Journal of Counseling Psychology Copyright 2000 by the American Psychological Association, Inc.

2000, Vol. 47, No. 3, 283-300 0022-0167AXW5.00 DOI: 10.1037//0022-O167.47.3.283

Dynamic Processes Underlying Adult Attachment Organization: Toward an

Attachment Theoretical Perspective on the Healthy and Effective Self

Frederick G. Lopez Kelly A. Brennan

Michigan State University State University of New York College at Brockport

Contemporary attachment theory is inspiring a burgeoning literature on adult attachment. This literature
offers counseling psychology a compelling framework for understanding the healthy and effective self.
In this article, the authors review studies probing the cognitive processes, affect self-regulatory dynamics,
and relationship behaviors associated with secure (primary) and insecure (secondary) adult attachment
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

strategies. They then offer a portrait of the healthy, effective personality that is consistent with theory and
This document is copyrighted by the American Psychological Association or one of its allied publishers.

research on adult attachment. Finally, in an effort to understand key characteristics of successful

movement toward the healthy and effective self, the authors review empirical extensions of attachment
theory to the therapeutic context.

What is the "healthy and effective" self? How can we explain psychology lacks an understanding of individual difference vari-
the life-affirming drive toward optimal functioning? Decades of ables that predict adaptive, or optimal, functioning across situa-
psychological research reveal that certain individuals possess sig- tions and throughout the lifespan. Gelso and Fassinger proposed
nificant intra- as well as interpersonal strengths (Cattell, 1973; several historical reasons for this lack of connection between
Vaillant, 1977). These strengths include, but are not limited to, counseling and personality psychology, one of which is especially
positive attitudes (e.g., optimism), adaptive cognitive appraisals salient to this discussion: the scarcity of theories effectively inte-
(e.g., identifying stressful events as challenges instead of threats), grating personality and developmental themes within a common
constructive methods of coping with emotional distress (e.g., seek- framework.
ing social support instead of withdrawing or using drugs or alco- In line with arguments presented by Lopez (1995a), we believe
hol), insight into one's personality and emotional dynamics (e.g., that attachment theory represents a perspective with considerable
understanding one's own emotions and emotional reactions), and potential for advancing counseling psychology. For example, as
extensive social skills (e.g., empathy, sensitivity to others' feel- outlined by John Bowlby (1969/1982, 1973, 1980, 1988), attach-
ings, assertiveness, and effective conflict-resolution strategies). ment theory effectively juxtaposes personality and developmental
Such exceptional individuals have been labeled "zestful" (Fore-
themes within a broad lifespan framework. In particular, we con-
man, 1966), "self-actualized" (Maslow, 1970), "fully functioning"
tend that the literature in adult attachment is contributing to an
(Rogers, 1962), or as having a "hardy personality" (Kobasa, 1982),
increasingly comprehensive understanding of the healthy and ef-
among others. Whatever the label, these persons appear to not just
fective self. We examine this argument in two ways. First, we
survive but thrive. They are resilient in adapting to the vicissitudes
identify key attachment theoretical concepts and assumptions.
of life, and they have a remarkable capacity to develop, maintain,
Second, we review recent empirical studies testing theory-driven
and enlarge their networks of supportive, intimate relationships.
questions and hypotheses concerning cognitions, emotions, and
Does counseling psychology have a good understanding of these behaviors relevant to attachment strategies. Such strategies appear
myriad competencies, their developmental antecedents, and the to vary in the extent to which they preserve stability and continuity
dynamic means by which they optimize human functioning? As while accommodating personal and interpersonal change. Thus, it
Gelso and Fassinger (1992) noted, "our efforts to form a coherent is our intent to go beyond prior reviews of the adult attachment
understanding of the healthy and effective self have been ham- literature (e.g., Collins & Read, 1994; Feeney, 1999; Fuendeling,
pered by a marked 'ambivalence' between the fields of develop- 1998; Hazan & Shaver, 1994; Mikulincer & Florian, 1998). These
mentaiypersonality psychology and counseling psychology—de- reviews, albeit comprehensive, stop short of organizing findings
spite their 'profound connections' (p. 275). In short, counseling into a discussion about optimal personality development. In addi-
tion, because empirical extensions of attachment theory to coun-
seling and therapeutic relationships have only recently emerged,
Frederick G. Lopez, Department of Counseling Psychology, Michigan they have not yet been included in prior reviews.
State University; Kelly A. Brennan, Department of Psychology, State Our article has three parts. In Part 1, we examine why adult
University of New York College at Brockport.
attachment dynamics and relationships are important to under-
We thank James Fuendeling, Eric Sauer, and Kim Thomas for their
comments on an earlier draft of this article. standing the development of the "healthy personality." We build
Correspondence concerning this article should be sent to Frederick G. the conceptual basis for this argument by first outlining Bowlby's
Lopez, 438 Erickson Hall, Michigan State University, East Lansing, Mich- original ideas regarding a lifespan, relationship-centered perspec-
igan 48824. Electronic mail may be sent to [email protected]. tive of healthy personality development. We then review recent
283
 284 LOPEZ AND BRENNAN

empirical findings that specifically assess adjustment-related dy- pathways along any one of which he might travel. The one chosen, it
namic processes associated with one's dominant strategies for is held, turns at each and every stage on an interaction between the
managing adult attachment-related insecurity. To achieve a sharper organism as it has developed up to that moment and the environment
focus on the key theoretical construct of adult attachment organi- in which it then finds itself. (Bowlby, 1988, pp. 64-65)
zation, we limit our review to studies using measures of adult
Primary and secondary attachment strategies. Attachment
attachment, and, for the most part, exclude studies that focus solely
theory further assumes that the human attachment system is an
on associations between early parent-child bonds and adult func-
innate and unique behavioral system, independent of other systems
tioning. We distinguish between primary versus secondary strat-
such as mating and feeding. Unlike these other systems, the
egies and then use this distinction to discuss findings from the
primary (and evolutionary) function of the attachment system is to
contemporary literature on adult attachment. In Part 2, we offer an
integrated attachment-theoretical perspective on the healthy per- enhance the infant's survival by successfully recruiting relational
sonality that is consistent with Bowlby's original ideas as well as care and protection in the face of threats to his or her security.
with key findings from contemporary research. In Part 3, we Bowlby reasoned that the infant's early interactions with his or her
review theory and research on attachment and the counseling primary caregiver quickly form a goal-corrected partnership that
relationship and derive implications for understanding therapeutic initially calibrates the infant's coordination of proximity-seeking
and proximity-avoidance behaviors. Sensitive, reliable, and re-
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

movement toward the healthy and effective self.

This document is copyrighted by the American Psychological Association or one of its allied publishers.

sponsive care from the adult figure promotes the infant's experi-
ence of felt security. A secure parent-child bond facilitates the
Part 1: Attachment Theory—A Relationship-Centered infant's development by providing him or her with a "secure base,"
Perspective on Healthy Personality Development or a predictable relational context, for confidently exploring the
environment and progressively mastering early developmental
Conceptual Perspectives
tasks within tolerably safe limits. Secure attachment is thus ef-
Counseling psychology has historically favored theoretical fected by the unfettered operation of this primary strategy for
frameworks that offer lifespan, context-sensitive views of person- managing insecurity (Main, 1990). This strategy, in turn, promotes
ality and development—frameworks that, while emphasizing the the development of an interdependent orientation to human rela-
centrality of person-environment interactions in understanding tionships—one that permits both separateness and exploration as
normal adjustment, are capable of envisioning how these dynamics well as connectedness and care seeking.
lead to higher order forms of competence and adaptation (Gelso & On the other hand, in early rearing environments lacking ade-
Fretz, 1992). More recently, our discipline has entertained con- quate, reliable caregiving, the infant may learn a secondary strat-
structivist models that describe how individuals cognitively rep- egy (Main, 1990) for enlisting caregiver responsiveness or for
resent and interpersonally negotiate their current identities and internally managing his or her responsiveness to threatening stim-
realities (Mahoney & Lyddon, 1988). In short, we seek to under- uli. These secondary strategies portend the formation of an inse-
stand human beings as both the products and producers of their cure bond with the caregiver, one marked by either a chronic
own development (Lerner, 1991)—as active, purposive organisms hyperactivation or deactivation (suppression) of the infant's at-
interacting with their surrounding contexts, using fundamental tachment system. In short, the early formation of an insecure
capacities to represent their experiences to themselves and others attachment bond with one's primary caregiver(s) begins to bias the
in ways that affect the course of their own development, for good infant's relational development toward either an excessive preoc-
or ill.
cupation with maintaining proximity or toward a persistent avoid-
Developmental pathways. Consistent with these historical and ance of closeness. Early observational research on mother-infant
contemporary preferences, attachment theory seeks to understand dyads in a controlled setting designed to activate the infant's
personality formation and development in the context of one's attachment behavior (the "Strange Situation") initially identified
continuing embeddedness in close, affectional relationships— three attachment patterns (secure, anxious-ambivalent, and
often with romantic partners but also with close friends, siblings, avoidant). These three styles were associated with the operation of
and parents. According to Bowlby, our inborn and evolutionarily primary and secondary attachment strategies (Ainsworth, Blehar,
designed dispositions to seek and maintain the experience of Waters, & Wall, 1978). Infants who are classified as secure (i.e.,
security with our primary caregivers, coupled with the nature and who rely on the primary strategy of neither hyperactivation nor
quality of our caregivers' responsiveness, set the stage for predict- deactivation) exhibited distress when separated from their mothers
able cognitive, affective, and relational behavior processes that but were comforted on reunion and subsequently resumed their
shape both personality and development along one of multiple, independent exploratory behaviors. Infants who are classified as
potentially adaptive or maladaptive pathways. According to the-
anxious-ambivalent (i.e., who exhibit the secondary strategy of
ory, one's early experiences with environments that promote feel-
hyperactivation), by contrast, demonstrated considerable distress
ings of security or threat result in internalized dispositions that
on maternal separation, were not promptly comforted by their
largely determine a person's resilience and vulnerability to stress-
mothers' return, and instead frequently exhibited angry protest
ful life events. These dispositions, in turn, establish pathways
behaviors. Finally, infants who are classified as avoidant (i.e., who
leading toward healthy or problematic development.
display the secondary strategy of deactivation) overtly appeared
Within this framework, human personality is conceived as a structure undistressed by their mothers' departures (although subsequent
that develops unceasingly along one or another of an array of possible research showed that their heart rates are elevated; Sroufe &
and discrete pathways. All pathways are thought to start close together Waters, 1977) and did not seek proximity with her on her return to
so that, at conception, an individual has access to a large range of the observational room. Subsequent research (Main & Hesse,
 DYNAMIC PROCESSES 285
1990) using the same methodology revealed a fourth disorganized/ decided to use categorical language in our reviews, for two rea-
disoriented pattern that appears to represent a combination of both sons. First, dimensional methodologies are not always used. Sec-
hyperactivating and deactivating strategies (Hesse, 1999; Lyons- ond, when used, researchers have continued to discuss attachment
Ruth & Jacobvitz, 1999). in terms of "types," or "styles," not dimensions.
It should be noted that Ainsworth et al. (1978) did not use the Internal working models: From attachment strategies to repre-
language of activation and deactivation per se. They carefully sentational models. Bowlby believed that, within the first years
calibrated the Strange Situation laboratory test to activate the of life, these early relational patterns of proximity seeking and
young child's attachment system because individual differences in proximity avoidance become cognitively represented by the child
attachment patterns are most apparent when the child is mildly in the form of an internal working model, or schemas regarding
stressed. (When not stressed or stressed too much, such individual close relationships. By serving as the primary cognitive circuitry
differences are not readily apparent.) The Strange Situation coding responsible for organizing one's attachment behavior, this model is
system includes many rating scales for evaluating infant behavior believed to function as the person's internalized "template" for
during the separations and reunion episodes (e.g., proximity seek- close relationships. The internal working model is presumed to
ing, separation protest). These ratings were then used to classify incorporate two discrete yet interrelated components: a self model
the three (and later four) parent-infant attachment categories. regarding one's own sense of worth and lovability and an other
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Ainsworth et al. also conducted a discriminant function analysis to model embodying core expectations of the availability and trust-
This document is copyrighted by the American Psychological Association or one of its allied publishers.

