Valérie-Anne O'Callaghan
For decades, the dominant explanation for social difficulties in autism took a very concrete form: autistic people lack the ability to infer the mental states of others. This was called theory of mind, or rather, its absence. From this emerged entire intervention programs dedicated to teaching autistic people to read emotions in neurotypical faces.
Nobody considered measuring the opposite.
Milton, 2012
Damian Milton, an autistic researcher at the University of Kent, published in Disability & Society An article that reformulated the approach. Its proposal, the problem of double empathy, states something seemingly simple: when two people with very different ways of processing the world interact, the difficulty of understanding is mutual.
The autistic person doesn't fully understand the non-autistic person. And the non-autistic person doesn't fully understand the autistic person. The difference was that only one of the two parties had been examined, diagnosed, and received intervention.
Milton also pointed out something uncomfortable: the asymmetry was not accidental. The majority defines what counts as normal communication, and the minority is described in terms of a deficit relative to that standard.
What experimental research found
The proposal was theoretical and contained a verifiable prediction: if the difficulty comes from the mismatch between ways of processing, then it should disappear when the two people share the same one.
Crompton et al. (2020) tested it in Autism. They designed information transmission chains with eight people, similar to the game of telephone: the first person receives a story and passes it on to the next. They compared three conditions: chains of only autistic people, chains of only non-autistic people, and mixed chains.
The result contradicts clinical prediction. Chains formed only by autistic individuals transmitted information as effectively as those formed only by non-autistic individuals. Information loss occurred in the mixed chains.
If the difficulty lay within the autistic person, autistic chains would have been the worst. They were equivalent.
A second study by the same group found that the perceived quality of the bond depended on neurotype matching, not on being autistic. Autistic individuals reported better rapport with other autistic individuals, and the same was true among non-autistic individuals.
The other half of the problem
Edey and colleagues (2016) published in Journal of Abnormal Psychology A study reversed the usual question. They presented non-autistic adults with animations of geometric figures whose movements had been generated by autistic and non-autistic people, and asked them to interpret the intentions behind those movements.
Non-autistic adults correctly identified animations created by non-autistic individuals but incorrectly identified those created by autistic individuals. In other words, they demonstrated precisely the difficulty in interpreting others' mental states that had previously been attributed exclusively to the other group.
Subsequent studies with video-recorded interactions found something similar, with one added detail: autistic people were just as expressive, with a different expression.
What does this change?
Several things, and not all of them obvious.
The location of the problem. It ceases to be within one person and becomes between two. That doesn't deny that real difficulties exist, but it does change where intervention is possible.
The purpose of the intervention. Teaching only an autistic person to communicate in a non-autistic way means placing all the translation work on them. Training their support network also distributes that burden.
The value of peer contact. If attunement improves with a shared neurotype, then access to the autistic community ceases to be a secondary support and becomes a necessity. The literature on well-being points in that direction.
The cost of misunderstanding. Repeated misunderstandings accumulate. Decades of interactions going wrong without knowing why produce a reasonable but mistaken conclusion: that one is the problem.
What this means in the body
I add the reading that corresponds to my profession, and which the literature on double empathy does not usually develop.
An interaction where one is unsure whether they are being understood keeps the nervous system in a state of continuous evaluation. Porges (2011) described neuroception: the organism automatically assesses whether the environment is safe, before any thought. A conversation in which the other person's signals are difficult to interpret produces signals of uncertainty, and sustained uncertainty closely resembles a threat.
This explains the disproportionate exhaustion after social encounters that, from the outside, seemed perfectly pleasant. The expense isn't in the content of the conversation. It's in the continuous work of translation.
And it explains why many people describe interactions with other autistic people as relaxing. A considerable portion of the expense disappears.
A caution
It's important not to treat this as a definitive new certainty. Research on dual empathy is recent, most studies use small samples and involve autistic individuals without associated intellectual disabilities, and replication is ongoing. The theory has growing empirical support and remains a theory.
It's also unwise to assume that communication difficulties disappear among autistic individuals. Studies show equivalent information transmission and better perceived attunement, which leaves considerable room for improvement.
To conclude
What interests me most about this body of work is where it came from. Milton is an autistic researcher. The question that opened all this up—why only one of the two sides was being measured—is quite obvious from within, but had gone decades unposed from the outside.
And perhaps that is the most useful lesson: who asks the questions determines what gets investigated.
Sources and references
Crompton, C.J., Ropar, D., Evans-Williams, CV, Flynn, E.G., and Fletcher-Watson, S. (2020). Autistic peer-to-peer information transfer is highly effective. Autism, 24(7), 1704-1712. University of Edinburgh and University of Nottingham.
Crompton, C.J., Sharp, M., Axbey, H., Fletcher-Watson, S., Flynn, E.G., and Ropar, D. (2020). Neurotype-matching, but not being autistic, influences self and observer ratings of interpersonal rapport. Frontiers in Psychology, 11, 586171.
Edey, R., Cook, J., Brewer, R., Johnson, M.H., Bird, G., & Press, C. (2016). Interaction takes two: Typical adults exhibit mind-blindness towards those with autism spectrum disorder. Journal of Abnormal Psychology, 125(7), 879-885.
Heasman, B. and Gillespie, A. (2019). Neurodivergent intersubjectivity: Distinctive features of how autistic people create shared understanding. Autism, 23(4), 910-921.
Milton, DEM (2012). On the ontological status of autism: the 'double empathy problem'. Disability & Society, 27(6), 883-887. Researcher, University of Kent.
Porges, SW (2011). The Polyvagal Theory. WW Norton. Neuroscientist, Indiana University.