Our perspective of early Indian trade has taken a few major leaps in the last two decades or so. One of these has been a rigorous delineation of the external trade of the Indus civilization: the minutiae of relevant artefacts, their contexts and chronology, the routes involved, the problem of ‘Overland trade versus maritime trade’, and the context of this trade in the overall pattern of historic and still continuing trade over the entire region. It was important that these issues were sorted out from the Indian point of view before the significance of this trade could be properly evaluated in the Harappan context. Regarding India’s external trade during the historic period, the primary emphasis has so far been on the textual sources, both of India and the Mediterranean world, and on the Roman coin finds. In recent years, however, there has been an increasing attempt to explore different types of non-numismatic data beyond these traditional sources. Occasionally, such attempts have been marred by more enthusiasm than is necessary, but we are now better informed about the different types of Roman artefacts in India as trade goods. By the time of Periplus of the Erythraean Sea (c. middle of the first century AD), the Indian coastline and ports and the products exported and imported through them were well known to the Romans, and it is not unreasonable to take the beginning of this knowledge and contact a couple of centuries earlier. The peak point should be the first two centuries AD when there are a large quantity of Roman or Rome-inspired antiquities in different parts of India, especially in west, south and east India. This is the time when one finds potters in the Bhagirathi delta turning out terracotta heads of males with laurel-leaf wreaths around their heads. The idea that the Bengal coast did not play a role in the Roman trade flies in the face of the archaeological data, inclusive of a complete Roman amphora, from this region. Further, the finds of innumerable Roman coins of gold, silver and even copper can now be better appreciated historically, thanks to D.W. MacDowall’s clear assessment. He refers to the currency reforms undertaken by the Roman emperors Nero in AD 64 and Vespasian in AD 70 and argues that
Fig. 56 Complete Roman amphora from the Contai area of coastal West Bengal (reported first by Chakrabarti et al., 1994)
… most of the gold, silver and later Roman copper coins found in India were exported from the Empire after major currency reforms. It was a direct result of these reforms that older coins still in circulation from earlier issues came to have a significantly higher metal content than the same denominations currently being issued, even though both had the same nominal value in the Roman monetary system. Following the principles of Gresham’s law, the old coins of higher intrinsic value would in due course be withdrawn by the mint for recoinage, hoarded, melted down by jewellers and others for bullion, jewellery, plate, etc. or exported where the recipient valued it for its bullion.
… we must now revise substantially the dating of the periods when these categories of coins in silver, gold and copper were exported to India—the aurei and denarii of the Julio-Claudians struck before Nero’s reform in AD 64 to the Flavian period (AD 70–98) and Trajan (AD 98–117); Republican denarii to the principates of Trajan (AD 98–117) and Hadrian (AD 117–138); aurei struck after Nero’s reform mostly to the middle and later second century AD; and later Roman copper coins to the fifth century AD. It will be seen immediately that these periods do not correspond either with the dating evidence of trade contacts derived from the amphorae and sigillata finds or with the literary evidence of trade contacts and incidence of embassies.
This is hardly surprising. The dating of the amphorae finds properly indicate the period of trade in wine, oil and other products for which amphorae were employed; the dating of the sigillata indicates the periods when the rare sets of dinner plates were imported from the west; and the revised chronology of the export of these coinages indicates the periods when Roman traders found it more profitable to export coinage (much of it worn and demonetized) instead of metal in other forms. Pausanias, writing in the mid second century AD, describes the non-monetary character of trade with India, confirming what one would have expected that trade was by barter and did not use money as a medium of exchange even when coins were being traded.
In short, the trade in coins was simply part of the much wider trade in metals, at limited periods when traders found that the use of gold coins, exceptionally, the most advantageous way for them to supply gold, silver and copper.10
This is a very important step in understanding the significance (or, in a sense, the lack of it) of the distribution of Roman coins in India. If these are parts of only a trade in metals, it is not necessary that the find of a Roman coin somewhere in the subcontinent should indicate direct or indirect contact of that area with the Roman world. The evidence of such contacts, which no doubt existed in great measure, should be sought in routine archaeological materials.
Archaeologically, the understanding of Roman trade cannot be said to have significantly advanced even after the recent excavations at Arikamedu and Kaveripattinam, both on the Tamil Nadu coast. However, a very large part of ancient occupation at Arikamedu is still undisturbed and lies unexcavated.
The study of India’s historic trade with south-east Asia has, regrettably, not gone much beyond the traditional sources of literature and inscriptions. Despite attempts by scholars, the archaeological basis of this trade is not well defined and is limited principally to the finds of etched carnelian beads, rouletted ware, etc. This sharply contrasts with the intensity of contact between these two regions, that we know of from art, architecture and inscriptions. However, the find of etched camelian beads, typical of the early historic Indian assemblages, in some quantity in the fourth century BC context at the western Thailand site of Ban Don Ta Phet is significant and has at least offered an early chronological marker of India–south-east Asia contact.
The overseas trade network is no doubt exciting and conjures up a rather romantic vision of a world operating, so to speak, on its own in the high seas—the Indian Ocean in this case. The prospect of ‘Indian’ coastal trade, i.e. the trade pursued by Indian coastal vessels and merchants in different sections of the Indian coast, is, of course, far less romantic but could be equally significant in relation to the internal trade network of the subcontinent. This aspect has more or less been ignored in studies of ancient Indian trade. V.D. Gogte’s premise that rouletted ware was made of the clay of the lower Ganga plain raises important possibilities.
