9.

35am Group 1 (Compliments)

Caela Bowman

Claudia Harper

Sharon Hunt

Lungelo Nkosi

Introduction:

Conversational analysis can be described as the study of social interactions amongst people. The

manner in which gender (variable) affects the responses to compliments received in isiXhosa

compliments is central to this analysis. We will closely look at how verbal and nonverbal

conduct occur in our everyday conversations amongst isiXhosa speakers and what variable

affects responses receive in isiXhosa. Central concepts in conversational analysis will be

explained and the manner in which linguists go about collecting research will be described. This

research looks at theories of conversational analysis and how we to use the theory to analyse data

we obtained from interactions with isiXhosa first language speakers.

Comment

• What is CA?

Conversation analysis (CA) is a substrate of linguistics which has been defined by many people

in slightly differing ways. CA, according to Psathas (1995), studies the organization of action in

particular social actions that occur in normal day to day interaction, in discursive practices.

Another way in which CA has been described can be seen in Ten Haves’ (2007) definition, that

1
being, CA can be seen as the study of talk-in-interactions, regardless of their character or setting.

Sidnell (2010), describes CA as an approach belonging within the social sciences in which the

aim is to analyse, describe and comprehend everyday talk as being a fundamental and important

feature of our social life as humans.

According to Psathas (1995), there are three main assumptions in CA. The first of these

assumptions is that “order is a produced orderliness” (Psathas, G. 1995, 2). This means that

certain adjacency pairs and turn taking rules are not given laws that must be followed, rather the

order seen in conversations is one that is created, or produced by the speakers themselves. The

second assumption made by CA is that the order that is created, by the participants, to fit the

certain situation and occasion. For example, in a courthouse, the order of the conversation will be

vastly different to that of the conversation between two friends on a night out. This means that

context of situation plays an important role in determining the order in the conversation that is

decided, unconsciously, on by the participants. The final assumption is that the participants

decide on the order the conversation takes on due to instinct rather than some pre-formulated

convention or theory. As mentioned in the previous examples this assumption refers to the fact

that, when having a conversation, a person does not consciously think “The rules say I must say

X and then they will say Y” and so on, rather, people act on an unconscious and instinctive

knowledge of the order which should occur.

In summary, CA is a branch of linguistics concerned with studying everyday verbal

communication, talk-in-interactions or conversation as a part of human social lives.CA assumes

2
that the order of conversation is one that is produced by the participants, according to the

context, on an unconscious or instinctive level.

There are three steps involved in the process of conducting research in CA. These include the

collection of data in various forms, the transcription of the data including various symbols to

depict linguistic features that may not be obvious from pure written text, and finally the analysis

of the collected data. These steps are described in more detail below.

• How to do research in CA

Comment

– data collection

Conversation analysis has a more descriptive approach to analysing conversation, this means that

it focuses mainly on how language is used on a daily bases by average people (Hunt, 2015).

Therefore, conversation analysis uses naturalistic methods of data collection (David and Sutton,

2004). In conversation analysis the data sources may take various forms such as audio

recordings or videos of conversations (Irranejad and Khonamri, 2014). The data collection is

seen as naturalistic as often the collection of data is done without the researcher being directly

involved, for example using a hidden recording device to record natural and unrehearsed

conversations (Irranejad and Khonamri, 2014). Due to the data sources being collected in a

naturalistic way consent is required from the subjects (Hunt, 2015). This is mainly because the

subjects are at the time of the data collection, unaware of the fact that they are being observed.

3
During data collection there may be possible pitfalls when using recording such as a poor quality

recording device and loud background noises (Sidnell and Stivers, 2012).

Once data has been collected, the next stage is transcription, explained below.

Comment

– transcription

Transcription fixes verbal and nonverbal actions and makes it possible for researchers to

examine them in greater detail than is possible when ‘on the hop when talking to people,

observing their interactions in the field or viewing recordings of those interactions. (Cassell,

2008).

Linguist are not able to accurately recall exact details of what people say (Cassell, 2008). They

enlist a set of audio and often video recordings of the interactions that they are in interested in.

But often recordings have their own problems. It seems much easier to see what is going on than

hear it. Thus in almost all cases, analysts also choose to work from a written record of what is on

tape. It is called a transcript or a transcription of an interaction. Transcriptions typically look like

an act or play.

Transcription conventions are usually based on a system developed by Gail Jefferson. Some

writers choose to deviate from this convention. However whatever convention that the writer

chooses to use it is important that they always provide a listing of their conventions so that

readers easily interpret the transcriptions.

