Re: interpretations, time, and HTTP [was: checked RDF semantics...]

>On Sat, 2002-12-14 at 01:31, pat hayes wrote:
>[...]
>>  >| It assumes, implicitly, that urirefs have the same
>>  >| meaning whenever they occur.
>>  >
>>  >I think that overstates the case; the semantics only
>>  >assumes that each interpretation gives one denotation
>>  >to each uriref; interpretations corresponding to different
>>  >times might give different denotations to the same uriref.
>>
>>  No. How can an interpretation correspond to a time
>
>Hmm... this seems so straightforward
>to me that I must be missing something...
>
>Consider: interpretation I1, which binds
>	my:dan to me in the present
>	my:age to A1
>In I1, EXT(A1) is { <me in the present, 35> }
>
>and interpretation I2 which binds
>	my:dan to me ten years from now
>	my:age to A2
>In I2, EXT(A2) is { <me ten years from now, 45> }
>
>
>or even simpler: in I1,
>	my:thisYear denotes the year 2002
>and in I2,
>	my:thisYear denotes the year 2012
>
>i.e. one interpretation might be the world
>today, and another might be the world tomorrow.
>Another might be the world as of some HTTP
>request, and another might the the world as
>of a later HTTP response.

Well yes, all true if these are just different interpretations; but 
there is nothing in that to actually connect these interpretations to 
any assertions made at a particular time. The times in the 
interpretations, if there are any, and the times that things are 
posted on webpages have got nothing to do with one another. And 
interpretations can carry all kinds of extra baggage so there might 
be time-sensitive interpretations, but there are also interpretations 
in which, say, a property from 1066 is applied to a subject-object 
pair from 1997, and all kinds of stuff like that. Whereas I think you 
have the picture that the interpretations actually are *at* times, in 
some sense, taken as a coherent whole. Which does make sense, of 
course, but only has any actual bite, more than being just a kind of 
mathematical conceit, when one has some machinery in place which 
makes some semantic sense of this temporal indexing, eg say a tense 
logic. Then it makes sense to require that everything is located in 
some timeframe and interpretations are temporally coherent and so on.

>  > or the set of
>>  interpretations change with time?
>
>'the set of interpretations'... umm... I don't know what
>to make of that... the set of all possible interpretations?
>I don't suppose that changes over time (though my brain
>hurts to think of what that set looks like).

Like making sausages, some things are best not looked at.

>
>>  The truth conditions do not change
>>  with time.
>
>Right; of course not.
>
>>  What you say only makes sense if interpretations are
>>  somehow defined to be temporally parameterized, but they aren't.
>
>I'm not saying they're temporally parameterized; I'm just
>saying that each interpretation is its own world, and
>some of those worlds might be sorta like snapshots
>of the the one we know and at different times.

Some of them might be, but lots more of them won't.

>
>
>>  >WRONG.
>>  >
>>  >(this would need a different sort of test than
>>  >the ones in our test document, but I think it
>>  >could be observed objectively.)
>>
>>  I would like to see it.
>
>umm... this goes well beyond what goes
>in the RDF spec, but if you will,
>consider a semantic extension that
>binds my:now to the current time
>(or perhaps the time when some request is received),
>and includes a class :YearsBefore2010.
>Then take the formula
>
>	my:now rdf:type :YearsBefore2010.
>
>It'll be entailed by the empty formula
>this year, but not in 10 years.

Then we had better not have my:now be a uriref, seems to me. What you 
just said would give you a nonmonotonic logic if 'now' is treated 
just like a name. I think we need to start getting explicit about 
indexicals at some point, but we aren't there yet. If urirefs 
themselves are indexical then we need to re-write RDF from the ground 
up; that will be a major seismic change.

>
>Maybe this testing point is a total red herring...
>
>
>>  Heres an argument why not: suppose that
>>  something is entailed by a graph at one time. Then it is entailed by
>>  the same graph at any other time, since entailment makes no reference
>>  to time (same applies to place, manner, person thinking about the
>>  entailment, whatever). Any graph entails itself. So, any graph at one
>>  time entails the same graph at any other time.
>
>Yeah, but entailment is about stuff that works
>for *all* interpretations.

