This article was downloaded by: [Carol George]On: 25 August 2011, At: 09:45Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK

Journal of Personality AssessmentPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/hjpa20

The Adult Attachment Projective Picture System:Integrating Attachment Into Clinical AssessmentCarol George a & Malcolm West ba Department of Psychology, Mills Collegeb Department of Psychiatry (Retired), University of Calgary, Canada

Available online: 22 Aug 2011

To cite this article: Carol George & Malcolm West (2011): The Adult Attachment Projective Picture System: IntegratingAttachment Into Clinical Assessment, Journal of Personality Assessment, 93:5, 407-416

To link to this article: http://dx.doi.org/10.1080/00223891.2011.594133

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions

This article may be used for research, teaching and private study purposes. Any substantial or systematicreproduction, re-distribution, re-selling, loan, sub-licensing, systematic supply or distribution in any form toanyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contentswill be complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses shouldbe independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims,proceedings, demand or costs or damages whatsoever or howsoever caused arising directly or indirectly inconnection with or arising out of the use of this material.

Journal of Personality Assessment, 93(5), 407–416, 2011Copyright C© Taylor & Francis Group, LLCISSN: 0022-3891 print / 1532-7752 onlineDOI: 10.1080/00223891.2011.594133

SPECIAL SECTION: Clinical Applications of the Adult Attachment Projective

The Adult Attachment Projective Picture System: IntegratingAttachment Into Clinical Assessment

CAROL GEORGE1 AND MALCOLM WEST2

1Department of Psychology, Mills College2Department of Psychiatry (Retired), University of Calgary, Canada

This article summarizes the development and validation of the Adult Attachment Projective System (AAP), a measure we developed fromthe Bowlby–Ainsworth developmental tradition to assess adult attachment status. The AAP has demonstrated excellent concurrent validity withthe Adult Attachment Interview (George, Kaplan, & Main, 1984/1985/1996; Main & Goldwyn, 1985–1994; Main, Goldwyn, & Hesse, 2003),interjudge reliability, and test–retest reliability, with no effects of verbal intelligence or social desirability. The AAP coding and classification systemand application in clinical and community samples are summarized. Finally, we introduce the 3 other articles that are part of this Special Sectionand discuss the use of the AAP in therapeutic assessment and treatment.

Attachment theory has been most characteristically the do-main of developmental psychologists concerned mainly withthe study of generalized normative development. Increasingly,this frame of reference is a source of considerable influence inadult clinical research and psychotherapy. Its rise to influencein clinical psychology is consonant with Bowlby’s original goalfor attachment theory as a foundational construct for under-standing personality functioning and psychopathology acrossthe life span (Ainsworth, 1989; Bowlby, 1969/1982, 1988), anapplication that has been substantiated by 40 years of normativeand clinical research (Cassidy & Shaver, 2008). Attachment the-ory is consonant with all assessment and treatment approachesthat evaluate childhood experiences as an important contrib-utor to adult functioning (e.g., Bakermans-Kranenburg, vanIJzendoorn, & Juffer, 2005; Blatt & Levy, 2003; Buchheim& George, 2011; Buchheim, George, & Kachele, 2008; Dia-mond, 2004; Wallis & Steele, 2001; Zegers, Schuengel, vanIJzendoorn, & Janssens, 2008).

The rapidly increasing clinical interest in attachment makesit important that we be clear about the differences betweenrepresentational measures of adult attachment and attachmentstyle. Developmental representational measures of adult attach-ment assess attachment status, that is, patterns of thinking andthe effects of childhood attachment experiences. Attachmentstyle is a personality measure that conceptualizes attachmentas two personality dimensions, avoidance and anxiety or posi-tive and negative self, essentially an attachment reinterpretationof Millon’s (1969) model of personality (e.g., Bartholomew &Horowitz, 1999). The content domain of attachment style as-sessments is limited to traits emerging from an empirical factor

Received March 25, 2010; Revised April 27, 2011.Address correspondence to Carol George, Department of Psychology, Mills

College, Oakland, CA 94613; Email: george@mills.edu

analysis approach rather than attachment theory. Indeed, the ex-pansive discussion of the attachment theory underpinnings forattachment style was added retrospectively after the measureshad been developed (Bartholomew & Horowitz, 1999; Hazan &Shaver, 1987). Following Bowlby (1969/1982) and Ainsworth(1964), developmental attachment assessments place an empha-sis on activating attachment and “seeing attachment in action”(i.e., in attachment activating contexts). By contrast, the attach-ment style approach asks individuals to report generically ontheir feelings about self and attachment figures, but attachmentis never observed in action. Further, recent attachment styletheorists have extrapolated this model from being a model ofromantic (i.e., sexual system) attachment to building a theoryof general human bonding, including coworkers or others notspecified by human biology as attachment figures, thus develop-ing a model of human intimate relationships more general thanthe specific relationship Bowlby postulated was central to men-tal health (e.g., Coan, 2008). Theorists and researchers fromthese two perspectives agree that these models do not over-lap, even though they both make claim to attachment theory(Crowell, Fraley, & Shaver, 2008; de Haas, Bakermans-Kranenburg, & van IJzendoorn, 1994; George & West, in press).The Adult Attachment Projective Picture System (AAP) frame-work we describe here represents the developmental attachmenttradition. The two main strengths of this approach are the eval-uation of unconscious defensive processes and assessment ofattachment disorganization and trauma (George & West, inpress; Solomon & George, 2011a).

At present, the Adult Attachment Interview (AAI; George,Kaplan, & Main, 1984/1985/1996; Main & Goldwyn,1985–1994; Main, Goldwyn, & Hesse, 2003) is the predom-inant method used to assess mental representation of attach-ment in adults. The AAI is a semistructured interview duringwhich the individual is asked to describe past and present re-lationships with parents and attachment-relevant events during

407

Dow

nloa

ded

by [

Car

ol G

eorg

e] a

t 09:

45 2

5 A

ugus

t 201

1

408 GEORGE AND WEST

childhood. The content of the interview is not just taken atface value, but rather subjected to a complex semantic analy-sis requiring several steps. Each interview is scored on several9-point scales; five on the actual experiences with parents andnine on aspects of the adult’s “current state of mind with respectto attachment.” From these scales, each interview is judged forthe overlying unresolved pattern and for one of three primaryattachment patterns: secure/autonomous, dismissing/avoidant,and preoccupied/enmeshed.

The AAI is a well-established measure with considerablestrengths. Its focuses on the assessment of the representationalmodel and assumes developmental continuity of the attach-ment system. It is methodologically precise in the evaluationof coherency and therefore has respectable interrater reliability.Finally, its validation rests on more than 25 years of devel-opmental and clinical research (Bakermans-Kranenburg & vanIJzendoorn, 2009).

Despite the strengths of the AAI, it has substantial draw-backs. From a practical standpoint, the interview takes 1 to2 hours to administer and its verbatim transcription and lengthycoding and classification process is time consuming, cumber-some, and costly for the result that is actually used in mostapplications; that is, the identification of individuals as secureor insecure or the assignment of them to one of four attach-ment groups. From a clinical standpoint, the AAI’s ability toassess attachment trauma is limited. The only validated assess-ments are for loss through death and physical abuse. The AAIrating scales and classification groups assumed to be associ-ated with trauma were developed because of problems apply-ing the AAI system to individuals in risk populations. Thesenew “trauma-related” classification groups have not been sub-ject to validity testing (George & West, in press; Hesse, 2008).Further, the administration instructions for the AAI requireinterviewers, rightfully, not to pursue material that individu-als do not wish to discuss. We and others have found thatsome trauma experiences are “edited out of the AAI tran-script,” so to speak, by the interviewee failing to discuss cer-tain experiences or to truncate the discussion (George & West,in press). Finally, designating state of mind using AAI failsto provide important information about attachment defensiveprocesses.

The AAP circumvents these practical and clinical prob-lems. As described below, the AAP provides clinicians witha construct valid measure of adult attachment status basedon the four-group classification model used by developmen-tal researchers that, as noted earlier, is trauma sensitive (e.g.,Buchheim & George, 2011; George & West, in press). Froma practical standpoint, the advantages of the AAP are con-siderable. Administration time for the AAP is relatively shortand coding by trained and reliable judges is expedited for theAAP.

