Gender and Grammar in Chinese With Implications for Language Universals

Gender and Grammar in Chinese With Implications for Language Universals
中文里的性别和语法 ——对语言共性的影响

Dr. Catherine S. Farris
University of Washington

First published in Modern China, Vol. 14, No. 3. (Jul., 1988), pp. 277-308

Translated by Maggie Li July 2019

The significance of the encoding of sex and gender for questions about the relationship between language and culture, and between language and thought, is just beginning to be exposed systematically.1

Language and gender research arising within the tradition of Western feminist theory, has overrelied on examples from Indo-European languages and cultures and has, moreover, not integrated this research very well with the current discourse on the nature and meaning of language in culture and society.2


In much of the literature on language and gender there is an assumed but not well documented relation between the extent of patriarchal bias in a culture and what has become known as sexism in language structure and language use. Whether language is merely a mirror of thought or helps structure thought, or, both, it is clear that there is no one-to-one correspondence between cultural and linguistic systems. Nonetheless, research on language and gender in Indo-European languages (specifically, the Germanic and Romance branches) suggests possible universals in the linguistic marking of the feminine, setting it off as Other, and letting the masculine form serve also as the generic or sex neutral one. This phenomenon has obvious implications for person perception, that is, for the ideas and beliefs we hold about women and men, and how these values get transformed into social process. If Man equals human in the linguistic and cultural codes, then Man will always be the standard against which Woman is measured, both cognitively and socially.

Chinese culture and society is well known for its extreme patriarchal bias, which twentieth century social and economic revolutions have left remarkably intact.3 Thus the Chinese language is an obvious candidate for an examination of the linguistic encoding of sex roles. Since Chinese, like English but unlike many Indo-European languages, does not have grammatical gender, gender is a covert category in both English and Chinese. In English, gender is marked by the pronominal system. But Chinese does not inflect or vary pronouns for gender, so the marking of gender is less obvious. Cultural context can mark a generic term for the masculine gender. For example, sunzi literally meaning “grandchild,” comes to mean specifically “grandson,” and a new form sunnu must be introduced for “granddaughter.” The previously sex-neutral term, sunzi, has acquired the semantic feature <+masculine> as a covert category.

Many people have wrestled with the problem of how language is related to culture and to thought. For me, the most useful way to conceptualize culture is as a learned and shared semiotic system of signs and meanings, of which language is the most significant, if not unique, subset (see Singer, 1978, for a discussion of a semiotic anthropology). As I will try to show in this article, signs acquire meaning only (1) in relation to other signs and (2) in the process of their deployment in the communicative context. Consider, for example, the Chinese word nu-shi, “lady,” “gentlewomen.” To understand the underlying gender asymmetries in the address system of which it is a part, we must know that xiansheng, “sir,” “gentleman,” is the masculine equivalent, not just of nu-shi, but also of taitai, “Mrs.,” and xiaojie, “Miss.” Or consider the phrase pei-qian-huo, literally, a “compensate-money-commodity,” referring to a daughter, for whom expenditures are wasted, as she will leave the natal home, taking a dowry with her. It is in the common usage of this phrase to refer to women that it covertly acquires the semantic feature <+feminine>. In a semiotic chain, these linguistic signs acquire new meanings in parole, that is, speech, which are then appropriated by members of the speech community and used by them in the cognitive organization of social categories. Many other examples of this process are evident in the Chinese language, which points to the need to identify systematically the dynamics of gender in Chinese and to integrate that investigation with theories of linguistic and cultural universals.

The research upon which this article is based was conducted in Taibei, Taiwan, the capital city of the Republic of China (ROC). It is the northern, urban Taiwanese speech community that I seek to characterize with respect to language and gender. Since the establishment of the ROC in 1911, there has been a national language policy aimed at linguistic unification and national development. In 1932, Guoyu (literally the “national language,” known in the West as Mandarin) was adopted as the national language, and today it is the official language of both the ROC and the PRC (where it is known as putonghua, the “common language”). The Taiwanese speech community is multilingual. The majority of the population is bilingual in Guoyu and at least one other language. For the majority of speakers, that second language is Taiwanese or Taiyu, a variety of southern Min, one of the seven major branches of the Chinese language family. Guoyu is the official language of government, the military, and education, while Taiyu, a vernacular language, is spoken in the home and temple, among the generation born before World War II, and in agriculture and petty commerce. Guoyu is much more evident in northern than in southern Taiwan, where it is the primary language used in urban areas, especially among people educated after the war. (For a more detailed look at the sociolinguistic situation in contemporary Taiwan, see Cheng, 1985, 1978; Jordan, 1969, 1973; Tse, 1982.)
本篇论文的研究是在中华民国的首都台北进行的,因此我在论证语言与性别的关系时使用的例子都是来自台湾北方城市的言语社区。为了实现语言的统一和国家的发展,中华民国自1911年成立以来便一直推行一项国家语言政策。1932年,国语(西方所熟知的普通话)地位被确立,至今仍然是中华民国及中华人民共和国的官方语言。台湾地区的言语社区是使用多种语言的,大部分人在掌握国语的同时还会至少另一种语言。最常见的第二语言是台语,即各种闽南语,闽南语是中文语系中七大主要分支之一。国语是政府、军事、教育等领域的官方语言,而台语作为方言,则在二战之前出生的一代人中间使用广泛,通常在家中,寺庙里,或是农业和小商业场景中得到运用。国语在台北的应用比台南更普及,在台北,国语是城市地区的主要语言,尤其是在战后受过教育的人口中间。(关于台湾当代社会语言学现状的更多详细描述,请参见Cheng, 1985, 1978; Jordan, 1969, 1973; Tse, 1982。)

In the sections below, I will first explain the three theoretical constructs utilized in the article, namely, linguistic gender, covert categories, and markedness. This explanation is followed by the presentation and analysis of 10 data sets. Sociocultural meanings of male and female in Chinese are shown to inform the linguistic and cultural codes, and motivate social behavior. I conclude with some thoughts on language, culture, and social process.

Linguistic Gender, Covert Categories, and the Theory of Marking

Linguistic theory distinguishes between grammatical and natural gender, regarding the former as structural or formal phenomena, and the latter as semantic or content phenomena. In an article on gender-marking in American English, Stanley (1977) explains the distinction between the two kinds of gender in linguistics. Grammatical gender refers to the three main noun classes, as recognized in Greek and Latin, namely, “feminine,” “neuter,” and “masculine.” Classification of nouns into three genders accounts for pronominal reference and adjectival concord. Theoretically, it is independent of sex. Natural gender, in contrast, “refers to the classification of nouns on the basis of biological sex, as female or male, or animate and inanimate” (Stanley, 1977:43). In this view, pronominal agreement in English is not a matter of gender concord, but, rather, is determined by natural or biological sex. However, Stanley (1977:44) asserts, and I concur, that the concept of natural gender “fails to accurately describe noun classifications and reference in American English.” I believe this is so because English,4 like Chinese, which also does not have grammatical gender in the classical sense, nonetheless possesses covert gender, which operates on surface structure phenomena in both languages.

