Chapter 30 The Circle of Security

CHAPTER 30
All rights reserved. May not be reproduced in any form without permission from the publisher, except fair uses permitted under U.S. or applicable copyright law.
The Circle of Security
Joe Coyne
Bert Powell
Kent Hoffman
Glen Cooper
I n the third edition of this handbook, the au-

thors of this chapter used Oliver Wendell
sense. Rather, we feel that parent is a term that
represents the emotional intensity and salience
Holmes’s quotation about the “simplicity on the of being the “go-to” person for children’s needs,
other side of complexity” as a goal for trans- and we use that term for such people irrespec-
lating attachment research into sound, well- tive of biological connection.
designed models of intervention (Powell, Coo-
per, Hoffman, & Marvin, 2014). This remains
a focus and strength of the Circle of Security Communicating Attachment within
(COS) model in providing early intervention to the COS Model
support parents to be a secure base/safe haven
for their young children. The distillation of complexity into simplicity
We begin this chapter by reviewing the COS is largely achieved through the COS graphic,
model, highlighting the importance of the COS which provides a simple pictorial representation
graphic (Figure 30.1), and its origins in the COS of the attachment system as described by Bowl-
Intensive (COS-I) program. Next, we review by (1969/1982). The two key functions of the par-
the underpinning principle of relation-based ent, to act as a secure base and a safe haven, are
therapy and the core skills for effective deliv- represented by the Hands. The COS graphic acts
ery of any COS intervention. Then we present as a map by which the parent learns to identify
some background on a recent development in the child’s attachment needs by tracking where
the COS project with the development of the on this Circle the child is during a particular
Copyright @ 2019. The Guilford Press.
Circle of Security Parenting (COS-P) program. interaction. As parents learn to read the child’s
This parenting program implements the core cues and reflect on what is needed, they develop
principles by reflecting on specifically prepared the capacity to respond in ways that meet the
video material. In the final section, we consider need (Cooper, Hoffman & Powell, 2009).
the current evidence for the various forms of the In their capacity as a secure base, parents
intervention and directions for future research support the child on the top half of the Circle
and development. Throughout this chapter, we in exploring the world and developing his or
prefer the word parent to primary caregiver or her capacity for autonomy and mastery. As the
carer. This is not intended to ignore that many safe haven, parents provide support for the child
significant people meeting the needs of chil- when he or she has a need to seek proximity, ei-
dren are not biological parents in the traditional ther when feeling threatened or through a desire
500
EBSCO : eBook Collection (EBSCOhost) - printed on 7/2/2019 3:16 PM via TUFTS UNIV
AN: 1843598 ; Zeanah, Charles H..; Handbook of Infant Mental Health, Fourth Edition
Account: s3579704.main.ehost
30. The Circle of Security 501
FIGURE 30.1. Circle of Security®: Parent attending to the child’s needs. Copyright © 1998 Cooper, Hoffman,
Marvin, and Powell.
to reconnect. The capacity of a parent to fill this as the primary learning tool for parents to see,
dual role in relation to the child’s attachment understand, and support their children’s attach-
needs is considered to promote security within ment and exploratory needs. The Circle is used
the attachment relationship (Ainsworth, Blehar, conceptually in all types of COS interventions.
Waters, & Wall, 1978).
When the parent does not provide the full
range of these capacities to the child, then The COS-I Protocol
“Limited Circles of Security” (see Figure 30.2)
emerge that are analogous to insecure catego- The intensive COS intervention aims to apply
ries of attachment (Cooper, Hoffman, Powell, & knowledge derived from attachment theory and
Marvin, 2005). research, as well as psychodynamic models of
When a parent has difficulty meeting a intervention, to enhance parental reflective ca-
child’s needs on the top of the Circle (i.e., being pacity. COS-I involves a 20-week therapy group
a secure base for exploration), then the child is program for up to six primary parents of a child
less secure in regard to separation and auton- age 12 months to 6 years meeting weekly for
omy, the insecure style identified by research around 75 minutes per session. The structure of
as insecure ambivalent/resistant. If a parent has these sessions is a mix of psychoeducation and
a limited ability to act as a safe haven for the psychotherapy.

