essai

A metalinguistic approach to deconstructing the concepts of 'face' and

'politeness' in Chinese, English and Japanese*

Michael Haugh and Carl Hinze

Abstract

In this paper we investigate the concepts of 'face' and 'politeness'. We introduce a metalanguage which we believe is a framework for simplifying the analysis of 'face' and

'politeness'. This metalanguage is based on the observation that both 'face' and 'politeness' involve external evaluations of people. This common element is represented in the metalanguage as "what A shows A thinks of B" and "what B thinks A thinks of B". The implications of the metalanguage for the analysis of Chinese mian and lian ('face') and

English face are then discussed. This is followed by an analysis of examples of politeness in

English and teineisa/reigi-tadashisa ('politeness') in Japanese. We conclude that the metalanguage may be further developed for use in comparisons of 'face' and 'politeness' across cultures.

Keywords: Face, politeness, Chinese, English, Japanese, metalanguage

* The authors would like to thank Associate Professor Ping Chen, Dr Yasuko Obana, Dr

Simon Patton, the two anonymous reviewers and others for all their helpful comments.

Biographical note: Carl Hinze has completed his PhD in the School of Languages and Comparative Cultural Studies at the University of Queensland.  His research interests are the semantics of 'face' across cultures, and more generally sociolinguistics.

Michael Haugh has also completed his PhD in the School of Languages and Comparative Cultural Studies at the University of Queensland. His research interests are politeness and implicature, and more generally cross-cultural pragmatics.

The concept of 'face' has been used to explain various social phenomena by a wide range of social science researchers, including social anthropologists, pragmalinguists, sociolinguists, sociologists and psychologists. 'Face' has been applied in an academic sense as an explanatory mechanism in the study of politeness discourse (Brown & Levinson, 1987; Mao, 1994; Nwoye, 1992; Scollon & Scollon, 1995), compliance-gaining and request/persuasive

strategies (Leichty & Applegate, 1991), impression management (Tedeschi & Riess, 1981), negotiation and conflict management  (Trubisky, Ting-Toomey, Lin, 1991), courtroom dramas (Penman, 1990), management practice and organisational behaviour (Earley, 1997), and communicative competence (Lim, 1994).

As a consequence of this rather prolific application, the figurative use of the term 'face' is becoming increasingly popular in both academic and non-academic English discourse. The value of 'face' to numerous academic considerations has proliferated to such an extent that the academic usage of this term is being progressively dichotomised from its non-academic

usage. Beginning with Goffman's (1955) realisation of the potential of 'face' to explain interactional behaviour, 'face' has acquired a mosaic of complex academic explications and applications that are more and more removed from the term’s original figurative meaning. These complex academic renderings of 'face' have in turn been used to explain a range of complex social phenomena, including 'politeness'. However, there is considerable controversy in the literature as to the exact nature of both 'face' and 'politeness', and the scope of phenomena which they should properly encompass.

In this paper we propose a metalanguage which we believe can be used to deepen our understanding of the basic mechanisms underlying both 'face' and 'politeness' in languages such as Chinese, English and Japanese. In our analysis of examples of 'face' in Chinese and English, and 'politeness' in English and Japanese, it seems that an element which is common to all these phenomena is external evaluations of people. This common element lies at the heart of the metalanguage that we use to further our understanding of the scope of 'face' and

'politeness'.

1. Introducing a metalanguage for describing 'face' and 'politeness'¹

When interacting, most members of society are concerned about what others think of them.²

This concern is manifested in interactions ranging from conversations with our family and friends through to transient meetings with strangers.  This concern about what others think of us results from an omnipresent vulnerability to the fact that other people's perceptions of us can be incongruent with what we consider them to be.  For example, I think you, my colleague, think highly of my research.  However, when you are given the chance to review some of my work and do so very critically, what I consider you think of my research changes. We believe that this ubiquitous concern (about what others think of us) can be used as the basis for developing a metalanguage which facilitates a deeper understanding of the notions

of 'face' in Chinese and English, and 'politeness' in English and Japanese.

In order to describe and analyse the cross-cultural phenomena of 'face' and 'politeness' we believe that a metalanguage based on simple expressions from natural language rather than complex technical jargon avoids as much as is ever possible circularity and obscurity.  As John Lyons (1977) once put it, "…any formalism is parasitic upon the ordinary everyday use of language, in that it must be understood intuitively on the basis of ordinary language…" (p.12).  The metalanguage which we propose here has similarities to the Natural Semantic Metalanguage (NSM) (Wierzbicka, 1991, 1996), but it is crudely influenced by the NSM

rather than faithfully adhering to the complex theoretical principles of the NSM.  Although the metalanguage we are using in this paper is English, the words that we employ in the metalanguage are semantically simple, and therefore readily translatable across languages.

It should be noted that we are not proposing the basis for a model of interpersonal interaction.³   We will not be using the metalanguage to describe in detail the complicated cognitive and behavioural processes that constitute an interaction.  Nor are we attempting to describe how 'face' and 'politeness' are co-constructed in interactions.  Our intention is to apply the metalanguage in such a way as to create a simple architecture, on the one hand, for

analysing the Chinese concepts of mianzi and lian, and the English concept of face, and on the other hand, for explicating politeness in English and Japanese conceptualisations of

'politeness' such as teineisa and reigi-tadashisa.

Before outlining a detailed explanation of the metalanguage, let us develop a hypothetical example.  Michael already has an established friendship with Carl, and following from this Michael has the notion of "what Michael thinks Carl thinks of Michael".  Michael has developed this notion on the basis of Carl's previous behaviour towards Michael.  Carl has treated Michael's work with the respect that is to be expected from an academic colleague. Recently Carl was given some of Michael's research to review as part of a peer evaluation process. In Carl's review he condemned significant portions of Michael's work.  This behaviour indicated to Michael that Carl does not think highly of Michael's research skills. This is the notion of "what Carl shows he thinks of Michael".  Carl's condemnation of Michael's research gives rise to incongruence between "what Carl shows he thinks of Michael" and "what Michael thinks Carl thinks of Michael". Michael will therefore re- evaluate "what Michael thinks Carl thinks of Michael" such that Michael thinks that Carl no longer regards Michael as highly as he previously did.

Let us consider another hypothetical example where Carl is the Head of the Asian Languages Department and Michael is a graduate student.   Through Carl's behaviour towards Michael, Carl has shown that he considers Michael to be an unexceptional student.  On the basis of this Michael infers what Carl thinks of him, which is "what Michael thinks Carl thinks of Michael".  Recently Michael had a research paper accepted for publication in an internationally renowned academic journal.   Carl lauds Michael's achievement at a departmental meeting.  This is "what Carl shows Carl thinks of Michael".  There is now a disequilibrium between "what Carl shows Carl thinks of Michael" and "what Michael thinks Carl thinks of Michael". Michael therefore revises "what Michael thinks Carl thinks of Michael" such that Michael now thinks Carl thinks more highly of Michael than Carl did before.

Let us consider one more hypothetical situation in which Michael is meeting Carl for the first time.  Michael is at a party and he is introduced to Carl whom he has not met before.  Before Carl and Michael are introduced there is no basis for a notion of "what Michael thinks Carl thinks of Michael".  Throughout the interaction Carl curses and swears incessantly, a fact which may be deemed as "what Carl shows Carl thinks of Michael".  That is, Carl shows that he thinks that Michael is the kind of person with whom he can break the social norm of avoiding offensive language in public.  From this the notion of "what Michael thinks Carl thinks of Michael" arises. This newly formed notion will then inform future interaction between Carl and Michael.  However, it is forever poised for change resulting from potential disequilibrium between "what Carl shows Carl thinks of Michael" in future interaction and "what Michael thinks Carl thinks of Michael" at the commencement of that interaction.

The key aspect of these examples which we wish to focus upon is the relationship between "what Carl shows Carl thinks of Michael" and "what Michael thinks Carl thinks of Michael". This is because we believe these two notions are underlying constructs common to the concepts of 'face' and 'politeness' in Chinese, English and Japanese. These two notions form the core of the metalanguage as depicted in Figure One.

Figure One: A simple architecture for describing 'face' and 'politeness' As we suggested above, the key aspects of this diagram are the constructs of "what A shows

A thinks of B" and "what B thinks A thinks of B".  "What A shows A thinks of B" is a behavioural construct (that is, A is showing something).⁴   It is an aspect of A's behaviour.  In example one, Carl criticizes Michael's research. This in terms of the metalanguage is "what A shows A thinks of B".  In simple terms, Carl (A) says he (A) thinks Michael's (B) research is not good.  Since Carl has criticized Michael's research, Michael thinks that Carl thinks that

Michael's research is not good.  This in terms of the metalanguage is "What B thinks A thinks of B".  This is a cognitive construct (that is, B is thinking something).

For the example outlined above, when Carl shows what Carl thinks of Michael, Carl is showing what Carl thinks of Michael's research, which is just one dimension of Michael.  In turn "what Michael thinks Carl thinks of Michael" is more specifically "what Michael thinks Carl thinks of Michael's research".  In other words, "what Carl shows Carl thinks of Michael" and "what Michael thinks Carl thinks of Michael" do not necessarily pertain to Michael as a person or everything about Michael, but rather they can denote particular dimensions of Michael that are salient at that point in the interaction. Thus "what A shows A thinks of B"

and "what B thinks A thinks of B" are constructs that can pertain to various dimensions of B, such as B's competence, B's physical appearance, B's wealth or B's moral integrity, among others.

In Figure One the various constructs represented in the "cognitive clouds" are a non- exhaustive list of the various cognitive constructs that can potentially influence "what A shows A thinks of B" and "what B thinks A thinks of B".⁵   "What A shows A thinks of B" is influenced by A's thought cloud, while "what B thinks A thinks of B" is influenced by B's thought cloud and "what A shows A thinks of B".  Among the constructs that can influence

"what A shows A thinks of B" are "what A thinks of B", "what A thinks A should (show A) think(s) of B", "what A thinks of social norms/values/attributes/principles", and "what A thinks A shows A thinks of B".⁶

In the first example, "what A thinks of B" is what Carl thinks of Michael's research, while "what A thinks A should think of B" is what Carl thinks he should think of Michael's research.  "What A thinks A should (show A) think(s) of B" and "what A thinks of B" are sometimes different, but at other times may be relatively indistinguishable.  "What A thinks of social norms/values/attributes/principles" relates to what A thinks of such things as social norms (for example, apologizing when you bump into someone by accident), social values (for example, the importance of filial piety), social attributes (for example, the importance of wealth), and social principles (for example, treating others as you would want them to treat yourself).  In example three, Carl thinks it unnecessary to abide by the social norm that one does not use offensive language when meeting someone for the first time.  "What A thinks A

shows A thinks of B"⁷ is Carl's perception of what he is showing he thinks of Michael.

