Re: [tied] PIE Stop System

From: etherman23
Message: 25956
Date: 2003-09-23

--- In, Jens Elmegaard Rasmussen <jer@...>

> There is ample material reflecting aspirated tenues from the whole
> range of IE. The superstition that /th/ etc. is non-existent is
> based on nothing but an undignified consensus to turn a blind eye
> on the relevant evidence.

I have to disagree. Unless, of course, you can come up with
conditioning environments in the various daughter languages that
would account for aspiration sometimes correlating to Sanskrit
aspirates and sometimes not (and sometimes both aspirated and
unaspirated forms existing). It looks totally chaotic to me.

> So, /th, t, d/ instead of /t, d, dh/, in that order? That is,
> Sanskrit /t/, Greek /t/ were aspirated, and Sanskrit /dh/,
> Greek /th/ were unaspirated?? Can you really mean that?

No. Aspiration in Greek, Sanskrit, and probably Italic were
introduced later as dialects were forming.

> > Such a system would be typologically
> > acceptable (it's found in Ancient Greek).
> - but not in that order.

Correct, but such a three way distinction exists.

> The second point is very wrong: Why would your *t take on voice in
> Indo-Iranian, Greek, Italo-Celtic, Albanian and Balto-Slavic?

??? Unvoiced stops cannot become voiced?

> And why would it behave as a lenis in Anatolian? And why would your
> *d take on aspiration in Indo-Iranian, Greek and Italic while your
> *th distinctively [!] loses its aspiration in the same languages?

This is only partially true in Indo-Iranian because some aspiration
was retained. Likewise for Greek. In all of the langauges you mention
the unvoiced unaspirated stops became voiced and there was no need to
have aspiration as a distinctive characteristic in unvoiced stops.

> Why would there be vowel lengthening in Balto-Slavic before your
> *t, but not before your *d and *th?

One could equally ask why, in the Traditional interpretation, there
would be lengthening before *d but not *t or *dh.

> And why would Indic, Greek and Tocharian dissimilate *dVd to *tVd,
> while keeping *thVth intact?

Why shouldn't it?

> And why would *tVt be banned in the root structure?

There is no such ban, except possibly as a ban against identical
stops appearing in the same root (but I'd need to double check on
that). I've read that roots of the form TVT (where T is a voiced
unaspirated stop in the Traditional reconstruction) are allegedly
banned, yet it's not difficult to find counter-examples (at least in
Pokorny's list). Ditto for roots with unvoiced stop-voiced aspirates.

> This surely explains none of the problems raised by the actual
> material.

It explains the typological problems more easily than the Glottalic

> The typological defects of PIE are mended if /th/ is added, and
> there are enough etymologies to demand this. The voiceless
> aspirates are far less frequent than the other series, but so they
> are in Sanskrit, which still must have existed in the shape we know
> it.

There are mainly three areas where aspirates occur: when preceeded by
*s, in onomatopeia, and in stop+laryngeal combinations. This suggests
to me that these were the conditioning environments where aspiration
was retained. It seems to me that there is very little reason to
suppose a fourth series based on evidence outside of Indo-Aryan.
There the evidence is scattered very thinly.

> The traditional phonetic values must have been reached at some point
> *before* the dissolution of PIE, since there was time to integrate
> a few examples of initial /b-/. I suppose this also explains the
> modest number of roots with other vowels than /e/.

The vowel system is a whole other can of worms.