Tantra Yukti in the Sushruta Samhita: 32 Devices for Reading a Classical Text, and the 33rd That Slipped In

📖 25 min read · Classical text study · Sushruta Samhita, Uttara Tantra, Chapter 65

Where Chapter 65 Sits in the Sushruta Samhita

The Sushruta Samhita is arranged in six sthanas. Five of them — Sutra Sthana, Nidana Sthana, Sharira Sthana, Chikitsa Sthana and Kalpa Sthana — form the original body of the work. The sixth, the Uttara Tantra, is the long supplement, and it runs to sixty-six chapters covering the eye, the ear, the nose, the disorders of infancy, the graha material, fevers and much else besides.

By the time a reader reaches chapter 65 the surgical business of the text is finished. What is left is housekeeping — and the chapter that Sushruta puts here, second from last, is not about a substance, an organ or a condition at all. Its title is Tantra-Yukti-Nama-Adhyaya: the chapter naming the technical devices used in this treatise.

Set that placement against what the same section spends its other chapters doing and the oddity is immediate. Chapter 40 works through atisara at exhausting length; chapter 65 explains what a sentence is doing. One is medicine, the other the instruction manual for reading the medicine. It is easy to skip — most translations bury it and it contains not one preparation. It is also one of the most quietly radical chapters in the classical corpus, because it takes the position that a treatise is a machine with named parts, and that a physician who cannot name the parts cannot be trusted with the contents.

What Tantra Yukti Means, and Why a Surgical Text Ends in Grammar

Tantra here means the treatise itself — a system, a woven thing, a body of organised teaching. Yukti means a device, a joining, a contrivance, a piece of skilled fitting. Together, tantra yukti means something close to the joinery of a treatise: the named techniques by which separate statements are fastened into a structure that carries meaning.

Sushruta asks the obvious question himself and answers it in a single line. What is the necessity of these devices? The answer given is: for connecting words together — that is, for making up sentences and giving a sense or a meaning to them.

Then the chapter supplies its reasoning in what the text calls memorable verses, and here the tone shifts from grammar to something more combative. By the use of technical devices in a scientific treatise, it says, the points of argument of the opposing party are frustrated and the points of one's own argument are established. The meanings of words — whether clearly used or not, whether direct or indirect, whether only partially used — are made distinct.

The lamp adds nothing to the room. It only makes the room visible.

And then the simile that gives the chapter its character. Just as the sun shows a cluster of lotus at its best, and a lighted lamp shows the inside of a room at its best, so the technical devices used in a treatise show the intended meaning clearly.

Notice what that image concedes. The lamp does not add anything to the room. The sun does not improve the lotus. Both merely make visible what was already entirely there. Sushruta is not claiming these devices generate meaning — he is claiming that meaning is already fully present in a well-made treatise, and that without the devices you simply cannot see it. The failure, in his framing, is never the text's. It is the reader's equipment.

The Thirty-Two, in Sushruta's Own Order

The chapter opens by naming all thirty-two before defining any of them. That opening list is the control against which everything after it should be checked — a point that becomes important later. Here it is, in the order the text gives, with a plain-English gloss.

Thirty-two names, stated once, in a fixed order. Keep the order. It is the evidence.

Named once, in a fixed order. That order is the evidence.

Adhikarana to Atidesha: Naming, Joining and Inferring

The first ten devices are the basic carpentry — how a subject is set up, how words are held together, and how the text reasons from one case to another.

Adhikarana, the subject, is simply what a passage is about. Sushruta's examples are as bare as they could be: on the subject of rasa, or on the subject of dosha. It sounds too obvious to name, until you meet a chapter where three subjects are running at once and the text never says which one a given verse belongs to.

Yoga, joining, covers the case where two words that must be read together sit far apart in the sentence. The example the chapter gives is a line in which the word meaning “prepared” sits in one half of the verse and the word meaning “should be drunk” sits well away from it in another — and the two must nevertheless be joined for the line to mean anything. Sanskrit verse scatters words for the sake of metre; yoga is the reader's licence to gather them back up.

