Simultaneous interpretation in interpreter-mediated remote legal proceedings: some observations from a forum theatre study 1

: This article concerns the findings of a study into participant responses to remote interpreter-mediated court proceedings using Welsh-English simultaneous interpretation. The study employs the forum theatre techniques of Augusto Boal, to explore how counsel, witnesses, mock-jurors and interpreters participate in hearings conducted remotely. The study was conducted in 2022, when the early difficulties of conducting hearings remotely because of the Covid pandemic had been resolved, and legal systems were considering how remote hearings could operate in the future


Introduction
This article is based on a study conducted in 2022 into the use of simultaneous interpreters in remote hearings.The Covid pandemic of 2020 meant that courts began conducting a number of hearings remotely (Grieshofer 2023: 768), but even prior to this, some hearings, e.g.taxation hearings, has been heard remotely (HMCTS 2021).Following the withdrawal of the Covid restrictions, remote hearings continue to be used for e.g.tribunal hearings (Jones 2023: 14), and hearings involving deported asylum seekers (Goodwin 2013), remand prisoners (Rowden and Wallace 2018) and witnesses outside the jurisdiction (Mondada 2011).In a sense, hearings involving vulnerable witnesses are also remote hearings, in that they may give video evidence from another location within the court building (Grieshofer 2023: 776).Accordingly, questions persist regarding the use of remote hearings (Pearce 2023), one aspect of which is interpreter-mediated proceedings.This article explores the findings of our study, and makes recommendations regarding the safeguards for fairness when remote proceedings are conducted via an interpreter.

Purpose of the research 1.Context
In many ways, interpreter-mediated remote hearings offer considerable advantages (Roman et.al. 2023).A larger pool of interpreters may be sought from a wider geographical area.This may be beneficial if; a language is not widely-spoken within the locality of the court; or if the witness may be known to local interpreters; or where a language has a number of varieties, and there is a need to ensure that an interpreter speaks the same dialectical variety as the speaker (Pawlosky 1996).
Another advantage of remote hearings for proceedings involving an interpreter is that many video communications platforms offer the opportunity to set up an interpretation channel, which allows the listener to choose which channel they hear.
Nevertheless, conducting interpreter-mediated remote hearings also presents challenges (Grieshofer 2023).It is difficult for sign language interpreters to ensure that their head, arms, and torso are visible (Hughes, Hudgins, and McDougall 2004) so that the person for whom they are interpreting can understand them.It is also difficult for the interpreter to indicate if a person is speaking too quickly, or if they have not heard what the speaker has said.If the interpreter is interpreting for both questioner and respondent (as is common in officially bilingual settings, including Wales), it may be more difficult for the listeners to follow which person is speaking.
Where the interpreter has their camera switched off, it is more likely that the listeners will conflate the characteristics of the speaker with those of the interpreter.Previous research has demonstrated that aspects of speech style, including pitch, pace, politeness indicators, pauses and hesitations, articulateness (De Jongh 1991) all affect how a speaker is judged (Berk-Seligson 2017).Gender, and its impact on the weight of their voice, may also be relevant if the case concerns physical violence, as a softly-spoken interpreter may be perceived as being an unlikely perpetrator of an offence requiring bodily strength, simply because the listeners fail to recollect that the voice of the interpreter is not the voice of the alleged perpetrator.It is for these reasons that we sought to conduct an experiment into how remote hearings conducted using interpreters operate, and how the participants (witnesses, counsel, interpreters, and listeners) respond to interpreter-mediated remote hearings.

Scope of the research
Our specific context is that of Wales, where Welsh may be spoken by "any party, witness or other person who desires to use it" (Welsh Language Act 1993 s22), and the model of interpretation used is simultaneous interpretation (Judicial College 2023).For remote hearings, some video communications platforms allow a separate interpretation channel, and listeners can choose whether to listen to the original language, or to the interpretation.Although our specific focus is simultaneous interpretation, many of our findings are equally applicable to consecutive interpretation.
For the purposes of this study of Welsh-English bilingual remote hearings, we used the forum theatre techniques developed by Augusto Boal, in order to elicit how court-users perceive proceedings conducted remotely via an interpreter.

