Speaker-Versus Listener-Oriented Disfluency: A Re-examination of Arguments and Assumptions from Autism Spectrum Disorder.

The study included only 13 adults with high-functioning autism and 13 controls in each comparison group, so it cannot support broad conclusions about autism speech. The participants were adults with existing diagnoses, so the findings say little about children or people with more support needs. The speech task was memorizing and repeating written sentences, not natural conversation, so the results may not transfer to everyday communication. The researchers did not assess language or reading ability, so some recall errors could have come from reading or encoding rather than memory. Only three filled pauses occurred, so the study could not test claims about um and uh as listener-oriented disfluencies. Unfilled pauses were defined as silent pauses of 250 ms or longer, and other studies use different thresholds, so comparisons between studies need caution.

The paper states that the funder was the British Academy. No other funding or conflict-of-interest information is given.

Do not read this as showing that people with high-functioning autism repeat themselves more or that their speech is generally less helpful to listeners. Repetitions were not significantly different, filled pauses were too rare to analyse, and the clearest difference separating the autism group from both control groups was more repairs.