Cognitive control of attention, emotion, and memory : an ERP study

Unwanted retrieval of negative memories can be problematic for many clinical populations. The Think/No-Think (T/NT) task (Anderson & Green, 2001) is a new paradigm for studying cognitive control during cued recall. In this task participants view a cue item and are asked to consciously retrieve (think) or interrupt retrieval (no-think) of the associated target item. Eyer (2009) found that self-reported mindfulness was correlated with T/NT cued recall, suggesting a relationship between control of memory retrieval and a general cognitive control skill. The current study measured event-related potentials (ERPs; i.e., electrical brain responses time-locked to cue presentation) for negative and neutral stimuli on the TNT task to assess cognitive control during retrieval. Method: Participants (N = 35) completed questionnaires (e.g., mindfulness, intrusive thoughts) and cognitive tasks related to cognitive control (e.g., attention, working memory span). Then, ERPs were recorded during the TNT task, followed by a final cued recall test. Results: Analyses of ERPs found evidence to support somewhat separable neural networks for control of memory retrieval and for processing the emotional content of the target pictures, with some time windows only exhibiting a main effect of strategy or of emotional valence. However, there was widespread evidence for interactions of these subsystems across a range of time latencies post-cue presentation. Of particular note was a significant Strategy x Valence interaction for the early P1 component (125-164 ms). The overall size of the N2 (250–324 ms) peak was correlated with a wide range of self- report and cognitive test measures of cognitive control at frontal electrode sites. Discussion: The present study adds to knowledge of the timing of control processes during performance of the TNT task through its use of ERP methodology. The effect of the emotional valence of the to-be-recalled target on the early P1 ERP component suggests surprisingly early emotional processing during memory retrieval. The present results also suggest that at least some of the control processes used during the TNT task are part of a larger general-purpose cognitive control system. These results suggest that individual traits provide important and varying influences on the cognitive control of emotional memories