本論文藉由檢視臺灣政論節目中所使用的語用策略（直接或間接）、語言行為類別、及語言行為目的，探討政治意識型態對於語言行為產生的影響。本研究以Grice（1975）的合作原則與Searle（1969）的語言行為理論作為分析框架，並以Leech（1983）的禮貌原則解釋語用策略及語言行為在不同政論節目中的分布差異。 　　本研究從臺灣主流的政論節目當中，政治光譜兩極的政論節目「大話新聞」（泛綠）與「2100全民開講」（泛藍）採集語料；並以Searle（1965年）的語言行為理論進行分析，總共識別12類直接語言行為和26類間接語言行為。 　　研究結果顯示，（一）語用策略方面，說話者在政論節目中使用間接語言行為的頻率比直接語言行為頻繁。（二）語言行為類別方面，直接語言行為中，排序則為：斷言（Assertive）、表述（Expressive）、指示（Directive）；而間接語言行為中，各類別使用頻率由高至低排序為：表述（Expressive）、斷言（Assertive）、指示（Directive）。（三）在政治意識型態的影響方面，支持執政黨的政論節目行使較多「直接且事實導向」的語言行為，而支持反對黨的政論節目則有較多的「間接且評論導向」的語言行為。（四）「譴責」是政論節中最常使用的語言行為，並且以間接方式表達居多。（五）推論過程中，推論步驟較多的「間接譴責」語言行為在政論節目中較常被使用。（六）「大話新聞」與「2100全民開講」雖偏向不同的政治意識型態，但是它們皆以斷言類（Assertive）或指示類（Directive）的語言行為來包裝，藉以間接達到「譴責」執政黨疏失的目的。 / This thesis investigates the political-ideological influence on speech acts in Taiwan political talk shows by examining the pragmatic strategies (directness and indirectness), the speech act categories, and the illocutionary purposes performed in the talk shows. In this thesis, Gricean maxims (1975) and Searle’s theory of speech acts (1969) are adopted as the analytic frames to study how speech acts are conducted, and Leech’s (1983) notions of politeness are the theoretical basis for explaining the distributional difference of pragmatic strategies. 　　The data analyzed in this study is composed of dialogues extracted from 6 episodes of two political talk shows with opposite stances on political issues, namely DaHuaXingWen (大話新聞), the pan-green talk show, and QuanMinKaiJiang (2100全民開講), the pan-blue one. Following Searle’s scheme of speech acts (1965), this study identifies the illocutionary act of every clause in the data and recognizes 12 types of direct speech acts and 26 types of indirect speech acts in the collected data. 　　The results of quantitative analysis show that, (1) indirect speech act is generally performed more frequently than indirect speech act in political talk shows. (2) The order of frequency in direct speech acts represents as: Assertive > Expressive > Directive; and in indirect speech acts, the order of frequency is: Expressive > Assertive > Directive. (3) In terms of the political-ideological influence, the political talks show supporting the ruling party (QuanMinKaiJiang) performs direct fact-orientated speech acts more, while the show that standing in the opposition to the ruling party (DaHuaXingWen) has more indirect opinion-oriented speech acts. (4) Condemnation is the most often used illocutionary act in political talk shows, and mainly done indirectly. (5) Indirect condemnations with longer length of inferential process are preferred in political talk shows. (6) Despite that DaHuaXingWen and QuanMinKaiJiang held different political stance, they share the same pattern of expressing indirect condemnation—wrapping it in speech acts of Assertives or Directives.

