Viśvarūpa

[ Previous Page | Next Page ]

Vaikuṇṭha-Viśvarūpa Vol. III

There is a prevailing confusion in some of the recent secondary literature concerning the correct nomenclature which should be applied to, on the one hand, 3-headed Viṣṇu images of the Mathura type, and, on the other hand, to 3-headed or 4-headed images of the Avantipura type. I append some observation and theories which seek to clarify this matter.

No image of 3headed Viṣṇu, made between the 5th century and circa AD 850 can be prior to AD 850. There is no textual evidence for the application of the name Vaikuṇṭha to the 3 headed Viṣṇu images which existed prior to AD 850; and no authentic canonical name has yet been traced which applies specifically to these images.

Images of 3headed Viṣṇu, in the Kashmiri style but made prior to AD 850, do not represent Vaikuṇṭha but the Kashmiri version of the Mathura 3headed Viṣṇu.

(It was this icon which circa AD 850 was adapted, by the addition of a fourth face on the back of the nimbus, to represent the 4headed Vaikuṇṭha concept. This was done by Avantivarman of the Utpala dynasty of Kashmir, in his first temple in the city that he founded, Avantipura)

Consequently all 4 headed Viṣṇu images made at Avantipur and elsewhere in Kashmir, (and any which may be found in Himachal Pradesh, Madhya Pradesh, and Rajasthan), after AD 850, must be considered to represent Vaikuṇṭha or a local variation of Vaikuṇṭha.

Three headed-Viṣṇu images, which were made anywhere after AD 850 could be either:

1. abbreviated versions of the 4-heded Viṣṇu.
2. versions of the earlier Mathura 3-headed Viṣṇu.

The style in which these images are made, indicates the cultural tradition from which they emanated; if they are Kashmiri in style, they emanate from Kashmir and ultimately from Avantipura; if they are North Indian in style, they emanate from North India and ultimately from mathura.

Such icons made in Kashmir or Himachal Pradesh, after AD 850 and in the Kashmiri style, were probably intended to be Vaikuṇṭha in abbreviated form, not the Mathura 3headed Viṣṇu.

Such icons made in Himachal Pradesh in the North Indian style are probably intended to represent the older Mathura 3-headed Viṣṇu, not the Kashmiri Vaikuṇṭha. The Mathura icon of 3-headed Viṣṇu penetrated the Sutlej and the Vyas Valleys, where images of it were made in the local sculptural style, but imitating the Mathura iconography in the representation of the animal heads.

Three-headed or 4-headed Viṣṇu images made in the Candella kingdom in Central India after AD 954 are to be identified as Vaikuṇṭha because the first multiheaded Viṣṇu image, which appeared there at about that date, is named Vaikuṇṭha in the Lakṣmaṇ temple inscription at Khajuraho. The sculptural style of the multiheaded Viṣṇu images Khajuraho is that of the local school. The iconographical style, however, which concerns chiefly the manner of representing the animal heads, is that of Avantipura and Kashmir in general, not of Mathura, Kanauj or North India in general.

There-headed or 4-headed Viṣṇu images made in the Rajput kingdoms of North India (in Madhya Pradesh, Rajasthan, Haryana) from the 10th century onward are also to be identified as Vaikuṇṭha. This is because the Rajput images were largely made imitation of the Candella icon of multiheaded Viṣṇu, known at Khajuraho as Vaikuṇṭha. Like the Khajuraho images of Vaikuṇṭha, the Rajput images of this god depict the animal heads in the Kashmiri, not the North Indian manner. I propose two causes for this, one factual and the other hypothetical. First, the three-headed Viṣṇu of the Mathura type no longer existed in North India in the 10th century when the Candellas came to power; multiheaded Viṣṇu was introduced into Khajuraho in the form of kashmiri Vaikuṇṭha. Secondly, the Rajputs' iconographic imitation of the Khajuraho image was a concrete expression of political emulation (whether real or desired) of the Candella achievement; after the Candellas made themselves independent of the Gurjara-Pratihara empire, other Rajput areas of the empire also asserted their independence as separate kingdoms. The icon of Candella Vaikuṇṭha became the symbol of that sense of independence, and in the late 10th centuries it was widely adopted in the Rajput areas beyond the boundaries of direct rule from kanauj, to the sou5h and west of Khajuraho, in territories as far away as Haryana, Rajasthan, and northern Gujarat.

The fact remains, however, that the larger number of Rajput Vaikuṇṭha images depict the god, not standing (as at Khajuraho), but seated, either on anthropomorphic Garuḍa (as largely in Himachal), or on a lotus (as chiefly in Central India); and this remains to be explained (see $6.1).

