A Discourse Study of Insurance Sales Agent-Client Interactions in China ’ s Rural Areas — From the Perspective of Topic Management

Based on 10 authentic audio-recorded data, the study aims to explore the dynamics of the insurance sales agent-client interactions in transformational China’s rural areas from the perspective of topic management. The analysis of the interactions aims to understand the interactants’ roles in midst of social change. The relationship between sales agents and clients is ambiguous and conflicting. By identifying three types of talk, namely, friendship talk, institutional talk and task-oriented talk, we find both parties are consciously or unconsciously initiating, maintaining and managing different topics to their respective advantages strategically. Through the analysis, we hope to depict the transforming nature of the modern Chinese rural society, where old and traditional value system has been, for a large part, demolished, while new value system is yet to be established.


Introduction
This paper analyzes the dialogues between insurance sales agents and clients in China's rural areas from the perspective of topic management.It analyzes how both agents and clients consciously or unconsciously shift among topics to their own respective interactive advantages through assuming topic control.More specifically, three different types of talk namely friendship talk, institutional talk and task-oriented talk are employed in the analysis of the dialogue, representing the mediation of different frames (Goffman, 1967) and the two different discourse systems (Scollon & Scollon, 2001) at large.Scollon & Scollon (2001) propose a framework of Discourse System (Confucianism VS Utilitarianism), which "has the capacity to provide a more complete and integrated description of the phenomena that contribute to the formation of social structures in response to the specific interactional communicative and practical needs of individual" (Duszak, 2002).In this discourse system, ideology, socialization, face system and forms of discourse are interrelated to one another.And it's a system of thinking, social practice, and communication.
Insurance is a core institution of a society increasingly organized around managing risk.As Beck (1992) argues, risk society is also an individual society.The utilization of insurance sales agents' identity to achieve economic goals interact with key contemporary social tendencies: the responsibilization of the individual consumer, the erosion of the social safety net, fragmentation, individualism and the attenuation of family ties, the growth of a "flexible" labor force, and the downloading of regulatory responsibility from the government.While in Chinese business culture, Guanxi (Interpersonal Relationship) is one of the major dynamics of the Chinese society and has become a key business determinant of a firm's performance.More importantly, Guanxi is an important cultural feature that affects the Chinese business communication as argued by many anthropologists (Yang, 1994), historian (Yang, 1994), and sociolinguists (Pan, 2010).Most of Guanxi is built on the basis of kinship, familiarity, trust, etc. in the Chinese "society of the familiars" (Zhai, 2007(Zhai, , 2009) ) with the feature of "differentiated mode of association" (Fei, 2007).Numerous scholars study Guanxi from different perspectives with different methodologies (Kong, 2003;Chen, 2004;Michailova & Worm, 2003).
Until recently, the hidden world of insurance has received surprisingly little attention from social scientists (Strange, 1996, p. 122;Baker & Simon, 2013).Most academic researches on insurance have been done by economists and researchers in business and finance.Nonetheless, insurance studies have recently emerged as a blossoming field for other researchers, especially in sociology, socio-legal studies and criminology (Baker & Simon, 2013;Ericson & Doyle, 2003;Ericson et al., 2003;Ericson & Doyle, 2004a;2004b;2004c).In China, research on insurance is mostly done from economic, business and financial perspective, their primary interest is to investigate the current situations in China's insurance industry and offer certain suggestions to improve the efficiency and profitability of the insurance institutions from political, economic and marketing perspectives (Sheng, 2009;Zhu, 2009;Hu & Wang, 2009;Shu & Li, 2006;Yuan, 2008;Guo, 2009;Li, 2008;Ding, 2008;Cao, 2009;Xu, 2009).Similar interests in this area also emerge (Hu & Wang, 2009) from a sociological perspective, the studies mostly put insurance into social, economic and cultural contexts, then study the inter-relationship among them to consider the interaction between insurance and many other social, economic and cultural factors.However, such studies are still rather on the macro aspect and lack of empirical evidence, which gives rise to this study to investigate the more micro-aspect of insurance sales interactions in transformational rural China.Moreover, there are several problems in their research.To begin with, their research seems rather general and adopts a top-down approach.Secondly, most of their research is based on academic reasoning and previous literature, with very few offering empirical evidence, which might create a problem in its convincing effect.Thirdly, the stance they take is on the behalf of the organizations; they neglect the other side, that is, the agents in action and their clients.They take them as strictly passive recipients of the actions, which might not be exactly the case in reality.On the other hand, another type of research tends to be focusing on the extreme micro aspect of discourse.As Sacks et al. (1974), Heritage (1984), Jefferson (2004) and many other researchers (Duranti, 1998;Heritage, 1984;Hutchby & Wooffit, 1998;Zimmerman, 1988), many analysts have studied mundane social action and have achieved desired results in such areas of discourse organization as turn organization (Boden, 1983;Roger, 1989;Schegloff, 2000aSchegloff, , 2001)), Repair organization (Schegloff, 2000b).Other scholars study the structure of social interaction (Eggins, 1997;Yang, 1994;Lin, 2004), intra-cultural communication in various languages (Hayashi, 2003) and intercultural communication studies (Cheng, 2003;Scollon & Scollon, 2001;Tanaka, 1999).Regarding Chinese data used in CA approach, most of it focuses on Cantonese (Kong, 2001(Kong, , 2002(Kong, , 2003;;Luke, 1990).However, their findings on Chinese communication may not satisfactorily explain the interaction pattern and linguistic features specific to the Mainland Chinese due to the data they collect and use for their analysis.
