Co-construction of desired behavior through health talk radio program in Ethiopia

Abstract This article presents an analysis of a health talk radio program in Ethiopia. Drawing on a conversation analytic framework, it describes how interaction in the health talk radio is organized structurally and demonstrates how participants in the talk show work collaboratively to the realization of the program’s goal—promoting social behavior change (SBC) relating to a public health concern. It reveals the interaction resources the participants employ from the opening through the closing phases in the organization of the talk show to clearly mark and add to the accomplishment of its institutional feature. The analysis is based on more than 8 h of recorded and transcribed talk data from a public FM radio station. The article contributes to research on media discourse and health education as its analysis has focused on a talk radio show from a linguistically and socio-culturally distinct context (Ethiopia) and in a less studied sub-genre of talk radio (public health).


PUBLIC INTEREST STATEMENT
Talk radios are used, among others, to accomplish participatory health communication.Participatory health communication involves community members as participants in the process of meaning construction and development efforts.It constitutes a communicative effort that enables community members to weave together individual and collective discourses on health and living.Recently, the setting up of talk radio programs that enable citizens to freely talk, share opinions, and experiences relating to public health issues, among others, have become a commonplace.Despite the growing interest in the production of such kind of programs, whether talk radio allows for the representation of a truly two-way equal interaction needs further empirical observation across contexts for scholars' findings in the area are not devoid of variations.Focusing on a health talk radio show from a distinct context, one that uses Amharic as language of interaction, this study has demonstrated how participants organize their interaction and collaboratively accomplish the SBC end of the program.The study has also shown that the structural organization of the talk radio analyzed reveals variations in some features with the practice in other contexts as described in previous studies of talk radio shows.The findings support the micro-analysis of talk as a reliable instrument to make sense of participants' initiatives.This study has implications for the various areas it has touched upon.It has demonstrated that the investigation of communication practice in a health talk radio program is a rewarding project, enriching the fields of studies on media discourse and health communication/education.

Introduction
To study media discourse is to work to make sense of a great deal of what makes up our world.This is because media are found to have a profound influence in the lives of human beings (Fairclough, 1995;Scannell, 1991;Thornborrow, 2015;Tolson, 2006).Of the different forms of mass media, talk radio would appear to have far more impact on the lives of the public at large mainly because they are not only means of getting access to information but also are assumed to create opportunity to the audience to involve in the production work quite regardless of the audience's literacy level (Rubino, 2016) as in print media.Talk radio is one form of radio programs whereby audience are invited to call in to the radio studio and take part in discussion on range of issues with the program host and sometimes with the host along with invited guest or expert (Dori-Hacohen, 2014;Jautz, 2014;Matwick & Matwick, 2018;Thornborrow, 2015).Different forms of talk radio have been the focus of studies in various countries, (Hutchby, 1991(Hutchby, , 2006;;O'keeffe, 2006;Thornborrow, 2015;Tolson, 2006) with heath talk radio receiving little attention.The contribution of health talk radio, a genre of talk radio, in creating platform for public participation on a number of public health matters is widely acknowledged (Nyirenda et al., 2018;Radoff et al., 2013).However, how participation in such a talk show is organized remains less clear.This study analyzes how talks that aim to promote social behavior change (henceforth SBC) are organized and accomplished talking a health talk radio program in Ethiopia as a case.It, thus, contributes to the research of media discourse in two ways.First, its focus on the sub-genre of health talk is different from forms previously analyzed, such as public affairs or political phone-ins (Dori-Hacohen, 2014;Fitzgerald & Housley, 2002;Hutchby, 1991Hutchby, , 1995Hutchby, , 1996Hutchby, , 2006;;Jautz, 2014;Rubino, 2016;Thornborrow, 2001aThornborrow, , 2001b)); second, this article differs in its focus on the practice from a linguistically and socio-culturally distinct context.
A significant number of studies on the discourse of talk radio constitute the practice in the western context.For example, Hutchby (1991) analysed the organization of radio phone-in interactions using data from a London-based phone-in program.In the following years, he extended his scrutiny into how arguments are organized and managed in the context of talk radio (1996) and media talk (2006), with a focus on radio and television.Grounding her analysis primarily on the methodological frameworks of Conversation Analysis and Participation Frameworks, Thornborrow (2001bThornborrow ( , 2002) ) has also studied the organization of talk-in interaction in the context of talk radio programs in the UK.She has demonstrated empirically that the talk radio program is a context for talk where the relationships between participants are such that the interactional status of the people "doing the questions" is not accompanied by a correspondingly powerful institutional status.Fitzgerald and Housley (2002) examined the radio phone-in talks in more detail documenting a variety of categorial and sequential resources, both routine and specialized, used and relied upon by participants when offering their opinions and debating a topic in the context of UK.Using a corpus of radio and television extracts from around the English-speaking world, O'keeffe (2006) emphasized on illustrating how they are managed, how "pseudo-relationships" are established and maintained, and how "others" are created.In one of his analytical chapters, Tolson (2006) paid attention to the analysis of talk radio programs on "sports talk".Through his analysis of this particular form of production, Tolson showed how features of talk radios exhibit variation by comparing his observations with Hutchby's (1996) seminal work.In addition, Dori-Hacohen (2014) describes the differing overall structural organization of public affair or political radio phone-in interactions in Israel and the USA.Also, Bücker (2013), taking the example of a German talk radio show, discusses the forms and functions of "'position offerings'".Moreover, Matwick and Matwick (2018) explore the construction of pseudo-intimacy in the interactional dynamics in an American food call-in radio program.They examine the interactional strategies that the host and callers employ to build solidarity.
There have also been studies dealing with the phenomenon of broadcast advice across a range of different programmes-particularly call-in radio shows that are either dedicated to a specific topic, or that feature more general helplines for emotional or social problem-solving.For example, in their study of advice-giving on talk radio in the USA, DeCapua andDunham (1993, as cited in Thornborrow, 2015, p. 158), observed that "a basic similarity among all radio talk shows is that the interactions taking place are not occurring solely for the benefit of the interlocutors, but are instead functioning as entertainment and sources of information for a wider listening audience".Hutchby (1995) has also dealt with advice-giving in a public context: calls to a radio advice line.He has too demonstrated analytically how experts formulated the advice so as to make it relevant to both the caller and others in the overhearing audience.Thornborrow's (2015) study is also among the notable ones.Making a comparative analysis of advice-giving talk radio programs with relatively different participation frameworks (e.g. from gardening to personal issues), she has revealed that the local interactional context plays an important part in both the development of the advicegiving sequence and the design of advice-giving turns.According to Thornborrow's (2015), there are features that advice giving talk radio programs share as well as differ in.
Overall, several studies made their focus of analysis on talk radio programs, but the increasing tendency to produce such programs in a range of sub-genres across countries hint the need to undertake further examination as analytical results can neither be similar for each sub-genre nor for productions from other contexts with distinct linguistic and socio-cultural background (Thornborrow, 2015;Tolson, 2006).This study, hence, aims to add to the world of knowledge on the discourse of talk radio by analyzing communication practice in the context of health talk radio in Ethiopia.To this end, it raises such research questions as: how are talks about public health concerns structured, and how do the participants co-construct and reveal their orientation to the "SBC" end of the program?

