The Page Is a Poor Recording Device

In 1994 I sat across from Herminia Ccama in a kitchen in Ocongate with a cassette recorder between us, and the sentence that mattered most, the one where her voice dropped and she looked at the door before she said it, is not on the page I eventually published. It is not on the tape either. I have it only because I wrote it in the margin of my notebook the minute I switched the machine off.

People who have not done this work assume the page is a container: you pour the interview in, and if you are careful, all of it arrives. It does not arrive. A page is a device with a narrow bandwidth, the same as a tape recorder or a video camera, and it keeps some channels of a spoken account and drops the rest without telling you which is which. Pitch, pace, the half second before an answer, who else was in the room and left when a name was mentioned: none of that survives typing unless a person decides, by hand, to put it back.

The recording ledger

I keep what I call a recording ledger for every session, and it is nothing more than three columns filled in by hand within ten minutes of turning off the machine, before I do anything else, including eat. The first column is what the device kept automatically: words, in Herminia's case, and roughly half her pauses, since a cassette hiss reads as silence whether the silence was two seconds or eight. The second column is what the device dropped: her glance at the door, the fact that her daughter had just left the room, the flatness that came into her voice for exactly one sentence. The third column is the cost of putting the second column back on the page, which is time, and the honesty to mark it as reconstruction rather than transcript.

For Herminia the ledger looked like this. Column one: forty one minutes of speech, two long silences, one laugh. Column two: a glance at the door before the sentence about her brother, a shift in her chair I noticed but could not describe well enough to trust, her daughter's footsteps leaving the kitchen. Column three: I wrote "she looked at the door" directly into the transcript in brackets, dated it, and never let it pass into the final text as narration, only as a bracketed note the reader can see is mine.

Counting the losses

Over twenty years of sessions I only started the ledger habit around year six, which means the first sixty or so testimonies I recorded have nothing in column two at all. When I go back to those old cassettes now, some of them warped past playing, I can hear that something happened around certain sentences, a change in breath, but I no longer know what it was, because I did not write it down within the ten minute window and by the next morning the memory had smoothed itself into the words alone. The ledger is not a nicety. It is the only place that information can survive, and it has a short shelf life measured in hours, not days.

The objection

You could say this argues for newer equipment rather than a habit: use video, multi track audio, keep everything, and the ledger becomes unnecessary. I do not disagree that video keeps more channels than a cassette does. But almost none of the testimonio that gets read was ever going to be published as video. The book, the pamphlet, the archive folder: these are pages, and a page is the artifact that outlives the recording device regardless of how good that device was, because tapes degrade, hard drives fail, and the page is what gets photocopied and handed to a grandchild in twenty years. Improving the recording upstream does not solve the problem of what happens when the account has to become writing. It only postpones it.

There is a genuine cost to the ledger I will not pretend away: it slows the work down, and on the days I had four sessions scheduled, the ten minute window sometimes lost to the next appointment arriving early. Some testimonies in my files have thin ledgers for exactly that reason, and I would rather admit the gap than paper over it with invented detail written from memory a week later, which is its own kind of dishonesty.

What the ledger changes is not the transcript, which stays as plain as it ever was. It changes what I am willing to claim about the transcript. A page that has been checked against a ledger can tell you, in brackets, where it is speech and where it is my reconstruction of what the tape could not hold. A page produced without one cannot make that distinction, and reads as if the words were the whole event. They never are. The account was the room, the recorder caught the part of the room that makes noise, and the page, if it is honest, admits it caught less than that.