India: A History

Chapter One

The Harappan World

C3000-1700 BC

THE BREAKING OF THE WATERS

In Hindu tradition, as in Jewish and Christian tradition, history of a manageable antiquity is sometimes said to start with the Flood. Flushing away the obscurities of an old order, the Flood serves a universal purpose in that it establishes its sole survivor as the founder of a new and homogeneous society in which all share descent from a common ancestor. A new beginning is signalled; a lot of begetting follows.

    In the Bible the Flood is the result of divine displeasure. Enraged by man's disobedience and wickedness, God decides to cancel his noblest creation; only the righteous Noah and his dependants are deemed worthy of survival and so of giving mankind a second chance. Very different, on the face of it, is the Indian deluge. According to the earliest of several accounts, the Flood which afflicted India's people was a natural occurrence. Manu, Noah's equivalent, survived it thanks to a simple act of kindness. And, amazingly for a society that worshipped gods of wind and storm, no deity receives a mention.

When Manu was washing his hands one morning, a small fish came into his hands along with the water. The fish begged protection from Manu saying `Rear me. I will save thee.' The reason stated was that the small fish was liable to be devoured by the larger ones, and it required protection till it grew up. It asked to be kept in a jar, and later on, when it outgrew that, in a pond, and finally in the sea. Manu actedaccordingly.

[One day] the fish forewarned Manu of a forthcoming flood, and advised him to prepare a ship and enter into it when the flood came. The flood began to rise at the appointed hour, and Manu entered the ship. The fish then swam up to him, and he tied the rope of the ship to its horn [perhaps it was a swordfish], and thus passed swiftly to the yonder northern mountain. There Manu was directed to ascend the mountain after fastening the ship to a tree, and to disembark only after the water had subsided.

Accordingly he gradually descended, and hence the slope of the northern mountain is called Manoravataranam, or Manu's descent. The waters swept away all the three heavens, and Manu alone was saved.

    Such is the earliest version of the Flood as recorded in the Satapatha Brahmana, one of several wordy appendices to the sacred hymns known as the Vedas which are themselves amongst the oldest religious compositions in the world. Couched in the classical language of Sanskrit, some of the Vedas date from before the first millennium BC. Together with later works like the Brahmanas, plus the two great Sanskrit epics, the Mahabharata and the Ramayana, they comprise a glorious literary heritage whence all knowledge of India's history prior to c500 BC has traditionally been derived.

    Brief and to the point, the story of Manu and the Flood served its purpose of introducing a new progenitor of the human race and, incidentally, explaining the name of a mountain. Such, however, was too modest an interpretation for later generations. Myth, the smoke of history, is seen to signal new and more relevant meanings when espied from the distance of later millennia. In time the predicament of the small fish liable to be devoured by larger fish became a Sanskrit metaphor for an anarchic state of affairs (matsya-nyaya) equivalent to `the law of the jungle' in English. Manu's flood, like Noah's, came to be seen as the means of putting a stop to this chaos. And who better to orchestrate matters and so save mankind than Lord Vishnu? A minor deity when the Vedas were composed, Vishnu had since soared to prominence as the great preserver of the world in the Hindu pantheon and the second member of its trinity. Thus, in due course, the Flood became a symbol of order-out-of-chaos through divine intervention, and the fish (matsya) came to be recognised as the first of the nine incarnations (avatara) of Lord Vishnu. Myth, howsoever remote, serves the needs of the moment. So does history, in India as elsewhere.

    Some historians have dated the Flood very precisely to 3102 BC, this being the year when, by elaborate computation, they conclude that our current era, the Kali Yug in Indian cosmology, began and when Manu became the progenitor of a new people as well as their first great king and law-giver. It is also the first credible date in India's history and, being one of such improbable exactitude, it deserves respect.

    Other historians, while conceding the importance of 3102 BC, have declared it to be not the date of the Flood but of the great Bharata war. A Trojan-style conflict fought in the vicinity of Delhi, the war involved both gods and men and was immortalised in the Sanskrit verse epic known as the Mahabharata, the composition of whose roughly 100,000 stanzas constituted something of an epic in itself. This war, not the flood, was the event that marked the beginning of our present era and must, it is argued, therefore belong to the year 3102 BC. Complex astronomical calculations are deployed in support of this dating, and an inscription carved on a stone temple at Aihole in the south Indian state of Karnataka is said to confirm it.

