Mikulincer Shaver Attachment Adulthood
Mikulincer Shaver Attachment Adulthood
Mikulincer Shaver Attachment Adulthood
Attachment Processes
and Couple Functioning
299
300 INTERPERSONAL MANIFESTATIONS OF ATTACHMENT-SYSTEM FUNCTIONING
disapproval or rejection, and self-focused attention to needs and worries. Avoidant ten-
dencies are likely to manifest themselves in detachment, inhibited self-expression, lax
attention, and overemphasis on sexuality (see Chapter 12). These signs of avoidance are
intended (perhaps unconsciously) to protect a person from threats to self-worth and pres-
sures to self-disclose and become intimate or dependent.
There is evidence that attachment orientations are active at the very early stages of
a couple relationship. For example, studies have shown that a person’s chronic attach-
ment working models are transferred to new dating relationships and partners (Ahmed
& Brumbaugh, 2014; Brumbaugh & Fraley, 2006, 2007). Brumbaugh and Fraley (2007)
also found that the greater the overlap in characteristics between past and new relation-
ship partners, the greater the transference of a person’s attachment-related insecurities
from past relationships to the new relationship. Moreover, Schindler, Fagundes, and Mur-
dock (2010) found that, although attachment insecurities were not predictive of singles
progressing from not dating to dating during a 1-year period, avoidance was predictive
of not moving from casual to committed dating. Moreover, Poulsen, Holman, Busby, and
Carroll (2013) found that attachment insecurities reduced the chances of entering into
a committed dating relationship over a 32-week period, and Pittman, Kerpelman, Soto,
and Adler-Baeder (2012) found that avoidant attachment was associated with less inter-
est in acquiring information about a potential dating partner.
similarity where no such similarity has been explored or established, enlist other people’s
sympathy and support prematurely, and focus on reducing fear of rejection rather than
enhancing reciprocal intimacy.
In support of this view, attachment anxiety has been found to be associated with
stronger desire for high-intimacy disclosures with a dating partner (Vohs, Baumeister, &
Ciarocco, 2005) and higher levels of actual relationship-focused disclosure (talking about
closeness-related issues) during the early stages of a relationship (Tan, Overall, & Tay-
lor, 2012). S. A. Bradford, Feeney, and Campbell (2002) also found that more anxiously
attached people rated the disclosures they made to a dating partner over a 7-day period
as more negative in tone, and their partners also felt less satisfied with these disclosures.
In a laboratory experiment, Cameron, Holmes, and Vorauer (2009) provided evi-
dence that the tendency of people with negative models of self to avoid self-disclosure
can be due to their negative expectations about the relational consequences of such dis-
closure. Participants were told that they had failed a competence test, and half of them
were asked to reveal the failure to their dating partner. When asked to disclose their
failure, participants with lower self-esteem, which is typically correlated with attachment
anxiety (see Chapter 6), held more negative expectations about the future of their dating
relationship than those told to conceal their failure. In contrast, participants with higher
self-esteem who were asked to disclose their failure reported more positive relational
expectations than those who were asked to conceal it.
Insecure people’s difficulties with self-disclosure were also manifested in K. M.
Carmichael and Tyler’s (2012) study of the effects of disclosure on available cognitive
resources (i.e., the extent to which self-disclosure is an effortful task). Participants were
instructed to disclose to a dating partner high- or low-intimacy topics, and the subse-
quent availability of cognitive resources was measured as the length of time participants
worked on anagram puzzles. Avoidant attachment was associated with less persistence
in attempting to solve the anagrams only in the high-intimacy condition, reflecting the
strenuous effort for avoidant individuals of discussing high-intimacy topics. In contrast,
anxious attachment was associated with less persistence only in the low-intimacy condi-
tion. This finding reflects the problems that anxious people may have in low-intimacy
situations due to the instruction to inhibit their tendency to disclose intimate topics.
In an examination of insecure people’s disclosures during attempts to entice a poten-
tial dating partner, Brumbaugh and Fraley (2010) videotaped participants who were
competing with others for a date with an attractive individual and asked raters to code
participants’ behaviors and statements during the interaction. More anxious people were
more likely to present themselves as warm, engaging, and humorous when communicat-
ing with potential mates. However, they also tended to present themselves as less secure
and to express more worries. More avoidant individuals were more likely to use physi-
cal touch to convey interest in the partner. Since we know that avoidant people tend to
withdraw from emotional closeness (see Chapter 9), it is possible that they use touch to
misrepresent themselves as comfortable with closeness and may use physical closeness
as a superficial proxy for psychological closeness. However, raters judged more avoidant
participants as less secure and more neurotic (like the more anxious participants).
Secure individuals’ self-disclosures are guided by the goals of mutual enjoyment
and intimacy. Mikulincer and Nachshon (1991), Keelan et al. (1998), and Grabill and
Kerns (2000) found that secure study participants disclosed more personal information
to a high- than to a low-disclosing partner (indicating sensitivity to the partner’s behav-
ior). They were able to elicit more disclosures from their partner and were more atten-
tive to thoughts and feelings expressed by the partner. This combination of reciprocal
302 INTERPERSONAL MANIFESTATIONS OF ATTACHMENT-SYSTEM FUNCTIONING
Mating Preferences
Beyond their disadvantageous position when dating, insecure people tend to possess
problematic mate preferences, which further jeopardize their chances of establishing a
Attachment Processes and Couple Functioning 303
TABLE 10.1. A Summary of Findings Concerning Partner’s Attachment Orientation and Consequent
Attractiveness
Main finding
Attachment Manipulation of Type Dependent for partner’s
Study measure partner’s traits of partner variable attachment style
Pietromonaco & HS ratings Behavioral vignettes Imaginary Positive Sec > Anx = Avo
Carnelley (1994) of three styles emotions
Negative Avo > Anx > Sec
emotions
Baldwin et al. HS ratings Questionnaire Potential Attractiveness Sec > Anx > Avo
(1996, Study 3) responses for each
of three styles
Latty-Mann & RQ Prototypical Ideal Attractiveness Sec > Anx > Avo = Fea
Davis (1996) description of four
styles
P. A. Frazier et al. HS ratings Questionnaire Potential Attractiveness Sec > Anx = Avo
(1996, Study 2) responses for each
of three styles
P. A. Frazier et al. AAS Questionnaire Potential Attractiveness Sec > Anx > Avo
(1996, Study 3) responses for each
of three styles
Chappell & Davis RQ Behavioral Imaginary Positive Sec > Anx > Avo = Fea
(1998) vignettes matching emotions
four styles Negative Fea = Avo = Anx > Sec
emotions
Preference for Sec > Anx = Avo = Fea
date
Klohnen & Luo RQ, ECR Behavioral vignettes; Imaginary Attractiveness Sec > Anx > Avo = Fea
(2003, Study 1) four styles
Klohnen & Luo RQ, ECR Behavioral vignettes; Imaginary Attractiveness Sec > Anx > Avo = Fea
(2003, Study 2) four styles
Klohnen & Luo RQ, ECR Questionnaire Potential Attractiveness Sec > Anx > Avo = Fea
(2003, Study 3) responses; three
styles
satisfying relationship. Some studies (Baldwin et al., 1996; P. A. Frazier, Byer, Fischer,
Wright, & DeBord, 1996; Klohnen & Luo, 2003; Latty-Mann & Davis, 1996; C. Strauss
et al., 2012) indicate that, even though secure partners were favored overall, insecure
people were more favorable than secure people toward insecure partners. Specifically,
anxious individuals were more attracted to anxiously attached partners, and avoidant
people were more attracted to avoidant partners. However, Pietromonaco and Carnel-
ley (1994) and C. Strauss et al. (2012) found some support for a complementary pat-
tern of mate preferences, by which avoidant people preferred an anxious partner. In any
case, these patterns of attraction are likely to undermine insecure people’s relationships,
because relationships between two insecure people tend to fare less well, on average, than
ones that include at least one secure person (as shown later in this chapter).
There is also evidence that insecure people tend to date less secure partners. In a
sample of adult women who had suffered abuse during childhood, G. McCarthy (1999)
found that those with an anxious or avoidant attachment style were more likely than
secure women to cohabit with a socially deviant (e.g., delinquent, drug-abusing) part-
ner. Moreover, Collins et al. (2002) reported that avoidant attachment (assessed during
304 INTERPERSONAL MANIFESTATIONS OF ATTACHMENT-SYSTEM FUNCTIONING
Dating Outcomes
Attachment insecurities seem to create many obstacles to dating success. For example,
using a speed-dating paradigm, McClure, Lydon, Baccus, and Baldwin (2010) found that
attachment-anxious participants were relatively unpopular and unselective during speed
dating (they missed fewer opportunities but made more failed attempts, a strategy that
has been found to increase rejection in speed dating; Eastwick, Finkel, Mochon, & Ari-
ely, 2007). It seems likely that the hunger for love that characterizes attachment-anxious
people leads them to adopt an unselective, indiscriminate approach to dating even at the
cost of increasing the likelihood of rejection. In another study, Birnie, McClure, Lydon,
and Holmberg (2009) found that avoidant attachment was associated with both expecta-
tions of failure in a dating interaction and aversion to commitment. These findings sug-
gest that avoidant people enter new relationships with detailed scripts for commitment
aversion and expectations for relationship failure, which in turn may reduce the likeli-
hood of moving toward a more committed relationship.
Recently, McClure and Lydon (2014) systematically documented the obstacles that
attachment anxiety creates during initial interactions with a potential romantic partner
and that can lead to interpersonal failure when relational opportunities are presented. Spe-
cifically, they assessed behaviors and interpersonal outcomes while participants engaged
in speed dating; made a video introduction of themselves to an attractive, opposite-sex
confederate who would ostensibly be deciding whether to meet them; or interacted with
an attractive, friendly, preferred-sex confederate for 40 minutes. Across the various con-
texts, independent judges rated more anxiously attached people as less attractive and less
interpersonally appealing. They were also rated as displaying more social disengagement,
anxiety, and interpersonal awkwardness. These behaviors were also found to mediate
anxious people’s negative dating outcomes.
However, despite findings showing that attachment insecurities can impair the
quality of dating interactions, there is some evidence for potential beneficial effects of
attachment anxiety during relationship initiation. Eastwick and Finkel (2008) found that
experimental manipulations of state-like experiences of attachment anxiety heightened
behavioral tendencies that contribute to the success of dating interactions, such as prox-
imity seeking, expressions of warmth and love, and other approach behaviors. It seems
that attachment anxiety may signal to a potential partner that one is interested in form-
ing an emotional bond. In addition, Tomlinson, Carmichael, Reis, and Aron (2010) asked
people, before they entered a dating relationship, to predict their likely level of happiness
once they had entered such a relationship. Findings indicated that participants scoring
higher on attachment anxiety predicted that they would be happier, although their actual
happiness after first dates was relatively low. This suggests that attachment-anxious
Attachment Processes and Couple Functioning 305
to perceive relational rewards and benefits (Spielmann, Maxwell, et al., 2013), anxious
people are more likely to emphasize relational threats and losses (Boon & Griffin, 1996;
Spielmann, Maxwell, et al., 2013). In addition, more anxious people are less confident of
being able to establish a successful relationship than secure people (Carnelley & Janoff-
Bulman, 1992; Pietromonaco & Carnelley, 1994; Whitaker, Beach, Etherton, Wakefield,
& Anderson, 1999) and hold more dysfunctional relationship beliefs, such as that their
behavior in relationships can be destructive (J. Feeney & Noller, 1992; Mohr, Crook-
Lyon, & Kolchakian, 2010; Stackert & Bursik, 2003; Whisman & Allan, 1996). They
are also more likely to perceive their relationships as inequitable because their partners
are not contributing as much as they should (Grau & Doll, 2003). Moreover, experimen-
tal priming of anxious attachment intensifies pessimistic expectations and perceptions of
inequity (Grau & Doll, 2003; Whitaker et al., 1999).
