THE ŚAMALĀJī Viśvarūpa
type of sculpture was invented in western India at a date which is still
disputed but which must, to judge by the iconography, be very close to
that of the Mathura fragments. There remain four pieces of sculptural
evidence, all in the same style, of this type.1
They are said by the institutions and individuals who own them to come
and adjacent Devni Mori, in Gujarat. For reasons of space, I shall here
deal only with the most nearly complete sculpture (Plate 57), which has
recently been reconsecrated, with a new identity, in a modern shrine at Śamalājī.3
The main figure has three human heads, the trirekha on the single throat, eight arms, and is seated in a near-squatting position, with knees apart and feet turned outward, upon a simple low bench, This seated posture (Plate 58) is derived partly from kuṣāṇa royal portrait statuary4 and partly from images of Viṣṇu riding Garuḍa of the same period.5 Gupta sculpture at Mathura also copied this position: there are at least two squatting Viṣṇu images in the Mathura Museum, one single-headed6 and another which has one remaining side-head on the left representing the Varāha. As I have shown above (pp. 123-24), the latter is the only Gupta image of Viṣṇu with an animal side-head growing from the god upon its own neck. Both these examples of the posture in Gupta art are rendered in relief upon nearly square plaques, complete images in themselves; they are not fragments of larger, multiform compositions of the Śamalājī type. Whether the Mathura and Śamalājī sculptors each derived this posture independently from kuṣāṇa iconography, or the latter copied it from the former, remains an open question; but the imitation at Śamalājī of Gupta iconography and anatomical form in the image under discussion in this chapter suggests that Mathura was the source. Referring to the plinth figure holding a blossom to his chest (Vijaya?), in this sculpture (Plate 59), U. P. Shah remarks: 'in the Śamalājī Mahā-Vishṇu, a number of smaller figures show beautiful modelling known to Gupta art, especially the standing male figure at the left end of the sculpture is a beautiful specimen of the Gupta idea of male form ....' 7
The god is naked but for a lower garment, the central folds of which fall to a point between his feet. A very narrow fold of the top of this garment is turned down over the waistband below the navel. This minor detail is perhaps significant in dating the image, for as Harle notes:
only element which can be said to be common to these figures [i.e. a
number of sculptures representing Śiva, the Mātṛkās and Skanda in
addition to the torso to be mentioned below, a Viṣṇu image and a head
from Kutch] and those of the Madhyadeśa is the little roll
down of the dhoti below the navel, seen in the Unchdih and Jhusi figures.8
latter are two Viṣṇu images from Allahabad District dated by Harle to
the mid-fifth and early or mid-fifth centuries respectively.9
girdle is intricately rolled and a loop of it hangs between the god's
thighs. His body ornaments consist only of a single-strand necklace and a
triple-wound keyūras. An almost
identical Śamalājī torso, standing with the same trirekha,
necklace: waistband and lower garment, is dated by Harle to the
'fourth or early fifth century', but this appears to be qualified by the
three folds around the neck, however, the way the pleats of the dhoti
are shown, and particularly the way in which a central loop falls down
from the rolled girdle are all identical none the less, except that the
girdle is thicker, to the Śamalājī sculpture of a later date.10
'later date', Harle suggests, is the sixth century at the earliest.11
on the other hand, believes that the Śamalājī Viśvarūpa
sculptures in the village itself and
at the Baroda Museum 'should date from an age not later than the sixth
The three Viśvarūpa .'-type
sculptures and one associated fragment from the Śamalājī area13
belong, in my opinion, to different periods. The image we are discussing I
believe to be the earliest and so similar in style to the standing torso
mentioned above that I accept a late fifth-century date for it, thus
almost reconciling the views of the two scholars cited.
the god's feet are the intertwined serpentine bodies of two nāgas
which have human torsos with heads barked by a triple cobra-hood. They
rise beside the legs of the god with their faces turned upward in
adoration (Plate 59). Their hands, now lost, were probably joined in the namaskāramudrā.
These human torsos rise in an elegant continuation of the sinuous
curves of their snake-bodies; and there is a line double line marking the
junction of human and serpentine forms, like a lip, giving the impression
that the personifications are emerging from a sloughed snakeskin.
pair of human figures stands at each end of the plinth, behind the nāgas.