Therefore, there is an urgent and imperative need which calls for a concern in studying insurance that is based on sociological theories or concepts, yet can be analyzed into greater details and offer empirical evidence.The dialogues between insurance sales agents and clients offer an exciting arena for analysis, and this paper is such an endeavor intended to fill the gap and tries to strike a balance between macro-and micro-studies for the purpose of revealing the hidden world of insurance sales discourse in transformational China's rural areas.More specifically, how is topic managed in the dialogues to the respective advantages of the interactants.

Methods and Procedures
10 authentic naturally occurring conversations between insurance sales agents and their prospective clients are collected and transcribed for analysis.And the chief methodology is discourse analysis, focusing on conversational analysis at the micro-analytic level.In sociolinguistics or business English studies (Cameron et al., 1988;Planken, 2005) and cross-cultural/intercultural pragmatic studies (Bargiela et al., 2009), data are generally collected in a controlled and selective way and the focus is always on one or several types of speech acts.As Huo (2004, p. 69) noted, it is preferred that the type of data for such research be collected in a natural, uncontrolled way in that we did not confine ourselves to a certain type of data but include all instances of dynamics in our corpus.The data collected are recordings of conversations between insurance sales agents and their prospective clients at their homes, tea houses, meeting rooms or other places where the conversations take place during 2009 to 2013 (about 80 hours, 50 conversations in total).The selected data for the analysis are 10 conversations between the 5 agents and their clients.The rest of the conversations, interviews and retrospections are used for coding the recurring pattern and triangulation purpose.By analyzing the different types of data, the validity could be better guaranteed (Marshall & Rossman, 1999).The data have been transcribed by the author himself according to the conventions of Conversation Analysis for discourse analysis purposes.Some of them involve two agents (the more experienced agent accompanies the less experienced one to visit their clients); some of them involve more than one client (their families or relatives may be present).In general, some of them involve two participants; some others may involve several participants since the conversations are rather informal in nature and there could be the prospects' relatives or friends present during the conversations.Their names, and some specific names of their products and their prices were replaced by pseudonyms or "X"s.
The procedure of analysis mainly starts from the transcriptions, followed by transcription study to identify some recurrent phenomena across the transcribed conversations.The procedures include: Listening to the data-classifying the data-selecting the data-transcribing the selected data-studying the data for recurrent phenomena-in-depth discourse analysis of the excerpts by focusing on the phenomena-making cultural and linguistic interpretations of the analyzed phenomena-studying the interpersonal preferences-generating implications from the analysis.
The analysis is mainly qualitative in nature and the aim is to discover and reconstruct patterns in insurance sales discourse.The data are examined for emerging patterns and language use that reflect the most striking features of the Chinese insurance sales agents in terms of discourse conventions, cultural patterns and values and linguistic manifestations in the sales process.The qualitative findings help to provide background information on the context and subject, acting as an aid for scale construction (Punch, 1998).The frequency rate of each feature/phenomenon was applied to provide the basis for qualitative study, with an aim to generate and frame the preferred patterns and linguistic choice in insurance sales discourse.In the analysis, the author studied the organizational structure and corresponding linguistic features of the data, then the author examined the data by focusing on identified phenomena and counting the frequency rate of each feature/phenomenon to generate the global structure of insurance sales interactions and generic socio-pragmatic preferences of the participants.

Results and Discussion
We found both insurance sales agent and prospective client are attempting to assume topic control to fulfill their respective objectives in insurance sales dialogues.The following will present the dynamics of insurance sales in contemporary transformational rural China.
It has been known that in our daily life communication, we either consciously or unconsciously manage topic in order to fulfill different purposes; while in insurance sales dialogues, such a feature is even more salient: in general, on the part of the agent, s/he tries to keep the conversation in track and tries every means possible to conclude the deal; while on the part of the prospective client, s/he is also managing to assume topic control to his/her advantage.The shifting or maintaining of the different topics reflects the mediation of different frames, thus the mediation of different discourse systems at large.

