Enumeration Is the Same Instrument in Both Hands
- The Asura Boon Loophole — a demon lists every way he could die and is killed in the one gap he did not list.
- The Asura Boon Loophole — a demon lists every way he could die and is killed in the one gap he did not list. The vault has read this as a story about asuric cleverness getting what it deserves. - The Nārāyaṇa-Kavaca as Threat Model — and the same volume's defensive enumeration, the armour the gods themselves wear: protections partitioned by place, by kind of harm, and by hour of the day. Nobody in the text calls it legalism. It is a hymn. - The Seam-Defeat — the pattern at n=6, and the finding that no contract, boon or vow in six Skandhas fails head-on. Every one fails at an unenumerated case.
The moral reading needs the demon to be doing something characteristically demonic. He is not. He is doing what his enemies do, in the same book, with the same tool.
Put the two specifications side by side and you cannot separate them by form. Hiraṇyakaśipu enumerates killers, times, places, instruments. The Nārāyaṇa-kavaca enumerates domains, harms, watches of the day. Both work by naming cases and covering each. Both are exhaustive in intent and finite in fact.
And the part the volume never says: the armour is not shown to be better. It is never tested against an unlisted case. The one time it operates outside a battle it knocks a gandharva king out of the sky for flying over a dead practitioner's bones — a man in none of its categories, treated as an attack because the thing has no way to represent intent.
So the text runs one structure twice, hands it to opposite sides, and stages a defeat only for the holder it disapproves of.
The seam-defeat is not a verdict on the specifier. It is a property of specification — and the corpus supplies its own control.
If enumerated defence failed because asuras are proud, the gods' version would work. Instead it goes untested, which is a different thing from working, and the difference is exactly where a moral reading hides.
What generalises: any defence that reaches coverage by listing cases inherits two failure modes at once, usually discussed apart. It cannot cover the unlisted case — the demon's problem, under-firing. And it cannot tell a listed case from one that merely resembles it — Citraratha's problem, over-firing. Both are the same absence: no model of what is actually coming. A threat framework that asks four questions about an adversary needs no list of situations. The kavaca lists situations and never asks who the adversary is.
Cheap test: wherever a corpus supplies both an offensive and a defensive enumeration, the defensive one will go untested — because testing it would show the flaw was never moral.