Where did the footnote go? How the change of mode in sight translation affects meaning-making
Randi Havnen, Oslo Metropolitan University
ABSTRACT
This study investigates how interpreters affect meaning making when translating a written text into speech. The analyses are based on Halliday’s theory of linguistic metafunctions from a multimodal perspective. The data are obtained from a strictly controlled design that isolated textual factors related to typical written resources other than language to explore how interpreters addressed them. By placing sight translation in Kress and van Leeuwen’s model of the four domains of meaning-making – discourse, design, production and distribution – the interpreter becomes a contributor as a re-producer and a re-distributor, adding more layers to the meaning-making process. Findings reveal that the interpreters vary in how they adapt the text to the spoken mode, and that their exploitation of semiotic resources affect the meaning potential. Although the interpreter´s code of ethics advises against altering the meaning potential, the findings show that such shift is unavoidable because of the modal shift. This study contributes with new insight about semiotic differences in written and spoken texts and shows that meaning potential related to all metafunctions changes through the sight translation process. This change is most evident at the interpersonal level in face-to-face encounters. These new insights should inform interpreters’ sight translation practice and training, as well as future research.
KEYWORDS
Sight translation, multimodality, shift of mode, interpreting, interaction.
1. Introduction
Sight translation, or interpreting from writing to speech, is a method that is frequently required as part of interpreters’ assignments in face-to-face interactions, for example, translation of reports after interviews, verdicts, decisions or orientation about rights (Čeňková 2015; Chen 2015). Existing research on sight translation does not reflect this practice; rather, studies are predominantly conducted on monologues analysing pedagogies, efforts, linguistic challenges and skills (Havnen 2019; Vargas-Urpi 2019). More recently, scholars have shown interest in exploring the multimodal and interactional aspects of sight translation (Felberg and Nilsen 2017; Vargas-Urpi 2019) and how the introduction of a written text influences the dialogical pattern (Davitti and Pasquandrea 2017; Defrancq and Verliefde 2018; Ticca and Traverso 2017). In Translation and Interpreting Studies, including translation between modes, there has been a movement toward including multimodal aspects in theories and methodologies (Boria and Tomalin 2020; Davitti 2019; Gonzalez 2014; O’Sullivan 2013; Tuominen et al. 2018). Surprisingly, the modal shift in sight translation and its effect on meaning-making have so far not been examined in empirical research (Havnen 2019).
The need for sight translation emerges when the intended reader of the written text does not have access to the document in the original language. A written translation might be time-consuming or expensive, so sight translation becomes the solution to overcoming the language barrier. This practice does not seem to consider the effect of the modal shift, possibly assuming that writing and speech create meaning in similar ways, as language is the dominant resource in both modes. Sight translation as a communicative practice does not have a monolingual counterpart. In a literate society, the expectation is that everyone can read, and a person is seldom in a situation in which something needs to be read aloud, if not for the sake of pleasure. One can think of this kind of translation as an adaptation for a person who has a temporal literacy impairment, hence related to accessibility issues.
From a multimodal perspective, mediating from one mode to another, or the very move from one meaning-making domain to another, is considered semiotic in itself. The mode shift affects the meaning potential. A typical example is a teacher’s execution of a syllabus in the classroom. However detailed the design is, the teacher’s choices affect what is foregrounded, hence altering the meaning-making (Kress and van Leeuwen 2001: 7). According to the interpreter code of ethics, an interpreter cannot alter the meaning. However, in practice, since the so-called cultural turn in translation studies in the early 1990s, translators’ discursive presence in translated texts have been well-document and shifts in meaning can occur when moving from one language to another. There is not “‘a pane of glass’ or a ‘black box’ through which ideas flow unchanged” (Rudvin 2006: 21). However, the ethical codes promoted by translation service providers advocate an illusion about equivalence, accuracy and neutrality (Lambert 2018).
In interpreting, including sight translation, time constraints in the production phase add to the limitations involved in any form of interlingual translation, oftentimes leading to loss of information (according to Gile’s effort model 2009). A study comparing bilingual speakers without interpreting training and educated interpreters’ performances show differences in their production. Thus, training helps to reduce loss; however, even the most educated never score 100%. This is expected, understandable and also accepted in interpreting assessments (Hale et al. 2018). Several studies on sight translation have documented problems in production that are related to interpreters’ competencies and/or to linguistic challenges (Agrifoglio 2004; Jiménez Ivars 2008; Lambert and Clas 2004; Sherve et al. 2010, 2011), but few studies have examined the meaning-making implications of the modal shift (Havnen 2019).
My motivation for analysing the modal shift in sight translation springs from my own experience as a schoolteacher for six years, an interpreter on and off for 26 years and an interpreter teacher for 15 years (amongst other experiences in teaching sight translation in the last four years). In this sense, I am what Gile (2018) refers to as a practisearcher. As a schoolteacher, I sometimes read aloud for my pupils and found myself explaining and explicating structural and referential information to keep the pupils’ attention. I wondered how interpreters facilitated written texts into speech when they were not as free to adapt the text as teachers are when they are reading. I saw students who focused exclusively on the verbal text when practicing, with little or no attention to, for example, paratextual resources or the listeners’ needs. Students oftentimes hid behind the document or buried themselves in the text, never looking up. Through a critical review of studies on sight translation, I reveal that little attention has been given to the modal mediation process and to listeners’ perspectives (Havnen 2019).
In this article, I address how interpreters treat typical written meaning-making resources other than language when translating a written text into speech. I discuss how interpreters’ renditions affect ideational, interpersonal and textual metafunctions when moving from one mode to another. Likewise, I briefly describe some listeners’ perspectives and argue for the usefulness of incorporating multimodal interactional perspectives into practice and training.
