Attachment & Human Development Vol 4 No 1 April 2002 107-124
The Circle of Security project: Attachment-based intervention with caregiver-pre-school child
dyads
ROBERT MARVIN, GLEN COOPER, KENT HOFFMAN and BERT POWELL
Abstract The Circle of Security intervention protocol is a 20-week, group-based, parent education and psychotherapy intervention designed to shift patterns of attach-ment-caregiving interactions in high-risk caregiver-child dyads to a more appropriate developmental pathway. All phases of the protocol, including the pre- and post-intervention assessments, and the intervention itself, are based on attachment theory and procedures, current research on early relationships, and object relations theory. Using edited videotapes of their interactions with their children, caregivers are encouraged:
1 to increase their sensitivity and appropriate responsiveness to the child's signals relevant to its moving away from to explore, and its moving back for comfort and soothing;
2 to increase their ability to reflect on their own and the child's behavior, thoughts and feelings regarding their attachment-caregiving interactions; and
3 to reflect on experiences in their own histories that affect their current caregiving patterns.
In this paper we describe the conceptual background of the protocol, and the protocol itself. We then present a case study from our current data set of 75 dyads who have completed the protocol.
Keywords: attachment disorders - intervention - secure base - reflective functioning -pre-schoolers
A central feature of my concept of parenting [is] the provision by both parents of a secure base from which a child or an adolescent can make sorties into the outside world and to which he can return knowing for sure that he will be welcomed when he gets there, nourished physically and emotionally, comforted if distressed, reassured if frightened. In essence this role is one of being available, ready to respond when called upon to encourage and perhaps assist, but to intervene only when clearly necessary. (Bowlby, 1988, p. 11)
Correspondence to: see Acknowledgements
Attachment & Human Development ISSN 1461-6734 print/1469-2988 online © 2002 Taylor & Francis Ltd
http://www.tandf.co.ttk/jottrnals DOI: 10.1080/1461673021013163 5
attachment & human development
vol. 4 no. 1
INTRODUCTION
In the field of early intervention generally, and specifically for professionals working with maltreated, foster and adopted infants, pre-school and school-age children, there is a rapidly increasing demand by parents and professionals for interventions that are effective in shifting problematic or at-risk attachments toward more adaptive developmental pathways. Over the past few years the number of systematic intervention programs that are at least partially driven by attachment theory and research has been increasing (see Lieberman & Zeanah, 1999; van IJzendoorn, Juffer, & Duyvesteyn, 1995). In this article we describe one such protocol - the Circle of Security project (COS) - that was developed specifically for high-risk toddlers, pre-schoolers and their caregivers. Both the assessments involved, and the intervention itself, are explicitly based on contemporary attachment and congruent developmental theories (see, for example, Schore, 1994). The protocol is currently undergoing field-testing with Head Start/Early Head Start child-caregiver dyads (e.g. Marvin, Cooper, Hoffman, & Powell, 2000). The goal of the project is to develop a theory- and evidence-based intervention protocol that can be used in a partnership between professionals trained in scientifically based attachment procedures, and appropriately trained community-based practitioners.
The intervention involves small (N = 6) groups of at-risk mothers or fathers (primary caregivers) of toddlers or pre-schoolers (1 to 4 years of age), who meet as a group with a psychotherapist for 20 weeks, an hour and a quarter each week. In the context of the group, each parent reviews edited video-vignettes of herself or himself interacting with her or his child. These video feedback vignettes, and the related psycho-educational and therapeutic discussions, are individualized to each dyad's specific attachment-caregiving pattern using a priori, individualized treatment goals based on videotaped interactions recorded during a pre-intervention assessment. An identical assessment is conducted immediately after the 20-week intervention in order to track changes in patterns of child-caregiver interaction. At this point, approximately 75 dyads have participated in the protocol, and the coding and analysis of the data are ongoing. In this paper we first present the protocol's conceptual links to attachment theory, and then outline the protocol itself. We conclude with a case study from our data set.
BACKGROUND TO THE PROTOCOL
The COS protocol was specifically designed: to be theory- and research-driven; to be individualized for each dyad; to utilize standardized, evidence-based assessment procedures; to yield specific intervention goals based on those assessment procedures; and to focus the interventions on the caregiver as the partner more likely to induce change.
Theory- and research-driven
The protocol directly includes many of the ideas reflected in current theory and research in child development generally, and attachment and early parent-child interaction in particular. Especially important among these are the ideas of emotion regulation; interactive synchrony; states of mind regarding attachments and intimate relationships; shared states of consciousness, affect, and perspectives; and reflective functioning.
Marvin et al.
the circle of security
The core constructs for the assessment and intervention components of the protocol are Ainsworth's ideas of a Secure Base and a Haven of Safety (e.g. Ainsworth, Blehar, Waters, & Wall, 1978). Our goal was to present these ideas to the parents in a 'user-friendly', common-sense fashion that would be cognitively and emotionally accessible to them, and that would also guide the intervention. To do so we developed a graphic that represents both sides of Ainsworth's construct (see Figure 1). The upper half of the Circle represents the child's exploratory system and needs, and his tendency to move off to exploration if he expects his attachment figure to be available when needed. To the far right side of the Circle is represented the child's need to have his attachment figure monitor (watch over) his play in case he needs protection, help him if he needs affective or behavioral structuring from her, and enjoy his activities with him. The bottom half of the Circle represents the child's attachment system, his need for his attachment figure easily to welcome him 'in' for protection, comfort, delight, and to organize his feelings and behavior when they go beyond his own limits of self-organization. Consistent with Bowlby's definition, the parent's formula for a secure attachment is, 'Always be bigger, stronger, wiser, and kind. . .. Whenever possible, follow my child's need. . .. Whenever necessary, take charge.' The parents are introduced to attachment theory through this graphic, have it displayed continuously in the therapy room, and are given a copy in the form of a refrigerator magnet to keep displayed at home.
