The pragmatic functions of the marker sawa in spoken Swahili1

This study examines the pragmatic functions of the marker sawa in spoken Swahili. The data have been obtained from informal conversations made by Swahili speakers in informal social settings. These settings include “vijiwe vya kahawa” (setting of informal conversations created around people drinking coffee) and “vijiwe vya mamantilie” (setting of informal conversations created around women preparing and selling food on the streets). The analysis of the data, performed within the framework of the contextualization theory (Gumperz 1982), shows that, apart from its basic connotation of agreement, the marker sawa conveys other meanings, depending on these very contexts of communication, and therefore acquires also various pragmatic functions. The pragmatic functions identified in this work include: to show that the speaker agrees with what has been said but on a condition (I agree, but...), as a receipt marker, as a tag-positioned-comprehension check, as an answer to the question showing that the speaker has understood what has been said (Yes, I understand), as a continuer, as a negative releasing marker, and as a gap filler. Interestingly, the study shows also that intonation and other paralinguistic features (like gestures) play a role in determining the pragmatic functions of this marker. The article concludes that the pragmatic markers in spoken Swahili are rich in meanings and are used to show speakers’ attitudes and emotions, therefore manifesting a deep and meaningful interconnection between the language and its contextualized experience.


Dar es Salaam University Dar es Salaam University College of Education (DUCE)
The pragmatic functions of the marker sawa in spoken Swahili 1 Abstract This study examines the pragmatic functions of the marker sawa in spoken Swahili. The data have been obtained from informal conversations made by Swahili speakers in informal social settings. These settings include "vijiwe vya kahawa" (setting of informal conversations created around people drinking coffee) and "vijiwe vya mamantilie" (setting of informal conversations created around women preparing and selling food on the streets). The analysis of the data, performed within the framework of the contextualization theory (Gumperz 1982), shows that, apart from its basic connotation of agreement, the marker sawa conveys other meanings, depending on these very contexts of communication, and therefore acquires also various pragmatic functions. The pragmatic functions identified in this work include: to show that the speaker agrees with what has been said but on a condition (I agree, but…), as a receipt marker, as a tag-positioned-comprehension check, as an answer to the question showing that the speaker has understood what has been said (Yes, I understand), as a continuer, as a negative releasing marker, and as a gap filler. Interestingly, the study shows also that intonation and other paralinguistic features (like gestures) play a role in determining the pragmatic functions of this marker. The article concludes that the pragmatic markers in spoken Swahili are rich in meanings and are used to show speakers' attitudes and emotions, therefore manifesting a deep and meaningful interconnection between the language and its contextualized experience.

Introduction
Recently, studies on spoken language in real life contexts increased substantially. As a result, some of the features previously considered "empty", "superfluous" and "redundant", are now regarded as crucial aspects of interpersonal communication (Alami 2016: 250). These aspects include what we call Pragmatic Markers (PMs). PMs are linguistically encoded clues which signal the speaker's potential communicative intentions (Fraser 1996: 168). They provide information for interlocutors how to interpret the relation between the current utterance and the previous utterance, or the other way around. PMs have one core meaning which is procedural rather than conceptual, although their specific meanings can vary depending on the context (thus becoming pragmatic meanings) (Nasir 2017: 15). Since the invention of the PMs by Schiffrin (1987), who worked on discourse markers in English, research efforts have expanded and PMs have been researched across languages, such as English (Beach 1995, Guthrie 1997, German (Barske 2009), Dutch (Hoek 2013), Modern Greek (Archakis 2001), Chinese (Chen & He 2001), Italian (Banzanella 1990), Spanish (Durán & Unamuno 2001), Turkish (Yilmaz 2004), and Indonesian (Nasir 2017), to mention but a few. PMs have been explored within a large number of frameworks reflecting a variety of research interests, methods and goals (Schourup 1999). Despite more general acknowledgement that PMs serve a purpose within utterances, the definition of PMs itself is disputed among researchers. Even the term PMs is not completely uniform. They have been studied by different researchers under different labels, like, for instance, pragmatic markers (Brinton 1996;Fraser 1996), discourse markers (Schiffrin 1987), discourse connectives (Blakemore 1989), discourse operators (Redeker 1991), discourse particles (Abraham 1991), pragmatic particles (Ostman 1983), and pragmatic expressions (Erman 1987). The various terms used to name these features are illustrative of the diversity of functions PMs perform (Alami 2015). In this study I will refer to sawa as a marker; a general term to cover both lexical-semantic and pragmatic functions.