Method 2.1. Augusto Boal and Forum Theatre
Brazilian director Augusto Boal's work on forum theatre has been used in a number of contexts to explore assumptions about behaviour, and how these may be modified.Boal works extensively with marginalised groups, and his forum theatre techniques allow those people to articulate the constraints on their lives and behaviour.He has titled his work variously as theatre of the oppressed (Boal 2008) and legislative theatre (Boal 1998), thus underlining that his work aims to give people the opportunity to articulate their experiences, and to suggest changes.Boal explains that "the theatre of the oppressed is always seeking the transformation of society in the direction of the liberation of the oppressed" and that in this sense it is not simply to "interpret reality" but rather, "to transform it" (Boal 2006: 15).Accordingly, it can be used across all sectors and age groups to explore assumptions regarding behaviour and to seek to change them.It has therefore been used in theatre in education settings to explore issues affecting children, such as bullying (Lencastro et.al. 2020) and also as a legislative tool, in order to explore experientially what mischief new legislation might seek to ameliorate, and how that legislation might work in practice (Boal 1998: 8).
We consider that it is a particularly appropriate tool to use in relation to interpreter-mediated court proceedings because, in the real world, the court administration e.g.HMCTS in England and Wales arranges for an interpreter to be present, and the case otherwise proceeds as usual.There is little opportunity to articulate what the challenges and difficulties are.Counsel is rarely aware of the need to modify their style of cross-examination in order to accommodate the interpreter.Witnesses speak, but some of the listeners do not hear them, because they are listening to the interpreter.Listeners may be confused by having to switch between listening to some participants directly, and then to others via the interpreter.
Forum theatre gives the participants the opportunity to articulate how they experience interpreter-mediated proceedings.It is empowering, because it does not begin from the standpoint that an authoritative 'we' is right and knows what is best.The research is led by the ideas and the suggestions of the participants.In some instances, it can lead to the realisation that 'how things are done' is based on incorrect assumptions.In other cases, there may be a realisation that though the current approach is imperfect, wider factors may inhibit the viability of working in other ways.There may also be a realisation that modification for the benefit of one group disadvantages another group.It may also cause participants to realise the intended meaning of what is said is perceived completely differently by the listener.Allowing participants to articulate their experiences may lead to a realisation that there may be aspects that work well, provided that they are explained in advance.This awareness may therefore help to ensure that the process is as fair as possible, and that safeguards are built into the process to ensure this.
At the centre of the process is the joker.As this is a theatrical technique, the joker is often a theatrical director, who acts as a "cultural animator" for the process (Boal 1998) Its equivalent is the jester or the zany of classical theatre (Boal 2008: 135).The joker is neither comedic nor deceitful, but rather an exposer of truth.The joker assumes a Comparative Legilinguistics 2024/57 subversive rôle (Boal 2008: 169) by intervening in the scene, and by permitting the spect-actors, as Boal terms them (because they observe and participate) to alter existing structures of power.In court for example, counsel will be accustomed to controlling the flow of information, while the interpreter will be accustomed to being unobtrusive.Witnesses and jurors will expect that they are to follow the rules of the court.However, the joker's rôle is to be transgressive, and to shake these established foundations (Boal 2008: 135) by acting as the catalyst for the expression of the participants' ideas.The joker is not, however, a director of the proceedings in the conventional sense -they exist to facilitate, not to lead.
The process begins with a series of warm-up activities, which are aimed at fostering trust and collaboration.A scene is performed (in our study, the cross-examination of a witness by counsel in a staged remote court hearing) and then the participants are encouraged to identify and debate problems, and formulate alternatives and solutions.The aim is to explore what is possible or what may be possible, rather than to assume that the current approach is the only approach, or that it is the best approach.The scene is then replayed, incorporating the modifications.Boal explains: Forum Theatre is a reflection on reality and a rehearsal for future action.In the present, we re-live the past to create the future.The spect-actor comes on stage and rehearses what it might be possible to do in real life.Sometimes the solution to the spect-actors' problems depends on themselves, on their own individual desire, their own efforts-but, equally, sometimes the oppression is actually rooted within the law.In the latter case, to bring about the desired change would require a transformation or redrafting of the law: legislation.How could that be done?There ends the power of the theatre.(Boal 1998: 8).
In this sense, as Boal (2006: 15) explains , forum theatre is concerned with the question 'what if' because within the theatre anything is possible for the purposes of explanation, irrespective of what external constraints may be imposed.Thus Boal (2008: 135) claims that forum theatre "is the rehearsal of the revolution" because it allows new solutions to be explored within the privacy of a theatrical space.

