1 Introduction1
The transmission of ‘Golden Isles’ Acala praxis from Maritime Southeast Asia to the Himalayan Plateau and thereon to Central Asia was a singular event in the premodern world. This trail, once blazed, seems to have been never followed again, and this raises doubts about whether it existed at all. The task of tracing the Serlingpa Acala back to its nominal source then begins by clarifying the position of the Golden Isles. The Sumatran highlands, long renowned for being rich in alluvial gold, are a natural referent of the term ‘golden earth’ (Skt. suvarṇabhūmi). Over time this term was applied, by extension or appropriation, to gold-trading emporia located around the Malacca Strait and the wider Southeast Asian region. The related term ‘Golden Isles’ (Skt. Suvarṇadvīpa, Tib. gSer gling) is linguistically applicable not only to Sumatra and its surrounding islands, but also to the Malay Peninsula. While there is no scope to review every possible referent of Suvarṇadvīpa and Suvarṇabhūmi here, it can be understood that both terms refer primarily to the Malay Archipelago, and secondarily to the South-East Asian mainland.2
Within the broad region of Suvarṇadvīpa, it is Kedah, on Malaysia’s western coast, that can be pinpointed with most confidence as the base of the Serlingpa guru. In the Sanskritic world he was known by the name Dharmakīrti, and for the Durbodhālokā [Light on the Hard-to-Illuminate], a major work of Prajñāpāramitā exegesis. The colophon to the Durbodhālokā, as conveyed in its Tibetan translation, states that it was completed in the 10th regnal year of King Cūḍāmaṇivarman (d. 1006, Tib. rGyal po dPal ldan gTsug nor bu).3 Cūḍāmaṇivarman is otherwise known as the ruler of Kedah (Skt. Kaṭāha), which was one of the Malay world’s prime entrepots during his reign. The primacy of Kedah has been revealed by the discovery, ongoing since the mid-twentieth century, of dozens of period sites in the present-day Lembah Bujang region.
The Durbodhālokā, which is preserved in Sanskrit,4 confirms that its author was called Dharmakīrti. He has previously been given the artificial name *Dharmakīrtiśrī, a back-translation from Tibetan Chos kyi grags dpal used in secondary literature, which is not attested in Sanskrit texts. There would, of course, have been no need for the qualifier ‘Serlingpa’ (Skt. Suvarṇadvīpīya) if the author’s name had been clearly distinct from that of the famous 7th-century logician Dharmakīrti, who is quoted throughout the Durbodhāloka.5 Although the Durbodhālokā is wholly scholastic in orientation, its author certainly knows about Buddhist tantras6 and could well have written about Tantric Buddhist figures such as Acala. The most renowned students of Serlingpa Dharmakīrti,7 including Ratnākaraśānti (ca. 970–1045) and Dīpaṃkara, wrote several works on tantric and non-tantric topics.
The Serlingpa Acala comes into view in the wake of Dīpaṃkara’s journey to the Golden Isles (1013–1025) and subsequent mission to Tibet (1042–1055). Nearly all of the extant literary and artistic sources for this form of Acala are end products of this transfer process. The Serlingpa vision is supposed to have been involved in this transmission from the outset. Dīpaṃkara calls upon “Acalanātha endowed with the gang of ten furies” together with “thou, the Dharma King guru”8 to overcome obstacles on his troubled sea journey to the Golden Isles. This episode, as told by Dīpaṃkara to his main disciple Dromtönpa Gyelwé Jungné (ca. 1005–1064, Tib. ’Brom ston pa rGyal ba’i ’byung gnas), appears to bring in Dīpaṃkara’s prospective teacher and āryācalayogin Serlingpa Dharmakīrti ahead of their meeting in the Golden Isles. The role of Acala glimpsed here—to stand firm against the arising of obstacles—represents a typical scenario of use, according to classical Buddhist texts.
2 Acala’s Roots in South and Southeast Asian Buddhism
Among the scriptural sources for Acala, it is the 7th-century Vairocanābhisaṃbodhi that provides the most definite inspiration for the Serlingpa tradition. Acala is described in the Vairocanābhisaṃbodhi as holding a sword and noose and having a single hair braid and one (open) eye. He is situated in a halo of primal light, with a fierce disposition and youthful body.9 Much of this description is paralleled in and probably based on a prose passage in the Trisamayarāja, a predecessor of the Vairocanābhisaṃbodhi.10 Both of these early tantric scriptures were known to Dīpaṃkara,11 but his short hymn to Acala—examined further on—follows the verse description in the Vairocanābhisaṃbodhi almost word for word.12 It should be emphasised here, as Dīpaṃkara does in his writings, that Acala is integral to salvation-oriented praxis and does not personify ferocity alone. Acala’s role is to ward off ‘obstacles’ (Skt. vighna) to the thought of awakening (Skt. bodhicitta), which manifest externally as vexatious gods and internally as disturbances in meditation.13
2.1 South Asian Acalas in the Tradition of the Vairocanābhisaṃbodhi
A variety of forms of Acala accompanies the transmission of the Vairocanābhisaṃbodhi praxis tradition throughout Asia. The depiction of Acala can vary depending on the interpretation of the statement that he ‘stays’ or ‘is stationed’ (Skt. *-stha, Tib. ’dug, Chin. zhu zai
The only Indian depiction of Acala that is linked unambiguously to the scriptural tradition of the Vairocanābhisaṃbodhi is a stone sculpture in Ratnagiri Temple 4 in Odisha. Acala is positioned here to the left of and underneath Mahāvairocana,17 standing with the right leg bent and the sword swung back. This is one of the few known South Asian representations of any form of Acala. There are also three Acala statues from Licchavi-era (ca. 450–ca. 750) Nepal that are consistent with the directives of the Vairocanābhisaṃbodhi, apart from the loss of their swords. One such statue depicts Acala sitting on a rock (Himalayan Art Resources, hereafter HAR, 58341). Another two statues feature left-leaning standing poses (Los Angeles County Museum of Art, hereafter LACMA, AC1995.39.1, and Metropolitan Museum of Art, New York, hereafter MMoA, 1982.220.13, fig. 11.1). One of the last representations produced in the heartland is a miniature in an illuminated manuscript (MMoA 1985.400.10) of the Aṣṭasāhasrikā Prajñāpāramitā. This manuscript, presumed to have been produced in Bengal at a relatively late 12th-century date, has not been preserved in full, and accordingly, its iconographic program and relationship to the Vairocanābhisaṃbodhi—which may be no more than vestigial—remain undetermined. It likewise depicts a standing Acala, who in this case leans to the right.
Two standing Acalas with lost swords and nooses that otherwise conform to the directions of the Vairocanābhisaṃbodhi. Nepal, 7th–10th c.
Left: LACMA AC1995.39.1. Right: MMOA 1982.220.13
On the basis of the small number of South Asian depictions, standing Acalas can be seen to be the norm on the subcontinent. This pose finds agreement in the extant Sanskrit corpus, although much of the early tantric literature, in which Acala is most often discussed, is lost in Sanskrit. In a ritual of sympathetic magic set out in Ānandagarbha’s (fl. 8th c.) Sarvavajrodaya, the āryācalayogin visualises himself standing with left knee bent on the head of an effigy of the ‘obstacle.’ Then,
as he scorches [the effigy] with fire, or smears [it with poison,] no doubt he soon burns even Brahmā or Śakra at that moment.18
Acala also appears as a retinue figure in miscellanies compiled long after the heyday of the Vairocanābhisaṃbodhi. Within this literature, two Uṣṇīṣavijayā-sādhanas stand out for their relatively archaic vision of Acala as a sword-and-noose-wielding door guardian (Skt. dvārapāla) posed in the right-knee-bent (Skt. pratyālīḍha) stance.19 Many of the extant artworks related to these Uṣṇīṣavijayā-sādhanas are, incidentally, products of the Tangut Empire and show close visual affinities with Serlingpa–Dīpaṃkara Acala images created there.20
Distinctive features of the Serlingpa vision of Acala, such as the surrounding ten furies and the figure(s) trampled underfoot, have not been found in the art of the subcontinent. However, one figurine assigned to tenth-century Kashmir, catalogued as Vighnāntaka in a recent publication,21 strongly resembles the Serlingpa Acala, complete with the trampled Gaṇapati. This extraordinary sculpture depicts a muscular, sword-wielding figure struggling with his right foot to control a bulky, squirming, cowering, elephantine deity with four arms. The graphic depiction of Gaṇapati’s torment—rarely seen in India even in Tantric Buddhist art—may reflect unusually strong levels of adversarial sentiment in Kashmir.22 Portable metal sculptures such as these are, however, difficult to locate in place and time on the basis of style alone. Other figurines of Acala trampling Gaṇapati that have been classified as South Asian (e.g. Alain Bordier Foundation ABS 03523) are not readily distinguished from artworks coming out of Western Tibet.
The South Asian repertoire of Acalas includes a posture in which a knee is planted on the ground, with the sword swung as though attacking an enemy in battle.24 However, this precarious pose is neither specified in nor sanctioned by the Vairocanābhisaṃbodhi. Kneeling figures are instead identified with the name and form of Caṇḍamahāroṣaṇa, who represents a divergent tradition consolidated in the Caṇḍamahāroṣaṇatantra. This tantra preserves elements such as mantras from the prior Acala cultus, while departing radically from it by recommending transgressive sexual praxis. In Tibetan nomenclature, however, Caṇḍamahāroṣaṇa is still referred to as the Unwavering (Tib. Mi g.yo ba) in some contexts.25 This has led to widespread confusion of two separate figures that have quite different origins and scriptural bases. The Caṇḍamahāroṣaṇatantra itself only reached Tibet at the end of the period of interest, in the 13th century, and this late tantric tradition needs no further discussion here.