One detail is of particular interest because of its originality, which is
often a mark of true inventiveness: I believe this sculpture to be the
earliest of the three remaining and that the other two are copies of later
date. The detail concerned is the stance of the inner figure of the pair
on the proper right-the tip of the snake-body of the nāga
on the left side whips across his left foot, which he is evidently
attempting to free by straining away from the serpent with his bent right
leg. The resulting posture is that termed pratyālāḍha14
which is repeated in the depiction of certain figures higher on the stele
who similarly, but with less obvious effort, disengage them-selves from
the central vertical axis of the main composition. This figure, although
it has no wings, may well he a personification of Garuḍa, archenemy of
snakes and conventionally the vāhana
of seated Viṣṇu, here held in captive attendance only by the great
tail of the serpent. The reason for this small but deliberately sculpted
detail is, in my view, to emphasize the significance of the nāgas'
support of Viṣṇu in his Viśvarūpa
role and the consequent displacement of Garuḍa. This minor drama being
enacted beneath the god's feet, though concerned with explaining a minor
reorganization of conventional symbolism, assumes the form of an
the opposite side of the plinth, the female figure in the corresponding
position-untouched by the other serpent's tail-is presumably the consort
of Viṣṇu, Śrī Lakṣmī. Unlike the trapped Garuḍa, she stands
casually with her weight on her left leg, right foot pointing forward, her
figure turned slightly toward the god. She wears a lower garment and a
long shawl is draped carelessly behind her shoulders. Like those of Viṣṇu,
her ornaments are simple: a girdle, nūpuras,
keyūras and a single-strand
necklace. In her left hand she holds an object that seems to be a blossom
with no stalk. Her face is smashed but the elongation of the broken area
suggests that her hair was piled high on top of her head. The god is thus
flanked by his consort on his left and his usual vāhana
(rather unwillingly) on his right.
figures at the two extremities of the curved plinth, both male and turned
away from the god, are well preserved. Both are dressed in a fashion
similar to Viṣṇu (including the turned-down fold of the lower garment
below the navel) hut have no crowns, wearing instead the curly 'Gupta
The figure on the proper left of the god holds a quatrefoil to his heart,
the other hand resting on the low-slung horizontal folds of his lower
garment; that on the right holds in his right hand a large fruit. There is
a broken piece of sculpture above both their heads which may have been a śiraścakra.
Viṣṇu is known to have had two dvārapālas
named Jaya and Vijaya.16
Flanking his throne and facing outward, these figures might represent
them, although they have not the alert attitude of guards, nor are they
body of the god has four pairs of arms, which might suggest that a fourth
head is to he imagined at the bark. All the arms are broken off at or near
the elbow except the uppermost left, the hand of which holds the disc of
Viṣṇu, the cakra. Although the
face of the disc and most of the hand supporting it are smashed, the index
finger survives, curled around the rim. Above the stump of the
corresponding right arm, the remains of an almost vertical sword-blade can
be seen. The mace (gadā) might be thought more appropriate here in normal
iconographical terms, but three of the chief Vaiṣṇava attributes-saṅkha,
cakra and gadā are held by
the almost undamaged figure of the Varāha avatāra in the nimbus above
(see figure AL in Figure 3.1). The shape of the mace as depicted by the Śamalājī
sculpture is clearly seen there to thicken towards the top,
which is square and has projections, whereas the sword in the right hand
of Viṣṇu narrows slightly to a rounded end. It might be suggested,
conversely, that the round object in the left hand is a shield, to balance
the sword. But this circular object was, to judge from the position of the
remaining index finger, held in the same grip as that in which the Varāha
figure mentioned (AL) holds the spoked disc, namely balanced upright on
the upturned palm with the index curved around the rim and the other
fingers curled across the front of the weapon, the thumb being presumably
at the back. Only the cakra-certainly)
not a shield--would be held in this fashion: it is a grip from which the
weapon can be hurled, spinning, either upright with an overarm or underarm
motion, or else horizontally with a forehanded or backhanded movement. The
nature of the correspondence between
these two objects held in the uppermost hands seems, therefore, to be that
they are both cutting weapons.
