HPSG varia

[Tim Florian Jaeger]

== sorry for the format of my email (my first attempt to send to the list
bounced ===

>Hi y'all,
>
>I wanted to reply to some of Daniele's comments:
>
>> (1) Submissions to conferences.
>[...]
>> Does that mean that students and others turn to semantics (or the
>> syntax-semantics interface), because they want to eschew the feuds going
>> on in the field of syntax? Does that mean that people working in the
>> dominant 'framework' (whatever reality it covers does not matter) in
the
>> US and some
>> european countries, are not interested in presenting their work to a
>> general audience? In any case, people working in lexicalist-non-dominant
>> approaches have not sent more abstracts over the years.
>
>as a student who has presented HPSG/LFG research at non-HPSG conferences,
>I find that it is hard (or even impossible) to convey a formal solution
in
>HPSG to an audience that simply doesn't have ANY background in that
>framework (i.e. in my experience 45 out of 50 people). I once saw - I
>think it was - Bob Borsley (??; please correct me if I am wrong) give an
>excellent talk about HPSG solutions to certain syntactic problems at the
>Formal Approaches to Slavic Linguistics. It was an invited talk, so he
had
>an hour. That talk stimulated a lively debate which nevertheless showed
>(to me at least) that many had difficulties following the HPSG
>terminology. A typical conference talk is 20-30 minutes which allows for
>maybe 5-10 minutes of introduction, not 15-25 .... that makes presenting
>in a non-mainstream framework at a non-framework specific conference
>difficult. On the other hand, if we do NOT present at non-HPSG
>conferences, this will never change.
>
>> (2) The attraction of typology.
>[...]
>> The morale is that we need both types of works (if done seriously)
[i.e.
>typological work and detailed empirical coverage of one lg; FJ]. And
>> we
>> will see.
>
>One more reason to see that - at the current state of affair - it
>shouldn't be an "either or" of frameworks. To the best of my knowledge,
>there isn't much typological work in HPSG (a few exceptions come to
mind).
>Regardless of whether one agrees/fancies MP/GB, it has generated a lot
>cross-linguistic research (as has LFG). I don't find it surprising that
>most researchers who are primarily interested in the restrictions across
>languages choose to work in other frameworks (although the HSPG formalism
>may ultimately be much more precise in expressing cross-linguistic
>constraints). There is a reason (no, I don't mean ignorance ;-) why many
>researches working in other frameworks than HPSG perceive the HPSG as to
>not restrictive enough (which may not disturb you, but it does disturb
>them).
>
>I see the HPSG research paradigm as having the primary goal to arrive at
a
>detailed description of one language (or a few lx) first and then move on
>to other lx. This way, cross-linguistic constraints are something that
>(presumably after a rather long time) emerges from the description of
many
>languages. In a research paradigm like GB/MP where cross-linguistic
>constraints are (one of) the primary subject(s) of research, these
>constraints are rather formulated prior to the description of a single
>language and therefore guide which empirical phenomena are perceived
>as "relevant" in a way very different from the HSPG paradigm. While I
>favor what I called the HPSG paradigm (since I believe that is better
>suited to empirically evaluation), I understand why the relative lack of
>cross-linguistic predictions in HPSG is disconcerting for "non-HPSGlers".
>
>Florian