The effect of task complexity and task conditions on foreign language development and performance. Three empirical studies

It has been argued that tasks constitute a valid alternative unit to sequence the language learning process, as opposed to linguistically defined syllabuses. Implementing this claim presupposes that it is possible to access the cognitive and linguistic demands of tasks, so that they can be sequenced in such a way that they optimally support and promote the L2 learning process. Knowing what demands a task will make opens up the possibility of using task design to manipulate the learner's attention between form and meaning in ways that may help interlanguage development. In this article, we present three empirical studies, which have tried to manipulate task complexity in order to study the effect of differing levels of task complexity on (L2) performance. We situate our studies within the Triadic Componential Framework for Task Design (TCFTD), elaborated by Robinson (1995; 2001; 2005) and interpret our findings in the light of two alternative theories, trying to explain effects on L2 performance arising from task manipulation, namely the Limited Attentional Capacity Model (Skehan & Foster 2001), and the Cognition Hypothesis (Robinson 2001; 2005). Apart from yielding evidence against or in favour of these theories, our studies demonstrate that manipulating L2 learners' attention while performing a task is anything but straightforward. The studies also illustrate how task conditions appear to interact with task complexity.