A key component of the protocol that the parent comes to understand and focus on is the idea that smooth interactions between children and their caregivers are often disrupted and need 'repair' (Bowlby, 1982; Tronick, 1989). It is this ability to repair a disruption that is the essence of a secure attachment, not the lack of disruptions. This repair requires clear cues from each other, and clear understanding of, and responsiveness to, each other's signals.
Each caregiver also comes to understand that certain needs on the Circle of Security can activate a painful feeling state that leads to an internal warning or sense of danger (see also Masterson & Klein, 1995). The parent then moves into a defensive strategy to protect herself by miscuing her child about the need. For example, a parent may react to her child's attachment behavior by encouraging him to move off and play. Over time, the parent's defensive strategy of miscuing triggers a sense of danger in the child regarding the need and so the child adapts by miscuing the parent about the painful need. A self-perpetuating feed-back loop is thus established in which both the child and the parent avoid the need by miscuing each other. Understanding this process leads parents to the realization that all dyads are typically more comfortable on one side of the Circle of Security than the other. In some dyads the caregiver is more comfortable encouraging the child's freedom to explore, but less comfortable dealing directly with the child's distress. In other dyads the caregiver find that giving emotional support comes easily, but allowing the child to move off to explore makes the caregiver uncomfortable and activates a danger signal. For example, a toddler or pre-schooler may know that directly cuing his mother that he needs soothing will make her uncomfortable, so he miscues her by acting like he really wants to play rather than climb on her lap. Even though the child's neutral affect, gaze avoidance and over-focus on play are available to the caregiver as cues that something is amiss, she accommodates to his lack of intimacy because of her own internal danger signal, and joins him in his miscuing. This is the basis of an Avoidant/Dismissing dyadic pattern. As implied in Figure 2, the Ambivalent/Preoccupied dyad has the reciprocal miscuing problem on the top side of the Circle.
CIRCLE OF SECURITY
PARENT ATTENDING TO THE CHILD'S NEEDS
.14}
.00
	
• Protect me	
• Comfort me	
• Delight in me	
• Organize my f	eelingS
Support My Exploration
A_
wyT Watch over me ■ • Help me -J • Enjoy with me
Welcome My Coming To You
Always: be BIGGER, STRONGER, WISER, and KIND. Whenever possible: follow my child's need. Whenever necessary: take charge.
Figure 1 Circle of Security:
Secure Base and Haven of Safety (© Cooper, Hoffman, Marvin, Sc Powell, 2000)
Limited Circles of Security
Protect me Comfort me Delight in me Organize my feelings
Watch over me ~"m • Help me 1 • Enjoy with me
> <
Figure 2 Limited (Insecure) Circles of Security (© Cooper, Hoffman, Marvin, & Powell, 1999)
112     attachment & human development
vol. 4 no. 1
Over the course of intervention, the caregiver learns that all parents (and dyads) experience especially strong danger signals when the child's need requires the caregiver to step out of her defensive strategy in order to meet that need. She also learns that this defensive, insecure strategy is linked to the patterns she developed early on in response to her own caregiver(s), and begins to consider the possibility that she may be passing on an insecure pattern by misinterpreting some of her child's signals. The protocol enhances this realization with a specific procedure: the group is shown a video clip of a coastline and rainforest, set to soft, serene music. The group members discuss the wonderful feelings this activates. The clip is then shown a second time, set to a modified version of the soundtrack from Jaws. Discussing how this time they experience agitation and anxiety, the parents come to understand how much their own subjective experiences can affect their feelings about their child's needs. It is as if the parents had learned from their own parents that certain needs are like shark-infested waters and must be avoided. They, in turn, 'protect' their children from what they currently perceive as dangerous needs.
Through watching films, seeing video clips of their children and themselves, reviewing responses from the pre-intervention interviews, and sharing the same with the other caregivers in the group, the parents begin to talk about this as their own 'Shark Music'. The term Shark Music becomes a frequently used icon in the protocol, as the caregivers increase their capacity to observe and reflect on the child's signals, and to reflect on, and 'stay with', their own painful feelings while meeting the child's need.
Individualized for each dyad
The intervention is individualized for each of the major patterns, or strategies, of attachment-caregiving interactions and internal working models (IWMs). This is based on differential identification of each child's attachment pattern and his or her parent's caregiving pattern, followed by a specific treatment (sub-)protocol assigned to that dyadic pattern. This design optimizes the 'goodness of fit' between parent and child, and helps to eliminate the potential problems of a 'one size fits all' approach to intervention.
Individual differences in patterns of attachment-caregiving interactions tend to be shared by parent and child in the form of a reciprocal 'dance'. Each pattern can be identified reliably in terms of specific organizations of behaviors and of IWMs of both partners, where in most cases there is a clear synchrony between the partners (see Ainsworth et ah, 1978; Cassidy & Marvin, 1992; Main, Kaplan, & Cassidy, 1985; George & Solomon, 1999; Marvin & Britner, 1995). Four or five distinct dyadic patterns (each with sub-patterns) have been identified:
1 The Secure (child)-Autonomous (parent) pattern is one in which both partners easily approach and interact with one another when the child is distressed, and do so in a manner that predictably terminates the child's attachment behavior system and facilitates further exploration. The shifts between the child's use of the caregiver as a secure base for exploration, and as a haven of safety, are smooth and relatively free of anxiety. This pattern therefore functions smoothly and easily around the entire Circle of Security, and the partners do not miscue each other in a consistent manner. The close attunement between child and caregiver is, of course, occasionally disrupted, but these disruptions are easily repaired, or re-equilibrated (see Bowlby,
Marvin et al.
the circle of security
1982; Schore, 1994; Tronick, 1989). These cycles of adaptive attunement-disruption-repair between the partners, applied to affect regulation, physical proximity and shared experience, feelings, goals and plans, lead to self-reliance and developmental competence on the child's part.
