TRoPICALS: A Computational Embodied Neuroscience Model of Compatibility Effects

Recent evidence shows that the perception of objects automatically activates the representation of their affordances. For example, some experiments found compatibility effects between the size of objects (small/large) and the kind of grip (precision/power) required to categorise them as natural or as artefacts, and between common location of object parts (top or bottom) and the kind of movement (up and down) required to indicate whether or not these parts belong to a whole object. This article presents a neural-network model that reproduces these results and also provides a general framework to account for several other types of compatibility effects. This model is based on four general principles: (a) visual perception and action are organised along a dorsal neural pathway (encoding affordances) and a ventral pathway; (b) within the ventral pathway, the prefrontal cortex biases action selection based on context and goals; (c) action selection results from neural dynamic competitions that cause variable reaction times; (d) words trigger “internal simulations” of their referents. The model was designed within a methodological approach that aims at developing it cumulatively so as to furnish increasingly general and comprehensive accounts of compatibility effects. The approach imposed four types of constraints on the model: (a) neuroscientific constraints on their architecture and functioning; (b) reproduction of specific psychological experiments; (c) functioning within an embodied system; (d) reproduction of the learning processes that result in the target behaviours. The claim on the generality of the model is supported by a critical comparison with other models that are related to the above four principles and by an analysis of how the model could be developed to account for other compatibility effects. The heuristic power of the model is also shown by presenting two testable predictions