Brahmā is narrating Kṛṣṇa's life to Nārada, in advance — the whole biography delivered as prophecy, from the killing of Pūtanā to the death of Kaṃsa, hundreds of years before any of it happens. The tense throughout is future: he will do this, he will kill that one.
And then one verse, describing an event from the Rāmāyaṇa, slips. Debroy notices, fixes it, and records the fix:
There is a minor tense issue in this shloka, where the future tense is used, the dialogue between Brahma and Narada having occurred prior to the Ramayana incidents. However, we have stuck to the past tense of the other shlokas.1
Three things happen in that note. He identifies an inconsistency in the source. He states which way he resolved it. And he leaves the record of both in a place the reader can find.
The chronology is worth walking through, because the problem is genuinely confusing and the note is compressed.
Brahmā's conversation with Nārada sits before the Rāmāyaṇa's events in the narrative timeline. So when Brahmā describes Rāma's deeds — the ocean granting passage, the tusks of Indra's elephant shattering against a chest — he should be speaking about the future, and future tense would be correct.
But the surrounding verses in this section are rendered in past tense, describing Rāma's deeds as accomplished. One verse breaks the pattern and uses future. Debroy's judgment is that the pattern is what to preserve — so he normalizes the outlier to match its neighbours, converting a future-tense verse to past.
The irony is that the outlier may have been the chronologically correct one. Debroy isn't claiming his fix restores accuracy. He's claiming it restores consistency, and saying so plainly enough that a reader can see the trade he made.
This is the most precisely documented intervention in the volume's apparatus, and it's worth isolating what makes it different from the others.
The blanket disclosure concedes that simplification occurred across three chapters, without saying where. The seven-mothers note reports a failed arithmetic check and offers an alternative reading. The Vidura note contradicts the narrative on external grounds. Each is honest, and none is fully auditable — a reader can't reconstruct exactly what the translator did.
Here they can. The note specifies the location (this shloka), the problem (future tense where the pattern is past), the decision (normalize to the pattern), and implicitly the alternative that was rejected (preserve the future tense, breaking the pattern but matching the chronology). A reader who disagrees with Debroy's choice has everything needed to reverse it.
That completeness is rare, and it's rare for a structural reason: full disclosure is only practical when the intervention is small. You can document a single tense change exhaustively. You cannot document three chapters of simplification the same way without producing a second book. The volume's most transparent note concerns its least consequential change, and that relationship is probably not a coincidence.
The tense problem sits inside a genuinely strange structural choice worth naming: this passage delivers Kṛṣṇa's entire biography as forecast. Pūtanā, the cart overturned by a kick, the arjuna trees uprooted, Kāliya, Govardhana held up like a mushroom, the rāsa dance, Kaṃsa — all of it narrated by Brahmā, in advance, in a conversation that predates the events by an age.
That means the Tenth Skandha, the most-read portion of the entire Bhāgavata and the section this whole tradition is built around, is summarized in the Second Skandha as prophecy before it's told as narrative. The tense slip Debroy repairs is a small crack in a passage that's already doing something temporally unusual — and a passage narrating events across multiple different avatāras, spanning ages, from a fixed point before most of them, is exactly where tense confusion would be most likely to arise in the first place.
The note opens by calling this a "minor tense issue," and that characterization is doing quiet work worth noticing. Whether a tense inconsistency in a prophetic passage is minor depends entirely on what you think the passage is doing.
If Brahmā's forecast is simply a narrative convenience — a way of summarizing later material early — then yes, a tense slip is a cosmetic blemish. But if the passage is making a claim about how divine foreknowledge relates to time, a verse that breaks from prophecy into retrospection might be doing something rather than failing at something. A speaker who slips into past tense while describing the future could be read as a speaker for whom the distinction genuinely doesn't hold.
Debroy's "minor" forecloses that reading before it can be raised. That's a defensible editorial judgment — the simpler explanation is almost certainly right, and a translator who entertained every such possibility would never finish. But it's a judgment, delivered in an adjective, and worth marking as one.
One clarification worth making explicit, since the note compresses it: the reason future tense would be chronologically correct is that the Brahmā–Nārada conversation is itself embedded several layers deep in the volume's frame structure. Sūta is narrating to the Naimiṣa assembly what Śuka told Parīkṣit, which includes what Brahmā told Nārada, which concerns events spanning multiple ages.
