The Heritage Society Presents... Back to Heritage F.I.E.L.D - First Ismaili Electronic Library and Database

The medieval Tamil-language inscriptions in Southeast Asia and China

Early inscriptions written in Indian languages and scripts abound in Southeast Asia. Literacy in the very early states of Southeast Asia - aside from the portion of north Vietnam annexed by China - began with the importing, by local rulers, of modified cults of Buddhism or Hinduism, and the attendant adoption of Sanskrit or Pali language for the writing of religious texts. Later, in the seventh century, a broader range of texts began to appear on permanent materials, written in indigenous languages. Given the importance of religion in spearheading the development of indigenous literacy in Southeast Asia, it is not surprising that the north Indian languages of Sanskrit and Pali have had considerable long-term impact upon the linguistic and intellectual cultures of Southeast Asia.

The script in which the earliest Southeast Asian inscriptions were written, however, suggests that these religious ideas and texts were transmitted to Southeast Asians via the southeast coast of India, where religious use of Sanskrit and Pali also sat alongside the more mundane use of very different indigenous local languages. The fact that southern Indian languages didn't travel eastwards along with the script further suggests that the main carriers of ideas from the southeast coast of India to the east - and the main users in Southeast Asia of religious texts written in Sanskrit and Pali - were Southeast Asians themselves. The spread of these north Indian sacred languages thus provides no specific evidence for any movements of South Asian individuals or groups to Southeast Asia.

The same is not true, however, of the handful of medieval inscriptions written in Tamil language and script that have been found in Southeast Asia and China, mainly in Sumatra and peninsular Thailand. These texts arose directly from trade links between south India and certain parts of Southeast Asia and China, which involved the residence in those regions of Tamil-speaking Indians. Several of these overseas Tamil inscriptions mention well-known medieval Indian merchant associations. Since they were so intimately linked to sea trade connections between South and Southeast Asia, these texts - their locations, their contents, timing, and the contexts in which they were written - provide an interesting sidelight on an important period in the economic history of the region. It is, therefore, worth examining these texts against the broader historical background of medieval Southeast Asia - the cycles of trade affecting Asian coasts, the role of south Indian merchant associations in this trade, and the power politics within maritime Southeast Asia at different phases in these economic cycles.


Merchant associations or guilds were features common to many medieval economies. The major Asian trade boom of the tenth to thirteenth centuries appears to have stimulated a significant restructuring of the administration of trade in most major economies of Asia, North Africa and Europe. The Middle Eastern commenda-type investment contract (qirad, etc.), which had become such an important tool of commercial activity in the Middle East by the tenth century, was introduced into Italian seaports by the early eleventh century, thus providing a major impetus for the expansion of trade in medieval Europe.(1) In China, the hang developed from a trading quarter into a nearly autonomous trade association, and the organization of collective financing and management of overseas shipping ventures became increasingly complex.(2) And in southern India a series of merchant associations developed powerful networks and vertical monopolies that included not only tied manufacturers but also private armies. These associations received a good deal of local political encouragement.

Southern India, because its tracts of cultivated land were more dispersed and limited than those of north India, was never the locus of large agrarian states. For much of the region's history, power was fairly evenly balanced between three or more medium-sized states with different linguistic characters: Tamil on the southeast (Coromandel or Cholamandala) coast; Malayalam along the southwest (Malabar) coast; Telegu on the eastern (Andhra) coast north of the Tamil heartland; and Kanarese on the western (now Konkan) coast to the north of the Malabar region. Many of these states relied upon overseas trade to enhance their limited agricultural income. The Pallavas were the first Tamil rulers of real consequence. They grew in power from the mid-sixth to the mid-eighth centuries, and their contacts with Southeast Asia are reflected in many of the early Hindu remains of the maritime region. Pallava power began to decline in the later eighth century, and by the later ninth century they had been replaced as the leading power in the southeast by the Cholas. The Cholas built upon the economic networks developed by the Pallavas, and expanded their links with merchant associations.

Much of southern India's sea trade during the late first and early second millennia was controlled by these merchant associations, the most important of which appear to have first formed in southwest India. During the eighth and ninth centuries much of southwest India's trade had been directed towards the west, and a number of Muslim, Jewish, Zoroastrian and Christian merchant groups from western Asia and the Middle East had established trading enclaves on the Malabar coast. It was in this context that at least some of the early Indian merchant associations began to form. The Manigramam, one of the earliest of the south Indian merchant associations to appear in local records, was mentioned in the late ninth century Kottayam (Sthanu Ravi) copper plate inscriptions found near Quilon on the Malabar coast. These record a contract between the local authority and a group of resident Christians from the Persian Gulf allowing them access, free of certain taxes, to the fort which protected the port market, apparently operated by the merchant group.(3) This merchant association appears to have extended its influence to the east coast shortly afterwards, probably with political encouragement, since the Pallava rulers of the southeast had maintained close relations with the Chera state of the Malabar coast.

During the same period another merchant association had begun to form further to the north in the western Deccan, at Aihole (called Ayyavole in Kannada and Aryapura in Sanskrit) in the Chalukya heartland behind the Konkan coast.(4) They were known either as the Ayyavole/Aryapura association (after their eighth or ninth century place of origin), or more generally as the Ainnurruvar ("The Five Hundred"). This group rapidly became the most powerful of the merchant associations. A number of other merchant groups are known to have affiliated themselves with the Ayyavole, those most prominent in sea trade being the Manigramam and the Nanadesi. In addition, at its height, the Ayyavole appears to have occupied a dominant position in relation to at least forty-six other professional bodies, including major associations of artisans and mercenaries.(5)

The trend in southern India towards merchant domination of certain crafts appears to have accelerated after the turn of the millennium, under Chola rule, with the effective creation of vertical artisan-merchant monopolies. Knock-on effects of the Asian trade boom that were experienced in India included such developments in the weaving and dyeing industries as the introduction - possibly by the eleventh century - of the draw loom, and of the spinning wheel by the thirteenth century.(6) Block-printed textiles, resist-dyed with indigo or mordant-dyed with madder or morinda, were exported in large quantities to both the Middle East and Southeast Asia by the thirteenth century.(7) On the back of this expanding manufacturing and trading base, merchant groups gained a large degree of control over the non-agrarian economy in southern India, assuming the role of major patrons of religious establishments, and being drawn into the collection of certain tolls and taxes on behalf of the state.(8)

Under Chola rule merchant associations appear to have dominated the internal and external trade of south India and parts of Sri Lanka from the late ninth century to the late thirteenth century. The general Sanskrit term for such corporate trading communities was vanigrama ("community of merchants"), and more than one association appears to have used this term or a local variant in their titles: Vanigrama was taken as their title by a group of merchants mentioned in a tenth-century Sri Lankan inscription,(9) and the term Manigramam was a local variant.

The slow decline of the Chola state had already begun before the end of the twelfth century. During the thirteenth century, this decline accelerated into a collapse, as the Pandyans in the far south, the Hoysalas in Karnataka, and local chiefs in northern Tamil Nadu carved up the Chola territory.(10) Zhao Rugua (Chau Ju-kua), writing in the port of Quanzhou in 1225 A.D., recorded reports that the Chola state was at war with its neighbours, although it was still - like several states on the western Gujarati, Konkan, and Malabar coasts, and a number of states on the east coast of India - an active exporter of cotton textiles.(11) This impression of changing fortunes is reinforced by the three early thirteenth century hoards of Chinese coins found on the coast of Tamil Nadu,(12) which were probably abandoned by the mid-thirteenth century due to local upheavals. The mid-thirteenth century efforts of the Kakatiyan ruler Ganapati to revive trade in the southeast Indian port of Motupalli by promising greater safety and stability reflect the growing turbulence in the region.(13)

The remnants of the Chola state were eventually incorporated into the fourteenth-century kingdom of Vijayanagar. Although foreign trade continued under Vijayanagar patronage, the broad consumer base and integrated regional economy of the later Chola period were lacking. By the fourteenth century the great merchant associations were issuing fewer inscriptions, suggesting that - although these associations continued to exist into the seventeenth century - either their structures and modes of operation were changing, or their economic and political power were declining. Despite these changes, individual traders on India's coasts continued to export the cotton textiles and dyestuffs that were increasingly in demand both in the Middle East and in maritime Southeast Asia.


South Indian merchant associations appear to have begun to extend their activities abroad during the ninth century. At first, despite their apparent involvement in the portage trade across the Thai peninsula (see below), the main overseas trade focus of the Ayyavole and the Manigramam seems to have been the western sector of the Indian Ocean and the Middle East. By the tenth century, however, the Asian sea trade boom had begun to draw more East and Southeast Asian trade into the Indian Ocean. The domestic and port trade activities of both of these associations spread during the tenth and eleventh centuries across the interior and to the east coast regions ofTamil Nadu, Karnataka, southern Kalinga, the Pandyan south and into northern Sri Lanka,(14) largely under the auspices of the expanding Chola state. The associations remained powerful along the east coast into the thirteenth century, but lost ground as the Chola state was dismembered towards the end of that century.

The direct impact of the south Indian merchant associations in Southeast Asia was limited both temporally and geographically, probably because of the strength of competing local vested interests in trade. It is also important to note that not all South Asians trading in maritime Southeast Asia during the Asian sea trade boom were clearly associated with South Indian merchant groups. Although the surviving records are too fragmentary to provide a clear picture of the structure of trading communities in early Malay and Sumatran states, epigraphic evidence from Java and Bali of this period suggests that South Asians were present in the local ports in large enough numbers to be classified by region. These inscriptions also indicate that while the term banigrama (the Javanized version of vanigrama) was used in Javanese and Balinese port texts for a time, these banigrama appear to have been rather fluid, port-based groupings that incorporated both local and foreign merchants, some of whom were acting as tax farmers under licence from local rulers.

Javanese inscriptions are particularly rich in information about overseas trade connections. Shifts over time in the composition of the lists of foreigners who were resident at ports in Java appear to reflect changing patterns of trade.(15) The earliest dated original inscription to contain such a list is that of Kalirungan(16) from Kedu in central Java, dated 883 A.D. The list of foreign traders in this inscription includes mainland Southeast Asians from Campa (Champa), Remman (Ramanyadesa, in Mon Lower Burma), and Kmira (Cambodia); the listed South Asians came from Kling (Kalinga), Aryya (Aryapura/Ayyavole), Pandikira (in Kamataka), and perhaps Singhala (Sri Lanka) - although this last reading is uncertain. The one surviving original inscription of the early tenth century to contain a list of foreigners - the Palebuhan charter of 927 A.D., which concerns a community on the banks of the Solo river near Madiun(17) - is, unfortunately, only partially legible. The names Singhala, Pandikira, and Ramman have, however, been preserved, suggesting that the composition of the lists had not changed markedly in the forty-four years following the Kalirungan inscription. The Indians in Javanese ports during the later ninth and early tenth centuries appear to have been drawn either from the east coast districts of what are now Tamil Nadu, Andhra Pradesh and southern Orissa, or from the western Konkan coastal region already dominated by the Ayyavole merchant association. These three regions remained a constant in the Javanese lists from the ninth to the fourteenth century. The two east coast regions of India - particularly that of Kalinga - appear to have been important sources of cotton textiles at an early date. It was perhaps because of the region's dominance of the early textile exports to Southeast Asia that the term "Kling" was widely adopted in the Indonesian archipelago as a blanket-term covering all South Asians, and in some cases all foreigners from the Indian Ocean.

