CLOSED WORLD ASSUMPTIONS

	It seems to me that Reiter (1978) overdoes the advantages in
compactness of the closed world assumption.  My intuition is that it
has other advantages, but I'm not sure what they are.

	Suppose we want to list a qA set exhaustively.  We then write

qall x.(x qin A qequiv x = a qor x = b qor x = c) qand card A = 3.

	We should list the set of flights of Air Canada (to use
Reiter's example) by listing an abstract set qA as above and then
asserting the identity of the set of flights with qA.

	In this way we can get the compactness of the closed world
assumption without assuming that we know all, since we can put some
qualification on the assertion that qA is precisely the set of
Air Canada flight.