The Unbearable Automaticity of Being . . . in a Close Relationship

Automatic tracking

Satisfying close relationships are critical to both mental and physical health (Proulx, Helms, & Buehler, 2007; Robles, Slatcher, Trombello, & McGinn, 2014). Yet remaining satisfied with such relationships is notoriously difficult. Not only do divorce rates in many countries hover between 30% and 50% (Amato & James, 2010), many people who marry evaluate their relationships less positively over time (Lavner & Bradbury, 2010). What is particularly puzzling is that relationship satisfaction declines despite people’s strong motivations to perceive their relationships positively and the numerous cognitive strategies that appear to effectively help them do so at any one point in time (Murray, 1999).

According to interdependence theory (Kelley & Thibaut, 1978), people evaluate the quality of their relationships by (a) considering their positive and negative experiences with those relationships, referred to as rewards and costs, and (b) utilizing the sum of those experiences to make decisions about their satisfaction and desire to stay in the relationship. However, interdependence theory is mute on how people summarize their numerous experiences and update their evaluations on a daily basis. According to several dual-process models of evaluation and related research (see Fazio & Olson, 2014; Gawronski & Strack, 2004; Jones, Olson, & Fazio, 2010; Zajonc, 1980), people automatically and unintentionally track positive and negative experiences with evaluative objects by forming associations in memory between those objects and the feelings experienced in relation to those objects. These affective associations are summarized as attitudes (Fazio, 2007), and strong attitudes are activated whenever one encounters or thinks about the attitude object (see Fazio, Sanbonmatsu, Powell, & Kardes, 1986). According to this perspective, people may track experiences involving their partners automatically as affect derived from positive and negative experiences shared with a partner becomes automatically associated with the partner in memory to form and update one’s automatic partner attitude.

Given that such associative evaluations appear to be less susceptible to motivated reasoning processes than are more deliberative judgments (e.g., Gawronski & Strack, 2004), automatic partner attitudes may track interpersonal experiences more closely than do deliberate relationship evaluations. Indeed, several studies indicate that motivated reasoning processes can bias deliberative judgments of experience (e.g., Luchies et al., 2013). For example, several longitudinal studies of dating couples demonstrated that compared with people who reported lower levels of trust in their partners, highly trusting people recalled fewer transgressions and reduced the negative emotions associated with those transgressions (Luchies et al., 2013).

In contrast, studies that have utilized both deliberative and automatic evaluations suggest that automatic evaluations may be less susceptible to motivated processes (Hicks et al., 2016, 2018; McNulty et al., 2017; Murray et al., 2010; Zayas & Shoda, 2015). For example, Zayas and Shoda (2015) demonstrated that participants’ automatic partner attitudes are marked by both positive and negative evaluative associations, although their subjective reports do not capture this ambivalence, which is consistent with the idea that most interdependent relationships involve experiences that are both rewarding and costly. Further, Hicks et al. (2016) showed that the frequency of various sexual behaviors was prospectively associated with automatic partner evaluations assessed with an implicit measure but not with self-reported relationship satisfaction. Likewise, Murray et al. (2010) demonstrated that participants with less responsive partners who felt less connected to their partners during conflict-of-interest situations had more negative automatic partner attitudes at a subsequent assessment but not more negative subjective evaluations. Moreover, in a direct test of the associative process behind automatic partner evaluations, McNulty et al. (2017) recently demonstrated that pairing the partner with positive affect from other sources (i.e., positive words and pleasant images) can increase the positivity of automatic partner attitudes relative to a control group.

Finally, in a direct test of the role of motivated processes in these evaluations, Hicks et al. (2018) demonstrated that the frequency of orgasm during sex was associated with automatic partner attitudes, whereas sexual frequency was associated only with deliberate reports of relationship satisfaction for participants who believed frequent sex was important for relationship health. Such evidence supports the idea that people’s automatic partner evaluations’ may track rewarding and costly experiences related to their partners in a way that is less susceptible to their motivations and preexisting beliefs.

Automatic influence

Of course, situating automatic partner attitudes within interdependence theory suggests that they eventually have implications for people’s relationship judgments and outcomes. According to interdependence theory, the purpose of creating evaluative summaries from relationship rewards and costs is to make subjective decisions about one’s satisfaction and desire to stay in the current relationship. Consistent with this idea, the dual-process models that guide this work suggest that automatically activated attitudes affect numerous downstream processes, such as perceptions of an object, evaluative judgments of the object, and behaviors toward the object (Fazio & Olson, 2014; Gawronski & Bodenhausen, 2006). Indeed, several studies have linked automatic partner attitudes and other automatic processes to subsequent evaluative judgments and behavior in relationships (Faure et al., 2018; LeBel & Campbell, 2013; McNulty et al., 2013, 2018; Murray, Gomillion, Holmes, & Harris, 2015). For example, more positive automatic partner attitudes have been linked to perceiving fewer problems over time (McNulty et al., 2013), self-reporting fewer destructive (Murray et al., 2015) and more cooperative relationship behaviors (LeBel & Campbell, 2013), and displaying attitude-consistent nonverbal communication in conflict discussions (Faure et al., 2018).