Padartha is the sense a word carries in this particular place. The chapter's own illustration is a small masterpiece of pedantry: take the phrase “we shall discourse on the origin of the Veda.” Which Veda? There are several, and the word alone cannot tell you. Only the surrounding words settle it — and once they do, it becomes clear the subject is the origin of Ayurveda. The text notes that words such as sneha, sveda or anjana each carry two or three possible senses, and only one will agree with what stands around it.

Hetvartha is a meaning carried indirectly, usually by analogy. The example runs: as earth is moistened by water, so an ulcer is moistened by certain foods. The comparison is not the claim; the comparison is the vehicle for a claim never stated outright.

Uddesha and Nirdesha are a pair, and they are the reason so many classical chapters seem to say the same thing twice. Uddesha is the brief statement — the chapter's example is the single word salya, an obstructing foreign matter. Nirdesha is the same thing stated in full: salya is of two kinds, sharira and agantu. Announce, then unfold. Once you see the pair, the apparent repetition in the samhitas turns out to be an index followed by its own expansion.

Upadesha is a direct instruction on doing something a particular way. The chapter's example is a line of daily regimen: one should not sit up at night, and one should give up sleeping in the daytime. Readers who know the classical dinacharya of Vagbhata will recognise the register.

Apadesha is the statement of a reason — the chapter's illustration being the specification that shleshma is increased by articles of sweet taste. It names a cause rather than merely reporting an outcome.

Pradesha and Atidesha are the two inference devices, and they run in opposite directions in time. Pradesha settles a present case from a past one: this person's salya was extracted by this practitioner, therefore that person's will be too. Atidesha projects forward from a present case: vayu courses upward under such and such an action, therefore that action may bring on udavarta. The second is precisely the reasoning behind the classical warnings about suppressing the natural urges, which we have taken apart separately in our study of vega dharana and the adharaniya vega.

Apavarga to Anumata: Exception, Presumption and Reversal

The middle stretch of the list is where the chapter becomes genuinely useful, because these are the devices that account for most of the apparent contradictions in classical literature.

Apavarga is the exception carved out of a wider rule. The chapter's own example: fomentation should not be applied to those suffering the effects of poisoning — excepting those affected by insect poison. A general prohibition with a hole cut in it. Anyone who has watched two people argue about what “Ayurveda says” about a given practice has usually watched one of them quote the rule and the other quote the apavarga.

Vakya-shesha is the word omitted because its absence costs nothing. When the text says the head, the hands, the legs, the sides, the back, the abdomen and the chest, it does not need to add “of a person” — that much is evident. Classical Sanskrit is relentlessly economical, and a great deal of what reads as gnomic compression in translation is simply vakya-shesha doing its job.

Apavarga: the general rule, and the one case cut out of it.

Arthapatti is presumption: the sense that is not stated but cannot be avoided. The chapter's example is beautifully ordinary. If one person tells another “this solid food may be taken,” it becomes evident that he is not willing for a liquid gruel to be taken. Nothing was said about gruel. Everything was said about gruel. The translator's footnote adds the classic textbook case from Sanskrit philosophy — the man who is fat and does not eat during the day, from which it follows that he certainly eats at night, or he could not have become fat.

Viparyaya is reversal: words that convey the opposite of what they name. When the text says that emaciated, weak and frightened persons are very difficult to manage, the opposite sense becomes evident — that strong persons are readily managed. The text states one half of a proposition and expects the reader to hold both.

Prasanga, connected reasoning, covers the case where the same point resurfaces in a different place and in different words. The chapter's illustration is a genuine cross-reference within Sushruta's own work: the first chapter of the Sutra Sthana, on the origin of the Veda, states that purusha is the sum of the mahabhutas together with the soul, and that purusha is the subject matter of every action — and the point is then repeated in the Bhuta-vidya material much later. We have taken both halves of that idea apart separately, in our studies of the pancha mahabhuta and of the karma purusha in Sharira Sthana chapter 1.

Ekanta and Anekanta are the certainty pair, and they are the sharpest tools on the list. Ekanta marks what holds in every case: the chapter's examples are that trivrit produces purgation and that the madana fruit produces vomiting — two of the most heavily documented dravyas in the corpus, each of which we have covered at length in our reading of trivrit and the two root colours and of madanaphala in the Kalpa Sthana.