The Mediation of Different Frames
About half a century after Malinowski (1923), Laver (1974Laver ( , 1976) ) provides detailed analyses of the sequential features and pro-social functions of phatic talk (Holmes, 2003;Coupland, 2000).Interaction, in Laver's view, can be generally divided into three phases sequentially: the opening, the medial, and the closing.And marginal phases of interaction, phatic talk, as Laver observes, is restricted only to the openings and closings of encounters, Holmes (2003), however, has shown that small talk may also appear in the middle of an interaction, especially when speakers intentionally get away from the ongoing business.The characteristics of the distribution of small talk are generalized as "elasticity", "flexibility", and "adaptability" (p.7-48), with the medial phase reserved for "big talk" where business is done.Thus small talk serves a transitional or "time-filler" function between different activities.Phatic talk in both the opening and the closing phase serve as a transition, though in different directions.In the opening phase, phatic talk serves to establish relationship and achieves a transition from non-interaction to interaction.In the closing phase, however, phatic talk is adopted once again to make another transition from interaction to non-interaction.Though phatic talk may happen sequentially at the opening and endings of conversations, its importance for a successful interaction can be felt throughout the whole process.This function of lubricating and smoothening everyday conversation has been widely acknowledged.Thus, small talk fills a time gap with acceptable and valuable relationship-maintaining social interaction and avoids problematic and embarrassing silent moments before the "main activity" can be started or resumed.
Laver also discusses three pragmatic functions of phatic talk in the opening phase.In the initial phase of a conversation, phatic talk is expected to fulfill a "propitiatory function of defusing the potential hostility of silence in situations where speech is conventionally anticipated (Laver, 1975, p. 220)".Here, phatic talk helps to establish a friendly atmosphere among participants, who may be acquaintances or strangers, and to make further interaction possible.Phatic talk also has an "explanatory function", "allowing the participants to feel their way towards the working consensus of their interaction" (p.221).This uncertainty-reducing function facilitates the revelation of each other's personal information, such as social status, personality, and mood.This function is also closely related to the choice of topics in the following interaction.Another function of phatic talk is the "initiatory function".Phatic talk allows the speakers to cooperate in getting the interaction under its way.By talking about conventional and uncontroversial topics, the speakers contribute to the initiation of an interaction, avoiding abruptness of speaking.In closing sequences, by picking up phatic communication once again, one can reduce the possible sense of rejection and consolidate the relationship, thus end the talk in a friendly way.
Similar to Malinlowski, Laver also recognizes the relational value of phatic talk.Apart from being a simple semantic phenomenon, it functions to create ties of union through subtle and complicated means rather than a mere exchange of words.As a relational phenomenon, phatic talk varies in its topics and ways of conducting interaction with people of different social relationships.Depending on their social status in relation to the other party, speakers may choose either self-oriented or other-oriented expressions to initiate talk.When an inferior opens an encounter, s/he is likely to open with self-oriented remarks.But if a superior initiates the encounter, s/he may choose other-oriented topics or questions about the inferior's personal matters.If the speakers are equals, a neutral topic is a probable choice.This speculative predictive mechanism specifies how speakers use phatic talk to demonstrate solidarity/intimacy and status relationships among them (Coupland et al., 1992, p. 212).This mechanism introduced by Laver is very similar to the politeness strategies proposed by Brown & Levinson (1987).
Having introduced phatic talk and business talk by previous scholars, it is imperative to examine their application in the analysis of the interactions between the insurance sales agents and their (prospective) clients.In the following part, the three dominant types of talk in the interactions will be identified for their content, organization, and pragmatic functions.There are three main sections: (1) Data of Analysis; (2) Three Types of Talk in Insurance Sales Interactions; and (3) Summary.
In rural insurance sales settings, most of the prospective clients are either friends, relatives or at least familiars introduced through a mutually acknowledged third party, the relationship between the agents and their clients are complicated and dynamic.On one hand, prospective clients are insurance agents' friends, it is expected that as friends, they should be supportive, honest, loyal, faithful and considerate; on the other hand, how much an insurance agent earns depends on how much the prospective client buys, according to a particular commission formula, thus it is profit-driven and practical goal-oriented.Therefore, the insurance sales agent-client relationship is a dynamic, ambiguous and conflicting one.Nevertheless, a harmonious relationship between agent and client is important for both of them.Agents need the clients' cooperation and decision to make the purchase, clients need agents' advice, both practical and spiritual, in order to facilitate their lives in the local community, as for most of the time, agents possess much greater resources in the local settings.These dilemmas and the potential conflicts motivate insurance sales agents to invoke various identity categories in order to achieve their interactional goals.
