According to attachment theory (Bowlby, 1973, 1980, 1982), human beings are born with a psychobiological system (the attachment behavioral system) that motivates them to seek proximity to supportive others (attachment figures) in times of need for the sake of gaining a sense of safety and security. However, although every human being is born with this propensity to seek support and rely on others as a source of protection and security, people differ in the way their attachment system functions, primarily as a result of their history of interactions with attachment figures (Bowlby, 1973). Interactions with figures who are available, sensitive, and responsive in times of need promote effective support-seeking strategies and encourage the development of a stable sense of security. This sense of security includes implicit beliefs that the world is generally safe, that other people are “well-intentioned and kind-hearted” (Hazan & Shaver, 1987), that one is valued, loved, understood, accepted, and cared for by others, and that one can explore the environment with interest and engage rewardingly with other people. These beliefs are associated with and rooted in positive mental representations of self and others, which Bowlby (1973) called internal working models. These models shape a person’s expectations for future interactions with the same or other relationship partners over time, especially in times of need.

Unfortunately, when a person’s attachment figures have not been reliably available, sensitive, and supportive, he or she has not learned that seeking proximity to others relieves distress. Moreover, the person is not likely to have developed a core sense of security. Rather, negative working models of self (as not sufficiently lovable) and others (as unaccepting, unreliable, and unresponsive if not downright abusive or cruel) are formed, and affect-regulation strategies other than confident proximity seeking are developed. These secondary attachment strategies are called “anxious” and “avoidant” in attachment theory. According to the theory, individual differences in working models and distress-regulation strategies eventually become trait-like attachment “styles” or orientations—characteristic patterns of relational expectations, emotions, and behavior (Fraley & Shaver, 2000).

Beginning with Ainsworth, Blehar, Waters, and Wall’s (1978) studies of infant attachment, and then followed by hundreds of developmental studies of children and social psychological studies of adults (reviewed by Mikulincer & Shaver, 2003, 2007a), researchers have found that attachment orientations can be measured along the two roughly orthogonal dimensions already mentioned: attachment-related anxiety and attachment-related avoidance (Brennan, Clark, & Shaver, 1998). A person’s position on the attachment anxiety dimension indicates the degree to which he or she worries that a partner will not be responsive in times of need and is afraid of being rejected or abandoned, partly because of his or her own self-doubts and self-criticism. A person’s position on the avoidance dimension indicates the extent to which he or she distrusts relationship partners’ goodwill and their capacity to help, and the extent to which he or she defensively strives to maintain independence and emotional distance from partners. People who score low on both dimensions are said to be secure with respect to attachment (Brennan et al., 1998). In extreme cases, attachment insecurities can result in a “disorganized” attachment pattern—a mixture of contradictory approach and avoidance tendencies or a form of psychological and behavioral paralysis (Simpson & Rholes, 2002). A person’s location in this two-dimensional space can be measured with reliable and valid self-report scales (e.g., Brennan et al., 1998) and is associated in theoretically predictable ways with many aspects of relationship quality and psychological adjustment (see Mikulincer & Shaver, 2007a, for a review).

Although attachment orientations are initially formed in relationships with primary caregivers (usually parents) during infancy and childhood (Bowlby, 1982), as confirmed by several decades-long longitudinal studies (reviewed in Cassidy & Shaver, 2008), Bowlby (1988) also claimed that meaningful interactions with later attachment figures (e.g., friends, romantic partners) can alter the sense of security and move a person from one region of the two-dimensional space (defined by anxiety by avoidance) to another. Moreover, although attachment orientations are often conceptualized as single, stable, global patterns, they may also be considered in terms of the most accessible working models at a given moment, selected from a complex network of many episodic, context-specific, and relationship-anchored memories and schemas (Mikulincer & Shaver, 2003). Research indicates that a person’s sense of attachment security can change, subtly or dramatically, depending on naturally occurring or experimentally induced contexts (Mikulincer & Shaver, 2007b; Shaver & Mikulincer, 2008), making it possible to study the psychological and relational effects of both experimentally primed senses of security or insecurity and actual partners’ sensitivity and responsiveness.

Individual differences in attachment anxiety and avoidance reflect both a person’s sense of attachment security and the ways in which he or she deals with stress and distress (Mikulincer & Shaver, 2003, 2007a). People who score low on these dimensions are generally secure, hold positive representations of self and others, tend to employ constructive and effective distress management strategies, and generally engage comfortably and rewardingly in intimate relationships and in exploration of many aspects of the physical, social, and cultural environment. In contrast, people who score high on either attachment anxiety or avoidance, or both, suffer from a variety of insecurities, including doubts about their safety and lovability and about other people’s intentions and goodwill. Insecure people tend to use secondary attachment and emotion-regulation strategies that we, following Cassidy and Kobak (1988), call “hyperactivation” or “deactivation” of the attachment behavioral system in an effort to cope with threats.

