Você está na página 1de 17

Chapter 19

The Attachment System in Adolescence


Joseph P. Allen

Adolescence is, of course, a transitional period, and questions also apply at earlier stages of develop-
this is particularly true with respect to the attach- ment, adolescence brings them all to the forefront
ment system. The early adolescent is beginning a simultaneously.
Herculean effort to become less dependent on pri- This chapter begins with consideration of the
mary attachment figures (i.e., parents). The late normative developmental changes in attachment
adolescent has the potential to function complete- in adolescence. It then moves on to consider the
ly independently of parents and even to become an meaning of individual differences in attachment
attachment figure to his or her own offspring. Yet phenomena. The chapter concludes with a discus-
adolescence is not simply a way station between sion of the implications of what we know about
these two kinds of involvement with attachment. continuities and discontinuities in attachment
Rather, it is a period of profound transformation in phenomena, both during adolescence and between
emotional, cognitive, and behavioral systems sur- adolescence and other phases of development.
rounding attachment relationships, as the adoles- Within this overall structure, I highlight nine pri-
cent evolves from being a receiver of care to becom- mary “developmental transformations,” each of
ing a self-­sufficient adult and potential caregiver to which reflects critical changes in the attachment
peers, romantic partners, and offspring. system or in its conceptualization in light of the
The psychosocial development that takes changes occurring in adolescence.
place in adolescence brings profound changes in
the meaning and expression of attachment pro-
cesses. As a result, one cannot begin to consider Normative Development
the attachment system in adolescence without of the Attachment System in Adolescence
immediately confronting fundamental questions
about its nature. Is attachment in adolescence best Developmental Transformation 1: Moving from
conceptualized in terms of ongoing attachment re- Attachment Relationships to Generalized States
lationships with parents? In terms of relationships of Mind Regarding Attachment
with peers or a romantic partner? As a combina-
tion of multiple relationship experiences and be- By adolescence, a milestone is reached: The at-
haviors? As a way of approaching new attachment tachment system can be assessed in terms of a sin-
relationships? As a way of thinking or talking about gle overarching attachment organization that has
attachment experiences? Although some of these developed, displays stability, and predicts future
419
420 IV. ATTACHMENT IN ADOLESCENCE AND ADULTHOOD

behavior and functioning both within and beyond tachment relationships are no longer important or
the family (Hesse, 1999). Although adolescence worthy of independent assessment. Although the
is the first period during which the primary mode move to the level of representation (Main et al.,
of assessing this overarching organization becomes 1985) and consideration of states of mind regard-
available (via the Adult Attachment Interview ing attachment in adolescence have been incredi-
[AAI] or its downward revision, the Attachment bly productive, they capture only certain aspects of
Interview for Childhood and Adolescence—­ the attachment system’s operations in adolescence.
Ammaniti, van IJzendoorn, Speranza, & Tambelli, As much as adolescents may at times deny the im-
2000; George, Kaplan, & Main, 1984, 1985, 1996), portance of relationships with attachment figures
the organization itself undoubtedly begins to de- (and researchers may find these relationships dif-
velop well before adolescence. Nevertheless, there ficult to assess), primary attachment relationships
is reason to believe that “states of mind” regarding continue in adolescence, develop dramatically, and
attachment—the ways that individuals conceptu- provide important issues worthy of study. Regard-
alize attachment experiences and relationships—­ less of whether a single overarching attachment
develop significantly during adolescence because organization exists in adolescence, individual re-
of teenagers’ rapidly growing capacities for formal lationships can bring out related, but nevertheless
operational thinking, including logical and ab- potentially quite distinct, attachment “styles” dur-
stract reasoning abilities (Keating, 1990). These ing adolescence (Furman & Simon, 2004; Furman,
growing capacities allow an adolescent to begin to Simon, Shaffer, & Bouchey, 2002). And although
construct, from experiences with multiple caregiv- we now have increasingly well-­validated measures
ers, a more integrated and generalized stance to- to assess qualities of such relationships (see, e.g.,
ward attachment experiences (Hesse, 1999; Main, Bartholomew & Shaver, 1998; Brennan, Clark, &
Kaplan, & Cassidy, 1985). Shaver, 1998; Crowell & Owens, 1996; Furman et
In addition, the cognitive and emotional al., 2002; Roisman, Collins, Sroufe, & Egeland,
advances of adolescence allow an adolescent to 2005; Treboux, Crowell, & Waters, 2004), it will
reflect upon and modify his or her states of mind be noted below just what a small proportion of re-
regarding attachment. For example, the dramatic search on adolescent attachment has actually fo-
increases in cognitive differentiation of self and cused on identifying characteristics of individual
other that characterize this period (Selman, 1980) adolescent attachment relationships.
allow the teen to begin to establish a more con- In sum, the construct of attachment in ado-
sistent view of the self as existing apart from in- lescence seems best viewed not as either an in-
teractions with caregivers. The advent of formal trapsychic or a relationship construct, but rather
operational thinking also allows an adolescent to as an organizational construct that is likely to be
contemplate abstract and counterfactual possibili- reflected both in intrapsychic development and in
ties, which in turn allow him or her to compare re- multiple aspects of ongoing attachment relation-
lationships with different attachment figures, both ships (Sroufe & Waters, 1977; Thompson, 1997).
to one another and to hypothetical ideals. Thus The move to the level of representation that has
the adolescent gains the capacity to “de-­idealize” occurred by adolescence is a fundamental transfor-
parents—to see them in both positive and nega- mation that has inspired an explosion of produc-
tive ways (Steinberg, 2005)—which research has tive research in adolescence and adulthood, but
linked with the security of adolescents’ attachment the relational perspective that has so energized
representations (Allen et al., 2003). Adolescents attachment theory earlier in the lifespan remains
not only demonstrate a capacity to think about at- critical to consider.
tachment in a general way, which extends beyond
any single relationship; they also have the capacity
Transformations in Parental Relationships
to operate metacognitively on this thinking—to
begin to reconstruct (or at least tinker with) their Developmental Transformation 3: Achieving
own states of mind regarding attachment. Independence from Attachment Figures
as a Developmental Goal
Developmental Transformation 2: The Continuing
Role of Relationship Assessment Kobak and Duemmler (1994) noted that one of
the most important characteristics of an adoles-
Having the AAI as a means of assessing attach- cent’s attachment relationship with a parent is its
ment phenomena in adolescence does not, how- potential to become increasingly goal-­corrected.
ever, imply that specific qualities of individual at- As the adolescent gains communication and
19. The Attachment System in Adolescence 421

perspective-­taking skills, it becomes possible for homeostatic, balancing safety and exploration, it
both parent and teen to modify (or correct) their makes sense that as increasing maturity increases
attachment-­related behavior when necessary, to safety, exploration will increase and overt attach-
meet the teen’s evolving attachment needs while ment behavior will decrease. Nevertheless, most
balancing other needs as well. The increasingly adolescents still turn to parents under conditions
goal-­corrected nature of the parent–­adolescent re- of extreme stress (Steinberg, 1990), and parents
lationship provides an important context for con- are still often used as attachment figures even in
sidering one of the most important and intriguing young adulthood (Fraley & Davis, 1997). Ado-
changes of adolescence: the decreased reliance on lescents may be on the edge of tears far less often
parents as attachment figures. than infants, but when they are highly distressed,
Whereas the goal-­corrected partnership in their likelihood of turning to parents for help still
infancy might be described as reflecting a coordi- increases dramatically. In this respect, the attach-
nated effort between parent and child, in adoles- ment system operates much as it always has, albeit
cence it seems more appropriate to consider this with a different and rapidly changing balance be-
as a negotiated effort. To a degree unparalleled else- tween attachment and exploratory behaviors.
where in childrearing (with the possible exception An important secondary effect of the chang-
of early toddlerhood), the adolescent struggle for ing balance between attachment and exploratory
autonomy becomes an omnipresent background behavior in adolescence is an increase in the
against which attachment processes play out. At adolescent’s capacity to reevaluate the nature of
this age, the many years of prior operation of the his or her attachment relationship with parents.
attachment system—and the habitual patterns of With increased independence from parents as at-
responding that have become established—­present tachment figures, there comes increased freedom
a significant threat to the adolescent’s efforts to es- from the need to monitor and assure parents’ avail-
tablish autonomy. Put simply, it is exquisitely dif- ability to meet attachment needs (Kobak & Cole,
ficult for a teen to strike out from his or her parents 1994). Main and colleagues (Main & Goldwyn,
and establish independence while feeling pulled 1984; Main, Goldwyn, & Hesse, 2003) refer to
by both habit and the attachment system to retain this cognitive and emotional freedom as “epistem-
their shoulders to cry on. ic space,” and suggest that it allows individuals to
Yet, without such exploration, accomplish- evaluate their parents as attachment figures more
ing the major tasks of social development in objectively.
adolescence and young adulthood would be dif- This epistemic space is likely to be as impor-
ficult if not impossible. In principle, this conflict tant to the emerging capacity to think autono-
is analogous to the competing influences of the mously about attachment relationships as are the
exploratory and attachment systems on infant be- developing cognitive capacities discussed above.
havior, although the press for autonomy in ado- For even with fully developed cognitive capaci-
lescence may be more relentless and more directly ties, it is likely to be difficult to attain the critical
in competition with the attachment system than distance needed to begin objectively evaluating
is the case during infancy (Allen, Moore, & Ku- the qualities of an attachment relationship on
perminc, 1997; Steinberg, 1990). It seems likely which one feels totally dependent. As indepen-
that a secure attachment relationship, with its dence increases, however, so too will the emo-
goal-­correcting capacities, will smooth this tran- tional distance needed to put developing cogni-
sition in adolescence—but it seems equally likely tive capacities to work in reevaluating the nature
that there will still be significant bumps in the of the attachment relationship with parents. Un-
road that affect the teen’s views of attachment and comfortable as this critical distance and objective
social relationships into the future (Allen, Hauser, evaluation may be to parents, it is likely to be fun-
O’Connor, & Bell, 2002). damental to an adolescent’s capacity to develop
The issue is primarily one of developing a an accurate, thoughtful response to attachment
new balance between attachment behaviors (and experiences. This in turn may be crucial to resolv-
cognitions) regarding parents and the adolescent’s ing attachment difficulties in relationships with
exploratory needs. The adolescent’s rapidly devel- parents in ways that allow some adolescents to
oping competence decreases his or her need for form more secure relationships with others in the
dependence on parental attachment figures, and future, and it may be necessary for allowing them
the strong need to explore and master new envi- to reconsider and alter their own states of mind
ronments promotes healthy growth in the explor- regarding attachment (Pearson, Cohn, Cowan, &
atory system. Given that the attachment system is Cowan, 1994).
422 IV. ATTACHMENT IN ADOLESCENCE AND ADULTHOOD

