Abstract
To recognize a familiar object, incoming sensory information is matched against memory representations. Previous findings have shown that these representations are not stable but adapt flexibly to recently encountered perceptual input, resulting in measurable aftereffects (i.e., prolonged exposure to one sensory input influences the perception of – and shifts the response to – the subsequent one). So far, evidence for such flexible representations has largely been limited to human faces, due to their high social relevance. However, research shows that the cognitive system prioritizes animate objects in general. Therefore, our aim was to use biologically plausible stimuli to test whether adaptation aftereffects generalize to visual representations of alternative categories of animate objects. We created a standardized image set of crustaceans by modelling their evolution on a morphing continuum between endpoints “lobster” and “crab” by systematically modifying their physical features (e.g., carapace shape and the angle of claws). The endpoint images served as adaptors and seven intermediate morphs as target stimuli. In Experiment 1, we manipulated the adaptors´ low-level features (size and orientation) in three conditions (same-sized upright, enlarged by 25%, and inverted), and found robust aftereffects in each. Upon comparing their magnitude, the effects were the largest in the first condition, smaller in the second, and relatively smallest in the third, indicating an only minor contribution of low-level adaptation and thus the involvement of higher-level mechanisms. In Experiment 2, the manipulation of the inter-stimulus interval (ISI) between adaptor and targets (0, 1200, or 2400 ms) also resulted in significant aftereffects, these being the largest at 1200 ms, intermediate at 0 ms, and comparatively smallest at 2400 ms, demonstrating their temporal robustness. Taken together, our results – in line with prior findings on human faces – imply high-level influences, and generalize the phenomenon of flexible representations to animate objects beyond faces.