Friday, June 30, 2006

Is The Bear Exhibiting 'Theory of Mind'?

The other day, the Bear, Momma Bear and I were watching a video together – Baby Einstein’s Baby da Vinci™. In one scene there is an owl, located in front of some trees, who hoots a couple of times. A second owl comes up within the trees, unseen by the first owl. The second owl hoots, and then ducks and hides, while the first owl turns towards the sound, looking for its source. The second owl appears in a different location - again unseen by the first owl - watches the first owl for a couple of seconds, and hoot and hides again. This occurs a third time, and then while the first owl is looking in the last place he heard the hoot, the second owl 'walks' up behind him, taps him on the shoulder, and they both giggle. Fade to black.

The Bear thinks this is absolutely hysterical. She giggles away, gets excited, flaps, climbs up on the table we put in front of the TV to keep her from sitting too close to the screen (sigh), giggles and flaps again, and then climbs down and walks over to Momma Bear or myself to either pull our hand or even hand us the remote to rewind the scene so that she can see the whole thing again. This is repeated over and over until she tires of it.

A subsequent scene involves two hippos. The first hippo yawns. The second hippo begins a yawn, and appears to produce a noise like an elephant’s trumpeting. Both hippos are startled. Again the first hippo yawns. The second hippo yawns, with the same elephant sound as a result, and the same sense of surprise. Then the second hippo yawns first, and yawns normally. The first hippo yawns and produces the same elephant sound. Both hippos look at each other, shake their heads, and walk away. Once they are gone, up pops an elephant, giggling and trumpeting away, pleased with its joke. Fade to black.

In this example too the Bear cracks up, and while she is not as insistent on getting us to rewind this segment to watch it again, we end up repeating the scene a few times.

I’ve also described in a previous post an example of playing peek-a-boo with the Bear, in which she clearly outsmarted me (not that you have to get up too early in the morning to accomplish this) by turning the tables and sneaking up on me.

It struck me that the common thread between these three occasions was that the Bear was able to understand the difference in perception between the various characters (in the latter case the characters being the Bear and me), i.e. that each saw the world differently, and drew humour from the differences in perception. Then it occurred to me, are these not demonstrations of Theory of Mind?

From Wikipedia, Theory of Mind is generally described as:

“a specific cognitive capacity: the ability to understand that others have beliefs, desires and intentions that are different from one's own.”


“This theory of mind covers two separate concepts:
1. Gaining the understanding that others also have minds, with different and separate beliefs, desires, mental states, and intentions
2. Being able to form operational hypotheses (theories), or mental models, with a degree of accuracy, as to what those beliefs, desires, mental states, and intentions are.“

The Bear has very recently turned three (Happy Birthday Bear!), so I should not be surprised if she did not yet demonstrate Theory of Mind.

A common ToM test is the Sally-Ann test:

"The experimenter uses two dolls, "Sally" and "Anne." Sally has a basket; Anne has a box. Experimenters show their subjects (usually children) a simple skit, in which Sally puts a marble in her basket and then leaves the scene. While Sally is away, Anne takes the marble out of Sally's basket and puts it into her box. Sally then returns and the children are asked where they think she will look for her marble. A child is said to "pass" the test if he realizes that Sally will first look inside her basket before realizing that her marble isn't there. This is based upon their developing the notion that she "cannot watch..."

"Normal children under the age of four
(emphasis added), and most autistic children (of all ages), will cheerfully and confidently answer "Anne's box"-- they do not conclude that Sally cannot know that her marble has been moved."

"Children who pass the test (presumably) understand that there are two different sets of beliefs:
• their own beliefs, based on what they have personally seen, heard, remembered, imagined, reasoned, etc., and
• the beliefs of others, based on what they have seen, heard, etc.."

Those who fail the test are said by some psychologists to lack a ‘theory of other people’s minds’, although there are problems with the test (discussed in the link) that may call this interpretation into question.

In the three examples above, does the Bear exhibit an understanding that the various characters see the world differently?

In the owl example, while she has a clear sense of excitement and anticipation, she does not react when the first owl hoots. She also does not react when the second owl repeatedly appears. It is only when the second owl hoots and hides that she really starts to laugh, suggesting – to me at least - that she clearly understands that the joke is the deception of the first owl by the second owl. She is also amused when the joke is revealed, but not as much as when the prank is being executed. To understand this scene, I’d suggest that she has to understand that the first owl has a different perception of the world than either the second owl or the viewer.

