On the origins of rock-cut architecture in India
The Elephanta Caves are a unique monument. Various people formed different opinions on them, especially in the initial years, and many foreign travellers have visited India just to view these caves. Visitors were often surprised by the nature of these monuments: since they are rock-cut, questions were often raised about whether India really had the technology, understanding and artistic knowledge to create such monuments. It was a surprise to many that in a country like India, which was supposed to be pagan and lacking in a higher order of intellectual, philosophical and ethical thinking, that such a monument should exist.
With regard to the origins of these rock-cut monuments, [James] Fergusson argues that the art style may have developed through contact with the Egyptians, Romans or Persians, all of who had a tradition of rock-cut architecture. Because an ambassador from a Greek king, Seleucus Nicator, visited India, and Chandragupta Maurya married the daughter of the Greek king, Fergusson argues for the possibility of this kind of architecture being derived from the west. But it is possible that the art form developed in India too on its own. We have ancient caves belonging to the Mauryan era in Barabar, in Gaya district of Bihar. Inscriptions there connect it with Ashoka. Scientific discoveries can also be made in two different parts of the world without one hearing of the work of the other. In central India, in the forests of the Vidarbha area—that is, in the districts of Chandrapur and Gadchiroli—rock-cut caves have been found. Scholars have concluded that megalithic architecture was practised here between 1000 BC to about 600 BC, much before the advent of the Mauryan empire. The settlements found here have been studied by teams from Nagpur University and the state of Maharashtra, and they have come to the conclusion that these people knew how to use iron. The iron was of very high quality, about 96 per cent pure. The megalithic people have a tradition of putting the carbon remains of their dead in such caves. Similar caves from a later period have been found even in Kerala. These caves are called 'bottle graves' as it is easy to descend into them, but very difficult to come out again.
These practices are associated with proto-historic people who did not have any script. These people predate the early historical period, when the Elephanta caves were carved. It can be argued on the basis of this evidence that this craft was not completely unknown to the Indian people. And we even have evidence now—for example, some caves which can be dated to the Satavahana period carry the traditions of the megalithic period. Even though the earliest caves were found in the Gangetic valley, not many of them were rock-cut caves. About 70 per cent of all rock-cut caves in India are located in the Deccan. This is because the rock here is more suitable for cutting.
The history of the Elephanta caves
Elephanta was named thus because of the giant elephant sculpture which was found there. It is popularly known as Gharapuri, which has its origins in a Sanskrit word—Agraharapuri. An agrahara is a place where pandits practise their spiritual and religious traditions. The revenue of that particular area is assigned to them by the king. But it was not simply an Agraharapuri as we would like to etymologise that word. It was also probably a royal seat. Harihar Thosar, who has produced a very good volume on important inscriptions in Maharashtra, says that it was probably a capital through successive periods. It was probably the capital of the Traikutakas, the Mauryas, and other dynasties that ruled over the western Konkan coast. It was a prosperous harbour and a religious place, as we know from the monuments. The archaeological remains which have been found at Mora-Bandar and Shet Bandar lead us to believe that these were probably places where jetties were located and ships docked; the descriptions in the inscriptions also testify to this.
The monuments at Elephanta
Archaeologically speaking, there are two types of monuments at Elephanta: Buddhist caves and the caves built by the followers of Pashupata cult (a Shaivite Hindu school). Some of the rock-cut caves which appear to be very simple now have been interpreted by scholars as having once been Buddhist cells. Beyond the main area where the big caves associated with Pashupata cult are located, there are remains of many structures including a stupa on a hill beyond. Visitors will be able to find the remains of the stupa, as well as other remains including a lion figure, which according to Gorakshankar goes back at least to the Kushana period (second century AD). There must have been a stupa in the early historical period and also probably a settlement of monks who were supported by the island, which was probably both a prosperous city and a royal seat. The cells as well as the stupa clearly testify to the existence of a Buddhist community there.