The analysis is based on a detailed study of an approximately 20-minute conversation between an insurance sales agent and a prospective client, and three other interactions between agents and clients.Only female agents are represented to avoid gender differences (actually most of the local agents are female, probably due to the fact that females could be more suitable for insurance sales); their ages range from 30 to 50.All the agents are experienced insurance sales agents with at least five years' experience.The clients are mostly new to insurance; few of them have prior knowledge of insurance.This phenomenon has some important implications, on one hand, since most of the prospective clients have little prior knowledge about insurance, it will take the agents great efforts to introduce and explain the concept, rationale, function and necessity of insurance; while on the other hand, it leaves much room for the agents to tap into the unexplored huge potential market, which also makes the prospective clients potential victims of certain manipulation or misbehaviors if the agents build their sales solely on the blind trust that the clients have in them.It could be safe to say, whether or not there exists serious manipulation largely relies on two factors: One, the sense of conscience and responsibility the agents have; and the other, the willingness and need of the clients to purchase the insurance in question.All the interactions occur in the working environment of insurance sales-in the clients' living room, the courtyard, store or corridors.All participants are aware that they are being recorded.

Three Types of Talk
In their interactions, three main types of talk-friendship talk, institutional talk, and task-oriented talk-have been identified, mainly by their content and their linguistic features.Among the three, the task-oriented talk is the most complex due to the fact that it contains the features of both friendship and institutional talk.The following explores the three categories of talk in insurance sales discourse and the detailed examination of how identities are inter-actionally managed within the three types of talk.

Friendship Talk
The friendship talk (the talk among friends), is identified by both its content and its organizational features.Talk plays a significant role in the indexing, managing, and maintenance of friendship, as Simmel (1961) argues, talk is "an end in itself" in social gatherings where our human instinct for "sociability" is achieved.Because of the special nature of friendship talk, its content is quite varied.It could be anything about the agent or client, from their full-time jobs to trivial matters, such as buying things and seeing a doctor.Most friendship talk is conducted for its own sake, without any explicit purpose of the participants, nor are there any of the practical outcomes that characterize most institutional outcomes.Nonetheless, it is not suggested that friendship talk is not goal-oriented, indeed, friendship talk can be highly pragmatic in nature.This is especially true of the friendship talk in institutional settings, where the socio-emotional aspects of human behavior are brought to play for various purposes.While most institutional talk is goal-oriented in the sense that the participants' behavior is highly contingent upon their relevant identities in an institution.On the other hand, friendship talk does not have the constraints of institutional talk.
In the interactions of insurance sales, friendship talk is fairly frequent and plays a significant role.Before turning to the functions of friendship talk in insurance sales, we need to focus on some excerpts of friendship talk in the data corpus.The following is an opening of a conversation between an experienced insurance sales agent and his prospective agent who was identified to have the potential need for insurance coverage.
A: Agent, female, an experienced insurance sales agent with over five years' experience B: Client, male, a potential customer of insurance policies They are familiar with each other but haven't seen each other for a while.The above interaction contains the features of friendship talk identified by Tannen (2005) and Lakoff (1980)-such as a preference for personal topics, turns of equal lengths, and mutual initiation of topics.Yet, some striking differences are found.First of all, as both Tannen and Lakoff point out, conversations among friends are reciprocal and cooperative.However, the greetings of the interactants are not a preferred sequence of question-answer.The agent initiates the conversation with "Long time since I saw you last time.You must have got rich?You have such a large and beautiful house."The preferred sequence is not only the "answer" to this question (Long time no see, it has been quite a while), but some modesty expected concerning the compliment to his house; instead, the client asks the same question "where have you been?" as a counter-question, and after that, initiates a new topic.In fact, this "counter-question" sequence is quite common in the data, not only at the opening of the conversations but also in the middles and closings of them.What is operating here is the two Maxims of Friendship Engagement: the maxim of intimacy and control (Rogers & Millar, 1988).Nevertheless, what they are doing is far from being "unfriendly" to each other.The maxim of intimacy is the maxim of dependence, whereas the maxim of control is the maxim of independence (ibid).Being friends involves not only showing care about each other, but also respecting personal territories when necessary.As a result, when the agent first initiates the question in turn 1, the client does not answer it but instead asks the same question.In fact, when the client asks the question, he doesn't expect an answer, because he has initiated a new topic by inviting the agent into his house.What they are doing here is negotiating the maxim of Intimacy and maxim of Control, in other words, the dependence and interdependence between friends.
