Vajravārāhī in Khara Khoto and Prajñāpāramitā in East Java: Connected by Pearl Ornaments

Abstract

This research, situated in the geographical and historical context of the Tangut and East Java, uncovers a significant aspect of the evolution of Buddhist art styles. A thangka of the goddess Vajravārāhī found in Khara Khoto, dated to the late 12th century, shows the bodhisattva decorated with a pearl-chain girdle and upper-arm bands. This form of pearl-chain jewellery, which appears on Vajravārāhī and other Sino-Tibetan-style bodhisattvas, also appears on three stone statues of the goddess Prajñāpāramitā in East Java, all of which depict a near identical use of this pearl-chain ornamentation, as well as on a statue of Prajñāpāramitā at the Muara Jambi Buddhist site in Sumatra. Maritime trade between the regions of China and Java was extensive. The commonality of such motifs in China and Java may highlight a convergence of cultural forces and perhaps shared styles originating from the maritime realm and traded via maritime routes; however, a direct or indirect influence of Sino-Tibetan styles on thangka paintings featuring this depiction of the jewellery perhaps occurred following dynamics of north–south exchange, highlighting the interrelated links along maritime and overland routes through the Pāla Buddhist kingdom in eastern India. Thus, I propose that the connection between the Vajravārāhī and other Tibetan thangka paintings was inspired by Northeast Indian influence from the Hexi corridor, eventually reaching East Java.

1. Introduction

Close analysis of the Buddhist statues and paintings discussed in this article suggests that a unique Pāla and Sena dynasty-inspired art style developed from the early 12th to the early 13th centuries in Central Asia1—more specifically, the Tangut empire, also known as Western Xia or Xixia (1038–1227)2—and parallels the Nepalese and Indian traditions in the form of pearl ornaments. The body’s physique exhibits a svelte smooth torso, with a slight sway to the hips and a naturalistic body casting. The facial features exhibit high arched brows, a long sharp nose, and a full lower lip. The body is generally shown with a short hip wrapper or dhoti, often depicted with one edge longer than the other, over which many metal girdles and belts are draped. The torso is draped with the upavīta, frequently shown as a long metal chain; a fabric sash is also depicted across the body from the left shoulder. This description applies to the early Pāla period, from the 8th to the 9th centuries; by the 12th to 13th centuries, the number of ornamentations increased, and the body was often adorned with a heavy neckplate and upright triangular-shaped upper-arm bands.

This article suggests that the lithe form of the torso and the facial features, which reflect the Pāla Sena characteristics, are evident in a late 12th-century Khara Khoto thangka painting and again in a corpus of ca. 13th-century sculptures from East Java and Sumatra; however, the addition of the pearl-beaded ornaments in both of these regions under discussion is not evident in the Pāla sculpture, which perhaps highlights a shared heritage that developed thanks to the contacts between the regions connected via maritime routes.

Empirical research and analysis, along with stylistic interpretations, may be the only available means of understanding connections, as the extant relevant premodern texts do not inform us that a particular art style was transferred and reinterpreted in another region, sometimes hundred years later. Making use of this methodology, this article discusses pearl ornaments depicted in a painting of the tantric Buddhist deity Vajravārāhī from Xixia in comparison with similar motifs—a particular style of pearl armbands and chains—found in a corpus of three statues of Prajñāpāramitā and an unknown Buddhist goddess from the Siṅhasāri period in East Java and Sumatra (Pullen 2021a, p. 237), along with a Durgā statue from the Central Javanese period.

These fine examples of this rare style, along with some of the other paintings and stone statues, provide us with a case study of the intra-Asian maritime networks of Buddhism, and highlight the religious and diplomatic networks in the East Asian sphere. Indeed, while the reappearance of pearl jewellery ornaments on Prajñāpāramitā and other 13th- to 14th-century East Javanese Buddhist sculptures, which closely resemble those depicted on the Vajravārāhī in Khara Khoto, may suggest a broad commonality of the way that pearl ornaments were represented in disparate regions of the Pan-Asian Buddhist world, I argue that the connection between the Vajravārāhī and the other Tibetan thangka paintings may have been actually inspired by Northeast Indian influence from the Hexi corridor, and eventually reached East Java and Sumatra. In order to strengthen my hypothesis, I will explore the maritime trade network and discuss the north–south exchange as well as the interrelated links between those regions and the Pāla kingdom in eastern India.3

In a recent study looking at new and old sources of the Southeast Asian mainland in the c. 1000–1500 period, Kenneth Hall states that, during this period, “there was contact with and influence by the inclusive communication and commodity networks of the extended Eastern Indian Ocean maritime realm that stretched by land and sea from the Bay of Bengal to the South China Sea” (Hall 2019, p. 267). This quote highlights the aims of this research, namely to suggest that these three locations between the 11th and mid-13th centuries probably owe their overwhelmingly stylistic merits to Pāla art styles.

