The Research Hub for Language in Forensic Evidence was established to address injustice arising from the legal handling of indistinct speech recordings admitted as forensic evidence in criminal trials.
The main sources of injustice the Hub has identified in the use of forensic speech evidence are
- use of unreliable transcripts and translations to assist the court in understanding indistinct audio recordings
- admission of audio that has been ‘enhanced’ with unreliable processes, and
- unreliable attribution of utterances in audio recordings to specific speakers, in English and other languages.
The Hub’s overarching goal
To build the research base needed to ensure that all audio used as evidence in criminal trials is
- processed in accountable ways that have been demonstrated not to alter how the content is understood
- accompanied by a reliable transcript, produced via accountable, evidence-based methods with reliable translation and speaker attribution where needed, and
- provided to jurors in ways that enable them to reach a reliable interpretation of the content of the audio evidence.
Our innovative approach
It is common to assume that forensic transcription research would focus on analysing the audio itself – and of course audio analysis is important. However long-standing scientific findings indicate that, with indistinct recordings, the listener is at least as important as the audio. This is shown by the fact that the same audio can be heard in dramatically different ways by different listeners, or under different listening conditions.
In its travels through the legal process, forensic audio is heard by multiple listeners, including the transcriber, and, crucially, the lawyers, judges and jurors evaluating the transcript later, in court. We have to be sure that all these listeners understand the content correctly.
For that reason, the Hub’s approach draws on scientific knowledge about speech perception, an interdisciplinary field incorporating findings from speech science, linguistic science, and various branches of psychological and social science.
Finally, and perhaps most importantly, the Hub bases its methods on detailed understanding of exactly how forensic audio and transcripts are handled in the legal process, not just in open court, but also behind the scenes.
Current projects
Establishing accountable methods for producing demonstrably reliable transcripts of forensic audio, whether in English (including non-mainstream varieties) or in other languages
This project runs experiments using human transcribers, rigorously tested for aptitude in deciphering indistinct audio, and a bespoke transcription platform called ‘SoundScribe’, custom-built for the Hub by Melbourne Data Analytics Platform. SoundScribe has been designed to collect transcripts from listeners, and enable expert analysts to compare and evaluate multiple transcripts produced under different conditions.
Think you might have aptitude for deciphering indistinct audio? Interested to be on our experiment participant list? Contact us – we’d like to hear from you!
Disseminating reliable information about the capabilities and limitations of forensic speech enhancement and automatic speech recognition, especially in relation to methods that use AI or machine learning
This project is developing a database of forensic-like audio, and testing a range of cutting-edge enhancing techniques to determine the extent to which they help listeners decipher forensic-like audio with and without reliable or unreliable transcripts. The results are already enabling the Hub to overcome many common misconceptions about forensic speech enhancing that affect the legal process.
Providing guidance on the best ways to present audio and transcripts to juries so as to ensure they reach a reliable understanding of the audio content
With forensic speech recordings, it is not enough for an expert process to reach a reliable conclusion about a transcript or an enhancement. The Hub aims for an end-to-end solution that ensures the court as a whole, and especially the jury, reach a reliable understanding of the audio content.
Finding effective ways to counter misconceptions about speech, speech perception and transcription that affect the legal process
One of the Hub’s main findings is that injustices with forensic audio arise from legal procedures developed on the basis of misconceptions about the nature of speech, how human speech perception works, and how speech can be represented with a transcript.
Theoretical implications of forensic transcription and enhancing
The Hub’s projects need practical experimentation, but they also require advances in theoretical understanding of how speech is perceived, and how what is perceived is represented in a transcript. The Hub approach departs from standard computational models of perception and cognition (useful as these can be for their own purposes).
Contact us
For more information on any aspect of the Hub’s work, including casework requests, please contact the Director, Prof Helen Fraser.
For general enquiries, please email the School of Languages and Linguistics: soll-info@unimelb.edu.au
Professor Helen Fraser
Director
Helen is the founding Director of the Hub. She has a strong background in linguistics, phonetics and speech perception. Over the past twenty years, she has pioneered a new interdisciplinary approach to forensic transcription, with a major focus on theoretically grounded solutions that deliver needed practical results, i.e. ensuring that juries understand forensic audio evidence accurately.
Eleanor Kettle
PhD Student
Eleanor is undertaking her PhD on transcription of forensic recordings featuring non-mainstream varieties of English, under the supervision of Professor Helen Fraser, Dr Olga Maxwell and Dr Debbie Loakes. She also has a valued role as a part-time research assistant for the Hub.
