This project seeks to understand how children acquire the grammar of their first language. Theories of language acquisition are developed by considering the structure of adult languages and the linguistic input available to children, formulating processes that are capable of discovering languages from input, and testing to see if children use such processes. Techniques include analyzing children's spontaneous speech, and testing them in experiments in which they act out sentences, describe pictures, or judge the acceptability of other speakers' utterances. This phase of the project focuses on the acquisition of inflection (e.g. walk/walked), and important component of human languages. This is a central topic in research on the nature of the mind, specifically, whether it is like a computer processing rules (e.g. "to form the past, add -ed to a word"), like a network of associations (where mental units standing for sounds of the word, such as 'wa' and 'ak', are associated through practice with units standing for sounds of the past form, such as 'akd'). The project seeks to answer empirical questions relevant to this debate: Why do children use correct irregular forms like came before they start to make errors such as comed- is because the early correct forms are learned by rote and the errors are to due to rule use? What causes children to make mistakes like comed - is due to a change in the relative amounts of evidence available to the child for different ways of forming the past tense, to the mastery of the semantic distinction between present and past, to age, or to overall linguistic development? What is the nature of the mental operation that children use to create past tense forms - associations from the sounds of present tense verbs to sounds of past tense verbs, based on analogies to similar-sounding forms encountered beforehand, or a rule that is sensitive to the word's grammatical structure? What makes children stop making errors like comed: frequent exposure to correct forms like came, matching up came and come as two versions of the same word, realizing that came means 'come in the past', or parental corrections? The answers will help our understanding of how the brain processes complex rule-governed information. It could also lead to a precise characterization of normal children's abilities to learn inflection which can then be applied to syndromes of language disorders, helping in the diagnosis, characterization, and remediation of such syndromes. In addition, understanding how learning succeeds in children may help in the design of second language teaching curricula.