Arts in ancient Bengal

Ancient Bengal developed not only in literature, but also in the various arts. In this page, a few of these achievements are discussed.

Folk Art

The folk traditions of painting floors with rice paste for special occasions has been noted in naiSadhacarita. However, it is difficult to find details of mud dolls, the manasA and gAjI paintings, the colourful pictures and drawings on mud and vase, the needlework on the blankets, the arrangement of hanging storage, and bamboo work in ancient Bengal ... one can only imagine that they existed even then.

Music and Dance

The tribal songs and dances, and those of the women of lower classes, the dance of the rAYabe~ze and lAThiYAls; as well as the bAula, bhATiYAla, and jhumur songs point to a folk culture of music about which little is known.

The proper history of music starts only in the 10th century, when the caryAgItis were supposed to sung in rAgas. The most important rAga seems to be paTamaJjarI. gavaD.A/gauD.A could be a gauD.I tradition, and locana mentions a gaurI rAga. It may be a mixture of this and mAlasI (could be mAlavazrI) that was called mAlasI-gavuD.A. Other such local rAgas could be gurjarI/guJjarI and its variation kAhNa-gurjarI, and vaGgAla. zavarI could be of tribal origin, and dezAkha (which may have given rise to desha) suggests a fold origin as well. rAmakrI probably gave rise to rAmakeli, dhanasI/dhAnazrI to dhAnuSI, mallArI to mallAra. Of the rest, kAmoda, bhairavI, and valADDI/varAD.I are known from later times, but aru and devakrI are difficult to fathom. The idea of a sthAYI (called dhruva) and antarA also goes back to this period.

Similarly, gItagovinda mentions rAgas and tAlas: mAlava (which may be related to the earlier mAlasI) rAga with rUpaka and yati tAlas; gurjarI with niHsAra, yati, and eka; vasanta, rAmakirI, karNATa, and bhairavI with yati; dezAga(/kha) and vibhASa with eka; deza-varAD.I with rUpaka and aSTa; and varAD.I and goNDakirI (which one imagines came for goNDakrI, and may be of tribal origin) with rUpaka. In a quotation in gurugranthsAhib, we also find jaYadevas songs in gujarI/gurjarI and mArU rAgas. The southern influence in Bengali music is also perceptible in this period.

Even though the tumburunATaka is not available any more, we do know that it talked about singing dezAkha, lalita, and paTamaJjarI every morning in the zukla pakSa preceding durga pujA, that it believed that tiny variations in rAgas and tAlas made for an essentially unlimited number of those, and it did not believe in fixed times for the various rAgas. locana talks about the notes in use, and he divides the octave into 22 zrutis and 7 notes. He allows for sharp variants of gAndhAra, madhyama, and niSAdha, and flat versions of RSabha and dhaivata. He, however, seems to use a sharp dhaivata in his pUravA. The tAlas he talks about, for example, caJcatpuTa and cAcapuTa, are not known elsewhere. He claims (with later interpolation since rAgas like iman and phirdost are not known before Amir khusrau at the end of 13th century) that there are twelve generating rAgas: bhiravI that generated 2 rAgas, gaurI 27, karNATa 20, kedAra 13, iman 4, sAraGga 5, megha 10, dhanAzrI 2, ToD.I 1, pUrvA 1, mukhArI 1, and dIpaka 1, for a total of 86 rAgas. Some disagreement is apparent in locana's times itself: thus locana used pure tones in bhairavI, but others use the flat dhaivata. Similar disagreement about the proper times for the various rAgas also arose.

In the medieval period baD.u caNDIdAsa in his zrIkRSNakIrtana used the following rAgas: koD.A, koD.A-dezAga, varAD.I, deza-varAD.I, kakU-gujjarI, vibhASa, vibhASa-kakU, vaGgAla, vaGgAla-varAD.I, gujjarI, pAhAD.iYA, dezAga, Ahera, rAmagiri, dhAnuSI, mAlava, velavalI, kedAra, mallAra, bhATiYAlI, lalita, mAhAraThA, zaurI, vasanta, bhairavI, zrI, sindhoD.A, and paThamaJjarI; and the tAlas included yati, eka, aSTa, rUpaka, aDhukka, laghuzekhara, and krID.A.