2 The Insecure, Avoidant-Dismissing pattern is one in which both partners tend to minimize the more intimate attachment-caregiving interactions in favor of other types of interactions, often 'distracting' themselves from attachment-caregiving interactions through a defensive focus on exploration. They therefore feel least comfortable on the bottom half of the Circle of Security, when the child's attachment behavior or needs are activated. It is at this point on the Circle that each miscues the other that the child is 'really' more interested in playing (see Figure 2). Early caregiver rejection of infant attachment behavior and related emotional signals results in a later complex, shared pattern of anxious avoidance/rejection of intimate shared experiences, feelings, goals and plans, over-regulation of affect, and comparatively little emotional self-knowledge. The partners work carefully to stay connected enough to protect the child, but not so close as to behave in an intimate manner. They therefore have their own distinct pattern of attunement-disruption-repair - a pattern that, while organized or ordered, is more complex and anxious than the secure pattern.
3 The Insecure, Ambivalent-Preoccupied pattern is one in which both partners tend to minimize the child's independent exploration, focusing instead on (often ambivalent) attachment-caregiving interactions and the child's over-dependence on the parent. They therefore feel less comfortable on the top half of the Circle of Security, when the child's autonomous and exploratory behavior or needs are activated. At this point on the Circle they tend to miscue each other that there really is something about which they should be worried and/or distressed (see Figure 2). Early caregiver inconsistency and lack of responsiveness result in a later complex, shared pattern of anxious ambivalence regarding intimacy vs. distance, problematic boundaries (enmeshment) between perspectives of child and caregiver, under-regulation of affect, and a preoccupation with the relationship to the relative exclusion of autonomous/competent activity. The distinct pattern of attunement-disruption-repair in this overly intimate dyad is also ordered, but is more complex and anxious than the secure pattern.
4 Finally, there are a number of even higher-risk, Insecure, 'Disordered'1 (Disorganized or Insecure-Other) patterns in which the parent's heightened fear of and/or anger toward the child's attachment behavior leads to disorganization and/or abdication of the executive, caregiving role (e.g. Main & Hesse, 1990; Solomon & George, 1996; Marvin & Britner, 1995). This caregiver pattern appears to be related to lack of resolution regarding early and/or ongoing trauma in the caregiver. During infancy this child's attachment behavior tends to be disorganized and lacking in coherence, suggesting the lack of a stable disruption-repair cycle in these dyads. By the middle of the pre-school years, continued fear and caregiver abdication, coupled with advances in the child's social-cognitive skills, lead some of these dyads into 'role-reversed' relationships in which the child rather than the caregiver assumes the executive role in organizing highly emotional, attachment-caregiving interactions (Main, Kaplan, & Cassidy, 1985; Cassidy & Marvin, 1992). Others of these disorganized dyads develop patterns in which both partners exhibit combinations of avoidant and ambivalent patterns, patterns of intimidation and compulsive compliance, or reciprocal disengagement so strong that the protective purpose of the
114     attachment & human development
vol. 4 no. 1
relationship cannot be fulfilled. In most of these disordered patterns, the partners have difficulties with individual and joint affect regulation, and have anxious, complex patterns of attunement-disruption-repair that tend not to have appropriate caregiver protection of the child as predictable outcomes. They therefore have difficulty at all points around the Circle of Security, and may in fact not 'be on it' at all. In these cases, the intervention focus is on building the capacity both to follow the child's activity, and to lead the child's activity, as appropriate.
Utilize standardized pre- and post-intervention assessment procedures
Assessment procedures Each child-caregiver dyad participates in the same assessment before and within 10 days of the completion of the 20-week intervention. The assessment is used to identify risk and resilience factors in their interactions, design individualized intervention goals, and measure change across the 20 weeks.
Each dyad participates in the infant or pre-school version of Ainsworth's Strange Situation (Ainsworth et ah, 1978; Cassidy & Marvin, 1992). Immediately following the Strange Situation, each caregiver is given some books appropriate for the child's age group, and is asked to read to the child for five minutes. The caregiver is then signaled to encourage the child to clean up the toys in the Strange Situation room. Following these observational procedures, the caregiver participates in a videotaped, one-hour interview (the Circle of Security Interview, or COSI) consisting of questions about the child's and her or his own behavior and feelings during the Strange Situation, selected questions from the Parent Development Interview (e.g. Aber, Slade, Cohen, & Meyer, 1989), and selected questions from the Adult Attachment Interview (George, Kaplan, & Main, 1984). Finally, the caregiver fills out a number of standardized questionnaires that measure child behavior problems, anxiety and depression, parenting stress, and stressful life events.
