Learning an Ordered Series: An fMRI Study

Results from lesion studies in animal research and imaging studies in human research have shown the importance of hippocampus (HC) in the representation of ordered stimuli. More specifically, HC has been found to be involved in judgments about non-adjacent pairs (A-C; A-D; B-D;...) after being trained only on adjacent pairs (A-B; B-C; C-D;...). fMRI studies in humans suggest that also intraparietal cortex is involved: Parietal cortex activation is observed during transitive inference tasks (e.g. Acuna et al., 2003; Marshuetz et al., 2000). However, parietal activation could be confounded with task difficulty. The aim of this experiment is threefold. First, we want to integrate previous findings on hippocampus and parietal sulcus in one study. Second, we want to differentiate whether hippocampus is responsible for learning new associations or for transitive inference specifically. Third, we want to disentangle parietal activation from its possible confound with task difficulty. Experiment Procedure Sixteen male, right-handed participants had to learn an ordered series of six arbitrary stimuli while being scanned. They were shown 24 blocks of triplets. Each triplet consisted of a learn, test, and control phase. In each phase two stimuli were shown simultaneously. During the learn and test phase participants were instructed to select the ‘last’ stimulus in the order. Only adjacent pairs where presented in the learn phase, whereas the test phase consisted only of non-adjacent pairs. Importantly, during the last four blocks, new figures were presented in the learn and test phase, meaning that the participants had to restart learning