The role of tag questions in medical encounters

The discourse of medical encounters is deemed to be an excellent example of
 both institutional talk and discourse of power. Asking questions is probably
 the most prominent characteristic of doctor- patient interaction and this
 paper deals with tag questions as one of the question types that can be
 found in almost every medical encounter. We will explore tag questions by
 reviewing current research results in the field of medical discourse and by
 comparing and discussing examples from transcribed medical encounters in
 English and Serbian. It will be discussed how often tag questions are used
 in both corpora, whether doctors and patients use them in the same way and
 what role these questions have in a medical encounter. Finally, having in
 mind that getting to ask any question in institutional talk requires a
 certain amount of power, we will also try to determine if using tag
 questions affects doctor and patient?s positions in a medical encounter.


INTRODUCTION
Institutional communication has been explored from various aspects and in a multidisciplinary manner by scholars from different fields (i.e. linguistics, sociology, anthropology, psychology, etc.). This type of communication usually involves interactions such as teacher-student, doctor-patient, police officer-witness or journalist-interviewee, where one participant in communication is an official representative of an institution and the other one is the client. While everyday communication usually implies a symmetrical interaction between the two parties that have equal rights, institutional communication is commonly characterized by a conspicuous asymmetry and an unequal distribution of power [1]. For this reason, Fairclough believes such interactions should be called unequal encounters [2].
Doctor-patient communication is believed to be an excellent example of a specific distribution of power in the field of institutional communication. It may be explored from various aspects, but the use of questions is probably the best way to illustrate some significant differences between the roles of an institution representative and of its client. This paper will first briefly present the classification of questions and general reasons for asking questions in this type of institutional communication. In next section, the category of tag questions will be defined and explored in more detail. Examples from two corpora (i.e. English corpus and Serbian corpus) will be used to illustrate the results that have been provided by comparing and contrasting recordings of doctor-patient communication in two languages. This paper aims at exploring how often tag questions are used, how doctors and patients use this type of questions and what role these questions have in a medical encounter in two different languages and two different cultures. Finally, we will try to determine to what extent the use of tag questions affects doctors and patients' roles in a medical encounter.

QUESTIONS IN MEDICAL ENCOUNTERS
Different authors offer different definitions of questions asked in medical encounters and the same is true for the types of questions and the role they have in doctor-patient interaction. However, multiple authors have reported that most of the questions are asked during the phase of history taking and that they are mostly asked by doctors, not patients [3,4].
Mischler was one of the first authors who dealt with this aspect of medical encounter and in his study he identified several types of questions -yes/no questions, 1 polar/ disjunctive questions, 2 restrictive WH-questions, 3 moderately open How-questions 4 and open How-questions 5 [4]. He also came to a conclusion that most of the questions asked by a doctor restricted patients' answers because of 1 E.g. Do you have temperature? The only two possible answers to such questions are yes or no. 2 Two alternative answers are provided in this type of question, e.g.
[...] this cough, are you producing anything or is it a dry cough? [4]. 3 E.g. When did you first notice that symptom? The patient is expected to provide a specific answer. 4 Questions beginning with How about, e.g. How about your legs?
They provide an interlocutor with an opportunity to tell their story about a particular thing (i.e. their legs), without being limited to a specific answer. 5 E.g. How are you feeling? the way these questions were formed, which consequently allowed the doctor to control both the interview and the patient's contribution to it to a great extent [4].
Ainsworth-Vaughn opted for a slightly different classification where she identified the following types of questions: WH-questions, 6 inverted auxiliary, 7 "search" WH-questions, 8 elliptical questions, 9 tag questions and rising intonation pattern. 10 She also explored the so-called rhetorical questions that are more like requests or statements than real questions as there is no need for an actual answer [5].
Boyd & Heritage and Hayano explored doctors' questions and they emphasized the triple function doctors' questions had in a medical encounter -(1) to limit the patient's answer by imposing a specific agenda, 11 (2) to assume various aspects of the patient's health and their medical knowledge, 12 and (3) to favour one type of an answer over another 13 [3,6]. Boyd & Heritage also insisted on the optimization principle -the fact that doctors' questions during the phase of history taking are formulated in such a way that they always predict the best-case scenario [3] 14 .
