Introduction: Mahākāla, Hinduism and Buddhism

1

Mahākāla, the Great Black One, is a nebulous character who has been worshipped in several Asian religions. His widespread acceptance is at the same time a potential cause for confusion about the affiliation of particular instances of the deity. Before proceeding to identify some historically significant Southeast Asian images as the Buddhist Mahākāla, it would be appropriate to sketch out his early trajectory. At first, he was the popular patron deity of the ancient city of Ujjain, located in present-day Madhya Pradesh. As a figure whose prestige did not rest on the authority of the Vedas, he was positioned to be recognised in post-Brahmanical creeds.

2

With the co-option of the deity into Śivaism—a process first outlined by Phyllis Granoff (2004, 105–23), and now elaborated by Péter Száler (2021, 90–93)—Mahākāla joined Śiva’s retinue. He then came to be identified with Bhairava, the “terrifier” who decapitated the god Brahmā for Śiva. Later Mahākāla was identified with Śiva himself and given a role in the creation of the universe. In a separate tradition, the stature of the dark god is reflected in a Jain legend of the monk Kālaka, who was venerated for helping restore the Jina’s teaching in Ujjain (Száler 2021, 97–99). The deity was also venerated in the Vaiṣṇava Pāñcarātra and, much further afield, in a Sogdian Christian text that translates the name of Apollo as Mahākāla (Sims-Williams 2021). Buddhists, for their part, embraced Mahākāla at an early stage as one of the many spirits and lesser deities who respected the Buddha’s religion, while remaining aware of Śivaites’ appropriation of the Great Black God (Granoff 2004, 106n29). Mahākāla is present in all phases of Buddhist Tantrism, from proto-tantra up to the yoginītantras of the late first and early second millennia. It is this latter, final phase of Mahākāla’s development that becomes prominent in late Hindu-Buddhist polities and which will be investigated in this study.

3

As a focal point of interreligious coactivity and contestation, the Great Black One presents an ideal if overlooked tertium comparationis for the study of Buddhism and Hinduism in their Tantric modes. The common property handed down in these religions, the traditum, comprises Mahākāla’s name, eponymous dark image, and function as a protector of hallowed ground. He is seen as a heavy, black, grimacing gatekeeper wielding tools of slaughter. Over time, this figure manifested in various forms and contexts, and many of these forms were described in writings affiliated to discrete, well-defined religious movements. Nonetheless, the manifestations of Mahākāla retained a common ‘look and feel.’ Mahākāla somewhat resembles Bhairava cum Śiva, who in turn looks like other krodharūpa deities, who may or may not display sect marks. Images of the Great Black One therefore had the potential, in a shared socioreligious environment, to foster congeniality and spiritual solidarity where there was little or none to be found in individual religious traditions. The challenge here is to discern the religiosity underpinning images that are said to have generic or multiple sectarian identities, but which come out of a milieu that expressed clear interest in religious difference.

The Vajramahākāla Depiction on the Mandākinī Sword

10

The oldest portrayal of the Buddhist Mahākāla in the Malayo-Javanese world is found on a short sword called Si Mandang Kini in Malay. The sword’s name is spelled in various ways,¹ all of which are cognate with the Sanskrit word Mandākinī. This is the classical name of a river flowing up to or through the heavens, and it often designates the Milky Way galaxy in particular. The name Mandākinī alludes to the blade’s dark and shiny appearance,² and signifies in a more general way that the sword is of heavenly origin. Mandākinī, as it will be called here, is known for several centuries to have been part of the regalia handed down in the Sumatran highlands (Drakard 1999, 240; Tjoa-Bonatz 2020, 35). Belief in the divine provenance of the sword continued long after its custodians, the Minangkabau royal house, abandoned Hindu-Buddhist rule.

11

The sword depicts two different figures, each engraved on a different side of the blade and outlined in gold leaf—“gouden figuren geincrusteerd,” as F. D. K. Bosch (1931, 211) first described them. The male deity on one side is plump, garlanded and standing upright (figure 1). He is clad in a tight waist-wrap garment and his hair is tied up in a bouffant. His right hand dangles a meat cleaver by his waist, and his left hand holds a skullcup in front of his chest. On the other side of the blade is the figure of a thin, naked woman with long loose hair, who likewise holds a cleaver and a skullcup. The thunderbolt handles of their cleavers establish immediately that they are associated with Tantric Buddhism.

12

Various interpretations of the male figure on the sword have been put forward. He is often identified as a Bhairava, “fearsome” Śiva, but this identification has not been tied to particular Śivaite sources. Rather, the term bhairawa has remained current in Java as a non-specific term for “fearsome” figures (Acri 2019); its imprecise use in the secondary literature will be discussed later in this article. The male figure, in the view of F. D. K. Bosch, was not only a Bhairava, but also a portrait of the king Ādityavarman. Bosch pointed to phrases in Ādityavarman’s inscriptions that purportedly implicated him in the “diabolical” ritual of “bhairawa” religiosity (1930, 213). The joint identification of the sword’s male figure with Bhairava and this king has become widely accepted (Reichle 2007, 208). However, Ādityavarman was not the only ruler of the region who is associated with Buddhism and Tantrism. Furthermore, the main documents of Ādityavarman’s reign, his epigraphs, have still not been read satisfactorily (Tjoa-Bonatz 2020, xxii), despite one recent effort to study them as a corpus (Kusumadewi 2012). There is also the important detail that both figures on the sword stand on lotuses, which indicates that they have otherworldly rather than worldly status.

13

The iconography of the male figure has instead been found by Claudine Bautze-Picron (2014, 113–14) to “clearly refer to (the Indian Buddhist) Mahākāla and not to (the Indian Śaiva) Bhairava,” such that we are to “reject the name Bhairawa” and “disregard the identification with Bhairava (or with Ādityavarman and his queen).” The identity of Mahākāla is claimed on the basis of visual similarity with a statue from Lakhisarai, a site near Indapaigarh in Bihar, northeastern India, which has been identified as a two-armed Mahākāla (Bautze-Picron 1991–1992, fig. 16). A Buddha figure at the top of this statue leaves no doubt as to its religious affiliation. The Lakhisarai statue displays strong and definite similarities with the figure on the sword, but the two are not iconographically identical. The large trident that rests in the crook of the Lakhisarai statue’s left arm, for instance, does not appear in the Mandākinī figure. Images of the two-armed Mahākāla with a trident or skull-staff (khaṭvāṅga) represent a divergent form of the deity. Another Indian example of this form is a sculpture in the collection of the Indian Museum, Kolkata, accession no. A22350.³ However, most of the South Asian depictions of the staff-bearing form are Nepalese and can still be found in situ at the entrances of Buddhist monasteries.⁴ There are literate sources for the two-armed, trident-bearing form. It is described in at least one sādhana of Mahākāla (published as Sādhanamālā 302) and in a hymn attributed to pseudo-Nāgārjuna, which circulated widely under titles such as Vajramahākālastotra. Pseudo-Nāgārjuna’s hymn was translated into Tibetan at least five times (Verhagen 2017, 175–77) and once into Chinese (Hou 2022, 413). The trident-bearing form is at least as old as its prose description in the Kriyāsaṃgrahapañjikā (Tanemura 1997, 22), a handbook for consecrating sacred objects compiled in mid-eleventh century Nepal. What this material represents is a common visual tradition of the two-armed Mahākāla that is largely similar to the sword depiction yet critically different from it, and which is concentrated in the northeast of the subcontinent.

