Doctoral Dissertation Research: Behavioral and Computational Investigation of Transitivity in the Acquisition of Non-Basic Syntax

在生命的最初几年里，在没有明确指导的情况下，儿童获得了一种语言系统，允许他们组合声音来表达意思。研究这一过程阐明了像语言这样复杂的认知系统是如何在人类头脑中发展起来的，并有助于澄清在发展中学习的本质。一般来说，学习需要建立在先验知识的基础上：如果孩子们不知道如何数数，他们就无法学习算术。这个项目询问婴儿如何在他们发育的每个阶段利用他们能够理解的有限的语音量，建立他们先前的语言知识，以便更多地了解他们的语言结构。通过提高我们对婴儿在发展的每个阶段如何学习语言的理解，该项目提供了重要的一步，有助于理解为什么以及如何在婴儿期听到的语言的数量和类型对儿童在学校和以后的生活中后来的词汇和语言表现的差异做出贡献。本研究的重点是习得“非基本”从句，以此作为了解语言学习增量性质的窗口。像这样的句子：“玛丽修了什么？”在句法习得的最早阶段，可能会误导儿童，他们可能不会认识到What代表FIX的宾语，并可能认为这个从句是不及物的。这项工作追求的假设是，儿童可以使用关于动词论元结构的知识来识别他们语言中的非基本从句类型。本研究整合了不同研究领域的研究方法，通过行为实验考察非基本从句习得的发展轨迹，并通过计算模型评估所提出的机制是否适用于从儿童导向的言语中进行学习。通过这一案例研究，本研究旨在提供一个更全面的图景，了解在任何一个发展阶段，输入的哪一部分对儿童是有用的，儿童如何在现有语法知识的基础上感知输入，以及在推断语言语法时，哪些内部机制使他们能够在输入之外进行概括。该项目还包括向当地家庭和来自代表不足的背景的高中生推广语言发展的问题。这一奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

Within the first years of life and without explicit instruction, children acquire a linguistic system that allows them to combine sounds in order to express meanings. Studying this process illuminates how a cognitive system as complicated as language can develop in the human mind, and helps clarify the nature of learning in development. Learning in general requires building on prior knowledge: children can't learn arithmetic if they don't know how to count. This project asks how infants build on their prior linguistic knowledge at each stage of their development, using the limited amount of speech they are able to understand, in order to learn more about the structure of their language. By improving our understanding of how infants learn about their language at each stage in development, the project provides an important step toward understanding why and how the quantity and type of speech heard in infancy contribute to differences in children's later vocabulary and language performance, in school and later in life.This research focuses on the acquisition of "non-basic" clauses as a window into the incremental nature of language learning. Sentences like "What did Mary fix?" are potentially misleading for children in the earliest stages of syntax acquisition, who might not recognize that "what" stands for the object of "fix" and might think that this clause is intransitive. This work pursues the hypothesis that children may use knowledge about verb argument structure in order to identify non-basic clause types in their language. The research integrates approaches from different research communities, using behavioral experiments to investigate the developmental trajectory of non-basic clause acquisition, and computational modelling to evaluate whether the proposed mechanisms are feasible for learning from child-directed speech. Through this case study, this research aims to provide a fuller picture of what portion of the input is useful to an individual child at any single point in development, how children perceive that input given their current grammatical knowledge, and what internal mechanisms enable them to generalize beyond that input in inferring the grammar of their language. The project also includes outreach about issues in language development to local families and high school students from under-represented backgrounds.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.