中英雙語者的空間用語與空間認知：眼動儀研究

Spatial Terms and Spatial Cognition of Chinese-English Bilinguals: An Eye-Tracking Study

根據強語言相對論，記憶歷程的空間認知理應受到不同語言的空間提示用語的影響或限制。弱語言相對論則對於語言的影響力持保守觀點，認為語言在認知歷程中扮演的是形塑分類系統的調節角色（Newcombe, 2005）。為探討此一替代觀點，本研究利用眼動儀測量中英雙語者從事空間記憶再認作業編碼過程中的眼球活動，並選擇以語意分界上不完全一致的中英空間用語做為提示，同時操弄不同作業複雜程度，以探討這兩種語言的空間用語，如何對中英雙語者之空間認知表現產生不同的影響。本實驗為2（提示語言：英文空間用語 vs. 中文空間用語） × 2（作業複雜度：多內容 vs. 少內容）完全受試者間設計，依變項為以眼動儀測量記憶編碼階段之興趣區總停留時間與總凝視次數。共有73位中英雙語成人參與實驗，隨機分為四組，分別接受四種實驗情境之一。過程當中以眼動儀配合電腦程式，紀錄按鍵與眼動訊息。結果顯示，英語特定空間用語的提示效果，僅在記憶目標複雜度較高時能彰顯。本研究結果支持弱語言相對論，即語言對於認知的影響，並非徹底改變認知表現，而是藉由調整分類界線的方式，扮演調節的角色。

According to the strong version of language relativity theory, spatial terms should affect or constrain the performance of spatial recognition tasks when cued by different languages. However, a weaker version of linguistic relativism proposes a “language as a category shaper” moderator point of view (Newcombe, 2005). To further explore this alternative viewpoint, the current study aims to investigate how categorizing boundaries in spatial terms between two languages may affect behaviors differentially as a function of task complexity. To achieve this goal, Chinese- English bilinguals' spatial cognition was assessed with eye-movement tracking techniques when they underwent visual recognition tasks cued with either Chinese or English spatial terms. The present study took on a 2 (cueing languages: Chinese vs. English) × 2 (task complexity: Less vs. More) complete between-subject experimental design. Seventythree Chinese-English bilinguals were randomly assigned into 4 groups, ChLess, ChMore, EngLess and EngMore. Their eye moving patterns during visual encoding stage were observed and examined to see how they were affected by cueing languages under various degrees of task complexity. Eyetracking equipment was programmed to present task materials and record task performance, including eye-movement information. Total dwell time and total fixation count in specific areas of interest during the encoding phase were measured and compared among the 4 groups. We predicted that only when the visual information is complicated enough can the characteristic features of the cueing languages be shown. Our findings support this prediction, which is derived from the weaker version of language relativity theory. That is, our language does not change our cognition in a thoroughly one-way fashion. Instead, it moderates our thoughts as a category shaper.