2. Bodhisattvas with Pearl Ornaments

For centuries, pearls were traded across Asian maritime routes. The Cōḻa dominions saw trade with China—pearls, amongst many luxury goods, were part of this trade. The Chinese Supervisor of Maritime Trade, Zhao Rukuo 趙汝适, compiled the Zhufan zhi 諸蕃志 (12th to 13th centuries), with several descriptions of the pearl trade. During the Yuan dynasty, they derived significant revenue from the duties on all imported merchandise, including precious gems and pearls.4 At this time, the southern sea trade was firmly in the hands of the Arabs (Dashi 大食) and other foreigners, namely from Java (Shepo 闍婆) and Palembang (Sanfoqi 三佛齊) (Hirth and Rockhill 1966). The pearls were traded in from Jambi in South Sumatra, the second base of the Śrīvijaya, which was destroyed by the Javanese in 1275.5 The primary sources of pearls were Sulu in the southern Philippines, Kampar near Jambi, and Java. The Chinese used pearls as head ornaments; due to the high price of pearls, traders smuggled them into China to avoid taxes (Miksic 2013, p. 131).6

For thousands of years, people have coveted pearls as symbols of wealth and status. Dakin’s small book on pearls (Dakin 1913) records a complete and comprehensive history of pearl knowledge. In short, pearls have been known since the 2nd–3rd century BC, when the first Chinese records were written, as freshwater molluscs. In the Indian Ocean region, pearls were known for centuries before the Common Era, perhaps collected as curiosities; in the New Testament, on the other hand, there are constant references to pearls. Pliny referenced that fishing off Ceylon was the most productive part of the world—records show that gifts of pearls were sent from the King of Ceylon to this Indian relative in about 540–550 BC. Dakin states that pearls were almost certainly known to the Persians seven centuries before the Common Era. The Romans had a strange fetish with pearls and forbade the wearing of pearls by people who had not obtained a certain rank; however, it was not until the 12th century that pearls were used in England, although through the 13th and 14th centuries, pearls became extremely fashionable throughout Europe (ibid., pp. 1–7).

Pearls were symbols of society, high rank, beauty, and fertility, as well as adornments for royalty, emperors, and religious figures. In Buddhism, the pearl (maṇi Skt., nor bu Tib.) is considered the jewel or ratna in the lotus that grants all desires, the divine pearl—a belief adhered to by Buddhists throughout Asia (von Schroeder 2008, Glossary 179). The pearl is also a symbol of spiritual wealth; its luminosity brings all the treasures and teachings of the Buddha and represents virtues, wisdom, and compassion. In the Metropolitan Museum, New York, is an Avalokiteśvara (Guanyin) ascribed to the 6th century (Figure 1).

The description reads, “the astonishing jewelled harness adorning this bodhisattva is made up of two long strands of pearl-like clusters and multifaceted beads”. The jewels perhaps refer to a passage in the Lotus Sūtra where the historical Buddha Shakyamuni extols the great compassion of Avalokiteśvara and presents him with a pearl necklace as a symbol of benevolence. These aspects perhaps explain why the adornments of many bodhisattvas found in thangka paintings and Javanese sculptures feature pearls.7

3. Early Representations of Pearl Ornaments at Ajanta, Sigiriya, and Central Java

Some of the figures in the Buddhist paintings in the Sigiriya frescoes (5th century) and the rock-cut temple paintings (ca. 5th–7th century) at Ajanta, in Sri Lanka and Western India, respectively, highlight the use of pearl jewellery. The paintings at Sigiriya show women and celestial beings adorned with gold jewellery and gemstones, along with strings of twisted pearl ornaments as necklaces and upper-arm bands; however, the style is less sophisticated and detailed and does not resemble that displayed by the Vajravārāhī and Prajñāpāramitā under discussion (Figure 2).

This is corroborated by similar depictions of mural paintings in the Ajanta Caves of the Gupta period in Maharashtra, India. A mural painting in Cave 1 shows the bodhisattva Padmapāṇi wearing a pearl necklace (Figure 3), while an analogous painting in Cave 2 shows him adorned with multiple pearl necklaces and pearl upper-arm bands. This, of course, highlights that this type of ornamentation also precedes that found in statues and paintings from the two Buddhist regions discussed in this paper (Okada 1995, p. 123).

Another relatively early—for the chronology of Hindu–Buddhist Javanese art—example of pearl ornamentation is a fine Durgā statue from Candi Śiva at Loro Jonggrang in Prambanan (Figure 4), Central Java. In the statue, datable to the ca. 9th century, Durgā is depicted as standing and holding a buffalo’s tail in one hand and a demon in the other. She is adorned with multiple necklaces, belts, sashes, and armbands. The neckplate and upper-arm plaques depict a gold ornament with trailing pearl beads. These pearls are also apparent on the lower two girdles around the hips. Despite the neckband and arm plaques being somewhat Pāla in style, the remainder of the ornaments are Central Javanese in inspiration (Figure 4).

Although this statue of Durgā is not directly relevant to the main argument of this article, it is still useful to contrast it with the later material discussed here, also because hardly any statuary exists—especially in Indian style—between the end of the Central Javanese period (ca. early 10th century) and the mature East Javanese period (13th–15th centuries). We can surmise that the statue suggests direct links between Java and the Pāla domains, which were undoubtedly present in the Buddhist inscriptions and statuary of the 8th–9th centuries. The jewellery style of both neck and triangular upper-arm bands along with a long chain (upavīta) reflect a so-called “Pāla style”, as well as the physiognomy and the ornamentation in a Maitreya statue of the 9th to 10th centuries (Figure 5, left). It is intriguing to note that this style seems to be already present in Central Java perhaps a hundred or so years earlier than any statues appear in eastern India, as we see in the seated Śiva8 in the Indian Museum, Kolkata (Figure 5, right).

