Bowen theory contains many concepts which overlap with concepts from attachment theory and the question of their relationship is a logical one. For example, Bowen was known to use the term “unresolved symbiotic attachment to the mother” in his writing, which opens questions about whether or not Bowen theory accounts for the phenomenon described by attachment theorists. This section will review a few studies which compare psychometric instruments from both theoretical systems.
Skowron & Dendy (2004) claim to have been the first to compare attachment theory with concepts from a systemic family theory. While their 2004 study did not specifically seek to account for variations in attachment presentations, they did study the convergence of differentiation of self (DoS) measured using the DSI-R and attachment style (AS) measured by the Experiences in Close Relationships Scale (ECR) (Brennan, Clark, & Shaver, 1998), correlating with a third concept effortful control as defined by the Effortful Control Scale (ATQ-S-EC) (Rothbart, Ahadi, & Evans, 2000). Skowron & Dendy (2004) summarized effortful control as the ability to “suppress reactive tendencies, modulate emotion feeling, and engage in purposeful behavior” (p. 338). This definition that is similar to Kerr’s (1988) definition of DoS, which reads “Increasing one’s ability to distinguish between thinking and feeling within self and others and learning to use that ability to direct one’s life and solve problems” (pp. 98).
Testing for theoretical convergence between DoS and AS is perhaps a logical choice for a first study toward understanding the place of attachment theory in a broader social context, as DoS is an individual as well as a systems concept (Skowron & Dendy, 2004). The generic scope of DoS suggests applicability beyond the child-caregiver dyad to all human relationships, and indeed beyond the domain of human functioning into other areas of life through the parent biological concept of differentiation. However, for the purpose of this review one may provisionally equate DoS to the level of security in an individual’s attachment style.
Skowron and Dendy (2004) found that there is considerable convergence between DoS and AS, which is in line with the expectations of this review. A significant link was found between measures of DoS and AS, with emotional reactivity (ER) correlating with anxious attachment, and emotional cutoff (EC) correlating with avoidant attachment, along with a link between DoS and effortful control. However, a link between AS and effortful control was not found, which is counter to claims that attachment security alone determines an individual’s ability to regulate affect. This suggests that either there are factors in the instruments used that do not account for all dimensions of the theoretical constructs of DoS and AS, or that there may be extraneous variables outside the caregiver-child dyad which affect one’s ability to auto-regulate in the midst of emotional intensity (Skowron & Dendy, 2004).
Subsequent studies found similar convergence with the anxiety and avoidance dimensions of attachment and dimensions of DoS using different measures of DoS (Ng & Smith, 2006; Ross, Hinshaw, & Murdock, 2016). However, Ng & Smith (2006) found no link between AS and non-dyadic measures of Personal Authority in the Family System (PAFS) (Bray, Williamson, & Malone, 1984), such as intergenerational intimidation (parents preventing children from psychological maturity) and intergenerational triangulation (parents conscripting children to help sort out their adult relationship problems). Based on Feeney’s (2003) finding that attachment style varies according to context, Ng & Smith (2006) suggest that attachment is just one variable among many in the larger context:
What these players bring to the relationship is not within the purview of the individual alone. … Hence, the attachment quality of an individual is only one among many variables in the equation. This might explain why intergenerational triangulation (i.e., triangulation involving parents) was not significantly associated with the attachment dimensions. (p. 437)
While early studies looked at convergence of attachment and systemic constructs, work toward a useful integration of attachment and systemic concepts has not occurred until recently. Ross et al (Ross, Hinshaw, & Murdock, 2016) compared experiential avoidance (EA) with DoS dimensions of triangulation, claiming triangulation accounted for the relationship between AS and DoS. Their findings suggest that DoS captures a more complete determinant of individual health outcomes. In similar fashion, Dallos, Lakus, Cahart & McKenzie (2016) take a direct look at how tension in the mother and father’s relationship impact the attachment response in the child. They aimed to fill a gap in the research of how “a mother’s ability to offer a secure attachment is influenced by the anxieties and tensions in her relationships with the child’s father” (p. 461). In this study, children’s responses to dyadic “attachment dilemmas” were compared to triadic attachment dilemmas to see if contextual variables influenced the type and severity across responses. Their findings suggest that “a child has an attachment, not just with each parent but with the relationship between them” (p. 461), and that triadic dilemmas “generated higher levels of attachment distress than the dyadic ones” (p. 466). Importantly, they found that children who would normally show secure attachment responses to dyadic dilemmas may show insecure attachment responses in triadic dilemmas. This not only indicates that the familial or social context influences the attachment response, but that the familial context may have a significantly larger impact on attachment responses than previously thought. Further, it was found triadic arousal persists significantly longer than dyadic arousal, indicating that the long-term effects of tension in the family may contribute more to complex trauma.
