Wim Notebaert

Post-error slowing: An orienting account

It is generally assumed that slowing after errors is a cognitive control effect reflecting more careful response strategies after errors. However, clinical data are not compatible with this explanation. We therefore consider two alternative explanations, one referring to the possibility of a persisting underlying problem and one on the basis of the low frequency of errors (orienting account). This latter hypothesis argues that infrequent events orient attention away from the task. Support for the orienting account was obtained in two experiments. Using a new experimental procedure, Experiment 1 demonstrated post-error slowing after infrequent errors and post-correct slowing after infrequent correct trials. In Experiment 2, slowing was observed following infrequent irrelevant tones replacing the feedback signals.

Adaptation by binding: a learning account of cognitive control

Cognitive control refers to the ability to repress our instantaneous urges in favor of more appropriate responses. Current debate concerns whether cognitive control effects that are studied in the laboratory (e.g. Stroop tasks) actually reflect the operation of a cognitive control system (adaptation theory) or instead merely reflect side effects of feature binding processes (binding theory). The two perspectives can be integrated by conceptualizing cognitive control as resulting from interactions between binding processes (as instantiated in Hebbian learning) and arousal. Conflict situations such as Stroop incongruent-stimuli lead to arousal and noradrenalin release throughout the brain, which facilitates binding between task-relevant cortical areas. Our proposal emphasizes an intimate link between cognitive and emotional processing.

Hebbian learning of cognitive control: Dealing with specific and nonspecific adaptation.

The conflict monitoring model of M. M. Botvinick, T. S. Braver, D. M. Barch, C. S. Carter, and J. D. Cohen (2001) triggered several research programs investigating various aspects of cognitive control. One problematic aspect of the Botvinick et al. model is that there is no clear account of how the cognitive system knows where to intervene when conflict is detected. As a result, recent findings of task-specific and context-specific (e.g., item-specific) adaptation are difficult to interpret. The difficulty with item-specific adaptation was recently pointed out by C. Blais, S. Robidoux, E. F. Risko, and D. Besner (2007), who proposed an alternative model that could account for this. However, the same problem of where the cognitive system should intervene resurfaces in a different shape in this model, and it has difficulty in accounting for the Gratton effect, a hallmark item-nonspecific effect. The authors of the current article show how these problems can be solved when cognitive control is implemented as a conflict-modulated Hebbian learning rule.

Top-down and bottom-up sequential modulations of congruency effects.

Several studies have demonstrated reduced congruency effects after incongruent trials. The conflict monitoring hypothesis (Botvinick, Braver, Barch, Carter, & Cohen, 2001) assumes that this sequential modulation is based on top-down cognitive control and suggests that more control is engaged after the detection of conflict. An alternative account is based on repetition effects of stimulus and response features and can be considered bottom up. This study investigates both modulatory sources. In a Stroop task with two response-stimulus intervals (RSIs), we demonstrate that top-down modulation does not occur with a very short RSI, suggesting that it takes some time before the system can be reconfigured. Bottom-up modulation is observed for both RSIs. This finding demonstrates that two different sources simultaneously reduce congruency effects after incongruent trials.

Cognitive control acts locally

Cognitive control adjusts information processing to momentary needs and task requirements. We investigated conflict adaptation when participants are performing two tasks, a Simon task and a SNARC task. The results indicated that one congruency effect (e.g., Simon) was reduced after conflict in the other task (e.g., SNARC), but only when both tasks used identical relevant information. On the other hand, when both tasks used different relevant information, a larger congruency effect was observed after conflict in the other task. The results are explained in terms of a local control mechanism.

Outcome Expectancy and Not Accuracy Determines Posterror Slowing: ERP Support

A considerable number of studies have recently used event-related potentials (ERPs) to investigate the mechanisms underlying error processing. Nevertheless, how these mechanisms are associated with behavioral adjustments following errors remains unclear. In the present study, we investigated how posterror slowing is linked to outcome expectations and error feedback. We used an adaptive four-choice reaction time task to manipulate outcome expectancy. Behaviorally, the results show posterror slowing when errors are unexpected and postcorrect slowing when correct responses are unexpected, indicating that outcome expectancy is crucial for posterror slowing. ERP analyses revealed that the error-related negativity and the feedback-related negativity were not correlated with the behavioral reaction time pattern, whereas the P3 was. The results support the hypothesis that posterror slowing is caused by attentional orienting to unexpected events.

Stimulus- and response-conflict-induced cognitive control in the flanker task

Recently, several studies have been conducted to investigate the top-down adjustments made after incongruent trials during conflict tasks. In the present study, we investigated conflict monitoring with different types of conflict. In a modified version of the flanker task, a distinction was made between stimulus—stimulus conflict and stimulus—response conflict. Six colors were mapped to three responses in order to exclude all sequences in which a relevant or an irrelevant stimulus- or response-related feature was repeated from trialn−1 to trialn. An analysis of the effect of the congruency of the previous trial demonstrated that conflict adaptation was present. The stimulus congruency effect was reduced after both a stimulus-incongruent trial and a response-incongruent trial. The response congruency effect did not vary as a function of previous congruency. These findings are discussed in relation to the distinction between conflict detection and conflict regulation.

Reward Modulates Adaptations to Conflict.

Both cognitive conflict (e.g. Verguts & Notebaert, 2009) and reward signals (e.g. Waszak & Pholulamdeth, 2009) have been proposed to enhance task-relevant associations. Bringing these two notions together, we predicted that reward modulates conflict-based sequential adaptations in cognitive control. This was tested combining either a single flanker task (Experiment 1) or a task-switch paradigm (Experiment 2) with performance-related rewards. Both experiments confirmed that adaptations after conflict were modulated by reward. In the flanker task, this resulted in increased conflict adaptation after rewarded trials. In the task-switching experiment, reward increased the conflict-modulated switch cost. Interestingly, both adaptations to conflict disappeared after no-reward trials. Moreover, individual differences in participants sensitivity to reward predicted these reward modulations of trial-to-trial adaptations. These findings shed new light on the exact role of cognitive conflict in shaping subsequent behavior.

Dissociating conflict adaptation from feature integration: a multiple regression approach.

Congruency effects are typically smaller after incongruent than after congruent trials. One explanation is in terms of higher levels of cognitive control after detection of conflict (conflict adaptation; e.g., M. M. Botvinick, T. S. Braver, D. M. Barch, C. S. Carter, & J. D. Cohen, 2001). An alternative explanation for these results is based on feature repetition and/or integration effects (e.g., B. Hommel, R. W. Proctor, & K.-P. Vu, 2004; U. Mayr, E. Awh, & P. Laurey, 2003). Previous attempts to dissociate feature integration from conflict adaptation focused on a particular subset of the data in which feature transitions were held constant (J. G. Kerns et al., 2004) or in which congruency transitions were held constant (C. Akcay & E. Hazeltine, in press), but this has a number of disadvantages. In this article, the authors present a multiple regression solution for this problem and discuss its possibilities and pitfalls.

Shared spatial representations for numbers and space: the reversal of the SNARC and the Simon effects.

In 4 experiments, the authors investigated the reversal of spatial congruency effects when participants concurrently practiced incompatible mapping rules (J. G. Marble & R. W. Proctor, 2000). The authors observed an effect of an explicit spatially incompatible mapping rule on the way numerical information was associated with spatial responses. The authors also observed an effect of an incompatible numerical mapping rule (if smaller than 5, press right; if larger than 5, press left) on the Simon effect. This effect was observed only when both tasks used the same effectors. The results point to a shared spatial representation for explicit spatial information (locations) and implicit spatial information (numbers).