2.2 The Maritime Southeast Asian Acalas
It can be assumed from Tibetan sources that the form of Acala taught by the Golden Isles guru was also known in the Golden Isles per se.26 But has any figure resembling Acala been found there? In present-day Malaysia, there is a paucity of precolonial Buddhist artefacts, whereas in neighbouring Indonesia, there are abundant remains but few attempts at systematic study and identification with reference to primary sources. Helena A. van Bemmel’s 1994 survey of Indonesian door guardian statues identified no images of Acala in Java and Bali. Yet several sword-bearing statues counted in this survey bear strong resemblances to Acala.27 These statues, all from East Java, are much later than the period of interest, but could have been based on old models. The text of the Vairocanābhisaṃbodhi is known to have reached Java by about the 9th century, as several of its verses—drawn from the second chapter, which includes the classical description of Acala—had by then been incorporated into the Sanskrit–Old Javanese Sang Hyang Kamahāyānikan Mantranaya [Holy Mantra Way of the Mahāyāna].28
Of particular interest is a Malaysian statue of a standing, sword-bearing figure unearthed at Batu Lintang in the heart of old Kedah. This defaced, weathered stone sculpture has been variously identified as a door guardian, Durgā or ‘dancing man’, but it may well depict Acala (fig. 11.2).29 It displays a blazing halo and youthful body, and it resembles the MMoA Nepalese Acala (fig. 11.1) in its overall appearance. This statue may bring us as close as we can now get to confirming the existence of Acala worship in Suvarṇadvīpīya Dharmakīrti’s homeland. Other period sculptures from the same region, now too deteriorated to identify, could have depicted Acala.30
Figure identifiable as Acala. Stone, Batu Litang, Malaysia, 8th–10th c., displayed at Muzium Arkeologi Lembah Bujang
Photograph: ‘Bujang Valley Museum,’ Marufish, CC BY-SA 2.0
3 The Conception of the Serlingpa–Dīpaṃkara Acalasādhanas
3.1 The Serlingpa Sādhana (Derge 3059)
The primary sources for the Serlingpa vision of Acala are connected to the Kadam (Tib. bKa’ gdams pa) teaching tradition originating with Dīpaṃkara and Dromtön. The main authority for this vision is the *Āryācala-sādhana (Tib. ’Phags pa mi g.yo ba’i sgrub thabs, Derge 3059), hereafter referred to as the Serlingpa sādhana. Its colophon attributes it to ‘the Golden Isles Guru’, a moniker that presumably refers to Suvarṇadvīpīya Dharmakīrti, but which the author himself is unlikely to have used. The translation is co-credited to Dīpaṃkara and Gönpawa Wangchuk Gyeltsen (1016–1083, Tib. dGon pa ba dBang phyug rgyal mtshan). The Serlingpa sādhana comprises fifty verses on the process of invoking Acala, which includes fifteen Sanskrit mantras. It is meant for a regular meditation routine or for use in contingent scenarios, such as the quelling of disasters. It is Acala’s physical appearance that is of interest here, as well as the question of the sādhana’s authorship.
In the Serlingpa sādhana, Acala is visualised manifesting out of a sun disc on top of Gaṇapati on top of a universal lotus. He has the spotless body of a youth and a black complexion. His bound orange hair hangs down and he stands with the right leg bent and the left leg almost fully outstretched. He has three moon-in-sun-like eyes, looking up and down in the ten directions. His sharp, bright white teeth are clenched. His tongue darts like a flash of lightning. His brow is concentrated in a deep frown. He holds a sword and serpent noose. He is adorned with a tiger pelt and serpent ornaments and a diadem incorporating Akṣobhya and the rest of the buddha pentad. Furious figures emanate from Acala’s body, surrounding him at the meditator’s whim.31 The names—tentatively reconstructed in Sanskrit32—and iconographic features of these ten kings of fury (Skt. krodharāja) are summarised in table 11.1. In short, Acala’s appearance in the Serlingpa sādhana fleshes out the brief of the Vairocanābhisaṃbodhi in considerable detail.
The ten furies in the Serlingpa sādhana: names in Tibetan and reconstructed Sanskrit, colours, and weapons
© Iain Sinclair
Suvarṇadvīpīya Dharmakīrti was active at a time when authors of literary and philosophical works were paying more attention to tantric practice. The contents of the Serlingpa sādhana show some consistency with the oeuvre of the Golden Isles Dharmakīrti, the author of the Durbodhālokā and specialist in the works of Śāntideva (fl. mid. 8th c.). The Durbodhālokā opens with a statement to the effect that those interested in scripture start by honouring their personal deities in order to quell obstacles.33 A key term here, “quelling of obstacles” (Skt. vighnopaśānti, Tib. gegs med par bya ba), is clearly tantric in nature; it occurs shortly after the Acala ritual set out in the Sarvavajrodaya. The sādhana also transmits a mantra that is also given in Śāntideva’s Śikṣāsamuccaya, where it is called the Acalahṛdaya.34 Its stated use is to purify food eaten with the hands. Throughout the known textual corpus, only the Śikṣāsamuccaya has a complete parallel for the mantra in the Serlingpa sādhana; other textual traditions transmit the first part of the mantra separately from the last part. This is an indication that the Serlingpa sādhana is the work of someone who was familiar with Śāntideva’s teaching, as the Suvarṇadvīpīya Dharmakīrti is said to have been.
The classical connection between the will to awaken and the invocation of Acala as its protector is also preserved in the Serlingpa sādhana. It is dedicated to “Acala the protector, king of fury, manifested from the minds of all buddhas of the three times”; it exhorts the practitioner to increase the flow of bodhicitta.35 The altruism of the bodhisattva—a topic treated at length in the Śikṣāsamuccaya—is to be cultivated by the practitioner in the preliminary phase.36 These bodhisattva preliminaries appear to be common elements of procedure from the perspective of mature sādhana praxis, but they may have been more unusual and distinctive when the Serlingpa sādhana was composed, especially in view of the emphasis placed on bodhisattva religiosity in Kadam transmission narratives.
3.2 Ratnaśrījñāna’s Acalasādhana (Peking 4895)
The Serlingpa sādhana states that it was written on the basis of “tantra and the guru’s advice.”37 This raises the question of who taught the praxis of Acala to the Golden Isles teacher. The early biographical account of Dīpaṃkara conveyed in the rNam thar gyas pa [Extensive Life-story] states that he and Serlingpa Dharmakīrti were taught mantra and philosophy by two persons called Ku su lu (supposedly Skt. *Kuśalin, or *Bhusuku).38 However, little is known about these teachers and their teaching outside the Tibetan lineage literature. An ācārya who is said to have met Serlingpa Dharmakīrti at Bodhgaya, and to have later given him his name, is referred to in the Extensive Life-story as *Mahāśrīratna and Rinchenpel (Tib. Rin chen dpal).39 He can be identified with the Sinhalese literatus and mahāpaṇḍita Ratnaśrījñāna (fl. ca. 900–980) in part on the basis that he not only had the same effective name—Rinchenpel being equivalent to Ratnaśrī—but was also active at the right place and time to have been in contact with Serlingpa Dharmakīrti.40 The Sinhalese Ratnaśrī is distinct from the Ratnaśrīs of the Indian subcontinent who collaborated with Tibetans on translations of tantric texts in the early second millennium.41 One Ratnaśrījñāna “extracted”42 a short Acalasādhana preserved in Tibetan translation (Tib. Khro bo mi g.yo mgon gyi sgrub thabs, Peking 4895). The identity of its author and the possibility of it informing the Serlingpa tradition have not yet been explored.
Ratnaśrījñāna describes Acala’s form in a few words: “Acalanātha, dark grey, is one-faced, two-handed; a sword is in the right [hand], the left holds a blood-filled skull.”43 The reference to the skull is odd, and may have resulted from a misreading (e.g. of Skt. *khaḍgakaṃ pāśaṃ as *khaḍga-kapālaṃ). The skull-carrying Acala is likewise a rare sight in the artistic corpus. There is a depiction in the frescoes of Cave 463 of the Mogao Caves (Chin. Mogao ku
The information in the Acalasādhana of Ratnaśrījñāna is too limited to resolve the question of whether it is part of the overall trajectory of Serlingpa Acala praxis. It teaches a simple procedure of sympathetic magic, which is arguably more germane to mid-tenth-century Sri Lanka than to the advanced tantric yoga known to the later Ratnaśrīs. Non-tantric writings of the Sinhalese Ratnaśrījñāna are known to have reached Tibet, although they were not translated into Tibetan.45 The possibility that Acala was worshipped in Sri Lanka during the 10th century, when the Sinhalese Ratnaśrījñāna was active, is concretely indicated by a statuette excavated from the Jetavanavihāra of Anurādhapura. Previously identified as a “Form of Vajrapāṇi or Dharmapāla,”46 it probably depicts Acala. The statuette displays attributes that are consistent with the Serlingpa visualisation: plumed hair, standing posture with right leg bent, a long weapon (a sword or club) swung back with the right hand, and a bundled rope in the left hand. Compelling though these circumstantial associations may be, they are not enough to establish that the Sinhalese Ratnaśrījñāna worshipped Acala and passed on his technique to Serlingpa Dharmakīrti, in view of the fact that the Acalasādhanas of the two authors represent quite different forms of praxis.
