Revision C Removed The Word 'Experimental'. The Printer Kept All Four Drafts.
A jobbing lithographer in Canonsburg filed every proof for sixty years. Form SG-4 exists in four printed states, dated by the shop's own billing: two guardians become one, then a parent's right to take the child home is deleted.
Background written by the news staff. Contains no argument or recommendation.
CANONSBURG, Pennsylvania — Between 1966 and 1974 the form a parent signed at the door of a state training centre was reprinted four times, and each reprint took something out of it.
Revision A required two guardians to sign. Revision B needed one. Revision C deleted the word "experimental" in nine places. Revision D deleted the paragraph that let a parent take the child home.
All four survive because a printer in Canonsburg, Pennsylvania never threw a proof away.
The job file
Pfander & Sons, Lithographers, printed the intake stationery for the Bureau of Special Placement from 1961 until the account closed in 1981. The firm kept every proof, every plate order and every delivery note for sixty years, in flat files, in the basement, under the presses.
Marguerite Pfander-Osei, who owns the works now, let this newspaper photograph the lot. The SG-4 material runs to 61 sheets, all published here.
A plate order is a dull document, which is what makes it useful. The shop generates it for its own billing — a date, an ordering office, a purchase-order number, a quantity — and the customer never touches it.
The sequence is dated by the printer's accounts department, not by the Bureau.
The word
The Revision C order is dated 11 January 1972 and it is billed to "Special Placement — Developmental Services Branch".
The Revision B order, from 1969, carries the same street address and the same purchase-order series under a different name: "Special Placement — Experimental Services Branch".
The branch and the form changed the same word in the same quarter. The marked-up proof for Rev C is in the file, and the corrections are all one correction, made nine times.
Ordered twice, by two methods
Dr. Ines Halloran-Adeyemi was sent the four printed forms and nothing else — no dates, no ledger, no agency name — and asked to put them in order.
She works from watermark, fibre, brightener content and plate wear. Optical brighteners appear in that mill's book stock from about 1970, which splits the four into a pre-1970 pair and a post-1970 pair. Within each pair she orders by wear on the rule at the foot of the sheet.
Her order is A, B, C, D. It matches the billing file she had never seen. She is careful about her limits and we print them: the brightener line is a date range, not a date, and she would not defend it inside eighteen months.
What two lawyers say it is
Prof. Adaeze Rooke-Villanueva in Ann Arbor and Prof. Halvard Nyeko-Barrantes in Toronto were each sent the same four photographs, blind to each other, and asked one question: what is this instrument?
Both answered within nine days, and both used the same phrase. It is a transfer of guardianship. It is not a consent to treatment.
Here is the operative paragraph of Revision D, entire:
The undersigned, being the parent or lawful guardian of the above-named minor, hereby commits the said minor to the care, custody and control of the Bureau for the duration of the placement, and authorises such developmental training, care and correction as the superintendent shall direct.
No drug is named. No procedure is named. No duration is given, because "the duration of the placement" is set by the party receiving custody. And after May 1974 there is no clause anywhere on the sheet by which the parent can bring it to an end.
The printer's own index agrees with them, which nobody planned. Pfander files SG-4 in the SG series, alongside SG-1 (guardianship transfer, by order of a court) and SG-2 (guardianship transfer, voluntary). The shop never filed it with the medical forms. It was never in that drawer.
What a parent was actually signing
Plainly: from May 1974, a parent who signed SG-4 handed the Bureau care, custody and control of their child, for a period the Bureau defined, authorising training and "correction" that the form does not describe, on the signature of one adult, with no stated way back.
Before 1966 the same transfer would have gone in front of a judge. The form is how it stopped needing one.
The run
Two thousand, five thousand, five thousand, twelve thousand. Then two reprints from the unchanged Rev D plates, in 1977 and 1980, for nine thousand more.
The protections come out as the volume goes up. That is not an inference about motive and we are not making one. It is the shape of the delivery notes.
Everett Lindqvist-Baruah, who initialled Revision C, has answered at length and his reply is at the foot of this page. He says the word came out because the programme had stopped being experimental, and the second guardian because most of the children did not have one. He also says he did not read the nine changes individually, because there were nine of them and they were all the same word.
We asked him who drafted the amendment. He does not have the name.
Neither do we, yet. It sits behind a requisition index the successor office says was destroyed in 1989, and that answer is under appeal. Everything else on this page is dated, billed and delivered.
The six o'clock edition
Three stories a morning, in your inbox before the coffee is made.
Sources & Method
We were not looking for a printer. A 1974 delivery note surfaced in an unrelated tranche of stores paperwork, carried a letterhead, and the firm on the letterhead was still trading. We wrote and asked. Then we did the part that makes it worth anything: the four printed states went out to two family-law scholars who did not know each other, and to a paper examiner, with the ledger, the dates and the agency name stripped out — so that the sequence would have to be established twice, once from the billing and once from the paper itself.
Who we spoke to
- The Pfander & Sons job file, 1961–1994, Plate ledger, proofs, press-room dockets and delivery notes for the Bureau of Special Placement account. Photographed in the firm's basement over four days with the owner present; 2,140 exposures, of which the 61 sheets bearing on SG-4 are published entire with this story June 2026 Nothing was removed from the premises. The firm holds the originals and has been asked, in writing, not to destroy them.
- Marguerite Pfander-Osei, Owner, Pfander & Sons Lithographers; granddaughter of the founder. Approached cold by letter, then interviewed at the works three times and by telephone twice May–July 2026 Was not told what we thought the form was until after she had given us access, and has not asked us to withhold anything since.
- Prof. Adaeze Rooke-Villanueva, Family law, Ann Arbor; writes on the history of voluntary committal. Sent photographs of the four printed states with the job file, the dates, the quantities and the agency name withheld; asked one question — what is this instrument July 2026 Replied in nine days. Her opinion runs in full and she is paid the same fee as the second reader.
