Item fifteen of the thirty-three transient states is mṛti, and the word means death.
The definition is unambiguous: giving up the life air due to disease, despondency, fear, physical debility or an assault. The symptoms are indistinct speech, change of colour, shallow breathing and hiccups.
Then Rūpa, cited by verse:
In Bhakti-rasāmṛta-sindhu 2.4.102, Śrīla Rūpa Gosvāmī explains that the
vyabhicāri-bhāvaknown asmṛtidoes not refer to actual death. Because the symptoms resemble the condition of the heart just prior to death, it is calledmṛti.1
An item in a devotional taxonomy is called death, defined as dying, and then explicitly disclaimed.
Which raises a question worth more than the disclaimer: why keep the word?
Mṛti is not the only borrowing that needs a note.
Vyādhi — disease. Defined as a feverish condition produced by separation or due to an excess in the humours of the body. And then the correction: in actuality it is the bhāvas, or spiritual transformations of the heart, arising from separation and severe distress, that are called vyādhi.2
Apasmṛti — the word means epilepsy. Defined as bewilderment of the heart that occurs due to an imbalance of the elements of the body, with symptoms including falling on the ground, foaming from the mouth, throwing the hands and legs into the air.
Unmāda — madness. Bewilderment of the heart arising from extreme bliss, calamity or acute separation, with loud laughter, dancing, singing, futile action, incoherent speech, running, shouting.
Moha — fainting. Glāni — debility from the waning of oja, the body's vital principle.
So the list of thirty-three transient devotional states includes death, disease, epilepsy, madness, fainting and debility — and the tradition has to explain, at least twice, that it does not mean the medical thing.
The obvious answer is that the vocabulary was inherited, and it is right as far as it goes: the thirty-three are nāṭya-śāstra's, and a dramaturgy describing what happens on stage needs terms for a character dying, sickening, going mad.
But the borrowing was not forced. The tradition could have renamed them on transposition, as it renamed a good deal else.
It kept them, and the reason is visible in the definitions. The pathological word is doing descriptive work no devotional coinage would do. Bewilderment of the heart arising from acute separation, with loud laughter, running and incoherent speech — there is no polite word for that. The tradition wanted a term that carried the full disorder of the state, and the medical vocabulary is the only place such terms exist.
So the borrowing buys precision and incurs a debt, and the disclaimers are the interest payment.
Notice where the disclaimers are actually needed.
Nobody needs telling that harṣa — joy — is not the medical thing, because there is no medical thing. The disclaimers cluster on exactly the terms where a bystander could mistake the devotional state for a clinical emergency, and where the mistake would have consequences.
Mṛti gets one because a person who looks to be dying is treated as dying. Vyādhi gets one because fever is a thing you treat.
Unmāda and apasmṛti — madness and epilepsy — get no disclaimer at all, and they are the two where the confusion would be most consequential and hardest to resolve. A devotee running, shouting and speaking incoherently, or falling and foaming, is describable in exactly two ways, and the tradition supplies the devotional reading without excluding the other.
Which is a real gap rather than a quibble. [n=1 OBSERVED] on the asymmetry: two disclaimers, four candidates.
Then a different kind of item entirely.
Āvega — agitation — is defined ordinarily enough: tremendous outburst of emotion and bewilderment of the heart. But it comes with an enumerated cause-list and a symptom-set for each cause:
A pleasing object → horripilation, comforting words, fickleness, standing to welcome. A displeasing object → falling on the ground, screaming, dizziness. Fire → disorderly movement, trembling, closing the eyes, shedding tears. Wind → covering the body, rapid movement, wiping the eyes. Rain → running, taking an umbrella, contracting the body. Calamity → change of facial colour, astonishment, trembling. An elephant → running, trembling, fear, looking behind oneself repeatedly. An enemy → putting on armour, taking up weapons, leaving home.3
Eight causes, and five of them are weather, animals and warfare.
Taking an umbrella is the detail that gives the whole thing away, and in the best way.
