以語體變異現象驗證認知框架之存在 - 政大學術集成

全文

(1)國立政治大學語言學研究所碩士論文 National Chengchi University Graduate Institute of Linguistics Master Thesis. 指導教授：詹惠珍治博士. 政. 大. 立 Dr. Hui-chen Chan Advisor: ‧. ‧ 國. 學. Nat. io. sit. y. 以語體變異現象驗證認知框架之存在. n. al. er. Stylistic Variation as Surface Evidence for Frame. Ch. engchi. i n U. 研究生：高子貽撰 Student: Tzu-yi Kao 中華民國一百零四年四月 April, 2015. v.

(2) STYLISTIC VARIATION AS SURFACE EVIDENCE FOR FRAME. BY Tzu-yi Kao. 學. n. al. y. er. io. sit. Nat. A Thesis Submitted to the Graduate Institute of Linguistics in Partial Fulfillment of the Requirements for the Degree of Master of Arts. ‧. ‧ 國. 立. 政治大. Ch. engchi. April, 2015. i n U. v.

(3) 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. i n U. Copyright © 2015 Tzu-yi Kao All Rights Reserved. iii. v.

(4) Acknowledgements. 在語言所求學期間，拓展了我對語言學此學科的眼界，大部分的人聽到語言所，會認為語言所的學生會講各國的語言，剛進入語言所的我，認為語言學是研究語法、語音等，侷限在語言層面的一門學科，在剛踏入語言所時我覺得疑惑，為何有些語言學家強調語言學是一門科學，在進入語言所後，逐漸明瞭其理。語言學涉及的領域甚廣，與其他學科領域皆有關聯，例如近期的新興領域大數據(Big data)，許多語言學的相關研究也採用大數據做為分析依據，講求一分證據說一分的話，又如我的論文中所研究的認知框架理論，竟然可以運用在人工智慧上，由此可見，語言學是一門跨足各領域的科學。此外，許多人認為語言學理論大多只能運用在語言教學上，其實不然，醫學上的語言治療與語言學習習相關，書店架上暢銷的談判、溝通、說話術等書籍，也是語言學涉及的領域，語言學涉獵的範圍相當廣闊，並且非常實用。. 立. 政治大. ‧ 國. 學. ‧. 能夠完成這本論文，主要要感謝我的論文指導老師詹惠珍老師以及我的父母親，感謝詹老師的引導與督促，不僅教導我學術方面的專業知識，也常講述一些待人處事的道理，感謝我的父母親一路支持我就讀研究所，並給予經濟上的協助使我能夠專注在課業上，另外，感謝我的論文口試委員曹逢甫老師和蘇席瑤老師願意擔任口委並給予精闢的意見，此外，感謝賴惠玲老師和徐嘉慧老師平時的鼓勵與照顧。. n. er. io. sit. y. Nat. al. Ch. engchi. i n U. v. 還要感謝助教學姊在行政事務上的提醒及協助，還有我的同班同學們，謝謝你們讓我在研究所期間度過了一段非常美好的學生時光，並在我有困難時給予打氣與協助，感謝你們，我才能完成這本論文。. iv.

(5) TABLE OF CONTENTS. Acknowledgements ....................................................................................................... iv LIST OF TABLES ........................................................................................................ ix LIST OF FIGURES ...................................................................................................... xi LIST OF ABBREVIATIONS ......................................................................................xii Abstract ........................................................................................................................ xv Chapter 1 Introduction ................................................................................................... 1 1.1. Motivations ..................................................................................................... 1 1.2. Research Questions ......................................................................................... 2 1.3. Hypotheses ...................................................................................................... 2 1.4. Outline of the Thesis ....................................................................................... 3. 立. 政治大. ‧. ‧ 國. 學. Chapter 2 Literature Review .......................................................................................... 4 2.1. Frame Analysis................................................................................................ 4 2.1.1. The Definition of Frame ...................................................................... 4 2.1.2. Footing and Frame ............................................................................... 5 2.1.3. Message and Metamessage .................................................................. 7 2.1.4. Interactive Frame V.S. Knowledge Schema......................................... 7 2.1.4.1. Types of Frame.......................................................................... 7 2.1.4.2. Interaction of Frames and Schemas .......................................... 8 2.1.5. Evidence of Frame ............................................................................... 9 2.1.5.1. The Peer-story Experiment ....................................................... 9 2.1.6. Frame Shifting ................................................................................... 11 2.1.6.1. Reframing the Frame .............................................................. 11 2.1.6.2 Register Shifting ...................................................................... 12 2.1.7. Interaction of Frames ......................................................................... 13 2.1.7.1. Yield or Resist the Frame ........................................................ 13 2.1.8. Levels of Frames ................................................................................ 13 2.2. Discourse Structure ....................................................................................... 15 2.2.1. The Structure of a Narrative .............................................................. 15 2.2.2. The Structure of a Conversation ........................................................ 17 2.2.3. Components of Communication ........................................................ 18 2.3. Illocutionary Acts .......................................................................................... 19 2.4. The Cooperative Principle (CP) .................................................................... 20 2.5. Stylistic Variation .......................................................................................... 21. n. er. io. sit. y. Nat. al. Ch. engchi. v. i n U. v.

(6) 2.5.1 Notions of Stylistic Variation .............................................................. 21 2.5.2. Code-switching as a Strategy of Stylistic Variation ........................... 22 2.5.2.1. The Definition of Code ........................................................... 22 2.5.2.2. The Definition of Code-switching .......................................... 23 2.5.2.3. Linguistic Aspect of Code-switching ...................................... 24 2.5.2.4. Functional Aspect of Code-switching ..................................... 26 2.5.2.4.1. Functions of code-switching ........................................ 26 2.5.2.4.2. Situational, metaphorical and conversational code-switching. ............................................................ 27 2.5.2.4.3. We Code and They Code................................................ 29 2.6. Summary........................................................................................................ 29 Chapter 3 Methodology ............................................................................................... 30 3.1. Data Collection ............................................................................................. 30 3.1.1. Data Sources ...................................................................................... 30 3.1.2. Sampling ............................................................................................ 30 3.2. Measurement ................................................................................................. 31 3.3. Criteria for Data Classification ..................................................................... 31 3.3.1. Linguistic Criteria .............................................................................. 31 3.3.1.1. Lexical Level .......................................................................... 31 1. Semantic density of word......................................................... 31 2. Formality of words ................................................................... 35 3. Word frequency ........................................................................ 40 3.3.1.2. Syntactic Level........................................................................ 41 1. Sentence complexity ................................................................ 41 2. Sentence completeness............................................................. 43 3. Sentence patterns ..................................................................... 44 3.3.2. Functional Criteria ............................................................................. 45 3.3.2.1. Discourse Structure ................................................................. 46 3.3.2.1.1. Narrative structure ....................................................... 46 3.3.2.1.2. Conversational structure .............................................. 50 3.3.2.2. Illocutionary Acts .................................................................... 52 3.3.2.3. Cooperative Principle.............................................................. 54. 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. i n U. v. Chapter 4 Data Analysis and Discussions .................................................................... 59 4.1. SVF by Linguistic Strategies in General ...................................................... 59 4.1.1. SVF by Lexical Devices .................................................................... 60 4.1.1.1. SVF by Shifting Semantic Density of Words ......................... 61 vi.

