The Routledge Handbook of Translation and Pragmatics

Pragmatics and agency in healthcare interpreting

Claudio Baraldi

Introduction

The concept of agency became first important in sociological theory. According to Giddens (1984), society is constituted by a combination of (1) social structures, which predefine the range of possible actions, and (2) agency as the contribution of individual actions to structural change. The concept of agency has also been used for interactional change (Van Langhenove & Harré, 1999), and the analysis of the interplay between agency and social structure may be applied to both interaction and society. Agency is defined as the choice of a specific course of action among various possible ones (Van Langhenove & Harré, 1999: 24; see also Giddens, 1984: 9). The availability of different choices for action can enhance social change in the interaction and/or in society.

Since the late nineties, a variety of studies on interpreting have stressed that interpreters’ availability of choices for action can change interpreter-mediated interactions (Wadensjö, 1998; Mason, 1999; Bolden, 2000; Davidson, 2000). These studies have highlighted the interpreters’ different ways of choosing actions in different social contexts, and the effects of their choices on interactions and participants. This chapter deals with these issues, focusing on healthcare interpreting.

While in some cases healthcare interpreting is provided “ad hoc” by healthcare professionals (Bridges etal., 2015; Meyer, 2012), it is most frequently provided by professional interpreters or, in some countries, by cultural mediators hired to deal with “cultural differences” between healthcare professionals and patients (Baraldi & Gavioli, 2012). Some important contributions have questioned the professional competence of cultural mediators (Pöchhacker, 2008) and the effectiveness of “mediated” interpreting (Hale, 2007). However, it is widely recognised that interpreters act as cultural mediators (Wadensjö, 1998; Davidson, 2000; Angelelli, 2004; Inghilleri, 2005; Pöchhacker, 2008) and, vice versa, cultural mediators act as interpreters (Pittarello, 2009; Baraldi & Gavioli, 2012, 2017a; Penn & Watermeyer, 2012). The relevant difference, therefore, is not between interpreters and mediators, but between the ways in which interpreting may be provided by either interpreters or mediators, in particular the ways in which interpreters or mediators exercise agency. In what follows, “interpreting” and “interpreter” will be therefore used as umbrella terms.

The adoption of the concept of agency is one specific case of a more general interest in sociological approaches developed in the past ten years in translation and interpreting studies (Wolf & Fukari, 2007; Angelelli, 2014; Tyulenev, 2014, 2016; Buzelin & Baraldi, 2016). The concept of agency is in line with a pragmatic approach which analyses the use of language “from the point of view of users, especially of the choices they make, the constraints they encounter in using language in social interactions and the effects their use of language has on other participants” (Crystal, 1985: 240). This chapter analyses the relevance of interpreters’ agency as shown through action, in the context of interpreter-mediated interactions (Mason, 2006) together with their contextual effects in the interaction (Carston, 2002, 2004). Interaction is seen as a site in which utterances and their effects are evidenced. In this sense, use of language in interactions, including interpreter mediated interactions, is considered as conditioned in society from both a pragmatic (Mey, 2001) and a sociological point of view. The actual ways in which interpreter-mediated interactions can be seen as conditioned in society have recently gained some interest in the literature (Buzelin & Baraldi, 2016; Baraldi, 2017) but it is beyond the scope of the present chapter to discuss them.

This chapter aims to clarify the ways in which interpreters’ agency affects and is affected by interpreter-mediated interactions involving healthcare professionals and patients. On the one hand, it analyses the ways in which interpreters’ utterances are relevant in these interactions; on the other hand, it analyses the ways in which the recipients treat interpreters’ utterances, thus showing their contextual effects in the interaction.

