A FEW YEARS AGO, Bahata Ansumali Mukhopadhyay, a Bengaluru-based software designer who has been working independently on decoding the mysterious script of the Indus Valley Civilisation (IVC), began to wonder if there wasn’t another way of approaching the question of what language was spoken in the IVC.
Most researchers have tried to find this answer in the Indus script, the mysterious inscriptions found on seals and terracotta tablets that depict pictographic signs, and human and animal motifs, including even a unicorn. But despite over a century of sustained attempts, the script has remained stubbornly undecipherable.
Mukhopadhyay argues in a paper in Palgrave Communications that the Indus script does not phonologically encode words. She believes it is logographic or semasiographic in nature, its many symbols representing concepts and ideas, the iconographies on the seals for instance functioning in the way the Sarnath lion emblem adorns our coins.
So, a few years ago, Mukhopadhyay came up with another idea: since IVC was known to have had elaborate trade links with farflung places such as Mesopotamia and the Persian Gulf, why not instead look up the languages spoken at that time there, keeping an ear out for potential loanwords from the IVC? Could words have travelled along with goods from the IVC? Could they be hiding in plain sight?
“My idea was since the Indus Valley residents were trading so heavily, if there is a commodity found here but not in places like Mesopotamia, that is to say those commodities were exclusively from the IVC, then the names [of those commodities] could have been used there,” she explains.
Mukhopadhyay made a list of products that Mesopotamia, in all likelihood, sourced from the IVC and began to reference works on ancient Eastern grammar and archaeological works. And then she hit a potential jackpot.
She found that the words for elephant (in Bronze Age Mesopotamia) and ivory (in Old Persian documents dated to the 6th century BCE) were connected to words currently in use in modern-day Dravidian languages in south India. These findings thus suggested the likelihood of the IVC being peopled by—or at least substantially peopled by—a proto-Dravidian-speaking population.
“The logic is that when we import a foreign commodity not locally produced, we usually call it by its foreign name. This intuitive approach has been duly rewarded, as it is found that the words ‘‘pīru’/‘pīri’ and their various dialectal variations, which signified elephant in Akkadian and ivory (‘‘pīrus’) in Old Persian, are perfect tools for the present endeavour…In several Dravidian languages, ‘pīlu’, ‘pella’, ‘palla’, ‘pallava’, ‘piḷḷuvam’, ‘pīluru’, etcetera, signify elephant,” she writes in her paper in Humanities & Social Sciences Communications.
The words pīru/pīri and their various dialectal variations, which signified elephant in Akkadian and ivory (pīrus) in Old Persian, were originally loanwords, Mukhopadhyay claims, borrowed from pīlu, an ancient Dravidian word for elephant. The variation in spelling with the ‘r’ replacing the ‘l’ in Akkadian and Old Persian is explained by the fact that the phoneme ‘l’ was often spelled and pronounced as ‘r’s in ancient Iranian languages. The people of ancient Persia functioned as intermediaries between Mesopotamian and IVC traders, she explains, and they passed along that variation in spelling.
She further points out records that show the Salvadora persica tree, commonly known as the toothbrush tree (miswak or tooth-cleaning stick in Arabic countries) was called pīlu after the decline of the IVC, reinforcing the idea that pīlu is related to the proto-Dravidian word for tooth. “When speakers of several Indic languages call Salvadora persica as ‘pīlu’, we need have no doubt that just like the elephant-word ‘pīlu’, this phytonym too is related to the Proto-Dravidian tooth-word,” she writes.
Could the people of IVC thus have been an ancient Dravidian-speaking people? It is a suggestion that has been made in the past, sometimes based on some evidence (such as attempts at deciphering the script or archaeological finds, like the discovery of an axe with IVC inscriptions in Tamil Nadu), and sometimes on very little. It remains an explosive suggestion because of its implications politically. What language was spoken at IVC is really a way for many to get at who were the people at IVC. Were they the forebears of the people who composed the Vedas, which would imply that the so-called Aryans were an indigenous race of people, some of whom moved out of the subcontinent towards Central Asia and Europe? Or were the people of IVC a distinct group, whose descendants we now find among Dravidian-speaking populations in the south?
It is to an extent a narrow way of looking at this issue. Since genetic studies indicate that all Indians, except for the people who remained secluded on the Andaman and Nicobar Islands, are affected by the mixing of two distinct ancestral groups—the Ancestral North Indians who are composed of people from IVC who mixed with the incoming migrants from the Steppe (who probably brought the Indo-European languages with them), and the Ancestral South Indians who were composed of people from IVC who mixed with Southeast Asian hunter-gatherers. The current diversity of Indians, from the highest to the lowest castes, including non-Hindu tribal populations, is a result of a mixture of these two ancestral groups, although the percentages of the mixture vary depending upon geography, caste and language groups.
A decipherment of the Indus script or the discovery of what language IVC people spoke could go a long way in illuminating the darkness that still remains.