Conversation analysis as theoretical and analytical framework
Conversation Analysis (henceforth CA) is one of the widely used ethnomethodology-based research approaches on language use (Hutchby & Wooffitt, 1998;Hutchby, 1996Hutchby, , 2006;;Thornborrow, 2001aThornborrow, , 2001bThornborrow, , 2002;;Tolson, 2006).Conversation analysis starts from the observation that language as a social process is sequentially organised into recognisable procedures and units (Hutchby & Wooffitt, 1998;Hutchby, 1996;Tolson, 2006).What CA is primarily concerned with are not singular acts carried out through singular utterance, but the patterns and structures of interaction built up in conversation or other verbal exchanges among two or more participants.According to Hutchby and Wooffitt (1998), all aspects of social action and interaction can be examined in terms of conventionalized or institutionalized structural organizations which analysably inform their production.These structural organizations have at least the following key features: a) they operate through sequences of talk; b) they are oriented to by the participants as normative standards; and c) such structures are pervasively present in all interaction (Hutchby & Wooffitt, 1998).The structures of interaction concern primarily the relations of successive utterances and the actions of which these utterances are vehicles.What action a given utterance performs is, however, a matter to be defined interactionally.
CA focuses in on how conversations are structured and organized locally turn by turn, and from this it makes inductive comments about social organization (Hutchby, 2006).It offers the possibility of fine-grained descriptions of how participants orient themselves towards mutual goals and negotiate their way forward in highly specific situations.As Hutchby and Wooffitt (1998) point out, CA aims to uncover the relevances that participants themselves orient to.It tries to do this by taking advantage of the fact that its materials are "interactive", in the sense that what the participants do next implements and displays their understanding of the previous action.What participants say or do next depends crucially on their understanding of what another participant has said previously-and furthermore displays how those prior utterances have been taken (i.e., understood).The fact that participants themselves display their understanding to each other can be used by researchers as a "proof procedure" (Hutchby & Wooffitt, 1998) for their professional analysis.In short, CA is a methodological framework used to analyze the collaborative accomplishment of participants in talk-in-interaction both in mundane conversations as well as in the context of institutional talk.
In general, taking into account the extensive work on broadcast talk done by CA researchers, the emphasis on the ways roles and meanings are made relevant in interaction and the mechanisms of interpretation and the social values attached to discursive practices, CA's framework is utilized in this study.It particularly builds on Hutchby's (1996Hutchby's ( , 2006) ) CA framework.This is because, firstly, this framework has been proven vital for it has been adopted and used widely in various talk radio studies (Dori-Hacohen, 2014;Rubino, 2016;Thornborrow, 2015), and, more importantly, the framework has been found to be of direct application to the current study's analytical goal.

The context
Located in the eastern part of Africa, Ethiopia, is the second populous country in the continent with an estimated population of more than 110 million.According to Mekasha (2005), radio, the earliest broadcast media to appear in the country, came into being in 1941.The broadcasting sector in Ethiopia has broadly three different kinds of broadcasters: 1) publicly funded television and radio services owned by national and regional mass media agencies 2) private sector radio stations; and 3) community broadcasters (Ward, 2011).FM broadcasting was launched for the first time in the country by Ethiopian Broadcasting Corporation's, formerly called Ethiopian Radio and Television Agency's, FM Addis 97.1 in 2000 (Zinaye, 2010).The launch of the FM programs opened the airwaves in that it started interactive radio by establishing live talk radio productions.These productions are believed to engage the audience in discussions over significant issues and problems of the public, including health matters.
In the Ethiopian context, the setting up of talk radio programs which enable citizens to freely talk, share opinions, and experiences relating to health matters has been one of the responses made to curb the expansion of health threats like HIV/AIDS.Despite the existence of a number of radio productions and the growing interest in producing such kind of programs by nearly every radio stations in the country, to the researcher's observation, there has been no specific study that aimed to explore how health talk radio programs are accomplished with a conversation analytic perspective.The absence of empirical observation relating to the interaction practice of health talk radio in the local context, hence, proved the current investigation vital.This is because undertaking a fine-grained analysis relating to how interaction is accomplished in this discursive site provides useful insights into the nature of the communication practice and the relationship between the media and the audience as part of the understanding of contemporary media culture in the country.

