A: My early writings were a little premature, I now realize. I was over-enthusiastic. Like most other scholars in the field, I realized as time went on that it is an extraordinarily difficult problem. And to begin at the end, I may say that I realize that I have not deciphered the Indus script and if I may so, it is extremely unlikely that I may do so in the remaining years of my life. But this is not to say that we do not know anything about the Indus script or that the script is indecipherable.

Q: What were your first experiences of looking at the signs?

A: The first time I came across specimens of the Indus script was by accident. I had taken a book out of the library of the Central Secretariat, Govt. of India, and that happens to be an important book, Hunter's concordance [Script of Harappa and Mohenjodaro and its connection with other scripts, 1934], which he published for his doctoral thesis at Oxford University. [G.R.] Hunter was an Englishman who was in the Indian Educational Service. He went to Harappa and Mohenjo-daro when the excavations were going on and he copied all the freshly excavated material. He was an extraordinarily careful draughtsman. Even now it is realized that his eye copies have been the best until Parpola and his colleagues brought forth the modern magnificent editions with the photos of the seals themselves. When I went through Hunter, I realized that it was an interesting problem in Indian epigraphy. My earliest attempt was that if it could be deciphered, it should represent only one of the Indian languages, either the Indo-Aryan group or the Dravidian group.

Aryan Immigration and the Brahuis

Now most scholars of the world believe that the Aryan immigration is later than the mature part of the Indus civilization. That ruled out Indo-Aryan as a possibility. Now we also know that before the Aryan advent into the subcontinent, the Dravidians lived in the north and northwest parts of the country. We have the Brahui, which is still spoken in Baluchistan and parts of Iran and nearby areas. It is a Dravidian language. Of course there has been a suggestion that the Brahuis could have migrated from south of the peninsula in later historical times, but experts discount this possibility because it is found by Burrow, Emeneau and others that Brahui is part of the North Dravidian group of languages which still survives in pockets like Kurukh and Malto in the north-east part of India. Therefore the chances are that the ancestors of the Brahui lived there during the time of the Harappan civilization. We know from recent excavations at Mehrgarh in Pakistan and elsewhere that the roots of the Indus culture lay in the soil, that they have been there for millennia, from at least the 8th millennia B.C., from the late Neolithic through the pastoral period this Bronze Age civilization blossomed. This also makes it unlikely that this civilization could have been brought in by immigrants.

There are still scholars, like Prof. Kinnier-Wilson of Cambridge, who believe that Harappan is a form of Sumerian. But there are not many takers for this theory, because whatever we know of the Indian languages in the northwest, there are and there have been Dravidian and Indo-Aryan languages. We see no evidence of a Sumerian or Akkadian substratum to the modern Pakistani and north Indian languages.

So I still very strongly believe that the Indus civilization language was in all probability an early form of Dravidian. Having said this, let me also sound a word of caution. This is still a theory. We haven't had final proof, we haven't been able to crack the code primarily because we do not have a bilingual [inscription in two languages] and also because the available inscriptional materials are all in the form of repetitive tablets and seals which are extremely small, not more than an average of five symbols strung in a row and thus very unlikely to have anything more than names and titles. Unlike the clay tablets from Babylon we do not have long narratives, which makes it very difficult to reconstruct the code.

It is true, as an axiom of cryptography that given adequate material no code can resist decipherment. But the problem here is that the material is not adequate.

Today you have brought to me from Harappa as a gift from Prof. Kenoyer, drawings of a number of new tablets from Harappa. Looking at them I find that most of them are repetitions of finds already well known and included in Parpola's volumes. So the chances of a long connected narrative seem to be slim. There is always the possibility that somewhere along the Makran coast [in Baluchistan] or even in the Middle East a bilingual seal or even a bilingual clay tablet could be found. It is known that traders from Meluhha [the Mesopotamian word for the Indus Valley] went to the Middle East and set up colonies there, but this is looking into the future. As of now all the work that has been done can only be said to be of a tentative character.

We can say that the language is most likely to have been a form of Dravidian, and we can also say - and this with certainty - that the script is written from right to left, like Arabic for example and unlike the Indian [Sanskrit] script, though there are the occasional cases of left to right and a few cases of boustrophedon where the writing alternates from right to left, then left to right and again from right to left. But about 90 percent of the writing seem to be from right to left and when I say from right to left I mean the direction you see on the impression as all seals are negatives and are meant to take an impression from. Now this is also proved by the boss on the back of the seal. Obviously every Harappan of any consequence carried a seal on his person, strung it on a cord and hung it around his neck and used it to attest his documents, his sale or the nobles, their orders. In this sense the seals of Harappa and Mohenjo-daro are no different from those found in Sumerian, Akkadian and Egyptian cities. That about sums up the present position.