conclusion relation

[Eduard Hovy]

Hi all,

Juliano, would you agree if one argued that there are three relations 
operating in parallel here between A and B?

A: [[the guy puts integrals wherever he wants,]
      [and does whatever he wants the way he wants,]]
B: [[he doesn't need the theory,]
      [he doesn't need maths.]]

1. reasoned conclusion by the author -- an interpersonal relation:
       "A so/therefore I conclude B"
2. result of state or process in the world -- a semantic relation:
       "B, and as a result A "
3. presentational sequence of the text -- an organizing relation:
       "A then B"
       (also, equivalently, "B then A")

Early RST tended to mix up the various 'dimensions' in which these 
relations are organized, and so obscured the fact that clauses are 
often related in multiple ways at once.  The problem is made worse by 
the fact that English (and probably also Brazilian Portuguese) uses 
the same words for different relations as well.  Your example is a 
nice way of illustrating the need to keep the dimensions apart.

Regards,
E


At 3:28 PM -0700 8/20/08, Maite Taboada wrote:
>Thanks for the interesting example. I'm wondering why a result 
>relation is not appropriate here. Is it because you think the "he 
>doesn't need the theory" part is the nucleus? If so, why not cause 
>either?
>- Maite
>
>Juliano Desiderato Antonio wrote:
>>In Brazilian Portuguese, words like "então" (then) and "logo" (so) are used
>>to signal a relation of conclusion, as in the following example:
>>
>>..  o cara pega e bota a integral onde ele quer,
>>.. e faz tudo que ele quer do jeito que ele quer, .. então não 
>>precisa de teoria,
>>.. num precisa de matemática,
>>
>>.. the guy puts integrals wherever he wants,
>>.. and does whatever he wants the way he wants,
>>.. THEN he doesn't need the theory,
>>.. he doesn't need maths.
>>
>>Actually, it has been very difficult to find a RST relation for cases like
>>this. So, I have been thinking of creating a new relation like "conclusion".
>>I'd like to hear from you about that.


-- 
Eduard Hovy
email: hovy at isi.edu            USC Information Sciences Institute
tel: 310-448-8731            4676 Admiralty Way
fax: 310-823-6714            Marina del Rey, CA 90292-6695
http://www.isi.edu/natural-language/nlp-at-isi.html