Bliss beyond All Limit: On the Apabhraṃśa Dohā in Tantric Buddhist Texts

Abstract

The Apabhraṃśa dohā is a literary medium from Indian antiquity, with early examples appearing in Kālidāsa’s plays around the 5th century and continuing in later Hindi-language Jain and Bhakti works in the early modern period. However, it was within Tantric Buddhist texts and traditions that the dohā truly came into its own as a literary genre. Particularly within the “Yoginī Tantra” strata of the Tantric Buddhist canon, Apabhraṃśa dohās appear in notable and formulaic ways, used within ritual contexts and other significant junctures, signaling the underexamined use of this literary form and its language of composition. This paper examines the use of dohās attributed to the mahāsiddha Saraha as they are used in the Hevajra Tantra, the Buddhakapāla Tantra, and some associated texts. In doing so, this paper demonstrates that as a literary genre, Apabhraṃśa dohās perform a similar function to mantras and dhāraṇīs, but are unique in their attention to phonology and discursive meaning. By examining the uses of these dohās during particular moments of Tantric Buddhist ritual syntax, this paper will then reflect on the later trajectory of these verses after the death of institutional Buddhism in India, and the reasons for their survival.

1. Introduction

Scholars have long remarked on the presence of Apabhraṃśa1 verses within Tantric Buddhist texts, and yet their use has been undertheorized. They are usually glossed (or rather explained away) as examples of twilight language (Skt: sandhyā-bhāṣa), and while this explanation is not incorrect, it is inadequate. This is made especially clear in the Hevajra Tantra, which contains an extended discourse on twilight language that consists entirely of esoteric meanings of specific Sanskrit words, without any Apabhraṃśa words at all2. Furthermore, the Hevajra Tantra contains so many Apabhraṃśa verses within ritualistic contexts that they must be examined alongside other examples of nonstandard language, such as mantras and dhāraṇīs. Within the Buddhakapāla Tantra as well, we see Apabhraṃśa dohās used as capstone verses, either summarizing or challenging the import of chapter topics. These verses are coarse yet mystical, with didactic and exhortatory charge, expressed in a medium that stands out in both form and content from its surrounding text. These verses also tend to be used formulaically, and as such tend to fall into specific genres, including dohās quoted in collections attributed to mahāsiddhas,3 verses sung in offering rituals,4 “password” verses,5 verses sung in initiations,6 “goddess songs,”7 and verses sung in worship.8 Given their ubiquity and underexamined significance in Tantric Buddhist texts, a study of Apabhraṃśa verses remains a desideratum in Tantric Studies.

Dohās as a literary genre predate Tantric Buddhism and they survived long after its demise in India. This continuation spans the transition from Śaurasenī Apabhraṃśa into early Hindustani, a linguistic development that also maps onto the rise of the Nāth Siddhas and the later Nirguṇī bhaktas. Strongly anchored by their shared anti-establishment rhetoric, marginal social position, and linguistic practice, the religious trajectory from Tantric Buddhist mahāsiddhas to Nirguṇī bhaktas is tangled but distinct, a fact noted by other scholars.9 So it was that the term used to refer to the dohās sung by the Nirguṇī Sānts10 in Northern India was “sākhī,” from the Sanskrit term sākṣin (lit. “seeing with the eyes”), used by Buddhist Siddhas to refer to their direct experience of enlightenment. This term underwent a semantic shift under the successors to the Buddhist Siddhas, the Nāth Siddhas, where it came to refer to the authoritative words of someone who has experienced enlightenment.11 The literary vehicle of the dohā offers a concrete example of how spiritual traditions and communities often transgress the neat religious boundaries conceptualized by high sectarian stakeholders, to the effect that “Buddhist” literary forms and religious ideas can inflect and inform “Hindu” traditions centuries after Buddhism disappeared from the subcontinent.

