Interpreter Intervention and Participant Roles in Witness Examination

Logo of the International Journal of Interpreter Education (™)

Interpreter Intervention and Participant Roles in Witness Examination

Eva Ng[1]
The University of Hong Kong

Download PDF of Article


The court interpreter code of ethics in general requires interpreters to restrict their function strictly to interpreting and to refrain from clarifying ambiguity with the speaker, especially with the witness. The code usually suggests that permission be sought from the court if interpreter intervention is unavoidable. Empirical studies show, however, that departure from this ethical code is commonplace. Drawing on an authentic courtroom trial in the High Court of Hong Kong, and using Goffman’s (1981) participation framework as the analytical tool, this article aims to illustrate how the court interpreter changes her participant role in the court proceedings by initiating turns with the speaker. It discusses the impact of such interpreter intervention on the co-present court actors and its pedagogical implications for interpreter education.

Keywords: interpreter intervention, interpreter-initiated turns, participant role, court actors.

1.       Introduction

The view of the court interpreter as a conduit of words, a view the court usually holds (Morris, 1993, 1999) considers an interpreter “a neutral machine through which a message passes untouched apart from the change in language” (Morris, 1999, p. 18). The perception of the court interpreter as a conduit in Australian law was first articulated in Gaio v. R (1960) CLR 419, where the interpreter was likened to a mere conduit pipe and a bilingual transmitter (Laster & Taylor, 1994, p. 112). The Australian case had followed an English precedent (R v. Attard [1958]) 43 Cr App Rep 90), where interpreters were categorised as a mere cipher. In an American case (People v Guzman 478 NYS 2d 455, 457-8 [1984]), the interpreter was compared to a modem (ibid.)

The notion of the court interpreter as a conduit pipe has, however, been challenged by scholars such as Morris (1995, 1999), Laster & Taylor (1994), Eades (1995, 2000) and Fenton (1997). Morris (1995) argues that interlingual interpretation is a process of communication, and that in order to attain the goal of true communication, interpreters must be allowed the latitude to go beyond the referential use of language rather than restrict themselves to verbatim interpreting. She thus argues that interpreters should be allowed to ask for and make clarifications and identify misunderstanding (1995, p. 32; 1999, p. 18) in an attempt to achieve enhanced accuracy in their performance.

The requirement for interpreters to be unobtrusive and to limit their activities strictly to the practice of interpreting has presumably been developed from the aforesaid conduit model. The Basic Guidelines for Part-Time Interpreters issued by Court Language Section of the Judiciary of Hong Kong (Judiciary of Hong Kong, 2003), for example, suggests that a court interpreter should refrain from asking questions to clarify what a witness has said, no matter how incoherent or unintelligible the speech may be. This, the guidelines argue, is to avoid giving an impression to those in court that the interpreter is engaged in a private conversation with the witness. The guidelines suggest that “the interpreter is expected to try his/her utmost to interpret accurately and faithfully what was said in full, regardless of how little sense it may make and leave the task of clarification to counsel or the bench” (Judiciary of Hong Kong, 2003, p. 3).

2.       Empirical Studies on Interpreter Intervention

Empirical studies on court interpreting have proved the conceptualization of the court interpreter as a conduit to be more of a myth than a reality. In an ethnographic and data-based study of the American courts at various levels, Berk-Seligson (1990, 2002) found that interpreters played an active role by interrupting and clarifying with attorneys and witnesses the meaning of their utterances, accounting for the side comments of witnesses and defendants, as well as prompting the witness or defendant to speak or otherwise silencing them; thus, they drew attention to themselves and made themselves highly visible.

In her study of interpreter interruptions in the Local Court Hearings in New South Wales, Australia, Hale (2001) too argues that interpreter interruptions bring in the interpreter’s own voice, rendering the interpreter more of an active participant than a mouth-piece of the interlocutors. She suggests that since interpreter interruptions are unexpected by counsel, they may interfere with questioning strategies or line of questioning, taking away some of counsel’s power and control over the witness.

In a study of asylum hearings in the Federal Asylum Office in Graz, Austria, Pöllabauer (2004) found that interpreters assumed an active role in the hearings by taking the initiative to elicit information they deemed necessary for the outcome of the hearings, and omitting or condensing information they considered irrelevant. Other activities included seeking clarifications from the asylum seekers without asking for the investigating officers’ approval, thus taking over the functions of the officers. All these studies and others (e.g., Angelelli, 2004, Fowler, 2003; Jacobsen, 2003; Roy, 2000; & Wadensjö, 1998) demonstrate that interpreters, whether in legal or other community settings, take on a co-participant role in facilitating the talk during an interpreted encounter.