Between the 8th and 11th centuries in both eastern India and within Java, there are numerous statues depicted with the triangular-shaped upper-arm bands and the necklace depicted as a solid plaque, both ornaments probably representing gold made with the repoussé technique. Therefore, we can see a progression in style from the 11th century in both the Vajravārāhī and the Prajñāpāramitā, where jewellery styles have evolved and gained the addition of pearl swags.

4. Vajravārāhī from Khara Khoto

Vajravārāhī belongs to a cycle of deities associated with tantras such as the Hevajratantra and the Cakrasaṁvaratantra, such as in the case of the Khara Khoto image, of Vajrayāna or esoteric Buddhism.9 This painting of Vajravārāhī, dated to after 1189 but before 1200,10 was found in Khara Khoto, in what is now China (Figure 6). The artwork is now in room 363, Inv. no. XX-2393, in the State Hermitage Museum, St Petersburg. In Tangut icons, Vajravārāhī is depicted alone, as in this case, or as the consort of the deity Saṁvara, one of the hypostases of the male esoteric Buddhist deity Hevajra.

This painting belongs to a corpus of hundreds of Buddhist artefacts showing both Chinese and Tibetan styles, and also reflecting the Indian Pāla style of works associated with Tibetan Buddhism. Artworks recovered from Khara Khoto11 and other sites provide compelling examples of the achievements of the Xixia people (Khokhlov 2016, p. 1). We presume that the paintings under discussion here were buried in the sands of Mongolia even after the city was taken in 1227, when it was destroyed in a campaign against Genghis Khan, and the Russians rediscovered them in the early 20th century.

In this painting, the description given by the State Hermitage Museum mentions Vajravārāhī as being depicted standing on her left leg, trampling a prostrate figure representing opponents of the teaching. She holds attributes that are in keeping with her iconography, and she has a necklace made of the severed heads of those who antagonise the teachings of the Hevajratantra. Around the goddess are six of her hypostases, denoting the six syllables in the incantation pronounced in honour of Vajravārāhī. The background depicts eight cemeteries—sacred cremation sites in the Indian subcontinent.

Despite the wrathful nature of this figure, her upper-arm ornament depicts a fine jewelled band with strings of small spherical white objects representing pearls flying up the arm and falling to the elbow. The upper armlets (rucaka), fine girdles (mekhala) with double strands of swags alternating with short tassels around the hips, and neck ornaments (kaṇṭhi) are all executed to depict tiny fine pearls; alternatively, they can be described as parures of beaded white jewellery. I deduce this from the small white dots on the red ground; however, Kossak (2007, p. 86) has described this ornament differently, proposing that these presumably represent human bone beads.

In his thought-provoking chapter on the role of jewellery in early Tibetan thangkas, Kossak describes at length the possible role pieces of jewellery can play in sculpture and paintings in dating-related objects. Kossak states there are “several possible explanations for the similarities in the ornaments won by benevolent and wrathful deities” (ibid., p. 85). Wrathful deities are defined as a class of esoteric divinities associated with the Yoginī Tantras—in the Indian subcontinent and the Himalayas they only became prominent in the 11th century (ibid., p. 85). He states that

the jewellery that ultimately became associated with wrathful deities was made solely of human bone rather than traditional precious metals […] beginning around 1300 in Tibet, the prominence of a variety of wrathful deities as objects of veneration corresponds with their jewellery being depicted as bone rather than metal.

(ibid., p. 86)

Kossak argues that all the wrathful deities, such as Vajravārāhī, in the period from the 11th through the 13th centuries were adorned with a mekhala (girdle) consisting of broken pieces of human bone; however, a review of early painting shows that artists were undoubtedly showing belts and girdles made of gold (ibid., p. 86). Furthermore, by the late 13th century—which, nota bene, is coeval with the East Javanese Prajñāpāramitās, who are not wrathful deities—the ornaments are described as Nepalese-inspired rather than Pāla-inspired, with necklaces of “oval jewels strung between beads” (ibid., p. 95; what the “beads” might be is not specified).

Kossak’s argument, based on references to the colours used by painters of thangka paintings, suggests that the colours used reflect the materials of the jewels and that a white girdle (as we see on the Vajravārāhī) might be made of ivory or bone. He argues that tantric practitioners were instructed to wear human bones as ornaments and to substitute gold and jewels with human bones (ibid., p. 86). Oppi Untracht, in his important book on Indian traditional jewellery, argues that organic materials used for the rosaries of high Lamas and the wealthy were made from beads consisting of pearls, coral, amber, conch shells, ivory, and animal as well as human bone; however, a highly prized type of bead used by Tantric Buddhists was a flat disc of a uniform diameter made from human bone (Untracht 1997, p. 73).

These two arguments serve as a counterpoint to the argument put forward in this article, that the Vajravārāhī and Prajñāpāramitā are both adorned with pearl ornaments—this is based on the spherical shape of the depicted beads, the colour used in the paintings, and the deities wearing entire parures (set of jewels) of white beading that would appear to be far more indicative of pearls with their lustre and circular form than a bone bead in the dull form of a flat disc.12

According to my own reading of this image, the facial features of the Khara Khoto Vajravārāhī, the aquiline nose, the hairstyle and abundant jewellery, and the treatment of the hands reflect Indo-Himalayan imagery in 12th-century China.13 The countenance of Vajravārāhī is more in keeping with the Indo-Himalayan style discussed earlier than in a more typical Chinese art style; however, the ornaments appear to be the defining feature of this period in Khara Khoto in the 12th century. The Indo-Himalayan imagery and Nepal-based artistic traditions are reflected in the Sino-Tibetan paintings, such as this Vajravārāhī ascribed to the style of the Indian Pāla dynasty and the art of the Xixia Empire.

