What’s Out There
← Back to swiping
Language & textFreeAcademic

WALS (World Atlas of Language Structures)

Max Planck Institute for Evolutionary Anthropology

Does a language put verbs first? Have tones? A grammar map of 2,600+ languages answers it.

A database of structural features of the world's languages (sound systems, word order, grammatical gender, and more) compiled from descriptive grammars by 55 authors. Each value records how one language behaves on one feature, and every language has coordinates, so you can map grammar the way others map rainfall.

Access

Cost
FreeFree web access and downloads.
License
Open licenseCC BY 4.0
Sign-up
No sign-up
How you get it
Bulk downloadCSV, CLDF

What's in it

One row is
One row per language per feature value
Covers
Languages worldwide, each with a point location
Years
2013 edition (version 2020.x on Zenodo)
Updated
Irregular
Size
192 features across 2,662 languages (sparse matrix)
languagefamily and genuslatitude and longitudefeaturevaluesource

Quality

4 of 5

Expert-compiled and widely cited, but the language-by-feature matrix is mostly empty and largely frozen since 2013.

  • Most languages are coded for only a few features
  • Feature sample languages differ by chapter, complicating comparisons
  • Little updating since the 2013 edition

Connects to

Questions it could answer

  • Are languages with tones clustered in particular regions?
  • How often do verb-object languages also put adjectives after nouns?

Similar data sets

  • Glottolog

    A stable ID and family tree for every known language and dialect, plus 471,000 works describing them.

    Arts & cultureMax Planck Institute for Evolutionary AnthropologyFree

  • Universal Dependencies Treebanks

    Sentences in 150+ languages, each word tagged and linked to its grammatical head the same way.

    Language & textUniversal Dependencies consortiumFree

  • Ethnologue: Languages of the World

    The standard count of the world's living languages, with speakers, status, and location for each.

    Arts & cultureSIL GlobalFree tier, paid upgrades

  • Corpus of Contemporary American English (COCA)

    A billion words of American English, balanced across genres from TV scripts to academic journals.

    Language & textEnglish-Corpora.org (Mark Davies)Free tier, paid upgrades

  • OpenAssistant Conversations (OASST1)

    161,443 chat messages written and ranked by volunteers in 35 languages, with no model in the loop.

    Language & textOpenAssistant (LAION-affiliated volunteers)Free

  • Multilingual parallel text corpora for East African languages

    English text translated into five East African languages, for training machine translation.

    Language & textMakerere University NLP groupFree