Interestingly enough, most of the studies deal only with doctors' questions and do not take patients' questions into consideration [3,4,6]. Some authors, like Mischler, do not even believe a patient can ask a question during a medical encounter, as they consider this to be an exclusive privilege of a doctor [4]. Having analysed over 2500 recordings, Byrne & Long came to a conclusion that it was asking questions that gave doctors such a great power over patients whose contribution was entirely insignificant and limited to the very end of the encounter [7]. Fairclough claims that the entire encounter is based on doctors' questions and that a doctor has a complete control over the turn-taking system in this type of communication [8]. He 6 The questions that Mischler [4] classified as restrictive WH-questions. 7 Yes/no questions in Mischler's classification [4]. 8 In this type of questions, a question word never occupies the initial position, but it is usually located at the very end of the question, in the position that matches its syntactic role. E.g. You have an appointment when? 9 These are not real questions, but they have a communicative value of a question. [9] In these questions, the topic from a previous context is emphasised, so that both interlocutors know what the question is about, e.g. Everything else remains unchanged? (here both the doctor and the patient know exactly what everything else is referred to). 10 At first sight, these are not questions at all, as they have the structure of a statement. However, owing to its rising intonation, they have an interrogative communicative value [9], e.g. You've got the previous report? 11 Imposing a specific topic or offering an agenda concerning a specific activity expected from the patient (i.e. to give an affirmative or a negative answer, to give an explanation, etc.) 12 This is more often the case in WH-questions than in yes/no questions. In the example What medications do you take? the patient is not expected to say whether they take medications or not, but they are just supposed to give an answer to a restrictive question. 13 This actually means that a question is formulated in such a way that an affirmative or a negative answer are naturally expected (e.g. You don't suffer from heartburn?). 14 For example, doctors would always rather ask Is your father alive?
than Is your father dead? also believes that the patients' contribution is absolutely negligible and that patients actually speak only when they are encouraged or prompted to do so by the doctor's questions. However, some other authors have proved that patients do ask questions during a medical encounter and that they even do so regularly, but certainly less frequently than doctors [5,10]. 15 Why do doctors and patients ask questions in a medical encounter? Klikovac states several reasons for doctors to ask questions -(1) they want to find out about everything they might consider important prior to giving a diagnosis or deciding on a treatment, (2) they wish to direct the patient towards a particular response or to encourage them to continue talking and say more about a particular topic or (3) they do this in order to make a conclusion [11].
Cordella studied both doctors' and patients' questions and she identified several different voices both groups used according to a situation or a personality type [12]. Concerning the patients, Cordella determined only one group of patients (i.e. the voice of initiator) who actually asked questions and they did so in order to find out more about their health, to understand their situation better and take a better care of their health [12]. McKenzie conducted a study that involved pregnant women carrying twins and most of them admitted they believed they should make a list of questions they wanted to ask their doctors in advance, before going to their surgery, as well as that they were unsatisfied with the fact their doctors did not share information with them [13]. Adler et al. even stated that some patients 16 thought it was inappropriate to ask any question whatsoever as it would indulge the doctor-patient relationship [14].

TAG QUESTIONS
Tag questions are commonly used to check if something is true or to ask for an agreement [16]. In English, tags are usually composed of an auxiliary verb or the verb do, depending on the form of the entire sentence, and they are positioned at the very end of the question, behind a comma. A negative question tag is normally used after a positive sentence, whereas a positive question tag is used after a negative sentence [17]. 17 In Serbian, tag questions have not been explored much, but there are two terms that could be used as equivalents to the English term question tags -dopunska pitanja [18] and finitivne rečenične upitne forme [9]. Just like in English, an interrogative word is located at the end of the sentence and such position is considered to change the communicative status of the sentence turning a statement into a question [9]. It is important to say that only an 15  affirmative answer is expected to these questions and they are related to something that was previously stated [18].
Most frequently used interrogative markers in Serbian tag questions are the following: je l'?, je l' da?, zar ne?, je li?, a? and unlike English tag questions they never change their form depending on a tense or a structure of the sentence 18 . These tags are called (dopunski) upitni operatori [9] or privesci [19]. Interestingly, Hudeček & Vukojević state that tag questions are found exclusively in everyday communication, which could be important in determining the genre of a medical encounter [20]. 19 Tag questions are interesting to explore, as they are quite different from most of the other types of questions. As it is always an affirmative answer we expect to such questions, they are often not even seen as questions; this could be an explanation for the reason why only Ainsworth-Vaughn counted them as questions [5]. We thought it would also be useful to see how often doctors and patients asked tag questions and if they used them to the same purpose. Since tag questions are commonly used to check a particular piece of information or ask for an agreement, they could be seen as a power-claiming tool (i.e. expecting someone to agree with you, never to oppose you or state their own opinion) or simply as a way to encourage another person to say more about something.