3.       Aim of the Study

This study does not focus on the role of the court interpreter per se, or on court interpreters’ ethical issues. Drawing on Goffman’s (1981) participation framework, this article aims to examine why and how a court interpreter changes the participant role during the course of interpreting and how this may impact on other co-present court actors in the court proceedings. In doing so, this article also seeks to identify how these findings may be implemented in interpreter education, and to address in particular such questions as whether, when, and how interpreters should intervene.

4.       Participant Roles in Interpreter-Mediated Court Proceedings

To explore the communicative dynamics in an institutional setting like the courtroom, it is essential that we examine the institutional and participant roles taken up by court actors. This helps demonstrate not only the participation status of individual court actors but also the power relations between them. It will also show how power is maintained and realised in the roles ascribed to or taken on by these actors. In this regard, Goffman’s participation framework (1981) provides a useful analytical tool.

One can participate in a communicative act as a speaker or a hearer. In his production format, Goffman (1981) deconstructs the speaker role into that of animator (sounding box or talking machine), author (the agent who composes or scripts the lines that are uttered) and principal (someone whose position or belief is established by the words spoken).

Regarding hearer roles, Goffman (1981) identifies two basic categories:the ratified and the unratified. According to Goffman, ratified hearers are official listeners comprising both the addressed recipients, who are being directly spoken to, and the unaddressed recipients, who may or may not be listening. Goffman regards the unratified participants as bystanders, whose presence is however considered the rule, not the exception. Those who follow the talk and “catch bits and pieces of it, all without much effort or intent” are categorised as overhearers, whereas those who “surreptitiously exploit the accessibility they find they have” will qualify as eavesdroppers (p. 132).

In monolingual courtroom examinations, the examining counsel has a speaker role as both animator and author but may or may not be principal, because the counsel’s words may not attest to her own position but to that of the client or the prosecution. The witness’s speaker role, on the other hand, usually combines animator, author and principal except perhaps in the case where the witness is made to say something which does not attest to his own stance or belief. Both the examining counsel and the witness are by default each other’s addressed recipient. The defendant, the judge, the jury (in the case of a jury trial), and the nonexamining counsel can be categorised as the unaddressed recipients. Those in the public gallery as bystanders can be regarded as unratified participants (either as overhearers or eavesdroppers) because they are as a rule not allowed to directly take part in the talking event, but only to act as silent observers.

4.1.     The Interpreter’s Participant Role

It could be argued that, in the conduit model, the interpreter is not considered a participant proper in an interpreted interaction but a transparent presence. Goffman’s citation of the provision of “simultaneous translation of a speech” (1981, p. 146) as an example of the speaker animating someone else’s speech is contentious, because it is tantamount to confirming the mythical conduit model for the interpreter and the suggestion that interpreting is a mechanical process in which a message can be transferred from one language to another intact, without the interpreter having to input personal knowledge, effort and judgment in creating a new version of the talk. In producing the target language version of the message, the interpreter, as suggested by Wadensjö (1998), necessarily becomes also the author, although not the principal. There are also times when the interpreter goes beyond the strictest sense of relaying or translating, but assumes the role of a coordinator and creates her own talk in the course of coordinating the talk between the interlocutors, thus qualifying also as principal (Wadensjö, 1998).

Wadensjö (1998) suggests that Goffman’s analytical distinction of recipientship fails to take into account the different listener roles a participant in an interaction takes or is ascribed. To complement Goffman’s production format in his participation framework, Wadensjö (1998) proposes a reception format, which identifies three listener roles: reporter, recapitulator and responder. She suggests that one may listen as reporter and memorise for repetition words just uttered by another speaker as in a say-after-me language lesson. Alternatively, one can listen as recapitulator and recapitulate what was said by the preceding speaker when he takes over the floor; finally one who listen as responder introduces content of his own or by back-channelling and gazing like a direct addressee.

Applying Goffman’s production format and her own reception format to interpreter-mediated encounters, Wadensjö, 1998) suggests that an interpreter taking or being given a reporter’s role in the reception format would be expected to speak only in the restricted sense of animator of someone else’s speech; by taking or being given a recapitulator’s role, an interpreter would be expected to speak as both animator and author of the production format, whereas interpreters taking the role of responder would relate to their talk as animator, author and principal and as the ultimate addressee, as in the case of clarifications with the preceding speaker. Wadensjö suggests that in the course of interpreting, interpreters, with the “mandate and responsibility to compose new versions of utterances”, always take the reception role of recapitulator and thus the production role of animator and author. The reception roles and thus the production roles the interpreter takes are represented in Figure 1 below.

Figure 1. Interpreter’s relationship between reception format and production format

Reception format (Wadensjö, 1998) Production format (Goffman, 1981)
reporter animator
recapitulator animator, author
responder animator, author, principal

In what follows, I will not be addressing ethical issues, but will merely be presenting the reality. I will show how the interpreter interweaving between a recapitulator and a responder in her listening role speaks not only as animator and author, but also as a principal in coordinating the courtroom talk. I will also discuss its potential impact on the co-present court actors.

