sitavaka's invasions in the 1580s; and the latter itself dissolving completely with Rajasinha's death in 1593.Footnote 38 We might invoke here the increasingly militiarised milieu brought about by the Portuguese presence and the arrival of firearms—a potential driver of state consolidation, to be sure, but an unstable one that could quickly become an agent of state fragmentation. By the 1580s and 1590s, we find lascarin units whose principal loyalties often appear to be with their unit commanders. This accelerated already strong fragmentary tendencies by endowing officers lacking royal blood with greater control of manpower, and the ability to pursue personal objectives comparatively un-beholden to any one court tradition. The exemplar here is Manamperuma Mohottala, a Saivite favourite of Rajasinha I, who was left clutching the reins of the military in Sitavaka after his patron's death, but then defected to Kotte and thus helped to engineer the Portuguese conquest of the lowlands.Footnote 39 Whether by virtue of some conjunctural force or by simple coincidence, the Lankan kingdoms’ experience of external interference had ended in the same collapse as mainland states such as Ayudhaya and Pegu.Footnote 40 It was only from the 1590s onwards that the imposition of a ‘foreign conquest elite’ would appear to wrench its history off in another direction entirely.
So much for the Liebermanesque perspective. There are, however, other viewpoints to consider. If Lieberman has demoted the early modern period in favour of a story reaching back to the tenth century, the historian of Sri Lanka is traditionally invited to push on much further back in time. As we have already remarked, take a second look at Polonnaruva and its ‘charter-polity’ clothes begin to look like a somewhat baggy hand-me-down from its elder brother, Anuradhapura, who had, after all, been wearing them for a millennium or so. Some might question whether a Lankan early modernity can be discerned via an enhanced role for trade. If Sri Lanka's geopolitical position connecting the two halves of the Indian Ocean only made it conclusively vulnerable to military threats after 1500, it seems to have brought the sort of opportunities of seaborne commerce to the states of the first millennium that we might otherwise associate with the early modern era.Footnote 41
As for administrative structures, in Lieberman's modelling there is a sort of regression between the charter states of Pattern A and the decentralised states of Pattern B, which works for us: the attempts of the twelfth century kings, particularly Parakramabahu I (1153–1186), to centralise institutional power strike one as considerably more ambitious than those of successive states.Footnote 42 But they did not establish any lasting administrative unity and may even have worked, perversely, to un-knit some of the fibres of state power through the damage they inflicted on local nobility.Footnote 43 What was there to un-knit? The decaying fabric of Anuradhapura. Ineluctably drawn to a perspective on the whole of two millennia, Polonnaruva can appear as a transitional phase between the Anuradhapura era and the more heterogeneous polities of the second millennium. In the reflections of Bardwell L. Smith, it is presented as a period of fragmentation and pluralism lent a superficial unity by the military achievements of its kings, almost as if it were exhibiting some of the symptoms of post-charter collapse at the same time as charter establishment.Footnote 44 Such, perhaps, are the tricks of perspective of the too longue durée. It is in this light that Tambiah's vision of an archetypal Theravadin polity, evincing in its history less progression than oscillation—between the brief flashes of strong statehood under a Parakramabahu I or VI and the habitually loose assemblage gathered under lesser rulers—can acquire some appeal.Footnote 45 Nevertheless, the ideological precedents set by the first of that name were pivotal.
Persisting Boundaries
The feeling remains then that there is more to this sense of continuity than the synchronic conceits of structuralist anthropology or the creeping epoch-extending of a certain kind of historiography, forever chasing antecedents further and further back in time. Is it just the assumptions of nationalist storytelling which might incline us to trace a single line between the two classical irrigation civilizations, and between these and the polities of the early modern period? Alighting on these stepping stones of state power popping up in different regions of the island at various times is, after all, to hop over centuries of political disintegration, over little-known marginal powers in the vanniyas or the east, over periods of domination by South Indian regimes, over deep transformations in their socio-economic base or relationship to religious institutions. Moreover, the tendency for new polities to enhance their legitimacy by evoking classical images of kingship, by asserting lineage from grand ancestral states, is to be found all over the world and is highlighted by Lieberman's theory through the very term ‘charter polities’.Footnote 46 In what sense is it meaningful to argue that Sinhalese rulers were doing anything different to analogues in Southeast Asia or elsewhere? We might be impressed, for example, that the kings of Sitavaka or Kandy were so intent on capturing Kotte-Colombo, even when the great majority of the rest of the island was under their rule, in order to capture a traditional seat of imperial power and the titles to go with it.Footnote 47 Yet is this really any different to the desire of the Toungoo kings to capture Pegu, or the Siamese appropriation of Angkor?
One idiosyncrasy of the Lankan situation is the unusual continuity in the political significance of a geographic boundary, the island itself.Footnote 48 Conquest of the whole island, the attainment of unopposed sovereignty over it, was therefore seen as a principal target of a Sinhalese ruler's ambition and the natural right of those succeeding to the imperial title. Bhuvanekabahu VII (1521–1551) never gave up this ambition even when the contrast with his actual political authority was frankly astonishing. General precedents for such ambitions were very ancient indeed. Already in the fifth century, the Mahavamsa was celebrating the achievements of Dhatusena (455–473) in throwing off the foreign yoke and drawing the whole island under his rule through an epic retelling of the reign of Dutugamunu (161–137 BC) who was supposed to have been the first king to manage this feat. In the tenth century the island was brought under a single royal canopy by the Okkaka kings, who promulgated an interpretation of the vamsa texts that heralded their special genealogical descent from mahasammata, and unique claim to overlordship of the island—and indeed beyond.Footnote 49 A poem from the tenth century embedded in a commentary on the Mahavamsa personifies the island of Lanka as a beautiful woman, her various features made up of sites of religious and political fame, ‘. . . clothed in the blue clothes [called] the still waters of the ocean that surround [her]’.Footnote 50 (In the spirit of connected history one might note how much this eroticisation and feminisation of the land shares with Cola images of the king as lover to the earth).Footnote 51
The first evidence linking island-wide supremacy to a special status and title expressed through a higher form of coronation derives from the reign of Parakramabahu I (1153–1186), while his heir Nissanaka Malla (1187–1196) was the first to refer to himself as a cakravartin.Footnote 52 That part of the Culavamsa composed in this era takes Parakramabahu I as its hero and forsakes any real drive towards regional (‘world’) supremacy in political terms for a more parochial vision.Footnote 53 After this point no other Sinhala ruler would pursue any serious overseas ambitions. The increasingly important role of the tooth relic over the second millennium, as the item of legitimisation par excellence with a unique power to compel others to obeisance, is also highly significant, for by definition only one king at a time could hold it.
It needs to be emphasised again that this ambition of unity was rarely realised in practice, and that its relevance waxed and waned: it was far from being a political vision that effaced all others. Certainly in the second millennium, the majority of kings were primarily concerned with the political boundaries within the island, which were often shifting in response to the activities of rivals. Therefore, an assessment of how Sri Lanka might fare according to the early modern characteristic of ‘territorialisation’ must be two-faced: an early territorialisation on the plane of imagination, and a perennially deferred counterpart on the plane of institutional reality.Footnote 54
There remains a very particular form of territorial boundedness to much of Sri Lankan political history that stands in contrast with the more shifting activity of mainland Southeast Asia.Footnote 55 Yet the next time that unity was actually achieved would be under Parakramabahu VI. If it was so rare, why does it make sense to give it our attention? There are at least two answers. First, because it was no mere febrile fancy of the odd generation of courtiers. It is apparent from the literary record that the vision of sovereign unity became a potent organising master symbol, drawn upon to invoke ideal attainment in many arenas of life. A monastic hierarch guiding the Sangha, a literary figure skilled in all the arts, or a king with all the attributes of man: all are likened to the comprehensive mastery of the cakravartin.Footnote 56 If Parakramabahu I's reign (1153–1186) is presented as seminal in this essay, it is because it was then that this vision of unity—in all its guises—was so firmly established. Second, because it is very likely that Parakramabahu VI looked back to the first king of his name, just as Parakramabahu I in turn looked back to his forefathers in Anuradhapura.
