國立政治大學研究所碩士論文提要 研究所別：語言學研究所 論文名稱：中文交談中修復之社會語言學分析 指導教授：詹惠珍 博士 研究生：魏詩婷 論文提要內容： 在交談情境中，當說話者(Interlocutor)無法清楚表達其訊息時就必須進行修復(Repair)。修復的種類不僅止於句法結構，在說話者發現語意不足時，就有做語意修復的必要性。而修復的種類也會依其所行使的語用功能(Pragmatic Function)而有所不同。修復的方式更有可能因交談者雙方的年齡差距大小而改變。因此，本研究針對說話者行使的語用功能及交談者的年齡差距來探討修復的語言形式與使用情形。 本研究所使用的語料來自十份日常生活會話。其中，三份來自於20-30歲的說話者，三份來自於40-55歲的說話者；其餘四份來自於跨年齡層的交談者。另外，交談者的關係若非好友則為親戚，因此較沒有距離感。每份語料長約三十至四十分鐘，主題均與日常生活相關，以便蒐集到最自然的語料。 語料又根據修復的方式分為句法層次（Syntactic Level）及語意層次(Semantic Level)。其中句法層次又包含刪除（Deletion與添加(Addition)兩重種策略；語意層次則包括替換(Replacement)及添加(Addition)兩個策略。語用功能則分為釐清(Clarification)、確認(Confirmation)、解釋(Explanation)、贊成(Agreement)、強調(Emphasis)及弱化(Alleviation)六種。 所有語料先經分類後，再加以統計檢定。研究結果發現：(一)說話者使用語意修復的頻率顯著高於句法修復，(二)在語意修復中，又以縮限(Narrowing)的使用情形最多。(三)修復大多用來行使釐清語意的功能。(四)說話對象的年齡對修復的使用有部分顯著的影響。(五)年紀小的說話者較少對年紀大的聽話者進行修復。除了量化分析之外，本研究亦追加訪談，以便雨量話分析的結果做初步比對。訪談內容發現，大部分的說話者認為他們的確會因為不同的語用目的而使用不同的修復，但交談者的年齡並不會完全影響修復的使用情形。 關鍵字：社會語言學，交談分析，語用學，修復 / Abstract Perfect utterances do not occur all the time during the conversation. An unclear message is usually repaired to maintain the clarity of meaning. Repair forms at Syntactic and Semantic levels are examined in this study. Also, it is proposed that formal distribution of repair forms are conditioned by pragmatic and social factors. On pragmatic aspect, the principles of Clarity and Expressivity are conformed. For social constraint on repair forms, the influence of interlocutors’ age is suggested. Data analyzed in this study are collected from ten dyadic, face-to-face daily conversations, with each lasting more than 30 minutes. Subjects in the ten conversations share the same ethnic background—Taiwanese. In addition, they are from two different age groups, with half of them at the age between 20 to 30 and the other half between 40 to 55. Among the ten conversations: three of them are conducted by both interlocutors being old, three by both interlocutors being young, and four by interlocutors from different age groups. Repair forms found in the data are categorized into two linguistic categories. In Syntactic aspect, repairs are derived from Deletion and Addition strategies, the former consists of Word-Deletion, Ellipsis, and Marker Deletion, and the latter Modal Addition, Addition of Marker for Focus Changing, and Addition of Marker for Attitudinal Adjustment. As for Semantic repairs, they are those resulted from strategies of Replacement and Addition. The former includes Synonyms, Substantializatoin, Hyponymy, and Hypernymy; while the latter is composed of Narrowing. Results of quantitative analysis yield several patterns. First, repair forms at Semantic level score significantly higher than those on Syntactic level. Moreover, within the Semantic realm, Narrowing is the strategy most frequently used. Second, there seems to be a significantly stronger preference for repairs to conform Clarity principle than to comply Expressivity principle. Among pragmatic functions under Clarity principle, Clarification is the pragmatic function that recieves first priority. Interlocutors’ age is only partially influential to a speaker’s choice of repair forms with interlocutors from younger age group being noticed to be putting more emphasis on the importance of hearers’ age than those from older age group. Follow-up interviews suggest that interlocutors manifest repair differently for certain purposes. However, most interviewees point out that the consideration of both interlocutors’ age does not influence the choice of repair forms. Instead, it is hearers’ age, solely, that lay significant effects on the use of repairs for pragmatic functions. Key Words: Sociolinguistics, Conversation Analysis, Pragmatics, Repair

