Anatolians

[Miguel Carrasquer Vidal]

Steven Schaufele <fcosw5 at mail.scu.edu.tw> wrote:

>Miguel Carrasquer Vidal wrote:

>> It's a generally recognized fact that a PIE stop system *t, *d,
>> *dh is typologically unacceptable.

>What about a stop system distinguishing between voiceless, voiced, and
>aspirated (unmarked for voicing)?  Or am i being naive?

Well, that's exactly the Greek/Proto-Italic system *t=/t/,
*d=/d/, *dh=/th/.  It's also one way of interpreting the
Armenian/Proto-Germanic system, with *t=/th/, *d=/t/ and *dh=/d/.
But it cannot be the original PIE system, as we can't derive the
"Greek" from the "Armenian" system, nor vice-versa, nor the
"Sanskrit" system from either.

The problem is that "aspirated (unmarked for voicing)", just like
"ejective [glottalized] (unmarked for voicing)", are purely
theoretical constructs.  Aspirated and ejective stops are
voiceless by definition: adding voice turns them into very
differently articulated phones, namely murmured /d"/ and
implosive /d'/.

In fact, that's exactly what may have brought about the
Greek/Italic and Armenian/Germanic systems.  If we imagine an
earlier system that distinguished lenis (but not voiced) ejective
*d from aspirated *dh, then phonetically the opposition may have
been realized as:

1) plain [t] vs. aspirated [th] ("aspirating dialects");
2) ejective [t'] vs. plain [t] ("glottalizing dialects");
3) ejective [t'] vs. aspirated [th] ("aspirating-glottalizing
dialects").

If the fortis-lenis opposition then tended to develop into
voiceless-voiced (fortis *t [tt] > [t]), aspirating dialects
would have developed a "Greek" system (because of the resistance
of aspirated [th] to become voiced/murmured), and glottalizing
dialects would have developed an "Armenian" system (because of
the resistance of ejective [t'] to become voiced/implosive).

"Aspirating-glottalizing" dialects would have developed, with the
aid of stop + laryngeal clusters, a new 4-way opposition /t/ ~
/th/ ~ /d'/ ~ /d"/ (with implosive /d'/ as in Sindhi and murmured
[half-voiced] /d"/ which develops into [t] + low tone or high
tone + [d] in Punjabi).

=======================
Miguel Carrasquer Vidal
mcv at wxs.nl
Amsterdam