Transformations in Peer Relationships acteristics help us to clarify the ways in which


enduring peer relationships do and do not serve
Developmental Transformation 4:
attachment functions in adolescence. These char-
Extending Attachments beyond
acteristics are (1) proximity seeking, (2) distress
the Child–­Caregiver Relationship
upon inexplicable separation, (3) pleasure or joy
upon reunion, (4) grief at loss, and (5) secure-
The adolescent is not simply becoming more in-
base behavior (comfort and freer exploration in
dependent of parents, however. In important
the presence of an attachment figure). Kobak,
respects, he or she is beginning the process of
Rosenthal, and Serwik (2005) note an additional
transferring dependencies from parental to peer re-
requirement implicit in the characteristics Ains-
lationships. Within the constellation of an infant’s
worth described: The person must view the attach-
attachment figures, a hierarchy is believed to exist,
ment figure as having an enduring commitment to
with the most preferred attachment figure serving
being available if needed, regardless of changes in
as the “principal” attachment figure and others as
time or context.
secondary figures (Bowlby, 1969/1982; Cassidy,
Ainsworth’s list of features makes clear the
1999). If we begin with the premise that most
ways in which childhood playmates differ from
adults are able to form attachment relationships
attachment figures. It also makes clear the extent
with a peer, whereas almost no children can do so,
to which these distinctions begin to blur when
then we must recognize that adolescence will be a
adolescent peer relationships are considered. By
period during which peer relationships will gradu-
late adolescence, long-term relationships can be
ally take on more and more of the qualities of full-
formed in which peers (as romantic partners or
blown adult attachment relationships. What is not
close friends) potentially serve as attachment figures
yet known is how this ontogenetically driven press
in all senses of the term. The change that is occur-
to shift attachment allegiance to a peer or roman-
ring is not just a gradual transition from one class
tic partner unfolds in adolescence.
of attachment figures (parents) to another (peers).
By midadolescence, interactions with peers
Rather, the very nature of what it means to define
have begun to take on many of the functions
a relationship as an attachment relationship must
they will serve for the remainder of the lifespan—­
also change as both parties now take on adult or
providing important sources of intimacy, feedback
adult-like capacities. Unlike in infancy and even
about social behavior, social influence and infor-
childhood, what spurs an adolescent to approach
mation, and ultimately attachment and sexual re-
an attachment figure will only rarely be fundamen-
lationships and lifelong partnerships (Ainsworth,
tal physical safety needs, extreme distress, or risk of
1989; Collins & Laursen, 2000; Gavin & Furman,
imminent emotional disorganization. Rather, the
1989, 1996; Hartup, 1992). The development of
spurs are now usually other, more subtle needs that
peer relationships in adolescence is characterized
may depend on complex interpretive contexts, and
by the gradual emergence of the capacity for adult-
may be more flexibly attended to or ignored in a
like intimacy and supportiveness (Buhrmester,
given situation. In addition, a relationship among
1996; Collins, van Dulmen, Crouter, & Booth,
equals (both of whom may be quite interested in
2006; Hartup, 1992). These well-known changes
preserving their independence) may be a context
create an important challenge for our efforts to un-
that fundamentally alters the meaning and expres-
derstand the operation of the attachment system
sion of attachment behaviors that were previously
during this period. The critical question is this:
directed toward a caregiver.
When do peer relationships become attachment
We now know that there is a broad array of
relationships? This question can be more fruitfully,
different neural, physiological, and psychologi-
though less conventionally, phrased as follows:
cal systems underlying attachment behavior (see
How do peer relationships gradually take on at-
Coan, Chapter 11, this volume). These range
tachment functions?
from managing physiological arousal to establish-
ing a sense of emotional security, to deactivating
Developmental Transformation 5: Moving from
primitive centers of the brain that support “flight”
Attachment Relationships to Attachment Processes
under conditions of stress (Coan, Schaefer, & Da-
in Relationships
vidson, 2006; Cummings & Davies, 1996; Hofer,
1994, 2006; Taylor et al., 2000). Given the infant’s
Ainsworth (1989) delineated five characteristics
extreme vulnerability, limited capacities for self-
that distinguish attachment relationships from
help, and intense relationships with caregivers,
other enduring social relationships. These char-
these functions all come together simultaneously
19. The Attachment System in Adolescence 423

and powerfully under conditions of stress as a fully questions can begin to build an understanding of
functioning attachment system. In adolescence, in the development of the attachment system during
contrast, it is not clear whether the more primitive adolescence, in a way that a static dichotomizing
neural circuits in the brain responsible for detect- of relationships as “attachment relationships” or
ing and physically responding to threat (e.g., the not does not allow.
amygdala and the caudate/nucleus accumbens) are One of the most important endpoints of de-
fully activated by the psychosocial stressors that veloping peer relationships in adolescence is the
adolescents typically experience. attainment of romantic relationships that may
Main (1999) has posited that the attachment eventually become lifelong attachment relation-
system evolved partly because of the value of asso- ships (Ainsworth, 1989; Collins, van Dulmen, et
ciating with others as a source of safety, given our al., 2006; Furman et al., 2002). Romantic relation-
vulnerability as a ground-­living species. Clearly, ships do not result solely from developing interests
this specific function applies only rarely to modern in forming attachments with peers, of course. They
adolescents, even if many other attachment pro- also reflect the operation of a sexual/reproductive
cesses still remain active. With these differences system that is every bit as biologically rooted and
comes the likelihood that the multiple functions critical to species survival as the attachment sys-
and features of the attachment system—which tem (Furman, Brown, & Feiring, 1999; Hazan &
often operate powerfully and in unison during Shaver, 1987; Shaver, Hazan, & Bradshaw, 1988).
infancy—may begin to operate less synchronously However, the sexual and attachment systems
in adolescence, particularly in interactions with both push toward the establishment of romantic
peers. In addition, with the adolescent’s increased relationships characterized by sufficient intensity,
cognitive capacity comes increased flexibility in shared interests, and strong affect to begin to take
directing attachment behavior. Adolescents in- over some of the functions of prior parent–child
creasingly gain the capacity to be “opportunis- relationships. The sexual component of these re-
tic” in seeking out a potential attachment figure, lationships may also help advance the attachment
whether it be a fellow camper at a 4-week summer component by providing consistent motivation
camp or a close-in-age sibling that a teen turns to for interaction, experience with intense, intimate
as parents go through a divorce. Such relationships affect, and a history of shared unique experience.
occur with enough frequency to anchor many pop- By midadolescence, we may begin to see romantic
ular books and films depicting adolescents (see, relationships that at times meet all of the crite-
e.g., Hinton, 1967; King, 1986). ria Ainsworth proposed for attachment relation-
Given these changes from infancy to adoles- ships—from proximity seeking to strong separa-
cence, trying to decide whether a given adolescent tion protest. And romantic partners may take
relationship is or is not an “attachment relation- turns being “stronger and wiser” for each other and
ship” may be overly simplistic. Rather than wading thereby serving as attachment figures (Ainsworth,
into the semantic quagmire of delineating the pre- 1982). Whether adolescent romantic relation-
cise conditions under which a given relationship ships do these things enough to be called “attach-
beyond infancy becomes an “attachment relation- ment relationships” may or may not be merely a
ship,” it may make more sense to recognize that matter of semantics, but that the development of
adolescent relationships increasingly take on criti- attachment processes within these relationships is
cal attachment functions, even if such functions worthy of study is beyond question.
are neither as synchronous nor as intense as they
were in earlier relationships with parents.
A focus on peer attachment processes (as op- Individual Differences
posed to the presence or absence of peer attach- in Adolescent Attachment Strategies
ment relationships) allows us to formulate fruitful
research questions. How does the process of “try- Having established some notion of the transfor-
ing out” close relationships to see whether they mations in attachment-­related cognitions, feel-
might serve attachment functions work? Does this ings, and behavior that occur during adolescence,
process differ for individuals with different attach- it is now possible to consider what we know about
ment histories? Which facets of attachment be- individual differences in the functioning of the
havior come “online” earliest in adolescent peer attachment system in that age period. Somewhat
relationships? What happens when adolescents ironically, given the intense relationship focus
make precocious (or delayed) transitions to using propounded by Ainsworth and Bowlby (Ains-
peers to serve attachment needs? Addressing these worth, Blehar, Waters, & Wall, 1978; Bowlby,
424 IV. ATTACHMENT IN ADOLESCENCE AND ADULTHOOD

