Towards Meaningful Information Processing: A unifying representation for Peirce’s sign types

Abstract

An open problem in AI is the definition of meaningful information processing. That human interpretation and information processing by current computers can be different is well illustrated by Searle’s famous Chinese room argument thought experiment. In this paper we suggest that an answer to the above open problem of AI can be given by introducing a model of information processing which is embedded in a Peircean theory of (meaningful) signs. Peirce’s sign theory, that he systematically derived from his concept of a category, is seen by many as a theory of the knowable (the types of distinctions that can be signified by signs). We show that our model of information processing has the potential for representing three types of relation that are analogous to Peirce’s three classifications of sign, consisting of 10, 28, and 66 elements.