Inside Cham society: land given to the gods "together with its inhabitants" (c. 400), slaves endowed to the Buddhist monastery at Đồng Dương (875), a reported custom of royal widow-burning (1307). From the Vietnamese advance: ancestral temples destroyed in 982; by the annals' own count 30,000 killed in 1044 and over 40,000 killed and 30,000 captured in 1471; 5,000 families fleeing to Cambodia in 1692–93; annexation of the last principality in 1832 and two risings crushed; Mỹ Sơn heavily bombed in August 1969.
CONNECTIONS · 400–1832 · RELIGION · From Early medieval Indic religious traditions → Cham (Champa)

Śiva on the Cham coast (c. 1000) — and the march that ended Champa

Around 400 a Cham king founded a liṅga of Śiva at Mỹ Sơn and gave it his own name. For a thousand years the kings of the coast built for Indian gods. Then a Vietnamese state pressed south, took the sanctuary's land in 1471 and the last principality in 1832.

Around 400, in a valley behind the central Vietnamese coast, a Cham king named Bhadravarman founded a liṅga of Śiva, called it Bhadreśvara after himself, and gave it the valley's land "together with its inhabitants". Mỹ Sơn became the dynastic sanctuary of a seafaring Austronesian people who took Sanskrit, Śiva and the goddess of Po Nagar as the language of kingship, without any Indian conquest. From 982 the Vietnamese state to the north pressed south: by its own annals 30,000 Cham were killed in 1044, and in 1471 over 40,000 were killed and 30,000 captured when Vijaya fell. Worship at Mỹ Sơn ended, the last principality was annexed in 1832, and in August 1969 American bombing laid waste to the sanctuary. The religion survives among the Cham Bà-la-môn of Ninh Thuận.

A tall red-brick temple tower with a saddle-shaped roof and pilastered walls, beside a lower, partly collapsed tower, with trees behind, low brick foundations on a lawn in front, and a blue sky.
Brick towers of Mỹ Sơn groups B, C and D, in the valley where Bhadravarman founded the liṅga Bhadreśvara around 400. Most of the surviving towers were built in the tenth century, at the height of Champa's Śaiva royal cult; the sanctuary was abandoned after the Vietnamese conquest of 1471.
Chainwit. Mỹ Sơn sanctuary, groups B, C and D (Champa, c. 7th–13th century), photographed 2024. In situ, Mỹ Sơn, Quảng Nam, Vietnam. CC BY 4.0 via Wikimedia Commons. · CC BY 4.0

Before: a seafaring people and its gods of the soil

Around 400 CE the coast of what is now central Vietnam, from the Hải Vân pass south to the plains of Nha Trang and Phan Rang, belonged to a people who had come by sea. The Cham spoke a language of the Austronesian family, and Michael Vickery, whose 2005 reappraisal is the starting point for any modern history of Champa, puts the point without qualification: "There is no longer doubt that Cham is an Austronesian language, closely related to Achenese and Malay and more distantly to other languages of Indonesia, the Philippines, Polynesia and Madagascar" 3. Archaeologists have long linked the iron-age Sa Huỳnh culture of the same coast to the first Cham settlers, though Vickery cautions that "ongoing research may modify this, and it is probable that there was more than a single landfall by the first seaborne Cham" 3. Their country was a string of narrow river plains between the Trường Sơn mountains and the South China Sea, each with its own estuary, its own harbour and its own chief. Wet rice was grown on the plains, fish taken from the sea, and the harbours lay on the sea road between the Indian Ocean and the ports of south China 8.

Chinese records of the period know a kingdom on this coast called Linyi (林邑). By their account it was founded around 192 CE, when the people of the Xianglin district, the southernmost outpost of Han rule, rose against their Chinese overlords and founded a state of their own 28. The ICOMOS evaluation that accompanied Mỹ Sơn's World Heritage nomination in 1999 gives the date as the beginning of the Champa kingdom 8. Vickery doubts the equation: in his reading the early Linyi of the Chinese sources was probably a separate polity, perhaps not even Cham-speaking, and the Sanskrit-using kingdoms of the coast should be studied from their own inscriptions rather than from Chinese descriptions of a neighbour 3.

Archaeology has begun to fill the gap between the Chinese texts and the first inscriptions. Majumdar, following the French geographer Léonard Aurousseau, placed the early capital at Trà Kiệu, on the Thu Bồn river a few kilometres downstream from the later sanctuary of Mỹ Sơn 1. Excavations there since the 1990s, and at the nearby site of Gò Cấm, have found brick structures, roof tiles and imported goods of the early centuries CE, and the 2011 volume edited by Bruce Lockhart and Trần Kỳ Phương gathers the results 25. What they show is a coastal town already in contact with both the Han world to the north and the trading networks of the south seas, before any Sanskrit was written there. The Indian religion did not arrive in an empty place. It arrived in a society that already had towns, chiefs, foreign trade and gods of its own.

Gods held in stones

What religion did the Cham coast have before Śiva? No text survives from it, and the answer has to be read back from later practice. The fullest attempt was made in 1933 by Paul Mus, in an essay on "Indian and indigenous cults in Champa" that remains the most influential French study of the subject 2. Mus argued that the religion beneath the Sanskrit inscriptions was a cult of the soil: a god of a particular place, present in a raw stone, a post, a tree or a mound, and called out of it by the right rites. The cult, he wrote, addressed not a true "spirit" but "un sol-dieu plutôt qu'un dieu du sol", a god-soil rather than a god of the soil 2.

