How Do You Learn Surgery When You Are Not Allowed to Cut a Body?

A student in this tradition made his first cut into a gourd, and the text specifies which gourd.

The Sushruta Samhita gives a list. Cutting is practised on the body of a pushpaphala gourd, an alavu or bottle gourd, a watermelon, a cucumber or an ervaruka. Opening a tense swelling is practised on a full water-bag, on the bladder of a dead animal, or on the side of a leather pouch filled with water or slime. Scraping goes on a piece of skin with the hair left on it. Puncturing goes on the vein of a dead animal or on a lotus stalk. Probing goes on worm-eaten wood, on the reed of a bamboo, or in the mouth of a dried gourd. Extraction is rehearsed by taking the seeds out of a vimbi, a bilva or a jackfruit, and by pulling teeth from the jaws of a dead animal. Suturing goes on pieces of cloth, skin and hide. Bandaging goes on a full-sized doll of stuffed linen. Cautery and caustic alkali go on a piece of soft flesh. Passing a tube into the bladder is practised on a pitcher full of water with a slit cut in its side.

For anatomy, where no substitute will serve, the text gives a second workaround: prepare a dead body, cage it, sink it in a river for seven days, then abrade the softened layers away with a brush of grass roots and look at what lies beneath. At no point is a knife used on the corpse.

Eight operations, and eight things that are not a patient

The Samhita divides surgery into eight acts: chedya (cutting out), bhedya (laying open), lekhya (scraping), vedhya (puncturing), eshya (probing), aharya (extraction), visravya (drainage) and sivya (suturing). The chapter assigning a practice material to each is the ninth of the Sutrasthana, headed practical surgical instructions in Kaviraj Kunjalal Bhishagratna's translation of 1907 to 1916. Its Sanskrit name is the Yogyasutriya, the chapter of rules for practice, and it works down the list without ceremony.

The Sutrasthana elsewhere compares a physician who has read without practising to a bird attempting to fly with one wing. The gourds and bladders are what the second wing was made of.

What makes the list worth reading closely is that each material preserves one physical property of living tissue and discards the rest.

A gourd has a firm rind over soft pulp on a curved surface. Cut too shallow and the rind does not part; cut too deep and you are through into the pulp. That is depth control, and the fruits named have different resistances, so the student meets a range rather than a single feel.

A water-bag, a dead animal's bladder and a leather pouch are containers under tension, holding contents that will escape if the opening is made carelessly. An abscess, a cyst, a hydrocele and a distended bladder all behave that way. The feedback is instant and unambiguous, because the failure mode ends up on the floor.

Hide with the hair left on is a scraping target with a built-in progress indicator. Where the hair is gone, the surface has been taken down; where it remains, it has not. Evenness becomes measurable by eye, which is what a beginner cannot yet judge by feel.

The lotus stalk is the sharpest piece of thinking in the chapter. It is a thin-walled hollow tube with poor tensile strength, so it tears if the entry angle is wrong and holds if the angle is right. Venesection mattered enormously to the surgery of the period, and the Sharira Sthana gives it three chapters: the seventh on the description of the sira, the eighth on the method of venesection, the ninth on the arteries and ducts. A vein taken from a dead animal supplies realistic tissue. A lotus stalk supplies an unforgiving one.

The plank of shalmali wood coated with beeswax, used for drainage practice, is a soft layer over a hard stop. Open the wax without scoring the wood beneath and you have controlled your depth against an unyielding base.

Cloth, skin and hide give three tensile behaviours for stitching. Cautery and alkali go on soft flesh because both are irreversible and both must be judged by the appearance and smell of real tissue. The tube passed into a slit pitcher, or into the mouth of a gourd, teaches catheterisation of a lumen you cannot see, with fluid return announcing success.

Not one of these substitutes resembles a human being. Each one isolates a single mechanical problem: depth, tension, angle, evenness, blind entry. That is the design principle behind every task trainer in a modern skills laboratory, arrived at by a text that had no vocabulary for it.

The bandaging exercise is the outlier, and deliberately so. The Samhita specifies different bandage forms for different regions of the body, so the practice object cannot be a fragment. It has to be a full-sized doll of stuffed linen, because the skill is not local to one wound. The chapter closes by promising that an intelligent physician who has served this apprenticeship on models will never lose his presence of mind in practice.

