Shiva's Karanas in the temples of Tamil Nadu: the Natya Shastra in stone

by Liesbeth Pankaja Bennink, and Kandhan Raja Deekshithar, Jayakumar Raja Deekshithar, Sankar Raja Deekshithar

August 20, 2013

text © asianart.com and Liesbeth Pankaja Bennink, and Kandhan Raja Deekshithar, Jayakumar Raja Deekshithar, Sankar Raja Deekshithar;
images © Liesbeth Pankaja Bennink, and Kandhan Raja Deekshithar, Jayakumar Raja Deekshithar, Sankar Raja Deekshithar

(click on the small image for full screen image with captions.)


The origin of dance: Bharata's Natya Shastra

Karana means 'action' and in the context of dance it indicates a coordinated action of the body, the hands and the feet [1]. 108 such karana or units of dance are named and defined in the Natya Shastra, the most ancient text on the performing arts composed by Bharata Muni. This text is dated to a period of around 2000 years ago, within a margin of 500 years and has been the most influential in defining and shaping Indian performing arts [2].


Fig. 1At the request of Indra, Brahma extracted the essence of the four Vedas and created the fifth, the Natya Veda, or drama Veda. It is accessible to all. Its purpose is not only to entertain, but to mirror the world, and to offer counsel, courage and comfort. Bharata Muni is given the responsibility to produce the first play. After the first successful performance it was presented before Shiva, the Cosmic Dancer, on Mount Kailasha. Shiva is deeply moved and is reminded of his Sandhya Tandava, his dance at twilight. “Brahma and Bharata supplicate Shiva for a knowledge of dancing. Thereupon the Lord of the Worlds bids Tandu initiate His devotees into the secrets of the difficult art.” Tandu proceeds to teach Bharata the 108 karanas. These are combined into angaharas and into larger choreographies. After Tandu initiates Bharata the art of dance is given to humanity through Bharata’s sons and the apsarasas, the heavenly dancers.

Therefore according to the Natya Shastra all Indian dance has its origin from Shiva's tandava. Tandava is a vigorous, masculine and divine dance. Shiva is pre-eminently a god whose divinity expresses through dance. The Sandhya Tandava is his divine dance performed at twilight on Mount Kailasha, his sacred abode. Witnessed by his consort and accompanied by all devatas playing various instruments. Other tandava's performed by Shiva are Tripura, after the destruction of the Three Cities, and of course the Ananda Tandava, the Dance of Bliss performed in Chidambaram [3].

Karanas and temples


Fig. 2The art of dance in India has always been intertwined with sculpture, architecture, ritual and doctrine. For this there is no better illustration than the karanas (figure 1). Not only do we know these 108 dance movements from their description in the Natya Shastra, we also have sculptural illustrations of them in ritually significant locations in temples in South India. Five temples are well known for the depiction of the karanas in their sculptural program. They are the Rajarajeshvara temple in Tanjore, the Nataraja temple in Chidambaram, the Sarangapani temple in Kumbakonam, the Arunachaleshvara temple in Thiruvannamalai and the Vriddhagirishvara temple in Vriddhachalam.

The oldest of these five examples is the Rajarajeshvara temple in Tanjore (figure 2). Built by Rajaraja Chola around the year 1000 CE, this temple features an incomplete karana frieze found in a circumambulatory passage around the grabhagriha. A four-armed Shiva is shown dancing the first 81 karanas in a horizontally placed relief which is on one's right hand side as the passage is followed in the clock-wise direction. The series has been left incomplete. The place where it is found was never intended for public viewing.


Fig. 3In Chidambaram the passages through all four gopurams have been decorated with complete representation of all 108 karanas (figure 3). We are sure of this because in the east and west gopuram the panels are accompanied by the relevant verse from the Natya Shastra. Here a female dancer accompanied by two musicians performs the movements. The east, south and west gopuram have been dated to the 12th and 13th century, the north gopuram somewhat later [4]. The series are largely identical in all four gopuram's.

