[en] Animal and human brains contain a myriad of mental representations that have to be successfully tracked within fractions of a second in a large number of situations. This retrieval process is hard to explain without postulating the massive modularity of cognition. Assuming that the mind is massively modular, it is then necessary to understand how cognitive modules can efficiently represent dynamic environments – in which some modules may have to deal with change-induced novelty and uncertainty. Novelty of a stimulus is a problem for a module when unknown, significant stimuli do not satisfy the module’s processing criteria – or domain specificity – and cannot therefore be included in its database. It is suggested that the brain mechanisms of incentive motivation, recruited when faced with novelty and uncertainty, induce transient variations in the domain specificity of cognitive modules in order to allow them to process information they were not prepared to learn. It is hypothesised that the behavioural transitions leading from exploratory activity to habit formation are correlated with (and possibly caused by) the organism’s ability to counter novelty-induced uncertainty.