任务切换机制研究的理论争议与整合

任务切换是执行控制的一种重要功能。切换代价是量化执行控制持续时间的指标。并广泛应用于各个研究领域。但是切换代价在多大程度上体现了执行控制的作用,仍然还存在争论．当前主要存在任务设置重构、任务设置惯性和联结竞争三种理论。本文首先介绍了三种理论的基本观点和实验证据,以及对任务切换加工机制的分析;其次阐明了对三种理论进行整合的相关研究;最后分析了进一步研究的方向。     

预先信息的不确定性与准备时间对任务切换的影响

为揭示任务切换中切换代价的本质, 考察了预先信息与准备时间对任务切换的影响, 结果发现: (1) 预先信息对任务切换有显著的影响效应, 全部信息的反应时和切换代价显著低于无信息和部分信息的反应时和切换代价, 而后二者之间无显著差异; (2) 预先信息的影响效应受到准备时间的调节作用, 在准备时间为100ms时, 预先信息无显著的影响效应, 而在准备时间达到600ms以后, 预先信息的准备效应显著。这说明, 预先信息影响的认知加工是任务设置重构, 并且这种影响效应受到准备时间的调节作用。     

刺激类型与预先信息对任务切换的影响

为揭示联结竞争与任务设置重构在切换加工过程中的关系,以切换代价为指标,研究了刺激类型与预先信息对任务切换的影响效应,结果发现:(1)无信息的切换代价显著高于部分信息和全部信息的切换代价,而后二者之间无显著差异;(2)单向刺激的切换代价显著高于双向刺激的切换代价;(3)预先信息和刺激类型对切换代价的影响无交互作用。这表明任务设置重构与联结竞争是两个具有相加效应的序列认知活动。     

On the difficulty of task switching: Assessing the role of task-set inhibition

This study assessed whether or not the difficulty of task switching stems from previous inhibition of the task set. A predictable sequence of univalent stimuli (affording performance of one active task) and bivalent stimuli (affording performance of two tasks) was used in two experiments. Experiment 1 used an alternating-runs paradigm (AABB) and Experiment 2 used a strictly alternating sequence (ABAB). The critical variable was whether the incentive for task-set inhibition was strong (on bivalent trials) or weak (on univalent trials). The question was whether it would be more difficult to switch to a task that previously needed to be inhibited than to a task that did not need to be inhibited. This pattern was not observed in either experiment. Thus, the data provide no evidence that task switching is difficult because of the need to overcome recent task-set inhibition.     

任务切换理论争议新证据

任务切换是研究认知控制的主要范式之一。大量研究发现切换试次比重复试次的反应时更长,错误率更高,这种差异称为切换代价。任务切换时所产生切换代价的理论解释主要有惯性论、重构论和联结论。近十年来,这些理论均获得新的实验支持和发展,但其争议依旧,没有哪一理论能成功解释任务切换的所有效应。未来研究可以建立整合模型,以准确描述切换代价产生的认知机制。     

Control processes in voluntary and explicitly cued task switching

Explicitly cued task switching slows performance relative to performing the same task on consecutive trials. This effect appears to be due partly to more efficient encoding of the task cue when the same cue is used on consecutive trials and partly to an additional task-switching process. These components were examined by comparing explicitly cued and voluntary task switching groups, with external cues presented to both groups. Cue-switch effects varied in predictable ways to dissociate explicitly cued and voluntary task switching, whereas task-switch effects had similar characteristics for both instructional groups. The data were well fitted by a mathematical model of task switching that included a cue-encoding mechanism (whereby cue repetition improves performance) and an additional process that was invoked on task-switch trials. Analyses of response-time distributions suggest that this additional process involves task-set reconfiguration that may or may not be engaged before the target stimulus is presented.     

Adjustments of task-set control processes: Effect of task switch frequency on task-mixing and task-switching costs

The present study tested the hypothesis that task-switch frequency triggers adjustments of task-set control processes. A mixed-task condition where task switches are frequent should promote flexibility—thus improving task-switch performance—whereas a condition where task repetitions are more expected should favour stability—thus improving task-repeat performance. In two experiments, participants performed single-task and mixed-task blocks. In mixed-task blocks, tasks varied randomly on a trial-by-trial basis. For half of the mixed-task blocks, the frequency with which the task changed was 25%, for the other half, it was 50%. Overall, depending on the task-switch frequency, performance on both task-repeat and task-switch trials was modified. Switch cost was reduced and task-repeat performance was altered by the increase in switch probability. This study demonstrates context-sensitive adjustments of task-set control processes. These results further support the view that mixing cost reflects sustained and endogenous components of cognitive control.     

