Testimonials in Stone - Pallava Rock-cut Architecture - II

Saurabh Saxena

Second Phase - The Riddle of Atyantakama

Note - All images in this article are by Saurabh Saxena. For the 1st part of this article, see this link.

Simhavishnu Pallava in Adivaraha Cave 

Monuments of the second phase of the Pallava rock-cut architecture are mostly concentrated in a small town of Mahabalipuram. There are many theories about the foundation of this town. Kaskkudi grant of the Pallava king Nandivarman II mentions about a naval expedition sent by an earlier Pallava king, Narasimhavarman I, to help his friend Manavarman, an ousted prince of Ceylon. Many scholars have suggested that this naval expedition was probably sent from the port of Mahabalipuram. Narasimhavarman I (630-668 CE), who also assumed the title ‘Mamalla’, founded and consecrated this town as Mamallapuram¹.  Another theory suggests that the town was in existence before Narasimhavarman I and was known as Kadal-mallai. It is believed that saint Bhutattalvar was born in this town around fifth-sixth century².

Atiranachanda Mandapa

Dharmaraja Mandapa

Mahabalipuram provides an advantage in studying evolution of architectural styles however it also presents various complexities for which no easy explanations are available.  There are about eleven rock-cut shrines at Mahabalipuram constructed in various architectural styles. An architectural study relied upon two aspects, first is epigraphy and the second is style. Though the monuments of Mahabalipuram excel on both theses aspects however instead of clarifying things it baffles more.

Pallava inscriptions are written in regnal years of their kings instead in any Common Era like Vikram Era or Saka Era. This results in difficulties for age determination of these monuments. Another difficulty arises due to usage of titles by these kings instead of their coronation names. As many kings assumed same titles it becomes difficult in fixing authorship. Evolution in architectural style is usually not very sudden that artists start adopting new style leaving the old style aside. There are monuments in Mahabalipuram where amalgamation of old and new style is seen further intriguing the matter.

 The first phase of the Pallava rock-cut architecture is characterized with massive and huge pillars which were square on top and bottom with an octagonal shaft in middle. This style is popularly known as the Mahendra-order. Scholars in majority have assigned the Pallava king Narasimhavarman I, sond and successor of Mahendravarman (580-630 CE), as the initiator of the artistic activities at Mahabalipuram. Great-grandson of Narasimhavarman I, Narasimhavarman II Rajasimha (700-728 CE), was the last king who contributed to these activities at Mahabalipuram. This provides a window of less than hundred years during which art and architecture flourished in this small town.

Varaha Mandapa

Mahishasuramardini Mandapa

Though the construction phase at Mahabalipuram was started with Narasimhavarman I, however few rock-cut shrines here depict striking resemblances with the Mahendra-order. Front façade of Kotikal Mandapa is supported on two pillars and two pilasters. The pillars are constructed in Mahendra-order. The shrine is simple and resemble very much to Mahendravadi cave of Mahendravarman. However there is one unique feature, the main cell on rear wall has two female dvarpala suggesting that the shrine was dedicated to some goddess, most probably Durga. None of the early phase shrines of Mahendravarman were dedicated to any goddess. There are two probabilities, either this shrine was constructed during the reign of Mahendravarman as suggested by its architectural style or it was constructed in the reign of Narasimhavarman I as he was the founder of this town so no monument earlier than his are supposed to be found here.

Though Kotikal Mandapa does not help much from epigraphical aspect, Dharmaraja Mandapa contains a very important inscription. This simple and very ordinary shrine is very similar to other earlier Mahendra shrines. Its front façade is supported on two pillars and two pilasters forming three bays. It is divided into two sections, ardha-mandapa and mukha-mandapa, by two rows of pillars. All the pillars are carved in typical Mahendra-order. There are three cells in its rear wall; central cell has two dvarpalas however these have been scraped probably during the Vaishnava resurgence. From their outline they do not look very huge as with that of Mahendra shrines and also they do carry any club. But the most important feature of this cave is a long inscription which referred this shrine as Atyantakama-Pallaveshvara-griham, The Temple of Atyantakama Pallava King.

