Located along the western coast of India are a series of caves, some of them richly decorated, that were cut into living rock (Fig. 1). The primary function of these caves was to serve as Buddhist chapels and monasteries. In this article, we will not only describe these monuments, but also investigate a second aspect of their construction and use: their role in the flourishing trade between India and the west that took place between ca. 100 B.C. and ca. A.D. 300
Construction and Design of the Caves
A single Early Buddhist monastery in the Deccan region (Figs. 2, 3) of western India might consist of several hundred separate ‘caves which were often decorated with sculpture and finished in plaster that is carved and painted. Individual caves fell into two basic types: the caitya or chapel for congregational worship (Fig. 4) and the vihara or residential hall.
A typical chapel was apsidal in plan and divided into a nave and side aisles by rows of pillars. The arched facade of this type of cave was carved of stone, and the broad entrance was filled by a wooden screen with a doorway. In the cave at Karle, the vaulted ceiling of the nave still retains its original 2000year-old wooden ribbing (Fig. 5). The object of worship in these early chapels was a stupa at the far end of the cave (Fig. 6). In the early (Hinayana) doctrine, the Buddha was not deified and was worshiped only through symbols such as the stupa, which was an emblem of the Buddha’s death and his release from earthly life. The residential hall was generally quadrangular in plan, with cells cut into three sides (Fig. 7). Each cell contained a raised rock-cut bed, sometimes complete with a rock-cut pillow. These monastic complexes had cisterns hewn out of rock that collected water from underground springs or from rain percolating through fissures; many of these are still in use over two millennia later.
It is intriguing to speculate as to how the resources for such monumental projects were mustered. A large number of individuals clearly contributed funds (see below). At Carle, for example, each pillar in the chapel bears the name of a different donor. But in spite of the number of separate donations, the cave presents a unified scheme, and it is unlikely that the pillars were carved haphazardly as donations.
The establishment and growth of monasteries in the western Deccan took place in two distinct phases. The earliest activity extended from about 100 to 20 B.C. and was followed by a lull of about seventy years. The second phase thus commenced around A.D. 50 and continued until the beginning of the 3rd century A.D. This break in the cutting of caves was also the period during which the fortunes of the ruling Satavahana Dynasty were at their lowest. It would seem that although large donations were made by lay devotees, monastic expansion depended to some extent on political stability.
It has been suggested that a system of work prevailed in ancient India which was very similar to that used until recently in Buddhist Tibet. When it had been decided that a certain edifice was to be decorated or a piece of sculpture executed, artists were gathered from the leading religious institutions of the day. They were retained at the new site, probably a monastery, until their commission was completed: Then these artists either returned to their monastic institution or traveled to another religious establishment that required their artistic services (Dehejia 1972:139).
The unfinished caves at various sites give us a fairly good idea of the actual process of excavation. Presumably an outline was first marked out on the rock. The cutting proceeded from the top downwards, and several activities such as rough cutting, sculpting, and polishing were undertaken simultaneously. An inscription at Canheri refers to several categories of craftsmen—stone masons, polishers, carpenters, and the like. The navakarmika or monk in charge of new works would probably have been responsible for formulating the plan for the religious facilities and the iconography of the monument.
As compared to other contemporary sites (for example, Sanchi or Amaravati), the Early Buddhist chapels and residence halls of the western Deccan are marked by their simplicity, even austerity. There are, for example, few representations of Jataka stories, early folk tales of the Buddhist tradition. Popular images include rather restrained motifs like dwarf yaksas, or couples and winged animals (Figs. 9, 10). Another favored ornamentation was the sculpting of pillar capitals with images of animals with riders (Fig. 11).
The couple or mithuna representation is a common motif at several early Buddhist sites on the Deccan. This motif develops from rather simple, graceless figures to later examples which have a suppleness of body and grace of movement that reaches a climax in the chapel at Karle (Fig. 10). On the whole these early human forms depict broad shouldered athletic men and curvaceous women in a variety of moods and postures. Unlike the couples of later periods, however, these images rarely have an explicitly erotic aspect.
The profusion of paintings in later monastic establishments at places like Ajanta (Fig. 6) has raised the question of whether the early Deccan caves might not have been similarly decorated. Indeed, traces of early paintings do survive in a few caves at the site of Junnar and in the earliest caves at Ajanta itself. On the whole, however, it seems that the early monastic caves were not painted and that it was only after the 2nd century A.D. that the practice of simply smoothing the walls gave way to plastering and painting them (Nagaraju 1981:66).