This purpose of this article is to provide the essential back-ground for readers to understand the development and use ofthe AAP, including a summary of the AAP classification cod-ing system. This discussion also serves as the foundation andintroduction for the three other articles that form this issue’sSpecial Section, “Clinical Applications of the Adult Attach-ment Projective.” These articles present for the first time in theclinical literature the application of the AAP to clinical casesand multimethod psychotherapeutic assessment.

THE ADULT ATTACHMENT PROJECTIVE PICTURESYSTEM

It could be argued that a selection of Thematic ApperceptionTest (TAT; Murray, 1943) cards might have been used to elicitattachment-related material instead of “starting from scratch” bycreating a new set of picture stimuli. This way of proceeding,however, would not have been theory-driven. Although the TATstories are likely interesting with respect to attachment, thesepictures were not derived a priori from theoretical considera-tions. That is, the TAT scenes were not generated by attachmenttheory. Separation and loss are the conceptual foundations of at-tachment theory. Bowlby proposed that fear of separation, lossof attachment figures, and being left alone without protection arecentral to human evolution (Bowlby, 1969/1982, 1973, 1980).

We were influenced by Klagsbrun and Bowlby (1976), whofirst applied attachment thinking about separation and loss todevelop a projective test that assessed children’s representa-tional responses to these phenomena. This approach has morerecently been used to develop child assessments more con-sonant with contemporary attachment theory than the origi-nal Klagsbrun–Bowlby assessment (Bretherton, Ridgeway, &Cassidy, 1990; Solomon, George, & De Jong, 1995). We de-cided, therefore, to follow this construct-oriented approach anddevelop an attachment theory-based set of picture stimuli thatportray important features within the definition of attachment.By using stimuli that represent the potential threat of separa-tion, loss, and aloneness, the AAP could then be validated andused as a developmental measure of attachment in conjunc-tion with other validated attachment assessments (Ainsworth,Blehar, Waters, & Wall, 1978; Cassidy & Marvin, 1987–1992;Main & Cassidy, 1988; Solomon et al., 1995).

The AAP picture selection process was based on three coreattachment features (Bowlby, 1969/1982). The first, and per-haps most important feature, is that attachment can only be as-sessed when the attachment system has been activated. There-fore, in developing the picture set, theory-defined scenes thatelicit attachment distress depicting the most prominent attach-ment activators in the Bowlby–Ainsworth approach, such asseparation, solitude, fear, and death, were included. The secondfeature is variations in the perceived accessibility of attach-ment figures; some stimuli portray individuals alone and othersportray individuals in attachment dyads. Perceived availabilityof an attachment figure is a defining feature that differentiatessecure versus insecure patterns of attachment. Attachment fig-ure availability, combined with responsive and effective care,is central to internal working models of attachment security(Ainsworth et al., 1978; Bakermans-Kranenburg & van IJzen-doorn, 1997; Bakermans-Kranenburg, van IJzendoorn, & Juffer,2003; Sroufe, Egeland, Carlson, & Collins, 2005). Attachment isat risk of becoming disorganized when children feel abandonedand appraise attachment figures as absolutely unavailable to pro-vide protection and care (George & Solomon, 2008; Solomon& George, 2011a, 2011b). Infants and young children requireattachment figures to be physically present, accessible, and re-sponsive. By contrast, adolescents and adults predominantly use“psychological” or representational proximity instead of phys-ical access to the attachment figure (Allen, 2008; Weiss, 1982;West & Sheldon-Keller, 1994). The third feature is the portrayalof attachment across the life span. Bowlby proposed that at-tachment is important across the life span, from infancy through

Dow

nloa

ded

by [

Car

ol G

eorg

e] a

t 09:

45 2

5 A

ugus

t 201

1

ADULT ATTACHMENT PROJECTIVE PICTURE SYSTEM 409

old age (Bowlby, 1969/1982, 1973, 1980). AAP scenes depictcharacters that range in age from children to the elderly.

The task begins with a warm-up stimulus. This is a pictureof two children playing ball, which is neutral with regard toattachment. The task then presents the individual with seven at-tachment stimuli : child at window (window)—a child looks outa window; departure—an adult man and woman stand facingeach other with suitcases positioned nearby; bench—a youthsits alone on a bench; bed—a child and woman sit oppositeeach other on the child’s bed; ambulance—a woman and a childwatch ambulance workers load a covered stretcher into an am-bulance; cemetery—a man stands by a gravesite headstone; andchild in corner (corner)—a child stands askance in a corner.The stimuli were drawn to contain only sufficient detail to iden-tify attachment contexts; actions, facial expressions, postures,and background elements are ambiguous. The characters arediverse with regard to ethnicity, gender, and age. Examples areprovided in Figures 1 and 2 (see George & West, in press, for thecomplete AAP picture stimuli set and description of stimulusdevelopment).

AAP administration takes approximately 30 minutes and isdone in a private setting. The administration instructions inte-grate traditional elements of free response tasks with semistruc-tured interview techniques. The individual is given the drawingto hold and asked to describe what is going on in the pic-ture, including the background of events, characters’ thoughtsor feelings, and the outcome. The responses are audiotaped fortranscription and verbatim analysis; transcripts are typically twoto three pages. Administrators must be trained in administrationtechnique but do not need to be attachment experts or trained inthe AAP coding and classification scheme. Coding and classifi-cation by a trained reliable judge typically takes about 1 hour.

The AAP coding and classification scheme, which is sum-marized in the following section, identifies the theoreticallydefined core attachment constructs integrated into AAP codingthat were developed following the Ainsworth–Bowlby tradi-tion. We also integrated into the AAP coding scheme salientfeatures of other validated child and adult attachment assess-ments that contribute to differentiating patterns of attachment,especially parent–child interaction on reunion (Strange Situa-tions; Ainsworth et al., 1978; Main & Cassidy, 1988; Marvin

FIGURE 1.—Adult Attachment Projective System stimulus: Bench. © George& West (in press).

FIGURE 2.—Adult Attachment Projective System stimulus: Departure. ©George & West (in press).

& Britner, 2008) and defensive processes in child attachment-based doll play and parent’s caregiving assessments (George &Solomon, 1989, 2008; Solomon et al., 1995).

The AAP was validated in a large, racially diverse commu-nity sample that was recruited from September 2002 to August2003 from community and college settings using newspaperand Internet advertisement.1 An overview of the psychomet-ric study results are provided here. The details of this researchand findings are published in George and West (in press). Theresults of independent studies bearing on AAP validity arealso published in other journals.2 The main validity study in-cluded 144 participants that made up a racially diverse, educated(M = 14.7 years of schooling), community sample of womenand men living in Canada and the United States. Participantsranged in age from 18 to 72 years old (100 women, M age =36.2; 44 men, M age = 26.4). The AAP classification distribu-tion was as follows: 25 (17%) were secure (F); 37 (26%) wereclassified as dismissing; 30 (21%) were preoccupied; and 52 ofthe 144 participants (36%) were unresolved. Age, gender, na-tionality, and years of education were not related to attachmentgroup classification.

We validated the AAP following a design analogous to thatused originally to validate the AAI, the international goldstandard of adult attachment (Bakermans-Kranenburg & vanIJzendoorn, 1993). We evaluated AAP concurrent validity

1Preliminary validation of the AAP, which included a portion of this samplecombined with two development samples, was reported in George and West(2001).

2AAI concordance and interjudge relability has also been established inseveral studies using the AAP in the German language. Reliable blind judgesclassified AAPs and AAIs from German transcripts and AAPs from Englishtranscripts (see Buchheim & George, 2011). Psychometric validity data fromstudies independent of collaborative work with the AAP codevelopers is avail-able in other publications (e.g., Aikins, Howes, & Hamilton, 2009; Beliveau &Moss, 2005; Benoit, Bouthillier, Moss, Rousseau, & Brunet, 2010; Van Ecke,2006).

