What is Ayurveda?

In 1792, in the closing weeks of the Third Anglo-Mysore War, a bullock-cart driver named Cowasjee, who had been working for the British, was captured by Tipu Sultan’s men. They cut off his nose. They cut off one of his hands as well, and then they let him go.

A year later, in Pune, a local practitioner rebuilt his face. He cut a flap of skin from Cowasjee’s forehead, left it attached at one end so the blood kept flowing, swung it down over the ruin and sewed it into place. Two British surgeons at the residency, Thomas Cruso and James Findlay, watched and wrote it up. The account reached London in October 1794, in The Gentleman’s Magazine, with an engraving of the restored face. A London surgeon named Joseph Carpue read it, studied it for twenty years, performed the operation himself in 1814, and published in 1816. European plastic surgery was born under a name it still half carries: the Indian method.

Here is what almost every retelling leaves out. The man who rebuilt Cowasjee’s nose was almost certainly not reading Sushruta. He was doing what his family had done for generations. And the flap he used, from the forehead, is not the flap the Sushruta Samhita describes, which is taken from the cheek. The technique had changed. It had been improved. It had moved in hands, not in books.

That is the thing about Ayurveda that neither its worshippers nor its debunkers will tell you. It was never a fossil.

What the science says it is for

Ayurveda comes apart into two words. Ayus, life, and veda, knowledge. The knowledge of life, and pointedly not the knowledge of disease. The name is already a claim.

The Charaka Samhita, the oldest surviving textbook of the tradition, states its purpose in a single line at Sutrasthana 30.26: prayojanaṃ cāsya svasthasya svāsthya-rakṣaṇam āturasya vikāra-praśamanaṃ ca. “The purpose of this science is to preserve the health of the healthy and to cure the disease of the diseased.”

Read the order of the clauses. The healthy come first, grammatically first, in the founding sentence of the whole discipline.

And what is health? Sushruta answers at Sutrasthana 15.41: one whose doshas are balanced, whose agni, the digestive fire, is steady, whose tissues and wastes are doing their work, and then he keeps going, and whose self, senses and mind are serene. Such a person is svastha, from sva, self, and stha, standing. A healthy person is one who is standing in himself. By that definition a man with perfect bloodwork and a wretched mind is not healthy.

Charaka binds the whole enterprise to the four aims of a human life, calling health the supreme root of dharma, artha, kama and moksha. Ayurveda wants you well in the world.

The two doshas nobody counts

Everyone has heard of the three. Vata, born of air and space, the principle of movement. Pitta, born of fire and water, transformation and heat. Kapha, born of water and earth, cohesion and structure. Dosha means, literally, “that which can spoil”: they are at once the principles that run the body and the things that can go wrong with it. Same word, both jobs. One of vata’s five sub-types is prana-vayu, where Ayurveda and the yogic map of the body share a vocabulary.

But Charaka does not count three. He counts five.

At Sutrasthana 1.57: vāyuḥ pittaṃ kaphaś coktaḥ śārīro doṣa-saṃgrahaḥ | mānasaḥ punar uddiṣṭo rajaś ca tama eva ca. “Vayu, pitta and kapha are declared the collection of bodily doshas. The mental doshas are said to be rajas and tamas.”

Rajas and tamas. Agitation and inertia. Two of the three gunas, the strands from which Samkhya weaves the whole of nature, reclassified here as pathogens and listed beside wind, bile and phlegm as things that make a person ill.

This is the verse nobody quotes, and it changes the shape of the system. A medicine that names restlessness and dullness as diseases is doing something stranger than humoral balancing. Follow it and you arrive somewhere unexpected: in Sharirasthana 1, Charaka teaches yoga and liberation outright, and defines moksha as the elimination of rajas and tamas. The very things he calls mental doshas are the obstacles to freedom. Medicine and soteriology run on one machine. The physician and the yogi are fighting the same two enemies.

The three books, and the man who admitted he was copying

The core library is the Brihat-trayi, the “great triad”, though that label is a later convenience the authors never used: the Charaka Samhita on internal medicine, the Sushruta Samhita on surgery, and Vagbhata’s Ashtanga Hridaya, “the Heart of the Eight Limbs.”

