1
The Tantric Age
Tantra and Bhakti in Medieval India
n order to understand the rise of bhakti in early modern North India and
its historical significance, we must first look back to India’s early medieval
period (ca. 600–1200), a time we can characterize as “the Tantric Age.”1
From roughly the seventh to the thirteenth century, the thought, ritual practice, and institutional presence of tantric traditions played a major role in the
life of South Asians. As Gavin Flood remarks, “The cultural, religious and political history of India in the medieval period cannot be understood without Tantra.”2 Critically, however, tantra’s rise to prominence was inseparable from the
growth of popular traditions of devotion, or bhakti, with which tantra forged
symbiotic relationships. In this chapter, I examine the tantric tradition in early
medieval India—particularly its relationships with state power and popular
forms of devotional religiosity—in order to set the stage for the book’s consideration of the relationships between bhakti, tantra, and yoga that emerged in
late Sultanate and Mughal India. Tantra first arose as an esoteric tradition for
initiated elites seeking liberation (mokṣa) or extraordinary powers (siddhi), but
it later became deeply involved with royal power and with India’s public temple cult (and the political and agrarian expansion linked to it), making tantric
ritual, institutions, and ideals of sacred power—epitomized in the figure of the
tantric yogī/guru—a fundamental part of mainstream Indian social, religious,
and political life.
Scholars have often emphasized the esoteric and fundamentally transgressive nature of tantra, yet transgression was quite marginal to the “mainstream”
tantric tradition I focus on here. This mainstream tantra was simultaneously
both esoteric and popular, brahmanical and folk. This chapter demonstrates how
I
burc19032_1st_i-412.indb 29
—-1
—0
—+1
12/12/18 6:14 PM
30 9 Part i: FroM MEdiEval tantra to Early ModErn Bhakti
tantric monastic orders and their institutions became integral players in an
early medieval religiopolitical economy that linked lay bhaktas, tantric yogīs, and
kings in exchanges of economic, sociopolitical, moral, and spiritual capital. In
the process it reveals how, in sharp contrast to the bhakti of early modern North
India, bhakti in this period is regularly subordinated or assimilated to tantric
ritual or yogic values and practices ( jñāna, dhyāna, etc.).
What Is Tantra?
-1—
0—
+1—
The tantric traditions rest on the foundation of a vast body of tantric scriptures,
primarily termed Tantras, Āgamas, and Saṃhitās, that were composed in
Sanskrit between the fifth and ninth centuries—in Śaiva, Vaiṣṇava, Saura,
Buddhist, and Jain contexts—as well as on a number of other important (usually more exegetical) tantric works that were produced into the thirteenth
century. 3 As several tantric studies scholars have made clear, these three
designations—Tantra, Āgama, and Saṃhitā—were synonymous and interchangeable terms for tantric scriptural revelation, thus in the pages to come I follow
common practice in using the term “Tantras” to refer to the tantric scriptures in
general.4 In the earliest phase of the tradition, the Tantras were concerned primarily with the various ritual techniques used in the initiated practitioner’s individual quest for spiritual liberation or occult powers. Certain branches of early
tantric scripture (e.g., the Bhūta Tantras and Gāruḍa Tantras) also concern themselves with protection against and treatment of demonic possession, poison, disease, and other dangers or misfortunes related to the health and livelihood of
individuals and communities. In the later, post–eleventh century development
of the tradition in South India, many tantric scriptures came to focus on aspects
of public religious and political life, such as the building of temples, consecration of kings, and conducting of public rites of worship.
The earliest extant tantric Śaiva scripture that we know of is the
Niśvāsatattvasaṃhitā, the oldest sections of which were composed probably
between 450 and 550.5 The text’s central innovation is the teaching that liberation (mokṣa) can be gained through tantric initiation (dīkṣā) itself. In this early
scripture we can already see the core features that would come to characterize
tantra more generally—namely, (a) tantric initiation (a liberating initiation,
given by an enlightened guru and available to householders and all castes);
(b) the ritual divinization of the body (i.e., the “consubstantiation” of the practitioner with the deity “in a transforming infusion of divine power”);6 (c) the use
burc19032_1st_i-412.indb 30
12/12/18 6:14 PM
thE tantric agE = 31
of tantric mantras; and (d) a conception of the Divine as immanent, accessible
power that can be employed for bhukti or mukti.
The Tantras claim to be supremely authoritative teachings descended straight
from the mouth of the gods. Medieval Hindu tantric communities typically recognized the Vedas as a legitimate but lower echelon of scriptural revelation
that the Tantras include and transcend.7 In order to access the “higher” truths
and practice the “more powerful” ritual methods taught in the Tantras, one first
had to be initiated. Initiation into tantric teachings had great appeal because
they offered new ritual techniques and potent tantric (non-Vedic) mantras that
were understood to be more efficacious in—and, indeed, entirely necessary for—
achieving the goals of spiritual salvation (mukti) or extraordinary powers and
enjoyments (siddhi/bhukti). Certain initiatory forms of Śaivism preexisted tantra, but these Atimārga Śaiva traditions focused exclusively on the goal of liberation, demanded renunciation from initiates, and typically admitted only
brahman males. Tantric traditions opened up initiation to all caste classes, and
even women, and did not require the renunciation of family life and traditional
social obligations.8 Hindu tantric traditions typically claimed that their major
initiation ritual was unique in itself effecting salvation. In this tantric initiation rite, the guru uses the power of non-Vedic mantras to destroy the previous karma of the initiate, purifying his soul of all impurities and stains (mala)
and allowing him to identify with God and realize the power of the Divine. As
Elaine Fisher explains, “The implications of this assertion—that a mere ritual,
in and of itself, possesses the means to sever the bonds that tie the individual
soul to transmigratory existence—radically recast the sociological implications
of elite Indic religion.”9 In offering this ritual initiation to a wide array of social
groups (i.e., not just brahmans and renouncers), tantric Śaivism “effectively circumvented the strictures of varṇāśramadharma, providing both kings and
Śūdras with access to liberation.”10
The Śaiva Āgamas came to articulate four basic classes of tantric initiates:
(1) the samayin, or entry-level community member; (2) the putraka, who has
received the primary, liberating initiation (nirvāṇa-dīkṣā) and whose only goal
is liberation; (3) the sādhaka, who is authorized to practice a special discipline
in order to acquire extraordinary powers (siddhis) and heavenly enjoyments; and
(4) the ācārya, or guru, a community leader granted the privilege and power to
give initiations, perform temple worship (pūjā) and installations (pratiṣṭhās), and
comment on tantric scriptures.11 In tantra, the guru is a spiritually realized
adept in and through whom the Divine acts (i.e., who is the vessel of, or even
nondifferent from, God) and who—in a direct relationship with his disciples—
transmits the knowledge necessary to conduct tantric ritual.
burc19032_1st_i-412.indb 31
—-1
—0
—+1
12/12/18 6:14 PM
32 9 Part i: FroM MEdiEval tantra to Early ModErn Bhakti
-1—
0—
+1—
In most tantric systems, regular ritual action is required to maintain the
purity and power attained in the main tantric initiation (nirvāṇa-dīkṣā) and to
thereby ensure liberation. The daily ritual worship (pūjā) of the tantric initiate
involves the systematic use of mantras and intricate visualization meditations
to purify and empower a subtle body understood to have homological connections to the rest of the entire cosmos and to be, at its core, inherently divine—i.e.,
suffused with the same energy and pure consciousness as the Divine. Tantric
ritual most differentiated itself from mundane brahmanical Śrauta and
Smārta rites in offering a method for divinizing the body and infusing oneself with divine power through consubstantiation with a deity.12 As Alexis
Sanderson has pointed out, this method is remarkably uniform across tantric
traditions, as all forms of tantric religion share a single ritual system whose
deeper structural unity is not significantly affected by differences such as
the choice of deity invoked and the character of the visualizations, mantras,
and maṇḍalas used.13 The general ritual structure found in the practice of all
tantric traditions consists of (a) the purification (bathing) of the external,
physical body (snāna), then (b) purification of the cosmic elements within
the subtle body (bhūtaśuddhi or dehaśuddhi), followed by (c) the divinization
of the body by placing mantras upon it (nyāsa), then (d) internal worship of
the deity (antara/mānasa-yāga) utilizing only visualization and the power of the
mind/imagination, and, finally, (e) external worship (bahya-yāga) of the deity
with ordinary devotional offerings such as fruit, flowers, incense, and bells.
As Gavin Flood has remarked, the notion “that to worship a god one must
become a god is a notable feature of all tantric traditions.” More specifically,
he has stressed that “the ritual construction of the body as the deity through
the use of . . . mantras is prototypically tantric.”14 In both its soteriological
aim and its ritual method, then, tantra was all about “becoming God.” It was
especially in this goal (and its associated ritual technologies) that tantra
“definitively shifted the paradigms of Indic religious practice and theology
for centuries to come.”15
If initiation and the divinization of the self mark two essential elements of
tantra, just as important to tantric religiosity is the mantra. The fundamental
religious instrument of the tāntrika is the mantra, understood as the sonic
form—the sound body—of a deity or aspect of the Divine. While mantras were
important in other traditions of South Asian religiosity, the Tantras were
unique in conceiving mantras as the vibrational forms of deities. As Shaman
Hatley states, “This ontological identification of efficacious sonic formulae
with divinities is distinctive to the tantric traditions.”16 It was the potency and
agency of these non-Vedic mantras that made tantric ritual so efficacious, thus
the use of mantras is often considered the most fundamental component of
burc19032_1st_i-412.indb 32
12/12/18 6:14 PM
thE tantric agE = 33
tantric practice. Indeed, tantric works distinguished and described their
teachings as “the Way of Mantras” (mantramārga or mantrayāna).17
In addition to the foundational features just discussed—initiation, divinization of the self, and tantric mantras—there is a particular understanding of the
sacred and an assumption about the nature of the cosmos that seems to structure all tantric practice. In the tantric traditions, the Divine or sacred is conceived in large part as tremendous, immanent, and accessible cosmic power or
energy. In tantric Śaivism and Śāktism, specifically, this power is typically
understood as feminine in nature and identified as śakti, an awesome, infinite
force (potentiality, capacity, energy) that pervades the universe in many forms—
circulating throughout the human body, the social body, and the body of the
cosmos—and that can be harnessed for any variety of purposes.18 As André
Padoux remarks, “the Tantric vision is that of a world issued from, upheld
and completely permeated by, divine energy (śakti), which is also present in
the human being who can harness and use it (her, rather) for worldly as well
as ritual aims and for liberation.”19 Similarly, Douglas Brooks states, “The
Tāntrika conceives of the world as power. The world is nothing but power to
be harnessed.”20
For most medieval Indians, behind victories in battle, successful pregnancies, and good harvests, as well as illnesses, droughts, floods, and untimely
deaths, was sacred agency of some kind. But this cosmic power typically was
not understood abstractly; rather, it was usually conceived to manifest as the
power of very specific divine or semidivine beings or energies. In practice, then,
to harness divine power often meant to control, pacify, or gain the favor of specific invisible beings or forces so as to protect and further one’s own interests.
The literature of the medieval period assumes a universe made up of a great
spectrum of beings and energies—ranging from great gods like Śiva and Viṣṇu
to a vast array of goddesses, nature spirits, and malevolent demons to the śakticharged life forces (e.g., prāṇa, kuṇḍalinī) within one’s own body—whose power
the tantric practitioner could realize, take hold of, and manipulate through the
use of distinctively tantric mantras and ritual practices.21 Clearly, tantric ritual practices were not aimed simply at spiritual liberation; just as often they
also—or instead—sought to control and employ divine power for this-worldly,
pragmatic reasons.22
With this in mind, it is important to note the importance that the quest for
extraordinary powers, or siddhis, held in the tantric tradition. The pursuit of
siddhis by practitioners of asceticism (tapas), yoga, and sorcery was a timehonored one, but one given a uniquely privileged place in tantra. Unlike orthodox traditions, tantra encouraged the pursuit of occult powers and heavenly
pleasures as valued goals alongside the aim of liberation. In particular, the
burc19032_1st_i-412.indb 33
—-1
—0
—+1
12/12/18 6:14 PM
34 9 Part i: FroM MEdiEval tantra to Early ModErn Bhakti
-1—
0—
+1—
sādhaka class of tantric initiates—who aimed quite deliberately at the acquisition of siddhis—might be considered as the tantric institutionalization of the
age-old path of the siddhi-seeking ascetic-yogī.23
India has an ancient and well-known tradition of ascetics whose practices
are believed to result in the possession of special knowledge and extraordinary
powers. Indian epic and purāṇic literature is filled with ascetics who, through
their tapas (ascetic “heat”), have earned the power to accomplish any desire, to
give curses, grant boons, and even to coerce the gods. A fifth of Patañjali’s Yoga
Sūtra is dedicated to the topic of siddhis—understood to result naturally from
yoga’s forms of intense mental concentration—including invisibility; superhuman strength, hearing, and sight; knowledge of past and future lives; control
over hunger and thirst; knowledge of the thoughts of others; becoming tiny or
gigantic, light or heavy; entering into the bodies of others; reaching any place
by willing it; and controlling natural elements and animals. The ascetics and
yogīs thought to possess these siddhis often also played the roles of sorcerer
(vidyādhara), shaman, and healer, offering services to the wider populace including protective amulets, generation of wealth, magical harming of enemies,
love potions, exorcism/healing, and divination, among others. Siddhis, then,
were more than mere entertainments or proofs of sanctity; ascetics and yogīs—
who might be shaven-headed monks as often as scantily clad, dreadlocked
wanderers—often relied upon the (perception of) possession of these “superpowers” to perform the variety of tasks desired by their patrons and employers. The powers that yogīs and ascetics were thought to possess garnered them
the fear and respect of others, but more importantly they made them valuable
service providers at every level of society who were sought out to ritually effect
a fruitful harvest, a successful pregnancy, or a victory in battle. In other words,
these ascetics’ occult powers could be employed to empower the actions, achieve
the desires, and protect the health and well-being of their clients, be they
regional kings or village peasants. Much of tantra’s growth seems to have been
a function of how tantric gurus and sādhakas came to “corner the market” in
supernormal power and were widely sought out for their renowned ability to
harness sacred power for pragmatic this-worldly purposes.
For those in search of extraordinary power, initiation as a tantric sādhaka
offered access to a new body of uniquely efficacious techniques (centered especially on the repetition of powerful tantric mantras) for acquiring siddhis.
Furthermore, since the Divine was conceived especially as power in the tantric traditions, the siddhis were not seen simply as a natural by-product of—but
potential obstacle to—one’s spiritual growth (as the classical Pātañjala-yoga tradition would have it) but were considered by many tāntrikas as the very essence
of that spiritual development, a sign of the unveiling of divine omnipotence
burc19032_1st_i-412.indb 34
12/12/18 6:14 PM
thE tantric agE = 35
within one’s self. The sheer amount of tantric literature dedicated to the
sādhaka’s pursuit of siddhis indicates that tantric mantra-centered rites became
a very important practice among the power-seeking ascetics of early medieval
India.
Marion Rastelli’s research on the tantric Vaiṣṇava (Pāñcarātra) scripture, the
Jayākhya Saṃhitā (ca. ninth century), gives us a better sense of the religious life
of the siddhi-seeking tantric sādhaka. After many years of dedicated, isolated
ascetic practice, the sādhaka comes to master his mantra, i.e., to possess and use
its power (mantra-siddhi).24 At this point, the sādhaka can perform rites for himself or others, specifically bhaktas (in this case Vaiṣṇava bhaktas) who have
requested his help and are not themselves able to master a mantra.25 Here the
text alludes to the important interaction of lay devotees and initiated tantric
adepts, explored later in this chapter, with bhaktas seeking out the magicoreligious services of professional tantric ascetics who themselves depended in
significant part on patronage from the bhaktas. The sādhaka might use his
mantra-siddhi to provide a variety of services, including the performance of rites
to exorcise, pacify, or protect against illness-causing demons; to treat poison;
to bring about good health, longevity, contentment (tuṣṭi), prosperity (puṣṭi), or
dominion over other beings (including defeat of enemies); to prepare pills giving special powers like flight or invincibility; to cause or stop rain; to produce
food; or to bestow fertility and good luck in pregnancy and childbirth.26 The
intense demands of the sādhaka’s rituals must have ensured that few took
up this path; however, it is clear that this small group of tantric elites—in
the services they provided and the possibility of extraordinary power they
represented—were a crucial part of tantra’s authority.
Considering all of this, I am now in a position to concisely articulate this
book’s approach to tantra. In the pages to come, I shall understand tantra as
the tradition of specifically tantric ritual techniques used to worship, realize,
and exercise sacred power. What makes these ritual techniques specifically tantric is that they are authorized by and taught in tantric scriptures, their practice requires tantric initiation, their primary effective instruments are tantric
mantras (i.e., non-Vedic mantras understood as the sound bodies of deities), and
they typically involve the ritual self-deification of the practitioner. This strict
definition of tantra suggests an esoteric tradition consisting only of dedicated
initiates and specialist ritual performers, but as I discuss, these individuals
might be better understood as a single stratum—though the earliest and most
essential one—in what would become a larger, popular culture of tantra whose
rituals, institutions, and cosmological understandings pervaded much of the
early medieval social world. Behind this larger culture of tantra lay a distinctly
tantric outlook or sensibility whose central element, in White’s words, was “that
burc19032_1st_i-412.indb 35
—-1
—0
—+1
12/12/18 6:14 PM
36 9 Part i: FroM MEdiEval tantra to Early ModErn Bhakti
human practitioners can empower, and even deify, themselves to manipulate
and dominate the entire spectrum of beings and energies that make up the tantric universe.”27 This tantric perspective and many of its associated practices
would prove quite distasteful to the outlook and sensibility at the heart of early
modern North India’s bhakti movement.
The Bhakti in Tantra and the Tantra in Bhakti
-1—
0—
+1—
What was the place of bhakti in medieval tantric religiosity? It is important to
note that bhakti was not at all absent from the tantric ritual process I have been
discussing, for devotion was often key in developing the closeness necessary for
the practitioner to identify with the deity. As André Padoux notes, “Tantric texts
often say that a given practice or rite is to be performed with devotion
(bhaktyā),”28 and Sanderson describes the heart of medieval tantric religious life
as “routinized ritual duty more or less qualified by the sentiment of devotion,”
or bhakti.29 Yet in medieval tantric communities bhakti was generally understood
not as passionate, emotional love so much as faith, reverence, and service and
was typically subordinated to ritual actions, techniques of self-empowerment,
and the quest for liberating knowledge ( jñāna).
