Biblical Hebrew Forum

[Kimmo Huovila <kimmo.huovila AT helsinki.fi>]

Dear Rolf,

Thank you for your reply. I just comment on some theoretical and
methodological issues, because I have much more competence in these
areas (I am a student of Hebrew, but I major in general linguistics).

You wrote to Peter:

> This would be helpful, because I agree that
> given the definintion of Broman Olsen and Comrie it *is* forced to view
> WAYYIQTOLs as imperfective, but not if we use the definition of Carlota
> Smith, Carl Backe [KH: Bache?] or myself.

Jouko Lindstedt wrote (Nested Aspects, in Aspect Bound: A voyage into
the realm of Germanic, Slavonic and Finno-Ugrian aspectology, p. 23-38,
edited by Casper de Groot and Hannu Tommola, 1984, Foris Publications -
Dordrecht):
"In aspectological studies, there has been a long tradition of what can
be called subjectivism. By this term I mean theories according to which
the choice of aspect in the co-called aspect languages is to a
considerable degree independent of what kind of situation the speaker is
referring to.....To subjectivism we can oppose denotative theories:
aspect forms in various languages denote something and the choice
between them is not primarily bubject to the speaker's "stylistic" aims.
In denotative theories, aspect systems reflect the human way of
classifying states of affairs (or situations)."

Do you perhaps refer to this kind of distinction above? (Perhaps not, as
I am not sure Carlota Smith would be a subjectivist, but I do not have
her book at hand to check.) I believe in a denotative view of aspect, as
it explains why a change in aspect always changes the situation type
(Vendlerian classification). I know that this is controversial, but for
a reference (besides my thesis), Alexander Mourelatos has some helpful
comments on Vendler's classification along these lines (Events,
Processes, and States, in Tense and Aspect 191-212. Syntax and Semantics
14, edited by Philip J. Tedeschi and Annie Zaenen. Academic Press: New
York). (You probably have already collected a good bibliography, so
perhaps there is nothing new in these references, but I give them just
in case they might prove helpful.)

Rolf Furuli wrote:

> As I already have stated in this thread, I believe that in the study of
> Hebrew verbs, morphology is the fundamental property by which to start, and
> then the use of the forms. This means that  I will not without strong
> evidence accept that YIQTOL is imperfective in most contexts and perfective
> in others and that QATAL is perfective in most contexts but imperfective in
> others. This would be tantamount to linguistic anarchy.

It is not such a rare phenomenon that a linguistic construction has some
specific meaning in one context, but not in another. Take Greek
imperfect, for example. It grammaticalizes past time and imperfective
aspect, but not in every context (pastness can be canceled in e.g.
hypothetical conditions). This is where I find prototype theory useful
(as expounded by cognitive grammarians, e.g. Taylor: Linguistic
Categorization: Prototypes in Linguistic Theory, Langacker: Foundations
of Cognitive Grammar, Volume 1), as it allows to build hierarchies of
interrelated (more and less
schematic) prototypes. This is also quite in the spirit of construction
grammar.

I think that linguistic anarchy would result only if one never studies
the contexts in which a form may grammaticalize imperfectivity or
perfectivity. If the contexts are clear cut, the theory can be very
precise. It is not exceptional if there is a linguistic opposition
between two constructions in certain contexts that does not surface in
other contexts, if the analysis is detailed enough. Actually many
construction grammar analyses focus on this kind of stuff.

Of course we should strive for finding unviolable generalizations that
would hold for all examples (possibly rather abstract, as in the case of
English genitive), but such a characterization would very possibly be
too
inclusive. Thus it is possible that in this case there is no TMA
category that these Hebrew forms grammaticalize in every case, but if
this is found to be so, the analysis must not stop there. The result
need not be linguistic anarchy, unless the general theory and
methodology are too restrictive.

Kimmo