Syntactic Complexity In Different Task Types

Over the past few decades, many researches have dwelt upon the phenomena and characteristics of syntactic complexity demonstrated in the second language (L2) written and oral production from various perspectives. At present, syntactic complexity witnesses more researches in written language than in oral performance, and studies on oral production have different focuses, so there remains much space for further researches. The present thesis, based on the classification of L2 development measures into three categories:fluency, accuracy, and complexity (both grammatical and lexical) proposed by Wolfe-Quintero et al., has adopted descriptive analysis, independent samples T-test, one-way ANOVA, multiple comparison methods and nonparametric tests to analyze the syntactic complexity features displayed in the three task types (Task 1:Story Retelling; Task 2:Impromptu Talk; Task 3: Role-play) of the 2009 Graded Test for English Majors-Band 4—Oral Test (TEM-4 Oral Test) by some English majors from one key university of science and engineering in Southwest China with a quantitative approach. The current study has altogether employed nine syntactic complexity indices covering three aspects:unit length, clausal density, and sentence patterns. It attempts to answer the following two research questions:(1) What are the differences of syntactic complexity features demonstrated by learners of high- and low-proficiency groups? (2) What are the syntactic complexity features demonstrated in different task types?The major findings of the research are:(1) In most cases, high-proficiency students outperform low-proficiency students in the three aspects, namely, unit length, clausal density, and sentence patterns, but some syntactic complexity indices show no statistically significant differences. The results above reveal, that high-proficiency learners have a larger size of vocabulary and possess their own proficiency advantages; and that the subjects, due to the limited cognitive and attentional resources and the consideration of "playing it safe", are likely to use too many dependent clauses and avoid using some structures such as reduced structures in the real TEM-4 Oral Test testing environment. It is also proved that the frequency of passive forms is very low in oral production. It is supposed that this phenomenon results from first language transfer.(2) In the aspects of unit length and clausal density, both high- and low- proficiency groups observe a gradual increasing tendency from Task 1 to Task 3 with relation to all the indices, except that clause length (W/C) sees a fall from Task 1 to Task 2, and a rise from Task 2 to Task3, and as far as all the nine syntactic complexity indices are concerned, not all the pairs of two tasks display significant differences; in the aspect of sentence patterns, both high- and low- proficiency groups observe a progressively decreasing trend from Task 1 to Task 3 with respect to almost all the indices, except that DC (dependent clause) witnesses an increase from Task 1 to Task 2, and a drop from Task 2 to Task3, and still, in terms of all the nine indices, not all the pairs of two tasks display significant differences. The following might account for these phenomena:There are originally a large number of simple sentences, three reduced structures (R), and two sentences using passive voice (P) in the audio scripts of the story to be retold in Task 1, while much fewer dependent clauses (DC) are noted in this task; time for planning exerts positive effects upon the performances of both proficiency groups, that is, the more time for planning, the more complex the syntax will be, though under real testing conditions, the testees may avoid the use of some more advanced structures; different tasks demand different degrees of complexity, so that test takers in both groups resort to more complex sentences (such as dependent clauses) to narrate their experiences in Task 2 which is thought to be the most difficult and complex task.The current thesis has enriched the present studies available on oral data and has widened the range of researches on syntactic complexity. Meanwhile, it shows positive significance in oral English teaching for English majors.The findings listed above indicate the following pedagogical implications for oral English teaching:It is advisable that teachers of English involve the practices of condensed structures, such as reduced structures (R) and passive voice in non-finite clauses, in oral (and also written) English teaching, by starting from some simple narrative tasks closely related to daily experiences; due to the possibility that students avoid the use of some advanced structures, teachers are supposed to help students overcome the negative effects of affective variables in order that they win more cognitive resources; it is also suggested that teachers, in different periods of time, attract students'attention to different aspects of oral language through adjusting difficulty of task-based approaches according to different teaching objectives; level-based English teaching is also recommended to be borrowed into teaching English majors in order to cater for different needs of students at different proficiency levels and thus to avoid the waste of intellectual resources.