Resolving Alliance Ruptures from an Attachment-Informed Perspective

Abstract

In this article, we examine how the different attachment patterns enable or hinder the resolution of ruptures in the therapeutic alliance. We try to show that secure and insecure patients alike may experience ruptures in the therapeutic alliance, but that their ability to participate in resolving such ruptures differ markedly. Recent findings with the Patient Attachment Coding System (PACS) show that attachment classifications manifest in psychotherapy as distinct ways of communicating about present internal experience. Secure patients disclose their present experience openly and invite attunement from the therapist, while insecure patients either minimize their contributions to the dialogue (avoidant) or the contributions of the therapist (preoccupied). Using examples from session transcripts, we demonstrate how secure patients are particularly responsive to resolution strategies that focus on here-and-now experience, while insecure patients' characteristic ways of communicating pose significant challenges to the resolution process.

There is a growing consensus that working through conflicts, tension, and misunderstandings, sometimes referred to as alliance ruptures, is a key component of psychotherapy (Barber, Muran, McCarthy, & Keefe, 2013). Clinical wisdom and empirical research suggest that resolving alliance ruptures serves not only as a means of re-establishing a collaborative therapeutic alliance, but also as a mechanism of change in itself (see e.g., Safran, Muran, & Eubanks-Carter, 2011; Zetzel, 1956). A rich body of work over the past twenty-five years has focused on developing empirically-based techniques for recognizing and resolving alliance ruptures (see Safran & Kraus, 2014). In order for therapists to make optimal use of these techniques, it is critical that we understand the specific patient factors that influence the success or failure of the rupture-resolution process.

Attachment has often been invoked as one such patient factor (Lingiardi, Holmqvist, & Safran, 2016). Numerous studies conducted over the last two decades have found a positive association between secure attachment and stronger therapeutic alliances (Bernecker, Levy, & Ellison, 2014; Folke, Daniel, Poulsen, & Lunn, 2015; Mallinckrodt, & Jeon, 2015). Research has also found a relationship between the frequency of alliance ruptures and insecure attachment (Eames, & Roth, 2000). Far fewer studies, however, have examined how attachment impacts the process through which alliance ruptures are resolved. Because early attachments are the context in which we learn how to negotiate relationships and to express our emotions, attachment may affect the rupture resolution process in significant ways.

In this paper, we examine how the different patterns of attachment enable or hinder the negotiation of the therapeutic alliance and argue that secure, avoidant, and preoccupied¹ patients’ approaches to resolving ruptures differ radically. Although this paper is not an empirical study per se, the exploration we present here draws from recent empirical findings in attachment research. Recent research shows that attachment classifications are related to distinct ways in which patients communicate their present internal experience (Talia, Miller-Bottome, & Daniel, 2015). The ways in which secure patients openly express their internal experience make them particularly responsive to rupture-resolution strategies. Insecure patients, on the other hand, either do not disclose their experience of the rupture in present terms, or they communicate their experience in a confusing or exaggerated way. Both ways of communicating pose distinct challenges to the process of repairing ruptures. These challenges are by no means the only obstacles to rupture resolution. For instance, transference and countertransference configurations, patients’ particular emotion regulation tendencies, as well as ambiguous nonverbal communications and behaviors are all aspects that can affect the rupture-repair process. Yet, the particular ways in which attachment affects rupture resolution have rarely been explored in the literature, and we believe a greater knowledge of this variable would be of value to clinicians.

We begin with an overview of alliance ruptures: how they manifest and some different approaches to resolving them. We focus in particular on Safran, Muran and their colleagues’ work on the importance of exploring present experience in the process of resolving ruptures (Safran & Muran, 2000). We then provide an overview of attachment as it has typically been conceptualized in psychotherapy research and introduce the Patient Attachment Coding System (PACS; Talia & Miller-Bottome, 2015), the first observer-based measure of attachment developed for the clinical context. We proceed to describe the three patterns of expressing present internal experience that in the PACS are associated with secure, avoidant, and preoccupied attachment classifications. We then discuss, using examples from transcribed sessions, how these communication patterns can facilitate or impede the rupture resolution process and the clinical implications of these observations.

Therapeutic alliance, alliance ruptures, and the process of resolution

The alliance can be defined as the degree of collaboration regarding the tasks and goals of the therapy as well as the quality of the personal bond between patient and therapist (Bordin, 1979). In this view, alliance ruptures may result from a strain in any of these domains (Safran & Muran, 2000). For example, a patient may not want to discuss her relationships with her parents, she may disagree that the source of her problems is a lack of self-esteem, or she may be angry towards the therapist for not showing enough empathy or concern. In all of these cases, what defines a rupture in our view is an experience of some degree of discomfort or affective discord, although the extent to which this is present in the patient’s conscious awareness may vary greatly.

