When it comes to love, it seems most everything we know, we did learn before kindergarten. We learn how to relate to people -- and, especially, how to love -- from our mothers, or from whoever takes care of us the most.

 
But not everyone is lucky enough to get the right kind of mothering, the kind that reassures the undeveloped baby mind that love and protection will always be there.

 
Some mothers don't feel confident they can take care of a baby; some didn't learn how to love from their own mothers. Sometimes, a woman feels angry and trapped by a pregnancy she never wanted. Although oxytocin associated with labor and birth provides an opportunity for forming a deep bond between mother and baby, circumstances or the mother's own needs or fears may prevent a secure bond from forming.

 
In the 1940s, psychologist John Bowlby realized that not every child formed a deep and secure attachment with his mother.

 
Bowlby's idea was that we learn "attachment styles," the way we behave in close relationships, from our mothers -- or whoever is the primary caregiver during our infancy. He and collaborator Mary Ainsworth identified three attachment styles: secure, anxious and avoidant. Anxious kids are never sure what response they'll get from their mothers, so they expend a lot of energy trying to get affection and attention. Avoidant kids have decided they're not going to get much love, so they do their best to do without.

 
These coping strategies are encoded at the neurochemical level: The stress response and the oxytocin response seem to be set by the relative amounts of anxiety and love a baby experiences.

 
According to Bowlby's and Ainsworth's theory of attachment, we tend to use the same style of relating to others that we learned with our mothers. But the brain is plastic; it continues to grow in response to experience. How much do we automatically apply the way we relate to Mommy and Daddy to the friends and lovers we meet?

 
Claudia Chloe Brumbaugh is a psychologist studying attachment, working in the lab of R. Chris Fraley at the University of Illinois. Fraley was a student of Phil Shaver's, one of the modern attachment theorists who took John Bowlby's ideas and extended them to adult relationships.

 
Fraley and others have expanded the list of attachment styles to provide more nuance.

Fraley has been tracking attachment styles for several years, via studies and Internet surveys. He and Brumbaugh hope to understand just how set these styles are, whether they change as we look outward from our mother's breast.

 
In a recent study, Brumbaugh and Fraley asked people to describe one of their parents and their current love partner. They also determined how secure each person felt in these relationships, as well as their general attachment styles. Next, they created fictional profiles of people, some of them designed to be similar to each person's parent and lover, putting them into the form of a personals ad.

 
In a separate session two weeks later, participants were asked to read the ads of people who were similar to their parent and to their lover, as well as ads that were like neither; these were the controls.

 
In a similar early study that only looked at love relationships, Brumbaugh and Fraley found that when people read the ads of individuals who were like their romantic partners, they tended to apply the same emotional strategies to the new person.

 
It looks like Mom and Dad -- and especially Mom -- provide the basis a global working model that we start with when trying to connect with someone new. But as we meet and get to know more and more people, we edit the model until it's more of an average of all our experiences of others. We may also develop some more specific models: parental figure, friend, lover. Ideally, we continue to edit all these models based on our ongoing experiences in the world.

 
Brumbaugh and Fraley have a paper in press that describes their latest study. I spoke with her in April 2007.

 
HUG: Tell me about the concept of multiple working models for relationships. Some are general and others represent a specific type?

 
Brumbaugh: Most attachment researchers measure a more global attachment style: How we approach close relationships in general. But people can have different attachment orientations to different people in their lives. You don't necessarily have the same feelings toward your mother as toward a boyfriend. I'm interested in the specific models of relationships.

 
HUG: In your paper, you wrote that "attachment orientations toward parental and romantic relationships have been shown to be only modestly correlated to one another." Does that mean that when we start to move out of the family circle, we start from scratch to build new models?

 
Brumbaugh: They're called working models for a good reason. The working aspect means they're open to change and revision. It's hard to conclusively say what's going on, but it seems we do use our parental models as the basis for new relationships. However, they are up for revision with life experiences and new relationships. In childhood, your attachment orientation and relationship-specific working models are the same thing; for most people, it's their parents. Over time, we build relationship-specific models for other people in our lives. The modeling becomes more complex and multidimensional.

 
HUG: Can we completely change our attachment style?

 
Brumbaugh: I'm not aware of any research, but I've heard people tend to move in a more secure direction over time; they're more likely to become secure than insecure over time.

 
HUG: I know it’s an unconscious process, but how do we “decide” which model to use?

 
Brumbaugh: It can be a variety of things, it seems. In my research, I ask people to provide an open-ended description of their attachment figures. They write down 14 sentences about someone, whether parents or a romantic partner. They can describe anything, from day-to-day behavior, such as enjoying cooking, to fundamental characteristics, like being supportive. I use these descriptions to create profiles. It seems that most any aspect of a person can trigger the use of these working models.

 
HUG: One of your hypotheses was that when similarity exists between a new person we meet and someone who's the basis for an existing relationship model, that model is activated, so that we use the same style in the new relationship. Did that happen in your study?

 
Brumbaugh: Even when there was no similarity between the new relational target and an existing model, people applied their partner models to both the partner target (that is, the person who resembled the partner) and the parental target. They also did use their parental model, but not in a selective way. There was more selectivity in the way people applied their partner models.

 
HUG: What about in the control condition, the person who didn't resemble either the parent or the partner?

 
Brumbaugh: People applied the parental model significantly more. It may have something to do with how parental models are foundations of working models; they might have a minimal but persistent way of being applied. When no resemblance exists between the past and present situation, the parental model may be more of a default.

 
HUG: You found that the subjects who were avoidant toward their parents didn't apply that strategy to the targets, not even those who resembled their parents.

 
Brumbaugh: Correct. It seems like people start relying more and more on their partners. But you have to take into consideration the demographic I'm working with. These people are falling in love; they may be having their first romantic experience at age 18 or 19. People who were not in romantic relationships were excluded from my study. I have the feeling it's a crucial time period for them. They've just left home for college, and maybe there's some distancing from their parents going on. At the same time, these romantic relationships are becoming really important.

 
HUG: But anxiety toward the parent seemed to pervade new relationships.

 
Brumbaugh: The thing I found most interesting was that people were more positive to new people who resembled their partners. At the same time, they were more anxious toward the partner target. They were seemingly more attracted to and interested in the person who resembled the partner but also afraid of them at the same time. Obviously, attraction makes sense. Possibly, the anxiety is because partners can break up with you and parents can't. The more important the relationship was, the more they transferred anxiety to the partner target. They seemed to feel, "You're the one, please don't leave me."