Walter M. Spink | Revised version (September 2008)

NOT SO LONG AGO, in 1956, I started publishing my ideas about the Ajanta caves for the first time. But like most scholars, I trusted conventional wisdom to do most of my work for me. We all knew, from inscriptional evidence, that some of the caves at Ajanta were made in the fifth century, but why did we all believe that others, such as the beautiful Cave 1, with its splendid bodhisattvas should be assigned to the mid-seventh century.  Today, I am trying to decide if these same bodhisattvas should be dated, not to the mid-seventh century, but to 477, or possibly 476, or 478—some two centuries earlier.

How did we all make such a significant mistake, which so deeply confounded our understanding of history during the period of India’s so-called Golden Age. For the final flowering of that great age came, if I am correct, not during the ascendancy of the famed Guptas, but during the reign of the little-known Vakataka emperor, Harisena, in the 460s and 470s. And if it possible to prove that this is true, we must clarify Ajanta’s development, for that is where the key to the problem is hidden.

The confusion about Ajanta started in 1879, when James Fergusson, speaking before the Royal Asiatic Society, proposed that this scene, located on the front wall of Cave 1 at Ajanta and clearly showing some kind of royal audience, might depict the visit of an ambassador from the court of the Sassanian monarch Chosroes II to that of the Calukyan ruler Pulakesin II, an event that took place in about 625 AD.  Scholars were so eager to know when the more developed Ajanta caves such as Cave 1 were made, that they soon convinced themselves that this suggestion was reasonable; and from there it was only a short distance to the conclusion that it was true.  For this reason the important Cave 1 soon came to be conventionally dated to the mid-seventh century.

This assumption clarified other vexing problems as well.  Since the famous but undated Great Cave at Elephanta was clearly more developed than Cave 1 at Ajanta, it was now assigned to the eighth century. Before too long, of course, a few scholars did start asking: what is a painting of a Persian delegation to a Hindu king far to the south doing in a Buddhist cave at Ajanta?  But by then the (presumed) eighth century dating for Elephanta had itself become conventional, and this in turn supported the seventh century dating for Ajanta. Thus the scholarly establishment had built an elaborate circular structure of argument on shifting Sassanian sand, and as a consequence both Cave 1 at Ajanta (which really belongs in the fifth century), and the great cave at Elephanta (which really belongs in the sixth century) were dated about two centuries too late.

The fact is, the painting that caused the confusion probably is an “embassy scene”; but surely not a Persian Embassy as we once thought. Professor Schlingloff, famous for work on the Ajanta paintings, has now identified it as a representation of the Buddha in a previous birth as a king called Mahasudarsana.  The fact that the crowd of vassals includes people in Sassanian garb is hardly surprising. Many of the murals at Ajanta, like this luxurious palace scene in Cave 17, often show the exotic foreigners (here a Sassanian servant with a Sassanian ewer) who came to India along the great trade routes flourishing during this happy period, linking India with stylistic sources both to the north and west, and to the southeast.

Our so-called Embassy Scene is in fact a Jataka Tale; and the fact that it shows him as a king is because this is a royal cave.  It was given by the great Vakataka emperor Harisena—Ajanta’s most important sponsor. All of the murals in this particular cave—Harisena’s Cave 1—belong to a carefully organized royal program, showing the Buddha in his previous births, not as a noble deer or as a self-sacrificing elephant, but always, in this cave, as a king.  Here the future Buddha, in another representation of one of these auspicious previous births, (but of course depicted in contemporary dress) is seen in the midst of the royal delights which he is about to renounce in search of enlightenment.

It was not until the about 1960 that I stumbled across the lucky evidence that proved how wrong I was in ascribing Ajanta’s highly developed later style, typified by the scenes we have just seen in Cave 1, to the seventh century.  And since those early visits, I have been trying to figure out how the site developed for almost fifty years—far longer, my colleagues remind me, than it took the excavators and the sculptors and the painters to make the caves.

