中文形容詞呈現之性別概念

全文

(1)國立台灣師範大學英語學系碩士論. 文. Master Thesis Department of English National Taiwan Normal University. 中文形容詞呈現之性別概念. Gender Representations Manifested in Chinese Adjectival Compounds. 指導教授：蘇席瑤博士 Advisor: Dr. Hsi-Yao Su 研究生：張建斌 Student: Chien-Pin Chang. 中華民國九十七年六月 June, 2008.

(2) 摘要. 在中文裡，當帶有某種性別特徵之詞素與其他字結合為複合詞，此複合詞通常帶有此性別特徵。但事實上，部份複合詞並非如此呈現。此論文旨在探討此種中文複合形容詞中所呈現之不對稱現象與其所代表之意涵。此研究選擇「狂」、「豪」、「柔」、「溫」、「正」五個詞素所構成之二十三個複合形容詞作為研究對象。研究結合了語料庫與問卷兩種方法，同時採用此兩種方法使得研究更為全面性。此研究選擇網路作為語料庫來源，Google 則為語料蒐集之工具。複合形容詞與男性或者女性共現之語料（共蒐集 690 筆）為語料庫研究分析對象。此外，共有 108 名男性與 127 名女性參與問卷研究。根據年齡，受試者之後也被區分為兩組。受試者在問卷的三個部份中，分別被要求判斷每個複合形容詞之性別連結偏向、表達對於複合形容詞之感受程度、提供這些複合形容詞之代表人物。語料庫研究之結果發現，複合形容詞之性別偏向可能受到多種原因影響，例如形容詞本身之語義偏向、網路及媒體之影響、社會風氣之改變。另外，傾向女性之複合形容詞時常帶有負面意涵。相同的複合形容詞與男性或女性連結，亦可能產生不同的意涵。這些結果某種程度上反映出一般人所抱持之刻版印象。在問卷方面，結果顯示刻板印象仍深入影響人們對複合形容詞之看法。對於複合形容詞與性別之連結，女性與年輕人看法較顯彈性。另外受試者大多偏好男性傾向之複合形容詞與排斥女性傾向之複合形容詞。然而，女性雖排斥女性傾向形容詞，但對此類形容詞接受度仍較高。此現象亦反映了女性在語言中的弱勢地位。問卷中不同年齡層之受試者選擇偏好亦有其差異，而此差異也反映了某種程度之社會變遷。在代表人物部份，女性受試者較常為某些被偏好的男性傾向複合形容詞提供女性作為代表人物；兩年齡層相互比較亦可發現男女意象之改變。比較兩種研究方法結果顯示網路與大眾媒體可能是造成兩種研究結果差異之主因。期望藉由中文複合形容詞與男女之共現研究，我們能更深入的探知詞義反映社會價值之現象。 i.

(3) ii.

(4) ABSTRACT. In Mandarin Chinese, when a morpheme biased towards a given gender combines with another word to form a compound, this compound tends to maintain the gender connotation of the biased morpheme.. However, some compounds do not. reveal this tendency. This thesis aims to investigate such unequal phenomena found in some Chinese gendered adjectival compounds and their linguistic effects and connotations.. We focus on the following five characters and their adjectival. compounds (23 in total): kuang (狂), hao (豪), rou (柔), wen (溫), and zheng (正). This thesis combines two approaches, including a corpus-based approach and questionnaire collection. The two approaches complement each other. The search engine “Google” was chosen as the tool to access our corpus—the web. Tokens in which a compound was used to modify people were included for analysis, and there were 690 tokens collected in total.. In addition, we recruited 108. men and 127 women, and the participants were further divided into two age groups. The participants were asked to judge a compound’s likelihood to collocate with men and women, express their feeling about a compound when it is used to describe them, and provide a representative figure for each compound. The results of the corpus-based study show that each adjectival compound has its collocational preference, which may result from diverse factors, including the original semantic features of the compounds, the influence of the Internet and mass media, social trends, etc. In addition, it is observed that female-inclined compounds carry negative meanings more often.. When a compound co-occurs with different genders,. different meanings and connotations might be generated.. The results to some degree. reflect the stereotypes people generally hold. The questionnaire reveals the great influence of gender stereotypes on people’s iii.

(5) perceptions of the compounds.. The female participants’ and the junior group’s. answers are also found to be more flexible in that the boundaries between masculine and feminine compounds are not as absolute as before.. Additionally, it is found that. male-inclined words are usually preferred, while most female-inclined words are dispreferred. However, women are adapted to these words more often.. Women’s. disadvantageous language position is observed. The different preferences of the junior group and the senior group also suggest the change of social trends. Furthermore, it is also discovered that the female participants are more likely to provide female representative figures for a preferred male-inclined compound. Through the comparison between the junior group’s and the senior group’s answers, it is also found that the images of men and women are changing.. Finally, the. discrepancies between the two approaches display the influence of the Internet and mass media. It is hoped that by observing some gendered adjectival compounds in Mandarin Chinese, we can gain a deeper understanding of how lexical meanings can reflect social values.. iv.

(6) 謝辭在放空了很長一段時間之後，我才有力氣再回來回想已經發生過的這一切！寫論文，真的是一段看不到盡頭的路，很難知道自己何時才會觸到終點！最殘忍的事情莫過於：在這條路上四處埋著「此路不通、退回起點」的地雷，常令人不禁發出「人生就像大富翁遊戲般鬧劇」的悠悠嘆息！在整段路程的有著極大的艱辛，很容易就因為巨大孤獨來襲而逃避與迷失。然而，我是個幸運的孩子，在這一路上，總是有著一些面孔，時而嚴厲、時而溫柔的催促著我，讓我能大步勇敢走著！而終於，我也走到這裡了！如果我從未得到這些有力的扶持，我這個傻孩子的路一定走得更蹣跚！因此，我要獻上我最由衷的謝意給這些我最堅強的後盾！首先，我要誠摯的感謝我的指導教授—蘇席瑤老師。老師是個內外兼俱的新時代女性，在學生中有著無與倫比的高人氣！認識老師，是從碩二擔任老師助理開始。還沒正式和老師會面之前，我還偷偷的擔心很「正」1、又很有「氣質」2 的老師是不是難以親近，但這個錯誤想法很快的就被老師「豪邁」 3 的笑容破除了！老師十分平易近人，是個陽光美女，對於我的任何大小疑問總是很樂意與我討論，給予我中肯的建議、全力的協助、不吝嗇的鼓勵。從老師身上，我學到了許多，其中不單單只是學術方面，還包括了對人、對事的態度。如果沒有老師的指導，我想我的論文是不可能完成的。我也要感謝我的兩位口試委員：張妙霞老師與陳振寬老師。我的論文的初步構想，是在妙霞老師的「章法分析」課程產生的。妙霞老師一直都十分樂意給我許多建議，鼓勵我從各種不同的角度切入，使我對於論文不斷產生新的想法，也讓我找到努力的方向。我論文有完成的一天，妙霞老師絕對是功不可沒。而在先前與我素未謀面的陳振寬老師，願意在百忙之中撥冗擔任我的口試委員，也讓我振奮與感激不已。自第一次以電話和老師直接接洽時，陳老師就對我展現十分和善的態度，也對我論文表達強烈的興趣。而在口試中老師提出很多有用的建議，另一方面也慷慨的給予我這小毛頭讚美與勉勵，使我有受寵若驚之感！接下來我想感謝師大英語系的其他曾經教導過我的老師：李臻儀、林雪娥、陳純音、馮和平、黃自來、謝妙玲、謝國平等 4 。因為有這些老師的細心教導，扎實我的專業能力，我才能夠順利完成論文。除了專業能力的教導，我認為師大老師最大特色之一是對於學生無微不至的關懷與照顧。我所看到的所有老師都是以非常真誠的態度面對學生，把學生當作自己子女一般的看待。即使是生活上的小事，老師們都十分願意犧牲寶貴時間在學生身上，這點總是讓我感到溫暖不已。在師大的這三年，我所獲得的比我能想像的遠超過太多了，這也讓我心懷感 1. 我和老師都認為「正」是一種新型態的稱讚，和「正點」比起來比較不會產生負面意涵。氣質，是老師研究的主題之一。 3 記得有次和老師討論論文時，老師說過如果有人用「豪邁」來描述她，她可能會蠻開心的，所以我要在這邊拍個小馬屁！ 4 按筆劃順序排列。 2. v.

