Managing affect in interpreter-mediated institutional talk: examples from the medical setting
Letizia Cirillo, University of Modena and Reggio Emilia and University of Siena
The present paper sets out to explore ways in which affectivity surfaces and is dealt with in interpreter-mediated doctor-patient talk. Section 1 briefly discusses previous work on doctor-patient talk with and without the mediation of an interpreter. Section 2 sheds some light on the concept of affect. Section 3 includes a description of the data analysed and the methodological approach adopted. Section 4 offers a close investigation of five interactional sequences containing affective cues. Finally, Section 5 draws some conclusions regarding the role of interpreters in managing affective displays within conventionalised forms of interaction.1
Affect, doctor-patient interaction, interpreting, institutionality, applied conversation analysis, English-Italian
1. Doctor-patient interaction, affect, interpreting
Doctor-patient interaction is one of the communicative situations in which the presence of an interpreter is increasingly required. Like all kinds of lay-professional encounters, it has often been described as an institutional type of interaction, displaying a certain degree of formality or conventionality and some recognisable features (see Levinson 1992). Focusing on institutionality, however, is a double-edged sword: indeed, in the vast literature on doctor-patient interaction (for a review see Cirillo 2005), the focus on the lay-professional relationship, while providing for a useful analytical categorisation, has often led to overemphasise the asymmetric character of doctor-patient encounters in terms of unequal participation, know-how, and (access to) knowledge (see Heritage 1997). This has in turn resulted in frequent premature categorisations, whereby doctor and patient have been identified as, respectively, the 'strong' and the 'weak' party within the dyad. Doctors and patients have even been described as speaking with two conflicting voices, which Mishler (1984) has labelled "the voice of medicine" and "the voice of the lifeworld," the former prevailing over the latter and conferring on the medical interview the status of a discourse type with specific features (e.g. doctors overwhelmingly asking questions, patients refraining from offering spontaneous elaborations of topics, etc.). As to the medical literature, most accounts have highlighted the role of doctors as objective professionals withholding expressions of involvement in response to patients' accounts, claims, etc., thus generally neglecting the issue of affect. A significant exception is some work in oncology and palliative care, which has seen affect in connection with emotionally challenging situations and delicate issues involved in the treatment of life-threatening illnesses (see Faulkner and Maguire 1994; Maguire and Pitceathly 2002, 2003; Kissane et al., forthcoming, among others). Being essentially practice-oriented and didactic in purpose, however, this work is concerned with providing healthcare practitioners and students with practical guidelines on how to deal with outcome variables like patient compliance and satisfaction, improving patient quality of life, and minimising stress and legal risks for doctors.
Similarly, studies on interpreter-mediated doctor-patient talk have not systematically investigated the affective dimension of the interaction, reflecting the traditional bias towards an alleged neutrality of interpreters, who, exactly like doctors, are trained to refrain from showing involvement of any kind (e.g. surprise, sympathy, encouragement, etc.). A substantial body of research has even failed to recognise interpreters as ratified participants in the interaction (not just in medical settings), and to acknowledge the "interpreting voice" (see Merlini and Favaron 2007) as a third, equally visible voice interpenetrating the voice of medicine and the voice of the lifeworld. And yet Wadensjö (1998) had already emphasised the coordinating aspect of interpreters' role, who, by virtue of their unique middle position and immediate access to "almost everything available to ears and eyes," have the hard task of "establishing, promoting and controlling connections between primary parties in conversation." (ibid.: 148). This task also includes managing the emotional character of interlocutors' talk by making the cues conveying it more or less accessible to co-participants, with the effect of either encouraging or inhibiting participants' mutual attention. Mutual attention is key in doctor-patient communication, especially with the emergence, starting from the 1980s, of a patient-centred approach. Ideally, such attention would imply patients reporting more than just their symptoms, and doctors paying special attention to patients' expressions of their concerns and expectations, sustaining in this way a collaborative relationship in which decision-making responsibilities are shared. Clearly, the presence of an interpreter in the medical encounter adds to the complexity and already delicate balance of this type of interaction, causing the borders between who is actually taking the responsibilities associated with the delivery and reception of healthcare to become blurry.
The role of interpreters within the medical encounter has been extensively examined by Davidson (2000) and Bolden (2000). Moving from the assumption that interpreters cannot be neutral (let alone invisible), as they have to bridge the gaps between different linguistic and cultural systems, and are themselves social agents co-constructing the meaning of the interaction in which they take part, both authors reach similar findings. In particular, they observe that interpreters edit patients' contributions, filtering out affective displays in order to make such contributions relevant to physicians' questions. In this way, they act as informational gatekeepers keeping the interview "on track" (see Davidson 2000: 400), ultimately sharing the physicians' normative, i.e. goal-oriented, tendency to collect as much objective, i.e. diagnostically relevant, information in the shortest possible time (see Bolden 2000: 414). In her extensive work on medical interpreting, Angelelli (2004) describes a rather more varied picture, in which interpreters become visible "by replacing one of the interlocutors, by aligning with the parties to channel information, by communicating affect, by exploring, by expanding or summarizing [...], and by controlling the flow of information." (ibid.: 132). Although Angelelli sees interpreters through different lenses and describes them using various metaphors depending on the situation (detectives, multi-purpose bridges, diamond connoisseurs, or miners; ibid.: 26-43; 129-132), she addresses the issue of affect management only sparingly.
The first researchers to have brought the affective dimension of interpreter-mediated medical consultations to the fore from an interactionist perspective are Baraldi and Gavioli (2007). In their data on mediated consultations with Arabic-speaking patients, the authors analyse instances of dyadic affective interaction between patients and interpreters. They find that interpreters engage in monolingual conversations with patients following the latter's emotional expressions, leaving out the healthcare provider and therefore hampering direct contact between this and the patient. While in line with Davidson's (2000) and Bolden's (2000) conclusions, Baraldi and Gavioli's findings present an important difference, in that the authors note that patients' affective contributions repeatedly project interpreters' affiliative responses. In a recent volume edited by Gavioli (2009) a number of contributions explicitly address the issue of affective communication, noting that triadic affective interactions are rather infrequent and the emotive involvement of all three parties is quite difficult to achieve. In particular, as observed by Zorzi and Gavioli (2009), when the third, momentarily left-out party is (re)involved in conversation, her/his contribution usually expresses alignment with the co-participants from a cognitive point of view, either rejecting or avoiding affective alignment.2 As we will see in Section 4, the data discussed in the present paper provide further evidence of this difficulty of managing a three-party affective communication within institutionalised forms of interaction like doctor-patient talk. Before moving to the description and analysis of our sample, however, let us take a closer look at what is intended by 'affect.'