predict the three categories from the rating scales. The ratings worthiness of intimate others in one's social world. By definition,
formed two major functions, or dimensions, that significantly working models are inherently relational schemas that embody not
differentiated the three categories. The two major dimensions were only self-perceptions and other-perceptions but also implicit action
anxiety and avoidance. Secure infant-mother pairs appeared to be strategies for responding to security-related threats and violations.
lower in anxiety compared with either insecure infant-mother Said differently, the internal working model comprises a set of
pairs. The insecure pairings were distinguishable in terms of avoid- overlearned, abstracted, generalized, and not necessarily fully con-
ance—the avoidant mother-infant pairings appeared to be higher scious appraisals and expectations, as well as relational scripts for
in avoidance compared with the anxious-ambivalent pairings. choreographing one's social behavior during circumstances per-
These two dimensions became important in later work on adult ceived as threatening to the self. These features also imply that,
attachment, in the realms of romantic or peer attachment (Brennan, beyond incorporating differing (generalized) expectations of self
Clark, & Shaver, 1998), which is why we are taking the time to and others, working models may also embody subtle cognitive
discuss them further. We see the familiar cognitive-affective- biases that can affect the direction and duration of attentional
behavior triad apparent in these dimensions. First, cognitions about processes, the level of perceptual vigilance for possible environ-
perceived danger, threat, stress, or illness are likely to lead to mental threats, the manner in which emotional arousal is managed,
attachment-system activation and the felt experience of anxiety. and the degree of cognitive flexibility that can be deployed when
Anxiety, in turn, is an emotion in response to perceived stress that one is confronted with new or discrepant information about self or
signifies attachment-system activation. Avoidance is the behav- others.
ioral outcome of such attachment-system activation. Said another Finally, these schemas are assumed to function as "working"
way, avoidance represents one's behavioral strategies for manag- models because their particular information processing is both
ing the relationship with one's caregiver in order to reduce the responsive and reactive to threats to felt security. As such, one's
threat of insecurity and to return the attachment system to a dominant strategy for managing attachment-related security should
quiescent state. We agree with Mary Main (1981), who theorized be especially salient during periods of life change and transition
about avoidance as an adaptive strategy for maintaining proximity and particularly influential in foreshadowing one's responses to
to less than optimally responsive caregivers without provoking these adaptational challenges (Lopez, 1993). Relevant to this point,
them; if provoked, such caregivers might become abusive toward Bowlby (1969/1982) asserted that our underlying attachment or-
the infant or abandon the infant altogether. Eventually, the re- ganization functions as a stress-activated system; thus, the influ-
peated use of avoidant strategies (e.g., averting eyes away from ence of the internal working model becomes especially prominent
caregiver when angry at caregiver, not seeking help from caregiver when the individual is distressed, fatigued, or ill. Because of this
when one is scared, stressed, or ill) is likely to lead to chronic unique function, the internal working model guides the perceiver's
attachment-system suppression. For infants with unresponsive, attributional and explanatory analyses of his own and of the other's
rejecting, or abusive caregivers, activating the attachment system responses precisely in those contexts that are highly fluid, unpre-
in response to any but the most dangerous cues is likely to be futile dictable, and unstable. In other words, by imposing its cognitive
if not counterproductive. Thus, for us, focusing on the underlying organization and perceptual biases on the interpersonal dynamics
dimensions provides a more useful way to think about attachment within these contexts, the working model constructively shapes the
dynamics than merely analyzing the attachment categories. In nature and meaning of the context itself and thus the quality of the
addition, subsequent research has shown that neither the infant- individual's experiences therein.
parent categories (Waters, Beauchaine, & Brennan, 2000) nor the Progressive developments in the conceptualization and mea-
adult romantic/peer categories are truly taxonomic in nature (Fra- surement of adult attachment styles and orientations, using self-
ley & Waller, 1998; Waters et al., 2000). The lack of taxonicity report and interview methods, have accelerated this general line of
probably accounts for the finding that attachment dimensions inquiry. For example, Hazan and Shaver (1987) first proposed that
typically account for four to five times the variance as attachment adult romantic love could be conceptualized as an attachment
categories (Brennan et al., 1998). Despite the theoretical and process; these authors also fashioned a self-report measure of adult
empirical lack of justification for using categories, we nonetheless attachment styles that conceptually paralleled Ainsworth et al.'s
 286 LOPEZ AND BRENNAN

(1978) original three-category classification scheme (i.e., secure, divorce. In line with these results, Vaughn and Bost (1999) have
anxious-ambivalent, and avoidant). More recently, Bartholomew reviewed the relevant literature and concluded that differences in
and Horowitz (1991) have validated a self-report measure based on attachment (secure vs. insecure) are not isomorphic with "temper-
a cross-classification of positive and negative models of self and ament" (defined as early appearing personality traits, possibly
other that yields four adult attachment style groups: This measure genetic in origin). These reviewers note, however, the neonatal
identifies persons with secure (positive self, positive other); pre- irritability is a risk factor for insecure attachment, perhaps by
occupied (negative self, positive other); dismissing (positive self, affecting parents' ability to provide the kind of sensitive, respon-
negative other); and fearful (negative self, negative other) styles. sive caregiving that results in secure attachment; such linkages
Griffin and Bartholomew (1994) supported the construct validity disappear when other risk factors, such as socioeconomic status are
of this typology in a study using the multitrait/multimethod strat- taken into account. Vaughn and Bost (1999) also concluded that
egy, obtaining data derived from participants' self-reports, friends, the literature does not support the idea that differences among
intimate partners, and from independent judges. As mentioned, children with insecure attachment are due to temperament (the
other investigators have developed reliable, continuously scaled hyperactivation vs. deactivation dimension we discuss in this ar-
measures of adult attachment-related anxiety and avoidance di- ticle). All in all, the theory and the research literature emphasize
mensions (Brennan et al., 1998), which, if desired, can be used to the centrality of critical events experienced in the context of close
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

place people into one of the four categories identified by Bar- relationships in altering both model revision and the trajectory of
This document is copyrighted by the American Psychological Association or one of its allied publishers.

tholomew and Horowitz. Still others have classified adults into one developmental progress (Bowlby, 1988). (In Part 3 of our presen-
of three or four attachment groups on the basis of responses to a tation, we more closely consider the implications of this assump-
semistructured interview regarding early experiences with parental tion to our understanding of therapeutic change.)
caregivers (the Adult Attachment Interview [AAI]; George, For now, we turn to the findings of recent empirical investiga-
Kaplan, & Main, 1985). tions probing the dynamic processes underlying adult attachment
The past as prologue: Stability and continuity in adult attach- organization. In line with our overarching purposes, we organize
ment organization. Of particular interest and relevance to our and discuss these findings in terms of their elaboration of the
disciplinary interests and goals have been these conceptual and cognitive, affective, and relational behavior dynamics associated,
methodological linkages that extend attachment theory's assump- respectively, with operation of primary and secondary attachment
tions about early personality formation to later (adult) adjustment strategies among adults. We precede each section with a brief
and development. These linkages have stimulated a highly pro- theory-derived synopsis supporting the particular thematic line of
ductive line of inquiry, beginning with the seminal work of Hazan inquiry, followed by a review and summary of key findings.
and Shaver (1987) and Main and her colleagues (George et al.,
1985) that continues unabated. More importantly, these linkages
Empirical Perspectives on the Primary and
offer a unique glimpse of how stable features of personality
structure are likely maintained by internal and external self-
Secondary Strategies
regulatory actions. As elaborated in the following sections, this Attachment theory contends that a key benefit of one's early
glimpse has far-reaching implications for advancing a more com- (and continuing) experience of attachment security in intimate
plete understanding of the dynamic person-context relations that relationships is an enhanced capacity to adaptively regulate one's
undergird personality stability and continuity, particularly with own cognitive and affective processes. Such adaptive regulation
respect to one's orientation to close relationships. should occur without chronically hyperactivating or deactivating
This is not to say, however, that attachment theory presumes the attachment system. We believe that the literature linking adult
that the qualities of early infant-caregiver relationships and of attachment security to various indexes of cognitive and affective
working models have developmental effects that are predeter- dynamics supports this view. In addition, these findings paint an
mined. Attachment theory concedes that powerful environmental integrated portrait of the healthy personality as someone capable of
events, especially those that disconfirm, overwhelm, or otherwise reflecting on his or her own thought processes, forming reasonably
erode the integrity of the working model may precipitate signifi- accurate and differentiated assessments of self and others, access-
cant revisions of the model toward either greater or lesser security ing and coherently organizing context-relevant affects and mem-
and thus result in discontinuities in the impact of earlier models on ories, and deploying generally flexible interpretive and coping
current and future development. For example, children appear to strategies. In our view, these enhanced capacities collectively
show different attachment styles with different caregivers early in enable the securely attached adult to function as a more adaptive
development (12 months), but their attachment styles in later self-organizer of his or her internal perceptions, attributions, and
childhood appear most closely predictable from that of their pri- affective states. Individuals who exhibit secondary attachment
mary attachment figure (Main & Cassidy, 1988; see also Howes, strategies may have one of several insecure attachment styles. The
1999). Also, Waters and his colleagues (Waters, Merrick, Treboux, research discussed in this section shows that individuals with
Crowell, & Albersheim, 1999) have discovered 70% secure- insecure, or secondary, attachment strategies share cognitive, af-
versus-insecure continuity in attachment to parents from infancy to fective, and relational behaviors that are generally diametrically
adulthood and 64% continuity among three attachment categories opposed to their secure counterparts. Individuals with insecure
(secure, anxious-ambivalent/preoccupied, and avoidant/dismiss- styles, or secondary attachment strategies, tend to think, feel, and
ing). More to the point, greater continuity (78% for secure/insecure behave in ways that reflect the more-or-less chronic state of their
and 72% for the three-group comparison) was observed in indi- attachment system (hyperactivated vs. deactivated) in combination
viduals who had not experienced major disruptive life events, such with their internal working models of self and other (positive vs.
as major illness (of parent or child), child abuse, parental death, or negative). Individuals with the preoccupied orientation, for exam-
 DYNAMIC PROCESSES 287

