Attachments to other people – you can’t live with them, you can’t live without them. Kevin Bacon in his classic role as Ren McCormack in “Footloose” is a bad boy teenager who reminds the straight-laced minister (played by John Lithgow) and his quiet town that dancing is also important in leading a fulfilling life. A social species such as humans needs to be social, and to be social we need to have attachments.
The notion of attachment was introduced by John Bowlby, a British developmental psychologist and psychiatrist who spent much of his career trying to define attachment and its various outcomes. Influenced by ethologists such as Konrad Lorenz and Nikolaas Tinbergen, other psychoanalysts, like the (in)famous Sigmund Freud, and evolutionary theorists, such as Charles Darwin, Bowlby believed that attachments were critical to survival. He postulated that very young children have an innate need to develop a close emotional bond with a caregiver. Originally believed to develop only between mothers and their offspring, attachments are emotional and physical bonds that promote the survival of the offspring. If this biological need is not met, then abnormal behavioral development (i.e., difficulty forming relationships with others, immature social skills) is expected to occur.
Attachments involve Dancing
Attachments are so fundamental to survival that they can even be formed even in the worst of situations, such as those involving abuse and neglect as Mary Ainsworth and Mary Main observed. Human attachments take a few years to fully form, unlike imprinting, which can take up to a week for some species but is usually completed within 24-72 hours.
The significantly longer time frame needed for an attachment to form is related to offspring development and the quality of interactions between the infant and the mother (and other caregivers), which is built upon coordinated and fluid dances between the caregiver and infant and should be performed with sensitivity.
Attachments are categorized into secure and insecure attachments, which have several variations. Attachments are assessed using the “Strange Situation,” a series of separations from and reunions with the primary caregiver and the 12-24 month old in the presence/absence of a stranger. Attachments are fully formed between two and three years, corresponding to the development of “internal working models.” These mental representations are believed to set the stage for future relationships.
Predicting the Future?
Research has yielded mixed outcomes about the validity of early attachments predicting future attachments or relationship quality as adults. Generally, children with insecure attachments form insecure-type relationships as adults while securely attached children develop healthier relationships. One reason the predictive validity of early attachments has been questioned is that little follow-up research has been conducted on the continuing development of attachments beyond school-age years.
Moreover, children form multiple attachments across their lives, which when combined with life experiences, can impact their relationships in the future and make it difficult to predict future attachment quality. What is known about adolescents is that those with secure attachments tend to have better outcomes, such as having more positive relationships with others, including their parents, less conflict, and better social skills.
Unveiling Adolescent Secrets (or Maybe Just Their Brain Activity)
A recent article in the Psychonomic Society’s journal Cognitive, Affective, & Behavioral Neuroscience examined the neural underpinnings of the consequences of attachments in adolescents. The authors, Martin Debbané and colleagues of the University of Geneva, had previously examined the relationship between two insecure attachment styles (avoidant and anxious) and how adolescents processed inaccurate social feedback using fMRI.
The results of that initial study indicated two different patterns of activation dependent upon what type of attachment was reported: (1) adolescents with avoidant attachments tended to have lower levels of activity across the emotional centers of the brain and the reward pathway whereas (2) adolescents with anxious attachments had increased activity levels in many of the same areas.
Attachments Influence Processing of Information in Adolescents
Armed with this knowledge, Debbané and colleagues recorded the brain activity of 44 healthy adolescents (males and females, 12-18 years) as they assessed characteristics about themselves or another individual, a close same-sex friend. Participants responded to adjectives that could be descriptive of themselves or their friend (e.g., “mean”, “cold”, “friendly”, “curious”) as part of the trait-adjective evaluation task (TAET), a well-validated test with fMRI-based studies. Participants judged how well each adjective described themselves or their friend, using a 1 (not at all) to 4 (completely) Likert scale. Finally, the participants also completed the Relationship Questionnaire (RQ), which the researchers used to categorize attachment styles of the participants: anxious or avoidant.
Per the researchers, these “derived self- and other-models directly represent a person’s view of the self and others in terms of expectations about the worthiness of the self and the availability of others.” Thus, the self-model was represented by a calculation on the RQ in which negative views of oneself (i.e., fearful and preoccupied) are subtracted from positive views of oneself (i.e., secure and dismissing), with lower scores indicating that a participant views him/herself as more negative and is undeserving of help or love from close others (i.e., anxious style). For the other-model, a similar calculation is conducted, but negative views of others (i.e., fearful and dismissing) are subtracted from positive views of others (i.e., secure and preoccupied). If a lower score is produced, then the participant has a representation that significant others are unhelpful and unreliable (i.e., avoidant style).
Participants spent about 16 min completing the task, receiving 30 blocks of words, 10 per condition (self, close other, and a control condition in which participants counted the number of syllables in the word presented). The figure below summarizes the research design.
