On the axiomatic foundations of the integrated information theory of consciousness.

The integrated information theory (IIT) is a prominent scientific framework for studying consciousness and developing a consciousness meter. This paper critically examines the axiomatic method that IIT's creators use. The author argues that the method is open to multiple interpretations and that none of IIT's five alleged axioms can play the required role: each either fails to be truly axiomatic or fails to impose a substantive constraint on a theory of consciousness. The article concludes by outlining an alternative methodology for the science of consciousness, called the natural kind approach.

Study at a glance

Abstract

The integrated information theory (IIT) is one of the most influential scientific theories of consciousness. It functions as a guiding framework for a great deal of research into the neural basis of consciousness and for attempts to develop a consciousness meter. In light of these developments, it is important to examine whether its foundations are secure. This article does just that by examining the axiomatic method that the architects of IIT appeal to. I begin by asking what exactly the axiomatic method involves, arguing that it is open to multiple interpretations. I then examine the five axioms of IIT, asking: what each axiom means, whether it is indeed axiomatic and whether it could constrain a theory of consciousness. I argue that none of the five alleged axioms is able to play the role that is required of it, either because it fails to qualify as axiomatic or because it fails to impose a substantive constraint on a theory of consciousness. The article concludes by briefly sketching an alternative methodology for the science of consciousness: the natural kind approach.