A fragile legacy


back to issue

IN 1999, Manager Rajdeo Singh, the ASI chief of conservation and head of science at Aurangabad, began work on the restoration of the murals in Caves Nine and Ten at Ajanta.

Manager Singh, as he is always known, had been in charge of conserving the murals of Ajanta for a number of years, but the work in Caves Nine and Ten was, he knew, especially difficult, and of the greatest importance. This was partly because these two caves contain the most severely damaged of all the Ajanta frescoes: ‘The paintings were so fragile that in some places there was a great fear even to touch them with the hand,’ he wrote later. ‘At some places the pigment was found completely detatched from the ground plaster and stone surface.’

But largely Manager Singh was concerned because the murals in those two caves are recognized to be not only the oldest images at the site, but the oldest Buddhist paintings in existence – dating from only 300 years after the death of the Buddha. These masterworks of early Buddhist art are, in other words, the prototypes of the forms which would later spread with Buddhism over the Himalayas to Afghanistan, China and Japan, or by sea to Sri Lanka, Burma, Thailand, Cambodia, and the rest of South East Asia.

More remarkable still, with the exception of a few prehistorical pictograms of stick men and animals left by paleolithic hunters at Bimbedkar in the wilds of Madhya Pradesh, they are also the oldest pictures of India and of Indian people to have survived from the ancient world.

The work took over a decade and proved to be even more difficult than guessed. Early British art historians who had worked on copying the murals between 1844 and 1885 had coated the murals with layers of varnish to bring out the colours, and they left the varnish in place after their work was finished, leaving a thick layer of discoloured glaze intermixed with soot and dirt. Moreover, these earliest murals were not only more fragmentary, they were also considerably more smoke and incense-blackened in antiquity than the relatively pristine later murals elsewhere in the site, and perhaps for this reason seemed, blackboard-like, to invite the attention of early graffiti artists and tourists who wanted to leave an inscribed record of their visit. By the time the Nizam of Hyderabad had sent the leading art historian of his state, Ghulam Yazdani, to produce the first photographic survey of the murals in the late 1920s, the murals of Caves Nine and Ten had already been irreparably damaged.


At the same time as the Nizam despatched Yazdani to study the murals, he also sent two Italian conservationists to help restore them. Unfortunately their efforts only obscured the murals further: they coated the pigments with a thick layer of unbleached shellac which sat on top of at least two existing Victorian layers of varnish. The shellac attracted grime, dust and dried bat dung and quickly oxidised to a dark reddish brown which totally obscured the images from both travellers and scholars. Less than a century after being rediscovered by a British shooting party in 1819, the figures of Caves Nine and Ten had been lost again. For the entire length of the 20th century they remained effectively hidden, invisible to the naked eye, forgotten by all.

However, a slow and painstaking restoration of the paintings by Manager Singh from 1999 onwards using infrared light, micro-emulsion and cutting edge Japanese conservation technology, succeeded in removing 75% of the layers of shellac, hard soot and grime from ten square metres of the murals. ‘Particular care and precautions were taken not to alter even a grain of pigment,’ he wrote.

Manager Singh’s remarkable work revealed for the first time since the 1920s the extraordinary images which lay beneath and are now on open display. I happened to stumble across them on a visit to the caves in March this year. The ASI does not have much of a tradition of PR work or public outreach, and even internally there is perhaps little recognition of what Manager Singh has actually achieved and uncovered. For his work is nothing short of a revelation: peeling off the successive layers of shellac, varnish, and dirt and bat dung, Manager Singh has uncovered the oldest paintings of Indian faces in existence, and restored to something approaching their former glory, murals which predate by perhaps six hundred years the better known work elsewhere on the site.


More exciting still, this earliest phase of work is not just very old, but very fine indeed and painted in a quite different style, and using markedly different techniques, to that used in the rest of Ajanta. The murals of Caves Nine and Ten, reproduced here for the first time, represent nothing less than the birth of Indian painting.

Anywhere else in the world a rediscovery of this importance would be the subject of nationwide head-lines, TV documentaries and triumphant exhibitions; but in India the remarkable work of Manager Singh has gone virtually unnoticed.

The King of Varanasi draws his bow to shoot a deer. From the Shyama Jataka – Ajanta cave 10 mural, 1st century BC.

Ever since the work of James Ferguson – the pioneering archaeologist, art historian and early scholar of Indian Buddhism – in the 1840s, scholars working at Ajanta have recognized that there were actually two quite distinct phases of work at the caves, separated by as much as six hundred years: the same gap as separates the Lodhi Garden tombs from the tower blocks of Gurgaon.

