I am looking for a dataset describing (nearly) all languages, and where they are spoken.
Bonus if the areas are linked to Geonames or another mainstream geographical dataset. Polygons are also OK.
It does not need to be a 1-1 mapping, as many places have several languages.
Official languages is OK, but actual usage (more than 1% of the locality's population has it as its mother tongue) of non-official languages would be even greater.
I am looking for sub-country granularity.