pie, who are likely to have hyperactivated attachment systems Still other characteristics fall under the rubric of generic infor-
along with a negative self model and positive other model, tend to mation processing skills. Individuals with a secure style appear
be consumed with thoughts of obtaining and retaining access to more flexible and open-minded in their information processing:
partners, are likely to focus on their intense negative emotions, and They tend to accommodate their cognitive structures to fit with
to regularly seek support from partners; oftentimes, however, these new information—as opposed to assimilating information to fit
individuals seek more support than partners are willing or able to with preexisting structures (Mikulincer, 1997; Mikulincer & Arad,
give, thereby compounding their distress with attachment-related 1999). Compared with individuals with insecure styles, those with
anxiety over abandonment. Individuals with a dismissing orienta- a secure style tend to possess greater tolerance for ambiguity,
tion, in contrast, tend to suppress attachment-related thoughts, disorder, and uncertainty (Mikulincer, 1997) and are more capable
possess a chronically deactivated attachment system and, even of recalling (across several distinct situations) information about
when stressed, are more likely to use coping strategies that involve their relationship partners that is incongruent with their prior
distancing rather than seeking support from partners. Individuals expectations of partner behavior (Mikulincer & Arad, 1999). The
with a fearful orientation usually report a great deal of emotional secure style has also been associated with higher levels of con-
distress (i.e., a hyperactivated attachment system) and, as a result, structive thinking, which is defined as the ability to use a range of
appear highly focused on negative emotions. At the same time, and flexible, reality-based cognitive processes and appraisals when
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

unlike their preoccupied counterparts, persons with a fearful style dealing with everyday problems (Lopez, 1996). By definition,
This document is copyrighted by the American Psychological Association or one of its allied publishers.

appear unwilling or unable to seek support from others, presum- these individuals' AAIs are characterized by flexible, collabora-
ably due to their internalization of negative self and other models.1
tive, and coherent speech (Main & Goldwyn, 1994). In addition,
Cognitive processes. Empirical investigations over the past their speech is more likely to be characterized by "reflective
decade reveal common cognitive processes among individuals self-awareness," or the capacity of the person to monitor his or her
who use the primary, or secure, attachment strategy. Not surpris- output during the course of the hour-long AAI while concurrently
ingly, most of this research pertains to social-perception processes. bearing in mind the mental state of the interviewer (Fonagy et al.,
Relative to individuals who do not use the primary attachment 1995, 1996). To test hypotheses regarding this "reflective self'
strategy, secure individuals possess more accurate and highly function, Fonagy et al. (1996) devised an additional scale for
differentiated assessments of others (Levy, Blatt, & Shaver, 1998); scoring AAI transcripts to assess reflective self-awareness (RSF).
more integrated self-structures (i.e., a more internally organized The RSF scale showed high interrater reliability (.91, Fonagy et
system of self-representational attributes; Mikulincer, 1995), par- al., 1996). High scorers on the RSF scale demonstrated greater
ticularly when under stress (Lopez, Gover, et al., 1997); make awareness and sensitivity to both their own and others' mental
more positive and accurate attributions for partner behavior (J. A. states and explicitly recognized the existence of diverse perspec-
Feeney, 1998; Mikulincer, 1998c); have more positive expecta- tives on the same event. Fonagy concluded that this higher order
tions about their partners' behavior (Pietromonaco & Carnelley, cognitive capacity emerges fully only in the context of a secure
1994; Collins, 1996; Feeney, 1998); make more flexible and
attachment relationship and that the inhibition of reflective self-
benign appraisals/interpretations of partner behavior (Pistole,
awareness (a) reduces people's ability to cognitively resolve early
1996); and demonstrate (in reaction-time tasks) attention that is
trauma and (b) increases the likelihood that they will exhibit
biased toward greater trust of others (Baldwin, Fehr, Keedian,
borderline psychopathology should such trauma be experienced
Seidel, & Thomson, 1993; Baldwin, Keelan, Fehr, Enns, & Koh-
(Fonagy et al., 1996).
Rangarajoo, 1996; Miller & Noirot, 1999).
Finally, other organizational features of self-structures appear to
Other cognitive characteristics pertain to memory for
differ as well. Individuals with the secure style do not appear to
attachment-related events: Individuals with a secure orientation
endorse rigid rules for maintaining self-worth (Roberts, Gotlib, &
appear to possess greater access to memories of early childhood
Kassel, 1996). Instead, they possess balanced self-views that ap-
events, especially affect-laden ones, using the AAI (Hesse, 1999;
parently derive from both interpersonal and achievement sources
Main, Kaplan, & Cassidy, 1985). (The AAI, among other things,
(Brennan & Morris, 1997). In contrast, individuals with a dismiss-
probes the interviewee's memories and affects associated with
ing style appear to derive their (high) self-esteem from achieve-
early experiences with parental figures.) In addition, using other
methods, individuals with a secure style have been shown to
possess greater memory for positive, trust-related events (Miku- 1
Less is known about either kind of avoidants. Before 1990, there was
lincer, 1998c). Memory patterns discriminate reliably among one only one kind of avoidant in the literature: Main's dismissing type and
or more insecure strategies, too. Using the AAI, individuals with a Hazan and Shaver's avoidant type. Bartholomew (1990; Bartholomew &
dismissing style consistently demonstrate poorer memory for Horowitz, 1991) noticed that Hazan and Shaver's avoidant was too "soft"
attachment-related events, particularly involving negative memo- and nondefensive, prompting her to define two types of avoidance resulting
ries of childhood, whereas those with a preoccupied style demon- from combinations of self and other models. As mentioned, both dismiss-
strate greater memory for negative childhood events (Hesse, 1999; ing and fearfuls have negative models of others, but the two differ in
Main et al., 1985). It has been noted that the AAI was not designed models of self. Most researchers now utilize the four-category typology
to assess the accuracy of memory for childhood events but rather (although a minority still prefer the three-category typology), and most of
the extant research reviewed here was already conducted using the three-
the manner in which such events are recounted to the interviewer;
category typology. Thus, it is impossible to determine what proportion of
thus, patterns of defensiveness, cooperation, or inappropriate bids research samples contain fearful versus dismissing individuals. In discuss-
for help in response to interview questions are all scored indepen- ing the self-report literature, where possible, we distinguish among the two
dently of the content of the interview itself. (See Hesse, 1999, for types of avoidants; otherwise, we simply refer to "avoidants," which
a review.) indicates the original Hazan and Shaver avoidant.
 288 LOPEZ AND BRENNAN

ment/competence sources. Unlike their dismissive peers, those the hyperactivating attachment strategy exhibited more overt dis-
with a secure style evidence more cohesive and integrated assess- tress and less self-reliance, whereas persons with a deactivating
ments of self and others that are resistant to decompensation under strategy were more likely to disavow the experience of distress
stress (Lopez, Fuendeling, Thomas, & Sagula, 1997). Kennedy while correspondingly demonstrating (relative to their secure
(1999) found that individuals with a secure style scored signifi- counterparts) less optimal scores on other measures of well-being.
cantly higher compared with their peers with dismissing, preoccu- Adopting a more process-oriented focus, other studies have
pied, or fearful styles on a measure of ego identity achievement. clarified how the primary attachment strategy affects the experi-
Thus, in line with Bowlby's (1980) theorizing, the primary (se- ence, retrieval, and defensive processing of specific affects and
cure) attachment strategy appears to entail an adaptive equilibrium affective memories (Mikulincer & Orbach, 1995), including un-
between the attachment and exploration behavioral systems. pleasant affects such as anger (Mikulincer, 1998b), shame and
Affect self-regulation. Empirical investigations over the past guilt (Lopez, Gover, et al., 1997), jealousy (Sharpsteen & Kirk-
decade also reveal common affect self-regulatory characteristics patrick, 1997; Rapoza & Malley-Morrison, 1997), grief reactions
associated with varying attachment strategies. Beyond promoting (Pistole, 1996), and depressive self-criticism (Zuroff & Fitzpatrick,
thought processes that permit flexible, benign, and self-reflective 1995). For example, relative to their insecure peers, those with the
cognitive operations, adult attachment security appears to facilitate secure style have greater access to emotion-laden childhood mem-
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

the adaptive self-organization of internal experience by concur- ories (Mikulincer & Orbach, 1995), are more competent at dis-
This document is copyrighted by the American Psychological Association or one of its allied publishers.

rently enhancing the person's capacity to manage and modulate criminating between related but distinct emotions (e.g., shame vs.
affective arousal in the face of actual or potentially distressing guilt; Lopez, Gover, et al., 1997), and are better at expressing and
information or events. With roots firmly lodged in the study of managing negative emotions in an appropriate manner (e.g., di-
dynamics within close relationships around threats to security, recting jealous anger at a straying partner instead of at oneself or
attachment theory clearly speaks to affect regulation as a signifi- the interloper; Sharpsteen & Kirkpatrick, 1997). Mikulincer's
cant mediator of attachment-related behavior. Indeed, some con- (1998b) research on anger revealed that persons with the secure
temporary researchers regard attachment theory as essentially a style were low in anger-proneness, endorsed more constructive
theory of affect regulation (Kobak & Sceery, 1988) and have anger goals, and expected more positive outcomes from anger
proposed that, in the course of human development, the principal episodes than did those with an insecure style. An important
function of affect regulation is the maintenance of emotional
implication of these findings is that the primary attachment strat-
security (Chorpita & Barlow, 1998; Cummings & Davies, 1996).
egy contributes to the formation of the healthy, adaptive person-
A number of recent empirical studies probing the links between
ality by enabling individuals to make finer grained distinctions
adult attachment styles and various aspects of affect arousal and
between distressing negative affects, thus enhancing their pros-
coping lend support to this view (Fuendeling, 1998).
pects of both effectively modulating their autonomic arousal and
Several studies attest to "secures"' enhanced capacity to cope engaging in more differentiated, affect-appropriate behaviors. By
with stress and negative emotions. Individuals using the primary contrast, in response to anger-eliciting scenarios, individuals with
attachment strategy tend to seek social support when stressed and either the anxious or avoidant style reacted with more negative
tend not to use alternative, maladaptive methods to cope, such as affect, higher heart rates, and stronger attributions of hostile part-
distancing from others, or focusing solely on negative emotions ner intentions than did securely attached persons. Mikulincer's
(Brennan & Shaver, 1995; Cooper, Shaver, & Collins, 1998; J. A. findings also revealed a discrepancy within the avoidant pattern:
Feeney, 1998; Lussier, Sabourin, & Turgeon, 1997; Mikulincer & low self-reported anger but high physiological signs of anger
Florian, 1995). Brennan and Shaver (1995), for example, found arousal, which suggests dissociated anger. Interestingly, however,
that individuals with avoidant or anxious-ambivalent styles de- when presented with paired drawings depicting either joy, anger,
scribed themselves as more likely to use food, sex, and alcohol to
or sorrow, individuals with the avoidant style remembered depic-
quell negative emotions compared with those with a secure style.
tions of anger significantly better than did individuals with either
Avoidance was associated with preferences for sexual activity that
the secure or anxious-ambivalent style (Kirsh, 1995).
does not involve emotional intimacy (e.g., one-night stands). Both
avoidance and anxiety were associated with "drinking to cope" and Using an experimental methodology that manipulated the use of
"bingeing under stress." Avoidance in particular was associated defensive strategies, Fraley and Shaver (1997) found that adult
with higher scores on measures of anorexic eating patterns as well attachment styles influenced both the cognitive and physiological
as a lack of bodily, or interoceptive, awareness, which is thought components of thought suppression. Their findings revealed that
to lead to eating disorders (Brennan & Shaver, 1995). explicit instructions to participants to not think about a loss image
Similar associations have been observed between adult attach- (i.e., the thought of their partner leaving them) led to a decrease in
ment orientations and the manner in which participants coped with autonomic arousal among students with the dismissing style but to
acute and normative stressors such as the "leaving home" transi- an increase in physiological activity among students with a pre-
tion of college students (Mayseless, Danieli, & Sharabany, 1996), occupied style. Fraley and Shaver concluded that those with a
the divorce adjustment process (Birnbaum, Orr, Mikulincer, & dismissing style are capable of actually suppressing the experience
Florian, 1997), and partner separations due to military assignment of arousal and are not simply concealing or disavowing latent
(Cafferty, Davis, Medway, O'Hearn, & Chappell, 1994) or per- distress; they speculated further that (a) this competency may
sonal travel (Fraley & Shaver, 1998). In each instance, persons function as a defensive strategy that results from habitually focus-
endorsing the primary (secure) strategy demonstrated more self- ing attention away from thoughts and memories that might trigger
reliant responses while concurrently acknowledging appropriate attachment-related affective memories, and (b) this strategy may
and well-modulated separation distress. In general, persons using not be useful when circumstances do not permit such suppression.
 DYNAMIC PROCESSES 289

