Publication summary: “Transcribing and translating forensic speech evidence containing foreign languages—An Australian perspective” by Miranda Lai

Another paper has recently been added to the Frontiers research topic Capturing Talk: The Institutional Practices Surrounding the Transcription of Spoken Language.

This paper, called Transcribing and translating forensic speech evidence containing foreign languages—An Australian perspective, is by a colleague of the Hub, RMIT senior lecturer in translating and interpreting Dr. Miranda Lai. It contributes to the dearth of literature on what happens behind closed doors when covert recordings contain languages other than English. This study is an Australian first large-scale survey that was conducted with professional interpreters and translators in an attempt to understand the status quo.

The survey looked at the following themes, via 19 specific questions:

1. Who among interpreters and translators are engaged to undertake forensic transcription and translation (FTT?) and what is the required training and/or credential?
2. How do they perform their FTT tasks?
3. What have been their reflections about their FTT experience?

356 Australian translating and interpreting (T&I) practitioners responded to the survey. They worked in a range of language backgrounds (49 total), the five most common being Mandarin, Arabic, Persian, Vietnamese and Cantonese. Respondents were largely NAATI Certified Professional Interpreters, followed by NAATI Certified Provisional Interpreters, and more than half had over ten years professional experience.

The study found that the task of FTT is performed by 58% of respondents less than once a year, while 36% said they engaged in FTT assignments between 1-12 times a year, and 19% said they engage in the task more than once a month. 80% of respondents noted that they perform the task alone, and 10% of respondents noted that they normally work as part of a team.

Respondents also reported on whether or not they had any briefing prior to engaging in FTT assignments. 36% noted that they always receive briefing from police officers, while another 36% noted that they never received any briefing. Those who did receive briefings would have various information relayed such as the type of case, location, people involved, and how the recording was done.

Lai found that the majority of respondents (72%) said they translated directly from the source language of the forensic recording into English, while 20% first transcribed the recording, and then translated it. The remaining respondents reported engaging in both practices. An interesting point drawn out here by Lai (p.10) is that “four respondents explicitly said that they wrote the words down first to enable a better translation into English, suggesting the utilitarian focus of this step for their translation process, rather than from the point of view of providing a traceable record for legal processes”.

Another topic covered in Lai’s paper is the additional kinds of information law enforcement tries to obtain about speakers. As noted in the paper (p.10) “[t]he respondents were also asked whether they had been asked to “profile” the speaker, i.e., “to give an opinion about what dialect they speak, or what region they might be from”. 4% of respondents said they were always asked to provide this information, while 22% said they were sometimes asked. The kinds of profiling T&I practitioners have been asked about includes accent location (for example, does the speaker come from the north or south in a country?), tribe/ ethnicity, social status and education level. As reported in the paper, a number of practitioners felt uncomfortable about this practice, with one observing that this “can be fraught with danger/traps. Best not to jump to conclusions”. (p.11).

Similarly, the study asked respondents about their experiences around voice identification. T&I practitioners gave information on whether they had been asked “to say who [speakers] are by comparing their voices to other voices either within the same recording, or in a separate recording.” Around 13% of respondents said this sometimes happened, and 3% said they always attempted this. Survey participants also varied in their opinions about voice identification – the paper gives an example of one participant saying “I deny [sic] to do that giving the reason that I am not a voice expert” but another saying “voice recognition is an important part of our work.”

Respondents also reported on their overall confidence in performing FTT, the typical kinds of audio they are faced with, and also about whether they testify in court. There was also a section of the survey where participants were asked to enter “free text” and 72 participants responded here. The main concerns from participants were poor audio quality, and the subsequent impact on their performance. Lai notes there are three themes coming out of this open-ended part of the survey, which are:

1. Working conditions and remuneration – for example conditions can be uncomfortable for T&I practitioners attending law enforcement institutions, and pay is often low;
2. Lack of translation guidance and protocols for FTT;
3. The need for specialist training.

Overall, Lai’s study reveals the mismatch between the level of competence required by FTT and the lack of training, and the absence of processes to safeguard the quality and reliability of the final translation products to serve the ultimate purpose of justice. Parallel to what has so far been well documented in monolingual settings in Australia and the concerns about police being allowed to produce transcripts of indistinct covert recordings without any training, the current practice of translators and interpreters being engaged and briefed (anywhere from next to none, to a significant amount of case information) by the investigating officers runs the risk of these practitioners being (inappropriately) primed. The paper calls for establishing FTT specialization in translating and interpreting studies, as well as the urgency to develop protocols for law enforcement when engaging translating and interpreting services for FTT. It also calls for evidence-based methodology for transcribing and translating in forensic contexts, in particular, the need for T&I professionals to be trained in transcription.

Blog post by Debbie Loakes and Miranda Lai

References

Lai, M. (2023). Transcribing and translating forensic speech evidence containing foreign languages—An Australian perspective. Frontiers in Communication. 8:1096639