2. Affectivity: Definition and contextualisation
The word 'affect' has been used by many different people to mean many different things, raising a wide range of theoretical and methodological issues in numerous research areas (e.g. developmental and social psychology, linguistic pragmatics, psycholinguistics, sociolinguistics, rhetoric), and causing a somewhat uncontrolled proliferation of a variety of blurry definitions and umbrella terms (e.g. "sensations," "emotions," "moods," "attitudes"). The complexity (and confusion) associated with the notion of affect lies in its being at the interface between inner states and external expressions, as it clearly emerges from Stern's (1985) idea of "attunement," and Wiener and Mehrabian's (1968) conceptual category of "immediacy"—to mention but two psychological accounts. To further complicate matters, the language resources conveying affect are various and variously classified (e.g. modality, deixis, emphasis, etc.). A categorisation of the (linguistic) phenomena comprised under the head 'affect' and a thorough discussion of the analytical approaches adopted to categorise them are beyond the scope of the present paper (for details see Caffi and Janney 1994 and Caffi 2007; for a classification of 'affect' see also Martin and White 2005). However, a few clarifications and caveats are in order to better outline the rationale for this study.
First, affect can be metaphorically seen as a scale, whose ends are a 'hot' and a 'cold' pole respectively. In other words, and moving from the folk psychological category of involvement (see Caffi and Janney 1994: 344ff.), it is possible to distinguish between 'more involved' and 'less involved.' This rough distinction is embraced and elaborated by Hübler (1987: 373), who argues that, to be analytically useful, the concept of involvement must be regarded as a continuum and therefore include both detachment and attachment as communicatively relevant modes. It is precisely as a continuum that I will consider affect in the discussion in Section 4. Second, in this paper a broad working definition of affect has been chosen, one that includes expressed feelings, attitudes, and relational orientations of all kinds (see Ochs 1989). However, I am mainly concerned with language as a vehicle for social action and will therefore move away from an individual psychological perspective, to focus instead on an interpersonal social perspective, where what really matters is not participants' sincerity or intentions, but the local negotiation of affect as displayed and oriented to by participants themselves throughout the interaction.
According to the view adopted here, affect is an attribute of sociality, and affective displays are "conventionalized ways of establishing rapport" (Tannen 1984: 371). Conventionality implies that the relationships existing between specific affective displays and specific interactional settings are analysable. In other words, the conventionality of affect goes hand in hand with the institutionality of interaction (see Section 1). This, in turn, means that affective displays may be described as either appropriate or inappropriate, and, as pointed out by Coulter (1986: 127; emphasis in original), "[t]ypes of situation are paradigmatically linked to the emotions they afford by convention," thus entailing that emotions "are not mere eruptions independent of appraisals and judgements, beliefs and conceptualizations." (ibid.: 126). In other words, emotions are normatively explicable, i.e. they are made contextually relevant by participants in the interaction. Emotions are actually visible through "contextualisation cues" (see Gumperz 1992), a body of verbal and nonverbal signs, including prosodic features (e.g. intonation and stress), paralinguistic indices (e.g. tempo, laughter, and hesitation), formulaic expressions (e.g. opening or closing routines), and extralinguistic behaviour (e.g. gesture). These cues are assigned context-bound meanings, and support speakers' foregrounding processes and listeners' inferential processes. Contextualisation cues are thus fundamental in order to interpret utterances in their particular locus of occurrence, i.e. to contextualise them, ultimately understanding what is going on in the interaction. It is by looking at these cues that I will explore how affectivity surfaces and is dealt with in interpreter-mediated doctor-patient talk.
3. Data and method
The examples discussed in the present paper are taken from a corpus of interpreter-mediated doctor-patient interactions recorded between 2004 and 2006 in hospitals and family support centres in the provinces of Modena and Reggio Emilia (in North-East Italy). The corpus includes 220 multilingual encounters involving speakers of Italian, English, Arabic, Chinese, Igbo, Urdu, Punjabi, and Hindi. For the purposes of the present paper, only the Italian-English subset was considered, which comprises 131 consultations (first visits, follow-ups, and routine discharge examinations). The length of consultation varies from less than five minutes to over one hour depending on the aim of the visit (from a simple prescription to an extensive examination). Since the data were mainly collected in obstetrics and gynaecology wards and family support centres/planning clinics (consultori in Italian), most patients are women and the issues discussed have to do mainly with women's reproductive health (e.g. contraception, pregnancy, voluntary abortion). The very few exceptions are exchanges involving young male outpatients seeking help for orthopaedic problems, respiratory tract infections, and other common pathologies often associated with occupational medicine. All patients belong to minority groups and use English as either their second language or a lingua franca, showing varying proficiency levels. Some of them also know Italian, although again with varying competence. Most patients come from West Africa and, in few cases, from either the Indian subcontinent or Southeast Asia. Healthcare providers are doctors (gynaecologists or other) and other staff (e.g. obstetricians, nurses, trainee doctors). They are native speakers of Italian, although some of them have some knowledge of English. The interpreters involved are three trained professionals, who have attended ad-hoc cultural mediation courses. Like many patients, they are from West Africa (one from Ghana and two from Nigeria), and have themselves experienced the process of immigration. The interactions were audio-recorded and subsequently transcribed using conversation analytical conventions (see Appendix) and inductive rationale.3 Out of the 131 consultations transcribed and analysed, five excerpts were selected for discussion (see Section 4). The extracts chosen are representative of the English-Italian subset in terms of type of visit, participants involved, type of sequences (dyadic vs. triadic), use of affective cues by primary parties and interpreters, and ways in which such cues are dealt with by co-participants, especially interpreters.