The data
The data for the present study were collected by recording verbal interactions on the health talk radio program of FM Addis 97.1.The talk show is aired every Sunday from 6:00 pm to 8:00 pm.It usually is conducted based on a pre-identified health matter and held between the talk hosts and callers to the program.The talk show also brings an expert to the studio to reflect and send SBC messages on points raised (usually on ones that seek experts' voice) during the discussion between the host and callers in the last 30 min or so of the final episode.In the particular talk radio show, discussion on a pre-set topic lasts for four episodes, an episode consisting of nearly 2 h talk.The programs recorded and transcribed for analysis involve participants' (primarily talk hosts and callers) discussion mainly relating to HIV/AIDS.They include talk radio programs broadcast in the months of September and October 2017.In short, the data considered for analysis comprise more than 8 h of talk recordings.The written representation of the talk data is done following Jefforsonian's tradition as reflected in Hutchby and Wooffitt's (1998) transcription protocol.The presentation of the data consists of two lines: the first line for the transcribed Amharic utterance and the second line for the English translation.The analysis is carried out primarily on the basis of the Amharic utterance.The English version does not follow the conventions of English composition writing as the aim was just to provide a corresponding translation of the Amharic utterance.

Structural organization
The structure of participants' contribution in talk radio follows an identifiable sequence (Dori-Hacohen, 2014;Hutchby, 1991;Matwick & Matwick, 2018).The communication practice in the talk radio data analyzed was found to constitute three broader phases, consistent with the practice in Israeli political phone-ins (Dori-Hacohen, 2014).These are the openings, the "talk", and the closings phases.Each phase is observed to have distinctive feature which can be unraveled by focusing on the turn constructions of the participants in the interaction.The characteristic feature of each phase is described as follows.

Openings
In the organization of talk in the Ethiopian health talk radio, the openings could be described as involving general and specific openings.The general opening is one that serves the broader introduction of distinct features of the show including name of the broadcasting station and the program, the topic selected for discussion and the participants that constitute the show, among others.It is used only at the beginning of the talk show.The specific opening entails the beginnings of interaction with each participant in the live show and occurs throughout the show.Once the general opening has been produced by the host, the program proceeds to the specific opening routines, which constitute the interaction practices that initiate the host's encounter with the callers.The practice follows a fairly repetitive pattern.This pattern typically comprises three moves in which each participant takes turns.It includes: the announcements, the pre-greetings (on-line cuing), and the greetings.
5.1.1.1.Announcements.These moves are organized in ways that tend mainly to accomplish dual purposes.These are, firstly, the business of informing or announcing to the audience that hostcaller dialogue is due, and, secondly, bringing the callers into the participatory framework (Thornborrow, 2001a) or orienting the callers that they are about to be positioned as "ratified" participant (Goffman, 1981).The following example illustrates this.

Hello hello
As can be observed, the host uses his first turn to mark alignment with the audience and the caller at a time.In excerpt (1), for example, we can see how the host manages shifting alignment by moving between addressing the audience and the caller.He uses "wuyeyetachen tejemerual" in line 1 to make the audience feel ratified in the participation framework and then says "helo" to mark the ratification of the caller to which the caller replied saying "helo" and thereby hint her understanding of the discourse function of the host's initial turn construction.Goffman notes that a persistent feature of natural talk is participants' shifting "alignment of speaker to hearers" and he calls this phenomenon footing (Goffman, 1981, p. 128).The announcement move in the opening sequence hence can be seen as mainly serving to bring both the audience and caller to the participation framework.Once this is done, the host and the caller would switch to another routine in the organization of openings phase-pre-greetings.

Pre-greetings.
This can be characterized as a move in the opening stage of the health talk radio where the hosts cue that "on-air" interaction between the host and the ratified caller is on.Here, the hosts usually say "helo" to mark that interaction with the ratified caller has begun.The callers take the hosts' "helo" as a signal that the line is open to them and displays their uptake of the floor by responding, in most cases, in the same way as can be seen in excerpt (2) and at times using the device "abet" as in excerpt (3) and excerpt (4) below, having more or less same function with that of "helo".

Yes
While these examples show that hosts use the pre-greeting move "helo" in a separate turn following the announcement move (as in excerpts ( 2) and (3), they also, at times, construct a turn that consists the pre-greetings and announcements moves altogether (as in excerpt (4) below).

H:→ ademach mesmer lay ale helo
A caller is on air hello

Yes yes
Here, the host constructs a turn (see line 1) that enables him address both the audience and the incoming caller as the first part of the utterance "ademach mesmer lay ale" is meant to draw the listening audience's attention and the later one "helo" to bring the caller to the on-air talk as well as hint both that the host-caller interaction is on.As can be seen in line 2, the caller affirms his understanding of the host's "helo" as ratifying him to participate in the on-air interaction through his initial turn construction using "abet abet".Overall, as can be seen in the examples, the pregreetings moves are mainly used to indicate that host-caller interaction is on.In the corpus of the talk radio data, the hosts' pre-greetings move is usually succeeded by greetings.

Greetings.
The greetings moves in the target interaction data appear to be designed within the range of two-to six-turn sequences.Such variation could be accounted for the participants' different tendencies in whether to include callers' identification or not.In the present data, while some greetings comprise turns that accomplish callers' identifications, others do not.
In the examples below, the greetings are constructed in a two-turn sequence.This is a recurrent feature of greetings exchange in the data where identification moves are missing.

