3.1. Tracing the Aryan migrants

Though the question of Aryan origins was much disputed in the 19th century, the Aryan invasion theory has been so solidly dominant in the 20th century that attempts to prove it have been extremely rare in recent decades (why prove the obvious?), until the debate flared up again in India after 1990. In his attempt to prove the Aryan invasion, Bernard Sergent uses the archaeological record, which, paradoxically, is invoked with equal confidence by the noninvasionist school (e.g. B.B. Lal: New Light on the Indus Civilization, Aryan Books, Delhi 1997).

The crux of the matter is: can archaeologists trace a population migrating through Central Asia and settling down in India? There seems to be new hope to pin down this elusive band of migrants: "Today, thanks to the extremely rich findings in Central Asia in the past twenty years, the discovery of the 'pre-Indian Indians' has become possible." (p.33) Sergent has tried to identify a crucial stage in this itinerary: the 3rd-millennium Bactrian culture as the base from which the Indo-Aryans invaded India.

Bactria, the basin of the Amu Darya or Oxus river, now northern Afghanistan plus southeastern Uzbekistan, is historically the heartland of Iranian culture. In an Indian Urheimat scenario, the Iranians left India before the heyday of the Harappan cities. The next waystation, where they developed their own distinct culture, was Bactria, where Zarathushtra lived (in the city of Balkh). In that framework, it is entirely logical that a separate culture has been discovered in Bactria and dated to the late 3rd millennium BC. However, Bernard Sergent identifies this Bronze Age culture of Bactria, "one of the most briliant civilizations of Asia" (p.157), as that of the Indo-Aryans poised to invade India.

Though not figuring much in the development of his own theory, evidence for similarities in material culture between Harappa and Bactria is acknowledged by Bernard Sergent, e.g. ceramics resembling those found in Chanhu-Daro. This Harappan influence on the Bactrian culture proper is distinct from the existence of six fully Harappan colonies in Afghanistan, most importantly Shortugai in Bactria, "a settlement completely Harappan in character on a tributary of the Amu Darya (...) on the foot of the ore-rich Badakshan range (...) with lapis lazuli, gold, silver, copper and lead ores. Not one of the standard characteristics of the Harappan cultural complex is missing from it." (Maurizio Tosi: "De Indusbeschaving voorbij de grenzen van het Indisch subcontinent", in UNESCO exhibition book Oude Culturen in Pakistan, Koninklijke Musea voor Kunst en Geschiedenis, Brussels 1989, p.133)

Logically, the close coexistence of Harappan colonies and Bactrian settlements was a conduit for mutual influence but also a source of friction and conflict. Indian-Iranian conflict has been a constant from the Bronze Age (replacement of Harappan with Bactrian culture in Shortugai ca. 1800 BC, Genèse de l'Inde, p.180) through Pehlevi, Shaka and Afghan invasions in India until Nadir Shah's sack of Delhi in the 18th century. Any Bactrian-Harappan antagonism would fit this pattern of hostility between Indo-Aryans and Iranians. Sergent's first job is to disprove the Iranian and prove the Indo-Aryan character of the Bactrian culture; the second is to show a Bactrian immigration in late- or post-Harappan India and a subsequent overwhelming Bactrian cultural impact on Indian society.

Sergent cites Akhmadali A. Askarov's conclusion that the Harappan-Bactrian similarities are due to "influence of northwestern India on Bactria by means of a migration of Indus people to Central Asia after the end of their civilization". (p.224, with reference to A.A. Askarov: "Traditions et innovations dans la culture du nord de la Bactriane à l'âge du bronze",Colloque Archéologie, CNRS, Paris 1985, p.119-124) The acknowledgment of a Harappa-to-Bactria movement is well taken, but this poses a chronological problem, for the Bactrian culture was not subsequent to but contemporaneous with Harappan culture. Sergent solves the problem by pointing out that Askarov and other Soviet scholars who first dug up the sites in Margiana (eastern Turkmenistan) and Bactria, used an obsolete form of C-14 Carbon dating, and that newer methods have pushed the chronology of these sites back by centuries, making Bactrian culture contemporaneous with Harappa. (p.160)

For Sergent, this chronological correction is essential: if the Bactrian culture was that of the Indo-Aryans who brought down the Indus civilization, it is necessary that they lived there before the end of the latter. But this synchronism is equally compatible with a dim pre-Harappan kinship between the Bactrian and Harappan cultures, which were different yet partly similar, a similarity which Askarov and Sergent attribute to Harappa-to-Bactria influence (which must inevitably have existed), but which may also owe something to a common origin.

Sergent then mentions a number of similarities in material culture between the Bactrian culture and some cultures in Central Asia and in Iran proper, e.g. ceramics like those of Namazga-V (southern Turkmenistan). Some of these were loans from Elam which were being transmitted from one Iranian (in his reconstruction, Indo-Iranian) settlement to the next, e.g. the so-called "Luristan bronzes", Luristan being a Southwest-Iranian region where Elamite culture was located. Some were loans from the "neighbouring and older" (p.158) culture of Margiana: does this not indicate an east-to-west gradient for the Indo-Iranians?

