विमर्श Vimarsh · Itihāsa

The Ones Who Wrote the Method

From Kautilya’s seven parts of a state to a milk cooperative in Anand, India’s path-breakers share one habit. They did not leave an idea. They left a procedure. Twenty-three centuries of it, with the myths taken out.

Nikhil Sharma · इतिहास Itihāsa · 9 October 2026 · 21 min read

Ask what India’s great minds have in common and the usual answer is genius. Read what they actually left behind and a plainer answer appears. They wrote the method down. A grammar with four thousand rules. A state in seven parts. A rule for zero. A procedure for a nose. A pattern for a cooperative. The genius was in the writing, and the writing is what survived.

This essay is a tour of those methods, from the one that came first to the ones running on your phone this morning. It has a rule of its own. Every claim in it rests on a text or a document that can be opened, and every claim that cannot be opened is left out and named at the end so you know what was refused. India has enough real achievement that it does not need the invented kind, and the invented kind is what keeps the real kind from being believed.

The state in seven parts

Begin with the one the request began with. The Arthaśāstra is fifteen books, a hundred and fifty chapters, about five thousand three hundred sentences, and it was lost for most of the last thousand years. A manuscript reached R. Shamasastry at the Mysore Oriental Library in 1905. He printed the Sanskrit in 1909 and the English in 1915, and the modern reputation of Kautilya dates from that decade, not from antiquity.

Two honesties before the admiration. First, the text and the man are not the same thing. Tradition puts Cāṇakya at the Mauryan court around 320 BCE. The surviving text is dated by most scholars to the first centuries of the Common Era, in layers, and it mentions Chinese silk and Roman coin that a Mauryan minister would not have seen. Patrick Olivelle’s edition lays this out. Kangle alone defended a single Mauryan author. Second, the name Cāṇakya does not appear in the book at all. The book calls its author Kautilya and a closing verse names Viṣṇugupta. The three names were stitched into one man by Gupta-era writers. So read it as the tradition reads it, as Kautilya’s, and read the dates as the scholars read them.

Now the method. Book six opens by naming the parts of a state: the king, the minister, the country, the fort, the treasury, the army and the ally. Seven prakṛtis. It is the first surviving text anywhere that treats a state as a thing with components that can be inspected one at a time, and it then does the inspecting. The next chapter builds the maṇḍala, a model of neighbours: the would-be conqueror, the enemy on his border, the middle king, the neutral, and from them twelve kings and seventy-two elements of states. Book seven gives the six policies a king can run between them, peace, war, waiting, marching, alliance and the double game. This is not wisdom. It is a decision table.

Book two is the largest and it is about money. Mines belong to the state and may be leased to private operators. Agricultural tax on private land is conventionally a sixth. New irrigation works are exempt for five years. Traders are taxed in kind, in labour or in cash. The intelligence service in books one and five is a standing organisation of disguised agents, ascetics, merchants, physicians, entertainers and women, paid on results and aimed, the text says, at traitors and the unrighteous and not at others. Twenty-three centuries later the structure of a modern state, its revenue code, its intelligence agency, its foreign-policy doctrine, sits in those chapters in recognisable form.

In the happiness of his subjects lies his happiness; in their welfare his welfare; whatever pleases himself he shall not consider as good, but whatever pleases his subjects he shall consider as good.

Book one, chapter nineteen. The line this house has quoted before and will keep quoting, because it makes the ruler’s wellbeing an output of administration rather than a mood. What is refused here: the Indian Machiavelli tag, which mistakes a bureaucratic manual for a cynic’s mirror, and the Cāṇakya Nīti aphorisms that circulate online, which are a separate and much later collection and are not in this book.

The grammar and the metre

Before Kautilya there was Pāṇini, and Pāṇini is the purer case. The Aṣṭādhyāyī is eight chapters of four quarters each, three thousand nine hundred and fifty-nine rules in the standard count, written somewhere between the sixth and fourth centuries before the Common Era by a man from Śalātura in Gandhāra, near the modern Lahor in Pakistan. It describes Sanskrit completely. It does so with ordered rules that operate on abstract underlying forms, with a metalanguage of markers and abbreviations, and with metarules that say what to do when two rules collide. The fourteen Śivasūtras at its head compress the sound inventory so that a two-letter code can stand for every vowel. It is the oldest surviving generative grammar of a natural language and Cardona called it one of the greatest monuments of human intelligence.

