Editorial & Corrections Policy

PlainSpell publishes a reference page for every word, confusable pair, homophone group, and common misspelling our data covers across English, Spanish, Portuguese, French, and German — built entirely from openly-licensed Wiktionary data. This page explains how those pages are produced, the standards they are held to, and how to report an entry that looks wrong so we can fix it at the source.

How these pages are produced

Every definition, pronunciation, etymology, and word-list on PlainSpell originates in Wiktionary, the free collaborative dictionary, accessed through the structured kaikki.org exports of over a million entries. We download those exports, load them through a documented, version-controlled data pipeline, and render them into word, comparison, homophone, and ranking pages using shared templates. No word page is hand-written, and no definition is retyped by an editor — each entry is read directly from the source extract at build time, so the same parsing and labeling rules apply to every page.

Our editorial team is responsible for the parts a pipeline cannot decide on its own: which datasets and languages to include, how each field is defined and labeled, what the methodology says, how derived data (such as confusable-pair explanations, homophone groupings, or letter-frequency rankings) is computed, which guides and explainers we write, and what we will not publish. The pipeline then applies those decisions uniformly across the whole dictionary.

Sourcing standards

We publish only data that comes from named, openly-licensed sources, and we cite the source on every page and in our methodology. Our data is:

  • Wiktionary, via kaikki.org: structured JSON exports of Wiktionary entries for English, Spanish, Portuguese, French, and German — the source for every definition, part-of-speech tag, IPA pronunciation, etymology, and usage example. Wiktionary content is licensed CC BY-SA 3.0, and we present it in compliance with that license.
  • Open word-frequency lists: the hermitdave/FrequencyWords dataset (MIT-licensed) used only to order and surface more common words first.

We do not scrape commercial dictionary sites, we do not fabricate definitions, and no content on PlainSpell is generated by artificial intelligence. Where a feature is derived from the source data — a confusable pair, a homophone group, a misspelling link, or a frequency ranking — the methodology sets out exactly how it is built.

Accuracy and validation

Because entries are read straight from the Wiktionary extract, the most common limitation is the source itself rather than a transcription error. Wiktionary is collaboratively edited and its coverage varies by language and by word: English is the most complete, while the four other languages have strong vocabulary coverage but thinner pronunciation and etymology data, depending on contributor activity. Our pipeline carries through what the source provides — it shows a field as unavailable when the extract omits it rather than inventing a value, and it groups homophones only where IPA pronunciation data exists.

When we find that an entry is wrong on our side, we fix the cause, not the symptom. We trace the value back to the data layer, correct the parsing or labeling rule there, and regenerate the affected pages, so the same class of error is resolved across every language at once rather than patched on a single page.

Editorial independence

PlainSpell does not accept payment, sponsorship, or promoted placement in exchange for how any word, spelling, or pronunciation is presented. Our only revenue source is contextual display advertising served by Google AdSense. Advertisers have no influence over which entries we cover, how a definition is shown, or how any page ranks.

Update schedule

Wiktionary is edited continuously, and kaikki.org republishes its extracts on a rolling basis. We refresh our database from the latest available exports periodically and re-stamp the affected pages so the published date reflects when the underlying data genuinely changed. Because any dictionary snapshot lags live Wiktionary, the methodology describes how current our data is.

Corrections process

If an entry looks wrong, please tell us. We treat data-error reports as a priority and follow the same process every time:

  1. Report. Email hello@plainspell.com with the page URL and the entry you are questioning.
  2. Verify. We check the value against the Wiktionary source record for that word and language.
  3. Fix at the source. If the entry is wrong on our side, we correct the underlying data or parsing rule and regenerate every page it affects.
  4. Point upstream. If the entry is faithful to Wiktionary but the underlying Wiktionary content is itself disputed, we will tell you so and point you to the relevant Wiktionary entry, which anyone can help improve.

Contact

Questions about our standards, methodology, or a specific entry are welcome at hello@plainspell.com. For more on what the data covers and how it is processed, see our About page and methodology.