When honing in on the particular genre of the dohā, it is necessary to distinguish it from other Tantric Buddhist songs with which it is commonly grouped, i.e., caryā songs. While the two canons share thematic content, their forms are strikingly different, in both structure and language. All of these songs are commonly grouped together as examples of “Old Bengali,”12 and yet the distinct morphological features characteristic of Bengali plainly evident in the caryā songs are completely absent in dohās, a fact noted almost a century ago by the preeminent Bengali linguist Suniti Chatterji.13 Indeed, the language of the dohās is so closely related to the various dialects of Hindustani that it is clearly their immediate ancestor.14 Furthermore, dohās are by their very nature muktaka, independent and free-floating, amenable to quotation ad hoc.15 To use the language of Bakhtin, each dohā comprises a single independent utterance, whereas caryā verses are explicitly subsections of larger utterances.16 When one disentangles these two canons in this way, the continuities between Tantric Buddhist dohās and the later dohās composed by Nāth Siddhas and Nirguṇī bhaktas come into sharp relief, and the Śaurasenī Apabhraṃśa of these Tantric Buddhist dohās is enshrined as the rightful progenitor of the Hindustani language and literature (rather than Bengali). As such, when considering dohās as a Tantric Buddhist literary form, it is necessary to group them with their Nāth Siddha and Nirguṇī successors.

This article will focus on dohās attributed to the mahāsiddha Saraha (Saroruha) in particular. Saraha is an important figure in the history of Tantric Buddhism in India, and yet the concrete historical details of his life/lives are virtually inaccessible to the scholar.17 Accounts of his life/lives are the work of hagiography and legend, and there is reason to believe that there were multiple personages with this name. For this reason, it would be far more useful to consider the personage of Saraha not as a historical figure but as an author function. For Foucault, the author function serves as a literary device to demarcate a special mode of discourse above and beyond everyday speech, with particular value within a given culture or society.18 Using this lens, the dohās of Saraha are not the words of a historical personage, but rather a bounded set of utterances “authorized” within a particular cultural context. They possess a particular currency and value, operate at a higher level of discourse, and are amenable to quotation and use within a variety of contexts. As we will see below, within the universe of Tantric Buddhism, the dohās of Saraha are mobilized in distinct ways.

2. Saraha’s Dohās in Hevajra Literature

The Hevajra Tantra is a classic Tantric Buddhist text, and its associated corpus of commentaries and explanatory Tantras have been massively influential in the history of Tantric Buddhism in India and Tibet, with the Hevajra cult even having a presence in Southeast Asia as well.19 Dated to between the 8th and 10th centuries,20 the text has been described as the most important Tantric Buddhist text in India,21 and shows a mature blossoming of Tantric Buddhism in its niruttarayoga phase.22 Most significantly, the text also bears witness to many of the doctrines and practices of the Siddha movement, including the doctrine of the four joys, culminating in the “sahaja”23 association with low-caste and tribal peoples, and anti-institutional rhetoric. Most importantly, the Hevajra Tantra contains numerous Apabhraṃśa verses, more numerous than most other tantras,24 falling into numerous distinct genres, including “goddess songs,”25 verses sung in offering rituals,26 verses sung at gaṇacakra feasts,27 and other miscellaneous verses.28 In particular, the Hevajra Tantra features dohās that also appear in collections (Skt: dohākoṣa) attributed to the mahāsiddha Saraha (Saroruha), with which we concern ourselves below.

The first dohā to be discussed here appears in the chapter describing the Hevajra maṇḍala visualization, chapter five of the second section of the text (Skt: hevajrābhyudayapaṭalaḥ pañcamaḥ). This chapter also includes four “goddess songs”, uplifting and motivational verses sung by yoginīs to inspire the practitioner to arise from enstatic emptiness and complete the sādhanā. These goddess songs are outside of the scope of this paper, but also occur at a particular juncture that underscores the performative and specific use of these songs, in both form and content. This chapter is a fairly standard description of a maṇḍala. Here, the Buddha Hevajra describes how he is to be visualized, his maṇḍala, and his retinue.29 After dissolving into emptiness, the four yoginīs Pukkasī, Śavarī, Caṇḍalī, and Ḍombī call out to the practitioner in dohās,30 and Hevajra continues his description of the assembled yoginīs, mantras, and other visualizations.31 Finally, the maṇḍala visualization instructions culminate in the goddess Nairātmyā asking Hevajra to describe the nature of the secret essence contained within all Tantras that is revealed by the practice of the maṇḍala.32 Hevajra replies with the following dohā:33

āi na anta na mahyu tahiṃ natra bhava natra nirvāṇa |

ehu so paramamahāsuha nau para nau appāṇa || (HT II.5.67)

No beginning, no end, and no middle there; Nor existence or liberation.