In the second example, she gets excited when the hippos appear to produce the elephant sound, and when the elephant appears, but she does not appear to have a particularly strong reaction when the elephant trumpets triumphantly (i.e. I don't think it is the elephant trumpet itself that she is reacting to). This example is a more open to question, in that she may be reacting to the incongruity of the elephant sound emanating from the hippo, but I really think she is getting the joke, that the elephant is deceiving the hippos (to whom the joke is never ultimately revealed) and that therefore the two sets of protagonists see the world quite differently.

In the third example, I’d suggest that the Bear understands that I am visibly ‘sneaking up’ on her, and by surreptitiously sneaking up on me she understands that she and I do not perceive the world in the same way. Otherwise, there would be no point in sneaking up on me because I would know that this is occurring.

So, am I overanalyzing, or is the Bear demonstrating Theory of Mind, based on the notion of ‘cannot watch’ mentioned above? If she is, then this demonstrates that a child diagnostically labeled as ‘autism, at the severe end of the spectrum’ is capable of Theory of Mind, and at a young age. Given that 'normal' children don’t usually pass the Sally-Anne test until four (which may be an issue with the test as much as with their cognitive abilities), this would be an interesting result.

If she is not yet demonstrating Theory of Mind in these examples, I would not interpret this as meaning that she is not capable of this ability and/or would not be capable of demonstrating it in the future. Any issues she, and other autistics, demonstrate in this area could presumably be related to sensory integration issues as much or more than due to cognitive weaknesses. If one has difficulty processing incoming stimuli, especially input from other peoples’ faces, then it stands to reason that the ability to analyze the expressions of others - as cues to their state of mind - may be impacted long before this ability fails at a cognitive hurdle.

Those are my thoughts. I’m interested in hearing yours.


Camille said...

HI, Ian.

This is a great post.

There have been some very serious challenges made to the interpretation of the Sally/Ann test.

Let me try to give you a description of some of them. There are papers that explain these, but I don't have them, they would be from developmental psych and "human development" journals that I don't usually follow... I found out about these challenges to Theory of Mind (ToM) through discussions with professors who are expert in ToM type questions.

One problem with many of the Sally/Anne tests is the way the questions are phrased. If you try to describe a situation in which one person knows something another person does not know, it's kind of hard to do that without using very complicated grammar. When the questions are posed with less complicated grammar, kids who fail the complicated version can get the simple version.

It's hard not to write a "leading" question. Kids will pick an answer that they think you want them to pick.

In doing the Sally/Ann test with 3, 4 and 5 year olds there is a problem. When the tester asks the child "what does Sally think Ann thinks"... some kids will say, "Now this grown-up knows the answer, and the answer is obvious to me, but the grown-up is asking like she doesn't know... so I will say ____ because that is the answer that the grown-up seems to want." This has shown in some kind of controlled way.

Deaf kids fail the Sally/Ann test as badly as autistic kids do. Deaf kids don't have problems with ToM, they have problems with understanding complex grammar. This is described in something that Dr. Gernsbacher wrote... I can't remember which paper it was.

I had a social psych textbook that I sold back to the bookstore, in it there was a description of an ideal kind of classroom where the kids were taught to work like a team all the time, to be valuing each other and each other's knowledge. This paradigm was tried out with some poor inner-city kinds of kids to see if it would change the way they acted toward other kids. They administered a ToM test at the beginning of the school year and then later. The kids did much better on the ToM test later and the scientists apparently said the kids had developed empathy. But... the kids also got lots of high quality teaching during that time and might have become better at understanding the complex grammar of the ToM questions. At any rate, the kids who were failing ToM at the beginning were like 4th grade age and not autistic. So how specific is the problem of "ToM" to autistics?

It's not.

One of the big problems with talk about ToM is that they will test a little slice of a child's cognition and that doesn't give one the whole picture.

When you watch your child and she imagines that you are thinking a particular thing, then you know she knows you have a mind separate from hers. If you watched her super carefully, you might see that she is responding to you differently according to whether you are in a good mood or bad moood. It's not likely that she thinks of you as a moving piece of furniture.

There could conceivably be problems with "empathy" among autistics, but is it any worse that it is for NT's?

Is one of the problems just that the autistics are not showing their response on their faces, or that they are overwhelmed by emotion or senses and so not able to offer "empathy" in real time?

Is one of the problems with autistics showing empathy that they so rarely are shown empathy? (in general, not your child)


The big problem with saying that autistics don't have ToM is that ToM is one of the key traits of being a human. It is used to describe what humans have and animals do not have. Animals may observe, smell you, hear you and react, but most people think that animals can not think about what another animal or person is thinking.

