Abstract

The approximately 7,000 languages of the world is currently divided into no less than 422 lineages (= families + isolates) by the orthodox evidential criteria of Glottolog (glottolog.org).

Should we believe this number?  To what extent is it subjective, consistent in meta-properties and dependent on the amount of research and documentation? The time-depth of the families so far established do not reach more than 10,000 years back in time whereas human language arguably goes back at least 300,000 years (probably further).

The inherent rate of linguistic change puts a limit on how deep linguistic data alone can probe, but are there combinations using evidence from other disciplines and general geospatial considerations that may take us further?