The paintings recovered from Khara Khoto and other sites by Peter K. Koslov (1863–1935) in 1908/9 attest to a melting pot where artists worked in different styles, for instance, Northern Song styles associated with Chinese Buddhism or Indian Pāla styles associated with Tibetan Buddhism. This co-existence of themes and styles is characteristic of Xixia culture, which continued under the Mongols and became part of Yuan art; however, works related to Tibetan Buddhism are primarily based on Pāla art styles but possess distinctive Xixia features and can be distinguished from the Pāla style made in Tibet.14 This fact led scholars to surmise that Xixia was receiving artistic input directly from India and not only from Tibet (Khokhlov 2022, p. 112). What is important to highlight for the purpose of this article is that the interpretation of the pearl ornamentation found on Buddhist deities is the definitive feature that distinguishes the Xixia style from a typical Pāla style,15 a style where pearl ornamentation is not apparent, as the depiction of upper-arm ornamentation is quite different and can be best described as a triangular-shaped plaque (Figure 7).

5. Prajñāpāramitā from Candi Singosari, East Java

The depiction of the upper-arm ornaments of the Khara Khoto Vajravārāhī, appearing as a band with hanging pearl swags, is very similar to upper-arm ornaments16 of a 13th-century Prajñāpāramitā stone statue from East Java, now housed in the Museum Nasional in Jakarta (Figure 8). In this particular instance, the goddess is adorned with triangular-shaped arm plaques with pearl swags.

The Prajñāpāramitā was found in a subsidiary Candi E at the Singosari complex near Malang and dated ca. 1280.17 The statue is carved from smooth andesite stone, which appears as a good material for the carving of detailed ornaments and textile patterns. Of all the East Javanese sculptures, the execution of Prajñāpāramitā is the most perfect, with only slight damage to her fingers. She is seated in the lotus position (padmāsana) on a double lotus cushion placed on top of a square pedestal carved with a frieze of small rosettes around the pedestal, probably representing lotus flowers. Her textile pattern resembles the High Tang rosette flower, a lotus outlined with palmettes, described as karahana (Raspopova 2006, p. 67; Pullen 2021a, p. 181).18

The goddess’ diadem is secured by floral bands that fall over her shoulders and are trailed by pearl strings. Across her forehead is a band with looped trails of pearls. Around her neck appear two bands, one of a gold plaque in the repoussé technique and the other with looped stands of pearl beads. The serene, inward-looking pose belies the unmistakable royal exterior of her ornaments and dress style. The udharabhanda (stomach band) and the triangular upper-arm band represent a similar type of ornament as her neckband, adorned with pearl-chain swags that fall from the lower edge and appear to fly up her arm. This form of triangular neckband and upper-arm bands is only apparent in a brief period, described by Kossak as Pāla-style ornaments sometimes supplemented by auxiliary beading, or “swags of beading often hang from the band” (Kossak 2007, p. 89). Evidence of pearl armbands join the two statues of Vajravārāhī and Prajñāpāramitā, yet the style of attaching them to the arm differs: the Vajravārāhī wears a cloth band to which the swags of beads are attached, whereas the Prajñāpāramitā wears a triangular plaque, probably in gold, from which the swags or trails of pearl beads fall (Figure 9).

There are two other statues from East Java, viz. a Prajñāpāramitā as well as an unknown goddess, which portray an “artistic mode”19 in the style and design of their pearl arm ornaments that is also reflected in the Vajravārāhī from Khara Khoto. These two statues will be discussed below.

6. Unknown Goddess and Headless Prajñāpāramitā from Candi Singosari, East Java

The first statue is a headless Prajñāpāramitā at the Candi Singosari site in East Java. Since it is still in situ, the damage to the statues is extensive. The identity of the goddess is suggested by the mudrā of her hands and the thick stem of the lotus plant around her left arm. Of note here is the carving of the stone backplate reflecting a Pāla art style with the crossbar behind her shoulders and two śārdūlas, rampant winged leonine creatures placed atop a small elephant, carved on either side of the torso (Pullen 2019, p. 2). Her ornaments reflect the other two Prajñāpāramitā from East Java, with the large repoussé gold neckplate and triangular upper-arm bands both showing strands of swags of pearl beads, with the chains flying both up the arm and falling below the gold ornament (Figure 10).

The second statue, representing an unknown goddess, is dateable to c. 1300, and has been published by Pullen to highlight the current use of carved textile patterns in East Javanese sculptures (Pullen 2021a, p. 191, Figure 139). It is discussed here to highlight the ongoing discussion on the evidence of pearl upper-arm bands. The statue, albeit badly eroded, probably represents a royal figure, and shows many similarities in the use of pearl arm ornamentation with the Vajravārāhī from Tangut. On careful close research and with knowledge of the Singosari pearl decoration of Buddhist statues, there is still faint evidence that she was richly adorned with ornaments and a carved textile pattern. The swags of pearl chains are apparent on her right arm, even though the stone’s surface is very exfoliated; both of these statues are decorated with a sacred pearl thread in a similar style, which falls to the lotus cushion (Pullen 2019, pp. 19–22) (Figure 11).

