Formative sources for ritualized charnel asceticism

[Full title: Formative sources for ritualized charnel asceticism in iconography and practice]

Charnel asceticism can be understood as the historical innovation of Buddhist and brahmanical śramaṇa, or renunciate, traditions which emerged in the social geography of south Asia over the course of the first millennium BCE.^[1] As these communities grew and diversified, they cultivated methods and modes of observance which instrumentalized transgressive, antinomian or left-handed (vāmācāra) qualities of pollution, impurity and social marginality.^[2] The kāpālika observance to carry a skull or skull-topped staff (khaṭvāṅga) is first recorded amongst these śramaṇa groups as the religious performance of a legal punishment for killing a brahmin, first documented in the Yājñavalkyasmṛti (c. third-fifth centuries C.E) as a twelve year period of exile and charnel asceticism, after which one is restored to society.^[3] Here, rather than a distinct sectarian affiliation, the name or title kāpālika can be understood as a ritual identity defined by the socially transgressive and ascetic instrumentation of charnel materials and human remains.

As a religious practice, the kāpālika vow is first documented in the Pāśupata corpus where the khaṭvāṅga is described as one of the implements taken by initiates during a period of public observance.^[4] This was done in order to identify with Rudra—associated in the Vedas with disorder and transgression—by which they would be liberated from rebirth through union with the deity Maheśvara or Śiva-Pāśupati.^[5] However, the kāpālika use of khaṭvāṅga is only one of the ways in which the Pāśupata performed their religious identity with other observances included feigned madness, singing and bathing in ash.The Pāśupata came to dominate the brahmanical atimārga, or ascetic path, by the fifth and sixth centuries and within this growing community, the Lākula distinguished themselves as a sub-group of charnel specialists, adding to the implements of the kāpālika vow an initiatory thread (upavīta) made of hair and charnel ornaments made from pieces of skull, further instrumentalizing the notoriety of these materials by identifying themselves as a community through them (figure 2.6).^[6]

Figure 2.6: Two ascetics under the seat of Pāśupata founding teacher Lakulīśa, at Taleśvara in Bhubaneśvar, 7th century. The kāpālika, on the left, holds a skull-topped staff or khaṭvāṅga while the figure on the right holds a triśūla. Image from Donaldson 1986.

After the sixth century, the Pāśupata tradition was a demonstrably active and diverse proportion of the Śaiva communities then expanding across south Asia and who would further re-contextualize these charnel ascetic observances within the brahmanical mainstream. Reflecting a division between orthodox Pāśupata ascetics (e.g. Pañcārtha) and more inclusive Śaiva practitioners of the mantramarga, or ritual path, narratives like the Skandapurāṇa —sections of which are dated as early as the late sixth or early seventh centuries—would describe the kāpālika vow as one of the methods with which to identify with Śiva-Maheśvara in a text oriented towards laity as well as ascetic specialists.^[7] This work moreover describes the legendary second century Pāśupata founder Lakulīśa as an incarnation of Śiva, reinforcing a fundamental relationship between these modes of charnel asceticism and Śaiva sources.^[8]

Other narratives from this period further these associations between Śiva and the kāpālika vow: A number of Śaiva monuments constructed between the seventh and tenth centuries in or around Bhubaneśvar in the eastern region of Orissa reflect the prestige of Pāśupata modes of charnel asceticism by juxtaposing Lakulīśa with images of Śiva-Bhikṣāṭanamūrti, illustrating the account of Śiva as a wandering kāpālika ascetic found in the Matsyapurāṇā (c. third to sixth centuries) as well as the Skandapurāṇa.^[9] As Bhikṣāṭanamūrti, Śiva performs the kāpālika observance after killing the deity Brahmā, whose skull he carries for twelve years until he is released from his expiatory exercise in the charnel grounds of Varanasi, a center for Pāśupata activities during these centuries.^[10] In an image preserved at the temple of Paraśurāmeśvara, the ithyphallic, two-armed Śiva solicits a female donor by holding the empty skull towards her with his left hand (figure 2.7).

Figure 2.7: Śiva-Bhikṣāṭanamūrti with female donor and dancing figure at right, mid-7th century, Paraśurāmeśvara temple at Bhubaneśvar. Śiva holds a skull in his left hand and a peacock-feather topped staff on the right. Image from Donaldson 1986.

Figures 2.8 and 2.9: Reverse and obverse of Śiva-Bhūteśvara (Rudra addorsed to Maheśvara) with a single skull in the hair of both Rudra and Aghora/Bhairava (obverse, proper right), Fattehgarh, Kashmir, c. 6th century. Image from Granoff 1979.