See also the page on culture for information on musical instruments and dance.


The preserved portion of ancient Bengali sculpture is mainly of the stone variety: but this was done only for the big temples. However, most of the art was probably made of mud, occasionally fired. Even much of the temple decoration was probably made of mud. Metal examples haven't survived either; most have been melted down for their value. In fact, little survives from the period before sixth-seventh century AD. Large scale stone carving in India can be traced back merely to the maurya period, but it probably was not widespread in Bengal for a few hundred years after; metal sculpting probably started no earlier than a couple of centuries AD. The sculptures found prior to the gupta period are all portable, mainly of terracotta and small stone, do not show much regional peculiarities, and could have been produced almost anywhere in north India. The bust of maurya and zuGga period yakSiNIs found in tAmralipta and candraketugaDh.a are vitually identical to those found at pATalIputra. The explicit sexual depiction on terracotta plaques coupled to portrayal of grains and fish probably attests to their symbolic attachment to generation. The dress sometimes reflects a gAndhAra style, and the human forms and footwear recalls the kuzAna and graeco-roman styles. Some buddhist jAtaka stories also found in these first=second century plaques.

The period starting between the 2nd century BC and 3rd century AD marked the emergence of a distinctive style of sculpture in north India. The artefacts fall into two major classes: (i) depictions of youth, especially the female form (and the child), often with foreign facial characteristics, hair styles, and dress; and (ii) plaques depicting stories. In Bengal, the plaques (often with holes on top probably to hang from walls) found at pokharNA (in bÃkurA), tamaluka, maYanAmatI-lAlamAI, candraketugaDh.a, and mahAsthAnagaDh.a concentrate on the first of these. The ornamentation is heavy and unusual, the hair is long and worn in a variety of styles, clothes are coarse but expected for the shunga and later period, and the sexual representation is exaggerated. They were probably used to decorate the houses in the newly emerging trading cities in Bengal; in style they were close to the plaques found at other places in north India (e.g. kauzAmbI, pATalIputra, basAra) around 1st century BC - 2 century AD. A few stone sculptures from the kusAna period has also been found. Most of these are from northern bengal, and are not in the mathura style of red sandstone. Two sun idols from niYamatapura (rAjazAhI) and a viSNu idol from hÃkarAila (mAldaha) are very similar to each other; and show clear influence of kuSAna art, though they are almost certainly local. A few terracotta plaques showing women in typical in finds in mathurA (though somewhat coarser portrayal) have been found in vANagaD.a

The gupta style, from its centre in sAranAtha had influenced almost all of northern India. A sandstone buddha from vihAraila is clearly in this style, though ever so slightly coarser. The sun idol found in deoA (district baguD.A) is an example of the strong but comparatively unadorned and, yet, beautiful eastern gupta style. The same can be said for the sun idol found in sundaravana-kAzIpura and the gold-plated bronze maJjuzrI from the valAidhApa stUpa. The sandstone cakrapruSa from mAlAra (district murzidAbAda) is also an example, though the craftsmanship is not as nice. Overall, the puNDravardhana (north bengal) area shows the greatest number of these artefacts.

So also in Bengal, this classical period came to an end starting around the sixth century AD and a new medieval style of rigid postures found maturity in the pala period. Early examples of this style is found in the aSTadhAtu sarvANI from deulabAD.I (with inscriptions by prabhAvatI, wife of; in the small sun idol sitting on a chariot drawn by seven horses found in the same place; and in the bronze standing shiva found at maNirahATa (district 24 parganAs).

