Biblical Hebrew Syntax

Rolf Furuli furuli at
Wed Apr 4 16:33:29 EDT 2001

Dear Trevor,

I would like to give a few comments to your points. See below:

>Just a couple of thoughts. (See below.)

>   >
>   >
>   [snipped]
>   >
>   > When we theorize about the older stages of a language, we may be
>closer to
>   > metaphysics than to linguistics. I am not aware of any compelling
>   > showing the existence of a preterite YIQTOL (a form YIQTOL which was
>   > grammaticalized past tense), and I doubt that YIQTOL ever had such a
>   > meaning.
>I suppose "compelling" is too subjective a category to address here. But I
>would think that the existence in both Akkadian and Hebrew of a short
>prefixal form and a long prefixal form, combined with the fairly
>comparable distributions (that the short form is used to narrate past
>events, while the long form is used to show non-past or modal action),
>constitutes at least some sort of evidence. I'll freely admit that I'm not
>all that familiar with the actual arguments for the connection. (It's
>usually been presented to me as a more or less accepted conclusion.) But
>it does seem to explain a good portion of the evidence.

I drop the word 'compelling', and would rather say that the the reasons for
believing in a very old preterite (grammticalized past tense),(e.g in
Ugaritic and Accadian) is paradigmatic (in the Kuhnian sense) rather than
evidential. When the supposed Stammbaum of man is constructed in
biology,comparative anatomy is used as a tool, and the paradigm (scientific
model) which is used, is that organic evolution is a fact. Based on this
assumption, bones are compared and the relationship between homo habilis
and others are worked out. At least two times in the last decades has the
Stammbaum been completely restructured when new bones have been found. This
is in my view pure speculation, and paradigms will in time be rejected and
new ones formed. Comparative linguistics applied to 'Proto-Semitic' (if
such a language ever existed) and the stages thereafter is just as
comparative anatomy, almost pure speculation, and it is also built on a
particular paradigm, namely that Classical Hebrew had four different
conjugations (at least most workers appear to build on this model). I for
one dispute this model.

What really was a shock for me in my introduction to Semitic languages, was
that that 'fundamental truths' that we were taught, and which all the
authorities believed, was just *assumed* and had never been tested. It was
just as we read in Aboth chapter 1. "Moses received the Law from Sinai, and
he delivered it to Joshua, and Joshua to the elders, (and so forth)".

To give an example: The same short prefix-form in Ugaritic is used both
with past meaning and with modal meaning. So it is assumed that i
represents two different conjugations, one modal and the other preterite.
Why is it assumed? Not because anybody has tested it with modern linguistic
techniques, but because the paradigm (model) demands it. And are the
Ugaritic verbs with past meaning really preterites (grammaticalized past
tense)? I am not even aware that anybody in a printed paper has asked the
question, but it is taken for granted that past reference is the same as
past tense (sic!). The following would be a simple linguistic test (which
to the best of my knowledge never has been performed either for Hebrew or
Ugaritic): Outline principles for how to distinguish between factors which
are pragmatic and which are semantic. Study all (or at least a great
portion) of Ugaritic verbs (or Hebrew verbs) in the light of their deictic
point, their event time and their reference time, and see if a certain
pattern can be found. Only when past reference is an uncancellable property
of a certain verb form can it be termed as preterite. This approach will
not give definitive answers, but it is at least a scientific approach and
not only assumptions!
>   > And similarly with Accadian stative: I am not aware of any
>   > evidence which should make me believe that Hebrew QATAL developed
>   > from this
>   > form. I think that a semantic comparison between Hebrew QATAL
>   > and Accadian
>   > perfect is more natural than a morphologic comparison between Hebrew
>   > and Accadian stative.
>Perhaps you could give some examples. I'm thinking particularly of what
>seems to be a more technical nature in the Akkadian perfect. From my
>limited experience, it seems like it is often used in conjunction with the
>preterite to mark the main point in the discourse. Maybe it's just that I
>haven't been looking for it, but I don't see the same force in the Hebrew
>   [snipped]

I have just finished a translation of Atrahasis into Norwegian and have
been working with the Gilgamesh Epic and with Enuma Elish as well. It is
clear to me that Accadian stative has almost no resemblance with Hebrew
perfect except when the Hebrew root itself is stative. In the mentioned
documents we find what is called 'consecutio temporum', two or more IPRUS
(socalled 'preterite') forms followed by a perfect, all with past meaning.
In the later stages of the language, the perfect takes over more and more
for IPRUS and it becomes the normal narrative form. So the use of Accadian
perfect resembles the use of Hebrew perfect, much more than does Accadian
stative. A challenge to those who believe that Accadian stative is the
basis for Hebrew perfect: Write down all similarities and differences and
count them. The only similarity I can see is that both can be called

>   > We should also add that Accadian has both a stative
>   > conjugation and a perfect. If the stative developed into the Ge'ez
>   > (and Hebrew QATAL) what about the Accadian perfect?
>If the perfect is as specialized in Akkadian as I've thought, I wouldn't
>see much of a problem here. Little would be lost in its absence, and
>little would be gained with the extra complexity. You say that the stative
>is unique to Akkadian; so why could it not be the other way around?

Everything is possible because what we are making, are just educated
guesses. However, I see little or no resemblance between Accadian stative
and Ge'ez perfect, but several similarities between Accadian perfect and
Ge'ez perfect.