In sections 1 and 2, I have referred to the conversationalist, and more generally, interactionist perspective characterising this paper, and have restricted the rather fuzzy notion of affect to that of displayed emotionality, focusing on the effects this has on the ongoing interaction. In line with this approach, in analysing the data I have tried to identify affective displays that are significant for participants themselves. To do so I have applied what Sacks et al. (1974: 729) call "next turn proof procedure," i.e. I have closely inspected single turns at talk to see, on each occasion, how the current speaker is treating what has been uttered before. In Section 1 I have also subsumed doctor-patient talk (be it mediated or non-mediated) under the rubric "institutional interaction." In this respect, it is important to stress here that, consistently with the approach adopted by applied conversation analysis (see for instance ten Have 1995: 251; ten Have 1999: Chapter 8), institutionality itself is considered as "talked into being" (Heritage 1984: 290) rather than predetermined. In other words, roles and activities are 'co-constructed' by participants as the interaction unfolds.4 Therefore, to see how well (or badly) affective displays fit in with interpreter-mediated doctor-patient interaction—that is ultimately how affect is negotiated—I have looked at the same loci indicated by Heritage (1997) to probe the institutionality of talk, namely turn-taking organisation, sequence organisation, overall structural organisation, turn design, lexical choice, and interactional asymmetries. Unfortunately, the limited number of the interpreters involved in the recordings does not allow generalisations about the interactional organisation of mediated doctor-patient encounters, particularly regarding how affectivity is managed. Moreover, the fact that videos were not available for analysis has meant little or no access to non-verbal behaviours, including gaze and gesture, which are important cues of affective communication. Despite these limitations, some regularities are observable in the data, as we will see in what follows.
4. Managing affect in mediated medical encounters
The present section sets out to discuss some examples of dyadic and triadic affective communication in interpreter-mediated interactions between healthcare providers and patients. Throughout the discussion of the data various affective cues will be examined focusing on who produces them and how they are made relevant (or not made relevant) to the ensuing talk. Special attention will be paid to the ways in which effective cues are treated by interpreters.
Excerpt 1 is an example of dyadic affective interaction between patient and interpreter. Monolingual two-party conversations between patients and interpreters are rather frequent in the corpus. They often occur at the end of the medical encounter, as in this case, when the visit is over, the healthcare provider has either left the room or is engaged in other activities (such as filing charts), and the interpreter 'is left with' the patient to provide further clarifications or instructions (usually concerning bureaucratic procedures). Here the clinician is physically present, but the interpreter does not do anything to involve her in the affective interaction. In fact, the encounter terminates immediately after the reproduced conversation with a closing sequence in which the participants exchange routine thanks and final greetings. The excerpt is taken from an exchange in a neonatal ward, where the patient, who is HIV positive and whose newborn baby is a little underweight, is being given news on the conditions of her son and instructions on where to buy powdered milk to feed him once he is discharged from the hospital.
The patient has just received an exemption form to get the milk for free, but does not seem 'at ease' with it, and asks the interpreter for clarifications. In particular, she seems worried about the possibility of not obtaining the milk from the chemist's by simply showing the form (lines 1 and 4). The interpreter tries to reassure her (see especially line 13) and offers her support if problems arise (line 15). After the interpreter's initial answer to her question (see line 5), the patient utters a "change-of-state token" (Heritage 1984) followed by "okay" in line 6; however in subsequent lines she only provides minimal acknowledgement tokens (ll. 8, 12, 14, 16) and a continuer (l. 10) in response to the interpreter's expansions (ll. 7 and 9), request for confirmation (l. 11), and offer of help (l. 15). In line 17 the interpreter's "mh" uttered with lengthening of sound and rising intonation invites a stronger display of understanding and agreement with the solution proposed. However, the patient, instead of confirming understanding and uptake of the course of action projected by the interpreter, expresses further doubts (l. 18). The interpreter reassures her once again in line 19, but her contribution is followed by the patient's continuer in line 20 (note the suspended intonation) and a long conversational silence in line 21. At this point the interpreter elaborates on her previous answer (ll. 22-23), and the patient formulates the gist of her previously mentioned concerns (l. 25). She does so by using what Local and Walker (2008: 729) call a self-attribution of affectual state, which reveals the patient's fear of getting embarrassed.
Attributions of affectual states, can be considered as a specific type of formulations,5 in that they offer a candidate reading for what has been said or done before (see Heritage and Watson 1979, Heritage 1985). To be more precise, they make "something explicit that was previously implicit in the prior utterance," or make "inference about its presuppositions or implications" (Heritage 1985: 104). Here the initial laughter token and the 'smile voice' in which the formulation is produced seem to denote that either the patient may in fact already feel embarrassed in having to explain her concerns to the interpreter, or, since she has repeatedly requested clarifications in the previous turns, she is providing a justification for being so insistent. In any case, the self-attribution results in the interpreter explicitly reassuring the patient and reiterating her offer of help (l. 26). It should be noted here that this use of affective formulations (or, to be more precise, attributions of affectual states) seems to be an instance of what Caffi and Janney (1994: 328) call "emotive communication," i.e. the "intentional, strategic signalling of affective information in speech and writing (...) in order to influence partners' interpretations of situations and reach different goals." (ibid.). In this specific case the self-attribution of an affectual state seems to be instrumental in obtaining precise instructions and/or reassurance for not being inconvenient. Interestingly, affect is made relevant to possible practical problems related to the post-visit phase by both participants: it is not just voiced by the patient, but it is also subsequently taken into account and addressed by the interpreter.
1 P: a:h woul- will there be any problem (with those)?
2 I: no!
3 ((baby crying loud in the background))
4 P: with the pack or (something) written or you know:?=
5 I: =no >(slb slb slb slb) (coz they have)< stamp one you know?
6 P: ah okay.
7 I: mh be- before there was no stamp.
8 P: mh mh.
9 I: now they stamp.
10 P: mh mh,
11 I: you know the stamp?=
12 P: =mh.
13 I: if you take it there will be no problem.
14 P: ºo:kayº.
15 I: if there's any problem let me know.
16 P: o:kay.=
17 I: =m:h:?
18 P: and i just want with no stamp,
19 I: no problem.
20 P: mh,
22 I: no no this one is [(slb slb slb)]
23 P: [no but they] know it's from the ºhospitalº mh,
24 ((incomprehensible conversation for 5.8 sec))
25 -> P: .hh £i don't want to: be get embarrassed.£=
26 I: =no no no dont' worry. if there's any problem just let me know e:h?
Excerpt 2 also illustrates the concept of emotive communication (see above), but is rather different from excerpt 1. Here the affective sequence is triadic and includes an instance of other-attribution of affectual state (see Local and Walker 2008: 729). The exchange takes place at a family support centre for immigrants, where a patient is having a blood sample taken. She is upset and screams throughout the procedure. The nurse (O) invites her (in English) to look at the interpreter (l. 2) (and away from the needle), and the latter reinforces the invitation (l. 3), but the patient continues screaming (l. 4).
