Involuntary refreshing of mental representations.

Laboratory tasks have revealed that mental representations (e.g., mental imagery) can enter consciousness in a manner that is involuntary, reliable, and insuppressible. These effects illuminate the capacities of involuntary processes as well as the function of voluntary, conscious processing. The Reflexive Imagery Task was developed a decade ago to investigate these involuntary effects systematically. Can refreshing yield such involuntary effects? Refreshing is the reactivating in mind of a mental representation that was activated moments ago. It is associated with mental rehearsal and executive function. We investigated whether a mental representation (subvocalization of an object name) can arise in consciousness involuntarily after a delayed interval, when the relevant stimulus is no longer present, and in response to a cue. In Experiment 1, participants were instructed not to refresh a previously presented (6 s before) stimulus in response to a cue. Involuntary refreshing occurred on a substantive proportion (0.56) of the trials. Experiment 2 replicated and extended this finding (proportion of the trials = 0.53) with a refreshing task that was more challenging than that of Experiment 1. Our findings suggest that mental representations arising from processes such as refreshing can occur involuntarily. We discuss the theoretical implications of this conclusion.

The Generation of Involuntary Mental Imagery in an Ecologically-Valid Task.

Laboratory tasks (e.g., the flanker task) reveal that incidental stimuli (e.g., distractors) can reliably trigger involuntary conscious imagery. Can such involuntary effects, involving competing representations, arise during dual-task conditions? Another concern about these laboratory tasks is whether such effects arise in highly ecologically-valid conditions. For example, do these effects arise from tasks involving dynamic stimuli (e.g., simulations of semi-automated driving experiences)? The data from our experiment suggest that the answer to our two questions is yes. Subjects were presented with video footage of the kinds of events that one would observe if one were seated in the driver's seat of a semi-automated vehicle. Before being presented with this video footage, subjects had been trained to respond to street signs according to laboratory techniques that cause stimulus-elicited involuntary imagery. After training, in the Respond condition, subjects responded to the signs; in the Suppress condition, subjects were instructed to not respond to the signs in the video footage. Subjects in the Suppress condition reported involuntary imagery on a substantive proportion of the trials. Such involuntary effects arose even under dual-task conditions (while performing the n-back task or psychomotor vigilance task). The present laboratory task has implications for semi-automated driving, because the safe interaction between driver and vehicle requires that the communicative signals from vehicle to driver be effective at activating the appropriate cognitions and behavioral inclinations. In addition, our data from the dual-task conditions provide constraints for theoretical models of cognitive resources.