Re: v > ?

Vasiliy Chernov <bc_@...> wrote:

> On Tue, 11 Dec 2001 13:24:45 -0500, Aidan Grey <grey@...>
> wrote:
>
> [...]
>
> >   Stops:  p t k' [k] kw
> >               [b] d g' [g] gw
> >              bh dh g'h [gh] gwh
> >
> >   fricative: s [th]
>
> Sorry, what do you mean by fricative [th]? "Brugman's spirants?" There
> seems to be a very simple explanation for them.
Yes, Brugmann's [thorn] is as dead as a dodo.  And /T/ merged with /s/
in Exilic Quenya anyway, though the stay-at-homes in Valinor preserved
it
(this change has thus been dubbed "The Shibboleth of Feanor").

> >   laryngeals: h1 h2 h3
> >   (and does anyone have any resources on recent laryngeal theory?)
>
> Oh, I wish it existed :(
>
> Too much argument around them. Just for example, Starostin maintains that
> laryngeals are OK only where Hittite has {X} or there's a secondary
> aspiration elsewhere (e. g. in Indo-Iranian). He insists that the
> Saussurean "sonantic coefficients" in general don't correspond to Hittite
> {X} and have no consonantal parallels in the rest of Nostratic.
I have seen claims that there *are* correspondences,
e.g. PIE *-h2 ~ Proto-Uralic *-k.

>       He also
> thinks one laryngeal is perfectly enough.
This is also Szemerenyi's position.  One laryngeal that corresponds to
Hittite /X/ or otherwise unaccounted aspirations in Indo-Aryan and
Greek,
and has nothing to do with Saussure's coefficients.
Szemerenyi reconstructs it as /h/.  However, I think there is evidence
for three laryngeals.

>       - While some multilaryngeal
> extremists are said to reconstruct up to 18 laryngeals (IIRC) ;)
Well, I have heard of ten, but that doesn't mean no-one proposed even
more.

If you ask me, there were three laryngeals, and they were [h], [x]
and [x^w], respectively.  Or something like that.  But some linguists
reconstruct laryngeals in places where there might never have been any.

What to make of them when applying IE sound laws to Quenya?
Well, we know that they all went to never-never land, so one could
simply ignore them, and Quenya doesn't have them, either...
...no, wait, if *h3 was [x^w], it comes close to Q. /hw/, so one could
treat Q. /hw/ as if it was PIE *h3, i.e. it recolours adjacent /e/
to /o/ and deletes, as in _hwesta_ -> _osta_.  And as Q. /h/ seems to
reflect an earlier */x/, one could treat it as *h2 with a-colouring
effect: _heru_ -> _aru_.

> [...]
>
> >   My guess is for _v_ > _bh_ or _w_, but I have no reasoning for these
> >choices.
>
> [v] > [w] or [b] seem the simplest changes. With [bh] you'll need an
> additional explanation for the aspiration - which, however, doesn't seem
> too difficult (e. g. in line with the various relatively recent revisions
> of PIE consonant inventory).
But who of them claim that the traditional voiced aspirated stops
have been fricatives?

I'd treat Quenya /v/ as PIE *w, or just leave it intact.

And then, there are other consonants in Quenya without PIE counterparts:
what happens with /f/?

One cannot blindly apply a set of sound laws to another language unless
the phoneme inventories are the same; if that's not the case, one has
to interpret the rules creatively rather than applying them in a
mechanical
manner.  (See my treatment of Q. /h/ and /hw/ as *h2 and *h3, above.)

Jörg.
.