Besides the negotiation of dependence and interdependence in the above interaction, the disagreement of the agent with the client's claim that he is rich, in turn 7, is also an implicit indication of his current situation and very covert intention of hoping the client makes the decision of making the purchase of the insurance policies under discussion.This ambivalence and contradictions, very common and frequently found in the data, will be further explored later in greater details.
Another striking difference between the friendship talk identified and the findings of Tannen and Lakoff is the rate of speech.As a prelude to the task-oriented talk later on, the friendship talk is slower in speech compared with the task-oriented talk that follows.The same is found in most of the other conversations.The contrast in the rate of speech distinguishes friendship talk and other types of talk.Rate of speech is not a static feature of talk, it is one of the dynamic strategies which can be used to negotiate social identities during interactions.
Although friendship talk is one of the distinctive features of insurance sales interactions, it is also quite frequently seen in other institutional settings as well.For example, business deals are also intermingled with business and conversational talk (Scollon, 1998); moreover, in doctor-patient interactions, socio-emotional talk is also widely employed at the beginning of the consultations as a lead-in to the professional talk in question, in order to address the increasingly holistic concern of patients in the health-care setting (Coupland et al., 1992); it has also been found that in classroom settings, teachers may relax the institutional constraints from time to time in order to build rapport with their students.Nonetheless, within these institutional settings, the goal of friendship talk is only auxiliary to primary institutional goal in question.To be more specific, in doctor-patient interactions, there are restrictions constraining participants' topic contributions.For example, after a doctor greets the patient with a simple "Hey" ("How are you?"),although a patient can initiate any topic in his turn, yet his response is still largely constrained by the relevance to the institutional frame.That is why most patients' responses are themselves related to the medical concerns, which can be used by the doctor as a lead-in to the medical consultation (Coupland, 1992).
It is the friendship talk in insurance sales which makes use of the socio-emotional aspect of institutional talk to the fullest extent.The function of friendship talk is evident now.As insurance sales intermingles friendship and business, the insurance sales agents' main strategy is to exploit these seemingly contradictory ideologies.One of the important tasks of insurance sales is to exploit the intersectional space of friendship and business to the greatest possible extent and at the same time to reconcile the contradictions in this mediated interface.As talk is the most effective means of transferring ideas and concepts, friendship talk certainly plays its role in socializing between agents and prospective clients.Through friendship talk, prospective clients can experience the friendly atmosphere of the interactions, in contrast with the contractual business goals in the deal itself.As experienced agents, they are well socialized about how to intermingle the personal and business domains, and manage potential or imminent conflicts from this mediation, as an agent confesses in his interview: "As matter of fact, in my view, the essence of the business is to make friends, real friends with our clients, as long as they can trust you, it is almost for sure he will buy from you given there is a need, so the key is to make friends and gain trust."(Qishiwomen zuo baoxian jiang daodi jiushi zuoren, ren zuo haole, qita dongxi du haoban, yiban biru jiang kehu you zhege xuqiu, ta yaoshi neng xinren ni, nata kending zhaoni zuo, suoyi zui guanjian de dongxi wo juede jiushi zuoren, rang renjia nenggou xinren ni, zhe yidian shi zui zhongyao de.)Given this phenomenon, friendship talk or friendship identity can be strategically employed by the insurance sales agents to serve various purposes, as will be elaborated and analyzed later.

Institutional Talk
The criteria for determining the institutionality of a segment of speech are the three features of institutional talk identified by Drew & Heritage (1992).Based on their work in conversation analysis, the three features are correlated with activities and goals, constraints, and inferences: (Activities and Goals) Institutional interaction involves an orientation by at least one of the participants to some core goal, task or identity (or set of them) conventionally associated with the institution in question.In short, institutional talk is normally informed by goal orientations of a relatively restricted conventional form.
(Constraints) Institutional interaction may often involve special and particular constraints on what one or both of the participants will treat as allowable contributions to the business at hand.
(Inferences) Institutional talk may be associated with inferential frameworks and procedures that are particular to specific institutional contexts.(Drew & Heritage, 1992, p. 22) Being different from the ordinary conversations among friends, institutional talk in insurance sales is institutionally goal-oriented, that is, the participants' goal achievement is contingent upon their institutional affiliation, constrained by turn-taking rules, and made possible through institutional inferential procedures in question.The features can be unveiled by the following excerpt from a conversation between an experienced insurance sales agent and her prospective client.The agent is selling a kind of policy by showing the benefits, the policy being introduced is claimed to have both the function of savings and investment.In the introduction, she shows how the policy works by presenting a table which shows the annual earning and other benefits possible.