I have borrowed the idea of covert categories from Whorf (1956). According to Whorf, language does not exist in words or morphemes, but in the patterned relations between them, which he termed “rapport.” Whorf was an early proponent of the semiotic claim that signs (linguistic or otherwise) acquire meaning only in relation to other signs. He asserts that “any scientific grammar is necessarily a deep analysis into relations,” and he distinguishes between overt and covert classes, or phenotypes and genotypes (Whorf, 1956:69).
隐性范畴的概念是我从Whorf(1956)的研究中借用的。Whorf认为,语言并不是依托于词语或词素而存在,而是依托于其遵循的模式化关系,Whorf将这种关系称为“默认关系”。Whorf是最早支持“符号(语言学或其他)只有在与其他符号产生关系时才会获得含义”这一理论的学者之一。他宣称,“任何科学化的语法都必然会对关系进行深入研究,”此外,他还为显性和隐性范畴,表现型和基因型做出了区分(Whorf, 1956:69)。

An overt category is one in which a formal mark is present, whereas a covert category is one in which the marking is present only in certain types of sentences, and not in every sentence in which a word belonging to the category occurs. Class membership is not apparent until it is referred to in one of these special sentences, then we find that the word belongs to a class requiring distinctive treatment. Whorf asserts that gender is a covert class in English. For example, in the English sentence “The nurse has an important role to play in patient care,” the class membership of “nurse” is not apparent (i.e., not overt). However, in a following sentence calling for an anaphoric pronoun, “She has more contact with the patient than the doctor does,” the covert gender category of “nurse”—<+feminine>—is manifested. English gender is a linguistic classification that as no overt mark actualized along with words of the class, but “operates through an invisible ‘central exchange’ of linkage bonds in such a way as to determine certain other words that mark the class” (Whorf, 1956:69). A covert concept such as gender in English is as definite from a meaning standpoint, as a lexical concept like “female.” It is not an analog of a word, however, but of a rapport system, and “awareness of it has an intuitive quality: we say that it is sensed rather than comprehended” (Whorf, 1956: 70). Both overt and covert categories are understood as conveying meaning. In English, this rapport of the covert category gender can be seen as the total pronominal linkage pressure of all the male class words, or all the female class words, that function in meditation, and not a lexical concept like “male” or “female.”
显性范畴拥有正式标记,而隐性范畴则在特定句子中才有标记存在,而非该词语出现的所有句子都有标记。直到在特定句子中发生指代作用,类别特性才会明显显现出来,因此我们才意识到这个词属于需要特殊处理的类别。Whorf坚称,性别在英文中属于隐性性别。例如,在英文句子“The nurse has an important role to play in patient care”中,“nurse-护士”一词的类别范畴并不明显(即非显性)。然而,在下面这个使用前指代词的句子“She has more contact with the patient than the doctor does”中,“nurse”的隐性性别范畴——<+阴性>,便体现出来。英文中对性别的语言学分类没有既定的显性标记系统,而是“通过关系之间的隐形‘核心交换’,从而推算出其他对该类别有标记作用的词语”(Whorf,1956:69)。在英文中,性别这一类的隐性概念从含义的角度来看,与“female-雌性”这个词汇概念一样,是确定无疑的。然而,这类词却不是因为相似性而被归为一类,而是因为其同属于一个默认系统,“要意识到其中差异,几乎需要靠直觉:因此我们只能感受而不是理解它们的词性”(Whorf,1956:70)。显性和隐性范畴都是用来传递含义的。在英文中,隐性性别类别的默认系统不是“雄性”或“雌性”一类的词汇概念,而是所有阳性或阴性类代词之间的紧密联系,潜伏于思维深处。

Whorf (1956:69) notes that in a language without sex gender (in the pronominal system), such as Chinese, thinking in terms of sex classification could not be of the same nature as in English gender: “it would presumably operate around a word, or a feeling, or a sexual image, or a symbol, or something else.” It is interesting to speculate what this rapport, or “central exchange” of linkage bonds in sex classification of Chinese, might be for native speakers. Below I hope to show that gender in Chinese is a covert category with few overt surface markings, which is nonetheless organized around a central and universal principle of “femaleness” or “maleness,” and which native speakers call upon to organize their thoughts about the sexes and to act toward and about, significant social others.

Whorf has been classified as a linguistic determinist, because of his emphasis on the formal and semantic uniqueness of individual languages, and the implications of that for our perceptions of reality. But I find in his emphasis on the deep relations of form and content in languages—which underlie and motivate surface phenomena—a systematic attempt to identify and explain what may actually turn out to be semantic universals, and their pervasive influence on all aspects of language. The concept of covert categories is a powerful tool with which to examine the phenomena of sex and gender in Chinese, and indeed in all languages.

In his discussions of markedness in natural languages, Greenberg (1966) demonstrates that the linguistic concept of marking has a high degree of generality in that it is applicable to the phonological, grammatical (morphosyntactic), and semantic aspects of language.5 He asserts that the tendency to take one of the members of an oppositional pair as unmarked so that it represents either the entire category or its opposite member par excellence is pervasive in human thinking. That is, the unmarked category is the culturally supposed “usual” case. For example, in logic we speak of the “truth value” of the set of which “truth” and “falsity” are the members. A “day” can either stand for a 24-hour period or indicate the opposite of “night.” Similarly, the Chinese word sunzi has two meanings: “grandchild,” and also the meaning “grandson.” Because <+masculine> is the unmarked category, the word sunzi is used chiefly but not exclusively to indicate <-feminine>. Because it is polysemous, sunzi is an ambiguous sign.
Roman Jakobson (cited in Greenberg) said of unmarked category that it has “zero” expression. To exemplify this, consider the phenomenon in Chinese of marking (i.e., expressing overtly) occupational terms for the feminine, for example: yisheng (“doctor”) versus nu yisheng (“woman doctor”). The process in Chinese is exactly parallel to the English gloss. That is, both “doctor” in English and yisheng in Chinese are not overtly marked for the masculine, nor are they examples of gendered nouns, that is, nouns with the semantic feature <+masculine>, as the words boy and husband are. Nonetheless, they are both covertly categorized with the semantic feature <+ masculine>.
Roman Jakobson(Greenberg在论文中引用)认为,未标记的类别具有“零”表达性。例如,中文里需要为女性职业词汇专门做标记(显性表达),如“医生”和“女医生”的对比。此过程在中文和英文中是完全平行对应的,即在两种语言中“医生”都没有阳性的显性标记,同时它们也不是自带语义特色的性别化词汇,如英文中的“boy”和“husband”。然而,它们都拥有隐性范畴的<+阳性>语义学特色。

Because the masculine is usually the unmarked term of a correlative pair, it assumes the role of the ambiguous term of the pair. Greenberg explains that the speaker interprets this form as unmarked or general and pervasive in reference (encompassing the pair) at the lexemic level, but as marked when the context demands it (1966:66). So, for example, “doctor,” or yisheng, means a physician of either sex, but male par excellence, because most doctors (in Chinese and American societies) are male, and the term evokes a male referent. That is, the semantic feature <+masculine> occupies canonical status in Chinese and English, and this fact is both a reflection of social reality and helps recreate that reality.

