| Summary: | 碩士 === 國立高雄師範大學 === 英語學系 === 92 === The study was conducted to investigate how the position of relative clauses (RCs), the humanness of head nouns, the function of RCs, the grammatical function of relative markers, and the type of relative markers interacted with the students’ use of RCs in their writing. To carry out the study, a database of RCs was first set up by collecting the compositions written by 102 English majors in National Kasohsiung Normal University, 51 male and 51 female. Then, the frequency count of RCs categorized along the five variables was put under the scrutiny of goodness-of-fit and chi-square analyses to figure out the frequency distribution of the five variables as well as the interaction between the five variables and each others’ frequency distribution. The following were the major findings of this study:
(1) In consistency with the prediction of Kuno’ Perceptual Difficulty Hypothesis (PDH), right-embedded RCs appeared more frequently than center-embedded RCs.
(2) Relativization was of higher probability to occur with non-human head nouns than human head nouns.
(3) RCs were defaulted to be used as restrictive and non-restrictive RCs tended to appear at the end of the matrix clauses.
(4) The frequency distribution of RC of different functions accorded with the prediction of Keenan and Comrie’s Noun Phrases Accessibility Hierarchy (NPAH) and RCs with subject relative markers were the most frequent. However, the humanness of head nouns might interact with the frequency distribution. When head nouns were human, the chance for relative markers to be subject increased, and when head nouns were non-human, the chance for relative markers to object increased.
(5) Relative makers were inclined to be Wh (i.e. which and who). When RCs were center-embedded, chances were that relative markers appeared as Zero (i.e. relative markers omitted). Moreover, non-human head nouns added to the chance for relative markers to be That or Zero.
On the basis of the findings, it was inferred that a module combing PDH, NPAH, and the humanness of head nouns could account for the difficulty of RCs. PDH predicted that right-embedded RCs (i.e. head nouns as object) were easier than center-embedded RCs (i.e. head nouns as subject); NPAH predicted that the RCs with subject relative markers were the easiest form of RCs. Combining the prediction of PDH and NPAH, the following difficulty order of RCs was derived:
OS > OO > SS > SO (“>” stands for “easier”)
However, when head nouns were non-human, the difficulty order of RCs would be:
OO > OS > SO >SS
|
|---|
