Guest Blog: Researching autism with Conversation Analysis

Conversation Analysis (CA) is a particularly attractive research tool for investigating the communication of people with autism: its careful, moment-by-moment method picks up just how it is that their interactional moves fail (and succeed). In this very welcome guest blog, Tom Muskett explains the promises of CA in working with the complexities of this often distressing disorder.

Tom Muskett, now at Leeds Beckett University

Tom Muskett, now at Leeds Beckett University

I was a speech and language therapist with an interest in working with children on the autism spectrum. By definition, this diagnosis is associated with so-called deficits in social interaction and communication, but I had become increasingly dissatisfied with the accepted representations of these, particularly within research and clinical writing.

On one hand, most of this work seemed to argue that the linguistic and interactional behaviours associated with autism can be modelled as the straightforward and inevitable surface manifestation of underlying cognitive / neuroanatomical / genetic deficits.

Who’s got the ‘autism’?

But on the other hand, I was spending time with real children who, regardless of diagnosis, were variable in their interactional behaviour depending upon who they were talking to, where, and to do what – with this variability also evident in any behaviours that were (diagnostically) describable as ‘symptomatic’ of autism.

In fact, I began to feel that the asocial, atomistic and hyper-rational understanding of “communication” that underpins most accepted theory about the condition was more ‘autistic’ than the children themselves!

Unsurprisingly therefore, I first began to use CA to analyse recordings of myself doing practical things with children with autism. Through CA’s methodological gaze, the language, communication and other manifest issues that had been of clinical interest to me were recast not as individual reflexes of underlying disorder, but instead as socially and contextually situated. And of course, once behaviour is viewed in context, many presuppositions about its meaning can be challenged.


For example, in one of our papers, my colleagues and I analysed interactions involving myself and Jennifer, an 8 year old girl with a diagnosis of autism. Jennifer’s interaction style could sometimes feel repetitive, rigid and inflexible – terms often cited as dispositional tendencies associated with autism. But when we examined the interactions involving Jennifer, we identified that these ascriptions were likely to be a consequence of her active (albeit unusual) use of a particular interaction device (the “d’y know what?” question) to retain the conversational floor following certain kinds of prior turn by me.


Strikingly, she used this device when I, as the adult participant, made interactional moves that did not quite fit my positioning in our activity (e.g. when she apparently telling a story, but I tried to turn it into play instead – see the example extract above). Hence, analysis indicated that what could be considered on one hand an individual dispositional tendency actually played out as a dynamic phenomenon that was variably emergent from the moment-by-moment twists and turns of social interaction.

CA findings imported back into practice

Evidently, such findings can straightforwardly come back into practice – in the above case, enabling reflection on aspects of my conduct that jointly contributed to the emergence of Jennifer’s apparent ‘inflexibility’. But there are other powerful overlapping stories to be told here – about adult/child interaction, cultural representations of children’s development and dis/ability, and the definition and psychologisation of social problems. It is striking that CA can generate empirical accounts of autism that sit remarkably comfortably alongside theoretical discussions of these kinds of big issues within critical psychology and disability studies. In fact, I consider that mobilising CA to destabilise accepted thinking about complex constructs such as “autism” could itself represent a powerful contribution to practice and policy.