Since January 2002, the prominent Indian Newspaper, THE HINDU (OPEN PAGE, published on every Tuesday) has carried an extensive discussion on the relation between the Harappan (Indus) Civilization (2600-1900 BCE) and the Vedic Civilization (c. 1500-500 BCE).

This discussion has dealt among others, with the question of an overlap or even an identification between both cultures, as recently championed by some rewriters of Indian history. One of them, the scientist turned  historian, N.S. Rajaram, of  Harappan Horse Fraud fame (FRONTLINE, 2000), fired the first volley. All articles of this year's exchange are listed below.

Note: Reliable information on the Vedas is rare to find on the internet. Therefore it is useful to refer to the 100 pp. survey, linked above (Vedic civilization). It was written in 1992/3, but has not been published to far beyond some samizdat distribution among friends and students. About half of it will now come out in a volume  edited by A. Sharma, Studies on Hinduism, Univ. of S. Carolina Press 2003. The present version will be updated, eventually.

1/22/02   N.S. Rajaram : Historical divide: archaeology and  literature

1/29/02   M. Witzel   Indus Civilisation and Vedic society

2/5/02     Clarence Maloney : Vedic-Indus debate: save Indian  civilisation today

2/19/02   N.S. Rajaram: Theory and evidence

3/5/02     M. Witzel :  Harappan horse myths and the sciences

3/12/02   R. Nagaswamy: Harappan horse

5/21/02   M. Witzel (assisted by Richard Meadow): Horses, logic, and evidence

6/18/02   David Frawley: Vedic literature and the Gulf of Cambay discovery

6/25./02   M. Witzel :  A maritime Rigveda? How not to read ancient texts

7/2/02 R. Nagaswamy : From Harappan horse to camel

7/9/02     Patrizia Norelli-Bachelet :   Cosmology in Rigveda the third premise

7/16/02 D.,Frawley: Witzel's vanishing ocean

8/6/02  M. Witzel Philology vanished: Frawley's Rigveda I

8/13/02  M. Witzel: Philology vanished: Frawley's Rigveda II

8/20/02 D.Frawley: Witzel's philology

The last article by Frawley, especially his various allegations, selective quotations and the resulting distortions, certainly is in need of  some answer;
I will come back to the whole question at an opportune occasion. -MW.  -- PS. :   see now below, Feb.11, 2003, and in detail: April 1, 2003

Rajaram and Frawley, however, continue their quest for the remotest imaginable, hoary date of "Indian Civilization"
(cf. the Antiquity Frenzy elsewhere),  as seen at:

12/3/02: The Hindu, (Nation): New light on south Indian civilisation

Answered on 12/4//02 by M., Witzel:  Rewriting  History

1/21/2003 D.Frawley: An ecological view of ancient India

2/11/2003 M. Witzel  Ecology, rhetoric or dumbing down? -- I

2/18/2003  M.Witzel :  Ecology, rhetoric or dumbing down? --II

3/18/2003  N.S. Rajaram:  Paradigm shift in history

4/1/2003  M.Witzel :'Paradigm shift' in history? I

4/8/2003  M. Witzel: 'Paradigm shift' in history? II

Full uncut text of:

Answer to Nagaswamy,
by Michael Witzel (assisted by Richard Meadow):
Horses, logic, and evidence


Horses, logic, and evidence

The vexed question of the 'Harappan Horse' does not seem to go away. I am glad to see that my last paper in OPEN PAGE has elicited such a strong reaction, this time from a professional archaeologist, Dr. R. Nagaswamy, the former Director of Archaeology, Tamil Nadu. He has narrowed down the long debate about the Indus and Vedic civilizations to the question of horses in the archaeological record, about which more below.

To begin with, Nagaswamy and I certainly agree about the need of 'rewriting' ancient Indian history from time to time. This is a normal procedure in historiography when new data have been discovered that result in the need for new interpretations.

I also agree that this "should be entrusted to an unbiased and balanced academic body free from racial, religious or political bias" as well as to individual, well prepared scholars who can contribute their insights. Certainly, we can also agree that "South Indian history [has] receive[d] inadequate representation" so far. However, it will be difficult to "confine such debates to post-graduate community of the country and [as to assure] our children are told only the factual history." What is factual history? And how does one decide this? Obviously, interpretation of various types of data is involved here, and the results of such discussions printed in history books will be based on extensive debate and a certain amount of scholarly consensus.

