Neural mechanisms behind imagined experience

Abstract

Mental images seem to involve a sensory experience of our thoughts, yet one which seems inferior to the sensory experience of ordinary veridical perception. While accounts have proposed how this difference might arise, none have first specified what exactly a quasi-sensory experience involves. This thesis argues that an absence of frameworks delineating the perceptual composition of imagined experience has prompted the field to rely on flawed methods of quantifying imagery quality, limiting mechanistic explanations into why mental images are quasi-sensory while other percepts are not. Consequently, Chapter 1 investigates the specific ways in which visual and auditory mental images are experienced, providing a wealth of insights into the nature of mental image phenomenology and how it varies depending on modality-specific idiosyncrasies in perceptual processing. Chapter 2 continues by describing a framework for perception which allows for quasi-sensory experience, then shows how interference between real and imagined stimuli could induce such an experience. This chapter argues that mental imagery should always be viewed in light of competition from external sensory inputs, using a computational model to show how interference could prevent imagery from influencing early sensory cortices and becoming visible. To understand how external stimuli interfere with mental images, Chapter 3 uses neural decoding to probe whether features of imagined and real stimuli constructively or destructively interact in the brain. Findings indicate a small constructive interaction effect consistent with the view that imagery is a late-stage modulator of perceptual processing with little impact on early-stage perception. Finally, this thesis concludes with a discussion of how to reconcile the consensus view that mental imagery is a depictive process involving depictive phenomenology despite evidence that early-stage perceptual processes often seem unnecessary for its construction.