Hub associates
Indispensable advisors
- Professor Lesley Stirling, School of Languages and Linguistics, University of Melbourne
- Professor Diana Eades, School of Humanities, Arts, and Social Sciences, University of New England
- Professor Marilyn McMahon, Dean of Law, Deakin University
- Professor Jane Goodman-Delahunty, School of Psychological Sciences University of Newcastle
- The Honourable Peter Gray, Retired Justice of the Australian Federal Court
- Professor Gary Edmond, School of Law, Society & Criminology, University of New South Wales
- Professor Kate Burridge, Linguistics and Applied Linguistics, Monash University
- Professor Stephen Cordner, Victoria Institute of Forensic Medicine
- Professor Peter French, Forensic Speech Science, University of York
Hub friends and collaborators
- Alex Bowen, School of Languages and Linguistics, University of Melbourne
- Dr Kate Haworth, Spoken Interaction in Legal Contexts, Aston Institute for Forensic Linguistics
- Dr Miranda Lai, Translating and Interpreting, RMIT University
- Professor Georgina Heydon, Criminology and Justice Studies, RMIT University
- Dr Cindy Schneider, School of Humanities, Arts, and Social Sciences, University of New England
- Dr Laura Smith-Khan, School of Law, University of New England
- Dr Debbie Loakes, School of Languages and Linguistics, University of Melbourne (former Research Fellow with the Hub)
- Professor Paul Foulkes, Forensic Speech Science, University of York
Frequently asked questions
-
The video demonstrates how perception of indistinct audio can be powerfully influenced by a transcript, even one that is thoroughly inaccurate. Why does that matter? Because our current law allows indistinct audio evidence to be admitted with an unverified transcript. These are typically produced by detectives investigating the crime (like the one in the video), and often have errors. The only safeguard protecting the court from being misled is that lawyers, judge and jury are expected to check the transcript against the audio. Demonstrations like the one in this video show how ineffective that safeguard is, motivating the Hub’s call for law reform in this area.
Learn more by reading these FAQs, watching demonstrations on https://forensictransription.net.au, and/or reading the article Injustice Arising from the Unnoticed Power of Priming: How Lawyers and Even Judges can be Misled by Unreliable Transcripts of Indistinct Forensic Audio. Fraser, H., Kinoshita, Y. (2021). Criminal Law Journal, 45(3), 142–152.
-
The Hub deals with language that is admitted as an ‘exhibit’ to be examined as evidence in court. This makes our topic a little different from other branches of forensic linguistics, which study how language is used in administering the legal process itself, for example during court proceedings or police interviews. The edges do get blurred however, as the Hub is very interested in how language is used in court to discuss forensic evidence, especially linguistic evidence. That's why we are called the Research Hub for Language IN Forensic Evidence, not the Research Hub for Language AS Forensic Evidence.
-
Any kind of language can be admitted as an exhibit for the jury to consider in reaching their verdict. It just has to be captured in a way that lets it be observed and discussed in court. One way language evidence can be captured is in writing – traditionally in a paper document, but now often in electronic forms such as text messages or social media. However, the Hub is most interested in addressing problems with the use of spoken language as evidence in criminal trials.
Speech evidence is usually captured in an audio recording. Typically, the audio is obtained via covert (secret) surveillance during a police investigation, but increasingly it comes from sources like mobile phones or body-worn cameras. Sometimes the 'capturing device' is simply the memory of an 'ear witness', but it is now recognised that this kind of evidence can be very unreliable.
-
Speech recorded secretly (and legally) can allow the court to hear speakers making admissions or even confessions that they would not make openly. This provides extremely powerful evidence, capable of affecting the jury's interpretation of other evidence, and ultimately their verdict of guilty or not guilty.
The problem is that forensic audio is often of very poor quality, to the extent that listeners need the assistance of a transcript to hear what is said, and who is saying it. Of course, it is essential that any transcript used for this purpose should be reliable. Otherwise, it might 'assist' the court to hear words other than those that were originally spoken.
That’s why a major strand of the Hub's research is about 'forensic transcription', the science of how to provide demonstrably reliable transcripts of poor-quality audio, in English (including non-mainstream varieties) or in other languages.
-
Forensic transcription research does not take place in a vacuum. The law has its own long-standing procedures for handling poor-quality forensic audio – and these create a range of problems that are known to cause injustice.
The first problem is that the courts allow transcripts to be produced by, or edited by, detectives investigating the case. Of course, these police transcripts are not given directly to the jury. The law has safeguards intended to protect juries from being misled by inaccurate transcripts.
The deeper problem is that these legal safeguards are ineffective. They rely on lawyers, judges and juries checking the transcript against the audio. Long-standing research shows that the very act of checking can 'prime' listeners to hear in line with the transcript, even if it is thoroughly inaccurate.
That means that, rather than protecting juries from transcript errors, the legal safeguards can propagate errors to the whole court. Multiple case studies have demonstrated the serious injustices that can arise. It is important to note that these injustices can go both ways: innocent defendants can be found guilty, or guilty defendants set free.
-
Recent developments with large language models have brought amazing advances in audio enhancing and automatic transcription – for reasonably clear speech. Poor-quality forensic speech recordings take us into a different realm completely. While the Hub is carefully monitoring the latest audio enhancing techniques, so far we have found none that can consistently and reliably transform indistinct forensic-like audio into clearly intelligible speech. And so far, even the best current automatic speech recognition systems often fail with forensic-like audio.
For the time being, then, the Hub's methods rely on expert human listeners. We aim to develop tools that enable forensic audio to be processed and transcribed in responsible, evidence-based ways that assist with the end-to-end goal of ensuring that juries gain a reliable understanding of the spoken content of forensic recordings.
-
It is common to blame transcript injustice on police or others producing inaccurate transcripts; or on lawyers failing to detect significant inaccuracies. But careful analysis suggests a deeper cause.
