The case of Stephen “Shorty” Jamieson and insights into the legal misconceptions that cause transcript injustice in forensic contexts. How the flaws live on, despite the reforms of the 1990s, writes HELEN FRASER, Director of the Research Hub for Language in Forensic Evidence at the University of Melbourne.
ABSTRACT
Stephen “Shorty” Jamieson is 37 years into a life sentence arising from an alleged confession documented only by an uncorroborated “verbatim record of interview” typed by detectives who arrested him in 1988. The fact that he was convicted on such flimsy evidence is not the most serious problem: this form of injustice was prevalent before mandatory electronic recording of police interviews. Far more serious is ongoing failure by the law to recognize the unfairness of his trial, despite reforms enacted, just a few years later, specifically to redress this kind of injustice. This article (extract below) describes the confession evidence, and the ineffectual challenges to it in the trial (1990), appeal (1992–1993) and application for inquiry (2001). It then turns to analysis of the record of interview, revealing multiple reasons to question its authenticity. Next it examines the legal misconceptions about transcription that underlie this injustice, and shows how these live on in current times, despite the reforms of the 1990s. Finally, it amplifies calls for the law to recognize that forensic transcription requires scientific expertise – and for linguistic science to recognize that ensuring fair handling of transcripts of spoken language used as evidence requires theoretical innovation within the discipline.
EXTRACT
It is comfortable to assume that the introduction of electronically recorded interviews put an end to transcript injustice. Electronic recording of interviews (ROI) is essential, and the current process is certainly a vast improvement over typed ROIs. However, it is no panacea. Problems arise even when interviews are electronically recorded (Deamer et al., Citation2023; Dixon, Citation2008; Haworth, Citation2020). This is because electronic recording is neither necessary nor sufficient for reliable representation of spoken language in written form. As the long history of court records shows (Scharf, Citation1989), reliability depends not on electronic recording as such, but on the motivation to ensure accountable practice, and understanding of the factors that affect reliability (Fraser, Citation2022a).
The point is: the reforms of the 1990s did not get to grips with the real causes underlying the admission of misleading ROIs. They blamed the problems solely on police misconduct, when the real culprit was the law, specifically the legal misconception that evaluating transcripts requires only educated common knowledge. It is this misconception that gives judges confidence they can collectively develop procedures for the fair admission and use of transcripts in trials, and personally assess the reliability of transcripts that come under dispute in particular trials.
The situation is much the same as for other misconceptions of common knowledge that have become the basis of legal procedure; for example, the misconception that eyewitnesses provide reliable “direct” evidence of past events (Doyle, Citation2014), or that genuine rape victims immediately report the offence to authorities (Schwendinger & Schwendinger, Citation1974). The law has responded positively to academic research countering these misconceptions (Judicial Commission of NSW, Citation2025; Victoria Police, Citation2019), albeit with considerable distance still to go (Albright & Garrett, Citation2022; Quilter et al., Citation2023). By comparison, it has been slow even to recognize the misconceptions underlying transcript injustice, let alone address them. This has allowed the misconceptions to live on to cause ongoing problems in multiple ways.
They certainly cause ongoing problems for Stephen Jamieson. The fact that such flawed confession evidence was able to be used with such devastating impact in his 1990 trial was not his biggest problem. That reflected flaws in the ROI procedures that were in place at the time, but have now been recognized and reformed. Far worse is the fact that those flaws were not acknowledged in his 1992–1993 appeal process, or, worse still, via his application to review the conviction in 2001 – long after they had been officially acknowledged by royal commissions, and reformed by legislation. Indeed, as of the time of writing, the Crown is still resisting calls to review Jamieson’s case (§5). Similar reluctance to investigate historical claims of verballing are known to have affected other current and former prisoners in Australia who claim they were convicted on fabricated or coerced ROI evidence – and it seems clear that these are the tip of a large iceberg, with real numbers far exceeding the known cases (§4).
a new kind of verballing
Beyond individual cases, failure to deal with the legal misconceptions underlying ROI procedures has left open possibilities for ongoing transcript injustice across multiple sectors of the legal process (Fraser, Citation2018b). In particular, the same set of legal misconceptions underlie the use of police transcripts to assist the courts in understanding poor-quality covert recordings used as evidence in court. Of course with covert recordings, the audio is available to enable evaluation of the transcript. The problem arises when the audio is indistinct. In such cases the court allows police to provide a transcript, on the grounds that the court can readily check its accuracy against the audio (Fraser, Citation2021b). However, as shown by previous case studies in this sequence, the checking process is manifestly inadequate. The effect is a new kind of verballing that can result in acceptance of false confessions without even the conscious intention to mislead (Fraser, Citation2013). And, just as ROI procedures continued long after flaws had been demonstrated, so use of police transcripts continues to this day, despite flaws revealed not just by academic research but by acknowledged failure of legal safeguards intended to protect juries from being misled by inaccurate transcripts (McMahon & Fraser, Citation2023).
