Online edition of India's National Newspaper
Tuesday, Mar 05, 2002

About Us
Contact Us
Open Page Published on Tuesdays

Features: Magazine | Literary Review | Life | Metro Plus | Open Page | Education | Book Review | Business | SciTech | Entertainment | Young World | Quest | Folio |

Open Page

Harappan horse myths and the sciences

The horses found in the early excavations at Mohenjo-Daro and Harappa do not come from secure levels and such `horse' bones, in most cases, found their way into deposits through erosional cutting and refilling, disturbing the archaeological layers.

In the Open Page of February 19, N.S. Rajaram posits a truism "A theory must not contradict empirical facts," but he then does not deliver on the `empirical facts.' As a scientist, he must suffer to be corrected, bluntly this time, by a mere philologist and Indologist. Philology, incidentally, is not the same as linguistics, as he says, but the study of a civilisation based on its texts. In order to understand such texts, one must acquire the necessary knowledge in all relevant fields, from astronomy to zoology. It is precisely a proper background in zoology, particularly in palaeontology, that is badly lacking in Rajaram's, the scientist's, account. Instead, it is he, and not his favourite straw man, the Indologist, who has created some new "myths and conjectures ... through the force of repetition." Let us deconstruct them one by one.

Harappan horses?

To begin with, he claims that "both the spoke-wheel and the horse were widely used by the Harappans." He quotes S.P. Gupta, without naming him, from a recent book (The Dawn of Indian Civilisation, ed. by G.C. Pande, 1999). According to Gupta the horse (Equus caballus) "was widely domesticated and used in India during the third millennium BC over most of the area covered by the Indus-Sarasvati (or Harappan) Civilisation. Archaeologically this is most significant since the evidence is widespread and not isolated." Nothing in this assertion is correct, even if — or rather because — it comes from an archaeologist and inventive rewriter of history, S.P. Gupta. For example, the horses found in the early excavations at Mohenjo-Daro and Harappa do not come from secure levels and such `horse' bones, in most cases, found their way into deposits through erosional cutting and refilling, disturbing the archaeological layers.

Indeed, not one clear example of horse bones exists in the Indus excavations and elsewhere in North India before c. 1800 BCE (R. Meadow and A. Patel 1997, Meadow 1996: 405, 1998). Such `horse' skeletons have not been properly reported from distinct and secure archaeological layers, and worse, they have not been compared with relevant collections of ancient skeletons and modern horses (Meadow 1996: 392). Instead, well recorded and stratified finds of horse figures and later on, of horse bones (along with the imported camel and donkey), first occur in the Kachi plain on the border of Sindh/E. Baluchistan (c. 1800-1500 BCE), when the mature Indus Civilisation had already disintegrated.

Even more importantly, the only true native equid of South Asia is the untamable khur (Equus hemionus, onager/half-ass) that still tenuously survives in the Rann of Kutch. Both share a common ancestor which is now put at ca. 1.72 million years ago (while the first Equus specimen is attested already 3.7 mya.). The differences between a half-ass skeleton and that of a horse are so small that one needs a trained specialist plus the lucky find of the lower forelegs of a horse/onager to determine which is which, for "bones of a larger khur will overlap in size with those of a small horse, and bones of a small khur will overlap in size with those of a donkey." (Meadow 1996: 406).

To merely compare sizes, as Rajaram does following the dubious decades old Harappan data of Marshall, and then to connect the long gone "Equus Sivalensis" with the so-called "Anau horse", resulting in the "Indian country" type, is just another blunder, but Rajaram, the scientist, is not aware of it.

Proper judgment is not possible as long as none of the above precautions are taken, and when — as is often done — just incomplete skeletons or teeth are compared, all of which is done without the benefit of a suitable collection of standard sets of onager, donkey and horse skeletons. Rajaram and his fellow rewriters of history thus are free to turn any local half-ass into a Harappan horse, just as he has already done (see Frontline, Oct./Nov. 2000) with his half-bull.

