The Conundrum of the Udayagirl Mahl~asuramardlni

lndlra S. Aiyar

Gupta kings dominated the political scene in the north in the 4th and
5t11 centuries (320-467 CE). The art style that came into being during this
period, continued its influence on sculptures even up to the 7th century.
Answering his own query as to what is peculiarly Gupta art, J. C. Harle
gives two points 1 : that the sculptures had an inward-looking stance which
conveyed a high spiritual evolution, and secondly, their elegance which reflected
the quality of that civilization.

Udayagiri, the famous archaeological site of the Gupta period, is 5 km.

north of Sanchi in Madhya Pradesh. The caves of Udayagiri are all at ground
level, with the lone exception of a Jaina cave situated half way up a hill.
The religious affiliation of these caves seems eclectic, with equal spread of
Saiva and Vaisr:iava images. Though the caves are of little importance
architecturally, some special features - ornamented door lintels, and the
scroll-and-leaf decorations, so uniquely Gupta, appear here for the first time.
Elsewhere, there is almost no inscriptional evidence of this period directly
relating to any particular king. But at these caves, there exist some inscriptions
of great importance, which can be connected personally to a Gupta king.

There are three representations of the goddess Durga Mahisasuramardin1

in the cave precincts. The first one is on the north wall of the courtyard
in front of Cave no. 6, the second on the outer wall of Cave no. 17, and
the third on the fac;:ade of Cave no. 6 to the proper left of its entrance.
The first two are thought to be of the transitional period between the Kusana
and Gupta times, and the last one belongs to the early Gupta period, circa
401/402 CE. Of the two inscriptions on Caves 7 and 6, relating to Candragupta
11, one is dated to 402 CE. Each of these representations of the goddess
is an improvement in its delineation on its previous one. All are of buff sandstone.

The two inscriptions mentioned above throw up some interesting insights

about the donors. The first one on Cave 7 says it is commissioned by one
Saba-Virasena, a hereditary minister in Candragupta ll's court, who describes
himself as one who is well versed in (or felicitous with) words, practical
matters, human nature, and a poet of Pataliputra. The inscription on the north
wall of Cave 6 on the other hand, says that it was a pious gift by the
king of Sanakanikas, a vassal of Candragupta. Joanna Williams remarks that
these Sanakanikas are the same tribe conquered by Samudragupta. She notices
2 Indira S. Aiyar

a significant difference between these two donors 2 from the two inscriptions.
One belongs to the upper echelons of lhe elitist central political hub of the
Gupta government; the olher a humble devotional offering by a vassal tribe,
which is still powerful enough to donate the cave. Studying other iconographical
features of the two caves, such as the stance of the dvarapa/as and the
lintel ornamentations of the door, Williams observes that the carvings of Cave
7 were done by a sculptor well aware of the Mathura style of the period,
whereas the carvings of Cave 6 were done by a tribal sculptor who followed
the style of representation of the previous one. Further, all the sculptures
on the fa<;:ade of Cave 6 belong roughly to the same period; and taken
together both the caves must have been executed in the first twenty-five
years of Candragupta's reign. Though a stable political climate resulting in
good roads and communications gave a great fillip to art, pilgrimage was
a great geographical force and played an even greater part in disseminating
art from Mathura to other regions. It is well worthwhile to quote Williams
as she waxes eloquent over the art of this period: "It is as if the courtly
poetics of kavya, running throughout all Indian religions, at this point demanded
a corresponding visual representation of the great literary themes and their
rich embellishment or afamkara". 3

The first one of the three representations of Durga, by its position is

totally eroded by weather and is barely visible. But we can ascertain its
features, from a look at the second icon, since it is seen to be a copy
of the earlier one. Here Mahi~. as a buffalo, is leaping in front of the goddess.
DevT is pressing him down with one of her right hands, while a trisOfa held
in another right hand is used to kill him. A left hand, whose fingers are
missing, held near the muzzle of the animal, is thought to pull out and hold
his tongue. 4 She is also seen holding a single lotus flower in one of her
left hands, a gift from the Lord of the Ocean. 5 '

The third configuration shows a unique handling of Mahi~. He is thrown

on the ground, with the Goddess' right foot pressing his head, DevT standing
in an afic/ha pose. One of her left hands is holding his left hind leg. It has
been commented lhat the animal has been whirled around, and thrown on
the ground. 6 All the three sculptures of the Goddess are twelve-armed, and
there is the interesting object held aloft in her upper two hands. The third
icon (Plates I, II) is also much weathered, but most of it is discernible. Though
some of the arms, and the weapons they hold are broken, the object held
aloft is very well preserved.

