Wednesday, September 14, 2011

Simon Baron-Cohen Responds to Criticism from an Autistic Blogger - Part I

EXECUTIVE SUMMARY: Many autistic and feminist bloggers have criticized Simon Baron-Cohen's theories about autism and gender differences, but not very many have had him respond to their criticisms personally. Rachel Cohen-Rottenburg is one of those lucky people; she posted a critique of his E-S theory on her blog a couple years ago, and he has just now written a ten-point response to it. In this part of the post, I talk about Rachel's objection that the characterization of autistic people as poor empathizers ignores the differences between cognitive and affective empathy, and that even autistic people who lack cognitive empathy often have great capacity for affective empathy. I also talk about Baron-Cohen's response to this objection, and how his thought on empathy and autism has undergone a further evolution in the two years since Rachel wrote her post.
Sandrine (a French woman living in Turkey, who has an autistic son) at The Paris Ankara Express pointed me to this guest post by the famous autism researcher Simon Baron-Cohen at the Autism Blogs Directory.

In his guest post, Prof. Baron-Cohen responds to criticism of his Empathizing-Systemizing (E-S) theory of autism from Rachel Cohen-Rottenburg, who writes the blog Journeys with Autism and also curates a web-anthology on Autism and Empathy.

He identifies ten points that Rachel makes in her essay, and writes a paragraph (or two or three) addressing each one. They seem to be split about half and half between questioning Baron-Cohen's conception of empathy (as in, autistic people actually tend to be quite sensitive to other people's emotions, particularly distress) and questioning whether his model is actually any more descriptive, or helpful, than older "deficit" models.

(You could argue --- and many people do --- that the E-S/Extreme Male Brain theory is still a deficit model since it considers a deficit in empathizing skills to be a defining characteristic of autism; it offsets this by also including normal or superior systemizing skills as part of the definition, but if you look at most of the psychological, educational, or behavioral literature on autism, you find all sorts of articles on remediating, or sounding the depths of, autistic deficits and not very much on nurturing or characterizing autistic strengths. So the offset deficit model isn't actually offset all that much, in practice.)

Rachel's post is a response to this paper from 2009 (Rachel's post is also from 2009) spelling out what the E-S theory says about autism. The article describes the E-S theory, particularly its evolution from Baron-Cohen's earlier "mindblindness" theory of autism. He lists five things the earlier model failed to explain that he thinks the newer one does adequately explain: 1) nonsocial aspects of autism, like attention to detail, love of patterns or routines, stimming etc., 2) empathy isn't just "mind-reading", it's also feeling something for the other person, and the "mindblindness" model only deals with mind-reading, 3) autism isn't the only condition that can produce mind-blindness, 4) some studies have failed to find any differences in Theory of Mind between autistic and non-autistic subjects, and 5) autistic strengths.

Rachel responds that, no, actually the E-S theory doesn't do any better at explaining these things. She thinks he totally misreads stimming, for one thing: to her, it's not about "systemizing" at all, it's about self-calming. It's a way to handle emotions, not a form of empirical investigation.

She also takes issue with his contention that we lack both cognitive and affective empathy:

The everyday experience of many autistic people, all across the spectrum, contradicts the professor's theory. Many of us experience such a high degree of empathy that we are constantly putting ourselves in other people's shoes and trying to see all sides in any controversy or conflict. Many of our problems with sensory and emotional overload derive from an excess of this ability, not a deficit.
From my contact with autistic people, it's clear to me that our empathy leads many of us to constantly question the impact of our words. While I am far from perfect, choosing my words carefully may very well rank as one of my Aspie obsessions. However, the professor believes that "the typical 9-year-old can figure out what might hurt another's feelings and what might therefore be better left unspoken. Children with Asperger syndrome are delayed by around 3 years in this skill." (Baron-Cohen, 69)
Baron-Cohen goes on to say that, in addition to not empathizing well, we don't know how to respond to someone even after the person tells us what's wrong.

News flash: Once someone tells me how he or she feels, I don't usually have a problem with an empathetic response. Sometimes, I'll make sure that my response is welcome, out of respect for the other person's boundaries. For instance, if a person is crying, I might ask whether the person would like a hug, or whether the person would like to talk. Some people want hugs, and some people want to be left alone. I consider it courteous to ask. Once I know people fairly well, however, and I know what works for them, I simply respond. Just ask my husband, my daughter, my daughter's friends, my friends, my former co-workers, my neighbors, and all the animals I've ever helped care for in various stages of illness.

