Exploring Attachment Incoherence in Bereaved Families’ Therapy Narratives: An Attachment Theory-Informed Thematic Analysis

Abstract

Attachment theory predicts that family bereavement leads even securely attached individuals to experience temporary attachment insecurity. This article explores how incoherence, a narrative marker of attachment insecurity, is displayed in the talk of families undergoing bereavement family therapy. This study uses the lens of attachment theory, and specifically the Adult Attachment Interview (AAI), to explore how members of five families talked about the loss of a close family member and how interactions between therapist and family members could hinder a coherent dialogue about the death. The families were recruited through Winston’s Wish, a UK-based family bereavement charity. The analysis centers on the “Telling the story” intervention, used at the start of therapy, in which family members together tell the story of the death. Families also retold the story in an extra family session toward the end of the therapy for the purposes of this research. Transcripts of the therapy sessions were analyzed using thematic analysis, with some codes developed directly from the unresolved loss codes of the AAI and other codes generated from analyzing the transcripts through the lens of attachment theory. Through a micro focus on therapy process, the study provides tentative support for suppositions in attachment theory about the psychological importance of (fostering) coherent speech as well as information about potentially helpful versus unhelpful therapist actions in family bereavement therapy. The findings have relevance for bereavement therapy interventions, therapy training, and research practice.

Keywords

Adult Attachment Interview attachment insecurity process research therapeutic microprocesses unresolved loss

Background

Research suggests the death of a family member has a significant negative impact on both children and adults, including negative effects on physical and mental health, poorer educational and employment prospects, financial disadvantages, and secondary relational loss (Akerman & Statham, 2011; Dowdney, 2000). A focus on therapeutic interventions for family death is important because research suggests that between 10% and 20% of bereaved adults and children show evidence of “chronic” grief (Bonanno & Kaltmann, 2001; McClatchey, Vonk, Lee, & Bride, 2014). Empirical support for the efficacy of grief interventions is equivocal (Larson & Hoyt, 2007; Neimeyer & Currier, 2009), and there is also a lack of research focused on interventions for families (Hooghe, De Mol, Baerens, & Zech, 2013; Stroebe, 2010). However, research has also shown that narrative coherence is a valid marker of progress and the “effectiveness” of bereavement therapy for clients (Neimeyer, Holland, & Currier, 2008).

Attachment Theory and Loss

In attachment theory, the (in)coherence of speech is important because the organization of speech is understood to reflect the intrapsychic organization of a person’s attachment schemas. Main’s work with the Adult Attachment Interview (AAI), an interview-based assessment of attachment in adults in which participants are asked about their parental relationships and childhood experiences of loss, posits that coherence of speech is a key marker of attachment security (Main, Kaplan, & Cassidy, 1985). The empirical support for the AAI is strong, for example, the security of pregnant mothers as assessed by the AAI has been found to predict attachment status of their infants at age 1 (Benoit & Parker, 1994; Steele, Steele, & Fonagy, 1996).

(In)coherence of speech has also been found to be a key marker of a particular type of attachment insecurity that is related to loss. Attachment theory suggests that when people experience the death of an attachment figure such as a close family member, their relational cognitive schema—what Bowlby termed their internal working model (IWM)—for this relationship is disturbed (Bowlby, 1980). Bowlby argued that coming to terms with a loss normatively requires reorganizing this schema to accommodate the death and that this process takes time. Main’s coding system for loss in the AAI posits that the disorganization of attachment schemas following loss creates particular types of incoherence in talk about the deceased (Main et al., 1985). These types of disorganization are catalogued in the coding system for what is termed “unresolved loss” (“U”). The assumption in the coding manual is that a bereaved person’s narrative within a year of their loss would normatively display these markers of incoherence. Again, the empirical support for the U-coding system is strong, for example, unresolved loss has been found to predict the “disorganized” form of attachment insecurity in infants (Ainsworth & Eichberg, 1991; Main & Hesse, 1990).

The Current Study

The current study applies the AAI coding system, not to AAI interviews in which an adult participant talks about their childhood experiences of loss to the AAI interviewer but to extracts from family bereavement therapy, in which the story of the family member’s death is recounted by the family to the therapist. Using the AAI codes on non-AAI data is not entirely without precedent (Thomson, 2010); however, there are limited examples of the AAI codes being used outside of this context and specifically on therapy data (Muscetta, Dazzi, Decoro, Ortu, & Speranza, 1999; Thomson, 2010). Although some researchers would disagree that attachment patterns influence discourse outside of an AAI interview (e.g., Hughes, Hardy, & Kendrick, 2000), there is a small literature that suggests that attachment processes impact on narrative formation in other contexts (e.g., Bishop, Steadmon, & Dallos, 2015), which provides support for the use of AAI codes in the current study.