The ruins at pAhAD.apura of a buddhist vihAra built during the reign of dharmapAla, provide examples of both the old gupta style and this new style. The former include a fine, clearly eastern gupta, and a much more coarser sculping styles, the former probably reflecting reuse of material from an older period. The topics focussed on the traditional brahminic portrayals of idols, including a couple, possibly a rAdhA-kRSNa form (though more likely, the lady should be identified as rukmiNI or satyabhAmA), and idols of yamunA, ziva, and balarAma. In contrast, the new style portrayed people in everyday situations: for example, dancing girls, men and women having sex, and a tired guardsman standing resting on his stick. The portrayals are strong but pretty stiff, the clothes and ornamentation is light, and full pleasures of simple life are evident in them. The same observations holds generally for the ruins of both pAhAD.apura and maYanAmatI. The topics covered stories from rAmAYaNa and mahAbhArata, of kRSNa, from paJcatantra and vRhatkathA, and, drawn directly from life. Mythological characters like gandharva, kinnaI, and human-animal chimeras abound; as do clear portrayal of the animal world around us, sculptures of mother and child, wrestlers, guards, women drawing and carrying water, women returning home, fighters male and female, archers on chariots, bearded mendicants and beggars, farmers, women cooking fish, singing and dancing women, hunters, men playing musical instruments, brahmins, traveling mendicants, many incidents from stories, and cock- and bull-fights. Some portray gods like brahmA, viSNu, gaNeza, bodhisattva padmapANi, maJjuzrI, tArA, and mainly the very bengali ziva.

The sena period saw the rise of conspicuous consumption, and the sculpture reflects the same. The patrons of the various temples were royalty or rich, though not necessarily highly ranked, people in society. It is obvious that the artisans were common people, and probably not very respected in society: in fact, very few names of the craftsmen were mentioned. When it comes to stone and metal working, we do find a few names like dhImAna and his son viTapAla, bhogaTa's grandson zubhaTa's son tAtaTa, resident of satsamataTa zubhdAsa's son maMkadAsa, vimaladAsa, sutradhara viSNubhadra, vikramAditya's son mahIdhara, mahIdharadeva's son zazIdeva, karNabhadra, tathAgatasAgara, dharma's great grandson manadAsa's grandson vRhaspati's son rANaka zUlapANi. It is very likely that some of the people who carved the royal inscriptions were also sculptors.

Other than terracotta, the medium of expression was usually stone, brass, or aSTadhAtu; though a few wooden sculptures are also known. The sculptures are mostly attached to a back panel, and idols of gods have an aura behind their head. A very few stone sculptures, mainly of gods like kRSNa-balarAma, indra, yama, kuvera, gaNeza, but a few dancing women as well, are also found at pAhAD.apura. Little has been found otherwise from the eighth century: two idols of goddesses from barddhamAna barAkar, one idol from mAnbhUma borAma, and one viSNu idol from dinAjapura kAkadIghi are probably from this period. A viSNu idol found in bAghAurA in tripura from the third year of mahIpAla's reign, a ganeza idol from his fourth year, a viSNu and a sUrya idol from govindracandra's reign, a sadaziva idol from gopAla III's reign, and a caNDI idol found in DalabAjAr in DhAkA from lakSmaNasena's rule are the other firmly dated examples known from ancient bengal. There are also examples from stone and metal sculptures found in vihAra dating to the reigns of devapala, zUrapAla, nArAYaNapAla, and mahendrapala. The ninth century examples show a certain stiffness in style and lack of vitality: strong lines and soft muscular interiors characterize this period; they are usually translations of stone sculptural styles into metal. In the tenth century, exemplified by the RSabhanAtha found at surohara in dinAjapura, the buddha found at ujAnI in pharidapura, and varAhavatAra found in at silimapura in baguD.A, show an increased strength and beauty within the same strong lines. This period was probably the height of bengali idol artistry. In the 11th century idols, for example the viSNu from mahIpAla III's time mentioned before, the umA-mahezvara from agradiguNa, the navagraha from kaGkanadIghi in sundaravana, and vINAvAdinI sarasvatI from sundaravana, one finds a tendency towards slimmer figures with much more ornamentation and additional figures, with much more emphasis on beauty. Twelfth century, when the society found itself inspired by sena ideals, found the vitality ebb away, this time from the weight of ornamentation and appeal to earthly pleasures. The metalworking examples from jheoArI (11–12 century) are clearly similar to contemporary ArAkAnI styles.

