SUO: Re: Monosemy, Semantics, and Natural Language

["John F. Sowa" <sowa@bestweb.net>]

Danny and Edward,

Since your notes address related aspects of similar
issues, I'll comment on both of them together.

ED> ... if I recall it correctly, our hero C. S. Peirce
 > has remarked that nothing is a sign unless it is
 > interpreted as a sign.  What troubles me is word
 > 'interpretation'.  Is it a process? - and if it is
 > then what kind of process is it and what are the
 > building blocks of such process?

Yes, it is a process, and the building blocks are
the triadic sign relations.  The attached diagram,
yojo3.gif, is taken from my ontometa paper:

    http://www.jfsowa.com/ontology/ontometa.htm
    Ontology, Metadata, and Semiotics

The three triangles illustrate different kinds of
signs:  names, symbols, and the character strings
that make up the symbols.  Such triangles could be
concatenated or stacked on top of one another
in many different combinations.

DA> Hmm, so returning to the temperature sensor analogy,
 > what if that was hooked up to a heating system? Aren't
 > we dealing with the same kind of (errrm) actors even
 > if there isn't a true cognitive agent involved. In this
 > case a sensor, a communications path, a receiver and
 > an actuator. A cognitive agent could observe the system
 > and however their personal interpretation varied, the
 > internal interpretation of 'temperature' by the agent
 > that controlled the heat would be constant.

Yes, you could consider a thermostat to be a very simple
kind of agent.  Such things nver occurred in nature,
except in living things, until human engineers invented
ways of constructing devices that incorporate an essential
triadic relation.

The diagram feedback.gif illustrates a triad for cruise
control in an automobile.  The engineer who designed that
device built a tiny interpreter that could recognize one
specific type of sign and respond to it in the way that
a human would respond under similar circumstances.  In
effect, the device embodies a frozen interpretation.

Until modern programmable computers were invented in the
1940s, every such triad required an elaborate ad hoc device.
But now, we can program them with loops and recursive functions.

DA> I suppose it's getting into Chinese Room territory -
 > but in that scenario, can the signs be treated as parts
 > of an ontology?

Certainly.  Peirce developed an elaborate classification
of signs, starting with icon, index, and symbol.  And by
the way, I apply Peirce's semiotics to an analysis of
Searle's Chinese Room in Ch. 6 of my KR book.  I believe
that Peirce's classification should be fundamental to any
upper ontology.  Following are two papers that discuss it:

    http://web.clas.ufl.edu/users/jzeman/peirces_theory_of_signs.htm
    Peirce's Theory of Signs

    http://www.digitalpeirce.org/p-190fre.htm
    The classifications of signs (II): 1903

ED> And speaking of interpretation, Umberto Eco expresses this
 > idea: "We explore plurality of possibilia to find a suitable
 > model for realia." (Umberto Eco, The Limits of Interpretation,
 > Indiana University Press; Reprint Edition (March 1994)[ p68 ]).
 > From his statement (or at least in my interpretation) emerges
 > that  'explorative' and  'comparative' elements are required
 > in our interpretation process.
 >
 > Also I agree that "The word "potential" is important" I wonder
 > what is it that determines the 'potentiality'? Should our
 > interpretation also be provided with a 'focusing' element?

All those words (explore, compare, focus, and interpret)
depend on triadic relations, and they should certainly be
an important part of any upper ontology.  But that gets
into a very long story.

John