We both would also agree that such interpretations should not be long discarded ones, e.g., those of M. Wheeler ("Indra stands accused" referring to the destruction of the Indus cities). However, they should also not be such as the recent fantasies of N.S. Rajaram (et al.), whose "decipherment" of the Indus script and all interpretations that flow from this are a priori wrong because he uses the wrong direction of the script, a direction that was securely established decades ago by the former Dir. Gen. of Archaeology, B.B. Lal (see also I. Mahadevan, at the Indian History Congress, Dec. 2001, and in EJVS 8-1,

"Digs" at the writings of persons such as Rajaram thus have their well-justified reasons. Especially so, as he has recently been appointed by the Minister of HRD as a member of the ICHR, just after I sent in my last piece. Such appointments spell the end of scientific history writing and open the door to Raamraaj fantasies galore. However, since I am neither an Indian citizen nor a professional historian, I refrain from further comment.

Second, there is the vexed question of the "Harappan horse". Though merely being a philologist and something of linguist, I have read the relevant scientific literature, and I have additionally asked some scientist colleagues, just to be 100% sure. For my last piece in OPEN PAGE I consulted archaeologists and archaeozoologists, upon whose advice I have changed two or three sentences. I showed my piece, for example, to my colleague at Harvard, R. Meadow. He is both an archaeologist and an archaeozoologist (one who studies animal bones from archaeological sites). In addition, he happens to be a Director of the Harappa Archaeological Research Project, where he has been digging for more than a decade; he also knows first hand many other early sites, e.g., Mehrgarh and Pirak in Baluchistan.

Amusingly, it is exactly the very sentence that the archaeologist Nagaswamy criticized that I changed based on input from the Harappan archaeologist R. Meadow, namely that horse bones are likely to have "found their way into deposits through erosion cutting and refilling, disturbing the archaeological layers." Nagaswamy complains that "neither does he say how he arrived at this conclusion nor has he cited any report in support of his view." Unfortunately my references for this were cut by editors; here are the details: R. Meadow in: The Review of Archaeology, 19, 1998, 12-21; R. Meadow in: D.R. Harris (ed.) The Origins and Spread of Agriculture and Pastoralism in Eurasia.1996, 390-41. UCL Press, London; R. Meadow and A. Patel in: South Asian Studies,13, 1997, 308-315.  Here for Nagawamy then are the references to the still elusive "Harappan horse."

If Nagaswamy had read such relevant technical literature himself he would have seen why the archaeozoological situation is as bad as I described it. But since he has not, he perfectly exemplifies the case I made, namely that archaeologists do not necessarily know enough of archaeozoology, and their judgments thereof can be flawed. For example, G.M. Pande's book (The Dawn of Indian Civilization,Delhi 1999, quoted earlier) mentions many times just "horses," without any further specification of equid type or any discussion. But, this kind of statement is believed at face value, as I have seen, by both the general and scholarly public.

Unfortunately Nagaswamy has also not checked the archaeological sources well enough. When he says that "there are three major excavations conducted at Mohenjodaro and Harappa namely by Marshal, Mackey and Mortimer Wheeler... George F. Dales, who was the last in the series to investigate the sites..." (1982), apart from contradicting himself within a few words, he simply overlooks decades of excavations at Harappa (and elsewhere!), -- excavations that have been reported in archaeological publications and even in a recent generally available synthesis by J. M. Kenoyer, Ancient Cities of the Indus Valley Civilization. Oxford 1998.

All of this should perhaps not surprise. As far as I can see, R. Nagaswamy has proficiently published on Tamil literature, culture, and archaeology, -- including even one excavation report (about Vasavasamudram), written together with Abdul Masjeed in 1978. This lacuna may be the reason why he resorts to a discussion of "logic" instead.

