Introduction¹

1

A fascinating and still lingering problem in the religious history of Nusantara² is the question of what tantric Buddhist systems were adhered to by Siṅhasāri King Kṛtanagara (r. 1268–1292) of the Rājasa dynasty (East Java).³ This ruler, who has rightly been called an “empire builder,”⁴ held an important place in Javanese history, as suggested by the praise found in Mpu Prapañca’s fourteenth-century Buddhist-flavoured Old Javanese versified travelogue-cum-chronicle Deśavarṇana (or Nāgarakṛtāgama, 1960),⁵ and especially by the many significant East Javanese artworks and monuments associated with him, like the Buddhist Candi Jago (built by Kṛtanagara to commemorate his ancestors), the Śaiva Candi Singosari, and the syncretic royal mausoleum of Candi Jajava/Jawi. Kṛtanagara passed down to history as the king who engaged in a diplomatic and then military confrontation with the Mongol Emperor Kublai Khan (r. 1260–1294), the founder of the Yuan dynasty in China, who carried out an unsuccessful maritime invasion of Java in 1292/1293.⁶ No less importantly, Kṛtanagara is also regarded as a crucial innovator in religious matters, being the unprecedented promoter of a Śiva-Buddha cult located at the intersection of Śaiva and Buddhist domains, and centred around his own apotheosized persona (Hunter 2007; Acri 2015).

2

In this article we discuss some key aspects of the historical and religious background of the period of Kṛtanagara and his aftermath in East Java and Sumatra. Through a comparative study of Old Javanese, Sanskrit, and Tibetan textual sources as well as artistic vestiges, we intend to highlight the transregional networks of tantric Buddhism that may have contributed to shaping the king’s religio-political agenda. Taking as a point of departure the enigmatic colossal statue at Padang Roco/Sungai Langsat in Dharmasraya (Central-Western Sumatra) that has been variously called a Bhairava, a Mahākāla, or a ‘Buddhist Bhairava,’ we advance an identification of this remarkable vestige in the light of possible Indian and Sino-Tibetan prototypes and Siṅhasāri-period East Javanese statuary. Reviving Moens’ (1924, 557) idea of an association between the icon and Kṛtanagara, we discuss the statue’s origin, thereby complementing, and fine-tuning, the discussion independently provided by Sinclair (2022b) in this special issue. Adding to the discussion on the Eastern Indian-style icon of Arapacana Mañjuśrī from Candi Jago, we highlight further parallels that support the idea advanced by previous scholars about the commonality of the Buddhist paradigms practiced at the courts of Kṛtanagara and Kublai Khan. We then propose that their legacy was adopted by the political elites of the subsequent two or three generations, for instance Ādityavarman (fl. c. 1320–1378), who became chief minister in East Java and then King in Sumatra, and Yesün Temür (1293–1328) and Tugh Temür (1328–1332) in China. While confirming the view that the esoteric ritual systems adhered to by Kṛtanagara were centred around fierce tantric deities, including Bhairava, we argue that a specific Hevajra-related form of Mahākāla was part of the sovereign’s personal cultic pantheon in the final years of his reign.

3

The case study presented here contributes to the overarching issues to which this special issue is devoted, namely inter-religious contact on the one hand, and the transfer and localization of belief systems and practices across the medieval Buddhist world (i.e., the vast region including the Indian subcontinent, the Himalayas, East Asia, and Southeast Asia) on the other. Our comparative analysis and trans-regional contextualization suggests that the peculiar iconographic program characterizing the statuary produced under Kṛtanagara and in the aftermath of its reign—especially the use of a ‘charnel ground imagery’—should not be regarded as the expression of a fuzzy Javanese, Sumatran, or Southeast Asian ‘local genius,’ as previous scholarship has tended to do.⁷ Rather, this program reflects a politically driven adoption and reconfiguration along regional lines (as well as Kṛtanagara’s ‘syncretic’ attitude) of Indic religious motifs that may be traced to specific tantric currents, against the background of analogous and largely coeval developments attested in other areas of the wider Indic world.

Kṛtanagara’s Religious Agenda and its Political Ramifications

4

Following up on Moens’ (1924) remark that Kṛtanagara’s Buddhism was similar to the tantric Buddhism of Kublai, Berg (1965, 98–99) argued that the Siṅhasāri king adopted the particular form of Buddhism that he knew to be practised by the Mongol ruler in order to acquire supernatural powers to assist him in a military engagement with Java, as well as in his conquest of Tibet. Several scholars have subscribed to this view ever since.⁸ While these claims seem to be compelling in the light of the prevalent religious paradigms and political contingencies across the Buddhist world in the thirteenth and fourteenth centuries, and in particular the late phase of tantric Buddhism, the question of what specific tantric system(s) Kṛtanagara was initiated into remains unanswered. The Deśavarṇana (43.3) laconically refers to the fact that as he grew older, Kṛtanagara held to all sorts of (Buddhist) esoteric rites, including the gaṇacakra, which in its usual technical meaning indicates a transgressive ‘tantric feast.’⁹ Siṅhasāri-period art historical evidence as well as later Old Javanese sources suggest that the ruler, besides sponsoring icons of wrathful manifestations of both Hindu and Buddhist deities (the latter especially in the final years of his reign), implemented an eclectic—or properly syncretic—agenda merging the figures of Śiva and Buddha into a unified cult revolving around his persona, in a way that was as much religious as political.

5

The art historical record of the reign of Kṛtanagara is permeated by an imagery of a markedly horrific aesthetic character, which is associated with transgressive practices of the ‘charnel ground culture’ found in the Bhairavatantras and the Buddhist Yoginītantras. The most notable artistic remains include the ‘dancing Bhairava’ and the ‘tantric Gaṇeśa’ of Candi Singosari, the gigantic dvārapāla locally referred to as Totok Kerot, and a Cāmuṇḍī group dated 1292 now preserved at the Trowulan Museum. All these icons feature a demonic look and an abundance of skulls—carved in their basements, crowning their foreheads, and again adorning their bodies as colliers, bracelets, earrings, diadems, and sacred threads. As far as the textual archive is concerned, although both religious and belletristic literature that can be firmly dated to the Siṅhasāri period is scant (Hunter 2007, 32), a handful of (often remarkably well-crafted) Sanskrit hymns preserved in Balinese manuscripts mention fierce deities displaying iconographical features that are in harmony with those characterizing the aforementioned art historical vestiges, and are therefore likely to be roughly coeval (see Acri, forthcoming).

6

While Kṛtanagara’s fascination for wrathful divine icons might have been dictated by his personal credo and initiatory practice, which emphasized the encounter with the powerful, horrific, and transgressive aspects of the numinous, the cults associated with those icons are likely to have also served his political purposes. Foremost among these was the success of his expeditions against neighbouring polities, among which Malayu in Sumatra. But no less important was the protection of the realm from Kublai’s aggression. As remarked by Hadi Sidomulyo (2010, 98), quoting Hunter (2007, 38), the major shift in religious orientation detectable after Kṛtanagara’s accession to the throne “should probably be linked in part to the threat posed by the Mongol emperor Kublai Khan in the last quarter of the 13th century.” This concern seems to be hinted at in a passage of the Deśavarṇana (Canto 42.3) in praise of the sovereign:

7

tuhun nṛpati tan pramāda luput iṅ mada makin atiyatna riṅ naya
apan tәtәs i kevәhiṅ bhuvanarākṣaṇa gavayәn i kālaniṅ kalī
nimittaniran aṅrәgәp samaya len brata mapagәh apākṣa sogata
tumīrva saṅ atītarāja riṅ usāna magәhakәna vṛddhiniṅ jagat.

8

Nevertheless the King was not negligent, was free of intoxication, and was
more and more energetic in his policy
For he had realized how difficult it is to protect the world in the age of Kali.
This is why he held fast to esoteric doctrines and observances, and was firmly committed to the sect of the Buddhists,
In order to imitate the kings of old, and to guarantee the continued prosperity of the world. (trans. Robson 1995, 155)

9

Canto 43.1 adds that in the Kali age the “world became violent and turbulent. Only the divine being who concentrates on the six supernatural faculties of the Buddha could protect the world as god-king” (trans. Robson 1995, 155). While this statement may reflect the general attitude of what may be referred to as ‘Kali age Buddhism’ (Verardi 2018), here it is likely to specifically refer to Kṛtanagara’s political concerns and his commitment to esoteric Buddhist cults and apotropaic rituals, especially to counteract the power of the expansion-driven Yuan emperor.¹⁰ Indeed, in Canto 43.3–4, Prapañca describes the sovereign as “very energetic in the rites for application of magic” (Deśavarṇana 43.3–4).¹¹

10

Evidence of Kṛtanagara’s eclectic adherence to both Śaiva and Buddhist tantric systems for political purposes, and his interest in ‘war magic,’ is provided by a reference in the Sukhāmṛta copper plate inscription of 1296 (Poerbatjaraka 1940, 33–49) to the special patronage accorded by Kṛtarājasa (aka Raden Wijaya, r. 1293–1309) to Mpu Kapat, a priest of Śaiva persuasion adhering to the Bhairava observance, whose father already served Kṛtanagara in the same capacity. The inscription mentions the manifestation of Kṛtanagara’s “divine nature” (kadevātmakan) as a powerful means to terrorize and subdue his enemies, and specifies that the consecration name of this ācārya was astrarāja (“Chief of Weapons/Weapon-Mantras”),¹² which suggests that his duties included war rituals (Jákl 2020, 36). Hunter (2007, 35–36) argued that Kṛtanagara may have patronized the family of Patipati precisely because of their role as instructors in Bhairavika Śaivism, which assisted the ruler in the art of warfare.