The tenth-eleventh century shift in focus of the merchant associations from the west coast of southern India towards the east, stimulated by increasing trade with the Southeast Asia and China, was accompanied by a broadening of the range of commodities traded. South Indian trade with the Middle East had been dominated by such exports as black pepper, areca nuts and iron.(18) Iron remained a major export to the West into the thirteenth century and may have played an important role in trade with such Southeast Asian states as Java and Bali, which had few of their own iron sources. However, with the expansion of merchant association activities to the east coast of India, particularly in the cotton-producing region stretching from northern Tamil Nadu through Karnataka to the southern coast of Kalinga, cotton textiles began to gain more prominence in the list of exports.

Changes in the textile patterns illustrated on Javanese statuary of the later tenth and eleventh centuries reflect the rapid rise in popularity of Indian export cloths at this time.(19) Four original Javanese inscriptions containing lists of foreigners survive from the eleventh century. They range in date from the 1020s to the 1050s. The Cane inscription of 1021 A.D.,(20) from the Brantas delta region of east Java, lists foreigners from Kling, Aryya, Singhala, Pandikira, Drawida, Campa, Remen, and Kmir. The inscription of Patakan,(21) from the same district, lacks a legible date, but was commissioned in the same period and contains an identical list. The new name in these two lists, Drawida, refers to the east coast Dravidian region of Tamil Nadu, where the Cholas had overthrown the Pallavas late in the ninth century. The Cholas - who gained power within southern India towards the end of the tenth century, and then extended their influence into Sri Lanka, the Maldives and Laccadives by the first decade of the eleventh century - had by this time joined the merchants operating in the larger ports in maritime Southeast Asia. The Chola state had clearly become, by the 1020s, a major player in the eastern sector of the Indian Ocean.

Chinese records indicate that textiles gained in importance among Indian exports to the East in the twelfth and thirteenth centuries.(22) Further changes in the textiles illustrated on Javanese statuary reflect the growing influence of certain densely-patterned block-printed Indian export textiles during this period.(23) By the 1040s and 1050s, when the Turun Hyang(24) and Garaman(25) inscriptions were written, two changes had taken place in names already contained in the lists: Drawida had been replaced by Colika (Chola), and Pandikira by Karnataka. The Garaman inscription of 1053 A.D. also adds to the list the name Malyala (the Malayala-speaking region of the Malabar coast). By this time most of the merchants arriving from the named South Asian regions, including those from southern Kalinga, were probably operating from regions controlled by the Chola empire, and may well have been members of the dominant or subsidiary south Indian merchant associations. However, those residing in Javanese ports appear to have been connected to the Javanese court as individuals - rather than as representatives of merchant associations - in their roles as tax farmers in east Java. They appear to have operated on the same footing as local merchants or those from elsewhere in Southeast Asia.(26)

There appears to have been a lengthy hiatus in the writing of inscriptions in coastal regions of east Java between the middle of the eleventh century, when the state of Mataram broke in half, and the end of the thirteenth century. Of the two states formed from the division of Mataram, the coastally-oriented Javanese state of Janggala, though active in trade, left few records on permanent materials - unlike its sister state, Kadiri, in the interior. The state of Singhasari, formed in 1222 by reuniting the two halves of former Mataram, also left few written remains. It was not until the Majapahit court replaced that of Singhasari in about 1292 A.D. that inscriptions were again written in any number on permanent materials - though many of these were merely reissues of earlier charters. The Balawi inscription of 1305 A.D.(27) is the only original Majapahit inscription found to date that contains a list of tax-farming foreigners. The list reads as follows:

"...wargga kilalan (tax-farmer group): Kling, Aryya, [.....], Singhala, Karnataka, [.....], Cina (China), Campa (Champa), Mandisi (?), Caremin (Ramanyadesa), Kmir (Cambodia),..."

The significant addition found in this list is China, whose merchants had, by the twelfth century, been encouraged by the financially-pressed Chinese court to carry trade abroad - rather than sit and wait for it to arrive. By the thirteenth century Chinese traders appear to have begun to outnumber Indian traders in some ports, just as Chinese-style currency had already begun to displace the earlier forms of indigenous currency in Java and Bali.(28) During the Mongol period at least some of these Chinese merchants may have felt more comfortable operating from non-Chinese ports. At the same time, the regions of India represented in the legible portion of the Balawi list were those that had been there from the ninth century onwards - the east-coast regions of Andhra Pradesh and Coromandel, and the west-coast Aryapura/Ayyavole heartland of the dominant (though now fading) merchant association. The decline of the Chola empire and the partial disruption of trade in eastern waters during the late thirteenth and early fourteenth centuries clearly had not led to an abandoning of trade links between Indian ports and those of maritime Southeast Asia.

The Nagarakertagama,(29) a Javanese text completed around 1365 A.D., mentions merchants and others from China, various parts of Southeast Asia, and Jambudwipa (India) - specifically from Goda (Gaur in Bengal) and Kamataka, these two regions having by then pulled away from the control of the sultanate of Delhi. The continued importance of the Tamil-speaking Kamataka coast, in particular, in Javanese lists of Indian traders may have had much to do with the textile trade. Patterned textiles represented on thirteenth- and fourteenth-century statuary from east Java bear a very close resemblance to some of the patterns preserved on fragments of Indian export textiles of those centuries found at Egyptian sites.(30) Although some of the patterns on Javanese statuary may represent local batik copies of Indian export cloths, there is no doubt that Indian printed cottons were hugely popular all around the Indian Ocean - and beyond - by this time. Even during this century of relative trade depression in parts of maritime Southeast Asia, India's cottons found a ready market.


Medieval south Indian interest in trade with Southeast Asia and China is reflected not only in a number of inscriptions in southern India, but also in several clusters of Tamil-language inscriptions and south Indian-style religious remains that have been found on the eastern fringes of the Indian Ocean and on the coast of China. Eight of these inscriptions, dating from the mid-ninth to late thirteenth centuries, written on stone wholly or partly in Tamil language and using Tamil script, have so far come to light: one in Burma, two in peninsular Thailand, four in Sumatra, and one on the central coast of China. The south Indian-style religious remains associated with some of these inscriptions can almost certainly be placed within the same time frame.

This handful of overseas Tamil inscriptions can be divided chronologically into roughly three periods:(31) the Pallava-Chola transition period of the ninth and tenth century, the high Chola period of the eleventh and twelfth centuries, and period of accelerating decline of Chola power in the late thirteenth century. Curiously enough, most of the overseas Tamilinscriptions that have been found appear to belong to the last of these periods.


Historical Context

The period between the late eighth and the early tenth centuries was a turbulent one in much of Asia. The major markets of South and East Asia were affected by political destabilization, and this economic uncertainty in turn produced knock-on effects in the international trade economy. Trade in Southeast Asian waters declined later in the eighth century and remained rather uncertain through the ninth century, as disturbances hit the markets in China and India.

The increasingly weak Tang government of China had gone into retreat in the mid-eighth century, and - following the disastrous battle of Talas against the expanding Muslim empire and a series of uprisings involving foreigners in the major ports of the southeast coast - the administration turned inwards on itself, becoming so xenophobic that foreign traders experienced increasing difficulty in gaining access to Chinese goods. However, China's market was so important to Southeast Asian trading economies that trade did continue, albeit at a lower level, even as Tang China fell apart late in the ninth century. Furthermore, the decline in trade appears to have hit some of China's trading partners harder than others. Chinese court records for period between the late 760s and about 900 list only a handful of trade missions. Of these, eight were from Java, two from the port-state of Jambi in Sumatra, one from Cambodia, and one from Champa. No missions were received from the trading empire of Srivijaya (which had earlier incorporated Jambi) between 742 and 904, and none was officially received from South Asia between 762 and at least 960.(32)

During this period, northern India was embroiled in a three-cornered struggle for power over the vast agrarian heartland of the Doab region and the Ganges-Brahmaputra delta. While much of north India's political and economic energy was directed either inwards, or westwards in response to the expansion of the Muslim empire, the Buddhist religious culture of the Bengal/Bihar region - then under Pala rule - continued to influence states both to the north and the east, and the great Buddhist centre of Nalanda drew pilgrims and investment from a number of states in Southeast Asia. However, north Indian traders are not mentioned in Chinese or Southeast Asian sources during the ninth century. Much of the trade of the Ganges delta ports appears to have been carried by south Indians or Southeast Asians at that time. Southern India during this same period was also in turmoil: Pallava rule in the south disintegrated. However, although Indian trade with the East clearly suffered as a result of the political upheavals of the period, the internal troubles in Indian states do not appear to have led to the deliberate severing of trade links that characterized the Chinese government's reaction to political turmoil. While links between Indian ports and specific sites on the coast of Sumatra during this period are not easy to demonstrate, continuing Javanese links with India are easy to substantiate in the Javanese epigraphic and archaeological records.

The late eighth and ninth century disturbance in Asian sea trade appears to have been sufficient to cause the trading empire of Srivijaya to unravel for some time, as the separate port-states of which it was composed began competing with each other for the declining trade. The few mentions, in Chinese records, of the Malacca Straits during the ninth century indicate that some fundamental changes had occurred in the political structure of the region: although archaeological evidence indicates that Palembang (probably the main port of early Srivijaya) continued to handle some trade during the ninth century,(33) Chia Tan's itinerary of about 800(34) suggests that Srivijaya had already lost control of a number of key ports on the coasts of the peninsula and northeast Sumatra. Then in 852 and 871, the southeast Sumatran port-state of Jambi (formerly at the heart of Srivijaya's empire) sent its own trade missions to China.(35) The impression that Srivijaya had broken down during the later eighth and ninth centuries is reinforced not only by epigraphic data from the peninsula(36) and northeast India,(37) but also by archaeological evidence that a portage route across the Isthmus of Kra was in use for some decades during that century? suggesting that order had so broken down in the Straits that piracy had become a severe problem.

Java, on the other hand, appears to have weathered the decline in trade - its agricultural underpinnings lent it stability, and the spices and sandalwood that its trade network controlled were always in demand, both in China and in the Indian Ocean. In fact, several Javanese missions were welcomed during the ninth century by the otherwise xenophobic late Tang government.(39) Java appears also to have taken advantage of the political vacuum in maritime Southeast Asia to expand its sphere of influence.(40) The first Arab sources to describe maritime Southeast Asia appeared in the 840s and 850s, and with them the first mention of the great island empire of Zabaj - which at that time clearly referred to Java and its sphere of political influence.(41) For the Javanese state of Mataram, this period of relative trade decline in Southeast Asian waters was one of political consolidation and increasing economic sophistication, enhanced by the creation of maritime Southeast Asia's first domestic coinage system.(42)

Tamil Text

a. Takuapa, Peninsular Thailand

The earliest of the eight overseas Tamil inscriptions was found on the hill Khau Pra-Narai (Brah Narayana), about ten miles upstream on the Takuapa river, on the west coast of peninsular Thailand.(43) It was associated with the remains of a small structure and three large stone figures - Siva and two companions - of apparent south Indian inspiration or origin.(44) The Takuapa inscription was written entirely in Tamil language, in south Indian script of about the ninth century.(45) The revised translation published by K.A. Nilakanta Sastri(46) reads:

[The first line is largely illegible, but contains the fragments of what appears to be a royal name: ...varmakku...]