Research has also documented attachment-related variations in “love styles” (i.e.,
kinds of attitudes about romantic love and couple relationships). Authors of eight studies
have used Hendrick and Hendrick’s (1986) Love Attitudes Scale, and three findings recur
across most of these studies (Collins & Read, 1990; J. Feeney & Noller, 1990; Fricker &
Moore, 2002; Galinha, Oishi, Pereira, Wirtz, & Esteves, 2014; Heaven, Da Silva, Carey,
& Holen, 2004; Hendrick & Hendrick, 1989; M. B. Levy & Davis, 1988; Zamora, Win-
terowd, Koch, & Roring, 2013). First, secure attachment is positively associated with
eros (romantic, passionate love) and agape (selfless, altruistic, all-giving love), the love
styles that facilitate the consolidation of lasting romantic bonds. Second, avoidance is
positively associated with ludus (game-playing, noncommitted love). Third, attachment
anxiety is positively associated with mania (possessive, dependent love). In addition,
some of the studies suggest that avoidance is associated with lower agape scores (Collins
& Read, 1990; J. Feeney & Noller, 1900; Hendrick & Hendrick, 1989; M. B. Levy &
Davis, 1988) and higher pragma (logical, “shopping-list” love) scores (Collins & Read,
1990; Heaven et al., 2004).
Supplementing these findings, there is evidence that more avoidant people score lower
on scales assessing passionate (erotic) and companionate (agapic) forms of love (Årseth,
Kroger, & Martinussen, 2009; Cruces et al., 2015; R. W. Doherty, Hatfield, Thomp-
son, & Choo, 1994). In addition, J. Feeney and Noller (1990) and Sperling and Berman
(1991) found that attachment anxiety was associated with higher scores on scales mea-
suring love addiction and desperate love. This desperate need for love may also explain
Aron, Aron, and Allen’s (1998) finding of more frequent unreciprocated love experiences
among people with an anxious attachment style.
Intimacy
Attachment orientation is likely to affect the progression of intimacy during the consoli-
dation phase of romantic relationships. Based on a theoretical analysis of secure people’s
working models, Pistole (1994) reasoned that secure people would be unlikely to expe-
rience much conflict related to closeness and distance. These people’s well-developed
interpersonal sensitivity and conflict management skills (Chapter 9) should allow them
to accurately gauge the amount of closeness sought by their partner and be able to toler-
ate and communicate effectively about any momentary violations, in either direction, of
desired personal boundaries.
Pistole (1994) also predicted that insecure people would be less able to negotiate
issues related to closeness and distance. Avoidant people’s preference for interpersonal
distance was expected to interfere with both their own intimacy-promoting behavior and
Attachment Processes and Couple Functioning 307
their responsiveness to a partner’s bids for intimacy. Anxious people’s unmet needs for
love were expected to cause them to seek closeness to such an extent that it would make
their partners uncomfortable. Moreover, fear of rejection was expected to cause them to
misinterpret a partner’s desire for privacy or autonomy as a sign of rejection, which could
tempt them to escalate demands for intimacy to such an extent that, paradoxically, it
might cause their partner to withdraw or flee. Pistole (1994) noted that this kind of intru-
sion would be most unwelcome to avoidant partners, who view even normal intimacy as
intrusive. This relational pattern, called “demand–withdrawal” (Christensen, 1988) or
“pursuit-withdrawal” (Bartholomew & Allison, 2006) in the marital research literature,
is one of the major predictors of relationship distress.
These attachment-related differences in construal of intimacy and closeness were
found in two studies that involved asking dating partners to talk freely about their rela-
tionship (J. Feeney, 1999b; J. Feeney & Noller, 1991). A content analysis of their remarks
revealed that secure people emphasized the importance of establishing a balance between
closeness and independence; avoidant people emphasized the need to place limits on
closeness; and anxious people emphasized the importance of closeness but not the impor-
tance of independence.
Additional evidence comes from self-report studies of couple intimacy (see Table
10.2 for a summary). Across numerous studies, secure people reported greater intimacy
in their relationships than either anxious or avoidant people. This pattern appeared
whether intimacy was assessed concurrently or as much as 6 years later (Collins et al.,
2002). In addition, Mikulincer and Erev (1991) found that avoidant people underesti-
mated their partner’s sense of intimacy, which may have served a defensive function. If
avoidant people can downplay their partner’s sense of intimacy, they can more easily
maintain their own independence and distance. In addition, anxious people desired more
intimacy than they were experiencing (Mikulincer & Erev, 1991). It wasn’t clear whether
anxious people were simply unable to elicit the extreme level of intimacy they desired
from their partners, or whether their intrusive style of relating caused their partners to
establish greater self-protective distance.
Attachment-related deficits in intimacy have also been documented in studies exam-
ining synchrony in emotional experiences between relationship partners and the extent to
which one partner’s momentary moods influence the other partner’s emotions (emotional
transmission). For example, Butner, Diamond, and Hicks (2007) recorded daily ratings
of emotions in both partners of dating couples and found that partners’ attachment inse-
curities inhibited emotional synchrony within couples. Whereas more anxiously attached
people showed lower synchrony of negative affect with their partner, more avoidant peo-
ple were less influenced by their partner’s positive affect. Recently, Randall and Butler
(2013) used daily diaries and second-by-second measures of emotional experience from
dating couples and found that partners’ attachment anxiety was associated with lower
levels of second-by-second emotion transmission within the couple. Overall, these find-
ings indicate that attachment insecurities thwart emotional closeness within romantic
couples and prevent partners’ sense that they are sharing their emotions.
Attachment-anxious people’s difficulties in regulating closeness and distance within
romantic relationships are also documented in studies examining various manifestations
of intrusiveness. For example, there is evidence that attachment-anxious people are more
likely to engage in stalking and unwanted pursuit of an ex-spouse mainly when they did
not want the breakup to occur (De Smet et al., 2013; Derlega et al., 2011; L. B. Dutton &
Winstead, 2006; Tassy & Winstead, 2014; Wigman, Graham-Kevan, & Archer, 2008).
Studies have also linked anxious attachment to cyberstalking (Ménard & Pincus, 2012;
308 INTERPERSONAL MANIFESTATIONS OF ATTACHMENT-SYSTEM FUNCTIONING
Strawhun, Adams, & Huss, 2013). In addition, there are both cross-sectional and diary
evidence that attachment anxiety is associated with heightened Facebook-related surveil-
lance of a romantic partner (Fox & Warber, 2014; T. C. Marshall, Bejanyan, Di Castro,
& Lee, 2013).
Lavy, Mikulincer, and Shaver (2010) also examined associations between attach-
ment insecurity (ECR) and intrusiveness in couple relationships (engaging in intrusive
behavior, perceiving a partner as intrusive, subjective experiences of being intrusive, and
reacting to intrusive behavior). Attachment anxiety was associated with more intrusive
behavior, more ambivalent reactions to partner intrusiveness, and greater emotionality
when being intrusive. Avoidance was associated with perceiving a partner as intrusive,
reacting critically and establishing distance in response to partner intrusiveness, and feel-
ing concerned and caring when being intrusive. In a subsequent study, Lavy, Mikulincer,
and Shaver (2013) asked members of dating couples to complete daily measures (over a
14-day period) of relationship satisfaction and intrusive behavior. Results indicated that
daily intrusiveness was associated with one partner’s attachment anxiety and the other
partner’s avoidance. Moreover, anxious people reported more intrusive behavior mainly
when they had been dissatisfied with their relationship the previous day, implying that
intrusiveness may be part of their effort to restore relational closeness. Like other anxious
behaviors, however, these closeness regulation efforts are likely to backfire if they cause
a partner to feel imposed upon.
Commitment
Attachment insecurities can interfere with relational commitment. Both avoidant needs
for independence and anxious approach–avoidance ambivalence combined with doubts
about a partner’s trustworthiness can interfere with committing oneself to a lasting rela-
tionship. Moreover, avoidant distancing and anxious intrusiveness can deter partners
from committing themselves to what they fear might be a troubled, unsatisfying relation-
ship.
A large body of evidence supports the hypothesis that attachment insecurities are
cross-sectionally and prospectively associated with lower relational commitment (see
Table 10.3 for a summary). Attachment insecurities are also associated with lower scores
on scales measuring engagement in relationship maintenance behaviors (Adams & Bap-
tist, 2012; Baptist, Norton, Aducci, Thompson, & Cook, 2012; Canary, Stafford, &
Semic, 2002; Edenfield, Adams, & Briihl, 2010; Pistole, Roberts, & Chapman, 2010).
Moreover, Gillath and Shaver (2007) found that security priming (recalling a warm and
loving relationship) caused participants to adopt more relationship maintenance behav-
iors, and Duemmler and Kobak (2001) found that appraisals of a dating partner’s sup-
portiveness predicted increases in relationship commitment over an 18-month period. In
another diary study, Overall and Sibley (2008) found that more avoidant people were
more likely to be attracted by alternative partners (another sign of low commitment) in
the course of their daily interactions.
In a behavioral test of relational investment, Vicary and Fraley (2007) asked people
to choose between two options in several “Choose Your Own Adventure” dating story
tasks. One option was always a relationship-enhancing option; the other option was
detrimental to the relationship. They found that attachment security was associated with
choosing relationship-enhancing options. In a subsequent study, Turan and Vicary (2009)
replicated this finding and examined the effects of priming security by manipulating the
310 INTERPERSONAL MANIFESTATIONS OF ATTACHMENT-SYSTEM FUNCTIONING
supportiveness of the fictitious partner. Findings indicated that priming a sense of secu-
rity (expecting a partner to be supportive) contributed to the choice of more relationship-
enhancing options.
Another aspect of commitment that is disturbed by attachment insecurities is the
willingness to make sacrifices for the relationship. In a 14-day diary study of dating
couples, Impett and Gordon (2010) found that avoidant attachment was associated with
less frequent sacrifices for the relationship, less willingness to sacrifice to improve rela-
tionship quality and the partner’s well-being (approach goals), and greater willingness
to sacrifice for preventing relational tensions (avoidance goals). Attachment anxiety was
associated with greater willingness to sacrifice if it served one’s own self-focused goals.
These findings were conceptually replicated in subsequent studies (Etcheverry, Le, Wu, &
Wei, 2013; Mattingly & Clark, 2012) and were supplemented by Hui and Bond (2009),
who found that avoidance interfered with another pro-relational sacrifice: improvement
in personal domains that a partner finds deficient. In addition, Ruppel and Curran (2012)
reported that the positive effect of daily sacrifices on relationship satisfaction over a
10-day period was stronger for more secure participants. That is, relational sacrifices
might be most beneficial to relationship quality for secure people who seem to value and
endorse this kind of pro-relational behavior.
