was accompanied by a demographic shift from upland settlements to the alluvial and easily watered soils of the deltas. ‘In the Chola country, watered by the Kaveri, it was said that the space in which an elephant could lie down produced enough rice to feed seven’ (people, presumably, rather than elephants).10 But here a surplus laboriously realised from agriculture, and then partially squandered on oblations designed to ensure its repetition, was second-best to the surplus on offer from the export of marine and forest produce (especially pearls and pepper) and the re-export of luxury items from further afield. Such options, not open to the Gangetic states, propelled the peninsula from Stone Age to statehood in record time.
Before the first century BC the southern extremity of the subcontinent scarcely features in India’s history. Today’s southern states – Karnataka, Andhra Pradesh, Tamil Nadu and Kerala – correspond to the languages spoken in each, respectively Kannada, Telugu, Tamil and Malayalam. All belong to the Dravidian family, which is quite distinct from the Indo-Aryan whence Sanskrit and most of north India’s contemporary languages derive.
Dravidian-speakers are thought to have preceded Indo-Aryan-speakers in the subcontinent. It has yet to be proved that the Harappans’ language was some form of Dravidian, but the survival of a pocket of proto-Dravidian-speakers in Baluchistan, the Pakistan province which borders with Iran, does suggest that the language was in use west of the Indus and could have emanated from there. It also once enjoyed a wide currency in Gujarat and Maharashtra, though whether in the course of a ‘descent’ to the south, or an ‘ascent’ from it, is uncertain. By the mid-first millennium BC, it was certainly well established in the south, perhaps as a result of dissemination by Dravidian-speakers possessed of the horses and iron weaponry usually associated with the Sanskritic arya, or possibly very much earlier as the language of those responsible for the megalithic sites found in upland regions of the peninsular interior.
The four Dravidian languages must have developed from proto-Dravidian at an early stage since they were already distinct from one another in prehistoric times. Each, too, was already confined to the region represented by today’s states. In fact the continuity of such geo-linguistic entities is the outstanding feature of south Indian history. Here, unusually, definable linguistic units seem to predate the states into which they would become integrated.
Megasthenes in around 300 BC knew of the Pandya kingdom; then, as subsequently, it occupied ‘the portion of India which lies southward and extends to the sea’; and it had 365 villages, a not incidental number in that each was expected to supply the needs of the royal household for one day in the year. Ashoka was even better informed. In the Major Rock Edicts he lists his southern neighbours as the Cholas and Pandyas (respectively the northern and southern Tamil-speaking peoples), the Satiyaputras (whose identity is disputed), the Keralaputras (or the Malayalam-speakers of Kerala), and the people of Sri Lanka. Although none formed part of the Mauryan empire, all, according to the Beloved of the Gods, acknowledged the superiority of dhamma and had imitated the Ashokan provision of roadside shade trees and of medical care for men and animals.
Later the all-conquering Kharavela, king of Kalinga (Orissa), also noticed the southern kingdoms. In his one extant inscription he typically pretends to have defeated a confederacy of Tamil states and to have acquired a large quantity of pearls from the Pandyas. Pearls and shells, along with the fine cottons of Madurai, the Pandya capital, are also mentioned in the Arthasastra. There, in a discussion on how to maximise the state’s revenue, Kautilya’s mentor rashly suggests that north India’s most valuable trade is that with central Asia. The know-all brahman puts him right in no uncertain terms: the trade via the Daksinapatha (the ‘Southern Route’) is the more valuable and, besides, the route is very much safer. Thus trade with the south, albeit in prestige goods, was well-established in Mauryan times; and by way of the secure – and, no doubt, well-shaded – Daksinapatha prestigious ideas also travelled down the peninsula.
