[ Back to publications page ]
Proceedings of the Ninth North American Conference on Chinese Linguistics (NACCL-9, May 1997), edited by Hua Lin. 1998. Los Angeles: GSIL Publications, University of Southern California. Volume 2, pages 35-52.
GENDER DIFFERENCES IN THE CHINESE LANGUAGE:
A PRELIMINARY REPORT *
Marjorie K.M. ChanOhio State University
Research on language and gender interaction is well into its third decade and yet there have been surprisingly few contributions from the Chinese language to the explosion of cross-linguistic literature on the topic. This paper brings together both scattered observations and detailed published works on Chinese to provide a preliminary report on gender differences in the Chinese language. |
1. INTRODUCTION
In the general linguistic literature, hundreds of popular and academic treatises have been published on language and gender since the early 1970's in the United States, prompted by the women's movement. For the Chinese language and its dialects, during that period and into the 1980's, scant attention was paid to gender-differentiated speech aside from language variation research, in which sex is an important independent variable (e.g. Shen 1987). Studies devoted specifically to issues involving language and gender were rare. Notable exceptions include Light (1982), probably the very first to approach the subject and primarily on Cantonese; Shih (1984), the first general overview (in Chinese); and Farris (1988), the second overview (in English). These early studies paved the way for subsequent exploration into gender-related differences in Chinese. This paper's aim is to give a short report on some earlier and recent findings, including the author's, of sex-based differences at different levels of linguistic structure in the Chinese language. Section 2 reports on gender differences observed with respect to the phonetics and phonology of the language. Section 3 deals with the lexical level, both with respect to differences in vocabulary choice between men and women, and gender-differentiated vocabulary items about men and women. Section 4 concentrates on syntax and pragmatics. Section 5 concludes with a preview of new research in gender differences in conversational interaction in Chinese. This report focuses on standard (Mandarin) Chinese, with side references to Cantonese.2. PHONETICS AND PHONOLOGY
There have been numerous observations of speech differences between men and women. The most fundamental difference is the average pitch of voice of females versus males that is largely, but not entirely, due to differences in the vocal anatomy. Much of the differences in pitch of voice between the sexes is socially learned. While there is a great deal of overlap between males and females in their pitch range, speakers use only a small part of that range.1 Cross-cultural studies, for example, show that pitch and voice quality vary from culture to culture. Majewski (1972) reports the average speaking pitch of 103 Polish men to be 137.6 Hz, whereas that of a comparable (though much smaller) group of American men was 118.9 Hz. Loveday's (1981) study compares the average pitch of voice used by Japanese and (British) English men and women. There is a dramatic difference in the average pitch of Japanese men and women, with women's voices reaching as high as 400 Hz. Nonetheless, the latter's high pitch has been dropping significantly in recent years (Kristof 1995).Comparable studies are not available for pitch differences between Chinese men and women. The average pitch of Chinese female voices has probably dropped somewhat during the past few decades. At the same time, there may be pitch differences associated with different socio-cultural contexts for women, with formal situations, such as public speaking, dictating a slightly higher overall pitch and clearer enunciation. Different pitch ranges may also be operative for female bilingual speakers, with a higher, overall pitch and larger pitch range for speaking Chinese versus a non-tone language such as English. Future research is needed for a systematic investigation into these proposals.
While information is lacking on what is the average pitch of Chinese males and females, for speakers of (standard) Mandarin Chinese, Professor Chin-chuan Cheng's (1995) software, Speech Analysis for Windows, can very accurately identify the sex of the speaker. The sex-identification capability of the program is based on analyzing F0's along the time dimension in a given stretch of recorded speech and obtaining an average fundamental frequency (F0, perceived as pitch) for that sequence. From his test with numerous Mandarin speakers during the development of his software program, he arrives at the ultimate selection of 150 Hz as the F0 that distinguishes male and female Mandarin speakers. A speaker's utterance of Mandarin that yields an average above 150 Hz would be identified by the software as the speech of a female speaker, and anything at or below 150 Hz would be identified as that of a male speaker.2
Aside from average pitch differences, there are other pronunciation differences between Chinese men and women. Lehman et al. (1975:35), for instance, reported on the American Linguistics Delegation that visited the People's Republic of China in late 1974, and noted that while very few speakers had mastered the standard pronunciation of Putonghua, those who did fell into three groups: (1) some university professors, (2) some female high school teachers of Chinese, trained in Beijing, and (3) female guides as at museums and exhibition halls. Thus, on the whole, women were generally more sensitive to the prestige standard pronunciation than men, exhibiting patterns that are found in the U.S. and other countries (James 1996). In contrast, those who paid least attention to the standard pronunciation were younger men who were leading members of Revolutionary Committees in charge of educational institutions. It appears that men with position and authority may be in less need of elevating their status by learning correct, standard pronunciation. The power and prestige of China's national leaders have never been diminished because of strongly accented Putonghua. There may well be covert prestige for men of status not to spend hours trying to master standard Chinese pronunciation.