1969/1982), by far the bulk of research on indi- One predictor of a secure adolescent state
vidual differences in adolescent attachment in the of mind is the presence of a parent who is highly
Ainsworth–­Bowlby tradition uses the AAI to as- sensitive to the adolescent’s internal states. When
sess adolescent internal states of mind regarding sensitivity is assessed in terms of a mother’s accu-
attachment. Far less research has addressed indi- racy in predicting her teen’s responses on a self-
vidual differences in adolescents’ actual attach- ­perception inventory, it is robustly linked to ado-
ment relationships, a point that will be considered lescent security in the AAI (Allen et al., 2003). In
further below. Both intrapsychic and relational ap- fact, the correlation of security with this marker of
proaches to adolescent attachment are considered sensitivity was slightly higher (r = .35) than is typi-
below, however, and the preponderance of atten- cally found in studies of parental sensitivity toward
tion given to the intrapsychic approaches should infants (De Wolff & van IJzendoorn, 1997). One
not be taken as an endorsement of the primacy of possible explanation is that adolescence research-
those approaches, but only as a reflection of their ers have an advantage over their infancy counter-
place in the research literature to date. Unless oth- parts: In adolescence, sensitivity can be assessed
erwise noted, all references to security or insecuri- directly in the form of concordance between pa-
ty as a property of the individual teen are based on rental report and adolescent behavior. The infancy
assessments with the AAI; references to research researcher, in contrast, has the burden of trying to
on properties of specific relationships will note the judge an infant’s needs with sufficient accuracy to
measures used for each study considered. allow the presumption that any parental deviation
from the researcher’s judgments reflects parental
insensitivity. Efforts to assess maternal sensitiv-
Individual Differences in Family Relationships
ity in adolescence by using the more demanding
The potential tension noted above between the third-party approach used in infancy have been
adolescent’s developmental push to gain autonomy less successful (Beckwith, Cohen, & Hamilton,
and the operation of the attachment system gives 1999). Infant research on sensitivity does not ap-
rise to important individual differences in how this pear to have taken the approach of asking parents
tension is managed. Just as the balance of explora- to predict their infant’s behavior, although such an
tion from a secure base has been highly informa- approach appears promising.
tive with respect to individual differences in infant Although it is quite plausible that parental
attachments, so too the balance of autonomy and sensitivity leads to adolescent security, another
attachment processes in adolescence appears as explanation for the link between attachment and
a robust marker of the quality of an adolescent’s parental sensitivity during adolescence is that se-
internal state of mind regarding attachment. A cure adolescents allow parents to be more sensi-
secure goal-­corrected partnership potentially allows tive by communicating their emotional states to
both parent and teen to recognize the teen’s au- the parents more accurately. Becker-Stoll, Delius,
tonomy strivings and to support these while main- and Scheitenberger (2001) used fine-­grained emo-
taining the relationship. For such a partnership tion coding during discussion tasks to observe, at
to be successful, however, two key ingredients are a micro level, the degree to which teens are af-
required: a strong capacity to communicate across fectively communicative with their mothers. They
the increasingly divergent perspectives and needs found a reliable association between adolescents’
of the parent and teen, and a willingness among security and the degree to which they were affec-
both parties to allow the adolescent to seek au- tively communicative.
tonomy while maintaining the parent–teen rela- Further supporting this idea, we (Berger, Jodl,
tionship. Allen, McElhaney, & Kuperminc, 2005) reported
that adolescent insecure states of mind were linked
Developmental Transformation 6: to greater overall discrepancies between adoles-
Open, Full Dyadic Communication by Adolescents cents’ and parents’ reports of the adolescents’ psy-
with Secure Attachment States of Mind chological symptoms. Notably, these discrepancies
appeared not only when teen reports of symptoms
Cassidy (2001) has suggested that a secure attach- were compared to parental reports, but also when
ment organization beyond childhood may be re- teen reports of their own symptoms were com-
flected in part by efforts to communicate truthfully pared to the reports of their closest peers regarding
and fully in intimate relationships about impor- the teens’ symptoms. This suggests that security is
tant topics. Several lines of research on adolescent linked to properties of adolescents’ dyadic commu-
security bear this out. nication both with parents and with close peers.
19. The Attachment System in Adolescence 425

With respect to the dismissing attachment may perceive disagreements as threats to other-
organization, we (Berger et al., 2005) found that wise shaky relationships. In addition, an insecure
communication was just generally poor and that adolescent may have a history of less than positive
discrepancies were not directional in nature (i.e., experiences with attachment figures when they
there was not a consistent pattern of over- or were called upon in times of need, which is likely
underreporting symptoms by any party). With to color future interactions with them.
respect to the preoccupied attachment organiza- One of the more consistent findings in the
tion, a specific form of bias in communication adolescent attachment literature is that teens with
was found: Adolescents with preoccupied states secure attachment states of mind tend to handle
of mind regarding attachment in the AAI con- conflicts with parents by engaging in productive,
sistently reported the presence of symptoms at problem-­solving discussions that balance auton-
levels that were significantly higher than those omy strivings with efforts to preserve relation-
recognized by either parents or peers. These find- ships with parents (Allen & Hauser, 1996; Allen,
ings make sense if we view preoccupation as a McElhaney, Kuperminc, & Jodl, 2004; Allen et
manifestation of hyperactivation of the attach- al., 2003; Allen, Porter, McFarland, McElhaney,
ment system, and if we recognize that symptom & Marsh, 2007; Becker-Stoll & Fremmer-­Bombik,
reports can readily be seen as cries of distress (e.g., 1997; Kobak et al., 1993). These discussions may
as attachment behaviors). From this perspective, be heated or intense at times, but in families with
we see that insecure-­preoccupied adolescents are secure adolescents they are tempered by behaviors
reporting their distress to a high degree, but not that maintain and support the parent–teen relation-
having these reports heard (or fully believed) by ship. In particular, the relationship-­maintaining
the people closest to them. behaviors in the midst of conflict are most consis-
Clearly, both an adolescent’s attachment or- tently linked to adolescent security, and typically
ganization and the quality of his or her relation- the behavior of the adolescents (rather than their
ship with parents and peers exist as parallel on- parents) is most predictive of adolescent security.
going transactional processes, making any effort In some sense, these relationship-­supporting be-
to tease out directional causal influences excep- haviors may be viewed as “secure-base” behaviors,
tionally difficult. What can be said, however, is in which an adolescent is revisiting and refreshing
that adolescents’ difficulties communicating their the attachment bond even in the midst of explor-
internal states accurately to others appears as a ing autonomy from parents.
robust marker of adolescent insecurity—­whether Most of the research in this area has been
these difficulties stem from adolescent communi- done with mothers, although recent work sug-
cative difficulties, problems in selecting receptive gests that similar patterns exist for fathers as well
peer partners, parental lack of sensitivity, or some (Allen et al., 2007). Research on fathers adds one
combination of all these factors. other behavior to the mix—use of harsh conflict
tactics—as a marker of adolescent insecurity. As
Developmental Transformation 7: with research on mothers, the focus is less on
Resolution of the Attachment–­Autonomy Tension whether an adolescent can establish autonomy
in Adolescents with Secure Attachment States in the disagreement than on the autonomy chal-
of Mind lenge as a backdrop against which relationships
are either actively maintained or significantly
As noted above, the negotiation of attach- threatened.
ment and autonomy issues may be problematic for In terms of specific types of insecure states of
all families at some point, but it may be especially mind, Becker-Stoll and Fremmer-­Bombik (1997)
difficult for the family of an insecure adolescent. report that dismissing adolescents show the least
The moodiness, changing relationships, tension, autonomy and relatedness in interactions with par-
and growing emotional and behavioral indepen- ents of all attachment groups observed. This sug-
dence from parents that characterize adolescent gests that a dismissing individual’s characteristic
development may conspire to create chronic ac- withdrawal from engagement with attachment ex-
tivation of the attachment system, thus increasing periences may particularly hinder the task of rene-
the impact of an insecure attachment organization gotiating parent–­adolescent relationships. Reimer,
on the adolescent’s behavior. Insecure adolescents Overton, Steidl, Rosenstein, and Horowitz (1996)
(and parents) may be overwhelmed by the af- also noted that families of dismissing adolescents
fect brought on by disagreements (Kobak, Cole, tend to be less responsive to their adolescents than
Ferenz-­Gillies, Fleming, & Gamble, 1993) and do families of preoccupied adolescents.
426 IV. ATTACHMENT IN ADOLESCENCE AND ADULTHOOD