Almost all the work in ToM in autism, and this work is borrowed by those who work in brain damage... has been done by a small group of scientists in England who all agree with each other, it's like they are all a little fan club of each other... Uta Frith, Simon Baron-Cohen and Francesca Happe. If you read other work on ToM you find these guys cited as authorities... seems a little flimsy to me.

People seem to be in a big hurry to dehumanize autistics, this accusation of no ToM is also launched at stroke victims with "right hemisphere syndrome." So any of us could face a day when we are impaired and people say that we have no empathy and no ToM and therefore are the same as a dog. How does that impact the likelihood that we will be abused, and at least, not be valued as a human?

I would say your daughter has very nice Theory of MInd, and she sounds fabulous all 'round.

Kristina Chew said...

There's a good, and critical, analysis of the Sally/Ann test in Send in the Idiots by Kamran Nazeer, who is autistic. He also points out the huge pressures of the testing situation and offers an interpretation from the point of view of an autistic child who is trying to get the right answer (I can see Charlie in this).

Ian, I do think Bear is responding to the owl and hippo (elephant) deceptions as....deceptions. That she sees that the joke is all about the misplaced expectations. Charlie used to think some Teletubbie segments were hilarious for the same reason----Po would have the Dipsy's hat on or something like that. But also, he is well aware of how to trick us, like Bear in Peek a Boo. This is not the happiest of examples but sometimes he'll shake his head as if to knock it, pause, note us looking at him, laugh, and stop. Joke's definitely on us!

Anonymous said...

I also agree that in order to actually "test" for theory of mind a lot of receptive language is needed by the child since it involves the processing of complex grammar.(and use of it by the child) There are some parents who think that Simon Baron Cohen and others are wonderful, yet their children probably did not have a lot of language deficits to begin with so they did well on the theory of mind "drills." These parents also act as if their children are recovered and much more "normal" because of the "theory of mind" existence in their children. Keep in mind, when I read stories of parents who push the importance of theory of mind, they present as people who will never accept their children unless they recover from autism, so I am leary of them. Basically, I think if these programs are going to be run, a great deal of language and understanding of language is needed, but it does not mean that your child does not have "theory of mind" now. She absolutely seems to.

Ian Parker said...

Hi All,

Thanks for the great comments.

Hi Camille. Thanks for the compliment.

I was aware from the Wikipedia article and before that from the Gernsbacher site (I love that site - Thank You again, Michelle Dawson, for drawing that site to my and others attention) about the language issues with ToM testing, but had not come across the ‘willingness to please’ or answer what the child thinks is expected rather than what they think is correct. That makes perfect sense but is also ironic, in that in trying to figure out what the experimenter wants to hear, the child is demonstrating ToM in concept (i.e. the experimenter thinks differently from me) if not in practice (misreading what the experimenter sees as the correct answer).

I first heard about ToM from Gutstein’s RDI-related book, 'Solving the Relationship Puzzle'. Prior to reading this, it never occurred to me that the Bear would be any different from anyone else in terms of understanding others. I figured that people quickly understood that others perceived the world differently, but that their ability to ‘read’ others correctly varied between individuals and developed with practice and experience, albeit with some being more capable than others in learning this skill.

I was very pleased to read (or was it see?) Gernsbacher’s criticism of ToM deficits in autistics, and her view definitely squares with the reality of what I see in my daughter. As I’ve said elsewhere, I really see no issue with autistic thought (and as I also said elsewhere, I too am at minimum part of the BAP and probably share this thought), and believe that any issues the Bear faces in this regard are likely due to SI rather than cognitive issues. I also believe that understanding ToM is an important component of understanding autism (e.g. any intervention takes a different form if ToM difficulties are due to practice and experience rather than cognitive inability). As such it bothers me a) that so many researchers may have this one wrong, and b) what ‘why they got it wrong’ may say about them and their perceptions of autistics (especially when casual observation by amateurs such as myself can see things that they do not).

And yes, she is “fabulous all 'round.”

Hi Kristina. I haven’t read 'Send in the Idiots', but I’ve heard enough good things about the book that I will order it. I can also see how the test environment could also be stressful and negatively impact results.

I’ve read examples before that indicate that Charlie has ToM. Perhaps we as parents need to get more ‘explicit ToM’ examples in circulation to help refute this notion that our children are not capable of ToM, as distinct from potentially being not as practiced in it (if we can even go that far). As I’ve read Amanda saying, we can’t assume that autistics (including our children) are not watching and understanding.