Mus found the evidence in the Cham villages of his own day, where the objects of worship in the old towers included Indian statues, liṅgas and plain stones alike: "Les Chams ne font aucune différence entre ces trois espèces d'objets : tous leur figurent également le corps même des divinités" 2. From this he drew the conclusion that has shaped every discussion since. The liṅga, Śiva's stone emblem at the centre of a Sanskrit temple, did not replace the soil-god's stone; it succeeded it. "Le linga est l'héritier de la pierre magique, représentation abstraite et tangible du dieu du sol" 2. And Mus went one step further. The god, though always one with the stone, was also present "dans la personne du chef ou du prêtre qui l'évoque" 2. A ruler who summoned the god of the land shared in it. That is the point at which the Indian religion would find its purchase. A Sanskrit cult that made the king the founder and namesake of the land's highest god did not ask the Cham chiefs to give up their claim to sacred power; it gave that claim a larger vocabulary, a written form and a place in a cosmos that reached from the Himalaya to the sea.

What the coast held around 400

The religious and political order on the eve of the Sanskrit inscriptions can be summarised in five features, each of which the later record preserves:

  • A seaborne Austronesian people settled in separate river plains, with a language related to Malay and Acehnese 3.
  • Gods of place held in stones, posts and mounds, in Mus's reconstruction from later Cham practice 2.
  • Chiefs who were also the gods' summoners, so that ritual authority and political authority were joined before any Indian text arrived 2.
  • Overseas trade as the economic base alongside wet-rice farming, with the coast on the main sea route between India and China 28.
  • Many polities, not one: Vickery's argument that "there was no single kingdom or state of Champa" applies with special force to this early period 3.

Mus himself believed that "des relations commerciales, des voyages individuels et peut-être des déplacements plus massifs avaient dû déjà relier l'Inde à la côte d'Annam" before the first inscription was cut 2. The ships that brought pepper buyers and Buddhist pilgrims along this coast also brought people who knew Sanskrit, and the Cham chiefs found in their language a way of saying what they already believed about themselves.

The transmission: Bhadreśvara, the god who bore the king's name

The first inscriptions

The oldest Sanskrit inscription on the coast is a stele found at Võ Cạnh, near Nha Trang. It records a gift by a descendant of a king named Śrī Māra, and it was long dated to the second or third century on the strength of George Coedès's identification of Śrī Māra with a ruler of Funan 10. The date is not secure. The Indian epigraphist D. C. Sircar argued in 1939 that "the language, style and the meter used in the record prove, in my opinion, that the inscription is later—at least not earlier than the fourth century A.D." 5. Which polity set it up, Linyi, Funan or an early Cham principality, is still argued. The Cham inscriptions as a whole are now being re-edited by Arlo Griffiths and his colleagues, whose 2012 catalogue of the inscriptions in the Museum of Cham Sculpture at Đà Nẵng, published in English and Vietnamese, is the first modern instalment 21.

The firm ground begins around 400, in a narrow valley some 40 kilometres south-west of the later port of Hội An. There a king who called himself Bhadravarman founded a temple to Śiva in the form of a liṅga, and named the god after himself: Bhadreśvara, "the lord of Bhadra[varman]". Georges Maspero, the first historian of Champa, identified Bhadravarman with the Fan Huda whom Chinese sources place on the throne of Linyi from 380 to 413; R. C. Majumdar, who published the Cham inscriptions with English translations in 1927, called the identification "probable but not certain" 14. The king's own stele records the gift that founded the sanctuary:

"a perpetual endowment has been given to Bhadresvara by our king Bhadravarman, who is devoted to the feet of Bhadresvarasvami; To wit, the land within (the boundaries viz,) Sulaha mountain in the east, the great mountain in the south, the Kucaka mountain in the west and the Great River in the north, together with its inhabitants, has been given … He who does not maintain but destroys, will himself be destroyed. Bhadresvarasvami knows." 1

Three things in this text set the pattern for the next eight centuries. The god carries the king's name. The god owns land, defined by the mountains that ring the valley. And the land is given "together with its inhabitants": the people who lived there passed to the god with the soil. Majumdar notes that the inscription calls the king "versed in the four Vedas" 1. Whether Bhadravarman read Vedic texts or simply claimed the learning is beside the point; he wanted to be described in the categories of an Indian king.

Śambhuvarman rebuilds

The valley, which the Vietnamese would later call Mỹ Sơn, "beautiful mountain", became what Majumdar calls "the national sanctuary of the Chams" 1. Its first temple was of wood, and it burned. A later inscription records that under King Rudravarman the temple of Bhadreśvara was destroyed by fire, in a Śaka year that falls between 479 and 577 CE 1. Rudravarman's son Śambhuvarman rebuilt it and renamed the god Śambhu-Bhadreśvara, "thus adding his own name to that of the original founder" 1. Śambhuvarman also sent embassies to the Chinese court in 568, 572 and 595 1. The pattern was now fixed: each king who restored or enlarged the sanctuary joined his name to the god's, so that the dynastic Śiva of Mỹ Sơn carried the memory of the kings who served him.

