Two ways a drink can be ruined and they are not the same failure.
The first: something is added that fights the drink. Salt in nectar. You taste the conflict and the whole thing is spoiled — not weakened, spoiled, because the added thing is actively at war with what it went into.
The second: something falls in that does nothing at all. A blade of grass. It is not bitter. It does not fight anything. It is simply in your mouth, and it was not supposed to be, and now you are thinking about a blade of grass instead of about nectar.
This Text names both, keeps them apart, and builds a twelve-by-twelve table on the first one.
Viśvanātha's root of Text 25 is compressed almost to a telegram:
Śāntaanddāsyaare mutually friendly.Sakhyaandvātsalyaare neutral.Vātsalyahas friendship with none.Ujjvalaanddāsyaare enemies.1
Four claims about five rasas. Then the commentary immediately widens the frame — the above description is a very brief summary… based only upon the consideration of mukhya-rasa — and reproduces a full chart over all twelve, sourced to Bhakti-rasāmṛta-sindhu 4.8.2–62.2
Which resolves the root's hardest line. Vātsalya has friendship with none would be false against the chart, where parental love is compatible with laughter, compassion, fear and astonishment. Scoped to the five primaries, it holds exactly: vātsalya's compatible list contains no primary rasa at all.
Parental love is the one primary relationship that cannot be combined with any other primary relationship. Stated in the root, unexplained.
The chart's most important feature is that it does not behave like a compatibility table.
A compatibility table is symmetric. If A goes with B, B goes with A. This one does not, and the commentary says so and gives an example:
Differences in compatibility are based upon which rasa is acting as the
aṅgi-rasa. For example, whendāsyais theaṅgi-rasa,mādhuryais incompatible. But whenmādhuryais theaṅgi-rasa,dāsyais neutral. This is becausedāsyacannot accommodate the sentiment of amorous love, whereasmādhuryacan accommodate the sentiment of servitude. Other similar peculiarities may be noted.3
So the entries are not statements about pairs. They are statements about what a given host can take in without being deformed — and a servant's love cannot hold an amorous element while an amorous love can hold a servile one.
Other similar peculiarities may be noted. The book supplies one and invites the reader to find the rest.
Read across the chart and the asymmetries are not occasional. Counted against each other, entries disagree in roughly fifteen places out of the sixty-six pairs.
Sakhya calls vātsalya incompatible; vātsalya calls sakhya neutral. Vātsalya calls bhayānaka compatible; bhayānaka calls vātsalya neutral. Adbhuta calls karuṇa neutral; karuṇa calls adbhuta incompatible. Vīra calls raudra neutral; raudra calls vīra compatible.
And then one that is not a peculiarity but a structure.
Every one of the five primary rasas lists raudra as incompatible. Śānta, dāsya, sakhya, vātsalya, mādhurya — all five, without exception.
And raudra, as host, refuses only one of them. Its own incompatible list is hāsya, mādhurya and bhayānaka; it is neutral to śānta, dāsya, sakhya and vātsalya.4
So anger admits almost everything and is admitted by nothing. As a guest it is incompatible in seven of eleven cases; as a host, in three of eleven.
That is a complete enumeration over a printed table, not a sample, and it needs no hedging as arithmetic. The most permissive host in the system is the least welcome guest, and no rasa in this book is more structurally isolated than anger.
Put it beside what the anger material actually contains and the isolation stops looking like a technicality.
Raudra-rasa — Rohiṇī furious at Yaśodā, Yaśodā at the boys, Lalitā at Rādhā, the gopīs at Akrūra — is, on this build's reading of Text 24, a map of positions in Kṛṣṇa's household wearing an emotion's name. Nothing about the anger varies; only its target does.
And the compatibility chart says that map cannot be laid over any of the five relationships without producing virasatā.
Which is a strange result for a rasa whose every worked example takes place inside those relationships. Yaśodā's anger occurs in vātsalya. Lalitā's occurs in mādhurya's inner circle. The chart says neither combination is admissible.
The reconciliation the system supplies is not in Text 25 but in Text 26: krodha-rati of raudra-rasa is included within one aspect of vātsalya- and mādhurya-rasa. Included, not combined. A demoted state absorbed into a primary is a different operation from two rasas mixing — and it is the operation the retraction at the close of Text 24 had already prescribed.
