The task-switching paradigm offers enormous possibilities to study cognitive control as well as task interference. The current review provides an overview of recent research on both topics. First, we review different experimental approaches to task switching, such as comparing mixed-task blocks with singletask blocks, predictable task-switching and task-cuing paradigms, intermittent instructions, and voluntary task selection. In the 2nd part, we discuss findings on preparatory control mechanisms in task switching and theoretical accounts of task preparation. We consider preparation processes in two-stage models, consider preparation as an all-or-none process, address the question of whether preparation is switchspecific, reflect on preparation as interaction of cue encoding and memory retrieval, and discuss the impact of verbal mediation on preparation. In the 3rd part, we turn to interference phenomena in task switching. We consider proactive interference of tasks and inhibition of recently performed tasks indicated by asymmetrical switch costs and n-2 task-repetition costs. We discuss stimulus-based interference as a result of stimulus-based response activation and stimulus-based task activation, and response-based interference because of applying bivalent rather than univalent responses, response repetition effects, and carryover of response selection and execution. In the 4th and final part, we mention possible future research fields.Keywords: task switching, cognitive control, interference Human behavior is highly adaptive and flexible in response to changing environmental demands. This flexibility requires complex cognitive control processes, which allow humans to not only respond reactively but also to behave in a more proactive way to achieve goals and to perform tasks. The exploration of the proAndrea Kiesel, Department of Psychology, University of Wiirzburg, Wiirzburg, Germany; Marco Steinhauser, Department of Psychology, University of Konstanz, Konstanz, Germany; Mike Wendt, Experimental Psychology Unit, Helmut Schmidt University, Hamburg, Germany; Michael Falkenstein, Leibniz Research Centre for Working Environment and Human Factors, TU Dortmund, Dortmund, Germany; Kerstin Jost, Department of Psychology, University of Marburg, Marburg, Germany, and Department of Psychology, RWTH Aachen University, Aachen, Germany; Andrea M. Philipp and Iring Koch, Department of Psychology, RWTH Aachen University.This article was supported by the Deutsche Forschungsgemeinschaft (DFG) in the context of the priority program "Executive Functions" (Grants Fa 211115-1,2; Ho 1301/8-1,2,3; Hu 432/8-1,2,3; KI 488/6-1,2,3; Ko 2045/4-1,2,3; and Ro 529/17-1,2,3). We thank Stephen Monsell for very helpful comments on earlier versions of the article.Correspondence concerning this article should be addressed to Andrea Kiesel, Department of Psychology, University of Wiirzburg, Rontgenring 11, 97070 Wiirzburg, Germany, or to Iring Koch, Lehrstuhl fur Psycholo!,>ie I, RWTH Aachen University, Jiigerstrasse 17-19,52056 Aachen, Germany. ...

While it is well known that working memory functions decline with age, the functional reasons for this decline are not well understood. A factor that has proven critical for general individual differences in visual working memory capacity is the efficiency of filtering irrelevant information. Here, we examine to what degree this factor is also responsible for age differences in working memory. Young and old participants performed a change-detection task where some items in the encoding display were marked as irrelevant. The contralateral delay activity of the electroencephalogram was used to assess individual participants' filtering efficiency (see Vogel EK, McCollough AW, Machizawa MG. 2005. Neural measures reveal individual differences in controlling access to working memory. Nature. 438:500-503.). Older adults showed smaller filtering scores than young adults, but only early in the retention interval, suggesting that efficient filtering was delayed. In contrast, age-independent individual differences in filtering were reflected primarily later in the retention interval. Thus, age and individual differences in filtering are reflected in different ways showing that old adults are not simply like less efficiently performing young adults.

