Dollard & Miller (1950)
According to Dollard & Miller (1950) attachment is a learned behaviour that is acquired through both classical and operant conditioning
Before attachment is learned, the infant gains pleasure through being fed.
When the infant is being fed, the infant associates the person providing the food with the food.
When the attachment has been learned, the infant gains pleasure when the primary caregiver is present.
When an infant is hungry it is in an uncomfortable state. Relieving the uncomfortable state will make the infant more comfortable, and so anything it does to make itself more comfortable will be learned through negative reinforcement
- A hungry baby will cry because it is distressed.
- Feeding the baby makes it more comfortable, and so crying is learned through negative reinforcement.
- Over time the pleasure of being made comfortable by being fed becomes associated with the primary caregiver.
- The baby has now learned to cry to get the primary caregiver’s attention, and it feels pleasure when the primary caregiver is present. Attachment has now been learned.
Dollard & Miller (1950) used the term secondary drive hypothesis to describe the processes of learning an attachment through operant and classical conditioning. Secondary drive hypothesis explains how primary drives which are essential for survival, such as eating when hungry, become associated with secondary drives such as emotional closeness. They extended the theory to explain that attachment is a two way process that the caregiver must also learn, and this occurs through negative reinforcement when the caregiver feels pleasure because the infant is no longer distressed.
Evaluating learning theory of attachment
Learning theory provides a very plausible and scientifically reliable explanation for attachment formation. It seems highly likely that simple association between the provision of needs essential for survival and the person providing those needs can lead to strong attachments. However the theory is extremely reductionist and there is evidence that infants can form attachments with a person who is not the primary care-giver.
Schaffer & Emerson (1964) studied the attachments formed by 60 infants from birth. They found that a significant number of infants formed attachments with a person other than the one doing the feeding, nappy changing, etc. and that the primary attachment was often with the father and not the mother. They found that it was the quality of interaction with the infant that was most important - stronger attachments were formed with the person who was most sensitive and responsive to the infant's needs.
Harlow (1958) experimented with the attachments formed between rhesus monkeys and surrogate mothers. In this case the surrogate mothers were wire framed models that provided food and therefore satisfied the monkeys' primary needs, or ones that were comfortable and padded but provided no food. The findings were that the monkeys would cuddle up to and be more distressed at losing the comfortable padded surrogate mother that provided no food than they were the uncomfortable wire-framed surrogate mother that fed them. It would be easy to conclude from this that conditioning does not explain attachment in infant monkeys as they are not linking food with pleasure, but clinging could actually be essential to the survival of infant monkeys whose mothers may be swinging through trees and so conditioning could still adequately explain their attachment formations - clinging is not so essential for survival to humans.
Learning theory has its strengths and weaknesses, but on its own it does not explain attachment adequately.