(7) 4.1.1.2. SVF by Shifting Word Formality ............................................ 62 4.1.1.3. SVF by Shifting Word Frequency ........................................... 63 4.1.2. SVF by Syntactic Devices ................................................................. 63 4.1.2.1. SVF by Shifting Sentence Complexity ................................... 64 4.1.2.2. SVF by Shifting Sentence Completeness ............................... 64 4.1.2.3. SVF by Shifting Sentence Patterns ......................................... 65 4.2. SVF for Discourse Structure ......................................................................... 65 4.2.1. SVF for Narrative Structure ............................................................... 65 4.2.1.1. SVF for Narrative Structure by Lexical Devices .................... 66 4.2.1.1.1. SVF for Narrative Structure by Shifting Semantic Density of Words ......................................................... 67 4.2.1.1.2. SVF for Narrative Structure by Shifting Word Formality ...................................................................................... 69 4.2.1.1.3. SVF for Narrative Structure by Shifting Word Frequency..................................................................... 71 4.2.1.2. SVF for Narrative Structure by Syntactic Devices ....................... 72 4.2.1.2.1. SVF for Narrative Structure by Shifting Sentence Complexity................................................................... 73 4.2.1.2.2. SVF for Narrative Structure by Shifting Sentence Completeness ................................................................ 74 4.2.1.2.3. SVF for Narrative Structure by Shifting Sentence Patterns .... 74 4.2.2. SVF for Conversational Structure.......................................................... 75 4.2.2.1. SVF for Conversational Structure by Lexical Devices ................. 76 4.2.2.1.1. SVF for Conversational Structure by Shifting Semantic Density of Words ........................................................... 76 4.2.2.1.2. SVF for Conversational Structure by Shifting Word Formality ...................................................................... 78 4.2.2.1.3. SVF for Conversational Structure by Shifting Word Frequency..................................................................... 80 4.2.2.2. SVF for Conversational Structure by Syntactic Devices ........ 81 4.2.2.2.1. SVF for Conversational Structure by Shifting Sentence Complexity ................................................................... 82 4.2.2.2.2. SVF for Conversational Structure by Shifting Sentence Completeness ................................................................ 82 4.2.2.2.3. SVF for Conversational Structure by Shifting Sentence Patterns ......................................................................... 83 4.3. SVF for Illocutionary Acts ............................................................................ 84 4.3.1. SVF for Illocutionary Acts by Lexical Devices ................................. 84. 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. vii. i n U. v.

(8) 4.3.1.1. SVF for Illocutionary Acts by Shifting Semantic Density of Words .................................................................................................. 85 4.3.1.2. SVF for Illocutionary Acts by Shifting Word Formality ........ 87 4.3.1.3. SVF for Illocutionary Acts by Shifting Word Frequency ....... 89 4.3.2. SVF for Illocutionary Acts by Syntactic Devices .............................. 90 4.3.2.1. SVF for Illocutionary Acts by Shifting Sentence Complexity 90 4.3.2.2. SVF for Illocutionary Acts by Shifting Sentence Completeness .................................................................................................. 91 4.3.2.3. SVF for Illocutionary Acts by Shifting Sentence Patterns...... 92 4.4. SVF for CP ..................................................................................................... 93 4.4.1. SVF for CP by Lexical Devices............................................................. 93 4.4.1.1. SVF for CP by Shifting Semantic Density of Words .............. 95 4.4.1.2. SVF for CP by Shifting Word Formality ................................ 99 4.4.1.3. SVF for CP by Shifting Word Frequency ............................. 103 4.4.2. SVF for CP by Syntactic Devices .................................................... 105 4.4.2.1. SVF for CP by Shifting Sentence Complexity...................... 107 4.4.2.2. SVF for CP by Shifting Sentence Completeness .................. 109 4.4.2.3. SVF for CP by Shifting Sentence Patterns............................ 111 4.5. SV for Structure of Frame........................................................................... 113. 立. 政治大. ‧. ‧ 國. 學. n. al. er. io. sit. y. Nat. Chapter 5 Conclusion ................................................................................................. 117 5.1. Summary of the Major Findings ................................................................. 117 5.2 Conclusions .................................................................................................. 120 5.3. Limitations and Suggestions ....................................................................... 120. Ch. engchi. i n U. v. References .................................................................................................................. 121. viii.

(9) LIST OF TABLES Table 1. SVF by linguistic strategies ........................................................................... 59 Table 2. SVF by lexical devices ................................................................................... 60 Table 3. SVF by shifting semantic density of words ................................................... 61 Table 4. SVF by shifting formality of words ............................................................... 62 Table 5. SVF by syntactic devices ............................................................................... 63. 政治大. Table 6. SVF by shifting sentence complexity ............................................................ 64. 立. Table 7. SVF by shifting sentence completeness ......................................................... 64. ‧ 國. 學. Table 8. SVF by shifting sentence patterns .................................................................. 65. ‧. Table 9. SVF for narrative structure by lexical devices ............................................... 66. sit. y. Nat. Table 10. SVF for narrative structure by shifting semantic density of words ............. 68. io. n. al. er. Table 11. SVF for narrative structure by shifting formality of words ......................... 70. i n U. v. Table 12. SVF for narrative structure by shifting word frequency .............................. 71. Ch. engchi. Table 13. SVF for narrative structure by syntactic devices ......................................... 72 Table 14. SVF for narrative structure by shifting sentence complexity ...................... 73 Table 15. SVF for narrative structure by shifting sentence completeness ................... 74 Table 16. SVF for narrative structure by shifting sentence patterns ............................ 75 Table 17. SVF for conversational structure by lexical devices .................................... 76 Table 18. SVF for conversational structure by shifting semantic density of words .... 77 Table 19. SVF for conversational structure by shifting formality of words ................ 79 Table 20. SVF for conversational structure by shifting word frequency ..................... 80 ix.

(10) Table 21. SVF for conversational structure by syntactic devices ................................ 81 Table 22. SVF for conversational structure by shifting sentence complexity ............. 82 Table 23. SVF for conversational structure by shifting sentence completeness .......... 83 Table 24. SVF for conversational structure by shifting sentence patterns ................... 83 Table 25. SVF for illocutionary acts by lexical devices .............................................. 84 Table 26. SVF for illocutionary acts by shifting semantic density of words ............... 86 Table 27. SVF for illocutionary acts by shifting formality of words ........................... 88. 政治大. Table 28. SVF for illocutionary acts by shifting word frequency................................ 89. 立. Table 29. SVF for illocutionary acts by syntactic devices ........................................... 90. ‧ 國. 學. Table 30. SVF for illocutionary acts by shifting sentence complexity ........................ 90. ‧. Table 31. SVF for illocutionary acts by shifting sentence completeness .................... 91 Table 32. SVF for illocutionary acts by shifting sentence patterns ............................. 92. sit. y. Nat. io. al. er. Table 33. SVF for CP by lexical devices ..................................................................... 94. v. n. Table 34. SVF for CP by shifting to low semantic density lexical items..................... 96. Ch. engchi. i n U. Table 35. SVF for CP by shifting to high semantic density lexical items ................... 98 Table 36. SVF for CP by shifting to low formality lexical items .............................. 100 Table 37. SVF for CP by shifting to high formality lexical items ............................. 102 Table 38. SVF for CP by shifting word frequency..................................................... 104 Table 39. SVF for CP by syntactic devices ................................................................ 106 Table 40. SVF for CP by shifting sentence complexity ............................................. 108 Table 41. SVF for CP by shifting sentence completeness ......................................... 110 Table 42. SVF for CP by shifting sentence patterns .................................................. 112 x.

(11) LIST OF FIGURES Figure 1. Overriding frame and subordinate frame…………………………………15 Figure 2. Hierarchical relationship among frames in discourse structure……….…114 Figure 3. Hierarchical structure of frame…...............................................................115. 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. xi. i n U. v.