In the east gopuram of the Sarangapani temple in Kumbakonam a more or less complete series is depicted as danced by a male dancer (figure 4). The panels are positioned in a horizontal band around the outside of the gopuram. Here many of the panels are captioned with inscriptions written in grantha script. Again this is the main reason it is sure the reliefs are intended as representations of Bharata's karanas. In this temple Vishnu is the presiding deity and it has been suggested these dance reliefs may have originally belonged to a Shiva temple and for some reason moved here. Among the karanas we find a relief of Shiva dancing the Urdhva Tandava and also goddess Kali dancing, possibly referring to the myth of the dance-contest between them [5].

• 
  Fig. 4
• 
  Fig. 5
• 
  Fig. 6

Two more temples are known to have the karanas from the Natya Shastra represented. In Vriddhachalam (figure 5) and in Thiruvannamalai (figure 6). In both temples the karanas are found in the gopuram passages. In the Vriddhagirishvara temple depictions of the karanas are found in all four gopurams, but are incomplete. Only 101 karanas are represented and in a strikingly unusual ordering. The construction date of the gopuram is not exactly known. On the basis of architecture, sculptural style and costume a date in the 14th century is likely, which would place this gopuram after the construction of the Chidambaram gopuram and before the gopuram of Thiruvannamalai.

In Thiruvannamalai all the karanas are systematically arranged in the east gopuram passage. Besides the 108, possibly copied from Chidambaram, there are many more dance movements depicted, besides panels depicting deities, rishis and other mythological figures. Depictions are found on 20 pilasters with 9 panels arranged vertically, making 180 panels in all, of which 108 constitute the karanas as defined in the Natya Shastra. This Raja Gopuram of the Arunachaleshvara temple in Thiruvannamalai was constructed in the 16th century.

Karanas and the history of dance

The karanas have always fascinated me, as a dancer and as a historian. They were the reason to visit Chidambaram, and to connect to Raja Deekshithar. My guru Smt. Rajamani taught me about them and raised my curiosity and also academic interest. 108 Dance movements described in an ancient text and depicted in sculpture constitute a unique source in the history of dance. They seemed to be lost for the most part with only fragments included in some choreographies. My first idea was to try and reconstruct them.

Several attempts at reconstruction have been undertaken. Dr. Padma Subrahmanyam's was an early one. Raja Deekshithar and I did not pursue this direction of research because we came to the conclusion early on in our research that reconstruction was impractical. The text and the image, a frozen moment in a sequence of movement, allow for many different alternative interpretations. Every dancer will reconstruct them differently, based on training and insight. This was also the ultimate conclusion of Dr. Alessandra Lopez y Royo [6].

• 
  Fig. 7
• 
  Fig. 8
• 
  Fig. 9

The subject never lost its fascination though. Many questions remain unanswered. Why, from all the millions of movement combinations the human body is capable of, were these 108 were considered to be the constituents of Shiva's tandava, the divine dance? What is the relationship between the karanas and the units of movement called adavus, which constitute Bharata Natyam today? How did the various regional dance styles develop from this Marga, or main tradition? What is the meaning of their placement within the temple complex? Why were they included in these five temples? And what is the meaning of their inclusion in these gopurams where in other gopurams (figures 8 and 9) we find only miscellaneous dance figures or mythological scenes?

Thiruvadigai


Fig. 10When Raja Deekshithar and I visited the Veerateshvara temple in Thiruvadigai in 2005 we were on a field trip related to his research into the history of the sphinx in the Indian traditions [7]. The Deekshithar almost immediately discovered a relief of the sphinx-purushamriga on one of the pillars in the vestibule of the gopuram. But he also realised the passageway of this gopuram was decorated with 108 dance reliefs, reminiscent of the karanas depicted in his own temple, the Shiva Nataraja temple of Chidambaram.

The gopuram is a pleasant but unimposing gateway with quality artwork in its inner passage (figure 10). The architectural features and style indicate it was built in the Nayaka period in the 15th or 16th century. This is among other things indicated by the base, the design of the podikai or corbel and the style of the reliefs. This would make it a contemporary of the Rajagopuram of Thiruvannamalai. It has a granite base with a superstructure of six levels of tapering brick and lime work topped with a barrel roof and seven shikaras.


Fig. 11The walls of the passage way which offers access to the temple courtyard is decorated with 108 panels of dancing figures (figure 11). They are organised in four groups of three pilasters very similar to the organisation of the karanas in Chidambaram, Thiruvannamalai and Vriddhachalam. Each pilaster has 9 panels. Each group covers one of the four wall spaces in the passage. Besides the dance panels there are several other interesting reliefs on these walls, and there are two decorated pillars in the vestibules.