Dow

nloa

ded

by [

Car

ol G

eorg

e] a

t 09:

45 2

5 A

ugus

t 201

1

410 GEORGE AND WEST

against the AAI, and also evaluated interjudge reliability andtest–retest validity. Concurrent validity was established by com-paring AAP classifications with independent blind AAI classifi-cations.3 The concordance rates for four-group classification (se-cure, dismissing, preoccupied, unresolved) and secure–insecuregroups were strong—90% and 97% classification matches re-spectively (all kappas and phis, p < .001; Pearson’s r = .84 and.88, respectively). Interjudge reliability was established com-paring the classifications of three expert AAP judges who wereblind to all information about participants, such as AAI clas-sifications, age, and gender. All transcripts were scored by atleast two judges. The concordance rates across all AAP classi-fications were well above accepted rates for interjudge reliabil-ity for four-group and secure–insecure classification agreement(all kappas and phis, p < .001 level; Pearson’s r = .70–.89).Test–retest reliability was established on a random sample ofparticipants. Our study design was to randomly contact 50% ofthe sample to return for testing 3 months following the originalassessment. Forty eight percent of the sample participated (n= 69; 39 women, 30 men). The main reasons for withdrawalfrom the retest portion of the study were that a participant hadmoved away or did not have time to come into the laboratory.Stability from the first testing was significant (kappa and phi, p< .001; Pearson’s r = .70, p < .001); 84% of the retest samplewas classified in the same main categories.

Following the Bakermans-Kranenburg and van IJzendoorn(1993) AAI validity study, we also evaluated verbal intelligenceand social desirability as related to AAP responses. Verbal in-telligence and social desirability had been shown not to influ-ence AAI responses (Bakermans-Kranenburg & van IJzendoorn,1993), so we similarly expected that these dimensions would notinfluence the AAP responses. We first resolved all AAP classi-fication disagreements through conferencing prior to evaluatingthese dimensions. The results showed no influence of verbalintelligence (WAIS–R) or social desirability (Balanced Inven-tory of Desirable Responding [BIDR]; Paulhaus, 1998) on theAAP classifications (George & West, in press). There were nostatistically significant differences among attachment groups onWAIS scores on the Verbal (means ranged from 12.40–13.45)and Similarities subtests (means ranged from 11.50–13.43).There were no statistically significant differences among at-tachment groups on BIDR scores for impression management(IM; means ranged from 6.95–8.73), self-defensive enhance-ment (SDE; means ranged from 2.06–3.31), or the BIDR total(means ranged from 9.19–12.40).

The AAP Coding and Classification System

Attachment, as originally defined by Ainsworth’s assessment,involves evaluating patterns of responsiveness (Ainsworth et al.,1978). Developmental attachment status, in contrast to person-ality attachment style (Crowell et al., 2008), cannot be assessedusing a single or paired set of orthogonal dimensions. Attach-ment classification using the AAP is determined by evaluatingpatterns of responses using a set of coding dimensions that

3The original AAI study used a verbal intelligence measure on a Dutch sam-ple that was equated with the 1972 Wechsler Adult Intelligence Scale (WAIS;Wechsler & Matarazzo, 1972). We used the Vocabulary scale in the 1981 versionof the Wechsler Adult Intelligence Scale–Revised (WAIS–R; Wechsler, 1981)in an effort to parallel the original AAI validation study. Social desirability wasmeasured using the Inventory of Desirable Responding (Paulhaus, 1998).

TABLE 1.—Adult Attachment Projective System coding system.

Dimension Stimuli Definition

Story contentAgency of self Alone Story character takes steps

to address the situation.Integrated; internalized secure

base, haven of safety;capacity to act; no agency.

Connectedness Alone Renew or makeinterpersonal

connections inrelationships with

others.

Makes connections;connections are blocked orconnections to stranger; no

connection with others.

Synchrony Dyadic Reciprocal and mutualgoal-corrected

partnership.

Mutual enjoyment or sensitivecare; functional interaction

or interactive failure; norelationship present.

DefenseDeactivation All Shift attention away from

attachment; deflectawareness of distress.

Rules, power, achievement,authority, social scripts,

distance, romance,rejection, neutralize.

Cognitivedisconnection

All Splintered or fractureddescriptions of

attachment events andemotions.

Uncertainty, opposite themes,anxiety, heightenedemotional arousal,

withdraw, withhold, glossover.

Segregatedsystems

All Overwhelmed byfrightening and

threatening experiences.

Helplessness, frightened, outof control, isolated,

unprotected, abandoned;disturbing themes, one’s

own traumatic story;constriction.

DiscourseSelf–otherboundary

All Story includes own lifeexperience in response.

Present; absent.

evaluate the attachment story content related to the hypotheti-cal characters portrayed in the stimuli, defense, and self–otherboundaries in narrative discourse (George & West, 2001, inpress; George, West, & Pettem, 1999).

Story content. These dimensions evaluate the attachmentmeanings and features of self and relationships that are revealedin the response regarding the hypothetical character(s) and story(see Table 1). Agency of self and connectedness are coded foralone stories; synchrony is coded for dyadic stories.

The agency of self dimension describes the character’s abil-ity to take steps to address the situation that confronts him orher in each picture. The development of this self-characteristicis influenced by an individual’s experiences in childhood withattachment figures (Sroufe et al., 2005) and later in life, includ-ing possibly psychotherapy or relationship with a long-termadult partner (Sroufe et al., 2005; Waters & Hamilton, 2000).We view agency as “the capacity to engage in behavior thatproduces change, the use of relationships to re-establish attach-ment equilibrium, and/or the capacity to enter and actively ex-plore one’s own internal working model” (West & George, 2002,p. 281). Agency of self is required for the “alone” self to pre-serve integrity under stress (e.g., problem solving, separation,relationship repair following an argument) and also contributesto maintaining organized thought and behavior when the individ-ual is flooded by feelings of desperation, helplessness, isolation,fear, or threat.

The AAP coding system identifies two levels of agency ofself. One level, integrated agency, refers to the qualities ofintegrity of self that are associated with balance and confidencethat underpin attachment security (Ainsworth, 1989; Bowlby,1969/1982, 1980; George & Solomon, 2008; Solomon etal., 1995). A form of integrated agency is haven of safety,

Dow

nloa

ded

by [

Car

ol G

eorg

e] a

t 09:

45 2

5 A

ugus

t 201

1

ADULT ATTACHMENT PROJECTIVE PICTURE SYSTEM 411

evidenced in the AAP when the character receives sensitive careor reaches out to repair or reintegrate a relationship. Anotherform is internalized secure base, a new attachment concept thatwe developed for the AAP to capture the capacity for thought-ful self-exploration (George et al., 1999). We derived thisconcept from the secure base phenomenon, which Ainsworthdefined as the ability to explore based on the confidencethat the attachment figure is available and sensitive to thechild’s attachment needs (Ainsworth et al., 1978; Waters &Deane, 1985). Attachment representation affords the childpsychological proximity to the attachment figure even in theparent’s physical absence (Bowlby, 1969/1982). We define theconcept of internalized secure base to refer to that state in whichsecurity and self-integrity are derived from the individual’sinternalized relationship to the attachment figure. It is evidencedin the AAP by themes that describe a character using solitudeto explore the thoughts and feelings in his or her internal world.

A second level of agency reflects the self’s capacity totake constructive action in attachment situations, termed ca-pacity to act. This action successfully removes the individ-ual from threat, includes self-protective behavior, and can oc-cur irrespective of the individual’s willingness to think aboutthe source of distress. It is evidenced in the AAP by descrip-tions of the character engaging in functional or problem-solvingbehavior.

The connectedness dimension evaluates the propensity of theindividual to renew or make interpersonal connections in inti-mate relationships. In addition to attachment relationships, otherbiologically defined relationships thought to be important inhuman development include friendship and adult pair bonds.Indeed, Ainsworth and others suggest that these relationshipsare important sources of emotional support (Ainsworth, 1989;Allen, 2008; George & Solomon, 2008; Zeifman & Hazan,2008). Connectedness is evidenced in the AAP when a char-acter who is alone demonstrates the desire and capacity to seekattachment figures, friends, or adult partners. Characters are notconnected when they are blocked from individuals in these re-lationships, or when they only seek the help or company ofstrangers. Relationships with others are sometimes absent in re-sponses, suggesting representations of self as alone and lackingthe capacity or desire to seek out others.