Dispute about the dating of the Shaivite caves at Elephanta
Side-by-side to these Buddhist remains, we have the big Shaiva caves which are the most beautiful caves that can be found in western India. Scholars have dated them to different periods. There are some scholars, such as Fergusson, who go to the extent of saying that these are Rashtrakuta caves; while [R.S.] Gupte, Brahmanand Deshpande and [K.V.] Soundara Rajan, who is a great authority on south Indian architecture, think that these are Chalukyan caves. However, [Walter] Spink, Bob Lee and Colin Smith believe that these caves might be assigned to what is called the Kalachuri period.
Now, what is the Kalachuri period? Two or three important kings followed the fall of the Vakataka dynasty—in eastern India. These were the Vishnukundins who succeeded the Satavahanas. Similarly, the Kalachuris also came up as successors of the Vakatakas. According to [V.V.] Mirashi and Spink, the Kalachuris probably had blood relations with the Vakatakas. A Vakataka princess, probably the granddaughter of the great Harisena, married one of the Kalachuri princes, argues Spink.
Those who say that the Shaiva caves at Elephanta are Chalukyan monuments will have to accept that the Chalukyan monuments from Aihole, Badami and Pattadakal are very different stylistically from the sculptures found at Elephanta. If we look at it from a finer point of view, the monuments at Aihole, Badami and Pattadakal appear to have been influenced by the Pashupata or Shaiva cults and bear similarities with the great cave at Elephanta, Cave 29 at Ellora and the caves found at Jogeshwari. However, though there may have certain similarities, scholars will have difficulty proving a common identity and patronage. It is important to note that these three monuments are very similar to each other from an iconographic point of view. The scheme of the icons and their placement on the monuments are very similar and is key to understanding the monuments. However, the monuments at Ellora, especially Caves 15 and 16, are definitely different from the other Chalukyan monuments.
The influence of the Pashupata sect on the Elephanta caves
The caves at Mandapeshwar and at Jogeshwari, Cave 29 at Ellora and the great cave at Elephanta share a similar plan, explains Stella Kramrisch. In her seminal article published in Ancient India, Kramrisch says that the architecture of these caves is totally different because the requirements of religious worship were peculiar in Elephanta, and the architecture at Elephanta needs to be studied in consideration of these peculiarities. According to Kramrisch and Collins, the architecture of the caves meets the requirements set down in the Pashupata Siddhantha, the tenets of the Pashupata sect. According to the Pashupata Sidhantha, explains Collins, it is necessary for devotees to start their spiritual progress after their initiation in a particular way. It has been told, for example, in one of the commentaries dated to the eighth century AD, that a Pashupata initiate, or sadhaka, can be married or may be a monk. At least at the initial stage, it is not necessary for him to be a monk. After his initiation, he has to spend some time in the house of his acharya [teacher], or the person who has given him the deeksha [initiation]. Then after that, he lives in a Shiva temple for some time. In the third stage, he lives in a cave. And in the last stage, he lives in a shamshan [cremation ground]. And during the course of this development, the initiate needs to meditate on the five forms of Pashupati. First of all, it is Sadyojata, followed by Vamdeva, Aghora, Tatpurusha and finally Ishana. In the last stage, while he is living in the cemetery, he has to meditate upon Ishana, whose form is usually not depicted. Whenever there is a Sadashiva shilpa [artistic representation] that seeks to depict all these five forms, only four faces will be made visible. The fifth face, which is on the top, is supposed to be present but not in a manifest form.
So if we enter the cave (main cave) from the northern side, we see a huge three-faced statue of Sadashiva which is more than 14 feet tall. The fourth face is supposed to be inherent, not to be seen, but it was supposed to be there. There are many Shiva lingas which have got all these four faces, such as the one at Chaumukhnath. But then the worshipper is aware that the fifth face of Ishana is there even if it is not to be seen. So a Pashupata sadhaka has to undergo this set of experiences and develop himself spiritually by meditating upon the five aspects of Shiva.
