Deepak Kannal

Retd. Prof. Deepak Kannal is a sculptor and art historian with expertise on Ellora caves. Former Dean of the Faculty of Fine Arts and Head of the Department of Art History and Aesthetics, M.S. University, Baroda, his two publications on the subject are "Ellora: An Enigma in Sculptural Styles" and "Ellora Caves: Sculpture and Architecture".

In conversation with Deepak Kannal on the Ellora Caves, Baroda, February, 2016


Ellora is perhaps one of the most fascinating sites in India. Nowhere else would you be able to find the monuments of all the major religions of ancient India in one place. There are Buddhist (Mahayana and Vajrayana), Brahminical and Jain caves. So all these things together make it very exciting. It is about 20 kms from a place called Aurangabad, and there are many other caves in the vicinity. Ajanta and Pitalkhora are not far from Ellora. The small caves at Bhokardan are also very close. So it seems that this particular area was quite busy with this kind of architectural activity right from fifth century till about eleventh century. We are not very sure about the exact date when work on Ellora commenced.


Some scholars are of the opinion that it started with Buddhist sculpture. Some believe that the Shaivite caves were the earliest caves at this site. I personally think that the Vaishnavite activity right next to the waterfall near cave number 29 is one of the earliest at Ellora. Now it is a little difficult to prove it, but on the basis of the available data, and considering the iconography of the sculptures in those caves, one can date the cave back to a very early age.


Ellora was not lost in amnesia. It seems to have been a living site ever since its inception, because it has been mentioned in ancient literature, medieval literature, pre-independence literature, and many European scholars have also talked about it. The earliest references to Ellora can be found in texts from the Yadava period. References to a cave temple, not exactly Ellora, can be found in a thirteenth century text called Jnaneshwari, perhaps one of the earliest texts in Marathi. It speaks about a temple which has been carved from top to bottom. Now we do not know whether the reference is to Kailasanath or not, but there is every possibility that it could be Kailasanath because Jnaneshvara’s home town was not far away from Ellora.


But there is another text which is almost contemporaneous to this—Sthana Pothi by Leela Charitra—where you find specific references to Ellora. It seems that the leader of that cult, Chakradhar Swami, spent four months in Ellora and most probably stayed in cave number 11. It is possible to visit the place where he stayed in Ellora. In Sthana Pothi, he and his disciples discuss Ellora. They visit different places and different caves and keep asking questions. In one of the dialogues, these disciples of Chakradhar walk through different caves for several hours, and when they come out from some other cave, they see a village in front of them and they ask Chakradhar, “Which village is this?”. Chakradhar answers, “It is Ellora village.” The disciples ask, “How is it that we have been walking for hours and we have not moved at all?” Chakradhar replies, “This mountain has been carved thoroughly from within, and though we walked several miles, we are in the same place.” This explanation that the mountain was totally hollow could be an exaggeration, but it seems that many of these caves were interconnected. If you visit the Jain caves, you can see that they are interconnected. Perhaps similar interconnections were there in the Buddhist caves and Brahminical caves too. It also seems that in this particular period, some esoteric activities were going on in these caves. In one of the dialogues, these disciples go to a cave where they see some people engaged in some frightening activities. This is narrated in ancient Marathi. The text describes it as, “Nake phul phula vithi, aakhe ghode mila viti unpiyaa…”. Their teacher tells them that that particular time was allotted for these esoteric practices and that they should not visit those caves at that hour. This means that Ellora was an active esoteric site in thirteenth century.


In the fourteenth and fifteenth centuries it continued to be a living site, because we know that Ahalyabai Holkar renovated this site in the eighteenth century. If you go to Kailasanath, you will find some very crude paintings on the outer walls. These paintings were made during the period of Ahalyabai Holkar. From the eighteenth century onwards, there are also many travelogues by European visitors like Seely, where they talk about the grandeur of Ellora. Of course, they don’t have a very high opinion about the sculptures because they find them ghastly. They were seeing sculptures with several arms and animal heads for the first time, and since their aesthetic ideas were quite Euro-centric or rather Greco-Roman-centric, they compared these sculptures with Greco-Roman sculptures and opined that Greco-Roman sculptures were far superior.