Well, one effect of Sergent's chronological correction is that what seem to be influences from elsewhere on Bactrian culture, may have to be reversed: "From that point onwards, the direction of exchanges and influences gets partly reversed: a number of similarities can just as well be explained by an influence of Bactria on another region as one of another on Bactria." (p.160) Note that this fits the Iranian east-to-west expansion implicit in the Avestic data and in the first chapter of the Zoroastrian Vendidad, which puts Afghanistan in the centre of the Iranian world, with the Caspian region hardly on the horizon yet. So, even for the relation between the Bactrian culture and its neighbours, the proper northwest-to-southeast direction required by the AIT has not been demonstrated, let alone a movement all the way from the northern Caspian region to India. And if there was transmission from other cultures to Bactria (as of course there was), this does not prove that the Bactrians were colonists originating in these other cultures; they may simply have practised commerce. Conversely, if they were colonists from elsewhere, they may have been colonists originating in pre-Harappan India.

At any rate, all the sites related in material culture to the prototypical Bactrian settlement of Dashli are in present-day Tajikistan, Uzbekistan, Turkmenistan, Afghanistan or Iran proper, without exception regions which were Iranian at the time they made their appearance in written history, mostly in the last millennium BC. While migrations are obviously possible, this Iranian bias says something about the burden of proof. It is entirely reasonable to accept as a starting hypothesis that the Dashli settlement, like its sister settlements, was Iranian. Those who insist it was something else, should accept the burden of proving that Dashli was Indo-Aryan, that migrations took place in which the Indo-Aryans there made way for Iranians.

3.3. Bactria vs. Harappa

A new insight based on archaeology and detrimental to the stereotypical Harap-pan/Aryan opposition, is that the Harappans were not matriarchal pacifists after all, that they did have weapons and fortifications, "just like" the Aryans (see e.g. Shereen Ratnagar: Enquiries into the Political Organization of Harappan Society, Ravish Publ., Pune 1991; note that Prof. Ratnagar is a virulent critic of all Indocentric revisions of the Aryan question, as in her article "Revisionist at work: a chauvinistic inversion of the Aryan invasion theory", Frontline, 9-2-1996, an attack on Prof. N.S. Rajaram). Yet, Sergent insists that the old picture still holds good: relatively unarmed mercantile Harappans versus heavily armed Aryans preparing their invasion in Bactria. The Bactrian settlements abound in metal weaponry, and this does present a contrast with the relative paucity of weapons in Harappa. The latter was a well-ordered mercantile society, Bactria a frontier society.

This contrast actually reminds us of a contrast between Iranian and Indian in the historical period. In pre-Alexandrine Iranian royal inscriptions, we come across truly shameless expressions of pride in bloody victories, even defiantly detailing the cruel treatment meted out to the defeated kings. By contrast, in Ashoka's inscriptions, we find apologies for the bloody Kalinga war and a call for establishing peace and order. Far from being a purely Buddhist reaction against prevalent Hindu martial customs, Ashoka's relative pacifism presents a personal variation within a broader and more ancient tradition of Ahimsa, nonviolence, best expressed in some sections of the Mahabharata. Though this epic (and most explicitly its section known as the Bhagavad Gita) rejects the extremist non-violence propagated by Mahatma Gandhi and also by the wavering Arjuna before the decisive battle, Krishna's exhortation to fight comes only after every peaceful means of appeasing or reconciling the enemy has been tried, whence the Hindu dictum Ahimsa paramo dharma, "non-violence is the highest religious duty".

True, the Vedas seem to be inspired by the same martial spirit of the Iranian inscriptions, but in the Indocentric chronology, they predate the high tide of Harappan civilization, belonging to a pre-Harappan period of conquest, viz. the conquest of the northwest by the Yamuna/Saraswati-based Puru tribe. Their westward conquest was connected with a larger westward movement which included the Iranian conquest of Central Asia (later continued into the Caspian area and West Asia). By way of hypothesis, I propose that Ahimsa was a largely post-Vedic development, and that the Iranians (who had a taste of it through Zarathushtra's strictures against animal sacrifice and the like) missed its more radical phase, sticking instead to the more uncivilized glorification of victory by means of force. This would concur with the finding of a more military orientation of Bactrian culture as compared with the post-Vedic Harappan culture.

3.4. The Bactrian tripura

In the principal Bactrian site of Dashli, a circular building with three concentric walls has been found. The building was divided into a number of rooms and inside, three fireplaces on platforms were discovered along with the charred remains of sacrificed animals. In this building, its Soviet excavator Viktor Sarianidi recognized an Iranian temple, but Sergent explains why he disagrees with him. (p.161) He argues that the Vedic Aryans were as much fire-worshippers as the Iranians, and that they sacrificed animals just like the early Iranians did (prior to the establishment of Zarathushtra's reforms, and even later, cfr. the bull sacrifice in the Roman-age Mithras cult), so that the excavated fire altars could be either Indo-Aryan or Iranian.

Of course, India and Iran have a large common heritage, and many religious practices, mythical motifs and other cultural items in both were the same or closely similar. But that truism will not do to satisfy Sergent's purpose, which is to show that the Bactrian culture was not generally Indo-Iranian, and definitely not Iranian, but specifically Indo-Aryan. There is nothing decisively un-Iranian about the Dashli fire altars, and I think Sarianidi's identification of Dashli as Iranian remains undisproven.

In fact, there may well be something un-Indic and specifically Iranian about it. First of all, roundness in buildings is highly unusual in Hindu culture, which has a strong preference for square plans (even vertically, as in windows, where rectangular shapes are preferred over arches), in evidence already in the Harappan cities. Moreover, Sergent notes the similarity with a fire temple found in Togolok, Margiana. The Togolok fire altar has gained fame by yielding traces of a plant used in the Soma (Iranian: Haoma) sacrifice: laboratory analysis in Moscow showed this to be Ephedra, a stimulant still used in ephedrine and derivative products.