國立政治大學英國語文學系在職碩士班 碩士論文提要 論文名稱：使用真實語料做為高中生英語聽力輔助教材之個案研究 指導教授：彭欽清 研究生：林宜如 論文提要內容： 本研究旨在分析探討以廣播真實語料 (authentic broadcast materials) 做為台灣高中生英語聽力輔助教材之可能性，並兼論不同題型設計 (task types) 對不同程度學生聽力表現上的影響，以期能夠提供高中英語教師在英聽教學教材上的另一選擇及英聽測驗的方針。 實驗對象為41位高一學生。學生在學期初以全民英檢聽力測驗結果分為高、中、低三種程度，並接受為期十週的廣播英聽訓練。實驗過程中搭配以不同測驗題型來協助並檢視學生英聽理解力。學期結束前再以同樣之英檢試題做為後測。 本研究採用質與量的分析方法。質的分析涵蓋：（一）實驗對象的英文學習背景；（二）學生英聽困難的分析；（三）學生對選材及不同題型的看法。量的分析則以SPSS統計法，包括（一）前後測統計數值的差異；（二）取材（text types）與後測英聽成效的關係；（三）題型（task types）與後測英聽成效的關係。 研究結果發現：受試者在收聽廣播真實語料上最大的困難在於速度及連音。但經過廣播真實語料的英聽訓練後，高程度組不但在後測表現上有明顯進步，且較習慣於外國人的速度及連音，並能接受更長篇的真實語料。至於題型設計對英聽成效影響不大，但是，從學生問卷發現，高程度者認為克漏字題型較有助於英聽訓練，而中、低程度者則都認為選擇題較有助於英聽訓練。最後，本研究提出一些英語聽力教學上的建議：（一）教學初期只要求學生達到重要但簡單的基本目標；（二）教導學生分辨中英文語言的差異以及口說與書寫英文的不同特性；（三）設計多樣化的題型。最後，希望本研究能對台灣的高中英語聽力教學有所助益。 / ABSTRACT This study aims chiefly at the feasibility of using authentic English broadcast programs as supplementary materials in listening comprehension training for EFL senior high school students. It also discusses the effect of different task types on different proficiency levels in English listening comprehension performance, in the hope that the research would provide other alternative of supplementary listening materials and test designs as a reference for senior high school teachers. The research is conducted among 41first-year senior high school students. In the beginning of the first semester in fall, 2002, the subjects were divided into three proficiency levels (high, middle and low) according to the results of their performance on the GEPT listening comprehension as a pretest. Then they were given a ten-week authentic English broadcast listening training. During the course of training, different task types were provided to evaluate students’ listening proficiency and to help them to the focus of the test. At the end of the semester, the same GEPT listening task was administered as a posttest. The research adopts both qualitative and quantitative data analyses. Qualitative analyses consist of (1) the subjects’ English learning background, (2) the English listening comprehension difficulties encountered by the subjects, and (3) students’ views on different test and task types. By using SPSS, the quantitative analyses incorporate (1) the difference of statistical values between the pretest and posttest, (2) the correlation between text types and proficiency performance on the posttest, and (3) the correlation between task types and proficiency performance on the posttest. The research has found out that in listening to authentic English broadcast, the major difficulties encountered by the subjects are speed and linkage. Nevertheless, after the ten-week systematic training, the high-achievers are found to have improved obviously in their posttest performance. Little by little the subjects, especially the high-achievers, are not only adapted to the speed and linkage in authentic speech but also used to longer utterance by native speakers. Furthermore, it is also found in the research that the task types have little effect on the subjects’ listening performance. However, from the analysis of the subjects’ feedback questionnaire, the research has found that high-proficiency students reveal that cloze test types are helpful for the training of listening comprehension, while the middle- and low-achievers prefer the multiple choices. It is suggested that in teaching English listening comprehension teachers should (1) start with requiring students to achieve the basic goals the teacher considers essential in comprehension certain types of tasks, (2) teach students to differentiate the features between English and Chinese as well as those between spoken English and written English, and (3) design diverged types of tasks for students to practice listening skills. It is hoped that this research would make some contribution to English listening comprehension teaching in senior high schools in Taiwan.

