| 摘要: | 本計畫旨在延展對詞彙文法構式之探究，以及對第二語言學習者的挑戰。詞彙文法構式為語言中多於一字的特定構造，既不像語序公式(formulaic sequence)般完全固化，也不符合語法創造原則，能隨意套入字詞組合出無限句子。因此，語言辭典和文法參考資源皆無法呈現全貌，且往往被語言課程的教學法所忽略。本團隊於108年度計畫中，探究了相對於直接教學法，以學習者為本位的探索式學習如何提升第二語言學習者對構式的敏銳度，本計畫將針對前案之成果，繼續深化研究。先前的實驗設計中，針對構式三大特質：整體語意不具組合特性、語法創造原則部分性、以及構式內部的詞彙有其語意約束性，本團隊以大學英語課程250位受試者進行準實驗研究，檢視何種語言輸入或情境能有效地促發學習。依據初步成果分析，本團隊提出以下假設以進而深化探究：這三大特質並非獨立存在，而使得語言習得窒礙難行，相反的，他們共同構成探索式學習的條件，提供線索使學習者得以推論構式之意涵。換言之，詞彙文法構式為不可分割的整體，內部構造及元素有其相互參照之影響性，進而構成學習者自我提升的要件，使之從不了解構式的特質或是不清楚構式內的陌生詞彙，轉化為將構式視為整體，並能理解內部的陌生詞彙意義。對於構式內部和構式之間的交互作用具敏感度，是母語學習者成功的關鍵，而現今的外語課程教學卻往往忽略此一關鍵。然而，欲培養第二語言學習者之敏感度，探索式學習的環境卻能提供解套方式，幫助學習者在未經指導的情況下，推動學習者主動建構構式和其內部詞彙的意義。本兩年期計畫即著眼於此關鍵，擬深化實徵性研究之設計以探討如何重新定位語言課程教學。計畫首年將以108年度之研究為基準進行準實驗，除了將原實驗設計所囊括的構式增多，也將語言輸入由單句擴增為篇章段落，以檢視在相對非控制且輸入增多的語境下，對於陌生的構式，學習者的敏感度是否依舊能提升。計畫第二年將進而檢視學習者在進行閱讀時對構式的敏感度，藉由泛讀以及深度閱讀的情境，探究其偶發性學習和有意識學習。於此同時，本計畫也擬於兩年期間將StringNet優化至5.0版本，以協助語言教師及學習者進行構式的教與學。 ;The present proposal continues a long-term investigation into lexico-grammatical constructions and the challenges they pose to second language learners. Lexico-grammatical constructions are multiword patterns that are neither frozen like formulaic expressions on the one hand (e.g., “by the way”, “believe it or not”, “as a matter of fact”), nor do they allow the open variation expected by fully productive grammar rules on the other (e.g., wh-questions; the passive construction). As such, they fall outside the scope of both dictionaries and reference grammars and escape the attention of most approaches to classroom language education. The current project investigates to what extent and under what conditions learners can become sensitive to the presence of lexico-grammatical constructions in their input as the first steps to acquiring them by learner-centered discovery rather than direct instruction. This entails investigating learners’ sensitivity to three distinctive properties of lexico-grammatical constructions encountered in context: their non-compositionality, partial productivity, and semantic selectivity. Preliminary results from this current work with 250 students in university Freshman English courses have led to the following novel hypothesis that the proposed research is designed to investigate: These three features of lexico-grammatical constructions are not simply independent properties that make these patterns difficult to learn; rather they conspire to make them discoverable by learners and to render them into informative contexts for inferring the meaning of unknown words within those patterns. These three features implicate the lexico-grammatical pattern as a whole into a reciprocal, mutually illuminating relationship with its parts (i.e., with the words in it). This has the striking consequence of enabling learners to bootstrap incrementally from a state of not understanding the non-compositional construction nor some unfamiliar word in that construction to a state of grasping the meaning of both the construction as a whole and the unknown word appearing in it. While sensitivity to this reciprocity is a key to the success of first language learners, it is largely absent in classroom second language learners. It can, however, be fostered in second language learners as impetus to uninstructed discovery learning of both words and constructions. The proposed two-year research submits this novel hypothesis to empirical investigation in the first year and then explores its implications for re-orienting language pedagogy in the second year. In the first year, the current year investigation is to be duplicated with an expanded the inventory of constructions. Further, the contexts in the first year study are to be expanded incrementally to larger stretches of discourse in less controlled settings to see if learners’ sensitivity to constructions fostered under controlled conditions transfers to novel conditions, with other unfamiliar constructions, and in larger contexts. In the second year, this expansion of context is developed further by investigating learners’ engagement with lexico-grammatical constructions both incidentally and intentionally within cycles of extensive and close reading. Over the two-year funding period and in parallel to these investigations, the next version of StringNet (5.0) will be developed to make lexico-grammatical patterns even more readily discoverable to learners and teachers. |