"The tank dug [by] Nangg[u]r-[u]dai[yan] [and] called Sri-A[vani]-Naranam [is placed under] the protection of the members of Manigramam, the residents of the military camp (senamuga[m]), and [...]"

Nilakanta Sastri suggested that the digging of the tank occurred at the behest of the Pallava king Nandivarman III, who ruled at Kancipuram in southeast India during the mid-ninth century. One of this king's epithets was Avani-Naranam ("Visnu on earth"), even though he was better known as a devotee of Siva. If this inscription does, indeed, date to the middle of the ninth century, then it is almost as early as the first known inscription to mention the Manigramam - the Tamil/Malayalam (with Arabic and Persian) language Kottayam inscription of Quilon on the Malabar coast of India, issued in about 849, during the reign of that region's ruler Sthanu Ravi.(47) However, since the Pallavas were, by the mid-ninth century, already ceding their dominant position in southeast India to the Cholas, it may be worth comparing the Takuapa inscription with the slightly later Tamil language inscription from Munasandal, in the early Chola heartland, issued under the early Chola ruler Parakesarivarman. This inscription - issued in about 870, during the early stages of the Chola expansion - records endowments made by the Valanjiyar (merchant) Five Hundred connected with a tank, this one named after the ainnurruvar (the Five Hundred [of Ayyavole]) merchant association.(48) If the Cholas were supporting the Ayyavole merchant association from the Konkan coast at this time, they are also likely to have had connections with the nearer Malabar coast and its Manigramam merchant association.

The remains of a tank of possible south Indian design have been found at the port site of Ko Kho Khao, not far from Takuapa; it is possible that this tank formed part of the infrastructure of a settlement associated with a south Indian merchant group. Ku Kho Khao was apparently the western terminus of an overland portage route, which, for several decades in the ninth century, linked this short-lived west coast port to a port at Laem Pho, near Chaiya on the east coast of the peninsula. Both of these former port sites have yielded large quantities of Middle Eastern glass and Chinese ceramics of the ninth century.(49) It thus seems that this portage route linking the Bay of Bengal to the Gulf of Siam - which is the only one across the peninsula for which clear archaeological evidence exists - was in active use only in the ninth century. Since those using this overland portage route were forced to tolerate a high rate of breakage of the glass and ceramics they carried, they must have had a compelling reason for doing so. The substantial loss in profits through breakage must have been outweighed by difficulties encountered in using the route through the Malacca Straits at this time, possibly because of the dangers presented by piracy during a period of political breakdown, as the trading empire that had been built by Srivijaya in the late seventh century unravelled for a time.

The apparent link between the operation of the ninth century transpeninsular portage route and one of the south Indian merchant associations most active in overseas trade is of considerable interest, but its implications for subsequent activities of Indian merchant associations are limited. The ninth century involvement of a south Indian association in this portage route appears to mark the high point in the direct involvement of south Asian merchants in the transit trade through the region. The route was in use for less than a century, during the period when the breakdown of the Malacca Straits trading empire of Srivijaya appears to have made alternative routes to that through the Straits attractive. When, in the tenth century, Srivijaya reformed itself, trade moved back to the Straits route, and the portage route fell into disuse.


Historical Context

During the tenth century, the tempo of Asian sea trade increased sharply, as the situation in China began to stabilize, and first the Southern Han and then the Song government in China encouraged trade from Chinese ports. Trade in southern and eastern India began to benefit from political consolidation under the Cholas, and in the Middle East under Fatimid rule the economic stability and prosperity of Egypt added to the overall increase in Indian Ocean trade. Levels of sea trade between the Indonesian archipelago, the Indian Ocean and the South China Sea remained relatively high from the early tenth until the middle of the thirteenth century. Southeast Asian traders were major beneficiaries of the sea trade boom. Until the end of the eleventh century - when Chinese merchants were actively encouraged by their government to carry their trade abroad rather than wait for others to carry it to them(50) - trade in the South China Sea appears to have been dominated by Southeast Asian shipping.

Intensifying Chola contacts with Southeast and East Asia soon after the turn of the second millennium are reflected not only in Chinese records, but also in the numbers of eleventh- to thirteenth-century Chinese ceramics found at south Indian port sites,(51) and in the south Indian epigraphic record. Inscriptions in south India indicate that during the first two decades of the eleventh century, relations between the Chola state and Srivijaya had intensified, as the Southeast Asian state made overtures to the Cholas in an effort to consolidate trade relations.(52) In 1015, while these overtures were taking place, the first Chola mission was dispatched, via the Malacca Straits, to the Chinese court, carrying a number of valuable commodities.(53) In the following year, the trading states that were accorded first class status by the Chinese court were the Arabs, the Chola state, Srivijaya, and Java.(54)

Then, in 1023, the Chinese emperor advised Middle Eastern merchants trading with China to avoid the overland silk route, which was growing increasingly dangerous, and to send missions to China by sea instead.(55) This advice appears to have further stimulated the flow of traffic across the Indian Ocean and through the Malacca Straits. This may, in turn, have precipitated the Chola raids on Southeast Asian ports in 1025. The list of thirteen ports that Rajendra I claimed to have raided included Kadaram (Kedah), several other peninsular ports, a number of Sumatran ports, and the main port of Srivijaya itself.(56)

The effects of the Chola raids appear, for the most part, to have been minimal and transitory: Srivijaya sent a well-received mission to China in 1028,(57) and the quantities of Chinese ceramics and other archaeological remains found at the extensive site of Muara Jambi (Malayu) on the lower Batang Hari river - which was apparently the main port of the state by the eleventh century - indicate that Srivijaya grew increasingly wealthy on trade during the later eleventh and twelfth centuries. Since the lull in trade missions sent by the state to the Chinese court after 1028 - which has given rise to speculation that Srivijaya went into a decline following the Chola raids - was part of a stoppage that affected all states, it appears to have reflected a decision on the part of the Chinese court to restrict the burgeoning number of trade missions to port areas, rather than a decline in Srivijaya's trade wealth or power. This diplomatic lull seems not to have affected trade, in which the emperor expressed an interest in 1028,(58) and which was on the rise from maritime Southeast Asia and the Indian Ocean during the period between the 1030s and 1050s.(59)

In the 1060s and 1070s Chinese records report that the imperial court's reception of missions from the Chola state, Srivijaya, and Java was resumed. Following this, competition between the Chola state and Srivijaya again manifested itself. The Chola ruler Virarajendra claimed in 1068 to have conquered Kedah on behalf of a local ruler.(60) Since, at this time, the Chinese court was under the impression that the Chola state was subordinate to Srivijaya,(61) this second claim of conquest must also be treated with caution. The archaeological remains found at the sites of Sungei Mas and Pengkalan Bujang in Kedah indicate that these ports, between the tenth and the thirteenth centuries, were engaged in trade in both directions, but although Indian potting styles may have affected local wares, and Indian merchants may have resided in the ports, the few religious remains from the ports are so similar to those of parts of Sumatra and east Java at the time that there seems little doubt that the port population was largely of local extraction and closely linked to maritime Southeast Asian trade networks. There is no local evidence that Chola power actually extended to this coast.

By the last decades of the eleventh century, the Chinese court had begun to encourage Chinese traders themselves to venture to sea, and particular rewards were offered to those who imported Middle Eastern frankincense. Chinese interest in some products of the Indian Ocean region was intense, and south Indian merchants appear to have responded to these favourable market conditions. But the Chinese court import lists do not record any sudden influx of commodities of specifically Indian origin. Many of the items that were presented to the Chinese court by the Chola missions of the eleventh century were either Middle Eastern or maritime Southeast Asian in origin, and were very similar to the range of goods offered by Srivijaya in the same period.(62) Although Indian printed textiles are not included in the few recorded lists of goods brought by Chola missions to China in the eleventh century, it seems likely that they formed an important part of the Chola trade with the East. Indian printed textiles had already grown so popular in Java that they had begun to provoke a change in Javanese textile production. The absence of black pepper from surviving Chola tribute lists, however, suggests that Java may already have undercut south Indian exports to China.

Much of the Chola trade with China may have been submerged in the tribute missions carried by Srivijaya. The fact that the mission sent by the Cholas to China in 1077 was the last mission from that state recorded by Song documents may indicate that most of the south Indian state's trade with China thereafter was either indirect or carried onwards from the Malacca Straits by Indian merchants using Srivijayan ships. Two years after this last Chola mission, Srivijaya consolidated its position in the port of Guangzhou through the reconstruction, paid for by the ruler of Srivijaya, of a local Taoist temple, to which this ruler made a donation of land, in addition to a donation of land made to a local Buddhist temple.(63) Srivijaya's apparent success in its eleventh century struggle with the Chola state over transit trade through the Malacca Straits may have had important repercussions within the Java Sea network.

During the last two decades of the eleventh century, officially-sanctioned Chinese imports from the southern seas continued to mount, thus stimulating further Srivijayan missions to China(64) and to south India.(65) Early in the twelfth century official ties between Java and China - which appear to have suffered for a time due to instability in Java - intensified once more.(66) Levels of Javanese and Srivijayan trade with China then remained high through the twelfth century, despite fluctuations in Chinese import duties and growing restrictions upon the export of Chinese copper cash.(67) During the mid-twelfth century the Chinese court bestowed further honours upon Java, but noted at the same time the number of wealthy merchants from Srivijaya resident in the port of Quanzhou.(68) A Chinese report of 1178 noted that at that time the most important states bringing precious and varied goods to China were those of the Arabs, the Javanese and Srivijaya - in that order. Java was said to be important because of its control of its own and the eastern Indonesian spice trade, and Srivijaya because of its control of the thoroughfare to the Indian Ocean.(69) There were no recorded missions from the Chola state during that century, although a Chinese itinerary of traders sailing from the Middle East to China late in the twelfth century noted that travellers transferred from the smaller Arab ships to larger vessels at the port of Quilon (Ku-lon) on the Malabar coast.(70)

There were apparently no ports in Java or Bali in which south Indian merchant associations either operated as extraterritorial groups or traded independently of locally-organized trading systems. This continuing absence of foreign groups acting extra-territorially may have been reinforced in the 1070s, when Srivijaya appears to have succeeded in restricting their activities in the ports of the Malacca Straits, but it was also a reflection of the degree of financial interest that the Javanese and Balinese rulers had in port trade, and the strength of local trading groups. After the middle of the eleventh century, as the Javanese tax system began increasingly to focus on individuals rather than groups, the term banigrama disappeared from tax collecting lists. The fact that this restructuring appears not to have affected the role of foreign traders in the tax-farming system reinforces the impression that they had, all along, been approached as individuals by the Javanese court.