From the Kusar:ia period onwards, DevT Mahi~asuramardinT is a popular

theme. Ku~ar:ia art is of no mean importance in Indian iconography. Innumerable
objects of Buddhist, Jaina and Hindu religions belonging to this era have
been unearthed in the Mathura region. These artefacts give us a glimpse
 The Conundrum of the Udayaglrl Mahl~suramardlni 3

of the glorious period of the religious iconography of the era. The plaques
of the Mahi:;;asuramardini theme recovered from this area are preserved in
the Mathura museum. They are all of them not more than 20 cm high, made
of red mottled sandstone, and excluding one, the rest are all very badly
damaged. The stance of the Goddess and her attributes are similar in all
of them. The goddess as seen in the best preserved plaque is six-armed
in a samapiida sthiinaka posture and relaxed in her attitude. She holds an
object aloft in her two hands. There are two prominent aspects of this icon
which engage our attention by their very absence: Devi's vahana, the lion
is never shown; and the tautness of the moment of overcoming the asura
is not portrayed. One may comment that there is a tranquil strength in the
Ku:;;iir:ia reliefs whereas the later icons show the vibrant dynamism in the
Goddess' stance. P. K. Agrawala 7 calls it the 'ponderous feeling of Kushan
plastic tradition', as compared to the 'suave elegance and artistic discipline'
of the early Gupta art. There is simplicity in the depiction of the moment
of victory of the Goddess over the buffalo, calmness being the main rasa
of the event. The sculptor has kept the theme in its bare form, 'an iconography
in process of formation.' But this became the basis for its development into
its future sophisticated representations, in the Gupta era. Harle has commented
that these may be the "first Hindu images .... that belong to a firmly established
iconographic type." 8

Returning now to the Udayagiri representation on the fa9ade of Cave

no. 6, the icon shows the Goddess for the first time placing her right foot
firmly on Mahi:;;a's head, while holding his hind leg by one of her left hands.
Her hands hold clockwise from her proper right, a broken object, which may
be a mace (gada), arrow, thunderbolt ( vajra), sword, the two hands holding
an object aloft; on the left side from top, a shield, bow, and a discus.

The particular object held aloft by the Goddess' upper two hands is the
subject of controversy amongst scholars. Over a period of time this has elicited
comments from various eminent art historians and there is no consensus
of opinion as to its identity. It is not a familiar object in Hindu iconography,
and it is held aloft in a peculiar way-in two upraised hands with the back
of the hands showing towards the viewer.

J. Ph. Vogel, examining the Ku:;;ar:ia plaque had thought that this object
held aloft was a serpent, though he was not certain. 10 Diskalsar, examining
another four-armed terracotta icon of the same period, thought it was the
skin of the slain buffalo. 11 Alexander Cunningham 12 seems to have seen
the icon at Udayagiri in a better state of preservation and he has listed most
of the above weapons; but omitted the top object either because he was
baffled by it as Harle comments, or, he might have listed it as the cakra
which is not visible now. He has also missed out the stJ/a. V. S. Agrawala
4 Indira S. Aiyar

identifies it as a bowl. 13 Odette Viennot, while examining the Udayagiri icon

under question, feels it is definitely a long oblong drum, and therefore the
Ku~r:ia artists must also have tried to depict such an object. 14 R. C. Agrawala 15
opines about the Udayagiri Durga, that what she holds in the upper hands
looks most probably like an iguana ( godha).