Synchrony evaluates whether or not the attachment–caregiving relationships portrayed in the dyadic stimuli areviewed as a “goal-corrected partnership.” Following Bowlby’s(1969/1982) discussion of the goal-corrected partnership,synchrony is evidenced in dyads by descriptions of recip-rocal interaction and demonstrates a commitment to theattachment–caregiving relationship (see also Marvin & Britner,2008). The most integrated form of synchrony is descriptions ofreciprocal interaction that indicate either the characters’ mutualenjoyment of being together or the attachment figure’s sensitiv-ity to the other character’s distress and vulnerability. Functionalsynchrony will keep attachment distress at bay, even when com-plete relationship reciprocity is not desirable or possible (e.g.,the adult explains the situation to the child; a teacher inter-venes on the child’s behalf). Functional synchrony sometimesentails the attachment figure responding to the situation but notto the individual’s attachment need (e.g., mother turns on thenight light but does not respond directly to the child’s cue for agoodnight hug).

Defensive processes. In Loss, Bowlby (1980) discussed atlength his view of defense, representation, affect regulation, andthe relation between defensive processes and psychiatric symp-toms. His unique approach was to consider defensive processes,a core psychoanalytic concept, from an information processingperspective. He viewed defense as a set of automatic, uncon-scious,4 attentional processes that select, exclude, and trans-form behavior, thought, and emotions to shift attention awayfrom attachment distress and prevent psychological breakdown(see also Hesse & Main, 2006). Bowlby (1980) delineated threeforms of defensive exclusion: deactivation, cognitive discon-nection, and segregated systems. Although the importance ofdefense is acknowledged in the field of attachment (e.g., Brether-ton, 2005; Bretherton & Munholland, 2008; Cassidy & Kobak,1988), George and Solomon are the only researchers who haveoperationally defined and validated Bowlby’s approach to makeit accessible for assessment (George & Solomon, 1989, 2008;Solomon et al., 1995). Further, they have shown that identifyingan individual’s primary defensive strategy is a key step to attach-ment classification; identifying patterns of defensive strategiesprovides important insight into differences between attachmentclassification groups and the unique characteristics of individu-als placed in the same group. Following this work, we developeda set of defining criteria for identifying Bowlby’s three formsof defensive processes in the AAP. This approach to assess-ing attachment is unique and the AAP is the only attachmentassessment that provides descriptions of the defenses of adultattachment “in action.” Evidence of defense is coded from thewords and images produced in the story narrative (see Table 1).

Deactivating defensive processes are defined as attempts toshift attention away from attachment events, individuals, or feel-ings. Deactivation works to deflect and prevent the individ-ual from becoming conscious of attendant attachment distress.Deactivation is coded for story themes that emphasize the im-portance of rules, social scripts, power, achievement, authority,distance, or romance (diversion of the attachment system to thesexual system). If distress does enter consciousness and charac-ters are described as distressed or in need of care, deactivatingdefenses produce evaluations of individuals as not deservingcare and attachment needs as rejected or neutralized.

Cognitive disconnection processes literally disconnect the el-ements of attachment from their source, thus undermining con-sistency and the capability of holding in one’s mind a unitaryview of events, emotions, and the individuals associated withthem. The result is confusion and representational shifts and

4Bowlby (1980) defined the term unconscious as processes occurring out ofconscious awareness drawing from theory and research defining the mid-20th-century cognitive revolution in psychology. Bowlby’s goal was to describe pat-terns of unconscious processes that filtered or excluded certain experiences andaffects that were directly related to an individual’s experience with attachmentfigures. These processes became the attachment approach to defense. Unlikethe cognitive information processing models from which he drew, however,Bowlby’s view emphasized affect (Bowlby, 1973, 1980). Bowlby’s approach tothe unconscious is only one of the attachment theory postulates derived fromredefining psychoanalytic concepts based on empirical research in cognitive anddevelopmental psychology and ethology. Other examples include his reformu-lation of the child’s internal world as a “cognitive-affective” internal workingmodel and his reformulation of the foundation of mother–baby relationshipsbased on Darwinian evolutionary theory and nonhuman primate research.

Dow

nloa

ded

by [

Car

ol G

eorg

e] a

t 09:

45 2

5 A

ugus

t 201

1

412 GEORGE AND WEST

oscillations evidenced in the AAP by uncertainty of thought.For example, the individual is unable to make decisions aboutcharacters or events, sometimes to the point of being inextri-cably caught between opposing themes (e.g., it is nighttime ordaytime; the girl is happy or sad). Disconnection is also as-sociated with story elements of anger and frustration that aredealt with by withdrawal, withholding, or attempts to gloss overattachment difficulties in hopes that they will simply go away.

George and Solomon viewed deactivation and cognitive dis-connection as normative forms of defense (George & Solomon,2008; Solomon et al., 1995). What is meant by “normative” isthat these processes successfully divert attention from or splinterattachment events, memories, and feelings sufficiently to keepattachment organized. These forms are the primary defensivestrategies associated with organized attachment classifications(i.e., secure, dismissing, or preoccupied). Normative defenseshelp prevent individuals from becoming flooded and immobi-lized by levels of fear or threat that disorganize attachment(Lyons-Ruth & Jacobvitz, 2008; Main & Hesse, 1990; Main& Solomon, 1990; Solomon & George, 2011a; Solomon et al.,1995).

Disorganized attachment is associated with Bowlby’s (1980)third form of defense, segregated systems (George & Solomon,2008; George et al., 1999; Solomon et al., 1995). The seg-regated system is defined as an intensified defensive processthat locks away the painful “package” of memories and af-fects associated with a particular threatening attachment re-lationship from conscious awareness (in information process-ing terms, working memory). Analogous to the psychoanalyticconcept of repression, viewing defense from an informationprocessing perspective provides a new level of mechanism clar-ity. Segregated systems are thought to provide a greater pro-tective shield than the normative exclusion processes of de-activation and cognitive disconnection. Bowlby proposed thatactivation of these segregated models resulted in such an in-tensity of emotion that thought and behavior become disorga-nized and chaotic. When segregated system defense that hasbeen heavily relied on begins to fail, we can anticipate thebreakthrough of traumatic or frightening material describedlater.

Bowlby’s approach in defining segregated systems helps usunderstand more clearly the juxtaposition of signs and symp-toms of rigid control, frozen constriction, frightened hypervig-ilance, oversensitivity, and dysregulation, and the psychiatricsymptoms of individuals who are classified in the child dis-organized/controlling and adult unresolved attachment groups,especially when their experiences encompass severe abuse orterror (e.g., Carlson, 1998; Hesse & Main, 2006; Liotti, 2004;Solomon & George, 2011a).

The first step in coding segregated systems in the AAP isto mark the transcript for evidence of segregated systems. Thenext step is to evaluate whether or not the segregated materialis reorganized or contained in the individual’s story response.In the instance of failure of segregated system defense and thefailure to contain the breakthrough, the individual is designatedas unresolved, the AAP group that is analogous to the AAIunresolved attachment group (Hesse, 2008). Evidence of seg-regated systems in the AAP includes any feature of the indi-vidual’s response that connotes helplessness, fear, being out ofcontrol, being isolated, being unprotected, or being abandoned.Some indexes of segregated systems have an eerie quality, a fea-

ture that parallels the attachment research links to dissociativesymptoms. Other indexes include individuals describing theirown traumatic stories, a discourse dimension that is similar tothe lapses in monitoring associated with AAI lack of resolution(Hesse, 2008). Sometimes individuals freeze while taking thetask, handing the stimulus picture back to the administrator orsaying in no uncertain terms that they can’t provide a responseto that stimulus. This response, termed constriction, was firstidentified in children’s doll play and is indicative of dysregu-lated attachment (Solomon et al., 1995). The responses are nextevaluated with regard to whether or not the individual is ableto contain or reorganize representational dysregulation, called“resolution” following the nomenclature in the field. This reor-ganization occurs when there is evidence of agency of self orfunctional assistance from others. The failure to reorganize orcontain segregated systems indicates that attachment remainsdysregulated and the classification is designated unresolved.