Most scholars agree that the people of the IVC could write, but there is very little agreement on anything else. Various theories have been proposed from the Indus script representing a proto-Dravidian language, an ancient Indo-European language, to an early Munda language. Some have speculated a link with the cuneiform script of Mesopotamia; others drawn parallels with the glyphs found in the Pacific Ocean on Easter Island; while some have even claimed they wrote the script of a language that has gone extinct. In 2004, historian Steve Farmer and Harvard Sanskrit and Indian studies scholar Michael Witzel caused a stir when they claimed that the people in the IVC were illiterate and the Indus script did not represent a written language.
A few scholars like Mukhopadhyay believe that the Indus script does not encode a written language but a logographic one whose purpose was administrative control. According to her, the inscribed objects represented tools like tax tokens, trade licences or metrological records that were used to control the complex trading economy. “The inscriptions logographically encoded a commercial sublanguage to convey information about what kind of taxes/tithes were paid to which entities, using which rates and modes, and which activities… were licensed through such tax payment,” she writes in one paper.
“There is evidence, such as the brief way of communication, the portability of the object. Across the 1 million square km expanse of the IVC, across objects, you will usually have the same [type of] inscription. It seems like the seals and tablets were issued by some sovereign authorities and (or) trade-guilds who had control over several Indus settlements. The situation resembles a stamp issued by the central government of India, where even when several things differ between two states, something that belongs to the central government will have the same stamp,” says Mukhopadhyay.
She grouped the Indus symbols into classes based on their function and then looked for patterns that determined which signs couldn’t appear together. In a syllabic or alphabetic script that phonogically encodes a language, we see “phonological co-occurrence restrictions” because the sound of certain alphabetic letters coming together cannot be enunciated or perceived. However “phonological co-occurrence restrictions” mostly work in the domain of adjacent syllables. In Indus script, the co-occurrence restrictions are visibly meaning-driven, she says, with certain signs of similar functionality never co-occurring in the same inscription as only one meaning from a meaning domain may apply to that context. She offers the example of how an inscription if it were mentioning a commodity that is measured by dry measure, will not contain a sign for liquid measure, and vice versa. In the rare occasions when certain signs restricted from coming together did come together in the Indus script, they occurred adjacent to one another. “In a phonological script, this would never happen,” she says, leading her to believe that the script did not phonologically spell words.
This idea, however, does not enjoy a wide consensus. “The Indus script is quite versatile. It has been found on various types of objects, so it probably had different functions, not just one,” says Nisha Yadav, a researcher at the Tata Institute of Fundamental Research (TIFR) who has spent several years working on the script. “There have been suggestions that the stroke signs may represent numbers. But we don’t know whether they were actually numbers. Of course, they must have had some system of counting. And some component of the script might have been used for that.”
A few years ago, Yadav was part of a team of researchers from TIFR and the University of Washington which tried to find if the symbols constituted a language. They analysed the statistical pattern of the script using mathematics and computer science, calculating the degree of randomness in successive symbols of a sequence, and compared it to non-linguistic systems such as human DNA, protein sequences and also four linguistic scripts—English, Old Tamil, Rig Vedic Sanskrit and Sumerian, apart from also comparing it to Fortran, a computer programming language. The results, published in the journal Science in 2009, showed that the Indus inscriptions were indeed linguistic in nature, displaying the same level of randomness and patterns as the natural languages used for comparison, and differing from Fortran and other non-linguistic systems.
YADAV HAS SINCE moved on from focusing primarily on the grammar of the script to looking at its contextual aspects. She has been trying to see if there could be differences in its structure depending upon the location of the sites where it was found or the object types it was inscribed upon. “Broadly, the grammar remains the same,” she says, although she found some differences such as certain symbols being used more frequently in inscriptions found at certain sites compared to others.
“Perhaps, it shows differences in content or the style of writing in that area; maybe, it was connected to an event like a token being distributed at a function or the symbol was linked to a certain occupation, or cultural or economic aspect of that site,” she says.
The idea that the people of the IVC spoke an ancient Dravidian language and later moved to south India does not have a wealth of archaeological evidence to back it up.
But since the people in the IVC were known to be in contact with people from as far as Mesopotamia, it is only likely that they would have exchanges with those in south India too. A script is also distinct from a language. Seals with Indus symbols have been found in modern-day Iraq and Bahrain but in sequences that are different from IVC, suggesting that the script may have been used as a part of a different language or information system in those regions. Thus, if an Indus script discovered in south India is one day found to bear resemblance to a Dravidian language or culture, it would not necessarily mean that speakers of Dravidian languages are descendants of the IVC residents.
Mukhopadhyay’s discovery that an ancient Dravidian language was probably being spoken in IVC makes for a fascinating discovery, partly because it does not rely on the Indus script. Her argument does not convince everyone, since it rests mostly on a single discovered word and its tooth-based etymology. (Although linguists classify ‘tooth’ as part of a basic ‘ultraconserved’ vocabulary list, words which remain stable and are rarely borrowed from other languages.) A lot will thus depend on whether more such words can be discovered, words that have gone silent in the IVC region but whose echoes can be found elsewhere.
Mukhopadhyay claims she has found several such words, some of which can even be tied to the symbols in the script. She is working on these words now and will reveal them in the coming months. About her likely discovery of the IVC word for elephant and ivory, she says, “Even if we take out the Dravidian part [that it belonged to a proto-Dravidian language], the discovery is pretty exciting. We have found one Harappan word at last.”