pre-IE k > H

[Robert Whiting]

On Sat, 25 Mar 2000 CONNOLLY at LATTE.MEMPHIS.EDU wrote:

<snip>
> Carol Justus wrote:

>> I wonder if it is not important to distinguish between a *k > h (such as
>> happened in Germanic (e.g., cornu, 'horn') and *k- > largyngeal written in
>> Hittite with a sound transcribed as 'h' with a diacritic under it?

> Absolutely!

Actually it isn't.  The writing of Hittite /h/ with the diacritic is just
a historical accident.  Hittite cuneiform was adapted from Akkadian
cuneiform.  In Akkadian, the only h-type sound is etymologically /x/, and
the transliteration system devised for Akkadian reflects this. (In writing
foreign words and names, the Akkadian h-type signs are often used for a
wide range of laryngeals [h, he, ayin, ghain].) However, the Akkadian
signs for this were adapted from Sumerian and we don't really know what
h-type sound Sumerian had (but most suspect that it was plain /h/).
Since Akkadian was the first of the cuneiform languages to be read, the
h-type signs were written with the h with diacritic that is used to
transcribe the etymologically equivalent sound (/x/) in Arabic.  Again, we
don't know what the h-type signs represent in Hittite since the sound is
not preserved in any other IE languages, but the transliteration system
for Hittite simply uses the values of the signs that were established in
Akkadian because the transliteration system is supposed to be a one-to-one
mapping of cuneiform signs to the latin script (i.e., no interpretation is
allowed and arbitrary changes to the sign values are not permitted).  So
the diacritic on the h has no particular significance for Hittite or IE
phonology.  It is only significant in identifying cuneiform signs, and
since Akkadian has only one h-type sound, the diacritic is often dropped
in publications to save typesetting costs and trouble since there is
nothing else that it can be confused with.  But the value of the
Hittite sound represented by the cuneiform cannot be extrapolated from
the Akkadian value except to conclude that it is some kind of laryngeal.

This being the case, Carol is still quite right in saying that
k > laryngeal is better terminology than k > h.  But it has nothing
to do with the diacritic under the h.

Bob Whiting
whiting at cc.helsinki.fi