The different forms of Shiva
The other aspect is that the popular conception of Shiva is based on the legends from the Puranas. This form of Shiva is called Maheshwara. He is usually iconographically depicted as having 10 arms and 10 attributes. If you go to the Prince of Wales Museum (CSMVS), there is a replica of the Shiva image. Placed above it is an image of Maheshwara, who has 10 arms. Similarly, the Parel Shiva statue [sixth century], exhibited at the Prince of Wales Museum, is not properly understood. According to me, the Parel Shiva image depicts the three aspects of Shiva which were mentioned by Stella Kramrisch in her seminal article. How is he to be depicted? He is conceived in the non-manifest form (nishkala), then in the manifest and non-manifest form (sakala-nishkala), and then in a manifest form (sakala). Now, Shaiva worshippers depict this non-manifest form as a Shiva linga, which is an abstract form of Shiva. It doesn’t have any face or limbs, and therefore can be called nishkala.
The part manifest and part non-manifest form is Sadashiva, which is usually known as Maheshwara Murti, and is popularly known as Trimurti. And there are eight sculptural panels which depict the different legends of Rudra Shiva who is an ambivalent deity—ambivalent in the sense that he is malevolent as well as benign and benevolent. This is also called Sowmya Murti, Raudra Murti or Samhara Murti. The studies of George Michell, Carmel Berkson and Stella Kramrisch show us that all these opposite characteristics of Shiva have been depicted in the eight panels.
By studying these panels, one becomes aware of the different legends connected with Shiva, as known from the epics, the Ramayana and the Mahabharata, and later on also the Puranas. By saying 'later on' I am in no way giving less importance to the evidence from Puranas. It is very clear that the epics were written and finalised by the third century AD, whereas in the case of the Puranas, though their antiquity goes back sometimes to even the sixth century BC, they were not finalised before the fifth to the sixth century AD. So the tradition of Shiva’s ambivalent form—as both malevolent and benevolent—comes from our legends. Even if a person is an asura, a demon—not quite complementary to the working of the gods—he is given a boon after practising penance when Rudra is benevolent. Sometimes even Shiva himself has to suffer because of the boon he has given—for example, in the legend of Jalandhar, the asura tries to kill Shiva himself.
Shiva is shown to have this type of personality in all the legends where Rudra is married. When we think about Shiva, Parashiva, who is the first principle of creation—we cannot think about his marriage. So here he is, a Shiva who is on a lower level and is worshipped by the common people for the sake of their worldly desires and also sometimes for their spiritual progress. So these three levels of Shiva—Rudra Shiva as you can call it—that is sakala-nishkala, nishkala and sakala, all these aspects have been very intelligently depicted in the iconographic scheme seen at Elephanta.
There are two entrances to this cave. When one enters from the east, one can see the nishkala Shiva, the higher aspect. The shrine has four doors—sarvatobhadra—so it can be seen from all sides. To see the lower aspect of Shiva, that is sakala-nishkala, the visitor has to enter from the north door. And therefore, to accommodate all the requirements laid down in the Pashupata Shastra, these caves were made.
When an initiate has to first live in the house of his acharya, it is possible that there might have been an ashram (monastery), probably built in brick, where the acharya may have been living. When it comes to cave temples, as the name suggests, they are both caves as well as temples. The sadhaka can come there. As it is a cave, he remains there to practise the next stage of his penance. So this way, cave temples were very suitable for the practises of the Pashupata sadhaka. In Ellora, for example, similar caves where sadhakas probably lived have been found. There is a small group of caves carved into the side of the Mhaismal Plateau, which are referred to as the Ganesh Lena caves. There is a stream here during the rainy season, and the caves have been carved on both sides of that stream. The stream falls into what is called the 'Sita ki Nahani', a pond near another cave which is very much similar to the cave at Elephanta, that is Cave 29.
And there is another Shaiva settlement or teertha which is very important, and that is Grishneshwar, one of the 12 jyotirlingas. In those small caves in the Ganesh Leni group, many such Sadashiva images have been detected. Pashupata sadhakas probably lived in these caves and meditated upon the different aspects of Shiva.