Asymmetric switch cost and backward inhibition: Carryover activation and inhibition in switching between tasks of unequal difficulty.

Asymmetric switch cost, observed when switching between tasks varying in difficulty, shows that the difference between repeat and switch trials is greater when switching to the easier task. Early explanations of this effect attributed this pattern to both positive priming of the difficult task and negative priming of the easier task, but more recent models have focused only on activation processes. The role of inhibition in asymmetric switch cost was examined using backward inhibition, a more direct measure of task-set inhibition. The results indicated asymmetric backward inhibition, with greater sequential inhibition of the easier task (i.e., easy-difficult-easy sequences). Switch costs, however, showed both typical and reversed asymmetry (greater cost when switching from the easy to the difficult task), depending on the relative difficulty of task pairs. This pattern of results indicates that switch costs are attributable to both activation and inhibition processes.     

延迟时间与预先信息对任务切换的作用

对切换代价的本质存在任务设置惯性和任务设置重构两种观点。为揭示二者在切换代价中的关系, 研究了不同延迟时间条件下预先信息对任务切换的作用, 结果发现：(1)延迟时间对切换代价有显著影响, 100 ms的切换代价显著高于1000 ms的切换代价, 而600 ms分别与100 ms、1000 ms之间的切换代价无显著差异; (2)预先信息对切换代价的影响显著, 全部信息的切换代价显著低于无信息和部分信息的切换代价, 而后二者之间无显著差异; (3)延迟时间与预先信息无交互作用。这表明, 延迟时间与预先信息分别作用于任务设置惯性和任务设置重构, 切换代价反映了这两种具有相加效应的加工过程。     

自主任务转换中的重构和干扰

任务转换是研究执行功能的常用范式。任务转换通常伴随着转换代价：执行转换任务比重复任务的反应时更长、错误率更高。转换代价可能反映了任务设置重构(重构理论), 也可能表明任务之间存在干扰(干扰理论)。与任务线索范式相比, 自主任务转换范式更具生态效度, 而且不仅能获得转换代价这个传统结果, 还引入了任务选择比例、任务转换率等新指标, 其结果倾向于支持重构理论。此外, 新近研究指出自主任务转换可能也包含干扰的作用。未来, 应通过进一步改进实验范式等方法, 实现两大理论的融合。     

Reconfiguration of task-set: Is it easier to switch to the weaker task?

Switching between two tasks afforded by the same stimuli results in slower reactions and more errors on the first stimulus after the task changes. This “switch cost” is reduced, but not usually eliminated, by the opportunity to prepare for a task switch. While there is agreement that this preparation effect indexes a control process performed before the stimulus, the “residual” cost has been attributed to several sources: to a control process essential for task-set reconfiguration that can be carried out only after the stimulus onset, to probabilistic failure to engage in preparation prior to the stimulus, and to two kinds of priming from previous trials: positive priming of the now-irrelevant task set and inhibition of the now-relevant task-set. The main evidence for the carry-over of inhibition is the observation that it is easier to switch from the stronger to the weaker of a pair of tasks afforded by the stimulus than vice versa. We survey available data on interactions between task switching and three manipulations of relative task strength: pre-experimental experience, stimulus-response compatibility, and intra-experimental practice. We conclude that it is far from universally true that it is easier to switch to the weaker task. Either inhibition of the stronger task-set is a strategy used only in the special case of extreme inequality in strength, or its consequences for later performance may be masked by slower post-stimulus control operations for more complex tasks. Inhibitory priming may also be stimulus specific. Received: 31 March 1999 / Accepted: 23 July 1999     

The effects of recent practice on task switching

Four experiments investigated the effect of recent selective practice on the cost of switching between 2 tasks afforded by letter-digit pairs: alphabet arithmetic and shape comparison. Experiments 1 and 2 found a greater cost associated with switching to the more recently practiced task: evidence that task-set inertia contributes to switching costs. Experiment 3 found this effect to be limited to trials on which a recently trained stimulus followed another such stimulus: a result problematic for all current theories of task-set priming. Experiment 4 showed that the effect of recent practice was eliminated by active preparation for a task switch: It appears that endogenous task-set preparation reduces the effects of task-set inertia. ((c) 2003 APA, all rights reserved)     

