perm filename KNOW.ART[F75,JMC]3 blob sn#554808 filedate 1981-01-03 generic text, type C, neo UTF8
COMMENT ⊗   VALID 00002 PAGES
C REC  PAGE   DESCRIPTION
C00001 00001
C00002 00002	.require "memo.pub[let,jmc]" source
C00035 ENDMK
C⊗;
.require "memo.pub[let,jmc]" source;
.cb AN AXIOMATIZATION OF KNOWLEDGE AND THE EXAMPLE OF THE WISE MAN PUZZLE

.cb John McCarthy

[Note: This memo was written in the Fall of 1975.  As of September 1977,
the situation had changed in the following respects: (1) Chris 
Goad did an axiomatization in which one could assert that a person
knows only what follows from a certain set of facts.  This allows
proving that the first and second wise men don't know the colors of
their spots rather than assuming it.  However, this axiomatization
seems unsatisfactory in various ways. (2) Morgan (1977) axiomatized
modal logic within first order logic (IEEE Computer Society
special issue on theorem proving).  (3) The original McCarthy axiomatization
can be treated just as readily in first order logic as the axiomatization
involving possible worlds. (4) Sato's thesis has contributed much
to the semantics of knowledge systems.  (5) The method of circumscription
has indicated that the way to circumscribe a
person's knowledge to a given set of facts should involve an axiom
schema.]

	This memo reports results  of a part of long  continuing work
on  the axiomatization  of knowledge for  the purposes  of artificial
intelligence.  Unfortunately, little of that work has  been published
even informally, and this is a piece out of the middle, so to speak.

	It  is  derived from  the  author's  modal axiomatization  of
knowledge, but  is more directly based on a Kripke type semantics for
that modal  system given  by M.  Sato of  Kyoto Univerity.   It  owes
considerable to  discussions with Professors Takasu  and Igarashi and
Messrs. Hayashi and Sato of Kyoto University.

	That axiomatization  is  distinguished  from  others  in  the
literature by the  presence of an  individual called FOOL,  such that
anything FOOL knows, he knows that every one knows.

	We work  in the first order logic  system associated with the
FOL proof-checker of Weyhrauch.

	There  are  three  sorts   of  individuals  called   persons,
propositions, and worlds.  Each variable  is confined to one of these
sorts,  and  the  sort  mechanism of  FOL  restricts  the  meaning of
quantification of a variable to the appropriate sort.

	The first  predicate is T(proposition,world)  and means  that
the  proposition  is  true  in  the  world.    There  is  a  function
K(person,proposition) which is  a proposition  asserting that  person
knows the proposition serving as argument.   The propositions of this
system are  individuals with respect to the  first order logic rather
than truth values or wffs.  Therefore, we have to expicitly introduce
proposition-valued   functions  AND,   OR,  IMP,   EQUIV,  and   NOT,
corresponding to the logical connectives ∧, ∨, ⊃, ≡, and ¬.  Finally,
KW(s,p) is introduced as an abbreviation for K(s,p) ∨ K(s,NOT(p)) and
is read "s knows whether p".

	Here is  the axiom system in FOL  notation.  The declarations
establish the types of the variables  and constants.  The axioms  are
essentially all  definitions except for  REFLEX which asserts  that a
world is accessible from itself for any person.

DECLARE INDCONST FOOL ε PERSON;

DECLARE INDVAR S S1 S2 S3 S4 ε PERSON;
DECLARE INDVAR W W1 W2 W3 ε WORLD;
DECLARE INDVAR P P1 P2 P3 P4 Q Q1 Q2 Q3 Q4 R R1 R2 R3 R4 ε PROPOSITION;

DECLARE OPCONST K(PERSON,PROPOSITION) = PROPOSITION;
DECLARE OPCONST KW(PERSON,PROPOSITION) = PROPOSITION;
DECLARE OPCONST AND(PROPOSITION,PROPOSITION) = PROPOSITION;
DECLARE OPCONST OR(PROPOSITION,PROPOSITION) = PROPOSITION;
DECLARE OPCONST IMP(PROPOSITION,PROPOSITION) = PROPOSITION;
DECLARE OPCONST EQUIV(PROPOSITION,PROPOSITION) = PROPOSITION;
DECLARE OPCONST NOT(PROPOSITION) = PROPOSITION;

