05: A Substantive Intervention

Chapter 5: A Substantive Intervention: Reading the Material Culture of the Western Deccan Across Region and Dynasty

More than one hundred miles from the Cāḷukya capital at Kalyana in the district of Bhoom, Osamanbad, lies a small village of scarcely ten thousand people. Despite its remoteness, it remains for certain caste communities an important site for pilgrimage, attracting devotees from as far away as Pune. Though in recent years the tradition has been emended to conform to the schedules of working people participating in a capitalist economy, it has long been customary that when a male child is born—on the sixth day he is alive—his parents would take him before Mother Sātwai, the goddess of the six.

Sātwai Devī in situ on the side of Māṇakeśvara Temple.

Sātwai Devī’s duties include inscribing the fate of the child upon on his forehead, a task she carries out on the midnight of his sixth day. Indeed, there seems to be some sense that through preemptive propitiation, the goddess of the sixth day can be coaxed into adopting a more magnanimous attitude in fixing a child’s destiny. For this reason, before a stone smeared in turmeric and vermillion, families come to Sātwai Devī with offerings of liquor, child size bangles, and, traditionally, a goat or chicken.

[[P592]]

As the careful reader of Kiss of the Yogini has by now probably recognized, from the perspective of a historian of religion, the goddess Sātwai¹ as found at the Māṇakeśvara temple in Bhoom represents a thinly veiled transposition of the pan-Indic goddess Ṣaṣṭhī, also primarily a goddess of childbirth, with whom our Devī even shares her name. Already described in the Mānavagṛhyasūtra, where she is identified with “Śrī, the great goddess of royal sovereignty,” as White aptly notes, Ṣaṣṭhī has been worshipped by “nearly every Hindu in India. . . since Kushan times” and thus “hers is no more a ‘folk’ cult than those of the great Goddesses Durgā or Lakṣmī.”² So, in a general sense, Sātwai’s turning up somewhere within

[[P593]]

the religious social imaginary of the rural Marathawada is hardly unexpected; her veneration is effectively part of a perennial religion. With such knowledge already in hand, when we ask, “where does this goddess come from” and “what is she doing here,” our task is not that of reconstructing intellectual genealogies. When we speak of “here,” what we are setting out to locate is not “the place of a cult within the culture.” Instead, we are posing a question much more like the ones asked by spatial geographers. If one carefully attends to the larger material context in which this particular image of the goddess of the sixth day is situated, it soon becomes evident that, for Sātwai, the “here” where she resides, rather incongruously, is on the side of an early medieval temple complex.

Left side of Māṇakeśvara temple. The image today worshipped as Sātwai Devī is on the far left, marked by vermillion run-off. The garbhagṛha is under the dome.

Under uncountable layers of vajralepa, this Sātwai almost certainly commenced her own life as a largely incidental decoration, most likely not even the focus of any sort of worship, embedded within a wider and distinctive religious field.

[[P594]]

Blurring the imagined lines between “folk and elite religion,” between practices kept in the family and those exposed to a range of publics, the application of this particular religious imaginary at this particular place poses a different sort of puzzle.

The village of Mankeshwar is named after a single rather spectacular temple,³ which, in the usual fashion, is understood by the local people as executed in the “Hemadpanti” style.⁴ Apart from the fallen śikhara, which has been replaced by a dome and pinnacle from the Maratha period, the structure, though defaced in places, is mostly intact. Despite the assigned moniker, simply on stylistic grounds Māṇakeśvara is indisputably a product of Kalyāṇi Cāḷukya period workmanship. Fortuantely, new evidene, in the form of an unpublished inscription in old Marathi located inside Māṇakeśvara’s maṇḍapa puts any further doubts about provenance to rest.⁵ While both the condition of the text as well as the style of [[P595]] inscriptional shorthand in which it is composed pose many difficulties, the identity of the patron in question, “the great heir to the throne” (yuvarāva maṭhīṃ) is thankfully not among them. As the date and listed title make plain, this is a śāsana (sāsana) issued by the Cāḷukya crown prince Mallikārjuna, the eldest son of Vikramāditya VI. As his fate would have it, Mallkārjuna predeceased his exceptionally long-lived father and thus never assumed the throne. Perhaps the Cāḷukya emperor would have benefited from a visit to the goddess of the sixth day.

yuvarāva sāsana maṭhīṃ yati (2 akṣaras missing) saghaṭa-pasime liṅgīṃ kāyā dārūpāla datena (sma) tribhuvana sailaṃ / [*] da-maṅgalā drumagaja manā. . . . /

[[P596]]

This chapter sets out to show that, even when our intellectual focus is thinking with nuance about a mostly long departed medieval world, texts are not the only sources we have for reconstructing contexts. Indeed, when it comes to posing questions not about ideas themselves, but about relations and interactions, or the play of meaning as it manifests within a lived multifunctional spaces and landscapes, sometimes texts are in fact the wrong sort of tools for our thinking. In this spirit, our present discussion offers three case studies in which textuality is largely decentered, as hermeneutics gives way to a phenomenology of space. To put the matter in simpler terms, we will be thinking about space, its impact on people and things, and how it can serve as a vehicle for instantiating values. Through taking space and not text as our context, like any other viewer, we are forced to confront what we tend to think of as isolated objects of knowledge or distinctive places as always already existing within a multitude of relations.

Much as we saw in chapter two when we treated texts as material objects, the translation of protean ideas into delimited physical forms that come to us already supplied with particular affordances leads to a reassessment of the range of often transactional functions that particular spheres of meaning evoke in daily life. However, instead of thinking with texts, however utilitarian, our main focus will be material culture. The medieval Deccan is home to some of the most abundant and also the most poorly documented temple complexes and sculptural traditions on the subcontinent. Most of these have never been the subject of academic study within a Western context. Apart from texts, along with even more thoroughly neglected archaeological evidence, they make up virtually all of what survives of early medieval culture. Despite the fact that, as we will see momentarily, such things represent probably the largest and, in a certain fashion, most readily contextualizable archive that is largely the product of a Tantric social imaginary, in thinking through the history of pre-Islamic religious tradition, they have scarcely received more than a few footnotes.

[[P597]]

As corrective to such tendencies, the present chapter offers three interrelated case studies. The first, with which we will continue in a moment, engages with the temple complex at Māṇakeśvara as an exemplum of how an unabashedly Tantric ritual complex is embedded in space in a manner that translates the types of epistemic values we have been exploring into the medium of stone. In the second, through chasing images Bhairavas closely linked with the perpetuation of royal power across the Deccan, we will see how iconographic transformations and spatial transpositions over the longue durée begin to suggest a very different story about the changing place of cremation ground culture within society. Finally, as prelude to chapter 6, which will offer an insider’s perspective of how explicitly Śākta values animated the types of social agents who produced our surviving material culture, we will travel from the ruins of the Cāḷukya capital at Kalyāṇa to the workshops of one the artisan communities that made the material culture.

Irreconcilable Images in Irresolvable Space: The Changing Cāḷukya Temple

Though the hundreds of near life-size images that adorn the outside of Māṇakeśvara are exceptional in their execution as well as essentially undocumented, the temple’s most unexpected and noteworthy feature is far more inconspicuous. Indeed, captivated by the more vivid parts of Māṇakeśvara’s iconographic program, as you circumambulate the temple— perhaps in a counter-clockwise fashion—most likely you will miss them entirely.

Mankeśvara, Maharashtra: Seated images of goddesses encircle the entire circumference of the Māṇakeśvara temple. .

[[P598]]

Small, seated goddesses, bearing a range of accoutrements—encircle the entire circumference of the temple, all of them placed at about the level of a person’s thighs. Should you take the time and tally them up, they number sixty-four. In other words, despite initial appearances, Māṇakeśvara, a temple constructed at the express direction of the heir apparent to the Kalyāṇi Cāḷukya throne, is dedicated to the sixty-four (catuḥṣaṣṭiḥ) yoginīs.⁶

[[P599]]

In the scholarly imagination, we tend to envision this quintessential expression of Tantric material culture—the yoginī temple— as a distinctively hypethral and circular structure placed on top of hill at least on an elevated plane. At places like Jabalpur in Madhya Pradesh or Hirapur in Orissa, while the external walls of the temple are nondescript, once one crosses through the gateway into the temple, the visitor is virtually assailed by the visceral highly individuated renderings of yoginīs. These deities are installed all around the internal walls of the site, facing inward and with the intention of encircling a free-standing image of Bhairava, in most cases no longer extant. Thus, outside the Deccan, the more familiar instantiations of the yoginī temple instantly present themselves to the observer as hyper-differentiated spaces, utterly dissimilar from the common temple in both form and function. The absence of such structures in the southern Deccan has generally been taken as an indication that the worship of yoginīs, and with it the Śākta social imaginary, never really found a home in the region.⁷

External wall of Māṇakeśvara: Baṭuka Bhairava and two female attendants. One is dancing, kapāla in hand.