The study
In 2022, we devised a study using Boal's techniques.Two variables were incorporated; interpreter-mediated cross-examination conducted where both questioner and respondent spoke Welsh, and interpretermediated cross-examination conducted where the questions were given in English and the answers provided in Welsh.In order to maintain consistency across the different iterations of the scene, we developed a transcript of a cross-examination from a recent case, with names and locations being changed in order to preserve the parties' anonymity.The document was translated, in order to create a bilingual Welsh/English text.The case concerned a claimant who alleged that he had been injured by the negligent actions of a bus company employee.Crossexamination by counsel for the bus company was seeking to prove that the claimant's injuries were exaggerated, as he had not disclosed them to his doctor until some time after the incident, despite having several opportunities to do so.
A Welsh-speaking (male) barrister was engaged to conduct the cross-examination, and a (male) actor was engaged to play the rôle of the witness.Two (female) interpreters who are qualified to conduct Welsh-English/English-Welsh interpretation in legal proceedings were engaged to interpret the cross-examination.A public call was put out for mock-jurors, and following the classification of the respondents according to linguistic ability, twelve people were selected at random from those who volunteered to participate.Three of these were fluent Welsh-speakers, three others indicated that they were able to understand some Welsh.The others were not Welsh-speaking.
There were several aspects of the cross-examination that are significant when mediated by an interpreter namely; interruptions from counsel, unfinished sentences and repeated words, pauses and hesitations.The witness also becomes frustrated by counsel's apparent failure to understand the explanations he gives for his conduct.
All the speakers -the witness, counsel, and the interpreters were communicating remotely, and none of them were in the same location.The mock jurors were all in the same room in order to facilitate the discussion.Each juror used a laptop to listen to the evidence in their language of choice, but the speaker was also projected onto screens in the room in which the study was conducted.This meant that we were able to control to some extent what the participants saw, and ensure that even if they chose to minimise their screens, or view the proceedings as 'gallery' they were also able to see the same thing as each other.
Following the cross-examination, a theatre director asked the participants to comment on the experience, and to identify what was problematic, challenging, or confusing, and to suggest modifications, with the cross-examination -or sections of it, being repeated with the suggested modifications incorporated.

Findings 3.1. Language use
One of the objectives of our work was to explore how participants responded to interpreter-mediated evidence when both questioner and respondent spoke Welsh, compared with when the questions were given in English and the answers were given in Welsh.The monolingual jurors considered that it was easier to distinguish who was speaking when the questions from the male barrister were given in English, and the answers, interpreted by female interpreters were given in Welsh.The more limited peripheral cues, such as the fact that the sound was coming from the same source, and the more limited visual inputs regarding who was speaking (especially if participants were visible using 'gallery' view) meant that the identities of the questioner and the respondent were more readily distinguished, because they were speaking in different languages and with very different voices.
Nevertheless, the participants also recognised that the dialogue flowed more readily between the questioner and the respondent when they were both speaking the same language.There was therefore a greater understanding of the fact that what would make the process simpler for the listeners, would make the process more difficult for the speakers.