In a more refined analysis of the attachment–commitment link, Pistole, Clark, and
Tubbs (1995) found that although both anxious and avoidant people reported relatively
low levels of commitment to their partners, the anxious ones reported the highest rela-
tionship costs while the avoidant ones reported the lowest investments. This implies that
anxious people’s lack of commitment stems from disappointment, pain, and frustration,
whereas avoidant people’s lack of commitment stems from unwillingness to invest in a
long-term relationship. This conclusion is supported by Himovitch (2003), who found
that anxious people have quicker mental access to memories of episodes in which a part-
ner decreased commitment to them, which we interpret as another example of hypervigi-
lance to signs of rejection and betrayal, whereas avoidant people had quicker access to
memories of episodes in which they decreased their commitment to their partner, which
we interpret as a case of attachment-system deactivation. Recently, Dandurand et al.
(2013) reported that anxious attachment was associated with both approach and avoid-
ance commitment-related goals—the desire to maintain a relationship due to benefits
inherent within the relationship and the desire to maintain a relationship to avoid poten-
tial losses resulting from relationship dissolution. In contrast, more avoidant people were
less likely to endorse approach commitment-related goals, probably because they doubt
or discount the benefits of relationship maintenance (Spielmann, Maxwell, et al., 2013).
There is also evidence that attachment-anxious people hold ambivalent attitudes
toward relational commitment. For example, Mikulincer and Erev (1991) found that anx-
ious people were more likely than avoidant ones to want a highly committed relationship,
and Senchak and Leonard (1992) found that anxious men acquired marriage licenses
much sooner than secure or avoidant men. These findings raise the question of why anx-
ious people score relatively low on commitment scales even though they apparently value
commitment. It is possible that their tendency to commit too early, often before they
know their partner very well leaves anxious people more vulnerable to entanglement with
a hurtful, uncommitted partner who reduces their relational satisfaction and sense of
commitment (Morgan & Shaver, 1999). This possibility received some support in a study
by Joel, MacDonald, and Shimotomai (2011), who found two indirect antagonistic paths
linking anxious attachment and commitment. On the one hand, attachment anxiety was
associated with relationship dissatisfaction, which in turn lowered commitment. On the
312 INTERPERSONAL MANIFESTATIONS OF ATTACHMENT-SYSTEM FUNCTIONING
other hand, attachment anxiety was associated with overdependence on the relationship
partner, which increased commitment. These complex findings are indicative of anxious
people’s ambivalence: Whereas the upward pressure exerted by relational dependency
can increase anxious people’s commitment, dissatisfaction can reduce relationship com-
mitment.
This line of thinking is further supported in a recent series of studies conducted
by Gere, MacDonald, Joel, Spielmann, and Impett (2013), who examined whether per-
ceptions of threats and rewards in couple relationships can explain the attachment–
commitment link. Findings indicated that avoidant attachment was uniquely associated
with lower perceptions of rewards in couple relationships, which in turn explained why
more avoidant people are less committed to their relationships. Anxious attachment
was uniquely associated with stronger perceptions of relational threats, which in turn
contributed divergently to lower levels of relational satisfaction and higher levels of rela-
tional investment. It seems that heightened perception of relational threats underlie the
ambivalent attitude of anxiously attached people toward commitment by concomitantly
reducing relational satisfaction and increasing investment.
Although secure individuals are often more committed than insecure individuals to
their relationships, they can adaptively reduce commitment if a relationship is unsatis-
factory (e.g., when a partner fails to meet their legitimate psychological needs). In both
a longitudinal study and a laboratory experiment, Slotter and Finkel (2009) found that
when a partner failed to help participants fulfill their relatedness and autonomy needs,
more securely attached people reported declining commitment over time and an increased
risk of breakup. This finding implies that securely attached people’s commitment to a
relationship is more a means for assuring the maintenance of a satisfactory relationship,
not a sign of unmitigated hunger for love and care.
However, anxious and avoidant people differ in the transfer of attachment functions
to a romantic partner during the initial stages of a relationship. For example, Eastwick
and Finkel (2008) found that whereas attachment anxiety accelerated the appraisal of a
new dating partner as an attachment figure, probably due to previously unmet needs for
safety and security, avoidance reduced the tendency to make such appraisals. Similarly,
Fagundes and Schindler (2012) reported that more anxious people preferred their part-
ner for proximity earlier than less anxious people, and more avoidant people reported
less preference for their romantic partner as an attachment figure even 2 years into the
relationship. In addition, avoidance seems to deter reliance on a need-based, noncontin-
gent norm for security provision in young married couples and to favor exchange norms
regarding support that can interfere with the consolidation of a stable and satisfactory
relationship (M. S. Clark, Lemay, Graham, Pataki, & Finkel, 2010). In addition, Sadikaj,
Moskowitz, and Zuroff (2015) found that more avoidant people reported lower levels of
felt security during daily interactions with their dating partner over a 20-day period. At
the same time, partners of more avoidant people also reported less felt security during
these interactions.
Attachment-anxious people’s tendency to satisfy unmet needs for security is also
evident in their tendency to engage in excessive reassurance seeking from a romantic
partner. In Chapter 9, we reviewed studies yielding strong correlations between anxiety
and excessive reassurance seeking. The correlations were corroborated by Shaver et al.
(2005), who assessed daily variation in reassurance seeking over a 2-week period. More
anxious participants were more likely to seek daily reassurance from their dating partner.
In addition, anxious men sought more reassurance on days following relationship con-
flicts. Thus it seems that anxious people seek reassurance as a way of quelling worries
that interfere with relationship-specific security.
Dyadic Communication
Verbal and nonverbal exchanges in which both partners feel free to express their thoughts
and feelings in an affectionate and loving way seem to be critical for maintaining high-
functioning long-term relationships (Sprecher et al., 2015). These open and comfortable
exchanges can be undermined or distorted by attachment insecurities. Avoidant attach-
ment may reduce partners’ interest in engaging in affectionate interactions and may pose
difficulties for expressing concerns and feelings and for responding sensitively to a part-
ner’s needs and comments. Attachment anxiety may pose difficulties for attending accu-
rately to a partner’s thoughts and feelings because of self-focus and worries about being
rejected.
Studies based on self-reports of couple communication indicate that insecure people
get involved more often in destructive patterns of communication, including demand–
withdrawal interactions (Crowley, 2008; Ebrahimi & Ali Kimiaei, 2014; Fitzpatrick, Fey,
314 INTERPERSONAL MANIFESTATIONS OF ATTACHMENT-SYSTEM FUNCTIONING
Segrin, & Schiff, 1993; Givertz & Safford, 2011; Mohr, Selterman, & Fassinger, 2013).
In addition, more avoidant people report higher levels of emotional control (bottling
up emotions; J. Feeney, 1995a, 1999a) and less frequent use of affectionate expressions
in conversations with romantic partners (Bombar & Littig, 1996; Dillow, Goodboy, &
Bolkan, 2014). J. Feeney (1999a) also found that attachment anxiety was associated with
greater self-perceived suppression or control of negative emotions, perhaps reflecting
reluctance to express neediness after having been criticized for doing so too often or too
intrusively in the past. Insecure people’s problems with dyadic communication have also
been documented in a diary study of newlywed couples (J. Feeney, Noller, & Callan,
1994). Couples in which the husband was avoidant reported lower levels of conversa-
tional involvement and satisfaction, and couples in which the wife was anxious reported
higher levels of conflict during the conversation.
Avoidant people’s cool, distant, and disinterested communication style has been
observed in studies of actual conversations between dating partners. For example, Guer-
rero (1996) found that avoidance was associated with lower levels of facial gaze, facial
pleasantness, general interest in the conversation, and attentiveness to the partner’s com-
ments during a videotaped discussion of personal problems. In another study, Tucker and
Anders (1998) found that more avoidant people laughed less, touched their partner less,
looked less at their partner, and smiled less during a videotaped conversation about posi-
tive aspects of the couple relationship. Le Poire, Shepard, and Duggan (1999) reported
similar findings in a couple-level analysis of dyadic communication. Avoidant people’s
lack of involvement has also been found in studies assessing text messaging and email
usage with dating partners (Jin & Peña, 2010; Luo, 2014; Morey, Gentzler, Creasy, Over-
hauser, & Westerman, 2013; Weisskirch, 2012).
Observational studies also provide evidence for the tense climate that attachment-
anxious attached people establish during dyadic conversations (Guerrero, 1996; Roisman,
Holland, et al., 2007; Tucker & Anders, 1998). For example, Guerrero (1996) found that
attachment anxiety was associated with more vocal and physical signs of distress during
videotaped conversations. Given anxious individuals’ strong desire for intimacy, it was
surprising that they didn’t seek greater proximity by gazing at and touching their partner.
According to Tucker and Anders (1998), “preoccupied [anxious] individuals are inclined
to engage in intimate behaviors, but learn that relationship maintenance demands, in
part, a suppression of their overt attempts to engage their partners, as ‘clingy’ behaviors
might cause their partners to withdraw” (p. 121).
Attachment insecurities also tend to be associated with misperception of a relation-
ship partner’s signals. For example, Noller and Feeney (1994) and J. Feeney, Noller, and
Callan (1994) asked both members of newlywed couples to send a set of nonverbal mes-
sages expressing particular emotional states and intentions to their spouse (e.g., sadness,
anger, support seeking) and to decode the feelings and intentions expressed by the spouse.
The more anxious husbands and more avoidant wives were less accurate in decoding their
partners’ nonverbal messages. This finding was prospectively replicated when accuracy
was assessed 9 months later. In a dyadic experiment, L. A. Beck et al. (2014) reported
evidence that more avoidant people also tend to misperceive signs of their own and their
partner’s responsiveness. Newlywed couples were videotaped during a conflict resolution
discussion and then reported perceptions of their own and their partners’ responsiveness
during the conflict. Observers also coded both partners’ responsive behaviors during the
conflict. As compared to observers’ ratings, more avoidant participants underestimated
both their own and their partner’s responsiveness.
These attachment-related inaccuracies are also evident in the perception of a romantic
Attachment Processes and Couple Functioning 315
partner’s emotions. Overall, Fletcher, Simpson, and Fillo (2015) assessed participants’
perceptions of partners’ emotions and partners’ actual emotions during couples’ conflict
discussions and during daily interactions over a 3-week period. Using partners’ reports of
their own emotions as the accuracy benchmark, they found that, although no attachment
effect was found in the accurate detection of shifts in partners’ emotions, more avoidant
participants overestimated the intensity of their partners’ negative emotion. Importantly,
this overestimation of negative emotions explained avoidant participants’ heightened
hostile and defensive responses to partner’s expression of emotions during the recorded
interactions.
In a study of empathic accuracy (the ability to infer a partner’s feelings correctly),
Simpson, Ickes, and Blackstone (1995) found that people who described their dating rela-
tionship as less secure were less accurate when trying to infer their partner’s feelings from
a videotaped discussion of pictures of opposite-sex people with whom the partner might
later interact. Using a similar procedure, Simpson, Ickes, and Grich (1999) unexpect-
edly discovered that more anxious people scored higher on empathic accuracy. However,
among anxious participants, greater accuracy was associated with subsequent negative
changes in relationship quality. That is, for anxious people empathic accuracy may have
negative relational consequences, because they are particularly able to detect partner’s
feelings that pose a threat to the relationship.