In the seventh century the cult reached its first full development. Prakāśadharma, who reigned as Vikrāntavarman I from about 656 to at least 687, is the author of at least eight inscriptions 1. In 657 he installed a new god, Prabhāseśvara, and endowed Īśāneśvara, Śambhu-Bhadreśvara and Prabhāseśvara together; he also built a temple to Viṣṇu and one to Kubera, the god of wealth 1. One of his inscriptions at Mỹ Sơn traces the royal line back to an Indian brahmin. It tells how Kauṇḍinya, "the foremost among Brahmanas", planted a spear he had received from Aśvatthāman, son of Droṇa, the hero of the Mahābhārata, and married "a daughter of the king of serpents, called Soma, who founded a family in this world" 1. The legend is set at Bhavapura in Cambodia, and it joins a Sanskrit epic to a serpent-princess of the land: Indian learning married to local sovereignty, which is the story the Cham kings wanted told.

Who carried it

No inscription describes an Indian conquest of the Cham coast, and none ever happened. The carriers were the people of the sea route: merchants, brahmins who travelled with them, Buddhist monks bound for China, and above all the Cham rulers who chose the Indian idiom. The historian O. W. Wolters gave this process its modern name, "localization": Southeast Asian elites selected Indian materials and refashioned them for their own purposes, rather than receiving a culture handed down from India 11. Coedès's older model, set out in The Indianized States of Southeast Asia, had imagined an Indian civilisation expanding across the Bay of Bengal 10. Mus had already, in 1933, described something closer to Wolters: an Indian religion grafted onto a local one, which "la correspondance des deux versants de l'Asie des Moussons" made possible 2.

The participants who can be named are all Cham or claimed by Cham kings:

Name Date Act recorded
Bhadravarman (Fan Huda?) c. 380–413 Founds the liṅga Bhadreśvara at Mỹ Sơn and endows it with land and people 1
Rudravarman 6th century Reign in which the first temple burns 1
Śambhuvarman late 6th century Rebuilds; renames the god Śambhu-Bhadreśvara; embassies to China 568–595 1
Prakāśadharma-Vikrāntavarman I c. 656–687 Eight inscriptions; new gods in 657; Kauṇḍinya genealogy 1
Satyavarman 784 Re-installs the Po Nagar god after a sea raid 1
Indravarman II 875 Founds the Buddhist monastery of Lakṣmīndra-Lokeśvara at Đồng Dương 1
Parameśvaravarman 1050 Sanskrit hymn to the goddess of Po Nagar 2

The goddess of Kauthāra

The second great sanctuary of the coast stands on a hill above the mouth of the Cái river at Nha Trang, in the Cham region the inscriptions call Kauthāra. Its deity was a goddess. The inscriptions call her Bhagavatī, "the Goddess of Kauthara", and describe her "body sparkling in beauty on account of an excellent coating of gold" 1. Mus identified her with Umā, Śiva's consort, while seeing in her an older earth-mother of the land: Bhagavatī Kauthāreśvarī 2. The sanctuary, known today as Po Nagar, is still a place of worship 6.

The sanctuary's history shows how exposed these temples were to the sea that had brought their gods. In 774, according to a stele of King Satyavarman, raiders "came in ships, took away the Mukhalinga of the God, and set fire to the abode of the God"; the inscription calls them "ferocious, pitiless, dark-coloured people of other cities, whose food was more horrible than that of the Vampires" 1. Satyavarman re-installed the god in 784. In 787 a second raid burned the temple of Bhadrādhipatīśvara further south, in the Phan Rang region, which Indravarman I restored in 799 1. The raiders are usually identified as Javanese. Around 950, as a separate record in this atlas describes, a Khmer army carried off the goddess's gold image again.

By 1050 a Cham king could address her in Sanskrit verse of real accomplishment. Mus, reading Parameśvaravarman's hymn, wrote: "Voilà comment un Cham pensait et comment il s'exprimait en sanskrit en 1050 A. D." 2.

The Buddhist interval at Indrapura

Śiva did not hold the coast alone. In 875 Jaya Indravarman II, ruling from a new northern capital he called Indrapura, at the site now named Đồng Dương, installed an image of the bodhisattva Avalokiteśvara and called it after himself, Lakṣmīndra-Lokeśvara 1. Beside it he founded a monastery "for the perpetual enjoyment of the Bhikṣu-saṃgha" 1. The Đồng Dương complex was the largest Buddhist foundation in Champa, and its sculpture is among the finest of the period. The king's grant shows that the Buddhist foundation worked on the same terms as the Śaiva ones: "the king Sri Indravarma has given these fields together with their corns, male and female slaves … to Lokesvara, for the enjoyment of the community of monks and for the sake of the propagation of Dharma" 1. The ICOMOS evaluation notes that from the beginning of the tenth century Buddhism waned and the influence of Mỹ Sơn, "where Hinduism had always been strong", returned; most of the finest towers in the valley date from that century 8.

