Medical terminology circulation and interactional organisation in interpreter-mediated medical encounters

Two medical encounters taking place in a Northern I talian hospital are analysed in this paper from a qualitative point of view, based on the author’s previous research. The aim is to reveal th e strategies adopted by medical interpreters, in these two specific cases, to translate medical terminology and promote/exclude interlocutors’ acti ve participation. This latter aspect is influenced by the way the interact ion is socially and linguistically organised and, in particular, by how interlocutors’ utterances are translated. The prevalence of dyadic or triadic sequences and especially the shifts between such communication ex changes are pivotal in fostering or hindering interlocutors’ participation . Furthermore, medical interactions, as a form of institutional talk, ensh rine specific expectations, which are mainly of a cognitive nature but may also be affective, as in the two encounters observed. By conveying such expectat ions and expressions of personal interest, interpreters have proved to c ontribute to the fair distribution of active participation among primary interlocutors. Hospital ethical approval and subjects’ written informed con se t have been obtained.


Introduction
This study aims to illustrate the strategies adopted by interpreters in medical settings to convey medical terminology and to promote or, alternatively, exclude the interlocutors' active participation in the encounter.For these purposes, two mediated encounters are here examined from a qualitative point of view.Attention will be first paid to how medical terminology circulates in the consultation (Bersani-Berselli, 2009a, 2009b), by also exploring the doctors' use of medical terms and, second, to how the mediated interaction is organised in terms of turn-taking, sequences and communication exchanges, as well as shifts between such exchanges (Baraldi, 2009a(Baraldi, , 2009b)).The specific interactional organisation namely promotes, or else excludes, the participation of primary interlocutors in the encounter.In the author's mind, such investigation is quite timely in view of the few in-depth studies on discursive interactions, able to show the contribution of all participants to the encounter's success or failure.
The analysis will start from observations on the dialogic process and the relationships between interlocutors during medical consultations.In this respect, the linguistic analysis of transcribed interactions, as proposed hereafter, is crucial to detect structures of discursive behaviours.

Different linguistic varieties at play during medical encounters
An issue to which great attention has been devoted by researchers is the use of specialist terminology in mediated medical encounters.As stated in the introduction, its circulation in healthcare settings is one of the two underlying threads of this paper too.According to Bersani-Berselli (2009a, 2009b), a sort of "non professional" variety of medical language coexists with the "professional" one.It derives from the extension of common language to include medical terms and phrases.The analysis of a sample of medical consultations revealed that doctors mainly resort to the "professional" variety when addressing other professionals directly, when writing/reading medical reports or in the treatment phase.In such cases, the doctor is addressing a peer who shares the same technical knowledge.No further explanation is therefore needed.On the other hand, the large sample of data collected-approximately 100 consultations- (Bersani-Berselli, 2009a, pp. 462-463), indicates that healthcare staff, in most of cases, adopt a "non professional" linguistic variety and avoid technical terminology when their utterances are to be conveyed to patients and hence when addressing the interpreter, in order to minimise the risk of communication failure.This happens for example during the complaint, the examination and the treatment phases, as confirmed by the excerpts reported hereafter.
The author observed that shifts from this identified pattern may occur on certain-limited-occasions (Bersani-Berselli, 2009a, p. 466).For example, hedges may be introduced, especially in the form of diminutives, as is the case in the first encounter.This results in a significant lowering of the linguistic register.Such shifts may signal the healthcare staff's attempt to maintain control of the conversation, which is the case, according to Bersani-Berselli (2009b, pp. 166-167), when they perceive that their "leadership" has been undermined by the presence of interpreters or by patients.
Notably, out of the six phases identified by ten Have (1989, p. 118)-opening, complaint, examination or test, diagnosis, treatment or advice, and closing-technical jargon is used more frequently during diagnosis and treatment.It is almost absent, on the other hand, during complaint and verbal and/or physical examination so as not to hinder or stop the information flow and to avoid misunderstandings (Bersani-Berselli, 2009b, p. 155), as confirmed by the cases examined below.An explanation might be that the patient is only the "immediate" addressee in diagnosis and treatment, whereas the true addressees are, indirectly, other professionals.