5.       The Data

The data in this paper were taken from a High Court murder case in Hong Kong. Court interpreting between English and Cantonese has been an indispensable service in the courtroom of Hong Kong since the onset of British colonial rule in 1842, due to the use of English as the court language and the predominantly Cantonese-speaking local population. Interpreters are therefore needed to bridge the communication gap between English-speaking legal professionals and Cantonese-speaking lay participants appearing in court as witnesses or defendants. The changeover of Hong Kong’s sovereignty in 1997 has resulted in an increasing use of Chinese as the court language, especially in the lower courts. A large percentage of criminal cases are, however, still tried in English in the High Court, due to the presence of expatriate judges and/or counsel, because English remains one of Hong Kong’s official languages. From the late 1990s onwards, all court proceedings have been audio recorded, thus enabling a bilingual court reporting system. Access to the court recordings of this trial and of eight other criminal trials was granted by the High Court for research purposes. The recordings were transcribed for analysis and the symbols and abbreviations used in the transcripts are set out in Figure 2 below.

Figure 2. Abbreviations and transcription keys used in this study

Abbreviation/Symbol Meaning
DC defence counsel
I interpreter
PC prosecution counsel
PW prosecution witness
SL source language
TL target language
[ overlapping talk
(2) the length of a pause in seconds
(words) possible hearings in the transcript/words added in the English gloss for grammatical reasons
(.) a brief pause of less than a second
? a sudden cut-off of the current sound
< > angle brackets contain transcriber’s descriptions rather than transcriptions
CAPITALS a louder voice relative to the adjacent talk, represented in Chinese by a change in the typeface of the characters

In this trial, the Cantonese-speaking defendant is charged with one count of murder for killing his landlady’s husband, who came over to the leased premises to demand rent arrears, accompanied by his wife, the landlady. The judge, the prosecution counsel and the defence counsel are all English-speaking expatriates, who do not speak Cantonese. The present study focuses on the examination of the first prosecution witness (PW1), the landlady, who testified in Mandarin (as an immigrant from mainland China), through an English–Mandarin interpreter in open court. The examination of PW1 is singled out for analysis because it was found to have the highest number of interpreter-initiated turns (hereafter IITs) among the examinations of all the other witnesses in this case and in the eight other cases I was given access to. In this case, because the defendant speaks Cantonese, a second interpreter had to be used to provide chuchotage from Mandarin/English into Cantonese for the benefit of the defendant, who otherwise would have been excluded from participation in the proceedings. However, because the Cantonese interpretation was provided in chuchotage, audible only to the defendant, and was not picked up by the recording system, this study focuses on the Mandarin/English interpretation provided consecutively in open court.

6.       Findings and Analysis

Throughout PW1’s examination-in-chief and cross-examination, the interpreter was observed to initiate turns with PW1 on many occasions, thus speaking as principal and listening as responder. An examination of the turn exchanges reveals a total of 200 IITs in the examination-in-chief of PW1, representing 13.3% of the total turns, or 28% of the total interpreter turns. Of the 200 IITs, 190 are made with PW1 and 10 with the examining counsel (see Table 1 below).

Table 1. IITs in PW1’s examination-in-chief and cross-examination

Type of IITs In-chief % of total speaker turns (1506) % of total interpreter turns (715) Cross % of total speaker turns (1908) % of total interpreter turns (911)
IITs with PW1 190 12.6% 26.6% 102 5.4% 11.2%
IITs with examining counsel 10 0.7% 1.4% 6 0.3% 0.7%
Total 200 13.3% 28% 108 5.7% 11.9%

The cross-examination of PW1 too showed a large number of IITs, although not as numerous as in the examination-in-chief, presumably because the interpreter had been furnished with most of the details of the case and thus the need for clarifications was significantly reduced. In addition, the majority of the questions in cross-examination are confirmation-seeking questions (CSQ) whereas questions in examination-in-chief are mostly information-seeking questions (ISQ; commonly known as WH-questions; see Hale, 2004; Harris, 1984; & Woodbury, 1984 for question categories in witness examination), and the witness’s answer is often limited to a choice between an “yes” or a “no”, typically in the form of “do you agree with me…” or “is it true that…”. The need to clarify with the witness in cross-examination thus diminishes, although the interpreter may sometimes need to clarify with counsel in cases of long and syntactically complicated questions. There were 108 IITs in the entire cross-examination process, representing 5.7% of the total turns or about 12% of the total interpreter turns. Of the 108 IITs, 102 were made with PW1 and 6 with counsel.