At the same time, an integration and popularization of charnel ascetic observances emerges in the iconography for comprehensive representations of Śiva as well. A series of sixth to eighth century images of Rudra addorsed to Śiva-Maheśvara—a figure identified as Bhūteśvara—have been found in Kashmir, in which region brahmanical traditions dominated after the seventh century (figures 2.8 and 2.9).^[11] Here, Rudra has a single skull ornament in his hair, above which the figure of Lakulīśa carries his characteristic club (Skt. laguḍa, lakula), while the deity holds a staff topped with a Śaiva trident (triśūla) horizontally across his body.^[12] On the obverse, the figure of Aghora—also called Bhairava, to the proper right of Mahādeva at center—has a skull ornament and wide-eyed, wrathful visage like Rudra on the reverse.

Figure 2.10: Śiva-Maheśvara as Śādaśiva, with Aghora/Bhairava at left with a single skull in his hair, Elephanta, 6th century. Image from Kramrisch 1981.

Figure 2.11: Śiva-Andhakāsura with skull in the left hand and single skull ornament in the hair, Elephanta, c. 6th century. Image from Kramrisch 1981.

The broader integration of charnel ascetic materials into Śaiva iconography is seen as well at the site of Elephanta, the construction and iconography of which has been explicitly related to Pāśupata sources and patronage after the seventh century.^[13] The central image of this temple complex is a five-headed Sadāśiva, a comprehensive form of the deity as Maheśvara related to the expansion of the Śaiva community from an ascetic to ritual tradition accessible to a diversity of social groups (figure 2.10).^[14] Here, a single-skull ornament is found—as above in fig. 2.9—in the hair of Aghora/Bhairava to the proper right of Mahādeva; in the slightly later figure of Śiva-Andhakasura at the same site, however, the wrathful deity—now a solitary central figure—holds a skull in the front left of his many hands under the damaged figure of Andhaka, with a single-skull ornament in his pile of hair (figure 2.11).32 Śiva is elsewhere illustrated at Elephanta as the ascetic Yogeśvara, or lord of yoga, who is also identified with Lakulīśa.^[15]

This shift in iconography towards centralized images of Śiva with kāpālika implements corresponds to a further expansion of the Śaiva mantramārga, within which the goal was the ritual cultivation of siddhi, or accomplishment. This ritual corpus is preserved in two substantial traditions, the mantrapīṭha and vidyāpīṭha, which were collectively identified as śāstra or āgama (rather than tantra) of Bhairava.^[16] However it was in the latter collection (pīṭha) of applied ritual knowledge (vidyā) that kāpālika methods and materials were especially prominent.^[17] And though yoga has its roots in the oldest śramaṇa traditions, the practice of a siddhayogin—one who ritually instrumentalizes the body through performance, visualization and other actions in order become accomplished or empowered—became central to the expansion of the mantramārga and the Śaiva cultivation of what would come to be the characteristic methods of tantra, including mantra, maṇḍala and mudrā.^[18] Moreover, the identification of a siddha as one who is accomplished by means of deity yoga, antinomian behaviors and ascetic observances was historically associated with the (non-tantric) Pāśupata community.^[19] Therefore, kāpālika siddhas were not an innovation of tantra (or Śaiva āgama), but rather an integration and valorization of charnel asceticism as part of its expanding ritual methodologies.

After the seventh century, siddhas were an increasingly valued category of skilled ritual practitioner whose utility was known to Śaiva and Buddhist communities as well as a variety of political patrons.^[20] Building on the historical repertoire of the vidyādhara—one who possesses applied ritual knowledge, often translated as sorcerer, magician or wizard—siddhas diversified their ritual methodologies with means to attain practical siddhi, including powers of flight, control of enemies, and invisibility.^[21] In some of these applications, the volatile associations of the skull—rather than indicating identification with a legal or brahmanical transgression—were instrumentalized as a vessel for extraction, refinement or purification, including the production of elixirs for immortality (amṛta) and the intoxicant soma.^[22] Where integrated into these methods of ritual empowerment, kāpālika implements came to define ritual methodologies which incorporate charnel materials as well as a specific type of vow-holder.^[23]

Figure 2.12: Cāmuṇḍā from an unknown site in Orissa, 9th century, now at the British Museum (1872,0701.83). This multi-limbed kāpālika deity has a skull in the lower left hand as well as the severed head of Brahma, a garland of skulls, the emaciated form of an ascetic and corpse as a seat. Note also the vajra in the top left corner.