Discourse: Self–other boundaries. The responses are alsoevaluated for the individual’s ability to maintain self–otherboundaries while responding to the AAP stimuli. The task is de-fined as telling a story about hypothetical characters and events.Although we know that one’s representation of attachment isbuilt on real experience, the AAP is not an autobiographical in-terview. The “invasion” of material about one’s own experience,therefore, demonstrates a failure to maintain the boundaries be-tween self and other during the task. We call this discourse intru-sion personal experience, and make a designation in coding foreach story in which “personal experience” material is includedin the response. This material contributes important informa-tion in determining whether or not the individual’s attachmentdistress is unresolved; descriptions of one’s own frighteningor traumatic experiences that are not contained or reintegratedduring the response are indicative of an unresolved attachment.

Classification. Attachment classification group is assignedon the basis of an analysis of the coding patterns across the entireset of seven attachment stories. The classification assignment isdetermined using the hierarchically integrated series of decisionpoints illustrated in Figure 3. Each decision point is defined inrelation to the overall patterning of coding dimensions for thealone and dyadic pictures.

Unresolved Attachment: The characteristic quality of un-resolved attachment is the individual’s inability to contain andreorganize stories that contain segregated systems markers.What this means is that unresolved individuals are, even mo-mentarily, flooded by their attachment fears and they cannotrecover. The story content and underlying defensive processesmight be similar to any of the other attachment groups.

Secure Attachment: What characterizes individuals whoare secure in the AAP is their ability and willingness to thinkabout attachment distress or reach out to attachment figures forassistance and comfort. In response to the alone stimuli, secureindividuals demonstrate the ability to think (i.e., internalized se-cure base) when they are alone, which then results in the charac-ters taking constructive action (e.g., the girl on the bench thinksabout why she was fighting with her parents and goes home totalk to them). Secure individuals also demonstrate a view thatattachment relationships involve a goal-corrected partnership in

Dow

nloa

ded

by [

Car

ol G

eorg

e] a

t 09:

45 2

5 A

ugus

t 201

1

ADULT ATTACHMENT PROJECTIVE PICTURE SYSTEM 413

StartUnintegrated or uncontained segregated systems marker(s)

Yes

Unresolved

No

Evidence of integrated attachment. Integrated agency, Connectedness Synchrony Scores (and, if present, segregated systems markers are integrated or contained)

Yes Secure

No

Deactivation markers in 3 or more stories (and, if present, segregated systems markers are reintegrated or contained)

YesDismissing

No

Cognitive disconnection markers (and, if present, segregated systems markers are reintegrated or contained)

YesPreoccupied

→→

→

→

→

→→

→

FIGURE 3.—Adult Attachment Projective System classification rules.

response to the dyadic stimuli. This can be described as sen-sitivity to distress that results in comfort or mutual enjoyment.Secure transcripts are not “perfect.” Integration and mutualityare rarely demonstrated in every story. Rather, the judge seesfrom the pattern of story responses that the person genuinelyvalues and represents the capacity for integration of self andrelationships above all other qualities of the response. What isalso visible is that secure individuals use defensive processes tohelp integrate attachment feelings and events, not to exclude andtransform them into something else. Secure individuals rarelyshow blurring of personal experience and hypothetical storyboundaries in their responses. This seems only to occur whensecure individuals are highly distressed.

Dismissing Attachment: Dismissing attachment is definedby the predominance of deactivating defensive processes thatmaintain distance in relationships. Stories often shift attentionaway from distress to a focus only on achievement and explo-ration. When distress is evident, dismissing individuals tend todescribe feelings of negative evaluation, such as being punished,and rejection that keep characters at a distance. The stories ofsome dismissing individuals suggest that they are quite savvyabout how to describe what looks on the surface as thinking,but the focus is typically superficial or focuses on the self andnot on the attachment relationship. Other dismissing individu-als simply solve problems by action (agency = capacity to act)and not necessarily thinking about them (i.e., no internalizedsecure base). Attachment figures are portrayed either as pro-viding functional care (e.g., goes to school to talk to the boy’steacher instead of comforting the boy) or as authoritarian (e.g.,tells the child the rules to be followed). As such, connectednessand synchrony codes typically reflect functional interactions aswell. Blurring personal experience with the hypothetical storyis unusual, demonstrating how in most dismissing individualsdeactivation keeps strict self–other boundaries in relationships.

Preoccupied Attachment: Preoccupied attachment is de-fined by the predominance of cognitive disconnection defensiveprocesses that confuse and obscure attachment relationships.Preoccupied attachment typically focuses on emotions relatedto problems rather than the problems themselves. Stories attemptto shift attention away from distress by making the situation oroutcome overly cheery (i.e., glossing) so as to prevent havingto face negative affect. Preoccupied individuals are more often,though, confused about attachment events, feelings, and thoseinvolved, and their responses have several undecided themes,story endings, and sometimes many different complete stories.Preoccupied individuals are often sentimental without perspec-tive and focus on what was rather than what is (e.g., the manat the cemetery remembers his many happy years before hiswife died). Agency in the alone stories is rare—the charactersmight engage in minimal agency that does not get them veryfar (e.g., the girl on the bench gets up but this action does nottake her away from the source of her problems, the bench itself).Characters in the alone responses often remain alone, not mak-ing connections with anyone. As with dismissing attachment,attachment figures might be portrayed as providing functionalcare (e.g., the mother in the bed cannot figure out what the boywants, but finally gives him a hug). Attachment figures mightalso be portrayed as not responding at all (e.g., the boy on thebed wants a hug and his mother leaves the room, never acknowl-edging the child’s signal). Blurring personal experience with thehypothetical story is quite common in preoccupied responses,demonstrating how difficult it is for preoccupied individuals tomaintain self–other boundaries in relationships.

Training to Administer and Use the AAP

Ethical use of the AAP requires training in administration andcoding. Following the standard of all attachment assessments,individuals are considered reliable judges in the AAP whenthey have reached 80% concurrence on standard reliability sets.Training begins with an 8-day workshop that first provides afoundation in attachment theory and subsequently builds on thisfoundation to teach AAP coding and classification. Trainees areprovided with practice cases and reliability cases as part of thetraining. Reliability is established as 80% on a minimum of 30cases. Another alternative in using the AAP is to be trainedin administration and have reliable judges code and classify thecases. Administration training is done by providing feedback onpractice case administration. Information about reliable judgesavailable for coding is provided as part of the administrationtraining process.

Other Work Using the AAP

AAP validity and interrater reliability with English andwith non-English-speaking participants have been demonstratedby independent investigators (e.g., Beliveau & Moss, 2005;Buchheim et al., 2006; Webster & Hackett, 2007; Webster &Knoteck, 2007). The AAP has been used in basic researchand hypothesis testing, establishing predictive validity in study-ing mothers’ adult attachment status in relation to caregivingand their children’s adjustment and developmental risk, moth-ers of children diagnosed with Attention Deficit Hyperactiv-ity Disorder, the attachment underpinning related to immigra-tion, and the developmental correlates of unresolved attachmentin adolescent maltreatment (Aikins et al., 2009; Beliveau &

Dow

nloa

ded

by [

Car

ol G

eorg

e] a

t 09:

45 2

5 A

ugus

t 201

1

414 GEORGE AND WEST

Moss; Kissgen et al., 2009; Van Ecke, 2005, 2006; Webster &Hackett, 2007; Webster, Hackett, & Joubert, 2009; Webster &Knoteck, 2007). Clinical studies have used the AAP to exam-ine posttraumatic stress disorder symptoms following trauma,depression, and emotional development in psychiatric patients,and as part of a parent evaluation battery in custody evalua-tion (Benoit et al., 2010; Buchheim & George, 2011; Buch-heim, George, Liebl, Moser, & Benecke, 2007; Isaacs, George,& Marvin, 2009; Subic-Wrana, Beetz, Langenbach, Paulussen,& Beutel, 2007; West & George, 2002). The AAP has alsobeen shown to be useful in studying the neurobiological andemotional expression correlates of attachment in nonclinicaland clinical samples (Buchheim & Benecke, 2007; Buchheim,Erk, et al., 2008; Buchheim et al., 2006; Buchheim et al.,2009; Buchheim et al., 2007; Fraedrich, Lakatos, & Spangler,2010).

