Chapter 2 Theoretical Background In this chapter we lay the theoretical groundwork for under-standing the Dynamic-Maturational Model's (DMM) approach to the AAI in two ways. First, we review briefly the accretion of ideas included in the DMM method for analyzing discourse in the AAI. In addition, we differentiate the array of patterns used in the DMM method from those used in the Main and Goldwyn method. Second, we describe the conceptual organization of the ABC strategies. That is, in Chapter 1 we described the patterns whereas here we present their underlying structures. Ainsworth offered three,distinct categories (A, li. and C); in the DMM method,.these.are reconceptuahzed as reflecting two opposite psychological processes (Types A and C) and their cooccurrence (Types B and A/C). ATTACHMENT AND PATTERNS OF ATTACHMENT Understanding attachment in adulthood requires considerable understanding of attachment theory itself. Because much has been written about attachment theory already, only the essential rudiments are provided here in the form of a very brief discussion of the major contribu- Assessing Adult Attachment 30___ tions of Bowlby, Ainsworth, Crittenden, Main, and Fonagy to attachment theory, particularly as it is operationalized in the DMM approach to the Adult Attachment Interview (AAI). References to fuller discussion of these ideas are included. Bowlby Attachment theory, as introduced by Bowlby (1969/1982, 1973, 1980), is an organizational, systemic theory regarding the function and development of human protective behavior. Bowlby's theory integrated ethological, evolutionary, psychoanalytic, and cognitive theories. Attachment theory postulates that humans are innately predisposed to form attachment relationships to their primary caregivers, attachment relationships function to protect the attached person, and such relationships exist in an organized form by the end of the first year of life. The attachment relationship itself is defined as a tie, that endures across time and space, to a specific person to whom one turns when one feels vulnerable and in need of protection (Ainsworth, 1973). Bowlby presented considerable evidence indicating that separation from, or loss of, an attachment figure is associated with a variety of psychological and physical disorders, including anxiety, depressive disorders, and criminality (Bowlby, 1944,1958,1973,1980). He believed that such disorders were relatively stable but amenable to change through treatment (Bowlby, 1979). In his later work, Bowlby proposed the construct of internal representational models to explain how prior experience was retained over time and used to guide expectations and future behavior. He further suggested that there were multiple internal representational models tied to (a) different relationships and (b) different memory systems, such as semantic and episodic memory (Bowlby, 1980, Chapter 3). In the same chapter, Bowlby discussed mental integration of information held in different memory systems from the perspective of cognitive theory about information processing (Tulving, 1979). Bowlby also introduced the notion of developmental pathways. Such pathways were not trajectories, meaning that, once the direction of the pathway was initiated, it was not necessarily maintained throughout the life-span. On the contrary, the metaphor of pathways was used explicitly because it contained the notion of change points and intersections where one's direction could be modified in ways that were not necessarily predictable from the original path. This aspect of mental ^Theoretical Background 31 and behavioral organization is especially important because Bowlby was dedicated to the development of theory that would be clinically relevant to initiating change. The notion of pathways is relevant to the AAI classifications that are "earned" or "reorganizing," that is, changing from a former Type A or C strategy to a Type B strategy. , Possibly the most useful of Bowlby's contributions was his example of integration of the best of empirical science and other theories into attachment theory. By doing so, he created a state-of-the-art theory and modeled the means by which its relevance could be maintained. Ainsworth Ainsworth's primary contribution to attachment theory was the notion of individual differences in patterns of attachment (Ainsworth et al., 1978). Patterns of attachment reflect" individuals' unique expectations regarding the availability and responsivity of specific attachment figures to meet their need for protection. This notion was developed through Ainsworth's anthropological observation of infant-mother dyads in Uganda (Ainsworth, 1967), then replicated and expanded in her longitudinal research in the United States (Ainsworth et al., 1978). Her second and crucial contribution was to tie these outcome categories to differences in maternal home behavior during the 11 months