Moreover, kāpālika methods were cultivated as part of a brahmanical tradition oriented towards the group of female deities known as mātṛkā. Though there is archaeological and literary evidence for this assembly from the early centuries of the Common Era, after the sixth century these figures came to be primarily associated with Śaiva practices of charnel asceticism.^[24] Foremost in illustrating this connection was the deity Cāmuṇḍā, whose iconography and characterization became of increasingly complex and central importance in the expansion of kāpālika practices of the Śaiva vidyāpīṭha oriented towards female deities, or śākta traditions (figure 2.12).^[25] At the tenth century temple of Someśvara at Mukhalingam, for example, near the Śaiva religious center of Bhubaneśvar, kāpālika and other ascetics are prominent and numerously depicted—and in veneration of Cāmuṇḍā, emaciated and seated on a corpse—in an iconographic programs which also feature images of the Pāśupata teacher Lakulīśa (figures 2.13-2.16).^[26]

Figure 2.13 (above): Eight-armed Cāmuṇḍā seated on a corpse as the object of kāpālika devotion at Someśvara, near Bhubaneśvar, 10th century. Image from Donaldson 1986.

Figures 2.14 and 2.15: Kāpālika practitioners in the iconographic program at Someśvara, near Bhubaneśvar, 10th century. Charnel ascetic yogins are distinguished by a single skull ornament in the hair, khaṭvāṅga and/or skull, shown with female companions and students. Image from Donaldson 1986.

Figure 2.16: Another kāpālika figure with khaṭvāṅga and skull from Someśvara, likely a teacher and/or ritual master (at left) instructing a student from his raised seat. Image from Donaldson 1986.

In a tenth century free-standing relief from Puri, another site of tantric activity in Orissa, Śiva as Gajāsūrasaṃhāramūrti—a form of the deity historically associated with mātṛkā iconography —holds a skull in the left hand with a single skull in his hair and wears a mālā of skulls, similar to Cāmuṇḍā in the eighth century representation in fig. 2.12, above.^[27] In fig. 2.17, an emaciated companion is positioned beneath Śiva’s feet as—and with Andhakasura above in fig. 2.11—the figure holds a skull to catch the blood of the slain demon who has been liberated at death in a narrative which supports the Pāśupata soteriological model of liberation in union with the deity by adopting the form of a transgressor. Multi-limbed, variously empowered forms of Śiva-Bhairava became increasingly central in the monuments of these brahmanical communities though after the tenth century, Śiva was less explicitly represented as a wrathful, kāpālika deity and charnel materials are integrated into the deity’s expanding assortment of martial and ritual implements (see fig. 2.19, below).^[28]

Figure 2.17: Śiva-Gajāsūrasaṃhāramūrti from Puri, 10th century, now at the Indian Museum in Kolkata. The figure has a single skull ornament in his hair and—like Andhakāsura, above in fig 2.11—lifts a skull with a left hand to catch the blood of the vanquished demon, December 2017.

Early vidyāpīṭha texts reflect how kāpālika methodologies and groupings of mātṛkā were integrated into the foundations of what would later be known as yoginī tantra. In the sixth to eighth century text of the Brahmayāmala (also called Picumata), charnel materials are made part of a ritual system which uses mantra, maṇḍala and mudrā, and is centered on a male and female deity pair whose primary forms—Kapāliśabhairava and Caṇḍakapālinī —are each recognized as skull-holders (kapālin/ī).^[29] In this text, methodologies of ritualized charnel asceticism—including the use of skulls as vessels for the preparation or refinement of substances which facilitate siddhi —are derived from earlier Lākula sources in the Niśvāsa corpus, and the kāpālika vow is described as an especially valued and refined means for embodying the deity as Bhairava.^[30] However, while Lākula sources acknowledge mātṛkā figures in their cosmological models and at times used sexual fluids in their prepared substances, they were primarily celibate ascetics concerned with practical siddhi like invisibility, the controlled reanimation of corpses and, ultimately, liberation through Rudra.^[31]