When discussing the relationship between bhakti and tantra, Alberta Ferrario
points out, it is crucial to recognize the diversity of tantric traditions (e.g.,
Pāñcarātra, Śāiva Siddhānta, non-Saiddhāntika Śaivism), which did not necessarily conceive of bhakti in the same way, as well as the historical change within
these traditions, since the conception and role of bhakti in certain tantric traditions changed significantly over time.30 The work of Sanjukta Gupta, Gerhard
Oberhammer, and Marzenna Czerniak-Drożdżowicz, for instance, demonstrates that the Vaiṣṇava tantric tradition of the Pāñcarātra gave a central
place to emotional bhakti only in its later development.31 In the early (preninth-century) phase of the tradition, Pāñcarātra texts focus not on humble,
passionate devotion but on ritual acts of worship and yogic meditation (on/
with mantras), tantric ritual practices that were only for initiates.32 Later, a
devotion of emotional self-surrender (prapatti) enters into the Pāñcarātra tradition, but it only very gradually (over centuries) comes to take a predominant
position over and above ritual and yogic modes of worship. Similarly, it was
only in the post-twelfth-century Śaiva Siddhānta of South India, under the
influence of the Tamil devotional tradition, that bhakti took on a new and central role in the path to salvation, as well as a passionate, emotional quality absent
in earlier Śaiva Siddhānta sources. 33 While the modern- day Śaiva Siddhānta
burc19032_1st_i-412.indb 36
12/12/18 6:14 PM
thE tantric agE = 37
tradition understands the bhakti hymns of the Tevārām as a fundamental part
of its canon, in fact these vernacular bhakti songs were not in any sense considered to be works of the Śaiva Siddhānta tradition in the period in which they
were composed, or for several centuries afterward, being incorporated into the
tradition only in the thirteenth century, nearly five hundred years after their
composition.34
Ferrario’s research demonstrates that bhakti in pre-twelfth- century Śaiva
tantric traditions was ordinarily conceived of as attitudes and actions of reverence, obedience, faith, and service but not as a cultivation or expression of emotion, passion, or love. When one examines the context in which the term is
used in pre-twelfth-century Śaiva Tantras and exegetical works, she explains,
it is clear that bhakti generally refers “to a devout attitude that manifests as the
desire to receive instruction from a Śaiva teacher; faith in the Śaiva scripture;
good disposition towards the Śaiva community; and the choice of Śiva as one’s
deity.”35 In these tantric texts, the expression “devotion to God/Śiva” (deve
bhaktiḥ; Śive bhaktiḥ) is usually found together with bhakti for the guru, the Śaiva
Āgamas, or one’s fellow Śaiva devotees and carries a meaning “closer to the
semantic field of terms including paricaraṇa (attendance, service), śraddhā (faith),
and viśvāsa (belief, faith), rather than love and affection.”36 Ferrario also shows
that in pre-twelfth-century tantric Śaiva sources, bhakti was typically not considered a means to salvation (as it is in bhakti traditions) but rather a sign of the
descent of Śiva’s grace (śaktipāta) upon a person, which was a prerequisite for
initiation. In other words, whether God’s power had descended upon individuals, awakening them to the potential of their true nature and thus making them
eligible for full tantric initiation, could be inferred from the quality of their
devotion.37 For full tantric initiates, even on the rare occasions when devotion
was impassioned and ecstatic, it was viewed as a sign of—or an affective experience bound up with—realization of the Divine but was not conceived of as a
method or path for achieving that realization. As Ferrario explains, even the
uncommon mentions of passionate, emotional bhakti in the literature of tantric communities—for example, in Utpaladeva’s Śivastotrāvalī (ca. tenth century) and some other stotra collections38—conceive bhakti as an experience
equivalent to (or concomitant with) the end goal of liberation, but not as a
means to that goal. 39
While the bhakti of the lay devotional tradition seems to have seeped into tantric texts on occasion—see, for instance, Jason Schwartz’s discussion of the
Mataṅgapārameśvara (ca. 600), a supplementary Āgama associated with the Śaiva
Siddhānta tradition40—orthodox tantric exegetes clearly subordinated bhakti to
the performance of ritual or the attainment of gnosis. In the dualistic tantric
tradition of the Śaiva Siddhanta, ritual alone (namely, the soul- cleansing,
burc19032_1st_i-412.indb 37
—-1
—0
—+1
12/12/18 6:14 PM
38 9 Part i: FroM MEdiEval tantra to Early ModErn Bhakti
-1—
0—
+1—
karma-wiping rite of initiation) is the means to liberation, whereas in the
nondual tantric traditions of Kashmir, in addition to ritual it is knowledge
( jñāna)— of one’s identity with the Divine—that has primary soteriological
value. In both cases, as a natural result of tantric soteriology, bhakti becomes
marginal and subordinate.
At the same time that we note the “official” marginalization and subordination of bhakti in the (pre-twelfth- century) texts of these major tantric
Śaiva traditions, it is important to remember that realities on the ground
were undoubtedly different from what Sanskrit treatises on theology and
yoga reflect, especially as tantric traditions opened up to admit a wider social
world.41 As I discuss in the following pages, mainstream tantric Śaivism grew
up in dependence upon and interaction with preexisting traditions of templebased lay Śaiva devotion, and the early medieval religious landscape saw regular
exchange between initiated tantric adepts and lay devotees. If tantric texts
tend to present an idealized picture of tantric religiosity oriented to full tantric initiates and thus give a particular impression of bhakti (in which bhakti is
marginalized and subordinated), communities of tantric initiates were nevertheless not far removed from a different world of bhakti, a lay community of
bhakta-jana whose religious lives centered on temple worship as well as the giving of material support to their fellow bhaktas and to the professional ascetics—
yogīs and ācāryas—who served them as teachers, objects of devotion, emblems
of spiritual authority, and key service providers.
My focus in this book is on North India, but it is instructive to briefly consider the ways in which the “emotional devotion” of early medieval South India
was inflected by and linked to the developing tantric tradition. Drawing on
Friedhelm Hardy’s classic, path-breaking work,42 Radha Champakalakshmi
argues that the concept of bhakti as an emotional, intimate personal loving relationship with the Divine was initially developed by the Tamil Vaiṣṇava Ālvārs
and Śaiva Nāyanārs, whose poems drew on the love theme of Cankam poetry as
they sought to carry purāṇic forms (dominated by northern, Sanskritic elements) “to the Tamil masses in their own idiom, namely an ‘intensely human
religious awareness,’ and in the vernacular, namely Tamil.” 43 These early medieval Tamil bhakti traditions emphasized ritual worship and the temple as the
house of God, features she says were “closely related to the teaching and ethos
of the Āgama and Tantra.” 44
Bhakti in medieval South India was, in fact, “forged in dialogue with Śaiva
Tantrism,” as Karen Pechilis has stated. Pechilis discusses how the Tamil bhakti
saint Kāraikkāl Ammaiyār (ca. 550) intertwines bhakti and tantra in her poetry
and describes her “nearest of kin” as the Śaiva tāntrikas.45 Relatedly, Indira
burc19032_1st_i-412.indb 38
12/12/18 6:14 PM
thE tantric agE = 39
Peterson’s research on early medieval Tamil Nadu has shown Śaiva bhakti’s
close links with temple life and ritual worship as laid out in the tantric scriptures known as the Āgamas.46 While the early bhakti hymns of the Tēvāram
emerged outside the tantric tradition, in them the Śaiva bhakti saints known
as the Nāyanārs regularly praise the Āgamas and refer their devotion to Śiva to
his wisdom and grace as manifested in these tantric scriptures. For example,
Appar sings, “My tongue will continue to utter the Āgamas in the presence of
its companion (the mind)” (129.1); Campantar states, “They are praising the
Lord of Tiruvorriyūr who is the wealth of the Āgamas” (3.57.10); and Cuntarar
says, “Indeed he is the mother, giving grace to one who preserves the wisdom
of the Āgamas” (7.96.6).47 By the end of the twelfth century, the Āgamas would
prescribe nearly all aspects of ritual religious life in South Indian Śaiva temples, a sign of tantra’s full integration with public devotional religion.
The devotion of the Ālvārs, the early medieval Tamil Vaiṣṇava bhakti saints,
also seems to have been closely linked to tantric scripture and ritual culture.
Dennis Hudson has argued that the tantric liturgy and yogic-meditational practices of the Pāñcarātra Āgama underlie the poems of the Ālvārs.48 In one essay, he
discusses how the ninth-century Ālvār bhakti saint Āṇṭāḷ “performed Tantric
rites of the Bhāgavatas and described them in her poems.” 49 Similarly, the poetic
corpus of Nammālvār (late eighth to early ninth century), most important of all
the Ālvārs, includes many verses about yogic meditation and ritualized tantric
visualization. In a poem from the Tiruvāymoli (I.9), for instance, Nammālvār
seems to describe a tantric laya-yoga visualization meditation, praising the Lord
in a series of passages in which he describes Viṣṇu in ascending locations within
his body (loosely corresponding with the cakras of the yogic subtle body): in his
lap, within his heart, upon his shoulders, on his tongue, in his eye, on his brow,
and finally at the crown of his head.50 With poems such as this in mind, Hudson
describes Vaiṣṇava bhakti in medieval South India as, in general, a “disciplined
devotion according to Bhagavata Tantra”—that is, a surrendering to (taking refuge in) God combined “with the ceremonial activities of a Tantrika liturgical
discipline.”51
By the tenth century tantric monastic orders had thoroughly integrated
themselves into India’s booming temple culture, and devotional life thus had
become “tantra inflected” in many respects. Throughout medieval India, brahmanical, temple- and monastery-based forms of tantra “blended easily with”
bhakti religiosity, “such that the two became indistinguishable.”52 Popular medieval devotion had considerably different emphases than the religious life of most
tantric initiates, but by the twelfth century (if not well earlier), bhakti seems
to have generally occurred within the frame of—or in necessary interaction
burc19032_1st_i-412.indb 39
—-1
—0
—+1
12/12/18 6:14 PM
40 9 Part i: FroM MEdiEval tantra to Early ModErn Bhakti
with—fundamentally tantric principles, institutions, and ritual prescriptions.
The forms of bhakti that arose in North India beginning in the fifteenth century would have a considerably different relationship with tantra than this.
The Tantric Age
-1—
0—
+1—
As Alexis Sanderson has demonstrated, Śaivism was unquestionably the most
successful and influential religious tradition of early medieval South Asia. 53
While Buddhism and Vaiṣṇavism had flourished in the preceding centuries,
between the fifth and seventh centuries, Śaivism gradually emerged as the
dominant religion of South and Southeast Asia.54 Sanderson terms this period
(ca. sixth to thirteenth century) “the Śaiva Age,” showing how Śaivism rose
to preeminence in early medieval South Asia and beyond as the principal beneficiary of royal patronage, a fact demonstrated “by the epigraphical record
of pious donations, by the preponderance of Śaiva temples at this time, and by
abundant evidence that Śaivism’s Vaiṣṇava, Buddhist, and Jain competitors
developed systems of ritual observance during this period” closely paralleling
that of tantric Śaivas.55 While Sanderson’s work forcefully demonstrates the
dominant presence and impact of Śaivism—especially tantric (or mantra-mārga)
Śaivism—in the early medieval period, it is clear that Buddhism and Vaiṣṇavism
each also had a large and influential presence in several areas of the subcontinent during this time, particularly in their tantric (Vajrayāna and Pāñcarātra)
forms. Indeed, most of the religious communities of early medieval India—
whether Śaiva, Buddhist, Vaiṣṇava, or Jain—came to share a parallel repertoire
of tantric rituals for initiation, installation (pratiṣṭhā), and regular worship
while also sharing patronage relationships in which virtually the same powers and protections were offered to the same royal clients. 56 As Christian
Wedemeyer has stated, “those communities centered around various Śivas,
Viṣṇus, Buddhas, and (Jaina) Tīrthaṅkaras in the late first millennium (A.D.)
participated mutually in a pan-Indian religious culture, most of whose structuring assumptions were the same and in which a variety of ritual forms were
shared and developed across traditions.”57 This pan-Indian culture was in significant part a tantric one, and it was as much “political” as it was “religious.”
The first major tantric system to emerge in South Asia seems to have been
the Śaiva Siddhānta, a school that was well established in the subcontinent
by the seventh century and by the tenth century had become a tradition “of panIndian scope enjoying close ties with the political order and often exercising
decisive control over the principal religious and social institutions of the time.”58
burc19032_1st_i-412.indb 40
12/12/18 6:14 PM
thE tantric agE = 41
A number of Vajrayāna Buddhist traditions, with a repertoire of rites modeled
on tantric Śaiva ritual forms and methods, developed at this time as well and
received widespread royal patronage and competed with tantric Śaiva traditions, particularly in eastern India. The Vaiṣṇava counterpart of the Śaiva
Siddhānta, the orthodox tantric tradition of the Pāñcarātra, was also an
important presence in this period that, like Saiddhāntika Śaivism, became integral in the operations of royal power and public, temple-based tantric devotional religiosity.59 While these tantric traditions, in which transgressive ritual
played little to no role, were predominant religious communities of early medieval India, constituting a sort of tantric mainstream, scholars such as Shaman
Hatley and Dominic Goodall have pointed out that “much of the scholarly literature has assumed an artificial distinction that, at times, goes so far as to
exclude the Vaiṣṇava Pāñcarātra and/or Śaivasiddhānta from the category of
‘Tantra,’ ” when, in fact, they are at the very center of the tantric tradition.60
While scholarship has tended to emphasize the esoteric and transgressive
dimensions of the tantric tradition, here I want to focus attention on the “mainstream tantra” of India’s Tantric Age—i.e., the tantric tradition as it manifested
itself pervasively in medieval Indian social life, both in elite political and popular quotidian spheres. Drawing on Alexis Sanderson’s work, among others in
the rapidly advancing fields of tantric studies and Śaiva studies,61 I argue that
we should conceive tantra (at least by the tenth century) as a broad, pervasive
mainstream tradition—only marginally concerned with transgression—whose
institutions, cosmological assumptions, and ritual forms were key elements in
the social life and religiopolitical structure of early medieval India. A significant amount of tantric literature does, of course, discuss radically transgressive rites,62 including the frequenting of cremation grounds, ritual sex with
“untouchable” women, and the consumption of shit, piss, and sexual fluids.63 It
would be absurd to sweep this corpus of tantric discourse under the rug as
unimportant. Yet the crucial fact is that the texts concerned with these radically transgressive rites were directed toward a small minority of advanced tantric practitioners. The mistake we must avoid is taking the history and practice
of a minority stream of professional tantric ascetics and sādhakas as the history
and practice of tantric religiosity as a whole. While the dedicated tantric adepts
who performed transgressive sexual and mortuary rites had a significance well
beyond their small numbers, their story is but one piece of a tantric tradition
that includes, more importantly, the preeminent religious communities of
the medieval period—Śaiva Siddhānta, Vaiṣṇava Pañcarātra, and Buddhist
Vajrayāna—and their pervasive institutional networks (which played a key
role in the larger sociopolitical order), as well as a widespread community of
tantric practitioners of healing, exorcism, and “practical magic” (i.e., various
burc19032_1st_i-412.indb 41
—-1
—0
—+1
12/12/18 6:14 PM
42 9 Part i: FroM MEdiEval tantra to Early ModErn Bhakti
practices and rites intended to effect this-worldly health, wealth, protection,
or harming).64
Observing the historical development of the Hindu tantric tradition, we find
on the one hand a set of more mainstream and orthoprax schools (e.g., Śaiva
Siddhānta, Vaiṣṇava Pañcarātra) that generally avoided transgressive practices
and operated within standard brahmanical purity codes and, on the other hand,
a diverse array of heteroprax Śaiva and Śākta communities that gave a more
central place to goddesses, transgressive behaviors, and impure substances
in the ritual life of their sādhaka initiates. Unlike the Śaiva Siddhānta tradition,
these “non-Saiddhāntika” Śaiva tantric cults—based in the Bhairava Tantras of
the Mantrapīṭha and Vidyāpīṭha—were generally not involved in mainstream
public and lay devotional (temple) religiosity but were oriented toward individuals performing rituals in the private domain for their own or their clients’
benefit.65 While these non-Saiddhāntika tantric cults gave more importance to
antinomian mortuary and sexual rites involving polluting substances (alcohol,
blood, sexual fluids), they seem to have grown out of the Śaiva Siddhānta tradition, accepting its legitimacy and assuming its ritual paradigms even as they
claimed to offer more powerful methods to reach the common goals of liberation and siddhis.66
If the misplaced scholarly tendency to see tantra as synonymous with transgression (especially antinomian sexual and mortuary rituals) has been one
factor making it difficult to conceive tantra as mainstream, there are other factors as well. In particular, one might reasonably ask how a fundamentally esoteric, initiatory tradition—one that originated to serve the interests of only the
most dedicated individual seekers of mokṣa and siddhis— can be considered a
mainstream, popular tradition integral to the larger religiopolitical order and
pervasive in South Asian social life more broadly. In the following pages, I aim
to explain just this.