Asko Parpola has tried to identify the Togolok temple as Indo-Iranian and possibly proto-Vedic, citing the Soma sacrifice there as evidence: the Rg-Vedic people reproached their Dasa (Iranian) enemies for not performing rituals including the Soma ritual, so Parpola ("The coming of the Aryans to Iran and India and the cultural and ethnic identity of the Dasas", in Studia Orientalia, vol.64, Helsinki 1988, p.195-265) identifies the former with the "Haumavarga Shakas" or Soma-using Scythians mentioned in Zoroastrian texts. However, every testimony we have of the Scythians, including the Haumavarga ones in whose sites traces of the Soma ceremony have been found, is as an Iranian-speaking people. It is possible that the sedentary Iranians included all nomads in their term Shaka, even the hypothetical Vedic-Aryan nomads on their way to India, but it is not more than just possible. The use of Soma was a bone of contention within Mazdeism, with Zarathushtra apparently opposing it against its adepts who were equally Iranian. There is nothing against characterizing the Togolok fire temple as Iranian.

And even if Thomas Burrow ("The Proto-Indoaryans", Journal of the Royal Asiatic Society, 1973, cited with approval by Sergent: Genèse de l'Inde, p.232) were right with his thesis that the Mazdean religion originated in a sustained reaction against the Indo-Aryans present in Bactria and throughout the Iranian speech area, making the non-Zoroastrian faction in Greater Iran an Indo-Aryan foreign resident group, it remains to be proven that these dissident Indo-Aryans made way for Zoroastrian hegemony in Iran by moving out, and more specifically by moving to India, somewhat like Moses taking the Israelites out of Egypt. There is neither scriptural nor archaeological evidence for such a scenario: the normal course of events would be assimilation by the dominant group, and the only emigration from Iranian territory (if it had already been iranianized) by Indo-Aryans that we know of, is the movement of the Mitannic Indo-Aryans from the southern Caspian area into Mesopotamia and even as far as Palestine.

In the Dashli building, Asko Parpola recognized a tripura such as have been described in the Vedic literature as the strongholds with three circular concentric walls of the Dasas or Asuras (Asura/Ahura worshippers), whom Parpola himself has identified elsewhere as Iranians ("The coming of the Aryans", Studia Orientalia, vol.64, p.212-215, with reference to Shatapatha Brahmana 6:3:3:24-25; and: "The problem of the Aryans and the Soma", in G. Erdosy ed.: The Indo-Aryans of Ancient South Asia, p.368 ff.). So, chances are once more that the Soma-holding fire-altars, like the tripura structures around them, in both Togolok and Dashli, were Iranian. Parpola (in Erdosy, ibid.) makes this conclusion even more compelling when he informs us that a similar building in Kutlug-Tepe "demonstrates that the tradition of building forts with three concentric walls survived in Bactria until Achaemenid times" -- when the region was undoubtedly Iranian.

Moreover, Parpola points out details in the Vedic descriptions of the tripura-holding Dasas and Asuras which neatly fit the Bactrian culture: the Rg-Veda "places the Dasa strongholds (..) in the mountainous area", which is what Afghanistan looks like to people from the Ganga-Saraswati-Indus plains; it speaks of "a hundred forts" of the Dasa, while the Vedic Aryans themselves "are never said to have anything but fire or rivers as their 'forts'. The later Vedic texts confirm this by stating that when the Asuras and Devas were fighting, the Asuras always won in the beginning, because they alone had forts. (...) The Rg-Vedic Aryans described their enemy as rich and powerful, defending their cattle, gold and wonderful treasures with sharp weapons, horses and chariots. This description fits the Bactria-Margiana Archaeological Complex in Bactria, with its finely decorated golden cups, weapons with ornamental animal figurines including the horse, and trumpets indicative of chariot warfare." (in Erdosy, ibid.)

This may pose a chronological problem to those who consider the Rg-Veda as pre-Bronze Age, or perhaps not, e.g. Parpola notes that the term tripura was "unknown to the Rg-Veda" and only appears later, "in the Brahmana texts" (in Erdosy, p.369) which noninvasionists date to the high Harappan period, contemporaneous with the Bactrian Bronze Age culture. At any rate, it affirms in so many words that the Bactrian Bronze Age culture was Dasa or Asura, terms which Parpola ("The coming of the Aryans", Studia Orientalia, vol.64, p.224) had identified with "the carriers of the Bronze Age culture of Greater Iran". It also constitutes a challenge to those who make India the Urheimat of IE or at least of Indo-Iranian: if the presumed tripuras are a distinctly Dasa/Iranian element, identified as such in Vedic literature, and if the Vedic Aryans fought the Dasas in India, as the Rg-Vedic data indicate, should we not be able to find some tripuras in India too? Or did the Iranians only develop them after leaving India but while still waging occasional wars on the Indian border?

3.5. Were the Bactrians Indo-Aryans?

Other artefacts in Dashli have the same Iranian/Indo-Aryan ambiguity with a preference for the Iranian alternative. A vase in Dashli shows a scene with men wearing a kind of shirt leaving one shoulder uncovered. In this, Sergent recognizes the upanayana ceremony, in which a youngster is invested with the sacred shirt or thread. (p.163) But this is both a Vedic and a Zoroastrian ritual, with the latter resembling the depicted scene more closely: in India, only a thread is given, but among Zoroastrians, it is an actual shirt.