Insecure preoccupation, in contrast, appears One long-term longitudinal study has shown
to be best predicted from heightened and unpro- that infant security with mothers was more predic-
ductive overengagement with parents in argu- tive of observed qualities of autonomy and related-
ments that ultimately undermine an adolescent’s ness in adolescent–­mother interactions than it was
autonomy. We (Allen & Hauser, 1996) reported of adolescent states of mind regarding attachment
that one indicator of preoccupation with attach- (Becker-Stoll & Fremmer-­Bombik, 1997). These
ment in young adulthood—use of passive thought findings suggest that success in negotiating auton-
processes, reflecting mental entanglement between omy issues in adolescence may be a stage-­specific
self and caregivers—was predicted by adolescents’ manifestation of a long-term secure attachment
overpersonalized behaviors toward fathers in ar- relationship with parents. Notably, continuity in
guments 10 years earlier and by adolescents’ lack the qualities of an attachment relationship were ob-
of simple withdrawal from or avoidance of argu- tained, even in the absence of continuity from an
ments. This overengagement appears to extend early attachment relationship to an adolescent’s
well into late adolescence, as research also suggests later internalized state of mind regarding attach-
that adolescents with insecure-­preoccupied status ment. This at least raises the possibility that se-
on the AAI are likely to have more difficulty leav- curity in adolescent–­parent relationships may be
ing home successfully for college, displaying high- distinct in important ways from security in ado-
er levels of both conflict and contact with parents lescents’ states of mind, and that both are worthy
during the transition (Bernier, Larose, & Whipple, of study.
2005). These effects were not found for adoles-
cents who were not leaving for college, suggesting
Individual Differences in Peer Relationships
that these attachment dynamics were likely to be
activated mainly in the presence of a significant Developmental Transformation 8: Peer Relationships
stress to the attachment system. as the Critical Context in Which Individual
Family interaction patterns may also predict Differences in Attachment Processes Emerge
changes in adolescents’ states of mind over time.
When autonomy-­undermining, enmeshed behav- There are several reasons to expect close links be-
ior between mothers and their adolescents was ob- tween an adolescent’s attachment organization and
served at age 16, this predicted relative decreases in qualities of ongoing peer relationships. A secure
levels of security from age 16 to age 18 (Allen et state of mind regarding attachment, characterized
al., 2004)—a notable finding, given evidence that by coherence in talking (and presumably thinking)
insecure-­preoccupied mothers may be more likely about attachment-­related experiences and feelings,
to engage in such entangled conversations (Kobak, should permit similar experiences and feelings in
Ferenz-­Gillies, Everhart, & Seabrook, 1994). As peer relationships to be processed more accurately
in other attachment and family interaction studies and coherently as well. In contrast, the defensive
of enmeshed and overpersonalizing behavior, the exclusion of information that is characteristic of
key element was the adolescents’ behavior, not the insecure organizations may lead to distorted com-
mothers’. Adolescents who engage in such behavior- munications and negative expectations about oth-
al strategies may be struggling with autonomy issues ers, both of which have been linked to problems in
in a highly confused way that leaves them mentally social functioning at various points in the lifespan
and emotionally entangled in their relationships (Cassidy, Kirsh, Scolton, & Parke, 1996; Dodge,
with their mothers. Such mental entanglement 1993; Slough & Greenberg, 1990). Similarly, dis-
during a developmental period characterized by the comfort with attachment-­related affect and ex-
need to establish autonomy seems likely to produce periences may lead adolescents with dismissing
enormous emotional stress—­stress that might typi- attachment strategies to push away peers, particu-
cally lead to seeking out an attachment figure. The larly those who could become close friends (La-
catch in this case is that the stress obviously cannot rose & Bernier, 2001; Spangler & Zimmermann,
be easily assuaged by mothers, given the role of the 1999). This is consistent with the finding that for
maternal relationship as the source of the stress. This college students, hostility and lack of social skills
unassuaged stress seems likely to make it difficult for as rated by close friends are linked to students’ in-
the adolescents to step back sufficiently to engage secure attachment organizations (Kobak & Sceery,
in the thoughtful and balanced reevaluation of pa- 1988). A related mechanism by which attachment
rental relationships that is characteristic of secure- organization may be linked to peer relationships
­autonomous individuals (Main, 1999). is that insecure attachment organization co-­occurs
19. The Attachment System in Adolescence 427

with (and serves to mark) problematic ongoing re- of relationships with parents (and with friends).
lationships with parents, which make it difficult to Furman (2001) also noted that many of the same
move freely beyond these relationships to establish features of cognitive appraisal and representation
successful new relationships with peers (Gavin & that apply to thinking about relationships with
Furman, 1996). parents and peers are likely to extend to romantic
A large, rapidly growing body of research sug- relationships as well. It thus seems likely that prior
gests a fairly tight linkage between a secure ado- attachment experiences and current patterns of
lescent attachment organization and competence approach to attachment thoughts and feelings will
with peers. Research indicates, for example, that in turn shape the nature of developing romantic
secure adolescents are more comfortable with the relationships (Hazan & Shaver, 1994).
intimate emotional interactions common in close In late adolescence, security as assessed with
friendships (Allen et al., 2007; Lieberman, Doyle, the AAI has been linked to the subsequent quality
& Markiewicz, 1999; Sroufe, Egeland, Carlson, of interactions with a romantic partner, as coded
& Collins, 2005b; Weimer, Kerns, & Oldenberg, from videotaped observations (Roisman, Madsen,
2004; Zimmermann, 2004). Observational data Hennighausen, Sroufe, & Collins, 2001). Assess-
suggest that this competence is a result of gener- ments of the degree of self-­reported insecurity in
alized comfort in handling one’s own emotional romantic relationships has been linked to greater
reactions in challenging situations (Zimmermann, anxiety among relationship partners, as well as to
Maier, Winter, & Grossmann, 2001). Similarly, greater incidence of sexual intercourse but with
using Hazan and Shaver’s (1987) three-­category less enjoyment derived from it (Tracy, Shaver,
prototype measure of attachment style, Cooper, Albino, & Cooper, 2003). Moore (1997) reported
Shaver, and Collins (1998) found that anxious- that among sexually active adolescents, insecurity
­ambivalent adolescents were prone to interperson- was associated with having more sexual partners
al hostility. And although some research suggests and with less frequent use of contraception. We
that security is more relevant to functioning in (Marsh, McFarland, Allen, McElhaney, & Land,
close relationships than in broader peer relation- 2003) found that an interaction of adolescent pre-
ships (Hazan & Shaver, 1987; Lieberman et al., occupation and mothers’ focus on their own (as
1999), adolescent security has also been linked to opposed to their adolescents’) autonomy was a
such measures of broader social competence such predictor of adolescents’ early sexual activity. Pre-
as overall friendship quality, popularity, and social occupied adolescents whose mothers were focused
acceptance (Allen, Moore, Kuperminc, & Bell, on their own (maternal) autonomy were more
1998; Allen et al., 2007). likely to engage in early sexual activity, whereas
In some cases, peer relationship qualities preoccupied adolescents whose mothers were rela-
(particularly the ability to seek emotional support tively unfocused on their own autonomy had strik-
from a peer while maintaining autonomy with re- ingly lower rates of early sexual activity (Marsh et
spect to peer pressure) are more strongly linked al., 2003). Of course, in thinking about qualities of
with adolescent attachment states of mind than dating relationships, partners’ attachment organi-
are some of the best markers of maternal and pater- zation will also be important to consider, as even
nal relationship qualities (Allen et al., 2007). This later in adulthood only modest evidence of assor-
is not necessarily surprising, given that mastering tative mating with respect to attachment exists
the realm of peer relationships may be the single (Owens et al., 1995; Treboux et al., 2004). In this
greatest social-­developmental challenge faced by sense, predictions from the AAI-based security of
most adolescents, and it does not mean that these one party in a relationship to overall qualities of
peer relationships are causally influencing adoles- that relationship are likely to underestimate the
cent security. It does suggest, however, the extent role of attachment, given that attachment is being
to which peer relationships have become the cen- assessed for only one of the two partners who to-
tral arena in which attachment processes are likely gether determine the qualities of the relationship
to play out during adolescence and beyond. (see J. Feeney, Chapter 21, this volume).
Individual differences in attachment organi-
zation are also linked to behavior in romantic and
Attachment and Adolescent Mental Health
sexual relationships in adolescence. Furman and
colleagues (2002) found that adolescents’ work- A number of recent studies suggest substantial
ing models of relationships with romantic partners links between adolescent attachment organiza-
exhibit substantial similarities to working models tion and mental health. Among the most highly
428 IV. ATTACHMENT IN ADOLESCENCE AND ADULTHOOD