It is worth noting that the way interpreters translate medical terminology, which tends to respect the tenets illustrated above, might depend on several intertwining factors, as evidenced in the following consultations.This justifies the need to analyse real interpreting sessions in the medical field, so as to evaluate interpreters' strategies and investigate the underlying reasons which might have caused them.

Promotion/exclusion of interlocutors' active participation
The second issue of this paper, that is, interactional organisation and, consequently, implicit or explicit strategies adopted by medical interpreters to favour/hinder interlocutors' active participation, is here discussed on the basis of Baraldi's research (2009aBaraldi's research ( , 2009b)).Baraldi notices a direct link between the way utterances are translated by interpreters and the distribution of participation among interlocutors.He observes that any mediated interaction is characterised by a rather complex social and linguistic structure.With reference to the social organisation of the interaction, this is to be understood as a communication system based on adjacency pairs and enabling interlocutors to act.From a linguistic point of view, since any mediation/mediated interaction involves two languages, it comprises both dyadic (monolingual) and triadic (bilingual interpretermediated) exchanges (Valero Garcés, 2007, p. 35).The latter ensure coordination among interlocutors speaking different languages.In mediated interactions, moreover, interpreters frequently shift between dyadic and triadic exchanges.This leads to a specific "structure" or form of translation (hereon referred to as "translation structure"), that is, a specific organisation of interactive sequences, which affects interlocutors' participation.The interpreter's translation may thus foster/hinder interlocutors' participation.
In this specific context, as already proposed in previous paragraphs, the term "mediation" consequently refers to the process enabling interlocutors to take part in the interaction.It is in itself a form of interaction and communication system, which is required when the understanding is hindered by linguistic and cultural barriers, among others.The mediation process, moreover, is to be viewed as part of the wider system of the institution where it occurs, that is, the hospital/healthcare centre (Baraldi, 2009b, p. 48).In institutional-and hence in medicalsettings, interactions which took place at different times are frequently connected, since a doctor may refer to something which was already stated in a previous medical consultation or to previous actions.
(Mediated) medical interactions, like all kinds of lay-professional encounters, belong to institutional talk, as they embody its three basic features: specific goal orientations, linked to the institution-relevant identities; special organisational constraints; and institutionally-specific interpretative frameworks for the interaction (Heritage, 2005, p. 106).Since doctor-patient interactions are goal-oriented, interlocutors understand the meaning of actions performed and words uttered by referring to the institutional tasks or ultimate objectives of the interaction itself (e.g., to cure and treat patients).This understanding is based on expectations on the nature of the event and participants' roles in it.
In the case of medical consultations, for example when advice is provided to patients, cognitive expectations seem to prevail (Baraldi, 2009b, p. 53).These are based on observations about how things work and what is going or not going to happen: the doctor will for example expect the patient to have a chief complaint.Baraldi observed that in mediated medical encounters, practitioners tend to entrust interpreters the task of reporting to patients the instructions they generated.This is based on the expectation that by doing so the patient will more likely accept the treatment requirements, as evidenced by the analysis of real interpreting sessions (Baraldi, 2009b, p. 53).In medical contexts, interpreters are called on to report the healthcare staff's utterances so that patients adhere to doctor's recommendations/prescriptions.
If in the everyday practice interpreters tend to align themselves to the doctor's point of view, on certain occasions, however, they might give voice to the "personal expression (or self-expression)" of patients (Baraldi, 2009a, p. 8, italics in the original), which reveals the presence of affective expectations.These reflect the self-expression of participants, whose personal views are encouraged or accepted within the interaction.Despite being rather rare in medical consultations, affective expectations might be formed as well, as evidenced in the two cases below.
