The Attachment Spandrel: A Speculation

When people talk about evolution of structures and functions or of behaviors through natural selection, they sometimes use an intriguing metaphor. They suggest that human evolution can have a an aspect that can be compared to a spandrel—an architectural feature like an arch that was originally developed as a way to support a building, but that later was liked for its attractive appearance and put to work as a way to decorate a structure. It seems possible that human evolutionary change involves spandrels too, with traits that began with one function later becoming useful in other ways. For instance, eyebrows might originally have been useful for keeping sweat out of our eyes, but now are used in facial expressions and gestures that convey our moods and intentions to people we encounter.

Can young children’s emotional attachment to their parents be considered as a trait that can be “spandrelized”? (Maybe there’s a proper word for this but I don’t know it.)   This would certainly not be the way that many people, including quite a few unconventional psychologists and counselors, think about it.  It’s commonly, but incorrectly, assumed that once an attachment relationship develops in early life, that attachment stays the same unless something happens to stop it. For example, the death of a parent and subsequent mourning would deactivate that attachment; some “parental alienation” advocates hold that one parent can destroy or at least suppress a child’s attachment to the other parent by emotional manipulation. People who believe in these ideas apparently miss the fact that both attachment behavior and cognitive abilities that are part of thinking about attachment change in the normal course of development, and this contributes to changes in the nature of relationships that originated in toddler’s emotional connections to familiar adults.

So where does the spandrel part come into this? What I am going to say is speculation, just like most of the Just-So stories that are put forth as ideas about the evolution of behavior. But it is possible to make the case that attachment began as a result of natural selection for certain advantageous behaviors, and then later was put to work for other purposes.

Many birds and mammals show some form of “attachment” to parents. Ducklings and goslings, for example, follow moving objects that they see soon after hatching, and those objects are most likely to be their mothers. Whatever kind of object they initially follow, they will during early life follow that same object, and in adulthood will court and attempt to mate with a similar object. This initially serves a protective function and encourages survival, because although a mother duck might not be able to fight off a fox, her behavior against a predator does increase the young duck’s chances of escape. The same consideration applies for young humans, whose curiosity and exploratory behavior could lead them straight to an enemy if they were not “attached” so that they stayed close to familiar adults and avoided strange creatures. Whereas ducks and some other animals form their “attachments” shortly after birth or hatching, humans, who are very immature at birth and cannot move independently for many months, do not show attachment behavior until roughly the time when they might be able to wander away.

Looking at it this way, we can see the enormous survival value for the young of preferring to stay near familiar adults and being wary of strangers and unfamiliar creatures. But within a few years after birth, human children become much more able to judge danger and to keep themselves safer (perhaps not in today’s high-speed traffic, but in the environment of early adaptation). They don’t need attachment for survival in the same way, yet they retain strong relationships with familiar adults under many circumstances, and do this right up until their own adulthood, often modulating into a “reversed” relationship when an aging parent needs the adult child’s care.

Feelings and thoughts about attachment thus lose much of their survival value after the child gets to “school age”, but those feelings and thoughts can be seen as serving new, useful purposes for both the individual and the group--  in other words, attachment has become a spandrel. For the individual, behaviors learned from experiences in the attachment relationship can provide a foundation for understanding other social relationships—that is, they form the internal working model (IWM) of relationships that John Bowlby proposed. The IWM allows the older child and adult to participate in any social relationship with some confidence in assumptions about how social interactions work. The individual does not have to develop new rules from scratch about interacting with any new person he or she meets, and this saves time and energy.

The community also benefits from the attachment spandrel for several reasons. One is that the social rules learned in the early attachment relationship help young adults care for their own children effectively and enlist other community members to help in this task when needed.  In addition, the community survives and thrives better when social interactions among members are orderly and constructive, when conflicts and aggressive impulses are modulated by attachment-influenced social rules. For just one more, community values and beliefs are more easily shared and passed on in the context of existing close relationships where they are modeled and implied, than if they had to be taught through direct instruction.

Thinking about attachment after the toddler period as a spandrel can be a helpful way to understand how some aspects of attachment work. But the really basic point is that attachment changes through the course of development, and that toddler attachment, fascinating though it is, cannot be considered to define the nature of attachment later in life.