台灣母語人士對外語口音的評價與成見－以英、泰母語的華語學習者為例

新住民已為台灣第五大族群。台灣母語人士對新住民等外語口音者的語言態度尚未充分研究，因此啟發本研究動機。本論文以規範化研究（norming）篩選出英、泰母語者的輕、重口音樣本及台灣母語者樣本，再將篩選後的口音樣本與歐美、東南亞、台灣等族裔臉孔搭配，設計與地位向度相關的評價性題目，由台灣母語者對不同臉孔的外語口音者以七級量表進行口音評價。本研究結果顯示，聽者依口音辨識而得的族裔訊息，顯著地影響其對各題的評價。臉孔因素僅在部分題目造成影響。七成的參與者（共75位台灣母語者）可正確判斷說話者的族裔。正判者多以口音歸類族裔，口音因素顯著地影響各題評價。正判者對英母語口音評價較正面，對泰母語口音評價較負面。三成的參與者誤判說話者的族裔，其評價顯著受到說話者本身的口音程度影響。誤判者對輕口音評價較正面，對重口音評價較負面。對正判者而言，說話者的族裔訊息明確，因此族裔成見影響聽者的評價。對誤判者而言，說話者的族裔訊息不明確，因此聽者以口音程度評價說話者。語言成見說認為說話者的言語形式觸發聽者的成見。倒轉語言成見說認為族群成見扭曲聽者對說話者的感知。兩學說皆強調族裔辨識的因素（口音或臉孔）。然而，本研究結果顯示，倒轉語言成見說忽略了聽者無法從臉孔辨識說話者的族裔訊息時，聽者將使用口音作為辨識線索藉以評價說話者。當口音辨識的族裔和臉孔辨識的族裔衝突時，聽者最終採用口音而非臉孔為辨識線索進行評價。因此，本研究結果支持口音成見說。本論文對華語教學儲備教師提出反思語言態度的建議活動，也對大學校園中本地學生和外語口音學生的互動提出研究啟示。希能喚起華語教學界對口音成見的重視。

關鍵字

外語口音；口音成見；語言態度；口音評價；新住民

並列摘要

New immigrants have formed the fifth largest sub-group in the population of Taiwan. This study is conducted to explore language attitudes of Chinese native speakers toward foreign-accented speakers, such as new immigrants in Taiwan. A norming study was utilized to select mild non-native-accented, heavy non-native-accented and native-accented speech samples from English, Thai, and Chinese native speakers. The selected speech samples were matched with photographic depictions of Caucasian, Southeast Asian and Taiwanese faces. A questionnaire consisting of status-related evaluative questions was employed to measure the accent stereotypes and language attitudes of Chinese native speakers in Taiwan. The results of this study indicate that ethnicity identified through accents significantly affected the listeners’ evaluations of the speech samples. The effects of speaker’s face and its ethnic indicators only acted on some, not all, evaluative questions. Seventy percent of the respondents accurately identified the speakers’ ethnicity and ascribed ethnic stereotypes to foreign-accented speakers. Positive evaluations were exhibited towards English-native-accented speakers, whereas negative evaluations were given to Thai-native-accented speakers. Thirty percent of the respondents incorrectly identified the speakers’ ethnicity and reacted differently towards speakers with different degrees of accentedness. Positive evaluations were given to mild-accented speakers, while negative evaluations were attributed to heavy-accented speakers. Linguistic Stereotyping Hypothesis holds that speech patterns of accented speakers activate the listeners’ stereotypical judgments. Reverse Linguistic Stereotyping argues that ethnic stereotypes distort the listeners’ perception of accented speakers’ speech. Both Hypotheses stress the role of ethnicity identification when listeners evaluate foreign-accented speakers. However, this study addresses the aspects neglected by Reverse Linguistic Stereotyping. Namely, when the ethnicity cannot be identified through the face, the accent will be the cue for identifying ethnicity and for processing evaluations. In addition, when the ethnicity identified through accents is inconsistent with the ethnicity identified visually through photographic depictions, listeners will categorize and evaluate the speakers based on the auditory information, i.e. the accents, and not on the visual cues. Therefore, this study supports Linguistic Stereotyping Hypothesis. At the end of this study, an activity for prospective Chinese-teaching teachers to reflect on their language attitudes is suggested. Implications concerning the interactions between native-accented and foreign-accented undergraduates in the campus are also discussed. With the increase of contact between Chinese native speakers and foreign-accented speakers brought on by globalization, this study aims to draw the attention of Chinese-teaching professionals to the accent stereotypes exhibited towards foreign-accented speakers.