After introducing the multimodal approach I chose for these analyses, I describe the design I made through strict input control to give prominence to the modal aspects. The results section focuses on the communicative implications of the modal shift in sight translation; this is followed by some concluding remarks.
2. Theory
The multimodal approach in this article is based on the semiotic perspectives of Halliday’s systemic functional linguistics and the concept of metafunctions. Halliday (1978) describes the ideational, textual and interpersonal metafunctions of language use, a theory that Kress and van Leeuwen (1996) developed for visual modes and which is recently explored in the realm of touch as a communicative mode (Jewitt 2018). The ideational meaning in a mode, which relates to reality, is also considered as presentational or logical. The social relations between the signer and the sign are expressed through interpersonal modal resources; the ideational and interpersonal meanings draw on the textual metafunction – the organisational meaning (Jewitt 2009: 24).
It is accepted that all interactions are multimodal, and that no mode creates meaning alone (Norris 2004). Language should not be considered a separate communicative mode, as it has to be realised through speech or writing that exploits quite different semiotic resources in meaning-making – in other words, language does not make meaning alone. Kress (2010, 2020) argued that writing and speech must be considered as separate modes; otherwise, communicative implications might be ignored. From a multimodal perspective, not only do communicative modes have specific affordances, but they are also chosen because of their affordances (e.g. writing for its recursive possibilities and speech for immediacy).
Sight translation as a semiotic practice can be studied through the four-strata model involving discourse, design, production and distribution, as Kress and van Leeuwen (2001) described. Discourse in this context is understood as the socially constructed knowledge of some aspects of reality. Design is a realisation of discourse and is semiotic in the sense that it, for example, exploits a genre to realise meaning. Production refers to the material through which the communicative event is expressed. Distribution is how the text reaches the receiver – the interpreter of the articulation (Kress and van Leeuwen 2001:4-9). These configurations are to be understood as layers in the communicative practice that are neither hierarchically ordered nor linear. However, they are separate in the sense that they all contribute to meaning-making in different ways. In the case of sight translation in face-to-face encounters, the text is distributed as a written text to the interpreter, who re-produces the writing into speech for a listener in person. In other words, the final text has even more layers that are semiotic. I presume that the shift in the original distribution, and the added (re)production layer, contributes to a shift in the metafunctions of the start text.
Interpreters do not have the same freedom as teachers or public servants do when adapting a text for listeners. Interpreters are guided by their professional code of ethics, including guidelines about not adding, omitting or changing the text along with neutrality and loyalty claims (Lambert 2018). There is a source–target text continuum in translation practice, as shifts in translations are inevitable; the solutions will place themselves somewhere on the continuum of being oriented toward the source text, or the target text. Or, as Pym (2018) (and I) prefer, the start text and the target text. Another continuum is at play in sight translation: the written–spoken language continuum. Both continuums are relevant to the understanding of how sight translation should sound: as if a written text is read aloud or adapted to the spoken mode (Havnen 2019).
I set up a controlled experiment to analyse professional interpreters’ mediation of a written text into speech. In Section 3, I describe the design constituting the basis of my analyses.
3. Method
I designed the experimental setting as a role play scenario in a room that was set up like a typical public office. Another study using role play found that there were no differences in the results when the role plays were done in a real police station versus in a simulated interrogation room (Hale et al. 2018). In the office in the present scenario, a representative for the Norwegian Labour and Welfare Administration met with a Serbian-speaking client, and they needed an interpreter to communicate.
Experimental designs are common in interpreting research for several reasons, such as difficulties in accessing real situations and unpredictability in the occurrences of what one wants to study. Experiments have been criticised for having too little external validity and/or little ecological validity (Liu 2016). In sight translation, experiments are also criticised for a lack of controlling variables (Li 2014).
I designed the experiment executed in this study to address some of the above-mentioned challenges. I also wanted to shift focus from the study of sight translation of monologues in language labs, to sight translation of documents as part of face-to-face interaction.
All the participants signed informed consent forms. The public servant who helped with this work is employed in a social welfare office and has extensive experience talking to clients with and without an interpreter; she is also an experienced role player in practical exams for interpreters at Oslo Metropolitan University. It was important that she was confident and secure in the experiment to keep participants’ focus on the communicative event and so that she would be able to answer domain-specific questions that could arise in the interaction. Her instructions were to meet a client who had contacted the office to get some general information about sick leaves. At some point, she was to hand the interpreter an information sheet for translation. She was to play along with the situation in whichever way it might turn out, but she was not to interfere during the sight translation unless the interpreter or the client-initiated contact. According to students in courses on sight translation, my own experiences as an interpreter and as reported by Felberg (2015), it is common practice to leave a document with the interpreter and oftentimes exclude oneself from the interaction. The public servant was already in the office when the interpreter and client arrived, accompanied by me acting as the secretary. After each role play, the public servant wrote down her immediate reflections (100–150 words). Three speakers of Serbian with as little knowledge of Norwegian as possible were recruited amongst newly arrived working immigrants through cultural organisations. In other similar experimental settings (Felberg and Nilsen 2017; Hale et al. 2018; Vargas-Urpi 2019), and to the best of my knowledge, researchers have not used speakers who do not also know the other language as a variable. The interlocutors in this experiment had a real need for an interpreter in order to understand one another, and they had a realistic communicative goal. There were different listeners for each interpreter because if they had been the same, the listener would have gained knowledge and experience and would not have been able to listen without presumptions. Two of the participants were already employed in jobs in which there was no demand for Norwegian skills; however, they had not yet been in a situation where they needed knowledge of welfare issues. The third participant was on a beginner’s course in Norwegian. All of them were educated and had work experience. Their instructions were that they had initiated a meeting with the social welfare office to obtain information about rights and duties in the case of illness in an employment relationship. I told them that even though it was a role play, the public servant was a real professional and the information was true. An interpreter would be present for them to communicate – the participants had not previously used interpreter services. I encouraged them to react and act according to whatever came up in the situation, but I did not give them any examples of what that might be. I did not want them to turn their attention to any issue that could bias their behaviour. I interviewed them after the role play, and I audio recorded their interviews.
