Home > Cultural Informatics > Visvarupa > The Visvarupa Iconographic Traditions > Book on Visvarupa by Prof. T. S. Maxwell


Chapter 2


Development in the Gupta Period 


The existence of a classical period in Indian art and culture is generally held to have coincided with, or to have been created by, the Gupta dynasty, whose rule lasted from A.D. 320 to about the mid-sixth century. Romila Thapar is cautious in her use of the word 'classical', on account of its being a term coined by non-Indian historians. She therefore finds it necessary at several points to advance definitions of classicism in relation to Indian history.1 From her political, social and artistic observations, it is evident that in speaking of Gupta classicism, one is considering North India, its ruling aristocracy and imitative upper classes both in the immediate vicinity of the king and in the provinces which enjoyed a certain degree of autonomy, and the art forms sponsored by the social elite.

I am here concerned only with religious art and, within this category, with the particular multiple icon types which are the subject of this work. The following types of icon belonging to this class are known from archaeological evidence to have been prevalent in the Gupta period, being executed in a recognizably Gupta sculptural style:2

1 . Brahmā

2. A form of Viṣṇu having the added side-heads of his Man-Lion (Nṛsiṃha or Narasiṃha) and the Boar (Varāha) avatāras, the multiple icon so formed being commonly termed 'Vaikuṇṭha' by historians of Hindu iconography.3

3. The same or a similar form of Viṣṇu, that is, the 'Vaikuṇṭha' type, with an enlarged and populated halo (śiraśkara) or mass of radiance (prabhāmaṇḍala) around the head, resulting in a yet more complex icon which I shall term Viśvarūpa for reasons to be given in due course.

The list is brief and the only two innovatory forms are Vaiṣṇava, the Brahms icon having already been developed in the Kuṣāṇa period. The Caturmukhaliṅga was perpetuated in Gupta sculpture, but multiheaded anthropomorphic images of Śiva, an iconographic formula for which had been arrived at in the Kuṣāṇa-Gupta transitional phase, as the Rang Mahal terracotta plaque (Plate 30) demonstrates were evidently discontinued. The one innovatory example of a multiheaded Śaiva divinity in the Gupta style represents Kārttikeya or Skanda, the six-headed son of Śiva.

There was in Gupta art a distinct reluctance to conjoin subsidiary figures intimately with the anthropomorphic figure of a deity. What might be termed 'split images', such as the androgynous form of Śiva, Ardhanārīśvara, The Lord Who Is Half Woman, constitute a different category entirely. For the two polarities of the same divinity are wholly subsumed in a single figure. The densely populated Varāha images, such as the Eran sculpture, indicate the ability of Gupta-period artists to create highly complex sculptures, but very little of the theologically dictated experimentation with ramifying and multiheaded iconographic form of the preceding Kuṣāṇa period was taken up and developed in Gupta religious art.

Of all the multiple and multiheaded anthropomorphic icon types that were created in the Kuṣāṇa period, only that of Brahmā continued into the Gupta period. There is otherwise a noticeable hiatus in the development of the ramifying and multiheaded imagery of the experimental Kuṣāṇa phase, a hiatus which lasted until the late-Gupta and post-Gupta periods, when the types invented in that earlier phase reappeared in more accomplished artistic styles, their multiple compositions imbued with a new confidence and vigour. It seems, in fact, that during the Gupta phase there was an aversion to the depiction of the gods with conjoined multiple forms or aspects, to which the images to be discussed in this chapter were exceptions.

Images of Brahmā

One of the earliest surviving Brahmā images made at Mathura in the Gupta style is a small, probably domestic image of red sandstone (Plates 33 and 34). 15.25 centimetres high.4 It is broken off at the knees, but its original height is unlikely to have exceeded 22.5 centimetres. The icon is not cared in the round, having clearly been intended to be seen from the front while it stood against a wall or within a shallow niche, as the back consists of a slightly curved, plain surface. The heavy, somewhat pendulous appearance of the chest and abdomen was to remain a consistent feature of icons of the god throughout most of the period covered by this study. Other fairly standard iconographical details of subsequent Brahmā images-a yogapaṭṭa, yajṅopavīta or a black antelope (kṛṣṇājina) skin slung transversely across his tom-are, however, absent. This figure wears only a loose garment, its rolled upper part passing around the left shoulder and hanging in loose curving folds under the belly, emphasizing its rotundity: there are no body ornaments (as there were none visible on the front of our sculpture 12, Plates 31 and 32) an indication, perhaps, of the ascetic nature of the god despite his bulk. His right hand is raised nearly to shoulder level in the abhayamudrā; the fracture lines of the missing lower Ieft arm indicate that it was lowered to hip level, where the hand probably held the symbol of an ascetic or sage, the waterpot (kamaṇḍalu).