(7) 恩！以前的我，總是把師大當作我ㄧ直以來夢想的第一志願；而我現在更要驕傲的說，這是個絕對正確的選擇！因此我也要感謝師大給了我圓夢的機會！另外，我還要感謝我最親愛的同學們：何信昌、沈正嵐 5、施姵如、孫安霖、許展嘉、黃郁欣、蔡貴如、謝孟璇 6 。在苦悶的研究生生涯中，我很感謝有一群總是能帶給我無盡歡笑、讓我再繼續邁進的優秀同學們。我以前常聽說過了大學之後就很難再找到能夠深交的好友，但我很幸運到了研究所還能結識到這群超棒、超有特色的同學。一起度過的歲月，我們有太多的值得細數的回憶，深刻到我想我一輩子都會想念！我要特別感謝何信昌與許展嘉，這兩位和我情同手足的好同學對於我的論文總是提供許多寶貴建議；甚至到了最後的修改階段，他們也費了很多心力仔細的協助我校訂。我也要感謝師大英語的其他同學、學長姊與學弟妹 7 。我還要感謝研究所這幾年伴著我的貴人、師長、好友們，包括在高雄大學的師長 8 與一起走過「拓荒時期」 9 的同學、學弟妹們 10 、擁有革命情感的傑青夥伴們 11 、還有一些在我每個其他學生階段結交的多年好友。我還要感謝我應該感謝卻不小心漏掉的人，我想我有太多道不盡的感謝，希望你們都能感受到！最後，我要由衷的感謝總是在背後默默支持我、深愛著我的家人與親戚們。在我最失意，覺得沒辦法繼續走下去的時候，老媽寄來的E-mail 12 、姊姊們打氣的MSN訊息、與老爸的電話，總是讓我再次加滿元氣。謝謝你們一直以來的陪伴，如果我已經達成什麼成就，或者將來有任何的成就，我想都是你們賜予我的。因此我要將這本論文獻給你們──我最深愛的親人！. 5. 沈正嵐是我的好「兄弟」，她即將前往美國就學，特別祝她一切順利(包括她最想要的感情)！按筆劃順序排列。 7 特別感謝林盈萱，總是義無反顧地在我需要幫忙時伸出援手；尤其是在我論文問卷的蒐集上，她幫了我非常大的忙！ 8 特別感謝在西洋語文學系曾經教導過我的彭淑惠、林雅惠、倪碧華、廖招治、陳麗青、曾銘裕、顏淑娟、狄建世、陳英輝、張秀珍老師。 9 我是國立高雄大學第一屆的畢業生，也以作為「高大永遠的老大」為榮。 10 特別是西洋語文學系的學弟妹與吉他社學弟妹。 11 感謝馮國華、李松濤、戴光廷、陳惠名、李德俊、張巍耀、陳長風、顏江龍，你們讓我在研究所課業之外有了很不同於一般人的研究生生活。另外要特別感謝民革的汪叔、楊姊、董哥。 12 我有個很酷的老媽，即使她沒唸過太多書，也可能不是太聰明，但她到了五十多歲還願意為了兒女們學電腦，花個一兩個小時用一指神功打出一小段的E-mail來噓寒問暖。當看到暱稱「小英」的老媽E-mail來信之時，我總是會有幸福破表的感覺！ 6. vi.

(8) TABLE OF CONTENTS Chinese Abstract English Abstract Acknowledgements Table of Contents List of Tables List of Figures. i iii v vii xi xii. Chapter 1 Introduction 1.1 Motivation 1.2 Background 1.3 Research Questions 1.4 Significance of the Thesis. 1 1 2 3 5. 1.5. 6. Organization of the Thesis. Chapter 2 Literature Review 2.1 Gender Studies 2.1.1 Wollstonecraft (1792) 2.1.2 Mill (1869) 2.1.3 Woolf (1929, 1938) 2.1.4 De Beauvoir (1949) 2.1.5 Money et al. (1968), Money and Ehrhardt (1972) 2.1.6 Other Related Studies 2.2 Gendered Language and Gender Biases in the Western Society 2.2.1 Lakoff (1975, 1990) 2.2.2 Schulz (1975) 2.2.3 Tannen (1993, 1994) 2.2.4 Sutton (1995) 2.2.5 Fowler (1991) 2.2.6 Caldas-Coulthard (1995) 2.2.7 Other Related Studies 2.3 Gendered Language and Gender Biases in Chinese 2.3.1 Shih (1984) 2.3.2 Shieh (1997) 2.3.3 Liao (2000) 2.3.4 Lin (2004). vii. 7 7 7 9 9 10 11 12 13 13 14 15 16 17 18 19 20 20 22 23 23.

(9) 2.4. 2.5. 2.6. 2.7. Gender Stereotype 2.4.1 Basow (1992) 2.4.2 Zou (1986) 2.4.3 Visser (2002) 2.4.4 Litoselliti (2006), Caldas-Coulthard (1996) Social Change 2.5.1 Luo (1981) 2.5.2 Lin (1998) 2.5.3 Other Related Studies Collocation 2.6.1 Firth (1957) 2.6.2 Benson et al. (1986, 1997) 2.6.3 Sinclair (1991) 2.6.4 Stubbs (2001a, 2001b, 2002). 25 25 25 26 27 28 28 28 29 30 30 31 32 32. Chapter Summary. 33. Chapter 3 Methodology 3.1 The Corpus-Based Approach (Discourse Analysis) 3.1.1 Adjectival Compounds 3.1.2 Tool 3.1.3 Sampling Method 3.2 Questionnaires 3.2.1 Participants 3.2.2 Questionnaire Design 3.2.3 Procedure 3.2.4 Analysis Procedure 3.3 Chapter Summary. 35 35 35 37 40 41 41 43 44 45 46. Chapter 4 Results and Analyses 4.1 The Corpus-Based Approach (Discourse Analysis) 4.1.1 Kuang (狂) 4.1.1.1 Kuangwang (狂妄) 4.1.1.2 Kuangye (狂野) 4.1.1.3 Kuangfang (狂放) 4.1.1.4 Fengkuang (瘋狂). 48 48 49 49 51 52 52 53 54 54. 4.1.2. 4.1.1.5 The association with men and women Hao (豪) 4.1.2.1 Haoqi (豪氣) viii.

(10) 4.1.3. 4.1.4. 4.1.2.2 4.1.2.3 4.1.2.4. Haomai (豪邁) Haoshuang (豪爽) Haofang (豪放). 4.1.2.5 Rou (柔) 4.1.3.1 4.1.3.2 4.1.3.3 4.1.3.4 4.1.3.5 4.1.3.6. The association with men and women. 4.1.3.7 The association with men and women Wen (溫) 4.1.4.1 Wenliang (溫良) 4.1.4.2 4.1.4.3 4.1.4.4 4.1.4.5. 4.1.5. 4.2. Jiaorou (嬌柔) Roumei (柔媚) Rouruo (柔弱) Roushun (柔順) Wenrou (溫柔) Yinrou (陰柔). Wenhe (溫和) Wenhou (溫厚) Wenwen (溫文) Wenshun (溫順). 4.1.4.6 The association with men and women Zheng (正) 4.1.5.1 Zhengdian (正點) 4.1.5.2 Zhengzhi (正直) 4.1.5.3 Gangzheng (剛正) 4.1.5.4 Zhengjing (正經). 4.1.5.5 The association with men and women Questionnaires 4.2.1 Part One 4.2.1.1 The overall tendency of the 23 adjectival compounds 4.2.1.2 The discrepancies between male and female participants 4.2.1.3 The discrepancies between the junior group and the senior group 4.2.1.4 The interaction between gender and age factors 4.2.2 Part Two 4.2.2.1 The overall tendency of the participants’ preference 4.2.2.2 The discrepancies between male and female participants 4.2.2.3 The discrepancies between the junior group and the ix. 55 55 56 57 58 58 58 59 59 60 61 62 63 63 64 65 66 67 67 68 68 69 69 70 71 71 71 72 73 74 76 77 78 78.