Some examples of ancient bengali sculpture (this will be reorganized someday):


The flood plains in which Bengal is situated makes it much easier to make buildings out of bamboo, wood, reed, straw, leaf, mud and brick rather than from stone: this also has led to the non-preservation of most dwellings. For household dwellings, brick itself was probably rare: from depictions one finds the typical mud and bamboo fencing houses on a square plan with poles of bamboo or wood and thatched in bow shaped two four or eight layers. In fact, this style is still very common in bengali villages, and it gave rise to the later gauD.IYa or bAGlA riti in construction, and ultimately to the word ‘bungalow’ in anglo-Indian usage. Settlements even in the ancient period were beautified with temples, stupas, vihAras, parks, lakes, playgrounds, forests, flowers, creepers, trees and plants.


More permanent constructions were the religious houses, but few temples have survived. At least one jaina vihAra and stupa in pAhAD.apura and many buddhist ones are known to have existed. Reference to them abounds in writings of travelers like Fa Hien (5th century) and Hieuen Tsang (7th century), in contemporary pictures and texts. However, few of these exist today: apart from the ravages of weather, thieves, and forests, religious intolerance in later periods have sometimes led to destruction of temples and creation of mosques over them.

There were mainly four types of temples in ancient bengal:

bhadra or pIra deula

The roof of the central portion goes up in three, five, or seven sloping pyramidal steps with a crown at the top. This is related to the oriYa jagamohana or bhogamaNDapa construction. This was perhaps the most common style as exemplified by depictions such as on the bronze plaque from Asrafapura. The floor plan was usually square augmented as in a chariot. The roof style was in use also in the medieval period, and examples from the 16th or 17th century survive (as in the nadImaNDapa of ekteshvar temple in bÃkurA). Depictions from the ancient period are also found in the kalYANasundara shiva plaque from hili, sUrya plaques from kuladiYA (24 parganas) and bariYA (rAjashAhI), ratnasambhava plaque from vikramapura, buddha plaque from madhyapAD.A (DhAkA), umAmaheshvara plaque from birola, and a depiction in kumArapura (rAjashAhI).

rekha or shikhara deula

The roof of the central portion goes up slightly curved into a point topped by the crown. This is related to the north Indian and oriYA nAgara style, exemplified by the zatrughnezvara, parazurAmezvara, and muktezvara temples at bhuvanezvara. However, the bengali temples were not as elaborated with a jagamohana or bhogamaNDapa, they just had the central garbhagRha. Similarly the decoration in the liGgarAja and later oriYA temples has no counterpart, the bengali temples at most had little caitya like windows on the top.

Some examples of this form survive till today. The 4th temple at barAkara (varddhamAna), a stone temple from the eighth century, is an early example of this flowing style. Two stone ‘dedication’ temples at dinAjapura and nimadIghi (rAjazAh) and a bronze one at jheoArI (caTTagrAma) are examples of a later stage of evolution, and brick temples at deuliA (varddhamAna), at bahulAD.a (bÃkuD.A) for siddhezvara, at dehAra (bÃkuD.A) for sarvezvara and sallezvara, and the original jaTA' deula in sundaravaNa are examples of a still later stage; probably from th1 10th or 11th century. The 1st, 2nd, and 3rd temples in barAkara (varddhamAna) and ichAighoSa's deula at gaurGgapura are probably from the later medieval period (14–15 century).

bhadra or pIra deula with stupa

This one has a stupa just below the crown. No examples of this exist now, and none have actually been found. We know of a lokanAtha temple at nAlendra from a painting in a manuscript, and this style likely was from the tradition which led to the abhaYadAna and pATothyAmA temples at pAgAna (Myanmar).

bhadra or pIra deula with spire

This one has a point just below the crown. No examples of this exist anymore, but from illustrations in manuscripts (a buddha temple at puNDravarddhana) and carvings, it does not seem to have been particularly rare.