In general, a direct way for patients to resolve a rupture is to communicate their distress openly and without hesitation. For example, a patient who feels misunderstood or wounded by an off-the-mark interpretation from the therapist might say that she feels frustrated by the therapist’s comment. Another patient may openly express her reservations about a therapeutic task that she does not feel engaged in. In these scenarios, the patient initiates a requisite step towards repair by expressing to the therapist her experience of the rupture and conveying her present internal state. The therapist can then offer a reparative response, such as proposing a different task, clarifying a misunderstanding, or empathizing with the patient’s frustration.

Rather than openly conveying their experience of the rupture, many patients may resort to maneuvers that defend against the open expression of emotions or needs. Safran & Muran (2000) refer to these behaviors as rupture “markers”, as they often signal to the therapist that a strain in the therapeutic relationship has occurred, even without the patient saying so explicitly. Some patients may “move away” from the rupture through subtle disconnections from their internal experience, diminished engagement in the interaction, or overly compliant assurances (withdrawal rupture markers; Eubanks-Carter et al., 2014). Other patients may make critical or dismissive remarks that oppose either the therapist or the treatment (confrontation rupture markers; Eubanks-Carter, Safran & Muran, 2014). Following Harper (1989a and 1989b), Safran & Muran (2000) see withdrawal and confrontation as two broad categories of rupture behaviors that indicate that the patient perceives a problem in the relationship but does not feel safe communicating his or her feelings about it (see Table 1 for examples of withdrawal and confrontation rupture markers).

Table 1.

Examples of withdrawal and confrontation rupture markers (adapted from Eubanks-Carter, Muran, Safran, 2014)

	Rupture marker	Example
Withdrawal	(1) Denial: Patient denies a feeling state that is clearly evinced by the patient’s affect or non-verbal behaviors	T: You look upset…  P: No, I’m not, can we talk about something else?
	(2) Minimal response: Patient becomes silent or gives a minimal responses to the therapist’s questions or statements	T: That sounds like it was hard. How did it make you feel? P: (shrugs, avoids therapist’s gaze)
	(3) Content/affect split: Patient uses positive affect to soften or withdraw from a complaint or concern about the therapist or the therapy	T: It’s so hard for you to tell me that you’re really disappointed in here.  P: I mean, I’m fine (cries) I’m completely fine! (cries)
	(4) Overly compliant: Patient withdraws by submitting to the therapist in an overly compliant or deferential manner.	T: Something about our session wasn’t so helpful to you. P: I mean it’s not your fault! This is part of your process, you know better than me, I’m just a slow learner.
Confrontation	(5) Complaints about therapist as a person: Patient criticizes the therapist or expresses doubts about the therapist’s competence.	P: I thought about our session last week. Not so much what you said, actually, more the way you said them. You were pushing me into a corner. I wouldn’t have thought that was the way to go about helping people.
	(6) Complaints about therapy: Patient expresses dissatisfaction or doubt about the progress that can be made or has been made in therapy.	P: I’ve been coming here for weeks now and I haven’t seen any changes. I thought therapy was supposed to help, but I guess I was wrong.
	(7) Complaints about therapy activities	P: What does talking about my past have to do with my present? I don’t get what it has anything to do with my depression.
	(8) Controls / Defends against therapist: Patient responds in a hostile or defensive manner in response to perceived slights by the therapist.	P: I keep asking you for a direct answer and you won’t give it to me. Can you tell me why you asked the question or not?

The topic of how to work through impasses and negative process in the therapeutic relationship has been written about extensively (see e.g., Aron, 2006; Etchegoyen, 2005; Greenson, 1965; Kohut, 1984). This paper focuses on the work of Safran, Muran and colleagues and their empirically informed model for successfully resolving ruptures (Safran, Muran, & Samstag, 1994; Safran et al., 2011). According to Safran & Muran (2000), both patient and therapist must move through a series of “critical tasks” that focus on drawing out the patient’s present, internal experience. In doing so, the therapist pays close attention to the ways in which the patient might be avoiding or moving away from expressing their emotions or asserting themselves. First, the therapist calls the patient’s attention to the confrontation or withdrawal marker. Then, the therapist helps the patient move through any reticence or misgivings about expressing his or her feelings or needs related to the rupture, exploring underlying fears and wishes uncovered during this process. Although Safran and Muran demarcate two distinct strategies for dealing with withdrawal and confrontation rupture markers, both entail exploring patients’ untapped internal experience - namely, anger towards the therapist, feelings of sadness and vulnerability, or needs for agency and self-assertion.

