Morphological acquisition in early infancy: Experimental and computational studies

形态后缀，如英语-S，-ed和-ing是语法的基本组成部分。限制使用形态后缀是患有SLI和自闭症等发育障碍儿童的语音特征。因此，为了促进典型和非典型人群的语言学习，我们有必要了解婴儿是如何以及何时获得形态的。这个项目主要研究英语和西班牙语中形态习得的起源。通过整合对6个月大的婴儿进行的实验、语料库分析和最先进的计算机建模，该项目试图确定婴儿何时以及如何开始获得形态。该项目还将重点放在西班牙语上，这是一种收购时间表远不如英语的语言，该项目还将增加参与研究活动的机会。将这个项目的行为和计算结果结合起来，将有助于揭示形态习得的表征、学习时间和机制。具体地说，婴儿实验将为形态习得背后的语言表征提供证据，并记录形态词缀习得的发展轨迹。下一步，使用计算模型，将对形态学习者进行逆向工程；其性能将根据婴儿实验中发现的形态习得的发展轨迹进行评估。计算模型随后将被用来精确定位最有可能区分不同理论的婴儿实验；这一点至关重要，特别是因为婴儿研究是时间和资源密集型的。最后，作为该项目的一部分，我们预计将生成一个大型的、带注释的婴儿语料库，该语料库将公开提供给调查语言习得的研究人员。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。

Morphological suffixes like English -s, -ed and -ing are basic building blocks of grammar. Limited use of morphological suffixes characterizes the speech of children with developmental disorders like SLI and autism. To promote language learning in typical and atypical populations, it is therefore necessary for us to understand how and when infants acquire morphology. This project is focused on investigating the beginnings of morphological acquisition in English and Spanish. By integrating experiments with infants as young as 6-months, corpus analysis and state-of-the-art computational modeling, this project seeks to establish when and how infants begin to acquire morphology. By additionally focusing on Spanish, a language in which the acquisition timetable is far less known than in English, this project will also increase access to participation in research activities. Integrating the behavioral and computational results from this project will shed light on the representations, time course of learning as well as mechanisms involved in the acquisition of morphology. Specifically, infant experiments will provide evidence for linguistic representations underlying morphological acquisition and document the developmental trajectory for the acquisition of morphological affixes. Next, using computational modeling, a morphological learner will be reverse engineered; its performance to be evaluate against the developmental trajectory of morphological acquisition uncovered in the infant experiments. Computational modeling will subsequently be used to pinpoint infant experiments that are most likely to distinguish between alternate theories; this is crucial particularly because infant research is time and resource intensive. Finally, as part of the project we anticipate generating a large, annotated corpus of infant-directed speech which will be publicly available for researchers investigating language acquisition.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.