I recruited three interpreters in Serbian – labelled A, B and C – from my professional network. They are all interpreters with substantial experience; they all hold a state authorisation in interpreting and have masters’ degrees in the humanities. Further details about their background would immediately identify them, as the interpreter community in this language is small. One of my priorities for the experiment was that they had similar qualifications, that they were familiar with the interpreting settings and that they would not have problems with basic interpreting skills or language. This last factor was important so that any language issues would not overshadow the modal aspects. None of the interpreters had attended the course in sight translation offered at Oslo Metropolitan University, as the course was not yet offered for that language. The interpreters wrote some immediate reflections after they completed the role play (100–200 words). The participants’ reflections and my interviews with the clients did not play a significant role in the current analyses – I will analyse them more closely in a future study about attention and awareness in the interaction. In this study, the reflections are supplementary to the analyses of the translations. For the purpose of the experiment, I developed a text (Appendix 1) with 413 words containing different areas of interest (AOIs) (inspired by Shreve et al. 2010). These authors developed texts with predefined problems and measured problem areas by hesitation; their main concern was complex syntactic constructions. For my study, the AOIs were typical written structural resources (headlines, lists), graphic resources (punctuation, font style) and visual elements (logo, illustrations). In addition to the predefined areas, the material generated an issue related to deictic references; I added these as AOIs for analysis.
The text I developed for my experiments was a manifestation of the social welfare discourse in Norway. I designed it as information about regulations about sick leave and the need to disseminate such information. I produced written text to be distributed to people who were in need of it. It contained general information about the duties and rights of employees in the case of illness in a working relationship. I chose this theme for its relevance to people who were working immigrants; they would have some personal interest in the information and therefore be more likely to stay engaged, which they did. The content was realistic, as I took it from the social welfare office’s website. I printed and distributed the information on an information sheet for reading, and I included AOIs, such as illustrations and a footnote. To illustrate how the modal shift influences the text, the analysis is a comparison of two texts: the written start text and the spoken target text. I focus on how and if the ideational, interpersonal or representative metafunction is affected by the modal shift through the mediation of non- verbal elements in the texts. The written text consists of visual resources, such as illustrations, graphic resources, structure and language; the spoken text is aural and visual, including speech, gestures and gaze.
The interpreters were handed the document several minutes before the meeting so that they could familiarises themselves with the content and layout. They sat in a separate room from the person acting as the client. They had no prior knowledge of the public servant or the client. The participants were randomly paired. There were two cameras – one focusing on the interpreter and the listener (client), and one focusing on the interpreter and the public servant. I was not in the room.
No unexpected factors affected the experiment. The tight experiment design generated material that was both comparable and suitable for analysing the AOIs.
I watched the video recordings several times and transcribed renditions connected to the AOIs in a table. There is no existing typology for annotating renditions of, for example, graphic resources into speech. I have developed some categories inspired by Wadensjö (1998) and her descriptions of various rendition types. I have adapted these to describe the renditions of non-verbal resources. The categories are: not rendered, explicated as in the text, explicated through gesture, tone, binder or gaze and verbalising the semiotic meaning of the sign/reference.
4. Results
I conducted my analyses on the AOIs and not on the translation of the verbal text, even though the latter was the focus in the interpreter’s reflections they wrote after the experiment. The text did not constitute comprehension problems for the interpreters, and some of the challenges called for the use of translation strategies that affect meaning-making. Such strategies are thoroughly discussed in existing research (Agrifoglio 2004; Akbari 2017; Jiménez Ivars 2008; Lee et al. 2012; Şulha 2014). My interest is in how and if typical written resources are transferred and how the modal change influences meaning-making and the interaction.
Tables 1–4 show the renditions categorised as 0: not rendered; 1: explicated as in the text (i.e. parentheses as parentheses); 2: explicated through a) gesture, b) tone, c) binder or d) gaze; or 3: verbalising the semiotic meaning of the sign/reference.
Interpreter |
A |
B |
C |
Headline |
3 |
1 |
0 |
Headline |
2c (3) |
2c |
2c |
List |
0 |
0 |
0 |
Bullet point 1 |
0 |
2c |
0 |
Bullet point 2 |
0 |
2c |
0 |
Bullet point 3 |
2 b |
0 |
2 c |
Page turning |
3 |
2c |
2c |
New headline |
3 |
3 |
2c |
Numbered list |
3 |
0 |
2c |
End of list |
2c |
0 |
2c |
Table 1. Structural elements
Interpreter |
A |
B |
C |
«.» 1 highlighting |
3 |
1 |
0 |
«.» 2 highlighting |
0 |
0 |
2 |
Bold highlighting |
2b |
2b |
2d |
Parentheses |
3 |
1 |
3 |
Parentheses |
3 |
1 |
3 |
Parentheses |
3 |
1 |
3 |
Parentheses |
3 |
1 |
3 |
Italics |
2b |
0 |
0 |
Parenthesis (synonym) |
3 |
3 |
0 |
Underlined 1 |
0 |
0 |
0 |
Underlined 2 |
2b |
2b |
0 |
Table 2. Graphic semiotic resources
Interpreter |
A |
B |
C |
Logo |
0 |
0 |
0 |
Illustrations |
0 |
0 |
0 |
Table 3. Images
Some of the deictic references did not fit into the above mentioned categories, so they were written as they were rendered (Table 4).