Current legal procedures are founded on the concept that understanding spoken language is a matter of common knowledge, safely left to judges and juries. However, while understanding spoken language is certainly a common skill, common knowledge about how that skill operates has been shown to rest on a network of misconceptions.
One example is the misconception that responsible listeners, like judges and juries, can reliably evaluate a transcript of indistinct audio simply by checking it carefully against the recording. Another is the misconception that responsible listeners can easily determine if ‘enhancing’ has made audio ‘clearer’.
-
The Hub argues that the way to address this source of injustice is for the law to treat forensic transcription as a science. That way audio can be transcribed via accountable, evidence-based practices like those used by other responsible forensic sciences, not as a matter of common knowledge, with all the misconceptions that involves.
Making this change is a challenge for the law, and the Hub is in active discussion with judges and legal scholars, helping to discover how it can best be brought about. But forensic transcription also poses interesting challenges for linguistic science.
-
Linguistic science has a great deal of relevant knowledge, collected over many decades, to explain why current legal procedures are bound to create injustice. However, to this date, our field has not developed a set of evidence-based methods specifically designed for creating demonstrably reliable transcripts of indistinct forensic audio.
That is one of the projects the Hub is currently engaged in; developing accreditation tests to identify listeners with high-level aptitude for deciphering poor-quality audio, and rigorous, evidence-based methods that enable them to produce either a demonstrably reliable transcript, or a demonstrably reliable opinion that the audio is not capable of transcription to a level of certainty suitable for use in court.
Research findings about forensic audio challenge current theories that see human speech perception as a kind of computation, so the Hub looks in new theoretical directions. We are especially interested in developing Bayesian theories that see the human listener as a purposeful, meaning-seeking agent, actively updating hypotheses about the words that are contained in incoming speech.
Publications
These are a selection of few key publications to let interested readers dive deeper into the Hub’s work.
Short introductions
Fraser, H. (2019). New linguistics research casts doubt on decades-old murder conviction. The Conversation
Burridge, K. (2017). The dark side of mondegreens: How a simple mishearing can lead to wrongful conviction The Conversation
Fraser, H. (2019). Don’t believe your ears: ‘Enhancing’ forensic audio can mislead juries in criminal trials. The Conversation
Fraser, H. (2013). Covert recordings as evidence in court: the return of police “verballing?” The Conversation
McMahon, M., Fraser, H. (2023). Transcription of indistinct forensic audio: Time for reform. Law Institute of Victoria Journal, (August), 20–23.
Fraser, H. (2021). How misconceptions about transcription affect the criminal justice system. Tiro: The Journal of Professional Reporting and Transcription, (3).
Fraser, H. (2020). Introducing the Research Hub for Language in Forensic Evidence. Judicial Officers’ Bulletin, 32(11), 117–118.
Non-technical articles
Fraser, H. (2025). “You’ve got the wrong Shorty!”: Further insights into the legal misconceptions that cause transcript injustice in forensic contexts. Australian Journal of Linguistics, 1–29.
Fraser, H. (2021). The development of legal procedures for using a transcript to assist the jury in understanding indistinct covert recordings used as evidence in Australian criminal trials: A history in three key cases. Language and Law/Linguagem e Direito, 8(1), 59–75.
Fraser, H., & Kinoshita, Y. (2021). Injustice arising from the unnoticed power of priming: How lawyers and even judges can be misled by unreliable transcripts of indistinct forensic audio. Criminal Law Journal, 45(3), 142–152.
Fraser, H. (2020). Enhancing forensic audio: What works, what doesn’t, and why. Griffith Journal of Law and Human Dignity, 8(1), 85–102.
Fraser, H. (2020). Forensic transcription: The case for transcription as a dedicated branch of linguistic science. In M. Coulthard, A. May, & R. Sousa-Silva (Eds.), The Routledge Handbook of Forensic Linguistics (pp. 416–431). Oxfordshire: Routledge.
Fraser, H., & Loakes, D. (2020). Acoustic injustice: The experience of listening to indistinct covert recordings presented as evidence in court. Law Text Culture, 24(1), 405–429.
Fraser, H. (2019). The reliability of voice recognition by ear witnesses: An overview of research findings. Language and Law/Linguagem e Direito, 6(2), 1–9.
More detailed treatments
Fraser, H. (2023). The Eastman transcripts: A case study calling Australian linguists to action against legal misconceptions about language in forensic evidence. Australian Journal of Linguistics, 43(4), 314–413.
Fraser, H. (2022). Forensic transcription: Legal and scientific perspectives. In C. Bernardasci, D. Dipino, D. Garassino, E. Pellegrino, S. Negrinelli, & S. Schmid (Eds.), Speaker Individuality in Phonetics and Speech Sciences: Speech Technology and Forensic Applications (pp. 19–32). Milano: Officinaventuno.
Fraser, H. (2019). Enhancing and priming at a voir dire: can we be sure the judge reached the right conclusion? Australian Journal of Forensic Sciences.
Fraser, H. (2018). Forensic transcription: How confident false beliefs about language and speech threaten the right to a fair trial in Australia. Australian Journal of Linguistics, 38(4), 586–606.
