Intervalio · v1 · Dutch

Learn the words
that do the work.

A vocabulary trainer built around two findings the research keeps confirming: practice retrieval, and meet new words in real sentences.

Start practicing →Browse sets

Abstract

Intervalio drills the highest-frequency Dutch words inside real example sentences, scheduled so each word returns just before you would have forgotten it. The approach combines Pimsleur’s graduated interval recall for short-term consolidation with FSRS for long-term scheduling. Active retrieval (being forced to produce a word from memory) builds retention that passive review and last-minute cramming do not.

Why high-frequency words

A small set of words does most of the work.

Word frequencies follow a steep distribution: a relatively small number of words appear over and over, and the long tail is rarely encountered. Knowing the most frequent words first is the most efficient way to start understanding speech and text.

≈ 80%

of written text

covered by the first 1,000 word families (Nation, 2006).

≈ 95%

of spoken language

covered by 2,000–3,000 word families (van Zeeland & Schmitt, 2013).

≈ 1 in 20

words unknown

the inference threshold at 95% coverage; the rest you fill in from context.

In Nation’s analysis of English corpora, the first 1,000 word families cover roughly 78–81% of written text, and the second 1,000 add another 8–9% on top. Combined, the first 2,000 word families account for around 86–90% of running words in writing¹.

For everyday spoken language, van Zeeland and Schmitt estimated that a vocabulary of 2,000–3,000 word families is associated with roughly 95% lexical coverage, which they propose as the threshold for adequate listening comprehension². Higher comprehension targets (98% coverage) require a much larger vocabulary of 6,000–9,000 word families¹.

II.

Methods

Two schedulers, working at different timescales.

Two algorithms run in parallel during a practice session, each addressing a different problem.

In-session

Pimsleur's graduated interval recall.

When you miss a word, it returns later in the same session at increasing offsets. The principle, set out by Pimsleur in 1967, is that recalling a word just before you would have forgotten it strengthens the memory more than re-reading does, and each successful recall lengthens the interval before you forget again³. The original schedule begins at five seconds and grows logarithmically out to years.

Across sessions

FSRS (Free Spaced Repetition Scheduler).

Between sessions, scheduling is handled by FSRS, an open-source algorithm based on the Difficulty–Stability–Retrievability model with parameters fit to large-scale learner data⁴. Each card is shown again at the moment its predicted recall probability drops below ~90%, which keeps reviews efficient without letting items slip into forgetting.

The combination is grounded in the testing effect: in Roediger and Karpicke’s experiments, repeated retrieval produced substantially better long-term retention than the same time spent re-studying⁵. Cramming feels productive in the short term and underperforms on delayed tests; spaced retrieval is the opposite.

III.

Practice

Say the words out loud.

Producing a word, by speaking it, is reliably more memorable than reading it silently. MacLeod and colleagues named this the production effect and replicated it across modalities; the benefit is largest when the learner speaks the word themselves rather than hearing someone else say it⁶.

The practical recommendation is simple: read each Dutch prompt and your answer aloud, and read the example sentence aloud once you’ve seen it. Don’t worry about accent. The gain comes from the act of producing speech, which encodes the word through an additional channel and helps with later recall.

IV.

Material

Words live in sentences.

Webb showed that the quality of the surrounding context affects how well a learner acquires the meaning of a new word, sometimes more than the raw number of encounters does. Informative contexts produce larger gains in form-meaning knowledge than impoverished ones⁷.

Every Intervalio card pairs the target Dutch word with a real example sentence drawn from Tatoeba’s CC-licensed corpus, with the English translation revealed only on demand so you can attempt to infer meaning first. Frequency ranks come from SUBTLEX-NL, a 44-million-word corpus of Dutch film and television subtitles, which approximates everyday spoken Dutch better than book corpora do⁸.

Scope

What's here today, and what isn't.

v1 is deliberately small. The current release is Dutch only, with the first 200 words translated and reviewed by hand against SUBTLEX-NL frequency rankings. Each word is paired with vetted example sentences from Tatoeba.

More words and more languages will follow, in that order, if there is demand. Spanish has a placeholder slot and the scheduler and persistence layers are intentionally pluggable, so adding a language is mostly a data exercise rather than a rewrite.

If you want to see this expanded, the most useful signal is showing up and practicing.

Try a session.

Ten words takes a few minutes. Your progress is saved locally in your browser.

Start practicing →

References

1.Nation, I. S. P. (2006). How large a vocabulary is needed for reading and listening? Canadian Modern Language Review, 63(1), 59–82. link
2.van Zeeland, H., & Schmitt, N. (2013). Lexical coverage in L1 and L2 listening comprehension: The same or different from reading comprehension? Applied Linguistics, 34(4), 457–479. link
3.Pimsleur, P. (1967). A memory schedule. The Modern Language Journal, 51(2), 73–75. link
4.Open Spaced Repetition. Free Spaced Repetition Scheduler (FSRS). Algorithm and academic references including Su, J., Ye, J., Nie, L., Cao, Y., & Chen, Y. (2023). Optimizing spaced repetition schedule by capturing the dynamics of memory. IEEE Transactions on Knowledge and Data Engineering. link
5.Roediger, H. L., & Karpicke, J. D. (2006). Test-enhanced learning: Taking memory tests improves long-term retention. Psychological Science, 17(3), 249–255. link
6.MacLeod, C. M., Gopie, N., Hourihan, K. L., Neary, K. R., & Ozubko, J. D. (2010). The production effect: Delineation of a phenomenon. Journal of Experimental Psychology: Learning, Memory, and Cognition, 36(3), 671–685. link
7.Webb, S. (2008). The effects of context on incidental vocabulary learning. Reading in a Foreign Language, 20(2), 232–245. link
8.Keuleers, E., Brysbaert, M., & New, B. (2010). SUBTLEX-NL: A new measure for Dutch word frequency based on film subtitles. Behavior Research Methods, 42(3), 643–650. link

Set in DM Serif Display and DM Sans. Example sentences from Tatoeba under CC BY 2.0. Frequency data from SUBTLEX-NL. No claims here are stronger than the cited literature; if you find a stat misrepresented, please flag it.