It has been argued that an infant under three who is cared for outside the home may suffer because of the separation from his parents. The British psychoanalyst John Bowlby maintains that separation from the parents during the sensitive 'attachment' period from birth to three may scar a child's personality and predispose to emotional problems in later life. Some people have drawn the conclusion from Bowlby's work that children should not be subjected to day care before the age of three because of the parental separation it entails, and many people do believe this. But there are also arguments against such a strong conclusion. Firstly anthropologists point out that the secluded love affair between children and parents found in modern societies does not usually exist in traditional societies. For example, we saw earlier that among the Ngoni the father and mother of a child did not rear their infant alone--far from it. But traditional societies are so different from modern societies that comparisons based on just one factor are hard to interpret. Secondly, common sense tells us that day care would not be so widespread today if parents, caretakers or pediatricians found that children had problems with it. But Bowlby's analysis raises the possibility that early day care has delayed effects. The possibility that such care might lead to, say, more mental illness or crime 15 or 20 years later can only be explored by the use of statistics. Statistical studies of this kind have not yet been carried out, and even if they were, the results would be certain to be complicated and controversial. Thirdly, in the last decade, there have been a number of careful American studies of children in day care, and they have uniformly reported that day care had a neutral of slightly positive effect on children's development. But tests that have had to be used to measure this development are not widely enough accepted to settle the issue. According to Bowlby, children under the age of three ______.