Vikas Saraswat
Indus Script: The no script theory is a non starter
This article originally appeared in CRI content has now been subsumed in The views expressed here are personal and do not necessarily reflect those of the editors of

Did the Harappans of Indus Valley Civilization know writing? The question might come as a shock for many of us. The first retort to the question might well be “what then are the numerous Indus inscriptions retrieved from archaeological excavations?” Most of us interested in the subject have just been waiting for a consensus on the decipherment of script but none has ever contemplated the possibility of a “no script” theory. But the importance of this “no script” theory was validated by expectations and anticipations from a research paper published some time back.

A team of researchers led by University of Washington associate professor and computer scientist Rajesh Rao confirmed in a recent study that the Indus script did encode a language. The findings of the paper titled “Entropic Evidence for Linguistic Structure in the Indus Script”published in the Science journal Vol. 324 Issue 5926, April 24, 2009 and co-authored by Rajesh P. N. Rao, Nisha Yadav, Mayank N. Vahia, Hrishikesh Joglekar, R. Adhikari and Iravatham Mahadevan would have come as a mere statement of the obvious but for the significance the paper assumed in the light of an earlier  controversial research paper titled “The Collapse of the Indus Script Thesis: The myth of a literate Harappan civilization” published jointly by Steve Farmer, Richard Sproat and Michael Witzel in the Electronic Journal of Vedic Studies of 02/11/2004.

The second mentioned research paper would have also gone unnoticed had it not carried the inordinate credence that comes easily with the authority of high positions in prestigious Universities. Steve Farmer holds a Ph. D. in comparative cultural history from Stanford University. His colleague Richard Sproat of University of Illinois and the Beckman Institute is one of the leading computational linguists with a number of publications to his credit. The third of the trio, Prof. Michael Witzel, Wales Professor of Sanskrit at Harvard and editor of the prestigious Harvard Oriental Series needs no introduction amongst scholars or lay readers with the slightest interest in Indology.

However, Farmer and Witzel, besides their scholarship, are also known for their haughty demeanour and aggression in academic circles. Both of them are arduous supporters of Aryan Invasion Theory. Witzel in particular has also become notorious for his consistent anti Hindu rants too. He has remained as much in news for his academics as for championing the cause against “communal” Hindutva forces in India and their “obnoxious” limbs in USA. The political beliefs of authors and the bearings of their convictions on scholarship can be debated but their importance, particularly in the context of Indian History establishment, cannot be denied.

The enormous influence these authors carry with our establishment Historians is demonstrated in an instance where Prof. R.S. Sharma, founding chairman of ICHR, lapped upon a mistranslation of a verse from Baudhayana Shrauta Sutra by Prof. Witzel to claim the hitherto missing testimony of an Aryan Invasion from the corpus of Vedic literature. In a major embarrassment to Prof. Sharma and other AIT supporters, Witzel himself was forced to concede the mistake, which he later blamed on a printing error, when confronted by Belgian scholar Koenraad Elst but not before Prof. Sharma had gleefully reproduced it as “the most explicit statement of an immigration into the subcontinent” in his Advent of the Aryans.

Coming to the paper “The Collapse of the Indus Script Thesis”, the trio makes a sensational claim that Harappans of Indus Valley Civilization were illiterates, incapable of encoding language in the form of a script. Whatever Indus Valley inscriptions have been retrieved on clay seals, steatites, pots, potsherds, metal seals, copper plates, terracotta, ivory artefacts, rocks etc., are according to the authors, nothing more than “abstract religious-political signs,” serving the utility of a multilinguistic society. The authors are also convinced that these inscriptions would have been used for invoking magic and depicting sacrificial rituals.

The main contentions of their thesis are one, “the brevity of inscriptions”; second, the absence of archaeological evidence for possibility of writings on perishable materials and third; the “paradox between high sign repetition rates in the Indus corpus against the low sign frequency in individual inscriptions”.

It is a fact that the average length of Indus inscriptions, according to Mahadevan’s concordance arrives at a measly 4.6 signs length. The longest inscription on one surface carries only 17 signs and less than one percent of all catalogued inscriptions have a length of more than 10 signs. In fact the typical briefness of Indus inscriptions is indeed a major impediment in the decipherment of Indus Script. Farmer et al contend that the brevity of Indus inscriptions “is unparalleled in any literate civilization represented by even a fraction of the number of inscriptions in the Indus corpus.”

