About these notes
A notebook that happens to be public, rather than a publication that happens to be personal.
Wordledger is a handful of pages about learning vocabulary in a second language. It was written for one reader — me — after the third or fourth time I rebuilt the same routine from scratch, having forgotten what I had already worked out. Putting it on a page seemed cheaper than working it out a fifth time.
The examples are English because English is the language the notes were made in. Nearly all of it transfers; the frequency figures and the specific collocations do not.
What is deliberately missing
- An account. There is nothing to sign up for. These are static files.
- An app. Any spaced-repetition tool will do, and arguing about which one is a much more enjoyable way to spend an evening than making cards. That is precisely the problem with it.
- Comments. A notebook does not need a comment section.
- Analytics and trackers. No scripts run on these pages at all. Nothing here needs them, and the site loads faster without.
- Precision I do not have. Where the research gives ranges rather than numbers, the text says so. Vocabulary studies measure different things under similar names, and the tidy figures that circulate are usually tidier than their sources.
Where the ideas come from
None of the substance is original. The main threads, for anyone who wants the primary material:
- Forgetting and spacing. Hermann Ebbinghaus, Über das Gedächtnis (1885) — the origin of the forgetting curve, and still surprisingly readable in translation. The modern experimental literature on the spacing effect is enormous and consistent.
- Retrieval practice and desirable difficulty. The work of Robert and Elizabeth Bjork on why conditions that slow acquisition often improve retention. This is the idea that most changed how I run a session.
- Vocabulary size, coverage and what knowing a word involves. Paul Nation, Learning Vocabulary in Another Language — the standard reference, and the source of the form/meaning/use framing used on the word lists page.
- Interference between semantically related words. A line of studies from the 1990s onward, associated with Tinkham and Waring among others, showing that thematic sets are harder to learn together than apart — the finding that contradicts how almost every beginner's list is organised.
For looking things up while reading, a good monolingual dictionary beats a bilingual one as soon as you can tolerate it, and Wiktionary is unreasonably good for etymology and for the senses that dictionaries prune.
Corrections
If something here is wrong — particularly if I have overstated a research finding, which is the easiest mistake to make in this area — I would rather know. There is no form to fill in; the domain is easy enough to guess an address at.