On the right side upon entering, one relief depicts Shiva as Tripurantaka, Destroyer of the Three Cities, the presiding deity of the temple (figure 12). Another relief below the first depicts Shiva as Gajantaka, Destroyer of the Elephant Demon (figure13). Whereas the Tripurantaka seems to be a sculpture from the Nayaka period, the Gajantaka could be much older, based on the pose of the body and the representation of the elephant [8]. Opposite these two reliefs we find a relief of Murugan on his peacock (figure 14). Crossing the threshold and the vestibules towards the temple courtyard we find on the northern wall Shiva with Parvati on the sacred bull Nandi (figure 15). On the Southern wall a king is standing with hands folded in worship (figure 16).

• 
  Fig. 12
• 
  Fig. 13
• 
  Fig. 14
• 
  Fig. 15
• 
  Fig. 16
• 
  Fig. 17

On one of the pillars in the vestibules which define the centre of the gopuram we find among the sculpture a Nayaka period sphinx or purushamriga (figure 17).

A dancing figure in a temple gateway is not automatically a depiction of a defined karana (figures 8 and 9). Of course every dance movement can be called a karana, a coordinated action. But the karanas of Bharata Muni's Natya Shastra are specifically defined and ordered. From the known examples it can be concluded there is a connection between the 108 karanas and gopurams. Even a connection between gopurams and dance in general [9]. Many gopurams include depictions of dance in their sculptural program. Only four of those are known to be the series of defined movements as described by Bharata Muni. The question to be answered here is whether the 108 dance reliefs in Thiruvadigai represent the karanas as defined and ordered in the Natya Shastra.

The number of the reliefs, the way they are organized, and a handful of recognizable poses led Raja Deekshithar to the hypothesis the 108 dance reliefs in the east gopuram of Thiruvadigai could be Bharata's 108 karanas.

The karanas which can be readily identified are [10]

Karana 50: Lalata Tilaka: Thiruvadigai 91 (figure 18)

Karana 53 Chakramandala: Thiruvadigai 90 (figure 19)

Karana 52 Kuncita: Thiruvadigai 20 or 93 (figure 20)

Karana 57 Argala: Thiruvadigai 33 (figure 21)

Karana 107: Shakatashya: Thiruvadigai 27 or 29 (figure 22)

Karana 108: Gangavatarana: Thiruvadigai 27 or 29 (figure 23)

• 
  Fig. 18
• 
  Fig. 19
• 
  Fig. 20
• 
  Fig. 21
• 
  Fig. 22
• 
  Fig. 23

Some more karanas can be identified tentatively:

Talapushpaputa (1) or Talasamsphotita (69): Thiruvadigai 66 (figure 24)

Alata(18) or Urdhva Janu (25) Nagasarpita (106) (figure 25a) could be : Thiruvadigai 104, 106, 92 (figure 25b)

Bhujangatrasita (24), Bhujangancita (40) en Talasanghattita (93) (figure 26a) could be either one of Thiruvadigai 35, 62, 99 (figure 26b).

Sannata (75) and Nitamba (85) could be Thiruvadigai 59 (figure 27)

Katibhranta (43) or Urudvritta (98) could be Thiruvadigai 18 or 88 (figure 28)

• 
  Fig. 24
• 
  Fig. 25a
• 
  Fig. 25b
• 
  Fig. 26a
• 
  Fig. 26b
• 
  Fig. 27
• 
  Fig. 28

Shiva's Karanas?

Recognising many of the poses also found in the other karana series, together with the number 108, and the location in the sculptural program gave us the anticipation this would be another karana series. It would be a wonderful discovery because it would give us one more window into the history of dance in the subcontinent. Further analysis raises question though.

From the sculptural style and the dress worn by the dancers these reliefs seem to be from the later period and, like the rest of the gopuram, probably belong to the earlier phase of the Nayaka period (1300-1600). It could be argued on the basis of the dance costumes these reliefs are somewhat later then the panels in the East gopuram at Thiruvannamalai. The dancers wear pyjamas covered with a pleated, skirt-like or apron-like, garment, but don't seem to wear blouses, as came into usage later. In earlier panels we see dancers wear draped lower garments worn in a fashion similar to a dhoti worn as 'katchai', draped between the legs in a fashion creating a kind of trousers. The costumes worn by the dancers in Thiruvannamalai closely resemble those worn in Chidambaram and Kumbakonam. Pleated fan-shaped skirts are still part of today's dance costume.