Simpson et al. (2011) reported similar attachment-related differences in empathic
accuracy. In one study, married couples were videotaped discussing a severe or a less
severe intimacy-related problem. In the second study, dating couples were videotaped
trying to resolve a relationship conflict. Across the two studies, more avoidant people
were less accurate in inferring a partner’s feelings during the conversation, whereas more
anxiously attached people scored higher on emphatic accuracy. However, this heightened
accuracy was mainly found when discussing issues that posed a threat to the relation-
ship and when participants were rated as more distressed when discussing a relationship
conflict. That is, anxious people’s empathic accuracy can be a manifestation of hyperac-
tivation strategies aimed at regulating distress, ones that paradoxically amplify worries
about relationship security.
Guerrero and Burgoon (1996) provided evidence of another difficulty encountered
by anxiously attached people during couple conversations: the inability to withdraw from
a dissatisfactory interaction. Couples participated in two conversations, and after the
first, one partner was asked to decrease involvement in the second conversation. More
anxious people became more involved in the conversations following a scripted decrease
in a partner’s involvement, probably reflecting their compensatory efforts to restore the
partner’s interest. This tendency to persist in an unpleasant interaction, which might on
its face seem laudable or beneficial, can take a destructive turn (as discussed later in this
chapter) when anxious people prove unable or unwilling to leave a dysfunctional relation-
ship.
(Gaines et al., 1997, 1999; Gaines, Work, Johnson, Youn, & Lai, 2000; Gaines & Hen-
derson, 2002; Pizzano, Sherblom, & Umphrey, 2013; Scharfe & Bartholomew, 1995).
In general, attachment insecurities are associated with less “voice” (active attempts to
solve a problem) and less “loyalty” (understanding the temporary nature of a partner’s
behavior and waiting for improvement), the two most accommodative, constructive
responses to a partner’s transgressions. Attachment insecurities tend to be associated
with more “exit” responses (attempts to harm the partner or leave the relationship) and
“neglect” responses (ignoring the partner and refusing to discuss the problem), the two
most destructive responses. Recently, Arriaga, Capezza, Reed, Wesselmann, and Wil-
liams (2014) found that more insecure people reacted more negatively to episodes in
which their dating partner ostracized them during a Cyberball game. Specifically, more
avoidant people evaluated their relationship more negatively as a function of a partner’s
ostracism, and more anxious people were more likely to report that their basic needs were
thwarted following a partner’s ostracism.
Conceptually similar findings were obtained in studies of reactions to a relationship
partner’s betrayal of trust (Jang, Smith, & Levine, 2002; Mikulincer, 1998c). Less secure
people were less likely to talk openly with their partner about his or her deception. In
addition, anxious people were more likely to ruminate about their partner’s betrayal,
avoid discussing the partner’s deception, and react with strong negative emotions. Avoid-
ant people, in contrast, increased their distance from the transgressing partner and denied
the importance of the threatening episode, which exemplifies their general attempt to
keep their attachment system deactivated. However, Mikulincer (1998b) measured lexical
decision times to the word “worry” and found that avoidant people implicitly activated
the word “worry” when primed with a trust violation story. This discrepancy between
self-reports of not being bothered by betrayal and implicit signs of worry may hint at the
fragile nature of avoidant people’s defenses, a finding that recurs throughout this book
(see Chapters 7, 9, and 13).
In a study of hurt feelings in couple relationships, J. Feeney (2004a) asked partici-
pants to recall and describe an event in which a dating partner said or did something
that hurt their feelings. Attachment-anxious participants were more likely to report that
the hurtful event had negative long-term effects on their self-esteem, and this association
was mediated by the report of more distress and more negative self-perceptions (“I’m
stupid”) following the event. That is, attachment-anxious individuals tended to react to
hurtful events in couple relationships with relatively strong distress and negative self-
views, which in turn seemed to exacerbate their self-related doubts and worries. Avoid-
ant attachment was associated with more negative perceived effects of the hurtful event
on the relationship, and this association was mediated by lower perceptions of partner
remorse and more destructive behavioral reactions to the hurtful event. That is, more
avoidant individuals were less likely to accept their partner’s remorse and more likely
to act destructively, which in turn seemed to aggravate relational tensions and conflicts.
Overall, it seems that hurtful events in couple relationships tend to exacerbate both anx-
ious people’s negative self-views and avoidant people’s negative views of others and rela-
tionships.
In a recent study, Overall, Girme, Lemay, and Hammond (2014) focused on a spe-
cific strategy that anxiously attached people can use in response to relational transgres-
sions that can concomitantly express their frustration and hurt feelings and move the
transgressing partner closer to them: guilt-inducing responses. Findings indicated that
on days when they received a partner’s criticism or during a videotaped conflict dis-
cussion with their partner, more anxiously attached participants exhibited exaggerated
Attachment Processes and Couple Functioning 317
expressions of hurt feelings and more guilt-inducing verbal and nonverbal strategies (as
coded by independent observers). More important, partners of more anxious participants
actually reported higher levels of guilt, and their guilt level was positively associated with
anxious participants’ use of guilt-induction strategies. In addition, more anxious indi-
viduals reported more positive appraisals of their partner and the relationship when their
partner felt more guilt, but these benefits were accompanied by the partner’s heightened
reports of relationship dissatisfaction. These results illustrate the manipulative, control-
ling stance that anxiously attached people adopt in response to relationship transgres-
sions, which can momentarily create an illusion of intimacy and commitment (due to a
partner’s compensatory efforts to reduce his or her guilt feelings) but can in the long run
erode the partner’s satisfaction and relationship quality.
Studies have also provided evidence that the attachment–forgiveness link reviewed in
Chapter 9 is observed within couple relationships. In two correlational studies of dating
and married couples, Kachadourian et al. (2004) found that relatively insecure people
were less likely to forgive their romantic partners. In a 21-day diary study of married
couples, Mikulincer et al. (2006) found that attachment anxiety and avoidance predicted
lower levels of daily forgiveness toward a spouse. Moreover, whereas secure people were
more inclined to forgive their spouse on days when they perceived more positive spousal
behavior, insecure people reported little forgiveness even on days when they perceived
their spouse to be supportive. In other words, attachment insecurities not only interfered
with forgiveness, they also interfered with the ability of a partner’s positive behavior to
restore relational harmony. In a series of experimental and longitudinal studies, Luchies,
Finkel, Kumashiro, and McNulty (2010) showed that the beneficial effects of forgive-
ness on relational quality and partners’ self-esteem are inhibited by a dyadic source of
insecurities— the extent to which a partner is perceived to be unresponsive. Overall, in
insecurity-arousing circumstances, forgiveness has negative effects, which may explain
why dispositionally insecure people are often reluctant to forgive an offending partner.
Attachment insecurities also tend to shape psychological reactions to real or imag-
ined cases of a partner’s infidelity. Attachment-anxious people tend to report higher levels
of jealousy, suspicion, and worries about relationship exclusivity; experience high levels
of fear, guilt, shame, sadness, inferiority, and anger; and cope by expressing strong disap-
proval and engaging in intense surveillance (Belus, Wanklyn, Iverson, Langhinrichsen-
Rohling, & Monson, 2014; Buunk, 1997; Collins & Read, 1990; A. L. Costa, Sophia,
Sanches, Tavares, & Zilberman, 2015; Donovan & Emmers-Sommer, 2012; D. G. Dut-
ton et al., 1994; Guerrero, 1998; Hazan & Shaver, 1987; Karakurt, 2012; Knobloch,
Solomon, & Cruz, 2001; Kolb & Owen, 2014; Leak, Gardner, & Parsons, 1998; Maraz-
ziti et al., 2010; Miller, Denes, Diaz, & Buck, 2014; Radecki-Bush, Farrell, & Bush,
1993; Sharpsteen & Kirkpatrick, 1997; Strawhun et al., 2013; C. D. C. Wang, King, &
Debernardi, 2012). Anxious people’s heightened jealousy has also been documented in
the analysis of Facebook responses to jealousy-inducing messages or posts (Fleuriet, Cole,
& Guerrero, 2014; T. C. Marshall et al., 2013). In addition, there is initial evidence that
jealousy is more destructive to one’s own and a partner’s relationship satisfaction mainly
among more anxiously attached partners (Dandurand & Lafontaine, 2014). Avoidance
is not associated with heightened distress in reaction to a partner’s infidelity, but tends
to interfere with constructive behavioral strategies to address the relational crisis and
restore relationship quality (Leak et al., 1998; Redecki-Bush, Farrell, & Bush, 1993;
C. D. C. Wang, King, et al., 2012). Recently, Selterman and Maier (2013) found that
security priming (as compared to a neutral prime) buffered jealousy in response to a
hypothetical scenario involving a participant’s partner behaving flirtatiously with a rival.
