Shiva Nataraja Shifting Meanings of An Icon
Shiva Nataraja Shifting Meanings of An Icon
Shiva Nataraja Shifting Meanings of An Icon
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms
CAA is collaborating with JSTOR to digitize, preserve and extend access to The Art Bulletin
This content downloaded from 103.25.231.2 on Sat, 29 Feb 2020 14:08:22 UTC
All use subject to https://about.jstor.org/terms
Shiva Nataraja: Shifting Meanings of an Icon
Padma Kaimal
((
7 (
~ ?
?~
~CE~i~~ ~
i?
r~ ~'~?::
~2
??
qB
~4! 1 ?
f W
4,
(.'''
~' Ir 3
?,
!? r
.e
'' '?
c-?'
?r
-? ~? ??
*?r
r. ~4~?
Y' L
j;
I ' ~ k? -~R
This content downloaded from 103.25.231.2 on Sat, 29 Feb 2020 14:08:22 UTC
All use subject to https://about.jstor.org/terms
SHIVA NATARAJA: SHIFTING MEANINGS O()F AN ICON 391
This content downloaded from 103.25.231.2 on Sat, 29 Feb 2020 14:08:22 UTC
All use subject to https://about.jstor.org/terms
392 ART BULLETIN SEPTEMBER 1999 VOLUME LXXXI NUMBER 3
?F
. or
A.A-
141#
ioi
trig,
AND
t?
pm~lp .
This content downloaded from 103.25.231.2 on Sat, 29 Feb 2020 14:08:22 UTC
All use subject to https://about.jstor.org/terms
SHIVA NATARAJA: SHIFTING MEANINGS OF AN ICON 393
Delhi (
anges R
PONDICHERRY
Kave River
'AREA OF DETAIL
Bay of Bengal
Kalugumala .
AVricddhachalam
Indian Ocean /er.
Kollidam R.
Udaiyargudi O TTiruviIa'kku*i
Tirumananjeri A j. unjai
Gangaikondcacholapuram M aram
Tirukkodik ava KtaLam Tirukkurukavur
Tiruvidaimarudur Tiruvauturai
Other temples
Other temples bearing
bearing minor large Nataraja images G-
Nataraja images
even
that dance, Shiva stands on his deeply bent right as while
leg the godhis
himself. As such it receives the honors one
would bestow
slightly flexed left leg lifts to waist height and crosses on the body of a king: it is awakened each
the hips.
morning,
Parallel to the dramatic transverse line of the lifted leg,bathed,
one ofoffered meals, delighted with music and
dance, and
his four arms crosses the body, its relaxed "elephant put to sleep in the evening. Here, as at other
hand"
(gajahasta) suspended above an equally relaxed ankle.
Hindu Shiva
temples, the icon at the temple's center embodies the
places one of his right hands just above that loose wrist
deity's in presence.
living the It must not be photographed.
Although
open-palm gesture (abhayamudra). Behind these boldlyin the present essay I use the Sanskrit term
dis-
posed limbs, the hips twist slightly to Shiva's Nataraja
right, but the "Lord of the Dance") for this image type,
(meaning
Chola sources
torso remains erect. Long, matted tresses alternating with refer to these images and the deity they
flower garlands stream outward from Shiva's head. On
represent his the language of southeastern India. There
in Tamil,
he is Adavallan or Kuttaperumanadigal, which also mean
right and among these locks sits a diminutive personification
of the river Ganga (the Ganges), her body human "Lord above
of the Dance."
the Modern authors also refer to him as
waist and piscine below. She presses her palms together in the
Anandatandavamurti, meaning in Sanskrit "he who perform
gesture of respectful address (anjalimudra). A the dance of furious
crescent moon, bliss."
This content downloaded from 103.25.231.2 on Sat, 29 Feb 2020 14:08:22 UTC
All use subject to https://about.jstor.org/terms
394 ART' BULLETIN SEPTEMBER 1999 VOL()IUME I.XXXI NUMBER 3
This content downloaded from 103.25.231.2 on Sat, 29 Feb 2020 14:08:22 UTC
All use subject to https://about.jstor.org/terms
SIIVA NATARAJA: SIIIFTING MEANINGS OF AN ICON 395
relief plane. In the other three reliefs, the dwarf and the
platform appear in profile. In those later reliefs, Nataraja is
flanked by figures playing percussion on a drum or clay pot
(ghatam) and, at Tiruverumbur only, by saluting celestial
attendants-all characters who seem part of the same time
and space as the dancing god-whereas symmetrical arrange-
ments of foliage and fantastic crocodilian creatures (makaras)
isolate the Tiruchchenampundi figure in a timeless vacuum
of ornament. Note that none of these flanking figures is yet
identifiable as the characters who attend Nataraja's dance by
about 971 at Konerirajapuram (Fig. 11): Bhringi, the danc-
6 Shiva D
ing, emaciated, and sometimes three-legged male devotee;
the main
Karaikal Ammaiyar, the fanged female devotee who plays the
Rajasimh
finger cymbals and who is also emaciated; or Tillai Amman,
the fleshier, multiarmed goddess who also dances.
The three-dimensional model that inspired the flatter
stone rendering at Tiruchchenampundi would almost cer-projectin
tainly have been made of bronze. The shapes of all Nataraja and filig
molten
figures strongly imply that this sculptural form was originally
conceived for the lost-wax method of bronze casting and metal. A
copied into stone only later, and with limited success. The
ill-suited
This content downloaded from 103.25.231.2 on Sat, 29 Feb 2020 14:08:22 UTC
All use subject to https://about.jstor.org/terms
396 ART BULLETIN SEPTEMBER 1999 VOLUME LXXXI NUMBER 3
-.q
u *"
I
I : jo.t"'. - J h,i%
relief on an ornamental crowning arc
(at upper center of photo), south wa
of the Sadaiyar temple, Tiruchchena
pundi, ca. 920 (photo: courtesy of
Artibus Asiae)
jor #11
lk?
8 Sh
just
crow
Nalt
This content downloaded from 103.25.231.2 on Sat, 29 Feb 2020 14:08:22 UTC
All use subject to https://about.jstor.org/terms
SHIVA NATARAJA: SHIFTING MEANINGS OF AN ICON 397
I k` ,
.. '.. ?
44
- SH~~1t47*.4 .
Nataraja's earliest representation must surely have been in Locating Nataraja: Chidambaram
Chidambaram, since this is the earliest site where we find him Since his genesis around the year 900 in sculpture and song,
in worship and this continues to be understood as his cultic Nataraja has been associated with Chidambaram, now a large
home, the place where the god actually resides. No such town 152 miles (244 kilometers) south of Chennai (formerly
figure of that age survives in Chidambaram, and the Nataraja called Madras) near India's east coast (Fig. 4). The faithful
bronze currently under worship there, by Barrett's estima- understand Nataraja as a living entity and Chidambaram as
tion, is no older than the thirteenth century.25 But Chidam- the home in which he physically resides.26 Chidambaram's
baram has been famous since the seventh-century poems of temple complex (Fig. 13) now covers some fifty-five acres.
Appar for enshrining a dancing figure of Shiva instead of the The specific structure in which Nataraja dwells, in his bronze
abstract, cylindrical stone symbols of Shiva (linga) that serve sculptural form, lies within the innermost enclosure wall of
as the central object of worship in most Shiva temples. this complex and is called the Hall of Consciousness, the
As its earliest known stone "copy" was carved at Tiruchche- chid-sabha (Fig. 14).
nampundi about 920, the original Chidambaram Nataraja By the seventh century the hymnists Appar, Sambandar,
would have been in place by then, but perhaps not for long, Sundarar, and Manikkavasagar had sung of Chidambaram's
given the unfamiliarity with details of Nataraja's iconography already ancient sanctity.27 They called the site Tillai and
registered in the early stone reliefs. At any rate, the form Puliyur ("Town of the Tiger" in Tamil), for the legendary
seems to have been new to stone. Bronze "copies," such as hazards of the place, a forest of poison-sapped tillai trees and a
perhaps the Karaiviram bronze, probably reprised the Chidam- fierce tiger that roamed the precinct. These devotional poems
baram icon more precisely and slightly earlier, being of the also indicate that multiple deities were worshiped there.
same material as the original icon. The songs of traveling Some songs mention a hall to Vishnu there by the eighth
hymnists like Appar would have acquainted audiences outsidecentury.28 Legends about a goddess of Tillai and a sacred
Chidambaram with its anthropomorphic, dancing god, allow- pond may also be quite old.29
ing such copies in bronze and stone to evoke his cultic home The structures of the Chidambaram complex also indicate
explicitly for such viewers. I am suggesting, then, that Nata- that this place has served as home to other deities for at least
raja images throughout the Tamil country were intended to as long as it has for Nataraja.30 Sharing the most sacred space
function as deliberate recapitulations of the original cult icon of the inner compound with Nataraja's chid-sabha is a shrine
at Chidambaram. for Vishnu as the cowherd Krishna (also called Vishnu-
This content downloaded from 103.25.231.2 on Sat, 29 Feb 2020 14:08:22 UTC
All use subject to https://about.jstor.org/terms
398 ART BUll.ETIN SEPTEMBER 1999 VOIUME LXXXI NUMBER 3
?? ..+~
!~tlp?
j i
.J
~`
r%
10 Shiva Nata
Attendants, g
lintel of an or
wall of the Pi
ca. 952
This content downloaded from 103.25.231.2 on Sat, 29 Feb 2020 14:08:22 UTC
All use subject to https://about.jstor.org/terms
SII\'VA NATARAJA: SIIIFTING MEANINGS OF AN ICON 399
;k ~t?r,
?t~l 7??
ii r !'r
.. ? q
.,i ?-?
?-;
'? ~
'-,.
Il
'1
~d~ef?rO ; L d
?r ?'
s\ ~ -C~d?
?r
'r
'? ' )?
4,;
rAr
..:
i??
ii
r", .
-f~;
crowning its ridge pole. The roof is constructed of wood and recapitu
sheathed in gold-colored metal. The interior space of the built be
chid-sabha extends from east to west, perpendicular to the wooden
direction in which the doorway invites visitors to approach.40 the chid
In its humped roof, the chid-sabha's nearest surviving relative the nint
is the seventh-century Draupadi Ratha in Mamallapuram mented
(Fig. 16), which is, however, completely of stone and square in of Shiva.
plan. The lower edges of its humped roof do not flare. In A mural in King Rajaraja I Chola's temple, the Rajarajesh-
Kerala, too, one finds temples with sloped wooden roofs vara in Tanjavur, demonstrates that by 1010, when that temple
sheathed in metal, though these are considerably younger was consecrated, the chid-sabha at Chidambaram already had
than the chid-sabha.4' its unusual shape (Figs. 18, 19). This dry fresco, which is the
Several other features reinforce the impression that Chidam- earliest painting I know depicting Shiva as Nataraja, portrays
baram's original shrine, of which the present chid-sabha is one him under the Chidambaram temple's distinctive humped
in a long series of selectively faithful "copies," was built and golden roof (compare Figs. 18 and 14). The mural even
before the tenth century. The main walls, uprights, and captures the slightly flared eaves and the roof's fine vertical
beams are made of wood,42 the material that inspired and striations. The mural includes a man and three women
thus preceded the lintel and pilaster decorative motifs of worshiping Nataraja from under a similar though slig
tenth-century stone architecture (Fig. 17). The humped roof lower roof (Fig. 19), indicating that the subordinate kana
This content downloaded from 103.25.231.2 on Sat, 29 Feb 2020 14:08:22 UTC
All use subject to https://about.jstor.org/terms
400 ART BUIILLETIN SEPTEMBER 1999 VOLUME LXXXI NUMBER 3
Shivaganga
Tank
Shivakamasundari
Temple
d~ East Gopura
chid-sabhha
West Gopura
nritta-sabha
This content downloaded from 103.25.231.2 on Sat, 29 Feb 2020 14:08:22 UTC
All use subject to https://about.jstor.org/terms
SHIVA NATARAJA: SHIFTING MEANINGS OF AN ICON 401
Esinuhri-L
1L
E aa]hrir
oha Cnndeshva
Govin araja
shrine
Vishnu-rI
E I r---,--
This content downloaded from 103.25.231.2 on Sat, 29 Feb 2020 14:08:22 UTC
All use subject to https://about.jstor.org/terms
402 ART BULLETIN SEPTEMBER 1999 VOLUME I XXXI NUMBER 3
'.r r?':J
iPj~bf
~fc:
~~D~jr
I ?"
features are part of Nataraja's visual form. In the language of r.:t u;P:
`!ii
tA
Indic dance both then and now, deep flexions of the joints ?c~
?c,
ii: ?~
!cr .e
leg, are lethal vermin that haunt deserted places. The flames 17 Lintel and pilaster motif and the triple niches of an
embracing the entire composition and the smaller flame in ardhamandapa funded by Sembiyan Mahadevi, south wall,
his upper left hand could allude to the funeral pyre. The skull Apatsahayeshvara temple, Aduturai, ca. 985
This content downloaded from 103.25.231.2 on Sat, 29 Feb 2020 14:08:22 UTC
All use subject to https://about.jstor.org/terms
SHIVA NATARAJA: SHIFTIN(; MEANINGS OF AN I(()ON 403
L4
?~-~s
C
PL~-~ii~~
d '~
~ Lr I ~F
i??vc!:e 1?
c? I
-r
~ ?~
..