Social Identity Terms

From Claude Levi-Strauss’s (1969) study, The Elementary Structures of Kinship, to the writings of ethnosemanticists (see Tyler, 1969; Casson, 1981), anthropologists have viewed social identity terms, especially kin terms, as a linguistic window onto cultural forms. Thus we can usefully examine such terms in Chinese for what they may reveal about the sociocultural construction of the genders. Life-cycle words for females and males in Chinese appear symmetrical in their surface manifestations. As in English, these words are common nouns that must be marked with nu (“female”) or nan (“male”). Table 1, below, provides terms for male and female statuses in Chinese. The words for “baby” and “child” are symmetrical; to indicate “girl” or “boy,” one adds nu or nan. However, in usage, haizi (“child”) is not really symmetrical, because women continue to be referred to as nuhai, “girl-child,” until marriage, while men are referred as nanren (“male-person”) after adolescence and regardless of marriage status. The pair nan-de and nu-de (“male” and “female”) are symmetrical. However, while the word for “man” is nanren, the logical feminine equivalent is not nuren (“female-person”). That term is perceived by native speakers as having sexual connotations that nanren does not have. The situation is akin to the connotation “woman,” until recently, is alleged to have had in English.
从Claude Levi-Strauss(1969)的研究《亲属关系的基本结构》到民族语义学家们的著作(见Tyler,1969;Casson,1981),人类学家们一直以来都将社会身份词汇,尤其是亲属类词汇,视作窥探文化形式的语言学渠道。因此,我们可以有效地检测中文里的此类词汇,以研究它们可能体现出的性别的社会文化学构造。中文里与男性和女性的生命周期相关的词汇从表面上来看都是对称的。而在英文中,这些词汇则是必须以“女”和“男”做标记的普通名词。下方的表格一列出了表明中文里男性和女性地位的词汇。代表“婴儿”和“孩子”的词汇是对称的;要想指明性别,必须要加上“女”或“男”。但在实际应用中,“孩子”这个词却不是完全对称的,因为女性直到结婚为止,都被称作“女孩”,而男性则无论婚姻状况,在青春期之后便被称作“男人”。“男的”和“女的”是对称词汇,然而按照上面的逻辑,“男人”的对立面却不完全是“女人”。“女人”对于以中文为母语的群体而言,带有隐含的性含义,而“男人”却没有。英文中也有类似现象:“woman”一词直到最近都被认为拥有相似的隐含意。


Women’s sexuality is evidently much more problematic than men’s in Chinese society, as we see from the word for virgin, chunu, overtly marked with the female character. The logical masculine equivalent chunan, is relatively rare, and has completely different connotations from chunu. A woman is shamed if she is not a chunu upon marriage, while a man is embarrassed, and not quite a nanren, if he has reached social maturity and is still a chunan. Finally, the terms zhongnianren, “middle-aged person,” and laonianren, “old person,” are symmetrical in reference to men and women. However, the term zhuangnianren, “in one’s prime,” while not overtly marked masculine, is so marked covertly; the term is used only to refer to men.

Address and reference in Chinese shows gender asymmetry. Terms for women and girls inevitably encode their relative age and/ or marital status, while men’s status is usually unmarked for such features after puberty. Some reference terms for husband and wife are complementary, while others have no masculine equivalent. The marking of the feminine is shown in Table 2 below, which lists address and reference terms for women and men in contemporary Taiwanese society.

As noted in the introduction, the presence of xiaojie (“Miss”) and taitai (“Mrs.”) as possible oppositions to xiansheng (“Mr.,” “Sir,” “gentleman”) make it clear that nushi (“lady,” “gentle-woman”) is not a member of gender-symmetrical reference term set. The multiple terms for spouses, especially “wife,” are particularly, “inside-person” and “outside—one,” refer to the two spouse’s traditional spheres of influence, the so-called domestic/public domains (see Sanday, 1973). These words are the most formal terms that people use to refer to their spouses, followed by (in degree of formality) xiansheng/taitai (“Mr./Mrs.”) and laogong/laopo (literally, “old husband’s-father” and “old husband’s mother”***). The latter terms are used in casual conversation among friends to refer to one’s spouse.6 Although concubinage and secondary wives are illegal in the ROC today, many men continue to have “little wives” (xiao laopo) if they can afford them. There is no linguistic need for the masculine alternative, xiao laogong, “little husband,” since the practice does not occur. Only waiyu, “outside interest” (i.e., an extramarital affair) can be used to refer to either the man’s or the woman’s adulterous partner.
如前文介绍中提到的那样,由于“小姐(Miss)”和“太太(Mrs.)”都是“先生(Mr., Sir, gentleman)”的潜在对应词汇,因此“女士(lady, gentle-woman)”就不能归类于性别对称的指称系统中。指代配偶的诸多词汇中,尤其是“内人”和“外子”这一组,表明了传统观念中配偶双方的影响力范围,即女主内男主外(见Sanday,1973)。除了这一组最正式的词汇以外,人们还会使用(按照正式性递减)“先生/太太”和“老公/老婆”,最后一组通常是在朋友之间的闲聊中使用6。尽管中华民国的法律禁止妾制和一夫多妻制,许多男性仍然在经济情况允许的情况下养“小老婆”。鉴于对立的情况基本不会发生,从语言学角度而言,创造“小老公”这个词就没有必要。只有“外遇”这个词,可以用来同时指代男性和女性的婚内出轨对象。

Kin terms are an important area for analysis of covert gender in Chinese. The relative value that Chinese culture places on girl and boy babies is evident from, among other things, two traditional expressions. Nong zhang (literally, “to play with a sceptre” ) means “to make jade,” that is “to have a son.” Whereas, nong wa (literally “to make earthware”) means “to have a daughter.” As Perry Link (personal communication) points out, a respectful term to refer to another’s daughter is qianjin “a thousand pieces of money,” a term with positive connotations. However, the clear majority of address and reference terms mark the feminine term as the lower status one, or have negative connotations.
血缘关系类词汇对于分析中文里的隐性性别属性非常重要。中国文化赋予女孩和男孩的相对价值高低可以从很多方面看出来,但下面两个传统表达尤其能够说明这一点。“弄璋(“璋”为某种权杖玉器)”的含义是“生儿子”,而“弄瓦(“瓦”为陶泥土器)”的含义是“生女儿”。在一次谈话中Perry Link指出,用来指代女儿的尊称是“千金”,该词拥有积极的隐含意义。然而,大部分的称呼和指代词汇都将女性置于地位较低的一方,或者赋予其消极的隐含意义。

The terms for grandfather and grandmother, husband’s father and mother, father and mother, brother and sister, aunt and uncle are all gender-symmetrical. However, in kin terms for descending generations in Chinese, the masculine form serves par excellence as the generic or unmarked term, as can be seen in Table 3.

Ambiguously, the unmarked form can stand either for an entire class, or, here, in the case of descending generational terms, for just the masculine half of the pair. What has occurred to produce the asymmetrical references in these Chinese kin terms has to do with the covert coloring of a common noun, -zi, glossed: “seed,” “offspring” (no longer a free morpheme) with the semantic feature<+masculine>, so that dictionaries now gloss –zi as “son.” This covert principle can readily be seen from the forms—overtly marked with the female character—which complete the correlative pairs. In some contexts, the –zi forms still function as common nouns, so, for example, zisun can either be, generally, “posterity,” or, specifically, “male descendants.”

Another common form, -er, glossed as “child” (now a bound morpheme), combines with the common noun –zi to form the lexical item erzi, “son,” while nu (“female”) is prefixed to it to form the word “daughter,” nuer (the –zi is deleted now as redundant). The form –er remains a common noun in such combinations as ertong, glossed, “children” (in general). In certain chengyu or “proverbs,” it also takes on the feature of <+masculine>. For example, the proverb, er nu qing chang (literally “male female feelings long”) means “long is the love between man and woman.” This also occurs in the little used compound erma, “male horse,” “stallion” (the more common term for stallion is gongma).

Occupational Terms

As occurs in English and other European languages,7 in Chinese many common nouns referring to persons in various occupations covertly bear the semantic feature <+masculine>, so that intended feminine reference must be overtly marked with the nu (“female”) affix. In a discourse context, the “female” affix need to be mentioned only initially, that is, as new information, and is then dropped in subsequent reference. Table 4 provides a partial list of such occupational terms in Chinese.