4
The final stage to be completed is the analysis stage. Analysis can be done in various ways. The

methods of analysis are discussed below.

Comment

– analysis

The focus of the wiki is analysis of speech acts in isiXhosa. The two themes to analysis are

categorization and sequential organisation (Hunt, 2015). Categorization involves labelling parts

of a conversation, such as whether it is a question or compliment (Hunt, 2015). Sequential

organisation involves what the action specifically does according to its sequential position (it

makes a difference to the meaning) (Hunt, 2015). An example would be the question naturally

comes before the answer sequentially. In order to analyse speech acts conversations must be

structured, must be adaptable to context and the conversation needs to be natural (Hunt, 2015).

The main contributors to Conversation Analysis are Harvey Sacks, Emanuel A. Schegloff and

Gail Jefferson (Ten Have, 1994). Sacks discovered the rules of conversational sequence are that

the first person to speak has control over the form of address expressed first and this affects the

form of address to which the other responds (rules for conversational sequence) (Ten Have,

2007). Schegloff developed the sequences to how conversations open (Ten Have, 2007).

Jefferson helped develop methods of transcribing recordings such as openings and closings (rules

of adjacency pairs) (Ten Have, 2007).

5
Central to CA is the concept of a speech act and the Speech Act Theory. In the sections that

follow both these concepts will be defined as well as provide an explanation to the approach

taken in Speech Act Theory, a definition of the specific speech act to be examined

(compliments). This section also provides insight into the characteristics of the compliment and

includes extracts from previous research done in other languages.

• Speech Act Theory

Comment

– approach

Most conversational analysts tend to oppose an approach that requires them to be heavily

engaged in theoretical texts. Conversational Analysts favour an approach that is based on

empirical research. However the empirical evidence based approach is quite problematic as its

content tends to leave a lot of readers confused. They will use their habitual expectations,

derived from established social scientific practice, as frames of reference of understanding in this

unusual species of scientific work (Ten Have, 1986). This is the difficulty that arises when one

analyses empirical based texts. Theoretical texts are not the favoured approach because they are

deemed as not taking account of reality.

As mentioned previously, the specific speech act that we will be focusing on is a compliment.

The following section provides a linguistic definition of a compliment as well as a brief

discussion of what the Speech Act Theory and a speech act is.

6
Comment

-Definition of Compliments

According to Searle, Kiefer & Bierwisch (1980), the initial assumption in the theory of speech

acts is that the minimal unit of communication in humans is the performance of particular kinds

of acts, or utterances,, rather than a sentence. These acts can be classified as locutionary acts,

illocutionary acts and perlocutionary acts (Hunt, 2015). According to Hunt (2015), locutionary

acts are those that involve the physical utterance of an act, and the literal meaning of that act.

Illocutionary acts are the expected force attached to the normal form of the utterance or the what

the speaker intended to be understood from the utterance (Hunt, 2015). Finally perlocutionary

acts refer to what the actual effect of the speech act is, whether this effect be the intended effect

or the unintended effect (Hunt, 2015).

The speech act we are interested in is the compliment. A compliment is a speech act which

directly or indirectly gives credit or positive attributes to an individual, typically not the speaker,

who displays some quality or possession which both the speaker and hearer value (Holmes,

1986).

The definition of a compliment is fairly easy to understand. However, the definition does not

shed light on the specific characteristics that make a compliment. These characteristics will be

described below.

Comment

7
-Characteristic elements of Compliments

Compliments, like other linguistic phenomena contain several communication rules which will

determine the response they receive (Mahmoud and Al-Khatibb, 2001). These rules appear to be

related to the ethnography of speaking, this means the appropriate use and response of a

compliment may display the presence of the interactants in a conversation (Mahmoud and

Al-Khatibb, 2001). Compliments appear to be status dependent and are more often paid to an

individual of equal status such as an acquaintance, a casual friend or a colleague (Trosberg,

2010). Compliments paid to intimate individuals such as family members are far less frequent

(Trosberg, 2010). It has been seen that compliments on a person’s appearance and/or

possessions can be given quite freely, whereas compliments on an individual's achievements

and/or abilities appear to be less frequent and limited to individuals of unequal status.

Compliments on a person’s appearance and/or possessions can be given quite freely,

compliments on an individual's. These compliments flow from individuals of higher status to

individuals with lower status and not vise versa as most assume (Trosberg, 2010).

Although the definition and characteristics of a compliment appear to be fairly universal, certain

factors may influence the ways in which they are realised. To illustrate these differences the

following section sets out to provide examples of previous research done in this field and show

how these factors influence results.