Some vases display horned snakes or dragons carrying one or more suns inside of them: according to Sergent, this refers to an Indo-Iranian dragon myth, attested in slightly greater detail in the Rg-Veda than in the Avesta (but what else would you expect, with Vedic literature being much larger, older and better preserved than the Avestan corpus?), about Indra liberating the sun by slaying the dragon Vrtra, or in the Avesta, Keresaspa killing the snake Azhi Srvara, "the horned one". (p.163-164, ref. to Rg-Veda 1:51:4, 1:54:6) The sources which drew his attention to this picture, both Soviet and French (Russian articles from the 1970s by Viktor Sarianidi and by I.S. Masimof, and Marie-Hélène Pottier: Matériel Funéraire de la Bactriane Méridionale à l'Age du Bronze, Paris 1984, p.82 ff.), are agreed that it is specifically Iranian, and we have no reason to disbelieve them. What Sergent adds is only that, like with the fire cult, it could just as well be Indo-Aryan; but that does not amount to proof of its Indo-Aryan rather than Iranian identity.

Several depictions (statuettes, seals) of a fertility goddess associated with watery themes have been found. Sergent points out that they are unrelated to Mesopotamian mythology but closely related to the "Indo-Iranian" goddess known in India as Saraswati, in Iran as Anahita. Which shall it be in this particular case, Iranian or Indian, Avestan or Vedic? Sergent himself adds that the closest written description corresponding to the visual iconography in question is found in Yasht 5 of the Avesta. (p.163)

Of course we must remain open to new interpretations and new findings. In this field, confident assertions can be overruled the same day by new discoveries. But if Sergent himself, all while advocating an Indo-Aryan interpretation of the known Bactrian findings, is giving us so many hints that their identity is uncertain at best, and otherwise more likely Iranian than Indo-Aryan, we have reason to believe in the Iranian identification established by other researchers. On the strength of the data he offers, the safest bet is that the Bactrian Bronze Age culture was the centre of Iranian culture.

This happens to agree with the evidence of Zoroastrian scripture, which has dialectal features pointing to the northeast of the historical Iranian linguistic space, meaning Bactria, and which specifically locates Zarathushtra in Bahlika/Balkh, a town in northern Afghanistan. It tallies with the list of regions in the opening chapter of the Vendidad, corresponding to Bactria, Sogdia, Pamir, Margiana, southern Afghanistan and northwestern India (Hapta Hendu, the Vedic Sapta Sindhavah), which happens to put Balkh near the geographical centre. Iran proper was iranianized only well after Zarathushtra's preaching. As Sergent notes, in ca. 1900 BC, the Namazga culture in Turkmenistan changes considerably taking in the influence of the then fast-expanding Bactria-Margiana culture (p.179): I read that as the Iranian expansion from their historical heartland westward into the south-Caspian area. From there, but again only after a few more centuries, they were to colonize Kurdistan/Media and Fars/Persia, where their kingdoms were to flourish into far-flung empires in the 1st millennium BC.

It is only logical that the dominant religious tradition in a civilization is the one developed in its demographic and cultural metropolis: the Veda in the Saraswati basin, the Avesta in the Oxus basin, i.e. Bactria. That Bactria did have the status of a metropolis is suggested by Sergent's own description of its Bronze Age culture as "one of the most brilliant in Asia". Though provincial compared with Harappa, it was a worthy metropolis to the somewhat less polished Iranian civilization.

3.6. Clarions of the Aryan invaders

Another distinctively Aryan innovation attested in Dashli was the trumpet: "Bactria has yielded a number of trumpets; some others had been found earlier in Tepe Hissar and Astrabad (northeastern Iran); Roman Ghirshman proposed to connect these instruments with the use of the horse, with the Iranian cavalry manoeuvring to the sound of the clarion. (...) In ancient India, the trumpet is not mentioned in the written sources". (p.162) Would it not be logical if the same type of cavalry manoeuvres had yielded the Aryans both Iran and India? In that case, we should have encountered some references to clarions in the Vedas. But no, as per Sergent's own reading, the Rg-Veda, supposedly the record of Aryan settlement in India, knows nothing of trumpets; though post-Harappan depictions of riders with trumpets are known.

All this falls into place if we follow the chronology given by K.D. Sethna and other Indian dissidents: the Rg-Veda was not younger but older than the Bronze Age and the heyday of Harappa. So, the trumpet was invented in the intervening period, say 2,500 BC, and then used in the subsequent Iranian conquest of Bactria, Margiana and Iran.

The comparatively recent migration into Iran of the Iranians, who supposedly covered the short distance from the Volga mouth to Iran in the 3rd or 2nd millennium BC (losing the wayward Indo-Aryans along the way), has not been mapped archaeologically, in contrast with the successive Kurgan expansion waves into Europe. Jean Haudry reports optimistically: "Since the late 3rd millennium BC, an undecorated black pottery appears in Tepe Hissar (Turkmenistan), together with violin-shaped female idols and esp. with bronze weapons, the horse and the war chariots, and -- a detail of which R. Ghirshman has demonstrated the importance -- the clarion, indispensible instrument for collective chariot maneuvers. We can follow them from a distance on their way to the south." (J. Haudry: Les Indo-Européens, p.118, with reference to R. Ghirshman: L'Iran et les Migrations des Indo-Aryans et des Iraniens, 1977) But this is not necessarily the entry of "the" Iranians into Iran, and if it is, it need not have the Kurgan area as its starting-point.

