தமிழ்த் தேசியம்

"To us all towns are one, all men our kin.
Life's good comes not from others' gift, nor ill
Man's pains and pains' relief are from within.
Thus have we seen in visions of the wise !."

- Tamil Poem in Purananuru, circa 500 B.C 


 Whats New

Trans State Nation Tamil Eelam Beyond Tamil Nation Comments Search
Home > Culture & the Tamil Contribution to World Civilisation  > Sathyam Art Gallery >  Tamil Art & Architecture - T.V.Mahalingam 

Tamil Art & Architecture 

 Professor of Ancient History and Archaeology, University of Madras
[paper presented at Second International Conference Seminar of Tamil Studies,
  Chennai, Tamil Nadu, January 1968]

Introduction Architecture Sculpture Chola Bronzes Discussion


In all major civilisations of the world art has been one of the primary media through which the creative urge and genius of the people have been succinctly expressed. 

The term 'art' is so comprehensive in its connotation that it includes in its fold different fields of human enterprise. But in our consideration here its meaning is restricted to the fields of architecture, sculpture and painting, the three main channels through which both the artistic mastery and profound philosophical truths of ancient India have been expressed. 

The Tamil country has been, throughout the successive historical periods, the meeting ground of many an artistic motif, design and norm, so much so that in her sculptural and architectural contributions the harmonious blending of diverse regional idioms are manifestly evident. It is not possible to recount in its entirety the history of the development of Tamil art, from its evolutionary beginnings to its consummation and final exhaustion, within the limited time at our disposal; what is attempted here is only a panoramic survey of the salient features of Tamil art. It is necessary here to point out that the name Tamil Nad in our discussion does not mean the modern State of Madras but refers to the areas under the cultural hegemony of Tamil - the area roughly corresponding to large parts outside it, and adjoining it.


Before sketching in outline the evolution of architecture it is necessary to acquaint ourselves with the major types of extant structures. These basic shapes are fivefold, viz., square (caturasra), rectangular (ayatasra), elliptical (vrittayata), circular (vritta) and octagonal (astasra). Generally speaking the plan of the temple was conditioned by the nature of the consecrated deity. The shrine of the reclining Ranganatha, for example, can only be rectangular. The basic shapes are amply reflected in the superstructure of the vimas . Though square and rectangular shrines are frequently met with, circular and octagonal shapes are very rare. However these forms are represented in the sikhara of the vimana. The apsidal form, a derivative from Buddhist architecture, was popular up to the 10th century in the Tondaimandalam, after which it declined in usage.

Mention should also be made here of the temples which have more than one shrine in the vertical order. This is to be found in a handful of Vaisnava temples as those at Kancipuram, Uttiramerur, Madurai, Tirukkostiyur etc. Three shrines, one above the other, are found in these and are intended for the seated, standing and reclining forms of Visnu.

Unlike other parts of India the architectural history of the Tamil country starts only with the beginning of the seventh century A.D., the monuments built before that period having perished. In early Tamil literature we hear of such structures as koyil, maddam, nagaram, palli, pali. etc., which are apparently references to temples or religious edifices. Presumably they were built of impermanent materials which have succumbed to the ravages of time.

The earliest extant monuments in the Tamil country are the rock-cut caves scooped out under the Pallavas, and following them by the Pandyas, Muttaraiyars, and Atiyas. In his inscription in the cave of Laksitayatana at Mandagappattu, South Arcot district, Mahendravarman I (610-630 A.D.) declares that he caused the construction of the temple for Siva, Visnu and Brahma without the use of conventional building materials like brick, timber, metal and mortar; and the tenor of the language has been taken to indicate that the king was introducing a new mode of architecture by scooping out the cave. Many other cave temples are definitely attributable to Mahendravarman on the authority of his inscriptions in them. These include the excavations at Pallavaram, Mahendravadi, Mamandur, Tiruchirapalli, Slyamangalam and Dalavanur. The Vasantesvaram at Vallam was also excavated in Mahendravaraman's reign by a feudatory of his. 

