Zenith Grant Awardee

Karim Jerbi

University of Montreal

Project Title

Quale weaving: Probing integrated information theory through neurofeedback

Project Summary

For all the achievements of modern science, it still has not been able to make much sense of this thing we call consciousness. This internal life of sights, sounds, thoughts and emotions was long written off as unfit for scientific study, but now even physicists acknowledge that in order to fully understand the world we observe, we need to understand what it means to observe it. A promising hypothesis known as integrated information theory (IIT) proposes that consciousness arises wherever a physical system gains integrated knowledge about itself, over and above any of its parts. However, the measurement of integrated information (Ф) in a system as large as the brain would require more computational power than we could ever muster, and so the hypothesis has remained untested. In this project, we propose a workaround: instead of measuring Ф in the brain, we use the predictions of IIT to change the brain’s consciousness itself. We will do this by reading magnetic brain signals and feeding them back in real-time through sound and vision. If IIT is true, the subject should suddenly become conscious of the neurofeedback loop as if it’s a part of their brain, with profound implications for modern science.

Technical Abstract

Integrated information theory (IIT) hypothesizes that consciousness arises wherever a physical system emergently constrains its own past and future states. In spite of having attracted much cross-disciplinary interest, progress in validating the theory has been slowed by the computational and measurement intractability of its central metric of integrated information (Ф) for systems as large as the brain. We propose a novel and unconventional test to circumvent these limitations. Instead of attempting to measure or approximate Ф and then correlate it with objective measures of consciousness, we use the unique predictions of IIT to devise an experiment that should alter consciousness itself in otherwise inexplicable ways. Drawing on the latest advances in neural decoding, we will read a subject’s conscious activity with magnetoencephalography and, in real-time, feed this information back to the subject’s consciousness through the senses. IIT predicts that this causal weaving of disparate nodes of a conscious system, within the right timescale, will result in a rearranging of the system dynamics that will be reflected in an abrupt and dramatic change in conscious experience, which we will measure with self-report. If successful, this experiment would strongly validate IIT with profound implications for physics, epistemology and braincomputer interface technology.