Among the constructs that can influence "what B thinks A thinks of B" are "what A shows A thinks of B", "what B thinks A should (show A) think(s) of B", "what B thinks B should think of B", "what B thinks of social norms/values/attributes/principles" and "what B thinks of B" and so on.  In the second example, "what A shows A thinks of B" is what Carl shows he

thinks of Michael's ability as a postgraduate student.  "What B thinks A should (show A) think(s) of B" and "what B thinks B should think of B" are what Michael thinks Carl should (show Carl) think(s) of Michael's ability and what Michael thinks he should think of his ability respectively.  In this example, "what B thinks of social norms/values/attributes/principles" concerns what Michael thinks of the social norms/values/attributes/principles that are associated with evaluating the ability of a postgraduate student.  One would expect that there would be some overlap between "what A thinks of social norms/values/attributes/principles" and "what B thinks of social norms/values/attributes/principles", some kind of 'shared' or 'common' knowledge, but this is

not an issue which we deal with in this paper.  "What B thinks of B" is what Michael thinks of himself, which is Michael's self construal.

In example one there was an incongruity between what Carl shows Carl thinks of Michael's research ("what A shows A thinks of B") and what Michael thinks Carl thinks of Michael's research ("what B thinks A thinks of B") arising from Carl's condemnation of Michael's research. In a pre-existing relationship, such as the one in this example, "what Michael thinks Carl thinks of Michael" has been formed on the basis of what Carl has shown Carl thinks of Michael in their past interactions. However, at any point in interactions between Carl and Michael there is potential for disequilibrium to arise between what Carl shows Carl thinks of Michael's research and what Michael thinks Carl thinks of Michael's research. A

disequilibrium, like that in the example above, is resolved by Michael altering what he thinks

Carl thinks of his research.

In example two, what Carl shows Carl thinks of Michael ("what A shows A thinks of B") is determined on the basis of a dimension of Michael, which is his ability as a postgraduate student.  Carl shows that Carl thinks highly of Michael's ability by lauding his achievements to others in the department.  Prior to this, what Michael thinks Carl thinks of Michael's ability was informed by Carl's previous behaviour towards Michael.  Carl had shown that he thought of Michael's ability as nothing special.  In the new interaction the disequilibrium between what Carl shows Carl thinks of Michael's ability and what Michael thinks Carl thinks of Michael's ability is resolved by Michael coming to a new understanding of what Carl thinks

of Michael's ability.

From examples one and two we can see that there are two possible kinds of disequilibrium between "what A shows A thinks of B" and "what B thinks A thinks of B".  In example one there was a negative disequilibrium. "What A shows A thinks of B" is worse than "what B thinks A thinks of B" before the event took place. In order to re-establish equilibrium between "what A shows A thinks of B" and "what B thinks A thinks of B", B has to change "what B thinks A thinks of B". It is now worse than before. In example two there was a positive disequilibrium. "What A shows A thinks of B" is better than "what B thinks A thinks of B" before the event took place. In order to re-establish equilibrium between "what A shows A thinks of B" and "what B thinks A thinks of B" B has to change "what B thinks A thinks of B" such that it is better than before.  If "what A shows A thinks of B" is neither better nor worse than "what B thinks A thinks of B" there is a state of equilibrium. This is, however, constantly vulnerable to disruption by future events.

In our analyses of 'face' in Chinese and English we have found that what is common to these culture-specific constructs is that they emerge from the dynamic relationship between "what B thinks A thinks of B" and "what A shows A thinks of B".  We believe that 'face' can be examined with the assistance of the two key elements of Figure One, "what A shows A think of B" and "what B thinks A thinks of B".  In the next section we will see how "what A shows A thinks of B" and "what B thinks A thinks of B" can facilitate our understanding of the collocational possibilities of 'face' in Chinese and English.  [We will also note that 'face' must be distinguished from construals of self.  While aspects of self construals influence 'face', they are not one and the same.]

In analyses of other social phenomena such as 'politeness', different aspects of this metalanguage need to be foregrounded.  In Section Three we will see how "what A shows A thinks of B" and "what B thinks A thinks of B" can be combined with "what A should show A thinks of B" to explain the underlying dynamics of 'politeness' in English and Japanese.

Throughout the remainder of this paper we will be concentrating on describing the 'face' of just one person at a point in an interaction, and describing one person being 'polite' to another at a particular point in an interaction; that is, B's 'face' and A being 'polite' to B.  In order to describe 'face' and 'politeness' as concepts the first step is to observe one person's 'face' and one person being 'polite' to another person.  Arising from this first step one could pursue an understanding of the complex dynamics of interpersonal interaction, but this is beyond the

scope of this paper.⁸

2. The metalanguage and 'face' in Chinese and English

2.1 The metalanguage and 'face' in Modern Standard Chinese Discourse

_{It is generally considered axiomatic that the figurative use of the word face was borrowed into}

English from Chinese in the late 19^th century (Ervin-Tripp et al, 1995: 45-7). More precisely,

'face' (in the relevant sense) is a literal translation of the two Chinese characters mian and lian (Mao, 1994: 454). Therefore, it is with a degree of reluctance that we employ "face" to describe mian, lian and related Chinese terms. After all, 'face' in this sense owes its existence in English to Chinese. However, for the sake of convenience, the Chinese terms will be referred to here as "face in Chinese".

In their study of 'risk', Fillmore and Atkins (1992) advocated the benefits of examining a concept on the basis of collocations. Following this, Ervin-Tripp et al (1995) took the first steps towards an analysis of the concept of 'face' based on collocations. In this section, we will apply the metalanguage to make a preliminary analysis of 'face' in Chinese based on its

collocational possibilities. We do not have the scope in this paper to offer an in-depth analysis of every collocational possibility of 'face' in Chinese. However, we will group common Chinese expressions into four main groups and various sub-groups and attempt a rudimentary analysis of a collection of these expressions using the metalanguage and excerpts from our data.  The object of this study is not to make a detailed empirical analysis of Chinese 'face',

but to show that the metalanguage can be used to explain various Chinese 'face' expressions. Indeed, a comprehensive analysis of Chinese 'face' expressions is something that has been attempted, to various degrees, by previous researchers (see, for example, Chu, 1983; Cole,

1989; Cheng, 1986; Hinze, 2002; Hsu, 1996; Hu, 1944; Huang, 1986, La Barre, 1946) and is something that can be further pursued in future research.

Unlike Indo-European languages, Asian languages have a wide range of expressions

involving 'face' terms (Ervin-Tripp et al, 1995: 48). In this section, we will focus on particular Chinese expressions. We are using data compiled from a survey responded to by 42 native speakers of Chinese. We asked the respondents what various expressions involving 'face' terms mean and requested them to make sentences using these expressions. Additional data is sourced from text searches of ten contemporary Chinese short stories by renowned author

Wang Shuo,⁹ and a selection of other contemporary Chinese short stories. These two data sources have provided over 300 instances of the use of Chinese expressions involving 'face' terms.

Chinese expressions involving 'face' terms can be categorized into four main groups: negative changes in, and states of, 'face'; positive changes in, and states of, 'face'; managing one's own and others' 'face'; and attitudes towards or judgments about one's own and others' 'face'. The groups can be further categorized into various subgroups. For the sake of brevity, we will not outline all of these here (see Hinze, 2002). However, we will examine a selection of expressions from all four groups and indicate how the metalanguage can be applied to explain these expressions.

2.1.1 Negative changes in, and states of, 'face' in Chinese

There are a number of expressions in Modern Standard Chinese that are used to describe negative changes in, and states of, 'face'. Here, we will select only a couple of these expressions and show how the metalanguage can be used to explain them.

Losing mian/ lian

Diu mianzi ('to lose mianzi')

Ervin-Tripp et al (1995: 56) assert that diu mianzi "is used to describe very minor losses of face." They add, "A specialist may diu mianzi if he/she is unable to answer a question in his/her field of specialization." Cheng (1986: 335) declares that "to lose lian (diu lian) is far more serious than to lose mianzi (diu mianzi) for to lose lian means dishonour and disgrace, while to lose mianzi

means merely that one's honour is not honoured or one's honour is not recognized." (See also

Earley, 1997; Kornacki, 1995; Swi, 1997: 6). Consider the following examples:

(1) Jintian shangxue shi dang zhe Zhang laoshi de mian shuai le yi jiao, zhen diu mianzi.¹⁰

'On my way to class this morning, I slipped over in front of Teacher Zhang. It was such a loss of

mianzi.'

(2) Gei bie ren song liwu de shihou, bu neng song tai pianyi de dongxi, fouze ni yiding diu mianzi.

'When you give a gift to a friend, you can’t give them a cheap one, otherwise you will certainly lose mianzi.'

(3) Ta zheng chuixu ta de jisuanji jishu ruhe ruhe liao bu de, bei ta de tongxue dangchang chaming bu hege. Ta zhen diu mianzi.

'He bragged about his accounting skills but, in front of his classmates, they were found wanting. He really lost mianzi.'

(4) Shang yici kaoshi zhihou wo zhi kao le 50 fen, ke qitaren quan jige le, zhen diu mianzi.

'In the last exam, I only scored 50% but everyone else did very well. It was a real loss of mianzi.' (5) "Ni bie juede diu mianzi, zan mei shenme bu hao yisi de. Lai zher chifan jiu shi rang tamen

cihou de, zan hua le qian bu neng mai qisheng." (Wang Shuo, Gei wo dingzhu [Don’t Give In], p.18).

"Don’t feel embarrassed (lit, don’t feel you have lost mianzi). We don’t have anything to feel embarrassed about. When we come here to eat we are paying to be served by them. We will not pay to be angered."

