IE, Genetic Data, Languages of Anatolia

[Peter &/or Graham]

Miguel asked about the Latin canis with its irregular /a/

I read Peter Schrijver's vast book today on the reflexes of PIE laryngeals
in Latin.   He suggests a delabialisation of *o to a in Latin in certain
contexts, including after /u/.   His suggestion for canis is therefore:

      Accusative PIE *k'uon-m  >  *kuonem  >  *kuanem  >  kanem.
The /a/ forms then generalise throughout the paradigm, just as /k/
generalised, replacing /kw/ in vox, vocare etc.

He also notes the PIE word (s)kenH "young dog"  (I haven't had time to check
it)   This would give the Latin form can-.

Not totally convincing, but food for thought!

Peter