More rapid reversal learning following overtraining in the rat is evidence that behavioural and cognitive flexibility are dissociable

Abstract

Cognitive flexibility is a term used to describe the brain processes underlying the phenomenon of adaptive change in behaviour in response to changed contingencies in the internal or external environment. Cognitive flexibility is often assessed in complex tasks measuring perceptual attentional shifting or response or task switching, but, arguably, reversal learning is a simple assay of cognitive flexibility. Reversal learning requires the detection of a changed outcome, the cessation of a previously-rewarded response and the selection of an alternative, previously-unrewarded, response. This study addressed the issue of the relationship between reversal learning and cognitive flexibility. In a single testing session, rats completed a series of 2-alternative forced-choice discriminations between digging bowls. The bowls differed according to both the medium within the bowl and the odor of the bowl. Having learned which cue (one of the odors or one of the digging media) indicated the food-baited bowl, half the rats were given additional trials of “over-training”. To test reversal learning, the meaning of the cues predictive of reward/non-reward was then switched. There was a robust effect of over-training, with over-trained rats performing reversal learning in fewer trials than rats trained to criterion only. The pattern of errors supported the hypothesis that more rapid reversing results from the formation of an attentional set. This is the same attentional mechanism that results in less rapid shifting or switching. We conclude that the behavioural flexibility demonstrated in reversal learning does not provide a scale on which cognitive flexibility can be measured.