Further, several studies demonstrate that automatic partner evaluations predict subsequent evaluative judgments, such as relationship satisfaction and dissolution (Lee et al., 2010; McNulty et al., 2013, 2017; Scinta & Gable, 2007). For example, Lee et al. (2010) demonstrated in two studies that more positive automatic partner attitudes were associated with a decreased likelihood of dissolution the following year, even after controlling for explicit relationship satisfaction. In a more long-term test of the prognostic superiority of automatic evaluations in predicting relationship outcomes, McNulty et al. (2013) found that newlyweds’ implicit partner evaluations but not explicit satisfaction within the first few months of marriage were more predictive of their satisfaction 4 years later. These studies provide compelling evidence that automatic partner attitudes do have implications for relationship outcomes.

Future directions

Considering the role of automatic interpersonal processes in this way offers numerous compelling hypotheses for future study. Most notably, the possibilities just described suggest that future research may benefit from examining whether stress and other threats to cognitive resources moderate the link between automatic partner attitudes and subsequent relationship judgment. Likewise, future research may benefit from examining the potential mechanisms through which such links emerge other than perceptions of problems (McNulty et al., 2013) and nonverbal behavior (Faure et al., 2018). For example, given that automatic attitudes direct attention (Roskos-Ewoldsen & Fazio, 1992), it is possible that automatic partner evaluations may direct attention toward positive characteristics of the partner and away from positive characteristics of alternative partners (McNulty et al., 2018). Further, given that automatic attitudes affect attribution processes (Fazio, Roskos-Ewoldsen, & Powell, 1994), automatic partner attitudes may lead to more benevolent attributions for partner-negative behaviors (Bradbury & Fincham, 1990).

Future research may also benefit from addressing questions regarding the speed with which automatic evaluations change over time as new associations are formed (see Cone, Mann, & Ferguson, 2017) and whether recent events weigh more heavily on automatic attitudes than do past events. Early theory suggested that implicitly assessed evaluations are slower to change and should therefore be less sensitive to more recent experiences (Greenwald & Banaji, 1995; Rydell & McConnell, 2006; Wilson, Lindsey, & Schooler, 2000). Yet recent evidence challenges this idea, demonstrating that automatic attitudes may be overturned with only one piece of counter-attitudinal information (Cone et al., 2017; Gawronski, Morrison, Phills, & Galdi, 2017). In the case of close relationships, it seems plausible that the experiences shared with one’s partner 2 days ago should have more power over automatic evaluations than those that occurred 2 decades ago, though both may be represented in some way.

Nevertheless, there is no strong evidence that new information completely overrides old information in cases of strong existing attitudes, and close relationships should be an ideal context to examine how past and recent experiences shape established automatic evaluations. Additionally, given the numerous contexts in which people encounter their partners (e.g., sex, household management, parenting) and the numerous roles those partners may espouse, future researchers might benefit from examining the extent to which the valence of automatic partner attitudes is similar or different across such contexts. Finally, researchers may benefit from considering the potential bidirectional association between automatic and deliberative relationship processes (Gawronski & Bodenhausen, 2006) by examining the extent to which deliberate evaluations may influence automatic partner attitudes.

Beyond relationship science

We would be remiss if we did not also highlight that studying automatic evaluative processes in close relationship may offer insights beyond relationship science. Much of the research on automatic processes involves capturing attitudes toward novel targets or exemplars of minority groups with whom participants may not have much experience. One notable exception is a longitudinal study by Towles-Schwen and Fazio (2006). This study followed interracial college roommates over the course of one semester (~15 weeks). The researchers found that White students’ automatic racial attitudes predicted the longevity of their relationships with Black roommates, whereas their motivations to control prejudiced reactions did not. As this study illustrates, close relationships offer an opportunity to study automatic evaluations that have important implications and involve a target other than the self that is motivationally relevant and encountered repeatedly in various contexts that involve both pleasant and unpleasant affect.

Even before Bargh and Chartrand (1999) drew attention to the powerful and pervasive influence of automatic processes, research utilizing dual-process models of social cognition had resulted in valuable advances across the field of psychology. But perhaps one of the most important applications of these models in terms of the number of people affected and the implications for individual well-being has only recently been realized. By applying dual-process perspectives to the study of close relationships, we can reveal a clearer, more nuanced understanding of how people think about their partners and behave toward them, as well as why negative automatic partner attitudes ultimately become unbearable.