The next section analyses the ways in which the concept of agency is used to give meaning to interpreters’ activity. The third section focuses on interpreters’ ways of exercising agency in healthcare interactions. The fourth section focuses on the social construction of interpreters’ agency in healthcare interactions, as well as on the relation between agency and social structure. The third and fourth sections include several extracts from interpreter-mediated interactions showing the pragmatic aspects of interpreters’ agency. The final section draws some conclusions on the meanings that can be assigned to interpreters’ agency in healthcare settings.

1 The meaning of interpreters’ agency

Cecilia Wadensjö (1998) was the first to combine a sociological approach with the analysis of contextual effects of interpreters’ utterances in interactions. Drawing on the theory of participation framework (Goffman, 1981), Wadensjö suggests that interpreting in the interaction largely involves talk coordination. The latter concept, coordination, is particularly relevant for understanding interpreters’ choices for action, although it is not explicitly associated to the concept of agency. Coordination is: (1) explicit when interpreters enact non-renditions, e.g. by requesting clarification, complying with the turn-taking order, commenting on translations, inviting participants to start or continue talking; (2) implicit when interpreters either provide renditions of talk, which are summarised, expanded, reduced, provided in more than one turn (multi-part), or omit renditions (zero renditions).¹ In all these cases, interpreting shows a course of action among various possibilities.

Drawing on this type of analysis, and using Bourdieu’s sociological theory, Inghilleri (2005) defines the context of interpreting as a zone of uncertainty, in which interpreters can either reproduce existing power relations, by acting in favour of the institution, or create opportunities for improving understanding of migrants’ expressions. The concept of agency highlights “the potential for interpreters to exert equal or greater control over interpreting activity, even where this involves the disruption of pre-established power relations” (Inghilleri, 2005: 76). Inghilleri identifies several ways in which interpreters can exercise agency, based on their knowledge of migrants’ unclear utterances or false claims, institutional providers’ poorly phrased questions, incompetence and incorrect interpreting. When the interpreter becomes aware of this knowledge, s/he “potentially has the power to act on it” (ibid.: 81).

Tipton (2008a, b) draws on Wadensjö’s concept of coordination, exploring its meaning in terms of agency. In her view, interpreters are always competent social actors, implying that the distinction between interpreters’ agency and social structure cannot be “framed in terms of the cultural broker/language conduit dichotomy” (Tipton, 2008b: 5). However, in the social context, interpreting may be set as either coordination of the interaction or taken-for-granted language conduit. Therefore, the interpreter can make different choices turning into an active agent or playing a “more passive or a priori more impartial role” (Tipton 2008b: 12). Since interpreters’ choices of action can show their impartiality, the interpreter may be defined as an agent of neutrality (Tipton, 2008a). Interpreters’ agency must be recognised by the institutional providers; this need of recognition gives relevance to the social context of interpreting, which may fix the practices to which interpreters should commit and either jeopardise or reinforce interpreters’ commitment to these practices (Tipton, 2008a).

According to Tipton (2008b), interpreters’ agency is particularly important, in that it can create the opportunity to recognise migrants as social (or knowledgeable) agents. Mason and Wen (2012) confirm the importance of interpreters’ agency as empowerment of disadvantaged parties. Interpreters can adopt strategies “to enable a disadvantaged party to have better access to information, to take a turn to speak, to decide on their own to do or not to do something”, thus, interpreters may “bring change to the original network of power relations” (Mason & Wen, 2012: 125). Interpreters’ agency is based on the interplay of conversational moves of all participants in the interaction (Mason, 2009). By virtue of these conversational moves, all participants in interpreter-mediated interactions “position themselves and others and are, in turn, affected by each other’s positionings” (Mason, 2009: 56). Ways of positioning show the variety of ways in which agency can be exercised by interpreters. For instance, Mason (2006) shows that in order to promote the interlocutors’ understanding, the interpreter may provide contextual information to the institutional provider, when the migrant’s utterance is “underdetermined”, i.e., when it does not include such information. Mason draws the concept of underdeterminacy from the thesis that “the hearer has to undertake processes of pragmatic inference in order to work out [. . .] what proposition she [the speaker] is directly expressing” (Carston, 2002: 20). The interpreters’ inferences may thus be made explicit in rendering by displaying the utterances’ contextual assumptions, which are clear in the interaction but may not be as clear when rendering interactional items in another language, for another interlocutor.

