COMMENT ⊗   VALID 00002 PAGES
C REC  PAGE   DESCRIPTION
C00001 00001
C00002 00002	ideas[w87,jmc]		virtual mediation
C00006 ENDMK
C⊗;
ideas[w87,jmc]		virtual mediation

An article representing a virtual mediation between the Israelis and
Arabs - perhaps the report of a virtual mediator.  Also between the
South African Gov't and various groups of blacks.

jan 25 - multiple vector states or situations

Suppose we have two state vectors  xi1  and  xi2, each of which has
many components.  The processes changing  xi1  and  xi2  may be
 {\it relatively asynchronous}.  Each of them is subject to assignments.
It need not be required that the components of  xi1  or  xi2  actually
have the values assigned to them at any actual time --- merely that
subsequent states are assigned as if they did.

mar 7

from bicycle notebook

Uses of counterfactuals
	excuses - If x hadn't happened ...
	blames - If you had done x ...
Would you have taken the course for 4 or 5 units if ...
(note that this leads to a generalization of ordinary counterfactuals,
because there can be a parameter, e.g. How does whether the number of
units you would have taken the course for depend on the number of papers
that might have been assigned and the grading system).
	superfluous use? - "If that building weren't there you could see
it from here" meaning "it's behind that building".

Elephant representation of procedure calls.
procedure foo(u,v) begin
	w ← u+v;
	return w↑2 end;

called by y ← foo(x,y);

pc(t) = 5 ⊃ xi(t+1) save(a(pc,foo,
		a(arg1,c(x,xi(t)),
			a(arg2,c(y,xi(t)),
				a(ret,5',xi(t))))))

pc(t) = 5' ⊃ xi(t+1) = inc(a(y,retval,xi(t)))

where

inc(xi) = a(pc,c(pc,xi)+1,xi).

pc(t) = foo ⊃ xi(t+1) = inc(a(u,c(arg1,xi(t)),a(v,c(arg2,xi(t)),xi(t))))

pc(t) = foo+1 ⊃ xi(t+1) = inc(a(w,c(u,xi(t))+c(v,xi(t)),xi(t)))

pc(t) = foo+2 ⊃ xi(t+1) = unsave(a(pc,c(ret,xi(t)),a(retval,
				c(w,xi(t))↑2,xi(t))))

save(xi) = a(stack,c(ret,xi) . c(stack,xi),xi)

unsave(xi = a(stack,cdr c(stack,xi)),a(ret, car c(stack,xi),xi))

Perhaps this isn't quite right.  Maybe more needs to be taken on and
off the stack.

Queries
	1. discuss
	2. find unintended models
	3. circumscribe and make first order
		a. and b. depend on what's varied
	4. axiomatize (towers?). What is being reified?
	5. Kowalski's event calculus

Towers
	∃x.on(x,table,result(prog,S0)) ∧ is-traffic-light(x)
		∧? exists(x,s)

X0 = {A,B,C,D,color(A,Red),color(B,Green),color(C,Brown),color(D,Yellow),
	ontop(A,B),ontop(B,C),ontop(C,D),ontop(D,Table)}

is-traffic-light x ≡ ...