A better theoretical explanation of task switching may need to contain all three sources of switch costs (an integrated model). With regards to the TSR model (Monsell, 2003),it suggested that RT switch costs reflect the time needed by performing task-set reconfiguration .It claims a task per se demands its own ‘task-set’ (Rogers & Monsell, 1995) or ‘recipe’ ( Allport, Styles, & Hseih, 1994), comprising the regulations of the task and stimulus-response (S-R) mapping (e.g. searching for targets in one position, suppressing interference in another position; if color red appears press right lever, if green press another) (Brand,2007). To change tasks, a new task-set need to be established before conducting appropriate task-specific processes (Monsell, 2003), so the theory proposes that switch costs is contributed by operating the TSR processes. The main empirical evidences of TSR have stems from studies investigating the influences of time available for preparing task. If preparation time is given, endogenous TSR (goal-driven or ‘top-down’, voluntary and intentional) can be proceeded before stimulus onset then the RT can be reduced (Meiran, Chorev & Sapir 2000). So the preparation effect indicated the existence of TSR in task switching. Some studies had found this preparation effects do exist (e.g., Rogers & Monsell, 1995; Monsell & Mizon, 2006). However it just work in a certain interval (0-600ms), when the preparing time longer than 600ms, the switch cost magnitude seems to very close to a substantial asymptote but not abolished even time last to 5ms or more (Monsell, 2003; Monsell & Mizon, 2006).
If the TSR processes are a whole story how can they account for this ‘residual cost’? Some researchers (Monsell, 2003; Rogers & Monsell, 1995) suggested that TSR not just includes endogenous control but also exogenous (stimulus-driven or 'bottom-up', automatic) control. Such endogenous TSR like attention switching, can be triggered by cue before stimulus. But the exogenous TSR like response setting and response settings adjustment (Monsell, 2003), cannot be accomplished till task-relevant attributes are caught from stimulus. Which means the residual cost actually reflects time consumed by exogenous TSR in response-stimulus interval (RSI); the reduction in switch cost (RISC effect) will appears by extending the cue-stimulus interval (CSI) (Monsell & Mizon, 2006).
However, the TSR model cannot explain the asymmetric switch cost that observed in studies with Stroop stimulus (Allport & Wylie, 2000). Such studies, for instance, asked subjects switching between colour naming task (not denoted by the words name) and colour words naming task (Allport, Styles& Hsieh, 1994; Allport & Wylie, 2000); a surprising result shows that the switch cost is larger when switching from colour naming to word naming task than switching from word naming to colour naming task (Allport, Styles& Hsieh, 1994; Allport & Wylie, 2000). It is widely considered that disparity between colour word and its display colour, the colour words naming interfere much less than the display colour naming (van Maanen, van Rijn & Borst, 2009). So word naming task is more proficient which possesses ‘stronger’ task-set (Cohen et al, 1990; as cited in Monsell, 2003), and the result suggests that switching to the stronger task increasing the switch cost. This asymmetric switch cost is also found in language Stroop study (Meuter & Allport, 1999), subjects were asked to name Number either using second language or mother language. The RT switch cost is increased when switching to a task of using second language to name numbers, compared with switching to a task of using mother language to name numbers (1999). The easier shift in less familiar task is showed. These findings raise a question to TSR model that why it requires extended time to reconfigure for the more practised task?
Although the theory provides explanation of switch cost in studies with congruent stimuli (Rogers & Monsell, 1995), it does not account for incongruent switch cost but the TSI theory do. It proposes that subjects must conduct additional inhibition to the practised task-set in order to implement the unskillful task (Allport & Wylie, 2000). This inhibition will carry over to the subsequent task (carryover effect) (2000), so when switching from the weaker to ‘stronger’ task, the RT is delayed for dissipating inhibition of ‘stronger’ task-set (Monsell & Mizon, 2006). It is supported by the study of backward inhibition (Mayr & Keele, 2000). It examined task inhibition by exacting subjects shifting between three tasks (ABA/CBA). When subjects in the ABA trials, temporal decay in performing the task A (second) was observed; whereas better performance of the task A was showed in the CBA trial (2000). It evidences that the poorer performance in second task A is because residual elements of first task A are inhibited enabling to perform task B, this inhibition leads a negative effect in retrieving the original task-set when requiring to switching to (practiced )task A again (2000), although the task-set had been established before; while in the CBA trial, the inhibition of task-set A did not occur, therefore, the time is just taken to reconfigure task-set of A which short than the time taken by overcoming inhibition and task-set retrieval. The effect of inhibition of the carryover effect is attested, suggesting the existence of TSI in switching.
