Can native-speaker corpora help explain L2 acquisition of tense and aspect?

This study investigates the distributional patterns of verb form frequencies in the Spanish past tense (the Preterit and the Imperfect) in the Corpus del Español (Davies, 2002). Following a usage-based approach, we examine the potential influence of input frequency as a driving factor in L2 learning. Results provide support for the existence of a distributional bias in L1 Spanish, with several telic predicates more often occurring in the Preterit and several atelic predicates in the Imperfect. Results also demonstrate that the distribution of verbs in the Preterit and the Imperfect is Zipfian, with the most frequent verbs overall accounting for the majority of all the tokens. Finally, an analysis of the different meanings of the Imperfect demonstrates clear differences in frequency of use with the continuous meaning as the most frequently expressed meaning, followed by habituality and progressivity. Implications for research and teaching are discussed.