Conversational turn-taking
The importance of conversational turn-taking was emphasised, as is also demonstrated in the work of Brun and Taylor (2012).Where both the questioner and the respondent spoke the same language, the need to pause between each speaker was identified as being important in order to allow the listeners to distinguish between each one, particularly when they were being interpreted by the same interpreter, as is the usual practice in Welsh courts.The normal exchange of dialogue, and in particular the parry and riposte of cross-examination in person and without an interpreter, cannot occur in online hearings, and in hearings where there is an interpreter present.Again, the more limited visual and aural cues means that clearer turn-taking is required.
In bilingual cross-examination, clearer pauses between utterances are required, because the listeners are obliged to switch between the original language (for the questions) and the interpretation channel (for the responses).In order to ensure that nothing is missedparticularly the crucial distinction between an 'is' and an 'is not,' clearer pauses are of crucial importance.Although a simultaneous interpreter is generally able to keep a reasonable pace with the speaker, there will often be small delays as the interpreter will have to listen until the end of a sentence in order to ascertain its meaning because of variations in sentence structures between different languages.Furthermore, dialogue in remote hearings will also suffer from time delay because of computer processing speeds, and the demands placed upon it (Mollo 2006).Clearer breaks between each utterance are therefore imperative, and this is something the judge must explain to the speakers, and reiterate if necessary.

Number of interpreters
It is accepted that the usual practice in legal proceedings is for one interpreter to be engaged (Huws, Jewell and Binks 2022).However, the nature of remote interpreting is different.Fewer visual cues, the tendency of video to flatten facial expressiveness (Nadler 2020) and the intensity of the face on screen (Shockley et.al.2021)makes interpretation more challenging online.Technical aspects such as timedelay, buffering, and the possibility of technical problems make the experience of remote hearings more intense, and potentially more time consuming and in terms of health, it is recognised that screen-time causes several physical and mental strains (Moser-Mercer 2003).
We therefore considered that the norms experienced in inperson interpretation would not necessarily transpose to remote hearings, and that further and different safeguards may be needed in order to ensure that remote hearings are conducted appropriately.The additional cognitive burden on the interpreter also means that to presume that one interpreter is sufficient is not reasonable.We therefore experimented with using two interpreters, and did so with two objectives in mind: to explore the appropriateness of using two interpreters when both the questioner and the respondent are speaking the same language; and to ensure that each interpreter has appropriate rest periods (Edwards 1995:74).
In other linguistic contexts (i.e.where interpretation is needed in order to ensure a fair trial), where the questioner is likely to speak the State language, and the interpreter is only required to interpret for the respondent, the situation where an interpreter interprets for both questioner and respondent is unlikely to arise.However, in such cases, the interpreter will be required to interpret bidirectionally -into and out of the language of the courts, and again the need for multiple interpreters is significant.
The participants in our study considered it to be extremely beneficial to have two interpreters, one to interpret for each speaker, in same language cross examination, as it was easier to distinguish between the questioner and the respondent, because they could hear two distinct voices.
Nevertheless, when the interpreters took turns to interpret the witness's responses, and periodically exchanging which of them was undertaking the interpretation, the listeners found this confusing.We therefore consider that one single interpreter should be used for each discrete section of the process e.g.cross-examination of a single witness.Where the cross-examination is likely to be protracted, such as where a defendant or a key witness is cross-examined, there may be a need for counsel to identify the themes of the cross-examination to the interpreter, in order to identify where there are appropriate opportunities to switch between interpreters.Alternatively, it may be necessary to incorporate more breaks into the process in order to allow for the interpreters to be relieved.This will commonly occur in court proceedings, as the judge will need to identify when it is an appropriate juncture to allow a lunch break, or to bring the proceedings to an end for the day.However, its importance is even greater in remote hearings because of the additional cognitive and physical challenges of remote proceedings, and particularly significant in interpreter-mediated proceedings, because of the nature of the interpreter's rôle and its additional complexity.

