Anemone's Research Emporium

A Brief History of Theory of Mind (ToM)

Part 2: Enter the little children.

Anemone Cerridwen
April 17, 2009

It is not necessary to have any fancy tests to see if children have theory of mind. All you have to do is listen to what they say. Any statement attributing wants or beliefs to another entity, e.g. "the dog wants out", is sufficient evidence that the child attributes mental states to other beings. Distinguishing between something done on purpose and something done by mistake is also diagnostic. However, once you have a fancy test you can try out, of course you're going to use it.

The standard false beliefs task was first suggested by three different researchers in response to Premack and Woodruff's (1978) paper on ToM in chimpanzees (see history of ToM part 1). Chimps don't talk, so in order to determine if they have ToM, you have to set up your experiments carefully. For those that didn't find Premack and Woodruff convincing, a more complex test involving two chimps, one who is in the know and one who isn't, would be more convincing.

Heinz Wimmer and Josef Perner (1983) tried this out on children. They had two scenarios that they acted out with paper cut-out dolls and a mock room with cupboards. In one, Maxi's mother puts chocolate in one cupboard when he's present, but moves it once he leaves. He then goes to look for it, or in other versions he lies about where he thinks it is in order to keep his brother from getting it, or tells the truth about where it he thinks it is so his grandfather can reach up and get it for him (the original location is out of reach). In the other scenario, a little girl hides a book so she can finish reading it, but while she's out of the room the teacher puts it back on the shelf where it belongs. In both cases the main character thinks the object is in one place but it's really in another. What they believe is wrong, hence the term "false belief task". Children were asked where the chocolate/book was, where it used to be, where the character thinks it is, etc. Wimmer and Penner found that the ability to understand the problem and answer all the questions correctly emerged between 4 and 6 years of age.

Since then, other researchers have created alternate versions of the test. A common one is the Smarties test, in which the child itself is first deceived. This test has a lot less information to keep track of. You show a child a box or tube of Smarties, "a desirable candy highly familiar to all subjects" (Perner et al., 1987, p.133). (Though when I did this test with a three-year-old acquaintance as part of a school assignment in the early 1990s, he didn't recognize it. His mother, looming over me, informed me that he wasn't allowed to eat candy, so of course he wouldn't know. I was mortified - me, a scientist-in-training, pushing junk food on young impressionable minds. Americans: Smarties are like M&Ms, but Commonwealth.) You shake the container and ask what's in it. The kid says "Smarties!" in eager anticipation. (Or "I don't know" in polite puzzlement.) You open the container and some or all of the Smarties have been removed and replaced with pencils or something else that makes a noise when shaken. The child realizes it's been fooled. Then you ask what another person who hasn't seen inside the container would think. Most three-year-olds say "pencils". They often say that that's what they thought themselves the first time, and they're serious. They don't remember being fooled. Most five-year-olds say "Smarties", and remember being fooled themselves. Four-year-olds are somewhere in between in terms of ability.

Three-year-olds have problems with false belief tests no matter how easy the researchers make the problem for them (Perner et al., 1987), although how the questions are phrased can make a difference, since children are still acquiring language at this age (Lewis and Osborne, 1990). And children obviously understand pretence (e.g. in pretend play) from a much earlier age. What's more, children acquire a number of parallel abilities at the same age, typically between the fourth and fifth birthdays (see Astington, 1993, for a summary of most of these).

• They learn to play hide and seek properly, for one thing. A three year old will hide like an ostrich with its head in the sand: if I can't see you then you can't see me. Five-year-olds will realize that you might still be able to see me even if I can't see you, and will take this into account in their hiding strategy.
• They start acquiring basic seriation: big, bigger, biggest.
• They start acquiring basic classification. This includes that there are two sexes. Up to this age, kids think you can change from boy to girl or vice-versa by changing your appearance. Five-year-olds, on the other hand, understand that boys are boys and girls are girls. They can be quite sex role stereotyped at this age. (It wears off as their logic gets more sophisticated.)
• They can understand that something can be one thing and look like another at the same time: a sponge can be a sponge and look like a rock simultaneously.
• They can understand that a picture of a turtle that is right side up for them is upside down for you if you're on opposite sides of the picture.
• They can deduce how a balance beam will settle based on simple comparisons between the two sides (Siegler, 1978). Before this, they score according to chance. At this point, they get the simplest situations right, but answer according to simple comparisons (one variable only) even for more complex situations. Kids don't get this completely right until formal operations (abstract reasoning) in early adolescence.
• Autobiographical memory generally kicks in at this time: that is, the ability of me today to remember what I did yesterday. This is holding two mental states in mind at once, in that remembering (holding in mind) today is one mental state and what I did yesterday is another one.

All this suggests that what is happening between the fourth and fifth birthdays is not that children acquire theory of mind per se, but that they acquire the ability to compare two mental states at the same time. Before this they have theory of mind, but only for one mental state at a time. Think of it as being like an upgrade to a computer's operating system. Computer OS's used to be able to work with only one program at a time (e.g. word processing or spreadsheet or graphics, but not more than one at a time). When Apple brought out System 7 in the mid-1980s, you could all of a sudden have more than one program open at the same time, which not only allowed you to copy data between programs, but also meant you could have desktop publishing programs that combined data from a variety of programs. What happens between the fourth and fifth birthdays is a system upgrade. Now you can process two mental states (two theories of mind) at the same time. Woo hoo!

In part 3: Autism and ToM

References (in chronological order)

David Premack and Guy Woodruff, 1978. Does the chimpanzee have a theory of mind? Behavioral and Brain Sciences 4:515-526. See also comments and reply pages 555-629 in the same journal issue.

Siegler, Robert S., 1978. The origins of scientific reasoning. In Siegler, Robert S. (ed.), Children's thinking: What develops? Lawrence Erlbaum Associates, Hillsdale, NJ, 109-149.

Heinz Wimmer and Josef Perner, 1983. Beliefs about beliefs: Representation and constraining function of wrong beliefs in young children's understanding of deception. Cognition 13: 103-128.

Josef Perner, Susan R. Leekam and Heinz Wimmer, 1987. Three-year-olds' difficulty with false belief: The case for a conceptual deficit. British Journal of Developmental Psychology 5: 125-137.

Charlie Lewis and Amanda Osborne, 1990. Three-year-olds' problems with false belief: conceptual deficit or linguistic artifact? Child Development 61:1514-1519.

Janet Wilde Astington, 1993. The child's discovery of the mind. Harvard University Press.