The current study adds to the limited literature on bereavement process research and aims to demonstrate the usefulness of the “U” loss codes for understanding incoherence in family therapy narratives. The Strange Situation Test (SST; Ainsworth & Wittig, 1969) and the AAI (George, Kaplan, & Main, 1985) deliberately both “stress” the attachment system (e.g., separating babies from their parents [SST] and asking adults to talk about things like separations from parents when they were children [AAI]). The therapy data in this study are also focused on a similar (very) stressful point—families retelling the story of the death of a family member—and therefore it can be assumed that it is at these moments that the researcher will be able to see the attachment system most clearly in operation. Utilizing the coding system of the AAI to examine the therapy data will make it possible to look for subtle linguistic markers of unresolved loss that attachment theory would predict should be present in the recent aftermath of bereavement.

It has been noted in the counseling and psychotherapy literature that there is little research based on directly analyzing what happens in therapy sessions (Finlay, 2014), and there have been calls for further process research based on therapy data (Henton, 2012; Mallinckrodt, 2011). Concurrent with a shift in assumptions about the “best” way to measure the efficacy of grief interventions, there have also been calls for researchers to adopt narrative qualitative methodologies to consider the bereavement experiences of young people and adults (Dowdney, 2000; Ribbens McCarthy, 2007) as well as to increase understanding of the operational implementation of interventions (Currier, Holland, & Neimeyer, 2007) and the critical mechanisms within interventions (Ahn & Wampold, 2001). Midgley (2004) correspondingly called for greater use of qualitative methodologies in child therapy process research. The current study answers such calls and asks the following research questions:

Is there evidence of incoherence (as described by the AAI) in the families’ narratives?

Are there other ways that incoherence is demonstrated or systematically enacted in the family sessions?

Method

Design

This study employed a longitudinal qualitative design using recordings of the “Telling the story” intervention at the beginning and toward the end of family bereavement therapy. This involved the recording of “naturalistic” therapeutic data.

Participants

Participants were recruited through Winston’s Wish, a family bereavement charity based in the UK. The charity offers support to bereaved children and their families on a local and national level through helplines, literature, drop-in services, and more formal therapeutic work. For families that engage with face-to-face work, there are various interventions offered. Telling the story, the focus of the current study, is a key intervention used with all families, usually in their first session. The aim of this intervention is to allow families to together tell the story of family life before the death, the death itself and how life is now, after the death. Families’ stories inform the planning of further interventions and the choice of support offered. Families repeat the storytelling intervention within peer groups (adults and children) at a residential weekend if they chose to attend, and, additionally, for the current study, families retold the story in an extra family session toward the end of the therapy contract.

Participant families were recruited by the team supporting families bereaved through illness or accidents. Families were excluded from recruitment if the therapeutic team felt that participation would be detrimental to a family’s therapeutic progress. Five families were recruited (a total of six adults and eight children) and 13.5 hr of audio-recorded data were collected over 10 sessions (2 sessions per family, 1 at the beginning and 1 toward the end of therapy). All the families were White British. The average time between the death and the first recording was 9.4 months (range 6–18 months), and the average time between the first and second recording was 7.2 months (range 1–14 months). The average time between the death and the second recording was 16.2 months (range 11–26 months).

Procedure

Ethical approval was granted by the first and third author’s Faculty Research Ethics Committee. The study adhered to the British Psychological Society’s (BPS, 2009, 2014) ethics code. After obtaining informed consent from the therapists, parents, and children, the Telling the story intervention was audio recorded. The recordings were transcribed orthographically (Braun & Clarke, 2013); transcription also followed guidance from Main and Goldwyn (1984), who stated that AAI interview transcriptions should be transcribed “verbatim” with all “errors” and hesitations transcribed, meaning that mispronunciations, gaps/silences, or stutters are noted. Accordingly, prolonged silences were timed and noted.

Data Analysis

This study utilized a flexible approach to theory-informed thematic analysis (TA; Braun & Clarke, 2006, 2013). Unlike many examples of theory-driven TA, this study was not underpinned by a positivist epistemology, which emphasizes coding reliability. This has been dubbed “small q” (Kidder & Fine, 1987) qualitative research—qualitative research conceptualized as the use of tools and techniques rather than the use of such tools and techniques within a qualitative paradigm. Instead, this study aimed to develop a more fully qualitative “deductive” approach to TA that prioritized researcher subjectivity and interpretation. This methodological approach also reflected the authors’ values of intersubjectivity and “professional artistry” in research, values which are core to the field of counseling psychology in the UK (BPS, Division of Counselling Psychology, 2005).

