Someone spends a chapter constructing seven rooms. Full specification each — what goes in them, what stimulates them, what the body does, which transitory states pass through.
Then, on the last page of that chapter, a notice goes up: these seven are not rooms. For practical purposes they are furniture belonging to the five rooms next door.
And then three more chapters arrive, and they are about how the seven rooms connect to each other, which of them may be opened onto which, and how to tell a real one from a fake.
That is the actual sequence, and it is not the sequence this build has been carrying.
The PRD and the SIGNAL both place the retraction at Text 27, described as the close of the rasa material — five Texts of apparatus, then the withdrawal.
That is wrong, and the source is unambiguous. The retraction is stated twice, and both statements are inside Text 24.
The first is mid-Text, in the bhayānaka section, reported by Jīva Gosvāmī:
Śrī Viśvanātha Cakravartī's own opinion, which is in accordance with Śrīmad-Bhāgavatam, is that
hāsyaand the other secondary rasas are, for practical purposes,vyabhicāri-bhāvas. The purport is that Śrī Kṛṣṇa is theviṣayaand the devotee is theāśrayaof the five primary forms ofrati. This is the general rule accepted everywhere.1
The second closes Text 24 entirely:
In Bhakti-rasāmṛta-sindhu (4.7.13–14), Śrīla Rūpa Gosvāmī concludes the discussion of
gauṇa-rasaby saying thathāsyaand the other secondary emotions have been accepted as rasa in accordance with the opinion of past authorities like Bharata Muni. This point should be understood by insightful persons. The five primary rasas are factually accepted ashari-bhakti-rasa.Hāsyaand the other secondary emotions are, for practical purposes,vyabhicāri-bhāvasof these five rasas.2
Texts 25, 26 and 27 come after this. What sits at Text 27 is rasābhāsa, and Appendix 2's own concordance confirms the split: Text 24 → Sindhu 4.7.13–14; Text 27 → 4.8.85 and 4.9.1.
The correction matters because it changes the shape of the finding. Not a system that dismantles its own second half at the end. A system that dismantles its second half at the halfway point and then spends three more Texts operating it.
And a second correction: this is Rūpa's conclusion. The page's own slug says Viśvanātha's dissent, and that is right about the dissent and needs qualifying about whose it is.
The reclassification is attributed, in the second passage, to Rūpa Gosvāmī, concluding his own gauṇa-rasa discussion at Sindhu 4.7.13–14. Viśvanātha's own opinion is reported to agree with it.
So this is not the summariser overruling his source. Both hold the same position, and the dissent is from Bharata Muni.
Which sharpens rather than softens the finding, because it means the seven-rasa scheme was carrying a disclaimer from the beginning of the tradition's own foundational text, and the Bindu's rasa half — the half that, per this build's census, contains not one verse of Rūpa's as text — reproduces it faithfully as quotation.
Both statements name their warrants, and they are not the same kind of thing.
The seven are rasas in accordance with the opinion of past authorities like Bharata Muni — a claim from precedent.
They are really transitory states in accordance with Śrīmad-Bhāgavatam — a claim from scripture.
So the structure is: an inherited authority establishes the categories, a scriptural authority demotes them, and the text prints both without adjudicating between the kinds of warrant.
And the demotion has a hedge that does almost all the work of keeping the system usable: for practical purposes. Not are. Not were never. For practical purposes.
Read the hedge as load-bearing and the three following Texts stop being strange.
Text 25 gives a full compatibility matrix over all twelve rasas — every secondary listed with its compatible, incompatible and neutral partners, and the aṅgī/aṅga machinery for how a secondary may rise to prominence.
Text 26 assigns each secondary a home inside a primary.
Text 27 grades semblances across all twelve, with anurasa defined specifically as a defect of hāsya or any of the seven gauṇa-rasas and śānta-rasa.
None of that is inconsistent with the retraction, because the retraction never abolished the seven. It ranked them, and then left them in service. The book's position, stated whole, is: these are not independent rasas, they are states of the five; you will nonetheless need to work with them as rasas, so here is how.
That is a coherent editorial position and the book never states it as one. It states the demotion, and then behaves as though it had not.
Text 26 is the retraction, executed. The clearest evidence that Texts 25–27 are the retraction's continuation rather than its contradiction is Text 26, which is six lines long and is usually skipped.
The seven
gauṇa-rasasmay be included within the fivemukhya-rasas. For example,hāsyaandyuddha-vīramay be included insakhya.Adbhutamay be included in all five.Dāna-vīraanddayā-vīramay be included invātsalya.Bhayānakamay be included invātsalyaanddāsya, andbībhatsawithinśānta.Krodha-ratiofraudra-rasais included within one aspect ofvātsalya- andmādhurya-rasa.3
That is not a new topic. That is they are vyabhicāri-bhāvas of these five, turned into a table. Each demoted category is given a forwarding address.
