In the Behaviourist approach the main theory on attachment formation is the Learning theory, where there are two types of “learning” or conditioners, classical — the child associates an action or sound to being fed or to being attended to, e.g. the child associates the sound of footsteps coming towards it with being fed and so begins to salivate and feel hungry, and operant — the child associates being certain behaviours with being either rewarded or being punished, e.g. the child has a tantrum and is “punished” this results in the child being less likely to have another or the begins walking and is rewarded this results in the child being more likely to repeat the behaviour. Classical and operant conditioning are non-exclusive of each other. Overall the theory is too reductionist to be of much use in explaining attachments but it is very useful in modifying or adapting the behaviour of animals e.g. Pavlov’s dogs.
On the other hand the Evolutionary approach states that infants are born with a drive to become attached and so have many innate abilities called social releasers i.e. smiling, giggling, and crying which elicit “caregiving” tendencies from those nearby. Though these alone do not make the bond, the caregiver must interact with the infant for the bond to be formed without this interaction the child would become privated (no bond forms in the critical period). The Evolutionary approach also says that infants form one specific bond (monotropy) due to this interaction, this means the child becomes attached the person who interacts the best with them, the person then becomes the child’s Primary caregiver and the relationship between the PCG and the infant is the basis for all the later relationships the child may have.
The extent to which the Learning theory and Freud’s Psychoanalytical theory have been successful in explaining attachments is quite limited because they are too reductionist and try to explain a difficult process using very simplistic measures and whilst it may work for “lower” species, they very difficult to apply with great accuracy to humans because neither account for or explain why the child has innate social releasers or how this prepares the child for later life. E.g. a child is conditioned to think that footsteps coming toward it mean food is on the way or that not throwing a tantrum is a good thing to do (Learning theory), 20-30 years later the “child” hears footsteps approaching it and begins to salivate or the “child” is annoyed but thinks not to throw a tantrum, but neither of these two are transferable to any other aspect of the child’s development. An example for the psychoanalytical approach would be the child’s id needs to be sated and so someone does so this person then becomes the love object and an attachment is formed, 20-30 years later the id again needs to be sated and so the “child” now capable of doing it themselves does so do they become the love object?
The extent to which the Evolutionary approach has been successful on the other hand is quite substantial, because unlike the other two it puts forward a more believable and proven view, that infants form attachments through interaction and through reciprocal bonds with its caregivers and primary caregiver. An example of the Evolutionary approach would be the infant begins to cry after falling or hurting itself, the “mother” then responds to this “social releaser” and begins to soothe the baby and interacts with it, the baby then begins to feel happier, the baby then begins to interact more with the “mother” and a stable bond is formed, 20-30 years later the “child” hurts itself again, it turns to the “mother” and “she” comforts the “child”, the bond is still present and the “mother” can still calm the “child”.
Overall all the theories have valid points because in some societies the bond may only form because the child is told not to do certain things and that others are allowed to be or must be done or because the child’s id is sated. It is possibly argued that both the Learning theory and the Freudian approach have levels of interaction in them because the child is told to do or not do something or the child’s id is met and neither process can happen with out some interaction between the CG and the infant, and no one denies that they may have some interaction, they just believe it is because the child is fed or the child is “taught”. And it is from this argument that Bowlby’s theory arose and has now taken “centre stage”.
In conclusion it is probably best to say that all the afore mentioned theories may have some basis of truth but it is with Bowlby’s that most psychologists would start a new theory or build upon.