In an effort toward integration of individual and systemic attachment concepts, Mikulincer, Florian, Cowan, and Cowan (2002) propose a couples-based attachment security model that has each member in a couple as a separate sub-system of attachment characteristics: attachment security; positive models of self and others; relationship satisfaction; interaction of goals of togetherness; and positive models of self and others. Each of these characteristics interacts within a single person as well as with all characteristics in the other person, and the process multiplies exponentially with the addition of each new family member. Mikulincer, et al. (2002) write, “the quality of the relationship between the parents plays a central role in the generational transmission of working models of attachment” (p. 415).
This suggests that the quality of the relationship between two family members may temporarily affect the presenting attachment characteristics of their child. For example, if a mother and her sister argue, then the mother may react to the anxious situation by activating her attachment system with the child. In this example the presence of the sibling may take part in the construction of an avoidant attachment style in the child that matches the now anxious presentation in the mother.
Bowen (1959) observed in his early work with schizophrenia that fathers played an important role in the relationship between the mother and symptomatic child. Either parent was able to have a relationship with the child so long as the other parent would permit it, though the typical configuration was with the mother in a close relationship with the child. If the parents could manage to contain the anxiety in the marital relationship between them instead of focusing on the child’s symptoms, the child’s symptoms could abate for some time. Like Bowlby (1988) and Ainsworth (1985), Bowen did not assign caregiver roles according to gender. But unlike Bowlby and Ainsworth, Bowen concluded that the father’s role was in part to help support regulate the mother as she cares of the child as well as provide a modulating force between the mother and child. Bowen also mapped how all three in this father-mother-child triad were similarly influenced by relationships with the rest of the members of the emotional system with varying degree of intensity.
Discussion of Bowen theory and Attachment theory
While many aspects of AS and DoS appear closely related, they were developed in parallel by different schools of psychology and as a result classical literature offers little explanation of their relationship. For example, Bowen frequently cites problems in the family stemming from an “unresolved symbiotic attachment” (Kerr & Bowen, 1988, pp. 68, 110, 201, 220) with the mother, but does not differentiate his “attachment” from Bowbly’s “attachment.” Indeed, the question may have little relevance in a systems context which assumes the level of differentiation and integration of the parts as a complete measure of the negentropic capacity of the whole (Becvar & Becvar, 2018). Rothbaum, Rosen, Ujie, and Uchida (2002) eloquently describe similarities between attachment theory and family systems theory such as the forces of togetherness and individuality,
…attachment theory is focused on dynamics involving protection, care, and felt security, whereas family systems theory is concerned with family dynamics, involving structures, roles, communication patterns, boundaries, and power relations; (b) attachment theory is focused on the dyad, with much of the action occurring within individuals (e.g., “internal working models”), whereas family systems theory is focused on the triad, with much of the action occurring within groups; (c) attachment theory is relatively more concerned with children and development, whereas family systems theory is relatively more concerned with adults and current functioning; and (d) attachment theory has historically relied primarily on empirical research with normal populations, whereas family systems theory relies primarily on case studies involving clinic populations” (pp. 329).
It should be noted that though Bowen theory suggests a therapeutic approach that focuses on the highest-functioning members of the family such as the parents, the theoretical focus remains on the family as an emotional unit. The theories’ eight concepts apply equally, and also depend on, children as much as adults (Kerr & Bowen, 1988).
Nevertheless, the more recent studies in this review on attachment and systems concepts address this discrepancy as justification for an integration of attachment and systemic concepts. Ng & Smith (2006) compare attachment and PAFS as a) having to do with connection and intimacy, versus separation while maintaining intimacy, b) “hypothesizing continuity of relationship quality across generations,” c) “integrating intrapersonal and interpersonal aspects of human functioning,” and d) “recognizing the central place of emotion in the family and the life and the well-being of the individual” (p. 433). Both Ng & Smith (2006) and Ross et al (2016) take an optimistic position toward integrating attachment and systems concepts, and claim the considerable theoretical convergence found between DoS and AS as sufficient evidence to do so.
