【作者】 陈永强；

【导师】 陈安涛； 郑涌；

【作者基本信息】 西南大学 ， 基础心理学， 2024， 博士

【摘要】 “吃一堑,长一智”、“前事不忘,后事之师”,这些成语都强调了个体从经验中进行学习是人生成长的重要途径。同时,这些成语也精炼地概括了认知心理学研究中的一个重要现象——冲突适应。在实验室研究中,个体在完成一致性任务(如Stroop、Simon和Flanker任务)时,先前经历过的冲突(吃一堑)有助于提升当前对冲突的解决效率(长一智)。个体能够对连续出现的冲突形成“适应性”反应,这一现象被命名为冲突适应效应(简称冲突适应)。冲突适应对于阐明认知控制的适应性机制有着重要的意义。冲突适应也已经作为个体适应能力的重要指标,被用于大量的应用研究中,以判别适应不良个体的行为和神经特征。然而,冲突适应的产生机制及其相应的神经基础还不清楚。这不仅阻碍了对冲突适应的深入理解,也会限制进一步的应用研究。因此,从行为和神经层面,阐明冲突适应的产生机制有着重要的理论意义和实践意义。研究者在冲突监测理论框架下来解释冲突适应的认知机制及神经基础。该理论认为,冲突适应是认知控制系统自上而下调整的结果,认知控制系统先监测到冲突,然后触发控制调整系统,增强任务相关信息的加工,和/或减少任务无关信息的加工,从而在后续任务中提高对冲突的解决效率。然而,冲突监测理论中还有一些假设,如冲突产生的必要条件,冲突产生在哪种水平上等亟需阐明。首先,以往研究忽视了对冲突产生的必要条件的考察。以往研究大多在包含反应执行的情境中考察冲突适应并检验冲突监测理论,但是冲突是否在反应执行之前产生,或者反应执行是否是冲突产生的必要条件,这还是一个亟待澄清的问题。其次,冲突产生于信息加工的哪个水平,是在早期刺激水平还是在晚期反应水平上?关于这一问题,冲突监测理论强调反应冲突触发控制调整,同时也认为刺激冲突可能也会被监测,并同样触发自上而下的控制调整。但由于以往分离刺激冲突和反应冲突的研究包含了促进效应的混淆,因此,刺激冲突是否可以被监测到,并触发控制调整还需要更进一步地考察。最后,对于冲突适应是否是领域一般性的,以往研究得到了不同的结论,其中一个原因可能是,以往研究都包含了反应执行,而反应执行可能会带来时间学习的混淆。研究者认为,时间学习是继特征整合和偶发学习的混淆之后,另一个可能混淆冲突适应的记忆学习因素。然而,对于控制时间学习因素的效果并不理想。因此,在特征整合和偶发学习之外,如何进一步控制时间学习的影响,也是一个值得探究的问题,且对于阐明冲突监测理论的隐含假设有着重要的促进意义。为了解决上述理论问题,本研究在重新梳理观察-反应范式的设计逻辑基础上,考察了以下三个具体问题:(1)冲突观察是否诱发冲突适应;(2)观察刺激冲突和反应冲突是否可以诱发冲突适应;(3)观察冲突诱发的冲突适应是否具有领域一般性。研究一旨在考察冲突观察是否可以诱发显著的冲突适应。对这一问题的考察,可以验证在排除特征整合、偶发学习和时间学习的基础上,冲突适应是否仍然存在;可以澄清冲突适应产生的必要条件;可以检验冲突监测和控制调整对冲突适应的不同贡献。实验1沿用唐丹丹等人的观察-反应设计,即个体在前一试次中仅观察Stroop(非)冲突刺激,在当前试次再呈现一个Stroop刺激(与前一试次不存在刺激或反应的重复),并要求个体进行按键反应,目的是重复验证冲突观察诱发的冲突适应,结果与预期一致。实验2对实验1的设计进行改进,排除偶发学习的混淆,考察冲突适应是否出现,结果与预期一致。实验3a进一步改进实验2的设计,缩短观察阶段的刺激呈现时长,再次考察冲突观察是否诱发冲突适应,结果与预期一致,即出现了显著的冲突适应。实验3b是对实验3a的改进,去除实验3a中的探测刺激,结果发现冲突观察可以诱发纯净的冲突适应。实验4采用实验3b的设计,并结合ERP技术,在神经层面验证冲突观察足以产生冲突,并触发控制调整,最终形成行为和神经层面的冲突适应。实验4还采用多变量模式分析(Multivariate Pattern Analysis,MVPA),检验了冲突监测和控制调整对冲突适应的不同贡献。结果发现,观察冲突在N450和SP成分上表现出了显著的Stroop效应,且诱发了随后反应任务在行为和SP成分上的冲突适应;另外,MVPA在620～760ms,以及800～1200ms时间窗内成功解码了表征冲突适应的cI和iI条件,且解码效果600～1200ms时间窗内表现出神经活动的跨时间稳定性。总之,研究一表明观察-反应设计在研究冲突适应中的可行性,并验证了不相容的表征激活就足以产生冲突,并触发随后的控制调整,最终出现冲突适应。而且,监测到冲突后,触发的持续而稳定控制调整对冲突适应产生至关重要。研究二考察观察刺激冲突和反应冲突是否可以诱发显著的冲突适应。对这一问题的考察,可以验证是否可以分离出纯净的刺激冲突和反应冲突;冲突监测系统是否对刺激冲突也敏感;观察刺激冲突和反应冲突是否都可以诱发冲突适应。实验5a以一致试次为基线分离刺激冲突和反应冲突,结果发现,观察刺激冲突和反应冲突都可以诱发显著的冲突适应,但主要反映在当前的反应冲突上。需要指出的是,以一致试次为基线,可能导致分离的刺激冲突和反应冲突包含促进效应的混淆。实验5b以与颜色无关的中性词为基线分离刺激冲突和反应冲突,结果与实验5a基本一致,另外实验5b还发现了观察反应冲突诱发当前刺激冲突中显著的冲突适应。这说明与颜色无关的中性词可以成功分离上述两种不同性质的冲突,且两种冲突都诱发了当前试次上的冲突适应,但未发现观察刺激冲突诱发当前刺激冲突上的冲突适应。实验6沿用实验5b的设计,通过ERP技术,考察监测系统是否对刺激冲突敏感,以及是否可以在神经层面诱发冲突适应。结果发现,N450对刺激冲突和反应冲突都敏感;而且,在神经层面上,观察刺激冲突和反应冲突都可以诱发当前刺激冲突和反应冲突上的冲突适应。总之,研究二发现采用与颜色无关的中性词可以成功分离纯净的刺激冲突和反应冲突,且监测系统(以N450成分为代表)对不同性质的冲突都敏感。重要的是,在神经层面,不同性质的冲突之间可以相互诱发显著的冲突适应。需要指出的是,研究一和研究二都采用了Stroop任务。那么,冲突观察诱发的冲突适应是否可以扩展到其他冲突任务中,以及冲突适应是否可以在不同的冲突任务之间相互迁移。研究三旨在考察冲突观察诱发的冲突适应是否具有领域一般性。实验7采用研究一中实验3b的设计,但是采用字母版Flanker任务,验证Stroop任务中冲突观察诱发的冲突适应是否可以扩展到Flanker任务中。结果与预期一致,即,在Flanker任务也出现了冲突观察诱发的冲突适应。实验8采用转换设计,让参与者观察Stroop或Flanker刺激(冲突或非冲突),然后在下一试次中对Flanker或Stroop刺激(冲突或非冲突,且与观察阶段的刺激或反应特征无重复)做按键反应。结果发现,观察Stroop可以诱发反应Flanker上的冲突适应,同样地,观察Flanker也可以诱发反应Stroop上的冲突适应。实验9则沿用实验8的设计,通过ERP技术,为冲突适应的领域一般性提供神经层面的证据。结果发现,在观察阶段,观察Stroop任务中,在N450和SP上表现出了显著的Stroop效应;观察Flanker任务中,在N2和P3上表现出了显著的Flanker效应。这再次验证了实验4中的结论,即观察冲突就可以产生冲突,并触发随后的控制调整。在反应阶段,观察Stroop任务后,诱发了反应Flanker任务中P3成分上的冲突适应;观察Flanker任务后,诱发了反应Stroop任务中N450和SP成分上的冲突适应。而反应Flanker中没有发现N2成分上的冲突适应,可能Stroop冲突出现较晚,且冲突调整时间较长,因此掩盖了出现较早的Flanker冲突,但未影响对Flanker冲突的调整加工。总之,研究三在Stroop和Flanker任务中再次验证了观察冲突就足以产生冲突,并触发随后的控制调整。重要的是,观察冲突诱发了不同冲突任务上的冲突适应。即,冲突观察诱发的冲突适应具有领域一般性。这种一般性可以用通过抽象水平的相似性和基于干扰的工作记忆模型解释。综合来说,当前研究通过三个研究的9个实验,在严格控制的行为实验中,结合ERP技术和MVPA分析,对冲突观察是否诱发冲突适应,观察哪种冲突可以诱发冲突适应,与冲突观察诱发的冲突适应是否具有领域一般性,这一系列问题进行了行为和神经层面的系统考察。当前研究较为清晰地表明了,不相容冲突表征的同时激活是冲突产生的必要条件,冲突可以产生于刺激水平和反应水平,且都可以触发自上而下的控制调整,最终形成具有领域一般性的、相对纯净的冲突适应。此外,当前研究还将分离纯净刺激冲突和反应冲突的设计首次应用冲突适应研究,并得到了具有启发性的结果。需要强调的是,当前研究的主要结果,均是在排除特征整合、偶发学习和时间学习因素的基础上得到的,因此支持并扩展了冲突监测理论,同时也为未来整合冲突适应的现有研究提供了一个可遵循的实验设计方案。最后,本研究的发现从理论上厘清了冲突监测理论的内涵假设,拓宽了冲突监测理论的外延,对于全面深入地理解冲突监测理论和冲突适应效应有着重要的理论意义。同时,为冲突适应的应用研究夯实了理论基础,对于判别适应不良个体的适应功能特征,促进个体的适应能力,提升个体的心理健康水平有着重要的现实意义。更多还原