child and to clearly welcome the child coming
back, then the child will be less secure in regard
Psychoeducation in COS-I
to intimacy and proximity seeking, the insecure
style identified in research as insecure–avoid- A core goal of the program is for the parents to
ant. The various elements described on the develop a clear knowledge of their child’s needs
Circle are derived from attachment research for secure base and safe haven dimensions, and
and theory, and expressed in a way to facilitate how to be the “Hands on the Circle.” Good-
communication to parents, including those in enough parenting represents the effective amal-
disadvantaged populations (Powell, Cooper, gam of being bigger, stronger, and kind, while
Hoffman, & Marvin, 2014). The Circle serves simultaneously possessing the wisdom to bal-
502 V . I ntervention
FIGURE 30.2. Limited Circles: Child attending to parent’s needs. Copyright © 1998 Cooper, Hoffman, Marvin,
and Powell.
ance these functions, allowing the child to feel The COS-I Assessment Process
secure within the relationship. It is possible for
A distinctive and core feature of COS-I is the
parents to get out of balance in providing Hands
functions by being overly harsh (mean), ineffec- application of attachment theory and research
tive (weak), or psychologically absent (gone). within a structured assessment process for
Good-enough parenting is also the ability to understanding a child’s and parents’ specific
follow the child’s need wherever possible but to strengths and struggles within relationship.
take charge and organize the child/relationship Therapists use the COS graphic as a guide in
whenever necessary. the assessment of relationship needs with the
COS helps parents see that children who can parent–child dyad. The aim of this assess-
use their caregivers all around the Circle are ment process is to determine what we call the
more secure in their attachment, and this secu- “linchpin struggle,” which is the core interac-
rity confers many advantages across the course tional process and accompanying state of mind
of development. Lyons-Ruth (2003) likens se- in the parents that maintains insecure and dis-
cure attachment to developing a “psychological organized relational patterns. To achieve this,
immune system”—a coping mechanism that al- COS-I uses an interactional assessment, as well
lows the child to use self and others to recover as an assessment of the parents’ state of mind.
from distress effectively and maintain adapta-
tion across time. Substantial evidence has ac- COS Interactional Assessment
crued that children who are securely attached
function better across a broad range of develop- The Interactional Assessment focuses on the
mental outcomes (Kobak, Cassidy, Lyons-Ruth, three categories of attachment interaction
& Ziv, 2006; Siegel, 2012; Sroufe, 2016; Sroufe, within the parent–child relationship: “top half
Egeland, Carlson, & Collins, 2005). of the Circle,” “bottom half of the Circle,” and
“Hands” (see Figure 30.1). The interactional moments of shared delight are more difficult to
assessment is an adaptation of the Ainsworth see during the interactional assessment. We see
Strange Situation Procedure (SSP), using ei- delight as a parental expression that is all about
ther the protocol for children younger than 24 loving the child for who he or she is. Having
months (Ainsworth et al., 1978) or that for chil- your accomplishments as a child validated by a
dren 24–60 months of age (Cassidy & Marvin, parent’s positive response is helpful; however,
1992). The SSP consists of a series of seven if this is the only delight a child sees in the par-
episodes, each lasting approximately 3 minutes, ent’s face, then the child’s self-esteem is at risk
with episodes designed to emphasize explor- of being contingent on only what he or she does.
atory or attachment needs in the child. We add Seeing a face that shows delight because of who
two further episodes to the original SSP pro- the child is provides the core of a noncontingent
tocol: a book reading (5 minutes) and clean-up feeling of self-worth. It is also important that
of the toys (3 minutes). The latter two episodes the child see this expression both when explor-
are clinically useful, as the book reading invites ing the world and building accomplishment,
moments of asking the parent to organize an and when he or she is seeking closeness and
activity that can promote closeness and enjoy- building intimacy. This supports a robust sense
ment. The clean-up provides an opportunity to of self that is sustaining both when life is going
observe a parent’s ability to take charge in an well and when there are struggles. Most inse-
everyday parenting experience. (See Powell et cure parents found delight to be in short sup-
al., 2014, for details of the procedure.) ply during their own development and therefore
lack an internal register of this emotional expe-
COS Needs. The assessment clarifies the rience to provide it for their child. Increasing
dyad’s strengths and its struggles negotiating and supporting the presence of delight is a key
needs on the Circle. A key question is whether marker of success within a COS-I intervention.
the child can clearly show need by giving a di-
rect “cue” or is there evidence that when need Top Half of the Circle. On the top of the
is present, the child has learned to “miscue” Circle there are five secure-base needs that
(Powell et al., 2014) as a way of managing the promote uninhibited exploration (see Table
relationship with the parent. A cue within the 30.1). Is the child supported by the parent in the
COS protocol is a direct or indirect signal from transition from proximity to the parent to ex-
the child for a specific need to be met. A mis- ploring the environment? This support can be
cue conceals a need and protects the child when demonstrated by subtle signals, including facial
the need is one that parents may find difficult to expression and tone of voice, that serve to reas-
meet because of the emotional struggle it may sure the child that he or she can engage with the
activate in them (Cooper et al., 2005). Insecure world, knowing the parent is there for him or
children are expected to be more reliant on mis- her. As the child moves away from the parent,
cues to manage relationship, notably, when the what emotional tone does the parent communi-
attachment behavioral system is activated. For cate? Calm or anxious? Does the parent become
example, a child who is clearly upset by a par- withdrawn, rejecting, or intrusive when the
ent’s separation in the SSP has a need for com- child shows interest in the play environment?
fort. If, during the reunion, instead of clearly How does the transition from being close to the
showing this need, the child turns away from parent to moving out to explore affect the child?
the parent to play with a toy, there is a miscue Can the child cue the parent directly about
that serves to manage the relationship by keep- needs related to an interest in the environment?
ing both participants comfortable within the

parameters of what they have learned is accept- Bottom Half of the Circle. There are also
able in the relationship. five safe-haven needs on the bottom of the Cir-
“Delight in me,” a need that is found on both cle (see Table 30.1). Is the child welcomed by
the top and the bottom of the Circle, is consid- the parent in the transition from exploring the
ered an important marker for positive relating world and returning to proximity? Exploration
that builds security. The presence of delight in cannot last forever, and a child will eventually
a dyad can be considered a positive prognostic be disturbed by something in the environment
sign, as this mutual, positive affectivity sug- or run low in the “emotional fuel” that only con-
gests a core healthiness within the pair. In dyads nection with a parent can provide. At that point,
in which there is significant relational struggle, the child may feel the need to seek proximity
TABLE 30.1. Needs around the Circle

Top of the Circle Bottom of the Circle
Support my Exploration: A child’s need to have his or Welcome my coming to you: A child’s need to
her parent encourage and support the child’s developing reconnect with an approachable parent when
autonomy in mastering his or her world. exploration has been curtailed for any reason.
Watch over me: A child’s need to know that his or her Protect me: A child’s need to be protected physically
parent is present and available for the child as he or she and emotionally, allowing him or her to feel safe.
explores the world.
Delight in me: A child’s need to feel that he or she is Comfort me: A child’s need for soothing and comfort
pleasing to the parent just for who the child is as he or from the parent when experiencing distress.
she explores.
Help me: A child’s need for a parent to scaffold learning, Delight in me: A child’s need to feel that he or she is
pitching the parent’s help at the level required to build pleasing to the parent just for who the child is as he or
the child’s competence, navigating the extremes of she seeks nurturance.
taking over or leaving the child on his or her own.
Enjoy with me: A child’s need to feel that exploration is Organize my feelings: A child’s need for help in
a shared and enjoyable experience for the parent also. regulating intense emotions in which a parent supports
by being present with the child’s intense feelings.
and closeness to the parent as an emotional safe ments, a parent may abdicate and look helpless
haven. When this happens, how does the par- or, conversely, become punitive and aggressive,
ent respond? As with the transition into explo- using fear to get the child to comply. At other
ration, the parent’s signals to the child can be times, a child needs the parent to function as the
quite subtle. Being able to clearly cue the parent Hands to alleviate uncertainty and anxiety. If in
about these needs makes for a simple transition, these moments the parent abdicates by acting
without undue stress for the child, thus reducing helpless, disengaged, neglectful, or rejecting,
distress and promoting security. the child is left without resource. Similarly, par-
ents who fear strong emotion in their children
Hands. The “Hands” function is to be “big- (or themselves) may be tentative in interactions,
ger, stronger, wiser, and kind,” fulfilling safe- leading a child to feel that he or she must direct
haven and secure-base functions reliably, while the relationship in lieu of the parent’s confident
ordering the basic structure of relationship presence. For example, a child upon reunion
through sensitively following a child’s needs may try anxiously to soothe the parent when the
and taking charge confidently when required. child is the one who is upset. A more subtle ex-
Deficits in the ability to be the Hands for chil- ample might be a child upon reunion who tries
dren are emblematic of what, in attachment to keep a parent happy by being amusing, acting
research, are classified as disorganized attach- “cute,” or “putting on a show.”
ment relationships (Cassidy & Marvin, 1992; Role distortion/reversal is a common fea-
Main & Solomon, 1990). Parent behavior in dis- ture of disorganized dyads in which the child
organized relationships is frightening (Mean), takes on the developmentally inconsistent task