Praise to God
As can be seen in the excerpt, the greetings moves are accomplished in excerpt (5) in two-turn sequences (lines 3 and 4) as no attempt of seeking identifications was made.Similar to Schegloff's (2007) description of the mundane phone opening, the host initiated a greeting (line 3), which is reciprocated by the caller (line 4).It seems to be a norm for the hosts to initiate greetings though cases whereby the callers take a lead in the initiation of greetings were observed in the corpus.The callers tend to make this feature evident by awaiting such a discursive action to the hosts through reliance on longer pre-greetings turns as in the example below.This exemplar demonstrates how callers, by postponing the initiation of the greetings soon after the first pre-greetings move, display their orientation to the program's feature.As can be seen in the example, the caller initiated greetings just in line 6 following the absence of the host's greetings initiating move in lines 3 and 5, which appear to be the potential turns for the initiation of greetings by the host.The caller's turn design in line 4, addressing the host by name, seems to hint that the caller was confused following the host's move in line 3 that seemed to breach the recurrent practice in the target talk radio production and wanted to make sure whether or not he was interacting with the host.It was only after he became sure of this issue that the caller took a lead in the initiation of the greetings sequence, observing the host's "abet" in lines 3 and 5 as ratifying him to make a discursive move that keeps the interaction progressing.
In the data analysed, the greetings moves with longer turn sequences (usually four-or moreturn sequences), as has been pointed out, tended to be those that comprise the participants' identification tokens.The excerpt that follows illustrates this feature.

Praise to God
In this example, the participants are seen constructing longer greetings turns mainly because identifications tokens are included.As can be seen in line 4, the caller identifies himself.The host, in turn, relies on the caller's identification in addressing the caller by his name (line 5) and do extended greetings.In the majority of the cases in the corpus examined, identifications tokens, unless revealed by the callers as in this example, appear to be less utilized.While the greetings moves with relatively shorter number of turn sequences may imply that less value is given to the relational aspect of the interaction, those moves with turn sequences extended to include identifications tokens might suggest that due weight is given to the relational end.
Generally, in the talk radio examined, the openings sequence encompasses the general opening component and the specific ones.The general opening component constitutes the hosts' broader introduction set to identify the program, the issue, and the people who are entitled to have contributions on it, whereas the specific opening comprises the interaction practices that publicize the commencement of the hosts' encounter with the callers.In terms of the interactional consequence, the opening phase of the encounter in the target program could be observed as a sequence that constrains the callers from what they are able to accomplish in practices of ordinary form of telephone conversation (Schegloff, 1986) thereby legitimizing what counts as normative in the context of interaction in the talk radio show.The opening sequences of the interaction on the talk radio program, thus, appear be a hybrid of what we expect of conversation in interpersonal communication and what we expect of production for the mass communication.
The accomplishment of the openings in the Ethiopian health talk radio data differs from the practice in USA (Dori-Hacohen, 2014) and UK (Hutchby, 1991(Hutchby, , 1999;;O'keeffe, 2006), but resembles the Israeli political phone-in interaction (Dori-Hacohen, 2014).As much of the previous talk radio studies show, participants open-up the interaction with the hosts addressing the callers in their names and places and then turn on to the topical talk (Dori-Hacohen, 2014;Hutchby, 1991Hutchby, , 1996Hutchby, , 2001;;Thornborrow, 2001aThornborrow, , 2001b)).In contrast, the openings in the Ethiopian health talk radio data involve more elaborate sequence structure.They mainly serve such functions as: announcing the audience that host-caller exchange is due, signalling the line is open for the caller, exchanging greetings.It is after such interaction moves in the openings stage that the participants begin topical discussion.In what follows, we shall see how "talk", which is primarily produced to realize the goal of the program, is sequentially developed.

Talk
Though all of the turns from the opening through the closure add to the accomplishment of the program's main goal, the interaction at this stage tends to be of major significance as interaction on the selected health issue (pre-set topic) is primarily carried out.This study takes as "talk" the speech exchange between the hosts and callers on the target topic (topic selected), which Dori-Hacohen (2014) refers to as "topical talk".The talk between the host and the caller at this phase could better be described as mainly involving two components: initial moves and follow up moves.While the former refers to turn designs used to open-up topical talk, the latter involves turn constructions used to develop the topical talk.Hosts accomplish these, respectively, using talk initiating devices and follow up turn constructing devices.It is to these interaction practices that we now turn to.

Initiating talk.
In the talk radio data examined, the completion of openings phase is marked with the participants' turn constructions meant to initiate topical talk.As has been pointed out, talk initiation is not on the agenda for the callers as the structural arrangement of the call exchanges renders such a role to the hosts.This can also be understood, as in the participants' pregreetings accomplishment, in the callers' tendency either to extend the greetings exchanges until the hosts construct a turn that initiates the topical talk, as in excerpt (8), or when at times the callers manage to initiate topical talk, to design their turn in a way that seeks to accomplish permission, as in excerpt (9).The excerpts that follow are just examples of the callers' orientation to such (institutional) routine.