Around 1000: a capital moves south, or does not

By the conventional account, the Indic kingdom of Champa reached a hinge around the year 1000. In 982 a Vietnamese army sacked the northern capital (described in the cost section below), and in the following two decades the Cham court moved its centre south, to Vijaya in the plain of modern Quy Nhơn. Maspero built this account on Chinese and Vietnamese chronicle notices of 989, 1000 and 1007 34. Louis Finot, reviewing Maspero's book in 1928, already judged that "le livre n'est pas entièrement up to date" 27.

Vickery rejects the move. The "Buddha city" of the Vietnamese chronicles, he argues, is most probably Đồng Dương itself, and "it is easy to see that there was no move of Cham kings or capital south of the Indrapura - Huế region" 3. He follows William Southworth, whose doctoral thesis on the origins of Champa insisted, in Vickery's words, that it "was neither a unitary polity nor even a federation", and who notes that the inscription C101, of about 1153–84, is "the earliest surviving reference to Vijaya", and that the name's "application to earlier periods in the academic literature should therefore be considered an historical anachronism" 326. The debate is summarised in the collected volume on the Cham that Bruce Lockhart and Trần Kỳ Phương edited in 2011, which reprints Vickery's paper alongside new work on the archaeology and art of the coast 325. This record takes the year 1000 as a convenient midpoint of Champa's Indic high period, not as the date of a proven event.

What changed and what was replaced

The god-soil becomes a god-king's estate

The first change the Indian religion made was in the scale of the god. Mus's soil-god belonged to one place and one community. The Bhadreśvara of Mỹ Sơn belonged to a king, carried his name, received grants of land with defined boundaries and the people on them, and was served by priests who could read the texts that made such grants binding. The gods of place were not abolished: Mus's argument is precisely that the liṅga was understood as the heir of the stone 2. But they were drawn into a hierarchy with a royal god at its head.

Majumdar's translations show the new vocabulary the Cham elite acquired 1:

  • Kingship in Sanskrit: royal names ending in -varman ("protected by"), titles of mahārāja, and genealogies running back to Kauṇḍinya and the Mahābhārata.
  • Gods as owners: land grants to named deities, with boundaries and curses on whoever disturbs them.
  • The royal-named liṅga: Bhadreśvara, Śambhu-Bhadreśvara, Prabhāseśvara, Indrabhadreśvara, each a Śiva bearing a ruler's name.
  • Temple dependants: inhabitants and "male and female slaves" given to gods and monasteries.
  • Śaka-era dating: events dated by the Indian Śaka calendar, such as the raid of Śaka 696 (774 CE).
  • Sanskrit as a literary language: hymns and eulogies composed by Cham authors, alongside inscriptions in Old Cham.
A square stone pedestal carved with a frieze of small figures around its base, topped by a lotus moulding, a broad flat basin with a broken spout, and a short cylindrical stone pillar, displayed in a glass case.
Sandstone altar from Trà Kiệu, seventh to eighth century, now in the Museum of Cham Sculpture, Đà Nẵng: a liṅga rising from a yoni basin on a pedestal carved with a frieze of small figures. The liṅga of Śiva, set at the centre of every royal temple, was in Paul Mus's phrase the heir of the older Cham soil-god's stone.
Daderot. Altar with liṅga from Trà Kiệu (Altar 22.2), 7th–8th century. Museum of Cham Sculpture, Đà Nẵng. CC0 via Wikimedia Commons. · CC0

Towers on the Indian plan

The second change was in building. The ICOMOS evaluation says plainly that "the predominant style of the architecture and sculptural decoration of the My Son temples derives directly from India" 8. Each temple complex centred on a main tower, the kalan, standing for Mount Meru; the interior held a liṅga on a plinth; the tower rose in three storeys to represent Kailāsa, Śiva's mountain; a gate-tower (gopura) and a long hall (maṇḍapa) stood in front 8. Henri Parmentier, the French architect who surveyed the valley at the start of the twentieth century and published the first inventory in 1904, recorded nearly seventy monuments and divided them into the groups A to H that are still used 819. A1, the main tower of Group A, was the highest in the valley at 24 metres 8. The parts of a Cham temple complex, as the evaluation describes them, map the Indian cosmos onto Cham brick 8:

Element Form Meaning
Kalan Main brick tower with false doors and a sanctuary chamber Mount Meru, the centre of the universe
Base (bhurloka) Square or rectangular platform in brick or stone, carved with reliefs The human world
Tower body (bhuvakola) Brick shaft with applied columns, a door facing east, the liṅga on its plinth inside The spirit world
Roof (suarloka) Three storeys rising as a stepped pyramid, once often covered in gold or silver leaf Kailāsa, Śiva's mountain
Gopura Smaller gate-tower in brick with stone pillars The threshold of the sacred precinct
Maṇḍapa Long tiled hall in front of the gate-tower Space for the ceremonies of the cult

Andrew Hardy and his Italian and Vietnamese colleagues, whose 2009 volume collects the modern archaeology of the site, describe the starting point: "The kings of ancient Champa … started building sacred temples in a circular valley more than 1500 years ago" 7. The ICOMOS evaluators were struck less by the scale than by the persistence: "Of especial importance is the association with the Hindu religion, imported from India, which stubbornly persisted here in the face of Buddhist competition for many centuries" 8.

