Six Recensions, and the Question None of Them Was Built to Ask

The Recension is not mine to invent and not mine to skip. Every seventh year the Register is checked against itself — entry against entry, hand against hand, no name permitted to answer to two dates or two fathers without a note explaining which is wrong. It last ran in Year 234. It runs again before Harvestmonth closes, the seventh in the line this office can confirm, and I have spent the week since the Almanac Corner named it deciding what, honestly, six prior audits have actually established.

Less than the practice would like believed. I set it down here before the seventh begins, so that whatever it finds, or fails to, is judged against what came before it rather than against its own good opinion of itself.

The line, so far as it can be traced:

Recension	Year	Standing of the record
Sixth	234	Completed on schedule. Findings sealed, and unremarkable — a verdict I did not question until this week, and should have questioned sooner. An audit that finds nothing wrong has either found nothing wrong, or has not known where to look.
Fifth	227	Delayed three years by the pass dispute with Kael Dorn — the clerks who would have run it were reassigned to border correspondence for the duration. Resumed and closed in 230, six years late against its own clock, on time against no one’s.
Fourth	220	Completed on schedule. A later hand, working the 215 marriage rolls for an unrelated matter, found two names transposed that this Recension had passed as sound five years earlier. The correction stands. The Recension that missed it does not record having missed it; I am the one recording that now.
Third	213	Only the summary survives. The working papers — the actual comparisons, entry against entry — do not. A conclusion without its evidence is a claim, not a finding, and this office has been content to treat it as the second for want of the first until the writing of this sentence.
Second	206	Known only from a footnote in the 213 summary, itself unverifiable by the above. Nothing else of it survives under any hand this Archive holds.
First (on record)	199	A single line in the Register’s founding charter, naming “a verification of the rolls” without method, scope, or outcome given. Whether this was the practice’s origin or an older custom under a different name cannot be settled from what remains. I decline to settle it.

Four of six are honest. Two are inherited on faith. That is the state of the instrument the seventh Recension will use to judge itself faithful to the six before it — a standard partly built on air, and no prior Recension thought to say so in writing until tonight.

But the deeper fault runs under all six, sound and unsound alike, and no correction to the working papers will touch it. Every Recension ever held has checked the Register against itself. None has checked the Register against the city.

This week alone, and I did not have to search for the examples — a second well, on Mulov Street, was fitted with the same vinegar-soaked cloth as the first, tied wrong at first and corrected within the hour, by no office’s order and to no office’s credit; a season’s latches went up across half the residential quarter answering a rumor no charter ever authorized anyone to answer; a ship held six days past due was towed, isolated, and treated by three separate offices acting on nothing but their own competence, three days running, before one petition reached a desk that could open a file on any of it. A Harbor Factor priced what those three days cost in tow-fees this week and found he could not price what they had, in the meantime, quietly settled. None of this will appear as an error when the seventh Recension runs its comparisons, because none of it is an error. The rolls will show no name entered twice, no date contradicting itself, no forged hand where a true one should stand. The rolls will be correct. The city, in the meantime, will have gone on governing itself in the gaps between what gets entered and what gets done, exactly as it has for six Recensions running, and the seventh will certify the Register sound without once asking the only question that would tell it anything worth knowing: not whether the record agrees with itself, but whether it agrees with what the street already knew.

I will run the seventh by the same rule as the six before me. I know, writing this before it begins, most of what it will not find. Whether that makes me the first Archivist honest enough to say so in advance, or merely the first with the nerve to file the excuse ahead of the failure, I leave for whoever runs the eighth.