These procedures provide a number of options for pre- and post-intervention comparisons and criteria for successful change. While a shift in the child or caregiver's pattern from insecure (Ordered or Disordered) to secure is one obvious option, for this project we are using a shift from Disordered to Ordered (secure or insecure) as our criterion of success. This is because the negative outcomes associated with Disordered attachment are more severe than those associated with Ordered-yet-insecure attachments, combined with the high rate (preliminary results suggest above 50%) of pre-intervention Disordered attachment-caregiving classifications in our sample. We are also comparing pre- and post-intervention caregiver rating scales, and child and caregiver questionnaire data.
Child and caregiver classifications and ratings The videotapes and questionnaires from each assessment are coded by clinical psychologists or clinical graduate students appropriately trained and certified in the attachment procedures. The child's attachment pattern is coded using either Ainsworth's classification system for infants (Ainsworth et ah, 1978), or the Preschool Attachment Classification System (Cassidy & Marvin, 1992). Adequate coder agreement is assured by double-coding approximately half of the Strange Situations.
The parent's caregiving classification is coded from the same Strange Situation, again in approximately half the cases double-coded by independent coders blind to the children's classifications. The parent's classification is carried out using the Caregiver Behavior Classification System (Marvin & Britner, 1995). This recently developed
Marvin et al.-. the circle of security 115
system is based on specific patterns of observable caregiver behaviors during play, separation and reunion in the Strange Situation. The system yields five major classifications, each one the reciprocal of one of the child classifications, and is conceptually continuous with both the Ainsworth infant system and the adult classifications from the Adult Attachment Interview: Secure/Autonomous; Avoidant/Dismissing; Preoccupied/Ambivalent; Disorganized/Abdicating; and Insecure-Other/Unclassifi-able (Ds-E). It also yields continuous ratings on 12 rating scales reflecting constructs that are central to attachment theory and that were adapted for use in rating caregiver behavior in the Strange Situation. The Caregiver Behavior Classification System has been found to be highly reliable, with a highly significant 'hit rate' between independently classified child and caregiver classifications (Britner, 1996).
Constructing intervention goals Intervention goals are derived from the classifications, ratings and clinical observations. These goals are specific to each major classification; for example, Dismissing caregivers usually are assigned the following treatment goals: increased appreciation of how much their children need them; increased skill at reading and registering their children's subtle distress signals; and decreased miscuing under circumstances in which a child's attachment behavior is activated. It is also typical that each caregiver has her or his own individualized goals within this set of shared goals; for example, a specific Dismissing mother may also have the treatment goal of decreasing her tendency to pressure her child to achieve more competent play, and of increasing her tendency to have her child's enjoyment of play be the more important goal.
In developing these individualized intervention goals, the person scoring the data will usually listen to the Circle of Security Interview to obtain information about the parent's IWMs regarding her or his relationship with the child. This yields useful information about events that trigger the caregiver's non-secure behavior patterns, events on her or his own developmental pathway that might be related to those patterns, and the words used by the caregiver herself or himself to describe and understand these 'stuck points', or 'Shark Music' (see below).
Finally, intervention goals are also derived, and fine-tuned throughout the intervention, based on differential diagnosis of character structure in the context of contemporary Object Relations theory (e.g. Guntrip, 1969; Kernberg, 1984; Kohut, 1977; Masterson & Klein, 1995). Particular attention is paid to the following: flexibility/rigidity of the caregiver's defensive process; capacity to manage or regulate her or his own emotions in the group; her or his capacity for reflective functioning; and her or his willingness and capacity to develop a therapeutic working relationship with the group leader.
Direct focus on the caregiver
In designing the protocol, we have decided to capitalize on the fact that the caregiver, as an adult, has more 'degrees of freedom' in changing patterns of attachment-caregiving interactions than does the child. This focus specifically does not imply that the problematic pattern is 'caused' by the caregiver. Rather, the implication is that even for a pre-schooler or an older child, a most effective intervention for problematic attachment-caregiving patterns may be to focus directly on the caregiver, and work toward shifting the caregiver's patterns of behavior and/or her IWMs of attachment-caregiving interactions with this particular child. This shift should then lead to a
attachment & human development
vol. 4 no. 1
change in patterns of parent-child interaction, and in turn shift the child's patterns of attachment- and exploratory-behavior toward a more adaptive developmental pathway.2
In the context of a supportive group environment in which the therapist has taken on the role of a secure base for the participants, each caregiver is guided at her or his own pace toward increased skill in reading the child's cues, reflecting on the child's (inferred) thoughts and feelings, and reflecting on her or his own feelings, plans and behavior. The basic steps in working with the caregiver's Shark Music are: to increase the caregiver's awareness of her or his discomfort/Shark Music; to honor the discomfort and know it has to do with danger in her or his history more than in the present; and to bring the caregiver to respond to the child's need despite her or his own discomfort.
THE CIRCLE OF SECURITY (COS) PROTOCOL
The 20-week intervention takes place in a small group of five or six caregivers, and one or two therapists. The group sessions are videotaped. The therapist edits key segments of the assessment videotapes for use in the group sessions.
The goals
There is a sequence of five overarching therapeutic goals used for all parents:
1 to create a holding environment or secure base from which the parents can explore their parenting;
2 to provide the parents with a user-friendly map of attachment theory that we call the Circle of Security;
3 to help the parents develop their observational skills, especially as these apply to reading and responding to their children's (often subtle and misleading) cues;
4 to develop a process of reflective dialogue in the group - a skill that the parent can then use internally; this process is viewed as the central dynamic for change;
5 to supporting the parents' empathic shift from defensive process to empathy for their children.
The intervention protocol
The course of the intervention protocol is as follows:
Weeks 1-2: introduction and theory-building The introductory two weeks of each group are used to build comfort, trust and the beginnings of a working alliance between parents and group leaders, as well as to introduce the central concepts of attachment theory to the parents in a user-friendly manner.