Ethnic Continuity and Historical Consciousness Before 1600
For we have been avoiding the eye of a contentious subject: the continuity of Sinhala culture and identity. It must somehow be important—both to us as analysts and to them as agents—that the rulers of Kotte were Sinhala-speakers, and they had this in common with those of rulers in Anuradhapura, Polonnaruva, Dambadeniya, Gampola etc. The Burman Toungoo kings may have drawn on Burmese precedents (late Pagan) but they saw it necessary to take Pegu and adopt its Mon heritage too: Tabinshweithi had his primary coronation there, married a Mon woman and favoured Mon hairstyle. Indeed his east-west coastal empire avoided the Burman heartland, taking in Shan and Mon areas.Footnote 57 The mainland Southeast Asian polities which rose and fell over the second millennium can be seen as Venn diagrams which contained sub-sets or presented intersections with a variety of tribes or ethnic groups who were distinguished by language, style of dress, political tradition, religious predilection, particularities which in turn were sustained by topographical features.Footnote 58 And these sub-sets often formed the various administrative units (provinces or sub-kingdoms) of the polity.
Lieberman argues that each successive post-charter administrative cycle saw the dominant cultural hue shade over these sub-set lines, and it was particularly in the early modern period that single colours began to creep over the map: the chief mainland societies of Southeast Asia promoted cultural integration (an etic evaluation) and ethnic consciousness (an emic quality).Footnote 59 Both of these are imagined to move vertically and laterally: vertically to include all the social levels of the dominant ethnic or linguistic group, and laterally to spread over and incorporate other ethnic groups. Cultural integration was promoted by diverse mechanisms including the promotion and regulation of state-sponsored Buddhism, the harnessing of the population through royal service, tribute-presentations that brought the regions to the centre, through rituals and pilgrimages and through continuing war with outsiders. As for ethnic self-awareness, this is presented as arising among an elite perhaps in the charter era (Burma) or the fourteenth and fifteenth centuries (Ayudhya, Laos, Northern Thailand) then diffusing both vertically down the social scale and laterally across to be appropriated by diverse other groups over the early modern period, ‘accelerating sharply in the chief mainland empires after c. 1770’.Footnote 60 Lieberman is very aware that these tendencies were combined with a paradoxical flourishing of universalist discourses of politics (world conquering or imperial visions) and religion (the appeal to foreign sources of religious legitimation etc), not to mention a voracious appetite for cosmopolitan fashions among the central elite.Footnote 61 It is for these reasons, and others that he eschews the terminology of ‘proto-nationalism’ for that of ‘politicised ethnicity’.Footnote 62
How does this compare to Sri Lanka? The etic evaluation is difficult to make at present, because we lack the appropriate kind of assimilative and conceptual work.Footnote 63 However, it does seem that by the thirteenth century—and even perhaps by the eighth to ninth centuries—there was already a certain cultural homogeneity to the Sinhala-speaking majority in the island as expressed particularly in a remarkable unity of language (that is to say, with few major differences in regional dialect) and to the predominance of a capacious Buddhism in their religious lives.Footnote 64
Sri Lanka is profitably seen in the light of Sheldon Pollock's account of a general shift across many parts of Eurasia early in the second millennium from the ‘universalistic’ to the ‘localised’—as seen most clearly in the increasing popularity of writing in the vernacular as opposed to the classical languages such as Sanskrit, Latin and Greek and so on.Footnote 65 Lieberman has developed Pollock's analysis to remark that vernacularisation was most vigorous on the periphery of the older centres from which the universal languages disseminated.Footnote 66 Sri Lanka is of course on the furthest periphery of the North Indian heartland in South Asia, and sure enough here we find early and vigorous vernacularisation. Sinhala as a literary language dates back to the eighth century, when the Brahmi script was replaced by the round-shaped script capable of representing long vowels.Footnote 67 The graffiti from the Sigiriya site of the eighth and ninth centuries reveals a Sinhala literary culture in which a quite surprising variety of social groups participated: courtiers and villagers, men and women, monks and laymen.Footnote 68 Charles Hallisey sees the fullest realisation of literary Sinhala around the turn of the millennium.Footnote 69 In one sense then, some sort of vertical integration was probably achieved early on in comparison with Southeast Asian societies; in another sense one might only detect a directed administrative push towards homogenisation in the late eighteenth or nineteenth century.Footnote 70 Again then, Sri Lanka can seem either too premature or too late-developing.
Be that as it may, Lieberman is equally as concerned with lateral integration, with what happens when the territorial expansion of the state brings different ethnic or cultural groups under its authority. The Sinhalese kingdoms already manifested a long-standing acceptance of cultural heterogeneity, as seen in their relationships with the vadda people, or with the Tamils of the north and east whenever they were brought under the canopy of the cakravartin. Given that canopy could not extend beyond the island, of more relevance perhaps is the relationship with immigrant groups, who came in substantial numbers in the Dambdadeniya and Kotte eras. These new groups were lent a place, often a subordinated one, within an overarching socio-political framework rather than being immediately assimilated.Footnote 71 By the sixteenth century, for example, we find various fishing castes populating coastal regions and preserving their Hinduism, or the Karavas settled in the southwest as a fighting force, or the recently-arrived chieftains ruling the sparsely-populated vanniyar region in the north.Footnote 72 But if the initial concern was to contain and master difference rather than simply erase it, the latter was promoted over the long term by the organic forces of ‘Buddhicisation’ and ‘Sinhalacisation’ (as particularly explored in the work of Tambiah and Obeyesekere).
If the etic question is beset by lack of directed analysis, the question of self-conscious identity is beset by an abundance of controversy. We can begin by being clear that ethnicity is not defined here as implying a firm or articulated notion of shared ancestry. Apart from the difficulties of finding evidence for this sort of understanding, it seems too narrow for the task at hand, which is to work out if—and when, and to what extent—it was meaningful for people to see themselves as being a member of the ‘Sinhala’ group.Footnote 73 In Sri Lanka, a Sinhala identity coalesced at least as an elite or dynastic category early in the first millennium. Exactly when and how it subsequently underwent a vertical diffusion down to the very lowest rungs of the social ladder is still a matter of great debate, famously pursued by K. N. O. Dharmadasa and R. A. L. H. Gunawardana—but one contained within the latter half of the first millennium and the first two centuries of the second.Footnote 74 One disputed text from the tenth century, the Dhampiya Attuva Gatapadaya, apparently asserts that the term for the ‘Sinhala’ (or helu) language is derived from the name for the residents of the island, which is in turn derived from the dynasty ruling over them.Footnote 75 However, even the most historicist of the disputants, Gunawardana, admits that the subsequent Cola occupation may have provided conditions for an ‘archaic ethnicity’, that is to say one relevant at least to small groups of the literati. Thus he suggest that the crucial terminological shift to an inclusive definition of ‘Sinhala’ had happened by the reign of Parakramabahu I's successor, Nissankamalla (1187–1196).Footnote 76 It was this same Nissankamalla who broadcasted the first explicit arguments that a king of Lanka must be of ksatriya lineage and Buddhist fidelity, and who made first use of the cakravartin title . . . .Footnote 77 At least in terms of political rhetoric (or bombast!), Nissankamalla's brief reign looks rather seminal for the ensuing millennium.Footnote 78
However, Gunawardana still wants to argue that a sense of Sinhalaness remained of interest only to a small minority, and did not emerge as strongly politicised and widely relevant until the shift to a ‘mature’ ethnicity under the British—an epistemic watershed shared with much recent historicist work. Towards the end of this essay we shall comment on how this underestimates the implications of military struggle with earlier European power, but it may also underestimate the potential for a broader Sinhala consciousness in the late Polonnaruva era and its immediate aftermath, assembled against a common and heretical Cola enemy. We ought to think of such sentiments as constituting a ‘soft’ identity, one sometimes more relevant to elites than to the masses and whose salience rose and fell according to political context—rather than as marking the birth of an eternal realisation. One such context may have arisen with the so-called ‘Sinhala rebellion’ against Bhuvanekabahu VI (1469–1477). When he brought down his army from Jaffna to take control of Kotte city, he appears to have sent a wave of disgruntled displaced Sinhalese courtiers out to the provinces to raise opposition against the newcomers.Footnote 79 It is possible then that subsequent interactions with the Portuguese only reinforced an identity that already had some resonance with a wider population.