In the methods of accomplishment (sādhana, Tbt. sgrub thabs) which were innovated in vidyāpīṭha texts like the Brahmayāmala, the yogin applies kāpālika practices primarily in order to ritually engage with volatile female deities called yoginī, through which they attain siddhi and become empowered in the form of the deity Śiva-Bhairava. No longer restricted to the historic grouping of seven or eight mātṛkā, in these practices the yoginī becomes an ever-expanding category of figures organized into clans (kula) oriented around the primary deity whose form is actualized through mantra, maṇḍala and mudrā.^[32] At the same time, the Brahmayāmala describes the yoginī as a figure complementary to that of a siddha or yogin whose most extreme observances incorporate the implements of the kāpālika vow: In this text, yoginī—like the deity Cāmuṇḍā, above in figs. 2.12 and 2.13—are represented as empowered charnel ascetic figures, naked with a corpse as vehicle or seat, holding a skull and khaṭvāṅga, wearing a skull mālā and raising the right hand in a gesture of munificence.

The models of ritual engagement cultivated in practices of the Śaiva vidyāpīṭha expanded the definition and functions of kāpālika implements by identifying them within the tantric (viz. āgamic) systems used to become empowered and attain siddhi. The Brahmayāmala, for example, describes the skull and khaṭvāṅga —as well as a set of charnel ornaments worn on the head, ears, neck, hands, arms and hips, and an initiatory thread—as a comprehensive set of mudrā used to embody the deity as the skull-bearer Bhairava and engage yoginī.^[33] Though Lākula kāpālika practitioners also ornamented themselves with skulls and a thread of hair in their vrata —in addition to their use of skull and khaṭvāṅga and in order to assume the form of Rudra—there is no indication that these were recognized as a set of mudrā.^[34] Moreover, by integrating the yoginī as a ritual consort—also interpreted as a mudrā—the Brahmayāmala indicates the combination of charnel asceticism and sexual yoga which would characterize the yoginī tantra corpus.^[35]

The integration of yoga and charnel methodologies would be further refined and popularized in the formalization of kaula traditions which, after the ninth century, promoted a ritual model which internalized within the body of the yogin the transgressive dynamics of kāpālika and yoginī practices.^[36] In this corpus, the instrumentation of impurity, polluting materials, ritualized sex and the performance of deity yoga are increasingly systematized through visualization and the tantric methods of mantra, mudrā and maṇḍala in order to facilitate ecstatic knowledge or gnosis (jñāna), with a decreased emphasis on gaining siddhi through encounters with yoginī consorts and the use of charnel materials.^[37] As a result, the embodiment of Bhairava—as well as the actualization of the deity’s maṇḍala and its assembled clans of yoginī—could be accomplished as a yogic exercise, rather than an externalized practice.^[38] At the same time, in kaula yoginī methods, the kāpālika vow was further systematized as one of a number of antinomian observances which included having sex or eating with consorts from non-brahmanical groups, eating impure substances, wearing animal skins, drinking alcohol in excess and the habitation of liminal social spaces, including charnel grounds.^[39]

The refinements of kaula ritual methodologies—internalized yoga and a shifting focus from empowerment to gnosis—encouraged the expansion of yoginī traditions within a broader range of social groups, facilitating the patronage necessary for the monumentalization of these practices.^[40] At Hirapur—also near to Bhubaneśvar in Orissa—a tenth century yoginī temple suggests a regional transition towards kaula imagery, supported in part by the changing political dominance from the Buddhist Bhauma-Kara regime of previous centuries to the Śaiva Somavaṃśī, who promoted kaula teachings (figures 2.18-2.22).^[41] At the same time, this site illustrates the emergence of yoginī tantra as a distinct iconographic program and monumental strategy, with the primary deity (no longer in situ) and assembly at the center of a circular array of numerous yoginī.^[42]

Figure 2.18: Plan of the yoginī temple at Hirapur with interior niches for 60 yoginī, four couples surrounding the main deity at center (now missing) and nine exterior yoginī with charnel implements, 10th century. Image from Dehejia 1986.

Figure 2.19: Unidentified form of Bhairava from the central pavilion at Hirapur seated on corpse with yoginī holding a knife and skull in the lower register, c. 10th century. Image from Donaldson 2001.

Figure 2.20: Cāmuṇḍā as one of the 60 yoginī on the interior of the Hirapur temple, 10th century. Image from Hatley 2014.

Figure 2.21: One of the few charnel implements from the interior of Hirapur, yoginī drinking from a skull, 10th century. Image from ASI/ Dehejia 1986.

Figure 2.22: Yoginī with knife in the raised right hand, skull and staff on the left, standing on a decapitated head with animals of the charnel ground; one of nine such figures on the exterior of Hirapur, 10th century. Image from Donaldson 2001.

