著者
ジェイソン モーザー
出版者
大阪樟蔭女子大学
雑誌
大阪樟蔭女子大学学芸学部論集 (ISSN:18807887)
巻号頁・発行日
vol.47, pp.191-201, 2010-01-29

In this paper I will discuss learner language complexity in oral tasks and how it has been conceptualized in task-based learning research. One of the key tools for measuring complexity in spoken discourse is the AS-unit (Analysis of Speech Unit). The AS-unit is a main clause and any attached subordinate clauses or sub-clausal units. I will first discuss the problems involved in codifying AS-units in data from communicative pair work tasks. I will demonstrate that subordination is not necessarily characteristic of communicative tasks, nor is it easy to identify in conversation, especially with beginner learners. In response I will argue that measuring learner productivity by words per AS-unit is an effective alternative. I will also demonstrate an AS-unit complexity benchmark based on AS-unit word count, in which units above a certain word count are deemed complex. The rationale for this benchmark will be discussed, and supported with examples from four beginner learners' data.