On Mar 17, 2009, at 4:00 AM, Richard H. McCullough wrote:
> On Mar 16, 2009, at 9:46 PM, Christopher Menzel wrote:
> > On Mar 16, 2009, at 5:13 PM, Richard H. McCullough wrote:
> >> mKR does describe actions, and reasoning about actions.
> >
> > Actually, no, it does not. That is what you *intend* the relevant
> > expressions of mKR to mean, but mKR has no semantics (the intentions
> > locked in your own head don't count), so there is nothing that makes
> > it objectively the case for all users that mKR's expressions mean
> what
> > you intend. Compare the semantics for OWL. There is a document
> that
> > everyone can read to find out precisely the semantics of each of
> OWL's
> > constructs. Similarly, google around for "semantics of programming
> > languages" for the same point vis-a-vis programming languages. See,
> > e.g., the book by Winskel (http://tinyurl.com/cnjo9j).
> >
> > Chris Menzel
>
> I have already said that I have not presented a formal semantic
> theory for mKR. (01)
And thus the problem. (02)
> When you continue to harp on that subject in every email, the only
> effect is to suppress communication between us, and other members of
> this forum. (03)
Dick, I am "harping on" that one point because it is fundamental to
your own project. Lack of a formal semantics makes mKR unusable as a
general ontology language. (04)
> I think you totally underestimate the context which members of this
> forum share, as fellow human beings. Consider this mKR proposition.
>
> at view = Ontolog Forum {
> I do say to you od {
> at time = 10 am yesterday {
> I do walk from my house to Raley's done;
> };
> } done;
> };
>
> I think every member of this forum will understand the meaning of
> this proposition. They will know the meaning of "I", "you", "walk",
> "my house", etc. They may not know exactly where "my house" is
> located, but if that is important to them, they can ask me to make
> my context more explicit. (05)
That everyone would understand the meaning of the construct above in
the same way is far from clear. (E.g., is 10am the time of your
saying or the time of your walking? How are we to know? What *makes*
it the case that it is the latter rather than the former? These are
the sorts of questions a semantics answers clearly, objectively, and
unambiguously.) But, be that as it may, what possible advantage is
the above over simply telling me "I walked from my house to Raley's
yesterday at 10am"? Why all the dos and ods and dones? This looks
like exactly the sort of thing that Kautz lampoons in the limerick
that John sent around yesterday. As far as I can see, these
constructs don't enhance people's ability to understand the intended
proposition; if anything, it seems to me, they hinder it. If,
however, you wish mKR to be used in support of sharing, integrating,
and reasoning upon large bodies of information on open networks (the
Web, in particular) -- which is when all is said and done the
fundamental theme of this forum -- then the intended meanings of both
its declarative and procedural primitives and the expressions built up
from them need to be fixed by a formal semantics. (06)
> You seem to imply that mKR is of no value without a formal semantic
> theory. I totally disagree. (07)
That much is clear. :-) (08)
> Further, I assert that there are many meaningful exchanges taking
> place in this forum, even though we have no formal semantics for
> English. (09)
Of course, but we have all learned English by being socialized into
various linguistic communities wherein the semantic conventions of
English have already been assimilated. It is, of course, logically
possible that large communities of people sit down and hash out the
meanings of the mKR constructs informally and start adding them to the
way they speak and write -- although, as noted above, there seems
little reason to do so, as English appears to be doing just fine as a
medium of spoken and written communication between human agents in
most contexts. But once again, if you want mKR to be of general use
as an ontology language that helps us to leverage high-speed computer
networks to share, integrate, and reason upon large bodies of
information, then mKR needs a formal semantics that (i) fixes the
meanings of its primitive constructs and (ii) assigns definite
meanings to complex expressions recursively in terms of the meanings
of their simpler parts. Without this it is impossible to guarantee
that information has been exchanged and integrated accurately and that
inferences drawn on the basis of that information are sound. (010)
Chris Menzel (011)
_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J
To Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx (012)
|