RESEARCH RESULTS
In order to check how often doctors and patients use tag questions and what role these questions have in a medical encounter, we compared recordings of medical encounters in English and Serbian.
The English corpus consists of 19 recordings, whereas the Serbian corpus contains 80 recordings, but both corpora were approximately 5 hours long and could be considered equal in length. The recordings from the English corpus are a curtesy of Prof. Richard Frankel, they have been made in a tertiary referral hospital in the USA and they belong to the fields of orthopaedic surgery (14 recordings) and internal medicine (5 recordings). The 18 E.g. To je sve, je l' da? 19 While some authors firmly believe that a medical encounter is a typical interview with strictly defined roles of an interviewer and an interviewee [4,10], some other authors claim it is actually a combination of two genres -an interview and everyday communication [5,21,22,23].
recordings from the Serbian corpus have been made in a tertiary referral hospital in Belgrade, Serbia, and they belong to the fields of pulmonology (37 recordings) and paediatrics (43 recordings).
Out of the total of 334 questions asked in the English corpus, there were only 8 tag questions (2.4%); doctors asked 4 tag questions and patients asked 4. In the Serbian corpus there were 40 tag questions out of the total number of 497 questions (8%); 30 tag questions were asked by doctors (75%) and 10 (25%) were asked by patients. So, we can conclude that tag questions are rather rare in English corpus, whereas in Serbian corpus they are more frequently used, but more by doctors than by patients. The results are presented in Table 1.
Doctors in English corpus always asked tag questions simply to check a particular piece of information (Example 1). In this example the doctor used a tag question to check a piece of information -the dosage of the medication the patient is currently taking. By opting for this particular type of the question, the doctor also directs the patient to a particular answer he considers adequate. So, by asking this question the doctor does not seek a new piece of information, but he just wants to confirm his assumption of the information he already possesses.
Unlike their American colleagues, the doctors from the Serbian corpus used tag questions in several different ways. They most often chose these questions to check a particular piece of information (Example 2), but they also opted for them when they wanted to check if the patients understood what they had previously told them (Example 3), to make the atmosphere more relaxed (Example 4), to get more time or make the patient change their mind (Example 5). Paediatricians also used tag questions to direct their patients towards a particular answer (Example 6). The doctor obviously asked this tag question either to make a statement or to check something he already knew, so he did not really expect the patient to answer. By asking this question, he hopes for a confirmation, not for a negative answer. The tag don't you (Serb. je l') which is used here is the most frequent tag in the entire Serbian corpus. As expected, the patient confirms the doctor's statement by using a minimal response yeah (Serb. a-ha).
Example 3: D: Now, when you start using the new dosage 80, you should follow what's going on. If you see it's stable, that nothing changes, then go on with it. If you see something's going on, come before the end of 3 months (..) Or maybe better 2 months. As we've got a low dosage, so we should follow your reaction (.). We've understood each other, haven't we? P: Yes.
In this citation, the doctor is giving instructions to the patient, concerning a new dosage of the medication he has already been using. For the most part of his turn, the doctor is determining the patient's behaviour by telling him what to do and by revealing the two possible scenarios. Interestingly, the doctor suddenly starts using the 1st person plural -we've got a low dosage (Serb. imamo malu dozu), we should follow your reaction (Serb. da vas ispratimo). He uses the 1st person plural either to denote both himself and the patient in an activity that only seems to be mutual but actually depends only on the doctor or to denote himself and other doctors, as professionals [11]. Finally, the tag question We've understood each other, haven't we? (Serb. Razumeli smo se, je l' tako?) is used only to check if the patient has heard and understood everything, but it is also a good example of the discourse of power in institutional communication -the doctor only expects a confirmation (so, the patient is supposed to understand well everything he's been told) and by using the 1st person plural he may put himself in the position of an authority. Example 4: D: It's salty, the sea. It's too salty, isn't it? P: Yeah.
