Top–down contingent attentional capture during feed-forward visual processing

Abstract

Theeuwes (2010) summarizes an impressive number of studiesdemonstrating interference by irrelevant visual singletons in com-puter experiments with humans. In these studies, if participantssearch for a relevant singleton target, such as the single diamondamong circles (i.e., a shape singleton), an irrelevant singletondistractor, such as the single red circle among the green stimuli (i.e.,a color singleton), delays the correct response to the target (cf.Theeuwes, 1991, 1992). If this distractor – as was the case in therelevant experiments – does not predict the likely target position, it issaid tobeirrelevant.Theeuwes(2010)arguesthattheinterference bythe irrelevant distractor reflects stimulus-driven or exogenouscapture of attention by singletons or feature contrasts, which heassumes to occur early during the feed-forward phase of visualprocessing.According to this explanation, a stimulus initially capturesattention exogenously (i.e., in a stimulus-driven way) to its position,to the degree to which it is salient. Salience in turn is defined as thesummed local contrast in terms of color, luminance, and orientationdifference to the surrounding stimuli or the background (cf. Bergen J Itti & Koch, 2001). Theeuwes (2010) assumes that thissalience-driven capture of attention is fast and occurs within 150 mssince singleton onset, during the feed-forward phase of visualprocessing (cf. Lamme, 2003).In contrast to Theeuwes (2010), we think that top–downcontingent capture is the rule and explains initial and fast attentioncapture effects in the first feed-forward phase of visual processing.Duringa laterphaseand undersomeconditions exogenouscaptureofattention possibly follows. At the same time, we propose that theevidence presentedbyTheeuwesfails to supportexogenousorientingbecause it fails to exclude a top–down contingent capture explana-tion. We present our arguments in two sections. First, we review theevidence directly supporting ourclaim that saliencecapture is subjectto top–down control. From this evidence we derive the exogenous-capture criterion that must be met by experiments for demonstratingexogenousspatialattention,andshowthatthiscriterioninnotmetbythe studies reviewed by Theeuwes. Second, we review studiesshowing attentional capture by stimuli during the feed-forwardphase, and conclude that only top–down contingent capture but noexogenous-capture can be observed during the feed-forward phase ofvisual processing.1. The exogenous-capture criterionA number of visual search studies demonstrated that participantsexert top–down control over attentional capture (reviewed recentlyby Burnham, 2007). In the corresponding studies, the experimentersecures the participant's top–down control over attention bydeclaring a specific visual feature as defining the target, or at leastas being informative for finding the target. The amount of attentionalcapture can then be compared under two conditions: with stimulusfeatures (or feature contrasts) that match the search templates of theparticipants (i.e., their top–down control settings for attentionalcapture) and with stimulus features (or feature contrasts) that do notmatch the top–down control settings. In these studies, researchersinferred top–down control over attentional capture from stronger oreven exclusive attentional capture by stimuli with matching featuresor feature contrasts (cf. Bacon & Egeth, 1994; Folk, Remington, J Eimer & Kiss, 2008; Folk & Remington,1998).The finding that attentional capture depends on the matchbetween the attentional control settings and features of theattention-capturingstimulushasimplicationsfor proofsof exogenousattention, too. That is, from these studies follows that demonstratingexogenous or stimulus-driven attentional capture requires that norelevant task set to search for the attention-capturing feature orfeature contrast of a distractor stimulus exists (cf. Bacon & Egeth,1994; Folk et al., 1992). This is the exogenous-capture criterion.Importantly, the exogenous-capture criterion is not met by thestudies that have been reviewed by Theeuwes (2010).Thisisbecause all of Theeuwes' reviewed studies used singletons – astimulus which is salient on a particular dimension – as the relevant