preceding the administration of the Strange Situation. Without this empirical basis, it is unlikely that attachment theory would have been taken seriously. Instead, with an elegantly designed, ecologically valid data collection and careful, detailed data analysis of a short longitudinal study, Ainsworth put attachment front and center in development psychology. Ainsworth also introduced the idea that patterning, and not quantitative variation on multiple dimensions, was the basis of organization. Three major patterns of attachment in infancy were identified and empirically tied to their roots in maternal sensitive responsiveness: • Infants whose mothers were consistently and sensitively responsive to infants' attachment behavior were labeled Type B (Secure). • Infants whose mothers were predictably rejecting of attachment behavior were labeled Type A (Avoidant). • Infants whose mothers were inconsistently or insensitively responsive to attachment behavior were labeled Type C (Ambivalent). 32 Assessing Adult Attachment Theoretical Background 33 (I Early studies indicated that approximately two thirds of infants could be classified as securely attached (Type B) and almost one third as avoidantly attached (Type A) with a small percentage as ambivalently attached (Type C) (Ainsworth, 1979). Nevertheless, when the classificatory procedure was applied to infants who varied widely in socioeconomic status, culture, and quality of childrearing, the proportion of securely attached children was lower, sometimes very substantially lower. In addition, there were three expansions to the original Ainsworth system: Bell's identification of the B4 classification (Bell, 1970), Crittenden's identification of an A/C classification (19S5a, 1985b), and Main and Solomon's identification of a disorganized category (Main & Solomon, 1986,1990). The latter two expansions reduced even further the proportion of securely attached children, even in low-risk samples. Finally, the use of videotape permitted more precise and detailed observation of infant behavior. By permitting identification of fleeting behavior and subtle discrepancies among behaviors, this technological advance further reduced the proportion of children classified as Type B. Thus, a better estimate of security would be that no more than half of a nonrisk, middle-class, Anglo (specifically, American, Australian, Canadian, and English) sample would be expected to be classified as Type B (see van IJzendoorn, Goldberg, Kroonenberg, & Frenkel, 1992). The distributions for other cultures are less clear (see Crittenden, 2000b). For high-risk groups, the proportion of Type B children is substantially lower to nonexistent (cf. Pleshkova & Muhamedrahimov, 2010). Discussion of the use of terminology is relevant here. The terms secure and anxious (including both anxious-avoidant and anxious-ambivalent) tend to be used more often than the letters A, B, and C. Nevertheless, when Ainsworth first identified the three patterns, Bowl-by's advice was to give them simple letter identifiers until meanings of behaviors were understood (Ainsworth, personal communication, 1980). As work expanded to include an increasing range of cultures and subcultures, it became apparent that attachment nomenclature is sometimes perceived as offensive and pejorative. This is true of both the labels given by Ainsworth and those that Crittenden and others have applied. In addition, the terminology overlooks the contextual validity and adaptiveness of the non-B patterns. Put another way, the terminology may be prematurely evaluative. For these reasons, we use the letter notations frequently. Nevertheless, there remains a tension between the theory offered and the pattern labels. Hopefully, this discrepancy will function to instigate further research into the meaning of the patterns in varied contexts. To summarize, Ainsworth's central contribution to attachment in adulthood is the notion of individual differences in patterns of attachment, expressed as the ABC classificatory system that she identified among infant-mother dyads and tied to the infants' developmental experience. Main applied this to the AA1 in terms of both the classificatory system and the expected distributions of the classifications.1 Ainsworth's classificatory system is applied in the DMM with modification of both the array of classifications and the distributional expectations. It should be noted that Ainsworth ultimately concludedjthat an_ expanding, anay^f^rj^izedpattons better captured the essence of attachment than did the idea of disorganization, which she thought would only be a transitory state (Crittenden & Ainsworth, 1989). In addition, Ainsworth exemplified a model of theory development and exploratory research that is open-ended and anthropologically descriptive and that encourages expansion and modification of earlier work. Thus, in the terminology of attachment