Thus the case of Jamieson, far from being an isolated aberration left over from a bygone era, is just one more example demonstrating the injustice arising from the law’s acceptance of misconceptions about language and speech embodied in educated common knowledge (Burridge, Citation2022).
Conclusion
Responding to the 2001 application for an inquiry into Stephen “Shorty” Jamieson’s 1990 conviction, Justice James acknowledged that it was “highly improbable” the ROI containing the confession on which he was convicted was an exact verbatim record of what had been said during the interview (§5 above). However, he seems not to have followed his finding to its implication that the 1990 jury reached their verdict on the basis of highly misleading evidence, namely the testimony by police, on oath, that the ROI represented the exact words with which Jamieson had confessed to one of the worst murders in NSW history. Even in the context of a plausible case that Jamieson was the victim of mistaken identity, along with the lack of any substantive evidence that he had had any involvement whatsoever in the crime, this implication seems not to have given Justice James sufficient disquiet about Jamieson’s conviction to accept the application. To the contrary, he excused the detectives for their false testimony, in a breathtaking demonstration of the kind of Catch-22 documented by decades of confession research around the world (§4) that has left Jamieson languishing in prison for a further 25 years.
It seems there is now some hope, despite ongoing opposition from the Crown, that examination of the bandana DNA might eventually bring clarity (§5). While such clarity will be welcome, when it finally comes, it will be essential to counter any ensuing narrative that this fresh scientific evidence overturned a previously reasonable conviction. As in the case of Kathleen Folbigg (Cunliffe et al., Citation2025), Jamieson’s conviction was blatantly unjust from the start. If made at all, it should have been subject to proper review, at the latest by 1995, when it had been acknowledged at the highest levels that exactly the type of confession evidence that convicted Jamieson had been the cause of systemic injustice at the time of his trial.
Both these cases, each characterized by unconscionable delays, support the institution of a Criminal Cases Review Commission (Cordner & Breen, Citation2023; Hamer, Citation2014). However CCRCs come with their own problems (Rayner, Citation2018; Thornton, Citation2025), especially regarding evidence whose acceptance relies on misconceptions of common knowledge (Ruyters & Bartle, Citation2024b). In any case, surely it is better to avoid the need for review in the first place. One of the most efficacious ways to achieve that is by reforming the legal procedures so frequently used for handling transcripts of spoken language admitted as evidence in criminal trials, with special attention to legal misconceptions as a source of injustice. Of course, such reforms have substantial initial costs, but we can take heart from the experience of reforming the ROI procedures in the 1990s. Despite strong prior resistance on grounds of cost, and restriction of police powers, the massive effort that went into instituting mandatory electronic recording was quickly rewarded by widespread approval from all branches of law and law enforcement. In following this model, however, it will be important for transcript reforms to address not just police practice but also legal procedure (§9 and §10).
In conclusion, Jamieson’s case provides yet another example (cf. Fraser, Citation2018a, Citation2023) arguing for the law to recognize forensic transcription as a (human) science – not just in the sense that the opinions of individual experts can be consulted in individual cases, but in the sense that the discipline as a whole should be consulted for assistance in establishing accountable, evidence-based procedures capable of ensuring that language evidence admitted in criminal trials is understood by the jury in a way that promotes fairness to all parties.
From the discipline of linguistics, this requires, among other developments (Fraser, Citation2023) recognition that forensic transcription is fundamentally different from transcription as practised in other branches of linguistic science. While extensive disciplinary knowledge about transcripts is highly relevant, traditional transcription methods are not directly applicable (Fraser, Citation2022a, Citation2022b). To help solve the problems that cause transcript injustice, we need a new purpose-driven approach, recognizing a transcript not just as a tool for research but as the product of a process affected by multiple factors, worthy of being a dedicated field of linguistic theory and research in its own right (Fraser, Citation2020). The reward is an opportunity to fill gaps in knowledge with implications beyond the justice application.
Published in Taylor & Francis Online (peer-reviewed journal articles)
Also see Did he say I shot the prick or I can’t breathe by Helen Fraser