Fig. 3.1 Śamalājī nimbus. Free-hand sketch. Not to scale
substitution of the sword for the mace, which was more conventionally held
in the upper right hand of four-armed north Indian Viṣṇu images (from
times onward at Mathura), might be seen as a mistake in copying northern
prototypes at Śamalājī, but this seems most unlikely in view of the
accuracy of iconographic detail evident in the smaller figures in the
nimbus. The Śamalājī sculptors appear to have been fully versed in the
iconographical formulae of the northern sculpture centre-so much so that
they felt free to experiment with them. The three (standing for four)
human heads of Viṣṇu, for example, is innovatory, while the entire
composition is a further development and coalescence of Kuṣāṇa multiple
iconography, representing sculptures of Types A and B, which was almost
certainly conceived and most fluently realized at Mathura. The
sword-for-mace substitution must therefore be considered to have been
deliberate: a reason for it is suggested in my stage-by-stage analysis,
below, of the whole image.
upright, forehanded grip on the cakra is to be seen at its best, perhaps, in certain relief
sculptures of Viṣṇu -seated upon the serpent above the doorway and
mounted on Garuḍa in the Gajendramokṣa panel-on the Deogarh temple of
about A.D. 600. It can there be seen to be very much an 'in action' grip,
differing completely from the static display of the disc as an emblem
rather than as a weapon which dominated north Indian individual cult-icons
of Viṣṇu to such an extent that the object became personified by another
standing figure (the rotund Cakrapuruṣa), as did the mace (in the
graceful form of Gadādevī). The Śamalājī image, although an
individual cult-icon, is far from being a still-life. It is a portrayal of
Viṣṇu in action: not as a warrior-he has abandoned his aerial equivalent
of a battle chariot, Garuḍa. for a bench with a footstool of writhing nāgas-but
as the growing and expanding universe. The action is conveyed in such
anatomical details as the tension of the bent legs, the mass of extended
arms and the multiplicity of faces. The life-like wielding of the sword
and brandishing of the disc add to the impression of a god seething with
life and suppressed energy seeking an outlet, which indeed it finds in the
which does not, properly speaking, mean 'creation'17)
of manifold life-forms into the upper half of the stele above him.
central face of the god is completely destroyed but from the staring
profiles of the heads to left and right of it, facing away at right angles
over the quadruple shoulders on each side, I infer that its eyes were wide
open. The deity, though seated, was intended to appear awake and alert,
regarding the four quarters of the horizon simultaneously like caturmukha
Brahmā. The pierced and elongated ears of the main face conceal those
of the (Source no. 9A) smaller side-heads which emerge without necks of
their own. A tall cylindrical crown, slightly wider at the top than at the
rim and densely ornamented, is worn by each of the heads.
the mouth of each of the god's side-heads emerges a vidyādhara
carrying what appears to be a sword in the right hand while the left holds
on to his streaming robe. These figures are represented in the full flying
position, the toes of the backward-pointing foot of the trailing leg in
each case being just clear of the god's gently smiling lips. Vidyādharas
being airborne creatures, I assume that part of their significance here is
to convey the idea that air (as prāṇa,
the breath-of-life) is the breath of God. They might also be seen as
heralds of the particular Vaiṣṇava doctrine (vāda)
declared by the god-his vidyā
or wisdom-which is symbolized by the multiplicity of figures above them.
Artistically, they mark the upper limit of antarīkṣa,
which is vertically connected to the aquatic nether-world of the nāgas
by the body of the god and horizontally pervaded by his eight arms. Above
them, the heaven-world teems with divine figures, covering the greater
part of the relief.
