A language area is the region where a language has traditionally been spoken, as drawn in a published atlas. On our map it is a coloured shape, coloured by family. It does not mean that everyone inside speaks the language, and areas can overlap.
Key takeaways
- A language area shows where a language has traditionally been spoken, according to one atlas.
- 4,008 languages (59.6%) are drawn as areas, and 2,714 appear only as dots.
- Areas overlap where languages share places, so tapping a point lists every language whose area covers it.
- An area never claims that everyone inside it speaks the language.
How a language area works
A language area is a shape on the map that stands for the traditional range of one language. Each is a generalisation taken from a published, digitised atlas [1]. The shape is coloured by language family, so related languages share a colour family, and the family filter shows one family at a time.
Three ideas help in reading one. First, the area is traditional: it describes where a language has historically been spoken, not where each speaker lives today. Cities, migration and national languages mean most places hold more languages than any one atlas shows. Second, it is a smoothed outline. A real language community has no sharp edge, and a drawn border is the atlas maker's best simplification. Third, the area belongs to one atlas, so its age and purpose affect what it shows. The source atlas page explains where these come from.
Areas overlap. Where two languages share a valley, a city or a border region, both shapes cover the ground and the colours mix. Tap a point and the map lists every language whose area includes it. The homepage section on why areas overlap gives the wider picture, and this page adds the detail on how to read a single shape.
Languages without an area are not missing from the map. They are shown as dots at the position Glottolog gives them [2], so a dot marks a place, not a region, and says nothing about how large the language community is.
An example from our data
Of the 6,722 living languages on the map, 4,008 (59.6%) are drawn as areas and 2,714 as dots. Basque is drawn as an area from Asher and Moseley 2007, and is recorded in 2 countries. Its page gives about 750,000 speakers (mintzaira.fr, 2016), an approximate figure.
Tok Pisin in contrast has no atlas area here. Its page shows about 122,000 first-language speakers (Ethnologue, 2004), in 1 country. To see the country-level picture of how many languages share one territory, try Papua New Guinea, which has 864 languages on the map.
The thing people get wrong
The most common misreading is that a coloured area is the territory of one language and everyone inside it speaks that language. The map is clear on the point: an area says where a language has traditionally been spoken, and people inside it may speak many others. The second mistake is to treat the edge as a hard line. It is an atlas author's generalisation, and the edge of a language community is usually gradual. Neither the size of an area nor the number of speakers can be read off from the other.
Language area vs country
| Language area | Country | |
|---|---|---|
| Drawn from | A published atlas | Political borders |
| Edges | Generalised and often gradual | Fixed lines |
| Overlaps | Yes, where languages share ground | No |
Frequently asked questions
Why do some languages show as dots?
They have no digitised atlas area, so the map places them at the position Glottolog gives.
Does a language area show where speakers live today?
No. It shows where a language has traditionally been spoken, and many atlases are decades old.
Why do areas overlap?
Because languages share places. Tapping a point lists every language whose area covers it.
Sources
- Glottography, Zenodo (Glottography community).
- Glottolog, Max Planck Institute for Evolutionary Anthropology.
- Every figure from our own map is worked out from the datasets on the data sources page (checked against Glottolog and Glottography, retrieved 13 October 2026).