In March 1967 P. Z. Ingerman wrote to the Communications of the ACM proposing that the Backus–Naur Form used to specify programming languages be renamed Pāṇini–Backus Form, since the Sanskrit grammarian had used an equivalent notation two and a half thousand years earlier. The letter exists and can be read. What does not exist is the viral descendant of it, that Sanskrit is the best language for computers or that NASA uses it. Nobody who has read the letter believes that.

Piṅgala is the case where the honest version is smaller than the popular one and still remarkable. His Chandaḥśāstra, around the second century before the Common Era, is a manual of poetic metre. Its last chapter gives six procedures for listing and counting every pattern of light and heavy syllables, and in sūtras 8.28 to 8.31 it gives a recursive rule for computing powers of two by halving and squaring that uses the word śūnya as a marker. Kim Plofker calls this the first known explicit reference to zero. The listing of syllable patterns is isomorphic to listing binary strings. That is the achievement. The triangle of binomial numbers that is usually credited to him appears in Halāyudha’s commentary a thousand years later, and the Fibonacci-type sequence is Virahāṅka’s and Hemacandra’s. Piṅgala did not invent binary. He wrote a procedure that happens to be one.

The number

Āryabhaṭa dates himself. The Āryabhaṭīya says three thousand six hundred years of the Kali age had passed when its author was twenty-three, which is 499 CE and a birth in 476. One hundred and twenty-one verses. In the tenth verse of the chapter on calculation he gives the circumference of a circle of diameter twenty thousand as sixty-two thousand eight hundred and thirty-two, which is π to four places, and he calls the figure āsanna, approaching. In the ninth and tenth verses of the chapter on the sphere he says the stars appear to move west the way the bank appears to move backward to a man in a moving boat, which is the daily rotation of the earth stated as a fact two hundred years before anyone in Europe would say it in writing. He explained eclipses by shadows and said the moon shines by reflected light. He did not invent zero, he did not propose a sun-centred system, and the satellite that carried his name in 1975 was named for him rather than by him.

Brahmagupta dates himself too. The Brāhmasphuṭasiddhānta was written in 628 at Bhillamāla in what is now Rajasthan when he was thirty. In chapter eighteen he writes the rules. A debt minus zero is a debt. A fortune minus zero is a fortune. Zero minus zero is zero. The product of a debt and a fortune is a debt and of two debts a fortune. Zero times anything is zero. It is the first surviving text anywhere to treat zero as a number with arithmetic rather than as a gap in a column, and it also says zero divided by zero is zero, which is wrong, and leaves division by zero undefined, which is right. Chapter twelve gives the area of a cyclic quadrilateral in a formula that contains Heron’s as the special case. H. T. Colebrooke put it into English in 1817.

Bhāskara, born 1114, wrote the Siddhānta Śiromaṇi in 1150 at thirty-six. In its algebra he gave the cyclic method, the first general procedure for the equation x² − Ny² = 1, and worked the case N = 61, whose smallest answer is a ten-digit number. Fermat set the same case as a challenge to the mathematicians of Europe in 1657. Five hundred years is the gap. The Līlāvatī says a quantity divided by zero is infinite, and then says that infinity times zero gives the quantity back, which it does not. What is refused: that he invented calculus. He has a relation between the change in a sine and the cosine, used for one astronomical purpose, and no derivative, no limit and no general method.