I feel like I have to back up now, because Prof. Baron-Cohen says explicitly in his response to her (he even draws a picture!) that he thinks autistic people are only impaired in cognitive empathy, and that we are just as capable of affective empathy as anybody, which is pretty close to what Rachel was saying, too.

His thinking has shifted somewhat since he wrote the paper Rachel was responding to; in that paper, he not only doesn't say that autistic people have affective empathy, but he also implies that we are deficient in both kinds of empathy --- he mentions the EQ as a measure of both affective and cognitive empathy, and also mentions that autistic people tend to score lower on the EQ than non-autistic people:

Most people regard [Theory of Mind] as just the cognitive component of empathy in that it simply involves identifying someone else's (or your own) mental states. ... [M]issing from ToM is the second component of empathy, the response element: having an appropriate emotional reaction to another person's thoughts and feelings. This is referred to as affective empathy (Davis 1994). On the Empathy Quotient (EQ), a questionnaire is filled out either by an adult about themselves or by a parent about their child, both cognitive and affective empathy are assessed. On this scale, people with autism spectrum conditions score lower than comparison groups.
So, while Rachel's objection --- yes, we do have affective empathy --- might seem redundant to someone who has read Baron-Cohen's more recent writing on the subject of empathy (his most recent book, Zero Degrees of Empathy, contrasts autistic people with psychopaths: the former group has poor cognitive empathy but normal affective empathy, while the latter has normal cognitive empathy but poor affective empathy), she is actually bringing up something that is not addressed in the text she's working with.

(I also think there is still a problem in Baron-Cohen's work with measuring empathy --- measuring it at all, much less measuring each component separately --- in that most of the tests he uses that directly measure some aspect of empathy, like the Sally-Anne test, the Faux Pas test, or the Reading the Mind in the Eyes test, are measures of cognitive empathy only. The indirect measures, which are usually questionnaires, have their own problems. With any self-reported measures, there's a risk of stereotype threat, where the test takers' awareness of stereotypes about some group to which they belong biases their answers toward whatever the stereotype predicts: women/girls and members of some ethnic minorities, like African-Americans, Latin@s, or Native Americans, tend to do worse on math tests when they are reminded of their group membership, even by something as innocuous as a check-box at the top of the page for race and gender. Since autistic people are often taught from early childhood that their way of speaking, acting, feeling, or relating to other people is weird and wrong, I would not be surprised if they tended to rate themselves low on things like fitting in, participating in conversations or getting along with other people. Also, a lot of the questions on things like the EQ measure some mixture of cognitive and affective empathy; they ask about intuitively knowing what a person feels and then reacting to it, like knowing to comfort someone who is upset. There is one measure that has subscales geared just toward having intense feelings on other people's behalf, and that is the Interpersonal Reactivity Index, which one study found autistic people scored either very close to, or even better than, non-autistic people on all but one subscale.)

So his position on affective vs. cognitive empathy in autism has shifted somewhat since he wrote the paper Rachel was responding to, and he doesn't explicitly acknowledge that in his response to her; instead, he makes it sound like that's what he was saying all along:

Rachel says "once someone tells me how he or she feels, I don't usually have a problem with an empathic response." This is exactly the point. For most people, they don't need to be told by the other person, "I am upset." They can just read this information in the other person's facial expression, vocal intonation, or 'body language', and they can make inferences about what the other person might be thinking, in the absence of being directly told. For many people with autism these non-verbal cues may be hard to read and instead they may only know how someone feels if they are told explicitly. The evidence for this again comes from many scientific studies documenting difficulties by people with autism in reading the mind in the eyes, the face, the voice, or in action (e.g., film). Rachel's own self-description seems consistent with this: "Now, I will readily admit that I cannot infer a person's mental state by reading nonverbal cues."