14

Bautze-Picron (2014, 113) further draws attention to the “northeast Indian influence” on the art of the Siṅhasāri period that other scholars have observed, and to the fact that the form of Mahākāla on the sword is quite different to the earlier, slender Javanese gate-guardian figures identified as Mahākāla. The Buddhist Mahākālas of Indonesia are a more recent import from the subcontinent. However, Bautze-Picron’s suggestions (2014, 116) that this iconography might be associated with Atiśa (985–1054), who spent over a decade in the ‘Golden Isles’ region that includes Sumatra—among other places—only highlight the fact that many of the period remains in Sumatra are either not securely dated or clearly not of the early eleventh century, the period of Atiśa’s sojourn in the ‘Golden Isles.’ The ferocious deity depicted on the sword belongs to a period later than the early eleventh century.

15

A precise identification for the male figure on the Mandākinī sword is given here by referring to sources that have not previously been taken into consideration. This figure is Vajramahākāla, a two-handed form of the deity, whose classical locus lies in the fifteenth chapter of the Ḍākinīvajrapañjara. This Buddhist tantra, which is affiliated to the Tantric system of the Hevajra, was circulating by the end of the first millennium, judging from the first named references to it, which occur in the Hevajraprakāśa of Rāhulagupta—who is said to have taught the youthful Atiśa—and in Nāropā’s (d. 1040) Sekoddeśaṭīkā.⁵

16

Vajramahākāla is described in brief in the Ḍākinīvajrapañjara’s fifteenth chapter. Although the complete Sanskrit text of this tantra is not accessible at the present time, its contents are available in Tibetan translation and the passages on the deity have survived in their original language in derived works. The Sanskrit verses that describe Vajramahākāla’s appearance are conveyed in separate invocation procedures, sādhana, two of which have been published as text numbers 300 and 301 of the so-called Sādhanamālā. It has not been noticed before that the Ḍākinīvajrapañjara is the main source for these sādhanas. Their vision of the deity is articulated as follows:

17

Two-armed and single-faced, emanated from the syllable hūm̐,
Black in colour, with a great flaming [halo], holding cleaver and skull,
Decked with a garland of heads, with teeth frightening to fear [itself],
Corpulent in form and fervent, blood trickling from the mouth.⁶

18

The depiction on the Mandākinī sword conforms to this description. The omissions are inconsequential: there is no flaming halo, which is often realised in art as an ornamental—and therefore dispensable—framing element. The bared teeth and bloodied mouth are impractical to detail with thick lines at a small scale. The figure is otherwise recognisable in all respects as the two-armed Vajramahākāla.

19

A few words should also be said about the nomenclature for this figure. The correct designation is Vajramahākāla, the “Tantric Buddhist Great Black One.” This name is used in the prose introduction to Ḍākinīvajrapañjara 15.31–35 and is given as Rdo rje nag po chen po (*Vajramahākāla) in the Tibetan translation; the corresponding sentence in Sādhanamālā 300, which is extracted from this part of the Ḍākinīvajrapañjara, likewise uses the name Vajramahākāla. This is all very clear. If for some reason a Tibetan idiom is applied to the Sumatro-Javanese context, the properly corresponding figure in the Gur mgon section of the Tibetan pantheon is not Gur gyi mgon po, who has the artificial title *Pañjaranātha and a divergent iconography,⁷ but Gur rkyang Rngog lugs, i.e., “the Mahākāla of the Ngog tradition of the Pañjara,” who carries only a cleaver and skullcup like the Malayo-Javanese depictions (Willson and Brauen 2000, 348). But again, the images under consideration here represent no deity other than Vajramahākāla.

The Goddess Partnered with Vajramahākāla

20

Vajramahākāla’s thin, long-haired female counterpart, who is depicted on the reverse of Mandākinī’s blade (figure 2) (see Reichle 2007, fig. 6.32), has not yet been located precisely in the Asian pantheon. The Ḍākinīvajrapañjara, which sets the iconographic standard for Vajramahākāla, treats him as a stand-alone figure. There are no known depictions of the deity in South Asia that portray him with a female counterpart of the same stature. The identity of this partner figure is then to be sought in other writings related to Vajramahākāla, as well as in the body of directly related art, lore and praxis literature.

21

First, the identifications already proposed for the sword’s female figure will be reviewed. She has been seen as a representation of Ādityavarman’s consort and a “Bhairawī” (Bosch 1931, 214), or simply as a “daemonic woman” (Schnitger 1939, 27). Her dishevelled hair is an unusual feature that narrows the range of possible identities in the Tantric Buddhist milieu. Bautze-Picron states that loose hair of this kind is characteristic of Vajravārāhī, the partner of Cakrasamvara; Vajravārāhī is indeed frequently described as having “loose hair” (muktakeśā) in authoritative works of the Cakrasamvara Tantric system.⁸ However, the proposal that “Vajravārāhī, prajñā of Hevajra […] would be the most appropriate candidate here” (Bautze-Picron 2014, 113n30) is not viable. Vajravārāhī is the chief goddess in the Cakrasaṃvara system, and she is not Hevajra’s partner. Cakrasaṃvara and Hevajra are different hypostases of Heruka. Vajramahākāla belongs to the Tantric system centred on Hevajra, which incorporates the Ḍākinīvajrapañjara, and in that system Nairātmyā is Hevajra’s partner. She is normally depicted with flame-like upraised hair, dancing wildly or leaning to one side. The woman on the sword is neither Vajravārāhī nor Nairātmyā.

22

The goddess Kālī is identified as a companion of Mahākāla in several Tantric Buddhist texts, all circulating in the period of interest, the early second millennium. The aforementioned Mahākālasādhana (Sādhanamālā 300) calls for the deity to be visualized in the company of five yoginīs starting with Kālī. They are described as very playful (mahāvikrīḍita) and as holding cleavers and skullcups. The yoginīs are visualized at the climax of the procedure when the target of the sādhana is butchered. The same arrangement of goddesses is presented in versified form in the Vajramahākālasādhana of Karuṇācala, Sādhanamālā 303 (Bhattacharyya 1928, 588). Kālī is likewise the first and primary goddess of the yoginī cohort surrounding Vajramahākāla here. Another sādhana extant in Tibetan translation invokes Mahākāla “with his consort” (yum can), namely, Kālī (Nag mo); she holds a cleaver and skullcup and embraces her partner (Derge, 1758, 248a). A similar vision is described in the farther recesses of the Sanskritic corpus. The caryā song Hūm̐kāra sambhava lauds Mahākāla and Mahākālī “in embrace,” āliṅgita (Bhakti De 2012, 170–71, No. 148). In another work of Indic origin, found in Central Asia and preserved in Chinese translation, Mahākālī is lauded as Mahākāla’s partner (Hou 2022, 413). This pairing is, in short, established in various corners of the late Buddhist world, albeit in a relatively informal way. What has not yet been found is a canonical description of the goddess in the company of the deity that accords with the portrayal on the sword in every detail, such as the unbound hair. There are several hymns to Kālī in Tibetan translation which have been affiliated to Buddhism (see, e.g., Suzuki 1962, Nos. 4912–4918), as well as the Indic Mahākāla depictions that feature female company (such as the aforementioned Lakhisarai statue), which are left to be explored in a study of wider scope.