There was no Pallava king with name Atyantakama however three different kings held this title, Narasimhavarman I Mamalla, his grandson Parameshvaravarman I (670-700 CE) and his great-grandson Narasimhavarman II Rajasimha (700-728 CE). As this shrine is very simple so this could be the earliest one on the site hence should be assigned to Narasimhavarman I. But it is very intriguing that the similar inscription is also found in few other monuments such as Ganesha Ratha which are built in later style. These later monuments cannot be assigned to Narasimhavarman I but are most likely built in the time of Parameshvaravarman I or Rajasimha. In that case is there a possibility that though Narasimhavarman I excavated this shrine but a later king inscribed this inscription, which is very hard to believe.

Koneri Mandapa is a fine example of amalgamation of two different architectural styles. This five bay rock-cut shrine is built into two sections, ardha-mandapam and mukha-mandapam, separated by two rows of pillars. Pillars of the front row are constructed in characteristic Mahendra-order however those of the inner row are of different style. These are slender and more ornamented in comparison to massive pillars of Mahendra-order. Constructed with multi-faceted circular shaft tapering in circumference from bottom to top, they support a round abacus on top of the capital. This distinctive change is style is often quoted as transitional phase from Mahendra-order to Narasimha or Mamalla-order. Front façade of this shrine has kudu (horse-shoe shaped ornamentation on cornice), sala (oblong shrines arranged horizontally above cornice) and cloisters connecting these salas. This ornamentation is a deviation from Mahendra phase. Why the artists selected two different pillar schemes for the same shrine, unless they wanted to point out this evolution phase specifically?

Koneri Mandapa

Kotikal Mandapa

Ramanuja Mandapa

Mahishasuramardini Cave is perhaps the first milestone where Mahendra-order was completely replaced with a new architectural style, Mamalla-order. The front façade of this cave is supported on four pillars and two pilasters. Pillars are carved in similar transitional style as seen in Koneri Mandapam, circular, multi-faceted with ornamenting bands in middle. A cell is provided in middle of the rear wall accompanied with a mandapam in front. This mandapam is supported on two pillars. The pillars are almost similar as those of the front façade but with one striking difference, the lion base. These lion base pillars became the characteristic statement of the Mamalla-order. If this is true then Mahaishasuramardini Cave would have been a later construction then that of Koneri Mandapam. However the front façade of Mahishasuramardini cave is left unfinished where a scheme similar to Koneri Mandapa is tried. If this scheme was already there by the time of Koneri Mandapam then why the façade of this later cave is left unfinished?

Ramanuja Mandapa rock-cut shrine depicts the Mamalla-order of lion bases in its front pillars. Though this shrine is simple in its style and architecture however there are certain peculiarities observed. Instead of dvarpalas on either extreme of front façade there are two miniature shrines carved. These miniature shrines display true Dravidian architecture style. This triple celled shrine was once the most completed monument however it lost most of its characteristics in remodeling during Vaishnava resurgence. There was a Somaskanda panel, now scrapped, in the middle cell which suggests that the shrine was probably dedicated to Hindu Trinity with Shiva housed in central cell. It is believed that Somaskanda icon was started during Parameshvaravarman I and reached to its zenith in Rajasimha period. The style of Somaskanda icon here points to Rajasimha period. There is an incomplete inscription consisting of last three lines of the Dharmaraja Mandapam inscription. Authorship of this shrine is not easy to determine. On architectural style it suggests that it was constructed during Narasimhavarman I. However an incomplete inscription praising Shiva puts this shrine to Parameshvaravarman I as similar inscription is found in Dharmaraja Mandapa. But Somaskanda panel in central cell suggests that this shrine is related to Rajasimha. Is there a possibility that the shrine was constructed by Narasimhavarman I and modified by later kings?