frightened (Weak), or disengaged or unreliable of providing emotional functions that are the
(Gone). province of the parent. This may take the form
There are a number of questions that assist a of disruption in the hierarchical relationship,
clinician in determining whether “Hands” are in which a parent may act more as a playmate,
an area of struggle for the parent. Is it the parent peer, or use the child as a confidant. Other par-
or the child who takes the greatest responsibil- ents may want the child to be completely in-
ity for organizing the relationship? Clean-up, a dependent, with no need for the parent. In all
section of the COS-I interactional assessment, of these examples, the adult looks to the child
is designed to elicit a “take charge” situation to help him or her feel emotionally stable. The
for the parent, and it is illuminating if the par- child associates the activation of attachment
ent lacks capacity here. In “take charge” mo- needs with fear and copes by taking control
of the relationship. The child can do this by orders of personality on the other. In the COS
being controlling–punitive, acting aggressive model, a core sensitivity includes defensive
and dominant toward the parent, or by being views of self and others that apply to parents
controlling–caregiving and soothing, amusing, seeking treatment, but in no way are derogatory
and taking care of the parent. In short, the ques- or judgmental.
tion of whether a parent can provide the role Each core sensitivity represents a set of non-
of Hands is central to the development of an conscious procedures that a person believes
effective treatment plan to support the child must be adhered to in order to avoid painful
and have enough predictability, structure, and memories of previously unregulated affect
safety to form an organized attachment. (“When I show I am smart, she stays nearby,
so I’ll keep doing that. When I ask for comfort
COS State of Mind Assessment she gets chilly or backs away, so I’ll stop ask-
ing”). The need for connection and the fear of
We have based the COS model on the recog- abandonment are recognized to be the central
nition that all parents use what John Bowlby organizing processes in the development of de-
called an “internal working model” (IWM; fensive personality structure (Masterson, 1976).
Bretherton & Munholland, 2016). IWMs are Even though these nonconscious rules are ame-
considered a crucial element in the transmis- nable to change, without reflection and support
sion of attachment, as this internal template is they tend to remain constant and to drive prob-
used to understand current and future relation- lematic parent–child interactions.
ships, including those with children in our care. The COS assessment of core sensitivity is
We combine Bowlby’s IWM concept with the derived from the parent’s responses during
work of James Masterson (1976; Masterson & the standardized Circle of Security Interview
Klein, 1995), Otto Kernberg (1975) from ob- (COSI). The following is a brief description to
ject relations theory, and the self psychology of help clarify the nature of these sensitivities:
Heinz Kohut (1971) to facilitate the differential
assessment of each parent’s specific relational • Separation-sensitive parents have come to
model. This understanding is predicated on the believe that to avoid experiences of perceived
belief that each parent approaches attachment abandonment, they must comply with what oth-
relationships with a particular lens regarding ers, including their children, want, need, and
sense of self and sense of other. This lens, first feel, while disavowing their own wants, needs,
learned within significant past relationships, and feelings. The underlying belief is that if
becomes the viewpoint that organizes current they feel and act on their own behalf, they will
interaction on the part of the parent with his or be abandoned by those they most need. They
her child. Having a series of potential templates tend to feel incapable of living without feeling
to understand a given parent’s worldview con- the continual availability of significant others.
cerning relationship is useful for effective inter- They often believe that their job is to focus on
vention, particularly when the therapeutic focus another’s needs and appear to be helpless re-
is on problematic representations that the parent garding their own. Separation-sensitive parents
holds regarding both self and child. struggle to allow their children an experience of
The COS model focuses on distinguishing autonomy while exploring (the top of the Circle)
among defensive strategies rather than using and simultaneously are unable to genuinely
DSM-5 terminology for personality disorders. support a full experience of comfort (the bottom
Thus, we label three distinct but predictable half of the Circle). The children tend to show
patterns to be “core sensitivities” and describe an ambivalent/resistant response upon reunion

