The habits of both European and Northeast Asian historians have encouraged a quest for early state formation in Southeast Asia, as the principal theme for the proto-historic period before written sources become abundant. Chinese sources from the third century of the Common Era report the existence of states that sent tribute from the tropical regions in the south, and scholarship focused on how to connect the Chinese toponyms so recorded with material remains in Southeast Asia, to determine the site of these kingdoms. Recent archaeological work in Viet Nam, Thailand, and Malaysia in particular has identified relatively complex societies from about the beginning of the Common Era, leaving evidence of pottery, metalwork, elaborate funerary arrangements, and trade as represented by imported beads, coins, and metalware. Until the sites speak for themselves, however, through inscriptions in Indian languages from about the sixth century, the naming of sites, let alone “kingdoms,” is highly speculative. Even in the period of relatively abundant inscriptions and temple remains from the eighth to thirteenth centuries, the labeling of kingdoms, cities, and ethno-linguistic communities on the map of Southeast Asia rests on rather fragile foundations. Economic and religious activity is well documented in the material evidence, but political structures and identities much less so.
This chapter will take a more conservative approach to the identifying of states than is common among historians, preferring to trust the evidence of the archaeologists as to what can be known. In particular, the Chinese record on the arrival of missions from barbarian southern “kingdoms” is extremely important in regard to commercial networks, production, and exchange, but should not be translated into modern assumptions about states. Traders understood that the best means of access to the huge market of China was to present themselves as tribute-bringers from a barbarian kingdom, particularly one that was recognized at the Chinese court from older documents regardless of whether it represented any continuing reality.
The deciphering of inscriptions in scripts, imagery, and often languages derived from India began a major debate about the nature of the “Indianization” which appeared to have transformed Southeast Asia between the sixth and fourteenth centuries. The Greater India school of the 1930s raised excitement in India about the spread of Indian culture outward through colonization – a wording particularly unacceptable to nationalist Southeast Asian of the 1950s and 1960s. Much post-war scholarship was devoted to seeking to show that it was local genius and adaptation rather than foreign colonization that built the states and their temples. This debate had the unfortunate effect of dividing Indian from Southeast Asian scholarship, and these have only begun to find each other again in the current century.
For western scholarship the classic text was George Coedès’ 1944 synthesis of the history of the états hindouisés, usually translated as “Indianized states,” of the region (Coedès 1968). It was widely relied upon by D.G.E. Hall (1968) and later authors. These pioneers constructed a coherent narrative of how “kings” may have succeeded one another, at the cost of beguiling later writers into a simplified idea that the same few “states” built the temples, left the inscriptions, and conducted relations with China.
An exemplary case is Funan, hailed by Coedès, Hall, and subsequent textbook-writers as “the first Indian kingdom.” Funan appears in Chinese records from the third century CE, though no such place-name occurs on the ground in Southeast Asia. Coedès (1968, 36) proposed a derivation from the old Khmer word bnam (modern phnom), “mountain,” making this an essentially Khmer state and antecedent of Angkor. Others argued it was a maritime state and therefore Austronesian. Coedès ingeniously linked the various stories about Funan in Chinese sources of the third to sixth centuries with later echoes in inscriptions from the sixth century, to propose the existence of a kingdom born in the first century CE in the Mekong delta region, but expanding through conquests in the third century until it dominated much of southern Indochina and the Peninsula (Coedès 1968, 36–42). Later historians and textbook writers developed this idea to the point where the first six centuries of the Common Era are explained as “the Funan Age,” as if the whole introduction of Indian religious ideas and writing was centered on this state.
The success of this Funan idea owed much to the fact that the first archaeological site of that period to be analyzed scientifically was at Oc Eo, today about 25 km from the Gulf of Thailand in the westernmost frontier of southern Viet Nam. Louis Malleret explored this walled city in the early 1940s as Coedès was preparing his book, and the two men agreed that this site, with its abundance of Roman and other coins, glassware, and medallions, must have been the port of the state of Funan. Since then there has been much more archaeological work, including other sites nearly as old as Oc Eo in the Mainland, the Peninsula, and West Java. There was trade and exchange, but were there states? Even the later temple-building civilizations that scholars labeled Angkor, Champa, Sriwijaya, and early Mataram have become less unitary and state-like as we learn more from archaeology. Oliver Wolters came to believe that in the early centuries of the Common Era “a patchwork of small settlement networks of great antiquity stretched across the map of Southeast Asia” (Wolters 1999, 16). There is plenty of evidence for a sophisticated bronze culture but little for states being responsible for it.
The remarkable dispersion of large bronze drums around Southeast Asia gave rise to some of the earliest scholarly work on the distinctive pattern of Southeast Asian proto-history. How metals technology spread in Southeast Asia in the earliest periods is a matter of long-standing debate, more intense since the excavations at Ban Chiang near the middle Mekong River in the 1970s raised the possibility that bronze-working there may have been as old as that in China. The earliest Ban Chiang periodization has now been largely discredited, however, and a consensus is emerging that the Southeast Asian bronze age, including vitally important sites in what is now southern Yunnan but was then very distinct from the older “Chinese” tradition of the Yellow River, began less than 4,000 years ago. Unlike the other major bronze traditions of Eurasia, the spread and development of sophisticated metallurgy in Southeast Asia involved not states nor even urban settlements but village communities linked by complex exchange networks. Bronze was adopted for ornamental and functional forms already familiar in the region in stone, ceramics, or shells. Bronze items joined the repertoire of prestige goods buried with the dead. Yet these burial sites suggest not hereditary chieftainship let alone kingship, but rather the honoring of individual men and women of achievement – including female potters buried with their tools.
The first millennium of this distinctive Southeast Asian tradition included production sites in Yunnan, the Red River area, the Khorat plateau, the middle and lower Mekong, and upper Chao Phraya. There is much overlapping of stone, bronze, and iron tools, although bronze axes, arrowheads, chisels, and adzes dominate in the first millennium BCE. Many of the finest surviving bronze and iron artefacts, together with ornamented pottery, have been found in burial sites, which demonstrate some complexity of ritual even in the small and scattered communities that are the hallmark of the region. There were settlements that specialized in pottery or bronze casting, and others that mined the copper and tin necessary for bronze casting from deposits hundreds of kilometers apart. Although there is no evidence of substantial cities before the Common Era, or of any overarching political authority, there must therefore have been extensive networks of trade and exchange between specialized settlements over long distances. The reciprocities that existed between coastal and interior producers, agriculturalists, fishermen and hunters, miners and metalworkers provided the basis for “civilization without cities” that was a distinctive long-term feature of Southeast Asia.
In the humid center of Southeast Asia, including Sumatra, the Peninsula, and western Java and Borneo, there is predictably little evidence of bronze working at all before the Common Era, as conditions for settled agriculture were difficult except in some highland valleys. The earliest evidence of bronze artefacts in Island Southeast Asia occurs in the last centuries before the Common Era, as traded items from the Mainland. The most spectacular finds dating back to this period are the large “Dongson” bronze drums apparently traded to settled agricultural communities in the uplands of Sumatra and Java and in Bali. The drums found in Sumatra were in some of the lakeside highland valleys furthest from the east coast, where they appear to have given stylistic inspiration to a number of megaliths and stone tombs assumed to be of the same period. These drums were therefore traded over large distances, from the Red River valley production centers by sea to Palawan, Borneo, and as far as Java. Some went to Trengganu and Pahang in the Peninsula. From there they were evidently carried by portages across the Peninsula, where others have been found in Selangor, and again by ship and land across the Malacca Straits and up the long eastern rivers of Sumatra to Besemah (Pasemah) and other highland centers of agriculture. Through this arduous trade route they must have assumed great value, which implies relatively complex societies to buy them. Numerous smaller drums, perhaps made in the Islands, were traded throughout the south-easternmost islands (Sumbawa, Flores, Timor, Alor) at a later period, perhaps beginning in the second century CE.
The finest productions of these non-state societies of the Islands before the Common Era were ceramics, with distinct traditions probably dispersed southward from Taiwan along with the Austronesian seaborne migrations. A cave site in Palawan (western Philippines) revealed a “soul boat” burial jar dated to about 800 BCE, linked in style to the findings of a similar period in other islands and the adjacent Indochina mainland. In this period the Islands were importers of complex bronze drums from the Mainland centers of the Red River, but not yet producers.