Real practice also demonstrates that when doctors express their personal interest or appreciations of participants' experience, interpreters may either omit to convey these expressions to patients and respond directly to doctors, or they may report such information to patients and vice versa (Baraldi & Gavioli, 2007).A failure to translate interlocutors' personal expressions causes distance between doctor and patient.Interpreters play a crucial role in this respect, since they may facilitate or else inhibit participants' personal expressions.By conveying them, they contribute to the fair distribution of active participation, addressing interlocutors' interests and needs.
Accordingly, medical interpreters may promote cultural adaptation to the institutional setting, that is, acceptance of explanations/instructions provided within the hospital system, and exclude patients from actively participating in the interaction.Conversely, they may enable all interlocutors to be involved in the encounter, by intervening and expressing their voices (Baraldi, 2009b, pp. 59-73).Specific patterns of expectations and turn-taking sequences are consequently implied (Baraldi, 2009a, pp. 11 and 13).
The shifts between sequences and communication exchanges (Baraldi, 2009b, pp. 72-73), and hence between different "translation structures", reflect the promotion/exclusion of interlocutors' participation.Some general trends can be summarised in this respect, which will be useful in the analysis of the two encounters below: 1) dyadic exchanges incorporating cognitive expectations are likely to hinder interlocutors' active participation; 2) the shift from a triadic structure to a dyadic one including affective expectations creates the conditions for empowering dialogue; 3) the shift from a dyadic to a triadic exchange aims at promoting active and fair participation of primary interlocutors.Real practice shows that when a shift from a triadic to a dyadic exchange takes place and affective expectations are involved, the interpreter supports the interlocutor by resorting to different tools. 1  In mediated medical encounters, interpreters are also frequently addressed directly (dyadic exchanges), as emerges in the first encounter below, or else included in the interaction in any case, even when not receiving the speaker's visual attention, as is the case in triadic exchanges.They consequently seem to be recognised as active participants in the encounter by primary interlocutors (i.e., healthcare staff and patients), if compared to interpreters working in other settings, who tend to be more "invisible", as demonstrated in the author's previous studies (Pittarello, 2009, pp. 78-79 and 104).
Interpreters' active participation in medical encounters and their shifting between different translation structures are confirmed by the following interpreting sessions.The two encounters selected are now presented against the background of the research context illustrated above, in order to exemplify the strategies adopted by interpreters to convey medical terminology and promote/exclude interlocutors' participation.

Corpus features and data analysis
The research further builds on the material collected in 2008 in selected health care units of a Northern Italian region (Pittarello, 2008(Pittarello, , 2009)), in line with the "case study" research method (Pöchhacker, 2002).The material comprised a questionnaire administered to 85 respondents (15 community interpreters and 60 healthcare personnel), 18 recorded interviews, the participant observation of 26 mediated encounters and the corpus-based analysis of four observed encounters.The aim, on that occasion, was to compare expectations and needs of the healthcare personnel on the medical interpreters' role in Italy with the opinions of community interpreters who work in the field.Interpreting practices were then analysed based on results achieved.
For the purposes of this paper and owing to space constraints, the analysis focuses on two out of the 26 mediated encounters which were observed in the previous study (Pittarello, 2008(Pittarello, , 2009)).These consultations were chosen as they better illustrate the two underlying issues of this paper: the circulation of medical terminology during the encounter as well as the specific organisation of the interaction, which fosters or else hinders interlocutors' active participation, as explained in previous paragraphs, through the shifts between dyadic and triadic exchanges.The sessions also depict, to some extent, some key characteristics of medical talk which are worth mentioning in this context: conversational asymmetry between interlocutors, particularly evident during the verbal examination and whenever medical issues relevant to the main topic of the encounters are treated; pursuit of an external goal; and doctors' unilateral control of the interaction, as evidenced by the information content and the turn-taking (Cambridge, 1999).The conversation, moreover, follows a pre-established pattern, corresponding to the standard phases identified by ten Have (1989).