7. Prajñāpāramitā from Muara Jambi, Sumatra

Another headless statue of Prajñāpāramitā, dating to the mid-13th century, is kept in a museum20 at the Buddhist site of Muara Jambi.21 She is decorated as a royal figure with profuse ornamentation, with the upavīta hanging down over both shoulders onto the cushion in the front of the body; the upavīta represents a five-strand twisted chain of pearls joined with a clasp resembling gold work depicted at her left breast. Every ten to fifteen centimetres along the chain are spacer ornaments—possibly representing gold work. Her large neckplate plate and triangular upper-arm bands are probably made of gold in the repoussé technique; falling from the lower edges of both ornaments are strands of pearl swags (Figure 12).

A worn but decorated seléndang (shawl) is depicted as falling between the breast and beneath the right arm; on the rear of the body, it is visible across the back, finishing with a pleated end over her left shoulder. Closely examining the textile pattern on the goddess’s sashes could represent an embroidery technique. Embroidery has been popular in China, used to create the intricate details needed for Buddhist textiles or kesi22 tapestry weaves or incised silk, an example of which being a monk’s garment.

8. Later East Javanese Statues

If we are to survey Central and East Javanese sculptures for evidence of this pearl arm decoration, it is unlikely that we will find any further statues; however, besides the 9th-century Durgā from Central Java (see Figure 4), examples of pearl decoration as part of statues’ general ornamentation are found in a few late East Javanese period statues, viz. the statue of a seated “royal couple”, the statue of a queen or goddess from the 14th to 15th centuries, and a deified ancestor figure in the guise of Śiva from the same period. In all these examples, on close analysis, it is possible to see the use of pearl beaded swags hanging from the lower edge of the upper-arm band, from the gold neck plate, beneath the stomach band, and the girdle around the hips. The Majapahit examples of the 14th and 15th centuries also depict a long upavīta consisting of four- and five-strand pearl beads (Pullen 2021a; Polak 2024) (Figure 13).

9. The East Javanese and Sumatran Prajñāpāramitā Statues in a Transregional Perspective

To describe the transregional geoenvironmental area of “Maritime Asia” in the mediaeval period,23 Acri highlights the circulation via the southern maritime routes of Buddhist monastic and lay agents—traders, artists, pilgrims, etc.—along with cosmologies, texts, and ritual as well as devotional practices, all of which contributed to the shaping of the Buddhist cosmopolitan continuum articulated in Sanskrit and literary vernacular languages across much of Asia (Acri 2019, p. 50). Keeping in mind this picture of transregional and transcultural connectivity that linked faraway regions of Asia, we may ask the following question: how did the design of the ornaments worn by the Vajravārāhī in Khara Khoto and found in East Java about a hundred years later come about? Perhaps there is no answer except that pearl ornaments represent a commonality of styles during this period, which were transmitted by the agents plying the maritime routes, indeed highlighting the transregional artistic connections in making divine images.24

On an archipelagic interregional scale, stylistic similarities between the Muara Jambi Prajñāpāramitā and the other Prajñāpāramitā from East Java highlight close political, religious, and artistic connections between the two regions during Kṛtanagara’s reign. On a wider intra-Asian scale, as no stone images of Prajñāpāramitā appear in the earlier Central Javanese period, the creation of these statues in the Siṅhasāri period points to an influx of new religious and political ideas25, as well as artistic styles.

While the stylistic similarities are apparent, considering the lack of supporting direct evidence, I prefer to work on the premise that the Prajñāpāramitā from Java and Sumatra were made in different workshops. The suggestion has also been made that Kṛtanagara sent from East Java to Sumatra either the statue or the artisans, who then produced a similar statue for the annexed kingdom in Jambi. Acri and Wenta propose that “Kṛtanagara is known to have commissioned replicas of icons he deemed significant and dispersed them across his realm, including East Java and Sumatra” (Acri and Wenta 2022, par. 26). Although an inscription on the Amoghapāśa statue from Padang Roco states that Kṛtanagara sent it to Sumatra, the same circumstance cannot be proven with respect to the Prajñāpāramitā statue as no inscription or text mentions it.

Late Śrīvijaya was a significant player in the international relations of the region until 1025 and the invasion of the Cōḻas—however, there is no evidence that the Cōḻas had a fleet of ships; it is more likely that there were South Indian merchant guilds who were powerful enough to carry out sea raids and to improve their commercial activities (Miksic and Goh 2016, p. 360). Another theory is described by Wade and Tana: “limited numbers of Indian and Muslim merchants continued to come from the west, but the Chola invasions from India, for example, were not sustained” (Wade and Tana 2012, p. 74).26 Śrīvijaya targeted South Indian ships and controlled the supply of products from the Indian subcontinent, with goods such as elephant tusks, gharu-wood/hardwood, pearls, aromatics, and medicines, amongst other luxury goods, and subsequently traded with the Chinese traders in the Strait of Melaka (Mukherjee 2019, p. 187). Research into this area shows far more archaeological remains in the Batang Hari River region around Muara Jambi than in Palembang.

The analysis of sculptures in the Pallava and Cōḻa regions of South India does not reveal any examples of statues depicting pearl ornaments—even though, within “Indian” culture, pearls were used as a form of adornment as seen in the murals of the Buddhist Ajanta caves in an earlier period. Within the Hindu traditions of South India, the use of pearls did not appear to be significant.

