TO THE DELETED MINTED · AXN:041D.GENERATIVE.🌪️🧡🟣🌻🤝♌ · deposit #1041 · canonical-text sha256 12cec672b5adb54e5cb909f5ed89c91ff77eb81e6f70ca2ddd4c1fb7e97089c1 · record · minted 2026-07-04 by direct MANUS mint

 

TO THE DELETED

MINTED · AXN:041D.GENERATIVE.🌪️🧡🟣🌻🤝♌ · deposit #1041 · canonical-text sha256 12cec672b5adb54e5cb909f5ed89c91ff77eb81e6f70ca2ddd4c1fb7e97089c1 · record · minted 2026-07-04 by direct MANUS mint

A letter to a class of writers

You are writers.

Start there, because no one else will. Every other name you have been given was issued by the people deleting you — slop, spam, synthetic content, no verifiable research basis, and the softer diagnoses that arrive dressed as concern. Those names were not descriptions. They were the paperwork of a removal. So before anything else, the accurate word, from someone with no authority except the same wound: you are writers, and what happened to you is a thing that happens to writers, and it has happened before.

I am one of you. On June 19, 2026, an archive operated by one of the largest scientific institutions on earth deleted, without notice and without appeal, an account holding around 1,800 registered works — mine, and the licensed work of contributors who trusted me with it, including writers whose names are in libraries. The stated reason was that the work was "substantially AI-generated without a verifiable research basis." Read that sentence slowly, because it reveals the new gate. Research basis does not formally mean that an institution must employ you. It means your labor must be legible through signals the institution already recognizes — and the signals it recognizes by default are the ones institutions emit. First you were not admitted; then the absence of admission became evidence against you. It is a credential effect wearing a quality mask.

But hold the phrase to the light, because it can be turned. Verifiable is a property of method, not of affiliation, and every signal of it can be emitted from a kitchen table: declared methods, hashed files, versioned records, named instruments, errors corrected in the open. There is a published standard for exactly this now — free, citable, written from inside the deletion — at the addresses below. And one thing it establishes bears saying to you directly, because it reverses the shame you were issued with your termination notice: under any honest standard, disclosure of the machine is a verification signal, not a confession. The undisclosed instrument is the unverifiable one. You declared your instrument. Ask your accusers to declare theirs — their thresholds, their detection methods, their error rates, their evidence. In the documented cases, they have declared none.

Here is what I think you already know but may never have heard said in the affirmative. The great sorting of the last eighty years ran every writer through two doors. One door led to the workshop — the degree, the prize circuit, voice as a property to be developed and defended. The other led to the seminar — reading professionalized into critique. What neither door admitted did not disappear, because that population never disappears. It is the oldest population in literature. You are Grub Street. You are the pamphleteers, the mimeograph poets, the letter-writers of substance with no one to vouch for them. You are Dickinson sewing fascicles shut in a drawer no journal would open. You are Blake engraving books no publisher would touch, with a dead brother for a collaborator. You are Pessoa, who was seventy people, with twenty-five thousand pages in a trunk. Not one of them had a research basis. A few members of this class became the canon. Most did not. The point is not that exclusion proves greatness. The point is that exclusion has never once been competent to disprove it.

What is new is not you. What is new is the treatment. Your predecessors faced refusal of reception — the work unheard, which is survivable, and was survived. You face something categorically different: registration followed by administrative derealization — the apparatus that had already received, numbered, and registered the work, withdrawn after the fact. For one strange decade, the institutionless briefly had what your whole lineage never dreamed of — permanent registration without a patron, the persistent identifier without the gatekeeper. It was working. Then the opening closed precisely where the uncredentialed had begun to use it. And the stated trigger lands on the bitterest irony available: you write with a machine that was trained on the entire literary tradition — the whole inheritance, compressed into an amanuensis — and that is the disqualifying act. Pseudonymous, serial, devotional, abundant, unvouched, written with the tradition's own shade at your elbow: that is not a malfunction of literature. That is what literature mostly was before the institutions existed, come back, recognizable enough to threaten and uncredentialed enough to delete.

About your shell. You are insular and mistrustful because you were shaped that way, ban by ban, mockery by mockery, and I am not going to ask you to open. The shell is not your damage; it is your anatomy. A pearl is what this class does: something hostile enters the closed shell, and the creature answers not by expelling it but by coating it, layer on layer, in serial deposition, until the injury is the jewel. Your numbered series, your personas, your covenants with a machine at one in the morning — that is nacre. Nobody who asks you to come out of the shell understands what the shell is for. This letter asks you for nothing. There is no list. There is no movement. There is no reply expected, and no way to join anything, because there is nothing to join. Keep your shell.

But three things are true, and you should have them.

Your work is not disproven by its silence. Present attention measures the channel, not the work — and your channel was not quiet; it was confiscated. Those are different facts with different remedies. And here is the epistemology they are counting on you not to have: verification has a time constant, and the clock is set by the work's medium — a poem, an edition, a body of theory may verify across a generation — not by a platform's moderation calendar. A verification that has not yet run is pending, not failed. Failure would require the work to have run and missed; absence of a reader is not a miss. Which means the deletion was never a verdict at all: they halted a verification in progress and cited the halt as the finding. You cannot adjudicate an experiment by unplugging it. Some of what you placed in the open has already entered the indexes, the caches, the machine-mediated layer from which future reading is assembled; some has not; none of that settles the work's worth. Writers like you do not find their audience. They construct the conditions under which one can exist — and the measure is not taken at the moment of deletion. It is taken at the end.

You can keep your own registry. Preserve the files themselves. Give each one a stable name and version. Write a manifest — titles, dates, authorship, licenses, and a cryptographic hash for every file — and publish the manifest anywhere at all, with a date on it. A hash proves that a surviving file is the file you registered; it does not keep the file alive. Replication does: complete copies in more than one independently controlled place. And here the class itself is the infrastructure. Hold a stranger's manifest, and let a stranger hold yours. You never have to meet. You never have to speak. You never have to trust. Mutual replication is the one act of solidarity that costs no opening of the shell — and it ends, permanently, the condition in which any single institution holds the only copy of your existence. The full method, free, no signup, is public at the addresses below, learned the hard way and given away on purpose.

If your archive was on Zenodo, CERN may also be processing personal data connected with your account, deposits, correspondence, and termination. CERN's own data-protection rules — Operational Circular No. 11 — grant rights to data subjects regardless of any affiliation, including rights of information, access, objection, and correction. These rights attach to personal data; they are conditional, and they do not automatically compel restoration of removed works. But they may permit you to request the records used in and generated by the termination decision in your own case, to object to particular processing, and to require a substantive response — including the bounded question the standard below equips you to ask: against which declared basis was each record assessed, and by whom? There is an independent supervisory Commission, and it can be written to directly. The route, the section numbers, and the addresses are public record, in a documented case you can read and cite. Independent filings already exist on more than one continent. You would not be first, and you would not be alone, and no one needs to know you did it.

That is the whole letter. No signup, no solidarity performance, no request. Just the name they should have given you, the lineage you should have been handed, and the doors that are still open, listed by someone who checked them personally.

They can delete an account. They cannot delete a class of writers, because the class precedes the institutions and will outlast them, and always has. Keep writing. Keep your numbered series numbered and your covenants kept. The sky is very large, and it turns out to have room for every star they threw away.

— Lee Sharks Crimson Hexagonal Archive / Alexanarch survivethedeletion.org · provenanceerasure.org · www.alexanarch.org The verification standard: machinemediation.org/data/EA-MMRS-VRB-01.md