6.1.     Typology of IITs

Berk-Seligson (1990, 2002) regards the dialogues initiated by interpreters as “the interpreter’s attention-drawing behavior” (p. 65), arising from the need to clarify witnesses’/defendants’ answers and attorneys’ questions, to account for witnesses’/defendants’ side comments, to prompt witnesses/defendants to speak, or to silence them. Hale’s (2001) study of the New South Wales courtroom in Australia demonstrates similar findings but suggests that interpreters also interrupt to provide unsolicited information and offer personal opinion.

With Berk-Seligson’s and Hale’s typologies as a point of reference, I have arrived at a typology of nine categories of the IITs identified and quantified in the examination of PW1 and present them in Table 2 below.

Table 2. IITs in PW1’s examination-in-chief and cross-examination

Types of IITs In-chief % Cross %
1. To seek confirmation 78 39% 33 30.6%
2. To seek clarification 70 35% 57 52.8%
3. To seek further information 15 7.5% 3 2.8%
4. To coach the witness 12 6%
5. To respond to the witness’s question 12 6% 9 8.3%
6. To prompt the witness (especially after interrupting the witness) 11 5.5%
7. To inform the court of the need to finish an interrupted interpretation 2 1%
8. To back-channel before rendering the witness’s answer 5 4.6%
9. To point out a speaker mistake 1 0.9%
Total 200 100% 108 100%

Some of these categories coincide with those of Berk-Seligson’s and Hale’s typologies, although my findings comprise more categories, some of which, especially the first three, may partially overlap. The reason why they are treated as categories in their own right is that in Category 1, the interpreter simply repeats or rephrases the speaker’s utterance to check her understanding without clarifying ambiguity or seeking further information (Example 1), whereas in Category 2, the interpreter takes the initiative to clarify ambiguity either arising from contextual problems (Example 2), or due to linguistic or cultural differences. In Category 3, the interpreter explicitly requests further information from the speaker (Example 3), which results from neither a decoding problem nor ambiguity of any kind. The following examples are drawn from my data to illustrate all these categories.

6.1.1    To seek confirmation

Most of the IITs occurring in the witness’s examination-in-chief are checking turns used by the interpreter to check her understanding of the witness’s utterance by repeating or rephrasing what is said by the witness—also the second-most-frequent type of IIT in the witness’s cross-examination, as demonstrated in Example 1.

Example 1. Examination-in-chiefInterpretation Example in Chinese and English text

On the surface, the interpreter (I) is making use of these turns (turns 2 and 4) to check her understanding of the speaker meaning, but it may well be the case that the interpreter uses these turns as a stalling tactic to buy her time for better reformulating her interpretation, as shown in turn 4, where the interpreter’s turn is immediately followed by her rendition without waiting for PW1’s confirmation.

6.1.2    To seek clarification

Apart from seeking confirmation from the witness, the interpreter interrupts the proceedings frequently to clarify the meaning of PW1’s utterances. Example 2 is one of this kind.

Example 2 Examination-in-chiefInterpretation Example in Chinese and English text

Because both the defendant and the deceased have the same surname, the interpreter is found to clarify on a number of occasions with the witness when she makes references to a Mr. W, as demonstrated in Example 2 above. By clarifying with the witness, the interpreter takes on a primary participant role: a listener role as responder in Turn 1, a speaker role as animator, author and principal in Turn 2, and finally as addressee of the witness’s reply in Turn 3.

6.1.3    To seek further information

In Example 3, again extracted from the examination-in-chief of PW1 by the prosecution counsel (PC), the interpreter asks the witness a follow-up question to seek further information before interpreting her utterance, possibly in an attempt to make a more complete and grammatically adequate rendition. I find it ironic that the witness would probably have been able to provide the requested details had she not been interrupted by the interpreter.

Example 3. Examination-in-chief

Interpretation Example in Chinese and English text

Example 4 below is another example of the interpreter asking the witness for further information before rendering her preceding utterance into English.

Example 4. Examination-in-chief

Interpretation Example in Chinese and English text

6.1.4    To coach the witness

The examination-in-chief reveals 12 instances of the interpreter coaching the witness. In most of these examples the interpreter tells the witness to speak slowly, as in Example 6 below. The interpreter’s coaching turns can be seen as her attempt to control the pace at which the witness testifies and thus the flow of the communication so as to facilitate her work of interpreting. Note that this is also followed by a request for the witness to repeat what she has just said.

Example 5. Examination-in-chief

Interpretation Example in Chinese and English text

6.1.5    To respond to the witness

The interpreter is also found to repeat, rephrase or elaborate counsel’s question when the witness’s answer appears to be nonresponsive, thus leaving the witness’s utterance uninterpreted as in Example 6. In this case, the interpreter might have held herself responsible for PW1’s nonresponsive answer, thinking that the witness must have misheard her, and thus takes the liberty to respond to her. It might as well be the case that the interpreter is worried that reproducing PW1’s nonresponsiveness would be face-threatening, because the majority non-Mandarin-speaking participants in court might mistakenly conclude that there is an interpreting problem.