The first textual reference to Ellora is found on a copper plate which is known as the Siddhamshi copper plate. Now, the Siddhamshi copper plate is controversial because the place mentioned in that particular copper plate was not identified as Ellora for a long time. Bhandarkar was the first person to identify the place mentioned as Kailasanath and then it was recognised as this particular cave temple. The Siddhamshi copper plate has two beautiful poems on it. One says that when the gods were travelling through the skies, they saw this temple and they wondered if it could be the work of mortals. And after a long debate, they finally reached the conclusion that Lord Shiva himself must have created this temple as his abode on earth. The other stanza says that the sculptor who carved it tried to do it again and obviously failed. He then kept wondering how he had been able to do it. He finally concluded that though he had been instrumental in carving the temple, he was not the one who had done it. If you go up the hill near the Kailasanath temple, you will find the ruins of an attempt at carving another cave temple. It seems that this myth must have been born from those attempts which can be found right next to Kailasanath. So Kailasanath is mentioned though the other caves were not.  Kailasanath seems to have been a living monument right from the Rashtrakuta period till today.


Unfortunately, unlike Ajanta, Ellora doesn’t have many inscriptions. Rather, there are three or four major inscriptions. One is in the Dashavatara cave, cave number 15, and is identified as Dantidurga’s inscription. It is a long inscription that talks about the genealogy of the Rashtrakutas and Dantidurga’s visit to that place. The other inscriptions are very insignificant. One of them is a small inscription next to the Gajalakshmi image in cave number 16 (Kailasanath). And there are a few more inscriptions in the Jain caves. They are important inscriptions because they also mention the date.


The Buddhist caves are considered to be the earliest caves by some scholars. The physical sequence of the caves is definitely not the same as the chronological sequence of these caves. We don’t know how exactly they chose the sites. But it seems that a cave was carved at a particular place, then another cave was carved to the south of it, and a third cave was carved to the north of it. We don’t know why exactly these sites were chosen. There is no question of dynasty while talking about the Buddhist caves. They were governed by the Buddhist sanghas. The dynasty which was responsible for the Brahminical activity could be the Kalachuris, because after the fall of the Vakatakas, the Kalachuris established their presence in this space and they were Pashupata Shaivites. They keep mentioning it time and again in their inscriptions. They call themselves Parmo Maheshwara and they are the people who carved the Elephanta Caves also.


If you see cave number 14, quite a few sculptures there divulge their Chalukyan origin. This doesn’t mean that the Chalukyas ruled this region or that the activity received patronage from the Chalukyas. Perhaps it was patronised by the Kalachuris, but the Chalukyan guilds worked on it. This was a very common practice in those days—guilds from other regions were invited to work on projects. For example,   so many sculptors from Andhra Desha worked on the Badami caves. Srinivas Padigar, who has researched the Badami inscriptions, has proved that more than 50 per cent of the sculptors at Badami came from another region. They have Telugu names. The same could be said of Ellora. The carvers must have come from Karnataka, though some were local practitioners. This activity went on for 60–70 years.


Caves carved in the later phase, that is eighth century onwards, are obviously Rashtrakuta caves. There is a theory that at some juncture the Rashtrakutas embraced Jainism.  Most of these theories are debatable, but the Jain caves too are attributed to the later Rashtrakutas by some scholars. But again, unlike the Hindu caves, the administration and execution of the Buddhist and Jain caves were mainly handled by the sanghas or religious groups.


Buddhist Caves


I personally feel that cave number 2 is a very important cave because it shows a marked similarity with the plan of the Ajanta caves. It divulges the Ajanta connection to Ellora, and that is why it is a key cave.The cluster of caves from cave number 3 onwards is a very complex cluster—cave numbers 3, 4, 5, 6 and 7 are clustered together, and you find images of Mahamayuri, Tara and Bodhisattvas here, which reveal the Vajrayana connection, and that is why they are very significant. This could signify the introduction of Vajrayana cult in Ellora.


Cave number 10 is obviously one of the most important caves for two reasons. One, it is the only chaitya in Ellora unlike Ajanta. Ajanta has many chaityas, but Ellora has only one. And the other is that it is perhaps the last rock-cut chaitya in the history of India. There are no more rock-cut chaityas after this. It is a very elaborate chaitya. It was a two-storeyed chaityagriha. The rest of it is quite similar to Ajanta. The ground plan and the central nave with the flanking aisles and the beautiful image of the seated Buddha emerging from the stupa show a marked similarity with cave number 26 of Ajanta.


Cave numbers 11 and 12, and also cave number 15, are very interesting caves. Caves 11 and 12 are multi-storeyed caves and they are almost like hostels. These caves have been carved with tremendous skill. The architects understood structural engineering very well.  The pillars are placed one above the other. And these are very spacious caves, which were probably used as hostels. Cave numbers 11 and 12 have some beautiful sculptures. Cave number 11 has some Vajrayana sculptures in it. Cave number 12 has a beautiful panel of Dhyani Buddhas and Manushi Buddhas flanking the garbhagriha. It is a very elaborate and huge panel.


