Biblical Hebrew Forum

[Henry Churchyard <churchyh AT ccwf.cc.utexas.edu>]

> From: Vincent DeCaen <decaen AT chass.utoronto.ca>
> Date: Fri, 23 Jul 1999 12:46:29 -0400 (EDT)

> we do assume vowel length oppositions underlyingly. we explain this
> as a mismatch between underlying representations and surface outputs
> (we follow the idea that surface length is predictable and so
> nonphonemic).  some common alternations.

>                       qal *u          hifil *i

> long /VV/             YFQW.M          YFQIYM
> short stressed /V/    YFQOM           YFQ"M
> unstressed            WAY.FQFM        WAY.FQEM

> in this case, a three way opposition is explained as a contrast
> between /VV/ and /V/, and also in quality/length distinctions for
> /V/ under stress. vowel length is phonemic in underlying
> representations.

Well, active phonological alternations between "long" h.ireq or
h.ireq-yod and s.ere/s@ghol are not actually very frequent (perhaps
mainly relics from an earlier diachronic stage of the language), nor
are alternations between "long" shureq/qibbus. and h.olem/qames..

By contrast, alternations between s.ere and s@ghol, and between h.olem
and qames. (or between h.olem and "short" qibbus. which precedes a
geminate consonant) are pervasive, and are _not_ strictly correlated
with stress -- for example, po`alo "his dead" (h.olem,h.at.eph-pathah.,
h.olem) vs. po`olo (qames.,h.at.eph-qames.,h.olem), two variant forms
of "his deed" which are both stressed on the last syllable.

The most prominent alternation between h.ireq-yod and s.ere/s@ghol
(besides that of "short" verb forms which you mentioned above) is the
alternation between s@ghol and h.ireq-yodh in lamedh-he forms and the
word for "mouth" -- but in that case s@ghol actually prefers to be in
word-final _stressed_ position...

--
Henry Churchyard   churchyh AT ccwf.cc.utexas.edu   http://www.crossmyt.com/hc/