Task complexity and grammatical development in ESL

Abstract

This study investigates whether second language learners' interlanguage (IL) systems change according to the tasks they perform. This is a long-debated issue in the fields of SLA and language learning pedagogy, as it may have implications for language assessment and syllabus design. Pienemann's (1998) Steadiness Hypothesis states that the basic nature of an IL system does not change across different communication tasks, provided they involve the same skill type. Pienemann claims that it is the learner's L2 developmental stage rather than the nature of different tasks, which influences linguistic competence. Tarone (1985, 1988, 2014) and Bayley and Tarone (2012), on the other hand, claim that IL is systematically and predictably variable across tasks as a result of style shifting due to shifts in social and contextual variables such as the topic of the interaction. One important element missing in the debate is the definition of 'different tasks'. In recent years, however, Robinson's Cognition Hypothesis (2003) has provided some clear criteria for classifying tasks according to their cognitive complexity. The present study, therefore, tests the two competing positions on IL variability by applying the Cognition Hypothesis for task evaluation. The main research question in the study is whether learners' IL systems vary with tasks of different degrees of cognitive complexity. In order to answer the question, tasks were designed by manipulating the task complexity variable of ± few elements, ± here-and-now and ± planning time. Tasks used in studies which apply Pienemann's Processability Theory (PT, 1998) are used in this research. In this study, 30 adult Chinese L1-English L2 learners in Australia were recruited based on their IELTS scores: 10 were from IELTS band 7.0 or above; 10 were from IELTS bands 5.0"" 5.5 and 10 were from IELTS band 4.5. First, the issue of competence in relation to tasks was approached by assessing the competence of learners by using traditional PT profiling tasks, such as 'spot the difference' tasks. The second step was to use Robinson's (2007) cognitive complexity criteria to assess learners in each group while they performed a set of picture description tasks. Each learner's performance was measured in terms of its accuracy and syntactic complexity based on the learner's PT developmental stage to check whether the IL system across tasks was invariant. The results of this experimental study showed that each learner was quite stable across tasks in terms of morphological and syntactic complexity. The results of the accuracy analysis showed some, but not significant, differences between variables. This suggests that a learner's IL system remains steady across tasks and within tasks of different degrees of cognitive complexity. The results of this study thus support Pienemann's Steadiness Hypothesis (1998).

Date of Award	2017
Original language	English