Notes for paper on use of mental terms by programs and in referring to programs.

The discussion of this in (McCarthy and Hayes) is to be updated.

Questions:

	1. What are the various levels of consciousness that a program may
posess.  (e.g. it may make sense to say that a certain program believes
that a tape is rewound but not that it believes that it believes it.

	2. What terms are there and can we find a complete set in the sense
that ordinary language text can be expressed using only these terms.

	3. What are the rules of inference?  Can they all be expressed
in logic neatly or are some of them essentially complex programs that refer
to a whole state of knowledge?


Here are some terms:

but - "but" signals that something is different from what might be expected.
"John went back to his house, but he didn't go in."  It is tempting to suppose
that it is only a vagary of English that "and" feels awkward in this sentence.
Are there languages without "but"?   Vaguely, it seems that biblical Hebrew
may lack "but".

knows and believes - We need to distinguish knowing a fact and knowing the
value of an expression as in "John knows Tom's telephone number".  This is
different from the "knows how" of philosophers.
The right formalism that covers both knowledge and belief is
value(expression,person,situation).

wishes, hopes, fears,regrets
Are these all cases of wishing?

however - Is this just "but".,although

therefore, so, since

consciousness and self-consciousness - Simple consciousness need involve
only beliefs about the world and perhaps about the self as an actor but
not as a thinker.  Thus it is plausible that a dog is conscious that
it will be punished if it steals the steak, but there is no evidence that
a dog regrets its infirmity of will and indecisiveness in the matter.

such as