Anekanta marks the opposite — and the example Sushruta chooses to illustrate uncertainty is extraordinary, because it is a live disagreement at the foundation of his own discipline. Many authorities hold that dravya, the substance itself, is the principal factor. Some hold that rasa, taste, is principal. Some hold virya, potency. Others hold vipaka, the post-digestive transformation. Four schools, no verdict. The chapter files the entire question under “certain in some cases, uncertain in others” and moves on. Our article on rasa, virya and vipaka walks the same four-way disagreement from Charaka's side.

Purva-paksha and Nirnaya are the debate pair: an objection raised, then settled. The chapter's example asks why the four vataja varieties of prameha are counted as asadhya — beyond the reach of the measures available — and then answers it by describing how vayu spreads through the body and carries the vitiated vasa, medas and majjan downward. The classification of prameha is set out at length in our study of the twenty types of prameha in the Nidana Sthana. What matters here is only the shape: a question that looks like a doubt is actually a rhetorical scaffold, and the answer was always coming.

Anumata is the subtlest device in the chapter and the one most often missed. It marks an opinion belonging to someone else, quoted but not refuted — and therefore, by the silence, allowed to stand. The chapter's example: some authorities hold that there are seven rasas. Sushruta does not agree. Sushruta does not disagree either. He records the view and passes on, and the classical convention is that unrefuted means sanctioned. Our article on the six tastes and the classical thali rule covers the majority position that six became the standard count.

Vidhana to Uhya: Time, Doubt, and What You Must Supply

The last stretch handles the management of a long text — how it points forward and back, how it flags doubt, and what it expects the reader to do without help.

Vidhana is stating at the outset what will be established, so the reader knows where the argument is going; the chapter's example is the count of marmans, the vital points, in the thigh — a figure declared in advance and then demonstrated.

Anagata-vekshana and Atikranta-vekshana are the two signposts, stated almost as templates. The forward reference: in the Sutra Sthana one may say “this will be dealt with in the Chikitsita Sthana.” The backward reference: in the Chikitsita Sthana, “this has already been said in the Sutra Sthana.” Unglamorous, and completely essential — without them a work of this size could not be navigated at all.

Samshaya, doubt, is the citing of two opposite and dissimilar cases side by side so that the difficulty is felt rather than described. The chapter's example places a fatal injury to one marma beside the survivable loss of a whole limb — two facts that sit badly together and are meant to.

Vyakhyana is explanation of detail, and Sushruta's example is doctrinal self-positioning: purusha as the twenty-fifth factor is dealt with in this book, whereas other works count only the twenty-four constituting the body. He is marking where his count differs from the neighbours'.

Niyoga is the enjoining of a duty; its example is the shortest line in the chapter — only what is pathya, what is suitable, should be taken. Samuchchaya joins several independent items into one group, illustrated by grouping named items from the text's own dietary material under a single heading. Vikalpa is the alternative — either this or that — which is why so many classical instructions offer a choice rather than one option.

Which leaves Uhya, the last of the thirty-two, and the most demanding. Uhya is what an intelligent reader is expected to work out although it was never definitely stated. The chapter gives it a full page and we will come back to it, because the example Sushruta chooses is his own book's table of contents.

Two signposts in one stack, one pointing forward and one back.

The Thirty-Third: Uddharana and Dalhana's Note

Now count.

The chapter names thirty-two devices in its opening list. It then defines them in order — adhikarana, yoga, padartha, and onward down the sequence exactly as listed. It reaches Sva-samjna, number twenty-six on the list. And then, before it arrives at Nirvachana, number twenty-seven, an extra entry appears:

The translator flags it in a footnote on the page, and the note is unusually blunt. Uddharana has been recognised here as a technical device, it says, but it should not have been, since it was not included in the list. Had it been included, the number would have been thirty-three and not thirty-two.

And then the footnote reaches for the commentary. Dalhana — the twelfth-century commentator whose Nibandha Samgraha is the standard gloss on the Sushruta Samhita, and the reason a great many of its difficult passages are readable at all — prefers to regard the passage as an interpolation, and reads the material as a continuation of the preceding example rather than as a device in its own right.

