Abstract: The generality and specificity of conflict adaptation across various domains can enhance our understanding of how the brain processes information and adapts to the environment. However, it remains unclear whether conflict adaptation is domain-general or domain-specific. Limitations in task design and insufficient control for confounding factors have contributed to this unresolved issue. In factorial designs, conflicts that compound within a single trial typically interact rather than operate independently, making it difficult to draw conclusions about domain specificity or generality. In contrast, task-switch designs can largely prevent such interactions by presenting different conflicts in separate trials. However, results obtained using task-switch designs have been inconsistent. Furthermore, feature integration, contingency learning, or temporal learning may have confounded the outcomes of previous factorial and task-switch designs. The present study aims to determine whether conflict adaptation is domain-general or domain-specific by employing a LOOK-DO transition, a task-switch design, and a confounder-minimized procedure in three experiments.
To control for feature integration and contingency learning, the stimuli were divided into two groups. Stimuli within each group were paired randomly, with no combinations made across groups. Stimuli from different groups were presented alternately. The LOOK-DO design required participants to observe in the previous trial and respond in the current trial, thereby helping to eliminate the influence of temporal learning on conflict adaptation. In Experiment 1, participants were asked to identify the central letter among five horizontally aligned letters in a letter version of the Flanker task. Prior to the presentation of these letters, a cue (“*” or “+”) was displayed for 500ms, followed by a jittered blank screen for 300~500ms. If the cue was “*”, participants needed to observe whether a conflict was present. Conversely, if the cue was “+”, participants were instructed to identify the central letter and provide a keypress response. An inter-trial interval of 800~1200ms preceded the next trial. In Experiment 2, participants completed both the Flanker task and a color-word Stroop task in separate trials within one of eleven blocks. The other settings in Experiment 2 were consistent with those in Experiment 1. Experiment 3 involved the same tasks as Experiment 2 but included the collection of EEG data from participants.
The main results indicate that conflict observation induced conflict adaptation in the Flanker task. This finding suggests that merely observing a conflict can trigger subsequent control adjustments, resulting in conflict adaptation. Furthermore, the LOOK Stroop conflict prompted the conflict adaptation in the DO Flanker task, and vice versa. This demonstrates that conflict observation can trigger cross-task conflict adaptation. The replication of the behavioral findings from Experiment 2, coupled with the significant conflict adaptation observed in the N450/SP during the Flanker-Stroop sequence and in the P3 during the Stroop-Flanker sequence, strengthens this conclusion. Overall, the above results indicate that conflict signals can initiate a general control adjustment process that leads to cross-task conflict adaptation at both the behavioral and neural levels. Furthermore, the significant conflict effects observed in the N2 and P3 components during the LOOK Flanker trials, as well as in the N450 and SP components during the LOOK Stroop task, provide compelling evidence that mere observation can generate conflict signals and trigger control adjustment processes. This finding aligns with the original definition of conflict in conflict monitoring theory, which defined conflict as the coactivation of incompatible representations. The generality of conflict adaptation may stem from the similar selective attention mechanisms present in both the Stroop and Flanker tasks. Additionally, observing conflict may enhance response readiness and improve the efficiency of the selective attention processes required to resolve the current conflict, ultimately facilitating the generality of conflict adaptation. Future research should operationally define similarity in different ways and utilize observation-response designs to further explore the relationship between similarity and conflict adaptation.

Key words: conflict adaptation effect, domain-generality, LOOK-DO transition, task-switch, conflict monitoring theory

摘要： 受限于任务设计的局限且未充分排除混淆因素的干扰,已有研究尚未说明冲突适应是领域一般的还是特定的。运用观察-反应设计和任务转换设计,同时严格排除特征整合和偶发学习等混淆因素,研究通过三个实验考察了这一问题。实验1发现冲突观察在Flanker任务中诱发了冲突适应。实验2发现冲突观察在Stroop-Flanker和Flanker-Stroop序列中均诱发了跨任务的冲突适应。实验3在重复实验2结果的基础上进一步发现,观察Flanker冲突在随后的反应Stroop任务中诱发了冲突适应,表现在脑电N450和SP成分上;同样地,观察Stroop冲突在随后的反应Flanker任务中也诱发了冲突适应,表现在脑电P3成分上。上述结果汇聚地表明冲突适应具有领域一般性特征。

关键词: 冲突适应, 领域一般性, 观察-反应设计, 任务转换设计, 冲突监测理论