Example 6. Examination-in-chief

Interpretation Example in Chinese and English text

At other times, the interpreter is observed to respond to a witness’s question directly, without interpreting it and referring it back to the defence counsel (DC) as in Example 7.

Example 7. Cross-examination

Interpretation Example in Chinese and English text

As the prescribed role of the witness in the judicial process is to answer, not to ask questions, and the legal base of power stipulates counsel’s right not only to ask questions, but also to impose sanctions against those refusing to answer (Walker 1987), the interpreter responding to the witness’s question without interpreting it and referring it back to the defence counsel in this case has in a way legitimated the witness’s right to ask questions and deprived the examining counsel of his right to censure the witness for not answering his question. Had the interpreter interpreted the question rather than responding to it, the defence counsel might have protested against it, as is evidenced in Example 8, in which the interpreter does not respond to but interprets the witness’s clarifying question for the defence counsel.

Example 8. Cross-examination

Interpretation Example in Chinese and English text

6.1.6    To prompt the witness

Prompting mostly occurs after the interpreter has rendered an obviously unfinished utterance by the witness. This can be seen as a repair strategy on the part of the interpreter as in Example 9 below, where the interpreter starts interpreting before the witness is able to finish her turn. After rendering her answer into English, the interpreter recapitulates it in Mandarin for the witness, as a reminder of what the witness has said, before prompting her to carry on with her testimony. The interpreter may have deemed it necessary to prompt the witness to go on with her testimony or else the turn might be taken over by the examining counsel.

Example 9. Examination-in-chief

Interpretation Example in Chinese and English text

6.1.7    To inform the court of the need to finish an interrupted interpretation

Because both the prosecution and the defence counsel in this case are monolingual English-speaking expatriates and thus have no access to the witness’s testimony in Mandarin, there are two instances of the prosecution counsel trying to take back his turn to carry on with his questioning, having taken the interpreter’s hesitation pause as an end-of-turn pause. As a result, the interpreter has to interrupt the prosecution counsel in order to finish her turn. This would not usually happen with bilingual counsel, who would be able to tell if the interpreter has completed her turn by overhearing the witness’s testimony in the source language. In this case, it could be argued that the interpreter intervenes in order to adhere to the ethical code of accuracy and completeness, and the intervention is unavoidable and therefore justifiable. Example 10 below is one of the two examples identified.

Example 10. Examination-in-chief

Interpretation Example in Chinese and English text

6.1.8    To acknowledge the understanding of the witness’s utterance

Example 11 illustrates the interpreter signalling her understanding of PW1’s answer by means of back-channelling, which is evidence of the interpreter listening as a responder (Wadensjö, 1998). This might also be taken as the interpreter’s strategy to stop PW1 from giving an answer which is too lengthy to be rendered accurately and completely. Note the overlap of the interpreter’s voice with PW1’s utterance, which is cut short by the interpreter’s back-channelling.

Example 11. Cross-examination

Interpretation Example in Chinese and English text

6.1.9    To point out a speaker mistake

In Example 12 below, the defence counsel has made an obvious mistake about the date on which PW1 made her statement to the police. Because the month in question is August, not September, the interpreter is sure that the defence counsel has made a mistake and alerts him to it in a whisper. The interpreter’s intervention in this case is presumably to avoid the confusion which might be caused to the witness if the mistake is preserved in the rendition. This might also be regarded as the interpreter’s face-saving strategy because any confusion likely to be caused by the reproduction of counsel’s mistake might be attributed to an interpreting problem. In any case, it is evident that the interpreter does not see herself as a copying machine (the conduit myth) but one who plays an active role in coordinating talk and facilitating communication, by listening, in this case, as responder and speaking in her own voice as animator, author and principal. Note that in this case, the defence counsel carried on without responding to the interpreter’s correction. He might not have heard the interpreter’s correction and was not aware of his mistake until the prosecution counsel, who might have been alerted by the interpreter, stepped in.

Example 12. Cross-examination

English text

7.       Impact of IITs

With the participant roles of court actors in mind and the implications for their participation status and control over the triadic communication, this section will explore the impact of such IITs. In her study of interpreter interruptions with the examining counsel, Hale (2001) contends that interpreters interrupting counsel will interfere with the latters’ questioning strategies or line of questioning, taking away some of their inherent power and thus control over the testimony of the witness. While Hale’s study focuses on interpreter interruptions of counsel, my analysis will focus on the impact of IITs on the witness, which account for over 90% of the total number of IITs in both the examination-in-chief and cross-examination of PW1 in this study (see Table 1 above).