The earliest evidence of bronze casting in the Islands has been found in Bali and Palawan, dated to the third and second centuries BCE. Copper does occur in Palawan and nearby northern Borneo, and is distributed widely also in Sumatra, west Java and western Borneo, but tin, the other essential for making bronze, is available only in the seam down the whole Peninsula and ending in Bangka and Belitung off eastern Sumatra. For bronze to be cast in Bali and Palawan, therefore, there must have been exchange networks already in place around the Archipelago. As well as copper and tin, gold and iron were widely mined, smelted, traded, and worked into ornaments and utensils by this period. We must wait for the first centuries of the Common Era, however, for specialist bronze production centers close to sources of copper and tin to emerge in the Archipelago, producing such gems as the bronze weaver discussed in Chapter 1. Iron-smelting may have been a little older, since the discovery at one of the major portages across the Peninsula, in Bujang Valley, Kedah, of a major iron-smelting site active from the third to sixth century CE suggests already considerable complexity in the smelting and trading of iron tools.
In early times, Austronesians were among the Asian pioneers in boat-building and navigation, and there seems little doubt that their activity linked the islands of Southeast Asia to each other, to the Mainland, and to India and China well before the Common Era. Austronesian traders from Southeast Asia appear to have reached Madagascar and the east coast of Africa by the onset of the Common Era, gradually colonizing the former in the following centuries. Enough of these pioneer navigators may have carried with them the gold of Sumatra and the Peninsula for the region to become known as Suvarnabhumi (gold-land) in the Indian epics as early as the third century BCE.
With the exception of this gold, and the tin of the Peninsula, metals remained less exploited in the region than in its better-populated neighbors, China and India. Travelers from Europe and China found that their everyday nails, knives, and needles were in great demand from the locals, chiefly because the sparse population in the vicinity of most of Southeast Asia’s minerals required a smaller scale of production, and therefore less efficient methods of both mining and smelting than developed in China. As trade developed in the second millennium of the Common Era, the flourishing Southeast Asian ports found it cheaper to import everyday metal items from afar, and their own local traditions of mining and smelting gradually retreated to the less accessible interiors.
Bronze statuary is one of the defining legacies of Indic religion everywhere, but many of the earliest Buddhist and Hindu images produced in Southeast Asia were in gold or terracotta. Bronze in the region was by no means first or primarily used for these universal religious purposes, but for everyday tools, ornaments, and celebrations of the dead.
The process once called the “Indianization” of Southeast Asia has been better explained by Sheldon Pollock (2006) as the rapid spread of a Sanskrit cosmopolis from a few priestly centers in northern India throughout all the commercially linked power centers of South and Southeast Asia. Having been for centuries a sacred “language of the gods,” taught to a few priestly men so that they could recite the activities of the gods in the Vedas, Sanskrit embarked on a new career as a universal legitimating language of power around the beginning of the Common Era. We might explain this as the first stage of globalizing religion, contemporary with the Roman Empire giving birth to Christian universalism in the West and with Han power extending a Chinese cosmopolis through Northeast Asia. While no comparable political power center arose in Southeast Asia at this or any other time, those engaged in the multiple maritime exchanges which ultimately linked Roman and Han Empires must have been unusually receptive to universal ideas. To judge by the inscriptions, by the fifth century CE Sanskrit had expanded not only throughout Southern India and Sri Lanka but also to Southeast Asian sites in Borneo, Java, the Peninsula, the lower Mekong and the centers of Chamic (Austronesian) settlement in what is today central Viet Nam. So successful did it become as a “celebration of aesthetic power” that its two new genres of courtly poetry (kavya) and royal eulogies (prasasti) spread to peoples of multiple languages and habits throughout South and Southeast Asia. It marked the beginnings of writing, of kingship (presented as universal, not bounded power), of scriptural religion (primarily through Buddhism), and of literature.
This expansionary impetus was facilitated by the mobility of Buddhism. Although early Buddhism rejected Sanskrit as identified with the established hierarchical order, by the third century CE there had been a return to using the universal language, now purged of its offensive features. For various reasons early Buddhism established a symbiosis with trading networks which made possible a considerable expansion of Indian Ocean trade in the first centuries of the Common Era. Buddhist advocacy of detachment from worldly engagement had the effect of rendering caste distinctions less restrictive, and breaking the caste taboo against travel by sea. The organization of the sangha (monkhood), and various practices of lay support for it, proved able to be transplanted readily across barriers of place and culture. Although the earliest contacts of China with Buddhism were through Central Asian overland routes, some of the earliest Chinese pilgrims discovered in India that it was easier to return by sea. Faxian did this round trip in the early fifth century, and by the time of Yi Jing (635–713) the sea route was well established. In 671 he set out by sea from Guangzhou, and traveled along the trading networks already established by Southeast Asia–based traders. Of the 24 years (671–85) he devoted to studying, traveling, and collecting scriptures along the route to the holy places in India, ten were spent at the place he called Shilifoshi, evidently the most important base for the Buddhist trading network in Southeast Asia. There, he wrote, “there are more than a thousand Buddhist priests whose minds are bent on study and good works … their rules and ceremonies are identical with those in India” (cited Coedès 1968, 81). He therefore recommended Chinese pilgrims to use this multicultural Buddhist center to learn their Sanskrit and Pali.
Historians have equated Yi Jing’s Shilifoshi with the Sriwijaya mentioned in some inscriptions, and estimated modern Palembang, 60 km up the Musi River in south Sumatra, to be its likeliest location. Certainly there was a major trade center there, at an important junction between sea routes through the two major straits and river supplies of rice from the well-cultivated south Sumatran highlands. It remains puzzling, however, that Palembang offers no major Buddhist remains comparable to Java’s Borobudur or even other Buddhist sites in Sumatra – Muara Jambi, Muara Takus, and Padang Lawas. Certainly there were religious and commercial connections between various Sumatran, Javanese, and Peninsula sites which shared the cosmopolitan cultures of early Sriwijaya, and Palembang may be best understood as primarily a node of interaction between rice-growing centers in highland Sumatra, Java, and Cambodia and the trade and pilgrimage to China and India. Buddhism made Sriwijaya a key player in Asian interaction from the seventh to tenth centuries. Some of the foreign monks expelled from China in 741 traveled to Southeast Asia and Sri Lanka, of whom the most famous, Amoghavajra (Bukong in Chinese, 705–74), returned to China in 750 with hundreds of texts and became one of China’s most influential Buddhist teachers. A ninth-century copperplate found at the great center of Buddhist scholarship and devotion in Nalanda, northeast India, records the endowment of five villages there for the upkeep of a monastery built by the Sailendra king of Suvarna Dvipa (Sumatra), to accommodate visiting monks from the “four quarters.” The famous Bengal-born monk who established Vajrayana Buddhism in Tibet, Atisha (980–1054), had apparently learned that doctrine while studying in Sumatra, at the feet of a famous teacher. Southeast Asian ports and sacred sites in this period were part of a pan-Asian ecumene of Buddhism. As a ninth-century Javanese Sanskrit text put it, “a constant flow of people” came to Java from places in India and Cambodia, “bowed low with the devotion to the Buddha” (Pollock 2006, 127).
The earliest Buddhist imagery understandably occurs along the major trade routes connecting China and India, notably at some of the trans-Peninsula portages and in southern Sumatra and tin-producing Bangka. By the sixth century, Buddhist networks were established from India through Southeast Asia to China. Buddhist statues and inscriptions found in Southeast Asia in this first phase (up to the seventh century) are exclusively religious in character. No king or state is mentioned as commissioning them, though one from the Peninsula notes that it was engraved by a ship-master (mahanavika). They share a number of features, including a small range of favored Buddhist texts repeated in many inscriptions, and three common genres of standing Buddhas and Avalokitesvaras. Their Sanskrit writing also evolved in a common way distinct from the Indian original, demonstrating the importance of intra-Southeast Asian Buddhist networks and pilgrimage.