As it happens, the chronology of a socially-comprehensive Sinhalaness is not a matter that we need to resolve here, because the perspective of the ruling elites themselves is important with regard to our comparison with Southeast Asia. For we can say that by the fifteenth century, those who built their capitals in Dambadeniya, Gampola, Kotte, Sitavaka and Kandy would have thought of themselves as of the same people who had built Anuradhapura, and this feeling would have amounted to much more than a tenuous proprietary myth; it was built on a rich and apparently continuous narrative.
This may seem like a striking statement to make, but then the antiquity and persistence of the Pali and Sinhala chronicle traditions is such a striking and unusual fact. If Polonnaruva seems to some today as a reform or revival of civilization, it is a sentiment that finds its echo in the perspective of the late twelfth/early thirteenth century Culavamsa, which compares Parakramabahu I with Dutugamunu (circa. second-century BC), a trope which was in turn a resurrection of a comparison that the fifth-century Mahavamsa invoked for Dhatusena (455–473 AD). Whatever their origins and their shifting political relevancies, it is difficult to argue that these vamsa discourses were only ever meaningful to very limited literate circles as specific projects of dynastic propaganda when we find their myths and messages in quite different kinds of texts, such as the Sinhala Pujavaliya, which was intended for broad oral dissemination.Footnote 80 And beyond the chronicles there was a rich literary heritage about which, as Charles Hallisey, has remarked, ‘Sinhala literati have been self-conscious . . . for more than a millennium’.Footnote 81 Indeed Hallisey has argued that the conventional overviews of Lankan history, which rupture the past according to the shifting locations of the capital cities, can serve to efface patterns and continuities in literary culture that remained apparently untouched by such upheavals.Footnote 82
From the beginnings of the post-charter period onwards, texts are routinely determined to place their subject within an historical overview reaching back into the earliest times.Footnote 83 As a brief example from the sixteenth century we could refer to the Sinhala Rajaratnakaraya composed in the small and relatively recent kingdom of Kandy, probably in the 1540s when it was penned in by Portuguese-influenced Kotte on the one side and Sitawaka on the other. But it allowed its ruler to trace his descent all the way back to Vijaya and the Anuradhapura kings and is clearly dependent on the Pujavlaliya for its reading of the vamsas.Footnote 84 In the 1580s, the Sitavaka Hatana reminded its listeners of Dutugamunu's martial feats against the Tamils.Footnote 85 A few years later, when scions of the ruling families of Sitavaka and Kotte fled to Goa in the 1590s and converted to Christianity they were able to chant versions of the Rajavaliya to a Portuguese writer such as Agostinho de Azevedo, who used them to flesh out the early history of the island.Footnote 86 By the mid-seventeenth century, the Portuguese were well aware of the Sinhalese consciousness of Anuradhapura history.Footnote 87 And Fernão de Queirós, a Portuguese Jesuit, would assert in the 1680s that ‘as for the character of the Chingalas, they are generally proud, vain and lazy . . . because of the antiquity of their Kingdom and people [nação] and the liberty in which they were always brought up’.Footnote 88 In short, Sinhalese ruling elites had a sense of historical self-consciousness, of their vicissitudes as over centuries.
By this time, European observers were making similar comments about mainland Southeast Asian peoples and their robust pride for their distinctive laws and customs.Footnote 89 Yet the chronicle tradition there—which owed a lot to Sinhalese models—was much less ancient.Footnote 90 Naturally, it would not do to imagine that the ideology of the vamsas and other texts corresponded to the mentalité of a whole society, but we would surely be missing something distinctive about Sri Lankan history if we failed to see the characteristic way in which the realities of instability and fragmentation were combined with stubborn visions or aspirations of unity (of the sasana, of the political realm). This must lend a vital element of subjective and objective continuity to the cyclical rise and fall of political centres in the island, and one that is much less evident in the various administrative cycles of mainland Southeast Asia. As so often with reflections on the slippery issue of identity, the distinction can seem vanishingly slender from some perspectives and yet awkwardly critical from others; it is perhaps, to make an analogy with the West, the difference between the relation towards Ancient Rome borne by early medieval Byzantium and that claimed by the Carolingian kings.Footnote 91
Lieberman has reflected on the circumstances under which ‘cultural or political interdependence translated into a degree of self-conscious, articulated solidarity’, and highlighted two factors: ‘First was a stable political centre—in turn dependent on favourable geography and external security—that could coordinate patronage, validate cultural norms, and provide a focus of identity . . . . The second requirement was a sustained sense of alterity or competition rooted in struggle for scarce resources . . .’.Footnote 92 In Lieberman's later work, these become some of the core features of the ‘protected rimlands’. It is immediately apparent that after the abandonment of Anuradhapura, Sinhalese continuity is not shepherded by some equally continuous and stable political centre, and this was in part because its ‘external security’ was increasingly corroded, first by land-based South Indians and then from 1500 more emphatically by seaborne Europeans. But, before 1500 there were important limits to this vulnerability, for at no point did a minority conquest elite establish a stable long-lasting rule over the majority of the Sinhalese inhabitants. Indigenous rule survived, even if outside pressures combined with the inheritance of the Indic model of hierarchical sovereignty to ensure that it would usually do so in fractured and polycentric form. Its geography was critical in a further sense, in that it helped to precipitate a stable political vision of the ambition for island-wide overlordship.Footnote 93
Buddhism Before 1500
Before we move on to consider the European interruption, we need to reflect on what Lieberman's theory has to say about religion and its role in the grand drama of integration. Established as a state religion in Burma from the eleventh century and elsewhere from the thirteenth to fifteenth centuries, the implantation of Theravada Buddhism in mainland Southeast Asia antedates that of monotheism by many centuries, yet Lieberman and Anthony Reid have argued that it underwent important transformations during the early modern period.Footnote 94 Theravada Buddhism in the ‘charter’ phase was essentially a cult of the ruling elite, and one which fluidly co-existed with promiscuous acceptance of Mahayana, Saivite and Vaisnava motifs. There are impressively large landed monastic foundations which become the chief centres of Pali learning, but little attempt is made to penetrate into peasant religion which retains animist characteristics.
But as the millennium wore on, and particularly from the fifteenth century, rulers of newly ambitious centralising states became increasingly single-minded in their promotion of Theravada and the Mahaviharan tradition. Claiming dominion over such heterogeneous regions, it was natural for them to seek legitimation in a universal textually-based form of religious authority: the moral superiority of the ruler and the cultural superiority of the ruling elite were affirmed as local cults were undermined.Footnote 95 A network of monasteries spread throughout the land and ordinary people were brought into the Theravada world by entering into exchange with the village-dwelling monks, giving alms, listening to sermons and jatakas, participating in merit-making rituals, absorbing Buddhist art and going on pilgrimages or to festivals. And the meanings of all this religious activity were increasingly being controlled by the centre as rulers took it upon themselves to disseminate texts and ritual norms, and to make the Sangha subject to a single hierarchy under their supervision, exercising the right to purge and reform it at will. These latter changes were particularly apparent from the seventeenth century, as the various Sanghas were endowed with tight organisations, often under the leadership of mahasangharajas. Monasteries began to provide an education for lay-folk that became routine for the elite and would slowly come to encompass lower social circles too.
