Or & Ong 08 Social Distance and Expressions of Disagreement_Report

GEK 1036 CROSS-CULTURAL COMMUNICATION & DISCOURSE SOCIAL DISTANCE AND EXPRESSION OF DISAGREEMENTS NAME: OR HUIXIAN OLIVIA, PATRICE ONG HWEI YUAN MATRICULATION NO.: U063344W, U063312J COURSE: PHARMACY YEAR 3 DATE: 16TH NOV 2008 Project Report: Social distance and expression of disagreements 1. Introduction Disagreements are part and parcel of daily interactions, functioning to hone one s social skills, build rapport and solidarity with others, and define one s own identity, to name a few. In addition, this speech act has been acknowledged as a component of the learning process; it has been used extensively in the classroom setting as an educational tool (Rees-Miller, 2000). Clearly, disagreements are indispensable in conversations, and it is interesting to note that it can be expressed orally, as well as in the unspoken form. In our scope of study, we will be focusing on spoken forms of disagreement. There have been many studies done on disagreements over the years, and topics covered by them span a wide variety. Among other things, literatures have studied (1) factors influencing disagreement production (such as power, severity, and social distance) (Rees-Miller, 2000), (2) the structure of disagreements, as determined by their use of linguistic markers and their disagreement patterns (Scott, 2002), (3) classifications of disagreements (Turnbull and Muntigl, 1998), (4) educational and political uses (Price V.; Cappella J. N.; Nir L, 2002), (5) different definitions by authors (in relation to power and social distance) (Spencer-Oatey, 1996), and (6) how inter- and intra-cultural aspects can affect disagreement strategies (Georgakopoulou, 2001, and Garcia, 1989). Theories on disagreements and social distance have been put forth by these literatures, but whether such theories are applicable cross-culturally is a big question. In addition, few, or even none of these studies have focused their research in the Singaporean context. There is a lack of data concerning Singaporeans in their use of disagreements. This gap in linguistic research will be addressed by our project, where we will be looking into spoken disagreements made by Singaporeans. More specifically, we have decided to focus on one important factor which plays an important role in disagreement expression - social distance. It is essential to first establish whether there is a relationship between social distance and expression of disagreement among Singaporeans, before thoroughly analyzing the effect of social distance on expression of the speech act. If a relationship does exist, how will social distance affect the frequency of use of aggravated and mitigated disagreements? Comparing the trend from our findings with theories that have been established in other literatures will allow us to conclude the applicability of the theories cross-culturally. It has been expressed in earlier studies that aggravated disagreements tend to dominate when social distance is less. (Brown and Levinson, 1987) Analysis will be carried out using two methods (I) Prevalence of linguistic markers, adopted and modified from numerous studies, including Rees-Miller s (Rees-Miller, 2000) and (II) types of disagreement strategies employed, adopted from Muntigl and Turnbull s study (Turnbull and Muntigl, 1998). The use of both methods will provide us with a broader view of our findings. 2. Definition 2.1. Social distance Social distance can be defined as the measure of the degree of friendship/intimacy or lack thereof between interlocutors. It is an important factor that determines the level of comfort or the amount of politeness and deference that exists in a verbal conversation and as such determines the constraints felt and the liberties taken during the particular speech exchange (Boxer, 1993). Social distance therefore plays a huge part in determining how interlocutors will converse with each other. Social distance in this context can be difficult to measure as it is difficult to determine the closeness of any particular relationship. According to (Spencer-Oatey, 1996), there is no single factor that is responsible for the relative distance/closeness of a relationship and it most likely involves a variety of factors. However no agreement has been reached as to what these factors are. In our report, we intend to interpret social distance as dependent on a few characteristics: frequency of interaction between interlocutors; the amount of trust; affection as well as the interdependence of the participants. It is generally assumed that persons involved in closer relationships tend to interact more often over a greater range of settings. While this generally holds true for many close relationships, exceptions do arise. In some cases, the participants involved in the relationship are not always able to interact frequently due to time and space restrictions. For example, best friends studying in different countries may not have the opportunity to interact regularly by meeting up or conversing due to time zone as well as distance considerations. Conversely, two people (e.g. colleagues, fellow students etc.) may be forced to interact with each other almost every day even though they are antagonistic towards each other. Basing social distance on just frequency of interaction alone is not sufficient. As such, it is important to consider another important factor: affection. According to the article by (Hays, 1984: 78) reviewed in (Spencer-Oatey, 1996), affection is defined as the expression of any sentiment (positive or negative) felt towards each other, or any expression of the emotional bond between partners. Having a positive sentiment such as liking a person in the first place is regarded as a prerequisite in the decision to pursue a current relationship/friendship and strengthen the bonds between the interlocutors. We believe this mutual affection to be the foundations of a close relationship and therefore an important criterion in determining social distance between interlocutors. However, care must be taken not to fall into the simplistic assumption that close relationships are categorized by positive shared sentiments alone. Many strong and stable relationships are also characterized by negative sentiments apart from positive. In fact some relationships which are characterized mainly by negative sentiments may be surprisingly stable over time. (Spencer-Oatey, 