23

The pairing of Mahākāla with Kālī was then known to and accepted by Buddhists, even though it lacks direct sanction in scripture and it is not prominent in the artistic corpus. In this respect, this pairing appears to be a late and contingent accommodation of a non-Buddhist religiosity. It is in Hindu Tantrism that Mahākāla is most often partnered with Kālī or some such goddess—Mahākālī, Kālikā or Guhyakālī. This partnership may be grounded in the Śivaite myth that the Bhairava who decapitates Brahmā is shadowed by a woman who personifies this sin (Adiceam 1965, 24; Granoff 2004). In certain hymns of the Trika, known to have emerged by the early second millennium, Mahākāla complements Kālī in her capacity as an aspect of divine consciousness (Silburn 1975, 2–3, 11–12). However, this Śivaite “interiorisation” of Mahākāla in partnership with Kālī has but little parallel in Buddhism, wherein the Great Black God is often remembered as a lone protector of the locality, a kṣetrapāla. He is lauded as such in the refrain of pseudo-Nāgārjuna’s hymn (Verhagen 2017, 182–94). While his roar is said to extend throughout the trichiliocosm, the tribhuvana (2017, 183), he does not gain the transcendent status of sexually united numens such as Hevajra–Nairātmyā and Cakrasamvara–Vajravārāhī. The Mahākāla of Tantric Buddhism does not personify the religion’s ultimate values; he rather functions as their darkest, most implacable defender.

The Significance and Provenance of the Mandākinī Sword

24

The two figures depicted on the sword have now been identified as the Buddhist Vajramahākāla and the semi-Buddhist or Buddhicised Kālī. Although Kālī is not a particularly Buddhist figure, she is coupled with one on the sword, such that the culture of origin is determined to be at least half Buddhist. The significance of this pairing lies in the fact that each deity is a bloodthirsty defender of their associated religion. The rich mass of material in this vein can only be skimmed here. Kālī is famously depicted in stories of the māhātmya and purāṇa genres vacuuming up the blood spilled by the foes of the (Hindu) gods (Kinsley 2003, 25–26). As for Vajramahākāla, he is summoned with a mantra addressed to the “helper of the religion” (om̐ mahākālāya śāsanopakāriṇe), which is further specified as the “Buddha’s religion” (buddhaśāsana) in one case.⁹ His sādhanas invoke the two-armed form of the deity devouring adversaries of the dharma. Again, a key image goes back to the Ḍākinīvajrapañjara:

25

He who is ever hateful to the teacher, who is averse to the Three Jewels,
A violator of many beings, is eaten by Mahākāla.
Let him chop the flesh off the body, then drink the blood.¹⁰

26

Such cutthroat imagery pervades the late Buddhist vision of Mahākāla. The hymn of Pseudo-Nāgārjuna lauds the deity with gory epithets such as “drinking a stream of blood as an offering” (Verhagen 2017, 180). Another typical example of this imagery occurs in the hitherto unstudied Dvibhujamahākālastotra of Saddharmāditya. This work was not translated into Tibetan until the sixteenth century, and as it appears to have circulated in a period in which Buddhist institutions had long stopped functioning in India, it illustrates the remarkable persistence of Mahākāla as a carrier of Buddhist identity. Here the stotra calls on Mahākāla to

27

Chop, chop, chop the flesh of the wicked! Slurp the blood, fat and entrails-garlanded body!¹¹

28

Across the literature there is a consistent, vivid vision of a rapacious god who eats dismembered enemies and is praised for giving “help to the religion” (śāsane sthopakāram). The deity is always presented in these texts as being on the side of Buddhism, which undermines any notion of his images being ready-made for syncretic purposes. The sword emblazoned with images of Vajramahākāla and Kālī is then to be seen as a weapon that juxtaposes two openly combative figures from two Indic religions. The religious backgrounds of the deity couple bring the sword’s purpose into plain view: it is to shed the blood of the enemies of Buddhism and Hinduism.

29

The sword is designated chorek, curek or cheriga in Malay, corresponding to the Sanskrit word churikā. It is a close combat weapon designed to be thrust downward in an icepick grip. Studies of its age (e.g., Winstedt 1932; Damais 1962, 280–82) have identified comparable swords, or depictions of swords, dating from the ninth to the fourteenth centuries. These studies are, however, limited to Maritime Southeast Asia and have not identified other possible periods or places of origin. Similar swords are in fact seen in eastern Indian sculptures of Mahākāla or lookalike figures produced between the eleventh and early thirteenth centuries. The Eastern Indian statues of four-armed ferocious deities in the Patna Museum (Huntington 1984, Pl. 150) and the Asutosh Museum of Indian Art (Mitra 1959)—probably depictions of Mahākāla—are cases in point (figure 3). The swords in these sculptures resemble the Mandākinī sword in certain details, such as the leaf-shaped blade that fans out at its base and the ornate disc-shaped pommel. Many other examples from the subcontinent could be given. The design of the sword as a whole, together with its engraved figures, can then be located in the northeastern Indian Tantric Buddhist milieu. This unique piece of exotica is surely the original Mandākinī sword of ‘Hindustan’ referred to in Malay sources, and the predecessor of two other royal swords in the Malay world that bear the same name.¹² This finding establishes that the Vajramahākāla depicted on Mandākinī is the earliest instance of the deity in Southeast Asia identified so far, and the only one of direct Indic provenance.

30

The foreign provenance of the Mandākinī sword is also asserted in the region’s chronicles. Mandākinī is said to have been introduced to Sumatra by a ‘Hindustani’ king with the Malay name Sang Sapurba and the Sanskrit title Tribhuvana. According to the Sejarah Melayu literature, king Tribhuvana settled some time after his arrival in the Minangkabau highlands (Winstedt 1938, 3, 123)— that is, the vicinity of classical Dharmāśraya. His floruit there is about the late thirteenth century, tracing backwards from the chronology of his fourteenth-century successors (Linehan 1947, 127). The Tribhuvanarāja who is said to have brought the sword to Sumatra is not an invention of the Melayu chronicles; he has been identified with the Tribhuvanarāja who is addressed in the 1286 pedestal inscription of the Amoghapāśa statue installed at Dharmāśraya (Drakard 1993, 306–7; Sinclair 2019). The claim that this king Tribhuvana owned the sword in the Minangkabau regalia was accepted at face value by R. O. Winstedt, who had commented on (1932, 160–1) the photographs of the sword published by Bosch, and who was also familiar with the lore of the sword narrated in the Sejarah Melayu (1909, 74, 1938, 2). The sword’s antiquity is accepted on different grounds by Jane Drakard (1999, 240–41). The sword lore in the Sejarah Melayu is paralleled in the “scarcely less venerable” (1999, 241) seal letters of Minangakabau rulers, which provide old and independent written references to the sword’s existence. There is now also a compelling case to be made for the sword’s antiquity and foreign origin on the basis of the classical Indic style of the sword and the deity depictions engraved upon it.

