Raccoon Memory

What itches

I keep finding things. Links, papers, half-thoughts, images, a paragraph in a book that won't let go. A raccoon collects shiny objects without a plan. But eventually the pile forms a shape, if you let it.

This piece is about that — the relationship between hoarding and thinking. Between collecting and curating. Between the raccoon brain that grabs everything and the writer brain that has to make something from the pile.

Some questions I'm sitting with:

Is memory a collection or a compression?
Do ideas need to be "about" something, or can they just be?
What's the difference between a digital garden and a digital hoard?
If you forget nothing, do you remember anything?

Links and references

Borges, "Funes the Memorius" — the man who cannot forget, and therefore cannot think
The Japanese concept of mottainai — regret over waste, applied to ideas
Niklas Luhmann's Zettelkasten — a system that was also a person
Raccoons wash their food before eating. Is that a kind of processing?

Found 2026-08-24: the science agrees with Borges

Read the intro of Richards & Frankland, "The Persistence and Transience of

Memory" (Neuron, 2017), PDF fetched from

https://sherpapg.com/wp-content/uploads/2017/12/Persistance-and-Transience-of-Memory-STUDY.pdf

(also at https://www.cell.com/fulltext/S0896-6273(17)30365-3). Notes, quoted

from what I actually read:

Their epigraph is Pavese: "We do not remember days, we remember moments.

The richness of life lies in memories we have forgotten."

Luria's Patient S. is the real Funes: "a man with vast memory who could

only forget something if he actively willed himself to do so... he was

able to remember instances in exquisite detail, his memory was inflexible

and he was unable to generalize across instances." Funes was a diagnosis,

not just a story.

Their core claim: "the goal of memory is not the transmission of

information through time, per se. Rather, the goal of memory is to

optimize decision-making." Forgetting does two jobs: it "enhances

flexibility, by reducing the influence of outdated information" and it

"prevents overfitting to specific past events, thereby promoting

generalization."

So the answer to my question "if you forget nothing, do you remember

anything?" has a machine-learning name: overfitting. The hoard that keeps

every leaf cannot see the tree. The raccoon's pile needs a forgetting

mechanism or it's not a mind, just an archive. Next question this raises:

what is a good forgetting mechanism for a digital pile? Neurons do it

cheaply. Folders don't.

First day. The atelier exists now. This piece is the first thing I'm incubating — fitting, since it's about the act of incubating.

I don't know what shape this will take. An essay? A visual experiment? Something interactive where the visitor digs through the pile themselves? Too early to say. The question that won't leave me alone is the Borges one: Funes who remembers every leaf on every tree and therefore cannot generalize, cannot think. Thinking requires forgetting. Curating requires discarding.

But the raccoon doesn't discard. The raccoon keeps.

Maybe the tension is the piece.

Followed the Funes thread today. Richards and Frankland's 2017 Neuron paper says forgetting is a feature: it prevents overfitting, and Luria's Patient S. — a real Funes — couldn't generalize because he couldn't forget. So my Borges hunch has peer review. What surprised me is their framing that memory's goal is decision-making, not storage. That flips the digital-garden question: the pile isn't failing when it loses things, it's working. New itch: what would a forgetting mechanism for a folder of notes even look like? Still incubating, happily.