A toddler hugging R2-D2--- what a cute image! And if a small child met a talking robot, a hug would be very likely, because toddlers do like to put their arms around big things and lift them if they can. They hug and try to pick each other up, and their parents often keep the waste baskets on top of tables so they won't be carried throughout the house.
But the article by Professor Noel Sharkey (Science, 19 December 2008, Vol 322, pp. 1800-1801), which I mentioned last week, in which he talked about the possibility of "nanny-bots" or child care robots, contained a reference to the possibility that young children could "bond" or "become attached" to robots. This statement referred to a research report saying that toddlers in a day care center had not only played with and tried to "help" but hugged a robot they became familiar with.
What does Dr. Sharkey's reference mean? Are "bonding" or "attachment" the same as hugging? If they are not, how would we know if such relationships between children and toddlers had developed?
"Bonding" is a word that used to have a highly technical meaning, but has lost or broadened it. In the 1970s, some researchers interested in mother-infant relationships began to use "bonding" to mean the very intense interest and concern many parents feel for their babies within the first weeks and months of life. However, other people soon began to use the word in different ways, especially to refer to the feelings the children had toward their parents. Nowadays we hear "bonding" used to describe any pleasant sort of social interaction. For instance, my husband has just told me he is going to do some "male bonding" on a trip with some boating friends, and I don't suppose he will wake up at night to check whether they are breathing, as new parents might! In this modern sense, we might well say children had "bonded" to robots if they just had a good time playing with them, but of course the original technical meaning would not apply, as that usage meant how parents loved their babies and wanted to care for them.
The term "attachment" refers to the more important issue here. "Attachment" means somewhat different things as we look at ways people change over the lifespan, but when we are talking about toddlers, it means the positive feelings children have toward familiar caregivers, the children's distress about separation from familiar people, and the efforts they make to stay near adults they know and away from strangers. If children are playing calmly in a familiar place, they may not show much interest in their "attachment figures", but a perceived threat like the arrival of a stranger will quickly bring the children near the familiar person, showing us evidence of their attachment. We can also see evidence of attachment if a child uses an adult as a secure base. This means that a young child who is in an unfamiliar place or with unfamiliar people can gradually explore the new place or people, as long as he or she has a familiar person to return to, or even just to look at, periodically. Without a "secure base" (a familiar person) present, the toddler is likely to be frightened, withdrawn, and unwilling to try out unfamiliar toys no matter how attractive they are.
Could a toddler be "attached" to a robot? It seems possible, as all toddlers show some degree of attachment to their homes, and behave differently there than they do in strange places. Many toddlers are also quite concerned with a special blanket or teddy that they carry with them; if it goes into the laundry or disappears, the child is very distressed. So the fact that the robot is not really alive would not interfere with the development of a special emotional attitude toward it.
However, a hug or play with the robot does not seem to be enough to support the idea that a toddler has formed an attachment to the robot of the same kind he or she has to familiar adult caregivers. To test whether such an attachment did exist, we would need to test the child in a situation where no familiar adults (parents or teachers, for instance) were present. Then we could see whether the child behaved toward the robot in the same ways he or she would usually act toward adults. If a stranger approaches, does the child run to the robot? If the robot leaves and then comes back, does the child rush to it for a reunion? And can the child explore a strange place with the robot acting as a secure base? These questions can only be answered through systematic observation--- which has not happened yet.
