Biblical Hebrew Forum

[Rolf Furuli <furuli AT online.no>]

Peter Kirk wrote:


>Rolf Furuli wrote: "...My work is in progress, but I can reveal that I
>have found a completely different pattern for the aspects in Hebrew
>than for English.  Some glimpses: (1) Only yiqtol and wayyiqtol
>intersect the event *before* ET (conative situations). (2) yiqtol,
>wayyiqtol, qatal and weqatal intersects ET at the nucleus or at the
>beginning of ET. (3) Yiqtol, wayyiqtol, qatal, and weqatal intersect
>the event/state *after* ET (This is particularly seen in Piel because
>it is resultative and factitive). Conclusion: Aspect is subjective and
>not objective! The evidence I gather on the basis of the cancelability
>principle indicating that there is just a pragmatic difference between
>yiqtol and wayyiqtol and qatal and weqatal and not a semantic one, is
>increasing."
>
>Thank you for your clear presentation of your argument. But have I
>understood this part correctly? Are you identifying "objective" with
>"semantic" and "subjective" with "pragmatic"?  It looks like you are
>saying that:
>1. The method that you have chosen for distinguishing between the
>semantic meanings of yiqtol, wayyiqtol, qatal and weqatal fails to
>find an objective distinction between them.
>2. Therefore, there is no objective distinction, everything is
>subjective and pragmatic.

>
>I fear that there is a flaw in your logic here. You assume that the
>method you have chosen is the only possible objective way of
>distinguishing the verb forms, or at least that every possible method
>will give the same result. Your method is no doubt a good one and
>helpful for analysis of many languages. But it is clearly failing to
>analyse Hebrew. You would get a similar result if you tried to analyse
>in terms of tenses a purely aspectual language, or vice versa. I would
>suggest that you try to analyse Hebrew from other directions, using
>distinctions other than those based on reference time and event time.
>If they all seem to fail, then maybe you are right that everything is
>pragmatic. But as some of the more traditional analyses have come
>closer to finding objective distinctions (though granted no-one has
>found the near 100% that you are seeking), that suggests that they are
>closer to finding whatever objective semantic distinctions may exist
>between verb forms.

Dear Peter,


No, I am not identifying "objective" with "semantic" and "subjective" with
"pragmatic", although there are some similarites. I define "semantic
meaning" as a property of something which under no circumstances can be
changed, while "conversational pragmatic implicature" is something that
must be construed on the basis of the context. However, what is pragmatic,
may play an important role in the verbal system, such as for instance
statitivy, which plays an important role in Hebrew but still is pragmatic.
My differentiation between what is objective and what is subjective
revolves around another axis. Let me explain.

When we are going to analyze a dead language, we must use certain
parameters. As Randall repeatedly has pointed out, the fundamental
properties which can, and are applied in most linguistic studies are
tense/mood/aspect. It is methodologically legitimate to use a definition of
one or more of these to test a dead language. I do not claim that my method
is "the only possible way of distinguishing the verb forms", but it is the
best method I have found.

Both yiqtol and wayyiqtol are prefix forms, and most workers would agree
that the conjunction "and" is present in wayyiqtol; and some would say that
another "undefined" element also is present. Two basic viewpoints are
possible: (1) wayyiqtol is nothing but yiqtol + the conjunction (my view).
(2) The way-element signals that the verb belongs to a different
conjugation semantically different from yiqtol.
Most workers subscribe to (2), and three meanings have been proposed: (a)
wayyiqtol is a preterite, (b) it is the perfective aspect, and (c) it is a
combination of the perfective aspect and past tense.

My first step in applying my model is to test the claim that wayyiqtol is a
preterite, and for this purpose Broman Olsen's model is excellent. Her
definition of tense is simple:  (+ past) and (+future) has a semantic
relationship to a deictic point (C) which is pragmatically fixed (it is not
always speech time) - events which are (+past) happens before C and those
which are (+future) happens after C.  Broman Olsen's model is even more
excellent for testing Hebrew since she also has a very clear definition of
aspect in English, and her aspects account for most of the features that
Reichenbach and Comrie ascribes to relative tenses. In this way will tense
and aspect in her model for English combine into a coherent system; and the
combination of tense and aspect will give a uniform interpretation of
English verbs. A superb system!

When I apply the semantic properties (+past) and (+future), I can hope to
achieve one thing: To find whether the *past meaning* of wayyiqtol which is
seen in narratives, is semantic or pragmatic. The reason why I do not  use
the distinction subjective/objective in connection with this investigation,
is that the distinction (+past)/(+future) is intuitively understandable for
all people and can be universally applied. The test is simple: A reasonable
number of wayyiqtols with non-past meaning (when the reason for the
non-past meaning cannot be explained one way or another) do falsyfy the
claim that wayyiqtol is a preterite; so the past meaning must be pragmatic.

