Secure attachment, designated as the "B" classification in Ainsworth's Strange Situation paradigm (Ainsworth et al., 1978), describes a relational pattern in which individuals are generally comfortable with emotional intimacy, able to depend on others, and confident that their partners will be available and responsive when needed. The concept originates in John Bowlby's foundational work on attachment theory (Bowlby, 1969/1982), which proposed that early interactions with caregivers shape internal working models — cognitive-affective schemas of self and others that guide expectations and behavior in close relationships across the lifespan. Securely attached individuals tend to hold positive working models of both self ("I am worthy of love") and others ("Others are generally reliable and well-intentioned"), a configuration that Bartholomew and Horowitz (1991) formally mapped in their four-category model of adult attachment.
The extension of attachment theory from infancy to adult romantic relationships was pioneered by Hazan and Shaver (1987), who demonstrated that the three infant attachment patterns identified by Ainsworth — secure, anxious-ambivalent, and avoidant — had meaningful parallels in adult love. In their landmark study, adults who endorsed the secure description reported relationships characterized by trust, friendship, and positive emotion, and they were more likely to describe their most important love experience as happy, friendly, and trusting. Subsequent research using dimensional measures (Brennan, Clark, & Shaver, 1998; Fraley, Waller, & Brennan, 2000) has consistently found that secure attachment corresponds to low scores on both the attachment anxiety dimension (worry about abandonment and partner availability) and the attachment avoidance dimension (discomfort with closeness and dependence).
Population estimates suggest that approximately 50 to 60 percent of adults in Western samples report a predominantly secure attachment orientation (Hazan & Shaver, 1987; Mickelson, Kessler, & Shaver, 1997), making it the most commonly reported style. However, this prevalence figure warrants careful interpretation. Cross-cultural research has demonstrated meaningful variation: van IJzendoorn and Sagi-Schwartz (2008) found that while secure attachment is modal across cultures, the distribution of insecure subtypes varies considerably, suggesting that cultural norms around caregiving, independence, and emotional expression shape the behavioral expression of attachment security. The universality of secure attachment as the "normative" or "optimal" pattern remains an active area of scholarly discussion.
It is important to note that contemporary attachment research increasingly favors dimensional models over categorical classifications. Fraley and Waller (1998) demonstrated through taxometric analysis that attachment is better modeled as continuous dimensions rather than discrete categories, meaning that "secure attachment" is not a fixed type but rather a region of low anxiety and low avoidance on a continuous space. Furthermore, the concept of earned security (Roisman et al., 2002) demonstrates that individuals who experienced insecure early attachment can develop secure functioning through corrective relational experiences, psychotherapy, or sustained self-reflection — a finding that challenges purely deterministic readings of attachment theory and underscores its fundamentally dynamic nature.
Common Patterns
Researchers have identified these characteristic patterns in individuals with secure attachment:
- Comfort with emotional closeness and physical intimacy without losing sense of self
- Consistent and reliable communication patterns, including during conflict
- Ability to clearly express needs, boundaries, and feelings in relationships
- Tendency to interpret a partner's ambiguous behaviors charitably rather than catastrophically
- Effective repair after disagreements, including willingness to apologize and reconnect
- Balanced capacity for both autonomy and togetherness within partnerships
- Relative ease with trusting others and being trusted in return
Research Foundation
The research base for secure attachment is among the most robust in developmental and social psychology, spanning more than five decades of empirical investigation across multiple methodological traditions. Beginning with Bowlby's ethological-evolutionary framework (1969/1982) and Ainsworth's observational studies of mother-infant interaction (1978), the secure attachment construct has been validated through behavioral observation, self-report questionnaires, interview-based assessments (the Adult Attachment Interview), physiological and neuroimaging studies, and longitudinal research tracking attachment patterns from infancy into adulthood. However, this extensive body of work is not without methodological tensions and ongoing debates that warrant transparent acknowledgment.
Ainsworth, Blehar, Waters, & Wall (1978)
Identified the secure ("B") classification in infants using the Strange Situation procedure. Securely attached infants used the caregiver as a secure base for exploration, showed distress upon separation, and were effectively soothed upon reunion. Maternal sensitivity — consistent, contingent responsiveness to infant signals — was identified as the primary predictor of secure attachment.
Methodology: Structured laboratory observation (the Strange Situation) with 26 mother-infant dyads in the original Baltimore sample, later replicated across multiple samples. Involved two brief separations and reunions in an unfamiliar room, with infant behavior coded across several dimensions.