1996) Apart from mutual affection, trust is another important component of any close relationship. Trust determines how much the interlocutors are comfortable in revealing certain aspects of themselves to the other party. In Wierzbicka (1991), the author defined closeness as the readiness to reveal to some particular persons some aspect of one s personality and of one s inner world which one conceals from other people; a readiness based on personal trust and personal good feelings. Examples are: sharing feelings, opinions, secrets/confidences or personal topics with each other. Lastly, we chose to take into account the degree of interdependence between the interlocutors. This is important in considering how dependent the two interlocutors are on each other and the extent they are able to impose on each other. Based on the definition by (Hays, 1984:78), we will view each interlocutor as a helper in providing goods, services or support; expressing concern for the other s well being. In fact, the closer the relationship between the two interlocutors, the greater the amount of benefits they will obtain from each other. 2.2. Disagreements Disagreements are specific face-threatening acts that have the potential to jeopardize the solidarity between the speaker and the addressee (Rees-Miller, 2000). Disagreements can be produced either verbally or non-verbally in retaliation to a prior statement made by the first interlocutor. This speech act can take place as two turn sequences, 3 turn sequences or even more. For the ease of our investigation, we will define a disagreement act as a verbal exchange consisting of two turns; similar to that done in Rees-Miller s (2000) paper: T1: Statement or claim put forth by the speaker T2: Addressee disputes the T 1 statement or claim In T1, the Speaker, S will utter a claim/statement. Following this, in T2 the Addressee, A will react to the T1 statement which he/she considers to be untrue by producing an alternative view that is in opposition to the original T1 claim. Disagreements made in T2 can be carried out by A in a variety of ways: aggravated, neutral, or mitigated. These disagreement classifications actually exist on a continuum, with aggravated and mitigated disagreements lying at the two extreme ends and neutral disagreements occupying the centre. 3. Guiding Principles According to Brown and Levinson s study, each person possesses both a positive and a negative face. The positive face refers to the person s need to be approved of by others, to maintain solidarity with others. Negative face speaks of the person s need to remain autonomous and unimpeded. According to Turnbull and Muntigl (1998), disagreement acts are inherently face-threatening as they express disapproval of another person and hence jeopardize their positive face. For example in the disagreement act (1), the addressee expresses disapproval of the speaker s personality: (1) T1: So I told him off... Thought that was the only [right thing to do] T2: [Told him off?] That s not the only thing you can do. Why are you always so hot-tempered? Disagreements also threaten negative face by having the potential to limit the Speaker, S s options as in the example (2) below: (2) T1: That shirt is cheap... I think I should get it. T2: No. It s ugly. Don t waste your money on this. In (2), the addressee in T2 is seen to constrain the speaker s next move in terms of what he/she should or should not do (spending money in this case). Disagreements are therefore determined by Turnbull and Muntigl (1998) as crucial interactional activities that involve facework (the negotiation of face). It is assumed that both interlocutors will be mutually interested in maintaining each other s face via the usage of politeness considerations that both inter ocutors would feel the need to preserve social harmon disagreements that were less aggravating to face. Levinson and Brown put forth a formula that showed the relation of social distance with weight of a § face-threatening act (i.e. amount of politeness considerations required) Weight of face-threatening act = D (social distance of the interlocutors) + P (relative power of interlocutors) + R (Rating of imposition). The weight of the face-threatening act will then determine the disagreement strategy chosen by the interlocutor (softened or aggravated). This relationship implies that more direct face -aggravating disagreement strategies will be employed when social distance is lesser and vice versa (Brown and Levinson 1987). ¤ ¨ The chart below (Fig. 1) will illustrate the relationship between social distance and type of disagreement act used more clearly: Social Distance Politeness Considerations Type of Disagreement Act Fig. 1: Relationship between social distance and disagreement acts This is supported by Boxer (1993) who maintained that social distance is regarded as one of the important factors that will determine the way interlocutors converse because it will affect the lev of el politeness and deference in the conversation. This will in turn determine the constraints felt and liberties taken in speech exchanges. n the case of disagreements, social distance will determine whether the disagreement act used is softened or aggravated. © ¤ ¦ ¥¥¥ ¤ show consid r tion for each other s face needs. ( ees -Mi er 2 £ ££ ¢ ¡  ). The author went on to state that through the use of softened Therefore when social distance is greater between the two interlocutors the use of politeness considerations increases and this results in the preference for more softened disagreements. Conversely when social distance is less, there is a tendency for the use of aggravated disagreements as interlocutors adhere less to politeness considerations. 4. Methodology In this section, three aspects of our methodology will be elaborated upon chosen, the data collection method, and the data analysis methods. 4.1. Subjects It is imperative for one to understand that social distance is a continuum (F   the subjects that we have 2). To fully account for the effects of social distance, our findings should reflect data from different parts of the spectrum. Hence, taking into consideration the definition of social distance that has been explained earlier    (S t 2 1), we focused our attention on three social groups with differing relative magnitudes in Acquaintances, Friends, and Family. Among them, Acquaintances have the most social distance social distance, because they are likely to have the lowest frequency of interaction, lowest degree of interdependence and trust, and neutral affection, since meetings between such interlocutors are usually quite transient. On the other extreme, Family will have the least social distance. Disagreements were made by the participants in the English language. Friends Acquaintances Fi . 