storytelling

Whoever writes the knowledge sets the language

If India wants to lead the world, the most direct route runs through language. Pick one Indian language, and my choice is Sanskrit, and make it the language in which new knowledge is first written. Translate the world's most important work into it. Then make sure the next breakthroughs appear there before they appear anywhere else.

By Tejas

If India wants to lead the world, the most direct route runs through language. Pick one Indian language, and my choice is Sanskrit, and make it the language in which new knowledge is first written. Translate the world’s most important work into it. Then make sure the next breakthroughs appear there before they appear anywhere else.

Anyone who wants that knowledge would have to learn the language. Nobody learns a language without absorbing its vocabulary, its way of dividing up the world and some of the culture that built it. That is what influence actually looks like: other people thinking in your terms because your terms are where the useful ideas live.

This is not a fantasy. It is roughly how English came to hold the position it holds today, including in India. The difference is that India was on the receiving end.

How English won India

English won in India in one decision, and the man who made it wrote down his reasoning. In 1835 the East India Company’s Committee of Public Instruction was split down the middle over how to spend its education grant. Half wanted to keep funding Sanskrit, Arabic and Persian learning. The other half wanted English. On 2 February 1835 Thomas Babington Macaulay settled it with his Minute on Indian Education, and Governor-General William Bentinck approved it that March.

Macaulay’s argument is worth reading closely, because it is the argument of this essay pointed the other way. He said Indians should be taught “those foreign languages in which the greatest mass of information had been laid up”, and that English was that language. Whoever learned it, he wrote, would get access to the accumulated work of the wisest nations. He did not argue that English was beautiful. He argued that it was where the knowledge was.

He was also open about the purpose. The Company could not educate the whole population, so it would form “a class who may be interpreters between us and the millions whom we govern”, people “Indian in blood and colour, but English in tastes, in opinions, in morals and in intellect.” That is the clerk class: Indians who could run the administration in the rulers’ language.

Was killing Sanskrit part of the plan? We don’t need to guess. Macaulay recommended that the government “at once stop the printing of Arabic and Sanscrit books”, abolish the Sanskrit College at Calcutta, and stop paying stipends to students of Sanskrit and Arabic anywhere. He claimed that “a single shelf of a good European library was worth the whole native literature of India and Arabia,” while admitting in the same document that he could read neither Sanskrit nor Arabic.

The uncomfortable part is the evidence he leaned on, and it is the most useful part for anyone who wants to reverse the outcome. Macaulay pointed out that Sanskrit and Arabic students had to be paid stipends to study, while students of English paid fees. The Committee had printed some 23,000 volumes of Sanskrit and Arabic books that almost nobody bought, while an English school book society was selling seven or eight thousand volumes a year at a profit. A decade earlier, Rammohun Roy had himself written to the Governor-General objecting to government money for a new Sanskrit college. Contempt and policy did real damage. But Sanskrit also lost because English was where the jobs and the new knowledge were, and Indians could see it.

Two locked doors

Most Indians today are shut out of two bodies of knowledge at once. The first is the world’s, which sits in English. In the 2011 census, about 129 million people, roughly 10.6 percent of the country, said they spoke English as a first, second or third language. The second is India’s own older learning, which sits in Sanskrit. The same census counted 24,821 people who gave Sanskrit as their mother tongue, and only about a quarter of one percent who claimed to speak it at all.

That leaves the large majority inside their regional languages, with access only to what someone has bothered to translate. The size of that gap is easy to see. Hindi has several hundred million speakers, and as of September 2026 the Hindi Wikipedia had 171,575 articles. The English one has about forty times as many. Many Hindi speakers who are online simply read the English edition instead.

A translated library is always behind. Worse, very little original work is produced inside these languages, because Indian researchers, engineers and founders mostly publish and build in English. So the regional-language world is an echo chamber with no source of its own. It hears late, partial versions of what was said elsewhere, and adds little that the outside world would need to come in for.

Macaulay predicted that his English-educated class would carry Western knowledge back into Indian languages and “enrich those dialects with terms of science.” Nearly two hundred years later, that mostly has not happened. The interpreters stayed interpreters.

It has been done before

Languages have become gateways to knowledge several times in history, and in each case you can see the mechanism at work. Four examples are worth knowing.

The first involves Sanskrit directly. In the early 770s an Indian scholar brought astronomical texts to the court of Caliph al-Mansur in Baghdad, and the caliph had Brahmagupta’s Brahmasphutasiddhanta translated into Arabic. A generation later, al-Khwarizmi built on these Indian methods and wrote about Indian numerals and arithmetic. When Europe picked up the system centuries later, it got it from Arabic books in Latin translation, and it still calls the digits “Arabic numerals.” Baghdad did exactly what this essay proposes: it translated the world’s best work into its own language, then built on it. Arabic became the language scholars had to learn. India supplied the idea and lost the name.