Open the Charaka at the end of any chapter and read the colophon, the line most readers skip. It says the treatise was made by Agnivesha, redacted by Charaka, completed by Dridhabala. Three men. Three centuries at least. The book confesses its own editing in its signature line. Dridhabala even tells us how he worked: a third of the treatise had been lost, he says, so he supplied it, forty-one chapters, exactly one third of the Charaka’s hundred and twenty, gleaned, in his own farmer’s image, grain by grain from other books.

Vagbhata is franker still. At Ashtanga Hridaya Sutrasthana 1.4 he announces that from the earlier treatises, “scattered far too widely”, he is making “a gathering of what is most essential, neither too brief nor too extensive.” He is telling you he is an anthologist, and that the books he condenses were already fixed, already authoritative, and already too many.

This is not how a revelation behaves. This is how a science behaves.

The dating is as unglamorous. G. Jan Meulenbeld, whose History of Indian Medical Literature remains the reference work, puts the Charaka and the Sushruta we can actually read between roughly 100 BCE and 500 CE. The oldest surviving object containing Indian medical writing is the Bower Manuscript, birch bark dug from a Buddhist stupa at Kucha, in the Taklamakan desert, copied around 500 CE. It was found fifteen hundred kilometres from India, in what Hoernle took to be a monk’s memorial stupa, because by the Gupta age Ayurveda had already gone travelling.

Eight limbs, and a fork in the lineage

Charaka lists the eight branches at Sutrasthana 30.28, which is why the tradition is called ashtanga, eight-limbed: internal medicine, the diseases above the collarbone, surgery (shalya, “the removal of the arrow”, a name that is itself an archaeology, since the discipline began on battlefields), toxicology, possession and what we would now call psychiatry, paediatrics, rejuvenation, and virility.

The genealogy is stranger than it looks. Both lineages begin identically, in Brahma, passed to the Ashvins, the twin physicians of the gods, and then to Indra. At Indra they fork. Down one branch Indra teaches Bharadvaja and Atreya, and the knowledge becomes internal medicine; its origin story is a council of sages convening in the Himalayas to discuss disease. A committee. Down the other he teaches Dhanvantari, incarnate as Divodasa, king of Kashi, and the knowledge becomes surgery; its origin story is a god ruling as a king, teaching disciples who ask to be taught the knife.

The tradition encodes the split between physician and surgeon in its own mythology, and the split is not decorative. The two books cannot even agree on what the five purifications of panchakarma are. Charaka’s are emesis, purgation, two kinds of enema, and nasal cleansing. Sushruta’s drop one enema and add raktamokshana, therapeutic bloodletting, because Sushruta treats blood, in effect, as a fourth dosha.

Which is exactly what a surgeon would do.

Where it came from, and who says so

Sushruta calls Ayurveda an upanga, a subsidiary limb, of the Atharvaveda, and Charaka tells the physician to declare his allegiance to that Veda, on the good ground that it is the Veda concerned with staying alive. The claim has warrant. The Vedic hymns are full of healing plants, and the Atharvaveda addresses fever as a person, ordering him away to a neighbouring tribe and telling him to take his brother Consumption and his sister Cough with him. But read those hymns beside the Charaka and the discontinuity is glaring. There is no dosha theory in the Atharvaveda. No agni, no dhatus, no anatomy, no surgery, no dietetics. It is Ayurveda’s ancestor in vocabulary and plant lore. It is not its ancestor in doctrine.

So where did the doctrine come from? The most forceful answer came from an Indian philosopher, Debiprasad Chattopadhyaya, in Science and Society in Ancient India (1977), and he made it from Sanskrit legal sources before the argument became famous in the West. Manu and his commentators, he showed, treat the physician’s learning as suspect, grouped with the logicians’ against Vedic learning. The medicine of the Charaka Samhita, he concluded, is “remarkably secular, free from supernaturalism, religion and scripture-orientation”, and its Vedic framing is later plaster over a rationalist core.

Kenneth Zysk built the philological case in 1991: healers pushed out by orthodoxy found a home among the wandering shramana ascetics, Buddhism institutionalised their medicine in the monastery, and Hinduism later absorbed the result and gave it, in his phrase, a brahmanic veneer. Nobody has refuted the philology. What is contested is the sharpness of the split, and the best evidence against it sits inside the tradition. At Sutrasthana 11.54 Charaka names three legitimate modes of therapy: the divine, meaning mantra, rites and pilgrimage; the rational, meaning drugs, diet and regimen; and the psychological. The magico-religious never left. It was written in as one leg of three.