2: Illu t ti t i t t li Fri , Family are i t t i l t t i t i i t. T ti uu , it l tiv iti Family i f A qu i t i l , % &" &" &"' ' %" " % ! "! !) ' ' '# #  '# ! ''!) '" '#  '%"  & $ $ %! "  '%'& # ( " "! %  '%"  & %! # !# ! "!    i ated. 4.2. Data Collection In a period of three weeks, field observation of disagreements occurring around us was carried out. To collect data for the groups of Friends and Family, we observed disagreements that occurred among our own friends and family members, additionally participating in some of the disagreement ourselves. Once a disagreement act was made, it was noted down. Acquaintances were observed for their use of disagreement in two settings: the classroom, and co-curriculum activities (CCA) meetings. Data from the classroom setting were taken from linguistic classes, and Pharmacy tutorials, where there was an extensive exchange and debate of ideas. The advantage of using field observation was that it was the least intrusive method that allowed us to gain an insight into naturally occurring interactions (ReesMiller, 2000). While a direct recording would capture the interlocutors attention, and may result in modified actions, field observation is done discreetly so as not to alert the interlocutors, hence avoiding a situation where they may subconsciously choose to respond differently from what they initially planned to do. In addition, this method of data collection was more favourable as compared to discourse completion tasks (DCTs), questionnaires and role plays, because unlike the former, these methods may provide data that do not correspond with natural data. To record down the disagreements taking place, we employed the use of pen and paper recording, similarly to that done in Rees-Miller s (2000) study where she analyzed the effects of power and severity on disagreement expression. This method was highly recommended in her study because it has been acknowledged by other linguists as a method suitable for recording short segments of speech, which disagreement acts are. The data recorded was then transcribed and the turns of disagreement acts (T1 and T2) were identified. On the whole, there were 75 disagreement acts, 27 from Acquaintances, and 24 each from Friends and Family. The use of linguistic markers, including rising intonation, stress, and increased volume, were taken notice of during field observation and indicated in the findings. 4.3. Analytical Framework Qualitative and quantitative analysis of our findings were carried out using two methods: (I) Prevalence of linguistic markers, and (II) the disagreement strategies employed. Method I was essential as it appeared to be a well-established and proven method, since many linguists, apart from Rees-Miller, employed the use of linguistic markers successfully in their analysis of disagreement expression, such as Scott (2002). In order to carry out the project further, and give us a broader picture of our findings, method II was also used. Method II, proposed by Turnbull and Muntigl (1998) in their study on conversational structure and facework in disagreements, was an unconventional way to analyze the findings, because it had yet to be used by other linguists in this field of research. 4.3.1. Method I: Prevalence of linguistic markers Linguistic markers are devices that speakers use to convey their message in different contexts (Scott, 2002). They are important elements for our data analysis because of their exclusive function. In expressing one s disagreement, interlocutors make use of various kinds of linguistic markers to strengthen and soften the speech act, thus producing aggravated or mitigated disagreements respectively (Rees-Miller, 2000). Following the footsteps of Rees-Miller (2000), we identified linguistic markers in the 75 disagreement acts, and classified them into three broad categories based on whether they strengthened or softened the speech act: Aggravated, Neutral, and Mitigated. An overview of our classification is depicted in F 432 10 3. Linguistic Markers Aggravated Neutral Mitigated ‡ ‡ ‡ ‡ ‡ ‡ ‡ ‡ ‡ ‡ No pause Overlaps Absolutes Personal, accusatory you Rising intonation Stress on words Increased Volume Rhetorical Question Profanity Intensifier ‡ Verbal shadowing ‡ ‡ ‡ ‡ ‡ ‡ ‡ ‡ ‡ Discourse markers Modulators Hesitation Question Partial/Pseudo agreement Personal Experience Humour Friendship terms Singlish - + Fi . 3: Classifi ati f li uisti markers To aggravate disagreements, interlocutors can employ the use of the following linguistic markers. After the T1 claim, the T2 speaker either follows up his/her disagreement immediately with no pa s , or may choose not to wait for the T1 turn to be finished by talking simultaneously to assert his/her own point of view (overlaps). Absol tes, such as at all and every, tend to give the impression of disallowing any further negotiation of the topic. The use of the personal accusatory you may be seen as a personal attack on T1. In addition, interlocutors can also use rising intonation, stress on words and increased volume, as well as rhetorical questions. Our findings had uses of bullshit, siao (which means crazy in a dialect, Hokkien, used in Singapore), and nonsense, which we classified as profanity. Such usage may make the T2 speaker come across as being rude, and may appear offensive to the T1 speaker. Intensifiers, which included so, very, and a few others, were also noticeable in our findings. In the category of neutral linguistic markers, there was only one linguistic marker identified verbal 5 6 5 shadowing. This is where the disagreement in T2 repeats the claim in T1, but with alte ration in words and intonation. (Rees-Miller, 2000) To soften disagreement, both positive and negative politeness strategies were carried out by interlocutors. In the category of negative politeness, discourse markers, modulators, hesitation and questions are used. These linguistic markers serve to respect the T1 speaker s negative face. Discourse 8 7@ 9 @9 8 7 markers are forms of mitigation as they announce that a disagreement act is about to be delivered (Stalpers, 1995), and