However, the evaluation of archaeological / archaeozoological detail cannot be reduced to "logic" by comparing various sentence from my paper, as he does: "Such arguments are brought under reductio ad absurdum by logicians". Instead, to state it clearly one more time: what we need is absolute security (to the degree scientifically possible) with regard to identification of the bones themselves, radiocarbon dates directly on the bones in question, and a detailed understanding of the archaeological context from which they come. It is evidence that counts, not a "logic game". Nagaswamy even distills from my paper that "it only shows that horse bones were actually found in the excavations at Harappan sites." But, this is precisely what was and still is under discussion! Even if we accept the identifications as true horse of material from the old excavations (and this still needs to be rechecked by specialists using the original material), we lack the other two pieces of information -- context and direct date -- that are necessary to securely interpret the cultural meaning of those identifications.

As scholars know, archaeology just as paleontology (or linguistics) is all about location, context, and stratification. An isolated horse skeleton, complete or not, just shows that a horse was present in that location at some time in the past. As an example, a nearly complete camel bone (humerus) was found at Harappa about a meter below the surface together with artifactual remains that could be assigned by archaeologists to the last part of the "mature" Harappan period (ca. 2200-1900 BC). A direct radiocarbon date on the bone itself, however, came out to be 690 CE, more than two thousand years later (Meadow 1996 & 1998): a clear case of a deposit effected "through erosion cutting and refilling." If, in addition, that find is not recorded in detail (and actually reported in print!) it is almost worthless, just as in linguistics a modern Sanskrit word like kendra"center," could go back to a formation based on the modern English 'center' or could be derived from a Greek word, (kentron)-- which is the case, as it is already attested by Varahamihira at 550 CE. The levels are clear here.

But bones from old and not well recorded excavations (Nagaswamy himself quotes some doubts in Dales' review of Indus excavations!), bones that in addition have not been radiocarbon dated and may not have been correctly identified through comparison with modern specimens, just do not provide suitable evidence. The alleged find of a horse or camel bone at Harappa without details of the identification, the context, and a direct date then is open to question: yesterday's Afghan or Sikh sipahi'shalf-horse can be turned into tomorrow's full Harappan horse... In sum, in the sciences we cannot work with data that do not conform to strict procedures and methods, such as those delineated above.

Third, as for the discussion of various equids (true horse, ass, half-ass/onager), Nagaswamy complains that I do "not show whether R. Meadow ... based his views on a full skeleton or full sets of onager, donkey, or horse skeletons." Again Meadow (and I, quoting him) has explained fully in his 1996-1998 papers why such detailed comparison is necessary, and that good collections of modern specimens, necessary to compare such finds, are only just now being established in South Asia itself.

Fourth, when true horses are indeed found, finally, in the Kachi Plain of East Baluchistan at c. 1800 BCE, this does not make them Harappan, as Nagaswamy maintains: "...1800 BCE, which still falls in the late Harappan period". A study of the Pirak data would show that these sites are post-Mature-Harappan and have a very different material culture inventory than the earlier Harappan sites in the same area. To lump all cultures from 3500 to 1500 BCE together as "Harappan" (or even as "Sarasvati" -- before its actual naming!) is not correct, and opens the door to all sorts of unscholarly rewritings of ancient history, in other words, to new myth making.

In sum, in my recent OPEN PAGEarticle, I intentionally preferred to err on the side of caution: Advised by specialists, I specified the conditions that are necessary to identify horses, and I did not simply follow the conclusions of various writers as to the nature of one set of bones or the other. To repeat it one final time: we need the bones in contention 1) to come from well stratified deposits the formation of which is understood, 2) to be carefully identified with the reasons for their identification published in detail, and 3) to be directly radiocarbon dated if possible. At the present time, these conditions have not been met well, and all conclusions must be explicitly preliminary or else they can only be called speculative or misleading.

If this stance is called "taking a partisan view", then I may gladly be called partisan. This certainly is better than to find, with Rajaram et al., horses (or "fire altars") all over the early subcontinent (see Frontline,Oct./Nov. 2000). Such writers see horses everywhere, just as some see Krishna everywhere: "He who sees me everywhere and sees everything in me..." (Bhagavad-Gita 6. 30) Or at least, Rajaram did so. By now, in messages sent to his followers, he has twice has given up pursuing any further discussions.

To sum it up once more, we can certainly agree to rewrite sections of ancient Indian history, but this has to be done on the basis of new facts, not of new myths or of incomplete archaeological or zoological data. It is the duty of scholars to point out such new myths before they enter the new textbooks. And with this, I leave the field to the scientists.

Michael Witzel
(assisted by Richard Meadow)
Harvard University