Abstract

It has been often claimed that the generalizations involved in neuroscientific explanations cannot qualify as laws, as they are exception-ridden. This thesis is hardly compatible with the idea that computer or robotic simulations can serve as experimental testbeds for neuroscientific hypotheses on the mechanisms underlying adaptive and intelligent biological behaviours. It will be argued here that neuroscientific generalizations should be interpreted as exceptionless, and that nomic predictability is a necessary feature of neuroscientific explanation. This view will be supported by the analysis of simulation-based studies on sensory-motor coordination mechanisms of biological systems.