In line 5 the interpreter uses a formulation to make the presuppositions of the patient's screams explicit: she is scared. The interpreter does not translate anything the patient has said, but employs a "non-rendition" (Wadensjö 1998: 108) to attribute an affectual state to her, as if glossing what is happening.6 This attribution elicits the nurse's empathic response in line 6, which is also not translated by the interpreter. In line 7 the patient expresses her helplessness using the question "what shall we do here?," which is unlikely to be a genuine request for information, not simply because she has probably already had a blood sample taken on other occasions, but also because of the creaky tone and the louder volume of her voice. The interpreter, however, takes the patient's question as a request for instructions, and tells her not to move her hand, adding that if she does she will feel pain (ll. 8 and 11). The interpreter's response is possibly aimed at discouraging the patient from 'interfering' with the effective and efficient execution of the procedure, and thus projects a trajectory that is in line with the tasks being performed. In line 12, however, the nurse does not align with this new trajectory but reiterates the affective display the interpreter has not translated, and again the latter does not provide any translation (at least an understandable one). In line 15 the format of the patient's question invites agreement, but both the nurse and the student in charge of recording the encounter (S; see note 3) produce strong signs of disagreement in lines 16 and 17 (note that their dispreferred answer is not delayed by any mitigating device and is uttered with an animated tone; see Sacks 1987). At this point the interpreter (re-)affiliates with the patient by claiming agreement with her (l. 18), but attaches some laughter token to her claim, somehow disclaiming responsibility for what she has just said. The ensuing ambiguity about the authorship and seriousness of the utterance could explain the dry answer of the nurse in line 19, which takes the conversation back "on track," as also shown by the interpreter's reaction to the patient's nth scream (see lines 20 and 21).
Once the blood has been drawn, the patient is still complaining, this time about the quantity of the blood taken (l. 24). In the last few lines of the excerpt, while the 'smile' echo and the laughter by the student (ll. 25 and 27) seem to make fun of the patient's concern, who further expresses her annoyance and astonishment (l. 26), the voice of the interpreter is totally absent. She does not provide any further upshot or gloss for what has happened, maybe because she has been 'reproached' by the nurse (l. 19) or because the co-participants can understand each other rather well, as demonstrated by code-switching and short monolingual sequences like the one in lines 28-29. The result of this absence is a direct contact between the nurse and the patient (ll. 28-29), whereby the former addresses the latter directly for the first time (note the shift from "her" to "you") and explicitly reassures her (although it is not clear whether she is being ironic). The patient replies in Italian, presumably understanding what she has just been told, although not necessarily showing uptake of the nurse's empathic display; she will in fact keep saying that she has been taken too much blood (data not shown).
To sum up, in using another-attribution of affectual state the interpreter seems to be 'speaking for' the patient and trying to justify her behaviour to the nurse. This is also shown by a later contribution whereby the interpreter reinforces the patient's position ("of course it hurts"), although the subsequent laughter somehow discounts what she has just said. Overall, by resorting to an affective formulation, the interpreter may want the nurse to help her calm the patient down so that the visit can be carried out smoothly, thus making affect relevant to the goal of the encounter. Such a hypothesis may also partially explain why the interpreter does not translate the affiliative but apparently not goal-oriented response by the nurse. The translation, however, may also be deemed unnecessary by the interpreter, as the patient displays some active competence of Italian (see above), and is therefore likely to understand the nurse's response. The same competence may also account for the missing translation of the interpreter's autonomously produced formulation to the patient.
1 P: m.h! hu hu hu .hh
2 O: look at jamila!
3 I: look at me.
4 P: u:::h hu::! hu hu hu. e::h e::h!
5 -> I: ha tanta molta paura lei.
she's very very scared.
6 O: [ la ca pi sco, ]
I understand her,
7 P: [#(co me FACCIA]mo# [qui?)]
what shall we do here?
8 I: [ al ]lora don't don't move [ your hand.]
9 O: [ (slb slb) ]
10 (slb slb slb)?
11 I: if you don- if you move your hand [it's gonna hurt you.]
12 O: [ io la ca pi sco. ]
I understand her
13 £(slb slb slb slb slb)£ heh heh .hh.
14 I: (slb slb slb slb)
15 P: fa male no?
it hurts doesn't it?
16 O: no!=
17 S: =no!
18 -> I: no altroché se fa male! he he .he.
no of course it hurts!
19 O: ºma smettila via!º
come on stop it!
20 P: A:H!
21 I: don't move don't move.
22 ((38 lines omitted))
23 I: okay [press.]
24 P: [ m:: ]:h! eh troppo!
25 S: £troppo!£ [he he.]
26 P: [mam ma] mia!=
27 S: =he .h=
28 O: =no:! ne hai ancora tanto [ di san]gue stai tranquilla.
no! you still have a lot of blood don't worry.
29 P: [ sì:! ]
Like Excerpt 2, Excerpt 3 contains an example of other-attribution of affectual state. This time however it is the doctor, rather than the interpreter, who formulates the patient's affectual state. Excerpt 3 also includes non-renditions by the interpreter which, similarly to what happens in Excerpt 2, seem to alternately encourage and discourage direct contact between the doctor and the patient and engagement in a three-party affective sequence. The interaction takes place between a ten-week-pregnant patient, a gynaecologist and an interpreter at the beginning of a routine check-up.