H: admach mesmer lay seletegegne ahun lela ademach yemenasetenaged yehonal eskahun altemeremerenem helo
As a caller is on line we will be hosting now another caller Haven't we been tested (for HIV) yet hello In excerpt (8), although a chance for the caller was there to take the talk initiatory role in line 6, s/ he prefers to await that role to the host by designing an extended greetings turn, which is responded to by the host along with a talk initiator.In line 7, the host made evident that they have been exchanging greetings in a more than sufficient manner by moving from the greetings token on to the initiation of the talk proper at a turn.Likewise, in excerpt (9), even though the caller manages to be in an initiatory role by responding to the host's pre-greetings turn via greetings initiator in line 2, which in turn is properly responded to by the host in line 3, s/he used the interactional space available to initiate the talk proper in line 4 by designing a turn that instead serves to position her/him as permission seeker.This, in addition to the caller's turn framing in line 4, can be understood in the host's confirmatory remark "yechalal", which means "possible" in line 5.But the host, in his turn in line 5, manages to regain his institutionally imprinted role of talk initiator/questioner by designing his turn in a way that simultaneously enables him to respond to the caller's permission seeking request as well as to raise identification seeking question.While demonstrating the presence of possibilities for the callers to be in a interactionally more powerful position, these two callers' turn construction preferences, respectively, in line 6 and line 4 of the two excerpts hint the participants' subscription to the institutionalized role construction when it comes to taking the talk initiatory role.The institutional affordances of the host category underpin the conventional norms of interactively positioning the caller (Fitzgerald & Housley, 2002) in this way.They vest the host with a greater degree of structural control than any caller.

Developing the talk.
Once the discussion on the chosen topic has been initiated, what follows is developing the talk in a way that enables to realize the program's goal.Central to the organization of the transaction at this stage, thus, appear to be the hosts' talk eliciting moves.While a range of discursive resources are utilized to make the conversation to develop, the ones to be introduced here are only those that have direct bearing to and are consistently practiced for the achievement of the target talk show's goal-i.e., promoting social behaviour change (SBC) concerning the health matter focused.To attain the primary goal of the productions in the turns that follow the initiatory exchanges, the hosts heavily rely on driving the talk towards the elicitation of callers' first-hand experience, observation and/or knowledge.This corresponds with what Hutchby (2001) calls "witnessing".
Witnessing is a term that refers to "a range of actions associated with making claims to personal knowledge, personal experience, direct perceptual access, or categorical membership in respect of an event or topic under discussion" (Hutchby, 2001, p. 485).To elicit the callers' first-hand knowledge and/ or experience, as part of their follow-up turns, in connection with the theme set for discussion, the hosts' broadly mobilize two interaction approaches: designing a framing questions and allowing the callers frame their productions in their own preferred manners.Hosts design their follow-up turns drawing mainly on two interaction resources.One of the widely relied on method in moving the interaction going and thereby elicit data that is primarily related to the callers' theme relevant category is through the use of questions that begin with whwords.The following excerpt makes such a practice visible.

So what of the (blood) testing which aspect of the (blood) testing
As can be seen in this example, the host sought to elicit the caller's reason for getting frightened, as an attempt to develop talk on the subject of discussion, through a question designed using whwords (lines 1 and 3).Such a resource seems to predominate in the corpus of talk radio data examined.It is put to use following callers' stance vis-à-vis the topic of discussion which is usually made available in the participants' talk initiatory turns.
The second approach the target talk radio hosts mobilize in the management of topical talk development is by giving space to the callers to organize their production in their own preferred manner.The excerpt that follows demonstrates how the hosts to the health talk radio show accomplish elicitation of topical talk in such a way.

Ok
In this example, the host sought to make the caller contribute to the program simply by allowing him to share his opinion (line 1) to which the caller responds saying "eh e:shi", literally means "ok" and proceeds on sharing his experience relating to the topic (line 2).The exemplar illustrates how the hosts take advantage of the callers' willingness to reveal their first-hand observations just by designing their turns simply with continuers "ehe" (line 3) and "eshi" (line 5).
The hosts usually stick to the construction of turns that seek to elicit the callers' first-hand data as their follow-up moves on conditions where the callers show the tendency to delay making an explicit address of the phenomenon that specifically relates to their first-hand observation.The callers, through delaying the overt presentation of their first-hand experience until requested by the hosts, may find a solution to two interactional inconveniencies.First, they may avoid the potentially morally dubious activity of volunteering details on "delicate" issues (e.g.telling or narrating about one's sexual practices) (Silverman, 1997).Second, having lacked the awareness or knowledge concerning what is appropriate in such a setting, they may give the hosts an option about whether specification is in order.Overall, the "talk" phase in the production of the health talk radio program appears to be the stage where the main goal of the program is primarily sought to be accomplished.This stage features longer sequences, formulations, and possible shifts in "footing" (Goffman, 1981) and topic as the hosts and the callers attempt to make sense of the talk.Although the hosts predominantly take a lead in initiating and eliciting topic-related talk, the callers are at times observed initiating the talk and presenting their topic-relevant points.This could stem from and be seen as the callers' observance of and orientation to the institutional feature of the talk, as revealed in previous studies (Drew & Heritage, 1992;Thornborrow, 2002).The fact that the hosts pursue the discursive accomplishment of authenticating the callers' talk via a range of knowledge eliciting devices seems to imply that such warranties are expectable on these occasions (Hutchby, 2001;Thornborrow, 2001a).In the examples above, either when the discursive work of grounding the callers' talk was not produced in the earlier phase, or conversely when it went on digression, then the hosts work either to establish those grounds, or frame them.Once the elicitation of talk that adds to the program end is done, the participants bring their interactions to a closure.Let us turn now to see how they accomplish the closing.

