perm filename KNOW[F78,JMC] blob
sn#393147 filedate 1978-11-06 generic text, type C, neo UTF8
COMMENT ā VALID 00003 PAGES
C REC PAGE DESCRIPTION
C00001 00001
C00002 00002 "Pat knows that Mike knows what Joan asserted" has the following readings
C00005 00003 Perhaps many of the problems arise in the simpler context of
C00007 ENDMK
Cā;
"Pat knows that Mike knows what Joan asserted" has the following readings
supposing that Joan asserted that Jimmy and Billy are brothers.
1. Pat knows that Mike knows whether Billy and Jimmy are brothers.
Neither Pat nor Mike may know that Joan said it.
k(pat,Kwh(Mike, up asserted joan))
2. Pat knows that Mike knows whether Joan asserted that Billy and Jimmy
are brothers. It need not even be true that Jimmy and Billy are
brothers.
k(pat,Kwh(Mike, up Asserted Joan))
3. Pat knows that Mike knows whether what Joan asserted is true. Pat
may not himself know what she asserted.
k(pat,Kwh(Mike, Asserted Joan))
4. Pat knows that Mike knows the truth-value of the proposition
Joan asserted, although Mike
may not know whether she asserted it.
Notes:
1. We use "asserted" rather than "said", because we are talking
about propositions not sentences, although the distinction may not be
important here.
2. The distinctions are not easily made even in English. The
auxiliary sentences explaining the cases seems to be the primary
way of making the assertions precise even in English. This suggests
that the problem may be a pedantic one for AI, but we'll press on in
the hope that the understanding gained may be useful.
3. Our goal is to modify the first order language of concepts
in my concepts paper so as to express all these readings by distinct
logical sentences. We want to preserve the pure first order logic
property of the theory. We do not wish to use any quantifiers in
the sentences themselves, and we wish to retain substitutivity of
equals for equals. It is not clear whether this requires merely an
extension of the language in the concepts paper or major modifications
to it. At present the latter seems more likely.
Perhaps many of the problems arise in the simpler context of
"Mike knows what Joan asserted".
Suppose again that Joan asserted that Jimmy and Billy are brothers,
and we write this
asserted joan = Brothers(Jimmy,Billy).
k(mike,asserted joan)
kwh(mike, Asserted Joan)
is the assertion that Mike can tell us what Joan asserted, i.e. Mike
will tell us that Joan said that Jimmy and Billy are brothers.
Presumably
denot Asserted Joan = asserted joan
But what proposition there is true. It seems that
true Kwh(Mike, Asserted Joan)
might mean something different or else how do we write the first one
true K(Mike,asserted joan).