In healthcare services, “the interpreter exercises agency and power, which materialize through different behaviors that alter the outcome of the interaction” (Angelelli, 2004: 10). Angelelli highlights several ways in which interpreters can exercise agency in healthcare interactions. They may (1) become co-participants and co-constructors of meanings; (2) set communication rules and control the information flow; (3) paraphrase or explain terms or concepts; (4) slide the message up and down the register scale; (5) filter information; (6) align with one of the parties; (7) replace one of the parties. Angelelli shows that interpreters’ agency can have either positive or negative consequences on interlocutors’ participation. Angelelli (2012) also shows that interpreters’ agency may bridge linguistic and cultural communities. For instance, she analyses the ways in which an interpreter translates the healthcare professional’s questions about the “degrees” of pain felt by the patient, adapting questioning based on the North American culture of precise measurement to the language based on the much less precise Latin American culture.

Angelelli makes clear that, in healthcare services, interpreters’ agency is associated to a great variety of positionings. Other studies highlight the various ways in which interpreters’ agency can be enacted, e.g. as advocacy of migrant patients (Greenhalgh etal., 2006), as linguistic, system, integration and community agency (Leanza etal., 2014), as cultural brokering, for instance by adapting the language of Western medicine to Zulu patients’ ways of expressing, by encouraging side conversations, adding details, simplifying jargon, and paying attention to the patient’s lifeworld (Penn & Watermeyer, 2012).

Summing up, in healthcare interactions: (1) interpreters both exercise and recognise agency, thus showing their positioning; (2) mutual positioning of interpreters, healthcare professionals and patients leads to recognition (expectation) of interpreters’ agency; (3) interpreters’ agency affects information flow and participants’ relationships in a variety of ways and empowers both parties; (4) interpreters’ agency means dis-aligning “from what is normally prescribed in the interpreters’ profession”, thus enhancing alternative actions (Baraldi & Gavioli 2015: 39); (5) interpreters’ agency may involve cultural brokering and bridging of cultural communities.

In interactions, exercising agency means designing turns of talk in specific ways. Turn design shows two types of speakers’ selection: (1) “the selection of the action which someone wants to accomplish”; and (2) the selection “among alternative ways of performing an action” (Heritage & Clayman, 2010: 46). These two types of selection show that speaker have a range of choices that are available to them. The very possibility of selecting and exercising selection shows their agency in the interaction. Interpreters’ turn design may or may not show their exercise of agency and may be either effective or non-effective for the healthcare interaction.

The meaning of both interpreters’ design of agency and its contextual effects is constructed in interpreter-mediated interactions highlighting a structure of mutual positioning (Mason, 2009) and corresponding mutual (or reflexive) expectations (Luhmann, 1995). The correspondence between positioning and expectations highlights the convergence between Interpreting Studies, on the one hand, and Social Psychology (Harré & Van Langhenove, 1999) and Sociology (Luhmann, 1995), on the other. The next sections include the analysis of several extracts from available research on healthcare interpreter-mediated interactions, showing both the design of interpreters’ agency and the structure of interactions.

2 Designing agency through non-renditions

According to Wadensjö (1998), explicit coordination is based on interpreters’ autonomous initiatives highlighting the availability of the choice of “non-rendering”, through actions, such as requests for clarification, comments on translations, requests to comply with the turn-taking order, invitations to start or continue talking (“non-renditions” in Wadensjö’s terms). The analysis of some extracts from interpreter-mediated interactions may help to understand what this means in terms of agency and positioning.