Kings and Gurus, Temples and Monasteries
-1—
0—
+1—
With the collapse of the Gupta dynasty in the mid-sixth century and the end of
the relative stability it had provided, India saw the emergence of multiple competing regional centers, which led to a culture of militarism and frequent warfare. As Ronald Davidson has explained, in this environment, warlords, “seeking
legitimacy and identity, began to increase their patronage of literature and to
strategize their support for religion, searching for religious counselors that
could bolster their political and military agendas.” 67 Beginning in the seventh
burc19032_1st_i-412.indb 42
12/12/18 6:14 PM
thE tantric agE = 43
century, these aspiring rulers turned increasingly to the emerging tantric tradition and its rituals of empowerment. We have inscriptional evidence of at least
three major kings taking Saiddhāntika Śaiva tantric initiation in the second half
of the seventh century, and “during its first half the Buddhist philosopher
Dharmakīrti (c. 600–660) goes to the trouble of attacking the Tantric practice
of initiation as a means to liberation.” 68 Between the seventh and twelfth centuries, tantra rose to prominence in South Asia as Hindu rulers increasingly
embraced relationships with tantric gurus and their communities. 69 If rulers
in early medieval India were turning more and more to tāntrikas, these tāntrikas
were also turning to them, adapting to meet their needs. Within the tantric tradition, a new cadre of religious specialists developed who could ritually consecrate power-seeking warlords with tantric mantras, “transforming them into
divine kings and their conquered territories into equally consecrated maṇḍalas
of royal power.”70 In order to extend their influence, tantric communities devised
a new class of initiate, exemplified by the king, who, because of his demanding
social duties, was given a special form of initiation, nirbījā dīkṣā (initiation without seed), which exonerated him from the time-consuming program of daily
rituals required of most full initiates (putrakas and sādhakas) while still ensuring liberation.71 Tantric gurus claimed that their initiation and consecration
(abhiśeka) rites endowed kings with a power beyond that of their rivals, intensifying their brilliance, ensuring their victory against enemies, and allowing them
to have long and distinguished reigns. These tantric dīkṣās and abhiśekas not only
infused the king with a deity’s immense power but also offered access to a wealth
of potent tantric mantras that could be performed on demand by tantric adepts
to protect and benefit the realm, promote a royal patron’s success, and frustrate
his enemies.72 At the same time, tantric traditions largely sought to accommodate and embed themselves within the orthodox brahmanical tradition that had
sanctified and legitimated royal power in India for centuries. As Sanderson has
shown, tantric traditions flourished, in significant part, by co-opting brahmanism, taking over many of the positions, functions, and ritual services that had
previously been exclusive to orthodox brahmans.73
In return for the empowerment and legitimation that tantric initiation provided them, newly made kings patronized their gurus’ sectarian communities,
building and sponsoring monasteries, or maṭhas, for tantric monastic orders and
promoting their interests throughout the kingdom. India’s early medieval period
is well known by scholars as “the great era of Hindu temple building”74 and the
time when “the temple became the dominant religious institution of South
Asia,”75 but the key role of monasteries is not as widely understood. While
temples—big and small, urban and rural—served as centers of religious community and devotion, symbols of royal authority, and key motors of economic
burc19032_1st_i-412.indb 43
—-1
—0
—+1
12/12/18 6:14 PM
44 9 Part i: FroM MEdiEval tantra to Early ModErn Bhakti
-1—
0—
+1—
growth, agricultural development, political integration, and brahmanical cultural expansion,76 in many ways maṭhas became just as important in the early
medieval religious and sociopolitical world. Tamara Sears has shown that “monasteries often not only preceded but also served as a stimulus for the development of larger temple towns.”77 Maṭhas were frequently built close to temples,
and the members of these monastic communities played a key role in supporting ritual activity within the temple and facilitating the social and economic
services offered at temple centers. Maṭhas were established in both the peripheries and hubs of the kingdom, together making up a network of interlinked
monastic centers that effectively tied disparate localized communities into the
kingdom’s central administrative and social framework.78 These maṭhas served
as seats of brahmanical learning, dissemination centers of sectarian theology
and philosophy, and sites supporting intensive, unhindered practice of meditation and yoga, sometimes through the establishment of hermitages for pilgrims or those seeking a base for isolated ascetic practices.79 By the ninth and
tenth centuries, many organized monastic orders—most of them tantric and/or
Śaiva—began to function also “as landed, self-sustaining administrative institutions responsible for the collection and redistribution of taxes and agricultural
revenue.”80 Some maṭhas even maintained armaments and served as sites for
training and garrisoning military forces.81
Like temples, mathas were also crucial institutions in lay devotional life. Gurus
and professional ascetics in monasteries “served as key agents in the growth of
wide-scale devotional activity, both through their role as temple priests and
attendants as well as in their function as foci for ritual in their own right.”82 Initiated members of tantric communities were often expected to worship not only
God but also their initiating guru (who was seen as not simply a respected teacher
but also a revered manifestation of divinity and a vehicle for liberation), and it
was at the maṭha that tantric gurus could personally receive such homage.83 As
mentioned, medieval kings sought out these tantric gurus for the initiation and
empowerment they could provide and, in exchange, rewarded them with maṭhas
and land grants. In this fashion, some tantric gurus accumulated enough wealth
and land that they were able to use their resources independently to establish
new maṭhas, and even “to behave like royal patrons themselves, not only founding new monasteries but also bestowing land-grants on Brahmins, rewarding
poets, founding temples and new settlements, and providing the means of irrigation.”84 Certain tantric gurus, particularly of the Śaiva Siddhānta tradition, actually “came to exercise a transregional authority whose geographical extent could
be greater than that of any contemporary king.”85
Gurus and kings, temples and maṭhas, these were the key institutional figures and spaces that made possible the spread and sustenance of tantric ritual
burc19032_1st_i-412.indb 44
12/12/18 6:14 PM
thE tantric agE = 45
and worldview in early medieval India. Beginning in the seventh century, rulers increasingly looked to tantric rituals and gurus for empowerment and legitimation and sponsored the institutional growth of tantric communities. By the
tenth century, temples (administered by tantric monastic orders) and maṭhas
(housing professional tantric ascetics) had become vital economic, political, and
religious hubs in the institutional network of medieval society.86 Of course, not
all the temples and monasteries patronized by the kings and local communities of medieval India were tantric, but a great many were. These sites embodied, expressed, and widely disseminated tantric ideology and ritual; they were
the institutions upon which mainstream tantra—as a key player in the medieval religiopolitical order—depended. This is important because in later centuries, when those institutions became threatened, damaged, or destroyed,
the stage would be set for a major transformation in India’s religiopolitical
landscape.
In seeking to understand the significance of tantric Śaivism, and of tantra
more broadly, in the early medieval period, we must keep in mind the presence
and influence of popular lay traditions of devotion. While Sanderson’s work has
centered on tantric Śaiva traditions, he makes the crucial observation that these
traditions “were successful in no small measure because Śaiva devotion had
become the dominant religious idiom in the population at large.”87 The rich and
powerful of early medieval India were increasingly aligning themselves with
tantric Śaiva initiatory lineages, in significant part because doing so was “particularly efficacious in the eyes of a predominantly Śaiva population, not only
among the brahmins but among all social strata, down to and including the lowest.”88 In other words, it seems that tantric Śaivism achieved its great success
largely because it “hooked onto” and was “parasitic” upon a preexisting, templebased tradition of lay Śaiva bhakti, a tradition I now turn to.89
The Bhakti of the Śivadharma: Lay Śaiva Religion in
Early Medieval India
The traditions of tantra and bhakti grew up alongside and in dialogue with each
other in the early medieval period. The massive success of tantric Śaivism, specifically, was dependent upon the vitality of a coexisting tradition of lay Śaiva
devotion. A brief investigation of this tradition through its literature—the
Śivadharma corpus90—is crucial to establishing the relationships between professional asceticism, popular devotion, and political economy in medieval India,
and more specifically to understanding the relationship of bhakti, tantra, and
burc19032_1st_i-412.indb 45
—-1
—0
—+1
12/12/18 6:14 PM
46 9 Part i: FroM MEdiEval tantra to Early ModErn Bhakti
-1—
0—
+1—
yoga during this period. The earliest texts of the lay Saiva tradition seem to
be the circa sixth-to-seventh- century Śivadharma (Śivadharmaśāstra) and
Śivadharmottara, likely composed in North India but widely known throughout
medieval India.91 These scriptures claim descent from Śiva and present themselves as an “easy, affordable set of teachings and rites that would allow common
people to fulfill all their wishes,” in contrast to expensive and ultimately unprofitable Vedic rites.92 The lay Śaiva religion prescribed in these texts was open to
śūdras, untouchables (cāṇḍāla), and foreigners (mleccha). The Śivadharma states,
“The one who knows the four Vedas is not dear to me [Śiva]. Even a dog cooker
who is my devotee, one may give to him or take from him. And he is to be worshipped just like I am (to be worshipped).”93
In key respects, this lay Śaiva religion centered on bhakti. As Śivadharma 1.29
states, “The essence of the Śivadharma is Śiva-bhakti.” The question is, What does
bhakti mean here? In a fascinating passage, somewhat incongruous with the
text’s overall representation of devotion, the Śivadharma describes bhakti as having eight limbs, which Śiva characterizes as (1) affection (vātsalya) for my
devotees (mad-bhaktajana); (2) taking pleasure in (seeing) my worship (pūjā)
(performed by others); (3) worshipping (abhyarcana) me oneself with bhakti;
(4) exerting one’s body (with bhakti) for me (i.e., performing physical activity
[labor] for my sake [mamārthe cāṅgaceṣṭanam]); (5) listening to my stories;
(6) trembling (vikriya) of one’s voice, eyes, and limbs (on hearing such stories);
(7) constantly remembering (anusmaraṇa) me, (8) always depending upon (living
for) (upajīvati) me. The passage concludes, “In whomever this eightfold bhakti
grows, even if a mleccha [foreigner], he is a chief of Brahmans, a glorious sage
[muni], an ascetic renouncer [yati], and a learned man [paṇḍit].”94 While many
scholars associate the sort of bhakti described here with later Vaiṣṇava traditions, here we find it expressed in a Śaiva scripture composed in Sanskrit
around the sixth century.95 Of particular note here are the Śivadharma’s
(a) stress on listening to the stories of God; (b) valuing of the embodied, affective dimensions of bhakti religiosity (its “modulations” of voice, eyes, and limbs);
and (c) emphasis on caring for—and celebrating the virtues of—one’s fellow devotees (bhaktas), a community of devotion explicitly embracing all caste classes
and even those outside the varṇa system.
The Śivadharma’s stress on listening to stories of God is paralleled in the
purāṇic literature that was proliferating at this same time and that, of course,
was filled with such stories. Well known for their sectarian character, the
Purāṇas seem to have been composed by and for the purposes of “the new sectarian theistic movements that were emergent in the early centuries CE: early
Śaivas and Vaiṣṇavas such as the Pāśupatas, Bhāgavatas and Pāñcarātras, operating on the frontiers of brahmanical orthopraxis” who, Travis Smith explains,
burc19032_1st_i-412.indb 46
12/12/18 6:14 PM
thE tantric agE = 47
were “combin[ing] Vedic orthodoxy with a new temple-based devotionalism
directed toward an expanded pantheon of deities (including most prominently
Viṣṇu and Śiva).”96 It is quite likely the same group—an educated lay Pāśupata
community—that authored both the Śivadharma texts and (many of) the Śaiva
Purāṇas. The broader lay population would have heard and absorbed the
teachings in these scriptures in particular ways. A cult of the book—as part of
“the gift of knowledge” (vidyādāna)—was important in this period and involved
the ritualized sponsorship (by kings or other laymen), worship, and public recitation of manuscripts containing sectarian religious teachings.97 In this context, listening (śravaṇa), devotion (bhakti), and scriptural knowledge (vidyā) were
understood to depend on one another and to work together in dynamic relationship.98 Since the Śivadharma texts were composed for the laity, they were “generally written in undemanding Sanskrit that could be expected to be readily
understood by a larger public”; nevertheless, there was clearly a concern that
teachings and stories in Sanskrit would not be fully comprehended by the lay
populace. With this in mind, the Śivadharmottara recommends “that it be taught
to its audiences in the languages of their regions.”99 Thus, when the Śivadharma
lists “listening to [Śiva’s] stories” as an essential feature of bhakti, we can imagine
a world of oral vernacular translations and retellings of sectarian religious stories and teachings in Sanskrit texts that—however little information we have to
describe it—must have been a vital dimension of the social and religious landscape of early medieval South Asia.
The Śivadharma’s description of an embodied, emotional bhakti—the trembling (vikriya) of one’s voice, eyes, and limbs (under the influence of bhakti)—is
especially striking, seeing as numerous scholars have argued that this sort of
embodied, passionate, ecstatic devotion is first expressed in the Bhāgavata
Purāṇa, a text that was by most estimates composed roughly five hundred years
after the Śivadharma. Among others, J. N. Farquhar, Jan Gonda, S. N. Dasgupta,
Paul Hacker, and, perhaps most influentially, Friedhelm Hardy have asserted
“that the passionate and ecstatic bhakti expressed in the Bhāgavata Purāṇa constitutes a distinctive new form of devotion that is markedly different from the
more intellectual and contemplative forms of bhakti that find expression in different ways in the Bhagavad- Gītā, the Viṣṇu Purāṇa, and Rāmānuja’s teachings.”100 Barbara Holdrege adds that what is perhaps most new and distinctive
in the Bhāgavata Purāṇa’s representation of bhakti is “its embodied nature,” with
“the bhakta’s internal ecstatic state . . . often described as manifesting through
the external body, overflowing into the senses and limbs and erupting in spontaneous bodily manifestations such as the bristling of body hair, stammering
speech, weeping, laughing, singing, and dancing.”101 We should not view a single verse in the Śivadharma describing and valuing “modulations of voice, eyes,
burc19032_1st_i-412.indb 47
—-1
—0
—+1
12/12/18 6:14 PM
48 9 Part i: FroM MEdiEval tantra to Early ModErn Bhakti
-1—
0—
+1—
and limbs [under the influence of bhakti/stories of God]” as somehow equivalent in significance to the multitude of verses in the Bhāgavata Purāṇa that
praise and describe in detail an embodied, emotional bhakti; nevertheless, the
Śivadharma verse once again points toward the existence of a world of bhakti—of
vernacular storytelling and affective, embodied devotional experience—that
many scholars would not have imagined for this period of Indian history, or
in this Śaiva context. At the very least, the Śivadharma should cause us to be far
less confident in our current scholarly assumptions about bhakti, specifically
regarding the historical origins of embodied, ecstatic, emotional expressions
of bhakti. We might see here the danger of seeking out or claiming origin points
and moments of genuine novelty, particularly when the available sources give
us such a limited view of on-the-ground historical realities. While further
research on the Śivadharma corpus will almost certainly not uncover any such
true origins, it very well may force a fundamental reassessment of our understanding of the historical phenomenon of bhakti.
Also of special note in the Śivadharma’s passage on eightfold devotion is the
very first component in Śiva’s description of bhakti: “affection [vātsalya] for my
devotees [mad-bhaktajana].” This verse clearly conceives a distinctive form of
community united by its devotion to God—“Śiva’s bhaktajana”102—a community
including śūdras, women, and even mlecchas.103 As the text states, “In whomever
this eightfold bhakti grows, even if a mleccha, he is a chief of Brahmans, a glorious sage, an ascetic renouncer, and a learned man.” Timothy Lubin has shown
how the Śivadharma redefines varṇa and āśrama categories in such a fashion that
hereditary status “is subordinated to a ritually mediated spiritual kinship,”
thereby enabling women and śūdras to “partake of the ritual entitlements otherwise identified with Brahmanical status.”104 Lubin is right to conceive this
Śaiva community in terms of “kinship,” for a familial form of care and affection for one’s fellow bhaktas seems to be central to the Śivadharma’s understanding of bhakti. This becomes more tangible in chapter 11 of the text, which
praises giving (material resources) and rendering service to fellow Śaiva bhaktas when they are tired, ill, or otherwise in need.105 The word I have translated
as “affection” is the Sanskrit vātsalya, a word used especially to refer to a mother’s selfless love, tenderness, and care for her children. Thus, here we get the
sense that the Śivadharma is envisioning (and perhaps even reflecting back the
existence of) a community whose members attend to and care for one another
in the intimate and tender manner of a family.
This brings us back to one of the Śivadharma’s key opening remarks (1:29):
“The essence of the Śiva-dharma is Śiva-bhakti.” It is worth reflecting on the very
fact that in this verse dharma and bhakti are made virtually equivalent. What
does this imply? Alf Hiltebeitel’s research on the Sanskrit epics offers insights
burc19032_1st_i-412.indb 48
12/12/18 6:14 PM
thE tantric agE = 49
here. Hiltebeitel points out that in the period of the epics—not long before the
Śivadharma’s composition—“bhakti is to be mapped with dharma.”106 He shows
that at the very heart of both dharma and bhakti in the Mahābhārata and Rāmāyaṇa
are the values of hospitality (ātithyam) and friendship (sakhya), both of which
are crucial in the construction of community.107 Hiltebeitel’s comments on the
epics, in combination with what we have seen in the Śivadharma, shed light on
an understanding—and a lived practice—of bhakti anchored in significant part
in a constellation of terms (e.g., vātsalya, ātithyam, sakhya) that all have to do with
social ethics and the building and “upholding” of community. Moreover, just as
important in this early medieval conception of bhakti as familial affection, hospitality, and friendship was the discourse and practice of the gift (dāna), whose
supreme recipient, notably, is the Śaiva yogī.
The Śivadharma’s eightfold bhakti thus offers some enticing suggestions about
a world of lay Śaiva devotion whose existence in this period—in its communal
listening (to stories and teachings), embodied emotionality, and casteless ethic
of care for fellow Śaiva bhaktas—many scholars would not have imagined. Nevertheless, as the research of Florinda De Simini demonstrates, a closer study of
the Śivadharma and Śivadharmottara illustrates that the devotional religiosity
extolled in the Śivadharma tradition centers not on the cultivation of emotion
but on faith in the spiritual authority of Śaiva scriptures and professional ascetics, and on practices of ritual worship and gift giving (dāna)—in particular,
offerings of material support to the community of initiated Śaiva ācāryas and
yogīs.108
In the Śivadharmottara, probably composed in the seventh century (not long
after the Śivadharma), bhakti is still important but seems clearly overshadowed
by the term śraddhā, “faith” or “trustworthiness.” Indeed, śraddhā, a word with
considerably less “emotional” and “participatory” connotations than bhakti, is
conceived as “constituting the essence of all Śaiva teachings and the only means
through which Śiva can truly be attained.”109 At the same time, the text claims
that the power of these Śaiva teachings is embodied in the six-syllable sectarian mantra oṃ namaḥ śivāya. Both the Śivadharma and Śivadharmottara extol the
great benefits of uttering this mantra, with Śivadharmottara 1.38–39 stating, “One
in whose heart this mantra ‘oṃ namaḥ śivāya’ constantly dwells, he has learned
[all] the knowledge that has been taught, and performed all [rituals],”110 a sentiment not too far removed from the Bhāgavata Purāṇa’s (and many other bhakti
texts’) later stress on the power of reciting the name of God. Overall, the most
common topics in these two earliest Śivadharma texts seem to be (a) instructions for and praise of the ritual worship of the liṅga (a sphere of devotional practice that would be adopted and adapted as the core of the tantric Śaiva ritual
repertoire); (b) praise of (and merits accrued by) constructing and maintaining
burc19032_1st_i-412.indb 49
—-1
—0
—+1
12/12/18 6:14 PM
50 9 Part i: FroM MEdiEval tantra to Early ModErn Bhakti
a Śaiva temple; and (c) rules, fruits, and proper recipients of dāna.111 Thus, the
window that Śivadharma texts give us onto early medieval lay Śaivism suggests
that, in fact, bhakti was understood most centrally as performing rituals of worship to Śiva (in the aniconic form of the liṅga) and as offering material support
to (patronizing) the larger Śaiva community (especially its professional ascetics). These devotional practices were believed to generate sufficient merit to
bring the Śiva bhakta (and their family members) success in this life, a long
afterlife in heaven (śivalokaḥ), and a desirable rebirth in which they might then
be able to attain final liberation.112 If the ritual worship of localized, material
forms of God in temples, combined with gift giving, constituted the heart of the
practice of this early medieval lay devotional tradition, its soteriology and spiritual ideals were focused squarely on ( jñāna- and dhyāna-) yoga and the yogī. In
fact, lay Śaivas’ relationships with communities of professional Śaiva ascetics—
namely, Pāśupatas and (later) tantric Śaiva yogīs—seem to have been a key
piece of their devotional lives.