This is the supreme Great Bliss where there is neither self nor other.34

This verse is significant for its apophatic description of nonconceptual awareness, termed the “Supreme Great Bliss” (ABh: paramamahāsuha), a common term used in Tantric Buddhist dohās to describe the attainment that Tantric Buddhist practitioners strive for, conceived as being beyond all distinctions and limitations whatsoever, the mind abiding in its original blissful state. The diglossic shift to Apabhraṃśa also signals a different level of discourse, in which the relatively standard and discursive Sanskrit gives way to a more fundamental, coarse, direct, and yet mystical level of communication. This shift in language is important and underscored by the dohā’s wide attestation throughout other Hevajra sādhanās, where the dohā is recited in key junctures as a central aspect of practice.

The significance of this verse is highlighted by its place in other Hevajra sādhanās, where it is quoted at key junctures. For instance, the 183rd sādhanā in the Sādhanāmālā is a Kurukullā sādhanā based on the limited instructions given in the Hevajra Tantra,35 and includes the above verse in a significant way. The sādhanā is devoted to identifying with Kurukullā so as to bring all beings under the aspirant’s control. It begins with the preliminaries, taking refuge in the three jewels, emanating and generating compassion towards all sentient beings, then proceeds into elaborate visualizations of seed syllables, maṇḍalas, and Kurukullā herself as identical with the practitioner.36 After another separate Apabhraṃśa verse recited at an important juncture (discussed below), the practitioner engages in more visualizations, recalls the teachings of the lineage (Skt: guru-parampara), entering into the teachings, and so on. Finally, the practitioner views the entire triple world as identical with their own body, and “in that very moment” (Skt: tat-kṣaṇam), one recalls the verse (here referred to as a gāthā):37

āi ṇa anta ṇa majjhu nahi na u bhava ṇa u nivvāṇa |

ehu so paramamahāsuhao ṇa u para ṇa u appāṇa ||

The reader will notice that this verse is largely identical with its previous appearance, with minor yet significant differences. Snellgrove’s edition of this verse is very Sanskritic in its phonology and differs from this version in many places, such as the use of mahyu instead of majjhu, and the Sanskrit word nirvāṇa instead of the Prakritic nivvāṇa.38 The reader will also notice the preference for the retroflex ṇa over na, characteristic of Prakrit languages in general. However, the import of this verse is the same, highlighting the emphasis on semantic content over form, contrasting with the use of dhāraṇīs and mantras.

This verse also appears in three sādhanās within Saroruha’s practice lineage, at similar key junctures. In the Hevajrasādhanopāyikā the verse occurs in the mantra and visualization phase of the sādhanā, where after visualizing the rows of letters, and entering himself and everything becoming one (Skt: ekībhūtvā),39 the practitioner recalls the same verse quoted earlier.40 In the Vajrapradīpa, the verse occurs as a description of the union with mahāmudrā in which experience itself ceases,41 and by contemplating this verse, the practitioner also enters into luminous clarity.42 The verse also appears in the Dveṣavajrasādhana, occurring after the practitioner withdraws the troop of ḍākinīs, realizes the goddess of emptiness, Nairātmyā, within their own consciousness, and meditates until there is no more perception.43 In each case, the dohā occurs at a crucial stage in the ritual, where the practitioner abides within emptiness, in union with mahāmudrā, where perception itself ceases.44