The excerpt opens with an empathic formulation by the doctor, whose language echoes baby-talk (note the diminutive faccetta, lit. "little face," in line 1). The interpreter does not align with the doctor, providing a response which somehow 'discounts' the doctor's hypothesis on the patient's emotional state, and therefore the patient's concerns (l. 2). At the same time, her contribution is a non-rendition, which responds directly to the doctor's observation, without translating it for the patient, and therefore not giving her the opportunity to reply by herself. After a pause and a partially unclear stretch of talk, in which the patient presumably starts reporting on her health conditions and the interpreter starts translating (ll. 3-6), the doctor asks for clarifications (l.7). In line 8 the interpreter makes the doctor's request explicit, formulating a direct question to the patient, maybe in an attempt to (re-)involve her in the conversation, but the patient remains silent (l. 9). In lines 10-13 the interpreter explains the reasons why the patient feels unwell, making reference to the patient's job and elaborating her own account (note the adverb forse, "maybe"). The interpreter's candidate explanation—again a non-rendition—triggers a fairly long account on the part of the doctor (ll. 14-21), which the interpreter rephrases in a postponed translation to the patient (ll. 22-28). Both contributions are expressed with the voice of the lifeworld (see Section 1), i.e. both the interpreter and the doctor speak in a language that is closer to the patent's experiential knowledge than it is to the technicalities of the medical realm. Interestingly, while the doctor employs impersonal constructions in Italian (e.g. è facile sentirsi), the "you" employed by the interpreter in English does not seem to be generic and is repeated a number of times (note especially the "you know" inviting a response from the patient). Clearly, this does not imply that the doctor's style is detached,8 as shown by the initial formulation and by the very last line of the excerpt, where, despite moving back to 'business as usual,' she talks directly to the patient and even addresses her by her first name, as if trying to open up a space for direct contact. Unfortunately, since the patient's contributions are only minimal throughout the exchange, it is not clear whether or not the doctor's and the interpreter's attempts at establishing rapport with her are to any extent successful.
As mentioned above, here the affective communication is initiated by the doctor. Although her attribution of affectual state to the patient is not translated by the interpreter, it influences the trajectory of the ensuing interaction. This affective display is treated at a later stage by the interpreter, who addresses it because the doctor invites elaborations on the patient's state. Such invitation results in a voluntary non-rendition by the interpreter, which somehow 'compensates' for her brusque conclusion on the patient's condition (another non-rendition) after the doctor's initial affective formulation. The second non-rendition triggers in turn reassurance by the doctor via an explanation of how people usually feel in a pregnancy. This explanation finally makes a translation by the interpreter relevant.
1 -> D: ha una faccetta un po' preoccupata a dire il vero ma,
she looks a bit worried to tell the truth but,
2 -> I: no ma lei è sempre così.
no but she's always like that.
4 ?: hhh
5 P: (slb slb slb slb slb slb)
6 I: dice che non sta bene non si sente [bene.]
she says she is not she doesn't feel well.
7 D: [cioè?]
8 I: what do you feel?
9 (9.4) ((people talking loud in the background))
10 -> I: perché lei, devi sapere che lei è fa la parrucchiera.
coz she, you know she's a hairdresser.
11 D: mh,=
12 I: =e non riesce più a stare in piedi. si sente debole (.) spesso.
and cannot keep standing any more. she feels weak often.
13 stanca forse. hh a stare in piedi.
tired maybe. when she stands.
14 D: all'inizio della gravidanza, ((throat clearing)) i primi due tre mesi
at the beginning of pregnancy the first two three months
15 è facile sentirsi molto stanche anche se non c'è la pancia stanno
you're likely to feel very tired even if there's no belly
16 succedendo talmente tante cose dentro che è il periodo pi- più
so many things are happening inside that it's the most
17 impegnativo per il corpo.
difficult time for the body.
18 I: mh.
19 D: ed è normale sentirsi più stanchi.
and it's normal to feel tired.
20 I: [mh.]
21 D: [si ] abbassa anche un po' [la pressione.]
the pressure goes also down a bit.
22 I: [sh said tha::] at the beginning of the
23 pregnancy, you know, t's normal: that you feel we:- that you feel
26 I: and your pressure go:: down. you feel ve:ry off.
28 I: it's normal. (slb slb slb slb slb) you feel?
29 P: mh,=
30 D: =adesso jane ti dò gli esami del sangue da fare,
now Jane I'll give you some blood tests to do,
Excerpt 4 is taken from an interaction recorded at an orthopaedic practice during the examination of a young patient who has had his arm injured in a car accident. As in excerpt 3, it is the doctor, rather than the interpreter, who acts as a promoter of affective communication and tries to establish contact with the patient through the interpreter. However, the excerpt differs significantly from both example 3 and example 2, in that here the interpreter's "zero-renditions" (Wadensjö 1998: 108), i.e. the originals left untranslated, are not compensated by any non-rendition. In fact, the presence, or rather absence, of the interpreter in specific points of the interaction seems to hamper affective communication between the doctor and the patient. The excerpt can be divided into two parts. In the first, which stretches from line 1 to line 12, the interpreter does not translate an affective cue produced by the patient; whereas in the second (corresponding to lines 14-30), she does not translate an affective display produced by the doctor.
At the beginning of the excerpt, after the patient's complaint about the pain in his arm (l.1), the interpreter utters a continuer (l.2), which, however, is not followed by any expansion on the part of the patient but by a one-second pause (l.3). In line 4 the interpreter does not translate the immediately preceding complaint but summarises in Italian (for the doctor) some of the information that she has received from the patient in the waiting room prior to the visit. In line 5 the doctor inquires about the patient's job and his question gets translated by the interpreter in line 6. In lines 7 and 9 the patient answers the question and attaches an assessment to it ("very very hard!"). Assessments are often employed in everyday conversation as displays of alignment and affiliation. As stated by Goodwin and Goodwin (1992: 155), they involve taking up a position towards the event or entity being assessed and displaying the utterer's experience of that event, including her/his affective involvement in it. From a sequential point of view, assessments elicit responses from co-participants, typically displays of agreement and second assessments (see Jefferson 1978, Pomerantz 1984). In this case, the interpreter does show agreement and affiliation (see line 10), but, by omitting to translate the information provided by the patient and the assessment he attaches to it, she prevents affective communication between so-called 'primary parties' from happening; in particular, she excludes the doctor from the affect-sharing sequence which has just occurred between herself and the patient. Despite her zero-rendition and the communicative gap it may create, the doctor, who understands a little English, offers his understanding of the patient's contribution by rephrasing it in Italian (l. 11). In so doing he uses the word pesante, which follows the patient's "hard," functioning in fact as a second assessment (besides being its Italian equivalent). In addition, the marker "eh" uttered with a rising intonation at the end of the doctor's turn corresponds to a request for confirmation. This is provided by the interpreter in line 12, but not by the patient, although he understands a little Italian.
