OK, so people with ASD lack social skills. But what ARE social skills?

As you may have noticed, I don't feel entirely comfortable with theories about ASD based solely or mainly on social skills explanations. It's not that I think people with ASD don't socialize differently, or that it doesn't cause problems for them or for others. To me, social differences are the most obvious signs of possible ASD: it's really easy to spot a person who makes poor eye contact, enjoys monologuing without asking people questions about themselves or following up on what they say, or periodically interjects quotes from TV shows or books that have nothing to do with the ongoing conversation. It's more that, especially at higher levels, it's not clear how to define "social skills" so we can research them or teach them to others.

Are social skills a species of hypothesis-testing ("social cognition") as many psychologists seem to think, or are they something intuitive and unconscious? If they involve processing information unconsciously, what sort of information do people use to make these sorts of social assessments? How do we deal with the fact that social skills are necessarily situated in a cultural or subcultural context? A crude example would be a gesture that means something different in two different countries, but many books have been written about expressive and interpretational differences between WASPs and Jews in America, or between men and women within the same culture. Is there some broad set of social skills that exists across culture, gender, age, ethnicity, and personality? Or should we define social skills as the ability to adapt to any new social "culture" within which one comes into contact? If we do (which I think we should), how would we test such flexibility or determine whether a person suspected of ASD lacks it?

Finally--and tied to the preceding point about culture--is there a normative aspect to social skills? i.e., suppose there are two people, Abby and Mitchell, who are both equally good at interpreting others' behavior and projecting the sort of thoughts and emotions that they intend. Abby hates conflict, while Mitchell enjoys argument and debate (they energize him). When Abby and Mitchell get into disagreement with others, Abby will tend to behave in ways that avoid conflict and improve the relationships of the people involved, even if it means her own concerns may not be addressed and her frustration continues. Mitchell will seek out argument when it might be more productive to pick his battles. Both of these approaches have their strengths and weaknesses, but how does one determine which is the more "socially skilled" response? If "social skills" are judged by the ability to make people comfortable and have harmonious relationships, Abby will seem more socially skilled. If "social skills" are judged by the ability to achieve one's goals, then either one may seem more socially skilled, depending on whether Abby succeeds in avoiding conflict and Mitchell succeeds in not antagonizing important people in his life. Personally, I think true "social skills" are not normative, but the normative aspect necessarily comes in for parents and teachers when they try to teach a child with ASD how to behave towards others, and is often overlooked.

My biggest misgiving, however, is that a lot of evidence suggests that neurotypicals (NTs), the supposedly "socially skilled" group, really don't have particularly great "mind-reading" abilities. That is, their interpretations of others' thoughts and feelings, and the reasons for their behavior, are usually wrong. For instance, we have a systematic bias to see ourselves in more detail than others see us, and others in less detail than they see themselves. This may affect the attributions we make in explaining others' behavior. And consider the following, from this article:
Strangers (who are videotaped and later report their second-by-second thoughts and feelings, as well as their assessments of their counterpart's thoughts and feelings) read each other with an average accuracy rate of 20 percent. Close friends and married couples nudge that up to 35 percent. And "almost no one ever scores higher than 60 percent," reports psychologist William Ickes, the father of empathic accuracy, who is based at the University of Texas at Arlington.

Furthermore, "past research has shown, for example, that people often do little better than chance when it comes to accurately identifying those who find them attractive, intelligent, or likeable."

Research on the mind-reading abilities of NTs consistently finds that they are biased because they have an egocentric perspective. And yet, most of the time, the average NT manages to make friends, find and keep a romantic partner, get through a job interview well enough to get a job, and get along with coworkers well enough to keep a job. I don't think this is because any but the most perceptive NTs have a particular knack for understanding others' thoughts or feelings, either on an intuitive or an intellectual level. Like people with ASD, they are stuck in their own heads all the time. The difference is that the majority of people think, feel, and express themselves similarly enough that they can guess accurately--at least, up to 35% of the time, which apparently is good enough. When they meet NTs of other genders, personalities, cultures, or nationalities, they stumble a bit, but after all, "it's a small world after all," the fundamental neurology is largely similar, so after some false starts, most can learn to "translate" from one social language to the other.