Interpreters |
A |
B |
C |
Reference «you» |
Passive |
I |
I |
Reference «here» |
3 (in this brochure) |
1 |
1 |
Reference (single provider) |
Single mother |
Single father |
Single provider |
Footnote |
0 |
0 |
0 |
Web page |
2a + 1 |
1 |
3 |
See point |
N |
you |
we |
Page turning |
we |
we |
we |
Table 4. Deictic elements
There were 30 instances of zero renditions of semiotic resources belonging to the written mode – most of these were by Interpreter C. There were 20 instances indicated through gesture, gaze, verbal discourse markers or prosody, with verbal discourse markers being the most frequent. There were 17 instances of verbal explication, most frequent in Interpreter A, and there were 9 instances of explicating by describing the written resource (in parentheses). Interpreter B exhibited all these instances.
The most obvious finding in the material was that the three interpreters addressed the text differently. This was evident from the beginning in terms of how they presented the document to the listener; later, it was manifested in how much they adapted the text to the spoken mode and in their use of discourse markers. Interpreter A consistently used gaze and voice as a means of engaging the listener, whereas Interpreter B focused on the text but lifted his or her eyes up after finishing a sequence and used a discursive verbal sign to wrap up the paragraph. Interpreter C only looked at the listener three times and focused very intensely on the text, simultaneously gesturing a lot with one hand, which can be understood as a thinking gesture for the interpreter but can also function as a cohesive gesture for the listener because it is rhythmic.
4.1 Structural elements
Interpreter A was clearly oriented toward the listener, presenting the document by describing it as an information brochure about what to do in the case of illness. Interpreter C just read the headline out loud, whereas Interpreter B was in a middle position, saying, ‘the headline says’. Something similar occurred when the page had to be turned: Interpreter A said that they must now go to the next page and mentioned that it was about sick leave, whereas the two other interpreters just said that they would now turn to ‘sick leave’.
There were two different lists in the material, one with bullet points and one with numbers. None of the interpreters indicated that the first was a list or how long it was, but Interpreters A and C both signalled that the list had ended. Interpreter B explicated the first two bullet points as numbers and stated that the list ended by not articulating the last number. For the numbered list, Interpreter A presented the list as consisting of four measures, Interpreter B read without explicating and Interpreter C connected the first number or item to its referent (measure). Interpreters A and C indicated that the list ended, whereas B did not.
4.2 Graphic signs
Interpreter B rendered parentheses and inverted commas by saying ‘parentheses’ and ‘inverted commas’, whereas the two other interpreters explained functions, such as ‘so-called’ (i.e. ‘that means’). However, even when inverted commas were used twice in a row, the interpreters only rendered one. Their function was to emphasise two concepts. ‘So-called’ is frequently used by interpreters for culturally specific concepts, which one of these concepts were. In other instances, parentheses were mostly rendered through their function in the text, but Interpreter B articulated them mostly as parentheses. This is a practice that is common in police interviews when translating transcripts, as text within parentheses refers to the author’s comments, this is explained in the introduction to the report/transcript. Interpreter B did not explicate the parentheses when they indicated a synonym, mentioning both concepts. Interpreter C chose the synonym only.
There were two cases of underlining. The first one was underlining an aspect of duty, and the second one was underlining the goal for an employee to get back to work as soon as possible. These represent political guidelines in the social welfare system. None of the interpreters emphasised the first idea; Interpreters A and B indicated the second idea through their tone. They all emphasised numbers in bold letters.
A part of the text was written in italics; their function here was that there was an exception to the rule. Only Interpreter A signalled this by highlighting the word ‘if’ at the beginning of the sentence. The two other interpreters read this part in the same manner as they did the other parts of the text.
4.3 Ignored resources
Some resources were left out totally, such as the logo and illustrations. None of the interpreters showed the listener the paper. However, Interpreter A showed the client the hyperlink at the end of the text.
The final resource that they all ignored was the footnote. They did not focus on it when the reference number appeared in the text, nor when they reached the bottom of the page. The footnote said that the rules did not apply if one was freelancing or was self-employed.
4.4 Deictic elements
The deictic references proved to be particularly interesting, as I purposely did not put them in the material as an area of interest. They did not only appear naturally in the text, but they also entered the discourse in the explication of structural elements. One instance was a ‘here’ in the text, which Interpreter A described as ‘in this brochure’, and which Interpreters B and C described as ‘here’. Interpreter A connected the neutral single provider to a mother (the listener was female), Interpreter B to a father (the listener was male) and Interpreter C kept it neutral. The neutral ‘see the point about sick leave on the next page’ was rendered as neutral by Interpreter A, as ‘you’ by B and as ‘we’ by C. When turning the page, they all used ‘we go to’. The name of the document was ‘What shall you do in case of illness?’. Interpreter A rendered this as neutral and changed the question to ‘What to do . . .’, whereas the other two interpreters changed ‘you’ to ‘I’ for some reason.
Another difference was the reference to the hyperlink. Interpreter A said, ‘for additional information, you have a web page mentioned here (shows the document, points at it) nav dot no, with three double w’s before’. Interpreter B said that ‘you can find more information on the site www dot nav dot no’, explicating that it was a website. Interpreter C said, ‘you can find more information on NAV’s webpage’.