The data analysis involved six phases of coding and theme development, beginning with data familiarization and identifying items of potential analytic interest. The second phase of TA, “generating initial codes,” involved creating a “codebook” in two stages. First, codes were developed directly from the unresolved loss codes from the AAI. AAI coding was developed for single-person research interviews and as such does not capture systemic aspects of the family bereavement session. Thus, in an iterative analytic process, involving all three authors reading through and discussing the data together, further codes that captured these aspects were generated, informed by attachment theory. These additional codes were added to the codebook (see Table 1), and all the data were recoded by the first author using the final version of the codebook.

Table 1.

Examples of Codes From the Codebook.

Codes Taken Directly From Adult Attachment Interview	Codes Created Through Analyzing the Data
Indications of disbelief that the person is dead (e.g., slip of the tongue to present tense)	Giggles/laughter/incongruous emotion when talking about the death/difficult event
Deceased and speaker living parallel lives in the present	Family member physically leaves the room or goes significantly off topic during distressing content
Being dead is an activity	Therapist co-constructing disorganized narrative (e.g., using “Daddy” instead of “Daddy’s body” after the death)
Change of pronouns/attributing deceased actions to self	Disjuncture between child and adult memory of death
Time line confusion (e.g., confusion about dates/events leading up to death/own age/when death occurred)	Children left vague about details or with gaps in time line

Examples of the data are shown in the Analysis section tagged by family number/recording number (1 or 2)/line number. For example: Family 2/1 Line 832 refers to the second family’s first session and line 832 in the transcript.

Analysis

Two themes were identified in the data: evidence of unresolved loss and creating incoherence.

Evidence of Unresolved Loss

This theme describes the way that unresolved loss, as defined by the AAI codes, is demonstrated in the families’ stories through incoherence. Most of the codes captured by this theme have been taken directly from the AAI. There were signs of unresolved loss in all of the family’s stories, both at first and second tellings, and the signs took three forms, which are reflected in the titles of the three subthemes.

Lapses in the monitoring of reason

Most of the families’ stories evidenced “indications of disbelief that the person is dead.” Examples of this can be seen throughout the families’ stories in both first and second tellings. First, there were “slips of the tongue to the present tense.” An example of this is shown in Family 2, when Andrew talked about his wife who died of cancer 6 months previously in the present tense: “The first pain (.) her worst pain is always in the morning” (Family 2/1 Line 832). Other indicators of “disbelief” were talking as if “the deceased is living a parallel life in the present” and “being dead is an activity” (which involved referring to a dead loved one as having animate living characteristics in the present). Examples of both of these codes were particularly evident when families described seeing their loved one’s body for the first time after the death. In Family 1, Katie described her partner’s body as follows: “He was quite cold and different wasn’t he? So we put him in the blanket to keep him warm” (Family 1/2 Line 78). The idea that Katie’s partner could feel the cold and needed to be kept warm, even though he was dead, illustrates the attribution of living characteristics to a dead body. In Family 3, when the father’s body was returned from the hospital to the undertakers, the mother used language that created the impression that he was still alive and perhaps coming back from being away on a trip rather than having died:

Mum:

I was working on the Thursday and I just thought I just got this feeling (.) it was going to be the Thursday he’s going to come back and I really don’t to be at work when he comes back and family will wanna go up and see his Dad (.) and Wednesday night I was thinking do I wanna phone up work and say I don’t wanna come in because I know I just know that their Dad is going to come back tomorrow (.) (Family 3/1 Line 372–374)

Other aspects of the lack of monitoring of reason expressed by the families were disorientation with respect to time and space—family members described days passing in a “blur” (Family 1/2 Line 245)—and psychological confusion was seen in statements that were paradoxical or impossible, for example: “(.) I knew something was wrong but I didn’t know” (Family 4/1 Line 218).

A very common indicator of incoherence in all of the families’ stories was confusion around the time line of the death itself and the time line of events leading up to the death. Family members were confused about what happened when; children were confused as to how old they were at certain stages, particularly over longer illnesses, and confused about events surrounding the death itself. This was seen throughout all the stories told and led to a sense of incoherence in the narratives.

The daughter in Family 1, Alice, displayed significant disruption in the time line as she could not remember a time before her stepfather was ill. The illness and death had become uncontained and stretched across all of her memories of her stepfather and their life together, leading to a huge confusion about the time line: “I can’t say how he was I can’t say any think about how he came became before he came ill and died ‘cause he was already ill (.) even though I met him before” (Family 1/1 Line 18–19).

Confusion about the time line was also displayed through events that were forgotten or partially remembered. Events and details being forgotten left gaps in the time line and resulted in a lack of detail and depth to the stories of the family members’ death, creating uncertainty and confusion. Episodic memories are important in creating clear and coherent time lines (Ehlers & Clark, 2000), so missing details such as these are also important markers of incoherence.