Regarding the categories that I call "procedural characteristics" (the
Vendleran categories), the three fundamental semantic properties (+telic
/the end is conceptually included/), (+dynamic /change/), and (+durative
/continuance/) also are intuitively understandable and can be applied
universally, but this is not the case with the properties of aspect which
Broman Olsen lists as (+perfective) and (+imperfective) for English. To use
concepts such as "pragmatic" and "semantic" we need definite parameters by
which we can test languages, and this is lacking as far as aspect is
concerned. Therefore I use the distinction objective/subjective regarding
aspects.

It seems that most or all workers agree that aspect is a kind of viewpoint
and that the imperfective aspect has another scope than the perfective one.
There is also general agreement regarding the end of an event or state,
that this is what distinguishes the aspects. However, Comrie says that the
perfective aspect covers the whole situation with beginning and end while
Broman Olsen says that only the end is important. What I do, therefore, is
to use the properties that are connected with aspect in most models,
namely, "reference time", "event time" and "deictic point" and apply them
to Hebrew. I have never seen a linguist criticizing such an approach. There
may also be other methods, but this is a legitimate linguistic approach -
and I even believe it can give great results.

Let me give some thoughts regarding aspect:

The terms "durative" and "punctual" which are often used as definition of
aspect are Aktionsart terms and not aspect terms. Broman Olsen has
completely avoided Aktionsart definitions of aspects by using the term "an
intersection of ET"; this means that it is the focus of the reporter that
is important. But how can we investigate this "focus" in different
languages? The most important point is whether aspect is an expression of
*the internal time* of an event or not. In Broman Olsen's model tense is an
expression of deictic time and aspect is an expression of internal,
non-deictic time (Comrie agrees). This gives aspect
(in English) an objective meaning; i.e. the interpretation of the different
combinations of tense and aspect gives a uniform interpretation of the
event (deictic time and internal time).

If we can show that the Hebrew conjugations do not represent tense, the
alternative seems to be that they represent aspect. However, because of the
seemingly haphazard use of the conjugations in non-narrative texts, we are
allowed to post the following question: Is it possible to think of aspects
that do not represent internal time? In other words, can we retain the view
that aspects represent different viewpoints, and that the end is important,
while we discard the notion that the aspects represent internal time in an
objective way? To put the question in a practical fashion: Must we conclude
that past events which are portrayed with the imperfective aspect were not
terminated at the point when the reporter focussed upon them (the RT)?
These questions are crucial, and I think they all can be answered in the
affirmative. The disticntion subjective/objective, therefore, has to do
with whether or not aspect represents internal time, not whether aspect is
semantic or pragmatic!

I would like to add an illustration. Each language must be studied in its
own right, and even cognate languages can only illustrate a certain way of
thinking and can not directly help us in our understanding of a related
language. Languages may, however, illustrate possibilities, and I will use
some examples from Sumerian. This is a unique language of the
split-ergative type which has no clear relations with other languages.
There are two "classes" of verbs - "hamtu" and "maru" and these can, from
one point of view be compared to Hebrew prefix- and suffux forms. The verb
in Sumerian consists of a string (often long) of prefixes and affixes
before and after the verb. Some of these are conjugation suffixes. However,
neither the meaning of hamtu/maru  nor of the conjugations are known. The
texts are translated without this knowledge, and this suggests that
fundamental forms may exist in a language, the understanding of which is
not "necessary" for a translation of the language into languages with
TMA-systems. These forms of course have a meaning, but this meaning may be
somewhat different from what we expect. In the UR III-period around 2000
BCE, most verbs with past meaning in the inscriptions were of the
hamtu-class and were of the mu-conjugation, while verbs with present and
future meaning were of the maru-class and other conjugations. Some have
suggested that this implies tense, others have suggest Aktionsart and
others aspect (often with definitions unheard of in Hebrew). However, most
workers have been much more causcious to ascribe a certain meaning to the
two classes and to the conjugations than is the case in Hebrew. The example
illustrates that it is possible to think differently than the tradition
have taught us regarding dead languages.

Regarding Hebrew conjugations, I think we can handle them as aspects, in
the linguistic sense of the word, keeping most of the traditional
descriptions while we discard any notion of internal time. Both the
imperfective and the perfective aspect has a certain meaning, but if they
are subjective expressions, in many instances both aspects can be used for
exactly the same event without any difference in meaning. In other
instances there is a clear difference.
The linguistic tools at the disposition of a reporter do not exactly match
the objective situation. Therefore he or she uses the combination of
different semantic plaenes to convey meaning. Such planes are: Aktionsart,
Procedural traits, Diathesis, Aspect, Mood, Linguistic convention, and
different pragmatic machanisms. The two first together with Linguistic
convention are the more important, and if aspect is subjective, it plays a
minor role. In particular situations, however, is the aspect choosen
crucial.

 By help of the model sketched in this and in my previous post, I think
that the use of *all* the verbs in Classical Hebrew can be accounted for;
but still the model is not so general that everything can be included



Regards
Rolf


Rolf Furuli
Lecturer in Semitic languages
University of Oslo