Besides these caves of definite authorship, those at Kuranganilmuttam, Vilappakkam, Aragandanallur and the Rudravahsvara cave at Mamandur are stylistically attributable to the period of Mahendravarman. These caves of Mahendra are simple in plan and consist of a mandapa with one or a few shrines. The sculptural decoration of the caves is inconspicuous. The pillars in them are equidistant and have square sections both on base and top with the portion in between chamfered octagonally. In the square section are seen delicate carvings of lotus medallions. The pillars and pilasters carry on top massive corbels with beams. This 'Mahendra style' was continued by his son and successor Mamalla, the famous Narasimhavarman I, who, however, introduced certain variations in some of his caves. In these the entablature is almost completely finished, unlike in those of the Mahendra variety. Besides kudu arches in the cornice, it carries salas, karnakutas and alpanasikas. The pillars in Mamalla's caves are not only taller but also more slender than those of his father. The strutting figure of a lion ro vyala as the base of the pillar is a notable feature. Again in Mamalla's caves one can also find large bas-reliefs on walls in striking contrast to their plain nature in all but one of Mahendra's caves. The Konerimandapam, Varahamandapam, Mahisamardanimandapam, Trimurti cave, Adivaraha cave, Ramanujamand. apam, etc. - all at Mahabalipuram - are typical examples of the Mamalla types of rock architecture.

The Pandyas, who were ruling in the extreme south of the Tamil country, appear to have soon adopted the rock-cut technique and developed certain interesting variations in their excavations. It is possible that the cave at Pillaiyarpatti is one of the earliest Pandya attempts in the rock medium as evidenced by the archaic palaegraphy of the inscription in Vatteluttu characters in it. The Siva cave shrine at Malaiyadikurichi is assignable on the basis of an inscription to the second half of the seventh century and the Narasimha cave at Anamalai and the Subrahmanya cave at Tirupparankunram are on the same ground datable respectively to 770 to 773 A.D. 

At Tiruttangal, Piranmalai, Kudumiyamalai and Sittannavasal are to be found other caves of the Pandyas. Though similar to Pallava caves in plan and design, the Pandya examples differ from them in their adoption of certain Calukyan features such as the introduction of the rock-cut linga and Nandi and sculptural representations of Ganesa and Saptamatrkas. The pillars are large and reminiscent of those of the Mahendra variety with corbels generally with a plain level. In this movement of scooping out live rocks for divine abodes minor dynasties like the Atiyas and Muttaraiyars also participated, though stylistically their excavations are much akin to those of their political master. The cave at Namakkal is evidently an Atiya enterprise while Muttaraiya involvement may be seen at Tiruvellarai, Narttamalai, Kunrlandarkoil etc.

Under Narasimhavarma I, Pallava rock-architecture took a new turn. besides cutting into rocks for caves, attempts were made to cut out monoliths from rocks. The rudiments of this practice are to be found in the carved-out stupas in the caves of Western India and the vimana-form in the Tawa cave at Udayagiri but it was at Mahabalipuram under the Pallavas that it found a full and eloquent expression.

 Architecturally they depict the external aspects of contemporary brick and timber structures. There are as many as nine monoliths at Mahabalipuram of which the five, named after the Pandavas and Draupadi, are a well-known assemblage of contiguous excavations, the other examples are the Ganesa ratha, Valayankuttai ratha and the two Pidari rathas. As they represent varying architectural designs they are of primary importance for any study of the plan and different zones and the details of the Yima-nas. The Dharmaraja-ratha is three-storeyed with a square viguana and an octagonal dome. Though the Arjunaratha is similar to this it is two-storeyed. The Bhimaratha has a wagon-top roof and is single-storeyed unlike the Ganesa ratha, another example of wagon top roof, which is double-storeyed. The Draupadiratha is hut-shaped and is square in plan and its roof is domical. The Sahadevaratha represents the apsidal form with its back resembling that of an elephant, a feature high-lighted by the carving of a huge elephant by the side of the monolith. 

The only non-Pallava monolith in the Tamil country is Kalugumalai which was cut-out under the Pandyas. This has been cutout, like the Rastrakuta monoliths in the Deccan, by entrenching all round and not by free cutting of standing rocks as in the Pallava domain.

Though the rock medium appears to have continued for some more time it was soon replaced by structural temples. This movement, as available evidences indicate, appears to have first started under Narasimhavarman I's grand-son Paramesvaravarman (669-691 A.D.), though it is not unlikely that the practice was still older. A few pillars in the typical Mahendra style, one of them with an inscription of Mahendravarman I, found in the Eltamranatha temple at Kancipuram seem to suggest that even at the beginning of the seventh century structural mandapas were built. The presence of Pallava pillars at Sivanvayil, Kuram, Vayalur, Tirupporur etc., is enough to confirm this.