A rigorous analysis of how the evolution of Sri Lankan Buddhism might correspond to this model calls out for specialist attention. Here, I shall merely offer some initial pointers and make the observation that the unusual continuity from antiquity is even more marked in the religious sphere than in the political, and as a result the concept of the early modern period, even its long version, loses much of its appeal.Footnote 96 Sheer antiquity is most obviously apparent in the establishment of Buddhism as the religion of the ruling elite certainly by the first century AD.Footnote 97 By this time the Sangha had been positioned as the privileged representatives of transcendent order and therefore the chief articulators of royal legitimacy. Of course, over the first millennium the same could be said for other regions across South Asia, as monastic orders existed in a cosmopolitan world of intellectual exchange, diversity, and competition encompassing different forms of what we now know of as Theravada as well as Mahayana.
The most compelling interpretation of what happened next in the progress of Buddhism across Asia has been advanced by Reynolds and Hallisey, who refer to a general shift from a ‘civilizational’ phase, in which Buddhism was essentially a cosmopolitan phenomenon of monastic elites patronised by imperial elites in urban centres, to a much more localised ‘cultural’ phase, in which the monastic orders settled within cultural or state boundaries and began to extend deeper roots among society at large. They see this as beginning in the fifth century and reaching culmination in the ninth to thirteenth, a vague and roomy chronology to be sure, but one which seems tailor-made for the awkward dimensions of Sri Lankan history. For our purposes, we can extend its application to a range of related developments: (1) the promotion of the Mahaviharan brand of Theravada by ruling dynasties and a corresponding concern for orthodoxy and purity; (2) localisation; (3) vernacularisation and popularisation; (4) royal control over the Sangha. What is striking about these factors, stated baldly in this way, is that they seem close to some of the defining features of the early modernity of the Southeast Asian states.
Mahaviharan Orthodoxy
By taking us right back into the heart of the first millennium in this way, Reynolds and Hallisey's formulation allows to appreciate how Sri Lanka became the crucible for the ideological transformations that played such an important part in this shift, and which can already be glimpsed in embryonic form in the Mahavamsa of the fifth century AD. It was around this time that Sri Lanka played host to the revival of Pali—and therefore the cultivation of the Pali imaginaire in Collins’ terms—which began with the commentarial work of Buddhagosa in the fifth century AD.Footnote 98 One particular Theravadin lineage, that of the Mahavihara nikaya, came to prevail. In one sense, this was in fact a victory for orthodoxy itself, since the Mahavihara, when set aside the other two Sinhala nikayas, the Abhayagiri and Jetavana, had an unusual preoccupation with the purity of doctrine.Footnote 99 In the ninth and tenth centuries AD, various kings began to promote the Mahaviharan lineage above the others and align themselves firmly with its rhetoric, a movement which reached its culmination in the abolition of other lineages in the twelfth century.Footnote 100 With all due regard for ‘historicist’ insistence on the mutability of ideology embodied in subsequent texts, there was now a distinctive Mahaviharan Theravada approach, and one which was readily exported to southeast Asia. In the fifteenth century, in the reigns of Tilokaracha in Chiang Mai (1441–1487) and Dammazeidi in Pegu (1472–1492), it became dominant as an instrument of state centralisation.Footnote 101
Localisation
Part of that Mahaviharan package was the notion of a particular connection between Buddhism and a territorial or political space, concretely represented in the accounts of the bodily visits of the Buddha himself.Footnote 102 Originally this was expressed in the idea that the island of Sri Lanka and its kings had been ordained to preserve the dhamma. This is clearly a key ideological component of the general localisation of Buddhism that Reynolds and Hallisey describe, and it is present as early as the Mahavamsa, most strongly in the symbolic resonances of the Dutugamunu myth.Footnote 103 Yet it may not have become a pre-eminent feature of royal discourses until the ninth century, and it is not until the later Polonnaruva period that it might be seen to correspond to a thoroughgoing localisation or acculturation of Buddhist institutions and practice.Footnote 104 Mainland Southeast Asian states then produced their own versions of this process—a paradoxical observation perhaps until we remember that exclusivist approaches can be the most exportable of cultural goods.Footnote 105
Again, the reification of territory in this way did not mean that the homeland became the only appropriate site for political or religious domination. On the contrary, an apprehension of divine destiny is likely to elicit a conviction of a right to just arbitration over the destinies of less favoured peoples or lands. Lieberman argues that ‘territorial versions of universal religions provided a critical source of legitimacy’ in the early modern period.Footnote 106 Indeed he remarks of all of his Eurasian protected rimlands that in this epoch, they ‘were frequently imagined as unique fields of holy sites, guarded by these deities, to whose protection the ruler ensured access by his piety and morality’.Footnote 107 In Burma, such convictions allowed its kings to represent invasions of other Theravada realms as pious reinstatements of pure Buddhism. Lieberman presents this ‘fusion of spatially-bounded doctrines and ethnicity’ as a key continuity between early modern and properly modern (i.e. nationalist) group sensibilities.Footnote 108 But in the case of Sri Lanka we surely then have to push back the continuity even further: we have seen that it was articulated as early as fifth century, and attained a certain pre-eminence when the Okkakas instituted the Mahaviharan worldview, as Jonathan Walters would have it, in the tenth century.Footnote 109 It is not surprising that this happened as part of an imperial courtly vision. Nor is it surprising that the thirteenth-century Pujavaliya would claim that Parakramabahu I set out to conquer overseas Indian lands because he wanted to rid them of false beliefs.Footnote 110
Vernacularisation and Popularisation
We already mentioned that Sri Lanka fits tolerably well into Sheldon Pollock's model of the ‘vernacular millennium’, in which the early centuries of the second millennium are critical. The religious manifestation of this tenth-twelfth century conjuncture is emphasised by Walters, who sees these early vernacular centuries as pivotal for ‘the deep integration of Buddhism into popular consciousness and practice’, the establishment of ‘domesticated Buddhism’.Footnote 111 This is when our corpus of texts becomes less dominated by Abhidharmic concerns as we find writings reaching out to the lay community in order to inculcate Buddhist principles and motifs. This is also when the sacred geography of the island is carved out. In other words the ‘newly coherent geocultural space’ that we see in the literature is part and parcel of a new mobility sustained by and sustaining the spread of Buddhist norms.Footnote 112 Royal centres are now drawing people in to partake in tooth relic festivals, or promoting the pilgrimage to the Buddha's footprint.Footnote 113 All of this is reminiscent of the modes of development that writers have proposed for early modern Southeast Asia.