Comment

-Previous research on Compliments in other languages

Compliment strategies tend to differ across languages as a result of cultural influence (Golato,

2005). An example would be in Spanish, Spanish men comment on a woman’s beauty and

8
attributes (Moore, 1996). It is part of their culture and is taken well by the women (Moore,

1996). However, in western society it could be seen as sexist (Moore, 1996). Another example is

in Turkish, compliments tend to be predictable and come about in friendly circumstances

(Bayraktaraglu and Sifianou, 2001). Furthermore, the influence of age and marital status affect

what compliments are given, such as the young Turkish men give compliments more on

women’s physical attributes and the older men comment more on the women’s internal attributes

(Bayraktaraglu and Sifianou, 2001). It is necessary to notes that women tend to be complimented

more than men (Bayraktaraglu and Sifianou, 2001). A further example would be in Chinese,

Chinese compliments are far fewer and tend to be restricted (An-Kun, 2008). As a result of the

Chinese having a culture of humility and modesty, the responses to compliments tend to reject

the compliments or deflect it back to the complimenter (An-Kun, 2008). In German, very few

compliments are given (Golato, 2005).

As shown above, compliments vary across cultures and languages. There is however, no formal

research that has been conducted regarding the isiXhosa language. This research is a starting

point for a new area of research. Provided below is the data that was collected with regard to

compliments in isiXhosa.

Comment

Data: audio files

In ‘voice recordings file’ because cannot upload them onto this specific document

9
Data: transcription

Voice recordings from two phones have been transcribed in order to analyse the data. Jefferson’s

transcription technique was used. The recordings were taken around the Kaif and around

Checkers. It is necessary to note that some responses are not in isiXhosa as a result of certain

speakers choosing to respond in English even though they were spoken to in isiXhosa. The

isiXhosa is first directly translated and then translated underneath in the English structure.

Sharon’s phone

1 Lungelo molo (.) bendithi yinhle ihat yakho (.5)

hello I am saying nice hat your

hello I’m saying your hat is nice

2 Woman thank you:::

1 Lungelo molo (.) bendithi unamehlo amahle (.5)

hello I am saying you have eyes nice

hello im saying you have nice eyes

2 Woman huh? (.8)

3 Lungelo unamehlo amahle (.5)

10
 You have eyes nice

You have nice eyes

4 Woman oh (.) enkosi?

oh thank you

oh thank you

1 Lungelo ndiyayithanda lento eniyenzayo (.) maziwe lez' tetue (.5)

I like this thing what all of you are doing, they must fall these

statues

I like what you are doing, these statues must fall.

2 Men yaaaaah? ((handed EFF pamphlet))

1 Lungelo molo sisi (.5)

hello sister

hello sister

2 Woman hello (.5)

11
3 Lungelo bendikuxelela ukuba umuhle nyani (.5)

I am telling you you are beautiful really

I’m telling you that you are really beautiful

4 Woman oh okay (.) enkosi

oh okay thank you

oh okay thank you

1 Lungelo molo ma (.) igama lami nguLungelo (.) bendithi ndiyazithanda

izinwele

zakho (.5)

hello ma’am the name mine is Lungelo I was saying I love hair

yours

hello my name is Lungelo and I like your hair

2 Woman >enkosi enkosi enkosi?<

thank you thank you thank you

thank you thank you thank you

12
1 Lungelo molo (.) bendithi zinhle izihlangu zakho (.5)

hello I am telling you nice they shoes yours

hello I’m telling you that you shoes are nice

2 Man izihlangu? (.5)

my shoes

my shoes

3 Lungelo yah (.5)

4 Man ndiyabulela

I am grateful

I am grateful

1 Lungelo molo (.) ndithanda ijacket yakho (.5)

hello I like the jacket yours

I like your jacket

2 Woman thank you

13
1 Lungelo molo sisi (.) ndithanda iscarf sakho (.5)

hello sister I like the scarf yours

hello sister I like your scarf

2 Woman thank you

1 Lungelo molo ma (.) ndiyawuthanda umbala we ((pointing at her skirt))

(.5)

hello ma’am I like colour of ((pointing at her skirt))

hello ma’am I like the colour of your ((pointing at her skirt))