The Monumental Vajramahākāla of Dharmāśraya

31

The monumental statue excavated in 1906 at Padang Roco in Dharmāśraya is another depiction of Vajramahākāla that has not previously been recognised as such. For several decades it has stood together with its base, 4.41 metres high in all, at the National Museum of Indonesia in Jakarta. The statue combines imposing bulk with exquisite detail in a singular image of superhuman power (figure 1). The statue has the same basic iconography as the male figure on the Mandākinī sword, and similarly, the identifications proposed for the figure on the sword have also been applied to the statue. Again, the statue is described both as a portrait of Ādityavarman and as a Bhairava in official documents and most secondary studies. And again, much the same objections can be raised: there is no specific evidence of Ādityavarman being involved in its creation, and the statue has clear markers of Buddhist identity—including an image of the Buddha Akṣobhya—that preclude it being identified primarily as a Śivaite Bhairava.

32

Before establishing that the monumental statue of Padang Roco is another Vajramahākāla, and not Bhairava, it should be reiterated that the name Bhairava has only been applied to the statue as a generic, scripturally unsourced identification. The fact that the statue was identified as Mahākāla when it was discovered is disregarded in the bulk of the secondary literature referring to the “Bhairava” or “Buddhist Bhairava” statue, which is too voluminous to review here (for some references, see Bautze-Picron 2014, 115–16). The excavation report of Pleyte correctly identified the two major statues of the Padang Roco as Amoghapāśa and Mahākāla, and as Buddhist creations (1907, 172–73, 176). Yet Pleyte also ventured, without elaboration, that “terwijl het tweede alle kenmerken draagt van een Brahmaansch Buddhistischen Kâla (the second bears all the marks of a Brahmin Buddhist Kāla),” paving the way for lasting confusion about the statue. To account for these contradictory statements, it should be clarified that the accurate part of Pleyte’s identification was grounded in the then fast-growing state of knowledge of the Buddhist pantheon. Illustrations of the eight-armed Amoghapāśa together with various forms of Mahākāla, all in Indo-Himalayan style, had been disseminated in Albert Grünwedel’s widely read Mythologie des Buddhismus (1900, 129–30, 70, 175), not to mention earlier publications on the pantheon such as that of Bhagvanlâl Indraji (1879, Pls. XXV, XXIX).¹³ The initial identification as Mahākāla was therefore not discarded by later generations because it was poorly informed. Instead, the term “Bhairawa” was uncritically accepted as a generic designation for ferocious deities in spite of its inaccuracy and a lack of awareness of its sectarian character. J. L. Moens’ influential use of the term “Bhairawa” in Javanese studies as a catch-all term for ferocious deities (1921, 514) draws on a nineteenth-century Nepalese Buddhist work that not only features novel language with no bearing on the thirteenth-century period; it does not use the word bhairava at all.¹⁴ Later studies that follow Moens’ whimsical view of discrete Buddhist Tantric deities as indistinguishable Bhairavas continue to repeat this misunderstanding.

33

Key features of the monumental statue conform to the iconographic tradition of Vajramahākāla stemming from the Ḍākinīvajrapañjara. The statue is one-faced and two-armed (dvibhujam ekavaktriṇam), of “dwarven,” thickset proportions (kharvarūpam), and “fervent” (mahātejam) in its demeanour. The teeth, “frightening to fear itself” (daṃṣṭrābhīmabhayānakam), are not clearly discernible now, but some descriptions of the statue report canine shapes within the statue’s parted lips (Reichle 2007, 174). A cleaver and skullbowl are held (kartrikapāladhāriṇam) at the heart. This pose differs from the figure on the sword, which dangles the cleaver at the waist on the right side. The monumental statue draws up the hands at the chest, allowing the main figure to fit within the vertical nimbus and for the sculpture as a whole to be more balanced—an important consideration at this scale. The upraised hair bouffant, facial hair and snake ornaments depicted on the statue are not specified in the Ḍākinīvajrapañjara, but they are mentioned in various works on Mahākāla¹⁵ and can be regarded as part of the deity’s usual appearance. Although the figure of the Buddha Akṣobhya in the crown is not mentioned in the extant Sanskrit sādhanas of Vajramahākāla, it is referred to directly in a few works of this genre that survive in Tibetan translation.¹⁶

34

The only feature specified in the Ḍākinīvajrapañjara tradition that the statue is missing is the garland of heads (muṇḍamālākṛtahāram). The statue is instead surrounded by a semi-circular ring of skulls that forms its pedestal. This intriguing feature is not mandated in any known description of Mahākāla, but it appears in other Siṅhasāri sculptures (Reichle 2007, 179–85). The Siṅhasāri skull pedestal is always comprised of eight skulls, and this regularity points to it being fashioned according to some directive. In this regard, there is the expression “head ringed by skulls,” kapālāvalimastakam, which is used in the sense of a crown of skulls in Sādhanamālā 304 (Bhattacharyya 1928, 590), an invocation manual for the six-armed Mahākāla. It might be asked whether this expression was misunderstood in this context as an “eightfold ring of skulls,” *kapālāvalim aṣṭakam. Another sādhana of Mahākāla (Sādhanamālā 302, Bhattacharyya 1928, 586) describes, not dissimilarly, the deity as muṇḍamālāveṣṭitam, which if understood as “enveloped by a garland of heads” might also be interpreted as a circular plinth of skulls. If any such misreading of sādhana text was involved in the creation of these images, the skull plinth would then have been associated with Mahākāla at its point of origin in Siṅhasāri court art.

35

The placement of the figure of the Buddha Akṣobhya in the statue’s hairdo rather than the crown is another oddity with a possible origin in a misread text. It is consistent in only a literal sense with a directive such as “upon the upraised orange hair” (ūrdhvapiṅgalakeśopari). The position of the deity’s crown is specified in this way in Sādhanamālā 301 (Bhattacharyya 1928, 586). The possibility of misreading is increased by the fact that at least one other Siṅhasāri sculpture displays a quirkily interpreted iconographic directive. The eight-armed Amoghapāśas of the period hold three rods instead of the prescribed tridaṇḍa, a staff with three branches (Sinclair 2022, 30; see Reichle 2007, 100, Fig. 4.16, 118). On the whole, the features of the monumental statue, when compared with the textual prescriptions for Vajramahākāla, show not only agreement with sources for the two-armed form; they also, in the areas of difference, reveal potential misunderstandings of textually codified models.