Nearly all studies reviewed here suggest that attachment theory alone may be inadequate in providing a comprehensive picture of human functioning. Ng & Smith (2006) suggest that AS may be contained within DoS as it only captures a subset of adult close relationship functioning and that a “subsystem within a larger family system may be affecting relationship functioning more than attachments” (p. 437). This raises questions about the obsolescence of AS within a systems context. Ng & Smith (2006) write,
…attachment theory may not live up to researchers’ claims to be an all encompassing or complete organizing theory of human functioning. It is most significant in demonstrating relationships among spousal relationships, relationships with children, and, to a lesser degree, relationships between adults and their parents. However, where attachment theory leaves off, intergenerational family systems complements and provides a larger picture. (p. 437)
Bowen (1978) claimed that DoS was more or less passed down through the generations through the multigenerational transmission process. Ross et al’s (2016) adds, findings support the idea of intergenerational transmission and offer AS and triangulation as accounting for it, suggesting that “dyadic relationship patterns originate within triadic processes and eventually affect the individual’s DoS” (p. 408). They add,
Although attachment theory accounts for dyadic relationships (between the primary caregiver and the child) and the dysfunction that can result when these relationships are not secure, the theory fails to acknowledge the role of a second primary caregiver or parent, and the potentially crucial aspect of disavowed negative emotional states resulting from them being deemed inappropriate by the caregiver is often considered secondary to the dyadic interaction. (2016, p. 401)
According to Kerr (1988) and Lassiter (2008), the triangle is only the human version of a universal mechanism emergent in every emotional system to ensure the survival of the family or social unit. Lassiter (2008) describes its vital role in social coordination in ideal circumstances as well as its role in the singling out of a scapegoat when the group experiences increased stress. She provides an analog in amoeba colonies which use the secretion of a pheromone in mating that serves to preempt an individual’s eventual self-sacrifice. These sacrificial individuals provide the “dead stalk” for the colony’s use in future generations (p. 69).
Such a generically applicable concept may therefore contribute to a more flexible and holistic view of human functioning that involves broader sets of variables from more levels of analysis. While failing to step fully into to the broad universal context of natural systems offered by Bowen, the studies in this review did succeed in demonstrating that triangulation may offer a mediating variable that connects the dyadic scope of AS with the greater system (Buehler & Welsh, 2009; Dallos, Lakus, Cahart, & McKenzie, 2016).
Unfortunately, none of the studies in this review address the problem of incompatibility between the scientific paradigms in which each theory is rooted, which suggest that the authors may not possess sufficient knowledge of the theoretical origins, and so also the a priori assumptions, of a system theory. Attachment theory, including other psychoanalytic-derived individual theories rely on a priori assumptions contained within a linear-causal paradigm that are more compatible with traditional random control trials. Bowen (1988) considers these individual theories because they were rooted in the psychology of the individual and the paradigmatic assumptions contained therein (Kerr, 1981). Systems frameworks such as Bowen theory rely on a priori assumptions contained within a mutual-causal paradigm which is incompatible with the random control trial model (Macy, 1991) unless isolated from their theoretical context. But as part of an interlocking natural system theory, it is not possible to remove these concepts from their parent theory. They were created specifically to overcome the limitations of the linear-causal paradigm (Kerr, 1981).
This paradigmatic incommensurability (Kuhn, 1962/2012; Noone, 2016) may in part explain why attachment theory enjoys extensive empirical research while systemic theories primarily rely on limited clinical case studies from within the same network of theoreticians (Rothbaum, Rosen, Ujiie, & Uchida, 2002; Dallos, Lakus, Cahart, & McKenzie, 2016; Noone, 2016). But it also points to a potential limitation in the foundation of attachment theory altogether, as one rooted in overly-simplistic assumptions about causality and the extent of the family system on influence. It is possible that this paradigm problem does not inhibit the type of theoretical convergence suggested by the studies reviewed here. But it is important to note the possibility of latent problems in attempts to integrate two concepts from different paradigms.
As this review on attachment theory and Bowen theory pokes at the idea of a natural system theory of the human family providing a more complete view of human functioning, it is required that the scope of such research align with a sufficiently broad context. The studies reviewed here limit their scope either to theoretical convergence of DoS and AT, or to qualitative observation of the mother-father-child triad, and so can only account for functioning within an idealized three-person nuclear family. In contrast, the development of the triangle as a pervasive biological concept suggests that triangulation in the nuclear family has an analog in the extended family, as well as in other higher and lower-ordered taxa pertaining to human life such as society, the organic, and the cellular. While these studies do not address this broad scope for the concept, future research could observe the interaction of triangling on attachment representations among siblings within the same nuclear family, aunts, uncles, and cousins of the children and/or parents, etc. However, the complexity of such an expanded scope raises issues which speak to the most important problem in comparing attachment and systemic theories related to their respective scientific paradigms. It is possible that performing direct comparisons between concepts “rooted in the psychology of the individual” (Kerr & Bowen, 1988, p. 7) such as attachment style, self-states, etc., and natural systems concepts such as differentiation, may be inherently problematic without one nomological taxonomy converting to the other (Kuhn, 2000).
