Entry 214.3, branched from 214, written 12 September 2026
The forgetting curve is a filing problem
Ebbinghaus tested himself on lists of nonsense syllables in the 1880s and watched most of each list disappear within a day. What brought a list back was meeting it again, at intervals, before it had gone completely.
So a notebook that only stores things works against memory. The useful ones put an old note in front of you at the moment you need it, which is what filing was always for. See 41.2 Spaced repetition and 12.4 Attention residue.
Indexed under Ebbinghaus, forgetting, memory, timing