The Double Empathy Problem – A Paradigm Shift in Thinking About Autism

How many times have you heard something of the following sorts about autism?

“People with autism lack social skills.”

“People with autism can’t communicate appropriately.”

“Deficits in theory of mind is a core symptom of autism.”

“People with autism lack the ability to feel empathy.”

As you can tell from the person-first language, these statements are not ones coming from autistic people. It’s sickening, isn’t it? Especially the last two. At least I hope it is, as I say that knowing there there are a large number of people out there who not only actually believe those things, but continue to advocate for them, publish them in scientific journals, and argue fiercely with anyone who dares challenge these presuppositions.

Of course, the statements above either only tell part of the story from one perspective or are just flat out false. As for the last two, autistic people don’t lack theory of mind (meaning the awareness that other people have minds that are separate from ones own), I can assure you that we are aware that other people have minds. I can also assure you that autistic people can and do feel empathy, in fact we often feel such intense empathy that it is often overwhelming. It’s just that we express it differently. I swear that I’ve written those same two things over and over again yet I have to keep writing it because misconceptions still abound.

As for the first two, one of the overarching themes of what I’ve been writing on this blog is the idea that difference does not necessarily equate to deficit, and so it’s not fair to say that autistic people “don’t communicate,” in fact many of us have spent large portions of our lives screaming out in communication attempts that no one seemed to recognize. And I have written before that it’s not really fair to say that autistics lack social skills, rather we simply have different social skills. In fact, anecdotal reporting and scientific data is every day more and more supporting the idea that autism is not, in fact, a social disorder but rather a difference in sensory processing and integration, with which differences in social-communication and reciprocity are a surface-level symptom. So-called “deficits in social reciprocity” seem to be a good way of identifying autistic people from one perspective, which is why it’s one of two big diagnostic criteria for autism in the DSM-5, but itself they are not the core of what autism actually is, nor does that description alone accurately portray the whole picture.

Differences in sensory processing do of course affect what might be called “social skills,” for example many autistic people don’t integrate incoming visual information in a way that allows them to easily notice the tiny subtle differences in body position and facial expression that are used in neurotypical communication, thus may “miss” so called “social-cues.” Sensory integration also affects motor movements, and so many autistic people may not show the body language that non-autistic people expect for the way they are feeling. Notice, however, that these are simply differences and only come across as deficits from a non-autistic perspective. Autistic people actually communicate really well with other autistic people. From the autistic perspective, it’s non-autistic people who lack social skills.

Thus, we have the observation that autistic people communicate relatively well with and understand other autistic people, non-autistic people communicate relatively well with and understand non-autistic people, but there’s a breakdown in communication between neuotypes. This is called:

The Double Empathy Problem

The Double Empathy Problem is a paradigm shift in the way we, as a society, think about autism. It is the idea that rather than viewing autistic social differences as inherent deficits, that we instead recognize that perhaps non-autistic miscommunication has just as much role in autistic social difficulties as autistic people themselves do. That non-autism is just as baffling to autistic people as autism is to non-autistic people (though autistics have the advantage of living in a world dominated by people not like us, so we quickly get used to it). That instead of insisting that autistic people continuously adapt to neurotypical communication norms, that non-autistic people have just as much responsibility to adapt to autistic communication styles. It is recognizing that people autistic people may communicate in different ways, but this doesn’t mean they aren’t communicating. I’m happy to report that the way of thinking surrounding the Double Empathy Problem is gaining traction not just in neurodiversity and advocacy circles but also in the broader autism community.

Let’s demonstrate the Double Empathy Problem with an informal experiment. You can participate too.

Look at the image below. Notice the crying woman in the middle (the one with red hair holding a tissue). Answer this question: How are the people sitting around the crying woman making her feel? What should the people sitting around her do to make her feel better? Imagine you are the crying woman. (You can post your answers in the comments if you want. Make sure to mention whether you’re autistic or not!)

This exact image was posted on Twitter with the same questions being asked. What is completely fascinating (though expected) and demonstrates the Double Empathy Problem quite well are the differences in responses between autistic and non-autistic people.

Non-autistic responses contained several recurring tropes. Answers included things like “the people around the crying woman are helping to comfort her and are making her feel better. They should stay there and keep talking to her, helping her work through it.”

Autistic responses also contained several recurring tropes, but they are very different from the non-autistic ones. Autistic responses tended to be something along the lines of “the people sitting around the crying woman are overwhelming her and making her feel worse. They should walk away and leave her alone. They shouldn’t all be crowded around her like that, and it’s making the woman uncomfortable that the person next to her is touching her shoulder.”