Contesting their argument, Dr Gunter Dryer, an Egyptologist holds that the mean word length of “comparable” Egyptian texts is 6.94 as against that of the Indus texts which is 7.39 showing no statistical difference. Dryer has also found phonetically readable Egyptian hieroglyphic texts with as few as two symbols. Asko Parpola, Finnish linguist from University of Helsinki with a formidable Corpus of Indus seals to his name is equally convinced that a logo-syllabic (combination of logograms and syllables) script, which many scholars hold the Indus script to be, of Sumerian type with an average sign length in Indus Script is quite sufficient to convey linguistic messages. Parpola says that even single character logo syllabic inscriptions representing composite signs consisting two or more components are enough to convey titles and nouns. Iravatham Mahadevan, the famous Dravidian linguist also dismisses the short inscription argument of Farmer et al. In his words, “seal texts tend to be short universally. Further, the Indus script appears to consist mostly of word signs (logograms). Such a script will have a lesser number of characters and repetitions than a (completely) syllabic script.”

Farmer et al’s argument concerning the ‘absence of expected archaeological evidence to support the “lost manuscripts thesis” also doesn’t impress many scholars. Farmer et al contend that if any lengthier documents on perishable materials had existed, as suggested by archaeologists and palaeographers to compensate for the short inscriptions on seals etc., some by-products of these Indus artefacts or markers such as writing paraphernalia or representations of scribes etc. should have been found in archaeological excavations. The manuscript tradition of Indus Valley writing, however, has been attested by the appearance of Harappan signs on later Indian pottery dating to 9th century B.C.E. at least.

BB Lal discovered that almost 90% of graffiti marks on megalithic red and black ware had affinity to Indus Valley signs. In fact it was this discovery which recommended the reading of Indus Script from right to left, a view later confirmed by Mahadevan. Parpola in his criticism of the paper mentions Yajnavalkya Smriti as well as the testimony of Alexander’s admiral Nearchus to cite the usage of cotton cloth as writing material from 500 B.C.E. onwards at least. He points out that the preserved examples of writing on cotton cloth appear, however, only after thirteenth century C.E. Similarly from the long Asokan inscriptions carved on pillars it can be safely presumed that similar or lengthier writings would have existed on perishable materials also, though no such specimens have ever been discovered.

There is nonetheless an interesting aspect to the argumentum ad absentium employed by scholars such as Farmer and Witzel. They take recourse in it when they find it convenient to promote their pet theories and prejudices as in the above case or as in the touted case of missing Horse seals (enough evidence from excavations at Surkotada-AK Sharma, Lothal-SR Rao, Kalibangan-Sharma, Aravalli Hills-Ghosh etc. has proved the domestication of equus caballus by Harappans ). On the other hand they refuse to consider the complete absence of any archaeological or literary evidence in support of Aryan invasion or migration in India.

In the third contention titled “paradoxical sign frequencies” of the paper which is dealt at quite a length, Farmer et al assert that “the high sign repetition rates in the Indus corpus overall contrast sharply with low sign repetition rates in individual inscriptions, which suggest that little if any sound encoding existed in the system”. The authors also note that a large number of signs in the Indus corpus (27% of Mahadevan’s concordance) occur only once in 13,372 sign occurrences and that 52% in the same concordance show up only five times or less. The authors suggest that in any evolving script the percentage of these singletons should be dropping progressively whereas this percentage, over the time, appears to be rising instead.

Yet again this assertion is contradicted by the observance of a high percentage of signs in Chinese writing system which are rarely ever used in Chinese newspapers. The high number of singletons has also been held by many scholars as representations of derivative signs inserted for the purposes of a clearer understanding of the texts. But it isn’t only the thrilling if not altogether incontrovertible arguments propounded by Farmer et al which merit serious criticism; their curt dismissal of “positional statistical regularities” in individual inscriptions as exaggeration also does injustice to the script theory. In fact “positional statistical regularities” of signs in Indus inscriptions strongly approves of a written script. Further, the authors ignore that the twenty or so high repetition signs in the Indus script are all very simplified signs clearly implying a higher stage of evolution of written script.

Moreover the “illiterate Harappans” model proposed by Farmer, Sproat and Witzel claims a rather asymmetric development of Harappans of Indus Valley Civilization. These Harappans had the best town planning of their times, fairly good knowledge of astronomy, an efficient transportation system as depicted in the terracotta figurines of the times, a wonderful irrigation system of canals, a perfect weights and measure system (the anna and pi) which was being followed till recently and extensive business with its literate Mesopotamian neighbours among various other accomplishments to their credit. But according to the authors, somehow, they neither could develop a written script of their own nor borrow it from their literate Sumerian or Akkadian neighbours.