18 Shiva i-
Nataraja
Enshrined in the
Chid-sabha, fresco ??: ?
?;3:
mural in the western
chambers of the inner
ambulatory, Rajara-
*rr
jeshvara temple
10
elephant-headed Ganesha who attend Shiva's dance in these appears beneath Nataraja's raised foot in reliefs at Koneriraja-
scenes also appear in representations of his home in puram the about 970 and Gangaikondacholapuram in the mid-
Himalayas and his adventures in various places. The seated eleventh century (Fig. 12), he has been understood as
Bhringi, a passionate devotee whose exclusive attentions to
musicians in these early reliefs fix only the associations with
dance. A small, emaciated dancing figure is present at someShiva
of earned him the Goddess's jealousy and her curse that
these early reliefs, and Coomaraswamy may have taken him he lose his flesh."2 Perhaps such figures were intended to
for one of the starving demons (pishacha) who accompany represent starving ghosts at Sirpur, Ellora, Elephanta,
Shiva's wrathful Bhairava manifestation and who haunt the Mukhalingam, and Kundaveli,63 and thus inspired the story of
cremation ground in search of human flesh. Such demons
Bhringi and later the story of Karaikal Ammaiyar, the emaci-
ated female devotee who accompanies Nataraja's dance on
accompany Bhairava and Chamunda, wrathful manifestations
of Shiva and the Goddess.61' When this emaciated figure
her finger cymbals.64 But this possibility is slim evidence on
This content downloaded from 103.25.231.2 on Sat, 29 Feb 2020 14:08:22 UTC
All use subject to https://about.jstor.org/terms
404 ART BULLETIN SEPTEMBER 1999 VOLUME LXXXI NUMBER 3
The Cholas
notes that the Chidambaramahatmya, the legend book com- points west of India.69 The Cholas played an increasing
posed by temple priests at Nataraja's cultic home between thedominant role in south Indian politics by the eleventh
tenth and twelfth centuries, refers to Apasmara as a dwarfcentury, the height of their dynasty's power, when Chola king
(bhuta) rather than a demon. At the center of Nataraja's cult,claimed hegemony from the southern tip of the subcontinen
This content downloaded from 103.25.231.2 on Sat, 29 Feb 2020 14:08:22 UTC
All use subject to https://about.jstor.org/terms
SIIVA NATARAJA: SHIF'TIN(; MEANINGS OF AN ICON 405
~t~ L?
.L1
Yt ~.~i c~ r-,
~??r-?i; IIIIL-SC~l~~i.~;Llc~-?
:3C
U
~5;~*)
~' c c. I u r. ... ?.
,,~ __ ~ r~ ?.
" ri ;t?i ~
rZ
?? I :? ,
: r
C
.e
*
r
r?
..
a r .e
This content downloaded from 103.25.231.2 on Sat, 29 Feb 2020 14:08:22 UTC
All use subject to https://about.jstor.org/terms
406 ART BULLETIN SEPTEMBER 1999 VOLUME LXXXI NUMBER 3
of Shiva known across India through Sanskrit texts (puranas) Nataraja icon. This myth made that icon an emblem worthy of
and distanced him from local, wrathful deities. Kulke suggests kings, with greater emphasis on his royal qualities than Tillai's
that the Chola kings' selection of Tillai's dancing god for their earlier literary versions of Shiva had displayed. The objects
emblem catalyzed these revisions.80 surrounding Nataraja became potential threats, triumphantly
A major vehicle for these revisions was the Chidambaramahat- overcome. By dancing with these trophies, he exhibited an
mya, a priestly redaction of old and new stories designed to easy confidence in his ability to meet the angry sages and their
present Chidambaram as a worthy destination for pilgrims.81 weapons, which Zvelebil reads as emblems of rival Tillai
Many South Indian temples had mahatmyas, in oral and deities.86 In this way, Nataraja's achievements paralleled the
sometimes written form, which constructed a temple's history tasks that lay before early kings of the Chola dynasty: the
and traditions out of local legends, historical events, and subjugation of rival kings and the establishment of a geo-
appropriated myths, edited and reframed to serve the contem- graphic and psychic center for the vast dominion they aspired
porary interests of its compilers. Much of the Chidambarama- to claim. In royal as well as cultic terms, Nataraja could serve
hatmya was written down about the twelfth century, but as a god of conquest, brandishing his challengers' failed
portions of it had been composed over the preceding two weapons much as kings would exhibit regalia they had looted
centuries or more.82 from their defeated rivals.87
The Chidambaramahatmya adapted the older and more Kulke identifies other transformations effected in the
widely known Myth of the Pine Forest to suit the chid-sabha's Chidambaramahatmya that would have enhanced this temple'
Nataraja icon. This story entered the Patanjali section of theattractiveness to pilgrims and royal patrons. Sanskritization
text probably during the tenth century and, in Kulke's view, localis place names made them easier for outsiders to recogni
the earliest textual reference to Nataraja's dance in its fullyand approach. This text presents Chidambaram as a new
canonic form.83 According to the Chidambaramahatmya's recen- name for Tillai-Puliyur. Chidambaram, a Sanskrit term mean
ing "the heavenly abode of the Spirit," replaced the similar-
sion of this myth, Shiva visited sages, or rishis, living in a pine
forest to punish them for the inadequacies of their devotions. sounding Tamil word citrambalam, meaning "little hall,"
He came as the naked, handsome beggar Bhikshatana, term that had for some time labeled one of the early
dancing, laughing, and making himself sexually irresistible to structures in that complex.88
the sages' wives. He brought with him Vishnu in his seductive That text gave Tillai yet another Sanskritizing name b
female form, Mohini, who drove the sages to distraction. interpolating the character Vyaghrapada into Nataraja's story
Perceiving that Bhikshatana had thus humiliated them, the A minor character from the Sanskrit Mahabharata epi
sages pelted him with weapons: a tiger, a skull, a drum, snakes, Vyaghrapada enters the Chidambaramahatmya when he com
fire, and Apasmara, the demon of ignorance. Shiva easily to the tillai forest to practice asceticism. He prays for tiger
overcame all these threats, turning them into trophies to feet so that he may climb trees and harvest their best flowe
adorn his body or to dance over in fierce joy. His dance took for Shiva. Those feet give him his name, Vyaghrapada, whic
on cosmic dimensions, drawing into itself all of creation. The means "tiger foot" in Sanskrit; his name gives the place a new
sight of his beautiful, awesome dance enlightened the sages. name, Vyaghrapura, Sanskrit for "town of the tiger." Vyag
At Shiva's urging, the sages worshiped him in his aniconic (or rapura was, then, a Sanskrit translation of the Tamil nam
abstract) columnar form, the linga.84 Puliyur. The name Vyaghrapura also integrated the town in
The Chidambaramahatmya thus attaches a mythological the sacred geography of the Mahabharata, through its associa
explanation, as it were, to Nataraja images, adapting them tions with a character from that epic. A descriptive local ter
both to a narrative of conquest, supremacy, and absorption. thus became a link to a pan-Indic legend.89
The Myth of the Pine Forest rationalizes Nataraja's icono- The prominence of tigers in Tillai's reputation-as an
graphic markers as the sages' futile threats: the skull in his ancient totem, as a weapon of the sages, and as the garment
hair, the snakes around his arm and in his hair, and the dwarf around Nataraja's hips-may have enhanced Chola intere
under his foot. Fine stripes incised in the little garment in this place. The tiger had been a Chola emblem for as long
around Shiva's hips depict the remnants of the attacking tiger. as anyone could remember. It adorned the flags, coins, an
Shiva catches the fire calmly in his hand and permits it to the royal seal of the Chola kings.90
envelop his entire body as the aureole. He uses the drum to The Chidambaramahatmya further described the ancien
beat the rhythm of his victory dance. In all these ways, he temple town as a center for widespread pilgrimage by defin
safely absorbs their weapons into his body and meets the ing Chidambaram as the center of the universe and of ever
sages' challenges in unruffled triumph. human heart, and as the only site at which Nataraja wou
This new story could distance Shiva's dance from the perform his beatific ananda-tandava dance. Vyaghrapada and
cremation ground, transmuting what may have been a dance another sage, Patanjali, resided in Tillai yearning to see this
about death into a dance about triumph. And yet Shiva dance, which would fulfill the spirit of all who witnessed i
remained energetic and aggressive, still not quite the tranquil Nataraja agreed to perform the dance there and only the
deliverer to be found in thirteenth-century texts. Perhaps the for eternity. This story constructed Tillai-Chidambaram as t
sharpness of that contrast was what deflected Coomaraswa- center of all devout concentration, the core of the psychicall
my's attention from the significance of the Pine Forest perceived universe, integral to the world around it. Thi
legend, which, he concluded, had "after all, no very close devotional philosophy asserted a universal significance for th
connection with the real meaning of the dance."85 town and its new god, as well as for the priests and kings wh
I would agree instead with Zvelebil and Kulke that this myth protected his temple.
played a significant part in transforming the meaning of the When Shiva answered Vyaghrapada's request to perform
This content downloaded from 103.25.231.2 on Sat, 29 Feb 2020 14:08:22 UTC
All use subject to https://about.jstor.org/terms
SHIVA NATARAJA: SHIFTING MEANINGS OF AN ICON 407
the ananda-tandava dance in Tillai, the goddess who already domesticated and overcome a goddess much like the on
reigned there, Tillai Amman, resented the encroachment on associated with his still-localized scope of kingship.
her precincts and challenged Shiva to compete through The placement of structures at Chidambaram implies that
dance for spiritual hegemony over that place. Shiva defeated an interaction between a dancing male god and a godde
her by taking up the raised leg (urdhva-janu) pose that may have been an early part of religious practice there.
embarrassed the virginal goddess and that modesty forbade goddess shrine, the nritta-sabha, stands just south of Nataraja
her to take up herself. In defeat, she was bifurcated. In her chid-sabha and kanaka-sabha and opens toward them. Younge
fierce, virginal aspect she ceded the Tillai grove to Shiva, argues that in eleventh-century inscriptions the Tamil name
withdrawing to a shrine outside the walls of her former for the nritta-sabha was "the opposite hall" (etirampalam
temple, while her benign aspect was domesticated to the role stressing the significance of its juxtaposition with the chid-
of Shiva's wife and housed in the chid-sabha, where she still sabha.94 This would suggest that the innermost precinct wall,
which now cuts between the nritta-sabha and the chid-sabha,
watches in adoration his performance of the ananda-tandava
dance. was introduced deliberately to establish Nataraja's indepen
dent significance by disguising an earlier architectural di
This segment of the Chidambaramahatmya narrative exposes
logue between his shrine and this goddess shrine.
the theme of competition among deities, a recurrent theme
at Chidambaram and one surely stimulated by the long Spatial relations at the temple also imply the bifurcation of
tradition at that site of multiple deities in worship sidea by
defeated goddess. Shivakamasundari, the adoring, prett
side.91 Shulman notes that the Chidambaram version of the appears inside the chid-sabha with Nataraja, while a
wife,
Dance Contest story subjugated the goddess with special eight-armed goddess bearing a trident stands in the nritta-
sabha beside an image of Shiva who dances in the raised-leg
force. Shulman argues that the narrative fate of Tillai Amman
posture through which he defeated Tillai Amman in th
articulated a broader tension between Tamil religious ideals
dance contest. These divergent halves of the goddess also
and those of the Sanskritic tradition of the Upanishad texts.
appear to have been later manifest through larger structure
Tillai Amman represented Tamil mythology's preference for
the Shivakamasundari temple, which lies within the walls of
highly localized, distinctive deities and for mythic action that
the temple complex (Fig. 13), and the Kali temple, which lies
tied them to particular places in the Tamil countryside. In
outside the temple walls at the city's northern edge.95
opposition to this stood an Upanishadic search for universal
Thus, Nataraja is in Chidambaram's literature embedde
truths, for cosmic rather than local geography, and for an
in myths that transform him, Tillai, and its legends, and se
all-encompassing deity who could embody multiple selves
him atop a hierarchy among its multiple deities. Those
within a unified, supreme whole. Shiva could embody Upani-
revisions displaced or transformed local deities, or integrate
shadic ideals more readily once the marginality and violence
them into pan-Indic mythologies. Traces of those revision
with which he was associated transferred to Tillai Amman; in
may be still visible in the Chidambaram temple complex. In
defeating her, he exiled those "primitive" characteristics
their redesigned forms, these deities constructed Nataraja a
beyond the sacred precinct.92
a more appropriate emblem for royal patrons.