A particular style of speaking and pose from the Cultural Revolution period was also noted by the Delegation (Lehman et al. 1975:35):
| (1) | The precision of articulation of many of these female speakers was quite striking, and was associated with a very erect posture, compressed lips, and a bright, serious and earnest expression. It is worth noting that almost all announcers in China are women, who use a similar clipped and forcefully articulated style over the radio, and at stage performances; this style is imitated by little girls who announce school plays and dances. |
In Chan (1996), noted in connection with sound symbolism are two other cases of feminine speech style in Beijing Mandarin. One is Hu's (1991a) observation that school girls in Beijing possesses what he dubs a 'feminine accent' (女國音) that is, these girls produce the palatal 
as dental sibilants [ts, ts', s], or as more fronted palatals. Hu (p.51) explains that fronted palatals and dentals sound more 'fragile' and 'piercing' to Chinese ears, and so more 'feminine', while alveolars tend to be more 'blunt' and 'masculine.' Hu (p.53) further remarks, the requirement for girls and young women to display good manners by avoiding laughing and talking with their mouths wide open. Thus, socially, girls need to be lady-like, and this is accomplished by producing a more feminine form of articulation. The more fronted articulation produces a sound with higher acoustic frequency, thereby adding to an already higher average pitch of voice for females in general.3
The second study is Shen's (1987) sociolinguistic investigation of syllable onset, /w/. The labial approximant, [w], is one of the phonetic realizations of /w/, another is the labiodental approximant, 
. Shen's large-scale variation study reveals that is used significantly more frequently by female speakers than by male speakers. This variant, , which is in Beijing speech and not in Putonghua, is produced with spread lips, and the teeth and lower lip closer together. The sound thus produced has higher acoustic frequency than the plain labial approximant, [w], which is produced with lip-rounding and lip protrusion. The variant, which generally occurs in syllables wei, wen, weng, and optionally in wan (Hu 1991b), is undoubtedly perceived as more feminine and may have contributed to its emulation elsewhere in the country, including southern China in Guangzhou.4 While the labiodental variant is often used by female news broadcasters in China, male news broadcasters sometimes use it, also.
Interestingly, in Taiwan as well, one frequently hears news broadcasters using in their speech, and this is typically (though not exclusively) produced by females. Such production is not accidental, as one trainee for television news broadcasting in Taiwan recalls.5 In her news broadcasting class at TTV in 1989, trainees were separated by sex, with female trainees taught by female instructors (and presumably male trainees taught by male instructors). In her all-female class, the trainees were asked to repeat and imitate their instructor, who used in such words as yi wan 一萬 'ten thousand' and xinwen 新聞 'news'. Those who pronounced such words using the plain labial approximant, [w], were corrected by her. The instructor had consciously used the variant in those phonological contexts where she wanted the trainees to do likewise. In Taiwan and mainland China, news broadcasters are often females. Shih (1984:224) attributes the greater use of female announcers to their more standard pronunciation and clearer enunciation. In the United States and other English-speaking countries, in contrast, national news broadcasters are typically male, the reason undoubtedly being that for covering world news, male figures exude solemnity and authority. In fact, radio and television announcers tend to have a pitch of voice that is lower than the general population in order for them to sound authoritative (Ohala (1994:327). For Chinese female news broadcasters, they compensate by speaking with steadier pitch (less pitch flunctuations) and in a lower and deeper voice in announcing the news (Shih 1984:225).
