B.2.4. Overview of the experiments

One basic goal of the present experiments was to evidence, in one single experiment, that both endogenous attentional orienting and resistance to interference depended on executive WM. Lavie and de Fockert (2005) showed that a memorizing task had detrimental effects on a visual search when a salient distractor was present. However, the "irrelevant singleton" task (Theeuwes, 1992) they used might be problematic for the study of resistance to interference. Indeed, Bacon and Egeth (1994; see also Leber & Egeth, 2006) suggested that this paradigm could actually induce participants to engage in a "singleton search mode", that is, in a strategy consisting of orienting attention toward any singleton that appeared, independently of the target-defining features. Participants probably used this singleton search mode in the experiments of Lavie and de Fockert (2005). Since the salient distractor was a singleton, participants were likely prompted to orient attention toward this salient distractor. Thus, one cannot be completely confident that this study did genuinely or strictly address the issue of resistance to interference. At best, the tendency to avoid it and to resist the AC might have been rather weak.

A visual search task, derived from Michael et al. (2001b), was proposed in the present experiments. The target location was systematically cued validly, so that target selection relied strongly on endogenous processes. In addition, a salient onset distractor could appear in some trials. In these experiments, participants had no reason to rely on salience for orienting their attention, or to engage in a "singleton search mode". Indeed, first, the onset item was never the target and thus, was completely irrelevant, as stressed by the instructions. Second, the target could be selected easily without being salient. These manipulations allowed to evaluate more strictly the resistance to interference, as well as the efficiency of endogenous attentional orienting. To load WM, an "executive" auditory task was proposed, which taxed memory quite weakly. First, this discarded the possibility of modality-specific interferences, since the two tasks were presented in different modalities. The interference might only arise via interference in more general WM resources. Second, the use of an "executive", rather than "mnesic" task followed the reasoning of Han and Kim (2004), that attentional orienting critically depended on the executive component of WM. Finally, the perceptual load was manipulated by varying the number of items in the visual search. Increasing this number should not lead to a longer search, nor modify the task difficulty, as was usually the case in visual search tasks. Indeed, the target was systematically validly cued by an arrow, leaving the "search" stricto sensu, useless.

In the present experiments, we hypothesized, on the basis of a biased-competition account of executive functioning (Duncan, 2001; Duncan et al., 1996; 2008; Miller & Cohen, 2001; Sala & Courtney, 2007) that both endogenous orienting and resisting interference would be disrupted by loading WM with a concurrent executive task. Moreover, consistent with the load hypothesis of Lavie (2005), increasing the perceptual load was hypothesized to have the opposite effect, that is, to decrease the AC (see also Torralbo & Beck, 2008). Two experiments were proposed, only varying according to the single-task conditions (presence vs. absence of the auditory list).