Tamil Texts

a. Lobo Tuwa, West Sumatra

There is a temporal gap of almost two centuries between the ninth century peninsular inscription of Takuapa and the earliest Tamil-language inscription found in Sumatra. This latter inscription - dated 1010 Saka, or 1088 A.D.(71) - was found at the early port site of Lobo Tuwa, just to the north of Barus on the west coast of the island. The date of the text falls within the reign of the Chola ruler Kulottunga, which marked the apogee of the Chola state's power. It states that:

"In the Saka year 1010 current, month Masi, we, the Nanadesa-Tisaiyayirattu Ainnurruvar,(72) having met at the velapuram(73) in Varosu (Barus), also called 'the pattinam (commercial town) for the welfare of the merchant body blessed by Siva', decided to grant as follows to 'our sons', the nagara-senapati Nattu-cettiyar,(74) to Patinen-bumi-desi-appar(?), and to the mavettugal (elephant-trainers?): [On each of the] ships' [cargoes?], the ship's captain(75) and crew will pay the fee anju-tundayam(76) in gold, pegged to the price of kasturi (musk),(77) and [then only] may 'step on the cloth spread' (ie. enter the settlement to trade). Thus we, the Five Hundred of the Thousand Directions, known in every direction in all Eighteen Lands, had the stone inscribed and planted. Do not forget charity; charity alone will help you."

The contents of this inscription are similar to those of a number of inscriptions produced in the Chola heartland of south India during the high Chola period. However, it appears that the grant covered only the velapuram settlement of the south Indian merchants, that was under south Indian administration, rather than the entire pattinam (major port) of Barus, which served a more mixed community of merchants. This port was famous from the Middle East (where it was known as Fansur) to China for its camphor, and it appears, during this period, to have drawn traders not only from Sumatra, Java and south India, but also occasionally from the Middle East, via south Indian ports.(78) The site of Lobo Tuwa appears to have been the focus of the eleventh to thirteenth century phase of settlement connected with the port of Barus: it has yielded large quantities of Middle Eastern glass and Chinese ceramics dating from the eleventh to early thirteenth centuries, as well as two fragments of another inscription, which is too worn to be legible, but which was written in the kawi script of Java, and can be roughly dated palaeographically to the same period.(79) Indigenous gold and silver coinage of this period found in the Barus area belongs to Java's sphere of monetary design influence, as do the gold rings with inscriptions in Javanese kawi script.(80)

b. Neusu Aceh, North Sumatra

A slightly later Tamil-language inscription has recently been found at Neusu Aceh,(81) at the eastern end of Kreung Raya Bay, on the east coast of Aceh, north Sumatra. The date of the inscription is illegible, but it has been dated palaeographically to about the twelfth century, The entire front of the stone is illegible, aside from the isolated word mandapam, presumably relating to a temple foundation or endowment. A provisional translation(82) of the legible portion of the back reads:

"[....] that [we] would not hold (kaikkollakadavadalla); for this decision. [If any violation comes or situation arises against it (?)...those] who put the [royal] emblem (ilachchinal, Sanskrit lancana) will come to take and go (seize?) [....] that [we] would not get or collect (kolludal) interest [in kind] (polisai) [...even] afterwards our people (makkal) would do in accordance with this inscription (ikkalvettuku kokka[okka]). Let there be prosperity."

This provisional reading suggests that the inscription dealt with trading regulations covering losses of goods, the waiving of collection of interest, and perhaps of royal fees. The legible portion of the text does not bear the name of any merchant association, but a reference to "our people" (makkal) may indicate that an association was involved: several inscriptions of this period in southern India connect merchant associations - particularly the Ainnurruvar - with the construction of mandapa or other additions to both Hindu and Buddhist foundations.

The nearby sites of Lhok Cut and Ladong(83) have yet to be properly explored, but they appear to represent the remains of a series of ports dating to the eleventh century and later, overlapping temporally with other ports on the same coast that have produced some evidence of south Indian contact. A port at Pulau Kompei, on Aru Bay, was apparently most active during the eleventh and early twelfth centuries, judging from the dates of the Middle Eastern glassware and Chinese ceramics and coins found at the site.(84) The port at Kota Cina, near Medan further south on the same coast, has produced several stone statues of clear south Indian inspiration, if not origin, along with large quantities of Chinese ceramics and coins of the twelfth and thirteenth centuries, a scattering of Middle eastern glass, and a few Sinhalese coins of the early thirteenth century.(85) Some of the twelfth-century hard red carinated pottery from the Kota Cina and Lhok Cut sites - and from the Pengkalan Bujang site in Kedah, on the other side of the Straits - may have been Indian influenced.


Historical Context

The boom in sea trade involving maritime Southeast Asia began to dissipate after the middle of the thirteenth century. Following the Mongol conquest of China and Central Asia, much of the trade between the Middle East and China shifted away from the sea to the overland silk route. Within China, the uneasy relations between Chinese traders in the southern ports and their new Mongol overlords - the Yuan dynasty - created further problems. At much the same time, the south Indian Chola state, which had provided a political and military platform for the activities of south Indian merchant associations, began to disintegrate. These political upheavals and shifts in trade routes combined to create a trade depression in the eastern sector of the Indian Ocean and in the South China Sea.

Although the volume of sea trade in Southeast Asian waters appears to have dropped sharply during the late thirteenth century, it did not cease, and certain export sectors and some states' economies were more affected than others. With the Mongol conquest of the south of China by the 1270s, the political situation stabilized once more, and commerce was encouraged. The Yuan administration made efforts to reopen sea trade connections, by reestablishing a maritime trade bureau in Quanzhou in 1278.(86) The Yuan government at first encouraged trade with India - early in 1281, a Mongol envoy was dispatched to India from the port of Quanzhou(87) - although at the same time it began taxing goods carried in by foreign traders at twice the rate it charged Chinese traders.(88) In response, missions were sent to China from the Malabar coast between 1279 and 1314.(89) The bilingual Tamil and Chinese language inscription of 1281 connected with the Hindu temple in Quanzhou indicates that south Indians still (or once again) maintained a trading enclave in the port.

But Yuan policy on overseas trade was notable both for its singleness of purpose in maximizing government profits, and for the inconsistencies in, and reversals of, the regulations enacted to achieve this end. In 1284, the Yuan government reversed its open-door policy, banning private trade and making the Yuan state the sole agent legally entitled to trade overseas.(90) Foreign traders were still allowed to trade in nominated Chinese ports, but under increasing government restriction, even in relation to their trade outside of China.(91) This must have given a boost to foreign-carried trade for a time, and, according to Marco Polo, Indian traders were active in Quanzhou in the late 1280s and early 1290s despite the increasing tax burden imposed upon them.(92) Between 1284 and 1330, however, Yuan regulations connected with port trade went through a series of unsettling shifts, involving not only fluctuations in tax, but also the abolishment of private trade with foreigners, a series of closures and reopenings of mercantile trade offices, and interdictions upon the export of metals, bullion and coinage. Further complicating the relations between the Yuan government and foreign traders was the fact that Mongol attacks on their southern neighbours - including not only the contiguous portions of mainland Southeast Asia, but also Champa and Java - peaked in the 1280s and early 1290s.

The increasingly prominent role in maritime trade played by Central Asian Muslim merchants and their ortaq partnerships at this time(93) was reflected in the official refurbishing of Quanzhou's mosques in 1310.(94) This official favouring of Muslim merchants may also have reflected the desire on the part of the Yuan administration to encourage trade with the increasingly Muslim-dominated Indian Ocean - particularly after their failure to capture Java's pepper trade by force in 1292-93.(95)

In maritime Southeast Asia, it was once again the Malacca Straits trading empire of Srivijaya that suffered most from the turbulence in sea trade. As the volume of trade passing through the Malacca Straits dropped, Srivijaya's member states appear once again to have begun to compete with each other for the declining trade income, and the multi-port empire disintegrated. Java, with its mixed economy and spice exports, rode out the trade depression. The balance of economic and political power in maritime Southeast Asia shifted in favour of Java, as Srivijaya faced financial difficulties. Marco Polo, passing through maritime Southeast Asian waters in the 1290s on his journey home from China via the Indian Ocean, mentions the great pepper-exporting state of Java, but he appears to have passed through the Malacca Straits without coming into contact with Srivijaya. Instead of stopping at a south Sumatran port, he laid over at the port of Samudra-Pasai in northeast Sumatra,(96) which fell within the circuit in which south Indian merchants operated, and from whence he travelled onwards to India.

During the later thirteenth and early fourteenth centuries - possibly in response to stresses between the Mongol administration and private merchants in southern China - the numbers of Chinese residents in east Java and elsewhere in Southeast Asia increased. Java took advantage of the political vacuum in the Malacca Straits region, expanding its sphere of political influence in southern Sumatra, and it continued to dominate much of the region until the beginning of the fifteenth century, when an upswing in sea trade stimulated the formation of another trading empire in the Malacca Straits, this time focused upon Malacca.

In southern India, turmoil attendant upon the conquests of the sultanate of Delhi during the late thirteenth and early fourteenth centuries seems not to have caused irremediable disturbance to established and lucrative sea trade networks. Several of the Tamil-language inscriptions found in western Southeast Asia, and that in China, were written during these rather unsettled times. And the Old Javanese language Balawi inscription(97) of east Java suggests that the long-established links between east Java and the Kalinga, Karnataka and Konkan (Aryya) coasts survived the political upheavals.

Tamil Texts

The last five of the known Tamil-language inscriptions of Southeast and East Asia appear to date to the second half of the thirteenth century. Two have been found in north and west Sumatra, one in peninsular Thailand, one in Upper Burma, and one in Quanzhou in eastern China. The timing and distribution of this group of overseas Tamil inscriptions is interesting, since the late thirteenth century was the period of the decline not only of the Malacca Straits trading empire of Srivijaya, but also of the Burman state focused upon Pagan. This was also a period of expansion of Javanese influence in Sumatra, of Thai influence in peninsular Thailand, and Mongol power in China.

a. Batu Bapahat, West Sumatra

A late thirteenth-century bilingual text has been found inscribed on the rock wall of the waterworks at Batu (or Bandar) Bapahat,(98) near Suruaso in the Padang Highlands of west Sumatra, about 250 km to the south of Barns. The left-hand text comprises ten lines written in the Javanese-influenced central Sumatran script, similar to that used for the inscriptions of king Adityawarman, who ruled in the Padang highlands in the mid fourteenth century. This text is written in a local version of Sanskrit. The right-hand text comprises thirteen lines written in Tamil language using south Indian script. No transcription or full translation of this text has as yet been published, but the Sanskrit portion apparently mentions the grama (settlement) Sri Suravasa (modem Suruaso). The inscription lacks archaeological context, but may have been connected with Minangkabau trade in camphor and gold through a west coast port in the Padang area. No merchant association appears to have been mentioned in the portions of the inscription that have been read.

b. Porlak Dolok, North Sumatra

The other late thirteenth-century Tamil inscription found in Sumatra is also bilingual. It is carved on a stone pillar bearing the head of Ganesa, found at Porlak Dolok near Paringginan in the Padang Lawas area of northeast Sumatra.(99) The two languages of this inscription are Tamil (written in south Indian script) and a variety of Old Malay heavily influenced by Old Javanese language (and written in Javanese kawi script). The inscription is extensively damaged, and only a portion of the text in each language is still legible, making the reading of the contents somewhat uncertain. The first part of the inscription - that in Old Malay/Javanese-language - has been transcribed by Edi Sedyawati.(100) Based upon this transcription and the notes made by Karashima,(101) a rough, provisional translation seems to read thus:

[Old Malay/Javanese text]: "Om! The auspicious Saka year evolved is [ ............ sa], [of] the bright half of the month the eleventh [day], the day (wara) of Sak[r]a, Rewati (the fifth) Naksatra (lunar mansion). At that time (tatkala) [............pan] senapati rakan Dipangka[..]da[.]a [.....?] Gangga sena[...] being(?) [....... ][may this] offering(?) [of/to the] nagara be considered as a meritorious act (akan punya) [of] paduka sri mahara[ja......] guya(?)."

