提取干扰对内隐记忆的影响及其机制研究

“Implicit memory” typically refers to a nonconscious, unintentional form of memory retrieval. Experiments on memory have revealed that previous experiences can facilitate performance on indirect tasks that do not require intentional memorizing. Such a facilitation effect is typically referred to as “priming.” Traditionally, researchers have agreed that implicit memory reflects certain automatic processes. This notion has been supported by studies on divided attention (DA), which showed that DA at encoding results in priming on subsequent indirect tests. However, our preliminary experiments have indicated that DA at retrieval significantly reduces priming on lexical decision-based tasks, whereas DA during encoding leaves priming intact. Since automatic processes are traditionally thought not to be prone to interference from other processes, our preliminary experiments cast doubt on the traditional notion that implicit memory is an automatic process. However, several other studies have failed to record a retrieval-phase DA effect on indirect memory tests such as perceptual identification, word-stem completion, and category exemplar generation. Thus, the relationship between DA at retrieval and implicit memory remains an open question.. This research will systematically investigate the effects of retrieval-phase DA on implicit memory, and the relationship between implicit memory and automatic processing. The project is comprised of a series of three studies. The first will explore what types of implicit memory might be affected by retrieval-phase DA. Implicit memory can be divided into perceptual and conceptual priming, depending on indirect tests based on perceptual or conceptual processes. Alternatively, implicit memory can be divided into identification and production priming, depending on indirect tests requiring identification of a stimulus or production of a response from one of several candidate options. More importantly, dissociation has been observed between the different categories of implicit memory. Lexical decisions are classified as perceptual identification, whereas the tasks employed in those studies that failed to observe retrieval-phase DA effects tended to be production tasks. This begs the question of whether different types of implicit memory are affected differently by retrieval-phase DA. Experiments 1 to 3 will explore this question.. The second study will consider why retrieval-phase DA interferes with unconscious implicit memory. Due to limited resources, available processing capacity must be shared between tasks, rendering the processing of both less efficient. Studies have found that retrieval-phase DA has no effect on explicit memory performance, but is accompanied by significant secondary task costs. This implies that explicit retrieval could be a more obligatory process that runs relatively uninterrupted, while drawing on significant attentional resources. We posited that unconscious implicit retrieval would not be an obligatory process, and might be affected by competition with secondary task demands. This competition could be for specific resources or general, non-specific items. Three hypotheses were then proposed for the retrieval-phase DA effect on implicit memory: (1) competition for resources to access the perceptual representational system; (2) competition for resources to trigger a response; and (3) competition for general, non-specific resources. These hypotheses will be tested in Experiments 4 through 10.. The third study will explore the neural mechanisms of the retrieval-phase DA effect on implicit memories that use event-related potential (ERP), based on the results of the first and second studies. A thorough analysis of DA and implicit retrieval will then be conducted. Based on this investigation, we will close by comprehensively discussing the relationship between implicit memory retrieval and automatic processing.

“无意识的内隐记忆应该是自动化加工”这一观念一直统治着记忆领域，其证据主要源于内隐记忆通常不受编码时注意控制如干扰的影响。然而我们前期研究发现，虽然编码干扰不会影响内隐记忆，但提取时设置的干扰却会减少内隐记忆。这是否意味着无意识的内隐记忆并不是完全的自动化加工，容易受到提取干扰的破坏？本项目拟通过三方面研究系统考察提取干扰与内隐记忆的关系及其产生机制。研究一主要探讨提取干扰是否对不同类型内隐记忆产生不同影响，确定内隐记忆提取干扰效应的产生条件。研究二进一步探讨提取干扰是如何影响内隐记忆，以及这种干扰效应与注意资源的关系，分析内隐记忆提取干扰效应的产生机制。研究三采用事件相关电位技术，考察提取干扰影响内隐记忆的时间进程和大脑分布，了解内隐记忆提取干扰效应的认知神经机制。最后，综合三方面的研究结果，系统分析内隐记忆与提取干扰、注意资源之间的关系，为“内隐记忆提取是否自动化加工”提供判断依据。

“无意识的内隐记忆应该是自动化加工”这一观念一直统治着记忆领域，其证据主要源于内隐记忆通常不受编码时注意控制如干扰的影响。然而本项目的前期研究发现，虽然编码干扰不会影响内隐记忆，但提取时设置的干扰却会减少内隐记忆。这是否意味着无意识的内隐记忆并非完全的自动化加工？.本项目通过两方面研究系统考察提取干扰与内隐记忆的关系及其产生机制。研究一主要探讨提取干扰是否对不同类型内隐记忆产生不同影响，确定内隐记忆提取干扰效应的产生条件。研究二进一步探讨提取干扰是如何影响内隐记忆，以及这种干扰效应与注意资源的关系，同时结合事件相关电位（ERP）技术，分析内隐记忆提取干扰效应的产生机制。.本项目得到的主要结果如下：.一是发现内隐记忆提取干扰效应是稳定存在的。提取干扰对识别式或产生式、知觉或概念内隐测验中的启动效应都会产生影响，并且不受提取意识的调节。.二是发现内隐记忆提取干扰效应是源于双任务产生的一般认知资源竞争。项目曾提出三种假设：知觉加工资源竞争假说、反应资源竞争假说、一般注意资源竞争假说，并通过系列实验加以验证。结果发现，当干扰刺激与记忆刺激先后呈现，或设置为不同类型，以去除同时呈现或相同类型刺激所引起的知觉表征加工资源竞争；在提取阶段设置ONE-BACK继时反应干扰任务去除动作反应资源竞争，但均未能影响提取干扰效应。这些结果意味着提取干扰效应应该是源于双任务操作所引起的一般的、非特异性的注意资源竞争。基于这一推测，实验采用双任务继时操作范式，将干扰任务改为go/no-go 任务，或延长双任务的时间间隔，以降低双任务操作引起的注意资源竞争。结果发现，这些设置均可以调节提取干扰对启动效应的影响，验证了一般注意资源竞争假说。.综上，本项目验证了“内隐记忆的提取加工并非完全自动化，也会受到注意资源的调节”，该一结论将促使人们重新审视内隐记忆的本质。同时研究结果可为内隐记忆如何更好地应用于解释和改变人类的行为提供指导作用。