Thirty-Three Pension Files. The Same Eleven Minutes, Sworn In Nine Cities.
Twenty-nine of thirty-three disability boards recorded the same causative event on the same date, 28 October 1943, in the same nine administrative words. The men filed decades apart, before examiners who never saw each other's work.
A months-long reporting project. Documents cited below are held in The Vault and available to readers.
PHILADELPHIA, United States — Thirty-three men. One trial. Twenty-nine disability boards, sitting in nine cities between 1947 and 1981, wrote the same nine words on the same line of the same form.
Exposure incident to shipboard equipment evolution, nature not established.
Beside it, on all twenty-nine, the same date: 28 October 1943.
The men filed as much as thirty-four years apart. They filed in Philadelphia, Norfolk, Pittsburgh, Akron, Mobile, Wichita, Fresno, Camden and Providence. No examiner had sight of another examiner's work, because the adjudication system of the period had no mechanism by which he could.
We started from the hull number
The ship has a Navy list name, and that name carries eighty years of published narrative with it. Search on it and you get the narrative. You do not get the men.
So we searched on the yard's own hull number, 1738, and on the yard's 1943 muster returns, which are a builder's record rather than a service record and are indexed separately.
That produced a crew. Thirty-three names, with service numbers, for the men aboard on the day of the trial.
Twenty-nine of their disability-board files came from the federal records centre after 214 days and an appeal. Four came from families who had kept their father's copy in a drawer.
The paper trail is not the evidence
The identical phrase across twenty-nine files is a fact about a filing system. It could mean a clerk copied a form. On its own it proves nothing about a deck.
What is underneath the phrase is different. Each file contains the veteran's own sworn statement, taken under oath, at the time he filed.
We had all seventy-one of those statements stripped by a third party of names, dates, places, ship and unit. We then mixed them into a batch with sixty unrelated wartime exposure claims drawn at random from the same records series, and sent the batch to two forensic readers who were not told that any statement in it was related to any other.
Dr. Solveig Auberton-Reiss picked the set out on her second pass, before anyone told her there was a set to find.
Eleven details
The trial-crew statements converge on eleven specific things. The controls converge on almost none of them.
A green light in the air rather than on the water: twenty-seven of thirty-three, and none of the sixty. Cold coming up through the deck plates: twenty-six against two. Eleven minutes, timed by the deck watch: twenty-four against none.
Two of the eleven have never appeared in print anywhere we can find. Sixteen men describe an order to unclip and stand clear of all ironwork. Fourteen describe two boats standing off at about four hundred yards with their engines stopped.
A man filing in Fresno in 1971 cannot have read those details, because they were not there to read.
What the agency says the words mean
Lorne Kestrelby-Ahearn's reply is at the foot of this page, entire.
He begins by saying the thing this newspaper wants on the record too: these men were believed. Twenty-nine boards found for the claimant on the causative question, and not one of them wrote that the veteran was unreliable or mistaken.
He then describes the phrase as a routine catch-all, which is a reasonable position and would dispose of the whole matter.
And then, asked what it codes for, he cannot say. The 1946 adjudication manual that defined the causative phrases is not in the agency's retention series. Three staff looked for six weeks. The succeeding manual cites it by paragraph number and does not reproduce it. The national archives finding aid does not list it.
He also confirms, in writing, before we could ask him to: the phrase appears forty-one times in the whole indexed history of the agency. Twenty-nine of the forty-one are these files.
A catch-all that catches one thing is not a catch-all. It is a name for one thing.
What we are not saying
Prof. Ephraim Tallow-Nkemdi read the same batch and agrees the convergence is real. He does not agree about what convergence can carry, and his objection runs with this story: shared exposure to one unexplained event and shared exposure to one unexplained explanation can look the same on paper thirty years later.
Part of that we can answer, and we should say which part.
An explanation has to be given to a man, and a given explanation leaves traces — in print, on tape, in the questions other men get asked afterwards. Two of the eleven details leave none that we can find. The order to unclip and stand clear of all ironwork is not in the published account of this ship. Neither are the two boats standing off with their engines stopped. Sixteen men and fourteen men describe things that were not available to be told to them.
