COMMENT ⊗   VALID 00003 PAGES
C REC  PAGE   DESCRIPTION
C00001 00001
C00002 00002	.# Many concepts are meaningful only in an approximate theory.
C00005 00003	Main points of MENTAL
C00006 ENDMK
C⊗;
.# Many concepts are meaningful only in an approximate theory.


	He who insists on an ordinary language definition of the %2internal
energy%1 of a sample of gas
is sure to be disappointed.  %2Internal energy%1 is a term of thermodynamics
that isn't defined separately from the rest of the theory, and its
connection to observation is through the rest of the theory.

	Such an epistemological status for scientific terms is accepted,
but I contend it is true of ordinary language terms as well, and with
an additional complication not often present in the scientific case.
Namely, the theory may be only an approximation, a there may be
no way to define the term with greater accuracy than the theory itself.
This is true of counterfactuals.


	This is true not only of scientific
terms but also of many terms used in common language.  Moreover,
the "theories" that give many terms their meanings are only
approximate and will not bear close scrutiny.  For eample, if
it is asked when Pat went out the door an answer "about two
minutes ago" may be given.  If more precision is demanded, an
answer accurate to about a second may be given in a way sufficiently
objective so that two observers with stopwatches would agree without
discussion.  If .01 second were demanded, the observers would
have to use instruments and would have to agree on what part
of Pat was meant.  If a microsecond were demanded, new conventions
would have to be developed on exactly what matter comprises Pat
at a given instant and exactly what "out the door" means.

Main points of MENTAL

	Utility of mental concepts applied to machines

	Inadequacy of definitional methods

	New methods

		approximate theories

		second order definitions

		approximate cartesian products

	Metaphilosopy or philosophical modelling