The first cave to be discovered, which Ferguson named Cave Ten, lay in the centre of the cliff face, and along with the five others flanking it, were dateable by inscription to the first or even early second century BC.


The great majority of the Ajanta caves, and almost all the murals, particularly the rich cycles of wall painting in Caves One and Two, date from the second phase of construction. This later phase coincided the height of India’s Golden Age, when in the Gangetic plain, the fifth century Gupta dynasty was filling their capital of Kannauj with some of the greatest masterpieces of Indian sculpture, and when Kalidasa was writing his plays. At this time too the philosophers Asanga and Vasubandu were producing their greatest works, the most profound ideas of Vedanta were being compiled, and the Buddhist university library of Nalanda, then the most admired repository of knowledge East of Alexandria, was at its scholarly apex, its wisdom and learning sought by scholars and pilgrims from across the East.


It was also, apparently, a time of great achievement in painting, almost all of which is now lost. Under the patronage of the Gupta’s southern rival, the Vatakata king Harisena (c475-500 CE) ruler of the Deccan, Ajanta reached the peak of its fame. According to the 7th century Chinese traveller-scholar, Xuanzang, Ajanta’s ‘lofty halls and deep chambers painted with great fineness, and stretching through the rock, storey above storey, up the scarped precipice through tiers of peaks and sheer summits’ was not just a centre of great art and craftsmanship: it was also a major centre of learning and was said to be the home of Dignaga (c480-c540), the celebrated Buddhist philosopher and polymath.

This Aquinas of Ajanta lived here in the 5th century composing books which subtly formalized Buddhist traditions of logic and inferential reasoning. The most famous of these, the Pramanasamuccaya or Compendium of Valid Means of Knowledge, came to occupy a crucial place in later Indian and Tibetan philosophical analysis. Ajanta was then probably home to several hundred teachers and pupils – a number that no doubt increased during the monsoon when itinerant pupils would return here for a rainy season filled with theological debate.

Archer guard attending the King of Varanasi watches the king as he prepares to shoot. From the Shyama Jataka – Ajanta cave 10 mural, 1st century BC.

It was at this period that most of the greatest masterpieces of Ajanta were painted. Unlike the flatter and more stylized art of much later Indian miniature painting, here the artists used chiaroscuro and foreshortening, and produced images that embraced at once both the profoundly spiritual and the strikingly sensuous. There are no panels or boundaries in the Ajanta paintings beyond the physical borders of the cave, and the artists likewise move from the world of the ascetic’s cave to the pleasure gardens of the royal court and back again without recognizing any essential separation between the two. The artists of Ajanta clearly saw nothing odd in this juxtaposition of monk and dancing girl.


For in Ajanta the Buddha tends to be shown not just in his monastic milieu, after his Enlightenment, but in the princely environment in which he grew up. Here among handsome princes and bare-chested nobles, princesses with tiaras of jasmine and raat-ki-rani, languish lovelorn on swings and couches, as narrow-waisted dancing girls of extraordinary sensuousness, dressed only in their jewels and girdles, perform beside lotus ponds, swaying to unheard music.

Nearby are painted very different images of stark ascetic renunciation – a shaven-headed orange-robed monk lost in meditation, a hermit seeking salvation in the gloom of a rock-cut grotto, or a group of wizened devotees straining to hear the words of their teacher. Dominating everything are the famous portraits of Bodhisattvas of otherworldly beauty, elegance and compassion, eyes half-closed, inward-looking, weightlessly swaying on the threshold of Enlightenment, caught in what Stella Kramrisch described wonderfully, as ‘a gale of stillness’. Even today, the colours of these astonishing murals glow with a brilliant intensity: topaz-dark, lizard-green, lotus-blue. Indeed such was the celebrity of these 5th century masterworks that most scholars, and almost all modern accounts of the Ajanta caves, have more or less ignored the earlier 1st-2nd century BC caves and their picture cycles.


Because of the dirt covering them, to this day the murals of Caves Nine and Ten have still attracted only passing scholarly attention and, remarkably for so famous a site – Ajanta is after all one of a handful of World Heritage Sites in India, attracting 5000 visitors a day in season and is the subject of many a PhD as well as the six volumes (so far) of Walter Spink – their wall paintings have never before been properly photographed. It is true that Ghulam Yazdani included four almost totally illegible shellac-obscured black and white shots of the Cave Ten cycles in his exhaustive four volume work, Ajanta, published in 1930; but up to now the early mural cycles have never before been published in colour, nor have any close-ups appeared in print before the illustrations that appear alongside this article. The true beauty and importance of these frescoes have therefore been almost completely missed.

