At a picnic yesterday, I saw a friend eating from a paper plate he held up close to his face, while his dog watched eagerly. The master said to the dog, "What are you looking at? You've never had anything to eat from a plate!" And although some pet owners are unreliable about this kind of rule, I'm pretty sure that this family of veterinarians had not slipped up and given the dog a tidbit from what they were eating.
So, what was the dog looking at, or for? A recent article by Michael Tomasello and Juliane Kaminski ("Like infant, like dog", Science, 4 September, pp. 1213-1214) offers some explanations about why dogs watch what we do-- and why human babies do too.
Tomasello showed that if a piece of food was hidden in one of several containers, and a person pointed at the right one, a dog could use that information to find the food. Sounds easy, doesn't it? But most animals can't do it, even those we usually think of as "intelligent".
The hidden-food test comes out of a test of early cognition that was developed by the Swiss developmentalist Jean Piaget, back in the 1920s. Piaget was interested in what's called object permanence-- the ability to understand that something still exists even when we can't see it. He noticed that babies of 6 or 7 months, or even more, seemed to forget about an interesting object as soon as it was hidden. (If you have a baby of that age around, you can try this with an object like your car keys. Hold the object up for the baby to see. Then put it down within the baby's reach. When the baby reaches for the keys, quickly place a scarf or a piece of paper over them. What will probably happen is that the baby will stop reaching, turn away, and look for something else to do-- as if the keys just vanished.) In a few more months, the babies could find the hidden object.
Piaget devised a more complicated hiding problem, and used a procedure often referred to as the "A-not-B problem". There are two containers, A and B, that the keys can be placed in. The baby can't see into the container, so he or she can't see directly where the keys are. An adult hides the keys in container A first-- and does not just hide them secretly, but calls the baby's name, gets the baby's attention, and makes sure the baby sees what's going on. The baby then reaches for A and finds the keys. This is repeated several times. But then the adult goes through the whole script again, and rather than putting the keys in A, makes a point of putting them in B. The baby reaches-- for A again!
I suppose an obvious conclusion is that although babies are smart for their age, they aren't as clever as adults. But a more important point may be that babies look to adults for instruction. What Mom or Dad thinks is right seems to take precedence over something the baby saw for himself. When an adult has appeared to be teaching the baby about a hiding place, and has done this several times, the baby takes this information seriously. The baby believes that adults can make mistakes, but what they tell you over and over again is probably correct. What adults tell you with the direction of their gaze, with their facial expressions, and with their voice tones are especially important to babies too young to understand much speech. We see this not only in the "A-not-B problem", but in social referencing, the infant behavior of looking at an adult's face when something unusual happens. The infant searches for a facial expression that tells what the adult feels about a new person or event. If the adult looks scared, the baby backs off from the unusual situation and is reluctant to explore it; if the adult looks relaxed and pleased, the baby is much more likely to approach the novel person or thing.
As Tomasello has shown, dogs are like babies in that they pay attention to things human adults communicate, whether by pointing or in other ways. My friend's dog figures his master knows some very important things, especially about food, and what he does with food could be significant information for a dog to have. So the dog keeps watching, even though he has never had food in those circumstances-- he's paying attention to his man's actions with food. That's worth learning about even though it hasn't brought any food--- yet.
To come back to human beings: over the last 100 years or so, there have been many recommendations for teaching children by allowing them to learn on their own. From Maria Montessori's day onward, learning by exploration and discovery has been put forward as the best method of schooling. Yet the attentiveness of babies to adult examples suggests that direct instruction may really chime best with natural human learning tendencies. But--- can we necessarily generalize from what babies do to what older children and adults do best? That might be a false analogy. The jury is definitely still out on the meaning of "A-not-B".