(12) LIST OF ABBREVIATIONS. Stylistic variation = SV Stylistic variation for frame = SVF. 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. xii. i n U. v.

(13) 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：以語體變異現象驗證認知框架之存在指導教授：詹惠珍研究生：高子貽論文提要內容：本研究主要討論認知框架(Frame)的存在可藉由語體變異現象驗證，框架的概念. 政治大. 可藉由說話者語體的轉換辨識，與語體變異現象相關並可驗證認知框架的語言策略. 立. 和功能性策略皆被討論。本篇將歸納出語言策略與功能性策略的對應關係，並藉由. ‧ 國. 學. 語體變異現象驗證框架的結構是具階層性的。. ‧. 本研究以語料庫為本，文內所分析的七筆語料皆來自政大國語口語語料庫，七筆. sit. y. Nat. 語料皆是面對面、包含兩位參與者的對話，分析的過程以說話輪替(Turn)為計量單位，. io. n. al. er. 進而討論語體變異的目標。語料分類條件主要為語言形式策略(包含詞彙及句構)和. i n U. 功能性策略(包含言談結構、言說行動、語用合作原則)。. Ch. engchi. v. 研究結果顯示(一)框架的概念可藉由語體變異現象驗證。語言形式策略包含詞彙的語意密度、正式性、詞頻，以及句型的複雜性、完整性、特定句型模式(Sentences patterns)等；(二)在敘述架構(Narrative Structure)中，語體變異現象主要用來標示闡述(Elaboration)，其次是評價(Evaluation)部分；在對話結構(Conversational Structure) 中，語體變異現象則主要用來標示話題延續(Topic Continuity)(三)語體變異現象主要可用來辨識言說行動中的斷言行為(Assertive)，其次為表述行為(Expressives)和指示行為(Directives)；(四)語體變異現象可用來辨識語用合作原則中的量的準則和方式. xiii.

(14) 準則；(五)語體變異現象驗證框架具有階層性的概念，包含主要三大階層— Denotative level、Metalinguistic level、和 Metacommunicative level。. 關鍵字；認知框架、語體變異現象、階層性框架. 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. xiv. i n U. v.

(15) Abstract. The aim of this study is to investigate how frame can be manifested through stylistic variation. Linguistic devices and functional strategies which related to stylistic variation for the manifestation of frame are discussed; the distributions between linguistic devices and functional strategies of stylistic variation for frame are also patternized. Last, stylistic variation can help to identify hierarchical structure of frame is verified.. 立. 政治大. This study is corpus-based that all of the data are face-to-face, spontaneous,. ‧ 國. 學. dyadic conversations, extracted from NCCU Corpus of Spoken Mandarin. Seven. ‧. samples are analyzed. “Turn” is using as measurement of linguistic unit to count the. sit. y. Nat. amount of stylistic variation for frame. In addition, only the goals of stylistic variation. io. n. al. er. are counted and categorized. Criteria for data classification includes linguistic criteria. i n U. v. (includes lexical devices and syntactic devices) and functional criteria (includes. Ch. engchi. discourse structure, illocutionary acts, and Cooperative Principle). The results of data analysis show that (1) frame can be identified through stylistic variation of lexical choices based on semantic density, word formality, and word frequency, as well as through shifting of syntactic devices which include sentence complexity, sentence completeness, and sentence patterns. (2) On discourse level in a narrative, stylistic variation is most frequently used to signal elaboration, less is evaluation; on discourse level in a conversation, stylistic variation is most frequently used to signal topic continuity. (3) Among the five types of illocutionary xv.

(16) acts, stylistic variation is applied most frequently for assertives, less for expressives and directives, and never for commisssives and declaration.(4) Among Cooperative Principle, stylistic variation is applied most frequently for Maxim of Quantity and Maxim of Manner. (5) Last, hierarchical structure of frame, including subordinate denotative level, metalinguistic level, and dominant metacommunicative level, are verified in this study.. Keywords: frame, stylistic variation, hierarchical structure of frame. 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. xvi. i n U. v.

(17) Chapter 1 Introduction. 1.1. Motivations Stylistic variation (SV, hereafter), the phenomenon of shifting from one linguistic form to another, is commonly used in daily communication. People may shift stylistically on the level of a word, a phrase, a clause, a sentence, or even beyond. 政治大. sentence level. As theories of sociolinguistic variation prescribe, stylistic variation of. 立. every kind serves functional requirements in verbal communication. It has been. ‧ 國. 學. noticed that various linguistic devices, on lexical level as well as on sentence level,. ‧. are applicable not just to serve for different communicative purposes and to reflect or highlight contextual characteristics, but also to manifest frame — the structured. sit. y. Nat. io. al. er. expectation on meta-thinking level－behind all functional demands.. v. n. Although the sociopragmatic functions of stylistic variation have been widely. Ch. engchi. i n U. discussed, the notions of “function” and “context” are quite confusing. To be specific, first, for the concept of “function,” social functions and pragmatic functions are not differentiated, and frequently they are examined within sentence boundary. Next, for the thought of “context,” it is treated as one single entity on one single layer, which is insufficient from the stance of function. Instead, the concept of context is multi-layer and can further be categorized into linguistic context, physical context, situational context, and social context, within each of which it is likely that different linguistic devices for stylistic variation are selected. In addition, rare of the current studies 1.

(18) analyzes stylistic variation from the perspective of frame. For the above reasons, this study intends to examine how stylistic variation may serve as surface evidence for frame.. 1.2. Research Questions The aim of this study is to investigate how frame can be manifested through stylistic variation. The following research questions are included in this study. (1) Can stylistic variation serve for the realization of frame? If yes, what are the. 治政大purpose? linguistic devices of stylistic variation adopted for this 立. ‧. ‧ 國. frame?. 學. (2) What functional strategies are related to stylistic variation for the manifestation of. (3) What are the distributions between linguistic devices and functional strategies of. y. Nat. n. al. er. io. related?. sit. stylistic variation for the manifestation of frame? What are the linguistic features. Ch. i n U. v. (4)Is there hierarchical structure in a frame? If yes, how can stylistic variation help to identify it?. engchi. 1.3. Hypotheses In this study, four hypotheses are given below. (1) Frame can be identified through stylistic variation. Also, linguistic devices applicable include lexical choices (based on semantic density, word formality, and word frequency) and syntactic selections (based on sentence complexity, sentence completeness, and sentence patterns). 2.

(19) (2) Frame on pragmatic level－including discourse structure, illocutionary acts, and Cooperative Principle－can be manifested by stylistic variation. (3) The distributions between linguistic choices (of both strategies and features) and selection of functional strategies for surface representation of frame are patternized. (4) Frame does have hierarchical structure, the existence of which can be verified by stylistic variation.. 1.4. Outline of the Thesis. 立. 政治大. ‧ 國. 學. This study is composed of five sections. The first chapter introduces the motivation of this study, the research questions, and the related hypotheses. The. ‧. second chapter reviews definitions and theories related to frame analysis,. y. Nat. io. sit. pragmatic functions related to frame, and stylistic variation. The third chapter. n. al. er. describes the research design of this study, which includes data collection,. Ch. i n U. v. measurement, and criteria for data classification. The fourth chapter presents data. engchi. analysis and discussions. The last chapter summarizes the major findings and describes limitations and suggestions of the study.. 3.