them in the following way: separation sensitive during the SSP.
(which in a rigid and pervasive form can be-
come borderline personality disorder), esteem • Esteem-sensitive parents have come to be-
sensitive (which in a rigid and pervasive form lieve that who they are, just as they are (imper-
can become narcissistic personality disorder), fect, flawed, average), is not enough to be val-
and safety sensitive (which in a rigid and perva- ued. Therefore, to protect themselves from the
sive form can become schizoid personality dis- fear of criticism and judgment, they continually
order) (Masterson & Klein, 1995; Powell et al., attempt to prove that they and their children
2014). The core sensitivities form a continuum are worthy (unique, special, exceptional, any-
from flexible and adaptive (mildly defensive) thing but average, etc.) through performance
strategies on one end, to rigid and pervasive dis- and achievement. Perceptions feel all impor-
tant for the esteem-sensitive parent. Hence, sensitivity as a means for the parent to protect
they tend to be vigilant about any implication him- or herself from unregulated affect evoked
of having failed or being inadequate as parents, by attachment experiences. By better under-
or having others see their child as inadequate. standing precisely what defensive approach is
These parents often pressure their children for being used by any given parent, the clinician is
achievement and performance (the top half of able to design and aim the intervention at the
the Circle) and struggle to provide comfort and precise struggle most in need of change. We
organization of feelings (the bottom of the Cir- see this individualized understanding of parent
cle). These children can present as avoidant or state of mind, coupled with an appreciation of
ambivalent during the SSP. the specific dyad’s strengths and struggles, as
the core premise of providing treatment in the
• Safety-sensitive parents believe that the COS-I model.
cost of being connected and emotionally close to
their children is the loss of self-determination.
This leads to perceptions of being controlled Individualized Treatment Planning in COS-I
and/or intruded upon by their children. There- The clinical value of COS-I is based on this
fore, they believe that the only way to have an capacity to provide individualized treatment
intact sense of self is to maintain a position of plans for each parent–child dyad. Cooper and
self-sufficiency, and to expect self-sufficiency colleagues (2005) described assessment of the
from their children. The children of safety-sen- strengths and struggles of the dyad on the top
sitive parents are often frustrated by how their of the Circle as the key task of the therapist in
parent continually seeks to control the level of treatment planning the top of the circle, and the
closeness. These children are often forced onto parent’s capacity to provide the “Hands” for
the top of the Circle (toward self-sufficiency) the child on the bottom of the Circle. One key
when their genuine need is for comfort and con- struggle then comes to be the “linchpin” focus
nection on the bottom of the Circle. The result- during treatment. A central task of the group fa-
ing lack of fulfilment in the child often leads cilitator is to maintain the treatment focus on the
to demands that reconfirm the parent’s view of linchpin for each parent, assisting the parents in
significant others as demanding and control- identifying their child’s needs and reflecting on
ling. These children can present as avoidant or how well their parenting meets those needs.
ambivalent during the SSP. This linchpin struggle stabilizes the relation-
ship by protecting both parent and child from
The COS approach uses the COSI as a way unregulated emotional distress; however, this
to gain initial insight into a parent’s core sen- comes at the expense of secure connection
sitivity. While a decision tree for building an around the needs of the child. It is within the
accurate differential assessment is too complex linchpin struggle that the child’s miscueing of
for this chapter, answers offered to the question the parent and the parent’s miscueing of the
“How do you think your child responded to the child has been learned, becomes reinforced,
separation?” can provide insight into the intra- and helps sustain the defensive struggle. It is
psychic world of the parent. A more esteem- important to bear in mind that this linchpin is in
sensitive parent might answer with a sense of large part a procedural pattern, and thus oper-
pride that his or her child played creatively and ates outside of the dyad members’ awareness.
was not upset. A safety-sensitive parent might The COS-I model uses a range of therapeutic
comment about how he or she is relieved to see attitudes and skills to deliver this individual-
his or her child being so self-sufficient. A sepa- ized treatment plan within the group context.
ration-sensitive parent might mention that he or We recommend Powell and colleagues (2014)
she is always worried that his or her child does for a complete discussion of the COS-I treat-
not really need him or her. Greater detail on the ment process.
COSI, sensitivities, and differential assessment
are available in Powell and colleagues (2014).
In conclusion, the COS model uses a system-
Relationship-Based Therapy Processes in COS
atic differential diagnosis of each parent’s non- The therapeutic relationship is crucial to the
conscious defensive schema as a means to bet- success of any COS intervention. Parenting is
ter understand the core themes for intervention. an intimate relationship in which significant
More specifically, we understand each core vulnerability is present. The therapist needs to
possess and model a style of sensitive engage- willing and able to “hold” and “be with” our
ment to successfully invite the parents to expe- emotional needs. Feelings that have chronically
rience their vulnerability in the group process. interfered with parent’s ability to respond to his
In the early stages of the protocol, parents are or her child can now be “held” by the therapist,
asked to make observations and conclusions the group, and the parent. This experience of
about their child’s needs and feelings. Later, shared understanding and coregulation of dif-
they are asked to focus more deeply on their ficult emotion lays the foundation for the parent
own experience, including moments of difficul- to provide more secure parenting.
ty in their relationship with the child in which
• The last “R” stands for reflection, the
they may experience distress when called on to
capacity to reflect on both the child’s and the
meet the child’s attachment needs.
parent’s thoughts, feelings, needs, and behav-
Just as parents come to use the Circle graphic
ior, and ultimately make new choices. Fonagy,
as a map for relationship, so too does it become
Gergely, Jurist, and Target (2002) propose that
a guide for therapists to become a secure base
secure relationships are the optimal environ-
and safe haven for the parent, providing an
ment in which children can come to understand
overarching organization for relationship-based
minds and mental states, both their own and
treatment and intervention. Only through pro-
other people’s. By providing a secure, holding
gressing with a deeply respectful and attuned
environment within the context of the group,
approach to each parent can the therapist cre-
where parents can safely experience their pro-