[The final portion, of roughly equal length, is written in Tamil, using south Indian script. It apparently contains more or less the same information. No merchant association is mentioned.]

The date given in the local-language portion is an abbreviated version of the type of elaborate date used in Java at the time, here listing only the day in the Sanskrit seven-day week, and omitting the specifically Javanese weekday. The partially effaced year probably falls somewhere around 1265 A.D.(102) Both portions of this inscription relate to an official, senapati rakan Dipangka[..]da[.]a, who appears - according to the Old Malay/Javanese version of the text - to have made some sort of offering (?bhali) involving the nagara,(103) which was to be considered as a meritorious act (akan punya) of a ruler. This ruler's name is, unfortunately, illegible, but he was accorded a title (paduka sri maharaja)(104) current in east Java at the time.

It is worth noting that Ganesa figures, some bearing inscriptions, were common features of the east Javanese landscape during the thirteenth century, but were rare in Sumatra. The Porlak Dolok inscription was, in fact, found in association with another of these rare Sumatran Ganesa statues, as well as a cut stone lintel piece from a now vanished religious structure. Other, minor inscriptions found in the Padang Lawas area confirm the religious links between this area and east Java from the eleventh to fourteenth centuries.(105) The text of the Porlok Dolok inscription, and the offering to which it refers, must have been connected in some manner with the possibly Batak community of northeast Sumatra that produced the east Javanese-influenced tantric Buddhist monuments of Padang Lawas in the thirteenth and fourteenth centuries, during the period of Srivijaya's decline and dismemberment. Padang Lawas may have formed part of the state of "Dagroian", about which Marco Polo was so uncomplimentary in the 1290s.(106)

The contents of the Porlak Dolok inscription provide no immediate explanation for the use of Tamil language or for the clear Javanese influences. Of the two, however, the Javanese influence is the easier to explain. The probable date and the title used by the Javanese king in the Porlak Dolok inscription suggest that this inscription was issued at the behest of king Krtanagara (1254/68-1292 A.D.) of the Singosari state of east Java. Another inscription - on the stone base belonging to a relief statue of the Buddhist deity Amoghapasa and his thirteen companions, found further south at Padang Rotjo, was also written in a form of Old Javanese language,(107) and inscribed in the mainstream east Javanese kawi script of the period.(108) It is dated 1286 A.D. (1208 Saka), and bears a fuller version of the elaborate Javanese date than the Porlak Dolok inscription, including the day in the Javanese week. This particular text states that the Amoghapasa statue was brought to Suwarnadwipa (Sumatra) from Bhumi Jawa (Java), and erected at Dharmmasraya as a meritorious act (akan punya) of Sri Viswarupakumara, with the support of the Singosari ruler of Java, paduka sri maharajadhiraja sri Krtanagara - this meritorious act causing delight to the people of Bhumi Malayu (Jambi/Srivijaya) and their ruler, sri maharaja srimat Tribuwanaraja Mauliwarmadewa. The Javanese ruler Krtanagara, according to the fourteenth century Javanese text, the Nagarakrtagama, had sent a military expedition against Malayu in 1275, as a result of which the Sumatran state appears to have acknowledged its subservience to the east Javanese state. The relief statue that was sent to the Sumatran ruler was apparently one of several copies made of the statues of Amoghapasa and his companions at Candi Jago in east Java, the burial place of Krtanagara's father.(109)

According to the Padang Rotjo inscription, one of the officials accompanying the relief statue from Java to Sumatra was a member of the royal family, dyah samgat Payana hang(?) Dipangkaradasa. This name is very similar to that of the official appearing in the Porlak Dolok inscription, the text of which also concerns a meritorious act performed on behalf of a Javanese king bearing the same title. There are thus clear links - in content, language and possibly personnel - between the Old Javanese-language inscription of Padang Rotjo (dated 1286 A.D.) and the bilingual Tamil and Old Malay/Javanese language inscription of Porlak Dolok. Both appear to relate to a period of intense Javanese interaction with the east coast of Sumatra. Again, the use of Tamil language in a portion of the Porlak Dolok inscription is less easy to explain at this point.

c. Nakhom Si Thammarat, Peninsular Thailand

Also belonging to this late thirteenth-century group of t Tamil-language inscriptions is the fragmentary bilingual Tamil and Sanskrit inscription held in the collection of the Wat Mahadhatu at Nakhom Si Thammarat, south of Chaiya on the east coast of peninsular Thailand.(110) The legible portion of the damaged date may read 1183 Saka (1261 A.D.).(111) The text is too damaged to provide much information, but it appears to record a gift presented by Danma (Dharma) Senapati to the Brahmanas. The river Gangga is mentioned, and an imprecatory formula appears to occupy about half of the inscription. There are some similarities between this text and that of Porlak Dolok in northeastern Sumatra.

Although the original provenance of this inscription is unknown, there may be some connection between it and three pieces of what appear be south Indian statuary of about the same period that have been found in the Vieng Sra-Chaiya region.(112) Two of these - Visnu and Surya - appear to be Hindu; the third - Batuka-Bhairava - may be connected with the tantric Buddhist cult popular at the time in the Straits region.

d. Pagan, Upper Burma

The only other Tamil-language inscription that has so far to come to light in Southeast Asia is the bilingual Sanskrit and Tamil inscription found at Myinpagan, a mile to the south of Pagan in Upper Burma.(113) This Vaishnava inscription, which Hultzsch dated to the thirteenth century on palaeographic grounds, reads:

(Sanskrit): "Hail! Prosperity! [I have] no regard for merit, none for a heap of wealth, none at all for the enjoyment of lust. Whatever is to happen, O God! [will happen] in accordance with previous actions. This [alone] is to be prayed for [and] highly valued by me. In every other birth also let [me] possess unswerving devotion to the pair of thy lotus-feet!"

(Tamil): "Hail! Prosperity! Let the wealth of [this] temple increase! [In] the temple of Nanadesi-Vinnagar-Alvar at Pukkam (Pagan), alias Arivattanapuram (Arimaddanapura), I, Irayiran Siriyan alias Sri-Kulasekhara-Nambi of Magodayarpattanam in Malaimandalam, made a sacred mandapa, gave a sacred door, and gave one fixed lamp to hum constantly in this mandapa. [Let] this prosper! This meritorious gift [was made by] a native of Malaimandalam."

The opening Sanskrit verses were taken from a Vaishnava religious text. The Tamil portion of the text records a gift made by an individual from Malaimandalam (Cranganore on the Malabar coast) to the Visnu temple of the Nanadesi merchant association - possibly the Vaishnava structure which still exists at Pagan. Arimaddanapura was the ceremonial name of Pagan, which was capital of Burma from the eleventh to the late thirteenth century. The contents of this inscription bear some resemblance to those of the Tamil-language Telegu-Choda inscription of 1279 A.D. from the textile centre of Nellore in Karnataka,(114) which also mentions Malaimandalam involvement with the Nanadesi group, and in international trade. This forms one of a series of inscriptions, of particularly the twelfth and thirteenth centuries, that mention gifts to temples from merchants of this region. It seems likely that the Tamil inscription and temple-building at Pagan date to some time before 1279, since by the 1270s Pagan was already intermittantly under threat of Mongol attack. Pagan finally fell to the Mongol armies in 1287.

e. Quanzhou, Eastern China

A late thirteenth-century bilingual Tamil and Chinese-language inscription has been found associated with the remains of a Siva temple of Quanzhou. This was one of possibly two south Indian-style Hindu temples(115) that must have been built in the southeastern sector of the old port, where the foreign traders' enclave was formerly located.

The first six lines of the inscription are written in Tamil language, in Tamil script, A final half line is written in Chinese.(116) The translation made by T.N. Subramaniam(117) reads:

[Tamil]: "Obeisance to Hara (Siva)! Let there be prosperity! On the day Citra in the month of Chittira in the Saka year 1203 (1281 A.D.), the Tavachchakkarvarttigal Sambandhap-perumal (a Saiva religious leader) caused, in accordance with the firman (written permission) of Chekachai Khan (the Mongol ruler), to be graciously installed the God Udaiyar Tirukkadalisvaram Udaiya-nayinar (Siva), for the welfare of the illustrious body of the illustrious Chekachai Khan."

[The short Chinese text appended to the inscription has not yet been published.]

The timing and location of this Tamil foundation are important. As noted above, following the Yuan establishment of a maritime trade bureau in Quanzhou 1278, an envoy was dispatched to India early in 1281. This inscription, dated March/April 1281, must have been written shortly after the dispatch of the Yuan envoy to India, at the peak of south Indian trade in the port. Marco Polo states that Indian traders continued to be active in Quanzhou in the late 1280s and early 1290s.(118) However, after 1284, Yuan government policies relating to foreign trade began a series of disruptive shifts and reversals, and progressive restriction of Chinese merchant contact with foreigners continued through the fourteenth century.

The Indian temple in Quanzhou, referred to in the Tamil inscription, may also have been built and endowed during the brief window of relatively unencumbered trade opportunity that appeared between the setting up of the Yuan maritime trade bureau in 1278 - which both regulated and encouraged overseas trade - and the progressive restriction of that trade that began in the mid-1280s. The Indian merchants frequenting the port of Quanzhou clearly set out to woo Yuan government support through the dedication of their temple to prayers for the health of the emperor. They were also careful not to tread on local bureaucratic toes. Although the Indian community using the port must have organized itself in some manner, and members of the surviving south Indian merchant associations were undoubtedly involved, the Tamil text written in 1281 does not associate any particular merchant association with the temple foundation. It is unlikely that the Yuan government would have tolerated the existence of rival trading cartels - in any sector overseas trade connected with Chinese ports - operated by foreign merchant associations.


The direct impact of the south Indian merchant associations in Southeast Asia was limited both temporally and geographically. Not all South Asians trading in maritime Southeast Asia during the Asian sea trade boom were clearly associated with South Asian merchant groups. The evidence from Java and Bali during this period suggests much more complex groupings evolved, involving very mixed personnel and structures that were more Southeast Asian than South Asian.(119)

Of the eight Tamil-language inscriptions so far discovered in Southeast and East Asia, only three definitely mention south Indian merchant associations: the ninth century inscription from Takuapa on the west coast of the peninsula, the inscription dated 1088 A.D. from Lobo Tua on the west coast of Sumatra, and the thirteenth century inscription from Pagan in Upper Burma. The earliest of the associations to have left a record was the Manigramam, which appears to have been actively involved in transit trade bypassing the Malacca Straits during a period of local political turmoil. The later two mentions - separated by perhaps a century and a half - involve the Nanadesi branch of the Ayyavole, the major association which had also attached the Manigramam in some way by the tenth century. Most of the thirteenth-century Tamil inscriptions abroad do not appear to mention merchant associations, perhaps reflecting the sharp decline in the economic power of these associations within south India during the course of the thirteenth century.