DECLARE PREDCONST T(PROPOSITION,WORLD);
DECLARE PREDCONST A(PERSON,WORLD,WORLD);

AXIOM REFLEX: ∀ S W.(A(S,W,W));;

AXIOM KNOW: ∀ S P W.(T(K(S,P),W) ≡ ∀W1.(A(S,W,W1) ⊃ T(P,W1)));;

AXIOM KW: ∀ S P W.(T(KW(S,P),W) ≡ T(K(S,P),W) ∨ T(K(S,NOT(P)),W));;

AXIOM FOOL: ∀ S P W.(T(K(FOOL,P),W) ⊃ T(K(FOOL,K(S,P)),W));;

AXIOM AND: ∀ W P Q.(T(AND(P,Q),W) ≡ T(P,W) ∧ T(Q,W));;
AXIOM OR: ∀ W P Q.(T(OR(P,Q),W) ≡ T(P,W) ∨ T(Q,W));;
AXIOM IMP: ∀ W P Q.(T(IMP(P,Q),W) ≡ T(P,W) ⊃ T(Q,W));;
AXIOM EQUIV: ∀ W P Q.(T(EQUIV(P,Q),W) ≡ (T(P,W) ≡ T(Q,W)));;
AXIOM NOT: ∀ W P.(T(NOT(P),W) ≡ ¬T(P,W));;

	As an example of the use of these axioms, we do the following
puzzle of the three wise men.

	"A  certain king paints white  spots on the  foreheads of his
three wisemen who can see each  other and are told that at least  one
spot is  white and the  others black.   He asks  if any can  tell the
color of  his spot and after a while the wisest affirms that his spot
is white.  The puzzle is to tell how he knows."

	The solution is  to say that  the wisest reasons as  follows:
"If my spot  were black, the smarter of the  remaining two would have
seen it and reasoned that if his  spot were black, the slowest of  us
would have seen two black spots and reasoned that his must be white."

	In this form, the puzzle requires the  wisest to reason about
how fast his colleagues think.  We can get rid of this at the cost of
unfairness by  having the  king as  each wise  man in  turn with  the
answer being audible to the others.   Another solution would have the
king  not accept  volunteered answers  but repeat the  question three
times,  getting the  correct  answer  from  all three  on  the  third
repetition.

	Here are the axioms describing the puzzle:

DECLARE INDCONST WISE1 WISE2 WISE3 ε PERSON;
DECLARE INDCONST RW ε WORLD;
DECLARE INDCONST WHITE1 WHITE2 WHITE3 εPROPOSITION;

AXIOM WISEMAN:
	T(WHITE1,RW)∧T(WHITE2,RW)∧T(WHITE3,RW),
	T(K(FOOL,KW(WISE1,WHITE2)),RW),
	T(K(FOOL,KW(WISE1,WHITE3)),RW),
	T(K(FOOL,KW(WISE2,WHITE1)),RW),
	T(K(FOOL,KW(WISE2,WHITE3)),RW),
	T(K(FOOL,KW(WISE3,WHITE1)),RW),
	T(K(FOOL,KW(WISE3,WHITE2)),RW),
	T(K(FOOL,OR(WHITE1,OR(WHITE2,WHITE3))),RW),

	T(K(WISE3,K(WISE2,NOT(K(WISE1,WHITE1)))),RW),
	T(K(WISE3,NOT(K(WISE2,WHITE2))),RW);
;

	Next follows an annotated version of  the FOL proof. Included
in the  proof are the proofs  of four lemmas that  depend only on the
knowledge axioms and not on the specifics of the wise man puzzle.

	Steps 1 thru 11 are  a proof of the result of  step 11, which
is  a lemma to the  effect that if S  knows whether p and  p is true,
then S knows p.