2 Woman oh (.) enkosi kakhulu bhuti

oh thank you very much brother

OH THANK YOU VERY MUCH BROTHER

1 Lungelo molo timer (.) inhle ileather jacket yakho (.5)

hello timer its nice the leather jacket yours

hello timer your leather jacket is nice

2 Man thank you

14
1 Lungelo molo ma (.) ndithanda ijacket yakho (.5)

hello ma’am I like the jacket yours

hello ma’am I like your jacket

2 Woman >enkosi<

thank you

thank you

1 Lungelo molo timer (.) ndithanda ijacket yakho (.5)

hello timer I like the jacket yours

hello timer I like your jacket

2 Man sure boss (.5)

3 Lungelo niwine ileague mos (.5)

y’all won the league hey

you guys won the league hey

4 Man yeah (.) =hahh hahh hahh

15
Claudia’s phone

1 Lungelo molo ma (.) oh yinhle i headband yakho (.5)

hello ma’am oh it’s nice the headband yours

hello ma’am your headband is nice

2 Woman oh (.) >thank you thank you<

1 Lungelo molo (.) inhle ibhant yakho (.5)

hello it’s nice belt yours

hello you have a nice belt

2 Man huh? (.5)

3 Lungelo yinhle ibhant yakho (.) ulithengephi? (.5)

it’s nice the belt yours, where did you buy it

your belt is nice where did you buy it

4 Man oh (.) andiyazi ndandithengelwe utata

oh I do not know I was bought the father

oh I don’t know because my dad bought it for me

16
1 Lungelo oh ma (.) bendithi yinhle ijacket yakho (.5)

oh ma’am I was saying it’s nice the jacket yours

oh ma’am I was saying you have a nice jacket

2 Woman <oh (.) thank you (.) enkosi>

oh thank you thank you

oh thank you thank you

1 Lungelo heita bra (.) inhle ijacket yakho (.5)

hello bra it’s nice the jacket yours

hello bra you have a nice jacket

2 Man enkosi

thank you

thank you

These transcriptions can be analysed with regard to four different approaches. Searle’s analysis,

Austin's analysis, analysis of the politeness strategies and analysis with regard to culture.

Comment

17
• Analysis

Comment

Searle’s Analysis:

Compliments are usually adjacency pairs, the first turn of initiation is the compliment, and the

second part of the pair is the compliment response (Cruz, 2002). The research project is focused

more on the compliment response acquired. According to Searle, compliments’ propositional

content part requires the speaker’s reaction to a particular situation where the hearer can take

either an active or passive part (Cruz, 2002). Thus, one can argue that compliments are

expressive speech acts. This can be seen throughout the data where all the compliments were

responded to, either in English or in isiXhosa. Therefore one can argue that in isiXhosa hearers

tend to take an active role in compliment responses, supporting Searle’s argument. However,

compliments and their responses can also be classified as representatives (Ruhi, 2006). Searle

argues that a compliment response is a positive reaction and therefore compliment responses can

be viewed as representative/assertive speech acts (Ruhi, 2006). This positive reaction can be seen

throughout the data. Most of the responses included the phrase “enkosi” which translated means

“thank you”. This supported Searle’s argument that typically compliments receive positive

responses. Although Searle’s research was mainly about English compliments, one can see this

trend in African languages such as isiXhosa.

Comment

18
Austin’s Analysis:

For each little conversation recorded, Lungelo complimented the men/women in order to get a

natural isiXhosa response. The compliments given are performatives, which means the

compliments can change how the person now perceives themselves in relation to the compliment

(Austin, 1962). An example would be the woman was called beautiful could have the result of

her having better confidence in her beauty. All the little recordings have felicitous conditions,

which are needed in order to be performatives (Austin, 1962). The felicitous conditions that have

been met are: the people in the recordings are appropriate for the procedures (recordings) and

have the conventional effect, the procedures were carried out correctly and completely, and the

people had the necessary intentions and behaviour (Austin, 1962). The illocutionary act of

Lungelo’s actual complimenting (locutionary act) was to get a natural response in isiXhosa.

Lungelo intention succeeded with only half replying in isiXhosa and the other half replied in

English. In terms of the perlocutionary act, the recipients tended to make the listener grateful

(e.g. fifth recording on Sharon’s phone), want to explain (e.g. second recording on Claudia’s

phone), agree (e.g. third recording on Sharon’s phone), happy (e.g. first recording on Sharon’s

phone) or surprised (e.g. second recording on Sharon’s phone).

