furuli at online.no
Wed Apr 10 10:26:55 EDT 2002
Dear Peter (and Rodney),
The "minimal- pair" situation where there is just a one graphic
difference between two entities is the easiest situation to handle in
linguistics, because all linguistic differences must be caused by
this single graphic difference. The situation that is farthest away
from that ideal situation is discourse analysis because the one
working with this has to handle hundreds of factors at once (scores
of words in sentences and paragraphs, lexical semantics, grammar,
syntax, linguistic convention, idioms, linguistic convention etc) So
there is a great danger of circularity and wrong conclusions in the
attempt to deal with so many words and factors at the same time. I am
not rejecting the method, to the contrary, I myself use discourse
analysis. But the method should be used where it belongs and not
In my view the book of Waltke and O'Connor is the best modern
treatment of Hebrew syntax, and my experience is that it is very
helpful for studens. Any work has a particular angle of approach, and
I think that the authors have succeeded very well in achieving their
goals. I am very happy that the book is not marred with doscourse
analysis, because it has no place in such a work.
If we approach the Hebrew verbal system in a systematic way, there
are at least three areas that need to be singled out, and discourse
analysis will only give results in one of the areas. The areas are 1)
fundamental assumptions, 2) meaning of the morpho-syntactic forms,
and 3) linguistic conventions and communication.
The assumptions used in almost all modern studies of Hebrew verbs,
including W and O'C, stems from the Middle Ages; they were made on
the basis of a tense view when the nature of aspect was unknown. The
basic assumptions are that there are four different conjugations
YIQTOL, WAYYIQTOL, QATAL, and WEQATAL. Discourse analysts and W and
O'C should not be criticized for usind these assumptions, because any
study must use some assumptions. We should however keep in mind that
the results of a study will not be better that its assumptions.
One important approach to the understanding of Hebrew verbs should be
to question these assumptions and deal with them in a scientific way.
Discourse analysis is completely impotent in this area, because these
assumptions are the very axioms of discourse analysis.
Words have meaning, but in the lexicons we find only glosses and not
the lexical meaning of words. Strictly speaking the lexical meaning
of a word cannot be defined (only a general outline can be made), it
must be *known* - by native speakers having the same presupposition
pool. Polysemy exists, but generally a verb or substantive signals
one concept in the minds of native speakers. Different sides of this
concept is made visible in different contexts, and this is understood
by native speakers. Similarly, groups of words (participles,
infinitives, YIQTOLs QATALs etc) signal a common concept each, in the
minds of native speakers, and to try to understand discourse analysis
is again impotent, because it does not address such meaning.
If I ask: "What is the meaning of an English 'present participle'?"
you may try to give a definition. But to give a very specific one,
covering all its functions and uses, is not easy. After some
thinking, however, you, as a native speaker, would come up with a
generalization that would answer the question. How would you achieve
this? Not by the help of discourse analysis, but by the help of your
knowledge of English sentences where the present participle is used.
A similar approach is necessary to understand the meaning of Hebrew
verbs - and this is what W and O'C does in an excellent way.
Discourse analysis would just cloud the picture.
The third area of study is communication, that is, the study of the
linguistic conventions used by the people familiar with Classical
Hebrew, to convey their thoughts to others. In this area discourse
analysis can be a fine tool, but even here would I prefer to use
other approaches first. The reason why I view discourse analysis as a
secondary tool only, is that it cannot pinpoint *meaning*, just
*patterns*, that it to a great degree is open for circularity, and
that there are few if any controls that can be used to test the
conclusions. What I see is that patters of foreground and background
information X-QATAL etc are established, and this is good and well
as a pattern. But then other situations are interpreted in the light
of these patterns as if the patterns were universal, but this is
circular reasoning. Often the results seem to be forced as well.