(20) Chapter 2 Literature Review. This chapter reviews definitions and theories related to frame analysis, pragmatic theories, and stylistic variation.. 2.1. Frame Analysis 2.1.1. The Definition of Frame. 政治大 To define what a frame is, the concept of expectation should be examined first. The 立. ‧ 國. 學. concept of expectation has been discussed in many fields. Related scholars, to name few, include Bartlett (1932), Rumelhart (1975), and Abelson (1975) in psychology; Bateson (1972),. ‧. sit. y. Nat. Hymes (1974), and Frake (1977) in anthropology; Goffman (1974) in sociology; Minsky (1975). n. al. er. io. in artificial intelligence; Fillmore (1975) and Chafe (1977) in linguistics. Different terms－. i n U. v. including schema, script, and frame－are used to refer to the concept of expectation.. Ch. engchi. Bartlett (1932) proposes that schema is dynamic, which is always active and developing through time. Hymes (1974) regards the notion of frame as a means of speaking. In order to interpret utterances correctly, hearers must know what frame they are engaged in. Schank & Abelson (1975:151) give a classical example to describe the notion of script－the restaurant script. The example is given below:. John went into the restaurant. He ordered a hamburger and a coke. He asked the waitress for the check and left. 4.

(21) In this example, definite article “the” is used to refer to “the waitress” and “the check” which do not mention before. Schank & Abelson take this definite article “the” as linguistic evidence of the existence of script. Chafe (1977) proposes that the process of verbalization is composed of three stages. The first stage is to identify what the event is, and to determine what frame will be applied, including determining what roles interlocutors play. The second stage. 政治大. is concerned with construction of syntactic structure. At the last stage, lexical choice. 立. is determined. All these terms can be summed up to Bateson’s notion－frame and. ‧ 國. 學. Ross’s (1975) concept－structure of expectation, with which people use their prior. experiences.. ‧. experience in a given culture to predict and interpret new information, events, and. sit. y. Nat. io. al. er. Goffman (1974) points out that frame can help people to understand the. v. n. incoming message and respond to events. Tannen (1993) adopts Goffman and other. Ch. engchi. i n U. scholars’ notions and develops the definition of frame. According to Tannen (1993:14), “In order to function in the world, people cannot treat each new person, object, or event as unique and separate,” indicating that people, based on their prior experiences, make sense of the world. In other words, with the frame in mind, people perceive, interpret, and verbalize things around them. 2.1.2. Footing and Frame Goffman (1981:128), in his investigation of footing shifts within interaction, finds that footing not only can shift from one to another, but also can be embedded 5.

(22) within one another, which is called lamination of experience. Based on Goffman’s notion of footing, Tannen (1991) purposes that footing is also a kind of frame that can be used to identify the relationship between interlocutors; therefore, footing shift can be regarded as frame shifting. Tannen (1986:91) also suggests that “frames are constantly evolving lines of interpretation, continually negotiate footings.” Tannen (1986) gives an example to explain footing and frame. Imagining a card. 政治大. checker of swimming pool does not let you go into the swimming pool when you. 立. forget to bring the card. He says, “How do I know you’re not trying to sneak in?”, or. ‧ 國. 學. he may say “I wish I could let you in. I don’t think the policy makes sense either, but I. ‧. can’t go against policy.” In the first one, the footing of the card checker is “me and the policy against you,” which leads to opposition; while in the latter one, the footing of. sit. y. Nat. io. al. er. the card checker is “you and me against the policy” that he tries to show empathy, not. v. n. opposition. This example shows that different footings may identify different. Ch. relationships between interlocutors.. engchi. i n U. Another example is Hoyle’s (1993) study that she investigates the sportscasting speech activity which constructed by three 8- and 9-year-old boys when they play games. Hoyle finds that these children can manipulate footing shifts in their play. In the sportscasting frame, the boys play the role of sportscasters. However, when they conduct footing shift, they pretend they are interviewing an imaginary player, playing the role of interviewer. Their footings are continuously changing with the ongoing game. In other words, the changes of their footings display their shifts of frames. 6.

(23) 2.1.3. Message and Metamessage Hoyle (1993:114) suggests that all messages involve implicit metamessage. Bateson (1972) uses the term “metacommunicating” to refer to the notion of frame. Tannen (1986:86) proposes that “metacommunicating itself carries a metamessage of involvements.” Therefore, meaning is composed on at least two layers, one is message (or surface meaning), and the other, metamessage (or intended meaning). Tannen (1986:88) gives an example, in which a woman takes a trip to London. 政治大. to visit her friends on Christmas holiday after she gets divorced. When the holiday is. 立. over, a male friend of her says, “You don’t have to go all the way to London not to be. ‧ 國. 學. alone on Christmas. Next year you can spend Christmas with us.” The woman thanks. ‧. for her friend’s kindness, but she feels offended. The message of the male friend’s speaking is meant to be an invitation, but the metamessage implies the woman’s. sit. y. Nat. io. al. er. pathetic attempt to avoid being alone on Christmas Day. Tannen concludes that such a. v. n. confusing communication is owing to the conflict bind of message and metamessage.. Ch. engchi. 2.1.4. Interactive Frame V.S. Knowledge Schema. i n U. 2.1.4.1. Types of Frame According to Tannen & Wallat’s (1987) study, the concepts of frame can be divided into two categories, one is interactive frame (i.e. frame), and the other is knowledge schema (i.e. schema). Interactive frame, being regarded as “dynamic”, refers to a definition of knowing “what is going on in interaction” or “what activity is being engaged in” (Tannen & Wallat, 1987:59). Bateson (1972) suggests that people have to know which frame is being applied 7.

(24) in the discourse, so they can decode the message without misunderstanding. Likewise, Ortega y Gasset (1959:3) points out that before understanding a statement, people should know which frame is being selected in the communication. As for knowledge schema, it refers to participants’ “expectations about people, objects, events and settings in the world” (Tannen & Wallat, 1987:60). Compared with interactive frame, which is dynamic, knowledge schema is rather static. In order to understand the meaning in discourse, people should “fill in unstated information. 政治大. which is known from prior experience in the world” (Tannen & Wallat, 1987:60).. 立. 2.1.4.2. Interaction of Frames and Schemas. ‧ 國. 學. It is necessary to elaborate how frame and schema can interact with each other. ‧. and influence the way people communicate. Tannen & Wallet (1987) discuss interaction of frames and schemas by giving Tannen’s own experience as example.. sit. y. Nat. io. al. er. One time, she is talking on the phone with a male friend, her friend suddenly yells. v. n. “YOU STOP THAT!” She knows that this interjection is not indicating to her, but the. Ch. engchi. i n U. friend’s dog. She can distinguish such reference because her friend uses a specific prosodic representation which is only used to address the dog. Besides, “YOU STOP THAT!” is more likely to appear in the frame “disciplining a pet” rather than “talking to a friend.” She can also infer that her friend may talk to a misbehaving child owing to her knowledge schema since she knows that the friend has a child. These “expectations” about what the friend might be speaking to indicate that frames and schemas interact with each other then influence people’s comprehension. Another example about interaction of frame and schema shows that mismatch 8.

(25) of knowledge schema will lead to frame shifting. In Tannen & Wallet’s (1987) study, they investigate a pediatrician’s register shifting during the examination, and they find out that frame shifting can be triggered by the mismatch of knowledge schemas. When the pediatrician is reporting some typical symptoms of cerebral palsy to the camera what she thinks that is normal for people who have the disease, the patient’s mother who has no professional medical knowledge schema will interrupt her, asking whether it is symptoms of illness. The pediatrician adopts register shifting, shifting. 政治大. from technical terms to simple words, to explain the ordinary symptoms to the mother.. 立. It is this shifting from examination register to consultation register that exemplifies. ‧ 國. 學. the matching between interactive frame and knowledge schema.. ‧. 2.1.5. Evidence of Frame 2.1.5.1. The Peer-story Experiment. sit. y. Nat. io. al. er. It seems like frame interplays with people all the time, but it is hard for people. v. n. to be conscious of its existence. A film-telling experiment reveals the existence of. Ch. engchi. i n U. frame. Chafe (1977) conducts an experiment by showing a six-minute short film to a group of participants, and asks them to tell the content of the film to someone. In this experiment, it is found that the ways which the participants organize and describe the film are varied. Besides, some informants change the content of the film. Tannen (1993) contends that how participants organize their narrations and why they change the content of the film are related to their structure of expectation (or frame). Also, in comparing the oral narratives of Greek and Americans participants, Tannen notices the ways they do the narratives are frame determined which is culturally determined. 9.