In the account of Roman Ghirshman and Jean Haudry, the proto-Iranians with their clarions travelled "to the south". Rather than Indo-Iranians on their way from South Russia to Iran and partly to India, these may just as well be the Iranians on their way from Bactria (and ultimately from India), via the Aral Lake area, to Iran and Mesopotamia. Indeed, viewed from Iran, entrants from Russia and from India would arrive through the same route, viz. from the Aral Lake southward. A look at the map suffices to show this: rather than go in a straight line across the mountains, substantial groups of migrants would follow the far more hospitable route through the fertile Oxus valley to the Aral Lake area, and then proceed south from there.

Even in Bernard Sergent's erudite book, I have not found any data which compel us to accept that a particular culture can be identified with the very first Indo-Iranian wave of migrants; Central Asia was criss-crossed for millennia by variegated Iranian-speaking tribes. Nonetheless, Haudry's clarion-wielders of "the late 3rd millennium BC" and Sergent's occupiers of Namazga "in ca. 1900 BC" may of course be the first Iranian intruders into Turkmenistan and Persia, but that would serve the Indocentric theory even better, for Sergent's data show that these intruders came from Bactria, not from Russia.

3.7. Bactrian invasion in India

Thus far, the archaeological argument advanced by some scholars in favour of an Aryan invasion into India has not been very convincing. Consider e.g. this circular reasoning by Prof. Romila Thapar ("The Perennial Aryans", Seminar, December 1992): "In Haryana and the western Ganga plain, there was an earlier Ochre Colour Pottery going back to about 1500 BC or some elements of the Chalcolithic cultures using Black-and-Red Ware. Later in about 800 BC there evolved the Painted Grey Ware culture. The geographical focus of this culture seems to be the Doab, although the pottery is widely distributed across northern Rajasthan, Panjab, Haryana and western U.P. None of these post-Harappan cultures, identifiable by their pottery, are found beyond the Indus. Yet this would be expected if 'the Aryans' were a people indigenous to India with some diffusion to Iran, and if the attempt was to find archaeological correlates for the affinities between Old Indo-Aryan and Old Avestan."

Firstly, if no common pottery type is found in Iran and India in 1500-800 BC, and if this counts as proof that no migration from India to Iran took place, then it also proves that no migration from Iran to India took place. In particular, the PGW, long identified with the Indo-Aryans, cannot be traced to Central Asia; if it belonged to Aryans, then not to Aryan invaders. So, if substantiated, Prof. Thapar's statement is actually an argument against an Aryan invasion in ca. 1500 BC. Secondly, if the absence of migration in either direction in the period from 1500 BC onwards is really proven, this evidence remains compatible with an Indo-European emigration from India in another time bracket, say between 6000 and 2000 BC.

In spite of the impression created in popular literature, archaeology has by no means demonstrated that there was an Aryan immigration into India. Even the new levels in accuracy do not affect the following status quaestionis of the Aryan Invasion theory: "The question of Indo-European migrations into the subcontinent of India can, at best, be described as enigmatic." (David G. Zanotti: "Another Aspect of the Indo-European Question", Journal of Indo-European Studies, 1975/3, p.260) Thus, among those who assume the Aryan Invasion, there is no consensus on when it took place, and some AIT archaeologists alter the chronology so much that the theory comes to mean the opposite of what it is usually believed to mean, viz. an affirmation of Aryan dominance in Harappa rather than an Aryan destruction of Harappa: "[This] episode of elite dominance which brought the Indo-Aryan branch of the Indo-European family to India (...) may have been as early as the floruit of the Indus civilization (...)" (C. Renfrew: "Before Babel", Cambridge Archaeological Journal, 1, p.14)

Enter Bernard Sergent. He builds on a corpus of findings (some of them already used by Asko Parpola) pertaining to the apparent entry of elements from the Bactrian Bronze Age culture into late- and post-Harappan northwestern India. He also offers a theory of how these Bactrians may have caused the downfall of the Harappan civilization, parallel with the contemporaneous crisis in civilizations in Central and West Asia.

Civilization and urbanization are closely related to commerce, exchange, colonization of mining areas, and other socio-economic processes which presuppose communications and transport. When communication and transport cease, we see cultures suffer decline, e.g. the Tasmanian aboriginals, living in splendid isolation for thousands of years, had lost many of the skills which mankind had developed in the Stone Age, including the art of making fire. One of the reasons why the Eurasian continent won out against Africa and the Americas in the march of progress, was the fairly easy and well-developed contact routes between the different civilizations of Europe, Egypt, Mesopotamia, India and China. So, one can force decline on a culture by cutting off its trade routes, a tactic routinely used for short periods (hence only with limited long-term effect) in wartime, but which seems to have troubled the ancient civilizations in ca. 2000 BC with devastating effect for several centuries. It was in reaction to this destabilization of international trade links that the civilizational centres started building empires by the mid-2nd millennium, e.g. the Kassite empire in Mesopotamia where there had only been city-states (Ur, Uruk, Isin, Larsa, etc.) prior to the great crisis.