Yogīs and Devotees in the Early Medieval
Religiopolitical Economy
-1—
0—
+1—
All indications are that the texts of the Śivadharma scriptural corpus were
probably composed by a lay segment of the Pāśupata Śaiva community.113 The
Pāśupatas, whose cult centered on Śiva as Lord (pati) of Beasts (paśu), emerged
in the second century CE114 as a “Hindu” response to Buddhist and Jain monastic traditions, an ascetic order—exclusive to brahmans—that proselytized lowcaste and tribal populations while maintaining and propagating brahmanical
values. Most scholars have narrowly identified the Pāśupatas with the system
of lifelong renunciation and rigorous asceticism outlined in Kauṇḍinya’s fourthcentury commentary on the Pāśupata Sūtra; however, as Peter Bisschop has
demonstrated, celibate ascetics were in fact only one strand of a far broader tradition that developed to include a lay community of Śiva bhaktas (Māheśvaras)
faithful to the Pāśupata ācāryas and teachings.115 The Pāśupatas were the first
“Hindu” (non-śramaṇa) ascetic group in South Asia to emulate and compete with
the Buddhists’ zealous proselytization of marginalized Indian and non-Indian
peoples.116 Indeed, as Hans Bakker explains, “the Pāśupatas had had a good look
at their Buddhist counterparts and had copied their formula for success, namely
a standing organisation of professional religious specialists—yogins, ascetics,
and ācāryas—supported by a following of ordinary devotees, the Māheśvara
community at large, to whose spiritual needs it catered.”117
burc19032_1st_i-412.indb 50
12/12/18 6:14 PM
thE tantric agE = 51
The Pāśupatas arose in southern Gujarat as a Śaiva initiatory order of brahman ascetics. Over time they spread across India, integrating themselves into
the temple-based lay Śaiva devotional tradition118 and thereby successfully
bringing together the realms of ascetic-yogic Śaiva practice and popular-level
Śiva bhakti.119 As Travis Smith explains, the Pāśupatas “were able to establish a
vast lineage-based Śaiva network, loosely organized yet ideologically coherent.
In each region, new monasteries were built and local ascetic and lay communities were gradually assimilated into the overarching Pāśupata worldview, spread
through texts of the order and through the circulation of teaching lineages.”120
The available evidence suggests that lay Śaivas who were members of, or associated with, a larger (loose and weakly bounded) Pāśupata community probably
composed not only the Śivadharma corpus but also the Īśvara Gītā121 and many of
the Śaiva Purāṇas, including the sixth-to-seventh-century Skanda Purāṇa. In significant part through these texts, they were able to skillfully accommodate and
incorporate diverse “peoples and practices on the margins of elite society into
the fold of Śaiva orthopraxis.”122 Clearly then, far from an antinomian ascetic
order on the fringes of society, from the sixth to the tenth century, the Pāśupatas
were an important part of the Indian religious mainstream, receiving considerable patronage from kings as well as local collectives of merchants, traders, and
artisans,123 and operating monasteries and temples across India, but especially in
Gujarat, Maharashtra, Madhya Pradesh, Rajasthan, and the Gangetic valley.124
For the lay population of Śiva bhaktas, the professional Pāśupata ascetic—the
yogī—was a crucial figure, a locus of spiritual authority and an object of devotion. As mentioned, Śivadharma texts extol the immeasurable fruits gained by
gift giving, but both the Śivadharma and Śivadharmottara state repeatedly and
emphatically that the best recipients of a gift “are those who are identified with
Śiva and whose cult is thus equivalent to his own, that is the Śaiva yogins
(śivayogin).”125 Indeed, according to Śivadharma 12.35–38, at the very top of the
hierarchy of beings stands the yogī—namely, the Śivayogin.126 The text (12.31)
defines the Śaiva yogī as “one who is endowed with knowledge and freedom from
desires thanks to a mind pacified by Śiva” and who is “committed to the practice of the sixfold yoga.”127 It is clear in the Śivadharma tradition that, whatever
the merits of (the ritual worship and gift giving that primarily constitute) Śiva
bhakti, it is especially dispassion, knowledge ( jñāna), and yoga that lead one to
liberation.128 As a passage from Śivadharma 10 states, “Detachment [vairgāya]
comes from indifference to worldly things; the arising of knowledge [ jñāna]
comes from detachment [vairāgya]; Yoga proceeds from knowledge [ jñāna]; and
one obtains the end of suffering because of Yoga.”129
In early medieval India, not only the Śaiva lay community but also the
Vaiṣṇava and Buddhist lay communities saw the professional ascetic—the yogī
burc19032_1st_i-412.indb 51
—-1
—0
—+1
12/12/18 6:14 PM
52 9 Part i: FroM MEdiEval tantra to Early ModErn Bhakti
-1—
0—
+1—
(saṃnyāsi, bhikṣu)—as the human religious exemplar. Notably, Śiva, Kṛṣṇa, and
the Buddha (and bodhisattvas) all are regularly presented as masters of yoga
(i.e., models of yogic dispassion, knowledge, and power/attainment) in the
first-millennium literature of their respective devotional communities.130 It is
unclear whether the power, perfection, and charisma ascribed to Śiva, Kṛṣṇa,
and the Buddha were theorized “in terms of powers already attributed to yogis”
or whether the abilities and qualities credited to yogīs were modeled after the
attributes of these divine figures.131 In either case, in the lay communities of all
these traditions, the yogī emerges as an authoritative spiritual model worthy of
respect, awe, and devotion. In the context of lay Śaiva religion, to give to yogīs
(or ācāryas) was to give directly to Śiva and thus to receive the corresponding
fruits/merit. All of this suggests an early medieval economy of spiritual and political power driven in large part by exchanges between lay devotees and initiated professional ascetics. Yogīs mediated spiritual knowledge and power to
laypeople (while also providing a number of ritual and practical services) and
gave sacred authority and political capital to kings and other political elites,
while kings and other lay devotees provided for the yogīs’ material support. This
same sort of religiopolitical economy would continue as a new form of initiatory religion burst onto the scene—tantra.
The success of Śaiva tantra (and thus tantra in general) was dependent upon
the existing tradition of lay Śaiva devotion but also on the crucial role that communities of initiated Pāśupata ascetics had established for themselves in early
medieval India’s religiopolitical economy, its particular system of exchanges in
spiritual, political, moral, and economic capital. In a number of key respects,
the activities of the Pāśupatas seem to have laid an ideological and institutional
foundation for the rise of the first major tantric community, the Śaiva Siddhānta.
In the ninth and tenth centuries, lineages of Siddhānta Śaivism “rather swiftly
replaced the Pāśupatas as the most extensive and influential transregional Śaiva
sect,” on the one hand by enticing patrons and followers with their novel and
uniquely efficacious tantric ritual technology and the promise of liberating initiation for all caste classes and, on the other hand, by appropriating and building further upon the Śaiva networks and infrastructure established by the
Pāśupatas.132 As Smith explains, the Śaiva Siddhānta community “assumed control over the administration of particular maṭhas and temples that were built
and maintained by Pāśupatas”133 while also following the Pāśupatas’ lead in disseminating their ritual and doctrine through the popular textual medium of the
Purāṇas. Evidence from Madhya Pradesh, Rajasthan, Gujarat, Kashmir, Maharashtra, Orissa, Andhra Pradesh, and Tamil Nadu attests to the fact that by
900, “a broadly extended network of interrelated Śaiva Siddhānta lineages [had]
spread itself out over much of India, acting frequently as spiritual preceptors to
burc19032_1st_i-412.indb 52
12/12/18 6:14 PM
thE tantric agE = 53
kings, constructing and presiding over temples and monasteries, and propagating the teachings of the āgamas.”134
The members of Śaiva Siddhānta monastic orders took on the same roles as
the Pāśupata professional ascetics had, serving as objects of devotion (and
recipients of dāna), spiritual teachers, and service providers who also administered the temples and performed the temple worship rites of popular devotional
religion. By offering initiation to householders and members of all castes, Śaiva
tantric orders worked to tie the lay population even more closely to their communities of professional ascetics than had the Pāśupatas. This brings us to
another important piece of tantra’s medieval success story, one oriented toward
the tradition’s broad social reach. In order to explain tantra’s pervasive presence in the social life of both elite and nonelite populations, I turn to the manner in which the tantric tradition participated in the larger medieval process
of agrarian and political expansion, incorporating cultural and religious
phenomena on the frontiers into brahmanical culture.
The Social Dynamics of Tantra
Tantra has regularly been described as a tradition centered on tribal, indigenous, pre-Aryan religious practices and folk traditions of shamanistic possession, healing, magic, and the worship of goddesses, nature spirits, and spirits of
the dead. At the same time, many scholars have stressed the brahmanical, conservative, elitist, esoteric, and scriptural aspects of tantra as its defining sociological features. Travis Smith has insightfully remarked that “while the debate
between positing either an elite or a non-elite origin of Tantra seems at first
glance irresolvable, it may itself hold the key to the problem of identifying the
fundamental characterization of Tantra.”135 Along these lines, rather than highlight either the “elite origins” or “nonelite origins” scholarly position, here I
aim to demonstrate how tantra was simultaneously both brahmanical and folk,
the product of a dynamic and mutually transforming encounter between a transregional Sanskritic, dharmic culture and a variety of unbrahmanized local
Indian subcultures peripheral to state structure and settled agriculture.
B. D. Chattopadhyaya has argued that one of the core elements in the transformation of early medieval India was “the appearance of state society in areas
that had long been peripheral.”136 Similarly, Sanderson has explained that “the
territorial expansion of brahmanical society into new regions . . . was one of the
salient features of the early medieval period,” and one that was inseparable from
the growth of tantra.137 During India’s early medieval period, forest regions of
burc19032_1st_i-412.indb 53
—-1
—0
—+1
12/12/18 6:14 PM
54 9 Part i: FroM MEdiEval tantra to Early ModErn Bhakti
-1—
0—
+1—
hunters and rudimentary agriculturalists on the peripheries of state power and
Sanskritic culture were increasingly transformed into well-irrigated, settled
agricultural regions under royal power.138 It is clear that a major factor in the
growth of both tantric and purāṇic traditions at this time was the extension of
brahmanical culture139 into rural, tribal, local areas alongside the expansion of
state institutions and the growth of settled agriculture. Early medieval political
integration and agrarian expansion were, then, inseparable from a religious
assimilation—seen in both the Purāṇas and the Tantras—in which local cults,
rituals, and deities were subsumed into more cosmopolitan, supralocal Sanskritic traditions. When a king expanded his realm into tribal or peripheral
areas, to successfully integrate the people in the area into the kingdom, it was
necessary to integrate local deities— often goddesses—who were the primary
objects of worship, into the more cosmopolitan brahmanical religious order.
Alternatively, when a local chief began to expand his power over a larger territory, in order to assume authority he would adapt his original religion, which
may have been centered on the local goddess, to conform to the larger religious
order.140 Clearly, a key factor in the success of tantric Śaivism was its ability to
effectively incorporate the worship of the goddesses who were so popular among
many South Asian people.141 Indeed, tantric Śaivism rose in conjunction with
Śāktism, for many Śaivas were, in practice, primarily worshippers of the Goddess (theologically conceived as Śiva’s inherent power, or śakti), often in the form
of various tutelary goddesses (kuladevīs) of land and clan.142
The role of the Purāṇas in the early medieval integration of new populations
(through the assimilation of local deities, mythologies, and ritual practices) into
a reenvisioned, transregional, brahmanical tradition is well known,143 but far
less attention has been drawn to the important overlap between tantric and
purāṇic traditions in this process. In his study of the Bengal Purāṇas, Kunal
Chakrabarti remarks that “large chunks” of the Tantras’ ritual forms and
practices—e.g., tantric mantras and nyāsa procedures—were incorporated into
the ritual prescriptions of the Purāṇas.144 Sanderson similarly points out “the
inconstant character of the boundary between [initiatory tantric] traditions and
Purāṇic forms of religious observance,”145 noting that, in fact, “a substantial
amount of Saiddhāntika [Śaiva tantric] ritual material has been propagated
within Purāṇas.”146 Nirajan Kafle gives clear evidence that the Śivadharmasaṅgraha,
a lay Śivadharma text composed in the ninth to tenth century, borrows heavily
from both tantric and purānic sources.147 While much about early medieval religiosity is still quite murky, it seems that, on the ground, the realms of the “tantric” and “purāṇic” often blurred into one another, especially in lay religion.
Drawing on Donald Davis’s pathbreaking historical research on Indian
law, Jason Schwartz has argued persuasively that the growth of the tantric
burc19032_1st_i-412.indb 54
12/12/18 6:14 PM
thE tantric agE = 55
traditions must also be understood in connection with medieval legal practices.148 Schwartz argues that the principle of legal pluralism in medieval
India—e.g., the king’s protection of the right of different autonomous corporate
groups to manage their own affairs in accordance with community-specific
values, codes, and customs—was, in fact, the necessary precondition for “the
defining feature of the medieval Indic religious landscape: namely the mass
institutionalization of Tantric communities openly recognized and patronized
by the state.”149 As he insightfully suggests, to become a tantric initiate was
not simply to gain access to the benefits of specific tantric religious rituals and
teachings but also to take on a new legal status, one in which orthodox brahmanical social codes often were no longer applicable. Samayin (entry-level) initiates became subject to the distinctive, community-specific laws (samaya) of
the tantric community they were initiated into (e.g., nonrecognition of caste
differences, the giving of a proportion of one’s material wealth for use by the
community), with the guru—over and above the state—now their final binding
legal authority.150
As noted, tantric traditions changed the rules of the game by opening up initiation (and its liberating benefits) to all caste classes and not requiring renunciation. Tantric communities likely grew especially by offering initiation to the
increasing numbers of śūdras—recruited from local populations—who were
essential in the cultivation of land in newly settled peripheral areas.151 While it
seems safe to assume that the lay populace would have significantly outnumbered the community of tantric initiates, Nina Mirnig’s research has suggested
that, in at least some cases, the modes of worship prescribed by the Śaiva
Siddhānta tradition for entry-level (samayin) tantric initiates were the same as
those prescribed for lay devotees.152 Furthermore, epigraphic evidence suggests
that tantric initiation (especially among heads of household) was considerably
more common than many have assumed and was even conducted in semiregular, prearranged mass ceremonies. Sanderson discusses an inscription from
Senakapāṭ in Chattisgarh that shows that, already in the seventh century, maṭhas
were receiving endowments to perform tantric initiation in regular ceremonies
on predetermined days (indicating a steady stream of people who would present themselves for initiation) and that it was typical to initiate multiple people
in a single ceremony.153 According to the Senakapāṭ inscription, doctrinal teachings were to be expounded on these occasions, implying that the new and previous initiates gathered at these ritual events were not professional ascetics but
householders living lives dominated by worldly concerns.154 Considering the
unique spiritual (and even legal) benefits to which it provided access, the tantric initiation of householders would have generated a strong sense of debt and
commitment to the guru and larger tantric community and called for a gift or
burc19032_1st_i-412.indb 55
—-1
—0
—+1
12/12/18 6:14 PM
56 9 Part i: FroM MEdiEval tantra to Early ModErn Bhakti
-1—
0—
+1—
investment of some kind in return. As Indrani Chatterjee reminds us, “All initiates paid for their learning and assimilation in some form,” even if “the quality, size and nature of these payments separated the humbler initiates from their
wealthier counterparts.”155 Initiation fees and the contributions of initiated
householder patrons (whether of wealth, labor, sustenance, or land)—which
were probably required by the community-specific laws (samaya) one became
subject to upon initiation—provided a crucial economic foundation for tantric
communities. Much further research is needed to better illuminate the historical relationships between professional tantric ascetics, initiated householders
(samayins and putrakas), and uninitiated lay devotees, but what we do know suggests that the religious lives of tantric initiates and lay devotees were on something of a continuum and, at a general level, that initiatory tantric traditions
and lay devotional traditions interacted with and influenced each other in vital
ways.
As brahmanical culture spread outward in connection with political and
agrarian expansion, in the overlapping tantric, purāṇic, and Śivadharma traditions there was a reenvisioning of orthopraxy to accommodate new social
groups and practices. In this process, a number of originally nonbrahmanical,
“folk” religious practices—possession, exorcism, nature spirit and goddess
worship, etc.—became important parts of the tantric tradition by being
incorporated into tantric ritual forms and authorized in and by Sanskrit tantric scriptures. In this sense, we might understand much of tantra as brahmanical appropriations, translations, or domestications of folk religious practices and traditions. As Travis Smith states, “While many of the elements that
make up Tantra were associated with non- elite, marginal peoples and sects,”
tantric practices were not “accurate representations of folk or non- elite
praxes, nor were they intended as such.” Rather, these elements “were bounded
and transformed through a remarkably rigorous ritualization, leading to a
profuse elaboration of precise rules for ritual conduct. . . . Tantric practices, it
is clear, were not the actual practices of these groups, but deliberate transformations and elaborations of them.”156
In all likelihood, many tantric scriptures were composed in areas into which
brahmanical state society had only recently entered, and some may not have
been composed by brahmans at all, but in order for the texts to carry cosmopolitan authority they had to be in the brahmanical language of Sanskrit. As
Flood remarks, many of the authors and redactors of tantric scriptures “were
not completely at home” in an elite Sanskritic milieu but still “thought it imperative to locate these texts and traditions within the wider, ‘high’ literary culture
of the Sanskrit cosmopolis.”157 The textualization—in the language of Sanskrit—of what were in some cases originally vernacular, oral, nonbrahmanical
burc19032_1st_i-412.indb 56
12/12/18 6:14 PM
thE tantric agE = 57
traditions of knowledge and practice is one illustration of how the growth of
tantra—whatever its folk or tribal dimensions—generally occurred in the hegemonic terms of brahmanism.
Frederick Smith’s discussion of the relationship between folk traditions of
possession and the core tantric ritual procedure of nyāsa (the tantric method for
divinizing the body) offers further insights that may help to explain how folk
traditions of practice were domesticated into more controlled, brahmanical
modes as they were assimilated into the tantric tradition. As has been noted, the
divinization of the body (through ritual imposition of mantras and visualizationbased yoga) is a distinguishing feature of tantric practice. In tantric ritual, the
self becomes divine, as Flood puts it, through the ritual “mapping of the body in
tradition-specific and text-specific ways.”158 In other words, the tantric divinization of the body is not separate from its systematic entextualization. This paradigmatic tantric self-deification typically occurs through the practice of nyāsa,
the imposing of mantras on the body, with the practitioner touching the
requisite part of the body and reciting the correct mantra. As Smith writes, “The
intent of nyāsa is to impose or place the power of the mantras, and perforce
the deities, and so on, which they inscribe, on or within various body parts,
either one’s own or that of an image of a deity.”159 The divinization of the body is
also a fundamental part of the folk practice of possession, but unlike nyāsa, it
occurs in a manner that is nontextual, far less systematic, and far more spontaneous (though still typically following certain ritual protocols). Smith argues
that the Tantras domesticated possession ritually and conferred philosophical
credibility on it, “apparently sensitive to its historical prominence as a popular
and legitimate mode of religious experience and expression.”160 He states, “Nyāsa,
we can say, is brahmanical possession.”161 In the process of nyāsa, “deities, powers, and so on are invited to take possession of the body. But they are invited in a
brahmanically programmatic, that is, ‘textual,’ way, one that emphasizes purity
at the expense of spontaneity and danger.”162 Thus Smith provocatively suggests
that, in nyāsa, we have a tantric brahmanical domestication of popular religious
possession. From this perspective, nyāsa, one of tantra’s most distinctive practices, served to exercise “programmatic control” on the practice of possession
while also conferring brahmanical legitimacy upon it.