Comment

Data analysis in terms of the Politeness Theory:

The politeness theory states that politeness can either be positive or negative. Positive politeness

is expressed by satisfying positive face by indicating the similarities amongst the interactants or

by appreciation of the interlocutor's self-image. (Brown and Levinson, 2000). Negative

19
politeness can also be expressed by saving the interlocutor's face by mitigating face threatening

acts such as giving advice and disapproval or by satisfying negative face by indicating respect

for the addressee’s right not to be imposed on. (Brown and Levinson, 2000). The trend from the

data is that Lungelo is imposing positive politeness as he is constantly appreciating the

interlocutor's self-image. This can be seen in all of the transcription as Lungelo is constantly

giving out compliments. It seems as if Lungelo is satisfying the positive faces of the people that

he is interviewing by pausing for a moment after giving them compliments. In doing this

(pausing for a moment), he gives the interlocutor the opportunity to reply as most of the people

interviewed show gratitude by uttering: “Enkosi” (Thank you). Positive politeness was expressed

by the addressees as they at times uttered seemingly polite responses i.e. "Thank you, thank

you". One would expect a single "Thank you" to be acceptable. Also utterances such as "Enkosi

kakhulu bhuti (Thank you very much brother) indicate characteristics of positive politeness.

Comment

Data analysis in terms of cultural dimensions:

Although we initially hypothesised that the variation in responses would be based on gender

differences, a very interesting and unexpected trend appeared. This trend related to the age of the

hearer. It was found that people who were under the age of about twenty five responded to the

isiXhosa complements in English, occasionally followed by the IsiXhosa response, whereas

those over this age responded in isiXhosa. This is culturally interesting as it shows a younger

generations preference for English, even when being addressed in isiXhosa. It would seem that

20
being educated in an English medium university has a profound effect on the use of languages

other than English.

This assumption was reiterated by the fact that the only member of the under twenty fives who

responded in pure isiXhosa was not a student at the university. This trend was again found

regarding over twenty fives. The only person in this age group who gave an English response

was an ex-student at the university.

Clearly, living and being educated at a mainly English medium institution has an effect on the

use of other languages. This effect is particularly clear in the younger generation. It would appear

that languages other than English are becoming endangered in an English medium academic

setting. Although our data suggests a decrease in the of isiXhosa in this environment, further

studies would have to be conducted to examine if this trend extends to other South African

languages, and whether it only applies to language or extends to culture too.

Comment

Conclusion

The initial hypothesis was that gender was the most important variable in compliment responses

in isiXhosa. After the data collection and analysis it was shown that rather age played a larger

role in compliment response. this may be due to a number of factors such as different politeness

theories regarding age as well as cultural dimensions. It can be seen that usually people older

than 25 typically answered in isiXhosa, whereas persons under the age of 25 would respond in

21
English as they are surrounded with the language more in schools and universities. It can

therefore be argued that age tends to be a defining variable in compliment responses in isiXhosa.

Comment

• References

An-Kun, L. 2008. “On English and Chinese Compliments”. US-China Foreign Language, 6(1),

65.

Brown, P. & S.C. Levinson. 1987. Politeness: Some universals in language usages. Cambridge:

Cambridge University Press.

Cassell, J.l. 2008. Introducing Language in Use. Cambridge: MIT Press Place.

Cruz, M. Compliments: A Relevance-Theoretic Approach. Granada: University of Granada.

David, M. & C.D. Sutton. 2004. “Ethnomethodology”. In David, M. & C.D. Sutton, eds. Social

Research: The Basics, 1-14. Great Britain: Sage Publications.

Farghal, M. & M.A. Al-Khatibb. 2001. “Jordanian college students’ responses to compliments:

A Pilot study”. Journal of Pragmatics, 33(9), 1485-1502.

Hunt, S. 2015. Conversational Analysis Slides.

22
Khonamri, F. & F. Irannejad. 2014. “The Effect of Dynamic Feedback on the Control and

Recognition of Untreatable Errors of Iranian EFL Learners: A Case of Group Dynamic

Assessment”. International Journal of Arts and Commerce, 3(2), 51-69.

Psathas, G. 1995. Conversation Analysis: The Study of Talk-in-Interaction. London: Sage

Publications.

Ruhi, S. 2006. “Politeness in Compliment Responses: A Perspective from Naturally Occurring

Exchanges in Turkish”. Pragmatics, 16(1), 43-101.

Sacks, H. 1992. Lectures on Conversation. Oxford: Blackwell Publishing.

Sidnell, J. & T. Stivers. 2012. The Handbook of Conversation Analysis. United States:

Wiley-Blackwell Publishing.

Sidnell, J. 2010. Conversation Analysis: An Introduction. West Sussex: Blackwell Publishing.

Ten Have, P. 2007. Doing Conversation Analysis: A Practical Guide. London: Sage

Publications.

23