      *****∀E KW S,P,W;

      1 T(KW(S,P),W)≡(T(K(S,P),W)∨T(K(S,NOT(P)),W))

      *****∀E REFLEX S,W;

      2 A(S,W,W)

      *****∀E KNOW S,NOT(P),W;

      3 T(K(S,NOT(P)),W)≡∀W1.(A(S,W,W1)⊃T(NOT(P),W1))

      *****∀E NOT W,P;

      4 T(NOT(P),W)≡¬T(P,W)

      *****ASSUME T(K(S,NOT(P)),W);

      5 T(K(S,NOT(P)),W)  (5)

      ***** 3,5;

      6 ∀W1.(A(S,W,W1)⊃T(NOT(P),W1))  (5)

      *****∀E ↑ W;

      7 A(S,W,W)⊃T(NOT(P),W)  (5)

      *****⊃E 2,↑;

      8 T(NOT(P),W)  (5)

      *****⊃I 5⊃↑;

      9 T(K(S,NOT(P)),W)⊃T(NOT(P),W)

      ***** 1,4,9;

      10 (T(P,W)∧T(KW(S,P),W))⊃T(K(S,P),W)
      *****LABEL (KW1 . 11);

      *****∀I ↑ S P W;

11 ∀S P W.((T(P,W)∧T(KW(S,P),W))⊃T(K(S,P),W))  

	Next comes a proof of the lemma on line 18 which asserts that
what S knows is true.

      *****∀E KNOW S,P,W;

      12 T(K(S,P),W)≡∀W1.(A(S,W,W1)⊃T(P,W1))

      *****ASSUME T(K(S,P),W);

      13 T(K(S,P),W)  (13)

      ***** 12:13;

      14 ∀W1.(A(S,W,W1)⊃T(P,W1))  (13)

      *****∀E ↑ W;

      15 A(S,W,W)⊃T(P,W)  (13)

      *****⊃E 2,↑;

      16 T(P,W)  (13)

      *****⊃I 13⊃↑;

      17 T(K(S,P),W)⊃T(P,W)
      *****LABEL (KNOWTRUE . 18);

      *****∀I ↑ S P W;

18 ∀S P W.(T(K(S,P),W)⊃T(P,W))  

	We begin  the proof proper,  and our  first objective is  the
assertions at step 23 and 24 that WISE3 knows WHITE1 and WHITE2, i.e.
he knows the spots of the other wisemen are white.

      *****∀E ↑ FOOL,KW(WISE3,WHITE1),RW;

      19 T(K(FOOL,KW(WISE3,WHITE1)),RW)⊃T(KW(WISE3,WHITE1),RW)

      *****∀E ↑↑ FOOL,KW(WISE3,WHITE2),RW;

      20 T(K(FOOL,KW(WISE3,WHITE2)),RW)⊃T(KW(WISE3,WHITE2),RW)

      *****∀E 11 WISE3,WHITE1,RW;

      21 (T(WHITE1,RW)∧T(KW(WISE3,WHITE1),RW))⊃T(K(WISE3,WHITE1),RW)

      *****∀E 11 WISE3,WHITE2,RW;

      22 (T(WHITE2,RW)∧T(KW(WISE3,WHITE2),RW))⊃T(K(WISE3,WHITE2),RW)

      ***** WISEMAN6,WISEMAN1,19,21;

23 T(K(WISE3,WHITE1),RW)

***** WISEMAN7,WISEMAN1,20,22;

24 T(K(WISE3,WHITE2),RW)  
*****LABEL (ASS1 . 25);

	Since our goal is an assertion that WISE3 knows WHITE3, which
must  therefore be  proved true  in any world  accessible to  him, we
assume that W1 is  an arbitrary such world.   This assumption is  not
discharged until step 131.

*****ASSUME A(WISE3,RW,W1);

25 A(WISE3,RW,W1)  (25)

	We now  must prove the  properties of W1.   Many of  them are
just properties of RW carried over as knowledge of FOOL or properties
that WISE3 knows in RW.  The  unindented lines 32, 33, and 40 and  41
are proved in this way.

