| Because of its universality, unique syntactic properties and its frequency in our daily use of language, the relative clause has received considerable attention in the linguistic research field. Research on its acquisition difficulty mainly focuses on two aspects: relative clause type and matrix position. The former refers to the function the head noun serves in the sub-clause, including subject, direct object, indirect object, object of preposition, genitive and object of comparison. The latter refers to the position where a relative clause embeds, including, matrix subject position and matrix object position. Although the acquisition difficulty of relative clauses has been widely investigated and explained from different perspectives abroad, few satisfactory explanation has been got. Firstly, processing language comprehension and production involves so many various constraints that it cannot be satisfactorily explained from one angle leaving others untouched. Secondly, only three or four common relative sentence patterns are investigated. Thirdly, the learners' relative clause proficiency levels have seldom been taken into account.Based on the previous findings, the present study attempts to make a complementary investigation into this issue by doing research on the Chinese EFL learners at two English relative clause proficiency levels. A grammaticality judgment task is employed to preliminarily test participants English relative clause proficiency level, and a sentence combination test is conducted to collect data. The data from the latter test are statistically analyzed to get an idea about the processing difficulty order and to find out how the relative clause type and matrix position exert effects on the production of relative clauses. Then the present results are compared with the previous relevant hypotheses to obtain a deep and thorough understanding of Chinese learners' production difficulty order of English relative clause. The main findings are as follows:Firstly, as for the effect of relative clause type, the mean accuracy scores obtained in these two level groups show that the processing difficulty orders of the six relative clause types are largely in the direction of the prediction of the Noun Phrase Accessibility Hypothesis only with one exception to the ordering of object of preposition and genitive. Although object of preposition and genitive are more complex than direct object syntactically, the participants from both groups do better in object of preposition types than in direct object types, which indicates that the processing difficulty of different relative clause types is not always totally determined by the syntactic complexity. Secondly, the effect of matrix position manifests itself in the results of both English relative clause proficiency level groups. The results also show that within, not across, each relative clause type, the scores obtained for the object position are higher than those for the subject position without any exception. This is completely consistent with the prediction of Kuno's Perceptual Difficulty Hypothesis. Thirdly, the ordering of O-S and S-OCOMP relative sentence patterns is fully consistent with each other between these two groups. This, to some extent, lends support to Izumi's view that the Noun Phrase Accessibility Hypothesis and the Perceptual Difficulty Hypothesis should be treated complementary to each other. Additionally, the orders of four common relative sentence patterns extracted from the results of both groups are the same. Their order is greatly consistent with the prediction of the Subject-Object Hierarchy Hypothesis and partially with that of the Perspective Shifting Hypothesis. To sum up, the procession and production of relative clause are not only subject to the syntactic complexity, but also to other constraints such as, working memory limitation, lexical perceptual saliency, some processing strategy, and linguistic experience. These constraints may come to play hierarchically and simultaneously.
|
PDF Full Text Request