Thus, the process by which certain folk practices became “tantric” (i.e.,
became part of the tantric tradition) was inseparable from the process by which
these practices became “brahmanical” and “Sanskritic.” A more specific example of the brahmanical nature of tantric practice and the tantric domestication of folk possession can be found in the practice of the tantric healer known
as the gāruḍika described in the body of scriptures known as the Gāruḍa Tantras. Michael Slouber has conducted trailblazing research on the Śaiva Gāruḍa
burc19032_1st_i-412.indb 57
—-1
—0
—+1
12/12/18 6:14 PM
58 9 Part i: FroM MEdiEval tantra to Early ModErn Bhakti
-1—
0—
+1—
Tantras, an important but understudied corpus of tantric scriptures focused
especially on healing snakebite and curing poisoning, while also including
broader material related to health care, astrology, possession, and sorcery.163 As
a class of scripture, these texts were known as early as the sixth century, and
by the tenth century the twenty-eight Gāruḍa Tantras had been canonized as
one the five “streams” of Śaiva revelation.164 The teachings of Gāruḍika medicine articulated therein were, like all the mantras and ritual methods in
tantric scripture, considered to be divine in origin, with a chain of command
flowing from “Śiva to Garuḍa [Lord of Birds and archenemy of snakes] to the
Gāruḍika practitioner who embodies him.”165
The ritual divinization of the body characteristic of tantric practice—
described by Smith as “brahmanical possession”—is a crucial element in the
gāruḍika’s healing ritual. The Gāruḍika practitioner ritually identifies with
(becomes possessed by) Garuḍa in order to heal. Slouber explains that “ ‘becoming’ Garuḍa was the fundamental act of the ritual, judging by its frequent mention in the literature.”166 This procedure involved the “mental construction of
the element maṇḍalas and their deposition on both the hand and body of the
practitioner” and “the visualization of oneself as Garuḍa.”167 In other words,
what likely was originally a nontextual folk tradition of practice, in which the
healer became possessed by a locally specific deity, has here become a tantric
practice of controlled and programmatic identification with a Sanskritic deity
(agency and control lying not with the possessing deity/spirit but with the practitioner), authorized by and performed in accord with the prescriptions of Sanskrit tantric scripture.168
The tantric healer of snakebites is a figure that will be encountered again in
early modern bhakti and Sufi literature and is worth a bit more attention here,
particularly insofar as his ritual activities speak to the wide range of functions
to which relatively standardized tantric ritual techniques could be applied.
Gāruḍikas drew on the power of various (textually specified) mantras to heal
snakebite, drive out snakes, and cure poisoning, but in addition to these functions, they made amulets and yantras (from mantra syllables) that could be worn
on the body for protection from snakes, dangerous animals, or thieves, to ward
off fear in dangerous places, to make barren women fertile, or to defend against
possession by hostile spirits.169 Passages from the Jayadrathayāmala suggest that
gāruḍikas were also involved in weather magic and crop protection, conducting
tantric rituals and reciting mantras to ward off thunderstorms, lightning, and
pests.170
The ritual practice of the gāruḍika gives us a sense of the more pragmatic
dimensions of tantric religiosity and the ways in which tantra could be simultaneously esoteric and popular. As Lawrence A. Babb has discussed, in Indian
burc19032_1st_i-412.indb 58
12/12/18 6:14 PM
thE tantric agE = 59
religious life we often find, on the one hand, a “transcendental complex” concerned with the ultimate purposes of man, long-term welfare, institutional continuity, and the stability of society, which is usually in the hands of hereditary
brahman priests, and, on the other hand, a “pragmatic complex” concerned
with local and personal welfare and desires, which is usually in the hands of
typically nonhereditary and lower-caste exorcists and shamans.171 The tantric
ritual practice of the gāruḍika would seem to fall under the pragmatic complex,
but what is far more important to highlight is that the basic ritual forms and
techniques a gāruḍika would use to harness divine power for “pragmatic” functions were no different from those used to access and employ sacred power for
“transcendental” purposes. Both “high” and “low,” elite and popular forms of
tantric practice sought to harness divine power through essentially the same
mantra-focused ritual techniques. As a modus operandi, tantric ritual bridged
the transcendental and pragmatic aspects of South Asian religious life in its ability to access and utilize cosmic power. Whether healing snakebite victims or
consecrating kings, exorcising illness- causing demons or installing temple
images, practicing black magic or striving for liberated consciousness, worshipping bloodthirsty goddesses or pure, brahmanical Sanskritic deities, the same
general ritual system and form—with self-divinization and the use of tantric
mantras—was utilized.
Furthermore, this tantric ritual form was used across sectarian lines. While
Gavin Flood has shown us that the tantric body was “mapped” in sect- and textspecific ways, tantric ritual forms were incredibly similar regardless of the
specific sect or text (and its cosmology and doctrine) informing the ritual perfomance.172 The Vajrayāna Buddhist monk and the Pāñcarātra worshipper, the
Saiddhāntika Śaiva priest and the transgressive Śākta ascetic, the gāruḍika
healer and the tantric rājaguru—all seem to have moved and understood their
bodies in remarkably similar ways in their ritual practice. Indeed, Sanderson
has described the ritual of the various tantric sects—Buddhist, Śaiva, and
Vaiṣṇava—as essentially different “dialects of a single ‘Tantric’ language.”173
The range of uses to which tantra’s transsectarian ritual techniques could
be applied was remarkably wide; however, knowledge of the specialized ritual
techniques themselves was of limited access and very time- consuming to
acquire. In the case of the gāruḍika, for instance, “patients would have known
little about the operations of the specialist except that they were renowned as
highly effective” and that “the power of the mantra to heal the envenomation
stems from Garuḍa.”174 To use White’s apt metaphor, tantric scriptures were
“classified documents” and their uniquely efficacious mantras and ritual techniques were “secret codes” given only to those with “top-secret security clearance.”175 Though specialist tantric ritual knowledge was esoteric—guarded by a
burc19032_1st_i-412.indb 59
—-1
—0
—+1
12/12/18 6:14 PM
60 9 Part i: FroM MEdiEval tantra to Early ModErn Bhakti
system of layered initiations—participation in tantric religious life became a
popular phenomenon in early medieval India, a period that saw pervasive faith
in the power of tantric mantras and rituals to accomplish one’s desires, widespread access to and demand for initiated tantric specialists, and a public life
in which tantric figures and institutions often played a key role.
In sum, in conjunction with the expansion of state society and settled agriculture in early medieval India, a brahmanical tantric civilization extended into
unbrahmanized tribal areas, beginning a multidirectional interaction in which
both the tribal areas and brahmanical culture were transformed. The development of the tantric tradition occurred in this context and, in key respects, was
characterized by the assimilation of local, nontextual, and nonbrahmanical traditions of practice (e.g., possession, exorcism, blood sacrifice) into the controlled,
transregional, brahmanic, textualized ritual forms of tantra. These tantric ritual
techniques served to harness cosmic power for purposes both spiritual and mundane, transcendental and pragmatic, with the same basic ritual technique used
by tantric practitioners, be they Śaiva, Vaiṣṇava, or Buddhist, rural or urban,
high caste or low caste, orthodox or heterodox. The specialized, guarded knowledge and training required to effectively utilize tantric mantras and perform
tantric rituals made tantra, at one level, essentially esoteric, but participation in
the tantric ritual world, interaction with tantric specialists and institutions, and
acceptance of tantric cosmological presuppositions and ritual logic were all so
widespread that, in toto, tantra must be considered a popular tradition.
The End of the Tantric Age
-1—
0—
+1—
The preceding discussion has shown how tantra became a popular, politically
powerful, mainstream tradition. It did so by allying with kings, by co-opting
orthodox brahmanism even as it assimilated frontier folk traditions, by disseminating highly esteemed (flexible but standardized) ritual forms and techniques,
and by integrating itself with temple-based lay devotional traditions. By the
tenth century, the institutions and ritual networks of tantric monastic orders
had achieved a huge and influential presence throughout India. The continued
existence of these institutions and the tantric communities associated with
them depended largely upon royal patronage and the existence of a religiopolitical environment in which transregional tantric ritual forms and Sanskritic aesthetic expressions were assumed as the shared mediators of sovereign power.
Beginning especially with the Delhi Sultanate period (1206–1526), this would no
longer be the case in much of northern and central India. The political order and
burc19032_1st_i-412.indb 60
12/12/18 6:14 PM
thE tantric agE = 61
cultural world of India—particularly North India—underwent a major change
at this time with the rise of new Persianate forms of state power, in which institutional tantric religion had no real place or authority.
It would undoubtedly be a mistake to reduce the decline of mainstream, institutional tantra in late medieval India solely to the impact of Persianized Turks in
India, though this was clearly a decisive factor in North India. Might the rising
influence of Vedānta (especially in South India) also have played a role in the
“decline” of politically influential, institutionalized forms of tantra? Further
research is needed to determine if and how these two historical trends were
related, but it is a question well worth asking. Beginning in the twelfth century,
there was a proliferation of different Vedāntic philosophical systems in southern
India, often integrated with new sectarian bhakti theologies.176 While mainstream, institutional tantra appears to have disappeared rather suddenly in
North India, in South India the influence of public, royally sponsored tantra
seems to have waned more gradually at the same time that the ideological and
institutional presence of Vedānta—most especially Advaita (nondual) Vedānta—
was continually expanding. Over the course of the late medieval and early
modern periods, the growing Vedāntic tradition seems to have assimilated or
significantly impacted a variety of devotional and tantric-yogic modes of religiosity.177 With some Indian social elites perceiving the traditional brahmanical
social order to be under threat (whether directly or indirectly) from Persianate
Turks, perhaps Vedānta offered something the tantric tradition could not: a
scripturally based philosophical axis around which (and a shared conceptual
language through which) a unifying, orthodox “Hindu” intellectual identity
could arise and cohere.178 As Elaine Fisher has shown, in late medieval and early
modern South India, Vaiṣṇava and Śaiva traditions ceased to operate as independent religious systems (each with their own separate, sect-specific ultimate
scriptural authority)—as they had in the Tantric Age—and instead increasingly
sought to position themselves, through Vedāntic exegesis, as ideal representatives of Vedic orthodoxy.179 The expanding influence of the Vedāntic tradition
had its own internal logic and should not be viewed simply as “a direct response
to Islam”;180 yet brahmanical perceptions of and concerns regarding the presence and expanding political influence of Central Asian “mlecchas” in India likely
gave real impetus to developments already under way in the rising Vedāntic
tradition, which in turn may have contributed to shifts in the position of institutional tantra.181
Whatever role Vedānta (and its rise) may or may not have had in the process,
with the spread of Turkish power across North India beginning in the twelfth
century, we see the rapid decline of the relationship between ruler and tāntrika—
political authority no longer relying upon tantric ritual and institutions—which
burc19032_1st_i-412.indb 61
—-1
—0
—+1
12/12/18 6:14 PM
62 9 Part i: FroM MEdiEval tantra to Early ModErn Bhakti
brings about a significant change in the religious landscape. As White writes,
“The rise and fall of Hindu Tantra as a religious ‘mainstream’ is directly linked to
the rise and fall of its royal patrons. In north and central India, Hindu Tantra
thrived as the royal cultus under the Kalacuri, Somavamshi, Chandella, Calukya,
and other dynastic lines, until their lands fell into the hands of Muslim rulers in
the 12th century.”182 While identifying these Turkish rulers first and foremost as
“Muslim” is problematic (essentializing them in terms of religious identity when
Indians at the time did not understand them in this light), White nevertheless
hits on a crucial point. As he states, “For so long as [the] relationship between
kings and tantric specialists remained in force, Tantra persisted as a sanctioned
religious force in India, with the ceremonial life of the kingdom being conducted
in a tantric mode. When that relationship was dissolved, as Hindu kings were
overthrown or reduced to vassal status by Muslim [Persianate Turkish] rulers (or,
from the 16th cent. forward, increasingly opted for a devotional religious style),
Tantra disappeared.”183 Here White does not mean that all tantric religiosity vanished but that tantra as a mainstream, public tradition—one held together as
such by royal patronage and participation in an overarching politico-religious
ideology and institutional culture—largely disappeared.
Tantric religiosity in North India would live on, but it would do so primarily
(a) in more private, secretive contexts among Hindu royal families now subordinate to Turkish rulers and (b) among lineages of tantric practitioners that were
less dependent on the royally patronized institutional structure of temples and
maṭhas. Tantric ritual forms would maintain a key place in North Indian Hindu
religious life, but with the end of the Tantric Age these ritual procedures and
techniques were often detached from tantric religiosity per se, persisting in new
contexts (e.g., Vedāntic, bhakti, and yogic frameworks) not linked to sectarian
tantric traditions or institutions. In the new Persianate political culture of Sultanate India, the infrastructure of mainstream, public, institutional tantra
largely collapsed, and the sphere of tantric religion underwent a transformation and contraction into one constituted largely by less-institutionalized lineages of yogīs, warrior ascetics, and rural tantric healers and “magicians.”
R
-1—
0—
+1—
In this chapter, I have articulated an understanding of tantra as a broad religiopolitical tradition whose ritual culture, institutional figures and spaces, and
cosmological presuppositions were pervasive in early medieval India. The rest
of this book seeks, in large part, to explain the rise of bhakti in early modern
North India and the concomitant marginalization of Śaiva-Śākta tantra, a phenomenon that cannot be understood apart from two interrelated factors: (a) the
burc19032_1st_i-412.indb 62
12/12/18 6:14 PM
thE tantric agE = 63
spread and influence of Persianate political culture and popular Sufism in the
Sultanate and Mughal periods and (b) bhakti authors’ and performers’ critique
and attempted delegitimation of the tantric outlook. The Tantric Age comes to an
end with the Delhi Sultanate, but tantric ritual technique would persist in and
beyond the thirteenth, fourteenth, and fifteenth centuries, sometimes in new
Sufi and bhakti contexts. The tantric outlook would also persevere, with tantric
yogīs now as its primary representative. In the early modern period, however,
tantric paradigms of thought and behavior—no longer having the sociopolitical context and state support that had sustained them—would often find
themselves marginalized and subordinated to Hindu religious practices and
perspectives that were more congenial to the new social environment and its
increasingly prevalent Islamicate worldviews. In the next chapter, I explore
this process, looking at the ways in which Sultanate political culture and the
spread of popular Sufism paved the way for the emergence of the great bhakti
poet-saints of North India.
—-1
—0
—+1
burc19032_1st_i-412.indb 63
12/12/18 6:14 PM
1. thE tantric agE = 321
72.
73.
74.
75.
76.
77.
78.
79.
80.
81.
82.
83.
84.
85.
words on a page but digested in the embodied terms of sonic vibrations and animal affects.
To privilege text (and its discursive content) above performed song (and its affective impact)
is to privilege sight over sound in a fashion characteristic of Western modernity and to
diminish and obscure the affective power of the aural.
Meyer 2009, 9.
Schaefer 2015, 78–79.
Warner 2002, 81.
Hirschkind 2005, 41.
Hirschkind 2005, 40. While I speak in general terms here, of course there would have been
differing registers of participation and resonance among members of the bhakti public,
therefore we must avoid conceiving bhaktas as unitary or homogeneous in the way they
responded to or participated in performances of bhakti songs and stories.
Hirschkind 2009, 2.
This formulation owes much to Joel Lee’s wonderful work on the sensuousness of caste; see
Lee 2015, 42–80.
Lofton 2011, 16.
Wedemeyer 2013, 6.
Wedemeyer 2013, 38–42.
Doniger 2011, 166– 67.
Keune 2015, 71.
Jonathan Smith 1982, 35.
Gottschalk 2012, 338.
1. The Tantric Age
1. Benoytosh Bhattacharyya seems to have coined the term “Tantric Age” in An Introduction to
Buddhist Esoterism (1932) to describe South Asia’s early medieval period. More recently,
Christian Wedemeyer has taken up the term in Making Sense of Tantric Buddhism (2013) to
describe the late-first-millennium period of Indian religions. Gavin Flood (2006) has similarly spoken of a “tantric civilization” that flourished “during the medieval period before
the rise of the hegemony of the Delhi Sultanate” (71). This period also generally corresponds with what Alexis Sanderson (2009) has called the “Śaiva Age,” and my characterization of it as the “Tantric Age” relies heavily on the evidence he has marshaled.
2. Gavin Flood 2006, 73.
3. Sanderson 1988, 661, 663. While the tantric scriptures are most often termed tantras,
āgamas, and saṃhitās, they also include texts called sūtras, yāmalas, nigamas, and siddhāntas,
among other names ( jñānas, vidyās, ḍāmaras, āmnāyas, arṇavas, rahaysa, etc.).
4. Goodall 1998, xxxvi–xxxix; Hatley 2007a, 7n20. A general tendency emerged for nonSaiddhāntika (Śaiva-Śākta) tantric texts to be termed tantras, Vaiṣṇava (Pāñcarātra) tantric
texts to be termed saṃhitās, and orthodox Śaiva (Siddhānta) tantric texts to be termed
āgamas, but this was not a hard-and-fast rule. Śaiva Siddhānta scriptures were designated
as both āgamas and tantras (and occasionally as saṃhitās), and Pāñcarātra scriptures were
termed tantras in addition to their more common designation as saṃhitās.
5. For more on this important text, including a translation of its first three (of five) sūtras, see
Goodall 2015.
burc19032_1st_i-412.indb 321
—-1
—0
—+1
12/12/18 6:14 PM
322 9 1. thE tantric agE
-1—
0—
+1—
6. Sanderson 1988, 660.
7. Nevertheless, the tantric tradition depends heavily upon the Vedic ritual framework and is
in many respects continuous with the earlier Vedic tradition (and, just as much, the Atharvedic tradition); Goodall 1996, xxxi–xxxii.
8. See Sanderson 2012–2013, 12–13. He notes that women were usually made “passive beneficiaries of initiation rather than . . . active initiates with access to office.”
9. Fisher 2017, 36.
10. Fisher 2017, 36.
11. The samayin receives samayadīkṣā, which allows him or her to take on the community’s
identifying marks and to follow its basic rules of conduct. This is usually followed by the
special initiation (viśeṣa-dīkṣā) in which the disciple completes his ritual rebirth as a samayin and thus becomes eligible to worship Śiva on his own behalf and to study the Śaiva tantric literature. One becomes a putraka after undergoing the quintessential tantric initiation
rite of nirvāṇadīkṣā, open only to those who had proven their spiritual “ripeness” to the
guru and which ensured liberation at death. In the nirvāṇadīkṣā rite of Śaiva Siddhānta tradition, the guru—via the power of tantric mantras—ritually transforms himself into Śiva,
enters the disciple’s body, extracts his soul, homologizing it with the various core elements
and levels of the cosmos and (in a process involving ritual gestures, detailed visualizations,
and the uttering of appropriate mantras) gradually purifying it of all previous karma and
impurities—revealing its full divinity—before reinstalling it in the initiate, thereby virtually ensuring his liberation at death. Sādhakas receive the sādhakābhiṣeka, allowing them to
seek out siddhis and bhukti, while the ācārya goes through a special consecration, the
ācāryābhiṣeka. See Brunner 1994, 431–32.