The list could be extended to all titles that refer to positions usually occupied by men must be added the feminine affix in order to produce the feminine equivalent. The glosses indicate the parallel process that occurs, to a lesser extent, in English. Notice that in Chinese, as in English, the profession usually associated with women—prostitution—must be marked for the masculine. The terms jinu (“prostitute-female”) and nan-ji (“male-prostitute”), appear symmetrical, but jinu is a lexical item, while nan is prefixed to it produce the masculine alternative (the –nu dropping out, being logically incompatible).8 In addition, the indefinite plural form changji, as well as the cognate that nan-ji and jinu share (i.e., ji) are both bound forms meaning “singing-girl prostitute.” Further, in the writing system, the “woman” classifier nu clearly marks these words as <+feminine>.

The Covert Gender Marking of Ren (“Person”)

Like the bound forms -zi and -er, discussed above, ren, glossed, “person,” “people,” “humans,” or “other,” is overtly a common noun, just as the English glosses suggest. It is a free morpheme that, combining with other nouns and verbs, forms compound words and phrases. Usually, it functions as a generic term, but in some expressions it covertly acquires the semantic feature <+masculine>, or, less commonly, <+feminine>. Table 5 provides examples.

In the first cases above, ren functions as a generic noun, as the glosses indicate. An interesting exception is the use of renjia (literally “person-family”), which usually means “other,” but which is also used as a first person pronoun. Chao (1968) explains how renjia has come to be so used. It usually means “someone other than I (or we).” Both a special usage is “someone other than you,” or “those who are other than you,” and “thus gets to be only a rhetorical way of saying “I” (p. 645). Interesting for our purposes is the fact that renjia is a modest metaphoric distancing device used for the first person pronoun, and in this usage it is stereotypically girl-children and young women who so use it. Once again, we see that an overtly common noun has a covert gender, in this case, <+feminine>.

Remaining examples in Table 5 are all covertly marked for gender. In cai zi jia ren, “a brilliant man and a beautiful woman,” two common nouns (zi and ren) are covertly marked for gender, the first for <+masculine>, the second for <+feminine> (the word order is invariant, about which, see below). Caizi or cairen is glossed, “a man of ability,” “a talented person,” seemingly a generic or gender neutral term. However, the presence is the language of cainu, “a woman of ability,” forces the interpretation that, once again, ren and zi are par excellence <+masculine>, and only a specific context will allow them to be read as <+feminine>. In the proverb, ren jie di ling, “(when) the people are outstanding, the land is auspicious” (Link’s translation) or “the birth of heroes makes the place glorious” (Mathews’s translation), ren is once again, <+masculine>. Since women as heroines are practically absent from Chinese history, the overwhelming preponderance of heroes forces the interpretation that the ren in this case refers to men and not women. Another common proverb is ren jin ke fu, to describe a woman, meaning, “any person (ren) can act as her husband (fu),” that is, “a promiscuous woman.” The masculine alternative, “any woman can act as his wife,” is not possible. Finally, the classical term bu ren, or in modern Guoyu, bu neng ren, “cannot be a man,” that is “impotent,” refers of course, only to males.
表格五中其余的例子都拥有隐性性别标记。在“才子佳人”中,两个普通名词(“子”和“人”)都被赋予了隐性性别,前者是<+阳性>,后者是<+阴性>(两者的顺序是不可变的,原因见下文)。“才子”或“才人”原本的注释应该是中性的“有才能的人”,然而正是因为“才女”一词的存在,表明在此例中“人”和“子”是偏向于<+阳性>属性的,只有在特定语境下才能够被解读为<+阴性>。“人杰地灵”这个俗语,Link对此的翻译是“当人才杰出时,土地也会因此而带有灵性”,Mathews 的翻译是“英雄的出生能为土地增辉”,因此这里的“人”又属于<+阳性>。由于在中国历史中,女英雄少之又少,占绝对优势的男性英雄迫使这里对“人”的解读趋向于男性,而非女性。另一个常见的成语是“人尽可夫,”是用来形容淫乱的女性,而其男性对立面“人尽可妇”则并不存在。最后,传统词汇“不人”,或国语的说法“不能人”,指“性无能”,则当然是用来指代男性。

Ren can also bear the covert feature of <+feminine>, as may be seen in the last two examples in Table 5. Social roles typically associated with women allow ren to take a <+feminine> reading, as we see in meiren (a “marriage go-between”) and mei ren (a “beautiful woman”).

Pejorative Terms

The tendency in English for words descriptive of women gradually to acquire pejorative connotations, with a similar process not occurring with regard to masculine descriptors, has been documented by Schulz (1975). In an article on gender-linked differences in the Chinese language, Shih Yu-hwei (1984:216) points out that there are more pejorative or derogatory terms in Chinese referring to women than to men. Table 6 lists some of the most common gender-linked pejorative in currency in the Taiwanese speech community today.
Schulz(1975)在研究中详述了英文中描述女性的词汇是如何逐步获得贬损含义的,而相似的过程却没有发生在描述男性的词汇上。在一篇论证中文里与性别相关的差异的文章中,Shih Yu-hwei(1984:216)指出,中文里指代女性的贬义词要多于指代男性的。表格六列出了一些当代台湾言语社区中最常见的与性别相关的贬义词汇。

The first pejorative term for women, pofu, or po fu ma jie, “a shrewish woman curses (in) the streets” (literally, “shrew woman curse streets”), is a common description for women that has no masculine equivalent. The terms changshefu, “a garrulous or over-talkative woman,” and duozuipo, “a big-mouthed woman,” reflect other common stereotypes about women’s speech styles. Just as researchers in English speech communities have noted, women in Chinese society are thought to chatter meaninglessly; but one of the four womanly virtues is propriety in speech.9 These stereotypes are closely related to shi-san-dian (literally, “thirteen o’clock”), “a silly acting woman,”10 usually one who laughs and giggles inappropriately. Still another stereotype about women is evident in the phrase fu ren zhi jian, “a woman’s perspective,” that is, narrow and subjective. This is usually used to describe women, but it can also be used for men, and then the insult is greater, similar to telling a man in American society that he “thinks like a woman.”

Two terms used to describe women overlap with masculine terms, and the differences between the contrasting pairs are informative of the stereotypes of women and men in Chinese society. Biaozi, literally, “a prostitute,” is no longer used as a word for that profession, instead, it is a word used to curse women. Thus a masculine pejorative, similar to the English glosses, becomes biaozi yang-de, “raised by a bitch,” or biaozi erzi, “son of bitch.” This second set of terms refers to sexuality. A saonuren is “a woman of loose morals,” “a slut,” while a saolaotou, is “a lecher,” “an old wolf.” The two terms, seemingly equivalent, differ from one another in much the same way the English glosses differ from each other in the relative value society places on the unrestrained sexuality of men as opposed to women, that is, a loose, immoral woman versus a lecherous old man. Although men’s sexuality may be the object of a pejorative term, as in boqinglang, “heartless lover,” such terms for men also refer to their virility and sexual prowess, as in selang, “old wolf.” Thus linguistic asymmetry mirrors cultural values and social mores.
有两组用来形容女性的词汇与男性有重合部分,而其间的差别则充分表明了中国社会对男女的刻板偏见。“婊子”原本含义是“妓女”,而现在却是用来咒骂女性的脏字;与英文相似,贬义的男性对应词汇是“婊子养的”(英文“son of a bitch”)。第二组词汇与性欲相关,“骚女人”指道德观念低下的女人,而“骚老头”则代表“色鬼”,“色狼”。这个两个词表面上似乎对等,但与英文相同,其反应了社会对男性无节制的性欲与女性相比赋予的相对价值,即道德败坏的女人对应好色的老男人。当然男性的性欲也可能体现在一个贬义的词汇中,例如“薄情郎”。但此类指代男性的词汇也会暗示他们的生殖能力和性技巧,例如“色狼”。因此,语言学中的不对称现象映射了文化价值和社会传统。