(A: Insurance sales agent, female, around 35 years old, experienced with over 5 years' experience; B: Prospective client, male, around 30 years old, relatively well-off, having a small privately owned business in Guangzhou, now at home having his days off for the Spring Festival) It can be clearly seen that this dialogue is marked with turns of unequal lengths, with most turn initiated and maintained by the insurance sales agent whose goal is obvious-to sell the policy in introduction and to promote the advantages, values and positive aspects of the policy.While in friendship talk, turns are normally equally distributed and maintained, not allowing participants to dominate a particular turn, nevertheless, in promoting business and selling products, a salesperson should have longer turns to enable him/her to have more detailed introduction.In the above excerpt, the agent's upholding of much longer turns and her prospective client's cooperation in not interrupting indicates the agent's institutional affiliation to one of the major insurance companies as an expert, which can invoke the frames of superiors' authority and subordinates' obedience, to the agent's interactional advantage.Moreover, the change of footing is also achieved through the use of the inclusive "we" and "our product" in T(1), in which the institutional role of the insurance sales agent can be resumed.By invoking the institutional affiliation, the agent can resume the discourse identity of an expert (being an author and principal of an institution) and can more easily accomplish her objective of persuasion.On the other hand, in friendship talk, or in task-oriented talk (will be examined later), the discourse identities a salesperson can play are rather limited to the principal of his/her personal roles, that is, as a friend and the animator of the company he/she is representing.Being only an animator of the company, his/her selling is obviously less convincing and more subject to contention and challenge, especially in the context of rampant overuse of such promotional efforts as commercials and advisements nowadays, in which people may resist or feel obnoxious.
Institutional talk not only allows the agent to adjust appropriate discourse identities, but also puts constraints on the participation of the prospective client, to the advantage of the agent.Resembling closely to the asymmetrical interactions of doctor/patient and teacher/student, institutional talk in insurance sales discourse allows the agents to question their prospective clients and to sustain a turn long enough to do their jobs.The question in the institutional talk in the above interaction serves a very important function.First of all, by asking questions to the prospective client, the agent is able to understand quickly how much technical knowledge the client knows, thus, she would be able to adjust the technical level of her talk.Secondly, by engaging in a question-and-answer sequence, as in other asymmetrical interactions, the agent is in a better position to control the introduction of topics and to convince her prospective client of what she is saying-establishing the impression that she is a person with technical/professional knowledge, as in a doctor/patient or teacher/student interaction.To put it another way, the inferential procedures of those kinds of asymmetrical interactions are being invoked to the advantage of the agent, in place of the more familiar buying/selling inferential procedures in which the buyer and seller are also in an asymmetrical relationship, but with advantage reversed.Apart from adjusting the discourse identity for selling, institutional talk can also be used to alienate the responsibility of the insurance agent when there is trouble, which will be addressed later.The following excerpt of a conversation between an agent and client can illustrate this phenomenon.
The agent is a female in her 40s, and she is an experienced insurance sales with more than four years experience; about a month ago, the client has already purchased from her an insurance policy worth of 100,000 Yuan with a term of ten years.Due to some reasons, the client wants to withdraw from the contract.The conversation happens when the agent is called to the client's home.
A: agent B: client 1) A: XX, nihao a. (Hello, XXX) 2) B: XX, ni nage zenme gaode o? shangci wo genni qianle, huilai wogen wojia ta yijiang, tajiu juede buhao, xianzai women bugaole.(XX, how is it?Last time after I signed the contract, I talked to my husband, and he didn't like it, so now we are quiting.)the contract is still valid, but the coverage will be reduced.For your case, if you choose so, the coverage will be reduced to…) From the above interactions, it can be clearly seen the agent employs institutional talk to persuade the client to give up the idea of terminating the insurance contract, marked by unevenly distributed turn length (Turn 9, 11, and 13), focused core goal (the agent tries to persuade the client to give up the idea of terminating insurance contract, the client wants to terminate it anyway), and most of the agent's talk is associated with the insurance company and insurance policy in question.In short, turn 9, 11 and 13 are the clear embodiment of being informed by goal orientations of a relatively restricted conventional form.