disturbed adolescents—those requiring residential likely to be ignored or rebuffed, they are likely to
treatment—three studies have found links to either react with externalizing behavior.
concurrent or future attachment insecurity, and There is, of course, a third environmental
to heightened prevalence of insecure-­unresolved possibility for preoccupied teens. They may be ex-
attachment status (Allen, Hauser, & Borman- posed to positive social relationships and behav-
­Spurrell, 1996; Wallis & Steele, 2001). Even with iors, and in this case the outcomes look far more
less severe forms of disturbance, both the preoc- positive. When preoccupied and secure teens are
cupied and dismissing strategies have been im- exposed to positive friendships, they are at lower
plicated in problems of psychosocial functioning, risk for delinquent behavior (McElhaney, Immele,
although the two are associated with somewhat Smith, & Allen, 2006). In addition, when exposed
different patterns of problems. to effective maternal behavioral control strategies,
Adolescents’ use of preoccupied strategies both preoccupied and secure teens have lower lev-
has been most closely linked to internalizing els of delinquent behavior than do dismissing teens
problems, particularly to adolescents’ self-­reports exposed to the same maternal behavior (Allen et
of depression, anxiety disorders, and internaliz- al., 1998).
ing symptoms and stress during transitions (Allen Taken together, these results suggest that the
et al., 1996; Bernier et al., 2005; Cole-Detke & hyperactivated state of the attachment system in
Kobak, 1996; Kobak, Sudler, & Gamble, 1991; preoccupied adolescents leads them to be exqui-
Larose & Bernier, 2001; Rosenstein & Horowitz, sitely sensitive to their social environments. In
1996). In addition, preoccupied attachment states enmeshed environments, internalizing symptoms
of mind may interact with a wide array of psycho- may be a way of seeking responses from attach-
social and environmental factors to predict critical ment figures (similar to an infant’s cries of dis-
outcomes. When preoccupied teens are confront- tress). When these relatively unobtrusive internal-
ed with intrapsychic states or environments that izing behaviors are likely to be completely ignored
are confusing or enmeshing, higher levels of inter- (in less responsive environments), an adolescent
nalizing symptoms are found. For example, Adam, may “raise the volume” by turning to more dra-
Sheldon-­Keller, and West (1996) reported that matic externalizing behaviors as a way to engage
suicidality in adolescence was related to a combi- parental attention and to express anger and resis-
nation of preoccupied and unresolved attachment tance. Finally, a hyperactivated attachment sys-
status. Similarly, preoccupied adolescents who tem, although clearly problematic, is nonetheless
had mothers who could not exercise their own a system that continues to function; in cases where
autonomy in discussions (i.e., were passive and it brings the adolescent into contact with positive
enmeshed) displayed higher levels of depression social interactions, it appears to leave the teen re-
(Marsh et al., 2003). sponsive to these as well.
Conversely, preoccupied teens are more In contrast to preoccupied adolescents, Cole-
likely to display externalizing behaviors under Detke and Kobak (1996; Kobak & Cole, 1994)
some circumstances. For example, when they have suggest that adolescents with dismissing strategies
mothers who display extremely high levels of their may take on symptoms that distract themselves
own (maternal) autonomy in discussions (perhaps and others from attachment-­related cues. When
asserting themselves to the point of ignoring their examining psychiatrically hospitalized adolescents,
adolescents), preoccupied teens have higher levels almost all of whom were insecure, Rosenstein and
of drug use, engage in precocious sexual activity, Horowitz (1996) reported that dismissing strate-
and exhibit higher levels of delinquent behavior gies were associated with externalizing symptoms:
over the following year (Allen et al., 1998). Simi- Dismissing adolescents were more likely to exhibit
larly, when preoccupied adolescents are exposed substance abuse and conduct disorder. Similarly,
to poverty (perhaps another situation in which my colleagues and I found that dismissing attach-
many of their needs are ignored) and/or are male, ment strategies were predictive of increasing de-
there is also an increased likelihood of delinquent linquency and externalizing behavior over both
behavior. Although the pattern is not yet well es- shorter and longer spans of adolescence (Allen,
tablished, the initial research findings collectively Marsh, et al., 2002; Allen et al., 2007).
suggest that when preoccupied adolescents are ex- In addition, dismissing attachment strategies
posed to passivity or enmeshment, an internaliz- have been linked to difficulty getting assistance
ing, anxious/depressed pattern emerges; however, from peers and teachers, as well as to peer-­reported
in situations where their attachment entreaties are social withdrawal during the transition to college
19. The Attachment System in Adolescence 429

(Larose & Bernier, 2001). Consistent with this exist at completely different levels of analysis—
finding, insecurity that manifests itself primarily the intrapsychic and the dyadic/relational—and
as dismissal of attachment in early adolescence thus by their very nature differ in fundamental and
has been linked to relative decreases in social skills irreducible ways. The question that adolescent re-
over time (Allen, Marsh, et al., 2002) and with search brings into focus most strongly is this: How
less active coping strategies (Seiffge-­Krenke & do these two different constructs relate to one an-
Beyers, 2005). Cole-Detke and Kobak (1996) also other, both conceptually and empirically?
reported that eating-­disordered individuals in a From infancy to adolescence (and from the
college population were more likely to use dismiss- Strange Situation to the AAI), what we find is
ing strategies; the attention given to their eating that the continuities across measures of attach-
behaviors was hypothesized to distract them from ment processes are quite modest. These conti-
feelings of internal emotional distress. Unlike pre- nuities appear most robustly when environments
occupied adolescents, dismissing adolescents do are generally stable and benign, and they may
not appear to be particularly sensitive to parental disappear entirely under more challenging cir-
behaviors: For example, a factor such as paren- cumstances (although assessment of intervening
tal control of adolescent behavior, which is well environmental factors can account for some ap-
established as a buffer against delinquency, did parent discontinuities) (Hamilton, 2000; Waters,
not appear to serve this role for dismissing teens Hamilton, & Weinfield, 2000; Weinfield, Sroufe,
(Allen et al., 1998). & Egeland, 2000; Weinfield, Whaley, & Egeland,
Other than the few studies mentioned at the 2004). Notably, observed continuities between in-
outset of this section, disorganized or unresolved fant attachment status and other qualities of ado-
states of mind have been the subject of far less lescents’ close relationships make clear that infant
research on adolescent psychopathology. This at- attachment status undoubtedly has implications
tachment status has been linked to disrupted ma- for future relationship qualities, which in some
ternal behavior with infants in adolescent–­mother cases are even stronger than links to future adoles-
dyads (Madigan, Moran, & Pederson, 2006), but cent and adult attachment measures (Grossmann,
otherwise it remains a fertile, relatively untapped Grossmann, & Kindler, 2005; Sroufe, Egeland,
area for studying more severe forms of adolescent Carlson, & Collins, 2005b; Zimmermann, 2004;
psychopathology. Zimmermann, Maier, Winter, & Grossmann,
2001). Within adolescence, security assessed with
the AAI displays only very modest correlations
Continuity, Discontinuity, with maternal security (Allen et al., 2004). These
and the Central Question: continuities and discontinuities are considered far
What Is Attachment in Adolescence? more thoroughly in other chapters of this volume.
For the purposes of this chapter, the point is that
Thus far, this chapter has considered many of the continuities from infancy and from concurrent
the features of adolescent attachment strategies measures of parental attachment status to adoles-
while skirting the question raised at the outset: cent states of mind regarding attachment are rela-
Just what is attachment in adolescence? When tively modest.
we are considering continuities of attachment in Evidence from within adolescence, however,
adolescence to attachment at other developmen- suggests that the modest continuity with prior in-
tal stages, however, precision in defining attach- fant attachment and current maternal attachment
ment becomes critical. For example, considering is not due to any inherent instability in adolescent
the transition from the assessment of attachment attachment organization; nor is it probably due
relationships in infancy and childhood (albeit rela- to difficulties in assessing adolescent attachment
tionships that are also represented internally), to with the AAI. Attachment as assessed with the
the assessment of an internal state in adolescence AAI displays strong stability, with test–­retest cor-
and beyond that transcends any particular rela- relations between overall indices of security rang-
tionship, makes clear that the two most common ing from .51 to .61 from ages 16 to 18 in two dif-
measurement approaches assess fundamentally ferent samples in Europe and the United States
different constructs at these two time points. At- (Allen et al., 2004; Zimmermann & Becker-Stoll,
tachment as assessed with the AAI is not simply 2002). Similarly, even in research using a modified
a more developed version of what we see in the AAI (the Attachment Interview of Childhood
Strange Situation. Rather, these two constructs and Adolescence) with a far younger sample, simi-
430 IV. ATTACHMENT IN ADOLESCENCE AND ADULTHOOD