    But the Aihole memorialist, endowing his temple 1600 kilometres from Delhi and nearly four thousand years later, may have got it wrong. According to the genealogical listings in the Puranas, a later collection of `ancient legends', ninety-live generations passed away between the Flood and the war; other evidence based on sterner, more recent, scholarship agrees that the war was much later than the fourth millennium BC. This greatest single event in India's ancient history, and the inspiration for the world's longest poem, did not occur until `C1400 BC' according to the History and Culture of the Indian People, a standard work of many volumes commissioned in the 1950s to celebrate India's liberation from foreign rule and foreign scholarship.

    Nevertheless, 3102 BC sticks in the historical gullet. Such are the dismal uncertainties of early Indian chronology that no slip of the chisel is going to deny the historian the luxury of a real date. Corroboration of the idea that it may, after all, apply to a Flood has since come from the excavations in distant Iraq of one of Mesopotamia's ancient civilisations. There too archaeologists have found evidence of an appalling inundation. It submerged the Sumerian city of Shuruppak, and has been dated with some confidence to the late fourth millennium BC. In fact, 3102 BC would suit it very well.

    This Sumerian inundation, and the local Genesis story in the Epic of Gilgamesh which probably derived from it, is taken to be the origin of the legend of the Flood which eventually found its way into Jewish and Christian tradition. Yet in many respects the Sumerian account is more closely echoed in the Indian version than in the Semitic. For instance, just as in later Hindu tradition Manu's fish becomes an incarnation of the great god Vishnu, so the Sumerian deity responsible for saving mankind is often represented in the form of a fish. `It is the agreement in details which is so striking,' according to Romila Thapar. The details argue strongly for some common source lot this most popular of Genesis myths, and scholars like Thapar, ever ready to expose cultural plagiarism, see both Manu and Noah as relocated manifestations of a Sumerian prototype.

    The tendency to synchronise and subordinate things Indian to parallel events and achievements in the history of countries to the west of India is a recurrent theme in Indian historiography and has rightly incurred the wrath of some Indian historians. So much so that they sometimes go to the other extreme of denying that any creative impetus, any technological invention, even any stylistic convention, ever reached India from the west — or, indeed, the West. And in the case of the Flood they may have a point. Subject to the annual deluge of the monsoon and living for the most part on the flat alluvial plains created by notoriously errant river systems, the people of north India have always had far more experience of floods, and far more reason to fear them, than their neighbours in the typically more arid lands of western Asia.

    Floods, though now associated more with the eastern seaboard of the Indian subcontinent and Bangladesh, still annually inundate vast areas of the Ganga and Indus basins. They have always done so. One such Gangetic flood, dated by archaeologists to about 800 BC, destroyed the town of Hastinapura which, after the great Bharata war, had become the capital of the descendants of Arjuna, one of the war's main protagonists. Since the flooding of Hastinapura is also recorded in Sanskrit textual tradition, and since the same tradition says that the town was then under its seventh ruler since the war, an approximate date for the war itself of about 975 BC has been postulated.

    Thus, for the titanic struggle recorded in the Mahabharata, we already have three dates: 3102 BC, C1400 BC and C950 BC. A couple of millennia one way or the other is a long time even in prehistoric terms. India's history, though undoubtedly ancient, leaves much room for manoeuvre. A mistranslated word from one of the many voluminous, difficult and defective texts wherein, long after their composition, the Vedic verses were eventually written down, can create havoc. Similarly a chance discovery of no obvious provenance can prompt major revisions.

    Another flood, later than the Sumerian one but much earlier than that at Hastinapura and so perhaps a serious contender for the one which Manu survived, is thought by some to have once inundated the plains of the lower Indus in what is now Pakistan. Geologists date it to some time soon after 2000 BC, and believe that it may in fact have been a succession of inundations. Whether they were the result of climate change, of tectonic action lower down the river resulting in damming and the formation of inland lakes, or simply the cumulative effect of annual siltation is not clear. But whatever the cause, the floods were bad news for those agriculturalists who had pioneered a highly productive economy based on growing cereals in the fine soil alongside the river. Managing the river's seasonal rise so as to enrich and irrigate their fields was the key to their success. An annual surplus had generated wealth, encouraged craft industries and fostered trade. Settlements had become cities. Along the lower Indus and its tributaries had grown up one of the world's first urban societies, a contemporary of those on the Nile and the Euphrates and a rival for the tag of `the cradle of civilisation'.

