Śrī Śaṅkuka rejects Lollaṭa's view outright: 'this is not so.' Then he piles up his reasons. If no vibhāvas are present, the sthāyī has no sign, so it cannot be known at all. Lollaṭa's view would make the bhāvas nameable even before they arise. In the fully developed state, defining rasa apart from the sthāyī would be pointless. Grade the sthāyī by intensity and you get an endless ladder — mild, milder, mildest indifference, on and on. The comic rasa would lose its recognized sixfold division; the ten stages of love would splinter into countless rasas and bhāvas. And intensity is not the mark anyway: grief is sharp at first and softens with time; anger, energy, and love fade once their supports — provocation, steadiness, attendance — fall away. In short, we see these moods decrease, not merely swell.

Now Śaṅkuka states his own view. From the signs — vibhāvas as causes, anubhāvas as effects, vyabhicāris as accompaniments — the audience infers a sthāyī seated in the actor. These signs are artificial, worked up by the actor's training, yet the audience never registers them as artificial. And the sthāyī inferred this way is not the real thing: it is a copy, an imitation of the sthāyī that lives in the actual Rāma. So rasa is an inferred, imitated mood — anumiti plus anukaraṇa.

And here is why it gets a new name. Because the inferred sthāyī is only an imitation, not the real sthāyī, it is called by a different name — rasa. The change of name tracks the change from real mood to imitated mood.

How is each sign grasped? The vibhāvas come from the poetry — the words give you the setting, the beloved, and the rest. The anubhāvas come from the actor's training — the practised gestures and expressions.

And the vyabhicāris, the passing feelings, are grasped through the anubhāvas the actor produces — outward signs that are artificial, got up on purpose, yet genuinely his own to display.

The sthāyī is different from the other three: it cannot be supplied even by the poetry. Take words like 'love' or 'grief.' As names, all they do is name love or grief — they make it a thing spoken of. They do not display it.

What such words do not do is make the sthāyī known the way verbal acting (vācika-abhinaya) does. Naming a mood is one thing; enacting it through speech is another.

A quick clarification of the term. 'Vācika' does not mean the words themselves. It means what is effected through the words — the enactment carried by speech. Compare 'āṅgika,' bodily acting: that too is not the limbs themselves, but what the limbs bring off. So verbal acting is the acting done through speech, not the speech.

He gives an illustration. Picture grief as an ocean — grown huge, bottomless, impossible to cross, immense. (The verse's point lands in the next line.)

The point: just as the submarine fire drinks up the whole ocean, so anger drinks up even this immense grief. Here grief is being talked about, described in words — it is named, not acted out.

A second example, from the Tāpasa-vatsarāja play. Grief has struck the man into a frozen stupor, and there he stands. Around him the ministers wail without stop. (The sentence completes in the next line.)

The ministers, terrified that his heart will burst from the pent-up grief, actually beg him to cry. This speech is Vinītadeva's in the Tāpasa-vatsarāja. Śaṅkuka's point: in verses like these, grief is not acted out — it is merely stated, put into words.

Now a contrasting verse, Ratnāvalī 2.11, in Udayana's voice. As she paints, a fine spray of her tears has fallen — and it shows on him. (Completed in the next line.)

The teardrops feel to him like sweat breaking out where her palm touched him. Now compare this with the grief verses. This sentence states its plain meaning — tears, touch, a thrill like sweat. But at the same time it enacts Udayana's rati, the joyful love that is his sthāyī: the mood is displayed, not just declared. That is Śaṅkuka's key point: abhinaya is a power of making a thing known, and it is a different power from a word's denotation. Naming and enacting are two separate things.

This is why Bharata's rasa-sūtra never mentions the word 'sthāyī' — not even in some oblique case-form. On Śaṅkuka's reading the omission is deliberate: rasa is the inferred, imitated mood, not the sthāyī itself, so the sūtra rightly leaves it out.

The upshot: love (rati) imitated is the rasa śṛṅgāra. And this settles an old tension. People say rasa 'is' the sthāyī, and also that rasa 'arises from' the sthāyī. On the imitation view both are right: the imitated love both has the nature of love and springs from love. No contradiction.

To back this, Śaṅkuka quotes Dharmakīrti (Pramāṇavārtika). Even a real, working result can come from a mistaken cognition. Take his example: two people each see a glow — one from a real gem, one from a lamp — and both run up thinking 'a gem!' (The point lands in the next line.)