lar stabilities have been found in overall security with the AAI is actually a fundamentally different
between ages 10 and 14 (kappa = .48) (Ammaniti (albeit related) construct from what we measure in
et al., 2000). the infant Strange Situation. Mary Main has been
Even the discontinuities in attachment states quite careful to describe what the AAI measures
of mind across adolescence that have been found not as attachment per se, but as “states of mind
appear to be explained largely by environmental with respect to attachment.” And she has de-
factors. Exposure to social and intrapsychic envi- scribed the analogue to infant security not as ado-
ronments that emotionally overwhelm adolescents, lescent or adult security, but as a state of mind that
while leaving them relatively unable to obtain is “autonomous, yet valuing” of attachment (Main
support from attachment figures, predicts relative & Goldwyn, 1984; Main et al., 2003). Although
increases in insecurity over and above baseline lev- thus far I have relied on the convenience of dis-
els across a 2-year period in adolescence (Allen et cussing AAI study results in terms of “security” of
al., 2004). These environments include poverty “adolescent attachment,” this usage, while handy
(which stresses adolescents while draining par- and defensible, may obscure several important dis-
ents), adolescent depression (a difficult-to-­soothe tinctions.
form of distress), and emotional enmeshment If one simply examines the origin of the AAI
(which stresses adolescents while making them and the validation research that led to its rapid
want to avoid seeking comfort from a parental fig- rise to prominence, it is difficult to avoid the con-
ure). Together, these factors account for substantial clusion that the AAI is most directly tapping not
variance in the change in security during adoles- the attachment system of the individual, but the
cence. They bring the R2 for security at age 18 (as caregiving system (Allen & Manning, in press).
predicted by security and environmental factors at Far more than any other measure, the AAI most
age 16) up to .72, which approaches the limits of strongly predicts success as a caregiver in raising a
the AAI’s reliability. Together with the underlying secure infant—as indeed it was primarily designed
stability of security, these findings suggest both the and intended to do. Obviously the caregiving and
robustness of attachment during adolescence and attachment systems are likely to be highly inter-
its continuing sensitivity to qualities of the adoles- related, as both draw upon many of the same cog-
cent’s psychosocial environment. nitive, behavioral, and emotional subsystems. It
If the lack of strong continuity between in- even appears defensible to use the terms “attach-
fant Strange Situation status and adolescent at- ment” and “security” broadly to refer to qualities of
tachment is not due to problems in measurement both kinds of systems. However, equally obviously,
in adolescence, or to any lack of meaningfulness of these systems are not isomorphic. Although one’s
the construct of adolescent attachment organiza- expectations about getting one’s own attachment
tion, then two explanations remain most viable. needs met may be linked to one’s ability to meet
First, intervening experiences between infancy the attachment needs of others, the two are clearly
and adolescence may play a large part in altering not identical (Collins, Ford, Guichard, & Allard,
the developmental course of the attachment sys- 2006). (See George and Solomon, Chapter 35,
tem for many individuals. Recent research raises this volume, for an extensive discussion of the
questions about whether in fact any internal stabil- caregiving system.)
ity in attachment actually exists from infancy to This issue comes to a head in adolescence,
adolescence; rather, the observed continuity over when individuals are making the transition from
this time span may primarily reflect stability in receiving care from attachment figures to gaining
parents’ attachment strategies (Belsky & Fearon, the capacity to be providers of such care to oth-
2002; van IJzendoorn, 1996). This view is sup- ers. Notably, by far the strongest correlations of
ported by evidence just reviewed showing that the adolescent attachment assessed via the AAI with
attachment system remains quite open to environ- any other indicator of functioning have been with
mental inputs well into adolescence. a caregiving outcome: the security of the infant
offspring of adolescent mothers (Ward & Carlson,
Developmental Transformation 9: 1995). The AAI may be more strongly linked to
The Emergence of the Caregiving System adolescents’ functioning with peers than to their
relationships with parents (as initial reports sug-
A second explanation for the relative lack of con- gest), because the AAI is most closely tapping a
tinuity in attachment measures from infancy to behavioral system—for meeting others’ attach-
adolescence is that the construct we are measuring ment needs—more relevant for relating to peers
19. The Attachment System in Adolescence 431

than to parents. The quality of close peer relation- cally associated with the attachment system. So
ships may well be determined by a teen’s capacity too does our view of the attachment system during
to meet a peer’s emotional/attachment-like needs, this period need to continue to grow and expand,
whereas a teen’s ability to meet the needs of a par- to reflect this increasing complexity and sophisti-
ent is of minimal importance in healthy families. cation.
Appropriately defining the boundaries of
the AAI as an indicator of adolescent attach-
ment processes has the advantage of potentially Acknowledgment
redirecting our attention to a gaping hole that re-
mains in research on adolescent attachment: We This chapter was completed with the assistance of grants
know remarkably little about the qualities of the from the National Institute of Mental Health.
actual attachment relationships adolescents form.
Knowing that adolescents with autonomous states
References
of mind regarding attachment form relationships
with certain qualities is useful, but it is not the
Adam, K. S., Sheldon-­Keller, A. E., & West, M. (1996).
same as being able to assess security or insecurity
Attachment organization and history of suicidal be-
in specific adolescent relationships (in part be- havior in clinical adolescents. Journal of Consulting
cause secure teens may at times form insecure rela- and Clinical Psychology, 64, 264–272.
tionships with other teens). Our efforts to develop Ainsworth, M. D. S. (1982). Attachment: Retrospect
sophisticated approaches to assessing relationship and prospect. In C. Parkes & J. Stevenson-Hinde
security in adolescence (and beyond) are only just (Eds.), The place of attachment in human behavior
beginning (Cooper et al., 1998). Attachment rela- (pp. 3–30). New York: Basic Books.
tionships in adolescence clearly differ dramatically Ainsworth, M. D. S. (1989). Attachments beyond in-
from those in infancy, as various new issues—from fancy. American Psychologist, 44, 709–716.
power and control, to gender and sexuality, to the Ainsworth, M. D. S., Blehar, M. C., Waters, E., & Wall,
S. (1978). Patterns of attachment: A psychological study
increasing mutual capacity to provide comfort in
of the Strange Situation. Hillsdale, NJ: Erlbaum.
relationships—all rise to prominence. Given these Allen, J. P., & Hauser, S. T. (1996). Autonomy and relat-
differences, it is not yet clear whether relationship edness in adolescent–­family interactions as predictors
assessments in adolescence will ultimately closely of young adults’ states of mind regarding attachment.
parallel Ainsworth’s categorizations of infant at- Development and Psychopathology, 8(4), 793–809.
tachment relationships, or whether they will need Allen, J. P., Hauser, S. T., & Borman-­Spurrell, E. (1996).
to expand upon and modify these to address the Attachment theory as a framework for understand-
complexities of adult relationships. ing sequelae of severe adolescent psychopathology:
From this perspective, an overly narrow focus An 11-year follow-up study. Journal of Consulting and
on issues of continuity in attachment from earlier Clinical Psychology, 64(2), 254–263.
Allen, J. P., Hauser, S. T., O’Connor, T. G., & Bell, K. L.
stages into adolescence, even of heterotypic con-
(2002). Prediction of peer-rated adult hostility from
tinuity, may risk oversimplifying the constructs autonomy struggles in adolescent–­family interactions.
in question. It may make more sense to think of Development and Psychopathology, 14, 123–137.
the attachment system as expanding and evolving Allen, J. P., & Manning, N. (in press). From safety to
into multiple new forms in adolescence—­ranging affect regulation: Attachment from the vantage point
from attachment functions appearing temporarily of adolescence. New Directions for Child and Adoles-
in passing relationships to complex new varieties cent Development.
of long-term peer relationships—than to see it as Allen, J. P., Marsh, P., McFarland, C., McElhaney, K. B.,
a homunculus moving from infancy into adoles- Land, D. J., Jodl, K. M., et al. (2002). Attachment
cence with only growth and a change in outward and autonomy as predictors of the development of
social skills and delinquency during midadolescence.
appearance.
Journal of Consulting and Clinical Psychology, 70(1),
To put this differently, the development of 56–66.
attachment theory now needs to recapitulate indi- Allen, J. P., McElhaney, K. B., Kuperminc, G. P., & Jodl,
vidual ontogeny. Adolescent development brings K. M. (2004). Stability and change in attachment
striking growth in capacities to maintain a diverse security across adolescence. Child Development, 75,
array of relationships meeting a diverse array of 1792–1805.
potential attachment needs, and triggering numer- Allen, J. P., McElhaney, K. B., Land, D. J., Kuperminc, G.
ous systems (from primitive neural mechanisms to P., Moore, C. M., O’Beirne-­Kelley, H., et al. (2003).
higher-order cognitive functions) that are typi- A secure base in adolescence: Markers of attachment
432 IV. ATTACHMENT IN ADOLESCENCE AND ADULTHOOD

security in the mother–­adolescent relationship. Child Self-­report measurement of adult attachment: An