The majority -- are indeed seated images (though they are not all seated on Garuḍa), and they are more widely diffused than the standing images, occurring in territories extending eastward from Rajasthan to Malwa and beyond, to Daksina Kosala south of Candella kingdom (at Harshnath/Sikar, Shahabad, two at Nilakantha, Nagda, jhalavadh, Candravati, Hinglajgadh, Pippaliya-Jatti, and Bihari). It is quite clear that the Garuḍa-mounted form of Vaikuṇṭha did not stem from areas immediately adjacent to the Candella kingdom and then spread westward into Rajasthan; rather, it arose and remained in Rajasthan (with the exception of the Bilhari Vaikuṇṭha, which represents a singular extension of the Garuḍa-mounted type form Rajasthan into Central India). This implies, of course, that the image of Vaikuṇṭha riding on Garuḍa reached the Rajput feudatories, by way of northern Rajasthan, from Himachal, were the type was already well known. In Malwa, lying immediately to the south-west of the Candellas, the lotus-seated Vaikuṇṭha image was developed, Probably in connection with the existing concept of meditating Viṣṇu as maunavratin, the silent sage; this form also penetrated westward into central and southern Rajasthan, where it overlapped with the Garuḍa-mounted form.

On the basis of these observations, one draws the following conclusion concerning the historical situation and the relationship between the Vaikuṇṭha images and the politics of the time.

The standing Vaikuṇṭha of Khajuraho was taken up first in those areas of Rajasthan which, in the second half of the 10th century, were at that time the most established and politically defined, and therefore already virtually independent; these were the kingdoms of the Tomaras (the standing Vaikuṇṭha from Rohtak), and of the Guhilas of Nagda (the Sas temple doorframe Vaikuṇṭhas) who also acknowledged the seated Vaikuṇṭha form. In the early years of their rise to power around Anahilapataka, the Caulukyas or solankis also represented Vaikuṇṭha as a standing figure (the Sandera Vaikuṇṭha).

At a slightly later date, towards the close of the 10th century and in the early 11th, those kingdoms which consolidated as independent states more slowly in the wake of the collapse of the Gurjara-Pratiharas' imperial power took up the seated forms of Vaikuṇṭha, of which those mounted on Garuḍa probably derived from Himachal Pradesh (or possibly from Kashmir itself), while the meditating forms seated on a lotus would have been natural extensions of the silent-sage cult which already existed in the Rajput regions. These rising powers were those of the Cahamanas of Sakhambari (the Harshnath/Sikar Vaikuṇṭha seated on a lotus), of the Cahamans of Partabgadh in south-east Rajasthan (the Garuḍa-mounted Shahabad Vaikuṇṭha and the lotus-seated Vaikuṇṭha of adjacent Candravati), of the Paramaras of Dhar (the lotus-seated Vaikuṇṭhas of Hinglajgadh and Pippaliya-Jatti near Dasapura, both in Malwa), and of the later Kalacuris of Tripuri (the Bilhari Garuḍa-mounted Vaikuṇṭha in Central India). The Guhilas of Mewar should also be included here, since the large Garuḍa-mounted image on the south wall of the Sas temple at Nagda clearly represents their primary concept of Vaikuṇṭha, the smaller standing Vaikuṇṭhas at the base of the sanctum doorframe being probably a reference to the older cult-form of the god. South of Nagda in northern Gujarat, the Caulukyas or solankis similarly preserved the earlier standing form in their small Viṣṇu temple at Sandera, but later took up the Garuḍa-mounted Vaikuṇṭha, as at Valam.

There is considerable variation in the groups, series, and individual figures represented in the haloes of north Indian Viśvarūpa images. This makes of each of these sculptures a unique work, not only in terms of style but also iconographically. It is this strong individuality which makes the treatment of these images as a group problematical: as whole creations they require individual study. Certain key features of the central Viṣṇu figure have therefore to be identified and isolated to provide the basis for a systematic examination. Since for the most part only the upper part of these images is preserved, and since the hands and attributes are for the most part broken and damaged beyond recognition, it is the multiple heads of Viśvarūpa which must be examined in order to introduce the first set of iconographical definitions.

The earliest Viśvarūpa images in which the heads are preserved come from the northern axis of imperial Gupta power in the Ganga-Yamuna plain: from Bhankari north of Mathura, and from Gadhwa south of Allahabad. The central figure of the Bhankari image is identical to the contemporary three-headed Viṣṇu icons of Mathura having the side-heads of Narasiṃha (right) and Varāha (left). The Viśvarūpa at the centre of the Gadhwa panel appears to have had the same heads: despite of the erosion of many details, the proper left side-head of Viṣṇu can be seen to represent the profile of Varāha. These images reflect an early phase in the Vaiṣṇava incarnation doctrine and its cult in which the boar and Man-Lion were regarded as the primary avatāras.

The Samalaji images of Viśvarūpa, produced in the mid-6th century, have human heads. These sculptures represent approximately the same phase in the incarnation doctrine, however, since the secondary emanations from the axis of gods above Viṣṇu's central crown are Narasiṃha (right) and Varāha (left), who are shown arising from the primary incarnation Hayagrīva.