4.5 The participants’ perspectives
In the interview material, the listener to Interpreter A (7 min and 38 sec), the interpreter who explicated the most (in a pedagogic manner), mentioned that the information provided was sometimes difficult to follow, and that it was easy to mentally drift away. The listener felt like she had been ‘sent back to school’. The listener to Interpreter B’s rendition (7 min and 15 sec) commented that posing questions along the way was difficult, as after finishing, the listener had already forgotten the questions. Interpreter B was attentive but distant. Interestingly, Interpreter A commented that interrupting the interpreter with questions seemed to be difficult, and the public servant also commented on the reading feeling lengthy, challenging any listener’s attention. In the case of Interpreter C, who took 5 min and 26 seconds and hardly looked at the listener, the listener could recall more information than the others and had no questions about the text. The public servant, however, felt excluded. The two participants that engaged in further conversation with the public servant after the sight translation preferred this interaction to the reading part of the experiment. All of the listeners remembered that the first part of the information related to individual rights best, whereas the more technical part on the next page was remembered as ‘the list with four points’.
5. Discussion
The text used in this experiment is developed as information for ordinary people. The interpreters did not have problems with the translation process, but they still did not render many semiotic elements that affect meaning-making.
5.1 Metafunctions of the text
Some of the ideational metafunctions were expressed in the text via highlighting, such as an individual’s responsibility to get back to work as soon as possible; this was articulated verbally and through the use of underscoring and bold characters. When the interpreters ignored the semiotic function of these resources and did not render them, it contributed to a shift that weakened the authority of the text and placed a stronger focus on the informative aspect. Not referring to the footnote excluded important information; if a person qualified for this exception from the rule, the information given would have been misleading. Furthermore, not showing the illustrations and the logo might have minor consequences, as the illustration had a supplementary function and did not give additional information; however, not showing it deprived the listener of the visual support as part of meaning-making, supporting the informative and structural properties of the text. There might be cultural aspects to the perception of semiotic resources, such as the use of cartoons in governmental information leaflets. If one expects serious information to be primarily written, cartoons might have a different function. Cultural aspects of literacy need to be explored further.
Relevant for the textual metafunctions are cohesive resources. The layout in the document and the visual elements were reader-friendly because of their combination. The interpreters’ renditions indicated a shared mental representation of the structure, but the listeners did not have visual access to, for example, the length of the lists; as a result, they cannot know what to expect. This way of rendering is in line with Felberg and Nilsen’s (2017) findings – two of the interpreters in their study did signal the existence of bullet points by use of their fingers, but they did not indicate the structure by, for example, telling/showing listeners how long the list was. The third interpreter in their study ensured the listener could see the document the entire time and used a pen to show where she was in the text. The interpreters in Vargas-Urpi’s (2019) study did the same – the text in that study was a short table and might have seemed more logical to show. In training situations, listeners express that they prefer visual access.
One of the interpreters, A, explicated the headline through its function as a structural element. Another, B, articulated the headline as it was. The third, C, pointed out that it was a headline. The two latter solutions do not combine as well as the first one does with the spoken mode; they also make the written structure salient with no visual support. They presuppose that the listener is literate and understands the written references in the same way as the interpreter does. Modal literacy, however, is socially and culturally shaped (van Leeuwen 2005).
The interpersonal metafunction of the texts shifted in more ways because of the shift in attention from the public servant to the interpreter, through the interpreters’ use of the deictic ‘we’ and through discourse markers (gaze and prosody). This shift was less evident in Interpreter C’s rendition, which mainly focused on the texts informative content. Interpreter B was in a middle position, both staying very close to the start text and some of the written resources, but also engaging the listener through gaze and discourse markers. Deictic elements, such as ‘you’ and ‘here’, were probably confusing, as the distributed written text was read by someone who was not the intended perceiver. Not changing it might lead to confusion about the reference, as Felberg and Nilsen (2017) reported. Here, an interpreter pointed at the listener when reading ‘convicted’ to ensure understanding. In training contexts, it is necessary to discuss deictic references and their function in the written compared to the spoken mode.
All of the interpreters’ renditions led to a change in the metafunctions of the text in different directions. The interpreters seemed unaware of this choice; they were not consequent through the text, and their renditions gave the impression that they were a result of personal style and intuition. This is in line with Jiménez Ivars’ (2008) and Felberg and Nilsen’s (2017) findings when exploring interpreters’ strategies in sight translation. I argue, however, that it is vital that interpreters consider all meaning-making resources, and that they are aware of how they influence meaning-making when translating a written text into speech. The change of mode affects perception and interaction. Attention must also be given to the accessibility issue – the intended reader who becomes a listener.
5.2 Listening and understanding
The listeners remembered the first part of the document best. This might be related to the content – it focuses on individual rights, as opposed to obligations and different possible sick leave arrangements on the next page. However, it can also be related to attention span. Sherve et al.’s (2010) study showed a clear decline in the interpreter’s concentration after only one paragraph during sight translation. The listeners in my experiment might have been exposed to saturation. Two of them mentioned that concentrating and remembering were difficult. The subsequent problems can also be related to the text, as the second part of the brochure about obligations was more technical. The interpreters did not have problems with this part of the text, but they had specialised knowledge in public sector discourse, whereas the listeners did not. Consequently, the listeners’ attention might have waned because of a lack of pre-knowledge, which affected their sense of coherence and memory. In the case of the migrants from Serbia, they do have experience with social welfare, especially if they were raised in the former Yugoslavia. This might have had an effect on why it was easier for the listener of Interpreter C to remember; he had a longer history of work experience and was older than the two other listeners.
Texts subjected to sight translation are oftentimes not as easy to interpret or understand, nor are they as short as the present text was. Documents can be dense, contain specialist terminology and be very detailed and lengthy. The listeners in my experiment, who were only listening for 5–7 minutes, did find it difficult to concentrate. Even if an interpreter understands and has no problems interpreting the content, one has to ask oneself whether it is the most beneficial way to convey information, or if it is just a pragmatic way to overcome a language barrier. Arranging for a more dialogical approach that involves the interlocutors more actively by, for example, telling the listeners that they are allowed to ask questions along the reading, is possible.