Three heads of the god are represented, each upon its own neck: the neck of the central head is visible despite its pointed beard, and the side-heads are clean-shaven. The face of the central head, which is carved in proportion to the rest of the figure, is missing; the outline of a piled jaṭābhāra, the ears, throat and beard extending like a long goatee as far as the centre of the chat are, however, still plainly distinguishable. The two side-heads, with only the throats of their brief supporting necks visible as they emerge from the shoulders of the central figure, are smaller than the central face, and are directed outward at an angle of about forty-five degrees (Source nos. 9B and C) on either side. The jaṭās, combed upward and hound atop each head by a hand, are well preserved; each face is represented with its own pair of ears-there is no overlapping of the ears adjacent to the central face by the larger ears of the latter (compare our sculptures 9 and 12, Plates 27 and 31-2), and those on the outer edges of the composition are not merged with the back slab, but are shown in full.

Despite this scrupulous individual treatment of each of the heads shown, there is no evidence (with the crucial exception of our sculpture 12 which is iconographically a transitional piece between Kuṣāṇa and Gupta methods) of sculptures in the round depicting Brahmā with all four faces. Even individual icons, such as this Mathura example, were treated as reliefs, however deeply cut, with the result that the fourth face was not represented. It is important to note that this iconographical 'abbreviation' is introduced at the very beginning of the standardization of the brahmanical pantheon in sculptural form under Gupta rule; the full 360 degrees around the image is no longer visually important, as it had been in the Kuṣāṇa art of Mathura-especially with regard to the symbolic tree on the reverse of multiple images-but only the 180 degrees which would normally be seen in frontal worship of an icon, as in a small temple, domestic shrine, or on a temple wall. This change suggests a transition from the worship of images in the round, in open-air hallows, to worship within an enclosed sacred environment. It further suggests, in the case of Brahmā, that the fourth face of the god is taken for granted; given the frontal approach to images which now becomes conventional, there is no need to represent the fourth head, for example atop the three which are aligned upon the shoulders (Source no. 3B). The conception of this deity as four-faced (caturmukha) was, through the literature-in particular, no doubt, the epics-so axiomatic that the three heads were entirely sufficient to suggest four-headedness. Iconographically, the treatment of the heads in this sculpture may be regarded as a further transitional stage following the Kuṣāṇa/Gupta Brahmā (Plate 31) in which the four heads are bearded; here, at Mathura, only one has a beard, and later Gupta examples have no beards at all. The absence of the beard accords with the absence of any mention of it in most, if not all, iconographical texts.5

For examples of Brahms images in the fully mature Gupta style one can do no better than examine the two well-preserved versions sculpted in relief on the outside wall surfaces of the so-called 'Daśāvatāra' Viṣṇu temple at Deogarh, Uttar Pradesh, some two hundred miles to the south of Mathura. These appear as minor, though symbolically important figures within two large panels, one on the south wall in a Vaiṣṇava depiction of the creation, and the other in a Nara-Nārāyaṇa scene on the east side. In the creation scene, Brahmā appears as active demiurge seated upon a lotus (Plate 35) which grows from the waters symbolized by the serpent Ananta or Śeṣa upon which Viṣṇu as the source of creation lies (as Śeṣaśāyyin) absorbed in his own being; while in the representation of Nara and Nārāyaṇa, the deity (Plate 36) occupies the centre of the frieze along the upper margin of the panel, presiding over and lending an air of orthodox sanctity to the hermitage of the two ṛṣi manifestations of Viṣṇu.

Brahmā in both roles appears in the same posture and similar iconographic forms. The god is seated cross-legged, contemplative, upon a lotus, his hands being in the same positions, displaying the same mudrā and holding the same waterpot as in the earlier standing icon. The number of arms has not increased, nor has the number of heads, which are treated in a manner similar to the earlier Mathura technique (though none is bearded), hut face more directly outward, appearing from the front as full profiles, turned at right-angles to the central head which they equal in size (Source no. 13A). In the Śeṣaśāyyin panel version of the god, he wears the black antelope skin most prominently across his body, in place of the cloth garment of the earlier version. The image of the god which appears in the Nara-Nārāyaṇa, panel, however, wears a wide sash with plain borders and incised cross-hatching which seems to represent the yoga-paṭṭa of a yogin, that is, a band of cloth or other material used in yogic exercises, for example in holding the knees up when sitting cross-legged. (Parts of this sculpture appear to have been subjected to later recutting by an inferior artist, but the sash or yoga- paṭṭa seems to be the work of the original sculptor.) The reasons for this difference between the two representations of the god on a single temple may have to do with the differing symbolical roles to which I have referred, that Brahmā plays in each tableau. As demiurge in the Śeṣaśāyyin panel, his wearing of the black antelope skin may have been intended to emphasize his omniscient ṛṣi character, since he is in the process of creating the world; whereas, in the other panel, his wearing a piece of the yogin's standard equipment is most suitable in a scene depicting a place of austerity, namely the āśrama of Nara and Nārāyaṇa.