Honest Assessment: The Strange Situation was developed and validated primarily with white, middle-class American samples, raising questions about cross-cultural generalizability. Subsequent cross-cultural studies (van IJzendoorn & Kroonenberg, 1988) found that while the secure classification was modal across cultures, within-country variation was often greater than between-country variation, complicating simple universality claims. Additionally, the laboratory setting may not fully capture the range of attachment behaviors observable in naturalistic contexts.
Hazan & Shaver (1987)
Demonstrated that Ainsworth's three infant attachment classifications could be meaningfully applied to adult romantic relationships. Adults who endorsed the secure description reported more positive relationship experiences, greater trust, and longer-lasting relationships. Approximately 56% of respondents classified themselves as secure.
Methodology: Self-report survey published in a local newspaper (Rocky Mountain News) with 620 respondents, plus a university sample of 108 students. Participants selected one of three paragraph-length descriptions that best characterized their experience in romantic relationships.
Honest Assessment: The forced-choice, single-item measure is a significant methodological limitation. Respondents selected from only three prototypical descriptions, which may not capture the complexity or dimensionality of individual attachment orientations. The newspaper sample raises concerns about self-selection bias. Brennan, Clark, and Shaver (1998) and Fraley, Waller, and Brennan (2000) subsequently developed multi-item dimensional measures that more accurately capture individual differences in attachment.
Brennan, Clark, & Shaver (1998)
Identified two fundamental dimensions underlying adult attachment: attachment anxiety (fear of rejection and abandonment) and attachment avoidance (discomfort with closeness and dependence). Secure attachment corresponds to low levels of both dimensions. This dimensional framework replaced earlier categorical models.
Methodology: Factor analysis of 323 items drawn from every published self-report attachment measure available at the time, administered to a large university sample. The resulting Experiences in Close Relationships (ECR) scale contains 36 items measuring the two dimensions.
Honest Assessment: Self-report measures of attachment capture conscious, explicit representations but may miss implicit or unconscious attachment strategies. Roisman et al. (2007) found only modest correspondence between self-report attachment measures (like the ECR) and interview-based measures (like the Adult Attachment Interview), suggesting these methods may tap different levels of the attachment system. Additionally, the ECR was developed primarily with young adult college samples, limiting generalizability to other populations and age groups.
Fraley, Waller, & Brennan (2000)
Provided strong evidence that adult attachment is better conceptualized as dimensional rather than categorical. Taxometric analyses found no evidence for discrete attachment types, supporting a model in which attachment security exists on a continuum rather than as a qualitatively distinct category.
Methodology: Applied Meehl's taxometric procedures (MAXCOV, MAMBAC) to data from large samples completing the ECR and other attachment measures. These statistical methods are specifically designed to test whether latent variables are categorical or dimensional.
Honest Assessment: Taxometric methods require large sample sizes and may be sensitive to the specific indicators selected. Some researchers (e.g., Fraley & Spieker, 2003) note that the dimensional finding does not necessarily invalidate categorical models for clinical or descriptive purposes — categories may remain useful heuristics even if the underlying structure is dimensional. The debate between dimensional and categorical models of attachment remains relevant for how attachment styles are communicated in popular and clinical contexts.
Mikulincer & Shaver (2007)
Provided a comprehensive synthesis of adult attachment research, documenting that secure attachment functions as a psychological resource that enhances emotion regulation, relationship quality, mental health, and prosocial behavior. Securely attached individuals demonstrate what the authors call a "broaden-and-build" cycle of attachment security, in which felt security expands the capacity for exploration, empathy, and caregiving.
Methodology: Integrative review and theoretical synthesis drawing on hundreds of studies across experimental, correlational, longitudinal, and neuroimaging methodologies. Included the authors' own extensive experimental priming research demonstrating that activating mental representations of security produces measurable effects on cognition, emotion, and behavior.
Honest Assessment: The security priming literature, while extensive, has faced scrutiny in the broader replication crisis context. Some priming effects may be smaller or less robust than originally reported. Additionally, the "broaden-and-build" framework, while compelling, may overemphasize the benefits of security and understate its potential costs — for instance, securely attached individuals may be slower to detect genuine relational threats or may maintain relationships that are not in their best interest due to high commitment thresholds.