This thinking is line in with Main's (1981) analysis of avoidant phasis on relatedness or an overemphasis on differentiation. Secure
infants, mentioned earlier. attachment histories and models, however, enable a person to chart
Mikulincer and his associates (Mikulincer, 1998a; Mikulincer, an optimal course between relatedness and differentiation, thus
Orbach, & Iavnieli, 1998) have more deeply probed the affective- making one's ongoing participation in secure, intimate relation-
cognitive interface of adult attachment dynamics in a series of ships the clearest instantiation of one's adaptive self-development.
studies exploring affect regulation strategies associated with vari- Empirical support for relational characteristics associated with
ations in self-views and in self-other similarity judgments. Their the primary attachment strategy can be found in recent studies that
findings indicate that insecure persons' self-views may be cogni- have specifically explored associations between adult attachment
tive manifestations of attachment-related affect regulation strate- styles, on the one hand, and social behavior, partner preferences,
gies. For instance, the positive self-views of persons with an and communication practices on the other. The majority of these
avoidant style were strengthened by prior or concomitant distress investigations are correlational in design, some are experimental
arousal and closely associated with efforts to reassert self-reliance; using hypothetical partners, and still others use observational or
the self-views of persons with an anxious-ambivalent style, on the experimental studies using actual partners. Across all the different
other hand, reflected a hyperactivation of inner distress, overem- types of research, however, a common theme emerges: The pri-
phasized personal deficiencies and imperfections, and were linked mary attachment strategy is optimal for enjoying and maintaining
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

with efforts to obtain others' compassion and support. Adult at- relationships and for managing conflict and stress both in and out
This document is copyrighted by the American Psychological Association or one of its allied publishers.

tachment style differences also predicted patterns of subjective of the relationship.

self-other similarity as well as the "cognitive maneuvers" partici-
In one study, participants were asked to maintain social inter-
pants used for changing self-other similarity upon distress arousal.
action diaries, and attachment security was shown to be associated
More specifically, persons with an anxious-ambivalent style over-
with activity patterns and immediate perceptions, especially in
estimated and persons with an avoidant style underestimated self-
high-conflict, attachment-relevant contexts (Tidwell, Reis, &
other similarity. These differences were exacerbated by negative
Shaver, 1996). Individuals with a secure style showed high levels
affect and reduced by positive affect.
of intimacy and enjoyment, lower levels of negative emotions, and
Relational behaviors. The emergent literature of the past de- greater efforts to promote social interaction, especially interactions
cade also reveals common relational behaviors characteristic of with members of the opposite sex. In addition, these individuals
individuals differing in attachment strategies. These findings fur- showed more clearly differentiated activity patterns in their inter-
ther contribute to an integrated portrait of the healthy and effective actions with romantic and other opposite-sex partners (Tidwell et
self by supporting Bowlby's contention that adaptive self-reliance al., 1996). Not surprisingly, in another study in which participants
is both the product and the most noteworthy expression of intimate were asked to describe the attachment styles of both their actual
relationships characterized by free, openly negotiated, and mutu- and "ideal" romantic partners, all participants—including all types
ally interdependent communication. For Bowlby, adult psycholog- of insecure individuals—preferred ideal partners who were secure
ical health and effectiveness is largely defined by one's capacity to (Latty-Mann & Davis, 1996). Hypothetical ideal partners por-
co-construct and collaboratively maintain a web of secure, inter- trayed as avoidant (measured differently but closely akin to Bar-
dependent relationships with others. tholomew's "dismissing" avoidant type) or ambivalent (akin to
Bartholomew's "fearful" avoidant type) received the lowest rat-
A well-based self-reliance, we may conclude, is usually the product of
ings (Latty-Mann & Davis, 1996). Thus, when asked their prefer-
slow and unchecked growth from infancy to maturity, during which,
through interaction with trustworthy and encouraging others, a person
ences, even individuals with insecure orientations want to be with
learns how to combine trust in others with trust in himself. (Bowlby, people who have secure orientations. Given the research showing
1979, p. 125) that those with secure styles tend to pair up with each other
(Brennan & Shaver, 1995; Collins & Read, 1990), insecure part-
These striking differences in the degree to which communication is
either free or restricted are postulated to be of great relevance for ners may be less likely to maintain relationships with secure
understanding why one child develops healthily and the other be- partners.2
comes disturbed . . . For a relationship between any two individuals to From diary studies to correlational/cross-sectional studies to
proceed harmoniously each must be aware of the other's point-of- experimental studies about actual or hypothetical partners, indi-
view, his goals, feelings, and intentions, and each must so adjust his viduals using secondary attachment strategies appear to suffer in
own behavior that some alignment of goals is negotiated. (Bowlby, their relationships. For example, these individuals show low levels
1988, p. 181)
of intimacy and enjoyment, more distress, and make fewer efforts
to interact with potential romantic partners (Tidwell et al., 1996).
In essence then, attachment theory views healthy personality
The literature that follows illuminates some of the interpersonal
development in the context of an unfolding relational journey that
dynamics associated with these problematic outcomes.
requires the person to navigate a course between the opposing
Brennan and Bosson (1998), for example, found that individuals
developmental pressures toward differentiation (i.e., self-
who use various secondary attachment strategies differ in the
definition, or separateness from others) and relatedness, or con-
extent to which they value partner feedback; they also differ in
nectedness to others (Blatt, 1990; Guisinger & Blatt, 1994).
Bowlby (1988) characterized this dilemma in terms of behavior
systems: the exploration behavior system and the attachment be- 2
Again, less is known about the two kinds of avoidants' actual partner
havior system. In general, insecure attachment histories and inter- pairings—the research by Brennan and Shaver (1995) and Collins and
nal working models are presumed to bias relationship functioning Read (1990) was conducted before the distinction was made between the
(and thereby personality development) toward either an overem- two kinds of avoidants.
 290 LOPEZ AND BRENNAN

their typical reactions to such feedback. Compared with those with likely perpetrator of violence was the anxious or hyperactivating
insecure styles, those with a secure style tend to value and rely on partner who may use physical aggression out of sheer, misguided
feedback from romantic partners, react positively to such feed- desperation. Similarly, preoccupied and fearfully attached persons
back, and feel that they are truly known by their partners. In other (both of whom rely on the hyperactivating strategy) are most likely
words, secure individuals appear to possess a more open working to report the occurrence of aggression in their premarital dating
model with regard to receiving self-relevant partner feedback relationships (Bookwala & Zdaniuk, 1998).
compared with their insecure counterparts. Although insecure, In laboratory research, couples composed of secure partners are
individuals with a preoccupied orientation appear to value feed- more likely to engage in constructive problem solving, to rate their
back and report that they desire feedback from their partners, but discussions positively, and to report positive changes in their
these individuals report that they react negatively to such feedback. perceptions of their partners (e.g., Simpson, Rholes, & Phillips,
Individuals with a dismissing orientation display a studied indif- 1996). Attachment security has been associated with more favor-
ference to partner feedback and appear neither to value nor seek able expectations of partners' motives and behaviors during an in
such feedback. Those displaying a fearful orientation display a mix vivo role play of a potentially anxiety-provoking "distancing"
of preoccupied and dismissing styles—desiring feedback but re- interaction task (J. A. Feeney, 1998). B. C. Feeney and Kirkpatrick
acting with both anger and indifference to such feedback. Finally, (1996) examined the effects of the presence or absence of a
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

individuals with secondary attachment strategies reported feeling romantic partner on college women's physiological responses to a
This document is copyrighted by the American Psychological Association or one of its allied publishers.