This literary transformation of Tillai's deity may have
mirrored a comparable shift in the reception of the temple's
bronze icon as well. That icon's wild dance and fire may have
Nataraja as a Chola Emblem
suggested the marginality and violence of older Tamil deities.
This strategy may have played a role in forging the strong
The Chidambaram versions of the myths of the Pine Forest bonds that would connect Chidambaram and the Cholas
and the Dance Contest could then have neutralized these
from the eleventh to thirteenth centuries. Whatever the
associations. The Myth of the Pine Forest cast the fire causes,
and thethose bonds clearly existed by the reign of Rajaraja
dance as part of Shiva's calm response to the threats ofChola
foolish
(r. 985-1014), and images of Nataraja constituted o
sages, and it identified Chidambaram's god as a later form of through which Rajaraja expressed that affiliation.
language
the more widely known Bhikshatana. The story of the Dance
propose that the Chola queen Sembiyan Mahadevi was
Contest then aligned Bhikshatana/Nataraja againstRajaraja's
the an-predecessor in deploying the Nataraja image as a
cient Tamil goddess of the locality and credited him with
Chola emblem. The dozen temples she had built between 970
exiling her from the sacred precinct. and the early eleventh century were the first to feature
The incorporation of Tillai Amman into Nataraja's storyin full-scale wall niches. Their geographic distribu-
Nataraja
could also have strengthened Chola identification with Nata-
tion across the deltaic region the Cholas aspired to dominate
raja. As a fierce goddess of local and ancient fame, Tillaithat one important function of her temples was to
suggests
Amman shared several features with the goddess Nishumbha-
expand the fame of the Chola dynasty in the localities. I see
sudani, possibly the favored deity of Vijayalaya Chola, the
the prominent Nataraja sculptures on their walls as playing a
earliest known king of his line who ruled during the role
thirdin that expansion and in the innovative iconographic
quarter of the ninth century. If Nagaswamy's identification
programs of her temples introduce.96
In proposing a political role for this group of Nataraja
Vijayalaya's goddess is correct, both goddesses were powerful
and wild, and aspects of their appearance were horrific. Both
images, I share the theoretical orientations of Kulke, Nicholas
goddesses had eight arms and in one of these wielded Dirks,the
and Arjun Appadurai, as well as their conviction that
kings
trident.93 Vijayalaya's descendants' ambitions to expand be-and temples were closely connected during much of
yond the geographic confines of his small kingdom may India's
have
history.97 As those two institutions produced much of
the art
encouraged them to identify with a god-Nataraja-who hadsurviving from medieval times, interconnections
This content downloaded from 103.25.231.2 on Sat, 29 Feb 2020 14:08:22 UTC
All use subject to https://about.jstor.org/terms
408 ART BULLETIN SEPTEMBER 1999 VOLUME LXXXI NUMBER 3
r
?;
which earlier temples of the region placed
rK?~ ?c-
Shiva that are not dancing.105
Rajaraja's declarations of devotion to Nat
edented in their scale, but earlier Cholas h
~
r;?
this deity. Gandaraditya Chola (r. 949-5
hymn to the dancing god of Tillai.o06 I w
?P
?u
+ ~1M~~CrC~s7s~-c, ? T;
*-
wife's prodigious sponsorship of templ
~'''
participated in the Cholas' pronouncem
Nataraja. Queen Sembiyan Mahadevi beca
a .?
to invest heavily in the construction of t
,,
~Y.' ~C
*.
.?r ??
of the twelve temples attributable to her
appeared for the first time as a full-sc
carved essentially in the round and housed
niche (Figs. 11, 2). In her 981 temple in Se
~ El ~3; village, he dominated that wall by occupyi
r ,?~15 4 ii In the same year, the iconography of these
;JI r 't
the form that would persist for Nataraja im
1L~IP"'
LL
1 ?j)f
like figure of Ganga, which at her earlier t
970s) was missing or on Shiva's left side, c
her standard
22 Shiva Nataraja, granite, in the westernmost niche, place
southon his right side.107
wall
of the Rajarajeshvara temple Sembiyan Mahadevi's inscriptions on t
rituals they prescribed, and the many gif
lavish on the enshrined deities guaranteed
between the political and the sacred the are likelywhere
localities to have she built her temples
influenced art considerably.98 temples with the Cholas.s 08Nadu residen
recognized the
Rajaraja I Chola manifested his devotion to Chidambaram Nataraja figures as refe
baram thanks to the
Nataraja clearly in sculpture, painting, and inscriptions at his poets Appar, Sundar
personal temple, the Rajarajeshvara in Tanjavur, which he
who had sung of Tillai's dancing god and
through the
had built between 1003 and 1010.99 Featuring Kaveri
Nataraja atregion
his these towns sti
Nataraja's
royal temple in the Chola capital, Rajaraja placedunprecedented
the dancing prominence on
god of Chidambaram at the geographic devi's temples
center of the could
king's then have signaled
dents an and
political realm, declaring Chidambaram's association between himself, Chid
Nataraja's
Cholas. His
importance in the world of Chola politics.100 The presence could declare the C
mural in an
Chidambaram,
ambulatory aisle at that temple depicts Nataraja being and his sculpted image c
wor-
emblem of
shiped by a man and three women (Figs. 18,the Cholas.
19). The deity is
thus ensconced close to the temple'sThe prominence
sacred and consistency
center, against with which Sembiyan
the very wall that encloses the stoneMahadevi's
linga in temples
whichincorporate full-scale stone Natarajas set
rituals
would incarnate Shiva himself. The themdistinctive
apart from temples humped,
built around the same time by other
golden roof over him and the similar patrons.
but Twelve
slightlyof herlower
thirteen roof
temples (Fig. 23) 109 present
over his royal devotees locate the Nataraja
sceneas unambiguously atniche on the south wall
a large figure set in its own
of the small
Chidambaram. If, as most scholars suspect, vestibule
the male (ardhamandapa)
figure that leads to the inner
represents Rajaraja himself,1'0 theshrine
painting fixes
(vimana). (The himtemple
thirteenth in is well north, in the
Tondaimandalam region,
eternal worship of Nataraja and demonstrates where Nataraja is absent from
that Rajaraja
had chosen Nataraja as his favored, personal
temples built
deity.'102
during the second half of the tenth century by
Cholas that
According to inscriptions at this temple and non-Chola
refer topatrons
him alike.)
as In contrast, temples
This content downloaded from 103.25.231.2 on Sat, 29 Feb 2020 14:08:22 UTC
All use subject to https://about.jstor.org/terms
SHIVA NATARAJA: SHIFTIN(; MEANINGS OF AN ICON 409
Sembiyan Konerirajapuram
Mahadevi Anangur
(confirmed) Tirukkodikaval
Sembiyan Mahadevi
village
Aduturai
Tiruvarur
Kuttalam
Vriddhachalam
Sembiyan Mayuram
Mahadevi Tirunageshvaram
(attributed) Tirumananjeri
Tirunaraiyur
TONDAIMANDALAM:
Sembiyan Tiruvakkarai
Mahadevi
Mahadevi's patronage
This content downloaded from 103.25.231.2 on Sat, 29 Feb 2020 14:08:22 UTC
All use subject to https://about.jstor.org/terms
410 ART BUI.I.ETIN SEPTEMBER 1999 VOLUME IXXXI NUMBER 3
?. .;; , c
Ir k?
40'?~'
SO-
A-
This content downloaded from 103.25.231.2 on Sat, 29 Feb 2020 14:08:22 UTC
All use subject to https://about.jstor.org/terms
SIIIVA NATARAJA: SHlIFTING MEANINGS OF AN ICON 411
top.
i. 4
AL fi?
i.ed
tana's exposed genitals, which also captured the gaze of the south gopura (towered gateways) at Chidambaram establish a
sages' wives. Diminutive figures carved beside his niche counterpoint between Shiva's married forms (Kalyana-
represent these smitten wives, who loosen their clothing as sundara and Somaskanda), which face north, and his unmar-
they gaze upward at the beautiful mendicant (Figs. 27, 28). At ried forms (Bhikshatana and Kankala), which face south. The
some temples, water pots abandoned at their feet evoke the precise orientation of Shiva's various manifestations vary
domestic responsibilities lust has driven from their minds."116 among these examples, but the opposition of north and south
While Bhikshatana is the narrative precursor of Shiva's remains a consistent device for articulating the domesticated
dance of furious bliss in Chidambaram's myths, that sequence or independent nature of Shiva's sexuality.
is inverted on the walls of Sembiyan Mahadevi's temples, Nataraja, Bhikshatana, and Lingodbhava evoke Chidam-
where worshipers encounter Nataraja near the beginning of baram in yet another way on Sembiyan Mahadevi's temples.
circumambulation and Bhikshatana on the last wall they
While Nataraja and Bhikshatana radiate from the temple core
meet. One reason for that inversion may lie in a desire to
toward the south and north, Shiva as Lingodbhava faces to the
engineer the viewer's experience of the temple. Whereas
west, and the Shiva-linga within the temple sanctum emanates
Nataraja's aspect is distant and awesome, absorbed in blissful
toward the east through the temple entrance."19 Through this
frenzy, Bhikshatana reveals an irresistibly personal face. Such
arrangement, Shiva recalls the Nataraja of the Chidambarama-
a progression from the remote to the enticing could draw the
hatmya, who radiates in all directions from the sacred center.
worshiper into a state of emotional and psychological receptiv-
Sembiyan Mahadevi's many temples propagated Nataraja's
ity ideal for one who is about to enter the temple and
cult with an intensity that was new to the Kaveri region, but
experience the charged intimacy of the god's home.117
even they may not have been the first to use Nataraja icons as
The spatial opposition between Nataraja and Bhikshatana
Chola emblems. Before 945, her father-in-law, Parantaka I
on Sembiyan Mahadevi's temples invites one to recognize
contrasts between their characters. In his languid posture and Chola (r. 907-54), funded completion of the Gomuktishvara
nude beauty, Bhikshatana embodies a charged sexuality temple in Tiruvaduturai, where he may have been responsible
outside marriage. Nataraja, who is enshrined at Chidam- for the small Nataraja relief discussed above (Fig. 9).120 This
baram beside his wife, turns his back on Bhikshatana at these figure is cut in much higher relief than are the early Nataraja
Kaveri temples as he inspires the devotion of the ascetics reliefs at Tiruchchenampundi, Tiruverumbur, and Punjai,
Bhringi and Karaikal Ammaiyar. David Smith notes that a suggesting that the god was especially important to Tiruvadu-
tension between Shiva's married and unmarried forms also turai's patron. It is also the only relief of the four to depict an
umbrella over Nataraja's head, frequently a marker of royal
permeates Chidambaram's temple complex and myth. Shiva's
ananda-tandava dance celebrates the ascetic's victory over
status in Indian art.121 In having Nataraja carved in deep relief
and on a vimana wall, Parantaka would have publicized
married sages, and yet Shiva receives worship in the chid-sabha
as a married man."8 Smith also observes that the north and that new god where that deity had not yet become a com-
This content downloaded from 103.25.231.2 on Sat, 29 Feb 2020 14:08:22 UTC
All use subject to https://about.jstor.org/terms
412 ART BULLETIN SEPTEMBER 1999 VOLUME IXXXI NUMBER 3
w y~*e
I~ 1)
jl ~~3;r ,
,U. ,...
.
Chidambaram's fame, as this in turn could spread the fame of
-?: 2., Nataraja's royal protectors across the increasingly large re-
?)
It.fi r
'I
gion over which the Cholas claimed political authority.
L?.? C 2-
?u:~
4r ~ ? '?
:-i
r
C;i
h?
?~~c?-' Nataraja would thereby have exemplified what Kulke calls a
`?' ~1~I)LLh~ "royalizing" deity, one whose identity and worship contrib-
" ?.
