Research in 5 Minutes
Title and Authors
Stein, R., Goodman-Delahunty, J., & Sourdin, T. (2025). A communication intermediary, an autistic defendant and cross-examination: a novel Australian case. Psychiatry, Psychology and Law. Published online 17 March 2025.
The Research
This Australian case study examined the use of a communication intermediary with an adult autistic defendant during evidence-in-chief and cross-examination in a sexual assault trial. The intermediary assessed the defendant's communication before trial, prepared recommendations for the court and lawyers, attended ground rules hearings and helped the defendant practise the process of giving evidence using neutral material and communication aids.
The researchers then analysed transcripts of the defendant's questioning to explore how the intermediary's recommendations, counsel questioning and communication aids affected the clarity and coherence of his evidence. The paper is a single case study, and the first author was the intermediary in the case, so the findings should not be treated as automatically generalisable to all autistic or vulnerable defendants.
What Did the Researchers Find?
- Lawyers largely followed the intermediary's recommendations and used mostly simple, comprehensible questions. The authors report that this reduced the need for intermediary intervention and supported clearer evidence.
- The defendant was able to use visual aids and independently indicated when he needed clarification. The aids included ways to communicate that he did not understand, did not know, could not remember, needed repetition or needed a break.
- Difficulties became more apparent when questioning or material increased the communication and cognitive demands. For example, the defendant became flustered when asked to work with telephone call records that he struggled to comprehend.
- Behaviour alone was not a reliable indicator of understanding. When unsure what was expected, the defendant could smile and remain silent, illustrating the risk of misinterpreting behaviour without knowledge of the individual's communication profile.
- The study illustrates the potential value of assessing communication before trial and agreeing adaptations in advance, rather than relying only on interventions after communication has already broken down.
Practice Takeaways
- Assess the individual, not the diagnostic label. Identify the defendant's own receptive and expressive language abilities, understanding of different question forms, attention, literacy, ability to seek clarification and ability to challenge inaccurate propositions.
- Reduce unnecessary linguistic and cognitive load. Use short questions containing one main point, allow sufficient processing time, and be alert to difficulties created by negatives, tag questions, multiple concepts, complex vocabulary or large amounts of information presented at once.
- Give defendants an explicit way to communicate difficulty. Depending on the individual, visual aids or agreed signals may help them indicate that they do not understand, cannot remember, need a question repeated or need a break.
- Plan adaptations before substantive questioning. Communication recommendations can be discussed through ground rules and counsel can consider the form and structure of proposed questions in advance.
- Do not assume that apparent compliance, silence, smiling or another behavioural response means that the defendant has understood. Check understanding in a way appropriate to that person's assessed communication profile.
Practice note: These are transferable practice considerations drawn from the case, not a standard set of adaptations for every autistic defendant. The paper itself stresses that its assessment and recommendations were individualised.
Read the Original Paper
Open-access original paper: https://pmc.ncbi.nlm.nih.gov/articles/PMC13215397/
Source note: Summary based on Stein, Goodman-Delahunty and Sourdin's published case study. The article was published online in 2025; later bibliographic databases may display it within the journal's 2026 volume/issue.