Understanding autistic communication

Dr Brett Heasman from the Centre for Research in Autism and Education (CRAE) at UCL and Dr Alex Gillespie, Department of Psychological and Behavioural Science at the London School of Economics, discuss their research into how autistic people interact with each other and create shared understanding.

Download a PDF of this article

Authors: Dr Alex Gillespie, Dr Brett Heasman

Understanding autistic communication

One of the curiosities of how autism research has developed over time is that it has traditionally focused on individual autistic abilities in an attempt to explain observed difficulties in social interaction. It’s curious, because social interactions are a two-way process (i.e. involving other people), and many other research disciplines (e.g. anthropology and sociology) have understood human interaction as a socially situated activity for decades. 

To be social requires a combination of:

  • individual skills, such as anticipating what people are thinking and feeling
  • an environment which supports the way people interact, such as the cultural norms which govern how we interpret what is meaningful and appropriate behaviour. 

Understanding one’s social-cultural environment for interacting is important because it includes rules and norms which guide people in how to develop shared understanding. However, we already know that many aspects of what might be termed a “neurotypical cultural norm” (i.e. the predominant way of interacting in culture) are problematic for autistic people. 

A classic example (in Britain at least) is the unwritten rules about managing eye contact – too much in an interaction is deemed obsessive and too little is deemed untrustworthy. For autistic people who may have difficulties processing facial expressions or interpreting non-verbal communication, this presents a challenge – but critically it is a challenge situated in the relationship, not the individual, because it arises from the mismatch of norms and expectations that autistic and non-autistic people use to make sense of each other. 

This mis-alignment between people who hold different norms and expectations of each other has been termed the ‘double empathy problem’ by Damian Milton (Milton, 2012). It has been the subject of a rapidly growing body of research in recent years (Heasman & Gillespie, 2018b; Sasson & Morrison, 2017; Sheppard, Pillai, Wong, Ropar, & Mitchell, 2016) because it underpins many of the reciprocal challenges in social interaction autistic people experience.

Our research: understanding neurodiverse intersubjectivity

We wanted to study autistic interactions in an environment only involving autistic people, where there was no imposition of neurotypical norms (Heasman & Gillespie, 2018a). Our key question was to examine how shared understanding is achieved between autistic people in a naturally occurring daily activity. We also took many methodological steps to reduce the potential for bias in interpreting autistic behaviour from a neurotypical viewpoint.  


We mapped shared understanding between 30 young adults taking part in a naturally-occurring activity at a charity supporting autistic adults. The activity was video-gaming, which was the most popular social activity at the charity, and we filmed 20 sessions. Each session was transcribed verbatim and then each utterance in a conversation was scored in terms of how it related to the previous utterance. 

This follows the principle of dialogue (Linell, 2009) and conversation analysis (Seedhouse, 2004), where each utterance by a speaker is part of a wider sequentially organised unit of meaning-making. In essence, when we communicate we externalise our thoughts and aim to build a shared understanding that both parties have access to, and which informs our future actions. This is known as building intersubjectivity (literally, the bridging of subjectivities; Gillespie & Cornish, 2010).

Importantly, we studied three aspects of how conversations related to each other, because there are many ways in which people may interact with each other. 

For example:

1. Coherence, which relates to how logically connected an utterance is to the previous utterance in a conversation.
2. Affect, which relates to how similar in emotion an utterance is to the previous one.
3. Symmetry, which relates to how assertive or submissive the turn is in comparison to the prior turn.
These three dimensions are common properties of all social interactions. Using the scores we gave each utterance, we were able to develop a map of each interaction over time.


Our study identified two common properties across all interactions that enabled shared understanding to be established. 

  • autistic participants could make very generous assumptions of common ground
  • autistic participants demonstrated a very low demand for social coordination

Generous assumptions of common ground

This may for example involve specific references which would only make sense with certain prior knowledge. Although often leading to many ignored turns, it could also have the social benefit of leading to fantastic rapport.

Example 1: assumptions of common ground (pseudonyms used)

Low demand for social coordination

Participants would not be overly troubled if the interaction became disconnected and turns were not reciprocated. Far from being problematic, this provided participants with the freedom to switch between individual and cooperative ways of making sense of their experiences. 