(26) For instance, it is showed that American participant are more aware of being a film viewer than Greeks, many of them give criticism to the film. In addition, in this film-telling experiment, participants are involved in interview that they are aware of being recorded. Both participant’s expectation about the film and the expectation of being as a film viewer would influence how the film is to be perceived and verbalized. Using the concept of frame to discuss this issue, several frames are interplayed with each other in this film-telling event, includes. 政治大. film-telling frame, interview frame, storytelling frame, and so on. For example, some. 立. participants mention that there is no dialogue in this film that they are conscious of. ‧ 國. 學. the film-telling activity. When the film-telling frame is salient, participants may. ‧. mention or comment on film viewer’s point of view when do the narrations, such as mentioning the sound track, sound effect, and verisimilitude in the film. They. sit. y. Nat. io. al. er. apparently feel that they should talk about the point of the film or give some. v. n. comments when telling about a movie. This is owing to the frames in their mind that. Ch. engchi. i n U. direct them what they should see or what they should talk about. Investigating the film-telling narrations, Tannen (1993) categories several linguistic phenomena which can be regarded as evidences of frame, includes omission, repetition, hedges, and negatives. Take omission for example, when telling the film, some participants may mention or emphasize specific details (such as a goat in the film), while the others, due to their culture backgrounds, would not. Using the concept of frame to explain these differences, “a man with a goat in the country” is a common scene in Greek’s frame which is viewed as unmarked, so it is less important to report. 10.

(27) While in American’s frame, seeing a man with a goat may not so common; the scene is marked and is worth-mentioning. Frame acts like a filter which operates a selection process, and determines what should be emphasized and what could be omitted. Moreover, take evaluative language for example, when describing the pear picker in the film, some Greek informants portray him as a “tall” pear picker, while American participants do not mention this characteristic. This may be owing to cross cultural differences in frame: Americans may regard the pear picker as average height. 政治大. according to their frame, but for Greeks, the height of the pear picker is not so. 立. common in their frame. Tannen (1993) concludes that frame can not only influence. ‧ 國. 學. the way people what and how to perceive things, but also the way they verbalize. ‧. things which they perceive. 2.1.6. Frame Shifting. sit. y. Nat. io. al. er. 2.1.6.1. Reframing the Frame. v. n. Tannen (1986) suggests that frame can be reframed. She gives a personal. Ch. engchi. i n U. example about how she reframes the frame to avoid confrontation. One time when she is lecturing, a couple sitting at the front of the room keep interrupting her. She thinks the better way to deal with the situation is not to start a battle frame, debating and arguing with the couple, but to jump out of the frame and say, “There are seventy-five people in this room. You’ve already asked a lot of questions; let’s give some of the others a chance (Tannen, 1986:86-87).” She reframes the frame to shift her attention from the couple to the other audiences. This effectively blocks the interruption from the couple, and more importantly, avoid confrontation. 11.

(28) 2.1.6.2 Register Shifting To understand how people manipulate with frames, it is needed to investigate linguistic forms as evidence to discuss people’s underlying expectation. Goffman (1974) proposes that linguistic forms can be used as cues or markers to see how frame is manifested. In Tannen & Wallet’s (1987) study, they investigate a pediatrician’s register 1 shifting during the examination as the linguistic evidence of frame shifting. In the. 政治大. pediatric interaction, the pediatrician has to deal with three addressees: the patient. 立. which is a little girl, the girl’s mother, and the future audience of the recorded. ‧ 國. 學. videotape who may watch it for research usage. It is found that three kinds of registers. ‧. are applied by the pediatrician in order to talk appropriately to different addressees. First of all, the pediatrician uses motherese to address the little girl, and she. sit. y. Nat. io. al. er. applies teasing register that sounds like they are playing games. She exaggerates shift. v. n. in pitch, lengthens vowel sounds, and with a smiling facial expression to get the little. Ch. engchi. i n U. girl’s attention. Afterwards, when she explains the examination to the mother, she shifts to conversational register. She uses simple words to explain the result of examination which are easy to understand for normal people. And then, she applies reporting register when she reports the findings to the camera. She uses many technical terms with flat intonation and applies third-person pronoun “her” to refer to the girl when doing the report. During the examination, the pediatrician shifts from. 1. Ferguson (1985) defines the term register as “variation conditioned by use;” that is, people use particular lexicons, syntactic structures, prosodic aspects depending on the contexts, so that such linguistic features are regarded as “appropriate.” 12.

(29) one register to another; register shifting in these speech activities illustrates frame shifting. Tannen & Wallet (1987:65) propose that “each of the frames entails addressing each of the audiences in different ways.” They also discover that sometimes frame can’t be embedded with each other: the pediatrician cannot apply both examination frame and consultation frame at the same time. Tannen and Wallet (1987) suggest that avoidance of such conflict verifies the existence of frame. 2.1.7. Interaction of Frames 2.1.7.1. Yield or Resist the Frame. 立. 政治大. There are two ways when dealing with the frame set by others: to yield the. ‧ 國. 學. frame, or to resist the frame.. ‧. Tannen (1986) gives an example to describe the two ways. A car approaches an intersection when two pedestrians, a women and a man, want to cross it. The driver. sit. y. Nat. io. al. er. stops the car and signals them to go first. In this situation, the frame set by the driver. v. n. is that it is the driver’s credit to let the pedestrians cross first. While in fact, it is not. Ch. engchi. i n U. the driver’s magnanimity since the law requires drivers to do so. The women who yields the frame crosses the road quickly to show her appreciation to the driver. In contrast, the man signaling the driver to go first resists the driver’s frame showing that it was his credit to let the car go first. Tannen points out that when dealing with the frame which is set by others, people can choose to yield or resist the frame. 2.1.8. Levels of Frames Goffman (1981) proposes the notion of lamination of experience that not only frame can shift from one to another, but it can be embedded within another. In Hoyle’s 13.

(30) (1993) study, she investigates the sportscasting speech activity which constructed by three 8- and 9-year-old boys when they play games. She finds out that one frame can be embedded with other frames which proves that there can be levels of frames existed in one activity. Tannen (1993) proposes that frames have levels. According to Tannen, there may be more than one frame intertwined or overlapped with each other. Tannen also suggests that a speech event is composed of a larger “context” and the “content” of. 政治大. communication. In the peer storytelling experiment, she proposes that the larger. 立. context is the interview frame that participants are aware of being recorded so they. ‧ 國. 學. have certain expectations about how to act in the interview. Interview frame is the. ‧. overriding frame of this speech event, and there are other subordinate frames (such as the film-telling frame). In other words, the former is that part of the frame on higher. sit. y. Nat. io. al. er. hierarchy, while the latter is on lower hierarchy. When participants try to retell the. v. n. film, their expectations about the story in the film and being a film viewer come into. Ch. engchi. i n U. play. Therefore, Tannen suggests that frames have levels which include overriding frame or other subordinate frames, and that they may intertwine and interact with each other. The illustration of overriding frame and subordinate frame is presented in Figure 1.. 14.