examples that appeared in our findings include wait, huh, but, and well. Modulators exist as a big category of linguistic markers which include erm, hm, maybe, probably, I think, and modal verbs (e.g. should). In the category of positive politeness, interlocutors use pseudo/partial agreement, personal experience, humour, friendship terms, and Singlish. All these attend to the positive face of the T1 speaker. In pseudo/partial agreement, the T2 speaker agrees with T1 first, with words like yes, that s true, before disagreeing. Personal experience and Singlish may enhance positive face as T2 uses strategies that implicitly tell T1 that the interlocutors are interconnected in a certain way, by the fact that they share the same memories and use the same kind of language. In our findings, friendship terms that appeared include dude and man. For each linguistic marker, examples from our findings are shown in the appendix. Initially, based on the type of linguistic marker present, we planned to classify the 75 disgreement acts into three categories: aggravated, neutral, and mitigated (similar categories as linguistic markers). However, as the findings unfolded, it was realized that such classification was not feasible. Most of the disagreement acts in our findings used a mixture of aggravated and mitigated linguistic markers, as shown in the example (3) here: (3) T1: I think that it s dangerous to ask a woman if they wanted a cake. T2: yeh yeh... well would you rather clean up the kitchen floor? ((Laugh)) No one agrees with you. In (3), T2 used a rhetorical question. Rhetorical question is seen as an aggravated linguistic marker. However, a second look at the intended meaning would make the observer realized that T2 was using humour at the same time. There was also use of partial agreement, and the discourse marker of well. All these are mitigated linguistic markers. The mixture of aggravated and mitigated linguistic markers in a particular disagreement act made this classification method complex. In addition, some disagreement acts used more linguistic markers than others, which we do not account for if this A classification method was used. The use of three or four mitigated linguistic mar ers may make a disagreement appear more softened, as compared to the use of only one. In order to avoid the limitations stated above, instead of classifying disagreement acts themselves, we chose to classify linguistic markers into the three similar broad categories of aggravated, neutral, and mitigated, which we have done so earlier. The emphasis here is that analysis would now be made looking at the different types of linguistic markers used by each social group, and not the type of disagreement acts that each social group frequently performed.(Fig. 4) eutral Mitigated eutral Mitigated Aggravated Aggravated DA LM E Fig. 4: Illustration to show initial analysis method that we originally planned to use and the final analysis method that we eventually used. To compare across the groups, the average number of aggravated, neutral, and mitigated linguistic markers used per disagreement act was calculated. An average calculation was used because the three groups did not have the same number of disagreement acts recorded, so averaging (by dividing with the number of disagreement acts) would make the comparison fairer. In addition, the proportion of aggravated, neutral, and mitigated linguistic markers used for each group was calculated as well. This allowed a better view of the type of linguistic marker each group used the most frequently. B D C 4.3.2. Method II: Disagreement strategies used In method II, we decided to analyze our data based on a method used by Turnbull and Muntigl (1998). In this method, we adopted their method of classifying disagreements into four categories: Irrelevancy claim (IR), Challenges (CH), Contradictions (CT) and Counterclaims (CC). In addition, the authors also identified an act combination category: Contradiction followed up with a Counterclaim. (CT + CC) Irrelevancy Claim (IR) - In the use of IR, the T2 speaker dismisses the previous claim as being irrelevant to the discussion and thus not an allowable contribution to the disagreement. IR s are considered by the authors to be the most face-aggravating disagreement act as it attacks the speakers competence as a skilful communicator who is able to make relevant contributions to the disagreement. Besides that, it tends to shut down negotiations or further discussions of the topic at hand. (4) T1: All guys are like that... Can t be trusted. NONE can be trusted. Just look at = T2: = EXCUSE ME! Just because one guy does it, doesn t mean that EVERY guy is like that. Just because you think this way, doesn t mean that it s true ok? In the above disagreement act (4), the addressee in T2 attacks the T1 claim before the speaker has completed the sentence by repudiating it as being irrelevant evidence for the disagreement topic at hand. Challenge (CH) - Challenges typically appear in the form of questions and may take on an interrogative form. Even though CH does not directly oppose the T1 statement, its usage implies that the addressee is not able to provide evidence for his/her prior statement. It is considered by the authors to be the second most face-aggravating act as it directly attacks the competency of the other speaker. (5) T1: I did call back last week mah T2: Where got? In this example (5), the addressee in T2 challenges the T1 claim by demanding that the speaker give evidence of the times when he/she actually called and at the same time shows skepticism regarding the truth of the T1 claim. Contradiction (CT) - The T2 speaker opposes the T1 claim by putting forth a claim that is in direct opposition to it. For example, if the T1 speaker says P, the T2 speaker will respond with a P. CT is considered the third most aggravating disagreement type as it directly repudiates the T1 speaker s claim. However it is still less face-aggravating than IR and CH as it does not attack the conversational skill or rationality of the first speaker. (6) T1: I don t think I have ever lent it to you before. = T2: = And I think you have lent it to me before. In (6), T2 directly opposes the claim in T1. There is also the use of a format known as verbal shadowing (as mentioned in Section 4.3.1). In this technique, the current speaker produces an utterance that is tied by means of semantic, syntactic, morphemic or morphological