The exploration of the role of naturally-occurring data in studies of second language acquisition begins by assessing the ways in which the term “naturalistic” may apply to second languages. Instead of classifying tasks and the data they elicit as natural or unnatural, this paper presents a view of tasks that is multidimensional, rather than oppositional, and highlights some of the features of tasks that appear to pose significant differences in the data they elicit. Following this review, three cases from the study of the second language acquisition of Spanish are presented. Each case demonstrates how data differ across tasks and builds support for the position that there is no single ideal method for eliciting data. Instead, these cases show that multiple means of elicitation are necessary to fully understand the acquisition of a given construct, with each elicitation task providing unique and essential information about learner language.
© 2001-2026 Fundación Dialnet · Todos los derechos reservados