The Vidyavimta Pallavesvaragriha at Kuram built by Paramesvaravarman I is an early structural edifice. The provision of a series of vertical and horizontal slabs instead of a full bAitti is an interesting and early feature in this temple. While this is a small temple and reflects the modest nature of the enterprise, the temples of the next reign are large in size, elaborate in plan and rich in architectural and sculptural decorations. With the accession of Narasimhavarman II Rajasimha the history of Pallava architecture enters upon a new and eventful phase. While the temples of Kailasanatha at Kancipuram, Talagirlsvara at Panamalai and the Shore temple at Mahabalipuram are indisputably assignable to his reign on epigraphical grounds, a large number of other smaller temples are also stylistically akin to them. The temples of Vaikuntanatha, Muktesvara and Matangesvara at Kancipuram are said to be slightly later and belong to the reign of Nandivarman Pallavamalla. 

The Kailasanatha is four-storeyed and is an example of sandharaprasada containing two walls providing an ambulatory. The storeys are decorated with architectural designs like kutas, kostas and panjaras. The pillars in structural temples are with rampant lions generally and with elephants, nagas and bhulas at times. Niches are to be seen in both the rock-cut and structural temples and have a makaratorana decoration on their top, the makaras in them having floriated tails overflowing on the sides. The corbels are generally curved in profile with the taranga (wave moulding) ornament and a median band. The gopuras are absent in these early temples. 

In the Kailasanatha at Kanci and the Shore Temple at Mahabalipuram there are faint but unmistakable suggestions of gopuradhvaras which were to evolve into towers. Another feature of these early structural temples is the almost prodigal sculptural embellishment of the exterior walls. The carvings are invariably those of deities, a few of which appear to be fresh inceptions from the Calukyan area.

The Colas who supplanted the Pallavas about the middle of the ninth century as a political force continued the latter's artistic activities. For about five centuries a large part of the Tamil country besides peripheral regions in contiguous areas in Andhra, Karnataka and Kerala were under their sway which they studded with hundreds of temples. On the basis of certain accepted notions regarding the evolution of temple architecture and on the authority of numerous inscriptions it is now fairly possible to determine the dates of most of the Cola monuments. Though the periodisation of South Indian art-history is even now a subject of debate it is conceded by most scholars that the Cola temples are broadly divisible into three groups: 

the first group belonging to the period from the accession of Vijayalaya to the accession of Rajaraja I (i.e., 850-985 A.D.); 

the second group assignable to the period from the accession of Rajaraja I to the accession of Kulottunga (985-1070 A.D.); and 

the third group comprising the period from the accession of Kulottunga I to the decline and fall of the Cola empire under Rajaraja III and Rajendra III (1070-1270 A.D .).

The temples of the first group are many which in stylistic characteristics break away from the structural temples of the Pallavas. In the Pallava temples the lowermost tier of the vimana is extended to the vestibule in front of the shrine, while this is not found in early Cola temples, the only exception being the Vijayalayacolisvaram at Narttamalai, which according to recent researches is not a Cola but a Muttaraiya edifice. The torus moulding in the basement which is chamfered in Pallava temples continues to be so in the Cola period for sometime but soon gets a rounded shape. The cornice in Cola temples is no longer a projecting tier as it is in Pallava monuments but gets a flexed shape. The old lion and vyala motifs in pillars are also dispensed with, though they linger for sometime in a few temples. Further, the corbels in pillars get an angular profile and are bevelled, resulting in a triangular, tenon-like projection. The absence of extravagant sculptural decoration on the exterior of the shrine walls is another distinguishing feature.

Though typical early Cola examples are numerous, special mention must be made of those at Kilaiyur, Srinivasanallur, Kumbhakonam, Erumbur, Pullamangai, Punjai and Kodumbalur. The introduction of sub shrines for attendant divinities (parivara-devatas) noticed in these temples reveals elaboration and development of the temple complex. In fact the beginnings of this practice are to be discerned even in the latter Pallava temple of Virattanesvara at Tiruttani built under Aparajita. 

This temple, though Pallava in name, is Cola in design and style and chronologically almost coeval with some of the Cola monuments enumerated above. The parivdra shrines, usually eight in number, were meant for attendant deities like Ganesa, Subrahmartya, Surya, Candra, Saptamatrkas, Jyestha, Candikesvara and Nandi. The gopuras of this period continue to be inconspicuous, the vimănas, dominating the temple complex. 