2026 Seminars
-
3 June 2026
This seminar examines the intersection of forensic linguistics, human rights evidence, and the admissibility of expert testimony within the court systems. While forensic linguistics has emerged as an important interdisciplinary field, those who are involved in presenting this as evidence in court settings are not always familiar with human rights laws relating to expert testimony.The seminar explores the legal standards governing expert witnesses and the admissibility of forensic linguistic evidence in courts. Particular attention is given to evidentiary frameworks such as the Daubert standard, and comparable international approaches that evaluate the reliability, relevance, and methodological validity of expert testimony. The seminar will discuss Human Rights law and the challenges surrounding objectivity, bias, translation accuracy, and the ethical responsibilities of forensic linguists serving as expert witness in legal proceedings.--
About the author
Dr Catherine Creamer is a Senior Lecturer in the School of Law and Society, University of the Sunshine Coast. She has 20 years postgraduate experience, providing psychosocial and clinical interventions to adults and children in clinical and forensic settings; both in-patient and community mental health, community outreach and providing expert testimony and advice to the Courts, Parole Board and Mental Health Tribunals.Catherine has been a Consultant Forensic Expert to the European Union, Council of Europe, since 2009 working on human rights projects relating to risk assessment, rehabilitation, mental health, and counter-terrorism. She has been a short-term consultant to the United Nations Office on Drugs and Crimes. -
by Eleanor KettleThe way that the law handles transcripts of spoken language evidence (such as police interviews, or covert audio recordings) has led to known cases of 'transcript injustice' (Fraser et al, 2025). Many well-known mis-transcriptions feature non-mainstream varieties of English (NMVE). For example, a speaker with a ‘very strong West Indian accent’ was mis-transcribed by police as saying he got on a train and shot a man to kill, but was later found to have said he got on a train and show[ed] a man ticket (Coulthard et al., 2017, p. 132).This talk will present findings from recent research relating to the forensic transcription of audio featuring NMVE. This research aligns with the Hub's aim of developing an evidence-based process for forensic transcription to ensure that only reliable transcripts of indistinct forensic audio are provided as assistance to the court.----About the authorEleanor Kettle is a PhD student and Research Assistant with the Research Hub for Language in Forensic Evidence in the School of Languages and Linguistics at the University of Melbourne.----Coulthard, M., Johnson, A., & Wright, D. (2017). An Introduction to Forensic Linguistics: Language in Evidence. Routledge. https://doi.org/10.4324/9781315630311Fraser, H. et al, (2025). New linguistics research casts doubt on decades-old murder conviction. The Conversation. https://theconversation.com/new-linguistics-research-casts-doubt-on-decades-old-murder-conviction-267425
-
Community presentation at Newling Gardens Retirement Village, Armidale NSW, on 19 March 2026.
By Professor Helen Fraser, Director of the Research Hub for Language in Forensic Evidence, University of Melbourne
It’s hard to imagine confessing to a crime you did not commit. That’s why a confession is considered the ‘gold standard’ of evidence. Yet worldwide hundreds of people who were convicted on the basis of a confession have later been proven innocent. How can that be? This talk looks at that question through the story of Stephen ‘Shorty’ Jamieson, from Sydney.
-
4 March
Stephen ‘Shorty’ Jamieson was a homeless 22-year-old when – on the basis of his nickname – he was charged with the brutal 1988 abduction, rape, and murder of Janine Balding.
Two years later, he was sentenced to life in prison, where he remains to this day.
Yet the major evidence against him was an apparent confession transcribed by detectives as a ‘record of interview’ – under conditions that have now been acknowledged at the highest levels to pose a significant risk of ‘verballing’ (presenting a confession that was never really made).
Join an expert panel to hear the facts of Stephen’s case and consider its ongoing relevance in 2026. Followed by Q and A.
The Panel:
Peter Breen
- Solicitor, former politician, long-time supporter of Stephen Jamieson.
- Author of Shorty: Mistaken Identity or Stitch-Up (Wilkinson Publishing).
- Recent guest on podcast Australian True Crime episode ‘Never to be released?’
Professor Michele Ruyters
- Director of Bridge of Hope Innocence Initiative (BOHII)
- Currently investigating Stephen Jamieson’s case.
Associate Professor Rod Gardner
- Linguistics expert, University of Queensland.
- Provided compelling evidence in Stephen Jamieson’s 2001 application to have his conviction reviewed – rejected by the reviewing judge for very unsatisfactory reasons.
Professor Helen Fraser
-
Read this recent piece in Pursuit where Professor Helen Fraser examines the Stephen ‘Shorty Jamieson’ case and how this case may not be unique.
2025 Seminars
-
by Prof Tammy Gales, Hofstra University, New York | 6 November 2025
Police reports are a cornerstone of the criminal justice system. They are the official record of an event, documenting the date, time, location, and participants involved; they serve as a reference point for further investigation; and they become vital narratives in court proceedings. Thus, their composition is a key responsibility of police officers.
In 2024, however, the law enforcement community was introduced to AI-assisted software that transfers that responsibility to an AI-assisted tool. Working from the audio recordings captured by body worn cameras (BWC), the tool first generates a transcript and then drafts a police report, all within minutes.
