以英語為外語學習者透過閱讀習得搭配詞及引申與過度引申之研究

• Main Authors: Chi, Pei-yin, 紀佩吟
• Other Authors: Truscott, John
• Format: Others
• Language: en_US
• Published: 2012
• Online Access: http://ndltd.ncl.edu.tw/handle/91467712692383925674

碩士 === 國立清華大學 === 外國語文學系 === 100 === This study aimed to explore whether learners acquire adjective-noun collocations through exposure, whether they generalize from what they have acquired to like collocations which share close semantic relations, and whether they overgeneralize at the same time. Ninety five non-English majors participated in the study, divided into an experimental group and a control group. The research items were seventy two adjective-noun collocations/pseudo-collocations, which fell into three categories. “Acquisition collocations” were included to examine learners’ acquisition through exposure. Each acquisition collocation had a corresponding collocation (referred to as “generalization collocation”) whose noun collocate was semantically related (e.g. heavy drinker and heavy smoker). “Pseudo-collocations ” were adopted as indicators of overgeneralization. The participants received a pretest, reading exposure, an immediate posttest and an identical delayed posttest. These tests asked them to judge the collocability of the target collocations/pseudo-collocations. The reading materials consisted of 48 isolated sentences which contained the acquisition collocations. The control group received no treatment but took the three judgment tests, which were administered to examine the internal validity of the tests used. The results showed that reading exposure had significant effects on learners’ acquisition of collocations. Based on the acquired collocations, learners generalized within the paradigmatic sets of the noun collocates (e.g. from the acquisition collocation heavy drinker to the generalization collocation heavy smoker). Overgeneralization occurred at the same time as a result of the exposure they received; in other words, learners accepted significantly more pseudo-collocations on the posttest than they had accepted on the pretest. The acquired collocations were significantly retained after a two-week delay. On the other hand, the learners showed even more generalization and overgeneralization two weeks later. Through item analysis, this study also probed into factors such as congruency and perceived semantic relations which might influence the extent of acquisition, generalization, and overgeneralization. It also showed that learners will generalize from the collocations they are exposed to, correctly accepting semantically related collocations but also wrongly accepting a great many pseudo-collocations. This and other findings of the study can serve as a basis for future research further exploring these issues.