3.3 Dīpaṃkara’s Acala Treatises (Derge 3060, 3061, Peking 4892)
Three works on Acala—two hymns and one invocation text (Skt. sādhana)—are ascribed to Dīpaṃkara in the Tibetan canon.47 The two hymns have the same title, Khro bo’i rgyal po ’phags pa mi g.yo ba la bstod pa [Hymn to King of Fury, Noble Acala] (Derge 3060 and 3061), but different lengths and contents. They are formulated in the same metre and are complementary works. The former, Derge 3060, a hymn in twelve stanzas, is concerned primarily with Acala’s appearance.48 The opening stanzas draw on the description of Acala in the Vairocanābhisaṃbodhi, as has already been mentioned. Other stanzas of this hymn echo the wording of the Serlingpa sādhana.49 The second hymn, Derge 3061, is nineteen stanzas long and focuses on the ten furies surrounding Acala. This distinctive set of ten furies is the same as that of the Serlingpa sādhana.
Dīpaṃkara’s Acala invocation manual is the *Krodharājācalasādhana (Tib. Khro bo’i rgyal po mi g.yo ba’i bsgrub pa’i thabs, Peking 4892). Its evocation procedure and mantra repertoire is quite different from that of the Serlingpa sādhana. In the latter Acala is invoked using a dedicated mantra set; Dīpaṃkara, by contrast, employs the generic yogic procedures of higher tantric praxis. There are also slight differences in the appearance of Acala. According to Dīpaṃkara, Acala has two eyes, but the Serlingpa sādhana specifies three eyes. These appear to be different interpretations of the Acala of ‘one eye’ described in the Vairocanābhisaṃbodhi. Whereas the Serlingpa sādhana requires Acala to be visualised “on top of Gaṇapati,” Dīpaṃkara’s sādhana has Acala “trampling the Wicked-Māra-obstacle-horde (Skt. *duṣṭamāravighnagaṇa).”50 Dīpaṃkara’s directive tends to be interpreted in art as two different figures: Māra and an Obstacle deity, namely, Gaṇapati.51 Dīpaṃkara’s sādhana also includes the distinctive Serlingpa feature of the furies emerging from the halo52 but does not specify their number or attributes. This information is relegated to the abovementioned second hymn, Derge 3061.
Dīpaṃkara’s Acala treatises together refashion the Serlingpa invocation process into a more systematic, doctrinaire presentation. These works are solid examples of Indic shastric and literary form, but they appear to have been created for Tibetan consumption. Both the Serlingpa and the Dīpaṃkara writings on Acala are only accessible today as canonical Tibetan texts, and no trace of them has been found in the Sanskrit Buddhist corpus. Likewise, there is little awareness of the Serlingpa Acala tradition in works by South Asian authors. The Acala surrounded by ten furies is briefly described by Mitrayogin (fl. ca. 1180–1197) in an entry in his iconographic compendium, the *Abhisamayamuktāmālā (Tib. Mngon par rtogs pa mu tig gi phreng ba, Peking 5022). Mitrayogin may have relied on his Tibetan associates for information about Acala,53 as he refers to the retinue of ten furies but gives no information about them, in contrast to his other, more elaborate descriptions in the same compendium.54 This leaves the impression that Mitrayogin’s account of Acala drew on informants or visual models rather than a literate South Asian tradition.
3.4 The Ten Furies Surrounding Acala and the *Āryācala nāma dhāraṇī
The Serlingpa and Dīpaṃkara visualisations of Acala are distinguished by the assembly of ten armed figures in the halo. This feature has some precedent in earlier traditions. In one arrangement transmitted in East Asia from the eighth century onwards, but most likely of South Asian provenance, Acala is surrounded by the eight youths Matijvala, Matisādhu and so on.55 They personify the eight syllables na, maḥ, sa, maṃ, ta, va, jrā, and ṇāṃ, which together form a formulaic prefix to the mantras of ferocious Buddhist deities: “homage to all vajras […]”. Independent Acala maṇḍalas such as these fill out the Vairocanābhisaṃbodhi’s requirement for Acala to be visualised in his “own maṇḍala,”56 which Dīpaṃkara reiterates in his first hymn.57
There are also resonances between the Serlingpa tradition and the imagery of ferocious Trailokyavijaya. A “frowning, protruding-fanged, big-headed, tantric follower (Skt. vajrānucara) emerges in blazing light” from Trailokyavijaya’s foot to force the trampled Maheśvara to obey, as described in an oft-cited passage of the Sarvatathāgatatattvasaṃgraha.58 This passage was ritually reenacted59 and there is at least one surviving depiction of it in Indian art, in which the upper bodies of ‘tantric follower’ figures are visible in Trailokyavijaya’s halo.60 This distinctive visual convention is also seen in a clay seal depiction of the Serlingpa Acala found at Tholing Monastery61 (Tib. Mtho lding mgon pa, Chin. Tuolin si
The ten furious figures of the Serlingpa vision are peculiar, as a set, to the worship of Acala. They are different from the well-known group of ten aligned with the Guhyasamāja, Māyājāla and other tantras of their ilk, whose members include not only Yamāntaka, Prajñāntaka, et al. but also Acala himself. The names, complexions and weapons of the ten furies emanated by Acala are a distinct set specified in the Serlingpa sādhana (Derge 3059, verses 20–24) and with more detail in Dīpaṃkara’s second hymn (Derge 3061, verses 9–18). Some of the ten names occur in other scripture—above all in the Sarvatathāgatatattvasaṃgraha—but as a set they are not part of the tantric tradition stemming from the Guhyasamāja.
The classical locus for the ten furies of the Serlingpa tradition is identified here as the third chapter of the long *Āryācala nāma dhāraṇī, which is preserved in a Tibetan translation, ’Phags pa mi g.yo ba zhes bya ba’i gzungs [Mnemonic of Noble Acala] (Derge 631). This work in nine chapters—it is sometimes referred to as such (Tib. gzungs le’u dgu pa)—is the longer of two canonical Tibetan translations with the same Sanskrit title. The shorter work, Derge 963, has no information on the ten furies.63 Comparing the Serlingpa specification of the ten furies with the parallel passage from the long *Āryācala nāma dhāraṇī, it is clear from a few examples that the sādhana shares most of its wording with the dhāraṇī:
This part of the Serlingpa sādhana appears to have been redacted from the long Mnemonic of Noble Acala—that is, from the Tibetan translation itself. Again, the question is raised of how exactly the Serlingpa sādhana was conveyed into the Tibetan language, and whether the text ever had an existence independent of Dīpaṃkara, its redactor and prime mover.
The long *Āryācala nāma dhāraṇī, now identified as the source for the visualisation of the ten furies, reveals other potential connections with the Serlingpa tradition. The description of Acala’s form in the second chapter, which expands on a description that opens the short dhāraṇī, is iconographically consistent with that of the Serlingpa sādhana. It is rich in similes that evoke the natural environment of Maritime Asia:
Space in its entirety is filled everywhere with light [from Acala’s body] like an island similarly permeated by sunrays. The hair on the head is bound up in one [coiffure] and hanging like the follicle of the crown flower tree [(Skt. alāka, Lat. Calotropis gigantea)]. He is squat yet swaying like an earth-rooted wishing tree [(Skt. kalpavṛkṣa)]. He stands with one leg outstretched and one drawn in, like a strongman. He has three eyes, bright red like the flower petals of a flame-of-the-forest tree [(Skt. palāśa, Lat. Butea monosperma)]. His eyes look up and down to burn obstacles above and below, like an asura [looking to] eat a tribute-offering. His bared fangs are half[-moon]-like. His tongue appears to dart quickly, like a flash of lightning in the sky. His brow is furrowed like the veins of the leaf of a white fig tree [(Skt. plakṣī, Lat. Ficus virens)]. His right hand carries a sword [glinting] like the rays cast by a great star. On the left he brandishes a serpent-noose [whirling] like the centre of an ocean maelstrom.64
The flora referred to here are distributed throughout both South and Southeast Asia. The likening of Acala’s splendour to islands in the sun, and his noose to “an ocean maelstrom”—or whatever may be meant by Tib. rgya mtsho’i mching65—are expressions that clearly pertain to a maritime region. As for the short *Āryācala nāma dhāraṇī, it prescribes a magic salve concocted from various spices, many of which are common in Southeast Asia: nutmeg, cloves, cardamom, long pepper and camphor.66 These are potential indications that the *Āryācala nāma dhāraṇī literature circulated in Maritime Southeast Asia or a closely connected part of the world. They add weight to the proposition that the Serlingpa sādhana, a derived work, was not only taught in the Golden Isles but also originated there.
4 The Serlingpa Acala as a Kadam Icon
Dīpaṃkara taught the praxis of Acala to several Tibetans during his lifetime, according to the Extensive Life-story.67 Dīpaṃkara’s Tibetan successors institutionalised it, such that Acala joined Tārā, Avalokiteśvara, and Śākyamuni in the group of Four Kadam Deities (Tib. bKa’ gdams lha bzhi).68 This form of Acala then came to be propagated as a sectarian deity, that is, in connection with a certain cohort of teaching traditions and institutions. Sectarian affiliation in this context does not necessarily entail exclusivity, and the possibility of transmission occurring without the accompanying sectarian framework is reflected in late Tibetan sādhana compendia, in which Acala is just one of hundreds of independently transmissible objects of veneration.69 The designation ‘Kadam Acala’ will then be applied to any figure that displays the distinctive traits of the Serlingpa or the Dīpaṃkara vision, as described in the texts of the ‘Jowo Acala protocol’ (Tib. Jo bo mi g.yo ba’i cho ga),70 namely, the above-mentioned sādhanas and hymns.
Tibetan Buddhism produced a large number of depictions of the Kadam Acala; only a few artworks, deemed representative of the whole, can be discussed here. Over sixty sculptures in bronze, brass or copper display the deity trampling gods; they are therefore identifiable as either Serlingpa or Dīpaṃkara Acalas. The majority are dated to the 12th or 13th centuries. Most are shorter than thirty centimetres in height and would have served as private objects of meditation. The proliferation of these statuettes points to the presence of a widespread, almost grassroots praxis tradition in Tibet during this period.