Scholars also argue on how PMs work in language. O'Neal (2013) for example argues that PMs are polysemous, that is to say they manifest different meanings in different contexts. According to Yang et al. (2006) intonation, pitch and duration can differentiate the intended meaning of the speaker. On the other hand Barske (2009) explains that embodied actions such as eye gaze, hand movements or facial expressions are important clues that help to understand the meaning of PMs. Therefore PMs have some unique linguistic characteristics in conveying meanings which need to be investigated taking into consideration not only the spoken language but also and importantly its unique communicative experience. The current study examines the pragmatic functions of the marker sawa in spoken Swahili. The marker sawa is equivalent to okay, yes, well and like in English. In this paper we will make reference on the PM okay 2 because it is more equivalent to sawa in Swahili.
PM okay has been researched more in English (cf. Schegloff & Sacks 1973, Guthrie 1997, Kovarsky 1989, Condon 2001, Schleef 2008, Gaines 2011) to name but a few. Outside the corpus of research on okay in English, only few scholars examine other language contexts, such as American English (Levin & Gray 1983), Canadian French (Heisler 1996), Nigerian English (Adegbija & Bello 2001), Israeli Hebrew (Maschler 2005), German (Barske 2009), Taiwan Mandarin (Wang et al. 2010) and Arabic (Azi 2018). The differences in meaning of the PM okay in two different varieties of English, as noted by Adegbija and Bello (2001), help us to understand that despite the similarity, the meanings of the PM okay are specific to specific language.
While okay is receiving growing consideration in literature, however, there has been little effort to investigate PMs in the Swahili language. Few studies have been done in spoken Swahili. The examples are the works by Habwe (1999) and Susan (1999). Since it was not their aim to investigate the marker sawa, we lack some useful information on how sawa is used in natural conversations. We only get few meanings from Swahili dictionaries like those of BAKIZA (2010), TUKI (2012), Gicharu (2015), BAKITA (2017), and Wamitila (2018). These include its use in showing correspondence, agreement and correctness. Therefore, this paper aims to present a systematic interactional use of this marker in natural settings. The marker sawa has been chosen because the results of a pilot study showed that sawa is frequently used by Swahili speakers in different conversational contexts and hence it performs various pragmatic functions. The study investigates the pragmatic functions of the marker sawa in spoken Swahili. The spoken language was chosen because language is designed for communication, so there are some important cues in conveying meaning that can be found in spoken language and not in written language. Also, spoken language is rich in PMs compared to written language (Stede & Schmitz 2000: 125, Fung & Carter 2007.

Theoretical and methodological background
This research adopted the contextualization theory, as intended by Gumperz in his Discourse Strategies (1982). The theory deals with natural language or natural interaction. Gumperz argues that any utterance can be understood in numerous ways, and that people make decisions about how to interpret a given utterance based on their definition of what is happening at the time of interaction. He insists on how utterances are anchored in contexts, contexts which in turn make the interpretation of these utterances possible. Aspects of a context may be the larger activity participants are engaged in (the speech genre), the small-scale activity (or speech act), the mood (or key) in which this activity is performed, the topic, but also the participants role (the participant constellation, comprising "speaker", "recipient", "by-stander", etc), the social relationship between participants, the relationship between a speaker and the information he/she conveys via language (modality), even the status of focused interaction itself. A context, therefore, is not just given in such an interaction, but is the outcome of participants' joint effort to make it available. It is about what is relevant for the interaction at any given point of time. A context can be revised, i.e. assumptions can be removed or added to it.
According to Gumperz, language is considered as an activity, emphasizing that although we are dealing with structured ordering of message elements that represents the speakers' expectations about what will happen next, it is not a static structure, rather it reflects a dynamic process which develops and changes as the participants interact. Therefore, meaning is also assessed as the activity which is shaped by and shaping context. Thus meaning cannot only be adequately described by the lexical items glosses used, but also with an attentive analysis of the whole communicative experience in which they occur as part of routinized interactive exchange. Gumperz (1982: 131) defines contextualization as a relationship between a speaker, a context (a cognitive construct like a frame, a schema), an utterance and a (non-referential) contextualization cue. Contextualization cue is, then, any feature of linguistic form that contributes to the signaling of contextual presuppositions; contextualization cues include features of language (i.e elements of linguistic structure such as words and syntax) and paralinguistic features such as pitch, tempo, intonation 3 , stress, rhythm, laughter, and nonverbal signals (gestural signs 4 ). I decided to use this theory because the functions of sawa depend much on the context of use. Thus, our research hypothesis is that the functions of the marker sawa depend on various contextualization cues such as what has been said before or after the PM sawa, the intonation used, and the embodied actions accompanying the utterance containing okay.