318 INTERPERSONAL MANIFESTATIONS OF ATTACHMENT-SYSTEM FUNCTIONING
319
Avoiding strategies Anx (ns) Avo (+) for W and M
Marchand (2004)a AAS CRBQ Attacking Anx (+) Avo (+) for W and M
Compromising Anx (–) Avo (–) for W; Anx (ns) Avo (ns) for M
Marchand et al. (2004)a AAS CRBQ Attacking Anx (+) Avo (+) for W and M
Compromising Anx (–) Avo (–) for W; Anx (ns) Avo (ns) for M
Heene et al. (2007) AAS CPQ Constructive patterns Anx (ns) Avo (ns for W; Anx (–) Avo (–) for M
Demand–withdrawal Anx (+) Avo (+) for W; Anx (+) Avo (ns) for M
Avoidance Anx (+) Avo (+) for W; Anx (+) Avo (ns) for M
Pearce & Halford (2008)a ECR-R CPQ Destructive patterns Anx (+) Avo (+) for W and M
Dominique & Mollen (2009) ECR-R CPQ Constructive patterns Anx (–) Avo (–)
Destructive patterns Anx (+) Avo (+)
Marchand-Reilly (2009) AAS CRBQ Attacking Anx (+) Avo (+)
Compromising Anx (ns) Avo (ns)
Bucx & Seifert (2010) PAQ CRQ Constructive patterns Security (+)
Givertz & Safford (2011)b ECR-R CPQ Constructive patterns Anx (–) Avo(–)
Destructive patterns Withdrawal Anx (+) Avo (+)
Anx (+) Avo (+)
Marchand-Reilly (2012) AAS-Anx CRBQ Attacking Anx (ns)
Compromising Anx (ns) (continued)
TABLE 10.4. (continued)
Attachment Conflict
Study measures measures Dependent variables Main findings
Castellano et al. (2014)a, b CRI ROCI Compromising/integrating Dominating Secure > Insecure
strategies Obliging strategies Avoiding Secure < Insecure
strategies Secure < Insecure
Secure < Insecure
Karantzas et al. (2014) ASQ CPQ Destructive patterns Anx (+) Avo (ns)
Rholes et al. (2014)a, b ECR CPSS Collaboration Anx (ns) Avo (ns)
Dominating Anx (+) Avo (+)
Avoiding Anx (+) Avo (+)
Verbal agression Anx (+) Avo (+)
Studies using observational measures of behaviors and feelings during conflict resolution discussions
Kobak & Hazan (1991)a Q-Sort IDCS Rejection behaviors Security (–) for W and M
Support validation Security (ns) for W; Security (+) for M
Cohn, Silver, et al. (1992)a AAI SC Positive conflict behaviors Security (ns) for W; Security (+) for M
Conflict escalation Security (ns) for W; Security (–) for M
Simpson et al. (1996) AAQ SC Observer-coded distress Observer-coded Anx (+) Avo (ns) for W and M
warmth Quality of discussion Anx (ns) Avo (ns) for W; Anx (ns) Avo (–) for M
320
Anx (–) Avo (ns) for W; Anx (–) Avo (–) for M
J. Feeney (1998) ASQ SC Observer-coded distress Observer-coded Anx (ns) Avo (+) for W
approach Quality of discussion Anx (–) Avo (ns) for W; Anx (ns) Avo (–) for M
Anx (ns) Avo (–) for W; Anx (–) Avo (–) for M
Paley et al. (1999)a AAI IDCS Positive affect Secure > Anxious for W; Ns differences for M
Withdrawal behavior Secure < Avoidant for W; Ns differences for M
Babcock et al. (2000)a AAI (only SPAFF Domineering, defensiveness Secure < Avoidant, Anxious
husbands) Stonewalling, contempt Secure < Avoidant
Roisman et al. (2001) AAI SC Positive conflict behaviors Security (+) for W and M
Negative conflict behaviors Security (–) for W and M
Bouthillier et al. (2002)a AAI, AAQ IDCS Support validation Secure > avoidant, anxious for W and M
Positive communication Withdrawal Ns differences for W; secure > avoidant, anxious for M
Negative conflict behaviors Ns differences for W; secure < avoidant, anxious for M
Ns differences for W and M
Creasey (2002a) AAI SPAFF Positive conflict behaviors Secure > anxious, avoidant for W; Ns differences for M
Negative conflict behaviors Secure < anxious, avoidant for W and M
Crowell et al. (2002)a AAI RMICS Positive conflict behaviors Security (+) for W; Security (ns) for M
Negative conflict behaviors Security (–) for W; Security (ns) for M
Roisman et al. (2002) AAI SC Positive conflict behaviors Security (+) for W and M
K. S. Wampler et al. (2003)a AAI Georgia Negative affect Security (–) for W and M
Marriage Respect, negotiation Security (+) for W and M
Q-sort Avoidance Security (–) for W; Security (ns) for M
Open communication Security (+) for W and M
Creasey & Ladd (2004) AAI SPAFF Negative conflict behaviors Secure < avoidant (controlling for gender)
Alexandrov et al. (2005)a Couple CPSSRS Conflict escalation Security (–) for W and M
Attachment Positive emotion Security (ns) for W and M
Interview Negative emotion Security (–) for W and M
Campbell et al. (2005) AAQ SC Observer-coded distress Conflict escalation Anx (+) Avo (ns) for W and M
behavior Anx (+) Avo (ns) for W and M
Creasey & Ladd (2005) AAI SPAFF Negative conflict behaviors Domineering, Secure < avoidant, anxious for W and M
Defensiveness Contempt Secure < avoidant, anxious for W and M
Ns differences for W and M
Roisman et al. (2005) AAI, CRI SC Positive conflict behaviors Security in the AAI, CRI (+)
Powers et al. (2006) ECR Self-report Distress Anx (+) Avo (ns) for W and M
Simpson et al. (2007) AAI SC Distress Security (ns) for W and M
Emotional Support Provision Security (+) for M, Security (ns) for W
Pearce & Halford (2008)a ECR-R SC Negative conflict behaviors Anx (+) Avo (+) for W and M
Gouin et al. (2009)a ECR SC Positive emotions Anx (ns) Avo (–)
Negative emotions Anx (ns) Avo (+)
Mehta et al. (2009)a CAI SC Positive emotions Secure > insecure
321
Tran & Simpson (2009) ECR SC Positive emotions Anx (–) Avo (–) for W and M
Negative emotions Anx (+) Avo (+) for W and M
Constructive responses Anx (–) Avo (–) for W, Anx(–) Avo (ns) for M
Destructive responses Anx (ns) Avo (–) for W, Anx(ns) Avo (ns) for M
Barry & Lawrence (2013)a, b RSQ SC Withdrawal Avo (+)
Ben-Naim et al. (2013) ECR SPAFF Positive emotions Anx (ns) Avo (–)
Sadness Anx (+) Avo (–)
Disgust Anx (+) Avo (ns)
Overall et al. (2013) AAQ SC Withdrawal Avo (+)
Anger Avo (+)
Winterheld et al. (2013) AAQ SC Affiliative humor Anx (–) Avo (–)
Aggressive humor Anx (ns) Avo (–)
Self-defeating humor Anx (+) Avo (ns)
Distress Anx (+) Avo (+)
Creasey, 2014a AAI SPAFF Negative Conflict Behavior Secure < insecure for M
Hertz et al. 2015 ECR PANAS Negative emotions Anx (+) Avo (+)
Note. a married couples; blongitudinal design; M, men; W, women; (–) significant inverse correlation; (+) significant positive correlation; (ns) nonsignificant effects; CPQ, Communica-
tion Patterns Questionnaire; CPSS, Conflict and Problem-Solving Scales; CPSSRS, Couple Problem-Solving Style Rating System; CRBQ, Conflict Resolution Behavior Questionnaire;
CRQ, Conflict in Relationships Questionnaire; HS, Hazan & Shaver; IDCS, International Dimensions Coding System; MCI, Margolin Conflict Inventory; RMICS, Revised Marital
Interaction Coding System; ROCI, Rahim Organizational Conflict Inventory; SC, Specific Coding; SPAFF, Specific Affect Coding System.
322 INTERPERSONAL MANIFESTATIONS OF ATTACHMENT-SYSTEM FUNCTIONING
People differing in attachment orientations also differ in the use and appraisal of
humor during conflict discussions with a romantic partner. In a study of dating couples
that were videotaped while trying to resolve a relationship conflict, Winterheld et al.
(2013) found that more avoidant individuals used more aggressive humor and less affili-
ative humor during their discussions, which in turn increased a partner’s distress during
and following the conversation. In contrast, more anxiously attached individuals used
more self-defeating humor, which also elicited negative responses from highly distressed
partners. In addition, insecure individuals, either anxious or avoidant, reacted less favor-
ably when they were the recipients of the humor styles they used most often. That is,
attachment insecurities are associated with humor styles that escalate conflict and dis-
tress rather than buffer them.
In a longitudinal study of 78 individuals who have been studied intensively from
infancy into their mid-20s, Simpson, Collins, et al. (2007) found evidence for a double-
mediation model explaining associations between attachment insecurities in the infant
Strange Situation and actual expressions of negative affect during conflict discussions
with a romantic partner 20 years later (rated by observers). Specifically, participants
classified as securely attached at 12 months of age were rated by their teachers as more
socially competent during early elementary school, and this social competence forecasted
their having more secure relationships with close friends at age 16, which in turn pre-
dicted less negative affect in conflict discussions with their eventual romantic partners.
That is, a trajectory of security from infancy to young adulthood underlies more adaptive
methods of conflict resolution. Using the same sample, Salvatore, Kuo, Steele, Simpson,
and Collins (2011) found that attachment insecurity (primarily avoidance) in infancy
predicted poorer conflict recovery during a 4-minute “cool-down” task that followed a
discussion of a relationship problem with a romantic partner at age 20–23.
In a 1-year prospective study of married couples, Sullivan, Pasch, Johnson, and
Bradbury (2010) highlighted the positive role that a partner’s responsiveness can play
in conflict management. Married couples were observed as newlyweds and again 1 year
later while engaged in conflict management interactions and support discussions. Initial
higher levels of partner responsiveness during the support discussion predicted 1-year
decreases in negative emotion during conflict conversations. Thus it seems that partner
responsiveness and support are key elements in reducing conflictual interactions with a
romantic partner. This conclusion is reinforced by Simpson, Collins, et al.’s (2007) find-
ings that one partner’s supportiveness during a conflict management interaction is effec-
tive in decreasing the contending partner’s distress and anxiety during the interaction.
Overall, Simpson, and Struthers (2013) provided more direct evidence on the buffer-
ing role of a partner’s responsiveness while assessing avoidant people’s reactions during
discussions with a romantic partner who was attempting to influence their behavior. As
predicted, avoidant people felt greater anger and displayed more coder-rated withdrawal
during the discussion, which resulted in less successful problem resolution. However,
when a partner displayed greater responsiveness (e.g., sensitivity to the other partner’s
autonomy needs, validation of his/her viewpoint), avoidant people exhibited less anger
and withdrawal, and their discussions were more successful. Salvatore et al. (2011) found
a similar buffering effect on conflict recovery. Insecure participants (rated as insecure
children 20 years earlier in the Strange Situation) were more likely to “reengage” the con-
flict during a cool-down task. However, when their partners displayed greater respon-
siveness (e.g., calmness, acceptance, affection), less secure participants expressed more
positive emotions during the cool-down task, and these couples were more likely still to
be dating 2 years later.
Attachment Processes and Couple Functioning 323
Barry and Lawrence (2013) have also observed the impact of a partner’s responses
during conflict-laden interactions on avoidant people’s withdrawal strategies. In their
study, more avoidant people were rated by independent observers as emotionally and
behaviorally disengaged during a conflict resolution discussion with their spouse mainly
when the spouse was rated as overtly expressing heightened distress and anger during
the interaction. The avoidance–disengagement link disappeared when a spouse did not
appear to be emotionally aroused during the discussion. In this case, avoidant people’s
defenses are less necessary and they can remain more calmly involved in a discussion.
There is evidence linking attachment insecurities to heightened physiological reac-
tivity during and after conflict discussions. For example, Gouin et al. (2009) found that
avoidant attachment was associated with immune dysregulation (increased production of
interleukin-6) after relational disagreement with a spouse. Powers, Pietromonaco, Gun-
licks, and Sayer (2006) and Laurent and Powers (2007) found that higher ECR anxiety
or avoidance scores were associated with higher levels of salivary cortisol (another index
of physiological arousal) after a conflict discussion with a dating partner. Hertz, Laurent,
and Laurent (2015) also found that ECR avoidance scores were associated with higher
levels of salivary cortisol following a conflict discussion. However, Brooks, Robles, and
Dunkel Schetter (2011) failed to replicate these findings and found that only women
with highly avoidant partners showed increased cortisol responses after a conflict discus-
sion. Laurent and Powers (2007) also found evidence for this avoidant partner effect,
noting that it was strongest when both partners scored relatively high on avoidance. In
a couple-level analysis of cortisol responses during conflictual marital interactions, L.
A. Beck, Pietromonaco, DeBuse, Powers, and Sayer (2013) found that couples in which
anxious wives were paired with avoidant husbands had the strongest cortisol responses.
Moreover, anxious wives had difficulty recognizing their husbands’ distress during the
conflict, and avoidant husbands were less likely to seek support from their wives.
Researchers assessing electrodermal (perspiration) responses during conflict discus-
sions with a romantic partner found that avoidance (assessed with the AAI) was associ-
ated with heightened responses in both dating couples (Holland & Roisman, 2010) and
married couples (Roisman, 2007). Interestingly, Roisman (2007) also found that par-
ticipants who became emotionally overwrought while discussing their early experiences
during the AAI (i.e., more anxiously attached adults) exhibited increased heart rate while
discussing a relationship conflict with their spouse. In a diary study of naturally occur-
ring couple conflict, Hicks and Diamond (2011) found that ECR anxiety scores predicted
more sleep disruptions and higher levels of cortisol following days of heightened quar-
reling.