All these styles are also seen in a larger cultural context including Myanmar, Java, and bAli. Neither is the list above exhaustive: the temple ruins at vaigrAma (dinAjapura) can possibly be identified with the ziva-nandi temple mentioned in 448–9 AD during the gupta period; and could have been the typical flat-roofed central site surrounded by a walkway, typical of this period.

A number of temples are found around the ancient site of mahAsthAna, as at vairAgI-bhiTA, govinda-bhiTA, and gokula pallI; but no temple city like bhuvanezvar, puri, or konAraka (Orissa), khajurAho, pAgAna (Myanmar), prAmbAnAm-pAnAtaram (Java), aGkora-thAma (kAmboja), or kAJcIpura is found. Most temples were probably small in size, but there were exceptions. The greatest example of temple architecture in ancient bengal was probably the burnt mud brick temple built on a mud foundation at pAhAD.apura, either converted from a previous jaina temple or built probably under dharmapAla. The plan was 356.5' in the north-south direction and 314.25' in the east-west direction. Each side was drawn out three times to give it many corners. There was a huge hollow square tower at the center, rising up to three or more levels including the courtyard level (this sarvatobhadra style is supposed to have sixteen corners, entrance on each side, and five levels, but no other example survives in India.) Its main entrance was to the north, and a stair took one from the courtyard to the first level. Crossing the main open space one reached the walkway surrounding the tower at this level, and surrounded by the temple walls. From any side, one could go up to the second level walkway. This level housed the four major garbhagRhas. The outside was decorated with terracotta plaques, similar to those found at gokula and govindabhiTA at mahAsthAna. Stylistically, this temple probably belongs to the tradition leading to temples like thATaviJu (or sarvaJja), zoYegu-jyi, Tih-lo-minh-lo in pAgAna (Myanmar) and loro-joMrAM and ziva atg prAmbAnAma (Java).


The concept of a stupa goes back to the vedic ritual of building an earthen mound on the remains of a person, but the buddhists took it over to save the relics of buddha. These were originally supposed to be a hemispherical structure on a circular base on the top of which a platform held the relcs covered by an umbrella. This concept got elaborated to the point that the main features got almost dominated by the lotus-like base and the huge ornamentation on top. Though Hieuen Tsang (7th century) mentions buddhist stupas in Bengal dedicated by the maurya king azoka, there is no other evidence for such antiquity. The oldest known evidence of stupa in bengal are the ruins at bharatapura (varddhamAna), probably from the ninth century. We do find little bronze dedication stUpas (for example, from c. 7th century Asraphapura in DhAkA, and later pAhAD.apura in rAjazAhI and jheoYArI in caTTagrAma), and though these don't count as architecture, they provide evidence of the style. Same goes for the paintings of mRgasthApana stUpa, ‘tulakSetre varddhamAna stUpa’, and another unknown one. In yogI-guphA, we find a real stone example of this elaborate ornate style. We also find ruins (for example, at pAhAD.apura in rAjazAhI and vahulAD.A in bÃkuD.A) from which the floor plan can be discerned: a central square flanked by small squares to form a cross. The inside was stacked with mud tablets with buddhist writings: symbolic of the religious body that was more deserving of preservation than the physical one. In short, little innovation is noticeable in bengali stUpa construction, for that one needs to look at the huge examples from the burmese capital, pAgAna.