Further, the archaeologists claiming to have found horses in Indus sites are not trained zoologists or palaeontologists. When I need to get my teeth fixed I do not go to a veterinarian or a beauty salon. Typically, S.P. Gupta (1999) does not add any new evidence, and just repeats palaeontologically unsubstantiated claims that are, to quote Rajaram, "myths and conjectures... through the force of repetition."

The Siwalik equid

In addition, Rajaram conjures up another phantom, the Siwalik horse: "fossil remains of Equus Sivalensis (the `Siwalik horse') show that the 34-ribbed horse has been known in India going back tens of thousands of years." Standard palaeontology handbooks (B.J. MacFadden, Fossil Horses, 1992) would have told him that the Siwalik horse, first found in the northern hills of Pakistan, is not just "going back tens of thousands of years" but is in fact 2.6 million years old. However, it has long died out during the last Ice Age, as part of the late Pleistocene megafaunal extinction of about 10,000 years ago (i.e. at the end of the Late Upper Pleistocene, 75-10,000 y.a.: it is reportedly found in middle to late Pleistocene locations in the Siwaliks and in Tamil Nadu, and recently, as a "Great Indian horse" in Andhra, 75,000 y.a.). But there is, to my knowledge, no account of a Siwalik horse that even remotely approaches the date of the Indus Civilisation — nor does Rajaram quote any authority to this effect.

Nevertheless, in order to bolster his claim for the antiquity of the "Vedic horse (as) a native Indian breed", he connects this dead horse with the Rigvedic one, which is described as having 34 ribs (Rigveda 1.162.18). But, while horses (Equus caballus) generally have 18 ribs on each side, this can individually vary with 17 on just one or on both sides. This is not a genetically inherited trait. Such is also the case with the equally variable (5 instead of 6) lumbar vertebrae, as found in some early domestic horses in Egypt (2nd. mill. BCE) and in the closely related modern Central Asian Przewalski horse (which shares the same ancestor, 620-320,000 years ago, with the domestic horse/Equus ferus).

As for the number 34, numeral symbolism may play a role in this Rigveda passage dealing with a horse sacrificed for the gods. The number of gods in the Rigveda is 33 or 33+1, which obviously corresponds to the 34 ribs of the horse, that in turn is speculatively brought into connection with all the gods, many of whom are mentioned by name (Rigveda 1.162-3). But this is mere philology, not worthy of "scientific" study...

In sum, even S. Bokonyi, the palaeontologist who sought to identify a horse skeleton at the Surkotada site of the Indus Civilisation, stated that "horses reached the Indian subcontinent in an already domesticated form coming from the Inner Asiatic horse domestication centers" — just as they were imported into the ancient Near East about 2000 BCE. Any zoological handbook would have told the scientist Rajaram the same (MacFadden 1992).

In addition, the identification the Surkotada equid as horse by S. Bokonyi is disputed by R. Meadow and A. Patel (1997). Even if this were indeed the only archaeologically and palaeontologically secure Indus horse available so far, it would not turn the Indus Civilisation into one teeming with horses (as the Rigveda indeed is, a few hundred years later). A tiger skeleton in the Roman Colosseum does not make this Asian predator a natural inhabitant of Italy. In short, to state that the "Vedic horse is a native Indian breed and not the Central Asian horse" is just another fantasy of the current rewriters of Indian history.

Nevertheless, Rajaram even repeats some of his own "myths and conjectures, (which) through the force of repetition, have come to acquire the status of historical facts," namely the old canard that "depictions of the horse are known at Harappan sites, though rare" — a case of fraud and fantasy that has been exploded more than a year ago in Frontline (Oct./Nov. 2000). Apparently, he thinks, along with other politicians, that repeating an untruth long enough will turn it into a fact.

Spoke-wheeled chariots

Rajaram, in dire need of `Rigvedic' horse-drawn chariots for the Harappan period, then introduces spoked wheels into the Indus Civilisation: "terracotta wheels at various Harappan sites. ... The painted lines (spokes) converge at the central hub, and thus leave no doubt about their representing the spokes of the wheel."