What makes people with ASD--at least, the "higher functioning ones"--different from NTs is that, from basic sensory processing on up, their brains don't work the same way. They feel the same emotions in different situations and use different body language to express them. They find very different things interesting or motivating. The differences are real, and so vast that a person with ASD may see himself, and be seen by NTs, as being from another planet. Conversation depends on common ground and there is very little common ground. The difference may be great enough that, without any difference in either social skills or egocentric perspective, people with ASD cannot learn to translate.

The evidence for this view is that NTs are often no better at understanding people with ASD than people with ASD are at understanding NTs. If NTs were really better at translating from one very different "social language" to another, that would not be the case.

I know, the idea that NTs have as big a stumbling block as people with ASD in their mutual interactions is probably shocking and seemingly unlikely, especially if you yourself are NT. But consider the autistic blogger ballastexistenz's description of what she, and NT members of the MIT Media Lab, observed when they watched the "Autism Every Day" video:
instead of watching it straight through, we stopped it and focused on the social behavior of the children in the video, and the parents in the video. The interesting part to me was that the social behavior of the children was not only often invisible to their parents, but often invisible to the people who worked at the Media Lab as well. I had to point out to them things like one child speaking to her mother and inquiring about her mother’s emotional state, another child’s affection, another child looking up at his mother’s face to gauge her feelings. We concluded that somehow through the camera person focusing on the mothers, combined with the mothers focusing on the camera people, the viewer’s focus was not on the social overtures of the children, who were then possible to describe as not engaging in social overtures even when they were very clearly affectionate, social, and concerned with their parents’ feelings.
So again, how do we measure innate level of social skill in this context? This is a context where autistic people’s parents are somehow (possibly by training from doctors, possibly through instinctively looking for a different set of social cues than the ones we use, possibly because of some other construction they have in their heads that overrides what’s in front of them) clearly not noticing our social approaches or our concern for them.
Ballastexistenz points out that the sort of overtures NT parents expect from their children are a sensory nightmare for children with ASD. Eye contact is often uncomfortable and overwhelming, making it hard to think. As this blogger puts it, "Imagine growing up somewhere where to be hit upside the head and locked in a room with a large predatory animal are the two highest forms of affection, and your emotional development is gauged on how well you learn to put up with those situations. To people who experience certain kinds of touch as pain and eye contact as a predator-style threat, that is some part of our experience growing up." This may sound melodramatic if you've never experienced it, but many first-hand reports by people with ASD experience similar discomfort (for instance, see here and here).

Lastly, if people with ASD were truly "mind-blind," they would not even be able to understand one another because they would, like children who fail to pass the false belief test, assume everyone else had exactly the same knowledge they did, whether autistic or NT. But, at least according to self-report, they can have comfortable, rewarding interactions with other people with ASD. Same principle as NT-to-NT socializing: we understand those who are like us, with whom we share common ground.

Given both the extreme sensory differences of people with ASD and the rather profound lack of social skills shown by the supposedly normative NTs, theories of autism based primarily on social symptoms are misleading.


  1. 若對自己誠實,日積月累,就無法對別人不忠了。........................................

  2. I love readding, and thanks for your artical. ........................................

  3. Thanks for commenting, I'm glad you liked it! It's so exciting to have readers halfway around the world.

  4. Thanks, Roh.

    Just for future reference, do you mean that you thought it was intellectually interesting or that it made some sort of connection with your life?

  5. 卡爾.桑得柏:「除非先有夢,否則一切皆不成。」共勉!..................................................

  6. "Very interesting post!!!!Thanks for the information!!!!!!!..................................................................

  7. Update: NTs are supposed to excel at interpreting ambiguous sentences, in large part by using background knowledge to interpret them. If computer programmers can be believed, NTs even prefer ambiguity. But this study found that NTs understood their spouses' ambiguous sentences no better than they understood strangers', despite presumably extensive knowledge about their partners. http://www.indianexpress.com/news/couples-communicate-no-better-than-strangers-study/740062/

  8. Hi, I have just arrived from Flappiness (hence the tardiness of the comment) and found this article very interesting.

    Coincidentally, I have just written a university paper which postulates something similar - that Theory of Mind (or lack of it), and the "Triad of Impairments", are as much or more applicable to NTs involved with autism as to autistics themselves. Your article has given me a lot of useful references on this theme, and I will pursue it in other work - thank you!

  9. Hi, thanks so much for commenting! Your paper sounds interesting--could I read it? mosaicofminds AT gmail