To be more specific, in turn 5, the agent sets the scene of "huge loss" if the client terminates the insurance contract, "Xiache o, nachangzi huabulai o, qishi ni zhege wohai zhen huale xinsi geini gaode, nuo, ni xianzai tuibao dehua, genju hetong, ni budan benjin nabu huilai, haiyao sunshi weiyuefei, nage shuzi ye buxiao ai…" (Nonsense, it is a good deal.Frankly, I've devoted a lot of efforts to your policy.Well, if you the contract now, according to the contract, you will suffer not only your deposit money, but also a default fee, your loss would be substantial…" in the actual selling practice, when the agent is introducing the policy, most of them would emphasize the benefits while downplaying the constraints, it can be clearly captured in the client's surprised response in turn 6 "A? Weiyuefei?"(What?Default fee?), which indicates she has no or little prior knowledge about such "default fee" if she wants to terminate the insurance contract.More importantly, the agent goes on with long segments emphasizing the detriments or serious consequence if the client still insists in the termination of the contract, besides that, she offers options for the client to rule out some possible reasons and difficulties the client might have.For example, turn 9, the essence of insurance, mid and long term investment, prior emphasis on guarantee ("Zhong changqi de touzi, qianqi zhongde shi baozhang, zheshi baoxian de zhendi"); turn 11, a possible loan (Baodan xianjin jiazhi daikuan), and turn 13, a possible reduction in insurance money (Jian'e jiaoqing); after such tactics, the client softens her determination and anger, becomes hesitant about the decision of terminating the contract, as in turn 12 "Aiya, ni zhege zhenshi gao siren le" (This is really troublesome...).
It may be argued whether there are rampant uses of manipulation on the part of insurance sales agent, nonetheless, our primary focus is to unveil the important functions institutional talk in the process of insurance sales, that is, not only for the necessary introduction of the insurance policy in promotion, but also in turning the client around if there is possible termination of the contract; Furthermore the dynamics between the agent and client and the underlying mechanism under which such sales interactions take place.
There are significant implications for such interactions.It will be argued as follows: On one hand, apart from the previous studies about the asymmetrical distribution of turns in expert/layman interactions (Heath, 1992;Linell & Luckmann, 1991) and employer/employee interactions (Gavruseva, 1995), Gavruseva (1995) points out that the asymmetrical pattern is discursively constructed and contested.Employees, despite their situational disadvantage, are able to resist and contest the existing asymmetrical structure by revoking changes in frames and footings.The interesting finding in the above analysis is that all institutional talk is asymmetrical, with the power going to the agent.Although sometimes the prospect may (un) intentionally raise objections to the institutional identity the agent is constructing.
On the other hand, Guanxi is the most important factor in insurance sales in transformational rural China, the success or failure of the insurance agents' selling efforts largely depends on how well the agents are managing the relationship; in other words, sometimes it is not a matter of the insurance policy itself, rather it's the personal trust or credibility the agents enjoy from the prospective clients.Therefore, it sometimes enables the agents some room to exploit such trust and credibility, this is also why there are heated discussions on the mal-practices of insurance sales nowadays in transformational rural China.No matter what, there is clear indication of inter-discursivity (Fairclough, 1989(Fairclough, , 1992(Fairclough, , 1995(Fairclough, , 2003) ) in the process of insurance sales, and a greater tendency of confusion between friends and business, or as the post-modernist view of "commodification of self".

Task-Oriented Talk
Task-oriented talk, the most predominant talk in my data, is different from institutional talk in several ways, although both have the same goal orientation.Task-oriented talk is embedded in the configuration of friendship talk, which in Fairclough's term as inter-discursivity (1989,1992,1995,2003), the constitution of a text from a text from a configuration of other text types.As a result of this inter-discursive nature, task-oriented talk has the features of both friendship talk and institutional talk, but deviates itself from the others in the following several ways.
Task-oriented talk is marked with equal turn distribution, mutual topic initiation, and cooperative interruptions, as well as a lack of asymmetrical features, such as the predominance of the question-answer sequence.Nonetheless, this type of talk has strong goal orientation as well, to get something done through the talk.The following is an example to illustrate task-oriented talk.Before this part, the agent has just finished presenting her prospective client some details of the policy in discussion regarding how is it like.She ends their conversation with a task-oriented talk in which the agent sounds more like a friend.But at the same time, she is doing her job, advising the time that her prospective client should come to the promotional meeting at a fancy hotel.
A This short segment of talk is significantly characterized by the predominance of features of friendship talk that is, almost equal turn length and topic contribution, lack of one-sided question-answer sequence, as compared with institutional talk analyzed above.Some interruptions are also common in both participants' turns.Nonetheless, the talk has a fairly strong institutional goal orientation-to get the prospective client to attend the gathering scheduled in a fancy hotel (where a normally substantial purchase could be made).The embedding of goal-oriented talk in friendship talk serves the pivotal goal of creating a friendly environment for the interactions in insurance sales.Compared with the traditional sales encounters, such tendency blurs the boundary of friendship and business, to a certain extent, business is integrated into friendship now.Therefore, the insurance sales agents enjoy the more egalitarian relationship.They are friends to each other even when they are talking about business.However, there is also a price to pay with the frequent occurrence of task-oriented talk, it is now becoming more difficult for the agent to control the talk, such as what to talk about, and where will the talk go, it becomes an intense struggle for both agent and his/her prospective client in determining which one is more "powerful" in negotiating a friendly relationship.Task-oriented talk appears frequently at the juncture of other types of talks, where more changes in footing and frames are involved to negotiate to the participants' advantage.