This is not devotional psychology. It is stage business — an actor's manual for how a body behaves under eight named provocations, so that a performance is legible. Rain: the character runs, hunches, puts up an umbrella. Elephant: runs, trembles, keeps looking back.
And the tradition has transposed it wholesale into a taxonomy of devotional emotion, including the umbrella.
Which is worth stating without mockery, because it is evidence rather than embarrassment. The vyabhicāri-bhāva list is the least-transposed part of the whole apparatus. Elsewhere the categories were reassigned — the āśraya moved from spectator to devotee, the viṣaya fixed on one entity. Here the inherited material came through nearly untouched, complete with props.
Take the umbrella seriously and it also changes how the list should be read.
The thirty-three are not thirty-three devotional states someone observed and catalogued. They are thirty-three dramatic states, inherited entire, whose devotional application is asserted rather than derived.
And you can see the tradition doing the applying, unevenly. Some entries are fully devotionalised — nirveda, self-disparagement, arises from great distress, feelings of separation, jealousy, non-performance of duty; viṣāda, despondency, from non-attainment of one's desired object, Śrī Kṛṣṇa, from inability to complete some endeavour begun for Kṛṣṇa, due to the appearance of some calamity that befalls Kṛṣṇa. Those are rewritten for the new setting.
Āvega's elephant is not. Śrama, fatigue, is half-way: vigorous movement in pursuit of Kṛṣṇa (like Mother Yaśodā's running to catch Kṛṣṇa) — an inherited category given a devotional example.
So the list is a stratified artefact: some entries rewritten, some exemplified, some passed through with their stage props intact. [n≥3 PATTERN] on the stratification — three distinct treatments visible across the entries read.
You encounter the list and try to locate your own states in it.
Most of it will not fit, and the reason is structural rather than a failure of introspection: these are dramatic states, sized for what is legible in a performance. Ordinary interior weather is finer-grained and less pictorial than any of the thirty-three.
What the list is genuinely good for is the opposite direction — reading the pastimes. It is a vocabulary for what the exemplars are depicted doing, and that is what it was built for.
Second scene, and it matters. Someone in a devotional setting displays what the list calls unmāda or apasmṛti — running, shouting, incoherent speech, or falling and convulsing.
The tradition supplies a devotional reading for both and disclaims neither. I would not use this list to decline medical attention, and the text gives no warrant for doing so: it disclaims mṛti and vyādhi precisely because a bystander would otherwise act, and its silence on the other two is a gap rather than a ruling.
Third. If you want one item from this list that is genuinely usable, take avahitthā — concealment of one's own condition — which is a legitimate transient state in good standing. It is the tradition's acknowledgement that hiding what is happening to you is itself one of the states, not a failure to have one.
Mṛti is defined as dying and declared not to be dying. What is left?
Not the referent — that was withdrawn. What survives is the shape: a state whose symptoms are the condition of the heart just prior to death.
So the term has been converted from a name for an event into a name for a resemblance. And that is a specific and unusual linguistic move, worth naming because this book does it more than once.
Compare nāma, which is declared not in the shape of letters and is nonetheless chanted in letters. Compare bhāva, which appears as though brought into manifestation by the mind and is not. Compare dīkṣā as bell metal transmuted to gold, in a procedure not completed in one day.
In each case a strong word is used, its literal application is denied, and the word is kept because nothing weaker would carry the weight.
Which is a defensible practice and it has a cost this tradition pays repeatedly: every such term needs a disclaimer, the disclaimers are in different Texts from the terms, and a reader who meets the word without the disclaimer gets the literal reading. The build has already logged the sharpest case — the fifth nāma-aparādha makes it an offence to treat the tradition's superlatives as exaggeration, which forecloses the deflationary reading of exactly this class of word.
So the tradition uses words beyond their literal reach and has made it an offence to read them as figures. [n=1 OBSERVED] on the combination; each half is separately documented.
Strongest support. The thirty-three are Rūpa's, quoted in Sanskrit at Text 15's root with citation to Bhakti-rasāmṛta-sindhu 2.4.4–6, and each definition is given in the Sanskrit before the commentary expands it. The mṛti disclaimer is attributed to Sindhu 2.4.102 by verse number.