And other definitions of what a popular Buddhism might mean render its history in Sri Lanka much older than that, certainly messier and more obscure. It is striking that even in the centuries BC there was a widespread dispersal of Brahmi inscriptions, usually signalling some form of monastic dwelling or stupa, throughout the island.Footnote 114 A network of Buddhist buildings had certainly penetrated the countryside by the end of the first millennium.Footnote 115 At least in terms of its geographical extension, Buddhism cannot be seen as a religion only of state capitals and great urban centres. By the ninth century, the increasing socio-economic predominance of the three nikayas based at Anuradhapura was transforming the lives, and perhaps therefore the religious sensibilities, of many layfolk through their monastic off-shoots dispersed throughout the island. Not only would many ordinary people have had monks as landlords, but the great monasteries were also becoming places where layfolk met for education, religious discussion and the arts.Footnote 116 The ninth is also the century from which Charles Hallisey charts the rise of newly intense forms of Buddhist personal devotion, arising in part from competition with and attraction towards the Saivite and Vaisnavite cults from the mainland.Footnote 117
Royal Control
When we come to consider the organisation of the Sangha and its relations with temporal power, we apparently have a much tighter focus on the twelfth century and the pivotal reign of Parakramabahu I (1153–1186). We have to cast Anuradhapura as the charter polity here, for it is only in the first millennium that the state existed alongside ‘autonomous religious institutions with extensive economic and social functions’.Footnote 118 Just as early modern Southeast Asian kings would see the re-organisation of the Sangha as a natural corollary of the centralisation and expansion of their rule, so Parakramabahu I's political feats were matched by his successful unification of the Sri Lankan Sangha. In a further foreshadowing of the manoeuvres of later Theravadin monarchs, this was pursued through the leadership of the ascetic forest-dwelling arannikas.Footnote 119 Here we do have a permanent transformation: after this point sectarian conflict per se ceases to be a major factor in Lankan history, as all monks are considered the heirs of the Mahavihara.Footnote 120 Gunawardana has argued that the three great nikayas had lost their institutional force during the chaos of the Cola invasions and the confiscation of their wealth by Vikramabahu I (1111–1132), and that a movement within this Sangha under threat called upon Parakramabahu I to bring them all under common leadership.Footnote 121 This then is the moment of the shift, in Reynolds and Hallisey's terms, from the largely autonomous and grandiose monastic centres of the ‘civilizational phase’ to the Sangha brought under the command of the centralising rulers of the ‘cultural phase’, ready to isolate monks from direct political power and intervene in their affairs at will.Footnote 122 A familiar corollary of this is that the Sangha was finally endowed with an organisational structure akin to that of the body politic, being answerable to the authority of the mahasvami (later sangharajas) whose position was similar to the mahasangharajas of seventeenth-century mainland Southeast Asia.
Even here however, it is easy enough to blur the outlines. There has been controversy, for example between Tambiah and Carrithers, over just how radical Parakramabahu I's approach was.Footnote 123 There certainly was ample precedent for the ‘traditional role’, as Gunawardana describes it, of purifying the Sangha, a royal prerogative exercised by many kings from the sixth to tenth centuries.Footnote 124 Equally, it is doubtful whether an effective organisational hierarchy was maintained throughout the subsequent disintegrations of central authority. Tambiah has argued that it was subject to the same tendency towards fragmentation as the galactic polity.Footnote 125 What did remain throughout the following centuries was the principle that the Sangha was always subject to the interventions of rulers. This meant setting up festivals for higher ordination (upasampada), purging the order of corrupt or incompetent monks, overseeing the maintenance of monastic regulations (vinaya), directing the production of texts, and so on. In the thirteenth century, for example, an influential set of monastic rules were drawn up under Parakramabahu II and recorded in the Dambadeni Katikavata.Footnote 126 In the subsequent periods of weakened central rule, high-ranking noble families took it upon themselves to discharge these regulatory and patronage functions.Footnote 127 The upturn of the fortunes of Buddhist institutions under Parakramabahu VI in the fifteenth century is then plausibly represented as a revival or a restoration, a particularly successful execution of very long-standing principles of royal conduct. The surviving records of katikavata (or vinaya rules) issued in his reign, for example, follow the Dambadeni Katikavata closely.Footnote 128
Parakramabahu VI appears to be an exemplary early modern monarch by virtue of the manner in which he oversaw the vigorous promotion of Theravada Buddhism: grand monastic centres flourished again (without reaching a point of financial advantage that threatened to undermine royal revenue), an ascetic or other-worldly-oriented fraternity vanavasi (‘forest-dwellers’), was supplemented by the gramavasi (‘village-dwellers’) fraternity, which had resident monks in most sizeable villages.Footnote 129 Centres of learning, pirivenas, were lavished with patronage, urged to copy and disseminate texts, and were to be found offering a range of secular subjects, presumably for lay consumption, beyond the sophisticated programmes for monastic neophytes.Footnote 130 Works such as the Saddharmaratnakaraya were commissioned that were intended to expound the dhamma in a manner accessible to laymen.Footnote 131 It is unlikely to be a coincidence that his reign is also marked by a renewed rush of contacts with Southeast Asian Buddhism.Footnote 132
Yet how innovative or distinctively ‘early modern’ all this activity really was, is a matter wide open for debate: much of it had been established practice for many centuries. And did this yet amount to a widespread immersion in what Steven Collins has referred to as the ‘Pali imaginaire’, that is, the mental world articulated and inspired by Pali texts? Collins sees this as an elite ideology that moved slowly downwards over the second millennium to become a peasant religion ‘at some point before or during the modern period’.Footnote 133 It is in a similar vein that some scholars have referred to the European Reformation and Counter-Reformation as the ‘Christianisation’ of Europe. Of course, in another sense, considering the massive presence of the Church in the lives of much of the populace in the preceding one thousand years, it would be peculiar to speak of medieval Europe as anything other than profoundly Christian.Footnote 134
The reign of Bhuvanekabahu VI (1469–1477) saw the leading elder Vidagama Maitreya defending the boundaries of Lankan Buddhism by pouring scorn on the Brahmanical forms filtering into all levels of religious practice.Footnote 135 But the Kotte kings did not align themselves consistently with calls for a rigid, exclusive orthodoxy. Anne Blackburn, for example, has eschewed the ‘early modern’ periodisation, partly because, even by the eighteenth century, there had not been ‘any attempt to codify the beliefs appropriately Buddhist to any degree than was customary from at least the twelfth century onward’, and nor is there any real indication that Lankan Buddhists became more concerned with ‘textually oriented forms of Buddhist practice and identity than they had been since at least the twelfth and perhaps the fifth century’.Footnote 136 In one sense the inter-dependent notions of orthodoxy and heresy are very old parts of the Sinhalese mental furniture, in that Pali and Sinhala texts have long argued about what constitutes right or wrong religious thoughts and deeds, pure or impure traditions and lineages, and are to be found castigating and abusing classes of opponents, and seeking to draw state power onto their side. But Blackburn is perhaps more concerned to guard against the application of any understandings of orthodoxy drawn from Christian or modern-day contexts.
Indeed it is important to consider that when state power and Buddhist patronage reach their apogee (when the galactic polity pulsated to its ‘strongest’ extent in Tambiah's terms), in the reigns of Parakramabahu I and VI, we also find religious and cultural pluralism. The Polonnaruva civilization which gave rise to the pivotal localisation of Buddhism, was also more open to the riches of Sanskrit tradition, to the aesthetics of Mahayana, to Brahmanic visions of sacred kingship than its Anuradhapura predecessor.Footnote 137 And we have now remarked many times on how the revival of institutional Buddhism under Parakramabahu VI coincided with an increasingly open and eclectic interest in other spiritual and aesthetic forms. Indeed the story of the thirteenth-to-sixteenth centuries could equally be written as an opening out to external religious and literary influences; current historiography emphasises ‘Hinduisation’ rather then the onward march of orthodoxy. That is, to adjust terms I have used elsewhere, boundaries were being endowed with rich significance just as they were being constantly traversed.Footnote 138 It is thus an enduring feature of Sri Lankan history that the sustained exposure to Indian or Hindu domination provoked both oppositional or exclusivist discourses as well as syncretic and inclusivist practices. The same insight has been observed in the relationship between Sinhala literary production and the great world of Sanskrit literature.Footnote 139 We know little about the extent to which Hindu cults were patronised by indigenous Sinhalese or on what terms. But it would be fair to describe Kotte—like Polonnaruva—as a state tolerating, even embracing many religious forms, while yet retaining a hegemonic position for Buddhism as the state-sanctioned religion par excellence.