Which leaves Mādhava of Saṅgamagrāma, near Cochin, around 1400, and the school that followed him for two centuries. None of his mathematical writing survives. His results survive in Nīlakaṇṭha and Jyeṣṭhadeva, and they are the infinite series for the arctangent, hence π/4 = 1 − 1/3 + 1/5 − 1/7 and onward, with correction terms that speed it up, and power series for the sine and cosine that Newton and Gregory would write two and a half centuries later. Jyeṣṭhadeva’s Yuktibhāṣā of about 1530, in Malayalam, supplies the derivations. Charles Whish read a paper on it to the Royal Asiatic Society in 1832 and the world mostly declined to notice for a hundred and fifty years. No document shows any of it reaching Europe. The Kerala school did not invent calculus either. It invented the series, which is the half of calculus nobody can see coming.

India has enough real achievement that it does not need the invented kind, and the invented kind is what keeps the real kind from being believed.

One paragraph on zero itself, because it is the claim most often inflated. Decimal place value with a placeholder was in regular Indian use by the seventh century: Brahmagupta’s rules in 628, Bhāskara I’s commentary the year after, and Khmer and Srivijayan inscriptions of 683 that carry a dot in the hundreds column. The oldest zero securely dated in stone in India is on the Chaturbhuj temple at Gwalior, 876 CE, in the numbers 270 and 50. In 2017 the Bodleian announced radiocarbon dates putting part of the Bakhshālī manuscript in the third century and the headlines called it the oldest zero in the world. Five specialists objected in print within months. In 2024 Oxford revised the dating to the ninth to eleventh century and the headline is withdrawn. The true story, zero as a number by 628, is good enough.

The body

The Suśruta Saṃhitā is a layered text whose final form is centuries later than the 600 BCE that gets printed on posters. The earliest manuscript is from 878 CE and the name is first linked to the tradition in the Bower manuscript of the fourth or fifth century. What is in it is what matters. It makes surgery the primary branch of medicine. It lists one hundred and one blunt instruments and twenty sharp ones and eight basic operations. It tells the student to practise on gourds and melons and dead animals. And in the sixteenth chapter of its first book it describes cutting a leaf-shaped flap from the cheek, turning it on a stalk of living tissue to cover the stump of a cut nose, and holding the nostrils open with two stems of the castor plant. Plastic surgeons still call the pedicled flap from the forehead the Indian flap. That is the first written description of reconstructive surgery and it needs no decoration. The Caraka Saṃhitā, in the same centuries, gives internal medicine its system and in the eighth chapter of its Vimānasthāna gives the physician a code: do not enter a house uninvited, keep the patient’s confidence, do not extort a fee. It is among the oldest medical codes anywhere and it is a set of instructions, not an oath.

Jump fifteen hundred years to a method with the same shape. In 1976 Dr G. Venkataswamy, fifty-eight and retired from government service, opened an eleven-bed eye hospital in Madurai. He had visited Hamburger University in Chicago and asked why a cataract could not be done the way a hamburger was: standardised, high-volume, the surgeon doing only what the surgeon must. Aravind now does more than seven lakh procedures a year and has done more than ten million, half of them free or nearly free, paid for by the half that pay. In 1992 it set up Aurolab to make intraocular lenses and the lens that cost about a hundred dollars to import fell to a few. Devi Shetty did the same arithmetic for the heart at Narayana from 2001: thirty to thirty-five major heart operations a day in Bengaluru, a break-even cost around twelve hundred dollars, a listed company that made a profit. Suśruta wrote down how to cut. Venkataswamy and Shetty wrote down how to cut ten thousand times without the quality or the poor falling off the table. Same instinct, different century.

The test

Some of the methods are for the mind. In the Kālāma Sutta the Buddha tells the people of Kesaputta not to go by reports, by legends, by tradition, by scripture, by logical conjecture, by inference, by analogy, by agreement with considered views, by probability, or by the thought that this one is our teacher. Ten grounds ruled insufficient. In their place a test of consequences: adopt what you know for yourself to be skilful and blameless and praised by the wise. The viral paraphrase, believe nothing unless it agrees with your own reason, is a twentieth-century rewrite and the sutta explicitly distrusts bare personal reasoning as well. Even so it is an epistemic procedure written down in the fifth century before the Common Era and it still works.