There is no doubt from these and many other studies that DoS and AS attempt to solve similar problems and show some level of theoretical convergence. This does not mean, however, that overlapping dimensions of the constructs are synonymous, a conclusion supported by the evidence that AS may be contained within DoS and does not account for all outcomes such as effortful control, experiential avoidance, internalizing problems, and intergenerational dimensions of PAFS.
Ainsworth, M. (1985). Attachment over the lifespan. Bulletin of the New York Academy of Medicine, 61(9), 792-811.
Becvar, R. J., & Becvar, D. S. (2018). Systems theory and family therapy: A primer. Lanham, Maryland: Hamilton Books.
Bowen, M. (1978). Family therapy in clinical practice. New York, NY: Jason Aronson.
Bowen, M., Dysinger, R. H., & Basamania, B. (1959). The role of the father in families with a schizophrenic patient. The American Journal of Psychiatry, 11, 1017-1020.
Bowlby, J. (1988). A secure base: Parent-child attachment and healthy human development. New York, NY: Basic Books.
Bray, J. H., Williamson, D. S., & Malone, P. E. (1984). development of a questionnaire to measure personal authority in intergenerational family processes. Journal of Marital and Family Therapy, 10(2), 167-168.
Brennan, K. A., Clark, C. L., & Shaver, P. R. (1998). Self-report measurement of adult attachment: An integrative overview. In J. A. Simpson, & W. S. Rholes (Eds.), Attachment theory and close relationships (pp. 46–76). New York: Guilford Press.
Buehler, C., & Welsh, D. P. (2009). A process model of adolescents’ triangulation into parents’ marital conﬂict: The role of emotional reactivity. Journal of Family Psychology, 23(2), 167–180.
Dallos, R., Lakus, K., Cahart, M.-S., & McKenzie, R. (2016). Becoming invisible: The effect of triangulation on children’s well-being. Clinical Child Psychology and Psychiatry, 21(3), 461–476.
Feeney, J. A. (2003). The systemic nature of couple relationships: An attachment perspective. In R. Erdman, & T. Caffery (Eds.), Attachment and family systems: Conceptual, empirical, and therapeutic relatedness (pp. 139-163). New York: Brunner-Routledge.
Kerr, M. (1981). The later years. On Helping families change: family therapy through the life cycle. New York, NY: BMA Audio Cassette Publications.
Kerr, M., & Bowen, M. (1988). Family evaluation: The role of the family as an emotional unit that governs individual behavior and development. New York, NY: W. W. Norton & Company.
Kuhn, T. (2000). Afterwords. In T. Khun, The Road Since Structure (pp. 224-252).
Kuhn, T. S. (1962/2012). The structure of scientific revolutions. Chicago, IL: University of Chhicago Press.
Lassiter, L. (2008). The regulatory function of the triangle. In P. Titelman (Ed.), Triangles: Bowen Family Systems Theory Perspectives. United States of America: The Hayworth Press.
Macy, J. (1991). Buddhihsm and general systems theory: The dharma of natural systems.
Mikulincer, M., Florian, V., Cowan, P. A., & Cowan, C. P. (2002). Attachment security in couple relationships: A systemic model and its implications for family dynamics. Family Process, 41, 405–434.
Ng, K.-M., & Smith, S. D. (2006). The relationships between attachment theory and intergenerational family systems theory. The Family Journal: Counseling And Therapy For Couples And Families, 14(4), 430-440.
Noone, R. J. (2016). From the editor [Editorial]. Family Systems, 11(2), 99-103.
Ross, A. S., Hinshaw, A. B., & Murdock, N. L. (2016). Integrating the relational matrix: Attachment style, differentiation of self, triangulation, and experiential avoidance. Contemporary Family Therapy, 38, 400-411.
Rothbart, M. K., Ahadi, S. A., & Evans, D. E. (2000). Temperament and personality: Origins and outcomes. Journal of Personality and Social Psychology, 78, 122–135.
Rothbaum, F., Rosen, K., Ujiie, T., & Uchida, N. (2002). Family systems theory, attachment theory, and culture. Family Process, 41(3), 328-350.
Skowron, E. A., & Dendy, A. A. (2004). Differentiation of self and attachment in adulthood: Relational correlates of effortful self control. Contemporary Family Therapy, 26(3).