should point out here what seems to me to be a clear parallel between the
organization of the main composition of the image and the form of a tree,
which first appeared on the reverse of Kuṣāṇa multiple sculptures of
Type B. Rooted underground in water where the god's feet rest upon the nāgas
(a tree is pāda-pa, 'drinking
through its feet'), the 'trunk' rises in the form of the body of the god,
his arms branching in the eight horizontal directions of spacc.1818
to proliferate in a mass of 'foliage' consisting of the divinities
spreading around and above the heads of the god.19
celestial area is depicted upon a semi-oval, a shape technically derived
from the enlarged and populated prabhā-maṇḍala of the
type otherwise exemplified only once in surviving Gupta statuary, at
Mathura (Plate 49). In the Śamalājī images, the massive nimbus was
vertically elongated and not closed at its base above the shoulders of Viṣṇu,
as at Mathura, but simply cut off by the horizontal flight of the vidyādharas, below which the edges of the backslab were cut
straight down to the plinth. The entire work is thus a round-topped stele
cut in deep relief, whereas the Mathura example seems to have been a
are twenty-one figures in the oval maṇḍala'.20
The organization is impressive. It was clearly part of the intention of
the designer that there should be three vertical rows of three overlapping
figures, fanning slightly outward from a central vertical row of three.
There is another pattern within this design, revealed by the physical
interconnections between the central row and the upper two figures of the
flanking rows. These connections are made by the trailing leg of the pratyālīḍha posture in which most of the figures appear: this leg
emerges from behind a particular figure near it. There is almost certainly
a deliberately implied relationship between images conjoined in this way.
I have pointed out the probable relationship between Garuḍa and the nāga
on the plinth of this sculpture, where the trailing leg is the connection.
It is certainly the link between multiple forms of Śiva in the slightly
later, almost certainly derivative sculpture at Parel (Chapter 4).
extension, as it were, of the 'trunk' of the tree-like form of the image,
three overlapping figures rise (Source no. 2A) above the central crown of
Viṣṇu. The first, his legs up to the thighs still concealed within or
behind the crown (Source no. 3B), is Hayagrīva. The outline of the long
equine face-its features are worn away-with pricked ears and a flaring
mane, is set upon a powerful human body dressed in the same kind of lower
garment as that worn by Viṣṇu.
The hands rest upon the ample hips.
and partly behind Hayagrīva rises Brahmā, seated upon a double lotus in
a way that suggests that his ankles (hidden by the horse-head) are
crossed; he is not depicted locked in the lull meditation posture (padmāsana).
A wide band, almost eroded, crosses his
ample figure from left shoulder to right hip; this may have been a yogapaṭṭa,
such as can be seen worn by Brahms in the Nara-Nārāyana panel on the
Gupta-period temple at Deogarh (Plate 36). Only the rolled waistband of
his lower garment can be seen, curving under his rotund belly; he wears keyūras
and a simple necklace like those of Viṣṇu. His three heads, by
convention standing for his traditional four, are represented in the same
way as those of Viṣṇu the two side-heads are slightly smaller than the
central face and are turned away from it at right angles. As in the Viṣṇu
figure, too, the neck of the central head supports all three and its
pendulous ears overlap those of the two profiles. The jaṭās
of the central head hang sideways to left and right, falling upon the
side-heads so that the latter appear to have upward sweeping locks, thus
uniting the hair-styles. The central crown of Viṣṇu is shared, by
multiplication, with the profiles on either side, creating a similar
impression of identity between the several heads. These similarities
suggest that the heads of the Viṣṇu figure were carved according to the
Brahms sculptural tradition and were similarly intended to represent four.
From this I infer that the multiheaded image of Viṣṇu in the Śamalājī
region was derived, sculpturally and probably ideologically, from that of
Brahms the creator, whereas the Mathura type, as I have shown in Chapter
2, was a syncretistic coalescence of the god with two of his animal avatāras,
developed from earlier vyūha iconography
invented in the Kuṣāṇa period.
on a lotus in the same position above and behind Brahmā is Śiva, third
in this vertical register, and apex of the whole sculpture. This image is
badly damaged, but the following identificatory details can be seen. The
remains of long, coiled braids of hair hang down the side of the
obliterated face, suggesting jaṭābhāra.