Four tidy rows, and one that belongs to no row at all.

That is the find, and it is worth being precise about why it is more than trivia. The chapter's entire claim is that a treatise is auditable — that its parts are countable, nameable and checkable, and that a competent reader can hold a text to its own declared structure. Chapter 65 states a number. Chapter 65 then fails to match it. The mismatch is detectable by exactly one method: taking the opening uddesha, the brief statement, and checking it against the nirdesha, the detailed unfolding — which is devices five and six on its own list.

The chapter can be audited by its own tools, and when you audit it, it does not balance. Whether a scribe added Uddharana centuries after the fact, or an author listed thirty-two and then wrote thirty-three, the discrepancy is real, it is visible in the text, and it was caught by a commentator who was doing nothing more than what the chapter told him to do.

There is a second, quieter arithmetic problem two chapters later, and it points the same way. Chapter 66 opens with Sushruta asking Divodasa about the sixty-two varieties of the doshas — and the footnote there observes that the reference is to chapter 63, where the number given is sixty-three, not sixty-two. The commentators reconcile it by noting that the state in which all three are normal is the sixty-third combination, and that three dhatus in their normal state cannot properly be called doshas at all. It is a reconciliation, not a coincidence — and the sixty-two/sixty-three question is one we have followed from Vagbhata's side in our study of the doshabhediya adhyaya and the sixty-two dosha combinations.

Two counting discrepancies within three chapters of each other, both caught by commentators and both resolved by careful reading rather than emendation. This is what a living scholarly tradition looks like from the inside.

Nirvachana: The Chapter That Derives the Word Ayurveda

Device twenty-seven, Nirvachana, is the derivation of a term from its parts — and Sushruta uses it on the name of his own discipline.

Ayus, life, is the subject matter of this work; a person obtains the means of ayus — longevity — from this work; and hence it is called Ayurveda. Ayus plus veda: the knowing of life.

It is a small moment and an unusually self-aware one: the text picks a device off its own list and turns it on its own name, demonstrating the tool and justifying the title in one sentence. Nirvachana is not modern historical etymology; it is a classical genre in which the parts of a word are read as an argument about what the thing is for.

The device shows up everywhere once you know to look for it. Tridosha parses as three (tri) things that can spoil (dush, to become vitiated) — which is why the standard gloss of dosha is not “humour” but something closer to “that which is liable to go wrong”; see our guide to the tridosha. Viruddha ahara parses as food set against itself, which is the whole thesis of our article on viruddha ahara. Read this way, much of classical terminology stops being vocabulary to memorise and becomes compressed argument to unpack.

Nidarshana, twenty-eight, sits immediately after it and is its companion: meaning supported by an illustration. The chapter's example compares the way fire in the koshtha increases in contact with vayu to the way an ulcer increases when assisted by the doshas. The agni–vayu relation, which our article on jatharagni in the Bhavaprakasha examines in detail, is being used here as a familiar case to make an unfamiliar one intelligible.

Svasamjna: When a Text Redefines a Common Word

Device twenty-six, Sva-samjna, means a text's own special term — a word this work uses in a sense not shared with other works. It is, for a modern reader, probably the single most practically useful item on the whole list, because it is the formal acknowledgement that classical vocabulary is not portable between texts.

The example Sushruta gives is exact and memorable. In medical works, the word Mithuna means two specific things: honey and clarified butter. In ordinary Sanskrit it means a pair or a couple. The medical tradition has taken a common word and assigned it a narrow technical value, and a reader who imports the ordinary meaning will misread every line it appears in.

Two ordinary things which, in this treatise only, are one term.

The implications run a long way. Rasa means taste among the six flavours, the first of the seven dhatus in tissue formation, and a mercurial preparation in later rasa shastra — three referents behind one syllable. Kalpa means a preparation, a procedure or a chapter-heading depending on where you stand.

Sushruta's point is that the responsibility for keeping these straight is not the text's. It is yours — and he has given the problem a name so you cannot pretend it is a translation error.

Uhya: The Chapter Title That Names Only Half Its Subject

The thirty-second and final device is Uhya — what can be understood by an intelligent person although it was never definitely stated. And the example Sushruta selects to close his list is his own book's chapter title.

