The contribution of proprioception to size constancy in perception and action: an fMRI study

Size constancy allows the visual system to maintain stable representations of object size despite changes in viewing distance, supporting accurate perception and action. Although extensively studied in vision, the mechanisms underlying size constancy remain only partially understood, particularly regarding the contribution of non-visual signals such as proprioception. Here, we investigate how proprioceptive depth cues are integrated with retinal information to support size constancy in both perception and action. We conducted an fMRI experiment using glow-in-the-dark spheres presented at two viewing distances (near vs. far) and two physical sizes (2.5 cm vs. 5 cm). The experiment was performed under restricted viewing conditions (darkness; monocular viewing through a 1-mm pinhole) to remove all depth cues. Participants were asked to either grasp or manually estimate its diameter with their right hand, while their left hand held the pedestal on which the sphere was placed, providing a measure of distance. Regions of interest were independently defined using an object localizer to identify ventral and dorsal stream regions, including the lateral occipital complex (LOC) and intraparietal sulcus (IPS), and a retinotopic localizer to map early visual cortex (V1). Preliminary findings reveal a functional dissociation between visual pathways: ventral stream regions within LOC are primarily sensitive to task demands and object size, whereas dorsal stream regions within IPS are preferentially modulated by viewing distance. These findings suggest that size constancy emerges from complementary contributions of ventral and dorsal pathways, integrating retinal and proprioceptive information to support stable perception and action across changes in distance.