THE NATURE OF LOVE: UNDERSTANDING THE ENNEAGRAM TYPES AS NINE EXPRESSIONS OF ATTACHMENT

Abstract This study examined the validity of an integrated typology that combines the nine Enneagram types with the four attachment styles. Attachment styles are derived from research on attachment theory, a theory of human development that focuses on how infants and adults establish, monitor and repair close emotional relationships. Support was found for a conceptualization of the Enneagram types as attachment strategies that organize attention in order to facilitate the establishment, monitoring, and repair of close emotional relationships throughout life. The results are discussed in terms of implications for understanding the development and maintenance of the passions, and implications for further study of the developmental process of spiritual transformation of the types. The distinguishing characteristic of the theory of attachment that we have developed is that it is an ethological approach to personality development (Mary Ainsworth and John Bowlby, 1991). Over the past several years, there has been increasing interest in understanding Enneagram type development in the context of already existing theories of human development. Chestnut's (2008) article addressed this interest by integrating the Enneagram typology with three object relations theories of development. In Chestnut's integration, the developmental challenges children face at different stages of growth are matched with the core emotions of the three Enneagram centers (anger, fear, and sadness). Killen's (2009) article sketches a developmental model of how the basic mammalian emotions might give rise to the nine Enneagram types, suggesting that, ". . .the structure of each of the nine Enneagram types is built around a particular pattern of emotion regulation relating to innate mammalian emotion systems regulating fear, anger, and social distress" (p. 40). Taken together, Chestnut's and Killen's models of type development describe a process in which the basic mammalian emotions of fear, anger and social distress are transformed in the context of early relationships with caregivers into the nine distinct Enneagram passions. However, the details of this transformation have still not been articulated. In particular, the evolving model of type development lacks a satisfying explanation of how one particular emotion becomes fixed as the passion for each type. Understanding this transformation of emotion into passion would contribute to the project of developing a sound, scientifically based theory of development of the Enneagram types. Perhaps even more importantly, understanding the process of transformation of an emotion into a passion could contribute to an understanding of the undoing of the passions, and the accompanying transformation the type experiences in spiritual development. The present study seeks to add to the understanding of the development and the transformation of the types by focusing on an integration of the Enneagram types with attachment styles. The construct of attachment styles has developed out of research on attachment theory, which seeks to understand how and why humans establish, monitor, and repair close relationships (Siegel, 1999). In essence, attachment theory is the study of love. Mary Ainsworth, the author of the seminal research studies on attachment theory wrote: We are here concerned with nothing less than the nature of love and its origins in the attachment of a baby to his mother. According to this viewpoint, attachment originates in a few specific patterns of behavior, some of which are manifest at birth and some of which develop shortly afterward. Attachment is not present at birth, however; it emerges gradually through a course of development .... Attachment is manifested through these patterns of behavior but the patterns do not themselves constitute the attachment. Attachment is internal. We can conceive of attachment as somehow being built into the nervous system, in the course of and as a result of the infant's experience of his transactions with his mother and with other people. This internalized something that we call attachment has aspects of feelings, memories, wishes, expectancies, and intentions, all of which constitute an inner program acquired through experience and somehow built into a flexible yet retentive inner mechanism (which we identify with central nervous system functions) which serves as a kind of filter for the reception and interpretation of interpersonal experience and as a kind of template shaping the nature of outwardly observable response (Ainsworth, 1967, p. 429-430).