Finally, we note the term yaojing (literally, “a supernatural spirit,” a “fox-spirit”), “a woman who bewitches men.” Vivien Ng (1987:64) points out that the fox-woman is a common supernatural spirit in Chinese folk tales who “is typically extremely beautiful and seductive and loves to prey on unsuspecting young men who are novices in the matter of love and sex.” She notes that “fox-possession tales are so universally known in China that, in the vernacular speech, seductresses are often referred to as ‘fox-spirits’” (Ng, 1987:64). In Taiwanese society today, the term yaojing is often used to curse the “other woman.” Because of a sexual double standard, married men in Taiwan who can afford it often have xiao laopo, “little wives,” with whom they usually set up a separate establishment. Or they will at least have a waiyu, “an outside interest.”11 But wives invariably blame the other woman for insinuating herself into the man’s affections, and thus this woman is said to bewitch men.
最后,我们再来看一下“妖精”这个词:迷惑男人的女人。Vivien Ng(1987:64)指出,“狐狸精”是中国民间传说中很常见的超自然生物,通常“美丽异常,擅长魅惑,并且喜欢勾引在爱情和性方面还是新手的、毫无戒备心的年轻男性。”Vivien还观察道,“狐狸精勾魂的故事在中国家喻户晓,因此在白话言语中,狐狸精就被用来指代勾引男人的女人。”在当代台湾社会,“妖精”一词经常被用来咒骂“第三者插足的女人”。由于性别上的双重标准,台湾的已婚男性通常会在经济允许的情况下找小老婆,并建立另外一个家庭。再不济,他们也至少会有外遇11。而家中妻子则不可避免地会撒怨气于另一位女性,怪其诱惑自己的丈夫,因此该女性是拥有勾引男性的特性的。

Word Order

Covert gender in Chinese also operates at the level of syntax. As Shih (1984:216) points out, normal—that is, unmarked—word order always places lexemes with the semantic feature <+masculine> first, <+feminine> second. In other words, there is a semantic hierarchy to nouns that is motivated by underlying metaphysical assumptions about the sexes. Hierarchical ordering of nouns also occurs in Navajo (Witherspoon, 1977), and, I suggest, in all languages. Table 7 is a list of contrast sets and common chengyu that refer to gender in Chinese culture and society. The subordinate status of females in Chinese society in apparent from the content or referent of the chengyu. Less apparent, because it appears so “natural,” is the invariant male-female word order,12 which signals, at the syntactic level—and thus reinforces at the semantic level—the symbolic subordination of the feminine.

Covert gender in word order also occurs in English, as the “naturalness” of the English glosses for the contrast sets makes clear.13 However, the word order constraints of covert gender in English are probably less binding than in Chinese. Thus “mother and father” and “girls and boys” are not infelicitous, whereas this order in Chinese always is. In fact, it simply does not occur, being so marked, for native speakers, that many perceive it as grammatically and not stylistically incorrect. Obviously, the classical injunction, fu zhe, hou ren ye (“women are those who come afterward”), still operates on the semantic field of Chinese language and culture today.
隐性性别在词序方面的体现也存在于英文中,这从表格中英文注释“自然而然”的“先男后女”顺序可以看出13。但是,英文中对于隐性性别的词序限制八成没有中文严苛。英文里如果说“mother and father”和“girls and boys”并不会显得不恰当,而在中文里这么说则显得突兀。事实上,这样的用法根本就不会出现,对于中文是母语的人来说,由于性别等级标记的根深蒂固,这样的表达不仅仅文体上欠妥,甚至算得上是语法错误。显然,传统的偏见“妇者,后人也”在当代的中国的语言和文化语义场中仍然根深蒂固。

Covert Gender in the General Lexicon

Covert gender in Chinese also operates in common nouns, that is, words without the semantic feature <+masculine> or <+feminine>, in contrast to gendered nouns, for example “husband/ wife” or “girl/boy.” The first set in this category to be discussed are words used as given names. The second set includes stative verbs (= predicate adjectives in English) and regular verbs. I will discuss each set below.

Given Names in Chinese

Unlike English, in which given names are no longer meaningful to the native speaker, people’s given names in Chinese are drawn from a subset of the content words in the lexicon. This occurs for a small number of female (but not male) names in English, for example Rose, April, May, Daisy, and so on. The majority of given names in English have no apparent content and moreover, most are, are Whorf noted, covertly <+feminine> or <+masculine>. For example, Betty, Cathy, Donna, and Helen are all covertly <+feminine>, while Tom, Dick, Harry, and Sam are all covertly <+masculine>. Only a small subset of given names in English can be used as male or female names, for example Leslie, Sydney, Terry. It is interesting, as Stevan Harrell (personal communication) points out, that many of these gender neutral given names in English acquire “sissy” connotations, and cease to be available as boys’ names.
英文中,对于母语者而言,名字已经不再具有实际意义,而中文名字则是从词库中选择了拥有具体含义的词语。同样的情形只适用于少数的女性英文名(而非男性),例如,Rose(玫瑰),April(四月),May(五月),和Daisy(雏菊)等。大部分的英文名没有明显的内容含义,除此以外,Whorf指出,它们大部分都是有隐性的<+阴性>或<+阳性>属性的。比如,“Betty”,“Cathy”,“Donna”,“Helen”等都带有隐含的<+阴性>属性,而“Tom”,“Dick”,“Harry”,“Sam”等则为<+阳性>。英文只有一小部分名字可以同时用做男性和女性的名字,例如“Leslie”,“Sydney”,以及“Terry”。有意思的是,正如Stevan Harrell指出的那样(某次谈话),英文里许多本来中性的名字带上了“女孩子气”,并且不再用作男生名。

Shih (1984: 217) notes that from the names parents in Chinese society choose for their daughters and sons we can see the hopes and expectations they have for their children, and how different these expectations are for girls and boys. Table 8 and Table 9 contain some of the common names for girls and boys in Chinese.

Many words used as given names in Chinese are not covertly gendered in other contexts of use (but xiong, “virile,” qiang, “strong,” xian, “refined,” and jiao, “delicate,” probably are.) However, when used as a person’s name, they are; native speakers generally know, when they hear or read a person’s name, what sex that person is.

Gender-Marked Verbs

The second set of words in this last category, as shown in Table 10, overlap with the set of covertly marked given names in that many of the latter function as native verbs. The words in Table 10 are marked for gender by common usage; that is, they are commonly used to refer to or describe males or (more usually) females. This list, as all the above lists, is by no means exhaustive, but merely meant to be representative.

The first word in this set, keai, “adorable,” “lovable,” has a broader descriptive range than the English gloss, and basically seems to be appropriate to describe anything that is diminutive, the relative size alone apparently taking on endearing connotations. All children can be described as keai, as well as small animals and insects, and also inanimate objects. While children of both sexes are often described as keai, at some times in early adolescence the term becomes covertly marked for reference to females, and boys are no longer described this way. In contrast, young unmarried women are often described as keai, and, indeed, consciously strive to elicit such as response by their dress and deportment.

The unmarked, polite formula for asking a person’s name is nin gui xing, da ming?, “your (polite) honorable surname (and) given name?” A young lady may also be asked merely for her fang ming, literally, “fragrant name,” that is “your given name?” This expression is used only to address a female. An informant explained to me that if one asked a man this question, he would think it a joke. This linguistic usage is reminiscent of the tendency in American universities (and other settings) for women professors to be referred to by their first names, while men professors are referred to by their last name (see, for example, Rubin, 1981). In both the Chinese and the American English example, the asymmetrical usage marks feminine the lower status address pattern.