operations to previous speaker s utterance. (Turnbull and Muntigl, 1998). It has the potential to be rather face -aggravating as it seems to use the T1 speaker s own words against him/her. Counterclaim (CC) - Counterclaims are the least face-aggravating of the lot as it does not directly repudiate the T1 claim and instead proposes an alternative claim. These alternative claims can then serve to open up further discussions and negotiations regarding the topic at hand. For example, in (7): (7) T1: It s not applicable if used as a solvent but instead you can use an organic solvent. = T2: = But manufacturers may not want to use organic solvents which may disrupt the integrity of the filter. Combination Act (CT + CC) - The authors discovered that some disagreement acts were in fact combinations of the various disagreement acts with CT + CC being most significant. Whilst it is considered less face threatening compared to CT, it is still face-aggravating as it contains a CT which opposes the other speaker s claim. However, this effect is softened by the accompanying CC which provides alternative claims which opens up discussion and allows negotiation of the two speaker s claims during a disagreement. As an example (8): (8) T : Maybe it s due to solidarity = T : = Actually I disagree. I think that sometimes it depends on the person you re interacting with. Like... if someone uses a lot of abbreviations, I would tend to use abbreviations also... so as to not make her uncomfortable. The authors next classified the above disagreement acts into 3 categories: (1) Low aggression, (2) Intermediate aggression and (3) High aggression. (Fig. 5) Low Aggression FF Fig. 5: Le els of aggression 5. i di g In this section, we will discuss the findings obtained from Method I and II separately, as these two methods employ a different approach towards analysis of the results. Additional noteworthy findings will also be briefly mentioned at the end. P H I F G S R Q Intermediate Aggression CC High Aggression IR CH 5. . Ave ge e of Agg vated i g i tic arkers er DA Acquaintances , 0.93 Ave ge Family, 1.67 e of e l i g i ic ke e DA Acquaintance s, 0.07 Frien s, 0.75 Family, 0.04 Average Family, 0.63 er of itigated i g istic arker er DA Acquaintance, 1.44 ien s, 0.08 Fig. 6: A erage n mber of Aggra ated, Ne tral, and Mitigated Ling i stic Markers sed per Disagreement Act Frien s, 1.17 Figure 6 shows the findings from Method I, which is based on the prevalence of linguistic markers. First, comparison is made between the groups in order to find a trend onexpression of disagreement as social distance changes. As shown in the chart, Acquaintances had the highest average number of mitigated linguistic markers per disagreement act (1.44 per DA), while Family had the least (0.63 per DA). In the category of average number of aggravated linguistic markers per disagreement act, Family used the most (1.67 per DA), almost twice that of each of the other two social groups (0.93 per DA for Acquaintances, and 0.75 per DA for Friends). All three groups displayed infreque t use of neutral n linguistic markers; hence discussion of results would be focused on only aggravated and mitigated linguistic markers. cf ` e d XWc` Y  ‘ i p i ih g W ba`Y ‘ ‘  V UT XW ‘ wr vu ` ed V hi y x p i gsrq t x ba`Y ‰ ˆ‡ „‡ † € € i hi „ƒ‚ 5. . et i … V  € Li g istic arkers Used for Acq ai tances Aggravated 38% Mi igated 59% Neutral 3% Ling istic arkers Used for Aggravated 38% Ling istic arkers Used for amil Mitigated 27% Neutral 4% When the proportion of the different types of linguistic markers used by each social group was considered (Fig. 7), an unexpected observation was made. Statistics for Acquaintances and Friends appeared similar. For Acquaintances, 59% of the linguistic mark they used were the mitigated type, ers and 38% were aggravated ones. This goes the same for Friends, where 58% were mitigated linguistic markers, and 38% were aggravated ones. For Family, results were significantly different from the other social groups. 71% of the linguistic markers they used were the aggravated type, while mitigated linguistic markers only made up 27%. 5. . . Anal sis Using the findings (as presented earlier), there were two observation made. Across the groups, it was seen that all three groups had different uses in the type of linguistic markers, though Acquaintances and Friends do display results that were slightly similar. Nonetheless, this shows that social distance does have an effect on disagreement expression. If such is the case, how exactly does it affect expression Referring to the chart, as social distance increased (from Family to Acquaintances), the h g fe j j Aggravated 71% j i Neutral 2% Fig. 7: Proportion of Aggra ated, Ne tral, and Mitigated Ling i stic Markers sed in each social gro p. ˜ — – j d ™ ’ “ ” • • ” “ ” riends Mitigated 58% use of mitigation increased. At the same time, use of aggravated linguistic markers decreased. This trend was made by comparing mainly the statistics belonging to Family and Acquaintances, because it was noted that the statistics for Friends, being similar to that of Acquaintances (these similarities will be explained subsequently), made the trend less obvious and noticeable. The observation made was in line with what we expected. As explained earlier in guiding principles ( section 3), Brown and Levinson had included social distance in the formula that decides one s expression of disagreement. In addition, literatures have proposed that more aggravated disagreements are used when social distance is lesser. The disagreement act, as we have defined, is face-threatening. (Rees-Miller, 2000) As a result, in performing this speech act, there is always a risk of jeopardizing the interlocutors relationship. This risk, however, varies among the different social groups. For family members, the face-threatening act is less likely to jeopardize the relationship because the ties are strong. Furthermore, instead of softening the disagreement, use of mitigation with family members may sometimes come across as being sarcastic. As social distance increases, the risk of the act jeopardizing the relationship increases at the same time. This occurs because such relationships are relatively less stable. Interlocutors here would perform more facework and politeness considerations in relation to the high risk factor (Stalpers, 1995). One may argue that since interlocutors in the groups of Acquaintances