(11) 4.3 4.4. senior group 4.2.2.4 The interaction between gender and age factors 4.2.3 Part Three 4.2.3.1 The overall tendency of the 23 adjectival compounds 4.2.3.2 The discrepancies between male and female participants 4.2.3.3 The discrepancies between the junior group and the senior group The Discrepancy between Two Approaches Chapter Summary. 80 82 84 84 86 88 89 92. Chapter 5 General Discussion 5.1 The Corpus-Based Approach (Discourse Analysis) 5.2 Questionnaires. 94 94 97. 5.2.1 5.2.2 5.2.3 5.3 5.4. 97 99 100 102 104. Part One Part Two Part Three The Comparison between Two Approaches Chapter Summary. Chapter 6 Concluding Remarks 6.1 Summary of the Major Findings 6.2 Implications of the Study 6.3 Suggestions for Future Research. 106 106 109 110. References. 111. Appendix A Appendix B. 117 122. Questionnaire Corpus Data. x.

(12) LIST OF TABLES. Table 1. Table 2. Table 3. Table 4. Table 5. Table 6.. Participants (Men and Women) Participants (Junior and Senior Groups) The results of the corpus-based study Adjectival compounds which tend to be used with men Adjectival compounds which tend to be used with women Compounds with statistically significant differences between male and female participants Table 7. Compounds with statistically significant differences between the junior group and the senior group Table 8. Compounds with statistically significant differences among different age + gender groups Table 9. Preferred adjectival compounds. 42 43 48 72 72. Table 10. Dispreferred adjectival compounds Table 11. Compounds with statistically significant differences between male and female participants Table 12. Compounds with statistically significant differences between the junior group and the senior group Table 13. Compounds with statistically significant differences among different age + gender groups Table 14. The participants’ answers in Part 3 Table 15. Occurrence of male figures and female figures in the participants’ answers Table 16. Adjectives with statistic significance in the comparison of the male and female participants’ responses Table 17. Adjectives with statistic significance in the comparison of the junior group’s and the senior group’s responses Table 18. Comparison between the corpus-based study and questionnaires. 78. xi. 73 75 76 78. 79 81 83 84 85 86 88 90.

(13) LIST OF FIGURES Figure 1. Collocational spectrum of the 23 adjectival compounds Figure 3. Collocational spectrum of kuang-compounds Figure 3. Collocational spectrum of hao-compounds Figure 4. Collocational spectrum of rou-compounds Figure 5. Collocational spectrum of wen-compounds Figure 6. Collocational spectrum of zheng-compounds Figure 7. Collocational spectrum of the 23 adjectival compounds. xii. 49 54 57 63 68 71 85.

(14) Chapter 1 Introduction. 1.1 Motivation In Chinese, a character is usually a morpheme carrying a specific meaning. When it combines with other words to form compounds, these compounds have their own meanings, but they usually maintain some features of the original character. For example, the character jun (俊) ‘handsome’ in Chinese usually collocates with a [+human, +male] lexical item, and its compounds, such as junmei (俊美) ‘handsome, good-looking’, junqiao (俊俏) ‘handsome and dainty’, and yingjun (英俊) ‘handsome and sharp’, are all used to modify men as well. The “male” feature of the word jun is inherited by the compounds with it. However, this is not always the case.. Language, though being systematic and. routinized, is more complicated and mysterious than we expect.. The character hao. (豪) ‘heroic’ in Chinese is also a morpheme connoting masculinity.. Some of its. compounds, such as haoqi (豪氣) ‘heroic and powerful’ and haomai (豪邁) ‘bold and generous, heroic’, inherit its masculine feature.. These compounds are typical. masculine adjectives and traditionally collocate with male figures. However, some adjectival compounds of hao like haofang (豪放) ‘bold and unconstrained’ tend to collocate with female figures, despite the fact that the morpheme fang (放) ‘lively’ is not regarded as a typical female trait in Chinese. It appears that the collocations of such compounded adjectives are not merely predicted by the meaning of the respective compounded elements. In Chinese, gendered adjectival compounds sharing an identical character do not necessarily have similar collocational profiles. In this study, five characters with the tendency are selected (Section 3.1.1), including kuang (狂) ‘crazy or mad’, hao (豪) 1.

(15) ‘heroic’, rou (柔) ‘soft’, wen (溫) ‘mild’, and zheng (正) ‘upright’.. Given the. variable collocational phenomena of these adjectival compounds, the present study aims to probe into the gender connotations of the adjectival compounds formed with the above five characters, focusing on the collocations of these compounds in terms of the spectrum between masculinity and femininity. In the following, some general backgrounds of language and gender studies are delineated in Section 1.2. presented.. In Section 1.3, research questions of this study are. Section 1.4 concerns the significance of this study, and Section 1.5. introduces the organization of this thesis.. 1.2 Background Gender differences, an always intriguing topic, have attracted flurries of research and debate.. In addition to physical differences, men and women are often regarded. as distinct in many other aspects, such as their personalities and their attitudes. As a result, the images of men and women in people’s mind, both their outer appearances and inner traits, could be extremely dissimilar. No wonder there goes a famous proverb, “men are from Mars, women are from Venus” (Gray 1992).. This saying. humorously expresses many people’s idea that men and women are very different from each other, just like two beings from two planets. “Language” is also an aspect that laypeople generally agree to be discrepant between men and women.. The discrepancies lie not only in the language men and. women use, but also in the language used to refer to them.. In the past decades, the. interactions between language and gender have been extensively investigated (Lakoff 1975, Smith 1985, Poynton 1989, Holmes 1992, Swann 1992, Tannen 1994, Eckert & McConnel-Ginet 2003, among others). In Chinese, these phenomena have also been examined by quite a few researchers (Shih 1984, Yang 1998, Su 1999, Liu 2005, 2.

(16) among others).. Most of these scholars have pointed out a general situation that. women are in a disadvantageous position in language, for language usually tends to be more male-friendly.. The situation is strongly linked to women’s disadvantageous. position in society. The present study investigates the association of Chinese adjectival compounds with male and female figures, which is about “the language used to refer to men and women”.. Based on these adjectival compounds, this study further examines whether. the compounds of gender modification also reveal a similar tendency suggested in the previous studies that women are in a disadvantageous position in terms of language.. 1.3 Research Questions The present study adopts two different approaches to investigating 23 adjectival compounds, i.e., a corpus-based approach (or discourse analysis) and questionnaire collection.. With a corpus-based approach, data are collected from the web (through. Google), reflecting the common usage of the adjectival compounds investigated in this study. We collect 30 tokens for each compound, and count the frequency of the tokens’ collocation with male and female figures. The other approach, questionnaire collection, is beneficial to the examination of people’s thoughts and attitudes toward the compounds.. The participants, regrouped. according to their gender (male/female) and age (junior/senior), are asked to judge how often the compounds would be used to describe males and females, to express their feeling about the compounds, and to give representative figures for the compounds. The central issues the present study addresses include the common use of these adjectival compounds and people’s perceptions, attitudes, and gender images toward these compounds.. To investigate these issues, the following eleven research 3.

(17) questions are posed:. A. The corpus-based approach 1. In the corpus, what are the collocational tendencies of the 23 adjectival compounds containing the five characters kuang, hao, rou, wen, and zheng, and what do the tendencies mean? B. Questionnaires (Part 1) 2. How do people perceive these adjectival compounds’ inclinations to collocate with male and female figures and what do the collocational tendencies show? 3. Are there any discrepancies between men’s and women’s judgment of these adjectival compounds?. If yes, what do the gaps mean?. 4. Are there any discrepancies between juniors’ and seniors’ judgment of these adjectival compounds?. If yes, what generational meanings might the gaps. carry? C. Questionnaires (Part 2) 5. How do people themselves feel about these adjectival compounds when used to describe themselves?. What might the participant’s answers show?. 6. Are there any discrepancies between men’s and women’s feeling about these adjectival compounds?. If yes, what do the gaps mean?. 7. Are there any discrepancies between juniors’ and seniors’ feeling about these adjectival compounds?. If yes, what generational meanings might the gaps. carry? D. Questionnaires (Part 3) 8. Which gender is more likely to be the representative image of each adjectival compound? 4.

