Why Does PGmc *smalaz (from PIE *(s)mal-) have "a" (not "o")?

  • 4
    Why should it have *o? Long PIE (usually < eh2) becomes PG , but short *a generally remains *a. Apr 12 at 8:33
  • @JanusBahsJacquet Why Does Proto-Slavic have long ā malъ? Apr 12 at 9:04
  • 1
    That’s a much better question – not one I’m sure anyone knows the answer to (but I’m not a Slavicist!). Derksen chooses to reconstruct a root *(s)meh₁- with a (presumably diminutive) -lo suffix. He posits o-grade for the Sl. form, and e-grade for Gk μῆλον and OIr míl, but zero-grade for Gmc… which of course doesn’t work, because *sm̥h₁-lo- would become **sumla. Others have argued that the root is *(s)melH-, which works for Gmc, but less so for Gk and OIr, and not for Sl. at all. Perhaps they’re from separate roots, or perhaps there was a metathesis in Gmc… Apr 12 at 9:18
  • 1
    @JanusBahsJacquet zero-grade smh̥1-lo would give the correct Germanic reflex, and it's more typical for the coda of the root to become syllabic in zero-grades
    – Tristan
    Apr 12 at 9:42
  • 1
    that said, the reconstruction of this particular root is definitely problematic
    – Tristan
    Apr 12 at 9:43

1 Answer 1


Proto-Germanic lacked a robust distinction between the *o & *a vowel qualities.

At an early stage, the two qualities merged, with the short phoneme always being reflected as *a, and the long & overlong phonemes always as *ō and *ô. Later a new *ā developed from earlier *aja sequences, and borrowings, but it was quite restricted.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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