Executive control: an electrophysiological investigation of control processes

Everyday behaviour requires constant coordination and monitoring in order for our actions to be successful. Within cognitive science such coordination and monitoring of behaviour is termed ‘control’ and refers to a set of functions that serve to configure the mental system for performing specific acts. A system of cognitive control is thought to set high level goals and direct subordinate cognitive systems in order to accomplish those goals. This thesis utilises a cognitive electrophysiological approach to the study of executive control, addressing research questions concerning the mental processes that are modulated by executive control and the mechanisms underlying control-related processing adjustments. The first experimental chapter investigates the process of task switching. More specifically, how demanding is a proposed stage of endogenous task-set reconfiguration in terms of information processing? It was previously reported that the process of task-set reconfiguration constitutes a hard bottleneck delaying even the earliest processing stages (e.g. perceptual) (Oriet & Jolicoeur, 2003). Three experiments investigated this claim by manipulating stimulus contrast and RSI within an alternating runs task switching paradigm. Both RT results and measurements of P1 and N1 ERP component peak latency did not offer support to the claim that task-set reconfiguration delays perceptual processing. Experimental Chapters 3 and 4 used interference paradigms that are common within the study of executive control (e.g. Eriksen Flanker task and a Stroop task, respectively). Within such interference paradigms, separate stimulus dimensions (relevant and irrelevant) are manipulated, with RT being faster when both the relevant and irrelevant stimulus dimensions indicate the same response. This is termed the ‘congruency effect’ and is often attributed to a failure of selective attention, namely, an inability to ignore the irrelevant stimulus dimension. It has been demonstrated that such congruency effects are dependent upon task sequence with the effect being reduced (or absent) after an incongruent trial (Gratton et al., 1992). Such conflict adaptation effects are a popular measure of cognitive control processes. An influential model of cognitive control is the conflict monitoring model of Botvinick et al. (2001), with much evidence for this model being based on the conflict adaptation effect. Specifically, the model proposes that the ACC measures for the occurrence of response conflict within two response channels, and when detected, signals its occurrence to other brain regions (e.g. DLPFC) that are involved in implementing control. Such control may be implemented via a top-down biasing mechanisms of attention toward the task-relevant stimulus feature. Chapter 3 investigated the conflict adaptation effect within the Flanker task and examined, whether after the occurrence of conflict, attention is directed toward the task-relevant central target location. This was done by measuring P1 and N1 ERP component amplitudes. Although behavioural conflict adaptation effects were evident in overt behaviour, these were specific to response repetitions, consistent with a bottom-up priming account that excludes the necessity for a top-down control explanation (e.g. Mayr et al., 2003). In addition, P1 and N1 amplitude did not show any evidence of increased attentional focus toward the central target location after the occurrence of conflict. Chapter 4 investigated the conflict adaptation effect within a modified Stroop task, and again, examined whether after the detection of conflict, attention is directed toward the task relevant stimulus feature. This was done by measuring N170 amplitude - an ERP component proposed to index face processing - when a face stimulus served as the relevant and irrelevant stimulus dimension. Again, conflict adaptation effects were evident in overt behaviour, with this effect being driven by the occurrence of response conflict. Unlike the data from the Flanker task, the conflict adaptation effect within the Stroop task was specific to response alternations, and thus, a bottom-up priming account is not applicable in this instance. However, again the ERP results did not offer any evidence that the processing of the relevant stimulus dimension was enhanced after the occurrence of conflict. Implications of the present results are discussed in the context of executive control and in particular, in relation to models of task switching and models of conflict control