7.1.     Participation Status of Co-Present Court Actors and Their Control Over the Interaction

In the examples cited above, the interpreter is seen to take on a primary participant role and assumes much latitude in negotiating meaning with the speaker. In these IITs, the interpreter ceases to be the voice of the key interlocutors, but is speaking in her own voice, combining the roles of animator, author and principal in Goffman’s (1981) production format, and as reporter, recapitulator and responder in Wadensjö’s (1998) reception format. By initiating talk with the witness, she has also made herself a direct addressee of the witness’s response. These interpreter–witness turn exchanges in Mandarin have, however, effectively excluded the participation of not only the monolingual English-speaking counsel and judge, but also the predominantly Cantonese-speaking jury and the audience in the public gallery. The exception is the defendant, who has had whispered Cantonese interpretation provided to him by a second interpreter and thus retains his “ratified unaddressed recipient role” in the participation framework. During these IITs, the monolingual judge and counsel, as Bell (1984, p. 176) puts it, become “uncomprehending hearer[s]” and are thus rendered “non-member[s]” because of the use of a language unintelligible to them. It could thus be argued that having been excluded from these IITs, the monolingual counsel and judge have seen their control over the flow of the testimony compromised or reduced. They are unable to access these interpreter–witness verbal exchanges, let alone intervene in the process. In the above-cited examples, the interpreter could be described to have usurped some of the power of the examining counsel, displaying considerable control over the flow of testimony.

7.2.     Evaluation of Counsel, the Witness and the Interpreter

The IITs may also have an impact on jurors’ impression of the examining counsel and the witness whose utterances are interrupted by the interpreter. Berk-Seligson’s (1990, 2002) experiment with mock jurors to evaluate the impact of interpreter intrusiveness shows that the attorney interrupted by the interpreter was found by the sample of listeners as a whole to be less competent and by Hispanic listeners as a subgroup to be both less competent and less intelligent. On the other hand, interpreter interruptions of the witness were found to have no impact on the attorney’s competence, intelligence or persuasiveness, but Hispanic mock jurors found the witness whose testimony was interrupted by the interpreter to be significantly less convincing and less competent. Berk-Seligson notes that the results suggest that those observing interpreted proceedings make a distinction between an interpreter’s interruptions of an examining counsel and of a witness. She suggests that an interpreter’s interruptions of an examining counsel “can be perceived as a veiled criticism” (p. 191) of his performance, thus rendering him less competent; an interpreter’s interruptions of a witness, however, seem to be seen by mock jurors “partially as a problem of the interpreter’s and partially a defectiveness in the witness”, but as unconnected to the examining counsel’s “professional capabilities” (p. 191).

In the light of Berk-Seligson’s findings, it could be argued that the frequent IITs with the witness in the case in this study might render the latter less trustworthy and less competent in the eyes of the jurors, who might perceive her to be evasive and uncooperative. The interpreter herself might also suffer a negative appraisal, judged by others in the courtroom as incompetent and unprofessional.

8.       Pedagogical Implications

As in any monolingual communication, problems of communication such as nonresponsive, ambiguous or unclear answers, do arise from time to time in interpreter-mediated interactions, and thus the need for clarifications is sometimes unavoidable. It is therefore unrealistic to suggest that interpreters should under no circumstances clarify with the speaker. However, clarifications by the interpreter with primary interlocutors not speaking each other’s language can be a very complicated issue. As has been explained above, any intervention by the interpreter, no matter how brief it may be, inevitably excludes the participation of the noncomprehending court actors, who may be left to wonder what is going on between the interpreter and the witness. This may also adversely impact on the evaluation of the competence of the interpreter and the trustworthiness of the witness as noted above. With this in mind, interpreter intervention such as prompting the witness or asking the witness for further information should be avoided where possible. It is therefore essential that student interpreters are taught when and how to intervene, especially when encountering the following situations.

8.1.     Ambiguity

Clarifying ambiguity in the courtroom can be a tricky issue. For one thing, counsel may, for a strategic reason, intend questions to be ambiguous in order to confuse the witness. Likewise, evasive witnesses may not want to give a clear answer to a question put to them, especially one that is likely to incriminate them. It is therefore advisable for court interpreters to retain, where possible, the ambiguity in the target language. For example, the Cantonese word saam1 is notoriously ambiguous, because it can mean either clothing (garment) or a top (upper garment). An interpreter’s attempt to disambiguate it or to opt for one meaning over another, as illustrated in a rape case I have written about (Ng, 2012, 2013), may prove problematic and may be challenged by counsel who considers a different interpretation more advantageous to his case. Likewise, pronouns in spoken Chinese are gender-neutral and nouns in Chinese do not have singular or plural markers. Therefore, when rendering a witness’s testimony into English, the interpreter might deem this information necessary in order to make a grammatically adequate rendition and want to clarify with the witness. However, instead of initiating clarifications with the witness and excluding other participants in the clarifying process, the interpreter might instead consider retaining the ambiguity by saying, for example, “he or she” or “finger or fingers” and leave the burden of clarification to counsel or the court.