• 
  Fig. 29
• 
  Fig. 30
• 
  Fig. 31
• 
  Fig. 32
• 
  Fig. 33

The dancers occupying the lowest row of panels have been depicted with more ornamental costumes, including some large head-dresses and ornaments (figure 29). Some of these also hold some kind of implements or possibly weapons like knives or swords (figure 30). This is reminiscent of costumes worn in traditional dance-dramas and ritual dances even today [11]. In only 10 of the panels does a musician accompany the dancer. In contrast, in Chidambaram two musicians accompany the dancer in all the panels and in Thiruvannamalai at least one musician is seen in most of the panels. A few panels do not seem to depict a dancer at all. Two panels depict male drum players (figure 31), another male figure plays a kind of tambourine-like drum or kanjira [12] (figure 32). One female figure holds a string instrument (figure 33). Five female figures stand holding camaras or ritual fly whisks (figure 34). One unusual panel centrally placed panel depicts two male figures dancing while holding one another around the shoulders while at the same time holding a fly-whisk and other attributes (figure 35). Two figures are seated on a kind of stool.

• 
  Fig. 34
• 
  Fig. 35
• 
  Fig. 36
• 
  Fig. 37
• 
  Fig. 38

Some of the characteristics of the karana series as known from the other five temples are missing: movements where the dancer shows the back (karana 16, prsta svastika, figure 36) ; where the dancer is jumping (karana 99, madaskhalita, figure 37); where the dancer is depicted sitting on the floor (karana 55, aksipta, figure 38). There are few movements depicted showing svastika or crossing positions of the feet.

Shiva's Tandava and Tripurantaka

It would have been agreeable if we could state with definite confidence these 108 reliefs of dancers in the gopuram of Thiruvadigai represent the 108 karanas as enumerated in Bharata Muni's Natya Shastra. There are several reasons to support the hypothesis they were indeed intended as a defined series. The number 108, the identified karanas, the location. But these are not conclusive against the counter arguments given above. Therefore we cannot offer a final conclusion.

On the other hand there is one argument to support the idea these are indeed Bharata's karanas. This argument lies with the deity to whom this temple is dedicated, Shiva as Tripurantaka, the Destroyer of the Three Cities. The myth of Shiva's destruction of the Three Cities is an early one. The earliest reference is found in the Yajur Veda [13]. In Tamil texts known as Sangam literature Shiva is described as dancing after he destroyed the Three Cities. This dance is referred to as Kotukotti and also as Pantarankam. His dance is witnessed by his consort Uma who beats the rhythm [14].


Fig. 39The connection between Shiva dancing and Tripurantaka Murti has not received much attention. Bindu S. Shankar points to the importance of the Tripurantaka Siva in the iconography of the Rajarajeshvara temple in Tanjore. In her dissertation she makes a connection between the first appearance of the karanas in the sculptural program with the significance of Tripurantaka in the doctrine of this temple [15].

This connection is also significant with respect to the Nataraja temple in Chidambaram. In an oral tradition of this temple it is told that after the destruction of the Three Cities Shiva landed his chariot opposite the main sanctum. After getting down from the chariot he ascended into the Sabha and commenced his dance. In the place where the chariot landed is now the Edirambalam, the Opposite Hall, in the form of a stone chariot. Today this hall is known as the Nritta Sabha. It is situated exactly opposite the Sabha where Shiva is dancing the Ananda Tandava. From this Nataraja's dance in Chidambaram is also known as Shanti Kuttu, or Dance of Peace [16].

