Reconstruction talk:Proto-Germanic/bi

Etymology

"From Pre-Germanic *h₁pi, zero-grade of Proto-Indo-European *h₁epi, with voicing due to Verner's law. Cognate with Ancient Greek ἐπί (epí, “on, at”)".
Does anyone else find this a bit implausible? I know this has a reference but frankly I'm starting to mistrust Kroonen. I think this is better derived from a hypothetical preposition *bʰi, whence the Ancient Greek adverbial suffix -φῐ (-phi), the dative endings in *-bʰi seen in PIE personal pronouns, the instrumental/dative/ablative plural of PIE nouns, and the adverb *h₂m̥bʰi. It might have been then influenced semantically by a hypothetical Proto-Germanic preposition *ibi from Proto-Indo-European *h₁epi, perhaps even merging at an early stage in the language's development similar to how the Proto-Celtic prepositions *uss and *exs merge as Old Irish a. Dghmonwiskos (talk) 12:49, 3 July 2018 (UTC)

• Certainly, the currently proposed reconstruction does not sound very convincing. Nevertheless, we unfortunately lack sufficient data to reconstruct pIE *bʰi, because it's attested as an autonomous postposition /later preposition/ only in Germanic. I'll add that there were definitely pIE postpositions of the *bʰ- family, e.g. Balto-Slavic *ba (“for”) < pIE *bʰo, *be(z) (“without, besides”) < pIE *bʰe, which easily could have been extended with *-i in certain cases (it was a marker for dative/locative case). The problem is that we do not know what was the exact meaning and the relation between these postpositions. If we had more definite examples in other languages, I'm sure scholars would have reconstructed *bʰo(i) long ago, but unfortunately such have not been found so far. Bezimenen (talk) 17:07, 6 January 2019 (UTC)
  • The Proto-Indo-European Morphology chapter of latest Mate Kapović's book (p. 63) concurs with the things which Dghmonwiskos talks about. It seems the idea is not too far-fetched. Bezimenen (talk) 17:17, 6 January 2019 (UTC)