that their partners did not truly know them, compared with indi- stressful laboratory situation (involving a progressively difficult
viduals using the primary (secure) strategy. mental arithmetic task). Female participants' heartrate and blood
Elsewhere, more controlled studies have assessed the impact of pressure were monitored during both baseline and challenge
participants' attachment styles on their social responsiveness to phases of the stress task and while their partners were either
imagined encounters and prospects with different relationship part- present or absent. Previous research using this task shows that the
ners. In one study (Pietromonaco & Carnelley, 1994), participants presence of a (potentially evaluative) human friend is highly
varying in attachment style were asked to imagine themselves with stressful when compared with the presence of one's pet dog, and
a partner who displayed secure or insecure behavior. As expected, that such stress emerges objectively in the form of increased
those with a secure orientation displayed the most positivity re- physiological arousal (Allen, Blascovich, Tomaka, & Kelsey,
garding themselves and hypothetical partners and also reported the 1991). B. C. Feeney and Kirkpatrick found, however, that for
highest level of optimism about these relationships. secure women, neither their partner's presence nor absence was
Other studies with similar self-reporting procedures have ex- significantly associated with changes in their physiological state,
plored how couples' attachment models affect their reported and in whereas among insecure women, initial partner absence was re-
vivo management of conflict and affective arousal. These studies lated to significant increases in heartrate and blood pressure, and
lend additional support to the inference that important attachment that this increased physiological activity persisted even after their
style-related variations in how participants internally regulate their partner was subsequently introduced into the experimental context.
cognitive and affective states extends to how they interpersonally This pattern of results suggests that, under stress, persons with an
manage their distress. For instance, J. A. Feeney (1995) found that insecure style are more physiologically reactive to the initial
couples in which both partners endorsed secure attachment styles absence of social support.
reported less emotional suppression than did couples with two Summary and critique. The emergent literature on adult at-
insecure partners. More specifically, attachment-related anxiety tachment dynamics supports the view that a variety of healthy
was particularly associated with the inhibited expression of anger personality characteristics and positive adjustment outcomes are
among men. Anxiety was also associated with a higher likelihood associated with the use of the primary strategy for seeking and
of perceiving partners as inhibiting their (the participants') anger maintaining attachments with intimate others over time. Persons
and sadness while concurrently controlling their own sadness. In who use the primary attachment strategy not only report more
sum, these findings suggested that insecure attachment is linked favorable and coherently organized developmental histories, they
with idealized beliefs that relationship partners should project a also evidence more flexible and benign cognitive processes, more
happy and untroubled image and that the expression of sadness mature forms of affect regulation, and more sensitive and collab-
should be controlled. Elsewhere, Young and Acitelli (1998) found orative relational skills—all competencies which presumably en-
that among couples in married or exclusive dating relationships, hance adaptation to most life circumstances. This literature is also
both men and women with secure attachment styles held less consistent with the view that persons using secondary attachment
idealized images of their partners. Regardless of their marital/ strategies (i.e., those who experience the hyperactivation or deac-
dating status, men who classified themselves as anxious paired tivation of the attachment system) evidence notable impairments in
with women who classified themselves as avoidant appeared to one or more of these dynamic competencies, rendering these
have the most idealized perceptions of their mates. (Again, it is no individuals more vulnerable to affective distress, relational diffi-
wonder that individuals with insecure styles report feeling es- culties, and adjustment failures.
tranged from their partners.) Compared with participants with An ongoing problem with the literature just reviewed, however,
insecure styles, couple participants with a secure style have also is a lack of consistency with regard to the measurement of attach-
been found to make greater use of "accommodation" strategies ment. The number and names for attachment categories differ, as
(i.e., open discussion, patience with partner) over an 8-month does the use of categories versus dimensions; self-report methods
period (Scharfe & Bartholomew, 1995). Roberts and Noller (1998) dominate; and attachment is sometimes measured with regard to
observed that, among couples in which partners used opposing peers, sometimes partners, sometimes parents. In addition, despite
secondary strategies (hyperactivating vs. deactivating), the more evidence that adults use multiple models of attachment (e.g.,
 DYNAMIC PROCESSES 291

Baldwin et al., 1996; Collins & Read, 1994), most of the literature inclusive framework is essential to integrating personality and
has conceptualized and assessed attachment as a singular model. developmental themes within a conceptualization of optimal psy-
Given that attachment may show different properties depending on chological health and effectiveness. Furthermore, most prevailing
the target of attachment, this issue becomes even more complex. If views overemphasize individual characteristics at the expense of a
there are multiple models of attachment, then researchers must find broader, relational, and contextualized view of human effectance
out whether these models are organized hierarchically (Collins & and adaptation. Other perspectives on the healthy and effective self
Read, 1994) or by association^ networks (Baldwin et al., 1996). that limit their view to relatively domain-specific operations such
Also related to this issue is whether adult attachment characteris- as cognitive appraisals, coping strategies, or social competencies
tics are better understood as properties of individuals or of rela- but otherwise fall short of illuminating how these processes inter-
tionships (Kobak, 1994). Although we reviewed studies that ex- act to promote higher order forms of self-differentiation, integra-
amined the impacts of particular adult attachment style pairings on tion, and adaptation. Still other theorists fashion their conceptual-
relationship functioning, it is plausible that certain emergent and izations of the healthy, effective self using vague or esoteric
dynamic properties of relationships are responsible for attachment language (e.g., "self-actualization") that confounds a more
effects misattributed as intrapersonal characteristics. In short, the grounded understanding of how psychological coherence is
adoption of a common method of conceptualizing and assessing achieved, maintained, and elaborated over the life course. Future
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

attachment is needed to facilitate comparison of findings. We tried, research on the elusive nature of the healthy, effective self is more
This document is copyrighted by the American Psychological Association or one of its allied publishers.

where possible, to help readers think about the dimensions under- likely to benefit from frameworks capable of generating operation-
lying the attachment categories, by drawing attention to the un- ally definable constructs and testable hypotheses. Finally, many
derlying dynamics of the attachment system. Finally, this line of prevailing views on the healthy, effective self have at best indirect
inquiry has been limited by an overreliance on cross-sectional or oblique implications for therapeutic intervention and change.
designs (cf. Davila, Burge, & Hammen, 1997), underscoring a We submit that, to be maximally useful to counseling psycholo-
need for carefully conducted longitudinal studies capable of more gists, emergent views on optimal psychological functioning must
sensitively assessing the impacts of adult attachment organization have direct linkages to therapy-relevant processes and outcomes.
on developmental processes and outcomes.
For instance, Kobasa's construct of psychological "hardiness"
Despite these shortcomings, our literature review confirms that (Kobasa, 1982) restricts its attention to individual characteristics of
attachment insecurity is very much associated with three general (perceived) control, commitment, and challenge without consider-
problems: (a) an initially biased assessment of the interpersonal ing larger relational and developmental factors that give rise to
situation (e.g., misperceiving a partner's interest, loyalty, and these characteristics or that affect their display and adjustment
responsiveness); (b) either heightened or suppressed negative af- consequences. As a result, this view of the healthy and effective
fect in response to perceived threat (i.e., the anxiety dimension); self is descriptive and largely atheoretical. Moreover, it provides
and (c) behaviors designed to approach or avoid partners to man- little guidance in formulating interventions for promoting hardi-
age the stress (i.e., the avoidance dimension). The individual with ness. Ascribing the essence of psychological health to a limited set
a healthy personality could be described as avoiding the pitfalls of self-perceptions that have unclear developmental antecedents
associated with hyperactivated or deactivated response patterns, and that presumably yield consistent and unilateral environmental
either because of a relatively unscathed relational history or else by impacts offers a minimalist and static account of human growth,
having moved toward the primary strategy as a result of therapeu- change, and adaptation.
tic or other interventions. More than just avoiding these problems, Although more contemporary conceptualizations of the healthy,
however, we believe that the accumulating research portrays a effective self such as Seligman's theory of learned optimism
dynamic, proactive view of the healthy, effective self—one that (Seligman, 1990) and Bandura's theory of self-efficacy (Bandura,
entails a sense of (accurately perceived) self-efficacy and social 1986, 1997) consider the reciprocal interactions of agentic person
support, resilience and hardiness in the face of stress, and comfort characteristics with dynamic conditions in the person's social
in seeking support when under stress. These ideas are explored context, they are handicapped in other respects. For example,
next. despite the fact that Seligman's framework emphasizes the role of
parental responsiveness as a critical developmental antecedent of
optimism and the function of optimism as an adaptive "explana-
Part 2: Toward an Attachment-Theoretical Perspective on tory style," it offers only a minimally detailed analysis of how this
the Healthy and Effective Self dispositional style emerges in the context of close relationships or
how it interacts with the fluid nature of these social environments
Limitations of Traditional Models of Optimal
to promote higher levels of psychological competence over time.
Psychological Health and Effectiveness
In a similar fashion, whereas Bandura (1986) acknowledges that
As we noted in our introduction, over the years a number of "social environments may place constraints on what people do or
independent theorists and investigators have offered conceptual- may aid them to behave optimally (p. 437)," his theory clearly
izations of the healthy and effective self. Although these perspec- emphasizes a conscious, cognitive-mediational analysis that is
tives yield several convergent observations and inferences, each individually focused, with limited ability to address the intersub-
appears restricted in ways that limit a deeper, richer, and more jective nature of human competence and adaptation. As Lent and
integrative understanding of optimal human functioning and de- Lopez (2000) have noted, most of the research extensions of
velopment. For example, none of them provides a biologically self-efficacy theory to domains of relevance to counseling psy-
based, lifespan perspective that concurrently speaks to the origins chology (e.g., career development) have generally studied self-
and development of the healthy, effective self over time. Such an efficacy/criterion relations at the level of the individual, either
 292 LOPEZ AND BRENNAN

holding constant or without considering the input and reactions of effective self. To our knowledge, no other theory embraces as
other persons. Clearly, the quality of intimate human relationships, wide a range of personal factors (e.g., cognitions, affects,
as well as their potentialities as a medium of psychological growth attentional processes, and motivational goals) and relationship
and adaptation, are not unilaterally determined by the beliefs or dynamics (e.g., caregiving, perspective taking, collaboration,
actions of a single participant. Rather, the power of close relation- support seeking, and distancing) within its intellectual scope.
ships to shape the life course probabilistically emerges from a By emphasizing the centrality of one's ongoing participation in
complex interplay of the cognitive, affective, and relational behav- close relationships, the theory helps explain how human beings,
ior of the emotionally joined participants. as a consequence of their relational histories and expectancies,
Finally, whereas existential and humanistic perspectives on the typically adopt one of a limited array of strategies for managing
healthy, effective self (Maslow, 1970; Rogers, 1962) place con- the experience of insecurity. Beginning with one's earliest
siderably greater emphasis on the search for meaning, the achieve- interactions with primary caregivers, the experience of attach-
ment of accurate, congruent self-experiencing and authentic con- ment security enables the development and deployment of
nection with others, and the power of intimate relationships as a adaptive cognitive processing, affect regulation, and relation-
source of personal growth, the optimal self in these models is ship problem-solving processes. For example, Chorpita and
presumed to grow out of as opposed to being a continuing part of Barlow (1998) have proposed that attachment theory can func-
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

a dynamic relational context. Such perspectives once again privi- tion as a conceptual "bridge" linking early relational experi-
This document is copyrighted by the American Psychological Association or one of its allied publishers.

