Feedback-based gaze control in social and non-social information processing contexts

有效的凝视控制必然涉及对与眼动相关的效果(即反馈信号)的预期。基本的反馈环路是理解凝视控制的关键，凝视控制可以根据不同的目标进行分类。这些目标通过预期反馈信号来引导我们的凝视，包括认知知觉-信息目标(例如，通过确保视觉稳定性来优化信息获取，并通过凹陷来优化相关目标对象的处理)，与情感相关的目标(无论新信息获取是否有回报)，也包括社会目标(例如，环顾四周以评估和控制社会情境)。因此，与其他动作域类似，相同的动作(即眼球运动)可以在不同的解释域中产生结果。虽然眼动控制的理论很清楚反馈处理循环与知觉目标(例如，视觉稳定性)的相关性，但从未考虑过来自环境的反馈。除了利用凝视控制技术的人机界面(例如，现代智能手机中基于凝视的菜单导航)，凝视行为经常产生反馈的一个领域是社交互动，在社交互动中，凝视可以用来控制社交环境(例如，我可能会凝视他人，让他们回头看)。在这种情况下，社交背景特别适合研究基于反馈的眼动控制。在一项初步研究中，我们证明了预期反馈信号(即几何形状)取决于两个不同的眼球运动(向左/向右)，对凝视控制有显著影响。在目前的工作计划中，我们想要进一步研究这一现象的潜在机制。虽然初步研究只涉及非社会刺激，但我们在这里将核心思想转移到社会刺激(具有显著凝视的面孔)的处理上。具体地说，参与者的眼球运动(在社交条件下)会导致对面孔的感知，这些面孔要么是直接凝视，要么是回避凝视。在四个小节中，我们研究了a)外源性与内源性触发目标的影响，b)与凝视行为和反馈信号之间获得偶然性的过程相关的约束(例如，通过操纵偶然性和邻接性)，c)眼动控制中反馈信号处理的潜在神经特征，d)(情感和社会)控制水平之间的潜在交互，以及e)反馈回路中刺激维度之间的交互。通过参考相同的基于反馈的行动控制的多层次框架并通过学习相同的基本反馈原理，将工作方案嵌入到研究单位中。然而，与其他项目的一个核心区别是对眼球运动控制的关注，长期以来，眼球运动控制一直被视为与输入相关的注意过程的窗口，而不是其本身的一种行动方式。

Efficient gaze control necessarily involves the anticipation of effects (i.e., feedback signals) associated with eye movements. The underlying feedback loops are essential to understand gaze control on several levels, which can be classified with respect to different goals. These goals, which guide our gaze by means of anticipating feedback signals, include cognitive perceptual-informational goals (e.g., optimizing information uptake by ensuring visual stability and optimizing relevant target object processing through foveation), affectrelated goals (whether new information uptake is rewarding or not), but also social goals (e.g., looking around to evaluate and control a social situation). Thus, similar to other action domains, the same action (i.e., an eye movement) can have consequences in different domains of interpretation. While theories on eye movement control are well aware of the relevance of feedback-processing loops regarding perceptual goals (e.g., visual stability), feedback from the environment has yet never been considered. Beside human-computer interfaces utilizing gaze control techniques (e.g., gaze-based menu navigation in modern smartphones), an area in which gaze behavior regularly produces feedback is social interaction, where gaze can be used to control the social environment (e.g., I might gaze at others to have them look back). In this way, a social context is especially suited to study feedback-based control of eye movements. In a preliminary study, we demonstrated that anticipated feedback signals (i.e., geometrical shapes), which occurred contingent upon two distinct eye movements (to the left/right), had a significant impact on gaze control. In the present work program, we would like to further study the underlying mechanisms of this phenomenon. While the preliminary study only involved non-social stimuli, we here additionally transfer the core idea to the processing of social stimuli (faces with salient gaze). Specifically, participants´ eye movements (in social conditions) result in the perception of faces which either display direct or averted gaze. Within four subsections, we study a) the impact of exogenously vs. endogenously triggered goals, b) constraints related to the process of acquiring contingencies between gaze behavior and feedback signals (e.g., by manipulating contingency and contiguity), c) underlying neural signatures of feedback signal processing in oculomotor control, d) potential interactions between (affective and social) control levels, and e) interactions between stimulus dimensions in the feedback loop. The work program is embedded in the research unit by referring to the same multi-level framework of feedback-based action control and by studying the same basic feedback principles. However, a core difference to other projects is the focus on oculomotor control, which has long been regarded as a window to input-related attention processes only, and not as an action modality in its own right.