Visibility of interpreter
We were advised by the interpreters that the common practice in remote hearings is that the interpreter will have their camera switched off (see also Licoppe and Veyrier (2017).Their experiences of conducting court proceedings and other meetings remotely suggested that speakers found it very distracting if they could see another person (i.e. the interpreter) speaking while they were speaking, even though they could not hear what the interpreter was saying because the interpretation is set to a separate communication channel.
We therefore experimented with this in order to explore the listeners' perspectives, and to evaluate whether it was easier to ensure that the listener did not conflate the characteristics of the interpreter with those of the speaker.Our earlier work (Huws, Jewell and Binks 2022) had suggested that where a female speaker was interpreted by a male interpreter, the listeners had failed to consider that the speaker was a woman, and it became apparent that the listeners, because they were hearing a male voice was tending to perceive the speaker as male.We were therefore concerned that this effect would be increased in remote hearings because there are fewer visual cues for the listeners, and that, with their cameras switched off, the interpreter is less 'present' in the situation.
We were surprised to discover that the responses to this experiment were very mixed.The interpreters stated that they felt very uncomfortable having the cameras on (Licoppe and Veyrier 2017), as they felt far more noticeable and visible than they would be in an inperson setting where they would not be the focus of the listeners' Comparative Legilinguistics 2024/57 attention.Some of the participants, particularly when two interpreters were used, found that having the interpreters' cameras switched on, made it easier for them to connect each speaker with their interpreter, and to consider that the interpreter was speaking on behalf of another person, and not on their own behalf.However, this depended very much on the arrangement of the tiles on the screen.Some of the participants had the interpreter located above or below the person for whom they were interpreting, and therefore it was easier to remember that they were paired, while the arrangement of other participants' tiles was diagonally configured, and this made it confusing to recollect which interpreter was interpreting for which speaker.Where there was only one interpreter, as would be standard in court proceedings, the participants did not perceive there to be any advantage in requiring the interpreter to have their camera switched on.

Replicating speech style
Another factor we were keen to explore in our study is the extent to which the interpreter does, and should, impersonate the speaker.Interpreters in our studies have expressed divergent views on this issue.Some argue (participants in Huws, Jewell and Binks' 2020 study) that they should speak neutrally but naturally, and not attempt to replicate the emotion of the speaker.They argue that it is for the listeners to evaluate what emotion the speaker is expressing.
Others (participants in Huws et. al. 2022 study) argue that to speak in a monotone would not make for a good interpretation.They argue that they must vary the intonation of their speech in order to make the evidence comprehensible.When conducting simultaneous interpretation, they will need to replicate the speaker's pace of delivery (which may convey their emotional state).What is less clear, and this may be a matter of individual practice, is whether the variation in intonation replicates that of the speaker, or whether it is the interpreter's own speech style (which they may vary in order to distinguish between the two speakers in same-language cross-examination situations).
In order to explore this, we experimented with having the listeners listen to a section of the cross-examination.Some listened to Interpreter A, who intentionally spoke in a monotone, and others listened to Interpreter B, who attempted to replicate the witness's emotions.During cross-examination, the witness became frustrated with the barrister's failure to comprehend his explanations of why he had not disclosed the injury sooner.Some participants found the monotone helpful, because they were able to focus more on the words spoken, divorced from their emotional context.Others found it confusing, because without any intonation, it was difficult to understand the words.Also, while they could see that the speaker was expressing an emotion, it was difficult to interpret what that emotion was when there were no aural cues from the speaker to assist them.It must also be borne in mind that when the interpreter utters 'no, it wasn't like that' without emotional context, the time delay and sentence-order differences between two languages may mean that the speaker's facial expression is not simultaneous with the interpreter's utterance, which again means that the emotion is difficult to read (Kaplan et al. 1995).
Where the interpreter was attempting to replicate the speaker's emotional intention, our findings were inconclusive, as the participants listening to Interpreter B, who was speaking in a natural tone, incorporating variations in intonation, could discern that this was natural speech, but it was not obvious to them, in an online setting, that Interpreter B was intentionally replicating the speech style of the witness.We propose therefore further research into this aspect, in order to explore a mock-jury's evaluation of interpreter-mediated evidence when the interpreter speaks naturally, but using their own linguistic cadences, and comparing this with when the interpreter attempts to impersonate the witness in terms of the emotion conveyed.