People who score high on attachment anxiety rely on hyperactivating strategies—that is, energetic attempts to achieve proximity, support, and love combined with a lack of confidence that these resources will be provided, and with feelings of sadness or anger when they are in fact not provided (Cassidy & Kobak, 1988). Hyperactivating strategies include attempts to elicit a partner’s involvement, care, and support through begging, complaining, clinging, and attempting to control a partner (Davis, Shaver, & Vernon, 2003); overdependence on relationship partners as sources of protection (Shaver & Hazan, 1993); and perception of the self as relatively helpless at regulating emotion (Mikulincer & Shaver, 2003). These strategies also include exaggeration of threats, intensification of distress, and rumination on distress-eliciting thoughts as means of upregulating the attachment system and eliciting others’ attention, care, and support (Shaver & Mikulincer, 2002). These reactions stem from relationships in which an attachment figure is sometimes responsive but unreliably so, placing the needy person on a partial reinforcement schedule that rewards exaggeration of threats and persistence in proximity-seeking attempts until an attachment figure is perceived to be adequately available and supportive.

In contrast, people who score high on attachment-related avoidance tend to keep their attachment systems downregulated, to avoid the frustration and pain of being unlovable and rejected. This goal is achieved through deactivating strategies that include avoidance of dependence on close relationship partners and maintenance of emotional distance from others. These tendencies are supplemented by what Bowlby (1982) called compulsive self-reliance, a defensive stance that decreases attachment-system activation by denying threats, failing to acknowledge distress, vulnerabilities, and personal weaknesses, and suppressing distress-eliciting thoughts. These strategies develop in relationships with attachment figures who disapprove of and punish bids for closeness and expressions of need (Ainsworth et al., 1978).

According to attachment theory (Bowlby, 1982), the attachment system evolved because it increased safety (hence survival and the opportunity to reproduce) and taught vulnerable children how to regulate emotions effectively in times of stress and distress. This behavioral system is automatically activated by external threats or internal sources of distress, and when it functions appropriately and successfully, it leads to socially supported emotional stability. Optimal functioning is associated with a relational if-then script, which Waters and Waters (2006) called a secure-base script: “If I encounter an obstacle and/or become distressed, I can approach a significant other for help. He or she is likely to be available and supportive. I will experience relief and comfort as a result of proximity to this person, and I can then return to other activities.” Once activated, this script guides behavior, but even by itself it can mitigate distress, promote optimism and hope, and help a person cope effectively with problems (Mikulincer, Shaver, Sapir-Lavid, & Avihou-Kanza, 2009).

Recently, adult attachment researchers have designed experimental procedures to examine these regulatory properties of the attachment system. For example, in a series of laboratory experiments, Mikulincer, Gillath, and Shaver (2002) showed that mental representations of attachment figures (e.g., names of security-enhancing attachment figures) are automatically activated in a person’s mind when he or she is exposed to threatening stimuli, even unconsciously. Specifically, when a threat-related word (e.g., “death”) was presented very briefly (i.e., subliminally) on a computer screen, participants were faster to detect the name of one of their attachment figures when it appeared on the screen and were slower to name the color in which such names were printed on the screen—an indication that the words had been automatically activated in memory (Mikulincer et al., 2002). In other words, threats, even when arising unconsciously, can automatically activate mental representations of security providers.

In another set of studies of what we call “security priming,” Mikulincer, Hirschberger, Nachmias, and Gillath (2001) showed that activation of representations of security-enhancing attachment figures can automatically infuse a previously neutral stimulus with positive affect. For example, subliminal presentation of the names of people who were nominated by the participants as attachment figures, compared with mere acquaintances or close others who were not nominated as attachment figures, led to greater liking of previously unfamiliar stimuli. Moreover, subliminal exposure to names of attachment figures eliminated the detrimental effects that threats otherwise had on liking for previously neutral stimuli. These effects of security priming on positive affect have been replicated in subsequent studies (see Mikulincer & Shaver, 2007b, for a review). Combining our findings, we conclude that people automatically search for internal representations of security-enhancing attachment figures during times of stress, and mental activation of such representations results in positive affect that can facilitate effective coping and restore emotional equanimity. Moreover, people who generally feel safe and protected benefit from what we, following Fredrickson (2001), call a “broaden and build” cycle of attachment security (Mikulincer & Shaver, 2003), which supports psychological resilience and broadens skills and interests (by virtue of what Bowlby, 1969/1982, called the exploration behavioral system).