Indic customs at court

The court culture that grew around the temples absorbed Indian practices that left traces even in the records of Champa's enemies. When a Vietnamese army entered the Cham capital in 1044, it carried off the king's wives and "palace women skilled in dancing to Western Heaven melodies" (宫女之善舞西天曲調者), Western Heaven being the Chinese and Vietnamese name for India 9. The sack of 982 had already taken a hundred palace women and "one monk from Tianzhu", from India 9.

The most striking Indic custom is recorded in 1307. When the Cham king Jaya Siṃhavarman III, known in Vietnamese as Chế Mân, died a year after marrying the Vietnamese princess Huyền Trân, the Vietnamese annals explain the court's alarm: "占城俗國王卒,主后入火壇以殉", "by Cham custom, when the king dies, the queen enters the fire-altar to follow him in death" 9. The Trần court sent an envoy on the pretext of condolences, and he snatched the princess away by boat 9. It is an outsider's statement, without Cham corroboration, but it describes a widow-immolation practice of the kind Indian courts also knew.

What the coast kept

The Indian religion never displaced the older one completely. Mus's Cham informants worshipped liṅgas, Indian statues and plain stones as one category of divine bodies 2. The goddess of Po Nagar remained the mother of the country. And when Islam reached the coast through the Malay world, it was layered onto the same structure rather than replacing it.

Today the descendants of the Indic religion are the Cham of Ninh Thuận and Bình Thuận provinces, the heartland of the last Cham principality, Pāṇḍuraṅga. Trần Kỳ Phương and Rie Nakamura describe two communities 6:

  • The Cham Bà-la-môn ("Brahmanist"), or Ahier, "adherents of an indigenized form of Hinduism. They worship the god called Pô Yang and they deify kings in the old temples … They are supposed to observe a taboo on beef. Usually they are cremated when they die" 6.
  • The Cham Bani, or Awal, "adherents of an indigenized form of Islam. They worship Pô Alwah (Allah) at their mosques called thang muki" 6.
  • The Sunni Cham of the Mekong delta, An Giang and Hồ Chí Minh City, who follow a reformed Islam linked to the Malay world 24.

The two Ninh Thuận communities understand themselves as halves of one society: "Ahier means 'back, behind, or after' and Awar means 'front or before' … society must have Ahier and Awar to function" 6. The Vietnamese scholar Đổng Thành Danh describes the same "cấu trúc nhị nguyên", a dual structure, as the frame of Cham religion: "Awal/Bàni - Ahiér/ Bàlamôn" 20. Vietnam's 2019 census counted 178,948 Cham 24. They are among the very few peoples outside South Asia and Bali who still keep a Hindu priesthood of their own.

What the cost was

A neighbour that had freed itself

The cost of the Indic religion's transmission, in the strict sense, was paid inside Cham society: land and people given to the gods, slaves assigned to monasteries, a custom of widow-burning reported at court. The larger bill came from another direction. The Indian religion became the identity of the Cham polities, and those polities were dismantled over eight centuries by the Vietnamese state to the north. The Vietnamese did not attack Champa because it worshipped Śiva. But the destruction of Champa was also the destruction of the institutions that sustained its religion, and the Vietnamese annals record the temples among the targets.

The Vietnamese kingdom of the Red River delta had itself emerged from a thousand years of Chinese rule, which a separate record in this atlas follows, and had won its independence in 939 17. K. W. Taylor has warned against reading the following centuries as a single national "march to the south": "If we can clear our minds of 'Vietnameseness' as the object of our knowledge and instead look carefully at what the peoples we call Vietnamese were doing at particular times and places", he writes, we find "quite different kinds of peoples" grounded in "particular times and terrains" 16. Momoki Shirō's study of the Lý and Trần state (2011) places its wars inside the competition for the southern-sea trade of the tenth to fifteenth centuries 18. The wars ran both ways, and Cham kings raided the delta as well as suffering its raids.

The annals' record of war

The principal source for the wars is the Đại Việt sử ký toàn thư, the Complete Annals of Đại Việt, compiled by Ngô Sĩ Liên and completed in 1479 9. Its figures are the victors' own, and they should be read as such. Ngô Sĩ Liên finished his work at the court of Lê Thánh Tông eight years after that emperor had taken the Cham capital, and its account of the Cham wars is told throughout from the conqueror's side. They are not a census, and no Cham source survives to set against them; the Cham inscriptions record the wars, when they record them at all, as the victories and restorations of Cham kings 19. With those cautions, the annals' entries are summarised here:

Year Event in the annals Recorded toll
982 Lê Hoàn sacks the northern Cham capital; the Cham ruler beheaded in battle Captives "beyond counting"; 100 palace women and an Indian monk taken; walls razed; "its ancestral temples destroyed" (毀其宗廟) 9
1044 Lý Thái Tông's campaign; the Cham king's head presented by his own general 30,000 heads taken; over 5,000 captured alive; "corpses choked the fields" 9
1069 Lý Thánh Tông captures the Cham king Chế Củ (Rudravarman III) Captured "with 50,000 of his people"; three districts ceded as ransom 9
1306–1307 Princess Huyền Trân married to Chế Mân; the districts of Ô and Lý annexed Ô and Lý renamed Thuận and Hóa 9
1402 Hồ Quý Ly forces the cession of Chiêm Động and Cổ Lũy Territory divided into four new districts 9
1471 Lê Thánh Tông takes Vijaya Over 30,000 captured; over 40,000 beheaded; the king Trà Toàn taken alive 9