But I completely agree that once it is explicitly pointed out, people with autism are very capable of an empathic response. Rachel may be surprised to hear that I agree with her on this one, but it hinges on the distinction between 'cognitive' and 'affective' empathy. Cognitive empathy is the ability to identify another person's state of mind (not just through language) and affective empathy is the drive to respond with an appropriate emotion to another person's state of mind. A growing number of studies suggest that the empathy difficulties in autism are largely restricted to the cognitive component, whilst the affective component is often intact. For this reason, people with autism are often highly motivated not to upset others or hurt others, and are themselves upset to hear that they may have done this if it is pointed out. And once they know that someone else is upset or suffering, they are very often motivated to want to help or offer comfort.

(This seems like a really patronizing response to me. She has addressed theory of mind in other parts of her post; with the passage he's quoting, she is explicitly addressing the "emotional response" aspect of empathy. To make the point that her capacity to form an empathic emotional response is working just fine, thank you, she's describing that faculty in isolation. He ignores that context so that he can go, "Ah-ha, see, she has to be TOLD what someone is feeling!" Well, no, she doesn't always. She's just describing a particular scenario where her cognitive empathy cannot be called into question --- because she doesn't have to use it --- in order to focus on her affective empathy.)

No, I am glad he understands that we care about other people, too. I still disagree somewhat with his assessment of our difficulties with cognitive empathy; that, I think, stems as much from our being profoundly different from other people in terms of our sensory and emotional responses as it does from any objective inability to "read" other people. I think he ignores just how badly non-autistic people fail to notice the signs of our distress, or misinterpret our body language or tone of voice. I think they are just as bad at reading us as we are at reading them; it's just that because they're the majority, their failure to understand us is not as disabling as our failure to understand them.

I had more things I wanted to point out about his response to Rachel, but I've ended up spending so much time untangling this one little snag where they seem to be talking past each other that I think I will split this into multiple posts.


Tom Michael said...

I agree with some of these distinctions between theory of mind skills, and emotional empathy. Prof. Baron-Cohen does actually make these distinctions in the detail of his theories, but fails to make these distinctions in his empathy quotient (EQ) scale. This leads to some confusion, as the distinction is not clearly made in his use of the term empathy, which could cause some people to mistakenly feel that he is equating empathy deficits in autistic spectrum conditions with empathy deficits in psychopathy.

I recently read his latest book, Zero Degrees of Empathy, and have reviewed it on my blog here:

Xanthe Wyse said...

The irony is NT's usually don't understand autistics, so one could say they have no empathy.

I think SBC confuses empathy with compassion & altruism

Here's a post I wrote about SBC's response to Rachel

Cecile said...

I wrote about this yesterday - about how NT'd don't understand us and we don't understand them - why are we lacking something then and they are not? This empathy thing has IMO a lot more to do with differences than disability.

Lindsay said...

@Xanthe, @Cecile - yes, I agree. I think it has more to do with us not having enough in common to accurately predict how people in the other group will react to what we do than it does with us just being generally obtuse.

Lindsay said...

@Xanthe - I think Baron-Cohen does a pretty good job of spelling out that he doesn't think we're bad people, and that he knows we feel compassion, too --- he tends to think of us as well-intentioned emotional illiterates. Which maybe some of us are, but from what I've seen on the Internet, far more of us possess a very high degree of emotional sensitivity. We might not always have the language to describe it, but it's there.

I also think Baron-Cohen's IDEAS about our capacity for empathy have evolved faster than his tool for "measuring" empathy, the EQ. Like Tom Michael above says, the EQ measures both kinds of empathy at once, without giving any information about cognitive vs. emotional empathy.

Rachel said...

Lindsay --

Please note that, in his response to me, SBC provides a wholly inaccurate depiction of the way that he profiles autistics in his latest book "The Science of Evil." The chart that he provides in his response shows autistic people being positive for affective empathy and negative for cognitive empathy. In the book, however, he profiles us as negative for both cognitive and emotional empathy. And his book is relentless in its depiction of us as lacking in empathy altogether. Please see my response to him at:

You might want to correct your post to reflect its account of the profile in his book.

ammar said...

Can you help me to get to the studies or research and a measure of cognitive empathy,
Waiting for reply ..
Ammar - Iraq said...

Hi Lindsay,

Healthline is interested in contributing a guest post to We would be open to contributing any blog that would be of interest to your readers. Healthline bloggers have been featured on a variety of sites including:

Washington Times:
Natural News:

Please let me know if you have any questions. Thank you in advance for your consideration.

Warm Regards,