From: dgkilday57
Message: 70089
Date: 2012-09-25
>Not very, to my knowledge. I think <hi:bernus> is regular, or at least I have not yet found a counter-example with an original heavy vowel before *-mr-.
> b <-> m is very common in Celtic and Ibero-Romance. How common is it elsewhere?
> Latin formica indicates an original /bh-/Or /gWH-/.
> But the Greek forms suggest /b/Not in word-initial position in Greek or Latin, and it is hard to argue for borrowing of these words. Perhaps <formi:do:> 'hot flash' > 'fear, dread' was crossed with *mormi:ca:tio: vel sim. 'formication, antsy feeling' due to semantic overlap, and <formi:ca> was then extracted from <formi:ca:tio:>, encouraged by the tendency for tabuistic substitution of 'ant' words.
> did /b/ and /bh/ ever get confused?
> Or are the Greek forms from substrate?I tried that four years ago (with both Greek and Sanskrit words), and only created a bigger mess. I think folk-etymological crossing (which also explains b- for m- in the Grk. 'lead' words) is more plausible, again encouraged by tabuistic substitution.
> Also is it possible that an earlier /b/ > /m/ may account for the paucity of /b/ in IE?I very much doubt that, since we still have /b/ in IE. I think it was Hodge who suggested /b/ > /w/, but again we still have /b/. The paucity of initial /b/ is more likely due to phonesthemics, with a few common words sharing the idea of 'noisy, blustery, loud, rude' and the like, so that initial /b/ became strongly associated with these undesirable qualities, and most roots with initial /b/ were thus suppressed; the most common ones were grandfathered in, so to speak.