11

Another association with martially driven magical empowerment is suggested by an unusual iconographical feature of the statues of deities facing the main directions at Candi Singosari (i.e. Nandīśvara and Mahākāla [east], Durgā [north], and Gaṇeśa [west]; the ascetic Agastya [south] being the exception), namely the short sleeveless “military jacket” or “tantric jacket” (kavaca), which according to Stutterheim (1936, 308–09) has a magically protective purpose (cf. Pullen 2021a, 41–42, 163). An analogous martial function may be detected in the Cāmuṇḍī group, probably originating from Candi Singosari, featuring miniature replicas of the Bhairava and Gaṇeśa standing on skull-pedestals found in the same temple. The back of the statue carries an Old Javanese inscription in Javanese script (except the first line, in Nāgarī script), dated to the last year of Kṛtanagara’s reign (1292), attributing the erection of the image to the satisfaction of the king at his victory over the entire world (digvijaya riṅ sakalaloka) and the entire archipelago (sakaladvīpāntara; Santiko 1992, 123). Hariani Santiko (1992, 216–17), having noted that in Bengal and South India Cāmuṇḍā was associated with vaśīkaraṇa (“subjugation”) rituals aiming at defeating enemies (śatrubali), hypothesized a connection between this sculptural group and analogous rituals directed towards Kṛtanagara’s archenemy Kublai Khan. Invocations to Cāmuṇḍā also occur in mantras associated with the tantric magical acts of vaśīkaraṇa and uccāṭaṇa (“expulsion”) in the Saṁvarodaya, a Buddhist Tantra from Nepal belonging to the Cakrasaṁvara cycle, in a context belonging to the cult of śmaśāna containing an invocation to Mahākāla (Tsuda 1996, 102).¹³ The sculptural representation of a Goddess that is otherwise unattested in Javanese art suggests that her martial character was aligned with the sponsor’s needs.

12

Association of monarchs with antinomian and martial traditions of tantric Buddhism was common from the twelfth to the fourteenth century and beyond across the Buddhist world: Besides Kṛtanagara and Kublai, as well as the Sumatran king of East Javanese pedigree Ādityavarman (see Hunter 2015, 324–27; Bautze-Picron 2014, 107, 123; Reichle 2007, 139), other prominent sovereigns of the Khmer and Cam domains are likely to have been initiated into tantric systems dominated by wrathful deities to ensure their success on the battlefields.¹⁴ As noted by Bautze-Picron (2014, 123), in that period “esoteric Buddhism was tightly intertwined with politics and […] fierce characters like Mahākāla or Heruka/Hevajra were made the protectors of various kingdoms.” The existence of a transregional web of cultural, diplomatic, and religious nature is reflected by the fact that icons of wrathful deities found in East Java were integrated in a wider network that connected them to Sumatra, Bali, South Asia, and the Khmer and Cam domains in the eleventh and thirteenth centuries.¹⁵ Veneration of Hevajra/Heruka and Mahākāla was especially significant outside South Asia, and the latter deity in particular was a state-protector deity in Tibet, in the Dali kingdom (Bryson 2017), in the Tangut Empire/Western Xia, and in China under the Yuan (Weirong 2008). Bautze-Picron (2016, 180) argues that

13

Images exported by monks evidently reflected concepts developed in Eastern India, but they merged with local Southeast Asian spiritual and political considerations[;…] both Hevajra and Mahākāla were images of political power in regions outside the Indian Subcontinent, symbolizing the duty of the ruler to protect the country by destroying its enemies (Mahākāla), while illustrating the presence of the political power as universal (Hevajra/Heruka standing at the centre of a maṇḍala) and destructive of negative forces (Hevajra/Heruka dancing on corpses). One may thus wonder whether in India, as was the case in Southeast Asia or China, they had the leading position of protecting the state […].

14

Icons of fierce tantric gods and goddesses found in the Buddhist kingdoms of Sumatra, notably at the predominantly Hindu site of Bumiayu (South Sumatra), and at the Buddhist sites of Muara Takus (Central Sumatra), Biaro Bahal, and Padang Lawas (North Sumatra), display iconographic resemblances with statues produced in East Java during the Siṅhasāri period.¹⁶ The same holds true for the gigantic statue originally found at Padang Roco in Central-Western Sumatra, which we will discuss below.

The Padang Roco Statue: Bhairava or Mahākāla?

15

The colossal, 4.4-metre-high statue unearthed in 1906 at Padang Roco (see figure 1), now kept at the National Museum of Indonesia in Jakarta, represents a standing one-headed, two-armed demonic deity generally labelled by art historians as “Bhairava” or “a (tantric) Buddhist Bhairava.”¹⁷ Accepting the label given by van Stein Callenfels (1920, 63), Bautze-Picron (2014, 113–14) emphatically rejects the previous identifications, and regards it as the Buddhist deity Mahākāla instead. Noting that Bhairava and Mahākāla may be (and, indeed, were) easily confused due to their entangled iconographies,¹⁸ she argues that several iconographical features closely match those found in depictions of Mahākāla in Indo-Tibetan art. Bautze-Picron (2014, 114) correctly observes that the knife and the skull-cup held in the two front hands typically belongs to the Mahākāla iconography displayed by South Asian icons;¹⁹ yet, the Indian Mahākāla (now kept at the Berlin Museum) reproduced as fig. 3 in her article does not precisely show the same pose as the Padang Roco statue, as he holds the knife in the right hand dangling by his waist, separated from the skull in the left hand. Building on this identification, Sinclair (2022b) identifies the statue as a Vajramahākāla emanating from the (Ḍākinī)vajrapañjara (Derge No. 419) and related sources, comparing it with the standing Mahākāla image depicted on the undated Minangkabau ritual dagger named “Mandākinī” or “Si Mandang Giri,” whose form and iconography relates to both East Java and Bihar, Bengal, and Nepal (Bautze-Picron 2014, 115). While these identifications go in the right direction, there appears to be an even closer correspondence with the typical iconography of the one-headed, two-handed form of Mahākāla known as Gur gyi mgon po in Tibetan (“Lord of the Pavilion” or “Lord of Death,” Skt. *Pañjaranātha; see figure 1),²⁰ whose right hand holds the knife right above the skull, pointing towards it, at the level of the heart. A guardian-deity emanating from the Hevajra-related Vajrapañjaratantra and the Mahākālatantra, *Pañjaranātha protects the Hevajra maṇḍala and those who are engaged in its practice.²¹

16

Like the Padang Roco icon, this form of Mahākāla stands on a corpse, and usually displays an emblem-like image of Akṣobhya (of which Mahākāla is an emanation); however, while in the latter Akṣobhya is hosted in the crown, in the Padang Roco icon it is hosted above it, in the bun. As argued by Sinclair (2022b), this peculiar placement is an iconographical oddity, which might reflect “a misinterpretation of the directive ‘upon the upraised orange hair,’ ūrdhvapiṅgalakeśopari,” specified in the description of Vajramahākāla given by Sādhanamālā 301 (= Śrīmahākālasādhana: see infra). Mahākāla *Pañjaranātha also displays features that are not found in the Padang Roco icon, namely the semi-squatting posture with legs bent, the gaṇḍī (long wooden gong) held horizontally across the hands, the tiger skin, the sacred thread made out of a snake, garlands of severed heads, and a crown ornated by five skulls. While the majority of the Tibetan and Nepalese icons depicting this form of Mahākāla date from the fifteenth century and later (see Pal 1977),²² there exists an iconographical variant that Linrothe (2004, 267, Cat. 18) described as the ‘original’ *Pañjaranātha, i.e., the Lord of the early Sa skya lineage of Gayādhara paṇḍita (d. 1103) (ga ya da ra’i lugs), whose iconography derives from the fifteenth chapter²³ of the Vajrapañjaratantra (figure 2).²⁴ This icon, standing straight-legged in erect posture and not holding the gaṇḍī, is a virtually identical match of the Padang Roco Mahākāla—save for the crown, the coiffure, the thread, and the body’s ornaments. Worthy of note is also the similarity between the elongated ears of the corpse on which the Padang Roco statue is standing and those of the Indo-Tibetan icons. In view of the significant iconographic similarities, coupled with the centrality of the cult of Gur gyi mgon po/*Pañjaranātha as a tutelary deity associated with state-protection Buddhism at the Yuan court and with the Sa skya order, we identify this particular hypostasis of Mahākāla as a likely prototype for the Padang Roco statue.

17

However, we believe that there is more to say about this enigmatic icon. An aspect that strikes the eye is a set of iconographic features typically attested in East Javanese Śaiva statuary from the Siṅhasāri period. The most significant one is the skull pedestal on which the statue rests, which finds no counterpart in any known icons or iconographical descriptions of Mahākāla (Sinclair 2022b) but is regularly encountered in Siṅhasāri Śaiva icons—see in particular the remarkably similar pedestal of the Singosari Bhairava (figure 4). The same icon also holds a skull and a flaying knife; the latter displays no vajra in its pommel (just like in the Padang Roco statue, and unlike in most of the Indo-Tibetan icons), and is held in a different position than that characterizing *Pañjaranātha Mahākāla and other Buddhist krodha icons, such as Vajrabhairava.²⁵ Furthermore, instead of the crown of five skulls that characterizes the standard iconography of Mahākāla across the Buddhist world, which is indeed depicted on the Candi Jago Mahākāla (see figure 5), the Padang Roco statue’s head displays a prominent bulbous, turban-like bun with a small icon of Akṣobhya on it. This bun has a different aesthetic quality than the tall, often ovoidal standing hair (ūrdhvakeśa) displayed in icons of Mahākāla in Nusantara and their Indian cognates;²⁶ although a similar puffed-up bun is also found in many paintings and sculptures of fierce Buddhist deities from Eastern India, one is reminded of the ‘ascetic buns’ depicted on East Javanese reliefs and statuary, like the Agastya of Candi Singosari. A typically East Javanese element is the so-called candrakapāla (moon-skull) etched within each diamond of the pattern of the statue’s sarong (figure 6, see Reichle 2007, 169). The skull and the half-moon are motifs traditionally associated with Śiva. In East Java they are transformed into a single emblem where the fang-like moon encircles both sides of the skull, which is often represented on royal-sponsored Śaiva statues and temples produced in or after the Kaḍiri and Siṅhasāri periods (see Pullen 2021a, 166–76). Indeed, the candrakapāla constituted the official emblem of East Javanese King Bāmeśvara (c. 1117–1135), which is interpreted by Hadi Sidomulyo as a Śaiva symbol connected with the gradual growth in popularity of Bhairava cults in East Java since that period, as attested by epigraphic and literary sources (see Robson and Sidomulyo 2021, 150–51). As such, it could be regarded as a symbol of allegiance to a specific Bhairavika priestly lineage, an emblem of dynastic genealogy, or both.

