Biblical Hebrew Forum

[Rolf Furuli <furuli AT online.no>]

Dear Dave,

See my comments below,



>Rolf,
>Sorry I took so long to reply to this.  I just found it buried in my
>inbox...
>
>
>> Regarding the Masoretes, we have about the same view. They were extremely
>> faithful copyists who would not dream of changing the text in any way. They
>> faithfully copied what they heard in the synagogue. When I speak of a
>> possible  invention in connection with them, I just point out that we, in
>> the Masoretic text, for the first time see the difference in vocalization
>> between WAYYIQTOL  and WEYIQTOL*.This difference is the basis for the
>> modern view that there are four conjugations. So the Masoretes invented
>> *the points* which later were used to justify a four-component model, but I
>> don't thing they had semantic motives. So I ask: This difference, ist it
>> phonological, is it pragmatic, or is it semantic? What really surprises me,
>> is that people who have strong convictions regarding the Hebrew
>> conjugations never have asked this questions. And not only that, it has
>> never occurred to them that such questions should to be asked. They are
>> just parroting their teachers and their grammars!
>
>The answer is none of the above, Rolf.  The difference is syntactic.
>That's the one BIG factor that is lacking in what you're doing.  You
>talk about semantic properties, pragmatic, phonological, but you
>don't mention anything about syntax.  The difference in the four
>conjugations is a syntactic one, and I don't see you dealing with
>that.

Iam glad you brought up the point about syntax, because I do not neglect it
at all (just think statistics of prothasis/apodosis). However, syntax comes
under the heading "Pragmatics". In fact, I think I can account for the high
number of WAYYIQTOL and QATALs with past reference and YIQTOL/WEQATAL with
future reference on the basis of the combination of three factors: syntax,
topic/comment (also called theme/rheme), and linguistic convention. At
present I am mapping all the active participles of my corpus. I remember a
grammmar saying that HINNE + participle in the 1st and 2nd person (Look
I.., Look you..) always have future reference. So far in my study, this has
turned out to be true. But why is it this way? I will not analyse the
question in detail, only point out that there is nothing in the word HINNE
itself requireing a future reference - it is used for past as well with 3rd
person subjects. Apart from other explanations, this future use is a
*linguistic convention*; it is simply expected by persons having the same
presupposition pool. In Aramaic we have another linguistic convention: (NH
W)MR, two participles with the literal meaning "he answered and said". Why
are the participles used so extensively in this way? We do not know. But
they are consistently used throughout Daniel.

In my view, Classical Hebrew has no tenses. All the finite verbs with and
without prefixed WAW can be both indicative and jussive, so mood is not
grammaticalized in these forms. The infinite form imperative is a
grammaticalized mood, and the short and extra long imperfect *may* code for
modality. The basic difference among the finite verbs, in my view, is one
of aspect- YIQTOL, WEYIQTOL, and WAYYIQTOL code for the imperfective aspect
and QATAL  and WEQATAL code for the perfective one.
The Hebrew aspects are different from the English ones because they do not
objectively signal whether the event time continued or was finished at
reference time. This means that while the aspects are different, they are
not different in relation to temporal reference or mood. (both aspects can
be used for any temporal reference and mood). Because the aspects are
fundamentally different but their use can overlap as far as temporal
reference and mood are concerned, both aspects can be used for the same
event without any difference in meaning. This is even possible in English,
as (1) and (2) show:

(1) This morning Peter was singing in the bathroom.

(2) This morning Peter sang in the bathrom.

We find hundreds of examples in the Tanach (e.g. in paralellisms,
doublettes etc) where a prefix-verb and a suffix-verb have exactly the same
meaning. However, because there is a clear difference between the aspects
as well, the combination of aspect,binyan,lexical meaning, Aktionsart,
verbal arguments being indefinite/definite, singular/plural,and
countable/non-countable,particles, context, syntax,
topic/comment,foreground/background and linguistic convention etc may
convey particular nuances. So the uses of the verbs are not completely
free, but there are clear patterns and restrictions.

The time of an event is important in all languages, and also the mood. If a
language has neither grammaticalized tenses nor moods in the finite verbal
system, how are the time of actions and their mood expressed? Temporal
particles can be used. But there can be no doubt that linguistic convention
will play a great role in this connection. Just as in the case with HINNE
and future reference and participles and past reference in Aramaic, we
expect that certain patterns will evolve. Let us look at one example,
namely the modal use of WEYIQTOL  and WEQATAL. Most grammarians would say
that the two forms basically have the same meaning because WEYIQTOL is
viewed as a YIQTOL with prefixed conjunction (I view WEYIQTOL as
imperfective and WEQATAL as perfective). Let us look at some statistics.

                        WEQATAL                 WEYIQTOL

FUTURE MEANING          4100 (67,3 %)           244 (20 %)
MODAL  MEANING          1256  (20, 6 %)         783 (64 %)

IMPERATIVE MEANING       643 (10,5 %)             0
FINAL MEANING             31 (,5 %)             195 (15,3 %)
VOLITIONAL               147 (2,4 %)            426 (35 %)

While most grammarians view WEYIQTOL  and WEQATAL as having the same
meaning, their use is very different. WHY??? I do not yet have a complete
statistics of YIQTOL, which most grammarians view as having the same
meaning as WEYIQTOL  and WEQATAL. Using only Genesis, we find 163 (55 %)
YIQTOLs with volitional force, 12 (4 %) with final meaning, 19 (6,4 %) with
imperative meaning, and 12 (4 %) with future meaning. So again, why do
three groups, viewed to have the same meaning, function so differently? Why
is the percentage of futures three times as high in the WEQATAL  group as
in the WEYIQTOL  group and the reverse is true regarding modality? First of
all the WAW, prefixed to two of the forms and not prefixed to the third,
does play a role. Because it is a conjunction, it has a syntactic function,
and particular nuances are more natural when a verb has this conjunction
prefixed than when it is without it. But this cannot explain why the
percentage of futures is three times as high in the WEQATAL  group as in
the WEYIQTOL  group and that the reverse is true regarding modality. And
neither can it explain why 643 (10,5 %) of the WEQATALs are used with
imperative meaning but none of the WEYIQTOLs have this meaning.

