My PhD thesis - Predictive Perception in Autism

My thesis investigates how predictive perception, motor coordination, and neurotype differences influences social perception in autism. It explores how these mechanisms contribute to social difficulties faced by autistic people, through the lens of the double empathy problem and second-person neuroscience.

Firstly, I investigated how spoken intentions and visual evidence guide predictive social perception. Across four experiments, results supported a dual-process model whereby top-down expectations shape early perception but are overridden by visual input once motion begins. This led to the ‘kinematic dominance hypothesis’, suggesting motion cues dominate perception once movement begins due to their higher precision.

Secondly, I used high-resolution active motion capture to analyse autistic motor coordination. Autistic participants showed differences in speed and amplitude of motion during complex tasks when compared to non-autistic. These differences correlated with communication and camouflaging traits, suggesting motor differences may influence how autistic actions are interpreted in everyday contexts, with emphasis on differences in movment in achieving task goals, not ‘deficits in ability’.

Finally, I tested the double empathy problem through predictive perception and diagnostic framing. Non-autistic participants altered perception when told the actor was autistic, despite identical kinematics. Yet, autistic participants showed stable perception regardless of actor neurotype. These findings bridge predictive perception and the double empathy problem, highlighting that difficulties experienced in autism may stem from environmental and relational factors, including assumptions of autistic behaviour by non-autistic observers, rather than measures of ability.

Together, my thesis presents a cohesive account of how social misunderstandings in autism may arise from perceptual, motoric, and contextual divergence rather than deficits in ability. It supports a neurodiversity-informed model of autistic sociality grounded in predictive processing, action kinematics, and behavioural interpretation. It recognises that the difficulties experienced by autistic people are not due to a deficit in ability, but from trying to navigate a world and social circumstance that has not been built with neurodiversity in mind. “Grant the dignity of meeting me on my own terms. Recognize that we are just as alien to each other, and that my ways of being are not merely damaged versions of yours” (Sinclair, J. ,2010).