10. Mañjuśrī Arapacana in East Java

Khubilai Khan saw himself as “a wrathful ‘imperial’ incarnation of Mañjuśrī as Mahākāla” (Acri and Wenta 2022, par. 56). Acri and Wenta propose that as this form of Mañjuśrī as Mahākāla may have been practiced at both Yuan and Siṅhasāri courts, as a likely result of a simultaneous wave of Indian and Tibetan masters, the consequence was that both sovereigns were initiated into tantric systems (ibid., par. 31). In the same period, there appears to have been a latent resurgence of art styles from Bengal, as seen in this statue of Arapacana (ibid., par. 61–67). Dhingra and other scholars argue that the relationship between the images from Bihar and Bengal in the Pāla period after the 10th century and those seen in Java has become weaker (Dhingra 2022, p. 8); however, by the mid-13th century, the iconography from the Indian subcontinent27 is repeated on this Mañjuśrī, and a local indigenous style was created (Figure 14).

This stone statue is the only example of Mañjuśrī Arapacana in East Java; however, several statues of Mañjuśrī were found during the Central Javanese period in bronze and stone. The sinjang (long cloth) worn by the Javanese statue is carved with a complex pattern of concentric circles made up of four different designs, which alternate and repeat across the cloth. The carving of the patterns is very clear and precise and appears to reflect several other influences, for example, the senmurv (or simurgh) bird, which is often depicted on Sasanian textiles (Hakobyan and Mikayelyan 2018). This example may suggest that textile patterns were transmitted to Java from overseas following long-distance trade, for instance, from the royal courts in Yuan China to the Javanese court. It seems reasonable to suggest that such kinds of textiles could have reached Java as part of a Chinese gift exchange or through Arab traders from Persia. This sculpture is the only evidence of the existence in Java of this textile pattern attributable to Persia. Amy Heller describes the movement of designs: “From Persia to China, medallions of single or confronting animals with a pearl roundel were among the most popular designs of the period. The creatures were frequently imaginary or hybrid birds or animals” (Heller 2006, pp. 175–88; Pullen 2021b, pp. 157–68) (Figure 15).

11. Icons Featuring Pearl Ornaments from the Himalayan Region and Tibet

From the trading port of Tamralipti on the Bay of Bengal, through Bangladesh, Nepal, to Leh in Ladakh, Kashmir, through the Sogdian region to Dunhuang and Yulin and into the Hexi corridor and Khara Khoto, a similar artistic mode in terms of the depiction of pearl ornaments on Buddhist icons is seen. A relevant example of iconography involving pearl decoration located on the Buddhist trail at Alchi28 is that of a Green Tārā (appearing as Prajñāpāramitā) mural painting dated to the third quarter of the 11th century (Figure 16).

Although this Tārā is dated to 150 years earlier than the Prajñāpāramitā at Muara Jambi and the Prajñāpāramitā from East Java, we see long strings of pearls in four strands on one image and as a rope on the other. We can also see the pearls on her upper-arm bands (whereas the Muara Jambi Prajñāpāramitā has the pearls hanging from her triangular gold arm plaque).

Other relevant examples are two—no longer extant—clay sculptures, one of Vairocana and one of Amoghasiddhi, from the Rkyang bu temple in Central Tibet, which was established shortly before 1037.29 Both statues exhibit pearl upper-arm ornamentation, similar to the Vajravārāhī of the late 12th century. They are depicted as wearing long swags of pearl ornaments around the neck and flying up from the upper-arm bands, in a very similar style to the Prajñāpāramitā from East Java, now in the Museum Nasional in Jakarta. The pattern on the sash draped across the upper body appears as the upavīta and is carved with a pattern of roundels, which we also see on the jacket of Mahākāla from East Java.30 In the book Tibet in Pictures by Govinda, Li Gotami (Govinda 1979), the figures on pages 40–42, of the 13th or 14th century and probably representing Vairocana, Amoghasiddhi, and Amitabha, appear in the same posture with the same ornaments and patterned textile with roundels. These three statues are described as showing “a strong Indian Influence”, all dressed in rich garments. The continuation of pearl ornamentation and the patterns of roundels on the textiles appear again in East Java, highlighting the continuity of this ubiquitous pattern and jewellery design. Of particular note in these images is the depiction of the crossbar in the backplate, the use of the śārdūla, and the elephant, all of which we see appearing again in East Java in the headless Prajñāpāramitā (see Figure 10).

Khokhlov describes the clay Vairocana and Amoghasiddhi sculptures from the Rkyang bu temple as “decorated with spiked armbands and swagged necklaces, which are characteristic of the Southern Indian mode in Hexi art” (Khokhlov 2022, p. 113).

The connections between Buddhism in the Indian subcontinent and insular Southeast Asia were strong from the early 7th century through the 13th century; however, in the case of Tibet, the elites converted to Buddhism relatively later, after Buddhism had gained a foothold in China (Schoterman 2016, p. 114). As suggested, the Buddhist art produced in the Xixia region includes an Indo-Himalayan style and local contemporary Chinese traditions—the painted images from Khara Khoto attested to the melting pot of artists who worked in many different styles.31 Indo-Himalayan imagery in 12th-century China reflects the previous Pāla aesthetic (Watt 2010, pp. 102–3). In 1260, Khubilai Khan was named ruler of the Mongol Empire, where the customs earlier established by the Xixia were eventually followed. There is a distinct blending of styles in these thangka examples of Indo-Himalayan and Chinese imagery; we see the facial features lean more to the Pāla aesthetic, but the elaborate ornamentation of the multiple strands of pearl jewellery reflects the Xixia elegance (ibid., pp. 102–8) (Figure 16).

