WEQATAL vs. QATAL statistics: a question and a doubt
furuli at online.no
Wed Aug 30 16:38:03 EDT 2000
>Over the last few weeks I have been following with interest, but not always
>in great detail, your exchanges with Henry. As you invite me to, I have
>reached my own conclusions, that there is a clear distinction in temporal
>usage between QATAL and WEQATAL. I think you now agree. If list members
>disagree about whether this distinction is one of meaning or semantics, they
>are now differing more over the meaning or semantics of the terms "meaning"
>and "semantics" within their presupposed theoretical models.
I do not accept that QATAL and WEQATAL are two groups with distinct
meanings; in my view they constitute one single group. I view them exactly
as most grammars view YIQTOL and WEYIQTOL. In order to use the normal
terminology I use QATAL and WEQATAL. I am ready to admit an assymetry in
the time reference and different other functions between QATAL and
WAW+QATAL, and I can even call them "pragmatic" groups. However, all the
time references and different functions of QATAL are found in WEQATAL, so
the difference is one of quantity (pragmatics) and not one of qauality
>I have a question. You state that this set of data "means that neither tense
>nor aspect is grammaticalized in Hebrew." Please can you explain? Henry's
>analysis shows that your data is consistent, to a 90% level or better and
>with a very high level of statistical confidence, with the hypothesis that
>QATAL is a past tense and WEQATAL is a non-past tense, at least if modality
>is included as "non-past". On what basis do you reject this hypothesis? Is
>it that the fit is not 100% or 99%? We would of course not expect a 100% fit
>because of the way in which you have chosen the time category for each verb
>form (see below), and because of the possibility of textual corruption. Of
>course, if there are 10% of cases which do not fit, these need to be
>explained, perhaps by some kind of refinement of the hypothesis. Or is there
>some deeper theoretical reason why you understand these data as disproving
>the grammaticalisation of tense in Hebrew?
At least in two of the posts in the thread I have explained why I conclude
that Classical Hebrew does not have grammaticalized tense or
grammaticalized aspect (in the English sense). In short: a tense has a
uniform temporal distribution, this is not the case of any form in Hebrew.
The English aspects have a uniform distribution in the area
terminated/non-terminated action. No Hebrew form has such a uniform
I do not follow any deeper theoretical reason as a basis for my
conclusions. The case is extremely simple; a tense has a uniform temporal
distribution and an aspect (in the English sense) has a uniform aspectual
distribution. You should read the book of Mari Broman Olsen where these
principles are and point out where you think she errs. A reference to one
of her articles was given on b-greek: http://www.umiacs.umd.edu/~molsen/.
>Also, I have a doubt about your methodology. On what basis did you conclude
>that particular examples of QATAL or WEQATAL were past, present or future
>etc? If I remember rightly you were basing this largely on established
>translations. Did you follow one particular translation consistently, or did
>you use your own judgment in certain cases? If the latter, was your judgment
>independent of your own theoretical model of the Hebrew verb system? I ask
>because some circularity could have crept in here. Translators, of course,
>translate according to the verb forms they see. If they translate from the
>Masoretic text, and they have learned that QATAL is mostly past and WEQATAL
>is mostly future, of course they will translate most QATALs as past and most
>WEQATALs as future, except where the context demands something different.
I classify the forms according to my own judgement based on the context and
do not use translations for this. Regarding each verb I ask: Is reference
time (RT) prior to, contemporanous with or after the deictic point (C)?
Some judgements may of course be disputed, but by and large I think
hebraists will agree with my conclusions. Today we discussed Psalm 110 in
class, and v 6 can illustrate my way of doing an analysis. Here we find one
YIQTOL and two QATALs. Because of the lexical meaning of the words of the
clauses and their construction I conclude that the relationship between RT
and C is similar in all three. Because of the context I conclude that the
RT comes after C in all three instances.
>So can I suggest a further test. A main plank of your argument, Rolf, seems
>to be that the verb forms of the Masoretic text are not an accurate
>reflection of the Hebrew verb system as it was in ancient times e.g. the
>final centuries BCE. So I would suggest that you look at the verb forms not
>in a modern translation into a Western language but in an ancient
>translation into a language whose verb system is well known. I would suggest
>the LXX, though the Vulgate or the Syriac might be interesting alternatives.
>I think it would be very instructive to make a table of correspondences
>between verb forms in MT and in LXX, and then put them into the type of
>statistical grid which Henry has used. This may seem an arduous task, but it
>could be made easy by a clever computer program working with the existing
>Westminster etc morphological analyses of MT and LXX. Note that such a
>program would work with morphological forms only (though I accept that there
>are ambiguous cases where the verb form has been decided from the context)
>and so would not depend on any semantic or syntactic theory.
