Hi. I am playing around with basic Haskell overloading. What I'm interested in is how to do specialization in Haskell -- i.e., functions that work generically on all (or many) types but work more efficiently on certain types. So, I am trying to make a class of functions which can be fed into a partial sum calculator. Maybe something like so: class PartialSum f where -- params: term function, end index partialSum :: Integral b => f -> b -> a The most generic instance would be any function that takes an integer and returns a number. The specialized instances would things like, say, a wrapped function which is guaranteed to be linear (through safe constructors or something). But I'm having trouble figuring out how even to make the generic version. I'm thinking something like this: instance PartialSum (a -> b) where partialSum f n = foldl (\u v -> u + f v) 0 [1..n] But the compiler complains it can't prove that the input to the "f" function is the same type as the "n" parameter. That makes sense, but I'm not sure how to fix that.