Gender differences in pronunciation may also be studied in association with a particular communication style, such as sajiao (撒嬌), analyzed by Farris (1995) in present-day Taiwan's setting. The sajiao style, which she describes as the adorable petulance of a spoiled child or young woman who seeks material or immaterial benefit from an unwilling listener, is analyzed as being marked for the feminine gender. Farris (p.16) reports on a friend's observation of a very nasal style in young unmarried women's use of sajiao with their boyfriends. Farris argues that the sajiao style indicates women's indirect and informal power in Chinese society; at the same time, it serves as a means to create and maintain that form of power. For her general description of feminine speech at the phonological level, Farris (p.16-17) translates Shih (1984:224), excerpted in (2) below. Shih, who considers gender differences to be most salient at the phonological level, provides a description that is based on introspection, observations, as well as data collected from dormitories of male and female students in Taipei:
| (2) | The standard man's voice ... is inclined toward the low and heavy, thick and strong, while the standard woman's voice is inclined toward the young and immature, warm and respectful, sometimes having bashful overtones or even a petulant air (sajiao).6 ... Moreover, the more a woman's voice emphasizes natural and artificial feminine qualities, giving an impression of tenderness and warmth, the more it lacks authority. Whereas, a man's voice, which is low and deep, steady and calm, gives the impression of authority. |
On the question of sajiao specifically, Zhang (1995) prefers the definition in the Modern Chinese Dictionary (Beijing: Commercial Press, 1979), namely, to deliberately act like a spoiled child in front of someone because of the awareness of the other person's affection. Zhang observes that in both mainland China and Taiwan, sajiao is a communication style that is typically used by children to their parents (to refuse things demanded of them or to get permission to do things prohibited by them), and by adults to their lovers or spouses (as a kind of romantic play). Of particular importance here is that Zhang identifies two additional features of sajiao speech, namely, the prolongation of the vowels and softening of the consonants. Vowel lengthening is self-explanatory. For consonant softening, she describes three manners of articulation that would produce this. One, aspirated consonants may be articulated with less (i.e., weaker) aspiration) so that the aspirated-unaspirated distinction becomes blurred. Two, the contact between two articulators may be softer (less abrupt, perhaps?). Three, the contact between articulators may be shorter in duration. Zhang also notes that this consonant softening effect is most apparent with the dental sibilants, [ts], [ts'] and [s] (i.e. the aim is to reduce the stridency of these sibilants to make them sound less harsh - mc).
Often accompanying sajiao is the sentence-final particle, ma 嘛, a particle that is used to soften the tone of an utterance and is generally regarded as more typical of women's speech. (It should not be confused with the grammatical particle, ma 嗎, used in yes-no interrogative sentences.) In the sajiao style, the entire sentence is uttered slowly and when ma is added, the syllable is nasalized and noticeably lengthened. As Shen (1995) observes, such manner of pronunciation may be scorned as unmanly when it is produced in public by males, and given the derisive label of niangniangqiang 娘娘腔 'womanish, womanish accent.' The scorn is shown in (3) from Shen (with glosses added here). (3b) was originally produced by Shen's room-mate, who had also dismissed the male caller as someone untrustworthy, based solely on hearing him over the telephone.
| (3) | a. | Bié kū le!
Dà nánrén zěnme kěyǐ niángniángqiāng? don't cry PRT! big male-person how can this-way womanish-accent 'Stop crying! How can you, a big male, be so womanish?!' |
| b. | Gāngcái diànhuà nàge rénde niángniángqiāng zhēn ràng rén shòu bù liǎo. just-now phone-in that-CL person-PRT womanish make person endure not PRT 'That person just on the phone's womanish speech is really unbearable!' |
Niangniangqiang speech, as noted by Shen, is marked with high pitch and thinness of voice quality. It is a style that one normally associates with female speech and is not stereotypically male speech, which has lower pitch and deeper resonances. Adding of the softener ma particle with vowel lengthening and nasalization only further mocks such speech as womanish. Men do occasionally produce sentences with this ma particle, as it is by no means gender-exclusive. However, a man using it too frequently would certainly raise eye-brows, as in a case reported to the author of native speakers' amusement when a young man who had learned Chinese from his girlfriend loaded his sentences with ma!
It is worth noting that native speakers from the PRC have remarked on Taiwanese men sounding more effeminate than those on the mainland. For example, one female student who had never met anyone from Taiwan and only heard them over the telephone when they called long-distance to her father, identified some features that made these speakers sound effeminate to her: the Taiwanese men were very polite, spoke slowly and enunciated very distinctly, and delivered sentences with rising intonation where PRC males would not have.7 Thus, the more prototypical male speech should be relatively quicker in tempo than women's speech rate; their voices should be deep and low-pitched; their intonation should be flatter, steadier, and falling rather than rising. These characteristics combined together contribute to the perception of speech that is manly, spoken with authority, confidence, and decisiveness. It is not a wonder that grown men speaking in public with heavy nasalization, high pitch, slow speech rate and thin voice quality, as in sajiao or whining, would be scorned as being niangniangqiang, and deemed untrustworthy!8 Stereotypical Chinese male speech is succinct, direct, confident, and definite, in contradistinction to stereotypical women's speech (Shih 1984:221).