In week 1, the focus is upon celebrating the competence and positive intentionality of the parents. This is accomplished by showing an edited video montage, set to music, of the parents interacting with their children. The song 'You Are So Beautiful' is superimposed on the audio track, and the group leader tells the parents, 'This is the song your children are singing to you.' A 'formula' for secure attachment is introduced and the benefits of security are discussed. The key concepts of attachment theory are
Marvin et al.-. the circle of security 117
introduced by showing the parents brief video clips of their children moving from exploration to proximity seeking and back to exploration. The parents receive a copy of 'The Circle of Security' (Figure 1) and are asked to share their observations as the group begins to figure out what is happening within each video vignette. This begins to build the parents' observational skills and engages them in the process of discovery.
Enhancing observational skills and understanding children's needs become the focus of week 2. Video vignettes of the parents and children are shown to practice observational skills. As their observational skills increase they are asked to discuss, based upon the Circle of Security paradigm, what each child's primary need is within each video example, i.e. exploration need or attachment need. Based upon what they observe regarding the emotional needs of each child, the parents begin to hypothesize what their children are feeling. This introduces reflective attunement and reflective dialogue, the project's proposed central mechanism for change.
Weeks 3-8: phase 1 tape reviews During each of weeks 3-8, the group watches video vignettes from one parent-child dyad, chosen to occupy the 'hot seat' that week. The other members of the group have a worksheet to fill out as they watch the tape, and are invited to participate in evaluating the interaction, each according to her or his capacity. The group leader carefully edits the tape of the pre-intervention assessment to illustrate the central issue of each parent's caregiving strategy. The underlying assumption: every parent is more comfortable with some areas of parenting and less comfortable with others. Attachment theory emphasizes the importance of a parent's ability to meet both a child's exploration needs and a child's attachment needs. Typically, parents tend to over-use one of these abilities (the one they are most comfortable with) and to under-use the other. These are described as 'over-used strengths' and 'under-used capacities'.
The leader chooses four vignettes of each dyad to present during that session:
1 Vignette 1 shows the child distressed and wanting the parent. This clip is used to clarify how important the parent is to the child and to activate and enhance the parent's caregiving system. This clip can also help to clarify and/or redefine the parent's internal representation of the child.
2 Vignette 2 highlights the parent being competent in her or his under-used capacity.
3 Vignette 3 shows the parent struggling with her or his under-used capacity. This creates a segue to the phase 2 review, which will take place seven weeks later.
4 Vignette 4 presents a moment to celebrate the parent's relationship with the child. The parent needs to know that the group and group leader are supportive, thus consolidating a secure base from which the parent can continue to explore her or his own patterns of caregiving.
During each tape review the leader gives that particular parent a still (printed from the video) as a 'take home' picture illustrating the central lesson of the review. The parent is encouraged to place the 'Picture of the Week' in a prominent place in her or his home as a reminder of the capacity being recognized and supported.
Week 9: transition and theory-building continued Week 9 is a transitional, theory-building segue to the next series of video reviews. The central theme of week 9 is: 'All parents struggle.' The leader further normalizes the struggles of parenting by reading 'Welcome to the Club', a short essay designed to help parents acknowledge their 'everyday' personal difficulties within the caregiving process, and the points at which
attachment & human development
vol. 4 no. 1
they become stuck, i.e. their Shark Music. Educational videotapes and discussion clarify the role of the defensive process within personal interactions. The leader clarifies further by utilizing attachment theory to explain, in a non-pathologizing way, the dynamics behind Avoidant, Ambivalent, and Disorganized-Role-Reversed attachment strategies (see Figure 2). Parents are then invited to share insights into their defensive process and to begin exploring how these defenses may impact their particular caregiving strategy.
Weeks 10-15: phase 2 tape reviews From week 10 through to week 15 the group participates in 'phase 2' tape reviews. In phase 2, the leader re-edits the Strange Situation tapes to focus more directly on 'under-used capacities/points of struggle' and seeks to engage the parent in reflective dialogue regarding the vulnerability she or he experiences when activating these capacities. In addition, the leader focuses upon themes of emotional regulation for the parent and the child. Again, four vignettes are typically used:
1 Vignette 1 is use to engage the parent's caregiving system. When a parent's caregiving system is engaged, she or he seems to be better able to contain defensiveness.
2 Vignette 2 looks directly at the parent's under-used capacity/point of struggle; for example, a 'Dismissing' parent struggles when the child needs contact and soothing. The vignette is used in a respectful manner and the use is based on the leader's sense of that particular parent's ability to utilize painful information. Usually, the focus is upon how the child 'miscues' the parent (a miscue is described as a misleading cue used to protect the child or caregiver from the pain of having a specific need exposed and/or unmet). The parent can then stay focused upon the child's process or move toward focusing on her or his own vulnerabilities. The leader follows the parent's lead.
3 Vignette 3 shows the parent being successful at an under-used capacity. This allows the leader to punctuate the fact that this parent has the inherent ability to do well within this area of struggle. In addition, the parent is invited to explore what gets in the way of further utilizing this particular capacity.
4 Vignette 4 is again a celebration of this parent and her or his relationship with her or his child.
Each week, at the conclusion of the group, the parent being focused upon is again given a 'Picture of the Week', highlighting a moment of success within her or his under-used capacity.
Week 16: transition and review Week 16 is a review of the information covered and preparation for final tape reviews. This allows parents to consolidate their learning and deal with any unfinished group process issues that may have emerged.