In addition to demonstrating local material support for the kaula tradition, this temple suggests unique historical evidence for the monumentalization of a maṇḍala in the context of yoginī practices and is illustrative of their tantric ritual dynamics.^[43] However, within this composition, kāpālika implements are largely absent from the interior though consistently seen in the nine exterior yoginī figures, each standing on a severed head framed by two jackals, holding a skull in the left hand with a knife raised in the right, as in fig. 2.22. The diverse interior program and central assembly incorporate Śaiva and mātṛkā iconography (figs. 2.19 and 2.20) but without prominent charnel elements and only one of the inward-facing yoginī survives holding a skull (figure 2.21). Moreover, the narrow entrance at Hirapur suggests the esotericism and restricted access promoted in kaula teachings, as well as the division between an interior, ritually activated and refined space, and an exterior marked by the volatile impurities of charnel materials.^[44]

Altogether, this section has explored the refinement and integration of kāpālika implements from their earliest surviving evidences into the expansion of the Śaiva and brahmanical ritual corpus and its iconographies including changing concepts of deity yoga, the engagement and control of volatile entities like yoginī, and methods for empowerment including both practical applications (e.g. immortality, flight) and the cultivation of gnosis (jñāna). Moreover, within these sources, the significance of charnel implements is dynamic and diverse, from the public observation of transgressive ritual identity, to a comprehensive system of mudrā and finally within a monumental illustration of ritual methodologies and visual programs that would come to dominate Buddhist tantra.

Footnotes and references:

[back to top]

[1]:

Geoffrey Samuel describes the common origins of Buddhist and brahmanical śramaṇa ascetics through their relationship to historical urbanization patterns across south Asia, including their early associations with charnel settings and specialization in socially provocative behaviors in idem., The Origins of Yoga and Tantra (Cambridge: Cambridge University Press, 2008), 120-128.

[2]:

See Wedemeyer, Making Sense of Tantric Buddhism, passim for an extended text-based historical and semiotic study of the discursive process of identifying and engaging concepts of the antinomian in the Indian sources for Buddhist tantra.

[3]:

David Lorenzen, The Kāpālikas and Kālāmukhas (New Delhi: Thomson Press, 1972), 13.

[4]:

Wedemeyer finds this reference in the c. fourth century Pāśupatasūtra in Making Sense of Tantric Buddhism, 157.

[5]:

Peter Bisschop and Arlo Griffiths, “The Pāśupata observance (Atharvavedapariśiṣṭa 40),” Indo-Iranian Journal 46, no. 4 (2003), 331-332. See also Stella Kramrisch, The Presence of Śiva (Delhi: Motlilal Banarsidass, 1988 [1981]), 331-332. Kramrisch describes Rudra (later Śiva) as lord of the animals (pāśupati) for the Pāśupata, for whom the self was sacrificial like a beast.

[6]:

Wedemeyer notes this version of the kāpālika vow in the Lākula text Niśvāsatattvasaṃhitā, op.cit., 157-158. See also, ibid., 254n81 on the materials used for this thread by the Lākula; note that Sanderson elsewhere renders vālayajñopavīta as “snake skin” in idem., “Śaivism and the tantric traditions,” in The World’s Religions: the Religions of Asia, ed. F. Hardy (London: Routledge, 1988), 133. On the use of human remains for the upavīta as an alternative to the cotton thread of non-ascetic brahmanical groups, see Sanderson, “The Śaiva age”, 209n479. According to the same, Lākula practitioners are progenitors of the ninth-thirteenth century community of Kālāmukha charnel ascetics in southern India.

[7]:

In his study of the Skandapurāṇā and its historical context, Hans Bakker finds that, by the seventh or eighth century kāpālika was an increasingly and broadly applied identification for Śaiva ascetics, idem., The World of the Skandapurāṇā (Leiden: Brill, 2014), 149. Bakker moreover notes that the connection of Śiva with the kāpālika observance is foremost in this purāṇā (p.151).

[8]:

Bisschop and Griffiths, “The Pāśupata observance,” 323.

[9]:

Thomas E. Donaldson, “Bhikṣāṭanamūrti images from Orissa,” Artibus Asiae 47, no. 1 (1986), 51-66. See also Lorenzen, op.cit., for a comprehensive treatment of historical and narrative sources on the kāpālika vow including several purāṇās, inscriptions and satirical accounts. Other Śaiva sources for this narrative are given in Kramrisch, op.cit., 287.