The interactions took place respectively in the Emergency Department (ED) and the Healthcare Service for Tourists (HST) 2 of a Northern Italian hospital, frequently visited by foreign tourists in summer. 3 Interpreters and doctors' names have been replaced with fictitious initials to protect anonymity.Different doctors and interpreters were involved, thus giving an account of diverse personal attitudes and of their impact on the mediation itself.It should be noted that I2 is a trained interpreter and translator, whereas I1 has a degree in foreign languages.
The following table outlines the main features, indicating place, duration and language requested as well as the interpreter (I), doctor (D) and patient/user (P/U) involved and the main reason for complaint.English and German were used as vehicular languages, since patients/users were not English or German native speakers. 4The cases examined only aim to provide an example of the complex nature of medical mediation, without any claim of completeness.

Medical interactions in Emergency Departments: main features
The focus on mediated encounters occurring in an ED is due to their better highlighting the need to communicate with patients immediately and effectively.EDs are namely high pressure healthcare settings, where complex interactions are involved in order to provide urgent care to patients, with whom the medical staff need to interact rapidly.These medical encounters, if compared to consultations in other departments, tend to be rather fast and dynamic.Furthermore, the immediate and emergency nature of such encounters justifies the frequent lack of the opening and closing phases.Due to the speed at which medical consultations take place in an ED setting, the physical and verbal examinations also tend to be performed at the same time (Merlini, 2007, pp. 439-440) and patients may have to wait before being informed about the diagnosis or being given any advice.This is due to the temporal discontinuity of interactions taking place in institutional, and hence in medical, settings.The first encounter analysed only deals with communicating diagnosis and treatment procedures to patients, since both the "complaint" and "examination or test" phases had previously occurred.The second, on the contrary, comprises almost all the six ideal phases and well depicts the influence of interactional structures on fair participation distribution.
A further characteristic typical of ED encounters is the lack of doctor's preliminary knowledge on patients' case history.Consequently, patients are frequently questioned on their previous medical history.In the HST, interpreters also welcome patients and collect their personal data and information on symptoms.This might explain why these consultations at the HST mainly started with a dyadic exchange between the doctor and the interpreter.The latter is already familiar with the patient's conditions and reason for complaint and thus reports the relevant information collected to the doctor.The greater autonomy enjoyed by interpreters in the HST might also be due to the relatively low case severity.
Selected encounters will now be analysed separately, so as to provide real examples based on the two topics of medical terminology circulation and interactional organisation.Excerpts only refer to the most relevant turns and transcriptions are not reported integrally, owing to space constraints. 5

Encounter A: Swedish child suffering from neck pain
The first encounter, embodying the typical traits of an ED medical consultation, only comprises the final phases of diagnosis, treatment/advice and brief closure.The patient is a young girl from Sweden suffering from neck pain and accompanied by her father.They are admitted to the consulting room to hear the orthopaedist's diagnosis on the x-ray findings.
With reference to the encounter's translation structure, the prevalence of triadic exchanges is observed (lines 1 to 65 out of 91), since I1 immediately translates the primary interlocutors' utterances.The sole exception is a dyadic sequence between the father and I1 towards the end of the encounter, triggered by the father's request about whether booking the subsequent check-up visit with the orthopaedist is mandatory.The request is dealt with directly by the interpreter, as it regards routine administrative information (Pittarello, 2009, pp. 77-78).
