A comparative analysis of empirical theories of consciousness.

Abstract

The search for the neural correlates of consciousness has brought many theories of consciousness with it. The abundance of theories points to the need to examine the empirical background supporting each particular theory. This research identified the most common theories of consciousness (global neuronal workspace theory [GNWT], integrated information theory [IIT], and predictive processing) and the most widely used experimental paradigms in consciousness research: masking, filling-in, ambiguous figures, dichotic listening, tip-of-the-tongue phenomenon, binocular rivalry, attentional blink, inattentional blindness, change blindness, biological motion, phosphenes, no-report paradigm, the effect of expectations on perception, and ensemble perception. The selected theories and the underlying neurobiology were reviewed and the extent to which the theories have been supported by experiments using the selected experimental paradigms was investigated through a systematic search. Forty-one empirical articles were found in support of GNWT, 36 in support of predictive processing, and three in support of IIT. Different theories mostly used different experimental paradigms and partly different measuring instruments, which somewhat restricts the theories when aspiring for universality. The possible impact of these differences on the generalizability of theories is presented in the discussion, and the results are compared (a) with other empirical works not specifically targeted here and (b) with a similar systematic review.