There were probably many languages and dialects in the Indus Valley region, with one standard variety located at the centre of the civilisation. But it would make sense for many of the little languages and dialects of the region to have belonged to a larger family, though there would probably have been some unrelated languages also looking on from the sidelines. In such great cities there would definitely have been local minority groups, as well as diplomats and traders from neighbouring lands who left behind communities to keep the relationship going.
Much of the early search for a lost Indus Valley language centred on the seals that were found when the old cities were excavated by archaeologists. These were small reliefs, many of which depict animals or people interacting with animals, besides more abstract markings that would appear to be “writing” or some form of notation. What could these seals be? Could they be people’s names? Perhaps the names of important merchants who needed to put a stamp on their transactions? Edicts? They were too short to be full sentences. And if they were words, what did they sound like? Were these markings meant to represent sounds? Or were they just symbols for concepts, like numbers?
Steve Farmer, Richard Sproat and Michael Witzel did not think they were words. In 2004 they went so far as to say that the seals were not writing at all, noting the “extreme brevity”of the Indus Valley inscriptions, with the longest text having only seventeen symbols, something “unparalleled in any literate civilization”, which led them to declare the Indus Valley Civilisation pre-literate. They announced a prize of $10,000 for anyone finding “just one inscription that contains at least 50 symbols distributed in the outwardly random-appearing ways typical of true scripts contemporary to the Indus system.”
It isn’t as though relics have never been decoded. The ancient Egyptian hieroglyphics all over the walls of their tombs, temples and monuments, and in old papyrus scrolls, were eventually deciphered, but it was because of one amazing piece of luck. In 1799 French soldiers retrieved a granite slab from a pile of rubble in the town of Rosetta, on the Nile delta. Carved into the stone surface were three texts, written in three different scripts. On top was a section in Egyptian hieroglyphics. Below it was text in Demotic script, an ancient Egyptian script used for more secular commerce. And at the bottom was a section in ancient Greek. Was it possible that all three texts contained the same message? Could the Greek and the Demotic scripts be a key to deciphering the Egyptian hieroglyphics?
It took time. First, the Demotic script had to be matched to the Greek. Then the names in the text had to be found and deciphered from the Demotic script. But it wasn’t just a matter of finding one-to-one correspondences. The hieroglyphics were, after all, in a language, and that language had a grammar. The word order and grammatical markers in the three texts would not be the same. And the Demotic script, unlike Greek, turned out to be written from right to left, and was only partly phonetic, as it included symbols derived from hieroglyphs.
It was only when French linguist Jean-Francois Champollion approached the problem through his knowledge of Coptic, an old language still used by the Coptic Christian Church in Egypt, that he was able to make out that the syllables written in the Demotic script were words he recognised from Coptic. The language of the hieroglyphics was a close cousin of the one used in the sacred texts of Coptic Christianity. Champollion had found not only the meanings of the words written in the hieroglyphics, but also how they had been pronounced.
It is lucky for Farmer, Sproat and Witzel that their challenge was so specific, finding a corpus of “50 symbols”, as there probably aren’t any. But tucked into their argument is the assumption that “writing” must mean something like prose, or poetry, statements or narrative bound by rules of grammar. Documents like the Rig Veda were, after all, preserved orally for millennia through elaborate methods of memorisation, so much so that we are now fairly certain of how the hymns sounded. We can treat the oral Rig Veda as a “text”. It was quite feasible for a civilisation, at that time, to do without written stories and “literature”.
But was early writing really about that? Was “literature” the first thing a society would wish to record in imperishable stone? And if a civilisation did not choose to preserve its stories and hymns in writing, what sorts of things would it want to note down if it had a complex economy with multiple transactions involving exact measurements and currency? Who were the people who would want to set things like this in stone?
It took a radical approach to the issue to make headway on deciphering meaning in the seals. And in 2019, Bahata Ansumali Mukhopadhyay, a software developer in Bangalore, published an article in Nature describing the seals as “formalised data carriers”, but not in a phonetic script. Mukhopadhyay’s engagement with the Indus Valley seals began in 2014, and over the years her approach matured into an article in 2022, posted on the Elsevier SSRN preprint server, where she identifies a symbol for “carat” based on the “ratti” seed, known in English as the Rosary pea, which has traditionally been used in India as a standard measure for weighing gold and measuring its purity, and she also posits a symbol made from the image of a crucible with a goldsmith’s blowpipe inside it as meaning “gold”.
I first saw her findings as a presentation from a video conference. I myself come from a family of traditional Indian goldsmiths, and everything she said was instantly familiar. Of course ,it would have to be commercial people working with the substance linked to currency, and requiring exactness and certification, who would be the ones to first need to note down this kind of information. As Mukhopadhyay adds, the actual physical crucibles and blowpipes and gold testing stones depicted in the seals were found in Mohenjo Daro, not in some rich person’s home, but along with unfinished jewellery pieces in a goldsmith’s workshop, where the tools looked just like the ones I grew up with. I saw in my mind’s eye the Paṇi, a mercantile community already living in the Indus Valley before the arrival of the Vedic men, and concerned with currency and the sorts of licence documents that allowed them to practise their trade. Early literacy would have been about that, not about stories.
Imagine, in a future age, archaeologists poring over our present-day barcodes and QR codes, or even the dhobi marks on our clothes, and wondering if they were “meaningful”, and whether they might be an ancient form of “writing”!
The favoured approach to finding the Indus Valley language has been by linguists: philologists who bypassed the tempting Indus Valley seals and went straight to the words in the Rig Veda that did not seem to come from early Sanskrit, words for new things the Vedic men must have come upon when they reached the Indus Valley. Masica had his list of words and their possible sources. The Finnish linguist Asko Parpola mentioned a number of words which, from their structure, seemed to him to be of Dravidian origin, and proposed Dravidian etymologies for these Sanskrit words. Witzel, a professor of Sanskrit at Harvard (the same Witzel who declared the Indus Valley Civilisation pre-literate) saw instead a “Para-Munda” origin for the words he found that hadn’t come from Sanskrit. And Franklin Southworth and David McAlpin posited “Elamo-Dravidian” as an origin, arguing that Elamite, an extinct language in southern Iran, might have been part of the Dravidian family. Munda languages trace back through their maternal line to the First People who came out of Africa, hunter-gatherers who reached the subcontinent 65,000 years ago, and Dravidian words go as far back as the first farmers who came to the Mehrgarh area of Balochistan in the northwest of the subcontinent about 9000 years ago from the Zagros mountains in southern Iran and mixed with First Indians. These two families would have to have been the first ports of call in any search for Language X.
Excerpted with permission from Father Tongue, Motherland: The Birth of Languages in South Asia, Peggy Mohan, Penguin India.