,rt
L1
~i?
r
uted directly to the reputation and authority of the affiliated
''~!l1??y: king.130
~ rllrl
This content downloaded from 103.25.231.2 on Sat, 29 Feb 2020 14:08:22 UTC
All use subject to https://about.jstor.org/terms
SHIVA NATARAJA: SHIFTING MEANINGS OF AN ICON 413
crest of kondrai leaves completes the radial arc across the top
of the composition. Long sashes flutter toward his lower left
from a cord tied just above the waist, and the suspended leg
angles toward his lower right. The aureole around Nataraja
ties these radials together, often anchoring the flying sashes,
the outstretched hands, and the kondrai ornament. Small
clusters of flame stud the circle, projecting the composition's
_ -N. N- ,
I
4:Z
the drum-bearing arm. The loose sashes also trace the spokes
of this wheel, either horizontally or along a diagonal to the
lower left. The upper body and neck may tilt to one side, but
they tend to do so along a radial axis. Flexion of the arms and
the standing leg could be rationalized as a means for fitting
this body into a circle radiating from the navel: these elegantly
long limbs are thus shortened as the knot of matted hair and
29 D
the cluster of kondrai leaves extend the radial defined by the
com
head and upper body.134 Juli
Though I have mapped a wheel-shaped diagram onto only
one Nataraja sculpture, I perceive a similar structure underly-
ing every Nataraja image I have studied, as long as I allow for Pur
reasonable flexibility within that pattern. I suspect that then dan
as now artisans would have had room to extemporize, using The
their judgment and adjusting to the demands of materials. of t
Yantras would have served as guides, not as legislation. The core
bends in the arms and the raised leg may then align the they
forearms and the shin along other radials extending from the the
navel. Sashes may be absent; they may articulate radials that dom
are not symmetrical with the raised leg; or they may articulate out
radials with only some of their undulating lines. Some exp
aureoles are oval or almond-shaped instead of round. They Cho
stretch downward to encompass the prostrate dwarf, a portion lele
of the platform, or a longer or less deeply flexed leg on Shiva. beco
But even in these cases, the aureole's upper section usually logi
remains circular and focused on the navel.135 gal
This wheel-shaped structure is, furthermore, generally thr
absent from representations of Shiva's dance before the tenth that
century (for example, Figs. 5, 6).136 Few of these lift the leg draw
across the hips (bhujangatrasita pose), depriving those compo- bara
sitions of the radial defined by that suspended shin. Torsos
lean away from the weight-bearing legs rather than toward Rer
them. The arms, when they do bend, rarely align the forearm Sem
with the navel; neither are fluttering scarves thus aligned. Few acro
such figures display flying locks or the fiery aureole. The atic
radial composition may be one of the innovations that capa
the Cholas and Chidambaram held in common could have
distinguish Nataraja from preceding forms for depicting
Shiva in dance.
made Nataraja icons particularly effective as emblems of the
The radial composition might even have been one factor
alliance between these parties. As Nataraja's presence ex-
that made the Nataraja icon work well as an emblem of the panded throughout the cosmos along the lines of his body, so
Cholas, of the Chidambaram priestly community, and of the the king's presence permeated the Tamil plain, and the
alliance between them. The expansive force of its lines could
reputation of a holy city reached beyond the Tillai forest. This
express ambitions of both groups simultaneously, just as radial
the shape may have initially encouraged the Cholas' inter-
This content downloaded from 103.25.231.2 on Sat, 29 Feb 2020 14:08:22 UTC
All use subject to https://about.jstor.org/terms
414 ART BULLETIN SEPTEMBER 1999 VOLUME LXXXI NUMBER 3
est in Chidambaram, or it may have evolved over the course of Coomaraswamy, Ananda K., "The Dance of Shiva" (1918), in The Dance of
Shiva: Fourteen Indian Essays, 2d ed. (Munshiram Manoharlal, 1970), 83-95.
the tenth century as a response to the Cholas' increasing Dehejia, Vidya, Art of the Imperial Cholas (New York: Columbia University Press,
commitment to Chidambaram's dancing god. 1991).
This political meaning could have come to function for an Goswamy, B. N., Essence ofIndian Art (San Francisco: Asian Art Museum of San
Francisco, 1986).
icon that had previously signified a very different aspect of
Harleames C., Temple Gateways in South India (Oxford: Bruno Cassirer, 1963).
Shiva's nature. The bronze icon may have at first continued to Kaimal, Padma, 1996, "Early Cola Kings and 'Early Cola Temples': Art and the
represent the frenzied, destructive god whose worship at Evolution of Kingship," Artibus Asiae 66, nos. 1-2 (1996): 33-66.
Kulke, Hermann, Cidambaramahatmya, Freiburger Beitrdge zur Indologie, vol. 3
Tillai was already ancient. The first such Nataraja image- (Wiesbaden: Otto Harrassowitz, 1970).
surely the original cult image enshrined in an ancient , 1978, "Royal Temple Policy and the Structure of Medieval Hindu
precursor of the chid-sabha in Chidambaram-may have Kingdoms," in The Cult of agannath and the Regional Tradition of Orissa, ed.
carried these associations when it was made about 900 C.E. A. Eschmann, H. Kulke, and G. C. Tripathi (New Delhi: Manohar), 125-37.
Kulke, Hermann, and Deitmar Rothermund, A History of India (London:
That sculptural form may have continued to mean this forHelm, 1986).
Croom
some viewers, even as members of the Chola family Lippe, began to "Review of Early
Aschwin, Cola Bronzes by Douglas Barrett," Artibus Asiae
30, nos. 2-3 (1968): 267-74.
read it as their own. Then, in another semiotic shift, the
Meister, Michael, ed., South India: LowerDrdvidadjia 200 BC-AD 1324, vol. 1, pt.
meaning with which modern audiences are most familiar 1 of The Encyclopaedia of Indian Temple Architecture (New Delhi: American
accrued to this sculptural form by the thirteenth century, and Institute of Indian Studies; Philadelphia: University of Pennsylvania Press,
1983).
the elegant abstractions of that reading gradually displaced Nagaswamy, R., "The Temple of Nataraja (Chidambaram)," in Perceptions of
the icon's earlier associations with gruesome rampage. South Asia's Visual Past, ed. Catherine Asher and Thomas Metcalf (New
As Coomaraswamy's appealing reading of the Nataraja Delhi: American Institute of Indian Studies and Oxford and IBH; Madras:
Swadharma Swarajya Sangha, 1994), 179-85.
image type has probably had much to do with these images' Peterson, Indira, Poems to Siva: The Hymns of the Tamil Saints (Princeton:
present fame and abundance, it becomes especially important Princeton University Press, 1989).
to understand the ways in which his essay oversimplifies and Rangacharya, V., A Topographical List of the Inscriptions of the Madras Presidency,
3 vols. (1919; reprint, New Delhi, 1985).
even distorts the evidence available. We gain little by faulting
Rao, T. A. Gopinatha, The Elements of Hindu Iconography, vol. 2, pt. 1 (Madras:
Coomaraswamy for the state of scholarship he had to work Motilal Banarsidass, 1914).
with, and there is after all much that is profound and useful in Sastri, K. A. Nilakantha, The Colas, 2d ed. (Madras: University of Madras, 1955).
Shulman, David Dean, Tamil Temple Myths: Sacrifice and Divine Marriage in the
the interpretation he offers. But we need not let the limits of
South Indian Saiva Tradition (Princeton: Princeton University Press, 1980).
Coomaraswamy's interpretation continue to restrict our read- 1990, Songs of the Harsh Devotee: The Tevdram of Cuntaramurttindyandr
ing of the Nataraja form. By granting that those images (Philadelphia: Department of South Asia Regional Studies, University of
Pennsylvania).
signified more fluidly, we can restore some of the complexity Sivaramamurti, C., Nataraja in Art, Thought and Literature (New Delhi: National
that would have characterized their significance for early Museum, 1974).
viewers. As the preceding -speculations demonstrate, the Smith, David, The Dance of Siva: Religion, Art and Poetry in South India
(Cambridge: Cambridge University Press, 1996).
image type lends itself to a variety of other readings, and these SII: South Indian Inscriptions (Madras and Delhi: Archaeological Survey of
readings are linked more closely than Coomaraswamy's to India, 1890-).
the historical environment in which these sculptures first Srinivasan, P. R., "Siva Nataraja, the Cosmic Dancer," Roopa Lekha, no. 26
(1955): 4-12.
functioned.
Stein, Burton, 1980, Peasant State and Society in Medieval South India (New York:
Oxford University Press, 1980).
,1984, "All the King's Mana," in All the King's Mana: Papers on Medieval
South Indian History (Madras: New Era).
Padma Kaimal received her'Ph.D. in the history of Indian art from
Swamy, B.G.L., Chidambaram and Nataraja: Problems and Rationalization (My-
sore: Geetha Book House, 1979).
the University of California, Berkeley. Her articles in Artibus Asiae,
Ars Orientalis, and Archives of Asian Art apply contemporary , Tevdram: Hymnes Sivaites du Pays Tamoul, trans. and ed. T. V. Gopal Iyer
and FranCois Gros, vol. 1, Ndnacampantar, vol. 2, Appar et Cuntarar,
theory and revisionist history to the study of ancient Indian art. She
Publications de l'Institut FranCais d'Indologie, nos. 68,1 and 68,2
(Pondicherry: Institut FranCais d'Indologie, 1985).
teaches Asian art at Colgate University [Department of Art and Art
Venkataraman, B., Temple Art under the Chola Queens (Faridabad: Thomson
History, Colgate University, Hamilton, New York 13346]. Press [India], 1976).
Younger, Paul, The Home of Dancing Sivan (New York: Oxford University Press,
1995).
Zvelebil, Kamil V., Ananda-tdndava of Siva-Saddnrttamarti: The Development of the
Frequently Cited Sources Concept of Atavalldn-Kattaperumdnatikal in the South Indian Textual and
Iconographic Tradition (Tiruvanmiyur, Madras: Institute of Asian Studies,
1985).
Adiceam, Marguerite, "Les images de Siva dans l'Inde du Sud, pts. 3 and 4,
"Bhiksatanamfirti et Kafikalamt rti," Arts Asiatiques 12 (1965): 83-112.
ARSIE: Annual Report on South Indian Epigraphy (Madras), 1887-.
Appadurai, Arjun, Worship and Conflict under Colonial Rule: A South Indian Case
(NewYork: Cambridge University Press, 1981). Notes
Balasubrahmanyam, S. R., 1944, "The Oldest Chidambaram Inscriptions:
Chapters II and III,"HJournal ofAnnamalai University 13 (1944): 53-91.
I am very grateful to the Getty Grant Program and the American Institute of
,1979,Douglas,
Later Chola Indian Studies for generously supporting my research for this article in the
Barrett, EarlyTemples (Faridabad:and
Cola Architecture Thomson Press
Sculpture: [India]). (London:
866-1014 United States and India; to the Berkeley Center for South Asian Studies for
Faber and Faber, 1974). providing research facilities in California; and to the Colgate University
, 1976, "The Dancing Siva in Early South Indian Art," The Sixth Annual Research Council for underwriting the expenses of illustration. My deepest
Mortimer Wheeler Archaeological Lecture, Proceedings of the British Academy 62:
thanks go also to Rick Asher, Janice Leoshko, Richard Davis, April Masten,
181-203. Andrew Rotter, James Heitzman, Mary Ann Milford, and Lory Frankel for
, 1981, "The 'Chidambaram'. Nataraja," in Chhavi 2, Rai Krishnaddsa their careful reading of drafts of this essay and for their valuable suggestions
Felicitation Volume, ed. Anand Krishna (Varanasi: Bharat Kala Bhavan), for revision, and to Alan Swensen for translating and interpreting Kulke's
15-20. German.
This content downloaded from 103.25.231.2 on Sat, 29 Feb 2020 14:08:22 UTC
All use subject to https://about.jstor.org/terms
SHIVA NATARAJA: SHIFTING MEANINGS OF AN ICON 415
1. To make this essay legible to readers who do not know Sanskrit and Tamil, Connoisseurship, Essays in Honour of Douglas Barrett, ed. John Guy (Middletown,
I have chosen to omit diacritical marks on words from these languages and to N.J.: Grantha, 1995), 110-13. The inscription is in SII 13, no. 302 (1952),
transcribe those words into English spellings that capture the proper pronun- 160-61. Without the surviving sculpture itself, however, we cannot verify that
ciations as closely as possible. "Kuttaperumal" bore the same iconographic connotations in the 9th century.
2. My understanding of Coomaraswamy and of his contribution to South Note that the figure on the 7th-century monolithic shrine, the Dharmaraja
Asian art history owes much to papers and comments given by Mary F. Linda, Ratha, in Mamallapuram, which Sivaramamurti, 192-93, classifies as a Nata-
Daniel Enbohm, Janice Leoshko, Sonia Y. M. Rhie, Jeremy Caslin, and Allan raja, is instead a representation of Shiva Kalarimurti, who lifts his leg to kick
Antliffe at the session entitled "Coomaraswamy Reconsidered" in the 83rd Kala; Rao, 156-64; Sivaramamurti also illustrates a 7th-century fresco of Shiva
Annual Conference of the College Art Association, San Antonio, 1995. dancing from Panamalai (fig. 49), but the remaining fragments indicate that
3. Coomaraswamy, 83-95. The first publication of this essay in the United the pose there differs from Nataraja's; one arm is raised straight up in the air.