The authors expected several key results based on their previous work:
(1) adolescents with lower scores on the self-model would show increased activity in social-emotional brain areas during exposure to negative evaluations of the self, because such negative self-evaluations sustain negative self-representations.
(2) adolescents scoring low on the other-model would show decreased neural social-emotional processing of positive other-representations.
Not so Happy With Themselves But They Really Like Their Same-Sex, Close Friends
Participants attributed negative adjectives to themselves (light grey bar) more frequently and positive adjectives to the close other (close same-sex friend, dark grey bar), as shown in panel A of the figure below, and represented by a significant interaction. Participants also responded more slowly to negative than positive items (panel C), an indication that suggested they needed more time to process negative information about themselves and others. Finally, more positive responses in the other-model were associated with fewer negative adjectives attributed to the close other (panel B). The authors suggested that this negative relationship may be related to either self-serving attributions or a self-positivity bias, which are the tendencies to see the self as more positive and better than others.
Rating Themselves Positively or Negatively is Emotionally Linked to People’s Attachment Style
The results of the fMRI analyses indicated an association between an attachment-derived internal working model and brain activity. Specifically, significant associations were found between the self-model and brain activity in the left amygdala/parahippocampus, bilateral ATP/aSTG, (pre)cuneus, left DLPFC, and cerebellum when comparing self versus close other adjective evaluations. This is shown in the figure below.
Put more simply, participants with negative attachment-derived views of themselves (i.e., low positivity of self-concept) showed increased brain activity (yellow regions in the figure) in the left amygdala/parahippocampus, bilateral ATP/aSTG, (pre)cuneus, and left DLPFC (See panels A-C in the above figure), while attributing positive and negative adjectives to themselves. Anxiously attached individuals are more likely to attend and process negatively-valenced social information in the amygdala and surrounding areas, which the authors suggested was evidence that this information was generally relevant to the participant and processed more intensely.
For participants with anxious attachments, heightened activity was also observed in three areas with separate functions: anterior temporal pole (ATP), anterior superior temporal gyrus (aSTG), and the parahippocampus (Panel D above). Despite being associated with different functions, all three areas are involved in processing emotional information within a social (attachment-related) context and were associated with increased activity when participants ascribed negatively-valenced characteristics (i.e., prejudiced, unintelligent). Interestingly, similar activity was also observed for positive adjectives, leading Debbané and colleagues to suggest that the participants’ anxious tendencies biased their interpretation of the positive attribute by placing it into an attachment-based social context.
In the words of the authors: when evaluating positive and negative adjectives regarding either the self or a close other, the attachment-derived self-model is associated with (i) increased neural processing of positive and negative self-attributes (i.e., I think a lot about who I am – good and bad.), but (ii) decreased neural processing of negative close other-attributes (i.e., My best friend is awesome at everything!).
It Always Comes Back to Freud . . . or Kevin Bacon in a few Degrees of Separation
So what does all this say about one’s early childhood attachment? Not a lot, directly. However, if we believe that childhood has a significant effect on adults or developing adolescents, then it is safe to say that the attachments formed as children will have a lasting effect on how they view themselves as adolescents and consequently their brain activity during a task in which they endorse specific characteristics about themselves and their same-sex, close friends.
Ultimately, children who formed anxious internal working models (as was studied in the present article) were more likely to process information in the emotional areas of the brain regardless of the valence of the adjective used to describe themselves, and they were also extremely positive about their close friend. Avoidant adolescents, by contrast, did not show the expected brain activation pattern, likely because a close friend was rated extremely positively.
Let’s wrap up with the inevitability of Freud. Remember the fun game “Six degrees of Kevin Bacon” (“Kevin Bacon – who??” as my adolescent children asked)? Based on the idea that any two people on Earth are six or fewer acquaintance links apart, the Kevin Bacon variation challenges us to get to Kevin Bacon in as few steps as possible.
Let’s test your Hollywood Know-how: How many steps does it take to link Selena Gomez to Kevin Bacon? Go to the Oracle of Bacon for the answer. You might be surprised. I was!
To unravel the mysteries of the mind, especially the adolescent mind, additional research is needed to understand the influences of early experiences, such as biologically important ones like attachments, with more adolescents and for longer periods of time. Freud would be proud that his legacy has held, while Kevin Bacon would remind everyone to remain . . . “Footloose”!
Article focused on in this post:
Debbané, M., Badoud, D., Sander, D., Eliez, S., Luyten, P., & Vrtička, P. (2017). Brain activity underlying negative self-and other-perception in adolescents: The role of attachment-derived self-representations. Cognitive, Affective, & Behavioral Neuroscience, 17(3), 554-576. doi:10.3758/s13415-017-0497-9.