The gods of Hinduism were not left behind, however, in this expansion of Buddhist imagery and Sanskrit language. Images of Vishnu with a characteristic miter on his head occur at some of the earliest Buddhist sites in Java and Cambodia, and all the associated inscriptions make claims to royal power. It appears, therefore, as though Buddha and Vishnu traveled together along the trade routes, with Buddhism a more popular belief system uniting traders from different parts of Asia, while Vaishnavism was taken up by those seeking legitimacy for novel forms of hierarchy and kingship in Southeast Asia. Vishnu was the preeminent deity of the Gupta Emperors who ruled northern India from the fourth to sixth centuries, so it was not surprising that his royal cult should appeal to the ambitious. One of the earliest such Sanskrit inscriptions, from the late fifth century, was far up the Mekong in what is today southern Laos, about a king named Devanika who built a reservoir and claimed to have inaugurated a kingdom emulating the gods of the Mahabharata. Another such was Mulawarman, who left seven stone pillars at a site in Kutai, 30 km upriver from the eastern coast of Borneo, with Sanskrit inscriptions, estimated by style to be fourth or fifth century, claiming to have conquered other kings. The Pendawa king-conqueror of the Mahabharata saga is thus made the legitimation for a novel concentration of power in Borneo, perhaps inspired by a Brahmin traveler who could provide an Indic form of consecration. Similarly in Java, the earliest dated inscription, from 732 CE, records the erection of a lingga, the preeminent symbol of Shiva, in the Kedu plain of south-central Java by one King Sanjaya who evidently used this Sanskrit and Shaivite legitimation for the first definite claim to kingship in the Javanese-speaking area.
Sanskrit courtly genres such as poetry and eulogies spread to peoples of multiple languages and habits from Gujarat in the west to Java and Cambodia in the southeast, praising local kings in exemplary Sanskrit as learned, beautiful, brave, just, and generous. An exalted rhetoric threatening cosmic curses for defiance but rewards for obedience became the principal weapons of big men seeking to rise above the established networks of reciprocity and kinship. Even Tang China, which accepted “tribute” implying trade access only from a single Java “king,” acknowledged there were in fact 28 “small countries” in that island.
The earliest centers of Indian religious ideas in Southeast Asia were naturally along the routes of trade from the Indian Ocean to China. The watery world where the Mekong’s many mouths meet the sea was evidently a key part of this route in the earliest centuries of the Common Era, as were portages across various parts of the Middle and Upper Peninsula. From the seventh century, inscriptions became numerous here, both Sanskrit ones addressed to the gods and Khmer ones dealing with everyday affairs. One Chinese report told of “Funan” having been established by a Brahman from India, who sought to unite people into villages around a reservoir or tank, rather than digging individual wells. The religious ideas from India do indeed appear to have established a Khmer pattern of settlement clustered around a tank or reservoir which maintained essential water supplies through the dry season, and a sacred shrine, lingga, or stupa which might incorporate both the new idea of merit-making through religious devotion and reverence for the ancestor-figures who had founded a particular community. The inscriptions show that many communities included both an elite of priests and rulers who administered the rites to these sacred places and a variety of bonded people who served the former in a great variety of modes of slavery or loyalty.
In this complex world of interconnected but independent villages, one larger temple complex was built in the seventh century at Isanapura (now Sambor Prei Kuk, near modern Kampong Thom), midway between Angkor, the royal capital from the ninth century, and the greatest concentration of more ancient sites on the rivers below Phnom Penh. The main temple among the more than a hundred constructed within the complex is to Shiva, and there is a king, Isanavarman, associated with it in the inscriptions. But it was popular piety that built the temples, while the kings achieved their power by association with the sacred site. Perhaps the people were inspired by pasuputa, wandering ascetic Brahmans who became a feature of southern Asia in the fifth and sixth centuries, and appear to have preached devotion to Shiva as the ultimate form of cosmic protection. Along with Mahayana Buddhist monks, they were the agents of broader unities of culture, convincing an increasingly mobile population that in addition to local deities of place and ancestry, universal protectors were also needed to ensure fertility, security, and prosperity. Because cosmic forces were everywhere, the goal of all was to realize the divine within, and the king, who by definition must have concentrated more of the sacred, was thereby an exemplar.
In the Irrawaddy River region we know today as central Burma, colonial and nationalist scholarship had posited a city, kingdom, and center of Mon-speaking Theravada Buddhist civilization named in later chronicles as Thaton, somewhere in the Irrawaddy delta. Archaeological research failed to find it, however, but rather located a pattern of small walled towns centered in the dry zone of the middle Irrawaddy, conventionally known as the Pyu civilization. Michael Aung-Thwin (2005) has effectively demolished Thaton as a latter-day myth, and proposed rather that we accept the Pyu civilization in most of the first millennium CE as the builder of a Buddhist shared civilization in what later became Burma.
Many earlier scholars assumed that the wonderful ninth-century monuments of south-central Java, including the Buddhist Borobudur and the Shaivite Prambanan, must have been built by powerful kings. The problem of two religions was resolved through the notion of a Sumatran invasion under a Buddhist Sailendra Dynasty that gave rise to the Buddhist monuments. More careful study of the physical evidence, however, suggests Hindu and Buddhist shrines coexisted, while there was no great urban center or palace complex such as might elsewhere be associated with temple-building kings. Most of the early monuments of the Mataram area are in fact Buddhist constructions, and no inscriptions have been found attributing their construction to a king. The earliest, Candi Sari and Candi Kalasan, built in the area of later Prambanan in the late eighth century, are massive-walled shrines to the Buddha, or the female bodhisattva Tara, with two stories intended also as dwellings for monks, devotees, or pilgrims. Candi Sewu, a great mandala of 240 Mahayana Buddhist temples and stupas, was erected at a similar period, evidently through the piety and devotion of a numerous population in this fertile rice bowl of Java. The only inscription mentioning the greatest of all such mandalas, Borobudur, is one dated 842 near to the temple site, recording the name of a woman who bestowed land to sustain it. A high degree of social cohesion must have been required to organize the labor and skills for these vast and exquisite building projects, but it appears to have been provided by charismatic religious leaders rather than kings.
The extent of popular enthusiasm for making merit and escaping the wheel of suffering through the teachings of Buddhism can be measured by the exceptional proliferation of stone buildings that have survived from the eighth and ninth centuries. Rather than states building temples, some state formation probably arose on the back of the temple-building effort, by controlling access to the popular pilgrimage sites or manipulating their charisma. The Shaivite temple complex at Prambanan was built on the very edge of Candi Sewu mandala a few decades later, and, unlike the Buddhist complexes, it contains inscriptions lauding the king who presided over the building.
Religion appears the strongest unifying factor in civilizations such as that of Java, but irrigation cooperatives and market cycles also played a role. Except for the Chinese model, extended to the Red River delta (Tongking) during the first millennium, Southeast Asia appears the antithesis of a “hydraulic state” where the control of large river systems required a kind of despotism. Local cooperative arrangements channeled the watercourses of upland valleys rather than the mighty rivers themselves, as is still a marked pattern in Bali. Market cycles also operated in ancient Java and other areas according to a five-day cycle, so that the women who largely operated them could travel to five neighboring villages on successive days, at least one of which was connected to a broader cycle of large villages. Within this world the kings who associated themselves with the most sacred centers of pilgrimage and devotion were sustained much as the temples themselves were maintained, by particular villages accepting a divinely sanctioned obligation to provide rice, thatch, ceramics, or labor whenever needed.
The Buddhist ecumene had particularly flourished when China’s Tang Empire had stimulated both trade and pilgrimage through the passages of Southeast Asia. Sriwijaya around the Straits of Malacca, south-central Java, and Champa on the main route to China had been particular beneficiaries of this Buddhist era, though without a strongly established pattern of hereditary kingship. Recent discoveries show the Buddhist trade diaspora and Indic scripts as more widespread than once thought. The Laguna copperplate inscription in Luzon of 900 CE, in an Austronesian vernacular with old-Javanese script, released a lady and her brother from a debt. Buddhist images and inscriptions have also surfaced in Mindanao and in Sulawesi.
The major trade centers all suffered disruption in the tenth century, probably more because of natural disasters than the disruption of China trade at the fall of the Tang (907). The silencing of the Mataram civilization in south-central Java after 928 is difficult to explain without natural disruptions such as one or more major volcanic eruptions.
The next phase of the expansion of civilizational cores coincided largely with the prosperous Song Dynasty in China (960–1127) but appeared much less dependent upon it. The period from the tenth to thirteenth centuries was dominated in Mainland Southeast Asia by polities I will call nagara, as they called themselves. This Sanskrit term essentially means a city or town and all that pertains to urban sophistication. Angkor and Pagan on the Mainland, and Majapahit in Java, saw themselves as foci of civilization as well as sacred centers. They had no boundaries, and have been labeled mandala or galactic polities by others, because they dominated a penumbra of smaller centers that emulated and deferred to them. Lieberman (2003, 23–5) has suggestively labeled this the “charter era” because each nagara established a vernacular civilization that survived subsequent vicissitudes as a sense of collective identity. Each nagara was made possible by one or more compact centers of wet-rice cultivation that provided a secure surplus to feed it. This was the period of the “rice revolution” discussed in Chapter 1, when rice surpluses began to be produced by more intensive methods in the areas that gave rise to nagara. Coinciding with the medieval anomaly of warm climate in Europe, this period was also favorable for rainfall in Mainland Southeast Asia.