Nāgārjuna, around 150 to 250 CE, turned the procedure on philosophy itself. The Mūlamadhyamakakārikā takes every claim about a thing’s intrinsic nature, asks whether it arises from itself, from another, from both or from neither, and shows that each answer fails, without putting a claim of his own in the empty chair. In the Vigrahavyāvartanī he says it outright: if I had a thesis the fault you name would apply to me, but I have no thesis. Emptiness is dependent origination and is itself empty. It is the only philosophical method ever designed to be impossible to turn into a creed, and it was invented in India.

Kabir, a weaver in Banaras in the fifteenth century, built a method out of a medium. He put a critique of caste, ritual and both the Hindu and the Muslim clergy into couplets that a singer could carry and a weaver could remember. No Sanskrit, no Persian, no scribe. The Ādi Granth took more of his verses than of any other non-Sikh. The posters say he synthesised Hinduism and Islam. Linda Hess calls that picture false; he attacked both and the need for institutional religion itself. Pothī paḍhi paḍhi jag muā, paṇḍit bhayā na koi. Reading book after book the whole world died and nobody became wise. A line about method, from a man whose method was not to need the book.

The atom and the sky

The modern path-breakers wrote in journals rather than in sūtras and the shape holds. In 1894 to 1896 at Presidency College in Calcutta Jagadish Chandra Bose generated and detected waves between five millimetres and two and a half centimetres long and in 1895 sent one through two walls and twenty-three metres of room to ring a bell. The IEEE marked it in 2012 as the first millimetre-wave communication experiments in the world. His galena detector of 1904, the one patent he ever took and let lapse, is an early semiconductor junction; Nevill Mott said he was sixty years ahead of his time. What is refused: that he invented radio. Marconi built a telegraph. Bose built the physics a later century would use for radar and the chip.

In January 1913 a clerk in Madras with no degree wrote to G. H. Hardy that he had had no university education but had undergone the ordinary school course, and enclosed theorems. Hardy brought him to Cambridge. The Royal Society elected Srinivasa Ramanujan in 1918. He died in 1920 at thirty-two. In the spring of 1976 George Andrews found a hundred and thirty-eight pages in his hand in a box in the Wren Library at Trinity, the lost notebook, and the mock theta functions in it have fed a field for fifty years. The method here was the man’s refusal to accept that a proof was the only form a truth could take.

In June 1924 a reader at Dacca University sent Albert Einstein four pages deriving Planck’s law by counting indistinguishable quanta in cells of phase space. Einstein translated it into German himself, sent it to Zeitschrift für Physik and extended it to gases. Half the particles in the universe are now called bosons after Satyendra Nath Bose and he never won a Nobel. In 1920 Meghnad Saha published the ionisation equation that lets you read a star’s temperature from its spectrum; Rosseland said the impetus it gave astrophysics could scarcely be overestimated. On 28 February 1928 at the Indian Association for the Cultivation of Science in Calcutta C. V. Raman found that light scattered by a liquid changes colour by an amount that fingerprints the molecule. The Nobel came in 1930, two years later, the first in a science to an Asian. The story that the apparatus cost two hundred rupees is a newspaper invention; the historian Rajinder Singh itemised it at seven thousand six hundred and thirty, which was still a modest laboratory and makes the point better.

Then the institution builders of the sky. On 12 March 1944 Homi Bhabha wrote to the Dorabji Tata Trust asking for a school of physics comparable with the best anywhere, on the argument that when nuclear power arrived in a couple of decades India should not have to look abroad for its experts. The Tata Institute of Fundamental Research began the next year and the Atomic Energy Commission in 1948, before the country had a reactor; Apsara went critical in 1956, the first in Asia. Vikram Sarabhai founded the Physical Research Laboratory in 1947 at twenty-eight and launched the first rocket from Thumba in November 1963. His sentence on purpose is the clearest statement of method in the whole modern record: we do not have the fantasy of competing with the economically advanced nations in the exploration of the moon or the planets, but we must be second to none in the application of advanced technologies to the real problems of man and society. ISRO prints it on his page. And Satish Dhawan, when the first satellite launcher failed in August 1979, took the press conference and the blame himself, and when it succeeded in July 1980 put the project director in front of the cameras. The project director was Kalam and he told the story for the rest of his life as the whole of leadership in two acts.