According to adult attachment theory (Mikulincer & Shaver, 2003, 2007a; Shaver & Mikulincer, 2002), although secondary attachment strategies (anxious hyperactivation and avoidant deactivation) are initially adaptive in the sense that they adjust a child’s behavior to the requirements of an inconsistently available or consistently distant or unavailable attachment figure, they are maladaptive when used in later relationships in which support-seeking and relational interdependence could be rewarding and could help people to maintain a sense of well-being even in times of stress. These attachment strategies encourage repeated activation of negative working models of self and others that can interfere with social adjustment and mental health.

The early attachment experiences of insecure people (whether anxious, avoidant, or both) are characterized by unstable and inadequate distress regulation (Bowlby, 1973; Shaver & Hazan, 1993), which interferes with the development of inner resources necessary for coping successfully with stressors and maintaining mental health. This impairment is particularly likely to be noticed during prolonged, highly demanding stressful experiences that require active support-seeking and actual confrontation with a problem (Berant, Mikulincer, & Shaver, 2008). In such cases, anxious (hyperactivating) strategies may become extreme, damaging not only a person’s own mental health but that of key relationship partners, and avoidant (deactivating) strategies can collapse, resulting in a marked decline in psychological functioning. These negative outcomes of attachment anxiety and avoidance have been documented in hundreds of cross-sectional and longitudinal studies (see Mikulincer & Shaver, 2007a, for a review).

The mental health implications of attachment insecurities are important for understanding individual differences in the way people react to traumatic events. Traumatic experiences such as rape, assault, car accidents, floods, war, and a host of other natural and man-made disasters disrupt a person’s psychological stability and may place him or her at risk for serious emotional and adjustment problems. In some cases psychological well-being is maintained even during severe stress or is restored shortly after a traumatic experience ends. In other cases, however, attempted restoration of emotional stability fails. Studies of the psychosocial effects of traumatic events have identified posttraumatic stress disorder (PTSD) as the most common and debilitating outcome (see Ball & Stein, 2012, for a review). PTSD is characterized by repeated reexperiencing of a traumatic event (unwanted intrusion of trauma-related material into conscious thoughts, mental images, and dreams), numbing of responsiveness to or reduced involvement with the external world (trauma-related avoidance responses), and a variety of autonomic, affective, and cognitive signs of hyperarousal (American Psychiatric Association, 1994).

As with other stressful experiences, we would expect the attachment behavioral system to be automatically activated when a person is exposed to traumatic events. According to Horowitz (1982), a person’s state of mind when undergoing trauma includes overwhelming shock and intense feelings of panic, vulnerability, helplessness, and exhaustion. Such a state of mind should automatically activate the attachment system, impelling a person to search for external or internalized attachment figures who can protect him or her from trauma. This attachment-system activation is likely to be experienced as an intense cry for help.

When internal or external sources of support can be mobilized during a traumatic event, the traumatized person can maintain psychological well-being despite the external challenge, thereby making PTSD less likely when the trauma subsides. A secure person’s mental cry for help during trauma should result in mobilization of internal representations of security-providing attachment figures and/or actual external sources of support. As a result, the secure person is likely to activate optimistic and hopeful representations of self and others, rely on constructive strategies of affect regulation, deal effectively with the trauma, and restore emotional balance. In other words, the sense of attachment security should act, at least to some extent, as a protective shield against the formation of emotional problems, including PTSD, following trauma.

In contrast, disruptions in the sense of attachment security may prevent maintenance or restoration of emotional equanimity during and following trauma, thereby contributing to PTSD formation. In such cases, a traumatized person may fail to find inner representations of security or external sources of support and comfort, which then interferes with the regulation of distress. This regulatory failure may initiate a cascade of psychological processes, including strong feelings of loneliness and rejection as well as negative working models of self and others, intensification of distress, and reliance on less effective (hyperactivating or deactivating) strategies of affect regulation, which may prevent resolution of the trauma and enhance the likelihood of prolonged PTSD. In other words, an insecure attachment orientation (anxious, avoidant, or both) can predispose a traumatized person to PTSD.

According to Horowitz (1982), the posttraumatic process is defined by two kinds of intrapsychic manifestations of PTSD: intrusion and avoidance. Intrusion refers to unwanted and uncontrollable thoughts, images, emotions, and nightmares related to the traumatic event. Avoidance refers to psychic numbing, denial of the significance and consequences of the traumatic event, and behavioral inhibition. The relative salience of intrusion versus avoidance is not constant (Horowitz, 1982). Intrusion is generally experienced immediately after the trauma, but the two states can alternate during the posttraumatic period until successful “working through” of the trauma is achieved, if in fact it is achieved.

