Somebody gives you a piece of advice with two halves. Do the hard thing, and tell them why first.
You do the hard thing. You skip the telling-them-why, because you were busy and it seemed like the soft part.
And it goes badly in exactly the way the second half was there to prevent, and when you report back the person who advised you does that particular face. Not smug. Tired. They can see that you took the half you already agreed with.
This chapter has a name for that, an image for it, and a corollary that goes considerably further than most people would want to.
Ardha-kukkuṭī-nyāya — the logic of half a hen.
A man reasons that the front half of the hen eats grain and costs him money, while the back half lays eggs and earns him money. So he cuts the hen in half and keeps the profitable end.1
It is a joke, and it is a joke about a specific error: treating an integrated thing as a menu. The two halves of the hen are not two goods with separate accounts. They are one animal, and the accounting that separated them was a category mistake that killed it.
Kṛṣṇadāsa deploys this against a man who accepts Caitanya and is unconvinced about Nityānanda: "If you have faith in one but disrespect the other, your logic is like the logic of accepting half a hen."2
And then the sentence that makes this more than a proverb:
"It would be better to be an atheist by slighting both brothers than a hypocrite by believing in one and slighting the other."3
Whole rejection is ranked above partial acceptance.
That is a striking piece of doctrine and it should be felt before it is analysed. On the ordinary devotional scale, an atheist is at the bottom and a partial believer is somewhere above them, closer, further along. This inverts it. The atheist is at least coherent; the partial believer has taken the object apart, which is the specific error the hen died of.
There is something genuinely bracing about it. Most traditions flatter the half-convinced because the half-convinced are the recruitment pool. This one tells them they are worse than the door they came in through.
It is worth granting how good this is before noting what it does, because the argument is not a trick.
Selective adoption of an integrated system is a real failure mode and it is everywhere. People take the meditation and leave the ethics. They take the productivity system's capture step and skip the review, which is the step that made capture worth doing. They adopt a management framework's rituals and not its underlying transfer of authority, and then report that the framework doesn't work.
In each case the discarded half was load-bearing for the half that was kept, and the person doing the selecting could not see that, because the criterion they were selecting on was which half do I already agree with — and the parts you already agree with are, by construction, the parts that don't change anything.
The half-hen image nails that precisely. It is one of the more useful things in this corpus and it travels intact.
The argument's force depends entirely on the object actually being one thing, and that is the question it never asks.
A hen is unseverable. That's a fact about hens, and it is why the image is funny.
Whether a body of claims is unseverable is not a fact — it is a claim being made by the person holding the body of claims, and it is precisely what a partial believer is disputing.
So the argument has a hole in it exactly where it looks strongest. You cannot take half of me is unanswerable when said by a hen and is a position when said by a doctrine. The brother in this chapter was not denying that Caitanya and Nityānanda were connected. He was denying that they were the same object, which is the disputed point, and the hen image assumes it rather than arguing it.
That gives a usable test, and it is not the one the chapter offers. The question is not whether the parts are related. It is whether keeping one and dropping the other makes the kept part stop working. Meditation without ethics does degrade — that's checkable, and the half-hen charge sticks. A theology in which one figure is fully divine and another is highly exalted is not obviously broken; it was, in fact, a widely held position in sixteenth-century Bengal, held by people who functioned.
The corpus needed it to be broken and asserted a hen.
Which is where the page has to say what the argument is doing in its actual setting.
It is being used to close a middle position, in a chapter where the person occupying that middle position is then struck down and the man who argued against him is rewarded in a dream.4 The half-hen reasoning is not deployed abstractly. It arrives with a demonstration attached, aimed at a constituency the corpus knew was real.
Both things are true and neither cancels the other. The reasoning is good and its deployment here is a warning shot, and a reader who takes the image away for their own use — which they should, it is excellent — should also notice what it was built to do the first time.
Layer A gives the image and the corollary in the author's own voice, in the middle of his autobiography, without softening.
Layer B is at home with it, and extends the corollary into general instruction about the danger of Vaiṣṇava offence and the necessity of accepting the whole Pañca-tattva rather than a preferred member.5 That is a fair extension. It is also the reading that keeps the middle position closed for a twentieth-century audience with the same temptation.
The build's pattern holds: the verse makes an argument; the commentary converts the argument into a standing requirement.
You've adopted a system and dropped a step. Ask the diagnostic question rather than the loyalty question: does the part I kept still work without the part I dropped? If it does, you have edited sensibly. If it doesn't, you have bought a hen and eaten the wrong half.