The second is German. In 1905, by historian Michael Gordin’s count, about 30 percent of the world’s scientific literature was in German, roughly level with English. A serious chemist or physicist had to read it. Then came the First World War, a postwar boycott of German scientists by their British, French and Belgian colleagues, laws against German in American states, and later the expulsions of 1933. Within a generation German was finished as a language of science. A gateway language holds only while the people producing knowledge in it keep producing it.

The third is Japan. During the Meiji era, translators coined Japanese words for Western concepts, and the country built a full science education in its own language. In 2008 Toshihide Maskawa delivered his Nobel lecture in physics in Japanese, having done prize-winning work without speaking English. A country can do frontier science in its own tongue.

The fourth is the closest parallel to Sanskrit. Hebrew was for centuries a language of scripture and prayer, not of daily life or engineering. In October 1913 the board of a new technical institute in Haifa chose German as its language of instruction. Teachers and students walked out in protest, and in February 1914 the board reversed itself and chose Hebrew. The Technikum became the Technion. A liturgical language became a language of engineering because people decided it would be, and then built institutions to make it so.

What it would take

The plan has two halves, and Baghdad shows they have to come in order: first bring the world’s knowledge in, then produce knowledge nobody else has.

Start with the choice of language. Sanskrit has three things going for it. It belongs to no single state, so choosing it does not hand one region a win over the others the way Hindi does. Most Indian languages already borrow their formal and technical vocabulary from it, so a Kannada or Bengali speaker meets familiar roots. And its grammar, set down by Panini as a compact system of a few thousand rules, was built to generate new words in a predictable way, which is exactly what a language needs when it has to name ten thousand new technical concepts.

Then comes the work itself:

  1. Translate the core of world knowledge into Sanskrit: school and university textbooks, the standard references in each field, and the most cited research. Machine translation makes this far cheaper than it was for Baghdad’s translators or Meiji Japan’s.
  2. Teach it seriously, so that within one generation a large number of Indians can read a physics textbook or a legal judgment in it.
  3. Write India’s history in it, from Indian sources and Indian research, rather than reading ourselves through accounts written for someone else’s purposes.
  4. Publish new research in Sanskrit first. Build software in it, including programming languages, documentation and developer tools, so that working with the newest Indian technology means working in Sanskrit.
  5. Keep inventing. None of the rest matters if the knowledge behind the gate is not worth the climb.

The last point is the whole game. Macaulay won his argument with a market fact: people paid to learn English and had to be paid to learn Sanskrit. To reverse that, Sanskrit has to become the language where the jobs, the tools and the best ideas are. People will learn it when not knowing it costs them something.

The hard parts

This idea has real weaknesses, and anyone who wants it to work should face them now rather than after the money is spent.

The first is that it can build the very echo chamber I described above. If Indians learn Sanskrit instead of English, and the frontier of knowledge stays outside India for another few decades, we will have recreated the regional-language trap on a national scale. The answer is to add, not to replace. Japan teaches and thinks in Japanese and still sends its best work out in English. Even Maskawa, after his Nobel lecture, said he wished he could have spoken to his fellow laureates in English. Sanskrit should become a second door into the house, and eventually the main one. It cannot start as the only one.

The second is that language tends to follow power more than it creates it. Gordin’s account of how German lost and English won is a story of wars, boycotts and American scientific weight after 1945, not of anyone planning a language. Baghdad’s Arabic and Macaulay’s English both rode on empires. India would be attempting it the other way round, using the language to build the lead. That can work, but only if the research, products and institutions behind it are good enough that people come on their own.

The third is that locked-up knowledge is often routed around rather than learned. When valuable work sits in a language others don’t read, they tend to translate it, reinvent it or ignore it. Baghdad never hid its books. Its scholars translated freely, and Europeans later went to Toledo to translate Arabic into Latin. Arabic’s prestige came from being the best library in the world, not from keeping people out. A Sanskrit-first policy should aim for that kind of pull: be the source everyone has to cite, so that reading the original in Sanskrit becomes the mark of being serious.

The fourth is politics at home. India has 22 scheduled languages and a long history of resisting any one of them being pushed on the rest. For centuries, Sanskrit learning was also restricted by caste. A revival that looks like a return to that, or like a religious project, will be fought, and should be. It has to be offered to every Indian as a working language of science and engineering, open to anyone, the way Hebrew was offered to immigrants from dozens of countries.

The bet

Macaulay bet in 1835 that Indians would come to English because that was where the knowledge was. He was right, and we are still living inside his bet. The same logic runs in reverse for any country willing to do the work: translate the world’s knowledge into your language, teach it, write your own history in it, and then produce things the world cannot get anywhere else. The language comes last. It follows the ideas.

So the real question is not whether India should switch to Sanskrit. It is whether India will produce enough that the rest of the world finds it worth learning.

Sources

Written by

Tejas

Founder, writer, philosopher

Tejas writes, runs the satsang, and holds the structure of the community. He spent years in software before turning the same engineering instinct on the inner life. His mission is to make quality education available across every part of being human, and to help people find their purpose in life.