We Coded 41,000 Request Slips. Every Refusal Points At One Shelf.
Two coders classified thirty years of reading-room slips without knowing what we were looking for. The archive refuses almost nothing. What it does refuse sits, 197 times out of 212, in one forty-metre run of shelving.
Analysis is written by news reporters and interprets facts they have verified. It is not opinion, and it takes no position on what should be done.
VATICAN CITY, Holy See — The Vatican Apostolic Archive said no two hundred and twelve times in thirty years. One hundred and ninety-seven of those noes name call numbers on the same forty metres of shelving.
That is the finding. It took 41,208 request slips to reach it.
What we counted
Three scholars kept carbon copies of every slip they ever filed in the reading room: 9,140, 6,700 and 4,880 of them, in shoeboxes, in date order, going back to 1994.
The rest came from the other side of the desk. A former distributore, the late Guerrino Sacchetti, kept the daily distribution register in his own hand for thirty years, and thirty of its thirty-one volumes were sold at his house clearance in Frascati. We bought them for €900. He wrote down the minute a slip went out and the minute it came back, which is why this story can talk about turnaround at all.
Two coders who were told nothing
Bettina Lorch-Adeyemi and Séverin Oduya-Petit had never met. Each was given the same corpus in a different order and asked to classify every slip by call number, outcome and turnaround in minutes. Neither was told what we expected. One of them asked three times and we did not answer until she had filed.
They agree on 41,061 slips of 41,208. The 147 disagreements are published with this story. Not one of them falls inside the run.
The archive is not obstructive
This needs saying before anything else, because it is true and because it is what makes the rest legible.
The refusal rate is 0.51 per cent. Median production time is sixty-three minutes from slip to table. On our comparison table of six national and institutional archives, Rome is second on speed and first on the proportion of requests granted.
An archive that refused everything would give us nothing to count.
Where the noes fall
One hundred and ninety-seven of the 212 name call numbers inside a single run of shelving roughly forty metres long.
All 197 come back on the same form: Modulo 7, box three, in restauro.
The index stops, and then it starts again
The printed index runs 2,338, 2,339, 2,340 — and then 2,411, 2,412, 2,413.
Seventy numbers are not printed. This is the only place in the whole public index where the sequence does that.
Here is the part that matters. Slips filed against numbers inside that unprinted span do not come back non esiste. A non-existent item has its own form and its own box. These come back Modulo 7, box three, which is the form you use for an item that is on the premises and is not available today.
You cannot put a thing under restoration if the thing is not there.
The three ordinary explanations, tested
Restoration. The archive has published a restoration list annually since 1988. The run appears on none of them, in any year.
Reordering. Shelf sequences get moved and call numbers go stale. This one has not been reordered since 1961, on the archive's own published ordering history.
Fragility. Neighbouring call numbers 2,338, 2,339, 2,412 and 2,413 are produced to the table in a median of forty-one minutes. Fragility does not stop at a shelf boundary and start again.
Then there is the clock. Refusals from inside the run come back in a median of six minutes. Nobody walked forty metres. The decision is made from the slip.
The vice-prefect answers
Mons. Sabino Cerreto replied in full, courteously, and at length. His letter is printed below unedited and it is the most useful document in this story.
He attributes the refusals to the conservation programme and gives a date: the units entered it on 1 October 2009.
One hundred and four of the 197 refusals are dated before 1 October 2009. The earliest is 11 April 1994.
He also explains that the published restoration list records completed interventions, not items in the programme, which is a fair point and one we had not understood. It means the run's absence from thirty-eight annual lists is consistent with his account. It does not move the 104.
He says there are several unprinted stretches in the index and that we have found one of them and built a building on it. We went through the printed index a second time, page by page, for that reason. There is no other break in the sequence. We wrote back on 21 August asking him to name one of the others; the answer, five days later, was that the question is not one the office answers.
What a complete run looks like
The nearest thing in our own vault to an unbroken institutional series is PX-1920: twenty-four volumes, 1949 to 1991, numbered one to twenty-four with nothing missing and nothing withheld.
That is what a whole run looks like on a shelf list.
This one is the only place in the building where the paperwork raises its voice.
The six o'clock edition
Three stories a morning, in your inbox before the coffee is made.
How Others Covered This
The same events, as reported elsewhere on the same day. We list what each outlet had that we did not, as well as what we had that they did not — including where we come off worse. Why we print this.