Limitations
We acknowledge that there are some limitations to the research.The focus was simultaneous interpretation, which is less widely used than consecutive interpretation in courts.However, many of the findings are applicable to consecutive interpretation, and to contexts other than the legal process where simultaneous interpretation is utilised.
We concede also that we would make some modifications to our experiment were we to repeat the study in the future.Although having the speakers connect remotely to the discussion was advantageous in many ways, it did pose some limitations for the purposes of the study.Firstly, unless they were directly asked about their experiences, the speakers felt remote from the discussion.In future, we would recommend that the speakers are located in a different room from the mock jurors, but within the same building, so that when the scene is being discussed, they are able to participate more effectively.Furthermore, the confines of the space meant that the remote participants were not always able to see and hear the discussion because of how microphones were placed.Some of them felt that this restricted their ability to participate fully in the discussion.Nevertheless, being mindful of these, and making ad hoc adjustments where possible (e.g.asking participants to switch microphones off in order to reduce echo and feedback from multiple machines), a meaningful discussion was possible, and a number of valuable findings were identified.
The methodology, as with focus groups, may also be subject to the influence of the more vocal members of the group, who may lead the discussion more forcefully than more passive participants, and it is important for the joker to focus the discussion, identify consensuses, establish priorities.Non-engagement is also likely to be a problem, and the joker must be careful to manage the person who is not participating (Barbour 2007: 82).
We also feel that ensuring that the professional participants are adequately prepared for the event is important.We were able to work with some of the participants in advance, and prepare them for the study, but those who had been unable to attend the briefing session were less familiar and comfortable with the experimental nature of the study.We recommend that in future, attending the briefing session will be compulsory, and that funding applications need to factor this in as an additional cost.
We also feel that a further limitation was the pool of interpreters available to participate.Our earlier study had explored matching the interpreter's sex with that of the witness.However, participant availability meant that we engaged only female interpreters, even though counsel, and the witness, were both males.For future research, we will ensure that there is a broader mix of male and female interpreters.
Nevertheless, we do not consider that these limitations have compromised the validity of the research.The Welsh context is a particularly valuable area for study because, its statutory status, and the quality assurance safeguards for interpreters, means that the high standard of interpretation can be assured -something that is lacking in other contexts (Grieshofer 2023), and therefore exploring how listeners respond to interpreter-mediated proceedings can be undertaken without concerns about the quality of interpretation being an obstacle.

Originality
The originality and value of this work is threefold.Firstly, as outlined above, the Welsh context is an exemplary, but under-researched context.Much research has focused on contexts where individuals do not speak the language of the State (e.g.Spanish in the US (Angermeyer 2015)) or where the language of the legal system is not spoken by most citizens (e.g.Chinese in Hong Kong (Ng 1998)) or where the interpretation is inadequate (Abu-Risha and Paramaswari 2021).The Welsh context is one where multiple participants -including the judge -may speak Welsh, and where the interpretation is of an extremely high standard, because interpreters must meet Cymdeithas Cyfieithwyr Cymru's standards for accreditation (Judicial College 2023).The legal status of Welsh is within a wider context where, because of the obligations under the Welsh Language (Wales) Measure 2011 and the Welsh Language Act 1993, bilingual services are provided much more widely in society, including the deliberations of the Senedd.Our originality is therefore that we explore this under-researched territory where bilingualism is normalised.
The second way in which this research is original is its exploration of remote hearings, and specifically the fact that our study was conducted when the initial problems had been identified and resolved, and the courts had found ways of working remotely that were generally effective, but at a time when the restrictions imposed in light of the Covid pandemic had receded, and consideration was being given to whether, and when the use of remote hearings should continue.
Our third claim to originality is the method.By allowing the participants to articulate how they experience the legal process, and to identify solutions and explore them, we were able to develop a better understanding of the assumptions people make, and to realise how something they accept without question may cause difficulties for other participants.We are therefore of the opinion that this is a very important study that widens the understanding of the use of interpretation in court hearings, and particularly in remote hearings conducted using simultaneous interpretation.