In a recent study, Ben-Naim, Hirschberger, Ein-Dor, and Mikulincer (2013) recorded
physiological responses during a conflict discussion between dating partners while one
of the them was asked to refrain from expressing emotions (affect suppression), to think
about the positive aspects of the relationship during the interaction (positive mindset), or
to behave naturally. Both participants’ attachment anxiety (assessed with the ECR) pre-
dicted higher cardiovascular responses during the discussion. Interestingly, Ben-Naim et
al. (2013) also found that affective suppression instructions increased the cardiovascular
arousal of the nonmanipulated partner, especially if he or she had scored high on attach-
ment anxiety. That is, anxious people responded with elevated signs of distress during a
conflict discussion mainly when their partner attempted to suppress emotions (presum-
ably a sign of the partner’s emotional withdrawal or bottling up of feelings).
Studies have also found that attachment anxiety is associated with intensification
of the negative consequences of conflict discussions. For example, Simpson et al. (1996)
324 INTERPERSONAL MANIFESTATIONS OF ATTACHMENT-SYSTEM FUNCTIONING
found that more anxiously attached people reported a stronger decline in love and com-
mitment after discussing a major relationship problem with a dating partner. Gallo and
Smith (2001) also found that more anxious wives reacted to a conflict discussion with
more negative appraisals of their husbands. These findings were conceptually replicated
and extended in Campbell et al.’s (2005) diary study of daily conflicts among dating
partners. More anxious participants reported more conflictual interactions across 14
consecutive days. They also reacted to days of intense conflict with a sharper decline in
relationship satisfaction and a more pessimistic view of the relationship’s future. Inter-
estingly, although their partner did not report a corresponding heightening of relational
worries, anxious people thought their partner was less satisfied and more distressed on
days of intense rather than mild conflict.
In a correlational study, Fishtein, Pietromonaco, and Feldman-Barrett (1999) found
that more attachment-anxious people construed their romantic relationships in more
complex terms as the level of conflict increased. According to Fishtein et al. (1999), this
implies that anxiously attached individuals are attuned to the positive sides of relational
conflicts, because conflictual interactions provide an opportunity to elicit a partner’s
attention and responsiveness. This finding might also be explained as resulting from anx-
ious people’s tendency to ruminate about conflicts, which creates more associative links
between different mental representations of the relationship and increases the cognitive
complexity of the overall relationship schema.
Another process with positive implications for relationship satisfaction is what Gable
and her colleagues call relationship capitalization, “the process of informing another
person about the occurrence of a personal positive event and thereby deriving additional
benefit from it” (Gable, Reis, Impett, & Asher, 2004, p. 228). In a series of correlational,
experimental, and diary studies, Sofer-Roth (2008) found that one partner’s attachment
insecurities reduced their capitalization of positive events disclosed by the other partner;
that is, more anxious or avoidant participants were less likely to express happiness and
joy in response to a partner’s disclosure of a positive event and more likely to express envy
and criticism. This, in turn, reduced the partner’s positive mood and relationship satisfac-
tion. In another study of videotaped interactions, Shallcross, Howland, Bemis, Simpson,
and Frazier (2011) found that less securely attached participants were rated by external
observers as less responsive to a partner’s disclosure of a positive event in his or her life.
They also found that this failure to “capitalize” was especially notable when one partner
was avoidant and the other was anxious.
Gable, Gonzaga, and Strachman (2006) found that the capitalization process is
damaged by the disclosing partner’s attachment insecurities. Specifically, when asked
to share with their partner a personal positive event, avoidant people were less likely to
perceive that their partners were responsive to their disclosures. It’s possible that in order
to protect themselves from the heightened intimacy created by a capitalization interac-
tion, avoidant disclosers discount or downplay their partner’s responsiveness. Avoidant
people’s perception of partner unresponsiveness to their disclosure of positive events was
replicated in a 10-day diary study of romantic couples’ daily capitalization responses
(Gosnell & Gable, 2013). This study also provided evidence for the overdependence of
anxious people on a partner’s daily responses: Fluctuations in daily relationship satisfac-
tion as a function of partner responsiveness to disclosure of positive events were greater
among more anxious people.
Family Dynamics
With the transition to parenthood, a couple usually becomes transformed into a fam-
ily system. In such a system, family members participate in different kinds of relation-
ships (marital and parent–child relationships) and many tasks and interactions have to be
planned, coordinated, and integrated to facilitate the smooth functioning of the family.
Using the AAI, Dickstein and colleagues (Dickstein et al., 2001; Dickstein et al., 2009;
Dickstein, Seifer, Albus, & Magee, 2004) found that spouses classified as insecure scored
lower than secure spouses on self-report measures and clinical ratings of family function-
ing. Moreover, external observers rated insecure spouses as lower in family functioning
than secure spouses during a family mealtime interaction with their 14-month old infant.
Paley et al. (2005) found that the quality of family interaction (when mother, father,
and a 2-year-old child were videotaped during a 15-minute play task) was dramatically
impaired by fathers’ attachment insecurities, particularly when there were high levels of
marital discord.
Using self-report attachment scales, Diehl et al. (1998) found that insecure spouses
reported a less positive family climate than secure spouses, and four other studies
(Baiocco, Cacioppo, Laghi, & Tafà, 2013; Finzi-Dottan, Cohen, Iwaniec, Sapir, & Weiz-
man, 2003; Mikulincer & Florian, 1999b; Pfaller, Kiselica, & Gerstein, 1998) found that
less secure spouses scored lower on two dimensions of family dynamics: family cohesion
(the extent of emotional bonding between family members) and family adaptability (the
extent to which a family is able to adjust its rules in response to changes). In addition,
326 INTERPERSONAL MANIFESTATIONS OF ATTACHMENT-SYSTEM FUNCTIONING
Kor et al. (2012) found that self-reports of avoidance and anxiety were related to greater
emotional distance among family members and greater family chaos (i.e., lower organiza-
tion and control), and Kohn, Rholes, et al. (2012) found that higher attachment anxiety
was related to perceiving family responsibilities as overwhelming.
Leon and Jacobvitz (2003) found that mothers’ attachment insecurity (assessed with
the AAI prior to the first child’s birth) was associated with reports of less flexible family
rituals (e.g., family dinners, birthday celebrations) when the child was 7 years old. More-
over, couples in which both spouses were insecure attributed less meaning and impor-
tance to family rituals. Similarly, Crespo, Davide, Costa, and Fletcher (2008) found that
avoidant attachment (assessed with the ECR) was associated with lower engagement in
family rituals and that this disengagement partially explained avoidant people’s lack of
relationship satisfaction.
Summary
Our literature review confirms that attachment insecurities interfere with the forma-
tion, consolidation, and maintenance of lasting and satisfying couple relationships. For
example, avoidant defenses interfere with comfortable flirtation and progression of a
relationship toward intimacy, commitment, and productive conflict resolution. Attach-
ment anxiety engenders tension and ambivalence, suspicion and intrusiveness, and con-
flict escalation. Moreover, as we show in subsequent chapters, attachment insecurity
impairs one’s ability to provide support to a partner who is undergoing stress (Chapter
11), interferes with parenting tasks (Chapter 11) and a satisfying sex life (Chapter 12),
and puts people at risk for emotional problems (Chapter 13), all of which can obviously
damage and derail a couple relationship over the long haul.
Relationship Satisfaction
“Satisfaction” refers to having one’s needs met, and within long-term couple relationships,
the needs have to do with wishes for love, intimacy, affection, acceptance, understand-
ing, support, and security as well as more individualistic wishes for autonomy, growth,
and competence. In terms of attachment theory, relationship satisfaction can be expected
to increase as partners become available and reliable sources of closeness and intimacy,
effective providers of support and security (safe havens), and secure bases from which
they can engage in autonomous growth-oriented activities (J. Feeney, 1999c; Mikulincer,
Florian, Cowan, & Cowan, 2002).
Attachment theory also suggests that relationship dissatisfaction arises from attach-
ment-related worries and insecurities (Kobak, Ruckdeschel, & Hazan, 1994). For exam-
ple, attachment injuries (experiencing a partner’s unavailability, infidelity, abuse, or
rejection) can cause especially strong relationship distress whenever they activate defen-
sive patterns of attachment-system hyperactivation or deactivation, which foster either
Attachment Processes and Couple Functioning 327
angry, clingy demands for a partner’s attention or cold detachment from a disappointing
or frustrating partner. Moreover, these injuries can also interfere with the restoration
of relational harmony if they are incorporated into insecure working models that gener-
ate pessimistic expectations about the partner and the relationship (e.g., S. M. Johnson,
2011). Because every relationship is likely to include partner transgressions, misunder-
standings, and at least minor attachment injuries from time to time, dispositional attach-
ment insecurities put couples at risk for more lasting relationship dissatisfaction.
Relationship satisfaction has been examined in studies of dating and married couples
(see Tables 10.5 and 10.6 for summaries). Supporting an attachment-theoretical analysis,
less secure people—whether anxious, avoidant, or both—generally report lower satisfac-
tion with their dating and marital relationships. This attachment–satisfaction link has
been found in homosexual as well as heterosexual couples (see Tables 10.5 and 10.6).
Some of these studies have considered potentially confounded variables and found that
the attachment–satisfaction link cannot be explained by other personality factors, such
as the “Big Five” traits, depression, self-esteem, or sex-role orientation (e.g., Carnelley,
Pietromonaco, & Jaffe, 1994; Jones & Cunningham, 1996; Noftle & Shaver, 2006;
Shaver & Brennan, 1992; Whisman & Allan, 1996), increasing our confidence in the
uniqueness of the contribution of attachment-related variables to relationship satisfac-
tion.
It is worth noting that the vast majority of studies based on self-report assessments
of attachment style have found that insecure spouses have lower marital satisfaction than
secure ones. In contrast, more than half of the studies based on the AAI have not found
a significant link between attachment insecurities and marital dissatisfaction. Although
method variance (self-report vs. interview) might account for some of the discrepancies,
some of these studies have used interview methods (e.g., the CRI) to assess attachment
orientations in marriage and found higher marital satisfaction among secure than among
insecure spouses (e.g., Alexandrov et al., 2005; Roisman et al., 2005; Treboux et al.,
2004). This implies that current attachment insecurities related specifically to adult close
relationships (whether assessed by self-report scales or interviews) are better predictors
of marital dissatisfaction than “state of mind with respect to attachment” to parents and
other childhood caregivers.
Tables 10.5 and 10.6 provide answers to two important questions frequently asked
in the attachment literature. The first concerns the possibility that women’s relationship
satisfaction is more influenced than men’s satisfaction by attachment insecurities, because
of women’s greater investment in relationships and their interdependent self-construals.
However, as can be seen in the tables, there is no consistent gender difference along these
lines. In fact, the vast majority of studies have yielded a significant attachment–satisfac-
tion link in both women and men. This conclusion was supported by a meta-analysis of
118 independent samples (N = 21,602) conducted by T. Li and Chan (2012), who found
that both anxiety and avoidance were detrimental to relationship satisfaction and that
the moderating effect of gender was weak.
