Computer Science And Recursion Theory

• Authors: Marian Boykan Pour-El
• Venue: ACM '78 Proceedings of the 1978 annual conference
• Year: 1978

One of the basic concerns of classical mathematical logic has been a rigorous definition of “mathematical proof". A proof may be discovered by luck, genius or accident. But once discovered it is mechanically checkable. Thus the most general type of mechanically checkable procedure provides a definition of the most general kind of proof. An investigation of “mechanical checkability" leads naturally to the notion of “computable process". Recursion theory is that branch of mathematical logic which studies computability theory. From its very inception recursion theory has been so closely associated with theoretical computer science that it is sometimes difficult to tell where one begins and the other ends. Both appear to have a common origin in the 1930's and 1940's - when the fundamental results of Turing, Post, Herbrand-Gödel, Kleene and Church on computability were obtained [52,53,35,36,19,25,26,11,12,13]. Although the work of each of these authors is based on a different intuition as to the nature of computability, they all share a common feature. In each, the approach is discrete and atomic. No attempt was made to use machines with continuously varying parameters - e.g. analog computers - as a basis for understanding computability. The author finds this regrettable and is trying to remedy the situation [39]. At all events, all of these discrete, atomic definitions have been proved to be equivalent. It is now widely accepted that the concept “computable function of natural numbers" is correctly identified with “partial recursive function". Lack of space forces us to be selective rather than exhaustive. We discuss briefly five areas of interest to the computer scientist. The purpose of this paper is to explore the interaction of recursion theory with computer science in these areas. We consider, not only similarities between the two fields, but also points of difference. We discuss how recursion theory supplied computer science with specific mathematical definitions and techniques. We also consider how the computer scientist shaped (and is still shaping) this recursion-theoretic material to his own needs. The five areas are not mutually exclusive: there are numerous connections between them. In the interest of clarity and brevity, reference is made only to those works which are easily understandable to the beginner. No attempt is made to make the bibliography complete or up-to-date.