Varaha Mandapa is the grandeur of the Mamalla-order as this is the most complete shrine among all in Mahabalipuram. Excavated in similar mandapa style as other Pallava rock-cut shrines, its front façade is supported on two pillars and two pilasters. Pillars and pilasters are carved with lion (vyala, a lion like mythological animal) bases. Lions of pillars face front while those of pilasters are facing each other. There is a cell in middle of the rear wall with two dvarpalas on either side. These dvarpalas are very different from huge & massive dvarpalas of Mahendra’s time. These are slender, flexible in their body proportions and do not carry clubs. The shrine also boasts of four bas-reliefs which may be rated among the best art specimens of India. All the features of this shrine reflect progressive movement of art and style. However it is very surprising that this beautiful piece of superb artistic execution does not have any inscription. There are inscriptions on much simpler caves but not on this one, strange that the creator of this majestic piece remained to be unknown, why?

Adi-varaha Perumal rock-cut shrine has been extended with a later period mandapa hiding its external features. It is one of the most important shrines in study of the Pallava history as there are two relief sculptures of the Pallava royal personages inside. Though there are inscriptions on these sculptures however it only complicates the matter. ‘Sri-Simhavinna-pottr-athirajan’ and ‘Sri-Mahendra-pottra-thirajan’ are two inscriptions found on these two relief panels. Complexity arises in identification of who these two Pallava kings, Simhavishnu and Mahendravarman, are. Simhavishnu (580-630 CE) was the father of Mahendravarman I which suggests that these two reliefs might be representing the Pallava king Simhavishnu and his son Mahendravarman I. However scholars were not able to assign any rock-cut monument to Simhavishnu and it is believed that rock-cut architecture was started during the reign of Mahendravarman I as suggested by his inscription at Mandagapattu. Grandson of Simhavishnu, Narasimhavarman I, also held title of Narasimhavishnu or in short Simhavishnu as suggested by his inscription at Badami. Rao Bahadur H Krishna Sastri suggests that these two royal portraits represent Mahendravarman I and his son Narasimhavarman I³. However there is a little issue with this identification, king under Simhavishnu title is shown seated while king under Mahendravarman title is shown standing. Isn’t it strange that father is shown standing while his son in opposite panel is shown seated?

K R Srinivasan suggests that these two royal figures are Narasimhavarman I and his son Mahendravarman II⁴. He argues that as per a Chola inscription this shrine is referred as Parameshvara-Maha-Varaha-Vishnu-griham. It seems that Parameshvaravarman I completed this shrine which was probably started during the reign of his two predecessors. He put the portraits of those predecessors while gave his name to the shrine. This proposition is further strengthen by last two lines of similar epigraph which is found in Ramanuja and Dharmaraja mandapam. These last two lines are in praise of lord Shiva. This is quite strange to find such an inscription in a Vishnu temple. Thus as the temple was started during the reign of Parameshvara’s predecessors hence though it was meant for Vishnu, but as Parameshvara was a staunch Shaiva he left his signature inscription in this shrine as well.

To make the matter more complicated, R Nagaswamy proposed another explanation of this problem. He suggests that all the monuments at Mahabalipuram are constructed under a single authorship of Narasimhavarman II Rajasimha. Two strong points to support this theory are, first Rajasimha bore the title Atyantakama and as many monuments here are left unfinished it suggests that the activities were suddenly stopped at this site. This sudden stop in activities is hard to explain if the activities were started during Narasimhavarman I and continued till his great-grandson Rajasimha for almost hundred years. Hence the author of all the monuments is same. He further suggests that the two royal portraits represent the Pallava king Rajasimha and his son Mahendravarman III⁵.

Atiranachanda Mandapa is located at Saluvakkuppam, few kilometers away from Mahabalipuram town. This rock-cut shrine has two pillars and two pilasters in front. The pillars are in typical Mahendra-order. There is a cell in center of its rear wall which houses a Somaskanda panel inside. There are two inscriptions found in this shrine, same inscription written in two different scripts, Pallava-grantha and Devnagari. The inscription refers this shrine as ‘Atiranachanda-Pallaveshvara-griham’. Pallava king Rajasimha born the title Atiranachanda as suggested by his inscription at Vayalur temple. This strongly suggests that this shrine would have been constructed by the Pallava king Rajasimha. If it is true then why he used Mahendra-order for pillars when a more matured style, Mamalla-order, was already in practice?