Right, and what I said was about entailment. So if something doesnt 
follow in ten years then it can't follow now either, since the 
ten-year-from-now interpretations are still perfectly fine 
interpretations now, as well. If they aren't supposed to be, then we 
need some way to rule out interpretations at one time that aren't 
ruled out at another time.

>
>I'm just saying that one interpretation might look
>one way, and another interpretation might look another way.

I think that what you also want/need is to have the interpretations 
switching on and off as the clock ticks. You want time-dependent 
entailments, right? You don't get them in RDF at present, sorry.

>
>>
>>  >
>>  >| We do not make any assumptions about the relationship
>>  >| between the denotation of a uriref and a document or
>>  >| network resource which can be obtained by using that uriref
>>  >| in an HTTP transfer protocol.
>>  >
>>  >Again, that overstates the case. 'We' the formal semantics
>>  >editor don't make any such assumption.
>>
>>  Right, that is what is meant. Restate this to say 'the formal
>>  semantics does not assume any. particular relationship....'
>>
>>  >But we the RDF
>>  >Core WG do expect that URIs will usually be used in RDF
>>  >consistently with their use in HTTP, HTML and other conventional
>>  >contexts.
>>
>>  Do we?? That is news to me. I was under the very strong impression
>>  that we were not making that assumption, in fact, and that the use of
>>  urirefs in RDF was not at all aligned with their use in HTTP or HTML.
>
>Perhaps I distracted you with the word 'consistently'...
>I didn't mean it in its formal logical sense.
>The formal semantics of RDF aren't constrained by HTTP nor HTML.

Phew (mops brow).

>But informally, one can use HTTP and HTML to observe
>that interpretations that satisfy
>	<http://www.w3.org/> dublinCore:title "bananas"
>are not very closely related to the world in which we live.

>
>One can build semantic extensions of RDF that take
>HTTP into account, is what I'm saying, I guess.
>
>>  There is nothing anywhere in RDF that assumes that a uriref has
>>  anything at all to do with whatever happens when you use that uriref
>>  in an HTTP protocol.
>
>Well, there is for example this text in a draft of one
>of the RDF specs:
>
>"The social conventions surrounding use of RDF assume that any RDF URI
>reference gains its meaning from some defining individual, organization
>or context. This applies most notably to RDF predicate URI references. "
>  --
>http://lists.w3.org/Archives/Public/www-archive/2002Dec/att-0053/00-rc#section-authority

That doesn't refer to HTTP, though, right? The defining authority is 
using the urirefs in names in *its* RDF , same as everyone else.

>
>
>>  We have discussed this issue at length several
>>  times and nobody has ever suggested otherwise.
>
>
>>  >This is what the intro to the semantics says;
>>  >directly only briefly, but indirectly thru the concepts
>>  >doc more elaborately.
>>
>>  In that case, it seems to me that the docs as a whole say that RDF
>>  does not in fact mean what the semantics doc says it means. If that
>>  is really the case, then we should say so very clearly, indeed.
>>  However, I wish that you had mentioned this earlier. This wording has
>>  been in the semantics document unchanged now for many months, through
>>  many edits and rewrites, and you have not had any problems with it
>>  before.
>
>Sorry, I didn't notice it. Note that I didn't label it WRONG.
>I just said it overstates the case.
>
>>  And in fact the reason it is in there is because you, Dan,
>>  told me explicitly to NOT try to take account of stuff like this when
>>  doing the MT, but to treat urirefs as simple logical names.
>
>Yes, as a formal matter, they're just logical names.
>But we expect those logical names to get sorta
>connected to the world in which we live via HTTP
>and other communications mechanisms, as discussed
>in the "Meaning of RDF" section of the concepts document,
>right?
>

Well, maybe I was just being crabby. I will draft a rewording that 
tries to not say anything too strong.

Pat
-- 
---------------------------------------------------------------------
IHMC					(850)434 8903   home
40 South Alcaniz St.			(850)202 4416   office
Pensacola              			(850)202 4440   fax
FL 32501           				(850)291 0667    cell
phayes@ai.uwf.edu	          http://www.coginst.uwf.edu/~phayes
s.pam@ai.uwf.edu   for spam

Received on Saturday, 14 December 2002 12:57:09 UTC