Brain EM field theory of consciousness and neural synchrony

McFadden (2020) cites human studies that tie neural synchrony to conscious perception. Lutz and colleagues (2002) found that synchrony patterns correlated with conscious recognition in subjects shown optical illusions.¹ Steinmann and colleagues (2014) reported that conscious auditory perception correlated with long-range synchrony of gamma oscillations.² Leon-Carrion and colleagues (2012) showed that synchronization between anterior and posterior cortex correlated with consciousness levels in patients who had suffered traumatic brain injury.³ All three findings are correlations. McFadden counts them among many later studies showing that neural synchrony correlates with conscious perception in humans.¹

McFadden (2020) starts from the claim that consciousness represents bound, integrated information. He argues that, since information is physical, integrated information must be physically integrated.⁴ He holds that nearly all examples called integrated information, including neuronal information processing and conventional computing, are integrated only in time, because outputs are correlated with multiple inputs.⁴ On this view they cannot correspond to physically integrated information. McFadden concludes that only energy fields can integrate information in space.⁴ The cemi theory then places consciousness in the brain's global EM field.⁵

Friston, Wiese and Hobson (2020) compare the free energy principle with integrated information theory (IIT) and describe a gap between them. The formal distinction is that the free energy principle concerns probabilistic beliefs about some external thing, while IIT concerns probability distributions over the states of a system.⁶ On an informal review, they find that the information geometry implied by Markov blankets appears to meet the axiomatic criteria of IIT.⁷ Fields, Glazebrook and Levin (2021) predict that any system able to experience internal memories has a separable internal state and positive integrated information Φ. They treat this as a consequence of quantum information theory.⁸

Hoel, quoted by Bongard and Levin (2023), points out that there are many viable scales of description, including computations, and that all have some degree of integrated information.⁹ On this account the exclusion postulate is needed to get a single definite consciousness.⁹ Hoel calls it the most controversial postulate within IIT.⁹ Bongard and Levin raise this when discussing the ad hoc status of claiming one privileged perspective within a system.⁹