Development, 74, 292–307. integrative overview. In J. A. Simpson & W. S.
Allen, J. P., Moore, C., & Kuperminc, G. (1997). De- Rholes (Eds.), Attachment theory and close relationships
velopmental approaches to understanding adolescent (pp. 46–76). New York: Guilford Press.
deviance. In S. S. Luthar & J. A. Burack (Eds.), De- Buhrmester, D. (1996). Need fulfillment, interpersonal
velopmental psychopathology: Perspectives on adjust- competence, and the developmental contexts of early
ment, risk, and disorder (pp.  548–567). New York: adolescent friendship. In W. M. Bukowski, A. F. New-
Cambridge University Press. comb, & W. Hartup (Eds.), The company they keep:
Allen, J. P., Moore, C., Kuperminc, G., & Bell, K. Friendship in childhood and adolescence (pp. 158–185).
(1998). Attachment and adolescent psychosocial Cambridge, UK: Cambridge University Press.
functioning. Child Development, 69(5), 1406–1419. Cassidy, J. (1999). The nature of the child’s ties. In
Allen, J. P., Porter, M. R., McFarland, F. C., McElhaney, J. Cassidy & P. R. Shaver (Eds.), Handbook of at-
K. B., & Marsh, P. A. (2007). The relation of at- tachment: Theory, research, and clinical applications
tachment security to adolescents’ paternal and peer (pp. 3–20). New York: Guilford Press.
relationships, depression, and externalizing behavior. Cassidy, J. (2001). Truth, lies, and intimacy: An attach-
Child Development, 78, 1222–1239. ment perspective. Attachment and Human Develop-
Ammaniti, M., van IJzendoorn, M. H., Speranza, A. M., ment, 3(2), 121–155.
& Tambelli, R. (2000). Internal working models of Cassidy, J., Kirsh, S. J., Scolton, K. L., & Parke, R. D.
attachment during late childhood and early adoles- (1996). Attachment and representations of peer rela-
cence: An exploration of stability and change. At- tionships. Developmental Psychology, 32(5), 892–904.
tachment and Human Development, 2(3), 328–346. Coan, J. A., Schaefer, H. S., & Davidson, R. J. (2006).
Bartholomew, K., & Shaver, P. R. (1998). Methods of Lending a hand: Social regulation of the neu-
assessing adult attachment: Do they converge? In J. ral response to threat. Psychological Science, 17,
A. Simpson & W. S. Rholes (Eds.), Attachment theory 1032–1039.
and close relationships (pp. 25–45). New York: Guilford Cole-Detke, H., & Kobak, R. (1996). Attachment pro-
Press. cesses in eating disorder and depression. Journal of
Becker-Stoll, F., Delius, A., & Scheitenberger, S. (2001). Consulting and Clinical Psychology, 64(2), 282–290.
Adolescents’ nonverbal emotional expressions during Collins, N. L., Ford, M. B., Guichard, A. M., & Al-
negotiation of a disagreement with their mothers: An lard, L. M. (2006). Working models of attachment
attachment approach. International Journal of Behav- and ­attribution processes in intimate relationships.
ioral Development, 25(4), 344–353. Personality and Social Psychology Bulletin, 32(2),
Becker-Stoll, F., & Fremmer-­Bombik, E. (1997, April). 201–219.
Adolescent–­mother interaction and attachment: A longi- Collins, W. A., & Laursen, B. (2000). Adolescent re-
tudinal study. Paper presented at the biennial meet- lationships: The art of fugue. In C. Hendrick & S.
ings of the Society for Research in Child Develop- S. Hendrick (Eds.), Close relationships: A sourcebook
ment, Washington, DC. (pp. 59–69). Thousand Oaks, CA: Sage.
Beckwith, L., Cohen, S. E., & Hamilton, C. E. (1999). Collins, W. A., van Dulmen, M., Crouter, A. C., & Booth,
Maternal sensitivity during infancy and subsequent A. (2006). “The course of true love(s)”: Origins and
life events relate to attachment representation at pathways in the development of romantic relation-
early adulthood. Developmental Psychology, 35(3), ships. In A. C. Crouter & A. Booth (Eds.), Romance
693–700. and sex in adolescence and emerging adulthood: Risks and
Belsky, J., & Fearon, R. M. P. (2002). Early attachment opportunities (pp. 63–86). Mahwah, NJ: Erlbaum.
security, subsequent maternal sensitivity, and later Cooper, M. L., Shaver, P. R., & Collins, N. L. (1998).
child development: Does continuity in development Attachment styles, emotion regulation, and adjust-
depend upon continuity of caregiving? Attachment ment in adolescence. Journal of Personality and Social
and Human Development, 4(3), 361–387. Psychology, 74(5), 1380–1397.
Berger, L. E., Jodl, K. M., Allen, J. P., McElhaney, K. Crowell, J. A., & Owens, G. (1996). Current relationship
B., & Kuperminc, G. P. (2005). When adolescents interview and scoring system. Unpublished manuscript,
disagree with others about their symptoms: Differ- State University of New York at Stony Brook.
ences in attachment organization as an explanation Cummings, E. M., & Davies, P. (1996). Emotional se-
of discrepancies between adolescent-, parent-, and curity as a regulatory process in normal development
peer-­reports of behavior problems. Development and and the development of psychopathology. Develop-
Psychopathology, 17, 489–507. ment and Psychopathology, 8(1), 123–139.
Bernier, A., Larose, S., & Whipple, N. (2005). Leaving De Wolff, M., & van IJzendoorn, M. H. (1997). Sensi-
home for college: A potentially stressful event for tivity and attachment: A meta-­analysis on parental
adolescents with preoccupied attachment patterns. antecedents of infant attachment. Child Development,
Attachment and Human Development, 7(2), 171–185. 68(4), 571–591.
Bowlby, J. (1969/1982). Attachment and loss: Vol. 1. At- Dodge, K. A. (1993). Social-­cognitive mechanisms in
tachment. New York: Basic Books. the development of conduct disorder and depression.
Brennan, K. A., Clark, C. L., & Shaver, P. R. (1998). Annual Review of Psychology, 44, 559–584.
19. The Attachment System in Adolescence 433

Fraley, R. C., & Davis, K. E. (1997). Attachment for- Hofer, M. A. (1994). Hidden regulators in attachment,
mation and transfer in young adults’ close friendships separation, and loss. In N. Fox (Ed.), The develop-
and romantic relationships. Personal Relationships, 4, ment of emotion regulation: Biological and behav-
131–144. ioral considerations. Monographs of the Society for
Furman, W. (2001). Working models of friendships. Research in Child Development, 59(2–3), 192–207,
Journal of Social and Personal Relationships, 18(5), 250–283.
583–602. Hofer, M. A. (2006). Psychobiological roots of early at-
Furman, W., Brown, B. B., & Feiring, C. (Eds.). (1999). tachment. Current Directions in Psychological Science,
The development of romantic relationships in adolescence. 15(2), 84–88.
New York: Cambridge University Press. Keating, D. P. (1990). Adolescent thinking. In S. S.
Furman, W., & Simon, V. A. (2004). Concordance in Feldman & G. Elliott (Eds.), At the threshold: The
attachment states of mind and styles with respect to developing adolescent (pp.  54–90). Cambridge, MA:
fathers and mothers. Developmental Psychology, 40(6), Harvard University Press.
1239–1247. King, S. (Writer). (1986). Stand by me [Motion picture].
Furman, W., Simon, V. A., Shaffer, L., & Bouchey, H. New York: Columbia Pictures.
A. (2002). Adolescents’ working models and styles Kobak, R. R., & Cole, H. (1994). Attachment and
for relationships with parents, friends, and romantic meta-­monitoring: Implications for adolescent au-
partners. Child Development, 73(1), 241–255. tonomy and psychopathology. In D. Cicchetti & S.
Gavin, L. A., & Furman, W. (1989). Age differences in L. Toth (Eds.), Rochester Symposium on Developmental
adolescents’ perceptions of their peer groups. Develop- Psychopathology: Vol. 5. Disorders and dysfunctions of
mental Psychology, 25, 827–834. the self (pp. 267–297). Rochester, NY: University of
Gavin, L. A., & Furman, W. (1996). Adolescent girls’ Rochester Press.
relationships with mothers and best friends. Child De- Kobak, R. R., Cole, H., Ferenz-­Gillies, R., Fleming, W.,
velopment, 67(2), 375–386. & Gamble, W. (1993). Attachment and emotion reg-
George, C., Kaplan, N., & Main, M. (1984). Adult At- ulation during mother–teen problem-­solving: A con-
tachment Interview protocol. Unpublished manuscript, trol theory analysis. Child Development, 64, 231–245.
University of California at Berkeley. Kobak, R. R., & Duemmler, S. (1994). Attachment and
George, C., Kaplan, N., & Main, M. (1985). Adult At- conversation: Toward a discourse analysis of adoles-
tachment Interview protocol (2nd ed.). Unpublished cent and adult security. In K. Bartholomew & D. Per-
manuscript, University of California at Berkeley. lman (Eds.), Advances in personal relationships: Vol. 5.
George, C., Kaplan, N., & Main, M. (1996). Adult At- Attachment processes in adulthood (pp. 121–149). Lon-
tachment Interview protocol (3rd ed.). Unpublished don: Jessica Kingsley.
manuscript, University of California at Berkeley. Kobak, R. R., Ferenz-­Gillies, R., Everhart, E., & Sea-
Grossmann, K., Grossmann, K. E., & Kindler, H. (2005). brook, L. (1994). Maternal attachment strategies and
Early care and the roots of attachment and partner- emotion regulation with adolescent offspring. Journal
ship representations: The Bielefeld and Regensburg of Research on Adolescence, 4(4), 553–566.
longitudinal studies. In K. E. Grossmann, K. Gross- Kobak, R. R., Rosenthal, N., & Serwik, A. (2005). The
mann, & E. Waters (Eds.), Attachment from infancy to attachment hierarchy in middle childhood: Concep-
adulthood: The major longitudinal studies (pp. 98–136). tual and methodological issues. In K. A. Kerns & R.
New York: Guilford Press. A. Richardson (Eds.), Attachment in middle childhood
Hamilton, C. E. (2000). Continuity and discontinuity of (pp. 71–88). New York: Guilford Press.
attachment from infancy through adolescence. Child Kobak, R. R., & Sceery, A. (1988). Attachment in late
Development, 71, 690–694. adolescence: Working models, affect regulation and
Hartup, W. W. (1992). Friendships and their develop- representations of self and others. Child Development,
mental significance. In H. McGurk (Ed.), Child- 59, 135–146.
hood social development: Contemporary perspectives Kobak, R. R., Sudler, N., & Gamble, W. (1991). Attach-
(pp. 175–205). Hillsdale, NJ: Erlbaum. ment and depressive symptoms during adolescence:
Hazan, C., & Shaver, P. R. (1987). Romantic love con- A developmental pathways analysis. Development and
ceptualized as an attachment process. Journal of Per- Psychopathology, 3(4), 461–474.
sonality and Social Psychology, 52, 511–524. Larose, S., & Bernier, A. (2001). Social support pro-
Hazan, C., & Shaver, P. R. (1994). Attachment as an cesses: Mediators of attachment state of mind and ad-
organizational framework for research on close rela- justment in late adolescence. Attachment and Human
tionships. Psychological Inquiry, 5, 1–22. Development, 3(1), 96–120.
Hesse, E. (1999). The Adult Attachment Interview: Lieberman, M., Doyle, A.-B., & Markiewicz, D. (1999).
Historical and current perspectives. In J. Cassidy & Developmental patterns in security of attachment to
P. R. Shaver (Eds.), Handbook of attachment: Theory, mother and father in late childhood and early adoles-
research, and clinical applications (pp. 395–433). New cence: Associations with peer relations. Child Devel-
York: Guilford Press. opment, 70(1), 202–213.
Hinton, S. E. (1967). The outsiders. New York: Viking Madigan, S., Moran, G., & Pederson, D. R. (2006). Un-
Press. resolved states of mind, disorganized attachment re-
434 IV. ATTACHMENT IN ADOLESCENCE AND ADULTHOOD