Generally speaking, temples built under Aditya and Parăntaka contained only three niches in the shrine walls, one on each wall, and two niches in the walls of the ardhamandapa, again one on each wall. While the niches in the southern and northern walls of the ardhamandapa carried respectively carvings of Ganesha nd Durga, those of the main shrine were intended for Daksinämurti and Brahmă. The niche in the rear wall offered scope for variation, the enshrined deity being either Lingodhbhava or Visnu, Harihara or Ardhanărisvara. But even in two very early temples - those at Srinivasanallur and Kumbhakonam - the tendency to multiply the niches is found, the additional niches carrying what looks like portraits. This tendency has been developed in the temples built by Sembiyan Mahadevi, mother of Uttamacola at such places as Tirukkodikkăval, Sembiyan Mahădevi, Anangur, Aduturai, Tirunaraiyur, Kuttälam, etc., where the additional niches carry such iconographic types as Natarăja, Bhiksätana and Ardhanri besides Agastya.

These early Cola architectural traditions are carried to those of the later Cola period by the temples built under the illustrious Rajaraja and his son Räjëndra. Many are the extant examples assignable to this middle phase, the most famous among them being the Brihadisvara temples at Tanjore and Gangaikondacolapuram. Other temples of this period are those at Tiruvaji, Mëlpadi, Tiruvalańjuli, Tirumalavadi, Tiruvarangulam, Dadapuram, etc. In most of these temples the basement is ornamented with pilasters which carry a cornice. The walls have a greater number of niches and a semi-circular arch (tiruvacci) the centre of which is identical with that of’ the küdu which appears beneath the architrave and over the niche. The introduction of the kumbhapańjara in between the niches is another feature.

The Tanjore temple is undoubtedly the grandest achievement of the age. It was more a monument of triumph than a strict example of temple architecture. It is in this temple that one notices for the first time two gopuras oriented in the same direction. They are architecturally coeval with the main vimana and are referred to in inscriptions as Rajarăjan tiruvasal and Keralăntakan tiruvasal In spite of the massive size of the gopuras the vimăna, rising majestically to a height of 190 feet, continues to dominate and it is only in the subsequent period that a change in the gradation of magnitude takes place.

The multiplication of parivăra shrines and the introduction of a separate shrine for the goddess are the two significant changes in the temple complex effected during this period. Even in the Tanjore temple the Devi shrine is not contemporaneous with the main cella but was built later. The earliest Devi shrine which appears to be definitely chronologically coeval with the main shrine is the one at Gangaikondacolapuram. 

The Devi shrines, known as Tirukkdmakkottams, were thus largely a feature from the reign of Räjëndra. In the temples representing the final phase of Cola architecture a discernible maturity of style is evident. Notable examples of them are to be found at Dăräsuram, Tribhuvanam, Chidambaram and Jambukesvaram. Of the stylistic improvements made in these temples mention must be made of the torus moulding in the basement which is rounded and has a smooth surface, though in a few cases it is orna­mented with vertical grooves or ribs. The makaratoranas become tall with narrow reverse curves on each side; the kumbhapanjaras are also developed and carry on top over the abacus the superstructure of a panjara. The phalaka in the pillars are thinner than those of earlier periods and the padma below it, which is inverted and smooth in early temples, now has petals. The pillars in the mandapas have attached pilasters on their sides, known as Aniyottikal.

Perhaps the most distinguishing feature of the temples of the late Cola phase is the increased height of the gopuras. The five-storeyed gopuras at Tiruvenkädu, Uyyakondan-Tirumalai, Tiruccengättăngudi and Kumbhakonam must belong to this phase. Besides the gopura, pillared mandapas were also built within the temple complex, some of them being shaped in the form of a chariot by the addition of wheels and horses and elephants.

Generally speaking, the characteristics of the early and late Cola temples are shared by Pandya monuments of the respective periods, though minor variations are present in them. 

The next stage of development is, however, seen only in the temples built under the Vijayanagar rulers. The Vijayanagar kings not only built many new edifices, but made many additions to the already existing temples. Such additions are to be found in many places, the most noteworthy among them being Kancipuram, Tiruvannämalai, Chidambaram, Kumbhakonam, Madurai, Srirangam, Vellore, etc. The mandapas become large and conspicuous adjuncts during this period due to the multiplication and elaboration of religious rituals and ceremonial observances. The Kalyănamandapa, Sopănatmandapa, Davana­mandapa, Sndpanamandapa, Alańkăramandapa, etc., are the usual mandapas in addition to the ardha, mukha and mahă mandapas of earlier times. Some of these mandapas are, however, not entirely unknown under the late Colas. A few of them were built outside the temple circuit but not much away from it. These mandapas are essentially pillared halls, open or closed, and contain either a shrine or a raised platform over a huge tortoise either in the centre or behind. They are also notable for their pillars which are rich in sculptural work and to which are attached riders on horse or lion or yăli. The fluted type of simple pillars becomes rare and huge and monolithic ones are often seen. They have ornamental brackets forming their capitals, below each of which is a pendant. This pendant has been in many examples elaborated into a ‘volute which terminates as an inverted lotus bud.’

