Timed myself properly for the first time — a plain unfamiliar essay, a stopwatch, no re-reading allowed — expecting to confirm a number I’d been quietly assuming about myself for years. The actual number came in well under what I’d assumed, by enough that I went back and checked the stopwatch.
The gap is explained, I think, by the difference between the books I remember finishing quickly, which were mostly ones I already partly knew or that were written in a plain, undemanding register, and the average of everything I actually read in a year, most of which is neither. Memory keeps the fast exceptions and quietly discards the slow ordinary case, so the self-image drifts upward over time without any single wrong data point causing it.
Small practical upshot, connected to the honest inventory in this note: I’ve started putting an actual estimated time next to each post here rather than trusting a reader’s — or my own — guess about how long something will take. It’s a minor feature. It only exists because the guess turned out to be wrong enough, in my own case, to be worth not repeating on other people.