In Relationships
In romantic partnerships, individuals with secure attachment orientations tend to create what Bowlby (1969/1982) described as a secure base and safe haven for their partners. The secure base function involves supporting a partner's autonomous exploration, growth, and pursuit of individual goals, while the safe haven function involves providing comfort, reassurance, and emotional availability during times of distress. Research by Feeney (2004) and Mikulincer and Shaver (2007) documents that securely attached individuals are more effective at providing both functions, calibrating their responsiveness to their partner's actual needs rather than projecting their own anxiety or withdrawing from emotional engagement. This dual capacity — encouraging independence while remaining emotionally present — is consistently associated with higher relationship satisfaction across studies.
Securely attached individuals tend to approach conflict as a problem to be solved collaboratively rather than as a threat to the relationship itself. Empirical research documents that they are more likely to engage in constructive communication during disagreements, including expressing their own needs clearly, listening to their partner's perspective without becoming defensive, and initiating repair after ruptures (Simpson, Rholes, & Phillips, 1996). Levine and Heller (2010) note that securely attached individuals serve as natural emotion regulators in their relationships — their steady emotional availability tends to de-escalate conflict and reduce the intensity of a partner's attachment anxiety or avoidance. Johnson (2008) frames this pattern within Emotionally Focused Therapy, describing how secure partners remain emotionally accessible and responsive even during the inevitable moments of disconnection that occur in all intimate relationships.
It is important to note that secure attachment does not imply the absence of relationship difficulties or emotional pain. Securely attached individuals experience conflict, jealousy, and disappointment; the distinction lies in how they process and respond to these experiences. Mikulincer and Shaver (2007) emphasize that security provides a regulatory advantage rather than emotional invulnerability — securely attached individuals are better able to acknowledge negative emotions, seek appropriate support, and return to emotional equilibrium after distressing events. Tatkin (2012) adds a neurobiological perspective, suggesting that secure couples develop co-regulatory patterns in which each partner's nervous system helps stabilize the other, creating a dyadic system that is more resilient than either individual alone.
Communication Patterns
Research consistently documents that individuals with secure attachment orientations exhibit communication patterns characterized by directness, emotional coherence, and collaborative intent. In the Adult Attachment Interview tradition (Main, Goldwyn, & Hesse, 2003), secure-autonomous individuals produce narratives about attachment experiences that are coherent, balanced, and reflective — they can discuss both positive and negative childhood experiences without idealization or unresolved distress. This narrative coherence translates to interpersonal communication in adult relationships: securely attached individuals tend to express their needs and feelings in straightforward terms, provide context for their emotional states, and maintain an open stance toward their partner's communications, even when the content is difficult (Mikulincer & Shaver, 2007).
Levine and Heller (2010) observe that securely attached individuals are notably effective at what attachment researchers call "effective dependency" — the ability to turn to a partner for support in a way that is clear, non-coercive, and easy to respond to. Unlike anxious individuals who may escalate emotional displays to elicit a response, or avoidant individuals who suppress needs to maintain self-reliance, securely attached individuals tend to communicate needs at a moderate intensity and trust that their partner will respond. This communication style creates a positive feedback loop: clear expression of needs facilitates responsive caregiving, which reinforces felt security, which further supports clear communication. Tatkin (2012) describes this as the establishment of a couple system in which both partners develop shared understanding of each other's signals and reliable protocols for connection and repair.
Under Stress
The stress-buffering effects of secure attachment represent one of the most well-documented findings in attachment research. Mikulincer and Shaver (2007) synthesize extensive evidence showing that securely attached individuals demonstrate superior emotion regulation under stress, employing strategies such as seeking appropriate social support, cognitively reappraising threatening situations, and maintaining access to positive emotional resources even during difficult periods. Experimental studies using security priming — activating mental representations of secure attachment figures — have demonstrated that even brief exposure to security-related stimuli reduces physiological stress responses, increases tolerance for distress, and enhances cognitive flexibility in the face of threat (Mikulincer & Shaver, 2001). These findings suggest that secure internal working models function as an internalized regulatory resource that can be drawn upon even in the absence of an attachment figure.
In the context of relationship-specific stress, securely attached individuals are more likely to employ what Fraley and Shaver (2000) describe as constructive coping strategies. Rather than becoming hyperactivated (as in anxious attachment) or deactivated (as in avoidant attachment), securely attached individuals tend to acknowledge the stressor, communicate about it with their partner, and engage in collaborative problem-solving. Simpson, Rholes, and Nelligan (1992) demonstrated in an observational study that securely attached women who were told they would undergo a stressful procedure sought more comfort from their partners and reported less anxiety compared to insecurely attached women. Importantly, the partners of secure individuals also rated the support-seeking as appropriate and were more responsive in return, illustrating the reciprocal nature of secure attachment dynamics under stress.