Lapses in the monitoring of discourse

The second subtheme is illustrated by examples of the families finding it difficult to monitor how they were forming their narrative. A common sign of incoherence in the stories was unfinished sentences and prolonged silences. Unfinished sentences are understood in the AAI to be evidence of the speaker being overwhelmed by the thought of the death and unable to monitor or repair their speech (Main, Goldwyn, & Hesse, 2003). Prolonged silences can be understood as moments of preoccupation with the death and part of the “freeze” mechanism that is triggered when talking about distressing material (Hesse & Main, 1999, 2006). In this study, silences longer than 6 s were coded and considered to be indicators of incoherence, in contrast with the 20- to 30-s silences coded in the AAI. The rationale for this was based on interpretive judgment after listening to all the recordings and noting silences that seemed “appropriate” in terms of turn taking and the natural flow of talk between individuals. These contrasted with silences that were midsentence or were disruptive to the flow of talk. These silences were timed and found to be all 6 s or longer, so this became the criteria for coding. This is in line with other findings that silences over 5 s can be considered problematic in conversation (Jefferson, 1988) and that the “usual” length for silences in psychotherapy conversation is 2 s (Berger, 2011). Unfinished sentences and prolonged silences were spread throughout the narratives but concentrated in particularly difficult parts of the story as demonstrated below in this extract from the mother in Family 1: “They were just in hospital (.) umm ((long pause 6 seconds)) just checking on his body (.) making sure everything is working as it should have been (.)” (Family 1/1 Lines 281–282).

Unfinished sentences were also common in the stories, both in adults and children’s contributions. When Rosie described her children visiting their father’s body, she found it difficult to monitor her speech, and there was disruption as sentences were left unfinished: “and then they went to see- I think he’d- I was at work when he- when he came back (.) and I-” (Family 3/2 Line 314).

Family members also displayed incoherence by going “off topic” midstorytelling. An example of this came in Family 5’s first recording that takes place at the therapists’ office at a point where the father and daughter (Brenda, aged 7) were talking about the mother’s funeral and the details of how her body was dressed and put into the coffin:

Dad:

no no you dress her outside the box and then put her in

Brenda:

so outside the box

Dad:

uh huh (.) and then place Mummy in nice and cosy and comfy

Brenda:

why is there poo on the window

Therapist:

because there are some birds that fly by that window that’s why there’s poo on that window

Brenda:

is that Saint Greg’s church

Therapist:

ah I’m not sure (.) I think it’s Saint Martins (.) so after mummy died you went to (.) she was at (Family 5/1 Lines 935–938)

Brenda was unable to maintain the conversation about her mother’s body and she switches off topic, literally “out of the room,” to the bird droppings on the window. Main and Goldwyn (1984) consider that when a speaker wanders to irrelevant topics or suddenly changes topic when creating a narrative, this is due to a lack of monitoring of their own speech, something that results from the speaker losing touch with the present context because of the distressing content. This diversion away from the distressing content can be understood as a small dissociative act, regulating the affect of the speaker by changing the topic to something less distressing and manageable (Parkinson & Totterdell, 1999).

Behavioral reactions

This theme captures the codes developed from analyzing the data and focuses on how the family members tolerated and expressed emotion during the storytelling. There were moments of appropriate laughter and humor as families told their stories, such as when a mother and daughter laughed about their experiences of boxercise (Family 3/2 Lines 455–462). However, there were times when there were giggles and laughter when talking about the death, and these instances of laughter were coupled with evidence of incongruous emotion. The mother from Family 1 demonstrated this incoherence when describing who came to the house as soon as the father’s body was found: “(.) yeh (.) Jane came (.) Aunty Steph came (.) (Laugh) then luckily everyone went home. It was quite a (.) I mean cos everything was fine” (Family 1/1 Lines 374–375). She then continues to describe the father’s funeral saying: “(Laughs) I’ve never seen so many people in one place (Laughs) ha-ha” (Line 448). This laughter paired with this content creates a confusing mismatch between the events described and emotion. As found in other research, laughter when talking about a death is a sign of incoherence and unresolved aspects of the loss (Lyons-Ruth, Yellin, Melnick, & Atwood, 2005; Salvatore et al., 2006).