[10]:

Wedemeyer notes that many ascetic vrata —including those which pre-date the kāpālika vow—were also temporary, ranging in length from one month to a lifetime, op.cit., 153. See also Diana L. Eck on the historical association of Varanasi with Śaiva kāpālika practitioners and Śiva’s liberation at the place of Kapālamocana, idem., Banaras: City of Light (Princeton: Princeton University Press, 1982), 119. Eck describes the deity as skull-bearer in the form of Bhairava, according to the later, likely twelfth century material in the Skandapurāṇa, ibid. 190ff.

[11]:

John Siudmak,The Hindu-Buddhist of Ancient Kashmir and its Influences, (Leiden: Brill, 2013), 154. Both Siudmak and Phyllis Granoff relate the iconography of Rudra on the reverse of this figure to contemporaneous representations of Nandin, Rudra’s non-transcedent form; see idem., “Maheśvara/ Mahākāla: A unique Buddhist image from Kaśmir,” Artibus Asiae 41, no. 1 (1979), 64-82. Granoff, however, identifies this image as Buddhist. Another sixth century example of Bhūteśvara from the same region and presently at the Metropolitan Museum of Art (2014.687) has three skulls in Rudra’s hair on the reverse. Note also Rudra’s use of stick or club suggests his role as a charnel figure and protector, see below on the yamadaṇḍa.

[12]:

A similarly addorsed image to a figure of Maheśvara from Kashmir has been identified as Rudra with Lakulīśa in his crown by Pratyapaditya Pal in Bronzes of Kashmir (Graz: Akademische Druck-u. Verlagsanstalt, 1975), figs. 4 and 4a.

[13]:

Charles David Collins, The Iconography and Ritual of Śiva at Elephanta (New Delhi: Sri Satguru Publications, 1991 [1988]), 128.

[14]:

Sanderson notes that the figure of Sadāśiva is iconographically associated with the least unorthodox trends in Śaiva ritual methodologies during this time, idem., “Śaivism and the tantric traditions,” 136.

[15]:

Collins, op.cit., 33.

[16]:

Sanderson, ibid., 136. In another classificatory system, these texts are classified by having been revealed by the proper right face of Mahādeva/Sadāśiva, which is Aghora/Bhairava, see Alexis Sanderson, “History through textual criticism in the study of Śaivism, the Pañcarātra and the Buddhist yoginītantras”, in Les sources et le temps, ed. François Grimal (Pondicherry: École Française d’Extreme Orient, 2001), 19.

[17]:

Svacchandabhairava was the dominant kāpālika practice of the mantrapīṭha though “its position as the standard Śaivism of the Kasmirian householder had modified its heteropraxy” resulting in a decreasing emphasis on the instrumentation of impure materials and observances towards the tenth century, see Alexis Sanderson, “Purity and power among the Brahmans of Kashmir,” in The Category of the Person ed. Michael Carrithers, et al. (Cambridge: Cambridge University Press, 1985), 204.

[18]:

Samuel elsewhere describes the formation of tantra through its historical relationship to yoga, Origins of Yoga and Tantra, 223ff. Alexis Sanderson has made the case that these innovations were first recorded by Śaiva authors, with Vajrayāna formed in response as a Buddhist adaptation, see idem., “Vajrayāna: Origin and function,” 96.

[19]:

Ronald Davidson finds the term siddhāyogin in the Pāśupatasūtra describing a practitioner who is “un-smeared by ethical action or guilt”, see idem., Indian Esoteric Buddhism: A Social History of the Tantric Movement (New York: Columbia University Press, 2002),184. Samuel elsewhere explores how, through the expansion of ritual traditions after fifth century, the ascetic vrata was reformulated as the deity yoga practice of Buddhist tantra in idem., Origins of Yoga and Tantra, 238.

[20]:

For a social history of siddhas in south Asia in the seventh to eleventh centuries, the political applications of their knowledge and contributions to the formation of Buddhist tantra, see Davidson, ibid.

[21]:

Davidson, ibid., 194. In Buddhist literature, vidyādhara may also refer to non-Buddhist tantric practitioners and siddhas. See also David Gordon White, “Mountains of wisdom: On the interface between siddha and vidhyadhara cults and the siddha orders in medieval India,” International Journal of Hindu Studies 1, no. 1 (1997): 73-95.

[22]:

David Gordon White further discusses the integration of kāpālika methods with the foundations of tantric alchemy, the production of amṛta and the yoga of Nāth siddhas, c.f. idem., The Alchemical Body: Siddha Traditions in Medieval India (Chicago: The University of Chicago Press, 1996).

