Reading focus
Syllabus: How Do We Direct Attention and Control Thought?
Follow the move from a theory of monitoring to trial-by-trial behavioral and neural predictions. Relate it to reactive control while keeping prediction separate from causal intervention.
A question to consider. Which results support a conflict-to-adjustment account, and what alternative involving a common cause would remain even if ACC activity predicted later prefrontal activity?
Why this reading
An account of control needs to explain how an adjustment is recruited, not merely where difficult trials produce activity. Kerns and colleagues test whether activity during conflict predicts subsequent behavioral and neural changes.
Before you read
In the Stroop task, a word's meaning can compete with naming its ink color. Congruent trials align these responses; incongruent trials oppose them. The anterior cingulate cortex, or ACC, is a proposed conflict-monitoring region. The fMRI blood-oxygen signal is an indirect, delayed measure of neural activity.
Read for
- Reconstruct the previous-by-current trial comparison and the exclusion of repeated colors or words in the 23-participant study.
- Connect Figures 1 and 2 to distinct predictions about conflict, subsequent behavior, and prefrontal activity.
- Separate predictive associations from the proposed causal chain, attending to the authors' common-cause and mechanism qualifications.
Consider afterwards
- ACC activity predicts a later performance adjustment even after some repetition effects are excluded. What manipulation or comparison would strengthen the claim that this activity helps cause the adjustment?
Editorial context
The findings support predictions of conflict monitoring without identifying a causal signal from ACC to prefrontal cortex. The authors acknowledge that another region could influence both. Later work summarized by Duthoo and colleagues shows why repetition, learned response probabilities, and expectations must also be considered when interpreting trial-sequence effects. These alternatives call for discriminating designs, not an automatic dismissal of control.