Attachment insecurities are also important in regulating the intensity and frequency of posttraumatic intrusion and avoidance tendencies. Anxious hyperactivation can facilitate reactivation of, and mental rumination about, the traumatic experience and the frustrated cry for help, thereby encouraging intrusive responses. Avoidant deactivation predisposes a traumatized person to deny inner pain and avoid direct or symbolic confrontation with trauma reminders, thereby encouraging posttraumatic avoidance responses. As a result, an insecure attachment orientation (anxious, avoidant, or both) can intensify posttraumatic intrusion and avoidance tendencies. Moreover, contextual activation of mental representations of attachment anxiety or avoidance, due to symbolic or actual encounters with rejecting or unsupportive figures, during the posttraumatic period can further increase the likelihood of intrusive or avoidant responses.

Following this line of reasoning, infusions of attachment security during the posttraumatic period should reduce the intensity and frequency of intrusive and avoidant responses and lead to a successful working through of the trauma. Research has shown that the activation of mental representations of attachment security, due to symbolic or actual encounters with loving, accepting, caring, and supporting others (including a trustworthy and reliable therapist), results in feelings of lovability, hope, optimism, and self-esteem, and can move a person to a more secure location in the two-dimensional anxiety-by-avoidance space, and can activate more constructive and effective ways of managing distress (see Mikulincer & Shaver, 2007b; Shaver & Mikulincer, 2008, for reviews). In this way, activating a sense of security can help people work through trauma, with healing effects on PTSD symptoms.

So far in this chapter we have focused on the potential effects of attachment orientations on traumatic and posttraumatic responses. However, traumatic events and prolonged PTSD, which include mental reactivation of the trauma, can also have important effects on attachment orientations. Although attachment orientations are fairly stable over time (like many core personality traits), they can be altered by powerful experiences that affect a person’s beliefs about the value of seeking help from attachment figures and the feasibility of attaining safety, protection, and comfort (e.g., Baldwin & Fehr, 1995; Davila & Cobb, 2004). Persistent and pervasive PTSD can increase a person’s sense of helplessness and vulnerability and therefore heighten attachment insecurities, especially among individuals who entered a period of trauma with already existing attachment-related doubts and insecurities. The constant mental reactivation of a trauma, particularly a man-made trauma that shatters one’s trust in others’ goodwill and one’s sense of personal value and lovability, can gradually increase the strength of negative working models of self and other, thereby heightening attachment insecurities and reducing the likelihood of attaining a calmer, more secure mental state.

Thus, prolonged and pervasive PTSD may involve a reciprocal, recursive, amplifying cycle of reactivation of the trauma and attachment insecurities: Attachment worries and doubts can prevent successful working through of the trauma, and the resulting mental reactivation of the trauma can further erode the sense of security. Moreover, this gradual but persistent exacerbation of attachment insecurities during the posttraumatic period may lead to the disorganization of the attachment system and the disruption of the regulatory and healing benefits of attachment security. As a result, traumatized people with persistent and pervasive PTSD are likely to score high on both attachment anxiety and avoidance, may be less likely to activate security-related representations during threat exposure, and may be impervious to contextual augmentation of the sense of security. That is, activation of security-related representations may be less effective than might be expected in increasing positive affect and healing the symptoms of PTSD. For severely traumatized people, restoration of a sense of having a safe haven and a secure base may be a long, difficult, and complex process.

Our analysis of the role of attachment orientations in traumatic and posttraumatic processes can be summarized in four hypotheses. First, attachment insecurities will be associated with the formation of PTSD following traumatic events, with attachment anxiety being especially associated with posttraumatic intrusion tendencies and avoidant attachment being especially associated with posttraumatic avoidance tendencies. Second, mobilizing external sources of support and protection or activating internal representations of security during the posttraumatic period will help to reduce PTSD symptoms and facilitate working through the trauma. Third, persistent and pervasive PTSD will increase attachment insecurities over time, thereby contributing to a recursive, amplifying cycle of trauma reactivation and attachment-related worries and doubts. Fourth, persistent and pervasive PTSD will be associated with attachment-system disorganization and a disruption of the regulatory and healing powers of the sense of attachment security. In the following sections, we review empirical evidence for each of these hypotheses.

The first systematic attempt to examine whether and how attachment insecurities are associated with PTSD symptoms focused on reactions of young adults to Iraqi Scud missile attacks on Israel during the 1991 Gulf War (Mikulincer, Florian, & Weller, 1993). In this study, Israeli undergraduates were approached 2 weeks after the end of the Gulf War and asked to complete self-report measures of attachment orientations and PTSD symptoms. Attachment anxiety was associated with more severe posttraumatic intrusion and avoidance symptoms, and avoidant attachment was associated with more severe posttraumatic avoidance responses (assessed with the Impact of Events Scale; Horowitz, Wilner, & Alvarez, 1979).