Diu lian ('to lose lian')

Hu (1944: 46) construes diu lian as "a condemnation by the group for immoral or socially disagreeable behaviour." A number of subsequent studies have upheld and extended Hu's original point (see, Cheng, 1986; Mao, 1994; Ervin-Tripp et al, 1995; Kornacki, 1995; Swi, 1997). Some scholars, however, have preferred a description of diu lian that accredits somewhat less importance to issues of morality (see, for example, Hinze, 2002; Ho, 1994: 278-279; Hsu, 1996). Nevertheless, it is not within the scope of this paper to pursue this particular controversy. Let us consider the following examples:

(6) Ni conglai meiyou name diu guo lian, zai zhong mu kuikui zhi xia di san xia si de qi qiu erqie haoqiu zuoyong, nar shi ni de chulian dui ma? (Wang Shuo, Wo shi 'lang' [I am a 'wolf'], p. 9.)

'You have never before lost so much lian. You helplessly begged her in such a servile way. She is your first love, isn't she?'

(7) Meiguo zao dao le diu lian de shibai. (Wang Shuo, Dongwu xiongmeng [Ferocious animals], p. 5).

'America suffered a lian-losing defeat.'

(8) Congqian, wo de fumuqin shi bu zhun jiali de ren tiqi ta de, renwei ta diu le women jia de lian, muqin shuo. (Gu Qian (2000) Daolu [The Path], found in Hua Cheng, 1, p. 135.)

'Mother said that, in the past, my parents forbade the family from mentioning her because they thought she had lost lian for our family.'

(8) Weihun de nuer huaileyun, liangwei zuo fumu de juede nuer gei ziji diu jin le lian, rang tamen zai ren qian tai bu qitou lai. (Zhao (1995), p.12, cited in Hsu, 1996: 88).

'When the unmarried daughter became pregnant, her parents felt that she had lost all their lian and made them unable to raise their heads before people.'

Although we claim that all of the examples listed above can be explained using the metalanguage, we will analyze only two examples in detail. In example (1) the respondent considers that she has tarnished her image in the eyes of Teacher Zhang. In this example, the respondent thinks that she will be thought less of by her teacher as a consequence of her clumsy behaviour. This is quite a special case as there is no element of "what A shows A thinks of B". The teacher (A) has not shown the respondent (B) what he thinks of her. However, the respondent has unilaterally

declared a loss of mianzi for herself in the knowledge that her teacher will most likely think less of her as a consequence of observing her clumsiness. The respondent proclaims a loss of mianzi for herself because what she thinks Teacher Zhang thinks of her is now less than what she

thought Teacher Zhang thought of her prior to the incident. That is, "what B thinks A thinks of B" is now worse than "what B thinks A thinks of B" at the beginning of the incident. Consequently, the respondent has suffered a self-proclaimed loss of mianzi. Indeed, it is more than likely that

this loss of mianzi is inextricably linked to the respondent's expectations of what Teacher Zhang will say about her clumsiness ("What A shows A thinks of B" or, more specifically, "what A can

show A thinks of B").

In example (6), the naval commander is scolding the young soldier for appearing weak and emotional in public. The young soldier has asked his girlfriend to marry him but she has refused, so he has resorted to begging her in front of everyone to receive his hand in marriage. The naval commander considers this behaviour to be a major loss of lian.

The naval commander (A) has explicitly stated his disapproval of the young soldier's (B's) servile behaviour. Prior to this incident, the commander favoured the soldier. This is "what B thinks A thinks of B" at the commencement of the scenario. However, the commander has now shown that he thinks less of the soldier. In fact, he has actually declared that the soldier has lost lian. This is "what A shows A thinks of B". What the commander (A) now shows he thinks of the soldier (B) is worse than what the soldier (B) thought the commander (A) thought of him at the beginning of the incident. The soldier subsequently suffers a loss of lian

and will have to make a negative adjustment in what he thinks the commander thinks of him ("What B thinks A thinks of B") so as to determine the nature of his lian in future interactions with the commander. An important point to note is that "A" for the purposes of lian and mianzi could be one person (like the commander in this example) or a group of people (such as the soldier's comrades). We can see from this that in applying the metalangage to Chinese

'face' constructs, "A" is essentially an audience that witnesses B's behaviour or will potentially come to know about B's behaviour.

2.1.2 Positive changes in, and states of, 'face' in Chinese

Modern Standard Chinese commonly employs expressions to indicate that a person's mian, lian (and related terms) have been positively affected. Again, we will take a brief look at just two of these expressions and use the metalanguage to explain them.

Lustre on one's mian/lian

Mianzi shang you guang(cai) ('lustre on one's mianzi')

Mianzi shang you guang(cai) ('lustre on one's mianzi') is quite often used interchangeably with the other Chinese expressions pertaining to positive changes in, and states of, a person's mian and lian. According to Kornacki (1995: 69), mianzi shang you guang(cai) is used when many people cannot but admit that there is not only one, but many different things which make a person 'stand out'. Huang (1986: 60) points out that "mianzi shang you guangcai means that one's status has been confirmed, and self-respect enlarged." Consider the following examples:

(9) Jiang Li zhongyu zai gechang bisai zhong huosheng dabai duishou, women zhei qun pengyou ye mianzi shang you guangcai.¹¹

'Jiang Li finally defeated her opponent in the singing competition, so our group of friends shared in her glory (lit., our group of friends had increased lustre on our mianzi).'

(10) Guojia fuqiang le, renmin mianzi shang ye you guangcai le.

'With our country becoming rich and powerful, there is glory on the mianzi of the people (lit., there is also increased lustre on the mianzi of the people).'

(11) Ta qizi gei ta sheng le ge pang xiaozi, ta ke mianzi shang you guangcai le!

'When his wife gave birth to a healthy baby boy, he really had lustre on his mianzi.'

Lian shang you guang(cai)/ Lian shang zeng guang ('lustre on one's lian'/ 'increased lustre on one's lian')

Most scholars have adopted the notion that lian is not subject to positive influence (see

Brunner and You, 1988; Earley, 1997: 60; Ho, 1976; Hu, 1944; Huang, 1986: Kornacki,

1995: 81). However, this stance does not appear to be entirely accurate. Expressions such as lian shang you guang(cai) ('lustre on one's lian') and lian shang zeng guang ('increased lustre on one's lian') which pertain to positive changes in lian are used quite commonly in Modern Standard Chinese. Consider the following sentence:

(12) Zhongguo yundong yuan zai Aoyunhui de le name duo jinpai, women zhenshi lian shang zeng guang le.

'When the Chinese Olympic Team won so many gold medals at the Olympics, we felt that there was increased lustre on our lian.'

(13) Xiao bai ren ye lian shang you guang le. Ta wen, "Wo zhen shi zhei yang me?" (Wang

Shuo, Ni bu shi yi ge su ren, p. 13).

'The small white man had increased lustre on his lian. He questioned, "Am I really?'" (14) Wo de didi shouyu yi deng jiang, de rongyu, zhen lian shang zeng le guang.

'When my younger brother received first prize, he really won credit and increased the lustre on his lian.'

All of the examples listed in this subsection can be examined with reference to the metalanguage. Again, we will analyze only two examples in detail. The respondent in example (9) explained that she felt 'increased lustre on her mianzi' when her friend achieved success in a singing competition hosted by a Beijing television station. Example (9) reflects a fundamental feature of mianzi and lian. In Chinese society, when a person experiences a positive or a negative change in his or her mianzi/lian, this change can also be experienced vicariously by those with whom the person is closely associated. Jiang Li's success has resulted in a positive change in the mianzi of those with whom she is closely associated, including the respondent. This positive change in mianzi (increased lustre on mianzi) is grounded in the respondent's conviction that others will think more highly of her as a

consequence of her association with the successful Jiang Li. The respondent (B) is certain that other people (A) will think more highly of her (and Jiang Li’s other friends) as a consequence of Jiang Li's success. That is, "what B thinks A thinks of B" as a result of the incident is better than "what B thinks A thinks of B" before the incident. Once again, in this example "A"

(other people) has not yet shown "B" (the respondent) that they think more highly of her as a consequence of the incident, but "B" has made a self-proclaimed assertion of mianzi-gain on

the assumption that "A" will inevitably say positive things about her stemming from her friend's success.

Example (12) features the vicarious augmentation of lian. The respondent (B) considers that other members of the global community (A) think more highly of him and his fellow Chinese as a consequence of the Chinese team's success at the Sydney 2000 Olympic Games. In this example, the respondent is declaring a self-proclaimed lian-gain that he has experienced vicariously owing to his compatriots' achievements. In terms of the metalanguage, "what B thinks A thinks of B" is now better than the previous "what B thinks A thinks of B." The present "what B thinks A thinks of B" is made on the basis of what B (the respondent) expects A (the world community) will say about him (and his country) when his compatriots' achievements are inevitably considered in the public gaze.

2.1.3 Managing one’s own and others’ 'face' in Chinese

Given the importance of mianzi in Chinese social relationships (see Bond and Hwang, 1986; Chang and Holt, 1994, Earley, 1997; Hsu, 1996), it is not surprising that Modern Standard

Chinese lexicalizes a number of expressions that relate to managing one's own and others' mianzi. Lian, however, does not seem to feature as prominently as mianzi in Modern Standard Chinese discourse on social relationships. Let us consider one expression that relates to the management of one's own and others' mianzi.

Gei mianzi ('giving mianzi')

Existing scholarship suggests that there is a diverse range of actions and situations that can be classified as gei mianzi ('to give mianzi'). Broadly speaking, this range of actions and situations can be categorized into two groups – actions that make others look better in public and actions that prevent others from looking worse in public. In both limbs, 'giving mianzi' means doing something good for the 'recipients'. In the first limb, 'giving mianzi' is about showing others that they are valued by making them look better in public, and in the second limb 'giving mianzi' is about showing others that they are valued by not allowing them to look worse in public. Consider the following examples:

(15) Xin shang shuo tamen na bang laotou xianzai chou, xuan chu jiu juede zhe jiang gai gei nin, you pa nin qiao bu shang, jujue de jiang, suoyi xiang xian gen nin shangliang shangliang, gei tamen ge mianzi. (Wang Shuo, Ni bu shi yi ge suren [You are not a common man], p.27.)

'The letter stated that those old fellows are a little anxious. They have decided to give you the prize, but they are worried you won’t accept it. So I decided to come and discuss it with you and ask you to show them some consideration (lit., give them a mianzi).'