In extract 1 (Gavioli, 2012: 212–213), the interpreter selects minimal signals (mhmmh, mmh, ah okay, sì, ah, va bene) as a way of responding to the doctor’s ongoing explanation. The interpreter chooses to respond, rather than waiting for the doctor’s conclusion, and also chooses the type of response (minimal signals). Minimal signals are a simple and effective way to invite the co-participant to continue talking. They enhance a specific organisation of the interaction: the doctor is encouraged to continue to talk until the interpreter signals that the explanation can be translated (turn 14), and the patient is contingently excluded from the conversation.

Extract 1

The interpreter’s choice of responding and the way of doing so affects the interactional sequence. However, the interpreter’s positioning as a responder does not change the doctor’s design of turns and positioning, as the doctor continues to talk until his explanation is concluded. Therefore, the structure of positioning and expectations is confirmed by the interpreter’s minimal responses. The interpreter’s choice of responding does not change the structure of the medical interaction, i.e., her action does not display agency.

In extract 2 (Baraldi & Gavioli, 2008: 199; revised transcription), the interpreter guides the conversation on the patient’s future actions, while the professional is almost completely excluded from it (turns 2, 11). During this dyadic conversation, the interpreter chooses to introduce both the issue of birth control (turn 5), and the normative way of dealing with it (turns 9, 12, 15, 17, 18), thus displaying normative expectations about the patient’s behaviour. The patient reacts to the interpreter’s action with minimal responses: the interpreter’s action blocks the patient’s potential contribution to the definition of her condition of migrant, mother and wife. The interpreter’s contributions also highlight her “essentialist” positioning (Holliday, 2011) about “African” culture, defining Ghanaian men as not liking condoms (turn 15).

Extract 2

The interpreter’s actions display agency as they promote an important change of structure in the interaction: they replace the doctor’s actions and block the patient’s actions. By positioning the self as a guide, the interpreter instructs the patient on her needs and warns her about her behaviours. By blocking the patient’s participation and excluding the doctor, the interpreter’s agency has rather negative effects on the healthcare interaction.

In extract 3 (Angelelli, 2004: 94–95), the interpreter’s agency is made visible through the choice to expand the nurse’s invitation to ask for the patient’s chronic illnesses. The interpreter takes the nurse’s expression “all that” very seriously, starting an autonomous interview with the patient, in which he pursues details about diabetes (turn 2), high blood pressure (turn 4), heart disease (turn 6) and finally liver, kidney and stomach problems (turn 8).

Extract 3

In extract 3, as in extract 2, the interpreter starts a dyadic sequence with the patient, substituting the healthcare professional and thus changing the structure of positioning and expectations, but with a different effect: the interpreter’s agency enhances the patient’s opportunities to participate. The comparison between extracts 2 and 3 shows that interpreters’ agency in dyadic sequences with patients can have different effects on patients’ positioning.

In extract 4 (Baraldi & Gavioli, 2016: 45–46), the dyadic sequence is opened by the interpreter’s question in turn 4. This and the following questions clarify the patient’s answer to the doctor’s routine question “when was last menstruation”. The interpreter discovers that the patient’s initial answer refers to a date that is much earlier than a month before; her questions attempt to find out whether there is a problem of understanding or whether the patient is having a menstrual delay. In this dyadic sequence, the interpreter clarifies that the patient is having her period at the time of the examination and that she understood the doctor’s question as a request for the date of “the last menstruation before the current one” (turns 7–15). At the end of the sequence (turn 16), the interpreter provides a rendition for the doctor as an indirect but important answer to his question, thus re-involving the doctor in the interaction.

Extract 4

In extract 5 (Angelelli, 2004: 88–89), the interpreter translates the doctor’s questions (turns 1 and 4), then she takes the initiative of explaining to the patient what a TB test is and how it is performed. She does not involve the doctor’s in providing the answer to the patient’s question (turns 5–8); however, she answers the doctor’s second question and she reports to the doctor on what she has chosen to do (turn 9). The interpreter first enhances a dyadic phase of explicit coordination, replacing the doctor in constructing an informative interaction, then provides a final rendition in which not only does she perform implicit coordination but also clarifies the meaning of her explicit coordination.

