Dissertation - vowels, wayyiqtol, etc.

Henry Churchyard churchyh at ccwf.cc.utexas.edu
Tue May 11 15:38:31 EDT 1999

I've finally turned in my dissertation ("Topics in Tiberian Biblical
Hebrew Metrical Phonology and Prosodics") last Friday, and this final
finished version is available on-line for download as a
.ZIP-compressed Adobe Acrobat .PDF file, at URL:


(Warning: 1.5 meg download)

A smaller excerpt, which contains the sections of the dissertation
which would probably be of most interest to those on this list
(i.e. Section 1.4 on the phonological interpretation of the Tiberian
orthography adopted in the dissertation, especially focusing on the
vowels, and Chapter 4 on the history of Hebrew main stress and the
origins of the consecutive imperfect stress shift) is available at


(only 350k in size)

To view these, you need an "un-ZIP" program to decompress the archive
files that you've downloaded, then the Acrobat viewer, available from
the http://www.adobe.com/ site, to view the resulting .PDF file.

Here's some delayed quasi-responses, in continuation of the last
discussion on the diachronic and synchronic phonology of wayyiqtol
that I actively participated in on this list (last December), followed
by an abstract of the dissertation:

> From furuli at online.no Wed Dec 16 16:43:35 MET 1998
> Subject: Re: "Consecutive imperfect"...
> From: Rolf Furuli <furuli at online.no>
> Date: Wed, 16 Dec 1998 16:43:35 +0200

> I am not aware of any evidence (i.e. data where the question of
> semantics versus pragmatics systematically has been applied) that an
> old Semitic language had a prefix form, short or long which
> expressed past tense.  Could you give an example or two of the
> "awfully suspicious coincidences regarding stress position, "e"/"o"
> vowels instead of long "i"/"u" vowels in final syllables, etc. that
> need to be explained away." if the wayyiqtol originated with the
> Masoretes?

What the phonological evidence reveals is that the wayyiqtol goes back
to earlier *yaqtul (without final vowel), while normal imperfect
yiqtol goes back to *yaqtulu (or some other form with final short
vowel; the exact vowel "u" here is reconstructed from comparative

Now whether the reconstructed *yaqtul can be properly termed a
"preterite" or not is a question to be answered by those who study
comparative Semitic verbal semantics and morphology (of whom I'm not
one); nevertheless, the fact that wayyiqtol goes back to earlier
*yaqtul while yiqtol goes back to *yaqtulu shows that the distinction
between wayyiqol and yiqtol could not have been "invented" by the

> Could you please explain why the the forms of Psalm 21:2 and 16:9
> [(way)yaagel], have penultimate stress while the other three [yaageel
> Psalm 53:7, (w@)yaagiil Zeph 3:17, Hab 1:15] have ultimate stress?

Oops, I seem to have been mistaken (as pointed out by Al Silberman)
when I hastily said in a previous post that there were forms from two
roots here.  Instead, what's happening in these forms is that in
Psalms 21:2 the location of the stress is due to application of the
rhythm rule / n at siga / nasogh 'ah.or (as I should have known, since I
have a list of 1,320 cases of the Rhythm rule, mostly taken from
Revell 1987, in easily accessible form), while in Psalms 16:9 there's
consecutive imperfect stress shift, and the other forms show normal



                      METRICAL PHONOLOGY
                         AND PROSODICS

                       Henry Churchyard
            The University of Texas at Austin, 1999

This dissertation examines a number of linguistic issues in the phonology
of ancient Hebrew, using the extremely rich prosodic information available
in the Tiberian orthography of Biblical Hebrew.  In addition to a basic
analysis of syllabification and stress-assignment, and an examination of
the exact linguistic interpretation of the orthography, a detailed
analysis of Hebrew vowel reduction has been undertaken.  Explaining the
occurrences of reduced and unreduced vowels requires the creation of a
layer of foot-like constituents below the level of feet responsible for
stress in the usual sense.  Also, certain previously-unconsidered
linguistic patterns provide evidence that the distribution of syllable
types in Hebrew phonological representations obeys a trochaic ``Trimoraic
Law'' generalization, and that the metrical constituents governing vowel
reduction are systematically maximally trimoraic in certain contexts.

It is also shown that the Hebrew rhythm rule does not apply in a
fundamentally different way from that of rhythm rules in other languages --
though the particular nature of Hebrew metrical representations does
have certain effects on the rhythm-rule (casting some light on the nature
of rhythm rules in general).  The analyses of vowel reduction and the
rhythm rule together throw doubt on the necessity and desirability of
multiplanar metrical analyses of Hebrew (or in general).

The relationship between two distinct sources of phrasal prosodic
constituency information provided by the orthography is also examined in
detail.  The first is the system of Tiberian cantillational ``accents''
(Hebrew _t.@`aamiim_), which provides an Immediate Constituent parse
(largely prosodic in nature) for the Hebrew Bible.  The second is a set of
phonological phenomena which are sensitive to the greater prominence
connected with phonological-phrase final (or ``pausal'') position.  A
computer-assisted empirical textual-statistical study of the congruences
and discrepancies between the accentual system and such pausal phonology
was undertaken, which throws light on the linguistic nature of the
accentual system.  A number of biblical verses that show apparent
discrepancies between accentual and pausal constituency are collected and
individually discussed, and while the accentual and pausal constituency
systems are often parallel, they cannot always be fully reconciled as
variant orthographic manifestations of a single type of
linguistic/prosodic constituency.

Finally, the history of Hebrew main stress and the origins of the
consecutive imperfect stress shift are also discussed.

10 10 scale/M{rmoveto}def/R{rlineto}def 12 45 moveto 0 5 R 4 -1 M 5.5 0 R
currentpoint 3 sub 3 90 0 arcn 0 -6 R 7.54 10.28 M 2.7067 -9.28 R -5.6333
2 setlinewidth 0 R 9.8867 8 M 7 0 R 0 -9 R -6 4 M 0 -4 R stroke showpage
       % Henry Churchyard      http://uts.cc.utexas.edu/~churchh

More information about the b-hebrew mailing list