States appears to have been in 1918 (by Sunwise Turn, New York), but Rao, Barrett (1974, 79-80) described a second "four-armed Nataraja" at Tiruvadu-
231, notes that it had been previously published in India in Siddhdnta-Dipikd 12 turai above the Durga niche on the north wall of the ardhamandapa (the small
(1912). Among the many who have accepted his reading are J. N. Banerjea, hall before the inner sanctum). There, the postures of the arms and legs
The Development of Hindu Iconography (Calcutta: University of Calcutta Press, denote the dance posture called chatura-tandava rather than ananda-tandava.
1941), 472-73; P. R. Srinivasan; Douglas Barrett, Early Cola Bronzes (Bombay: Kulke, 1970, 119, proposed Tiruvalishvaram as the site of the earliest surviving
Bhulabai Memorial Institute, 1965); Sivaramamurti; Barrett, 1976; R. Naga- Nataraja figure. For illustrations, see Sivaramamurti, figs. 83, 84. The Tiruval-
swamy, "Nataraja and the Urdhva Saivas," in Kusumafijali: New Interpretations of ishvaram is in ananda-tandava, but the temple seems not to have been
Indian Art and Culture, vol. 1, ed. M. S. Nagaraja Rao (Delhi: Agam Kala constructed until the end of the 10th century; see Sastri, 706; Sivaramamurti,
Prakashan, 1978), 241-48; Swamy; Peterson, 100; Dehejia; B. Natarajan, Tillai 1974, 232; Meister, 122-23, pl. 116; and K. D. Swaminathan, Early South Indian
and Nataraja (Madras: Mudgala Trust, 1994); and Younger. Temple Architecture: Study of TiruvdlWivaram Inscriptions (Trivandrum: CBH,
4. Richard H. Davis, Lives of Indian Images (Princeton: Princeton University 1990). Their assessments are based on the absence of inscriptions older than
Press, 1997), 11. 996 on the temple and the richness of the iconographic program. Note that
5. Brendan Cassidy, introduction to Iconography at the Crossroads (Princeton: this site is also far to the south of the Kaveri delta, and thus not at the hub of
Princeton University Press, 1993), 6-10. Chola influence during the 10th century.
6. On the importance of that alliance, see Balasubrahmanyam, 1944; Harle, David T. Sanford proposes examples at Tirumeyjnanam and Karuntattan-
41; Kulke, 1978, 199; Sivaramamurti, 220-59; Sastri; and Younger, 137-42. gudi as early Natarajas; Sanford, "Early Temples Bearing lRmayana Relief
7. Shulman, 1980; Kulke, 1970. Cycles in the Chola Area: A Comparative Study," Ph.D. diss., University of
8. Barrett, 1965 (as in n. 3); Barrett, 1976; Barrett, 1981, 15-20. California, Los Angeles, 1975, 239-42. My own examination of these temples,
9. Coomaraswamy, 86. however, leads me to place them in the third quarter of the 10th century.
10. Coomaraswamy, 87-93. Smith, 17, provides a useful discussion of this 20. The Tiruchchenampundi figure stands over a niche containing Shiva
evidence. See also Unmai Vilakkam: The Exposition of Truth, Tamil text, trans. Vinadhara-Dakshinamurti; the Punjai figure stands over a Durga niche; and
with notes and index by C. N. Singaravelu (Tellipazhi: Singaravelu, 1981). the Tiruvaduturai and Tiruverumbur Natarajas stand over niches for seated
11. See Smith, 17; W. Graefe, "Legends as Mile-Stones in the History of forms of Shiva Dakshinamurti. Nagaswamy (as in n. 3), 241-48, notes that the
Tamil Literature," in Professor P K. Gode: Commemoration Volume, vol. 2 (Poona: location of Nataraja over Dakshinamurti conforms to Urdhva Shaiva philoso-
Oriental Book Agency, 1960), 145; Zvelebil, 46-47. For the text itself, see phy and deduces that Urdhva Shaiva philosophy contributed to the genesis of
Tirumantiram by Tirumular, trans. and notes B. Natarajan, gen. ed. N. the Nataraja form. In that philosophical system, Nataraja is seen as supreme
Mahalingam (Madras: Sri Ramakrishna Math, 1991). over other forms of Shiva, such as Sadashiva, Ishana, and Dakshinamurti. I
12. That is, the Tirumantiram incorporates Shaiva Siddhanta notions of would suggest that this philosophical interpretation may equally well have
god's fivefold acts, a principle central to Umapati Shivacarya's Kunchitangh- arisen to explain preexisting sculptural realities.
ristava of ca. 1300, which describes Nataraja's dancing form in precise detail; 21.John Irwin, as cited by Lippe, 272. In my view, the technical logic of this
Smith, 17. argument carries more weight than do the accidents of preservation on which
13. Rao, 231. some scholars have deduced that the first stone representations of this deity
14. Gupta examples of early forms of Shiva's dance appear at Nachna and preceded or coincided with bronze representations. For example, Barrett,
Bhumara. Later examples are found in East Bengal and Ellora. Further south, 1976, 201, relies on surviving examples when he hypothesizes that bronze and
early examples appear in Aihole, Badami, Mogalrajapuram, Pattadakkal, stone figures appeared simultaneously. Kulke, 1970, 119, argues that the
Hemavati, Siyamangalam, and the Kailasanath temple in Kanchipuram. On earliest stone form to which he can assign a date (Tiruvalishvaram) was the
these early examples, see Sivaramamurti, 168-222; Srinivasan, 4-12; and prototype for later Natarajas. And Dehejia's evidence (40) for deducing that
Zvelebil, 13-22. bronzes of Nataraja began to be cast under Sembiyan Mahadevi seems to be
15. On earlier dance forms as iconographic predecessors, see Zvelebil, that this is the period from which the earliest datable objects survive.
12-13. 22. See, for example, the objections of Dehejia, 131 n. 32, who points out
16. Smith, 9, identifies the earliest such text as the Suta Samhita, soon beforethat the inscription does not identify the female figure as a consort of Nataraja
the Chidambaramahatmya, in which the ananda-tandava dance is also present. specifically. Barrett, 1981, has worked out a chronology of three types of
Suta Samhitd 111.8 describes a dancing form of Shiva that is quite close to theNataraja images in which only the last type, which he believes to have been
ananda-tandava form; Suta Samhitd, ed. S. Ramachandra Sastri and K. Kup-initiated by Sembiyan Mahadevi, depicts all the locks of hair long and flying
puswamy Sastri (Madras: K. R. Sastry, 1916), 313-22. But Kulke, 1970, 132,outward. Dehejia, 46, attributes such formally arranged locks to iconographic
revisions of the 11th century. She also hypothesizes that the aureole's shape
notes that in it the left leg position is not yet established, and he considers this
poem to reflect an earlier icon at Chidambaram very close in form though notevolved from ovoid to circular at the beginning of the 11th century; cf.
yet identical to the ananda-tandava. Dehejia, 43.
17. Zvelebil (46-54) draws examples from Manikkavasagar (Tiruvdcakam IX, 23. R. Nagaswamy, "A Nataraja Bronze and an Inscribed Uma from
X, XXXV, Mdnikkavdcaka Cuvdmikal. Aruliya Tiruvdcakam, ed. Mu. Karaiviram Village," Lalit Kald, no. 19 (1979): 17-19.
Kanapatippillai [Chennai: Kantalakam, 1992]) and Tirumular (Tirumantiram 24. For illustrations of figural sculpture at these temples, see Barrett, 1976,
IX, 2674-758), sections of which he retranslates as a corrective to Kulke andpls. 10-12; and M. A. Dhaky, "Cola Sculpture," in Chhavi: Golden Jubilee Volume
Narayana Ayyar. For the Tamil text and another English translation, see (Varanasi: Bharat Kala Bhavan, 1971), figs. 422, 426, 433.
Natarajan (as in n. 11), 413-25. Younger, 202-14, presents his own English 25. Barrett, 1981, 17, assesses this date for the Nataraja currently under
paraphrases of the Tamil poets Appar, Sambandar, and Sundarar. Peterson'sworship at Chidambaram. Nagaswamy has since located three Nataraja
selections from these poets (118-22, 243-44) demonstrate the same limitedbronzes at Chidambaram, but he finds them to be no earlier than the late 10th
conformity to the Nataraja icon that Younger's selections show: Appar IV.2.6, century; R. Nagaswamy, "Chidambaram Bronzes," Lalit Kald, no. 19 (1979):
IV.80.1, IV.81.4, IV.92.9, IV.121.1; Sambandar 1.39.1,1.46,1.134.5; and Sundarar 14.
VII.90. See Tevdram, vol. 1, 41, 48-49, 142; vol. 2, 3-4, 79-81, 90-91, 492-93. 26. For further discussion of Hindu understandings of the living divine, see
Balasubrahmanyam argues that this Nataraja form was present in literatureDavis (as in n. 4), 15-50.
by the 7th century; S. R. Balasubrahmanyam, Early Chola Temples (New Delhi: 27. Sundarar, Teudram VII.90 (Shulman, 1990, 567-73; Tevdram, vol. 2,
Orient Longman, 1971), 287-99. Note, however, that he does not distinguish492-93); Appar, TievdramIV.15, IV.23, IV.80, IV.81, VI.216 (Peterson, 120, 130,
between Nataraja and other forms of Shiva's dance. 151-52, 214, 223, 232-22; Teudram, vol. 2, 16-17, 25, 79-81); Sambandar,
18. Tirumantiram 2799; for the Tamil text and English translation, see Tevdram 1.80, II.175 (Peterson, 154, 159-60, 190-91; T7evudram, vol. 1, 87-88);
Natarajan (as in n. 11), 430. In other words, even if Zvelebil (46-52) is correct Manikkavasagar, Tiruvacakam II.127-30, XII.14 (Tiruvdcakam [as in n. 17], 6,
in the very early date-7th or 8th century--he proposes for this text, I do not49). For discussion and partial English translations, see also Zvelebil, 53-54;
see the Tirumantiram demonstrating the presence of the complete Nataraja and Younger, 197-214.
image type that early. Chidambaram's sanctity may be yet more ancient than the 7th century.
19. Barrett, 1976, noted the last three. The Tiruchchenampundi example is Kulke, 1970, 221, alludes to an earlier folklore. Younger, 82-83, finds this place
mentioned by Lippe, 236. On the construction dates of these temples, see mentioned in stories of the Kalabhra kings who ruled the Tamil region ca.
Barrett, 1974. R. Nagaswamy also identifies an inscription of the year 893 that4th-5th centuries.
commemorates a donor's gift of two lamps to "Kuttaperumal"; he interprets 28. Tirumangai praises the 8th-century king Nandivarman II Pallavamalla
this as evidence that Nataraja images were already in worship in the late 9th for donating to a Vaishnava shrine at Chidambaram; Tirumangai Alwar,
century; Nagaswamy, "On Dating South Indian Bronzes," in Indian Art and Periyatirumoli 3.2.2b, 3.2.8b; for the Tamil text, see Balasubrahmanyam, 1944,
This content downloaded from 103.25.231.2 on Sat, 29 Feb 2020 14:08:22 UTC
All use subject to https://about.jstor.org/terms
416 ART BULLETIN SEPTEMBER 1999 VOLUME LXXXI NUMBER 3
56. Kulashekhara also mentions a Vishnu shrine at Chidambaram in Tirumdli; 43. Smith, 86, alludes to the shared orientation. Smith, 87, notes as well that
for the Tamil texts of this and Tirumangai in transliteration, see Younger, 118 these murals also demonstrate that the chid-sabha was originally designed to
n. 16. display the dancing god as an open stage displays performers, with the gilt
29. Though these legends survive only in the Chidambaramahatmya, Younger, roofs acting as a gleaming beacon to attract the eye.
84, 90, suspects they are older. 44. Compare Gerd Mevissen's discovery of an analogous use of visual arts to
30. I have derived the following description of the Chidambaram temple embody political geography. He finds that Cholas and their contemporaries
complex and its history from Smith; Younger, 81-124; Nagaswamy, 179-85; oriented representations of Shiva Tripurantaka to face the direction of their
David Smith, "Chidambaram," in Temple Towns of Tamil Nadu, ed. George enemies' kingdoms; Gerd J. R. Mevissen, "Political Geography as a Determi-
Michell (Bombay: Marg, 1993), 58-75; Harle; Natarajan (as in n. 3); Meister, nant in South Indian Temple Art? A Case Study of Tripurantakamfirti," South
298-99; Balasubrahmanyam, 1979, 23-213 passim; T. Satyamurti, The Nataraja Asian Archaeology 1993 (Helsinki: Suomalainen Tiedeakatemia, 1994), 483-95.