Despite “human-in-the-loop” safety protocols built into the tool, current linguistic research suggests that the resulting reports will still be problematic. For instance, in related AI contexts, scholars have found that transcripts produced by talk-to-text programs (such as those relied upon by the report-writing software) can be fraught with errors due to ambient noise, non-standard language use, unfamiliar accents, overlapping speech, and variations in voice due to trauma or fear. Additionally, while steps have been taken to reduce overt racism in AI-generated texts, covert racism remains. Finally, while chatbots have the capacity to process fairly transparent pragmatic meaning accurately (e.g., canonical apologies), they still have difficulties with non-canonical forms.
Examples generated by this tool, and supplemental user feedback, confirm the existence of some of these issues and raise additional concerns, demonstrating the need for more research and further discussion about the responsible use of AI in the composition of such a consequential document within the criminal justice system.
Dr. Tammy Gales is Professor of Linguistics and Director of Research at the Institute for Forensic Linguistics at Hofstra University, New York. She is currently the President of the International Association for Forensic and Legal Linguistics (IAFLL) and co-editor of the Elements in Forensic Linguistics series from Cambridge University Press.
Gales’ research interests cross the boundaries of forensic linguistics and language and the law. Within the first strand, she applies corpus and discourse analytic methods to the examination of authorial stance in threatening communications, cross-examinations of assault victims, and parole board hearings in which certain populations are disproportionately denied parole. Within the second, she applies corpus linguistic methods to the investigation of disputed meanings in legal statutes and trademark cases. Her recent projects include the newly-released Speaking of Crime: The Language of Criminal Justice, 2nd edition (with Larry Solan and Peter Tiersma), and the in-progress Corpus Linguistics and the Interpretation of Statutes across Time (with Kevin Tobia and Larry Solan).
Gales frequently serves as a linguistic consultant and expert witness on civil and criminal matters, has trained law enforcement and legal professionals across Canada and the US, and is a trained lexicographer, having worked in the UK for Longman and Bloomsbury Publishing.
https://www.hofstra.edu/faculty-profile/?id=3587; https://www.tammygales.com/
-
Read this recent piece in The Conversation where Professor Helen Fraser and other experts examine the case in terms of a fresh analysis of Jamieson’s case conducted by The Research Hub for Language in Forensic Evidence.
-
By Prof Georgina Heydon, Criminology and Justice Studies, RMIT | 8 October 2025
Within forensic linguistics, there is a long history of case studies and research exposing weaknesses in the use of written versions of spoken evidence. More recently, this has included work on transcription of unclear speech (Fraser 2022) and the interaction between transcription and translation (Gilbert and Heydon 2021), but some of the earliest forensic linguistic case work involved interview case notes (Coulthard 2002), with such concerns still relevant in forensic psychology (Milne, Nunan, Hope, Hodgkins and Clarke 2022) , particularly in the climate of global efforts to improve and standardise interviewing practices through the Mendez Principles. Nonetheless, note-taking continues to provide the main means of recording evidence and witness testimony around the world. This presentation begins with an expert witness report from a bank robbery trial about the reliability of written records of conversations between detectives and a key witness-turned-suspect in the case. This paper asks what our role might be in overturning legal practices that rely on linguistic fallacies, including, but not limited to, the assumptions about the accuracy of contemporaneous notes.
-
by Dr Zakeera Docrat | 3 September 2025
South Africa has 12 official languages that enjoy constitutional protection and promotion. Theoretically, the language rights framework provides extensive support and protection for indigenous language speakers accessing the legal system. From a point of practice, however the framework fails at various implementation stages. The 2017 English language of record decision marginalised the use of the indigenous languages while placing extra strain on a fragile interpretation profession.
In this presentation, I will focus on the criminal justice system, identifying the language-related challenges associated with police statement taking and its effects on witness credibility during oral evidence. The presentation focusses on the challenges facing police officers, court intermediaries and interpreters and the impact on complainants seeking justice.
I will discuss the cases of State v Pistorius (2014); State v van Breda (2019); State v Omotoso (2018); the Meyiwa murder trial as well as the ongoing Mbenenge Judicial Conduct Tribunal. These cases together with other examples will highlight the language complexities in the South African criminal justice system in relation to language, culture and power. As part of the recommendations, I will present the ongoing developmental Thoka App for police statement taking.
-
by Helen Fraser, Director of the Research Hub for Language in Forensic Evidence | 6 August 2025
When forensic audio is of such noisy quality that the content can barely be heard, noise reduction techniques seem to offer a foolproof solution. Strangely, however, though these 'enhancing' methods work brilliantly on intelligible but unpleasant recordings, when speech is hard to hear, they often have an effect opposite to the one intended.
Far from making indistinct speech clearer, even the latest AI models typically reduce its intelligibility, and can actually encourage inaccurate perception. Nevertheless, enhanced audio is routinely admitted in criminal trials, in the false belief that it can assist the court in understanding crucial evidence.
Today’s seminar investigates this paradoxical situation through case studies and experimental results. Why does noise reduction have this puzzling effect on indistinct audio? Why do the courts nevertheless put faith in it? What do the counter-intuitive results reveal about human speech perception? How can the answers be used to improve justice and increase scientific understanding?
This will be a fun one with plenty of audio examples. Beyond the ‘wow!’ response, it raises exciting challenges for both law and speech science, taking us to the heart of the Hub’s reason for existence.