To their credit the authors do not leave the issue unaddressed. Theorising on the possible reasons which could have forced the ruling Harappan elite (amounting to a rather amazing consensus among the large number of rulers, if the authors are not considering the entire one million square area of the Civilization as one Principality) to consciously oppose writing, they attribute it “to the threats (the writing could have) posed to whatever control the symbols gave them over Indus populations.” The reader is made to wonder if the introduction of written scripts in other ancient civilizations had caused the first “revolutions”!

The aggressive, provocative and sensational style of the paper and the brash conduct of the authors-they have announced an award of $10,000 for anyone finding an Indus inscription longer than 50 signs and termed the paper “Entropic Evidence…” as “…garbage in garbage out”-despite the faltering nature of their own hypothesis begs some explanation.

Western Indology with its roots in German nationalism, its concomitant anti Semitism, European Colonialism and Christian Evangelism, from its very inception has invested heavily in the Aryan Invasion Theory. In India it found ready cahoots in Marxist historians who had their own socio-political axes to grind. When the invasion model was found to be completely untenable, a variant called Aryan Migration Theory (more or less on the previous lines) hypothesising the immigration of Indo European Aryans from an unidentified location in Central Asia in small groups around 1500 BCE was mooted. These migrants then despite their low numeric strength, according to the hypothesis, somehow managed to impose their language and prevail culturally over the original inhabitants (Harappans) of Indus Valley. The descendants of these immigrant Aryans then went on to compose Rig-Veda in a language they had brought with them.

A number of bogus arguments such as the missing Horse seals, differences in the religions of Harappan and Vedic culture or the ignorance of iron amongst Harappans as against the knowledge of it and its prolific use by the Vedics have been advanced to create a cleavage between the Harappan age (before 1500 B.C.E.) and the Vedic age (after 1500 B.C.E.). However none of the arguments proffered for even an AMT can be validated. K.D. Sethna, David Frawley, Natwar Jha and N.S. Rajaram have shown that religious practices, rituals and cults of Harappans and Vedics were not unrelated and that the former, in fact, was a progression over the latter, thus completely overturning the carefully crafted chronology by invasionists and immigrationists. This reversed chronology offered by these scholars is supported by the recent multidisciplinary approach in the study of prehistoric India.

The missing Horse seal argument stands demolished in the light of Horse skeletons found in various excavations. The argument about iron presents an interesting case study in the kind of expedient speculation peddled as serious research by invasionists. The invasionist in this case is the Harvard Professor himself giving us a glimpse of his shoddy scholarship. For long, iron and its use in making weaponry was projected as an Aryan introduction to India. Recent excavations, however, have discovered iron artefacts at various Harappan sites and substratum. But Prof. Witzel who would not allow some obscure antique pieces to upset the reputations of invasionist Indologists claimed the iron of these artefacts to be of meteoric origin!

The entire Indo European invasion/migration theory is riddled with more of such absurdities and despite the Herculean efforts of committed scholarships the inherent paradoxes in the Harappan- Vedic relation refuse to be contained. One such apparent paradox which has come to be known as Frawley’s paradox points at a fundamental contradiction. David Frawley, Vedic scholar from U.S., points out that Harappans in the Indo European migration/invasion model have a wonderful civilization with sophisticated town planning, international trade, impressive craftsmanship, knowledge of metallurgy and what not but do not show up any literature to their credit.

On the other hand Vedic Aryans have a world class and prolific literature without a commensurate civilized Urheimat (a mythical homeland of the Indo European Aryans proposed by scholars obsessed with the Indo European model). The picture becomes clear when we view the two as one. Such a scenario however spells the collapse of AIT/AMT and doom for the upholders of this fantastic theory. Witzel, Farmer and Sproat’s assertion that the Harappans were illiterate is an exercise in reinforcing the delineation between Harappans and Vedic Aryans.

Rao et al’s paper uses the Markov model, a statistical method to estimate the meaning of unknown symbols in the context of known symbols, as a computational tool for investigating ancient scripts. The results of this study confirm that the Indus script signs show some order as well as flexibity. This according to the researchers is a typical feature of spoken languages which fall between the two extremes of strict order for pictorial representations and a random pattern for non-linguistic systems.

Most scholars agree that the paper, beyond its confirmation of a language in the Indus script, may not be much useful in deciphering the script as such. But the challenge to the decipherment comes from the entrenched political biases given to competing Proto Dravidian, Vedic Sanskrit and Proto Munda hypotheses for the Harappan language. Coming around a broad consensus in such a scenario is tough but the task calls for reappraisals of existing decipherments with open minds, arriving at common grounds to pursue further efforts and if needed, altogether fresh approaches rather than yield to the bunk “no script” hypothesis by Farmer, Witzel and Sproat.

(This response was written in 2009)