There is a delicious irony in the story of Sanskrit. In the land of its origin, it is summoned either for ceremonial obeisance or political combat around caste and religion. But thousands of kilometres away, in European universities, it has for two centuries been subjected to the considerably less glamorous business of being learnt, parsed, edited, catalogued, and researched. Every now and then, a newspaper report tells us that some foreign university has discovered its wonders and the news is greeted with pride.
The Germans, in particular, have been studying Sanskrit since the 19th century. That tradition continues with 14 German universities today teaching Sanskrit and Indology, compared with four in the UK. At Heidelberg, demand for Sanskrit courses has been so strong that its South Asia Institute expanded summer programmes beyond Germany. Students from across the world study Sanskrit not merely as a religious or cultural artefact, but as a gateway to Indian philosophy, history, linguistics, and intellectual traditions.
European Indology emerged in the 19th century, with German philologists, British orientalists, and French scholars producing grammars, dictionaries, translations, and critical editions of Indian texts. But much of this scholarship developed within colonial Europe’s assumptions about race, religion, language, and civilisation. India became not merely a country that had been conquered, but also a civilisation interpreted through categories developed elsewhere. Whatever the limitations and prejudices of colonial-era scholarship, it created enduring institutional infrastructure for studying our civilisation. The consequence is a curious geography of knowledge. India possesses the civilisation and much of its surviving archive; but the intellectual centres with the greatest global influence in interpreting that archive are often elsewhere.
Rajiv Malhotra’s The Battle for Sanskrit raised this provocative question of who possesses the adhikara to interpret Sanskrit and traditions that grew around it. While his “insider” versus “outsider” distinction is debatable, the underlying question is worth retaining. If a civilisation becomes primarily an object of study for institutions outside it, over time the categories of those institutions begin to shape how the civilisation sees itself.
That does not mean a foreign scholar cannot understand India. After all, scholarship has no passport. But a civilisation should have the confidence to study itself and disagree with interpretations made elsewhere. This becomes especially interesting when we arrive at artificial intelligence.
Panini, the great Sanskrit grammarian, was doing something extraordinary. He was trying to understand how language itself worked, and to describe it with extraordinary precision. In the Astadhyayi, he described Sanskrit through an intricate system of rules and conditions. Modern computer scientists naturally find such a system interesting.
But Sanskrit and computers are not a new romance. For more than three decades, scholars in India and elsewhere have been working on Sanskrit computational linguistics. They have built programmes that can analyse Sanskrit words, study sentence structures, understand how words join together and assist in translation. Scholars such as Amba Kulkarni and Pavankumar Satuluri, among numerous others, have spent years asking how ideas from the Indian grammatical tradition might be represented computationally. The conversation did not begin yesterday.
However, grammar is only part of understanding a language. A modern AI system can take a Sanskrit word, find its nearest English equivalent and produce a perfectly fluent interpretation, while quietly importing the assumptions that accompanied that English translation into the text. In other words, the machine may get the language right but the civilisation wrong. This is where decolonisation becomes more than a historical argument.
For generations, we have debated how Europeans interpreted India. The next question is whether we can build machines that understand Indian texts without simply inheriting those old interpretive habits. This does not mean throwing away Western scholarship. We need the best of it. But Indian grammatical traditions, commentaries, and ways of understanding our languages must also become part of the knowledge from which our machines learn. The two worlds—the Sanskrit scholar who has spent decades reading manuscripts and commentaries, and the computer scientist building language models—need to converge.
India’s languages have been speaking to one another for centuries. Sanskrit has contributed enormously to many Indian languages. Our AI should eventually be able to understand some of this traffic. Ancient texts could become searchable across languages rather than be left sitting in archives while we build the future elsewhere.
A Kannada speaker should not have to become English-literate before accessing knowledge that exists in Sanskrit, Bengali or Marathi. Nor should an Indian language have to surrender its particular vocabulary and cultural associations simply because the machine was trained primarily on English. This is the problem that we at NAAV AI have begun trying to address: how can technology help knowledge travel between India’s languages without making English the obligatory middleman? Sanskrit could form an important part of that connective tissue.
That is a more interesting future for Sanskrit than simply putting it on a pedestal. Europeans built institutions around it. Indian scholars continued the tradition, and over the past few decades some have begun teaching computers to work with it. Perhaps the next chapter should be ours. Not because Sanskrit needs to be rescued from the West. But because intellectual independence means having the confidence to study what we inherited, question how it has been interpreted, correct what needs correcting and carry it forward into a world that did not exist when Panini wrote his grammar.
The future of Sanskrit may not lie in getting more people to recite it. It may lie in getting our machines to understand, on our terms, what it has been saying for millennia.
vikram.sampath@fihcr.com
