From The Comprehensible Output Hypothesis To The Length Approach: The Role Of Output In Language Development

Language development can be measured in three dimensions, namely fluency, accuracy, and complexity. According to Swain's Comprehensible Output Hypothesis, it's quite natural that language output can enhance fluency development, especially when we refer to the proverb "practice makes perfect". Besides, the Output Hypothesis presents three functions of language output which relate more to accuracy development than fluency development. These three functions are: 1) hypothesis testing function, second language learning can be considered as a process of making hypotheses on the target language, testing the hypotheses, and modifying the hypotheses, and producing output is a way of testing a hypothesis about the target language; 2) metalinguistic function, "as learners reflect upon their own target language use, their output serves a metalinguistic function, enabling them to control and internalize linguistic knowledge"; 3) noticing function, output may enhance a language learner to notice the gap between his or her interlanguage and target language. The Length Approach, as a language teaching method theoretically based on the Output Hypothesis, encourages language learners to write longer compositions so as to help them build confidence on language learning and push them to produce more language output, and thus makes the language learners achieve more language development.Many researches (Pica, et al, 1989; Nobuyoshi & Ellis, 1993; Lapkin, Swain, & Smith, 2002; Swain & Lapkin, 1994; 2001 ;2002; etc.) on the three functions of language output and the role of output in language development had been carried out. However, some issues remain: a) fewer studies were done to research the noticing function of language output; b) few researches were based on written output; c) few researches were done on the relationship between output and fluency development; d) most of the researches used small samples.Using written output for analysis, this thesis adopted aimed at answering the following research questions:1. Can language output lead the language learners to notice the gap between their interlanguage and target language?2. Can language output lead to the accuracy development on the linguistic form?3. Can language output, i.e., the application of the Length Approach, promote the learners' language development in terms of fluency and accuracy? In order to answer the research questions, the author of this thesis carried out an experiment and an investigation.In order to answer the first two questions, this thesis partially duplicated the experiment of Izumi, et al (1999). Two groups of students majoring on science and technology participated in the experiment, and pretest showed that they were not statistically different in the command of the target language form (past hypothetical conditional). The experiment consisted of two treatment phases. In treatment phase 1, participants read a short passage, 70% of whose sentences contained the target language form. Then the EG was required to reconstructed it while the CG answer comprehension questions. After that, the EG was given a second exposure to the same input material and a second reconstruction opportunity. In phase 2, participants of the EG wrote on given topics which called for the use of target language form, followed by the presentation of a model essay. The EG then wrote a second time on the same topic. The CG wrote on different topics which did not call for the use of target form, and the model essay was provided to them as reading exercises. To test the noticing function of output, participants were asked to underline morphemes, words, or phrases which they thought were "particularly necessary" for subsequent (re)production or reading comprehension tasks. In order to test whether output can enhance the development of accuracy, three tests were held before the experiment, after treatment phase 1, and after treatment 2 respectively.In order to answer the third research question, this thesis investigated two groups of sophomore students, one of which was instructed by the Length Approach for nearly three terms while the other not. A questionnaire survey showed that, students instructed by the Length Approach did not increase their confidences in English learning, which excluded the role of "affect" in language development. A same writing task was assigned to the two groups. Their written output was analyzed in terms of fluency, accuracy, and complexity. The analyzing tools were T-unit based language developmental measures (Wolf-Quintero, et al, 1998: 119-121).Results of the experiment and investigation showed that:1. In the experiment, language output did not necessarily lead to the notice of the target language form.2. The EG only significantly improved their accuracy in the use of the target form in treatment phase 2. 3. Students, who were instructed by the Length Approach and pushed to produce more language output, showed higher accuracy developmental level than the students that were not instructed by the Length Approach, but they did not achieve higher developmental level in terms of fluency and complexity than the students who were not instructed by the Length Approach.This thesis partially proved the Output Hypothesis, because the study only proved output's role in accuracy development as was predicted by the Comprehensible Output Hypothesis, but the noticing function of language output and output's role in fluency development were not verified. Besides, this thesis found that the Length Approach could effectively promote the learner' accuracy development, which proved that the Length Approach was an effective teaching method in the sense of enhancing students' language accuracy development.