Apart from the inscription of uncertain provenience held in Nakhom Si Thammarat and the late thirteenth inscription from China, all of the Tamil language inscriptions, and religious or other remains, that suggest the establishment of south Indian enclaves (of as yet uncertain nature or duration) have been found to the west of the Malacca Straits. They appear to have been confined to regions accessible directly from the Indian Ocean. These were areas that lay outside of the core region controlled by Srivijaya, and over which even the later, similarly-structured Sultanate of Malacca had no direct control. This distribution is understandable: it is very unlikely, given the degree to which Srivijayan rulers depended financially upon their control of trade, that they would have invited into their core ports any competing groups with great economic leverage. The lesson of the history of later Malacca Straits states is that these trading polities were only stable as long as the majority of the wealth flowed through the hands of their rulers. If too much wealth was dispersed into the hands of resident orang kaya (rich/powerful men), political breakdown invariably ensued. Given the temporal and spatial distribution of south Indian remains in Southeast Asia, and the fact that Srivijaya remained wealthy and stable for over three centuries - even though the capital appears to have shifted from Palembang to Muara Jambi by the eleventh century - it appears that during Srivijaya's heyday, between the early tenth and the early thirteenth centuries, Srivijaya's hold over the trade passing through the Malacca Straits remained firm.

1 A.L. Udovitch, Partnership and Profit in Medieval Islam (Princeton: Princeton University Press, 1970), pp. 171-72.

2 Yoshinobu Shiba, Commerce and Society in Sung China (Ann Arbor: University of Michigan Center for Chinese Studies, 1970), pp. 2, 188; Andrew Watson (trans.), Transport in Transition: the evolution of traditional shipping in China (Ann Arbor: University of Michigan Center for Chinese Studies, 1972), pp. 16ff; Laurence J.C. Ma, Commercial Development and Urban Change in Sung China (960-1279) (Ann Arbor: University of Michigan Department of Geography, 1971), pp. 29-43.

3 Meera Abraham, Two Medieval Merchant Guilds of South India (Delhi: Manohar, 1988), pp. 5, 19-24, 29, 110; E.P.N. Kunjan Pillai, Studies in Kerala History (Kottayam, Kerala: National Bookstall, 1970), pp. 370-77.

4 Abraham, Two Medieval Merchant Guilds, pp. 41-42.

5 For discussions of the various associated merchant, craft and mercenary groups, see for example: S.R. Balasubrahmanian, "The Tisai Ayirattainurruvar and the Municandai Record", Tijdschrift voor Indische Taal-, Land-, en Volkenkunde 74 (1934): 613-18; K.R. Venkatarama Ayyar, "Medieval Trade, Craft and Merchant Guilds in South India", Journal of Indian History 25,3 (1947): 269-80; K.A. Nilakanta Sastri, The Colas, 2nd revised edition (Madras, 1955); Burton Stein, "Coromandel Trade in Medieval India", in Merchants and Scholars: Essays in the History of Exploration and Trade, ed. J. Parker (Minneapolis: University of Minnesota Press, 1965), pp. 49-62; K. Indrapala, "Some Medieval Mercantile Communities of South India and Ceylon", Journal of Tamil Studies 2,2 (1970): 25-39; Abraham, Two Medieval Merchant Guilds.

6 Abraham, Two MedievalMerchant Guilds, pp. 117, 162-66.

7 Ruth Barnes, Indian Block-Printed Cotton Fragments in the Kelsey Museum, The University of Michigan (Ann Arbor: University of Michigan Press, 1993), pp. 30-31; Jan Wisseman Christie, "Texts and Textiles in 'Medieval' Java", Bulletin de l'Ecole Francaise d'Extreme-Orient 80,1 (1993): 193-95.

8 Indrapala, "Some Medieval Mercantile Communities", pp. 30-32.

9 Abraham, Two MedievalMerchant Guilds, p. 135.

10 J. Heitzman, "State formation in South India, 850-1280", The Indian Economic and Social History Review 24,1 (1987): 38; see also the rather inflated Pandyan claims of conquest in about 1254 A.D., which included not only the Chola state, but also Kedah, Java, and China (Archaeological Dept., Southern Circle, Madras, Annual Report on Epigraphy 1911-12, p. 65; 1916-17, p. 111).

11 F. Hirth and W.W. Rockhill, Chau Ju-kua: His Work on the Chinese and Arab Trade in the Twelfth and Thirteenth Centuries, entitled Chu-fan-chi (New York: Paragon, 1966 reprint of 1911 edition), pp. 87-102.

12 Abraham, Two Medieval Merchant Guilds, p. 154.

13 K.A. Nilakanta Sastri, "Foreign trade under the Kakatiyas", Journal of Oriental Research 8 (1934): 319.

14 Abraham, Two MedievalMerchant Guilds, pp. 34-35, 51-60.

15 For detail, see: Jan Wisseman Christie, "The banigrama in the Indian Ocean and the Java Sea during the early Asian trade boom", in Seafaring Communities in the Indian Ocean, 4th c. B.C.-15th c. A.D., ed. J.F. Salles and H.P. Ray (Brussels: Brepols, in press).

16 M. Boechari, "Prasasti Kalirungan" (unpublished ms.).

17 H.B. Sarkar, Corpus of the Inscriptions of Java (Up to 928 A.D.), two volumes (Calcutta: K.L. Mukhopadhyay, 1971), inscription xciii.

18 S.D. Goitein, Letters of Medieval Jewish Traders (Princeton: Princeton University Press, 1973), pp. 175-229; S.D. Goitein, "From Aden to India", Journal of the Economic and Social History of the Orient 23,1&2 (1980): 46; R. Chakravarti, "Merchants of Konkan", The Indian Economic and Social History Review 23,2 (1986): 215.

19 Christie, "Texts and Textiles", pp. 193-95.

20 J.L.A. Brandes, Oud-Javaansche Oorkonden, nagelaten transscripties van wijlen Dr. J.L.A. Brandes, uitgegeven door Dr. N.J. Krom (The Hague: Martinus Nijhoff, 1913), inscription lviii.

21 Brandes Oud-Javaansche Oorkonden, inscription lix. The Kaladi inscription - for which, see A.M. Barrett Jones, Early Tenth Century Java from the Inscriptions (Dordrecht: Foris, 1984), pp. 178-94 - from the same region, which is dated 909 A.D., but which survives only as a later reissue, preserves an anachronous list identical to those of the Cane and Patakan inscriptions.

22 Paul Wheatley, "Geographical Notes on some Commodities involved in Sung Maritime Trade", Journal of the Malaysian Branch of the Royal Asiatic Society 32,2 (1959): 59-61; Hirth and Rockhill, Chau Ju-kua, pp, 88, 92-93, 96-98.

23 Christie, "Texts and Textiles", p. 194.

24 Brandes, Oud-Javaansche Oorkonden, inscription lxiv. From Truneng, in the Brantas delta region. Undated, but commissioned late in Airlangga's reign, probably in the 1040s.

25 M. Boechari, "Prasasti Garaman, 1053 A.D." (unpublished ms.).

26 Jan Wisseman Christie, "Javanese markets and the Asian sea trade boom of the tenth to thirteenth centuries A.D.", Journal of the Economic and Social History of the Orient 41,3 (1998).

27 R.Ng. Poerbatjaraka, "Vier oorkonden in koper", Tijdschrift voor Taal-, Land-, en Volkenkunde 76 (1936): 378.

28 A. van Aelst, "Majapahit picis: the currency of a 'moneyless' society", Bijdragen tot de Taal-, Land-, en Volkenkunde 151,3 (1995): 357-93; Jan Wisseman Christie, "Money and its uses in the Javanese states of the ninth to fifteenth centuries A.D.", Journal of the Economic and Social History of the Orient 39,3 (1996): 243-86.

29 Th.G.Th. Pigeaud, Java in the 14th Century: A Study in Cultural History, five volumes (The Hague: Martinus Nijhoff, 1960-63), canto 83.4.

30 Christie, "Texts and Textiles", p. 194; Ruth Barnes, "From India to Egypt: The Newberry Collection and the Indian Ocean textile trade", in Seafaring Communities in the Indian Ocean, 4th c. B.C.-15th c. A.D., ed. J.F. Salles and H.P. Ray (Brussels: Brepols, in press).

31 There is, in fact, another, much earlier Tamil-language inscription, not considered here, that has been found in Southeast Asia. It comprises two words written on a small stone that was apparently used as a goldsmith's touchstone, now held in the museum at Wat Khlong Thom, Krabi, on the west coast of peninsular Thailand, about 120 km. south of Takuapa. Khlong Thom was the site of a very early port and manufacturing centre specializing in the production of beads and other jewelry. The short inscription, written in Brahmi script of the third or fourth century A.D., reads perumpatan kal, meaning, "the [touch]stone of the master goldsmith". This was the personal property of an individual artisan, rather than an inscription set up for public attention. See Norobu Karashima, "Indian Commercial Activities in Ancient and Medieval Southeast Asia" (Paper delivered at the Conference of the International Association of Tamil Research, 1995), pp. 3-4.

32 Wang Gungwu, "The Nanhai Trade", Journal of the Malayan Branch of the Royal Asiatic Society 31, 2 (1958): 123.

33 Chinese ceramics of ninth and early tenth century date have been recovered from one of the early sites at Palembang - see Ho Chui Mei, "Ceramics from the Palembang Excavations", ACRO Update 2 (1995): 1; the Jambi site connected with that period has yet to be located.

34 O.W. Wolters, Early Indonesian Commerce: A Study of the Origins of Srivijaya (Ithaca: Cornell University Press, 1967), pp. 187-88.

35 Wang, "The Nanhai Trade", pp. 101, 104-105, 123.

36 The Sanskrit inscription on the front of a stone found in a collection in Nakhom Si Thammarat - dated 775 A.D. and lauding the king of Srivijaya - appears to have been superseded by an undated, but palaeographically similar Sanskrit inscription on the back of the stone, lauding a king of the Sailendra family that was in power in central Java at the time. See G. Coedes, Recueil des Inscriptions du Siam, Deuxieme Partie: Inscriptions de Dvaravati, de Crivijaya et de Lavo (Bangkok: The Siam Society, 1962), pp. 20-24.

37 The Nalanda inscription, commissioned in about 860 A.D., concerning a Buddhist foundation funded by a Sumatran ruler, mentions a king of Suvarnadvipa (Sumatra), but not the state of Srivijaya; this ruler's main claim to fame appears to have been the fact that he was the grandson of the Sailendra ruler of Java - H. Shastri, "The Nalanda Copper-plate of Devapaladeva", Epigraphia Indica 17,7 (1924): 310-27.