    Then, soon after 2000 BC according to the archaeologists, came the floods. If they did not actually overwhelm this precocious civilisation, they certainly obliterated it. In time, layer after layer of Indus mud, possibly wind-blown as well as water-borne, choked the streets, rotted the timbers, and piled high above the rooftops. The ground level rose by ten metres and the water table followed it. Meanwhile the river resumed its regular flow and found new channels down which to flood. On top of the cities, now consigned to oblivion beneath tons of alluvium, other peoples grazed their goats, sowed their seeds and spun their myths. A great civilisation was lost to memory.

    Not until nearly four thousand years later, in fact in the early 1920s, was its existence even suspected. It was pure chance that Indian and British archaeologists, while investigating later more visible ruins at Mohenjo-daro in Sind and at Harappa in the Panjab, made the prehistoric discovery of the twentieth century. They called their find the `Indus valley civilisation', and drew the obvious comparisons with those of Egypt and Sumeria. Indeed they thought that it might be an offshoot of the latter. Later, as its sophisticated and surprisingly uniform culture became more apparent, the Indus valley civilisation was accorded distinct status. And when the extent of its cultural reach was found to embrace a host of other sites, many of them well beyond the valley of the Indus, it was renamed after one of these sites as the Harappan civilisation.

    Suddenly India's history had acquired a rich prehistoric pedigree of archaeologically verifiable antiquity. Here, it seemed, was a worthy companion to that Sanskrit literary heritage of equally impressive, though maddeningly uncertain, antiquity as comprised by the Vedas and associated texts — the Brahmanas and Puranas as well as epics such as the Mahabharata. Perhaps these two very different sources, the one purely archaeological and the other purely literary, would complement one another. An ancient and immensely distinguished civilisation would thus be revealed in multidimensional detail.

    The Harappan finds included buildings, tools, artefacts, jewellery and some sculpture. Intimate details about Harappan housing, diet and waste disposal came to light. Maritime trade with Sumeria was attested and led to some cross-dating. The Carbon 14 process produced comparative dates accurate to plus or minus a century or so. Amongst the Harappans there was even what looked like a system of writing: some four hundred characters were identified, each, it was deduced, representing a single word; and they read from right to left. Sanskritists were soon clear that this was not Sanskrit, the language of the Vedic heritage. But it might be some kind of proto-Dravidian, the parent of south India's languages, while the script did suggest similarities with Brahmi, the earliest Indian script hitherto identified and read. It seemed only a matter of painstaking study before the Harappan language would also be understood and the secrets of its civilisation revealed.

    Unfortunately this script, despite the best endeavours of international scholarship and despite the code-cracking potential of computers, remains undeciphered. Totally lacking, therefore, is any intelligible record of the Harappans written by themselves. Who were they? What did they worship? Had they established a recognisable state or states? They tell us nothing. How did they come to be there? And what became of them in the end? We don't know. Here was history complete with approximate dates, cities, industries and arts, but absolutely no recorded events. Here too was a society with a distinct and extensive culture but, barring some not very helpful bones, no people, indeed without a single name.

    Names, on the other hand, were precisely what that Sanskrit literary tradition of the Vedas provided — in mind-boggling abundance. Kings and heroes, gods and demons, places and peoples, tumble from the Vedas, Brahmanas, Puranas and epics as if ready-made for the compilation of a historical index. Although no single site, no potsherd or artefact, can certainly be identified with the people who composed these verses, and although their chronology remains shrouded in that maddening uncertainty, we know that they called themselves arya — hence `Aryan' — and we know of their lifestyle, their social organisation, their beliefs and their innumerable antecedents and descendants. Here, in short, was a people proudly obsessed with the past, who defined themselves in terms of lineages reaching back through the generations to Manu, and whose records might therefore provide for the enigmatic Harappan civilisation precisely the human detail that it so notably lacked.

    Would that it were so. In fact, as will be seen, though the two civilisations — the Harappan and the Aryan — overlapped in geography and possibly also in chronology, no shred of coincidence certainly connects them. India's history starts with the apparently irreconcilable. Only in the last few years have sustainable connections between its Harappan and Aryan constituents been tentatively proposed. These connections, though tantalising, remain few and far from conclusive. India's history as currently understood must be seen as beginning with two woefully unconnected cultures.

    This state of affairs may, however, serve as a warning. Despite the pick-and-preach approach of many nationalist historians, geographical India is not now, and never has been, a single politico-cultural entity. In fact, its current three-way division between Pakistan, India and Bangladesh, far from denying some intrinsic unity, is a notable simplification of its traditional plurality. Analogies should be drawn, if at all, not with Egypt or with Greece but with regional constructs of a similar size like the Middle East or Europe. And just as in the Middle East those early civilisations in Egypt and Mesopotamia flourished simultaneously yet quite independently, or just as later in Europe the Byzantine and Carolingian empires could both claim pre-eminence without necessarily coming into conflict, so it is in India.