An attachment-informed perspective on the therapeutic alliance

Although attachment theory has its origins in the study of early relationships between children and their parents, the theory expanded to become a rich framework for understanding close relationships throughout the lifespan (Cassidy & Shaver, 2016). Attachment research grew out of the pioneering classification by Ainsworth (Ainsworth, Blehar, Wall, & Waters, 1978) of three different patterns of “proximity-seeking” behavior in early childhood (secure, avoidant and ambivalent). The Adult Attachment Interview (AAI; George, Kaplan, & Main, 1996; Main, Goldwyn, & Hesse, 2002) was later developed to assess the sequelae of these early behavioral patterns in adulthood by examining individuals’ narratives about their early attachment history. In the AAI, individuals classified as secure provide coherent and vivid descriptions of their childhood experiences. Individuals with a dismissing classification (corresponding to the infant classification “avoidant”) narrate their experiences in an overly succinct and unemotional way, while those with a preoccupied classification (corresponding to the infant classification “ambivalent”) speak in a confusing, emotionally exaggerated or vague manner. These three attachment classifications have become a popular way for clinicians and researchers to assess and think about attachment in adulthood (Slade, 2008).

Bowlby (1988) famously postulated that attachment differences in adulthood might be especially apparent in the therapeutic relationship, as the treatment bond parallels an attachment relationship in many respects. This inspired a rich body of research using the AAI and other measures establishing an association between secure attachment and strong therapeutic alliances (see meta-analysis by Bernecker et al., 2013). To explain this association, many influential attachment-informed clinicians commonly invoke the hypothesis that secure patients, who have a good enough history of care with their primary attachment figures, experience their therapists as capable of providing care and comfort (see e.g., Daniel, 2014, 2006; Eagle, 2013; Holmes, 2001; Wallin, 2007). Along the same lines, it is hypothesized that insecure patients, whose caregiving experiences were less consistent, have negative expectations of the therapist’s availability and consequently relate to the therapist in maladaptive ways. In this perspective, patients of the same attachment category share a certain degree of comfort with closeness and differentiation in the therapeutic relationship and thus perceive and relate to the therapist in predictable ways.

From a contemporary relational perspective, there are limitations of this attachment informed view of the therapeutic relationship and its emphasis on attachment as certain types of transference-like perceptions and expectations. While adult attachment classifications have been shown to be relatively stable over time (see Hesse, 2016), clinical wisdom suggests that patients’ experience and view of the therapist and their relationship is far more variable and continuously evolving (Wachtel, 2005). Thus, equating transference-like experiences with attachment classifications may obscure the more granular aspects of the therapeutic interaction that clinicians and researchers are most interested in (Marmarosh, 2015; Obegi, & Berant, 2009; Slade, 2016). This is especially relevant when thinking about the bearing of patients’ attachment on the rupture resolution process. Given the pernicious effects of unresolved ruptures (Muran et al., 2009), it is not enough for therapists to know that secure attachment is associated with stronger alliances. Therapists must learn how to recognize and intervene on the specific challenges that insecure attachment brings to the negotiation of a collaborative therapeutic alliance.

The introduction of the PACS (Talia & Miller-Bottome, 2015) advances the study of attachment in psychotherapy by examining the ways in which attachment as a patient factor influences the therapeutic interaction. With the PACS, we can assess attachment by monitoring the patients’ moment-to-moment verbal communications with the therapist². The PACS was developed from the discovery of a number of in-session discourse characteristics that were shown to be statistically associated with patients’ pre-treatment AAI classifications (Talia, et al. 2014). The occurrence of these characteristics in a transcript leads to assigning one of three PACS attachment classifications (secure, avoidant and preoccupied) that have been validated in a large-scale study as independent and reliable indicators of patients’ three AAI classifications (N=156; .87, κ = .82). These findings are especially robust given that they were obtained from samples from three countries and in different therapeutic modalities, including both psychodynamic and cognitive-behavioral treatments (Talia et al., 2015). With the large-scale validation of the PACS, we can not only assess attachment based on patients’ in-session behavior; we can predict which sorts of in-session processes will be associated with the different attachment classifications.