Path Toward Growth
For individuals who already exhibit predominantly secure attachment orientations, growth often involves deepening relational capacities and extending security into domains or situations where it may be less automatic. Mikulincer and Shaver (2007) describe a process they call the "broadening" of the security circle, in which felt security expands the individual's capacity for empathy, compassion, and prosocial behavior toward an increasingly wide range of others — not only intimate partners and family members, but friends, acquaintances, and even outgroup members. This expansion reflects the attachment system's broader function as a foundation for social engagement and moral development.
One area of growth for securely attached individuals involves recognizing and managing what might be called the shadow side of security. Because secure individuals tend to hold positive expectations about others' intentions and relationship outcomes, they may sometimes underestimate genuine relational threats, overlook boundary violations, or persist in relationships that are not serving their well-being. Johnson (2008) notes that secure individuals can benefit from developing greater discernment about when their trust is well-placed and when it may be functioning as a form of denial. Tatkin (2012) similarly emphasizes that secure functioning in relationships requires ongoing attentiveness and is not simply a fixed trait that operates automatically.
The concept of earned security (Roisman et al., 2002; Pearson, Cohn, Cowan, & Cowan, 1994) provides one of the most hopeful findings in attachment research: individuals who experienced insecure early attachment can develop secure attachment representations through reflection, corrective relational experiences, and psychotherapy. For those who are already securely attached, understanding the earned security pathway can enhance their capacity to serve as a secure base for others — partners, children, friends — who are working toward greater security. Levine and Heller (2010) note that the mere presence of a securely attached partner in a relationship creates conditions that facilitate the insecure partner's movement toward security, a process that represents one of the most significant interpersonal contributions a securely attached individual can make.
Frequently Asked Questions
Is secure attachment simply the absence of insecurity, or is it a distinct pattern with its own characteristics?
This question reflects an active debate in attachment research. Dimensional models (Brennan, Clark, & Shaver, 1998; Fraley, Waller, & Brennan, 2000) define secure attachment as the low end of both the anxiety and avoidance dimensions, which could suggest it is simply the absence of insecure features. However, Mikulincer and Shaver (2007) argue compellingly that security is a distinct psychological state with its own positive qualities — including effective emotion regulation, cognitive openness, empathy, and the capacity for both autonomy and intimacy — rather than merely the default when anxiety and avoidance are absent. Security appears to involve active regulatory processes and internalized representations of reliable care, not just the absence of defensive strategies. The resolution may lie in recognizing that security is both the low-anxiety/low-avoidance region of dimensional space and a coherent psychological organization with distinctive features.
Can a person become securely attached later in life if they were not securely attached as a child?
Yes. Research on earned security demonstrates that individuals can develop secure attachment representations despite insecure early experiences. Roisman et al. (2002) found that adults classified as "earned secure" on the Adult Attachment Interview — meaning they reported difficult childhood experiences but discussed them with coherence and reflective understanding — functioned comparably to "continuous secure" adults in their romantic relationships and parenting behavior. Corrective pathways include long-term relationships with securely attached partners (Levine & Heller, 2010), psychotherapy — particularly attachment-informed approaches such as Emotionally Focused Therapy (Johnson, 2008) — and sustained self-reflection. However, it is important to note that the process is typically gradual and requires substantial relational experience, not simply cognitive insight.
If secure attachment is the most common style, why does so much popular content focus on insecure attachment?
Population estimates from Western samples consistently place secure attachment at approximately 50-60% of adults (Hazan & Shaver, 1987; Mickelson, Kessler, & Shaver, 1997), making it the modal pattern. The disproportionate popular focus on insecure attachment likely reflects several factors: insecure patterns involve more visible distress and interpersonal friction, which generates greater demand for self-help content; the clinical origins of attachment theory naturally emphasize problematic patterns; and the experience of attachment insecurity is often what motivates individuals to seek information about attachment in the first place. From a research perspective, Fraley and Shaver (2000) note that understanding secure attachment is essential precisely because it provides the template for what effective attachment functioning looks like, serving as the benchmark against which insecure patterns are understood.