本文以客語移除類動詞為研究對象探索語意和形式的關係。動詞的語義透過詞彙分解化、概念結構、詞彙化、框架語義和顯像等機制帶出。後以Goldberg (1995, 2006)的構式理論為基礎，運用Iwata (2005a, b)建議的修正模式，對動詞和結構之間的整合，提出更精細、更詳盡的解釋。根據前述方法，我們初步將移除類動詞分為六個次分類，並呈現出他們在結構上被顯像的論元。最後帶入Iwata提出的模組，分析各次分類動詞和不同句子結構的結合情形和其後的語義表現。 / This paper aims to explore verbs of removal in Hakka with respect to the relationship between form and meaning. Following the constructional approach in the shape of Goldberg (1995, 2006) and Iwata (2005a, b), we display a finer mechanism for the integration of verbal meanings and constructions. Verbal meanings, L-meanings (Lexical Head Level Meaning), are carried out through the following concepts: decomposition, conceptual structure, conflation, frame, profiling. Constructional meanings or P-meanings (Phrasal Level Meaning), variants of L-meanings, are manifested by different constructions. In our preliminary analysis, we display six tentative subclasses of verbs of removal in Hakka, depending on their different lexicalized meanings. We also present the different roles they profile in phrasal expressions. Last, we demonstrate the integration of the verbal meaning and several constructions (e.g., BUN construction and LAU construction).

本研究主要探討台灣學生英語的拒絕語言行為(refusal speech act)與美國學生之異同，並探討社會地位高低及啟始行為(initiating acts)兩個變數對語言行為之影響。此外在拒絕時,所使用的句法修飾(syntactic mitigation)及詞彙修飾(lexical mitigation)之類型與頻率,其所用的句型是否與美國學生不同, 以進一步探究學習者母語對其中介語(interlanguage)是否有負面移轉作用。 本研究採用質性及量化研究方法，研究對象為40位以中文為母語的大學學生, 40位英語為母語的美國大學學生,及40位學習英語為外語的大學學生, 各組男女各半。研究工具為言談填充測驗(Discourse Completion Test), 包含12個不同情境(4個請求, 4個建議, 4個邀請),藉此問卷收集受試者語言行為之表現。 本研究的主要發現如下：間接拒絕策略是最常被三組學生使用的拒絕行為。其次,社會地位高低及啟始行為(initiating acts)兩個變數下對三組學生語言行為有不同影響, 而以後者影響尤鉅。除此之外, 台灣學生的英語中介語的語用能力,不僅反映在句法修飾及詞彙修飾兩方面, 同時在直接拒絕行為也與美國大學學生有顯著差異。本文進一步發現學習者母語對拒絕行為之影響, 拒絕之語序及直接與間接拒絕策略的負面移轉作用。本論文並提出教學上之應用建議，讓學生英語學習更有成效。 / Refusals can be regarded as a pragmatic universal, but refusal performance varies with cultures and is constrained by factors like status and initiating acts. However, the relative weight of these factors placed on each type of refusal strategy varies from culture to culture, or even within the same speech community due to intra-lingual variations (Blum-Kulka, 1987). In order to find out the complexity of cross-cultural differences in the realization of this face-threatening act, we investigated refusal speech acts performed by native speakers of Mandarin Chinese, EFL learners and native speakers of American English with the variation of social status and initiating acts. In addition, we compared EFL learners and Americans on linguistic structures including common directive forms, syntactic and lexical refusal mitigation as well as types of errors in the interlanguage of refusals. We also observed