Testing Major Linguistic Politeness Theories against the Marital Relationships of Bilingual (Urdu and Punjabi) Speaking Pakistani Couples

The study aims to investigate the first order account of (im)politeness in the intimate relationships of married couples in the context of urban Pakistani society and with respect to four competing (face-saving, discursive, frame-based and rapport management) models of politeness. The study participants comprised 21 of those urban Pakistani bilingual (Urdu and Punjabi speaking) couples who sought psychological marital counselling from the researcher after being affected mainly by linguistic impoliteness and their training in linguistic politeness helped them out. Placed within the constructivist qualitative research paradigm and grounded in ethnography and phenomenology, findings of the study reveal that the phenomena under investigation should be studied within a more general framework, the discursive model emerges to be the most robust in its applicability though.


Introduction
The word 'politeness' generally applies to any social behaviour, which is culturally appropriate and shows care for other people's feelings (McIntosh, 2013) whereas the term 'linguistic politeness' refers to such use of language in conversation as is appropriately considerate for the feelings and desires of the interlocutors and is intended to develop or maintain good interpersonal relationships (Huang, 2017;Sharifian, 2017;Van Olmen, 2017). The word 'appropriately' here denotes cultural sensitivity and cross-cultural differences in distinguishing the polite from the impolite (Hirschon, 2001;Lustig & Koester, 1999;Van Olmen, 2017). Linguistic politeness, both in its concept and the practice, has never been new to human beings, but serious research into this phenomenon only became evident in the 1970 and 1980s, mainly in the fields of pragmatics and sociology (Hirschon, 2001;Van Olmen, 2017). One seminal, and the most famous, contribution of that era is Levinson's (1987 [1978]) politeness theory that draws on Goffman's (1962) notions of face and facesaving, Austin (1962) and Searle's (1979) speech act theory and Grice's (1975) cooperative principle (Márquez-Reiter, 2000). The theory quickly caught on and inspired a plenty of politeness research in the CONTACT Iftikhar Ahmad Khokhar at diligentiftikhar@gmail.com upcoming years but several of its claims could not be tested against the accumulated data over the years (Huang, 2017;Márquez-Reiter, 2000;Mills, 2011;Van Olmen, 2017). The invariable sociolinguistic factors did not prove to be universal, as this theory had suggested, and its second-order, observer-oriented, 'etic' approach significantly differed from the first-order, participant-generated, 'emic' perspective (Armaşu, 2012;Lakoff & Ide, 2005;Mills, 2011;Van Olmen, 2017). Several other alternative theoretical accounts of politeness and impoliteness have been proposed since then, the most influential of whom are the discursive (or postmodern) model, the frame-based model and the rapport management model (Armaşu, 2012;Haugh, 2007;Huang, 2017;Mills, 2011;Van Olmen, 2017).
According to Brown and Levinson's (1987) politeness theory, politeness is a tactful act of showing awareness and consideration for the hearer's face, a public self-image, that we all have and want all others to recognize and respect. The act is tactful because it requires the speaker to tactfully assess 'weight' and accordingly choose an appropriate 'strategy' to successfully mitigate threat to both their own and the hearer's face. The weight refers to the accumulative effect of three sociocultural factors: power, distance and rank, that is, the comparative social status, the closeness of the relationship and the cultural sensitivity of the topic. And the strategy explains how the speaker chooses among the five available options in any given situation. The first option is silence: not to say anything and avoid the face threatening act altogether; the second is positive politeness: respond to the positive face wants of the hearer by expressing friendship and solidarity; the third is negative politeness: cater the negative face wants by treating the hearer with deference and avoiding any imposition on them; the fourth is off-record: produce an indirect statement by only hinting at a need and leaving full room for both the speaker and the hearer to avoid losing face and the fifth, but not the final, option is bald on-record: make a direct and open (or even imperative) statement, judging that there is no threat to the hearer's face involved whatsoever.