and Friends are more concerned for the effect of the speech act on the outcome of the relationship, why do they not use mitigation all the time? Use of directness (in the form of aggravated linguistic markers) may not always lead to an unfavourable situation; its use may often indicate the interlocutor s wish to be seen as socially close. This may explain why aggravated linguistic markers were still being used in the context of Friends. For Acquaintances, disagreements were observed in the classroom and CCA meetings. In such serious formal setting, it was acceptable to show conflicting views, as this action was seen and acknowledged by the interlocutors as part of the work/classroom culture (Stalpers, 1995). Because interlocutors do not take it personally when disagreed upon, the use of politeness strategies to mitigate and soften one s disagreement was not a big requirement. Interlocutors place priority on the conversational maxim of Clearness over politeness . Hence, they may not feel restricted in using aggravated linguistic markers in their speech. Apart from recognizing the link between social distance and disagreement, a second observation was also made. The statistics between Acquaintances and Friends did not appear to be significantly different. Results were quite alike, with only minor differences. This could be possibly due to the important fact that social distance is a continuum. It may not be possible for one to see each social group as being distinctively different. This is especially so between Acquaintances and Friends, where the line between them may be blurred. As a limitation of our field observation method, the relationship between interlocutors may not be what they appear as to the observer. As observers, we do not have the ability to know when interlocutors cross the line from Acquaintances to Friends, which could possibly lead to such similar results between the two social groups. The second reason would be the context in which disagreements were made by Acquaintances. As mentioned in the previous paragraph, such serious formal settings allowed more use of aggravated disagreements. 5.2. Method II 5.2.1 Findings Our findings in method 2 did not show the results we had expected: as social distance increases, the use of aggravated disagreements decreased. The results obtained for this method were inconclusive and a clear trend was not observed. 5. . A. i ee e t t te ie U e Figure 8 shows that acquaintances actually used the greatest percentage (48.15%) of CT + CC. This is followed by CC which is the next highest (29.62%). CT + CC is the most frequent strategy used probably because, as said earlier, most of our data for acquaintances were taken from more formal settings such as CCA meetings, classroom settings (tutorials and lectures). In these settings, it is the Fig. 8: Disagreement strategies sed b ac aintances norm for disagreements to be stated clearly and directly accompanied by a supporting reason statement to lend weight to the disagreement. This also allows the opening up discussions and further negotiations. It is also important to mention that no highly aggravating IR claim was seen to be used amongst acquaintances. v As shown in Fig. 9, friends actually used the highest number of CC (41.67%). This shows that friends actually favoured disagreement acts that allowed for further negotiations and discussions of the matter at hand in a bid to reach a compromise. It is also significant to note that a new combination category of CH + CC that was not mentioned in the author s study actually surfaced 45.00% 40.00% 35.00% 30.00% 25.00% 20.00% 15.00% 10.00% 5.00% 41.67% 20.83% 20.83% 0.00% 12.50% 4.17% 0.00% Disagreement Types Fig. 9: Disagreement strategies sed b friends in our data. As in the case of acquaintances, no highly face-aggravating IR claim was used. x x w w w y u n 60.00% 50.00% 40.00% 29.62% 30.00% 20.00% 11.11% 11.11% 10.00% 0.00% 0.00% Disagreement Types n p oq t s r q po n m lk 48.15% IR CH CT CT + CC CC IR CH CT CT + CC CC CH + CC In Fig. 10, contrary to what we expected, CT + CC and CC were used most often by family members. We had actually expected a trend where family members would use more IR, CH or CT disagreement strategies. We had expected more face aggravating strategies to be used. It is also important to note that the new category of CH + CC persists. And within family the use of IR was finally seen. Even then it is infrequently used. 5. . B. Aggression levels {z 35.00% 30.00% 33.33% 29.17% IR 16.67% 12.50% 4.17% 4.17% CH CT CT + CC CC CH + CC 25.00% 20.00% 15.00% 10.00% 5.00% 0.00% Disagreement Types Fig. 10: Disagreement strategies sed b famil 59.26% 60.00% 50.00% 40.00% 30.00% 20.00% 10.00% 0.00% Acquaintances Friends Family 11.11% 29.62% 41.67% 37.50% 29.17% 20.83% 20.83% High Intermediate Low 50.00% From the chart on the left (Fig. 11), by comparing across the groups it can be seen that friends and family used the same percentage of high aggression disagreement strategies used the whilst acquaintances no least. However, apparent trend is seen when comparing across the groups for low andintermediate aggression. When comparing within each group, friends are seen to use more low aggression disagreement strategies. Aggravated disagreements are used least amongst friends. When comparing the distribution of aggression levels within family with acquaintances, it is found that the shape and pattern of distribution is rather similar. The significance of this will be discussed later. } } | 5. . . Anal sis From the findings, insufficient evidence exists to prove the trend: that as social distance increases, the use of aggravated disagreement acts decreases and vice versa. Friends and family shared the same percentage for high aggression disagreement strategies. Family was expected to show a higher percentage than friends. However acquaintances did show the lowest percentage for high aggression disagreement strategies as expected. For low aggression and intermediate aggression strategies no obvious trend was seen across the 3 groups. Amongst friends, it is seen that low and intermediate aggression disagreement strategies arefavoured over high aggression strategies. In fact the most favoured disagreement strategy was the low aggression CC. A reason for this could be because friends do not want to jeopardi e their relationship. According to (Laursen, 1998), poorly managed disagreements are increasingly regarded