>   >
>   > My guess, which is just as metaphysical as yours, is that the short
>   > prefix-forms of Accadian, Ge'ez, and Hebrew (YIQTOL and WAYYIQTOL)
>have a
>   > similar meaning - they are connected with modality
>   > rather than with tense.
>But in a synchronic analysis, it's hard to find uses of the Hebrew long
>form that don't carry some type of modal spin. So where is the distinction
>with this definition? I don't see much of a uniquely modal connection in
>Akkadian, either. It seems that the preterite is used pretty regularly for
>simple past narration.

In a synchronic analysis we find the following: The Accadian IPRUS
(socalled 'preterite') is often used in narration with past reference but
it is also the stem used for the wish- and asservative forms (precative,
cohortative and vetitive). So it can both be used with past reference and
for expressing modality (and it can be used with present and future meaning
as well). In Ugaritic the short form YAQTUL can both be modal and can be
used with past reference. In Hebrew the short YIQTOL often is modal, and it
is often used as the stem in WAYYIQTOLs. In Ge´ez the short prefix-form
YINGER is often modal. It can be used with past reference, but this is not
very often. We may like it or not, but in the mentioned languages the short
prefix-form is connected with modality - and with past reference. A
plausible explanation of this form should therefore be able to account for
both these factors.

The long YIQTOL need not be modal. Look at the following statistics from
Genesis and Exodus:

1314 studied YIQTOLs of which 47 are apocopated and 25 are combined with
cohortative. (Many of the forms cannot be apocopated). Of these forms I
found the following characteristics:

57 (4,3%) past
62 (4,7%) present
635 (48,3%) future
2 perfect
245 (18,6%) modal
313 (23,8%) imperative or command

I am at present working of a complete mapping (temporal reference, modality
etc) of all the around 12.000 finite and infinite verbs in the Dead Sea
corpus. While there are grammatical differences in the language compared
with Biblical Hebrew, as shown in Qimron's grammar, I find no real
difference in the conjugations. More YIQTOLs are modal in this corpus than
in the OT, but while most cohortatives are modal in the OT, just a few have
this meaning in the DSS.

>   > The long prefix-forms in the three languages are
>   > connected with the imperfective aspect, and Ge'ez and Accadian
>perfect and
>   > Hebrew QATAL are connected with the perfective aspect. Accadian
>   > contains both verbs, adjectives and substantives, and is
>   > something which is
>   > peculiar for that language. It is more like the D-forms
>(Dopplungstamm) in
>   > meaning in the other languages than the conjugations.
>I guess I'm going to need more information on this one. Akkadian has a D
>stem, and it seems to overlap pretty well with the Hebrew D. Why would
>they need another system for conveying the same thing? And in what sense
>are they the same?

When we use the word 'stative' or speak about a 'state' in connection with
verbs, we speak of a characteristic which basically is *lexical* and not
morphosyntactic, as is the case with aspect. However,statitivity is not a
semantic property, because stative verbs can also have a fientic
interpretation. Stativity is in linguistic studies given a slot together
with 'procedural characteristics' such as 'accomplishment', 'achievement'
etc. and not with the imperfective and perfective aspects. Therefore is
statitivity in Classical Hebrew primarily connected with lexical form, and
a special case of it is connected with the stem of the verb, namely, the
D-stem. But I am aware of no reason to connect it with the *conjugations*
(prefix-form/suffix-form) in any of the Semitic languages, because these
conjugations operate on a completely different semantic plane. It can also
be questioned whether the stative in Accadian is a conjugation on the same
level as IPRUS, IPARRAS, and perfect, because adjectives and substantives
can also be parsed together with verbs. So basically the stative is a bare
form (verb, adjective or substantive) with an enclitic pronoun (or no
pronoun in 3.p.s.).

Waltke/O'Connor has a very fine discussion of Hebrew Piel, showing that
resultativity is a main characteristic of the verbs (just as is the case in
the Accadian D-form). This means that the subject leads an object through
the end of an action and into a resultant state, and the state is what is
stressed. My students often have problems to understand Accadian stative
when fientive verbs occur as statives. How do we for instance ascertain the
verb "run" in Accadian stative? The Accadian stative is more specialized
than the D-form,but it is also very similar to this form. Therefore, my
point was that Accadian stative resembles the D-form much more than
Accadian or Hebrew perfect or any other conjugation. So if Accadian stative
was the basis of anything in another Semitic language, I would expect that
it was the basis for a stem-form rather than for a conjugation.

>   >
>   [snipped]
>   > In other words, the WAWs of the
>   > WAYYIQTOLs and the WEQATALs are just conjunctions and signal no
>   > semantic meaning, and Hebrew has just two conjugations -YIQTOL and
>   > WAYYIQTOL on the one hand and QATAL and WEQATAL on the other.
>Then why the distinct form for WAYYIQTOL? Why does it always build from
>the short prefixal form? And what happened to the distinction that you
>outlined above between the short and long forms? Doesn't that amount to
>more than two conjugations? And why would it be necessary to use the modal
>form so consistently in past narrative?

WAYYIQTOL does not allways build from the short prefix-form. Even when a
short form is available, the long form is sometimes used. In my view the
WA(Y)- before YIQTOL and the WE- before QATAL are simple conjunctions; they
have absolutely no semantic meaning, and their use is based on linguistic
convention. As far as modality its concerned, it encompasses different
things in different languages, and it seems to me that the semantic range
of modality in Hebrew is slightly different from what is is in Germanic
languages. We should not therefore a priori reject the possibility that
'modal' verbs could be used in narrative.

>Trevor Peterson



Rolf Furuli
University of Oslo

More information about the b-hebrew mailing list