In one chapter of my forthcoming dissertation I discuss how the
communication of ideas took place in Classical Hebrew. I demonstrate
by examples that all the finite and infinite forms (the participles,
the infinitives, the prefix-forms (YIQTOL, WEYIQTOL, and WAYYIQTOL),
the suffix-forms (QATAL and WEQATAL) can be used for *any* temporal
reference (pre-past, past, "perfect", present, and future) and for
any mood. Then I show *why* there are different patterns when all the
forms of the verbal system can be used for any temporal reference and
mood. The answers are not based upon discourse analysis (that never
can answer any *why*-question). But they are based on an analysis of
the interplay of lexical semantics, Aktionsart, "procedural traits"
(the semantic characteristics dynamicity, durativity, and telicity,
and the pragmatic traits punctiliarity and stativity),"binyan", the
singularity/plurality, indefinitenes/definitenes, and
countability/uncountability of subject and object, adverbials etc.
The characteristic stativity may be used as an illustration. It is
defined as something that continues without an input of energy; thus
a state is durative but not dynamic. Any part of a state is similar
to any other part or to the state as a whole. Communication means to
make something visible to the listener or reader and to make other
things invisible; verb forms serve as tools to achieve this. Whereas
there is a difference in opinion regarding the nature of the
perfective and imperfective aspects, there is common agreement that
they represent different viewpoints. But what is the combination of
viewpoint and stativity? Because a part of a state is similar to any
other part, regardless of whether an infinitive, a participle, a
WAYYIQTOL, a YIQTOL, a WEYIQTOL, a QATAL or a WEQATAL is used,
exactly the same thing is made visible! Therefore we should not draw
the conclusion that two forms have similar *meaning* if they have
similar functions; first we must analyse the other factors. In some
connections, therefore, any of the forms can be used, in others just
one form can be used.
The only restrictions I have found after analysing the 60.000 finitie
and infinite forms of the Tanach is that conative situations
(something is attempted but not carried out) and situastions where
one verb intersects another verb (as in "when John entered, Peter was
reading the paper") can only be expressed by the imperfective aspect
(YIQTOL, WEYIQTOL and WAYYIQTOL), and not with the perfective aspect
(QATAL and WEQATAL). (BTW. In most states expressed by a WAYYIQTOL,
the state is intersected in the middle (imperfective aspect),and not
at the end, as in (1). The point in (1) is not that the slave "had
loved" but that he "does" love. Even when the reference of a state is
past, the intersection is usually in the middle.
(1) Deuteronomy 15:16 And it must occur (WAYYIQTOL) that in the case
he says (YIQTOL) to you, "I will not go out (YIQTOL) from your
company" because he does love (QATAL) you and your household
A very important problem with discourse analysis is the assumption
that YIQTOL, WAYYIQTOL, QATAL, and WEQATAL are four different
conjugations with different semantic meanings. If there are just two
conjugations, as I will give very strong evidence in favor of, one is
in fact analysing WAWs prefixed to verbs when one thinks one is
analysing different conjugations. As an example of the magnitude of
evidence against WAYYIQTOL being different from YIQTOL I have a list
of 86 infinitive absolutes, 760 infinitive constructs, 1.745
participles (all these are used as finite verbs with past reference),
and 1.020 YIQTOLs with past reference that fulfill the criteria that
is the basis for ascribing the perfective aspect to the WAYYIQTOLs.
In other words, if the WAYYIQTOLs are perfective, the same must all
these 3.611 forms be. but that is of course nonsense.
To Rodney: My scepticism regarding the axioms of discourse analysis,
and its usefulness to anything else than describing patterns, is the
reason that I refrain from giving advise regarding your "trial
thesis". I simply am not aware of how this method cangive us any
insights regarding the meaning of the verbal system of Classical
University of Oslo
>Another weakness of Waltke and O'Connor is that it rubbishes then
>ignores discourse analysis, and (in my opinion at least) its treatment
>of Hebrew verbs is thereby seriously flawed.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the b-hebrew