as forces that  € threatened close peer relationships. Disputes with close peers are more likely to involve compromise thus avoiding damage to the relationship . ‚ „ Fig. 1 : Where 1=Ac aintance, =Friends, =Famil on the x -axis † … ƒ ‚  ~~ 70.00% 60.00% 50.00% 40.00% 30.00% 20.00% 10.00% 0.00% 0 0.5 1 1.5 2 2.5 3 3.5 11.11% 41.67% High 37.50% 29.62% 20.83% 29.17% 20.83% Intermediate Low 59.26% 50.00% The lack of trend for low and intermediate aggression levels (Fig. 12) can actually be explained using Wolfson s Bulge Theory which was actually used by Boxer (1993). This theory states that: when the ways in which different speech acts are realized in actual everyday speech are examined, and when these behaviors are compared in terms of the social relationships of the interlocutors, it is found that the two extremes of social distance (minimum and maximum) seem to call forth very similar behavior, while relationships which are more toward the center show marked difference. This trend proposed by Wolfson is seen for low and intermediate aggression distributions across the 3 social groups. Wolfson proposed that social groups at both extremes of the continuum actually shared similarities in the relative certainty of their relationships. Acquaintances are regarded as distant relationships that may not be as important to the interlocutor and as such the interlocutor may not bother to use solidarity establishing behavior (low aggression disagreements) during interactions to preserve harmony. For family, interlocutors already have strong blood ties and close bonds and as such they see it as irrelevant to use solidarity establishing behavior such as low aggression disagreement acts. The same factors could also be used to explain the trend seen for intermediate level aggression acts. Such acts, which are considered face-aggravating, were used more frequently in Acquaintances and Family, because of the relative lack of solidarity consideration. However, whilst The Bulge theory is able to explain the results for low and intermediate, it cannot be used to explain the trend observed for high level aggression disagreement acts. As such, we believe that the results we have obtained are inconclusive possibly due to limitations of method 2. 5.2.3 Limitations of Method 2 The inconclusive results found in method 2 could be due to several limitations encountered. First of all, it was found that classification of disagreement strategies was rather limiting. There was a tendency to force a particular disagreement act to fit a particular category without considering if new, additional categories existed other than those proposed by the author. Apart from that, the author only took into consideration the CT + CC combination. In our study, we actually discovered a new combination: CH + CC which we considered an intermediate aggression disagreement act. However, we lack the means to justify our classification as there was no precedence for this combination act in the author s study. As such, the model used was too limiting and subject to exceptions. Besides that, method 2 also does not take into account the use of linguistic markers which can increase or decrease face threat. It also fails to take into consideration what is known as attributive disagreements, that is, it attributes some action, thought, or attitude to the speaker which is in opposition to what the speaker has said. (9) T1: I think I ll be getting that. Should come in handy = T2: = That useless stuff? Please lor. How many times have you bought something that you thought was useful, but ended up not using? In the above conversation (9), a CC (Counterclaim) is used. However it is more face-aggravating than a normal CC as the addressee in T2 attributes that the speaker does not have the capability to make useful decisions. It also puts the addressee in a position superior to the speaker in implying that he/she knows better. Therefore, CC does not automatically mean that a disagreement is softened. Apart from that, classification of the disagreement strategies can be very subjective. The sample size used was also rather small due to time and resource constraints and therefore the results obtained may not be representative of a true population. Lastly, this method is actually a novel method used by the authors for their study and has not been replicated elsewhere. As such, this method may only be applicable in their study. 5.3. Additional noteworthy findings In the group of Acquaintances, there were instances where the agreement maxim was observed even in aggravated disagreement acts. In these acts, there was use of mitigation. As demonstrated in (10), T2 uses the disagreement strategy of Challenge, which is seen as having a high level of aggression. T2 is showing his/her disagreement by questioning T1 and asking T1 to provide evidence for his/her claim. However, the disagreement act was also marked with mitigation. There was use of hesitation and the discourse marker Erm. In (3), T2 uses a rhetorical question to show his disagreement. Rhetorical questions can aggravate the disagreement act, but the T2 speaker chooses to use humour and partial agreement to mitigate the act. (10) T1: I think that maybe if you give Singaporean students a response form they will feel it s an exam and tend to give models answers T2: Erm upon what you based your assumptions? In addition, interlocutors in the category of acquaintances did not use Singlish (which is seen in the other two groups), and personal experience (which was only seen in Friends), both of which are forms of positive politeness. The absence of Singlish could possibly be due to the context that disagreements were observed in; it may not be appropriate to use Singlish in classroom discussions and meetings. Moreover, a simple calculation (Fig. 13) showed that Acquaintances used more negative politeness compared to Friends. One reason attributing to this behavior is that Acquaintances, having the most social distance among the three groups, may not see it appropriate to use positive politeness to enhance the relationship, since in the first place, the relationship is relatively less stable, and could be a one-time occurrence only. Negative politeness may be preferable to protect one s autonomy. Percentage of negative 87.18 politeness Percentage of positive 12.82 politeness Fig. 13 Proportion of use of negative and positive politeness among the groups (out of the total number of