Some Tibetan depictions were directly informed by the visual sensibilities of the subcontinent. At least one ‘Indian’ depiction of Acala is said to have accompanied Dīpaṃkara’s journey in Tibet.71 And one early painting of Acala, currently in a private collection, is outstanding for its richness, accuracy and apparent close connection to Dīpaṃkara’s circle. This painting was dated by Pratapaditya Pal—with the assistance of carbon dating—to an interval centred on the mid-11th century.72 Its donor portrait is understood to represent Dromtön. The main subject is not, as Pal states, Caṇḍamahāroṣaṇa, but the Acala described by Dīpaṃkara. Likewise, the figures surrounding Acala are not the eight subsidiary deities of the Caṇḍamaharoṣaṇatantra (Śvetācala, Pītācala et al.73) but the ten furies of the Kadam Acalasādhanas. Another early painting that can be regarded as a companion piece—because the proportions and bearing of the central figure are almost identical—depicts the Serlingpa Acala, complete with third eye and single trampled figure (Rubin Museum of Art, hereafter RMA, P1996.20.30, HAR 594).
A common convention in the artistic corpus is the use of fewer than ten figures to represent the group of ten furies. This convention is condoned in the Serlingpa sādhana’s requirement to diffuse the furies ‘at a whim.’ It is often seen where space for detail is limited. In two paintings in which Acala is not the sole focus,74 Acala is portrayed surrounded by four furies. One clay seal mould, from which several impressions were struck,75 depicts Acala together with hook-wielding *Vajratīkṣṇa (Tib. rDo rje snon po) to his left (fig. 11.3). In cases such as these, the smaller single figure or figures metonymically stand for the whole array of ten furies. An arrangement of this kind, which adds a donor portrait on Acala’s right, is fashioned on a plaque affixed to a manuscript cover (MMoA 2005.436.1b, fig. 11.4). Some space-restricted depictions, however, managed to cram in all ten furies, as seen on a wooden manuscript cover (The Walters Museum W.896) and the abovementioned clay sealing from Tholing.
Serlingpa Acala with *Vajratīkṣṇa, encircled by Sanskrit ye dharmāḥ formula. Clay seal impression, Tibet, 11th–12th c.
LACMA AC1993.239.12
Kadam Acala flanked by a krodharāja and a donor figure and surrounded by the seven gems of a Wheel-Turning King (Skt. cakravartin) and a Buddha. Plaque on wooden manuscript cover (detail), Tibet, 12th–13th c.
MMOA 2005.436.1B
5 Tangut Depictions of Acala
5.1 Kadam and ‘Golden Isles’ Teaching in the Tangut Empire
The transfer of Kadam religiosity to the Tangut Empire, which has long been evident in the artistic corpus, is becoming clearer from studies of historical and hagiographic documents. The Kadam seat at Reting (Tib. Rwa sgreng), founded by Dromtön in 1057, prospered under the patronage of a Tangut emperor identified as Renzong (r. 1139–1193, 
). A rich account of the teaching transmitted from Dīpaṃkara to Tangut teachers is the so-called Lyr ghá mý nga jon tshi ngwu 
[Quintessential Instruction of the Forty Banners of Emptiness (sic78)]. The Tangut Buddhist corpus includes works on subjects of interest to Kadam teachers, such as rousing the bodhicitta (Tang. poten ne show 
).79
Information about Dīpaṃkara’s sojourn in Southeast Asia was conveyed in the Tangut language. The Tangut translation of his Satyadvayāvatāra (Tang. Ny khan ghá ó she 
, Tib. Bden pa gnyis la ’jug pa) preserves the colophon of its Tibetan source stating that it was originally composed for the “Golden Isles guru, King (Dharma)pāla” (Tang. Ki lhé kurúr Palá né 
).80 Also recorded in the colophon is the first teaching of the treatise in the presence of Devamati (fl. first half of 11th c., Tang. Dzwyrvamatí 
), a monk who is known to have accompanied Dīpaṃkara to the Golden Isles. A Tangut commentary on the Satyadvayāvatāra even concludes “with a discussion of the circumstances of composition of Atiśa’s treatise in Suvarṇadvīpa.”81 Such documents reveal at least a faint awareness of the Golden Isles as one of the remote sources of Indo-Tibetan Buddhism propagated in Tangut temples.
Although there was a literate Tangut tradition of Kadam teaching, many Tangut depictions of Acala show clear signs of being produced through visual transmission. Their irregularities are more readily explained as graphic confusions than as outcomes of misunderstanding the associated verbal directives. No specific information on Acala has been found yet in either the Tibetan Kadam corpus transmitted in the Tangut Empire or in Tangut-language texts. A work called the Mi mú zi é 
[Mnemonic of the Unwavering] (Chin. *Budong zongchi
The ritual environments of many Tangut Acala depictions have no discernible connection with meditation. At the Eastern Thousand Buddha Caves (Chin. Dong Qianfodong
5.2 Iconographically Atypical Kadam Acalas in the Tangut Empire
A ferocious deity surrounded by ten furies, depicted in a xylograph from Karakhoto now kept at The State Hermitage Museum (x2537)—and often discussed and reproduced in scholarly literature87—resembles the Serlingpa Acala in most respects. The main subject of the xylograph has been repeatedly identified as ‘Acala or Vighnāntaka,’ as it lacks Acala’s characteristic sword and noose. The iconography of the central figure does conform to that of Vighnāntaka, as described in a dedicated sādhana text: “single-faced, two-armed, blue-black-coloured, [holding] a noose with the threatening gesture in the left and, a vajra high with the right hand, fearsome, orange hair upraised.”88 Vighnāntaka was also an independently depicted figure in Tangut art.89 However, no known source for Vighnāntaka’s iconography describes a halo with ten furies.
The subject of the xylograph can be recognised as the Serlingpa Acala, copied in a graphically confused way that omits the characteristic sword and noose. In many depictions of this form, Acala’s sword is posed swung back behind the head, the blade obscured by his plume of hair. Here the artist seems to have only noticed the hilt of the sword and rendered it as a vajra. The thin outline of the noose could likewise have been lost against the busy background of flames. There may be another factor in this accident: several early statues of Acala, as noticed above, have lost their fragile sword and noose elements due to wear and tear, and some Buddhist paintings were ‘portraits’ of pre-existing physical objects.90 These same irregularities, the missing sword and noose, are also seen in the Tangut painting x2376 of the Hermitage collection,91 which likewise depicts a figure identifiable as a Kadam Acala, with furies in the halo and trampled Gaṇapati.
There is no doubt that the central figure of the Hermitage xylograph was meant to represent Acala. In the header section Acala’s signature mantra Oṃ caṇḍamahāroṣaṇa hūṃ phaṭ is given in Rañjanā script. It is also conveyed just below in Chinese transcription: an zanda mohe langzhana hong bada (
The mantra of Acala in the xylograph’s header is flanked by two phrases: “Long live the emperor” (Chin. huangdi wansui
Another atypical depiction of Acala is the Hermitage Museum painting x2376.93 It displays some Kadam features, such as furies in the halo, but the main figure is kneeling in the dynamic pose. The iconography of this painting most likely absorbed the imagery of the kneeling form, Caṇḍamahāroṣana, which Tangut tapestry workshops had begun to produce at a relatively late stage.
5.3 Tangut Silk Tapestry Depictions of Acala
Several fine silk tapestries depicting Acala were produced in the Tangut Empire. There is a large secondary literature on these tapestries, to which just a few observations are added here. Firstly, it is clear that three of the four tapestries, for which various identifications have been offered,94 all portray the Dīpaṃkara form of Acala. The two-eyed face and the two different figures trampled underfoot identify it as such, as do the other common features of the Kadam Acalas. The three Tangut tapestry portrayals comprise the most accessible work, Cleveland Museum of Art 1992.72 (HAR 59257, fig. 11.5), and two works kept for a time in the Potala Palace (Tib. Pho brang Po ta la, Chin. Budala gong
Dīpaṃkara tradition (Tib. Jo lugs) Acala. Silk tapestry, Tangut make, 12th–13th c., detail (with digital colour enhancement).
Cleveland Museum of Art 1992.72
The patent skill of the Tangut textile workshops may have been a factor in attracting the patronage of the fourth tapestry (HAR 99102) depicting Caṇḍamahāroṣaṇa. This well-known tapestry, created for the Sakyapa hierarch Jetsün Drakpa Gyaltsen (1147–1216, Tib. rJe btsun grags pa rgyal mtshan) in around the year 121096 and currently in the Potala collections, is often touted as an exemplary “Tangut commodity,”97 although its subject is a late outlier in the mainstream of Tangut Acala worship.
5.4 Kadam Acalas in Tangut Cakrasamvara–Vajravārāhī Paintings
Paintings that feature Acala and Cakrasaṃvara–Vajravārāhī as retinue figures, or with the former in the retinue of the latter, have been found at a number of sites connected with the Tangut Empire. The erotically charged Cakrasaṃvara tantric system is associated much more with the Kagyü School (Tib. bKa’ brgyud pa) than the Kadampas, and although this tantric system was well known to Dīpaṃkara, it was not prominent in his Tibetan teaching career. The historical information encoded in the composition and portraiture of these paintings needs extensive unpacking; just a few remarks will be offered here. A superb painting recovered from the western Baisigou Square stūpa (Chin. Baisigou fangta
Two-armed Cakrasamvara–Vajravārāhī with five retinue figures including the Kadam Acala (middle, lower register). Painting (detail), Baisigou Square stūpa, 12th–13th c.