Closings
This is the final stage in the organization of talk in the health talk radio examined.As described earlier, the main purpose of the talk radio program is producing talks that promote social behavior change relating to a public health concern.When this purpose is thought to be realized by the participants, the conversation is brought to a closure.The accomplishment of closings in the health talk radio program appears to reveal a complex feature.Despite the complexity of interactional work, the closings of interaction in the target site could, however, be seen as comprising two moves: a pre-closing move and a terminating move.The pre-closing is the one where the complexity appears to be more pronounced.Owing to the complexity, the boundary indicator does not normally appear as the sole pre-closing device.In the data analysed, this feature tended to manifest in the format of either a "closing preface" or a "closing projection" (Clayman, 1988).The closing preface consists of the use of such boundary indicating devices as "melkam" and "selezih", which, respectively, means "good" and "therefore" whereas the closing projection comprises linguistic resources that either tacitly or explicitly suggest that the participants' exchange is due to end.As can be seen, in excerpt ( 13), the host uses the boundary marker "melkam" in line 2 which means "good" as preface to signal the interaction is due to terminate, whereas he relies on the explicit announcement of lastness in the design of his turn "koyetachenen enatenak", which means "let us finish our talk" in line 1 of excerpt ( 14).The linguistic resources that overtly mark initiation of closings, as in excerpt ( 14), seem to take the initiation of advice sequences.They, in most cases in the data analysed, tend to serve as closing projection: hinting that the main section of the talk is about to end and thereby give clue both to the ratified callers and the overhearing audiences that call closure is due.
What appears to be a routine in the organization of closings in the target data is that the hosts advance to the terminating sequence immediately after the enactment of the pre-closings.This is usually accomplished through the deployment of a thanksgiving act addressed at the caller.Such a practice takes two forms.While the first practice is characterized as one that allows the callers to respond to the hosts' acknowledgement tokens, the second can be described as one that does not allow the callers to respond to the hosts' thanking remark.The first terminating method renders the callers with the space to take part in the construction of their response token to the adjacency pair initiated by the host and thereby imply the valuation of the relational element in the encounter in contrast to the second one.

Excerpt (15)
1 H: melkam hasabun tekebelenal seledeweleshelen enamesegenalen Good we have taken your point (we) thank you for calling 2C: → eshi amesegenalehu As the example in excerpt (15) shows, the caller used her turn in line 2 to design a response token to the host's thanking remark through the devices "ameseginalehu", which means "thank you" while in other instances, for example in excerpt (13) above, a response token from the caller is missing.
In general, the final phase in the organization of the health talk radio interaction involves the closings.This phase in the encounter could better be described as a stage where much interaction complexity is witnessed.Despite the complexity, the closings of interaction in the target site in much of the data, however, comprise a pre-closing move and a terminating move.In the closing sequence, the hosts usually design their final turns in a way that expresses their well wishes; invites the callers to stay tuned to listen to others' views; highlights the theme of the program; reminds the name of the station that broadcasts the program; and sometimes announces the in-coming of another (the next) caller, which might be seen as aspects of the programming business.
In this section, effort has been made to demonstrate how communication in the target context is visibly structured.Such a structural observation has primarily been concerned with the turn-taking system that organizes the distribution of turns at talk into sequences of adjacency pairs.The analysis of the turn-taking organisation has allowed inquiring into the actions done in each of the turns at talk.It has also enabled to demonstrate that a wide range of interaction resources are generally used by the participants to mark the different stages in the organization of the program.As the participants in the target institutional setting have been observed as recurrently and pervasively organizing their turn-taking in a way that is somewhat distinctive from ordinary conversation, it can be proposed that they are organizing their communication conduct so as to display and realize its "institutional" (Hutchby, 1991(Hutchby, , 1996(Hutchby, , 1999(Hutchby, , 2006;;O'keeffe, 2006;Thornborrow, 2001aThornborrow, , 2001bThornborrow, , 2002Thornborrow, , 2015) ) character over its course.In the target interaction site, the talk is found to be produced to achieve the program's SBC end.In the following section, how participants orient to and make this feature visible will be emphasized.

Co-construction of the program's SBC end through advice
As has been pointed out, the talk show is primarily produced to foster social behavior change (SBC) communication with the object of building the community's agenda for prevention, and educating audience about specific health behavior.The participants in the talk show collaborate to the achievement of this central goal.The participants' adherence to the realization of this goal can be observed in their interaction moves that gear towards constructing advice.In this section, we will see how advice, a widely relied on resource to the attainment of the target productions' goal, is managed in the interaction site.The conversational environment in which advice is constituted and delivered by the parties involved in the communication practice will be highlighted with the help of illustrative excerpts from the corpus.
Drawing upon the definition proposed by Heritage and Sefi (1992), this study takes as advice those sequences in which one "describes, recommends, or forwards a preferred course of action", or in which one "approves or supports a past course of action or present state of affairs" (1992: 368).This section shall address the forms of advice sequences that recur in the communication practice in the health talk radio show.It also highlights when and how participants draw up on the deployment of this resource, i.e., advice giving and advice seeking, to organize their turns in the effort to realize the SBC end.
In the corpus analysed, advice manifests taking different forms.The variations evident in the forms could be seen as emanating from who initiates it and who it is primarily addressed to (or who are the target recipients).Seen against the first one, the form of advice evident in the data can be classified as ones initiated by the callers and those initiated by the hosts.The following excerpts, respectively, illustrate these features.
Excerpt ( 16) Just let others take care of themselves taking my experience as learning point (xxx) let everybody be responsible to her/his own life These excerpts, respectively, are exemplars of caller-initiated and host-initiated advice sequences.The construction of the advice sequence that commences in line 3 in excerpt 16 occurs following the caller's turn design in line 2 which markedly made evident the desire for advice.In excerpt 17 too, the caller's design of advice turn (line 2) began right after the host's turn (line 1) which was constructed to invite the caller to share his experience.
On the basis of their direct addressees, while some advice sequences appear to be of callercentred, others tend to be of audience-centred.Caller-centred advices are those forms of advice usually designed by the hosts either in response to callers' explicitly stated advice seeking turns as in line 2 in excerpt ( 16) above, where the caller overtly displayed advice seeking, or following some sort of confusions evidenced in caller's turn constructions as in line 2 in excerpt (18) below.Just let others take care of themselves taking mine as learning point (xxx) let everybody be responsible to her/his own life While excerpt 19 shows how a caller designs her turn in a way that makes explicit the audience as primary addressee, excerpt 20 illustrates how the host overtly designs his turn in a manner that overtly addresses the audiences.Having seen the forms of advice sequences in play in the interaction site in focus, a concern worth observing then is about when does in the organization of the communication advice sequences are initiated by the participants.
In the corpus analyzed, advice sequences emerge mainly in two ways.The first is when the advice sequence is initiated by the participants as part of their attempt to develop the conversation (see excerpt (19) above) whereas the second is initiated when the interactants are about to bring the unfolding talk to the close (see excerpt (20) above).In the construction of the advice sequences observed, it also appears to be the norm that when the hosts take the (discourse) role of advice givers, the callers are positioned as advice recipients.Although such a feature predominates in the corpus examined, there are also instances whereby the callers are positioned as advice giver.The example in excerpt 21 illustrates this.One thing to note here, however, is that though a possibility is there for the callers to be positioned in the discourse role of advice givers, it does not necessarily suggest that such positioning in return assigns the role of advice recipient to the hosts.In such cases, the hosts, instead, enact in the role of talk facilitator to benefit the audiences.In the excerpt that follows, in his turn design in lines 2 and 4, the host relies on the utilization of continuers that hint his footing in the role of just talk facilitator.In these turns, the host appears to simulate the role of advice recipient, positioning the audiences as primary recipient of the advice.