Contributing to the disruption of the older trade centers on the route to China was the only maritime military intervention from India into Southeast Asia known to pre-modern history (balancing the two from China in the 1290s and 1400s). This was from a southern and maritime Tamil polity, not from the empire-builders of the Gangetic plain. The Chola Dynasty since 985 had risen with the increase in trade from the Indian Ocean to Song China, although it was aggressively Shaivite rather than Buddhist. Its partner and potential rival in exploiting this trade was Sriwijaya, which sponsored the building of a Buddhist vihara (monastery) at the principal Chola port of Nagapattinam in 1005, as Tang China had done a couple of centuries earlier, to facilitate pilgrims following the trade route. But the south Indian traders sought the greater profits that would flow from direct access to China. The Chinese court recorded the first “tribute” from the Cholas in 1015, among a flurry of Sriwijayan ones. Ten years later the Chola king Rajendra made his move, claiming that he sent many ships to capture the treasures of Sriwijaya including its “gate of large jewels” (perhaps a metaphor for the gateway to China), among a dozen ports in the Malacca Straits area. Sriwijaya’s sway over rival ports relied on supernatural charisma (loyalty was sanctioned by Buddhist-tinged oaths), patronage, and privileged access to China rather than armies or laws. What little it had accumulated by way of administrative capital was lost with the Chola raid.
The new era favored centers that could combine a domestic rice surplus with control of trade arteries, without being exposed to direct naval attack. Popular Buddhism remained a key mobilizing factor, but kings were increasingly able to channel it for their own accumulation of sacred power. And for such kings, even while patronizing the Buddhist pilgrimage centers, the hierarchies of Hinduism were indispensable. They required a corps of Brahmins to conduct the rituals of enthronement and legitimation as godlike devaraja or “king of the gods,” elevated far above others by the accumulation of supernatural power.
In linguistic terms, to judge by the inscriptions, this period marked a vernacularization of the earlier “language of the gods,” Sanskrit, into what is recognizably Bama, Khmer, Cham, Javanese, and Malay, each with its own adaptation of an Indic script. This shift occurred throughout the Indic world, slightly earlier in south India than in Southeast Asia. By the end of this period each had congealed as a distinct high-culture tradition with which the literate elite of religious scholars and court officials had to be familiar. Sanskrit and Pali were still known to scholars, but no longer served as the essential mark of a universal civilization. On the ground, of course, numerous spoken dialects differed markedly throughout this sprawling region, from each other and from the written high culture. But the pattern of pilgrimage to sacred sites of piety and learning, rather more than the pull of charismatic kings, laid some basis for the broader cultural integrations that would follow.
Angkor and the Khmer. The most puzzling case for historians, and the first to emerge as a major power center in the new era, was the civilization we know as Angkor (the term itself a localization of nagara). The idea of cosmically powerful kings identified with Shiva and the lingga cult appears to have become established in the Khmer area of the lower Mekong and Tonle Sap Rivers in the ninth century. It may have come from the islands in the south, since the first Khmer to claim himself (according to much later inscriptions) as a devaraja representing the power of Shiva appears to have brought this idea from “Jawa.” In the time of one of his descendants called Yosadharapura (reigned c.889–910) a city was established at Angkor. This was not, or not yet, a “state” that could be inherited or captured by a king.
The astonishing development of Angkorean civilization in two centuries that followed, 950–1150, certainly requires more explanation than the sakti of Shaivite kings and their Brahmin advisors, however. The awe-inspiring temple complexes of Angkor Wat (c.1140) and Angkor Thom (c.1200), the vast reservoirs, and the straight roads fit for armies to march hundreds of kilometers, inspired earlier scholars to presume that here, at least, was a true oriental despotism organizing a complex hydraulic system. Subsequent research has shown, however, that Khmer agriculture rested rather on thousands of autonomous village irrigation systems. The core of the Angkor nagara in the plain north of the Tonle Sap lake was extremely favored by the lake’s prolific fish and its habit of doubling or tripling its extent in the wet season when Mekong flood waters filled it to capacity. The Khmer method of agriculture required dykes to slow the retreat of the water and direct it into bunded rice fields. Traces of more than 3,000 temples have been discovered in this flood plain, each served by a community that cooperated to build its dykes. The system did not require state control, yet at its peak may have contained one or two million people, probably more than 5% of the total Southeast Asian population of the time. The uniquely favorable location attracted people to the Angkor area where they flourished for two centuries.
This concentration gave Angkor the potential manpower to dominate over an unprecedentedly large hinterland. It became a militarized state only after a fleet of the more maritime-oriented Chams sailed up the Mekong and plundered the riches of Angkor in 1177. The leader who drove the Chams out, known as Jayavarman VII (r.1181–1220), was understandably obsessed with military power, and became the conqueror and builder par excellence of Khmer history. In religion he was also exceptional. He adopted Mahayana Buddhism, perhaps as a way to rise above the pluralities of the old Shaivite order. His reign was all conquests and building projects, estimated to have exceeded those of all previous rulers combined. While in many respects he does look the stereotype of the “oriental despot,” his power was not structural but personal, an extreme case of what Wolters (1999) called the “man of prowess” phenomenon in Southeast Asia, where dynastic and bureaucratic continuities were relatively low. His massive building projects were not sustainable, including the rapidly built Bayon (Figure 2.1) and what appear desperate enlargements of the water reservoirs. His Buddhism was rejected by his successors, though it later made a comeback. His excesses probably hastened the demise of the system.
Figure 2.1 Buddha-like faces on the Bayon, perhaps intended also to represent its patron, Jayavarman VII, the most powerful and most Buddhist of Angkor’s kings (1181–1218).
Source: Wikimedia Commons. Photograph by Mackay Savage.
Pagan and the Bama. In the Irrawaddy valley a typically Southeast Asian fragmentation of power continued through the tenth century, during which ethnically Bama migrants from the north (perhaps driven by a rising Nanchao in Yunnan) appear to have joined and eventually dominated the older Pyu settlements. They probably brought with them some of the new techniques of irrigating and ploughing rice fields. By the eleventh century, Pagan on the central Irrawaddy had emerged as the preeminent center of Buddhist pilgrimage and temple-building. It was adjacent to the Irrawaddy valley’s most sacred pre-Buddhist pilgrimage site, the volcanic cone now known as Mt Popa, from which the nats of the Burmese spirit world are believed to have emerged. Pagan’s great founder-king, Anawrahta (r.1044–77), was also by tradition the great synthesizer of Buddhism and nat-worship. He canonized 36 ancestral and nature spirits (almost equally male and female) from Mt Popa, and an additional one to represent the Hindu god Indra, by including their images at the base of the great Buddhist stupa, Shwezigon, he built near Pagan to contain a tooth relic from Sri Lanka. Pagan also had some practical advantages through its ability to attract rice supplies from upriver settlements such as Kyaukse and Minbu, as well as the Irrawaddy trade artery between the Indian Ocean and footpaths into Yunnan. Anawrahta for the first time appears to have established a kind of Buddhist peace over the whole of the Irrawaddy valley.
During the following two centuries, Pagan grew ever more important as a center of Buddhist piety. As many as 10,000 brick structures may have arisen in the plain of Pagan, and 3,000 remain as marvels of this charter period. The inscriptions show that most were built not by kings but by a landed elite, typically donating rice land for the building and upkeep of the temples. Not without reason did a Burmese envoy tell the Chinese Emperor in 1286 that Pagan’s glory derived more from religion than from military power. This temple building peaked around 1200, when tens of thousands of acres were donated each year, and collapsed quickly toward 1300. The cause of the rise and fall of this pious building is controversial, as in Angkor, but it must have accompanied an expansion of population and profitable rice-growing. A favorable climate and rainfall in the period had much to do with both phenomena, as did improved technologies of rice-growing and metal-working and the growth of trade during the prosperous Song Dynasty in China.
