Article

Sri Vijaya

From Munshipedia, the MBRAS digital historical encyclopedia

Sri Vijaya

Sri Vijaya was a maritime polity centred on the Musi river basin in southern Sumatra, active from the seventh to the thirteenth centuries, whose commercial and cultural reach extended across the Malay Peninsula, the Strait of Malacca, and into the broader Indian Ocean trading network. The polity is attested in Chinese, Arab, and indigenous sources, and its inscriptions—carved in Old Malay on stone—constitute the earliest surviving physical evidence of the decimal zero in its place-holding function, predating comparable Indian material by nearly two centuries [1, pp. 53–54]. Sri Vijaya’s political geography, however, remains one of the most contested in Southeast Asian history: its fame in foreign records stands in stark contrast to the near-total absence of indigenous monumental remains, a discrepancy that has driven successive waves of revisionist scholarship from the 1930s to the present [2, pp. 1–3].

Scope and Historical Definition

The identification of Sri Vijaya as a historical entity rests on a chain of philological and geographical reasoning that began in the early twentieth century. George Coedès, in 1918, identified the Arab toponym Sribuza and the T’ang Chinese rendering Shih-li-fo-shih as referring to the same polity, tentatively locating it at Palembang on the strength of Groeneveldt’s reading of fourteenth-century Ming sources [2, pp. 3–4]. That reading, however, was based on a misinterpretation: the Ming texts stated that after a Javanese attack c. 1377 the king of San-fo-ch’i resided at Jambi and that Palembang had become “the Old Harbour,” a distinction that Coedès and the French sinologist Pelliot both overlooked, leading them to conflate a later administrative arrangement with the seventh-century capital [2, pp. 3–6].

The location question was further complicated by a series of revisionist challenges. Majumdar, in 1933, proposed that the maritime empire was Peninsular rather than Sumatran; Moens, in 1937, attempted to relocate the seventh-century capital first to Kelantan and then to Muara Takus, on the basis of a misreading of the Chinese pilgrim I-ching’s sundial passage [2, pp. 4–5]. The 1974 joint Indonesian-American archaeological expedition to Palembang delivered what was, at the time, a devastating verdict: “the entire vicinity of Palembang does not contain enough pre-14th century domestic artifacts to make one small village,” and its members even suggested that the inscriptions and statues found there had been redeposited in later periods [2, p. 14].

The turning point came with Professor Boechari’s re-examination of the Kedukan Bukit inscription of 682, from which he recovered the toponym Upang—a village and river approximately twenty-five miles downstream from Palembang, halfway between the city and the Musi estuary—where Coedès had read mata jap and Krom had read Malayu [2, p. 15]. This finding re-established the Palembang area within seventh-century Sri Vijayan geography and anchored subsequent reconstruction efforts to a specific riverine landscape rather than to a generic “capital city.”

Political Culture and the Riverine Environment

O.W. Wolters, writing in 1979 on the eve of a multi-disciplinary research workshop in Jakarta, proposed that Sri Vijaya’s persistent elusiveness should be treated not as a deficiency in the evidence but as a meaningful sign in itself [2, pp. 1–2]. His “semiological” approach argues that the absence of land-grant inscriptions, the absence of large-scale monumental religion, and the absence of a walled city are all predictable consequences of a riverine environment with limited wet-rice agriculture and no settled villages available for temple endowment [2, pp. 16–17]. In such a landscape, royal presence was signified not by immovable monuments—as in central Java—but by mobile messengers in distinguished boats, and spatial integration was achieved through the responsive obedience of dispersed settlements to royal summons [2, pp. 17–18].

Wolters identifies the paddle as the key invisible sign of Sri Vijayan political culture. In the nineteenth century, local craft could cover the fifty miles from Palembang to the Musi estuary in a few hours on an ebbing tide, and goods from Palembang reached the Pasemah highlands in fourteen days by paddle over some three hundred miles to Muara Malang [2, p. 17]. The royal compound at the Sri Vijayan centre would have resembled a Malay kota—a settlement defended by earth, felled trees, and bamboo, relying on the river for reinforcement or evacuation—rendering permanent stone or brick fortifications superfluous in a paddle-based environment [2, pp. 19–20]. Bricks, in this context, were reserved for the gods in the form of stupas, not for the defence of men [2, p. 20].