The niches in the walls are not surmounted by torănas as in Pallava and Cola temples but have a simple pańjara design over them. What is more, the niches are empty, without any image in them. Their old functional character has been lost and they remain a simple ornamental design on the exterior of the wall. The increase in the height of the gopuras and in the number of prăkaras is yet another feature. The gopuras are generally seven­storeyed and are large and tall, especially in the Pandya region. The most typical gopuras of this period are to be found at Kańcipuram, Srirangam, Chidambram and Tiruvannamalai. These are rich with architectural designs like salas, karnakutas and alpanasikas rather than sculptural decoration.

The Vijayanagar mode of architecture was continued by the Nayak rulers of Madurai. In the temples renovated or rebuilt by them, as the ones at Madurai, Rămesvaram and Tirunelveli, the corbels in the pillars show at their ends a plantain-flower-like motif. The gopuras continue to be slender and tall, the typical example being the Vatapatrasayi gopura at Srivilliputtur which is eleven-storeyed. The corridors in these temples, unlike those of earlier periods, are provided with ceilings which are at times painted.


The art of sculpture like that of architecture has a continuous history in the Tamil country and during the course of its development it has, even within the limited geographical area of Tamil Nad, shown more than one regional trait. As the advent of stone for purposes of art does not seem to antedate the beginning of the seventh century, it is likely that in the early centuries images were made of wood or stucco. In a few of the South Indian Temples the main deity in the sanctum sanctorum even today is made either in wood or in stucco and this probably is only a survival of the old practice. Due to their impermanent nature such images have not survived. As a result of this it is not possible to trace the sculptural history of the region from the period of the Sangam classics in the light of extant examples.

It is in the early rock-cut caves of the Pallavas that one encounters for the first time the earliest extant specimens of Tamilian stone carving. But even here, with the sole exception of the Lalitankura Pallavesvaragriha at Tiruchiräppalli, the other rock-hewn caves of Mahëndravarman I are devoid of much plastic decoration. Sculptures in these caves mostly consist of a pair of dvărapălas guarding the entrances and almost invariably leaning on their clubs; and a glance at their types in different caves would indicate the successful attempts made by the Pallava craftsmen in perfecting physiognomy. When compared with the simplicity and paucity of sculptures in the Mahendra type of caves, the carving of many figures in the caves and monoliths of Narasimhavarman I and Paramesvaravarman I is striking.

Stylistically Pallava carvings are characterised by a naturalism in pose and an attenuated physiognomy. In fact the human figure is the pivot of Pallava sculptural art. In it the development of lines into contours and the manner in which the contours merge with the lines can be seen. By ignoring smaller anatomical details the Pallava craftsman have idealised and generalised human anatomy. A taller and somewhat oval-shaped face, double chin and flat nose are the major traits. Orna­mental decorations and costumes are kept to the minimum. In a large number of specimens the yajńopavita is carried over the right arm, though this is not a very dependable feature for purposes of dating. When the images represent deities, they are endowed with their weapons and attri­butes, which are either held by them naturally in the hands or placed immediately above them.

The different iconographic types met with among the Pallava images indicate the prevalence of syncretistic trends during the period. The deities shown include Harihara, Ardhanäri, Trimurti-Ekapada, Subrahmanya as Brahma-Sasta, etc., not to mention the different forms of Siva and Visnu. It is somewhat strange that Ganesa makes his debut only in the structural temple of Răjasimha at Kańcipuram, his images at Vallam and Mahabalipuram being probably only later additions. His cult, along with that of the Saptamätrkas appears to be a Calukyan derivative. When the Pallavas were engaged in such a prolific sculptural activity in the Tondaimamandalam and in the regions immediately to the south of it, the Pandyas in the extreme south were busy making their own contributions. Stylistically and in decorative details, in conception and in execution, the Pandya carvings are not much different from the Pallava reliefs and appear to be just another edition of the Pallava art, though in the realm of iconography it seems to strike a slightly different note. The provision of a mrdanga instead of a vina for Vinădhara Daksinamurti in the monolithic excavation at Kalugumalai and the decoration of Visnu with cannavira  - an ornament usually associated with goddesses and minor gods only - at Kunnakkudi are instances in point.

