Re: CG: SUO: RE: Link Grammar and Parser
>
>
>Another thing I like about RRG is the separate operator projection, and the
>guidelines as to the scope of various operators (operators: markers for
>tense, aspect, negation, evidentials, illocutionary force, modality,...).
>How this all works out computationally remains to be seen.
>
This was around I believe in Dik's functional grammar from around the same
period and is taken up again in versions of Cognitive Grammar... Dik's
grammar has even received computational implementations (toys I believe,
I could be wrong) but still... And if a grammar formalism has not
faced the computational crunch and come through, I would remain
skeptical.
I suppose the dependency-allusion comes from the role of the roles in
RRG, but
I would agree, as dependency grammar goes, RRG is really far out on the
edge of being a family member and I for one wouldn't want to argue the
case.
>By the way -- what do you mean by 'old'? As far as I can tell RRG
>developed during the 80s more or less, and would appear to be about the
>same age as the other things you mentioned -- but I'm no expert, and would
>appreciate any guidance.
>
Yes, times are also related to communities. I guess a grammar developed
in the 80s
must look pretty new in philosophical and philological discourses! But
surely in
computer science and certainly in AI, the 80s are a long time ago. There
were
just so many different grammar formalisms being suggested in the 80s because
there was less understanding of grammar as information and not as form. This
is something that HPSG was very helpful in promoting. Nowadays it is
more useful to think of different formalisms as bringing different
perspectives
on the linguistic information that is maintained for different reasons.
Some of these reasons can be theoretical, others can be practical. For
example: a lexicalised tree-adjoining grammar may be a good choice
for parsing, where a systemic-functional grammar may be a good choice
for generation. Some try to do both, which brings its own costs. This
is the background for a host of conversion activities, considering
for example, grammar specifications as HPSG, lexicalising them
and converting to TAG for faster analysis, perhaps even compiling
out finite state models to go the last inch (mile) in terms of
performance. Or, more recently, analysing corpora with
LFG and converting the results to a combined constituency
dependency representation for more "theory-neutral"
concerns. RRG was one of the formalisms that grew out
of people wanting to get their own perspective into the
picture: for certain typological work it certainly allowed statements
about some cross-linguistic relationships to be made more
conveniently than some others (e.g., X-bar) but nowadays...?
>This is it's explicit attempt to be a more or less universal
>grammar.
>
Yeah, well, there are no shortage of those around...
>For example, while devoting a good deal of space to thematic
>roles, they note that it is important to distinguish between the role a
>participant plays in a state of affairs, and the semantic interpretation of
>
>an argument in the logical structure of a sentence.
>
Also standard in stratified linguistic theories. (But often forgotten in
ontology work, you are right!)
>For those who have read this far and are curious, RRG doesn't use the
>traditional grammatical functions and phrase structures. It is organized
>around a clause nucleus (which contains the predicating element), the
>clause core (which contains the nucleus and the arguments), and a clause
>periphery (which contains modifying pp and adverbials).
>
Well, if it is really of interest, I suppose I could go through the
literature on similar
approaches, those which are already computationally implemented, those which
are not, and so on. There is nothing really unique here: the main
purpose and
value of RRG was, as far as I see it, the typological one and focusing
on a particular
kind of information that is undoubtedly useful for many many languages
(but that
is also why this organisation has occured in various approaches).
>Since I have no deadlines to meet, I'm slowly trying to develop a parser
>that is RRG compliant (more or less) that uses Attribute Value Structures
>
I'd certainly be interested in the result. I am certainly in favour of
more parsers
using more different frameworks rather than less--as long as someone has
the time
to work on it, great!
John B.
>and unification.
>