      *****∀E KNOW WISE3,WHITE1,RW;

      26 T(K(WISE3,WHITE1),RW)≡∀W1.(A(WISE3,RW,W1)⊃T(WHITE1,W1))

      *****∀E KNOW WISE3,WHITE2,RW;

      27 T(K(WISE3,WHITE2),RW)≡∀W1.(A(WISE3,RW,W1)⊃T(WHITE2,W1))

      ***** 23,26;

      28 ∀W1.(A(WISE3,RW,W1)⊃T(WHITE1,W1))

      ***** 24,27;

      29 ∀W1.(A(WISE3,RW,W1)⊃T(WHITE2,W1))

      *****∀E ↑↑ W1;

      30 A(WISE3,RW,W1)⊃T(WHITE1,W1)

      *****∀E ↑↑ W1;

      31 A(WISE3,RW,W1)⊃T(WHITE2,W1)

      *****⊃E 25,↑↑;

32 T(WHITE1,W1)  (25)

      *****⊃E 25,↑↑;

33 T(WHITE2,W1)  (25)

      *****∀E KNOW WISE3,K(WISE2,NOT(K(WISE1,WHITE1))),RW;

      34 T(K(WISE3,K(WISE2,NOT(K(WISE1,WHITE1)))),RW)≡∀W1.(A(WISE3,RW,W1)⊃T(K(WISE2,NOT(K(WISE1,WHITE1))),W1))

      *****∀E KNOW WISE3,NOT(K(WISE2,WHITE2)),RW;

      35 T(K(WISE3,NOT(K(WISE2,WHITE2))),RW)≡∀W1.(A(WISE3,RW,W1)⊃T(NOT(K(WISE2,WHITE2)),W1))

      ***** WISEMAN9,34;

      36 ∀W1.(A(WISE3,RW,W1)⊃T(K(WISE2,NOT(K(WISE1,WHITE1))),W1))

      ***** WISEMAN10,35;

      37 ∀W1.(A(WISE3,RW,W1)⊃T(NOT(K(WISE2,WHITE2)),W1))

      *****∀E ↑↑ W1;

      38 A(WISE3,RW,W1)⊃T(K(WISE2,NOT(K(WISE1,WHITE1))),W1)

      *****∀E ↑↑ W1;

      39 A(WISE3,RW,W1)⊃T(NOT(K(WISE2,WHITE2)),W1)

      *****⊃E 25,↑↑;

40 T(K(WISE2,NOT(K(WISE1,WHITE1))),W1)  (25)

      *****⊃E 25,↑↑;

41 T(NOT(K(WISE2,WHITE2)),W1)  (25)

	Now it  turns out that  we need two  more lemmas, one  to the
effect that  what FOOL knows, everyone knows  he knows, and the other
to the effect  what FOOL knows in  any world, he  knows in any  world
accessible from it  by any knower.  It would  have been more esthetic
to  have proved these lemmas at the  beginning, since, like the first
two, they don't  involve the  specifics of the  wiseman puzzle.   The
lemmas are lines 46 and 53.

      *****∀E FOOL FOOL,P,W;

      42 T(K(FOOL,P),W)⊃T(K(FOOL,K(FOOL,P)),W)

      *****∀E FOOL S,K(FOOL,P),W;

      43 T(K(FOOL,K(FOOL,P)),W)⊃T(K(FOOL,K(S,K(FOOL,P))),W)

      *****∀E 18 FOOL,K(S,K(FOOL,P)),W;

      44 T(K(FOOL,K(S,K(FOOL,P))),W)⊃T(K(S,K(FOOL,P)),W)

      ***** 42:44;

      45 T(K(FOOL,P),W)⊃T(K(S,K(FOOL,P)),W)
      *****LABEL (FOOL2 . 46);

      *****∀I ↑ S W P;

46 ∀S W P.(T(K(FOOL,P),W)⊃T(K(S,K(FOOL,P)),W))  