Temple, History, Art and Architecture (New Delhi: Classical Publications, 1978); 45. According to the Chidambaramahatmya; see Kulke, 1970, 166-70.
Swamy; J. M. Somasundaram Pillai, Siva-Nataraja--The Cosmic Dance in Chid- 46. Younger, 13-15, paraphrases a number of verses from Appar, Samban-
Ambaram (Annamalainagar: Annamalai University Press, 1970); Balasubrahman- dar, Sundarar, and Manikkavasagar that attest to the early presence of the
yam, 1944; and my own observations of the temple in 1985, although neither Three Thousand, or Antanar, at Tillai.
Harle nor I had access to the interior of the chid-sabha. 47. On this community and the issue of Vedic study, see Smith, 1996, 58-79.
31. This shrine's dedication to Brahma or Chandesha is discussed by HarleSee also Younger's research on this strictly endogamous group of some one
(36-37); and indicated on the site plans of B.V.N. Naidu et al., Tdndava thousand people and on the priests themselves, who call themselves di-ksitar
and who own the temple as its collective trustees (11-45).
Laksaniam; or, The Fundamentals ofAncient Hindu Dancing; Being a Translation into
English of the Fourth Chapter of the Natya Sdstra of Bhdrata, 3d ed. (New Delhi: 48. See Skanda Purdna 1.1.22.52-53, trans. G. V. Tagare, in Ancient Indian
Munshiram Manoharlal, 1971), fig. 1; and Younger, 95. Tradition and Mythology, vol. 49, ed. G. P. Bhatt (Delhi: Motilal Banarsidass,
32. On this as the current understanding of the mukhalinga here, see Smith, 1992), 201; and Shiva Purdna 2.2.26.14-15 and 2.3.27.27-31, in Ancient Indian
89. Tradition and Mythology, ed. J. L. Shastri (Delhi: Motilal Banarsidass, 1970), vol.
33. For a plan and elevation of the chid-sabha and kanaka-sabha, see Swamy,1, 396, vol. 2, 584. Rao (43, 50) quotes two similar passages but does not name
30-31. his primary source. For Tamil examples from before 900, see the poems of
34. These diagrams are derived from the sketches proposing the various Appar (IV.2.6, IV.92.9); Sambandar (1.46,1.134.5), and Sundarar (VII.72). For
the Tamil texts, see T-evdram, vol. 1, 48-49; vol. 2, 3, 90, 472-73. For English
phases of construction at Chidambaram by Swamy, 56-64; Younger, 83-110;
andJ. M. Somasundaram Pillai, as reproduced in Harle, 33. translations of Appar and Sambandar, see Peterson, 119-22. For an English
translation of Sundarar, see Shulman, 1990, 463-67. Smith, 148, 226-27,
35. Studies published so far are partial or focused on the temple's religious
comments that throughout the Tevdram Shiva is essentially a fierce goblin, and
or literary identity rather than its architectural history. The most recent and
the cremation ground is the only site for his dance. While the Sanskrit texts
useful text is Satyamurti (as in n. 30). Harle focuses on the gopura (gateways)
were part of an elite world of thought, the Tamil songs of the Tevdram were
and admits reluctance to sorting out a history of the entire complex. Natarajan
widely popular and accessible.
(as in n. 3), Swamy, and Pillai (as in n. 30) disagree on key points and are often
49. Snakes are a primary attribute of Bhairava in the Vishnudharmottarapu-
speculative. Plans contradict each other and published photographic documen-
rana, which mentions them in three verses: 59.2, 59.3, 59.6. For the English as
tation is sparse. The monument appears briefly in such surveys as Meister,
well as the Sanskrit text of this treatise on painting, see Pratimrdlaksana of the
298-99; and Balasubrahmanyam, 1979. Smith presents many useful points of
Visnudharmottara, trans. D. C. Bhattacharyya (New Delhi: Harman Publishing
information but does not address the temple's structural history. Younger
House, 1991), 89-90. Marguerite Adiceam also finds snakes identified as
proposes a sequence of construction and renovation but does not employ art
historical data or standards of evidence. Bhairava's sign in the Silparatna, I and III; Adiceam, "Les Images de Siva dans
l'Inde du Sud," pt. 2, "Bhairava," Arts Asiatiques 11, no. 2 (1965): 26. The text,
36. Speculation about the 11th-century project is based on an inscription
however, is of the 16th century and may have little direct bearing on
crediting Naralokaviran, a general during the reign of the Chola king
9th-century perceptions; The Silparatna by Sre Kumdra, ed. Mah-mahopndhyaya
Kulottunga I (1070-1118), with many construction projects. The entire
T. Ganapati Sastri, pt. 1, Trivandrum Sanskrit Series, no. 75 (Trivandrum:
inscription is published in SII 4, no. 225 (1923), 31-34; Swamy, 123-34
Government Press, 1922), 2.
(English and transliterated Sanskrit); and Balasubrahmanyam, 1979, 23-2650. For an illustration, see Carmel Berkson and Wendy Doniger et al.,
(English only). See esp. verses 8-11 and 30. The language of the inscription,
Elephanta, the Cave of Shiva (Princeton: Princeton University Press, 1983), pls.
however, is vague in identifying the structures involved. Younger (98-103)40, 44.
interprets it to say that Naralokaviran had built the third enclosure wall, two of 51. For an illustration of Shiva destroying Andhaka at Elephanta, see ibid.,
its massive gateways (gopura) and the innermost enclosure wall, the steps and
pl. 61. The Kurma Purdna associates the skull with the head of Brahma that
porches that transformed a pond into the formal tank, and that he expanded Bhairava cut off; Kurma Purdna, 11.31.23-108, trans. G. V. Tagare, 2 vols.
the Shivakamasundari shrine.
(Delhi: Motilal Banarsidass, 1983), 514-21. For further discussion of Shiva's
To another wave of building activity in the later 13th century Younger expiation of that sin as Kankalamurti, see Adiceam, 83-112. A bell, a skull,
(104-9) assigns the linga shrine, the second enclosure wall, and construction
cobras, and a long garland of bones and skulls are appropriate to this violent
on the seven-story gopura of the third enclosure wall. Harle (63-69) seesstory.
the See, for example, the "Bhairava Adored" from the collection of Thakur
west and east gopura as having been finished by 1250, the south gopura by 1272,
Sahib of Sohagpur, Shahdol, in Goswamy, no. 150.
and the north gopura perhaps begun at that time and finished during the 52. See the fragmentary image from Hingajgarh, now in the Central
Vijayanagara period, or fully rebuilt during the Vijayanagara period in Museum,
the Indore, no. 2 (195), in Goswamy, no. 161. A similar image remains
style of the earlier gopura. under worship at the Vaital Deul in Bhubaneshwar, Orissa; Vidya Dehejia, Early
37. Inscriptions credit the Govindaraja shrine to the Vijayanagar king Stone Temples of Orissa (New Delhi: Vikas Publishing House, 1979), 24.
Achyutadevaraya in 1539; see Rangacharya, vol. 1, 132-33, no. 64 (ARSIE, no. 53. Nagaswamy has identified as Nishumbhasudani the emaciated form of
272 of 1913), 149, no. 169 (ARSIE, no. 1 of 1915); and S. R. Balasubrahman- the goddess at Tanjavur who sits on the crumpled forms of two demons. On
yam, "Chidambaram Inscription of Acyutadeva Raya," Journal of Oriental her associations with violent death, see R. Nagaswamy, "A Note on Ni'umbha-
Research (Madras) 12 (1938): 168-78. Satyamurti (as in n. 30), 24, points sudani
out Installed by VijayTlaya Chola in Tanjore," Lalit Kal, no. 18 (1977),
that some shrine was already in place before Kulottunga II removed it in the 39-40.
mid-12th century. 54. Zvelebil, 17-18, notes that Shiva's posture is actually a conflation of two
38. Nagaswamy, 179-85, also believes that the present chid-sabha structure,postures delineated in the ancient dance manual the Na.tya Sdstra: the arms of
though rebuilt in 1648, adheres in many respects to the shrine that had stoodbhujangatrasita (Karana 24) and the legs of bhujanganchita (Karana 40).
there since perhaps the 7th century. Understanding of the posture's name in the 10th century may not, however,
39. Smith observes that of the temple's four large gopura, only the south- have been as multifaceted, as the term "frightened by a snake" more readily
facing one bears saffron flags and, in addition, is aligned with a shrine within,
describes what the legs might be doing. For the textual passages, accompanied
and that is the chid-sabha: Smith (as in n. 30), 65. Further indications of the
by illustrations from the Chidambaram temple, see Naidu (as in n. 31), 27, 31.
shrine's original southward orientation lie in the processional directions used
Rao , 227-28, also finds these postures in the Amshumedbhedagama, which
in local festivals; Younger, 97. includes a third variant, bhujangalalita.
Glenn Yocum notes one other example of a south-facing temple, that at 55. For an illustration, see Sivaramamurti, figs. 37, 38. Zvelebil, 17-19,
Avadayarkoil, which is associated with yogis, teachers, and, thus, Dakshi- argues the importance of this relief as an iconographic precursor to Nataraja.
namurti, another south-facing deity; Yocum, "Brahmin, King, Sanny~si and 56. Many earlier figures of Shiva's dance hold the drum. See, for example,
the Goddess in a Cage: Reflections on the Conceptual Order of Hinduism atEllora,
a Badami, Kanchi, and the Nallur bronze, in Rao, pls. LXIII, LXVI,
Tamil Saiva Temple," Contributions to Indian Sociology, n.s., 20 (1986): 24. LXFII, LXX. Rao, 177-78, finds the drum also listed as one of Bhairava's
attributes in the Rupamandana and the Vatuka-Bhairavakalpa.
40. Smith, 1996, 66, 87, suggests also that the chid-sabha and kanaka-sabha
are, unlike the more typical sealed inner shrine, designed to display imagery 57.
as Younger, 90, understands that legend to describe the present site of the
a stage would display a dancer. nritta-sabha, the area south of the chid-sabha, which was a cremation ground
41. Nagaswamy, 181-82, draws the parallel to Kerala. and the home of the local goddess.
42. According to Smith, 85; but Satyamurti (as in n. 30), 15, maintains the 58. Coomaraswamy, 89-90. Coomaraswamy did not pursue this point since
columns of the chid-sabha are of polished stone. For more on the materialshis ofconcern lay with similes generated later by this history, similes that picture
this shrine, see Smith, 85-89; and Harle, 37-38. the burning ground as the hearts of Shiva's devotees from which he burns
This content downloaded from 103.25.231.2 on Sat, 29 Feb 2020 14:08:22 UTC
All use subject to https://about.jstor.org/terms
SHIVA NATARAJA: SHIFTING MEANINGS OF AN ICON 417
away illusions and deeds. Note also that Coomaraswamy (84-85) identifies the the Tamil texts of these verses, see S. R. Balasubrahmanyam, "The Oldest
tandava dance as Bhairava's dance in cemeteries and as the dance performed Chidambaram Inscriptions: Chapter I," Journal of Annamalai University 12
by the Shiva figures at Ellora, Elephanta, and Bhubaneshwar, rather than the (1943): 114-15. Sekkilar's Periya Purdnam is a work of the 12th century.
ananda-tandava dance that is the focus of this essay and that he terms the The Periya Purdnaam reports Chidambaram priests refusing to invest the royal
"Nadanta" dance. Smith, 100, points out that Coomaraswamy does not define diadem on a local chieftain, saying, "Only the ancient family of the Cholas was
the connotations of the Nadanta dance, but that Dorai Rangaswamy identifies
the Nadanta as the dance in the crematorium. entitled
the Pallavatoking
thisNandivarman
high privilege"; Periya Purdn.am
II Pallavamalla gave gold,3933.
pearls,Tirumangai
and preciousreports that
59. Coomaraswamy, 84-85. stones to Govindaraja, whose worship the dikshitars conduct, i.e., the form of
60. For illustrations of Coomaraswamy's examples as well as others like Vishnu housed at Chidambaram. For the Tamil texts, see Balasubrahmanyam,
them, see Sivaramamurti, 172-79, and 292-96. 1944, 56. The temple was also reported to defend the Pallava kings' right to tax
61. See, for example, the skeletal figures drinking blood in a sculptural revenues. The poet Sundarar (ca. 780-820) declares, "The god of Puliyur
fragment from a Chamunda image, now in the State Museum, Bhopal (no. Citrampalam will cause trouble to kings who refuse to pay tribute to the
817), in Goswamy, no. 164. Bhairava himself may be emaciated. See Rao, 178, Pallavas who guard the world" Tevdram VII.90.4. For the Tamil, see Tevdram,
pl. XLII.1, 3; and Stella Kramrisch, Manifestations of Shiva (Philadelphia: vol. 2, 493; translated by Shulman, 1990, 569.
