Maurer and Maurer (1989) however, claimed that ''attachments are welded in the heat of interaction'', clearly suggesting that the formation of an attachment does not simply require two individuals to be close together, nor does it form due to the provision of food as proposed by Dollard and Miller. The creation of an attachment is rather reliant upon the interaction between the two. A study which strenghthens this view, is the research conducted by Harlow and Harlow (1962). This is the study of the formation of love in infant monkeys, which demonstrates the importance of care-giving and companionship (in social and cognitive development). It does not support the learning theory of attachment as the learning theory predicts that an infant's attachment will be to the person who gives greatest pleasure or drive reduction; however, in the study, this statement is proven to be wrong. In the study, we see the infant monkey cling onto the cylinder covered with a towel, rather than the cylinder providing milk. If food was the cause of the attachment, we would expect the monkey to cling onto the lactating cylinder. Also, the majority's of the monkey's time was spent on the towel covered cylinder, providing contact comfort. They would also jump to this cylinder when frightened which is a characteristic of attachment (seeking proximity for security). This experiment provides evidence for the fact that the formation of an attachment is not simply based on the supply of food, and therefore disapproves with the learning theory. Similarly, Shaffer and Emerson(1964), demonstated that the child's attachment is not always to the person who feed s it. This study involved 60 babies, who were observed over a period of time. The study showed that by 18 months, infants had five or more attachments, such as with siblings, and therefore goes against the learning theory, as it suggests that not all infatns attach to the person who provided them with food first, but rather to the person who contributed to their satisfaction, as a result of contact comfort and security (interaction).
Moreover, another explanation of attachment is seen with the evolutionary perspective of Bowlby's theory (1969). This theory consists of three different aspects including: a critical period, innate programming and continuity hypothesis. The theory is centralised around nature and argues that attachment is an inherited behaviour. According to Bowlby, children have a biological need to attach to their main caregiver, serving for the purpose of survival. This is achieved with the help of social releasers (crying). For example, because babies are physically helpless, they need to be looked after, without this support, they would be susceptible to death. This led Bowlby to conclude that humans have evolved in such way that they are born with this inherited tendency to form an attachment, which in trun increases their chances of survival, giving them an adaptive advantage. This implies that infants too, are inately programmed to form an attachment.
Furthermore, Bowlby also highlighted the fact that there is a critical period which an attachment is most likely to develop (first 2.5 years of their life). If the attachment is weak, or even worse- lost, it can result in several negative consequences, and although they are relatively short term, others can have an enduring long-term effect on the behavioural development of an individual.
Under the continuity hypothesis, Bowlby highlighted that the relationship with one special figure (monotropy) provides an infant with the internal working model of relationships. This internal working model can be posititve or negative and determine the perceptions, emotions and expectations in later relationships. Secure children are said to develop strong self-esteem as they grow older and tend to be more independent. They are also likely to be more successful both socially and accademically. Contrastingly, children with an avoident attachment have a tendency of avoiding parents/caregivers. They also have a working model of themeselves as unworthy. As a result, these types of children tend to exaggerate their emotional responses as a way of obtaining the attention they desire.
Bowlby's theory has been further studied with the work of Hazan and Shaver, Scroufe et al, and many others to determine whether this theory could be applied in reality. The study by Mccarthy (1999) involved women who were previously identified as ''avoidant-insecure'' as infants. The study discovered that being insecurely attached led to difficulty in forming relationships, which therefore supports Bowlby's theory of continuity hypothesis. Similarly Hazen and Shaver (87) found that secure attachment increased the chances of the ability to form good relationships, whereas insecure decreased it.
Rutter et al's study on the other hand, showed that the adoptees were perfectly able to form relationships after the first year of life, and made developmental progress following adoption. However, there was a strong correlation in that the later the adoption, the slower the progress was. Nevertheless, these children were still able to form an attachment out of the so-called 'critical period'.
It is worth considering that although several negative criticisms have been made on Bowlby's evolutionery theory, Bowlby's theory is still accepted as the dominant explanation of attachment. The learning theory is also accepted, however, it has been highly criticised on the grounds that it is quite reductionist.