There are a number of reasons to enlist the AAP in clini-cal practice. First, it is economical. Second, because it is notan autobiographical interview, the AAP lends itself especiallywell to a range of populations in which the discussion of attach-ment using an interview format might be difficult or threatening(e.g., Szajnberg & George, 2011). Third, the AAP is easily in-tegrated with other assessments, including interviews that tapautobiographical experience such as the AAI. In this regard, wetypically combine the AAP with the AAI and a trauma inter-view (Blake et al., 1996) in our work. The trauma interviewincreases our comprehension of real-life trauma situations andoften completes such gaps in the AAI. The AAP and AAI canbe administered by the same person because there is no overlapin their content or format. We examined AAP and AAI for ad-ministration order effects in our validity study and found none.As well, colleagues using the AAP in therapeutic assessmentoften combine it with the Early Memories Procedure (Bruhn,1992). Fourth, the AAP is amenable to repeated administrationto identify and elaborate indications of a client’s gains or set-backs in the course and outcome of therapy. Finally, the AAPcan be used with clients to help sharpen the agenda of therapyand facilitate the organization of the therapeutic process (seeFinn, 2011/this issue).

The three other articles that are part of this Special Sectionintroduce the reader to the use of the AAP in specific assessmentand therapy situations. Each article provides verbatim AAP re-sponses and extensive interpretation of those responses usingthe AAP coding system. The collection begins with an articleby Webster and Joubert (2011/this issue) on the use of the AAPas a significant aid in planning intervention for a maltreatedteen in foster care. Maltreatment and foster care are attach-ment traumas (George & West, in press); these experiencesthreaten safety and the attachment bond itself. Their articleshows through total AAP analysis how disruptive behavior isarticulated in attachment terms. Specifically, the pattern of thisadolescent’s representation of agency of self, synchrony, andconnectedness, together with her defensive operations, iden-tify and define the underpinnings of her behavior problems.This case also demonstrates how attachment information in atest battery (e.g., standardized cognitive and behavioral assess-ment and the Rorschach) is integrated with recommendations ofother professionals (e.g., court-appointed representative, thera-pist, teacher).

Contrary to what might be expected for most foster and mal-treatment cases, this adolescent was judged secure on the AAP

as indicated by her strong capacity to think about attachment, astrength not detected by other professionals. The authors offera clear analysis of what it means to be secure in the context ofdisruptive behavior or psychopathology.

In the second article, Finn (2011/this issue) describes the useof the AAP in the course of long-term psychotherapy with amiddle-aged man to enhance his commitment to therapy. Finnaddresses the processes associated with his client’s dismiss-ing attachment. In particular, he describes how the deactivatingdefense of dismissing attachment interfered with his client’s ca-pacity to complete the process of mourning and reorganize hissense of self in relation to childhood problems with his parents.A unique aspect of Finn’s article is his description of how theAAP was used in the feedback step that is core to the therapeuticassessment model (Finn, 2007). Working with the first author,Finn wrote his client a two-part letter. The first part describedin a client-friendly way the etiology and meaning of dismissingattachment. The second part discussed the personal meaning ofhis client’s pattern of dismissing attachment. This letter sen-sitively, yet firmly, captured his client’s problems and was animportant catalyst in helping him continue in therapy and makegenuine changes in his life.

The final article in the collection by Lis, Mazzeschi, Di Riso,and Salcuni (2011/this issue) describes the inclusion of the AAPin a multimethod assessment of a severely anorexic adolescentgirl and shows how attachment assessment strengthens the in-cremental validity in an assessment battery (see also Isaacs etal., 2009). In particular, Lis and colleagues discuss their client’sunresolved classification and describe how the AAP combinedwith other assessments to delineate thinking about a range oftreatment options.

These articles describe only a few of the ways in which attach-ment can be used in clinical settings. Attachment assessment isnot a measure of the veracity of the details of a client’s actualattachment past. The AAP assesses current views of self, at-tachment figures, and expectations about the productiveness ofattachment relationships, elucidating how current experience ac-tivates attachment accomplishment, disappointment, and traumafrom the past (George & West, in press; West & Sheldon-Keller,1994). It is this current state of mind regarding attachment that25 years of research has demonstrated to be an important pre-dictor for psychiatric and relationship health (Cassidy & Shaver,2008).

ACKNOWLEDGMENTS

This study was supported by grants from the Social Scienceand Humanities Research Council of Canada, the Barrett Foun-dation, and a faculty development grant from Mills College.

REFERENCES

Aikins, J. W., Howes, C., & Hamilton, C. (2009). Attachment stability andthe emergence of unresolved representations during adolescence. Attachmentand Human Development, 11, 491–512. doi:10.1080/14616730903017019

Ainsworth, M. D. S. (1964). Patterns of attachment behavior shown by the infantin interaction with his mother. Merrill Palmer Quarterly, 10, 51–58.

Ainsworth, M. D. S. (1989). Attachment beyond infancy. American Psycholo-gist, 44, 709–716.

Ainsworth, M. D. S., Blehar, M., Waters, E., & Wall, S. (1978). Patterns ofattachment: A psychological study of the Strange Situation. Hillsdale, NJ:Erlbaum.

Dow

nloa

ded

by [

Car

ol G

eorg

e] a

t 09:

45 2

5 A

ugus

t 201

1

ADULT ATTACHMENT PROJECTIVE PICTURE SYSTEM 415

Allen, J. P. (2008). The attachment system in adolescence. In J. Cassidy & P.R. Shaver (Eds.), Handbook of attachment: Theory, research and clinicalapplications (2nd ed., pp. 419–435). New York, NY: Guilford.

Bakermans-Kranenburg, M. J., & van IJzendoorn, M. H. (1993). A psychometricstudy of the Adult Attachment Interview: Reliability and discriminant valid-ity. Developmental Psychology, 29, 870–879. doi:10.1037/0012–1649.29.5.870

Bakermans-Kranenburg, M., & van IJzendoorn, M. (1997). Sensitivity and at-tachment: A meta-analysis of parental antecedents of attachment. Child De-velopment, 68, 571–591.

Bakermans-Kranenburg, M. J., & van IJzendoorn, M. H. (2009). The first 10,000Adult Attachment Interviews: Distributions of adult attachment representa-tions in clinical and non-clinical groups. Attachment and Human Develop-ment, 11, 223–263. doi:10.1080/14616730902814762

Bakermans-Kranenburg, M. J., van IJzendoorn, M. H., & Juffer, F. (2003). Lessis more: Meta-analyses of sensitivity and attachment interventions in earlychildhood. Psychological Bulletin, 129, 195.

Bakermans-Kranenburg, M. J., van IJzendoorn, M. H., & Juffer, F. (2005).Disorganized infant attachment and preventive interventions: A review andmeta-analysis. Infant Mental Health Journal, 26, 191–216.

Bartholomew, K., & Horowitz, I. M. (1999). Attachment style among youngadults: A test of a category model. Journal of Personality and Social Psy-chology, 61, 236–244.

Beliveau, M.-J., & Moss, E. (2005). Contribution aux validites convergente etdivergente du projectif de l’attachement adulte [Contribution to the conver-gent and divergent validity of the Adult Attachment Projective]. La RevueInternationale de l’Education Familiale, 9, 29–50.