【Abstract】 The proverbs,“Fall into a pit,gain in wisdom” and “Don’t forget past lessons,they serve as a guide for the future”,emphasize that learning from experience is an important pathway to personal growth.At the same time,these idioms succinctly encapsulate an important phenomenon in the research of cognitive psychology — the conflict adaptation.Researchers in the laboratory have found that individuals who have experienced conflicts in previous tasks(“fall into a pit”)contribute to an improvement in the efficiency of resolving current conflicts(“gain in wisdom”)when performing tasks involving consistency(such as Stroop,Simon,and Flanker tasks).Individuals can form “adaptive”responses to consecutively occurring conflicts,a phenomenon known as the conflict adaptation effect.The conflict adaptation effect helps to elucidate the adaptative mechanisms of cognitive control.The conflict adaptation effect has also been recognized as a crucial indicator of individual adaptive capacity,used extensively in applied research to discern the behavioral and neural characteristics of maladaptive individuals.However,researchers have not yet reached a consensus on the mechanisms underlying conflict adaptation and its corresponding neural foundations.This not only hinders a deeper understanding of conflict adaptation but also limits further applied research.Therefore,elucidating the mechanisms underlying conflict adaptation from both behavioral and neural perspectives holds important theoretical and practical significance.The researchers interpret the cognitive mechanisms and neural basis of conflict adaptation within the framework of Conflict Monitoring Theory.The Conflict Monitoring Theory posits that conflict adaptation results from top-down adjustments of the cognitive control system.According to this theory,the cognitive control system first detects conflict,triggering adjustments to enhance the processing of task-relevant information and/or reduce the processing of task-irrelevant information.This,in turn,improves the efficiency of conflict resolution in subsequent tasks.However,within the Conflict Monitoring Theory,there are still some assumptions that urgently need clarification,such as the necessary conditions for conflict generation and the level at which conflicts occur.Firstly,previous research has overlooked the examination of the necessary conditions for conflict generation.Most studies have predominantly investigated conflict adaptation in contexts involving response execution and tested the conflict monitoring theory.However,whether the conflict arises before response execution or if response execution is a necessary condition for conflict generation remains a question that urgently requires clarification.Secondly,at which stage of information processing does the conflict occur—during the early stages of stimulus processing or the later stages of response execution? The Conflict Monitoring Theory emphasizes that response conflicts trigger control adjustments,while also suggesting that stimulus conflicts may be monitored and similarly trigger top-down control adjustments.However,due to confounding effects in past research that separated stimulus conflict and response conflict,further investigation is needed to determine if stimulus conflict can be detected and trigger control adjustment.Given the premise of elucidating the theoretical implications of the Conflict Monitoring Theory,it is necessary