What the objection still reaches is the next thing along, and there we cannot help him or ourselves. We can show these men were describing something they were inside. We cannot show what it was, because nobody told them and the file does not say.
What we will say is what the record says. Thirty-three men were aboard Hull 1738 on 28 October 1943. Something happened to them that the boards accepted, paid for, and could not characterise. They described it independently, decades apart, in the same eleven particulars.
Two of the thirty-three were living when this reporting began. One died in March, aged 101. The other is 102, and his daughter answers for him. She asked us not to print his name.
The paper trail is thin, and the manual that would read it is missing. The testimony is not thin. It is thirty-three men, and this newspaper believes them.
The six o'clock edition
Three stories a morning, in your inbox before the coffee is made.
Sources & Method
The Navy list name for this ship carries eighty years of published narrative with it, and searching on it returns the narrative rather than the men, so we searched on the yard's own hull number and on the yard's 1943 muster returns instead. We then had every sworn statement stripped of names, dates, places, ship and unit by a third party, mixed with sixty unrelated wartime exposure claims from the same records series, and read by two forensic specialists who were not told that any statement in the batch was related to any other.
Who we spoke to
- Disability-board findings and sworn statements, 29 files, Federal records centre, veterans' adjudication series. Requested by yard hull number cross-referenced against the yard's own 1943 muster returns rather than by unit designation; released after 214 days and one appeal November 2025 to June 2026 Fee waiver refused. This newspaper paid $2,890 in search and copying charges.
- Four families, Daughters of three of the crew and the grandson of a fourth. Interviewed in person, each over two visits; each supplied their father's or grandfather's own retained copy of the board finding, which we photographed and returned February to July 2026 Three asked us to print the name. One asked us not to, and we have not.
- Dr. Solveig Auberton-Reiss, Forensic linguist; works on authorship and contamination in sworn testimony. Given all 71 statements stripped of names, dates, places, ship and unit, mixed into a batch with 60 unrelated wartime exposure claims and not told that any of them were related May to July 2026 Identified the trial-crew set as a set on the second pass, before being told there was one. Her report runs in full.
- Prof. Ephraim Tallow-Nkemdi, Psychologist; memory, contamination and convergence in group testimony. Given the identical stripped batch with no contact with Auberton-Reiss and no sight of her report May to July 2026 Agrees on the convergence. Disagrees sharply about what convergence can be taken to mean, and that disagreement is printed in his own words in the story.
- Lorne Kestrelby-Ahearn, Deputy Director, Adjudication Policy, Veterans Claims Adjudication Service. Two telephone interviews, the second recorded with consent, and a written reply to the full draft after twenty-two days July and August 2026 Printed entire under Right of Reply.
What we could not confirm
- The four files that break the pattern. Two use a different causative code entirely. One is water-damaged across the whole of the causative-event line and cannot be read. The fourth is missing its second sheet, which is the sheet the phrase is written on. We do not know what the fourth file said and it is wrong to assume it agreed with the other twenty-nine.
- What the phrase codes for. The agency has been unable to produce the adjudication schedule that defined it, and says the 1946 manual that would contain it is not in its retention series. The national archives finding aid does not list it either. Until that manual surfaces, nobody alive can say what the boards thought they were writing.
- What the apparatus was. Thirty-three men describe eleven things that were done to them, and not one of them was told what did it, so no statement in these files can say. The record that would say is an engineering record — a trial file, a fitting-out log, an equipment schedule for October 1943. The yard's custodian holds the muster returns we worked from and tells us it holds nothing else against hull 1738. Two federal custodians have been asked for the same three items: one says the series was not retained, and the other has not answered in 96 days. That is the door, and it is not the same door as whether these men are telling the truth.
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 Meridian TelegraphTHE CREW LIST EXISTS — AND THEY ALL SAID THE SAME THING
Printed the eleven details as a single continuous eyewitness narrative, in the present tense, with no indication that they come from thirty-three men who never met.
