I recall seeing somewhere a paper on the limited use of classical logic in functional programming, that somehow negation types had an interpretation under certain circumstances. I know I downloaded the paper, but I don't know where I put it. ANy pointers? Jim Apple