THE EVIDENCE EMERGED, for the first time, when I visited a remote cave in a beautiful valley, known as the Ghatotkacha vihara, not too far from Ajanta, but hardly ever visited except by thousands of bats. It is sadly neglected today, but in fact was made by the emperor Harisena’s Prime Minister Varahadeva who, probably more than any other person, was responsible for Ajanta’s renaissance in the 460s and 470s.  That was when about a dozen of Harisena’s courtiers and feudatories, impressed by the grandeur of the ancient Hinayana caitya halls throughout India and at Ajanta itself, as well as by the early viharas for the monks’ residence—all excavated some three hundred years earlier—decided to outdo their ancestors and to create one of the wonders of the world.  The Ghatotkacha vihara is a contemporary reflection of this same Vakataka energy, as are some of the caves at nearby Aurangabad, the small cave at Banoti in its huge gorge, and the now ruinous but once glorious caves at Bagh, near modern Indore.

I knew, of course, that the Ghatotkacha Vihara had been sponsored by the Vakataka Prime Minister, for the great scholar V. V. Mirashi had already published Varahadeva’s informative dedicatory inscription, cut on the left rear wall of the porch.  The inscription is of great interest, in part because it refers to Varahadeva’s monarch, the emperor Harisena, known to be a late fifth century ruler.  Of course we have no contemporary portraits of the emperor Harisena, but he well have been ceremonially adorned like this handsome bodhisattva attending the Buddha in the emperor’s cave.

But what my art-historian’s eye focused on was not the inscription, but an adjacent pilaster capital that, if I had seen it at Ajanta, I would have ascribed to the seventh century, on the basis of its highly elaborate character.  I would never have expected to find it in the earlier Ajanta excavations which, like Ghatotkacha itself, were connected by their inscriptions to Harisena’s reign.

So how could this fifth century Ghatotkacha inscription coexist with such a decorative motif which we all then assumed belonged to the seventh century?

It was to face this confusion that I first proposed that the caves of the main (Vakataka) phase at Ajanta probably had all been created during the period when the emperor Harisena was reigning in the last half of the fifth century.

TO PROTEST THIS DRASTIC REVISION, a number of critics insisted that these surely “late” motifs at Ghatotkacha must be much later additions—that the cave must be seen as having two distinct phases; a fifth century phase when it was originally defined, and another much later phase when the elaborate decoration was added.

But why would the Prime Minister have dedicated the cave when, without all these late features, it would have been only half-completed? It was clear that I would have to turn back to Ajanta itself, to see if such so-called “late” features could possibly be ascribed to Harisena’s reign, as I felt must be the case at Ghatotkacha.

Obviously, there was no better place to start than at Ajanta Cave 16, made by the same Prime Minister Varahadeva, who, according to his inscription, ”governed the country righteously...(shining brightly) with the rays of his fame, religious merit, and virtue”…. Even the energy in his architecture would seem to confirm that Varahadeva’s authority. As one of the founders of the site’s renaissance in the early 460s, he was able to choose the very best central location for his cave.  This is where he placed his famous Elephant Gate, known in ancient times as the Entrance to the Site.  Just within, the Lord of the Nagas, a snake divinity, surveys the whole ravine from his place of honor.

As we might expect, the prime minister’s cave, as one of Ajanta’s inaugural Vakataka monuments, contains some of the most revealingly early features at the site: the simplicity of the octagonal pillars, the unnecessarily high ceiling—useful in palaces but not in caves for air-conditioning; and the soon outmoded shape of the still undecorated windows.   The ceiling too, with its beautiful but simple floral forms further confirms its early date, when we see how far more complex the site’s ceilings will become a mere ten years later, as here in Cave 1.

ON ENTERING CAVE 16 one is immediately struck by the grandeur of its great Buddha image. When I first started my research on Ajanta, I believed that this was perhaps the earliest shrine Buddha at the site—for was not Varahadeva one of the site’s inaugural donors. But now I insist that this great Buddha is one of Ajanta’s latest shrine images, perhaps the very latest of all. How can this be?