Ainsworth's description of attachment is striking because it could serve perfectly well as a generic description of Enneagram type. In addition, there are strong similarities among the prototype descriptions of the two models. In particular, the "dismissing avoidant" attachment style provides a very good description of the Enneagram Type Five; the description of the "preoccupied" attachment style is similar to the description of the Enneagram Type Two; and the "fearful avoidant" attachment style is similar to the description of the Enneagram Type Six (Table 1). Organization of Attention as a Common Factor We believe that the reason for these strong parallels between attachment styles and Enneagram types is that organization of attention is central to both models of development. In particular, Palmer's (1988, 1995; Palmer and Brown, 1997) conceptualization of the Enneagram types is based on the observation that each type has a distinct focus of attention. Palmer conceptualizes the types as developing around these foci of attention and corresponding blind spots early in childhood in order to regulate strong emotions such as fear, anger, and sadness that the child experiences in the context of relationships with parents and other caregivers. The focus of attention serves the purpose of filtering and shaping information in the context of these emotions, and ultimately of shaping subjective realities that keep a specific emotion fixed in place, in the form of a passion. Similarly, attachment theory focuses on individual differences in how humans manipulate attention in order to regulate emotion in the context of the attachment relationship (Bowlby, 1987; Fraley, Garner, and Shaver, 2000; Hadley, 1985; Main, 2000; Mikulincer and Shaver, 2003). In agreement with Killen's (2009) model of development, attachment theory begins with the evolutionary functional need for animals to regulate their emotions, ". . . the notion that emotions can be regulated as an adaptive means, a strategy to achieve one's goals in a given context" (Cassidy, 1994, ? 228). In agreement with Chestnut's (2008) model, attachment theory focuses on the central role of repeated transactions between parent and child in shaping the child's personality. However, attachment theory departs markedly from both of these models in that it abandons drive theory as a foundation for understanding child development. As Ainsworth stated: [The infant's] potential for attachment is part of the equipment of the newborn, although obviously attachment itself is only gradually acquired in the course of development - acquired in the course of interaction with his environment and the feedback he experiences as a consequence of his actions. His experience of having his needs for food and warmth met is important, but its significance is within the context of feedback. Fulfillment and gratification, as well as privation and frustration, are aspects of the feedback that the infant receives. Furthermore, it is obvious that the infant's survival is impossible if his basic requirements are not met at least minimally, and healthy development is hampered by too much privation or frustration. These things are true, and yet it is a gross oversimplification to focus on gratification and frustration while ignoring the role of organismenvironment interaction and the fact that the feedback refers to the consequence of the infant's own actions and not to need-gratification (Ainsworth, 1967, p. 438-439, italics added). What Ainsworth is emphasizing in this passage is that the developing child takes a fully active, engaged role in shaping the parent-child relationship in order to maximize the attainment of attachment-related goals (Main, 1999). In order to accomplish the attainment of these goals, the child develops attachment strategies that best serve this purpose (Bowlby, 1969/1982; 1973; 1980). This is an important distinction because it means that the developing personality is not simply a buffer against the pain of unmet needs in a relationship. Instead, the child develops a fixed, orderly, stable pattern of attention, unconsciously oriented toward the attachment figure, in which strong emotions are both regulated by, and also regulate, the attachment relationship (Schore, 2001). We are suggesting that it is within this stable pattern of attention, developed and fixed in place by the attachment relationship, that the basic mammalian emotions are experienced and become regulated in the form of the passions. Attachment Related Anxiety and Avoidance As noted above, attachment styles can be conceptualized as complex strategies for attaining attachment-related goals (Mikulincer and Shaver, 2005). These strategies are based on manipulation of attention along two orthogonal dimensions of anxiety and avoidance. The avoidance dimension is associated with a strategy used by a child who is overwhelmed, physically or emotionally, by the parent, or alternatively, by a child who is consistently ignored (Ainsworth, et al., 1978; Siegel, 1999). Use of this strategy involves "deactivation" of the attachment system (Magai,