Hi Anonymous. In the Bear’s case she has no words that we can perceive (which is why we are pursuing PECS as a means of communications), but she clearly demonstrates a capability for receptive language. The absence of words makes it difficult to know exactly how much understanding currently exists, but I have no doubts that she is fully capable of learning. I take the approach that she understands much more than I can see (except in cases of potential danger, when I assume nothing). I don’t know if she has the understanding of receptive language required to pass a formal ToM test, and even if she does, that does not mean that her SI issues will co-operate when she is being evaluated.

Regardless, I see her ToM capabilities and potential as innate, and was surprised when I heard about perceived ToM deficits in autistics. That is why I’m now watching (although not that intently) for evidence that the theories are wrong. I definitely do not see the Bear’s ToM capabilities as being linked to ‘recovery’, except to the extent that an improvement in her SI capabilities may make it easier for her to integrate information and ‘practice’ this capability.

And to all, thanks again, it is nice to hear from others that my perceptions on the Bear’s ToM capabilities may be correct.

Wade Rankin said...

Fascinating post, Ian. (And as rare as I say this), I'm in agreement with Camille. ;-) Most testing of the kind she describes is so subjective in nature that any reported conclusions have to be taken with a grain or ten of salt.

Ian Parker said...

Hi Wade,

Thanks. This is one of those areas in which we all have a vested interest, in having research not underestimate or short-change our children or all autistics.

The Bear has been regularly underestimated - including by me at first, before I began to gain an understanding of both her and autism ('understanding' is still a work-in-progress). She continues to surpass expectations (and by that I don't mean 'discounted' expecations), and ToM is just one more example.

María Luján said...

Hi Ian
You are so right. Our children always surprised us. Also it surprises me many times how autism seems to be misunderstood by many of the scientists that are in the field, considered experts in the topics related to ASD, being ToM one- as you say-but also about emotions and empathy , face processing or non-verbal communication and lately in my son, implicit learning.
Again, thank your for your excellent post.
María Luján

Ian Parker said...

Hi María Luján,

Thanks. Given how different my daughter is from the autistic 'profile' that she is supposed to be adhering to, sometimes I wonder who it is that the researchers are studying?

Frog's mom said...

Some thoughts from the neurodevelopmental standpoint from conversations with Judith Bluestone- the supposed lack of empathy or lack of ToM may have more to do with an autistic person's prior experiences and preceptions under the influence of low muscle tone and/or incomplete inforamation from the proprioceptors. It is hard to recognize the feeling someone else is having from the look on thier face if you've not "experienced" the same look on your face. In other words, due to low muscle tone or low propriocetive input, you may not associate your feeling or experience with your body position and hence to not recall the feeling or emotion by "seeing" it on someone else. This may also account for the development of fewer mirror neurons in autistics. I also recall Lucy Blackman (Lucy's Story, Autism and Other Adventures) posing the question - Why, when I don't know what you're thinking, I lack Theory of Mind, but when you don't know what I am thinking it's because I am autistic?

Ian Parker said...

Hi Frog's Mom,

Thanks for stopping by and for the insight.

I'm also a believer that many of the autistic 'deficits' are experience-related, i.e. that SI issues can cause difficulties in processing and understanding certain inputs and make acting upon them difficult. The result is deficits in experience and learning that may sometimes be misinterpreted as the absence of even latent capabilities.

That is an interesting explanation for the deficit in mirror neurons too. Is it cause or effect?

Frog's mom said...

Now we are getting into theories by little ol' me, but I think it could be both cause and effect. You start with a brain that has a set number of mirror neurons (may be more or less than "normal"). With incomplete, limited or unprocessed sensory information the person does not "recognize" the action of another as something he has experienced, and the mirror neurons that exist do not fire correctly or often. Wouldn't this limited activity in the mirror neurons discourage growth of more mirror neurons and result in pruning of existing, but inactive mirror neurons? This in turn would lead to continuing and increasing difficulty with "empathy" and "ToM" as they are influenced by mirror neurons. If this theory were true, improvements should be seen by addressing low muscle tone and proprioception in young, developing minds.

This conversation came up with Judith when, at her lecture, she was promoting the idea of "mental rehersal" - visualizing yourself or watching someone else do something. I asked her how this would work for someone with low muscle tone who had difficulty associating the visual with the feeling. She told me that mental rehersal done while muscle tone was weak wouldn't strengthen low muscle tone, but it would make it easier to learn the activity once muscle tone was stronger.

Little frog did not like the face tapping (I think it had to do with sinus pain) She told me to tap my own face while he watched. He would cring when I tapped my own face - experiencing referenced pain - and could hardly stand to look at me. How is that for empathy / ToM.