The final and very powerful example of evidence of unresolved loss evident in the data is the code “family member physically leaves the room.” There are examples of family members disconnecting from the storytelling by moving out of the room in three of the families, all of these examples taking place in the first family therapy session. The first recording of Family 4 took place in the family’s home. At the point in the story when the father is talking about finding his son’s body, the mother leaves the room, taking the pet dogs outside. She gives no verbal indication for her reason for leaving the room. Later in the session, when the youngest son is talking about events that may have contributed to his brother’s death, the father is heard on the recording getting up from his chair and can be heard moving things and banging in the distance. He then returns a couple of minutes later. There is again no explanation given or permission sought from other family members or the therapist nor does the therapist comment on either of these “breaks” from the session. This physical response to distressing content can be understood as intolerance of difficult emotions and having to disconnect from the narrative as it is too overwhelming (Parkinson & Totterdell, 1999).

Creating Incoherence

The second theme reflects the actions of family members and therapists that maintain the incoherence that is already present in the stories and thwarts the creation of a coherent and collaborative story. This theme captures the codes developed through analyzing the data and is focused systemically on the impact of the interactions between those present in the session.

A clear example of noncollaboration between family members was parental resistance to providing clear and full information to their children, which leaves their children with gaps in their time line of the death or vague about the details of the death. Research shows that appropriate details and clear understanding of events are vital for a coherent story (Ehlers & Clark, 2000) and that stories with fuller detail and a clearer time line structure lead to better outcomes for families (Figley & Kiser, 2013). One example of this was the lack of clarity around details seen in Family 1 with the use of nonspecific language and words by the mother such as “horrible things” (line 365) and “nasty stuff” (line 342) to describe medical equipment and the failed attempts at resuscitation. Although this may be understood as a mother’s attempt to protect her daughter from details about the death, the daughter is left confused and without a clear story of her own. Not giving children full details of events in an age-appropriate way perpetuates mystery for them and does not allow them to create a coherent account of what happened. Children need accurate information about the death so they can avoid “magical thinking” or filling in the gaps with misinformation that may lead to self-blame for the loss (Howarth, 2011; Lampton & Cremeans, 2002). Withholding details also results in the creation of a story where a family member has exclusive insight or understanding about events, meaning there cannot be a cocreated story as details are not shared.

In most of the families’ stories, there were disjunctures between the children’s and the adult’s memories of the death, which is not surprising as family members likely remember different versions of the events. However, there were also examples of disagreement between the members of the family as they told their story of the loss, over both factual events and interpretations of the events. This is to be expected to some extent within a family group; however, it is potentially the resistance to allowing all parts of the story to be told that creates incoherence. For example, in Family 4, a significant factor contributing to incoherence was resistance from the father to the inclusion of both parts of the son’s narrative in the family story and the son’s perspective on events leading up to his brother’s death. The son Mike’s search for meaning and sensemaking is disrupted by the father’s disallowing of this part of the narrative. Research shows that concurrence between family members and having a congruence in the family story may be more important for adjustment and well-being than the interpretation given to the event, even if this is a positive one (Davis, Harasymchuk, & Wohl, 2012). A family’s ability to allow a variety of perspectives in a story is disrupted by trauma and loss (Kiser, Baumgardner, & Dorado, 2010), so by continuing to disallow a “full” story, even if this includes negative interpretations of the events, the incoherence is perpetuated.

Another feature of this theme of creating incoherence is parents reshaping the child’s story to match their own understanding. In Family 5, the children have been talking about seeing their mother’s body in the open casket, and Steve (aged 5) had already described his sister Brenda as being scared and not wanting to kiss the body, and Brenda agreed this was right, she had been scared and had been slower to kiss her mother’s body than her brother. However, the father then went on to tell this part of the story differently: “the kids got to go and say their goodbyes and they weren’t a bit scared and they were constantly kissing her” (Family 5/1 Lines 822–823). This “rewriting” of the story denies the children validity in their emotions around a particularly difficult point and creates dissonance between their experience and what is being told as the family story. This can be understood as a display of “misattunement” (Fonagy, Gergely, & Jurist, 2002) from the father to his children’s story and emotional state.

Dissonance was also created by parents not tolerating distress or being emotionally dismissive toward their children in sessions. In Family 1, Lucy, the youngest daughter (aged 3), although not actively engaged in creating the verbal narrative, is still present for the session and impacts on the storytelling process. There are some really chaotic passages in the session where Lucy is clearly very distressed, but her tears are not acknowledged and she is not involved in the storytelling. For example, at one point, Alice has to shout to make herself heard above Lucy’s crying when clarifying with her mom the actual date that dad died. At another point, Lucy asks “Where my Daddy?” (Family 1/1 Line 64), and her question is not answered or addressed by anyone, including the therapist. The mother’s response to Lucy is anger and discipline rather than comfort or involvement in the story in an appropriate way, which perhaps demonstrates the mother’s lack of ability to engage emotionally with her children at this time and to empathize with their distress in an appropriate way. The mother has to leave the session with Lucy, and this adds to the incoherence as they are no longer present for a part of the storytelling. Overall, across the session, there is a strong sense of emotional disconnection between the mother and her children, and there are no instances of the mother offering comfort to either child nor offering sensemaking to them. Although there is evidence of the mother actively encouraging Alice to engage in the process, the mother appears emotionally withdrawn and passive in response to both current distress and accounts of distress.

