Collaborative research: An integrated theory of syntactic acquisition - Realistic input, quantitatively defined target states, and computational models of the learning strategy

协作研究：句法习得的综合理论 - 现实输入、定量定义的目标状态和学习策略的计算模型

• 批准号: 1347115
• 负责人: Jon Sprouse
• 金额: $ 23.11万
• 依托单位: University of Connecticut
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2014
• 资助国家: 美国
• 起止时间: 2014-07-01 至 2017-12-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1347115&HistoricalAwards=false
• 关键词: Collaborative; research; integrated; theory; syntactic

Understanding how the human mind creates the intricate system of syntactic structure that underlies human language has long been a central question in cognitive science. A complete theory of the cognitive architecture scaffolding this complex process needs to specify at least three components: (i) the input children use to learn their language's syntax, (ii) the gradient adult syntactic behavior that serves as the target state for learning, and (iii) the learning mechanism that children use to map between the input and the target state. The field is closer than ever to constructing its first set of large-scale theories because both the computational and experimental methodologies necessary to specify these components are now available to researchers willing to combine their expertise.This project leverages a research framework that integrates realistic data about children's input, experimental syntax techniques, and computational modeling approaches to explain how children successfully acquire knowledge of multiple syntactic phenomena. The goal is to integrate the successful learning strategies for several complex syntactic phenomena into a cohesive theory of syntactic acquisition, identifying what the necessary elements of each strategy are and how children would know to utilize these strategy elements in just the right way. This represents a major advance for theoretical linguistics, developmental linguistics, and the broader psycholinguistics community, as it connects theoretical and experimental work in linguistics with computational modeling to yield broader insights than ever before.This project provides an empirical foundation for studies of syntactic acquisition and a template for future projects that involve a coordinated synthesis of theory, experimentation, and computation. In addition, it provides hands-on research experience in theoretical linguistics to multiple undergraduate students and a graduate student each year, giving them access to professors engaged in highly relevant research that promises to have a significant impact on the field.

长期以来，认知科学的核心问题一直是理解人类思维如何创造出构成人类语言基础的复杂句法结构系统。关于这一复杂过程的认知结构支架的完整理论需要明确至少三个组成部分：(i)儿童用于学习语言语法的输入，（ii）作为学习目标状态的梯度成人句法行为，以及（iii）儿童用于在输入和目标状态之间映射的学习机制。这个领域比以往任何时候都更接近于构建它的第一套大规模理论，因为明确这些组成部分所需的计算和实验方法现在都可供愿意结合他们的专业知识的研究人员使用。该项目利用一个研究框架，将儿童输入的实际数据、实验语法技术和计算建模方法结合起来，解释儿童如何成功地获得多种语法现象的知识。目的是将几种复杂句法现象的成功学习策略整合到句法习得的内聚理论中，确定每种策略的必要元素是什么，以及儿童如何知道以正确的方式使用这些策略元素。这代表了理论语言学、发展语言学和更广泛的心理语言学社区的重大进步，因为它将语言学的理论和实验工作与计算建模联系起来，产生了比以往任何时候都更广泛的见解。该项目为句法习得的研究提供了经验基础，并为未来涉及理论、实验和计算协调综合的项目提供了模板。此外，它每年为多名本科生和一名研究生提供理论语言学的实践研究经验，使他们能够接触到从事高度相关研究的教授，这些研究有望对该领域产生重大影响。