You catch yourself in a window you did not expect and for about a third of a second you see a person.
Not yourself — a person. Posture, expression, the way they are standing, all of it arriving as information about somebody rather than as a mirror. And then it resolves, and it is you, and the small shock of the interval is that you had briefly been reacting to yourself the way you react to a stranger.
Almost everyone has had this. What nobody has is the version where the third of a second does not end.
Madhya 8 has that, and the person in the glass is God.
Kṛṣṇa is in Dvārakā, in a hall with jewelled pillars, and he sees his own reflection.
And he wants to embrace it.
"Alas, I have never seen such a person before. Who is He?"1
The corpus's gloss is the important part: he is attracted exactly as Śrīmatī Rādhārāṇī is.2 Not similarly, not analogously. The comparison is to her specific state, and it is the reason the episode is in this chapter rather than filed as a curiosity.
The whole Caitanya-descent argument runs on a single premise: that there is an experience of Kṛṣṇa which Kṛṣṇa cannot have, and that he wants it.3
That premise needs evidence, and evidence about a deity's inner life is not easy to come by. The pillar supplies it, and it supplies it in the only form that would work: an occasion on which he is placed, briefly and accidentally, in the position of somebody looking at him.
The reflection does the impossible thing. For as long as it lasts, he is the āśraya rather than the viṣaya — the one who holds the love rather than the one it is aimed at — and the corpus says explicitly that what he feels there is what she feels.4
So the episode is not a marvel about vanity. It is the descent argument's one piece of evidence, and everything at Ādi 4 is an elaboration of the third of a second in the pillar.
And here the page has to be honest, because the episode does not quite do what the corpus needs.
The reflection is not another person. He is attracted to it while under the impression that it is somebody else — who is He? is a genuine question, and the attraction depends on the misidentification. The moment the mistake resolves, the state must end.
Which means the pillar gives Kṛṣṇa the experience of being a lover only by giving him a false belief. The corpus's three confidential desires are not requests for an illusion; they are requests for the real thing, which is why an incarnation was necessary and a mirror was not sufficient.
The evidence establishes that the state is desirable and not that it is available, and the distinction is exactly what the rest of the book is about. The corpus does not draw it.
Worth noting how much of this is available without any theology, because the episode's grip on a reader comes from something everybody has.
The reason you cannot fall in love with yourself is not modesty. It is positional: falling in love requires an object you do not have full access to, whose next move you cannot predict, and who might not be interested. You fail all three tests about yourself, permanently, by being yourself.
The mirror suspends the first two for a third of a second. And what people report from that interval is not attraction exactly, but interest — a brief availability of oneself as an object of curiosity, which is normally impossible and which people pay therapists to arrange more durably.
The corpus's episode is that interval, held open, and taken to its extreme. Which is why it lands even for a reader who thinks the theology is furniture.
Two features of the staging are doing work, and both are choices.
It happens in Dvārakā. The corpus is emphatic elsewhere that Dvārakā is the majestic register — the place where everybody knows exactly who he is, where reverence is at its highest and, by the corpus's own doctrine, love at its most crippled.5 Setting the episode there is pointed: even in the setting where the relationship is most formal, the desire for the other position surfaces. It could have been set in Vṛndāvana and would have proved nothing.
And it is a pillar rather than water. A jewelled pillar is architecture — installed, polished, part of a palace built for him. The reflection is produced by his own opulence. The apparatus of his majesty is what shows him the thing his majesty cannot give him, which is as neat a compression of the aiśvarya doctrine as the corpus manages anywhere.6
Neither detail is remarked on. Both are the kind of thing an author chooses.
Layer A gives the scene, the question and the comparison to Rādhā, briefly, as one rung in the dialogue.
Layer B connects it to the descent doctrine and to the hlādinī apparatus, which is the correct systematic placement and more than the verse says.7
The divergence is mild. The verse reports a moment; the commentary makes it a proof. Both readings are available, and the page's argument above is that the proof does not quite hold — which is a criticism of the systematic use rather than of the passage.