Philadelphia Museum of Art, 1981), 34. 77. Barrett, 1981, 15, perceives the limbs to have just stopped moving, with
62. On Bhringi, see Smith, 22-24. the hair not yet come to rest. Zvelebil, 34, reads the flying locks as dance in full
63. See Sivaramamurti, 172-75, 189, 293-94.
swing. The play of light across bronze can enhance the impression of figural
64. Karaikal Ammaiyar prays for "the form of a demoness (pey) who could motion. Kapila Vatsyayan argues forcefully that all sculptural renderings
stand by god ever in prayer"; Periya Puranam, 1765-66. For a recent English distort the postures of Indian dance since sculpture is necessarily static and the
translation, see St. Sekkizhar's Periya Purdnam (Part I), trans. T. N. Ramachan-
postures allude to flowing movements; Vatsyayan, "Siva-Natega: Cadence and
dran (Thanjavur: Tamil University, 1990), 358. Karaikal originates in the Form," in Discourses on Siva, ed. Michael Meister (Philadelphia: University of
Tiruvalangadu recension of the dance contest myth, a recension that probably
Pennsylvania Press, 1984), 193-200.
precedes the Chidambaram recension. Shulman, 1980, 213-328.
78. Zvelebil, 43-45. He does not, however, cite primary sources for these
65. Smith, 3, 186. He also cites D.H.H. Ingalls, who saw in Shiva's dance "the
propositions. Zvelebil dates the Sangam age from ca. 200 B.C.E.-C.E. 300
costume and practice of a tribal shaman"; Ingalls, An Anthology ofSanskrit Court
Shulman, 1980, 213 nn. 7, 8, also notes the tradition in Tamil literature of
Poetry, Harvard Oriental Series, vol. 44 (Cambridge, Mass.: Harvard University
Kotravai's and Murugan's battlefield dances.
Press, 1965), 69.
79. Cilappatikdram, bk. 6.4, lines 44-45, ed. Puliyur Kecikan (Madras: Pari
66. Smith, 222-27. Ren6 Grousset proposed that the dwarf is underfoot
nilayam, 1958), 80. For an English translation, see The Cilappatikdram ofil.anko
because "this dance is danced upon the bodies of the dead," and that the
Atikal: An Epic of South India, trans. R. Parthasarathy (New York: Columbia
dwarf's gesture then teaches that destruction leads to generation; Grousset,
University Press, 1993), 58, Book of Pukar, canto 6, lines 43-49.
quoted in Sastri, 730-31.
67. These are found at Melakkadambur in Tamil Nadu (ca. 1100), and in 80. Kulke and Rothermund, 142-44. For further summaries of this process
at Chidambaram, see Kulke, 1970; and Zvelebil, 66-72. On the idea of
Assam, Malhar, and the Pala kingdom (from the 10th-12th centuries); see
"Sanskritization," see M. N. Srinivas, "A Note on Sanskritization and Western-
Sivaramamurti, figs. 171-74, 179, 181-82, 236. On Kalugumalai and Vilinjam,
see ibid., figs. 60, 63. ization," Far Eastern Quarterly 15, no. 4 (1956). On the strategies implied by
such transformations and for other cults (the Minakshi cult of Madurai, the
68. This was true despite profound differences in their systems of kingship
Jagannath cult of Puri, and the Krishna cult of Mathura) similarly transformed
as well as despite the intervention of some five centuries of dynastic obscurity
(the 4th-9th centuries). On the history of the Cholas of the Sangam age, by see royal involvement into legends that reified kingship, see Kulke, 1978,
132-33; and Dennis Hudson, "Siva, Minaksi and Vishnu-Reflections on a
Sastri, 30-106. On the early Cholas' uses of that past, see Stein, 1984, 42; and
Popular Myth in Madurai," in South Indian Temples, ed. Burton Stein (New
Stein, 1980, 45-46. Stein, 1984, describes three distinct kingship systems
evolving over the first millennium c.E.:, the heroic, the moral, and the ritual.Delhi: Vikas Publishing House, 1978), 107-18.
69. Kulke and Rothermund, 122-23; Sastri, 110-284. On the two major81. Kulke, 1970, lays out the case for reading the Chidambaramahatmya this
trading guilds in the region, the Ayyavole and the Manigramam, see Meera way. For the Tamil text, see Cidambara Mdhdtmya, ed. Somasekhara Dikshitar
Abraham, Two Medieval Merchant Guilds of South India (New Delhi: Manohar, (Kadavasal: Sri Meenakshi Press, 1971). For a translation into English, see E. A.
1988). For more on the maritime trade, its almost global scope, and its Sivaraman, Chidambara Mahatmya (Bombay: Bharatiya Vidya Bhavan, 1993).
economic nature, see Janet Lippman Abu-Lughod, Before European Hegemony:82. According to Kulke, 1970, 221.
The World System A.D. 1250-1350 (New York: Oxford University Press, 1989). 83. Kulke, 1970, 46-50. Zvelebil, 72, points out that the Myth of the Pine
70. See George W. Spencer, The Politics of Expansion (Madras: New Era, Forest must have been in place in the Chidambaramahatmya before the year
1983). 1000, because it is told in the Skanda Purdna, which is redacted by that date,
71. For further discussion of incorporative kingship, see Stein, 1984, 28-52; with its earliest material deriving from the 8th century.
Stein, 1980, 366-488; Nicholas B. Dirks, The Hollow Crown: Ethnohistory of an84. For summaries of this story as it is rendered in the Chidambaramahatmya,
Indian Kingdom (NewYork: Cambridge University Press, 1987), 28-43; Appadu- see Kulke, 1970, 10-12; Younger, 170; and Zvelebil, 68. On other texts that
rai, chap. 1; Kulke and Rothermund, 134-38; Kulke, 1978. narrate this tale and for a summary of his sculptural representations in South
72. On the limits operating on the Cholas and their responding strategies, India, see Adiceam, 83-112; and Wendy Doniger O'Flaherty, Siva, the Erotic
see Stein, 1980, 254-365; George W. Spencer, "Religious Networks and Royal Ascetic (New York: Oxford University Press, 1973), 178-84.
Influence in Eleventh Century South India," Journal of the Economic and Social85. Coomaraswamy, 85.
History of the Orient 12 (1969): 42-56; idem, "Royal Initiative under Rajaraja I,"86. Zvelebil, 30-35, reads Nataraja's feminine earring as an allusion to the
Indian Economic and Social History Review 7, no. 4 (1970): 431-42; Kulke and
goddess whom Nataraja, as resolver of oppositions, embraces within his male
Rothermund, 135-36; Appadurai; andJ. C. Heesterman, "The Conundrum of
self. The tiger Shiva flays for a loincloth suggests the ancient totem that gave
the King's Authority," in Kingship and Authority in South Asia, ed.J. F. Richards
this place its early name, "Town of the Tiger." The pose Shiva strikes to win a
(Madison, Wis.: University of Wisconsin Press, 1978). These studies revise
Sastri's earlier models of a highly bureaucratic Chola government.
narrative, could also have strengthened the associations between Nataraja and
73. On the localism of authority in this region before 1000, see the
the linga's early phallic connotations. Referred to as urdhva-janu or "erect
introduction to Stein, 1980; Y Subbarayalu, Political Geography of the Chola
leg," this pose may imply the erect phallus (urdhva-linga) that is also Shiva's
Country (Madras: State Department of Archaeology, Government of Tamil
emblem. On the erotic nature of this dance, see Zvelebil, 31
Nadu, 1973), 30-40; James Heitzman, "State Formation in South India,
110.
850-1280," Indian Economic and Social History Review 24, no. 1 (1987): 35-61;
Nataraja's association with Bhikshatana, who is ithyphallic
idem, "Temple Urbanism in Medieval South India," Journal of Asian Studies 46,
could incorporate an early cult of the linga at Tillai. Young
no. 4 (1987): 791-826; Noboru Karashima, South Indian History and Society:
that linga worship was only introduced late to Chidam
Studies from Inscriptions (AD 850-1800) (Delhi: Oxford University Press, 1984),
construction of the still-standing linga shrine in the 13th
3-15; and Kenneth R. Hall, Trade and Statecraft in the Age of the Clas (New Delhi:
Abhinav, 1980). Zvelebil, however, see in local myth and in the use of th
74. I would compare early Chola kingship to the modes of kingship("original place") indications that the linga may have been
perceived by Dirks, Stein, and Kulke to have been practiced before the 8th before the Nataraja icon developed; Kulke and Rothermun
century; Nicholas Dirks, "Political Authority and Structural Change in Early 30-31, 67.
South Indian History," Indian Economic and Social History Review 13, no. 2Zvelebil, 34-36, would see yet more deities absorbed int
(1976): 143; Stein, 1984, 3-11; and Kulke and Rothermund, 130-34. Accord- though note that his evidence for the presence of Krish
ing to Kulke's model, kings in the process of expansion--extending cultiva- Kiratamurti rests on his unusual reading of the crest of
tion, defeating neighbors and treating them as tributaries--had not yet peacock feathers.
established a centralized administration. 87. On the role of loot and the display of loot in the reif
75. Kaimal. Indian kingship, see Davis (as in n. 4), 53-85.
76. The parents of Kochchenganan (Krccenkanan), a Chola king of the 88. On Kulke's analysis of this name change, see Kulke
Sangam era, finally conceive him after praying at Chidambaram. Chidam-143. He understands the "little hall" as the shrine of
understand it as the chid-sabha: Younger, 24; Smith, 80.
baram was also the site of Kochchenganan's birth. Periya Purn.am 4203-6. For
This content downloaded from 103.25.231.2 on Sat, 29 Feb 2020 14:08:22 UTC
All use subject to https://about.jstor.org/terms
418 ART BULLETIN SEPTEMBER 1999 VOLUME LXXXI NUMBER 3
89. On the Vyaghrapada section, see Kulke, 1970, 4-9, 31-44; Younger, 105. Nagaswamy, 177-78, points out thejoint significance of these aspects of
166-70. For the Chidambaramahatmya chapters (in English), see Sivaraman (as the temple. See ibid., fig. 187, for an illustration of the dancing Shiva from the
in n. 81), 22-28, 43-50, 64-172. west side of the vimana. The inscription about the dancers is in SII, vol. 2, pt. 2,
no. 66 (1892), 278-303. On the dancing figures sculpted on the second story,
90. Sastri, 20. It appears on the Karandai plates from the third quarter of the
see Venkataraman (as in n. 101), 131-47.
10th century and on the earliest Chola coin known to Sastri. See ibid., 159, pl.
XXXII. 106. Ninth Tirumurai, Tiruvicaippd. For the Tamil, see Balasubrahmanyam,
91. Kulke, 1970, perceives a competitive atmosphere among deities 1944,of59this
n. 4. For an English translation, see Younger, 215-16.
place. Zvelebil, 68-71, expands on Kulke's picture. Smith, 50-51, 107. suggests
This chronology of Nataraja's emergence in temple sculpture is from
sometimes the multiple deities of this place complement each other, but 1976.
Barrett, he For an illustration of Nataraja in Kailasanathasvamin temple of
also acknowledges continuing instances of competition among priestly 981 groups.
in Sembiyan Mahadevi village, see Kaimal, fig. 17.
92. Shulman, 1980, 40, 88, 213-23. 108. For a summary of her gifts and inscriptions, see Venkataraman, 52-58.
93. On Vijayalaya's dedication of the goddess Nishumbhasudani's image at synthesized this figure from data in Barrett, 1974; Barrett, 1976;
109. I have
Tanjavur and on her significance, see Nagaswamy (as in n. 53); and Kaimal,
Meister, 179-93; Balasubrahmanyam, Early Chola Art, Part One (Bombay: Asia
36-53. See Kaimal, fig. 4, for an illustration. On the features of Tillai Amman, House, 1966); idem (as in n. 17); Venkataraman, 11-58; and my
Publishing
see Shulman, 1980, 219. own visits to many of these sites between 1985 and 1990.