    Sadly, though, this is not a situation which makes for fluent narrative history. In a global landmass as vast and varied as the South Asian subcontinent an orderly linear progression from one cultural flowering to another, one dynasty to another, or one empire-builder to another will prove elusive. Only a still far from certain chronology, and not any sequential progression, demands that the Harappans and their archaeology take precedence ahead of the Aryans and their literature.

A VERITABLE EMPIRE

To anyone familiar with the Egypt of the Pharaohs, the warren of dun diggings which is an excavated Harappan site may seem unimpressive. It is hard not to sympathise with the first archaeologist to survey Mohenjodaro. `I was greatly disappointed,' wrote Mr D.R. Bhandarkar in his report. He was visiting the largely desert province of Sind in the winter of 1911—12 as Superintending Archaeologist of the Western Circle of the Archaeological Survey of India. `Mohenjo-daro', he noted, meant `the Mound of the Dead Men'. There was one big mound and six smaller ones. And in words that must subsequently have haunted him, the Superintending Archaeologist dismissed the lot as `not representing the remains of ... any ancient monument'.

According to local tradition, these are the ruins of a town only two hundred years old ... This seems not incorrect, because the bricks here found are of the modern type, and there is a total lack of carved terra-cottas amidst the whole ruins?

    Wrong in every detail, this statement must rank amongst archaeology's greatest gaffes.

    Today's less qualified visitors, though willing to forgive the absence of `carved terra-cottas', tend to bemoan that of more obvious features. For at Mohenjo-daro no pyramids or ziggurats, no sculpted towers or mighty henges frown over the deep and dusty thoroughfares. On first acquaintance it is as if the most extensive of the Harappan sites was never really a city at all, merely the footings and foundations of one.

    This, though, is decidedly not the case. Deep in `the Mound of the Dead Men' there was once activity and industry. Behind the extant façades of blank, featureless wall families lived, craftsmen plied their trades and vendors sold their wares. If there was an absence of eye-catching memorials it was not, as will appear, through any lack of civic pride or direction. It may tell us something about the nature of authority in the Harappan state and the organisation of its society; more certainly it indicates the limited materials available to the city's builders.

    Four thousand years ago stone was as scarce in the lower Indus region as it is today. Even the local timber, though more plentiful than now, and possibly able to meet the need for roof joists, seems not to have been sufficiently well-grown for major construction purposes. Instead, it was used as fuel to fire brick kilns. The Harappans built almost entirely in brick, both sun-baked and kiln-fired, and the excellence of their firing is well attested by the survival, albeit underground, of so many structures in such a comparatively friable material. In assuming their bricks to be `of a modern type', Bhandarkar was unwittingly paying the Harappan brickmakers a generous compliment.

    Brickwork, however, has its limitations, as the Harappans were no doubt aware. Large areas can be easily enclosed and conveniently partitioned; groundplans of some of the Mohenjo-daro houses compare favourably with those of today, while larger individual structures, presumably public buildings, cover areas equivalent to half a football pitch; some walls, obviously for defence, are as thick as thirteen metres. On the other hand bricks, unlike dressed stone, must be kept small for good firing and are therefore less suitable for towering elevations and long-lasting monuments. Sun, salt and wind play havoc with a mortar of mud; weight stresses cause bowing and buckling. Few if any buildings at Mohenjo-daro were of more than two storeys. Even supposing the Harappans had aspired to the monumental extravagances of their Egyptian contemporaries, it is hard to see how they could have achieved them.

    Of unremarkable profile, then, the mud-and-rubble mounds of the Harappan cities and settlements nevertheless made an impression on Bhandarkar's successors in the Archaeological Survey. Happily ignoring his report, R.D. Banerji and Sir John Marshall resumed explorations at Mohenjo-daro in the late 1920s. Ernest Mackay and Sir Mortimer Wheeler continued their work and also re-examined Harappa, a collection of mounds in the Panjab whence in the nineteenth century bricks similar to those at Mohenjo-daro had been removed by the wagonload as ballast for a 160-kilometre section of the Lahore—Multan railway line. After Independence and the Partition of the subcontinent in 1947 B.B. Lal, J.P. Joshi, S.R. Rao, M. Rafique Mughal and a host of others extended operations to numerous other sites with outstanding results. What amazed all these pioneers, and what remains the distinctive characteristic of the several hundred Harappan sites now known, is their apparent similarity: `Our overwhelming impression is of cultural uniformity, both throughout the several centuries during which the Harappan civilisation flourished, and over the vast area it occupied.'