Drawing on Foucault (1972Foucault ( , 1978 and Fairclough and Wodak's (1995) earlier works, postmodern discursive theorists (such as Locher (2004), Bousfield (2008), Geyer (2008 and Culpeper et al. (2008) etc.) shifted the focus of politeness research from isolated utterances to the longer stretches of interaction; from the second order politeness to the first order politeness; from stereotypical roles of the interlocutors to those interactants who are constantly co-constructing their identity and contesting for power; from the linguistic behaviour which was viewed as inherently polite or impolite to the inferential process through which the participants assign meaning to an utterance and from the individual choices of the speaker to those dynamic social factors that shape, provoke or limit their choices for action (Armaşu, 2012;Haugh, 2007;Locher, 2004;Locher & Watts, 2008;Mills, 2011;Sifianou, 2010). They also differed in regarding impoliteness as just the lack of politeness or as its polar opposite and asserted that, in the struggle for power, certain contexts will rather offer impoliteness as a more suitable device to exercise power and limit the action choices of the opponent Mills, 2011). That is why, investigating the role of impoliteness is far more important than the role of politeness in the relational work Mills, 2011). They count impoliteness as one of the available choices and assert that the labels in interactants' judgements are far more diverse than the dichotomy of polite and impolite Culpeper, 2011). People, sometimes, find 'impolite' and 'rude' quite different from each other, perceive 'over-politeness' as 'rudeness' and do not just use impoliteness for coercive purposes alone, but for affective and entertaining reasons as well (Culpeper, 2011;Huang, 2017). Finally, about the role of the social norms in shaping linguistic behaviour of the interactants, discursive theorists assert that the interactants are aware of these norms at large but their judgements keep negotiating and renegotiating their meaning and thus their role ultimately changes from one social interaction to another (Locher, 2006).
Another competing account of politeness came from Marina Terkourafi (2005) who developed her argument for a frame-based approach to politeness on empirical evidence from the polite use of imperatives in the modern-day Cypriot Greek (Lakoff & Ide, 2005;Terkourafi, 2005Terkourafi, , 2008. A frame, as defined by Minsky (1975), is a "data structure [in the mind] for representing a stereotyped situation" (Minsky, 1975). Stored in our memory, it is that knowledge of our past experiences that influences our perceptions and judgements about new situations (Brown & Yule, 1983;Tannen, 1993). We acquire frames during the process of our socialization in the early years of our life (Lakoff & Ide, 2005;Terkourafi, 2005Terkourafi, , 2008. She proposes that both interlocutors judge the conversational norms in a particular situation and then try to have an unmarked (appropriate) status within that frame, because being polite is both rational and cost-effective. So politeness does not make the processes of 'making meaning' and 'socially fitting in' difficult for the interlocutors but rather makes them simpler and easier.
The fourth influential account of politeness, Spencer-Oatey's (2000) rapport management model does not view impoliteness as a 'face-threatening act' but rather as a 'rights-threatening act' and suggests that politeness is an effort to develop, promote and maintain harmonious social relations (rapport), made on account of both interlocutors (Aoki, 2010;Félix-Brasdefer, 2008;Spencer-Oatey, 2000. Spencer-Oatey (2000 maintains that there is no such thing as 'negative face wants.' All members of a society only have positive face wants that revolve less around maintaining face and more around maintaining our sociality rights. For example, when we feel that we have been pushed to do something by a person who did not have any right to do so, we feel offended, uncomfortable or angry etc., but we do not necessarily feel that we have lost our face. She believes that politeness research should be concerned more with the (dis)harmony in social relations (which, according to her, is based on subjective social judgements and can vary from one culture or social situation to another) and less with the "linguistic strategies per se." She accepts 'face' "as a crucial base factor" but also suggests that the researchers should broaden their perspectives and should incorporate other elements (e.g. sociality rights) into politeness research, too (Spencer-Oatey, 2000, pp. 17-18).
