ONTOGENY of SEGMENTAL SPEECH ORGANIZATION

描述（由申请人提供）：该项目的长期目标是建立一个理论解释，说明在声信号中没有明确标志的情况下，儿童是如何获得分段语音结构的。所采取的方法假设成年听者通过应用语言特定策略来感知组织语言相关的声学特性来访问该结构。具体来说，所探索的模型表明，儿童最初根据通用策略将声学语音信号组织为一个频谱统一的整体，然后，通过早期的语言经验，学会以特定于语言的方式组织语音信号的动态成分，并在语音决策中权衡某些非动态的、频谱时间属性，这些属性在母语中具有语音信息。反过来，假设这些学习策略中的一种或两种有助于片段结构的恢复。从该模型衍生出的六个具体目标将检验以下假设：儿童最初非常重视动态信号成分（即峰运动），因为这些成分通常符合听觉场景分析的原则；2. 成熟言语知觉的习得涉及动态信号特性的知觉组织的变化，这一过程被称为发展性知觉重组；3. 幼儿比成年人更有强烈的义务将类似言语的信号组织成频谱统一的信号；4. 静态特性在恢复节段结构方面优于动态特性；5. 言语感知和对片段结构的敏感性的发展变化是由于早期（即上学年龄之前）使用母语的经验引起的，而不是其他因素（即一般听觉发展或识字教学）；和6。当儿童获得母语的经验时，其语言生成的手势组织会发生变化。收集到的数据应该促进我们对听者如何从声学信号中恢复片段结构的理解。这项工作的临床意义在于帮助我们了解我们应该做些什么来促进有沟通问题的儿童的口语发展。

DESCRIPTION (provided by applicant): The long-term goal of this project is to develop a theoretical account of how children acquire access to segmental speech structure given that there are no clear landmarks for that structure in the acoustic signal. The approach taken assumes that adult listeners access that structure by applying language-specific strategies for perceptual organization to linguistically relevant acoustic properties. Specifically the model explored suggests that children initially organize the acoustic speech signal as a spectrally unitary whole according to universal strategies, and then, through early language experience, learn to organize the dynamic components of the speech signal in language-specific ways, and to weight in phonetic decisions certain non-dynamic, spectro-temporal properties that are phonetically informative in the native language. In turn, it is hypothesized that one or both of these learned strategies facilitates recovery of segmental structure. Six specific aims derived from this model wilt test the hypotheses that: 1. children initially weight dynamic signal components (i.e., formant movements) greatly because these components generally adhere to principles of Auditory Scene Analysis; 2. the acquisition of mature speech perception involves changes in the perceptual organization of dynamic signal properties, a process termed Developmental Perceptual Reorganization; 3. young children are more strongly obliged than adults to organize speech-like signals as spectrally unitary signals; 4. static properties provide an advantage over dynamic properties in the recovery of segmental structure; 5. developmental changes in speech perception and in sensitivity to segmental structure arise due to early experience (i.e., before school age) with a native language, rather than from other factors (i.e., general auditory development or literacy instruction); and 6. children's gestural organization for speech production changes as they acquire experience with their native language. The data collected should advance our understanding generally of how listeners recover segmental structure from the acoustic signal. Clinical significance will derive from this work by helping us understand what we should be doing to promote spoken language development in children with communication problems.