The handful existing specimens of such terracotta disks may indeed look, even to a trained archaeologist, like a spoked wheel — especially when he wants to find Aryan chariots, just like Aryan fire altars, all over the Indus area. But, they may just as well have been simple spindle whorls, used in spinning very real yarn, not wild Aryan tales. Further, "spoked wheel patterns" occur in cultures that never had the wheel, such as pre-Columbian North American civilisations. In other words, all of this proves nothing as long as we do not find a pair of these "spoked wheels" in situ, along with a Harappan toy cart. Normally, the wheels of such toy carts are of the heavy, full wheel type (that is made of three interlocked wood blocks).

Rajaram then asserts, for good measure, that the "depiction of the spoke-wheel is quite common on Harappan seals." This refers to the wheel-like signs in Harappan script. Unfortunately, these "wheels" can easily be explained as unrelated artistic designs (like in the N. American case). Worse, they mostly are oblong ovals, not circles. A Harappan businessman using a cart with such wheels would have gotten seasick pretty soon. They are unfit for travel — and for the discerning reader's consumption.

Instead, the rich Rigvedic materials dealing with the horse-drawn chariot and chariot races do not fit at all with Indus dates (2600-1900 BCE) and rather put this text and its chariots well after c. 2000 BCE, the archaeologically accepted timeframe of the invention of the spoke-wheeled chariot in the northern steppes and in the Near East. Again, Rajaram's fantasised "Late Vedic" Indus people have scored a "first": they invented the chariot long before archaeologists can find it anywhere on the planet!

"Aryan" chariots

There is no need to go deeply into his building up the straw man of Aryan invasions (i.e. immigration of speakers of Indo-Aryan), involving a need to "prove that the Vedas are of foreign origin." No one today maintains such a theory anyhow. Instead, the Rigveda is a text of the Greater Punjab, indicating a lot of local acculturation but using a language and poetics that go back to the earlier Indo-Iranian period in Central Asia (c. 2000 BCE).

Equally misleading is his caricature: "without the horse and the spoke-wheel the Harappans were militarily vulnerable to the invading Aryan hordes who moved on speedy, horse-drawn chariots with spoke-wheels." As has been mentioned here a few weeks ago, nobody today claims that the Indo-Aryan speakers arrived on the scene when the mature Indus Civilisation still was flourishing and destroyed it, it in whatever fashion. Instead, there is a gap of some centuries between the two cultures, as the descriptions of ruins and simple mud wall/palisade forts (pur) in the Rigveda indicate. Vedic texts tell us that the pastoralist Indo-Aryan nobility fought from chariots, and the commoners on horseback and on foot, with the local people (dasyu) of the small, post-Harappan settlements who, like the Kikata, are said not even to understand "the use of cows." Next to warfare there also was peaceful acculturation of the various peoples in the Greater Punjab, as is shown by the Rigveda itself.

As for a chariot use, a brief study of ancient Near Eastern warfare would have done the `historian' Rajaram some good. It is clear to even a superficial reader that after c. 1600 BCE the Hyksos, Hittites, etc., used such chariots, not just for show and sport but also in battle, such as in the famous battle of Kadesh between the Hittites and Egyptians in 1300 BCE. Chariots were in fact used as late as in Alexander's battle with Poros (Paurava) in the Punjab, or by the contemporary Magadha army with its 3,000 elephants and 2,000 chariots. Why then all this diatribe about the "Aryan" use of chariots in favourable, flat terrain? (Not, of course, while "thundering down the Khyber Pass"!)

Foray into linguistics

Mercifully, Rajaram has spared us, this time, his usual assaults on the "pseudo-science" of linguistics, and instead tries his own hand at it, and teaches us some Dravidian: kudirai `horse,' which should prove that the horse has been native to South India forever. However, his foray into linguistics is incomplete and misleading.

First, Tamil kutirai, Kannada kudire, Telugu kudira, etc. have been compared by linguists, decades ago, with ancient Near Eastern words: Elamite kutira `bearer', kuti `to bear.' The Drav. words Brahui (h)ullii `horse' and Tam. ivuLi are derived from `half-ass, hemion' (T. Burrow in 1972). Both words, far from being `native South Indian', thus were coming in from the northwest.