theory, both Bowlby and Ainsworth demonstrated use of representations that were open to new information and readily modified on the basis of such information. Crittenden Based on several developmental theories, current work in the cognitive neurosciences, and work with risk families (particularly maltreating families), Crittenden began proposing expansions to Ainsworth's ABC model. As she prepared her thesis under Ainsworth (Crittenden, 1981), she proposed new organizations. She also proposed an A/C pattern for endangered infants while preparing her dissertation, again under Ainsworth (Crittenden, 1985b). Later, A3-4 and C3-4 patterns were added for the preschool years (Crittenden, 1992) and A5-6 and C5-6 were proposed for the school years (Crittenden, 1994). With Ainsworth, she coauthored a chapter on the self-protective strategies of maltreated children (Crittenden & Ainsworth, 1989). They argued that disorganization, if it occurred, would be transient and replaced by context-ajiaj^ed^^ategies^including A/C, compulsive Type A (i^e., A3-6), and ~~ extremely anxious strategies. l.Main retained both B4 and the disorganized category, but not A/C, in her AAI classificatory system. Hesse later included A/C equivalents under the heading of "Cannot Classify" (Hesse, 1996). Assessing Adult Attachment 34___ Later, the array of expansions became known as the Dynamic-Maturational Model of attachment and adaptation. The DMM describes variation and change across the life-span in the development of attachment relationships. This perspective proposes that maturation is in dynamic interactionjyith experience, yieldirigThe j^ffltjjďTOTTawfuT_ £hangejnpatterns of attachment, that is, "reorganization" (Crittenden, 1994, 1995, 2008)._Changes7m_pju1S^ particularly frequent near_peripds..of -rapid neurological change (i.e., near transitions in developmental "stage"). Thus, Crittenden expected the possibility of changes in pattern of attachment at several points prior to adulthood. Further, Crittenden suggested that a very great majority of infants, especially those who experience dangerous circumstances, have organized strategies for relating to their attachment figures. Indeed, she argued that danger is central both to the evolution of attachment processes in our species and also to the organization of specific attachment relationships in each individual (Crittenden, 1997c, 1999b). Consequently, she considers information relevant to predicting danger and protecting oneself from it to be the basis of pattern of attachment. She uses this perspective because it focuses attention on the function of attachment; distortions in the way information is processed preserve the function under conditions of thz'eat. The distortions are of considerable clinical relevance; indeed, they can be considered the means by which risk for psychopathology develops and maladaptation is maintained (Crittenden, 1996, 2002). If this perspective is accurate, understanding the organization and function of these distortions may generate new approaches to diagnosis and treatment. If the distortions can be identified in the AAI, then the AAI may become an important clinical diagnostic tool. Because maltreated children experience both threats of danger and actual danger, they provide an exceptional opportunity to study the development of adaptation to danger. In Crittenden's studies of maltreated children, she identified one complex pattern in infancy that protected children who were both abused and neglected (i.e., A/C, Crittenden, 1985a, 1985b). Other researchers have associated this pattern with bipolar depression in mothers, that is, another source of variable danger (Radke-Yarrow, Cummings, Kuczynski, & Chapman, 1985). In addition, she hypothesized that patterns of attachment become more complex with development and has offered evidence of one new major pattern of organization, coercion (Type C), and several new patterns Theoretical Background 35 within Type A, including compulsive caregiving and compulsive compliance, that first develop in the preschool years (Crittenden, 1992). Thisj^rocesj^g^ of patterns, resulting from the interaj:tion_of experience with maturation, became the basisjrf a life-Ispajijnodeljrf.,^ for coping with danger-_ ous conditions (see Figure 1.1 in Chapter 1). Be^iiseThis model has been describedaFlength in many places (Crittenden, 1995,2000a, 2000b, 2008), it will not be described in more than a cursory manner here. It does, however, form the backbone of the DMM method for analyzing AAI transcripts. The Dynamic-Maturational Model represents a conceptualization of self-protective strategies based on evolved aspects of information processing. Specifically, innate aspects of organic processing (somatic information), temporal order of stimulation (cognition), and intensity of stimulation (affect) are identified as tine three most basic forms of information about