12. Sanderson 2006b, 3; Sanderson 1988, 660.
13. Sanderson 2006b, 2–3.
14. Gavin Flood 2006, 11. While scholars have often identified the notion that “to worship a god
one must become a god” as distinctively tantric, in fact, as Walter Kaelber (1989, 51–52) has
shown, earlier Vedic ritual necessitated the divinization of the body as well (though not
through the imposition of mantras).
15. Fisher 2017, 37.
16. Hatley 2007a, 8.
17. As Annette Wilke explains, “What is most characteristic of Tantric mantras is the idea that
they incorporate—like a ‘seed’—in their mere sound pattern the respective god or goddess
(or another numinous force) in a very real sense” (2014, 135). Technically, mantras refer to
the sonic forms of male deities, while vidyās refer to sonic forms of female deities.
18. Śakti is originally a theological concept developed in tantric Śaivism, thus it can be problematic to project it onto other tantric traditions that used different terminology with different
metaphysical connotations. Nevertheless, śakti-like conceptions of the sacred as tremendous
cosmic power pervaded all tantric traditions. In the Buddhist context, for instance, Glen Wallis has argued that the principle underlying the practice of esoteric Buddhists in early medieval India “was the assumption of the Buddha’s power in the world, and that the design of the
practice was to enable the practitioner to mediate that power and manipulate it toward particular ends” (2002, 1).
19. Padoux 2017, 16.
20. Brooks 1992, xix.
21. In this sense, David Gordon White is quite right to have stressed the role of demonology in
tantric systems (2012b, 164– 66) and to have described tantra as, in large part, an array of
burc19032_1st_i-412.indb 322
12/12/18 6:14 PM
1. thE tantric agE = 323
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
ritual techniques for the control of various powerful spirit beings, “both for one’s own benefit and as tools to use against others” (2003, 13).
Relatedly, some aspects of tantric ritual seem to have developed out of post-Vedic, pretantric śānti (pacification/appeasement) rituals in which the non-Vedic astrological tradition
( jyotiḥśāsta) appropriated late-Vedic Atharvan ritual methods for protecting rulers (and
others) from omens and malevolent astrological forces (e.g., adbhuta, utpāta, bhaya, duḥsvapna).
If tantric initiation ceremonies were modeled on royal consecration ceremonies such as the
rājyābhiṣeka, as argued by Ronald Davidson (2002), Marko Geslani (2012) has shown that
these royal coronation rites had a core apotropaic function that has not been adequately
emphasized, making use of non-Vedic mantras and invoking the power of a wide array of
cosmic deities to protect the king (and his subjects) from misfortune and inauspicious
omens. For an argument (drawing on Geslani’s work) that the esoteric Buddhist maṇḍala
initiation ceremony modeled itself on the paradigm of these post-Vedic śānti rituals, see
Shinohara 2014, 64–90.
See White 2009, 196.
Rastelli 2000, 320–22.
Rastelli 2000, 340.
Rastelli 2000, 340–43, 354–55. The sādhaka who has mastered his mūla-mantra is said to have
also acquired the ability to successfully perform any of the infamous sinister six rites, or
ṣaṭkarman. The sādhaka achieves any of these various goals by ritually actualizing the mantra and its power through recitation ( japa) of the mantra, supported by visualization meditation (dhyāna) and imposition (nyāsa) of the mantra upon the body or a physical object
(e.g., an amulet) (349–50).
White 2011, 574.
Padoux 2017, 132.
Sanderson 1995, 22.
Ferrario 2015, 16–20.
Ferrario 2015, 20–24. See Gupta 1986; Oberhammer 2007, 37–54; Czerniak-Drożdżowicz 2003.
Gupta 1986.
Ferrario 2015, 28–29.
Dominic Goodall, email message to author, August 29, 2017.
Ferrario 2015, 30.
Ferrario 2015, 31.
Ferrario 2015, 24, 28–29, 59.
On stotra literature and its relevance to the study of bhakti, see Stainton 2019.
Ferrario 2015, 31–32.
The Mataṅgapārameśvara tells the story of a sage named Mataṅga who is meditating intently
on Śiva when he becomes distracted by the sweet sound of wind passing through stalks of
bamboo, loses his focus, snatches a stalk of bamboo, and fashions it into a flute, then fervently plays the flute with “supreme bhakti,” such that Śiva himself comes down and
appears before Mataṅga, who is then further possessed by the fervor of bhakti toward Śiva.
For a discussion of this text and how tantric Śaiva brahmans in the ninth century interpreted it in such a way as to minimize or eliminate the value of bhakti while reiterating the
power and salvific efficacy of tantric ritual, see Schwartz 2012a, 216–225. Schwartz offers an
insightful analysis of the Kashmiri Siddhāntin Bhaṭṭa Rāmakaṇṭha’s commentary on this
text and its subordination of bhakti to ritual; however, it seems to me that he errs in suggesting that the devotional voice of the lay Śivadharma tradition is present in the early
burc19032_1st_i-412.indb 323
—-1
—0
—+1
12/12/18 6:14 PM
324 9 1. thE tantric agE
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
-1—
0—
+1—
tantric tradition, since all our evidence suggests that bhakti had a very different, and less
important, place in early tantric Śaivism than it did in the Śivadharma. See Ferrario’s (2015,
50–58) detailed analysis and critique of Schwartz’s position.
Dominic Goodall, email message to author, August 29, 2017.
Hardy 1983.
Champakalakshmi (1996) 2004, 50.
Champakalakshmi (1996) 2004, 64.
Pechilis 2016. Pechilis is specifically interested in the way in which Kāraikkāl Ammaiyār
emphasizes the potency of the cremation ground, drawing on certain prototantric
(Atimārga) and early tantric (Mantramārga) practices and conceptions but reframing them
in the terms of a bhakti disposition averse to tantra’s elaborate ritual procedures.
Peterson 1989, 12.
Translated and discussed in Judith Martin 1983, 114–15.
Hudson 2008, 10.
Hudson 2000, 206.
I am indebted to Archana Venkatesan (email message to author, November 15, 2017) for
making me aware of, and sending me a rough translation of, this passage.
Hudson 2008, 25–26.
White 2011, 575.
Sanderson 2009.
Sanderson 2009, 58– 61.
Sanderson 2006b, 4.
Sanderson 2006b, 6.
Wedemeyer 2013, 31.
Davis 1991, 4.
An early form of Pañcarātra Vaiṣṇavism likely existed before tantric Śaivism, but this
early Pañcarātra did not have a tantric ritual system and later came to adopt the popular
ritual system developed within tantric Śaivism (i.e., the Śaiva Mantramārga); Sanderson
2009, 58–70.
The mistaken idea that the Śaiva Siddhānta and Pāñcarātra are marginal to the study of
Tantra proper is linked, in part, to the artificial distinction some scholars have made
between texts that designate themselves as Tantras, Āgamas, and Saṃḥitās. These three
designations were, in fact, synonymous and interchangeable terms for tantric scriptural
revelation, with both Pāñcarātra and Śaiva Siddhānta scriptures at times referring to
themselves as Tantras (in addition to Āgamas or Saṃhitās). See Goodall 1998, xxxvi–xxxix;
Hatley 2007a, 7n20.
Thanks to the research of Alexis Sanderson, along with scholars such as Dominic Goodall,
Harunaga Isaacson, Mark Dyczkowski, Diwakar Acharya, Shaman Hatley, and Somdev
Vasudeva, among many other tantric studies experts, as well as ongoing (mostly textcritical) studies of the lay Śaiva traditions of first-millennium India (by scholars such as
Peter Bisschop, Florinda De Simini, Timothy Lubin, Nirajan Kafle, and Nina Mirnig, among
others), our knowledge of tantric and Śaiva traditions, and of early medieval religiosity
more generally, has grown in leaps and bounds and will only continue to do so in the coming decades. Unfortunately, however, this specialized (and predominantly European) tantric studies scholarship has, for the most part, not made its way into the basic formulations
of nonspecialist scholarship on South Asian religious history and practice, something I
burc19032_1st_i-412.indb 324
12/12/18 6:14 PM
1. thE tantric agE = 325
62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
72.
73.
74.
75.
76.
77.
78.
79.
80.
81.
hope to address in some small way in this chapter. For a useful assessment of the field of
tantric studies, including discussion of key authors and publications that have advanced
scholarly understandings of Tantra in recent decades, see Goodall and Isaacson 2011.
For my present purpose, blood sacrifice and offering of alcohol will not be considered radically transgressive practices. Sanguinary rites did involve impure and polluting substances,
but they seem to have been so widely practiced in medieval India that they cannot be
understood to have violated fundamental social codes and shocked or aroused social censure in the same way that (the far more rarely performed) tantric sexual and mortuary
rites clearly did.
A major branch of Śaiva tantric scriptures known as the Vidyāpīṭha, for instance, discusses
the performance of mortuary and sexual rites centered on the offering of conventionally
impure substances (including blood, alcohol, and sexual fluids) to feminine deities (goddesses and yoginīs) in order to acquire their extraordinary powers. Later tantric texts of the
Kaula tradition offered domesticated, private versions (for householders) of such transgressive practices, detailing ritual copulation with “polluted” outcast women and consumption of alcohol, human feces, and sexual fluids as a means to achieve and perform
nondual spiritual enlightenment, with its transcendence of the dualities and arbitrary
moral codes of the social world.
The term “practical magic” is meant to refer to a wide spectrum of this-worldly ritualized
practices with more pragmatic, worldly aims, including making protective amulets, love
potions, and spells, or power-giving pills, performing weather and crop-related rites, and
conducting rites to defeat or harm enemies.
Sanderson 2013, 214. While generally not public or mainstream in the way that Saiddhāntika
rituals and institutions were, non-Saiddhāntika Śākta- oriented communities also conducted rituals for the protection of the king and state, especially in times of danger.
Sanderson 2006a, 146.
Davidson 2002, 26.
Sanderson 2001, 8–11. These three kings were from the Deccan, Orissa, and the Tamil south.
Around the same time (seventh to eighth centuries) it seems Pāñcarātra Vaiṣṇava tantra
began to be important in royal cults in Kashmir and the Tamil country.
White 2011, 579.
White 2011, 578.
Sanderson 2009, 254.
Sanderson 2009, 258–59.
Sanderson 2009, 249–52, 301–3.
Davidson 2002, 177.
Davis 1991, 8.
On the economic role of temples, temple patronage, and the ways that temple building was
linked to the expansion of brahmanical authority, agricultural development, and political
integration, see Talbot 2001, 94, 102–3, 117–20.
Sears 2014, 10.
Sears 2008, 26.
Sanderson 2009, 266.
Sears 2014, 6.
Misra 1997, 75–77. Misra’s research, relying predominantly on epigraphic records, suggests
that maṭhas of the Mattamayūra order—a branch of the tantric Śaiva Siddhānta school—in
burc19032_1st_i-412.indb 325
—-1
—0
—+1
12/12/18 6:14 PM
326 9 1. thE tantric agE
82.
83.
84.
85.
86.
87.
88.
89.
90.
91.
92.
93.
94.
-1—
0—
+1—
the Kalacuri state of central India (ca. seventh to thirteenth centuries) supported the state
by garrisoning war forces, manufacturing armaments, offering training in warfare, and
perhaps even themselves recruiting a combatant force.
Sears 2014, 10.
Sears 2008, 26.
Sanderson 2009, 268.
Sanderson 2009, 267.
Medieval kings allied themselves primarily with Saiddhāntika Śaiva tantric communities
and institutions, which subsumed and preserved the brahmanical social order; however,
the broad shift tantra (from esoteric, private contexts) into mainstream, public settings
also involved elements of the heterodox non-Saiddhāntika tantric traditions, particularly
in rituals conducted to assist and protect the king and state against enemies and calamities. India’s well-known yoginī temples, constructed mostly in the tenth through twelfth
centuries, are a case in point. Shaman Hatley remarks that, by the tenth century, the originally esoteric yoginīs “became prominent in the wider Indic religious landscape, as attested
by their entry into the purāṇic literature and the unique, circular, open-air temples
enshrining them across the subcontinent” (2012, 107). Citing evidence of the integration of
yoginī temples into major state-sponsored temple complexes, their proximity to royal capitals, and the royal patronage of these temples, Hatley argues that the yoginī temples reflect
“the adaptation of esoteric pantheons and secretive praxis systems to a more public, calendrical liturgy suiting the aspirations of elite patrons and performed in permanent structures” (2014, 216). In the public context of the yoginī temples, tantric and purāṇic modes of
worship melded. Rituals involving sex or sexual fluids seem to have been abandoned, but
offerings of blood (animal sacrifice) and alcohol, night vigils, and fire rituals were central.
Relatedly, Judit Törzsök (2012) has shown that Kalhaṇa’s Rājataraṅgiṇī, a Sanskrit work of
twelfth-century Kashmir, recommends a king’s engagement with tantric mantras and
magic but sees participation in polluting, transgressive (Kaula and Krama) tantric rituals
(including yoginīsādhana) as undesirable and dangerous, even as causes of a king’s downfall.
Sanderson 2013, 224.
Sanderson 2013, 224.
Sanderson 2015.
The Śivadharma corpus, which has not yet been critically edited, consists of the Śivadharma,
Śivadharmottara, Śivadharmasaṃgraha, Umāmaheśvarasaṃvāda, Uttarottaramahāsaṃvāda,
Śivopaniṣad, Vṛṣasārasaṃgraha, Dharmaputrikā, and Lalitavistara. The first two texts (the
Śivadharma and Śivadharmottara) of this corpus are broadly attested and were clearly composed against a Pāśupata (Śaiva Atimārga) background, but the full corpus is attested only
in Nepal, with the later texts of the corpus reflecting the concerns of other (non-Pāśupata)
communities; Florinda De Simini, email message to author, June 6, 2018.
De Simini 2016a, 22; De Simini 2016b, 236.
De Simini 2016a, 49.
Śivadharma 1.36; translation in Schwartz 2012a, 211.
Translation of this passage is mine, based on the Sanskrit text provided in Ganesan and
Sathyanarayanan 2010–2011, 54. I have consulted Ganesan and Sathyanararyan’s translation as well as that in Schwartz 2012a, 212, 227–28. I am grateful to Florinda De Simini for
checking the Sanskrit text used by these scholars against an earlier manuscript of the text
in her possession (Asiatic Society of Calcutta, G4077, dated 1036). She found only very minor
differences in the text that did not substantively change the meaning of the passage. The
burc19032_1st_i-412.indb 326
12/12/18 6:14 PM
1. thE tantric agE = 327
95.
96.
97.
98.
99.
100.
101.
102.
103.
104.
105.
106.
107.
108.
109.
110.
111.
112.
113.
114.
115.
manuscript used by Schwartz inserts kīrtan into the description of bhakti, a word not present in the other manuscripts used by De Simini or Ganesan and Sathyanarayanan.
The Śivadharma and its eight-limbed description of bhakti precede the Bhāgavata Purāṇa and
its well-known ninefold classification of bhakti by nearly half a millennium. It would seem
to be earlier than even the vernacular “emotional bhakti” expressed by the Vaiṣṇava Tamil
poet-saints. Nammālvār’s impassioned hymns come from the seventh century (i.e., most
likely after the Śivadharma), and the few Tamil Vaiṣṇava bhakti saints (Ālvārs) preceding
him seem to have expressed a more intellectual devotion related to the theistic yoga of the
Bhagavad Gītā. For some other brief comments upon the nature of the Śivadharma’s bhakti
and its relationship to that of the Bhāgavata Purāṇa and later bhakti traditions, see Schwartz
2012a, 212, 227–28.
Travis Smith 2016, 363.
On this topic, see De Simini’s outstanding study Of Gods and Books (2016a).
The Devīpurāṇa (91.23) states that “[from] listening, bhakti emerges; [urged] by bhakti, one
sits intent upon the guru, and this explains the scriptures of knowledge (vidyā-āgamān).
Knowledge resides in manuscripts (granthā), o king!” (De Simini 2016a, 79–80n222).
Sanderson 2012–2013, 4. He remarks that “the text probably envisages its being chanted in
Sanskrit with each verse or group of verses followed by an explanation in the vernacular.”
Holdrege 2015, 81, 82–84.
Holdrege 2015, 81; emphasis in original.
For a fascinating discussion of the bhaktajana as a devotional public representing the political
power of the quotidian world, and bhakti as “a fulcrum around which social cohesion adheres,”
see Novetzke 2016, 93–102. Novetzke discusses a series of Marathi inscriptions, 1189–1311, in
which the Yādava state of Maharashtra invokes a bhakti public—the bhaktajana—in a display of
beneficence meant to appeal to a quotidian populace of devotees (and thus acknowledging
them as a political force).
It appears that the lay Śaiva devotional tradition of early medieval India was more socially
liberal than its lay Vaiṣṇava counterpart, and this may have had something to do with
Śaivism’s great success. Timothy Lubin (2017) has discussed how the Śivadharma makes far
more significant moves toward social inclusion (particularly for women and śūdras) than
the Viṣṇudharma (the authoritative lay Vaiṣṇava scripture likely composed around the
same time as the Śivadharma), whose conception of Viṣṇu-bhakti mostly adheres to classical
Smārta brahmanical status hierarchy and its prerogatives.
Lubin 2011, XXX.
Hazra 1952–1953, 12.
Hiltebeitel 2012, 159.
Hiltebeitel 2012, 160
De Simini 2016a, 46, 66.
De Simini 2016a, 66.
De Simini 2016a, 67n194.
De Simini 2016a, 58.
Sanderson 2013, 212; De Simini 2016a, 50.
Bisschop 2014, 134–35; De Simini 2016a, 51–52.
On the dating of the Pāśupata sect (and debates surrounding it), see Lorenzen (1972) 1991b,
173–92. Some scholars date the Pāśupatas to the second century BCE, but Lorenzen gives
compelling evidence for a second century CE date.
Bisschop 2010, 484–85.
burc19032_1st_i-412.indb 327
—-1
—0
—+1
12/12/18 6:14 PM
328 9 1. thE tantric agE
-1—
0—
+1—
116. Davidson 2002, 85, 186.
117. Bakker 2014, 14.
118. According to Bisschop, “the earliest explicit epigraphical references to Pāśupatas that we
possess are at the same time among the earliest examples of copper-plate grants recording endowments for temple worship” (2010, 485). These are late fourth- century inscriptions that refer to Pāśupatas as recipients of grants for the performance of worship in
temples. This is not to say that Pāśupatas “created” the Śaiva temple cult, for the evidence suggests that the tradition of temple-based lay devotion was an independent and
preexisting phenomenon that initiatory communities like the Pāśupatas later integrated.