The Sutra Sthana has a chapter on Anna-pana-vidhi — the rules of food and drink. In it, anna is declared to be of four kinds:

Four kinds. But the chapter's own title names only two of them — anna and paniya, food and drink. The other two are simply understood.

And the chapter explains exactly how the understanding works. Bhakshya is included within anna, both being solid. Lehya is included within peya, both being liquid. Articles of food, though really of four kinds, are ordinarily spoken of in common language as being of two. So when two are named in respect of food and drink, the inclusion of all four is easily comprehended.

Look at what that does as a closing move. The last device on the list is the one saying some of the structure is not written down and you must reconstruct it — and it is demonstrated on a chapter heading, the very surface a reader trusts most to be complete. Our article on the classical classification of foods in Vagbhata rests on the same fourfold division.

Having taught you to distrust a chapter title, the chapter then ends — two paragraphs later — having quietly failed to match its own opening count. Whether that is design or accident, it is a good joke either way.

Why Dhanvantari Ranks the Reader Beside the Physician

The chapter closes with a verse, and the verse is the reason this material matters beyond philology.

Here, it says, have been fully described the thirty-two technical devices for the investigation into the essence of this treatise. And then: the intelligent person who is fully conversant with these devices — which work like lights, as it were — is to be regarded as the greatest physician, and to be held in great esteem. This, the text says, is what the sage Dhanvantari says.

Read that claim slowly, because it is a strong one. The qualification being described is not surgical skill. It is not knowledge of dravyas. It is not diagnostic experience. It is the ability to read the treatise correctly — and the text ranks that ability at the very top, and puts the ranking in the mouth of the divine physician who is the tradition's own source.

The lamp image returns here from the chapter's opening, and it has hardened. At the start the devices showed the meaning, as a lamp shows a room; at the close, the person holding the lamp is the greatest physician. The colophon then records what has ended: the sixty-fifth chapter of the Uttara Tantra, dealing with the technical terms used in this work.

Reading the Samhitas With the Thirty-Two in Hand

The practical payoff is that the chapter converts a vague instinct — “classical texts are hard” — into specific questions you can put to any passage.

Is this the brief or the detailed statement? If a line seems thin, look for its nirdesha further on; if it seems repetitive, you have probably found the uddesha it expands.

Is this the rule or the exception? A great many confident claims about what Ayurveda “forbids” are general rules quoted with the apavarga left off.

Whose opinion is this? If a view is recorded and not argued with, it is anumata — tolerated, not asserted. Do not quote it as the author's position.

Is this a real question or a scaffold? A purva-paksha looks like doubt and is actually setup. The answer is always already written.

Does this word mean here what it means elsewhere? Ask sva-samjna first, always — it is the most common source of confident misreading. And what has been left for me to supply? If a list seems short or a title narrow, uhya may be operating and the missing members may be recoverable from the structure.

These questions work on any of the great texts. Our walk through the Charaka Samhita and its eight sthanas uses them constantly without naming them, as do our overviews of the Sushruta Samhita and the Ashtanga Hridaya of Vagbhata. Chapter 65 is where the tradition writes the method down instead of assuming it.

The device-list genre is not confined to medicine, either — Kautilya's Arthashastra closes with a section on tantra yukti of its own, which suggests the apparatus belonged to Sanskrit systematic writing generally rather than to physicians alone.

Yesterday's article on gandha pariksha in Sharira Sthana chapter 2 is, in retrospect, an exercise in exactly this: it reads a chapter's arrangement — which register appears in the verdict and which does not — rather than its content. Chapter 65 is the licence for that kind of reading, written by the text itself.

Nirvachana on a Shelf: Three Names and What They Derive From

One device from the thirty-two travels out of the manuscript more easily than the rest, and it is Nirvachana — the derivation of a name from its parts.

Classical product names are not branding. They are descriptions, and they parse. In the spirit of device twenty-seven, here are three things from our shelf, described purely by what their names break down into, checked against what is actually on each ingredient list. No other claim is being made about any of them.

Questions Readers Ask

Shop All Ayurveda Hub Products →