A trident with damaged prongs is held upright in the rear right hand
and an object which may have been a khaṭvāṅga
in the rear left. The front right hand makes the abhayamudrā, while the front left hand, resting upon the raised
left knee, holds an object which is either an aṅkuśa or an axe. This figure was intended to appear ithyphallic:
there is a projection above the waistband below the navel, whereas the
other images in this series have a knot tied in the lower garment under
the waistband, covering the groin.
sequence of figures in the central axis of the whole sculpture is thus,
from the horizontal basis of intertwined nāgas
branching pattern of figures mentioned previously, its interconnections
made by the trailing leg of the 'pratyālīḍha' position, joins four figures to the second and
third (Hayagrīva and Brahmā) of this sequence.
away from the axis, the trailing leg rising at a forty-degree angle from
behind the left shoulder of Hayagrīva, is the Varāha avatāra.
The upward thrusting, elongated boar's head is unmistakable, set upon
a vigorous human frame. The boar has four arms. The front right hand holds
a mace, the rear right is raised above the head and turned palm upward in
an elevated tripatākā-mudrā. The
front left hand holds a conch upon the raised left thigh, and the rear
left wields an eight-spoked disc. Apart from the overhead gesture, this
incarnation holds the standard attributes of Viṣṇu. Its counterpart on
the right, emanating from Hayagrīva at the same angle and in the same
determined attitude (Source no. 3A) is the Nṛsiṃha avatāra.
The right hands hold a sword and make the tripatākā-mudrā,
but the two left hands are broken off. These two avatāras are depicted rising- from Hayagrīva to flank Brahmā
(Source no. 3A-C) for two reasons. Firstly, they are of course
incarnations of Viṣṇu (not of Brahms) and Hayagrīva is shown emanating
directly from that god. Secondly, they are evidently cast in the role of
warriors, armed with the weapons of Viṣṇu, protecting- the unarmed Brahmā
who creates the material and living universe (a parallel to the
vulnerability of Viṣṇu asleep prior to creation when his personified
weapons defend him from disruptive forces in the shape of demons21).
As mythological saviours of the creation from the Veda-based
Viṣṇudharma on the other,
interpreted in the sculpture as guardians of Brahmā, they cannot emanate
from the demiurge as part of his creation, but are sent from the source,
that is from Viṣṇu (in this case via
Hayagrīva) to save the world of man. They are therefore seen rising fully
armed from Logos, the Word of God incarnate, represented by Hayagrīva,
Knower and keeper of the Veda which is the orthodox core of Dharma,
to occupy forward defensive positions flanking the main central thrust of
the next evolutionary phase personified by Brahmā.
Above these two incarnations, rising at parallel angles from behind
the shoulders of Brahmā, are Candra, the Moon, on the left, and Sūrya,
the Sun, on the right. Both
figures have two arms and with their hands hold up their flowing robes in
a gesture similar to that of the left hands of the vidyādharas
at the base of the nimbus. The
lunar god has a crescent, horns upward, enclosing the base of his halo.
The sun-god has a plain prabhāmaṇḍala
with an incised inner rim. Haloes seem to have been included in this
sculpture only were they serve a definite symbolic function: a disc with a
crescent identifies the moon, a plain disc, the sun. None of the other
figures is nimbate, with the notable exception of Śiva, seated between
Candra and Sūrya. Sun and moon thus stand to the right and left of Śiva
respectively, appearing at a slightly lower level and springing from Brahmā.
There is evidence of three later copies of the Śamalājī sculpture which is to be discussed here: two are
in the Baroda Museum (no.2.550 and a fragment, no.2.372) and the
third, probably of the early 8th century A.D., has been moved from a
private collection in Ahmedabad to the National Museum, Delhi.
A pillar relief at Mandasor clearly imitates the Śamalājī
images, representing the easternmost direct influence of this Western
Indian sculpture. The
Mandasor pillar is dated to the late 6th or early 7th century by J.
Williams, 'The Sculpture of Mandasor', AAA26, 1972-73,pp.61ff. I have
dealt with this material separately: see T.S. Maxwell,
'The Evidence for a Viśvarūpa Iconographic Tradition
in Western India, 6th-9th Centuries A.D.', Artibus
Asiae XLIV.2/3,1983,pp.213-34 and Plates 1-20. For a more recent
appraisal, see S.L. Schastock, The
Śamalājī Sculptures and 6th
Century Art in Western India, Leiden 1985; I have not yet read
this work thoroughly, but Dr. Schastok's view of the developmental
sequence of Viśvarūpa sculptures from Śamalājī
clearly differs from mine, and readers might care to compare the two
studies for further debate in the future.
