semantics editors draft

The document (same uriref http://www.coginst.uwf.edu/~phayes/) now 
has another evening's work on it, almost entirely concerned with 
section 3.4 (datatyped interpretations), which has been extensively 
rewritten, following email discussions with Peter.  As noted in the 
document preamble, this doesn't change anything outside this section, 
but people still might want to check it over as the actual text has 
been almost completely rewritten and the conditions in the tables 
reformulated. It reads OK to me but I would welcome an outside eye to 
check the wording.

The D in D-interp is now a *mapping* from URIrefs to datatypes, not 
just a set of datatypes. But notice the use of the term 'datatyped 
with respect to', which was already in use elsewhere in the document, 
so I just borrowed it here to make the required connection to 
entailment. D-entailment still means exactly what it used to mean and 
is defined the same way as before.

This way we can have the datatype names attached to, but still not 
actually 'part' of, the datatype. This finally, I think, puts to rest 
all the worries about how to handle 'names' of datatypes, by the way.

Comments welcome.

Pat

PS:
FAQ: why the separation between D-interps and the rdfs:Datatype condition?
A: because the others allow D to get bigger harmlessly (cf lemma) but 
this one bounds it from above, so blocks upward 'datatype 
monotonicity'. So it makes sense to separate them.

PPS. I forgot, Ive also made changes in response to some of Graham's 
comments. Some of them are made moot by the rewrite described. Others 
are good ideas but I havnt done them yet. I will respond to Graham's 
message directly with details.
-- 
---------------------------------------------------------------------
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@ai.uwf.edu	          http://www.coginst.uwf.edu/~phayes
s.pam@ai.uwf.edu   for spam

Received on Monday, 14 April 2003 22:19:48 UTC