Abstract

This article is part of a special issue of Sistemi Intelligenti dedicated to the memory of Domenico Parisi. The work illustrates an original research method developed for the study of the brain and behaviour called ‘computational embodied neuroscience (CEN)’. The method integrates ‘computational neuroscience’ and the ‘embodied view’ on cognition which was greatly supported by Domenico Parisi’s research. The article first illustrates the principles of the approach, in particular the importance given to computational models able to integrate data from multiple experiments on brain and behaviour and to explain their adaptive function through embodiment. Then it illustrates the strengths of the approach. Next, the article describes a few examples of computational models in order to illustrate the outcomes of the application of the approach, and hence analyses its limitations. Finally, the article considers the future of the approach, in particular in relation to the recent success of deep neural networks, and draws conclusions in the form of an imaginary dialogue between Domenico Parisi and the author