Mind wandering (MW) is a pervasive phenomenon that occurs very frequently regardless of the task. A content-based definition of MW holds that MW occurs when the content of thought switches from an on-going task and/or external stimulus-driven event to self-generated or inner thoughts. A recent account suggests that the transition between these different states of attention occurs via an off-focus state. Following this suggestion, previous work relating MW to pupil size might have lumped attentional states that are critically different from each (i.e., an off-focus and a MW state). In the current study, both behavior and pupil size were measured during a Sustained Attention to Response Task (SART) to disentangle the content of thought (on-task or MW) from an off-focus state of mind. The off-focus state was operationalized by probing the intensity with which participants were on-task or MW. Results of two experiments showed that the behavioral and phasic pupillary responses were sensitive to changes related to the content of thought. Behavioral responses were furthermore related to the intensity of the thought. However, no clear relation between the different attentional states and the tonic pupillary diameter was found suggesting it is an unreliable proxy for MW.

Mind wandering (MW) is a highly prevalent phenomenon despite its negative consequences on behaviour. Current views about its origin share the idea that MW occurs due to changes in the executive functions system. Here, we argue that not all instances of MW are necessarily related to changes in that system. Combining results from MW and sleep research, we propose that MW could also be initiated by the depletion of resources in primary task-related networks. To test this hypothesis, participants performed four sessions of the texture discrimination task (TDT) on a day. The TDT is a perceptual learning task in which performance is related to the local build-up of sleep-pressure. During the TDT, MW was recorded in both a subjective (i.e., with thought probes) and an objective (i.e., phasic pupillary response) manner. Results showed that the pattern of behaviour in the TDT was mirrored in the objective measure of MW. Also for the subjective measure the pattern was similar to that of task performance, but it failed to reach significance. These results demonstrate that not all MW is necessarily related to changes in the executive system and support the hypothesis that MW can be related to the depletion of local resources.

Mind wandering (MW) or having thoughts unrelated to the task at hand is a very pervasive phenomenon. Although research on MW has exponentially grown during the last decade and a half, the mechanisms behind this omnipresent phenomenon remain largely unknown. In this review, we will discuss some factors that have been shown to contribute to the occurrence of MW: the quality of sleep, the time of day when the task is performed, the chronotype of the individual and the duration of the task. The intriguing commonality between these specific factors is that they all suggest a relation between MW and “fatigued” brain states. This suggestion nicely fits a recent proposal that related MW to local “fatigued” brain areas involved in task performance. We will argue that one of the mechanisms underlying the pervasiveness of MW might be the local build-up of homeostatic sleep pressure that inevitably occurs during task performance in the brain areas related to the task.