However, in a task-switching study (Allport & Wylie, 1999), the RT costs were found in performing merely one task, but before the subjects have implemented another task with same stimuli. These findings cannot well interpret by both TSR and TSI processes. A task-set priming model (Allport & Wylie, 1999) was presented, it proposes a long-term priming effect which is triggered by the ambiguous stimulus (e.g., bivalent).The stimulus stimulate associative retrieval of task-set and the competition will occur among task-sets retrieval that engendering switch cost.
Koch and Allport(2006) attempted to work out which part (or interval) of switch costs are produced by the stimulus-based priming effect. The experiment (2006) manipulated the activation of the priming effect by using task with equally difficulty (judging the digital size or parity).The each stimuli is associated to only one task in the first four block (e.g., judging the parity of No.3 only), but reversing the mapping in the fifth block to test the influences of item-specific stimulus-task association. The performance then measured in the fourth and fifth block, showing a considerable drop in the task-set priming effects when cue-stimulus interval CSI is delayed. It indicated that stimulus-based priming largely emerging in performance when competition between tasks is fierce and that cue-based advanced preparation lessens task competition. Inversely, elongating the RCI (decay time) only decreased switch costs but not the stimulus-based priming effect (Koch & Allport, 2006).
Each theoretical process has been indicated as a contributor of switch costs above. So far, an integrated model seems more likely to offer a better explanation of behavioural effects of task switching.
The brain regions of implementing of task switching
Knowing the three theories provides three fundamental mental processes of task switching; the basic behavoural effects of task switching can be better explained. To answer: how we perform task switching, the neuroimaging studies of task switching provides sufficient evidences for knowing how they implement in our brain.
Evidences from some neuroimaging studies demonstrate that the TSR and TSI processes implement in the different brain regions. For example, a study with curing paradigm, and found the switching positivity wave in the parietal lobe which is influenced by the length of the CSI (150ms-600ms) but not affected by the length of the RSI (750ms-1200ms) (Nicholson, Karayanidis, Poboka, Heathcote & Michie, 2005). This means that the switching positivity in the parietal lobe is affected by task preparation of upcoming task but not impacted by prior task activation, so the brain mechanisms in the TSR and TSI processes are different. Moreover, further study (Aron, Monsell, Sahakian & Robbins, 2004) has uncovered the active regions of the TSR and TSI processes by comparing brain activation in brain injury patients and health subjects (detecting by fMRI). The study shows active brain region of inhibition of task-set is right inferior frontal gyrus (IFG), whereas the active region of endogenous TSR is left middle frontal gyrus (MFG). Two studies indicated that the both TSR and TSI processes are operated in different brain region, with no interacted effect.
Apart from this, the operating intervals of the TSR and TSP processes are found by using ERP. A study attempted to examine the effect of predictability task sequence and stimulus types (bivalent vs. univalent).The result revealed that more positivity wave appeared in switching trial (during CSI) when the task is predictable than it in repetitive trial; after stimulus onset, the more negativity wave was evoked in switching task with bivalent stimulus, while the negativity wave does not observed in univalent stimulus condition. The positivity wave in switching task reflects the processes for rules set up and activation during preparation, whereas the negativity wave reflects the inhibition of carryover effect or irrelevant interference. So the result illustrated that the TSR and TSP processes performed in the CSI and RSI respectively.
Limitation & conclusion
In conclude, this essay suggests that the three types of processes (TSR, TSI&TSP) are all contributing to the magnitude of switch costs. Each theory can partially explains the switch costs, but the integrated processes which includes all three types of processes seems to better account for the behavioural effects of task switching (e.g., asymmetric switch cost). The experimental evidences were provided to support that one process is not enough to clarify the mechanism of task switching.
Then the neuroimage studies were presented to indicate the active regions of operating TSR and TSI processes. However, the active region of the TSP is unknown. The neuroinage evidence also located the performing intervals of the TSR and TSP processes. The further studies in neuroimaging are required to discover more active regions of task switching processes, so the how physically behavour of brain when implementing switching will be clear. Furthermore, the studies in locating the task switching processes to corresponding interval are also demanded. Importantly, even combined three task-switching processes, the residual cost still remains (Koch,2001), other contributors of switch cost require exploring.
