What's in a Voice? Mindreading and Prosody in Autism Spectrum Disorders

Background: Recent tests of Theory of Mind (ToM) demonstrate that individuals with an Autism Spectrum Disorder (ASD) struggle to recognise mental or emotional states conveyed in the voice (see e.g., Golan, et al., 2007; Rutherford, et al., 2002) but the exact relationship between ToM and vocal cue recognition requires further investigation. In particular, studies often mix various emotions types hence making it difficult to identify a selective impairment in processing vocal cues linked to ToM.

Objectives: This study aims to tackle this issue by drawing on psychological research and pragmatic theories in order to distinguish five vocal-cue categories based on the amount of mindreading they require: manners of speech (e.g. singing), physical states (e.g. being tired), basic emotions (e.g. happiness), social emotions (e.g. embarrassment), and speaker’s attitudes (e.g. irony). We argue that only the former two require ToM and predict that they will be especially challenging for ASD individuals.

Methods: Teenagers with a high functioning ASD and Typically Developing (TD) controls matched on chronological age, verbal mental age and basic auditory skills were included. In Experiment 1 (ASD: n=17, Mean CA=14;2, TD n=17, Mean CA=13;8), sentences with a neutral content and a marked prosodic contour were presented to the participant, who then had to pick the foil which best described the speaker’s state. For example, the item “Jane’s Mum wonders why Jane is not with Ben. Ben says: I told her to walk home from school!” was followed by two foils: “Ben is proud. He thinks he had a great idea!” and “Ben is sorry. He forgot he was supposed to pick her up!”. Each item could be uttered with two distinct intonation contours (appearing in different lists) so that content effects were overridden. In order to identify compensatory strategies, Experiment 2 (ASD: n=20, Mean CA=13;10, TD n=20, Mean CA=13;8) used the same material in the context of a dual task (detecting a sound in the target utterance). Finally, in Experiment 3 (ASD: n=16, Mean CA=13;10, TD n=16, Mean CA=13;11), the complexity of the dual task is increased (detecting the number of ‘Ts’ in the target utterance).

Results: Contrary to our predictions, ASD participants were not specifically impaired in conditions requiring higher order mindreading skills. In Experiment 1, ASD and TD participants had similar accuracy rates and reaction times across all conditions. This was confirmed in Experiment 2 despite the increased demands imposed by the dual task. Finally, in Experiment 3, ASD participants showed no ToM-specific impairment in a highly demanding dual task. On the contrary, we observed that they were slower than TD participants in all conditions, which suggests that, when placed under high cognitive load, they struggle to identify vocal cues in general, independently of underlying mindreading requirements.

Conclusions: Our study confirms that people with autism have difficulties dealing with emotional cues in challenging contexts. Yet, our results - together with past empirical findings - show a combination of competences and impairments that is inconsistent with the idea that atypical recognition of vocal cues is caused by impaired ToM.