(16) Mei Hua zai laoshi mian qian shuo le wo hao xie hao hua, zhen shi gei wo mianzi! ¹²

'Mei Hua said many good things about me in front of the teacher. She really gave me mianzi.' (17) "Yi ge zhuming zuojia dou bu lai, zhen bu gei mianzi." (Wang Shuo, Wan zhu [The

^{Operators], p. 12.)}

'None of the famous writers showed up. That really is a massive snub (lit., really did not give

mianzi).'

Let us examine example (15). In example (15), the speaker is requesting the hearer to 'give mianzi' to a group of old men who have organized a competition. The hearer can 'give mianzi' to the old men by accepting their suggestion that he receive the prize as winner of the competition. This is quite a satirical use of the expression 'to give mianzi' as the hearer would hardly be reluctant to show consideration to the competition organizers by accepting the prize. However, it is quite clear that the notion of 'giving mianzi' in this case is about doing something for others that one does not really want to do but nevertheless decides to do as a favour for those others. A can 'give mianzi' to B by showing that he or she thinks more highly of B than he or she did before ("what A shows A thinks of B" is better than "what B thinks A thinks of B"), or by showing that he or she thinks highly enough of B to want to prevent B from looking bad in the eyes of others. It is "what A shows A thinks of B" that enables B to determine whether or not he or she has been 'given mianzi'.

2.1.4 Attitudes towards and judgments about one's own and others' 'face' in Chinese

A significant proportion of the collocational range of both mian and lian is comprised of expressions that evaluate both the magnitude and nature of one's own and others' mian and lian. Below, we look at one expression that is employed as an enunciation of an attitude towards others' lianpi ('face-skin').

Lianpi hou; hou lianpi ('a thick skin on the lian')

Hu (1944: 54) suggests that "lianpi hou means the defiance of public censure or disregard for the injunctions of elders trying to impress on the young the moral standards of society." Hsu (1996: 78-9) further explains that "lianpi hou is applied to persons who have little or no concern about their social image. It amounts to a serious criticism of the person described." The image here is of a person who has little regard for the fact that others might avoid him/her in the future; he or she makes excessive requests unaware of the danger that being overly demanding might "break" the lian in the relationship with others. Seemingly resistant to being broken, his or her 'facial skin' is simply thick (see Kipnis, 1995: 131). Consider the following examples:

(18) "Wo..." Liu Huifang yi shi yu sai, xuan ji qing mou yi xiao: "Wo mei ni name hou lianpi."

(Wang Shuo, Liu Huifang, p.13.)

"Me…" Liu Hui-fang was lost for words. "I am not as shameless as you (lit., I don’t have

lian-skin that is as thick as yours)."

(19) Ta zai ye mei peng dao guo yi ge bi ni lianpi geng hou de ren. (Wang Shuo, Wo shi

"lang", p. 5.)

'He has never met another person as brazen as you (lit., a person with lian-skin as thick as yours).'

In example (18), Liu is talking to her mother about people who are not genuine. Her Mum says, "Yes, yes, I am fake. I'll say anything to anyone. Even now, I still do it. But when I am fake I'll admit to it. What about you?" Liu replies with the above statement. In this example, it is clear that the expression hou lianpi is used to describe a person who is not concerned about their image in the eyes of others. Liu describes her mother's lian as 'thick' on the basis that she behaves in a way that shows disregard for social values and the opinions of others. B's (Liu's mother) lian is thick because she does not care about what A (others) thinks of her ("what B thinks A thinks of B" is unimportant) and she is not concerned about what others say about

her ("what A shows A thinks of B"). In this way, having hou lianpi is used as a criticism in

Chinese social interaction.

Throughout this section, we have seen that various 'face' expressions in Chinese can be described in terms of a simplified metalanguage that uses notions such as "what A shows A thinks of B" and "what B thinks A thinks of B". As Ho (1976: 876) suggests "a person's face (mianzi and lian) is ascertained by conducting enquiries on the opinions that others have of him (sic). Stated simply, a person's face is assessed in terms of what others think of him; the assessment does not include what a person thinks of himself; but may include what he thinks others think of him." We have seen that B can lose (or gain) mianzi and lian when another person or group of people (A) shows that they think less (or higher) of B than B expects

("what A shows A thinks of B" is worse [or better] than "what B thinks A thinks of B"). Additionally, we have noted that B can declare a self-proclaimed loss (or gain) of mianzi and lian when he or she thinks another person or group of people (A) will think less (or higher) of him or her when they inevitably come to evaluate something B has done or something someone closely associated with B has done ("what B thinks A thinks of B" as a result of the occurrence is worse [or better] than "what B thinks A thinks of B" prior to the occurrence). We have also seen that the management of 'face' in Chinese and judgments or attitudes towards 'face' in Chinese can be explained using this metalanguage.

Public knowledge and external evaluation are crucial to the operation of 'face' in Chinese, even when losses or gains of 'face' are self-proclaimed. Although the audience may not be present to witness the face-losing or face-gaining behaviour, any self-proclaimed loss or gain of 'face' is inextricably linked to the probability that "what A shows A thinks of B" when A comes to know about B's behaviour will inevitably become worse or better.¹³ Mianzi and lian

are easily explained in terms of these fundamental elements.

2.2 The metalanguage and 'face' in English

As we mentioned earlier in this paper, it is accepted that the figurative use of 'face' in English was borrowed into the English lexicon from Chinese by missionaries and diplomats in the late

19th and early 20th centuries. In order to develop a crude preliminary analysis of English

'face' in the figurative sense of something like social image, we carried out a study into the collocational range of the term in everyday English. Initially, we gathered over 50 instances of the figurative use of 'face' in contemporary novels, cinema, music and media. This was

later supplemented with a search of the Literature Online database14, which yielded over 250 instances of the figurative use of 'face' in English poetry, drama, prose and on the World Wide Web (refer to Tables 2.1 and 2.2). Contrary to Ervin-Tripp et al's (1995: 45) conclusion that

"the usage for this term currently is educated and primarily appears to be diplomatic or

psychological", it seems that 'face' can be used in the everyday parlance of ordinary speakers of English. However, in accordance with the collocations given by the OED, it appears that the collocational range of 'face' in English is largely limited to 'saving face' and 'losing face'.

	Prose	Poetry	Drama	Cinema	Music	Media (newspapers, radio and popular magazines)	WWW	Totals
Save face	54	25	2	12	4	14	3	114
Saving face	32	12	0	3	2	3	0	52
Lose face	41	18	0	16	2	4	2	83
Losing face	19	8	0	6	0	3	0	36
Regain face	0	0	0	0	0	1	0	1
Give face	0	0	0	0	0	1	0	1

Table One - Breakdown of the instances of the figurative use of English face

	Prose	Poetry	Drama	Cinema	Music	Media (newspapers, radio and popular magazines)	WWW	Totals
Australian source	37	16	0	10	3	12	0	78
United States source	68	24	2	12	3	9	5	123
British source	32	20	0	15	2	5	0	72
Other sources (Incl. India, Canada)	9	3	0	0	0	0	0	12

Table Two - Breakdown of the sources of the figurative use of English face

In this section we will apply the metalanguage to briefly examine examples of 'saving face' and 'losing face' in English. Let us look first at 'saving face'. Consider the following examples:

(20) "The worst strain was on Wednesday afternoons, when you had to put on a brave face to get through the games period. Badminton sessions were the most scary: if you got beaten, you’d shoot to the top of the list for bullying; if you beat the girl on the other side of the net, she would be tempted to clonk you one afterwards, to get even and save face." (Andrea Ashworth (1998), Once a House on Fire, p. 126.)

(21) "We offer you mandatory retirement. Plus, of course, a settlement…a golden parachute of such magnitude that your feet will never touch the ground. We will, of course, delay the announcement, out of respect for our former chairman, until after the celebration of his birthday this weekend."

"Well thanks for allowing me to save face Drew." (This is from the American film Meet Joe

Black.)

(22) "Why should you trust a civilian? I wouldn’t if I was in your place."

That's bloody clever. She don't say, "Why should you trust a woman?" So now all present can save face. We're not bird bashin' or nothin' like that, just being ourselves, Vietnam vets not trusting no one but our own kind. (Bryce Courtenay (2000), Smokey Joe's Café, p. 131.)

(23) "The loss awoke us. We had to save face after such a defeat." (This was said by Brisbane

Lions Australian Rules Football player, Alastair Lynch, on Brisbane AM radio station 612

4QR on Friday June 15, 2001. He was referring to Brisbane's loss to Carlton.)

Like the Chinese examples presented in the previous section, we claim that all the examples of 'saving face' in English listed above can be explained using the metalanguage. Here, we

will analyze only example (20) in detail. In example (20), the author is describing her years as a secondary school student. In this example, "B" is the girl "on the other side of the net" who

has lost the badminton match to the author, and "A" is B's schoolmates, including the author. B's efforts to 'save face' are based on her desire to show that she is a domineering person even though she has revealed an embarrassing lack of sporting ability. B is concerned that "what A shows A thinks of B" will be worse than previous evaluations, and worse than "what B thinks A thinks of B", as a consequence of her defeat. In order to shift the focus from her defeat, and in order to reaffirm her status as a 'toughie', B hits the victor. B is determined to do anything

to ensure that there is no negative disequilibrium between "what A shows A thinks of B" and

"what B thinks A thinks of B".

Losing face

Although perhaps not as popularly employed as saving face, instances of losing face were quite easy to find in our search. Note the following examples:

(24)  "Now listen…I cannot be seen to lose face – that’d be suicide." (This is from the Irish film, My Name is Joe.)

(25) "It was not as if the conquering Normans never married members of the defeated Saxon noble class. They still did. She would lose a little face, but not too much…" (Edward Rutherford (2000), The Forest, p. 72.)

(26) "…yet she could see no way out of the situation which did not involve her backing down and losing face, and having learned to bare her face, my grandmother was most reluctant to lose any of it." (Salman Rushdie (1981), Midnight's Children, p. 43.)

(27) "Mandras knelt down and placed the barrel against the old man's head. He hesitated, appalled with himself somewhere in the back of his mind. He could not do it. In order to make it look as though he was doing something, he clicked back the hammer and took up first pressure. He could not do it. He closed his eyes tightly. He could not lose face. It was a question of being a man in front of other men, a question of honour." (Louis de Bernieres (1995), Captain Corelli's Mandolin, p. 193.)

