Abdul's explanation (below) leads me to the following thoughts:
No matter how "intelligent" the machines are, the symbolic code applied to
signify some aspect of reality is applied by humans to represent that reality -
based on the human's interpretation of what the represented or signified
elements of reality are and do. I don't want to get into the fray of whether one
should use the term "concept" or not - except here, the point is that the
application of the system in a machine is mediated through human interpretation
of reality - it is not direct from reality. If the machine "creates" its own
signification "directly" from reality, it is still done through human-designed
systems for making the correspondence based on human perceptions or rules for
the correspondence with reality, and is interpreted by the humans assessing its
appropriateness, or using the results.
Humans have ways of assessing how "accurately" the significations or
representations apply to or correspond with the represented reality - I've been
in discussions in this forum previously related to such correspondence and its
implications for "truth". But however divorced from human concepts or ideas the
symbols are, we cannot escape the fact that they have been devised and are
interpreted by some path through a human mind, and can only be assessed (so far,
barring a new level of machine "intelligence") by a human assessment of the
correspondence with a human interpretation of reality.
To put it another way, the signs are "sensible" (Abdul's word) only by
virtue of making sense to the human interpreter of them - which involves that
human's perception of the reality to which they apply (dare I say concepts?).
"Significance" can be theoretically independent of the human interpretation
(Azmat's "symbols ... get their significance without the mediation of the
conceptions of human intellect"), but practically speaking, one or more humans
assign and/or interpret the significance - the correspondence with their
perceptions of the signified reality. As long as humans are the ones using the
symbols, we cannot escape the passing of the "signification" through the human
mind, and therefore an inevitable "mediation of the conceptions of the human
intellect," despite Abdul's indication that they get their significance without
it.
Ken
In a message dated 6/15/2007 7:07:48 A.M. Eastern Daylight Time,
abdoul@xxxxxxxxxxxxxx writes:
Paola,
Try to explain your concerns in a more systematic way.
Knowledge systems, as
semantic web applications, thinking machines, etc.,
are all designed to be
using ''sensible'' signs (physical signals, codes,
or words) in order to
process and communicate information about things,
processes, facts, rules,
laws, feelings, ideas, thoughts, or
concepts.
Unlike the human brain, in the intelligent machines the
symbolic codes
signify things directly without the agency of concepts,
constructs, notions,
categories or abstractions. This means that the
nature of mechanical meaning
is dependent on the types of symbols and the
kinds of things these symbols
denote (symbolize, stand for or name) or
represent. And that knowledge
machines are devoid of mental experience or
meaningful mental constructs.
The symbols processed by the mechanical
intelligence are the signs of
entities and hence they get their
significance without the mediation of the
conceptions of human
intellect ( note, the signification, not meaning; for
the symbol
signifies, via denotation and representation, while the construct
means,
via sense and reference). That is why the significance of symbols is
rather to come directly from the real objects denoted and their
relationships connoted, thus leaving off all the conceptual troubles
discommoding human beings.
Kenneth
Cliffer, Ph.D.
|
_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Subscribe/Config: 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 Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx (01)
|