1999). The logic of this strategy is that it allows the child to maintain at least minimal contact with the parent without being overwhelmed with emotion (Siegel, 1999). The term "deactivation" is somewhat of a misnomer because the attachment system is not deactivated, but rather is dampened, to allow for a minimal attachment connection. In contrast to the avoidant dimension, the anxiety dimension is associated with a strategy employed by a child whose parent is inconsistently responsive to the child's needs (Ainsworth, et al., 1978; Siegel, 1999), and use of this strategy involves "hyperactivation" of the attachment system (Magai, 1999). This strategy is typically used by a child when the parent is inconsistent with regard to meeting attachment needs. In response to this inconsistency in the parent, the anxious child will basically remain anxious all the time, constantly asking for attachment responses from the parent, whether anything is needed or not. The logic of this strategy is that if the system is hyperactivated all the time, there might be a chance of receiving attention when it is needed. Where the avoidant strategy is used by a child to maintain attachment contact without being overwhelmed emotionally or physically, the anxiety strategy is used by a child in an attempt to maximize attachment related attention at all times, in order to counteract the inconsistent responses of the parent (Siegel, 1999). Researchers have identified adult attachment styles that indicate that the original attachment strategies developed in childhood continue to operate across the lifespan, though they are malleable under certain conditions. There is compelling evidence that adults continue to manipulate attention in order to facilitate attachment relationships, and that this attentional manipulation operates at both the conscious and the unconscious level (Mikulincer and Shaver, 2007). In addition, there is some evidence, and growing speculation, that people generally have a preferred attachment style, but also access other secondary styles under certain conditions (Mikulincer and Shaver, 2003). This movement between a primary style and several secondary styles is of course reminiscent of the Enneagram type movement to stress and security positions (Palmer, 1995). Integration of Enneagram Types and Attachment Styles Typologies can be successfully integrated when they share an underlying structure of some kind, and when their integration results in the development of a richer, more complex, and more effective typology (Mandara, 2003). For example, Lienert, Reynolds, and Lehmacher (1990) integrated Eysenck's (1953) personality typology with Cattell's (1972) personality typology by focusing on the underlying structure that contributed to characteristics of extraversion in both typologies. Hypotheses about how two typologies fit together are derived using logic and existing knowledge about the characteristics of the two typologies that are to be integrated (Mandara, 2003). In the present study, the typological analysis brought together the characteristics of the attachment style categories, defined by anxiety and avoidance, with the characteristics of the Enneagram type categories, defined by focus of attention. Hypotheses Mikulincer and Shaver (2003) stress the importance of understanding that anxiety and avoidance are two independent subsystems of the overall attachment system. As independent subsystems, attachment and avoidance operate in parallel to each other, but can also have feedback effects in which the action of one system affects the other system. For example, a person might avoidantly withdraw from an emotionally charged attachment situation, but this withdrawal could in turn increase attachment related anxiety. Mikulincer and Shaver (2003) conceptualize the difference between anxiety and avoidance in terms of intensity and direction. The anxiety dimension is associated with variations in the intensity with which a person monitors and appraises attachment-related information for relevance and meaning. The avoidance dimension is associated with variations in the direction of movement (towards or away) from attachment-related information. This conceptualization of anxiety as a function of intensity and avoidance as a function of movement served as the foundation for the development of our hypotheses. Avoidance In the attachment typology, the avoidance dimension is associated with deactivation of the attachment system (Magai, 1999). People who are high on the avoidance dimension organize their attachment-related thoughts, emotions manipulate their attention in ways that serve to move attention away from attachment-related information (Shaver & Mikulincer, 2003).

Similarly, the nine Enneagram types are associated with specific attentional directions ("toward", "away from" and "falling asleep"), according to their emotional triad. Using Killen's (2009) terms, the "away from" movement is associated with the FEAR triad, the "towards" movement is associated with the PANIC triad, and the "falling asleep" movement is associated with the RAGE triad (Palmer, 1988). Translating these three attentional movements into the two categories of high and low avoidance led to the prediction that Enneagram types positioned on the left side of the symbol (associated with "moving away") would be higher in avoidance, while Enneagram types positioned on the right side of the symbol (associated with "moving towards") would be lower in avoidance. Special Case of Categorizing Type Nine In the Enneagram typology, Type Nine is a special case. Type Nine is located at the exact top of the Enneagram symbol. In order to integrate the Enneagram typology, comprised of nine types, with the attachment typology, comprised of four quadrants, created