J. C. Harle has a similar figurine in his private collection, belonging to

the transitional interregnum between the Kusar:ia and Gupta periods. He states
that the object could be a snake, a lizard or the soft portion of any animal. 16
As for the Udayagiri goddess, he identifies the object reluctantly as a lotus
garland, 17 though the representation is unlike the garlands usually seen in
sculptures. 16 He also has an alternative suggestion that it could be a soft
basket ( karanr;la), or a piece of the matting used as a container for the
garland. J. N. Banerjea 19 has identified this object as an iguana, and quotes
Rupamanr;Jana to prove this animal's connection to the goddess GaurT. Grilli
Von Mitterwallner says that it is a garland. 20 Joanna Williams 21 calls this
a wreath (which, according to her also adorns the heads of the two dvarapa/as).
Doris Meth Srinivasan also identifies this as a wreath. 22 She is reluctant to
name this configuration as Mahi$asuramardini, because she sees no literary
support for this designation in that period. Myths need not always predate,
or even be coeval with their lithic representations. Religious literature is a
result of autochthonous ideas influencing oral traditions, which is very difficult
to gauge; and in particular, the Goddess myth has taken a long haul to
be Sanskritised.

There are two recent opinions on this object held by the Durga representation
in Orissa. Bhagabat Tripathy 23 commenting on a four-armed deity found at
Gandibedha in the Balasore district, states that the Goddess is holding a musa/a
(club). Pammasani Parandam and Veerabhadran Devadass 24 opine that such
an object is definitely a stone, which the Goddess is throwing at her enemy.
They support this view by the description of the battle in the Devi Mahatmyam,
where the Goddess throws amongst other things, stones to kill the general
of Mahi$a. Devotees of the Goddess seem to prefer the view that the object
is either a musa/a or a cakra.

Without further muddying the waters, this paper would prefer to lean
towards the views that the object is either an iguana, or a crown of snakes.
Lak$mi is th~ presiding devata of the Madhyama Caritam, of the Devi
Mahatmyam. Sri as one of the six forms of Gauri is said to have an alligator
as her vahana . Gauri herself has an iguana as her vehicle. 25

This concept in its elementary form as seen from its representations of

the Kul;iar:ia era culminates in this unique Udayagiri sculpture. The whole image
might be of a composite nature, putting two or three ideas together, a fore-runner
to its narrative form. Devi is shown holding not only the weapons she used
 The Conundrum of the Udayagirl Mahl~suramardlni 5

against Mahii,;a., but also those she used to vanquish other asura~. In the
Devi MaMtmyam, 26 the devas comment that Devi gives moksa Jo 1fie asuras
0

while killing them. Moreover, Mahi$a is hinted as being an arrf$a of Siva

in Sanskrit as well as folk traditions. 27 Thus the godha held aloft may symbolize
the Goddess transformed as Gauri I Sri, giving sayujya to Mahi$a, after killing
him. In the Devi Mti.hatmyam 3.38 28 the word 'yavat' is interpreted as "Till
I as Can<;1ika (in the turfya form), become transformed into Mahalak$mi". 29
If the object is identified as snake, then again in the same text Devi
Mahatmyam, in the Pradhanika Rahasyam of the Uttarabhaga, when the king
Suratha requests the 8$i Medhas to describe to him the nature and significance
of Devi, the R$i describes the Goddess as wearing a garland of snakes on
her head. 30 The Yama/a Tantra describes her as a beautiful damsel wearing
a garland of snakes on her head. 31 Thus, there is some evidence to support
the view that the object might be a godha or snakes. 32

Finally, it has to be said that any of the above views may be the right
one. Though Truth be One, it seemingly shows different hues according to
the perceptional capacity of its interpreter.

Notes and References

1. J. C. Harle, Gupta Sculpture, 1996, p. 8.
2. Joanna G. Williams, The Art of Gupta India, 1982, p. 40.
3. Op. cit, p. 62.
4. Gritli Von Mitterwallner, "The Kuear:ia Type of the Goddess Mahieasuramardini as
Compared to the Gupta and Mediaeval Types", in German Scholars on India II,
Bombay, 1976, p. 202.
5. Devi Mahatmyam, 2.29, (with Tamil commentary), Sri Ramakrishna Mutt, Chennai,
1973. All references to the verse numbers in this article are from the text of this
edition.
6. Mitterwallner, Op. cit. p. 201.
7. "An Estimate of Gupta Terracottas" in Hari Smriti-Sludies on Art Archaeology and
lndology, Vol. I, 2006, p. 46.
8. J. C. Harle, "On a Disputed Element in the Iconography of Early Mahi~asuramardinT
Images" in Ars Orienta/is, Vol. VIII, 1970, p. 14 7.
9. Doris M. Srinivasan, Many Heads. Arms and Eyes. Brill, 1997, p. 285, identifies
this as a bell. She misses the vajra, and identifies the shield as a cakra.
10. J. Ph. Vogel. Catalogue of the Archaeological Museum at Mathura, p. 97; R.
C. Agrawala, "The Goddess Mahi~asurarnardinT in Early Indian Art", Artibus Asiae,
Vol. 21.2, p. 123, f.n. 2.
11. "Some Brahmanical Sculptures in the Mathura Museum", Journal of the Uttar
Pradesh Historical Society, Vol. 5, 1932, p. 55 in R. C. Agrawala, Ibid, p. 123,
6 Indira S. Aiyar