This association with Saroruha (Saraha) is underscored by its appearance in a dohākoṣa attributed to him,45 which is especially significant in light of the fact that another dohā in this dohākoṣa makes an appearance in the Kurukullā sādhanā discussed above. Within this ritual, long before the practitioner recites the “āi na” dohā above, the practitioner is consecrated with the five nectars,46 whereupon the Buddha Amitābha emerges from the crown of their head; and after they worship the eight goddesses with their bodhicitta47, the universe dissolves in the seed syllable in one’s heart.48 Next, the practitioner recites the following dohā:

akkharamantavivajjiyao na u so vinda ṇa vitta |

eso paramamahāsuhao na u pheḍia ṇa u khitta ||

Devoid of syllables and mantras, bindu, and turnings [of the mind]

This is the Supreme Great Bliss, without a tiller or field.49

The reader will notice from the outset that this verse shares pada c with the previous dohā, with only a minor phonetic substitution, showcasing the ways in which these dohās can be built around other well-known phrases. However, the verse is still describing a similar state of being, the mind abiding in emptiness, in which all referents, even mantras and other elements of practice, dissolve into the Supreme Great Bliss.

As mentioned above, both of these dohās also appear in a dohākoṣa attributed to the very same Saraha. The “akkhara” verse appears with some significant changes in vocabulary:

akkhara-vaṇṇa-vivajjiaü ṇaü so biṃdu ṇa cittu |

eu su parama-mahāsuhaü ṇaü kheḍaü ṇaü khettu ||

Free from syllables and colors; with neither bindu nor mind.

This is the Supreme Great Bliss: there is neither the tiller, nor the field.50

Here the verse substitutes vaṇṇa (Skt: varṇa) for manta (Skt: mantra), along with cittu (Skt: citta) for vitta (Skt: vṛtta), as well as exhibiting phonetic differences (such as the elision of the consonant -h- in ehu, etc.). These changes in phonology and vocabulary, however, do not substantially change the meaning of the verse, for it still describes the Supreme Great Bliss as being beyond the reach of the conventional mind. This fluidity in phonology and vocabulary, paired with the fidelity in terms of inner semantic and didactic meaning, is characteristic of Tantric Buddhist dohās as a whole, and is shared by the āi verse appearing in the same collection:

āi ṇa aṃtu ṇa majjhu tahim̐ ṇavi bhavu ṇavi ṇibbāṇu |

ĕhu so paramamahāsuhaü̐ ṇavi paru ṇavi appānu ||51

The changes to this verse are more cosmetic than in the “akkhara” verse, employing the Prakritic vi/bi (Skt: api) with the negative particle ṇa, as well as rhyming with nasalization on tahim̐ and -suhaü̐. While differences in spelling within texts can often be attributed to scribal error, these phonetic differences attest rather to the primarily oral nature of these dohās. These songs are performed, either literally or in the imagined, and as such are prone to changes in phonology or terminology, whereas the didactic meaning is more static. It is common to substitute words or alter pronunciation in the performance and transmission of these verses, given that their purpose is primarily didactic. Their usage at key junctures within Tantric Buddhist texts highlights their performative nature, where they are used like mantras and dhāraṇīs. Like mantras and dhāraṇīs, these dohās are recited at key junctures in sādhanā, in this case during the experience of the Supreme Great Bliss, whereas others are recited during offerings, gaṇacakra feasts, or to impel the practitioner to continue in their practice. Furthermore, like mantras and dhāraṇīs, these dohās stand out in both form and content from the surrounding text, expressing another level of language beyond the simply discursive. They are all examples of ritual language, language whose purpose is not merely to convey information, but rather to disrupt ordinary discursive language to access a state of being beyond the ordinary and discursive. However, here dohās are distinct from mantras and dhāraṇīs in not honoring precise phonetic reproduction while still conveying didactic meaning.