The 982 entry is the plainest evidence that religious sites were deliberately destroyed: "夷其城池,毀其宗廟", the Vietnamese army "levelled its walls and moats, and destroyed its ancestral temples" 9. In 1044, after 30,000 heads had been taken in the pursuit, "blood smeared the blades, corpses choked the fields" (血塗兵刀,屍塞原野) 9. The emperor, Lý Thái Tông, then issued an order that anyone who wantonly killed Cham people would be executed without pardon 9. The Cham queen, Mỵ Ê, summoned to the emperor's boat, wrapped herself in a felt blanket and drowned herself in the river; the annals record that the emperor honoured her chastity 9. Cham captives of these wars were settled in the delta in villages that kept "the old Cham names" 9.

Champa strikes back

The Cham side of the war produced the most formidable king in the coast's history. Chế Bồng Nga, who ruled from about 1360 to 1390, and whom John Whitmore calls "the last great king of classical Southeast Asia", carried the war to the Vietnamese capital 23. The annals record his campaigns from the losing side 9:

  • 1371: the Cham fleet enters by the Đại An estuary, strikes Thăng Long directly, "burned the palaces and carried off women and silks" 9.
  • 1377: a Vietnamese counter-invasion of Vijaya ends in disaster; the emperor Trần Duệ Tông is killed in battle 9.
  • 1378: the Cham, finding the coast guarded, enter by another estuary and strike the capital again 9.
  • 1390: a disgraced Cham officer deserts and points out the king's green-lacquered boat; the Vietnamese commander Trần Khát Chân "ordered the firearms to fire together", and Chế Bồng Nga is pierced through the planking and killed 9.

With him went Champa's last period of military parity.

1471: the fall of Vijaya

The decisive campaign came eighty years later. Lê Thánh Tông, the most powerful ruler of the Lê dynasty, besieged Vijaya in the spring of 1471. The annals give the result: "闍槃城陷、俘獲三萬餘人、斬首四萬餘級、生擒茶全", "the city of Vijaya fell; over 30,000 people were captured, over 40,000 heads were taken, and Trà Toàn was taken alive" 9. The captured king died of grief on the march; his head was cut off, his body burned and thrown into the river 9. The conquered land became the Vietnamese province of Quảng Nam, and what remained of Champa to the south was divided among three vassal rulers 917.

The figures have drifted in modern retellings. Some secondary sources give "over 60,000 killed" and treat the 40,000 as a separate massacre of the city's inhabitants. The annals do not say that: they give 40,000 heads taken and 30,000 captives, and they are the victor's own count 9. What is not in doubt is the result. The ICOMOS evaluation dates the end of worship at Mỹ Sơn to the later fifteenth century 8. The national sanctuary that Bhadravarman had founded a thousand years earlier lay inside a Vietnamese province, and its god had no king to serve.

The religion moved south with the surviving Cham courts. The temples that the Cham Bà-la-môn of Ninh Thuận and Bình Thuận still use were built, Trần Kỳ Phương and Nakamura note, "between the 13th/14th and the 16th/17th centuries" in the southern region, and in them the community deifies its kings 6. The cult that had begun as a king's Śiva at Mỹ Sơn ended as the cult of the kings themselves, kept in the last province of the last principality. The northern sanctuaries were left to the forest. When Parmentier's survey teams reached Mỹ Sơn at the end of the nineteenth century, the valley had been empty of worshippers for four hundred years 819.

A stone liṅga on a stacked square pedestal standing on a carved platform, seen between mossy brick walls, with a ragged, broken brick wall and trees behind it.
A liṅga on its pedestal among the ruins of Mỹ Sơn Group A. The group's main tower, A1, was at 24 metres the tallest in the valley; the sanctuary was heavily bombed in August 1969, five centuries after worship here had ceased with the Vietnamese conquest of 1471.
Ms Sarah Welch. Liṅga on pedestal amid ruins, Mỹ Sơn Group A (Champa), photographed 2023. In situ, Mỹ Sơn, Quảng Nam, Vietnam. CC0 via Wikimedia Commons. · CC0

Pāṇḍuraṅga, 1692–1835

The southern Cham principality of Pāṇḍuraṅga, around Phan Rang and Phan Rí, survived for three and a half more centuries under pressure from the Nguyễn lords, who ruled the southern Vietnamese territories from the late sixteenth century. Li Tana's study of Nguyễn Cochinchina shows that the advance was religious as well as military: "the Nguyen rulers needed to provide an alternative to Cham beliefs that would help sustain Vietnamese immigrants spiritually and psychologically" 22. The last stages, in the sources:

  • 1692–1693: the Cham king Po Saut attacked the Nguyễn border and was defeated in early 1693 15. Nicolas Weber records that he "was deported, and died in jail less than a year after his capture" 14. "The Cham court was renamed Thuan Thanh Tran and occupied by Nguyen garrisons" 15.
  • 1692–1693: "The Khmer Royal Chronicles record that in 1692–3, 5,000 Cam families led by a 'member of the royal family' … requested to settle in Cambodia" 14.
  • 1694: the Nguyễn restored a Cham king, Po Saktiraydaputih, as a tributary "native king" after popular revolts 1415.
  • 1832: Emperor Minh Mạng annexed the principality. "The last semi-independent principality of Pāṇḍuraṅga disappeared in 1832" 14.
  • 1833–1835: two risings, the Islamic movement of Katip Sumat (1833–34) and the insurrection of Ja Thak Wa (1834–35), both suppressed; Po Dharma's two-volume study of these years remains the reference 12.