As far as medical terminology is concerned, no specific term is used.Three aspects are nevertheless worth mentioning.Firstly, the doctor prescribes a small neck brace ("collarino") for a couple of days.Presumably, the use of the diminutive is not to be automatically interpreted, as in Bersani-Berselli (2009a, p. 466), as a doctor's shift from the standard use of medical terminology, for him to maintain control of the conversation.The adoption of a "lower" linguistic register might be due to D1's attempt to get closer to the interlocutor, by using a short (and affectionate) form.The patient is namely a child, who will necessarily wear a smaller neck brace.Secondly, D1 recommends paracetamol, yet this information is not immediately transmitted to the user by I1, who is keen to convey information on how long the collar needs to be worn.D1 hence repeats this advice in English (line 22) and this time the interpreter transmits the information, also adding the word "tablets": )) 31 U syrup (.) we have the::: we have the paracetamol Interestingly, D1 chooses the word "paracetamol", rather than the brand names "Tachipirina" or "Efferalgan".The two latter are more in use in Italy, where the chemical name of the compound is less known because it is not widely used as an over-the-counter analgesic as in other, especially English-speaking, countries.This choice might be due to the doctor's willingness to transmit information more effectively by resorting to a term which is more likely to be understood by a Swedish patient, thus showing possible awareness of the cultural background.Furthermore, owing to the interpreter's zero rendition, D1 attempts a direct interaction with the user, by mentioning the term in English (line 22).Notably, I1 autonomously suggests administering tablets (line 26) and D1 intervenes to say that syrup is fine as well (line 29).In doing so, he demonstrates that he has understood the interpreter's suggestion and endeavours to gain control over the turntaking, especially to play his role of expert within the encounter.The additional question posed by I1 (line 23) and her attempt to explain the use of paracetamol might be interpreted as efforts towards a better understanding, probably because she is not sure whether the recipient is familiar with the term, as English is not his native language.
D1's willingness to communicate directly with the user is proved by his numerous interventions in English (lines 22, 47, 60 and 62-63) and the frequent use of feedback markers-especially "okay"-signalling the need for confirmation from either the user (line 22) or the interpreter (line 39).The doctor's behaviour, in the author's view, proves to be contrary to Bersani-Berselli's assumption (2009a) that it is extremely rare for doctors to address patients directly in the presence of interpreters.Bersani-Berselli observed that mediated interactions seem to be preferably composed of doctor-mediator pairs (and vice versa) and mediator-patient pairs (and vice versa) (p.461).In this specific case (Example 2), not only does D1 address the user directly in the vehicular language (lines 60 and 62-63), but he also resorts to gestures in order to be understood.He even adopts a simplified language, devoid of any medical term, in reply to the father's question on how he can recognise evidence of improvement in his child: Example (2) 60 D =THIS is better ((slowly shaking his head from right to left)) 61 U okay (.) that is better (.) without pain (.) without pain= 62 D =with less pain (.) with less pain (.) no-not zero pain (.) three days it 63 should be:: right The third interesting aspect can be seen in Example (3): when D1 reports that the child might have to undergo an "x-ray" (line 39), the father asks for confirmation by using the word "tomography" (line 44) and I1 will afterwards align with his linguistic choice (line 52 versus line 41).The term tomography-an advanced form of imaging-is more specific if compared to "x-ray" and less widely used by laymen, at least in Italy.Notably, the observer has no further hints to whether the child has undergone a "simple" x-ray or a (computed/x-ray) tomography.It seems nevertheless that by choosing a more specific and technical term rather than its hyperonym, the father shows good command of the medical language relevant to this specific case.He thus raises the linguistic register of the encounter since the interpreter aligns with the user's term.D1's choice is again in the direction of a more direct and immediate information flow.
Example (3) Notably, D1 supports his statement (line 40) by explicitly inviting I1 to translate that the medical staff hope an x-ray will not be needed (line 51).By doing so, D1 shows he is in sympathy with the father.Secondly, by resorting to the first person plural he demonstrates that he identifies himself with the institution. 6To conclude, by explicitly inviting the interpreter to translate through a verbum dicendi, he addresses the interpreter directly as a full participant in the encounter.

Encounter B: an unshared diagnosis
The second encounter, taking place in the HTS, involves, as patient, a small child suffering from gingival infection and accompanied by his mother.They live in Austria yet are not German native speakers.D2 is not an Italian mother tongue speaker either.Five sequences can be identified, which embrace all the six ideal phases of a medical encounter: 1) opening, complaint and verbal examination-all included in the same sequence because of their concision (lines 1-40); 2) physical examination (lines 41-99); 3) diagnosis (lines 100-191); 4) treatment and advice (lines 192-340); 5) administrative procedures and clarifications aimed at dispelling the mother's doubts (lines 341-421).