The institution

Jamsetji Tata did not found Tata Steel. He read a report on iron in the Chanda district in 1882, chased coking coal for twenty years, died in 1904 and his son Dorab floated the company in 1907 with about eight thousand Indian subscribers putting up two hundred and thirty-two lakh rupees in three weeks, the first large Indian industrial venture financed by Indian money. Nor did he found the Indian Institute of Science. He endowed it in 1898 with fourteen Bombay buildings and four estates worth about thirty lakh, roughly half his wealth, for a university that opened in 1911, seven years after he died. The method is a man writing a cheque to a future he would not see and attaching a plan to it.

Verghese Kurien did not found Amul. Tribhuvandas Patel did, in 1946, under Sardar Patel’s guidance. Kurien arrived in Anand in 1949 and wrote the pattern: farmers own the cooperative, the cooperative owns the processing and the brand, professionals work for the farmers and not the other way round. Operation Flood ran it across the country from 1970 to 1996, Phase One financed by selling donated European milk powder and butter oil, and milk production went from twenty-one million tonnes in 1968–69 to fifty-one million by 1989–90 and India passed the United States as the largest producer in 1997–98. Per-head availability had been falling since 1950. It has risen every decade since.

M. S. Swaminathan did not breed the dwarf wheat. Norman Borlaug did, in Mexico, and William Gaud of USAID named the Green Revolution in a speech in March 1968. What Swaminathan wrote was the Indian procedure: he took Borlaug through the wheat belt in 1963, brought in eighteen thousand tonnes of seed, adapted it, and organised the extension that took wheat from ten million tonnes in the drought year 1965–66 to twenty million in 1969–70. Foodgrain rose by half in five years. He spent the rest of his life warning that the method had a cost and coined evergreen revolution for the version that would not.

E. Sreedharan rebuilt the Pamban bridge in forty-six days after the cyclone of December 1964 had been given months, built the Konkan Railway, seven hundred and sixty kilometres and two thousand bridges, through a corporation funded partly by public bonds between 1990 and 1998, and delivered the first phase of the Delhi Metro in 2005 two years and nine months ahead of schedule and inside the budget. Nandan Nilekani’s UIDAI issued its first number in September 2010 to a woman in Nandurbar and by September 2025 had issued one hundred and forty-two crore of them, and the Supreme Court in 2018 upheld the core and struck down the private use and the mandatory linking in the same judgment, which is what a court is for. UPI went live for customers with twenty-one banks on 25 August 2016, seven months before demonetisation and not because of it, and in September 2026 carried twenty-four billion transactions, eight hundred million a day, the largest real-time retail payment system on earth, built as a public utility rather than a company.

And the two who wrote the method for the state itself. On 11 September 1906 at the Empire Theatre in Johannesburg Gandhi had a hall of Transvaal Indians take a vow before God never to submit to fingerprint registration, and later wrote that he only knew some new principle had come into being. The method was public, pre-announced acceptance of the penalty, so that the sufferer rather than the state chose the ground; he acknowledged Thoreau and Tolstoy and the novelty was the vow and the scale. B. R. Ambedkar, introducing the draft constitution on 4 November 1948, quoted George Grote on constitutional morality and added that it is not a natural sentiment, that it has to be cultivated, that our people have yet to learn it and that democracy in India is only a top-dressing on an Indian soil which is essentially undemocratic. He credited B. N. Rau and S. N. Mukherjee for the drafting. The Supreme Court revived his phrase from 2014. A method for a republic, with its own warning attached.

The pattern

Look at the list once more and the thing they share is not a field or a century or a caste. It is the habit of leaving a procedure that another person could run. A grammar that generates every correct sentence. A chapter that lists the parts of a state. A rule that says what a debt times a fortune is. A description of a flap with a stalk. A letter to a trust with a plan attached. A cooperative pattern. A seed procedure. A vow with a date. A public utility with a specification. The ideas were not the achievement. The ideas were everywhere. The achievement was the writing down, in a form that outlived the writer and could be checked.