(18) 9. Are there any discrepancies between men’s and women’s images of these adjectival compounds?. If yes, what do the gaps mean?. 10. Are there any discrepancies between juniors’ and seniors’ images of these adjectival compounds?. If yes, what generational meanings might the gaps. carry? E. Comparing the two approaches 11. Do the results of the corpus-based study conform to the results of the questionnaires?. If not, what does the discordance mean?. 1.4 Significance of the Thesis The present study investigates the collocations of adjectival compounds by adopting two different approaches: a corpus-based approach and questionnaire collection. Corpora are often adopted in recent studies on discourse analysis and lexical semantics. be clearly revealed.. With corpus data, the distribution of specific linguistic items may Since corpora are comprised of large amounts of data, people’s. genuine linguistic behaviors could be observed.. Also, with investigation of the. information connected to the terms, subtle semantic clues these terms carry could be obtained. Therefore, this approach is very helpful in defining the semantic features of the terms, especially when finer-grained differentiation between synonyms or near synonyms needs to be made. Questionnaire collection, on the other hand, is an approach often adopted by studies in sociolinguistics.. With questionnaires, participants’ ideas, feeling and. attitudes toward the target questions could be easily observed.. In addition,. participants usually have adequate time to contemplate and provide the best answers when filling in a questionnaire.. Therefore, if researchers aim to examine. participants’ thoughts and attitudes, questionnaires could be a rather useful tool. 5.

(19) Questionnaires can help researchers conduct not only qualitative studies by analyzing the collected answers, but also quantitative studies when more questionnaires are collected and the answers in the questionnaires are turned into statistically useful data. However, existing linguistic studies seldom integrate the two approaches.. The. present study attempts to pioneer the integration, covering the merits of the two approaches.. With the method, the results of the two approaches can be compared.. It is hoped that the method could provide more detailed and profound analyses from different viewpoints and shed new light on the issues of the interaction between language and gender.. 1.5 Organization of the Thesis This thesis consists of six chapters. study.. Chapter 1 briefly introduces the present. In Chapter 2, some previous relevant studies are reviewed.. presents the methodology the present study adopts.. Chapter 3. Chapter 4 displays the results. and analyses of the results, and Chapter 5 proffers the synthetic discussion of the results. Finally, Chapter 6 summarizes the major findings of the present study and makes concluding remarks.. 6.

(20) Chapter 2 Literature Review. This chapter is devoted to previous relevant studies, which are reviewed in the following six sections.. Section 2.1 introduces some pioneering gender-related. studies which have attracted numerous follow-up studies on male and female behavior, especially men’s oppression on women.. Sections 2.2 and 2.3 turn the attention back. to language-related issues—studies on language bias: the former concerns studies in the Western cultural context and the latter those in the Chinese cultural context. Some studies about gender stereotype are presented in Section 2.4, and Section 2.5 reviews some studies concerning the relationship between language change and social change.. Finally, Section 2.6 presents some important studies on collocation.. 2.1 Gender Studies Language use can reflect social trends to some extent. trends might influence the way people use language.. That is to say, social. Although the interaction. between language and gender is the main focus of the present study, it is still beneficial to probe into more general gender studies first.. These studies could. provide fundamental ideas about the modes the society treats men and women, and these modes could be strong arguments for diversities in men’s and women’s language.. 2.1.1 Wollstonecraft (1792) Mary Wollstonecraft, a renowned philosopher and political commentator at the end of the 18th century, is considered the “first feminist” or “mother of feminism”. Her essay on women’s rights, A Vindication of the Rights of Woman (1792), pioneers 7.

(21) early gender studies and is deemed to be a powerful argument for following feminist movements.. She suggests that women should be treated as an essential element of a. country, for they take the chief responsibility for educating children. Also, they act as men’s partners more than just as wives, helping men to handle things that they themselves might not be able to tackle. As a result, Wollstonecraft suggests that women ought to have equal rights to those of men, rather than act as men’s assets in marriage. Wollstonecraft maintains that women in her period are quite foolish and shallow. She even calls them “spaniels” and “toys” (p.144). However, she does not ascribe women’s stupidity to their intrinsic nature but to their lack of education. The lack, of course, is due to men’s refusal to offer any chances for them to receive education. They are not allowed to develop their own career and obtain more knowledge in the society controlled by men.. As a result, women become relatively powerless,. dependent and passive, while men are powerful, independent, and active. These characteristics gradually turn into stereotypes, although the formation of these stereotypes should be attributed mostly to the way men and women are brought up rather than to the natural differences between men and women. Therefore, social and historical backgrounds do play an important role in shaping traditional gender notions. The only thing women are taught to be concerned about is beauty.. From their. childhood, they are told that “beauty is woman’s scepter, the mind shapes itself to the body, and, roaming round its gilt cage, only seeks to adorn its prison” (p.157).. They. are not only “protected” but also “isolated” from challenges in the outside world, just like canaries in birdcages.. All they need to do is to maintain their beauty and grace. and please others, i.e., men.. If women’s appearance is not overemphasized, they. might also be successful in many other respects and fields. 8.

(22) 2.1.2 Mill (1869) As a male, John Stuart Mill is considered to be a rare scholar in early periods who dedicated his efforts to the study of women’s place.. Although men in his time. were dominant in the society and treated women as incompetent, he still suggests in his work The Subjection of Women (1869) that like men, women could also be mature adults and enjoy the same right as men did.. Mill does not hold the idea that women. should belong to families, although this idea was generally accepted by both men and women.. He suggests that the “tradition” can be traced back to history in which. women were usually limited to the domestic domain and subject to men’s control. However, women’s abilities are seriously ignored in their families, while men’s power is highly exaggerated in other fields.. Such division of labor forces males and. females to develop in different directions.. Hence, women live an oppressive life,. sacrificing themselves for men and their families and excessively caring about others’ opinions, especially men’s approval. Meanwhile, men become aggressive, proud and selfish, concerned only about their own will and emotions. Gradually, men refuse to accept others’ criticisms, especially women’s, and women’s voices become muffled.. 2.1.3 Woolf (1929, 1938) Virginia Woolf is generally agreed to be one of the most prominent and influential novelists in the twentieth century. Her novels A Room of One’s Own (1929) and Three Guineas (1938) criticize male chauvinism and reveal the obstacles women have to face.. In A Room of One’s Own, Woolf states that women do not have. their own space to develop their own interests and careers, even if they just want to write their own works. By creating a figure “Shakespeare’s sister” in her book, she 9.

(23) implies that if Shakespeare is a female, he might have died with his talent left to rust. Women are obviously denied to have equal opportunities to develop their talent by men. In Three Guineas, Woolf even more harshly attacks men’s dictatorship, glorification of male power and suppression of women. She argues that women are unfairly treated in both private (family) and public (education, profession, and politics) domains.. Some biased situations that women have to face in the English society at. that time include the limitation for women to receive higher education, the denial of women’s fundamental rights in laws, and the rejection for women’s taking higher positions in jobs.. Woolf taunts men’s stale notions and indicates that men do not. know how to introspect themselves.. 2.1.4 De Beauvoir (1949) Simine de Beauvoir, a French author and philosopher, is regarded as an important figure of contemporary feminism.. Her treatise The Second Sex (1949). provides a comprehensive analysis of female’s oppression, which accelerates women’s awakening and triggers more drastic movements to struggle for women’s rights. It is suggested that women always position themselves according to their relationships with men and take men’s ideas into consideration. do not exist as an independent individual.. Consequently, they. Here is a famous sentence in her book:. “One is not born, but rather becomes, a woman” (p.267).. She argues that gender is. not necessarily a product of biology. Instead, gender is socially constructed. In other words, she suggests that gender is something that one learns to do. Men, on the other hand, are always considered to be independent individuals; they do not have to identify themselves based on their relationships with women. The asymmetrical relationships between men and women reveal the fact that women 10.