The reason is that the Prime Minister’s Cave 16, like his Ghatotkacha vihara, developed—and developed with a remarkable vigor—in two distinct phases. But these rather distinct phases were separated by only a few years, not a span of one or two centuries, as we used to think.

It is clear that in the 460s Cave 16 was planned like the adjacent Cave 17, with a pillared porch, a large pillared hall surrounded with cells for the residence of the monks, and a conventional shrine.  But then, about a decade later, after a distinct break in the cave’s development, Varahadeva decided to drastically revise the plans for the shrine, and to put a dramatic new type of authoritatively seated image—never carved at the site before—into an equally innovative pillared pavilion.

In fact, this drastic change of plans, and the gap in time that separates the cave’s bipolar development was due to war—a war that was caused by the proud but impractical local king, Upendragupta.  And its consequences explain the enigma of the similar two-phase development at Ghatotkacha and indeed in most of other caves at Ajanta as well.

The very fact that the excavations at Ajanta could be begun very shortly after Harisena acceded to the throne in about 462 testifies to the stability of the extensive empire that Harisena had inherited. Besides his own domains in Western Vidarbha, he also reigned over the feudatory kingdom of the Ajanta region (ancient Risika), the large and eventually trouble-making area to the South, Asmaka, and extensive other regions.  And all were at peace, at least for the time being.  The stock market was rising. Fashion was flourishing, at least among the rich.  And the roads were safe. Inspired by the great Vakataka emperor, Harisena’s pious—and proud—courtiers decided that it was time to turn dreams into reality—to honor the Buddha, and to earn religious merit—and earthly praise—for themselves.

THUS AJANTA BEGAN in a great burst of activity. In the early 460s no less than twenty-two major excavations were started, as the site began to be developed after a gap of 300 years.  But now, instead of being a typical community effort, as in the earlier (Hinayana) caves at Ajanta itself, where a donor named Katahadi made “the gift of the façade”, and another donor made “the gift of a wall” and yet another “the gift of a cell”, each cave was sponsored exclusively by a single important patron. And what they created was totally elitist.  No one else—the local devotees from the villages around the site, or the merchants and pilgrims arriving at the site from the nearby trade route, or even the hundreds of monks beginning to gather at the site, to say nothing of the artisans, both skilled and humble, flooding into the site in search of work or even fame, could spend a single rupee of their own on the developing caves, and thus share the merit.  The caves were the exclusive province of the courtly patrons.

So, especially for these privileged patrons, it was a happy and abundant world, at least for a few years.  But by about the end of 468, the sky darkened.  The peaceful relationship between the luxury-loving Upendragupta, the local feudatory king, and the powerful feudatory ruler of nearby Asmaka suddenly ended when Upendragupta expelled the Asmakas from the Ajanta region.At this point work was halted abruptly on the remarkable complex of caves—still barely roughed out--that the Asmakas themselves were developing at the western extremity of the site.  And what is also significant is that Upendragupta, the local king, ordered work stopped on nearly every other cave at the site as well. Only Upendragupta’s own caves (17, 19, 20, and 29) and (not surprisingly) Harisena’s Cave 1—that is, only the royal caves—were exempted—were allowed to continue underway.

Such a drastic prohibition could only mean that the luxury-loving Upendragupta, the local king, was now so threatened by his aggressive neighbor’s power, that he realized that he should start preparing for war.

Upendragupta’s worries are clearly confirmed by the way he treated his own privileged caves during the next two or three years.  His lavish goal, as he says in his Cave 17 inscription, was to use “the power of the expenditure of wealth” to continue to develop his own caves with his typically spend-thrift vigor.  However, very soon we begin to see signs of stress, of things being rushed in the interest of economy.