evidence of negative pragmatic transfer in EFL learners’ performance in three main categories of refusal strategies and refusal sequences, and tried to explain the causes of the transfer. Three groups of subjects participated in this study: 40 native speakers of American English (NE), 40 native speakers of Mandarin Chinese (NC), 40 Chinese EFL low-intermediate proficiency learners. They were asked to react by writing in the format of discourse completion task (DCT), which consisted of scripted dialogues with the manipulation of the interlocutors’ social status and initiating acts. The results were coded based on the taxonomy developed by Beebe, Takahashi and Uliss-Weltz (1990), and were analyzed qualitatively and quantitatively. Results showed that indirect refusals were the most frequently used main strategy type in the three groups examined. The effect of social status and initiating acts affected the three groups’ performance to a different degree; initiating acts seemed to be a more decisive force in influencing groups’ refusal behaviors. In terms of refusal mitigation, we have found that significantly more syntactic mitigation like interrogatives, conditionals, conditional clause, etc. were employed by Americans to soften the force of refusals. However, similar to the findings in Chen (2006), EFL learners mainly limited their use of modality in their refusals on types of deontic and epistemic modalities whereas Americans were more flexible in using different kinds with the variation of context. Generally, EFL learners’ refusal performance exhibited deviations from native speakers’ norms. Therefore pedagogical implications were suggested which included that teachers should build up EFL learners’ linguistic knowledge, teacher-fronted talk can be supplemented by additional activities that broaden the range of speech acts and provide a broader opportunities for learners. Also, instructors should provide contextual information regarding the similarities and/or differences between L1 and the target language under the influence of social status. Types of initiating act should be included in EFL teaching so that students’ awareness of these social constraints could be activated and they can integrate such information to facilitate their interlanguage system.

感嘆詞在文學作品和口語中，佔據不可忽略的地位。因為感嘆詞本身帶有特殊的表現力和鮮明的情態意義，可以用來表達人類的情緒、感覺、態度、主觀評價，賦予話語生動的表現力，所以感嘆詞也可視為一種特別的修辭手法。儘管感嘆詞在口語和文學作品中佔有一席之地，但因其在現代俄語的詞類系統中，處於較模糊的地帶，對於感嘆詞的研究也相對比其它詞類少，因而教學上受限於研究文獻的不足，辭典中也常因部分感嘆詞的多義性及解釋的簡化而造成外語學習者的困擾。感嘆詞在俄語詞類系統中地位特殊，它既不屬於實詞，也不屬於虛詞，在詞彙文法意義上自然也有別於實詞和虛詞。大部分的感嘆詞在句中不屬於句子成分，此時其語意若是作為句子既有的整體情感或意志意義的輔助，便很容易被外國讀者或譯者所忽略；但事實上卻減損了作品中的細微情感表達而不自覺。有鑑於此，筆者挑選感嘆詞作為研究客體，希望藉由感嘆詞的研究，能使無論是俄語教學者或學習者，對於感嘆詞的說明或理解都能夠清晰明確。本論文的目的是歸納出感嘆詞在俄國幽默諷刺作品中的詞彙修辭意義及句法功能特徵，並針對字典中感嘆詞釋義提供之不規則性或不明確性，尋找本詞類系統化的解釋方法。