If we look at the four groups, we do not find that YIQTOL/WEYIQTOL/WEQATAL
have a uniform distribution of time reference, modality (including
imperative) etc, and that the same is true with WAYYIQTOL/QATAL. True, 95 %
of the WAYYIQTOLs and 46 % of the QATALs have past reference, and 37 % of
the YIQTOLs and 64  % of the WEQATALs have future reference, but they have
many other non-uniform meanings as well. So there is no clear pattern of
"semantic meaning" to be seen.


If a language neither has grammaticalized tense, nor grammaticalized mood
in its finite system, it is logical that patterns created by linguistic
convention will play an important role. In addition to the examples of
HINNE+participle and WEQATAL  as imperative, we can think of the fact that
a verb in first position in a clause is avoided in most situations. When we
find a QATAL in first position, we normally take it as a signal of direct
speech; when we find a YIQTOL, we normally take it as a signal of modality
(but there are many exceptions in both cases). This is a pattern of the
language that we cannot explain, we just take notice of it.

My working hypothesis regarding  WAYYIQTOL and WEQATAL is that the
distribution of them primarily is due to linguistic convention (there are
other factors as well; e.g. the nature of the aspects). Because no form
uniforely signals tense, it was natural to create some signals that could
help the hearers/readers (just as sentence-initial position could help
them). To move the narrative accounts forward, the imperfective YIQTOL  was
chosen. Each YIQTOL was connected with the conjunction WAW, and the
syntactic function of the conjunction in this position would signal that
each verb had a new reference time. And similarly with WEQATAL. For future
reference the perfective QATAL was chosen, it was connected with the
preceding text with a WAW which had its normal syntactic function. In this
way we got the system we have, which is misunderstood as having four
different *semantic* groups, while it actually has two groups of different
aspects, but verbs of each group combined with proclitic WAW has particular
syntactic functions. As to the difference in stress of some of the
WAYYIQTOLs and some of the WEQATALs, this is a part of the same pragmatic
pattern. Accents should assit the recitation in the synagogue, and to
contribute to the order I have previously described, it was felt natural to
have penultimate stress on the WAYYIQTOLs whenever possible and ultimate
stress on WEQATAL in a good many cases. If this turns out to be correct,
the stress patterns of the two forms just constitute a general signal of
the special use of WAW+YIQTOL and WAW+QATAL, just as sentence initial
position constitute such a general signal. The consequence is that this
stress pattern of WAYYIQTOL  and WEQATAL is not phonemic (it does not
distinguish different meanings), it is just a general signal of verbal use.

Henry has made a strong case against this view regarding the accents.
However, while being plausible, the weak point is the existence of an
original YAQTUL which represented grammaticalized past tense.



snip

>
>
>If your basis is just the difference in pointing, why don't you
>> conclude there are five conjugations because there are five groups?"
>
>What's the fifth one?

The fifth is WEYIQTOL. If the WAW  is the basis for distinction between the
conjugations why leave out WEYIQTOL? The only grammarian I am aware of who
believes that WAW is semantic and who is completely consistent, is Alviero
Niccacci. He believes there are 5 different conjugations, and even WEQATALs
and WEYIQTOLs with past meaning belong the the conjugations respectively.

Let all the list-members ask themselves? Why do I believe in four
conjugations?  Is it due to the prefixed WAW? If one believes in four and
not five, this cannot be the whole reason. Is it WAW plus temporal
reference that make the conjugations four? Is it WAW plus aspect? Is it
because you see four clearly different functions? In that case, how are the
functions different? (As shown in this and previous posts, the function of
each form is really diverse). And a last and most important question: Which
data speaks clearly against the view that YIQTOL,WAYYIQTOL, and WEQATAL all
represent the imperfective aspect, though having specialized functions; and
which data speaks against the view that QATAL  and WEQATAL  represent the
perfective aspect, though having specialized functions?



>
>> It seems to me that many hebraists have never systematically analyzed the
>> basis of their own view that there are four conjugations, but the view is
>> rather haphazardly woven together by taking a little here and a little
>> there. Others may refrain from such an analysis because of convenience. If
>> one gives a clear analysis of one's basis, one has to defend it as well,
>> and this may prove difficult in view of all the different meanings of the
>> verbs.
>
>Exactly!  That's where tense and aspect fail.  However, taking a
>view that combines Galia Hatav's observations of modality with the
>question "does this have a syntactic connection with what
>precedes or not?" produces a workable model that does indeed
>account for all the apparently-divergent uses that we find in the HB.
>I'd like to suggest that you take this model - I'll be glad to supply
>you with the details of what constitutes a "syntactic connection" -
>and try applying it to some of that mountain of evidence that you
>have amassed - an impressive feat in itself! - and see if it works as
>well as I think it does.


Dave Washburn


Data can be processed and worked on in different ways. I do not claim to
follow the only correct approach. But my approach is very different from
most others because I start with the smallest units and have a strict
distinction between semantic meaning and conversational pragmatic
implicature.


Regards

Rolf



Rolf Furuli

University of Oslo