Evidence for Task-Specific Resolution of Response Conflict
Kiesel, Andrea, Kunde, Wilfried, Hoffmann, Joachim, Psychonomic Bulletin & Review
When a target requires different responses to a relevant and to an irrelevant task in a task-switching paradigm, there is response conflict. This target-induced response conflict was combined with conflict caused by a subliminally presented prime presented prior to the target. We found that target-related conflict reduced prime-induced conflict effects within the same trial. However, target-related conflict modified prime-related conflict effects according to the irrelevant stimulus-response (S-R) rule, but not according to the relevant S-R rule. Moreover, trial-to-trial modulations of the target congruency effect were observed in task repetition trials, but not in task switch trials. These results indicate that conflict resolution mechanisms, at least under the present circumstances, operate in a strictly task-specific manner.
Humans are confronted with complex situations in which different stimuli require different responses. Just imagine a typical situation in which you are preparing dinner: Boiling water tells you that it is time to put the rice in the pot. At the same time, a sound of the timer might urge you to remove the turkey from the oven. And while you are removing the turkey, you hear the telephone ring and a knock on the door. The best strategy in such a situation would be to just concentrate on the most important task and to ignore other, distracting stimuli.
However, numerous experimental studies have shown that stimulus processing is not restricted to task-relevant information but that task-irrelevant information influences performance as well. In so-called conflict tasks, it has been shown that location (Simon, 1969), word meaning (Stroop, 1935), or distractor (Eriksen & Eriksen, 1974) information is processed despite being task irrelevant. In task-switching situations, performing the currently relevant task is hindered when stimuli are presented that belong to another, currently irrelevant task (Allport, Styles, & Hsieh, 1994; Meiran, 1996; Rogers & Monsell, 1995). And finally, in priming studies, it has been shown that task-irrelevant information influences performance even when it is presented not simultaneously with, but in close succession to, the target and when the irrelevant prime stimuli are presented subliminally-that is, too quickly to be recognized consciously (see Dehaene et al., 1998; Neumann & Klotz, 1994; Vorberg, Mattler, Heinecke, Schmidt, & Schwarzbach, 2003).
Conflict and Conflict Resolution
In all these settings, conflict is determined by comparing incongruent and congruent stimulus conditions: When the irrelevant stimulus requires a response different from that required by the relevant stimulus, response times (RTs; and quite often, also error rates) are increased. However, processing of irrelevant information is not as automatic as is suggested at first glance. Several studies have shown that congruency effects (measured as the difference between congruent and incongruent trials) depend on congruency sequence. After an incongruent trial, the congruency effect in the current trial is reduced and, sometimes, even eliminated (see Gratton, Coles, & Donchin, 1992; Kunde, 2003; Kunde & Wühr, 2006; Stürmer, Leuthold, Soetens, Schroter, & Sommer, 2002).
These sequential modulations of congruency effects have been taken as evidence for a conflict adjustment mechanism (such as that proposed by Botvinick, Braver, Barch, Carter, & Cohen, 2001; Botvinick, Cohen, & Carter, 2004). Presumably, the anterior cingulate cortex (ACC) registers the occurrence of conflict and triggers "compensatory adjustments in cognitive control" (Botvinick et al., 2004, p. 539). Thus, top-down processes are assumed to compensate for conflict. There are several conflict resolution models that differ regarding the nature of these top-down processes; for example, it has been proposed that conflict induces a more cautious mode of information processing by stressing accuracy over speed (Gratton et al. …