(31) The larger context Overriding frame. The content of communication Subordinate frame. The content of communication Subordinate frame. Figure1. Overriding frame and subordinate frame. 2.2. Discourse Structure. 立. 政治大. ‧ 國. 學. Since stylistic variation for frame by discourse structure is analyzed in this study, including both narrative structure and conversational structure, a review of. Nat. y. ‧. these two types of discourse structure is given below.. er. io. sit. 2.2.1. The Structure of a Narrative. al. Labov (1972:359) defines narrative as “method of recapitulating past. n. v i n Ch experience by matching a verbal sequence to the sequence of events which e n gof cclauses hi U actually occurred.” He proposes the structure of narrative; as he asserts, a fully-formed narrative structure is composed of six elements: abstract, orientation, elaboration (complicating action), evaluation, solution or resolution, and coda. To describe the elements of narrative structure in an easier way, Labov regards them as a series of answers to underlying questions:. 15.

(32) a. Abstract: what was this about? b. Orientation: who, when, what, where? c. Complicated action (Elaboration): then what happened? d. Evaluation: so what? e. Result: what finally happened? 1. Abstract According to Labov (1972:363), abstract is the part which “encapsulates the point of the story.” 2. Orientation. 政治大 situation of the story. It is 立 used to “sketch the kind of thing that was going on before Orientation is the part which sketches the time, place, persons, activity or. ‧ 國. 學. the first event of the narrative,” and it is always placed at the beginning of the narrative (Labov, 1972:364).. ‧. 3. Elaboration (Complicating action). y. Nat. al. er. io. sit. After the orientation, the narrator elaborates the narrative in a more detailed. n. way, namely, giving more (and more detailed) information when doing the elaboration.. Ch. engchi. i n U. v. 4. Evaluation Evaluation is defined as “the means used by the narrator to indicate the point of the narrative” (Labov, 1972:366). Labov compares a narrative of vicarious experience and a narrative of personal experience, and proves that the latter is an evaluated narrative which involves narrator’s self-aggrandizement; that is, the narrator uses subjective view to describe the experience.. 16.

(33) 5. Solution or resolution Solution is the consequence or outcome of the story; resolution is the final determination or decision of the story. 6. Coda Coda is used to signal that the narrative is finished. For instance, sentences such as “And that was that” or “That was it, you know” can signal the narrative is completed (Labov, 1972:365). Coda includes “general observation” or “the effects of. 政治大. the events on the narrator.” Coda is always placed at the ends of narratives. Labov. 立. finds out that “tense” can also be used to distinguish coda from others. The illustration. y. er. io. sit. Nat. And you know that man who picked me out of the water? He’s a detective in Union City And I see him every now and again.. ‧. ‧ 國. 學. is showed below.. al. The italics in these examples show that past tense and present tense are applied. n. v i n C h the change ofUthe tense from past to present in coda. As Labov (1972:365) mentions, engchi can “bring the narrator and the listener back to the point at which they entered the narrative. 2.2.2. The Structure of a Conversation According to Sacks, Schegloff & Jefferson (1974), the structure of conversation can be divided into four major parts, including opening, body (i.e., middle), pre-closing, and closing. Opening is the part that conversational exchanges are initiated. This part also establishes the relationship between interlocutors. Body, 17.

(34) which is so-called middle of the conversation, is the part that topic is developed and discussed by interlocutors in conversation. A pre-closing can signal that the conversation is near the end. Closing is the part that conversational exchanges are finished. 2.2.3. Components of Communication According to Hymes (1988), communication is composed of several components which includes genre, topic, purpose, setting, interlocutors, message form,. 政治大. message content, act sequence, rules of interaction, and rules of interpretation. Each. 立. of the components is illustrated below.. ‧ 國. 學. 1. Genre. 2. Topic. ‧. Genre refers to type of event, such as chatting, debating, lecturing, and discussing.. sit. y. Nat. io. al. er. Topic refers to the referential focus, ranging from daily affairs, through social. n. issues, to professional talks. 3. Purpose. Ch. engchi. i n U. v. Purpose indicates the function of the communication, such as gaining approval and showing repression. 4. Setting Setting includes the factors of time, location, and physical aspects of situation. 5. Interlocutors Interlocutors include speaker and hear. Both the speaker and the hearer carry social features (e.g. gender, age, ethnic background, race, educational level, and 18.

(35) interpersonal relationship), and there is always a social distance based on power and solidarity between them. 6. Message form Message form includes code (verbal code vs. nonverbal code) and channel (vocal form vs. nonvocal form). 7. Message content Message content includes the surface denotative reference of the verbal form. 政治大. and the underlying intention of the speaker.. 立. 8. Act sequence. ‧ 國. 學. Act sequence refers to the ordering of speech acts in verbal communication. For. ‧. example, in a Q-A sequence, it is expected that the prior speaker’s question would invite the next speaker’s answer.. sit. y. Nat. io. al. er. 9. Rules of interaction. v. n. Sociopragmatic rules such as Cooperative Principle and Politeness Principle are. Ch. engchi. i n U. considered and applied to define appropriateness of verbal behavior during communication. 10. Rules of interpretation Rules of interpretation may influence by individual’s personal frame which helps to perceive and interpret things.. 2.3. Illocutionary Acts Searle (1979) categorizes illocutionary acts into five major types, including 19.

(36) assertives, expressives, directives, commissives, and declarations. First, assertives are speech acts that commit a speaker to the truth of the expressed proposition. Second, directives are speech acts that are to cause the hearer to take a particular action. In addition, commissives are speech acts that commit a speaker to some future action. Fourth, expressives are speech acts that express the speaker's attitudes and emotions. Last, declarations are speech acts that change the reality in accord with the proposition of the declaration.. 政治大. Communicative purposes on pragmatic level are achieved through successful. 立. accomplishment of the illocutionary speech acts, which also help to display frame.. ‧ 國. 學. 2.4. The Cooperative Principle (CP). ‧. Grice (1975:45-46) proposes The Cooperative Principle to explain how people. y. Nat. io. sit. act in a cooperative way during communication. The four maxims and their. n. al. er. submaxims of CP are given below.. i n U. v. 1. Maxim of Quantity a. Make your contribution as informative as is required. b. Do not make your contribution more informative than is required. 2. Maxim of Quality a. Do not say what you believe to be false. b. Do not say that for which you lack adequate evidence. 3. Maxim of Relevance a. Be relevant. 4. Maxim of Manner a. Avoid obscurity of expression. b. Avoid ambiguity. c. Be brief (avoid unnecessary prolixity). d. Be orderly.. Ch. engchi. 20.

(37) 2.5. Stylistic Variation 2.5.1 Notions of Stylistic Variation Holmes (1992:245) proposes that style is the way people talk, that is, “the choice of code.” Holmes suggests that stylistic variation refer to language variation which is influenced by situational factors, such as addressee, setting, task, and topic. Stylistic variation can be influenced by two major factors, one is addressee, and the other is context. Addressee is an influential factor which affects speech style, which. 政治大. involves age, gender, and social status. When talk to a close friend or someone who. 立. you are quite familiar with, casual style is used; in contrast, when address an. ‧ 國. 學. unfamiliar person, much formal style is applied. For instance, people in Northern. ‧. Ireland tend to use Standard English, a formal speech style, when they talk to visitors. Another example is the “baby-talk” style; people tend to use baby-talk words such as. sit. y. Nat. io. al. er. doggie and with high pitch sound when they talk to young children.. v. n. Formality of the context is also one of factors which could determine choice of. Ch. engchi. i n U. style. For instance, business meeting and conference are much formal settings, so a formal speech style is applied. Wardhaugh (2010) also suggests that stylistic variation is influenced by the circumstance. For instance, people apply formal style of speaking when they are in ceremonial occasion, but they used informal style when conducting daily communication. Wardhaugh believes that certain stylistic features are applied for specific occasions. In addition, Holmes (1992) regards register as a kind of style and it is function-oriented. Register is a kind of jargon that is used by a group of people who 21.

