Perhaps the reason for a knee-jerk response that the problems of covert recordings are too hard to fix lies in false assumptions regarding what is needed to bring about a better system.
Here’s some brief examples responding to common but inaccurate assumptions about what is (and is not) involved. The intention is partly to counter these specific suggestions, but also to emphasise that it is not a good idea to leave the solutions to law and law enforcement agencies. As discussed further shortly, problems caused by errors of common knowledge can’t be fixed on the basis of common knowledge.
1. There is no suggestion that all indistinct recordings should be transcribed by experts in phonetic science
There is an important role for phonetic science. However, it is not the role of producing transcripts. Rather it is the role of designing, in collaboration with law and law enforcement professionals, an accountable process for the handling of indistinct covert speech recordings used as evidence in criminal trials.
2. There will always be a role for police transcripts
Police transcripts are a valuable source of evidence. We need a process that maximises the reliable information that can be derived from them, and minimises the potential for misleading errors or omissions to affect the fairness of trials. There are valid ways to achieve that – the current practice of giving police transcripts to the jury for evaluation is not one of them.
3. Presenting audio with no transcript at all does not solve the problem
Common knowledge might suggest that letting the audio be heard without a transcript would minimise the problems raised here. That is not true. Contextual expectations can mislead perception just as powerfully as a transcript, and their effects are even harder to control. Indistinct audio, if used as evidence, must be accompanied by a reliable and useable transcript.
4. Transcripts made with no knowledge of context are not more accurate
Problems are not solved by completely withholding context from the transcriber. With indistinct audio, knowledge of context is essential to reliable transcription. The important thing is to ensure the transcriber is independent of the case, and to manage the disclosure of context in appropriate ways.
So there’s one particularly bad way of trying to resolve the problem of inaccurate police transcripts that unfortunately looks a lot more sensible than it really is. That is to obtain a fresh transcript by someone who knows nothing about the context, and let the court decide which one it prefers. This is a complex issue – please contact me if you wish to discuss in more detail.
5. Personal evaluation of audio by the judge is not the solution
Phonetics is a science like other sciences. Indistinct audio must be evaluated and presented by scientists with appropriate knowledge and experience, just like other forms of forensic science evidence. Judges, for all their knowledge and experience of a wide range of topics, simply do not have the relevant expertise to evaluate forensic audio.
So what is needed?
Development, via close collaboration between phonetic science and the law, of a uniform, accountable process for the evaluation and presentation of covert recordings. It is far easier than you might imagine! Interested? Check recent developments under NEWS in the top menu.