Or so Sergent says. Dismissing the thesis of a climatological crisis (though such a crisis would by itself already trigger an economic crisis even in the areas not directly affected climatologically), he argues that only an economic crisis can explain the simultaneous decline of cities in widely different locations, some near rivers and some on hills, some in densely populated agglomerations and some overlooking thinly populated steppes or mountain areas, some in hot and some in colder areas. The ones to blame are -- who else? -- the Aryans. They, and "specifically Indo-Aryans" (p.198-99), played a role in the Hurrian and Kassite invasions disrupting Mesopotamia (while the IE or non-IE identity of the Guti and Lullubi invaders remains unknown, though attempts are made to link the Guti with the Tocharians); and from Bactria, they by themselves disrupted the economy of the Indus-Saraswati civilization.

They didn't physically destroy the Harappan cities, as Mortimer Wheeler and others of his generation thought: "No trace of destruction has been observed in these cities." (p.201) But by creating insecurity for the travelling traders, they bled and suffocated the economy which made city life possible, and thus forced the Harappans to abandon their cities and return to a pre-urban lifestyle. The declining and fragmented Harappan country and society then fell an easy prey to the Indo-Aryan invaders from Bactria.

This scenario has been attested in writing in the case of Mesopotamia. Sergent quotes other experts to the effect that "from ca. 2230 BC, (...) the Guti had cut off the roads, ruined the countryside, set the cities on fire" (p.199, quoting Paul Garelli: Le Proche-Orient Asiatique, PUF, Paris 1969, p.89-93), that the Assyrian trade system was disrupted by the Mitannic people, etc. But is there similar evidence for the Indus-Saraswati civilization?

Sergent cites findings that in the final stage of Mohenjo Daro, we see the large mansions of the rich subdivided into small apartments for the poor, the water supply system neglected, the roads and houses no longer following the plan. (p.200) This certainly marks a decline, the rich losing their power and the powerful losing their control and resources. Same story in Harappa, Chanhu Daro, Kalibangan, Lothal: a great loss of quality in architecture and organization in the last phase. Moreover, all traces of long-distance trade disappear (just as in Mesopotamia, all signs of commerce with "Meluhha"/Sindh disappear by 2000 BC), and trade is the basis of city life. So, "these cities didn't need to be destroyed: they had lost their reason for existing, and were vacated". (p.201) But that doesn't bring any Indo-Aryan invasion into the picture. Indeed, it is perfectly compatible with a hypothesis of Iranian Bactrians disrupting a Harappan economy manned by Indo-Aryans.

3.8. Aryan invader settlements in India

To Bernard Sergent, the "strategic" key to the Aryan invasion puzzle has been provided by the discovery, by a French team in 1968, of the post-Harappan town of Pirak, near the Bolan pass and near Mehrgarh in Baluchistan. Pirak was a new settlement dating back only to the 18th century BC. Culturally it was closely related to the societies to its north and west, especially Bactria. Sergent sums up a long list of precise material items which Pirak had in common with those non-Indian regions. (p.219 ff.) So, this was a settlement of foreign newcomers bringing some foreign culture with them.

Sergent will certainly convince many readers by asserting that in Pirak, "the horse makes its appearance in India, both through bones and in figurines", and this "connotes without any possible doubt the arrival in India of the first Indo-European-speaking populations". (p.221) That depends entirely on how much we make of the limited but real evidence of horses in the Harappan civilization. Note moreover that while the horse was important to the Indo-Aryans, the Bactrian two-humped camel was not; but in Pirak, both camel and horse are conspicuous, both in skeletal remains and in depictions.

If the Bactrian culture and those to its west were Iranian-speaking, which is likely, then Pirak is simply an Iranian settlement in an Indian border region, a southward extension of the Bactrian culture. Indo-Iranian borders have been fluctuating for millennia, while different groups of Iranians down to Nadir Shah have again and again tried to invade India, so the Iranian intrusion in Pirak (which may have ended up assimilated into its Indo-Aryan environment) need not be the momentous historical breakthrough which it is to Sergent. It would only be that if it can be shown that the Pirak innovations are repeated in many North-Indian sites in the subsequent centuries, where we know that the dominant culture was Indo-Aryan.

A related culture is the Cemetery H culture on the outskirts of Harappa itself. Sergent offers a detail which is distinctly non-Vedic and Mazdean (Zoroastrian): "The dead, represented by unconnected skulls and bones, were placed, after exposure, in big jars". (p.224; emphasis added) Exposure to birds and insects is still the first stage in the Zoroastrian disposal of the dead. Sergent also reports that the influence of the native Harappan civilization is much greater here than in Pirak. So, as the Iranian invaders moved deeper inland, across the Indus, they soon lost their distinctiveness. Considering that Afghan dynasties have ruled parts of India as far east as Bengal, using Persian and building in a West-Asian style, this post-Harappan Iranian intrusion as far as the Indus riverside is not that impressive.

Indeed, from the Indus eastwards, we lose track of this Bactrian invasion. Sergent himself admits as much: "For the sequel, archaeology offers little help. The diggings in India for the 2nd millennium BC reveal a large number of regional cultures, generally rather poor, and to decree what within them represents the Indo-Aryan or the indigenous contribution would be arbitrary. If Pirak (...) represents the start of Indian culture, there is in the present state of Indian archaeology no 'post-Pirak' except at Pirak itself, which lasted till the 7th century BC: the site remained, along with a few very nearby ones, isolated." (p.246-247) So, the Bactrian invaders who arrived through the Bolan pass and established themselves in and around the border town of Pirak, never crossed the Indus, and never made their mark on India the way the Indo-Aryans did.