The great shrine in his Cave 17 got plastered but its walls were never painted. The most drastic cutback of all is seen in the fate of his Cave 29, started as an insult to the Asmakas, just after their expulsion, to replace their own great Cave 26. However, work went on for a mere single year only, before it had to be abandoned.  Only the beautiful Caitya Cave 19, planned by Upendragupta as the devotional center of the site, was largely exempted from the general sumptuary restrictions. Here we see Upendragupta’s demanding vision at its best; indeed it becomes an inspiring source for all later work at the site.

Sure enough, by 475, the aggressive Asmakas had come back to the site in force.  At this point, as the tanks and the supply trucks—the war elephants and the heavily-laden donkeys—came over the hill, Upendragupta, the threatened local king, had to flee. He never even had time to get his laudatory dedication inscribed before he left. Not only was his beautiful cave declared “off-limits” by the invading Asmakas, but they insultingly cut a passage right through the monks’ cells in its courtyard, to make it easier to get to their own caves at the site’s western extremity.

Now, but under the firm rule of the conquering Asmakas, the site started to flourish once again, most notably in the Asmaka complex. If we wonder where the money came from,the cave’s patron, the great monk Buddhabhadra proudly claims in his inscription to have been the friend of the pious minister of Asmaka “through many previous existences”—they had been continually re-cycled together.  And the funds available must have been enormous.  Although in 468, when the Asmakas were banished from the site, their caves had been merely roughed out, now the Asmaka caves take the lead in the site’s development, being clothed in a marvelous tapestry of carved and painted forms, all added after the Asmaka takeover of the region, in 475.

We can now understand that both Cave 16 and the Ghatotkacha vihara, as well as the vast Cave 26 itself, and in fact nearly all of the caves at the site have two very distinct phases, separated by a short gap in time when the aggressive Asmakas were battling to take over the region.  Thus nearly all of these caves are both “early” and “late”; for all of them, started at the very inception of the site’s Vakataka phase, were still underway when work at the site suddenly broke off.

ANOTHER CAVE which is characteristically both early and late is the vast Cave 4, donated by a wealthy patron named Mathura.  With its severe and simple porch, Cave 4 is a typically inaugural undertaking, started about 464. Work on it progressed with stunning speed; the whole interior was roughed out in the next five years, before it was interrupted by the local king’s expulsion of the Asmakas in 468. As if this political crisis were not enough, the vast ceiling next collapsed due to a pernicious flaw in the layered basalt, laid down eon after eon by volcanic action.  Nonetheless, in 475, when the aggressive Asmakas took over the site, work at the site started up again.

The development that takes place between the simple early pillars, and the later ones toward the rear of the cave reveals a dramatic course of development, typified by the eight charming little dwarfs bringing music into the cave from their location near the shrine.  Even more important, the six Buddhas of the Past now attend the Buddha Sakyamuni in the shrine.  The fact that such work was generally interrupted, as a result of Harisena’s unexpected death, gives a revealing insight into such work in progress.

Work on the shrine was also direly affected by the great emperor’s death, but the patron Mathura, like many of his fellow patrons, did manage to get the shrine Buddha hurriedly completed early in 478, even though the great hall itself, never painted, was still in process when time ran out. Understandably, he did manage to get his inscription hastily written on the image base.  Let the merit be, he says, “for the attaining of supreme knowledge by [his] mother, father, and paternal grandmother—to whom belongs the principle share—as well as by all living beings.”

TT WAS THE DEATH of the great emperor Harisena late in 477 that put Ajanta into a state of shock, and indeed sealed its doom. We can best understand what happened, when we turn to Harisena’s own excavation, the sumptuous and dignified Cave 1.  This is the most splendid vihara in the whole of India—a cave indeed “fit for a king”.  Even though the emperor Harisena died before he could dedicate the cave, his authorship of it cannot be disputed.  His is the only cave at the site that continued to develop—and to develop with an unparalleled grandeur—first when the local king expelled the Asmakas from the site, and second when the Asmakas expelled the local king from the site.  Therefore, since both of these feudatories honored him, he was obviously their overlord.  The particular attention given to his cave confirms this view, as does the site’s frantic and troubled development after his death.