Once again, all the examples of 'losing face' in English that are listed above can be explained with reference to the metalanguage. However, we will only examine example (24) in detail. In example (24) former alcoholic, soccer coach and ordinary guy, Joe Kavanagh, has got

involved in a drug deal to save one of his favourite players, Liam. Joe has gone to see MacGowan, a drug baron, to sort out Liam's troubles. He requests MacGowan to ignore Liam's wrongdoings and to set him free. MacGowan makes the above statement in response

to Joe's proposal. MacGowan (B) is concerned that other people associated with the drug gang (A) will think much less of him if he pardons Liam. Such behaviour would invariably be viewed by other gang members and rivals as weak. In this way, the 'loss of face' that MacGowan refers to can be described as a potential disequilibrium between "what A shows A thinks of B" and "what B thinks A thinks of B" that would emerge if MacGowan did engage

in weak behaviour. "What A shows A thinks of B" would be worse than previous evaluations, and "what B thinks A thinks of B" would deteriorate from an evaluation that B is proud of to an evaluation much less favourable to B.

We have seen so far that the metalanguage introduced in this paper can be used to examine Chinese mian and lian and English face, and that the simplified language of "what A shows A thinks of B" and "what B thinks A thinks of B" provides a sound starting point for

deconstructing 'face' in Chinese and English.¹⁵ In the following section, we apply the metalanguage to an analysis of 'politeness' in English and Japanese.

3. The metalanguage and 'politeness' in English and Japanese

The concept of 'politeness' has received considerable attention in the past thirty years and numerous approaches and theories have been proposed to illuminate what 'politeness' is and what underlies this pervasive phenomenon. We believe that the metalanguage which has been introduced in this paper can go some way in further clarifying the dynamics underlying the conceptualisation of  'politeness', and thus provide a strong foundation for the development of a comprehensive theory of politeness. In this section the metalanguage is further developed in order to examine examples of politeness in English and teineisa and reigi-tadashisa ('politeness') in Japanese.

The focus in this paper so far has been on the dynamic relationship between "what A shows A thinks of B" and "what B thinks A thinks of B". In the previous section, examples of 'face' were examined using this core dynamic, while in this section it is applied to the analysis of

'politeness' in English and Japanese. 'Politeness' can arise when one shows what one thinks of others. More specifically, 'politeness' can arise when someone shows they think good (or do not think bad) of someone else, or in other words, when a speaker shows a positive evaluation of someone else through his or her behaviour. 'Politeness' also arises relative to social norms, or in other words, the dynamics underlying 'politeness' involve "what A should show A thinks

of B".¹⁶  This view of 'politeness' is summarized as follows, where A is the speaker, B is the addressee, and 'X' is the speaker's utterance.¹⁷

Politeness arises when: A says 'X'

'X' shows [A thinks good of B] or [A does not think bad of B]

'X' shows A thinks the same or better than what A should show A thinks of B B thinks [A thinks good of B] or [A does not think bad of B] because of this

For example, I might compliment a friend on her new haircut (A says 'X'), and thereby show that I approve of my friend ('X' shows A thinks good of B). Politeness arises when upon hearing the compliment my friend thinks I approve of her (B thinks A thinks good of B

because of this), and also thinks the compliment shows that what I think of her is the same or

better than she expects ('X' shows A thinks the same or better than what A should show A

thinks of B).

In this section we are only able to give a preliminary sketch of the implications of our metalanguage for theories of politeness. We are not attempting to give a definitive characterization of 'politeness' in English or Japanese, because there is still considerable work to determine which aspects of people are shown to be evaluated when 'politeness' arises. Instead, a number of examples are given to illustrate how the metalanguage can be applied to analyses of various types of 'politeness' found in English and Japanese. Examples are given to illustrate four types of 'politeness': compensatory politeness, stasis politeness, enhancement politeness, and demeanour politeness.

3.1. Compensatory politeness

Compensatory politeness involves showing that one does not think bad of someone else in spite of some utterance or behaviour that could be interpreted as implying that one thinks bad of them. The way in which compensatory politeness arises can be illustrated using the

following examples in which a student is trying to politely refuse a request from his teacher to do some part-time work.

(28) Teacher: Kono arubaito yat-te mora-e-nai?

this    job        do-Te  receive-Pot-Neg

(Won't you do this part-time work for me?)

Student: Sensei, sumimasen. Sono hi   moo   betsu no baito hait-te-ru-n-de…

teacher excuse me    that day already other of   job   get-Te-Prog-Nomi-Te(Conj) (M'am, I'm sorry.  On that day I already have some part-time work [so]…)

(29) Teacher: Kono arubaito yat-te mora-e-nai?

this     job        do-Te receive-Pot-Neg

(Won't you do this part-time work for me?) Student: Suimasen, hokani baito o     yat-te-iru   node,

excuse me other  job   Acc do-Te-Prog so

jikan ga    tor-e-nai-n-desu                      yo.

_{time  Nom take-Pot-Neg-Nomi-Cop(Pol) M}

(Sorry, I am doing some other part-time work, so I can't make time)  (Haugh, 2003)¹⁸

In both examples (28) and (29) the student refuses the teacher's request, but example (28) is more teinei ('polite') than example (29). The different ways in which the student refuses the request gives rise to this difference in teineisa ('politeness'). The student in example (28) is more teinei ('polite') than in example (29), because what the student shows the student thinks of the teacher by his reply in example (28) is closer than in example (29) to what the teacher

thinks the student should show the student thinks of the teacher.¹⁹

Both the act of refusal itself and the way in which the student refuses shows what the student

(A) thinks of the teacher (B) (that is, "what A shows A thinks of B"). The act of refusal

creates a potential negative disequilibrium between what the student should show he thinks of his teacher ("what A should show A thinks of B"), what the teacher thinks the student thinks of her based on the student's past behaviour towards her ("what B thinks A thinks of B"), and what the student shows he thinks of the teacher through the refusal ("what A shows A thinks of B"). The way in which the student deals with this potential negative disequilibrium is what gives rise to teineisa ('politeness').

In these two examples, what the student should show he thinks of his teacher is something like "the student should show he thinks the teacher's status is higher than the student's, and thus should try to comply with the teacher's requests within the bounds of the school environment". What the teacher thinks the student thinks of her based on the student's previous behaviour is something like 'the student thinks my status is higher than his'. However, a potential negative disequilibrium is created when what the student shows he thinks of the teacher by refusing her request indicates that the student thinks less of the teacher's status than what the teacher thought the student did ("what B thinks A thinks of B"), and less than what the teacher expects the student should show he thinks of her status ("what A should show A thinks of B").

The way in which the student makes the refusal compensates for this potential negative disequilibrium to a lesser or greater extent. In example (28) the student apologizes and then gives a reason, which implies that the student cannot do the part-time work requested. Through his apology and the indirect refusal implied by giving a reason, the student shows

that he does not think less of his teacher's status ("what the student shows the student thinks of the teacher"). The apology shows that the student regrets having to make the refusal,

indicating that it is not a deliberate snub of the teacher. It may also indicate that he knows that

the refusal of his teacher's request is going against social norms (what the student should show he thinks of the teacher), and thus he apologizes for this transgression. The indirect refusal shows the student's reluctance to make the refusal and also foregrounds a reason for making the refusal, which makes the refusal seem to be out of the control of the student. The student tries to show that he does not have the desire or intention to deliberately refuse the teacher's request, indicating that he does not think less of the teacher's status (in terms of compliance with requests).

In example (29) the student also apologizes and gives reasons implying that he cannot comply with the teacher's request. However, as noted above, the student's response in example (29) is less teinei ('polite') than in example (28). This is because the student gives an additional

reason (that the student does not have time), which is more strongly asserted (that is, it ends with the assertive particle yo), giving the impression that the student is not so reluctant to make the refusal. This stronger assertion shows that the student in example (29) does not

think as highly of the teacher (that is, what the student shows the student thinks of the teacher)

as the student in example (28). In both examples (28) and (29) the teineisa ('politeness') which arises can be characterized as a kind of compensatory politeness, because the speaker compensates for a potential negative disequilibrium between what the speaker shows he

thinks of the hearer ("what A shows A thinks of B") and what the hearer thinks the speaker thinks of the hearer ("what B thinks A thinks of B").

In the examples above, 'politeness' arises when the speaker compensates for a potential negative disequilibrium that has been produced through his or her own action. However, compensatory politeness may also arise when someone compensates for a potential negative disequilibrium created by the actions of someone else.  In the following example, Tanaka has arrived 20 minutes late for a pre-arranged meeting with Suzuki.

(30) Tanaka: Gomen osoku nat-ta.

sorry   late    become-Past

(Sorry I'm late)

Suzuki: Iya iya, maa kooyuu       koto   mo   aru yo.

no   no  well this kind of thing also exist M

(No, no, anyway, this kind of thing happens)  (Nishio, 1998: 59)

The fact that Tanaka is late could be interpreted as showing that Tanaka does not think good of Suzuki, because it could show a lack of respect for Suzuki's time ("what A shows A thinks of B"), and this gives rise to a potential negative disequilibrium. Tanaka apologises for being late to compensate for the potential negative disequilibrium created by his lateness. Suzuki also tries to downplay this potential negative disequilibrium (created by Tanaka) as he

reduces the force of Tanaka's apology by not accepting it, and he generalises that being late is

not so uncommon, and thus it is quite excusable for Tanaka to be late. In other words, Suzuki shows that he does not think Tanaka thinks bad of him ("what B shows B thinks A thinks of B"), which in conjunction with Tanaka's apology ("what A shows A thinks of B") gives rise to compensatory politeness. Suzuki's downplaying of the potential negative disequilibrium also shows that Suzuki does not think bad of Tanaka ("what B shows B thinks of A"), which

further contributes to the teineisa ('politeness') arising in this example.

Similar kinds of examples can also be observed in English, as illustrated by example (31). (31) Jill: Would you like to meet for coffee at three?

John: Oh, sorry, I have something to do.

John's implied refusal of Jill's invitation for coffee has the potential to show that John does not approve of Jill as a friend ("what John shows John thinks of Jill"). He compensates for

this, however, by apologizing and giving a reason for his refusal. His reason shows he is unable to meet Jill for coffee, which implies that it is not the case that he does not want to meet Jill or does not approve of Jill as a friend. In other words, his apology and reason shows that he does not think bad of Jill in spite of refusing her invitation.