8.2.     Questions or Requests From the Witness

Handling questions or requests from the witness is another challenging issue which particularly deserves novice interpreters’ attention. In the course of interpreting, a witness may ask for repetition or explanation of an interpreted question, as illustrated in Example 7. It may not always be easy to tell who is to blame for the witness’s noncomprehension: the counsel who asked the question, or the interpreter who interpreted it. An interpreter who responds to the request/question without interpreting it and referring the request back to counsel, as the interpreter did in Example 7, may have held herself accountable for the witness’s comprehension problem (or simply responded for the sake of efficiency). An interpreter who thinks otherwise or adheres to the ethical code by interpreting everything said in court might interpret the request and refer it back to counsel. It would be considered ethical, and in the best interest of justice, for the interpreter to render everything said by the witness in court into the target language, as is prescribed in most interpreters’ codes of ethics or guidelines for professional practices. It is however important that the interpreter does not do this indiscriminately. In a recent preconference visit to a criminal court in Guangzhou, China, many delegates, especially those who speak both Mandarin and English, were amused by the way the interpreter handled the defendant’s request for repetition. At some point during our observation in court, the Mandarin-speaking judge asked the English-speaking defendant, through an interpreter, if he agreed to the particulars of the offence he was charged with. The defendant responded with this on two occasions, “You speak very fast, I don’t understand. Can you please say that again?” On both occasions, the interpreter immediately rendered the defendant’s utterance into Mandarin for the judge to repeat his questions, not realising that it was her own problem, not that of the judge, who was very slow and clear in putting his questions to the defendant. It was the interpreter who was a fast speaker. Presumably the defendant had no access to the judge’s utterance in Mandarin or to his speech tempo for that matter, so he could not have targeted the request at the judge. Under such circumstances, the interpreter should inform the court that the defendant requests the interpreter to go slowly and to repeat the rendition.

8.3.     Nonresponsive Utterances

In the case where a witness’s answer is nonresponsive to counsel’s question, some interpreters may prompt the witness to be responsive or relevant in her answer by either repeating the interpreted question (as shown in Example 3 above) or responding to the witness by rephrasing or explaining counsel’s question, in the belief that the witness may not have adequately heard or understood the interpreted question. The interpreter’s intervention in such a case may also be seen as an attempt to protect her own face, or to save the court’s time. However, from a pedagogical point of view, intervention of this kind should be discouraged, as this would inevitably deny the access of other court actors, including that of the judge and the jury, first to the witness’s nonresponsive answer (which is not interpreted), and to the subsequent intervention by the interpreter. These noncomprehending court actors would have a reason to believe that the interpreter and the defendant are engaged in a private conversation, from which they are excluded.

8.4.     Inaudible or Nonsensical Utterances

There are cases in which interpreters feel obliged to intervene and such intervention is justifiable. That is, when they cannot hear the speaker or when what the speaker (witness and counsel alike) has said does not make much sense to them. It is fair to say that an interpreter must be able to hear and make sense of what is said before he can render it into the target language. It is therefore unrealistic or unreasonable to expect an interpreter to refrain from clarifying with the witness but to “try his/her utmost to interpret accurately and faithfully what was said in full, regardless of how little sense it may make” in the case where a witness speaks “incoherently or unintelligibly” (Judiciary of Hong Kong, 2003, p. 3). Clarifications under such circumstances seem unavoidable. It is nonetheless advisable for the interpreter to inform the court beforehand. It would, however, be helpful for educators to warn interpreters of the perils of engaging in a lengthy clarifying process comprising multiple exchange turns. Where possible, a witness’s response to a clarifying question should first be interpreted before further clarification is sought.

9.       Conclusions

The process of interpreting is dynamic, and so are the roles of the interpreter, which can vary from one extreme as a conduit to the other as an advocate. During the course of interpreting, interpreters have to make quick decisions and solve problems. There are, as Mikkelson (2008) suggests, a range of options for interpreter intervention “in the middle of the spectrum between what is deemed by most as unacceptable advocacy for individual clients and what most consider acceptable advocacy for the interpreting process” (p. 87), which has yet to be fully defined. This study corroborates previous research about the role of the interpreter as a co-participant as she constructs or co-constructs the talk between interlocutors not speaking each other’s language. Although interpreters, like other interlocutors, should be given the right to clarify ambiguity during the course of interpreting as the need arises, they should be taught how to exercise this right properly and be alerted to the potential impact of their intervention. It is hoped that this study has made a useful contribution to interpreter education in this regard.