      *****∀E KNOW S,K(FOOL,P),W;

      47 T(K(S,K(FOOL,P)),W)≡∀W1.(A(S,W,W1)⊃T(K(FOOL,P),W1))

      *****ASSUME T(K(FOOL,P),W);

      48 T(K(FOOL,P),W)  (48)

      ***** 45,47:48;

      49 ∀W1.(A(S,W,W1)⊃T(K(FOOL,P),W1))  (48)

      *****∀E ↑ W1;

      50 A(S,W,W1)⊃T(K(FOOL,P),W1)  (48)

      *****⊃I ↑↑↑⊃↑;

      51 T(K(FOOL,P),W)⊃(A(S,W,W1)⊃T(K(FOOL,P),W1))

      ***** 51;

      52 (T(K(FOOL,P),W)∧A(S,W,W1))⊃T(K(FOOL,P),W1)
      *****LABEL (FOOLTRANS . 53);

      *****∀I ↑ S P W W1;

53 ∀S P W W1.((T(K(FOOL,P),W)∧A(S,W,W1))⊃T(K(FOOL,P),W1))  

	With these  lemmas, we transfer  some of FOOL's  knowledge to
W1, and the results are lines 59 thru 63.

      *****∀E ↑ WISE3,KW(WISE1,WHITE2),RW,W1;

      54 (T(K(FOOL,KW(WISE1,WHITE2)),RW)∧A(WISE3,RW,W1))⊃T(K(FOOL,KW(WISE1,WHITE2)),W1)

      *****∀E ↑↑ WISE3,KW(WISE1,WHITE3),RW,W1;

      55 (T(K(FOOL,KW(WISE1,WHITE3)),RW)∧A(WISE3,RW,W1))⊃T(K(FOOL,KW(WISE1,WHITE3)),W1)

      *****∀E ↑↑↑ WISE3,KW(WISE2,WHITE1),RW,W1;

      56 (T(K(FOOL,KW(WISE2,WHITE1)),RW)∧A(WISE3,RW,W1))⊃T(K(FOOL,KW(WISE2,WHITE1)),W1)

      *****∀E 53 WISE3,KW(WISE2,WHITE3),RW,W1;

      57 (T(K(FOOL,KW(WISE2,WHITE3)),RW)∧A(WISE3,RW,W1))⊃T(K(FOOL,KW(WISE2,WHITE3)),W1)

      *****∀E 53 WISE3,OR(WHITE1,OR(WHITE2,WHITE3)),RW,W1;

      58 (T(K(FOOL,OR(WHITE1,OR(WHITE2,WHITE3))),RW)∧A(WISE3,RW,W1))⊃T(K(FOOL,OR(WHITE1,OR(WHITE2,WHITE3))),W1)

      ***** WISEMAN2,25,54;

59 T(K(FOOL,KW(WISE1,WHITE2)),W1)  (25)

***** WISEMAN3,25,55;

60 T(K(FOOL,KW(WISE1,WHITE3)),W1)  (25)

***** WISEMAN4,25,56;

61 T(K(FOOL,KW(WISE2,WHITE1)),W1)  (25)

***** WISEMAN5,25,57;

62 T(K(FOOL,KW(WISE2,WHITE3)),W1)  (25)

***** WISEMAN8,25,58;

63 T(K(FOOL,OR(WHITE1,OR(WHITE2,WHITE3))),W1)  (25)

	In W1 we have that WISE2 doesn't know  WHITE2, so we now find
a  world in which WHITE2  is false accessible  to WISE2 from  W1.  We
call it W2 and 69 and 70 give these results.

      *****∀E KNOW WISE2,WHITE2,W1;

      64 T(K(WISE2,WHITE2),W1)≡∀W.(A(WISE2,W1,W)⊃T(WHITE2,W))

      *****∀E NOT W1,K(WISE2,WHITE2);

      65 T(NOT(K(WISE2,WHITE2)),W1)≡¬T(K(WISE2,WHITE2),W1)

      ***** 41,64:65;

      66 ¬∀W.(A(WISE2,W1,W)⊃T(WHITE2,W))  (25)

      *****UNIFY↑;

      67 ∃W.¬(A(WISE2,W1,W)⊃T(WHITE2,W))  (25)

      *****∃E ↑ (W2);

      68 ¬(A(WISE2,W1,W2)⊃T(WHITE2,W2))  (68)

      ***** 68;

69 A(WISE2,W1,W2)  (68)

***** 68;

70 ¬T(WHITE2,W2)  (68)

	Now we  have to establish  the facts  about W2, getting  them
from the facts about W1 by the same methods as got the facts about W1
from RW.  The additional facts we need are steps 76, 80, 86 and 91.

      *****∀E 18 FOOL,KW(WISE2,WHITE1),W1;

      71 T(K(FOOL,KW(WISE2,WHITE1)),W1)⊃T(KW(WISE2,WHITE1),W1)

      *****∀E 11 WISE2,WHITE1,W1;

      72 (T(WHITE1,W1)∧T(KW(WISE2,WHITE1),W1))⊃T(K(WISE2,WHITE1),W1)

      *****∀E KNOW WISE2,WHITE1,W1;

      73 T(K(WISE2,WHITE1),W1)≡∀W.(A(WISE2,W1,W)⊃T(WHITE1,W))

      ***** 32,61,71:73;

      74 ∀W.(A(WISE2,W1,W)⊃T(WHITE1,W))  (25)

      *****∀E ↑ W2;

      75 A(WISE2,W1,W2)⊃T(WHITE1,W2)  (25)

      *****⊃E 69,↑;

76 T(WHITE1,W2)  (25 68)

      *****∀E KNOW WISE2,NOT(K(WISE1,WHITE1)),W1;

      77 T(K(WISE2,NOT(K(WISE1,WHITE1))),W1)≡∀W.(A(WISE2,W1,W)⊃T(NOT(K(WISE1,WHITE1)),W))

      ***** 40,77;

      78 ∀W.(A(WISE2,W1,W)⊃T(NOT(K(WISE1,WHITE1)),W))  (25)

      *****∀E ↑ W2;

      79 A(WISE2,W1,W2)⊃T(NOT(K(WISE1,WHITE1)),W2)  (25)

      *****⊃E 69,↑;

80 T(NOT(K(WISE1,WHITE1)),W2)  (25 68)

      *****∀E 53 WISE2,KW(WISE1,WHITE2),W1,W2;

      81 (T(K(FOOL,KW(WISE1,WHITE2)),W1)∧A(WISE2,W1,W2))⊃T(K(FOOL,KW(WISE1,WHITE2)),W2)

      *****∀E 53 WISE2,KW(WISE1,WHITE3),W1,W2;

      82 (T(K(FOOL,KW(WISE1,WHITE3)),W1)∧A(WISE2,W1,W2))⊃T(K(FOOL,KW(WISE1,WHITE3)),W2)

      *****∀E 53 WISE2,OR(WHITE1,OR(WHITE2,WHITE3)),W1,W2;

      83 (T(K(FOOL,OR(WHITE1,OR(WHITE2,WHITE3))),W1)∧A(WISE2,W1,W2))⊃T(K(FOOL,OR(WHITE1,OR(WHITE2,WHITE3))),W2)

      *****∀E 18 FOOL,KW(WISE1,WHITE2),W2;

      84 T(K(FOOL,KW(WISE1,WHITE2)),W2)⊃T(KW(WISE1,WHITE2),W2)

      *****∀E 18 FOOL,KW(WISE1,WHITE3),W2;

      85 T(K(FOOL,KW(WISE1,WHITE3)),W2)⊃T(KW(WISE1,WHITE3),W2)

      ***** 59,68,81,84;

86 T(KW(WISE1,WHITE2),W2)  (25 68)

      ***** 60,68,82,85;

      87 T(KW(WISE1,WHITE3),W2)  (25 68)

      ***** 63,68,83;

      88 T(K(FOOL,OR(WHITE1,OR(WHITE2,WHITE3))),W2)  (25 68)

      *****∀E KW WISE1,WHITE2,W2;

      89 T(KW(WISE1,WHITE2),W2)≡(T(K(WISE1,WHITE2),W2)∨T(K(WISE1,NOT(WHITE2)),W2))

      *****∀E 18 WISE1,WHITE2,W2;

      90 T(K(WISE1,WHITE2),W2)⊃T(WHITE2,W2)

      ***** 70,86,89:90;

91 T(K(WISE1,NOT(WHITE2)),W2)  (25 68)

	In W2,  WISE1  doesn't know  WHITE1, so  there has  to be  an
accessible world  from W2 in which WHITE1  is false, and establishing
this world W3 is accomplished at steps 97 and 98.

      *****∀E KNOW WISE1,WHITE1,W2;

      92 T(K(WISE1,WHITE1),W2)≡∀W1.(A(WISE1,W2,W1)⊃T(WHITE1,W1))

      *****∀E NOT W2,K(WISE1,WHITE1);

      93 T(NOT(K(WISE1,WHITE1)),W2)≡¬T(K(WISE1,WHITE1),W2)

      ***** 80,92:93;

      94 ¬∀W1.(A(WISE1,W2,W1)⊃T(WHITE1,W1))  (25 68)

      *****UNIFY↑;

      95 ∃W1.¬(A(WISE1,W2,W1)⊃T(WHITE1,W1))  (25 68)

      *****∃E ↑ (W3);

      96 ¬(A(WISE1,W2,W3)⊃T(WHITE1,W3))  (96)

      ***** 96;

97 A(WISE1,W2,W3)  (96)

***** 96;

98 ¬T(WHITE1,W3)  (96)

	The only additional properties we have to  transfer to W3 are
that  WHITE2 is false and  the disjunction WHITE1 ∨  WHITE2 ∨ WHITE3.
The first is established  at step 103  and the second is  essentially
transferred  at step  106 though  not isolated.   These  are used  to
establishe WHITE3 in W3 at step 109.

      *****∀E KNOW WISE1,NOT(WHITE2),W2;

      99 T(K(WISE1,NOT(WHITE2)),W2)≡∀W1.(A(WISE1,W2,W1)⊃T(NOT(WHITE2),W1))

      ***** 91,99;

      100 ∀W1.(A(WISE1,W2,W1)⊃T(NOT(WHITE2),W1))  (25 68)

      *****∀E ↑ W3;

      101 A(WISE1,W2,W3)⊃T(NOT(WHITE2),W3)  (25 68)

      *****∀E NOT W3,WHITE2;

      102 T(NOT(WHITE2),W3)≡¬T(WHITE2,W3)

      ***** 97,101:102;

103 ¬T(WHITE2,W3)  (25 68 96)

      *****∀E 53 WISE1,OR(WHITE1,OR(WHITE2,WHITE3)),W2,W3;

      104 (T(K(FOOL,OR(WHITE1,OR(WHITE2,WHITE3))),W2)∧A(WISE1,W2,W3))⊃T(K(FOOL,OR(WHITE1,OR(WHITE2,WHITE3))),W3)

      ***** 88,97,104;

      105 T(K(FOOL,OR(WHITE1,OR(WHITE2,WHITE3))),W3)  (25 68 96)

      *****∀E 18 FOOL,OR(WHITE1,OR(WHITE2,WHITE3)),W3;

106 T(K(FOOL,OR(WHITE1,OR(WHITE2,WHITE3))),W3)⊃T(OR(WHITE1,OR(WHITE2,WHITE3)),W3)  

      *****∀E OR W3,WHITE1,OR(WHITE2,WHITE3);

      107 T(OR(WHITE1,OR(WHITE2,WHITE3)),W3)≡(T(WHITE1,W3)∨T(OR(WHITE2,WHITE3),W3))

      *****∀E OR W3,WHITE2,WHITE3;

      108 T(OR(WHITE2,WHITE3),W3)≡(T(WHITE2,W3)∨T(WHITE3,W3))

      ***** 98,103,105:108;

109 T(WHITE3,W3)  (25 68 96)

	Now we have only to transfer WHITE3 back to W2 and finally to
W1  using the  facts  that WISE1  and WISE2  via  whose accessibility
relations the transfers have to  be made know whether WHITE3.   These
results are obtained at steps 119 and 130 respectively.

