Kretschmer identified both Illyrian and Messapic as satem languages due to the change from IE o to a.

Illyrian language - Wikipedia

That sounds like complete nonsense to me. How does the argument go?

How does he know that PIE 'o' always changed into 'a' in Illyrian? I have studied Croatian toponyms quite a bit, and a generally accepted etymology is that Colapis, the ancient name for the Kupa river, meant "winding water", that "col" (winding) comes from *kwol (an ablaut of *kwel, meaning to turn), so, if that's true, not all 'o'-s changed to 'a' in Illyrian.

And even if we take for granted 'o' changed to 'a' in Illyrian, how does it follow that Illyrian is a satem language?

  • 1
    How do you know that Kretschmer had the beliefs attributed to him on Wikipedia a half century after he died? Did you find an actual article written by Kretschmer: or are you asking for a bibliographic source, e.g. "what are the writings of Paul Kretschmer?".
    – user6726
    Mar 19, 2023 at 20:03
  • 2
    That entire Wikipedia article is really, really poorly written. A lot of it contradicts itself, and some parts don’t make any sense at all. If its representation of Kretschmer’s identification is accurate – and given the state of the article, that’s a big if – then it’s bunkum, yes. The change of o to a may have some level of correlation with the centum/satem split, but it is most certainly not a reliable predictor of whether a language is centum or satem. Mar 19, 2023 at 23:08
  • certainly if that is Kretschmer's argument it's bogus. An o > a shift is also found in Germanic, Tocharian, and Anatolian. Preservation of the distinction between o & a is only found within centum (specifically Greek, Latin, and Celtic), but the two shifts clearly do not always coincide
    – Tristan
    Mar 20, 2023 at 9:19
  • cp. Old English collenferhþ "proud," (full of ferhþ), German Quelle "well, source", Freudenquell (offspring), or differently the Waal, Wacal in Roman sources. Use the source, Luke. The idae that *kʷol- is generally agreed on has little merrit if hardly anyone knows anything about it for lack of attestation while Roman scribes are known to hear what they want to hear.
    – vectory
    Mar 23, 2023 at 23:01
  • To the opposite however, "roman. /ǎ/ [...] in slav. /o/" in loans, cf. pòreč (Bichlmeier, Sava). See also Labe, interlape etc.
    – vectory
    Mar 23, 2023 at 23:02

1 Answer 1


I assume Kretschmer has Albanian in mind when he speaks of Illyrian. Albanian presents the shift PIE *o > a: *nókʷts > natë etc., and it is a satem language.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.