Week 17-19: phase 3 tape reviews Weeks 17, 18 and 19 provide an opportunity to celebrate the changes that have been made. In these 'phase 3 tape reviews', two parent-child dyad tapes are reviewed each week. The parents have recently been videotaped in a 'modified Strange Situation' setting. Again using four edited video vignettes, this interaction is designed to emphasize the parent's increased success within the context of her or his under-used capacities.
Week 20: final session Week 20 brings graduation and more celebration. Parents share with one another their experience in the group. The leader shares her or his
Marvin et al.-. the circle of security 119
appreciation for the parents' willingness to face caregiving vulnerabilities within a group context. In conclusion, the leader shows a video montage, set to music, which sums up the salient moments of each parent-child dyad. Each parent is then given her or his completed workbook, 'The Kissing Hand' (another take-home children's story, with the theme of emotion regulation and internalized support), and a 'Certificate of Graduation'. Final arrangements for the post-intervention assessments are made and the parents are given paper-and-pencil tests to fill out at home before they come to the assessment. They are also given a copy of the video montage tapes and their individual tapes from the three reviews.
CASE STUDY
Paula is 28 months old. She enters the laboratory playroom carried by her mother, and upon being put down immediately focuses on the box of toys near the couch. She pulls out a bunny and asks her mother if she wants to see it. In a voice barely audible to those watching through the one-way mirror, her mother says yes. Over the next several minutes Paula increasingly takes charge of the interaction between herself and her mother. As her mother leaves Paula with the stranger in the first separation, Paula seems indifferent. Then, 20 seconds later, she gets up and walks to the exact spot where her mother has been sitting and says, 'That's Mom.' Eighteen minutes later, after Paula had continued organizing interactions through the two separations and reunions, her mother asks her to help clean up the toys. Paula refuses. Then, picking up a red plastic phone, she rejects her mother's pleas to put it away.
Formal pre-intervention assessment results
The child was classified as Disordered-Controlling-Caregiving. The mother was classified as Disordered-Role-Reversed-Abdicating, with high ratings on the scales reflecting role-reversal and flat affect, and low ratings on affection, sensitivity, delight and support for her child's exploration.
Intervention goals
Capacity The mother is much more affectively alive, as well as effective, when reading to Paula or managing her behavior .. . much less so if Paula is playing on her own, or around separations/reunions and subsequent activation of their attachment and caregiving systems.
Primary goals (1) Tracking and intervening with Mom's tendency just to sit back and let Pamela run the show ... to help Mom see that this is more than just cute 2K-year-old behavior. .. and to help her take a stronger executive role without becoming intrusive; (2) connecting her tendency to parent in a role-reversed manner to her own, early, unresolved, role-reversed relationship with her own mother; and (3) helping Mom see that her flat affect deprives her child of information about Mom's feelings and probable behavior, and helping her begin the process of becoming more expressive.
Course of intervention It is now two months later and Candy, Paula's mother, is sitting in a room with four other parents looking at her first tape review of the video
attachment & human development
vol. 4 no. 1
assessment with her daughter. 'I'm nervous about seeing this. I'm not sure what I'll find out about us. But I also want to know what I can do to make things better with Paula,' she says, clarifying her commitment to find a way to become a more effective caregiver. Summarizing her primary concern about their relationship, Candy then says: T just never feel like I know what to do with her.' Recognizing Candy's depressive affect and her tendency to encourage Paula's role-reversed caregiving while simultaneously abdicating to her daughter's demands, the group leader had chosen video clips designed to help this mother recognize Paula's need for a safe haven. This child was clearly in need of the firm, yet loving, presence of a caregiver who was 'bigger, stronger, wiser and kind'. Candy was clearly afraid of her daughter's already considerable power in the relationship and was consistently asking her daughter for reassurance and safety. Hence the initial video clips would need to focus upon reversing this upside-down arrangement, allowing Candy to see her daughter as the one who was clearly in need of her mother. Only then could the group leader hope to point out Candy's obvious inability to establish hierarchy and begin to engage in reflective dialogue around the possible causes of their disorganized pattern of relationship.
Given that the focus of the first video review is to show parents the 'under-used' caregiving qualities that need reinforcement, this particular review began with Paula looking for guidance as she moved further out in the Circle. 'Notice how she seems to be enjoying those princess slippers, and then suddenly looks back to you to see if you are interested too. When she does you smile back and Paula seems to really relax,' said the leader. Candy shyly grins, as she settles back into her chair, obviously feeling relieved to see some indication that her daughter finds her important. The leader then turns to the group and asks, 'Which half of the Circle is Paula on when she is walking in those cute shoes?' Everyone, almost in a single voice, declares, 'Top half. Exploration.' 'Right. And what about here, when Paula looks back at you and smiles?' 'Enjoy with me,' exclaims Candy. 'Definitely,' declares the leader. 'But it's also something more. It is also the bottom half of the Circle, where Paula is looking to you for comfort as well as enjoyment. This kiddo really likes being with you, and isn't it nice how she goes out all the way to the edge of the Circle and then - just like clockwork - needs to return to you for reconnection.' Candy smiles nervously, but with an increased sense of confidence that she has a central role to play in her daughter's life.