This research was conducted in Dar es Salaam and Tanga in Tanzania from August 2018 to September 2018. In Dar es Salaam, the research was conducted in Manzese ward found in Ubungo region. In Tanga the research was done at Ngamiani-Kati ward found in Tanga-Mjini Region. Data were obtained from "vijiwe vya mazungumzo" (informal conversation in specific settings). Two types of vijiwe were used to collect data. These were "vijiwe vya kahawa" (setting of informal conversations created around people drinking coffee) and "vijiwe vya mamantilie" (setting of informal conversations created around women preparing and selling food on the streets) 5 . I decided to use the two kinds of vijiwe because it was easy to get data on PMs as people always talk about different issues in their lives while drinking coffee or preparing food 6 . The symbols used in the transcription especially for the PM sawa are explained in the attachment.

The functions of the marker sawa in spoken Swahili
Data obtained from "vijiwe vya kahawa" and "vijiwe vya mamantilie" revealed that the marker sawa has various functions. It can be used as a receipt marker. 4 Gumperz further argues that in the act of talking eyes, face, limbs and torso all emit automatically produced signs which tend to go unnoticed yet nevertheless convey information. The non-verbal signs are language-like in the sense that they are learned through interaction, culturally specific and analyzable in terms of underlying processes. They are coordinated with the verbal signs both at the micro-level of syllables and at the level of clauses and longer discourse segments. They can be used to show speakers' emotional states. They can also be used to frame the interaction and simultaneously reflect and signal transition from one stage of an encounter to another. 5 From each region, 5 vijiwe were for kahawa and 5 vijiwe were for mamantilie. In order to get 20 vijiwe, Snowball sampling technique was used. Researchers got some information about the location of one kijiwe from one person, then participants of one kijiwe were used to get information for other vijiwe. 6 This research is not a gender based study but we decided to use two types of vijiwe (vijiwe vya kahawa which are dominated by men and vijiwe vya mamantilie which are dominated by women) so as to obtain the data that reflect the use of PM by both genders. There were some cases where women participated in vijiwe vya kahawa and men participated in vijiwe vya mamantilie. Therefore, as long as the needed data were available, they were used irrespective of gender.
Here the marker sawa is used to respond to the previous utterance showing that the speaker has accepted the prior instruction, it is well understood and there is no problem with it. The following data from "kijiwe cha kahawa" and "kijiwe cha mamantilie" show that function: From the data, it is evident that the speaker A (in both conversations) gives some instructions to the speaker B (to be given some coffee or to serve the customer) and the speaker B responds by using the marker sawa to signal that they have received and understood what has been said. Here, sawa serves to link between the verbal and non-verbal actions. This means that after using the marker sawa the speaker performs the requested action.
In other contexts, it was observed that sawa can be used to mean I agree with you but on a condition (I agree with you but….. ). Here, the speaker can use the PM sawa to agree with what has been said but also indicating that he/she either needs some more explanations or he/she wants to add something contrary to what has been said. In these two conversations, sawa has been used to mean I agree with you but on a condition. For example in the Data No. 3 speaker A wanted to know where somebody mentioned in the conversation lives. Speaker B seemed to give an unrelated answer -instead of explaining where that person lives, he explains that that person owns a house. Speaker A in line 7 agrees by using the marker sawa: it is true he has a house but he needed more information (where he lives).
In the Data No. 4 participants in the conversation are discussing the voting for election. Speaker A is accusing others of being responsible as they are the ones who voted for. Speaker B agrees that they voted for but asks whether the votes were not stolen; the implied meaning of the speaker is actually an accusation: even though they voted, the votes were stolen. Therefore, in this context the marker sawa is used to convey the meaning I agree with you but on a condition.
In this function the marker sawa is used to signal that the speaker agrees with what has been already said but is not happy about that. From the data above, I noted that the function of the marker sawa as a receipt marker was also observed (see lines 43 and 45). After every instruction speaker B responds with the marker sawa to show that she has heard and understood the instruction. Furthermore, here the marker sawa is also used as a tag-positioned comprehension check. In fact, sawa here is placed at the end of a sentence, therefore it is used to monitor the listeners' comprehension, to check and prove if the other speaker understood what has been said or not (Did you get me? Understood?). This function is always accompanied by rising intonation and is always followed by the response sawa (see line 45) to signal that the speaker has understood what has been said (yes I understand).