Second, other Indian language families have such `foreign' words as seen in Munda (Koraput) kurtag, (Korku) gurgi, kurki, (Sabara/Sora) kurtaa, (Gadaba) krutaa, which are all derived from Tibeto-Burmese, for example Tsangla (Bhutan) kurtaa, Tib. rta. We know that Himalayan ponies have always been brought southwards by salt traders and with them, of course, their names. There also is the independent and isolated Burushaski (in N. Pakistan) with ha-ghur, cf. Drav. gur- in Telugu guRRamu, Gondi gurram, etc., and the Austro-Asiatic Khasi (in Shillong) kulai, Amwi kurwa', etc., — all of which again point to a northern origin. (For details see: EJVS 5-1, Aug. 1999,, or: International Journal of Dravidian Linguistics, 2001).

Far from magically proving, with one Dravidian word, that the "native Indian horse" has been found in the South since times immemorial, the "man made theory" of linguistics --just as the hard facts of palaeontological science — rather indicate that the words for `horse' were imported, along with the animal, from the (north)western (Iranian) and northern (Tibetan) areas. Genetics now add another facet. The domesticated horse seems to have several (steppe) maternal DNA lines (Science 291, 2001, 474-477; Science 291, 2001, 412; cf. Conservation Genetics 1, 2000, 341-355), which fits in very well with the several northern Eurasian words for it mentioned above. The Eastern Central Asian words must be added; they all probably derive from Proto-Altaic *mori (as in Mongolian morin, Chinese ma, Japanese uma, and as surprisingly also found in Irish marc, English mare).

The Harappan "Sarasvati"

The case of the Vedic Sarasvati river (the modern Sarsuti-Ghagghar-Hakra) is complex and cannot be dealt with in detail (see, rather, EJVS 7-3, section 25). It must be pointed out, however, that the Rigvedic Sarasvati is a river on earth, a `river' in the sky (Milky Way), and a goddess, and as such Sarasvati is described in superlative terms, once as flowing `from the mountains to the sea' (samudra). However, this word has several meanings that must be kept apart: `confluence, lake, mythical ocean surrounding the earth'; the sky, too, is called a `pond'! To commingle all of this as samudra `Indian Ocean' is bad philology.

In addition, far from emptying into the Rann of Kutch then, the Harappan Sarasvati (`having lakes'), disappears as Hakra in the dunes around and beyond Ft. Derawar in Bahawalpur, after showing signs of a delta (playa) and of terminal lakes, just like its Iranian namesake in the Afghani desert, the Haraxvaiti (Helmand) with its Hamun lakes.

Further, simple satellite photographs also do not show when a river dried up, as the Ghagghar-Hakra has indeed done several times in its different sections in recent millennia. This was shown in detail for the Indus and Vedic periods by the former director of Pakistani archaeology, Rafique Mughal, in his book Ancient Cholistan (1997). Rajaram again is simply wrong as a scientist in asserting that the river conveniently "dried up completely by 1900 BC." Reality is much more complex.

Actually, much of this has been known since Oldham and Raverty (1886, 1892). (Thus, I myself have printed a Sarasvati map, based on a lecture of 1983, before the overquoted satellite photos of Yash Pal et al. were published in 1984). However, we need many more close observations such as Mughal's, with archaeologically vouched dates for the individual settlements along the various sections and several courses of the river.

Finally, the "oceanic descriptions" of the Rigveda imagined by Rajaram and many other rewriters of history (such as S.P. Gupta, Bh. Singh, D. Frawley) are based, again, on bad philology: their "data" are taken from Vedic mythology, floating in the night time sky, and the like! Or was Bhujyu abducted on another first, a Vedic airship?


Harvard University

Send this article to Friends by E-Mail

Open Page

Features: Magazine | Literary Review | Life | Metro Plus | Open Page | Education | Book Review | Business | SciTech | Entertainment | Young World | Quest | Folio |

The Hindu Group: Home | About Us | Copyright | Archives | Contacts | Subscription
Group Sites: The Hindu | Business Line | The Sportstar | Frontline | Home |

Comments to :   Copyright 2002, The Hindu
Republication or redissemination of the contents of this screen are expressly prohibited without the written consent of The Hindu