Cueing Cognitive Flexibility: Item-Specific Learning of Switch Readiness

The rich behavioral repertoire of the human species derives from our ability to flexibly reconfigure processing strategies (task sets) in response to changing requirements. This updating of task sets is effortful, as reflected by longer response times when switching a task than repeating it (switch costs). However, some recent data suggest that switch costs can be reduced by cueing switch readiness bottom-up, by associating particular stimuli with frequent switch requirements. This type of stimulus-control (S-C) learning would be highly adaptive, as it combines the speed of automatic (bottom-up) processing with the flexibility and generalizability of controlled (top-down) processing. However, it is unclear whether S-C learning of switch readiness is truly possible, and what the underlying mechanisms are. Here we address these questions by pairing specific stimuli with a need to update task-sets either frequently or rarely. In all 3 experiments, we observe robust item-specific switch probability (ISSP) effects as revealed by smaller switch costs for frequent switch items than for rare switch items. By including a neutral condition, we also show that the ISSP effect is primarily driven by S-C learning reducing switch costs in frequent switch items. Furthermore, by employing 3 tasks in Experiment 3, we establish that the ISSP effect reflects an enhancement of general switch readiness, rather than of the readiness to switch to a specific alternate task. These results firmly establish that switch readiness is malleable by item-specific S-C learning processes, documenting that a generalizable state of cognitive flexibility can be primed by a bottom-up stimulus. Public Significance Statement A hallmark of human behavior is cognitive flexibility, the ability to give different responses to the same stimulus, depending on our current task. Flexibly shifting between tasks is effortful, however, as reflected in switch costs: responses are slower when changing than when repeating tasks. Here we tested whether we could improve people's ability to switch by presenting particular visual stimuli (e.g., a picture of a dog) more often in a context where people were cued to switch tasks than to repeat tasks. Over three experiments, we found that people learned to associate these stimuli with a greater switch readiness, as reflected in reduced switch costs when having to shift tasks in the context of such frequent switch stimuli than of stimuli that were only rarely paired with the need to switch. This suggests that, through learning, particular stimuli can come to serve as external triggers of cognitive flexibility.