Based on the experimental data in this study, making a connection between the interpreter’s behaviour and the listener’s response was not possible, as listeners’ preferences might be individual. What is interesting is that they do have reactions and preferences, and that the act of sight translation does have an impact on the interaction at several levels – as the very act of reading also seems to.
The act of reading – as opposed to the cognitive process of reading – is a kind of social interaction, a way of being socially present in the here and now, which places participants in quite specific webs of mutual obligations to others who are socially present. (Scollon 1998: 281)
In the case of sight translation, the intended reader is deprived of the control (and power) one has as a reader. The reader becomes a listener who must rely on the interpreter’s reading: an immediate spoken rendition and his or her own memory. Future research should explore, for example, whether creating a mental construct for the listener before sight translating is beneficial, as well as investigate ways of involving the interlocutors. When role playing in the course on sight translation offered at Oslo Metropolitan University, listeners appreciate receiving some structural information before the translation starts. During the reading, they also value involvement. The public servant in Interpreter C’s situation said that she would not prefer to use that interpreter, as she felt excluded from the interaction, although the listener was satisfied with the information he or she received. The communicative goal, if it was a dialogue, was not reached; if it was strictly informative, it was achieved, although stripped of the social morality. Additional cognitive studies involving diverse listeners and various text types and lengths need to be conducted.
5.3 Sight translation: A specialised translation method
Sight translation is often treated as a hybrid – a practice that is between translation and interpretation, although several scholars have argued that it needs to be taught separately as a method and not only as a pedagogical exercise (Sampaio 2007; Lee et al. 2012; Felberg and Nilsen 2017). A categorisation as a ‘between practice’ does not clarify many of the unique challenges of sight translation. Sight translation has many similar challenges to other intersemiotic translation practices, such as audiovisual translation, subtitling and transcriptions that demand some choices related to the mode (medium). Chaume (2018) discusses the concept of translation and various interlingual, intralingual and intersemiotic translation practices that are often categorised as adaptations across medias. He concludes that although there are different norms, expectations and levels of adaptations from a source to a target when moving meaning across languages and/or media, all the practices could be embraced by the concept of translation. That would mean revisiting translation and its relation to linguistic transfer, which is especially evident not only in expectations about legal texts in general (Hale 2015), but also in sight translation (Lee et al. 2012; Mikkelson and Willis 1993; Weber 1990). There is an underlying expectation that the communicative barrier is solely related to language. This leads to the idea about the translator as a mediator who does not act as a participant. Even in literature on multimodality, translators are described as ‘mediational means’ as opposed to social actors (Norris 2019: 38). I trust that Norris does actually acknowledge the translator as contributing to meaning-making, especially between modes; however, such descriptions support the notion of the machinelike translator. Ethical codes promoting equivalence and accuracy underpin these expectations. Lambert (2018) argues that this illusion is misleading and unethical.
It is an everyday reality for interpreters to encounter these expectations, which they might also have about themselves. I argue that interpreters are in the most obvious position to argue and explain the limitations and advantages connected to translation methods. Language and mode shifts bring about a shift in meaning-making, and consumers of services should be actively included in making knowledge-based choices about methods. In the public sector, both legal safety and participation are at stake (Havnen 2019).
Interpreters become specialists through training, and they should be encouraged to use their knowledge and take a more active part in choices of method, especially when it comes to sight translation because of its many pitfalls and interactional challenges. Nilsen and Havnen (2020) found that interpreters reported that they sometimes compromised quality because of time pressure, for fear of not being collaborative or for fear of being regarded as unqualified when arguing against sight-translating certain texts. Määttä (2015) reports the dilemmas surrounding legal safeguarding and language practices in the intersection between written and spoken language. Interpreters might not consider themselves to be the responsible parties in these practices; however, I argue that their specialist competence matters. Measures to educate service users should also be implemented at other levels.
In our program at Oslo Metropolitan University, discussions on how to decline a task, ask for preparation time and explain limitations and possible angles to sight translation form part of the training. This is in addition to multimodal analyses of various texts, practicing through role plays with a listener’s perspective and experimenting with different strategies with peers and in language labs (Nilsen and Havnen 2020). As codes of ethics are created on the basis of interpreting between speech, the problem of how to transfer them to sight translation requires a multimodal focus. The meaning-making resources are different, and modal adaptation is necessary to maintain the focus of the (temporarily illiterate) listener.
An interesting negotiation of translation norms occurs in the area of audio description (for the blind and visually impaired), in which neutrally describing visual information has been a norm, for example, by focusing on narratives in theatre plays. However, in post dramas, the narrative is not necessarily the salient resource, so other signs might be the salient semiotic resource that need to be described. The audio-visual description is suggested to be a part of production processes and not of post-production to make an adequate description of the salient signs (Roofthoot et al. 2018).
6. Concluding remarks
Interpreters need knowledge about meaning-making resources beyond just the language in written texts. They also need insight into the functions of the semiotic resources exploited in writing and speech. The interpreter’s reluctance to say out loud what is not verbal in the start text might have its source in the interpreter code of ethics and the idea of accuracy, which prohibits adding, omitting or changing information. Not attending to all semiotic resources, however, is also omitting. The metafictional shift when going from one mode to another is inevitable and must be approached consciously and knowledgeably.
Knowledge about communicative potentials and the constraints of translation across languages and modes is necessary to maximise the use of the most adequate methods and measures in particular situations. Interpreters need these insights to make good choices, sometimes making the challenges transparent for the users, such as by recommending other measures to convey information.
