Preparing for hard times: Scalp and intracranial physiological signatures of proactive cognitive control.

Based on reward and difficulty information, people can strategically adjust proactive cognitive control. fMRI research shows that motivated proactive control is implemented through fronto-parietal control networks that are triggered by reward and difficulty cues. Here, we investigate electrophysiological signatures of proactive control. Previously, the contingent negative variation (CNV) in the ERPs and oscillatory power in the theta (4-8 Hz) and alpha band (8-14 Hz) have been suggested as signatures of control implementation. However, experimental designs did not always separate control implementation from motor preparation. Critically, we used a mental calculation task to investigate effects of proactive control implementation on the CNV and on theta and alpha power, in absence of motor preparation. In the period leading up to task onset, we found a more negative CNV, increased theta power, and decreased alpha power for hard versus easy calculations, showing increased proactive control implementation when a difficult task was expected. These three measures also correlated with behavioral performance, both across trials and across subjects. In addition to scalp EEG in healthy participants, we collected intracranial local field potential recordings in an epilepsy patient. We observed a slow-drift component that was more pronounced for hard trials in a hippocampal location, possibly reflecting task-specific preparation for hard mental calculations. The current study thus shows that difficulty information triggers proactive control in absence of motor preparation and elucidates its neurophysiological signatures.

Oscillatory signatures of reward prediction errors in declarative learning.

Reward prediction errors (RPEs) are crucial to learning. Whereas these mismatches between reward expectation and reward outcome are known to drive procedural learning, their role in declarative learning remains underexplored. Earlier work from our lab addressed this, and consistently found that signed reward prediction errors (SRPEs; "better-than-expected" signals) boost declarative learning. In the current EEG study, we sought to explore the neural signatures of SRPEs. Participants studied 60 Dutch-Swahili word pairs while RPE magnitudes were parametrically manipulated. Behaviorally, we replicated our previous findings that SRPEs drive declarative learning, with increased recognition for word pairs accompanied by large, positive RPEs. In the EEG data, at the start of reward feedback processing, we found an oscillatory (theta) signature consistent with unsigned reward prediction errors (URPEs; "different-than-expected" signals). Slightly later during reward feedback processing, we observed oscillatory (high-beta and high-alpha) signatures for SRPEs during reward feedback, similar to SRPE signatures during procedural learning. These findings illuminate the time course of neural oscillations in processing reward during declarative learning, providing important constraints for future theoretical work.

The sound of beauty: How complexity determines aesthetic preference.

Stimulus complexity is an important determinant of aesthetic preference. An influential idea is that increases in stimulus complexity lead to increased preference up to an optimal point after which preference decreases (inverted-U pattern). However, whereas some studies indeed observed this pattern, most studies instead showed an increased preference for more complexity. One complicating issue is that it remains unclear how to define complexity. To address this, we approached complexity and its relation to aesthetic preference from a predictive coding perspective. Here, low- and high-complexity stimuli would correspond to low and high levels of prediction errors, respectively. We expected participants to prefer stimuli which are neither too easy to predict (low prediction error), nor too difficult (high prediction error). To test this, we presented two sequences of tones on each trial that varied in predictability from highly regular (low prediction error) to completely random (high prediction error), and participants had to indicate which of the two sequences they preferred in a two-interval forced-choice task. The complexity of each tone sequence (amount of prediction error) was estimated using entropy. Results showed that participants tended to choose stimuli with intermediate complexity over those of high or low complexity. This confirms the century-old idea that stimulus complexity has an inverted-U relationship to aesthetic preference.

Signed reward prediction errors drive declarative learning.

Reward prediction errors (RPEs) are thought to drive learning. This has been established in procedural learning (e.g., classical and operant conditioning). However, empirical evidence on whether RPEs drive declarative learning-a quintessentially human form of learning-remains surprisingly absent. We therefore coupled RPEs to the acquisition of Dutch-Swahili word pairs in a declarative learning paradigm. Signed RPEs (SRPEs; "better-than-expected" signals) during declarative learning improved recognition in a follow-up test, with increasingly positive RPEs leading to better recognition. In addition, classic declarative memory mechanisms such as time-on-task failed to explain recognition performance. The beneficial effect of SRPEs on recognition was subsequently affirmed in a replication study with visual stimuli.

Occipital alpha power reveals fast attentional inhibition of incongruent distractors.

Recent associative models of cognitive control hypothesize that cognitive control can be learned (optimized) for task-specific settings via associations between perceptual, motor, and control representations, and, once learned, control can be implemented rapidly. Midfrontal brain areas signal the need for control, and control is subsequently implemented by biasing sensory representations, boosting or suppressing activity in brain areas processing task-relevant or task-irrelevant information. To assess the timescale of this process, we employed EEG. In order to pinpoint control implementation in specific sensory areas, we used a flanker task with incongruent flankers shown in only one hemifield (congruent flankers in the other hemifield) isolating their processing in the contralateral hemisphere. ERPs revealed fast modulations specifically in visual processing areas contralateral to the incongruent flankers. To test whether these modulations reflect increased or decreased processing of incongruent flankers, we investigated alpha power, a marker for attentional inhibition. Importantly, we show increased alpha power over visual areas processing incongruent flankers from 300 to 500 ms poststimulus onset. This suggests fast cognitive control by attentional inhibition for information disrupting goal-oriented actions. Additionally, we show that midfrontal theta earlier in the trial is also modulated by incongruency, and that theta power predicts subsequent alpha power modulations. This supports the hypothesis that midfrontal incongruency detection leads to control implementation, and reveals that these mechanisms take place on a fast, within-trial timescale.

The time course of cognitive control implementation.

Optimally recruiting cognitive control is a key factor in efficient task performance. In line with influential cognitive control theories, earlier work assumed that control is relatively slow. We challenge this notion and test whether control also can be implemented more rapidly by investigating the time course of cognitive control. In two experiments, a visual discrimination paradigm was applied. A reward cue was presented with variable intervals to target onset. The results showed that reward cues can rapidly improve performance. Importantly, the reward manipulation was orthogonal to the response, ensuring that the reward effect was due to fast cognitive control implementation rather than to automatic activation of rewarded S-R associations. We also empirically specify the temporal limits of cognitive control, because the reward cue had no effect when it was presented shortly after target onset, during task execution.