Integration of oculomotor and cognitive-dynamical modeling of eye-movement control during reading

在过去的15年里，关于阅读中眼动控制的研究是由单词识别和注意力分配的数学模型推动的。在控制眼球运动的水平上，运动控制的框架理论，即感觉运动整合的贝叶斯模型(Körding&Wolpert，2004)取得了进展，该水平决定了单词内注视位置的精确规划。在这一建议中，我们建议在认知-动态SWIFT选词模型(Engbert等人，2005)和面向过程的动眼运动选词模型(Krüel&Engbert，2014)的基础上建立一个综合模型。有了这样的模型，首次有可能对眼球运动过程和正在进行的认知之间的相互作用进行定量预测。在工作计划中，分为同等权重的理论和实验部分，我们将使用镜像颠倒/倒置文本的阅读实验，以检验我们的假设，即面向过程的两个模型组成部分的整合。由于实验结果依赖于凝视条件实验，最重要的是边界技术(Rayner，1975)，我们建议进行一系列相应的实验模型测试。

During the last 15 years, research on the control of eye movements in reading was driven by mathematical models of word recognition and attention allocation. On the level of oculomotor control, which determines the precise planning of fixation positions within words, advances were obtained by Bayesian models of sensorimotor integration, a framework theory of movement control (Körding & Wolpert, 2004). In this proposal, we propose to develop an integrated model based on the cognitive-dynamical SWIFT model of word selection (Engbert et al., 2005) and a process-oriented oculomotor model of word targeting (Krügel & Engbert, 2014). With such a model, quantitative predictions on the interaction of oculomotor processes and ongoing cognition will be possible for the first time. In the work program, which is divided into equally weighted theoretical and experimental parts, we will use reading experiments on mirror-reversed/inverted texts that permit the test of our assumption for the process-oriented integration of both model components. Since experimental results rely on gaze-contingent experimentation, most importantly the boundary technique (Rayner, 1975), we propose to run a series of corresponding experimental model tests.