COMMENT ⊗   VALID 00002 PAGES
C REC  PAGE   DESCRIPTION
C00001 00001
C00002 00002	co[w83,jmc]		correspondence, coherence and consensus
C00007 ENDMK
C⊗;
co[w83,jmc]		correspondence, coherence and consensus

	The various strongly held theories of truth are maintained
by strong intuitions on the parts of their supporters.  However,
these intuitions, often involving important collections of facts
to be taken into account, may not all be strictly relevant to
the concept of truth.  My current view (1983 feb) involves the
following:

1. correspondence theory of truth.  For reasons given in the phil-sci
discussions.

2. coherence theory of meaning, where meaning is considered intensional.
Not every sentence has a definite
truth value in a given referential state.
  This is because some collections
of terms are held together by coherence in such a way that only certain
terms have decided referents in the external world.
Whether a term has a referent in the external world is a distinct
question from whether its referent is directly observable.
Many useful terms are defined only relative to approximate theories.
  This requires defining the
notion "referential state" in a way that makes this possible.  Life World
physics may provide a good toy problem.
Namely, we can ask what referential state is required before the Life
World physicists can even conjecture that the physics of their world
is that of Life.
  There may be
isomorphisms as between field and action-at-a-distance theories,
but the more usual situation is less fancy.  Namely, the referents
can be decided in a definite way, but in the present mental state,
they haven't.  However, a term may have a definite referent even
when some of its subexpressions don't.  Thus "the lemon on the table"
may refer to a certain object even though a speaker or thinker
or database doesn't have a definition of lemon or table.
However, in other states of the world, the same term may have no
referent or an ambiguous referent because of the referential state.
We probably need to consider classes of states of the world.

3. consensus or perhaps dialectical theory of mental process.  This is vaguer
and (to me) less clearly of value.  The idea, emphasized by Jon Doyle)
is that in deciding whether to believe a proposition, an appropriate
mental process considers pro and con arguments.  An example is that
in entertaining a mathematical proposition of the form  (∀ x (P x))
one often simultaneously looks for a proof and a counterexample.
The trouble is that this process seems rather indefinite and is
often not used - the proposition being decided by simpler means.
Also there may be more than two competing alternatives.

	Consider the heuristic that there is always something worthwhile
in any constellation of views held by a large number of people, but
it may not at all be what they think it is.