Benoit, M., Bouthillier, D., Moss, E., Rousseau, C., & Brunet, A. (2010). Emo-tion regulation strategies as mediators of the association between level ofattachment security and PTSD symptoms following trauma in adulthood.Anxiety, Stress and Coping, 23, 101–118. doi:10.1080/10615800802638279

Blake, D. D., Weathers, F. W., Nagy, L. M., Kaloupek, D. G., Charney, D. S.,& Keane, T. M. (1996). Clinician-administered PTSD scale for DSM–IV:Revision of measure described in (1995) The development of a clinician-administered PTSD scale. Journal of Traumatic Stress, 8, 75–90.

Blatt, S. J., & Levy, K. N. (2003). Attachment theory, psychoanalysis, personal-ity development, and psychopathology. Psychoanalytic Inquiry, 23, 102–150.

Bowlby, J. (1973). Attachment and loss: Vol. 2. Separation: Anxiety and anger.New York, NY: Basic Books.

Bowlby, J. (1980). Attachment and loss: Vol. 3. Loss: Sadness and depression.New York, NY: Basic Books.

Bowlby, J. (1982). Attachment and loss: Vol. 1. Attachment. New York, NY:Basic Books. (Original work published 1969)

Bowlby, J. (1988). A secure base. New York, NY: Basic Books.Bretherton, I. (2005). In pursuit of the internal working model construct and

its relevance to attachment relationships. In K. E. Grossmann, K. Gross-mann, & E. Waters (Eds.), Attachment from infancy to adulthood: The majorlongitudinal studies (pp. 13–47). New York, NY: Guilford.

Bretherton, I., & Munholland, K. A. (2008). Internal working models in at-tachment relationships. In J. Cassidy & P. R. Shaver (Eds.), Handbook ofattachment: Theory, research, and clinical applications (2nd ed., pp. 102–127). New York, NY: Guilford.

Bretherton, I., Ridgeway, D., & Cassidy, J. (1990). Assessing internal workingmodels of the attachment relationship: An attachment story completion taskfor 3-year-olds In M. T. Greenberg, D. Cicchetti, & E. M. Cummings (Eds.),Attachment in the preschool years (pp. 273–308). Chicago, IL: University ofChicago Press.

Bruhn, A. R. (1992). The Early Memories Procedure: A projective test ofautobiographical memory, part 1. Journal of Personality Assessment, 58,1–15.

Buchheim, A., & Benecke, C. (2007). Mimisch-affektives Verhalten bei Pa-tientinnen mit Angststorungen wahrend des Adult Attachment Interviews:Eine Pilotstudie [Affective facial behavior of patients with anxiety disordersduring the Adult Attachment Interview: A pilot study]. Psychotherapeuticand Psychological Medicine, 57, 1–5.

Buchheim, A., Erk, S., George, C., Kachele, H., Kircher, T., Martius, P., . . .

Walter, H. (2008). Neural correlates of attachment dysregulation in borderline

personality disorder using functional magnetic resonance imaging. PsychiatryResearch: Neuroimaging, 163, 223–235.

Buchheim, A., Erk, S., George, C., Kaechele, H., Ruchsow, M., Spitzer, M., . . .Walter, H. (2006). Measuring attachment representation in an fMRI environ-ment: A pilot study. Psychopathology, 39, 144–152.

Buchheim, A., & George, C. (2011). The representational, neurobiological andemotional foundation of attachment disorganization in borderline personalitydisorder and anxiety disorder. In J. Solomon & C. George (Eds.), Disorgani-zation of attachment and caregiving (pp. 343–382). New York, NY: Guilford.

Buchheim, A., George, C., & Kachele, H. (2008). “My dog is dying today”:Attachment narratives and psychoanalytic interpretation of an initial inter-view. In D. Diamond, S. J. Blatt, & J. D. Lichtenberg (Eds.), Attachment andsexuality (pp. 161–178). Hillsdale, NJ: Analytic Press.

Buchheim, A., George, C., Liebl, V., Moser, A., & Benecke, C. (2007). Mimis-che Affektivitat von Patientinnen mit einer Borderline-Personlichkeitsstorungwahrend des Adult Attachment Projective [Affective facial behavior of bor-derline patients during the Adult Attachment Projective]. Zeitschrift fur Psy-chosomatische Medizin und Psychotherapie, 53, 339–354.

Buchheim, A., Heinrichs, M., George, C., Pokorny, D., Koops, E., Henningsen,P., . . . Gundel, H. (2009). Oxytocin enhances the experience of attachmentsecurity. Psychoneuroendochronology, 34, 1417–1422.

Carlson, E. A. (1998). A prospective longitudinal study of attachmentdisorga-nization/disorientation. Child Development, 69, 1107–1129.

Cassidy, J., & Kobak, R. R. (1998). Avoidance and its relation to other defen-sive processes. In J. Belsky & T. Nezworski (Eds.), Clinical implications ofattachment (pp. 300–323). Hillsdale, NJ: Erlbaum.

Cassidy, J., & Marvin, R. S., with the MacArthur Attachment Working Group.(1987–1992). Attachment organization in preschool children: Coding guide-lines. Unpublished manuscript, University of Virginia, Charlottesville.

Cassidy, J., & Shaver, P. R. (Eds.). (2008). Handbook of attachment: Theory,research and clinical applications. New York, NY: Guilford.

Coan, J. A. (2008). Toward a neuroscience of attachment. In J. Cassidy & P.R. Shaver (Eds.), Handbook of attachment: Theory, research, and clinicalapplications (2nd ed., pp. 241–265). New York, NY: Guilford.

Crowell, J., Fraley, R. C., & Shaver, P. R. (2008). Measurement of individualdifferences in adolescent and adult attachment. In J. Cassidy & P. R. Shaver(Eds.), Handbook of attachment: Theory, research and clinical application(2nd ed., pp. 599–634). New York, NY: Guilford.

de Haas, M. A., Bakermans-Kranenburg, M. J., & van IJzendoorn, M. H. (1994).The Adult Attachment Interview and questionnaires for attachment style,temperament, and memories of parental behavior. The Journal of GeneticPsychology: Research and Theory on Human Development, 155, 471–486.

Diamond, D. (2004). Attachment disorganization: The reunion of attachmenttheory and psychoanalysis. Psychoanalytic Psychology, 21, 276–299.

Finn, S. E. (2007). In our client’s shoes: Theory and techniques of therapeuticassessment. New York, NY: Psychology Press.

Finn, S. E. (2011/this issue). Use of the Adult Attachment Projective PictureSystem (AAP) in the middle of a long-term psychotherapy. Journal of Per-sonality Assessment, 93, 427–433.

Fraedrich, E. M., Lakatos, K., & Spangler, G. (2010). Brain activity duringemotion perception: The role of attachment representation. Attachment andHuman Development, 12, 231–248. doi:10.1080/14616731003759724

George, C., Kaplan, N., & Main, M. (1984/1985/1996). The Adult AttachmentInterview. Berkeley: University of California, Berkeley.

George, C., & Solomon, J. (1989). Internal working models of caregiving andsecurity of attachment at age six. Infant Mental Health Journal, 10, 222–237.

George, C., & Solomon, J. (2008). The caregiving system: A behavioral systemsapproach to parenting. In J. Cassidy & P. R. Shaver (Eds.), Handbook of at-tachment: Theory, research, and clinical applications (2nd ed., pp. 833–856).New York, NY: Guilford.

George, C., & West, M. (2001). The development and preliminary validationof a new measure of adult attachment: The Adult Attachment Projective.Attachment and Human Development, 3, 30–61.

George, C., & West, M. (in press). The Adult Attachment Projective PictureSystem. New York, NY: Guilford.

George, C., West, M., & Pettem, O. (1999). The Adult Attachment Projec-tive: Disorganization of adult attachment at the level of representation. In

Dow

nloa

ded

by [

Car

ol G

eorg

e] a

t 09:

45 2

5 A

ugus

t 201

1

416 GEORGE AND WEST

J. Solomon & C. George (Eds.), Attachment disorganization (pp. 462–507).New York, NY: Guilford.

Hazan, C., & Shaver, P. R. (1987). Romantic love conceptualized as an attach-ment process. Journal of Personality and Social Psychology, 52, 511–524.