Cave number 15 was originally a cave of this kind. It seems that it was abandoned halfway. It is still incomplete. The upper storey has been converted into a Brahminical cave. Cave number 15 perhaps was a laboratory. I call it a laboratory because activity on this cave started after a hiatus of several decades, and that is why the sculptures there are extremely inconsistent. The quality of the sculptures and their stylistic features are so inconsistent that it is obvious that a large number of guilds worked on these sculptures. It cannot be attributed to a single guild.


Brahamanical caves


Cave number 18 is a very small cave that has a small Lakulisa sculpture inside it. This makes it significant because it again speaks of the introduction of Pashupata Shaivism in Ellora.


Cave 21 is perhaps one of the most beautiful and one of the most classical caves of this complex. It is classical in every respect. When we talk about classicism in Indian art, it is usually attributed to Buddhist art. I think the same classical and contemplative quality, the same restraint and composure can be seen in the sculptures of cave 21. Even the ground plan is very interesting. It has a very well-lit veranda or hall which is parallel to the façade. And there are two upavarnakas, two side chambers on the lateral side of this hall and a garbhagriha in between. So there are not many sculptures in these caves, but the size of the sculptures and the quality of the sculptures is amazing.


There is a beautiful Nandipitha also in front of it. A sculpture of Aditi Uttanapada  is seen on that Nandipitha. A similar sculpture is found at Elephanta, which makes this very significant. It establishes a connection.


Cave 27 is very important because of its Vaishnavite sculptures. There is an incomplete Sheshashayin on one side. The most significant sculpture in the cave is that of Ekanamsha. This Ekanamsha sculpture, rather the deity itself, almost disappears from the Indian pantheon from the second and third centuries onwards. She is considered to be the yogmaya of Vrishni. What is yogmaya? You must have heard about the story of Krishna’s birth—the eighth child was exchanged and a girl was handed over to Vasudeva who carried it back to the prison. And when Kamsa tried to kill that child, the girl escaped from his hand and declared that his enemy is still alive.  Now that deityis identified with Ekanamsha. She is always shown standing between Baladeva, that is Sankarshan, and Vasudeva. So this is the iconography of Ekanamsha which has been described in Brihat Samhita and the earlier texts.


I am of the opinion that though India is full of some very ambitious monuments, the number of guilds was not very large. There were very few guilds, all of which worked at different places. So it seems that even at Ellora, while working on the Shaivite caves, they invited some guilds from Karnataka who worked on cave number 14. This exchange, I think, is very important for the further evolution of this style. Otherwise, when we talk about the evolution of a style, we talk about a rectilinear evolution, as if style   worked in isolation; this was not possible at a place like Ellora where there were so many guilds working at the same time. The interaction is inimitable. And this interaction can be seen in the later period where elements of the Chalukyan style were incorporated into the Kalachuri idiom, and Kalachuri elements were incorporated into the Chalukyan idiom. It gives rise to a very different style.


Theatricality in the Sculptures of Ellora


When it comes to the theatricality seen in the carvings at Ellora, I think it is borrowed from the Pattadakal sculptures. I observed that the Pattadakal sculptures depict an arrested movement. Like when we do an action, it is a combination of several actions. When I attack somebody with a sword, I raise my arm and then I bring it back down. So there is a fraction of inaction in between. The Pattadakal sculptures capture that particular moment. So the action is not there, the action is anticipated. And that gives a tremendous theatricality to that particular sculpture. Now the same theatricality can be seen in some of the sculptures at Ellora, particularly in cave number 15. As I mentioned, there were several guilds who worked in cave number 15.  The Pattadakal guild also worked there. The guild has worked on the Kailasanatha Temple also. Now, the similarities between the Pattadakal sculptures and the Ellora sculptures are so obvious that if you mix up these photographs, even a trained eye will find it very difficult to separate these images.


A particular person from Pattadakal, someone called Bala, travelled to Ellora and carved some of the sculptures, particularly the first two sculptures in cave number 15—the sculpture of Hiranyakashipu and Narasimha, and the sculpture of Trivikrama. It is also clear that all the sculptures in the upper register of the main shrine of Kailasanatha were carved by a single master. It could be Bala. Now, I take this theatricality to be the inauguration of Baroque in India, if we are keen on using this terminology borrowed from western art historical theory. If we really want to identify Baroque-ness in Indian sculpture, then this is the beginning of it. The flamboyance, the grandeur and the dramatic elements, all come after this particular period.  It travels to Ellora from Pattadakal and then further gets refined in Kailasanath.