If it is accepted that Rajasimha was the only author of all the monuments then most of the complexities can be explained easily. However a question arises that if Rajasimha was the author then what was happening in art activities between the period of Mahendravarman I (580-630 CE) and Rajasimha (700-728 CE). Were the Pallavas in hibernated mode for about hundred years? It is hard to believe as Mahendravarman I pioneered in art and architecture field and his successors would surely have followed him in this regard. Also if Narasimhavarman I is the founder of Mahabalipuram town then it is very obvious to find his monuments at this site. Many of the monuments at this town are assigned to be constructed by some king referred as Atyantakama. All the kings involved in Mahabalipuram authorship dispute born this title. The riddle of Atyantakama is still unexplained.

Contact - msg4saurabh@gmail.com 

"Portraits" of Mahendravarman and his two Queens in Adivaraha Cave

¹ G Jouveau-Dubreuil, Pallava Antiquities Vol I, pp 75

² Nagaswamy, R, Mahabalipuram – Monumental Legacy, pp 2

³ H Krishna Sastri, Rao Bahadur, Two Statues of Pallava Kings and Five Pallava Inscriptions in a Rock Temple at Mahabalipuram, pp 4

⁴ Srinivasan, K R, Cave Temple of the Pallavas, pp 173

⁵ Nagaswamy, R, Mahabalipuram – Monumental Legacy, pp 25

Read more...

Testimonials in Stone - Pallava Rock-cut Shrines (Part I - Mahendra Style)

Saurabh Saxena
Saurabh Saxena is primarily a computer engineer by profession. However, he is also interested in exploring the monuments of India on his own and he likes to discuss them with the experts in this discipline

Photo - Saurabh Saxena

Early temples of India were constructed with perishable material like mud, plaster, timber, brick etc that could not endure the ravages of time. Probably to leave the legacy of their art that could survive for a long time, our ancestors moved to rocks for their creations. Excavating caves out of a solid rock for living and worship purposes was a common practice in ancient India. A few of such caves exist in Bihar, excavated for the residence of Buddhist monks during the reing of Mauryan Emperor Ashoka. Ajanta and Ellora are world renowned sites where we see caves carved out for residence as well as worship purposes. There are other caves excavated in Maharashtra and Orissa, which predate Ajanta and Ellora caves.

When this movement was at its peak in Northern and Central parts of India, the Southern part of India was still untouched as far as the knowledge of such usage of rocks is concerned. The Chalukyas of Badami in modern Karnataka were the first to build a rock-cut shrine in South India in the latter half of 6^th century at Aihole. The Pallavas of Kanchipuram were the first who introduced this form of art in Tamilnadu about a century after the Chalukyas.

Mahendravarman (580-629 BCE), one of the great Pallava rulers, is attributed as the initiator of this form of art in Tamilnadu. He was a multi-talented personality. Apart from his contributions to architecture, he also contributed in the fields of music and literature. It is generally believed that Mahendravarman got his inspiration from the caves of Undavalli and Mogalrajapuram in Andhra Pradesh, excavated by Vishnukundins in the 6^th century. Whether it was the inspiration from those caves or it was his own ingenious idea, he put the foundation of an art which witnessed tremendous development and progress within a century, culminating finally into the majestic shrines of Mahabalipuram. 

Shatrumalleshvaralaya, Dalvanur, Lakshitayana, Mandagapattu and Mahendra-Vishnu-Griham, Mahendravadi (Photo - Saurabh Saxena)

Lakshitayatana rock-cut shrine at Mandagappatu, supposed to be the first shrine of Mahendravarman, is a very simple structure. We cannot term it as rudimentary. However, it lacks ornamentation and style in comparison to the shrines of later times. This triple celled shrine is dedicated to the Hindu Trinity: Vishnu, Shiva and Brahma.