To date, much is known about the neural mechanisms underlying working-memory (WM) maintenance and long-term-memory (LTM) encoding. However, these topics have typically been examined in isolation, and little is known about how these processes might interact. Here, we investigated whether EEG oscillations arising specifically during the delay of a delayed matching-to-sample task reflect successful LTM encoding. Given previous findings of increased alpha and theta power with increasing WM load, together with the assumption that successful memory encoding involves processes that are similar to those that are invoked by increasing WM load, alpha and theta power should be higher for subsequently remembered stimuli. Consistent with this assumption, we found stronger alpha power for subsequently remembered stimuli over occipital-to-parietal scalp sites. Furthermore, stronger theta power was found for subsequently remembered stimuli over parietal-to-central electrodes. These results support the idea that alpha and theta oscillations modulate successful LTM encoding.

We investigated if incongruent solutions of simple multiplication problems would elicit similar event-related brain potentials as inappropriate words in sentences. In Experiment I, 12 subjects verified the appropriateness of solutions of multiplication problems or of final words in short sentences. Both incongruent solutions and incongruent words evoked a phasic negative shift between 300 and 500 ms having a similar topography. In Experiment II, we tested with another sample of 13 subjects if the amplitude of this arithmetic N400 effect was affected differently by different stimulus onset asynchronies (SOA = 200 and 500 ms) and by errors that were either table-related or table-unrelated to the preceding operands. Again, incorrect solutions elicited an arithmetic N400 effect whose amplitude depended on both the relatedness of the solution and the SOA. The ascending part of the N400 effect was always larger for unrelated than for related errors independently of the SOA, whereas the maximum of the N400 effect was larger for unrelated errors in case of a long SOA only. This pattern of effects was similar to that observed with semantic material varying lexical associations. These results suggest that arithmetic incongruencies are handled by the system in a manner functionally similar to that of semantic incongruencies.

74The task-switching paradigm has become an important tool for study of the cognitive control processes needed to flexibly adjust to a changing environment (for a review, see Monsell, 2003). In a typical task-switching experiment, participants have to switch between two or more tasks, and differences in response times (RTs) and accuracy between task alternations and task repetitions-that is, the so-called switch costs-are typically interpreted as a reflection of the processing demands involved in changing task-specific cognitive configurations. However, recent evidence has posed a challenge to this interpretation of task-switch costs, as assessed in one variant of the taskswitching paradigm-that is, the task-cuing paradigm.In the task-cuing paradigm, tasks vary randomly from trial to trial, and a cue presented in advance indicates the relevant task. In contrast to alternative task-switching variants (e.g., that of Rogers & Monsell, 1995), the task-cuing p paradigm allows precise temporal control over both cueassociated preparatory processes and stimulus-associated p processes involved in resolving stimulus-triggered comp petition. However, as noted independently by Logan and Bundesen (2003) and , in the standard task-cuing paradigm, a task switch is always confounded with a switch of the task-indicating cue: Whenever the task switches, the cue switches as well; whenever the task stays the same, the cue also stays the same. To handle this problem, both groups introduced a new variant r in which two cues were mapped to each task, allowing for examinations of task switching independently from cue switching. From such 2:1 mappings, three different trand sition types result: regular no-switch transitions (cue and task repetition); regular task-switch transitions (cue and task switch); and a new type of transition, in which cue changes are associated with task repetitions (cue-switch condition). Using this method, both studies reported substantial costs for cue switching at short cue-stimulus intervals (CSIs), even when the task remained constant. The studies, however, differed in their findings concerning the actual task-switch costs. Whereas the data in the Logan and Bundesen (2003) study suggested that task-switch costs r are almost completely attributable to a cue switch, Mayr r and Kliegl (2003) observed actual task-switch costs over t and above the costs from a cue switch. These discrepant findings resulted in two different theoretical approaches to account for task-switch costs in the cuing paradigm. account starts with the simple assumption that working memory can contain only one coherent task set at a time, so that each task-switching trial requires updating working memory and the attentional configuration to reflect the current task demands. This updating process comes in two stages (for similar twoprocess approaches, see Meiran, 1996 Recent findings suggesting that switch costs in the task-cuing paradigm are largely attributable to a change in the task-indicating cue have been interpreted in terms of a pr...