Conclusions

Fig. 40The Veerateshvara temple dedicated to Shiva as Tripurantaka in Thiruvadigai was build by the Pallava king Mahendravarman in the 6th century. It is a large pyramidal structure similar to the Kailasanatha temple in Kanchipuram [17]. The pyramid shape replicates the form of Kailasha, the cosmic mountain [18]. Also the Rajarajeshvara temple in Tanjore with its pyramidal vimana is a representation of Kailasha, the Cosmic Mountain as Shiva's abode [19]. The presence of the karanas in the circumambulatory passage around the vimana underlines this identification and is a clear statement of the importance of this relationship to the king Rajaraja and his architect. The presence of 32 murtis of Shiva as Tripurantaka in the niches of the second elevation of the vimana is another statement pointing to the relationship between Kailasha and Shiva's Tandava dance in the conquest of evil.


Fig. 41A relationship of Tripurantaka, Kailasha, Shiva as divine dancer and the presence of the karanas is tentatively appearing. This relationship is pointed out by Bindu Shankar in her dissertation [20]. The discovery of 108 dance reliefs in a gopuram of a temple dedicated to Shiva Tripurantaka corroborates this relationship. And at the same time this connection between Dancing Shiva, Tripurantaka, Kailasha and karanas makes it very likely the 108 dance panels in the gopuram in Thiruvadigai represent the 108 karanas as conveyed to humanity by Bharata Muni.

There can be no doubt these 108 dance panels could represent the karanas as described and defined by Bharata Muni in the Natya Shastra as practiced at the time of the Nayaka dynasty. They thus represent a valuable document of a living and evolving dance tradition which the sculptor shaped on the basis of dancers performing for him. It allows us a glimpse into the art of dance in the 15th and 16th century. And offers insight into its evolution through this unique sculptural resource.

July 2013
Liesbeth Pankaja Bennink, and Kandhan Raja Deekshithar, Jayakumar Raja Deekshithar, Sankar Raja Deekshithar

Bibliography

Acharya, Guru C.R., Mallika Sarabhai, Understanding Kuchippudi, IGNCA and Darpana Academy of Performing Arts, New Delhi, 1992

Harle, James C., Temple gateways in South India: the architecture and iconography of the Cidambaram gopurams, 1963.

Lopez y Royo, Alessandra, ReConstructing and RePresenting dance: Exploring the dance/archaeology conjunction, Online publication, 2006 http://humanitieslab.stanford.edu/117/693 (retrieved 18-6-2013)

Naidu, N. Naidu, B.V. Pantulu, P.R. Tandava Lakshanam :The Fundamentals of Ancient Hindu Dancing. New Delhi,: Munshiram Manoharlal, 1971, 1980

Shankar, Bindu S., Dance Imagery in South Indian Temples: study of the 108 karana sculptures. Dissertation, The Ohio State University, 2004

Subrahmanyam, Dr.Padma, Karanas, Common dance codes of India and Indonesia, Chennai, 2003

Vatsyayan, Kapila. Dance Sculpture in Sarangapani Temple. 1st ed, Saher ; 9. Madras: Society for Archaeological Historical and Epigraphical Research, 1982.

Taittiriya Sanhita of the Yajur Veda, Kanda VI, prasna 2, anuvaka 3. http://www.sacred-texts.com/hin/yv/yv06.htm

Zvelebil, Kamil V., Ananda-Tandava of Siva-Sadanrttamurti, Institute of Asian Studies, Chennai, 1998,

Footnotes:

1. A karana in dance is defined as "the coordination of the movements of the hands and feet" , Naidu, N. Naidu, B.V. Pantulu, P.R. Tandava Lakshanam :The Fundamentals of Ancient Hindu Dancing. New Delhi,: Munshiram Manoharlal, 1971, 1980, p.19. It is a basic unit of movement from which choreographies are constructed and 108 such units are specifically defined by Bharata in the Natya Shastra. (http://en.wikipedia.org/wiki/Karana_dance, as recovered on 19-03-2013).

2. The Natya Shastra of Bharata Muni is a Sanskrit text giving detailed and elaborate descriptions, directions and definitions of drama, dance and music. This text is dated to the period around 2000 years B.P. (http://en.wikipedia.org/wiki/Natya_Shastra, recovered on 19-03-2013)

3. Various lists enumerating Shiva's tandavas exist.

4. Harle, James C., Temple gateways in South India: the architecture and iconography of the Cidambaram gopurams, 1963.

5. Vatsyayan, Kapila. Dance Sculpture in Sarangapani Temple. 1st ed, Saher ; 9. Madras: Society for Archaeological Historical and Epigraphical Research, 1982.