In the second part of the excerpt, the doctor is testing the functionality of the patient's hand by repeatedly asking him to open and close it and clench his fist (ll. 14-27). Seeing that he cannot hold it very tight, the interpreter empathically asks the patient if he feels much pain (l. 28), whereas the doctor accompanies his invitations with a humorous remark (l. 29), which the interpreter responds to with laughter (l. 30), but does not translate for the patient, who is left out of the dyadic affective sequence. Therefore, as with the assessment in line 9, the interpreter does align with her immediately preceding interlocutor, but leaves out, at least temporarily, the third party. Differently from what happens in the first part of the exchange, where the patient's assessment is matched with the doctor's second assessment despite the missing translation, here the untranslated phatic remark by the doctor is not responded to by the patient, and the analyst is left to wonder whether the latter has understood what the practitioner has just said. While in the first part the doctor's uptake makes the patient's affective contribution relevant—even to the goal of the visit9—and a delayed translation no longer relevant, it is as if the interpreter considered the doctor's joke in the second part of the exchange not important for the purposes of the medical encounter.
1 P: i feel very much pain.
2 I: u:h,
4 I: ha detto che è tornato a lavora:re,poi: è s[ta to solo d ue o: ]re,
he said he went back to work, then he stayed only two hours
5 D: [(e qua dove lavora)?]
and here where does he work?
6 I: what is your job (slb slb slb)?
7 P: metalmechanic [ at ]er::=
8 I: [mh,]
9 -> P: =castronovo. very [ ve ]ry hard!=
10 I: [mh!] =yea::h!=
11 D: =(quindi) comunque un lavoro pesa:nte eh?
so in any case hard work right?
12 I: sì sì sì.
yes yes yes.
13 ((12 lines omitted))
14 D: APRI E CHIUDI LA MA:NO:!
open and close your hand!
15 P: (apro)?
16 I: close it and open. close open.
18 D: STRINGI FO:RTE DA:I!=
hold it tight come on!
19 I: =close it.
20 D: STRINGI [FO:RTE!]
hold it tight!
21 I: [ he said] do it hard.
22 D: >FORTE FOR[TE FORTE FORTE!<]
tight tight tight tight!
23 I: [you cannot do it, ]
24 D: gli dica di stringere (il pugno).
tell him to hold his fist tight.
25 I: can you hol- hold it tight.
27 D: STRINGI:! FORTE:!
hold it tight!
28 I: a lot of ºpain eh?º
29 -> D: che non sono un datore di lavoro. [stringi.]
I'm not an employer. hold.
30 I: [eh he ]he he he noh: ho. .hhh
Similarly to what happens in Excerpt 4, in Excerpt 5 the interpreter refrains from translating affective displays (specifically assessments), which, in this case, are widely used by one of the healthcare providers taking part in the conversation. The patient is a young woman, who is visiting a family support centre with her newborn child and is being prescribed some tests for a menstrual block she has experienced for some time. Throughout the exchange the baby squeaks and chatters, virtually occupying turns of talk and triggering various reactions from the co-participants, especially the nurse (O) and the interpreter, as can be seen in lines 2-3, where they laugh at the baby's giggles. The nurse's inquiry about the baby's age in line 5 is met with the doctor's answer in line 6, followed by her request for confirmation and the doctor's affirmative answer (ll. 7-8). In line 9 the nurse initiates an affective sequence. She produces an assessment expressing surprise, which she reinforces in line 11, after the interpreter's minimal acknowledgement token in line 10. The interpreter expresses agreement with the preceding evaluation (l. 12), and the nurse shows further appreciation again using assessments (ll. 13 and 15). This time, however, she gets no response, apart from the baby, who continues giggling (ll. 14, 16 and 18), and whom she addresses directly with another assessment in line 17. After a long pause (l. 19), the doctor, who has been filling out the patient's chart, changes topic and resumes the interview to reassure the patient about her menstrual block—a contribution after which the interpreter resumes translating (data not shown).
In the affective sequence stretching from line 9 to line 17, the interpreter never translates the nurse's remarks, although she aligns with the trajectory projected by the nurse, at least at the beginning. The nurse and the interpreter somehow fill in the conversational gap occurring in the actual interview (as the doctor is leafing through some papers and compiling the patient's chart) to engage in affective communication. However, theirs is almost a side conversation, in which neither the doctor nor the patient get involved. It seems that the interpreter cannot either ignore affectivity displays or give up responding to them, but at the same time she somehow keeps affective and institutional communication separate. In this respect, example 5 is in line with Zorzi and Gavioli's (2009) claims, in that when the momentarily left out parties are re-involved, or in this case when two parallel conversations merge back into a single multi-party conversation, participants' contributions seem to converge on more visit-related topics, thus realizing a cognitive rather than affective alignment.
1 ((the baby giggles))
2 I: ha [ha ha ]
3 O: [mh mh] mh mh
5 O: quanto tempo ha:?
how old is he?
6 D: cinque mesi ha ºpiù o meno [di cia mo.º]
five months more or less let's say.
7 O: [ cinque me ]si?
8 D: ºsì.º
9 -> O: (ah se) è sveglio!
10 I: eh.=
11 O: =molto:.
12 I: sì sì. sì [ infatti. ]
yes yes. yes indeed.
13 -> O: [molto sve]glio! molto precoce.
very smart! very precocious.
14 ((the baby giggles))
15 -> O: poi sta dritto. sta seduto benissimo.
and he stays upright. he sits very well.
16 ((the baby giggles))
17 -> O: sei uno forte sei!
you're a strong one!
18 ((the baby giggles))
20 D: va bene lucy. quindi (.) stai assolutamente tranquilla perché il
okay lucy. then you can be absolutely sure because your
21 blocco della mestruazione non è una malattia. eh?
menstrual block is not a disease.
22 P: ºva bene.º
The excerpts discussed above exemplify various ways in which affect surfaces in mediated healthcare provider-patient interaction. In all examples affective cues of some kind are produced by the participants (patients, healthcare providers and interpreters alike). Overall, these cues do not appear as a "spontaneous, unintentional leakage or bursting out of emotion in speech" (Caffi and Janney 1994: 328), but are 'emotive,' rather than 'emotional,' in nature (ibid.), in that they seem to be closely connected to the attainment of the goals of the medical encounter. Regardless of the main 'supporter' or 'promoter' of affective communication being the interpreter (as in excerpts 1 and 2) or healthcare staff (as in excerpts 3, 4, and 5), affective displays seem to be only made interactionally relevant by participants—and thus also translated (or otherwise addressed) by the interpreter—when they may have practical consequences for the activities conducted during the visit or in the post-visit stage (compare excerpts 1-3 and the first part of excerpt 4 to excerpt 5 and the second part of excerpt 4).
5. Concluding remarks
The examples analysed in Section 4 show that, while affective communication is far from absent in interpreter-mediated doctor-patient interaction, the affective alignment of all three parties (healthcare provider(s), patient, and interpreter) is hard to achieve.