The verse finishes: though both run with the very same mistaken 'gem' cognition, the outcomes differ — one actually reaches a gem, the other a mere lamp. (This is Pramāṇavārtika 2.77.) Now Śaṅkuka turns to the theatre. The audience's cognition is not 'the actor by himself is happy' — that flat, literal reading is ruled out. He will now rule out three more.

Śaṅkuka now pins down the audience's cognition by ruling out the ordinary options. It is not the true judgment 'this man actually is Rāma.' Nor is it the false judgment 'this man is not really happy.'

Nor is it the doubtful 'this might be Rāma, might not be.' Nor is it the comparison 'this man merely resembles Rāma.' So all four ordinary cognitions — true, false, doubtful, and similar — are off the table.

What is left is a fifth kind of cognition, all its own. Unlike the true, false, doubtful, and similarity cognitions, and working like the way we take in a painted horse, the audience simply has the cognition: 'this is that happy Rāma.' That special cognition is what grasps the actor as the joyful Rāma.

Śaṅkuka clinches it with a verse of Dharmakīrti's. The painted horse just appears to us. That appearing is not doubt, not a true judgment, and not an error. (The verse runs on, but this first line already names the special cognition.)

The cognition that dawns is 'this is that Rāma.' But it is not the cognition 'this is literally the one and only Rāma.' Śaṅkuka's point: the spectator does feel the actor to be Rāma, yet stops short of asserting strict identity. It is the painted-horse cognition (citra-turaga-nyāya) — 'that is a horse' about a picture, without claiming it is a real horse.

The two opposed judgments — 'he is Rāma' and 'he is not Rāma' — cannot lock together into one settled cognition. So the experience remains a running-together of both that the mind never pulls apart. That very un-sortedness is what makes it a third kind of awareness, not doubt, not truth, not error.

By what reasoning could you ever interrogate an experience that flashes up this vividly? The verse's punchline: the actor-as-Rāma cognition is self-evident and immediate, so no logic can arraign it. Its being felt is its own warrant.

This position too, says the Upādhyāya, is empty at its centre and cannot survive a hard examination. Here the refutation of Śaṅkuka's inference-theory begins, and Abhinava frames it as his teacher's verdict.

The Upādhyāya splits the thesis into three. When you say 'rasa is imitation,' from whose standpoint do you mean it? From the spectator's cognition, from the actor's intention, or from the analysing critic who sorts out how things really are? He quotes the Pramāṇavārtika's own gloss — 'the commentators do analyse it this way' — to grant that a critic's-eye reading is a real option. Each of the three will now be taken up and broken.

Take the first horn — that 'imitation' is meant from the spectator's cognition. Judged by what Bharata actually says, this reading does not hold together. He now shows why.

Here is the governing rule. You can call something 'an imitation of that' only if 'that' — the original — has already been apprehended by some means of knowledge. No prior acquaintance with the original, no recognizing anything as its imitation.

Take the illustration. On stage a man drinks milk while it is announced 'see, he drinks liquor,' and we plainly watch the milk-drinking read as a mimicry of liquor-drinking. The mapping: the perceived act is drinking milk; the original imitated is liquor-drinking; and it registers as imitation only because we have all perceived real liquor-drinking before. It works precisely because the original is already known — which is what the actor's rati will fail to satisfy.

But now in the drama, what exactly, perceived in the actor, is it that shows up as an imitation of rati? That is the question to think through. In the milk case we could point to the perceived milk-drinking; here there is no obvious perceived item to point at.

Look at what is actually perceived in the actor: his body, the mask on it, the horripilation, the choked voice, the flinging arms, the turning, the eyebrow-play, the sidelong glances. None of this is the mental state that rati is. And only a mental state of rati could ever show up to anyone as an imitation of rati. So the perceived items are the wrong kind of thing to be that imitation.

Three reasons the body and gestures cannot pass as an imitation of rati: they are insentient matter; they are caught by a different faculty — the eye, whereas rati is felt within; and they sit in a different seat altogether. On every count they are utterly unlike rati. Something so wholly unlike it cannot register as its mimicry.

Restating the rule: an imitation appears as an imitation only when the original has itself been seen. Without prior sight of the real thing, nothing can show up to you as its copy.

