consonantal roots

Henry Churchyard churchh at
Tue Dec 18 13:04:48 EST 2001

[Tried to send this before, don't think it came through.]

> Subject: consonantal roots
> From: "M & E Anstey" <anstey at>
> Date: Sat, 15 Dec 2001 15:04:16 +0100

> There have been several posts lately about the (psychological)
> reality of pure consonantal stems in Hebrew, modern and biblical.
> For those interested in this issue, there are several points to
> remember: 1. Ussisshkin and Bat-El argue that vowels must be stored
> with consonants for certain derviational processes to work.  Their
> arguments are quite strong, suggesting that lemmas are stored as
> complete.  Breuning also argues that segolates are stored complete,
> as in m‚lek, not as malk-# or mlk.  These arguments suggest that
> consonantal roots ARE NOT basic.  2. Psycholinguistic research of
> aphasic Hebrew (and Arabic) speakers strongly suggests that
> consonantal roots ARE basic, since most of their errors involve the
> placement of incorrect vowels into correct consonantal skeletons.
> There are various other arguments, but the complete picture suggests
> to me that both are true.  Words are stored as wholes, with vowels,
> but that abstractions are made (mostly without awareness) of
> consonantal skeletons that are invariate across many related forms.

Some points:

 -- Ussisshkin basically only examines modern spoken Israeli Hebrew
(I haven't looked at the final form of his dissertation, but that
was true of his earlier writings).  It's doubtful whether some of his
arguments could be extended to Bibilical Hebrew, or to other Semitic
languages.  Also, no one ever has denied that _stems_ (with a fixed
syllable structure and specified vocalism) exist in Semitic languages
in addition to roots -- so /malk-/ is a stem, and m-l-k is an abstract
root, and both can co-exist within the same linguistic theory.

 -- If you're examining Biblical Hebrew by working within any
framework in which different inflectional forms such as _malkii_, unsuffixed _melekh_, etc. are all derived
from an invariant stem-shape within the synchronic grammar of the
language, then it doesn't really make very much sense to say that
surface unsuffixed _melekh_ has a bisyllabic synchronic underlying or
underived form.  If you're not working with a theoretical model
which derives phonetically-related allomorphs from invariant
underlying morpheme shapes within the synchronic grammar of the
language, then of course it's trivially true to say that unsuffixed
_melekh_ is stored as such within the mental lexicon, but this
conclusion is not then particularly insightful, and doesn't really
explain anything.

Henry Churchyard   churchh at

More information about the b-hebrew mailing list