I think if you compare *waithanjan to PG *burþinjō (“a burden”) and OGH Lua error in Module:parameters at line 110: The parameter "land" is not used by this template., it actually isn't that strange. Although, *waithinjan may be a better reconstruction.
It still begs the question why OHG has a class 2 weak verb, though. I mean, if that's what's actually attested, we should probably reconstruct it. Besides, French had no verb chass in -o- so it had to improvise on that.
I don't know. That may be true in stressed syllables, but the prefix ga- was unstressed so it may have given a different outcome.
Last edit: 02:37, 26 January 2016
If gaaiginōn came into Old French via Old Provençal, the hard /g/ could have survived.
But how could it have entered Old Provençal without it being Frankish in origin? Old French is kind of... in the way between them. Of course it's possible that the word came from Alemannic instead. But is it not possible for the g- to have survived? After all if garir has g-, why can't this word?
Ok, I see. But still... if w > g survived, then g itself could have survived too. Are there any examples Germanic words with g being borrowed into French with j?
When I say /w/ > /g/, I actually mean /w/ > /gu/ > /gu/, /g/. A really good example is gay, which came into Old French from Frankish as jai, but was displaced by gai via Old Provençal, which was from Gothic gaheis (gaheis), both ultimately from the same PGm source.
I'm guessing this is related to the oddities in the second paragraph of gain#Etymology 3? If French gagner > older gaaignier > Gothic gaaiginōn (gaaiginōn), then any /w/ > /g/ mutation seems irrelevant. And what of Spanish ganar?