Java-Bali. Although the Buddha-Shiva civilization of the Mataram area south of Mount Merapi in Java was abandoned in the tenth century, it had sown the seeds of a distinctive Java-Bali culture. The two great Indian epics were both “Javanized” into a poetic form of old Javanese, the Ramayana as early as the ninth century and the Mahabharata in the tenth. The Ramayana was also popular enough to be displayed graphically, at the Prambanan temple and on one of the gold items discovered in a buried hoard from this period. Elements of the same cultural complex began to appear in the intra-volcano valleys of east Java (around modern Malang and Kediri, divided by the active Kelut volcano), and in southeast Bali, of which the earliest dated inscriptions are tenth century. As the inscriptions cease in central Java they become more numerous in east Java and Bali. The period of this flowering of a Hindu-Buddhist culture in their upland valleys in the eleventh to fourteenth centuries can be seen as a charter era of similar type to that in the Mainland. It also profited from similar techniques of wet-rice agriculture, including the plough, and thereby built another pocket of concentrated population with a comfortable rice surplus. Some of the legendary hero-kings of this era continued to be celebrated in subsequent literature and performance. The fourteenth-century Pararaton rendered immortal the story of Ken Dedes, “the princess with the flaming womb” of an earlier era, and her low-born but supernaturally powerful lover Ken Anggrok. The first two state universities of the east were named in the 1950s after two eleventh-century “men of prowess,” Airlangga in Java and Udayana in Bali. Even the first of independent Indonesia’s universities, Gajah Mada in Yogyakarta, was named not for a personality of that region but for the most powerful chancellor of east Java’s fourteenth-century Majapahit.
A Javanese culture was being formed in these valleys through such stories, and the cycles of markets and pilgrimage to the sacred sites of Buddha and Shiva. The economy was commercialized and monetized in a trade boom of the tenth to thirteenth centuries, but never unified politically. The Brantas River that curls its way from the valleys between volcanoes to the coast near Surabaya was not able to unify as the Irrawaddy and the Tonle Sap began to do. The inscriptions of kings such as Airlangga and Kertanagara (1254–92) suggest that they promoted irrigation, conquered rivals, and attracted holy men and artists to their courts, but could build only a temporary primacy.
The unprecedented maritime expedition of “a thousand ships,” 20,000 men and horse-mounted cavalry that Kublai Khan sent from China to Java in 1292–3, evidently failed in its objective to establish Mongol authority. The Chinese force was exploited by a man claiming to be Kertanegara’s son-in-law, who used it to establish a new capital only 50 km from the mouth of the Brantas River, and thereby to control the foreign trade of all the fertile uplands. This became Majapahit, which in the fourteenth century established a true charter nagara by combining Java’s overseas connections with the politically autonomous upland valleys. The most remarkable document to emanate from Hindu-Buddhist Java was a text written in 1365 by Majapahit’s “superintendent of Buddhist affairs,” Mpu Prapanca, rediscovered in a single old Javanese text at the Dutch conquest of Lombok in 1894, and known in its two English translations as Desawarnana or Nagarakertagama. This lengthy poem is dedicated to honoring the king known to history as Hayam Wuruk (b. 1334; r.1350–89), as the divine figure who supernaturally unites all of Java and the known world. “He is Shiva and Buddha, embodied in both the material and the immaterial; as King of the Mountain, Protector of the Protectorless, he is the lord of the lords of the world” (Robson 1995, 26). Prapanca tells in vivid poetry the earliest history of Java we have, explaining that it is through women that the dynasty was established and legitimized. The founder of Majapahit who allied with the invading Mongols married all the daughters of the former king Kertanegara to establish his legitimacy. One of these wives, Rajapatni, is described as so adept in yoga and Buddhist meditation that she could arrange the succession of her daughter on the throne, and later of that daughter’s son, Hayam Wuruk himself.
This Mongol/Chinese intervention in 1292 helped move the balance of power in east Java toward the coast and commerce. The emergence of east Java and its north coast ports as a hub of Southeast Asian trade in the following century was made possible in part by the technology brought by defectors and captives from the Mongol fleet. A new type of large Sino-Southeast Asian trading vessel, known to sixteenth-century European observers as a junk (Javanese jong), came into use at this time. The literary evidence of Hayam Wuruk’s reign shows that Majapahit knew the trade centers of the Nusantara world (Peninsula and Archipelago), and regarded them as part of the universe protected by Majapahit’s divine king. These extended from the spice-producing islands of Maluku in the east to the newly Muslim port of Pasai (Sumatra) in the west. Siam, Cambodia, Champa, Dai Viet, and some Indian ports were listed as “always friends.” This brief interval of peace and prosperity consolidated a formidable literary, musical, and theatrical culture that would survive much subsequent disunity.
Prapanca and the other great writer/monk of the period, Mpu Tantular, celebrated not only their own Buddhist faith but the other two crucial elements of Java’s inherent balance – the Shaivite priests essential to Indic ideas of kingship, and the indigenous tradition which honored the sacred power of the volcanoes, the sea, and the ascetics (resi) who lived hermit lives in the forest. In southern Java and Bali royal rituals appeased both the male gods of the mountains and the female deities of the sea. The Shaivite kings of Singhasari positioned themselves between volcanoes, naming the most active one Smeru for the Meru of Indic myth, and the others Bromo (the Hindu Brahma) and Arjuna. A mark of Hayam Wuruk’s divine status was that the active Kelut volcano erupted at his birth in 1334, while the earth shook with earthquakes.
The Desawarnana tells us that the royal citadel of Majapahit was surrounded by the priests of Shiva in the east, under their Brahmaraja, the Buddhists in the south, under their Abbot, and the royal relatives and officials in the west. It became a theme of Javanese mysticism that different ritual paths to enlightenment were only external expressions of an inner oneness. The state motto of contemporary Indonesia (Bhinneka Tunggal Ika) derives from a fourteenth-century mystical poem, the Sutasoma, in which Mpu Tantular pondered the essential oneness beneath surface differences, such as those between Buddhism, Shaivism, mountain cults, and the new, still marginal presence of Islam: “the truth of Jina (Buddha) and the truth of Shiva is one; they are indeed different, but they remain one [Bhinneka Tunggal Ika], as there is no duality in Truth” (Santoso 1975, 578).
The Malay and Cham polities experienced not a “charter era” in this period but rather an intensification of rivalry between strategically placed ports along the trade routes supplying Song and Yuan China. Just as Sriwijaya fell before the Cholas in 1025, the Champa capital of Indrapura (modern Tra Kieu, near Hoi An), probably including its ridge-top temple sites of My Son, fell before a Dai Viet attack around 1000. The subsequent four centuries of increasingly lively trade were a time of opportunity for trade-based societies, but of a different type. These had never been nagara unified by a single river system and the rice surplus of an extensive irrigated core, but rather what we will call by the Malay term negeri – rival ports strategically situated as gateways for upriver communities along the trade route of the Malacca Straits and the mountainous eastern coast of the Mainland. The model of multi-centered loose polities has been described by such metaphors as mandala, archipelago, galaxy, or more prosaically federation, but these may exaggerate the political stability of the model. What held these societies together were essentially commonalities of religion, language, and material culture, and the pragmatic need to trade with China as if they were a single polity presenting “tribute” from a recognized barbarian kingdom. Kings continued to build Shaivite temples (in Champa) and Mahayana Buddhist, often Tantric, stupas in Sumatra, with inscriptions suggesting their aspirations to be protectors of the various regions within their realm. A 1070 Cham inscription notes that the king “has protected the ten regions from fear.” In times of outside threat, which Champa suffered repeatedly from Dai Viet and Angkor, kings did indeed attempt to act this way. But their sanctions remained supernatural for the most part, threatening cosmic retribution in the afterlife for those who disturbed the royal order. The ports that succeeded in this period were highly cosmopolitan, embracing Chinese, Indian, Arab, and Malay traders, Hindu, Buddhist, and Muslim teachers.
In its own inscriptions Champa appeared to recognize five Sanskrit-named regions or centers spread along the coast. Having lost the northernmost, Indrapura, to Dai Viet, a new successful center emerged in the twelfth century named Vijaya (near modern Quy Nhon), until this was also lost to Dai Viet in 1471. Chinese and Sino-Viet sources recognized at least ten centers, however, while one Champa mission to China claimed that Champa had “105 places” protected by the king. The polity continued to be among the most frequent “tribute”-bearers to China, but also traded increasingly eastward, to Java and Butuan (Mindanao). To a greater extent than the Malay ports of the Malacca Straits, each region had its own small rice-growing valley, profiting from the annual flooding of the riverine plain. Indeed, “Cham rice” was introduced to southern China during the Song period as a superior strain.