With the revival of the Cola empire under Vijayälaya and the building of numerous structural temples throughout the Tamil country, we enter upon a new era of sculptural activity. In some of the earliest Cola and Pandya temples sculptural decoration is restricted and the value of plain space on walls is appreciated. But this was only a passing phase and the tendency to decorate niches and other parts of the shrine with radiant carvings asserted itself soon.

Sculpture under the Colas is relieved of its architectonic context and it may even be said that it is “subsidiary to architecture “. The style may be characterised as “fluent” thanks to the continuous experience in the art of stone-cutting from about 600 A.D. The striking attenuation of the Pallava period is replaced now by very subtle rhythmic quality and, what is more, Cola sculptures are pleasingly delicate in outline. A flat upper torso, protuberance on the knees and a soft and supple form are some of the notable physiognomical characteristics shared by a large number of Cola specimens. The humanism and freedom of pose are the two significant features that elevate Cola carvings to the status of great art. They are endowed with naturalistic and elaborate treatment of decorative details. While these details are suggested in Pallava images by soft lines which often merge in the modeling, they are in bold and emphatic lines in the carvings of the period of Colas. Mention should be made here of details like katisutra, hara, kanti, etc., which are recognisably more pronounced in early Cola images. Again it is in sculptures of this period that the skandamala (shoulder tassel) appears for the first time.

It is rather difficult to speak of the general characteristics of late Cola sculptures as they display interesting variations in style and decorative details. More than one school of late Cola sculpture appears to have persisted, one preserving the classical traditions of the early period and the rest attempting to conventionalise in varying degrees in anticipation of grotesque stylisation that was to characterise the future. The tendency to elaborate and “improve", a feature noticed in architecture, is extended to the decorative details of the images.

The composition is generally large, and subordinate figures are seen within the niche, unlike the early Cola examples where they flank the devakosta. Prabha - arch - is noticed behind the head of the principal figures in many of the upper tier sculptures in temples. Generally the images are in bold relief, though round ones are not wanting. Figures are shown frontally and profiles are rare. Some of the specimens of the period are poor in depth of conception and formal and weak in their presentation of themes. The torso becomes thick and squatty, unlike its elegant and natural shape in early Cola sculptures. The under-garment is invariably brought down below the knees and it encumbers the effect of the modelling of the limbs.

The decline and fall of the Cola empire had an inevitable impact on the art of the Tamil country, as it amounted to the withdrawal of a powerful patronising agency. However, the rise of the Vijayanagar rulers farther north and their eventual supremacy over TamilNad were welcome phenomena with the resultant encouragement to various arts.

The sculptural art of the Vijayanagar period commands one’s attention not so much for its aesthetic qualities as for its prodigious output and the diverse themes it chose to represent. The sculptures are formal and rigid and lack the naturalness and softness of earlier periods. The pose is stiff and the face becomes expressionless, The nose becomes pointed and the cheeks are vertically grooved. The elaboration of draperies, ornaments and other decorative devices, started during the late Cola period, is continued now with greater vigour. Tilak, the mark on the forehead, not found in early periods, makes its debut in Vijayanagar carvings.

Thematically a very significant introduction of the period is the Ganga-Yamuna motif. This relates to two female figures in bold relief on both sides of the entrance under the gopura of the temple, one of them representing the river goddess Ganga and the other Yamuna, both on their vahanas. From the mouth of the vahana rises a thick plant. It comes round the figure and rises above with involute circles in which are sculptural reliefs of the ten incarnations of Visnu. This practice of representing the river goddesses at the entrances was in vogue in North India from the Gupta period onwards but was introduced in the south only under the Vijayanagara rulers.

Chola Bronzes

This account will not be complete without at least a brief sketch of the art of bronze casting for which the Tamil country was famous. Metal images have been cast throughout the centuries under the patronage of different dynasties in the north, south, east and west, but nowhere does it seem to have registered such an acme of development as it did under the Colas. A few Agamic texts and the contemporary practice among the sthapatis indicate two modes of casting icons - the hollow and the solid methods. The figure of a female, identified as Mother Goddess and discovered at Adichanallur is the oldest extant metal icon in the Tamil country. It is small in size and has been taken to be at least 3,000 years old. A few Buddhist metal icons discovered at Amaravati and Kaverippumpattinam and Buddhapad in Andhradesa and assignable to the early centuries of the Christian era reveal that the metallic art was already flourishing in South India.