The legs of Narasimha and Hiranyakashipu  in cave number 15 are interlocked with each other, which is a pose directly borrowed from a small sculpture from Virupaksha Temple, and their hands are raised. And very surprisingly, if you see their facial expressions, they are smiling at each other. They are trying to kill each other and still they are smiling. So it is almost a theatrical performance. If one keeps looking at this sculpture, one will be able to also hear the sculpture. Just add the sound of chenda to that sculpture and it becomes a performance in itself.




Kailasa is a real conundrum. There are so many different beliefs and theories about Kailasa. Some of them almost treat Kailasa as a miracle. There are some who attribute it to aliens, while others think that it must have taken at least 500 years to carve this temple. On the other hand, some scholars like Walter Spink say that it was carved over a period of 20 years. Earlier scholars believed that a trench of 100 feet depth was carved around the central shrine, the central block was separated from the living rock and then the temple was carved. Now this is simply ridiculous. If you carve out a trench of that depth, lifting stones out from that trench would be practically impossible. So the simple solution is that they must have started carving it from top to bottom, and the boulders which were carved out from around the main shrine were rolled down the slope of the mountain. That is the easiest way of doing it. If you go to any stone quarry in Rajasthan today, they use the same technique—where huge boulders are separated through drilling and then those lorry-sized boulders are rolled down the slope of the mountain. The same technique must have been used in Ellora also.


The masons and the sculptors worked hand-in-hand. The stone was scooped out and immediately the other team would start carving all the details since they don’t need scaffolding. And then the stone around the main shrine was gradually carved out and rolled down and they started working on the actual sculptures. This means that the sculptures on the top of the main shrine were the earlier sculptures, and the sculptures in the rear veranda were the latest sculptures. If a sculptor wants to carve from the living rock, he has to have enough place to sit and enough elbow space to hit the stone. It will take hours to explain the whole thing, but I can briefly say that it has been carved from top to bottom but in a diagonal fashion.


I am of the opinion that the work must have been started by Dantidurga and finished by Krishna Deva because the inscriptions attribute it to Krishna Deva. The total shrine must have taken about 30 years, but the work continued even in the period of later rulers. Some scholars are of the opinion that the Nandidevta shrine was carved after the Rashtrakuta conquest of the Triveni sangam. It seems that even in the ninth, tenth, and eleventh centuries, the work continued in this complex. The caves on the lateral side were certainly not a part of the original plan of that project and were carved much later. And that is why you find a marked similarity between the pillars of those caves and the pillars of the Jain caves. So there is enough reason, even on stylistic grounds, to believe that the work continued at least till the tenth and eleventh centuries.


Unlike Ajanta, there are not many paintings that have survived in Ellora, but it seems that Ellora was profusely painted. Jnaneshwar has said in one of his ovis—“Chi Kailasu parade thumrila”, or “Kailash is looking as if it has been smeared with mercury”. There are remains of paintings in the Kailasanath temple, particularly in the mukhamandapa. There are three layers of paintings. The lowermost or the earliest is perhaps from the Rashtrakuta period. But the Jain caves are full of paintings. Obviously, these paintings are from much later than the Kailasanath paintings, and you find the characteristics of Jain manuscript painting in some of those paintings.




There do not seem to be any attempts at vandalism at this site. This is very surprising considering Aurangzeb was responsible for pulling down so many Hindu monuments, and Khuldabad is just a stone’s throw away from Ellora and Aurangabad is not very far. Daulatabad too is not very far from Ellora, but surprisingly, there was very little deliberate vandalism in Ellora. Whatever has happened has mainly happened because of carelessness and complacency.


Last Excavations at Ellora


Obviously, the Jain caves have seen the last activity at Ellora, which means that after the downfall of Rashtrakutas nobody ventured into this arena. Rather, by that time, cave architecture in India had almost come to a close. There is a very interesting theory proposed by Dr. Jamkhedkar who is a very important scholar in this particular area. He said that the same carvers who worked on cave architecture started working on forts from the Yadava period onwards. For example, the Daulatabad Fort is a rock-cut fort. So they must have started working on the forts and the practice of cave architecture came to a close.


Marathi scholars believe that the inscription found in Jnaneshvara’s Kuta Kavya is puzzle poetry or  mysterious poetry. He says, “Chinchechya pani ek mandira bandhile, Adhi kalas mag paya re”, or “A temple was constructed on a tamarind leaf and the finial was built first and then the foundation was excavated.” This is very interesting imagery. People believe that Jnaneshvara must have visited this place. I personally think that Ellora itself is a Kuta Kavya. Ellora itself is a mysterious poem carved in stone.