Recommendations
In light of this study, we therefore make the following recommendations.

More time needs to be allocated for interpretermediated remote hearings
It may be assumed that remote hearings can be conducted more quickly than in-person hearings.Delays may be averted because there is no need for the parties to travel long distances.Interpreter-mediated remote hearings are particularly challenging, and when scheduling remote hearings, consideration must be given to addressing this issue.Our study demonstrated that the limited visual cues, and the need for a greater degree of intense concentration on the part of both the interpreters and the listeners made the need for more frequent breaks imperative.Furthermore, our finding that there needs to be clearer pauses between utterances by different speakers (to allow listeners to switch between the original language and interpretation channels in bilingual cross-examination, and to allow listeners to distinguish between speakers in same-language cross-examination), means that this is another reason to justify an additional allocation of time for interpreter-mediated remote-hearings.

Two interpreters should be engaged
We also recommend that two interpreters are engaged when remote hearings are conducted.In same-language cross-examination, the more limited visual and aural cues make it difficult to distinguish between the speakers, and therefore being able to hear two different voices will make the process clearer for listeners.In Ó Cadhla v The Minister of Justice and Equality (Ó Cadhla v The Minister of Justice and Equality and Others 2019) the court concluded that because an interpreter speaks on behalf of a person, they cannot interpret for two people whose interests are necessarily in conflict e.g. a witness and a judge.We also recommend the use of two interpreters for remote hearings because of the additional cognitive burden on the interpreter.Even more than with in-person hearings, the need for more frequent breaks is imperative.

Explaining the process to the listeners
A gap that we identified in our study is the need for participants to understand more about the interpretation process and how it works.Our earlier work (Jewell, Huws, and Binks 2022) identified that although jurors in in-person hearings are shown a video explaining to them what is expected, this does not explain the operation of proceedings conducted via an interpreter.The interpreters explained therefore that in in-person proceedings, they often have to take it upon themselves to explain the process to court users.However, this is more difficult in remote hearings.
Where the judge is bilingual (which will be arranged, as far as it is reasonably practicable to do so (Ministry of Justice 2018)), they will not experience the situation encountered with same-language cross-examination, of both questioner and respondent speaking with the same voice i.e. that of the interpreter.As our study demonstrates, similar issues arise with interpreter-mediated remote hearings -there is a need to switch between the original language and the interpretation channels in bilingual cross-examination, and a need to distinguish between questioner and respondent in same-language cross examination.This is something that must therefore be explained to listeners in court proceedings.
The invisibility of court interpreters (Licoppe and Veyrier 2017), and the unobtrusiveness of their work (Aguilar Solano 2012) is also problematic.Matters such as sentence structure, differences between languages requiring the interpreter to hear the end of a sentence before translating it, the danger of conflating the interpreter and the speaker, and the interpreter's impartiality are all matters that are not made explicit in court proceedings.Nevertheless, these are all factors that may influence listeners' evaluations, and need to be more clearly understood, by judges and jurors alike.