lege a person-centered view of autonomy, while subtly devaluing ences with the development of a cognitive style regarding
the bidirectional processes within close relationships that promote security and control of one's affective states. These writers note
increasingly complex states of interdependence. In our view, mod- that
els privileging either the person, or the environment in which he or
she is embedded, fall short of achieving a comprehensive, rela- an important connection can be drawn between the reciprocal social
tionally sensitive account of the healthy, effective self. influences inherent in secure attachment and the notions of consis-
tency and autonomy valued in the literature on the development of
control cognitions. (Chorpita & Barlow, 1998, p. 11)
Bridging Personality and Developmental Themes:
The Promise of an Attachment Theory Perspective It thus appears that the primary attachment strategy equips one
to construct relationships that maximize positive engagements and
In light of these limitations, we believe that attachment theory appropriate social investments. By contrast, as mentioned, the
offers an integrative perspective on the healthy and effective self, early experience of attachment insecurity deploys a pattern of
one capable of illuminating (in more pragmatic terms) the inter- cognitive processing, affect regulation, and relational behaviors
play of cognitive, affective, and relational behavior processes. We consistent with either a hyperactivated or deactivated attachment
believe this framework not only offers a functional historical system. Taken together, these secondary strategies are more fre-
account of the development of the healthy, effective self; more quently associated with distress and problematic engagements with
importantly, it enhances our appreciation of the interdependent others. As such, both primary and secondary strategies respec-
connections of personality and developmental themes in the social tively shape personality and development by orienting the individ-
construction of an increasingly well-adapted life course. We pre- ual to be differentially responsive (and reactive) to different types
cede these arguments with assumptions regarding the relationship of stress, different forms of performance feedback, and to be
of personality and development that are consistent with theory and differentially inclined to seek and construct relationships with
research on adult attachment. others.
To begin, we propose that personality and development are Of course, any number of life changes can provoke the
inextricably related constructs. We come to know personality only experience of insecurity and mobilize the person's coping ef-
insofar as we (as observers) can discern recurring, repetitive fea- forts. What then would be indicative of an optimal response
tures of the person's relationships with others. Similarly, we come pattern characteristic of a healthy, effective self? In addressing
to apprehend development when we can discern some favorable this question, we use attachment theory to holistically consider
change in how the person defines self (i.e., self-understandings) in the operation of cognitive, affect regulation, and interpersonal
the context of his or her ongoing relationships with others. In other processes. From this perspective, optimal health and effective-
words, personality and development are interdependent constructs ness is an inherently social construct that, at its core, is best
embedded in temporal and social contexts. For these reasons, we represented as a secure adult attachment orientation. An optimal
hold that concepts such as "adjustment" are less-than-adequate response pattern is characterized by our capacities to organize
terms for describing healthy, effective personality formation and our internal and relational coping processes in ways that facil-
development. In our view, optimal development is not defined by itate agentic problem solving while, at the same time, main-
the quality of a particular adjustment but rather by a higher order taining positive, enduring, and mutually satisfying relation-
orientation toward the management of change and uncertainty ships. An optimal response pattern also embodies a capacity to
(Marris, 1996). In a similar vein, the healthy self is not defined by reflect on the impacts of our actions on significant others, and,
a fixed configuration of traits but by a flexible patterning of if necessary, to make context-appropriate self-corrections in
competencies that promote consistent and constructive engage- order to maintain these relationships. In essence, an attachment-
ments with others. theoretical perspective urges us to conceptualize the healthy,
Given the interdependencies of personality and development, effective self as an evolving self-context relation that facilitates
we believe that attachment theory is the best available frame- favorable self-organization and reorganization processes. Opti-
work for achieving an integrative understanding of the healthy, mally adapted and effective adults are proactive, fully invested
 DYNAMIC PROCESSES 293
participants in their social worlds. They are capable of both Part 3: Adult Attachment Organization and the
giving and receiving care. They know themselves and others, Counseling Relationship—Implications for
and they allow themselves to be known by others, in deep, rich, Understanding Therapeutic Movement
and authentic ways.
Toward the Healthy and Effective Self
The accumulated research supports the idea that the consis-
tent operation of the primary attachment strategy is a hallmark Although the bulk of research on adult attachment has used
of the healthy, effective self. Furthermore, the research suggests nonclinical samples or else not addressed clinicallyrelevantissues
how the nature of one's adult attachment organization speaks and processes, a recently emerging body of literature is extending
not only to a basic human motivation to achieve "belonging- the theory toresearchand practice in counseling. We believe that
ness" (Baumeister & Leary, 1995) but also to issues of change this line of inquiry, by demonstrating the applicability of theoret-
and plasticity in human adaptation. In contrast to persons using ical concepts and assumptions to the study of client and counselor
secondary strategies, secure adults are more tolerant of unpre- characteristics, counselor-client interactions, and therapeutic
dictability and ambiguity and less inclined to form hasty judg- change, further underscores the promise of attachment theory as an
ments (Mikulincer, 1997); they are also more competent in integrative framework for understanding the healthy, effective self.
In severalrespects,this nascent literature has corroborated findings
processing expectation-incongruent relationship experiences
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

from the nonclinical literature by affirming that persons who use

(Mikulincer & Arad, 1999). By the same token, secure adults
This document is copyrighted by the American Psychological Association or one of its allied publishers.

secondary strategies are overrepresented in clinical samples and

are more likely than their insecure peers to entertain self-
more at risk for psychological disturbances. Such disturbances
discrepant feedback from their intimate partners (Brennan &
include depression (Carnelley, Pietromonaco, & Jaffe, 1994;
Bosson, 1998), engage in accommodative and collaborative Zuroff & Fitzpatrick, 1995), personality disorders (Brennan &
problem solving (Lopez, Gover, et al., 1997; Scharfe & Bar- Shaver, 1998), and interpersonal problems (Horowitz, Rosenberg,
tholomew, 1995), recognize the need to renegotiate their rela- & Bartholomew, 1993). More importantly, persons disposed to-
tionships when necessary (J. A. Feeney, 1998), and form accu- ward hyperactivating or deactivating strategies appear differen-
rate perceptions of others (Kobak & Hazan, 1991). In short, the tially oriented to seeking therapeutic help (Dozier, 1990; Horowitz
primary attachment strategy optimizes health and effectiveness et al., 1993; Lopez, Melendez, Sauer, Berger, & Wyssmann,
by means of cognitive-affective dynamics that keep the person 1998), are likely to display different levels of distress (Kemp &
behaviorally engaged with "minding" tasks deemed crucial to Neimeyer, 1999) and affectivity (Searle & Meara, 1998), and, once
the long-term maintenance of healthy, intimate relationships engaged in counseling, may pose distinct challenges to the forma-
(Harvey & Omarzu, 1997). Taken together, these diverse yet tion of effective therapeutic alliances (Kivlighan, Patton, & Foote,
integrated competencies embodied within the primary attach- 1998; Mallinckrodt, Coble, & Gantt, 1995; Mallinckrodt, Gantt, &
ment strategy may help explain the resilient adaptation of Coble, 1995; Mallinckrodt, King, & Coble, 1998). In this conclud-
secure adults over time to a wide variety of life disruptions and ing section, we review this literature and call attention to several
transitions (Birnbaum et al., 1997; Cafferty et al., 1994; Fraley unaddressed questions. We begin, however, by considering some
& Shaver, 1998; Mayseless et al., 1996; Pistole, 1996). of Bowlby's ideasregardingthe application of attachment theory
to the therapeutic context.
The available research further suggests that these competencies
are an outward expression of a coherent and flexible internal
process for accessing and managing attachment-related cognition An Attachment-Theoretical Perspective on Counseling
and affects. By using the primary strategy, the healthy, effective
person is better able to retrieve, differentiate, and organize Despite the fact that much of Bowlby's extensive writing fo-
attachment-related images, affects, and memories (Lopez, Gover, cused on the early childhood antecedents of later psychopathology,
et al., 1997; Mikulincer, 1998c), even painful ones, and is thus he was keenly interested in the potential applications of his ideas
to clinical practice and to the formulation of a theoretically
better equipped to integrate prior learning and immediate affective
grounded account of therapeutic change. Indeed, his final book on
states into his or her cognitive processing. This more complex and
attachment theory was largely devoted to these goals (Bowlby,
"self-reflective" information processing (Fonagy et al., 1995,
1988). For Bowlby, successful therapeutic relationships not only
1996) likely contributes to the development of relatively stable, yet
activate the client's attachment system, they concurrently promote
increasingly differentiated, less distorted, and more coherent as- insight into the client's attachment organization and strategies,
sessments of self and others (Kobak & Hazan, 1991; Levy et al., including their origins and problematic functions. Throughout this
1998; Mikulincer, 1995; Young & Acitelli, 1998). The secure, process, the overarching task of the therapist is to provide a secure
healthy, and effective self seemingly possesses a "learning to base for exploring and recovering the client's lost attachment-
learn" orientation to life and living: He or she has a higher order related thoughts and feelings.
perspective that encourages autonomous learning and coherence
seeking amidst life events and forces that instigate change, disor- The process of therapeutic change . . . can then be understood as due
der, and uncertainty. to the patient, thanks to the relatively secure base provided by the
analyst, developing sufficient courage to permit some of the informa-
In our closing section, we examine one final advantage of an
tion hitherto excluded to go forward for processing. This included
attachment-theoretical perspective on the healthy and effective both information stemming from the present situation, for example,
self: its ability to speak more directly to therapy-relevant processes evidence of the analyst's genuine concern to help the patient and the
and outcomes and to provide conceptual guidance regarding the conflicting thoughts, feelings, and behavior that arouses, and also
complex process of therapeutic change. information stored in memory, for example, memories of the very
 294 LOPEZ AND BRENNAN

painful events of childhood and the thoughts, feelings, and behaviors the general population (Dolan et al., 1993; Mickelson, Kessler, &
aroused by them. (Bowlby, 1988, p. 71) Shaver, 1997), and that persons with preoccupied and fearful
attachment styles are especially prone to report high levels of
True to his relationship-centered view of healthy personality symptomatic functioning (Carnelley et al., 1994; Feeney & Ryan,
development, Bowlby viewed the client-therapist relationship as 1994; Kemp & Neimeyer, 1999; Lopez et al., 1998). This pattern
the medium of change. By functioning as a secure base, the is not surprising in light of the assumption that these individuals
competent therapist invites the client to enter into a safe, trusting are assumed to have internalized negative self-models. Persons
relationship capable of tolerating the client's maladaptive strate- with a fearful attachment style, in particular, are likely to endorse
gies for managing insecurity while promoting the client's reflec- "self-critical" forms of depression (Zuroff & Fitzpatrick, 1995).
tion on the unmet emotional needs and cognitive biases from Certainly, a key question in using attachment theory to under-
which these behaviors derive. stand adult psychopathology is whether or not attachment-style
More contemporary writers have elaborated these ideas (Birin- differences are unique to more fundamental aspects of personality
gen, 1994; Rutter, 1995), whereas others have offered clinical and personality disorders (Bartholomew & Horowitz, 1991; Bren-
insights and case examples regarding how therapists appropriately nan & Shaver, 1998; Carver, 1997; Shaver & Brennan, 1992). The
function as autonomy-supporting caregivers (Pistole, 1999) who evidence suggests that, although interconstruct overlap exists,
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

can adopt appropriate "interpersonal stances" that complement knowledge of a person's strategy for managing closeness and
This document is copyrighted by the American Psychological Association or one of its allied publishers.

