Behavioral and Neurologic Factors in Speech Learning

描述（由申请人提供）：拟议的研究重点是影响口语（声音到单词）学习的行为和神经因素，特别是成年人在单词识别中学习非母语词汇音调和辅音。我们的行为神经方法将使我们能够追求我们的长期目标，寻求最理想的学习结果介导的行为训练。这项资助申请中计划的十项实验采用的技术包括行为训练、不同学习能力的学习者的神经解剖学特征以及与学习相关的神经生理学（脑血流动力学）变化的测量。具体目标是：1）比较强调训练刺激中的低刺激与高刺激可变性的非母语词汇音调和辅音训练程序的功效，并比较神经生理学反应（通过事件相关功能磁共振成像测量）与两个训练计划相关; 2）调查是否可以通过额外的语音训练来纠正糟糕的词汇学习（非词汇音高和辅音）学习，并检查与这种补救相关的神经生理学变化;和3）表征不同能力的学习者之间的神经解剖学差异。在我们新提出的综合口语习得网络（Integrated Spoken Language Acquisition Network）模型的指导下，我们提出以下假设：1）高刺激可变性的训练是最有效的，这表现在听觉关联皮层的脑反应增加和顶叶的募集; 2）由于缺乏语音表征而导致的单词学习不良，并且可以通过在单词学习之前首先训练以建立这种表征来补救;（3）学习不良者可以通过较小的左侧Heschl's Gyrus（由灰色或白色物质驱动，取决于声学特征）来识别;然而，通过适当的补救训练仍然可以实现成功的学习。通过了解不同训练计划的功效以及它们如何使不同学习概况的个体受益，以及通过识别训练前后不同学习者的神经特征，我们向能够将学习者置于可能是最具成本效益的并且可能导致最理想的学习结果的训练计划迈进了一步。在一个越来越多语言/多文化的世界里，许多人，包括成年人，都有兴趣学习一门外语，尽管他们很难做到这一点。学习一门外语需要我们大脑中的资源，学习的结果是发生变化。我们提出的研究旨在了解成年人学习在单词中使用外国声音时大脑中发生的变化，不同的训练方法对大脑有什么影响，以及为什么某些人可以更成功地学习（例如，他们在训练前后是否有不同的大脑组织）。

DESCRIPTION (provided by applicant): The proposed research focuses on the behavioral and neurologic factors influencing spoken language (sound-to-word) learning, specifically the learning of non-native lexical tones and consonants in word identification in adulthood. Our behavioral-neural approach will enable the pursuit of our long-term goal of seeking the most desirable learning outcome mediated by behavioral training. The ten planned experiments in this grant application employ techniques that include behavioral training, neuroanatomic characterizations of learners of different learning abilities, and measurements of neurophysiologic (cerebral hemodynamic) changes associated with learning. The specific aims are to: 1) Compare the efficacy of non-native lexical tone and consonant training programs that emphasize low- versus high-stimulus variability in the training stimuli, and to compare neurophysiologic responses (measured by event-related fMRI) associated with the two training programs; 2) Investigate whether poor lexical learning can be remediated by additional training that focuses on phonetic (non-lexical pitch and consonant) learning and to examine neurophysiologic changes associated with such remediation; and 3) Characterize neuroanatomic differences between learners of different abilities. Guided by our newly proposed model of spoken language processing and learning called the Integrated Spoken Language Acquisition Network, we hypothesize the following: 1) Training with high-stimulus variability is the most efficacious, which is indicated by increased brain responses in auditory association cortex and the recruitment of the parietal lobe; 2) poor word learning resulting from a lack of phonetic representation and can be remediated by first training to establish such representations before word learning; and 3) poor learners can be identified by a relatively smaller left Heschl's Gyrus (driven by grey or white matter depending on the acoustic feature); however, successful learning can still be achieved with the appropriate remedial training. By understanding the efficacy of different training programs and how they benefit individuals of different learning profiles, as well as by identifying the neural characteristics of the different learners before and after training, we move a step closer to being able to place learners into training programs that are likely to be the most cost- effective and will likely lead to the most desirable learning outcomes. In an increasingly multi-lingual/multi-cultural world, many people, including adults, are interested in learning a foreign language even though it is difficult for them to do so. Learning a foreign language requires resources in our brain and changes occur as a result of learning. Our proposed research seeks to understand the changes that occur in the brain as adults learn to use foreign sounds in words, what effects different training methods have on the brain, and why certain individuals can learn more successfully (e.g., whether they have different brain organization before and after training).