From: stlatos
Message: 71045
Date: 2013-03-07
>Hmmm, he doesn't use laryngeals either. You'd better just throw his book away.
>
>
> --- In cybalist@yahoogroups.com, "stlatos" <sean@> wrote:
> >
> > --- In cybalist@yahoogroups.com, "dgkilday57" <dgkilday57@> wrote:
> > >
> > > --- In cybalist@yahoogroups.com, "stlatos" <sean@> wrote:
> >
> > > > There's no reason to expect a regular change for either above; quadru- shows tr > dr , dr > tr in taeter , maybe vitrum , and definitely:
> > >
> > > In _quadri/u-_ we may have contamination with an old word for 'whetstone', *quadrum, replaced by _co:s_; cf. OE _hwaet_ 'keen, bold', OS _hwat_ 'id.', OHG _(h)waz_ 'sharp, rough, severe', ON _hvatr_ 'bold, vigorous'; OE _hwettan_, OHG _wezzan_, ON _hvetja_ 'to whet', from PGmc *xwatjanan. Whetstones are generally square and *quadrum could have passed into the meaning 'square', whence _quadra:re_ 'to make square', and squares have 4 corners.
> > >
> > That is certainly a possibility that I would never have thought of. There's a reason for that.
> >
> Note that Pokorny's root-shape *kWe(:)d-/*kWo(:)d- (IEW 636) is impossible.
>
>Whether you like it or not, PIE *kWo(:)- loses its labialization in Germanic (Moeller, PBB 7:483, 1882).
>There's no clear ev. for it; I don't believe it. What is clear is that * kYwo didn't, which is what you claimed. If * ixwa-z isn't good enough for you, how about *xwaitya- > hweiti OIc; hwæ:te OE; wheat E; weizzi OHG; ? That's another word whose root shows t/d alt. regardless of whether the "root-shape... is impossible".
>We are dealing with what Piotr calls a quasi-Narten root with PIE *-a:- in the strong grade, *-a- in the weak (pace scholae Lugdunensis Septentrionalis). The best example is *swa:d-/*swad- 'sweet', whose derivatives violate Lubotsky's Law in the RV. The root in question is *kWa:d-/*kWad- 'sharp'. All of Pok.'s good citations come from one or the other in Gmc.
>So, the ev. for -a- not -o- is in the dia. in which a/o can't be distinguished?
> The Skt. verb and Lat. _triquetrus_ are false friends.What origin for triquetrus lets your exp. of quadrus remain possible?
>
> > > I will accept (Sabino-Latin) dialectal -dr- > -tr-, also -lb- > -lp-, -nd- > -nt-, etc. (The latter explains _scintilla_ beside pure Latin _scindula_.)Then which Latin words that have -dr- are from IE -dr- ?
> > >
> > > > uter utri- = water-skin L; hudría = water pitcher G;
> > >
> > > Possibly Sab.-Lat.
> >
> > So all Latin words that would have -dr- were influenced by Sab.-Lat.; none survived?
>
> I never said any such thing.
>In my view this devoicing of word-internal media + resonant goes along with the aspiration of word-int. tenuis + res. in _lachrima_, _sulphur_, _mamphur_, _lympha_, _triumphus_ and a few other words which entered the literary language from Sabino-Latin. This peculiarity of the Sab.-Lat. dialect is to be attributed to unconscious hypercorrection when the first large group of Sabines acquired Latin (under the first Appius Claudius a.k.a. Attus Clausus ca. 500 BCE). The Sabine language itself, like Oscan, underwent voicing of word-int. tenuis + res. as we see in _Publius_, _publicus_ against pure Latin _Poplius_, _poplicus_. Sabine, like South Picene, also fortited /l/ in certain positions toward [d], so the Sab.-Lat. lateralization of /d/ in _lachrima_, _lympha_, and _le:vir_ also comes from hypercorrection.
> > >You're giving all that hypercorrection to create alt., but no simple opt. change can ever do so?
> > > > utur U; water E;If uter shows clear tr, I don't need to worry about how clear the other is; it wouldn't be my only ev. for -dr- > -tr- in that root.
> > >
> > > Mere graphy, since _utur_ was written in the Etruscan-based alphabet, with _t_ doing double duty for /t/ and /d/.
> > >
> > That's why I included uter at the same time.
> >
> Clarity is not one of your strong suits.
>
> > > > (showing it's Italic), and either dhr > tr or something odd in trahere, among some that are unclear.What vowel-shortening before /h/ is attested?
> > >
> > > I derive _trahere_ from *treh2-g^H-.
> > >
> > What reg. rule would you assume for hg() > g() ? Nothing I've seen looks reg.
> >
> Zero-grade *tr.h2g^H- > *tra:h- followed by vowel-shortening before /h/.
>