Here the paediatrician uses a tag question only to make her little patient laugh and help him feel more relaxed. She makes a joke about the sea being salty and she obviously does not expect to get a real answer to her question, she only expects her patient to agree with her and get the joke. In this example, the tag question is asked in order to make the patient re-think what he has just said or simply to provide the doctor with some more time to think about what he has just heard. It is also possible that the doctor expected to get a positive answer and was thus a bit confused, so he took some more time to think about the problem. Example 6: D: So, tell me, how often do you have these headaches? P: Well, twice a week. D: When you're on holiday and when you go to school? When you have obligations and when you don't? Or are they more frequent when you have obligations. Now during the summer holiday it happens less frequently, doesn't it?
P: Yes. This is a typical example of directing the patient towards a particular answer or helping them to answer properly. Paediatricians used tag questions to this purpose rather frequently, whereas there was not a single example of this usage in encounters with a pulmonologist. This might be connected with differences between paediatrics and almost all the other branches of medicine -paediatricians talk to children (and their parents) and they need to find a way to learn everything they would like to know, without intimidating the patient.
The patients in English corpus usually used tag questions only to check a particular piece of information (Example 7). In this quote the patient asked two tag questions. The first question was negatively formulated and even though the patient asked it hoping that there is an alternative treatment, he still expected to get a negative answer. The doctor hesitated, as if he did not want to disappoint the patient, but he finally admitted the medication the patient was currently using is the only option. The patient then asked another tag question, asking for a confirmation of what even he himself had noticed.
The patients in Serbian corpus used tag questions in the same way -to check information (Example 8), but also to check if they have understood their doctor correctly (Example 9). In this example the patient simply checked whether a piece of information he had was correct so that he would not make a mistake. Instead of asking a yes/no question (e.g. Will I have to get a referral once again?), he turned it into a tag question, as he believed it would be necessary for him to get a referral before seeing his doctor again. Surprisingly enough, the doctor gave him a negative answer. This is the final phase of the medical encounter and the doctor is discussing the therapy and giving instructions. So, the patient's tag question If I need to, right? (Serb. Po potrebi, je l'? has a role of finding out if he has understood the doctor correctly, to make sure he will do everything properly. This is not a typical tag question, as it is a bit elliptical, but we can certainly consider it a tag question, as an affirmative answer is expected.

CONCLUSION
Doctor-patient communication, as one of the most researched types of institutional communication, is explored from various aspects. Asking questions in a medical encounter is believed to be one of the most important features of this type of communication. According to some authors, the right to ask questions is an exclusive privilege of doctors, whereas some other authors claim patients can ask questions as well, but less frequently and to different purposes.
It is thought that having a privilege to ask questions provides doctors with a powerful tool for controlling the entire encounter -they use questions to direct patients towards a particular response, to encourage them to say more about a topic or to determine their behaviour during the encounter. Patients do ask questions as well, but they obviously do it in a significantly different way, usually with the purpose of finding out more about their health.
Tag questions have not been researched as much as some other types of questions and they were counted as questions in only one larger study [5]. These questions might not even look like questions, as only an affirmative answer is expected. So, posing a tag question basically means asking for a confirmation or checking information.
In our research, medical encounters from two different corpora, English and Serbian, have been compared in order to check how often tag questions are used, how doctors and patients use them and what role they have in a medical encounter. According to the research results, tag questions are almost four times more frequent in Serbian than in English corpus. In English corpus doctors and patients asked the equal number of tag questions and both groups posed these questions only in order to check a particular piece of information. In Serbian corpus, doctors asked tag questions much more frequently than patients. Apart from checking information, they used tag questions in several other ways -to help the patient relax, to get more time or make the patient re-think what they had said, to check if the patient had understood something or to direct the patient towards a particular answer. Patients in the Serbian corpus used tag questions to check particular information or to check if they had understood their doctor and instructions correctly.
When it comes to the discourse of power and the actual positions of doctors and patients in a medical encounter, we cannot agree that patients are completely deprived of the possibility to ask any kind of questions, including tag questions. In the English corpus there are almost no differences whatsoever between doctors and patients. However, in the Serbian corpus doctors ask these questions much more than patients and certain characteristics of the discourse of power can be recognized -doctors sometimes use tag questions to direct their patients to a particular answer they find most appropriate. On the other hand, patients never ask tag questions to this purpose, but exclusively in order to check a particular piece of information or if they have understood doctor's instructions correctly, which again proves the existence of asymmetry in doctor-patient interaction. [24,25,26]