This confirms the statement by the American archaeologist Jim Shaffer that "no material culture is found to move from west to east across the Indus" (personal communication, 1996), or more academically, that the demographic eastward shift of the Harappan population during the decline of their cities, i.e. an intra-Indian movement from Indus and Saraswati to Ganga, "is the only archaeologically documented west-to-east movement of human populations in South Asia before the first half of the first millennium BC", while the archaeological record shows "no significant discontinuities" for the period when the Aryan invasion should have made its mark. (Jim Shaffer and Diane Lichtenstein: "The concepts of 'cultural tradition' and 'palaeoethnicity' in South-Asian archaeology", in G. Erdosy, ed.: The Indo-Aryans of Ancient South Asia, p.139-140)

The Pirak people were not the Vedic Aryans conquering India. The Aryan invasion of India has somehow gone missing from the archaeological record, and this is admitted by Sergent himself in the very section containing his decisive piece of evidence for the Aryan Invasion Theory.

3.9. Scriptural evidence

To fortify his reconstruction of the Aryan invasion, Bernard Sergent repeats some well-known scriptural references. Indian authors are right in pointing out that this is systematically the weakest part in AIT argumentations, as the knowledge of Vedic literature among Western scholars is either too limited or too distorted by AIT presuppositions. Sergent's arguments at this point repeat well-known claims about the contents of the Vedas. Thus, the Rg-Veda was written by foreigners because it doesn't know the tiger nor rice nor "the domesticated elephant which existed in the Harappan Indus culture". (p.241)

As for the tiger, it is often said that India was divided in a lion zone in the west and a tiger zone in the rest. This image persists in the symbolism of the civil war in Sri Lanka: the Sinhalese, originating in Gujarat (where lions exist even today), have the lion as their symbol, while the separatists among the Tamils, originating in southeastern India, call themselves the Tigers. However, to judge from the Harappan seal imagery, tigers did originally exist in the Saraswati and Indus basins as well, overlapping with the lion zone. As Sir Monier Monier-Williams (Sanskrit-English Dictionary, p.1036, entry vyâghra) notes, in the Atharva-Veda, "vyâghra/tiger is often mentioned together with the lion". It is simply impossible that the Rg-Vedic seers, even if they were unfamiliar with the Ganga basin (quod non), had never heard of tigers.

As for the domesticated elephant, if it was known in Harappa, does anyone seriously suggest that it was not known in the same area in subsequent centuries by the Vedic Aryans?

While regression in knowledge and technology does sometimes happen, there is no reason whatsoever why people who could domesticate elephants would have lost this useful skill, which is not dependent on foreign trade or urbanization, when the Harappan cities declined. Isn't the mention of how "the people deck him like a docile king of elephants" (Rg-Veda 9:57:3, thus translated by Ralph Griffith: Hymns of the Rg-Veda, p.488) a reference to the Hindu custom of taking adorned domesticated elephants in pageants?

Rice, according to Sergent himself, made its appearance in the Indus basin in the late Harappan period, and was known to the Bactrian invaders in Pirak. (p.230) He identifies those Bactrian invaders as the Vedic Aryans, so why haven't they mentioned rice in their Rg-Veda? One simple answer would be that the Rg-Veda is pre-Harappan, composed at a time when rice was not yet cultivated in northwestern India. This chronological correction solves a lot of similar arguments from silence. Thus, there was cotton in Harappa and after, but no cotton in the Rg-Veda. Bronze swords were used aplenty in the Bactrian culture and in Pirak, but are not mentioned in the Rg-Veda; a short knife can be made from soft metals like gold or copper, but a sword requires advanced bronze or iron metallurgy. (Ralph Griffith uses "sword" twice in his translation The Hymns of the Rg-Veda, p.25, verse 1:37:2, and p.544, verse 10:20:6, both already in the younger part of the Rg-Veda, but in the index on p.702 he corrects himself, specifying that "knife" or "dagger" would be more appropriate.) Likewise, the core stories of the Ramayana and Mahabharata, the ones most likely to stay close to the original versions even in their material details (unlike the many sideshows woven into these epics, often narrating much more recent events), feature only primitive pre-Bronze Age weapons: Rama's bow and arrow, Hanuman's club.

Camels were part of the Bactrian culture and its Pirak offshoot, but are not mentioned in the Rg-Veda except for its rather late 8th book, which mentions Bactrian fauna, possibly in the period when the early Harappans were setting up mining colonies like Shortugai. It all falls into place when the Rg-Veda is considered as pre-Harappan. Incidentally, the late appearance of Afghan fauna in the Rg-Veda contradicts an Afghanistan-to-India itinerary, and argues in favour of an India-to-Afghanistan movement during the Rg-Vedic period.

For a very different type of scriptural evidence, Sergent sees a synchronism between the archaeologically attested settlement of Pirak and the beginning of the Puranic chronology, which in his view goes back to the 17th century BC, in "remarkable coincidence" with the florescence of Pirak. (p.223) Reference is in fact to Kalhana's Rajatarangini, which starts a dynastic lists of kings of Kashmir in 1882, i.e. the early 19th century BC. But if Kalhana can be a valid reference, what about Kalhana's dating the Mahabharata war to the 25th century BC? If Puranic history is any criterion, Sergent should realize that its lists of Aryan kings for other parts of India than Kashmir go way beyond 2,000 BC.