Ok
In the construction of the advice sequences discussed thus far, the relational works the participants tend to make evident in their turn designs could be characterized as involving both symmetrical and asymmetrical power relations.While the reciprocity observed in the interactants' footing in the roles of advice initiators and advice givers seem to position them in a symmetrical relation, i.e., both as capable of initiating the advice sequence and offering advice, the absence of reciprocity in the participants' footing in the discourse role of "advice recipient", which tends to be confined merely to the callers, is suggestive of the manifest of power asymmetry in the achievement of the interaction feature.
In the analysis of the corpus, what contributes to the co-construction of asymmetrical relations evidenced appears to be the organizational structure of the talk radio program.As has been pointed out, the hosts seem to take a lead in advising the callers either following the callers' overt advice seeking turn designs or having noticed some sort of confusion made available by them.Put differently, the callers' answers as follow up to the hosts' turns tend to serve to the coconstruction of the asymmetrical relations which specifically is referred to as "epistemic asymmetry" by conversation analysts (Clayman & Heritage, 2010).
In contrast, a possibility for the callers to be in an interactionally more powerful footing is also there.This usually happens when the callers, in their response to the hosts' turns, reveal their good knowledge of the subject in discussion via their utilization of authenticating resources.As has been pointed out earlier, there are different ways through which callers attempt to establish a claim to speak based on a sense of entitlement, and this is made evident as being relevant here.In the data of this study, it is apparent that callers, who draw on their first-hand experience, knowledge, observation, membership as "witnessing device" (Hutchby, 2006), tend to be ratified to offer advice to the audiences.It can, thus, be understood that it is the callers' strategic reliance on the mobilization of witnessing devices that win them candidacy to be positioned in an internationally powerful role.Such a discursive powerful footing in turn bestows them with an interactional role of advice giver.
Overall, in spite of the prevalence of shifts in their footings or positionings in advice sequences, participants in the talk radio show are found to engage in the collaborative achievement of the SBC end.They, to the attainment of this end, involve in the negotiation of discourse roles through their locally constructed turns.The shifts in their footings could, thus, better be observed as emanating primarily from the participants' orientation to the achievement of this goal.In short, in the talk show examined, much effort is vested to create awareness and bring about behavior change relating to unfavorable practices, and thereby promote desired health behavior.