Tension one. Mṛti and vyādhi are disclaimed; unmāda and apasmṛti are not, and they are the two where a devotional reading and a clinical one would conflict most consequentially.
Tension two. The vyabhicāri list is the least-transposed part of the apparatus — āvega's eight causes include fire, wind, rain, an elephant and an enemy, with an umbrella among the symptoms — while other parts of the machine were substantially reassigned. The unevenness is not remarked on.
Tension three. Vyādhi is glossed as in actuality the bhāvas arising from separation. If the state simply is the spiritual transformation, then the pathological term names nothing at all, and the disclaimer has dissolved its own entry. The taxonomy keeps the entry.
Open question. Are any of the thirty-three original to the devotional application? The list is Rūpa's and quoted as such; whether he added, dropped or renamed any against the received nāṭya-śāstra set is not recoverable from this edition, and the decisive comparison is with the Nāṭyaśāstra itself, which this build has not read.
The vault's Pollock-derived corpus holds the thirty-three vyabhicāri-bhāvas as a category of Sanskrit literary theory, read historically — which is exactly what the umbrella confirms them to be.
Where the two accounts meet is more interesting than where they differ. The literary corpus can say these are dramatic categories without embarrassment, because in a dramaturgy that is all they need to be. This book has to claim they are devotional states, and the claim is doing real work in the rasa machine — the thirty-three are what swell the sthāyibhāva.
So the transposition's cost lands here more heavily than anywhere else in the apparatus: a set of stage-directions is being asked to serve as an inventory of a devotee's interior life, and the entries that were not rewritten still have the props attached.
The tradition's own uneven treatment — some entries rewritten, some exemplified, some passed through — is the honest record of a transposition that was done thoroughly in places and not at all in others.
A devotional state and a clinical emergency described by the same words. Set this against Spiritual Emergency, whose whole subject is this boundary: episodes with the phenomenology of psychiatric crisis — disorganised speech, motor agitation, terror, collapse — occurring inside a religious or transformative process, and the question of which frame to apply and who decides.
The thirty-three are that problem four hundred years earlier and from the other side. This tradition already has unmāda, apasmṛti, moha and vyādhi in its inventory as devotional states, and already needs to say — twice — that two of them are not the medical thing.
What the pair produces that neither yields alone. The spiritual-emergency literature's core difficulty is that the frame cannot be chosen from the phenomenology, because the presentations overlap. This taxonomy demonstrates how old that difficulty is and how a tradition handles it when it has authority: it assigns the frame by context — these occur in a devotee, in a devotional setting, arising from separation from Kṛṣṇa — which is a provenance criterion, the same instrument this book uses three other times.
And it surfaces the risk both share, which is the same risk in both directions. Provenance-assignment of the frame will systematically mis-sort the cases where the context is wrong about the person — a devotee having a seizure, someone in genuine crisis inside a religious community. The tradition's silence on unmāda and apasmṛti is exactly where that mis-sorting would occur, and the modern literature's insistence on ruling out organic causes first is a direct answer to it.
Which is the usable conclusion: where two frames share a phenomenology, the one that can kill you has to be excluded first, regardless of how good the provenance argument is. That is not a modern squeamishness; it is what the tradition's own disclaimers on mṛti and vyādhi were already doing, applied consistently.
One handshake here rather than two. The word-kept-after-its-meaning-withdrawn finding is a point about this tradition's linguistic practice, carried at higher resolution in the Analytical Case Study.
Sharpest implication. The thirty-three are inherited stage-directions, applied to devotion unevenly — some rewritten, some exemplified, some passed through with an umbrella still attached. And the two entries where a devotional and a clinical reading would most consequentially collide, madness and epilepsy, are the two the tradition does not disclaim.
Generative questions.
vyādhi simply is the bhāva arising from separation, what is the entry naming?avahitthā, concealment, sits against the lowest intensity grade