Had that we did not
The convergence, and the fact that the boards used one phrase. Both correct.
Left out
The four files that do not use the phrase, and the nine of sixty controls that produce the metal taste. Neither is fatal and both belong in the piece.
-
The Continental WireAgency Says Wartime Claim Code Was Routine Administrative Practice
Ran the agency's characterisation of the phrase as the story, in 340 words, on the day the agency offered it.
Had that we did not
The deputy director's quote, accurately and in full, which is more than we would have got had they not asked first.
Left out
The follow-up question. The agency's own index shows the phrase used 41 times in its history, and 29 of those are these files.
-
They Buriedthis newspaperThirty-Three Pension Files. The Same Eleven Minutes, Sworn In Nine Cities.
Worked from the yard's hull number instead of the Navy list name, assembled the files, and had the sworn statements read blind against a control set by two forensic readers.
Had that we did not
The eleven converging details with their control counts, the four files that break the pattern, and the agency's reply entire.
Left out
We have written 1,100 words about a records methodology and roughly 90 about what it was like to be nineteen and on that deck. Four families gave us their fathers' papers and got a chart. — V. Ashcombe-Doyle, standards editor
Right of Reply
They Buried contacted Lorne Kestrelby-Ahearn, Deputy Director, Adjudication Policy, Veterans Claims Adjudication Service on 29 July 2026 by telephone, and in writing on 1 August with the full draft and the file list attached. Replied in writing 22 August 2026. Printed entire and unedited.
I want to begin with the thing your draft gets right, because you will not enjoy the rest of this and I would like to be fair first.
These men were believed. That is not a modern courtesy applied retrospectively; it is what the boards did at the time. Twenty-nine boards found for the claimant on the causative question. Not one of them wrote that the veteran was unreliable, exaggerating or mistaken, and I have read enough of these files to tell you how unusual a run of twenty-nine like that is. Whatever else is in dispute here, the men's credibility was not in dispute then and it is not in dispute now.
On the phrase itself. 'Exposure incident to shipboard equipment evolution, nature not established' is a catch-all. That is what a catch-all is for: an event a board accepts as service-connected, in circumstances the board is not equipped to characterise, where the alternative is to leave the causative line blank and send a man away. I would say the same about a dozen other phrases from the period. It is administratively unremarkable.
You then asked me what it codes for and I have to give you the answer rather than a better answer.
The schedule that defined the causative phrases is the 1946 adjudication manual. It is not in our retention series. I have had three people looking for six weeks. It is not in the succeeding manual, which cross-references it by paragraph number without reproducing it, and it is not in the two supplements we hold. I asked the national archives and it is not in their finding aid. So I can tell you the phrase was standard usage, because it was, and I cannot tell you what standard it was usage of.
Your researcher then put to me the index count and I confirmed it, and I will confirm it again here in writing, because I would rather it appeared in your story from me than from a leak. The phrase appears forty-one times in the whole of this agency's indexed history. Twenty-nine of those forty-one are the files you have.
I do not think that means what your draft implies it means. A catch-all that catches one thing is a catch-all that was written for one thing and then not needed again, and boards in that period wrote local phrasing constantly. But I am not going to stand here and tell you forty-one is a large number. It is not.
Where I part company with you entirely is the inference. You have established that twenty-nine boards, over thirty-four years, characterised the same event the same way. You have not established what the event was, and neither the phrase nor the number of times it was used can establish it. A shared characterisation is a fact about the adjudication system. Your two readers have found a convergence in testimony, which is a fact about the testimony. Neither is a fact about the deck of that ship on the twenty-eighth of October, and you should say so more plainly than your current sixth paragraph says it.
One request, and it is not a small one. Two of these men were alive when you began. One is still alive. If your story runs and the follow-up finds his door, please remember that he filed in 1963, was believed, was paid, and has spent sixty-three years not being interesting to anybody. He is entitled to go on that way if he chooses.
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.