by the intersection of two dimensions, we had to make a decision about how to conceptualize Type Nine in terms of the attachment style dimensions. We were at a loss for a long time about how to do this. Finally, in frustration, one of the authors asked a Nine simply, "What do Nines want?" The answer was immediate and forceful: "Nines want to be noticed and not noticed. Nines want to speak up but not speak up. Nines want to confront but not confront." These words impressed on us the bifurcated nature of the Nine experience, and based on that, we decided to split Type Nine conceptually according to its wings, so that Type Nines with a One wing are categorized with the right side of the Enneagram symbol, and Type Nines with an Eight wing are categorized with the left side of the Enneagram symbol. Hypothesis One. Because the types on the right side of the Enneagram are assumed to have a "moving toward" bias and the types on the left side of the Enneagram are assumed to have a "moving away" bias, we hypothesized that Enneagram types 1, 2, 3, 4, and 9(1 wing) would be significantly lower in avoidance than Enneagram types 5, 6, 7, 8, and 9 (8 wing) (Figure One). Anxiety In the attachment style model, anxiety is associated with hyper-activation of the attachment system. Individuals who are high on the anxiety dimension intensely engage in mental, emotional, and physical monitoring and appraisal of information that is relevant to concerns about whether attachment figures are available, close, and responsive. Availability, closeness, and responsiveness can be monitored either in physical or psychological terms, or both (Mikulincer & Shaver, 2003). High Anxiety Types. Qualitative descriptions of the Enneagram types indicate that some types appear to be higher in attachment-related anxiety than others. Specifically, we identified Types Two, Four, Six, Eight, and Nine (wing Eight) as more likely to be high in anxiety than Types One, Three, Five, Seven, and Nine (wing One). Types Two and Four. On the right side of the Enneagram symbol (associated with "moving towards") both Type Two and Type Four are described as being highly interested in relationships, and appreciative of and talented at understanding emotional communication (Palmer, 1988, 1995). The world view of Type Two is, "People depend on my help; I am needed" (Palmer, 1995, p. 22). This world view is supported by a focus of attention on others' needs, and a corresponding lack of attention to the Two's own needs. This focus on others' needs involves intense monitoring of others, in order to discern what is needed (Palmer, 1988). The world view of Type Four is, "Something is missing; others have it; I have been abandoned" (Palmer, 1995, p. 47). This world view is supported by a focus of attention on what's missing in the present, and a corresponding lack of attention to what's good in the present. This pattern of attention in Type Four is also associated with hyp er- vigilance about possible abandonment (Palmer, 1995). Both of these attentional concerns suggest that Type Four would engage in high levels of monitoring and appraising of attachment-related information. Types Six and Eight. On the left side of the Enneagram symbol (associated with "moving away"), Types Six and Eight are both described as being concerned with issues of power and authority (Palmer, 1988, 1995). The key role of attachment figures as sources of authority was identified by Bowlby (1979) but has received almost no research attention since then. Similarly, Palmer (1988) briefly summarizes each Enneagram type's relationship with authority, but does not provide a discussion of why authority might play a central role in development of the types. While Bowlby and Palmer seem to agree that authority plays an important role in development generally, in our analysis, Type Six and Type Eight stand out as having particularly distinct relationships with attachment figures as authority figures.

The world view of Type Six is, "The world is a threatening place; question authority" (Palmer, 1995, p. 55). This world view is supported by a focus of attention on hazard and a corresponding blind spot regarding the actual magnitude of power of authorities. In response, Type Six intensely monitors others in order to protect against being harmed or left helpless by untrustworthy authority figures. This concern with authority in Type Six predicts especially intense monitoring of attachment figures that are also seen as authority figures. The world view of Type Eight is, "The world is an unjust place; I defend the innocent" (Palmer, 1995, p. 71). This world view is supported by a focus of attention on power and a corresponding blind spot regarding an Eight's own impact on others. Type Eight's intense concern with issues of protection and justice predicts heightened monitoring of attachment figures that appear to be either in need of protection, or alternatively, threatening to dominate the Eight. Low Anxiety Types. In contrast to Types Two, Four, Six, and Eight, prototype descriptions of Types One, Three, Five, and Seven indicate that intense appraisal and monitoring of others is not a key part of their attachment strategies. Types One and Three. On the right side of the Enneagram symbol, Type One and Type Three are both described as energetic, hard workers (Palmer, 1988, 1995). The world view of Type One, is, "The world is an imperfect place; I work towards improvement" (Palmer, 