Thread:User talk:CodeCat/Could you check this over/reply (10)

I suspect that Old Irish must reflect Proto-Celtic, as a short -ei- would be reflected as é, later ía in Old Irish. So this would then have to come from. On the other hand, the -th- is palatalised, which is indicated by the preceding -i- in the spelling, and it's possible that palatalisation inhibits the é > ía change. You would have to ask someone more knowledgeable like User:Angr about it.

Germanic can't reflect, as this can't result in a short -i- in zero grade: *wih₁- would give a long ī. I'm less certain about the outcome of a zero-grade *wh₁i-, as presumably this would turn the preceding semivowel into a full vowel: *wh₁i- > *uh₁i- > *ūi-. On the other hand, it's also possible that it remained as a semivowel, in which case the expected result would be *wh₁i- > *wi-, with the laryngeal simply lost.

Lithuanian has an acute accent, which must reflect either a former laryngeal or the "new long grade" of Balto-Slavic. It can't reflect, as a Lithuanian y always reflects an "old" long vowel, generally from a laryngeal. A full grade -ey- gives -ei- or -ie- in Lithuanian. This also means that it probably can't reflect, which would not account for the length, although I have no idea what -eh₁i- gives in Lithuanian. Balto-Slavic developed a new type of long grade ablaut though, which could have been introduced in a zero grade form: *wh₁i- > *wi- > *wī-.

Latin ī must reflect -eyh₁-, -ih₁- or -ey-, so that rules out and its zero grade. The former would give if I'm not mistaken.

So while some possibilities can be ruled out, I can't say anything conclusive. It would be a good idea to take this to the ES.