Top-down influences on the attentional orienting to and disengagement from threat

Abstract

The rapid detection and processing of potential threats enables a fast response to danger and provides a survival benefit (e.g. LeDoux, 1996; Tooby & Cosmides, 1990). Theories of threat processing suggest that we have a specialized threat module that enables the automatic (i.e. bottom-up driven) detection and evaluation of stimuli with a high threat value (Bentz & Schiller, 2015; LeDoux, 1996; 2015; Ohman, Flykt, & Esteves, 2001). Such theories suggest that threat processing is (largely) unaffected by top-down cognitive influences, but it remains debated whether attention to threat is contingent on top-down influences. Furthermore, it remains unexplored whether top-down processes can differentially affect the pre-attentive orienting of attention to threat and the subsequent disengagement of focal attention from such a threat. Strong versions of the threat module would predict that both the early (i.e. orienting) and later processing (i.e. disengagement) of threat should be unaffected by top-down influences, whereas more moderate versions would propose only the early stages of threat processing (i.e. orienting) to be unaffected by top- down processes. Theories of attention on the other hand do not explicitly model attention to be contingent on a stimulus’ threat value. Such theories propose that only a stimulus its features (e.g. green) can be processed pre-attentively and thus a stimulus’ identity and its associated threat value should not drive orienting (Treisman & Gelade, 1980, Wolfe, Cave, & Franzel, 1989; Wolfe, 1994, 2006). Moreover, such theories propose that both orienting and disengagement should be subject to top-down modulation, albeit perhaps to different degrees (see Carrasco, 2011). In this thesis I assessed whether top-down task goals modulate attention to threat. I assessed specifically how the relevance of a potential threat to the current task-goal affects the orienting and disengagement of attention. To disentangle orienting and disengagement, I developed a novel cueing paradigm that controlled the allocation of attention before the threat presentation. On disengagement trials, attention would be pre- cued to the threat location; on orienting trials attention would be pre-cued to one of the non-threat locations. The goal-relevance of the threat was manipulated by rendering the threat distractor its identity similar or dissimilar to one of the targets. In Chapter 2, I assessed the influence of top-down task goals on the disengagement of focal attention. Specifically, I investigated whether disengagement from a simplistic threat distractor (spider silhouette) is contingent on the threat distractor’s relevance to the immediate task-goal. Across two experiments I varied which stimuli were the designated targets (coloured green) that required a response. The results show that the threat distractor delayed disengagement when its identity was relevant to the current task-goal (i.e. identify whether the target is a spider or cat), but not when it was irrelevant to the task-goal (i.e. identify whether the target is a bird or fish). Thus, chapter 2 provides initial evidence that disengagement from threat is contingent on the current task-goals. In chapter 3, I assessed the influence of top-down task goals on the orienting of attention to threat. Specifically, I investigated whether orienting to photorealistic threats (spider pictures) is contingent on their relevance to the immediate task-goal. Across two experiments I observed that attention was rapidly oriented to the threat distractor when its identity was relevant to the current task-goal and much reduced when it was task- irrelevant. Thus, chapter 3 indicates that, not only the disengagement of focal attention, but also the early pre-attentive orienting of attention to threat is contingent on the current task-goals. In chapter 4, I assessed whether orienting to a task-goal irrelevant threat is contingent on stimulus-specific fear and the time-on-task (especially the number of threat presentations). Specifically, I investigated whether the orienting of attention to a task-goal irrelevant threat (photorealistic spider) is enhanced in individuals who were more fearful of the threat and decreases with increased time-on-task. The results showed that orienting to a task-goal irrelevant threat was observed only for relatively more fearful individuals and that such orienting rapidly decreased with increased time-on-task. Thus chapter 3 indicates that orienting to a goal-irrelevant threat was contingent on elevated fear and limited time-on-task. In summary, the current thesis suggests that top-down processes can modulate both the pre-attentive orienting to and the disengagement of focal attention from threat. Thus, the results of the current thesis are not in line with an encapsulated fear module. Instead, the influence of current task goals on the initial orienting of attention to threat implies that even the earliest stages of threat processing are contingent on an organism’s top-down goals. I would like to speculate that such orienting may be feature-driven as specific stimulus features may have become imbued with a threat value and is thus not necessarily at odds with existing theories of attention.