Trade and Travel on the Deccan Plateau
The western Deccan plateau, the land behind Bombay, was quite different in antiquity from the open savanna landscape familiar to modern travelers. Around 500 B.C. it was still densely forested, and contemporary literature in Pali and Sanskrit refers to the Deccan as the abode of several tribes. The narrow coastal plain, today known as the Conkan, probably formed a somewhat distinct region from the interior Deccan. Settlements on the Konkaii participated in a looping maritime trade with distant centers along the Indian coastline and were also ports of call for Arab sailors and traders.
From about the 3rd century B.C. onwards, these regions changed rapidly. Expansion by the Mauryans, ancient India’s first great empire centered in modern Bihar, was motivated by a search for gold and semi-precious stones. This exploration and expansion led to the development of land and sea routes linking the peninsula to the cities of north India. The western Deccan thus became a sort of corridor in the ensuing communications system. The distribution pattern of a series of royal edicts of Cing Asoka, the great Mauryan king, may indicate the most important routes along which traffic moved (Fig. 2). There is also a cluster of Mauryan inscriptions at locations in the gold-rich area of peninsular India, as well as a sprinkling along the ports of the east and west coasts.
In the 1st century B.C., the volume of long-distance trade to the west was greatly expanded as a result of an increased demand for Indian products in Roman markets. The Emperor Augustus had established peace in the Roman Empire after his defeat of the fleet of Antony and Cleopatra in the battle of Actium (31 B.C.). Roman frontiers had expanded, and new markets were open for Oriental goods. One consequence was a greater involvement of Roman subjects in the maritime traffic to India, an enterprise that had formerly been controlled by middlemen from the Arabian peninsula.
The cities of Alexandria, Coptus, and Palmyra now became important markets and distribution centers for trade with South and Southeast Asia. Enclaves of (ethnic) Greek, Indian, and Arab merchants grew up along the coast of the Indian Ocean. Places like the island of Socotra at the entrance to the Red Sea became important trading centers. Arab and Indian merchants faced increasing’ competition from their Roman and Greek counterparts, especially those based in Egypt and Palmyra.
A contemporary mariner’s account of Roman trade in this period survives in the form of a handbook titled the Perilous of the Erythraean Sea. This document was written in the 1st century A.D. by a Greek sea captain resident in Alexandria. It tells the reader about sailing conditions in the Indian Ocean: where to land, what ports were available for safe haven, what goods could be procured there, and what might be traded in return. It warns of hostile tribes, dangerous reefs, rocks and shoals.
It is a remarkable work in detail and geographic scope and entries on India are especially important to our understanding of Roman sea contact with the subcontinent. For example, the Periplus mentions the two principal market towns in the Deccan: Paethana (Paithan in the distric of Aurangabad) and Tagara (Ter in the district Osmanabad). It tells us that merchandise was brought to Barygaza (the modern city of Broach) after traversing vast distances without roads. The goods bound for Rome included varieties of wood, such as ebony, teak, and sandalwood; semi-precious stones; ivory; cloth; aromatics; spices; and dyes such as indigo. In return, Indian merchants obtained wine, dates, glass, tin, coral, vessels of silver and gold, singing boys, and maidens, along with Roman coinage.
An Indian ivory figurine found at Pompeii, the prosperous Roman city destroyed by the eruption of Mount Vesuvius (A. D. 79), apparently came from the ports of western India (Fig. 12). Conversely, objects of Roman manufacture, including fragments of amphorae and glass bottles, have often been recovered during archaeological excavations in the western Deccan. An extraordinary discovery was that of a hoard of Roman metal objects from Brahmapuri (a suburb of Kolhapur) in 1944. This hoard contained, among other things, an exquisite statuette of Poseidon, two intricately carved handles of wine jugs, a candlestick base, and a mirror—all of Roman origin.
Monasteries on the Inland Routes
The advent of long-distance maritime trade with Rome took place at the same time as the cutting of the early Buddhist caves in the slopes of the Western Ghats or Sahyadris mountains along the west coast. The location of these monastic sites on land routes, and at the heads of important passes connecting the coastal ports with inland centers, tells u.s that there is a causal link between trading enterprises and the new monasteries.
A majority of the early caves are located in the geographical transition zone, where the gentle gradient of the western side of the mountains ends and a steep descent to the ports begins. This is also the place where a change would have been made in the mode of transportation, probably from wagons to pack animals or to human carriers for the journey down the steep slopes to the Conkan ports.
The Sahyadri range dominates the rugged terrain of western India and forms a barrier between the Deccan plateau and the narrow coastal plain. It can only be crossed at specific points where passes provide a break. Modern advances in communications have not diminished the importance of these passes, and even today national highways and railway lines follow the ancient tracks. Small wonder that they were chosen by the earliest Buddhists as sites for their monasteries. The pass at Thalghat connected Nasik cave with the coast, while Gunnar cave was situated at the head of the Nanaghat Pass; the caves at Shelarwadi, Karle, Bhaja, and Bedsa were in the vicinity of the Bhorghat pass and the descent to the coast near Thana and Bombay.