Throughout the encounter, triadic exchanges prevail, with only a few exceptions.Worthy of notice in the first sequence is D2's control over the turn-taking, which confirms the enhanced conversational asymmetry of the complaint and verbal examination phases.At this stage, it is D2's prerogative to give the turn to the other interlocutors, by posing questions aimed at collecting relevant information for the diagnosis, and to take the turn back without displaying any feelings or making any comments.Owing to his institutional role and command over two linguistic codes (the technical one and ordinary speech), he may choose between an authoritarian and an empathic conversational style (Merlini, 2007, p. 439).In this specific case, he opts for a rather authoritarian style, which will lead to the user's distrust and will consequently be softened by the doctor to counterbalance the lack of confidence he created.Notably, the rapid sequence of questions, replies and respective renditions builds a triadic configuration based on cognitive expectations, where almost all utterances are elliptical, with no verbs.Only a few lines of this sequence are reported to give an overview of what is illustrated above.
Example (4) The third sequence (Example 6) starts with the doctor's diagnosis: the child has developed gingivitis and herpetic stomatitis due to teething (lines 114-126).Notably, D2 introduces the medical term by using a relative clause where he resorts to the first person plural (lines 114-115).The same structure, which is also repeated later, hints at his sense of belonging to the medical class and the use of the verb "ho visto" (line 122) suggests that his diagnosis is the result of careful analysis during the physical examination.This strategy might be justified by the need to dispel the perceived mother's mistrust.In I2's rendition only the relevant medical information is conveyed, without reference to either of the personal pronouns used by D2.
What is worth noticing is the translation of the sole technical term in the whole encounter (herpetic stomatitis).The term is introduced by the doctor (line 117), who endeavours to communicate directly with the mother in English, as is frequently the case throughout the whole consultation, yet with no success, since the mother does not speak English.I2 asks her whether she knows the meaning of the term and explains it (line 120), actively translating a term which might be too technical and hence difficult to be understood.As the mother displays her disagreement with the diagnosis (lines 154-165), I2 interestingly shifts from the brief dyadic exchange with the mother to a triadic configuration conveying the emotional content (affective expectations) of the mother's utterance (lines 166-172), as might be inferred from the adversative conjunctions and the colloquial expression (lines 166-167).In her rendition she reformulates what is illustrated by the mother, stressing her feelings and perceptions (lines 167, 169 and 172) and resorting to the indirect speech and a different person perspective.This is due to the potentially negative impact of the utterance and denotes the need to clarify the utterance source.
Example (6) At the beginning of the following sequence (Example 7 below), the treatment is illustrated in a triadic structure.D2 again maintains control over the turn-taking and the mother only intervenes through feedback markers.Utterances are rather short and immediately translated.
Remarkably, D2 stresses that the syrup will heal both the gum infection as well as the spots on the tongue, thus showing his endeavour to take into account the mother's perspective and concerns.He also resorts to the imperative form while explaining the treatment.No technical term is used and the indications provided are rather simple.Notably, I2 does not translate the mother's repetition that teething is not the cause of the problem.D2 takes the turn back to shift topic and start the administrative procedure.The mother looks quite disappointed when reading the prescription (i.e., Zovirax syrup), since Zovirax was also prescribed, as a cream, by the doctor at home.Structures are here mainly triadic, yet when this misunderstanding arises in lines 192-195, probably owing to language difficulties, I2 attempts to convey the emotional load in a couple of renditions, by stressing the mother's main concern (line 276).