Extract 5

Extracts 4 and 5 show two important aspects of interpreters’ agency. First, non-renditions provide important opportunities to change the structure of positioning and expectations, as interpreters encourage patients’ participation and clarifications. Second, through the final renditions for healthcare professionals, interpreters select what they observe as important information and the way of reporting on it, thus re-involving the healthcare professional and re-stablishing the structure of positioning and expectations.

3 Designing agency through renditions

Extracts 4 and 5 show that interpreters’ agency can be exercised through their renditions. Any type of rendition may be seen as a potential manifestation of agency. However, it is difficult to observe agency when interpreters rigorously comply with the normative instruction of providing close renditions. Deviations from close renditions are therefore the clearest manifestations of agency (Baraldi & Gavioli, 2015). These deviations manifest the interpreters’ ability to display their own understanding of interlocutors’ linguistic and cultural knowledge, an ability that is founded on the interpreters’ pragmatic competence within the interaction (Jazsczolt & Allan, 2012: 20). The frequency of these deviations is so high that the question then becomes what close renditions are (Baraldi & Gavioli, 2014). These deviations change the structure of positioning, as the interpreter replaces the healthcare professionals in selecting the relevant information.

Designing agency through renditions is a risky business. Davidson (2000) claims that interpreters can enhance the structure of gatekeeping, which is based on interpreters’ selection of the contents to render, according to interpreters’ own criteria of what is relevant in the context of medical interactions. Thus, the relevance of interpreters’ assumptions is shown through their contextual effects (Sperber & Wilson, 1995). Specifically, interpreters keep patients on track by only translating what they believe is relevant information for healthcare professionals. Gatekeeping highlights the positioning of interpreters as institutional insiders, aiming to meet healthcare professionals’ expectations. However, gatekeeping also increases “the patient’s difficulty in making herself, or her agenda for the discourse, heard” (Davidson, 2000: 381). Therefore, the structure of positioning and expectations is changed, and this change leads to serious problems of interaction between healthcare professionals and patients. Extract 6 (Davidson, 2000: 388–389) shows a case of gatekeeping: the interpreter provides neither a rendition of the doctor’s question in line 199, nor a rendition of patient’s lines 201 and 205, although the patient has invited him to translate.

Extract 6

Bolden (2000) shows that interpreters’ renditions of doctors’ history-taking questions display interpreters’ orientation to obtaining medically relevant information from patients and conveying it to doctors in an efficient way. Interpreters show their sharing of doctors’ normative expectations by selecting the patients’ answers they believe to be diagnostically relevant. In extract 7 (Bolden, 2000: 404–405), the patient mentions that nitroglycerine helps a bit, but it “squeezes her head” (line 14). The interpreter modifies that statement as “a little head pressure headache” (line 16), which fails to convey the patient’s sensation.

Extract 7

In extract 8 (Zorzi, 2012: 143–144), the interpreter’s selection concerns the affective side of the doctor’s positioning. The interpreter provides a reduced rendition (turn 8) of the doctor’s utterance, including the doctor’s intention to visit the patient and to prescribe the blood test, and excluding the doctor’s appreciation of the patient’s condition (turn 5: good, well done, splendid).

Extract 8

In extracts 6–8, interpreters’ agency is displayed through either zero renditions or reduced renditions, which change the structure of positioning and expectations by excluding contents of either patients’ or healthcare professionals’ utterances. While Tipton (2008b) highlights interpreters’ neutral agency, Inghilleri (2005) highlights interpreters’ agency as disruption of power relations, and Mason and Wen (2012) highlight interpreters’ empowerment of disadvantaged parties, extracts 6–8 show interpreters’ negative selection of both migrant patients’ and healthcare professionals’ utterances, highlighting the risk of designing agency through renditions.