And the table is doing what the Text 24 close only asserted: it says which of the five each secondary belongs to. The retraction claimed the seven were states of the five; Text 26 assigns them.
Now the wrinkle, and it is checkable in two lines.
Viśvanātha's Sanskrit root of Text 26 reads: karuṇā-dāna-vīra-dayā-vīrāṇāṁ vātsalye — karuṇā, dāna-vīra and dayā-vīra, in vātsalya. Three items in the compound.4
The English renders it: Dāna-vīra and dayā-vīra may be included in vātsalya.
Karuṇā is dropped.
And the omission is detectable precisely because the table is otherwise a complete assignment. With karuṇā restored, all seven secondary rasas have a home: hāsya → sakhya; adbhuta → all five; karuṇā → vātsalya; vīra → sakhya and vātsalya by sub-type; bhayānaka → vātsalya and dāsya; bībhatsa → śānta; raudra → vātsalya and mādhurya. Without it, the list covers six of seven and the reader is left to wonder where compassion went.
Stated at the strength it carries: one compound, one omission, in the range read. [n=1 OBSERVED]. It joins the sixty-fifth limb — where the edition silently merged two limbs to make the arithmetic close — and the three-versus-four mādhuryas of Text 22, where the commentary supplies a fourth the root does not have. Three edition-level discrepancies now, all small, all in the same direction: the printed English tidies. [n≥3 PATTERN] on the edition's habit, which is a claim about this printing and not about Viśvanātha.
(One genuine incompleteness survives the repair and is not a translation slip: dharma-vīra is assigned no home in either the root or the English. Three of the four heroisms are placed; the fourth is not.)
The retraction reads as a considered dissent arriving at the end of a long construction. It is not, and this build established why before reaching it.
Text 17's definition of parārtha-rati, quoted back verbatim inside Text 24 itself:
When
rati, instead of nourishing its ownsthāyibhāva, recedes into the background and nourishes one of the seven secondary emotions, it is calledparārthā, nourishment-giving… Only whenparārtha-ratiin one of the five primary dominant emotions recedes into the background and nourishes the seven secondary emotions do those secondary emotions attain to the status ofgauṇa-rati.5
Only when. The secondary rasas have no standing of their own at any point in the system. Their rasa-status is conferred, on loan, by a primary that has stepped back to lend it.
So the demotion is an entailment of the definition, not a change of mind — and the definition was given ten Texts before the seven were built.
Which is the finding this build carried into Cluster J and the source confirms it in the strongest possible form: the retraction is quoted from the same page as the definition that guarantees it. Text 24 back-quotes Text 17 and then, twenty pages later, states the conclusion Text 17 had already made unavoidable.
You are trying to work out whether to use the twelve-rasa scheme or the five.
The book's answer, assembled from the pieces, is: use twelve to describe and five to explain. The seven are real as descriptions of what happens in a devotee — the taxonomy of laughter, the anatomy of anger, the three objects of grief are all good instruments. They are not real as independent relationships with Kṛṣṇa, and any claim that rests on their independence is resting on something the tradition withdrew.
Second scene. Someone tells you their rasa is vīra, or adbhuta, or karuṇa.
Text 26 gives you the correct response and it is not a correction. It is a question: which of the five is it a state of? Heroism in fighting belongs to friendship; heroism in giving and in compassion belongs to parental love; astonishment belongs to all five and therefore tells you nothing on its own.
Third, and this is the one that matters for reading the literature. When you meet a passage arguing at length about a secondary rasa — its subdivisions, its exemplars, its compatibilities — remember that the argument is being conducted inside a hedge. For practical purposes. The apparatus is being operated, not asserted, and the tradition told you so on the last page of Text 24.
Put the retraction beside the enumeration and something falls out that neither shows alone.
The seven secondary rasas are Bharata's eight minus śṛṅgāra — the complete inherited set, with the one item the tradition renamed and promoted removed from it.
The retraction removes exactly that set.
So the operation, stated in one sentence: everything this tradition took from the older aesthetics is granted rasa-status on the older aesthetics' authority, and then demoted on scriptural grounds; everything the tradition added or renamed is retained as primary.
There is a charitable reading and it is probably true. The five that survive are the five that are relationships with a person — peace, service, friendship, parenthood, love — and the seven that go are reactions to events. That is a real distinction, it is stated in the material (Kṛṣṇa is the viṣaya and the devotee is the āśraya of the five primary forms of rati; this is the general rule accepted everywhere), and it would sort the twelve the same way even if nobody knew where any of them came from.
The uncharitable reading is that a lineage kept its own contributions and demoted its inheritance.
Both readings predict the identical partition of the twelve, and this build has found no evidence that separates them. That is worth stating as the result rather than resolving. The relationship criterion is available, it is stated, and it is also exactly coextensive with provenance — which means the tradition never had to choose between the two justifications and never did.
[n=1 OBSERVED] on the reading. The enumeration itself is complete and checkable and needs no hedge.
Strongest support. Both statements of the retraction are quoted in full, one attributed to Viśvanātha through Jīva and one to Rūpa's own Sindhu 4.7.13–14. Both sit inside Text 24. Appendix 2's concordance independently assigns 4.7.13–14 to Text 24 and 4.9.1 to Text 27. The parārtha-rati definition is back-quoted inside Text 24 itself.
Tension one. The retraction is at the close of Text 24 and three further Texts operate the retracted categories in full. The hedge for practical purposes makes this coherent and the book never says so.
Tension two. The seven are demoted on scriptural grounds after being established on precedential grounds, and no criterion is offered for how the two kinds of warrant relate.
Tension three. The partition the retraction produces is exactly the provenance partition, and the stated criterion — relationships versus reactions — predicts the same result. Nothing in the range read separates them.
Open question. Dharma-vīra is assigned no home in Text 26, in root or translation. Whether religious heroism was thought to belong everywhere, nowhere, or simply got lost is not answerable from here.
The vault's Pollock-derived account presents the Gauḍīya move as the twelfth-rasa revolution: bhakti-rasa deployed as a meta-category, the canonical nine reinterpreted as modalities of devotion, the framework expanded rather than displaced.
That account has the retraction nowhere in it, and the omission is not an error — it is a difference in what the two sources are for. A historian of aesthetics records the framework a tradition published. The manual records what the tradition then did with it.
And what it did was publish the expansion and the withdrawal in the same chapter.
The convergence with this build's own earlier findings is close to total. The census: a book named for Rūpa's ocean whose rasa half carries none of his verses as text — and which nonetheless quotes him at the one point where the position is withdrawn. The ālambana dispute: Bharata credited by name and overruled in a subordinate clause, in the same paragraph as the first statement of this retraction. The sixty-fifth limb: an edition tidying an arithmetic that would not close.
Every one of those is the same move at a different scale, and this is the largest instance: an entire inherited taxonomy invoked, formally honoured, given three chapters of loving apparatus, and demoted in a clause containing the words for practical purposes.
A change that passes because everything around it looks unchanged. Set this against The Seamless Blend, where Greene's account of the Cambridge spies is three clauses long and mentions no tradecraft: they went undetected for years because they were apparently decent chaps, had gone to all the right schools, and fit the old-boy network perfectly. Not careful. Familiar.
The seven gauṇa-rasas keep every marker of belonging. Bharata's names, Bharata's count, Bharata's sthāyibhāvas, a full five-component apparatus each, a compatibility matrix, a semblance-taxonomy. Nothing on the surface signals that anything has been taken from them. The demotion is a single clause inside a chapter that otherwise looks exactly like the tradition it came from — and it survives, in the sense of being widely unnoticed, for the same reason the spies did.
The instructive difference is intent, and it is worth being fair about. Greene's blend is deliberate camouflage. There is no evidence at all that this is: Rūpa states the position plainly, Viśvanātha reproduces it, and the edition prints it in full. Nobody is hiding anything.
Which is what makes the pair useful, because the effect does not require the intent. Familiarity buys non-inspection whether or not anyone bought it on purpose. A qualification embedded in a structure that otherwise looks entirely conventional will be read past — by summarisers, by students, and by a vault build that carried the wrong Text number for two clusters until it read the passage. That last is not a hypothetical; it is what happened here, and it is the strongest available demonstration of the mechanism.
And deference performed rather than earned, at the level of a whole framework. Set this against Authority Dynamics and the Construction of Deference, whose central claim is that earned and constructed authority look identical from outside — a person defers in both cases — and that the difference is internal: earned authority must keep delivering, constructed authority need only maintain the performance.
Bharata's authority in this book is honoured in exactly the way the page describes as performance. He is named, credited, cited as past authorities and previous authorities on nāṭya-śāstra, and his seven-rasa scheme is given the full dignity of a chapter. What he is not given is the last word on any contested point: not on ālambana, not on whether the wicked can be a support, and not on whether his seven are rasas at all.
What the pair produces that neither yields alone. The vault page frames the earned/constructed distinction as a fact about a person's competence. This case shows the same structure operating on an inherited body of knowledge, where the question is not whether Bharata was competent but whether his authority is doing evidential work or ceremonial work. And here it is separable, because the text tells you: the seven are rasas on his authority and transitory states on scriptural grounds. Precedent supplies the categories; scripture supplies the verdicts. That is a division of labour, and it is the honest form of what looks from outside like inconsistency — a tradition can keep an inheritance in full working order while reserving every decision that matters, and it will look like reverence the whole time.
Sharpest implication. The retraction is at the close of Text 24, not Text 27 — and it is Rūpa's own conclusion, which Viśvanātha endorses. Three further Texts then give the demoted categories a compatibility matrix, a forwarding table and a counterfeit-grading, which is coherent only because the demotion is hedged for practical purposes. The set removed is exactly the inherited set, and the stated criterion predicts the same partition as provenance does.
Generative questions.
ālambana