The extreme antiquity of Caves Nine and Ten is not disputed, though their exact dates remain elusive: from the paleographic evidence they can be securely dated to between the second century BC and the first century CE, although most scholars – notably Vidya Dehejia and Walter Spink – believe that between 90-70 BC is the most likely period of construction. This makes Caves Nine and Ten – ribbed chaitya halls lined with tapering octagonal columns, ending in a rounded apse which encloses the perfect dome of a tall stone stupa – some of the oldest rooms in the world. These halls were already two hundred years old when Augustus started rebuilding Rome, or when Jesus of Nazareth led his first disciples into the fastness of the Judean hills.


Caves Nine and Ten were excavated shortly after the collapse of the Mauryan Empire and are therefore roughly contemporary with the spectacular early Buddhist stupas and gateways at Sanchi and Bharhut, and a little before those of Kanaganahalli and Amaravati. The closest monuments to the early caves of Ajanta are the other rock cut chaitya halls of western India in sites like Bhaja, Bhasa and Pitalkhora.

It is probable that the Buddha envisaged his monks as leading a permanently peripatetic life – the life of the wandering scholar and thinker that he himself led: ‘You cannot travel on the path,’ he said, ‘unless you become the path itself.’ His last words to his disciples told them: ‘Walk on!’ Yet, by the second century BC, when the great rock cut monasteries of western India began to be constructed, Buddhist monks had already begun to turn away from the road to embrace instead the more sedentary life of the hermit in his cell.

The moment of realisation as the King of Varanasi realises he has missed the deer but shot an innocent boy fetching water in the forest. Ajanta cave 10 mural, 1st century BC.

It was the monsoon that acted as the engine of change. For the rains made travel impossible, and since the Buddha did not want his monk drowning in floodwater, he allowed them a ‘rain retreat’. During this time the bhikkhus – literally beggars, as the Buddha called his followers – were allowed to congregate on higher ground and to live in huts of wattle or daub, or better still in natural caves in the rock face of the Himalayas and the mountains of the Western Ghats. It was from these sites, in time, that the great Buddhist monasteries arose.


The early rock cut cave monasteries of western India predate almost all the surviving written texts of Buddhism, and all we know about them comes from the Sanskrit and Pali inscriptions left on the rock walls by the monks, their patrons and their lay devotees. By then the great monasteries appear to have been as powerful as those in mediaeval Europe, and often had their own mints and owned landed estates, some of which were worked by slaves. The inscriptions show how suprisingly middle class and mercantile early Buddhism was: not only were the monasteries built along trade routes, but the patrons of these early monks were often merchants or bankers.


The chaitya halls executed at this time were some of the first spaces in Asia made for congregational worship. They were created as part of a momentous change in religious practice, and provided a setting for a new form of communal Buddhist worship directed at the stupa, which had come to be seen as the living embodiment of the Buddha. In these halls, the monks of Ajanta would gather together to chant and sing and take darshan of what they believed to be the living presence of the Buddha.

They were built, if not for eternity, then at least for the foreseeable future: as one inscription puts it, these chaitya halls were deliberately made to endure a kalpa – an entire cosmic age. A later dedicatory insciption gives an indication of the motives of the patrons who paid for this work: ‘Realizing that life, youth, wealth and happiness are transitory,’ it reads, Varahadeva, Prime Minister to the great King Harisena, ‘made this dwelling to be occupied by the best of ascetics. It resembles the palaces of the Lord of the Gods, clothed in the brilliance of Indra’s crown. As long as the sun shines this spotless cave may be enjoyed! Even a single flower offered here can yield freedom and the fruit known as paradise. The wise man will show reverence.’

The paintings in both Caves Nine and Ten appear to be a little later than the actual excavation of the cave – an incised dedicatory inscription on the left wall of Cave Ten was covered with a thick layer of plaster upon which the paintings were done, so the cave was clearly plastered after the excavation was complete. Technically they are slightly different from the later painting: they are painted on thin lime not mud plaster and they use only local pigments. There is for example none of the blue lapis lazuli imported from Afghanistan which appears in the later images.


On the façade of Cave Ten there exists a carved panel which mentions one Vasithiputra Katakadi as contributing to the excavation of the cave’s frontage. From his name he was probably a prince of the Satavahana dynasty, which controlled the Deccan after the break-up of the Mauryan Empire and the assassination of Brihadratha, the last of the Mauryas, in 187 BC. The style of the turbans, the costumes the figures wear, the absence of images of the Buddha, and the similarity of the figure style to that of other early sites like Sanchi also point to this period. In addition, the presence of very different 5th century CE over painting confirms this very early date.

Kings and deities look on as the Buddha preaches his first sermon. Ajanta cave 10 mural, 1st century BC.