18

Keeping in mind that Mahākāla, even in his tantric Buddhist manifestation, admittedly occupies a “liminal position at the bauddha–tīrthika divide” (Sinclair 2022b, par. 54), being ultimately a transformation of the originally Śaiva Mahākāla/Bhairava, it seems reasonable to ask whether the ‘contamination’ of a Buddhist icon with iconographic elements that are commonly found in Śaiva icons dated to the Siṅhasāri period intended to give it a subtle Śaiva ‘imprint,’ or simply reflected a typically East Javanese style. Either scenarios link back to King Kṛtanagara, whose ecumenic and/or syncretic religious agenda is reflected by some of the art and architecture produced under his reign. In the light of these outstanding questions, a re-examination of the provenance and function of the Padang Roco statue seems in order.

The Padang Roco Mahākāla as Kṛtanagara’s Wrathful ‘Deification Statue’

19

A general scholarly consensus attributes the Padang Roco statue to Ādityavarman, deeming it to be a representation of himself as a deity, ever since Schnitger expressed that idea in 1937.²⁷ This attribution, in spite of hardly being supported by any evidence, has eclipsed earlier alternative views, and has not been seriously called into question thus far.²⁸ While the common features shared by the statue with the Buddhist and Śaiva tantric imagery developed in East Java during the reign of Kṛtanagara may seem to support it—for one could argue that Ādityavarman, who spent several years at the East Javanese court, was influenced by the artistic and religious paradigm of the period—, one could also argue that Kṛtanagara himself sponsored this in many respects unique icon. Indeed, Moens (1924, 557) suggested just that, asking whether the East Javanese king did not send to Padang Roco this “portrait statue” (portretbeeld) along with the Amoghapāśa Lokeśvara replicating the deified portrait sculpture of Kṛtanagara’s deceased father Viṣṇuvardhana (figure 1) as a proof of his power.²⁹ Along similar lines, independently of both Moens and ourselves, Sinclair (2022b, par. 36) argues that the statue was not only an “iteration of a type that had already reached the Siṅhasāri royal domains and was, therefore, known to Kṛtanagara,” but also represented a portrait of the latter. In what follows we shall test this idea against relevant art historical and textual evidence.

20

Firstly, the iconographical variations displayed by the statue with respect to its possible Indo-Tibetan prototype suggest that the Nusantaran icon reflects a process of local adaptation, on the one hand, and responded to Kṛtanagara’s specific needs, on the other, namely to represent himself in the form of a protective deity displaying the attributes shared by Mahākāla and Bhairava. In particular, the frontal upright posture emphasizing verticality not only mirrors that of the Amoghapāśa-Viṣṇuvardhana of Padang Roco but is also strikingly remindful of that of Majapahit-period ‘deification statues’ representing divinized royal or aristocratic mortals. These East Javanese deification statues display an unusual combination of features, namely attributes and gestures related to meditation, ritual, and worship, and are depicted as being absorbed in/united with the supreme deity, which “is not specified and borrow aspects of the great gods of Hinduism and Buddhism […]. The equality of Śiva and Buddha, of Shiwaism and Buddhism, is stressed” (Lunsingh Scheurleer 2008, 300). As such, the Padang Roco statue would qualify as a possible precursor. In fact, no less than four statues displaying traits of the Siṅhasāri period-style associated with Kṛtanagara—including the merging of divinities belonging to different religious denominations—and recovered in the Candi Singosari compound have been regarded as early codifications of the formula recurring in deification statues sculpted in the Majapahit period (Lunsingh Scheurleer 2008, 321–22, figs. 23, 24 and 25).³⁰ Further, as pointed out by Sinclair (2022b, see in particular figs. 6 and 7), there appears to be a striking similarity between the physiognomic traits of the face of the Padang Roco Mahākāla and that of the so-called Joko Dolog statue portraying Mahākṣobhya—and possibly Kṛtanagara³¹—, a smaller replica of which, also bearing unusual individual traits, is now found in Malang.³² We also point out that the smaller damaged statue (now lost) of Mahākāla found at Candi Jago (figure 5) shares with the Padang Roco statue the unusual feature of facial hair.³³ While it may seem naïve to attribute the similar facial traits of various idealized deities to a conscious will by Kṛtanagara to reproduce his features in those divine icons, we do not think it is beyond the realm of possibility. But even if this view is rejected, one may still propose that those statues are products of the same East Javanese stylistic tradition or even of the same workshop.

21

Further, as noted by Reichle (2007, 187, 191), there are intriguing iconographical similarities between the Padang Roco Mahākāla statue and other icons found in East Java associated with Kṛtanagara, so much so that they “would lead one to suspect that it was made in East Java” (Reichle 2007, 191).³⁴ These include the detailed and intricate carving of the jewellery and the delicate patterning of the sarong’s fabric, which are reminiscent of the Prajñāpāramitā statue found near Singosari; the skull pedestal on which the statue rests, which, as we have noted above, closely resembles that of Śaiva statues from Singosari, such as the dancing Bhairava and the ‘tantric Gaṇeśa’ (and their smaller replicas in the Cāmuṇḍī and Ardhanareśvarī groups); and the snake bracelets, the cascading ringlet of hair, the sashes of the sarong tied in large bows at the sides of the waist, and the kāla-shaped belt buckle, which also characterize the Ardhanareśvarī group and the Candi Jago Mahākāla (Reichle 2007, 199). The last icon also displays similar attributes—a skull cup and a cleaver held pointing down along his right thigh, rather than in front of his chest—, and hosts an Akṣobhya in the central niche of the headdress (Reichle 2007, 199–201). Unlike that icon, however, the Padang Roco statue displays a typically ‘royal’ diadem (Reichle 2007, 172–73), devoid of skulls; its calves, arms and chest are covered with swirling hair—“a unique characteristic of this statue” (Reichle 2007, 169–71).

22

As mentioned above, the candrakapāla emblem is another motif associated with Kaḍiri and Siṅhasāri art, and statues produced under Kṛtanagara in particular, as is the flower ear-ornament motif displayed by the Padang Roco icon and other statues associated with Kṛtanagara, which according to Pullen (2019, slide 28) disappears in the fourteenth century. A similarity may also be detected between the rare textile pattern carved on the statue’s dress and the pattern displayed on three statues from Candi Jago (i.e., Śyāmatārā, Bhṛkuṭī, and Sudhanakumāra), which is otherwise unattested in Java.³⁵ Lastly, the colossal size of the Padang Roco statue, which exceeds that of most of the monumental statuary ever produced in Java, may also be relevant to determine its origin and dating: gigantic statues are unusual in Nusantara, but were commonly produced during Kṛtanagara’s reign.³⁶

23

Just like previous scholars, Reichle (2007, 201) connected these statues to Ādityavarman rather than Kṛtanagara, arguing that “it is tempting to think that Ādityawarman erected this small statue at Jago after his renovation of the site and then the monumental image after coming to rule in Sumatra.” However, we believe that the evidence mentioned above supports the association—firstly voiced by Moens, and also advanced by Sinclair (2022b) in this special issue on the basis of iconographic evidence—of the Padang Roco Mahākāla statue (as well as the other smaller icons at Candi Jago) with Kṛtanagara himself rather than Ādityavarman. Since the former did actually send to Sungai Langsat in 1286 the large statue of Amoghapāśa from Candi Jago that was later re-consecrated by Ādityavarman in 1347, the transfer of the Mahākāla statue, in spite of its even larger size, cannot be ruled out as an impossible feat.³⁷ Indeed, the possibility that the colossal Mahākāla sculpture might not have been locally produced on behalf of Ādityavarman but rather brought from East Java was recently voiced—independently from Moens—by Kulke in a short (and, by his own definition, “conjectural”) post scriptum to his 2009 article on Ādityavarman, where he argues that

24

in view of its Siva-Buddhist iconology, its style and excellence of workmanship and the fact that no locally produced stone sculptures of any size and of similar quality are known in Sumatra, its East Javanese origin should not be ruled out completely, particularly if the stone out of which it was produced is unknown in Dharmasraya. (Kulke 2009, 249)

25

Kulke concludes that, if this is the case, the rituo-political function of the statue would have to be reconsidered, representing “a mighty demonstration of Majapahit’s authority in Dharmasraya rather than […] a symbol of Adityawarman’s independence from Majapahit” (2009). The transfer of the grand icons of Amoghapāśa and Mahākāla from East Java to Sumatra, which must have been an extraordinary logistic feat, was then possibly meant to be an act of “ritual conquest” (2009, 230),³⁸ or rather, as already suggested by Berg (1965, 99) and de Casparis (1985, 247–48), a political alliance against the threat of an attack by Kublai Khan or other hostile forces.