(38) have common specialties or interests. For instance, the language used by engineer and athlete are registers. Wardhaugh (2010) also thinks that register is language that used by certain occupations or groups, but he distinguishes register form jargon that he thinks the term jargon may have some negative meaning. Last, Holmes and Wardhaugh both believe that stylistic features can reflect the context and the identity of speaker. Hausenblas (1993:52) regards style as “means that take part in the construction. 政治大. of communications described as linguistics or verbal.” Hausenblas believes that. 立. communication has layers. The notion of linguistic means is introduced that it is the. ‧ 國. 學. basic layer for construction of communication, including lexical and syntactic devices. ‧. which denote the basic meanings. Paralingual means are phonetic features that are parallel to linguistic means. Thematic means are the higher layers which are themes of. sit. y. Nat. io. al. er. the communication. For instance, the theme of condolence is to show grief, while. v. n. there are different linguistic means could use to express the grief. The highest layers. Ch. engchi. i n U. are textual means, which help to supplement the subordinate means and “combine parts of discourse into a whole” (Hausenblas, 1993:54). 2.5.2. Code-switching as a Strategy of Stylistic Variation 2.5.2.1. The Definition of Code Bernstein (1971) raises the notion of code, proposing that code is the symbolic system of a language; different languages have different code systems. However, the notion of code is not just symbols of a language; it also includes symbols of varieties within one language, such as dialect. Versehueren (1999) regards code as a kind of 22.

(39) language variety which is systematic and distinguishable; it may relate to social status, functional need, setting, or geographical factors. According to Wardhaugh (2010:84), “code” is a neutral term, unlike language, dialect and style, which are inclined to arouse emotion. He suggests that code is “any kind of system that two or more people employ for communication.” When people speak, they need to determine what language, dialect, style, or register they will employ; that is, what particular code they will choose. It is unusual for a speaker to. 政治大. use only one kind of code to communicate with other people. Therefore, code. 立. selection is an unavoidable process when people conduct verbal communication. To. ‧ 國. 學. sum up, the definition of code is “a system used for communication between two or. ‧. more parties,” ranging from language and dialect to style, register, and verbal forms of any other kinds (Wardhaugh, 2010:98).. sit. y. Nat. io. al. er. 2.5.2.2. The Definition of Code-switching. v. n. The usage of the term code-switching is first appeared in Vogt’s (1954) work. Ch. engchi. i n U. “Language Contacts.” Many scholars proposed that code-switching can be divided into two categories, one is cross-language code-switching; the other is within-language code-switching. 1. Cross-language code-switching Cross-language code-switching refers to shifting between languages. Gumperz (1982), Heller (1988), Myers-Scotton (1993), Timm (1993), Grosjean (1995), and Milroy & Muysken (1995) all propose that code-switching is the shifting between different languages in the same communication. 23.

(40) 2. Within-language code-switching Most of the researches discuss only the code-switching phenomena found in bilingualism/multilingualism and bilingual/multilingual societies. However, some scholars point out that code-switching should include the notion of within-language code-switching, such as style shifting and register shifting. Hymes (1974), Kirschner (1984), Romaine (1995), Wardhaugh (2010), and Thompson (2011) all regard code-switching as a form of style shifting.. 政治大. Woolard (2004) suggests that code-switching is the act of an individual who. 立. uses more than one varieties of a language within the same speech event or. ‧ 國. 學. communication. He proposes that “If only one code is behaved necessary to get the. ‧. communication job done, then the use of more than one needs explanation.” As Labov (1972) and Wardhaugh (2010) contend, there is no single-style speaker; people often. sit. y. Nat. io. al. er. switch codes in accordance with topic, participant, setting, which lead to different. v. n. styles. Both Woolard (2004) and Auer (1995) suggest that linguistic choice is. Ch. engchi. i n U. function-oriented. To be specific, different forms could convey different functional meanings. To sum up code-switching is the shifting between languages, dialects, styles, and registers. In order to speak appropriately or to achieve the intended communicative goal, it is inevitable that a speaker chooses particular linguistic forms. 2.5.2.3. Linguistic Aspect of Code-switching In this section, studies which categorize linguistic aspects of code-switching are reported, including Poplack’s (1980) and Muysken’s (2000) works. 24.

(41) 1. Inter-sentential Switching, intra-sentential switching, and tag switching Poplack (1980) classifies code-switching into three categories according to the syntactic structure: related inter-sentential switching, intra-sentential switching, and tag switching. First, the phenomenon of inter-sentential switching occurs between sentences or clauses. Speaker considers addressee’s language competence or other factors, thus code-switches and employs sentence or clause as linguistic unit of code-switching.. 政治大. Next, intra-sentential switching is the switching within one sentence, which. 立. involves intra-sentence and intra-clause switching. For instance, speaker may. ‧ 國. 學. code-switch at the beginning, middle, or end of a sentence in order to achieve the. ‧. discourse function of emphasis. According to Tay (1989), intra-sentential switching is also termed as “code-mixing,” which is the switching within the same sentence and. sit. y. Nat. io. al. er. most of them are words or phrases.. v. n. Third, tag switching is inserting a tag of distinctive code to a sentence or clause. Ch. engchi. i n U. in order to achieve the discourse function of emphasis. It also includes the insertion of discourse markers, such as “OK,” “all right,” “well.” Tag switching may occur at any position of a sentence. 2. Linguistic devices of code-switching Muysken (2000) proposes three linguistic devices of code-switching: insertion, alternation, and congruent lexicalization. Code-switching of insertion type takes one of the two codes being the matrix language which determines the syntactic structure, with the other only inserts a single 25.

(42) word or constituent into the sentence structure. Code-switching of alternation type refers to the usage of two matrix languages within one single utterance; that is, both languages involved are matrix languages, and syntactic structures from both languages are used. For code-switching of congruent lexicalization type, “the grammatical structure is shared by languages A and B, and words from both languages A and B are inserted more or less randomly” (Muysken, 2000:8). In other words, both languages. 政治大. can be matrix languages; the syntactic structure is alternatively changed within. 立. 學. ‧ 國. utterance.. 2.5.2.4. Functional Aspect of Code-switching. ‧. 2.5.2.4.1. Functions of code-switching. Many scholars have discussed the functions of code-switching. Auer (1995). sit. y. Nat. io. al. er. proposes eight major functions of code-switching which include: reported speech,. v. n. change of participant constellation, parentheses or side-comments, reiterations (so. Ch. engchi. i n U. called translation, repetitions, or recycling), change of activity type, topic shift, puns or language play, and topicalisation. Myers-Scotton (1993) propounds four motivations which cause speaker to employs code-switching, including emphasizing personal style and sociopragmatic meaning, using as discourse marker, using for convenience of expression, and using to fill the lexical gap. McClure (1977) proposes that more than often, code-switching serves to obtain hearer’s attention, to emphasize the manner of speaking, to emphasize topic or 26.