See Sanderson 2012–2013, 2– 8.
119. Travis Smith 2007, 85, 313.
120. Travis Smith 2007, 301–2.
121. The Īśvara Gītā, a circa eighth-century Pāśupata philosophical poem (part of the popular
Kūrma Purāṇa) studied and translated by Andrew Nicholson (2014), teaches a yoga (seemingly open to śūdras and women) in which visualization meditation exercises facilitate the
yogī’s self-deification. I have highlighted the divinization of the self as one of the distinctive
elements of tantric practice, thus the goal of the Īśvara Gītā’s Pāśupata yoga—to identify
with and “become” God (Śiva)—speaks to how fuzzy the boundaries were been between
“tantric” (mantra-mārga) practice and atimārga Pāśupata practice in the eighth century.
122. Travis Smith 2007, 295, 313.
123. For evidence of how the patronage of temples and monasteries that spurred the growth of
early Śaivism was not undertaken primarily or exclusively by political elites but was significantly dependent upon the religious giving (dāna) of local collectives of merchants,
traders, and artisans, see Cecil 2016.
124. Alvarez 1990, 407–14. Relatedly, Ronald Davidson identifies a list of more than a hundred
probable Pāśupata sites all over India (2002, 184, 341–43).
125. De Simini 2016a, 59.
126. Bonazzoli 1993, 348. Śivadharma 12.109 states, yathā śivas tathā yogī yathā yogī tathā śivaḥ //
127. De Simini 2016a, 51n46. The “sixfold yoga” mentioned by the text is likely the same as the
six-limbed Pāśupata yoga taught in the Skanda Purāṇa, a yoga that, S. D. Vasudeva notes, has
many parallels with the yogas of the Śaiva Mantramārga, an indication of tantric yoga’s
Pāśupata prehistory (2017, 3).
128. While the yogī is the spiritual ideal in these texts, the Śivadharma and Śivadharmottara do
sometimes give hints that others (i.e., not only initiated yogins but also lay devotees) can
attain liberation through the power of the śiva-jñāna; Florinda De Simini, email message to
author, June 19, 2018.
129. My translation from the Sanskrit text in Hazra 1952–1953, 13n28. tan nirvedāc ca vairāgyaṃ
vairāgyāj jñāna-saṃbhavaḥ / jñānat pravartate yogo yogād duḥkhāntam āpnuyāt // Thanks to
Hamsa Stainton for his assistance with this verse.
130. White 2009, 167–94; Coleman 2014.
131. White 2009, 168.
132. Travis Smith 2007, 304, 307–9.
133. Travis Smith 2007, 307.
134. Davis 1991, 15. Particularly influential was the Mattamayūra branch of the Śaiva Siddhānta,
based in central India but with a network of monasteries extending from Kashmir to the
seacoasts of the south. On this important tantric lineage, see Sears 2014, 12–33.
burc19032_1st_i-412.indb 328
12/12/18 6:14 PM
1. thE tantric agE = 329
135.
136.
137.
138.
139.
140.
141.
142.
143.
144.
145.
146.
147.
148.
149.
150.
151.
152.
Travis Smith 2012, 173.
Chattopadhyaya 1993, 37.
Sanderson 2009, 42.
Ronald Davidson (2002, 173, 225–26) notes a striking change in early medieval literature,
which contains far more positive attitudes toward, and depictions of, “tribal” peoples than
earlier literature, a trend correlated with state/brahmanical expansion into forest/jungle
areas, increased contact with “tribals,” and feudalization of “tribal” clans.
Brahmans, of course, are not a homogeneous community; there are enormous regional
variations and complex internal differentiations among brahman communities, yet there
is still great heuristic value in the notion of a “brahmanical tradition.” As Kunal Chakrabarti puts it, “We cannot challenge the proposition that if in India’s diverse cultural traditions there is one way of life and one medium of expression that can claim a semblance of
pervasive influence these are Brahmanism and Sanskrit. The brahmanical tradition has
decidedly the widest horizontal spread which cuts across regional boundaries. It is a continuous, overarching tradition with an essential unity of content and purpose compared to
the many localized traditions which, though rich in cultural content, are nevertheless
varied and fragmented” (2001, 95–96).
Yokochi 2004: 15. Jaya Tyagi describes how the inclusion of goddesses (and certain vratas) in
purāṇic and tantric literature of the period had much to do with a realization that women
were important patrons and promoters of religious cults (Tyagi 2014: 89–136).
Yokochi 2004, 16.
Sanderson 1988, 660; White 2003, 127.
See especially Chakrabarti 2001.
Chakrabarti 2001, 190.
Sanderson 2013, 217.
Sanderson 2006b, 15. Sanderson also states, “In Purāṇic texts such as the Uttarabhāga of the
Liṅgapurāṇa, the Kālikāpurāṇa, the Devīpurāṇa, and the Agnipurāṇa, the boundary between
Smārta and Tantric domains has almost completely dissolved” (2009, 250).
Kafle 2015, 33, 61. The dating of this text was proposed by Anil Kumar Acharya. Kafle’s
research shows that the Śivadharmasaṅgraha has significant parallels with the Skanda Purāṇa
and Vāyu Purāṇa, contains five full chapters that closely parallel the Niśvāsamukha (the
introductory text of the canonical Śaiva Tantra, the Niśvāsatattvasaṃhitā), and two chapters
identical to, or corresponding closely with, parts of the Guhyasūtra, the fifth and final sūtra
of the Niśvāsatattvasaṃhitā.
Schwartz 2018; Davis 2005.
Schwartz 2018, 7.
Schwartz 2018, 16–17.
R. S. Sharma argues that, in the context of early medieval agrarian expansion, śūdras who
had been enlisted by brahmans as cultivators (karśakas) of newly settled lands “naturally
came to have some rights” in those lands and “this became inconsistent with the traditional ritual status of the śūdras which had to be raised by providing initiation for them in
the tantric sects” (1974, 179–80).
Mirnig 2013, 292, 298. Mirnig’s research on the different modes of śrāddha (postmortuary ancestor worship) rites offered to householders by the Śaiva Siddhānta tradition in
pre-twelfth- century sources shows how tantric communities reached out to the broader
populace of Hindu householders while suggesting that modes of worship prescribed for
burc19032_1st_i-412.indb 329
—-1
—0
—+1
12/12/18 6:14 PM
330 9 1. thE tantric agE
153.
154.
155.
156.
157.
158.
159.
160.
161.
162.
163.
164.
165.
166.
167.
168.
169.
170.
171.
172.
-1—
0—
+1—
entry-level (samayin) tantric initiates were at times the same as those prescribed for lay
devotees.
Sanderson 2013, 237–39. On mass tantric initiation ceremonies, see also Nandi 1973, 81.
Sanderson 2013, 239.
Indrani Chatterjee 2013, 6. The “payments” or “gifts” made to a guru (and his monastic
order) ranged from labor services or foodstuffs (grain, wine, meat) to artisanal bronze, silver, or gold items to tax-exempt land grants. On the topic tantric initiates’ surrendering
portions of their wealth (to the guru) for use by the community they have just joined, see
Schwartz 2018, 19–20.
Travis Smith 2012, 173.
Gavin Flood 2006, 73.
Gavin Flood 2006, 28.
Frederick Smith 2006, 376.
Frederick Smith 2006, 369.
Frederick Smith 2006, 383.
Frederick Smith 2006, 385.
Slouber 2017a.
Slouber 2017a, 16. Diwakar Acharya (2016, 157–58) notes that the Gāruḍa Tantras are usually
paired with the Bhūta Tantras, probably because both are concerned with healing. The
Bhūta Tantras incorporated (into tantric scripture and ritual systems) and expanded upon
the branch of Āyurveda known as bhūtavidyā, which deals with rituals of appeasement to
various semidivine and demonic beings to pacify these possessing spirits and to heal and
free people from their grip.
Slouber 2012, 153.
Slouber 2012, 152.
Slouber 2012, 50. The Bhūta Tantras describe the same basic procedure for the exorcist,
though he is to visualize himself not as Garuḍa but as Bhairava (or Skanda) in order to
pacify or banish the demon/spirit possessing the patient. For details of the gāruḍika’s tantric ritual procedure, see especially Slouber 2017a, 67–74.
Indeed, in other work, Slouber (2017b) has suggested that tantric exorcists probably drew
on the practices of tribal shamans and sought to spread these practices in an altered form,
merging them with Sanskritic tantric ritual and developing them “in conjunction with the
idiom of Śaiva mantra-śāstra.”
Slouber 2012, 51; Slouber 2017a, 127–28.
Slouber 2017a, 127–28. Kṣemendra’s eleventh-century satirical poem Narmamālā (2.142.145)
describes an untouchable leather worker who raises his status by, at one point, “landing a
job as a protector of crops because he knew the Gāruḍa Tantras” (128).
Babb 1975, 178–79.
In emphasizing that the tantric ritual process and pattern were relatively constant and
shared, I do not mean to imply that sect- and text-specific doctrinal, theological, and metaphysical content (the specific deities, mantras, descriptions of the levels of the cosmos, etc.)
were not important. Indeed, they were absolutely central in teaching practitioners to
“inhabit a tradition-specific subjectivity”; however, what I wish to emphasize instead is a
shared pan-Indian tantric culture of ritual forms/methods and cosmological assumptions,
“a shared substrate of ritual and cosmology in spite of divergent metaphysical claims” (Gavin
Flood 2006, ix, 28). Specific sectarian philosophical commitments certainly influenced how
burc19032_1st_i-412.indb 330
12/12/18 6:14 PM
1. thE tantric agE = 331
173.
174.
175.
176.
177.
178.
179.
180.
these shared ritual technologies were used and understood, thus we might imagine a religious world characterized by the multinodal interaction of reinterpretations of a stable
core of tantric ritual procedures, with each sectarian reinterpretation inspiring further
competition and development.
Sanderson 2001, 38–39n50. Dominic Goodall’s (2011) study of the enthronement of a central
deity (in visualization meditation) as a central practice in nearly all tantric cults provides
further evidence for this notion of “a single ‘Tantric’ language.” In other work, Goodall and
Harunaga Isaacson (2016) have offered a detailed identification and discussion of the specific elements of a “shared ritual syntax” among tantric traditions.
Slouber 2012, 153.
White 2003, 150.
At the scholastic level, the rise of Vedānta centered on competing interpretations of
Bādarāyaṇa’s Brahmasūtras (a.k.a. the Vedānta Sūtra), a summary and systemization of the
Upaniṣad’s ideas regarding the nature of the universe and the path to liberating knowledge. Unfortunately, a proper analysis of the pivotal role Vedānta played in late medieval
and early modern religious developments (including especially developments in the
bhakti and yoga traditions) in both North and South India is a topic beyond the scope of
this book.
On Advaita Vedānta and yoga, see Schwartz 2017 and Mallinson 2014a; and on Advaita
Vedānta and bhakti, see, for example, Venkatkrishnan 2015 and Barua 2017.
Nicholson 2010, 2, 200–201. Andrew Nicholson has argued that “the perceived threat of
Islam” motivated Sanskrit intellectuals in the twelfth century to create, for the first time,
“a strictly defined category of āstika philosophical systems, systems that professed belief in
the authority of the Veda.” Before this the category of āstika had been an indistinct one
potentially admitting Buddhists and Jains, but between the twelfth and fourteenth centuries Advaita Vedāntins came to permanently classify Buddhists and Jains as “deniers”
(nāstikas), while simultaneously the category of nāstika underwent “a subtle blurring with
categories like ‘barbarian’ (mleccha), allowing foreigners to be classed together with Buddhists and Jainas” (200). Drawing on the earlier arguments of David Lorenzen (2005) and
Sheldon Pollock (1993), Nicholson thus suggests that it was the presence of Islam that
sparked these intellectual efforts to conceive a distinct, unified Hindu identity.
Fisher 2017, 38–48. While pre-twelfth-century tantric Śaiva and Vaiṣṇava traditions had
positioned themselves as independent religious systems whose basis of authority in tantric
revelation required no reference to the Veda, with the influential works of Rāmānuja and
Śrīkaṇṭha—which utilized Vedāntic exegesis to present their respective Vaiṣṇava and Śaiva
traditions as emblematic of Vedic orthodoxy—this began to change. Sectarian theistic
communities in South India that had until then relied solely on tantric scriptural authority
now increasingly deferred to norms and doctrines grounded in the authority of Vedic revelation; see Clark 2006, 215, 221–22. Using the philosophy of Vedānta, theologians were able
to variously meld bhakti and yoga (including the techniques of tantric yoga) with jñāna,
thereby constructing a host of new, competing sectarian Vaiṣṇava and Śaiva forms of
orthodox brahmanical religiosity.
Allen 2014, 883; Minkowski 2011, 218. Michael Allen (2014) argues that “the late medieval
developments discussed by Nicholson [2010]”—i.e., intellectual efforts to conceive a distinct,
unitary Hindu philosophical tradition—“might arguably be seen as a natural unfolding
of scholastic commentarial traditions—with their commitments to systematizing ideas,
burc19032_1st_i-412.indb 331
—-1
—0
—+1
12/12/18 6:14 PM
332 9 1. thE tantric agE
refining terms and categories, and resolving apparent contradictions—rather than as a
direct response to Islam” (883). Though Allen does not establish a meaningful intellectual
trend prior to the twelfth century, he complicates Nicholson’s argument (see n. 176) by
offering some examples of pre-twelfth-century scholars (e.g., Vācaspati Miśra, tenth century) who drew a sharp line between schools that accept the Vedas and those that openly
rejected them (Buddhists, Jains, Kāpālikas) and even associated the non-Vedic schools with
mlecchas. Relatedly, Christopher Minkowski (2011) has suggested that the key shift of the
powerful Śṛngerī and Kāñcīpuram maṭhas from tantric Śaivism to a brahmanical Advaita
Vedāntic Śaivism in the late medieval and early modern periods (see Clark 2006, 177–226)
was not any sort of reaction to Islam but rather was “probably in response to the growth of
the Śrī Vaiṣṇava institutions” (218).
181. Indian intellectuals and social elites, unlike much of the Indian populace at large, do seem
to have perceived Turko-Afghan invasions and the establishment of the Delhi Sultanate as
a threat—specifically a threat to the traditional brahmanical social and political order (i.e.,
to brahmanical privilege)—though they generally did not conceive them or the threat they
posed in religious terms. See, for instance, Talbot 2003 and Ernst 1992, 29–37.
182. White 2011, 577.
183. White 2011.
2. Sultans, Saints, and Songs
-1—
0—
+1—
1. It was Marshall Hodgson (1974, 59) who coined the term “Islamicate” to distinguish “the
social and cultural complex historically associated with Islam and the Muslims” from the
religion of Islam (e.g., its doctrines, rituals, etc.) itself. Hodgson also coined and elaborated
the term “Persianate” (1977, 293–314).
2. Eaton 2003a, 9.
3. Eaton 2003a, 9.
4. Eaton and Wagoner 2014, xxii.
5. Eaton and Wagoner 2014, xxv.
6. Pollock 2006.
7. Eaton and Wagoner 2014, 21.
8. Eaton and Wagoner 2014, 24.
9. Digby (1986) 2003, 235.
10. Eaton 2003a, 11.
11. O’Hanlon 2007a, 364.
12. Richard Eaton (2000b) has shown that the Turks’ destruction and desecration of temple
sites was generally not done for religious motives but out of strategic political considerations. Since temples were important symbols of political power for Hindu kings, violence
to temple sites was intended to delegitimate the previous royal authority in newly conquered realms. For these same reasons, competing Hindu kings engaged in desecration of
each other’s temples well before the Sultanate period. Thus, the temple destruction/desecration of the Sultanate period was a continuation of an established Indian militarypolitical practice and was not based in anti-Hindu religious sentiment but rather was a way
to seize resources/wealth and to undermine the authority of enemy kings (whose power
was considered to be embodied in the temples). Nevertheless, it is true that destructions of
burc19032_1st_i-412.indb 332
12/12/18 6:14 PM
2. sultans, saints, and songs = 333
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
Hindu monuments were sometimes celebrated in Afghan documents as a religious achievement as part of the rhetoric of Islamic rule; see also Davis 1993.
See Talbot 2003, 90–93.
Ernst 1992, 31.
See Chandra 1996, 122–30.
Willis 1993, 51.
Wink 1997, 294. Wink explains, “It was only in South India that the building of large temple
complexes remained embedded in a Hindu polity and continued to be organically linked to
the other institutions of kingship and social organization in a variety of complex ways. . . .
Nowhere in the North did the Hindu temple building tradition perpetuate itself without
hindrance” (324, 327).
Davis 1991, 17–18.
Davis 1991, 17–18.
Sears 2009, 27.
Sears 2009, 8.
Sears 2009, 10, 24.
Satish Chandra remarks, in the context of the changes wrought by Sultanate power, “it
would appear that the first beneficiaries of the diminished influence of the brahmans were
the Nathpanthi yogis. This sect seems to have reached the height of its prestige and influence during the 13th and 14th centuries” (1996, 124).
Eaton and Wagoner 2014, 25.
Alam 2004, 118.
Eaton and Wagoner 2014, 26.
See Wagoner 1996. It seems that in many cases, the “Islamicate” (Wagoner’s term of choice
in the article cited here) was, in fact, mediated by the “Persianate.”
Sheikh 2017, 38–39.
Eaton 2005, 25–26. Eaton explains how the Sultanate was able to use the iqtā‘ system to penetrate “the grass roots of local politics by co-opting and redefining local revenue systems
and personnel.” In some cases, the sultans in Delhi recognized lands held by entrenched
local chiefs as iqtā‘ and designated those chiefs as the overseers (iqtā‘ holders) of that land,
thus seeking to transform potential enemies into servants of the state.
Behl 2012a, 16.
Emblematic of this process is the great Persianate Indian poet and musician Amīr Khusrau
Dihlavī (1253–1325), patronized by both Khalji and Tughlaq sultans. Born of an Indian
mother, he proudly modeled a distinctly Indic tradition of Persian poetry, was a disciple of
the Chishti shaikh Nizām al-Dīn Awliyā’ and is famous for his influence on Indian musical
traditions, credited with begetting the South Asian Sufi musical form of qawwālī.
Stewart and McGregor 2018; emphasis in original.
Stewart and McGregor 2018.
Richards 1998b, 10.
Ernst 1992, 25.
Ernst 1992, 25–26.
Hawley 2015b, 90.
Ernst 1992, 38.
Lorenzen 2005, 53; Nicholson 2010, 2–3, 200–201.
Moin 2012a, 33–34.
burc19032_1st_i-412.indb 333
—-1
—0
—+1
12/12/18 6:14 PM
334 9 2. sultans, saints, and songs
-1—
0—
+1—
41. Eaton 2005, 45.
42. Digby 1975, 17–18.
43. On the spread of Sufi khānqāhs in lineage-based networks across India, see Nizami 1961,
175–77.
