Sources
Where the material on this site comes from.
A working-out of the question whose words are these, actually? Where a specific source is identifiable, it's credited. Where the site draws on a wide consensus of Dutch teaching material, the consensus is named rather than any one source reproduced.
A note on attribution
This is a personal study project, built by a learner from the materials a learner ends up looking at: textbooks, published wordlists, academic lexicons, and the small handful of public-domain reference works that have aged into reusable form.
None of those materials is reproduced wholesale. The intent is to credit honestly, and to lean on public-domain and openly-licensed material wherever it covers the same ground. The CEFR level assigned to each word is the author's own judgement — not a direct copy of any published list, but a reading of what the framework actually requires of a learner at each level.
Vocabulary & CEFR labels
The Dutch words on the site draw on a wide range of study and teaching material — including the Numo NT2 woordenlijst, the NT2Lex CEFR-graded academic lexicon (Tack et al., UCLouvain), the vocabulary chapters of Nederlands in actie (Coutinho), and the publicly distributed Staatsexamen NT2 reference materials. No single list is reproduced; these sources informed the word selection but were a starting point, not the final word.
The CEFR level assigned to each word is the author's own call. Rather than averaging labels across references — which can disagree at the margins — we read what the CEFR framework actually requires a learner to do at each level, and decided which words belong there on that basis. If you're used to a different list, you may find some words placed a level higher or lower than you expect. That's a deliberate choice, not an oversight.
Levels are a rough guide for choosing what to practise next, not a placement and not a certification — see About / The name.
Example sentences
The example sentences attached to each word and verb form were generated by a large language model (Anthropic's Claude), prompted to produce CEFR-level-appropriate Dutch with a natural English gloss. They are then reviewed for accuracy and level fit before entering the active list.
They aren't quotations from any human-written corpus. They're working examples — the kind a teacher might write on a whiteboard.
Etymology & decomposition
The way each word breaks into prefix / root / suffix is hand-curated. A rule-based morphological splitter proposes candidate decompositions; the author reviews each one and keeps the ones that are pedagogically useful.
The decompositions are not derived from any third-party etymology database. They aim to be useful learning aids rather than philologically definitive — see About / Etymology for more on the distinction.
Expressions & proverbs
The 500 expressions, idioms, and proverbs on this site are sourced from Nederlandsche Spreekwoorden, Spreekwijzen, Uitdrukkingen en Gezegden by F. A. Stoett, 4th edition (Zutphen, 1923 – 25), via the Digitale Bibliotheek voor de Nederlandse Letteren (dbnl.org).
Stoett died in 1936; the work has been in the public domain in the Netherlands since 2007 under the standard life-plus-seventy term. The etymology summaries on the site are paraphrases written by the author, drawing on Stoett's entries.
Illustrations
The small images that accompany some words were generated with Midjourney from prompts written by the author. They're not stock photography and not drawn from any third-party catalogue.
Images aren't on every word — they're added where a picture actually helps disambiguate a concrete noun. Abstract words are left to do their own work.
Pronunciation audio
Every Dutch word, verb form, and sentence example on the site is paired with synthesised speech using the Microsoft Azure AI Speech Service, accessed under a paid licence that permits use of the synthesised audio in a published product. The voice is nl-NL-ColetteNeural, a neural Dutch (Netherlands) voice from Microsoft's Text to Speech portfolio.
The audio is generated once and cached — your browser fetches a pre-built MP3, not a live API call. Nothing about what you're practising is sent to Microsoft's servers at practice time.
Typography & lexical data
The site is set in EB Garamond (by Georg Duffner and Octavio Pardo) for the body and Inter (by Rasmus Andersson) for the interface. Both are released under the SIL Open Font License 1.1 and are self-hosted from this domain — no third-party font CDN is contacted, and pages render the same in regions where Google's font servers are unreachable.
Open Dutch WordNet (Postma et al., 2016) is used in the project's curation pipeline to surface candidate near-synonym pairs for the differences feature; the published entries themselves are author-authored. ODWN is distributed under CC BY-SA 4.0.
If something here is yours
If you recognise material on this site as belonging to you and feel the credit is wrong, missing, or misleading, please get in touch and it will be corrected or removed promptly. This is a study project, run in good faith.
You can reach the author at cefr.nl@outlook.com.