Time course of spoken word recognition

口语单词识别的时间进程

• 批准号: 7586690
• 负责人: Michael Kehas Tanenhaus
• 金额: $ 25.16万
• 依托单位: UNIVERSITY OF ROCHESTER
• 依托单位国家: 美国
• 财政年份: 2001
• 资助国家: 美国
• 起止时间: 2001-08-01 至 2012-03-31
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/7586690
• 关键词: 10 year old; Accent; Acoustics; Address; Adult; Affect; Applications Grants; Area; Auditory; Child; Classification; Cognition; Communication; Communication impairment; Cues; Deterioration; Development; Diagnostic tests; Exposure to; Eye Movements; Frequencies; Goals; Grant; Judgment; Language; Learning; Link; Maps; Mediating; Modeling; Natural regeneration; Neighborhoods; Outcome; Pattern; Perception; Phonetics; Positioning Attribute; Process; Psyche structure; Relative (related person); Research; Series; Speech; Strategic Planning; Structure; System; Time; Treatment Protocols; Ursidae Family; Weight; experience; knowledge base; lexical; lexical processing; natural language; phonology; programs; research study; sound; theories

DESCRIPTION (provided by applicant): The overall goal of the proposed program of research is to characterize the process by which listeners recognize words in fluent speech. Word recognition in this context, which comprises nearly all of our exposure to words in sentences, requires a remarkably sophisticated system that can rapidly and accurately identify words in real-time from among the thousands of alternatives in our mental lexicon. This process entails mechanisms by which acoustic/phonetic information is combined to create auditory word-forms that, when linked to meanings, serve as the entries in the lexicon. Unresolved issues for theories of spoken word recognition are (a) the units used to represent auditory word-forms in the lexicon, (b) when during the temporal unfolding of speech these units are used to access lexical candidates, (c) how these units are organized to represent words in the lexicon, and (d) how the mapping of sounds to meanings is reorganized when confronted with systematic shifts in the sound patterns of the input. We address these questions in studies of adults and children using standard judgment tasks (e.g., lexical decision), and eye movements, which are remarkably sensitive to the time course of speech and lexical processing, thus allowing for a detailed mapping of how asynchronous acoustic cues are integrated and how lexical candidates are evaluated dynamically. Many of the proposed experiments use miniature artificial languages to enable precise control over the sound patterns and frequencies of occurrence of words that cannot be achieved in natural languages. Studies using artificial lexica also enable us to study the process of lexical learning to determine how new words enter into, and may reorganize, the current lexical representation. The proposed experiments focus on (a) how context modulates the effects of lexical competitors, and (b) the dynamics of cue integration and lexical reorganization. The expected outcomes of this program of research will reveal fundamental aspects of spoken word recognition in adults and children that have important implications for models of lexical organization and access, for the mechanism by which the efficiency of lexical access is optimized in normal listeners, and for the effects of experience on the reorganization of both mature and immature lexicons. Successfully addressing these specific aims will provide a base of knowledge from which diagnostic tests and treatment regimens for children and adults with delays or deficits in spoken word recognition can be developed. The research outlined in this grant proposal bears on two of the four research areas outlined in the NICDC draft Strategic Plan for 2006-08: II. Development, Deterioration, Regeneration, and Plasticity of processes mediating communication, and III. Perception, Cognition, and Sensorimotor processing in normal and disordered communication.

描述（由申请人提供）：拟议的研究计划的总体目标是表征听众在流利的语音中识别单词的过程。在这种情况下，单词识别几乎包括我们在句子中接触过的所有单词，都需要一个非常复杂的系统，可以从我们的精神词典中成千上万的替代方案中实时快速准确地识别单词。这个过程需要将声学/语音信息组合在一起的机制，以创建听觉单词形式，并在含义时用作词典中的条目。口语识别理论的未解决的问题是（a）用来表示词典中听觉单词形式的单元，（b）在言语的时间展开期间，这些单位被用来访问词汇候选者，（c）这些单元如何组织为词典中的词语，以及（d）在词汇中的绘制方式，均值映射到均值的绘制方式，即在均值映射的范围内，该单元是在均值映射的，即在均值映射的范围。 输入。我们使用标准判断任务（例如词汇决定）和眼动运动在对成人和儿童的研究中解决了这些问题，这些问题对言语和词汇处理的时间过程非常敏感，从而允许详细映射异步的声学提示如何集成以及词汇候选者的动态评估。许多提出的实验都使用微型人工语言来精确控制声音模式和出现单词出现的频率，这些单词无法在自然语言中实现。使用人工词典的研究还使我们能够研究词汇学习的过程，以确定新单词如何进入（并可能重组）当前的词汇表示。提出的实验集中于（a）上下文如何调节词汇竞争者的效果，以及（b）提示整合和词汇重组的动力学。该研究计划的预期结果将揭示成人和儿童对口语识别的基本方面，这些方面对词汇组织和访问模型具有重要意义，对于正常听众中词汇访问的效率以及经验对成熟和不成熟词典的重新组织的影响的机制。成功解决这些具体目标将提供知识基础，从中可以从中为有延迟或口语识别延迟或缺陷的儿童和成人提供诊断测试和治疗方案。该赠款提案中概述的研究对2006 - 08年的NICDC战略计划草案草案中的四个研究领域中有两个：ii。介导交流的过程的发展，恶化，再生和可塑性以及III。正常和无序沟通中的感知，认知和感觉运动处理。