26

More recently, Ni Wayan Ariati (2016, 159–63) has discussed the images of wrathful deities showing stylistic features associated with the reign of Kṛtanagara, including the Candi Singosari Bhairava (which to her portrays Kṛtanagara himself), the Padang Roco statue (which she calls a “Bhairava”), and a statue of Bhairava or Bhīma from the temple of Pejeng in Bali, and suggested that “the spread of Bhairava imagery throughout East Java and as far as Sumatra to the West, and Bali to the East, may have occurred in the late Singasari period due to Kertanegara’s desire to create a strong system of ‘magical defence’ in the archipelago.” Ariati concludes that the Padang Roco statue could reflect political purposes like those of Kṛtanagara, in the context of Ādityavarman’s reign in Sumatra and his concerns about possible threats from Yuan China as well as the rising power of the Islamic trading states in Northern Sumatra (Ariati 2016, 163).

The Candi Jawi Śiva-Buddha Statue: A Parallel Icon?

27

In support of the view that the Padang Roco colossal statue might have been crafted in a Siṅhasāri milieu, we would like to draw attention to some relevant Old Javanese textual evidence, i.e., a well-known textual passage in the Deśavarṇana that was already mentioned by Moens as being relevant to the issue of the statue’s origin. This is the description, in Canto 56, of a statue representing ‘Śiva-Buddha’ at the syncretic Candi Jawi:

28

ndan tiṅkahnikanaṅ sudharma riṅ usāna rakva karәṅə̄
kīrtti śrī kṛtanāgara prabhū yuyut nareśvara sira
tәkvān rakva sirāṅadhiṣṭhita śarīra tan hana vaneh
etunyaṅ dvaya śaiva boddha saṅ amūja ṅūni satatā. 56.1

29

cihnaṅ caṇḍi ri sor kaśaivan apucak kaboddhan i ruhur
mvaṅ ri jro śivavimbha śobhita halәpnirāparimitā
akṣobhyapratime ruhur mmakuṭa tan hanolyantikā
saṅke siddhinirān vināśa tuhu śūnyatatvaparamā. 56.2

30

Now the character of the foundation in the past is of course well-known:
It was a pious work of King Kṛtanagara, the great-grandfather of the King.
Moreover it was he who ruled over it in bodily form, and he alone.
And hence it was both Śaivas and Buddhists who always used to worship there. 56.1

31

As a sign of this the caṇḍi below was Śaiva, with a Buddhist pinnacle above.
And within was a splendid image of Śiva of limitless fineness;
A statuette of Akṣobhya above the crown was undoubtedly its highest point.³⁹
And it was through its supernatural powers that it was destroyed, being truly of the highest essence of Void. 56.2 (trans. Robson 1995, 165)

32

This passage contains several interesting points, two of which are worth discussing in detail here. First, the expression aṅadhiṣṭhita śarīra in line c of verse 1 has been translated by Robson as “ruled over it in bodily form,”⁴⁰ thus suggesting that the sanctuary must have been some kind of state temple adhering to the Śiva-Buddha cult espoused by Kṛtanagara. In harmony with this idea is Pigeaud’s (1960, 68) translation of the expression as “to erect […] in person, self, alone.” However, in the light of the meaning (among others) of adhiṣṭhita in Sanskrit as “inhabited,” one could also understand aṅadhiṣṭhita as having a meaning akin to that of the cognate Old Javanese verbs aṅadhiṣṭhāna, inadhiṣṭhāna, “(intr.) to take one’s place in st. material, embody os. in; (tr.) to give a material support to, make st. enter into st. material, give a visible form to” (Zoetmulder 1982, 18) (compare the substantive paṅadhiṣṭhānan, “that in which st. is embodied or made visible). One could therefore propose the translation “he embodied himself in it,” hinting at the fact that the temple hosted a statue regarded to be a divine embodiment of Kṛtanagara, thus suggesting that the temple was devoted to his personal cult while he was still alive.⁴¹ If that was the case, we would have another piece of evidence highlighting Kṛtanagara’s penchant for representing himself in divine form.

33

Second, this passage has been generally interpreted as describing two statues—one of Śiva (śivavimbha), and a smaller one of Akṣobhya (akṣobhyapratimā) placed above the crown (i ruhur mmakuṭa)—which were considered as together representing Lord Śiva-Buddha (see Pigeaud 1960, 152–53; Nihom 1986, 495; Hunter 2007, 41).⁴² This feature seems to be reiterated in stanza 3 of Canto 57 of the Deśavarṇana, which mentions the Akṣobhya as “very hard to make out there above” (atisūkṣma, lit. “very subtle,” immaterial, etc.) (Zoetmulder 1982, 157). The placement of a movable icon of Akṣobhya above Śiva’s crown, while not impossible, seems awkward; indeed, Krom (1920, 2:72–75) suggested that the Akṣobhya may have actually been hosted in a ‘secret chamber’ above the temple. In support of this view, we note that the word (m)makuṭa in 56.3 may have been used by Prapañca to refer (in the sense of “crest, point, head”) not to the statue, but to the supposedly Buddhist upper section of the temple, where the śikhara begins.⁴³ However, following Poerbatjaraka (1917), we deem the former scenario (i.e., an icon of Śiva with a small Akṣobhya carved above the crown) to be more likely. At the same time, we cannot avoid drawing a parallel with the colossal statue of Padang Roco, which also hosts a small image of Akṣobhya in the headdress—not in, but above the crown, in the bun.

34

The mention of a Śiva statue with an Akṣobhya image above the crown is remarkable, for the iconographical merger (whether eclectic or syncretic) of Śaiva and Buddhist features is rare in Nusantara (except during Kṛtanagara’s reign and its aftermath), and in the entire Indic world indeed, although comparable icons from both South and Southeast Asia are known to exist.⁴⁴ That this arrangement might have been deemed controversial is suggested by the stanzas of the Deśavarṇana following those quoted above, which relate the visit to Candi Jawi of an eminent Buddhist monk-scholar (munivara) of courtly pedigree and his conversation with the temple’s abbot, “at the time when the holy image of Akṣobhya disappeared” (təpək saṅ hyaṅ akṣobhyāvimbhan hilaṅ, 57.1a):

35

sira ta mahās atīrtha seccāmәgil riṅ sudharma ḍalәm,
praṇata marәk i saṅ hyaṅ arccātibhaktyāṅdharāṅastutī
ya taṅ amuhara śālyani tvasniraṅ sthāpakānaṅśaya
ri vәnaṅanira bhaktya ri hyaṅ sivārccātañā ṅakṣama. 57.2

36

munivara mavarah sire tatva saṅ hyaṅ sudharmmeṅ daṅū
mvaṅ i hananira saṅ hyaṅ akṣobhyavimbhātisūkṣme ruhur
ryy ulihiran umaluy muvah maṅhinәp riṅ sudharmmomarәk
salahāśa kavṅan sirān ton ri mukṣa hyaṅṅ ārccālilaṅ. 57.3

37

pilih anala śarārkka rakva śakābde hyaṅ ārccan hilaṅ
ri hilaṅira sināmbәr iṅ bajraghoṣa sucaṇḍi dalәm
pavaravarahiraṅ mahāśrāvakāvas ndatan saṅśaya
pisaniṅu valuya dhārmma tәkvan kadohan huvus. 57.4

38

He was travelling about visiting holy bathing-places and chose to spend the night within the sanctuary.
He humbly came before the holy statue and with great devotion he respectfully offered praises.
This caused the Abbot to feel pained, as he doubted
whether he was entitled to worship an image of Śiwa, and so he asked about it with an apology.

39

The sage informed him of the true nature of the sanctuary in the past.
And of the presence of the holy statue⁴⁵ of Akṣobhya, very hard to make out there above.
When on his way back he again spent the night in the sanctuary in the presence.
He was disappointed and dismayed to see that the statue had completely vanished.

40

It must have been in “fires-arrows-suns” of the Saka-era (1253, AD 1331) that the holy statue disappeared;
When it disappeared the fine caṇḍi within was struck by lightning.
The repeated account of the great monk is clear and leaves no room for doubt—
It could not possibly become a sanctuary again, and moreover had already been deserted. (Trans. Robson 1995, 165–66)

41

Prapañca informs the reader that the statue disappeared (vināśa, mukṣa, hilaṅ)⁴⁶ in 1331—that is, nearly forty years after Kṛtanagara’s death. While it may appear that Prapañca, especially in 57.1a, is referring to the Akṣobhya statuette only, the fact that he uses the term (hyaṅ) arcca, referred to the Śiva statue, instead of (saṅ hyaṅ) vimbha or pratimā, referred to the Akṣobhya image, suggests that in 57.4a he is hinting at the disappearance or physical removal of the Śiva+Akṣobhya statue as a single icon (see Nihom 1986, 495). Because of the association of Akṣobhya with wrathful manifestations of tantric Buddhist deities, Kṛtanagara’s penchant for wrathful deities (compare the Bhairava of Candi Singosari, possibly representing its central icon), and Prapañca’s notorious reticence to openly refer to, and provide the details of, the tantric nature of Kṛtanagara’s cult, one may assume, just like Moens (1924, 534, 558) did, that the Śiva icon enshrined at Candi Jawi would have been a wrathful one, i.e., a Bhairava.⁴⁷