The guillotine, and what grew back

For a decade the East India Company ran a genuinely plural medical education. The Native Medical Institution opened in Calcutta in June 1822, teaching in the vernacular, and from 1826 Ayurveda was taught at the Sanskrit College and Unani at the Calcutta Madrasa within the same system. Then a committee chaired by Dr John Grant reported in October 1834, attacking above all the absence of practical anatomy, and on 28 January 1835 the government abolished the Institution and shut the indigenous medical classes. Macaulay’s famous Minute on Education came five days later. The medical decision was first.

The Calcutta Medical College opened three weeks after that. Exactly one member of the old staff crossed over: Madhusudan Gupta, an Ayurvedic practitioner retrained in Western medicine, who in 1836 performed the first human dissection by Indians in modern India.

What grew back was not what had been cut down. Dabur was founded in 1884, the Arya Vaidya Sala at Kottakkal in 1902, the All India Ayurvedic Congress at Nasik in 1907, and the state took legal custody of the tradition in stages, from the Central Council of Indian Medicine in 1970 to a full Ministry of Ayush in November 2014. Along the way Ayurveda acquired colleges, journals, standardised syllabi, printed and edited samhitas, branded pharmacies and, as the Indian historian Projit Bihari Mukharji shows in Doctoring Traditions (2016), a pocket watch for taking the pulse. His word for the process is braiding. Rachel Berger argues that Ayurveda’s identification with a specifically Hindu genealogy was a political choice of that period rather than an inheritance. Much of what the nationalists called a revival was an invention, built in the shape of the thing it resisted. Both historians grant the counter-case, and it is real: the drug knowledge, the diagnostic framework, the classical texts and thousands of hereditary vaidya lineages ran straight through. What the revival lacked was not descent. It was purity.

Gandhi saw it early. Invited on 13 February 1921 to open the Tibbia College in Delhi, the temple of the revival, built to house Ayurveda and Unani together in defiance of the British, he told the assembled priests that their religion had gone stale. “They follow without question formulas,” he said of the hakims and vaidyas. “They carry on little investigation. The condition of indigenous medicine is truly deplorable.”

The contested ground

This is where an honest account earns its keep. And let it be clear what this page is: a description of a tradition, not medical guidance. Nothing here is a recommendation, and anyone considering a therapy or a preparation should consult a qualified physician.

The five-thousand-year claim does not survive contact with the evidence. It descends from Hoernle’s dating of Sushruta to 600 BCE, overturned more than a century ago. Philipp Maas, in the Cambridge History of Science (2018), places the first attestation of Ayurvedic concepts in the Buddhist canon of the mid first millennium BCE, and calls Indus Valley origin claims moot, since the script is undeciphered.

The heavy-metals problem is real, and it is not a problem about classical Ayurveda. Almost nobody says this, and it is the most important fact in the dispute. Early Ayurvedic pharmacy is overwhelmingly herbal. Metals and minerals enter the pharmacopoeia pervasively only after roughly the eleventh century, through rasa-shastra, the alchemical tradition of the Nath and Siddha lineages, and they reach the mainstream canon with the Sharngadhara Samhita around 1300 CE. Herbo-metallic bhasma is a medieval graft, a thousand years younger than Charaka. The sceptic who says Ayurveda poisons you and the traditionalist who says bhasmas are ancient Vedic science are arguing about the wrong millennium.

The indictment is solid. Robert Saper’s team, in JAMA in 2004 and again in 2008, found detectable lead, mercury or arsenic in roughly one in five Ayurvedic products tested, with rasa-shastra products more than twice as likely to contain them, and American-manufactured products just as likely to be contaminated as Indian-made ones. That last finding kills the “shoddy Indian factories” framing and the “Western scaremongering” framing at once. A co-author was Venkatesh Thuppil, director of India’s National Referral Centre for Lead Poisoning. The CDC has documented real poisonings, including in pregnant women.

The defence is also real, and mostly Indian. Practitioners hold that shodhana, purification, and marana, the repeated incineration that “kills” a metal, turn bulk metal into something chemically different, and the characterisation literature, much of it from the Department of Rasa Shastra at IMS-BHU in Varanasi, does show measurable phase changes and nanoparticulate products. The chemistry genuinely changes. But “it is a nanoparticle, therefore it is safe” is a non-sequitur, and nanoscale can raise bioavailability rather than lower it. There is no large, long-term human safety trial of lead- or mercury-containing bhasmas. That absence is the bottom line.