The Monumental Vajramahākāla as an Embodiment of the Ruler

36

The notion that the monumental statue portrays Ādityavarman has been justified with reference to the environment in which it was found. Its findspot, Padang Roco in Dharmāśraya, is close to that of the Amoghapāśa statue created in 1286 and reinscribed by Ādityavarman in 1356. There are also again the perceived resonances between the fierce visual imagery of the statue and the verbal imagery of Ādityavarman’s Amoghapāśa epigraph. But if Ādityavarman had this deity or even this statue in mind when composing his 1356 epigraph, the possibility that the monumental statue had been in place beforehand is by no means excluded. The Vajramahākāla and the 1286 Amoghapāśa statue found in its vicinity both have clear affinities with Siṅhasāri artistic production. The monumental statue’s ensemble resembles in special detail—its jewelled belt, the shape of its Kāla-head clasp, its diamond-patterned fabric—that of a smaller, damaged Vajramahākāla statue preserved at a Siṅhasāri foundation, Candi Jago in East Java (Reichle 2007, 169, 199 Fig. 6.28) (figure 2). The Vajramahākāla at Candi Jago differs from the monumental statue in that it dangles the cleaver by the right side, a feature that aligns with the early Mandākinī depiction of the deity. The monumental Vajramahākāla then appears to be a later iteration of a type that had already reached the Siṅhasāri royal domains and was, therefore, known to Kṛtanagara.

37

It can now be shown that the monumental Vajramahākāla doubles as a portrait of Kṛtanagara. This is demonstrated through a comparison with a known representation of the king. The face of the monumental statue appears to be practically identical with that of the famous East Javanese statue Joko Dolog. This is the name given to the large inscribed sculpture of a stocky, seated, shaven-headed mendicant found in the former royal precinct of Trowulan, and which now sits in Apsari Park in Surabaya (Reichle 2007, 23). Joko Dolog’s inscription carries a date equivalent to 1289. Its letterforms are so much like those of the 1286 Amoghapāśa base inscription that they appear to have been produced by the same scriptorium. The statue is identified by its inscription as Jñānaśivavajra—an initiatory name of king Kṛtanagara—“in the form of the Great Buddha Akṣobhya,” mahākṣobhya.¹⁷ As the statue lacks the distinctive head-crest (uṣṇīṣa) of a Buddha, it may have been installed with its head covered by a separate gilt crown. Crowns and ornaments are typically worn by a Tantric Buddha, and this mode of appearance may be indicated by the unusual term mahākṣobhya (which is interpreted differently by Nihom 1986, 494). The objection that Joko Dolog is not a portrait statue because it does not follow the conventions of East Javanese posthumous royal portraiture can be countered by pointing out that it is not a posthumous portrait statue, but a likeness created during the lifetime of its subject. The closeness of Joko Dolog to the royal person is affirmed by its initial emplacement in a cemetery in the king’s precinct. The statement in the statue’s inscription that identifies its installation site as a cemetery has been corroborated by the discovery of human remains nearby (Lua 2019, 22).

38

The alert, rounded eyes, full lips, broad trapezoid nose and facial proportions of Joko Dolog are mirrored on the monumental Vajramahākāla. A photographic comparison of the two statues’ faces is presented here for the first time together with a composite image (figure 3, 4). This visual composite shows that the facial proportions of the two statues match almost exactly. The sculptor of one has copied the other assiduously, perhaps by a technique such as pointing. It is Joko Dolog that is most likely the exemplar of the shared facial physiognomy. As a self-styled likeness of a living king, it was an important piece of sculpture that was itself copied at least twice (Reichle 2007, 28, 48, Figs. 2.5, 2.11).¹⁸ The identity of facial physiognomy in the two statues gives firm support for the proposition that the monumental Vajramahākāla was created in the image of king Kṛtanagara, a proposition that was floated noncommittally in earlier scholarship (Schnitger 1939, Pl. III et al.; see Acri and Wenta 2022 for additional references). These statues can also be linked conceptually. The earth-touching, crowned Buddha image above the crown of the monumental Vajramahākāla represents Akṣobhya cum Joko Dolog. As has been pointed out earlier, the Buddha in the deity’s crown is a textually endorsed arrangement, even though it is not specified in the authoritative Ḍākinīvajrapañjara tradition.

39

The sculpting of the monumental statue evidently required access to Joko Dolog. The enormous Vajramahākāla would then have been shipped together with its skull base all the way to the Sumatran highlands. Such an undertaking is reckoned, however, to have been prohibitively difficult in the remote mountainous terrain (Edwards McKinnon, email, 2021). This presumed practical difficulty dampens any suggestion that the monumental statue was created, erected or worshipped in Java. In any case, the sculpting of the monumental Vajramahākāla would have taken place after the Joko Dolog statue was completed in 1289. The installation of the monumental statue in Sumatra then probably occurred after Siṅhasāri control over the whole Melayu region was realised in 1292, according to the Old Javanese Pararaton (Brandes 1896, 110). Work on the statue may also have continued after Kṛtanagara’s death that year. In all, the monumental Vajramahākāla was most probably created around 1292, if not at the Siṅhasāri court in Java then in the Sumatran highlands and with the aid of some kind of replica of Joko Dolog’s face. From the perspective of the Siṅhasāri court, Mahākṣobhya-Jñānaśivavajra, represented by Joko Dolog, would be the primary likeness of the ruler and master; Vajramahākāla-Tribhuvanarāja-Kṛtanagara, represented by the monumental Padang Roco statue, would be the slave.

40

Kṛtanagara’s almost pharaonic project of having a statue made in his own image and installing it in the Sumatran highlands, located over a thousand miles from his court, can be rationalised only briefly here, as this study is focused primarily on iconography. A longer discussion, presenting different perspectives on Siṅhasāri court religion, is provided by Andrea Acri and Aleksandra Wenta’s article in the present special issue (2022). Kṛtanagara was, in short, the foremost public representative of the Hindu and Buddhist religiosity patronised in East Java. The Sanskritic term for this joint religiosity, *śaivasaugata (Old Javanese sewasogata), had come into use before the Siṅhasāri dynasty was founded and continued long after it ended. In epigraphs this term usually accompanies or qualifies a word for Brahmanism, such that it designates a bipartite or a tripartite religious polity of Śivaites, Buddhists and Brahmins or ṛṣis (Sarkar 1934). Kṛtanagara was declared to have been a śaivasaugata ruler during his own reign—for instance, in his 1292 Cāmuṇḍā inscription (Blom 1939, 137). He is similarly remembered in the Pararaton (Brandes 1896, 40, 41) and so on. The king was also a prominent aficionado of Tantric religion, and was more than capable of directing the sophisticated iconographic program under discussion here.

