LINGUIST List 6.1697

Sun Dec 3 1995

Disc: Unusual Sound Change t > h

Editor for this issue: Ljuba Veselinova <lveselinemunix.emich.edu>

Directory

1. benji wald , Re: 6.1691, Disc: Unusual Sound Change t > h

Message 1: Re: 6.1691, Disc: Unusual Sound Change t > h

The list messages on *t > h are most interesting. In addition I have
received a number of messages on Irish and other Celtic languages.
I'm waiting a little to see what else comes in before posting a
summary. The Siouan mentioned by Koontz and maybe the Iroquoian
(Northern) as well may be at a greater time depth than the Bantu
changes I originally mentioned. Therefore, a fuller understanding of
the Bantu changes, with an interesting array of reflexes of *t, may be
of some help in reconstructing the paths of the Siouan (at least)
changes. For example, I take it that the *th reconstructed by Koontz
and some others for Siouan was NOT dental, but post-alveolar (cf. the
*rh or *hr reconstruction he mentioned), while the Irish change was
uncontroversially from a dental "t". In the sum, I'll give further
information about the Bantu situation, which also includes a dental
vs. post-alveolar "t" as a phonemic contrast in some languages. I'll
also present a change which I think is even rarer, involving post-
alveolar "t", i.e., its UNCONDITIONED change to a voiceless palatal
affricate, having absolutely nothing to do with textbook
palatalisation. I'll explain what probably happened for this outcome
- and I'm sure it is a most unusual change, with implications for
consonantal shifts (which are clearly very rich and instructive in
Bantu).

Meanwhile, something that has given me pause about the Irish (dental)
t > h as I read up on it is that on one hand it seems that the
intermediate theta fell together with /s/, and both changed to /h/.
However, the merger seems to have only been partial, since
descriptions I have read indicate that the theta > h is affected by
the nature of the following vowel, but I have not noted such an
account for the s > h change. Maybe somebody can clarify this
situation for me before I post a sum. By the way, theta > h is
attested elsewhere, and, of course, s > h is a very common change.
Thus, in t > h, we want to know whether or not the starting point was
dental in order to reconstruct the path of the change. More to come.
Benji