This theme of creating incoherence also describes actions taken by adults: the therapists, parents, or sometimes parents and therapists together. In some stories, parents use exclusive adult language, which leaves the child out of the story, for example, medical terminology that is not understandable by the child such as specific names of drugs or treatments. Other instances of adult exclusivity are using phrases that create “in jokes” between parent and therapist. In Family 2, the father is talking about a cruise holiday the family took together before the death of his wife that had been recommended by his father-in-law. He says,

Dad:

Sarah’s father was (.) for want of a better phrase (.) a serial cruiser

Therapist 1:

Ha-ha (Family 2/1 Lines 273–274)

Although this may be understood as a harmless joke between adults, or even strengthening the therapeutic relationship, it has the possibility of disengaging the child from the process of storytelling as the language used is not understandable by them nor cocreated.

In two of the families’ stories, the therapist is involved in the cocreation and maintenance of a disorganized narrative. This is primarily through the use of language when talking about a person who has died or a dead body, which does not clearly identify the person or body as dead. In Family 1, the level of incoherence in the first story is high, and at points, the therapist is party to creating the incoherence. In this extract, Alice and her mother are talking about what happened to the stepfather’s body once he had died:

Mum:

And that was before, no that was after Daddy had gone

Alice:

yeh

Therapist:

And where did Daddy go to?

Alice:

hospital

Therapist:

ah so the ambulance took him?

Alice:

yeh

Therapist:

ah ok

Mum:

it was (.) it wasn’t the ambulance that took him, was it?

Alice:

what was did it?

Mum:

it was the funeral people, wasn’t it? (Family 1/1 Lines 144–153)

The language used gives a sense of “aliveness” to the father, and we could easily believe they are referring to him going to hospital because he is unwell until the mother mentions “funeral people.” The therapist is pulled into this incoherence, and instead of using a phrase such as “Daddy’s body,” which would clearly indicate that he was dead, she continued to use language such as “him” that perpetuates the incoherence and disbelief he is dead. This resembles Salvatore, Dimaggio, and Semerari’s (2004) findings regarding the impact of disorganized narratives on therapists: that they too experience feelings of confusion and chaos and at worst are pulled into behaviors that become antitherapeutic.

The final way in which families created incoherence in their storytelling was omissions in the second telling of the story, such as having different starting points (and therefore missing out significant events) or having significant details missing (such as the grandmother being in the house when the son died [Family 4]) or a notable reduction in the emotional richness and expressivity in the story. When significant details are missing, this creates distinct differences between the first and second telling that potentially foster incoherence.

Discussion

The aim of this study was to utilize the framework of attachment theory and specifically the unresolved loss codes of the AAI to explore how incoherence is expressed in bereaved family’s therapy narratives. The findings suggest that the AAI U coding can be productively used to analyze narratives of the death of a family member that are extracted from family bereavement therapy sessions. Furthermore, the findings also suggest that these codes can be used to identify the narrative incoherence that would be, within attachment theory, expected within about a year of a family loss. The findings also suggest the value of an attachment-informed TA to show how incoherence can be systemically created by family members and therapists in the process of telling the narrative.

Previous research has also shown that unresolved loss codes can be identified in transcripts of an individual’s clinical treatment sessions (Thomson, 2010); therefore, the findings from this study add to the body of evidence demonstrating the value of AAI coding beyond the AAI interviewing process and that attachment theory describes processes that occur not just in the AAI interview.

The findings also suggest that not only is incoherence demonstrated through individual narratives but also that incoherence is manifested in behavior and seen interactionally between family members, providing evidence of the microprocesses involved in the cocreation of attachment in families. Baradon and Steele (2008) identified behaviors (as well as narrative) in infant–parent psychotherapy that they understood as demonstrating representational knowledge (i.e., IWMs) through “action and enactment” (p. 209) in therapy sessions. These include frightening behaviors (FR)/parental anomalous behaviors (PAB) such as moving away from the infant without cause and dissociative “blind moments.” The findings of the current study show parallel examples of incoherence enacted between family members such as adults resisting giving further information or resistance to hearing or discussing certain parts of the story and adults reshaping the story or being emotionally dismissive. The behaviors that are demonstrated in these sessions can be likened to the FR/PAB identified by Abrams, Rifkin, and Hesse (2006) that are proposed to “transmit” unresolved loss from parent to child. Some of the interactive patterns found in the families’ narratives are also subtle, brief, and unmonitored by the parents, such as the prolonged silences or unfinished sentences (with parallels to the FR dissociative blind moments). However, some are more overt, such as disagreement about the facts or leaving the room during the session (with parallels to FR backing away from a child). Whether subtle or overt, these actions all create further incoherence. Moreover, these are aspects of incoherence that could not be present in an individual narrative (such as the AAI) as they are interpersonal manifestations of incoherence. Using an attachment-informed TA of therapy data has thus allowed these broader aspects of incoherence to be recognized and understood within the context of a family’s bereavement narrative. The findings also suggest ways that those in attachment systems can collaboratively create incoherence and perpetuate lack of resolution by resisting the process of resolution.