Lieberman does not attempt to force the development of Buddhism too neatly into the patterned advancements of his administrative cycles: its fortunes were less convulsive. Yet Sri Lanka arguably stands even outside this capacious modelling, insofar as that it makes little sense to see Polonnaruva, so redolent of the ‘charter’ civilizations of Southeast Asia in many other respects, as espousing a rarefied Buddhism of the courts and urban elites that left the masses untouched, or to have entertained an essentially lassez-faire attitude to matters of doctrinal correctness, monastic discipline and organisation. This is one reason why the history of Sri Lankan Buddhism over the second millennium does not take on a progressive or teleological character, and why the chronicle tropes of decline and revival, or the structuralist conceits of oscillations over the longe durée—apparently so suspect in their ahistorical patterning—yet retain some appeal.Footnote 140
The European Interruption
For all that Lieberman's sites of progressive cohesion may be ‘protected’ regions they are very far from secluded: that would be to fly in the face of the burgeoning scholarship revealing the early modern period as an epoch of intensifying communication, seafaring exploits, ambitious commercial networks, economic trends that transcended local or regional particularities, the virus-like spread of certain ideas and fashions, the rapid diffusion of socially-explosive technologies such as firearms, the circulation of mobile elites and specialists happy to ply their trade in the port cities springing up all around the Indian ocean. This is the world that Christopher Bayly evokes with his phrase ‘archaic globalisation’.Footnote 141 Subrahmanyam too has heralded these connective forces as the defining feature of early modern world, and sees them reflected in the new regimes of sixteenth-century Asia which were characterised by a multi-ethnic open elite organisation.Footnote 142
Lieberman has therefore been concerned to evoke elite cosmopolitanism as a mildly paradoxical concomitant of ethnic solidification. If we refer back to the composite sketch of the fifteenth or sixteenth-century Theravada state presented earlier, it is easy to see the internal drivers promoting the embrace of these early modern energies.Footnote 143 The perennial competition for manpower with rival kingdoms would have been a powerful reason to incorporate foreign groups, particularly those that offered specialist military assistance. This would have become an almost irresistible option when the survival of the kingdom was at stake. Moreover many of these were governmental systems that allowed little institutional expression of political pluralism; if kings had allies it helped if their provenance denied them a permanent role in the polity or if they could be found outside of the indigenous interest groups that might threaten sovereign authority.Footnote 144 Foreigners placed under the direct control of the king, whether they were Chinese captains, Portuguese gunners, Brahmin ritual experts, Malay merchants or Japanese Christian bodyguards, would have formed a valuable counter to the perennial pretensions of the native nobility.Footnote 145
These small colonies of maritime groups bestowed cultural capital too, opening up diplomatic exchanges with far-off glamorous powers, and exuding all the power of the exotic: having extensive dealings with such foreigners was a mark of greatness.Footnote 146 So during the sixteenth century most kingdoms had important foreign groups residing at their capital cities, with those of Ayudhaya and Cambodia perhaps the most cosmopolitan, Burma somewhat less so.Footnote 147 Arakan (or Mrauk-U), a lesser power which is not considered one of the chief mainland states in Lieberman's theory, was notably cosmopolitan and best exemplifies in its Islamicisation the willingness to adopt the aesthetics of foreign courts.
Beyond these Theravada states, this was also the way of things in much of the rest of the Indic world into which the monotheistic mercantile-military groups sailed. It is hardly surprising then that some rulers—the most beleaguered or the most opportunistic—became increasingly reliant on these groups, sometimes to the extent of alienating themselves from indigenous sources of power. When a coastal centre is weak it will scavenge strength from the seas. In this light, the behaviour of the kings of Kotte from 1513 onwards is seen to be entirely typical, approaching the Portuguese as both a means of salvation from the designs of their rivals and an irresistible tool for their own ambitions. The king's personal guard had long been entrusted to foreigners; it now switched from South Indians to Portuguese. Just as Buddhist Kotte had been rather Hinduised up to that point, now Portuguese culture too was welcomed and Christianity at least patronised. Indeed, even when the Portuguese became identified as enemies, ‘elite cosmopolitanism’ ensured that their aesthetics would remain in vogue into seventeenth century, helping to fashion the court of their foe Vimaladharmasuriya I of Kandy (1591–1604).
But when does cosmopolitanism become a bridgehead for imperialism? On whose terms would these foreign groups reside? In mainland Southeast Asia they were normally harnessed to indigenous advantage, only occasionally presenting a threat at moments of crisis. They were more likely to hasten the centralisation of authority, providing armies with a new cutting edge, building and manning fleets and bringing lucrative world trade to its ports.Footnote 148 We could take the Toungoo king Tabinshweihti (1531–1551) of Burma as an example, whose military success owed so much to the shock value of his gun-toting Portuguese mercenaries.Footnote 149 No king there was subjected to the indignities of vassalage.
It was during the rein of Bhuvanekabahu VII (1521–1551), king of Kotte and nominal overlord of Sri Lanka, that the Portuguese came to dominate Lankan politics. The successes of a Tabinshweihti were no doubt the stuff of Bhuvanekabahu's dreams as he pursued his alliance with the Portuguese, but in his case the conjuncture turned against him, fatally undermining the authority of Kotte. From early on Portuguese showed themselves unwilling to be just one part of a multi-ethnic community under local rule. For all Bhuvanekabahu's diplomatic manoeuvres, they would not be incorporated on any terms but their own. What lies behind this divergence with the Southeast Asian experience? Lieberman tells us that, ‘by and large, the chief mainland states were too removed from the chief sea-routes, too immune to naval pressures, and too well integrated to exhibit the same sensitivity to European pressure as their archipelagic counterparts’.Footnote 150 As we have remarked, as a general explanation it makes sense simply to see Sri Lanka an honorary member of the archipelago. The matter of political integration is also highly germane, but here we need to introduce an important qualification, one that derives from a pulsating rather than progressive vision of the galactic polity. For if the Portuguese had arrived in the 1460s they would have found Kotte pulsating into its ‘strong form’, with Parakramabahu VI's grip probably too tight to be released by the machinations of a few dozen Portuguese. Instead they arrived at a time when centrifugal pressures would concentrate their impact.