44. Digby 2004, 302–5.
45. The Chishti order began in central Afghanistan but became the largest and most popular of
South Asian Sufi traditions. It was established in India by Mu’in ud-Din Chishti (d. 1236),
who settled in Ajmer, Rajasthan, in the wake of the Ghurid conquest of North India.
46. Digby (1986) 2003, 242–52. A key factor in the success of Nizām al-Din and the Chishtis was
also the popularity and impact of Nizām al-Dīn’s master, Bābā Farīd (d. 1265), in the Punjab.
On Bābā Farīd, see Eaton 2003b.
47. For a historical introduction to the Chishti, Suhrwardi, Naqshbandi, and Kubrawi Sufi
orders in India, see Rizvi 1978, 114–300.
48. Bulliet 2004, 36.
49. Moin 2012a, 7–8.
50. Eaton 2003b, 263.
51. Eaton 2003b, 264.
52. Moin 2012a, 100.
53. Sears 2009.
54. Ernst 2005, 23.
55. Quoted in White 2009, 232.
56. Joel Bordeaux, email message to author, July 4, 2014.
57. Behl 2012a: 155.
58. Behl 2012a, 261– 62.
59. Digby 1975, 51.
60. Hatley 2007b, 367.
61. Shaikh Ghawth’s Baḥr al-ḥayāt (ca. 1550) is a Persian translation, revision, and expansion of
the most important Islamic work on yoga, the Hawd al-ḥayāt (The pool of life), which was
itself an Arabic translation (composed in Bengal in 1210) of a nonextant Sanskrit work on
tantric yoga known as the Amṛtakunda (The pool of nectar). The history of this text and its
uses attest to a long-running Indian Sufi interest in, and Islamization of, tantric yoga practice. See Ernst 2003.
62. Ernst 1996, 13.
63. Accardi 2015.
64. Dobe 2015, 33.
65. As Christopher Shackle has explained, Sufis understand “divine love as the core organizing
principle of the universe” and conceive of “a hierarchy extending upwards from the interpersonal loves of the phenomenal world to the transpersonal connection with the Divine
which is perceived as their real exemplar.” Sufis defined “the twin force of love as ‘phenomenal love’ (‘ishq-e majāzī) and ‘real love’ (‘ishq-e ḥaqīqī) in which the human is seen as the
mirror of the divine” (2006, 88).
66. Pinch 2006, 65–66.
67. See White 2001; Digby 2000, 140–220.
68. Stewart 2000, 47.
69. Stewart 2004, 14. Satya Pīr is prominent among middle- and low-class Hindus and Muslims
in West Bengal, Orissa, and Bangladesh; he has more Bengali texts dedicated to him than
anyone except Caitanya.
burc19032_1st_i-412.indb 334
12/12/18 6:14 PM
2. sultans, saints, and songs = 335
70. Stewart 2004, 15.
71. Hawley 2015b, 4.
72. See, for instance, Tara Chand’s Influence of Islam on Indian Culture (1936) and Muhammad Hedaytullah’s Kabir: The Apostle of Hindu-Muslim Unity (1977), in addition to the works cited hereafter
by Diana Eck, Shahabuddin Iraqi, Satish Chandra, and Aziz Ahmad. These authors rarely analyze specific instances of the “historical influence” they assert, do not provide explanations of
the actual dynamics of the “symbiotic/syncretic” relationship they suggest, and too often
posit bhakti and Islam as firmly bounded, discrete entities, the one responding to and reacting
against the other. For a useful discussion of some of this scholarship, see Hawley 2015b, 89–94.
73. Eck 2012, 90.
74. Iraqi 2009, 253, 258.
75. Chandra 1996, 127–28.
76. Ahmad 1964, 136, 140.
77. As Samira Sheikh (2017, 35–36) has argued—drawing on Simon Digby’s important essay
“Before Timur Came” (2004)—contrary to its common scholarly portrayal, Timur’s invasion
was more of a tipping point than a transformative cataclysm. The decentralization, increased
mobility, extension of trade routes, sprouting of new towns, and linguistic experimentation
usually associated with the (post-Timur) fifteenth century were already well under way in
the (pre-Timur) second half of the fourteenth century but were given a great deal of further
impetus by the devastation (and population dispersal) caused in Delhi by Timur’s invasion.
78. Moin 2012a, 25–26, 28–29.
79. Moin 2016, 127.
80. Occult practices such as lettrism, astrology, geomancy, divination, and spirit communication and subjugation were not unusual among medieval and early modern Muslims, and
post-Mongol Persianate rulers often seized upon the occult sciences to harness sacred
power for political purposes; see Melvin-Koushki 2016, 143. In the many manuscripts of
occult texts that circulated throughout the Islamic world at this time, Neoplatonic theories
are used to frame all occult powers as deriving from the one God and all occult operations
as “sanctioned and even effected by the power of God acting through his angels and the
spirits . . . of the celestial sphere,” conceived as the highest beings (divine ministers) with
whom men could be in contact; see Pingree 1980, 3–4.
81. As Carl Ernst (2009, 199–200) has clearly shown, Indian tantric teachings were circulating
within Islamic occultist circles in the Persianate world of Timur. Ernst explains that most
Persianate Muslims would have understood occult aspects of tantric practice—e.g., mantras and yogic practices for summoning yoginīs and acquiring their powers—in terms of the
familiar and largely accepted tradition of the Islamic occult sciences (al-‘ulūm al-gharība).
He posits that Persianate Muslim rulers probably engaged more with the occult dimensions
of tantric yogic practices than did any other sector of society because of the politicalmilitary use to which these occult arts could be put.
82. Digby 2004, 301–2. The historical shift away from a more centralized Sultanate and toward
the development of a set of distinctive regional centers was not, of course, the result of any
single political event. Rather, Timur’s invasion was one crucial event precipitating—and
serving as a convenient historical marker for—the rise of these vernacular politicalcultural nexuses and a broader Indo-Islamicate “shared local” culture.
83. In fact, the Bahmani sultanate in the Deccan had declared independence from the Tughlaqs
in 1349, beginning the decline and fracturing of the Tughlaq dynasty into multiple smaller
regional sultanates.
burc19032_1st_i-412.indb 335
—-1
—0
—+1
12/12/18 6:14 PM
336 9 2. sultans, saints, and songs
-1—
0—
+1—
84. Sreenivasan 2014b, 243–47.
85. Ramya Sreenivasan (2014b, 243–47) has demonstrated that some of the earliest literary
works in (local forms of) the North Indian vernacular of Hindavī—Dā’ūd’s Cāndāyan (ca.
1379), Nārāyaṇdās’s Chitāī-carita (ca. 1526), and Jāyasī’s Padmāvat (ca. 1540)—were patronized
by and addressed specifically to the particular political concerns of rural gentry and local
warlords in the hinterland; see also Behl 2012a, 48.
86. Orsini 2012, 227.
87. Orsini 2012, 228–29.
88. Orsini quite rightly suggests that in understanding the use of language in Sultanate and
Mughal India we should not think “purely in terms of High and Low but rather in terms of a
continuum, something that makes us more aware of the importance of recognizing registers both within High and Low: ‘ornate Persian’ versus ‘simple Persian’; ‘Persian-near’ or
‘Persian-far’ and ‘Sanskrit-near’ versus ‘Sanskrit-far’ vernaculars” (2014c, 407).
89. Orsini and Sheikh 2014, 14–15; Orsini 2005, 395.
90. Bangha 2014, 395–96.
91. Orsini and Sheikh 2014, 14–16.
92. Orsini 2014c, 404.
93. Though Nāmdev was not from North India and seems to have lived well before its great
bhakti saints, his memory—and Hindavī poetry attributed to him—became crucial features
of North India’s emerging bhakti public. Indeed, according to most sources, while Nāmdev
flourished in the early fourteenth century in Maharashtra, he traveled north and was a
founding figure of North India’s bhakti movement.
94. Hawley and Juergensmeyer 2004, 7.
95. Novetzke 2015, 176.
96. Novetzke 2015, 180.
97. Orsini 2014a, 199.
98. Orsini 2014a, 201–2.
99. Behl 2012a, 286.
100. Ernst 1992, 166– 67; Orsini 2014c, 409, 415.
101. Orsini 2014b, 223; emphasis in original.
102. Behl 2007, 321.
103. Behl 2012a, 16–22, 328–29.
104. Digby 2004, 351.
105. Williams 2014, 88; Behl 2012a, 324.
106. Behl 2012a, 13.
107. Behl 2012b, 36.
108. Orsini 2014b, 228, 232.
109. Sheikh 2010, 168.
110. Green 2015, 13. Green uses this language to refer to nineteenth-century religious economies, but I think it is certainly apt here as well.
111. Hawley 1984, 245.
112. Gāyā bin pāyā nahi, anagāvan se dūr / jin gāyā vishvās se, sahib hāl hazūr; Hess 2015, 32.
113. Hawley 1984, 249.
114. Hawley 1984, 247. Hawley explains that Sūrdās commonly concludes his compositions with
the following expression: sur dās bhagavant bhajan binu; that is, “Sūrdās says, unless one sings
to the Lord . . .” (246).
burc19032_1st_i-412.indb 336
12/12/18 6:14 PM
2. sultans, saints, and songs = 337
115.
116.
117.
118.
119.
120.
121.
122.
123.
124.
125.
126.
127.
128.
129.
130.
131.
132.
133.
134.
135.
136.
137.
138.
139.
Williams 2014, 110–11.
Hawley 1984, 258.
Hawley 1984.
Hirschkind 2005, 30. As Tyler Williams (email message to author, August, 26, 2016) has
pointed out to me, many bhakti hagiographies express ideal forms of community interaction and ethical behavior that “model the performance contexts, and consequently the
social contexts” in which they are performed. In this sense, the singing of bhakti stories
in a community “simultaneously reproduces both the community and ethical behavior
within it.”
Hawley 1984, 251.
Mallison 2000, 292.
Geertz 1973, 129.
Geertz 1973, 127. While his essay “Religion as a Cultural System” has received far more
attention (and its fair share of criticism), I find Geertz’s “Ethos, World View, and the Analysis of Sacred Symbols” (from which the preceding quotes come) considerably more useful in
thinking about religion and, particularly, the interrelated roles of aesthetics, emotion, and
ethics in religious life.
Mallison 2000, 291–92.
Lawrence 1992.
Nizami 1992, 10–15. Importantly, Nizām al-Dīn also stressed that miracle mongering has no
place in spiritual life and is, in fact, a sign of spiritual imperfection.
Digby 1975, 24.
Schofield 2015, 408.
Schofield 2015, 416; Brown 2006, 62.
Schofield 2015, 410.
Behl 2012b, 34.
In technical terms, Sheldon Pollock explains, “Rasa— and here is the common understanding—is produced when certain ‘stable’ or primary emotions (sthāyibhāva) of ours
are fully developed by stimulation from a suitable object (ālambanavibhāva) under appropriate external conditions (uddīpanavibhāva), and nuanced by more evanescent feelings
(vyabhicāribhāva) that are themselves made manifest by physical reactions (anubhāva). All
this activates our own latent dispositions (vāsanās, saṃskāras) to respond sympathetically” (2001, 208–9).
Miller 1977, 14–17; Behl 2012a, 66– 68.
Pollock 2016, 285–86.
Classically, sṛṅgāra usually refers to an embodied, romantic, sexual passion between two
comparable beings; e.g., a noble hero and heroine. The love between child and parent, disciple and guru, or, most certainly, devotee and God would not typically have been classified
as sṛṅgāra since these are loves between beings quite unlike (incomparable to) each other;
Hamsa Stainton, email messages to author, June 15, 2016. Bhakti rasa, then, was quite different from sṛṅgāra rasa.
Pollock 2016. 302–3.
Behl 2012b, 35–36; Behl 2007, 322; Behl 2012a, 74.
Behl 2012a, 81.
Behl 2012a, 64.
Behl 2012a, 64– 65.
burc19032_1st_i-412.indb 337
—-1
—0
—+1
12/12/18 6:14 PM
338 9 2. sultans, saints, and songs
140. The circulation of Kabīr’s poems clearly illustrates the complexities involved in distinguishing a general bhakti public from a general Sufi public (and in determining the boundaries of
any given public). Poems attributed to Kabīr circulated among Sufi, Sikh, nirguṇ Sant, and
saguṇ Vaiṣṇava communities (as well as in nonsectarian devotional and entertainment contexts). As Thomas de Bruijn has demonstrated, these poems shared images, semantics, and
cultural references that could be decontextualized and relocated into contexts of meaning
specific to particular religious communities, a fact instrumental to the circulation of such
poems “in a field where the boundaries between religious communities were fluid and where
culture was being transmitted and valorized in a dialogic exchange . . . [that] was not
intended to bridge the differences between religions” (2014, 157).
141. Rosenwein 2006, 2.
142. Rosenwein 2006, 25–26.
143. Cf. Finbarr Flood 2009, 5.
144. Cf. Lincoln 1989, 18.
3. Akbar’s New World
-1—
0—
+1—
1. O’Hanlon 2007b, 889.
2. The term “Rajput” is a status title/category to which a variety of local groups, of differing
ethnicity but often with warrior backgrounds, assimilated themselves over time. While entry
of clans into the Rajput status group seems to have been relatively open at first (ca. eleventh
to twelfth centuries), over time genealogical purity was increasingly stressed and by the seventeenth century Rajputs had essentially become a caste. See Tambs-Lyche 1997, 86–87; Kolff
1990, 71–116.
3. Moin 2012a, 21.
4. Moin 2012a, 93.
5. Alam 2000, 229.
6. Alam 2000, 239. Far from a fixed Islamic “orthodoxy,” Islamic legal prescriptions in Mughal
India were subject to considerable interpretation in their application, and judicial decisions were usually made according to local custom, not the injunctions of sharī‘a. See Eaton
2003a, 23.
7. Moin 2012a, 99, 103. This quote comes from the Tarikh-i-Rashidi (1540s), the chronicle of
Babur’s cousin, Mirza Haydar Dughlat, which also says, “Shaikh Pul had donned the guise of
a Sufi master and taught that spells and invocations were the best means to obtain one’s
true desire, and even that one’s true desire should be the attainment of these means. Since
[Humāyūn] had a temperament for such things, he soon became a disciple.”
8. We have clear evidence that the Mughal emperors Akbar, Jahāngīr, and Aurangzeb (and
probably Shāh Jahān as well) were interested in yogīs (and vice versa). We know that Akbar
visited and patronized certain Nāth yogī sites, summoned yogīs to court for lengthy private
conversations, and even took up some of their practices and customs (e.g., performing
alchemy with them and shaving the hair at the crown of his head); Jahāngīr regularly visited a gosain named Jadrūp; and Aurangzeb corresponded with yogīs and purchased quicksilver from them in the 1660s; see Pinch 2006, 33, 51–52, 55. These Mughal interactions with
yogīs should not be understood as instances of some sort of “interreligious dialogue,” for
they were driven by a curiosity and a pragmatic interest in supernormal power and esoteric
burc19032_1st_i-412.indb 338
12/12/18 6:14 PM
.
f. ion
oo ut
Pr rib
's ist
or D
th or
Au t f
o
N
A
Genealogy
of
Devotion
Bhakti, Tantra, Yoga, and
Sufism in North India
Patton E. BurchEtt
Columbia University Press
New York
burc19032_1st_i-412.indb 3
—-1
—0
—+1
12/12/18 6:14 PM
Columbia University Press wishes to express its appreciation for
assistance given by the Dean’s Office at the College of William & Mary,
as well as the Columbia University Seminars’ Schoff-Warner
Publication Funds, in the publication of this book.
Columbia University Press
Publishers Since 1893
New York Chichester, West Sussex
cup.columbia.edu
Copyright © 2019 Columbia University Press
All rights reserved
Chapter 8 contains some material previously published in
“My Miracle Trumps Your Magic: Encounters with Yogis in Sufi and
Bhakti Hagiographical Literature,” in Yoga Powers: Extraordinary
Capacities Attained Through Meditation and Concentration,
ed. Knut A. Jacobsen (Leiden: Brill, 2012).
Chapter 9 contains some material previously published in
“Bitten by the Snake: Early Modern Devotional Critiques of
Tantra-Mantra,” Journal of Hindu Studies 6, no. 1 (2013).
Library of Congress Cataloging-in-Publication Data
Names: Burchett, Patton, author.
Title: A genealogy of devotion : Bhakti, Tantra, Yoga, and
Sufism in North India / Patton Burchett.
Description: New York : Columbia University Press, 2019. |
Includes bibliographical references and index.
Identifiers: LCCN 2018036629 (print) | LCCN 2018043259 (ebook) |
ISBN 9780231548830 (electronic) | ISBN 9780231190329 (cloth : alk. paper)
Subjects: LCSH: Bhakti. | Sufism. | Yoga. | Tantrism. | India—Religion.
Classification: LCC BL1214.32.B53 (ebook) | LCC BL1214.32.B53 B87 2019 print) |
DDC 294.5/43609545—dc23
LC record available at https://lccn.loc.gov/2018036629
Columbia University Press books are printed on
permanent and durable acid-free paper.
Printed in the United States of America
-1—
0—
+1—
Add cover/jacket credit information
burc19032_1st_i-412.indb 4
12/12/18 6:14 PM
Contents
Acknowledgments ix
Notes on Transliteration and Translation xv
Introduction: Tantra, Yoga, and Sufism in
the Historiography of Bhakti 1
Part i
From Medieval Tantra to Early Modern Bhakti
1. The Tantric Age: Tantra and Bhakti in Medieval India 29
2. Sultans, Saints, and Songs: Persianate Culture, Sufism,
and Bhakti in Sultanate India 64
3. Akbar’s New World: Mughals and Rajputs in
the Rise of Vaiṣṇava Bhakti 99
Part ii
Yogīs , Poets, and a New Bhakti Sensibility in Mughal India
4. Between Bhakti and Śakti: Religious Sensibilities Among
the Rāmānandīs of Galta 129
burc19032_1st_i-412.indb 7
—-1
—0
—+1
12/12/18 6:14 PM
viii 9 contEnts
5. Nāth Yogīs and Rāmānandī Bhaktas: Styles of Yoga
and Asceticism in North India 169
6. Agradās and the Circulation of Mughal Bhakti:
Formations of Bhakti Community 195
Part iii
The Devotee Versus the Tāntrika
7. Yogīs and Tantra-Mantra in the Poetry of the Bhakti Saints 239
8. The Triumphs of Devotion: The Sufi Inflection of Early Modern Bhakti 276
Conclusion: Bhakti Religion and Tantric Magic 305
Appendix: List of Manuscripts Containing Compositions by Agradās 313
Notes 317
Bibliography 389
Index 000
-1—
0—
+1—
burc19032_1st_i-412.indb 8
12/12/18 6:14 PM