When investigated in the context of urban, middle-class, educated, Urdu and Punjabi speaking bilingual Pakistani couples' intimate relationships, to what extent four of the above stated politeness theories will test against the first-order account of the couples, is the objective of this study. The 'urban' in this study refers to five central cities of the Punjab province of Pakistan: Lahore, Gujranwala, Sialkot, Sheikhupura and Kasur. Rationale behind the delimitation of the study is based on three factors. First, marital relationship is different from other types of relationships (Robinson, 2003;Wilson, 2015). What makes married couples' relationship unique is the intimacy of the relationship, mutual interdependence and shared vision of goals (Cuyler & Ackhart, 2009;Mead, 2013). Wedlock, especially in the Pakistani society, is a very serious and constant commitment and the social fabric of the society makes it extremely hard for the either partner to break free of it (Malik, 2006;Mohiuddin, 2007;Rashid, 2015). It makes this relationship even more special, on the one hand, and usually taken for granted, on the other, and these two factors obviously tend to influence the attitude of one spouse towards another (Chapman, 2009;Smart, 2017). Second, Pakistan is a highly stratified society and also enormously diverse in cultural, ethnic and religious terms (Chandio, 2012;Paracha, 2014). We can expect to find one social group wildly different from another group with the addition of each single denominator as, for example, middle-class educated families can be significantly different from the upper-class educated families and the middle-class urban families may not be very similar to the middleclass rural families. Third, for urban middle-class families of the province of Pakistani Punjab, Urdu is the preferred language of communication but they find it easier to express their emotions in Punjabi than in Urdu, especially when they are romantic, excited or angry. So, although English is frequently code mixed into Urdu, the communication mostly takes place in Urdu with regular code mixing of and code switching to Punjabi (Rafi, 2017;Talaat, 2002;Umer-ud-Din, Iqbal, & Khan, 2011).

Theoretical Underpinnings
This study is placed within the constructivist qualitative research paradigm, in which the researcher emphasises "to understand a phenomenon from the perspective of those experiencing it" through observation and detailed interviewing (Fosnot, 2005;Given, 2008;Lincoln & Guba, 2016). Through participant observation of the linguistic behaviour of the study participants, within their culture and the natural settings, over an extended period of two to six years, the study is grounded in ethnography (Fetterman, 1998). And in its effort to understand the phenomenon of linguistic (im)politeness from the perspective of the study participants through detailed interviews with them, it is also grounded in phenomenology (Moran & Mooney, 2002). Constructivism emphasises that the understanding of a phenomenon is co-constructed between the researcher and the study participant(s) through dialogic interaction, within the natural research settings (Fosnot, 2005;Given, 2008;Lincoln & Guba, 2016). This interaction is important because reality does not exist independent of an individual and outside a particular context (Given, 2008;Mertens, 2014). The reality is not simply "out there" but "in there" in the mind of the individual who is experiencing it and who gives it a meaning through social interaction (Given, 2008;Lincoln & Guba, 2016). This meaning is shaped by the blend of internal and external factors (Berry, Poortinga, & Breugelmans, 2011). The internal factors include predispositions, past experiences and the existing knowledge while the external factors are social, natural, contextual and accidental ones (Berry, Poortinga, & Breugelmans, 2011;Mertens, 2014). Owing to the interdependence of the two types of factors, the reality is best understood when its understanding develops through the joint construction of etic and emic perspectives (Fosnot, 2005;Given, 2008;Mertens, 2014). That is why postmodern researchers (see e.g. Huang, 2017;Terkourafi, 2012 andWatts, Ide, &Ehlich, 2005) emphasise that first-order politeness (termed as politeness 1) and the second-order politeness (termed as politeness 2) should go hand-in-hand, and the same approach has been adopted in this study.

Study Sample and the Data Collection
The researcher's office of the clinical and counselling psychology provided a good natural setting for the investigation of the role and the nature of linguistic (im)politeness in the married couples' relationships. During the course of the study, 227 married couples sought (psychological) marital counselling from the researcher (a counselling psychologist and linguist). As a phenomenological study requires the selection of those participants who have experienced the phenomenon themselves (Creswell, 2013) and they serve the purpose of the study well by matching those characteristics which have been specified in the delimitation of the study (Johnson & Christensen, 2010), the purposive sampling technique was employed. So out of the available 227 married couples, those 21 couples were selected whose marital relationships were mainly affected by the linguistic impoliteness and their training in linguistic politeness significantly helped them out. (For detailed selection criteria, please see Appendix II).