3. Saraha’s Dohās within the Buddhakapāla Tantra

In the Buddhakapāla Tantra, we encounter more dohās attested in collections attributed to Saraha, used in similar ways to those in Hevajra literature but with a particular emphasis on capstone summary. The Buddhakapāla Tantra has been dated to the 9th century CE or later,52 and shares with the Hevajra Tantra its mortuary symbolism, transgressive rhetoric, and sexual practices, while being notable for its detailed instructions for curing snakebite in its fourth chapter.53

The first Saraha dohā considered here appears at the end of the ninth chapter of the text, on practice (Skt: caryā). In this chapter, the practitioner is enjoined to wander about in the mode of a Tantric ascetic, wearing no clothing and carrying a skull-bowl (Skt: kapāla), released from all hindrances and accepting everything given to them without hesitation. By means of such nonattachment, the practitioner will attain the state of mahāmudrā, which is tantamount to liberation.54 At the end of this description, the chapter ends with the following dohā:

jatha pi tatha pi jaha pi taha pi jena tena hua buddha |

saim viappe ṇāsiā saala sahāve śuddha || (9.9)

Wherever, in whatever manner and with whatever, one becomes awakened; due to conceptualization (the mind) is destroyed by itself; all things, [however] are pure by nature.55

This verse is attested in the dohākoṣa of Saraha discovered by Sankrityayana. In this version, phonetic differences abound; Luo’s critical edition is far more Sanskritic in character, while the version in Saraha’s dohākoṣa substitutes the Prakritic vi for pi (Skt: api) as well as endings in -u, in addition to substituting jagu (Skt: jagat) for saala (Skt: sakala) and saṃkappeṃ (Skt: saṃkalpena) for viappe (Skt: vikalpena).56

The next verse occurs in the thirteenth chapter, on “the complete purification of mind” (Skt: citta-viśuddhi-paṭalaḥ). This chapter is concerned with the senses and the apprehension of wisdom. The beginning of the chapter lists the different kinds of women the practitioner should desire, including family members as well as the wives of low-caste members of society. Far from advocating actual incest and adultery, this passage should be interpreted as extraordinary language (Skt: sandhyā-bhāṣa), and the following section defines the various women listed as embodied in the senses and impure mental states. By purifying one’s mind and perceiving the emptiness of the mind that examines, the practitioner can then be free of attachments, and experience all sensory experience as mahāmudrā.57 The chapter ends with this capstone dohā:

yāvan na viṣaya unmuliaï tāma budhataṇu kema |

searahianavaaṃkuraha tarusampatti na jema || (13.24)

So long as not nourished by external objects, how then could there be Buddha, the wish-fulfilling tree? Just as a tree cannot grow from a new sprout without water.58

This verse is also attested in the dohākoṣa discovered by Sankrityayana, in addition to a collection of verses edited by Cecil Bendall called the Subhāṣita-saṃgraha.59 The differences between these versions are even more stark than in the previous dohās; apart from the expected phonetic differences, the Buddhakapāla Tantra prefers Sanskritic yāvan (Skt: yāvat), while the version in the dohākoṣa uses the Prakritic jaï (Skt: yadi). The verb in the first pada is subject to variation; the Buddhakapāla Tantra uses unmuliaï (Skt: unmūlayati), which hardly makes sense, since its primary meaning is “to uproot.” Abhayākaragupta glosses this verb as paripālyate,60 which aligns more with the verbs in other versions: in the dohākoṣa, the verb is kīliaï (Skt: krīḍyate), while the Subhāṣita-saṃgraha version has ṇullanti (Skt: nudati).61 The differences here are profound; just as the differences in vocabulary within the Hevajra verses cannot be attributed to phonetic difference or scribal error, here verbs can be substituted ad hoc.

The last dohā, on recitation and requesting a wisdom consort, appears in chapter 14. The chapter begins with an extended panegyric to the goddess Citrasenā, the goddess of wisdom and Tantric consort of the deity Buddhakapāla. Afterwards, the chapter lists various mantras to be recited for particular goals (Skt: siddhi), and the procedure for initiates to request a wisdom consort, including maṇḍala offerings, fragrances, food, drink, and ornaments, upon which she herself is worshipped. Thereupon the chapter cites the following dohā:

indīyattha vilayagaü naṭṭa vi appasahāva |

so hale paramāṇandagaï phuḍa pucchaha gurupāa || (14.23)

The faculties and the objects have been dissolved, and the intrinsic nature of Self has also been destroyed. Hey good friends! This is the way of the Highest Bliss, [if you want to know it] ask clearly the venerable Guru!