This model is supported by multiple textual witnesses: the Kedukan Bukit inscription records a royal expedition; the Sabokingking inscription attests to the mobilization of forces; Abu Zayd’s account describes a thousand-ship assembly; and the Chinese merchant Chao Ju-kua records wartime gatherings [2, pp. 17–18]. Wolters further reconstructs I-ching’s list of Indonesian dvipas (island realms), identifying Mo-ho-hsin as Mukha Asin (Briny Surface), a zone of salt marshes and shifting tidal creeks south of the Musi estuary, and argues that the Palembang area belonged to this inhospitable coastal zone rather than to Malayu-dvipa proper [2, pp. 30–32]. This opens the possibility that three centres—Sri Vijaya, Malayu, and Palembang—were involved in the events of the seventh century, and that the Kedukan Bukit expedition of 682 represented Sri Vijaya’s expansion from Malayu-dvipa southward into the Musi basin [2, pp. 30–32]. Wolters also suggests that the Chinese toponym San-fo-ch’i of the Sung period means simply “the Three Vijayas,” reflecting a ruler who governed three neighbouring Sumatran states [2, pp. 23–24].

Diplomatic Relations and External Testimony

The earliest Chinese evidence for a polity identifiable as Sri Vijaya comes not from the T’ang period but from the Southern Sung court of the fifth century. William T. Kao, in 1956, demonstrated that the Yuan-chia Ch’i-chu-chu, a contemporaneous court diary covering the reign of 424–453, records five embassies from a polity called Ho-lo-tan to the Southern Sung capital [3, pp. 163–167]. The diary itself no longer survives but is preserved in fragments quoted in the T’ai-p’ing Yu-lan (compiled 977–983) [3, p. 163]. Kao showed that the standard official histories—the Sung Shu and the Nan Shi—contain internal inconsistencies in their treatment of these embassies, omitting the third entirely, misdating the fourth, and failing to identify the usurper who deposed King P’i-sa-pat-maw [3, pp. 163–166].

By turning to the Yuan-chia fragments, Kao recovered the name of the third embassy’s sender, Shi-li-p’i-jia-ya, and reconstructed the political sequence: the usurper seized the throne between the departure and return of the second envoy (c. 430–434), consolidated his position, and was eventually recognised by the Southern Sung emperor in 449 through an imperial edict conferring a title [3, pp. 166–167]. Kao argues that this recognition, coming fifteen years after the coup and coinciding with tribute from two other polities (P’an-huang and P’an-tat), implies that Ho-lo-tan had by then achieved sufficient regional power—possibly through control of Kalasapura and the China sea route—to warrant imperial honour [3, pp. 167].

Kao’s central philological intervention is his demonstration that Shi-li-p’i-jia-ya, read in the Amoy dialect, is a more consonant transliteration of “Sri Vijaya” than the later T’ang-era form Che-li-fo-che [3, p. 170]. He supports Moens’s hypothesis that the usurper’s personal name became an honorific for the kingdom, a plausible political development in which a dynastic name absorbs the name of the state [3, p. 170]. Kao further challenges the conventional dating of Che-li-fo-che embassies to 670–741, arguing that the later compilation Suh T’ung-tien (1767) extends the range to the K’ai-ch’eng period (836–840) on the basis of more careful citation practices that include the reigning emperor’s name alongside the year-style, reducing the likelihood of clerical confusion between K’ai-yuan and K’ai-ch’eng [3, pp. 172–173]. This extends the documented diplomatic relationship to a span of approximately 170 years rather than 70 [3, p. 173].

Kao also corrects a widespread misconception that the T’ang cartographer Kia Tan (d. c. 825) based his geographical descriptions on personal travel. He shows that Kia Tan never left China, compiling his forty-volume work from Pan Ku’s Han Shu, other reference books, and data supplied by returning traders and envoys who were interrogated by the Hung-lu sze upon arrival at the capital [3, pp. 176–177].

Cultural Influence and the Question of Reach

The extent to which Sri Vijaya’s Mahayana Buddhist culture penetrated its neighbours, particularly the Dvaravati kingdom in the Menam valley and Korat plateau, was a subject of sustained debate in the 1970s. H.G. Quaritch Wales, revisiting the question he had first raised in his 1969 monograph Dvaravati, argued that Sri Vijayan influence abroad was far more limited than recent scholarship—particularly that of J. Boisselier—had claimed [4, pp. 4–5].