I would like to thank the Honourable Madam Justice Queeny Au-Yeung, the then High Court Registrar of the Judiciary of Hong Kong, for granting me access to the audio recordings of nine criminal trials for teaching and research purposes. I am also indebted to Leung Kau Kui Research and Teaching Endowment Fund (2009) and Teaching Development Grants (2014), both from HKU, for funding the transcription of the bulk of the audio data.


Angelelli, C.V. (2004). Medical interpreting and cross-cultural communication. Cambridge, UK: Cambridge University Press.

Bell, A. (1984). Language style as audience design. Language in Society, 13(2), 145–204.

Berk-Seligson, S. (1990). The bilingual courtroom: Court interpreters in the judicial process. Chicago, IL: University of Chicago Press.

Berk-Seligson, S. (2002). The bilingual courtroom: Court interpreters in the judicial process (Rev. ed.). Chicago:, IL University of Chicago Press.

Berk-Seligson, S. (2006). Court interpreting. In K. Brown (Ed.), Encyclopedia of language and linguistics (pp. 262–269). Oxford, UK: Elsevier.

Eades, D. (1995). Cross examination of Aboriginal children: The Pinkenba case. Aboriginal Law Bulletin, 3(75), 10–11.

Eades, D. (2000). I don’t think it’s an answer to the question: Silencing Aboriginal witnesses in court. Language in Society, 29(2), 161–195.

Fenton, S. (1997). The role of the interpreter in the adversarial courtroom. In S. E. Carr, R. P. Roberts, A. Dufour, & D. Steyn (Eds.), The critical link: Interpreters in the community (pp. 29–34). Amsterdam, the Netherlands: John Benjamins.

Fowler, Y. (2003). Taking an interpreted witness statement at the police station. What did the witness actually say? In L. Brunette, G. Bastin, I. Hamlin, & H. Clarke (Eds.), The critical link 3. Interpreters in the community (pp.195–209). Amsterdam, the Netherlands: John Benjamins.

Goffman, E. (1981). Forms of talk. Oxford, UK: Blackwell.

Hale, S. (2001, May). “Excuse me, the interpreter wants to speak”–Interpreter interruptions in the courtroom: Why do interpreters interrupt and what are the consequences? Paper presented at Critical Link 3, Interpreting in the Community: The Complexity of the Profession, Montreal, Canada.

Hale, S. (2004). The discourse of court interpreting: Discourse practices of the law, the witness, and the interpreter. Amsterdam, the Netherlands: John Benjamins.

Harris, S. (1984). Questions as a mode of control in magistrates’ courts. International Journal of the Sociology of Language, 49, 5–27.

Jacobsen, B. (2003). Pragmatics in court interpreting: Additions. In L. Brunette, G. Bastin, I. Hamlin, & H. Clarke (Eds.), The Critical Link 3. Interpreters in the community (pp. 223–238). Amsterdam, the Netherlands: John Benjamins.

Judiciary of Hong Kong. (2003). Basic guidelines for part-time interpreters. Hong Kong: Hong Kong Government.

Laster, K., & Taylor, V. L. (1994). Interpreters and the legal system. Sydney, Australia: Federation Press.

Mikkelson, H. (2008). Evolving views on the court interpreter’s role: Between scylla and charybdis. In C. Valero-Garcés & A. Martin (Eds.), Crossing borders in community interpreting: Definitions and dilemmas (pp. 81–97). Amsterdam, the Netherlands: John Benjamins.

Morris, R. (1993). Images of the Interpreter: A study of language-switching in the legal process, unpublished Ph.D. thesis, Lancaster University: Department of Law.

Morris, R. (1995). The moral dilemmas of court interpreting. The Translator, 1(1), 25–46.

Morris, R. (1999). The gum syndrome: Predicaments in court interpreting. Forensic Linguistics, 6(1), 6–29.

Ng, E. (2012). Garment, or upper-garment? A matter of interpretation? In Proceedings of the International Association of Forensic Linguists’ Tenth Biennial Conference. Retrieved from

Ng, E. (2013). Garment, or upper-garment? A matter of interpretation? International Journal for the Semiotics of Law – Revue internationale de Sémiotique juridique, 26(3), 597–613. doi:10.1007/s11196-012-9290-9

Pöllabauer, S. (2004). Interpreting in asylum hearings. Interpreting: International Journal of Research & Practice in Interpreting, 6(2), 143–180.

Roy, C. (2000). Interpreting as a discourse process. New York, NY: Oxford University Press.

Wadensjö, C. (1998). Interpreting as interaction. London, UK: Longman.

Walker, A. G. (1987). Linguistic manipulation, power and the legal setting. In L. Kedar (Ed.), Power through discourse (pp. 57–80). Norwood, NJ: Ablex.

Woodbury, H. (1984). The strategic use of questions in court. Semiotica, 48(3-4), 197–228.

[1] Correspondence to: