Prototype Theory and Hebrew Tense/Aspect

yochanan bitan ButhFam at
Wed Dec 15 16:25:35 EST 1999

>> (buth)
>> well, if you're talking prototype (and network semantics) 
>> then you'll have to do something more with vayyiqtol. 

>Not necessarily.  That's what we're trying to sort out.

the whole idea is to start out within an intuitive ballpark. 
that is why past and perfective are included. 
that intuition needs to extend to the difference between vayyiqtol and
and it shouldn't be too hard. you paint with large strokes at the

this may point out some of the difference between what happens 
when one uses a language and one sits outside a language analyzing it. 
analysis can be as convoluted or counterintuitive (a great generative
as one likes. 
(like when degen [altaramaishe, 1969] calls a 
finite verb clause with fronting a 'nominal clause'. that produces a
non-generating, counterintuitive grammar.) 
back to the point, a prototype should start with broad intuition. 
i am happy to state that a vayyiqtol is not simple default order, that the
first position of the verb is 'marked' because it blocks almost any
fronting. (fronting produces an extra-clausal topic, like isaiah 6.1 or
moabit ve`omri melex israel vayye`anev ... .) 
but vayyiqtol does mark/imply a general thematicity and sequentiality. it's
just something that happens after processing tens of thousands of
occurrences as well as musing over options that were not chosen tens of
thousands of times.

>> the features listed above are the prototype of qatal
>> in narrative and direct speech, (+ ki-clauses and asher- clauses).
>> if you don't like sequence for vayyiqtol, and i wouldn't push it, 
>> then at least +thematic ('theme advancing saliency').
>> see how easy it is to talk about a prototype?

>Of course it's easy to talk about it, but demonstrating it is another 
>matter.  I find that beginning an investigation with a foregone 
>conclusion doesn't produce much in the way of improved 

well, if i remember correctly, a friend of mine, (kirk lowery, editor of
this list) did a dissertaion on vayyiqtol where close to 3/4 of the cases
of vayyiqtol were 'cleanly' sequential. the other 25% had various kinds of
overlaps and technical problems. anything with that kind of magnitude means
that it will become a practical starting point for decoding/expectation.

>And the set I suggested is hardly what you describe, 
>since it includes [-dependence -subordination] which is definitely 
>not a feature set of ki and asher clauses!

fair enough, but notice that they were put in parentheses. 
"(+ ki-clauses and asher- clauses)"
i was signally additional implicit information.
as for ki-clauses, they can occur independently, though maybe not
as for asher clauses, they are dependent. 
both are important to generative linguistics and qatal and yiqtol.
because ki- and asher-  allow for pre-verb 'fronting' 
yet more frequently do not use fronting, 
i.e. frequently do not show/need any pragmatic marking/manipulation 
and thus reveal a more default/basic order for simple finite verb clauses.
qatal clauses are in wider oppposition than to just vayyiqtol 
and will need this other imput that does not apply to the vayyiqtol
so i added the ki- and asher- clauses to remind one of the overall basic
word order
parameters for qatal within the whole language.

yisge shlamax
randall buth

More information about the b-hebrew mailing list