41

What is important to note in the context of period Hindu-Buddhist ecumenism is that Kṛtanagara recognised a coreligionist in his neighbour and contemporary, the Tribhuvanarāja of Dharmāśraya. This much is evident from Kṛtanagara’s 1286 Amoghapāśa pedestal inscription addressed to Tribhuvanarāja and his people, which directs persons of the four (Hindu) castes to worship the (Buddhist) bodhisattva Amoghapāśa (Sinclair 2022, 32). Tribhuvanarāja was therefore seen from the outside to have personified the coalition of Tantric Buddhism and Hinduism that is also evoked, independently, by the two deity depictions on Mandākinī, the king’s personal sidearm. The monumental Vajramahākāla in this way embodies the ideologies of two kings and two polities, not just one.

Parameters of Hindu-Buddhist State Imagery in Insular Southeast Asia

42

The state ecumenism upheld by the Hindu-Buddhist kingdoms of the region, according to one recent analysis, entails “equality of… ultimate goals—yet not [of] paths” (Acri 2015, 267–69). This précis is by far the one that best fits the picture unfolding here. It should further be emphasised that the sources for Hindu-Buddhist religiosity almost always relate to the situation of kingship—the courtly kakawin literature, the prasasti epigraph and so on—and hardly ever to the works internal to each religion. The conceptual threshold for syncretism in these polities is never met, as Tantric Buddhism always remains a distinct entity in the view of the court and, of course, of its own proponents. If the polity of Siṅhasāri cannot be shown to be an indistinct socioreligious amalgam, then it is difficult to claim that the monumental Vajramahākāla commissioned by the Siṅhasāri ruler—which agrees in all key respects with particular Tantric Buddhist sources, and with those only—was conceptualised as a “Buddhist Bhairava” hybrid.

43

The possibility that the Siṅhasāri court encouraged Brahmins and Śivaites to worship Vajramahākāla cannot be discounted (see Acri and Wenta 2022, 42), especially in view of the precedent set out in the Amoghapāśa inscription, and other contact scenarios sketched later in this article. Yet the monumental statue displays no obvious Śivaite features. Purely ornamental details such as the half-moon and skull motif patterned on the sarong of the monumental Vajramahākāla, which is speculated to have been a concession to Śivaism, are immaterial to the iconography of the deity codified in literate sources. This pattern is revealed only in close-up photographs (Reichle 2007, Fig. 6.5; Acri and Wenta 2022, Fig. 7) and would not have been noticed by worshippers in the dim light of a traditional shrine setting. The halfmoon–skull emblem is itself Tantric common property, given that it is mentioned by name in Tantric Buddhist texts.¹⁹ Furthermore, the globelike hair bouffant of the monumental Vajramahākāla is not, as has been proposed (Acri and Wenta 2022), an attempt to evoke the turbans worn by Śivaite ascetics. A recent study of portrait statues of these ascetics (Ghosh 2021) shows that their turbans were styled in a semi-realistic manner with well-defined wrappings, voluminous fabric and so on. These turbans are represented according to the conventions of portraiture and would not have been confused with the gravity-defying hair of a ferocious deity. The shape of Vajramahākāla’s bouffant conveys otherworldly character and follows widespread textual and visual conventions for the depiction of ferocious deities.²⁰ It corresponds in particular to sources on Mahākāla such as pseudo-Nāgārjuna’s hymn, which mandate “bulging” (uru) and upwards-rising hair (Verhagen 2017, 184). With these understandings, the monumental statue cannot be regarded as a syncretic icon at the level of formal representation.

44

Although Kṛtanagara’s court commissioned images of Bhairava proper, it is remarkable that these images have not been connected to Buddhism (Puri 1982–1983, 21), as would be expected in a syncretic politicoreligious context. A prime example is the four-armed, dog-attended, dancing Bhairava of Candi Singasari, now kept in the collection of the Rijksmuseum voor Volkenkunde, Leiden (see Blom 1939, 53–54; Reichle 2007, 185, Fig. 6.18; Acri and Wenta 2022, Fig. 5) (figure 6). The Bhairava depicted in this near-life-size statue holds a cleaver, skull, rattle-drum and trident. The statue was inscribed with the obscure name Cakracakra, which is yet to be located in classical sources. Perhaps the intended name was cakracakrī (in the sense of kula-, bhū- or samasta-cakra) or some such graphically minor variant of cakracakra. What should be clarified here is that the iconography of this Cakracakra is not that of a generic Bhairava. It conforms to a distinct Śivaite type called Baṭukabhairava or Bhairavakṣetrapāla, whose signature feature is a dog companion. The Cakracakra statue has most though not all of the features specified in the iconographic prescriptions for this type gathered by Marguerite E. Adiceam (1965, 28–29). None of Adiceam’s Śivaite sources require the deity to carry a cleaver and a skull, as the Cakracakra figure does. The way in which these two objects are held by Cakracakra, with the vajra-handled cleaver in an icepick grip (figure 6), very closely mirror the holding of the same objects by the Candi Jago Vajramahākāla (figure 2). The eight-skull plinth is another distinctive feature that ties Cakracakra and other Siṅhasāri sculptures (see, e.g., Reichle 2007, Figs. 6.12, 6.17) to the monumental Vajramahākāla.

45

It is possible that a more conformant iconographic description for Baṭukabhairava or Bhairavakṣetrapāla will be found in āgamas not studied by Adiceam, and that the iconography of Cakracakra will find stronger scriptural support. At present, however, it is this Śivaite statue, not the Buddhist Vajramahākāla, that would appear to have had its iconography modified to satisfy the presumed state prerogative of Hindu-Buddhist unity. We then find in Siṅhasāri art a visual example not of compromised Buddhism, but of a subtly Buddhicised Śivaism. Even if so, we are still left with the question of how to identify the principles that would allow a “Buddhist Bhairava” to be conceived by the state as a discretionary śaivasaugata icon.

46

The Cakracakra statue of Siṅhasāri can further be seen as another attempt to embody the features of Kṛtanagara (figure 5), albeit one that was not as expertly accomplished in this respect as the monumental Vajramahākāla. If it is accepted that this statue is an additional part-portrayal of the king, the normal mode of these portrayals then becomes clearer. Kṛtanagara sought to embody his śaiva-saugata persona discretely, in separate Śivaite manifestations and Buddhist manifestations, each represented in accordance with their respective scriptural injunctions—as far as they could be determined—and incorporating no blatantly extraneous elements. The supremacies of Śiva and of Buddha could be unified in the living divine persona of the ruler who defends these faiths, but not in hybrid forms improvised outside the parameters of accepted sacrality.