f. n. 7.
12. Archaeological Survey Reports.Vol. 10, p. 50.
13. "A Catalogue of Brahmanical Images in Mathura Art", Journal of the Utlar Pradesh
Historical Society, Vol. 12, 1949, p.158 in Odette Viennot. "The Goddess
Mahishasuramardini in Kushana Art", Artibus Asiae, Vol. 19, no.2, 1956, p. 371,
f. n. 8.
14. Odette Viennot, Ibid. p. 372.
15. "The Goddess Mahi~asuramardinT in Early Indian Art". in Artibus Asiae, Vol. 21,
no. 2, in Harle, op. cit p. 148, f. n. 8.
16. Op.cit., p. 153
17. Devi Mahatmyam. 28 ed. 29 ab.
18. Op. cit. pp. 152-153.
19. The Development of Hindu Iconography, New Delhi, 1974, p. 502.
20. Op. CJl, p. 203.
21. Ibid, p. 42.
22. Many Heads. Arms and Eyes. E. J. Brill, 1997, p. 289.
23. "Sakta Images in Northern Orissa" in Proceedings of /he ffh Session of Indian
Art History Congress, Hyderabad, 2000, p. 56.
24. "Early Sculpture on Sakti Forms", Op. cit, p. 131.
25. J. N. Banerjea, Ibid; T. A Gopinatha Rao, RrJpamal)f/ana in Elements of Hindu
Iconography, Vol. I, Part II.
26. 4. 18 & 19.
27. For references in Sanskrit and folk traditions, cf. Indira Aiyar, Durga as
Mahi$8suramardini, New Delhi, 1997, p. 282 ff.
28. rR rR PJuT ~ tfWllCl!ftlilflL~6'{ I
29. Commentary Guptavati of Bhaskararaya in Kandiyur Mahadeva Sastry Sri
Devimahatmyam (Malayalam) Kodungallur, 1994, p. 248.
30. Verse 9 : ~ cmrlm: ~<a,.d ~ I CflilfrQ61( fuw ~ fm: ~II
31. Kandiyur Mahadeva Sastry, Op. cit p. 546.
32. A similar icon of Siva holding an object (poesibly a wreath) above his head
is examined by Devangana Desai in her paper "A Remarkable Kushana Image-Unga
from Mathura Discovered by Bhagwanlal lndraji", Marg magazine, Vol. 58, No.
4, June 2007.
List of Illustrations
I. Durga Mahi~suramardini, fa<(ade of Cave 6, Udayagiri, c. CE 401/402
11. Close-up of Durga's raised hands and the object held by her, Udayagiri
 .. 1/

f'..;,-_
.;\, ..
- ·,

~~·~"
.. i

Durga-Mahi~suramardini. fa9ade of Cave 6. Udayagiri, c. CE 4011402.

II
....-
<I.)
..c
>-
..c
"..cQi
u<I.)
'E
0
<I.)
£
-g
co
""'cco
..c
"<I.)
"'
-~
"'.....
tco
Cl
::::J
Cl
0
Q.
::::J
I
<IJ
5
"'0
 Explorlng Versions of the Mahabharata