准备时间和预先信息对任务切换的影响

两个实验采用任务线索范式,考察了准备时间和预先信息对任务切换的影响。实验1的准备时间为可预测,实验2的为不可预测,以考察准备时间的可预测性对任务切换的影响。以被试执行重复任务、切到任务和切离任务的正确反应时为主要指标,结果：（1）在准备时间可预测和不可预测两种条件下,准备时间和预先信息之间存在显著交互作用。在可预测条件中,重复和切到任务的反应时随着准备时间的增加而减少,但切到代价无显著变化,切离任务的反应时不受准备时间的影响,切离代价表现出反准备效应;不可预测条件下,三种任务的反应时都随着准备时间的增加而减少,但是切到和切离代价表现出反准备效应;（2）切离任务的反应时和代价显著高于切到任务的反应时和代价;（3）在这两种条件下,都没有出现切换代价的准备效应     

执行控制研究的重要范式——任务切换

任务切换是当前研究执行控制的重要范式。被试完成切换任务要慢于完成重复任务,并且往往有更高的错误率,这种差异称为切换损失。该文对此领域的研究进行了回顾,介绍了任务切换研究范式的发展,重点阐述了对切换损失来源的三种主要解释：任务重建、联结竞争和任务设置的惯性;指出任务切换与其他范式的结合是更深入地了解执行控制的可行途径。最后进行了简要的总结与展望。     

Task switching and the measurement of “switch costs”

The measurement of “switch costs” is held to be of interest because, as is widely believed, they may reflect the control processes that are engaged when subjects switch between two (or more) competing tasks. [In task-switching experiments, the reaction time (RT) switch cost is typically measured as the difference in RT between switch and non-switch (repeat) trials.] In this report we focus on the RT switch costs that remain even after the subject has had some time to prepare for the shift of task, when the switch cost may be approximately asymptotic (so-called residual switch costs). Three experiments are presented. All three experiments used Stroop colour/word, and neutral stimuli. Participants performed the two tasks of word-reading and colour-naming in a regular, double alternation, using the “alternating runs” paradigm (R. D. Rogers & S. Monsell, 1995). The experiments were designed to test the hypothesis that RT switch costs depend on a form of proactive interference (PI) arising from the performance of a prior, competing task. A. Allport, E. A. Styles and S. Hsieh (1994) suggested that these PI effects resulted from “task-set inertia”, that is, the persisting activation-suppression of competing task-sets, or competing task-processing pathways. The results confirmed the existence of long-lasting PI from the competing task as a major contributor to switch costs. Non-switch trials, used as the baseline in the measurement of switch costs, were also shown to be strongly affected by similar PI effects. However, task-set inertia was not sufficient to account for these results. The results appeared inconsistent also with all other previous models of task switching. A new hypothesis to explain these between-task interference effects was developed, based on the stimulus-triggered retrieval of competing stimulus-response (S-R) associations, acquired (or strengthened) in earlier trials. Consistent with this retrieval hypothesis, switch costs were shown to depend primarily on the S-R characteristics of the preceding task (the task that was switched from) rather than the upcoming task. Further, the effects of the other, competing task were found to persist over many successive switching trials, affecting switch costs long after the stimulus overlap (and hence the principal S-R competition) between the current tasks had been removed. Switch costs were also found to be affected by recent, item-specific experience with a given stimulus, in either the same or the competing task. Finally, the results showed that switch costs were massively affected by the ratio of the number of prior trials, in response to the same stimuli, that had implemented either the currently intended or the competing S-R mappings. None of these effects are predicted by current models of residual switch costs, which appeal to the differences in control processes assumed to be engaged in switch versus non-switch trials. Received: 31 March 1999 / Accepted: 23 July 1999     

Differential effects of cue changes and task changes on task-set selection costs

A task-switching paradigm with a 2:1 mapping between cues and tasks was used to separate cue-switching processes (indexed through pure cue-switch costs) from actual task-switching processes (indexed through additional costs in case of cue and task changes). A large portion of total switch costs was due to cue changes (Experiments 1 and 2), and cue-switch costs but not task-switch costs were sensitive to effects of practice (Experiment 1) and preparation (Experiment 2). In contrast, task-switch costs were particularly sensitive to response-priming effects (Experiments 1 and 2) and task-set inhibition (Experiment 3). Results suggest two processing stages relevant during task-set selection: cue-driven retrieval of task rules from long-term memory and the automatic application of rules to a particular stimulus situation.     