From the point of view of minimizing the interference of intervening variables, (Creswell, 2013;Morrison, 2012), these couples were then divided into two sub-groups. 5 couples in Group 1 are those couples, in whose problems, linguistic politeness played the most crucial role and the interference of other factors was perceived to be minimal (by the researcher as well as the participants). The factors considered for the selection criteria include the involvement of other members of the family, infertility issues, difference of age, education and/or family backgrounds and the forced marriage etc. (For the detailed list of these factors and the selection criteria for the Group 1, please see Appendix II.) The factors were finalized after detailed discussion with 5 elderly couples of the urban, middle-class, Urdu and Punjabi speaking backgrounds. Couples in Group 1 are all aged between 28 to 35 years, have university level education, have been married for about 4 to 8 years, have 1 to 3 children and their marriages are both loved and arranged ones. All five hail from urban, middle-class, bilingual (Urdu and Punjabi) speaking, moderately Islamic families of Lahore. All five couples had 6 to 10 counselling sessions with the researcher with an interval of a minimum of 1 and a maximum of 6 weeks between the two sessions.
16 couples in Group 2, however, come from a variety of family backgrounds, are aged between 21 to 44 years, have been married for 1 to 16, have differences of qualification and age-gap and live in varied family structures. But all these couples also sought marital counselling having been mainly disturbed by the linguistic impoliteness and their training in linguistic politeness also helped them out. They belong to five urban areas of Pakistani Punjab: Lahore, Gujranwala, Sialkot, Sheikhupura and Kusur. Couples in Group 2 had from 4 to 13 counselling sessions with the researcher with an interval of a minimum of 1 and a maximum of 6 weeks between the two sessions.
Both ethnographic and phenomenological conventions require that the interaction between the researcher and the participants takes place with "an open mind but not an empty head" (Fetterman, 1998;Given, 2008) and the investigation is well-informed by the extensive literature review and well-controlled by semi-structured interview questions and observation protocols (Fetterman, 1998;Moran & Mooney, 2002). To fulfil this requirement, a list of 36 investigation questions was prepared after an extensive study on the politeness research. For more control and detailed investigation of the phenomena, these questions were then subdivided into 6 different categories. (For the list of these questions, please see Appendix I. It is also pertinent to note here that the technical jargon of the questions is due to the fact that they were used as the researcher's reference and not as the interview questions as such).
After each session of at least half an hour to a maximum of two hours with each couple, the researcher took notes for each of the 36 questions (separately) and prepared questions for the next session. After each next session, the notes were revised, edited and refined. During the sessions, the couples' communication included dialogic interactions (both between the counsellor and the spouse(s) and between the spouses themselves) and the spouse(s)' reflections, comments and interruptions. Paralinguistic communication was also observed and recorded. In most of these sessions, the couples had discussions with the counsellor together, with only a few exceptions, when the other partner was physically present, could see both the counsellor and the client through a transparent glass but could not hear them. The notes of the researcher were later analysed both qualitatively and quantitatively with the help of the software (NVivo 11 and Microsoft Excel 2016).

Ethical Considerations
Because the couples contacted a professional psychologist for counselling and not for the participation in the study, it was cautiously ensured that the time and the money of the clients were not wasted in any way, the research observation and/or the questions did not pose any psychological and/or social harm to them and their counselling was not compromised. As the ethics of both counselling psychology (Altmaier & Hansen, 2012; "Ethical principles of psychologists and code of conduct," 2017) and qualitative research (Halai, 2006;Soltis, 1989) require, the consent from each couple was formally obtained after explaining all details and potential hazards to them. The sessions were not recorded in the audio or the video form as none of the couples was comfortable with it, such recording could be a valuable resource for the researcher though. As each couple was offered to withdraw or refuse at any time, 3 out of 24 eligible couples preferred to withdraw from the study, due to which only 21 couples could be included in the study. In order to ensure the confidentiality of the couples, any such information (e.g. the name and the location of the counsellors' office, the identities of the couples and their demographics and the years during which they sought psychological counselling etc.) that can reveal their identity has been concealed in this paper.

Analysis of the Data and Discussion
To begin with the face-saving model, majority of the couples were found to be clearly aware of the concept of face and face wants (see Table 1 for the statistical description of the data and Examples for sample instances).