During the first millennium of the Common Era, the Cham ports had profited by being “not China,” free ports close to the great empire and its flourishing southern maritime province of Jiaozhi, but not subject to its restrictive regulation of maritime trade. Once a Viet polity emerged from the collapse of the Tang in the tenth century, it became a rival for Champa. From the eleventh through the thirteenth centuries there was much conflict between ambitious rulers of Angkor, Champa, and Dai Viet, each anxious for dominance in the flourishing trade with Song China. In the fourteenth century, however, when the nagara of Pagan and Angkor were in crisis, Champa, like Majapahit in Java, reached an unprecedented peak of external power. The key to this was evidently its large fleet and crucial location, not only for the trade bound for China, but in acting as the export center for Vietnamese ceramics, then at their peak of distribution virtually throughout Southeast Asia. When the ambitious Che Bong Nga came to the Champa throne in 1361 relations with Dai Viet broke down, and the great Cham fleet turned to raiding the Red River delta. In a sequence of ten separate raids before 1383, the Vietnamese capital was sacked three times. This aggression ended when the Vietnamese successfully deployed newly acquired firearms to kill Che Bong Nga as he was on the point of his greatest success in 1390. This loss proved crucial, since the Viet were better able ultimately to profit from the Chinese invasion and occupation of their homeland in 1406 to create a military/bureaucratic state of Chinese type, and thereby definitively to shift the balance against Champa in the fifteenth century (Chapter 9).
Malay ports around the Straits of Malacca also flourished during the trade boom of Song and Yuan, but were even less united than the Champa ones. After the Chola raid on the Sriwijaya capital at Palembang, the China trade shifted to the Batang Hari River further north under the pretext of continuing the “tribute” of Sriwijaya. One Buddhist temple complex was developed at Muara Jambi, near the mouth of that river, whereas others were built hundreds of kilometers up Sumatran rivers, at sites such as Pagarruyung in Minangkabau, Muara Takus, and Padang Lawas. By contrast, a major port site of the eleventh to fourteenth centuries has been excavated at Kota China, near modern Medan, without much evidence of sacred buildings. Marco Polo visited the north Sumatran coast in 1292, as Islam was beginning to make its mark in the Sumatran ports, and declared that there were eight “kingdoms” in Sumatra, each with its own language. In the Peninsula, meanwhile, archaeological sites at the portages reveal increased activity from the eleventh century, with seemingly autonomous Buddhist city-states around modern Chaiya, Nakhon Sithammarat, and Trang.
Sumatra had extensive contact and conflict with the Hindu-Buddhist centers of east Java in this period, giving rise to a cluster of inscriptions and Buddhist constructions found almost exclusively in the highlands. The highland areas today speaking dialects very similar to modern Malay, from Minangkabau in the north to Rejang in the south, had already established a written culture based on a south Indian type of script. About 30 inscriptions from this period relate to the figure of Adityavarman, a Buddhist prince evidently sent from Majapahit in the fourteenth century to exploit the gold-producing area Majapahit knew as Melayu. The majority of these inscriptions are in the higher reaches of the Batang Hari river system and at Pagarruyung, the later supernatural center of essentially stateless Minangkabau. They show Adityavarman attempting to accumulate cosmic power in this area, but do not mention his undoubted interest in the region’s gold and rice. The fourteenth century also marks the first Malay manuscript (on paper rather than as a bronze inscription), preserved on bark-paper in the highlands at Kerinci. This contained a compilation of laws, in Old Malay but with a Sanskrit introduction.
The fourteenth-century manuscript in question also contains a more everyday document of similar age in both the ancient Pallava script and the ka-ga-nga alphabet which became widespread in Sumatra. In short, Buddhist civilization, and writing for both sacred and prosaic purposes, had already extended to the highlands of Sumatra by the fourteenth century. So had notions of sacred kingship expressed in tantric Buddhist form, though this feature appears to have survived Adityavarman’s death in the 1370s only as the lofty charisma that continued to surround the “kings” of Minangkabau at Pagarruyung for several centuries. Once deprived of the opportunity to control trade through a single river-mouth or privileged access to the China market, however, the highlands returned to an essentially stateless mode of civilization. The many ports of eastern Sumatra and the Peninsula, in turn, appear to have developed Malay in the direction of a lingua franca between them, even before Islam and its Arabic loan-words transformed the language in the fifteenth century.
Dai Viet in the tenth to thirteenth centuries was also a charter polity in the sense of establishing the first distinct Vietnamese written identity, a permanent capital (Southeast Asia’s most enduring), and a concentration of population around irrigated rice cultivation. But far from vernacularizing an Indic language and script, its perilous achievement was to build in a Chinese written idiom a polity that was not China. Its predecessor, Jiaozhi, had been the “jewel in the crown” of the Tang Dynasty, occupying all three sides of the Tongking Gulf and making of it China’s vital southern trade centre. Indeed, throughout the first millennium CE the Red River delta was more clearly enmeshed in the Sinic world than most parts of what we now call China. As the Tang collapsed in the tenth century, however, Jiaozhi broke into competing fragments, among which only Dai Viet succeeded in having its separate legitimacy recognized by Chinese dynasties. It thereby drew a boundary against the further southward expansion of Chinese imperial order that proved decisive for the definition of Southeast Asia in the long term.
Dai Viet’s state aspirations were always Chinese/Confucian rather than Indic/Shaivite, so it cannot be called a nagara. Popular Mahayana Buddhism was, however, crucial to its earliest stage of state formation, as it was in the charter nagara. This was the achievement of some warrior chiefs from a redoubt at Hoa-lu, on the southern edge of the Red River delta, who harassed the remaining elements of collapsing Tang imperial control, gained recognition from the interim Southern Han Chinese regime, and established a degree of local order. But only their alliance with Buddhist monks literate in Chinese, and aware of the basic principles of Chinese bureaucratic order, made possible the re-establishment of a court at the old Tang-era capital of Thang-long (modern Hanoi) in 1010. The monks managed the essential tributary relationship with the Song court, supervised the building of Buddhist monuments, and enabled the Ly Dynasty of kings (1009–1225) to produce a body of laws and court rituals. Buddhism (rather than Confucianism) was essentially the mark of civilization, but like their equivalents at Pagan, the Ly systematized the powerful local spirits into a kind of royal cult. The Tongking Gulf maritime area profited from Song China’s prosperity on the one hand and relative military weakness on the other, to become a much more fluid zone of competing ports. The Ly also launched several raids on Champa to the south and on competing “bandit” groups to its north. The long-term consequence of China’s diminished control of the Tongking Gulf, however, was to make Guangzhou (Canton), and especially Quanzhou in Fujian, the favored access ports to China, which would lead by the fifteenth century to the isolation of Dai Viet and the whole Tongking Gulf from long-distance Asian trade.
Maritime trade was still vigorous enough in the thirteenth century, however, to enable the Tran clan to use its fleet to seize power and establish a new dynasty in 1225. Its great innovation was to inaugurate a Chinese-style examination system for officials. While Buddhism remained dominant, these examinations on Chinese literary classics provided a growing cadre of literati for whom the norms of Tang and Song Chinese statecraft were a model. These literati were instrumental in enabling Dai Viet to survive the threat from an aggressive Mongol China, which occupied Thang-long in 1257 before being driven back. The task of the literati was then to try to convince successive rulers of China in the language of the classics that Dai Viet was a separate place, equal in civilization and dignity to China itself. One of them, Le Van Huu, was commissioned to write the first narrative of Dai Viet history. His Dai Viet Su Ky of 1272 rejected as unsatisfactory past rulers who called themselves only by the Chinese term for king, and reached back to a legendary declaration of a southern ruler opposing the Han that he was equal to the great emperor himself. The status of an emperor, Le declared, lay in his virtue (duc), not his territory or armies.
The Tran Dynasty did not survive the deadly struggle with Che Bong Nga’s Champa in the 1370s and 1380s, and the takeover of “usurper” Ho Quy Ly in the 1390s provided the pretext for Ming China to invade in favor of the “legitimate” Tran in 1406, inaugurating a complete transformation of the country. Already for some decades beforehand, however, the order provided by the Buddhist monkhood and an alliance of warrior-chiefs was dissolving in Dai Viet’s own version of a fourteenth-century crisis. It had done enough, however, to mark a clear national identity for subsequent generations of Vietnamese.