Example 2: low demand for social coordination (pseudonyms used)

The advantage of these two approaches to conversation is that they enable diverse forms of social relating that would normally be restrained by conventional social norms. It provides the freedom to continually experiment with different ways of connecting with others without incurring a significant social cost if references or turns are not reciprocated. 

Future areas of potential development

For neurodivergent intersubjectivity to work, both parties must have the same broad expectation about communicative style. This raises an interesting question about whether neurodivergent intersubjectivity can be taught to non-autistic people, to reduce gaps in mutual understanding. 

Further research is certainly needed to understand how such patterns unfold in different social settings involving different activities. This can help us to understand, more accurately, what meaningful adjustments can be made to the context (e.g. the perspectives of non-autistic people, the physical environment, and group norms) that can reduce gaps in understanding between autistic and non-autistic people. 


Gillespie, A., & Cornish, F. (2010). Intersubjectivity: towards a dialogical analysis. Journal for the Theory of Social Behaviour, 40(1), 19–46.

Heasman, B., & Gillespie, A. (2018a). Neurodivergent intersubjectivity: Distinctive features of how autistic people create shared understanding. Autism, 136236131878517. 

Heasman, B., & Gillespie, A. (2018b). Perspective-taking is two-sided: Misunderstandings between people with Asperger’s syndrome and their family members. Autism, 22(6), 740–750. 

Linell, P. (2009). Rethinking Language, Mind, and World Dialogically: Interactional and contextual theories of human sense-making. Charlotte, NC: Information Age Publishing.

Milton, D. (2012). On the ontological status of autism: the ‘double empathy problem’’.’ Disability & Society, 27(6), 883–887.

Sasson, N. J., & Morrison, K. E. (2017). First impressions of adults with autism improve with diagnostic disclosure and increased autism knowledge of peers. Autism.

Seedhouse, P. (2004). Conversation Analysis Methodology. Language Learning, 54(1), 1–54.

Sheppard, E., Pillai, D., Wong, G. T. L., Ropar, D., & Mitchell, P. (2016). How easy is it to read the minds of people with autism spectrum disorder? Journal of Autism and Developmental Disorders, 46(4), 1247–1254.

Date added: 15 February 2019



Thu, 07/03/2019 - 20:11

Does this research part have any data on how middle aged autistic adults in their 50's and older as the information that I have read with great difficulity due to my dyslexia seems only talk about understanding autistic communication in children and adults. As an autistic adult of 55 who only diagnosed with ASD in 2014, I feel invisible as when I go to shopping centres staff are only aware of autism in children and not autistic adults especially middle aged autistic adults of our communction and sensory barriers and ways that we communicate.

Nick Chown

Thu, 07/03/2019 - 20:50

It is great to see Alex Gillespie and Brett Heasman mention Damian Milton's double empathy hypothesis. The acknowledgement that communication is a two-way process where autistic and neurotypical people are involved is very important indeed. But I am unsure about their interpretation of the second example of conversation between the two autistic young people. They say that in section 1 of this example Mark and David are '(talking) about two separate topics without responding to each other' and describe the turns in this section as being unreciprocated. (This is important because conversation analysis (CA) states that unreciprocated turns are conversational faults that need to be "repaired".) One could equally argue that 'vehicle attacks' in line 2 is David's example relating to Mark's point that effort has been put into designing the game, and that Mark's mention of 'buildings' in line 3 is his example of where effort has been put in. Just because Mark mentions buildings rather than following up David's mention of vehicle attacks doesn't seem to me to necessarily suggest that they are talking about two separate topics but different examples relating to the same topic. And I don't see this reinterpretation as being specific to autism. I think it would be just as applicable to two neurotypical young people. By the way, I am not saying I am right about this just that I think my interpretation is a valid one. My underlying point is that there remains a tendency to look for something 'wrong' when autism is involved, often unconsciously.


Wed, 08/05/2019 - 17:22

I made an error in the below comment, I meant to say, "the information that I have read with great difficulity due to my dyslexia seems only talk about understanding autistic communication in children and young adults as middle aged adults are being left out of most of the autism research surveys."


Thu, 09/04/2020 - 19:32

Most interesting research work on autistic communication here. The term 'neurodiversity' is often used rather vacuously but this research indicates how autistic styles of communication can function effectively. Much more to be learned from this line of research I've no doubt. Thank you!