ate the requisite feelings of safety, support, and
cedural struggles with affect in relationship, the
understanding for the parent to work effectively
therapist provides conditions in which a reflec-
in the group process. We see the principles of
tive stance can be taken toward relational infor-
treatment are encapsulated by the acronym
mation. In this way, aspects of relationship that
R-A-R:
were “hidden in plain sight” can become known
to the parent and facilitative of change in the
• The first “R” returns us to the importance
dyadic relationship (Powell et al., 2014).
of the relationship between therapist and par-
ent. Many parents seeking assistance to provide
The primary process of change in the COS
better care for their children, especially those
model may be summed up as regulating affect
parents considered high-risk, have few experi-
in relationship while facilitating reflective un-
ences in their own developmental histories that
derstanding. This process is a model of how
support having trust and confidence in another.
we want to encourage parents to relate to their
How then can such parents be expected to use
child. In COS, this is referred to as “being-with”
the support and assurance of a therapist to be
and refers to parents’ ability to tune in and be
vulnerable and to learn? It is for this reason that
available to their child in whatever affective
the essential initial step of all parent–child psy-
state is being experienced. In this way, coregu-
chotherapy is building a therapeutic alliance, a
lation is central to both the therapeutic relation-
safe and trustworthy relationship between the
ship and the shift that is sought for the dyad in
parent and the therapist. The therapist and the
treatment.
group become a secure base from which the
In all forms of COS treatment, video review
parent can explore his or her relationship with
is a powerful tool for helping parents to gain ac-
the child.
cess to information that may currently lie out-
• The “A” stands for affect regulation (Cas- side their awareness. The editing and selection
sidy, 1994). Each parent brings to the group his of appropriate video clips is considered crucial
or her own unique history of how emotions are to the impact of the COS-I program. The ini-
managed. During the group, some aspect of how tial SSP is examined closely by trained program
parents manage chronic painful affects will in- providers to determine the linchpin struggle
evitably be challenged. British pediatrician and for the dyad (with supporting information from
psychoanalyst Donald Winnicott (1965/1990) the COSI). The SSP then becomes the source
used the term “holding environment” to de- material for tape reviews in the intervention
scribe the central emotional requirement of all protocol. These reviews occur in two phases.
children and adults to face these painful af- The Phase 1 tape reviews seek to help parents
fects. Winnicott clarified that each of us needs identify the linchpin issue and the needs of their
the sensitive availability of another, someone child on the Circle, particularly how their child
may “miscue” them in relation to his or her at- ent. This enhanced reflective capacity is pro-
tachment needs. Secure relationships are char- posed, in line with research and theory, to allow
acterized by clear communication or “cues” be- parents to build new understandings of relation-
tween parent and child about attachment needs. ship through modification of prior implicit rela-
A chronic miscue, on the other hand, occurs tionship scripts.
in insecure dyads, in which the child and par- Prior to Phase 2 tape reviews, we use a simple
ent have learned a habitual way of relating to two-part video clip set to music to help parents
minimize anxiety in the relationship. The tape make sense of how our procedurally organized
reviews with each parent in this first phase of state of mind effects perception. The first part
treatment are designed to individualize the par- opens with a beautiful ocean view. This clip is
ents’ use of the COS as a map to understand set to music that has a soft tone and tends to
their child within the attachment relationship. elicit a calm, pleasant, safe feeling. The second
During these reviews, parents practice creating part of the clip uses the same videotape but is set
behavioral descriptions of the observed interac- to a musical composition similar to the theme
tions. The rationale for focusing on behavioral music from the movie Jaws and tends to evoke
observation is that parents’ procedural histories quite different feelings. Suddenly, the beach
of relationship lend them to make inferences is transformed into an eerie place with hidden
or “guesses” about relationship that are often danger. The beach evokes a sense of foreboding
defensive and inaccurate with respect to their and a strong desire to flee from the water. Par-
child’s needs. By encouraging them to develop ents quickly grasp that the music dramatically
an observational skill for behavior in the dyadic shifts the mood of these two identical visual
process, we aim to calibrate their view to seeing experiences. Thus, “shark music” is a metaphor
the attachment needs of their child, what is hid- for an experience with a child that is frighten-
den in plain sight. This process, called “seeing ing but not dangerous. Parents are told that it is
and guessing,” becomes the basic procedure for normal to experience “shark music” with some
viewing videotapes throughout the group. Once of their child’s needs on the Circle, and what is
parents can see more clearly, it is expected that important is the ability to be able to notice and
the inferences or guesses they make regarding reflect on it.
their child’s needs as represented by his or her Phase 2 reviews are focused on parents’
behavior will become more accurate and facili- reflection on their internal signals of “shark
tate secure relating. music” as they review the linchpin moments
To assist this process, the following sequence in their relationship. First, these signals are hy-
of questions is used throughout the tape reviews pothesised to originate from struggles in the
in COS-I. These questions form the skeleton of parents’ own developmental histories that are
what is referred to in the model as “reflective now triggered in emotionally salient interactions
dialogue”: with their child. The term “shark music” was
created to help parents recognize two aspects
• “What is your child doing?” (observation) of parenting and state of mind (SoM): (1) not all
• “What do you think your child is needing?” parents view the same behavior in the same way
(inference) and (2) for some, a behavior or need of the child
• “What do you think your child is feeling?” will trigger anxieties based in this procedural
(inference) history of relationship. Second, these anxieties
• “What are you doing?” (observation) are actually fears that emerge in the context of a
• “What are you feeling?” (self-reflection) situation that is now safe: meeting their child’s
• “What are you needing in this moment?” attachment needs. This is in contrast to the situ-
(self-reflection) ation in which the parents learned to fear the
• “What are you thinking about yourself as expression of need within their own attachment
you watch this?” (self-reflection) (Powell et history. The video segments reviewed in this
al., 2014, p. 261) phase are used to support the parent “turning
themselves in” by sharing their awareness of
The idea of reflective dialogue guides the fa- their responses within this SoM and to find new