[[P600]]

Produced under royal patronage, at the direction of the scion of the most influential ruler of his day outside the Tamil south, the temple at Māṇakeśvara offers a substantive rebuttal of such working assumptions. Indeed, when we reassess the material record, we find that that—while this sort of yoginī temple itself seem to be mostly a Maharashtrian feature, there is a virtually limitless canon of images and structures pointing toward a distinctly Deccani take on the Śākta past hiding in plain sight.

Detail: Baṭuka Bhairava (left), Bhairavī (right). External wall exterior of temple at threshold left and right side.

[[P601]]

At Māṇakeśvara, Bhairava and his consort Bhairavī occupy the place prescribed to them by scripture: the heart of the maṇḍala, encircled by sixty-four yoginīs. Indeed, the couple actually appear twice. First, expertly crafted images of the two deities are found on the exterior of the temple. At precisely the same point in the floorplan, respectively positioned on the left and right side of the temple, each is placed aligned with the other on the external wall of the threshold marking the entrance into the garbhagṛha. Additionally, albeit by a much less gifted hand, variations on the same images reoccur inside the temple, placed respectively on the right and left pillars flanking the entrance into the threshold-space before the garbhagṛha.

Temple interior details: Baṭuka Bhairava and Bhairavī.

[[P602]]

As we shall soon see, the changing presentation of this iconographic program in lived space has its own quite recoverable history intimately linked to the visual language assumed by the powerful in the Deccan by most political actors from at least the seventh century. What is distinctive about the Kalyāṇi Cāḷukya approach is precisely that these temples embed their focuses of worship in the midst of a play of disparate images that carry with them a competing array of semiotic referentiality.⁸

Details on the exterior of Māṇakeśvara. Top left: Woman with a drum and royal woman interspersed with images of Viṣṇu and the gods of the directions. Top Right: Piśāca with a curved knife in the middle of representations of Mahiṣāsuramardinī and Andhakāsuravadha. Bottom Left: Ascetics initiating a devotee. Bottom Right: woman with a mirror.

[[P603]]

Left side of the Umāmaheśvara temple at Umapur near Kalyana. Bhairava is in the alcove.

In considerable contrast, at the pre-twelfth-century Kalyāṇi Cāḷukya ritual center at Umapur in Bidar district, ten miles from the royal capital, this rich play of varied imagery is not yet in evidence. In its place, we find a structurally much simpler though quite elegant facade executed in a style somewhat reminiscent of that of the late Bādāmi Caḷukyas. The Umāmaheśvara Temple, which bears this name because it houses an image of Śiva and Parvatī

[[P604]]

in place of the expected liṅga in its garbhagṛha, is much smaller in scale than Māṇakeśvara and almost entirely free from ornamentation and figurative iconography. In place of hundreds, only three figures adorn its outer walls, each of them positioned flanking the three exteriors that encircle the garbhagṛha, Unmistakably, two of these images are virtually identical to the representations of Bhairava and his consort as they appear several hundred years later in the interior of Māṇakeśvara right down to the gesture of the goddess raising her finger to her lips.

Bhairava and Bhairavī from the left and right alcoves at Umapur.

Yet if the images themselves are iconographically identical, the visual fields in which they are embedded foster very different sorts of impressions. The bare surfaces at Umapura, though imposing, impose no specific cognitive content in the mind of a visiting devotee. The exterior of the temple has only three visual focal points to which the design directs the eye. All three of the deities face outward towards the observer and each of them is positioned at exactly [[P605]] the same height on the exterior of the temple. In this way, the worshipper is straightforwardly oriented, presumably to a program of pūjā centered around the veneration of three deities on the exterior as prelude to the worship of the central image of Umāmaheśvara. Though just ten miles from the palace, Umapur is situated on the outskirts of a network of religious centers linked to the Kalyāṇi Cāḷukya capital and seemingly intended for specialized worship conducted by the kings. It is, in other words, a space conceived of with a single focused purpose and highly limited audience.

Exterior courtyard at Umapur. Royal portrait of unidentified Cāḷukya ruler.

At Māṇakeśvara, in contrast, scarcely a surface is bare. Much like their counterparts in Europe, the Indian sculptor deployed a compositional strategy where each individual image [[P606]] was not only framed in relation to architectural elements and the surrounding negative space but also with the intention that there was a single optimal place for the viewer to stand if the individual image was to have its intended impact. Indeed, in many cases approaching the image from other angles would present to the viewer a distorted image. What is so unnerving and intentionally overwhelming about a temple like Māṇakeśvara is that there is no perfect vantage point for viewing the temple as a whole. To align oneself as intended with one image is of necessity to become disaligned with many others. Like in a symphony by Schumann, though the melody is compelling, one is left waiting and waiting in vain for resolution—for that anticipated return to a tonic key that never ever comes. In short, the temple offers a visual idiom that resists and rejects the possibility of reduction to an essential element, form, or point of view.

Temple exterior: Māṇakeśvara.

In place of the outward facing images at Umapur, each offered in isolation and directly connecting to the viewer, at Māṇakeśvara, the eyes of images gaze upward, offer obeisance, [[P607]] engage with the enemy, entice a partner, or cast at you a coquettish side long glance. We are forced to witness a play of technical perspective in which we are confronted in a single visual field by the presentation of multiple seemingly unconnected perspectives. Each image or grouping of image abides in its own reality. Each is attuned to a particular affective response. Thus, while this image of a mother cradling a nursing toddler who is tugging eagerly at her breasts and a skull-bearing guardian figure in concrete terms occupy adjacent physical spaces, they offer a jarring juxtaposition in terms of their affective space.

Māṇakeśvara: Detail of exterior wall.

[[P608]]

As you might have noticed, in this image, the plinth above the deity to the right of the nursing mother is crowned by a relief image of the śikhara of a simple single shrine temple in the nāgara style. Indeed, it is precisely the same sort of structure that adorns the top of the temple at Umapur. The great scholar of medieval Indian architecture Adam Hardy has offered a convincing explanation of the logic that animates these structural choices:

For me the key is the aedicule, defined as an “image of a building used as a compositional element” (Glossary, 244). Much of Indian temple architecture is not merely “decorated” with the “small models of large buildings” observed by James Fergusson, but composed entirely of temple-images that appear at a whole range of scales. . . . Aedicules are not arbitrary assortments but representations of familiar shrine forms. . . from the matrix of temple walls. Underlining the fact that they are meaningful entities, these various kinds of shrine-image appear not only as primary components, but also at lower levels of the compositional hierarchy, as niches and so on. . . . In many traditions the relationships between the aedicular components of a temple design increasingly express movement. More than just “cascading,” interpenetrating forms emerge one from another in emanatory sequences. . . . Parallels and connections can be shown between temple forms and, on the one hand. . . religious and philosophical ideas. The congruencies between these realms are vastly more apparent once the aedicular structure of temple designs is recognized. For example, in terms of religion, the simple idea of one god with many aspects is made visible when one sees how their many abodes are linked into a heavenly palace. If a temple is an image of the unfolding universe, a symbol of divine and cosmic manifestations, the aedicules are its disaggregating units.⁹

Hardy’s adept interpretation centered on the aedicule, an “image of a building used as a compositional element,” has considerable explanatory power, and yet it overlooks an important practical and conceptual dimension of a temple like Māṇakeśvara. For right alongside representing “an image of the unfolding universe,” the building block of which is the single free-standing shrine endlessly replicated in space, our temple also presents an unspecified quotidian domain, the sort of space where, without the elegant framing of nāgara [[P609]] śikhara, a young mother might pause momentarily—perhaps too long for the little one’s liking—and hold her own space before once again nursing her child. The peculiar genius of the Cāḷukya visual idiom is that for the first time, it allows for the mingling of such disparate sensibilities held in dynamic tension in shared but irresolvable space.