Faculty collections, Northwestern University Libraries, Rob Linrothe image collection, accessed July 26, 2021. ARK:/81985/N2TM73V1M
Most remarkable are the couplings of Acala—stalwart of the chaste Buddhist Tantrism advocated by Dīpaṃkara100—with Vajravārāhī, naked goddess of erotic tantric praxis propitiated by the Kagyü school. In a Karakhoto fragment of a Tangut painting that is presumed to have depicted Vajravārāhī or Vajrayoginī (x3550),101 the Serlingpa Acala is placed in the upper corner, a position of high respect.102 Another Vajravārāhī painting, probably of Tangut origin (HAR 35845), places Dīpaṃkara and Dromtön above the main subject and Acala in the lower register. To put this unusual pairing in context, it can be noted that one Vajravārāhī praxis lineage in the Tangut Empire claimed to have inherited teachings on the goddess from Dīpaṃkara via Dromtön. This lineage accepts the well-known story of Dīpaṃkara coming to Tibet to atone for expelling a Vajrayoginī/Vajravārāhī devotee from Vikramaśīla monastery. Tibetan sources identify this devotee as Maitrīpā (Advayavajra). But the Tangut transmission adds a twist: it was a manifestation of Vajrayoginī that encouraged Dīpaṃkara to go on his journey across the Himalayas. In this way Dīpaṃkara was repositioned as progenitor of a sensuous religious practice which, by all other accounts, he had opposed at Vikramaśīla.103
The Tangut recipients of Dīpaṃkara’s lineage also claimed to have inherited instruction on the Cakrasaṃvara-system Ḍākinī of the Four Syllables (Skt. akṣaracatuṣṭaya) stemming from a South Indian yogin who wandered throughout Central Asia in the 11th century. He is referred to in Tangut, Tibetan and Chinese hagiographies104 by various names: Padampa Sangyé (fl. mid–late 11th c., Tib. Pha dam pa sangs rgyas), Kṛṣṇapāda the younger (Tang. Tsy na réwr tsen 
, Chin. Xiao heizu 
, Chin. Gemalashila 
, Chin. Mingdeliwa
6 Conclusions
The ascription of a certain form of Acala to a teacher from the ‘Golden Isles,’ Serlingpa Dharmakīrti, is supported by several indications in artistic and textual corpora. Images identifiable as Acala are extant in present-day Malaysia and Indonesia; their classical source, the Vairocanābhisaṃbodhi, is known to have been studied there. The source for the ten furies invoked in the Serlingpa sādhana, the *Āryācala nāma dhāraṇī, shows signs of being composed in Southeast Asia or some other maritime region. The praxis of the Serlingpa Acala also has faint connections with the religiosity of the bodhisattva taught by Śāntideva, which the Golden Isles guru is said to have mastered. There is not yet enough information to determine whether Ratnaśrījñāna’s short Acalasādhana (Peking 4895) is the work of the Ratnaśrī who taught the Golden Isles guru. The evidence accompanying the transfer of Serlingpa Acala worship to the Tibetan Plateau is, overall, consistent with the claims that are made about the origins of this tradition in Tibetan writings.
The praxis of the Serlingpa Acala was amplified and reworked by Atiśa Dīpaṃkara during its transmission to Tibet, resulting in minor variations. Acala is visualised, by means of a more systematic invocation protocol, with two eyes instead of three, and two different figures underfoot instead of one. Acala’s classical role as a suppressor of ‘obstacles’ in the sense of hindrances to awakening was emphasised, and the sense of suppressing the ‘obstacle’ of Hindu competition de-emphasised, even as vivid depictions of Acala trampling Gaṇapati proliferated in Central Asia. When the worship of the Serlingpa Acala spread from Tibet into the Tangut Empire, its literate bases receded from view, but the continuation of a learned tradition is still visible in productions such as the iconographically precise Tangut tapestries.
The Golden Isles Acala experienced a late rise to prominence in the Hexi Corridor—well after the deity fell out of view in South and Southeast Asian Buddhism—and a subsequent rapid decline at the start of the Mongol era. Having made the jump into the Sino-Tibetosphere in the mid-11th century, the deity’s fortunes were tied to those of the Kadam sectarian tradition, which upheld the legacy of the Golden Isles guru and Dīpaṃkara. However, this form of Acala also was also accepted briefly among Kagyüpas who were in close contact with the Kadam tradition. In future explorations of the Golden Isles Acala’s journey, there is room for more comprehensive art-historical analysis, and for further work on the primary sources, which would benefit from the understanding that the deity’s prime movers belonged to a transregional Sanskritic world, which was larger and more diverse than ‘Indian’ or even ‘Indo-Tibetan’ Buddhism.
This chapter was written with support from a Käte Hamburger Kolleg fellowship at CERES, Ruhr-University Bochum. I thank the editors—Yukiyo Kasai, Henrik H. Sørensen, and Carmen Meinert—for encouragement to publish and for help with locating relevant sources. I also thank Gudrun Bühnemann, David Templeman and other colleagues, who are credited individually in the footnotes, for their comments on earlier drafts; all errors are, however, mine alone. This chapter is a companion piece to the author’s “Dharmakīrti of Kedah: His Life, Work and Troubled Times,” Temasek History Research Centre Working Paper 2 (Singapore: ISEAS–Yusof Ishak Institute, 2021). Transcriptions of Tangut use acute diacritics to denote rising tones and otherwise conform to Marc Miyake, “Tangut Phonetic Database Version 1.0,” 2015, accessed October 22, 2020, http://www.amritas.com/Tangut/tangutdb-1-0.pdf.
Peter Skilling, “Many Lands of Gold,” in The Golden Land Suvarnabhumi: The New Finding for Suvarnabhumi Terra Incognita, ed. Ian Glover et al. (Bangkok: GISTDA and BIA, 2019), 198, asserts that “The quest to localize Suvarṇabhūmi as a single static site or nation state goes against the logic of history.” However, the self-understanding of one region as the Golden Land/Isles is well attested in the period of interest, particularly in the well-known Sumatran highlands inscriptions of the Tantric Buddhist rulers Kṛtanagara and Ādityavarman, dated 1286 and 1356 respectively.
Peter Skilling, “Dharmakīrti’s Durbodhāloka and the Literature of Śrīvijaya,” Journal of the Siam Society 85.1–2 (1997): 191–192, refers to Derge 3794, 254a.
Francesco Sferra, “Sanskrit Manuscripts and Photographs of Sanskrit Manuscripts in Giuseppe Tucci’s Collection,” in Sanskrit Texts from Giuseppe Tucci’s Collection Part I, ed. Francesco Sferra (Roma: IsIAO, 2008), 51 n. 83, documents one manuscript kept in Tibet. The first three chapters have been edited by Guan Di
Quotations in the Tibetan translation of the Durbodhālokā have been listed and traced by Isoda, Hirofumi
The quarter-verse pañcakāyātmako buddhaḥ is quoted from Mañjuśrīnāmasaṃgīti 6.18a by name in the Durbodhālokā (Isoda, “On the Durbodhāloka,” 103, no. 8–2; Derge 3794, 251a), and there follows a reference to the *Yogatantra (Tib. rnal ’byor pa’i rgyud).
For further background see Sinclair, “Dharmakīrti of Kedah,” 5.
’Brom-ston rGyal ba’i ’byung gnas, Jo bo rje lha gcig dpal ldan A ti sha’i rnam thar bla ma’i yon tan [Merit of His Eminence, the Singular Deva Majesty Atiśa, Liberation guru] (Zi ling: mtsho sngon mi rigs dpe skrun khang, 1994), 30: mi g.yo mgon po khro bo bcu yi tshogs bcas dang/ […] chos rgyal bla ma khyed kyi dpung gis bzlog tu gsol. For an alternative translation see, among others, Hubert Decleer, “Atiśa’s Journey to Sumatra,” in Buddhism in Practice, ed. Donald S. Lopez Jr (Princeton: Princeton University Press), 536.
The relevant passage, not known to be extant in Sanskrit, has been conveyed both in Tibetan (Derge 494, 166a) and in Chinese (T. 848.18, 7b), with small differences in content. For translations into English, see, respectively, Stephen Hodge, The Mahā-Vairocana-Abhisaṃbodhi Tantra, with Buddhaguhya’s Commentary (Richmond: Curzon, 2003), 113; Rolf W. Giebel, The Vairocanābhisaṃbodhi Sutra. Translated from the Chinese (Taishō Volume 18, Number 848) (Moraga: BDK America, 2005), 31. In Sanskrit texts this scripture has the title Vairocanābhisaṃbodhi and it is usually designated a tantra. The title *Mahāvairocanasūtra is a back-translation of Chin. Dari jing
Derge 502, 219a: de’i ’og tu mi g.yo ba mig gcig pa zhes bya ba bgegs tshar gcod pa/ rin po che’i brag la bzhugs pa mig dmar po gcig yod pa/ kun nas ’bar bas bskor ba nag cing ljang ba/ khros shing ’khrugs pa/ gzhon nu ral pa gcig yod pa/ lag pa na ral gri dang/ zhags pa thogs pa bri bar bya’o. On the priority of this text over the Vairocanābhisaṃbodhi, see Tanaka, Kimiaki
Dīpaṃkara refers to the Vairocanābhisaṃbodhi as a representative caryātantra (Tib. spyod pa’i rgyud) and to the Trisamayarāja as a kalpa (Tib. rtog pa’i rgyud) in his *Bodhipathapradīpapañjikā (Tib. Byang chub lam gyi sgron ma’i dka’ ’grel, Derge 3948, 287a–b).
Derge 3060, 116a, verses 1–2c, parallel the Vairocanābhisaṃbodhi, Derge 494, 166a: mi g.yo zhes bya’i khro bo ni/ […] /byis pa’i gzugs can ’gying bag can.
See also Rob Linrothe, Ruthless Compassion: Wrathful Deities in Early Indo-Tibetan Esoteric Buddhist Art (Boston: Shambhala and Serindia, 1999), 152.