As has been noted for English as well as other European languages (see Thorne et al., 1983, for references), Chinese has many terms to describe the way women look, act, dress, their body parts, and their sexuality. Related to yaojing, “a woman who bewitches men,” (as noted above), there is also yaoyan, a stative verb meaning “provocative,” “voluptuous.” It is used to describe, for instance, a woman who dresses in a deliberately seductive manner. A yaojing looks yaoyan and lures innocent husbands away.
与在英文和其他欧洲语言中观察到的一样(见Thorne et al.,1983),中文里也有许多形容女性外貌、行为、着装、身体部位,以及性欲的词汇。与“妖精”——“勾引男人的女人”一词相关的(如上文),还有“妖艳”,代表“引起性欲的”,“性感的”的状态动词。例如,它可以被用来描述故意穿着香艳的女性。“妖精”通常外表“妖艳”,并且勾引别人单纯无知的丈夫。

An obvious clue to covert gender in Chinese lies in the character or logographic writing system. In so-called phonetic compounds, the ancient Chinese developed a recursive rule for generating new characters. A classifier, drawn from a finite list, contributes a meaningful element, while the phonetic, any character, including already compounded ones, adds the phonetic element; that is, the newly coined character is homophonous, or nearly so, with it (see Kalgren, 1923/1974, for an introduction to the writing system). Shih (1984:215) reports that the Shuo Wen Jie Zi, a lexicon of the Han dynasty (A.D. 100), lists over 250 characters with the productive female (nu) classifier. Several of the words that we have already examined (yao, “bewitching,” jiao, “delicate,” xian, “refined”) are written with this classifier. One other lexical item worthy of mention is jidu or duji “to be jealous”; both parts of the compound lexeme are written with the nu classifier. It is said that women shan du; they are “good at jealousy,” because of the practice of men having little wives and outside interests. This is not to imply that men are never described as jidu, but its unmarked usage will be <+feminine>. As for the many characters written with the nu classifier, it may be that when these words were standardized, the characteristics to which they referred were conceived of as typical of women.14 The fact that men generally are responsible for dictionaries is the subject of another article. (See, for example, Wolfe, 1980, for a discussion of patriarchal bias in the study of diachronic semantics of Indo-European languages.)

The final words to consider in this set are sajiao, salai, and shua piqi. The first two are related to each other through their cognate form, sa, meaning “to disperse,” “let loose,” or “exhibit,” “display.” Sajiao has two related meanings: (1) “to show pettiness, as a spoilt child,” and (2) “to pretend to be angry or displeased, as coquettish young woman.” Salai means “to pretend to be injured.” It is closely related to shua piqi, meaning “to indulge one’s temper,” “to act angry intentionally.” This contrasts with fa piqi, meaning “to fly into a temper,” “to become enraged.” That is, the last two verbs contrasts as descriptions of affected versus genuine emotions. Sajiao and salai are conceived of by native actors as behaviors or communication styles that spoiled children of both sexes, and young (particularly unmarried) women engage in when they want to get their way from an unwilling parent/boyfriend/husband. These behaviors are at once recognized as consciously affected ones that children and women engage in for certain strategic goals, and are at the same time thought of as natural or intrinsic to the cognitive makeup of the people occupying these social identities. Thus sajiao and salai bear the semantic features <+feminine> and <+child> at a covert level, and male actors (but not women) will deny that men engage in such behavior.


The demonstrated pervasiveness of marking as a structural principle in linguistic performance suggests the possibility of underlying universals in the organization of thought. Languages as a cognitive system avails itself of the marking principle as it comes into contact with social reality. Language is not some externally imposed restraint on thought; rather, it is the (primary) means through which thought is given expression in context. The evidence from the English and Chinese languages powerfully suggests that masculine is the unmarked or canonical gender in most circumstances. But it is not a cognitive imperative that marks the feminine; rather, it is the observable conditions of women and men in culture and society that so marks femininity, however it is conceptualized from culture to culture.

How are we to understand the link between language, culture, and reality? With the aid of Whorf’s cogent discussion of covert categories in language, I have shown how a language such as Chinese, which makes very few semantic distinctions in gender at the level of grammar, nonetheless possesses a pervasive covert gender system at the lexical level. It can be seen that sociocultural meanings inform langue (the grammars of individual languages), are appropriated by the native actors for use in parole (speech), which in turn allows other factors to appropriate such meanings from the social context and use them in ongoing cognitive organization of social facts. Such is sociocultural process.

I suggest the two primordial words/characters nan nu constitute a rapport system in Chinese, holding the structurally diverse gendered elements together. All the meanings of masculinity and femininity, of what it means to be boy or girl, man or woman in Chinese society, are contained in these two irreducible morphemes.15 The “natural” fact of maleness and femaleness, mediated by sociocultural meanings, is encoded in language, and appropriated by actors to talk to and about, “men and women” (nan nu). There is sexism in language. As long as people have sexist beliefs that get transformed into sexist social practices, those meanings will be encoded via the lexicon in language, and transmitted to and potentially transformed by, new generations. As Witherspoon (1977:3) notes, language and culture are symbolic codes through which messages are transmitted and interpreted. “But, more than a code, culture is a set of conceptions of and orientations to the world, embodied in symbols and symbolic forms. Through the adoption of and adherence to particular concepts of and orientations to reality, human beings actually create the worlds within which they live, think, speak, and act.” Chinese metaphysical assumptions about the nature of women and men inform their linguistic and cultural codes, and motivate their social behavior. A comprehensive examination of those underlying assumptions is in order.


Author’s note:

Earlier drafts for this article were presented at the China Program Colloquium, Jackson School for International Studies, University of Washington, March 1987; the Association of Asian Studies Meeting Individual Paper Session, Boston, April 1987; and as the winning student essay at the Puget Sound Anthropological Society meeting, University of Washington, June 1987. I am grateful for participant comments. I benefited from critical reviews of the article by Carol Eastman, Stevan Harrell, William Boltz, Eugene Hunn, Anne Yue-Hashimoto, Zhang Huiyang, Perry Link, Gary Witherspoon, Greg Guldin, Dru Gladney, Allan Barr, and an anonymous referee for Modern China.
The article is part of a larger project, my unpublished Ph.D. dissertation (Farris, 1988), which is concerned with how the meanings of gender are linguistically encoded in Chinese, and how these meanings are learned as part of the sex role socialization process. The research has been supported by the Graduate Student Research Travel Fund (University of Washington, Department of Anthropology), a Woodrow Wilson Women Studies Research Grant, a Northeast Center for Research on Women Mentor-Scholar Dissertation Fellowship (University of Washington), and a Pacific Cultural Foundation Research Grant (Taiwan, ROC). For the data examined in this article, I have relied primarily on native sources, Shih (1984), Mathews Chinese-English Dictionary, the Ci Yuan, the Xiao Shuo, and conversations with native speakers. The calligraphy of Alice Liou Hsiao-hsiun graces the tables.
作者备注:1987年3月,在华盛顿大学国际研究学院的中国项目座谈会上,本论文的早期草稿得到呈现;1987年4月,位于波士顿的亚洲研究协会举办的个人论文研讨会;随后,1987年6月,本文在华盛顿大学普及桑人类学学会会议上被选为获奖学生论文。我对于所有与会者提出的反馈评论都心怀感激。我从Carol Eastman,Stevan Harrell,William Boltz,Eugene Hunn,Anne Yue-Hashimoto,Zhang Huiyang,Perry Link,Gary Witherspoon,Greg Guldin,Dru Gladney,及Allan Barr等学者的研究成果中受益良多,同时,还有一位匿名朋友为我提供了有关当代中国的信息。

本篇论文是更大项目的一部分,也是我未经发表的博士论文(Farris,1988),主要研究性别的含义是如何被编码进汉语当中的,以及这些含义是如何作为社会化过程中性别角色的一部分,被语言使用者所习得的。本研究获得了研究生研究旅行基金的支持(华盛顿大学人类学学院),伍德罗·威尔逊女性研究奖学金,东北中心女性研究导师学者论文奖学金(中华民国台湾),以及太平洋文化基金会研究奖学金(中华民国台湾)。对于本文中所包含的语言学数据,我主要参考了Shih(1984)的研究,《马修汉英词典》,《辞源》,以及与中文为母语者的谈话等。我有幸请到了Alice Liou Hsiao-hsiun进行论文里表格中的汉字书写。

Translator’s Note:

It is hereby declared that the references to the Republic of China and its capital, Taipei, in this article are from the original English author’s text and do not represent any views of the translator.