1995, p. 14). This world view is supported by a focus of attention on error, and a complementary lack of attention to shades of grey. Type One works hard to be perfect, and by extension, to be worthy of love from attachment figures. This strategy involves intense focus on work and tasks, as opposed to intense focus on the attachment figure. The world view of Type Three is, "The world values a champion; avoid failure at all costs" (Palmer, 1995, p. 31). This world view is supported by a focus of attention on success, and a corresponding exclusion of attention to issues involving failure. Type Three works very hard to be successful at whatever role he or she inhabits, and by extension, to be valued by attachment figures. This strategy involves intense monitoring of the self, in terms of presenting a successful image, but less monitoring of the attachment figure. Types Five and Seven. On the left side of the Enneagram symbol, Type Five and Type Seven are both described as appearing to be unperturbed by the external world. The world view of Type Five is, "The world is invasive; I need privacy to think and refuel my energies" (Palmer, 1995, p. 47). This world view is supported by a focus of attention on intrusion, and an exclusion of information regarding having enough. In withdrawing from the external world, Type Five reports a strong reliance on, and attachment to, a vivid, interesting, rewarding inner life (Palmer, 1988, 1995). This indicates that Type Five would not be expected to engage intensely in monitoring and appraising attachment figures, and so would be expected to be low in attachment anxiety. The world view for Type Seven is, "The world is full of opportunity and options; I look forward to the future" (Palmer, 1995, p. 63). This world view is supported by a focus of attention on pleasant future possibilities, and an exclusion of information regarding actual limitations. This approach involves intense appraisal and monitoring of pleasant future options, but less intense monitoring and appraisal of attachment figures. Type Nine. As with the avoidance dimension, Type Nine is split between a high anxiety categorization and a low anxiety categorization. The world view of Type Nine is, "My efforts won't matter; don't make waves; keep the peace" (Palmer, 1995, p. 79). This world view is supported by the Nine's focus of attention on other people's agendas, and a complementary exclusion of information about the Nine's own agenda. The analysis of Type Nine with regard to intensity of monitoring and appraising is that Nines who are more One-like (having a One wing) will be more concerned with being worthy of love, like a One, and thus engage in lower levels of monitoring of the attachment figure. Alternatively, Nines who are more Eight-like (having an Eight wing) will be more concerned with attachment figures as sources of authority, like an Eight, and thus engage in higher levels of monitoring of the attachment figure. Hypothesis Two: Based on the preceding analysis of the Enneagram types in terms of attachment strategies, we hypothesized that Enneagram types 1, 3, 5, 7, and 9 (wing 1 ) would be significantly lower in anxiety than Enneagram types 2, 4, 6, 8, and 9 (wing 8) (Figure Two). Research Instrument This study used the Experiences in Close Relationships-Revised (ECR-R) (Fraley, Waller, & Brennan, 2000) measure of adult attachment-related anxiety and avoidance. Internal consistency reliability for the ECR-R is reported to be .90 or higher for both subscales (Sibley & Liu, 2005). Examples of anxiety items are, "I worry that romantic partners won't care about me as much as I care about them," and, "I'm afraid that I will lose my partner's love." Examples of avoidance items are, "I find it difficult to

allow myself to depend on romantic partners," and, "It helps to turn to my romantic partner in times of need" (reverse scored). Enneagram Type We chose to use a purposeful sample (Patton, 2002) of people who had already received extensive training in how to self-identify their type. Subjects were people who had attended at least one week of Enneagram training with the Enneagram Professional Training Program (EPTP). The logic here is that the ability to observe one's own thoughts and emotions is assumed to be a universal capacity of humans, but it is more easily studied using a sample of people who have received training in how to do this. In addition, this sampling method has been used successfully by previous Enneagram researchers, all of whom were able to identify significant relationships between Enneagram types and other personality measures (Brown 8c Bartram, 2005; Palmer, 1988; Wagner 8c Walker, 1983). Data Collection Data were collected in person at two sessions attended by EPTP participants (Dayton, OH, May 12-15, 2007 and Colorado Springs, CO, July 5-8, 2007). Approximately 150 questionnaires were distributed, and 75 (approximately 50%) of the questionnaires were returned. Six questionnaires were discarded from the sample because of incomplete answers, for a total sample size of 69. Sample Characteristics The sample consisted of 14 Type Ones (19%); 6 Type Twos (8%); 5 Type Threes (7%); 8 Type Fours (11%); 2 Type Fives (3%); 6 Type Sixes (8%); 9 Type Sevens (12%); 5 Type Eights (7%) and 14 Type Nines (19%). Mean anxiety for the sample was 3.22 (s.d. = 99). The minimum anxiety score was 1.22, and the maximum anxiety score was 5.83. Mean avoidance for