The metalanguage can also be applied to situations where what the speaker says does not have a potentially negative influence on the equilibrium holding between "what A shows A thinks of B" and "what B thinks A thinks of B".  In other words, it may be applied to situations

where there is no apparent disequilibrium (giving rise to stasis politeness), or even where there is a potentially positive disequilibrium (giving rise to enhancement politeness), as discussed in the following two sections.

3.2. Stasis politeness

Stasis politeness arises when "what A shows A thinks of B" is consistent with both "what A should show A thinks of B" and "what B thinks A thinks of B". In this situation, A's behaviour only serves to reinforce B's perception of A's evaluation of him or her. Stasis politeness is needed to avoid giving rise to a negative disequilibrium, but it is generally

unnoticed, because it only confirms what someone already thinks someone else thinks of them

(Arundale, 1993, 1999; Usami, 2001).

The use of honorifics in Japanese can give rise to stasis politeness in some contexts. Matsumoto (1988) gives the following examples to show how different forms can give rise to differences in politeness when saying something like Today is Saturday in Japanese.

(32a) (to a close friend)

Kyo     wa   doyoobi   da.

today Top Saturday Cop(NonPol) (32b) (to an acquaintance)

Kyo     wa   doyoobi  desu.

today Top Saturday Cop(Pol)

(32c) (to a guest on a formal occasion)

Kyo     wa   doyoobi  degozaimasu.

today Top Saturday Cop(SuperPol) (ibid: 415)

The "polite" forms of the copula in (32b) and (32c) both give rise to teineisa ('politeness'), because they show that the speaker respects the social position of the addressee as someone with whom they are not familiar/intimate ("what A shows A thinks of B"), and this is expected in the situations noted in examples (32b) and (32c) ("what A should show A thinks of B"). The "super-polite form" in (32c) shows greater respect than the "polite form" in (32b) to the social position of the addressee, so it gives rise to a greater degree of teineisa ('politeness'). The so-called "non-polite form" in (32a) shows the speaker approves of the addressee as someone with whom they are familiar, but generally does not give rise to teineisa ('politeness') unless it is being used to show a greater degree of approval of the

addressee (e.g. to show a shift from acquaintance level to a closer friendship).²⁰

The forms in (32b) and (32c) can both give rise to stasis politeness between people meeting for the first time, because they are used at the end of most utterances in those conversations (Usami, 2001, 2002), so the use of a "polite form" or "super-polite form" in one particular utterance only confirms what the addressee already thinks the speaker thinks of him or her (which is based on the use of "polite forms" or "super-polite forms" in previous utterances). In other words, the use of "polite forms" with acquaintances or "super-polite forms" with guests

at a formal occasion simply serves to reinforce the addressee's perception that "what A shows A thinks of B" is the same as both "what B thinks A thinks of B" and "what A should show A thinks of B". In this way their usage can give rise to stasis politeness in particular contexts.

In an example of stasis politeness in English, when meeting colleagues at the department in which we work it is normal to greet each other with something like Hi, how are you? The function of this greeting is not really to inquire about the addressee's well-being, but rather to show that the speaker still approves of the addressee as a colleague (that is, "what A shows A thinks of B" is the same as "what A should show A thinks of B" and "what B thinks A thinks of B"). If we were to neglect to greet our colleagues, although it could be interpreted as tiredness on the part of ourselves, it might also be interpreted as a lack of concern for our colleagues. In other words, not greeting colleagues can be interpreted as showing one does not approve of them, thereby giving rise to a negative disequilibrium.

Another example of stasis politeness in English involves the use of the address forms by children towards adults who are not relatives in Australian or New Zealand English. It is common for children to address adults with the form Mr/Mrs/Ms/Miss preceding the surname of the person in question. This shows the child respects the addressee's position as an "adult" ("what A shows A thinks of B"), but since this is normally the same as what the addressee already thinks the child thinks of him or her ("what B thinks A thinks of B"), and is also consistent with what is expected ("what A should show A thinks of B"), this kind of politeness is not normally noticed. Instead it contributes to maintaining the equilibrium between "what A shows A thinks of B" and "what B thinks A thinks of B".

3.3. Enhancement politeness

Enhancement politeness arises when "what A shows A thinks of B" is better than both "what B thinks A thinks of B" and "what A should show A thinks of B". In other words, it arises from a positive disequilibrium between the addressee's previous perception of how the speaker evaluates him or her and the evaluation of the addressee shown by the speaker. It differs from compensatory politeness because there is no apparent or potential negative disequilibrium, while it differs from stasis politeness because it involves a positive disequilibrium.

In example (33) a friend compliments another friend on her new haircut. (33) Mari: (to Yoko, a friend) Kami, kit-ta   ne. Yoku niau wa yo.

hair  cut-Past M  well  suit  M M

(You cut your hair huh?  It really suits you!)  (Aoki and

Okamoto, 1988: 205).

In this example Mari's utterance shows that Mari thinks Yoko looks more attractive with her new hairstyle (that is, "what A shows A thinks of B"), and Yoko may therefore revise "what B thinks A thinks of B" such that she thinks Mari thinks she looks more attractive with her new hairstyle.²¹   If the compliment is not expected (that is, it is not in line with "what B thinks A should show A thinks of B"), then Yoko will revise what she thinks Mari thinks of her such

that it is better than before.  In other words, what Yoko thinks Mari thinks of Yoko is enhanced, thereby giving rise to enhancement politeness.²²

Enhancement politeness may arise in English when one receives an invitation for the first time from someone. For example, if a landlord with whom a tenant has previously only had small-talk invites that tenant to dinner for the first time, enhancement politeness may arise.

This is because the landlord shows he approves of his tenant more than before through this invitation (that is, "what A shows A thinks of B" is better than "what B thinks A thinks of B"). Showing this degree of approval is not necessarily expected since landlords do not always invite their tenants for dinner (that is, "what A shows A thinks of B" is better than "what A should show A thinks of B"), so enhancement politeness arises as a consequence.

3.4. Demeanour politeness

While the focus of the metalanguage thus far has been restricted to external evaluations of people (that is, "what A shows A thinks of B"), 'politeness' may also arise by showing what one thinks of oneself ("what A shows A thinks of A"), a phenomenon that has been previously termed "self politeness" (Chen, 2001) or "demeanour politeness" (Goffman, 1967: 77). Demeanour politeness in both English and Japanese involves showing that one does not think good of oneself, especially in cases where one has received a positive evaluation from

someone else.²³ It thus contrasts with the previous three types of 'politeness', which involve showing what one thinks of others.

The definition of 'politeness' outlined at the beginning of this section requires modification if it is to be successfully applied to demeanour politeness, as illustrated below.

Politeness also arises when: A says 'X'

'X' shows A wants to show A does not think good of A

'X' shows A thinks the same or worse than what A should show A thinks of A B thinks A does not think good of A because of this

For example, if a colleague responds to my compliment about her new haircut by saying that it is nothing special, she shows that she does not think her haircut is particularly special ('X' shows A does not think good of A). Demeanour politeness arises if I consequently think that my colleague wants to show she does not think her haircut is particularly special (B thinks A does not think good of A because of this), and this is expected ('X' shows A thinks the same as what A should show A thinks of B).

Demeanour politeness in Japanese can arise from "token refusals", which are ritualistic interchanges where someone refuses an offer or invitation a number of times before finally accepting. In example (34), Tsune's apparent refusal of Tora's gift gives rise to demeanour politeness.

(34) Tora: Korya, omiyage tte    hodo no mon  janai ga,

this      gift      Quot extent of thing Neg but obachan chotto hame-te-mi-te-ku-re               yo. aunty     a little  wear-Te-try-Te-receive-Imp M

(This is not really a proper gift, but Tsune please try it on) Tsune: Ii     no      kai.   Konna        rippana    mono.

good Nomi Q     this kind of marvelous thing

(Is it okay?  (To receive) this kind of marvelous thing?) Tora: Aa, ii    yo.

oh  good M

(Oh, it's fine!)  (Aoki and Okamoto, 1988: 163-164)

Tsune's (A) initial refusal to accept Tora's (B) gift shows that she thinks she shouldn't casually accept gifts from Tora.  This indicates that Tsune thinks that her position in relation to Tora is such that she cannot expect to receive big gifts from Tora ("what A shows A thinks of A"). From this Tora infers what Tsune thinks of herself ("what B thinks A thinks of A").  As this is

apparently in line with what Tsune should show she thinks of herself (that is, "what A should show A thinks of A"), teineisa ('politeness') arises.

Demeanour politeness in Japanese also arises when showing modesty with regards to oneself in interactions upon receiving a positive evaluation from someone else (Kokuritsu Kokugo Kenkyujo, 1990). In the following example a friend of Otoboke's downplays a compliment made about his daughter who is playing the piano.

(35) (Otoboke is visiting his friend's house. His friend's daughter plays the piano for them both)

Otoboke: O-joozu       desu      ne.

Hon-skillful Cop(Pol) M (She is good isn't she?)

Friend: Ee, shoogakko          gonen      kara yat-te-masu kara…

well primary school year five from do-Te-Pol    so

(Well, she has been playing since year five of primary school so…) (Ueda, 1983: 118) The response from Otoboke's friend implies that since his daughter has been playing for a long time, it is only natural that she can play at a certain level, therefore her playing is not worthy of such praise. Otoboke's friend downplays the compliment, and shows that while Otoboke may think highly of the daughter's piano playing (and by extension his friend), the friend does not think more highly of his daughter's playing (and by extension himself) than he should. In other words, Otoboke's friend expresses modesty, by showing that he does not think more highly of himself than he should ("what A shows A thinks of A"). The expression of modesty by Otoboke's friend is also probably expected in this situation ("what A should show A thinks of A"), as otherwise he may show he thinks too highly of himself. Since Otoboke's friend downplayed Otoboke's compliment, Otoboke can see that Otoboke's friend does not think more highly of his daughter (and by extension himself) than he should ("what B thinks A thinks of A"), thus demeanour politeness arises in this situation.

