A binding perspective on task and language switching: Exploring the influence of a task-irrelevant context on episodic retrieval

人类行为的特点是具有极大的灵活性。例如，我们能够在交谈中灵活地从一个任务切换到另一个任务，或者从一种语言切换到另一种语言。然而，这种灵活性也伴随着成本。在任务转换和语言转换两种情况下，与任务或语言转换相比，受试者在任务或语言重复中表现得更好。从特征绑定的角度来看，这种重复益处可以用这样的假设来解释，即一个特征的重复提取了前一段情节(即任务、刺激和反应特征的绑定)，从而导致了任务或语言的正启动。重要的是，似乎不仅与任务相关的特征可以检索上一集，而且与任务无关的特征也可以检索到，我们称之为“上下文”特征。在本项目中，我们探索了任务无关语境在先前特征绑定的情景提取中的作用。在其他方面，我们对情节检索的时间方面感兴趣，例如，为了有效地检索先前的情节，必须在何时呈现与任务无关的上下文(例如，单词的字体颜色或彩色单词背景)的问题。此外，我们调查了与任务无关的上下文是否只能检索单个情节或多个情节。为此，我们在任务切换范式中引入了上下文-任务权变。因此，两个任务之一(例如，尺寸判定：是比鞋盒更大或更小的对象)更多地与一个上下文特征(例如，红色单词背景)组合，而不是与另一个上下文特征(蓝色单词背景)组合。以这种方式，很可能出现上下文和任务之间的关联，从而上下文特征的重复不仅可以检索最近的一集，而且可以检索相关联的任务。在各种实验中，我们研究了特征绑定和情景提取与偶然事件的关联学习之间的相互作用。此外，我们将通过努力促进新单词的学习，将功能绑定的观点转移到应用环境中。在这样做的过程中，我们的目标是使用上下文特征的重复或切换来具体支持对正确单词的检索，并减少对之前错误回答的检索。通过这种方式，我们的项目有助于通过功能绑定的视角来描述现实生活中的现象。

Human behaviour is characterised by its great flexibility. We are, for example, able to switch flexibly from one task to another or from one language to another in the middle of a conversation. However, this flexibility also comes with costs. In both task-switching and language-switching situations, subjects perform better in a task or language repetition compared to a task or language switch. In terms of a feature-binding perspective, such a repetition benefit can be explained by the assumption that the repetition of one feature retrieves the preceding episode (i.e., a binding of task, stimuli and response features) and thus leads to a positive priming of a task or language. Importantly, it seems that not only task-relevant features can retrieve a previous episode, but also task-irrelevant features, which we termed “context” features. In the present project, we explore the role of the task-irrelevant context on the episodic retrieval of a previous feature bindings. Among other things, we are interested in temporal aspects of episodic retrieval, for example, the question of when the task-irrelevant context (e.g., the font colour of a word or a coloured word background) must be presented in order to effectively retrieve a previous episode. Furthermore, we investigate whether a task-irrelevant context can retrieve only a single episode or multiple episodes. To this end, we introduce a context-task contingency in a task-switching paradigm. One of the two tasks (e.g., size decision: is an object bigger or smaller than a shoe box) is thereby more often combined with one context feature (e.g., a red word background) than with the other context feature (blue word background). In this way, an association between context and task is likely to emerge, so that the repetition of a context feature may not only retrieve the most recent episode, but also the associated task. In various experiments, we examine the interplay between feature binding and episodic retrieval on the one hand and associative learning of contingencies on the other. In addition, we will transfer the feature-binding perspective into an applied setting by trying to promote the learning of new words. In doing so, we aim to use the repetition or switch of contextual features to specifically support the retrieval of a correct word and reduce the retrieval of previous incorrect responses. In this way, our project contributes to describing real-life phenomena through a feature-binding perspective.