References
Allport, D., Styles, E. A., & Hsieh, S. (1994). Shifting intentional set: Exploring the dynamic control of tasks.In C. Umiltà, M. Moscovitch, C. Umiltà, M. Moscovitch (Eds.), Attention and performance 15: Conscious and nonconscious information processing (pp. 421-452). Cambridge, MA US: The MIT Press.
Allport, A., & Wylie, G. (2000) Task-switching, stimulus-response bindings and negative priming. In S. Monsell & J. Driver (Eds), Control of Cognitive Processes: Attention and Performance XVIII (pp. 35–70).Cambridge, MA: MIT Press
Allport, A. and Wylie, G. (1999) Task-switching: Positive and negative priming of task-set. In Attention, Space and Action: Studies in Cognitive Neuroscience
Aron, A. R., Monsell, S., Sahakian, B. J., & Robbins, T. W. (2004). A componential analysis of task-switching deficits associated with lesions of left and right frontal cortex. Brain: A Journal Of Neurology, 127(7), 1561-1573. doi:10.1093/brain/awh169
Brand, S. L . (2007).Task Switching and Distractibility. Doctoral thesis, University of London
Hallowell, E. M. (2007). Crazy Busy: Overstretched, Overbooked, and About to Snap! Strategies for Handling Your Fast-Paced Life. Ballantine Books. ISBN 034548244
Jersild, A. T. (1927). Mental set and shift. Archives of Psychology, 89,1–81
Koch, I. (2001). Automatic and intentional activation of task sets. Journal of Experimental Psychology: Learning, Memory, & Cognition, 27, 1474-1486.21
Koch, I., & Allport, A.(2006). Cue-based preparation and stimulus-based priming of tasks in task switching. Memory and Cognition,34 (2), 433-444
Monsell, S. (1996). Control of mental processes. In V. Bruce (Ed.), Unsolved mysteries of the mind.
Tutorial essays in cognition. Oxford England: Erlbaum (Uk) Taylor & Francis, Publ
Monsell, S., & Mizon, G. A. (2006). Can the Task-Cuing Paradigm Measure an Endogenous Task-Set Reconfiguration Process? Journal Of Experimental Psychology. Human Perception & Performance, 32(3), 493-516. doi:10.1037/0096-1523.32.3.493
Monsell, S. (2003). Task switching. Trends in Cognitive Sciences. 7, 134140.
Mayr, U., & Keele, S. W. (2000). Changing internal constraints on action: The role of backward inhibition. Journal of Experimental Psychology: General, 129 (1), 4-26
Meuter, R. I., & Allport, A. (1999). Bilingual language switching in naming: Asymmetrical costs of language selection. Journal Of Memory And Language, 40(1), 25-40. doi:10.1006/jmla.1998.2602
Meiran N, Chorev Z, Sapir A. (2000). Component processes in task switching. Cognitive Psychology. 41 (3), 211-253
Nicholson, R., Karayanidis, F., Poboka, D., Heathcote, A., & Michie, P. T. (2005). Electrophysiological correlates of anticipatory task-switching processes. Psychophysiology, 42(5), 540-554.
Rogers, R., & Monsell, S. (1995). The costs of a predictable switch between simple cognitive tasks. Journal of Experimental Psychology: General, 124, 207–231.
Rubin, O., & Meiran, N. (2005). On the Origins of the Task Mixing Cost in the Cuing Task-Switching Paradigm. Journal Of Experimental Psychology: Learning, Memory, And Cognition, 31(6), 1477-1491. doi:10.1037/0278-7393.31.6.1477
van Maanen, L., van Rijn, H., & Borst, J. P. (2009). Stroop and picture–word interference are two sides of the same coin. Psychonomic Bulletin & Review, 16(6), 987-999. doi:10.3758/PBR.16.6.987
Shiffrin, R.M., & Schneider, W. (1977). Control and automatic human information processing: II. Perceptual learning, automatic attending, and a general theory. Psychological Review, 84, 127–190.
Sohn, M-H, Ursu, S, Anderson, J. R, Stenger, V. A & Carter, C. S. (2000). The role of prefrontal cortex and posterior parietal cortex in task switching. Proc. Natl. Acad. Sci. U. S. A. 97, 13448–13453
Wylie, G., & Allport, A. (2000). Task switching and the measurement of “switch costs”. Psychological Research, 63, 212-233.