The charter civilizations that succeeded in combining a rice surplus from a core area with control of a vital artery of trade occupied only small enclaves of a vast, heavily forested region (Map 2.1). The cosmopolitan port-cities along the trade routes made even less impact on the overall environment. The overwhelming majority of Southeast Asian space, sheltering a probable majority of its people, comprised primary forest, grasslands in some drier areas, and the secondary forest resulting from the shifting agriculture described in Chapter 1. Population throughout this area was lightly dispersed around the coastal strand and upland waterways where food was relatively easily harvested from sea, river, and forest fringe, as well as the palm and fruit trees, vegetables, tubers, and hill rice planted by scattered village communities. As James Scott (2009) has demonstrated, irrigated rice baskets such as Kyaukse in upper Burma, the Red River delta, or the stream-fed upland pockets of east Java both required and facilitated a high degree of organization and peace, to ensure that the high yields of the harvest season were safeguarded for the remainder of the year. Coordination was required, if not through a state then through a kind of irrigation association like the Balinese subak. Peaceful conditions and high rice yields per hectare boosted the population of these rice bowls during the charter era, and permitted leisure time for a rich ritual and cultural life. The ease with which discontented or dispossessed farmers could open an alternative agricultural frontier outside state control created definite limits to oppression.
Map 2.1 Irrigated rice and the rise of polities, 700–1300.
Outside these denser enclaves, scattered village communities engaged with a network of other communities both through reciprocal trade, pilgrimage, feasting and the exchange of tribute for protection, and in sporadic raiding. Their freedom was real, but came at the cost of insecurity and constant readiness to fight or run, which kept populations low. The contrast between such mobile, stateless societies and the cities and irrigated-rice cores became progressively sharper after 1400, as Confucian, Islamic, and Christian notions of civilization created boundaries between those within and without – as detailed in Chapter 5. In the charter era, however, the civilization of written culture and shared ritual was carried in the first place by a Buddhist monkhood that knew few boundaries, and spread its habits of scholarship also to forest retreats and hilltop sacred places. As we have seen in the Sumatran case, the highlands were by no means isolated from the spread of these elements. Even those peoples regarded as most clearly outside the civilized world in the nineteenth century – highland Bataks (Sumatra), Torajans (Sulawesi), Ifugaos (Luzon), and Karens (Mainland) – absorbed key elements of Indic and Buddhist civilization, including Sanskrit terms for cosmic forces. Bataks kept alive through centuries of subsequent isolation their enthusiasm for chess, introduced before Islam from India and Persia.
Writing was an essential tool of states whether for law or for narratives of legitimacy. The many peoples of Southeast Asia who did not have a system of writing until the advent of the roman script at western hands were seen as stateless and backward. Yet many such peoples, including the Hmong, Karen, Chin, and Yao in the upland Mainland, and Kadazan/Dusun, Kayan, and Kenyah in Borneo, retain vivid stories of how their sacred book and the keys to reading it were lost, stolen, or destroyed in a mythical past. In earlier stages of the spread of writing at Buddhist hands, the barriers were low to sharing or accepting that writing across the still-embryonic boundaries between civilized and barbarian, believers and pagans. Once writing became associated with both alien religion and alien state power, on the other hand, those barriers became substantial. Nevertheless, the spread of a particular type of ka-ga-nga alphabet around stateless societies of Island Southeast Asia is a remarkable exception to the widespread association of writing with states and scriptural religions. That this script of straight lines to cut across the fibre of bamboo or palm-leaf with a stylus was first adapted from Indic models in Sumatra during the “charter era” of vernacularization is suggested by the fourteenth-century Sumatran text described above. A similar alphabet had spread throughout highland Sumatra, to south Sulawesi and much of the Philippines by the sixteenth century. It was used, Spanish missionaries in Luzon noted, not for religion or government, but by women even more than men for exchanging notes and commercial and ritual information. In parts of Sumatra the exchange of love notes between young men and women was an essential part of the mating game, and the probable reason for exceptionally high levels of literacy.
This pervasiveness of Buddhist civilization was even more evident in the Mainland during this charter era. Its first carriers appear to have been monastic networks expressing themselves through inscriptions in Mon (the Austroasiatic language of the Irrawaddy delta), and through a style of religious art known as Dvaravati, from the sixth to ninth centuries. This Buddhist culture influenced not only the nagara of Angkor and Pagan, but also the diversity of Tai-speaking peoples who gradually moved into Mainland Southeast Asia from the north during the eighth to twelfth centuries. They developed a predominately Theravada style of Buddhism, but suffused also with Mahayanist elements and the lingga cult of Shiva, present in most of the autonomous rice-growing Tai muang (a group of interrelated villages) of the Mainland. These muang were clearly very successful in organizing the irrigation of upland valleys for rice, and their population grew accordingly. They appear to have recognized each other as part of a coherent world probably sharing agricultural technologies, trade networks, and a similar pattern of dominating the irrigable valley floors while operating in symbiosis (mutual raiding as well as exchange of essentials) with longer-established swidden farmers of the hills.
As the nagara crumbled in the thirteenth and fourteenth centuries, Tai chao (lords), war leaders or men of prowess, emerged to challenge Angkor or Pagan through strategic raiding expeditions, while using some of the Shaivite rhetoric and imagery of the old charter order. Hso Hkan Hpa features as such a figure among the Shans of the upper and eastern Irrawaddy basin. In the period 1152–1205, the chronicles claim, he led Shan forces to attack “China” (probably meaning the Nanchao nagara of the Yunnan area), Assam, and eventually Pagan itself. Despite these legends, the essential Shan level of polity remained the autonomous muang, sharing a written culture and extensive market networks, but never developing into a bureaucratic state. Since the unifying factor of the Irrawaddy trade artery was in Bama hands, the more successful Shan muang, like Mogaung and Hsipaw, tended to acknowledge the Bama nagara when it was strong, and raid it when it was weak, rather than developing a unifying state of their own.
Mangrai, who created the Tai kingdom of Lan Na and founded its capital at Chiang Mai during an energetic reign (1259–1317), was clearly skilled both as war leader and diplomat. His chroniclers claimed that alone among the lords of neighboring muang, he had the legitimacy that came from Indic rituals of enthronement and supernatural regalia, while in the course of his reign he acquired famously potent images and relics of the Buddha from older-established Mon and Bama cities. In 1287, Mangrai mediated a dispute between two other ambitious chao, of Phayao and Sukhothai respectively, which ended in an agreed alliance to fight not each other but the threatening Mongol/Chinese armies from the north.
King Ram Khamhaeng, who ruled Sukhothai from 1279 to 1298, appeared from both chronicles and inscriptions to have created a still more remarkable sphere of influence within his short reign, extending to Tai muang throughout the Chao Phraya valley and even into the Peninsula as far as Nakhon Sithammarat (whose chronicle improbably claims that Ram Khamhaeng came to rule the city for two years). His secret appears to have been to offer to an array of diverse peoples a distinctly Theravada discipline of the Buddhist monkhood, and a model of consensual kingship more compatible with muang autonomy than that of overstretched and declining Angkor. His famous multilingual inscription of 1283, the authenticity of which has been doubted because it was almost too good to be true for its nineteenth-century modernizing discoverer, explained Ram Khamhaeng’s kingship by the Angkor-like things it did not do.
When any commoner or man of rank dies, his estate – his elephants, wives, children, granaries, rice, retainers and groves of areca and betel – is left entirely to his son … When he [the king] sees someone’s rice he does not covet it, when he sees someone’s wealth he does not get angry … If anyone riding an elephant comes to see him to put his own country under his protection, he helps him, treats him generously, and takes care of him (cited Wyatt 1982, 54–5).
Sukhothai has elements of a charter state in the model of Thai-ness (now more specifically referring to Siam rather than Tai in general) it left, notably including the vernacular Thai script and the close identification with Sri Lanka’s Theravada style of Buddhism. But it also bequeathed a loose personal pattern in which heads of small muang acknowledged one other’s legitimacy in an ever-shifting hierarchy of rivalry or temporary deference.
Strategically placed on the upper Chao Phraya, Sukhothai provided the Tai within its orbit with some of the distinctiveness of the people of a state, taken further in subsequent centuries by Ayutthaya near the mouth of the river. We may thus begin to call them Thai or Siamese – a distinct variant of Tai. The tributaries of the Irrawaddy were not similarly unifying for the Shan, and the middle Mekong also less so than the Chao Phraya. The Lao counterpart of Ram Kamhaeng, Fa Ngum, therefore remained in the category of a man of prowess, unifying ephemerally for war purposes. With his more successful but less remembered son Un Huan he built a major capital at Luang Prabang between 1353 and 1416. Fa Ngum invites comparison with his Sumatran contemporary Adityavarman in bringing some of the ritual and imagery of a Shaivite charter state into the stateless territory of Tai muang and upland shifting cultivators. Fa Ngum’s state, like Adityavarman’s, was too Angkor-like to last, and he was overthrown in favor of his son in 1373. Un Huan was more diplomatic, and in his cordial relations with more powerful neighbors he claimed to be king of 300,000 Tai (organized in irrigated rice-growing muang) and 400,000 other peoples presumably distinguished by their swidden cultivation. The prominence of these rulers died with them, but subsequent state-builders in the nineteenth and twentieth centuries cultivated their memory.