Another classic scriptural reference concerns everything relating to the enemies of the Vedic Aryans, such as the "aboriginal" Dasas. Very aptly, Sergent identifies the Dasas and the Panis as Iranian, and the Pakthas (one of the tribes confronting the Vedic king Sudas in the Battle of the Ten Kings) as the Iranian Pathans. (p.241-244) He specifically rejects the common belief that the Dasas were black-skinned, in spite of their occasional description as "black-covered" or "from a black womb", pointing out that even the fair-haired and white-skinned Vikings were called the "black foreigners" by the Irish, with "black" purely used as a metaphor for "evil". (This is even the case in some African languages, for there is no relation between colour symbolism and skin colour: white is the sacred colour to dark-skinned Indian tribals, while black is auspicious to the whitish Japanese, who consider white as the colour of mourning, just as Sanskritic Hindus do.)

Yet, Sergent doesn't identify the said Iranian tribes with the Bronze Age Bactrians, arguing that in Alexander's time, Greek authors locate the Parnoi and Dahai just south of the Aral Lake. But that was almost two thousand years after the heyday of the Bactrian Bronze Age culture and arguably even longer after the Rg-Veda. The only mystery is that these ethnonyms managed to survive that long, not that during those long centuries, they could migrate a few hundred miles to the northwest -- centuries during which we know for fact that the Iranians expanded westward from their Bactrian heartland across rivers and mountains to settle as far west as Mesopotamia.

Moreover, the Vedas locate the confrontations in the prolonged hostility between IndoAryans and Iranians not on the Saraswati (which could in theory be identified as the homonymous Harahvaiti/Helmand in Afghanistan) (p.242), but on the riverside of the Parush-ni/Ravi and other Panjab rivers, unambiguously in India. This is only logical if the Vedic Aryans were based in the Saraswati basin and their Iranian enemies were based in an area to their west near the Khyber pass: they confronted halfway in Panjab. So not only did these Iranian tribes (Dahai, Parnoi) move from Bactria to the Aral Lake area in 2000-300 BC, but they had started moving northwestward centuries earlier, in the Rg-Vedic period, in Panjab.

With every invasionist attempting to strengthen his case by appealing to the testimony of Hindu scripture, the collective failure becomes more glaring.

3.10. Comparison with archaeological reconstruction in Europe

The westward expansion of the Kurgan culture has been mapped with some degree of accuracy: "If an archaeologist is set the problem of examining the archaeological record for a cultural horizon that is both suitably early and of reasonable uniformity to postulate as the common prehistoric ancestor of the later Celtic, Germanic, Baltic, Slavic, and possibly some of the Indo-European languages of Italy, then the history of research indicates that the candidate will normally be the Corded Ware culture. At about 3200-2300 BC this Corded Ware horizon is sufficiently early to predate the emergence of any of the specific proto-languages. In addition, it is universally accepted as the common component if not the very basis of the later Bronze Age cultures that are specifically identified with the different proto-languages. Furthermore, its geographical distribution from Holland and Switzerland on the west across northern and central Europe to the upper Volga and middle Dniepr encompasses all those areas which [have been] assigned as the 'homelands' of these European proto-languages."

(J.P. Mallory: In Search of the Indo-Europeans, Hudson & Hudson, London 1989, p.108)

This is a very important insight for understanding the large common (partly pre-IE substratal) element in the European IE languages, distinguishing them collectively from Anatolian, Tocharian and Indo-Iranian: "The study of the lexicon of the Northern European languages, especially Germanic and Baltic, reveals that a large number of terms relevant to the ecology of the habitat of the early populations of the area and to their socio-economic activities have no plausible Indo-European etymology. (...) it is possible to ascribe to the pre-Indo-European substrate in the Baltic area a number of names of plants, animals, objects and activities characteristic of the Neolithic cultures." (Edgar C. Polomé: "The Indo-Europeanization of Northern Europe: the Linguistic Evidence", Journal of Indo-European Studies, fall 1990, p.331-337) Many of these terms also extend to Celtic, Slavic and sometimes Italic and Greek.

Examples include the words barley, Russian bor ("millet"), Latin far ("spelt"); Irish tuath, Gothic thiuda, "people", whence the ethnic names Dutch/Deutsch; German wahr, Latin verus, Old Irish fir, "true"; Latin granum, Dutch koren, English grain and corn; Lithuanian puodas, Germanic fata, whence Dutch vat, "vessel"; Dutch delven, "dig", Old Prussian (Baltic) dalptan, "piercing-tool"; Old Irish land, Old Prussian lindan, Germanic land; Latin alnus (<alisnos), Dutch els, Lithuanian elksnis, "alder", also related to Greek aliza, "white poplar"; Dutch smaak, "taste", Gothic smakka, "fig, tasty fruit", Lithuanian smaguricu, "sweet, treat"; from an ancient form *londhwos, Dutch lenden, Latin lumbus, "waist". Likewise, the Germanic words fish, apple, oak, beech, whale, goat, elm, (n)adder have counterparts in other European languages, e.g. Latin piscis, Old Irish aball, Greek aig-ilops or krat-aigos (possibly related to Berber iksir, Basque eskur, as suggested by Xavier Delamarre: Le Vocabulaire Indo-Euro