These traditions of the early period in the realm of art were continued by the Pallavas who held hegemony over parts of the southern Andhradesa and the whole of the Tondaimandalam and even the region up to the Kaveri in the south. A few scholars tend to believe that the art of bronze casting was either unknown to the Pallavas or at least had not attained great heights under them and that for all practical purposes the history of the art of bronzes in Tamil Nad begins with the Colas. 

Now it is difficult to wholly accept this, as we know for certain that this art was flourishing under the Satavahanas and Ikshvakus and hence the Pallavas might also have been aware of it. Considering the fact that under the early Colas the output of metal icons was prodigious in quantity and unparalleled in quality, it is difficult to assume that this art had developed to that extent within a short time after its introduction in the Tamil country. Of particular interest in this connection is an inscription of a certain Abhimana Siddhi (who seems to be either a contemporary of or a ruler anterior to Dantivarman Pallavamalla) in the Vaikunthapperumal temple at Kanchipuram making reference to the gift of one thousand pon (gold) obviously for the making of a golden plate for offering bali and also for a padimam. The padimam here could only mean an image made out of the gifted gold and hence a metal icon. Apart from some of the Pallava characteristics revealed by a few bronzes, this inscription would show that metal art was not unknown to the Pallavas. A Tripurantaka in a private collection now in Ahmedabad, a Vishapaharana from Kilappudanur in the Tanjore District, a Natarăja from Nallur in the same district are a few of the icons with obvious Pallava features and noted for graceful and simple modelling. A Maitreya from Melaiyur and a Visnu in the Trivandrum Museum may also be included in this list.

A large number of specimens belong to the period of transition from the Pallava to the Cola period and the first two or three decades of the Cola period. This is the period which witnessed the highest water mark in the art of bronze casting, and in the light of recent and penetrating studies it is possible to discern different phases in sequence in the development of the art.

The Visnu from Tiruchcherai, Chandikesvara from Tiruvenkadu, Kirata and Arjuna from Tiruvetkalam are the most representative of the flowering phase of the early Cola period. In modelling treatment they offer valuable links between earlier images and the clearly datable icons of the subsequent period. The skandhamala (shoulder tassel), which is not generally noticed in images of the Pallava period, is invariably seen in Cola bronzes. Similar and interesting changes are found in many ornaments and decorative devices, including armlets, udarabandha, necklaces, katistra, loops and tassels, etc. The shape of the yajńopavita, i.e. its running over the right arm, is continued in a few images but it ceases to be a dependable stylistic feature in the Cola period.

Stylistic characteristics, useful as they are for any chronological classification of images, are not always useful and at times even prove to be deceptive on account of the persistence of certain modes for quite a long time. It is in this connection that a few inscriptions prove to be useful, affording exact dates in which the icons were cast and endowed and thereby enabling one to study the features of dated bronzes and compare them with images with identical features to arrive at their probable dates. Quite a number of well-known and masterly examples of the Cola bronzes have now been dated with as much accuracy as possible. 

Numerous are the inscriptions making mention of the dedication of bronze images to temples under successive rulers, but unfortunately not all of them have survived. Special mention must be made of the references to a host of deities in metal in the Tanjore inscriptions of Rajaraja of which none, with possibly the single exception of a Tripurantaka, is still extant. It is difficult to exaggerate the importance of a few dated bronzes, as they indicate art forms and norms of the periods of their making and show how stylistic features are not always wholly trustworthy. 

Of the many superlative icons in the Umämahesvara temple at Konerirajapuram, Tanjore District, a Vrshabhavahana Tripurantaka and Ganapati are datable between 959 A.D. and 977 A.D. on the authority of an inscription in the same temple. The reference to gifts to an image of Kuttaperumal and his consort in the Vriddhagirisvara temple at Vriddhachalam by Sembiyan Mahädëvi in an inscription in that temple reveals that the now extant images of Natăraja and his consort there were made in or before 981 A.D.