(24) are always defined in terms of their gender instead of their rationalities.. They are. not complete beings due to the limitations the society imposes on them.. However,. de Beauvoir does not think the disadvantage of women should be totally attributed to men, for women should also take the responsibility to break these preconceived ideas. They should bravely maintain their dignity, denying the traditional subordinate positions.. 2.1.5 Money et al. (1968), Money and Ehrhardt (1972) Money et al. (1968) and Money and Ehrhardt (1972) state that the sex a baby is assigned to and raised determines his/her gender identity. That is, gender identity might be undifferentiated at birth. Latter on, if we give a baby a label of sex and raise him/her as such, the baby would also grow up according to it. The study conducted in 1968 investigates girls who suffered from “adrenogenital syndrome”, also known as congenital adrenal hyperplasia. Due to the abnormality of their genitals, they were mistaken as boys in their infancy. If they received surgery by the age of three, most of them could have adapted themselves to the gender as girls. However, if they did not receive surgery before three, they might have had difficulty identifying themselves as females.. Most of them would act like boys and agree to. remove their uterus when they grew up. The other study (1972) investigates two twin brothers. received circumcision when they were 7 months old.. The two brothers. One of them got hurt, and his. penis was seriously damaged. Afterward, the baby’s penis was replaced by a vagina after undergoing an operation at the age of 17 months, and the boy was then brought up as a girl.. The “girl” was originally as mischievous as a boy and liked to control. her brother.. After the operation, her parents tried to cultivate her female. temperament and dressed her up as a beautiful girl. 11. She was taught to be quiet,.

(25) polite, and tender as a girl. Gradually, she started to become feminine and weak, and her brother became the brave protector whenever someone attempted to bully her. It is manifested by the two cases that nurture on development could seriously influence children’s identity and their behaviors.. Children’s gender identity might. not be determined by their sex characters but led by social factors.. They learn to be. a boy or a girl according to their sexuality, learning dos and don’ts like what other boys and girls do.. Therefore, they gradually construct their identity toward a. specific gender.. 2.1.6 Other Related Studies Kate Milllett’s Sexual Politics (1970) more radically reflects the weakness of women, emphasizing men’s oppression and alienation. She wonders why the world is still controlled by men after women’s success in gaining the opportunities to receive education and to participate in politics.. The real problem, as Millet indicates, is that. women are constantly brainwashed by the world.. From their infancy, they are still. brought up and imbued with traditional rules and values. Their life environments also strengthen these ideas through different ways, such as language, laws and mass media. As a result, women are forced to conform to the stereotypes, acting as passive, kind, emotional, family-oriented, and beautiful daughters, wives, and mothers. However, these traits are traps men set for women to jump in, because by doing this men could easily control women and gain whatever they want.. If women choose not. to jump, they are accused of disobeying natural rules and told that they fail to be women.. In such a paternal society ruled by men, women’s conversion seems. impossible. Germaine Greer holds similar ideas to Milllett’s, disparaging the idea that women are innately different from men.. In her work The Female Eunuch (1970), 12.

(26) she argues that many traditional gender notions are created by males to dominate women.. Women, like eunuchs’ castration, are prevented from developing their own. features and strengths.. They are encaged, lose their passions, and become. sex-objects for men step by step.. 2.2 Gendered Language and Gender Biases in the Western Society Language is an aspect which may honestly reflect social values.. Since women. are usually regarded as historically disadvantageous in their social position, language should also be able to reveal the fact.. This section reviews some studies on language. biases in the Western cultural context.. 2.2.1 Lakoff (1975, 1990) When it comes to language and gender, Robin Lakoff’s (1975) enormously influential study Language and Woman’s Place is a must reading.. The study. discusses the most important obstacles women encounter in their lives, i.e., language. Her proposal attracts swarms of researchers to engage in gender studies, especially the interactions between language and gender (Smith 1985, Poynton 1989, Swann 1992, among others). She points out that language reflects the roles of men and women in their societies.. Men’s verbal behavior is often considered to be more powerful,. prestigious, and desirable, while women’s speech tends to be self-denial and subordinate.. For example, the use of hedges characterizes the language of those who. are “out of power” (p.53) in the society, and women’s language tends to contain hedges. From the language people use, it could also be found that language exhibits men’s higher status than women’s and that women’s language is even regarded as “worse” than men’s.. Men’s language is “the language of the powerful, which is 13.

(27) meant to be direct, clear, and succinct” (p.205), while women’s language seems to be tentative, powerless, and trivial, which makes them disqualified from possessing authority and power. control women.. As a result, language itself is sure to be the best tool for men to. Women are forced to listen instead of speaking and to approve. instead of opposing. Lakoff suggests that women undergo language discrimination in two ways: “in the way they are taught to use language, and in the way the general language use treats them” (p.4).. Following Lakoff’s study, numerous studies examine more. linguistic phenomena related to men and women.. 2.2.2 Schulz (1975) Schulz (1975) investigates semantic change of words, finding that a lot of abusive words are originally neutral or even positive. However, when they start to be used to modify females, they gradually obtain some negative connotations and became abusive.. Besides, the connotation of some terms depends on the gender of. the figures they depict.. They denote positive meanings when describing men;. however, when they collocate with female figures, negative meanings are generated. For example, a male pirate indicates one who infringes on rights of others or commits robbery on high seas, whereas a female pirate becomes an adulteress who chases other women’s man or partner. The words “callboy” and “callgirl” are also an example of asymmetric language for men and women; the former denotes just a male servant, while the latter means a whore. Furthermore, Schulz reports that there are far more abusive terms for women than men. “prostitute”.. For example, she investigates slang and collects 500 synonyms for However, for the synonyms of masculine sexual term “whoremonger”,. only 65 words are found.. Schulz suggests that the reason for the pejorative language 14.

(28) for women is men’s sexual fear for women; they worry about women’s menace to their power.. 2.2.3 Tannen (1993, 1994) Tannen states that men and women speak totally different “genderlects” in the same community, because the processes of their socialization are different (1993:43). She asserts that men’s and women’s focuses are different. emphasis on competition, independence, and position.. Men’s world lays. However, characteristics like. intimacy, solidity, and interdependence are accentuated in women’s world. the languages they use are also quite different.. Hence,. Tannen suggests that the different. attitudes and the genderlects are acquired from men’s and women’s same-sex peers when they are still young children. In other words, in a group of boys, children learn to contend with each other.. However, in a group of girls, children learn to tolerate. and help each other, sharing each other’s favorites and secrets.. As a result, the. languages they use to achieve their goals are different. In Tannen’s another work (1994), she discusses “marked” and “unmarked” behaviors and terms for men and women in her society. It is found that the default or unmarked ones always belong to men:. I asked myself what style we women could have adopted that would have been unmarked, like the men's.. The answer was none.. woman.. There is no unmarked (Tannen 1994:630). She adduces some interesting examples. It is indicated that suffixes like -ess and -ette mark words as “female”. Also, these suffixes also tend to carry extra negative connotations, such as frivolousness and sex. 15. Furthermore, when filling.

