[ri′kər·siv prə′sē·jər] (computer science) A method of calculating a function by deriving values of it which become more accurate at each step; recursive procedures are explicitly outlawed in most systems with the exception of a few which use languages such as ALGOL and LISP.