The coupling between spatial attention and other components of task-set: A task-switching investigation

Is spatial attention reconfigured independently of, or in tandem with, other task-set components when the task changes? We tracked the eyes of participants cued to perform one of three digit-classification tasks, each consistently associated with a distinct location. Previously we observed, on task switch trials, a substantial delay in orientation to the task-relevant location and tendency to fixate the location of the previously relevant task—“attentional inertia”. In the present experiments the cues specified (and instructions emphasized) the relevant location rather than the current task. In Experiment 1, with explicit spatial cues (arrows or spatial adverbs), the previously documented attentional handicaps all but disappeared, whilst the performance “switch cost” increased. Hence, attention can become decoupled from other aspects of task-set, but at a cost to the efficacy of task-set preparation. Experiment 2 used arbitrary single-letter cues with instructions and a training regime that encouraged participants to interpret the cue as indicating the relevant location rather than task. As in our previous experiments, and unlike in Experiment 1, we now observed clear switch-induced attentional delay and inertia, suggesting that the natural tendency is for spatial attention and task-set to be coupled and that only quasi-exogenous location cues decouple their reconfiguration.     

The nature of residual cost in regular switch response factors

Two experiments are presented that compare the residual cost found when switching from one task to another under predictable conditions. The aim of the study was to explore the roles played by the stimulus, the response, or both in the process of the mental set reconfiguration necessary to switch between two tasks. The experiments tested [Rogers, R. D., & Monsell, S. (1995). Cost of a predictable switch between simple cognitive tasks. Journal of Experimental Psychology: General, 124, 207-231] stimulus-cued-completion hypothesis and [Schuch, S., & Koch, I. (2003). The role of response selection for inhibition of task sets in task shifting. Journal of Experimental Psychology: Human Perception and Performance, 29, 92-105] hypothesis of response selection as the key factor in the nature of switch cost. In the first experiment, two conditions were created that varied in terms of a Go/No-Go signal: The Go trials were a replication of [Tornay, F. J., & Milán, E. G. (2001). A more complete task-set reconfiguration in random than in predictable task switch. The Quarterly Journal of Experimental Psychology, 54A, 785-803 Experiment 3]; The No-Go trials were identical to the first condition, except that participants did not execute a response in the trial n-1 (Schuch & Koch, 2003). In addition, the percentage of Go and No-Go trials was manipulated. The results showed that the cost was significant only in the high Go signal-frequency case (Experiment 2), with an abrupt offset in Go trials and a gradual offset in No-Go trials. Based on the results of these experiments, it was concluded that the crucial factor to complete a mental set reconfiguration is response-related and not stimulus-related.     

On the role of verbalization during task set selection: Switching or serial order control?

Recent task-switching work in which paper-and-pencil administered single-task lists were compared with task-alternation lists has demonstrated large increases in task-switch costs with concurrent articulatory suppression (AS), implicating a crucial role for verbalization during switching (Baddeley, Chincotta, & Adlam, 2001; Emerson & Miyake, 2003). Experiment 1 replicated this result, using computerized assessment, albeit with much smaller effect sizes than in the original reports. In Experiment 2, AS interference was reduced when a sequential cue (spatial location) that indicated the current position in the sequence of task alternations was given. Finally, in Experiment 3, switch trials and no-switch trials were compared within a block of alternating runs of two tasks. Again, AS interference was obtained mainly when the endogenous sequencing demand was high, and it was comparable for no-switch and switch trials. These results suggest that verbalization may be critical for endogenous maintenance and updating of a sequential plan, rather than exclusively for the actual switching process.     

Task-switch costs subsequent to cue-only trials

There is abundant evidence that there is a performance cost associated with switching between tasks. This “switch cost” has been postulated to be driven by task performance on the preceding trial, but recent research challenges any necessary role of previous task performance in driving the cost. Across three experiments, we investigated whether it is difficult to switch from a task that was prepared but never performed. We replicated the finding of a switch cost following cue-only trials (involving no task performance) whilst controlling for a potential cue-switching confound. This cost was larger than that following completed trials when preparation interval was short (300?ms), and it reduced significantly with a longer preparation interval (1000?ms) on the current trial. We also found that preparing only to attend to a particular visual dimension (colour or shape) was sufficient to drive a significant subsequent switch cost, which appeared to be residual in nature; we speculate that this cost may reflect the persistence of unfulfilled task intentions and/or a strategic slowing when consecutive intentions conflict.