There’s plenty of empathy in both of these responses. Though, the non-autistic version is probably more stereotypical extroverted “empathy.” But there is feeling for and with another person in both cases. But obviously, there are some clear differences in autistic empathy due to differences in neurology. Whilst non-autistic people may feel comforted by a touch on the shoulder, that could be a serious sensory aversion for many autistic people and would of course make them feel worse. The whole idea of “talking through” problems is likely to overwhelm many autistics who would prefer to be left alone to decrease sensory input so they can calm down and decompress.

Notice that if either the autistic or non-autistic person tried to apply their empathetic actions to the other group, good results would probably not occur. The Double Empathy Problem is a big reason I think so many people have drawn erroneous conclusions about autism. They observe that autistic people don’t show empathy the way they expect, or show the right body language in the right context, or react the way they expect. However, they misunderstand the autistic perspective and apply wild conjecture and guesswork, leading to the conclusions that autistic people lack empathy, emotions, etc.

This is still happening in autism research. Wild uninformed conjecture about autism passed off as fact isn’t limited to psychology of the 60s. Ann Memmott (Ann’s Autism Blog) is an autistic autism professional who does a lot of work trying to correct misconceptions about autism and change narratives that occur in professional fields and autism research. She once recalled on Twitter that she once read a published “scientific” paper where researcher claimed to have found strong evidence that autistic people lack empathy. I think I remember her calling it “The ‘Oops’ Experiment.” How did the experiment work? Researchers took a group of autistic children (because autism is a child’s condition, didn’t you know? /sarcasm/) and individually told them that they were going to do a test/game with a box of matches. The researcher then *accidentally* on purpose spilled all the matches out of the box and onto the floor. If the kids didn’t immediately bend down and start picking up the matches (with zero prompting) then it was said they had no empathy. Never mind the fact that most of those kids probably hadn’t even processed what had just happened before the researcher was writing down on their clipboard “lack of empathy.” Never mind the fact that you can’t expect to determine the entirety of empathy with one silly task with one expected response. Never mind that prompting, one of the most basic and useful accommodations you can make when working with an autistic person because it helps our brains initiate new tasks, wasn’t given, nor was executive functioning nor motor planning in general taken into account. Golly. This is the kind of stuff that passes for rigorous science in peer-reviewed behavioral health journals apparently.

The researchers in question could have corrected their methods and/or conclusions just by interviewing a group of autistic people, especially autistic autism experts. But honestly at this point I’m convinced that, even if subconsciously, a lot of autism researchers view autistic people as literally being less-than-human. They view us like lab rats. That’s the best explanation I can come up with as to why so many are so averse to collaborating with and talking with and listening to autistic people to aid their research. This isn’t an indictment of all psychologists and neurologists who study autism, of course, there are plenty of good autism scientists out there, some of whom are autistic themselves. It’s just a general trend that seems to be occurring; the complete misunderstanding of the autistic perspective and what’s actually going on. This heavily demonstrates the Double Empathy Problem. It’s not just autistic people who misunderstand non-autistics, it goes both ways. This is certainly a two-way street.

The concept behind the Double Empathy Problem has potential application beyond just research, of course. It affects everyday interaction. There’s been some preliminary research (that I linked to on my Facebook page but can’t find it again… links would be appreciated) that suggests that the reason why autistic people are often socially rejected by their peers has very little to do with the autistic person themselves and more to do with how the non-autistic people around them perceive them. It’s not that a lack of social skills is keeping them from making friends, it’s that most people around them are prejudiced against difference and so they reject advances from autistic people and/or have no idea how to actually communicate with the autistic person in a way they understand.

A close up of two people holding hands. One person’s torso is visible and is wearing a black and white striped shirt.

So many resources are put into trying to teach autistic children, teens, and adults “social skills,” often with ethically questionable methods. But if non-autistic people have just as much trouble communicating with autistic people, then why don’t we try to teach non-autistic people autistic social skills? It would certainly make the lives of autistic people much better. What if people likely to interact with many autistic people, like doctors, first responders, and therapists, received training on how autistic communication may be different from their own? Not just general autism “awareness” training, but practical skills?

As per usual, hopefully this has given you something to think about. I like to think. I like to make people think about new ideas. And hopefully more people thinking about the Double Empathy Problem will help to change the narrative surrounding autism, and help people practically interact better with autistics.