Prof Helen Fraser is the Director of the Research Hub for Language in Forensic Evidence. For more than a decade she has taken an active research interest in the perception of enhanced forensic audio, especially how it is affected by an accurate or inaccurate transcript.
https://forensictranscription.net.au; https://findanexpert.unimelb.edu.au/profile/31457-helen-fraser
-
by Emeritus Professor Stephen Cordner, Victoria Institute of Forensic Medicine | 4 June 2025
If one was devising a scheme to resolve a scientific or medical question, one would not immediately think that employing two barristers, a judge, and a jury of twelve people off the street was the best way to go. Given the huge number and variety of different areas of expertise, it is remarkable that that approach seems to work as well as it does. Or does it? Is it fair to leave knotty problems within specialised areas to non-experts, and if not, how should such problems be resolved for the purposes of particular trials. This conundrum in all its magnitude will not be solved in this presentation, but will be discussed with particular reference to forensic medicine and some recent cases.
Stephen Cordner graduated in Medicine from The University of Melbourne in 1977. After his internship at the Royal Melbourne Hospital and two years in the Department of Pathology at Geelong Hospital he took up an appointment in 1981 as Lecturer, and later Senior Lecturer, in Forensic Medicine at Guy's Hospital in London. He stayed there until 1987 working as a Home Office Pathologist. During this period he became a Fellow of the Royal College of Pathologists of Australasia and the Royal College of Pathologists of Great Britain.
Stephen was appointed Foundation Professor of Forensic Medicine at Monash University and Foundation Director of the Victorian Institute of Forensic Medicine in 1987. He retired as Director in 2014 and returned to operational forensic pathology work, teaching and training. He has worked as a Consultant to the International Committee of the Red Cross (ICRC), exploring the potential for humanitarian contributions from forensic pathology. This has involved missions to Iraq, the former Yugoslavia, Myanmar, The Philippines, Liberia, Nepal, Palestine and Afghanistan, not only for ICRC but also WHO and UNODC. Over the last 20 years one of his particular interests has been wrongful convictions, leading in late 2023 to the publication, with Adjunct Prof Kerry Breen, of Wrongful Convictions in Australia.
-
by Professor Michele Ruyters, RMIT | 7 May 2025
When determining the admissibility of police interviews, judges are often required to assess defendants’ language competence – an area outside their formal expertise.
Through an analysis of Australian case law, this seminar explores how judges’ assumptions about their capacities to evaluate linguistic competence and resist bias can compromise judicial decision-making and result in a miscarriage of justice.
Professor Michele Ruyters is Associate Dean of Criminology and Justice Studies at RMIT University and director of RMIT’s Bridge of Hope Innocence Initiative. Dr Ruyters’ practice and research interests are in wrongful convictions, official misconduct, policing accountability, parole reform, and the lived experiences of the wrongfully convicted.
2024 Seminars
-
by Prof Jane Goodman-Delahunty, University of Newcastle
Evidence that is misconstrued can result in wrongful convictions or acquittals. To avoid these errors, juries may be assisted by educational information presented by an expert witness. That does not mean juries get it right. To assess the effectiveness of this legal intervention, trial simulation research is necessary. Pretest-posttest research designs can capture knowledge change, and the impact of expert evidence. Comparisons of individual juror decisions with those of juries after deliberation reveal the effects of group deliberation on case outcomes. Qualitative and quantitative analyses of deliberations expose how juries make meaning from evidence, apply verdict-driven or evidence-driven strategies, and the presence and extent of threats to justice, such as common jury errors and unfair prejudice. This presentation is illustrated by examples drawn from trial simulation studies, and their contributions to legal reforms are discussed.
Jane Goodman-Delahunty, BA, TTHD, JD, PhD, is a transdisciplinary scholar, trained in Law and Experimental Psychology. She conducts empirical research supporting policies to advance justice. She is a Professor in the School of Psychological Sciences at the University of Newcastle.
-
On Sunday 24 November 2024, Prof Helen Fraser appeared on the national investigative journalism program, 60 Minutes (Channel 9), along with several other experts. Prof Fraser described serious problems with the unverified transcript that convicted Stephen ‘Shorty’ Jamieson for the 1988 murder of Janine Balding. Jamieson has in prison for over 36 years on the basis of a confession he allegedly made during a police interview transcribed by one of the detectives. His lawyers are now seeking an inquiry into his case. Watch the 90-second preview below.
-
by Miranda Lai
In criminal investigations, law enforcement agencies sometime have to rely on lawful interception and surveillance to access private communications. These covertly collected materials, if deemed relevant, may be admitted as forensic evidence in court, where prosecution simply let the defendant’s own words caught “on tape” speak for themselves. When such evidence includes content in foreign language(s), translators and interpreters—an "invisible" yet essential group of professionals—play a critical role in producing translations into English, so the trier of fact can understand the material. This process poses unique challenges, as the translation is admitted as evidence, in addition to the recording, regardless of audio clarity. Forensic translation in this context requires converting spoken input into written output, a distinct process from the conventional tasks for which translators and interpreters are typically trained for, that is to produce written output from written sources or oral output from spoken input. This presentation examines the current landscape in Australia regarding the involvement of translators and interpreters in this type of hybrid task, arguing for evidence-based enhancements to existing training and workflows to uphold the accuracy and reliability standards essential for the evidentiary materials they are tasked to produce.