The PACS shows that attachment classifications manifest in patients in psychotherapy as distinct ways of communicating about present internal experience. Secure patients convey their present experience openly and allow the therapist to participate in defining and elaborating on their experience. For example, secure patients disclose their emotions in the here and now and share vivid narratives of past experiences that clearly convey their feelings in the present. Secure patients also communicate their present intentions, autonomous reflections, and positive experiences. These speech acts, or markers, are rated on the PACS Proximity seeking, Contact Maintaining, and Exploring scales. Avoidant patients, on the other hand, tend to decline requests to express their here-and-now experience, or downplay the magnitude or importance of any experience that has been implied. These markers in the PACS are scored on the Avoidance scale. Finally, preoccupied patients share their experience in a one-sided, exaggerated, or confusing way that leaves little room for the therapist to respond, or actively disregard the therapist’s interventions entirely. Both patterns of communication, rated on the Resistance scale of the PACS, tend to limit the extent to which the therapist is able to participate in making meaning of the patients’ experience. (See Table 2 for some examples of PACS in-session attachment markers).

Table 2.

Selection of PACS in-session attachment markers.

Selected scales and markers	Examples of the markers used by patients during rupture resolution
Proximity seeking
(1) Asks for help	I mean I just - - - hope that you can help me (cries)
(5) Discloses distressful emotions experienced in the present	I’m feeling angry at you right now.
(6) Gives a vivid narrative of a past distressful event	Last session, at some point, I think we were talking about my relationship with my parents, and you made a comment that made me feel rejected.
(8) Discloses unmet relational needs	I need to talk openly and not feel judged. I feel like I’m not getting the support I really feel like I need.
Contact Maintaining
(11) Praises therapist’s ability to establish closeness	I’m glad I brought it up. Even though it’s hard, you’re listening, I can tell you really care.
Exploring
(16) Expresses independent will	I really want to be completely honest with you in here.
(17) Proposes tasks goals for therapy	Yeah, can we talk through a plan, ‘cause this is what I need, I think I need a plan.
(18) Expresses misgivings about the therapeutic tasks	The last couple weeks I felt like I wasn’t really getting to what I had been intending to talk about or the things that I thought were important to address here.
(24) Reflects in the present, assuming alternate perspective on self/other experience, beyond what is apparent	Even though we’ve finally established a space where you can give me real feedback on this, I feel like you’re wanting me to share right now and it is almost like - I’m pulling back.
Avoidance
(25) Fails to respond to inquiry about distress	T: What are your feelings right now?
	P: Well, it’s weird, I was just thinking about this situation…yeah, it’s a little strange.
(28) Dismisses offer of support	T: It seems like this is hard, talking about this issue we’re having. P: Yeah, maybe a little bit. But these things happen all the time.
(35) Attributes distress to an external cause	P: Yea, at times I’ve thought that I’m not sure therapy was useful, but I shouldn’t focus on the negative. I think it’s also the election and everything going on in the world right now that brings me down.
(36) Reject own complaints as pointless	P: It may be a little frustrating, but there’s no use in dwelling on it.
Resistance
(41) Does not respond to therapist’s support and changes to another distressing topic.	T: That was hurtful when I said that P: Most people just don’t really listen. The other day, my boyfriend yelled at me for forgetting the milk and it just hurt, y’know?
(44) Enlists other’s opinion to reinforce one’s own	P: I have a certain kind of depression that’s just inherited. My cousin the other day, she said you have a genetic depression, therapy won’t hurt you, but it’s not going to help you
(52) Changes topic abruptly	P: I thought that’s what therapy is for, to fix these things. Do you have any idea how to change this? The other thing is I told Marty at the store that we were out of supplies and then all of a sudden we had to go to four different stores to find what we were looking for.
(54) Speaks in a vague and confusing way	P: Given the choice between this therapy and what I had before … my dissatisfaction would be in that part of me was thinking this is the element that I said would happen all along. It of course nears my progress “in here” like here I am in therapy, sort of thing.

Note. Only the PACS markers relevant for resolving a rupture are presented in this table. The examples represent instances of the markers occurring in the context of a rupture, but each marker can be coded in reference to any topic. The numbers refer to the original numeration of the PACS markers (see Talia, et al., 2015)

PACS markers do not by themselves constitute ruptures in the alliance. The secure, avoidant, and preoccupied modalities of communicating about internal experience are pre-existing abilities that the patient “brings” to the therapy, given that they can be predicted on the basis of patients’ pre-treatment AAI and vice versa (Talia et al., 2015). These communication differences are not reactions to the therapist or the therapist’s remarks, nor do they emerge exclusively in response to discussing attachment-related experiences. The findings from Talia et al. (2014) and (2015) show that the in-session markers used by secure, avoidant, and preoccupied patients emerge in any given session, regardless of whether a rupture has occurred, and in the context of discussing a wide variety of topics.