(43) personal opinion, to reduce ambiguity, to shift the manner of speaking, to conduct topic shifting, to invite new participants, and to show interest to the topic. 2.5.2.4.2. Situational, metaphorical and conversational code-switching. Gumperz (1982) regards code-switching as a kind of social phenomena that people’s psychological thinking is involved. He proposes that code-switching can imply speaker’s motivation, and it can also label the change of relationship between interlocutors. Through such phenomenon, speaker can confirm and enhance the. 政治大. relationship with their interlocutors, so they can achieve their communicative. 立. purposes easier.. ‧ 國. 學. Blom & Gumperz (1972) investigate the phenomena of code-switching in the. ‧. Norwegian village of Hemnesberget. They find out that Norwegians switch between standard language and dialect in order to convey certain social information. They. sit. y. Nat. io. al. er. propose that code-switching is influenced by social, situational, and affectional factors,. v. n. based on which they categorize code-switching as situational code-switching and. Ch. engchi. i n U. metaphorical code-switching. Later, Gumperz (1982) proposes another category－ conversational code-switching. 1. Situational code-switching Situational code-switching is influenced by many situational and social factors, such as types of events (such as debating, chatting, and reporting), settings (such as class, workplace, and home), and interlocutors (such as advisor, family member, and boss). When one of these factors changes, it may cause speaker to do code alternation. For instance, Blom & Gumperz (1972) investigate the phenomena of code-switching 27.

(44) in Norway and discover that teachers tend to speak standard language－Bokmål when do the lecture, but use dialect－Ranamål while chatting or discussing with students. This proves that social and situational factors cause the emergence of code-switching. 2. Metaphorical code-switching Metaphorical code-switching is used to convey certain implicit social meaning. Topic is the crucial communicative factor which arouses metaphorical code-switching. That is, if situational factor changes, speaker would chooses an established code in their. 政治大. mind to convey implications or motivations. For instance, Blom & Gumperz (1972) find. 立. that bank staffs in Norway tend to use standard language, Bokmål, when they discuss. ‧ 國. 學. business affairs; but they use the dialect, Ranamål, when they talk about daily affairs.. ‧. Such linguistic choice reveals that when people discuss formal or serious topic, they tend to use standard language, while when they discuss casual or personal affairs, they tend to. sit. y. Nat. io. al. er. use dialect. It is shown that topic is an important communicative factor which influences. v. n. metaphorical code-switching. Rickford & McNair-Knox (1994) also emphasize that topic. Ch. engchi. is an influential factor which leads to code-switching.. i n U. 3. Conversational code-switching According to Gumperz (1982:59), “conversational code switching can be defined as the juxtaposition within the same speech exchange of passages of speech belonging to two different grammatical systems or subsystems.” He classifies the functions of conversational code-switching into six subcategories: quotation, addressee specification, interjections, reiteration, message qualification, and personalization versus objectivization. However, McClure & McClure (1988) propose that conversational code-switching is a 28.

(45) subcategory of situational code-switching, with both being caused by the factor of social identification. 2.5.2.4.3. We Code and They Code Gumperz (1982) adopts anthropological view of code-switching, and regards code-switching as a kind of social phenomena. He believes that the usage of code-switching is motivated by certain social purpose, and that it is also accompanied with the change of the relationship between the interlocutors. Therefore, he proposes the. 政治大. notion of we code and they code. As Gumperz suggests, code-switching is the shifting. 立. between the ethnically minority language and majority language, with the minority. ‧ 國. 學. language being regarded as we code (i.e. the in-group language that is associated with. ‧. familiarity, solidarity, or informal activities), while the majority language being they code (which is associated with formal, out group relationship). Therefore, code-switching can. sit. y. Nat. io. al. er. be used to reveal interlocutors’ identities and relationship.. v. n. Gal (1979:116) provides a classic example in which a Hungarian-speaking woman. Ch. engchi. i n U. in Austria uses the majority language, German, to show retort to her husband. In this case, the woman shifts from minority language (i.e. we code) to majority language (i.e. they code) in order to enlarge the social distance between her and her husband, and thus cause the effect of repression.. 2.6. Summary The above studies indicate that stylistic variation of every kind may serve as a tool to manifest frames on different functional layers. It is based on this point that this study is designed. 29.

(46) Chapter 3 Methodology. In this chapter, the research design of this study is described, including data collection, data analysis, and criteria for data classification.. 3.1. Data Collection 3.1.1. Data Sources. 治政 This study is corpus-based. All of the data are 大 from face-to-face spontaneous 立 ‧ 國. 學. conversations, extracted from NCCU Corpus of Spoken Mandarin. In NCCU corpus, conversations last between twenty to forty minutes. Within each conversation, at least. ‧. two interlocutors are involved. All of these conversations address to topics of daily. y. Nat. io. sit. affairs (including jobs, school life, friend’s marriage, and traveling), also some of. n. al. er. them include social issues (such as food safety issues in Taiwan), and also some. Ch. i n U. v. contain professional talks (such as global warming). In addition, the word frequency is based on Academia Sinica Corpus.. engchi. 3.1.2. Sampling Data analyzed in this study are all from face-to-face spontaneous communication between two interlocutors. Also, although effect of gender is not examined in this study, both same-gender and cross-gender conversations are included to secure data balance. In total, seven conversations are analyzed, including three cross-gender conversations, two male-to-male ones, and two female-to-female ones. 30.

(47) Following the same line, the interlocutors in the conversations are of similar age, ranging between twenty-three years old to twenty-five years old.. 3.2. Measurement Owing to the extracted data in this study are all spontaneous spoken data, it is clearer to use “turn” as measurement of linguistic unit to count the frequency of stylistic variation. For instance, in the same turn, when the speaker shifts form high density words to low density words, ignoring how many low density lexical items are, this. 治政大would count separately. wound count once. Furthermore, different linguistic criteria 立 ‧ 國. 學. In addition, only the goals of stylistic shifting are categorized and counted; the sources of stylistic shifting are ignored in this study.. ‧. The results of data analysis are presented in percentage as well as in frequency.. sit. y. Nat. io. n. al. er. 3.3. Criteria for Data Classification 3.3.1. Linguistic Criteria. Ch. engchi. i n U. v. In this study, both lexical and syntactic devices for stylistic variation were examined. 3.3.1.1. Lexical Level On lexical level, it is hypothesized that SVF involves shifting of semantic density of word, word formality, and word frequency. 1. Semantic density of word The measurement of semantic density is determined by how much information conveyed in the lexical item. In this study, semantic density of a lexical item is 31.

(48) divided into low density and high density. Both stylistic shifting form high density words to low density words and that in opposite direction were examined in this study. Linguistic features related to the judgment of lexical semantic density include ellipsis, pro-form, pragmatic particle, discourse marker, technical term & jargon, and syntactic particle. These features are illustrated one by one below. a. Ellipsis. 政治大. Ellipsis is one of the linguistic features which indicates low semantic density.. 立. Ellipsis includes abbreviation, blending, and acronym. In this study, only abbreviation. ‧ 國. 學. and blending are found in the data. Example (1) is an illustration of ellipsis.. ‧. (1) M: ..那你爸媽會煩惱嗎 F: (0)不會..因為我爸媽不會幫他出錢..但是我媽會催他趕快結婚 M: …那你爸媽會煩說以後還要…如果他不..不..就以後沒有什麼工作.. 啊以後結婚還要他幫他出錢..這樣. er. io. sit. y. Nat. al. In example (1), the two interlocutors are discussing a quite serious topic－job.. n. v i n C h from the wordU“煩惱” to an abbreviation “煩,” In this example, the male speaker shifts engchi which is taken as a lexical item of low semantic density. b. Pro-form Pro-form, as a substitute of a full form, the semantic content of which is subtracted, is regarded as a feature of low semantic density words. In this study, pro-form includes pronominalization and substitution, which replace the original noun or verb. The example of pro-form is showed below.. 32.