The sculptures at Elephanta
The sculptures at Elephanta have a great similarity with the sculptures that have been found at Shamlaji and Roda [Gujarat]. In fact, U.P. Shah, a great scholar in this field, has written a small monograph on these sculptures. Another scholar from the United States, Sara Schastok, has worked on the Shamlaji sculptures and has come to the conclusion that these are probably from the Kalachuri period (circa 550 AD). They are near their capital at Mahishmati on Narmada.
This is probably the peak of Indian sculpture. Of course, the Gupta sculptures which came much earlier are some of the best in India; in fact, it is called the classical sculpture of India. The Elephanta sculptures are from the post-Gupta period, but at the same time, they are very unique. And as U.P. Shah states, this was part of the western Indian school of sculpture. This has been mentioned by Taranatha in his History of Buddhism. And this school of sculpture, to some extent, has been influenced by the Gandhara sculptures.
Shilpa texts such as the Chitrasutra lay down rules for the art of painting and sculpture. It mentions the various proportions which images of different types of celestial beings should take. So all the celestial beings are not of the same proportions. Some are in nava-tala, some are in dasha-tala, while others are in ashta-tala, sapta-tala and even shat-tala. While the ganas are smaller in stature, the others are bigger. The consort of the god is almost, but not quite, the height of nava-tala. And when the god is to be shown as supreme, the sculpture is probably dasha-tala.
Texts have provided instructions regarding the various postures which an image should take. So there are six types of postures which have been mentioned—looking forward, looking backward, facing sideways, and also two more which are in between. If you carefully see the images at Elephanta, you will see that the sculptors were aware of these conventions. The texts were formulated in the Gupta period, and their best results can be found at Elephanta. So they not only meet the different requirements which are given in the classical texts, but perfect it. Fergusson has a wonderful observation to make: he says that the sculpture at Borobudur in Southeast Asia is a unique monument—it is three storeys tall and has no equal in Buddhist art anywhere in the world. But if you compare this sculpture with others, you will arrive at one conclusion—the artist must have travelled there from western India, from Ellora. He very clearly says that the type of sculpture found at Ellora and the type of architecture found at Borobudur and other places share common origins and roots located in western India.
Elephanta and Ellora are connected. As Prof. Spink explains, the first experiments were made at Jogeshwari, where a sub-terranean cave was created in order to depict the three types or three levels of images, as per the Pashupata doctrine. In the middle, there is one sarvatobhadra shrine which has got a Shiva linga. The ideas regarding the nishkala and sakala-nishkala aspects are clear, but at the same time, the placement is not very clear. At the same time, the type of Sadashiva image which you find at Elephanta is absent there. But panels have been found at Jogeshwari that are very similar to the eight panels at Elephanta. So he thinks that the theme was conceived at Jogeshwari. In my own article in collaboration with M.N. Deshpande, we come to the same conclusion: the caves at Jogeshwari and those at Lonad are interconnected and are Buddhist caves. The pillar order and the types of sculptures found at Lonad and Jogeshwari show they are of the same period roughly, about 480 AD.
Rivalry between the Buddhists and Pashupatas
There is one more factor which is to be emphasised, and that is my own personal observation—my hypothesis is that there was something like a rivalry between the Buddhists and the Pashupatas. It is not a coincidence.Near Shamlaji there is a Buddhist stupa called Dev Ni Mori, meaning a monument which is in front of the god. The god is Shamlaji and in front is the stupa. The stupa was excavated by Baroda University, and has been dated to the third century AD. The sculptures at Shamlaji and Roda are from the Kalachuri period. We find that the Buddhist monuments here date back to the second or third century AD and the presence of the Pashupatas was also there. If you visit Aihole, Badami or Pattadakal, you will find that near the Meguti temple, there is a small cave which we know to be Buddhist because of the depiction of Jatakas and of Avalokitesvara. But again, Shakta caves, Shaiva caves, Pashupata caves and Pashupata temples are also located there. Then, if you go to Andhra Pradesh, you will find that on one side there is a Buddhist monument and immediately on the other there is a Shakta monument. This rivalry can be found in Orissa to Central Asia. In the paintings in the Buddhist caves in Kizil (China), some ascetics have been depicted. In pursuit of their rivals, the Pashupatas travelled up to Khotan and Central Asia. My hypothesis is that these Buddhist monks remained confined to small pockets in the Deccan and did not increase in great number, in the number of caves constructed, or in terms of followers in the fifth century AD because of this rivalry. The Buddhist monks were losing their base among the masses. Cults like Lajja Gauri were assimilated into the Pashupatas. In Cave 21 at Ellora, Lajja Gauri has been depicted. By this period of time, Buddhism had become Sanskritised. At Ajanta, the followers of Buddhism were depicted as either big ministers or tradesmen—Buddhism appears to have had lost its mass base. The same thing happened in the Vidarbha area—the tribals there accepted the cult of Narasimha, that is, the Bhagavata cult. There is evidence showing the presence of Shaivism at places like Deulvada. So Buddhism lost its ground in Maharashtra, and all these people got converted to either Shaivism or Vaishnavism.