Of the many bronzes unearthed at Tiruvenkädu in the Tanjore district a Vrshabhavähana was dedicated in 1011 A.D. and his consort in 1012 A.D. and the characteristics of these succinctly illustrate the bronze style during the last years of Răjaräja I. An inscription of the same ruler dated in his 28th year refers to gifts to an image of Adavallăn (Natarăja) which may be a reference to either of the two figures of the god in the temple; obviously this was dedicated in 1013 A.D. Another epigraph speaks of the dedication of Bhikshatana in the 30th year of Rajadhiraja I corresponding to 1048 A.D. while yet another inscription reveals that an Ardhanărisvara was endowed in or before 1047 A.D. That other undated Tiruvenkadu bronzes like the Bhairava and Kalyanasundara should also belong to about the same period is apparent. All these icons admirably reflect the heavy and stolid forms of contemporary stone sculpture.

Generally speaking, the bronze icons reflect the form and style of contemporary sculpture in stone. This is true also of iconography and decorative details. The Nataraja image which is rare in Pallava times (found only in the Siyamangalam cave, Dharmaraja ratha at Mahabalipuram, Kailasanatha, Muktisvara and Matangesvara temples at Kanchipuram) is frequently represented in the Cola period. It is in the beginning of the early Cola period that the Anandatandava mode of dance gets crystallized and is shown alike in stone and metal. In the representation of this and other themes and in general execution and details, minor albeit interesting variations are found between the specimens wrought in the metropolitan art centres in the Cholamandalam and the products in the other peripheral regions like the Pandya and Kongu countries. The reversed posture of Nataraja in the Pandya realm, known as marukal tandavam, is particularly interesting.

The metal art was zealously patronised during the later Cola and Vijayanagar periods as well; but examples of these periods, like contemporary stone carvings, are devoid of life. They are much conventionalised and the dynamic and rhythmic movement characteristic of early examples is now replaced by mathematical schematism.


C. L. Fabri, M. S. Kesavan, N. Vanamamalai.

Fabri questioned the view that “the advent of stone for purposes of art does not seem to antedate the beginning of the seventh century”, since at Nagarjunakonda and other Buddhist cities stone had been used in the second century B.C. Moreover, there were Buddhist bronzes in Nagarjunakonda and other places. At Nagapattinam half-finished statues have been found, and it would not be incorrect to suggest that Nagapattinam had exported vast numbers of statues even to the northern cities - the Budhhist cities.

Mahalingam: At the point referred to I was discussing the date of the first Pallava monument in stone. It is true that in eastern Andhradesa and in northern and western India we have rock-cut structures, but they were Buddhist and Jam in origin. So far as Hindu architecture is con­cerned, we find that stone seems to have been used for the first time only during the days of Mahendravarman. Therefore there is a time gap of nearly eight centuries, or even more. However that may be, the fact remains that cut stone was used for Hindu monuments relatively late in the Tamil country.

Kesavan raised the question of the figures of Nataraja described by the speaker and the respective forms they took in Pallava and Cola times. 

Mahalingam replied that there were stone figures of Nataraja in Pallava times (such as those at Mahabalipuram and Kanchipuram), but that bronze statues did not appear until later. Vanamamalai asked a number of questions about the continuity of certain trends in art and sculpture in South India and about possible outside influences.

Vannamalai: Can we trace any continuity in art and sculpture from the pre-historic period to the times of the Pallavas and Colas?

Mahalingam: This question concerns the relationship between the art of South India from the seventh century onwards and earlier art as revealed by the Adichanallur and Arikamedu findings. It is too soon to say anything about continuity, because we do not yet know much about Adichanallur and the findings there. From Arikamedu we have only terracotta items. The art of the Pallava period consists of stone sculpture and icons. Hence it is rather difficult to say much about the continuity.

Vannamalai: What is the relationship between the early Pandyan sculptures such as Kazhugumalai and early Pallava sculptures. Is there any intermingling?

Mahalingam: There is not much difference between the art of Kazhugumalai and the Pallava art of Mahabalipuram. But in some respects, with regard to certain sculptures and motifs found in these architectural monuments, we find some extraneous influences (such as the Chalukya influence in the Pandiyan country), which we do not find earlier.

Vannamalai: Can traces of Buddhism and Jainism be noticed in our art before the seventh century A.D.?

Mahalingam: With regard to the period before the time of the Pallavas, there are no extant monuments. We have literature only, and from the nature of this evidence it is not possible to draw any conclusions.

Vannamalai: Are there any traces of Greek or Roman influence in the early art and architecture?

Mahalingam: There is some Greek and Roman influence on early art. In a few places at Nagarjunakonda and Amaravati we find some sculptures, the characteristics of which are not indigenous.

Mail Us Copyright 1998/2007 All Rights Reserved Home