The Impact of Early Interpersonal Experience on Adult Romantic Relationship Functioning: Recent Findings from the Minnesota Longitudinal Study of Risk and Adaptation

Abstract

Adopting an organizational view on social development, we have investigated how interpersonal experiences early in life prospectively predict how well individuals resolve relationship conflicts, recover from conflicts, and have stable, satisfying relationships with their romantic partners in early adulthood. We have also identified specific intervening interpersonal experiences during middle childhood and adolescence that mediate the connection between how individuals regulated their emotions with their parents very early in life and as young adults in their romantic relationships. We discuss the many advantages of adopting an organizational view on social development.

INTRODUCTION

One of the core ideas underlying many theories of personality is that early social experiences leave an indelible imprint on people, including how they behave in later relationships. This premise is a cornerstone of several major theories, including those of Freud (1940), Erikson (1963), and Bowlby (1969, 1973). For the past several years, we have investigated how certain interpersonal experiences early in life are systematically linked to how people think, feel, and behave in their adult romantic relationships two decades later. Our goal has been to identify and understand how early attachment-relevant experiences—some of which occurred before individuals could form memories of these events—affect the functioning of adult romantic relationships and the “interpersonal pathways” through which transmission patterns might have occurred. We have also examined the significant role that romantic partners assume in buffering certain people from the vulnerabilities posed by their early negative relationship histories.

Our studies have generated three key insights: (1) Certain early experiences exert small but apparently lasting effects on how most people think, feel, and behave in their adult romantic relationships years later; (2) There appear to be specific interpersonal pathways through which early-life experiences impact adult romantic relationship functioning; and (3) Some romantic partners are able to buffer individuals with early-life vulnerabilities, such as insecure attachment histories, from experiencing negative relationship outcomes in adulthood.

AN ORGANIZATIONAL PERSPECTIVE ON DEVELOPMENT

Our research has been guided by an organizational perspective on social development that has four basic principles (Sroufe, Egeland, Carlson, & Collins, 2005). First, the meaning of behavior depends on how it “fits with” other behaviors in a specific social context. For example, disengaging from conflict with a romantic partner when it is appropriate to do so ought to protect individuals from the deleterious effects of further conflict (Gottman, 1994), whereas failure to do so—especially when continued conflict is futile—should harm future relationship functioning (Gottman & Levenson, 1999).

Second, the way that individuals regulate their emotions in adult relationships should be associated with how they regulated their emotions with their caregivers earlier in life (Sroufe & Fleeson, 1986; Thompson, 2008). Synchronous and supportive relationships with early caregivers are the first context in which good, functional emotion-regulation skills are learned (Sroufe et al., 2005). Indeed, attachment security, which is a strong indicator of good interaction synchrony and effective emotion regulation in early childhood (Schore, 2005), predicts better emotion-regulation skills later in life (Thompson, 2008).

Third, mental representations (working models) of the self and relationship partners formed early in life guide interaction patterns in their later relationships, including romantic ones (Roisman, Collins, Sroufe, & Egeland, 2005; Sroufe & Fleeson, 1986). According to Bowlby (1973), the quality of caregiving provided by early caregivers is a template for what later relationships will be like, which in turn affects how people think, feel, and behave in their later relationships.

Fourth, experiences in early relationships with parents and in later relationships with adult romantic partners should jointly influence what happens later in development (Carlson, Sroufe, & Egeland, 2004; Sroufe, Egeland, & Kreutzer, 1990). For example, positive relationship experiences encountered later in life (such as becoming involved with a committed, well-adjusted partner) can counteract negative relationship experiences earlier in life (such as experiencing less responsive or less consistent caregiving during childhood; see Sroufe et al., 2005). Moreover, romantic partners may buffer “developmentally vulnerable” individuals (such as those who have insecure attachment histories) from experiencing bad outcomes in their romantic relationships (Rönkä, Oravala, & Pulkkinen, 2002; Tran & Simpson, 2009).

THE MINNESOTA LONGITUDINAL STUDY OF RISK AND ADAPTATION

In our research, we have tested how specific early interpersonal experiences are systematically linked to adult romantic relationship functioning by following a longitudinal sample of participants from birth into adulthood. The sample comes from the Minnesota Longitudinal Study of Risk and Adaptation (MLSRA; Sroufe et al., 2005). In 1976–77, new mothers receiving free prenatal services in Minneapolis public health clinics were recruited for the study. Their first-born child (whom we call “targets”) became the primary focus of the study. From birth, targets (N = 174) have been assessed at regular intervals at each stage of development using multi-method measures (e.g., interviews, questionnaires, teacher and parents ratings, behavioral observations, etc). Our research has focused on approximately 75 targets (and their romantic partners) who were involved in an established romantic relationship when targets were 20–21 years-old.