the sample was 3.02 (s.d. = .88). The minimum avoidance score was 1.14, and the maximum avoidance score was 5.35. Overall, the sample appears to have a reasonable distribution of Enneagram types, and to be similar to Fraley's (2008) very large sample with regard to distribution of anxiety and avoidance scores. Based on these observations, we concluded that the sample could be considered adequately representative of the nine Enneagram types, and of the general population in terms of anxiety and avoidance scores. Complete details regarding methods and data analysis for this study are available in Arthur (2008). Results The results of this study provide support for our hypotheses concerning the relationship between attachment styles and Enneagram types. Hypothesis One predicted that Enneagram Types One, Two, Three, Four, and Nine (wing One) (Group 1) would have a significantly lower mean avoidance than Enneagram Types Five, Six, Seven, Eight, and Nine (Wing Eight) (Group 2). This hypothesis was tested by comparing mean levels of avoidance for the two groups using a t-test for difference in means (Howell, 2002). Mean avoidance for Group 1 was 2.78 and for Group 2 was 3.35. As predicted, the t-test showed this difference was statistically significant (t = -2.67, p < .05). Hypothesis Two predicted that Enneagram Types One, Three, Five, Seven, and Nine (Wing One) (Group 3) would have a significantly lower mean anxiety than Enneagram Types Two, Four, Six, Eight, and Nine (Wing Eight) (Group 4). This hypothesis was tested by comparing mean levels of anxiety for the two groups using a t-test for difference in means (Howell, 2002). The mean level of anxiety for Group 3 was 2.88, and for Group 4, it was 3.63. As predicted, the t-test showed this difference was statistically significant (t = -3.35, p < .01). Discussion The integration of Enneagram types with attachment styles creates a typology which provides new information about important differences in attention patterns among the nine Enneagram types (Figure Three). Adding to already existing knowledge about the focus of attention of each type, the integrated typology provides information about how attention is organized in order to establish, monitor, and repair connections with attachment figures (Figure Three). Siegel (2009) defines "emotion" as a shift in one's state of integration. Negative emotions mark shifts to less integrated states, while positive emotions mark shifts to more integrated states. Transformation, according to Siegel, occurs when differentiated parts of a system become linked tiirough the process of integration (Siegel, 2010). In terms of the integrated Enneagram model presented in this paper, the passions can be understood as emotions that have become fixed in place by attachment strategies that develop in early childhood in order to ensure die attainment of

attachment-related goals, especially maintaining proximity to the parent. While the need for close physical proximity diminishes as children develop, die attachment strategy, including the passion, remains embedded in die neurobiologies substrate, in the form, we are suggesting, of the Enneagram types. This fixed passion prevents the individual from moving into states of greater integration with larger, more complex systems. Understanding the role of attention in maintaining a fixed orientation toward an attachment figure in type development and maintenance provides a useful starting point for exploring spiritual development of the types. In a sense, the types are developed and maintained through unconscious attention practices (Palmer, 1988). Attention practices that continually redirect attention back to die breath, as in mindfulness, or back to a specific mental object, as in some religious and psychotiierapeutic disciplines (Newman, 1996), can be understood within this integrated Enneagram attachment model as practices that facilitate a new kind of attachment relationship, a relationship with a greater reality, rather than a human figure. Conclusion If, as Ainsworth wrote over fifty years ago, the attachment relationship is the human vehicle for the expression of love, then within the integrated typology presented here, the nine Enneagram types are nine expressions ofthat love. Because of the evolutionary need for the attention patterns of attachment relationships to be fixed and orderly, these nine expressions of love remain fixed, undifferentiated, and unintegrated, and as such are experienced as passions within the human system. As expressions of love, attachment relationships are a source of unsurpassed joy in our lives. However, again because of the fixed and unchanging nature of tiiese emotional cognitive patterns, attachment relationships are also a source of untold sorrow and despair. Bringing attention to our attachment patterns means appreciating the joy these relationships bring us, and grieving the sorrow they bring as well. But through this process we can begin to prepare for the establishment of a new kind of attachment relationship, a relationship with the greater reality. We will take our first steps into this new relationship within the relationship patterns we have already developed and maintained from childhood into adulthood. The more we know about our own particular patterns of establishing, monitoring, and repairing our human relationships, the clearer our path should be as we begin to explore a relationship with the unknown.