An error in our thinking: The sensory conflict hypothesis of combined action observation and motor imagery

The capacity to learn and adapt movement and skills through non-physical forms of practice provides benefits which extend across disciplines such as rehabilitation andsports. Two non-physical forms of practice are action observation (i.e., watching an action; AO) and motor imagery (i.e., cognitively simulatingperforming an action; MI). Researchers and practitioners now often adopt an integrated approach, where one observes an action while concurrently imagining the feeling of performing the action (termed congruent AOMI). Congruent AOMI results in greater corticomotor activity and improved behavioral outcomes compared to its independent counterparts. Despite existing theoretical frameworks seeking to explain these effects, empirical findings remain inconclusive. This article critically evaluates previous hypotheses before advancing an alternative - the Sensory Conflict Hypothesis (SCH). Situated in an internal modelling framework, the SCH proposes greater corticomotor activity and improved behavioral outcomes to be driven by the occurrence and reduction of sensory prediction errors, respectively. The SCH is further discussed in the context of differing actions during AOMI, where actions are associated in some form (coordinative AOMI) or differ in both action kinematics and goals (conflicting AOMI), with implications for cognitive engagement discussed. The SCH presents new, testable avenues of research to further understandthe relationship between AO and MI, and how to optimize their integration.