Summary of Three Types of Talks
The characteristics of each talk mode are summarized in the following table, and will be discussed in greater detail in the following part.Meanwhile, it should be noted that the talk categories are only proposed to enrich the analysis, and for the purpose of quantification of interactional devices, they are proposed for analytical purpose.

Summary and Conclusion
The above analysis of talk in agent-client interactions unveils the dynamic process of insurance sales agent-client interactions through the three different types of talk, namely, friendship talk, institutional talk and task-oriented talk.Friendship talk reinforces the rapport between the agent and the client and reduces the potential tension between them.Institutional talk is mainly used by the agent to activate the institutional frame for controlling turns and initiating topics for persuasion.While task-oriented talk provides effective means for the agent in controlling and socializing the client, and for the client to assume more control in topics, moreover, it is also an arena rampant with intense struggle owing to the contestation between the institutional frame and friendship frame.From the above analysis, the study analyzes the discursive patterns and functions of moves and steps, in the interactions between insurance sales agents and their clients in transformational rural China.Specifically, we have summarized the specific communicative purposes of insurance sales dialogues.With these communicative purposes in mind, the analysis model for the generic structure of insurance sales dialogues is offered and analyzed.Consequently, the underlying cultural sources that account for the features are interpreted according to the functions in the collected data.
Secondly, the study has enriched the socio-cultural study of transformational rural China in examining insurance sales agent-client interactions in terms of exploring the discursive features in such interactions.In other words, money has become more important in rural Guanxi, as manifested in language, more specifically, topic management.As most of the current studies of transformational rural China, either focus on ethnographic study or logical reasoning.In contrast, this study manages to incorporate macro-analysis of the socio-cultural environment with micro-analysis of more detailed discursive features in insurance sales interactions in transformational rural China.
Thirdly, by integrating modern social theories and concepts (Fei, 2007;Zhai, 2007Zhai, , 2009;;Cao & Chen, 1997;Sun, 2001;Lu, 2002), we describe and generalize the communication patterns for insurance sales in transformational rural China from the perspective of topic control; meanwhile, we manage to uncover the hidden agenda of agent-client interactions with some interpretations from the perspective of socio-cultural interpretation with empirical data.
In conclusion, the study aims to investigate the inherent dynamics of insurance sales agent-client interactions in transformational rural China, and it is found that in rural areas, the decision of making the purchase of insurance policy can be summarized as the following: Firstly, there is an emergence of understanding of the necessity of such a service given the fact of the insufficiency of social security the government offers in such areas as education, medical expenses and pension plans.While at the same time, due to the rising standard of living of local residents, simply put, there is a present necessity and financial feasibility to purchase further services to improve their well-being.
Secondly, in transformational China's rural communities, Guanxi and Renqing (connections and interpersonal relations) are still quite rampant and pervasive in everyday life, from daily life to dealing with others.While currently, insurance sales in rural areas are still mostly conducted through the agents' extensive connections with the local residents, for most of the time, the sales are achieved through the trust of the clients in the agents, which requires almost impeccable credibility and reputation of the agents, that is also why almost all agents who are doing well (having good sales performance) are locally well-respected individuals.
Thirdly, the changing social conditions, or the decentralization of central authority, presupposes a downward responsibilization.Individuals are becoming more and more independent and assume more responsibility, not only greater freedom in earning, making a living, but more risks taken by themselves, compared with the state allocation of resources.It's unwise and dangerous to impose a simple moral judgment of which is better, however, it's a general fact and tendency in China's current transformational rural areas.
Fourthly, the selling efforts of the insurance agents' are almost irresistible.In the investigation, there are certain extreme cases that the agents' constant, inexhaustible and repeated visits and talks make the prospectives annoyed and thus the whole situation turned into a mess, which could be conducive to the fact that in current transformational China, many people have certain bias or prejudice against insurance sales agents.
Last but not the least, given the fact that people's living standard is getting higher and higher, more and more wealth accumulated through various means, many people just don't have the necessary knowledge or expertise on how to manage their wealth.Thus, they have to resort to someone who seems to have the relative knowledge and expertise, which consequently offers some room for the insurance agents to take advantage to make the sales.Moreover, the rapid improvement in life makes the local residents be eager to show off, just as an old Chinese saying goes, "Not returning home after getting rich, it's just like parading with beautiful dresses at night".The idea of "face" in the Chinese people's minds is deeply ingrained, therefore, the work of agents can, to some extent, satisfy the need and fill this vanity.

Table 1 .
Summary of the characteristics of the three types of talk