In other contexts, sawa was used as a tag question, but with slightly different functions from the one we have just observed before. In this "kijiwe cha mamantilie" speaker B keeps asking sawa? after making some arguments to emphasize what she is saying, to check for understanding and to gain attention from other interlocutors. When the PM sawa is used to perform this function it is also uttered with rising intonation like when it is used as tag-positioned comprehension check. There are some cases where the other interlocutors respond by saying sawa after every tag question and some where interlocutors just respond by nodding as it was the case with these data. When the PM sawa performs this function, it is always uttered repeatedly so as to make emphasis.
A further function of the marker sawa is to signal a sense of equality (or inequality). Mi-nafikiri suruhisho tubadili mfumo wa elimu. Kijana aandaliwe namna ya kujiajili ili akifika mataani maisha yaendelee… 'I think the solution is to change the education system. Youth should be prepared on how to employ themselves so that when they come to the community their life becomes easy.' In this discourse, participants are discussing the problem of unemployment, especially in youth. In line 51, speaker B uses sawa to ask a question whether the person selling groundnuts and the one with no job are equal or not. On the other hand, Speaker A in line 52 uses the sawa as an answer, to show that they are not equal. So it is clear that in this context sawa is used to convey the sense of equality. The marker sawa has also been used in line 56 and 58 as an affirmative releasing marker (as a continuer). Here it is used to indicate approval of the previous utterance and to allow the speaker to continue talking. The continuation and the intonation used are the contextual cues that help the researcher to identify this function. In this context sawa is spoken with some continuation of sound /a/ in the second syllable and after each utterance of this marker speaker B continues talking to show that the use of sawa by speaker C allows him to continue speaking.
The marker sawa can also be used as a negative releasing marker, i.e. to release the other speaker from having to continue speaking, but in this case it indicates dissatisfaction with the previous utterance. In this context speaker A is just wondering how the other speaker has managed to come early to Manzese. Then speaker B, instead of explaining himself why he manages to come early, gives the chance to speaker C (whom they call "uncle") to explain it. Speaker A is not satisfied with that as he asks why should he explain it instead of speaker B. Speaker C doesn't care about that and he continues explaining. After every turn, speaker A uses sawa as a negative releasing marker.
Here speaker A allows the speaker to continue speaking but also showing that he is not satisfied with what speaker C is talking about. Speaker A uses the marker sawa accompanied by some facial expression of annoyance with the conversation. Also, in this context, the PM sawa was said with the falling intonation that shows dissatisfaction.
With respect to the previous function, it was observed that sawa can be used to show dissatisfaction but also conveying an ironic sense of "I will catch you or I will deal with you later". Data from "kijiwe cha kahawa" and "kijiwe cha mamantilie" show this function: Data No. 11 (kijiwe cha mamantilie) In the Data No. 11, speaker A is ordering the other participant to bring back something. Speaker B refuses to do that and speaker A uses the marker sawa followed by the utterance: wewe si unajifanya mjanja ('"You think you are clever. We shall see!"), expressing a sense of dissatisfaction. Considering the relationship (being mother and son) it was expected the son could obey his mother. The act of disobeying brings dissatisfaction to his mother. Therefore the mother uses sawa as a warning she will deal with him later on (in this case the son might receive punishment). In these two different contexts, the marker sawa signals a sense of being in good health. In line 80, speaker A is asking speaker B by using sawa whether speaker B is in good health or whether there is any problem. In line 51, speaker B answer the question by using sawa to reply that she is in good health and there is no problem. Here sawa has been used as a question, uttered with rising intonation, and as an answer to a question, with normal intonation. In these examples sawa is used accompanied with verbs indicating state of being (like upo, nipo or utakuwa [sawa]) and others to mark the sense of being in good health. The same is also noted in line 85 where speaker B says utakuwa sawa to mean she will get better soon or she will be in a good health.
Related to the sense of being in good health, the marker sawa was also used as a predicative adjective or adverb to mean all is correct, all is right, is satisfactory, is good, is well and that everything is in order. In these four different conversations, sawa is used to signify that things are in good order and satisfactory. In cases where the speaker is asked to report about the recent events, speaker B responds using the sawa to signify that he/ she is trying to put things in order so that they can provide the report. Also in the Data No. 18 speaker B shows that he is trying to schedule the timetable so that everything could be okay for him to attend the meeting. Therefore, apart from showing the condition of being in good health, the marker sawa is used to signal that all is correct, all is right, is satisfactory, is good, is well and that everything is in order.
