The development of phonetic perception

这个研究项目通过对元音感知的研究来探索婴儿语言感知的发展。我们的目标是全面描述幼儿在对言语的反应中表现出的元音感知偏差，解释这些感知偏差在发展过程中是如何以及何时形成的，并确定它们在口语习得中的作用。为了实现这些目标，我们将测试并阐述一个新的概念模型，即自然指涉元音（NRV）框架。根据这一观点，具有特定发音声学特性的元音（也定义了传统元音空间中的外围元音）是一种天然的指称物：它们通过吸引婴儿的注意力和为学习者建立稳定的参考框架来支持和指导元音感知的发展。先前关于婴儿对不同元音的听力偏好的研究支持了这一观点，表明婴儿（4至12个月大）对周边元音有强烈的感知偏见。研究1将评估这种感知偏差是何时（以及如何）在婴儿早期最初形成的。我们将比较两个月大的婴儿和新生儿，以评估元音感知偏差通过婴儿神经听觉能力和语言暴露之间的相互作用在生命早期出现的假设。研究2和研究3将验证NRV关于周边元音支持元音感知和产生发展的说法。研究2通过检查婴儿对婴儿说话者发出的元音的听力偏好（使用一种新的元音合成工具），测试了元音感知偏差也影响婴儿如何感知自己的声音模式的预测。研究3将评估婴儿识别不同说话者（成人、儿童、婴儿）发出的相同元音的能力，以检验NRV预测的元音分类技能是逐步、分层发展的，从而使外围元音类别更容易获得，并且先于非外围元音类别。这将是第一个调查婴儿如何感知婴儿发出的声音信号的研究。这种方法对于理解说和听技能在婴儿语言发展中的相互作用是必要的。

This research program explores the development of speech perception in infancy through investigations that focus on vowel perception. Our goal is to fully portray vowel perception biases that young infants display in response to speech, explain how and when these perceptual biases form in development, and establish their role in the acquisition of spoken language. In pursuing these goals we will test and elaborate a new conceptual model, the Natural Referent Vowel (NRV) framework. According to this view, vowels with specific articulatory-acoustic properties (which also define the peripheral vowels in a traditional vowel space) act a natural referents: they support and guide the development of vowel perception by attracting infant attention and establishing a stable frame of reference for the learner. Prior studies of infant listening preferences for different vowels support this view by showing that infants (4 to 12 months of age) have a strong perceptual bias favoring peripheral vowels. Study 1 will assess when (and how) this perceptual bias is initially formed in early infancy. We will compare 2-month-olds and newborns to assess the hypothesis that vowel perception biases emerge early in life via an interaction between infant neuro-auditory abilities and speech exposure. Studies 2 and 3 will test the NRV claim that peripheral vowels support the development of vowel perception and production. Study 2 tests the prediction that vowel perception biases also impact how babies perceive their own vocal patterns by examining infant listening preferences for vowels produced by an infant talker (using a new vowel synthesis tool). Study 3 will assess infant ability to recognize the same vowel produced by different talkers (adult, child, infant) to test the NRV prediction that vowel categorization skills develop in a stepwise, hierarchical manner such that peripheral vowel categories are acquired more easily and prior to non-peripheral vowel categories. This will be the first study to investigate how infants perceive infant-produced vocal signals. This approach is needed to understand how speaking and listening skills interact in infant speech development.