Ferenc, please pardon me for having in my haste (in
my previous reply to you via the list) missed your real point below, which I
have at last seen as one concerning the machine understanding or at least
analysis of NL texts.
Yes, your observation is extremely relevant and
recalls to my mind what I have related on this list before, namely how I would
start our brainstorming sessions at Metaset in the early 1990s (when I could
validly talk of our team members in the plural) by writing up on the flipchart:
"The meaning of a message is (largely) implicit in its
context!"
By now I can happily report that in The Mainstream
Architecture for Common Knowledge that slogan is implemented very literally,
with appropriately precise technical definitions of "meaning", "message",
"largely", "implicit" and "context", all those definitions being of course
interdependent.
That is the meta-lesson. But the same applies
to all ordinary or detail messages in an architecture-canonical operational
environment. Everything fits together very tightly. Yet quite
naturally too. (That will - I hope - help greatly in overcoming what
might be seen as my dogmatism in these and many other matters.)
(However, for further detail once again I must
ask you to be patient while I prepare the relevant further pages still to go up
on http://TheMainstream.info.)
Thanks once again.
Christopher
----- Original Message -----
Sent: Monday, January 25, 2010 5:51
PM
Subject: [ontolog-forum] tighter control
of ontolog forum
I feel guilty for making a lot of noise on this forum.
My sincere appologies. On the other hand looking at the deadlock when
confronted with the problems to be sorted for further advance I cannot help
trying to contribute with ideas that I find relevant. For instance, one of the
problems that ontologies cannot solve yet as I see it is compatibility between
words used in texts, free-floating or otherwise with reality. All such texts
read projections, because we make statements by using phrases (word clusters)
that never get properly defined as a whole (and not checked against reality,
and parsed by computers as composites of parts) even not even afterwards – for
the sake of reflection. Neither are such phrases recorded with sufficient
annotation of meaning in terms, for instance, extension and
intension.
Random example from a recent posting, no criticism
intended:
(Phrases that are ambiguous or left to any voluntary
definitions of the audience, or evidenced by a single example as opposed to
providing more rigorous and chained definitions)
- the more
productive lists on the Internet that successfully concentrate on their core
themes and objectives
- some friendly
censorship
- too many who
are interested in those core themes are voting with their feet by choosing
not to take part or even listen further
- common
ontological background
- extreme
complexity of our given world that is not always amenable to being captured
by sharable concepts.
- the soft
issues you like raising, of infinities and and ineffabilities
- most members
of this list are fully aware of that background, and of many of the myriad
alternative and complementary points of view that so easily lead
- much fruitless
verbosity.
- further
brainstorming on Eastern philosophies (for example) might contribute
to
such practice.
- the task of
understanding the power of verbalization and its conceptual
constructs
- frequent
distortions of the truth
- impotence as
complete tools
- symbolic
quasi-realities.
- The scientific
worldview
- everyday
commonsense in general, of all kinds,
- mysticisms and
infinities
- many problems
confronting humanity
- the enormous
scale of the still unexploited opportunities
- only implicit
in complex reality,
- collaborativeness.
- extremely distributed and varied world of the
Internet
- conceptual
edifices
- scalable
teamwork and interoperability of social systems,
- optimize
communication,
- maximize
humanity's constructive resources.
- all the
information system challenges
- transcended by
a total disregarding of their respective strengths.
- better
theories.
- practically-constructive theorizing
- already-familiar wistfulnesses with little proven
record of socially-enriching achievement above a certain basic and rather
obvious level.
And so
on. Regards,
ferenc
__________ Information from ESET Smart Security, version of virus signature database 4804 (20100125) __________
The message was checked by ESET Smart Security.
http://www.eset.com
|