38 K.A. Nilakanta Sastri, "The Takua-pa (Siam) Tamil Inscription", Journal of Oriental Research 16 (1942): 326-27; see also the archaeological evidence discussed below.

39 Wang, "The Nanhai Trade", p. 123.

40 This impression of the relative importance of Java in maritime Southeast Asia during the late eighth and ninth centuries is reinforced by Cham, Vietnamese and Cambodian sources. See Jan Wisseman Christie, Patterns of Trade in Western Indonesia: Ninth through Thirteenth Centuries A.D. (Ph.D. diss., School of Oriental and African Studies, University of London, 1982), pp. 3031; G. Coedes, Les Inscriptions du Cambodge, eight volumes (Paris: Ecole Francaise d'Extreme-Orient, 1942-66), volume I, p. 42; G. Coedes, The Indianized States of Southeast Asia (trans. S.B. Cowing) (Honolulu: East-West Center Press, 1968), pp. 123, 315 n. 104.

41 G. Ferrand, L'Empire Sumatranais de Crivijaya (Paris: Paul Geuthner, 1922), pp. 50-52; G.R. Tibbetts, A Study of the Arabic Texts Containing Material on South-East Asia (Leiden: E.J. Brill, 1979), pp. 25-32. Given the shifting political situation in maritime Southeast Asia during the centuries between the ninth century, when these Arab reports were first written, and the time of their reuse in later compilations, there remains some doubt concerning the exact identity of Zabaj. Tibbetts (ibid., p. 112) is almost certainly correct, however, in his conclusion that the term Zabaj was first, in the ninth century, attached to Java and its economic and political sphere in the eastern islands, and that only later, after the port hierarchy of Srivijaya had reassembled itself, did confusion set in.

42 Christie, "Money and its uses", pp. 245-47.

43 Coedes, Recueil des Inscriptions du Siam, inscription xxvi.

44 S.J. O'Connor, Hindu Gods of Peninsular Siam. supplementum to Artibus Asiae 28 (1972), figs. 28-31.

45 E. Hultzsch, "Note on a Tamil Inscription in Siam", Journal of the Royal Asiatic Society (1913): 337-39; and "Supplementary Note on a >Tamil Inscription in Siam", Journal of the Royal Asiatic Society (1914): 397-98.

46 "Takuapa and its Tamil Inscription", Journal of the Malayan Branch of the Royal Asiatic Society 22,2 (1949): 29.

47 Kunjan Pillai, Studies in Kerala History, p. 371.

48 Balasubrahmaniam, "The Tisai Ayirattainurruvar", p. 618; Abraham, Two Medieval Merchant Guilds, pp. 45, 183, 248.

49 Bennet Bronson, "Chinese and Middle Eastern Trade in Southern Thailand during the 9th Century A.D.", in Ancient Trades and Cultural Contacts in Southeast Asia, ed. Amara Srisuchat (Bangkok: Office of the National Cultural Commission. 1996), pp. 181-200; H.A. Lamb, "Takuapa, the probable site of a pre-Malaccan entrepot in the Malay peninsula", in Malayan and Indonesian Studies, ed. J. Bastin and R. Roolvink (Oxford: Clarendon Press, 1964); Ho Chui Mei, Pisit Charoenwongsa, Bennet Brunson, Amara Srisuchat and Tharapong Srisuchat, "Newly Identified Chinese Ceramic Wares from Ninth Century Trading Ports in Southern Thailand", SPAFA Digest 11,3 (1990): 12-17.

50 Wheatley, "Geographical Notes", p. 25.

51 See, for example, the reports on the upper levels at Arikamedu, near Pondicherry: R.E.M. Wheeler, A. Ghosh and Krishna Deva, "Arikamedu: An Indo-Roman Trading Station on the East Coast of India", Ancient India 2 (1946): 91: on the site of Kanchipuram in Tamil Nadu, see Government of India, Indian Archaeology (Delhi: Government of India, 1969-70), p. 37; on the tenth-eleventh century site of Madilakam, near Cranganore on the Malabar coast, see Government of India, ibid., p. 15.

52 In 1005 A.D. the king of Srivijaya, who was also "lord of Kataha (Kedah)" built in south India a Buddhist vihara, for which the Chola ruler later granted revenues. See S.K. Aiyangar and R. Sewell, Historical Inscriptions of Southern India (Madras, 1932), pp. 57-58; and Epigraphia Indica 22. no. 34. A decade later, in 1014-15 A.D., the ruler of Srivijaya presented gifts to a Hindu temple in the Chola state (Annual Report on Indian Epigraphy 1956-57: 15. nos. 161 and 164). In 1018-1019 A.D., the ruler of Srivijaya and Kedah presented gifts of "Chinese gold" and other objects to the same Hindu temple (Annual Report of Indian Epigraphy 1956-57: 15, no. 166). These inscriptions have also been discussed by R.C. Majumdar, Ancient Indian Colonies in the Far East, volume 2: Suvarnadvipa (Calcutta, 1937), and more recently by J.N. Miksic, "Hubungan sejarah antara Seriwijaya, Palembang dengan Lembah Bujang", Tamadun Melayu 3 (1995): 894-917.

53 Grace Wong, A comment on the tributary trade between China and Southeast Asia, and the place of porcelain in the trade during the Song dynasty in China (Singapore: The Southeast Asian Ceramic Society, 1979), no. 15.

54 O.W. Wolters, "Tambralinga". Bulletin of the School of Oriental and African Studies 21,3 (1958): 605.

55 Wong, A comment on the tributary trade, no. 16.

56 Aiyangar and Sewell, Historical Inscriptions, pp. 65-66; Coedes, The Indianized States, pp. 142-43; Nilakanta Sastri, The Colas, pp. 211-20. Rajendra's list of conquests includes: Srivijaya, Pannai (north Sumatra), Malaiyur (Jambi), Mayirudingam (?), Ilangasogam (Langkasuka), Mappappalam (Lower Burma?), Mevilimbangam (Palembang?), Valaippanduru (?), Talaittakkolam (Takuapa), Madamalingam (Tambralinga), Ilamuridesam (Lamri, Aceh), Manakkavaram (Nicobar Islands), Kadaram (Kedah). It is interesting to note that Rajendra's son later claimed on his father's behalf only the conquest of Kidaram (Kedah) (Annual Report of South Indian Epigraphy 1912-13: 961, 26).

57 Wong, A comment on the tributary trade, no. 11.

58 Wheatley, "Geographical Notes", p. 25.

59 Paul Wheatley, The Golden Khersonese (Kuala Lumpur: University of Malaya Press, 1961), p. 61.

60 Epigraphia Indica 25, part 6: no. 25.

61 Hirth and Rockhill, Chau Ju-kua, p. 59. The Sung Shih states that the Cholas were subject to Srivijaya, a statement which may reflect the fact that the Cholas traded with China under the aegis of Srivijaya, and thus may have accompanied their missions to the court. For other views on this passage in the Sung Shih, see also: R.C. Majumdar, Ancient Indian Colonies in the Far East, volume 2: Suvarnadvipa (Calcutta, 1937), pp. 182-90; Tan Yeok Seong, "The Srivijaya inscription of Canton (A.D. 1079)", Journal of Southeast Asian History 5,2 (1964): 21.

62 Wheatley, "Geographical Notes", p. 25. Middle Eastern missions, which continued to arrive into the second half of the twelfth century, carried ivory, rhinocerous horn, frankincense, rosewater, pearls, glassware, etc. (Wong, A comment on the tributary trade, no. 16). Missions from Coromandel carded many of the same commodities (Wong, ibid., no. 15). Both also carried such Southeast Asian produce as camphor (probably from west Sumatra) and cloves (from eastern Indonesia, but bought in ports of Java or Srivijaya). Missions from Srivijaya carried an almost identical range of goods (Wong, ibid., no. 11), presumably to advertise its role as entrepot for Indian Ocean commodities.

63 The history behind the temple's reconstruction and the subsequent gifts of land is detailed in a Chinese-language inscription of 1079 A.D. (Tan, "The Srivijaya inscription").

64 Wong, A comment on the tributary, trade, no. 11.

65 There was further Srivijayan activity during the 1080s in connection with Buddhist foundations in south India (Nilakanta Sastri, The Colas, pp. 271-72; Epigraphia Indica 22: no. 35).

66 The Javanese state involved was by this time probably the east Javanese coastal state of Janggala, created during the period of civil strife that followed the death of Airlangga, which is referred to in an inscription of king Garasakan, dated 1053 A.D. (Boechari, "Prasasti Garaman"); on the favours shown to the Javanese mission by the Chinese emperor in 1129 A.D., see W.P. Groeneveldt, Historical Notes on Indonesia and Malaya compiled from Chinese Sources (Jakarta: Bhratara, 1960), p. 19.

67 Wheatley, "Geographical Notes", p. 32; Aelst, "Majapahit picis", p. 361; Christie, "Money and its uses", pp. 269-70.

68 T. Yamamoto, "Reexamination of historical texts concerning Srivijaya" (Paper presented at the 8th biennial conference of the International Association of the Historians of Asia, Kuala Lumpur, Malaysia, 1980), p. 5.

69 Hirth and Rockhill, Chau Ju-kua, pp. 23, 25.

70 Hirth and Rockhill, ibid., p. 24.

71 The inscription was first read by E. Hultzsch - Madras Epigraphy Report (Madras: Government of India, 1892) - who summarized its main points, but did not publish a transcription or full translation; his comments formed the basis of studies by K.A. Nilakanta Sastri - "A Tamil Merchant-guild in Sumatra", Tijdschrift voor het Taal-, Land- en Volkenkunde 72 (1932): 314-27; Balasubrahmanian, "The Tisai Ayirattainurruvar"; F.M. Schnitger, The Archaeology of Hindoo Sumatra, supplement to Internationales Archiv fur Ethnographie 35 (1937): 15; and Indrapala, "Some Medieval Mercantile Communities". Karashima ("Indian Commercial Activities", pp. 6-7) actually provides the first transcript and translation, made by Prof. Y. Subbarayalu of Tamil University. The version of the contents above derives from this reading.

72 "The Five Hundred of the Thousand Directions in All Countries", one of the branches of the Ainnurruvar (Ayyavole) merchant association. A number of inscriptions of the eleventh century in south India and Sri Lanka mention the Ainnurruvar (Ayyavole) merchant association, and a few mention the Tisaiyayirattainnuruva Nanadesa group (Abraham, Two Medieval Merchant Guilds, pp. 186-98).

73 This term, which means "the settlement on the seashore", also appears in the eulogy to ainnurruvar in several inscriptions in south India. These eulogies generally state that the merchant group was involved in commerce in 18 pattinam, 32 velapuram, and 64 kadigaitavalam (Karashima, "Indian Commercial Activities", p. 8). This velapuram enclave attached to Barus seems, thus, to have been a trading settlement of secondary rank; the main port at Barus, however, appears to have been classed as a pattinam, a commercial centre of first rank. This accords with the international significance at the time of Barus, as a major camphor-exporting port.