6. Lopez y Royo, Alessandra, ReConstructing and RePresenting dance: Exploring the dance/archaeology conjunction, Online publication, 2006. http://humanitieslab.stanford.edu/117/849

7. Raja Deekshithar, Sphinxes in Indian Art and Tradition, Asianart.com July 31, 2009, http://www.asianart.com/articles/sphinxes/index.html

8. Starting from the Chola period Shiva Gajasamharamurti or Gajantaka, the destroyer of the elephant demon, is depicted as dancing facing towards the left with his left leg raised up very high, the foot touching the inner thigh of the right, standing, leg, which is placed on the elephant skull. With the upper body turned back in such a way the chest is showing (http://en.wikipedia.org/wiki/File:Gajasamharamurti.JPG retrieved 19-6-2013). During the Pallava period Gajantaka was depicted as dancing facing to the right, with the left leg lifted up and resting on the elephant skull (http://www.temples-dravidiens.net/pallavas/kailashanatha/enceinte/mursud/21.gajantaka.html, retrieved on 19-06-2013). Here in Thiruvadigai Shiva dances facing right, left leg placed on the elephant skull, right leg slightly lifted. And the ribs of the elephant skeleton are clearly depicted, which is unusual.

9. Shankar, Bindu S., Dance Imagery in South Indian Temples: study of the 108 karana sculptures. Dissertation, The Ohio State University, 2004.

10. For the purpose of analysis the dance panels have been numbered. For the sake of systematization we have followed the organisation of the karanas as found in Chidambaram and Thiruvannamalai. There they are arranged from inside to outside, from below to high, in the direction of a pradakshina circumambulation. With the first karana Talapuspaputa found in the lowest panel of the first pilaster as one leaves the temple. From there from low to high upto the last pilaster on the left hand side, and then turning back into the gopuram, towards the inner courtyard of the temple. Upto the last one, 108 gangavatarana the topmost panel of the final pilaster.

11. For instance Teyyam from Kerala (http://en.wikipedia.org/wiki/Theyyam retrieved 19-6-2013), Terukuttu from Tamil Nadu (http://en.wikipedia.org/wiki/Terukkuttu, retrieved 19-6-2013), or Yakshagana (http://en.wikipedia.org/wiki/Yakshagana, retrieved 19-6-103).

12. In wikipedia it is mentioned this instrument was in use since the late 19th century. This gopuram is definitely older by one or more centuries. Personally I would date it to the 16th or 17th century.

13. Yajur Veda, Kanda vi, prasna 2, anuvaka 3: The Asuras had three citadels; the lowest was of iron, then there was one of silver, then one of gold. The gods could not conquer them; they sought to conquer them by siege; therefore they say--both those who know thus and those who do not--'By siege they conquer great citadels.' They made ready an arrow, Agni as the point, Soma as the socket, Visnu as the shaft. They said, 'Who shall shoot it?' 'Rudra', they said, 'Rudra is cruel, let him shoot it.' He said, 'Let me choose a boon; let me be overlord of animals.' Therefore is Rudra overlord of animals. Rudra let it go; it cleft the three citadels and drove the Asuras away from these worlds. The Yajur Veda (Taittiriya Sanhita) translated by Arthur Berriedale Keith, 1914, (http://www.sacred-texts.com/hin/yv/yv06.htm retrieved 19-6-2013).

14. Zvelebil, Kamil V., Ananda-Tandava of Siva-Sadanrttamurti, Institute of Asian Studies, Chennai, 1998, p34-36.

15. Shankar, Bindu S., Dance Imagery in South Indian Temples: study of the 108 karana sculptures. Dissertation, The Ohio State University, 2004, p 79-80, and p 235-274.

16. Oral communication from Raja Deekshithar to the author. He in turn learned this from his elders while young.

17. We would have liked to offer the reader an illustration of this Pallava temple but photography was not allowed.

18. The names Kailasha and Meru are often used interchangeably for the Cosmic Mountain.

19. Shankar, Bindu S., Dance Imagery in South Indian Temples: study of the 108 karana sculptures. Dissertation, The Ohio State University, 2004, p.249.

20. Shankar, Bindu S., Dance Imagery in South Indian Temples: study of the 108 karana sculptures. Dissertation, The Ohio State University, 2004, p 235-274.