The same thing happened in the later period. A last pocket of Vajrayana Buddhism can be seen at Panhalekaji. M.N. Deshpande substantially proves that the settlement at Panhalekaji, which was a Vajrayana Buddhist settlement, was converted to a Natha Panthi. So this process unfolded over the same route—the people in Andhra Pradesh became Shaktas or Vaishnavas, the Shaivas or Vaishnavas. If you go to Orissa, traces of the same process is observed. And after 1000 AD, when they were attacked by the Turks from Central Asia, the monasteries were totally lost. There were only two options: either become Baul sanyasis, who were affiliated to Vaishnavism, or become Shaivites. From that point of view, if you look at Cave 15 at Ellora, it began as a Buddhist cave but was virtually converted by the great warrior, Dantidurga, who was a vassal of the Chalukya sovereigns but later declared his independence. And then comes the great monument at Ellora, but it is preceded by Caves 21, 22 and 29, which are of the Pashupata lineage.
So Elephanta is also very crucial—not simply important—from this point of view of understanding the ups and downs in the histories of the different cults and religious traditions. I never call them religions. There were three religious traditions in India—Jainism, Buddhism and Hinduism, which generally included Shaivism, Vaishnavism and other groups. So it is very interesting to study this give-and-take relationship, the ascension and recession of all the groups, through the help of monuments; Elephanta is one of the best monuments through which these trends can be understood.
Curzon, whom we many times curse for his division of Bengal, was the person who was responsible for establishing the Department of Archaeology in India. And that is why we can proudly say that India is a unique country. We don’t know much about China, though only China can vie with us with regard to antiquity and the abundance of monuments. We are the only country where, according to my earlier estimates, there are 5,000 protected monuments. Of these, 3,500 are protected by the Archaeological Survey of India and 1,500 by the states. We will have to revise these figures because when I started my career as the first Director of Archaeology in Maharashtra, 168 monuments were therein the state. Today there are about 300 monuments that have been declared as protected. Possibly overall about 6,000–7,000 monuments are now declared as protected.
There were many good conservationists like M.N. Deshpande who reinstated the seven echo effect of Gol Gumbaz. The seven echo effect of the Gol Gumbaz was lost by the '50s. He was the superintending archaeologist of the southwestern circle, that is, the Aurangabad circle. He was a field archaeologist, an excavator and an art historian, but he was not trained as a conservationist. With the help of structural engineers, he restored it. People at all levels—right from conservation assistants to the director of conservation—everybody has to undergo training and learn things because so many things are happening. There are so many institutions of conservation, for example, the institute of conservation at Lucknow. Instead of criticising various organisations or the state, we should increase awareness at different levels—I take the media also as a part of that—among policymakers at the state as well as central level. One has to compliment Manmohan Singh ji. He included a provision in the central budget for the maintenance of state monuments. Earlier, it fell within the purview of the states only.
Not only that, you have to take into consideration the zila parishads (district councils) as well. They have to provide for the conservation and maintenance of monuments. Zila parishads should be party through provisions in the law. So all these factors have to be taken into consideration. For me, I am hopeful. Whatever defects or problems are there, we ourselves are responsible for it; we should introspect and go ahead. That is my position.