Dr Miranda Lai is a senior lecturer in interpreting and translating studies at RMIT University. She completed a PhD in interpreter-mediated police investigative interviews, looking into how interpreters facilitated or unknowingly interfered with such processes. Her research interests include public service interpreting and translation, police interpreting, forensic transcription and translation, and vicarious trauma for translators and interpreters.
-
by Frank Mollica
There is an abundance of evidence that human languages are shaped by principles of efficient communication. Yet, there is one notable exception: legalese. It’s widely acknowledged that the register of language we use to codify the numerous laws, agreements and obligations in modern civilizations is difficult to use and understand. In this talk, I’ll discuss our corpus and behavioural investigations with lawyers answering 1) how harmful is legalese to language processing, 2) which linguistic properties of legalese are doing the damage (hint: they’re in the abstract) and 3) why do lawyers use legalese. Luckily, our results suggest that the harm of legalese can be easily mitigated. The dominant barrier to change is the inertia of the existing system. If there’s time, I’ll discuss ongoing work investigating why we started using legalese in the first place? So, since the dawn of civilization we’ve needed laws, come to the talk to find out why have we laid them down so inefficiently
Frank Mollica is a new Lecturer in Computational Cognitive Science at the Complex Human Data Hub, School of Psychological Science, University of Melbourne. He remains an Honorary Fellow at the Institute for Language, Cognition at Computation, School of Informatics, University of Edinburgh and an associate member of the Center for Language Evolution at the University of Edinburgh. Frank uses computational and experimental techniques to investigate how children and adults construct rich conceptual systems that support everyday cognition and how these conceptual systems interface with language. Frank’s work on legal language has won an IgNobel Prize, for science that makes you laugh then think, and the Cognitive Science Society’s applied Cognitive Modelling Prize. On winter breaks, you’re likely to find Frank hiking somewhere in the Scottish highlands.
-
by Dr Vincent Aubanel
Forensic speech enhancement (FSE) appears to be a straightforward endeavour aimed at improving the quality of speech recordings presented in court. However, its execution has often fallen short of scientific standards, sometimes leading to negative outcomes and miscarriages of justice. In this talk I will present current issues in FSE from a speech research perspective, however highlighting its interdisciplinary nature. I will introduce ongoing projects and initiatives at the Research Hub for Language in Forensic Evidence, including the development of FSEEL (Forensic Speech Enhancement for Expert Listening), an innovative platform dedicated to the evaluation of FSE using quantitative and replicable methods. The aim of this talk will be to foster discussion on reforming FSE practices, with the overarching goal of ensuring more reliable and scientifically grounded methods in criminal justice.
Dr. Vincent Aubanel studied Phonetics, Music Technology and Computer Science before obtaining a PhD in Phonetics from Aix-Marseille University (Laboratoire Parole et Langage) in 2011. Since then he has held several postdoctoral positions in Spain (Ikerbasque), Australia (MARCS Institute) and France (GIPSA-lab) before joining the Hub for Language in Forensic Evidence at the University of Melbourne in 2024 as a Research Fellow. His main interests are forensic speech, speech perception in noise and speech rhythm.
-
3 July
Translation as Reframing: Recontextualisation of the Attitudinal Positionings in Bilingual Legal Judgments
By Wei (Wendy) YU
In the “uncommon” bilingual jurisdiction of common law in Hong Kong, judges can draft their legal judgments in both languages—Chinese and English. Judges can use either language, subject to the circumstances. This presentation will focus on the linguistic nuances of the two language versions via translation. As indicated in the title, I argue that translation, done by either legal translators or judges themselves, functions as a way of reframing and recontextualising the attitudinal positionings in the original version. In this presentation, I will cover the background of the linguistic landscape of Hong Kong’s legal system, the basic information of the selected cases for analysis, key findings and the interpretation, uncovering the underlying ideologies.
Wei YU is a PhD candidate from Asia Institute, under the supervision of Professor Yongxian Luo and Professor Jia Gao.
-
5 June
"Feel sorry for your ears": Exploring challenges in transcribing speech with unknown content in unfamiliar varieties
by
In many branches of linguistics, the researcher (transcriber) knows the content of recorded speech (what is being said), and the challenge is how to best represent the linguistic features being studied (eg. with IPA symbols). In forensic transcription, the situation is flipped: the transcription part is relatively straightforward (using simple spelling), but the challenge is that the content of the recorded speech is unknown (or even contested). So why is forensic transcription so difficult? This talk will explore the roles of bottom-up and top-down information in speech perception, with examples from recent experiments highlighting the crucial role of top-down information. This leads to the key question of which types of top-down information are relevant to the reliable transcription of forensic audio of unfamiliar varieties, focusing on one transcriber's experience with an unfamiliar variety of English.
Eleanor is a second-year PhD student in the School of Languages and Linguistics, supervised by Helen Fraser, Olga Maxwell, Debbie Loakes
-
by Helen Fraser
We have all seen the ‘priming’ examples** demonstrating the dangers of using police transcripts to assist the courts in understanding indistinct forensic audio, and the ineffectiveness of the ‘safeguards’ relied on by the law, such as instructing the jury to listen carefully using the transcript only as assistance. But we still have the question: ‘How should indistinct forensic audio be deciphered and transcribed, and how should transcripts be used in court?’. It is a more complex issue than many realise.