This verse quotation is slightly different from the others in this text, since it does not encapsulate the teaching of the chapter so much as describe the experience of bliss during the peak of sexual union. As with the other verses this dohā is attested in the dohākoṣa discovered by Sanskrityayana; however, it also makes an appearance in a manuscript discovered by Bagchi.62 As in other dohās, the different versions have numerous phonetic differences, while the Buddhakapāla version uses paramāṇanda, and the two dohākoṣa versions use sahaja [-aṇaṃda/-u], another example of ad hoc substitution.

Taken as a whole, the Saraha dohās in Hevajra literature and the Buddhakapāla Tantra display clear evidence of phonetic and terminological substitution throughout their various quotations, with a noticeably more static semantic meaning. Given the unlikelihood of these changes occurring through scribal error, this is clear evidence of their primarily oral nature, according to which they are quoted to fulfill didactic and ritual purposes. These dohās fall into various genres with particular ritual functions; in this case, either to summarize chapter contents or to provide apophatic descriptions of ultimate reality.

4. On the Uses of Dohās

In her work on dohās and Sānt teachings, Karine Schomer examined the uses of Apabhraṃśa and Hindi-language dohās, and arrived at a taxonomy of uses: independent verses quoted in other works, independent verses gathered together in collections, building blocks around which longer narrative cycles (prabandha) are constructed, the concluding verses for stanzas in other, narrative meters, refrains (ṭek) for sung lyrics (pad), lines of song lyrics, and in combination with other meters to create new metric forms.63 She also notes that these verses (sākhīs) were often used like the sūtras of philosophical discourse.64 Many of these uses hold true for Tantric Buddhist dohās, such as independent verses gathered in collections and concluding verses for larger works (as seen in the Buddhakapāla Tantra). Her observation that they are often used like sūtras in philosophical discourse is also apposite, although it is not quite a correct description of the Apabhraṃśa dohās in this paper. Rather, they are examples of ritual language in the diglossic Tantric Buddhist language ideology, used at key moments, like mantras and dhāraṇīs, to convey different layers of discourse and communication. Further study of these dohās will no doubt expand this model and taxonomy further.

Furthermore, Schomer emphasizes how the dohās became a particular vehicle for the teachings of the Sants and other low-caste nirguṇa bhakti poets. Indeed, the dohā was associated with and used by low-caste nirguṇī bhaktas,65 such that no less a figure than Tulsidās decries the use of dohās and other similar meters by “vile poets” who disrespect the Vedas and Purāṇas.66 This disrespect towards the Vedas and Purāṇas is probably inflated for rhetorical effect, but it also calls to mind the name used by Sānts to refer to dohās, sākhīs (Skt. sākṣin), which, as noted earlier, is a relic of Buddhist ideas within nominally Hindu traditions (Buddhists being nominally hostile to Hindu religion and coining the term sākhī). However, how did this Buddhist literary form enter the repertoire of the Nāth Siddhas, who themselves were the ancestors of nirguṇa bhakti traditions? As we will see, it resulted from the loss of Buddhist institutions in the 13th century.

5. Language and Social Topography

The social topography of late Indian Tantric Buddhism has proven to be a controversial topic. In their portrayals, the Buddhist mahāsiddhas come from all walks of life, consorting with low-caste and tribal peoples, disregarding normative purity injunctions, and taking on extreme transgressive practices, all in their search for an ultimate reality beyond the narrow habitus that keeps human beings in spiritual bondage. The stories of their lives also seem to align with the antinomian aspects of Tantric Buddhism itself, particularly the sexual and mortuary practices that would demand that a practitioner break their monastic vows. This leads to the obvious question of an extra-monastic source for these practices, outside of the monastery that had long been the backbone of Buddhist institutions in India.