The vihAras where the monks lived grew out of cave groups, and maintain the concept of a central courtyard surrounded by rooms, some of which (often the central ones on each side) bigger than others. A well in one corner and a single entrance completed the construction: sometimes it was multistoried. There were many buddhist vihAras in bengal, but all of them are now in ruins.

Near somapUra, for example, in bat gohAli (or goYAl bhiTA) a jaina vihAra was established around 478 under AcARya guhanandi, and the major dharmapAla's mahAvihAra was constructed there about the middle to end of the eighth century. The jaina vihAra is totally non-existant, but the plan of the buddhist one can de deciphered. It was an almost 900 feet square surrounded by walls and housing about 180 upper floor rooms surrounded by a verandah. Downstairs in the center was a vast courtyard with the temple in the center. The density of columns indicates that it may have been multistoried. The vihAra was entered from the north through a stairway which lead through a wide door into a large room with pillars, and then through a smaller door into a similar but smaller room and then to the verandah. There was a secondary entrance on the north east side, and maybe through the central room on the east side. The administrative office was next to the central entrance. The drainage system went from room to room and finally into a sewage pond. The temples, stupas, wells, bathrooms, and dining halls were spaced among the rooms and in the courtyard.

To be finished

Fine arts

Even though the traveler Hieuen Tsang mentions painting in Bengal around 738 AD, no actual examples of painting from before around the 10th AD has survived. No doubt, however, that the practice of painting on the ground, on cloth, and on the walls of temples has a long tradition ... it is that we know nothing about what forms it took in those ancient times. From the pala period, however, we find a number of examples of paintings on palm leaves which have survived till date: these were meant as illustrations for books. Though it is not clear whether all the paintings were done in Bengal, in this period, little difference can be found in the Bihari, Bengali, and Nepali styles: so we can discuss all of them together. The texts include eight aSTasAhasrikA prajJApAramitA's (from the fifth and sixth years of the reign of mahIpAladeva, from 1071 AD, from the thirtyninth year of rAmapAla, from the reign of harivarmA, from probably gopAladeva III's reign, from 1148 AD, and an undated 12th century one), kAraNDavyUha (probably 12th cent), gaNDavyUha (probably pAla period), bodhicaryAvatAra (probably 12th cent), paJcarakSA (14th year of narapAla) and few isolated ones (from the pAla period including one from 1015 AD, one from the fourth year of probably gopAladeva III, and one from the 18th year of govindapAla). In addition, there is a shaiva treatise where the pictures are on thin wooden planks, again probably from the pAla period, a treatise written and drawn on paper, and a few pictures scratched on metals. Most of the pictures are of buddhist gods and goddesses, but a few are of brAhminical gods like ziva (or his symbol the male generative organ). Even though these pictures were drawn to decorate accompanying texts, they are more ornamentations than thematic. Moreover, they are not traditional miniatures: the drawing style more closely resembles the style of wall paintings.

The colours used include yellow (haritAla), white (khaD.imATi), deep blue, black (pradIpa-zISa), red (si~dura), and green. The blue and green are different from those found at ajantA, the green here is probably made from blue and yellow. The saturation of the colour is controlled, and the color style is basically the same as that at ajantA, though of poorer quality. A sketch in thin red or black thin brush of varying thickness to emphasize features is filled with colours with a thicker one. The central image is bigger than the surrounding ones, and stands on an ornamented background or semicircular regions. The canvas shows little empty space, what is not taken up by the rows or circles of smaller images is filled with flying or walking figures, foliage, and plain ornamentation. The main feature of this period of Eastern India is the smoothly flowing lines: pure linear sketches are found in some copper plaques from the 11th century. Even in the same manuscript one often finds differences: in some, the portrayal is often stiff, and figures soft but emotionless; in others one finds expression through solid colours; and yet others are very lightly coloured.

Up to history of ancient Bengal

Valid HTML 4.0! Valid CSS!