This presentation is an opportunity to hear the recommendations being developed by the Research Hub for Language in Forensic Evidence, catch up on the Hub’s recent research findings, ask questions, and offer your input and suggestions. The discussion is aimed at those with a linguistics background, but should be accessible more broadly.
Prof Helen Fraser is Director of the Research Hub for Language in Forensic Evidence at the University of Melbourne, with a research specialisation in forensic transcription.
**Not sure about the priming examples? Check out forensictranscriptiion.net.au
-
By Lois Fairclough
The interaction between phonetics and the forensic setting is complex. Although forensic phonetics is an ever-growing and innovative field, only some of the research findings translate into changes in judicial ruling. My research is centred around the use of formant frequencies as an acoustic parameter for forensic phonetic application. Although research has shown they are good at distinguishing speakers from one another relative to some other acoustic parameters, they are by no means proven reliable - particularly in speech encountered in forensic casework. In the UK, a legal ruling from over 20 years ago [R v O’Doherty, NICA 2002] requires practitioners to present formant frequency measurements in their analysis. This seems problematic given their known variability and sensitivity to a range of factors encountered in forensic settings. Findings thus far show considerable variability within and between practitioners, as well as between different softwares when measuring formants using good quality spontaneous speech data. Further work will provide a thorough investigation of how the choice of parameters in speaker discrimination tasks can influence evidential output. By furnishing a thorough analysis of this parameter, for its duration in casework, it can be duly considered in the presentation of expert evidence.
Lois Fairclough is a PhD Student studying forensic phonetics under the guidance of Dr Georgina Brown, Department of Linguistics and English Language (LAEL), Faculty of Arts and Social Sciences (FASS) University of Lancaster.
-
The Australian Linguistics ‘Call to Action’
At the end of 2017, a ‘Call to Action’ was initiated by the Australian Linguistic Society (ALS), and endorsed by the Applied Linguistics Association of Australia (ALAA), the Australian Speech Science and Technology Association (ASSTA) and the Australian Institute of Interpreters and Translators (AusIT).
The Call to Action was a letter informing the judiciary that linguists had become aware of serious problems in the legal handling of indistinct forensic audio, and asking them to review and reform the relevant procedures.
The letter was delivered to the Australasian Institute of Judicial Administration (AIJA), from where it made its way to the Council of Chief Justices (CCJ) and was eventually taken up by the Judicial Council on Cultural Diversity (JCCD) – who have an impressive track record of working with linguistics scholars to improve awareness of language and speech issues that affect the fairness of legal procedures.
Communication with the judiciary was taken forward by a working party of committed individuals, led by Helen Fraser, and including Professor Diana Eades, Professor Lesley Stirling, Mr Alex Bowen, The Honourable Peter Gray, Associate Professor (now Professor) Georgina Heydon and Professor Sandra Hale.
At the end of 2019, a working party of four judges from the JCCD held a full-day consultation with the linguists, as well as representatives from police and prosecution agencies around the country. At the end of the consultation, the judges acknowledged the linguists’ concerns, and the need for the issues be further investigated. The University of Melbourne stepped forward to take on the role, and some months later, the Hub was born.
We are glad to live in a democracy where this kind of engagement between law, law enforcement and academia is possible.
Establishment of the Hub
The Research Hub for Language in Forensic Evidence was established in February 2020 as a strategic initiative of the Faculty of Arts and the School of Languages and Linguistics, brokered by Professor Russell Goulbourne and Professor Lesley Stirling.
The Research Hub for Language in Forensic Evidence was officially launched over zoom on Friday, 30 October 2020 by Professor Russell Goulbourne, Dean of the Faculty of Arts at the University of Melbourne. Speeches were presented by The Honourable Chris Kourakis, Chief Justice of South Australia and Chair of the Judicial Commission on Cultural Diversity; Professor Diana Eades, of the University of New England; and Professor Sandra Hale, of the University of New South Wales.
Further developments, and a view to the future
On Monday 24 October 2022, with the pandemic finally starting to ease, another important consultation was convened by Professor Helen Fraser, Research Hub for Language in Forensic Evidence and Professor Marilyn McMahon of Deakin Law School, in collaboration with Melbourne Law School’s Criminal Law Research Forum.
This time an impressive list of eminent judicial officers, legal scholars and linguistics researchers from around Australia and the UK considered problems with forensic audio. At the conclusion of the workshop, participants agreed that:
- Australian courts should not use police transcripts to assist jurors in understanding indistinct audio admitted as evidence in court;
- all indistinct audio admitted as evidence should be accompanied by a demonstrably reliable transcript produced via accountable, evidence-based methods designed and managed by experts in forensic linguistics;
- law reform should be implemented to ensure the above.
During 2023-4, Helen Fraser briefed the Forensic Evidence Working Group, an ad hoc group set up by Supreme Court of Victoria to help improve the handling of forensic evidence in general. A survey was circulated, with enlightening results, and recommendations were made. Further progress is awaited.
At the end of 2023, the Hub was extended for a further five years, and after an extensive interview process, our new research fellows were selected, with Dr Vincent Aubanel starting on 5 February 2024 and Dr Annina Heini on 9 September 2024.