Harisena’s supreme status is confirmed by the fact that his Cave 1 is the only vihara with a remarkably decorated façade. Fronted by a courtyard of “imperial” size, the facade describes the prerogatives of royal rule.  Thus we see splendid representation of scenes of battle, as well as complex descriptions of the royal hunt, and finally, when the battles and the hunts are over, the rewards of erotic dalliance.

Inside the cave, every one of the beautiful scenes in the main hall forms part of a program focusing on the virtues of kingship, as noted earlier.  Even the two famous bodhisattvas on the rear wall, Avalokitesvara representing compassion, and Vajrapani spiritual power, have kingly crowns—in fact the gloriously elaborate crown of the Vajrapani very likely duplicates the kind of crown made by the imperial jeweler for the emperor himself.

However, ominously, the decoration of Cave 1’s  main hall was never quite finished, and when we look into the shrine area, it is clear that the great image, although fully carved and painted, was never dedicated.  Its painted surfaces, although damaged by time, have not been darkened at all by the soot from the oil lamps used in worship.  The paintings in the antechamber are equally unsullied, while it is clear that garlands were never hung from the hook (now missing) at the center of the antechamber medallion. Had they been changed daily, as in Cave 2, the painting on their fragile mud-plaster base would have been damaged when the monks daily changed the garlands.

The fact that such an important undertaking was never used for worship, can only be explained by the assumption that the great emperor died, and died so suddenly, that he could not get it finished and dedicated—as was essential if he was to acquire the merit from his great endeavor. In fact Harisena’s cave, the most beautiful vihara in India, so full of life, is dead; because he never was able to get it dedicated.

The fact that Cave 1 is dead is confirmed by clear evidence. We know, for instance, that when the courtly patronage of the site so suddenly ended, in 478, the monks and devotees remaining at the site eagerly utilized the caves for their own donations.  But such intrusive offerings were placed only in caves that had been dedicated.  So the fact that there are no such intrusions in Cave 1 again tells us that the great cave is dead.

HARISENA'S DEATH, in roughly 477, was a tragedy— a death-blow—for his great Vakataka empire. From the moment that work stopped on the emperor’s Cave 1--that is, from the moment that the emperor suddenly died, perhaps from a heart attack, but more probably from poison or a knife held by an Asmaka terrorist--the situation changed.  Within the very year after Harisena’s death, the site, under the rule of his inept successor, Sarvasena III, went into shock.

When the Asmaka monk Buddhabhadra, overseeing the development of the site’s western extremity, ordered his inscription placed on the rear wall of the porch of the huge Cave 26, he praised his own Asmaka overlords, but he did not even mention this ill-fated new monarch, the useless Sarvasena III. This was a deep insult—the rejection of the Vakataka overlordship—and it clearly meant war.

Now, suddenly, the expectation of war put all of the patrons loyal to the Vakatakas—whose caves were those along the main, or eastern, end of the scarp—into a turmoil.  Immediately, they gave up any thought of finishing them properly, and turned all of their efforts and their diminishing funds into completing their shrine Buddhas alone, in order to get them dedicated before the Ajanta region was taken over by the Asmaka forces.

The situation in the first months of 478 was dire, and the rushed completion of Buddha after Buddha describes it.  What we have seen in the unfinished Caves 4 and 16--the image being hastily completed and dedicated in order to get the merit—was repeated over and over in the other caves by the anxious Vakataka donors.

The caves in the more remote Asmaka section of the site were also deeply affected by the expectation of war.  But since Asmaka was now in total control of the site, work on their own caves went on throughout 478 in a somewhat normal fashion.