to further investigate its theoretical extension,which can be examined through the generality of conflict adaptation.Whether conflict adaptation is general remains inconclusive in current research,and one reason might be that previous studies included response execution,potentially introducing confounding effects of temporal learning.Temporal learning is another memory learning factor,alongside Feature Integration and Contingency Learning,that may confound conflict adaptation.However,researchers have not achieved satisfactory control over the effects of temporal learning.Therefore,beyond Feature Integration and Contingency Learning,exploring how to further control the impact of temporal learning is a worthwhile endeavor.This inquiry holds important implications for elucidating the implicit assumptions of the Conflict Monitoring Theory.To address the theoretical issues mentioned above,this study,based on a reevaluation of the design logic of the observation-response paradigm,examined the following three specific questions:(1)whether conflict observation induces conflict adaptation;(2)whether observing stimulus conflicts and response conflicts can induce conflict adaptation;(3)whether conflict adaptation induced by observing conflicts exhibits cross-task/cross-conflict generality.Questions(1)and(2)aimed to clarify the theoretical implications of the Conflict Monitoring Theory,and on this basis,question(3)investigated the theoretical extension of the Conflict Monitoring Theory.Study 1 aimed to examine whether conflict observation could induce significant conflict adaptation.The investigation into this question served to verify whether conflict adaptation persists after excluding feature integration,contingency learning,and temporal learning.It sought to clarify the necessary conditions for the emergence of conflict adaptation and examine the distinct contributions of conflict monitoring and control adjustment to conflict adaptation.Experiment 1 adopted the LOOK-DO design by Tang et al.,wherein participants observed Stroop(non)conflict stimuli in the previous trial and then responded to a different Stroop stimulus in the current trial.The purpose was to replicate and validate the occurrence of conflict adaptation induced by conflict observation,and the results were as expected.Experiment 2 improved the design of Experiment 1 to eliminate confounding effects from contingency learning,and the results aligned with expectations,demonstrating the presence of conflict adaptation.Experiment 3a further enhanced the design of Experiment 2 by shortening the stimulus presentation duration during the LOOK phase.Once again,it confirmed that conflict observation induced conflict adaptation,showing significant conflict adaptation.Experiment 3b refined Experiment 3a by removing detection stimuli,revealing that conflict observation can induce pure conflict adaptation.Experiment 4 adopted the design of Experiment 3b,combining ERP technology to neurologically validate that conflict observation is sufficient to generate conflict and trigger control adjustment,ultimately resulting in both behavioral and neural conflict adaptation.Additionally,Experiment 4 used Multivariate Pattern Analysis(MVPA)to examine the differential contributions of conflict monitoring and control adjustment to conflict adaptation.The results indicated significant Stroop effects on the N450 and SP components during conflict observation,inducing subsequent behavioral and SP component conflict adaptation.Moreover,MVPA successfully decoded conditions representing conflict adaptation(cI and iI)in the 620～760ms and 800～1200ms time windows,demonstrating cross-temporal stability