[Lexicog] Semantic parsing

[Mike Maxwell]

Patrick Hanks wrote:
> I guess all this reinforces the point that semantic tagging is more 
> problematic than part-of-speeech tagging. 

Those of us who are old codgers will remember that Fillmore's "Case for 
Case", which came out in 1968, used basically the same sort of semantic 
tagging as is being discussed here--agent, patient, etc.  Over the next 
several years, any number of case systems were proposed, which offered 
more (usually) or fewer cases.  (SILers on this list may also remember 
that for a time, these semantic cases were an essential part of Tagmemic 
descriptions.  And even Chomsky referred to such things as Theta Roles, 
although it's hard to tell how seriously he took them.)

One of the strongest criticisms of this enterprise, I believe, was that 
there was never any principled way of choosing a universal case system, 
or even a language-particular one (with the possible exception of 
languages like those of the Philippines, where you could argue that the 
set of language-particular case roles was the same as the actual case 
distinctions made in the morphosyntax).

Another criticism of Fillmorean case grammars was that there were always 
borderline cases (pardon the pun), where you couldn't decide for a 
particular clause or verb which case to assign one of its NPs to.

In the end, the linguistic community largely gave up on Fillmore-style 
case grammar, AFAIK.

Of course, similar objections can be--and have been--raised about 
assigning parts of speech (= syntactic categories).  Linguists have 
argued, for example, about whether some languages (like the Tucanoan 
languages of Ecuador, Colombia and Brazil) have the category 
'Adjective', and part of the reason for the argument is that it is 
unclear what a universal definition of that category would be.

Also unclear is how different the morphosyntactic behavior of a word has 
to be before you assign it to a different category (including perhaps a 
unique category for that word or for that language).  Think of words 
like 'ago' and 'hence' in English, for instance, or the multiple 
subclasses of so-called "adverbs" in English.

Depending on your proclivities, you may find such debates depressing 
("There isn't any answer") or exciting ("I'm going to publish the 
answer") :-!
-- 

    Mike Maxwell
    CASL/ University of Maryland


------------------------ Yahoo! Groups Sponsor --------------------~--> 
You can search right from your browser? It's easy and it's free.  See how.
http://us.click.yahoo.com/_7bhrC/NGxNAA/yQLSAA/HKE4lB/TM
--------------------------------------------------------------------~-> 

 
Yahoo! Groups Links

<*> To visit your group on the web, go to:
    http://groups.yahoo.com/group/lexicographylist/

<*> To unsubscribe from this group, send an email to:
    lexicographylist-unsubscribe at yahoogroups.com

<*> Your use of Yahoo! Groups is subject to:
    http://docs.yahoo.com/info/terms/