口譯產出停頓時的認知歷程：以視譯眼動軌跡為證

Abstract

停頓研究推論位於句尾的結構停頓（juncture pause）和位於句中其他位置的遲疑停頓（hesitation pause）應分別具有獨特的功能，並象徵不同的認知處理歷程。然而目前為止鮮有研究專於分析講者/口譯員於停頓當下的認知反應，或採用第二組資料對口語產出進行三角檢測。本研究採用眼動資料以分析視譯產出中的停頓，因眼動資料已普遍應用於閱讀研究當中，作為認知歷程的指標。本研究主要檢視產出停頓當下的眼睛凝視位置和掃視方向，並比較眼動資料與口語產出內容。

本研究蒐集十一位受試者於中進英視譯時的口語產出及眼動資料。研究結果顯示產生結構停頓時，眼睛多位於第一次或第二次閱讀，且凝視在停頓後即將產出的段落，顯示受試者正在口譯的初期階段，理解並重組即將產出的內容。產生遲疑停頓時，眼睛多位於第三次閱讀之後，且停頓多以回視開始，顯示受試者正在偵測錯誤，解決造成產出困難的字詞、句法和翻譯策略問題。研究結果支持結構停頓和遲疑停頓代表不同的認知處理歷程。根據本研究結果，可修正口譯品質評量中的停頓標準，以及提出較系統性的方式，教導口譯學生如何控制口語產出的停頓。Pause studies have raised the assumption that juncture pauses, pause that occurs at sentence junctures, and hesitation pauses, pause that occurs elsewhere in the sentence, could possess distinct functions and indicate different cognitive processes. However, so far little research has focused on examining the speaker / interpreter’s cognitive reaction right at the moment of pausing, nor has additional data been used to triangulate the oral production analyses. The present study incorporates eye movement data, which has been commonly used in reading studies as indicators of cognitive processes, into the study of pausing in sight translation. The research focused on examining eye fixation location and saccade trajectory at the time of pausing, and comparing eye movement data with oral production.

Oral outputs and eye movements of 11 trainee interpreters were recorded during a task of Chinese to English sight translation. Results showed that the eye is usually in the first or second reading pass during juncture pause, and tend to fixate on the upcoming segment after the pause, suggesting the interpreter is engaged in early stages of interpreting, comprehending and reformulating the upcoming text. During hesitation pause, the eye tends to fall in the third reading pass and on, and the pause is likely to begin with a regression, which indicates the interpreter is involved in the error detection stage, solving lexical, syntactic and strategic problems causing production difficulties. The results support the assumption that distinct cognitive processes go on during juncture and hesitation pause. This study may help refine the pause criteria in assessment of interpretation quality, and produce a more systematic way of instructing the control of pauses to interpretation students.