in neural activity decoding from 600 ms to 1200 ms.In conclusion,Study 1 demonstrated the feasibility of the LOOK-DO design in studying conflict adaptation and validated that the activation of incompatible representations during observation is sufficient to generate conflict and trigger subsequent control adjustments,leading to conflict adaptation.The study also emphasized the crucial role of sustained and stable control adjustments triggered after detecting conflicts in the manifestation of conflict adaptation.Study 2 investigated whether observing stimulus conflict and response conflict can induce significant conflict adaptation.The examination of this question aimed to verify if it is possible to isolate pure stimulus conflict and response conflict,whether the conflict monitoring system is sensitive to stimulus conflict,and if observing both types of conflict can induce conflict adaptation.Experiment 5a used congruent trials as a baseline to separate stimulus conflict and response conflict.The results showed that observing both stimulus conflict and response conflict can induce significant conflict adaptation,primarily reflected in the current response conflict.It is worth noting that using congruent trials as a baseline may introduce confounding effects related to facilitating influences in the separated stimulus conflict and response conflict.In Experiment 5b,neutral words unrelated to color were used as a baseline to separate stimulus conflict and response conflict,and the results were generally consistent with Experiment 5a.Additionally,Experiment 5b found that observing response conflict induced significant conflict adaptation in the current stimulus conflict.This suggested that neutral words unrelated to color can successfully isolated the two different types of conflict,and both types of conflict induced conflict adaptation in the current trials.However,no evidence was found for conflict adaptation in the current stimulus conflict when observing stimulus conflict.Experiment 6 adopted the design of Experiment 5b and used ERP technology to examine whether the monitoring system was sensitive to stimulus conflict and if conflict adaptation can be induced at the neural level.The results revealed that N450 was sensitive to both stimulus conflict and response conflict.Importantly,at the neural level,observing both stimulus conflict and response conflict can induce significant conflict adaptation in the current stimulus conflict and response conflict.In conclusion,Study 2 found that using neutral words unrelated to color successfully separates pure stimulus conflict and response conflict,and the monitoring system(represented by the N450 component)is sensitive to conflict of different natures.Crucially,at the neural level,different types of conflict can mutually induce significant conflict adaptation.Noting that both studies 1 and 2 employed the Stroop task.Therefore,a question arises: can conflict adaptation induced by conflict observation extend to other conflict tasks,and can conflict adaptation transfer between different conflict tasks? Study 3 aimed to examine whether conflict adaptation induced by conflict observation is domain-general.Experiment 7 adopted the design from Experiment 3b in Study 1 but used the letter version of the Flanker task to verify if conflict adaptation induced by observing conflicts in the Stroop task could extend to the Flanker task.The results were as expected,indicating that conflict adaptation induced by observing conflicts also occurred in the Flanker task.Experiment 8 employed a transfer design,where participants observed Stroop or Flanker stimuli(conflict or non-conflict)and then made