However, designing agency through renditions can also have positive effects on patients’ and healthcare professionals’ participation. The following extracts show the positive effects of interpreters’ expanded renditions. In extract 9 (Baraldi 2012: 315), the interpreter expands the doctor’s short utterance concerning the first period of pregnancy, by providing more details about the “normality” of vomit, headache and “other problems like this”. This expanded rendition explicates the content of the doctor’s short utterance and provides contextual information (Mason, 2006) for the migrant, rather than for the healthcare professional.

Extract 9

In extract 10 (Baraldi, 2012, p. 314; revised transcription), the interpreter’s rendition concerns a doctor’s routine question about previous miscarriages or abortions the patient may have had. While the doctor’s question is quick and standardised, the interpreter’s rendition mitigates its directness by developing a question which shows sensitivity and attention to the patient’s perspective. The word “aborto” (meaning both miscarriage and abortion) is rendered with “any pregnancy that did not continue” and the interpreter’s rendition embeds the doctors’ question in the reassuring context of the will of God, which makes manifest to the patient and her husband (H) that the question does not have any worrying implications and that everything is fine with the current pregnancy. The rendition develops the topic in a way that is reassuring in the patient’s cultural and emotional perspective, thus combining information and affect.

Extract 10

To sum up, renditions show interpreters’ agency as either selections of information excluding patients’ and healthcare professionals’ voices, or summaries, explications and developments which give voice to them. Summaries, explications and developments can be considered as formulations, which have been described as interpretative contributions i.e., utterances that elaborate on what has been previously said. To be more precise, formulations find “a point in the prior utterance” and shift “its focus, redeveloping its gist, making something explicit that was previously implicit in the prior utterance, or by making inferences about its presuppositions or implications” (Heritage, 1985: 104). Interpreters’ formulations highlight structural changes in the interaction, giving meaning to the gist of other participants’ utterances (Baraldi & Gavioli, 2015). In extract 4 and 5, interpreters’ formulations are summarised renditions produced at the end of patient-interpreter dyadic sequences and successfully involving the healthcare professionals. In extracts 9 and 10, interpreters’ formulations are explications (extract 9) or developments (extract 10), which successfully expand on healthcare professionals’ explanations and questions for the patients. From a pragmatic point of view, these formulations may be considered as “explicatures” of professionals’ utterances, which are “much more specific and elaborated than the encoded meaning” of these utterances (Carston, 2004: 636).

4 Interpreters’ agency and social structures

The analysis of extracts 1–10 shows that interpreters’ agency is part and parcel of organised sequences of interpreter-mediated interactions, i.e., it is constructed through the chain of all participants’ actions. This analysis shows that interpreter’s agency is interdependent with patients’ unclear or confused turns (dyadic sequences and summarised renditions), doctors’ quick explanations and questions (expanded renditions), and also patients’ and professionals’ reduced participation (extracts 2, 6–8). This interdependence highlights the social structure of positioning and expectations of interpreter-mediated interactions. Specifically, it shows that interpreter-mediated interactions can take either the form of mediation, when interpreters’ agency is coordination of other participants’ agency, or an interpreter-centred form, when interpreters’ agency either blocks or excludes other participants’ agency.

The form of interpreting as mediation is particularly evident when interpreters’ agency is explicitly enhanced by other participants, for instance when healthcare professionals authorise interpreters’ initiatives, as in extract 3 (Can you ask her about chronic illnesses, diabetes . . . all that?) and in extract 10 (then ask her). These authorisations show the coordination of healthcare professionals’ and interpreters’ agency. In extract 11 (Gavioli, 2015: 173), the form of mediation is particularly evident. In turn 1, the doctor suggests that the patient takes a given medicine cautiously, introducing the recommendation with “diglielo” (tell her). The interpreter develops the doctor’s suggestion with a formulation: first, she tells the patient that the pills need to be taken carefully and in small quantities (turns 2, 4, 6); second, she reassures the patient that the doctors want to heal her (turn 6, “they want the stomach pain stop, and stop”).