Assessments were made at several pivotal points of social development when targets were transitioning to new roles and exploring how to balance autonomy and interdependence with significant people in their lives, especially parents, peers, close friends, and romantic partners. As children, targets were observed in myriad social situations to assess how well they regulated (managed) their emotions in different stressful interpersonal contexts. For example, when they were 12 and 18 months old, targets were videotaped in the Strange Situation (Ainsworth, Blehar, Waters, & Wall, 1978), a stressful lab procedure that involves separations and reunions with caregivers (targets’ mothers). Securely attached children use their caregivers as a source of comfort to reduce negative affect in the Strange Situation, which allows them to engage in other life tasks, such as exploring the environment. Insecurely attached children do not use and cannot rely on their caregivers to reduce negative affect. Consequently, their attachment systems stay activated, and they remain distressed throughout the Strange Situation procedure.

When targets were two years-old, they and their mothers completed a series of videotaped tasks during which mothers tried to teach targets (their child) skills that were above the child’s capabilities. When targets were in elementary school (ages 6–8), they were rated by their classroom teachers in terms of how socially competent they were in organized classroom situations compared to their classmates. When targets were age 16, they completed an interview during which they described the nature and quality of their relationship with their best friend, including the security of their relationship and how conflicts were resolved. When targets were 20–21 years-old, they and their romantic partners came to our lab and did a videotaped conflict resolution task, which was immediately followed by a conflict recovery task (see below). Each of these developmental measures assessed how well each target managed (regulated) her or his emotions with significant others at different points of social development in different types of stressful situations. When targets were 23 years-old, we assessed whether they were still dating the same romantic partner.

Emotion Regulation during Conflict

In an initial study examining emotion regulation during relationship conflict, we (Simpson, Collins, Tran, & Haydon, 2007) found that if targets had an insecure attachment relationship with their mothers at 12 months (assessed in the Strange Situation), they reported and behaviorally expressed more negative emotions when trying to resolve a major relationship conflict with their romantic partner at age 20–21. However, this lingering effect of early attachment was mediated by targets’ degree of social competence in elementary school (rated by their grade-school teachers) and the qualities of their relationship with their best friend at age 16 (e.g., the degree to which targets felt they could share any personal feelings with their best friend, how much they trusted and could count on him/her, etc.). This mediation pattern, which fit the data better than several other possible models, is shown in Figure 1. These findings reflect one “interpersonal pathway” through which infant attachment security versus insecurity is probabilistically related to the quality of emotion regulation in adult romantic relationships 20 years later.

Figure 1.

The best-fitting mediation model showing the interpersonal pathway through which attachment security versus insecurity at age 1 is associated with greater reported and displayed negative emotions in adult romantic relationships at age 20–21.

Recovering from Conflict

In a second study, we (Salvatore, Kuo, Steele, Simpson, & Collins, 2011) examined whether and how early interpersonal experience is related to the way in which individuals recover from romantic relationship conflicts. Conflict recovery refers to how quickly, well, and completely individuals can emotionally and behaviorally “shift” from a negative interaction (e.g., discussing a major relationship problem) in order to achieve another dyadic goal (e.g., discussing topics on which both partners agree). Recovering from conflict is likely to involve a different set of skills and abilities than resolving a conflict constructively and fairly (Salvatore et al., 2011). Targets who were securely attached at 12–18 months rebounded from conflicts with their romantic partners better at age 20–21, controlling for how difficult the conflict was. Their romantic partners also recovered better if targets had been securely attached earlier in life. Moreover, having a romantic partner who recovered better from conflict was associated with greater relationship satisfaction and more positive relationship emotions. Finally, targets who had been insecurely attached as infants were more likely to still be with their partners two years later (at age 23) if their partners displayed better conflict recovery at age 20–21. Emotionally well-regulated romantic partners, therefore, can protect individuals who have insecure attachment histories from certain romantic relationship difficulties in adulthood.