Standing Acalas portrayed in Japanese iconographic compendia are referenced in Lokesh Chandra, Dictionary of Buddhist Iconography, vol. 1 (New Delhi: International Academy of Indian Culture and Aditya Prakashan, 1999), 41.
The aforementioned Tibetan translation of the Trisamayarāja, Derge 502, refers to Acala “seated on a jewelled mountain” (Tib. rin po che’i brag la bzhugs pa). Acala is said to have a seated pose elsewhere in the Vairocanābhisaṃbodhi; see e.g. Hodge, Mahā-Vairocana-Abhisaṃbodhi, 228; Giebel, Vairocanābhisaṃbodhi, 97.
In addition to the Acala images from Odisha, Nepal, and Kashmir referred to in this chapter, at least four others in Eastern India should be noted. See Thomas E. Donaldson, Iconography of the Buddhist sculpture of Orissa (New Delhi: Indira Gandhi National Centre for the Arts, 2001), vol. II, figs. 76 (crouching), 95 (crouching), 344 (standing). I thank Samuel Grimes for bringing the former and latter to my attention. A statue of Acala(?) in crouching posture is kept in the State Archaeological Museum, Kolkata, accession number S249 (Rob Linrothe Image Collection, Northwestern University Libraries, ark:/81985/n2z89476x, accessed July 26, 2021).
The Ratnagiri guardian figure is positioned in effect to the southwest (Tib. bden bral gyi ni, Chin. nielidi
Sarvavajrodaya of Ānandagarbha, ed. Mikkyō Seiten Kenkyūkai
Marie-Thérèse de Mallmann, “Notes d’iconographie tântrique: III—A propos du «Fudō bleu»,” Arts Asiatiques 9.1–2 (1962): 76, in this regard notices Benoytosh Bhattacharyya, ed., Sādhanamālā (Baroda: Oriental Institute, 1925), I, 418–420, nos. 211, 212.
For examples see Carmen Meinert, “Creation of Tantric Sacred Spaces in Eastern Central Asia,” in Buddhism in Central Asia I, ed. Carmen Meinert and Henrik H. Sørensen (Leiden, Boston: Brill), 254–260.
Jan van Alphen, Cast for Eternity. Bronze Masterworks from India and the Himalayas in Belgian and Dutch Collections (Antwerp: Ethnographic Museum, 2005), 74–75, no. 16. The text itself states: “This bronze represents Acala.” See also HAR 35924.
Instances of ‘Hinduisation’ and their Buddhist counterreactions—often regarded as primarily Nepalese developments—are discussed with reference to precedents in Kashmir by Iain Sinclair, “The Appearance of Tantric Monasticism in Nepal,” (PhD diss., Monash University, 2016), 119–120, 157.
“Nila-Acala or “Blue Acala”—the “Immovable One”, Late Pala Style,” accessed July 26, 2020. https://tibetmuseum.app/index.php?w=coll&cat=S&id=234.
See e.g. Sindhu, “Acala, the ‘Krodharaja’,” 438, fig. 1.
Terms such as *Caṇḍācala (Tib. gTum mo mi g.yo ba) and *Yoganiruttara-Acala (Tib. Mi g.yo ba rnal ’byor bla na med pa) refer more or less clearly to Caṇḍamahāroṣaṇa, but the designations Two-Armed Acala (Tib. Mi g.yo ba pPyag gnyis pa) and Blue Acala (Tib. Mi g.yo ba sngon po) are ambiguous.
Helmut Eimer, “Life and Activities of Atiśa Dīpaṃkaraśrījñāna: A Survey of Investigations Undertaken,” Journal of the Asiatic Society 27.4 (1985): 8, infers from these sources that Dīpaṃkara first met the Serlingpa Guru at Bodhgaya, yet “all extensive biographies speak about a sea voyage to […] Suvarṇadvīpa and his stay there for twelve years,” where any significant teaching would have taken place.
Helena A. van Bemmel, Dvārapālas in Indonesia: Temple Guardians and Acculturation (Rotterdam, Brookfield: A.A. Balkema, 1994), figs. 53, 89, and 101, display strong resemblances. A tacit suggestion that the Indonesian door guardian repertoire may include depictions of Acala was offered by Ito Naoko
For details see Ishii Kazuko
Michael Sullivan, “Excavations in Kedah and Province Wellesley, 1957,” Journal of the Malaysian Branch of the Royal Asiatic Society 31.1 (1958). Regarding more recent scholarship on the “objek seperti lelaki yang sedang menari,” see Adnan Jusoh and Zuliskandar Ramli, “Makara di Lembah Bujang dan kaitannya dengan sosiobudaya masyarakat Nusantara [The Makara in Lembah Bujang and Its Relation with the Nusantaran Sociocultural Community],” in Seminar Antarabangsa: Sejarah Lisan dalam Perspektif Warisan dan Budaya Malaysia-Satun [International Seminar: Oral History in Cultural Perspective and Malaysia-Satun Heritage] (Satun: Persatuan Sejarah Lisan Malaysia, University Utara Malaysia and University Teknologi Mara, 2015), 5, 12.
See, for instance, Nasha Rodziadi Khaw, Mokhtar Saidin, and Zuliskandar Ramli, “Tinjauan Umum Berkenaaā Penemuan Arca Agama Di Kedah Tua [General Overview on Findings of Religious Sculptures in Kedah Tua],” Jurnal Arkeologi Malaysia [Malaysia Journal of Archaeology] 31.2 (2018): 64, Foto 1, for a stone sculpture in which only a head and sword-like shape can now be discerned.
Derge 3059, 113a, with readings silently adopted from Peking 3883 130ab: /de ’og gaṃ spros tshogs bdag ste/ /paṃ spros sna tshogs padmar blta/ /de steng raṃ las rta bdun pa/ […] /dri ma med pa gzhon nu’i sku/ […] /gzhon zhing sku rgyas sku mdog gnag/ /’od kyis nam mkha’ ma lus ’gengs/ /dbu skra dmar ser bcings shing ’phyang/ […] /g.yon brkyang g.yas ni cung zad bskum/ […] /spyan gsum nyi zla bzhin du dmar/ /steng ’og phyogs bcur blta bar mdzad/ /mche ba rno dkar rab tu gtsigs/ /ljags ni glog ltar rab tu ’gyu/ /smin ma khro gnyer rab tu bsdu/ /phyag gnyis ral gri sbrul zhags bsnams/ /stag dang sbrul gyi rgyan gyis brgyan/ /mi bskyod la sogs dbu rgyan can/ […] /khro rgyal sna tshogs sku las spro/ /ji ltar ’dod pas yongs su bskor/.
This set of ten is not known to be attested in Sanskrit texts. Its classical locus, the long Āryācala nāma dhāraṇī (Derge 631), which is now available only in Tibetan translation, is discussed in the following section. Some of the furies’ names resemble those of vajra deities in the maṇḍala of the Trisamayarāja (cf. Tanaka, “Trisamayarāja-tantra,” 118): Vegavajra, Caṇḍavajra, etc.
Guan, “Abhisamayālaṃkāravivṛti and Its Commentary Durbodhālokā,” 94: śrutādipravṛttānāṃ ca vighnopaśāntaye sveṣṭadevatān namaskaroti; cf. Derge 3794, 140b.
Śikṣāsamuccaya 7, ed. Cecil Bendall, Çikshāsamuccaya: A Compendium of Buddhist Teaching Compiled by Çāntideva (St. Petersburg: Académie Impériale des Sciences, 1902), 141, with readings silently adopted from Derge 3940, 79b (Chin. T. 1636. 32, 102a, omits bhrāmaya hūṃ): namaḥ samantavajrāṇāṃ trāṭa | amogha caṇḍamahāroṣaṇa sphoṭaya hūṃ | bhrāmaya [bhrāmaya] hūṃ | trāṭ hūṃ [hāṃ] māṃ | oṃ balaṃ dade tejomālini svāhā. For the mantra in the Serlingpa sādhana see Derge 3059, 114b–115a. Regarding the variant baliṃ dade see Giebel, Vairocanābhisaṃbodhi, 260 n. 162.
Derge 3059, 113b: /dus gsum sangs rgyas kun gyi bdag/ /thugs sprul khro rgyal mi g.yo mgon/ […] /bsam gtan ting ’dzin mos pa bya/ /byang chub sems kyi rgyun spel phyir/. The term *bodhicittadhārā (Tib. byang chub sems kyi rgyun) is suggestive of, though not definitively aligned with, non-celibate Tantrism.
Derge 3059, 113b: /byams dang snying rje dga’ ba dang/ /btang snyoms tshad med rnam pas ni/ /tshangs pa’i gnas bzhi bsgom par bya/.
Derge 3059, 113ab: /rgyud dang bla ma’i man ngag la/ […] /khro rgyal mi g.yo’i sgrub thabs [’]bri/.
rNam thar gyas pa, §§19–21, 24, 41, 44, 57, summarised and edited by Helmut Eimer, Rnam thar rgyas pa: Materialen zu einer Biographie des Atiśa (Dīpaṃkaraśrījñāna) (Wiesbaden: Harrassowitz, 1979), I, 158–160, 164–165, 169; II, 11–14, 16, 28–29, 38. It is far from certain that kuśalin is the word underlying ku su lu, which, as David Templeman has communicated (July 26, 2021), refers to an ‘eater, defecator and sleeper’, a layabout, i.e. bhusuku. See also Herbert Guenther, The Life and Teaching of Nāropa (London: Oxford University Press, 1963), 34, n. 1.
Eimer, Rnam thar rgyas pa, §§52–55, I, 167–168; II, 38.
Sinclair, “Dharmakīrti of Kedah,” 4.
Jean Naudou, Buddhists of Kaśmīr (Delhi: Agam Kala Prakashan, 1980), 169–170 n. 42, discusses these later Ratnaśrīs.