The 1st century BC was a period of political disruption across India but also of great artistic and intellectual ferment. The early caves of Ajanta, like the sculpture decorating the stupas and gateways at Sanchi, Bharhut and Amaravati, contain inscriptions which show that the caves were the result of numerous small contributions, mainly from pious monks and nuns, but also from local noblemen, guilds and craftsmen rather than from a single king as became the norm for most later Buddhist and Hindu art.


Inscriptions in Cave Ten record these, like sponsorship tags on television adverts today, mentioning a Kanhaka of Bahada, presumably a nobleman, and monks named Dharmadeva, Buddhinaga and Sikhabhadra, the latter ‘in honour of his mother and father.’ One pillar was ‘the meritorious gift of the teacher Sachiva. Whatever merit is in this, let that be for the good of all sentient beings.’ This was a period not of royal but community patronage, and appropriately the art that these guilds and pious individuals sponsored was a crowded, vibrant narrative art, teaming with people and alive with drama and incident. It lies in marked contrast to the later period when each cave had a single rich patron: men like the monk Buddhabhadra, the chief donor of the opulently appointed Cave 26, who was clearly a man of wealth – in the inscription he describes himself as ‘the friend of kings’. As Walter Spink has noted: it is unlikely ‘that he spent very much time humbly wandering from village to village with his begging bowl as his predecessors in the early days of Buddhism did.’


The murals of Cave Ten has recently been shown by Dieter Schlingloff to contain one supreme treasure that had not been previously identified: fragments of the oldest surviving painting of the Life of the Buddha. One scene shows a group of kings and deities venerating a Bodhi tree, a stupa and dharmic wheel of the law. This appears to be an image of the first sermon at Sarnath, with the tree taking the place of the Gautama Buddha.

Next to the first sermon lies a fragment of the legend of Udayana, a tale of two rival queens, one virtuous and one evil. The evil one places a snake in the vina the king liked to play one evening, deviously framing her virtuous competitor as a murderer; but when the enraged king tries to shoot her with an arrow, such is the power of her virtue that the arrow reverses and hits her accuser. The most dramatic and best preserved scenes, however, show two Jataka stories, both of which illustrate tales of killings by huntsmen in the jungles of ancient India. At the furthermost end of the apse, the artists drew images from the Shyama Jataka which tells of Shyama, a virtuous forest dweller who was fatally hit by the poisoned arrow of the King of Varanasi who was out hunting. Because he was sinless, his blind parents were able to call him back to life and he becomes the king’s guru and instructor in the virtuous ways of Dharma.

Next to it is the Chaddanta Jataka which tells of a virtuous six-tusked elephant who is killed at the instigation of a jealous and vindictive queen. When she sees his tusks, the queen is overcome with remorse and dies of guilt and heartbreak. The elephant is buried and commemorated in a stupa erected near the pond where he lived.


In these images there are no clear boundaries between these palaces and the jungle beyond, as royal processions shaded by elegant white parasols and princely hunting parties venture out past a phantasmagoria of rocks into thickets full of monkeys and herds of wild elephants. We see clearly both the holy stories and the exuberantly depicted natural world in which they take place: jungle tendrils curling, deer running, elephants lumbering through the forest. Even the biological boundary between the human and the natural world is blurred as friezes unroll like ribbons in Cave Nine to tell the tales of Naga princes who are at once both royal personages and snakes, crowned with diadems of curved cobra hoods.

One feels strongly here the Buddhist intuition that the natural and animal world are closely related to humankind through cycles of reincarnation: a neglected elephant queen in Cave Ten is reborn as the Queen of Varanasi – yet remains essentially the same soul. As a result the animals at Ajanta are depicted with the very same love and the same respect and individuality as the humans.

Attendant of the King of Varanasi watches the king as he prepares to shoot. From the Shyama Jataka – Ajanta cave 10 mural 1st century BC.

In illustrating these three stories, the early artists of Ajanta open wide a window on an age which remains otherwise dark and shadowy to us. We see the costumes of this very early period: the King of Varanasi, for example, wears a white cotton tunic of strikingly Central Asian appearance, wrapped around the waste with a cummerbund, while on his head he wears a turban wound around his hair and twisted into a top knot. He has a bow and a full quiver of arrows. His guard are bare-chested but wrap an antariya lungi around their hips and are armed with spears and bows and bell-shaped shields decorated with the emblems of half-moons and shining suns. The different turbans of the different ranks are shown with great care and seem to be an important indicator of status, the different materials – some with red or gold stripes, others pure white – and the different styles of wrapping delineated with great care.