Where is Bhairava? Or, The Changing Place of Power in the Deccan

While the Imperial Guptas still ruled, however tentatively, in Vidiśa, far to their south in the district of Bagalkot—even before they attained their independence from the Kadambas—the Bādāmi Caḷukyas were already beginning to build temples.¹⁰ By the early fifth century, archaeological traces point to the existence of numerous structures made of bricks. Though little unambiguous evidence exists to take us beyond speculation, these were perhaps comparable to the virtually contemporary temples of Uttarēśvara and Trivikrama, each executed in a different style, that miraculously still stand in the town of Ter on the edge of the Marathawada.¹¹ To this very limited and in many ways heterogeneous corpus, one might well add the few remaining Kadamba structures. Most of these are associated with the founder of the dynasty, Mayuravarma (r. 345–365), a Brāhmaṇa who famously responded to the disrespect directed at him when he went to continue his Vedic studies at Kanchi by taking up the sword and founding his own dynasty.¹²

[[P610]]

Prāṇaveśvara at Talagunda in Shimoga district, Karnataka. Circa fourth century.

[[P611]]

The temple of Prāṇaveśvara he had built at Talaguna, for example, apart from the representation of door guards in relief, resembles a large stone box.¹³ For our purposes, what is most pertinent is the formal simplicity of these buildings, their focused iconographic program, as well as their location. Where images are still in evidence, they contain either a four-armed standing Viṣṇu or an immense śivaliṅga. Furthermore, the archeological record at both Ter and Talagunda tells a story of almost continuous habitation in the region for many centuries predating the building of these structures. In other words, in both cases these are temples that, apart from their basic religious affiliation, communicate no particular array of messages to a visitor. In each instance, the temples are situated in close spatial relationship to the everyday world.

A possible counterexample ascribed to the Kadamba period is found at Chandravalli in central Karnataka, at the site where Mayurvarman took refuge upon his return from the Tamil country.¹⁴ Here too patterns of habitation allow us to document human occupancy of much of the surrounding terrain well back into the prehistoric era. Alongside the banks of the main lake, as one ascends up into hills, however, one finds two very different traces of the Kadamba past. Right alongside Mayurvarman’s most famous inscription, the subterranean cave temple of Ankali has functioned as a Śaiva maṭha, complete with its own shrines and library, since at [[P612]] least the fourth century. Even more intriguingly, at the top of the hills, there stands a small box shaped shrine. Dedicated to Bhairava. it contains an image of the deity that is not altogether dissimilar from our few surviving traces of the Kadamba sculptural tradition. If it is authentic, the earliest trace of this sort of religious system in Karnataka, where the veneration of Bhairava forms a part of royal religion but is placed at some distance from multipurpose space intended for diverse publics, may well be the product of the fourth or fifth century of the Common Era.

Map of the heart of the western Bādāmi Caḷukya imperial landscape.

[[P613]]

Should we choose, however, to discard the alleged Kadamba Bhairava as spurious, our story would instead begin, among the Bādāmi Caḷukyas themselves in the mid-sixth or early- seventh century. At this time, the Bādāmi Caḷukya capital was still located in town of Aihole. Like all of the eventual locations (Badami, Pattadakal) which serve as dynastic capitals in the coming centuries, Aihole is situated in the verdant Malaprabha Valley wedged between hills made of sandstone, an ideal strategic position for resisting outside invaders.

Given that this region contains one of the richest arrays of archaeological sites on the Indian subcontinent, many of them set against some of the most compelling landscapes in the Deccan, that the mid-seventh-century temple now called Nāganāthakŏḷḷa should have attracted very little attention is hardly surprising.¹⁵ Located in an out of the way corner of the Malaprabha Valley, even today to reach the temple one must pass through dense jungle. It was in this remote place however, that archaeologists associated with the Archeological Survey of India recovered a collection of tantalizing fragmentary images.

ASI Field Museum at Aihole; Left: Lower torso of seated goddess.Right: Larger than life size bust of Bhairavī. c. 650–700. Original location in Malaprabha valley unknown. Mid-seventh century.

[[P614]]

ASI museum, Aihole: Bādāmi Caḷukya skeletal. c. 650–700.

[[P615]]

The image above hardly does justice to either the grandeur or grotesque quality of this representation of Kaṅkāla Bhairava. Seated, the god is nearly six feet tall. Were he to rise he would likely tower over us. Recovered from the same location, all of them seated or squatting, the Aihole Field Museum also preserves the heads and torsos of the seven mothers, rendered in a similar style, which in situ would have encircled our macabre cremation ground couple. Plausibly related to the temple, which would have housed this pantheon, we also find a fragmentary panel from the same period. This appears to depict two distinctive though perhaps related scenes.

ASI museum, Aihole: Bādāmi Caḷukya skeletal Bhairava panel. c. 650–700.

On the right, a devotee performs worship before a large Śaiva liṅga of the sort we find still installed in many temples from the period. On the left, in contrast, an emaciated and horrific entity bearing what may be a long scalpel prepares to make a sacrifice. Depending on whether the entity in question, plausibly identified with Kaṅkāla Bhairava, for the head-dress and body [[P616]] are similar, was imagined to be as massive as our mūrtis, the offering is either a young child, or perhaps a full-grown human.[^616_620]

To make sense of such fragments, we must travel a considerable distance to a rather unexpected destination: the caves at Ellora, more than two hundred and fifty miles to the north- west in the heart of the Marathawada. In the modern imagination, Ellora is virtually synonymous with the Kailāsa Temple (cave 16), which was likely constructed by the Rāṣṭrakūṭa King Kṛṣṇa I (756–773 CE), perhaps with some contributions from his predecessor Dantidurga. By that time, however, most of the hundred caves at Ellora has already been excavated centuries beforehand. During the period with which we are concerned, the Kalachuris of Mahiṣmatī as well as the Bādāmi Caḷukyas seem to have been active contributors to developing Ellora’s religious ecology.¹⁶

[[P617]]

Detail: Indrāṇī, Ellora, Rāmeśvara cave.

At three of the caves (Rāmeśvara, Daśāvatāra, and cave 25) we find the very same eerily familiar—indeed downright eerie in general—iconographic program is at work. In each case, when you enter into the complex, the figures are to your right, but always situated in a manner so that they are scarcely visible from the main hall and are generally kept out of the light.

[[P618]]

B. Photograph courtesy of the ASI: The full panel of Māṭrkās, Ellora, Rāmeśvara cave.

At two of the sites, the visual anchor of the space is a standing skeletal Bhairava accompanied by an equally emaciated consort. To their left, one finds a panel of mothers. In each case, they are represented in a sensuous style, individually differentiated in their dress, assets, age as well as personality. The idiom is courtly, akin to what one finds in the renderings of the gods in the rest of the complex, and the scale is human.

[[P619]]

Bhairava’s representation, in contrast, is jarringly different. In the Rāmeśvara cave, raised up from the ground, the ten-foot-tall skeletal Bhairava towers not only above human viewers, but also the gods who have come to pay him homage and the ghouls that seek refuge by clinging to his skinless lotus-feet. His wife Bhairavī, though the same sort of creature, and thus bigger than the other figures in the scene, is positively petite beside him. Clearly the likely unrecoverable theology that underwrote this tableau belonged to a register of the Bhairava Siddhānta where the terrifying lord was still in charge with the goddess treated as complement or after thought.

Back right corner, Daśāvatāra cave, Ellora.

[[P620]]

A similar scene repeats itself at Daśāvatāra, where again a somewhat genial looking skeletal Bhairava, is accompanied by a retinue of mothers and a Gaṇeśa, who standing would just about reach his navel, as well as his doting skeleton wife. Here the images are not merely inconspicuous, but reside in the back corner of the cave in near total darkness, and in fact can only be perceived at all as one nears the completion of circumambulating the liṅga at the cave temple.

Detail: Kaṅkāla Bhairava, Daśāvatāra Ellora, Cave 25. Panorama view of the mothers. The seated Bhairava is on the far right.