    The ubiquitous bricks, for instance, are all of standardised dimensions, just as the stone cubes used by the Harappans to measure weights are also standard and based on a modular system. Road widths conform to a similar module; thus streets are typically twice the width of side lanes, while the main arteries are twice or one and a half times the width of streets. Most of the streets so far excavated are straight and run either north—south or east—west. City plans therefore conform to a regular grid pattern and appear to have retained this layout through several phases of rebuilding. In most cases the ground plan consists of two quite separate settlements, one apparently residential and commercial (`the lower town'), and the other elevated on a massive brick platform (`the citadel') and endowed with more ambitious structures. `The citadel' invariably lay to the west of `the lower town'. Clearly Harappan settlements were not just India's first cities and townships but its first, indeed the world's first, planned cities and townships. Town-planning not being conspicuous in the subcontinent's subsequent urban development, they have been hailed as the only such examples until, in the eighteenth century AD, Maharajah Jai Singh decided to lay out his `pink city' of Jaipur in Rajasthan.

    Harappan tools, utensils and materials confirm this impression of obsessive uniformity. Unfamiliar with iron — which was nowhere known in the third millennium BC — the Harappans sliced, scraped, bevelled and bored with `effortless competence' using a standardised kit of tools made from chert, a kind of quartz, or from copper and bronze. These last, along with gold and silver, were the only metals available. They were also used for casting vessels and statuettes and for fashioning a variety of knives, fish-hooks, arrowheads, saws, chisels, sickles, pins and bangles. As for the potters' production of dishes, bowls, jars, flasks and figurines, it was all that one would expect of master brickmakers — well made, competent if restrained as to decoration, and predictably uniform as to design. In short, the uniformity in technology `is as strong as in the town-planning, and so marked that it is possible to typify each craft with a single set of examples drawn from one site alone'.

    What made all this consistency even more remarkable was the area throughout which the Harappans sustained it. With Mohenjo-daro and Harappa nearly six hundred kilometres apart, it was immediately obvious that the `Indus valley' civilisation was more extensive than its contemporaries — Egypt's Old Kingdom and Mesopotamia's Sumeria. The Indus valley, however, has proved to be only the core area. Subsequent to the discovery of its two principal sites (Mohenjo-daro in Sind and Harappa in the Panjab) the Harappan civilisation has been steadily expanding by more than a province a decade. In Pakistan further sites have been found, not only in Sind and Panjab (where at Fort Derawar on the desert frontier with India a third major city stood), but as far away as the Iranian frontier in Baluchistan and in the North-West Frontier Province. India itself, not to be outdone, now boasts an important cluster of sites in Gujarat, another in Rajasthan, and more scattered settlements in the states of Panjab, Haryana, Uttar Pradesh, and Jammu and Kashmir. Lately, hundreds of kilometres away to the north-west, what seems to be a Harappan settlement, or `colony', has been identified at Shortughai near the river Oxus (Amu Darya) on Afghanistan's Russian frontier. From Lothal, a small but important settlement in Gujarat which may have been a port, to Shortughai in the mountains of Badakshan, where the Harappans probably obtained supplies of lapis lazuli, is a distance of over sixteen hundred kilometres; and east—west from Alamgirpur on the upper Ganga to Sutkagen-dor on the Makran coast is hardly less.

    Naturally such a bonanza of new sites has prompted some revisionism. The uniformity of Harappan culture, necessarily dented by local adaptations to the desert, upland and maritime extremities of such a vast area, is no longer taken for granted. Theories based upon it about the existence of a strong central authority, a pervasive administration and a heavily regulated and stratified society have also suffered. The easy assumptions made on the basis of a few partially and imperfectly excavated sites are dubbed `old platitudes' as a new generation of scholars and field workers gingerly sifts the incontrovertible from the fanciful.