whether, when, and where there might be danger. Reliance on cognitive information, to the relative exclusion of affect, is the basis for the Type A classifications, whereas reliance on affect, to the relative exclusion of cognition, is the basis for the Type C classifications. Type B is defined by flexible use and integration of both sources of information. Transformations of cognition and affect permit more precise identification of the most probable sources of danger as well as the organization of protective responses. Concurrently, however, the transformations distort information in ways that often lead to heightened expectation of danger and, thus, to the use °r,jjjfigrotectry^be-havior under safe circumstances. Change in the array of transformations reflects change in stFategyTn this model. Framed in this way, the pattern of attachment becomes a dimensional construct defined by a horizontal dimension of source of information and a vertical dimension of type of transformation of information (see Figure 1.1). Type B individuals, at the top of the model, integrate true affect and cognition whereas, at the bottom of the model, falsified and sometimes delusional information is integrated by psychopathic individuals while they exclude denied information from processing (Type AC). >. This model leads to at least two points of tension. First, it creates a •Siy£2ll!iS£M£Ili!2Si.as °PPose(i to categories) within all classifications (including within Type B). Consequently, declassifications become ap- ..EQ^teyations of unique individual patterns. Second^ in_this modeT~ every pattern carries both "adaptive and rr^dTaptLye aspects- Given this complexity, the verbal labels cannot reflect all the possi- Assessing Adult Attachment 36_ bilities within a pattern or even the primary function of the pattern and primary advantages and risks associated with it. Consequently, this model offers the advantages of being theoretically comprehensive and of differentiating a wider range of functioning than the original Ains-worth system and the early expansions of it. Nevertheless, it, too, leaves a range of differentiation (ultimately individual-specific, unique differentiation) unnamed. An important point regarding within-pattern variation is the pervasiveness of the strategy within the person's functioning. If one considers distance from the center of the model as indicative of greater penetration of the strategy into all aspects of an individual's functioning, then placement near the edges of the model should have implications for both greater risk of psychopathology and also more severe psychopathology. An important feature of the model is the inclusion of sexuality in adolescent and adult functioning. Protection and reproduction are proposed to constitute the two major organizing functions of human behavior (Crittenden, 1997a). The emphasis on danger harkens back to the earliest work in attachment theory, for example, Bowlby's study of 44 juvenile thieves (Bowlby, 1944) and his work with children displaced during World War II (Bowlby, 1951). Because reproduction does not motivate behavior until after puberty, distortions of sexuality are reflected only in the late developing patterns. Nevertheless, the central threat of failure to reproduce successfully is presumed to interact with aspects of attachment in ways that modify mental and behavioral organization. Further, even though the AA1 was not intended to address sexuality, in clinical populations speakers' responses frequently include references to sexuality. Therefore, distortions of sexuality are addressed in the coding method described here. Finally, Crittenden emphasized the notion that patterns of attachment reflect learned patterns of mentally managing cognitive and affective information so as to predict and adapt to dangerous circumstances and opportunities for reproduction (Crittenden, 2002). In particular, she focused on the notion that the brain transforms sensory stimulation into predictive information about dangerous and safe conditions and opportunities for sex. This idea is expanded in Chapter 4 in the discussion of transformations of information, representational models, and memory systems. With regard to the AAI, wej^ropose that adults' discoureejreflects mentalprocesses used both to focus on the mosF^áTíént andjneaning;_ Theoretical Background 37 ful predictors of danger and reproduction and to preclude awareness of information that increases danger and feelings of anxiety. Because maturation increases the range of mental and behavioral responses, the needfor ancVuse oj^self-protectiye organizations of thought and behavior may change with development, even when circumstances themselves are unchanging. This can result in a change in pathways as well as the^organizaticmofneT^trategj^sIZDie outcome of the changes can be observed in AAI discourse. These_changes involve a process of reof-ganiz