On this topic, Wedemeyer proposes that the divide within medieval Indian Buddhist society lay not between laypeople and monastics, but rather between professional and amateur practitioners, and that antinomianism is entirely operative within a normative Buddhist framework, and not the outcome of different social classes in conflict.67 Furthermore, regarding Tantric Buddhist language practice, he asserts that the Buddhist Tantras were composed by Buddhist professionals in “standard Sanskrit, a learned language”,68 with Apabhraṃśa being merely a “pan-Indic koine” used “to suggest rural simplicity and joyful vulgarity.”69 The observation that Apabhraṃśa was a pan-Indic koine is significant, for, as noted earlier, the Apabhraṃśa used in dohās is quite distinct from the proto-Bengali of caryā songs, showing the performative nature of this language. Apabhraṃśa was a conscious stylistic choice for the composers of Tantric Buddhist dohās; however, it is an oversimplification to reduce them to a mere invocation of “rural simplicity and joyful vulgarity,” given their use within Tantric Buddhist texts. Instead, as the examples in this article show, they are used as a common reservoir of songs and teaching, expressing the ineffable nature of enlightenment, in addition to their other uses (“goddess songs” and others).

Furthermore, while Wedemeyer’s centering of normative Buddhist institutions as the backbone of Tantric Buddhist practice is a useful corrective to any crude narrative ascribing purely extra-Buddhist origins to Tantric religion,70 it goes too far in the other direction by ascribing purely intra-Buddhist origins for Tantric Buddhism.71 Davidson and Burchett have both examined the medieval expansion of Brahmanical and Buddhist influence into previously peripheral areas, in which political expansion and centralization was inseparable from the assimilation of local deities and practices into the more cosmopolitan religions of Brahmanism and Buddhism.72 Here, high-status Brahmins and Buddhists would claim and utilize tribal practices, if within a Brahmanical or Buddhist framework. Burchett quotes from Frederick Smith, who sees in the Tantric practice of nyāsa a controlled and Brahmanical practice of possession, wherein the uncontrolled possession rites practiced by marginal and tribal peoples is replaced by a controlled and textual practice within a programmatic Brahmanical context.73 Therefore, the growth of Tantric Śaivism and Buddhism entailed the adoption of marginal and tribal practices, under the control of authorized elites.

However, at the same time, the question is begged whether it is even useful to posit a monolithic Tantric Buddhism. Wedemeyer’s centripetal impulse has the effect of reducing Tantric Buddhist practice to the sole province of elite Buddhist insiders; however, his invocation of Tantric Śaivism as a point of comparison belies his thesis.74 Just as in Buddhism, in Tantric Śaivism the higher antinomian traditions formed elite cults that relied upon the mainstream Śaiva Siddhānta as a ritual and institutional base, while transgressing the latter’s purity norms in ritual. However, there is no indication that adherents to the mainstream Śaiva Siddhānta acknowledged the superiority of the antinomian traditions over their own, and indeed maintained separate and opposed metaphysical views and ritual practices.75 This tacit but concrete admission of non-uniformity within putatively unified Tantric traditions is underscored by the porous boundaries between nondualist Tantric Buddhism and Śaivism, which share iconography, ritual syntax,76 and even explicit textual relationships.77 This is strong evidence for a stratified Tantric world, in which the normative dualistic core of the tradition is surrounded by more exclusive cults that nevertheless fraternized across sectarian divides and competed with each other in the spiritual marketplace.

This milieu of Siddhas (upon which Buddhists affixed the mahā- designation), would survive the death of institutional Tantra in India, for after the Arab and Turkic invasions in the 13th century CE, large institutions such as Tantric Śaivism and Buddhism lost the patronage and infrastructure necessary to sustain themselves,78 and in response, Tantric practices were stripped of their complex rituals and metaphysical trappings, emphasizing laya and haṭha yoga instead.79 These techniques were simple and portable, relying on the human body, breathing techniques, and physical practices, which allowed the practice of yoga to reach a far greater audience. I propose that the practice of reciting Apabhraṃśa dohās also survived this wide-scale collapse in Tantric institutions due to its own simplicity and portability. Singing dohās (or as the Nāths called them, sākhīs) required no complex ritual or institutional support, and could be used ad hoc at crucial junctures in Tantric practice to convey teachings and the experience of realization. It was this portability and rhetorical punch of Apabhraṃśa (and soon Hindi) dohās that made them so amenable to expressing nirguṇa experience, a communitas beyond the distinctions and boundaries created and policed by stakeholders in difference.