[Prev][Next][Index][Thread]

typed eta reduction



The analogue of the eta reduction

	\x:tau. fx   >     f
		      eta

in dynamic logic would be the equivalence

	[p?]q = q

This is unsound in dynamic logic.  Why should it be sound in a typed
lambda calculus?
-v