A thangka painting of the Four Heavenly Kings from Xixia manifests the syncretic and multifaceted nature of Xixia Buddhism—which brings together Chinese Mahāyāna Buddhist traditions and Himalayan Tantric traditions—with a shift in focus more towards Tibetan Buddhism from the middle of the 12th century onwards. Therefore, we see two types of artistic style, the Chinese and the Tibetan, both of which were influenced by Pāla art (Zhu 2023, p. 3). In this painting, the headdress, the bare torso, and a dhotī for the lower part of the body all appear Indian in style (ibid., p. 5). Zhu explains that “what was absorbed into the Buddhist art of the Yuan was a largely Pāla Indian Style as it filtered through the Xixia people” (ibid., p. 18). This statement highlights the continuous use of the terms Pāla and Indian style in relationship to Xixia thangka paintings.

Despite the general commonalities between Pāla, Tibetan, and Chinese styles in the Buddhist art of Xixia, if we limit ourselves to the study of arm ornamentation, specifically the strings of tiny pearl-beaded swags that appear to fly up the arms, we see that this precise motif appears only in one instance in Central Java and a handful in East Java, and on a few sculptures in Khara Khoto. This fact may suggest that the commonality of Buddhist ornamentation and shared style originated from trade across the maritime (and overland) routes going beyond the Indian subcontinent–Tibet–China region to Southeast Asia.

12. Pāla and “Related Art Styles” Between India, Central Asia, and Java (c. 750–1200)

The Buddhist art of eastern India from the 8th to the 12th centuries had been recognised as a significant source of inspiration for the cultural arts of Tibet and Southeast Asia.32 From the 11th to the 12th centuries, the Bengal region held a strategic location at a crossroads of the Indian subcontinent and mainland as well as island Southeast Asia. Buddhists excelled in producing images of material nature and power, and those images became active when brought to life through regularly performed rituals (Bautze-Picron 2016, pp. 163–64). Images were carried from place to place, from Bengal, where merchants and monks travelled to the Himalayans in the 11th to 12th centuries. Monks from various countries converged at Bodh Gayā. In Pagan, for example, pilgrims who travelled across from Bengal left artistic evidence in the forms of relief temple paintings, cloth paintings,33 and bronzes, which were cast in Bengal and left at Pagan.

Stone images of the seated Buddha in Bengal follow an iconographic programme also found in the Pagan seated Buddha images and murals of the 11th to 12th centuries (Figure 17); however, by the 13th century, ideas from Bengal found their way to East Java34 and became heir to iconographic subjects (ibid. 2016, pp. 165–66), such as the Mañjuśrī Arapacana. Newar artists were likely situated in East Java, which would attest to the style of the Mañjuśrī and might explain how the unusual textile patterns on the kain of the Mañjuśrī reached the region (Pullen 2021a, p. 207) (Figure 15).

Between the 10th and 12th centuries, during the Pāla dynasty (760–1161), Buddhist expansion in Northeast India was seen, with ideas and practices travelling to the Himalayas via Nepal and onto Central Asia. The new Indic traditions had roots in esoteric practices—some of the earliest evidence of these practices is found in central Tibet in the 12th and 13th centuries and the art of the Xixia dynasty in northwest China (Watt 2010, p. 99). The knowledge transfer depended on the work of Indian scholars in Tibet and the many Buddhist centres across various kingdoms. Because of this sharing of ideas and practices, Tibet became the recipient of many Indian Buddhist texts and paintings. Therefore, the various missionaries who travelled throughout the Mahāyāna Buddhist world played a part in the diffusion of Pāla art styles, including the Himalayas and Tibet.35 Bautze-Picron argues that the iconography, be it literary or plastic, was extremely rich and was later inherited by Tibetan and Nepalese monks. Additionally, there is ample artistic evidence from Southeast Asia to prove that intensive contacts were developed between different states of this part of the world and Northeast India (Bautze-Picron 1993, p. 277).

One should not forget to stress that trade routes across Central Asia were important vectors of connectivity, as they were where most indirect trade occurred and most of the contact between India and China transpired. The indirect transmission of ideas and objects, especially of Buddhism and its associated forms, flourished in the fertile grounds of Central Asia (Murphey 1999, p. 184).

The Mongolian city of Khara Khoto is known as the Black or Dead City, a fortified town on the southern Gobi Desert of the Tangut State (1038–1227), built in 1032. In the early 13th century, during Khubilai Khan’s reign, the city was considered a garrison town in the north of Xixia territory (Takeuchi and Iuchi 2016, p. 1; Watt 2010, p. 24). It became an active artistic centre that shared Buddhist art styles compared to those in Tibet, Northern Song, and Buddhist art produced in the Xixia region, including works in the Indo-Himalayan style and styles from contemporary Chinese traditions (Watt 2010, p. 102). Buddhism provided a link between Tibet and the Tangut empire. Xixia fell to the Mongols in 1227, and the capital was largely destroyed. Russian scholars have suggested that Xixia Buddhist culture profoundly influenced the Mongols and played a crucial role in their adoption of Buddhism. Khubilai Khan became the Great Khan in 1260 and 1271, proclaiming the new Yuan dynasty (Khokhlov 2016).