(29) out a form, women are forced to reveal their personal information, because they need to choose a “title”.. Choosing “Mrs.” not only shows women’s marital status, but. also implies that they are “conservative” in the form of address.. Although choosing. “Ms.” waves aside the danger to disclose their marital status, it may hint that they are either liberated or rebellious. The last title women can select—“Dr.”—tends to be interpreted as a show-off. Married women’s surnames are also marked. Tannen suggests that the surname not only reveals a woman’s marital status but also exhibits her obedience to traditional values—a woman’s affinity as well as affiliation to her husband. rejecting their husband’s name is still marked.. Nevertheless,. These contradictories reveal the truth. that women do not have the freedom to be “unmarked”.. 2.2.4 Sutton (1995) Sutton (1995) investigates “ugly names” of women, discussing slang used by youngsters.. She collects slang frequently used by undergraduates from diverse. ethnic groups at Berkeley. She finds that the number of negative terms for women is overwhelmingly larger than that of positive ones.. In addition, those negative words. mostly carry sexual connotations, overemphasizing women’s body attraction.. She. classifies the collected terms into several types: animal references (women as animals and as men’s pets, such as hellpig and nocturnal), women’s genitalia references (women as sexual organs, or even holes, such as clam, pink taco and cheesehole), objects references (women as attractive objects, especially as food, such as cheesecake and peach), terms for promiscuous women (such as skank and hobag). From these terms, it is concluded that women are expected to be thin, smart but not too smart, passive, sexually available but not promiscuous, while they are not expected to be fat, ugly, aggressive, or unattractive. 16. Women are in fact defined in.

(30) terms of their sexual relation with men. Sutton further investigates users of these terms and finds that some words, such as bitch and ho, are used by women as terms of affection when they are addressed to female close friends, but insults and offense are generated when men use them to portray women.. The author suggests that women do so not for talking like men or. African Americans; instead, they seek their identity in a male-dominated world as individuals and as a group. As a result, women’s addressing these terms to close friends represents solidarity of women. In addition, for women, using these terms can be an escape from traditional norms set by men.. Women feel more secure when. they are accompanied by their friends; hence they can disobey or even ridicule societal expectations for women and gain some power of control through the use of these taboo words.. 2.2.5 Fowler (1991) Fowler’s work (1991) discusses discrimination against women through lexical uses in news discourse.. He suggests that category labels in the media can reveal. ideologies in society. By labeling, discrimination from stereotypes might not only be attached to an individual, but also be imposed on all members of a group, such as “young married women” or “teachers”. Discourse is hence extremely powerful in enhancing or holding discrimination against specific members or groups, for it might set the boundaries for certain categories, and this might influence people’s conceptions through frequent use. In addition, he suggests that syntactic forms and transitivity are also important for the formation of ideologies. Fowler selects “the 1986 New Year’s Honours List” and also some other major stories and articles in a complete set of newspapers as materials of his study.. He. hypothesizes that a figure’s being located in a subject or object position might 17.

(31) implicate the authority of this person, and a figure’s being linked to specific verb types may also indicate the person’s ability and power.. Besides, the ways to. describe or to refer to a specific person are usually related to social values. He discusses power discrepancy between men and women in the material. It is found that some naming forms, such as diminutives, first names only, and titles related to marital status or kinship, are mainly used with women, but seldom with men. Also, the recorded actions of women in newspapers tend to reveal female stereotypes of hysteria or irrationality, and women are usually portrayed in terms of sexual stereotypes.. These frivolous uses are rarely connected with male figures.. Furthermore, gender-specific terms scarcely apply to men but frequently to women, and general terms sometime simply designate men.. Women are characterized. unfavorably, while successful men are highly esteemed in their fields. Fowler suggests that repeated uses of these expressions strengthen the distinctions between males and females, naturalizing the discrimination. distinctions also reflect different ways society treats men and women.. The. However,. women are often powerless in the media, and most media also regard the biased treatment to men and women as a natural phenomenon.. 2.2.6 Caldas-Coulthard (1995) Following Fowler’s study, Caldas-Coulthard (1995) also explores men’s and women’s voice in the press.. She explores men’s and women’s participation in the. news and finds that most texts are about men and also written by men.. She. investigates some lexical items, including address terms (Miss, Mrs., Ms. and Mr.), spokesperson. (spokesperson,. spokesman. and. spokeswoman),. chairperson. (chairperson, chairman and chairwoman), pronouns (he and she), genders (woman and man), and roles (wife, husband, widow and widower). 18. It is discovered that.

(32) female terms occur much less often in comparison with their male counterparts except terms of role, because women are usually labeled according to their relationship with men. Caldas-Coulthard also collects verbs of “saying” with high frequency occurrences in the press to find out which gender is quoted more often.. It is found. that quotations of men’s speech are eight times more than those of women’s.. Also,. these verbs express some gender stereotypes, emphasizing the negative image of women, as Fowler proposes.. For example, men “shout” and “groan” while women. “scream” and “yell” (p.235). Finally, she examines terms of addressing speakers in the press. As found in Fowler (1991), male speakers are usually addressed by their professional titles or position, while women speakers are often glossed with various information, such as their marital status, age, and diminutive.. Men speak as a. professional role, while women seem to speak as normal civilians, telling their personal affairs. Caldas-Coulthard asserts that women are members of the unaccessed voice group, and their language is less visible in the press.. She emphasizes that the invisibility of. women’s language in the press is a piece of powerful evidence for women’s deviant and powerless status. The situation that the press is male-oriented also represents that women are unconsciously imposed with certain values, i.e., men’s language and viewpoints.. If the press kept ignoring women’s language and contribution, biased. assumptions would be conveyed and reinforced constantly.. 2.2.7 Other Related Studies Men are powerful in language, and whether they intend it or not, they may use language to oppress women.. Since gender biases in language are matters of repeated. occurrences, the phenomena have piqued a lot of researchers’ interest (Smith 1985, 19.

(33) Poynton 1989, Holmes 1992, Swann 1992, Cameron & Kuklick 2003, Eckert & McConnel-Ginet 2003, among others). Holmes (1992) states that the norms for women’s talk are indicated as the norms for small group interaction in private contexts.. Eckert and McConnell-Ginet. (1992:90) also suggest that “women’s language has been said to reflect their conservativism, prestige consciousness, upward mobility, insecurity, deference, nurture, emotional expressivity, connectedness, sensitivity to others, solidarity. And men’s language is heard as evincing their toughness, lack of affect, competitiveness, independence, competence, hierarchy control.” It is even explicitly stated in the renowned and influential dictionary Oxford English Dictionary that the term “woman” can be used “with allusion to qualities attributed to the female sex, as mutability, proneness to tears, or physical weakness; also to their position of inferiority or subjection” (Swann 1992:36-37).. Again and. again, women’s disadvantage in language is obviously revealed.. 2.3 Gendered Language and Gender Biases in Chinese Like the Western society, language biases could also be discovered in the Chinese cultural context.. This section reviews some studies on language biases in. Chinese.. 2.3.1 Shih (1984) Discrimination against females could also be found in Chinese culture, and the situations are even worse, for the boundaries between men and women are much more rigid in Oriental culture. Shih (1984) considers that different linguistic behavior between Chinese men and women might be traced back to their traditional positions and roles in history, in which women were usually devalued. Although women’s 20.

(34) social status has become higher in these decades, the traditions are still rooted deeply in many people’s mind.. Language is also greatly influenced by traditional notions.. Most gender-biased linguistic features in English could mostly be found in Chinese.. Shih points out nine linguistic features of Chinese men and women, and. some of them are concerned with women’s depreciation in language. Firstly, biases could be found in the components of a large number of Chinese characters.. For. example, a lot of characters containing the radical nu (女) ‘women’, such as nu (奴) ‘slave’, jian (奸) ‘wicked’ and ji (妓) ‘prostitute’, carry negative connotations. addition, some Chinese lexical items also demonstrate discrimination.. In. For examples,. abusive terms for women, such as pofu (潑婦) ’shrew’, wanniang (晚娘) ‘stepmother’, and huoshui (禍水) ‘disaster’, are much more than those for men. Some Chinese proverbs and classics also hold gender discrimination, for example, Shige. furen, jiuge du. (十個婦人，九個妒。) ‘Among ten women, nine of. them are envious.’, Nanren. sanshi. yizhi. hua, nuren sanshi. laorenjia. (男人. 三十一枝花，女人三十老人家。) ‘At the age of thirty, men are handsome and prosperous, while women have become old.’, and Wei yang. nuzi. yu. xiaoren. nan. ye. (唯女子與小人難養也。) ‘Women and children are hard to get along with.’.. Address terms for men and women are also biased. It is demonstrated that women’s titles are often belittled.. For example, when women talk about their husbands in. front of others, they use waizi (外子) or wo xiansheng (我先生) to name them. However, when men talk about their wives, they use negative terms such as jiannei (賤內) or zhuojing (拙荊). Chinese naming is also a piece of evidence for gender bias. Boys’ names usually carry more expectation from their parents than girls’. For example, men’s names often entail that males are expected to bring honor to their family, such as guang (光) ‘light’, zong (宗) ‘clan’, yao (耀) shine, and zu (祖) ‘ancestor’, while 21.