Peking 4895, 248b: paṇḍi ta chen po […] rad na shrī dznyānas phyung ba rdzogs so.
Peking 4895, 248a, with glosses omitted: mi g.yo mgon po […] sku mdog […] dud kha zhal gcig phyag gnyis pa g.yas ral gri g.yon na thod khrag bsnams pa.
Peng Jinzhang
Dimitrov, Legacy of the Jewel Mind, 7–8, 555 n. 178, points out that at least two manuscripts of the Cāndravyākaraṇapañjikā attributed to Ratnamati alias Ratnaśrījñāna have been preserved in Tibet. The existence of this pañjikā and its author Rinchen Lödrö (Tib. Rin chen blo gros) was known to Tibetan scholars.
Ulrich von Schroeder, Buddhist Sculptures in Sri Lanka (Hong Kong: Visual Dharma Publications, 1990), 298–299, pl. 84E, reproduced with the kind permission of Ulrich von Schroeder in Sinclair, “Dharmakīrti of Kedah,” 17, fig. 8.
The praxis of the *Caṇḍamahāroṣaṇaparamārthasādhana (Peking 4896), purportedly written by Dīpaṃkara (slob dpon chen po bhangga la’i mkhas pa dī pangka ra shrī dznyā nas mdzad pa), involves subtle yoga and overt sexuality and generally shows no sign of connection to the Serlingpa tradition. Its authenticity should be examined in connection with other works translated into or otherwise created in Tibetan by bla ma Tāranātha (1575–1634) in conjunction with his paṇḍita Nirvāṇaśrī (fl. 1596–?).
The bulk of the text has been translated by Amy Heller, “On the Development of the Iconography of Acala and Vighnantaka in Tibet,” in Embodying Wisdom: Art, Text and Interpretation in the History of Esoteric Buddhism, ed. Robert N. Linrothe and Henrik H. Sørensen (Copenhagen: Seminar for East Asian Studies, 2001), 215–216. Verse 7d on the furies was not translated; see Derge 3060, 116b: /sprul pa’i khro rgyal sna tshogs sku las ’phro/.
Derge 3060, 116a, v.5b, parallels the Serlingpa sādhana, Derge 3059, 114a: /stag dang sbrul gyi rgyan gyis rnam par brgyan/ […] /mi bskyod la sogs rigs lngas dbul brgyan/.
Derge 4892, 245a: gtum pa’i bdud dang bgegs kyi tshogs du ma ’og du mnan pa.
In the 19th century Tibetan Rin lhan [Jewel Manual] Acala is to be visualised standing on two figures described as Maheśvara—this identification is not warranted in the Serlingpa-Dīpaṃkara tradition—and Gaṇapati. See Martin Willson and Martin Brauen, Deities of Tibetan Buddhism: The Zürich Paintings of the Icons Worthwhile to See (Boston: Wisdom Publications, 2000), 282–283, no. 174.
Peking 4892, 240ab: […] rang lus ’bar ba’i sgo nas ’byung ba’i khro bo’i rgyal po ma lus pa phyag mtshan bsnams pa sna tshogs pa thogs pa’i tshogs […] rnam par bsgom par bya’o.
Siglinde Dietz, Die Buddhistische Briefliteratur Indiens. Nach dem tibetischen Tanjur herausgegeben, übersetzt und erläutert (Wiesbaden: Otto Harrassowitz, 1984), 72, gives some of Mitrayogin’s dates and Tibetan connections.
On these descriptions see Masahide Mori
Their classical source is the Sheng wudong zun yizi wusheng da tongzi mi yao fa pin
On the svamaṇḍala directive see Linrothe, Ruthless Compassion, 153; Giebel, Vairocanābhisaṃbodhi, 56.
D 3060, 116a, verse 1d: rang gi dkyil ’khor ’od la bzhugs. See also Heller, “Acala and Vighnantaka.”
Sarvatathāgatatattvasaṃgraha 6, ed. Horiuchi Kanjin
The mantra Oṃ pādākarṣaṇa vajra hūṃ uttered during the trampling of Maheśvara is in the repertoire of the Shingon Jūhachidō
Linrothe, Ruthless Compassion, 205–207, identifies two figures in the halo of the Nālandā Museum Trailokyavijaya (Archæological Survey of India, accession number 00002; “Ghostly warrior-attendant near the proper left foot of damaged Trailokyavijaya. Detail,” Rob Linrothe Image Collection, accessed July 26, 2021) as Sumbha and Nisumbha. Mori Masahide
For a photograph see Li Yizhi
Eimer, Rnam thar rgyas pa, §§259–267, I, 249–266.
The Tibetan translation of the shorter dhāraṇī text was edited and translated into Japanese by Nakayama Shyorei
Derge 631, 70b (silently emended): nam mkha’ ma lus pa thams cad du ’od kyis ’gengs pa ni dper na rta bdun pa’i ’od kyis gling kun tu khyab pa bzhin no/ /dbu skra gcig tu bcings shing ’phyang ba ni dper na shing a la ka’i ’bras bu bzhin no/ /sku cung zhig ’gying ba dang g.yo ba ni dper na dpag bsam gyi shing yongs ’du sa brtol bzhin no/ /zhabs brkyang bskum du bzhugs pa ni dper na gyad kyi ’gros bzhin no/ /spyan gsum shin tu dmar ba ni dper na pa la sha’i me tog bzhin no/ /spyan steng ’og tu blta ba ni steng ’og gi bgegs bsreg pa ste/ dper na lha ma yin gyi bya gtor ma za ba bzhin no/ /mche ba gtsigs pa ni dper na zla ba phyed pa bzhin no/ /ljags shin tu ’khyug pa ni dper na nam mkha’i glog bzhin no/ /smin ma khro gnyer du bsdus pa ni dper na shing pa la kṣī lo ma bzhin no/ /phyag g.yas na ral gri bsnams pa ni dper na gza’ chen po’i ’od lhung ba bzhin no/ /sbrul gyi zhags pa g.yon na bsnams pa ni dper na rgya mtsho’i mching bzhin no. I thank Dylan Esler for some emendations to the text and for suggestions taken up in the translation.
The expression rgya mtsho’i mching is paraphrased by Dīpaṃkara as brgya mtsho’i dba’ rlabs in his hymn, Derge 3061, 117a. In the view of David Templeman (personal communication, July 26, 2021) it should mean “the ‘eye’ of the whirlpool.”
Derge 931, 58a: dzā ti pha la dang / […] / li shi dang / sug smel dang / […] / kakko la dang / […] / ga bur rnams […]. Each spice in this assortment has been identified and discussed by Nakayama, “Sei Fudō fundo (son) no darani,” 64–67.
Eimer, Rnam thar rgyas pa, §§369, 376, I, 274–275, 276–277; II, 313, 318.
These four are said to have been Dīpaṃkara’s primary (exoteric) yidams; Eimer, Rnam thar rgyas pa, §7, I, 153, II, 5. On their systematised sādhanas, see Willson and Brauen, Deities of Tibetan Buddhism, nos. 174, 133, 103, 39, respectively.
The entry on Acala in the sNar thang brgya rtsa [Narthang Hundred] of Chim Namkha Drakpa (ca. 1210–1285, Tib. mChims Nam mkha’ grags pa) is consistent with the Serlingpa sādhana in particular; see Jeff Watt, “Item: Achala (Buddhist Deity)—Blue, Standing,” HAR, 2019. The Kadam Acalanātha (Tib. Mi g.yo ba sngon po bka’ gdam lugs) described in the 19th-century Jewel Manual follows the Dīpaṃkara tradition (Tib. jo lugs); see Willson and Brauen, Deities of Tibetan Buddhism, 282–283.
Eimer, Rnam thar rgyas pa, I, 336; II, 318.
Eimer, Rnam thar rgyas pa, I, 293; Per K. Sørensen and Guntram Hazod, Rulers on the Celestial Plain: Ecclesiastic and Secular Hegemony in Medieval Tibet (Vienna: Verlag der Österreichischen Akademie der Wissenschaften, 2007), II: 375 n. 23*.
Pratapaditya Pal, “An Early Tibetan Mandala of Ekallavira Achala in a Private Collection: An Art Historical Analysis.” Asian Art, 09.09.2013, accessed May 3, 2020. https://www.asianart.com/articles/achala/index.html.
There is certainly a striking visual similarity between the Caṇḍamaharoṣaṇa surrounded by eight figures, as described in the eponymous tantra, and the Serlingpa–Dīpaṃkara Acala surrounded by ten furies. The possibility that the former vision draws on the latter—the opportunities for direct contact would have been limited—needs separate study.
See HAR 69454 (Asian Art Museum of San Francisco 1997.2), a thangka dated “approx. 1300–1400” but probably of the 12th century, and HAR 77196 (Musée Guimet, accession number unknown), reproduced in Linrothe, Ruthless Compassion, pl. 3.
Other sealings made from the same mould as LACMA AC1993.239.12 include HAR 44376, 65512 and HAR 65513 (RMA accession numbers unknown), Sothebys sale HK1038 lot 6056, a sealing in a private collection in Australia, and a sealing photographed by Rob Linrothe in Langmi, Zanskar (ark:/81985/n2f76823r).
Maho Iuchi, “A Note on the Relationship between the Bka’ gdams pa School and Mi nyag/Xixia,” Journal of Tibetology 8 (2012): 60.
Tsuguhito Takeuchi and Maho Iuchi, Tibetan Texts from Khara-Khoto in the Stein Collection of the British Library (Tokyo: The Toyo Bunko, 2016), 11–13. Karakhoto has also yielded Tibetan translations of works authoritative for the Kadam school: the Bodhicāryāvatāra (ibid. 129, no. 253), Dīpaṃkara’s *Bodhipathapradīpa (ibid. 56, no. 80, Derge 3947) and pañjikā (ibid. 90, no. 153, Derge 3948), and the Bodhisattvasaṃvaraviṃśikā pañjikā of Dīpaṃkara’s teacher Bodhibhadra (ibid. 121, no. 234, Derge 4083).