42

In the light of the above, we find it tempting to entertain the hypothesis that the statue at Jawi and the Padang Roco icon were one and the same. The pedestal of skulls, the flaying knife and the skull, as well as the candrakapāla and the kāla-kīrtimukha buckle displayed by the latter statue are all compatible with Bhairava’s iconography on the one hand, and with Śaiva statuary of Kṛtanagara’s period on the other, whereas one notes the absence of such typically Buddhist iconographical features as vajras. Could this statue, mainly reflecting an esoteric, unfamiliar Buddhist iconographical type newly introduced from overseas on the one hand, and Kṛtanagara’s wish to merge a Buddhist and Śaiva iconographical type on the other, have led astray lay worshipers and low-level religious agents in the aftermath of Kṛtanagara’s reign? Could Kṛtanagara’s wish to represent his divinized persona as a subtly ‘Śaiva-Buddhist’ icon of Mahākāla have been largely forgotten by the third decade of the fourteenth century except in elite milieus? This scenario, however far-fetched it may sound, seems to be supported by the fact that the caretaker of Candi Jawi apparently regarded the temple’s main icon as Śaiva, whereas the learned monk, coming from the courtly ‘inner circle,’ was aware of the icon’s (as well as the temple’s?) ‘true nature,’ and worshiped it accordingly. The detail about the Akṣobhya image as being “very hard to make out there above” in Deśavarṇana 57.3 suggests that, if the Akṣobhya icon was not easily visible above the statue’s crown, it must have been very small indeed, while the size/height of the Śiva statue must have been huge. The hypothetical sizes of the main statue and of the Akṣobhya image are compatible with those of the Padang Roco Mahākāla statue and the Akṣobhya above its crown. It seems at least theoretically possible that in the dimly lit sanctum of the temple the small Buddhist icon carved on the coiffure of the colossal statue, which was probably placed on a high altar or pedestal, could have gone unnoticed. Indeed, Poerbatjaraka (1917, 144–47) already noted that the only way to justify the dialogue between the Śaiva abbot and the Buddhist monk in canto 57 of the Deśavarṇana would be to suppose that the Akṣobhya image was not easily visible, and that the abbot was apparently unaware of its existence.

43

Adding to the mystery, the account about the disappearance of the statue due to its supernatural powers suggests that Prapañca may have intended to provide a fictional account to justify the removal of an icon that was deemed to be controversial during his time. Indeed, the Buddhist monk’s sibylline statement about “the true nature of the sanctuary in the past” could hint at the fact that it was frequented by both Śaiva and Buddhist worshipers on account of its syncretic statue, but also indicates that the sanctuary had been later repurposed to cater uniquely to the Śaivas. Could the main ‘Śiva-Buddha’ icon have been regarded and worshiped as a Buddha (i.e., Mahākāla) by the Buddhists, and as a Śiva (i.e., Bhairava) by the Śaivas, precisely on account of its subject—the intra-sectarian Mahākāla—and its overlapping Śaiva and Buddhist iconographic features?

44

Be this as it may, it does not seem coincidental that the statue disappeared after the Śaiva abbot comes to know about the Akṣobhya image from the monk, as a result of which the sanctuary is allegedly struck by lightning and abandoned (see Hunter 2007, 41n36). Of course, it is difficult to explain the complete disappearance of what was likely to be a huge statue, even as a result of the incident involving lightning mentioned by Prapañca (and from the Old Javanese wording it seems that the lightning was the effect rather than the cause of its disappearance). But that something worthy of being recorded did actually happen, whether or not along the lines mentioned in the Deśavarṇana, is suggested by an inscribed stone block with the date 1254 śaka (1332 CE), i.e., just one year after the alleged disappearance of the statue, which was recovered from Candi Jawi and is now kept at the Trowulan Museum (figure 1).⁴⁸ While one could only speculate about its function and the circumstances surrounding its placement, we find it remarkable that around that year Ādityavarman had already started his political career in East Java (under the reign of Queen Tribhuwanā, Kṛtarājasa’s daughter), and certainly had access to ships because of his prominent position at court—indeed, he might even have visited the Yuan court in the capacity as a diplomatic envoy in 1332, exactly one year after the disappearance of the statue (see infra, fn. 81). Assuming that the ‘Śiva-Buddha’ statue of Jawi was actually the Padang Roco icon, one could speculate that the statue was removed from the temple and sent to Sumatra by ship by Ādityavarman, just like the Amoghapāśa statue lying nearby in Padang Roco was sent by Kṛtanagara.

45

Admittedly, while the hypothetical scenario reconstructed above does provide a solution to the enigma of the syncretic Candi Jawi and Prapañca’s description of its missing statue, as well as the identity and provenance of the colossal statue of Padang Roco, it is impossible to prove at this stage. It is also partly undermined by the fact that neither the Śaiva icons recovered at Candi Jawi, nor the relatively small dimensions of the sanctuary and its cella match the colossal size of the Padang Roco statue. If the statue was indeed enshrined in an East Javanese temple, a better candidate would be Candi Singosari, where Deśavarṇana 43.5d states that Kṛtanagara was worshiped (post-mortem) as Śiva-Buddha: riṅke sthānanirān dhinarmma śivabuddhārcca halәp mottama, “His Majesty was laid to rest, people say; Here is the place where he is enshrined as a Śiwa-Buddha statue of imposing fineness” (trans Robson 1995, 156). Thus, to summarize, one could envisage the following alternative hypothetical scenarios (see table 1):

46

Any one of the scenarios outlined above is supported by the fact that Kṛtanagara is known to have commissioned replicas of icons he deemed significant and dispersed them across his realm, including East Java and Sumatra, for political purposes.⁴⁹ None of them makes the possibility that the Padang Roco statue was intended as a trans-sectarian Mahākāla, i.e., a subtly Śaiva-Buddhist wrathful icon representing a divinized Kṛtanagara, any less likely.

Kṛtanagara’s jinābhiṣeka

47

Given the precedent set by Kṛtanagara with the erection of a Śiva-Buddha statue in Candi Jawi, as well as other ‘deification statues’ in Singosari and other locations, including the Joko Dolog icon, one could imagine that the Padang Roco statue was meant to represent the ruler himself as a subtle merger of Mahākāla and Bhairava. In support of this view, and in order to cast some light on what tantric Buddhist systems Kṛtanagara would have adhered to, we shall now discuss his consecration name (i.e., jinābhiṣeka, “consecration as a Jina”). This is reported as Jñānabajreśvara in the Deśavarṇana (43.2c), as Jñānaśivabajra in the Vurare inscription engraved on the back of the Joko Dolog statue (1.11), and as Jñāneśvarabajra in an inscription of Singosari of 1351. What the exact meanings of these compounds are has been object of speculation over the past hundred years (Nihom 1986, 492–95).

48

The terms jñāna and bajra—a Javanese spelling of vajra—occurring in all three compounds, whether separated by śiva (as in the Vurare inscription) or īśvara (as in the Singosari inscription), are likely to have been understood as forming the word jñānavajra. This form is found not only in the Deśavarṇana, but also in a Sanskrit language/Nāgārī script dedicatory inscription engraved on the back of a statue representing a miniature version of eight-armed Amoghapāśa (in the same style as the icon from Candi Jago), describing Kṛtanagara as a “great Mahāyāna follower” (pravaramahāyānayāyinaḥ) and giving his official name as śrī kṛtanagaravikramajñānavajrottuṅgadevamahārājaḥ.⁵⁰ Jñānavajra is a technical term appearing in tantric Buddhist texts in association with a weapon usually emanating from a wrathful form of a buddha or bodhisattva.⁵¹ In the Yogatantra commentary on the Mañjuśrīnāmasaṅgīti, the Nāmamantrārthāvalokinī by Vilāsavajra, it denotes a circle of flames that makes the protected space “unapproachable by Māras” during the ritual of the maṇḍala (Chapter 4; Tribe 2016, 143); this jñānavajra is the same vajra, generated from the syllable HŪṀ, that in the Vajra-bhairavatantra transforms itself into the deity Vajrabhairava, who tames the wicked ones and protects sentient beings. In the mythic narratives denoting the origins of Vajrabhairava in the Tibetan tradition, jñānavajra is an important term for Vajrabhairava’s liṅga, which is considered to be a powerful weapon with which he subjugates and converts the evil ones, such as Yama(rāja), Śiva-Maheśvara, and their retinues.⁵²

49

The term bajrajñāna is attested in Old Javanese transmitted literature. For instance, the Sanskrit-Old Javanese tantric Buddhist manual Saṅ Hyaṅ Kamahāyānikan (Advayasādhana) (46a) identifies the body with eight petals of a lotus, which are stabilized by bajrajñāna, which in turn equals advayajñāna.⁵³ The related text Saṅ Hyaṅ Kamahāyānan Mantranaya (v. 12) mentions a vajra-wielding vow equated to Vajrasattva, through which the unsurpassed bajrajñāna penetrates the disciple and becomes his heart.⁵⁴ The word bajrajñāna also occurs in the Old Javanese Buddhist kakavin Sutasoma (or Poruṣādaśānta) by fourteenth century East Javanese courtly poet Mpu Tantular as the name of the supreme deity Mahāvairocana, whose manifestation is the prince and king-to-become Sutasoma. The deity is invoked in its opening hymn as follows: śrī bajrajñāna śūnyātmaka,⁵⁵ “When He emerges in the mind of the enlightened one, He is in all respects equal to the brilliance of sun and full moon together” (Sutasoma 1.1, sākṣāt candrārka pūrṇādbhuta ri vijilirān saṅka riṅ boddhacitta). The mention of sun and moon may be an echo of the Saṅ Hyaṅ Kamahāyānikan’s advayajñāna (42a), arisen from the “sun of meditation” (sūryasmṛti) and “tranquil moon” (śāntacandra), representing the syllables aṁ aḥ (see O’Brien 2008, 208–10).⁵⁶ This set of images pertaining to the meditative sphere finds a parallel in the mental generation of Mañjuśrī as the moon (peaceful) and the sun (as the wrathful Vajrabhairava or Mahākāla) in Indian sources, including the Vajrabhairavatantra.⁵⁷ We wonder whether the pattern of a blazing sun-disc incised on the back of the Padang Roco Mahākāla (which is also found on the Amoghapāśa at Padang Roco, as well as on the Gaṇeśa of Singosari: see Moens 1924, 557), and the half-moon behind the head of icons of Mañjuśrī from Central Java and China (e.g., Mogao Cave 98 dating to 925, see Miksic 2006, 209) reflect these symbolical associations.