And nobody has ever previously perceived the rati that lived in Rāma's own heart. It was his private inner state, long past and unseen. So there is no known original against which the actor's display could read as an imitation — the first horn collapses.

The same reasoning also throws out the wider claim that 'the actor imitates Rāma himself.' Since Rāma was never perceived, the actor cannot be recognized as imitating him either. So both forms of the first horn fall together.

Now try the other tack. Suppose it is the actor's own mental state that gets cognized, and that this is the 'imitation of rati' we name śṛṅgāra. Even then a question bites: cognized as having what nature? That is what must be examined next.

The objector answers s37's question. In ordinary life we read a mental state off three kinds of mark: its cause (a lovely woman), its effect (sidelong glances), and its accompaniments (contentment and the like). By exactly these marks, the actor's mental state shows up wearing the shape of that everyday rati. So it is cognized as having the nature of worldly love.

But then the objection defeats itself. If the actor's mental state is cognized in the very shape of rati, then it is taken to be rati — and calling it a mere 'imitation of rati' falls right away. Grasp it as rati, and it is no imitation of rati.

The objector tries again: the vibhāvas are real in Rāma, the one imitated, but not real in the actor, the imitator — and just that difference is what makes the stage-version an imitation. The Upādhyāya does not fight this; he grants it: fine, let it stand. The real difficulty he will press in the next line.

But press it. These vibhāvas are not the genuine cause, effect, or accompaniment of any real rati; they are artificial, got up by poetry and training. So ask the spectators: do you take them as artificial, or not? Take the first fork — if you do see them as fakes, then how could they ever yield knowledge of rati? A mark known to be counterfeit proves nothing. (The other fork, 'not grasped as artificial', is picked up in what follows.)

The objector counters: that is exactly the point. Because the display is an imitation of rati, it produces a naive, unanalytic cognition in the simple spectator. The artificiality is not a defect but the very engine of the ordinary viewer's response.

Reply with the rule of inference. From an effect you may legitimately infer 'some other thing' only in two situations: when the effect genuinely does arise from a different cause, or when a well-trained observer recognizes it as such. In those cases inferring the unusual cause is sound.

But the untrained person — the naive spectator of the objection — infers only the ordinary, well-known cause. He does not reach for 'a different thing.' So seeing the staged effects, a simple viewer would infer their familiar everyday cause, not any sophisticated 'imitation of rati.' The objector's mugdha-buddhi cuts the wrong way for him.

Take an example. When an untrained person sees a certain kind of scorpion, all he infers is cow-dung — because he has learnt that scorpions come from cow-dung. He reads the mark straight back to its one familiar cause. He does not infer some 'imitation' of anything.

Or else, beyond that, one may get a plain false cognition of a scorpion itself — as when a painted scorpion makes you think 'a scorpion!' Notice what you get: either the real cause (cow-dung), or a false grasp of the real thing (scorpion). What you never get is a cognition whose content is 'an imitation of a scorpion.'

Push it further: even when the mark itself is misperceived, you still do not infer a 'semblance of X.' A false mark makes you infer the real X (falsely), not something labelled 'a look-alike of X.' So mis-cognition of the mark does not open a door to 'imitation' either.

Here is the proof. When you mistake mere vapour for smoke, you go on to infer fire — real fire, falsely. You never infer 'an imitation of fire.' The false mark still points at the real thing, not at some copy of it.

Now grant Śaṅkuka his best case: suppose the mark itself already appears 'as an imitation of X.' Even then you cannot infer 'an imitation of X.' Appearing-as-an-imitation in the mark does not license inferring-an-imitation as the conclusion.

Here is the parallel. Suppose you take a bank of fog to be 'an imitation of smoke.' Does that ever make you go on and cognize 'a fire-imitating red hibiscus'? Never. Smoke-taken-as-imitation does not breed fire-as-imitation. Just so, imitation-vibhāvas cannot breed an inferred imitation-rati.

One may object: even when the actor feels no anger, he looks angry. Granted, says the upādhyāya. But all that means is that he resembles an angry man. 'Resembling' is not 'imitating' — the objector has quietly slid from mere likeness to imitation, and that slide is exactly what is denied.

And the likeness runs only through outward features — the knitted brow and so on. It is like the way a gavaya, the wild ox, resembles a cow through its face and build. The gavaya does not 'imitate' a cow; it merely looks like one. Likeness by such features, and nothing more, yields no imitation at all.