[[P621]]

Ellora, Cave 25: details of Bhairavī and Bhairava.

Finally, at the unnamed cave twenty-five, in a secluded cave on the right side of the entry to the site segregated from the rest of the complex, while the full array of the mothers as well as Bhairava and his consort are all depicted seated, once again a skeletal Bhairava and his consort are set aside from the rest and represented as considerably scaled up from their retinue. Thus, despite some variations in style, across a temporal span of a least several hundred years and geographical region comprised of at least several hundred miles, a distinctively cremation ground culture religious sensibility—centered around Bhairava, his consort, and the mothers— formed a consistent part of the idiom with which the powerful portrayed the sacred. From the sixth to eighth century, whatever its particular semantic content, set aside in space and style of presentation from other representations of the gods, Bhairava and his consort continued to be depicted in the same general manner, right down to their headgear and hats.

[[P622]]

Something different starts to happen among the Rāṣṭrakūṭas, who appear have held slightly different commitments. At Kailāsa itself, the gallery of the mothers once again is situated upon an elevated platform on the right side of the complex, positioned in a manner so that until one ascends most of the way up the stairs the figures remain invisible. As one reaches the landing, the viewer encounters representation of mothers in keeping with the thematic and

Gallery of the mothers at Kailāsa, Ellora.

stylistic elements we have already explored. Something is conspicuously absent however from this arrangement, for in the place where we would anticipate seeing Bhairava and his consort, seated on a throne made of corpses and ghouls there is only a single figure of somewhat indeterminate gender. Indeed, even in situ it is difficult to determine if the sizeable genitalia that hang over the deities emaciated leg are a part of its body or the corpse throne on which it sits. If it is the former, this may be an image of the skeletal Bhairava. Perhaps his consort has [[P623]] been demoted. If it is the latter, Bhairava has vanished from this scene, and in his place, we observe all the female pantheons that increasingly emerge with the later Śākta Tantras. Elsewhere in a distinctively Rāṣṭrakūṭa milieu, however, one encounters more substantial indications that the reinvention of this particular religious idiom, with Bhairava and the goddess becoming decoupled from each other without either fading in significance, is already well underway. The Rāṣṭrakūṭa capital at Mānyakheṭa was razed to the ground at least twice, once by the invading Cōḻa army and then again in the sack of city carried out by the Kalyāṇi Cāḷukya king Taila that brought his dynasty into power, so it is hardly surprising that nothing of substance remains. However, as my colleague Saili Palande-Datar has pointed out, drawing upon her work assisting Arunachandra Pathak, substantial Rāṣṭrakūṭa remains survive at their secondary capital at Kandhar, in Nanded District, Maharashtra, close to the Telangana border. To do the material justice would require a more extensive study than can be carried out in the present context. [[P624]]

Kandhar: Goddess encircled by seven mothers.

As a matter of fact, however, at Kandhar, Bhairava¹⁷ has not disappeared. Perhaps because he had outgrown his surroundings, he has simply been relocated. Apart from several bronzes depicting Śiva and Parvatī, Jain, and possibly Buddhist images, the Kandhar horde contains numerous individual life-size depictions of the mothers analogous to those found at Ellora. The repetition of several of the same figures suggests that multiple temples dedicated to the mothers must have existed in the vicinity. In contrast, the unidentified representation of supreme goddess surrounded by mothers in miniature (image left) offers another possibility: an [[P625]] all-female pantheon.

Top: Colossal head of Bhairava at Kandhar. Below, other parts of his body given to scale.

Nor was such a colossal image, the bridge of whose nose is a tall as a man, something unique to the royal center at Kandhar. At least six such figures have survived into the present day, though each is in a fragmentary form, and their locations are scattered throughout the western Deccan. In fact, almost two hundred miles to the north, outside the domain of Marathawada in the region of Buldhana in the town of Anandpur, again found in the company of separate representations of the goddess and the mothers, we find a substantively bigger image from the Rāṣṭrakūṭa period of the very same sort, though in this case the head has sadly been destroyed. [[P626]]

Remains of colossal Bhairava at Anandpur.

When such images are encountered at a remove from their original interpretive context either in the form of entextualized discourse or the interrelated spatial field where they were once embedded, an opportunity presents itself for ungrounded speculation and over- interpretation. Resisting such inclinations, let us stick with a minimalist even conservative reading of the evidence at hand inflected by our earlier explorations of related issues. From the sixth to the mid-ninth century, in a range of sites found across the western Deccan between the Narmadā and the Kāverī, “Hindu” dynastic powers articulated a vision of religious life in which a sizeable place was set aside for terrifying powers associated with the Bhairava Siddhānta. Representations of these figures were executed in a distinct visual and spatial style. As we have as seen, this was inseparable from the idiom of cremation ground culture. Considering the amplified scale of the central images and its recurrence in settings where royal power is everywhere else on display, for at least three hundred years, this theology was somehow intimately linked with a theory of power. Contrary to what we would expect from [[P627]] the textual record, which until very recently read the emergence of the worship of Bhairava as a tertiary subordinate phenomenon sparked by contact by between the purveyors of mainstream Hindu mythology with either tribal people or marginal communities,¹⁸ here Bhairava worship from its inception reads as an elite phenomenon. Instead of operating as an accrual that attaches itself to a pristine Śaiva base rather late in the historical record, which is precisely the sort of model implicit in the periodization of the scriptures presented by Alexis Sanderson and his disciples, if anything the evidence is pointing in the opposite direction. As mainstream Śaiva institutional life works its ways into the halls of power, it finds itself having to adapt to accommodate an archaic and originally perhaps entirely independent tradition of Bhairava worship intertwined with the performance of kingship. This dynamic seems to have been operation at least since the fourth century, for this is when the Vākāṭaka ruler Rudrasena I proudly introduced himself as bearing the title atyantasvāmīmahābhairavabhakta, and when the kings of Valkha, antecedents to the Kalachuris of Mahiṣmatī, as well as the Kadamba King Mayūravarman all within a period of a hundred years proclaimed their allegiances to the Terrifying Lord across a terrain encompassing most of the western Deccan.¹⁹ In each of these [[P628]] cases, the accommodation of this tradition animated by a divergent set of values, which the visual record would seem to suggest included the incorporation of at least occasional acts of human sacrifice, seems to be accomplished through acts of segregation that delineate a separate spatial and conceptual space where such values can be constructively performed.

From the mid-ninth century onward, in contrast, we begin to witness a somewhat different dynamic. Correlated it seems²⁰ with the rise of the Kalyāṇi Cāḷukyas, traditions of Bhairava veneration and their unmistakable iconography are integrated first into shared and then into multifunctional spaces. As we will see, during the reign of Vikramāditya VI, the visual idiom of the cremation ground culture is reimagined as one among many components of the sculptural program that adorns a temple. Instead of placing Bhairava and his consort in their own distinctive caves or temple sites situated on the edges of the social world, they are assigned a liminal spot within the microcosm of the temple itself.

[[P629]]

In general, from the early tenth up through the thirteenth century, we see several different arrangements at play concurrently within the same regions. In the first, large images of Bhairava, sometimes paired with his consort, are placed in a separate shrine alongside the main temple but within the same complex. In the second, smaller decorative or relief images of Bhairava and his consort are integrated into liminal spaces, often tucked in the midst of a more standard iconographic program depicting scenes from Indian mythology. Such figures could be carved on the exterior of the threshold before the garbhagṛha, an arrangement we have already seen at Māṇakeśvara. Small images could be also installed inside the temple in the niches before the entrance to the threshold of the shrine, as in Raṭṭa royal temples, as will soon become evident.²¹ They could be engraved on the pillars nearest to the main shrine, as occurs in the Mahādeva Temple at Itagi. Or, finally, especially in evidence in the corpus of Hoysala architecture that is transparently the product of Kālamukha patronage, Bhairavas are installed in elevated positions on the śikhara of the temple itself.