Overall, the conventionality dictated by routine medical procedures and the goal-oriented nature of the encounters on the one hand and the affectivity that is inherently linked to the interactants' need of establishing rapport on the other interpenetrate, exactly like the voice of medicine and the voice of the lifeworld. In this respect, the role of interpreters is crucial, in that they may guide the interaction by alternately favouring either voice. In particular, they may translate, not translate, or autonomously use affective cues, like affective formulations and assessments, thus either encouraging or inhibiting primary parties' involvement with each other. Such involvement, however, seems to be influenced by the institutionality of doctor-patient talk. In other words, affective communication is emotive (see above) in its being dependent upon the tasks performed, shaped by professional and organisational constraints, and associated with inferential frameworks as to what is appropriate to say and at what stage (see Levinson 1992). Although this hypothesis will need further investigation, it is possible to conclude that it is not just the relevance of affective cues that is negotiated by co-participants, but also the relevance of what needs to be translated is jointly decided.10
The examples presented also suggest that the use of evaluative language on the part of interpreters is not necessarily considered a taboo and the only censorship it is exposed to seems to be self-censorship. Clearly, this observation will have to be validated against a much larger and varied sample. What can definitely be gathered from the data, however, is that all participants in the interaction share responsibility for negotiating the meaning of affective cues locally. This also means that participants may independently react to their interlocutors' affective displays, by showing varying degrees of involvement (including understanding, sharing, and detachment), or even ignoring affective cues. As Wadensjö (1998: 148) puts it, "[a] primary party's need for the interpreter's assistance in understanding these kind of cues may vary." In other words, "[t]he interpreter is dependent on the interlocutors' interest in each other's emotions." (ibid.)
Sharing an interactionist approach to interpreting in institutional settings does not imply endorsing a light-hearted view, whereby interpreters can disclaim all responsibility for what happens between primary participants. On the contrary, a microanalytical approach to interpreter-mediated interactions, like the one presented here, may help interpreters themselves acquire a better understanding of delicate interactional mechanisms and equilibria, which are not predetermined and cannot therefore be governed solely by codes of conduct. Ultimately, an in-depth analysis of mediated talk can be an important resource for interpreter training by raising interpreters' awareness of their role and of the effects that their initiatives—be they translational or conversational—may have on the interactions in which they are involved.
- Anderson, Laurie (2009). "La commutazione di codice in incontri mediati in contesti legali e sociosanitari." Gavioli (2009), 259-298.
- Angelelli, Claudia (2004) Medical Interpreting and Cross-cultural Communication. London: Cambridge University Press.
- Atkinson, John M. & John C. Heritage (eds) (1984). Structures of Social Action. Studies in Conversation Analysis. Cambridge: Cambridge University Press.
- Baraldi, Claudio (2009). Dialogue in Intercultural Communities. Amsterdam/Philadelphia: Benjamins.
- Baraldi, Claudio & Laura Gavioli (2007). "Dialogue interpreting as intercultural mediation: An analysis in healthcare multicultural settings." Marion Grein & Edda Weigand (eds) (2007). Dialogue and Culture. Amsterdam/Philadelphia: Benjamins, 155-176.
- Bolden, Galina B. (2000). "Toward understanding practices of medical interpreting: Interpreters' involvement in history taking." Discourse Studies 2 (4), 387-418.
- Caffi, Claudia. (2007). Mitigation. Oxford: Elsevier.
- Caffi, Claudia & Richard W. Janney (1994). "Toward a pragmatics of emotive communication." Journal of Pragmatics 22 (3/4), 325-373.
- Ciliberti, Anna (2009). "Fenomeni di 'coinvolgimento' in incontri mediati medico-paziente." Gavioli (2009), 81-110.
- Cirillo, Letizia (2005), The Interactional Organisation of Talk in Naturopathic Interviews. PhD Thesis. University of Naples Federico II. [available at: http://www.fedoa.unina.it/880/1/Cirillo-Letizia_Tesi.pdf]
- Coulter, Jeff (1986). "Affect and social context: Emotion definition as a social task." Rom Harré (ed.) (1986). The Social Construction of Emotions. Oxford/New York: Blackwell, 120-134.
- Davidson, Brad (2000). "The interpreter as institutional gatekeeper: The social-linguistic role of interpreters in Spanish-English medical discourse." Journal of Sociolinguistics 4 (3), 379-405.
- Dijk, Teun van (ed.) (1985). Handbook of Discourse Analysis. New York: Academic Press, Vol 3.
- Drew, Paul D & John C. Heritage (eds) (1992). Talk at Work. Interaction in Institutional Settings. Cambridge: Cambridge University Press.
- Duranti, Alessandro & Charles Goodwin (eds) (1992). Rethinking Context. Cambridge: Cambridge University Press.
- Faulkner, Ann & Peter Maguire (1994). Talking to Cancer Patients and their Relatives. Oxford: Oxford University Press.
- Gavioli, Laura (2009). La mediazione linguistico-culturale: una prospettiva interazionista. Perugia: Guerra.
- Goodwin, Charles & Marjorie Harness Goodwin (1992). "Assessments and the construction of context." Duranti & Goodwin (1992), 147-190.
- Gumperz, John J. (1992). "Contextualization and understanding." Duranti & Goodwin (1992), 229-252.
- Have, Paul Ten (1995). "Medical ethnomethodology: An overview." Human Studies 18: 245-261.
- Have, Paul Ten (1999). Doing Conversation Analysis. London/Thousand Oaks (CA)/New Delhi: Sage.
- Heritage, John C. (1984). "A change-of-state token and aspects of its sequential placement." Atkinson & Heritage (1984), 299-345.
- Heritage, John C. (1985). "Analyzing news interviews: Aspects of the production of talk for an overhearing audience." Van Dijk (1985), 95-119.
- Heritage, John C. (1997). "Conversation analysis and institutional talk: Analysing data." David Silverman (ed.) (1997). Qualitative Research: Theory, Method and Practice. London: Sage, 161-82.
- Heritage, John C. & Douglas W. Maynard (eds) (2006). Communication in Medical Care. Interaction between Primary Care Physicians and Patients. Cambridge: Cambridge University Press.
- Heritage, John C. & Rodney D.Watson (1979). "Formulations as conversational objects." George Psathas (ed.) (1979). Everyday Language: Studies in Ethnomethodology. New York: Erlbaum, 123-161.