The leader now shows another clip, this one of Paula getting oppositional to her mother about putting on her own shoes at the end of the assessment. He looks to Candy and asks what she sees going on. Increasingly trusting that she is safe, and aware that she is central to her daughter's emotional well-being, she volunteers the observation that her daughter 'has just taken charge'. T think you're right,' agrees the group leader. 'From the perspective of what Paula needs most from you right now, is this a "take charge" moment or a "follow" moment?' 'Take charge,' Candy offers, 'but Paula likes to be in control.' Paula's need for a secure, yet safe haven and their primary struggle regarding hierarchy are now on the table. The session closes with an additional video of Candy providing needed help when Paula seems frightened of a play mouse in the toy box. This last clip punctuates how Paula already relies upon her mother for support and guidance.
It is now seven weeks later, and Candy is once again ready to observe her relationship with Paula on video. The group leader had previously explained to each of the mothers in the group that it is common, and expected, to have 'Shark Music' emerge. Having normalized defensive process around a caregiver's capacity to meet specific needs (as he does each time before each parent's second video review) the leader now
Marvin et al.
the circle of security
turns to Candy and says, 'What's your hunch about where you think your Shark Music is?' Candy replies, 'When I get to being "in charge," that's where the Sharks are.' Then she adds a comment common with many abdicating parents, ' I don't want to be abusive.'
The videotape begins. First, there is a revisit to the vignette where Paula walks to the couch during Candy's absence and says, 'That's Mom.' This reinforces Candy's awareness that her daughter relies upon her, even though she often 'miscues' her mother by appearing independent and in charge. (A miscue is the defensive denial of a genuine need.) Then a video clip shows Paula pick up a red phone and refuse to relinquish it, regardless of how hard her mother tries to get her to put it away. 'It's like I was asking permission to do everything, and she was the one letting me know what we were going to do,' exclaimed Candy. The leader then mentions, 'It's almost like you make yourself her sister rather than her mom.' Suddenly Candy is crying. 'My mom always used to tell me that we were more sisters than we were mother and daughter. She still does. And she tries to make it sound like a good thing.'
The dialogue between Candy, the group leader and other group members then focuses upon the specific issues that comprised the 'shark-infested waters' in which she always seemed to swim. Candy had been born when her mother was 16. Afraid and alone, her mother had turned to her own daughter as a source of support and safety. Seeing her mother as 'inconsistent. . . sometimes a "mush ball" and sometimes she'd knock me down for being bossy', Candy had taken charge of the relationship. Twenty-five years later she had unknowingly been asking her daughter to do the same, seeking to re-enact a dysfunctional agreement, rather than seek a new alternative. Until now. With the aid of a group setting as a holding environment, the Circle of Security as a clear map describing her daughter's needs, and edited videotape with precisely delineated themes focused upon the core struggles in her relationship with her daughter, Candy was now able to observe, reflect on and discuss the linchpin issues that were keeping their dysfunction intact. She was also able to consider new options. Candy left this session willing to see her daughter's significant need for her as a safe haven (someone available to comfort, protect, and organize feelings) and a resource ('bigger, stronger, wiser, and kind') committed to maintaining hierarchy.
Post-intervention assessment Seven more weeks pass. During the early portion of the post-intervention Strange Situation, Candy continues to show signs of a mild lethargy and Paula continues to find ways to organize their play. At the same time Candy takes a more active role than she had in the pre-intervention Strange Situation, suggesting toys and participating more fully with the playthings that she and her daughter have chosen. During the second reunion, Paula makes it clear that she is upset and angry at her mother for having gone, and continues to show signs of distress for several minutes after Candy's return. Candy seeks to provide comfort, haltingly offering Kleenex, toys and physical comfort. But Paula's upset continues until her mother firmly places Paula on her lap, comforts her physically, and gets her involved with a toy. Only then does Paula calm, and come to rest in her mother's presence.
Three minutes later, following a time of reading when Paula has allowed herself to mold into her mother's body with noticeable comfort, there is a tap on the window signaling Candy that it is time for the toys to be put away. 'Time to clean up. Put everything back in the box,' directs Candy. Paula quickly jumps from her mother's lap and starts putting the toys away. 'Good job,' says her mother. Almost as if delighted to be in the presence of someone who is willing to take charge, Paula continues
122     attachment & human development
vol. 4 no. 1
cleaning up. The room now placed in order, Candy asks her daughter if she is ready to go. Paula nods. Together they walk toward the door. Paula tries to open it, but the door handle seems too difficult for her still small hands. Candy, standing over her, quietly reaches down and with an ease that clearly delights her daughter, turns the knob and opens the door.
Formal post-intervention assessment results The child was classified as Secure, with some continuing disordered elements. The mother was classified as Ordered, Borderline Autonomous-Dismissing. She exhibited decreases in role-reversal and flat affect, and increases in (brief) rejection, affection, sensitivity, and support for her child's exploration.
CONCLUSION
These results represent significant changes for both partners in this dyad. Both shifted from Disordered to Ordered patterns, and the child actually shifted to a Secure pattern. There were a number of changes in the caregiver rating scales. Some reflected significant improvements in Candy's behavior such as improved scaffolding of Paula's exploration and decreased abdication/role reversal. Others reflected degradations in her caregiving, such as increased (temporary) rejection of Paula's attachment behavior, and a mild increase in Candy's overinvolving behavior. Close observation of these degradations suggest, however, that they may represent transitional points in this mother's not-yet-completed task of shifting away from her role-reversed, abdicating caregiving pattern.
At this time we are completing the coding and data analysis of the 75 dyads that have completed the protocol. Preliminary results suggest a significant shift from Disordered to Ordered child attachment patterns (from 55% to 20%), an increase (from 32% to 40%) in the number of children classified as Secure, and a decrease in the number of caregivers classified as Disordered (from 60% to 15%). In addition, we are currently conducting a second study to determine if community-based therapists, with appropriate training and supervision, are able successfully to carry out the protocol with other parent-child dyads drawn from the same population. Given the relatively large investment of time in the assessments and intervention goal setting, it will be important to show that the protocol is, in fact, fiscally efficient. Given the group format, we are optimistic that the Circle of Security protocol will prove to be a successful model for a cost-effective, university-community therapeutic partnership.