語用能力是語言習得成功重要的一環，但在臺灣高中的英語教學中卻未受到重視。教師的語言行為對於學生在語用和文化價值觀方面有很大的影響，所以檢視英語教師和外籍教師在語言行為是否有相似的策略有其必要性。本文著重在高中英語教師和外籍教師祈令語使用之比較，以瞭解臺灣英語教師的語言行為策略之使用和外籍教師有何異同，並進一步探討文化對語言行為之影響。 本研究利用語文完成測驗問卷(Discourse Completion Task Questionnaire)收集受試者之語料，共有50位本國籍及30位外籍高中英語教師參與研究。問卷資料主要以卡方檢定(Chi-square)，並輔以質性分析兩組受試者之回答。 研究結果指出兩組受試者均使用最多非傳統間接策略(non-conventionally indirect strategies)；但是，本國籍英語教師比外籍教師採用更多直接策略(direct strategies)，外籍教師則比本國籍教師常使用其他策略(other strategies)，例如接受學生的行為或私下約談處理。本研究證實文化的確在語言行為中扮演重要角色，更建議中外籍英語教師均有責任教導學生正確的語用知識，以協助他們在不同情境中和他人成功地溝通。

語調是各種語言中不可或缺的語言因素。在漢語中，由於聲調扮演較強大的辨義功能，相形之下，語調的重要性就容易被疏忽。而非屬聲調語言的俄語，語調則具有強大的辨義功能，俄語學習者與教學者不應掉以輕心。本論文所探討的辨義功能將包括：句法功能、語用功能、交際功能（即句子實際切分功能）以及修辭功能。 　　由於在翻譯研究中，鮮少學者將俄語語調和漢譯問題放在一起討論。有鑑於此，本論文以俄語語調和漢譯策略為研究主題，將分別在簡單句、複合句和上下文中，探討俄語語調的辨義功能與等值的翻譯手段。希望藉本論文之研究，除了深入了解俄語語調理論之外，亦擬定適當的漢譯策略，以供相關研究者與學習者之參考。

隨著網路技術的發展，目前網路電話(VoIP)已有逐漸取代傳統電話的趨勢。尤其能夠允許多人同時在線上進行會談是其最大的優勢之一。但在多人參與網路會談時，因為聲音在空間中傳遞或反射等因素，使得由喇叭發出的聲音再次被麥克風收回，造成回音的產生。會談中只要有一位使用者的裝置發生回音時，回音訊號就會在與會者之間擴散，使得所有使用者均會受到影響，進而嚴重影響網路通話的進行。此狀況在參與會談人數越多時，發生機率越高，且對通話品質影響越嚴重。 傳統電話在一對一通話時，通常使用遠端回音消除機制(Near End Echo Canceller)，由接收端在接收聲音後先暫存在記憶體中再播放，再將麥克風擷取的聲音與事先暫存的訊號反向後混合，以抵銷回音。網路會談的環境下，由於沒有標準的聽筒設備，使得回音發生的時間難以預估。且多人參與的網路會談中，由於收聽者所聽到的聲音可能混合多個使用者說話的聲音與回音，使得回音訊號難以偵測。另外，由於網路傳輸的特性，回音訊號到達的時間與順序都難以預估，這使得回音消除機制在多人網路回談中經常失效。 本研究提出藉由語音動態偵測(Voice Activity Detection-VAD)的方式分辨回音訊號，藉由本研究所提出的語音能量VAD判定機制，能夠有效區別正常語音與回音的差異，即可有效的消除回音，同時發揮靜音抑制（Slience Suppression）的效果，阻擋不含語音內容的封包，降低網路頻寬耗用。本研究以自行開發的VoIP軟體進行實地測試實驗，實驗中顯示，我們的方法能消除85%以上的回音。 / With the prosperous development of Internet technology, traditional phone service is being replaced gradually byVoice-over-IP (VoIP) technology. One of the critical problem that is yet to be improved is the echo problem. Due to the difference in working environment, conventional echo cancellation technology may not work well on VoIP system. The echo problem is becoming more critical as the number of participants in a talk session increases. As long as one user fails to depress echos, every other participant in the conference will be infected. The more participant, the higher probability of echo infection. We propose an energy based Voice Activity Detection (VAD) mechnism that effectively differentiate echo from speech signal. Our VAD algrouthm records a user’s speech volume, and based on this information to determine whether the frame is echo or not. By applying this mechnism to network conference, we can filter out echo frames and suppress slience at same time to save bandwidth consumption. We experimented on a self-developed VoIP software platform, the experiment result shows that our method can eliminate more than 85% of the echo.