cilitators’ behavior and communication through ways of being with their shark music as they re-
all interventions in the model. It refers to an at- late to the child.
titude and process of therapeutic conversation As trust builds in the group it is our hope that
that builds the reflective functioning of the par- parents will feel safe enough to manage the vul-
nerability of reflecting on their linchpin “shark gram concepts and develop insight into the at-
music.” This helps parents develop a pause in tachment-related scripts each parent is using to
the automatic feelings and behaviors embed- understand the child and the relationship. This
ded in their insecure procedural knowledge and information becomes key within the briefer
opens the possibility for emotional regulation form of the model for providing a degree of indi-
and choice. The new choice is to either con- vidualized treatment to the parent. Later in the
tinue to defend against nonexistent “sharks” or program, this becomes more relevant as parents
to manage their fear by reflection and “choose are introduced to shark music and will need the
security” for their child. support of the facilitator to see this influence in
their parenting relationship. To further support
parents in making sense of relationship, Hoff-
The COS-P Protocol man, Cooper, and Powell (2017) have recently
published a book that acts as a further resource
Cooper and colleagues (2009) developed a brief to support developing secure connection.
variant of COS-I that is intended to be more us-
able in primary care and community contexts.
The COS-P protocol is based on a prerecorded The Evidence Base for COS
video, with the material organized into eight
“chapters” that can be delivered in a flexible Hoffman, Marvin, Cooper, and Powell (2006)
manner. COS-P is designed to be used with reported on the initial COS-I intervention study
families over 8–10 sessions and can be deliv- that assessed effectiveness with a preinterven-
ered individually with parents or in a group tion–postintervention design using the SSP
format. COS-P retains the flexibility to be de- (Ainsworth or MacArthur). Of the 75 dyads that
livered in the home, clinic, or other pertinent began the program, 65 (86%) parents completed
settings, such as child care and school. the entire intervention. The hypotheses tested
Video review remains a central pillar of the were that after intervention there would be (1) a
COS-P program and in the program is provided significant decrease in disorganized attachment
by video modeling using footage of families who and (2) a significant increase in secure attach-
have participated in the program, who have vol- ment. Disorganized attachments (controlling
unteered their time, and, in some cases, actors and insecure/other) decreased from 60% pre-
(Cooper et al., 2009). This video material is po- treatment to 25% posttreatment, and secure at-
tentially less psychologically immediate to the tachments increased from 20% pretreatment to
parent, as it is communicating the key informa- 54% posttreatment—both significant changes.
tion of the treatment protocol via third-person Of the 13 children who were classified as se-
material unlike COS-I, which uses the parents’ cure on the preintervention SSP, 12 remained
own interactions with their child, offering a secure on the postintervention SSP. This stabil-
first-person account of relationship. Reflective ity of security suggests that the intervention not
dialogue remains central for processing video only does no harm but it also provides needed
material with parents in COS-P. Parents are en- support in high-risk populations for secure
couraged through facilitator-guided reflection dyads to remain secure. A key limitation of the
in the sessions to look for the key relationship study was the absence of a control group and a
needs being depicted and to develop an under- randomized design.
standing of the core concepts of the model. In Recent research in Australia has provided ad-
developing an appreciation of the attachment ditional data about efficacy. Huber, McMahon,
needs of their child through this third-party and Sweller (2015a) reported on the outcomes
presentation, the parent then does have to make of 83 parent–child dyads referred to a commu-
a relational leap to transfer learning developed nity clinic for concerns relating to the child’s
in the sessions to their own dyadic relationship. behavioral and/or emotional needs, or for par-
Without the individualized treatment process ent–child relationship issues. Four outcomes
and the first-person clips, the facilitator needs to were assessed: positive representations of care-
have an attuned awareness to how each parent is giving, frightening and frightened statements,
engaging with the presented material. Through reflective functioning, and child attachment
listening closely to parents’ progressive stories classifications (and dimensions of attachment).
of relating to their child, the facilitator needs to Results showed that the intervention was
gauge how well parents have understood pro- effective in this clinically referred sample if
participants completed the intervention. Care- dimensions (which had been rotated to form a
giver reflective functioning, caregiving repre- secure vs. fearful attachment dimension and a
sentations, and level of child attachment secu- dismissing vs. preoccupied attachment dimen-
rity increased after the intervention, and level sion) showed important interaction effects on
of attachment disorganization decreased for treatment. Results showed that highly irritable
those with high baseline levels. Changes from infants of highly secure mothers benefited
pre- to posttreatment were explained largely by significantly more from intervention than did
improvements in caregivers and dyads who had moderately irritable infants of highly secure
exhibited most disturbance on the various indi- mothers. When mothers were more dismissing,
ces of relationship functioning at pretreatment. highly irritable infants benefited significantly
Subsequent analyses examined the impact of from intervention compared to their highly ir-
COS-I on measures of behavioral and emotion- ritable counterparts in the control group. In
al outcomes drawn from the sample mentioned contrast, a treatment effect emerged only for
earlier (Huber, McMahon, & Sweller, 2015b). moderately irritable, and not for highly irri-
Broadly, the findings indicated that the inter- table, infants of highly preoccupied mothers.
vention produced significant improvement in Again, this study is suggestive of the benefits of
promoting protective factors and in reducing in- COS as an intervention for populations in which
ternalizing and externalizing behaviors, based risk is higher. For a full discussion of both these
on parent report. Once again, the parents who interventions, see Woodhouse, Powell, Cooper,
showed the most initial risk in terms of repre- Hoffman, and Cassidy (2018).
sentations of relationships, seemed to acquire
the greater benefit from participating in COS-I.
The COS-P Program
In another study, the COS Perinatal Protocol
(COS-PP) was delivered as a component of the Research on COS-P program has started to
Tamar’s Children jail diversion program with emerge, which is an important step given that
mothers who were pregnant at the time of in- this model has become increasingly popular
carceration or sentencing (Cassidy et al., 2010). and in need of evaluation (McMahon, Huber,