Extract 11

The form of mediation is also evident when patients take initiatives that invite interpreters’ exercise of agency. In extract 12 (Baraldi & Gavioli, 2017b), the patient takes the initiative explaining and commenting on her experience of gases associated to her pregnancy (turns 1, 3, 5, 7, 9, 13, 16), inviting the interpreter to participate in the conversation (turn 1, you see the gases?). This initiative enhances the interpreter’s listening (turns 2, 4, 17), clarification requests (turns 4, 8) and direct response (turn 12). Then, the interpreter formulates a summarised rendition of this conversation, including a report of her response to the patient (turns 18, 20). Finally, the midwife (M) provides her “expert” answer (turns 19, 21).

Extract 12

To sum up, healthcare professionals’ and patients’ actions contribute to the construction of interpreters’ agency, co-constructing the interactional organisation of the interaction. The interdependence of participants’ turn design shows the structure of mutual positioning and expectations of interpreter-mediated interactions and the corresponding form of interpreting.

Concluding remarks

This chapter has explored the ways in which a combination of approaches, namely pragmatics, sociological theory and interpreting studies, contributes to the analysis of interpreters’ utterances in the context of interpreter-mediated interactions. On the one hand the contextual effects of interpreters’ agency were examined, on the other hand the effects of interactional structures on interpreters’ agency were looked at.

This chapter has shown that, in healthcare interactions, interpreters’ design of agency can either facilitate or block/exclude healthcare professionals’ and patients’ participation. The analysis of extracts from interpreter-mediated interactions gives a precise meaning to the zone of uncertainty (Inghilleri, 2005) in which interpreters can dis-align from existing structures of positioning and expectations, also showing their level of pragmatic competence in understanding their interlocutors’ linguistic and cultural knowledge.

As Tipton (2008b) and Mason (2009) suggest, the structure of positioning and expectations, shown through the participants’ conversational moves, is associated to interpreters’ agency. The analysis shows that interpreters’ agency may change the existing structure of positioning and expectations in different ways, depending on the specific situation.

Interpreters’ agency can empower both patients’ and healthcare professionals’ participation: (1) in dyadic sequences, by offering patients the opportunity to clarify their answers or initiatives; (2) through renditions phrased as formulations, by providing contextual information and enhancing the healthcare professionals’ opportunity to question, explain or decide how to go on. The coordination of equal distribution of agency among all participants is at the core of interpreting as mediation. The empowerment of disadvantaged patients seems to be inextricably intertwined with the empowerment of healthcare professionals; therefore, the idea that interpreters’ agency means “disruption of power” (Inghilleri, 2005) in healthcare settings may be questioned. Moreover, since the change of existing positioning and expectations is important, the concept of interpreters as agents of neutrality (Tipton, 2008a) may also be questioned.

By positioning as coordinating agents, interpreters position both healthcare professionals and patients as principal agents. Thus, paradoxically, interpreters’ agency both coordinates and is subordinated to their interlocutors’ agency. On the other hand, an interpreter-centred form of interaction, based on conversational moves consisting in non-renditions, reduced renditions and zero renditions, creates serious difficulties in healthcare interactions. Finally, while mediation as intercultural adaptation (extract 10) and essentialist interpreter-centred positioning (extract 2) may be important consequences of interpreters’ agency, this agency has the general function of changing personal and role positionings and expectations, rather than bridging cultural communities.

It is important to recognise that the interpreters’ design of agency in healthcare interactions highlights a variety of possible changes in the structure of mutual positioning and expectations, i.e., ultimately, different forms of interpreting and different consequences for the quality of healthcare provision.

Note

1 See Wadensjö (1998) for details on her taxonomy of interpreted renditions.