From: dgkilday57
Message: 65263
Date: 2009-10-16
>Are any of Whitney's nouns feminine? (Not that it would exclude the possibility of a fem. *k(^)ontu'-.)
> [...]
>
> I just assumed there would be some feminine oxytone nouns ending in *-tú-. Actually, Whitney's Sanskrit Grammar lists 8 nouns in -tu- with accent on the suffix (a few in zero-grade, others whose root vowel could equally be *e or *o, but I suspect *o since *e usually bore stress). But I agree *handuz could derive its -u- from the instrumental plural or dual cases.
> As for 'tooth', Gmc *tanþ-/tunþ-, note that Old Saxon <tand> (also OHG <zant>) seems to show evidence of Verner's shift (unless it was influenced by the alternate OHG form <zand>; OS should have *tâth or *tôth). I don't understand why Gothic has <tunþus> instead of <*tundus>, unless it was a mixed form from an original declension with *tánþ- in some cases, and *tunðú- in other cases, with mixture causing by-forms *túnþ- and *tanðú-. Perhaps the same mixture explains the OS and OHG forms <tand>, <zant>.A mixture as you suggest may well be necessary here. Gothic thorn is no problem, since Gothic levelled out grammatical change in strong verbs. It would have performed the same levelling on a noun paradigm.
> As far as the deciles, the conventional explanation for the presence of long vowels in some of them before <-ginta:> and <-konta> in Greek is that the *-d- was absorbed producing compensatory lengthening (so Sihler, 1995). Is this what you're saying is ad hoc? Yes, I've always found that explanation a little unconvincing (along with the supposed loss of *d- in <vi:ginti:>, <vim.çatí>, <eíkosi>, etc.)Yes, this loss of *-d(w)- is ad hoc, no matter how it is packaged, and Sihler should know better than to attribute comp. leng. to *-d- when it does not occur in <lapis> and the like. The long vowel is probably due to a laryngeal. No *-d- is required for <vi:ginti:> etc. where we have either *wei- 'apart', or a dual *wi:- 'the two which are apart', with analogical "restoration" of a nasal in Skt.
> I'm reluctant to agree that composite forms of *ken- or *k^em- plus *dH- explain 'hand'. To me this is like saying that 'foot' comes from *po- plus *d- ('that which gives (the ability) to go away or depart'? -- *(a)po 'from, away' plus *deH3- 'give'). I know that *komtus would probably be composite as well, but maybe we have to give up that hypothesis as the origin of 'hand'. Maybe 'hand' is just a lone wolf with its own form.My primary parallel for *ken-dH- is *wei-dH- 'to separate'; another example is *dHen-dH- 'to make hollow, gnaw'. The weakness is not postulating a composite; it is the poor attestation of *ken- as an adjectival root 'compact' vel sim.