A Study Of Complexity, Accuracy And Fluency In English Learners’ Oral English Development In Light Of Complexity Theory

This paper reports a longitudinal study of complexity, accuracy and fluency(CAF) in oral English development in light of Complexity Theory. In the current study, individual learners’ oral English development is taken as a complex system. Complexity, accuracy and fluency(CAF), the three dynamic measures of oral English development, are considered as the central elements of this system. All factors influencing the initial conditions and development trajectories of CAF are considered as the environment of this system.By observing learners’ short-term and long-term CAF trajectories, this study first aims to find out the general trend of learners’ oral English development and major dynamic patterns of trajectories. By doing correlations and plotting scores geometrically, the study then focuses on CAF relations and dynamic patterns of CAF relations. Finally, using environmental factors, the paper attempts to explain individual differences in CAF initial level and CAF development trajectories.Six sophomores of English major(3 males and 3 females) participated in this longitudinal study. Each of them performed a topic-given narration task(different topics with equal difficulty) once a week over 15 weeks, and repeated the task of Week 1 after 6 and 12 months. Various scales, questionnaires and tests were employed to investigate “environmental factors”. The principles of CHAT were observed in transcribing the oral data and CLAN was used in doing analysis work. Major findings and contributions are as below.First of all, this paper presents 96 charts of short-term trajectories and 16 charts of long-term trajectories, from which three major trends of oral English development are drawn. i. Most participants made progress in complexity and accuracy, and experienced descending trajectories in fluency, but the decrease of fluency in long-term trajectories was not as pronounced as that in short-term trajectories. ii. For most CAF variables, their short-term trajectories carried the same trend with their long-term trajectories. iii. Syntactic complexity, lexical diversity, grammatical accuracy, pronunciation and intonation accuracy mainly took a growing trend, lexical density experienced little change, while positive measures of fluency F1(speed measure) and F2(MLU) mainly descended, and negative measures of fluency F5(self-repairs) and F6(total of repetitions, self-repairs and fase starts) mainly ascended.Five dynamic patterns of CAF trajectories were identified: “peaks/valleys”, “ascending in curve”, “descending in curve”, “quasi-horizontal line”, and “mixed mode”. These patterns reflected the two types of change in complex systems: stability and variability. The first four patterns only showed stability, and the last pattern, also the most common pattern, showed both stability and variability. There were individual differences in dynamic patterns. For instance, participant S4 never expericened the pattern of “peaks/valleys”, while participant S2 never had “quasi-horizontal line”. For some CAF variables, some dynamic patterns were more common than others. For example, four out of six learners had the pattern of “descending in curve” in variable F1.Secondly, the relationship between complexity, accuracy and fluency can not be overgeneralized into “compete” or “grow together” when multidimensional measures of CAF are involved. i. This study added pronunciation and intonation accuracy to the existing accuracy dimension, and found that grammatical accuracy negatively correlated with fluency, while pronunciation and intonation accuracy positively correlated with fluency. ii. The connection between complexity and accuracy was not applicable to every aspect of complexity or accuracy. Results show that grammatical complexity was related to grammatical accuracy, but lexical complexity was not found to be connected with lexical accuracy. iii. Besides, grammatical complexity and lexical complexity were related with different aspects of fluency. Results show that morphological complexity correlated with repairs while lexical diversity correlated with repetitions.In 15-week short-term trajectories, the relation of correlated variables was found to have three patterns: no variability, irregular variability and regular variability. In long-term trajectories, the early competition of fluency with complexity and accuracy tended to disappear with the development of participants’ spoken English.Finally, environmental factors could explain individual differences in CAF intial level and trajectories. For instance, high level of risk taking might explain learner’s less frequent repetition or repair. Factors that could explain the trajectories of CAF variables included initial conditions, CAF relations, state anxiety, weekly engagement and attitude, unconscious language input and special language experiences. CAF initial levels might seriously explain the development trajectories of CAF measures. Lower initial level tends to incur progress and higher initial level tends to cause decrease.