Somebody tells you that you cannot take part of their position. Notice they have just made a claim that needs support — the unseverability is the thing in dispute, and asserting it more firmly is not an argument for it.
You're the one with the integrated system and somebody is cherry-picking. Resist the urge to reach for the hen. Show them the failure instead: name the specific way the piece they kept degrades without the piece they dropped. If you can't name it, the pieces may genuinely be severable and your objection may be about loyalty rather than coherence.
You find yourself half-persuaded and being told that half-persuaded is worse than unpersuaded. That argument is always made by people who need the middle closed, and it is sometimes also true. Work out which by asking what the middle position actually costs them.
The evidence is a classical figure deployed in one sentence, plus the corollary stated in the next.6 It is compact and the reasoning is fully explicit — one of the few places this corpus argues rather than asserts.
The tension the source doesn't close: unseverability is assumed, not shown. The image works only if the object is one thing, and that is exactly what the person being rebuked denies.
A second one, and it is about the corollary. Better an atheist than a hypocrite is stated as a ranking of spiritual conditions. The corpus elsewhere holds that the name works irrespective of intent — that men who chanted in derision could not stop, and were delivered by it.7 Under that doctrine the mocking atheist is well placed and the partial believer is not, which the corollary would endorse; but the corpus has no account of how a contagion-theory of the name and a coherence-theory of belief fit together, and it uses both.
Open: does the half-hen charge apply to the corpus's own conduct? It routinely takes half of an opponent's position — accepting a scriptural authority while rejecting its commentary, accepting the Vedānta-sūtra while refusing Śaṅkara's reading. Nothing in the range read applies the figure reflexively.
This page is the argumentative instrument of The Author's Own Conversion, and neither should be read alone. There the image arrives with a man falling down attached; here it is examined as reasoning. Separating them lets the argument be assessed on its merits and lets the deployment be assessed on its own, which is exactly what running them together prevents.
It is the operational face of Six Diversities in One. That page's structural claim is that declaring categories identical rather than complementary removes proportionality — there is no small offence against something with no parts. Half-a-hen logic is that consequence stated as a rebuke: if the object has no parts, taking part of it is not partial acceptance but a category error.
Against All the Names Are Correct there is a real inconsistency worth flagging. That chapter permits devotees to hold materially different identifications of Kṛṣṇa — Nara-Nārāyaṇa, Vāmana, Kṣīrodakaśāyī Viṣṇu — and rules that "each is as correct as the others" because a sincere devotee's words cannot be false. That is maximal latitude on the identity of the object of worship. Two chapters later, a man with full faith in Caitanya and partial faith in Nityānanda is told he is worse than an atheist. The corpus's tolerance is wide on what its God is and nil on which of the five you may under-rank.
Both/And Thinking: The Capacity to Hold Contradiction — that page's argument is that most thinking runs on dichotomy, that dichotomy simplifies choice and falsifies reality, and that the mature capacity is holding two things at once: strong and sensitive, ambitious and at peace, a belief that is both true and limited.
Half-a-hen logic is the case against that capacity, and the collision is worth having rather than smoothing, because both are right about different objects.
Both/and describes holding a tension: two things that appear to exclude each other and both obtain. Half-a-hen describes taking a part: one thing severed, with the inconvenient portion discarded. Those look similar from outside — in both cases someone is refusing a simple either/or — and they are opposite operations. Holding a tension keeps the whole object and tolerates the discomfort. Cherry-picking removes the discomfort by removing the part causing it.
So the test that separates them is not how many positions you're holding. It is what happened to the part you found uncomfortable. Both/and thinkers are still carrying it. Half-hen thinkers put it down and kept walking, usually while describing themselves as nuanced.
And that has a sting in both directions. A tradition can use the half-hen charge to forbid genuine both/and — which is what happens in this chapter, where a man holding two figures at different valuations is told he has butchered one object. Equally, I'm holding the tension is the most respectable available cover for having quietly dropped the half you disliked. Neither page can tell you which you're doing. The only question that can is the one from the case study above: does the part you kept still work? A held tension keeps functioning under strain. A severed half stops laying eggs.
Sharpest implication. Whether you may take half of something is not a question about your integrity — it's a question about the object, and it is answered by whether the kept half still works. Every doctrine that forbids partial acceptance is asserting an answer to that question, and asserting it most loudly at exactly the point where it is least demonstrable.
Generative questions.