mitigated linguistic markers) ˆ In the group of Friends, interlocutors used positive politeness strategies, such as Singlish, friendship terms, and personal experience. The use of the highest percentage of positive politeness in Friends, as compared to the other two groups, could be due to the interlocutors desire to take the relationship ‡ Acquaintances Friends Family 60.71 6.67 39.29 13.33 up a higher level, because unlike Acquaintances, interlocutors here may have a relationship that is longer and less volatile. It was expected that frequency of aggravated linguistic markers used would have been higher, instead of being almost on par with that of Acquaintances. It is possible that the interlocutors might have been more concerned for the consequence of the disagreement act on the relationship; hence they may not wish to perform a face-threatening act that will affect the relationship (Laursen, 1998). For Family, it was observed that there was no use of partial agreement, personal experience, or friendship terms. On the whole, the use of positive politeness was lesser as compared to Friends. If the explanation for Friends holds true, (where their relationships are more stable as compared to Acquaintances, and thus they may desire to enhance the relationship), shouldn t Family, with the least social distance, use the highest number of positive politeness strategies? We propose that the less use of positive politeness strategies could be since the relationships are relatively more stable and ties are stronger, family members may not see the need to use such strategies to enhance the relationship. 6. Limitations of Our Project One of the limitations of our project is that we did not take into account the severity of the T1 claim in influencing how the addressee would respond. This variable was difficult to control whilst still maintaining a natural data collection at the same time. Besides that, it is also impossible to predict when a person would actually find a particular T1 claim severe/face-threatening as this is a subjective parameter. Apart from that, power could also not be standardized across all the 3 groups as power relations would tend to naturally differ for each social group. There is also the limitation of field observation. Field observation requires a lot of observational skill and as it was our first time conducting it, we may have failed to take down certain keywords or linguistic markers. Field observation also does not allow us to record facial expressions which in their own right serve as intensifiers or mitigation tools. For example, smiling can be used to mitigate a disagreement (Garcia, 1989). Apart from that, classification of disagreement acts is very subjective. For example, the observer may classify a disagreement between two interlocutors as being aggravated. However, the interlocutors actually involved in the disagreement may regard it as aggravated. Due to time and resource constraints we only managed to collect a limited sample size which may not have been representative of the true population. And lastly, there may have been statistical errors in our findings. Averaging which was used in our calculations tends to flatten potential outliers and make them fit the general trend. Therefore any statistically significant departures from the trend may have been missed. 7. Conclusion In spite of our limitations, we are able to answer the questions posed at the beginning of the project, which was (1) Is there a relationship between social distance and expression of disagreement? And (2) If there is, is the relationship similar to that established in other studies overseas? From our results in Method I, we can conclude that there is a relationship. Moreover, the trend was as expected: When social distance decreased, there was more use of aggravated disagreements, and vice versa. This follows the theories put forth by Rees-Miller (2000), Boxer (2000), and Brown and Levinson (1987). Therefore, these theories are applicable in the Singaporean context. There are several areas for further research. Method II can actually be reevaluated by addressing the limitations mentioned, to determine if this method is applicable to other studies besides the authors own study. For example, a larger sample size can be used, or more categories can be included to make the classification less rigid. Besides that, there were no positive comments noted in our data, which were present in Rees-Miller s (2000) findings. Positive comments are defined as in (11). (11) T1: T2: That sounds like a good idea, but... In fact, the nearest thing that came to it were partial agreements. This could be possibly due to cultural differences between Western culture and Singaporean culture, as Singapore may not be a complimenting culture. Positive comments could possibly be interpreted differently, for example, as sarcasm. However, this is an area of research that is beyond the scope of our study. Reference Boxer, Diana. (1993). Social Distance and speech behaviour: The case of indirect complaints. Journal of Pragmatics 1 , 103-125. Garcia, Carmen. (1989). Disagreeing and Requesting by Americans and Venezuelans. Linguistics and Education 1, 299-322. Georgakopoulou, Alexandra. (2001). Arguing about the future: On indirect disagreements in conversations. Journal of Pragmatics 33, 1881-1900. Laursen, B. (1998). Closeness and conflict in adolescent peer relationships: Interdependence with friends and romantic partners. In William M. Bukowski, Andrew F. Newcomb, Willard W. Hartup, The Friendships They Keep Friendships in Childhood and Adolescence. [Chapter 9]. Cambridge University Press. Muntigl P., & Turnbull W. (2005). Conversational structure and facework in arguing. Journal of Pragmatics, 2 , 225-256. Price V.; Cappella J. N.; Nir L. (2002). Does disagreement contribute to more deliberative opinion? Political Communication, Vol. 1 . (pp. 95-112). Routledge, part of the Taylor & Francis Group. Rees-Miller, Janie. (2000). Power, severity, and context in disagreement. Journal of Pragmatics 32, 1087-1111. Scott, Suzanne. (2002). Linguistic feature variation within disagreements: An empirical investigation. Text 22(2), 301-328. Spencer-Oatey, Helen. (1996). Reconsidering power and distance. Journal of Pragmatics 2 , 1-24. Stalper, J. (1995). The expression of disagreement. In Konrad Ehlich, Johannes Wagner, The Discourse of Business Negotiation. (pp. 275-291). Walter de Gruyter. ‹ ‰ ‰ Š ‰