The second question concerns differences between anxious and avoidant individu-
als (or between correlations involving anxious and avoidant attachment dimensions).
Although most studies suggest that both kinds of insecurity are associated with relation-
ship dissatisfaction (see Tables 10.5 and 10.6), detailed examination of the results within
genders reveals differential effects of anxiety and avoidance. Whereas anxiety and avoid-
ance are equally predictive of women’s dissatisfaction, avoidance appears to be more
consistently associated than anxiety with relationship dissatisfaction in men. It’s possible
that the male role’s emphasis on independence and emotional control exacerbates or is
328 INTERPERSONAL MANIFESTATIONS OF ATTACHMENT-SYSTEM FUNCTIONING
participants to be more reactive to daily fluctuations in their partner’s positive and nega-
tive behaviors. That is, compared to secure spouses, anxious ones reacted with greater
daily ups and downs in satisfaction depending on the extent to which their partner was
available and supportive. This result may mean that anxious partners’ doubts about their
lovability and their partner’s love cause them to assign greater significance to daily events
that signal approval or rejection. According to Campbell et al. (2005), this “myopic”
here-and-now focus on daily relationship events helps to explain anxious individuals’
dissatisfaction, because evaluating a relationship based on daily fluctuations exacerbates
doubts and insecurities (today’s available partner can be tomorrow’s rejecting partner)
and intensifies a maladaptive cycle of clinging behavior and relationship distress.
334 INTERPERSONAL MANIFESTATIONS OF ATTACHMENT-SYSTEM FUNCTIONING
Some investigators have examined possible mediators of the link between attach-
ment insecurities and relationship dissatisfaction (see Table 10.7). Findings indicate that
maladaptive ways of coping, negative beliefs about relationship partners, and problems
in conflict management underlie the heightened relationship dissatisfaction of anxious
and avoidant people. Moreover, whereas negative affectivity also contributes to anx-
ious people’s relationship dissatisfaction, lack of nurturance and deficits in interpersonal
expressivity and sensitivity are additional mediators of avoidant people’s dissatisfaction.
The issue of mediation is complex, and research is not likely to identify just one or two
isolated mediators. Nevertheless, exploring the mechanisms by which attachment inse-
curities play out in couple relationships is important for both theoretical and clinical
reasons.
Although the attachment–satisfaction link is a strong and pervasive finding, there
is evidence that it can be buffered or intensified by other factors. For example, a recent
meta-analysis of 57 cross-sectional studies (N = 14,340) conducted by Hadden, Smith,
and Webster (2014) showed that relationship duration moderated the association between
attachment insecurities and relationship dissatisfaction; this association was stronger as
relationship duration increased. Other studies indicate that the avoidance–dissatisfaction
link was buffered by religious faith (J. L. Lopez, Riggs, Pollard, & Hook, 2011), living in
an individualist culture (the United States rather than Hong Kong; Friedman et al., 2010),
and among couples in which both partners were perceived as contributing to housework
chores (Badr & Acitelli, 2009). There is also evidence (see Chapter 12) that having sex
can buffer the link between attachment anxiety and daily relational satisfaction (e.g.,
Little et al., 2010).
In a prospective study of relationship satisfaction trajectories across the first 2 years
of parenthood (6 weeks before the birth of first child, and then at 6, 12, 18, and 24
months postpartum), Kohn, Rholes, et al. (2012) found that the increasing dissatisfac-
tion of attachment-anxious people over time was buffered when they perceived their
partners as more supportive and as behaving more positively toward them. In addition,
the increasing dissatisfaction of avoidant people over time was buffered when they per-
ceived less work–family conflict and fewer demands from their partner. These findings
suggest that insecurities predict dissatisfaction on the part of new parents primarily when
parenting-related stressors block the pursuit of important goals (support availability for
anxious people, independence for avoidant people).
Up to this point, we have focused mainly on the detrimental effects of attachment
insecurities on relationship satisfaction. However, there is also evidence that attachment
security can protect relationship quality during life transitions and stressful periods. For
example, Amir et al. (1999) found that attachment security buffered the detrimental
effects of prolonged infertility on marital satisfaction. Similar effects have also been
noted during the transition to parenthood. Rholes et al. (2001) and Simpson and Rholes
(2002a) reported that wives’ negative prenatal perceptions of spousal support (assessed
6 weeks before the birth of a first child) predicted decreases in marital satisfaction 6
months later, but this detrimental effect of insufficient spousal support occurred mainly
among anxious women. Paley, Cox, Harter, and Margand (2002) found that couples who
exhibited escalation of negative emotions (one spouse’s negative emotion being followed
by an increase in the other spouse’s negative emotion) while discussing a marital problem
prior to the birth of their first child also experienced decreases in affection and intimacy
3 and 12 months after the birth. But again, the decrease occurred mainly among insecure
husbands and their wives.
Attachment Processes and Couple Functioning 335
Relationship Stability
Even if attachment insecurity is conducive to relationship dissatisfaction, does it also
contribute over time to breakups and divorces? The answer appears to be yes. In the
very first study of romantic attachment, Hazan and Shaver (1987) found that people
who described themselves as having an avoidant or anxious attachment style had shorter
relationships than secure people, and were more likely to report having been divorced.
Subsequent studies have also found that insecure people have briefer relationships and are
more likely to divorce (e.g., Birnbaum et al., 1997; R. W. Doherty et al., 1994; J. Feeney
& Noller, 1990, 1992; E. M. Hill et al., 1994). In a study of individuals who entered
another marriage following a divorce, the avoidant ones were more likely to divorce again
(Ceglian & Gardner, 1999).
Beyond these cross-sectional studies, several prospective longitudinal studies (with
the exception of Whisman & Allan, 1996) have indicated that attachment insecurity
predicts subsequent breakups of dating relationships (Crowell & Treboux, 2001; Duem-
mler & Kobak, 2001; Kirkpatrick & Davis, 1994; Kirkpatrick & Hazan, 1994; Shaver
& Brennan, 1992). In a 3-year longitudinal study, Kirkpatrick and Davis (1994) found
interesting gender differences in this respect: Couples that included an avoidant woman
and/or an anxious man were highly predisposed to break up within the 3-year study
period. In contrast, the pairing of an anxious woman with an avoidant man, even though
both partners reported relatively high levels of relationship distress, was resistant to dis-
solution. Unfortunately, this intriguing pattern of gender effects has not been replicated
in other prospective studies.
In a 4-year follow-up of Hazan and Shaver’s (1987) original newspaper survey study,
Kirkpatrick and Hazan (1994) found theoretically sensible differences between anxious
and avoidant people in their patterns of relationship stability or instability. Findings for
anxious respondents seemed to reflect both their frustrating search for love and their
reluctance to leave an unhappy relationship. People with an anxious attachment style
at the time of the initial survey were most likely to have broken up at least once since
completing the initial questionnaire, but they were also just as likely as secure people to
be with the same romantic partner with whom they had been involved 4 years before.
According to Kirkpatrick and Hazan (1994), this pattern of results implies that anxious
people tend to break up and then get back together with the same person, perhaps mul-
tiple times. D. Davis et al. (2003) also found that more anxious respondents were more
likely to feel sexually attracted to their former partner, and were more likely to become
re-involved with the person through sexual activities.
Attachment-anxious people’s reluctance to exit from unhappy relationships was fur-
ther documented by Davila and Bradbury (2001), who followed newlywed couples for 4
Attachment Processes and Couple Functioning 337
years and examined whether attachment insecurities within the first 6 months of mar-
riage predicted staying in unsatisfactory marriages. Although both avoidance and anxi-
ety predicted marital dissatisfaction over time, only attachment anxiety put spouses at
risk for staying in unhappy marriages. It thus seems that anxious adults’ overdependence
on relationship partners, fears of abandonment and separation, and doubts about their
ability to handle life challenges alone lead them to maintain relationships at all costs,
even if it implies staying in an unhappy marriage.
Kirkpatrick and Hazan’s (1994) findings for avoidant individuals suggested, in con-
trast, that these people were ready to exit a close relationship as soon as they experienced
relationship distress. They were the most likely of the three attachment “types” (assessed
in that study with a categorical self-report measure) to report no longer being involved
with the partner they had been with 4 years earlier. Moreover, they were most likely to
say they were not currently seeing anyone and not looking for a partner or were going out
with more than one partner (see Chapter 9 for related findings on loneliness and social
withdrawal).
Klohnen and Bera (1998) reported conceptually similar findings among women who
participated in their 31-year longitudinal study (see details in Chapter 5). Women who
described themselves as securely attached at age 52 were more likely to be married at age
52 than avoidant women. In addition, secure midlife women had reported higher com-
mitment to getting married and starting a family at age 21 than avoidant women, and
this early difference in commitment seemed to have been borne out in behavior 6 years
later at age 27, when secure women were more likely to be married and report fewer
marital tensions. Although we have to keep in mind that variations in relationship trajec-
tory might have affected women’s self-reported attachment style at age 52 (see Chapter
5), Klohnen and Bera (1998) conducted several analyses of other kinds of self-report and
observational measures showing that the avoidant women at age 52 would probably have
classified themselves as avoidant earlier in adulthood as well.
The instability of insecure people’s couple relationships is further highlighted by
findings of associations between attachment insecurities and being sexually unfaithful
to a partner (E. S. Allen & Baucom, 2004; Bogaert & Sadava, 2002; J. N. Fish, Pavkov,
Wetchler, & Bercik, 2012). In eight samples of dating couples, DeWall et al. (2011)
reported that more avoidant people had more permissive attitudes toward infidelity,
devoted more attention to attractive alternative partners, expressed greater daily inter-
est in meeting alternatives to their current relationship partner, perceived alternatives to
their current relationship partner more positively, and engaged in more infidelity over
time. In a prospective study of infidelity in two newlywed samples, Russell, Baker, and
McNulty (2013) found that attachment anxiety predicted more marital infidelity mainly
when the spouse was also relatively high on attachment anxiety. These effects emerged
controlling for marital satisfaction, sexual frequency, and personality.
Relationship Violence
Attachment insecurity contributes to abuse and violence within couple relationships, the
most painful and dangerous kind of relationship maladjustment. Insecure partners’ defi-
cient conflict management skills incline them to try coercive tactics, insults, and threats
(see Table 10.4), which can escalate conflict and set the stage for aggression and reciprocal
violence. Anxious individuals’ controlling behavior can escalate to coercion and aggres-
sion when a partner is not responsive to bids for proximity and loyalty, or can incite
a partner to behave aggressively in order to free him- or herself from overly insistent,
338 INTERPERSONAL MANIFESTATIONS OF ATTACHMENT-SYSTEM FUNCTIONING
TABLE 10.8. A Summary of Findings Concerning Attachment Orientations and Couple Violence
Attachment Violence
Study measure measure Main findings
Studies that compared attachment scores of violent and nonviolent samples
D. G. Dutton et al. (1994)a RSQ (M) PMWI Sec: V < NV; Anx, Fea: V > NV; Avo:
V = NV
Holtzworth-Munroe et al. AAS (M) CTS Anx: V > NV; Avo: V > NV
(1997, Study 1)a
Holtzworth-Munroe et al. RQ (M) CTS Sec: V < NV; Anx, Fea: V > NV; Avo:
(1997, Study 2)a V = NV
Bookwala & Zdaniuk (1998)a RQ (M, W) CTS Sec: V = NV; Anx, Fea: V > NV; Avo:
V = NV
Tweed & Dutton (1998)a RQ (M) PMWI Sec: V = NV; Anx, Fea: V > NV; Avo:
V = NV
Babcock et al. (2000)a AAI (M) CTS Sec: V < NV
Lawson (2008) AAS (M) CTS Anx: V > NV; Avo: V > NV
Wigman et al. (2008) ECR (M) CTS Anx: V > NV; Avo: V = NV
Buck et al. (2012)a RQ (M) CTS-2 Sec: V < NV
& Weintraub, 2015; Wekerle & Wolfe, 1998; Yarkovsky & Timmons, 2014), and others
finding elevations in avoidance (Bond & Bond, 2004; Kesner & McKenry, 1998; Kui-
jpers, van der Knaap, & Winkel, 2012; Miga, Hare, Allen, & Manning, 2010; Shechory,
2013; Wekerle & Wolfe, 1998). Using the AAI in a sample of battered women, P. C. Alex-
ander (2009) found that women who were unresolved with respect to important losses
or abuse (according to the AAI) were more likely to be multiply victimized in adulthood.