Danny Wong Tze Ken, of the University of Malaya, explains the 1832 rising as "a reaction against Emperor Ming Mang's harsh oppression of the Chams in reprisal for their support of Ming Mang's viceroys in Gia Dinh in the south"; Katip Sumat arrived in 1833 "accompanied by a large force of Malays and Chams from Kelantan" 15. Weber, working from Cham poems of the period, treats 1832–35 as "the disappearance of the kingdom of Campā — or more precisely its last independent principality of Pāṇḍuraṅga — and its integration into the Vietnamese realm" 13. His survey of Cham narrative writing ends on a sentence that is the shortest history of the people: "All Cam narratives commemorate an exile, a displacement or a migration" 14.

August 1969

The sanctuary's last destruction came from the air. After 1965 the Mỹ Sơn valley became a guerrilla base, and "the whole region was the target for American bombing and minelaying" 8. The ICOMOS evaluation records the result in one sentence: "In August 1969 the sanctuary itself was heavily bombed, causing great destruction" 8. Of temple B1, the largest at Mỹ Sơn, the evaluators wrote that it "has, however, been largely destroyed by bombing" 8. From 1980 a Polish conservation team worked with Vietnamese specialists to document and consolidate what remained, and the site was inscribed on the World Heritage List in 1999 78.

Weighing the bill

Cost When Evidence
Land "together with its inhabitants" given to the dynastic god c. 400 Bhadravarman's stele 1
Male and female slaves assigned to the Buddhist foundation 875 Đồng Dương grant 1
Temples burned and images seized by sea raiders 774, 787 Po Nagar and Phan Rang steles 1
A custom of royal widow-immolation reported at court 1307 Vietnamese annals 9
Ancestral temples destroyed; captives "beyond counting" 982 Vietnamese annals 9
30,000 killed, 5,000 captured 1044 Vietnamese annals 9
40,000 killed, 30,000 captured; end of worship at Mỹ Sơn 1471 Vietnamese annals; ICOMOS 89
5,000 families flee to Cambodia 1692–1693 Khmer chronicles via Weber 14
Annexation of Pāṇḍuraṅga; two risings suppressed 1832–1835 Po Dharma; Weber; Wong 121315
Mỹ Sơn heavily bombed August 1969 ICOMOS 8

The bill falls into two parts. The first was paid inside Cham society and belongs to the transmission itself: the gods' estates absorbed land and the people on it, monasteries received slaves, and a court custom made a queen's life depend on her husband's death. These are the ordinary costs of the Indian royal religion wherever it took root in Southeast Asia, and they are smaller than the comparable bill at Angkor. The second part was paid by the Cham as a people, and it was paid for something else: the expansion of a neighbouring state. That expansion destroyed the temples in 982, killed tens of thousands in 1044 and 1471, ended royal worship at Mỹ Sơn, and finally annexed the last principality in 1832.

What survived is not nothing. In the late towers of Pāṇḍuraṅga, the Cham Bà-la-môn still serve the god they call Pô Yang and the deified kings of the coast 6. The religion that a Cham king joined to his own name around 400 is now the faith of a minority within a minority, kept by people who know that their history is a history of displacement.

What followed

Where this lives today

The Cham Bà-la-môn (Ahier), an indigenised Hindu community of Ninh Thuận and Bình Thuận The Cham Bani (Awal), whose Islam is paired with Ahier in one dual society Mỹ Sơn Sanctuary, a UNESCO World Heritage Site since 1999 The Po Nagar towers at Nha Trang, still a place of worship The Museum of Cham Sculpture in Đà Nẵng Cham communities of Cambodia, descended from refugees of the Vietnamese advance