The next Viśvarūpa fragment in which the multiple heads of Viṣṇu are preserved almost certain comes from the Mathura region (National Museum No.66.25); on stylistic grounds it can be dated to the late 6th or early 7th century. The number of animal heads has here been increased to four by the addition of the profiles of Matsya and Kūrma, the incarnations, which in the developing avatāra-cult had now been given precedence over the Boar and Man-Lion. These were to constitute the standard four emanatory animal-heads of all Viśvarūpa images made in North India up to the 9th or early 10th century. But there is variation even within this simple typology, and it is this variation which provides the primary bases for the systematic analysis of Viśvarūpa images.

In the National Museum fragment of the 6th or 7th century, the Fish and Turtle profiles emerge from the Junction of Viṣṇu's neck and Shoulders, precisely replacing the heads of the man-Lion and Boar, which are moved up to the sides of the crown. The placement of these four heads clearly represents the introduction of an upward sequence alternating from left to right; Matsya (first, right) and Kūrma (second left) on the lower level, followed by Varāha (third, right) and Narasiṃha (fourth, left) on the upper. This alternating ascent is a continuation of the plant-growth analogy, which had been applied to multiple emanation iconography in Kusana Mathura.

In all known Viśvarūpa images dating between the 8th and 10th centuries, the sequence of heads seen in the National Museum fragment is changed.

The primary variation takes one of two forms: either (1.) the man-Lion and Boar heads revert to the positions which they occupied in the Bhankari Viśvarūpa of the 5th century, with the Narasiṃha face on the proper right and that of Varāha on the proper left; or (2.) both these heads are transferred to the proper left side.

The first alternative easily predominates in the archaeological record, but it also shows more secondary variation. While the lateral positions of the Narasiṃha and Varāha heads are now fixed, their vertical location can change in relation to the heads of matsya and kūrma. In the Viśvarūpa images from Tumain, Deogarh, Dalmau, Bhuali, and Bhuili, the Man-Lion and boar heads are placed above those of the Fish and Turtle, in Continuation of the disposition first shown in the national Museum fragment of the 6th/7th century.

In the Viśvarūpa sculptures from Manwadhi and Bajaura, however, the Narasiṃha and Varāha heads are returned to the junction of Viṣṇu's neck and shoulder, as in the three-headed Bhankari Viśvarūpa of the 5th century, and those of matsya and Kūrma are elevated above them.

Moreover, in these images the locations of the Matsya and Kūrma heads are interchangeable. Matsya on the proper right and Kūrma on the proper left, the relative position introduced in the National Museum fragment is continued only in the Viśvarūpas from Tumain, Manwadhi, and Bhuali.

In all other extent examples -- the Viśvarūpas from Deogarh, Dalmau,Bhuili, and Bajaura -- Kūrma is placed on the proper right and Matsya on the proper left.

The lack of system in the arrangement of the heads in these images makes it clear that the concept of a fixed sequence--whether vertically alternating or circular -- was either forgotten or deliberately abandoned after the end of the 7th century at the latest. The only conscious option appears to have been the lowering of the Narasiṃha and Varāha heads to their old position of Viṣṇu's shoulders, and this was acted upon only at northerly sites of Manwadhi and Bajaura. (The profiles of Matsya and Kūrma meanwhile became so reduced in size and indistinct in feature as to make their lateral location of very little account.) In the main, throughout the Viśvarūpa sites in Madhya Pradesh and Uttar Pradesh, the placement of the Man-Lion and boar heads corresponds to the development introduced -- probably from Mathura -- by the National Museum fragment, which must therefore be seen as representing the dominant tendency from the 7th century onward.

The second primary alternative, however, is strikingly different. All four extant examples show the Varāha head placed above that of Narasiṃha, both on the proper left side of Viṣṇu's face. Matsya and Kūrma appear on the proper right. This type of Viśvarūpa image appears to have been developed in Kanauj city from earlier forms made to the south in Madhya Pradesh (the earliest example is the Philadelphia Viśvarūpa.

On the proper right, two of these sculptures locate the profile of Matsya above Kūrma: the Philadelphia Viśvarūpa and the Kapoor Viśvarūpa.

Two other sculptures locate Kūrma above matsya on the proper right: the two Viśvarūpas in the Ram-Lakṣmaṇ temple in Kanauj.

$7.3

It would be noted that the two latest images of this group -- the two Ram-Lakṣmaṇ images, very similar to each other in size and style, and clearly representing the fully developed Kanauj Viśvarūpa type -- place the four heads of Viṣṇu in a new disposition ($7.2) which re-introduces the concept of a sequence. The profile of the Turtle now appears above that of the fish on the proper right, reversing the earlier pattern ($7.1), so that a clockwise circular order (pradakṣiṇā) is established around the central face of Viṣṇu: Matsya -- Varāha -- Narasiṃha. As Viśvarūpa images with the heads in this arrangement are known nowhere else, their exclusive association with the emperors of kanauj in their capital city is virtually certain.

This analysis is set out in Tables 1 and 2.

TABLE 1

[ Previous Page | Next Page ]

 Report Volumes  | Viśvarūpa Multimedia Project

[ Home | Search  | Contact Us |  Index ]

[ Sutradhara | International Dialogue Unit | Multimedia Development ]