A B S T R A C TSemi-structured interviews were conducted with 12 highfunctioning adults on the autism spectrum, in order to examine the nature of their personal experiences of music. Consistent with the literature on typically developing people's engagement with music, the analysis showed that most participants exploit music for a wide range of purposes in the cognitive, emotional and social domains, including mood management, personal development and social inclusion. However, in contrast to typically developing people, the ASD group's descriptions of mood states reflected a greater reliance on internally focused (arousal) rather than externally focused (emotive) language.
A D D R E S SCorrespondence should be addressed to: RO RY A L L E N , Department of Psychology, Goldsmiths, University of London, New Cross, London SE14 6NW, Over the last two decades, an increasing number of experimental studies have explored an association between autism and absolute pitch, autism and savant skills in music, and autistic traits in musicians with absolute pitch (Bonnel et al., 2003; Brown et al., 2003; Heaton, 2003; Heaton and Wallace, 2004; Heaton et al., 1998;1999b;2007; Mottron et al., 1999; Nettelbeck and Young, 1996; Treffert, 1988; Young and Nettlebeck, 1995). However, many of these studies report findings from prespecified subgroups of individuals, test specific components of music (e.g. absolute pitch), and do not therefore throw light on the nature of musical experience in the wider ASD population.Experimental findings that generalize to broader populations of individuals with ASD have explored sensitivity to emotion in music. For example, Heaton et al. (1999a) showed that children with ASD understood the affective connotations of musical mode sufficiently well to be able to pair schematic representations of happy and sad faces with extracts of music in major and minor keys. In a more recent investigation (Heaton et al., 2008), typically developing 4-to 10-year-old children, children and adolescents with Down syndrome and those with ASD matched musical extracts with pictures denoting a range of affective and non-affective scenarios; unimpaired performance was again noted in the ASD group. However, a limitation of both of these designs is that they specifically tested the ability to make conventional musical associations, and provided only limited insights into the nature of the participants' personal experience of music.Other, non-experimental, investigations into the broader impact of music in people with autism have explored the value of music in a therapeutic context. Recent studies (Boso et al., 2007; Kern et al., 2007) have shown benefits, and a meta-analysis (Whipple, 2004) claimed that music therapy is demonstrably effective in improving the condition of children and adolescents with autism. For example, Wigram and Gold (2006, pp. 540-1) found some evidence of positive effect on interpersonal communication, reciprocity and relationship-building skills. However, a number of methodological criticisms hav...