And even the resemblance-judgement is missing on the audience's side: they do not sit there thinking 'this man is like Rāma.' Nor do they grasp the dancer as an emotionless mimic — their apprehension of him is not drained of the emotion itself. So neither 'imitation' nor even 'resemblance' captures what the spectators actually experience.

So the upshot: to say 'there appears an imitation of the emotion' is just empty words. Nothing in experience answers to it — not an inference of imitation, not a look-alike, not even a resemblance-thought in the audience. The phrase is a label with nothing under it.

Now take Śaṅkuka's other claim — that a special cognition 'this is Rāma' occurs. Force it into the dilemma. It is a definite cognition. Then: if no later cognition ever cancels it, why isn't it plain true knowledge? And if a canceller does come, why isn't it plain false knowledge? There is no middle 'special' status. Worse, since the actor is not in fact Rāma, the cognition is false anyway — canceller or none. Śaṅkuka's third, in-between kind of cognition collapses back into true or false.

So Śaṅkuka's supporting line — 'contradictory cognitions do not merge into one' — falls with the cognition it was propping up. And here is a further point: put a different actor on stage and you get exactly the same 'this is Rāma.' The awareness is not fixed to any one performer.

The consequence is absurd. If any and every actor equally yields 'this is Rāma,' then 'being Rāma' has become a generic universal spread over them all — not the one unique person, Rāma. Śaṅkuka's cognition, taken seriously, dissolves Rāma into a mere class-feature.

Next, Śaṅkuka had said the vibhāvas are supplied to the mind by piecing them together out of the poem. That too, says the upādhyāya, we do not follow. Who does the connecting, and to whom does the vibhāva get connected? The next atoms press the question.

Take the actor first. He never thinks 'this Sītā is my own beloved.' For him the beloved-vibhāva is simply not connected up to anyone as his. So the 'connecting from the poem' cannot be happening in the actor.

So retreat to a thinner sense: 'connecting up' just means the vibhāvas are made presentable for the audience to catch. Very well — but then the durable emotion itself is 'connected up' even more plainly, since it is what the audience catches most of all. And that wrecks Śaṅkuka's own thesis, for he had insisted the sthāyī-bhāva is precisely NOT connected up from the poem.

And it is that very emotion the audience takes in first of all — 'in this actor, this feeling' — as the primary thing. The Nāṭyaśāstra passage beginning 'vāg-vācikam' already sets out a serious analysis of what acting really is, eager to mark off its divisions; Abhinava says that full treatment belongs later, in chapter 14, and he defers it to there.

So the first reading is dead: taken on the side of what the audience cognizes, 'rasa is an imitation of the durable emotion' does not hold up. The upādhyāya now turns to the second reading — that the imitation is on the actor's side.

Now the actor's side. He never actually thinks 'I am imitating Rāma' or 'I am imitating Rāma's state of mind.' The self-aware act of imitating, which the theory needs, is simply not there in his mind.

For what is imitating? It is making yourself resemble a model. But you cannot make yourself resemble a model you have never once perceived — and no actor has ever laid eyes on Rāma. So 'the actor imitates Rāma' is impossible on the very meaning of imitation.

Try loosening the definition: say imitation just means 'doing something afterwards.' Then everything in daily life that comes after something else would count as 'imitating' it — the definition swallows the whole world. Over-broad, and so useless.

One more fallback: don't tie the imitation to any particular man at all. Say instead, 'I imitate the grief of a noble character, whoever he may be.' Here prakṛti no longer means a fixed model but a type of person. The upādhyāya takes this up — and, in the atoms that follow, presses 'imitate it by what, exactly?'

So take the opponent's last move — 'I imitate the grief of some noble character.' Then by what does he imitate it? That is the question. Not by grief itself: the actor has no grief in him to imitate with.

Nor can he imitate grief by shedding tears and the rest. Tears are a bodily effect; grief is an inner state — the two are of wholly different kinds, so one cannot be the imitation of the other. This was already argued earlier.

Here the opponent falls back to his smallest claim: 'I do not imitate grief itself, nor imitate it with tears — I imitate just the visible effects of grief (śoka-anubhāvas) as a noble character shows them.'

But even that will not do. Whose grief-effects, exactly? If you answer 'anybody's, no one in particular,' then this featureless nobody cannot be held before the imagination at all. You cannot imitate a model you cannot even picture. An imitation demands a definite original.