That is also why the exaggerations matter and why this essay names them. Kaṇāda’s paramāṇu is a philosophical posit about divisibility and not Dalton’s atom. Piṅgala did not invent binary. Āryabhaṭa did not invent zero or move the sun to the centre. Bhāskara and Mādhava did not invent calculus. Suśruta did not do plastic surgery as we know it in 600 BCE. Raman’s apparatus did not cost two hundred rupees. Kurien did not found Amul and Swaminathan did not breed the wheat. Every one of those is a smaller claim than the real one, because the real one comes with a date and a text and a procedure, and the inflated one comes with nothing but a poster. A country that writes its methods down should be the last to need its history embroidered.

Kautilya first, then, because he wrote the state. But the thing to take from Pataliputra is not the sixteen-block day or the spies. It is the sentence on the first page of book six that lists the seven parts and then goes and inspects each one. Name the parts. Write what each must do. Check. Everything in this essay descends from that sentence and most of what fails in India today fails for want of it.

The wager

Dated 9 October 2026. By 9 October 2031 the two largest methods in this essay that are still being written, the identity layer and the payments layer, will be running in at least ten countries outside India under agreements that name the Indian specification, and the next Indian name added to this list will be a person who wrote one of them. If by then the only thing India has exported from the stack is a slide deck, I will say so here, on the shelf where the argument was made.

इति ह आस. So indeed it was.

Gurugram, 9 October 2026. The first long essay on the Itihāsa shelf and the sequel to the column on Kautilya’s timetable. The request was for all the path-breakers; the rule was that every one of them had to come with a document; the rule won every argument with the request.

Sources. Kautilya: Arthaśāstra 1.19 and Book 6 in Shamasastry’s translation; Shamasastry’s full text; Olivelle’s dating and the Trautmann analysis via the SOAS repository. Pāṇini: MacTutor; Ingerman’s 1967 letter and Wujastyk’s caveat on the Indology list. Piṅgala: University of Hyderabad analysis of the sūtras. Āryabhaṭa: Shukla–Sarma translation of Gola 9–10; Biographical Encyclopedia of Astronomers. Brahmagupta: MacTutor. Bhāskara II: Britannica; Indian Journal of History of Science 2023. Mādhava: MacTutor; India and the calculus of trigonometric functions. Zero: Plofker, Keller, Hayashi, Montelle and Wujastyk on the Bakhshālī dating, HSSA 2017; Bhāvanā on the controversy. Suśruta: JKIMSU 2016 on the instruments and the eight operations; Hektoen International on 1.16. Buddha: Kālāma Sutta, Thanissaro and Lost in Quotation. Nāgārjuna: Stanford Encyclopedia of Philosophy; Westerhoff on Vigrahavyāvartanī 29. Kabir: Encyclopaedia of Sikhism. J. C. Bose: IEEE Milestone. Ramanujan: MacTutor; the lost notebook, oral history. S. N. Bose: Physics World. Saha: arXiv 1810.10898. Raman: Nobel Prize; Rajinder Singh on the cost of the apparatus. Bhabha: Department of Atomic Energy history. Sarabhai: ISRO. Dhawan and Kalam: President Kalam’s address. Tata: Tata timeline; IISc heritage; Tata Steel 2006–07 report on the 1907 issue. Kurien: PIB backgrounder, Nov 2021; NDDB genesis. Swaminathan: Review of Agrarian Studies; Gaud’s 1968 speech. Aravind: Aravind, our story; McKinsey on the model. Narayana: KCUR 2015. Sreedharan: Konkan Railway, the beginning; ThePrint on Delhi Metro Phase I. Aadhaar: UIDAI; Oxford Human Rights Hub on Puttaswamy. UPI: NPCI launch release 2016settogoliveAugust252018.pdf); September 2026 volumes. Gandhi: Satyagraha in South Africa, chapter on the 1906 meeting. Ambedkar: Bar and Bench on the 4 November 1948 speech.