This place-name is here transliterated in accordance with the way in
which it is pronounced locally.
It is now installed in a small shrine - not its original temple -
above Viśrāmaghāṭ, on the east bank of the river Mesvo below a new
dam; this shrine is dedicated to Śiva in the form of a Liṅga named Nīlakṇṭha
Mahādeva, behind which the sculpture has been placed as its female
counterpart. The image is
now known as Kalasī Chokrā ni Mā, 'Mother of Sixteen Children'. My
photographs of it decked with flowers for pūjā
being useless for iconographical analysis, I have relied upon those
kindly supplied by Dr. U.P. Shah and the late Mrs Madhuri Desai,
supplemented by my own notes made on the spot. Excellent photographs
in Schastok, The Śamalājī
Sculptures, Figs.34, 35-37.
Rowland, The Art and
Architecture of India, Plates 43.
H. Hartel, Indische Skulpturen
I, Berlin 1960, Plate 15. A similar posture was adopted by other Kuṣāṇa
deities, mainly female, such as Hārītī and Lakṣmī(e.g. Mathura
Museum nos. 17.1301 and C.30), but a suatting goddess is unlikely to
have been the model for an image of Viṣṇu.
D.B. Disalkar, 'Some Brahmanical Sculptures in the Mathura Museum',
JUPHS V., 1 January 1931,Plate 3 and pp. 23-4; Disalkar also notes an
image of Viṣṇu (Plate 4) seated in a posture very similar to that of
the Musānagar Śaiva relief.
U.P. Shah, 'Some Early Sculptures from Abu and Bhinmal', BMPGB 12,
1955-56, p. 55.
Harle, Gupta Sculpture, p. 25.
Ibid., Plates 61 and 62 and notes,p. 46.
Ibid., Plates 86 and notes, p. 48.
Ibid., pp. 25-6.
U.P. Shah, Sculpture from Śamalājī
and Roda, BMPGB
13(Special Issue, 1960), p. 70.
Supra, Note 1.
The position may be ālīḍha;
thia is a vaxed question. See J.C. Harle, 'Remarks on Ālīḍha', Mahāyānist Art after A.D.900,
Colloquies on Art and Archaeology in Asia No.2, Percival David
Foundation of Chinese Art, University of London 1971.
This may not have been a wig in reality, but it resembles one. See
Harle, Gupta Sculpture,Plate
28 and note, p. 39.
T.S. Maxwell, 'Lākhamaṇḍal and Triloknāth: The Transformed
Functions of Hindu Architecture in two Cross-Cultural Zones of the
Western Himalaya', Art
International XXIV.1-2 (September-October 1980), pp. 15-18 and
note 5, Plates 4-7A.
See M. Monior-Williams, A
Sanskrit-English Dictionary, New Edition, Oxford 1989, p. 1245,
the root sṛij and noun sṛishṭi Also A. Bharti, The
Tantric Tradition, London 1965, p. 212: 'For 'creating' does not
exist in Indian thought in any sense of 'creatio;
i.e. 'ex nihilo':
'manifestation', 'emanation' or a similar tern should be agreed upon
instead of 'creation' whenever Indian thought is under discussion'.
E.W. Hopkins, Epic Mythology,
Strassburg, Berlin 1915,p.205, notes a Mahābhārata
reference to Viṣṇu in this connection: 'It is as the embodiment of
space with four or ten directions that he is called four and eight and
This comparison is not fanciful on may part. See again Note 202 to
A. Mookerjee, Tantra Āsana,
A way to self-realization,
Basel, Paris, New Delhi 1971, notes in a caption (p.47) that the 'golden cosmic egg... divides itself into two parts forming
the twenty-one regions of the cosmos.'
As illustrated in the Śeṣaśāyyin in panel on the Gupta 'Daśāvatāra'
temple at Deogarh, to take a roughly contemporary example.
©Oxford University Press 1988