On Wed, July 25, 2012 12:44, Pat Hayes wrote:
> ... (01)
> I attended several workshops and meetings on the theme of
> contexts and context reasoning. At one of them I took careful notes. Over
> three days of talks, no two speakers used the same sense of "context".
> They were all talking about different things, ranging in scope from a
> psychological state lasting a few miiliseconds to an entire human culture
> lasting millennia. When people say "context" they simply mean "all the
> stuff that I havn't got an explicit theory of". I have in fact offered
> this as a definition of "context" at the Heidelberg ontology meeting
> (which John Sowa was also present at): (02)
> Context
> The meaning of language is influenced by a very large number of factors.
> No theory of meaning is comprehensive enough to account for all of these;
> particular theories of meaning focus on some and analyze them in detail,
> and ignore others completely. For each such theory of meaning, the
> combined effect of the factors which the theory does not explicitly
> address is often called a "context".
> ... (03)
> Personally, I think we
> would make great progress (indeed, we ARE making great progress) by
> eliminating all talk of "contexts" entirely, and doing our utmost best to
> forget about the term as being pre-scientific and confusing. If you want
> to think about how time influences meaning, think about temporal logics.
> If you want to think about how beliefs influence meaing, think about
> epistemic logics. If you want to think about fiction, study literary
> theory. But don't think that by calling all these (and so many other
> things) all "contexts" that you have thereby achieved any kind of insight
> or clarity. All you have done is get time and belief and fiction muddled,
> ie created confusion. (04)
I agree with Pat that "context" is a very slippery term; that different
people use the term with different meanings, and that one can not
generally crisply define as specific context. (05)
That said, i do think that the concept is useful for reasoning purposes.
So many of the example statements that people have brought up
in this discussion (someone is alive or dead, a person holds a certain
office, someone believes something, ...) are not true under all possible
circumstances. (06)
One can define a reasoning context as a conceptual space in which
a consistent set of statements can be made, which are true "in that
context". (07)
The statements that are declared and derived within such a context are
propositions outside the context
(ist Context2107 Statement02734)
so that without intercontext rules, they have no effect on other contexts.
By creating a context and then making statements within that context,
the statements can be simpler, which makes the reasoning simpler.
Inter-context rules allow reasoning between contexts in a given set. (08)
This is how Cyc's microtheories, which have been compared to contexts,
work. (09)
A context for reasoning purposes can be underdefined. The amount of
definition that is necessary depends on what sort of inter-context
reasoning is intended. (010)
For example, a context for a literary work could state what is true in
that work as facts. The truth of such statements within the context
would have no bearing on their truth outside the context unless
inter-context rules made them so. A rule could specify that if a certain
set of conditions were true within the context of a literary work then
the work was of a certain genre. (011)
For temporal reasoning one could, e.g., write a rule
that if something has an age in one context, and exists in a second
context in the same set, then one can add the difference in the
context times to conclude its age in the second context. (012)
One could also use such inter-context rules for epistemic reasoning
and other sorts of reasoning that Pat mentioned. (013)
-- doug (014)
> Pat
> ...
> ------------------------------------------------------------
> IHMC (850)434 8903 or (650)494 3973
> 40 South Alcaniz St. (850)202 4416 office
> Pensacola (850)202 4440 fax
> FL 32502 (850)291 0667 mobile
> phayesAT-SIGNihmc.us http://www.ihmc.us/users/phayes (015)
_________________________________________________________________
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 (016)
|