A controlled experiment with strict input control makes it possible to focus on predefined areas of interest. In this experiment, it was of great importance that language was not a hindrance for the interpreters, and that they were confident practitioners. The authenticity in the setting made the participants behave quite naturally. It would be interesting to use a longer text as well as a text containing more challenges. For example, the hyperlink inserted was short and common, and the illustrations were supplementary and would be more challenging if they displayed additional meaning. In future analyses based on the same experiment, I will consider the whole interaction. A weakness in the design for such analyses is the input control of the client actors. Two of them had similar backgrounds, but the third had a different socio-cultural background and was older than the others, which may have had an impact on pre-knowledge and expectations in the interactions. In a future experiment, I would also try to gain more information from the other participants as their reflections give valuable information when seen in relation to each other and to their common communicative experience.
Bibliography
- Agrifoglio, Marjorie (2004). “Sight translation and interpreting: A comparative analysis of constraints and failures.” Interpreting 6(1), 43-67.
- Akbari, Alireza (2017). “Needed skills and strategies to improve the efficiency of sight translation in classroom context.” trans-kom 10(1), 22-34.
- Boria, Monica and Tomalin, Marcus (2020). “Introduction.” Monica Boria, Ángeles Carreres, Maria Noriega-Sánchez and Marcus Tomalin (eds) (2020). Translation and Multimodality: Beyond Words. London & New York: Milton: Routledge, 1-23.
- Čeňková, Ivana (2015). “Sight interpreting/translation.” Franz Pöchhacker, Nadia Grbić, Peter Mead, and Robin Setton (eds) (2015). Routledge Encyclopedia of Interpreting Studies. London: Routledge, 144-153.
- Chaume, Frederic (2018). “Is audiovisual translation putting the concept of translation up against the ropes?” The Journal of Specialised Translation 30, 84-104
- Chen, Wallace (2015). “Sight translation.” Holly Mikkelson (ed.) (2015). The Routledge Handbook of Interpreting. London and New York: Routledge, 144-153.
- Davitti, Elena and Sergio Pasquandrea (2017). “Embodied participation: What multimodal analysis can tell us about interpreter-mediated encounters in pedagogical settings.” Journal of Pragmatics 107, 105-128.
- Davitti, Elena (2019). “Methodological exploration of interpreter-mediated interaction: Novel insights from mulimodal analysis.” Qualitative Resarch 19(1), 7-29.
- Defrancq, Bart and Sofie Verliefde (2018). “Interpreter-mediated drafting of written records in police interviews: A case study.” Target 30(2), 212-239.
- Felberg, Tatjana Radanovic (2015). “Utfordringer ved prima vista tolking i asylintervjuer.” FLEKS: Scandinavian Journal of Intercultural Theory and Practice 2(1), 1-17.
- Felberg, Tatjana Radanovic and Anne Birgitta Nilsen (2017). “Exploring semiotic resources in sight translation.” The Journal of Specialised Translation 28, 230-249.
- Gile, Daniel (2009). Basic Concepts and Models for Interpreter and Translator Training. Netherlands: John Benjamins.
- Gile, Daniel (2018). “Research into translation as a specialism: an analysis and recommendations.” The Journal of Specialised Translation 30, 23-39.
- Gonzalez, Luis Perez (2014). “Multimodality in translation and interpreting studies: Theoretical and methodological perspectives.” Sandra Bermann and Cathrine Porter (eds) (2014). A Companion to Translation Studies. Oxford: John Wiley and Sons, 119-131.
- Hale, Sandra (2015). “Approaching the bench: Teaching magistrates and judges how to work effectively with interpreters.” MonTI 7(1), 163-180.
- Hale, Sandra, Jane Goodman-Delahunty and Natalie Martschuk (2018). “Interpreter performace in police interviews. Differences between trained interpreters and untrained bilinguals.” The Interpreter and Translator Trainer 13(2), 107-131.
- Halliday, M. A. K. (1978). Language as Social Semiotic. The Social Interpretation of Language and Meaning. London: Edward Arnold.
- Havnen, Randi (2019). “Multimodal and interactional aspects of sight translation – A critical review.” FITISPos International Journal - Public Service Interpreting and Translation 6(1), 91-106.
- Jiménez Ivars, Amparo (2008). “Sight translation and written translation. A comparative analysis of causes of problems, strategies and translation errors within the PACTE translation competence model.” FORUM 6(2), 79-104.
- Jewitt, Carey (2009). “An introduction to multimodality.” Carey Jewitt (ed.) (2009). The Routledge Handbook of Multimodal Analysis. London: Routledge, 14-27.
- Jewitt, Carey (2018). “Towards a multimodal social semiotic agenda for touch.” Sumin Zhao, Emilia Djonov, Anders Björkvall, and Morten Boeriis (eds) (2018). Advancing Multimodal and Critical Discourse Studies. Interdisciplinary Research Inspired by Theo van Leeuwen’s Social Semiotics. London: Routledge, 79-93.
- Kress, Gunther and Theo van Leeuwen (1996). Reading Images: The Grammar of Visual Design. London: Routledge.
- Kress, Gunther and Theo van Leeuwen (2001). Multimodal Discourse: The Modes and Media of Contemporary Communication. London: Arnold Hodder.
- Kress, Gunther (2010). Multimodality : a social semiotic approach to contemporary communication. London: Routledge.
