Legal and scientific perspectives
Keywords:transcription, perception, forensic, evidence, acoustic
Audio recordings are often used as forensic evidence in criminal trials. Unfortunately, they are often of very poor quality, meaning the court needs a transcript to be sure of their content. Many jurisdictions allow transcripts to be provided by police. This creates problems that can result in substantial injustice. Phonetic science is needed, but how can it best assist? Many recommend that transcripts should be produced, or evaluated, by experts in acoustic-phonetic analysis. However, this does not necessarily solve all the problems. The present paper argues that this is because forensic transcription is significantly different from established forms of phonetic analysis, and requires not just applying existing knowledge, but developing new knowledge, with a broader view of the evidence needed to ensure a transcript of indistinct audio is reliable.
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.