On Editing Someone Else's Testimony

Rosa Mamani ran a small radio archive in Sicuani for eleven years, editing testimonies from elderly campesinos for a local anthology, and she had a rule I have never seen written down anywhere else: nobody's testimony left the building until she had read the edited version aloud to them, in the room, and watched their face while she did it.

I visited her archive twice while researching my own methods, and the second visit changed how I edit permanently. Most testimonio editors treat consent as a signature: you show the subject a printed page, they nod or sign, and the editing is considered approved. Rosa considered a signature almost worthless as a check on the actual work, because a person looking at a printed page of their own words, especially an older person who did not grow up reading fluently, often nods out of politeness or fatigue rather than genuine agreement, and the page cannot show them what changed, only what remains.

The return pass

What Rosa did instead, and what I now call the return pass, is this: after editing a testimony down from a recording, she read the final version aloud to the speaker, at the speaker's own pace, stopping wherever they wanted her to stop, and she watched specifically for the moment their face registered that a sentence sounded wrong to them, not factually wrong, but wrong the way a word in someone else's mouth sounds wrong even when the meaning survives.

I watched her do this with a man named Ignacio, eighty one years old, whose account of a 1982 drought she had trimmed from ninety minutes to about twenty for the anthology. Partway through the return pass Ignacio stopped her at a sentence describing his cattle dying "one by one, over three weeks." He said that was not how it felt. Rosa asked what was wrong with it, since the fact was accurate; it had taken three weeks, and the cattle had died one at a time. He said the sentence made it sound organised, like a list, and it had not felt like a list, it had felt like waiting for the next one and not knowing which day. Rosa changed the line to "we kept waiting to see which one would go next," a sentence with less information in it and, by Ignacio's own judgment, more truth.

What a signature cannot catch

A written approval would have passed Ignacio's original sentence without incident, because nothing in it was false. The return pass caught something a signature structurally cannot: the difference between a sentence that is accurate and a sentence that matches how an experience actually sat in the person who lived it, which is a distinction only available to someone listening in real time to the person's reaction, not reading their nod.

Rosa told me the return pass added, on average, another full session to every testimony in her archive, sometimes two, since some speakers wanted multiple sentences reworked and some wanted to hear the whole thing twice. Over eleven years and roughly ninety testimonies, that is close to a hundred extra sessions of unpaid listening time, absorbed entirely by her, never billed to the small grant that funded the project.

The objection

You could argue the return pass does not scale, that a larger project with hundreds of testimonies and a real deadline cannot afford a second full session per subject, and that Rosa's method, admirable as it is, only survives because her archive was small enough to carry the cost. I think this objection is honest and I do not have a way around it that does not involve money or time that most projects, including several I have worked on, simply do not have. What I take from Rosa is not that every project must run a full return pass. It is that a signature was never doing the job we told ourselves it was doing, and knowing that changes what counts as thorough enough, even when a full return pass is not possible. At minimum, I now read the most consequential sentences aloud, even under deadline, because Ignacio's cattle taught me that accuracy and truth are not the same test, and only one of them can be checked from a face.