The body that may not be cut

The corpse presents a problem no gourd can solve. Bones, joints, muscles and vessels have to be seen, and the surgeon most of all needs to know where not to cut.

The obstacle was not squeamishness, and the explanation that follows is a historian's reconstruction rather than anything the text says. In brahmanical practice the dead body goes to the fire substantially intact, and contact with a corpse carries ritual pollution that must afterwards be cleared by prescribed means. Cutting a body open for study sits badly against both. The Samhita itself never raises the question and offers no justification. It gives a procedure that involves no incision, and that silence is the most that can honestly be claimed: the text supplies the workaround, and the constraint has to be supplied from what is known of the surrounding culture.

The instructions, in the fifth chapter of the Sharira Sthana, begin with the choice of subject. The body should be complete in its parts, not that of someone who lived to a hundred, nor of anyone who died of a protracted disease or of poison. The reasoning is medical before it is anything else: a body wasted by long illness or great age is a poor model of ordinary anatomy, and poisoning was held to alter the tissues.

The entrails are then cleared, the body is wrapped in the outer sheaths of munja grass, kusha grass, hemp or rope, secured inside a cage so that it neither drifts nor is eaten by fish, and left in the water for seven nights. By the end of that period it has softened throughout. The observer scrapes the decomposed skin away slowly with a whisk of grass roots, hair, a kusha blade or a strip of split bamboo, and examines each structure by eye, external and internal, working inward from the skin.

That is a dissection protocol. It specifies its subject, preparation, interval and instrument, it produces a sequence of exposed layers under direct observation, and it is repeatable. What it does not use is a blade.

"For a thorough knowledge can only be acquired by comparing the accounts given in the Shastras (books on the subject) by direct personal observation."

Sushruta Samhita, Sharira Sthana, on the anatomy of the human body, verse 49 in Bhishagratna's translation

Read that against the constraint beside it and the position becomes clear. The text is not conceding that dissection is forbidden and settling for books. It insists that books alone are insufficient, then engineers a way to see for yourself that the surrounding rules will tolerate. A few verses later the requirement is explicit: the qualified physician has both read the works on the subject and observed the interior of the body himself.

Two textual notes. Bhishagratna's English places the body in "the water of a solitary and still pool", and a good deal of secondary writing has followed him. The Sanskrit does not support it: the phrase is avahantyam apagayam, in a flowing river, and translations from the Sanskrit give a stream of gentle current rather than standing water. The cage and the grass wrapping, not stillness, are what hold the softened body together. Second, verse numbers differ between editions. The statement quoted above is verse 49 in Bhishagratna and 48 in some others, and the training chapter carries no verse numbers in Bhishagratna.

What the counts reveal about how the looking was done

The same chapter enumerates the body: 300 bones, 210 joints, 900 snayu, 500 muscles with twenty more in women, ten of those in the breasts, then 700 sira, 24 dhamani, 16 kandara and 107 marma.

Set that against the Charaka Samhita, which gives 360 bones. Two authorities inside one tradition, sixty apart. The Ayurvedic literature explains the gap through counting conventions rather than through observation: Charaka's total counts the tooth sockets as bone, along with teeth and nails, and Sushruta's does not, while the two divide the pelvis, sacrum and back differently. Sushruta's 300 still includes teeth and cartilaginous structures, which is why it exceeds the modern adult figure of 206 rather than falling short of it.

A disagreement about how to count is not a disagreement about what is there. It is evidence that counting was happening, and that its rules were contested.

The more revealing pattern is where the text is precise and where it is round. It gives seven layers of skin, each with a thickness expressed as a fraction of a grain of rice: the outermost at one eighteenth of a grain, then one sixteenth, one twelfth, one eighth, one fifth, then a full grain, then two. That resolution appears in the fourth chapter of the Sharira Sthana, the chapter on the development of the fetus. Bhishagratna's English renders the first five as twentieths rather than as this ascending series; the ascending reading, in which each layer is thicker than the one above, is standard in the Ayurvedic literature. Meanwhile the vessels come in figures like 700 and 24, and the ligaments in a flat 900.