Becoming the “Weak-Link” Partner

In a third study, we (Oriña, Collins, Simpson, Haydon, Salvatore, & Kim, 2011) investigated whether being treated poorly by others earlier in life leads individuals to “protect” themselves by becoming the less committed/less invested partner in their adult romantic relationships (i.e., the “weak-link” partner; Attridge, Berscheid, & Simpson, 1995). The weak-link partner in each relationship was defined as the one who scored relatively lower on commitment. We found that targets who either received less support from their mothers during challenging teaching tasks at age two or had more difficulty resolving conflicts with their best friend at age 16 were more likely to become the weak-link partner—that is, the less committed/less invested partner within their romantic relationship—at age 20–21. Furthermore, the less committed the weak-link partner was in each relationship compared to other weak-links in the sample and the larger the difference in commitment between the weak-link partner and the strong-link partner within each relationship, the more romantic couples behaved in an overtly hostile manner during their conflict discussions at age 20–21.

In all three studies, the longitudinal effects described above remained statistically significant when several measures of current romantic relationship quality and functioning were statistically controlled. These effects, therefore, are not attributable to the quality of the current relationship.

ADVANTAGES OF AN ORGANIZATIONAL PERSPECTIVE

Mental representations (working models) of past relationships tend to be carried forward into new relationships, which can change existing representations, depending on how an individual is treated in later relationships (Bowlby, 1973; Carlson et al., 2004). By adopting an organizational view on development, one can generate unique predictions about an individual’s future relationship functioning based on both his/her past and current functioning (Sroufe et al., 2005). For example, individuals can arrive at the same relationship outcome (e.g., they can have the same level of satisfaction) from different starting points; they can also arrive at different relationship outcomes from the same starting point. Cast another way, people who have different developmental histories can exhibit the same adaptation at one point in time, but different adaptations at later points of development (Sroufe et al., 1990). These differences can be predictable to the extent that an individual’s current relationship functioning partially reflects his/her cumulative developmental history up to the current relationship.

Consider a concrete example. John and Tom both feel and display strong negative affect whenever they interact with their current romantic partners. John has a secure attachment history, but he recently discovered that his partner has been cheating on him. John and his partner are trying to repair their damaged relationship in therapy, but the emotional strains of the betrayal continue to be reflected in John’s negative feelings and turbulent interactions with his partner. Tom also recently learned of his partner’s infidelity, and he too is in therapy with her working to repair the relationship. Tom, however, has an insecure attachment history. Tom, in other words, is carrying a lot of “emotional baggage” from his earlier insecure, unsupportive relationships into his current one.

When John’s and Tom’s relationships are viewed now (at a single time-point), they look very similar. Both are having problems with their current partners. A non-developmental perspective might predict that John and Tom will have fairly similar relationship trajectories and outcomes in the future. Very different predictions emerge, however, when John’s and Tom’s current adaptation is considered in combination with their divergent developmental histories. If John’s therapy is successful and he is able to trust his partner again, his relationship should show better and more rapid improvement in the future based on his secure working models. Tom’s relationship, in contrast, is likely to show poorer and slower improvement even if therapy is helpful given his insecure working models.

FUTURE DIRECTIONS AND CONCLUSIONS

There are several important directions in which future research should head. We need to learn more about how early social experiences “accumulate” across development to affect adult relationship outcomes, including non-romantic ones. We still do not know whether early social experiences occurring at specific time-points (e.g., during the first few years of life) are uniquely and independently linked to specific types of adult relationship outcomes. We also are just beginning to understand how current relationship variables statistically interact with representations of relationships earlier in life to influence adult relationship experiences and outcomes (see Salvatore et al., 2011). Future research also needs to determine whether and how specific gene x early environment interactions are associated with specific adult relationship outcomes.

In conclusion, an organizational perspective highlights the coherence of social behavior in relationships across development. The pattern of relationship-relevant thoughts, feelings, and actions—and especially their representation in working models—is what links early interpersonal experiences with caregivers to later interpersonal experiences with peers and adult romantic partners. Although the features of good versus poor emotion regulation look somewhat different at each developmental stage, their underlying functions and meaning remain fairly stable over time. Our program of research has revealed that important adult romantic outcomes are systematically related to relationship experiences that occur very early in life, well before individuals can form conscious memories. This connection, however, is affected by what happens in different types of relationships at intervening stages of social development, including relationships with romantic partners in adulthood.