Hesse, E. (2008). The Adult Attachment Interview: Protocol, methods of analy-sis, and empirical studies. In J. Cassidy & P. R. Shaver (Eds.), Handbook of at-tachment: Theory, research, and clinical applications (2nd ed., pp. 552–598).New York, NY: Guilford.

Hesse, E., & Main, M. (2006). Frightened, threatening, and dissociative parentalbehavior in low-risk samples: Description, discussion, and interpretations.Development and Psychopathology, 18, 309–343.

Isaacs, M., George, C., & Marvin, R. S. (2009). Utilizing attachment measuresin custody evaluations: Incremental validity. Journal of Child Custody, 6,139–162.

Kissgen, R., Krischer, M., Kummetat, V., Spiess, R., Schleiffer, R., &Sevecke, K. (2009). Attachment representation in mothers of childrenwith attention deficit hyperactivity disorder. Psychopathology, 42, 201–208.doi:10.1159/000209333

Klagsbrun, M., & Bowlby, J. (1976). Responses to separation from parents: Aclinical test for young children. British Journal of Projective Psychology, 21,7–21.

Liotti, G. (2004). Trauma, dissociation, and disorganized attachment: Threestrands of a single braid. Psychotherapy: Theory, Research, Practice, Train-ing, 41, 472–486.

Lis, A., Mazzeschi, C., Di Riso, D., & Salcuni, S. (2011/this issue). Attachment,assessment, and psychological intervention: A case study of anorexia. Journalof Personality Assessment, 93, 434–444.

Lyons-Ruth, K., & Jacobvitz, D. (2008). Attachment disorganization: Unre-solved loss, relational violence, and lapses in behavioral and attentionalstrategies. In J. Cassidy & P. R. Shaver (Eds.), Handbook of attachment:Theory, research, and clinical applications (2nd ed., pp. 666–697). NewYork, NY: Guilford.

Main, M., & Cassidy, J. (1988). Categories of response to reunion with theparent at age 6: Predictable from infant attachment classifications and stableover a 1-month period. Developmental Psychology, 24, 1–12.

Main, M., & Goldwyn, R. (1985–1994). Adult attachment scoring and classifi-cation system. Unpublished manuscript, University of California, Berkeley.

Main, M., Goldwyn, R., & Hesse, E. (2003). Adult attachment scoring andclassification system: Version 7.2. Unpublished manuscript, University ofCalifornia, Berkeley.

Main, M., & Hesse, E. (1990). Parents’ unresolved traumatic experiences are re-lated to infant disorganized attachment status: Is frightened and/or frighteningparental behavior the linking mechanism? In M. T. Greenberg, D. Cicchetti,& E. M. Cummings (Eds.), Attachment in the preschool years (pp. 161–182).Chicago, IL: University of Chicago Press.

Main, M., & Solomon, J. (1990). Procedures for identifying infants as disorga-nized/ disoriented during the Ainsworth Strange Situation. In M. T. Green-berg, D. Cicchetti, & E. M. Cummings (Eds.), Attachment in the preschoolyears (pp. 121–160). Chicago, IL: University of Chicago Press.

Marvin, R. S., & Britner, P. (2008). Normative development: The ontogeny ofattachment. In J. Cassidy & P. R. Shaver (Eds.), Handbook of attachment:Theory, research, and clinical applications (2nd ed., pp. 269–294). New York,NY: Guilford.

Millon, T. (1969). Modern psychopathology: A biosocial approach to maladap-tive learning and functioning. Philadelphia, PA: W. B. Saunders.

Murray, H. A. (1943). Thematic Apperception Test manual. Cambrige, MA:Harvard University Press.

Paulhaus, D. L. (1998). The Balanced Inventory of Desirable Responding.Toronto, ON, Canada: Multi-Health Systems.

Solomon, J., & George, C. (2011a). The disorganized attachment-caregivingsystem: Dysregulation of adaptive processes at multiple levels. In J. Solomon& C. George (Eds.), Disorganized attachment and caregiving (pp. 25–51).New York, NY: Guilford.

Solomon, J., & George, C. (2011b). Dysregulation of maternal caregiving acrosstwo generations. In J. Solomon & C. George (Eds.), Disorganization of at-tachment and caregiving (pp. 3–24). New York, NY: Guilford.

Solomon, J., George, C., & De Jong, A. (1995). Children classified as con-trolling at age six: Evidence of disorganized representational strategies andaggression at home and at school. Development and Psychopathology, 7, 447–463.

Sroufe, L. A., Egeland, B., Carlson, E. A., & Collins, A. W. (2005). The devel-opment of the person. New York, NY: Guilford.

Subic-Wrana, C., Beetz, A., Langenbach, M., Paulussen, M., & Beutel, M.(2007). Connections between unresolved attachment trauma and retrospec-tively remembered childhood traumatisation in psychosomatic inpatients.Journal of Psychosomatic Research, 61, 399.

Szajnberg, N., & George, C. (2011). Attachment in Ethopian immigrants livingin Israel: A three-year study of children and their parents. Manuscript inpreparation.

Van Ecke, Y. (2005). Immigration from an attachment perspective. Social Be-havior and Personality, 33, 467–476.

Van Ecke, Y. (2006). Unresolved attachment among immigrants: An analysisusing the Adult Attachment Projective. Journal of Genetic Psychology, 167,433–442.

Wallis, P., & Steele, H. (2001). Attachment representations in adolescent: Fur-ther evidence from psychiatric residential settings. Attachment and HumanDevelopment, 3, 259–268.

Waters, E., & Deane, K. E. (1985). Defining and assessing individual differencesin attachment relationships: Q-methodology and the organization of behaviorin infancy and early childhood. Monographs of the Society for Research inChild Development, 50 (1–2, Serial No. 209), 41–65.

Waters, E., & Hamilton, C. E. (2000). The stability of attachment securityfrom infancy to adolescence and early adulthood: General introduction. ChildDevelopment, 71, 678.

Webster, L., & Hackett, R. K. (2007). A comparison of unresolved and resolvedstatus and its relationship to behavior in maltreated adolescents. School Psy-chology International, 28, 265–278.

Webster, L., Hackett, R. K., & Joubert, D. (2009). The association of unresolvedattachment status and cognitive processes in maltreated teens. Child AbuseReview, 18, 6–23.

Webster, L., & Joubert, D. (2011/this issue). Use of the Adult AttachmentProjective Picture System in an assessment of an adolescent in foster care.Journal of Personality Assessment, 93, 417–426.

Webster, L., & Knoteck, S. (2007). Attachment in high riskpopulations—Implications and practice for schools. School Psychology In-ternational, 28, 259–263.

Wechsler, D. (1981). The Wechsler Adult Intelligence Scale–Revised. New York,NY: Psychological Corporation.

Wechsler, D., & Matarazzo, J. D. (1972). Wechsler’s measurement and appraisalof adult intelligence. New York, NY: Oxford University Press.

Weiss, R. S. (1982). Attachment in adult life. In C. M. Parker & J. Stevenson-Hinde (Eds.), The place of attachment in human behavior (pp. 171–184).New York, NY: Basic Books.

West, M., & George, C. (2002). Attachment and dysthymia: The contributionsof preoccupied attachment and agency of self to depression in women. At-tachment and Human Development, 4, 278–293.

West, M., & Sheldon-Keller, A. E. (1994). Patterns of relating: An adult attach-ment perspective. New York, NY: Guilford.

Zegers, M. A. M., Schuengel, C., van IJzendoorn, M. H., & Janssens, J.M. A. M. (2008). Attachment and problem behavior of adolescents dur-ing residential treatment. Attachment and Human Development, 10, 91–103.doi:10.1080/14616730701868621

Zeifman, D., & Hazan, C. (2008). Pair bonds as attachment: Reevaluating theevidence. In J. Cassidy & P. R. Shaver (Eds.), Handbook of attachment:Theory, research and clinical applications (2nd ed., pp. 419–435). New York,NY: Guilford.

Dow

nloa

ded

by [

Car

ol G

eorg

e] a

t 09:

45 2

5 A

ugus

t 201

1