Two features of this pattern of production are particularly essential for understanding what is new about the greater Cāḷukya world, whose conceptual frames structure social and religious realities from the mid-ninth to mid-thirteenth century throughout the Deccan. The first is that there is a retreat from the unapologetic representation of Bhairava and his retinue through the lens of his terrifying and utterly non-human strangeness that animates the earlier images of enormous emaciated Bhairavas, purposefully setting them apart from the more anthropomorphic sensibility used to represent the rest of the gods. If Bhairava is to share space with the rest of the religious imaginary, it must happen at the expense of muting his distinctiveness, except perhaps in spaces in the temple not easily accessible to the eyes of the [[P630]] majority of visitors. Equally importantly, at the very same moment that he is integrated into the general religious imaginary, images of Bhairava become ubiquitous, receiving veneration and patronage from virtually the entire spectrum of society. Indeed, it is precisely this proliferation of images of Bhairava and his retinue that distinguish the temples of the early medieval Deccan from most of the subsequent and previous iterations of Hindu temple culture. For our purposes, a few brief engagements with select examples illustrating the range of the phenomena and its occurrence in different contexts will suffice.

[[P631]]

[[P632]]

1. Tumkur, Karnataka, ca thirteenth century, five feet tall 2) Khopeshwar, Maharashtra, eleventh century, four feet tall 3) Khopeshwar, Maharashtra, tenth century, six feet tall 4) Hooli, Karnataka, tenth century, five feet tall.

Top left and right. Eight feet tall, twelfth century, ten feet tall, early eleventh century. Warangal, Kākatīya capital, Telangana.

[[P633]]

Bottom left, twelfth century, Belur, Karnataka, three feet tall, Udupi district, thirteenth century, four feet tall.

[[P634]]

Munavalli, five feet tall. Commissioned in 1226 CE by the Seuṇa Yādava King Siṅghaṇadeva II as a gift for his rājaguru Sarvēśvara.

[[P635]]

The gallery you have just viewed, chosen from my photos almost at random, encompasses images from the territories of what are now three different states. In some cases, these images are separated from each other by a distance of more than five hundred miles. And yet, despite the fact that they were produced in many cases under the patronage of dynastic powers engaged in bitter rivalries and open conflict with each other, they are all quite clearly products of the same iconographic conventions intended to evoke analogous religious visions. As we can see, with the notable exception of the piece from Hooli, the product of an earlier sensibility that persisted among the now much diminished Kadamba kings, in each of the images, Bhairava has been further anthropomorphized. His frame has filled out, sometimes to the point of corpulence, and he has taken on some of the more general trappings associated with warrior classes in the Deccan. These same traits are also in evidence in the placement of smaller paired representations of the deity and his consort at the heart of a small temple, a development best exemplified by the royal temples of the Raṭṭa dynasty. Though the Raṭṭas of Saundatti claim descent from the Rāṣṭrakūṭas, they belong to a line that traces its origins not to Manyakheṭa, but rather to Latur in southeastern Maharashtra, several hours from the Cāḷukya capital at Kalyāṇa.²² Exactly what sort of circumstances drove them to relocate the dynasty two hundred miles to the south in the center of Karnataka remain obscure, but by the early tenth century, we find them established alongside the bank of lake now called Renuka Sagar, less than an hour to the south of Shirshangi. The clan goddess of the Raṭṭas, who resides on a hill just outside of the city, is today understood to represent

[[P636]]

the transregional non-Brāhmaṇical “folk” goddess Yeḷlammā or Reṇukā Devī. Saundatti serves as the main ritual center of the tradition, and the place is closely associated in the popular imagination with traditions of so-called sacred prostitution and transgender Jogappa devotees. Indeed, Yeḷḷammā, whose worship is prevalent throughout the Deccan, is often perceived as goddess especially associated with the Dalit castes.²³ In light of this dominant narrative, which seeks to locate the goddess in an exclusively Dravidian and subaltern milieu, it is somewhat disconcerting that Yeḷḷammā is housed in an imposing eleventh-century temple executed in a typical Cāḷukya style. Indeed, both the inscriptional and archaeological record, while offering no traces of her subaltern histories, suggest that in the tenth through twelfth centuries, the temple on the hill was ritually linked to two smaller structures integral to the ritual life of Raṭṭas, both of which are located in their capital city and each of which embeds in situ an inscription documenting the conditions under which the temple and its pantheon were installed. Scarcely large enough to fit five people, the temple of Puradhēśvara is situated underground and today lies sunken down in the middle of a working- class neighborhood.

[[P637]]

Above: The Saptamātṛkās over the door to the garbhagṛha at Puradhēśvara, Saundatti. Below: Bhairava and Bhairavī in the left and right niches of Puradhēśvara temple, Saundatti.

Several hours to the west, at Itagi, a Pāśupata ritual center just a few towns over from Kukkunur we find the integration of Bhairava and his consort into shared public space is carried out more surreptitiously. Generally acknowledged as one of the first masterpieces of the “new” Kalyāṇi Cāḷukya style, the Mahādeva Temple, called the “Emperor of temples,” was built by Mahādeva Daṇḍanāyaka, a favored military commander of Vikramāditya VI, whose labelled portrait²⁴ adorns the canopy surrounding the pūrvaraṅga. The temple of Mahādeva was established in 1112 CE in the middle of space that had seen uninterrupted occupation by self-identified Pāśupatas for at least several hundreds of years beforehand.²⁵ The grounds where the temple was to be built already housed a series of shrines, some likely dating back to the eighth century (or even earlier), where Pāśupata yogins had taken samādhi, as well as a maṭha where the community resided, complete with a ghaṭikāśāla dedicated to the teaching of the Vedas and the śāstras. In other words, Mahādeva commissioned a temple in his name in a space with its own longstanding institutional culture that for centuries had been under the

[[P638]]

purview of a Śaiva community with an atypically Vaidika and Brāhmaṇical orientation.

Left: Itagi, Bhairava and two ghouls. Right: Bhairavī, surrounded by piśācas, left, and rearing dog, right. differences in coloring are a product of different camera filters.

It is thus a testament to the pervasiveness of the religious sensibility we are beginning to recover as well as its liturgical centrality in the medieval Deccan that even here, Bhairava and his consort as a framed pair appear on the lower pedestal of the columns before the entranceway leading to the garbhagrḥa.

In considerable contrast, in the very same region, religious centers that are both constructed and governed by Kālamukha preceptors frequently make a more ostentatious display of these commitments. The eleventh-century Cāḷukya-period Kaḷēśvara temple at Huvina Hadagali,²⁶ the city of “the flowery boat,” for example, offers a pristine presentation of the four sides of the śikhara which sits over the garbhagṛha. About a century later, an even more elaborate but decidedly more damaged variation of the same vision appears on the śikhara of the Somēśvara temple at Puligere.

[[P639]]

Huvina Hadagali, details. L: Emanation chain of five Bhairavas, Bhairavī, head damaged, top of sequence. R: Detail of first standing Bhairava.

[[P640]]

Huvina Hadagali, details. L: Bhairava, seated with bell. R: Bhairavī with jackal and ghouls, head damaged, detail from top of sequence Puligere/Lakṣmēśvara, details of a cycle of Bhairavas on one façade.

[[P641]]

Puligere/Lakṣmēśvara, Details of a cycle of Bhairavas on one façade. Details of emanated goddesses on separate façade of same temple.

Thus, at the very place²⁷ the Kavirājamārga praises as the heart of the Kuntaladēśa where the people speak the most exalted and literary Kannada, an unmistakably Śākta-Śaiva ritual program is in evidence.²⁸

[[P642]]

Puligere: Detail, seated six-armed goddess, probably nominally Kālī, in lalitāsana with kapāla.

[[P643]]

In fact, rather unusually, this Somēśvara does not merely derive its name from some vaguely aspirational institutional links with the temple of Somanātha at the Prabhāsakṣetra in Gujarat. Rather, as is suggested by documentary records and even more forcefully asserted by the material culture, its medieval pūjāris were themselves refugees from that original storied pīṭha who carried with them two images from the old temple, one of which, depicting Śiva and Pārvatī riding a bull remains the focus of worship down into the present day.

Puligere. Above L: Śiva and Pārvatī mūrti in worship in situ. R: Photograph of mūrti, presented without alaṃkāras. Bottom: Twelfth-century panel from gateway in temple showing worship of the same image in situ. Notice the tall hats on the two sādhakas. Allegedly, this mūrti was rescued from inside Prabhāsakṣetra, Somanātha.

[[P644]]

Detail of Bhairavī on the śikhara at Kēdārēśvara temple, Balligave, ur-maṭha of the Śaktipariṣad line of Kālamukhas.