74 The title nagara-senapati also appears in the Telegu language inscription of about 1090 A.D., from Vishakhapatnam on the Andhra coast of south-eastern India, as that of the recipient, along with the padinettu (18) bhumi, of a grant from a merchant group (Abraham, Two Medieval Merchant Guilds, p. 196). The term nagaram, in south India, referred to a commercial settlement dominated by merchants, but also containing artisans (Abraham, ibid., p. 84). The term senapati was derived from Sanskrit and referred to a military commander; it also appears in numerous Southeast Asian inscriptions, mostly in local languages, including several in Sumatra and a larger number in Java. The personal name or designation of this commander of the commercial settlement is of particular interest, since it appears to connect him with the Nattukottai Chettiar merchant community, who later traced their ancestry back to the Ayyavole association (Abraham, ibid., p. 5).

75 The term marakkala-nayan used here, may be related to the term marakkayar, which was attached to sea-faring Muslim merchants of both the Coromandel and Malabar coasts of south India in the fifteenth century (Karashima, "Indian Commercial Activities", p. 8).

76 Karashima (ibid., p. 8) suggests that this term might refer to the ship's owner; it may, however, have carried a meaning closer to the term tunda, which appears in east Javanese inscriptions of the tenth and eleventh centuries, referring to the deck of a vessel, in connection with the classification of boats by size for tax purposes.

77 This word was adopted into Malay and Javanese vocabularies as a term either for the musk of the musk deer (Moschus moschiferus, Linn.) or of the Malay civet (Viverra tangalunga), which was somewhat different from the Indian civet (V. zibetha, Linn.), Malabar civet (Moschothera civetina, Blyth), and the true African civet (V. civetta, Schreber), all three of which were imported by the Chinese, via Southeast Asia, during this period (Wheatley, "Geographical Notes", p. 105). South Indian merchant associations exported both deer musk and civet (Abraham, Two Medieval Mercantile Guilds, pp. 161, 172). It appears that, in the port of Barus, "entrance" fees were pegged to the value of the in-coming cargoes rather than that of the out-going cargoes of camphor.

78 Goitein, Letters of Medieval Jewish Traders, pp. 227-29.

79 B. Bronson, Basoeki, M. Suhadi and J. Wisseman, Laporan Penelitian Arkeologi di Sumatera (Jakarta: Pusat Penelitian Arkeologi Nasional, 1973), pp. 18-19. Samples of the glass and ceramics are held in the offices of the Pusat Penelitian Arkeologi Nasional, in Jakarta. One fragment of the kawi inscription is held by the Museum Nasional in Jakarta, while the other remains in the village of Lobo Tuwa.

80 Schnitger, Archaeology of Hindoe Sumatra, p. 15; R.S. Wicks, Money, Markets and Trade in Early Southeast Asia: The Development of Indigenous Monetary Systems to AD 1400 (Ithaca: Cornell University Southeast Asia Program, 1992), fig. 7.3; Christie, "Money and its uses", p. 246.

81 E.E. McKinnon, " Tamil Imagery in Northeast Sumatra", Oriental Art 40,3 (1994): 15-24; and "Medieval Tamil Involvement in Northern Sumatra, c11-c14 (The Gold and Resin Trade)", Journal of the Malaysian Branch of the Royal Asiatic Society 69,1 (1996): 85.

82 This provisional transcription and translation has been made by L. Thyagarajan (P.Y. Manguin, personal communication, 29 Jan. 1996). Further work on the text may produce more information.

83 McKinnon, "Medieval Tamil Involvement".

84 E.E. McKinnon and Luckman Sinar, "A Note on Pulau Kompei in Am Bay, Northeastern Sumatra", Indonesia 32 (1981): 49-73.

85 E.E. McKinnon, "Research at Kota Cina, a Sung-Yuan period trading site in East Sumatra", Archipel 14 (1977): 19-32; and "Tamil Imagery"; John Miksic, "Classical Archaeology in Sumatra", Indonesia 30 (1980): 43-66.

86 H.E Schurmann, Economic Structure of the Yuan Dynasty: Translation of Chapters 93 and 94 of the Yuan shih (Cambridge. Mass.: Harvard University Press, 1967), p. 230.

87 John Guy, "The Lost Temples of Nagapattinam and Quanzhou: A Study in Sino-Indian Relations", Silk Road Art and Archaeology 3 (1993-94): 300.

88 Schurmann, Economic Structure, pp. 223-25.

89 Abraham, Two MedievalMerchant Guilds, p. 150.

90 M. Elvin, The Pattern of the Chinese Past: A Social and Economic Interpretation (Stanford: Stanford University Press, 1973), p. 217.

91 Schurmann, Economic Structure, pp, 225-27.

92 R.E. Latham (trans.), The Travels of Marco Polo (Harmondsworth: Penguin, 1958), p. 237.

93 Schurmann, ibid., p. 224.

94 Guy, "Lost Temples", p. 299.

95 Coedes, The Indianized States, pp. 198-201.

96 Latham, Marco Polo, p. 254.

97 Boechari, Prasasti Koleksi Museum Nasional, vol. 1 (Jakarta: Museum Nasional, 1985/86), pp. 164-68.

98 It was reported in Oudheidkundig Verslag 1912: 46, no. 39, and in Notulen van het Bataviaasch Genootschap van Kunsten en Wetenschappen 1911: 128, 14e; it was referred to by N.J. Krom, in Hindoe-Javaansch Geschiedenis (The Hague: Martinus Nijhoff, 1931), p. 410, but no transcription or translation has been published. See also J.G. de Casparis, "Peranan Adityawarman, Putera Melayu di Asia Tenggara", Tamadun Melayu 3 (1995): 933-64; Satyawati Suleiman, The Archaeology and History of West Sumatra (Jakarta: Research Centre of Archaeology of Indonesia, 1977) [Berita Pusat Penelitian Purbakala dan Peninggalan Nasional, no. 12, p. 5].

99 The stone is housed in the Museum Nasional in Jakarta, where it is numbered D 181, but now apparently lacks a record of its provenance. See P.V. van Stein Callenfels, "Rapport over een diensreis door een deel van Sumatra", Oudheidkundig Verslag (1920): 70, for the initial report of the find and its context; see Karashima, "Indian Commercial Activities", p. 9, for discussion of the content of the text in both languages, based partly upon a provisional reading by Sukarto K. Atmodjo. A provisional transcription of the Old Malay/Javanese-language portion of the text is presented by Edi Sedyawati, Ganesa statuary of the Kadiri and Singhasari periods: A study of art history (Leiden: KITLV Press, 1994), p. 326, along with inscriptions on a number of Ganesa statues from east Java. The translation above is based on her transcription.

100 Edi Sedyawati, Ganesa statuary, p. 326.

101 Karashima, "Indian Commercial Activities", p. 9.

102 Callenfels ("Rapport over een dienreis", p. 70) read the date as 1167 Saka (1245 A.D.), but L.C. Damais later, in "Etudes d'Epigraphie Indonesienne III: Liste des Principales Inscriptions Datees de l'Indonesie", Bulletin de l'Ecole francaise d'Extreme-Orient 46,1 (1952): 100, corrected it to 1135 Saka (1213 A.D.); Karashima ("Indian Commercial Activities", p. 9), using the incomplete astronomical data given in the inscription, suggests three possible dates: 1187, 1258 or 1265 A.D., but on palaeographical grounds prefers the last two.

103 Since the term nagara appears here in the text written in Old Malay/Javanese, it may have been used in the Javanese manner (meaning "palace") rather than the Tamil manner (meaning "commercial centre"). Both meanings were later combined in the Malay term negri, derived from nagara, which referred to the port-capital of a Malay coastal trading state and, by extension, to the state as a whole. How early this fused meaning was adopted is unclear.

104 This title was the standard title accorded to kings of east Java from the eleventh century into the fourteenth century. The title is rendered as peritu sri maharaja in the parallel Tamil text (Karashima, "Indian Commercial Activities", p. 9).

105 See Rumbi Mulia, The Ancient Kingdom of Panai and the Ruins of Padang Lawas. (Jakarta: Research Centre of Archaeology of Indonesia, 1980) [Berita Pusat Penelitian Arkeologi Nasional, no. 14].

106 Latham, Marco Polo, pp. 255-56.

107 This inscribed base was found at Padang Rotjo, upstream from Sungai Langsat on the Batang Hari river, along with other remains of the thirteenth and fourteenth centuries, including the giant stone Bhairava statue that has been taken to be a portrait-statue of the mid-fourteenth century Sumatran ruler Adityavarman. The relief statue of Amoghapasa, to which the inscribed base was originally attached, was later moved to Rambahan, not far away, thus creating some confusion over the original provenance of the two pieces. See N.J. Krom, "Een Sumatraansche Inscriptie van Koning Krtanagara", Verslagen en Mededelingen der Koninklijk Akademie van Wetenschappen, afdeeling Letterkunde, Series 5, volume II (1916): 306-339; Ferrand, L'Empire Sumatranais, pp. 179-81; A.J. Bernet Kempers, Ancient Indonesian Art (Amsterdam: C.P.J. van Peet, 1959), pp. 87-88.

108 J.G. de Casparis, Indonesian Palaeography: A history of writing in Indonesia from the beginnings to c. A.D. 1500 (Leiden: E.J. Brill, 1975), p. 57.

109 Bernet Kempers, Ancient Indonesian Art, pp. 84-88: Pigeaud, Java in the 14th Century, vol. III, p. 47.

110 Coedes, Recueil des Inscriptions du Siam, inscription xxix; Karashima, "Indian Commercial Activities", p. 9.

111 Karashima, "Indian Commercial Activities", p. 9.

112 O'Connor, Hindu Gods, figs. 32-34.

113 E. Hultzsch, "A Vaishnava Inscription at Pagan", Epigraphia Indica 7 (1902-1903): 197-98.

114 Abraham, Two MedievalMerchant Guilds, pp. 232-33.

115 Jean Filliozat, "Research in South-East Asia and the Far East", Tamil Culture 12 (1966): 115; Guy, "Lost Temples", p. 296.

116 Guy, "Lost Temples", p. 298.

117 T. N. Subramaniam, "A color=Blue>Tamil< Colony in Medieval China", in South Indian Studies, ed. R. Nagaswamy (Madras: Society for Archaeological, Historical and Epigraphical Studies, 1978), p. 8.

118 Latham, Marco Polo, p. 237.

119 Javanese inscriptions indicate that tax farming involving mixed groups of foreign and local merchants was common in the Brantas delta by the eleventh century. None of these groups - which were labelled banigrama by the Javanese courts - was exclusively south Indian in composition, and none of them acted extraterritorially. The composition of these tax-farming groups of merchants shifted over time (most significantly to include Chinese members after the twelfth century), and although Indian merchants were always included, these Javanese tax arrangements long outlasted the heyday of both the Chola empire and the south Indian merchant associations - and they outlived many individual Javanese states. See Christie, "Banigrama"; and Jan Wisseman Christie, "Javanese markets and the Asian sea trade boom of the tenth to thirteenth centuries A.D.", Journal of the Economic and Social History of the Orient 41,3 (1998).

COPYRIGHT 1998 Singapore University Press Pte. Ltd. (Singapore)

COPYRIGHT 1999 Gale Group