So the two Texts do not contradict each other. They describe anger's two available routes: it cannot mix with a primary rasa, and it can be swallowed by one. [n=1 OBSERVED] on the reading; what is not a reading is that both statements are printed, three pages apart, and neither refers to the other.
And the opposite pole is worth naming, because the chart makes it exact.
Adbhuta — astonishment — is compatible with all five primary rasas, and all five list it as compatible in return. It is the only entry in the twelve with that property, and the symmetry is perfect.
Text 26 says the same thing in its own register: adbhuta may be included in all five mukhya-rasas.
Which is a satisfying result for a devotional aesthetics, and it is not obviously a doctrinal decision. Astonishment is the one reaction that does not compete with a relationship — you can be astonished at your master, your friend, your child or your lover without the astonishment displacing what they are to you. Anger cannot do that, because anger reassigns the relationship for as long as it lasts.
The book does not offer this explanation. It prints the two extremes at opposite ends of the same chart.
The chart needs a theory of what "combining" means, and Text 25 supplies one:
When there is a mixture of rasas, the rasa that receives nourishment is known as the whole, or
aṅgī-rasa, and the rasa that provides nourishment is known as the component, oraṅga-rasa. When there is a meeting of many rasas, the rasa that is tasted most prominently in a given situation is theaṅgi-rasa.5
Then the constraint that makes the whole apparatus personal rather than situational:
The devotees who are the
āśrayaof amukhya-rasathat becomes the prominent rasa in the development of a particular pastime are always theāśrayaof that rasa. For those devotees, whenever there is a mixture of differentmukhya-rasas, that particularmukhya-rasawill always be theaṅgi-rasaand the othermukhya-rasaswill be component parts.5
So the host is not chosen per scene. A devotee has one primary rasa and it hosts, permanently. Which means the chart is read down a single column for any given person: if you are vātsalya, the only row that will ever apply to you is the vātsalya row, and mādhurya, yuddha-vīra, dāsya and raudra are permanently unavailable as components of your experience.
That is a much stronger claim than the chart alone suggests, and it converts a compatibility table into a specification of what a given devotee's emotional life may contain.
You are combining material — in writing, in smaraṇa, in a talk — and something in the mixture is not working.
The Text gives you two diagnoses and they call for opposite fixes. If the added element is fighting the dominant one, that is virasatā, and the chart will usually tell you in advance: check whether the guest appears in the host's incompatible list. If it is merely sitting there, that is the blade of grass, and the chart will not help you at all, because the failure is not incompatibility but purposelessness.
Only when an
aṅga-rasaincreases the relish of theaṅgi-rasadoes it serve any purpose as a component part. Otherwise, its mixture with theaṅgi-rasais fruitless… like a blade of grass that accidentally falls into a drink of nectar.5
The test for the second failure is not compatibility. It is whether the addition increases the relish of the thing it went into. A compatible element that does nothing is still a defect.
Second scene. You want to bring anger into devotional material — righteous anger, anger at injustice, anger at God.
The chart's answer is unusually firm: not as a component. All five primary relationships refuse it. What the system permits instead is absorption — anger as one aspect inside parental or amorous love, which is where Yaśodā's and Lalitā's actually sit. The distinction is not pedantry: a combination keeps both rasas visible, and an inclusion does not.
Third. If you know your own primary rasa, the chart is read as a single row and most of it is inapplicable to you. That is the tradition's own claim, and it is worth knowing before treating the twelve as a palette. For any one devotee it is not a palette. It is one colour and a list of what may be mixed into it.
The chart's format is doing something to its content, and it is worth separating them.
Printed, it looks like twelve entries each with three lists. Read as data, it is a directed graph: sixty-six ordered pairs, each labelled compatible, incompatible or neutral, with roughly fifteen of them labelled differently in the two directions.
Nothing in the presentation signals this. There is no arrow, no note that the entries are host-relative until the final paragraph, and no marking on the fifteen pairs that disagree. A reader who assumes symmetry — which is what a table of this shape invites — will get fifteen answers wrong and will have no way to notice.
The commentary's closing line is an acknowledgement that stops short of a remedy: other similar peculiarities may be noted. True, and the noting is left entirely to the reader.
This is the same species of thing the build has been recording all cluster. The enumerations in this book have no mechanism for noticing relations among their own entries — residence nested in three limbs, four items appearing in two lists with incompatible framings, karuṇā dropped from an inclusion table that would otherwise be complete. [n≥3 PATTERN], long since established.