their client's attachment style (Dolan, Arnkoff, & Glass, 1993; distance may provide a particularly sensitive means of understand-
Farber, Lippert, & Nevas, 1995). For example, Farber et al. argued ing variations in presenting problems.
that attachment theory-guided therapeutic work "must focus on
Using the NEO Personality Inventory (Costa & McCrae, 1985),
creating a containing environment wherein the patient experiences
Shaver and Brennan (1992) found that insecure attachment styles
the constancy, consistency, and boundedness of the therapist and
were positively associated with neuroticism, whereas the secure
the therapeutic frame" (p. 209). Still others have considered how
attachment styles were negatively associated with neuroticism and
attachment theory can be extended to specific therapeutic contexts,
such as family counseling (Lopez, 1995b), group work (Pistole, positively associated with extraversion. More recently, these same
1997), brief psychotherapy (West, Sheldon, & Reiffer, 1989), authors explored whether adult attachment styles and personality
counselor training and supervision (Pistole & Watkins, 1995; disorders share a common underlying structure (Brennan &
Watkins, 1995), and the treatment of special populations, such as Shaver, 1998). They found that securely attached participants
the adult survivors of incest (Alexander & Anderson, 1994) and scored significantly lower than their insecurely attached peers on
persons with borderline personality disorders (Sable, 1997). A virtually every personality disorder scale, underscoring once again
common theme across these therapeutic extensions of attachment the adaptational superiority of the primary attachment strategy. For
theory is the theory's utility for conceptualizing the developmental example, important variations in the deployment of hyperactivat-
antecedents and interpersonal features of client problems, as well ing and deactivating strategies demonstrated noteworthy associa-
as their likely impacts on help seeking and on the formation of the tions with different personality disorders: Persons with preoccu-
therapeutic relationship. For example, Lopez (1995b) proposed pied styles scored higher on the Dependent and Historionic scales
that attachment theory can speak to such systemic family patterns (i.e., scales that are both indicative of hyperactivation) and lower
as intergenerational problems, affect regulation within the marital on the Schizoid scale (i.e., a scale indicative of deactivation)
dyad, dysfunctional family alignments, and problematic family relative to persons with the dismissing style. Individuals with a
responses to life cycle transitions. fearful attachment style scored high on a number of personality
disorders indicative of both attachment hyperactivation (Depen-
To be sure, the counseling relationship is generally regarded as
dent, Histrionic) and deactivation (Schizoid), disordered thinking
an intimate interpersonal context, one premised on remediating
patterns (Schizotypal, Paranoid, and Obsessive-Compulsive), and
problematic client behavior and restoring the client to a state of
scales representing admixtures of symptoms (Avoidant, Passive-
healthy, adaptive functioning. It is also a context for exploring the
client's relationship dynamics both within and outside the thera- Aggressive, Self-Defeating, and Borderline).
peutic dyad, as well as themes pertaining to experiences of loss, These findings also are corroborated by earlier studies linking
separation, and reunion (Slade, 1999). We contend that many attachment-related strategies toward hyperactivation or deactiva-
forms of presenting problems can be viewed as expressions and tion to different types of interpersonal problems. For example,
consequences of the clients' use of secondary attachment strategies using the Inventory of Interpersonal Problems, Bartholomew and
for managing insecurity. In a parallel fashion, many forms of Horowitz (1991) found that each insecure adult attachment style
therapeutic intervention can be viewed as efforts to minimize the was associated with a distinct pattern of interpersonal difficulty:
negative consequences of secondary attachment strategies (toward Persons with a fearful attachment style were more likely to report
either hyperactivation or deactivation) and to promote the use of problems of passivity and social inhibition, whereas those with a
the primary attachment strategy for solving problems and coping dismissing style were more likely to report problems related to a
with stress. lack of warmth in social interactions. Interestingly, persons with a
preoccupied style endorsed items such as "I want to be noticed too
much" and "it is hard for me to stay out of other people's busi-
Empirical Findings
ness," suggesting that although they are highly dependent on
Personality characteristics and client problems. There is ev- others to maintain self-esteem, they pursue this goal using a
idence that individuals with insecure adult attachment styles are controlling and dominating interactional style. A subsequent study
overrepresented in clinical samples relative to their base rates in largely replicated these findings within both an undergraduate
 DYNAMIC PROCESSES 295
sample and in a clinical sample receiving brief psychotherapy was related to stronger working alliances. In addition, comfort
(Horowitz et al., 1993). with closeness interacted with counselor experience level in pre-
More recently, Searle and Meara (1998) examined relationships dicting alliance scores: Among clients who were less comfortable
among adult attachment styles, gender, and several aspects of with intimacy, more experienced counselors (who presumably
subjective experience pertinent to the counseling situation: atten- were more adept at working with these reluctant clients) received
tion to emotion, emotional intensity, emotional expressivity, and more favorable alliance ratings (Kivlighan et al., 1998).
"intentness," defined as a combination of attention and intensity. Continued progress along this line of inquiry should be accel-
Their findings indicated that participants with a preoccupied at- erated by the recent construction and validation of the Client
tachment style displayed labile emotions and expansive emotional Attachment to Therapist Scale (CATS; Mallinckrodt, Gantt, &
self-presentations—a pattern diametrically opposite to that ob- Coble, 1995). Developed on a clinical sample, the CATS contains
served in respondents with a dismissing style. Persons with a three factor-analytically derived subscales (i.e.. Secure, Avoidant/
fearful style, by contrast, demonstrated a more mixed emotional
Fearful, and Preoccupied Merger) that reflect distinct client attach-
picture—exhibiting both features of emotional intensity along
ment orientations toward the therapist. Cluster analysis was used to
with restricted emotional expressivity. Similarly, Kemp and Nei-
identify four clusters of clients (i.e., secure, merged, avoidant, and
meyer (1999) found that, relative to their secure peers, individuals
reluctant) that appear to correspond with the secure, preoccupied,
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

with a preoccupied attachment style evidenced higher levels of

dismissing, and fearful styles identified by Bartholomew and
This document is copyrighted by the American Psychological Association or one of its allied publishers.

intrusive psychological symptoms and overall psychological dis-

Horowitz (1991). Mallinckrodt et al. found that the CATS sub-
tress. Persons with hyperactivating strategies are apparently more
scales correlated in expected directions with measures of object
likely to display distress in the counseling situation. Moreover,
relations, social self-efficacy, adult attachment, and client-rated
persons using either hyperactivating or deactivating strategies are
likely to exhibit immature ego defenses such as denial and pro- working alliance. For example, clients scoring highly on the Pre-
jection (Diehl, Elnick, Bourbeau, & Labouvie-Vief, 1998). These occupied Merger subscale (and by inference those using hyperac-
results suggest that knowledge of clients' attachment strategies tivating strategies) exhibited a number of object relations deficits,
may be informative of both the developmental antecedents and the and, although motivated to form early alliances with their coun-
specific interpersonal manifestations of more basic personality selors, sought to dissolve the "normal boundaries" of the counsel-
characteristics; in addition, knowledge of clients' attachment strat- ing relationship. By contrast, clients who scored high on the
egies may augment standard personality assessment by providing Avoidant-Fearful subscale (which included clusters of fearful and
a more elaborated view of clients' interpersonal problems. reluctant persons, both of whom presumably use deactivating
strategies) tended to distrust their therapists, feared rejection, and
Orientations toward counseling. Several studies have exam- exhibited deficits in social competence, particularly in their ability
ined relations among adult attachment styles, help seeking, and the to identify their feelings (Mallinckrodt, King, & Coble, 1998).
early formation of the therapeutic working alliance. There is evi-
dence that persons with deactivating strategies are least disposed to Counselor attachment orientation. It is intuitively reasonable
acknowledge personal vulnerability and to seek therapeutic assis- that the formation of therapeutic alliances reflects the joint contri-
tance. For example, among clients with serious disorders, adults butions of clients' and counselors' strategies for regulating inse-
with dismissing styles were more likely to reject treatment pro- curity. To date, however, few studies have examined the role of
viders, engage in less self-disclosure, and make poorer use of brief counselors' attachment styles and orientations on the therapeutic
treatments (Dozier, 1990; Horowitz et al., 1993). Lopez et al. process. Dozier, Cue, and Barnett (1994) found that securely
(1998) found that students' attachment styles interacted with prob- attached clinicians provided therapy experiences that challenged
lem levels to predict willingness to seek counseling: Among stu- their clients' expectations of the world, whereas the interventions
dents reporting high problem levels, those with positive other- of insecure clinicians provided reciprocal feedback that did not
models (i.e., secure and preoccupied attachment styles) were more incorporate such challenges. In another study, therapists who re-
willing to seek counseling, whereas those with negative other- ported greater comfort with closeness were more likely to have
models (i.e., dismissing and fearful attachment styles) were least clients who reported a strong, early emotional bond with them
likely to seek counseling. In short, the use of deactivating strate- (Dunkle & Friedlander, 1996). Sauer (1999) recently examined the
gies appears to orient the person away from help seeking. contributions of both counselor and client attachment orientations
Mallinckrodt, Coble, and Gantt (1995) found that clients' child- to the formation of the working alliance over seven counseling
hood bonds with parents and their current adult attachment orien- sessions. Twenty-eight counselor-client dyads participated in this
tations both uniquely and significantly predicted working alliance study, although only 17 dyads continued to the final assessment
scores (controlling for the length of treatment). This pattern was point. Of this group, 7 dyads were composed of secure counselors
particularly true for attachment-related anxiety, or hyperactivation: and insecure clients, whereas another 7 dyads were represented by
Clients who were less anxious and fearful of abandonment were insecure counselors and insecure clients. Sauer specifically exam-
more likely to form positive therapeutic alliances. Satterfield and ined change patterns in clients' ratings of the working alliance of
Lyddon (1995) also reported associations between attachment these 2 dyad groups. Interestingly, the clients of secure counselors
styles and ratings of the working alliance among first-time clients reported weaker alliances immediately after the first therapy ses-
seeking services at a university counseling clinic. In their study, sion but thereafter acknowledged strong alliances; the clients of
however, clients expressed that "trust in the availability and de- insecure therapists, by comparison, reported positive but virtually
pendability of others" was the most prominent correlate of work- unchanging working alliance ratings over the three assessment
ing alliance scores. In a sample of 40 counselor-client dyads, periods. Following the findings of Dozier et al. (1994), one plau-
Kivlighan et al., (1998) found that client "comfort with closeness" sible interpretation of this finding is that secure counselors may be
 296 LOPEZ AND BRENNAN

perceived by their clients as initially more challenging and less tolerate social proximity, and engage more fully in healthy social
accommodating in establishing the therapeutic contract. investments.

Unaddressed Questions Conclusions

Although the literature yields preliminary support for the value Several years ago, Lopez (1995a) proposed that attachment
of adopting an attachment theory perspective on the counseling theory might serve as an integrative "metaperspective" for re-
relationship, a number of important questions remain unaddressed. search and practice in counseling psychology. Since his claim, this
First, are therapeutic interventions based on attachment theory theory has guided a steady and impressive current of research on
effective? Second, do attachment-theory based interventions pro- adult attachment. Our review of much of this research particularly
mote, as Bowlby suggests, revision of clients' working models and affirms the theory's promise for advancing our understanding of
strategies for managing insecurity? We could find only one inter- the healthy, effective self, although admittedly, more work needs
vention study explicitly using attachment theory in an adult sample to be done. To date, social and developmental psychologists have
(Kilmann et al., 1999). In this study, women identified as at risk generated most of the empirical research on adult attachment.
for relationship distress were randomly assigned to a control Although valuable, this work has generally focused on single
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

condition or to a 3-day attachment-focused group intervention. dimensions or limited ranges of human functioning (e.g., affect
This document is copyrighted by the American Psychological Association or one of its allied publishers.