-
The Hollow PostForty Metres Of Silence: What Rome Will Not Let You Read
Took the run as proof of a specific hidden document and named three that it might be, none of which appear in any index we have seen.
Had that we did not
The 197 figure, correctly, and a clear account of how Modulo 7 works.
Left out
That the archive produces 41,000 items in thirty years and refuses 212, which is the fact that makes the 197 visible at all.
-
The Continental WireVatican Archive Says Conservation Programme Explains Refusal Pattern
Carried the vice-prefect's letter as the story and our count as the allegation, in that order, without checking the date he gives.
Had that we did not
Mons. Cerreto's statement in full, and earlier than we ran it.
Left out
That 104 of the 197 refusals predate the date in his own letter by up to fifteen years.
-
They Buriedthis newspaperWe Coded 41,000 Request Slips. Every Refusal Points At One Shelf.
Bought the distribution registers at a house clearance, collected three scholars' carbons, and paid two coders who were told nothing about what we expected to find.
Had that we did not
Both coders' full classifications, the 147 slips they disagreed on, and the fourteen months we do not have.
Left out
We coded 41,208 slips and named not one of the readers whose refusals we counted. Two of them told us a refusal ended a book. Neither is in this story, because a chart of 212 was tidier than three careers. — V. Ashcombe-Doyle, standards editor
Right of Reply
They Buried contacted Mons. Sabino Cerreto, vice-prefect of the Vatican Apostolic Archive on 6 August 2026, in writing and in Italian, with the full draft, both coders' classifications and the schedule of 212 refusals attached. Replied 19 August 2026. Printed in full, unedited, in our translation, which his office reviewed and did not amend.
I will answer your questions in the order you put them, and I begin by thanking you for putting them in writing and in Italian. That is rarer than you would think.
First, a correction of frame rather than of fact. This is a working archive of a sovereign institution. It is not the public record office of a state and it is under no legal obligation of access to anybody at all. It grants access anyway, without charge, to some fifteen hundred scholars a year, and it has done so continuously since 1881. Your figure of two hundred and twelve refusals in thirty years is close to ours. I will not confirm it to the unit, because I have not counted your way. I will say that in thirty years and something over forty thousand consultations we said no about two hundred times, and I would be grateful if you printed that sentence in the same size of type as the others.
On the sequence of shelving you describe. I recognise which units you mean and I do not accept your characterisation of them. Those units entered the conservation programme on 1 October 2009, following an assessment of their binding structures and of iron-gall corrosion in the text block. The programme is not fast. We do not have twelve conservators. We have four, and one of them is part-time, and the queue is measured in decades because the queue has always been measured in decades.
You ask why refusals from that sequence come back quickly. Of course they come back quickly. When a unit is in the programme the distributor does not walk to the shelf, because there is no purpose in walking to the shelf; the list is at the desk and it is consulted at the desk. I am genuinely puzzled that you present the efficiency of a clerk as though it were evidence of something. Would you prefer we kept the reader waiting an hour in order to look sincere?
You ask why the sequence does not appear on our published restoration list. Because that list is a list of completed interventions. Items appear on it when they leave the programme, not when they enter it. Nothing in that sequence has left the programme. I accept that this is not obvious from the way the list is titled and I have said so internally more than once.
You ask about the printed index. The index you are working from is a nineteenth-century instrument, and it is not a catalogue of holdings; it is a catalogue of what a particular prefect chose to print in a particular decade. Numbers not printed in it are not thereby secret. They are simply not printed in it. There are several such stretches. You have found one and built a building on it.
Finally, and I say this without heat, I would ask you to consider what your own figures mean. An archive that refused everything would give you nothing to count. It is precisely because we say yes forty thousand times that your two hundred noes are visible at all, and I would rather be an institution whose refusals can be counted than one whose refusals cannot. You will write what you write. For my part I take your table as an argument for opening more, not less. I have made that argument inside this house. I have not always won it.
Published unedited under our right-of-reply guarantee.
How was this story?
We publish the result, whatever it is. Reader verdicts appear on the front page and in our newsroom metrics.
Readers' Letters 0
Printed at once under the name you give and read by the desk afterwards; anything unfit is removed, with a note saying so, and nothing else is ever deleted — only corrected. Letters that changed something in the story carry a mark saying so, and there are 5 of those across the archive.