- Kress, Gunther (2020). “Transposing Meaning: Translation in a multimodal semiotic landscape.” Monica Boria, Ángeles Carreres, Maria Noriega-Sánchez and Marcus Tomalin (eds) (2020). Translation and Multimodality: Beyond Words. London & New York: Milton: Routledge, 24-48
- Lambert, Joseph (2018). “How ethical are codes of ethics? Using illusions of neutrality to sell translations.” The Journal of Specialised Translation 30, 269-290
- Lambert, Sylvie and André Clas (2004). “Shared attention during sight translation, sight interpretation and simultaneous interpretation.” Meta 49(2), 294-306.
- Lee, Jieun, Sylvie Vandaele and Georges Bastin (2012). “What skills do student interpreters need to learn in sight translation training?” Meta 57(3), 694-714.
- Li, Xiangdong (2014). “Sight translation as a topic in interpreting research: Progress, problems, and prospects.” Across Languages and Cultures 15(1), 67-89.
- Liu, Minhua (2016). “Putting the horse before the cart: Righting the experimental approach. Interpreting studies.” Claudio Bendazolli and Claudia Monacelli (eds) (2016). Adressing Methodological Challanges in Interpreting Studies Resarch. Newcastle upon Tyne: Cambridge Scholar Publishing, 87-105.
- Määttä, Simo K (2015) “Interpreting the discourse of reporting: The case of screening interviews with asylum seekers and police interviews in Finland.” The International Journal for Translation and Interpreting Research 7(3), 21-35
- Mikkelson, Holly and Jim Willis (1993). “Introduction to the Sight Translation Chapter.” The Interpreter's Edge: Generic Edition. Practical Excercises in Court Interpreting. Online: ACEBO, https://acebo.myshopify.com/pages/introduction-to-the-sight-translation-chapter-from-the-interpreters-edge-generic-edition.
- Nilsen, Anne Birgitta and Randi Havnen (2020) “Sight Translation: Best practices in Health Care and in Training.” Izabel E. T. de V. Souza and Effrossiny Fragkou (eds) (2020). Handbook of Resarch on Medical Interpreting: IGI Global, 375-396.
- Norris, Sigrid (2004). Analyzing Multimodal Interaction: A Methodological Framework. London: Routledge.
- Norris, Sigrid (2019) Systematically Working with Multimodal Data. Resarch Methods in Multimodal Discourse Analysis. Hoboken, NJ: John Wiley & Sons.
- O‘Sullivan, Carol (2013). “Introduction: Multimodality as challange and resource for translation.” The Journal of Specialised Translation 20, 2-14.
- Pym, Anthony (2018). “A typology of translation solutions.” The Journal of Specialised Translation 30, 40-65.
- Roofthooft, Hanne, Aline Remael and Luc Van de Dries(2018). “Audio description for (postdramatic) theater. Preparing the stage.” The Journal of Specialised Translation 30, 232-248.
- Rudvin, Mette (2006). “The cultural turn in community interpreting. A brief analysis of epistimological development in community interpreting literature in the light of paradigm changes in humanities.” Linguistica Antverpiensia, New Series: Themes in Translation Studies 5(1), 21-41.
- Sampaio, Glória Regina Loreto (2007). “Mastering sight translation skills.” Tradução & Comunicação: Revista Brasileira de Tradutores 16, 63-69.
- Scollon, Ron (1998). “Reading as social interaction: The empirical grounding of reading.” Semiotica 118(3/4), 281-294.
- Shreve, Gregory, Erik Angelone and Isabel Lacruz (2010). “Cognitive effort, syntactic disruption and visual interference in a ST task.” Gregory M. Shreve and Eric Angelone (eds) (2010). Translation and Cognition. Amsterdam: John Benjamins, 63-84.
- Shreve, Gregory, Isabel Lacruz and Erik Angelone (2011). “Sight translation and speech disfluency: Performance analysis as a window to cognitive translation processes.” Cecilia Alvstad, Adelina Hild, Elisabeth Tiselius, and Birgitta Englund Dimitrova (eds) (2011). Methods and Strategies of Process Research: Integrative Approaches in Translation Studies. Amsterdam: John Benjamins, 93-120.
- Şulha, Yrd Doç Dr Pelin (2014). “A study on the problems encountered in sight translation.” Celal Bayar Üniversitesi Sosyal Bilimler Dergisi 12(3), 1-13.
- Ticca, Anna Claudia and Véronique Traverso (2017). “Participation in bilingual interactions: Translating, interpreting and mediating documents in a French social centre.” Journal of Pragmatics 107(1), 129-146.
- Tuominen, Tiina, Catalina Jiménez Hurtado and Anna Ketola (2018). “Why methods matter: Approaching multimodality in translation resarch.” Linguistica Antverpiensia, New Series: Themes in Translation Studies 17(1), 1-17.
- van Leeuwen, Theo (2005). Introducing Social Semiotics. London and NewYork: Routledge.
- Vargas-Urpi, Mireia (2019). “Sight translation in public service interpreting: A dyadic or triadic exchange?” The Interpreter and Translator Trainer 13(2), 1-17.
- Wadensjö, Cecilia (1998). Interpreting as Interaction. London: Longman.
- Weber, Wilhelm K (1990). “The importance of sight translation in an interpreting training program.” David Bowen and Margareta Bowen (eds) (1990). Interpreting: Yesterday, Today, and Tomorrow. Amsterdam: John Benjamins, 44-52.
Biography
Randi Havnen is a PhD student and a lecturer in Interpreting Studies at the Department of International Studies and Interpreting, Oslo Metropolitan University. Her research interest is sight translation from a multimodal and interactional perspective, as well as the communicative and pedagogical aspects of interpreting. She is also a trained interpreter and an educated schoolteacher. She holds a master’s degree in Bosnian/Croatian/Serbian language, culture and literature.
Email: randi.havnen@oslomet.no