The debate centred on the Chula Pat’on bas-reliefs at Nak’on Pathom, discovered in 1968 during road construction. Boisselier had identified certain figures among the scenes as Bodhisattvas of Srivijayan style and concluded that Dvaravati underwent a period of Srivijayan Mahayanist influence at the end of the eighth and in the ninth century [4, pp. 4–5]. Piriya Krairiksh’s intensive study, however, showed that the style of the figures’ crowns was closer to a sixth-century Siva at Bhumara, India, than to the ninth-century Srivijayan style, and that all the scenes depicted at Chula Pat’on, as well as those at Muang Bon, were inspired by Hinayana Buddhism [4, p. 5]. Since no cult images of Bodhisattvas had been found, nor any relief showing a scene derived from a specific Mahayana sutra, Krairiksh concluded that it was Buddhism of the Hinayana school, though of the Sanskrit canon, that was practised at the temple [4, p. 5].

The large number of bronze Bodhisattvas from the Korat plateau, including the Prakhon Chai hoard of several hundred figures, had been classed by Boisselier as showing Srivijayan influence, particularly in the style of the jatamukuta (crown) and the pearls ornamenting the centre of each hair loop [4, pp. 5–6]. Wales argued that these bronzes were better explained as Pre-Angkorian Kompong Prah style, carrying central Javanese features, and that the Mahayana religion they represented had reached the Korat plateau from Cambodia rather than from Sumatra [4, pp. 5–6]. He noted that no fragment of a mitred Visnu had been found on mainland Sumatra, undermining the attribution of the Peninsula’s mitred long-robed Visnus to Srivijayan production [4, p. 9]. S.J. O’Connor had shown that these figures represented an accretion of Indian styles—Gupta, post-Gupta, and Pallava—accumulated from the sixth to the eighth centuries over the original Chaiya Visnu type, which O’Connor dated to no later than A.D. 400 [4, pp. 9–10].

Wales concluded that, while the finding of scattered Srivijayan images in Siam suggested that this art was appreciated, its influence was small and the Mahayana of Sri Vijaya was not accepted in Dvaravati [4, p. 11]. He noted that monks from Nak’on Si Thammarat introduced some Srivijayan architectural features to Subhodaya in the fourteenth century, but that this postdated the political extinction of Sri Vijaya itself [4, p. 11].

The Inscriptions and the Decimal Zero

Anthony Diller’s 1995 article addressed a gap between two scholarly communities: Southeast Asian epigraphers, who had long recognised the Sri Vijayan inscriptions as valuable for dating the polity’s history, and historians of mathematics, who had largely ignored the same inscriptions’ significance for the decimal number system [1, p. 55]. Diller traced the origin of the observation to Coedès’ 1931 article, in which Coedès cited the Saka-era dates 605, 606, and 608 on Old Malay and Old Khmer inscriptions as early examples of zero in its “keystone” place-holding role, using them to refute G.R. Kaye’s claim that decimal notation in India was a ninth- to tenth-century development [1, p. 55]. Despite this, standard reference works in the history of science continued to omit or misrepresent the Southeast Asian evidence, a disciplinary insularity Diller attributed to the article’s publication in a French-language journal of Oriental studies rather than in a mathematics history venue [1, p. 55].

The core of Diller’s argument moves beyond the simple chronological claim to explain why this evidence survived in Southeast Asia when comparable Indian material apparently did not. He contrasts classical Indic inscriptional conventions—Sanskrit verse, word-symbol chronograms for metrical flexibility, regnal-period dating, and a sacral culture centred on oral recitation—with the Old Malay practices visible in the Sri Vijayan corpus: prose form, direct decimal numerals, Saka-era dating, and a distinctive belief in the physical potency of inscribed texts [1, pp. 60–61]. The Telaga Batu inscription, carved on a naga-canopied stone with a drainage funnel so that ritual water could be poured over the text and “drunk” by appointed governors, exemplifies this sacral-textual sensibility [1, pp. 60–61]. Diller draws on D.C. Sircar’s observation that Cambodian eulogies of the same period regularly recorded accession dates by Saka era, a practice uncommon in contemporary Indian prasasti inscriptions, to support the case for a regional departure from the Indic prototype [1, p. 61].

The specific inscriptions at issue are the Kota Kapur inscription (Bangka Island), dated Saka 608 (686 AD), the Kedukan Bukit inscription (Palembang), dated Saka 605 (683 AD), and the Talang Tuwo inscription (Palembang), dated Saka 606 (684 AD), all of which contain the zero figure in decimal place-holding dates, predating the Gwalior inscription of 876 AD—the accepted first extant Indian example of physical zero—by nearly two centuries [1, pp. 54]. The Khao Rang inscription (Thai-Khmer border area), dated Saka 561 (639 AD), provides the earliest known inscriptional use of the word-symbol (chronogram) method in Southeast Asia, using “senses-tastes-moon” (5-6-1) to encode the date [1, p. 57]. The Mi-son Stele of Champa, dated Saka 609 (687 AD), represents the central zero in the date by the word amvara (atmosphere), confirming the zero function in a neighbouring polity within the same decade [1, p. 57].