In thinking about how these PACS communication differences might influence rupture interactions, the overlap between the in-session markers of secure attachment and the “critical tasks” of rupture resolution described in Safran and Muran’s (2000) model are readily apparent. According to Safran and Muran (2000), rupture repair process involves the therapist exploring patients’ emerging experience of the rupture and of the therapeutic relationship, focusing in particular on the patient’s experiences of anger, vulnerability, and needs for self-assertion. These are precisely the sorts of communications that secure patients engage in during psychotherapy. Secure patients facilitate repair by expressing whatever anger, shame, or anxiety have been evoked by the rupture and what they want to do about it. Avoidant patients tend to struggle to express any present, internal state, while preoccupied patients leave little room for the therapist to make sense of their experience. Because they present obstacles to a mutual exploration of patients’ present internal experience, the avoidant and preoccupied in-session communication markers can compromise rupture resolution in significant ways.

Avoidant and preoccupied patients’ communication markers present challenges to the therapy process that can differ from the challenges posed by rupture markers such as withdrawal and confrontation. The PACS avoidant and preoccupied communication markers are particular ways in which insecure patients communicate about their present, ongoing experience. When talking about feelings in the therapeutic relationship, for example, an avoidant patient may dampen or minimize her experience, while a preoccupied patient may convey her experience in confusing or exaggerated terms (see Table 2). In either case, the patient is communicating about their experience with the therapist, albeit in a problematic way. On the other hand, withdrawal and confrontation rupture markers are (in Safran and Muran’s framework) ways in which patients manage to not express their experience of a rupture by withdrawing from or controlling the interaction. For example, patients may fall silent, turn away from the therapist, or declare that they do not want to discuss the rupture further; other patients may become subtly hostile or commanding. These are ways in which patients respond when it does not feel safe to express their vulnerable or angry emotions towards the therapist, almost akin to fight or flight strategies. Secure, avoidant and preoccupied patients alike may resort to withdrawal and confrontation in the context of a rupture. However, while secure patients may initially engage in these strategies, with encouragement from the therapist, they remain more readily able to openly express their internal experience. Avoidant and preoccupied patients may still persist in communicating about their internal experience in non-collaborative ways even after any withdrawal or confrontation has been overcome.

In the following, we provide examples of how secure, avoidant, and preoccupied patients’ distinct ways of communicating come to bear upon the rupture resolution process. To illustrate our points, we include examples taken from transcribed sessions with secure, avoidant and preoccupied patients in which a rupture has occurred. The sessions were transcribed and the questionnaires were collected for previous research studies conducted out of the Brief Psychotherapy Research Program³. For all of the sessions from which the excerpts were taken, both patient and therapist reported on a post-session questionnaire (PSQ; Muran, Safran, Samstag, & Winston, 1992) the occurrence of a rupture – defined on the questionnaire with the statement “Did you experience any tension or problem, any misunderstanding, conflict or disagreement, in your relationship with your [therapist/patient] during the session?”. In each excerpt, we will point out where the PACS in-session communication markers occur, and discuss how these markers affect the therapist’s attempts to repair the rupture.

Secure patients

Our first excerpt was taken from a session with a patient named Rachel, 41 years old, in treatment for longstanding depression and classified as secure on the PACS. Rachel starts the session by telling the therapist about a phone call with her parents that occurred the previous week. The therapist attempts to probe for the patient’s thoughts and feelings during the phone conversation and Rachel struggles to recall exactly what those were. The therapist suggests that Rachel for the following week work on observing and maybe writing down her thoughts and feelings so that they have some information to work with in sessions. Rachel agrees, but begins to laugh nervously.

T: Why are you smiling?

P: No, I'm not smiling (the patient flips through notebook, noticeably smiling)

T: Is that a discomfort- is that an uncomfortable smile?

P: um, no uh, just the writing down thoughts - - I thought that you want me to do this because there’s something's wrong with me (wipes eyes).

In this interaction, the patient signals that a rupture has occurred through what can be recognized as a withdrawal marker (Table 1, marker 1). Her laughter, incongruent with her denial that anything is wrong, gives away her discomfort. The therapist attends to this marker (“Is that an uncomfortable smile?”) and attempts to explore the patient’s experience further. In response, Rachel discloses that she was hurt by the therapist (a Proximity seeking marker; Table 2, marker 6) and that she believed that the therapist suggested the writing exercise as a way to express a negative evaluation of her. The therapist continues:

T: That something's wrong- oh, so just now, a thought came up that the reason I want you to do this is that I think that there’s something wrong with you.

P: Yeah (laughs, wipes eyes)

T: And, and how did that make you feel? If we're using the thought?

P: Yeah. Um, (laughs, wipes eyes) anxious, nervous, upset.

T: mmhm. Okay. Well first of all thank you for be willing to share that with me. I’m not sure that a few sessions ago you would have, right? So that’s great. Second of all is, that’s not what I was thinking at all.