This triggers the doctor's alignment with the interpreter in the attempt to reassure the mother, as might be observed in the frequent use of explicit invitations to translate (e.g., "tell her", which is repeated eleven times throughout this sequence).Worthy to note is D2's use of verba dicendi on two occasions (lines 204 and 303), which underline his need to convey such information to the mother.I2 adopts two different strategies in this respect.The first consists of the use of the indirect speech which signals detachment from assuming responsibility for the utterance, owing to the potentially negative impact of its content.The second strategy is a reformulation of the doctor's statement into a sort of echo-question (Ciliberti, 2009, p. 98), implying major involvement of the interlocutor who can reply and express her view, by confirming or denying the doctor's perception.I2 further softens the impact by resorting to a rhetorical device named litotes, an understatement generated by denying the opposite ("nicht komplett überzeugt", line 304) of the adjective used by D2 ("perplexed", line 303).The suspicion enshrined in the doctor's utterance is hence moderately conveyed.I2's mediation shortens the relation distance between D2 and the mother: her affective involvement is transmitted to D2 who actively participates and tries to establish a direct and unmediated relation with her.He addresses again the mother in English and uses feedback markers even when I2 is translating his utterances ("no no no no", "mhm", "capito?", "ecco", "andiamo").Furthermore, despite rejecting the mother's questions on other possible infection causes, he slightly mitigates the diagnosis (lines 318-319 and 334-336) In the final sequence, not reported as not strictly relevant for the purposes of this study, a more active participation of I2 is to be noticed since administrative information is involved.Attention should however be paid to a personal intervention of I2, who asks for clarification on the medication dose prescribed, showing her active translation role ("di Zovirax quantoparliamo di milligrammi", "Zovirax how much, we 7 are talking about milligrams").D2 replies that he was referring to millilitres.Thanks to I2's personal intervention and consequent rendition ("er hat Milliliter geschrieben nicht in Milligramm", "he has written millilitres not in milligrams"), the mother eventually realises that the doctor is referring to syrup and not to cream as she previously thought.It is, perhaps inadvertently, the interpreter's mediation that dispels the implicit misunderstanding, shortens the distance and promotes alignment between the initially conflicting views of the primary interlocutors.

Conclusions
The qualitative analysis of two mediated medical encounters, involving foreign tourists as patients in a Northern Italian hospital, has proved useful to detect the way medical terms circulate and the organisation of the interaction.This latter aspect has influenced the preferred "translation structure" adopted by interpreters, consequently favouring or else hindering interlocutors' active participation.The two medical interpreters have deployed specific strategies, which are summarised below, in terms of translation of medical terminology, register variation and promotion of interlocutors' participation.Being aware of the limitations of the qualitative approach to corpus analysis, which focuses on a limited set of data, this study might, however, be useful for promoting further research in these two directions so as to validate results obtained on a larger scale.Most encouragingly, however, the overall trends illustrated in this article are corroborated by numerous examples in the corpus of data previously collected (Pittarello, 2008(Pittarello, , 2009)).
The results obtained show that medical knowledge is explicitly mediated by the interpreter for the good of patients through specific choices and strategies, in the attempt to shorten the distance and soften potentially conflicting views between primary interlocutors.With respect to the two cases observed, ensuring that information is conveyed effectively and accurately seems to be the hub around which the interpreters' choices and strategies have implicitly or explicitly revolved.
A prevalence of triadic exchanges was noticed in the encounters selected.The shifts from dyadic to triadic sequences, which included affective expectations, allowed for fairer participation of primary interlocutors and enabled patients to express their perspectives and emotional load.
In the cases examined, contrary to what emerged in previous research (Bersani-Berselli, 2009a, 2009b), doctors frequently endeavoured to interact with patients directly.By doing so, they show more a shift in relational patterns rather than a loss of confidence in the interpreter's translation skills (Merlini, 2009, p. 83).Such attempts are signalled by the doctors' use of English as vehicular language (both encounters) and of feedback markers aimed at obtaining confirmation of patients' understanding (B).Doctors were ready to renounce specialised medical terms (almost absent in the cases observed) and even resort to gestures or lower the register in order to be better understood (A) and thus ensure the effectiveness of the information flow.