Excavated out of a rock boulder, this cave has three bays formed by two pillars and two pilasters. Beyond the pilasters are carved two dvarpalas (doorkeepers) on either side. Pillars of Mahendravarman’s caves are composed of two cubical parts, one at the bottom and the other at the top. In between these, a middle octagonal shaft is carved out. The style of these pillars became the signature style of Mahendravarman and many scholars term this as ‘Mahendra Style’. Most of his shrines are triple bay, formed by two pillars and two pilasters, structures with a single cell either carved in the inner wall or on sides depending upon the direction the cave faces. His shrine at Pallavaram is a five celled shrine while the one at Trichy is with five bays single cell shrine.

Before we move forward, let’s have a look at the foundation inscription (1)

of Mandagappatu. It reads - 

Atdnishtakam-dru (malo)Hamsudham (vichitra-chi)-ttena Nimarpitan-nripe

(na)-Brahmo-Shvara-Vishnula(kshi)-tayanam

Translation - 

This brick-less, timber-less, metal-less and mortar-less temple, Lakshitayana, which is a mansion for the Brahma, Ishvara and Vishnu was caused to be made by the king Vicitra-citta.

Though the inscription does not state clearly whether it is the first of its kind, the enthusiasm of the king and mention of brick, timber, metal and mortal specifically in the inscription suggests that it is probably the first attempt in this direction, and hence the creator was overwhelmed at the success and so inscribed such words over the pillar.

It is believed that Mahendravarman was converted from Jainism to Shaivism by Saint Appar and that he dedicated his monuments to various sects regardless of his own affiliation. While his first shrine was dedicated to Hindu Trinity, another one called Mahendra-Vishnu-Griham at Mahendravadi was dedicated to Vishnu. Emerging like an outcrop in a vast land, Mahendravadi cave is excavated out of a free standing rock boulder. Here we find an advanced architectural style in the form of lotus medallions which are carved on the cubical face of the pillars. This lotus medallion motif later became an integral feature of his shrines. Upper front face of the shrine, above the pillars, is almost plain; however, we see some unfinished work in form of square blocks. This unfinished part gives us vital information on the process of excavating these caves. It seems that after the initial preparation of the rock face, such square blocks were made by carving the intersecting lines and after that these were chiseled off to get plain surface. Was there something planned to adorn on this face? We get an answer in another shrine of Mahendravarman.

Satrumalleshvaralaya rock-cut shrine at Dalavanur is a south facing cave with a similar front portal as that of Mahendravadi. Similar to Mandagapattu, we find dvarpalas (doorkeepers) on the front portal here but these were missing in the case of Mahendravadi. The shrine is dedicated to Shiva, and the inner shrine (antechamber) is carved on the west wall so that it could face east. Other than the regular characteristics of Mahendravarman’s style, we have two new features seen in this cave.

First is a row of kudus (horse-shoe shaped arch windows) on the front face, above the pillars, of the cave. This architectural feature is very prominently found in Buddhist shrines hence it is possible that the inspiration to include this here probably came from the Buddhist shrines. A human head, peeping out, is carved inside this arch. At few places, instead of human heads, other motifs like lotus flower are seen.

The second feature of interest is the inner shrine. As the cave faces south, to make the main cell facing east, a shrine was carved on the western wall of the cave. This shrine got its pair of dvarpalas as well. A Shivalinga is placed inside the cell; however this is a later addition. We do not find any image inside the cell, as is also the case with other rock-cut shrines of Mahendravarman, which suggests that during that time no stone image of deity was put inside the shrine. Probably, either the deity’s image or symbol was carved on some wooden plaque which could be attached to the back wall or the image was painted on the back wall of the cell for worship.

Shiva-Parvati, Avanibhajana Cave-Shrine, Siyamangalam (Photo - Saurabh Saxena) 

Avanibhajana-Pallaveshvara rock-cut shrine at Siyamangalam, another shrine of Mahendravarman, showcases many advanced features in the rock-cut sculptural art. For the first time we find relief sculptures carved on the faces of the pillars. One such sculpture, of Shiva as Nataraja, is probably the first representation of this form in Tamilnadu. One another sculpture represents Shiva as Vrishabhantika, which depicts Shiva standing on support of his mount, Nandi, accompanied by Parvati. Siyamangalam also has two magnificent warriors, carved out in niches, which can be termed as the first representation of dynamic forms. In a nutshell, Siyamangalam shows how this art flourished and advanced during Mahendravarman’s time.