Kirill Solonin, “Dīpaṃkara in the Tangut Context. An Inquiry into the Systematic Nature of Tibetan Buddhism in Xixia (Part 2),” Acta Orientalia Academiae Scientiarum Hungaricae 69.1 (2016): 20, states that the title is “relatively easy to reconstruct as *stong nyid rgyal mtshan bzhi bcu’i rnam pa’i man ngag.” However, Tang. jon (
) should be understood as rendering Tib. mtshan in an expression such as stong pa nyid kyi mtshan nyid, Skt. *śūnyatālakṣaṇa, rather than “banner, flag” (Tib. rgyal mtshan, Chin. chuang
See e.g. Nishida Tatsuo
Derge 4467, 6b: gser gling rgyal po gu ru pha la yis. This phrase is taken to mean ‘Golden Isles rājaguru Dharmapāla’ by Jan A. Schoterman, “Traces of Indonesian Influences in Tibet,” in Esoteric Buddhism in Mediaeval Maritime Asia, ed. Andrea Acri (Singapore: ISEAS Publishing, 2016), 118. For the Tibetan and Tangut texts, see Kirill Solonin and Kuowei Liu, “Atiśa’s Satyadvayāvatāra (Bden pa gnyis la ’jug pa) in the Tangut Translation: A Preliminary Study,” Journal of Indian Philosophy 45 (2017): 154.
Solonin, “Dīpaṃkara in the Tangut Context (Part 2),” 17.
Both the long *Āryācala nāma dhāraṇī (Peking 318, Derge 631) and the short version (Peking 588, Derge 963) are listed as parallels of SI 194 by Nishida, Seikabun Kegongyō III, 27, no. 102; however, the titular Mi mú (
) may refer to Akṣobhya. A dhāraṇī was transcribed from this manuscript by Nie Hongyin, “Tangut Fragments Preserved in the China National Institute of Cultural Heritage,” in Tanguty v Central’noj Azī. Sbornik Statej v Chest’ 80-letija Prof. E.I. Kychanova [Tanguts in Central Asia: A Collection of Articles in Honor of the 80th Birth Anniversary of Professor E.I. Kychanov], ed. I.F. Popova (Moscow: Bostochnaja Literatura, 2012), 276, as follows (the Sanskrit has been silently corrected): namo ratnatrayāya | tadyathā oṃ kākani kākani rocani rocani troṭani troṭani trāsani trāsani pratihana pratihana sarvakarmaparāṃparāṇi me svāhā 
[]
. This formula is a dhāraṇī of Tibetan Mi ’khrugs pa (i.e., Akṣobhya) in the Sarvakarmāvaraṇaviśodhani nāma dhāraṇī (Derge 743, 236a). For a published facsimile see Eluosi kexueyuan dongfang wenxian yanjiusuo
Zhang Baoxi
On the transmission of Mahākāla worship into the Tangut Empire and its association with war magic during the life of Ga Lotsāwa Zhönnupel (1105/1110–1198/1202, Tib. rGwa lo tsā ba gZhon nu dpal), see Haoran Hou’s contribution to the present volume. The pairing with Acala has some basis in the passing mention of Mahākāla in the Serlingpa sādhana (Derge 3059, 116a).
Derge 495, 322a: me nyag rgyal po’i bka’ lung gis/ rgya gar gyi mkhan po u pa sha ma rakṣi ta’i zhal snga nas lotstsha ba glan dge slong dar ma blo gros kyis bsgyur cing zhus te […] bi kra ma la shī la’i gtsug lag gi rgya dpe la gtugs nas.
Maho Iuchi, “The Bka’ gdams chos ’byung Genre and the Newly Published Ye shes rtse mo’s Bka’ gdams chos ’byung,” in The Historical Development of Tibeto-Himalayan Civilization, ed. Iwao Kazushi and Ikeda Takumi (Kyoto: Institute for Research in Humanities, Kyoto University, 2018), 349–351.
See e.g. Heller, “On the Development of the Iconography of Acala and Vighnantaka”; Kira F. Samosyuk, Buddhist Painting from Kharakhoto, XII–XIVth Centuries, Between China and Tibet. P. K. Kozlov’s Collection (St. Petersburg: The State Hermitage Publishers, 2006), 295, cat. no. 129; Grigory Semenov and Jin Yasheng, ed., Eluosi guo li Aiermitashi bowuguan cang Heishui cheng yishu pin
Vighnāntakasādhana, Sādhanamālā, ed. Bhattacharyya, vol. II, 558, no. 281: vighnān- takaṃ […] ekamukhaṃ dvibhujaṃ nīlavarṇaṃ vāmakareṇa tarjanikāpāśaṃ dakṣiṇa- kareṇodyatavajraṃ bhayānakaṃ piṅgalordhvakeśam; cf. Derge 3633, 250a; Derge 3363, 41a.
The State Hermitage Museum paintings x2378 and x2378 (Samosyuk, Buddhist Painting, 333, cat. nos. 162, 163; Semenov and Jin et al., Eluosi guoli Aiermitashi, pls. 162, 163, respectively, may depict this form of Vighnāntaka.
On the distinction between a physical ‘likeness’ (Skt. pratimā) and a ‘painted cloth’ (Skt. paṭa), see Sinclair, “The Appearance of Tantric Monasticism,” 259–260.
Samosyuk, Buddhist Painting, 331, cat. no. 160; Semenov and Jin et al., Eluosi guoli Aiermitashi, pl. 160.
Derge 3059, 116a: /rnal ’byor bsrung phyir ral gris bzlog/ /yul khams bsrung dang […].
Xie, Tibet-tradition Paintings in Xixia, fig. 37; Samosyuk, Buddhist Painting from Kharakhoto, 159, cat. no. 159; Semenov and Jin et al., Eluosi guo li Aiermitashi bowuguan, pl. 159.
For instance, Jisheng Xie, “Tibetan Buddhism and Tibetan Buddhist Art in the Xixia Kingdom,” in Faith and Empire, ed. Karl Debreczeny et al. (New York: RMA, 2019), 94, connects the Cleveland tapestry to the Five Great Vidyārājas (Chin. Wu da mingwang
See James C.Y. Watt et al., When Silk was Gold: Central Asian and Chinese Textiles (New York: The Metropolitan Museum of Art, 1997), 91, pl. 24, for the Cleveland tapestry, and 93, fig. 35, for the first Potala tapestry; both are misidentified as Vighnāntaka. For the second Potala tapestry (HAR 9157), now in the Tibet Museum, see Bernadette Broeskamp, “Lasa Xizang bowuguan yisi Budong jingang tangka zhi duandai
Tsangwang Gendun Tenpa, “Tibetan Buddhism and Art in the Mongol Empire According to Tibetan Sources,” in Faith and Empire, ed. Karl Debreczeny et al. (New York: RMA, 2019), 106. However, Broeskamp, “Lasa Xizang bowuguan yisi Budong jingang tangka zhi duandai,” understands this particular work to be a Yuan-era copy.
Sørensen and Hazod, Rulers on the Celestial Plain, II, 375, n. 23.
The painting was first noticed by Fanwen Li, “The Influence of Tibetan Buddhism on Xixia,” in Tibetan Studies. Proceedings of the 7th Seminar of the International Association for Tibetan Studies, Graz 1995. Volume II, ed. Helmut Krasser et al. (Vienna: Verlag der Österreichischen Akademie der Wissenschaften, 1997), 571, pl. 6.
See HAR 30914, which depicts Dīpaṃkara and Dromtön in the upper left corner, the related work HAR 10203/36447, and HAR 8069 and 61324.
Sinclair, “The Appearance of Tantric Monasticism,” 219–220.
Samosyuk, Buddhist Painting from Kharakhoto, 325, cat. no. 150; Semenov and Jin et al., Eluosi guo li Aiermitashi bowuguan, pl. 150. The fragment preserves part of a red goddess who occupies a corresponding position in Tangut Vajrayoginī paintings such as x2388 (HAR 18089) and x2394 (HAR 18112) from The State Hermitage Museum.
On the significance of this area of the composition, see Sinclair, “The Appearance of Tantric Monasticism,” 261.
Here I rely on Solonin, “Dīpaṃkara in the Tangut Context (Part 2),” 18–20, for this information from a Tangut commentary on Dīpaṃkara’s Satyadvayāvatāra, the Mýr mja e vó phí la 
. The aforementioned Tangut text Quintessential Instruction of the Forty Banners of Emptiness gives the usual story in which Dīpaṃkara’s meditational deity (Tib. yi dam) Tārā advises him to go to Tibet. Tibetan sources for this episode, starting with the Extensive Life-story (Rnam thar rgyas pa) §§195–196, were examined by Haruki Shizuka, “Expulsion of Maitripa from the Monastery and Atiśa’s Participation,” Journal of Indian and Buddhist Studies 63.3 (2015).
These names are taken from the Tang. Gu ma nu de la 
[Vajravārāhī Aural Transmission] (MS: Nishida, The Hsia-Hsia Avataṁsaka, III, 38, no. 172), ed. Sun Bojun
Dan Martin, “Padampa Sangye: A History of Representation of a South Indian Siddha in Tibet,” in Holy Madness: Portraits of Tantric Siddhas, ed. Robert N. Linrothe (New York: RMA, 2006), 25.
David Templeman, Aesthetic Path: Buddhist Art Collections (Hong Kong: Hollywood Galleries, 2019), 50–57, no. 20.
Solonin, “Dīpaṃkara in the Tangut Context (Part 2),” 11.