Conclusion
Drawing on conversation analytic framework, this article describes how communication in a health talk radio from a linguistically and socio-culturally distinct context in Ethiopia is organized structurally and how the participants in the particular talk show co-construct a desired health behavior.The data for the study were collected by audio recording and transcribing talks in a health talk radio program broadcast by FM Addis 97.1 in Ethiopia.The analysis was based on more than 8 h of talk data.On the basis of such data, effort has been made to demonstrate how communication in the target context is visibly structured.Such a structural observation has primarily been concerned with the turn-taking system that organizes the distribution of turns at talk into sequences of adjacency pairs.The analysis of the turn-taking organization has allowed inquiring into the actions done in each of the turns at talk.The analysis revealed that the communication practice (between the host and caller) in the talk show constitutes three broader phases: the openings, the "talk", and the closings.Each phase was observed to have distinctive features.The opening encompasses the general opening component and the specific ones.The general opening component constitutes the hosts' broader introduction set to identify the program, the issue, and the people who are entitled to have opinions on it whereas the specific opening comprises the announcements, the pre-greetings (on-line cuing), and the greetings moves.The announcements moves are organized in ways that tend mainly to accomplish dual purposes.These are, firstly, the business of informing or announcing to the audiences that host-caller dialogue is due, and, secondly, bringing the callers into the participatory framework or orienting the callers that they are about to be positioned as ratified participant.The pre-greetings moves can be characterized as moves in the openings stage where the commencement of "on-air" conversation between the host and the ratified caller is cued.The hosts' pregreetings move is then succeeded by greetings.The greetings sequence is put to use by the participants mostly to make exchanges of greetings and sometimes to accomplish identifications.
The "talk" or "topical talk" phase in the production of the talk radio program is the stage where the main goal of the program is primarily sought to be accomplished.This stage features longer sequences, formulations, and possible shifts in footing and topic as hosts and callers attempt to make sense of talk.The hosts tend to accomplish the interaction using talk initiating devices and talk developing turn constructing devices.Central discursive resources drawn upon by the hosts to make the talk to unfold involves: the elicitation of the callers' first-hand experience and/or knowledge, and the initiation of advice sequences.The final phase in the organization of host-caller interaction involves the closings.The accomplishment of closings in the talk radio program in focus is found to reveal a complex feature.Despite the complexity of interactional work, however, the closings of interaction in the target site could be described as consisting of two moves: a pre-closing move and a terminating move.The former is attained in the format of either a closing preface or a closing projection, whereas the latter is accomplished usually through the deployment of a thanksgiving act addressed at the caller.
In addition to the structural organization of interaction in the talk radio, this article has demonstrated how participants to the talk radio show co-construct the SBC end of the program using mainly advice sequences.It has described the forms of advice sequences that recur in the communication practice in the health talk radio show.It has also highlighted when and how participants draw up on the deployment of this resource to organize their turns in the effort to realize the SBC end.The conversational environment in which advice is constituted and delivered by the parties involved in the communication practice has been discussed with the help of illustrative excerpts from the corpus.Despite the variation observed in their form and occurrence, one thing tends to be invariably evident in the construction of the advice sequences.This is the participants' orientation to the utilization of the advice sequences as useful resource to the achievement of the target talk radio program's goal.This is evidenced in the resources they draw upon to depict the collaborative nature of their turn constructions.
To conclude, by following a CA approach, this study has attempted to develop description of communication conduct on an Ethiopian health talk radio show empirically derived from how participants in the interactions make sense of, and respond to, each other.It has also brought to light that the participants conduct their communication with an orientation to the institutional requirement.This is made evident in the overall organization of their interaction and co-construction of their roles and relationships.It has also been observed that the health talk radio show participants organize their communication in a fairly repetitive pattern, enact from multiplicity of roles, align, and disalign with some form of actions, and position each other symmetrically and asymmetrically to the attainment of the program's goal: modeling of a desirable health behavior and stance.

Excerpt
→egziaber yemesgen selam nachehu Praise to God are you fine 3 H:→ dehena nen We are fine 4 C:→ memeher (xxx) ebalalehu I am (called) teacher (xxx) 5 H:→ memeher endemen ameshu selam new How are you teacher how are things 6 C:→ egziaber yemesgen eshi eh selemermera guday lemenegager new Ok eh it is to talk about (HIV) testing 5 H: yechalal enetewawek man enebel Possible let us introduce who is calling 6 C: eshi ene seme yekoyegn kezih kesebeta akababi new emedewulew Ok let my name remain unmentioned I am calling from ena ahun ene temeremeriyalehu semeremer yerase yehone enten yenoregnal (.) e: yehelina weyem yechenkelat psychologically zegejet malet new For example I have been tested to get tested I have my own(preparation) or of mind that means psychological readiness 5 H: eshi

Excerpt
[eneza] negeroch yelum beleh emetaseb kehone sewoch seleferu new emeteferaw malet new belela amarigna If you think that those worrisome things are not your concern put differently does it mean that you feel frightened because that others are afraid 2 C:→ eko awo endeza aynet neger new Yes it is just something like that 3 H:→ lemanegnawum eh (.) zerzer yalu negerochen eh zetegn amsahulet lay magegnet techelaleh yenetsa yeselek mesmer new Any way eh (.) you can get detailed information eh on 952 it is a telephone (help) line free of charge 4C: eshi Ok Although these forms of advice are addressed overtly to the callers, they could also serve the SBC goal of the program by tacitly addressing the audiences.The other form of advice in terms of recipient design, addressee, is audience-centred.It is characterised as audience-centred for it explicitly marks the audiences as its primary recipient.Unlike the caller-centred form of advice sequences, such kind of advice can be initiated either by the callers or by the hosts.The examples that follow illustrate such practice.Excerpt (19) 1 C: → [sew] gen endimeremer new beka yalegn mele'eket ene All I want to suggest is just for [people] to get tested 2 H: sewoch endimeremeru kanchi emiyagegnut yehone neger yenoral belen enasebalen (.) kanchi yemeremera hidet= We think that in order to get tested, people would gain something out of the experience you had (.) from the process in your (blood) testing Excerpt (20) 1 H: → menalebat ahun bezih huneta wuset ersewo beneberubet huneta wuset yeneberu sewoch yenoraluna ahunem yaletemeremeru rasachewun yalaweku ezihem sus wuset yalu sewoch yenoraluna mendenew lenezih sewoch tadiya ersewo emiyasetelalefut Probably as there might be individuals who are in a condition similar to yours, who have not yet been tested, and who are indulged in addiction, what would you pass on to these individuals?2C: beka enen eyaye hulum yeketa (xxx) temekeru new ene emelew

Excerpt
. ..] tenesh yehone mereja lemasgebat vayresu kand sew enten keteyaze sew wedalteyaze yemitelalefebet gize sent new= [. ..] just to add some piece of information about the time it takes the virus to transmit from the infected to the non-infected 2 H: = ehe ehe Ok Ok 3C: lemilew tenesh eske huletshi sement almost eske huletshi zetegn yalew mereja The data up to two thousands eight almost up to two thousands nine 4 H: ehe