A major breakthrough has been achieved in recent years by a careful study of the distribution of various raw materials at protohistoric and some early historic Indian sites, the possible source/sources of these raw materials on the basis of the geological distributions of these materials, and the possible lines of movement of such materials both in the light of archaeological site distributions and the historical and ethnographic data on these lines of movement. What has emerged is a very extensive series of geographically well-defined networks of contact all over the subcontinent right from the protohistoric period to about 200 BC. The early historic pattern is strongly rooted in protohistory. By c. 200 BC and possibly considerably earlier, the whole of the subcontinent was within an integrated communication chain; there was no difficulty at all in moving down from Taxila to the deep south or from the sub-Himalayan tract to the Bhagirathi delta. The north-west had, as we have emphasized time and again, its network of communications with Afghanistan, central Asia, Iran and beyond. These routes are geographically constricted and well defined.
The locations of major sites or inscriptions in certain sectors also have their own geographical implications. For instance, the location of the Asokan edicts at Shabazgarhi and Manserah in the north-west is in an area where the famous Silk Route or the Karakoram route begins with central Asia as the objective. This suggests that in the Mauryan scheme of things central Asia and/or China were significant. The Mauryans must have had some knowledge of China. Similarly, the location of two Asokan pillars at Rampurwa near the Bhikna Tori pass on the Nepal border implies that the Kathmandu valley, the road to which went through this pass, must have been important in the Asokan administration. Again, the location of a Mauryan provincial capital at ancient Pundranagara or modern Mahasthangarh in Bagura district of Bangladesh underlines the possibility of the Brahmaputra valley being within the Mauryan kingdom, although there is as yet no direct evidence for it. The location of the early historic site of Wari Bateshwar in the Meghna delta of Bangladesh suggests that the ancient Samatata tract, in which this site is located, was a part of the Mauryan kingdom. This further means that the Mauryans would have had no difficulty in extending their power up to the Chittagong coast. The point is that the geographical nuances of archaeological locations have to be worked out area by area.
Of the major axes, the first is from the north-west to the Bengal delta, with many offshoots all along the way. One of its major offshoots goes in the direction of Sind from about the region of Taxila. From Sind, again, one can easily move to Kutch and the mainland Gujarat coast which gives access to central India and the Deccan, or, from Sind one can move towards Haryana-Delhi through the Hakra-Ghaggar system. Once the trunk-route from the north-west reaches Delhi-Mathura, it is easy to branch off in the direction of Rajasthan and Gujarat. Through the Malwa plateau adjacent to Rajasthan one moves towards the Deccan and west India. Moving further towards the east, a major line takes off from the area of Allahabad towards central India through the Baghelkhand region, while another line goes towards central India and the Deccan from the Banaras sector. From the Banaras sector it was also possible to enter the hills, reach Garhwa on the Son at the fringe of Palamau and move through the hills and some open valleys towards Keonjhar and the Orissan coast. At some point along this line one could break off and leave in the direction of modern West Bengal to reach the ancient port of Tamralipta. The trunk route to the Bengal delta followed the southern bank of the Ganga in Bihar and went through the narrow strip between the Rajmahal hills and the river to reach the delta. From this delta one could go to Orissa on the one hand and Assam or the Brahmaputra valley on the other, without running against any major geographical obstacle. The route to Arakan was open too from the Samatata area (roughly, the area of modern Comilla). From the Brahmaputra valley there could be several routes in the direction of Bhamo in northern Myanmar with south China in mind, whereas there could be communication lines with Tibet through the Chumbi valley, Bhutan and perhaps (he Sadiya region. If one targeted the Chhattisgarh sector of central India from the Allahabad-Banaras sector, it was possible to reach it by travelling across the central Indian hills and then moving from Chhattisgarh to north Orissa or Vidarbha. The routes in the interior of the Deccan and the west coast are clearly marked and in any case dotted with archaeological sites and monuments. From this sector there was no difficulty in reaching Hyderabad and eventually the Krishna-Godavari delta in Andhra, from where the route to the deep south was entirely open along the Tamil Nadu coast. There was another route to southern Deccan through western Andhra, and once in southern Deccan (part of modern Karnataka) the routes further south were easy. The Kerala coast was accessible through the Palghat gap.
It is along such lines that the tapestry of ancient Indian trade and trade routes (and some history as well) has been woven. This tapestry is, in the long run, far more exciting than following vessels across the sea and enumerating ports and coastlines. In a way, this tapestry is still in the field before us and all that one has to do is to go out and look for it. If the Mauryan king Asoka visited Lumbini, the birthplace of the Buddha, travelling from his capital Pataliputra, he must have travelled in the direction of Arrah and Buxar, crossed the Ganga at or near Buxar for Ballia, possibly stayed at the fortified early historic site of Pukka Kot in the Balliah area and gone in the direction of Khairadih, another fortified city site, on the bank of the Sarayu. He must have crossed the river at Khairadih, because on the other bank there is an uninscribed and broken shaft of a Mauryan column. From Khairadih he possibly passed through Sohgaura and Kopia to reach Basi, still the most important approach point for the Lumbini-Kapilavastu area. There was a major route also along the northern bank of the Ganga, which sent offshoots towards the sub-Himalayan tract in the north frequently along the way. The archaeological details of the history of all these internal routes of the subcontinent are known only in patches. Much remains to be done.
The journey from the subcontinent’s palaeolithic beginnings to its early historic foundations is no doubt very varied, and, of course, very long. Also, much of it can be seen only in patches. If this journey can be compared to travelling along the bank of a river, we have tried to follow its course in this volume as much as we can, sometimes moving close to the main flow and sometimes being forced away from it. This flow has no doubt meandered all over the land, taking many tributaries along the way. The gurgling of the regional tributaries does not, however, detract from the quiet and majestic flow of the main stream.