  1. The literature on language and gender generated by American scholars is quite different in its focus and theoretical assumptions from the European, and particularly French, feminist tradition. See Cameron (1985) for the first critical overview of the various theoretical bases for research on language and gender in the United States and France. For a recent collection of essays in the American tradition, complete with annotated bibliography of hundreds of papers, articles, books, theses, and dissertations dealing with this topics, see Thorne et al. (1983). A thorough review of the topic from a cross-cultural perspective is Philip Smith (1979).

    在语言和性别的研究方面,美国学者的关注点和理论假设与欧洲,尤其是法国的传统女性主义视角相比,差别很大。详见Cameron(1985)撰写的第一份关于美国和法国学者在语言和性别研究方面的理论基础综述。Thorne(1983)等学者的论文则呈现了采用美国传统学术角度的研究资料汇总,以及与该话题相关的带注解的数百篇论文、文章、书籍等。跨文化研究角度的话题综述可以参见Philip Smith(1979)的研究。

  2. An important exception to the Western language bias in language and gender studies is the research on Japanese women’s language. See Shibamoto (1985) for original research on gender-linked differences in syntactic choice in Japanese, as well as a review of the literature (much of it in Japanese) on women’s speech styles in Japanese. See also Penelope Brown’s (1979) unpublished Ph.D. dissertation on language and sex roles in a Tzeltal Mayan speech community.

    语言与性别研究中西方语言的偏见有一个很重要的例外现象,那就是对于日本女性语言的研究。详见Shibamoto(1985)关于日语中与性别相关的句法选择的分析,以及关于日本女性言语风格的研究综述(大部分由日语写成)。另外还有Penelope Brown(1979)未发表博士论文,研究了塔尔泽尔人的玛雅语社群中语言和性别角色。

  3. In labeling Chinese culture and society as “patriarchal,” I refer to its male supremacist ideology, which informs and buttresses masculine supremacy in the political economy of family, state, and society. By “remarkably intact” I mean that patriarchy’s persistence in China was unforeseen by feminist scholars who placed faith in the improvement of women’s status through either the socialist or the modernist transformation of Chinese society. The continuing patriarchal bias in contemporary Chinese society in Taiwan—despite the industrialization and urbanization that supposedly brings the sexes into a more equitable relationship—has been documented by Nora Chiang and Ku Yenlin (1985), Diamond (1975a, 1979), Farris (1986), Gallin (1984a, 1984b), Greenhalgh (1985), and others. Chinese women under socialism have also not been “liberated” from patriarchal bias, as a growing number of scholars concerned with mainland Chinese society has noted. See, for example, Croll (1978), Diamond (1975b), Stacey (1983), Wolf (1984), and others. See also Guldin’s (1986) interesting discussion of the “pseudo-emancipation” of Fujianese women in Hong Kong.
    我之所以将中国文化和社会总结为“父系”,是因为其大男子主义意识形态渗透并影响了家庭、国家,以及社会单位的政治经济氛围。“非常完整”的意思则是,女权学者们原本以为中国的社会主义转型或现代化转型能够为女性的地位带来实质改善,结果父权制却仍然顽固地在中国社会留存下来。尽管工业化和城市化理应将两性地位带入更加平等的关系中,但是中国当代社会里父权偏见仍然顽固持续着,这一现象被Nora Chiang和Ku Yenlin(1985),Diamond(1975a,1979),Farris(1986),Gallin(1984a,1984b),Greenhalgh(1985)以及其他学者的研究记录了下来。Guldin(1986)关于香港地区福建女性的“假解放”现象讨论也同样颇有趣味。

  4. The English language originally possessed grammatical gender, but that system disappeared during the Middle Ages when the inflectional morphology system in English collapsed. Gender distinctions survive in the morphological system of English, for example the suffix -ess is added to previously gender neutral nouns to mark them for feminine, as in waiter-waitress. See Baron (1986) for a fuller treatment of gender and grammar in English.

  5. Greenberg explains that the concepts of marking arose in Prague school phonology, in the context of the problem of neutralization and the archiphoneme. It was noticed that in certain environments the contrast between correlative sets (that is, groups of phonemes differing only in a single feature) was neutralized in that both could not occur. The archiphoneme—the unit defined by the common features—occurs in these environments. The feature which appears in these instances is the unmarked feature, and the contrasting feature, which does not occur, is the marked one. The unmarked feature is described by a term itself having a negative prefix, un-, while the marked feature lacks it. “It is as though the marked feature is a positive something, e.g., nasality, aspiration, while the unmarked feature is merely its lack” (Greenberg, 1966:14).

  6. Terms used to refer to one’s spouse in the ROC contrast with present day usage in the PRC in interesting ways. Fan Zhongying (1987) reports that airen (“loved one”) is used to refer to both sexes, instead of qizi (“wife”)/zhangfu (“husband”), or taitai (“Mrs.”)/xiansheng (“Mr.”). The introduction of airen supposedly signals the symbolic equality of the sexes in the new China. Allan Barr (personal communication) notes, however, that impressionistic evidence suggests airen is going out of fashion, probably because it was in artificially imposed term. Zhang Huiying (personal communication) points out that neiren (“inside person”) and waizi (“outside person”) are considered archaic in the PRC. However, Fan reports that in the countryside, one’s wife is referred to as jiali-de (“someone in the house”), even though the majority of women now work outside the home in “productive” labor. Wives are also referred to by their own children’s names plus ma (“mother”), that is, “someone’s mother” (Fan, 1978:16). Thus the same semantic message survives in the PRC as in the ROC.
    在中华民国和中华人民共和国,用来指代伴侣的词汇之间的差异别有趣味。Fan Zhongying(1987)在研究中指出,“爱人”一词可以代替“妻子/丈夫”或“太太/先生”,被用来指代夫妻双方。该词的引入本应该代表着新中国在性别平等方面的进步。在谈话中Allan Barr指出,从整体印象来看,“爱人”这个词正渐渐变得不再流行,大概是因为它本身就是生造词。Zhang Huiying在聊天中提及,“内人”和“外子”在中华人民共和国也已经陈旧弃用了。然而,Fan在研究中描述说,在农村地区,妻子会被称为“家里的”,尽管现在绝大部分的女性都会从事家庭以外的有价值劳动。对妻子的另一种称呼是以孩子的名字加上“妈”,即,“谁谁他妈”(Fan,1978:16)。因此,同样的语义场现象在中华民国和中华人民共和国都成立。
  7. See Hellingere (1984) for a discussion of occupational titles in English, Dutch, Swedish, Norwegian, Italian, French, and German, from a feminist language planning perspective. Hellinger notes that planners for the first four languages advocate a “generic strategy” for feminist language change, in which a neutral or a masculine term stand for the entire category (for example, “flight attendant” [neutral] or “chairman” [<+masculine>]). In contrast, for the latter three languages, language planners would employ a “visibility strategy,” in which productive morphological devices are used to derive feminine terms for example, Italian professore <+masculine>+ -essa=professoressa <+feminine>.
    Hellingere(1984)的研究从女权语言规划的角度关注英语、荷兰语、瑞典语、挪威语、意大利语、法语,以及德语里的职业名称。他指出,前四种语言倡导用“一般策略”推动女权主义变革,其中中性和阳性的词汇通常被用来代表整个类别(例如,英文中的“flight attendant[中性]”和“chairman[<+阳性>]”)。相比之下,后三种语言的规划者采取的是“可视性策略”,通过形态工具来产生阴性词汇,例如意大利语里的professore<+阳性>+ -essa=professoressa <+阴性>。