A series of rock-cut caves along the west coast are also in strategic positions, lying in close proximity to ports, from Chaul to Bankot, Rajapur, and Malvan. The largest monastic establishment is located at Canheri near Bombay, suggesting the enduring importance of this port. The coastline around Bombay has altered considerably with the development of the modern city, making it difficult to appreciate the strategic location of Buddhist caves in this area. There is, however, direct evidence for the importance of this area in the western sea trade in the form of a wharf built on one of the islands near Canheri during the early historic period. On Elephanta Island inside the modern port of Bombay, pieces of Roman amphorae and coins have been recovered; this island is also the site of a beautiful set of later caves, this time dedicated to the Hindu god Shiva (Fig. 13).
The way in which monasteries participated in the exchange of goods is also of interest. D. D. Cosambi, a noted historian of ancient India, has suggested that monastic establishments may have actively participated in trading ventures by advancing money to merchants, or by selling the surplus produce from their fields to them (1955:60-61).
The resources necessary, whether land or money, were obtained through donations. Inscriptions preserved in the early Buddhist caves document the fact that entire villages were given to monasteries for their support. Tied to the villages were large tracts of agricultural land which would have produced far more food than was needed for the associated monasteries. A majority of the inscriptions recording donations of land are at Gunnar, situated at the head of the Nanaghat Pass. This is a broad fertile valley on the right bank of the Kukdi River, a tributary of the Ghod. Investment in land in this region was thus a natural way to support the monastic establishment. By contrast, in an area that was not easily developed for agriculture, such as that surrounding the coastal site of Canheri, donations took the form of perpetual endowments of money, ranging from 200 to 1600 kahapanas.
An analytical study of royal inscriptions granting land to monastic establishments shows that these are concentrated at Nasik and Canle. Both of these centers were in critical, strategic locations, essential for control over regional trading routes from the interior to the coast. They were also located on “frontiers” between ruling dynasties and were important for maintaining a balance of power.
There is considerable debate among scholars concerning the degree to which the Buddhist monks were themselves directly involved in commerce. Inscriptions at man Buddhist establishments refer to endowments of money made by ascetics and nuns themselves and also to gifts for the cutting of caves and cisterns. This is certainly a violation of the early tenets of Buddhism, which stated that monks and nuns were neither to accept nor to own money. The gulf between theory and practice continued to grow with increasing prosperity, until in the 5th century it became a practice for Buddhist establishments to employ a body of officials to administer the monasteries and their wealth.
Even within the early Buddhist period there were a variety of sects with different attitudes toward money. Some seem to have been deeply involved in commerce, while others distanced themselves from it. Still others seem to have fulfilled material ambitions while maintaining spiritual principles. D. D. Cosambi cites practices in the Buddhist monasteries of China, which borrowed at least some of their basic organization from India. Monks scrupulously observed the letter of the law: forbidden to touch money, they used a piece of cloth as a kind of spiritual insulation.
The Role of the Monasteries within Society
In addition to their commercial function, the monastic establishments would have been important to the community as a whole as disseminators of information, in addition to serving as a nucleus of the social life within their region. The population of western India at this time was heterogeneous and included traders from many regions of the subcontinent, including merchants from cities and towns, and foreigners, as well as local tribal peoples. The organization of trade was an immense task, especially given the limited means of communication. Cooperation between the diverse elements of the system—producers, traders, transport specialists, money lenders, warehouse operators—meant that it was important for there to be nodes where reliable information on scheduling, market conditions, and the like would be available. Monasteries would also have been safe havens for travelers, a kind of predecessor to the caravanserai of later history.
The esteem in which monastic institutions were held is reflected in the lists of their donors. Although a few inscriptions record gifts from the ruling dynasties of the time (the Satavahanas and the Cshatrapas), a majority of the donations came from ordinary devotees. These included merchants and traders, weavers and oil-pressers, jewelers, iron-mongers, gardeners, fishermen, guilds of bamboo workers, and people known as “yananas.”
Originally the term yavana denoted an Ionian Greek, an ethnic group that entered the ancient Indian scene prominently in 326 B.C., with the invasion of Alexander the Great. By the beginning of the Christian era the term came to be used indiscriminately for all foreigners. The Sangam literature in Tamil Nadu, which is contemporary with the caves under discussion, refers to colonies of yauanas in south India. It is not certain if similar colonies existed in other regions of ancient India as well. In western India, a majority of the donations by yavanas occur at Carle, and some of the donors are known to have adopted Buddhist names.
Thus the early Buddhist monasteries of the Deccan, seemingly remote from the world, were not only integrated into the social and economic system of peninsular India, but were closely tied to the west through both trade and the piety of individuals.