Suppose instead he says, 'I imitate the one who weeps like this' — meaning himself. Then the actor becomes his own model, and the theory needs two poles: an imitated original and an imitator. If they collapse into one person, the whole imitation-relation dissolves. There is nothing left to be the imitation of.

And here is the positive account. Trained in his craft, the actor draws on emotional triggers of his own; the mental state is made common — lifted out of any one person's private grief; and it lands because the audience's heart answers to it. On that basis he simply displays the visible signs and recites the poem, dressed up with heightened intonation and the like, and acts. What we actually take in goes exactly this far — and it never once registers as 'imitation.'

At most the actor is copying the outward get-up — the lover's costume and bearing. That is not an imitation of Rāma's actual conduct. Dressing the part is not reproducing the man.

And we made this same point earlier, back in the first chapter.

Nor does the imitation-thesis hold if we take it by the analyst's third route — 'how things really stand.' Even measured against the plain facts, rasa is not the imitation of the sthāyī.

Rasa is something inwardly savoured, felt from within. That is exactly not the kind of thing that counts as an external, objective fact. So 'the facts of the matter' cannot make it an imitation either.

What the real state of affairs is, we will show in due course. And note this: nowhere does Bharata himself ever say 'rasa is the imitation of the sthāyī.' The doctrine is not even in the text it claims to expound.

There is not even an inferential hint in Bharata that would point toward the imitation-reading. Quite the reverse: the way he lays out the dhruvā songs, the variety of rhythm, and the limbs of the lāsya-dance actually points the other way — against imitation. We will work this out fully at the end of the sandhyaṅga chapter.

The opponent's one textual foothold is Bharata's phrase 'imitation of the seven continents' (NŚ 1.117) — proof, he thinks, that Bharata uses 'imitation.' But that word can be read another way there. And even if it does mean imitation, notice: nowhere does plain imitation — of a lover's costume, gait, or gestures — get a brand-new name like 'rasa.' So the term proves nothing.

The opponent presses a further analogy — the painted animal (his earlier painted-horse maxim). Only when the surface is worked over with pigments like orpiment do we call it 'a bull.' Just so, he says, only with the actor's added colouring does the sthāyī show up as rasa.

Now, if by that he means the pigments manifest a bull — reveal one that is already there — then he is wrong.

A lamp manifests: it lights up a real thing already sitting there. Vermilion and the other pigments do nothing of the sort — there is no real bull hidden in the surface for them to reveal. So 'manifest' is the wrong word.

What actually happens is this: the pigments produce a fresh arrangement that merely looks like a bull. They fashion a resemblance; they do not uncover a real animal.

Here is why the analogy fails. The pigments, laid out in the shape of a bull's limbs, are what the awareness 'this looks like a bull' actually fastens on. But nothing parallel happens with the vibhāvas: no awareness of the form 'this looks like love (rati)' ever settles on the assembled vibhāvas the way 'looks like a bull' settles on the paints. So the painted-bull comparison collapses.

So the verdict is settled: 'rasa is the imitation of the bhāvas' will not stand — Śaṅkuka's thesis fails. Now Abhinava turns to a quite different proposal made by someone else. On this view the outer assemblage of objects (viṣaya-sāmagrī) has the power to produce pleasure and pain; being external, and taken through Sāṅkhya eyes, it is itself of the nature of pleasure and pain — and that, the proponent says, is rasa.

The proponent fills in the recipe. Within that assemblage the vibhāvas are the raw base — the 'petals' or ingredients — while the anubhāvas and the passing vyabhicāri-feelings are the refiners that work the mixture up into its finished state.

The sthāyī-bhāvas, on this view, are what that outer recipe produces: they are the inner side, pleasure- and pain-natured. So the outer sāmagrī is rasa, and the sthāyīs are its inner yield.

This proponent is treated gently. In reading Bharata's line 'we shall bring the sthāyī-bhāvas to rasa-hood,' he grants that the equation is only figurative (upacāra), and he sees for himself that a literal reading would clash with the text. Because he concedes all this on his own, he spares judicious readers a noisy parade of refutations. So there is little left to say against him.

And the leftover objections — that this view forces on you an inconsistency of cognition, the sort of thing haggled over in other schools, and so on — how much of that need we say here? He waves them off: the main thesis is already broken, so these peripheral charges do not need answering one by one.