Another good exam question (Hmm!): What does last (last (map return [1..])) lastly return given that last (return (not True))? I also would prefer "unit". "return" makes sense for me as syntactic sugar in the context of a "do"-expression (and then please like an unary prefix-operat or with low binding power...). An alternative sugary would be "compute": When a monad represents a computation, "init" returns a computation with a result, not just the result: foo x = if x > 0 then compute x*x else compute -x*x By the way, an alternative for "do" would be "seq" (as in occam) to indicate that operations are sequenced: getLine = seq c <- readChar if c == '\n' then compute "" else seq l <- getLine compute c:l But such a discussion has probably already been taken place some years ago. It would be interesting for me to know the arguments that led to the choice of "return" (and "do"). Elke. --- "If you have nothing to say, don't do it here..." Elke Kasimir Skalitzer Str. 79 10997 Berlin (Germany) fon: +49 (030) 612 852 16 mail: elke.kasimir@catmint.de> see: http://www.catmint.de/elke for pgp public key see: http://www.catmint.de/elke/pgp_signature.html