[[P645]]

Finally, in the district of Shimoga, central Karnataka—scarcely two miles from the Prāṇavēśvara temple established by Mayūravarman in the fourth century—lies Balligave, quintessential hub of Kālamukha religiosity. In violation of virtually everything suggested to us by over a century’s worth of secondary literature²⁹ produced without much attention to its [[P645]] material culture, the visual repertoire that permeates this captivating town is irreducibly Śākta.

Detail, Bhairavī on the śikhara of Kēdārēśvara.

Kēdārēśvara, Balligave. Main ritual center of the Śaktipariṣad line of the Kālamukhas.

[[P646]]

In fact, likely the oldest intact structure at Balligave, which incorporates numerous elements from the Rāṣṭrakūṭa period, is a Śākta ritual center still in worship. Now called the Kāḷamma temple, its ritual focus is the goddess Kālikā. In her short study of the related inscriptions, Shanthamurthy tells us that in the twelfth century, this goddess “was the guild goddess of metal workers, who are described as ‘sakalaguṇagaṇālaṅlkṛta satya śaucācāra naya Vinaya śīla sampannarum, śrī kāḷikā dēvī labdha vara prasādaram, pañca lōhādhipatigaḷ.”³⁰ The pañcalōha in question are the five metals used especially for the

[[P647]]

making of mūrtis, and thus the ones who have sovereignty over such materials (adhipatigaḷ)

Balligave: Kāḷamma temple, detail of Rāṣṭrakūṭa-period plaque over door lintel. The image depicted, seated in a royal posture and bearing a sword and a shield, is the goddess now called Kāḷamma. She sits dead center, surrounded by her attendants, Gaṇeśa and a stack of human heads. Intriguingly, Śiva is nowhere to be seen. Stylistically, in terms of the rendering of the robust figures, their large scale, and the form of the piece as embedded in a large plaque affixed above the garbhagṛha that depicts a tableau as opposed to a single deity, this is unmistakably pre-eleventh-century Rāṣṭrakūṭa material culture.

[[P648]]

are the caste gurus of the artisan communities responsible for the making of images, whose ranks would have included, at the very least, metal workers, stone masons, goldsmiths, and wood workers. Apart from the stone masons, who have largely vanished from the landscape, these are the very same constituencies that continue to worship at the Kāḷamma temple today.

Balligave: Kāḷamma temple, main mūrti of the goddess Kāḷamma worshipped by artisans. The deity depicted has just decapitated an elephant, but otherwise resembles the one found on the lintel.

[[P649]]

Kāḷamma temple and its small maṇḍapa as it appears from the other side of the lake.

Kāḷamma lies on the bank of the main lake which, while some distance away from the more famous temples of Kēdārēśvara and Tripurāntakēśvara, seems to represent the original ritual center of Balligave.³¹ On the other side of the lake, one finds a cluster of Kalyāṇi Cāḷukya-

[[P650]]

period temples as well as the dilapidated ruins of some early medieval maṭhas. The most important of these, which local people often identify as “the pañcamaṭha,” now under the care of the Kāśī śākhā of the Pañcācārya Paramparā of Vīraśaivas, retains its association with the goldsmith castes who now essentially control the Kāḷamma temple. According to people in the neighborhood, though Virakta Liṅgāyats would dispute this, the grounds surrounding this complex are said to mark the birthplace of Allama Prabhu, the focus of the Vijayanagar-period

Balligave: Maṭha of the goldsmiths and likely other artisan castes, now locally identified as “the pañcamaṭha.”

[[P651]]

anthologies that go by the title Śūnyasampādanĕ, second only to Basava himself as the locus of later Vīraśaiva devotion. For our purposes, what is most interesting is that in Balligave at least, Allamaprabhu, the quintessential śivayogin, is said to have been born into the goldsmith caste within a family that lived inside and attended to the pañcamaṭha complex, whose preceptors served as the final authority among the artisan castes.

Balligave: Interiors of the two chambers of the “birthplace” of Allamaprabhu. While the image on the left is a twentieth-century memorial, the liṅga on the right rests in an elaborate Cāḷukya style pīṭhā.

[[P652]]

While a thorough study of Balligave’s material culture could easily run onward for several more dozen pages, we would be remiss to not engage, if briefly, with the rather spectacular evidence for Śākta commitments in evidence at Balligave’s other “famous” temple, Tripurāntakēśvara. In orienting us to the rather extensive textual records associated with the site, selections from Shanthamurthy’s summaries of the available evidence are most useful.³²

The temple is dated to early 12cAD or late 11cAD, and at least 30-40 years later than the construction of the Kēdārēśvara temple. It is said to be similar to, but probably pre-dating the earliest temples in Belur. We can reconstruct the history of the Tripurāntaka from 21 inscriptions. The earliest inscription referring to these kāḷāmukha ascetics belonging to the bhujaṅgāvaḷi of śakti pariṣe is from 1050AD. It is a grant to a disciple of Kaśmīra Paṇḍita Dēva for religious services of the deity Mallikārjuna of Bēgūru, which is a village about 20kms from Baḷḷigāve. The name of the disciple is defaced in this grant, but in another inscription of 1066AD found in the same place, we are told of a grant of land and a pond to Trilōcana Paṇḍita Dēva, disciple of Kāśmīra Paṇḍita Dēva, for feeding the students of the maṭha associated with Mallikārjuna. The praśasti of Trilōcana follows the usual kāḷāmukha pattern. The next reference to Trilōcana is in c.1070AD in a much damaged inscription found near the present day Tripurāntaka temple. It records a grant of a village by Barmadēva, feudatory of Cālukya Sōmēśvara II, to Trilōcana for the temple and maintenance of the ascetics of the maṭha. The praśasti of Trilocana in this inscription is unusual, in that while it identifies him as a kāḷāmukha, it is written in a less Sanskrtized Kannada and does not follow the formula that most other kāḷāmukha praśastis do; nor does it appear to recount his lineage. . . . An inscription of 1077AD from the village of Maṭṭikōṭe near Bēgūru records a land grant by some Baḷḷigāve merchants for religious services of Mallikārjuna of Mariyāse installed earlier by them, and for maintaining the ascetics. . . . In 1096AD we are told that Sarvēśvara Dēva, a general of Cālukya Vikramāditya VI, built a temple with a golden kaḷaśa for the god Sarvēśvara of Tripurāntaka and granted a village for religious services and maintaining its ascetics. The recipient of these grants was Varēśvara Paṇḍita, who had now become the ācārya of Tripurāntaka. He is described as the foremost disciple of Trilōcana Dēva, who himself belonged to the lineage of the kāḷāmukha-cakravarti Kāśmīra Dēva. . . . In an inscription from the village Cikka Māgaḍi, 1182AD, the ācārya of Tripurāntaka, Sūryābharaṇa makes a grant for the sthaḷa vṛtti of a jinālaya. This kind of grant, to an institution of another religious sect, is unusual; however, since the village belonged to Tripurāntaka, the ācārya may have acted in the capacity of landlord or local ruler and made the grant to satisfy the inhabitants, many of whom must have been jaina.³²

Shanthamurthy’s distillation of the contents of nearly thirty old Kannada and Sanskrit inscriptions show us a ritual center founded by an ācārya from Kashmir that nonetheless managed within a generation to forge a direct link with the Kalyāṇi Cāḷukya rulers themselves as well as their generals, on the one hand, and to establish constructive relationships with a range of local communities within Balligave itself, on the other. On the basis of Shanthamurthy’s investigation, however, we find little evidence, apart from the usual boiler plate, of what sort of commitments were held by its Kashmiri inspired Kālamukhas. In this regard, exquisite images preserved on both the interior and exterior of the Tripurāntakēśvara temple, some of which seem to be sponsored by the Kalyāṇi Cāḷukya kings and their ministers of war, prove most illuminating.

[[P653]]

Balligave. /gtemple. Exterior: Tripurāntakēśvara illuminating. Tripurāntakēśvara, exterior niche: Durgā with at least twenty-four arms, astride lion, suggesting an esoteric deity.