This is inference rather than a sourced finding, but a body softened for seven days and taken apart from the outside inward with an abrasive brush would be expected to yield roughly this profile. Superficial stratification measured in grain widths, because that is what the brush meets first. Round numbers for the deep and continuous systems, because a week of immersion is unlikely to leave fine vessels traceable or brain tissue readable.

The marma count belongs to a different logic again. The 107 vital points get a chapter of their own, the sixth of the Sharira Sthana, and they are defined by what happens when they are injured: knowledge from the battlefield and the operating couch rather than from the water cage. That reading follows from the chapter's organisation and is not a claim the text makes. For a surgeon, the marma are the map of where the knife must not go.

Why surgical educators keep returning to this chapter

The training chapter turns up regularly in the modern literature on simulation-based medical education. A 2022 review by Jared Bienstock and Albert Heuer in Medicine (Baltimore) calls the Sushruta Samhita the oldest written evidence of simulation in healthcare education, and the point recurs whenever anyone traces the field's history back past the twentieth century.

The continuity is one of logic rather than of descent. Nobody in a modern skills laboratory learned their curriculum from a Sanskrit text. People facing the same problem reached the same answer, roughly two thousand years apart.

Look at what surgeons still practise on. Grapes and banana peel for fine suturing. Chicken thighs and pig trotters for tendon and vessel work. Synthetic skin pads with layered densities. Box trainers with ports and instruments and no patient. Perfused cadaveric limbs. Virtual reality rigs with force feedback. Every one does what a gourd and a lotus stalk did: it separates acquisition of the hand from treatment of a person, and puts the cost of failure on an object that does not mind.

The sequence matches too. In the Samhita the student practises on models, observes procedures being performed, and only then operates. That is the shape of a modern surgical curriculum, and a more defensible shape than the apprenticeship maxim of watching one case and then doing one, which the profession has spent much of the last three decades dismantling. The dismantling shows up in the requirements: from the 2009 to 2010 academic year, the American Board of Surgery has required general surgery residents to pass the Fundamentals of Laparoscopic Surgery examination, a simulator-based skills test, before sitting the qualifying exam. Competence on a box trainer became a gate, and the gate stands where the Yogyasutriya put it.

What the text gets wrong, and why that is the expected result

The anatomy has real failures, and they are not embarrassing ones.

There is no closed circulation in the Samhita. The 700 sira and 24 dhamani form a classificatory scheme rather than a vascular map, they are described as carrying more than blood, and nothing in the text corresponds to a pumped circuit with a pulmonary loop. Structures called snayu and kandara are sometimes translated as nerves, which imports a category the text did not have; they answer better to ligaments, tendons and fibrous bands. Organ function is often assigned on grounds drawn from the theory of the dhatus rather than from anything the brush revealed.

Set that beside the method and it is close to what should be expected. Seven days of maceration trades depth of access for permission to look at all. It hands over the layers of the skin and the gross architecture of bones and joints in good order. It could never deliver circulation or nervous conduction, which need continuity of fine vessels, readable soft tissue, and a fresh specimen in which a ligature can show where blood is going.

The date needs stating honestly too. Gerrit Jan Meulenbeld, still the standard reference on the history of Indian medical literature, holds that the Sushruta Samhita contains several historical layers, that composition may have begun in the last centuries before the common era, and that the surviving form was produced by a later author who redacted the first five sections and added the long final section, the Uttaratantra. Ray, Gupta and Roy place the recension associated with Nagarjuna, which underlies Dalhana's commentary, in the third or fourth century of the common era. A single confident date is not something the evidence supports.

None of this weakens the training chapter, which is not an anatomical claim but a piece of pedagogy, and pedagogy either works or it does not. A student who has cut five kinds of gourd, opened a filled bladder without spilling it, punctured a lotus stalk without tearing it, stitched cloth and hide and skin, and bandaged every region of a linen doll arrives at his first patient with a hand that has already made most of its mistakes.

The chapter's own promise is narrower than the one modern retellings put in its mouth. It does not say the student will understand the body, and it does not say his patients will live. It says he will not lose his presence of mind. Given how much of the anatomy seven days in the river could never have shown him, that was the promise the text was in a position to keep.