In some ways Kotte had the worst of both the archipelagic and mainland worlds. It belonged, analogically, to the latter in that it seems to have derived the bulk of its revenue from the command of agriculture rather than commerce.Footnote 151 Commerce was obviously important to Kotte—indeed had determined its coastal location—yet the trade in the most profitable goods was left largely in the hands of foreign merchant groups.Footnote 152 In this it might bear some comparison to Mon and Toungoo Burma, but it does seem as if Kotte remained at the undeveloped end of the spectrum of early modern commercialisation: even Pegu appears to have been a more flourishing commercial centre, and one protected by its own navy.Footnote 153 While there was clear intent to exploit the value of cinnamon, neither Parakramabahu VI nor any of his successors had the maritime ambitions of his illustrious twelfth-century namesake. Yet neither could the sixteenth century kings of Kotte draw on a reservoir of agricultural resources of similar magnitude to their counterparts in Burma or Ayudhya. Indeed, some parts of the island were rice-deficient, and the island as a whole was a net importer of rice.Footnote 154 This dependency was a useful leverage for the Portuguese patrolling the Sea of Ceylon.Footnote 155
What Sri Lanka did have in common with many Southeast Asian islands was an attractive high-value crop, cinnamon, and a location in the middle of ‘chief sea-routes’. It was this that earmarked it as one of the bases of the first real programme of empire as envisaged by Afonso de Albuquerque.Footnote 156 Sri Lanka was also much closer to the heart of the Estado da Índia in Goa and was a more plausible candidate for imperial surveillance than the far-off lands of the shadow empire in the Bay of Bengal and beyond. In mainland Southeast Asia, by contrast, communities of private Portuguese were allowed to bed down under local authority. Arakan is a good example to turn to here, as a smallish state with a large and unruly settlement of Portuguese enjoying substantial autonomy from Goa. The Arakanese rulers struggled but ultimately succeeded in keeping a grip of the reins. The Portuguese could certainly pose a threat, most notably when they allied with the sub-king of Chatigão in the early seventeenth century. But the telling point here is that the Portuguese revolt was crushed without any reprisal from Goa. And the Portuguese at Siriam were useful at other times as a counter-balance to regional pretenders.Footnote 157
In the 1590s, the Portuguese in Sri Lanka abandoned their policy of indirect rule for one of major territorial conquest. We cannot embark here on the reasons behind this move, almost unique in the Portuguese East, but it can be seen in the light of a broader conjuncture of the Portuguese diaspora in the 1590s, in which ambitious frontiersmen dispatched schemes for the conquest of lands in Southeast and East Asia to the imperial centre. The Union of Crowns (1580–1640) with the Habsburg dynasty must be relevant here, as Castilian conquistadorism exerted a greater appeal.Footnote 158 The difference, of course, is that such arguments issuing from mainland Southeast Asia and the far east were generally ignored or allowed to issue in poorly-supported expeditions. In Sri Lanka the same conjuncture bore fruit in the rapid subjugation of the lowlands.Footnote 159
The Road not Taken? Some Notes on 1600–1800
Meanwhile in Lieberman's mainland Southeast Asia, the new states that arose out of the late sixteenth-century collapse would succeed in establishing a more thoroughgoing centralisation of power and integration of society: this heralded the shift to Pattern C governmental types.Footnote 160 In Burma this is represented by the administrative differences between the Toungoo empire and its successor state (the Restored Toungoo empire) based in the inland centre of Ava. Whereas Bayinnaung had governed the regions through closes relatives who replicated the insignias and rituals of kingship, Anaukhpetlun (1606–28), and Thalun (1629–48) moved to replace these bayin with lower-ranking governors, who were in turn supervised by spies and other central appointments.Footnote 161 And in Siam, King Naresuan the Great (1590–1605), managed to effect institutional reforms that had been suspended in the realm of theory since the fifteenth century.Footnote 162
What happened in Sri Lanka? Since we are at risk here of becoming overwhelmed by the imagery of trajectories and alphabetic sequencing, it may be profitable to give contingency its due and first of all briefly consider what might have happened. If Rajasinha I (1581–1593) had been able to take Colombo—as he came so close to doing—he may have been able to consolidate his sovereignty over the island and wrench his state onto a path towards something like Pattern C. He certainly had the dynamism and opportunism. The Sitavakan forces were quick to seize on the possibilities of Muslim and European firearms—these certainly played a role in their expansionary success.Footnote 163 Rajasinha also successfully reached out beyond the island by attempting to attract mercenaries from South India and elsewhere.Footnote 164 Most significant is his development of a navy, for which he drew on the expertise of Calicut and Aceh, and which became a real threat in the 1570s and 1580s.Footnote 165 Moreover, his commercial aspirations are certainly apparent in his control over the collection and trade of cinnamon. He was perhaps the first Lankan ruler to truly exploit its revenue potential; and this is another authentic marker of ‘early modernity’.Footnote 166 That the sea was becoming an important focus of activity for Rajasinha is indicated by Portuguese plans for a navigational blockade of the island 1587–1589.Footnote 167
Yet by this time, Rajasinha's willingness to provide tribute in exchange for gifts suggests that his maritime resources could be effectively neutered by such tactics. Sitavaka was always at risk of being isolated from the outside sources of strength whether they be the profits from diverse types of trade, the harbouring of large bodies of private foreign military specialists, or the availability of foreign models of administrative reform. We know little about his governmental programme; what stands out most from our sources is a polity directed largely towards war. In short, it was always more plausible that economic and political power would remain sundered as Lieberman suggests it was for island southeast Asia.Footnote 168
And indeed this was roughly the predicament of the only real candidate for Pattern C status, the Kingdom of Kandy whose rebirth came right on schedule in the first years of the seventeenth century. By this time the Portuguese were in command of the whole of the lowlands and the great bulk of the cinnamon trade. Kandy might be imagined at this point to be a ‘protected’ enclave—by virtue of the highland fastness at its core—in the centre of an otherwise unprotected island. I shall leave it for others to ascertain whether resurrected Kandy departs from its analogues by essentially reverting to traditional governmental strategies or whether it came to exhibit the integrative innovations of Pattern C. We might just begin the task with the following notes: in the seventeenth century, Kandy's administrative system appears to have attained impressive centralisation under the strong rule of Rajasinha II (1635–1687)—yet we ought to remember that this was a dominion not so much expanded beyond its core territories.Footnote 169 By the eighteenth century, however, historiographical summations start to sound undeniably familiar. John Rogers has recently written that ‘the general trend within Kandy in the eighteenth century was towards greater political and ritual integration of the kingdom's population . . . . The centralisation of power was also linked to the reconstruction of Kandyan Buddhism in mid-century, which resulted in . . . . a general tightening of the ritual and social order’.Footnote 170
The Kandyan kings, in common with their illustrious counterparts in mainland Southeast Asia, sought to attract foreign elites and were quite able to play the various monotheistic groups off one another. Again, the re-founder of Kandy, Vimaladharmasuriya I (1594–1604) affords an excellent illustration of the transferral of ‘early modern’ skills, for his military apprenticeship was spent fighting for the Portuguese under the baptismal name of Dom João d'Austria. When he assumed power in Kandy, he fortified the city, formalised the use and manufacture of firearms and sought alliances with the Dutch in an attempt to blockade Portuguese-held ports.Footnote 171 All this is true enough, but over time Kandy would suffer greatly from its isolation, its inability to develop a naval presence and to glean profits from seaborne commerce.Footnote 172
In its re-affirmation of Theravada Buddhism under the auspices of Vimaladharmasuriya, Kandy certainly had much in common with new dynasties in seventeenth-century Burma, Siam and the other protected rimlands, with their intent to ‘restore a cosmically sanctioned hierarchy’ after the ‘excessive mobility’ and chaotic bloodshed of preceding decades.Footnote 173 Further comparisons suggest themselves between the elevation of the person of the king above the intimacies of normal intercourse in seventeenth and eighteenth-century Kandy and representation of kingship during the height of royal absolutism in Ayudhya. The Kandyan period also provides us with the best evidence for the creation of a religious moral community with the king at its centre; a whole people, that is, with a stake in the religious efficacy and propriety of the monarch.Footnote 174 What requires much greater analysis is the extent to which all this represents a real transformation from the Kotte period.
It is in the sphere of the monastic institutions that the disruptive consequences of colonial activity are unmistakeable. This is why if we were to adopt a teleological language and judge Sri Lankan Buddhism by the progress of its peers in mainland Southeast Asia, it might appear slow-to-develop as well as remarkably precocious. The political context was hardly conducive to the progressive strengthening of monastic organisation; indeed it is possible that from roughly the mid-sixteenth century until the introduction of the Siam Nikaya in the eighteenth century the Sangha had no effective and comprehensive centralised hierarchy at all.Footnote 175 Let us be wary of assuming that this administrative disintegration would have fatally undermined the moral authority of monks, their relationship with village society, or even the persistence of Buddhist patterns of thought and behaviour. In 1638, after many war-ravaged decades in which institutional Buddhism was attacked throughout much of the country, we yet find a hostile Portuguese soldier reporting on the respect for and discipline of the ganinnanses (unordained monks).Footnote 176 Buddhist monasticism was not engineered to approximate the autonomous secular-style efficiency of the Christian Church. Tambiah has famously argued that if secular and spiritual authority did find a meeting-point, it was in the person of the king, who provided Buddhism with its sense of leadership and cohesion.Footnote 177
Nevertheless, for large periods of the sixteenth and seventeenth centuries, the kings of Kotte, Sitavaka and Kandy found it difficult to advance and develop that role. The decline of monastic institutions and the regular collapse of higher ordination lineages have been discussed often enough. Naturally, education suffered too as Buddhist textual production diminished and the large pirivenas fell out of view, along with any signs of a comprehensive curriculum.Footnote 178 This is then quite contrary to the trend in the ‘Pattern C’ states of Southeast Asia towards such procedures as centrally-regulated examinations for the Sangha—at least until the arrival in mid-eighteenth-century Sri Lanka of the Siam Nikaya from Ayuddhya.Footnote 179 ‘Connectedness’ would ensure that the hiatus of development was overcome at this point, and Sri Lankan Buddhism once more assumed a more obvious place within the larger flow of Theravadin history.
