Linguistic variation and micro-cues in first language acquisition

Children are often exposed to considerable variation in the input. Nevertheless, there is very little overgeneralization in child language data and children are typically found to make errors of omission, not errors of commission, a fact which is often referred to as conservative learning. In this paper, these findings are accounted for by a model of micro-cues, a generative approach to language acquisition arguing that children are sensitive to fine syntactic distinctions from early on. The micro-cues are small pieces of abstract syntactic structure resulting from parsing the input. This means that UG provides children with principles, features, and the ability to parse, but not the micro-cues themselves, which are considered to be part of the knowledge of a specific language. The model also considers children's errors to generally be due to economy and the language acquisition process to be development in small steps, from specific to more general knowledge.