-
Views
-
Cite
Cite
Roger C. Conant, A Vector Theory of Information, Communication Yearbook, Volume 3, Issue 1, December 1979, Pages 177–194, https://doi.org/10.1080/23808985.1979.11923760
- Share Icon Share
Abstract
Useful as it is, conventional information theory is designed for dealing only with average values of information in ergodic situations and in a way that does not take into account its effect on the observer. Starting with the premise that information is associated with the relation between message and observer rather than being a property of the message itself, a quantitative but non-probabilistic theory of information is derived based on the meaning of the message for the receiver. The theory serves as a framework for discussing information in non-ergodic situations (such as learning) and for single messages, as well as under conditions for which conventional information theory applies. Consistency, relevance, intent, importance, and other aspects of communication are discussed to illustrate the theory.