Diller situates Sri Vijaya within a wider Indian Ocean commercial network, comparing its position to that of the Euphrates crossing at Kenshra, where the Nestorian Bishop Severus Sebokht praised the Indian numerical system in 662 AD [1, pp. 59–60]. He suggests that Sri Vijaya’s “multicultural” commercial orientation, attested by the Chinese pilgrim I-tsing, made it a natural nexus for practical decimal notation, and that local court astronomers (hora) may have applied the system to both calendrical and commercial calculations [1, pp. 59–60]. The broader theoretical frame is one of “localisation” or “indigenisation”—Southeast Asian formations contributing subtle but crucial departures from Indic prototypes rather than passively receiving them [1, p. 55]. By 856 AD, however, Indic conventions of Sanskrit verse and word-symbol chronograms had been adopted in Old Javanese inscriptions, indicating that the Old Malay practice of prose with decimal numerals was a transient local configuration that fell out of use within a century [1, p. 62].

Research and Documentation

The Society’s literature on Sri Vijaya spans nearly a century and reflects the shifting priorities of Southeast Asian historiography. The earliest substantive contribution in the archival record is the abridged translation of Moens’ 1937 article “Srivijaya, Yava en Kataha,” published in 1939, which introduced to the Malayan scholarly community the Dutch revisionist challenge to Coedès’ Palembang identification [5]. Winstedt’s brief 1941 note on a Perak site connected to the Sailendra empire, and his 1949 article on the Ligor inscription, represent the Malayan engagement with the question of Sri Vijayan territorial reach beyond Sumatra [6], [7].

The 1950s and 1960s saw a shift towards philological precision in the treatment of Chinese sources, exemplified by Kao’s 1956 article, which corrected persistent misconceptions about the methodology of T’ang-era geographical compilation and extended the documented diplomatic relationship between Sri Vijaya and China by a century [3]. The 1970s brought two complementary interventions: Quaritch Wales’ 1978 article, which drew on his own fieldwork in the Malay Peninsula and eastern Siam to challenge the expanding claims of Boisselier regarding Sri Vijayan cultural influence [4], and Wolters’ 1979 survey, which proposed a fundamentally new methodological framework for studying a polity that left no conventional monumental trace [2]. Wolters’ article was written as a personal tribute to the organizers of the March 1979 SPAFA-sponsored workshop in Jakarta, which prepared an international multi-disciplinary research proposal for studying Sri Vijaya over the following five years, with delegates from Indonesia, Malaysia, the Philippines, and Thailand [2, p. 3].

The 1995 Diller article represents a further disciplinary shift, bridging the gap between Southeast Asian epigraphy and the history of science, and challenging the standard narrative that locates zero’s material origins exclusively in India [1]. The article draws primarily on the inscriptions published in the MBRAS Monograph 20 (1992), the Coedès-Damais anthology edited by Manguin and Sheppard, alongside secondary literature in the history of mathematics and Southeast Asian epigraphy [1]. Historiographically, the Society’s literature on Sri Vijaya has moved from the initial problem of identification and location, through the question of political and cultural extent, to the finer-grained questions of textual practice, numerical notation, and the relationship between local and Indic conventions—a trajectory that mirrors the broader maturation of Southeast Asian studies as a discipline.

MBRAS Sources

References

  1. A. Diller (1995). Srivijaya and the first zeros JMBRAS 68(1): 53–66.
  2. O.W. Wolters (1979). Studying Srivijaya JMBRAS 52(2): 1–32. Read on JSTOR
  3. W.T. Kao (1956). A primary Chinese record relating to Ho-lo-tan, and miscellaneous notes on Srivijaya and Fo-Che JMBRAS 29(1): 163–178. Read on JSTOR
  4. H.G.Q. Wales (1978). The extent of Srivijaya’s influence abroad JMBRAS 51(1): 5–11. Read on JSTOR
  5. J.L. Moens (1939). Srivijaya, Yava en Kataha. Tijdschrift voor Indische Taal-, Land– en Volkenkunde (Deel LXXVII Aflevering 3, 1937) uitgeven door het Koninklijk Bataviaasch Genootschap van Kunsten en Wetenschap-pen. An abridged translation R.J. de Touche JMBRAS 17(2): 1–108.
  6. R.O. Winstedt (1941). The Perak site of the Sailendra empire JMBRAS 19(2): 242.
  7. R.O. Winstedt (1949). The Ligor inscription JMBRAS 22(1): 176–177.