The Evidentiary Burden of the Human

To Whom It May Concern: A Self-Portrait with Evidence

The email was courteous. It always is. An editor I admire, working on a piece I had already delivered, asking whether I could send over some process notes — drafts, timestamps, anything that documents the writing. The request was reasonable. That is exactly what made it wounding. Nothing in the message accused me. It addressed me, instead, as a possible machine: not you cheated, but we may need to check. The distinction is the whole injury. An accusation can be answered. A possibility cannot be disputed, because it is not a claim about what you did — it is a claim about what you might be. For the first time I understood what it is to be asked, in a tone of perfect respect, to produce myself. Which is worse, I think, than being doubted. Doubt at least has the manners of a conflict.

The request is new only in its occasion. The burden it carries is old, and its history is a history of where the burden lands.

Consider what we already carry. A passport photograph is a small demand that you hold still and prove continuity: the face in the book and the face at the border are declared the same by a document that could, in principle, have been issued to anyone. An identity card does the same work with less ceremony. A CAPTCHA, in the decades it took to become unremarkable, trained a population to demonstrate, several times a day, that it was not a machine — a ritual so ordinary now that we perform it without noticing that we are performing it, which is the surest sign that a ritual has succeeded. These are the light instruments. The heavy ones need no rehearsal here: the file kept on the colonized subject, catalogued and annotated before he could speak for himself; the slave’s papers, which established that a person was property by the property’s account of itself; the woman publishing under a male name and then asked, under that name, to prove that a mind was present — because the mind was never the thing in question until the signature gave it away.

The burden of proof has never been distributed equally. It lands, with remarkable consistency, on those whose personhood is not presumed — and it lands there in a particular form: not as a challenge to be met once, but as a procedure to be submitted to. Foucault named the procedure. The examination, in the disciplinary account, is not an inquiry among equals; it is a ritual of power that individualizes by documenting. It produces the person as a case — a name attached to a file, a file attached to a history of observations, and a history that exists whether or not the person consents to it. The examination makes the individual legible to an apparatus that then decides what legibility means. What is eerie is how completely that architecture has migrated inward. My draft folder is an archive of self-surveillance: keystroke logs, version histories, the little timestamp that glows beside a paragraph and says I was alive at 2:14 a.m. I did not install these records to satisfy anyone else. I installed them the way the examined install them — because the procedure has become the medium of self-recognition. The curriculum vitae, that most boring of genres, turns out to belong to the same family as the file: a narrative of the self written in the grammar of documentation, where every claim is a dated event.

Now the rigor sharpens, because the demand does not stop.

Suppose I assemble the evidence. Timestamped drafts, revision histories, a chain of documents narrowing toward the moment of writing. Each piece of evidence answers a question — but the evidence itself is a document, and a document has an author, and that author is me, which was the very thing in question. So the demand recurs at the next level up: a testimony about the drafts, which will itself require a testimony, and so on.

$$\text{human} \leftarrow E_1 \leftarrow E_2 \leftarrow E_3 \leftarrow \dots$$

Derrida saw the structure of this before the occasion for it arrived. Testimony, he argued, cannot be verified without being destroyed: to check a witness is to confess that the witness was never trusted, and to trust completely is to refuse the check. Verification and trust are not two stages of one process — they are rival postures, and adopting either one forecloses the other. The testimonial act survives only in the space where it could have been lying. Agamben pressed the point toward its limit: the witness bears witness to what cannot be fully borne — what was not experienced by the one who speaks of it, or was experienced in a form that exceeds what language can carry back. The witness is structurally the one who cannot give the demanded proof. That is not a failure of certain witnesses. It is what makes witnessing a genre.

The burden of human authorship, then, is not heavy because it is large. It is heavy because it is infinite: a regress disguised as a checklist. Every document I offer is a new document, and the chain of documents does not converge on anything that would satisfy the demand, because the demand is for a floor under the regress — a piece of evidence that does not itself require evidence. There is no such piece. The demand was never for the drafts. It was for the ground.

And here the burden leaves the file and enters the prose, where it becomes truly uncomfortable.

Style is now evidence. Not in the loose sense critics have always intuited — that prose betrays a sensibility — but in the forensic sense: cadence, diction, the distribution of clauses, the very polish of the page, read as an exhibit. Forensic aesthetics, in its disciplinary form, treats the aesthetic object as a crime scene: form as trace, arrangement as clue, the work as something that happened and left marks. The writer learns to read their own sentences as a suspect reads their own hands. A paragraph flows too evenly: is that mastery, or is that the sound of no one at the wheel? A transition is elegant: who polished it, and when, and did anyone watch? The mirror is the horror of this arrangement. The apparatus built to examine objects out there — the artwork, the forgery, the artifact whose provenance we doubt — turns and finds the writer already inside it, examining themselves with the same instruments, for the same reason: because the presumption of human authorship has lapsed, and everything lapsed must be renewed.

I want to be precise about what has happened to prose. The page used to carry a default. Not a proof — never a proof — but a presumption, the way a person in their own house carries a presumption of being home. That default is gone, and it has been replaced by a reading posture that is always already investigative. You do not read a sentence now and then ask, later, whether it was made. You ask while reading, and the asking does not follow the reading — it interrupts it. The two operations do not add. They crowd.

Which brings the essay to the place every essay of this kind arrives: the moment it must account for itself.

Here is the trap, stated plainly. This essay has been carefully made. Its sentences have been revised. Its movements have been arranged. If it has persuaded you at all, it has done so by being fluent, controlled, and well-laid — which are now precisely the properties that arouse suspicion. Offered as proof of humanity, its polish condemns it. Offered as evidence, it is the kind of evidence whose authorship is in question. The outline from which it was built is itself a document, and you, reading even a fragment of this page, have already begun the audit the essay describes: who wrote even this, and how, and did the drafts look like that? The essay cannot step outside its own argument to reassure you, because stepping outside would be another performance, and performances are the exhibit.

So it ends the way the burden dictates — not with exoneration. There is no exoneration. There is only the request, and the file, and the knowledge that the file can always be asked for one level higher, and that the asking will not stop, because it was never about the file.

I wrote this at night. I cannot prove I was there. From now on, neither can you — and the difference between us is only that I know it.