人們常因為禮貌或其他因素避免對立的情況發生。然而，異議在我們日常溝通中又扮演了不可或缺的角色。之前，眾多對於異議及其相關語言活動的研究均未曾探究異議內容的本質(對於事實內容的異議或對於議題評估的異議)與異議的建構有何關係。此外，台灣鮮少研究社會因素對異議建構方式的影響。基於上述不足，本研究旨在探討何種異議(內容異議或評估異議)在日常生活中較常出現，不同異議類別的語言形式與語用策略為何，以及年齡是否會影響異議的數量多寡與建構方式。本研究採用言談分析(conversational analysis, CA)作為研究框架，並以言語行為理論(speech act theory),合作原則(Cooperative Principles)及禮貌理論(Politeness Principles)為理論基礎。 本研究以12份日常交談為語料，進行異議分析。在這12份語料中，8組對話者為同齡(4組年長者，4組年輕者)，4組對話者為跨齡。在分析過程中，先依異議的本質進行分類，進而分析討論異議中所使用的語言形式、語用策略、社會因素(年齡)，以及四者彼此之間的互動。 研究結果顯示，第一，人們使用評估異議的頻率為內容異議的兩倍之多。個人主觀式遠多於社會文化評估的異議。第二，就語言形式而言，在異議的建構中，否定句、預告詞及肯定句(依此順序)的使用頻率高於其他語言形式。然而，語言行式的選擇會隨異議的本質而有所改變。內容異議通常使用直接句型，如否定句與肯定句；評估異議則平均使用直接性的否定句與間接性的預告詞。第三，就語用策略而言，更正、解釋與質疑(依此順序)的使用頻率高於其他語用策略。語用策略的選擇亦隨異議本質的不同而有所改變。超過一半的內容異議使用更正策略，但在評估異議中，更正、解釋與質疑的使用頻率相當。第四，在評估異議中，在各個語用策略中，語言形式的種類比內容異議多。這個結果影射著評估異議對面子的威脅程度可能比內容異議來得嚴重。因此，在進行評估異議時，語言形式與語用策略的挑選用必須格外注意。第五，年齡與異議的建構有顯著的相關性。同齡組比跨齡組更容易產生異議。最後，在異議中，聽話者的角色比說話者的角色更具有影響力。 / Although people try to avoid opposition for the sake of politeness or other reasons, disagreement, which may threaten interpersonal relationship and the success of communication, is inevitable in our daily life. Previous studies on disagreement (including dispute, argument, conflict, etc.) have not probe into the nature of the referential content—whether it is content-based (in this study, C-disagreement) or evaluation-based (in this study, E-disagreement), and the influences of social factors on disagreement have rarely been examined in Taiwan. Therefore, the purposes of this study are to see what type of disagreement are most likely to occur in daily conversations and to examine whether age is an influential factor on linguistic choices for in disagreement in Chinese society. This study uses the framework of conversational analysis (CA), and adopts speech act theory (Austin, 1962; Searle, 1975), Cooperative Principles (Grice, 1975) and Politeness Principles (Brown and Levinson’s, 1978, 1987; Leech, 1983) as the theoretical foundations. 12 conversations by speakers of 8 same-age groups (including 4 old groups and 4 young groups) and 4 cross-age groups were examined for disagreement. Related data are categorized, analyzed, and discussed by types of disagreement, linguistic markers, pragmatic strategies, social variable (in this study, age), and the interaction among the four. The results of the data analyses show, first, people adopt nearly twice more E-disagreement than C-disagreement; moreover, E-disagreement based on personal judgment emerges more often than E-disagreement based on socio-cultural evaluation. Second, for linguistic markers, negation, pre-announcement marker, and affirmative (in this order) are adopted more in disagreement. However, preferences for linguistic markers change according to types of disagreement. In C-disagreement, direct syntactic markers, such as negation and affirmative, are used more frequently than the others; however, in E-disagreement, direct negation (syntactic) and indirect pre-announcement (lexical) are used with equal frequencies. Third, among pragmatic strategies, correction, account, and challenge (in this order) are adopted more frequently than the others. The usage of pragmatic strategies varies with types of disagreement. In C-disagreement, correction is highly adopted. But in E-disagreement, correction, account, and challenge are used with equal percentages. Fourth, the fact that more varieties of linguistic markers are used in each pragmatic strategy in E-disagreement than in C-disagreement may imply impoliteness, since face-threatening force is more serious in E-disagreement than in C-disagreement, which, in turn, indicates that more careful manipulation is needed in using E-disagreement. Fifth, age is influential in disagreement. More disagreements are found in the same-age groups than in the cross-age groups. Last, the hearer’s role is found to be more influential than the speaker’s role.