The therapist explores the rupture experience further by inquiring how this perception of the therapist made her feel. The patient responds with another prototypical PACS marker of in-session security and discloses her emotions in the here-and-now (“nervous, anxious, upset”; Proximity-Seeking Table 2, marker 5). Disclosing vulnerable feelings or the experience of being hurt in some way by the therapist is at once the open expression of the rupture and the very action that enables successful repair. These secure markers enable patient and therapist to work together to define the patient’s experience of the rupture so that the therapist can respond accordingly.

The next excerpt was taken from an article by Safran & Kraus (2014), which includes a transcription of a session with a patient named Ruth. In the paper, the therapist describes how he struggled to remain emotionally connected to Ruth, noting in particular her tendency to tell long and sometimes irrelevant-seeming stories. The passage from the rupture session begins with Ruth noting that termination was nearing and expressing doubt about how much progress she had made thus far (Table 1, marker 6). Acknowledging her tendency to become “sidetracked”, Ruth states that she wishes that the therapist could help her stay “on-topic.” Ruth admits that this felt uncomfortable for her to say. The therapist asks Ruth to elaborate:

T: Can you say any more about your discomfort?

P: Well...it’s like I’m being unreasonable and expecting too much...I have a tendency to blame myself when things aren’t going well in a relationship. And I don’t want to do that here.

T: Yeah. It’s not really fair for you to have to take all the blame if things don’t work out for you here. . . .

P: If I’m not to blame. I’m asking you to be really honest and tell me if I go off and start talking about a crack in the ceiling or whatever. Actually, as I’m saying that, I’m feeling stronger.

In this passage, Ruth conveys an active intentionality with regards to the therapeutic interaction by using two different markers from the PACS Exploring scale (Table 2, markers 16 and 17). She states her independent will in present terms (“I don’t want to…”) and suggests a therapeutic task for the two of them to engage in (“I’m asking you to…tell me if I go off…”). These markers have the effect of conveying a willingness to work collaboratively to address a tendency of hers that might otherwise obstruct her progress. The therapist seems to attune to these states as he affirms her expressed desire not to take the blame on herself. This is followed directly by another PACS Exploring marker (Table 2, marker 24) through which the patient reflects in the here and now on a feeling that arises as she is speaking. Ruth’s self-assertion and reflection on her experience as it emerges in the moment, both characteristic markers of secure attachment, set the rupture resolution process in motion.

As both Ruth and Rachel’s excerpts show, patients with a secure attachment classification can experience anxiety about closeness with the therapist and difficulty with aspects of the therapy process. This observation stands in contrast to the typical depiction of secure patients in clinical literature as being trusting and “secure” with their therapists (Wallin, 2007). The examples also show that therapists can also struggle to feel close to secure patients and may feel as though they at times avoid contact with the therapist. Furthermore, as we have highlighted elsewhere, although secure patients are capable of communicating a perceived strain in the bond with the therapist, they may also initially resort to confrontation and withdrawal. This is particularly notable in the session with Rachel, who at first appears to cover up her hurt feelings with nervous laughter and denial. While the signaling of the rupture itself may initially manifest through confrontation or withdrawal, a secure patient’s characteristic pattern of communicating facilitates repair in two major ways. In Rachel’s case, the patient expressed feelings of hurt stemming from some action or remark made by the therapist. In Ruth’s case, the patient initiated repair by negotiating alternate tasks and reflecting independently on the interaction. The patients’ communications of their present internal states invite the therapist to attune to the patient’s experience of the rupture and make efforts to address it.

Avoidant patients

The next excerpt was taken from a rupture session with Chuck, 38 years old, rated avoidant on the PACS and in treatment to address ongoing problems at work. During the fifth session of treatment, Chuck raises the concern that the therapist only “focuses on the past” in their sessions and that he has noticed feeling more depressed as a result (Table 1, rupture marker 7). The therapist attempts to explore the patient’s experience further:

T: But tell me a little more about what your experience is. Are you noticing some difficult emotions come up for you right now?

P: Hm, I guess um y’know some, little emotion - - I dunno - what about.

T: It can be painful, to start addressing and talking about this, I’m wondering if that is the feeling that you’re having?

P: it could possibly be, I can’t really put my finger on it. I have just noticed that outside this room - some days I just - some days I feel like I have a cloud over my head, I just feel more depressed. It’s probably just my - probably just my current circumstances. Y’know?

T: Yea, yea. So, you find yourself feeling depressed thinking about your life circumstances. I kind of heard some emotion in your voice when you said that.