Observations on the Critical Edition of the epic,

the edition giving Its bare

Constituted Text and Its later redacted versions

Krishna S. Arjunwadkar

Critical Editions: An Overview

The project of the Critical Edition of the MaM.bhiirata ( Mbh), the great
epic of India and the largest single epic of the world, was completed, after
long years of expert labour, by the Bhandarkar Oriental Research Institute
(BORI), Pune in 1966. This project, planned with great care and executed
with the editorial help of renowned Sanskrit scholars from all over the world,
gave inspiration to similar undertakings in India and abroad. A Critical Edition
of the Ramayana, another great epic of India, was brought out by the Oriental
Institute in Baroda with the help of a team of scholars associated with it.
Another project was floated by Professor Slaje of West Germany and his
team on the voluminous work, the Yoga-vasi$fha or Mahii-ramayaoa or
Mok$opaya. What I am here concerned with is ( 1} the Critical Edition of
the Mbh, (2) the edition of only the Constituted Text (Con. Text) therein printed
separately ( 1971-76), and the editions of the hypothetical, (3) original
( Jayasarhhha (JS), 1977) and ( 4} the intermediate ( Bhiirata-sarhhita (BS),
1998) versions of the great epic edited by Prof. K. K. Shastree (Bambhania)
and published by the Gujarat Research Society, Ahmadabad. My intention
in discussing this subject is to note down points concerning the editing guidelines
and occasional printing defects in the editions and versions noted above.
I must confess that my study is based on a sample survey of a limited
part of the editions mentioned. A line by line checking of the whole material
is beyond the ability of a single person. If this test study inspires some young
scholar to undertake a more ambitious plan, I would consider my labour
rewarded.

The Critical Edition of the Mahabharata

The full-extent Mahabharata itself contains references to its growth, through

ages, from a nucleus of about 8800 verses (reference committed to App.
1.39A in the Crit. Edn.}, through the middle version of about 24,000 verses
(1.1.61 ), to the last extensive version of about 1,00,000 verses (reference
8 Krishna S. Arjunwadkar

committed to Critical Apparatus in the Grit. Edn. under 1.29.4 ). What the
BOAi edition has done is to determine the last version of the text on lhe
basis of an objective criterion, viz. the key manuscripts (MSS). A critical and
comparative study of the huge MSS material collected from all over India
leading to the fixation of their broad genealogy prepared the foundation of
the basic principles of editing this text. This resulted in reducing the size
of the text to 78.675 verses (the total number in the Critical Edition) including
the Harivamsa-parvan (6073 verses) which is considered for generations
as its appendix.(Figures of verses in BORI Critical Edition are taken from
the Introduction to the Ahmedabad edition, discussed later in this article.)

Critical Edition of Mbh : Some readings :

1) Explanation of the name Mahabharata

To start with, the name Mahabharata is explained in the text itself as

· mahattvad bharavattvacca mahabharatam ucyate' (BOAi edn. 1. 1. 209),
implying a reference to its value and extent ('heaviness'). While I can appreciate
the pun on the words 'bharala' and · bharavat', I consider 'bharavatlvat' as
a misreading for · bharatatvat' founded on the similarity between the Deva-nagarl
letters ·ta' and · va'. My reasons for holding this view are : ( 1) the Con.
Text is not such as would support a light pun, howsoever amusing ii may
be; (2) the name · Bha.rata' occurs in the Con. Text (' Bhflrata-samhita, 1.1.61)
for a version of this text. The full statement would then be: 'As ii is extensive
( mahal), and as it is bharata, it is called Maha-bharata.'

If, however, the line cited is not taken as a pun on the name of the
epic, and is taken as a genuine explanation, then, hypothetically, the name
of this epic was originally Maha-bharavat. This hypothesis conflicts with all
references in the epic to its being the history of the dynasty of Bharata,
leading to the name Bharata for a shorter version of the epic. This makes
it more reasonable to presume the original reading bharatatvat wrongly spelt
as bharavatlvat in the existing manuscripts and editions including the Grit.
Edn., as shown above. As far as I know, Textual Criticism d~es not shun
doors to reasoning when MSS evidence is not conclusive. The most flagrant
illustration on this point (praha satpuru~dhamafl) is going to be discussed
soon. After all, it is the editor who has to break up words with spaces in
the edited text while ancient Sanskrit MSS on the whole show break-ups
only at the end of half verses by inserting a vertical line ( dal)r,/a), single
or double.
There is another explanation of the name Mahabharata in the text itself:

bhciratanartJ mahaj janma mahabharatam ucyate I (I. 56.31)

(The great birth of the scions of Bharata is called Mahabharata.) The