It is not what you expect: dissociating conflict adaptation from expectancies in a Stroop task.

In conflict tasks, congruency effects are modulated by the sequence of preceding trials. This modulation effect has been interpreted as an influence of a proactive mechanism of adaptation to conflict (Botvinick, Nystrom, Fissell, Carter, & Cohen, 1999), but the possible contribution of explicit expectancies to this adaptation effect remains unclear. The present study shows that it is possible to dissociate explicit expectancies from sequential adaptation effects in a Stroop task, in conditions in which feature repetitions are avoided, and in which the response-to-stimulus interval is set to 0 ms. We found a progressive adaptation effect that depends on the congruency of the previous series of trials, rather than exclusively on the preceding trial. This effect is independent from explicit expectancies (Experiment 1), and can even contradict these expectancies when participants are presented with informative patterns favoring either repeating or alternating congruency (Experiments 2a and 2b). The existence of a progressive adaptation effect independent from explicit expectancies and from repetition priming challenges the idea that conflict adaptation acts always on a top-down basis (Notebaert, Gevers, Verbruggen, & Liefooghe, 2006), and it rather indicates the existence of automatic sources of sequential adaptation, including the adaptation to the lack of conflict. Implications of these results on current understanding of some empirical phenomena of cognitive control, such as that of proportion of congruency, will be highlighted.