>You could then test such hypotheses as that THE LXX TRANSLATORS UNDERSTOOD
>(from their unpointed Vorlage and their knowledge of traditional
>vocalisation) QATAL and WAYYIQTOL as (mostly) past and YIQTOL and WEQATAL as
>(mostly) non-past. If these hypotheses were confirmed, at the 90% level or
>even perhaps at a much lower level, it would disprove any suggestions that
>these distinctions were introduced by the Masoretes, or by the chanting
>tradition in the centuries following the fall of Jerusalem. On the other
>hand, if these latter suggestions were true, I would expect to see a pattern
>of LXX to MT verb form correspondences very different from the hypotheses I
>have put forward.
I think that the consonants of the Masoretic text are an accurate
reflection of the Hebrew verb system in ancient times. I also think that
the vowel points were made on the basis of the recitation in the Synagogue.
These also probably are an accurate reflection of the verb system (with its
three radical basis) in ancient times. I am more uncertain as to the value
of the accents, to which degree they are semantic or pragmatic, e.g. what
the intention was of the ultimate pointing of some WEQATALs.
Your suggestion of a comparison of the MT and the LXX is very fine. (I am
at present making such a comparison between the Ethiopic version /which is
based on the LXX/ and the MT, because I teach Ge'ez this semester) I have
already planned to do such a comparison between the whole LXX and the MT.
However, the interpretation of the results of such a comparison are not
uncomplicated. Only interpretations which are based on quality and not only
on quantity should be used. This means that we have to make a strict
distinction between what is semantic and uncancelable and what is pragmatic
and cancelable. And even more important, we have to admit and keep in mind
which assumptions we make. Both points have in some instances been very
difficult to put across on this list - to my big surprise.
Let me take one example. We will find a high percentage of instances where
the LXX translates a WAYYIQTOL with an aorist. Suppose now that we compared
the LXX and the MT and found that the WAYYIQTOLs in 70 % of their
occurrence were translated with Greek aorist. What have we demonstrated as
far as semantic meaning is concerned? Absolutely nothing! Without any
*assumptions* we had shown WA(Y)+YIQTOL could be used with past reference
and the same is true with Greek aorist. But this is *function*
(pragmatics)! To make a quality examination, we have to ask several
questions, some of which are?
1) What is the meaning of aorist? Is it a tense or an aspect or a
combination of both? (I think it represents the perfective aspect but not
2) Why is aorist so often used to translate WAYYIQTOL?
3)Most WAYYIQTOLs have past reference, so the Greek form used to translate
it should also have past reference. What are the Greek alternatives?
(We find Greek aorist, imperfect (and pluperfect)).
4) Do we have reasons to think that each Hebrew aspect covers exactly the
same area as its Greek counterpart, or are there some differences?
5) How would the signals given by a Greek imperfect reflect the WAYYIQTOL
in its context compared with an aorist?
6) Is the meaning of Greek imperfect so different from narrative WAYYIQTOL
that it in most instances would distort the meaning to translate WAYYIQTOL
with Greek imperfect? If that is the case, is it possible that aorist is
chosen, not because it perfectly reflects the meaning of WAYYIQTOL, but
because it is the best possible choice, though an imperfect one? (In
Norwegian we don't have aspects, so the Norwegian Church Bible for the most
part translates WAYYIQTOLs, YIQTOLs, QATALs, participles and infinitives
with past reference with Norwegian preterite. This is the only choice but
a very imperfect one.)
7) Which other Hebrew verb forms are generally or often translated by Greek
aorist? What does the meaning of all the forms which are translated by
Greek aorist tell us about the meaning of aorist itself? Do we find a
uniform time or aspect reference?
There are other questions as well. But the point that cannot be stressed
strongly enough, is that *past reference* is not necessarily *past tense*.
So we need som parameter(s) to distinguish between the two - and I use the
ones mentioned above. To comment on your last paragraph, not even a 90 %
fit between WAYYIQTOL and aorist would prove that the Greek translators
viewed WAYYIQTOL as a past tense, without a thorough study of the questions
above. And as already mentioned, several researchers view aorist as
tenseless. So I return to my original position. Only a study of the
smallest linguistic units of Greek and Hebrew with a strict distinction
between semantics and pragmatics will help us understand their verbal
systems. Other languages or modern Bible translations can only give
University of Oslo
More information about the b-hebrew