RE: No Standard Semantic Web Pragmatics?

>  > From: Pat Hayes
>>  Sent: Thursday, June 17, 2004 3:30 PM
>>  >
>>  >I think there should be a means for creating names,
>>
>>  Right, me too. At present there is no official way to name - to give
>>  a name to, to 'baptize' - anything on the Web. All one can do is use
>>  names that are created elsewhere, or rely on the sagacity of the
>>  reader to figure out what a new name is intended to name from the
>>  context of initial use. (? ...all uses? ...all by the owner of the
>>  name? ...ignoring the owner of the name? Ah, the familiar debates)
>>
>>  >i.e., for creating
>>  >rigid designators, as in Kripke, that could be used in all
>>  >interpretations to designate the same thing. And I think there should
>>  >be a standard way to find out what it is intended that it be used for
>>  >with machine readable data, natural language, images, and anything
>>  >else necessary to fix its reference. That comes out sounding
>>  >like what URIs were said to be, at least the URL kind, at least
>>  >if you put related stuff there (also Pat's 'rather lame' example
>  > >http://www.ihmc.us/users/phayes/PatHayes What is lame about it?).
>>
>>  That it can only be understood by human readers, and has no meaning
>>  for any software agent. Which I think is exactly your point, right?
>
>Well, right, that is clear. But I thought maybe you saw a technical
>problem with it that you were referring to but not revealing. The thing
>about names, if I understand Kripke right, is that they denote by social
>authority, not by descriptive validity.

But surely the two are linked at some point. After all, social agents 
tend to operate in large part by reading and understanding 
descriptions.

>So even if you lied on that page,
>and gave the wrong SS number, or said you were a perfect being, it would
>still refer to you. It would name you by WILL, not by TRUTH, and give the
>rest of the semantic web an anchor point in social reality.

Hmm, I confess to being a bit doubtful about that. Not that I 
disagree, I just reserve an opinion. I think that when we get up 
against genuinely social issues like this, we have to defer the 
matter to genuinely social authorities like courts of law. If I had 
lied about myself, would the page with the lie really refer to me? 
What if I had lied about the information that was being intended o be 
used to do the identifying? I dunno, is my answer: but I'd follow the 
Supreme Court's verdict rather than Kripke's. After all, Kripke is 
just describing a social convention himself, and he might have got it 
slightly wrong.

>It could be
>used to resolve those debates you mention as well, by containing and
>communicating the intent of the creator.

Right, it was intended to be an illustration (as forceful and 
unambiguous as I could make it) of the possibility of doing this.

>
>Another pretty thing about this, and other pragmatic machinery, is that
>for the most part, as far as I can tell, they can peacefully coexist with
>the existing model theoretic semantics. They just complement and extend
>it. Or is that not true?

That seems to be the case, yes. They are essentially orthogonal. When 
it is necessary to explicitly link them, its seems to be kind of 
trivial to do so (as in the 4-author paper.)

Pat

>
>John Black
>
>>
>>  Pat Hayes
>>  --
>>  ---------------------------------------------------------------------
>>  IHMC	(850)434 8903 or (650)494 3973   home
>>  40 South Alcaniz St.	(850)202 4416   office
>>  Pensacola			(850)202 4440   fax
>>  FL 32501			(850)291 0667    cell
>>  phayes@ihmc.us       http://www.ihmc.us/users/phayes
>>
>>


-- 
---------------------------------------------------------------------
IHMC	(850)434 8903 or (650)494 3973   home
40 South Alcaniz St.	(850)202 4416   office
Pensacola			(850)202 4440   fax
FL 32501			(850)291 0667    cell
phayes@ihmc.us       http://www.ihmc.us/users/phayes

Received on Thursday, 17 June 2004 20:24:37 UTC