50

An interpretation of the word bajrajñāna as Kṛtanagara’s kingly power of world-protection—an activity that implies an occasional exercise of violence—was advanced by O’Brien (2008, 226–27, 236–37). Since bajrajñāna equals a divine vajra mind-weapon, an analogy can be established between Sutasoma and Kṛtanagara in the Sutasoma (e.g. 3.4cd): That text, while sponsored by Kṛtanagara’s great-grandson Rājasanagara, has been regarded as representing a political allegory whereby Sutasoma = Kṛtanagara and Poruṣāda = Kublai Khan (2008, 229, 236).⁵⁸ Sutasoma uses a thunderbolt-weapon emerged from the concentration on jñāna and the Bodhyagrī-mudrā against evil, namely Gajavaktra and his master Poruṣāda, a manifestation of Śiva as Mahākāla.⁵⁹ O’Brien takes bajrajñāna in Kṛtanagara’s jinābhiṣeka’s name and jñānabajra in the Sutasoma as having no appreciable difference. As for the element -śiva- and -īśvara-, she interprets them as an attempt to syncretically unite Śiva and Buddha (2008, 237–40). We find this interpretation plausible, also on account of the attestation of bhaṭāra śivabuddha as the abhiṣeka name of Kṛtanagara in the Old Javanese Pararaton (18.15–16), and in harmony with the hypothesis that Kṛtanagara might have been portrayed in the Padang Roco Mahākāla. Speaking about the Śaiva way, the Sutasoma (40.6) states that one of its dangers is to acquire supernatural powers and become an embodiment of the fierce deity Rudra, whose power is difficult to control. Kṛtanagara, as the sovereign who inspired the allegorical narrative of the Sutasoma, would have eclectically chosen to unite both ways to increase his power against the enemies of his kingdom, both during his life and post-mortem. It is, therefore, not surprising that Kṛtanagara would have chosen to represent himself as (Mah)ākṣobhya (of which Mahākāla is an emanation) in the Joko Dolog statue carrying the abhiṣeka name Jñānaśivabajra, given the repeated equations found in post-Kṛtanagara Old Javanese literature between Akṣobhya and Śiva or manifestations thereof, including Rudra,⁶⁰ and the Deśavarṇana’s reference to the Śiva (i.e., Bhairava?) with Akṣobhya above the crown enshrined at Candi Jawi.

51

Be this as it may, in view of the apotropaic and/or the martial implications of the term jñānavajra in the texts mentioned above, and considering the particularly hostile relationship between Kublai and Kṛtanagara that would have prompted the adoption by the latter of tantric initiations as magical means to combat the impending invasion of Java by the former, it may be suggested that Kṛtanagara’s initiation name jñānabajreśvara was borrowed from the tantric nomenclature found in both the Yogatantras and the more transgressive Yogottara and Yoganiruttaratantras.

52

The textual and art historical record suggests that Kṛtanagara adopted tantric technologies to repel Kublai’s hostilities. Given that the Yuan emperors were familiar with the cult of Mahākāla, Hevajra, Heruka/Cakrasaṁvara, and Yamāntaka/Vajrabhairava, it seems plausible that the Javanese king, who had openly challenged Kublai in a series of diplomatic incidents culminating in the invasion of Java by the latter’s army, had been initiated into similar cults to counteract the power of his rival on a magico-ritual plane.

Nusantara at the Crossroads: Transregional Buddhist Links in the Thirteenth and Fourteenth Century

53

Unfavourable socio-political developments—most notably the Muslim invasions and the demise of the Pāla-Sena dynasty in Northeastern India—caused the decline of Buddhism in Northern India in the period from the eleventh to the fourteenth century, triggering a diaspora of Buddhist masters and artisans seeking sponsorship in Southeast and East Asian courts. Some of these masters would have been prime vectors for the transmission of the tantric systems adhered to by Kṛtanagara and Kublai Khan. Art historical vestiges suggest that Northeastern Indian and Newar artists were hosted at the court of Kṛtanagara—just as they were by Kublai and subsequent Yuan rulers—to produce Buddhist icons (in particular, the statues found at Candi Jago).⁶¹ More circumstantial evidence indicates that prominent religious figures from the subcontinent may have been hosted by the East Javanese monarch and his successors, and possibly also by Ādityavarman in Sumatra.⁶²

54

Among this evidence is the graffiti inscription recording the passage at the Karimun islands in Central-Eastern Sumatra—right in the middle of the straits separating the island from the Malay Peninsula and the actual island that is now Singapore—of the Indian paṇḍita Śrī Gautamaśrībhadra (also known as Gautamaśrī), who hailed from Bengal and was active in Tibet, Nepal, and China between the early to mid-thirteenth century and at least 1279 (Sinclair 2018, 2021). Gautamaśrī is identified in Tibetan lineage texts as an expert in many tantric texts and practices, including the Kālacakra (Bu ston, Deb ther sngon po [Blue Annals], 802), which he taught to Tibetan disciples, and which were adopted by the subsequent Sas kya pa tradition. In 1253 a lama of the Sa skya pas, Dam pa Kun dga’ grags (Dan-ba Guoshi 膽巴國師, 1230–1303), who later became Kublai Khan’s state preceptor, helping him to establish the Yuan dynasty and performing war magic for him, met and studied with Gautamaśrī in Tibet (Sinclair 2018). Dam pa Kun dga’ grags played an important role in introducing the cult of Mahākāla among the Mongols and helped their armies during their military campaigns (Weirong 2008, 319, 2011, 542; Debreczeny 2014). Thus, the figure of Gautamaśrī may represent a plausible link between the East Javanese and Sino-Tibetan courts: even though positive evidence is lacking, it is hardly conceivable that he (and/or his acolytes) would have embarked on such a long sea journey and crossed the straits without spending some time in East Java—a major political and religious powerhouse of the time—either on the way to or back from China.⁶³ With regard to this, it seems significant that Gautamaśrī, as noted by Sinclair (2022b), is known to have transmitted an anthology of sādhanas that incorporates the Śrīmahākālasādhana (corresponding to Sādhanamālā 301), a work on the two-armed Vajramahākāla that might have been the source of the iconographical detail of the Akṣobhya placed upon the upraised hair (rather than on the crown) of the Mahākāla of Padang Roco, and perhaps also of the main statue of Candi Jawi—whether or not they were the same icon.

55

A contemporary of both Gautamaśrī and Dam pa Kun dga’ grags (and teacher of the latter) was ’Phags pa Blo gros rgyal mtshan (1235–1280), also known as Chos rgyal ’Phags pa. As the seventh patriarch of the Sa skya monastery and a powerful member of the ’Khon family, ’Phags pa was undoubtedly one of the most important masters in the Sa skya lineage. Regarded as a Tibetan emanation of Mañjuśrī, he earned his fame in the Tibeto-Mongolian Buddhist history as a national preceptor (guoshi) in 1261, and then as imperial preceptor (dishi) of Kublai in 1269 and 1270 (Petech 1990, 36–37). ’Phags pa—whose relationship with Kublai was modelled on the patron-lama liaison (yon mchod: see Ruegg 1995)—acted as the Mongols’ proxy in Tibet and was in charge of religious institutions throughout the Mongol empire. More importantly, he occupied the office of royal religious officiant (mchod gnas) and was responsible for granting the emperor tantric initiation rituals (abhiṣeka). ’Phags pa seems to have had a determining influence on Kublai and the religious practices in his court, despite some scholars claiming otherwise.⁶⁴ The Tibetan Script Edict (’Ja sa bod yig ma) records that ’Phags pa granted three Hevajra initiations to Kublai in 1253.⁶⁵ Chinese, Mongol, and Tibetan sources agree that “rites of Hevajra and Mahākāla seem to have become customary for every enthronisation of a Yuan emperor” (Franke 1978, 60).

56

The cult of Mahākāla (and Gur gyi mgon po/*Pañjaranātha in particular) was of paramount importance at the court of Kublai and his successors, especially in connection with state-protection rituals. Like Vajrabhairava (and its precursor Yamāntaka, “The Ender of Yama/Death”), Mahākāla is regarded as an emanation of Mañjuśrī,⁶⁶ and is also a protector of the Sa skya order, which was influential at the Yuan court. Various epigraphic and textual sources confirm that Kublai himself was conceived of as a wrathful ‘imperial’ incarnation of Mañjuśrī as Mahākāla,⁶⁷ through which he consolidated his spiritual and political authority in accord with the Buddhist ideology. Mahākāla was invoked by the armies during the Sino-Mongol military campaigns.