Only a single image, this rather superb Durgā, whose more than usual number of arms suggests her identity with a specific esoteric deity, survives in the niches on the exterior. Her identification as a Kaula deity is further suggested by panel of māṭrkās in the temple exterior, for here they number eight, with the emaciated Caṇḍikā being upstaged by an image of Mahālakṣmī.

[[P654]]

Balligave: Tripurāntakēśvara temple: Many armed dvārapāla flanking the entryway.

[[P655]]

As one enters into the initial sanctum of the temple, the entry is flanked by enormous door guardians bearing many of the insignia of the Kāpālikas, who once again have an unusually large number of arms, though these too have been deliberately defaced. The pillars inside contain a number of images, including this Kālamukha adept to your left, depicting goddess as well as a variety of types of male and female adepts.

Balligave: Tripurāntakēśvara. Detail of female adept seated on a serpent throne, her sword resting on a human head.

In the image above, which is paired with the male Kālamukha ascetic, an apparently human female, naked except for her jewelry, her hair done up in long matted dreadlocks that fall down to the small of her back, is seated in āsana upon the three and a half coils of a snake. She gestures, holding two objects, one of which may be a mātuluṅga, and to her side mounted on [[P656]] and again expressed in different registers, our pictures tell a story of surprising continuities, across region, dynasty, and religious formations.

Balligave: Tripurāntakēśvara, deity counterparts to the female and male adept.

a severed head rests her sword. Just as the male Kālamukha adept finds an esoteric counterpart in a four-armed Bhairava, our female Śākta adept shares much of her iconography with a seated four-armed Bhairavī, who, seated once again in a royal posture is also stepping upon a snake and wields a sword along with her triśūla, ḍamaru, and skull bowl. In other words, again Thought the remaining remnants are more subtle, now that we have an eye for what to look for, our visual archive allows us to retrace this same sort of story back to the site of the cultural articulation that time and again has proven so important to our story: the world of the Kalyāṇi Cāḷukyas as epistomized by the rājadhāni of Kalyana where Vikramāditya VI and his successor Someśvara III reigned for the better part of a century.

[[P657]]

Mainstreaming the Terrifying Lord: From Court to Monastery and Back Again

All that remains in the modern-day city of Kalyana associated with its non-śaraṇa past is a museum that prohibits photography and a fortification of primarily Indo-Muslim design. And though, as Richard Eaton and Phillip Wagoner have famously argued, later rulers in the Deccani Sultanate consciously appropriate certain visual features evocative of the Cāḷukya past, explicit expressions of the esoteric religious vision of the medieval period did not figure among their chosen patrimony.³³

Fort at Basavakalyana: Some examples of creative reuse. Notice that in the second image a standing figure has been mutilated.

Should you search scarcely a mile outside of the center of the modern town on the banks of the river, however, the avid explorer encounters two relatively intact Cāḷukya-period ritual and cultural centers, now called Shivapuri and Narayanpur. As the historian and epigrapher Dr. R. Shadakshariah of Dharwad University argues in his forthcoming monograph, unpublished inscriptional evidence suggests that the latter site in particular was the location of the main Cāḷukya palace.

[[P658]]

Shivpuri, Kalyana: Top Left and Right: Two images of the patron goddess of the Cāḷukya kings. Bottom Left: Portrait of female donor holding skull bowl. Bottom Right: Cāḷukya royal portrait, plausibly Vikramāditya VI.

[[P659]]

The ruins of the temple at Shivpuri lie alongside the bank of a small lake, the typical site prescribed by the Śākta Tantras for the construction of the type of small-scale goddess shrine that was the standard locus for esoteric worship. The central focus of worship, quite possibly the first image on the top left, appears to have been an emaciated ten-armed form of the goddess, vaguely reminiscent of Cāmuṇḍā. An unpublished inscription dredged up from the lake at Narayanpur, currently being edited by Dr. Shadakshariah not only identifies this deity as the clan goddess of the Kalyāṇi Cāḷukyas, but, for probably the only time in the inscriptional corpus, actually incorporates a relief image of an esoteric emaciated goddess on the top portion of the stela.

[[P660]]

In nearby Narayanpur, within the shattered remnants of an important ritual complex, in place of the expected image of Gajalakṣmī, over the door lintel we find this ferocious fanged goddess, regal, heavily armed, and big bosomed, framed on all sides by guardian Bhairavas. Scattered around the surrounding region in the open air, in what was once the maṇḍapa and garbhagṛha, lay several important images. Of particular interest are the figures we see below. Rather gentle in appearance, but the nonetheless seated on a corpse, the importance of this figure to the liturgical life of the complex is confirmed by its repetition, sans corpse, on a basalt shrine still in situ in an otherwise completely renovated shrine roughly to the northeast of the ruins of this temple. The significance of this placement in space is effectively inseparable from the actual identity of the deity. As we shall see in the next chapter, the exact nature of this mysterious figure has considerable implications not only for understanding the precise nature of the religious sensibility installed within the Cāḷukya ritual capital but more broadly for making sense of the practical dimensions of the social and institutional realities that animated the Kālamukha inflected Deccan.

Two images of the same for now unidentified deity, to be analyzed in the next chapter, from Narayanpur.

[[P661]]

All that remains of this Cāḷukya goddess temple is the doorframe and lintel images.

Narayanpur today is renowned for the “Śiva temple” that lies just to the east, an impressive structure that still operates as a living focus of worship, whose exterior is adorned with life-size images of seemingly secular women in various stages of undress.

Left: Outside edifice of Narayanpur temple, Right: Detail of life-size image.

[[P662]]

Despite its contemporary Śaiva associations, as a closer examination of its subtle iconography demonstrates, this too was originally a temple dedicated to the mothers.

Narayanpur “Śiva temple”: Detail of Saptamātṛkā panel on exterior. Narayanapur, “Śiva temple”: Detail of Saptamātṛkā panel over door lintel of garbhagṛha

In Power, Memory, Architecture. Contested Sites on India’s Deccan Plateau, Richard Eaton and Phillip Wagoner have suggested, seemingly on the basis of a relief image of Narasiṃha that adorns the stair way leading up to the temple, and some stray comments in Bilhaṇa’s mahākāvya, that the site we have been exploring is indicative of a shift in affiliation among the later Kalyāṇi Cāḷukyas towards an expressly Vaiṣṇava milieu. As we have already [[P663]] seen, this hypothesis³⁴ is not borne out by a close reading of the visual archive in situ, but had we any remaining doubts, they would more than alleviated upon entry into the temple’s interior. Once inside, the viewer is confronted by two types of images. On the one hand, affixed to the pillars are two exceptionally beautiful “secular” renderings of the female form.

[[P664]]

Narayanpur: Interior of the “Śiva” temple of the Kalyāṇi Cāḷukya emperors.

On the other, crudely stuck down into the concrete, one finds more than a dozen images, mostly of standing Bhairavas, emaciated goddesses, or dancing naṭeśas.

Interior of “Śiva temple,” Narayaapur, details. L: Caṇḍikā. R: Gallery of images, from right to left, three Bhairavas, the middle one is dancing, Caṇḍīkā, and, on the left, dancing Naṭeśa with khaṭvāṅga.