One correction cuts the other way. Pulse diagnosis is not classical Ayurveda either. Nadi-pariksha appears nowhere in Charaka and nowhere in Sushruta; it enters the corpus around 1300 CE. That is not a debunking. It is more evidence that the tradition kept developing.

What the evidence shows, and what it does not

Both of the loud claims are false. “Nothing in Ayurveda has any evidence” is false. “Ayurveda is scientifically validated” is equally false.

The most rigorous trial of classical Ayurveda yet run is Daniel Furst’s 2011 study in the Journal of Clinical Rheumatology: a double-blind, double-dummy comparison, in India, of classical individualised Ayurvedic treatment against methotrexate and against the two combined, in forty-three patients with rheumatoid arthritis. All three arms performed about equally. It is a real result, and a pilot far too small to settle anything, as its own authors, who included the Indian methodologist P. Ram Manohar, said in print. Meta-analysis finds modest benefit from Boswellia in osteoarthritis; ashwagandha trials find reductions in stress measures with low certainty. Curcumin is the cautionary tale. Nelson and colleagues argued in 2017 that as an isolated compound it is unstable, non-bioavailable and an assay artefact, and that no double-blind placebo-controlled trial of it has ever succeeded; Bahadori and Demiray answered the same year that single-target drug-discovery criteria are the wrong instrument for a multi-target polyphenol. It is a live argument among pharmacologists, and not a settled sceptic win.

The sharpest methodological indictment of Ayurvedic research comes from Ayurvedic researchers. A 2022 analysis in the Journal of Ayurveda and Integrative Medicine of the Ayurvedic COVID trials registered in India found “major flaws in terms of lack of randomization, blinding, allocation concealment, less sample sizes and less intervention duration.” India’s trials registry permitted retrospective registration until April 2018, a standing invitation to publication bias. Concede it. It is the most damaging thing an informed critic can say, and the people saying it loudest are inside the field.

The defenders have an argument worth taking seriously, and Bhushan Patwardhan of Pune has made it best. A system whose logic is individualisation, whose unit of treatment is the person rather than the disease, is awkward to test with a method whose logic is standardisation. Patwardhan is no apologist, and has attacked poor-quality AYUSH research himself. But the sceptic has a fair reply: if every negative trial can be answered with “you tested it wrong”, the claim has been placed beyond refutation.

One thing must be said flatly. There is no evidence that Ayurveda treats or cures cancer.

How Indians actually use it

Formal clinical use of AYUSH accounts for roughly five to seven per cent of India’s outpatient episodes. In the 71st round of the National Sample Survey it was 6.9 per cent, of which the Indian systems of medicine proper were about half. When Indians are ill enough to see a doctor, they overwhelmingly see an allopath.

And yet the 79th round, the first national survey dedicated to Ayush, found that around 46 per cent of rural and 53 per cent of urban individuals had used it for prevention or treatment in the previous twelve months. Both figures are true. They measure different things. (The claim that eighty per cent of Indians use Ayurveda has no Indian survey behind it. It is a garbled descendant of a WHO estimate about global traditional-medicine use.)

Indians use Ayurveda as kitchen, kitchen-garden and grandmother. Haldi doodh for a cold, ajwain for a bad stomach, tulsi, ghee, “hot” and “cold” foods, the injunction against curd at night, the seasonal turn of the diet. Almost none of that is formal Ayurvedic medicine, and almost all of it would be recognised by the texts as pathya and ritucharya, the wholesome regimen and the seasonal one. Ayurveda in India is less a medical system that people choose than a grammar of the body they think in.

The clinical lineage is thinner than the household one. In Kerala the Ashtavaidyas, the Namboothiri families held to have mastered all eight limbs, once numbered eighteen households; about four still practise. And many graduates of India’s Ayurvedic degree make their living on the peripheries of small towns prescribing antibiotics and analgesics, practising the medicine they were not trained in, because that is what their patients come and ask for.

The tradition’s own record remains the best answer to the myths told in its name. Charaka’s deepest theory of why people fall ill was not germs, which he did not have, and it was not fate. He called it prajnaparadha, the crime against wisdom: knowing perfectly well what is good for you, and doing otherwise anyway.

Two thousand years on, nobody has improved on the diagnosis.