There were 20 parents who completed this & Schneider, 2016; Mercer, 2015). In a recent
program based on the COS-I protocol, in that study, the COS-P intervention was evaluated in
it used Phase 1 and Phase 2 reviews and moth- a randomized controlled trial by Cassidy and
ers participating in small groups. Prerecorded colleagues (2017). This trial was conducted with
clips were also used (therefore giving a third- a Head Start population of parents of children
person review dimension) to introduce the ideas ages 3–5 years. Data were collected at baseline
of attachment theory and relationship patterns. and upon completion of the intervention on a
Of the 20 mothers participating, blind coding total of 141 dyads (from an original recruit-
at postintervention, when infants were approxi- ment of 164 dyads), with parents randomized
mately 12 months old, found 14 dyads to be se- into treatment in COS-P groups or a waiting
cure and six to be insecure/disorganized. list. The researchers hypothesized that COS-P
The COS—Home Visiting 4 (COS-HV4) would have a positive impact on parental sensi-
intervention was a randomized controlled trial tivity, measured by the Coping with Toddlers’
of a four-session version of the COS program Negative Emotions Scale (CTNES), and child
delivered in the home to mothers classified as behavior, measured by the Child Behavior
having economic stress and irritable infants. In Checklist (CBCL). They also assessed wheth-
this study, Cassidy, Woodhouse, Sherman, Stu- er intervention parents differed from controls
pica, and Lejuez (2011) found that intervention at postintervention on attachment, measured
effects were moderated by both infant irritabil- using the SSP, and on executive functioning.
ity and maternal attachment. First, examining The research also considered moderation of
only infant irritability as a potential moderator, outcomes by maternal depression and attach-
results showed the COS intervention was most ment style, as well as child gender.
effective at improving security for the dyads Following an intent-to-treat analysis, inter-
with the most irritable infants. For moderately vention and control groups did not differ on
irritable infants, there were no differences in measures of attachment category or dimensions
security between intervention and control. (security, avoidance, disorganization) at post-
Second, an analysis examining both infant ir- test. However, other results of the study provide
ritability and self-reported maternal attachment encouraging preliminary support. Unsupport-
ive responses to children’s distress were lower based intervention for families in which rela-
in the COS-P group than in the waiting-list tional issues are a primary presenting concern.
control group. Citing data that it is caregivers’ Over time, new variants of the program have
responses to distress that particularly contribute arisen to address various needs in the field.
to child attachment and to developmental out- COS-I remains the form of the model with the
comes, the authors concluded that if COS-P was strongest evidence base, though the recent de-
able to reduce unsupportive responses, it could velopment of COS-P raises hopes that a less in-
be valuable as a means for reducing risk of intensive protocol may yield benefits for families
secure and disorganized attachment, and asso- in a more readily disseminated form. At this
ciated psychopathology (Cassidy et al., 2017). point, we must acknowledge that the evidence
They note that COS is unique as an intervention base is limited to small studies, with few having
model in its explicit focus on enhancing paren- randomized designs or control groups, and, as
tal sensitivity to moments of distress rather than in the previous edition of this handbook, we are
overall parent sensitivity, and that other inter- motivated to develop and conduct further re-
ventions have not provided specific evidence of search into the model using well-designed clini-
impacting unsupportive responses to children. cal trials. Of particular interest is whether the
Other research has suggested that it is unsup- COS-P and COS-I represent a possible tiered
portive responses (and not supportive respons- program of intervention, given COS-I and its
es) by the parent that are most associated with variants’ success in positively impacting high-
child-related outcomes (e.g., Eisenberg et al., risk families. It may be that COS-P represents
2010). Inhibitory control was also shown to be a less intensive approach for families at lower
greater for children in the intervention group, risk or when setting constraints do not allow
once researchers controlled for marital status for first-person video review or more intensive
and maternal age. The relationship of inhibi- treatment. COS-P may also be an effective way
tory control to developmental capacities, such to engage parents who need more intensive
as school readiness, are supportive of COS-P’s therapy.
potential as an early intervention strategy. We see the strengths of the model, however,
There are two further studies in the literature continuing to be its clarity in expressing and
with small-scale evaluations of the COS-P pro- sharing the attachment relationship through the
gram. Horton and Murray (2015) conducted a COS graphic, and for COS-I, the support for
community treatment trial of 15 mothers in res- clinicians to develop impactful individualized
idential substance abuse treatment. They found treatment plans for dyads whose attachment
that mothers who attended sessions obtained struggles are a focus for treatment. Because
benefits with respect to self-reported emotion both parents and treatment providers who use
regulation, parenting attributions, and parent- the COS paradigm find within it an opportu-
ing behaviors. Kohlhoff, Stein, Ha, and Mejaha nity to reflect on their own attachment history,
(2016), in another sample of 15 mothers, this the model has a way of eliminating the distinct
time presenting with a child younger than age 2 demarcation between those providing and those
years to primary care parenting services, found receiving treatment. This increased empathy on
that COS-P reduced parent self-reported stress the part of professionals for their own attach-
on the Stress scale of the Depression Anxiety ment struggles translates directly into helping
Stress Scales and problematic parent attribu- parents build empathy for how it is that they
tions, measured on the Caregiving Helplessness also struggle within intimate relationships. The
Questionnaire and Parental Bonding Instru- human condition, which encompasses a strug-
ment from pre- to postintervention. Both these gle with painful and often unresolved memo-
studies are limited by the small sample sizes ries regarding dysregulated affect within close
and absence of a control group, yet they suggest relationships, unites us with similar challenges.
that COS-P may offer similar benefits to those From this perspective, it is not necessary to
found for the COS-I intervention. pathologize parents who are seeking help for
relationships with their young children. Indeed,
as we evolve as a species, what becomes criti-
Conclusions cal for creating a world in which empathy is a
central part of our experience is the increased
The COS approach has continued to attract sig- clarity about specific attachment needs and how
nificant interest in the field as an attachment- to respond to them better.
REFERENCES Hoffman, K., Marvin, R., Cooper, G., & Powell, B.

