Let N be a transitive model of ZFC. Let P be a partial order and G such that for some M, G is (M,P)-generic and N=M[G]. 1) Is M unique? 2) Is M definable in N?