Beyond top-down: Feature search as a serial clump-wise process

There has been a long-standing debate over whether physically salient stimuli capture attention automatically, independent of the observer's goals, or whether attentional capture depends on the match between the stimulus and the observer's task set. Recent evidence indicates that attentional capture can be prevented when observers adopt a so-called feature search mode, which enables top-down suppression of salient but irrelevant stimuli. The present study examined whether the feature search mode reflects genuine top-down control or whether it simply results from serial, clump-wise search induced by specific display characteristics. To investigate this, we employed a two-target compatibility manipulation. Our findings reveal that in feature search mode, there were almost no compatibility effects, indicating clump-wise serial search which in turn resulted in no attentional capture by the salient distractor. In contrast, when participants engaged in singleton detection mode, there were large compatibility effects and at the same time strong attentional capture by the salient singleton distractor. The current results challenge the notion that top-down suppression prevents attentional capture. Instead, they support the interpretation that, in feature search, the absence of capture arises from the inherently constrained nature of serial search dictated by the display's structure. Consequently, our findings call into question the functional relevance of search modes as distinct cognitive states and advocate for a reinterpretation grounded in attentional window size and display configuration.