The above only scratches the surface of what we still need to learn about gender differences in speech production, as well as gender differences in attitudes toward prestige forms and particular manner of speaking by males versus females.9 Farris (1991:201) aptly states that cultural stereotypes play a crucial role in developing and maintaining gender differences, and that speech stereotypes in particular serve to characterize the way that native speakers perceive how men and women normally speak. These stereotypes reinforce what is expected of speakers and the roles they play in society. Much is yet unclear as to the degree to which speech stereotypes reflect actual language use, and to what extent the two diverge. Nor do we have a clear idea of what cultural differences may exist with respect to how well stereotypes accurately reflect gender differences in language use. There may, for instance, be stronger pressure in the Chinese environment for individuals to conform to social expectations, such that stereotypical behavior as cultural norms may dictate language behavior to a greater extent in Chinese society than in western, English-speaking society. Issues such as these do not find any quick, ready answer, and will be equally relevant in the following sections, even though they will not actually be raised.
NOTES (number 1 through 9)
* This paper is a much expanded version of my NACCL-9 talk in May 1997. My thanks to Cat Farris, Shou-hsin Teng, and Sam Wang, among others, for helping me obtain some sources. [BACK]
1. Studies on gender differences that are socially learned and not attributable to anatomical differences between the sexes are reported in Smith (1979:123ff), Graddol and Swann (1989:18ff), and Coates (1993:146ff), among others. [BACK]
2. Professor C.C. Cheng's pitch of voice ranges from 90 to 170 Hz in speaking Mandarin. His average is usually around 130 Hz (e-mail of 8/25/97), which would be unambiguously identified as 'Male.' Where his software have made errors in judgment, so have humans. [BACK]
3. A somewhat similar gender difference may also be present in the production of affricates in Cantonese. Hashimoto (1972:120, fn.8) mentions D.C. Lau of the University of London, who has observed that male speakers tend to palatalize [ts] and [ts'] more than female speakers. [BACK]
4. News broadcasters as well as ordinary women are imitating the sound. In the Cantonese television weekly series, Maahnfa Tung (Kaleidoscope) produced in Guangzhou (Chan forthcoming), in one of the episodes, one young adult female speaker's greeting on the telephone was "Wai, wai" (in high rising tone), said with
onset, which is not part of Cantonese phonology. In fact, the greeting, wai, in Guangzhou and Hong Kong Cantonese also has a high front vowel variant, wei,
that generally only young females use (Jian (1994/1996:137). [BACK]
5. Thanks go to Lin Huey (e-mail of 8/25/97), who further remarked that in 1992, when she was in a training program for Chinese-English interpreters, "one of our instructors also made this obvious distinction. She was working for the largest radio broadcasting company in Taiwan [at the time]." [BACK]
6. Shih (p.224) also mentions in an aside that the sajiao style is used in particular with the father or sweetheart. (The sajiao style is frequently acted out in television series produced in Taiwan.) [BACK]
7. This observation was made by a graduate student during my trip to China in summer 1996. [BACK]
8. For English, noteworthy is the case of Margaret Thatcher, former British Prime Minister, who needed to improve her public image by taking training lessons both to lower her average pitch and to reduce her pitch range, in addition to keeping a steady pitch (Romaine 1994:105). [BACK]
9. Cf. Lung (1997) on pre-1997 research results, based on a 1994 study, that suggest women in Hong Kong leading in acceptance and positive attitudes toward Putonghua, the non-local, prestigious language norm. Interestingly, Bourgerie's (1990) variation study of Cantonese finds women in Hong Kong tending toward innovative, non-prestige forms in the local dialect. [BACK]
[ TITLE PAGE | MID SECTION | FINAL SECTION ]
To Publications Page.
Copyright (c) 1997 Marjorie K.M. Chan. All rights reserved.
Created 1 December 1997. Last update: 4 April 2004 (to convert to UTF8-encoding).
URL: http://people.cohums.ohio-state.edu/chan9/articles/naccl9.htm