Buddhabhadra himself sponsored the greatest of all Ajanta’s sculptures in this turbulent year: the massive Dying Buddha, the remarkably complex Temptation of the Buddha by Mara, the God of Desire, and the richly conceived panels of the Sravasti Miracle. These great sculptures are staggering achievements, giving great promise for the future.  But the future was not to be.  By the end of this turbulent year, 478, the Asmaka resources had been totally diverted to the increasing military needs.  It is clear that this was going to be a serious war; and as soon as this became clear, Buddhabhadra and his Asmaka associates, like the other patrons at the site, now also struggled to get their shrine Buddhas quickly completed, in order to get the merit from such donations.

If we review the situation throughout the whole site, we find that in the disastrous, but incredibly productive months immediately after the great emperor Harisena died, no less than twelve shrine Buddhas were rushed to their anxious completion and dedication, at which point all of their proud patrons had to depart, sadly and hastily, from the dying site.

When we analyze the development of the site, we find that the majority of the great excavations started in the early 460s were still incomplete and underway at the site’s sudden demise following Harisena’s death.  And except for Harisena himself, and perhaps also the exiled local king, Upendragupta, the very patrons who started the caves were apparently all still alive at this time; in six cases inscriptions prove this, and even where there are no such records, one can assume that most, if not all, of the original patrons were still working on their caves when time ran out.

FROM A PUERLY ACTAURIAL POINT OF VIEW, it is impossible that such a large group of presumably mature donors, having started such a project, would all still be alive at its ending, if it had taken as much as twenty years.  Thus actuarial data inarguably confirms the telling epigraphic and literary evidence that equally assigns the dating of Ajanta—and in fact all other late Vakataka monuments—to a single brief span in the 460s and 470s.

But if, with the departure of its major patrons in about 478, Ajanta’s doom was sealed, its end was not quite yet. Now the floodgates were opened, and anyone could add an image, whether painted or carved, anywhere, in a completely helter-skelter way; as for the now-stranded artists, although highly skilled from years of work, they must have been willing to work at bargain rates.  It had clearly become a “buyer’s market”, even though the buyers themselves were now in a hurry to get away.

Once on the now-dangerous roads, the best protection of all was the bodhisattva Avalokitesvara, the Lord of Travelers; he, in particular, could protect fleeing travelers from the dangers of the road—the brigands, the wild beasts and reptiles, even the hazards of boating across the dangerous river-crossings.

This spate of intrusive donations, itself soon affected by the threats of war, did not last for long. Suddenly, as if on some Tuesday morning late in December of 480, in areas where whole groups of intrusive images were still unfinished, we have clear evidence of how precipitously work stopped.  At this point, all of a sudden, the sculptors laid down their chisels, the painters laid down their brushes, and everyone hastily went away, leaving work that sometimes could have been completed in a mere few hours, suddenly abandoned.

If we have read the evidence correctly, within a matter of days almost everyone was gone—all moving away--hurrying back to their native places, or to safer places, as they heard the developing sounds of war.  This may have been at the very time, according to the Dasakumaracarita (the historically based Tale of the Ten Princes) that the great emperor’s doomed successor, Sarvasena III, “having the border of his kingdom invaded, mobilized his army to march   against (the Asmaka coalition)”.1 But he himself, as one translator puts it, “became mincemeat” on the banks of the Narmada river, where his forces finally fell.

We know that a few monks stayed on in the 480s, even after everyone else had gone;  signs of wear in the pivot-holes of the latest cell doorways, showing that they continued in use for at least a few years, tell us this. After all, where else were these remaining monks to go, with their once-glorious world falling apart? But even so, surely within the next decade, they too were gone; and in  Ajanta’s beautiful ravine, “resonant with the chirpings of the birds and the chatterings of monkeys” (Cave 26  inscr vs 18), the amazing caves, intended to “cause the attainment of well-being by good people as long as the sun dispels darkness by its rays” (Cave 17 inscr, vs 29), now lay untouched, protected during the many coming centuries by the sad, but preserving, circumstance of their centuries of isolation.