Demeanour politeness may arise in English when a sports star, after winning a gold medal at the Olympic games, praises his or her family and coach for all their support at the after-event interviews. By praising others upon one's success, the sports star shows he or she does not think more highly of him or herself than he or she should ("what A shows A thinks of A"),

and in Australia at least, this is expected ("what A should show A thinks of A"). The audience who is watching the interview will most likely think the sports star is appropriately modest ("what B thinks A thinks of A"), at least in public, and thus consider the sports star to be polite.

3.5. Implications of the metalanguage for studies of 'politeness'

We believe that approaching politeness using this metalanguage has a number of advantages for studies of 'politeness'. Firstly, politeness can be explained using non-technical terms which are less likely to diverge from intuitive conceptualisations of 'politeness'. All terms used are simple vocabulary, which are found across a wide range of languages. It thus has potential for development into a more widely applicable cross-cultural metalanguage. Secondly, different types and degrees of 'politeness' which arise in interaction can be explicated using this metalanguage. A theory of politeness must account for the fact that there are different types of

'politeness' and also different degrees of 'politeness'. Thirdly, disagreement about whether

'politeness' has arisen or not or the degree of politeness manifested in an interaction can be explained because of inherent variability in perceptions of "what A should show A thinks of B" and "what A should show A thinks of A" across native speakers.

There are, of course, many other dimensions of 'politeness' which have not been fully explored in this paper. One dimension which requires further consideration is what aspects of B (with regards to "what B thinks A thinks of B" and so on) are salient in interactions where

'politeness' arise. Arundale (1993) has argued that those dimensions of B salient to 'politeness' are culture-specific elaborations of the basic dimensions of "connectedness" (how people are connected to others) and "separateness" (how people are separated from others) (see also Hernandez Flores, 1999; O'Driscoll, 1996). How this distinction can be developed to account for 'politeness' in English and Japanese requires further research, but we believe the metalanguage is an important starting point for this endeavour.

Another issue that needs further investigation is determining the factors involved in the constructs "what A should show A thinks of B" and "what A should show A thinks of A", since these constructs play such a pivotal role in giving rise to 'politeness' (or not) in interactions. We suggested in Section One that dimensions such as "what A and B think of social norms/values/attributes/principles", "what B thinks of B", "what A thinks of A" and so on influence these constructs. These dimensions require further investigation and elaboration if we are to more fully understand the complex concept of 'politeness'.

4. Conclusion

We believe that the metalanguage proposed in this paper can be used to deconstruct the concepts of 'face' and 'politeness'. We have shown that the metalanguage can be used to describe the operation of mianzi and lian  in Chinese and face in English, and the mechanisms underlying 'politeness' in English and Japanese.  We have tried to avoid superimposing the semantic domains which accompany English terms such as face and politeness upon an analysis of potentially similar concepts from other cultures through the use of this simplified metalanguage.

This metalanguage illuminates the common dimensions that underlie 'face' in Chinese and English and 'politeness' in English and Japanese. Due to the constraints on the scope of this paper we have not made in-depth comparisons between mianzi and lian and the theoretical and folk usages of face in English, or comparisons of teineisa/reigi-tadashisa in Japanese with theoretical and folk usages of politeness in English, nor have we investigated the relationship between 'face' and 'politeness' across cultures.  However, we propose that the metalanguage provides an operational tool with which future research can make such comparisons. Future research may illuminate the salient dimensions of mian, lian and related terms in Chinese; mentsu, kao and related terms in Japanese; face in English; limao and related terms in Chinese; teineisa and related terms in Japanese; and politeness and related

terms in English. With an understanding of the salient dimensions of these concepts, it may be possible for future researchers to arrive at accurate explications of them, and to explore cross- cultural relationships between them.

References

Aoki, Haruo, Okamoto, Shigeko, 1988. Rules for Conversation Rituals in Japanese.

Taishukan, Tokyo.

Arundale, Robert, 1993. Culture specific assumptions and the concept of face: a proposal toward a cultural universal for studying face management in using language. Paper presented at the International Pragmatics Association Conference, Kobe, Japan (Revised August 1997).

Arundale, Robert, 1999. An alternative model and ideology of communication for an alternative to politeness theory. Pragmatics 9, 119-154.

Bond, Michael, Hwang, Kwang-Kuo, 1986. The social psychology of Chinese people. In: Bond, Michael (Ed.), The Psychology of the Chinese People. Oxford University Press, Hong Kong, pp. 213-266.

Brown, Penelope, Levinson, Stephen, 1987. Politeness.  Some Universals in Language Usage.

Cambridge University Press, Cambridge.

Brunner, James, Wang, You, 1988. Chinese negotiating and the concept of face, Journal of international consumer marketing 1, 27-43.

Chen, Rong, 2001. Self-politeness: a proposal. Journal of Pragmatics 33, 87-106.

Cheng, Chung-Ying, 1986. The Concept of face and its confucian roots. Journal of Chinese

Philosophy 13, 329-348.

Cole, Michael, 1989. A cross-cultural inquiry into the meaning of face in the Japanese and the United States cultures. Paper presented at the Speech Communication Association Convention, San Francisco, CA.

Chu, Rui-ling (Zhu, Rui-ling), 1983. Guanyu mianzi de shizheng yanjiu (Empirical research on the psychology of face). Unpublished Ph.D. dissertation, National Taiwan University.

Danton, G.H. 1938. The Chinese people. Marshall Jones Company, Boston. Douglas, Robert, 1895. Society in China. Innes, London.

Durkheim, Emile, 1933. The division of labor in society. Simpson, G. (Trans.). Free Press, New York.

Earley, Christopher, 1997. Face, harmony and social structure. Oxford University Press, New

York.

Ervin-Tripp, Susan, Nakamura, Kei,  Jiansheng, Guo, 1995. Shifting face from Asia to Europe.

In: Shibatani, Masayoshi, Thompson, Sandra (Eds.), Essays in Semantics and Pragmatics.  (Essays in honour of Charles J. Fillmore). John Benjamins, Amsterdam, pp. 43-71.

Fillmore, C., Atkins, B., 1992. Towards a frame-based organization of the lexicon: The semantics of RISK and its neighbours. In: Lehrer, A., Kittay, E. (Eds.), Frames, fields, and contrasts: New essays in semantics and lexical organization. Lawrence Erlbaum, Hillsdale, pp. 75-102.

Giles, Herbert, 1902. China and the Chinese. Columbia University Press, New York. Goffman, Erving, 1955. On face-work: An analysis of ritual elements in social interaction.

Psychiatry: Journal of the Study of Interpersonal Processes 18, 213-231.

Goffman, Erving, 1959. The Presentation of Self in Everyday Life. Penguin, Harmondsworth. Goffman, Erving, 1967. Interaction Ritual.  Essays on Face-to-face Behavior. Pantheon Books,

New York.

Gu Qian, 2000. Daolu [The Path]. In Hua cheng, 1, pp. 135.

Gu, Yueguo, 1990. Politeness phenomena in modern Chinese. Journal of Pragmatics 14, 237-

257.

Haugh, Michael, (2003). Politeness implicature in Japanese: A metalinguistic approach.

Unpublished PhD dissertation, University of Queensland.

Hinze, Carl, 2002. Rethinking 'face': Pursuing an emic-etic understanding of Chinese mian

and lian and English face. Unpublished PhD dissertation, University of Queensland. Hernandez Flores, Nieves, 1999. Politeness ideology in Spanish colloquial conversations: The

case of advice. Pragmatics 9: 37-49.

Ho, David, 1976. On the concept of face. American Journal of Sociology 81, 867-884.

Ho, David, 1994. Face dynamics: from conceptualization to measurement. In: Ting-Toomey, Stella (Ed.), The Challenge of Facework. State University of New York Press, Albany, New York, pp. 269-285.

Holcombe, Chester, 1895. The real Chinaman. Dodd, Mead, New York.

Hsu, Stephen Chuan-Hsi, 1996. 'Face': An ethnographic study of Chinese social behaviour.

Unpublished PhD dissertation, Yale University.

Hu, Xianjin, 1944. The Chinese concept of 'face'. American Anthropologist 46, 45-64. Huang, Shuanfan, 1986. Two studies on prototype semantics: Xiao (filial piety) and mei

mianzi (loss of face). Journal of Chinese linguistics 15, 55-89.

Hwang, Kwang-Kuo, 1987. Face and favor: The Chinese power game. American Journal of

Sociology 92, 944-974.

Ji, Shaojun, 2000. 'Face' and polite verbal behaviours in Chinese culture. Journal of

Pragmatics 32, 1059-1062.

Kipnis, Andrew, 1995. 'Face': An adaptable discourse of social surfaces. Positions: East Asia cultures critique 3, 119-148.

Kokuritsu Kokugo Kenkyujo [National Language Research Institute], 1990. Keigo Kyoiku no Kihon Mondai, Volume 1 [Fundamental issues in honorifics education]. Kokuritsu Kokugo Kenkyujo, Tokyo.

Kornacki, Pawel, 1995. Heart and Face: Semantics of Chinese Emotion Concepts. PhD

dissertation, Australian National University.

La Barre, Weston, 1946. Some observations on character structure in the Orient, II. The

Chinese, part one. Psychiatry 9, 215-238.

Latourette, Kenneth, 1942. The Chinese: Their history and culture. Macmillan, New York:. Leichty, Greg, Applegate, James, 1991. Social-cognitive and situational influences on the use

of face-saving persuasive strategies. Human Communication Research 17, 451-484. Lim, Tae Seop, 1994. Facework and interpersonal relationships. In: Ting-Toomey, Stella

(Ed.), The Challenge of Facework: Cross-cultural and Interpersonal Issues. State

University of New York Press, Albany, New York, pp. 209-229. Lyons, John, 1977. Semantics. Cambridge University Press, Cambridge.

Mao, LuMing, 1994. Beyond politeness theory: 'face' revisited and renewed. Journal of

Pragmatics, 21, 451-486.

Macgowan, John, 1908. Sidelights on Chinese life. Lippincott, Philadelphia. Macgowan, John, 1912. Men and manners of modern China. Unwin, London.

Matsumoto, Yoshiko, 1988. Reexamination of the universality of face: politeness phenomena in Japanese. Journal of Pragmatics 12, 403-426.

Morisaki, Seiichi, Gudykunst, William, 1994. Face in Japan and the United States. In:  Ting- Toomey, Stella, (Ed.), The Challenge of Facework: Cross-cultural and Interpersonal Issues. State University of New York Press, Albany, New York, pp. 47-93.

