Computable creativity: lessons learned from mental imagery research

There are similarities between the debate regarding whether creativity is computable and that which focused on the representational nature of mental imagery. While the scientific validity of studying imagery was rejected by the Behaviourist movement in the early twentieth century, there is now widespread acceptance that mental images can have functional significance. This article proposes two main lessons from the study of mental imagery can be applied to the study of creativity. First, computational models of creativity should adopt a “wet mind” approach that considers the neural structures and functions of the brain. Second, computational theories need to address the role of externalisation and external feedback during creative thinking, including the possibility that creativity may best be conceptualised as a form of extended cognition. The article concludes that recent advances in neurocomputational models of creativity provide real optimism for resolving the question of whether creativity is computable in the future.