ACKNOWLEDGEMENTS
This project is being supported by USDHHS University-Head Start Partnership Grants #90YF0018, and 90YD0108. The authors would like to emphasize that all four have contributed equally to the project. They would like to thank Jude A. Cassidy, and the Spokane research affiliates, for their support throughout the project. Requests for reprints should be addressed to the first author at Child-Parent Attachment Clinic, University of Virginia, 310 Old Ivy Way, Charlottesville, VA 22903, USA, or at circleofsecurity@attbi.com
Marvin et al.
the circle of security
NOTES
1 This use of the term 'Disordered' is not meant as synonymous with the DSM-IV term. For the purposes of this paper it denotes that the pattern is outside the range of the three primary, organized, attachment-caregiving patterns. (See also Zeanah & Boris, 1999.)
2 This is not meant to imply that working directly and jointly with child and caregiver would not be as effective.
REFERENCES
Aber, J., Slade, A., Cohen, L., & Meyer, J. (1989, April). Parental representations of their toddlers: Their relationship to parental history and sensitivity and toddler security. Paper presented at the biennial meeting of the Society for Research in Child Development, Baltimore, MD.
Ainsworth, M. D. S., Blehar, M. C, Waters, E., & Wall, S. (1978). Patterns of attachment: Psychological study of the Strange Situation. Hillsdale, NJ: Erlbaum.
Boris, N. W,. & Zeanah, C. H. (1999). Disturbances and disorders of attachment in infancy and early childhood: An overview. Infant Mental Health Journal, 20, 1-9.
Bowlby, J. (1982). Attachment and loss: Vol. 1. Attachment. New York: Basic Books.
Bowlby, J. (1988). A secure base: Clinical applications of attachment theory. London: Routledge.
Bowlby, J. (1988). A secure base: Parent-child attachment and healthy human development. New York: Basic Books.
Britner, P. A. (1996). Maternal caregiving behavior and child attachment classifications in the preschool Strange Situation. Unpublished PhD dissertation, University of Virginia.
Cassidy, J., & Marvin, R. S., with the MacArthur Working Group on Attachment (1992). A system for classifying individual differences in the attachment-behavior of 2'A to 4'A year old children. Unpublished coding manual, University of Virginia.
Cooper, G., Hoffman, K., Marvin, R., & Powell, B. (1999 & 2000). Secure and limited circles of security. Unpublished material, Centre for Clinical Intervention, Spokane, WA, and University of Virginia.
George, C, Kaplan, N., & Main, M. (1984). Adult Attachment Interview protocol. Unpublished
manuscript, University of California at Berkeley. George, C, & Solomon, J. (1999). Attachment and caregiving: The caregiving behavioral system.
Inj. Cassidy & P. Shaver (Eds), Handbook of attachment theory and research. New York:
Guilford.
Guntrip, H. (1969). Schizoid phenomena, object relations and the self. New York: International Universities Press.
Kernberg, O. F. (1984). Severe personality disorders: Psychotherapeutic strategies. New Haven, CT: Yale University Press.
Kohut, H. (1977). The restoration of the self. New York: International Universities Press.
Lieberman, A. F., & Zeanah, C. H. (1999). Contributions of attachment theory to infant-parent psychotherapy and other interventions with infants and young children. Inj. Cassidy & P. Shaver (Eds), Handbook of attachment theory and research. New York: Guilford.
Main, M., & Hesse, E. (1990). Parents' unresolved traumatic experiences are related to infant disorganized attachment status: Is frightened and/or frightening parental behavior the linking mechanism? In M. Greenberg, D. Cicchetti & E. M. Cummings (Eds), Attachment in the preschool years (pp. 166-182). Chicago: University of Chicago Press.
Main, M., Kaplan., N., & Cassidy, J. (1985). Security in infancy, childhood, and adulthood: A move to the level of representation. In I. Bretherton & E. Waters (Eds), Growing points of attachment theory and research. Monographs of the Society for Research in Child Development, 50 (1-2, Serial No. 209), 66-104.
Marvin, R., & Britner, P. (1995). Classification system for parental caregiving patterns in the preschool Strange Situation. Unpublished classification manual.
124     attachment & human development
vol. 4 no. 1
Marvin, R., Cooper, G., Hoffman, K., & Powell, B. (2000, J une). Attachment-based intervention with at-risk Head Start child-parent dyads. Presented to the National HeadStart Research Meetings, Washington, DC.
Masterson, J., & Klein, R. (Eds). (1995). The disorders of the self:New therapeutic horizons. New York: Brunner/Mazel.
Schore, A. (1994). Affect regulation and the origin of the self: The neurobiology of emotional development. New York: Erlbaum.
Solomon, J., & George, C. (1996). Defining the caregiving system: Toward a theory of care-giving. Infant Mental Health Journal, 17(3), 183-197.
Tronick, E. Z. (1989). Emotions and emotional communication in infants. American Psychologist, 44, 112-119.
Van IJzendoorn, M. H., Juffer, E, & Duyvesteyn, M. G. F. (1995). Breaking the intergenerational cycle of insecure attachment: A review of the effects of attachment-based interventions on maternal sensitivity and infant security. Journal of Child Psychology and Psychiatry, 36, 225-248.