If we were to conclude from all this that Sri Lanka appears in this way as something of a victim of the early modern conjuncture, its inhabitants were certainly not about to quietly adopt that role themselves. The most sustained and ambitious territorial endeavour of Iberian colonialism in Asia (with perhaps only the Philippines bearing comparison) produced also some of the most recalcitrant resistance. First, in mid sixteenth-century Sitavaka, and then certainly during the low country rebellions and hill-country resistance of the 1590s onwards, we see the result of sustained warfare with outsiders in an enhanced and politicised indigenist or ethnic consciousness.Footnote 180 This is one of the key developments that Lieberman highlights in his schema for the protected rimlands.Footnote 181 The Portuguese and Dutch were, of course, highly obvious outsiders, and all who rejected their dominion could and sometimes did regroup around Kandy. Moreover, the symbolism of cultural and religious antagonism became an integral part of rebel campaigns and sometimes Kandyan propaganda too.Footnote 182 Thus at the same time that we see in the northern end of exposed South Asia a certain limited commingling of political and religious antipathy as a result of Islamic conquests, at the southern end of the region—Sri Lanka now having become part of this exposed zone—we find the same effect produced by Christian imperialism.Footnote 183
By the time Queirós was writing his retrospective late in the seventeenth century it seemed to him that the Sinhalese ‘in obeying their native Kings they have always been various and inconstant, but most stubborn in not admitting foreign domination, and when the Portuguese entered Ceylon . . . they did not hesitate to submit to any bold rebel, in order to recover their liberty’.Footnote 184 Portuguese writers of the seventeenth century simply present the Sinhalese as a distinct people (nação) with particular national characteristics and an origin myth (for both kings and commoners) in the Vijaya story.Footnote 185 Whether or not Kandy ever shifted into a genuine Pattern C state, a certain king-centred patriotism can certainly be glimpsed.
Conclusion: Further Comparative Angles
The majority of this paper was written using mainland Southeast Asia as a comparative foil. From that perspective, it is the particular kinship of Lankan kingdoms with the states of this region that is striking. Even after 1500 Sri Lanka continues to be analogous to mainland Southeast Asia in terms of its basic religio-political matrix, just as it becomes analogous to island Southeast Asia in terms of its geopolitical predicament. And what, after the monotheistic interruption got under way, did this ultimately result in? A future that also looked both ways: before the final coup de grâce by the British it was allowed much less opportunity to develop the kinds of indigenous political integration and economic development that we see in the Pattern C and D polities until the eighteenth century, but it was ultimately able to resist the religious transformation. However cleverly maintained by the Oratorian mission, with the waning of Portuguese power there was little chance that the majority of Sinhalese would become a Catholic people, and as for Dutch and British Protestantism, that withdrew leaving barely a trace. The antiquity and entrenchment of Sri Lanka's Buddhist traditions would see to that.Footnote 186
However, in the broader Eurasian perspective that will be developed in Lieberman's forthcoming work, island and mainland Southeast Asia are presented as forming a single zone of ‘protected rimlands’ before 1500, following a roughly comparable chronology.Footnote 187 It is only after 1500 that the archipelagic region is forced down a different historical path by the arrival of newly successful seafaring groups. From this viewpoint it is quite clear that Sri Lanka follows the same historical logic as island Southeast Asia.Footnote 188 In both areas the sixteenth century saw an intensification of competition which produced some indigenous polities more than capable of repelling European advances, but thereafter serious territorial inroads were made nonetheless. It is only that in the archipelago the Dutch would become the major conquest elite, while in Sri Lanka the Portuguese would institute a rare policy of conquest. If we wanted to single out one island in particular, perhaps the most fruitful analogy here would be Java, whose people were culturally closest to the mainland states by virtue of their sustained Indianisation, and which also had a tradition of all-island overlordship and a sense of territorial distinctiveness.Footnote 189
Before 1500, then, we could think of Sri Lanka as an honorary protected rimland—if the only one from the southern limit of Eurasia, with the rest bunched along the eastern and western extremities. The only major difference between Sri Lanka and the other protected rimlands in this phase remains the very early flourishing of civilization in the northern zone and the subsequent striking sense of continuity dating back to that time. For Anuradhapura has to be seen as the charter state, of which Polonnaruva was merely a reformulation. What analogies can we find here? Sri Lanka's chronology bears some comparison with China, if one is tempted to make Anuradhapura the equivalent of the Chinese ‘charter’ period, the Han (third century BC), and Polonnaruva equivalent to the second period of imperial unity, the Song (960–1279).Footnote 190 Of course, the analogy quickly runs aground if one is concerned with the respective extents of land and population brought under political unity, or what ‘unity’ actually entailed in administrative terms, to mention only a couple of many points of contrast.Footnote 191
It should have become clear that Lieberman's model offers an exceptionally fertile means of thinking about national and regional histories in a new light, and I suspect it will be useful for the reframing of much grander historiographies than Sri Lanka. But can we refine Lieberman's categorisations at all? If we leave to one side Sri Lanka's unusual civilizational precocity, one might be able to pursue a sub-category of ‘island rimlands’.Footnote 192 This immediately raises the question of the comparability of Britain and Sri Lanka. Both seem to be protected in a rather distinctive way before 1500, that is to say, they were not distant enough from the mainland for the sea to form an insuperable barrier—therefore there were more than enough threats from outside (Danes, Normans, French for Britain; Colas, Pandyas etc for Lanka) for a profound sense of alterity to develop. But the sea lent these islands enough protection to ensure that indigenous polities would always have a redoubt from which they could re-emerge and eventually assimilate the conquerors. Moreover the size of these islands may be significant: large enough to generate reasonably powerful polities, but small and compact enough to allow for the rapid diffusion of ideas and institutions. I hope to explore at another time the coincidence by which in the mid-first millennium both islands produced a text (or textual tradition) that offered a certain localisation of religion conveyed through the vehicle of a universal language, which could serve as a blueprint for subsequent imagery of a chosen sacred people and land—Bede's Ecclesiastical History of the English People and Sri Lanka's Mahavamsa.Footnote 193 Both then saw early vernacularisation.Footnote 194
From 1500, the sea shifts more from a military barrier (to both invasion from without and expansion from within) to a military conduit. In England this conjuncture presented an opportunity, as it would soon be able to expand its influence in the outside world. And in Sri Lanka, lacking a navy and politically fractured, it became a threat. One is tempted to see how well Japan would fare in this category, given that these three islands are moored off the western, southern and eastern coasts of the Eurasian landmass.Footnote 195 If Japan was by far the most protected of these islands, Sri Lanka was the least. Combine this with Sri Lanka's inheritance of the Indic system of hierarchical kingship (a.k.a. ‘nested sovereignty’, ‘galactic politics’), and it is clear why all-island unity was comparatively rarely achieved in Sri Lanka and why it was not in a position to take advantage of the post-1500 conjuncture.Footnote 196
What did result, however, was a sense of continuity, both subjectively and objectively considered, that it is difficult to find parallels for elsewhere in South Asia. Indeed, if this paper began by referring to the—on the face of it, bizarre—fact that Sri Lanka can often seem moored in an entirely separate side-water from the main currents of South Asian historiography, we can end by restoring some sense to that fact. This is, of course, not to deny the massive influence of India, particularly South India, on Lankan history. But it is not simply on linguistic or religious grounds that Sri Lanka seems different. Rather, Lieberman's model allows us to see how it belonged to a different category of Eurasian politico-social development. In the rest of India there were strong geopolitical reasons why religion, ethnicity and politics would tend to become comparatively detached from each other.Footnote 197 Whereas in Sri Lanka, its relatively protected status before 1500 allowed the Sinhalese to develop certain territorial, ethnic and patriotic ties of affect, that—however apparently undermined by political polycentricity and the social divisiveness of caste—would be further enhanced by the experience of European conquest.Footnote 198