Although the medical language was rather simple, interpreters further explained the few technical terms, presumably due to their concern that patients might not fully understand them.It should be recalled that the vehicular language was used in both cases.
The conversational asymmetry (i.e., doctor's control over turn-taking and topic) was mainly registered when medical issues were dealt with.In such cases, when patients intervened or when a dyadic exchange took place between the patient and the interpreter, the doctor tried to regain conversational control so as to re-establish his conventional role.On these occasions (e.g., Encounter B), interpreters tended to counterbalance the asymmetry and mitigate patients' perceived mistrust.
Notably, utterances with a potentially negative content (B) have been reformulated by the interpreter in her renditions so as to soften the impact through different strategies (use of indirect speech, shift from affirmative to interrogative sentences and use of rhetorical devices).She thus acted as a "filter" between conflicting views and shortened the relation distance between primary interlocutors, promoting their active participation in the interaction.The result consisted of the doctor's alignment with the interpreter so as to reassure the patient/relative.
A conflict was also prompted by a misunderstanding due to the mother's misinterpretation (B).This contributed to her mistrust towards the doctor, but was unconsciously dispelled by the interpreter thanks to an autonomous intervention as "active translator".The interpreter also overcame the mother's lack of confidence by giving voice to her feelings (affective expectations) and thus acting as point of reference for her.The mother's perceptions were namely conveyed to the doctor, thus promoting interlocutors' active participation.
Interpreters' mediation has hence proved to aim at the above mentioned objective of ensuring the effectiveness of the communication flow, through different linguistic and interactional strategies.In both the cases analysed, interpreters displayed great solidarity with patients through active listening, feedback markers and numerous personal interventions.
The analysis suggested in this study confirms the high versatility of medical interpreters' roles and tasks as well as the numerous variables they have to deal with.Adequate support and proper training are essential in order to overcome the merging of linguistic, cultural as well as administrative tasks and to achieve the institutional tasks or ultimate objectives of medical interactions, whether they are mediated or not.
_____________________________ 1 Such tools include, among others, active listening, conveyance of information which takes into account the interlocutor's perspective on person and culture, feedback on the effects of one's own actions in terms of interlocutors' understanding, checking of interlocutors' perceptions and positions, etc. (Baraldi, 2009a, pp. 25-26, 2009b, p. 71 and73).
The HST is similar to an outpatients' clinic devoted to the handling of less severe cases. 3 For further details on the interpreting service provided in this healthcare unit, see Pittarello (2009, pp. 68-69). 4 This aspect might influence interpreters' renditions, especially in their use of medical terminology, which might be simplified or omitted owing to the interpreter's awareness of the patient's imperfect command of the vehicular language.The surrounding context nevertheless enabled the observer to detect possible reasons for such behaviours.Transcriptions follow, to a large extent, Atkinson and Heritage's (1984) graphical conventions.
The transcription of Encounter B is available in Pittarello (2008), whereas that of Encounter A is taken from the work of Sara Verdini, whom I warmly thank.The initials "D", "I", "P" and "U" refer respectively to: doctor, interpreter, patient and user, the latter term being hereafter used to indicate the person accompanying the patient to the encounter. 6 The same pronoun is also adopted by the interpreter in her rendition.The use of personal pronouns and specific address forms is useful to understand the alignment of interpreters with either of the parties.The third person singular mainly indicates detachment and intention to deny all responsibility for the utterance.The use of the first person, on the contrary, may suggest a cooperative attitude and the endeavour to share the responsibility about what is being said or, alternatively, it may express a strictly personal view and consequently the highest degree of autonomy and detachment from the original utterance.For further comments on the use of personal pronouns in the data collected, see Pittarello (2008Pittarello ( , 2009, pp. 80-84), pp. 80-84). 5

Table 1 :
Summary description of mediated medical encounters .