(1) Beevi: Saaryan de wich beh ke aihne meri bisti keeti. This finding raised a question: do other couples not have any such concept or they are not actually conscious of it? Detailed discussions with the couples in the later sessions revealed that they also had a clear concept of face and face wants but they were concerned more with sociality rights than face (some of them also viewed sociality rights and face as substitute concepts though). These couples do not feel insulted or demeaned as such, but feel angry, frustrated and/or depressed etc. instead (some of them experience mixed feelings, too, though).
( Linguistic choices of both types of couples and their interpretation of an act as facethreatening or rights-threatening, however, depends on their assessment of weight: the comparative social status, the closeness of the relationship and the sensitivity of the topic assigns their partner the right to speak a certain utterance and/or employ a certain politeness strategy. But the assessment of none of these three factors is universal or constant (as suggested by the face-saving model), nor do the majority of the couples consider any utterance inherently polite or impolite. Their assessment is very dynamic and keeps changing from one situation to another (as observed by the discursive theorists). It all depends on how the frame of mind of the interlocutors, at that particular moment, views and places one partner in relation to another and how they assign meaning to the social and/or the linguistic context of an utterance.
[Laughing] Ghussay main to tareef bhi tanz lagti hai. (Urdu) (Wife: It all depends on the situation, otherwise he habitually starts with a swear word. Look, for example, even praise sounds sarcastic when you are angry.) (7) Shohar: Tum ghar par aa kar baat karti to mujhe bura na lagta. Haaji Saab ke saamne tumhain nai kehna chahye tha. (Urdu) (Husband: Had you said that having been home, I wouldn't have minded it. But it was wrong to speak about it in the presence of Haaji Saab.) Next, negative face wants of the couples are stronger and clearer than the positive face wants as they seem to be concerned more with protecting their own face than saving the other's. They feel badly hurt when their own face is threatened and interpret it as totally unacceptable and unjustifiable, but when they themselves threaten the other's face, they find it totally reasonable and well-justified. Speech acts of apology and acknowledgement occur very infrequently and when they occur, they are usually preceded by long stretches of argument. Mostly performed speech acts are advice, rebuke and criticism. First, these findings are against the frame-based model that suggests that interlocutors try to have an unmarked status in the conversation. Then, it also indicates that couples' communication is not intended for seeking harmony in the relationship, as suggested by the rapport management model, but is rather directed towards negotiating their identity and contesting for power, as proposed by the discursive model. The conflict of interests and preferences, poor mutual understanding and lack of shared vision and goals are major reasons that compel the couples to contest for power and the construction of desired identity. In their struggle for power, most of the couples find impoliteness the preferred choice of action for exercising power and view politeness as an ineffective strategy to limit undesirable action choices of the other. As far as the politeness strategies are concerned, most of the communication of these couples takes place in 'bald on-record' mode. All other strategies are also employed but mainly not to avoid threatening the other's face but for coercion, expressing romantic love and/or entertainment. Secondly, these strategies do not seem to be mutually exclusive as the couples interpreted the same utterance with several different labels at the same time. Then, most of the couples do not regard lack of politeness as impoliteness. They label different instances of behaviour on the continuum of over-polite (e.g. Urdu: zuroorat se zayada izzat de rahe hain), polite (e.g. Urdu: izzat de rahe hain), neutral (e.g. Urdu: izzat to nahi de rahe lekin be'izzati bhi nahi kar rahe), cold (e.g. Urdu: bas aap apne mood main hain, doosron ko lift nahi karwaa rahe), impolite/rude (e.g. Urdu: mood dikha rahe hain; seedhay moon baat nahi kar rahe; ghuma phira kar be'izzati kar rahe hain) and insulting (e.g. Urdu: seedhi seedhi be'izzati kar di; Punjabi: laa paa ke hath parraa'ti) while polite and neutral being appropriate, over-polite, cold and impolite/rude being inappropriate and insulting being offensive. These three findings are similar to those of the discursive theorists Mills (2004), Roloff and Riffee (2004), Bousfield and Locher (2008), Culpeper (2011) and Johnson (2011 etc. So a better explanation, once again, does not come from the face-saving model but from the discursive account. (11) Shohar: Bad'tameezi nai keeti te aidha matlab ay nai ke tusi barri izzat keeti ay.