Dancing Shiva, Siyamangalam (Photo - Saurabh Saxena)

The pinnacle of Mahendravarman’s time is witnessed in his shrine, locally known as Upper Cave Temple at Trichy. This south facing cave has four pillars and two pilasters resulting in five bays. It has a single shrine inside, carved on the western wall of the cave. On the side wall, adjacent to this shrine, is a marvelous bas-relief of Shiva as Gangadhara. This sculpture has been considered as a masterpiece by many scholars.

Lalitankura Cave-Shrine, Trichy (Photo - Saurabh Saxena)

The whole canvas is turned into a theme of Ganga’s descent. Shiva is standing majestically and in graceful manner, resting one foot over a demon, in the center of the canvas. He is holding his tresses to support Ganga who is shown in a female form and descending into his tresses. His stance is tribhanga (bent at three places) but seems very firm to accommodate the force of Ganga’s descent. Many of Shiva’s features are carved very finely, be it crescent moon on his right side of his head or a skull on his left side of the head on the crown. Ornaments on his body are minimal, which is also one of the distinctive features of the Pallava sculptures. Drapery on his body is very nicely folded till his ankles. Ganga, shown in a small female form, is bending by her waist and her lower part of the body is blending with a cloud, depicting that she is descending from the heavens. Overall, this is a marvelous composition which is indeed a masterpiece of that time. 

Gangadhara Shiva, Lalitankura Cave-Shrine, Trichy (Photo - Saurabh Saxena)

Michael Lockwood was the first scholar to propose the argument that the image of Shiva here is a symbolic representation of Mahendravarman himself and his receiving the Ganga on his locks is a symbolic representation of the conquest of Mahendra over the Kaveri valley, since Kaveri is regarded as the Ganga of the South (Tenganga) in Tamil Nadu. Mahendra got this cave excavated after his conquest of this region. Lockwood gives the reference of the inscription carved on the wall of this cave, which says that because Parvati felt jealous seeing Shiva receiving another woman on his locks, she assigned Mahendravarman as the lord of Kaveri, which is the Ganga flowing in this region (2). Since then, many scholars have followed Lockwood's argument to show the symbolic parallel between the divine and the royal images at this shrine. The inscription cited by Lockwood is also the first reference of the Tamil version of the legend of Gangavatarana (albeit composed in Sanskrit here), which is a modified version of the Samskrit version prevalent in the North India. The element of Parvati feeling jealous is a special characteristic of the Tamil version of this legend, which doesn't exist in the Sanskrit versions. The image itself shows the Sanskrit version, as there is no jealous Parvati seen here, but the Sanskrit inscription adds extra dimensions to this image in terms of symbolic meanings. R. Nagaswamy later objected to the arguments of Lockwood (3), but Lockwood responded by saying that Nagaswamy has not been able to propose his case in a convincing manner.

There are many more such shrines of interest which were either carved by the order of Mahendravarman or were executed by his feudatories. In the next article, we will see the progress of this art during the time of Narasimhavarman I, successor of Mahendravarman, who converted the town of Mahabalipuram into a workshop of architecture.

 

Notes - 

1.    T A Gopinath Rao, Epigraphia Indica Vol XVII, Archeological Survey of India, New Delhi, 1983, pp. 14-17.

2. See Michael Lockwood, Gift Siromony and P. Dayanandan, Mahabalipuram Studies, The Christian library Society, Madras, 1974.

3. See Link here. 

References

1. Dubreuil G-Jouveau, Pallava Antiquities, Asian Educational Services, Madras, 1917.
2. Dubreuil G-Jouveau, Dravidian Architecture, Asian Educational Services, Madras, 1917.
3. Longhurst A H, Pallava Architecture, Archeological Survey of India, New Delhi, 1930.
4. Srinivasan K R, Cave Temples of the Pallavas, Archeological Survey of India, New Delhi, 1964.

 Saurabh Saxena can be contacted at msg4saurabh@gmail.com. 
 

Read more...