Much about the clothing is identical to that worn in the sculptures of Sanchi, Bharhut and Karle and there are some of the same compositional strategies – the fondness, for example, for ranks of figures looking from the left and right of the frame at some event taking place at the centre. There is also the same interest in the prose of the world: in describing the detail and variety of lives lived by both humans and animals in cities and villages and woods and fields. Yet the style is in feeling and temper subtly different from Sanchi.

In Sanchi, the images are animated with a wonderful joie de vivre. Here there is a sadness in the programme of painting, which is concerned with issues of justice, peace and non-violence: one image tells of a war breaking out over the Buddhas relics – something that went totally against the grain of everything the Buddha taught. This is followed by three images which all tell of the unjust killing of an innocent: successively, the loving wife of a king who is unjustly accused of trying to kill him; the shooting of a boy as he fetches water for his parents in the forest; and a noble elephant king murdered by hunters for his tusks.


There is also something very different from Sanchi in the style of these faces. Their physiognomy and the world in which they are depicted may be utterly Indian; yet in the artist’s long, bold brushstrokes in the depth and modelling, in the use of perspective and three-quarter profiles, in the play of light and shade across the large brown eyes and cheekbones, and in the technical aspects of the work with its choice of pigments and use of lime mortar – all show a strong hint of Hellenistic influence.

This is hardly a surprise: after all at this period Indo-Greek kings controlled as far South as the Punjab and were integrated into the Indian world: at least one of them, Menander or Milinda, actually converted to Buddhism, while another, the Bactrian Greek Ambassador Heliodorus, left his famous inscription at Besnagar, that recorded the fact that he was ‘a devotee of Vasudeva, the god of gods’, an early form of Vishnu. At other caves in Maharashtra there are references to donors who describe themselves as Yavanas, foreigners – probably Greeks.

This intimacy, classicism, and striking realism, combined with the haunting wistfulness of the features of these faces is not a million miles away from the melancholy world of the later 1st century CE encaustic wax mummy portraits from the al-Fayum region of Egypt, which were also the products of the Hellenization of the East. There are also distant echoes of the frescoes and mosaics of Pompeii which date from roughly the same time. Closer to home, there are interesting parallels with more recent finds of extraordinary textiles excavated with the Tarim Basin mummies on the southern silk road around the Gobi, especially a trouser leg recently identified as bearing the image, possibly a portrait, of a Parthian soldier.

As with these other images we are in a world so astonishingly realistic and lifelike that even today, even in reproduction, they can still make one gasp as one finds oneself staring eyeball to eyeball with a silently watching soldier who could have fought the Bactrian Greeks, or a monk who may have seen the sculpures at Sanchi being carved.


So realistic are the faces of the people depicted, so direct are their expressions, that one feels that these have to be portraits of real individuals, glowing still with the flame of eternal life. There is none of the idealization or otherworldliness one sees in the later images of the Bodhisattvas. Instead there is something deeply hypnotic about the soundless stare of these silent often uncertain Satavahana faces. Their fleeting expressions are frozen, startled, as if suddenly surprised by the king’s decision to loose his arrow or by the nobility of the great elephant breaking through the trees. The viewer peers at these figures trying to catch some hint of the upheavals they witnessed and the strange sights they saw in ancient India. But the smooth, clean humane Indo-Hellenistic faces stare us down.

Perhaps the most disconcerting thing about the people in these murals is that they appear so astonishingly familiar. Two thousand years after they were painted these faces convey with penetrating immediacy the character of the different sitters: the alert guard, the king caught in the excitement of the hunt, the obedient son fetching water. Indeed so contemporary are the features, so immediately recognizable the emotions that play on the lips, that one has to keep reminding oneself that these sitters are not from our world, that they are pigments attached to the wall of a cave, and depict a court and jungle world of hunters and hunted, and Buddhist monks and devotees, that vanished from these now bare Deccan hills more than two millennia ago.


Yet these are self-evidently the same people who inhabit western India today: looking at these images you cannot help but feel the great distance of time separating them from us; and yet we find in their eyes an emotional immediacy that is at once comprehensible. Some of them look like the guards who admit you to the caves: indeed while the glass coverings were being removed to allow the photography for this piece, the guards joked among themselves about which painted king looked most like which guard. The women on the cave walls wear the same bangles as the Banjara tribes of these hills still stack along their forearms, and their dupattas are decorated with fringes of taarkaam or Paithani, still popular in Maharastra today, as are the fishscale kham textiles which clothe the hunters in the Shyam Jataka.

It is odd and eerie to stare into the eyes of men and women who died more than 2000 years ago; but odder still to feel that their faces are somehow reassuringly familiar.


* Photographs: the author.