Here it takes its most quantifiable form yet. The relation being missed is not between two distant passages; it is between the two halves of a single table, and the book itself supplies the one example that proves it knew.
Which suggests the honest description of this chart is not a compatibility matrix with some peculiarities. It is a set of twelve independent host-profiles that has been arranged to look like a matrix, and the arrangement is the artefact.
Strongest support. The root's four claims, the twelve-entry chart, the aṅgī/aṅga definitions, the permanent-host constraint, the two failure modes and the dāsya/mādhurya asymmetry with its stated reason are all in Text 25. The chart is attributed to Sindhu 4.8.2–62, which Appendix 2 independently confirms.
Tension one. Every primary rasa refuses raudra, and raudra refuses only one of them — so the rasa whose every worked example occurs inside a primary relationship cannot combine with any of them. Text 26's inclusion clause is the only route back and neither passage cites the other.
Tension two. The root says vātsalya has friendship with none. The chart gives it four compatible partners. The commentary's scoping to the five primaries reconciles them, and the root read alone says something stronger than the chart supports.
Tension three. Roughly fifteen of sixty-six pairs are labelled asymmetrically, one is explained, and the rest are left to the reader with a two-word invitation.
Open question. Karuṇa's incompatible list names sambhogātmaka-mādhurya specifically — union-flavoured amorous love — where every other entry names a rasa flat. It is the one place the chart shows its working. Whether other entries carry unstated qualifications of the same kind is not answerable from the range read.
The vault's Pollock-derived rasābhāsa material treats incompatible combination as a creative resource: the semblance of rasa as something a sophisticated author deploys deliberately, a tool rather than a fault.
This Text has no such register. Virasatā here is salt in nectar — a spoiling, full stop — and the chart exists to help a practitioner avoid it rather than to help an author exploit it.
That difference is the whole distance between the two literatures, and it follows from the relocation this build has been tracking. When rasa is something an audience tastes, an author may sabotage it on purpose for effect. When rasa is something a devotee is in, there is no author and no effect to be had — spoiling it just spoils it.
Which also explains the permanent-host constraint, which would be unintelligible in a literary setting. A dramatist can put any rasa at the centre of any scene. A devotee cannot, because the host is not a compositional choice but a fact about who they are — and the vault's sāmānya/svaccha page has already recorded the book's own uncomfortable account of how that fact gets set: by whose company the practice matured in.
A pairing scheme for emotional states, and the question of whether pairing is symmetric. Set this against Pranayama-Rasa Pairings for Reception, the vault's own working method: each rasa matched to a breath practice that produces the physiological substrate in which that rasa is best received or produced, so that a reader tuned to śānta receives śānta more deeply.
The reflexive value is the point. The vault holds a rasa-pairing scheme and it is undirected — a rasa and a breath, matched, with no claim that the matching runs one way. This Text holds the tradition's own pairing scheme over the same categories, and it is directed, in fifteen places out of sixty-six, with the direction determined by which member is hosting.
Which raises a question the vault's page cannot ask itself: is the substrate relation directed too? The classical claim is that dāsya cannot accommodate amorous love while mādhurya can accommodate servitude — an asymmetry of capacity, not of similarity. If something like that holds for breath and reception, then a mādhurya-tuned nervous system might receive śānta material without loss while a śānta-tuned one distorts mādhurya, and the pairing table would need arrows.
That is checkable, and it is checkable in the direction that matters for practice: not do these two go together but which one is hosting. The classical table's entire structure is built on that question, and the modern adaptation has inherited the categories without inheriting the question. Which is the general risk in borrowing a classical taxonomy for contemporary use — the list transfers easily and the relations between list items do not transfer at all, because they were never printed as relations. They were printed as a table.
One handshake rather than two. The two-failure-modes distinction — actively spoiling against inertly present — has genuine purchase in editing and design, but the vault's craft neighbours on restraint describe cutting what is weak rather than distinguishing harm from inertness, and forcing the pair would restate the page's own plain language rather than sharpen it.
Sharpest implication. Compatibility here is a property of the host, not of the pair — printed as a symmetric-looking table that disagrees with itself in roughly fifteen of sixty-six places, with one example explained and the rest left to the reader. Every primary rasa refuses anger and anger refuses only one of them, which makes raudra the most permissive host and the least welcome guest in the system, and adbhuta its exact opposite.
Generative questions.
mādhurya — and the rest are flat. How many of the flat entries are compressed?