[23]:

Judit Törzsök notes the intersection of these definitions in evidence for a widespread kāpālika ritual tradition which centered on the cultivation and consumption of the intoxicant soma, idem., “Kapalikas”, in Brill’s Encyclopedia of Hinduism, vol. 3 (Leiden: Brill, 2011), 357-8.

[24]:

Shaman Hatley has identified earlier mātṛkā iconography with no representation of charnel associations from the Kuṣāṇa period while the Pāśupata Śaiva authored Skandapurāṇa (sixth to ninth centuries) describes both Skanda and Śiva as a kāpālika in the company of a group of fearsome mātṛkā, idem., “The Brahmayāmalatantra and early Śaiva cult of yoginīs,” (PhD diss., University of Pennsylvania, 2007), 48ff. At Elephanta, a group of seven mātṛkā accompanies Śiva as Yogeśvara/Lakulīśa; Collins, op.cit., 24.

[25]:

Sanderson has elsewhere detailed the development of tantric ritual traditions oriented towards fearsome central female deities—i.e. śākta tantra, Kālī and Krama practices—within and adjacent to the Śaiva vidyāpīṭha, including their integration of kāpālika materials and methods, see idem., “Śaivism and the tantric traditions,” 140ff. See also idem., “Maṇḍala and Āgamic identity in the Trika of Kashmir,” in Mantras et diagrammes rituels dans l’hindouisme: Table ronde, Paris 21-22 juin 1984, (Paris: Editions CNRS, 1986), 169-214 for the influence of the Kashmiri author Abhinavagupta (c.950 - 1016) on the reformation of kāpālika elements in later śākta tantra, or āgama.

[26]:

See Mary F. Linda, “Nārāyanapuram: A tenth century site in Kalinga,” Artibus Asiae 50, no. 3/4 (1990): 232-262 and Walter Smith, “Images of divine kings from the Mukteśvara temple” Artibus Asiae 51, no. 1/2 (1991): 90-106 for a comparison of style to other nearby monuments in the eastern region, as well as an exploration of the local political connections of the Pāśupata siddhas community in the tenth century. Linda notes a number of vajra in the iconography of Someśvara, a Pāśupata site.

[27]:

Michael W. Meister, “Regional variations in mātṛkā conventions,” Artibus Asiae 46, no. 3/4 (1986), 242. Meister documents forms of Śiva found with mātṛkā iconography in examples from the fifth to fourteenth centuries; Gajāsūrasaṃhāramūrti is the form of Śiva at the majority of these sites.

[28]:

Thomas E. Donaldson finds that images of Bhairava—who has many forms—are increasingly seen at śākta sites in Orissa during the seventh to tenth centuries; after the mid-tenth century his tantric, though not prominently charnel ascetic form becomes more common; c.f. Donaldson, “Bhikṣāṭanamūrti images from Orissa”, 62 as well as idem., Tantra and Śākta Art of Orissa, (New Delhi: D.K. Printworld (P) Ltd, 2001), 440.

[29]:

Hatley explores this text, its sources and iconographic innovations at length in his doctoral dissertation (op.cit.); this text is classified as a yāmala or mata, though the Brahmayāmala is also referred to as a matṛtantra in the roughly contemporaneous Skandapurāṇa (a non-tantric Pāśupata Śaiva text) (p.35).

[30]:

Hatley finds citations and repetitions in the Brahmayāmala from various Niśvāsa sources which were shaped by the charnel specialization of Lākula ascetics (ibid., 133ff). On the kāpālika vow as bhairavavrata or a mahāvrata in the same corpus, see ibid., 181. Hatley notes that, amongst Śaiva sources for kāpālika ritual methods, the Brahmayāmala is more explicit in its ritual instrumentation of impurity than the Lākula Niśvāsa corpus.

[31]:

On the celibacy and cosmologies of the Lākula see Sanderson, “Śaivism and the tantric traditions”,134 as well as Samuel, Origins of Yoga and Tantra, 173 on the fundamental relationship between celibacy and asceticism in śramaṇa traditions.

[32]:

See Hatley on the history, characterization and diversification of yoginī iconography, including its relation to earlier representations of mātṛkā, op.cit., 11ff. Hatley defines the yoginī as any number of figures with shared properties including multiplicity and formation into clans, the appearance of a female kāpālika (e.g. with charnel implements and ornaments), a volatile character and capacity to transmit teachings and empowerments, frequently including flight. Other methods for obtaining siddhi in the Brahmayāmala include visualizations (dhyāna), homa (offerings made by fire) and processes of consecration, ibid.,190ff.

