A Multi-Component Approach to Extinction in Pavlovian Learning

描述（由申请人提供）：这里提出的研究将探讨消除不想要的学习行为所涉及的心理机制。具体来说，我们将从巴甫洛夫学习的多成分模型的角度来研究灭绝行为现象。其长期目标是确定在何种条件下，灭绝治疗可能会对反应产生持久的影响，并提供对灭绝过程的更完整的理解。一个基本前提是，消退疗法可能对巴甫洛夫学习的不同组成部分（例如，其更多的“认知”或“情感/动机”组成部分）有不同的影响。本研究将通过使用将学习分解为单独的认知和情感/动机成分的方法来研究这一点，以便分别检查对这些不同成分的消退效应。研究方法将使用啮齿动物模型（褐家鼠）进行食欲学习范式（杂志方法条件反射），因为这种范式已经成功地用于研究巴甫洛夫学习的多个组成部分。结果将与心理健康问题相关，因为这项研究将确定消退治疗何时可能导致持久的反应丧失，并且更一般地说，将指出需要研究消退治疗对学习行为的不同方面（例如，认知和情感）的影响。该研究将探讨4个具体目标。基于初步的发现，Specific Aim 1将测试假设，即以某种方式受到“损害”的有限训练的刺激将特别容易灭绝。可以预见，基于认知的学习（即由感觉特异性关联控制）将被应用于这些脆弱刺激的灭绝所破坏。具体目标2将检验有限训练后的灭绝是持久的假设。如果是这样，在旨在促进这种恢复的各种情况下（即自发恢复、恢复、更新和反转学习），由感觉特异性关联引起的控制减退应该不能恢复。Specific Aim 3将直接测试灭绝是否比基于认知的学习更快地破坏基于情感的学习。这一结果将指出在评估灭绝时监测学习的各个组成部分的重要性。具体目标4将开始探索一种基本机制，提出解释灭绝，负预测误差。这将通过探索预测误差在一般“情感”价值货币中编码的假设来检验。这个想法将通过在灭绝期间结合食欲和厌恶训练的线索来验证，以确定它们的总结效应是否预测了灭绝的总体影响。总的来说，这些研究的结果强调需要评估灭绝中的多种反应系统，并确定何时灭绝可能起作用，何时灭绝可能不起作用，这将具有重要的理论和治疗意义。

DESCRIPTION (provided by applicant): The research proposed here will examine the psychological mechanisms involved in the elimination of unwanted learned behaviors. Specifically, the behavioral phenomenon of extinction will be examined from the perspective of a multi-component model of Pavlovian learning. Its long-range goal is to identify conditions under which extinction treatments may be expected to have durable effects on responding, and to provide a more complete understanding of extinction processes. One basic premise is that extinction treatments may have distinct effects on different components of Pavlovian learning (e.g., its more "cognitive" or "emotional/motivational" components). The present research will investigate this by using methods that dissect learning into its separate cognitive and emotional/motivational components in order to separately examine extinction effects on these distinct components. The research approach will be to use a rodent model (Rattus norvegicus) performing in an appetitive learning paradigm (magazine approach conditioning) because this paradigm has been used successfully to study the multiple components of Pavlovian learning to be dissected here. The results will be relevant to mental health concerns because this research will establish when extinction treatments may result in durable response loss, and, more generally, will point to the need to look at extinction treatment effects on different (e.g., cogniive and emotional) aspects of learned behavior. The research will explore 4 specific aims. Based on preliminary findings, Specific Aim 1 will test the hypothesis that stimuli given limited training o that have been "compromised" in some fashion will be especially vulnerable to extinction. It is anticipated that cognitively based learning (i.e., control by sensory- specific associations) will e undermined by extinction applied to these vulnerable stimuli. Specific Aim 2 will test the hypothesis that extinction after limited training is durable. If so, reduced control by sensory-specific associations should fail to recover in various situations designed to promote such recovery (i.e., spontaneous recovery, reinstatement, renewal, and reversal learning). Specific Aim 3 will directly test whether extinction undermines emotionally based learning more rapidly than it does cognitively based learning. This result will point to the importance of monitoring separate components of learning in assessing extinction. Specific Aim 4 will begin an exploration of one basic mechanism proposed to explain extinction, negative prediction errors. This will be examined by exploring the hypothesis that prediction errors are coded in a general "emotional" value currency. The idea will be tested by combining appetitive and aversively trained cues during extinction to determine if their summative effects predict the overall impact of extinction. Overall, the results from these studies will have important theoretical and therapeutic implications by emphasizing the need to assess multiple response systems in extinction and by determining when extinction may or may not be expected to work.