(2006). Changing toddlers’ and preschoolers’ attach-
Ainsworth, M. S., Blehar, M. C., Waters, E., & Wall, ment classifications: The Circle of Security Inter-
S. (1978). Patterns of attachment: A psychological vention. Journal of Consulting and Clinical Psychol-
study of the strange situation. Hillsdale, NJ: Erl- ogy, 74(6), 1017–1026.
baum. Horton, E., & Murray, C. (2015). A quantitative explor-

Bowlby, J. (1982). Attachment and loss: Vol. 1. Attach- atory evaluation of the Circle of Security–Parenting
ment. New York: Basic Books. (Original work pub- program with mothers in residential substance-abuse
lished 1969) treatment. Infant Mental Health Journal, 36, 320–336.
Bretherton, I., & Munholland, K. A. (2016). Internal Huber, A., McMahon, C. A., & Sweller, N. (2015a). Ef-
working model construct in light of contemporary ficacy of the 20-week Circle of Security intervention:
neuroimaging research. In J. Cassidy & P. R. Shaver Changes in caregiver reflective functioning, represen-
(Eds.), Handbook of attachment: Theory, research, tations, and child attachment in an Australian clinical
and clinical applications (3rd ed., pp. 63–88). New sample. Infant Mental Health Journal, 36, 556–574.
York: Guilford Press. Huber, A., McMahon, C. A., & Sweller, N. (2015b). Im-
Cassidy, J. (1994). Emotion regulation: Influences of at- proving child behavioral and emotional functioning
tachment relationships. Monographs of the Society after Circle of Security 20-week intervention. At-
for Research in Child Development, 59, 228–249. tachment and Human Development, 17, 547–569.
Cassidy, J., Brett, B. E., Gross, J. T., Stern, J. A., Martin, Kernberg, O. F. (1975). Borderline conditions and path-
D. R., & Mohr, J. J. (2017). Circle of Security–Par- ological narcissism. New York: Jason Aronson.
enting: A randomized controlled trail in Head Start. Kobak, R., Cassidy, J., Lyons-Ruth, K., & Ziv, Y. (2006).
Development and Psychopathology, 29, 651–673. Attachment and developmental psychopathology. In
Cassidy, J., & Marvin, R. S. (1992). Attachment organi- D. Cicchetti (Ed.), Developmental psychopathology
zation in preschool children: Procedures and coding (2nd ed., pp. 333–369). New York: Wiley.
manual. Unpublished manual, University of Virgin- Kohlhoff, J., Stein, M., Ha, M., & Mejaha, K. (2016).
ia, Charlottesville, VA. The Circle of Security Parenting (COSP) interven-
Cassidy, J., Woodhouse, S., Sherman, L., Stupica, B., tion: Pilot evaluation. Australian Journal of Child
& Lejuez, C. (2011). Enhancing infant attachment and Family Health Nursing, 13, 3–7.
security: An examination of treatment efficacy and Kohut, H. (1971). The analysis of the self. New York:
differential susceptibility. Development and Psycho- International Universities Press.
pathology, 23, 131–148. Lyons-Ruth, K. (2003). The two-person construction of
Cassidy, J., Ziv, Y., Stupica, B., Sherman, L. J., Butler, defences: Disorganized attachment strategies, unin-
H., Karfgin, A., et al. (2010). Enhancing maternal tegrated mental states, and hostile/helpless relational
sensitivity and attachment security in the infants of processes. Journal of Infant, Child and Adolescent
women in a jail diversion program. Attachment and Psychotherapy, 2, 107–119.
Human Development, 12, 333–353. Main, M., & Solomon, J. (1990). Procedures for identi-
Cooper, G., Hoffman, K., & Powell, B. (2009). Circle fying infants as disorganized/disoriented during the
of Security Parenting: A Relationship Based Parent- Ainsworth Strange Situation. In M. T. Greenberg, D.
ing Program: Facilitator DVD Manual 5.0. Spokane, Cicchetti, & E. M. Cummings (Eds.), Attachment in
WA: Circle of Security International. the preschool years: Theory, research, and interven-
Cooper, G., Hoffman, K., Powell, B., & Marvin, R. tion (pp. 121–160). Chicago: University of Chicago
(2005). The Circle of Security intervention: Dif- Press.
ferential diagnosis and differential treatment. In L. Masterson, J. (1976). The psychotherapy of the border-
J. Berlin, Y. Ziv, L. M. Amaya-Jackson, & M. T. line adult. New York: Brunner/Mazel.
Greenberg (Eds.), Enhancing early attachments: Masterson, J., & Klein, R. (Eds.). (1995). The disorders
Theory, research, intervention, and policy (pp. 127– of the self: New therapeutic horizons: The Masterson
151). New York: Guilford Press. approach. New York: Brunner/Mazel.
Eisenberg, N., Spinrad, T. L., Eggum, N. D., Silva, K. McMahon, C., Huber, A., & Schneider, S. (2016). Use of
M., Reiser, M., Hofer, C., et al. (2010). Relations Circle of Security approaches to early intervention
among maternal socialization, effortful control, and and prevention with families in Australia: Summary
maladjustment in early childhood. Development and of practitioner survey findings. Unpublished report,
Psychopathology, 22, 507–525. Centre for Emotional Health, Macquarie University,
Fonagy, P., Gergely, G., Jurist, E. L., & Target, M. Sydney, Australia.
(2002). Affect regulation, mentalization, and the de- Mercer, J. (2015). Examining Circle of SecurityTM: A
velopment of the self. New York: Other Press. review of research and theory. Research on Social
Hoffman, K., Cooper, G., & Powell, B. (2017). Raising Work Practice, 25, 382–392.
a secure child: How Circle of Security parenting can Powell, B., Cooper, G., Hoffman, K., & Marvin, B.
help you nurture your child’s attachment, emotional (2014). The Circle of Security: Enhancing attach-
resilience, and freedom to explore. New York: Guil- ment in early parent–child relationships. New York:
ford Press. Guilford Press.
Siegel, D. J. (2012). Developing mind: How relation- Winnicott, D. W. (1990). The maturational processes
ships and the brain interact to shape who we are and the facilitating environment. London: Karnac.
(2nd ed.). New York: Guilford Press. (Original work published 1965)
Sroufe, L. A. (2016). The place of attachment in devel- Woodhouse, S. S., Powell, B., Cooper, G., Hoffman, K.,
opment. In J. Cassidy & P. R. Shaver (Eds.), Hand- & Cassidy, J. (2018). The Circle of Security interven-
book of attachment: Theory, research, and clinical tion: Design, research, and implementation. In H.
applications (3rd ed., pp. 997–1011). New York: Steele & M. Steele (Eds.), Handbook of attachment-
Guilford Press. based interventions (pp. 50–78). New York: Guilford
Sroufe, L. A., Egeland, B., Carlson, E. A., & Collins, Press.
W. A. (2005). The development of the person: The
Minnesota Study of Risk and Adaptation from birth
to adulthood. New York: Guilford Press.

Chapter 30 The Circle of Security

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Chapter 30 The Circle of Security

Uploaded by

Copyright:

Available Formats

CHAPTER 30

The Circle of Security

I n the third edition of this handbook, the au-

a limited ability to act as a safe haven for the psychotherapy.

ing both participants comfortable within the

TABLE 30.1. Needs around the Circle

organized relationships is frightening (Mean), takes on the developmentally inconsistent task

patterns to be “core sensitivities” and describe an ambivalent/resistant response upon reunion

REFERENCES Hoffman, K., Marvin, R., Cooper, G., & Powell, B.

baum. Horton, E., & Murray, C. (2015). A quantitative explor-

You might also like