Against the Written Version as Authority

A university archive once asked me for "the definitive text" of a testimony I had recorded, edited, and published, as though the printed booklet were the original and the woman who gave the testimony were a source I had consulted along the way to producing it.

This gets the relationship backwards, and it is worth being direct about how completely backwards. The printed booklet is not the testimony. It is one recording of the testimony, made with the least faithful instrument available, a page, and produced last, after the cassette, after my notes, after her own memory had already reorganised the event at least twice by the time I sat down to write.

The authority stack

I rank every version of an account on what I call the authority stack, ordered not by how polished a version is but by how close it sits to the actual living of the event and how much of the original channel survives in it. At the top sits the speaker's own memory, which no device has ever touched and which continues changing after every recording is finished. Below that sits the audio recording, which fixes words and tone at one specific moment but nothing before or after it. Below that sits my notes taken during and immediately after the session, which add gesture and room but filter everything through my attention and what I happened to notice. At the bottom sits the published page, which fixes a single, permanent version of choices made at every level above it, and which most institutions treat as though it sat at the top.

The inversion is not accidental. A page is citable, shelvable, reproducible at no cost, and searchable, and every one of those qualities is an institutional convenience rather than an evidentiary one. Archives prefer the bottom of the stack because it behaves well as an object, not because it knows more than the woman it was made from.

What this changes in practice

When a researcher writes to me asking to cite "the authoritative version" of a testimony I published twenty years ago, I now write back with the full stack: the speaker's name, whether she is still alive and could be asked directly, where the original cassette is archived, and only then the page reference, with a note that the page reflects choices I made at a particular moment and should not be read as her final or fullest account. Several researchers have found this response unsatisfying, wanting a clean citation rather than a genealogy. I understand the frustration and I do not think it entitles anyone to a false simplicity.

I once revised a testimony I had published a decade earlier after the speaker, Delia, then in her seventies, told me a detail she had withheld in the original session because her mother in law was alive at the time and would have been implicated. The published page from a decade prior was, at the moment of its printing, the fullest honest version available. It stopped being that the moment Delia told me more, and treating the old page as permanently authoritative would have meant privileging a stale artifact over the living woman it was supposedly a record of.

The counter worth taking seriously

The case against this position is that scholarship needs fixed points to argue from, and a stack that can be revised whenever a speaker says something new makes citation unstable in a way that undermines the reliability historians and researchers depend on. I do not disagree that instability has real costs, and I do not revise published testimonies casually or often. But the alternative, treating the page as though it were more stable than the person, only hides the instability rather than resolving it. The living woman was always the moving target. Pretending the page had caught up with her, once and permanently, was never true. It only felt true because paper does not visibly change, and a person, obviously, does.

What I now put on the page itself

Since Delia, I have added a short line to the front of every testimony I publish, naming the date of the recording session and stating plainly that the speaker's own account, if she is living and asked, supersedes the printed text. It is a small thing to add and it costs the book nothing in polish. It does mean the printed page can no longer quietly stand in as the last word on a life, which is exactly the function I no longer want it to perform, because the woman it describes is still out there, still capable of remembering more, still the actual top of the stack whether or not an archive finds that inconvenient to catalogue.