[[P665]]

Mostly standing around three feet tall, it is almost certain that the majority of these images were not originally housed in their present location. Nevertheless, simply the sheer destiny of representation tells us something significant about the landscape surrounding long ago Kalyāṇa, which must have been simply saturated with such figures, rendered not only in stone, but in the much preferred mediums of clay, wooden, and precious metals. In contradistinction to an older logic where such figures and the often macabre sensibility they instantiated were only installed on the edge of social reality in special designated spaces that unapologetically proclaimed their commitments to those granted entry into the carefully demarcated domain, at Kalyāṇa, the “public” face of these temples offer extravagant sensuous presentations of a seemingly “courtly” idiom, instantiated in the repetition in space of the bodies of beautiful women, who present themselves to the viewer either heavily adorned in jewels or in a state of [[P666]] nature. All the while, even on the outside, in the margins, along the borders—even, as we shall see in the next chapter—through forms of visual śleṣa, Śākta systematicities are being hidden in plain sight. Even if one gains entry into the more select environs of the interior of such a temple, visual representations of the esoteric deities remain distinctly muted when compared with the older register we saw at Ellora. To speak of this as a simple domestication, in the sense of a taming, of wild goddesses and their consorts of the sort that will be far more in evident subsequent centuries is misguided. The gods have not been made docile; they have just become cosmopolitan, adapting themselves like any good city dweller to inhabiting a variegated social arena abounding in diverse sensibilities.³⁵

[[P667]]

As we have been seeing through visual representations whose iconographic regularity borders on monotony, a narrow band of shared visual signifiers invoking a much more deeply rooted and conceptually variegated panoply of Śākta traditions saturated the medieval Deccan. As we have begun to recognize, the use and reuse of the same sorts of imagery as platforms onto which an initiated adept might project his pantheon places significant limits on the analytical power of iconographic analysis, beyond its capacity to identify domains where esoteric Śākta practice of some sort has been thoroughly institutionalized and to distinguish them from those where no such cues exist. In much the same way that, when one is trying to sell a house, a broker will recommend the removal not only of personal items, but elements that communicate a particular use for a given space, so that potential buyers can more easily envision a multiplicity of arrangements that meet their own needs, icons and the spaces that house them often seem to be designed specifically for repurposing over the longue durée as well as context-specific multifunction use. Highly differentiated icons and temples are thus something like white elephants, only of relevance and interest to a very specific and narrow band of potential users. Thankfully, there are other more productive approaches to thinking with images. In contrast to svayambhū liṅgas, mūrtis do not just rise up spontaneously out of the ground. Like any craft, they are the product of specific communities of highly skilled artisans locateable in lived space. If the theology of images often seems somewhat murky, the sociology of images is less likely evasive. Though an exhaustive study of the relevant corpuses will have to await another occasion, even if in a cursory manner, we can begin to track stylistic choices along with the masons’ marks and signatures produced by individual artist. Additionally, through engaging both unpublished and neglected textual resources, we can begin to reconstruct some of the ethos that animated an essentially finite pool of craftsmen.

In this endeavor, the material traces found around Basavakalyana at Shivpur and Narayanpur provide us with additional assistance. Here, it is not the subjects being depicted that are of interest, but rather the style in which the figures have been rendered. With their rounded cheeks, solid arms, and curvaceous bodies, the images at Kalyana adhere to the recommendations made by Someśvara III himself in the Manasollāsa that images should be “pleasing to the eyes, and somewhat plump.”³⁶

[[P668]]

Narayanpur: Outside of Śiva temple. Details of three images.

Fifteen miles to the east in the vicinity of Latur—the southernmost part of the Marathawada—the tiny remote village Ramling Mudgal abounds in images with strikingly similar features to the ones we have seen near Kalyāṇa. Though this village has never been the subject of academic study in any language, the existence of a least six temples as well as its comparatively enormous collection of images, perhaps as many as a hundred, to say nothing of the several large unpublished śāsanas scattered around the village are indicative of the importance of the place. That whole runs of images, which were clearly crafted at the same time, repeat an identical iconographic program, including more than two dozen figures of Umāmaheśvara—typically encountered as isolated forms—as well the fact that a number of the images appear to be partially unfinished, suggests the possibility that during the Cāḷukya era, this village served as a key artisans’ workshop patronized by the Cāḷukya capital. This

[[P669]]

Ramling Mudgal: 1) Umāmaheśvara with mātuluṅga and khaṭvāṅga. 2) Dancing emaciated goddess. 3) Mahālakṣmī, the Kaula goddess of Kolhapur, headless. 4) Maheśvara seated on bull with mātuluṅga. Ramling Mudgal: Runs of images indicative of an artisan workshop. Many of these repeat the same form or vary only slightly.

[[P670]]

Ramling Mudgal: Top Left: An early image of Kāmeśvarī or Tripurasundarī. Top Right: Parā/Sarasvatī. Bottom: Iconographically identical images of four-armed Śiva and his bull.

[[P671]]

Ramling Mudgal: Umāmaheśvara with deer.

Today among people in southern Marathawada from certain caste communities, Ramling Mudgal is remembered primarily for its samādhis, the earliest of which seems to date from the Cāḷukya period and the last of which was established in the mid-1980s, when the last yogin associated with this tradition appears to have left his body. Near the Rāṣṭrakūṭa-period temple now called Magadeva at the entrance to the village, one finds four particularly impressive Peshwa-period jīvansamādhis. At least three of these, a certain Devajī, his successor Nāganāthjī and Nāganāthjī’s disciple, whose name the elderly women pilgrims visiting in the village to pay their respects apologized for forgetting, belong to a lineage of siddhas, all sonārī goldsmiths, revered at once as both miracle workers and caste gurus throughout the region. [[P672]]

Some of the more recent Maratha era samādhis of sonārī yogins at Ramling Mudgal. Detail of the Pādukas of Nāganāthjī

If such sites of memory and the still lingering religious communities they have until recently produced represent a chapter that marks the end of this part of Ramling Mudgal’s past, the village also presents us with ample—indeed, in an understated fashion, downright [[P673]] spectacular—indications of that story’s opening chapters.

Ramling Mudgal: Unique images depicting artisans, with guild sigil of chisel and a chattra, indicative of an elevated status. Some of the iconography prefigures the early modern Viśvakarman.

Presenting themselves at a range of scales, from miniatures to around three feet tall, we find at least twenty such representations at Ramling Mudgal offering slight variations on a figure that seems to have no analogues anywhere else in the medieval Deccan of the twelfth to thirteenth [[P674]] century. The figure is bearded, seated either in padmāsana or lalitāsana, and bears a japamālā in the right hand and kalaśa or other sort of vessel in the other. Alongside the larger representations, one finds the unmistakable sigil of the artisan castes, especially stone carvers, though it is possible these have been engraved into the stone by a different hand at a somewhat later date. Apart from the fact that these figures have only two arms, the images at Ramling Mudgal bear a strong resemblance to the early modern mūrtis fashioned to represent the god Viśvakarman when he assumes the role of patron deity of the artisan castes.

Two equally compelling possibilities present themselves. The first is that images like these served as key formative antecedents drawn upon by the early modern tradition of Deccani artisans that in subsequent generations codified the representation of Viśvakarman. If this is the case, instead of standing in for a god who, though certainly present as a trope in the medieval Deccan, had yet to become central to the religious imaginary of Deccani artisans, these images originally bore some different significance. In light of the distinctiveness in the rendering of the individual facial features of each image, perhaps they were even originally intended as portraits of specific artisan yogins. Alternatively, two-arms notwithstanding, these may well be the oldest representations Viśvakarman on the subcontinent where he is presented not as merely an element in a narrative panel, but as a focus of worship.

Regardless, both interpretations gesture in the direction of a largely unnoticed emic understanding with deep historical roots in which technical craftsmanship was virtually inextricable from the practice of yoga and founded on the values of the Śākta Tantras. In the next chapter, drawing on unpublished pratiṣṭhā texts in circulation within the world that the Cāḷukyas —or more to the point, the artisans they employed—shaped and made we will come face to face with just such an insider’s view—approximating some part of the self- [[P675]] understanding of the very sort of people that have been captured in stone at Ramling Mudgal. But before we return to looking out at their world through the lens of some rather peculiar texts, we should reflect more critically and precisely about the conditions of possibility behind the archives we have used so far to begin to evoke something of the greater Cāḷukya milieu of the medieval Deccan. With the telling exceptions of the ephemeral chancellery culture we encountered in chapter 2 and the compositions of the Cāḷukya King Someśvara III himself we encountered in the last section of chapter 4, all of our evidence bears the distinctive mark of this artisan culture. Every inscription we have examined was inscribed, if not composed by such people. They are the authors of every social space depicted in our texts, from the courts where legal cases were tried to the temporary wooden pavilion where Āvahamalla misguidedly attempted to offer elaborate dāna to secure an heir. If we imagine the institution as impressive buildings, then this is the pool of talent that designed and constructed them. If we define an age by its inspired aesthetic style, in this case some of the most evocative representations of human like forms in stone in the history of art, then these are our geniuses. Even in purely materialist terms, if one’s conception of what underwrites a culture is its economic system, then these are the very people who mint the coins and thus manage the money supply. In short, from whatever angle we might approach it, when we study the early medieval Deccan, by necessity—for this is what we have left—we move through a world almost entirely shaped by a single social demographic that till now has largely evaded the purview of scholarship.

[[P676]]