Of course, because of the cross-sectional nature of most of the studies, the findings might
indicate either that attachment insecurity puts people at risk for becoming victims of
partner abuse or that abuse increases attachment insecurity, or both (see Chapter 5 for a
review of the effects of abusive relationships on attachment style). However, Kuijpers et
al. (2012) found that avoidance predicted psychological revictimization 2 months later in
342 INTERPERSONAL MANIFESTATIONS OF ATTACHMENT-SYSTEM FUNCTIONING
a sample of victims of partner violence. Also, given the previously mentioned mutuality of
violence, most of the victims are also perpetrators. Therefore, logically the same variables
predict both perpetration and victimization.
Longitudinal studies indicate that abused women who previously scored higher on
attachment anxiety had more problems in resolving their feelings of separation 6 months
after leaving their romantic partner. For example, they engaged in more frequent sexual
contact and emotional involvement with their old partner after separation (Henderson,
Bartholomew, & Dutton, 1997; see also D. Davis et al., 2003). This finding fits with
Davila and Bradbury’s (2001) conclusion that anxious people are unable or unwilling to
leave unhappy relationships. More important, it suggests that such people may form a
“traumatic bond” with an abusive partner that puts them at risk for further abuse.
Partner Effects
Living with an emotionally detached partner or with a clingy, controlling partner can
easily erode a person’s relationship satisfaction and raise doubts about the wisdom of
maintaining the relationship. It’s therefore reasonable to expect significant effects of one
partner’s attachment insecurities on the other partner’s relationship dissatisfaction. More
than 50 published studies have examined these partner effects (see Table 10.9), and most
have found that partners of insecure people report lower relationship satisfaction than
partners of secure people. However, since most of the studies relied on cross-sectional
designs, the findings might imply that relational dissatisfaction on the part of one couple
member arouses attachment insecurities in the other. Fortunately, several prospective
studies provide support for the path going from one partner’s attachment insecurity to
subsequent decreases in the other partner’s relationship satisfaction (see Table 10.9). For
example, Davila et al. (1999) found that husbands’ attachment anxiety and avoidance
and wives’ avoidance within the first 6 months of marriage predicted decreases in the
other partner’s marital satisfaction over the 3-year study period.
A detailed analysis of Table 10.9 also provides information about a frequently dis-
cussed gender difference in partner attachment effects. Two of the early attachment
studies (Collins & Read, 1990; Simpson, 1990) found that men’s satisfaction was more
adversely affected by women’s anxiety than by women’s avoidance, and that women’s
satisfaction was more detrimentally affected by men’s avoidance than by their anxiety.
Koski and Shaver (1997) explained these findings based on gender differences in needs
for closeness and autonomy. Men’s relationship satisfaction might depend on their satisfy-
ing a need for autonomy, which can be more easily frustrated by overly anxious, clinging
female partners than by avoidant women. In contrast, women’s relationship satisfaction
Attachment Processes and Couple Functioning 343
might be heavily dependent on their desire for closeness, which can be more easily frus-
trated by avoidant than by anxious men.
However, as can be seen in Table 10.9, these gender differences have not been rep-
licated in subsequent studies. In the case of women’s satisfaction, the findings suggest
that men’s anxiety is the most consistent predictor of dissatisfaction. In the case of men’s
satisfaction, there is some evidence that women’s anxiety is also a more consistent predic-
tor than women’s avoidance, but several studies also find significant effects of women’s
avoidance. Overall, then, the best single conclusion we can reach from the reviewed find-
ings is that, although both major kinds of attachment insecurity are associated with
partner dissatisfaction, one partner’s attachment anxiety seems especially likely to erode
the other partner’s satisfaction, regardless of gender. Longitudinal studies (e.g., Robins,
Caspi, & Moffitt, 2002) have suggested that neuroticism (which is partly determined by
genes) has a similar negative effect on relationship satisfaction.
In a series of studies, Lemay and Dudley (2011) illustrated the relational costs of liv-
ing with an insecurely attached partner. In their first experiment, participants reported
on their privately felt global feelings (e.g., care, regard) toward a study partner (a friend,
in most cases), and then, following completion of other questionnaires, some of the par-
ticipants were unexpectedly asked to reveal their feelings toward the study partner. The
results showed that participants disclosed less authentic global sentiments to a less secure
partner. This bias was in the direction of exaggerating the positivity of sentiments that
were privately more negative. In an additional diary study, participants whose romantic
partner was less secure reported more daily vigilance about potential events that could
upset their partner and a heightened tendency to exaggerate affection for the partner.
Importantly, although participants’ affective exaggeration appeared to enhance chroni-
cally insecure partners’ perceptions of being valued, it also predicted participants’ reduced
relationship satisfaction. That is, the efforts of insecure people’s partners to modulate
their mate’s insecurities impaired their relationship satisfaction.
Couple-Type Effects
Few studies have examined couple-type effects on relationship satisfaction. But five stud-
ies (Berman, Marcus, & Berman, 1994; Dickstein et al., 2001; Kilmann et al., 2013;
Senchak & Leonard, 1992; Strauss et al., 2012) found that secure couples (in which
Attachment Processes and Couple Functioning 345
both partners were securely attached) reported greater satisfaction than mixed couples
(in which only one partner was secure) and doubly insecure couples. No difference was
found between mixed and doubly insecure couples, implying that attachment insecurity
in one partner has negative effects on couple satisfaction. This negative effect of one part-
ner’s attachment insecurity has also been noted in studies assessing regulation of close-
ness (J. Feeney, 1999b), dyadic communication (Tucker & Anders, 1998), responses to a
partner’s transgressions (Gaines et al., 1999), conflict management strategies (Bouthillier
et al., 2002), and relational violence (J. B. Wilson, Gardner, Brosi, Topham, & Busby,
2013).
However, Cohn, Silver, Cowan, Cowan, and Pearson (1992), Mikulincer and Florian
(1999c), and Wampler, Shi, Nelson, and Kimball (2003) found no significant difference
between couple types. Moreover, other researchers have concluded that a secure partner
can sometimes buffer the negative effects of an insecure partner on a relationship. Specifi-
cally, as compared with couples in which both partners are insecure, mixed couples in
which one partner is secure report higher marital quality and more intimacy-promoting
behavior (A. Ben-Ari & Lavee, 2005; Volling et al., 1998), higher emotional expressivity
(J. Feeney, 1995a), and calmer emotions and pro-relationship behaviors during a con-
flict discussion (Cohn et al., 1992; Paley, Cox, Burchinal, & Payne, 1999). In addition,
Powers et al. (2006) found that men interacting with a secure partner during a conflict
resolution task exhibited less physiological reactivity (assessed with salivary cortisol),
although women’s cortisol levels did not depend on their partner’s security. Overall, it
is difficult to reach a definitive conclusion because of the small number of studies and
the failure of some studies to include both dimensions of insecurity. As J. Feeney (2003)
noted, avoidant people will create very different relational problems for anxious and
avoidant partners.
Interaction Effects
In the research literature on couple relationships, there is evidence that interactional pro-
cesses resulting from both partners’ behavioral propensities predict relationship adjust-
ment (see Bradbury, Fincham, & Beach, 2000, for a review). In the attachment domain,
although some studies have failed to find interactive effects of partners’ attachment ori-
entations on relationship satisfaction (e.g., Jones & Cunningham, 1996; Kirkpatrick &
Davis, 1994) and on other relational cognitions and behaviors (e.g., Creasey, 2002a;
Mikulincer & Florian, 1999c; Paley et al., 1999), there is substantial evidence that two
combinations of insecure attachment styles interfere with relationship adjustment: (1)
pairing an anxious person with an avoidant one and (2) pairing two anxious people (e.g.,
Allison, Bartholomew, Mayseless, & Dutton, 2005; J. Feeney, 1994; N. Roberts & Nol-
ler; 1998).
Couples in which an anxious person is paired with an avoidant one tend to pro-
duce destructive pursuit–distancing or demand–withdrawal patterns of relating. In such
couples, the anxious partner’s needs and demands frustrate the avoidant partner’s prefer-
ence for distance, and the avoidant partner’s tendency to create distance frustrates the
anxious partner’s intense desire for closeness. As a result, both partners are dissatisfied
and can become abusive or violent when attempting to influence their partner’s undesir-
able behavior.
In a study of the early years of marriage, J. Feeney (1994) found that a wife’s attach-
ment anxiety coupled with a husband’s avoidance reduced both partners’ marital satis-
faction. N. Roberts and Noller (1998) and Allison et al. (2005) extended these findings
346 INTERPERSONAL MANIFESTATIONS OF ATTACHMENT-SYSTEM FUNCTIONING
to relationship violence and noted that the combination of an avoidant individual with
an anxious one intensified the effects of each partner’s attachment style on the perpetra-
tion of violence against the other partner. In addition, Babcock, Jacobson, Gottman, and
Yerington (2000) found that anxiously attached men engage in couple violence mainly
when their partner remains distant during conflicts.
Attachment studies also reveal the destructive effect of pairing two anxious part-
ners: One partner’s anxiety exacerbates the other partner’s anxiety, and the combination
erodes marital satisfaction (J. Feeney, 1994; Gallo & Smith, 2001), amplifies negative
responses to a partner’s distancing (J. Feeney, 2003), and increases couple violence (Alli-
son et al., 2005). J. Feeney (2003) described these anxious/anxious couples as engaging in
“mutual attack-and-retreat,” and Bartholomew and Allison (2006) labeled them “pursu-
ing-pursuing.” In such couples, both partners feel misunderstood and rejected, both are
excessively focused on their own insecurities, and both try to control the other’s behavior.
Overall, the evidence strongly indicates that attachment insecurities put couples at
risk for relationship maladjustment. At the individual level, attachment insecurities are
associated with relationship dissatisfaction and troubled relational behavior, sometimes
including violence. At the dyadic level, there is increasing evidence that partners of inse-
cure people also suffer from relationship distress. Moreover, whereas specific combi-
nations of anxious and avoidant partners within a couple (anxious/avoidant, anxious/
anxious) tend to exacerbate relationship dissatisfaction and violence, the presence of a
relatively secure partner within a relationship sometimes, but not always, buffers the
detrimental effects of the other partner’s insecurity.
Concluding Remarks