[33]:

Sanderson cites these mudrā—which are not enumerated—in the Brahmayāmala (ch. 80) in a passage which suggests a sustained historical association with Lākula and Śaiva ascetic sources by describing them as the implements of Rudra accompanied by a gathering of mātṛ(kā) (rudro matṛgaṇaiḥ sārdhaṃ), idem., “The Śaiva age,” 179n435. In the same note, he cites two passages in another vidyāpīṭha source for kāpālika practices, the Jayadrathayāmala, describing charnel ornaments as mudrā though again, without being enumerated: First as ornaments for a sādhaka (viz. siddha) who takes the form of Bhairava wearing earrings, bracelets, an ornament in the hair, a necklace and thread of hair; and elsewhere as the indication of a vow-holder with a skull ornament in the hair, charnel ornaments on the ears and limbs and with skull mālā, skull vessel and khaṭvāṅga.

[34]:

See Wedemeyer, op.cit, 157-8 and 253n81-83.

[35]:

Siddhas are moreover associated with ritualized sex by secondary sources after the fifth century, c.f. Davidson, Indian Esoteric Buddhism, 174.

[36]:

Hatley, op.cit., 157. See also David Gordon White on the kaula contributions to hatha yoga and tantric alchemy by Nāth siddhas, c.f., idem., The Alchemical Body, passim. The same author has explored the foundations of the kaula tradition through its reformed, internalized interpretation of sexual yoga in idem., Kiss of the Yoginī: “Tantric Sex” in its South Asian Contexts (Chicago: The University of Chicago Press, 2003). Hatley, unlike White, does not recognize the kaula as a formal sectarian identity, but rather a tantric literary and liturgical tradition cultivated by a diversity of authors and communities (op.cit., 20).

[37]:

Sanderson credits the kaula tradition—its authors, teachers and practitioners—with the reform and popularization of yoginī tantra and notes that kāpālika ritual methodologies are more prominent in earlier sources from the vidyāpīṭha, i.e. the Brahmayāmala, c.f. idem., “The Śaiva age,” 49 and idem., “Śaivism and the tantric tradition,” 147ff. White moreover notes that the legendary ninth century kaula author and Nāth yogin Matsyendra was against the open use of charnel materials, idem., Kiss of the Yoginī, 164. These reforms in gnostic ritual methodology and decreased reliance on charnel materials would be further developed in the Trika Kaula corpus by the eleventh century Abhinavagupta, resulting in the diminished popularity of Svacchandabhairava kāpālika practices, see Sanderson, “Purity and power,” 204.

[38]:

Samuel, Origins of Yoga and Tantra, 270. Samuel moreover notes that before the tenth century, this practice of embodying the deity at the center of an array of yoginī was more often referred to as kaula, rather than tantra.

[39]:

Wedemeyer, Making Sense of Tantric Buddhism, 138. Wedemeyer finds that it is primarily through yogic practice and observance (caryā and/or vrata) that socially and ritually impure materials like human remains are instrumentalized in Buddhist mahāyoga and yoginī tantra, and primarily in the latter. He moreover tabulates these behaviors in Buddhist sources in op.cit., 142.

[40]:

For a review of yoginī temple sites—round, open-roofed temples with numerous yoginī and a central deity, constructed predominantly between the tenth and fourteenth centuries across south Asia—see Vidya Dehejia, Yogini Cult and Temples: A Tantric Tradition, (New Delhi: National Museum, 1986). The number of yoginī figures varies at these sites from approximately sixty to as many as eighty.

[41]:

Donaldson, Tantra and Śākta Art of Orissa, 653.

[42]:

Shaman Hatley has explored this site in relation to the Śaiva yoginī corpus and notes that the earlier vidyāpīṭha was more oriented toward solitary practice than the construction of communal spaces, idem., “Goddesses in text and stone: Temples of the yoginīs in light of tantric and purāṇic literature,” in Material culture and Asian religions: Text, Image, Object, ed. Benjamin J. Fleming and Richard D. Mann (Abingdon: Routledge, 2014), 207.

[43]:

This is noted as well by Hatley in ibid., 214.

[44]:

Hatley observes that these monuments are inherently both “exoteric and esoteric”, and were suitable for the monumentalization of a ritual tradition which increasingly valued secrecy, “Goddesses in text and stone,” 217.