NEUROBIOLOGY OF LANGUAGE PERCEPTION--FMRI INVESTIGATIONS

这个项目的长期目标是绘制听觉语言感知的神经生物学，包括语言子成分和领域特异性。 功能性磁共振成像（fMRI）--一种非侵入性、高分辨率的脑功能成像技术--的出现使这一目标的实现在不久的将来成为真实的可能。 由于fMR成像的最佳方法程序尚未制定出令人满意的，我们提出了一个相对较小的一组理论动机的实验，我们将进行使用协议，可以告知方法以及理论问题。 关于语言的神经组织。 1.在句子水平上绘制参与处理听觉语言的神经组织的范围--这是一个在认知和语言学理论中非常突出的语言分析水平，但尚未使用功能成像技术进行彻底的研究。 2.确定语言与非语言听觉感知相关神经系统的特异性。 我们将句子的感知与一系列非语言听觉刺激的感知进行对比，其中每个刺激控制句子处理的特定方面，例如时间声学处理（向后句子），层次结构处理（音乐），处理有意义的声音串（环境声音“事件”）。 我们还研究了听觉处理中的专业知识的作用，以及将单个刺激作为语言与非语言信号进行听觉的神经相关性。 3.研究语言信号的不同方面（韵律、词汇语义内容、句法）在不同神经系统中的处理程度。 关于方法论的目标。 1.调查信度问题：重测信度，本地化的信度，和可靠性方面看到的重复效果。 2.为了实证地检验神经影像学中任务效应的认知基础，即回答这个问题，任务操纵能否为语言感知的神经基础提供一个窗口，或者我们只是测量特定任务的效应。 我们提出了一套选择性注意实验（广泛使用的任务操纵）来调查这个问题。 3. 我们的第三个目标是探索单一试验演示格式的实用性。 我们先前的工作已经证明，人们可以在单一试验中呈现刺激（而不是这种刺激的块），并获得可靠的激活，也许具有更高的灵敏度。这从根本上增加了fMRI实验设计的灵活性。 我们在几个实验中评估该技术的有效性。

The long range goal of this project is to map the neurobiology of auditory language perception, both in terms of its linguistic sub-components and in terms of its domain specificity. The advent of functional magnetic resonance imaging (fMRI)--a non- invasive, high resolution functional brain imaging technique--has made the realization of this goal a real possibility in the near future. Because optimal methodological procedures for fMR imaging have yet to be worked out satisfactorily we propose a relatively small set of theoretically motivated experiments that we will carry out using protocols that can inform methodological as well as theoretical questions. Aim Concerning the Neural Organization of Language. 1. To map the extent of neural tissue involved in processing auditory language at the sentence level--a level of linguistic analysis which has been prominent in cognitive and linguistic theories but has yet to be investigated thoroughly using functional imaging techniques. 2. To determine the specificity of neural systems involved in linguistic vs. non- linguistic auditory perception. We contrast perception of sentences with perception of a range of non-linguistic auditory stimuli each of which controls for specific aspects of sentence processing such as temporal acoustic processing (backwards sentences), hierarchical structure processing (music), processing meaningful strings of sounds (environmental sound "events"). We also investigate the role of expertise in auditory processing and the neural correlates of hearing a single stimulus as linguistic versus non-linguistic signal. 3. To investigate the extent to which different aspects of the linguistic signal (prosody, lexical-semantic content, syntax) are processed in distinct neural systems. Aims Concerning Methodology. 1. To investigate reliability issues: test-retest reliability, reliability of localization, and reliability as seen in terms of repetition effects. 2. To examine empirically the cognitive basis of task effects in neuroimaging, namely, to answer the question, can task manipulations provide a window on the neural basis of linguistic perception or are we simply measuring task-specific effects. We propose a set of selective attention experiments (a widely used task manipulation) to investigate the issue. 3. Our third goal is to explore the utility of the single trial presentation format. Our prior work has demonstrated that one can present stimuli in single trials (rather than blocks of such stimuli) and obtain reliable activations, perhaps with increased sensitivity. This radically increases the flexibility of experimental design in fMRI. We assess the validity of the technique in several experiments.