At the innermost layer, Brahmā speaking to Nārada, the Rāmāyaṇa hasn't happened. At the outermost, Sūta speaking to Śaunaka, it happened long ago. A verse's correct tense therefore depends on which narrative layer you take as the reference point — and in a text nesting speakers this deep, the answer isn't obvious. What Debroy calls a minor issue may be the natural consequence of a frame structure that gives any given verse several defensible temporal anchors.
You're editing, translating, or otherwise mediating a source, and you encounter a small inconsistency — a mismatched tense, a name spelled two ways, a figure that doesn't match its neighbours. The fix is obvious and minor. The temptation is to make it silently, because documenting it feels disproportionate to its size.
This note argues for documenting it anyway, and the reason isn't primarily about this particular fix. It's that a reader who finds one small, fully-documented intervention learns something about the mediator's practice generally: that they notice inconsistencies, that they choose deliberately rather than by reflex, and that they consider the reader entitled to know. That inference is worth far more than the specific tense change, and it's only available if the small fixes are on the record.
The corollary is worth stating too: a mediated text with no documented small fixes isn't necessarily a text that needed none. It may simply be one where the fixes weren't recorded, which leaves a reader unable to calibrate anything.
Placed alongside the volume's other disclosed moves, this note completes a taxonomy worth stating in one place. Across the apparatus, five distinct kinds of intervention get disclosed: blanket simplification over a range (endnote 4), divergence from another text (141, 314), arithmetic that fails a check (259), correction of the narrative's own premise (295), and — here — silent normalization with the receipt left in.
Only this last species involves the translator changing what the text says to match a pattern, rather than clarifying, comparing, or contradicting. The other four leave the rendered text as the Sanskrit gives it and add commentary around it. This one alters the rendering and then confesses the alteration.
That distinction matters for how a reader should weigh each species. Four of the five affect a reader's understanding of the text while leaving the text itself intact. The fifth affects the text. It's the smallest intervention in the volume and the only one that changes what's actually on the page.
The strongest evidence for this note's value is its completeness. Location, problem, decision, and implied alternative are all present, which makes it the only intervention in the volume a reader could fully reverse if they disagreed with it.
Tension — the fix chosen may be the chronologically wrong one. If Brahmā's conversation genuinely predates the Rāmāyaṇa, the future-tense verse was correct and the surrounding past-tense verses are the anomaly. Debroy normalizes toward the majority rather than toward the chronology, and doesn't argue for why consistency should win. Preserved.
Open question. Are there other tense inconsistencies in this same prophetic passage that went unnoticed or unrecorded? A passage narrating multiple avatāras across ages from a fixed prior point is a natural site for such slips, and one flagged instance suggests the possibility of others.
Set against the rest of this volume's apparatus, this note occupies the far end of a spectrum. Endnote 4 discloses at maximum scope with minimum specificity — three chapters, no locations. This note discloses at minimum scope with maximum specificity — one verse, fully documented.
Both are honest. The difference is what each makes possible for a reader: endnote 4 supports calibration ("be careful across this range"), while this note supports verification ("here is exactly what I did, check it if you like"). A reader who wants to audit the translation can only do so where interventions are documented at this level, which is to say almost nowhere — and that limitation is a fact about the practical economics of annotation rather than about this translator's willingness to disclose.
With The Kumarajiva Translation Workshop. That page describes a translation operation run as a workshop — a room of named pupils working under a master, producing the Chinese Buddhist canon through collaborative labour rather than solitary rendering. The workshop model has a specific consequence for documentation: with multiple people handling the same text, decisions have to be at least somewhat externalized to be coordinated at all.
Debroy's practice here is the solitary translator's equivalent — externalizing a decision that could have stayed internal, writing it down where a reader rather than a collaborator will find it. What the pairing produces: the workshop makes translation decisions visible by necessity, because coordination requires it. The solitary annotator makes them visible by choice, and only for the fraction small enough to document. That asymmetry suggests something worth carrying elsewhere: the transparency of a mediated text may depend less on the mediator's honesty than on whether the production process structurally required decisions to be spoken aloud. A workshop leaves a record because it must. A single scholar leaves one only where he decides the reader is owed it — and where the record wouldn't cost more to produce than the text itself.
Sharpest implication. The volume's most completely documented intervention is also its least consequential, and that inverse relationship is structural rather than accidental: full disclosure scales inversely with the size of what's being disclosed. Which means a reader's ability to verify a mediated text is systematically strongest exactly where verification matters least.
Generative questions.