      *****∀E KW WISE1,WHITE3,W2;

      110 T(KW(WISE1,WHITE3),W2)≡(T(K(WISE1,WHITE3),W2)∨T(K(WISE1,NOT(WHITE3)),W2))

      *****∀E KNOW WISE1,NOT(WHITE3),W2;

      111 T(K(WISE1,NOT(WHITE3)),W2)≡∀W1.(A(WISE1,W2,W1)⊃T(NOT(WHITE3),W1))

      *****ASSUME ∀W1.(A(WISE1,W2,W1)⊃T(NOT(WHITE3),W1));

      112 ∀W1.(A(WISE1,W2,W1)⊃T(NOT(WHITE3),W1))  (112)

      *****∀E ↑ W3;

      113 A(WISE1,W2,W3)⊃T(NOT(WHITE3),W3)  (112)

      *****∀E NOT W3,WHITE3;

      114 T(NOT(WHITE3),W3)≡¬T(WHITE3,W3)

      ***** 97,109,113:114;

      115 FALSE  (25 68 112)

      *****⊃I 112⊃↑;

      116 ∀W1.(A(WISE1,W2,W1)⊃T(NOT(WHITE3),W1))⊃FALSE  (25 68)

      ***** 87,110:111,116;

      117 T(K(WISE1,WHITE3),W2)  (25 68)

      *****∀E 18 WISE1,WHITE3,W2;

      118 T(K(WISE1,WHITE3),W2)⊃T(WHITE3,W2)

      *****⊃E ↑↑,↑;

119 T(WHITE3,W2)  (25 68)

      *****∀E 18 FOOL,KW(WISE2,WHITE3),W1;

      120 T(K(FOOL,KW(WISE2,WHITE3)),W1)⊃T(KW(WISE2,WHITE3),W1)

      *****∀E KW WISE2,WHITE3,W1;

      121 T(KW(WISE2,WHITE3),W1)≡(T(K(WISE2,WHITE3),W1)∨T(K(WISE2,NOT(WHITE3)),W1))

      *****∀E KNOW WISE2,NOT(WHITE3),W1;

      122 T(K(WISE2,NOT(WHITE3)),W1)≡∀W.(A(WISE2,W1,W)⊃T(NOT(WHITE3),W))

      *****ASSUME ∀W.(A(WISE2,W1,W)⊃T(NOT(WHITE3),W));

      123 ∀W.(A(WISE2,W1,W)⊃T(NOT(WHITE3),W))  (123)

      *****∀E ↑ W2;

      124 A(WISE2,W1,W2)⊃T(NOT(WHITE3),W2)  (123)

      *****∀E NOT W2,WHITE3;

      125 T(NOT(WHITE3),W2)≡¬T(WHITE3,W2)

      ***** 69,119,124:125;

      126 FALSE  (25 123)

      *****⊃I 123⊃↑;

      127 ∀W.(A(WISE2,W1,W)⊃T(NOT(WHITE3),W))⊃FALSE  (25)

      ***** 62,120:122,127;

      128 T(K(WISE2,WHITE3),W1)  (25)

      *****∀E 18 WISE2,WHITE3,W1;

      129 T(K(WISE2,WHITE3),W1)⊃T(WHITE3,W1)

      *****⊃E ↑↑,↑;

130 T(WHITE3,W1)  (25)

	Since W1 is an arbitrary world accessible to WISE3 it follows
that  WISE3  knows WHITE3  in  RW,  and four  more  steps  suffice to
establish this.

      *****⊃I 25⊃↑;

      131 A(WISE3,RW,W1)⊃T(WHITE3,W1)

      *****∀I ↑ W1;

      132 ∀W1.(A(WISE3,RW,W1)⊃T(WHITE3,W1))

      *****∀E KNOW WISE3,WHITE3,RW;

      133 T(K(WISE3,WHITE3),RW)≡∀W1.(A(WISE3,RW,W1)⊃T(WHITE3,W1))

      ***** 132:133;

134 T(K(WISE3,WHITE3),RW)