(35) women’s names often stress the importance of their beauty, such as mei (美) ‘beautiful’, xiu (秀) ‘beautiful’, and zi (姿) ‘appearance”. These features in the Chinese language reveal the fact that women are devalued in their social status.. 2.3.2 Shieh (1997) It is pointed out by Shieh (1997) that textbooks for students in Taiwan also show a lot of gender biases. He suggests that in textbooks, the roles, characteristics, and positions of men and women are often distorted. become misleading.. Thus, teaching materials might. For example, a lot of examples and illustrations in these. textbooks do reinforce gender stereotypes, ignoring other possibilities about gender. In addition, textbooks often emphasize differences between men and women. For example, in a textbook, men’s bravery and wisdom are often praised, while women’s tenderness and meekness are frequently marked.. Furthermore, male and. female writers and figures are discrepantly selected in the textbooks, with the former outnumbering the latter.. Li’s study (1988) is taken as an example.. Li investigates. all elementary school textbooks. It is found that among all figures appearing in elementary school textbooks, only 2.1% are females. are frequently used in textbooks.. Moreover, male generic terms. For example, words like chairman and policeman. which stand for both men and women are still taught at school. Shieh implies that gender biases in textbooks might prevent students from correctly constructing their conception of gender. Besides, the biased material in these textbooks would be very likely to become students’ models for playing their gender roles.. Therefore, these textbooks are subject to stricter censorship, and the. discriminations should be avoided.. 22.

(36) 2.3.3 Liao (2000) Biases against women can even be seen in fundamental language uses.. It is. observed that some Chinese women’s names reveal gender discrimination (Liao 2000). Liao conducts a comprehensive study on personal names in Taiwan and generalize 19 rules of naming.. It is found that some women’ names express their parents’. reluctance to have the children.. For example, some women born in the 1930s are. often named Wang-shi (罔市) and Wang-yao (罔腰), both of which mean ‘to raise the child reluctantly in Taiwanese’. Some women’s names also carry parents’ hopes that they would have a boy next time.. For example, names like Dai-di (帶弟), Chao-chi. (招治), and Chao-lan (招蘭) all signify ‘to expect younger brothers’. Names of the kinds reflect girls’ being extremely unwelcome by their family. Liao states that names sometime also mirror expectations parents have for children.. In Taiwan, boys’ and girls’ names are often distinct, which indicate. parents usually set different goals for boys and girls:. They want their sons to be excellent, heroic, moral, great, wise, prosperous, powerful, and lucky in life.. They like their daughters to be beautiful, frugal,. quiet, and pleasant to please people around them.. (Liao 2000:53). For example, boys are often named with hsiung, wei, and jie, which all display parents’ expectation that their boys would be outstanding.. However, women are. often named with flowers such as he (荷), lian (蓮), and lan (蘭), which project parents’ hope that girls would be as beautiful as these flowers.. 2.3.4 Lin (2004) Lin (2004) indicates that biased values are also rooted in songs often sung and 23.

(37) heard, whether in children’s songs or pop songs.. She points out that some lyrics of. children’s songs reveal gender stereotypes, and hence children are imbued with these traditional values unconsciously. Several famous children’s songs are discussed, such as the two famous songs Zhiyao wo. zhangda (只要我長大) ‘As long as I. grow up’ and Meimei bei zhe yangwawa (妹妹背著洋娃娃) ‘My younger sister is carrying a doll on the back’.. The first song praises males’ working hard outside for. the family on the one hand, and describes that women should stay at home and take all responsibility of household affairs on the other.. The second song also hints that. “dolls” belong to girls. Also, the song indicates that when a baby cries, the mother should take the responsibility to pacify him/her. That is, mothers should take care of babies.. Both of the two songs talk about the Chinese tradition that men’s work. centers around outside, and women’s work centers around the home (男主外、女主內). It is also possible that females are praised in children’s songs. expressions of admiration are limited to mothers’ tender sentiments.. However, most On the other. hand, songs related to fathers’ love are rarely found. This suggests that men are trained to be cold and strict with their children, because they need to achieve higher goals, i.e., outside works. However, mothers are expected to be mild, kind, and intimate to their children.. That is to say, the definitions of good father and good. mother are quite different.. Good fathers have to work hard outside, while good. mothers are supposed to rear their children without complaints.. The fixed traditions. burden great pressure on both men and women. Even in general pop songs, it is indicated that women are usually housewives, mothers and girlfriends. They are tender and shy, waiting for men’s care. They would never become men’s partners in their professions in these songs. Nevertheless, men are strong, positive, and active in pursuing anything they like. 24.

(38) Lin proposes that lyrics of pop songs reflect the social condition on the one hand, and exert some impact on the society on the other hand.. With agreeable melody, the. gender-biased language is repeatedly transmitted to people.. Therefore, gender. values and stereotypes are constantly enhanced.. 2.4 Gender Stereotype Gender stereotypes could be a key factor to influence people’s conceptions about gender. Many studies have paid attention to the issue. The following sections review some studies on gender stereotype.. 2.4.1 Basow (1992) Basow (1992) conducts a very comprehensive study on gender stereotypes and roles.. She discusses discrepancies between men and women from various. perspectives, including physical, cognitive, characteristic, social, sexual, historical, and generational viewpoints.. It is argued that most of the discrepancies actually are. not the reasons for different gender roles in society; on the contrary, they are the “results”.. Therefore, gender is gradually constructed in society instead of being. assigned from birth.. Basow suggests that gender stereotypes extensively influence. individuals, controlling their relationships with others. negative effects on the whole society.. These stereotypes also have. Although it is not easy to break the. stereotypes, Basow suggests that people should still work hard to change them, especially women.. 2.4.2 Zou (1986) Zou (1986) investigates how the roles of men and women are presented in soap operas.. She chooses three soap operas and explores 41 figures in the operas. It is 25.

(39) found that the traits of these figures correspond to several gender stereotypes, especially female figures. and emotional.. These women are portrayed as frail, sentimental, sensitive,. Many of them are married and have children, concentrating their. whole heart to their family.. They usually have no job.. Even though some of them. have a job, it is still shown that they care more about family affairs than their jobs. These women with jobs also tend to rely on or obey their male colleagues.. On the. other hand, male figures in the operas are usually unmarried and do not focus on their family life.. Many of them are independent, enthusiastic, rational, and have. prominent occupations. They can handle their jobs and family well at the same time. Zou believes that the way these figures are presented in the operas is greatly influenced by the society and culture. For example, men more often occupy higher positions than women, and they are respected by others in real life.. Therefore, the. situation tends to be duplicated in the operas. Although the operas are designed in some special context, a lot of people would believe that all conditions in the operas might happen in real life.. Similar misleading impressions and stereotypes are also. broadcasted and responsible for women’s disadvantageous fate.. 2.4.3 Visser (2002) Gender is a permeating social and cultural phenomenon which exists in everyone’s life.. It not only regulates our behavior but also affects our ways of. thinking and speaking.. Visser (2002) discusses the conceptions of gender. internalized in people’s subconsciousness.. She recruited freshman students as. participants, who were asked to rank some traits’ suitability to be categorized into the two categories: FEMININE and MASCULINE.. The participants were asked to. provide every listed attribute in the two categories a rank from 1 to 7, where “one” represents a best example, while “seven” stands for a worst one. 26. The experiment.