- Hutchby, Ian (2005). "Active listening: Formulations and the elicitation of feelings-talk in child counselling." Research on Language and Social Interaction 38 (3), 303-329.
- Hutchby, Ian (2007). The Discourse of Child Counselling. Amsterdam/Philadelphia: Benjamins.
- Hübler, Axel (1987). "Communication and expressivity." René Dirven and Vilém Fried (eds) (1987). Functionalism in Linguistics. Amsterdam/Philadelphia: Benjamins, 357-380
- Jefferson, Gail (1978). "Sequential aspects of storytelling in conversation." Shenkein (1978), 219-248.
- Kissane, David, Bultz, Barris, Butow, Phyllis & Ilora Finlay (eds) (forthcoming). Handbook of Communication in Oncology and Palliative Care. Oxford: Oxford University Press.
- Levinson, Stephen C. (1992). "Activity types and language." Drew & Heritage (1992), 66-100.
- Local, John & Gareth Walker (2008). "Stance and affect in conversation: On the interplay of sequential and phonetic resources." Text & Talk 28 (6), 723-747.
- Maguire, Peter & Carolyn Pitceathly (2002). "Key communication skills and how to acquire them". BMJ 325(7376), 697-700.
- Maguire, Peter & Carolyn Pitceathly (2003). "Managing the difficult consultation". Clinical Medicine 3(6), 532-537.
- Martin, James. R. & White, Peter R.R. (2005). The Language of Evaluation, Appraisal in English. London/New York: Palgrave Macmillan.
- Maynard, Douglas W. (1997). "The news delivery sequence: Bad news and good news in conversational interaction." Research on Language and Social Interaction 30(2), 93-130.
- Merlini, Rafaela & Roberta Favaron (2007). "Examining the voice of interpreting in speech pathology." Franz Pöchhacker & Miriam Shlesinger (eds) (2007). Healthcare Interpreting. Amsterdam/Philadelphia: Benjamins, 101-137.
- Mishler, Elliot G. (1984). The Discourse of Medicine: Dialectics of Medical Interviews. Norwood (NJ): Ablex.
- Ochs, Elinor (ed.) (1989). The Pragmatics of Affect. Berlin/Oxford: Mouton de Gruyter. Special Issue of Text 9 (1).
- Pomerantz, Anita. M. (1984). "Agreeing and disagreeing with assessments: Some features of preferred/dispreferred turn shapes." Atkinson & Heritage (1984), 57-101.
- Sacks, Harvey (1987). "On the preference for agreement and contiguity in sequences in conversation." Graham Button & John R.E, Lee (eds) (1987). Talk and Social Organization. Clevedon: Multilingual Matters, 54-69.
- Sacks, Harvey, Schegloff, Emanuel A & Gail Jefferson (1974). "A simplest systematics for the organization of turn-taking for conversation." Language 50, 696-735.
- Schenkein, James N. (ed.) (1978). Studies in the Organization of Conversational Interaction. New York: Academic Press.
- Stern, Daniel N. (1985). The Interpersonal World of the Infant. New York: Basic Books.
- Tannen, Deborah (1984). Conversational Style. Norwood (NJ): Ablex
- Wadensjö, Cecilia (1998). Interpreting as Interaction. London/New York: Longman.
- Wiener, Morton & Albert Mehrabian (1968). Language within Language: Immediacy, a Channel in Verbal Communication. New York: Appleton Century-Crofts.
- Zorzi, Daniela & Laura Gavioli (2009). "La gestione dell'affettività in incontri mediati da interprete." Gavioli (2009), 171- 201.
O other staff (nurse, obstetrician, etc.)
[ ] overlapping talk
( . ) time gap shorter than 0.2 seconds
(0.3) time gap in tenths of a second
wo- truncated word
: sound lengthening
. falling intonation
, rise-fall in intonation
? rising intonation
! fall-rise in intonation
¯ marked falling or rising intonational shift
<word> word uttered at a slower pace
>word< word uttered at a quicker pace
# creaky voice
£ smile voice
ºwordº word spoken more quietly
WORD word spoken more loudly
(word) reasonable guess at un unclear word
(slb slb) number of syllables in an unclear segment
((nodding)) non-verbal activity or transcriber's comments
-> phenomenon of interest
Letizia Cirillo graduated in translation and interpreting at the University of Bologna and received her PhD in ESP at The University of Naples Federico II. She worked as a research assistant at the University of Modena and Reggio Emilia and is currently research assistant at the University of Siena. Her research interests include conversation analysis applied to interpreted-mediated communication in institutional settings. She has also a longstanding experience as free-lance interpreter and university lecturer.
I would like to thank Laura Gavioli and Laurie Anderson for their precious comments on earlier drafts of this paper.
Return to this point in the text
Other relevant contributions in the volume are Ciliberti (2009), dealing with instances of 'emotive involvement' in interpreter-mediated doctor-patient talk, and Anderson (2009), dealing with code-switching as a resource for the management of affectivity in interpreter-mediated talk in both medical and legal settings.
Return to this point in the text
A first draft of the transcriptions was made by undergraduate students of languages, who were in charge of observing and recording the interactions, whereas subsequent drafts were prepared by researchers at the University of Modena and Reggio Emilia, including myself. References to people and places were replaced by fictitious names containing the same number of syllables as the originals.
Return to this point in the text
The co-constructive approach is best illustrated by some applied conversational contributions to the analysis of doctor-patient interaction contained in a recent volume edited by Heritage and Maynard (2006).
Return to this point in the text
Indeed, Baraldi and Gavioli (2007) use the phrase "affective formulations."
Return to this point in the text
"A 'non-rendition' is a 'text' which is analysable as an interpreter's initiative or response which does not correspond (as translation) to a prior 'original' utterance." (Wadensjo 1998: 108).
Return to this point in the text
The Italian bits in excerpts 2-5 are followed by an English translation. This, however, is not meant to be a gloss or literal translation, but a rough pragmatic equivalent of the original.
Return to this point in the text
The use of impersonal constructions may well be a way of further stressing the normality of the situation, thus contributing to reassure the patient.
Return to this point in the text
It may be of help to the doctor to know how hard the patient's job is to tell him precisely what he can or cannot do with his injured hand.
Return to this point in the text
Although a significant role may be played by participants' competence in the language(s) of their co-participants.
Return to this point in the text