References

  1. Majumdar, R. C. Ancient Indian Colonies in the Far East, Vol. I: Champā. Greater India Society Publication No. 1. Lahore: Punjab Sanskrit Book Depot, 1927. Book I (history) and Book III (Sanskrit inscriptions with English translations). en primary
  2. Mus, Paul. « Cultes indiens et indigènes au Champa ». Bulletin de l'École française d'Extrême-Orient 33 (1933): 367–410. fr
  3. Vickery, Michael. "Champa Revised." ARI Working Paper Series No. 37. Singapore: Asia Research Institute, National University of Singapore, 2005. Reprinted in Lockhart and Trần Kỳ Phương, eds., The Cham of Vietnam (2011), 363–420. en
  4. Maspero, Georges. Le royaume de Champa. Paris and Brussels: G. van Oest, 1928. fr
  5. Sircar, D. C. "Date of the Earliest Sanskrit Inscription of Champa." Journal of the Greater India Society 6, no. 2 (1939): 53–55. en
  6. Trần Kỳ Phương and Rie Nakamura. "The Mỹ Sơn and Pô Nagar Nha Trang Sanctuaries: On the Cosmological Dualist Cult of the Champa Kingdom in Central Vietnam as Seen from Art and Anthropology." ARI Working Paper Series No. 100. Singapore: Asia Research Institute, National University of Singapore, 2008. en
  7. Hardy, Andrew, Mauro Cucarzi and Patrizia Zolese, eds. Champa and the Archaeology of Mỹ Sơn (Vietnam). Singapore: NUS Press, 2009. en
  8. ICOMOS. Advisory Body Evaluation: My Son (Vietnam), No. 949. Paris: International Council on Monuments and Sites, 1999. en primary
  9. Ngô Sĩ Liên et al. 大越史記全書 (Đại Việt sử ký toàn thư, Complete Annals of Đại Việt). Completed 1479; Chính Hòa edition, 1697. Bản kỷ, quyển 1–3, 6–8, 12. Classical Chinese text, Chinese Wikisource. lzh primary
  10. Cœdès, George. The Indianized States of Southeast Asia. Edited by Walter F. Vella, translated by Susan Brown Cowing. Honolulu: East-West Center Press, 1968. en
  11. Wolters, O. W. History, Culture, and Region in Southeast Asian Perspectives. Revised edition. Ithaca: Cornell Southeast Asia Program; Singapore: ISEAS, 1999 (first edition 1982). en
  12. Po Dharma. Le Pāṇḍuraṅga (Campā) 1802–1835. Ses rapports avec le Vietnam. 2 vols. Publications de l'École française d'Extrême-Orient 149. Paris: EFEO, 1987. fr
  13. Weber, Nicolas. "The Destruction and Assimilation of Campā (1832–35) as Seen from Cam Sources." Journal of Southeast Asian Studies 43, no. 1 (2012): 158–180. en
  14. Weber, Nicolas. Exploring Cam Narrative Sources for History of the Cam Diaspora of Cambodia. Nalanda-Sriwijaya Centre Working Paper No. 17. Singapore: ISEAS, 2015. en
  15. Wong Tze Ken, Danny. "Vietnam-Champa Relations and the Malay-Islam Regional Network in the 17th–19th Centuries." Kyoto Review of Southeast Asia 5 (March 2004). en
  16. Taylor, K. W. "Surface Orientations in Vietnam: Beyond Histories of Nation and Region." Journal of Asian Studies 57, no. 4 (1998): 949–978. en
  17. Taylor, K. W. A History of the Vietnamese. Cambridge: Cambridge University Press, 2013. en
  18. 桃木至朗 (Momoki Shirō).『中世大越国家の成立と変容』(The Formation and Transformation of the Medieval Đại Việt State). 大阪: 大阪大学出版会, 2011. ja
  19. Parmentier, Henri. « Les monuments du cirque de Mi-so'n ». Bulletin de l'École française d'Extrême-Orient 4 (1904): 805–896. fr primary
  20. Đổng Thành Danh. "Awal – Ahiér: Một cách tiếp cận về cấu trúc tín ngưỡng, tôn giáo của người Chăm." Tạp chí Nghiên cứu Tôn giáo 4, no. 196 (2020). vi
  21. Griffiths, Arlo, Amandine Lepoutre, William A. Southworth and Thành Phần. The Inscriptions of Campā at the Museum of Cham Sculpture in Đà Nẵng / Văn khắc Chămpa tại Bảo tàng Điêu khắc Chăm Đà Nẵng. Hà Nội: EFEO; Hồ Chí Minh City: VNUHCM Publishing House, 2012. en primary
  22. Li Tana. Nguyễn Cochinchina: Southern Vietnam in the Seventeenth and Eighteenth Centuries. Ithaca: Cornell Southeast Asia Program, 1998, p. 103; quoted in Shawn McHale's review, H-Asia (H-Net), September 1998. en
  23. Whitmore, John K. "The Last Great King of Classical Southeast Asia: 'Chế Bồng Nga' and Fourteenth-century Champa." In The Cham of Vietnam: History, Society and Art, edited by Trần Kỳ Phương and Bruce M. Lockhart, 168–203. Singapore: NUS Press, 2011. en
  24. Nhân Dân (édition française). « Les Chăm ». Hanoi: Nhân Dân, 2020 (data of the 1 April 2019 Population and Housing Census). fr
  25. Trần Kỳ Phương and Bruce M. Lockhart, eds. The Cham of Vietnam: History, Society and Art. Singapore: NUS Press, 2011. en
  26. Southworth, William Aelred. "The Origins of Campā in Central Vietnam: A Preliminary Review." PhD thesis, School of Oriental and African Studies, University of London, 2001. en
  27. Finot, Louis. Review of Georges Maspero, Le royaume de Champa. Bulletin de l'École française d'Extrême-Orient 28 (1928): 285–292. fr

Further reading

Cite this article
OsakaWire Atlas. 2026. "Śiva on the Cham coast (c. 1000) — and the march that ended Champa" [Hidden Threads record]. https://osakawire.com/en/atlas/champa_hinduism_to_vietnam_1000/