Nishio, Junji, 1998. Mainasu taiguu hyoogen koodoo bunseki no kokoromi [Towards an analysis of minus interpersonal expressions]. Osaka Daiguku Nihongogakuhoo 17, 57-

69.

Nwoye, Onuigbo, 1992. Linguistic politeness and socio-cultural variations of the notion of face. Journal of Pragmatics 18, 309-328.

Obana, Yasuko, 2000. Understanding Japanese. A handbook for learners and teachers.

Kurosio, Tokyo.

O'Driscoll, Jim, 1996. About face: A defence and elaboration of universal dualism. Journal of

Pragmatics 25,1-32.

Pan, Yuling, 2000. Politeness in Chinese face-to-face interaction. Ablex, Stamford, CT. Penman, Robyn, 1990. Facework and politeness: Multiple goals in courtroom discourse.

Journal of Language and Social Psychology 9, 15-38.

Pike, Kenneth, 1967. Language in Relation to a Unified Theory of the Structure of Human

Behaviour. Mouton, The Hague.

Redding, Gordon, Ng, Michael, 1982. The role of 'face' in the organizational perceptions of

Chinese managers. Organization Studies 3, 210-219.

Scollon, Ron, Scollon, Suzanne Wong 1995. Intercultural Communication. Blackwell, Oxford. Spencer-Oatey, Helen, 2000. Rapport management: a framework for analysis. In: Spencer-

Oatey, Helen, (Ed.), Culturally Speaking.  Managing Rapport through Talk across

Culture. Continuum, London, pp. 11-46.

Strecker, Ivo, 1993. Cultural variations in the concept of 'face'. Multilingua 12, 119-141. Sueda, Kiyoko. (1996). Differences in the perception of face: Chinese mien-tzu and Japanese mentsu. World Communication 24, 23-31.

Swi, Hong, 1997. Facework in Chinese cross-cultural adaptation. Unpublished PhD

dissertation, University of Texas at Arlington.

Tedeschi, James, Riess, M., 1981. Identities, the phenomenal self, and laboratory research. In: Tedeschi, James (Ed.), Impression management theory and social psychology research. Academic Press, New York, pp. 3-22.

Ting-Toomey, Stella, Gao,  Ge, Trubisky, Paula, Yang, Z., Kim, Kwangsu, Lin, Sung-Ling, Nishida, Tsukasa, 1991. Culture, face maintenance, and styles of handling interpersonal conflict: A study in five cultures. The International Journal of Conflict Management 2, 275-296.

Ting-Toomey, Stella, Trubsiky, Paula, Lim, Sung-Ling, 1991. The influence of

individualism-collectivism and self-monitoring on conflict styles. International Journal of Intercultural Relations 15, 65-84.

Tracy, Karen, 1990. The many faces of facework. In: Giles, Howard, Robinson, Peter (Eds.), Handbook of Language and Social Psychology. John Wiley, New York, pp. 209-226.

Ueda, Masahi, 1983. Otoboke Kachoo, Volume 1. Hobunsha, Tokyo.

Usami, Mayumi, 2001. Poraitonesu no danwa riron - "Disukoosu poraitonesu" to iu toraekata [Discourse theory of politeness - an understanding based on "discourse politeness"]. In: Kokuritsu Kokugo Kenkyuujo [National Language Research Institute] (Ed.), Dai 7

Kai Kokuritsu Kokugo Kenkyuujo Kokusai Shinpojiumu Hookokusho - Danwa no Poraitonesu [Report from the seventh symposium of the National Language Research Institute]. Kokuritsu Kokugo Kenkyuujo, Tokyo, pp. 9-57.

Usami, Mayumi, 2002. Discourse politeness in Japanese conversation. Some implications for a universal theory of politeness. Hituzi Syobo, Tokyo.

Wierzbicka, Anna, 1991. Cross-cultural pragmatics: The semantics of human interaction.

Mouton de Gruyter, Berlin.

Wierzbicka, Anna, 1996. Semantics: Primes and universals. Oxford University Press, Oxford. Yang, Martin, 1945. A Chinese village. Columbia University Press, New York.

Zhao Hesong, 1995. Qinghai zhuoliu [Turbulent Currents in the Sea of Love]. Gushihui 4, 10-

19.

¹ The ideas in this section were partially inspired by Arundale's (1993, 1999) and Ho's (1976, 1994) work on

_'face'.

² Of course there are always some members of society who show a total disregard for what others think of them, but this is generally regarded as anti-social behaviour.  This issue is not, however, considered in our analysis of

'face' and 'politeness'.

³ For an example of a model of interpersonal interaction which deals with 'face' and 'politeness' issues across cultures see Spencer-Oatey (2000).

⁴ Of course B may come to know of "what A shows A thinks of B" vicariously through a third party.  For example, someone else may tell Michael that Carl has condemned Michael's research.

⁵ _{For example, what others think of us can influence what we think of me (that is, "what B shows A thinks of A").}

⁶ In a model of interpersonal interaction, the interactants' desires in relation to 'face', such "what B wants A to think of B" may be important tools for describing the emotional aspects of an interaction. However, we do not

aim to describe interpersonal interaction as a whole and we believe that 'face' and 'politeness' precede emotional

reactions and are independent of the interactants' own desires. Thus, the concept of 'face-wants' is unnecessary for describing 'face' and 'politeness'.

⁷ _{That is, "what A thinks (A shows A thinks of B)".}

⁸ In models that examine interpersonal interaction, the distinction between 'self-face' and 'other-face' is important. In this paper, because we are not dealing with interpersonal interaction as such, we do not emphasize this

distinction. However, 'other-face' can be explicated using our metalanguage: "what A thinks B thinks of A",

"what B shows B thinks of A", and "what B should show B thinks of A".

Another aspect of interpersonal interaction which is not included in this exposition of the metalanguage is potential differences in the perception of these constructs by the interactants (e.g.  "what A thinks A shows A

thinks of B" versus "what B thinks A shows A thinks of B"). If these two perspectives do not intersect then

miscommunication has occurred, but this is not the focus of this paper. We look at where the two perspectives intersect, so they are represented as simply one construct (e.g. "what A shows A thinks of B") in this metalanguage.

⁹ All references to the works of Wang Shuo are extracted from the website

_{<http://dir.sohu.com/literature/classes/novels/wenyi/author/w_x/wangshuo/index.html>.}

¹⁰ _{Examples (1) to (4) were provided by respondents from the People's University of China in Beijing.}

¹¹ _{Examples (9) to (12) and (14) were provided by respondents from the People's University of China in Beijing.}

¹² _{Example (16) was provided by a respondent from the People's University of China in Beijing.}

¹³ Some research has been carried out into the effect that audience size has on the magnitude of 'face' gain and loss in Chinese. Much of the work tends to suggest that a wider audience generally results in a more significant

loss or gain of 'face', and a narrower audience generally results in a less significant loss or gain of 'face'. Additional research also examines the effect that the comparative status of the audience has on the magnitude of

_{'face' gain or loss that can be attributed to a certain person in an event (for example, Hsu, 1996; Swi, 1997).}

¹⁴ Literature Online is a licenced database described as "the home of English and American Literature on the

World Wide Web". It is a fully searchable library of more than 330, 000 words of English and American poetry, drama and prose, plus biographies and key criticism and reference resources. It can be found at

_{<http://www.lion.chadwyck.com/home/home.cgi?source=config2.cfg>.}

¹⁵ Some scholars such as Brown and Levinson (1978, 1987) and Helen Spencer-Oatey (personal communication)

consider 'face' to be the want of an individual, in which case the idea of "what B wants A to think of B" is accorded a central focus in determining what underlies the operation of 'face'. However, we disagree with this stance and suggest that notions such as "what B wants A to think of B" or "what B thinks A should think of B" may, in some way, inform B's own estimation of A's evaluation of him or her ("what B thinks A can show A thinks of B"), but are not fundamental to the operation of face, mian or lian. An individual's wants or desires, in the final analysis, are independent from public opinion and external evaluation. B might want A to think highly of B, but if A shows A thinks lowly of B the result may be a 'loss of face' for B, regardless of what B wants. If B expects A to think highly of B, but A shows that A thinks lowly of B, then the 'loss of face' is often quite clear to

all parties involved. "What B wants A to think of B" may remain rather constant throughout B's interactions with A, while the dynamic between "what A can show A thinks of B" and "what B can think A thinks of B" may be forever changing and hence bringing about changes to B's 'face'.

¹⁶ These expectations about "what A should show A thinks of B" vary according to the situation, as recently argued by Pan (2000), who claims that 'politeness' is situationally-based.

¹⁷ This approach can be extended to encompass non-linguistic politeness as well by substituting "does" for "says"

_{as in A does 'X'.}

¹⁸ Both these examples were taken from responses in a discourse completion test administered to 122 students at

Tokyo University of Foreign Studies in January 2000.  The other examples of 'politeness' in Japanese are taken from a corpus gathered by the researcher as part of his doctoral research. See Haugh (2003) for further details.

¹⁹ The abbreviations used in the morphological gloss are based on those used by Obana (2000: 27): Acc = accusative, Cop = copula, Hon = honorification, Imp = imperative, M = mood marker, Neg = negation, Nom = nominative, NonPol = non-polite form, Pol = polite form, Past = past tense, Prog = Progressive, Q = question marker, Quot = quotation, SuperPol = super-polite form, Te = 'te-form', Top = topic marker.

²⁰ Ikuta (1988) has noted that 'politeness' can be conceptualised both as a "form" and as an "effect". The metalinguistic approach to 'politeness' outlined in this paper focuses primarily on 'politeness' as an effect, but in

discussing Japanese 'politeness', the way in which certain forms are considered 'polite' (in particular honorifics)

_{should also be considered (see Haugh (2003) for further discussion).}

²¹ In another context, however, this compliment may be meant in a more sarcastic manner, and 'politeness' would not arise in this case.

²² However, if this kind of comment is expected between Mari and Yoko because of the nature of the friendship

(that is, it is in line with "what B thinks A should think of B"), then no such a revision will take place, and stasis politeness will arise.

²³ Showing that one does not think good of oneself may imply that one thinks neither good nor bad of oneself, or that one thinks bad of oneself, but it is the former that is most common in interactions involving 'politeness'.