P: No, no, you see, I don’t think I’m that emotional. I mean I don’t think I am really depressed for long periods of time, I think I just have you know thoughts, yea.

In this excerpt, the patient responds to the therapist’s empathic conjectures with overly general statements that downplay any disclosure of his present experience. One PACS marker of avoidance appears as the patient begins to describe in general terms his depressed mood (Table 2, marker 25). His comments about his experience are located in the past (“some days I just…”) and then attributed to an external cause (“my current circumstances”) (Table 2, marker 35). These markers shift the attention away from the more specific aspects of the patient’s present experience and partly disengage him from the repair process. In failing to specify further what emotion or self-state he is experiencing, the patient leaves the therapist with too much room to guess and probe. What we see in this exchange is prototypical of the avoidant communication pattern: an overly “zoomed-out” and generalizing discussion of personal experience that evades deeper understanding. The consequence of this pattern in the context of a rupture is that the repair process is stymied without the patient’s communication of his here-and-now experience.

Preoccupied patients

Our final excerpt is from a session with a 50 year-old patient named Francesca, who is in treatment to get help coping with marital issues. The rupture in this exchange begins with the patient expressing reluctance about directly expressing to the therapist her feelings towards her husband. The patient vaguely implies that she is experiencing “performance anxiety” about “doing it right” for the therapist (Table 1, rupture marker 7). The therapist tries to explore the patient’s experience of this discomfort:

T: So when you say performance anxiety, you are always performing in front of an audience. And in this case I’m the audience. So you want to do right by me.

P: Yes. I was amazed in St. Louis how little I spoke to people. People who are friends of Shane’s family or whatever. I said enough so that I registered my presence, but that’s about it I think. There’s no real freedom in being able to speak, no real freedom to be able to think of anything to say or whatever.

T: So, just to bring it back to this situation again, let’s take it one step further, there’s a sense that you don’t want to disappoint me (P: Yes.). That you’re really stuck.

P: That I’m stuck. I’ve been stuck in this so long. I was thinking this morning that this is an attempt to break out of whatever spot I’m in. But there are other things I need too. I had jokingly mentioned that, just to learn skills of presentation that might be superficial. But perhaps in conjunction with a deeper emotional restructuring, some cosmetic and superficial techniques might be beneficial as well. I took Bob to practice last night and was looking at the catalogue for the YMCA, which has courses for interviewing…

The therapist makes two attempts to explore how the patient’s current feelings and perceptions about the therapist may be connected to the patient’s so-called “performance anxiety”. In response, the patient shifts the discussion to an unrelated topic (Table 2, marker 52), resisting the therapist’s attempt to attune to the patient’s experience. While a patient using withdrawal markers may change topics as a way to evade rupture-related feelings, preoccupied patients show these shifts in topic habitually, regardless of the topic under discussion (Talia et al., 2015). In the second patient speech turn, the patient’s discourse devolves into another PACS Resistance marker: vague, wordy descriptions of thoughts and intentions (related to being “stuck” in a “situation”; Table 2, marker 54), which leave it unclear what the patient is actually thinking, feeling or referring to. This vague marker is not to be confused with the overly general responses of avoidant patients (e.g., “I dunno, I’m just depressed”). Vague speech such as this has the paradoxical effect of drawing the listener in while simultaneously preventing any further definition of or attunement to the patient’s experience. The changes of topic and overly vague articulations of the patient’s experience restrict what the therapist can contribute to the dialogue and leave the rupture unattended and unresolved.

Conclusions

In this paper, we explored the ways in which attachment affects patients’ capacity to engage in the rupture resolution process. On the basis of patients’ PACS attachment classifications, we examined excerpts from rupture session transcripts and identified different communication patterns that may be especially facilitative or problematic for the resolution of ruptures. We attempted to show that communicating present internal experience openly, as secure patients do, comports well with the tasks of resolving a rupture, while the ways in which avoidant patients and preoccupied patients minimize their own contributions or the contributions of the therapist, respectively, pose particular challenges to the process. Although attachment has long since been established as an important factor in the building of a strong therapeutic alliance, until now there has been no empirically based account of how attachment manifests in the therapeutic interaction. The PACS answers important questions about how attachment differences affect the process through which the alliance is negotiated, and reveals distinct process markers that therapists should be mindful of when attempting to repair ruptures with insecure patients.