It is more difficult to describe the probable majority of Southeast Asians who continued to live the life of shifting cultivators, fishermen, and hunter-gatherers outside the fixities that irrigated rice brought. Occasional Chinese references suggest that the nagara did incorporate such people through various types of bondage. Zhou Daguan described how the Angkor elite used “savages” as slaves in the 1290s, and regarded sexual relations with them as disgraceful. In one intriguing passage he distinguishes between two kinds of savage, of which the first “know how to deal with people,” are useful as slaves and were probably in the process of assimilation into the dominant society. The second were “very ferocious” hunter-gatherers, who “have no homes to live in, but move from place to place in the mountains” (Zhou 1297/2007, 61).
The savage/civilized dichotomy came easily to Chinese, Arab, and later European observers, but Southeast Asian sources have no such sense of boundary. They mention different peoples distinguished by dress and language, and various types of subordination to a lord or patriarch. The dichotomies that do appear are those between upland and lowland, mountain and sea, forest and settled wet-rice agriculture. But the former categories of these dyads have their own spiritual potency, exemplified by mountain shrines, forest monks, and local spirits. Vernacular Buddhism drew no boundary to exclude them. Even in the later period when boundaries were drawn (see Chapter 5), the interdependence and synergy between these dyads would remain part of the fabric of Southeast Asia.
There has been much discussion of the reasons for the eclipse of the temple-building nagara in the thirteenth and fourteenth centuries, and the apparent fragmentation of power among trade-based negeri and relatively compact but autonomous Tai muang. Among causes favored are the invasions of Yuan (Mongol) China into Burma and Java in the 1280s and 1290s, the southward migration of Tai-speaking peoples under pressure of agricultural expansion in south China, and the growing importance of new religions. The Theravada Buddhist sangha (monkhood) in the Mainland, and Islamic teachers in the Islands, increasingly dominated the supernatural resources which kings had sought to attract to themselves through Shaivite cults and the Sanskrit cosmopolis.
Much progress has been made recently in understanding global patterns of climate, of disease, and of volcanic and tectonic disturbance. No historian of the long term can now fail to grapple with these expanding fields, even though most of the new data and the debates concentrate on temperate zones in the northern hemisphere. In that context it is now clear that the Eurasian expansion of Mongol power in the thirteenth century was followed by a Eurasian pandemic of bubonic plague in the 1340s, halving populations in some intensive-agriculture areas. It is likely that parts of Southeast Asia joined the common disease pool of Eurasia at about the same time, suffering initial heavy losses in the pockets of dense population, but eventually acquiring in them the status where smallpox and plague were endemic rather than epidemic, and therefore less destructive. The shift of rice-growing population to the swampy flood plains of the Mainland river deltas was long restricted by malaria and other water-borne diseases, but in the fourteenth century delta rice agriculture appears to have reached the point where the anopheles mosquito did not survive the harsh sunlight of continuous padi fields. An early Siamese chronicle records a myth that appears to link the fourteenth-century establishment of the capital on the lower Chao Phraya River with the conquest of a disease that looks like malaria. An exiled Chinese prince was leading his fleet from place to place in search of an appropriate new settlement. When he came to the island of Ayutthaya in the Chao Phraya River he was told by a hermit that no settlement could be built there unless the dragon (naga) in a marsh there was first vanquished, which “on being disturbed blew poisonous saliva from his mouth. This brought about such an epidemic that everybody around there died of the stench.” Once the dragon had been killed and the marsh filled in, “the land has been free from epidemics” (van Vliet 1640/2005, 200–1).
The “Mediaeval warm period” in about 900–1250 facilitated agricultural production in Europe and China in this period, and its reversal toward cooler weather contributed to northern-hemisphere famines in the fourteenth century. Tree-ring analysis, in its infancy in the tropics, now suggests that the warmer weather in northern Europe was connected with more reliable monsoon rainfall patterns in Mainland Southeast Asia. Thereafter there were some substantial periods of drought. One colossal eruption creating the crater of Rinjani (Lombok) in 1257 must have disrupted weather patterns and deposited poisonous ash on a wide area. These critical changes in weather disrupted agriculture in general and the complex systems that sustained Pagan and Angkor in particular. Thereafter markedly dryer conditions are shown by recent tree-ring analysis, notably in the mid-fourteenth century and the extremely dry years 1401–4. What made these worse was the occurrence also of extremely wet years, the wettest of them occurring in 1257–8 and 1453, the years of the two biggest tropical volcanic eruptions. Floods appear to have damaged the hydraulic system of Angkor so much that its agriculture was unable to cope with the dry periods. These conditions also made the dry-zone agriculture of Upper Burma less viable. There was a shift of population to the wetter regions of the Irrawaddy delta, which also profited more directly from the flourishing maritime trade. The Mons of the coastal areas by 1300 had their own polity no longer in awe of Pagan. The Pagan nagara itself had been distracted by Mongol/Chinese expansionism in its northern hinterland from 1277 to 1301, and eventually fell to Shan raids in 1359–68 much as Angkor succumbed to other Tai-speakers in the Southeast.
This climatic data confirms a long-held hypothesis that the population density that had developed in Angkor through reliance on regular rainfall and flooding of the Tonle Sap proved no longer viable in the worsened climatic conditions of the mid-fourteenth century. The overstretched military empire and construction program of Jayavarman VII was impossible to maintain, and royal power appears to have begun to decline even before his probable death in 1218. The Angkor described by Zhou Daguan in the 1290s offered an agreeable urban life with an honored place for learning, but the Shaivite cult of kingship had lost the power it had to inspire devotion. The people were all Buddhist of distinctly Theravada type, with a temple in every village. In the countryside “as a result of repeated wars with the Siamese the land has been completely laid to waste” (Zhou 1297/2007, 79). The Thai conquest and looting of the capital, probably in 1431–2, only confirmed the massive shift away from the pattern of agriculture and sacred statecraft already long untenable.
Did the archipelagos experience a similar period of environmental crisis separating Buddhist from Muslim-Christian eras? Unfortunately the best tree-ring sequence explored so far, in Java teak forests, began only in 1514, and in their absence caution is appropriate on how to relate humid tropical islands to northern-hemisphere climatic patterns. Tectonic mega-events including tsunamis, and volcanic eruptions that darken skies and induce major crop failures, are likely to be more important factors.
Since the devastating tsunami of 2004 along the northwest coast of Sumatra, however, the advances in understanding this “subduction zone” have been spectacular. By examining and carbon-dating layers of sand deposited by previous tsunamis, and dating the death of coral through the uplift that accompanies major tectonic releases such as that of 2004, geologists have given us a degree of certainty about the last major events of equal or bigger scale than 2004. These probably multiple events occurred in the period 1340–1450, affecting both the northern (Aceh and Nias) and central sections of the west Sumatran coast. This provides an additional reason for the absence of Hindu-Buddhist remains on the tectonic west coast of Sumatra, and probably also the south coast of Java, in addition to the strong surf that reduces the number of safe harbors. River-mouth ports probably did facilitate international interactions in the several-century intervals between major tsunamis (as again during the relatively benign nineteenth and twentieth centuries), but a major tsunami would have eliminated most of the evidence.
The newer scientific research makes clear that there was relatively high discontinuity in the long-term pattern of Southeast Asian history, caused by climatic disruption, volcanoes, and tectonic releases. To this we should probably add diseases, in the period from the tenth to the fourteenth century, when the concentrated pockets of population were in the process of incorporation into the great Eurasian disease pool. Unlike the Americas and Australia, even Island Southeast Asia did not suffer horrendous mortality with the coming of Europeans in the sixteenth century, meaning that smallpox and other diseases had become endemic earlier. But there are enough memories preserved in mythology about cataclysmic epidemics associated with the previous five centuries to make it likely that smallpox, plague, and cholera did arrive with the large Chinese fleets of the 1290s and early 1400s, or the smaller-scale Indian contacts, and became endemic after initial traumatic epidemics. This discontinuity is the background to Southeast Asia’s still-low population of the sixteenth century, at only one-eighth the density of India or China.