From: dgkilday57
Message: 63846
Date: 2009-04-18
>This can't be compensative lengthening at all; see for example 'hope', nom. sg. <elpis>, gen. sg. <elpidos>; 'shield', nom. sg. <aspis>, gen. sg. <aspidos>; these two show that accent has nothing to do with it. Likewise Lat. <pe:s> can't have comp. leng.; see 'stone', nom. sg. <lapis>, gen. sg. <lapidis>. Greek and Latin must have either inherited long vowels in the nom. sg. of 'foot' or independently lengthened them for some reason other than compensation.
> --- In cybalist@yahoogroups.com, "dgkilday57" <dgkilday57@> wrote:
> >
> > Doric does have <po:s>. Attic <pous> has probably been remodeled after the participle <didous>. That is also the usual explanation for Attic <odous> where other dialects have <odo:n>. As for the vowel-assimilation from *edo:n, we have the same thing in <orophe:> 'roof' from *eropha:, cf. <erepho:> 'I cover'.
>
> Doric has different rules of compensatory lengthening (either because its short e and o had a quality different from their Ionic counterparts, or there was a stronger pressure of the already existing system with lower e: and o: ). It seems that the question is: Why is there ANY compensatory lengthening in this word? Does Doric Auslaut behave differently?