This intervention was composed of psychoeducational modules regulation, relationship perceptions, etc.). As a result, the power of
addressing dysfunctional relationship beliefs, childhood factors attachment theory to address broader, cross-situational aspects of
influencing partner choices and relationship styles, relationship functioning characteristic of optimally effective people remains
skills, and relationship strategies. In spite of low power sufficient largely untapped. Furthermore, direct empirical extension of the
to detect only large effects, at a 6-month follow-up intervention, theory to the study of counseling processes has only recently been
group participants reported improved interpersonal styles, en- launched, and more theory-grounded intervention studies are
hanced satisfaction with family relationships, and a greater reduc- needed to evaluate the theory's potential as a guide for counseling
tion in unrealistic relationship beliefs and interpersonal problems practice. Nonetheless, we are heartened that several counseling
relative to their peers in the control condition. However, measures psychologists are leading this line of inquiry, and we are hopeful
of adult attachment orientation were not readministered at fol- that our review will stimulate others to join this endeavor.
low-up. Although the Kilmann et al. study represents a promising
Attachment theory offers a rich life span framework for under-
beginning, more work must be done to develop and evaluate other
standing how the activation and operation of a basic and unique
attachment theory-based interventions, particularly ones with a
motivational system sets into motion an integrated pattern of
more experiential skills-based component and a stronger client-
human responses capable of shaping the life course in favorable or
therapist relationship (Brennan, 1999).
unfavorable ways. Indeed, we believe that variations in adult
Third, are there client-by-treatment interactions? For example, attachment organization may represent a higher order class of
are clients who use hyperactivating versus deactivating strategies individual differences that more proximally affect the experience
differentially responsive to different forms of intervention? Cer- of stability, continuity, and coherence in human lives. Counseling
tainly, clients adopting deactivating strategies, given their height- psychologists are in a unique position to advance this line of
ened distrust of others, avoidance of intimacy, and social skills inquiry by using the theory to probe a wider array of human
deficits would seem poor candidates for early therapeutic inter- functions and competencies across a wider array of relationships
ventions that emphasize affective exploration and self-disclosure; and situations, including therapeutic encounters. We believe that,
instead, these individuals may respond more favorably to cogni- as this work unfolds, it will affirm Bowlby's central belief that
tively based approaches. In contrast, clients with hyperactivating optimal psychological health and development can be fully under-
strategies, given their stronger desires for intimate contact coupled stood only in the context of one's continuing attachments to others.
with their difficulties modulating their own emotional arousal and In sum, an attachment-theoretical perspective (and its attendant
self-presentation (as well as their inattentiveness to the impact of research findings) urges us to conceptualize the healthy, effective
their behavior on others) would seem better suited for relationally self as a particular self-context relation, one that optimizes both
oriented interventions capable of containing the client's spreading favorable self-organization and reorganization processes by pro-
affect while, at the same time, setting and maintaining appropriate moting continuing and creative engagement with others. In so
interpersonal boundaries. Clients' use of hyperactivating and de- doing, attachment theory implicitly offers a view of development
activating strategies also may be differentially related to that goes beyond an emphasis on adjustment to include a broader
transference-related distortions and idealizations of the therapist understanding of how attachment security links both persons and
and to distinctive reactions to alliance ruptures and therapy termi- their social environments in ways that maximize the quality of
nation. In short, clients' strategies for managing attachment secu- their reciprocal interaction. By incorporating a receptivity toward
rity in the therapeutic context, coupled with therapist responses to self-reflection, perspective taking, competent affect regulation, and
the countertransferential pulls (Dozier & Tyrell, 1998; Farber et relational competence, attachment security enables the person—in
al., 1995) of these strategic displays, may ultimately predict the the context of his or her continuing relatedness with others—to
course of therapeutic movement toward enhanced psychological socially construct developmental pathways leading toward those
health and effectiveness. Whatever the course, an attachment the- higher order forms of empowerment, autonomy, and relatedness
oretical perspective on therapeutic change would emphasize the that we associate with psychological health and effectiveness. The
target goals of helping clients regulate their affective states, ex- healthy and effective self continually seeks to achieve new inte-
amine and revise their inaccurate models of self and others, better grations and understandings (i.e., coherence) amid the welter of
 DYNAMIC PROCESSES 297

life forces that instigate change, disorder, and uncertainty in the plications for personality theory, psychopathology, and health (pp. 299-
course of daily living. 335). Chicago: University of Chicago.
In closing, we propose that the healthy and effective self, by Bookwala, J., & Zdaniuk, B. (1998). Adult attachment styles and aggres-
being embedded in secure, intimate relationships, operates at the sive behavior within dating relationships. Journal of Social and Personal
boundary of stability and change, neither rigidly maintaining its Relationships, 15, 175-190.
organization nor indiscriminately surrendering it. Rather, by Bowlby, J. (1973). Attachment and loss. Vol. 2: Separation. New York:
Basic Books.
achieving and maintaining an equilibrium between internal (self)
Bowlby, J. (1979). The making and breaking of affectional bonds. New
awareness and external (context) sensitivity, the healthy, effective
York: Routledge.
self optimizes the codevelopment of its capacities for differentia-
Bowlby, J. (1980). Attachment and loss. Vol. 3: Loss. New York: Basic
tion and relatedness (Blatt, 1990; Guisinger & Blatt, 1994). An Books.
attachment-theoretical perspective affirms this cybernetic process Bowlby, J. (1982). Attachment and loss. Vol. 1: Attachment. London:
by underscoring the dynamic power of secure working models to Tavistock. (Original work published 1969)
revise and update themselves in a manner that permits the accom- Bowlby, J. (1988). A secure base: Parent-child attachments and healthy
modation of new roles, experiences, and information while, at the human development. New York: Basic Books.
same time, preserving the "felt" experience of positive continuity Brennan, K. A. (1999). Searching for secure bases in attachment-focused
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

and self-sameness, or what Erikson (1968) referred to as "iden- group therapy: Reaction to Kilmann et al. (1999). Group Dynamics:
This document is copyrighted by the American Psychological Association or one of its allied publishers.

tity." Hence, the achievement and maintenance of attachment Theory, Research, and Practice, 3, 148-151.
security provides a fundamental basis for continued growth, Brennan, K. A., & Bosson, J. K. (1998). Attachment-style differences in
greater interpersonal connection, acceptance and tolerance of oth- attitudes toward and reactions to feedback from romantic partners: An
ers, creativity, expansion, coherence, autonomy, and joyful en- exploration of the relational bases of self-esteem. Personality and Social
Psychology Bulletin, 24, 699-714.
gagement with life.
Brennan, K. A., Clark, C. L., & Shaver, P. R. (1998). Self-report measure-
ment of adult attachment. In J. A. Simpson & W. S. Rholes (Eds.),
References Attachment theory and close relationships (pp. 46-76). New York:
Guilford Press.
Alexander, P. C , & Anderson, C. L. (1994). An attachment approach to Brennan, K. A., & Morris, K. A. (1997). Attachment styles, self-esteem,
psychotherapy with the incest survivor. Psychotherapy, 31, 665-675. and patterns of seeking feedback from romantic partners. Personality
Allen, K. M., Blascovich, J., Tomaka, J., & Kelsey, R. M. (1991). Presence and Social Psychology Bulletin, 23, 23-31.
of human friends and pet dogs as moderators of autonomic responses to Brennan, K. A., & Shaver, P. R. (1995). Dimensions of adult attachment,
stress in women. Journal of Personality and Social Psychology, 61, affect regulation, and romantic relationship functioning. Personality and
582-589. Social Psychology Bulletin, 21, 267-283.
Ainsworth, M. D. S., Blehar, M. C , Waters, E., & Wall, S. (1978). Patterns Brennan, K. A., & Shaver, P. R. (1998). Attachment styles and personality
of attachment: A psychological study of the Strange Situation. Hillsdale, disorders: Their connections to each other and to parental divorce,
NJ: Erlbaum. parental death, and perceptions of parental caregiving. Journal of Per-
Baldwin, M. W., Fehr, B., Keedian, E., Seidel, M , & Thomson, D. W. sonality, 66, 835-878.
(1993). An exploration of the relational schemata underlying attachment Cafferty, T. P., Davis, K. E., Medway, F. J., O'Hearn, R. E., & Chappell,
styles: Self-report and lexical decision approaches. Personality and K. D. (1994). Reunion dynamics among couples separated during Op-
Social Psychology Bulletin, 19, 746-754. eration Desert Storm: An attachment theory analysis. In K. Bartholomew
Baldwin, M. W., Keelan, J. P., Fehr, B., Enns, V., & Koh-Rangarajoo, E. & D. Perlman (Eds.), Advances in personal relationships (Vol. 5, pp.
(1996). Social-cognitive conceptualization of attachment working mod- 309-330). London: Jessica Kingsley.
els: Availability and accessibility effects. Journal of Personality and Carnelley, K. B., Pietromonaco, P. R., & Jaffe, K. (1994). Depression,
Social Psychology, 71, 94-109. working models of others, and relationship functioning. Journal of
Bandura, A. (1986). Social foundations of thought and action: A social Personality and Social Psychology, 66, 127-140.
cognitive theory. Englewood Cliffs, NJ: Prentice Hall.
Carver, C. S. (1997). Adult attachment and personality: Converging evi-
Bandura, A. (1997). Self-efficacy: The exercise of control. New York:
dence and a new measure. Personality and Social Psychology Bulle-
W. H. Freeman.
tin, 23, 865-883.
Bartholomew, K. (1990). Avoidance of intimacy: An attachment perspec-
Cattell, R. B. (1973). The measurement of the healthy personality and the
tive. Journal of Social and Personal Relationships, 7, 147-178.
healthy society. The Counseling Psychologist, 4, 13-18.
Bartholomew, K., & Horowitz, L. M. (1991). Attachment styles among
Chorpita, B. E., & Barlow, D. H. (1998). The development of anxiety: The
young adults: A test of a four category model. Journal of Personality
role of control in the early environment. Psychological Bulletin, 124,
and Social Psychology, 61, 226-244.
3-21.
Baumeister, R. F., & Leary, M. R. (1995). The need to belong: Desire for
interpersonal attachments as a fundamental human motivation. Psycho- Collins, N. L. (1996). Working models of attachment: Implications for
logical Bulletin, 117, 497-529. explanation, emotion, and behavior. Journal of Personality and Social
Biringen, Z. (1994). Attachment theory and research: Application to clin- Psychology, 71, 810-832.
ical practice. American Journal of Orthopsychiatry, 64, 404-420. Collins, N. L., & Read, S. J. (1990). Adult attachment, working models,
Birnbaum, G. E., Orr, I., Mikulincer, M., & Florian, V. (1997). When and relationship quality in dating couples. Journal of Personality and
marriage breaks up—Does attachment style contribute to coping and Social Psychology, 58, 644-663.
mental health? Journal of Social and Personal Relationships, 4, (AT>- Collins, N. L., & Read, S. J. (1994). Cognitive representations of attach-
654. ment: The structure and function of working models. In K. Bartholomew
Blatt, S. (1990). Interpersonal relatedness and self-definition: Two person- & D. Perlman (Eds.), Attachment processes in adulthood (pp. 53-90).
ality configurations and their implications for psychopathology and London: Jessica Kingsley.
psychotherapy. In J. L. Singer (Ed.), Repression and dissociation: Im- Cooper, M. L., Shaver, P. R., & Collins, N. L. (1998). Attachment styles,