Research

Original word-data research from the PlainSpell editorial team. Every analysis is backed by reproducible queries against the same 6,918,744-entry Wiktionary corpus that powers the rest of the site. We publish source queries so anyone can verify or extend the findings.

  • 2026-06-24

    The Most Polysemous English Words

    Which English words carry the most distinct meanings? Short, ancient, high-frequency verbs dominate, draw, take, go, run, ranked by Wiktionary sense count, with the frequency-drives-meaning law that explains why.

    EnglishPolysemyFrequency

  • 2026-05-14

    Wiktionary Coverage by Language: How Five Living Languages Compare

    How the open-license multilingual lexicon distributes entries across English, Spanish, French, German, and Portuguese, and what the skew reveals about open-data linguistics.

    WiktionaryCross-languageOpen data

  • 2026-05-14

    Longest English Words by Misspelling-Variant Count

    Which long English words generate the most distinct misspelling variants? Ranked from algorithmically generated edit-distance variants of Wiktionary headwords.

    EnglishMisspellingsEdit Distance

  • 2026-05-14

    The Largest English Homophone Groups in Wiktionary

    Which English IPA clusters carry the most homophones, and what the distribution reveals about phonological economy in modern English.

    EnglishHomophonesIPA

  • 2026-05-14

    The Most Confusable English Word Pairs

    The pairs most likely to be confused are short, high-frequency function words separated by a single character. We rank them and explain why frequency times small edit distance is the recipe for confusion.

    EnglishConfusablesFrequency

  • 2026-05-14

    Which Letters Begin the Most English Words and Confusables

    S, P, and C lead the English word count; S, C, and B lead the confusable pairs. We trace the pattern back to Latin and Greek prefix clustering.

    EnglishLettersMorphology

How does the PlainSpell research program work?

PlainSpell research is produced in-house by the PlainSpell editorial team. We do not accept payment from any organization in exchange for coverage. Source data is open-license (Wiktionary CC BY-SA 4.0). Every page documents its methodology so independent researchers can reproduce or extend the analysis.

Suggestions for future research topics: hello@plainspell.com.