Of these and much else about the south we know from anthologies of Tamil poetry and from an early Tamil grammar. The poems, of which the oldest date from about the time of Christ, were composed and first recited at marathon arts festivals, or assemblages (sangam), organised by the Pandyan court. They were collected into the ‘Sangam’ anthologies and committed to writing only very much later. Like the Sanskrit classics, they may therefore contain additions and revisions. On the other hand, unlike the Sanskrit classics, they were not the property of a particular caste and served no obvious ritual purpose. Moreover, they provide much reliable detail about social conditions. ‘It would be difficult to make too much of this fact,’ writes an American authority on Sangam literature. ‘Not only does ancient Tamil literature furnish an accurate picture of widely disparate classes; it also describes the social conditions of Tamil Nadu much as it was before the Aryans arrived in the south.’11
This verdict suggests some future Aryan mass-migration, for which the evidence is scant. Besides, Sangam literature was already aware of Aryan and Sanskritic ideals. The Tamil poets – and poetesses – knew the epics well and were keen to associate their patrons with the heroes of the Mahabharata. Place-names like Madurai, a variant of ‘Mathura’, reflect the early adoption of the sacred geography of the epics; and just as ‘Ayodhya’ travelled on to Thailand and central Java, so ‘Mathura/Madurai’ would make a further landfall in the crowded island of Madura off eastern Java. The Sangam poets also knew of the fabled wealth of the Nandas and of the one-time presence of the Mauryas in Karnataka. Brahmans were already well-established in the south and were the recipients of land grants; Buddhism and Jainism were also familiar; and the script used in the Tamil dedications of their caves was a form of north Indian Brahmi.
Caste distinctions were also observed in the south, but may well predate contact with the Sanskritic north. Certainly they did not conform to the hierarchical four-tier varna system; native ksatriya and vaisya are practically unknown in the south to this day. In caste functions, in hero-worship of the dead, and in the taboos and importance attached to relations between the sexes, there is indeed much that is non-Aryan. Equally unprecedented is the Sangam’s spirit of joyous celebration, which pervades both the endless wars between Cheras (Keralans), Pandyas and Cholas as well as the scenes of peaceful plenty and royal munificence which intervene. The impression given by these poems is not that of a society defying the rigid orthodoxies of inevitable Aryanisation, more of one voluntarily adjusting to prestigious new values and selectively adopting from them.
Patterns of Aryanisation were typically spontaneous and here, as outside India, Sanskritic innovations did not necessarily spread through direct contact with the Gangetic heartland. Thus it seems that the southern kingdoms derived as much from their seaward contacts as from landward intercourse. Literacy, for instance, ‘and indeed incipient civilisation in general’12 look to have originally spread not southward from the Gangetic valley but northward, from Sri Lanka. Heavily indebted to Ashoka’s missionising, Sri Lanka had stolen a march on the mainland. Its Buddhist chronicles provide the only cross-dating yet established for any of the kings mentioned in the Sangam poems. And from Sri Lanka the Brahmi script is thought to have crossed the straits to neighbouring parts of the Pandya country and thence on to Kerala and the Chola country. By this roundabout route other Aryanising traits may have followed.
The maritime dimension would continue to be crucial; in fact it is from their detailed descriptions of commercial life and foreign trade that the Sangam poems derive much of their authenticity. For in references to busy markets, bulging warehouses, ships from many lands, elaborate import/export procedures, and the Yavanas (not only the Bactrian Greeks, but foreigners in general) ‘whose prosperity never wanes’, there is an impressive convergence of Tamil testimony with what we know of south India in the first century AD from other sources, principally archaeology and copious references in the literature of the Roman empire.
This was the age of Rome’s commercial expansion. The new empire’s demand for exotica was insatiable, and the acquisition of Egypt in 30 BC had opened the maritime route to the East to Roman investors. A text written by a Greek of the first century AD, the Periplus of the Erythraean Sea, contains detailed navigational, commercial and even political information on the ports of the Indian Ocean, many of which have been reliably identified with maritime outlets on India’s coast. Ptolemy’s second-century