In other contexts the marker sawa performs slightly different functions from the two above-mentioned. In fact, it can be used to signal that something was not well understood but now is well understood, was not in good order or in proper manner but now it is in good order or proper manner. In these three different conversations, sawa is used to signal that something was not well understood or was understood differently but now is well understood (see line 83). Furthermore, it is used to show that something was not in good order or in proper manner but now it is in good order or in proper manner (see lines 99 and 102). When the marker sawa is used to convey this meaning, it is always accompanied by the word hapo (hapo sawa).
The marker sawa can be used to mark whether something is correct or not or whether what is said is correct or not. In this context of use, participants are discussing about the government evicting the citizens from their land to construct new pipelines. In line 111, speaker A is asking, using the marker sawa, whether for the other participants that action is correct or not. Speaker B answers the question using sawa to convey the meaning that that action is not correct (see line 112). Therefore in this context the marker sawa performs the function of conveying the meaning of whether something is correct or not.
Observing the following conversation made from "kijiwe cha kahawa", the marker sawa appears to convey the meaning of "acceptability", to mean it is acceptable to me or the decision is acceptable to me. Speaker A (see line 116) uses sawa as a question in order to get information if it is acceptable to meet on Wednesday. Speakers B and C (see lines 117 and 118) answer the question by using the PM sawa to mean that it is okay to them (the day is acceptable to them). In the same conversation, the marker sawa marks the closure of the conversation. After all members agree to meet on Wednesday, the speaker A (see line 119 ) says okay to mark the beginning of the closure of discussion (Basi tukutane Jumatano). Afterwards, speakers B and C use sawa to agree but also to conclude the conversation.
In a different contexts the marker sawa can be used as a gap filler: Speaker A wants to hear speaker B's stance about the target issue. Speaker B starts speaking by using the marker sawa which, in this context, does not signal agreement or is not a receipt, rather it is used as a gap filler. In this context it is uttered with the prolonged sound /a/ in the first syllable. Here the PM sawa provides the speakers the opportunity to better organize his thoughts, in fact, after using the PM sawa, speaker B continues to give his suggestions. This is the kind of gap-filling role the marker sawa assumes here.
It was also noted that the marker sawa can be used as a quotative marker. The following data from "kijiwe cha kahawa" suggest this function: The speaker uses sawa to quote what one is planning to say sometimes in the future. Besides, in other contexts, it was observed that sawa can also be used to quote words that were spoken by somebody else in the past. Therefore, from this context it is evident that the PM sawa is used as a quotative marker.
In other contexts, the PM sawa is used to convey the sense of "I don't care", like in the following conversation: By considering the contextualization cues (what has been said before and after the PM sawa), the way the PM was uttered (with a prolonged sound /a/ in the second syllable) and the gestures (like the hand movements) the marker sawa acquires the function of signaling that the speaker doesn't care (whatever will be done to him is correct).
In the following conversation too, the specific attention to the context (gestures, contextualization cues, and voice intonation) discloses how the PM sawa can be used as an answer carrying the meaning of "no problem". In this conversation participants are discussing whether cheating is a correct behavior. Speaker A uses the marker sawa to ask the question and speakers B and C use it as an answer conveying the meaning that it is not correct.

Conclusion
The preceding analysis examined the functions of the marker sawa in spoken Swahili. The findings reveal that sawa performs various functions depending on the context of use and specific conversational experiences. In this research I have underlined some interesting functions of the marker sawa: as a receipt marker, tag-positioned-comprehension check, tag-question used to check for understanding, continuer, negative releasing marker, gap filler, and quotative marker. It can be also used to make emphasis and gain attention, to signal whether things are equal or not equal, to signal acceptability (whether something is acceptable or not), to mark the transition from the state of not understanding to the state of understanding, to signal that everything is in good order or someone is in good health, to signal dissatisfaction, to mark ironic sense of "I will deal with you later", to signal that the speaker doesn't care ("I don't care"), to signal that the speaker doesn't want to hear anymore what the other speaker is talking about, to signal the sense of "no problem" and to signal the topic change. The results also show that, apart from relying on contextualization cues indicating what has been said before and after the PM sawa, also some paralinguistic features like gestures and intonation are among the contextualization cues that help to determine the functions of this marker. It would be certainly interesting if further research could focus on the marker sawa as used in other contexts in order to explore its more possible functions. It could also be worth investigating if there are any regional differences in the use of PM sawa and any possible codeswitching to replace it by English "OK" in some social contexts.