As Banerjea6 suggests, Brahmā, in the brahmanical religion of this and later periods, had become a subsidiary figure in the cults of the two major gods, Śiva and Viṣṇu; his presence was necessary, but adaptable to the symbolical necessities of cults other than his own, which was almost eclipsed by the rise of other deities. But there is a contemporary image which seems to contradict this state of affairs, at least in one area. It is a bronze (Plate 37), beautifully made, from Mirpur Khas, Sindh,7 some 800 kilometres west of Mathura and 500 kilometres south-west of Rang Mahal. Like the Kuṣāṇa/Gupta Brahmā from northern Rajasthan or Haryana, it is an image in the round, with all four faces of the god represented; and as in the Rang Mahal terracotta Śiva (Plate 30), the right hand is raised as if in abhyamudrā, but with the palm facing the body. The thin, snaking yajṅopavīta crosses the front of the corpulent torso from the left shoulder to the right hip; whether it traversed the back in continuation is not certain, due to erosion. A robe is draped over the left shoulder, leaving the left hand free, the position of which suggests that originally it held a kamaṇḍalu, which was probably the intention in the earlier Mathura sculpture (Plate 33).

Unlike the side-heads of the bust from Rajasthan/Haryana-or those of an)- other multiheaded image so far examined-in this bronze image they arc made to face backward, away from the central fare, at an approximate forty-five degree angle. From the rear, (Plate 38), they present a full profile. These heads, and that at the back, are smatter than the 'natural' one. Both front and rear heads have their own cars, elongated and pierced but not ornamented, which entirely conceal those of the narrow, compressed side-faces. On the back of the image, above the rear face and between the shoulder blades, is a projecting tenon (as there are also on the soles of the feet), either to support a large, open-work nimbus, or else to fit the image into a framework with other deities. In either event, it is apparent that this figure of Brahmā was designed to be seen primarily from the front; the inclusion of the fourth head at the back being regarded as an iconographical necessity rather than as a feature to be honoured during ritual circumambulation. The angle of the side-heads supports this surmise, for they would have appeared turned away from the observer but flush with a flat background. As to why they should face toward the rear rather than being angled toward the worshipper, I can only suggest that the idea of multiheadedness was being deliberately under-stated, the side-faces appearing as retreating aspects of the main face, rather than as bold and independent heads in their own right, each having its own symbolic identity. Such an explanation accords with the general Gupta aesthetic principle of avoiding monstrosity of appearance wherever possible, white remaining faithful to iconographical dictates. It was this principle which no doubt led to the virtual suppression of Kuṣāṇa multiplicity and the invention of so few new multiheaded images in the Gupta period.

  An image of Skanda Kārttikeya as Kumāra

Probably the earliest six-headed representations of this god-and, apparently, of his consort-occur upon coins minted by the Yaudheyas, a traditionally warlike people settled in modern Rajasthan who 'lived by their weapons' (āyudhajīvinaḥ) and had Skanda as their principal god.8 Although the Yaudheyas persisted as a social group during the rule of the Guptas, to whose suzerainty they submitted, and even later, it is possible that these particular coins antedate Gupta rule. This numismatic evidence shows the six heads of the god, and of a goddess on the reverse of one specimen, arranged in two rows of three set one above the other. In sculpture, such an arrangement of multiple heads does not appear until after the Gupta period.

The one multiheaded sculpture of Skanda which as far as I know survives from the Gupta period9 is a deep relief upon the upper part of the pilaster of a fragmentary gateway lintel from Pawāya (Padmāvati) in Madhya Pradesh.10 Bath sides of the lintel have been published by the Archaeological Survey of at India,11 but the photographs show little detail. The image is carved upon the pilaster directly behind the panel depicting the Vedic ritual in which the demon king Bali promises to the Vāmanāvatāra of Viṣṇu as much space as he can cover in three steps. The adjacent large scene, upon the curved lintel itself, shows part of the giant form of Viṣṇu; on the reverse, next to the Skanda image, appears the scene of the churning of the milk ocean by the gods and demons. The reason for Skanda's appearance in what is evidently a Vaiṣṇava sculpture is not clear (Plate 39).