The methodology of this study offers a new way of conducting research on therapy data that provides a more accessible method of using “live” therapy data and engaging in theory-informed process research than highly technical and demanding methods such as conversation analysis (e.g., Antaki, Barnes, & Leudar, 2005). TA is often (particularly outside of the United States) associated with inductive analysis, so the current study is unusual in that the use of TA is explicitly informed by attachment theory. The analytic process promotes both exploration of the data and the practical application of the theory to the data; as such, it seeks to link theories (of change, i.e., increasing/decreasing coherence) to actual practice in the room. This method thus creates opportunities for psychology and psychotherapy researchers to use this method to examine how practice-relevant theory can be understood at the level of therapy microprocesses.

There are limitations related to the design of the study. Audio recordings of the therapy sessions were used rather than video recordings, which limited the amount of information and data available for transcription and coding. There were elements of behavior that could be picked up on the recordings, such as family members leaving the room, but some behavioral aspect of sessions was not accessible. Although the AAI coding is based solely on verbal narrative, other methods of measuring attachment status (e.g., the SST; Ainsworth, Blehar, Waters, & Wall, 1978) place importance on physical proximity and positioning (proximity seeking) as well as touch and giving of comfort between parent and child. To address these limitations, further research could include using video recording of therapy sessions, which would allow for a wider analysis of the behavioral aspects that are important in attachment theory. Such analysis could draw on the Child Attachment Interview developed by Target, Fonagy, and Shmueli-Goetz (2003), which incorporates behavioral as well as linguistic markers of incoherence.

Conclusions

The study demonstrates the value of using attachment theory as a framework for understanding bereavement narratives and the therapeutic encounter in family bereavement therapy. The study highlighted processes that occur in family bereavement work and contributes new understanding about bereavement narrative processes, in particular how incoherence, a marker of attachment (in)security, appears to manifest dialogically and interactionally.

Footnotes

Acknowledgment

The authors would like to thank the team at Winston’s Wish in Cheltenham who helped make this research possible and all the families who allowed their stories to be heard.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Victoria Clarke

References

Abrams

K. Y.

Rifkin

Hesse

(2006). Examining the role of parental frightened/frightening subtypes in predicting disorganized attachment within a brief observational procedure. Development and Psychopathology, 18, 345–361.

Ahn

H. N.

Wampold

B. E.

(2001). Where oh where are the specific ingredients? A meta-analysis of component studies in counseling and psychotherapy. Journal of Counseling Psychology, 48, 251–257.

Ainsworth

M. D. S.

Blehar

M. C.

Waters

Wall

(1978). Patterns of attachment: A psychological study of the strange situation. Hillsdale, NJ: Erlbaum.

Ainsworth

M. D. S.

Eichberg

(1991). Effects on infant-mother attachment of mother’s unresolved loss of an attachment figure, or other traumatic experience. Attachment Across the Life Cycle, 3, 160–183.

Ainsworth

M. D. S.

Wittig

B. A.

(1969). Attachment and exploratory behavior of 1-year-olds in a strange situation. In Foss

B. M.

(Ed.), Determinants of infant behavior (Vol. IV, pp. 111–136). London, England: Methuen.

Akerman

Statham

(2011). Childhood bereavement: A rapid literature review. London, England: Child Wellbeing Research Centre.

Antaki

Barnes

Leudar

(2005). Diagnostic formulations in psychotherapy. Discourse Studies, 7, 627–647.

Baradon

Steele

(2008). Integrating the AAI in the clinical process of psychoanalytic parent-infant psychotherapy in a case of relational trauma. In Steele

Steele

(Eds.), Clinical applications of the adult attachment interview (pp. 195–212). New York, NY: Guilford Press.

Benoit

Parker

K. C.

(1994). Stability and transmission of attachment across three generations. Child Development, 65, 1444–1456.

10.

Berger

(2011). Support and evidence for considering local contingencies in studying and transcribing silence in conversation. Pragmatics, 21, 291–306.

11.