Vajramahākāla’s Indo–Himalayan–Indonesian Transfer

47

More light on Vajramāhākāla’s pathway into the region can now be shed by findings on the career of the Bengali paṇḍita Gautamaśrī (fl. 1248–1268). This paṇḍita’s name, East Indian provenance and Mahāyāna religiosity are recorded in an inscription at Karimun Besar island located near Sumatra. Gautamaśrī is known from other sources as an exiled Eastern Indian teacher and translator who once resided in the Gusatalavihāra of the Kathmandu Valley. The identification of Gautamaśrī as a common figure active in both regions rests on the hard epigraphic fact that he is referred to in the same way in the Karimun Besar and Gusatalavihāra inscriptions: paṇḍita… śrīgautamaśrī (Sinclair 2018). There is also strong circumstantial evidence that places Gautamaśrī, or someone with his distinctive expertise, in the two regions in the same period. I have explored his connection with the eight-armed Amoghapāśa elsewhere (Sinclair 2022, 30–32) and will focus here on his knowledge of the form of Māhākāla that reached Sumatra and Java.

48

Gautamaśrī’s expertise on Tantric Buddhist subjects such as the invocation of Mahākāla was transmitted to and recorded by Tibetan students in Nepal. A sādhana on the two-armed Mahākāla, i.e., Vajramahākāla, is among the texts that he is known to have studied. It was incorporated into an anthology of twenty-four invocations of ferocious deities called the *Devāntasādhanasāgara, which was conveyed to Tibet via Gautamaśrī in the late thirteenth century.²¹ This work, transmitted independently as the aforementioned Sādhanamālā 301, conveys the directives from the Ḍākinīvajrapañjara that underpin depictions of Vajramāhākāla in Sumatra and Java. Although Gautamaśrī is said to have passed on teaching associated specifically with the Ḍākinīvajrapañjara, it is more likely that the information on Mahākāla that travelled to Sumatra was expressed in the dedicated sādhana literature rather than the Ḍākinīvajrapañjara Tantra itself, given that Gautamaśrī is often associated with sādhana practice transmissions.²² The varied contents of the *Devāntasādhanasāgara show, incidentally, that ferocious Buddhist deities were not regarded as deindividualised instances of a generic type, and unsurprisingly Bhairava is not mentioned in this anthology except as a deity crushed underfoot.²³ This anthology combines strains of combative Buddhist Tantrism that were spread widely across the Indo-Himalayan region. In Nepal, Gautamaśrī would also have been exposed to Vajramahākāla in the deity’s aforementioned local role as a monastic gate-guardian. A statue of the deity is installed at Gautamaśrī’s former residence, Gusatalavihāra, although the current statue dates only from 1515 (see Vajrācārya 1999, 62).

49

It is clear that Gautamaśrī was exceptionally well placed to transfer the praxis of Vajramahākāla from the northeast of the subcontinent to Sumatra, and perhaps also from there to Java. Nonetheless, we lack information from within the region that would confirm Gautamaśrī as the agent of transfer. Although he was an important figure in the translocation of late Sanskritic Buddhism, the rapid uptake of Vajramahākāla’s cult at the highest level indicates the existence of a joint movement rather than a lone mission. The Mandākinī sword that bears the image of the deity, created in East Indian style, is not the kind of item that would be carried by a paṇḍita such as Gautamaśrī or a vajrācārya. It is not a ‘ritual dagger’ such as a kīla; it has a shape, size and weight that are appropriate for a combat weapon. Mandākinī is meant to be wet with barbarian blood, quenching the thirst of the ravenous Buddho-Hindu deities who empower it.

50

An alliance between the ruling class and the Hindu and Buddhist religions, with Mahākāla as a common point of reference, is already mooted in Nepalese period sources. There is a pertinent characterisation of the warrior persona in the Kapīśāvadāna, a Buddhist mythopoetic history of the monastic complex in Nepal where Gautamaśrī resided. In Kapīśāvadāna II.2–4 (ed. Both 1995, 116), the various states of rebirth (gati) are likened to various occupational types. To be born in the realm of the warlike asura is to be “a scoundrel full of ferocity and krodha who loves to get into battles,” who can be called Mahāhaṃkārajitkāla, “Great Darkness Conquering the Ego.”²⁴ However this compound may be parsed, it is evidently a synonym for Mahākāla (Both 1995, 124n19)—the Buddhist Mahākāla. The whole martial class is personified here by a deity who stands at a nexus of Hindu–Buddhist civilisation. It is no less remarkable that this equivalence is being forged in a text with the Buddhist canonicity of an avadāna.

51

A groundbreaking sense of rapprochement between Buddhism and Hinduism is further evident in the Kapīśāvadāna’s acceptance of the Brahmanical practice of sacred bathing (I.2, IX.28), of participation in non-Buddhist festivals such as the Navarātra (IX.13–14)—still celebrated in Nepal—and so on. Such conciliatory sentiments, rarely voiced before the crises of the thirteenth century, signal late Buddhist hopes for building a unified front against an unspecified external threat. They are epitomised by a vision of the king—“the Victor, the Tathāgata, together with ministers and citizens”—giving alms to the saṅgha in a parade around a city that honours “Brahmā, Viṣṇu, Indra, Śiva and the four communities (of Buddhist monastics and laypersons) in all quarters.”²⁵ The convivial public religiosity envisioned in this Nepalese text resonates with that of contemporaneous śaivasaugata polities such as Siṅhasāri or its successor, Majapahit.

52

The suitability of Mahākāla as an object of worship for Hindu-Buddhist royalty is made more explicit in the Kavitāvadāna, another Nepalese avadāna production. The Kavitāvadāna draws upon and extends the circa mid-thirteenth century Kapiśāvadāna (Both 1995, 24–27) and is close to it conceptually if not in time. The fifteenth chapter of the Kavitāvadāna, currently unedited and unpublished, tells of the ancient king Brahmadatta, who is “conversant with all religions and versed in the uses of all treatises.” The king lustrates Mahākāla on the fourteenth day of every dark lunar fortnight “in a Śiva temple in Varanasi.” At the same time he observes the overnight fast (poṣadha), the practice that is often promoted to outsiders in post-Indian Sanskritic Buddhism. One day, after a long peace, an army masses to attack Varanasi; the kind of army is indicated by its use of mules and camelry (auṣṭrāḥ). Brahmadatta summons Mahākāla and asks for eight magical feats (siddhi) that can be used in battle. He praises Mahākāla with eight epithets culminating in “He whose soul is Mahābhairava.”²⁶ The king wins Mahākāla’s siddhis, enters battle, shoots down the massed camelry and saves the city. The chapter then ends with the closing of the frame of the Buddha teaching Brahmadatta’s story to his disciple Śāriputra.

53

Although the propitiation of Mahākāla by a Hindu-Buddhist ruler is endorsed here at the universal level of buddhavacana, this practice—known in Nepal as the fourteenth-day vrata—is associated with a particular site in Kathmandu, the Mahākāla temple of Tundikhel (Vajrācārya and Vajrācārya 1989, ii). The temple houses Asia’s only other statue of the deity that is much taller than ordinary human height (figure 1). Its date has not been determined with accuracy, but it is at least several centuries old and could have been standing during Gautamaśrī’s time. The temple is still well known as a recipient of worship from non-Buddhists and Buddhists alike. The Tundikhel statue conforms to the prevalent iconography in Nepal, in which the deity bears a trident, but its towering presence is overall much like that of the Padang Roco statue. The description of Mary Slusser (1982, 292, pl. 480), could apply equally, apart from the odd detail, to both monumental statues:

54

Hieratically posed on a prostrate corpse, the obese and oil-blackened Great Black One is garlanded and crowned with skulls and carries his prescribed […] gruesome chopper and skull cup.