  8. I am grateful to William Boltz for pointing this out.
    我十分感谢William Boltz指出这一点。

  9. During the Han dynasty (206 B.C.-200 A.D.), in which the Confucian school of political philosophy gained the ascendancy, the nu jie (“Admonishments to Women”), advocated san cong si de, “the three obediences and the four virtues” for women. The three obediences are: in the natal home follow the father and brothers, in marriage follow the husband, in widowhood follow the son. The four womanly virtues are: womanly fidelity, physical charm, propriety in speech, and efficiency in work (cited in Shih, 1984).
  10. Zhang Huiying (personal communication) notes that shi-san-dian originates in the Wu dialect, where it can refer to male or female. According to Cheng (1985), Standard Chinese (“Mandarin”) as spoken on Taiwan has been influenced by speakers of southern Chinese dialects, most notably speakers of Wu, including the ruling Jiang family and the Shanghai capitalists who retreated to Taiwan after 1949.
    Zhang Huiying(私下谈话)指出,“十三点”是来自吴语的方言词汇,可以同时用在男性和女性身上。Cheng(1985)认为,在台湾地区使用的普通话已经被中国的南方方言所影响了,其中最明显的是吴语,尤其是1949年以后退居台湾的蒋氏家族和上海资本家。

    Zhang Huiying (personal communication) notes that shi-san-dian originates in the Wu dialect, where it can refer to male or female. According to Cheng (1985), Standard Chinese (“Mandarin”) as spoken on Taiwan has been influenced by speakers of southern Chinese dialects, most notably speakers of Wu, including the ruling Jiang family and the Shanghai capitalists who retreated to Taiwan after 1949.
    Zhang Huiying(私下谈话)指出,“十三点”是来自吴语的方言词汇,可以同时用在男性和女性身上。Cheng(1985)认为,在台湾地区使用的普通话已经被中国的南方方言所影响了,其中最明显的是吴语,尤其是1949年以后退居台湾的蒋氏家族和上海资本家。

  11. Married women, widows, and divorcees may also have a waiyu (see Arthur Wolf and Huang Chieh-shan, 1980). However, for a woman, a sexual liaison outside of matrimony definitely violates cultural ideals and social norms, whereas it is expected that men will consort with more than one woman, both before and after marriage.
    已婚女性、丧夫女性以及离婚女性也可能有外遇(见Arthur Wolf和Huang Chieh-shan1980年的研究)。然而,对于已婚女性而言,产生婚外性关系必然是违反文化范式和社会标准的,但男性在婚前和婚后都是允许与多个女性结交的。

  12. There is an important exception to the invariant male-female word order in Chinese, namely, the yin-yang terminology. Although in the above example (yang gang yin rou), “the masculine principle is hardness, the feminine, softness,” the order is consistent with the male-female order, the usual word order for this correlative set is yin-yang. Black (1985) examines how the masculine-feminine pair finds a place in the “metaphysical polarities associated with traditional Chinese cosmology.” While the feminine and masculine principles participate in yin-yang cosmology, they do not define it. Black agrees with Ortner (1974) that “conceptions of gender are themselves partly shaped by other and perhaps more fundamental categories of thought and experience,” and she posits that, in Chinese history “there is no warrant for assuming that gender was always present in the shaping of a cosmological system” (Black, 1985:189-190).

  13. English word order for matched pairs is often said to be phonetically motivated. That is, speakers have a cognitive preference for saving the shorter term first, so, for example: “bread and butter,” “salt and pepper,” “pain and suffering,” “man and woman.” This phonetic principle would explain “ladies and gentlemen” but not “husband and wife.” It is probably the case that the word order for matched pairs with the semantic features <+masculine> and <+feminine> is both phonetically and semantically motivated. Thus, “ladies” come first because of a linguistic and a cultural rule. “Husband and wife” is faithful to the semantic rule but not to the phonetic one, so we often get “man and wife” instead, which does not offend our aural or aesthetic sensibilities.
    英语词汇对的词序通常会被认为是由语音特征决定的,即,语言使用者从认知角度来说倾向于将较短的词汇放在前面,例如,英文里的“bread and butter-黄油面包/主要收入来源”,“salt and pepper-盐和胡椒”,“pain and suffering-痛苦与创伤”,以及“man and woman-男人女人”。语音原则能够解释“ladies and gentlemen-女士生先生们”的顺序,却不能解释“husband and wife-丈夫和妻子”的顺序。带有<+阳性>和<+阴性>语义场特征的词汇对中,其词序的确定通常与语音和语义场都有关系。因此,源于语言学和文化角色的规则,“ladies-女士”排在优先地位,“husband and wife-丈夫和妻子”遵循的是语义场规则而不是语音规则,人们也经常使用“man and wife”,如此一来,从听觉上和审美感知上都不会产生冒犯感。

  14. Shih (1984) asserts that the large number of characters written with the “woman” classifier (nü), including the character for surname (xing), as well as many kin terms, is evidence of a previous matriarchal society. Pejorative terms written with the “woman” classifier are taken as evidence of later encoding by the patriarchal-based society which followed (1984:215). While I cannot speak to the archaeological evidence for a previous matriarchal Chinese social system (but see Pearson and Underhill, 1987:815), I agree with William Bolz (personal communication) that the evidence from the writing system does not support this contention. An explanation closer at hand could be that birth is the experience most intimately connected with women, while the man’s contribution to reproduction has no phenomenologically verifiable basis. Thus, when the writing system was standardized, the “woman” classifier was employed in many kin terms, when it was necessary to disambiguate them, via the writing system, from homonyms.
    Shih(1984)宣称,带“女”字旁的大量汉字,其中包含姓氏以及许多代表亲属关系的词汇,都是证明先前母系社会存在的证据。而以“女”字为偏旁的贬义词汇,就能够证明后来演变出的父系社会是如何进行语言编码的(1984:215)。虽然我不能够通过考古学证据证明先前中国也存在母系社会系统,但是我同意William Bolz的观点,因为目前的书写系统无法完全支持这个观点。另一个现成的解释是,分娩经历是与女性联系最紧密的一种体验,而男性在繁衍后代这件事情上则没有什么可量化的实质性贡献。因此,当书写系统被标准化以后,“女”字旁便出现在许多跟亲属关系相关的词汇里,有必要通过书写系统消除其与同音异意词之间的歧义现象。

  15. As Perry Link (personal communication) points out, the traditionalists might argue that the written character nan—as distinguished from the morpheme nan—is reducible, to tian, “field” plus li, “strength.”
    Perry Link在个人谈话中说,传统主义者也许会这样争辩,“男”在作为汉字时,与作为词素时有所不同,是可以简化为“田”和“力”的。