- Prof. Halvard Nyeko-Barrantes, Family law, Toronto; second reader, blind to the first. Given the identical four photographs under identical conditions, with no sight of the first opinion and no contact with its author July 2026 Agrees on the characterisation and disagrees at length about what follows from it. Both opinions are published.
- Dr. Ines Halloran-Adeyemi, Forensic examiner of paper and printing, twenty-two years. Given the four forms and no other material — no dates, no ledger, no context — and asked to put them in order July 2026 Works from watermark, fibre, brightener content and plate wear. Her report and her stated margins are published with this story.
What we could not confirm
- Who drafted the Revision C amendment. Every plate order carries a Bureau requisition number in a single series, and the series index is the document that would turn a number into a person. The successor office confirms in writing that the index existed; it says it was among forty-one linear feet destroyed under schedule in 1989. We have appealed that answer and the appeal is live. This is the next locked door and it is one door, not a wall.
- How many of the 33,000 sheets were ever signed. A print run tells you what a shop delivered, not what a parent put a pen to. The Bureau's own returns would answer it; we have asked for them twice, been refused twice, and the refusals do not say the returns are gone.
- Whether Revision A's two-guardian requirement was ever enforced in practice. We have not found a single completed Rev A form anywhere. What the printer kept is the blank.
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 FORM THAT TOOK THE CHILDREN: THE PRINT SHOP THAT KNEW FOR SIXTY YEARS
Turned a jobbing printer into an accomplice, on the strength of a delivery note with a signature on it.
Had that we did not
All four revisions, in the right order, with the quantities correct.
Left out
That the signature on the delivery note belongs to a Bureau storeman receiving forty cartons, and that a printer who reads the copy he sets would be an unusual printer.
-
Signal & LedgerJob 7712: What Sixty Years Of Kept Proofs Are Worth To A Records Case
A trade piece for the print industry on retention as an asset, using this file as the worked example. Genuinely good on how a plate ledger is constructed.
Had that we did not
The purchase-order series, explained more clearly than we managed, including why a shop numbers its own orders and the customer cannot renumber them.
Left out
What was on the form. The words 'guardianship' and 'child' do not appear in eleven hundred words.
-
They Buriedthis newspaperRevision C Removed The Word 'Experimental'. The Printer Kept All Four Drafts.
Photographed the whole job file, had the four states read cold by two family-law scholars, and had the paper stock dated by somebody who was not shown the file.
Had that we did not
All four printed states in full, both scholars' opinions including where they differ, the examiner's working, and the reply of the man who initialled Revision C.
Left out
We have built a story about children out of billing paperwork, and there is not one child in it. We asked the Bureau's successor for the list of centres this form was printed for, so that we could begin finding people, and when it said no we printed the plate orders anyway rather than wait. — V. Ashcombe-Doyle, standards editor
Right of Reply
They Buried contacted Everett Lindqvist-Baruah, deputy director of the Bureau of Special Placement, 1968–1979 by letter on 19 June 2026, then in person at his home on 4 July with the four printed states, the plate orders and both scholars' opinions; given six weeks and no deadline. Replied in writing on 3 August 2026 and asked that it run whole. Printed unedited.
I initialled Revision C. My initials are on the requisition and I am not going to spend my last years pretending otherwise, so let us start there and you can decide afterwards what it proves.
You are treating the word 'experimental' as though it were a confession that somebody scrubbed. It was not a confession. It was a status. A programme is experimental while it is under evaluation and it stops being experimental when the evaluation is complete and the thing is adopted as standing practice. Ours was evaluated between 1969 and 1971 and adopted. The form was then wrong on its face, because it described the placement as something the placement no longer was, and a form that is wrong on its face is a form your lawyers make you fix. I would change it again tomorrow.
On the second guardian I will be blunter, because your two professors have not been near a referral in their lives. Consider who was coming to us. In 1969, of the children referred to that Bureau, fewer than one in five had two lawful guardians available to sign anything at all. One was dead, or gone, or in a hospital, or had never been on the certificate. The two-signature rule did not protect those children. It kept them out. It meant a boy whose mother was willing could not be placed, and the alternative on the table for that boy was not his own bedroom, it was a court order and a locked institution with a different name over the door. We took a rule that was excluding the children who needed us most and we removed it. You may think that was wrong. Say why, and say what you would have done with the boy.
The fourteen days is the one I have thought about, and I have thought about it more than you would credit. It came out on clinical advice and the advice was not stupid. A course of training broken off in week six is worse than a course never started; the child goes home mid-way, unsettled, and comes back six months later further behind. That is what we were told and that is what I accepted. What I did not do, and this is the sentence you want, is ask what the paragraph was doing besides the clinical thing. I have had fifty years to notice that a clause which stops a parent removing a child is not only a clinical instrument, and I noticed it somewhere in the middle of them.
Your professors say the document transfers guardianship. It does not, and the word 'custody' in that paragraph is a term of art. It meant the superintendent could sign for a broken arm at two in the morning without telephoning three counties. Nobody was transferring anybody's child and no court would have read it that way. That is my position and I understand it is not yours.
You asked me who wrote the amendment. I did not write it. It came up from the general counsel's office with the branch's new letterhead already printed on it, and I will tell you the truth because at eighty-nine there is no percentage in the other thing: I did not read the nine changes individually. There were nine of them and they were all the same word.
You will print that and it will look like the whole story. It is a description of an ordinary Tuesday afternoon in an office that was trying to do something difficult with the people and the money it had, and if you cannot hold both of those in your head at once then you should not be allowed near a form either.
I would give you the name from the counsel's office if I had it. I do not have it. I have looked.
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.