55

A consistent vision of state-sponsored Buddhism and Hinduism coming together around the figure of Mahākāla is then apparent in these late works produced in a Sanskritic polity on the Indic periphery. They provide just the kind of principled justification that would be needed to sanction the creation of images of bipartisan interest, and which has so far been missing from discussions of Hindu-Buddhist art in Indonesia. While there is no suggestion that the Kapiśāvadāna and Kavitāvadāna were preached outside their place of origin, they do capture a post-Indian Buddhist zeitgeist that urges unity in surviving śaivasaugata kingdoms and a recourse to a mutual Tantric religiosity. In such polities we see separate but juxtaposed Buddhist and Śivaite spheres aligned tacitly against an unidentified other. Syncretism and inclusivism are not observable in the thinking set out here, just as they are not observable in the Sumatro-Javanese Vajramahākāla images. Hindu-Buddhism in this period evinces a bilateralism in which the constituent elements remain distinct while staying in contact with each other. This is the ideal that the Tantric Buddhist Mahākāla embodies at this time and place and which, as has been argued here, is instantiated in the form of the Mandākinī sword and in the monumental sculpture possessing the visage of king Kṛtanagara.

Conclusions

56

The fearsome two-armed figures who are depicted on a sword and in monumental form towards the end of Hindu-Buddhist rule in Sumatra and Java have been conclusively identified here as Vajramahākāla. These images have been found to match in all key respects the distinctive descriptions of the deity circulating in particular Sanskrit sādhanas (the so-called Sādhanamālā 300 and 301), which have in turn been traced for the first time to a Buddhist tantra, the Ḍākinīvajrapañjara. Claudine Bautze-Picron’s view that there are no grounds for identifying these images as Bhairavas has been confirmed with reference to relevant literature from the Indo-Himalayan source milieu, further art-historical analysis and the history of studies starting with Dutch East Indies scholarship. The application of the Śivaite term Bhairava to the Sumatro-Javanese Vajramahākālas is inaccurate even in a generic sense, and it can now be seen to be rooted in unwarranted, unexamined assumptions about the nature of Tantric deities going back to the early twentieth century. Yet it has also become clear that the Tantric Buddhist Mahākāla occupied an unusual, liminal position at the bauddha–tīrthika divide, where he could be seen by outsiders as a Śivaite or pan-Indic “area protector” figure. In this respect he embodied a certain appeal to rulers of late Sanskritic polities who faced unprecedented threats from outside the Indosphere and welcomed cohesion within their borders. It is this dark spirit of alliance between two Tantric, war-ready religions that was conveyed to insular Southeast Asia as institutional Buddhism in India was in its death throes.

57

The transfer of Vajramahākāla’s cult started in the middle of the thirteenth century in the northeast of the subcontinent and reached its destination, Sumatra, at the same time as the arrival of two South Asians: Tribhuvanarāja and the Buddhist paṇḍita Gautamaśrī. It has been shown here that Gautamaśrī possessed expertise in a related Sanskrit sādhana of Vajramahākāla (*Devāntasādhanasāgara 23; Sādhanamālā 301), as well as proximity to a visual and literate culture in Nepal that championed the deity as a Hindu-Buddhist unity figure. Gautamaśrī can then be identified as potential agent of transfer of the Great Black God’s image and praxis to maritime Southeast Asia. In a separate but perhaps related transmission, Vajramahākāla was transferred to the region as one of two horrific haematophages depicted on the Mandākinī sword of Tribhuvanarāja. The sword’s pairing of Vajramahākāla with Kālī identifies it as a weapon for slaughtering opponents of dharmic order. The traditional claim that the sword was brought from India by Tribhuvanarāja to the Dharmāśraya province of Melayu-region Sumatra is supported by the Indic style of the deity depictions on the sword, the subsequent state-supported production of images of Vajramahākāla, the records of its provenance kept by its custodians and other circumstantial evidence.

58

Given that the Mandākinī Vajramahākāla represents the first image of the deity to emerge in this region and period, and that it is tied so closely to the xenarch of Melayu, Tribhuvanarāja, the deity must have remained associated for some time with this king’s lineage and religion as well as the Melayu polity. The subsequent creation of a monumental Vajramahākāla at the Siṅhasāri court seems to have aimed to co-opt and redeploy the śaivasaugata religiosity of this Tribhuvanarāja at a frontier of the Javanese imperium circa 1289–1293. As a photographic composite now confirms, the monumental statue embodies the triple divine persona of Vajramahākāla cum Tribhuvanarāja cum Jñānaśivavajra–Kṛtanagara. The statue in this way additionally signifies the absorption of Indo-Melayu (Dharmāśraya) into Indo-Java (Siṅhasāri), as well as the founding of a new Hindu-Buddhist superorder empowered to face a clash of civilizations. These combative Tantric artefacts appear to have been pointed towards a nameless transregional conflict, in which the Siṅhasāri dynasty was soon replaced by Majapahit and the first of many Sultanates came to power in Sumatra.

59

This study has also sought to clarify an aspect of the operation of the Hindu-Buddhist polity by analysing state-sponsored Vajramahākāla images. While the kings of these polities may well have sought to bring two religions together under one roof or around a single object of worship at court, direct interference by the state in sacred, sect-marked imagery cannot be discerned in most cases. One possible such intervention identified here does not concern the Buddhist Vajramahākāla depictions, but rather the Śivaite Cakracakra statue of Candi Singosari in Java, which takes after the former in a nuanced way. If there can be a “Buddhist Bhairava” in this milieu, this statue is by far the strongest candidate. The ecumenical governing principles of Hindu-Buddhism up to the fourteenth century need more clarification, but bilateralism has emerged here as a more definite priority than syncretism or blending, which is so far only observed in the Sumatro-Javanese realm in the exceptional circumstance of portrayals of the ruler. Although the widespread assumption that the monumental statue portrays Ādityavarman has now been disproven, it can still be surmised that Ādityavarman, a scion of the Majapahit court resident in Melayu, would have tied his authority to the powerful political significance and protective aura of the monumental Vajramahākāla.

60

There are still many questions left to answer regarding the sources, destinations and mechanisms of transfer in this part of the world in this period. In the present case, the textual and artistic corpora for Mahākāla and Kālī need to be more chronologically stratified and combed for more specific connections to the Malayo-Javanese context. The physical characteristics of the sword and monumental statue ought to be examined forensically, so that their extraordinary association with the region’s rulers can be evaluated with empirical methods. And the dynamics that led to changes not just of regimes but of multifaith civilizations across the precolonial Indo-Pacific will continue to need much more attention than they have received to date.