在現階段上古音的研究中，上古漢語複聲母的存在已普遍獲得學界的肯定，而親屬語的比較更是為複聲母的研究取得了一定的成果。然而，關於複聲母系統的研究在比例上卻顯得不足。上古漢語究竟有擁有哪些類型的複聲母，以及這些複聲母輔音群的搭配方式如何，學者們之間更是存在著各種不同的看法，可謂莫衷一是。在本文中，我們將藉由「語言的共性」以及人類發音的「一口原則」特點，企圖從藏緬語、苗瑤語、侗台語等親屬語，設法找出其原始複聲母類型，並透過時空投影法的比對，釐清層次區別與音變規律，重建原始音類，設法替這些原始音類的複聲母類型從漢語內部材料裡找到證明，若此即可為上古漢語建立此一類型的複輔音聲母形式。 本論文由觀察漢語之親屬語言出發，結合諧聲、通假、同源詞、反切又音、聲訓、重文、古籍注音、異文、聯綿詞、漢語方言、古文字、域外譯音等文獻材料，並結合印歐語系之語言，藉由「語言的普遍性」來全面探討上古複聲母的結合規律、類型甚至是演變。筆者嘗試從漢語的內部材料與親屬語等外部材料中歸納出上古漢語複聲母的系統性與複聲母的種類。此外，更用此所得出之結果檢視當前各家學者所構擬之漢語複聲母系統，看看這樣的擬構是否能真能符合語言演化的「規律性」、「普遍性」與「系統性」。 除了上述的方法外，我們將由現代「語音學」的角度去看音節的輔音配列是否合乎響度原則，利用音素的響度高低判斷輔音結合的可能，驗證於親屬語言的複聲母形式，從而認定上古漢語複聲母的可能形式。這種借助語音學的研究成果來檢測所構擬的上古漢語複聲母的類型，是一種新的嘗試。雖然，親屬語言的語音對應規律說明了即便是親屬語言也會因為社會狀態、系統制約等因素而呈現不平衡發展的情形。不過我們相信這種不平衡的差異在親屬語言裡仍呈現規律的對應，某些形式當是由某個最初的形式演變而來。透過時空投影，期盼可以找出語音的演變規律，重建上古漢語原始形式。 此外，我們也應當了解，上古漢語的形態研究和上古漢語語音研究有著極其密切的關係：語音研究為上古漢語的形態研究提供了紮實的基礎；同樣地，上古漢語的形態研究也推進了上古漢語語音研究。眾所皆知，純粹從事上古漢語語法研究的學者絕少涉及到上古漢語的形態，因此本文希望能進一步觸及上古漢語的詞頭問題，將學者們所提出的上古漢語詞頭作一個檢討，對於上古漢語前綴之形態、語法功能有所補充，最後更欲釐清複聲母與詞頭間錯縱複雜的關係，此二者彼此是如何演化？甚至消失在漢語語音的歷史舞台上。

本研究從認知語言學的觀點出發，以具有類似意義之日語複合動詞「～nuku」、「～kiru」、「～toosu」為研究對象，透過複合動詞「～nuku」、「～kiru」、「～toosu」所具有之多重意思，以及其具有之中心基模，來探討其語意之異同點。 過去的研究主要是以複合動詞之前項動詞的分類為基準，來探討具有類似意義之複合動詞之異同點，然而卻無法完全說明造成其異同點之關鍵原因。本研究著重於複合動詞具有之「中心基模」與「語意擴張構造」，並主張具有類似意義之複合動詞其最大的不同，主要即來自於各個複合動詞本身具有之不同的「中心基模」與「語意擴張構造」。 透過此分析方法可以得知，複合動詞「～nuku」、「～kiru」、「～toosu」其各自的「中心基模」分別如下： 「～nuku」為「克服行動過程中所具有之障礙」 「～kiru」為「由於某動作之影響，造成某對象消失之結果」 「～toosu」為「通過某路徑之起點與終點」 由以上所述之不同「中心基模」來探討可知，複合動詞「～nuku」、「～kiru」、「～toosu」雖然具有類似意義，但在「時間範圍」、「量範圍」、「程度範圍」、「認識範圍」、「有無意志性」、「有無障礙」、「有無目的」、「積極度」、「理解度」等項目中，仍可發現其不同之處。