key responses to Flanker or Stroop stimuli(conflict or non-conflict,with no repetition of stimulus or response features from the LOOK phase)in the next trial.The results showed that observing Stroop induced conflict adaptation in subsequent responses to Flanker,and vice versa.Experiment 9 followed the design of Experiment 8 and used ERP technology to provide neural evidence for the generality of conflict adaptation.The results revealed that,during the observation phase,observing the Stroop task exhibited significant Stroop effects on the N450 and SP components.Similarly,observing the Flanker task showed significant Flanker effects on the N2 and P3 components.This again validated the conclusion from Experiment 4,suggesting that observing conflicts was sufficient to generate conflicts and trigger subsequent control adjustments.During the DO phase,observing the Stroop task induced conflict adaptation on the P3 component in subsequent responses to the Flanker task,and observing the Flanker task induced conflict adaptation on the N450 and SP components in subsequent responses to the Stroop task.No conflict adaptation on the N2 component was found during responses to Flanker,possibly because Stroop conflicts appeared later,and the conflict adjustment time was longer,thus overshadowing earlier Flanker conflicts.However,this did not affect the adjustment processing of Flanker conflicts.In summary,Study 3,conducted with both Stroop and Flanker tasks,once again confirmed that observing conflicts was sufficient to generate conflicts and trigger subsequent control adjustment.Importantly,observing conflict induced conflict adaptation on different conflict tasks.In other words,conflict adaptation induced by observing conflicts exhibited cross-task/cross-conflict generality.This generality can be explained through the abstraction level of similarity and an interferencebased working memory model.To sum up,the current study,comprising 9 experiments across three study,rigorously examined behavioral and neural aspects of whether conflict observation induced conflict adaptation,which type of conflict triggered conflict adaptation,and whether conflict adaptation induced by conflict observation exhibited cross-task/crossconflict generality.The research consistently demonstrated that the simultaneous activation of incompatible conflict representations was a necessary condition for conflict generation.Conflict can occur at both stimulus and response levels,and both can trigger subsequent top-down control adjustments,resulting in a cross-task/cross-conflict generality of relatively pure conflict adaptation.Importantly,the study utilized a design separating pure stimulus conflict and response conflict,yielding insightful results for the first time in conflict adaptation research.It is crucial to emphasize that the main findings of the current study were obtained after controlling for feature integration,contingency learning,and temporal learning factors.Therefore,the study not only supported and extended the Conflict Monitoring Theory,but also provided a followable experimental design for integrating existing research on conflict adaptation in the future.Lastly,the theoretical contributions of this research clarify the internal assumptions of the Conflict Monitoring Theory and broaden its external scope,contributing significantly to a comprehensive understanding of the theory and the conflict adaptation effect.Moreover,the study establishes a theoretical foundation for applied research on conflict adaptation,aiding in identifying adaptive characteristics in maladaptive individuals,promoting adaptive capacity,and enhancing psychological well-being.更多还原