Neural mechanism of reading: Investigating rapid feedforward-feedback interaction

描述（由申请人提供）：我的长期目标是推进有关人类语言理解的认知和神经机制的知识。拟议的研究重点特别是视觉单词识别。理解视觉单词识别的进展对于理解熟练阅读及其无序状态的机制至关重要。我建议使用头皮记录的事件相关电位（ERPs）调查快速前馈反馈动态之间的低级别的单词形式和更高级别的词汇表征在最初的200毫秒的单词识别。许多先前的ERP研究发现，N400 ERP成分（刺激开始后约250 - 500毫秒）对与词汇访问或整合相关的变量敏感。大多数现有的研究表明，在N400之前的处理是专门用于视觉词形式的较低水平的感官分析。我最近的工作报告了新的证据，表明在单词识别过程中，低级单词形式和高级表征之间存在快速的前馈反馈交互作用。我们发现，在刺激开始后130毫秒（P130效应），由最小程度地歪曲上下文适当的真实的词的假词（例如，"她量了面粉，这样她就可以烤蛋糕了。").后来的影响，对N170而不是P1，是由上下文不支持的伪词引起的（例如，"她量了面粉，这样她就可以烤一吨。"）和非法非词（例如，"她量了面粉，这样她就可以烤一个SRDT。").我已经从快速访问的角度解释了P130效应 从刺激"ceke"的词汇表征CAKE。这种高水平的激活然后可以生成与输入"ceke"冲突的单词形式表示（"cake"）的自上而下的激励，并且P1调制可以反映这种冲突。因此，我们推断低层次的词形式和更高层次的表征之间的快速前馈反馈的相互作用。我们还不知道介导P130效应的更高层次表征的性质。本系列研究将探讨语义表征（具体目标1）、语音表征（具体目标2）和词频（具体目标3）在中介这些效应中的作用。这项工作将阐明复杂的认知过程之间的相互作用，服务于快速识别的一个字在约1/2秒后，视网膜刺激熟练阅读。了解这些基本过程的进展将为调查成人和儿童在阅读过程中的个体差异奠定基础，并可能发展成为阅读障碍的诊断工具。

DESCRIPTION (provided by applicant): My long-term goal is to advance knowledge about the cognitive and neural mechanisms of human language comprehension. The proposed research focuses specifically on visual word recognition. Progress toward understanding visual word recognition is critical to understanding the mechanisms of skilled reading and its disordered states. I propose to use scalp-recorded event- related potentials (ERPs) to investigate rapid feed forward-feedback dynamics between low-level word-form and higher-level lexical representations during the initial 200 msec of word recognition. Numerous prior ERP studies find that the N400 ERP component (approximately 250-500 msec post- stimulus onset) is sensitive to variables related to lexical access or integration. Most existing work has suggested that processing prior to the N400 is dedicated to lower-level sensory analysis of the visual word form. My recent work has reported novel evidence of rapid feedforward-feedback interactions between low-level word-form and higher-level representations during word recognition. We found enhancement of the P1 component of the ERP 130 msec after stimulus onset (P130 effect) by pseudowords that minimally malformed a contextually appropriate real word (e.g., "She measured the flour so she could bake a cake ..."). Later effects, on the N170 rather than the P1, were elicited by pseudowords that were contextually unsupported (e.g., "She measured the flour so she could bake a tont ...") and by illegal nonwords (e.g., "She measured the flour so she could bake a srdt ..."). I have explained the P130 effect in terms of rapid access of a lexical representation CAKE from the stimulus "ceke". This high-level activation can then generate top-down excitation of word-form representations ("cake") that conflict with the input "ceke", and the P1 modulation may reflect this conflict. Thus, we infer rapid feedforward-feedback interactions between low-level word- form and higher-level representations. We do not yet know the nature of the higher-level representations that mediate our P130 effects. The proposed series of studies will investigate the role of semantic representations (Specific Aim 1) phonological representations (Specific Aim 2) and lexical frequency (Specific Aim 3) in mediating these effects. This work will illuminate the complex interaction between cognitive processes that serve the rapid recognition of a word within about 1/2 second after the retinal stimulation during skilled reading. Progress in understanding these fundamental processes will set the stage for investigation of individual differences among adults and children in reading processes and can potentially be developed into diagnostic tools for reading disorders.