lationships, and disrupted interactions of adolescent Roisman, G. I., Madsen, S. D., Hennighausen, K. H.,
mothers and their infants. Developmental Psychology, Sroufe, L. A., & Collins, W. A. (2001). The coher-
42(2), 293–304. ence of dyadic behavior across parent–child and ro-
Main, M. (1999). Epilogue. Attachment theory: Eigh- mantic relationships as mediated by the internalized
teen points with suggestions for future studies. In representation of experience. Attachment and Human
J. Cassidy & P. R. Shaver (Eds.), Handbook of at- Development, 3(2), 156–172.
tachment: Theory, research, and clinical applications Rosenstein, D. S., & Horowitz, H. A. (1996). Adoles-
(pp. 845–887). New York: Guilford Press. cent attachment and psychopathology. Journal of
Main, M., & Goldwyn, R. (1984). Adult attachment scor- Consulting and Clinical Psychology, 64(2), 244–253.
ing and classification system. Unpublished manuscript, Seiffge-­Krenke, I., & Beyers, W. (2005). Coping trajec-
University of California at Berkeley. tories from adolescence to young adulthood: Links to
Main, M., Goldwyn, R., & Hesse, E. (2003). Adult at- attachment state of mind. Journal of Research on Ado-
tachment scoring and classification system. Unpublished lescence, 15(4), 561–582.
manuscript, University of California at Berkeley. Selman, R. (1980). The growth of interpersonal under-
Main, M., Kaplan, N., & Cassidy, J. (1985). Security in standing: Developmental and clinical analyses. New
infancy, childhood, and adulthood: A move to the York: Academic Press.
level of representation. In I. Bretherton & E. Waters Shaver, P., Hazan, C., & Bradshaw, D. (1988). Love
(Eds.), Growing points of attachment theory and re- as attachment: The integration of three behavioral
search. Monographs of the society for research in child systems. In R. J. Sternberg & M. Barnes (Eds.), The
development, 50(1–2, Serial No. 209), 66–104. psychology of love (pp. 68–99). New Haven, CT: Yale
Marsh, P., McFarland, F. C., Allen, J. P., McElhaney, K. University Press.
B., & Land, D. J. (2003). Attachment, autonomy, Slough, N. M., & Greenberg, M. T. (1990). Five-year-
and multifinality in adolescent internalizing and risky olds’ representations of separation from parents: Re-
behavioral symptoms. Development and Psychopathol- sponses from the perspective of self and other. New
ogy, 15(2), 451–467. Directions for Child Development, 48, 67–84.
McElhaney, K. B., Immele, A., Smith, F. D., & Allen, J. Spangler, G., & Zimmermann, P. (1999). Attachment
P. (2006). Attachment organization as a moderator representation and emotion regulation in adolescents:
of the link between peer relationships and adolescent A psychobiological perspective on internal working
delinquency. Attachment and Human Development, 8, models. Attachment and Human Development, 1(3),
33–46. 270–290.
Moore, C. W. (1997). Models of attachment, relationships Sroufe, L. A., Egeland, B., Carlson, E. A., & Collins, W.
with parents, and sexual behavior in at-risk adolescents. A. (2005a). The development of the person: The Min-
Unpublished doctoral dissertation, University of Vir- nesota Study of Risk and Adaptation from Birth to Adult-
ginia. hood. New York: Guilford Press.
Owens, G., Crowell, J. A., Pan, H., Treboux, D., Sroufe, L. A., Egeland, B., Carlson, E., & Collins, W.
O’Connor, E., & Waters, E. (1995). The prototype A. (2005b). Placing early attachment experiences in
hypothesis and the origins of attachment working developmental context: The Minnesota longitudi-
models: Adult relationships with parents and roman- nal study. In K. E. Grossmann, K. Grossmann, & E.
tic partners. In E. Waters, B. E. Vaughn, G. Posada, & Waters (Eds.), Attachment from infancy to adulthood:
K. Kondo-­Ikemura (Eds.), Caregiving, cultural, and The major longitudinal studies (pp. 48–70). New York:
cognitive perspectives on secure-base behavior and Guilford Press.
working models: New growing points of attachment Sroufe, L. A., & Waters, E. (1977). Attachment as an
theory and research. Monographs of the Society for Re- organizational construct. Child Development, 48,
search in Child Development, 60(2–3, Serial No. 244), 1184–1199.
216–233. Steinberg, L. (1990). Interdependency in the family:
Pearson, J. L., Cohn, D. A., Cowan, P. A., & Cowan, C. Autonomy, conflict, and harmony in the parent–­
P. (1994). Earned- and continuous-­security in adult adolescent relationship. In S. S. Feldman & G. El-
attachment: Relation to depressive symptomatology liott (Eds.), At the threshold: The developing adolescent
and parenting style. Development and Psychopathology, (pp. 255–276). Cambridge, MA: Harvard University
6(2), 359–373. Press.
Reimer, M. S., Overton, W. F., Steidl, J. H., Rosenstein, Steinberg, L. (2005). Adolescence. New York: McGraw-
D. S., & Horowitz, H. (1996). Familial responsive- Hill.
ness and behavioral control: Influences on adolescent Taylor, S. E., Klein, L. C., Lewis, B. P., Gruenewald,
psychopathology, attachment, and cognition. Journal T. L., Gurung, R. A. R., & Updegraff, J. A. (2000).
of Research on Adolescence, 6(1), 87–112. Biobehavioral responses to stress in females: Tend-
Roisman, G. I., Collins, W. A., Sroufe, L. A., & Egeland, and-­befriend, not fight-or-­flight. Psychological Review,
B. (2005). Predictors of young adults’ representations 107(3), 411–429.
of and behavior in their current romantic relation- Thompson, R. A. (1997). Sensitivity and security:
ship: Prospective tests of the prototype hypothesis. New questions to ponder. Child Development, 68(4),
Attachment and Human Development, 7(2), 105–121. 595–597.
19. The Attachment System in Adolescence 435

Tracy, J. L., Shaver, P. R., Albino, A. W., & Cooper, M. Weimer, B. L., Kerns, K. A., & Oldenberg, C. M. (2004).
L. (2003). Attachment styles and adolescent sexual- Adolescents’ interactions with a best friend: Associa-
ity. In P. Florsheim (Ed.), Adolescent romantic relations tions with attachment style. Journal of Experimental
and sexual behavior: Theory, research, and practical im- Child Psychology, 88(1), 102–120.
plications. (pp. 137–159). Mahwah, NJ: Erlbaum. Weinfield, N. S., Sroufe, L. A., & Egeland, B. (2000).
Treboux, D., Crowell, J. A., & Waters, E. (2004). When Attachment from infancy to early adulthood in a
“new” meets “old”: Configurations of adult attach- high-risk sample: Continuity, discontinuity, and their
ment representations and their implications for correlates. Child Development, 71, 695–702.
marital functioning. Developmental Psychology, 40(2), Weinfield, N. S., Whaley, G. J. L., & Egeland, B. (2004).
295–314. Continuity, discontinuity, and coherence in attach-
van IJzendoorn, M. H. (1996). “Attachment patterns ment from infancy to late adolescence: Sequelae of
and their outcomes”: Commentary. Human Develop- organization and disorganization. Attachment and
ment, 224–231. Human Development, 6(1), 73–97.
Wallis, P., & Steele, H. (2001). Attachment represen- Zimmermann, P. (2004). Attachment representations
tations in adolescence: Further evidence from psy- and characteristics of friendship relations during
chiatric residential settings. Attachment and Human adolescence. Journal of Experimental Child Psychology,
Development, 3(3), 259–268. 88(1), 83–101.
Ward, M. J., & Carlson, E. A. (1995). Associations Zimmermann, P., & Becker-Stoll, F. (2002). Stability of
among adult attachment representations, mater- attachment representations during adolescence: The
nal sensitivity, and infant–­mother attachment in a influence of ego-­identity status. Journal of Adoles-
sample of adolescent mothers. Child Development, 66, cence, 25(1), 107–124.
69–79. Zimmermann, P., Maier, M. A., Winter, M., & Gross-
Waters, E., Hamilton, C. E., & Weinfield, N. S. (2000). mann, K. E. (2001). Attachment and adolescents’
The stability of attachment security from infancy to emotion regulation during a joint problem-­solving
adolescence and early adulthood: General introduc- task with a friend. International Journal of Behavioral
tion. Child Development, 71(3), 678–683. Development, 25(4), 331–343.

Você também pode gostar