Memory Is Collective Before It Is Personal

Cirila Quispe described the day of a landslide near Lares in detail so vivid I assumed she had watched it happen. She had not. She was four years old and asleep in another village, and the account she gave me was assembled entirely from what her aunts had told her at every family gathering for the following thirty years.

I did not learn this until I asked a routine question, whether she remembered the sound of the mud, and she paused and said she thought her aunt had described a sound, not that she herself had heard one. This should unsettle anyone who treats a first person account as evidence that the speaker experienced the thing described. Cirila was not lying. She was telling me a memory that had always been collective, and I had almost published it as though it were hers alone.

The borrowed memory ledger

I now keep what I call a borrowed memory ledger for testimonies involving childhood or events a speaker was too young, too far away, or too unconscious to have witnessed directly. For every vivid detail, I ask a version of the question I asked Cirila: did you experience this, or did someone tell you this, and if someone told you, who, and how many times had you heard it before you began to feel it as your own. The ledger does not aim to strip these details from the account. It aims to mark, honestly, where a detail entered the speaker's memory from outside, since a memory assembled from repeated family telling is a different kind of fact than a memory laid down by direct experience, even when it arrives in the same confident first person sentence.

Cirila's ledger, once filled in, showed that of the eleven vivid details in her landslide account, only two came from her own limited experience that day, waking to her mother crying. The other nine, the sound of the mud, the neighbour who ran for help, the exact number of houses lost, all traced back to her aunt Herminia's retellings at family gatherings, repeated so often across three decades that Cirila herself no longer experienced them as secondhand.

Why this is not a debunking

I want to be clear that the ledger is not a device for catching people in error. Cirila's account is not less true for being collectively assembled; it is a completely accurate record of what a family told itself about a disaster for thirty years, which is its own real subject, arguably a more interesting one than a single eyewitness account would have been. What changed after the ledger was not whether I trusted Cirila. It was how I wrote the account: instead of "I remember the sound of the mud," the published version reads "my aunt always said you could hear the mud before you could see it," attributing the detail to its actual source inside the sentence, which costs nothing in vividness and gains a fact about how memory moves through a family that the first version erased.

The disagreement

Critics of this method argue that constantly flagging which details are secondhand undermines the emotional authority of testimony, turning a woman's account of her own family's disaster into a hedged, footnoted document that reads as less hers rather than more honest. I understand the worry and I have seen it happen in poorly done versions of this method, where every sentence gets an attribution and the text becomes unreadable. My answer is that attribution belongs at the level of the detail, not the sentence, and most of an account can stay in plain first person while the two or three details that carry the weight of collective memory get named for what they are. The alternative, letting every vivid detail read as eyewitness regardless of its actual source, produces cleaner prose and a quieter kind of dishonesty: it lets a family's thirty years of retelling disappear into one woman's borrowed certainty, as if she alone had stood in the mud.

What the ledger gave back

The strange result of filling in Cirila's ledger was that she did not experience the published version as a diminishment. When she read the line crediting her aunt Herminia with the sound of the mud, she added a detail I had not recorded anywhere: that Herminia had told the story so often precisely because she had felt guilty for not reaching the village in time, and the retelling was itself a kind of penance passed down as narrative. That detail exists only because the ledger had already made visible whose voice a given image belonged to, and once that ownership was named on the page, Cirila had somewhere to attach a second, truer layer of the story that an unbroken first person account would have had no room for.