You catch yourself in a window and for a moment see a person. Stay in it as long as it lasts. That interval is the only access you have to yourself as an object rather than as a vantage, and it does not come often.
You want to know how you land on people and you keep asking them. The asking produces descriptions, and descriptions are not the thing. The corpus's God had the same problem and could not solve it by asking either, which is why the book exists.
You are attracted to something and you notice the attraction depends on not knowing something. That is not a defect and it is worth knowing. Full access and attraction are hard to hold together, and most arrangements that promise both are managing the information.
Somebody's love for you seems to require a version of you they have not checked. Before you correct them, consider what the correction costs. This corpus's whole doctrine of majesty is that the correction is what does the damage.
The evidence is a short narrative moment with a quoted line and an explicit comparison to Rādhā's state.8 It is compact and its systematic significance is supplied by the corpus rather than inferred.
The tension the source doesn't close: the reflection gives him the lover's position by giving him a false belief. The corpus's descent argument requires the position without the error, and this episode cannot supply it.
A second one. The state is reported as ending — the misidentification resolves — and the corpus does not say what he concludes afterwards. An experience that terminates when the mistake is corrected is a different kind of evidence from one that persists.
Open: whether the episode is the corpus's own or inherited. It reads as a set-piece and nothing in the range read indicates whether Kṛṣṇadāsa found it or made it, which bears on how much of the descent argument is his construction.
This page is the evidential base of The Three Confidential Desires, and the two should be read in that order. The desires state what he wants; the pillar is the corpus's one demonstration that he wants it. Without the pillar the descent argument is an assertion about God's inner life with nothing behind it.
It sits directly on Āśraya and Viṣaya: The Abode Enjoys More as the moment that ratio is briefly inverted. He is the object; the reflection puts him in the holder's position; and the corpus says what he feels there is what she feels — which is the ten-million claim tested on the one being who could compare the two seats.
Against Majesty Weakens Love the setting is the argument. The episode occurs in Dvārakā, in a palace, in a reflection thrown by his own opulence. The apparatus of majesty produces the glimpse of what majesty forecloses, and the corpus stages this without comment.
See Yourself as Others See You — Temple Grandin cannot perceive her own effect on people, and builds a painstaking workaround: a deliberate capacity to review her own past behaviour as though watching a stranger, assembled over years because direct perception was never going to arrive.
The pillar is the same deficit and the accidental version of the same solution, and the comparison makes the corpus's argument sharper than the corpus states it.
Grandin's method works by temporal displacement: the stranger she watches is her past self, and the distance that makes the review possible is time. That is available to anybody, it is laborious, and its product is a reconstruction rather than a perception.
The pillar works by misidentification: the stranger is present-tense and immediate, and the distance is supplied by an error. Which produces something Grandin's method cannot — an actual first-person reaction to oneself rather than an assessment of oneself.
And the trade is exact. Grandin's version is repeatable and gives you an analysis. The pillar's version is unrepeatable, lasts a third of a second, and gives you the thing itself.
Which is why the corpus needed an incarnation and not a mirror. The reflection proves the state is desirable and delivers it only under a false belief; Grandin's review delivers information and never the experience. Neither route gets you what the three desires ask for, and the corpus's answer — change position permanently, at the cost of your original standing — is the only remaining option once both of these have failed.
That is a genuinely useful thing to have established, and it holds outside the theology: the two ordinary routes to seeing yourself from outside are reconstruction and error, one gives you data without experience and the other experience without duration, and there is no third.
Sharpest implication. The corpus's entire descent argument rests on a third of a second in which God mistook himself for somebody else — and the mistake is load-bearing, because the position he wanted is not available to anyone who knows who they are looking at. Every route to occupying the lover's position with respect to oneself runs through either a reconstruction or an error, and this corpus's whole project is the attempt to find a third.
Generative questions.