Clearer pauses and conversational turn-taking
Unlike the consecutive model of court interpretation used in some jurisdictions, simultaneous interpretation allows for the more ready flow of dialogue between the questioner and the respondent.With samelanguage cross-examination, the questioner and the respondent will be speaking the same language.However, even with bilingual crossexamination, the fact that Welsh-speakers generally have (or are assumed to have) high levels of bilingual fluency, it is often the case that they will understand the questions in English, even though they choose to answer in Welsh.Those who are able to listen to the crossexamination without interpretation will therefore experience a normal flow of dialogue between the questioner and the respondent.Where the respondent understands the questioner, or where both speakers understand each other, they will be able to speak without having to consider the interpreter.This is problematic for those who need to listen to the dialogue via the interpreter.With same-language cross examination, there is a need to ensure that the listeners are able to distinguish between the questioner and the respondent -the speakers know of course that they are different people, as do bilingual listeners, but those who listen via the interpreter, if only one interpreter is engaged, will only hear one voice.Similarly, with bilingual cross-examination, there is a need for the participants to switch between direct hearing and headsets, or in the case of remote hearings, between the original language and interpretation channels.Clear conversational turn taking is important (Skaden 2018) -the speakers must allow sufficient time between utterances to allow for the change of speaker to be marked.In remote hearings this will be difficult because the interpreter will not be visible to anyone, and the bilingual judge will be likely to listen to the evidence in the original language.
Some conventions need to be developed, and the participants must be instructed to adhere to them.In our study (at the suggestion of one of the participants), we adopted the convention of applying a 5 second pause between utterances.This allowed the interpreter to mark the change of speaker in a way that was clear to the listeners, and allowed the listeners to switch between the original language and the interpretation channels.Other conventions may be developed, and the interpreter is best placed to advise the court of how to control conversational turn-taking -and we suggest that the chat and reaction functions of many online video communication platforms may be used effectively for example in order to create a signalling system to indicate when a speaker may start speaking.

Conflating the speaker and the interpreter
A matter for concern for this research is the gap identified between the fact that listeners will often evaluate the credibility of a witness with reference to aspects of their speech style.Politeness markers, pauses, colloquialisms and slang, obscenities, and the characteristics of the speaker will all influence how a person is evaluated (Berk-Seligson 2017).In cases involving physical injury, the weight of the voice (and the significance of the speaker's sex in this) will be significant, as a large, heavy-sounding person may be assumed to have suffered less significant injuries than a small, light person.In cases involving violence, the softly-spoken, gentle voice may be perceived to be a less likely perpetrator than a gruff-sounding speaker.Politeness indicators may be more indicative of remorse, while the inclusion of pauses and hesitations may indicate a greater degree of truthfulness than someone who speaks too glibly.
In our earlier work, we explored the extent to which listeners conflate the interpreter with the speaker, and discovered that this was a significant factor -the listeners had overlooked that the speaker was female, even when she was facing them, because they were hearing her evidence in the voice of a male interpreter.This situation is compounded of course in remote hearings, because, if the interpreter has their camera switched off, the absence of even the visual cues reminding the listeners of the fact that the speaker and the interpreter are different people, and that they should not be conflated with each other, has a potentially significant impact on how the listeners interpret the evidence.How this is to be resolved requires further research into the extent to which listeners conflate speaker and interpreter, and the effect this may have on the proceedings.However, at this juncture, we recommend that, as a minimum, a judge should advise the listener, and remind them that the person they are hearing is the interpreter, and how the interpreter may affect what they hear, even though their interpretation of the words is (one assumes) accurate.

Conclusion
In general terms, interpretation in court proceedings introduces a number of challenges for the courts.Remote hearings can work well with interpretation.However, there are a number of aspects to consider and to make participants aware of, particularly where simultaneous interpretation is used.The risk of conflating the speaker with the interpreter, and the invisibility of the interpreter in the process raise particular questions.It is imperative therefore that these are explored in a meaningful way as remote hearings continue into the future.

Contribution statements:
Dr Catrin Fflûr Huws -principal investigator and article author Dr Rhianedd Jewell -investigator and article editor Dr Hanna Binks -investigator and article editor Non Shafto-Humphries -researcher