A Minimalist Account of Verbal Morphology in Child Grammar: Evidence from VP Ellipsis

Rosalind Thornton Proposal NumersSBR-9905199《儿童语法中动词形态的最简解释：来自VP省略的证据》根据最简方案，只有当需要推导出可被概念系统解释的表示(或对语音/发音部分来说是可读的表示)时，移动才会在句法/计算系统内发生。这样的接口条件使得一个句子不太可能产生一个以上的收敛派生。相比之下，对儿童早期语言的实证研究得出这样的观点：儿童早期语法的特点是可选的结构派生。这种观点与极简主义纲领的精神相悖。一个顽固的观察是，两岁半以下的儿童似乎在动词时态派生(如果语言需要的话，可以进行动词移位)和动词非限定形式(无限形式)之间摇摆不定。本项目试图在语言学理论和儿童语言研究之间找到更好的契合点。研究策略包括通过对几个年幼的儿童进行长期仔细的纵向实验，寻找儿童对语言差异的敏感性，这些差异在之前的研究中没有被注意到或没有得到解释。在早期儿童语法中，动词形态的明显可选性的一个核心假设是阶段级谓词和个体级谓词之间的语义区别；个体级谓词以动词形态出现，作为限定形式，而阶段级谓词以非限定形式(或不定式形式)出现。有人认为，普遍语法在早期儿童语言中的“纯粹”反映被掩盖了，因为特定语言的过程，如动词移位，被纳入了儿童的语法中。另一个需要研究的假设是，即使在所谓的可选不定式阶段，动词形态的表观可选性也会被句法过程所覆盖。VP省略是一个句法过程，其结果是必须对时态和可能的一致进行形态标记，只有在助动词所带的“强一致”的情况下才有可能。儿童在VP省略结构中必须使用时态和协议，这将对当前的许多理论提出挑战，这些理论认为可选性是语法发展的某种缺陷或生物诱导的限制的产物。本研究的积极发现也可以为寻找触发经验提供信息，这些经验可以指导儿童为他们的目标语言选择正确的参数设置。这些和其他关于VP省略和谓词类型的假设将在实验中使用诱发产生的方法进行验证，除了来自自发产生的数据和来自Childes数据库的数据。

AbstractRosalind ThorntonProposal Number SBR-9905199"A Minimalist Account of Verbal Morphology in Child Grammar: Evidence from VP Ellipsis"According to the Minimalist Program, movement takes place within the syntactic/computational system only when needed to derive a representation that can be interpreted by the conceptual system (or one that is 'legible' to the phonetic/articulatory component). Interface conditions such as these make it highly unlikely for a sentence to yield more than a single convergent derivation. By contrast, empirical investigation of early child language has led to the view that children's early grammars are characterized by optional structural derivations. This view is at odds with the spirit of the Minimalist Program. One recalcitrant observation is that children younger than two and a half years old appear to vacillate between a derivation in which the verb is tensed (undergoing verb movement if the language requires it) and one in which the verb is a nonfinite (infinitival) form.The present project seeks a better fit between linguistic theory and investigations of child language. The research strategy involves searching for children's sensitivity to linguistic distinctions that have gone unnoticed or unexplained in previous investigations, using careful longitudinal experimentation over an extended period with a few young children. One distinction that is hypothesized to lie at the heart of the apparent optionality of verbal morphology in early child grammars is the semantic distinction between stage-level and individual-level predicates; individual-level predicates appear with verbal morphology, as finite forms, and stage-level predicates appear as non-finite (or infinitival forms). The suggestion is that this "pure" reflection of Universal Grammar in early child language is obscured, as language-specific processes such as verb movement are incorporated into the childs grammar. Another hypothesis to be investigated is that the apparent optionality of verbal morphology is overridden by syntactic processes, even during the so-called optional infinitive stage. One syntactic process that results in obligatory morphological marking of tense and possibly agreement is VP ellipsis, which is licensed only by "strong agreement" carried by an auxiliary verb. Childrens obligatory use of tense and agreement in VP ellipsis structures would challenge many current theoretical proposals which view optionality as the product of some sort of deficit or biologically-induced constraint of developing grammars. Positive findings from the present research could also inform the search for the triggering experiences that guide children in selecting the correct parameter settings for their target language. These and other hypotheses concerning VP ellipsis and type of predication will be tested in experiments using the methodology of elicited production, in addition to data from spontaneous production and from the CHILDES database.