94. On this inscription and the significance of this structure, see Younger,
110. On renovations at Govindaputtur, Karuntattangudi, Nalur-Tirum-
89-90, esp. n. 12. Zvelebil, 60, understands the Mother cult to beeyjnanam,
the oldestand Tirupurambiyam, see Barrett, 1974, 96-105; and SII, vol. 6
element of the Tillai religious tradition. (1928), no. 21. On renovations at Tiruvelvikkudi, see Lippe, 237; and Gary
95. The Kali temple appears to have been built during the 13th century, review of Early Cola Architecture and Sculpture, by Douglas Barrett,
Schwindler,
while Kopperunjingan, a leader of the Kadava feudatory family, hadArtibus
authority
Asiae 39, no. 1 (1977): 98.
over this area. Twenty-two inscriptions on a prakara of the Chidambaram111. The oldest inscription at Tirurameshvaram is from 974; Meister, 181.
complex give the date in terms of his regnal years; Rangacharya, vol.Also
1, 127-29,
on the basis of inscriptions, Barrett, 1974, 100, attributes the Kuhur
nos. 30-51 (ARSIE, nos. 459-68 of 1902, nos. 390-401 of 1903). Onetemple
of these
to ca. 970; ARSIE of 1918, nos. 286, 287, 292 of 1917, 25-26.
records a sale of land for the construction of a temple to Pidariyar (Kali); see
112. Barrett, 1974, 113-14, assigns the Udaiyargudi temple to the beginning
ibid., 129, no. 51 (ARSIE, no. 401 of 1903). of his Third Phase in his classification system, that is, shortly after 969. Note
96. On the basis of her use of Nataraja figures on her manythat
temples,
no Nataraja sculpture survives at Udaiyargudi, but inscriptions there
Zvelebil, 24-25, sees Sembiyan Mahadevi as instrumental in the spread of the
indicate that an "Araiyan Geyavidangan" funded the Nataraja niche and the
Nataraja cult and in "pressing for its acceptance as the state-cultotherof the
two niches of the south ardhamandapa wall; ARSIE, no. 555 of 1920.
Cholas."
113. Barrett, 1974, 107-8.
97. On the productive interactions between the political and the sacred in
114. Note, however, that Lingodbhava's narrative links with Bhikshatana are
medieval India, see Kulke, 1970; Kulke, 1978; idem, "Legitimation and in the Chidambaram version of this myth than they are in other
weaker
Town-Planning in the Feudatory States of Central Orissa," AARP versions.(Art andZvelebil, 71, points out that the Chidambaramahatmya suppresses the
Archaeology Research Papers), no. 17 (1980): 30-40; Dirks (as in n. episode
71); andin which the angry sages curse Shiva's linga to fall to the ground,
Appadurai, chap. 2. where it grows infinitely (thus leading into Lingodbhava's manifestation).
98. On the importance of such interactions to the study of art,Instead,see, forthe Chidambaramahatmya integrates the linga theme by having Shiva
example, Susan Huntington, "Kings as Gods, Gods as Kings: Temporality and sages to worship the linga at Tillai. For a survey of this deity's
urge the
Eternity in the Art of India," Ars Orientalis 24 (1994): 30-39; and Michael D.
iconography, see T. Chandrakumar, "Iconography of Lingodbhavamurti in
Willis, "Religions and Royal Patronage in North India," in Gods, Guardians
SouthandIndia," East and West 41 (1991): 153-72.
Lovers: Temple Sculptures from North India AD 700-1200, ed. Vishaka N. Desai
115.andThe first case is the Umamaheshvara temple at Konerirajapuram,
Darielle Mason (New York: Asia Society Galleries, 1993), 48-65.
Sembiyan Mahadevi's earliest known commission. Bhikshatana figures are
99. On these dates, see N. Sethuraman, "Date of Birth, Date of Coronation
present at Srinivasanallur, Kumbakonam, and Tirukandiyur, but Barrett, 1974,
and the Last Day of Rja Rja C61a," in Rdjardja the Great: Seminar Proceedings
75, perceives the Tirukandiyur figure to be a later addition.
(Bombay: Ananthacharya Indological Research Institute, 1987), 17-32.
116. For further discussion on the iconography of Bhikshatana images in
100. Stein, 1980, 830-34, argues that the Rajarajeshvara was unprecedent-
South Indian sculpture, see Adiceam, 83-112.
edly political in nature, in that earlier temples in the area were built on sites
117. Compare, for example, the sculptural sequence of Ganesha to Durga
already sanctified by divine miracles or saintly song, whereas Tanjavur's
aiding the transformation of the viewer as Huntington hypothesizes at
historical associations were only with the Chola family.
Udayagiri; Susan Huntington, with contributions byJohn C. Huntington, Art
101. No inscriptions or other labels confirm this identification, and K. R.
of Ancient India (New York: Weatherhill, 1985), 191. Note also that the Tamil
Srinivasan (in Meister, 238) identifies them instead as an earlier Kerala king
text, the Vaikhanasa Agama, recommends the same sequence; see Gary
Cheraman and his queens. The painted figures, however, are given much
Michael Tartakov and Vidya Dehejia, "Sharing, Intrusion and Influence: The
emphasis through their large scale and eye-level placement, and most other
Mahisasuramardini Imagery of the Chalukyas and the Pallavas," Artibus Asiae
scholars interpret the figure as Rajaraja, including the first author to publish
45, no. 4 (1984): 330.
the murals, S. K. Govindaswamy, "Cola Painting," Journal of the Indian Society of
118. Smith (as in n. 30), 65-66.
Oriental Art 1 (1933): 73-80. See also Sastri, 741; S. R. Balasubrahmanyam,
119. My thanks to Dennis Hudson for helping me see this pattern.
Middle Chola Temples (Faridabad: Thomson Press [India], 1975), 33-34; and B.
Venkataraman, Rdjardjeivaram: The Pinnacle of Chola Art (Madras: Mudgala 120. An inscription of the year 945 on this temple credits Parantaka with
Trust, 1985), 119, 130. funding the completion of the building's upper layers; ARSIE, no. 143 of 1925.
102. That is, his ishtadevata; Kulke, 1978, 199. Nataraja's repeated Since the Nataraja is high on the vimana wall, it may have been created under
presence
in this Tanjavur temple and the mural that probably representsParantaka's Rajaraja patronage. Note, however, that the inscription describing the
worshiping him at Chidambaram lead me to disagree with Younger's view
extent of construction Parantaka funded uses a term (kht.appatai) whose exact
meaning is no longer known. The precise upper layer with which Parantaka's
(136-37) that Rajaraja turned his religious attentions away from Chidam-
baram. contribution began remains, therefore, a matter of speculation.
103. These inscriptions note gifts made to the image of his consort by 121. See Kaimal, 57-58, fig. 6, for an illustration. For a nearby comparison,
Rajaraja's chief queen, Cholamahadevi, and his sister in 1010: SII, vol. 2, pt. 1, see the umbrella over the portrait of King Uttama Chola at the Umamahesh-
no. 2 (1892), verse 1, 14; and SII, vol. 2, pt. 1, no. 6 (1892), verses 1-2, 68-69.vara temple at Konerirajapuram, illustrated in Balasubrahmanyam (as in n.
A number of scholars agree that Rajaraja himself donated this dancing 101), pl. 167.
figure and that it survives as the large bronze still housed at the Rajarajeshvara; 122. On inscriptional evidence and the date of the gift, see Kaimal, 58.
Barrett, 1976, 188-89; Dehejia, 46; Sivaramamurti, 1974, 229; and R. Naga-123. Chidambaram's political independence is suggested by its identifica-
swamy, "Iconography and Significance of the Brihadigvara Temple, Tafijaviir," tion as a taniyur, a self-governing region; see Rangacharya, vol. 1, 125, no. 4
(ARSIE, no. 118 of 1888). Here the name used for Chidambaram is
in Discourses on Siva, ed. M. Meister (Philadelphia: University of Pennsylvania
Perumbatra Puliyur. On this and the nature of taniyurs as more independent
Press, 1984), 177. Barrett, 1976, 188-89, would assign the consort now placed
beside this bronze to a later date. than the nadus composing the majority of the deltaic region, see Subbarayalu
104. Nagaswamy (as in n. 103); Nagaswamy (as in n. 53); Sastri, 741; and (as
V. in n. 73), 92-94.
Venkayya, introduction to SII, vol. 2, pt. 1, ed. and trans. E. Hultzsch et al.124. For the Tamil text, see Balasubrahmanyam, 1944, 55 n. 1. Note also that
(Madras, 1892; reprint, New Delhi: Navrang, 1983), 20. See, for example, SII, the act of gilding was repeated after the early 10th century, as, for example, by
vol. 2, pt. 1, no. 1, in which most verses give the weight of the gift in termsKulottunga
of I Chola's sister Kundavai; Venkataraman, 108-9.
"the stone called Adavallan." 125. On this myth, see Kulke, 1970, 156-58, adhyaya 22; Sivaraman (as in n.
81), 63 (chap. 21); and Younger, 173-76. Though this story enters the
Nagaswamy also finds inscriptions describing the temple's largest annual
festival, the Brahmotsava, in which this Adavallan bronze was taken outChidambaramahatmya
in in the 12th or 13th century, its roots may be much older.
processions over three consecutive days. R. Nagaswamy, "Saivism On the Tamil notion that identity inheres within the earth of a place, see
under
Rajaraja--A Study," in Rdjaridja the Great: Seminar Proceedings (Bombay:Samuel
Anan- K. Parker, "Contemporary Temple Construction in South India: The
thacharya Indological Research Institute, 1987), 58-60; and NagaswamySrirangam
(as in Rajagopuram," Res 21 (1992): 112; and E. Valentine Daniel, Fluid
n. 103), 177. In neither article does Nagaswamy name the particular Signs: Being a Person the Tamil Way (Berkeley: University of California Press,
inscriptions on this temple that are his source. 1984).
This content downloaded from 103.25.231.2 on Sat, 29 Feb 2020 14:08:22 UTC
All use subject to https://about.jstor.org/terms
SHIVA NATARAJA: SHIFTING MEANINGS OF AN ICON 419
126. On the singularity of Chidambaram's chid-sabha, see Smith, 87; and on the attention of other historians of Indian art as well. Rao, 571-78, has
its unusual rituals and management, see also Younger, 13-45, 85. proposed that an equilateral triangle underlies the composition of Shiva
127. Kulke, 1978, 130-31. Nataraja, with apices beneath the standing foot and at the palms of the
128. Nambi Andar Nambi and Sekkilar claim that Parantaka's father, Aditya outstretched hands. He deduces this from the kamikagama and karanagama
I Chola, roofed the Chidambaram hall in gold; for the Tamil verse, see texts, which give measurements but do not actually delineatea triangle. Note
Balasubrahmanyam, 1944, 59 n. 5. There are, however, two copperplate that many rules advocated in agama texts were descriptive aswell as prescrip-
inscriptions that assign the gilding of the chid-sabha to Parantaka alone. These tive, their "principles" derived from already long-established
practices.
are the Larger Leyden grant; Epigraphia Indica 20, no. 34, 256, verse 17; and 132. A wheel literally articulates the radial composition of two 8th-9th-
the Tiruvalangadu grant; SII, vol. 3, no. 205 (1920), 419, verse 53. The Larger century figures of Shiva's dance (in catura and urdhva-janu modes) from Bihar,
Leyden grant claims that Parantaka won this gold in his victories over Pandya published by Raju Kalidos, "New Dimensions of the Cosmic Dancer," East and
and Sri Lanka kings. In addition, Gandaraditya's Tiruvicaippa credits Paran- West 40 (1990): 335-39. Note that in these figures, the center of the wheel is
taka with gilding the chid-sabha. For the Tamil verse, see Balasubrahmanyam, aligned with the navel.
1944, 59 n. 4; for an English translation, see Younger, 216. 133. Vatsyayan (as in n. 77), 192, 200, also perceives the compositions of
129. In such matters, each medieval South Indian kingship may have Nataraja images to be focused on the navel, from which all energies emanate
undergone the evolutionary process separately instead of inheriting achieve- and on which they focus.
ments from preceding dynasties; Kaimal, 37. 134. I turn to geometric logic because the half flexion of the elbows
130. For his definition of "royalizing," see Kulke, 1978, 133. contradicts the principles of classical Tamil dance, which call for smoothly
131. See, for example, John F. Mosteller, "Texts and Craftsmen at Work," in extended or more sharply bent elbows when the arms are held out to the side.
Making Things in South Asia: The Role of Artist and Craftsman, ed. Michael W. 135. See, for example, the elongated aureoles on the bronzes from
Meister, Proceedings of the South Asia Seminar, 4, 1985-86 (Philadelphia: Kilakattur, Seynalur, and Tarangambadi, illustrated in C. Sivaramamurti, South
Department of South Asia Regional Studies, 1988), 24-33; and Kapila Indian Bronzes (Bombay: Lalit Kala Akademi, 1963), figs. 100, 103, 105.
Vatsyayan, The Square and the Circle of the Indian Arts (New Delhi: Roli Books, 136. See also the wider range of such figures illustrated in Sivaramamurti,
1983), 73-99. Given the importance of yantras as auspicious structures on 168-222; Srinivasan, 4-12; and Zvelebil, 13-22. The Bihari example cited by
which to organize images, deducing them from finished pieces has absorbed Kalidos (as in n.132) is a noteworthy exception.
This content downloaded from 103.25.231.2 on Sat, 29 Feb 2020 14:08:22 UTC
All use subject to https://about.jstor.org/terms