Source ingest: Bibek Debroy, tr., The Bhagavata Purana, Volume 1 · 93 chapters, 1,575 endnotes, 173 pages built, folded to Kṛṣṇa Bhakti — Map of Content at §17–§43.
What does a text do with a disagreement it cannot afford to settle?
That is what four Skandhas turn out to be about, and the answer is consistent enough to be called a method. The volume disagrees with itself seven times on numbered doctrine. It stages three confrontations it cannot let anyone lose. It issues two contradictory orders in one speech. It teaches a woman that women are the gate to hell in a discourse God took a body to deliver to her. Not one of these is resolved, and not one is presented as unresolved. Something is always moved — a share, a jurisdiction, a head-jewel, a local gloss — so the contradiction can be carried instead of adjudicated.
[CONTESTED] and Even a Kirāta Can Do This — the exclusion clause against the purification roster1. Seven internal disagreements, seven local glosses, no flag. Sheaths seven or eight; ignorances four or five with different names; the cosmic body mapped three incompatible ways with a fourth admitted to exist; two avatāra catalogues; two causal accounts of one curse minutes apart in one speech; two etymologies for Parīkṣit, both kept; two tattva schemes both totalling twenty-five with almost no shared membership. [n≥3 PATTERN]. Debroy annotates each occurrence to fit its own local number. Across his corpus the same behaviour was already recorded from the Śiva Purāṇa build, so the translator pattern is [n≥3] too.
2. Every great confrontation ends by amending an allocation. [n≥3 PATTERN] Kali, Rudra, Indra. Three antagonists, three ledger entries, no defeat. And the volume's rhetoric throughout is about destroying the wicked, which nothing in these three scenes does.
3. Recognition fails at every position in the hierarchy. [n≥3] Brahmā does not recognise the thumb-sized boar and debates with his sons what the thing in the sky is. The Sātvatas live with Kṛṣṇa and take him for a bull among the Sātvatas. The sages hear God's words attentively, find them deep in import, and do not understand. Brahmā cannot find his own root and says so in the first person.
4. The procedures are exact and the ontology is plural. Seven days, five nights, ten days, ninety-nine thousand yojanas, fourteen repetitions of a refrain — against sheath-counts and body-maps that will not sit still. Developed at The Directions Are Exact, the Destination Is Not, where the candidate explanation is that precision tracks testability rather than importance.
5. The volume's widest claims arrive in registers that require nothing afterward. The purification roster is inside an invocation. The strongest disclosure Debroy makes lands inside a speech condemning a father. Nārada's referral of Kālakanyā to Fear is a clause of connective narration. The load-bearing sentences are consistently in the positions readers skim.
A corpus that survives long enough stops treating contradiction as an error and starts treating it as freight — and it develops handling equipment: allocation, substitution, and the local gloss. All three preserve the contradiction perfectly while making it invisible to anyone reading in one pass.
The equipment is not deceptive and nobody in the chain intends concealment. The composer carries two accounts because both are authoritative. The translator glosses the passage in front of him because that is what an annotation is. The reader reads forward. Three innocent behaviours compose into a text that disagrees with itself seven times and reads as consistent.
A. The acquittal-refused shape. [WILDCARD] This volume has a man refuse an acquittal he is offered; the Śiva Purāṇa corpus has acquittals handed out and ignored. Two Purāṇas, one translator, structurally opposite handling of the same juncture. This is [n=2 SUGGESTIVE] and is deliberately not named as a signature. No third instance appeared anywhere in Skandhas 3–4. The thread is: what would a third instance look like, and which corpus is most likely to hold one — and until it arrives this stays a thread and does not become a page. See The Acquittal He Refuses.
B. ✅ RUN AND CLOSED the same day — and it refuted the claim it was written to test. Debroy's notes do cross-reference his own translations, repeatedly, including one that lands on this build's own 4(24) (Śiva Purāṇa Vol 3 n. 1334, "Rudra Gita is from the Bhagavata Purana (4th Skandha)"). But only backward. Bhāgavata Volume 1's 1,575 notes name no other Purāṇa at all, because it is the earlier book. A growing translated corpus is annotated backward only, which leaves its oldest volume sealed against everything that followed. See The Same Search, Two Books, One Translator, revised.
B′. The successor, and now the strongest item here. [WILDCARD] Do the later Debroy volumes function as a partial errata for the earlier ones? Vol 3 n. 1334 supplies a title for a passage the Bhāgavata never names — the phrase Rudra Gītā occurs nowhere in Volume 1. If there are more such notes, then material that corrects or names things in the books this vault cites most heavily is sitting unread in the apparatus of the books it cites least. A sweep of the three Śiva Purāṇa volumes' notes for Bhagavata and Skandha references is an hour.
C. Provenance of the settlement pattern outside devotion. [NEWSLETTER] Whether the licence-rather-than-verdict mechanism is findable in regulation, litigation and harm-reduction policy, and — the part that decides whether it is interesting — whether such an allocation has ever been withdrawn. See The Fight Ends by Amending the Allocation.
D. The world-size test. [NEWSLETTER] Withheld presence produces identical behaviour to intermittent reinforcement, and the two domains give it opposite moral valences with no way to arbitrate from inside either. The proposed external criterion is whether the subject's range of action expands. It has not been applied systematically and it would fail a lot of devotional material. See Does the Subject's World Get Bigger?.
E. The frame around the Purañjana allegory. [WILDCARD] The animal-vengeance argument brackets the allegory at both ends. Whether it recurs inside it is unchecked, and it is one read of five chapters. See The Allegory Was the Envelope.
Angle 1 — Nobody Loses Here
Angle 2 — The Name Was the Argument
[CONTESTED] and is not a punchline.Angle 3 — The Allegory Was the Envelope
Six filed, all speculative: the fight ends by amending the allocation · the teaching delivered to the person it disqualifies · canonised with the crime still attached · the sentence is shaped like the gesture · the man who refused her told her where to go · the same search, two books, one translator.
Strongest for promotion: the fight ends by amending the allocation, which has [n≥3 PATTERN] support inside one volume, a stated falsifier, and two independent vault corroborations already on disk.
Weakest, and named as such: the sentence is shaped like the gesture, which rests on [n=1 scene] containing three instances. It is a scene, not a regularity, and the page says so.
Four filed to META/open-questions.md on 2026-08-12: whether a settlement-allocation has ever been withdrawn; whether any Debroy note cross-references another of his own translations; what the vault does with a [CONTESTED] teaching whose recipient the text then liberates; and whether doctrinal precision in a Purāṇa tracks testability rather than importance.
Run the Debroy cross-reference grep (thread B). Done, 2026-08-12, and it refuted the hypothesis — see thread B above.
Next: run B′ — sweep the three Śiva Purāṇa volumes' endnotes for Bhāgavata and Skandha references. The one instance already found supplies a title (Rudra Gītā) that Volume 1 never uses for a chapter this build wrote a page on. If that is not isolated, the later volumes are an unread errata layer for the earlier ones, and that is a finding about how this vault reads its most-cited translator rather than about any one passage. An hour.