Cracking an ancient code: A Wknd interview with Sanskrit scholar Rishi Rajpopat

He spent days lounging on the banks of the River Cam, gazing at the Bridge of Sighs; went on late-night cycle rides through the streets of Cambridge; swam a lot; spent time in libraries with spectacular views. It was a warm, sunlit month.

“After nine gruelling months of researching my PhD thesis, I felt I was getting nowhere and was ready to quit. The disappointment over my potential failure was making it impossible to enjoy what I truly loved doing — asking questions, and decoding puzzles. So I decided to shut my books and take a month-long break that turned out to be a game changer,” says the 27-year-old.

The discovery he made when he returned to his research was the kind of thing one might expect in a Sherlock Holmes novel; certainly not in real life.

Rajpopat had been researching the grammar algorithm for Sanskrit created by the scholar Pāṇini about 2,500 years ago. It is a set of 4,000 rules that lets the user churn out grammatically correct Sanskrit words, and subsequently sentences, once the root and the suffix have been chosen.

Except, no one could get the system to work. The big problem was what to do when two of Pāṇini’s rules became simultaneously applicable. Pick the one that comes later, Pāṇini had said. But if one picked the rule that appeared later in the sequence, the result was often an ungrammatical mess.

This was the scholar considered the father of linguistics. Clearly, if he said the program worked, it worked. Why wasn’t it working? The question had begun to obsess Rajpopat.

Century after century, new metarules had been written, in attempts to “fix” the “problems”. “We had all begun to accept this unnecessary workaround as the norm, but the one thought that kept eating into me was, why would a genius like Pāṇini leave it to us to ‘fix’? What were we missing?”

The answer, it turned out, was hidden in plain sight.

After his summer in the sun, Rajpopat went back to his books and began to study the patterns in his work. “Pick the one that comes later…” He realised that Pāṇini hadn’t meant pick the rule that comes later in the serial order; he’d meant pick the rule applicable to the right hand side of the word.

Used that way, the algorithm works almost without exception, making it possible to construct grammatically accurate Sanskrit words. “I was ecstatic when I solved it,” Rajpopat says.

While he made his discovery in 2018, Rajpopat finally published his thesis, titled In Pāṇini We Trust: Discovering the Algorithm for Rule Conflict Resolution in the Aṣṭādhyāyī, this month. It has made him a celebrity overnight. “I expected this to draw the attention of a couple of academic magazines, but didn’t expect such an overwhelming response. My phone hasn’t stopped ringing; people are calling up my parents and friends hoping to speak to me. I feel extremely grateful and blessed.”

***

Rajpopat grew up in Mumbai, the son of engineers-turned-science tutors Atul Rajpopat and Bhavna Rajpopat.

He’d always loved languages and picked them up easily. He first studied Sanskrit in high school. He began studying Pāṇini’s grammar out of interest, while pursuing a degree in Economics. His Sanskrit tutor was a retired professor who lived in the neighbourhood.

Rajpopat says he was smitten by Pāṇini’s genius. By the time he graduated, he had “this burning desire to make sense of Pāṇini’s brilliance in my own way”. He signed up for a Master’s degree in Sanskrit at Oxford University and then a PhD at St John’s College, Cambridge, in 2017. He had a penchant for grammar, “but my heart always lay in puzzles and hidden meanings, especially in the linguistic context,” he says.

The code that he has now cracked could open new doors for an ancient tongue. The algorithm that runs Pāṇini’s 4,000-rule grammar could potentially be taught to computers. Rajpopat, meanwhile, comes away with a new respect for the simple question: Why? “I believe it is the willingness and commitment to unlearn and relearn that helped me do this,” he says.

Rajpopat, who has studied 12 languages (including Urdu, Persian and Sindhi), is now studying Ancient Greek at the University of St Andrews, Scotland, while working as an academic editor on the Encyclopaedia of Theology being compiled there.

He’s currently in India and says he was totally unprepared for the frenzy his thesis had caused here too. “Congratulatory calls are turning into unexpected marriage proposals. That’s been the trickiest bit so far,” he says, laughing. “But I am extremely grateful for all the love I’ve been getting. I hope this encourages other students to ask more questions and not settle for the easy answers.”