The image, approximately 46 centimetres in height, represents the god standing, with three smaller main attendant figures upon his right. Skanda Kārttikeya appears with five visible heads, the central one facing forward, the next two on either side angled at about sixty degrees outward, and the last pair as profiles at about 120 degrees to the central face (i.e. facing backward into the stone block). Above them is the damaged remainder of a mushroom-shaped parasol, its vertical shaft rising from behind the central head. That a sixth face was intended to be imagined at the back is implied by the six right arms which survive; a total of twelve arms would provide the god with one pair for each of six heads. The live out of six heads are all represented on one level, disposed at roughly 60-degree intervals around an imaginary common axis (the single neck of Skanda continued upward by the shaft of the parasol), unlike the two-tier arrangement on the Yaudheya coins, in which all the heads face directly forward.

Each of the heads has the hair arranged in a bound topknot (śikhaṇḍa śikhaṇḍaka), a necklace hangs around the single neck, and in the narrow space between each head ear-ornaments are visible. In such a tight cluster of heads, representation of the ears themselves was clearly impossible. The raised six right hands do not appear to have held any symbols. Apart from the number of heads and arms and the hair-style, there are no other standard attributes of Skanda in the image, such as the spear, cockerel bell or peacock vāhana. This absence of the standard symbols by which the god was, in the Gupta period, characterized in his single-headed form, 12 suggests that this is either an experimental iconographic prototype or a particular aspect of the deity. (Alternatively, the Skanda figure alone, devoid of Śaiva cult emblems but with its flurry of arms and circle of beads, may have been placed next to the churning scene to strengthen the impression of violent rotation; the borrowing by Vaiṣṇavism of symbols from the Śaiva iconographic repertory was not uncommon, as will become apparent in discussing other images.)

The iconographical section of the Viṣṇudharmottara (third Khāṇḍa) which is devoted to this god differentiates between four aspects, namely Kumāra, Skanda, Viśākha and Guha (3.71). Of these, only the form designated Kumāra is said to have six facts:


 caturmūrteḥ kumārasya rūpaṃ  te vacmi yādava/

kumāraś-ca tathā skando viśākhaś-ca guhas-tathā//3

kumāraḥ ṣaṇmukhaḥ kāryaḥ śikhaṇḍakavibhūṣaṇaḥ/ 4ab


skando viśākhaś-ca guhaḥ kartavyāś-ca kumāravat/

ṣaṇmukhāa te na kartavyā...............

I shall tell you, Yādava, of the appearance of the four-formed kumāra;

[The four forms are:]Kumāra and Skanda and Viśākha and Guha.

Kumāra is to be made six-faced, adorned with topknots.


Skanda, Viśākha and Guha are to be made like Kumāra;

[But] they are not to be made six-faced.............


1 R. Thapar, A History of India, Volume One, Harmondsworth 1966, pp.136-66, esp.136 and 157-8.

2 J.C. Harle, Gupta Sculpture, Indian Sculpture of the Fourth to the Sixth Centuries A.D., Oxford 1974,pp.7-8, gives a resume of the elements comprising the Gupta style.

3 So, far example, K. Desai, Iconography of Viṣṇu (In Northern India, Up to the Mediaeval Period), New Delhi 1973, pp. 37-47.

4 Mathura Museum reserve, no. 34.2481.

5 J.N. Banerjea, The Development of Hindu Iconography, Calcutta 1956, p. 516. Yet later still, in many post-Gupta styles, the beard is represented in North Indian images.

6 Ibid., pp. 512-14.

7 Now in the Karachi Museum.

8 P. K. Agrawala, Skanda Kārttikeya-A Study in the Origin and Development of Iconography, Banaras Hindu University 1967, pp. 40-1 and figs.7-10.

9 R. C. Agrawala, in 'Skanda from National Museum, New Delhi and U.P. Hills', EW 18, 1968, p. 319, notes and illustrates a possible exception (fig.1), a small bronze: 'His central main head is surrounded by five miniature heads shown on the circular halo.' Agrawala thinks it was probably made in the Chambā Hills and dates from the 6th-7th century A.D.

10 State Museum, Gwalior, no. 543, open gallery 7.

11 ASI, AR 1924-25, plate XLIII (d). R.C. Agrawala, 'Skanda from National Museum', p.321, notes that the Skanda panel on the Pawāya lintel had not (in 1968) been published. Part of the relief appears in J. Williams, Gupta India, Empire and Province, Princeton, 1982, Plate 52.

12 Harle, Gupta Sculpture, plates 10,65 and 90 and notes, pp. 34, 46 and 49.



[ Previous Page | Content List | Next Page ]

HomeSearchContact usIndex

[ Home | Search  |  Contact UsIndex ]

[ Digital Library | Electronic Books | Visvrupa Project ]

Oxford University Press 1988