    One mystery has certainly been solved. Pioneers like Marshall were puzzled how such a sophisticated culture could have sprung up from nowhere. Unaware of any other Bronze Age cultures in the region, not impressed by the Indian characteristics of Harappan architecture and artefacts, and wrongly assuming dates of about 3500—3000 BC, they duly looked to the west for an explanation, and suggested that the Indus valley civilisation must be a colony or offshoot of Mesopotamian or even Mycenaean civilisation. This idea is now quite untenable. At numerous sites to the west of the Indus in Baluchistan and Afghanistan, as well as in the Indus valley itself, sufficient pre-Harappan and Early Harappan settlements have been found to establish a local progression from hunter-gatherer to urban dweller by way of all the various stages of pastoralism, agricultural settlement, technological advance and cultural refinement. No such consensus exists about the Late Harappan and post-Harappan periods, but it is now possible to assign most Chalcolithic (Bronze/Stone) Age sites in the region to one of these categories and to give approximate dates for each.

    Designated by their find sites and principally distinguished by their pottery styles, the pre-Harappan peoples of C3000 BC had already progressed to building houses and tilling the land. They had some knowledge of metals and had access, through trading links, to other precious materials and manufactures. Some time around 2600 BC — the dating varies from site to site — the appearance of typically Harappan styles in pottery and tools announces the Early Harappan phase. Brick-built houses assume a regular design with a courtyard and rooms off it. Figurines anticipate later Harappan styles. Towards the end of the millennium, say 2300 BC, this Early Harappan style gives way to the Mature Harappan phase, in which appears the full inventory of Harappan artefacts — standardised bricks and pots; regular streets above a network of well-made sewerage ducts; typical terracottas; a notable production of decorative artefacts including beads, faïence and shell work; more copper and bronze hardware; and a plenitude of the mysterious seals (as well as the impressions made by them) whereon that enigmatic script features prominently. In some cases, to produce the typical grid layout of streets, sites were apparently cleared and then rebuilt. Other sites were briefly deserted before being rebuilt. Still others suggest a continuance of non-Harappan or pre-Harappan styles, particularly in ceramics, side by side with the Mature Harappan. It is thus far from clear what relationships — of tribute, migration, conquest, intermarriage or cultural attraction — underlay the transition to greater standardisation.

    Even worse inconsistency characterises the Late Harappan phase. Around 1900 BC Mohenjo-daro was gradually abandoned, possibly because of those floods and the associated salination of the soil. Kalibangan, an important town in Rajasthan, suffered a similar fate, but probably from desertification and the drying-up of the Ghaggar river. Elsewhere there is evidence of declining authority and of population decrease, possibly as a result of migration from the central settlements. Yet in some peripheral areas like Gujarat, Haryana and the Panjab, the decline is less marked and there may even have been an increase in activity and population.

    Dispersal or dilution are evident from the prevalence of non-Harappan pottery styles, impoverishment and disruption from the gradual disuse of the script and from the disappearance of the more fanciful manifestations of Harappan culture, including that obsessive standardisation. On the other hand, craft skills and agricultural expertise survived. The spinning and weaving of cotton, for instance, in which the Harappans seem to have been the world's pioneers, must have been gradually disseminated throughout India, since by the mid-first millennium BC it was commonplace. The finer textiles were by then an important item of trade and would remain so ever after, enticing to India Roman, Arab and eventually European merchants.

    A similar case might be made for the ox-drawn wagon, which was as much a cliché of the Harappan world as it is of the Indian subcontinent today. Again, the Harappans may have been the first in the world to use wheeled transport. Numerous toy carts in terracotta and bronze testify to their pride in this technological breakthrough, and the generous street widths of their cities were presumably dictated by the consequent traffic.

    Provisioning cities the size of Mohenjo-daro, with its estimated thirty to fifty thousand inhabitants, necessitated not only effective transport, both by river and road, but also a reliable rural surplus, a large labour force, and some means of crop storage. It has been conjectured that the largest structures at Mohenjo-daro, Harappa, Kalibangan and possibly Lothal may have been granaries, although their internal arrangements, consisting of carefully aligned brick plinths, await a satisfactory explanation.

    The only public building whose function is beyond dispute is the great bath at Mohenjo-daro. The size of a modest municipal swimming pool, carefully sealed with bitumen, and with steps down at each end, it was clearly designed to hold water and to be used for bathing. Quite probably the ablutions, or immersion, were of some ritual significance. The bath forms the inner sanctum of an elaborate building, although there is no clear evidence that, as with later temple tanks, it was a place of worship. In fact, we have no idea what part religion played in the lives of the Harappan people. No site has certainly been identified as a temple, and most suppositions about sacrificial fires, cult objects and deities rest on doubtful retrospective reference from the Hindu practices of many centuries later. Such inferences may be as futile as, say, looking to Islamic astronomy for an explanation of the orientation of the pyramids. In short, `these theories are all fanciful and do not bear scrutiny.'