(40) was conducted twice, with an interval of 5 years between the two experiments, in order to examine the conceptual changes which may be indicators of social and cultural changes during the period. The results show that the highest ranks of MASCULINE for young people is about independence, especially in their career, while FEMININE is mostly associated with the characteristic of outward appearance.. With the investigation of the other. non-salient traits, it is found that MASCULINE is less changeable and diverse than FEMININE.. Furthermore, changes of nonprototypical attributes rating also reflect. some trends.. For example, the rising of men’s awareness of their appearance, the. awakening of women’s sexual rights, and also the declining of women’s being unfree and dependent on others, can be observed in these attributes. With the overall investigation, Visser concludes that women are still expected to be family-oriented rather than career-oriented. Even though nowadays women may have their own careers, usually they still have to be a responsible family work organizer. The dual identities make them even strenuous. Finally, Visser suggests that any attempt to change gender stereotypes is ineffective. Only societal changes can lead to gender stereotype changes, for gender stereotypes strongly influence children and force them to develop according to the stereotypes.. 2.4.4 Litoselliti (2006), Caldas-Coulthard (1996) Litoselliti (2006) investigates women’s magazines and discovers that these publications declare women’s changes, such as discussions of the occurrence of “superwomen”, women who are successful in both their domestic lives and their professions.. On the other hand, the central topic, consumption, still keeps. encouraging women to conform to the stereotypes.. Caldas-Coulthard’s (1996). investigation on women’s magazines also reveals similar contradiction. 27. These.

(41) magazines discuss the transgression in sex, but in fact emphasize morality and traditional views of women’s role in society.. 2.5 Social Change Although women are generally regarded as socially disadvantageous, gender images are always fixed and might also change with time.. The following sections. review some studies on social changes, which concern the shifts of gender roles in particular.. 2.5.1 Luo (1981) Luo (1981) analyzes the special column “women’s mailbox” in newspapers, which aims to answer the questions that women encounter. Questions discussed in this column are diverse, such as sex, marriage, family, occupations, and partners. Luo suggests that it can be discovered from those articles in the columns that women’s position has been lifted, because what women need and want is expressed instead of being ignored.. Also, from the information provided in the column, it. could be observed that women’s life experiences start to be emphasized. However, Luo also points out a formidable obstacle to women’s lift in position. She suggests that many women still do not know how to fight for their rights, or they even do not want to fight at all. They have got used to tolerating the inequality occurring around them. reeducate women.. Luo believes that women’s mailbox could be a good way to. Through the column, the idea of gender equality could be. conveyed to change women’s traditional values and break the stereotypes.. 2.5.2 Lin (1998) In order to investigate the change of men’s roles in society, Lin’s (1998) study 28.

(42) focuses on advertisements in magazines, for she suggests that how men are displayed in these advertisements could reflect their social roles.. She points out that during the. period between 1988 and 1997, the advertisements in magazines kept demonstrating gender stereotypes.. Men are shown with the masculine image and appear mostly in. their professions instead of being in their family.. As for women, they are still. presented with the image of submission. However, it is observed that men’s roles become diverse with time, but the change is very slow. other interests.. Apart from their profession and family, they start to develop. Although men’s roles are changing, they are still those who own the. most of power.. 2.5.3 Other Related Studies In Section 2.4.3, it is pointed out that gender discrimination in language is still displayed in Litoselliti’s (2006) and Caldas-Coulthard’s (1996) studies, and that some actual changes in women’s features are still undoubtedly revealed, such as women’s valuing personal styles and women’s freedom to enjoy sex. This may indicate social change to some degree. Similarly, Benwell (2003) also states that men’s magazines also start to accentuate the concept and image of “new man”, a new form of masculinity. “New man” is presented as fashionable, healthy, leisure, stylish, and attaching importance to appearance.. Intriguingly, some of them are in fact the features of stereotypical. women. Visser’s (2002) study on gender stereotypes also reveals the trace of social change.. She asked the participants to categorize some terms into “femininity” and. “masculinity”.. As Section 2.4.2 has displayed, it is found that the participants’. preferences fit gender stereotypes well.. However, she further indicates that there are 29.

(43) also some changes when the old data and new data are compared, such as the rising of men’s awareness of their appearance, the awakening of women’s sexual right, and the independence of women.. The change in traits associated with men and women may. imply gender stereotype change in society. According to Lakoff (1975), social change predicates language change. By observing language changes, we might be able to find what concepts are undergoing changes with time.. 2.6 Collocation Collocation has been extensively investigated, for it is one of the conventions of language use.. The present study adopts a most general viewpoint on collocation and. simply treats collocation as an umbrella term, regarding collocation as a phenomenon that a word habitually occurs with some words or some words containing specific features.. The related studies on collocation could help us gain a better understanding. of word combination. The following sections summarize several researchers’ studies on collocation and word combination.. 2.6.1 Firth (1957) Firth (1957) may be the first person in the literature who investigated collocations. (p.196).. He claims that collocation is “the company words keep together”. Besides, it is indicated that meanings between words could be differentiated. at five levels, i.e., orthographic (spelling or word form), phonological (pronunciation), grammatical (part of speech), situational (social or pragmatic meaning), and collocational levels.. Collocation does play an important role in distinguishing. words’ meanings. In addition, Firth also states that there are two types of word associations, 30.

(44) paradigmatic and syntagmatic relations.. Paradigmatic relation refers to the situation. that words could be placed in an identical grammatical category, and syntagmatic relation indicates that words co-occur with each other in a linear order.. Therefore,. collocations are word relations at the syntagmatic level. Firth’s ideas are widely adopted by the following studies on collocation.. 2.6.2 Benson et al. (1986, 1997) Benson, Benson, and Ilson (1986) categorize lexical combinations into 5 types according to their degrees of cohesiveness: free combinations, collocations, transitional combinations, idioms, and compounds. The loosest combination is free combinations (e.g. to analyze/report/investigate a murder). Collocations are more fixed than free combination (e.g. to commit murder), and the meaning of a whole combination is quite related to the meaning of the parts. Therefore, it is relatively unacceptable when other words are used to replace any word in a collocation.. As for. transitional combinations (e.g. the facts of life), their degree of cohesiveness is higher than collocations because they are less alterable.. Idioms (e.g. to have one’s back to. the wall) are rather fixed expressions, and the meanings of their parts may not constitute the meaning of the whole.. The most stable combination is compounds. (e.g. floppy disk). Benson et al. (1997) further categorize collocations into two types in the BBI Dictionary of English Word Combinations: grammatical collocations and lexical collocations. Grammatical collocations refer to phrases that comprise a dominant word and a grammatical structure (e.g., adjective + preposition combination, such as proud of and satisfy with).. Lexical collocations are combinations composed by. nouns, adjectives, verbs or adverbs without prepositions (e.g., acquire knowledge).. 31.

(45) 2.6.3 Sinclair (1991) Sinclair (1991) proposes four word co-occurrence relations in extended lexico-semantic units, including collocations, colligation, semantic preference, and discourse prosodies.. A collocation is merely a lexical relation that makes no. reference to syntax, and it contains a pair of a node and a collocate. A node is the word form, lemma (a lexeme or a dictionary headword), or other pattern investigated, and it co-occurs with collocates (e.g., loud/thunderous/rapturous/spontaneous applause). A collocation relation also implies that the combination frequently occurs in the corpora. grammatical. Colligation is syntactic relation, indicating the combination of. categories. or. some/many/most/both cases).. the. pairing. of. lexis. and. grammar. (e.g.,. in. Collocation and colligation are often related to each. other, for specific grammatical patterns are sometime more frequent than other ones in a collocation relation. Semantic preference is the relation between a lemma (or word form) and a set of semantically relevant or similar words (e.g., large number/scale/amounts/areas, all the terms are related to the quantity and size). A discourse prosody is the additional meaning extended from the combination (e.g., cause often related to unwelcome events, such as to cause problem/damage/trouble). Discourse prosodies often reveal speakers’ evaluation of specific events.. 2.6.4 Stubbs (2001a, 2001b, 2002) Like what Firth (1957) has suggested, Stubbs (2001a, 2001b) maintains that collocation is a syntagmatic relation that words frequently appear together in a linear string, as shown below:. collocates…node…collocates span 32.