57

Tibetan Buddhist lamas were in charge of the ritual protection of the empire on Wutaishan—a major transnational ritual centre of Sino-Tibetan Buddhism, regarded as the place of birth of Mañjuśrī since at least the seventh century—in the Yuan era, and ’Phags pa administrated Wutaishan’s clergy in 1257–1258 before being appointed imperial preceptor in 1260 (Charleux 2015, 95–96). As mentioned above,’Phags pa’s disciple National Preceptor Dam pa Kun dga’ grags, after having met and studied with Gautamaśrī, took residence for almost ten years on Wutaishan at ’Phags pa’s request, and became abbot of Shouningsi before 1272, after which he became known as a tantric ritual specialist at Kublai’s court.⁶⁸ Dam pa Kun dga’ grags was a teacher in the lineage of the Mañjuvajra maṇḍala of the Guhyasamāja, a specialist in the Hevajratantra, and a devotee of Mahākāla (Franke 1994, 162)—in fact, he was even deemed to be an incarnation of the same (Debreczeny 2014, 135). He built a temple for the performance of state-protection rituals devoted to that god (Debreczeny 2014, 137), and specifically the form of Mahākāla represented in the golden statue sponsored by ’Phags pa, i.e., Gur gyi mgon po/*Pañjaranātha, the special protector of the Sa skya order on Wutaishan. ’Phags pa instructed the Nepalese artist Arniko, supervisor-in-chief of all artisans at the Mongol court, to sculpt a *Pañjaranātha in 1274 or 1276,⁶⁹ and install it in the temple of Zhuozhou, south of Dadu (Beijing), to which the Mongol victory over the Southern Song armies was attributed (Debreczeny 2014).⁷⁰ Another famous icon of Mahākāla was hosted in the temple to Vajrabhairava-Yamāntaka west of Huiquing pavilion in the Yuan palace compound (Watt and Wardwell 1997, 95).

58

As a wrathful incarnation of Mañjuśrī as Mahākāla, Kublai might also have been initiated into the cult of Vajrabhairava, one of the tutelary deities of the Mongols that was worshiped in China by the political elites and in Tibetan Buddhist milieus up to the Late Imperial and Republican period.⁷¹ Besides some shared iconographical elements,⁷² the association between *Pañjaranātha and Vajrabhairava is suggested by the fact that the former is the most prominent among the fourteen protectors guarding the latter, and they are often depicted together on Tibetan paintings.⁷³ Furthermore, both of them belong to the lineage of Akṣobhya, and reflect Buddhist appropriations of Śiva as Bhairava/Mahākāla. The cult of Vajrabhairava as a wrathful form of Mañjuśrī became fashionable in the Mongolian court of Kublai, and there are reasons to believe it was ’Phags pa or his direct disciples who introduced this cult there.⁷⁴ ’Phags pa was a lineage-holder of the (Western) Rwa lugs transmission of Vajrabhairava, and arrived at Ldon ston’s invitation when he was residing at Wutaishan in China to receive Yamāri teachings from him (Dhongthog 2016, 67). ’Phags pa composed at least one sādhana of Vajrabhairava based on the teachings of Rwa lotsāwa,⁷⁵ and he was also a recipient of the Raktayamāri cult of the mahāsiddha Virūpa in the lineage of Glo bo lo tsā ba Shes rab rin chen,⁷⁶ the thirteenth-century translator from the Mustang area (glo bo), who granted him the Yamāntaka initiation.⁷⁷

59

Circumstantial evidence suggests that ’Phags pa conferred the Vajrabhairava initiation on Kublai. Cuevas (2021, 64) notes that there was a distinctive teaching that merged certain practices of Vajrabhairava of the Rwa tradition and the Kālacakratantra, i.e., a unique practice of the special Yamāntaka wheel device (Gshin rje ’khor lo) that was apparently bestowed by Ldong ston Shes rab dpal to ’Phags pa. According to Tāranātha (Gshin rje gshed [II], 109), Ldong ston conferred these same teachings upon the Mongol emperor (65). Initiations into the Vajrabhairava system were conducted for the Yuan dynasty by the imperial preceptors in later years: the first Taidung emperor, Yesün Temür, was granted the Vajrabhairava-Yamāntaka initiation by the imperial preceptor in 1324, and his wife in 1326 (Watt and Wardwell 1997, 99n15). It is plausible that the same preceptor performed a Vajrabhairava-Yamāntaka abhi-ṣeka on Tugh Temür (great-great grandson of Kublai and the emperor of the Yuan dynasty from 1328–1332) and his wife Budhashiri in 1329, for it is also known that Tugh Temür was bestowed two tantric initiations as an emperor, however no details of these have been given (1997, 95, 98). The evidence in support of the theory that it was the Vajrabhairava initiation that was conferred on Tugh Temür (and possibly on Kublai as well) is the existence of the kesi-silk tapestry thangka (now at the MET Museum in New York) designed according to the fourteenth-century Sa skya style, usually commissioned for the imperial initiation rituals (1997, 98).⁷⁸ Although Kublai’s religiosity and that of his preceptors is relatively well-documented, one has to keep in mind the highly secret nature of these initiations, which were deemed dangerous and controversial even in Buddhist milieus. In that epoch these initiations, associated with war magic and the annihilation of enemies, must have formed the premodern counterpart of ‘weapons of mass destruction,’ and as such secretly guarded from potential competitors and the intelligence of foreign powers⁷⁹ (although, conversely, tantric initiations also had an aspect of ‘soft power,’ through which political establishments sought to increase their stature by garnering a reputation for magical efficacy in war).⁸⁰

60

In the light of the above, it is relevant to mention here that Ādityavarman, in the period of Kublai’s aftermath, was stationed in East Java as a minister, and might have visited the Yuan court in the capacity as a diplomatic envoy in 1325 and 1332.⁸¹ Whether or not Ādityavarman himself went to China, the existence of diplomatic channels⁸²—as well as a flourishing maritime trade⁸³—between the two kingdoms must have facilitated the transmission of the prevalent tantric systems of that period, centred on Mahākāla, Hevajra, Mañjuśrī, and possibly also Vajrabhairava-Yamāntaka, from China to East Java. But the flourishing intra-Asian diplomatic and commercial links of the thirteenth century also support the likelihood of a simultaneous parallel transmission via the earlier masters who, like Gautamaśrī, plied the maritime routes between India, Southeast Asia, and China. The cosmopolitan and expansion-driven sovereigns of the Yuan and the Siṅhasāri/Majapahit courts must have been keen to make use of the services of tantric Buddhist paṇḍitas looking for sponsorship, and some such ritual specialists might even have been part of their armies to support them on the battlefields. As noted by O’Brien (2016, 275), “Tension between China and Java may well have continued until the Yuan dynasty came to an end in 1368 […] much of the design and embellishment of Candi Jago reflects the tension of those times.”⁸⁴

The Cult of Mañjuśrī and its Religio-Political Ramifications in Siṅhasāri/Majapahit East Java and Yuan China

61

The intra-Asian religious links as well as the diplomatic and military contacts between East Java and China in the thirteenth and fourteenth centuries discussed above prompt us to (re)consider the possibility that the existence, or rather re-emergence, of an elite cult of Mañjuśrī in East Java in the same period could be the expression of the same religio-political dynamics that linked the tantric Buddhist systems adhered to by Kṛtanagara and Kublai Khan.

62

A majestic Mañjuśrī Arapacana statue probably stood at Candi Jago and is now kept at the Hermitage Museum. Counting among the finest examples of Indic statuary known from Nusantara, this Mañjuśrī, because of its size and outstanding quality, was probably housed in a sizeable separate shrine in the temple compound and must have been directly sponsored by a member of the political elite of the time (Miksic 2006, 207). While the statue carries two inscriptions dated 1343, and does not display the demonic attributes that are typical of Siṅhasāri icons, stylistic features suggest that it may have been crafted in the 1280s, hence well within the period of Kṛtanagara’s reign (Lunsingh Scheurleer 2008, 296).

63

This statue of Mañjuśrī Arapacana is notable insofar that icons of this particular form of Mañjuśrī are not common in East Java. Indeed, the cult of Mañjuśrī (as a political symbol) is mainly attested in Central Java from the eighth to the eleventh century (Miksic 2006). It has been suggested that this unique icon represents a latent resurgence of art styles from Bengal and, as other contemporary statuary, it could betray a possible direct influence from Eastern India (Lunsingh Scheurleer 2008, 297; Reichle 2007, 196; Bautze-Picron 2016, 166). More recently, however, Pullen has pointed at a number of features that would also suggest a Central and East Asian connection. For example, the bodhisattva brandishes a sword with a vajra pommel—a feature attested in modern Nepalese and Sino-Tibetan thangkas and bronzes, but unknown in north-eastern Indian art of the Pāla period (Pullen 2021b, 159). Further, she has argued that the pattern depicted on its dress does “not fit into any design category known today in Java, Sumatra or the Malay Peninsula” (Pullen 2021a, 40), but rather reflects Iranian and Chinese motifs, i.e., an Indo-Sasanian design incorporated by the Mongols. This fact suggests the possibility of borrowing and imitation through political and commercial channels that existed throughout the thirteenth century (Pullen 2021a, 38–39).⁸⁵

64

The inscriptions engraved on the two faces of the statue, although recording the same date, display important differences in their language and script. Bosch (in Kozok and van Reijn 2010, 142) argued that a religious official of the Bhairava sect, Saṅ Ārya Vaṅśādhirāja,⁸⁶ erected and consecrated “a statue of Bhairawa, the highest god of his sect, in the shape of Mañjuśrī” in 1343/Śaka year 1265 (the year recorded on the inscriptions) and made the inscription on the front of the statue, while the one on the back was inscribed later for Ādityavarman. Bosch noted that the relationship between the Bhairava priest Vaṅśādhirāja and Mañjuśrī would be explained by the fact that