    A much-cited example, depicted on some of the Harappan seals, is that of a big-nosed gentleman wearing a horned head-dress who sits in the lotus position with an erect penis, an air of abstraction and an audience of animals. He may indeed be an early manifestation of Lord Shiva as Pashupati, `Lord of the Beasts'. But myth, as has been noted, is subject to frequent revision. The chances of a deity remaining closely associated with the same specific powers — in this case, fertility, asceticism and familiarity with the animal kingdom — for all of two thousand years must raise serious doubts, especially since, during the interval, there is little evidence for the currency of this myth. Rudra, a Vedic deity later identified with Shiva, is indeed referred to as pasupati because of his association with cattle; but asceticism and meditation were not Rudra's specialities, nor is he usually credited with an empathy for animals other than kine. More plausibly, it has been suggested that the Harappan figure's heavily horned headgear bespeaks a bull cult, to which numerous other representations of bulls lend substance.

    Similar doubts surround the female terracotta figurines which are often described as mother-goddesses. Pop-eyed, bat-eared, belted and sometimes mini-skirted, they are usually of crude workmanship and grotesque mien. Only a dusty-eyed archaeologist could describe them as `pleasing little things'. The bat-ears, on closer inspection, appear to be elaborate head-dresses or hairstyles. If, as the prominent and clumsily applied breasts suggest, they were fertility symbols, why bother with millinery? Or indeed mini-skirls?

    These and other `folk' products, including numerous toys, scarcely merit comparison with the finest of Harappan sculptures. Indeed the latter are so fine and so exquisitely modelled that, `for pure simplicity and feeling' nothing comparable was produced `until the great age of Hellas'. They are, however, extremely few: Sir Mortimer Wheeler records just eleven `more or less fragmentary' stone statuettes and one bronze figure. They are also extremely small, indeed just a few centimetres high. This combination of rarity and pocket-size invites doubts as to their provenance. They could easily have come from somewhere further afield. Two perfectly modelled miniature torsos were found at Harappa — one decidedly male, the other probably female; both have socket holes by which their missing arms were attached. On this evidence they have been convincingly related to a similar technique used by artists of the contemporary Namazga culture which was discovered by Soviet archaeologists in the Ashkabad region of Turkmenistan. Namazga equivalents have also been cited for the formidable bearded figure in an embroidered toga, of which there are two examples, and even for the most famous of all Harappan works of art, the bronze `dancing girl'.

    Although probably not dancing, the `dancing girl' is unquestionably `a pleasing little thing'. Naked save for a chunky necklace and an assortment of bangles, this minuscule statuette is not of the usual Indian sex symbol, full of breast and wide of hip, but of a slender nymphet happily flaunting her puberty with delightful insouciance. Her pose is studiously casual, one spindly arm bent with the hand resting on a déhanché hip, the other dangling so as to brush a slightly raised knee. Slim and attenuated, the legs are slightly parted, and one foot — both are now missing — must have been pointed. She could be absent-mindedly surveying her wardrobe, except that her head is thrown back as if challenging a suitor, and her hair is somehow dressed into a heavy plaited chignon of perilous but intentionally dramatic construction. Decidedly, she wants to be admired; and she might be gratified to know that, four thousand years later, she still is. If there is one piece of Harappan fine art that one is reluctant to yield to the Namazga culture it is the `dancing girl'.

    Happily her local credentials are not insignificant. For one thing her features, including full lips and broad nose, are distinctly proto-Australoid, a type not usually associated with the Central Asian culture of Namazga. Skeletons unearthed in the Indus valley, however, attest that the Harappan people were of several different racial types, amongst them that, related to Australia's native people and still represented in parts of India, of proto-Australoid cast. Furthermore, although most of the surviving Harappan stone sculptures were found at Harappa itself, whence contacts with Namazga seem to have been closest, the `dancing girl' was found at Mohenjo-daro, whose external trade was more orientated to the Persian Gulf and Mesopotamia. A better case will need to be made before the Harappans are robbed of their most celebrated representative.

