Search for an integrated, comprehensive theory of information science (IS) has so far been unsuccessful. Appearance of a theory has been retarded by one central constraint, the large number of disciplines concerned with human communication. Crossdisciplinary interdependence occurs in two ways: theoretical relation of IS phenomena to a given science, and practical relation of the science to IS dissemination of its contributions. We are concerned here with the first. The main difficulty in making IS models is the above mentioned interaction of many sciences of communication. Prominent among these are physics, biology, psychology, library science, computer science, several social sciences, applied logic and mathematics, and not least, cybernetics. That cybernetics should apply to IS is not a new idea. But more precision has emerged recently. Cybernetics is now seen to underlie nearly all IS phenomena through two central concepts: variety, and the law of requisite variety. The last provides a quantitative approach to system regulation and control. IS concerns regulatory activity in typically large, goal-seeking systems. That is, propagation of meaningful human messages between sender and recipient on an "IS path" is a goal-seeking process. (Author)