65

In the lamaistic pantheon Bhairawa is the highest manifestation of Mañjuśrī in his dreadful shape (krodha), one of the most hideous divinities of tantric Buddhism, depicted with 16 feet, 34 arms and 9 heads (of which the lowest one is a bull’s head), trampling underfoot gods, men, quadrupeds and birds, having his abode in a cemetery populated with corpses, those sentenced to death, phantoms, vultures, dogs and jackals. Waṅśādhirāja did not erect and consecrate in this shape the main god of his sect, but in the gentle shape (śānta) of Mañjuśrī, the god of wisdom, equipped with a book and a sword. (Bosch in Kozok and van Reijn 2010, 142)

66

The deity to which Bosch refers is not Bhairava, but the ‘Buddhist Bhairava,’ namely the buffalo-headed deity Vajrabhairava, which has indeed a longstanding association with Mañjuśrī Arapacana (as a wrathful emanation thereof) as well as Akṣobhya.⁸⁷ Although Vajrabhairava as such does not appear to be attested in the extant textual and iconographical record of Nusantara,⁸⁸ an elite cult of this deity may very well have existed in East Java: its absence of evidence might be due to the fact that such esoteric icons, most probably cast in metal or painted on perishable supports, must have been exceedingly rare, and are therefore likely to have disappeared long since.⁸⁹ However, in the context of this inscription, the involvement of a (Śaiva) Bhairavika priest in the statue’s consecration may simply suggest that the eclecticism and/or syncretism promoted by Kṛtanagara was carried forward in the early Majapahit milieu in which Ādityavarman flourished. According to Kulke (2009, 233, 241, 245), Ādityavarman was moved by some sort of ‘tantric obsession,’ for which he took Kṛtanagara as an example, also with respect to his stance towards ‘deification in life’: In the fourteenth-century Kuburaja stone-slab inscription, Ādityavarman is praised by his son Advayavarman as an incarnation of Lokeśvara (avatāra śrīlokeśvaradeva) (2009, 241).

67

In the light of its inscriptions, previous scholars have regarded the sculpture as being commissioned by Ādityavarman, who was also responsible for the renovation of the temple. On this basis, Miksic (2006, 207) suggested that Ādityavarman “did espouse a form of Buddhist doctrine which included Mañjuśrī as an important figure.”⁹⁰ However, the importance of this bodhisattva for the form of Buddhism practiced by the Sumatran ruler could be equally true in the case of Kṛtanagara. Apart from the fact that the statue found at Candi Jago could be dated to Kṛtanagara’s reign on stylistic grounds, one should bear in mind that his enemy Kublai was considered an incarnation of Mañjuśrī as Mahākāla. Indeed, since Mahākāla was regarded as an emanation of Mañjuśrī, our Mañjuśrī Arapacana statue could have been part of a maṇḍalic configuration or royal cult including the two deities. Although none of the Mahākāla statues recovered at Candi Jago matches the one of Mañjuśrī in size and fineness, one could either hypothesize the existence of a lost icon, or indeed regard the Padang Roco Mahākāla as the ‘sister icon’ of our Mañjuśrī. Be this as it may, considering Kublai’s association with Mañjuśrī and the importance of this deity’s cult in Sino-Tibetan circles, the iconography and Northeastern Indian stylistic features of the Candi Jago icon, and the rarity of Mañjuśrī icons in East Java, it seems safe to assume that a transregional tantric cult, characterized by a markedly martial and political dimension, of Mahākāla (and perhaps also Vajrabhairava and Yamāri/Yamāntaka) as a wrathful form of Mañjuśrī may have been practiced at the Yuan and Siṅhasāri/Majapahit courts in the same period, as a result of the parallel influence of a wave of Indian and Tibetan masters who initiated sovereigns into tantric systems. This religious configuration, based on a prototypical tantric tradition that was transmitted to Java from overseas during Kṛtanagara’s reign, would still have been followed by Ādityavarman.

Conclusion

68

This article has gathered and linked together data drawn from textual sources and artistic remains to attempt to identify the tantric systems adhered to by Kṛtanagara against the historical background of transregional religious, artistic, and political contacts between Northeastern India, Tibet, Sumatra, East Java, and China in the thirteenth and fourteenth centuries. 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 control over part of Sumatra, we have advanced the hypothesis that the East Javanese ruler may have been familiar with cults revolving around the wrathful protector Mahākāla, and perhaps also Vajrabhairava-Yamāntaka, both being conceived of as fierce manifestations of Mañjuśrī. A growing body of evidence suggests that the circulation of tantric texts and icons was made possible by the diplomatic contacts existing between East Java and China in the thirteenth and fourteenth century, as well as by a diasporic wave of Eastern Indian, Nepalese, and Tibetan tantric Buddhist masters and artisans who sought political sponsorship overseas.

69

On the basis of iconographical and stylistic features, and in substantial agreement with the arguments advanced independently by Sinclair (2022b) in this special issue, we have argued that the wrathful deity represented in the colossal icon of Padang Roco was not meant to be a portrait of Ādityavarman, but rather a deified image of Kṛtanagara himself, and that it may have been transferred from Java to Sumatra either by the same sovereign or by Ādityavarman a few decades after Kṛtanagara’s death. Fine-tuning Bautze-Picron’s (2014) and Sinclair’s (2022b) identification of the deity as a Mahākāla rather than a Bhairava, we have pointed to an early form of *Pañjaranātha/Gur gyi mgon po, protector of the Sa skya order and tutelary deity of the Mongols, as the most likely prototype. Further, while we agree with Sinclair’s remarks on the “suitability of Mahākāla as an object of worship for Hindu-Buddhist royalty” (2022b), we do not deem Kṛtanagara’s religious attitude to have been just eclectic and bilateral, but rather syncretic, and have tentatively proposed that the Padang Roco icon reflected an attempt to either subtly merge Mahākāla and Bhairava, or to appropriate the Bhairava/Mahākāla of the Śaivas as the Buddhist Mahākāla.

70

In support of the above hypothesis, we have noted the appearance in an otherwise eminently Buddhist icon of iconographical features that are mainly attested in East Javanese Śaiva statues from Kṛtanagara’s period. Further, we have drawn attention to an iconographical parallelism existing between the Padang Roco Mahākāla and the ‘Śiva-Buddha’ statue representing Kṛtanagara at Candi Jawi described in the Deśavarṇana, namely the icon of Akṣobhya above the crown (i.e., upon the upraised hair). Since, according to Prapañca, this statue, worshiped by both Buddhists and Śaivas, disappeared in 1331, we think it necessary to consider, at least theoretically, the possibility that the two statues might have actually been one and the same; but we deem it more likely that they were separate icons reflecting a similar religious and political agenda. That Kṛtanagara’s eclectic attitude towards Śaiva and Buddhist cults might have climaxed in the unification of Mahākāla and Bhairava into his royal persona is suggested by the fact that a similarly syncretic merger of Śaiva and Buddhist features characterize the ‘deification statues’ produced under as well as after his reign; and that a mix of Śaiva and Buddhist terminology is detectable in his initiatory name, Jñānabajreśvara (or, according to the Pararaton, bhaṭāra śiva-buddha). The association of jñānavajra with weapon-like emanations of protective wrathful deities, such as Mahākāla and Vajrabhairava, in tantric literature appears to reflect Kṛtanagara’s political concerns, and this quest for magico-ritual martial efficacy may partly explain his syncretic stance.

71

The sudden appearance in East Java of a large and exquisite icon of Mañjuśrī, probably dating from the period of Kṛtanagara’s reign and later reconsecrated by Ādityavarman, which displays Eastern Indian as well as Central Asian/Mongol stylistic influences, has prompted us to explore the doctrinal, cultic, and mythical association existing between Mañjuśrī, Mahākāla, and Vajrabhairava (and his precursor Yamāntaka) in tantric literature. The fact that Kublai Khan was thought to be an incarnation of Mañjuśrī; that he and his court worshiped Mahākāla as a state-protector deity; and that he, just like his successors, might have received Vajrabhairava initiation by Tibetan masters, can be taken as circumstantial evidence to support the hypothesis that the same tantric systems may have been adopted also by Kṛtanagara, in an attempt to counteract the magical efficacy of his enemy. In fact, applying Sinclair’s (2022a) argument about Kṛtanagara’s co-option and absorption of Amoghapāśa as a divine protector of the Śrīvijaya–Melayu polity, one could also hypothesize that Kṛtanagara adopted a similar, perhaps more mimetic, attitude towards the chief protective deity of Kublai.

72

Just like subsequent Yuan emperors continued Kublai’s tradition of tantric initiations, the tantric systems adhered to by Kṛtanagara were carried forward by Ādityavarman, one of the last known Buddhist kings of Sumatra, who ‘appropriated’ the statues of his powerful ancestor to consolidate his political legitimacy as mahārājādhirāja and successor to the Siṅhasāri and Majapahit royal lines, and perhaps even to counter the advance of Islam on the island. This seems all the more plausible since the two-armed form of Mahākāla is often invoked in Buddhist Tantras, including the Ḍākinīvajrapañjara, as “devouring adversaries of the dharma,” chopping their flesh and drinking their blood (Sinclair 2022b).

73

Our discussion has highlighted the dislocation and decontextualization of tantric traditions through the migration across the Buddhist world of ritual specialists, as well as the circulation of medias such as texts and icons. The latter in particular, in the form of colossal statues, appear to have been adapted to the local contexts in Java and Sumatra, by the initiative of Kṛtanagara and his successors, as public statements expressing political power imbued with religious significance through the articulation of divine royalty and supernatural efficacy. Since analogous ideas and practices were current in the Yuan court as well as other coeval Asian milieus, it seems unlikely that these icons reflect local/indigenous ‘ancestral’ beliefs and practices, but rather represent a cross-fertilization between local and translocal tantric elements in the late phase of tantric Buddhism.