    Trade, both within the sprawling Harappan world and without, was clearly essential to the development of its culture. Bronze or tin (for making bronze), silver and certain precious stones like lapis lazuli and soapstone are not found within easy reach of the Indus valley, and must therefore have been imported from elsewhere. Likewise it is clear that the Mesopotamian cultures obtained numerous commodities from the Harappans, including copper, gold, timber, ivory and probably cotton textiles. Harappan sealings and seals have been found in Sumerian sites, and Sumerian documentation makes frequent reference to relations with the distant lands of `Dilmun', `Magan' and `Meluhha'. The first seems to have been in the Persian Gulf, possibly Bahrain, and to have been something of an entrepôt. `Magan' is usually identified with the coastal regions of Iran and Baluchistan, the modern Makran coast. And `Meluhha', by a process of deduction from the trade items associated with it, looks to have been the Harappan civilisation. There are objections to this hypothesis. The Mesopotamians claim to have once conquered `Meluhha', for which there is no archaeological evidence. And a later `Meluhha' was usually associated with the African coast. Notwithstanding, opinion still favours the idea that in Sumerian references to `the ships from Meluhha' which King Sargon the Great `made tie up alongside the quay of Agade' we have a positive identification of the Harappan world.

    The importance of Harappan, or `Meluhhan', trade, and the recent speculation about it, rests heavily on the evidence provided by the Harappan seals. Usually of soapstone, or steatite, the face of each is carved intaglio and in reverse so as to leave a legible impression on soft clay. Most are rectangular and about the size of a postage stamp; and typically they include an average of five characters, or word symbols, in that unintelligible script, plus one or more images. The latter are often of animals and, in the famous examples of a humped bull with pendulous dewlap, the Harappan genius for vivid depiction from life in the minutest and most demanding of mediums has been universally acknowledged.

    Several thousand seals and sealings have now been found. The seals appear to have been distributed throughout the Harappan world, not simply in its major population centres, and to have been carried about or worn, each having a boss or hole by which they could be threaded on a string. The distribution of the sealings suggests that seals may have been used to facilitate the exchange of goods over long distances. Thus the stamped image, attached to a consignment of goods, might have identified their owner, provenance, destination or contents, and so have served somewhat the role of a waybill or even a bar-code. Clearly, if this was indeed their purpose, their multiplicity and far-flung distribution argues for a vast and buzzing commercial network. Perhaps, instead of conspicuous expenditure on monuments and memorials, the Harappans pumped their surplus into commodity exchange. It has even been suggested that the Harappans were so dependent on this exchange that its apparent decline in the early second millennium BC was a cause, rather than an effect, of the disintegration of urban life.

    Although the script remains indecipherable, interesting conclusions have been drawn from the images which usually accompany it on the seals. These are often single animals, as with the humped bull, the elephant, the tiger and a magnificent rhino. Commonest of all, however, is a stocky creature unknown to zoology with the body of a bull and the head of a zebra, from which head a single horn curls majestically upwards and then forwards. In fact, `the "unicorn" occurs on 1156 seals and sealings out of a total of 1755 found at Mature Harappan sites, that is on 60 per cent of all seals and sealings.' Shireen Ratnagar, an authority on Harappan trade, also notes that, since the word symbols which accompany these images vary from seal to seal, image and text must have conveyed different information; and that, since the images recur frequently and look like totemic subjects, they may be the identifying symbols of different social groups. Assuming such groups were based on descent, as with the Vedic Aryans, Ratnagar calls them `lineages' or clans.

... we would therefore infer that the `unicorn' was the symbol of the dominant lineage which had expanded, or was expanding, by assimilation or alliance at the expense of other lineages, and administrative office and lineage affiliation would be closely connected. In other words, we may interpret the unicorn as the religious expression of a system of political control operating through lineage connexions.

    How this political control operated, and whether oppressively or consensually, it is impossible to say. Likewise, as noted, we have no clear idea what religious practices the Harappans subscribed to. Here, and in other researches, there is, though, a gradually emerging notion of a Harappan state. Ratnagar conjectures that it began to emerge when numerous ethnic and/or cultural groups were drawn together by alliance, intermarriage and agricultural or industrial specialisation. By the time of the Mature Harappan phase these groups formed not a federation but a single state. In fact `at this stage of knowledge it appears to me that we are dealing with a veritable Harappan "empire".'

    This being the case, the total, albeit gradual, eclipse of Harappan civilisation is all the more mystifying. Sumerian civilisation led on to that of Babylon, Egypt's Old Kingdom was succeeded by the Middle Kingdom and the New Kingdom, China's dynastic succession scarcely faltered. But in the Indian subcontinent the first great experiment in urban living, in political organisation and in commercial enterprise disappeared without trace beneath the sand and the silt. In the land of reincarnation there was to be no rebirth for the bustling and ingenious world of the Harappans. History would have to begin again with a very different group of people.