It is important to distinguish the PACS attachment classifications from the idea of attachment “styles” (e.g. Mikulincer & Shaver, 2007; Mallinckrodt, 2010; Brennan, Clark & Shaver, 1998), a term used in an important tradition of social and personality research informed by attachment theory. Self-reported attachment styles refer to how people tend to experience significant relationships, including the therapeutic relationship. For example, this view posits that avoidant patients experience more discomfort with closeness in the therapeutic relationship whereas preoccupied patients tend to be overly dependent on the therapist. The PACS, on the other hand, examines how patients convey their internal experience in dialogue with the therapist. While self-report researchers focus on the nature and quality of patients’ interpersonal experience itself, our focus is on the particular ways in which interpersonal experience is communicated. From our perspective, these are two dimensions of the therapeutic process that may overlap at times, but are otherwise conceptually distinct. According to our analyses of sessions of hundreds of patients assessed with the AAI, we have observed that patients of the same attachment category may experience the therapeutic relationship in many different ways. For example, secure patients may struggle with insecurities and experience ruptures in the therapeutic relationship as much as their insecure counterparts. The research evidence, which has found a low correlation between patients’ self-report of their attachment tendencies in relationships and their AAI classifications, seems to corroborate these observations (Roisman et al., 2007).

Our discussion of the role of attachment in the therapeutic interaction bears a resemblance to the literature on infant attachment and mother-infant interactions. In this literature, there is consensus that ruptures (or moments of affective mis-coordination) are pervasive in all caregiver-infant interactions, and that what is most vital for the infant’s developing sense of agency, mastery, and indeed, secure attachment, is a history of successful repair with the caregiver (see e.g., Beebe, & Lachmann, 2013; Tronick, 1989). In this paper, we adopt a similar view and claim that secure and insecure patients alike experience ruptures in their relationship with the therapist. While the rupture itself, its precipitant and the emotional climate in which it plays out may be idiosyncratic and arise out of the unique characteristics and interactions of a particular patient-therapist dyad, the capacity for repair that the patient brings to the interaction is not. According to the recent findings with the PACS, while secure attachment does not necessarily correspond with a problem-free therapeutic relationship, it does relate to an ability to openly express present experience and thus enable the successful repair of ruptures.

Some clinical implications of the analysis presented in this paper warrant further discussion. There seem to be two challenges in repairing a rupture in the therapeutic alliance. The first is patients’ initial reluctance to disclose their painful experience in the relationship, as signaled by withdrawal and confrontation rupture markers. A second challenge to repairing ruptures, in the case of insecure patients, is their habitual mode of communicating present experience. From our perspective, these are two separate aspects of resolving a rupture and it is likely that they require different strategies to be worked through.

In pointing out these possible clinical implications, we would be remiss not to acknowledge some limitations of our paper. For one, as we mentioned in the paper’s introduction, our paper focuses on the verbal level of patients’ communication and thereby leaves out other equally important aspects of the therapeutic interaction, such as non-verbal behaviors, transference-countertransference dynamics, as well as the domains of defenses and emotion regulation capacities. Second, although our paper discusses findings from recent attachment research studies and presents clinical case examples that are assessed using an empirically validated research instrument, it is not an empirical study in itself. Thus, the conclusions drawn in this paper represent clinical considerations that should be further elaborated and verified with empirical studies.

As Safran & Muran (2000) have shown, withdrawal and confrontation can be ameliorated through the therapist “meta-communicating” or talking about the enactment occurring within the relationship through tentative self-disclosures and empathic conjectures (Safran & Kraus, 2014). Unlike withdrawal and confrontation, markers of insecure attachment do not seem to be motivated by a conscious or unconscious experience of a lack of intimacy or trust with the therapist. Rather, they are a part of a general pattern of communicating present experience in more or less open and collaborative ways. Thus, changing these communication patterns may occur more at the implicit, procedural level of the therapeutic relationship, perhaps through the gradual pursuit of “moments of meeting” between patient and therapist (Stern et al., 1989). Therapists may perhaps begin by initially adapting their communication approach to the patient’s and then gradually prompting for more open disclosures and receptivity to the therapist’s responses over time. Future studies may use the PACS to refine the existing rupture resolution strategies and to identify the therapeutic techniques that seem to shift the communication patterns of avoidant and preoccupied patients in a more secure direction.

In clinical practice, it is all too common to encounter a patient who seems difficult to reach and on whom rupture resolution strategies do not seem to make an impact. The PACS can help narrow our focus on the particular, momentary obstructions in communication that can stall the repair process and compromise the therapeutic connection. In this paper, we have examined how patients’ approaches to communicating about present experience with their therapist are distinctly related to their attachment classification and the ways in which these communication differences can have immediate and profound effects on the resolution of ruptures in the therapeutic relationship. We suggest that using the PACS to track patients’ communications and to assess the effects of different therapeutic interventions could lead to further refinements of rupture resolution techniques and offer a promising new direction for alliance and attachment-focused research.