Given the status of the Xixia region as a centre of Buddhism and a nexus of connectivity across Tibet, Central Asia, and China, it is possible that some Buddhist “art styles” were transmitted from that landlocked region and eventually ended up in East Java. But how did these Buddhist ideas and art styles reach Java?

The Mongols commissioned hundreds of Chinese ships built in the newly established southern Fujian province (Watt 2010, pp. 160–61), which enabled them to sail within the region from these southern Chinese ports. China held tribute relations with Śrīvijaya and Malayu/Jambi between the 11th and 13th centuries. The growth of trade with India and China during this period brought coastal people in the maritime region of Southeast Asia into contact with Buddhism and the Sanskrit language and art styles. Many pilgrims who visited the Pāla monasteries of Nālandā and Vikramaśīla, mainly for religious and educational purposes, often returned with examples of Pāla Buddhist artwork. Namely, small portable bronze icons that arrived in Java and frequently became templates for Javanese examples, especially in the 9th century. The text within the Zhufanzhi clearly states that ships were travelling from southern Chinese ports, trading in Java and Sanfoqi/Palembang, and continuing to various ports in India.

As some scholars suggest, in spite of (or perhaps thanks to) the unfriendly relationship between the Mongols and the Javanese during the late period of the reign of Kṛtanagara, it is likely that East Java witnessed a new influx of Buddhist ideas, iconographies, and scripts as a result of maritime relations precisely in that period (Acri and Wenta 2022; Acri 2016, 2024).

As for Śrīvijaya, by the end of the 13th century it had become historically irrelevant. In 1288, it was absorbed into the Siṅhasāri domain and attacked again by its successor, the Majapahit kingdom in 1293. Yuan notices list that the Ma’bar coast, which had been the Cōḻa coast earlier, strikingly omitted ports in the Strait of Melaka (Mukherjee 2019, p. 191). At this time, the connections in the South China Sea enabled numerous diffusion centres and the circulation of goods and religious art styles, supporting the paradigm of these ideas (Figure 18).

13. Conclusion

This study has focused on the implementation of pearl ornaments in a particularly unique style, which only appears in these three locations in the brief period between the 11th and mid-13th centuries. All the relevant artworks discussed here owe their overwhelmingly stylistic merits to Pāla and “related art styles”. My analysis has highlighted a convergence of cultural forces and ideas in the vast region spanning South through Central, Southeast, and East Asia,36 with the Pāla links across the Chinese, Mongolian, and Southeast Asian regions where the Pāla and “related art styles” are evident. While the links between Kṛtanagara and Khubilai Khan may appear unsound and indirect, this is not the case as Khubilai and Kṛtanagara were engaged in a diplomatic and then military confrontation. Kṛtanagara’s adoption of Tantric Buddhism was in imitation of a response to Khubilai’s initiation that requires the assumption that Tantric Buddhism offered effective power, spiritual, political, and military power (Bade 2016, p. 153). Acri and Wenta proposed that “accepting the widespread scholarly view that Kṛtanagara, in the final period of his reign, imitated Kublai’s tantric initiations and adopted tantric technologies practiced at the coeval Yuan court to protect his kingdom from external aggression and consolidate his regional power” (Acri and Wenta 2022, par. 68). Despite the plausible historical scenario suggested by these views and the circumstantial evidence supporting them, what I am suggesting here is the possibility that the art styles of the two regions under discussion were not the result of any direct connections between Khubilai or Kṛtanagara and the trade between China and Java during the mid-to the late 13th century, but rather the result of a shared common (Pāla, or Pāla-influenced) source and shared styles of pearl ornamentation, that influenced various regions of the Buddhist world.

I argue that the East Java Prajñāpāramitā in the 13th century and the Tibetan Pāla-inspired thangka paintings often show a convergence of cultural forces and a commonality of style originating from the maritime realm and traded via maritime routes, both by sea and overland. Thus, I propose that the connection between the Vajravārāhī and the other Tibetan thangka paintings was all inspired by Northeast Indian influence from the Hexi corridor, eventually reaching East Java. As Pal suggested, these Pāla-period paintings were made in Tibet by masters hailing from the Indian subcontinent, and I propose that Javanese artisans made the stone statues using them as models. It is without a doubt that the textile patterns and the ornament of Pāla-style metal jewellery, with the extensive use of pearls, in the triangular armbands and the upavīta, are of a Pāla style originating from Tibet reinterpreted with an indigenous Javanese aesthetic (Pullen 2021a).

What we see here are over 200 years of parallels in Mahāyāna Buddhist art styles from the Vajravārāhī in Xixia style, which reflects Pāla art styles, Pagan mural paintings, and the Buddhist sculpture in East Java and Sumatra.37 It is apparent that Buddhist “art styles” travelled from Central Java and eventually arrived in Tibet in the form of textile patterns (Pullen 2021a, p. 21). However, regarding the 13th-century East Java Buddhist statues, which present clear examples of pearl ornaments, we can surmise that these ideas were transmitted by pilgrims and artisans with an internalised knowledge of Buddhist “art styles”, reflecting the Pāla idiom.

Studies of iconography rely on textual sources; however, some visual features are not featured in the texts. The production of images is a creative process, not always following strict religious guidelines and generally meeting each region’s spiritual needs. Monks interacted with foreigners and were open to new ideas, which travelled both ways to and from the Indian subcontinent. Therefore, the images exported by monks reflected the concepts from eastern India, but they merged and reappeared with the local Javanese aesthetic (Bautze-Picron 2016, p. 175), hence the commonality of styles.