| The central goal of this dissertation is to introduce and provide evidence for the Hypothesis of Context-Driven Learning, which states that linguistic forms and functions are first acquired in highly specific contexts before their usage becomes generalized. Support for this hypothesis comes from one comprehension and three production experiments with children aged 2--6. The experiments vary with regard to the complexity of the discourse context in which aspectual forms are embedded. Their analysis leads to four major conclusions.;First, Russian aspect is not an innate category or a category that is readily available from the beginning of language acquisition. The acquisition of this category, with its complex morphology, semantics and pragmatics, is a long process that is not completed even by age 6.;Second, aspect acquisition is directly influenced by a multitude of factors: Aktionsarten (lexical temporal specifications of verbs), morphology, discourse complexity, and narrative competence. For example, the developmental pattern is different for telic and ingressive Aktionsarten, for synthetic and analytic morphology, for isolated and concatenated utterances; and the development of aspect competence is tied to the development of narrative competence.;Third, the distribution of aspectual forms within each level of discourse complexity is approximately the same across age groups, but differs widely across these discourse levels. Thus, children are sensitive to adult-like contextual frequency distributions from early on. However, younger children do not yet master the full range of canonical functions of these forms. For example, the backgrounding function of the imperfective emerges relatively late, i.e., only when children's narrative competence has developed sufficiently.;Fourth, there is a stage in which children use aspect in a context-dependent way, without making generalizations across contexts. In some contexts their form/function mapping corresponds to that of the target language, while in other contexts it does not.;Together, these four results suggest that there are three stages involved in the acquisition of Russian aspect, and perhaps in acquisition in general. Stage 1 is tied to individual verb meanings, here Aktionsarten. Stage 2 is characterized by context driven learning. In Stage 3, the target stage, usage is no longer tied to specific contexts. |