Banda marra jaya mood vi wakha sakhda ay. Te zuroorat ton zayaada tameez vi ainj lagda ay banda meesna ay. (Punjabi) With regard to the frame-based model, the couples report that their partners' behaviour is mostly predictable, but at times, it also turns out to be surprisingly different. It indicates that frames do play an important role in shaping their behaviour but frames are also vulnerable to both sudden and gradual change under the influence of the current mood, new information, new experience(s) and the then relevant social and linguistic context. When a person actively behaves as a thinking being and when he/she stays passive subject to the internal or the external pressures, in both cases, habitual communication patterns are likely to become prone to occasional variations.
(12) Beevi: Jin dinon main in ki walida hamaaray haan ai hoti hain, un dinon main ye thorra different behave karte hain. (Urdu with English code-mixing) (Wife: He only behaves a bit differently during when his mother has come to stay with us.) As far as the existence of and the conformity with the social norms of conversation between the husband and the wife are concerned, the couples do not always accept that such norms exist or they do not always conform to them. As the discursive model explains, they keep negotiating and renegotiating their role and importance. Many couples, while talking about their partners' actions, complained that such norms existed and were violated by their partners but, while justifying their own actions, they asserted that no such norms existed or the norm was to "get even" with the other. Similarly, they accept that politeness usually goes unmarked as compared to the impoliteness but they do not usually try to have an unmarked status during their conversation.
With reference to the rapport management model, most of the couples admit the importance of developing and maintaining healthy relationships with their partners, but they do not usually make any such sincere effort. They tend to take their marital relationship for granted and mostly put the burden of this responsibility on the other's shoulders. In most of the cases, they do not keep from saying any such thing that can threaten harmony and find their action well-justified and reasonable. Once again, they seem to be struggling less for harmony and more for power. So detailed (qualitative and quantitative) analysis of the data reveals that each model of linguistic politeness explains couples' communication to a certain extent but not completely. One aspect fits into one model while the other corresponds to another. Discursive model, however, appears to be the most robust in its applicability and scope. But it does not take into account the role of frames, nor does it cover those instances of communication which are genuinely intended for rapport management. Most of the couples' strained relationships stemmed from unhealthy beliefs, poor interpersonal skills and annoying habits that they developed mainly during childhood socialization. And the psychological treatment of these variables by the frame-based and the rapport management theorists is far more in-depth, focused and coherent than the "relational work" of the discursive model (see e.g. Terkourafi, 2005;Haugh, 2007;Locher, 2012). Then, as a counsellor, I found face-saving model the most useful (and convenient) in assessing impoliteness and providing politeness training. It was hard for almost all couples to grasp rather philosophical concepts of identity construction, power contestation, rapport management and frames. It was, however, easy to talk in terms of face and politeness strategies: silence is best, negative politeness will always pay, bald off-record is safer, positive politeness is the last option and save bald on-record for only special occasions. But this training was only workable when couples realized the importance of developing and maintaining healthy relationships and when they found themselves powerful and sensible enough to challenge the integrity and the rationality of frames, contextual factors and individual utterances. This leads us to the conclusion that (im)politeness can better be studied within a general framework and its investigation should not be restricted to just one model. Several postmodern politeness researchers (see e.g. Gough, McFadden, & McDonald, 2013;Huang, 2017;Lakoff & Ide, 2005;Terkourafi, 2012;Van Olmen, 2017) have also had similar findings and recommendations.

Conclusion
When studying linguistic (im)politeness in the interpersonal communication of the intimate relationships, in general, and of the Pakistani, urban, bilingual (Urdu and Punjabi speaking), educated married couples, in particular, all four politeness theories provide a significant understanding of the phenomena, but discursive model appears to be the most comprehensive and the best suited description. But because discursive model alone does not encompass all aspects and nuances of these complex and multifaceted phenomena, it is better that they are studied within a general framework instead of being restricted to just one model.