Bishop

Stedmon

Dallos

(2015). Mothers’ narratives about having a child with cancer: A view through the attachment lens. Clinical Child Psychology and Psychiatry, 20, 611–625.

12.

Bonanno

G. A.

Kaltman

(2001). The varieties of grief experience. Clinical Psychology Review, 21, 705–734.

13.

Bowlby

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

14.

Braun

Clarke

(2006). Using thematic analysis in psychology. Qualitative Research in Psychology, 3, 77–101.

15.

Braun

Clarke

(2013). Successful qualitative research: A practical guide for beginners. London, England: Sage.

16.

British Psychological Society. (2009). Code of ethics and conduct. Retrieved from http://www.bps.org.uk/system/files/Public%20files/aa%20Standard%20Docs/inf94_code_web_ethics_conduct.pdf

17.

British Psychological Society. (2014). Code of human research ethics (2nd ed.). Retrieved from https://www.bps.org.uk/news-and-policy/bps-code-human-research-ethics-2nd-edition-2014

18.

British Psychological Society, Division of Counselling Psychology. (2005). Guidelines for professional practice in counselling psychology. Retrieved from http://www.bps.org.uk/sites/default/files/documents/professional_practice_guidelines_-division_of_counselling_psychology.pdf

19.

Currier

J. M.

Holland

J. M.

Neimeyer

R. A.

(2007). The effectiveness of bereavement interventions with children: A meta-analytic review of controlled outcome research. Journal of Clinical Child and Adolescent Psychology, 36, 253–259.

20.

Davis

C. G.

Harasymchuk

Wohl

M. J.

(2012). Finding meaning in a traumatic loss: A families approach. Journal of Traumatic Stress, 25, 142–149.

21.

Dowdney

(2000). Annotation: Childhood bereavement following parental death. Journal of Child Psychology and Psychiatry, 41, 819–830.

22.

Ehlers

Clark

D. M.

(2000). A cognitive model of posttraumatic stress disorder. Behaviour Research and Therapy, 38, 319–345.

23.

Figley

C. R.

Kiser

L. J.

(2013). Helping traumatized families. New York, NY: Routledge.

24.

Finlay

(2014). Qualitative methodologies. In Vossler

Moller

(Eds.), The counselling and psychotherapy research handbook (pp. 164–183). London, England: Sage.

25.

Fonagy

Gergely

Jurist

E. L.

(Eds.), (2002). Affect regulation, mentalization and the development of the self. New York, NY: Karnac Books.

26.

George

Kaplan

Main

(1985). Adult attachment interview protocol (2nd ed.). Unpublished manuscript, University of California, Berkeley.

27.

Henton

(2012). Practice-based research and counselling psychology: A critical review and proposal. Counselling Psychology Review, 27, 11–28.

28.

Hesse

Main

(1999). Second-generation effects of unresolved trauma in nonmaltreating parents: Dissociated, frightened, and threatening parental behavior. Psychoanalytic Inquiry, 19, 481–540.

29.

Hesse

Main

(2006). Frightened, threatening, and dissociative parental behavior in low-risk samples: Description, discussion, and interpretations. Development and Psychopathology, 18, 309–343.

30.

Hooghe

De Mol

Baetens

Zech

(2013). The measurement of couple and family interactions and relationship quality in bereavement research. Family Science, 4, 66–78.

31.

Howarth

R. A.

(2011). Promoting the adjustment of parentally bereaved children. Journal of Mental Health Counseling, 33, 21–32.

32.

Hughes

Hardy

Kendrick

(2000). Assessing adult attachment status with clinically-orientated interviews: A brief report. British Journal of Medical Psychology, 73, 279–283.

33.

Jefferson

(1988). Notes on a possible metric which provides for a “standard maximum” silence of approximately one second in conversation. In Roger

Bull

(Eds.), Conversation: An interdisciplinary perspective (pp. 167–196). Clevedon, England: Multilingual Matters.

34.

Kidder

L. H.

Fine

(1987). Qualitative and quantitative methods: When stories converge. New Directions for Program Evaluation, 1987, 57–75.

35.

Kiser

L. J.

Baumgardner

Dorado

(2010). Who are we, but for the stories we tell: Family stories and healing. Psychological Trauma: Theory, Research, Practice, and Policy, 2, 243–249.

36.

Lampton

C. C.

Cremeans

K. J.

(2002). Recent advances in the understanding of childhood grief and mourning. Marriage and Family, 5, 467–476.

37.

Larson

D. G.

Hoyt

W. T.

(2007). What has become of grief counseling? An evaluation of the empirical foundations of the new pessimism. Professional Psychology: Research and Practice, 38, 347–355.

38.

Lyons-Ruth

Yellin

Melnick

Atwood

(2005). Expanding the concept of unresolved mental states: Hostile/helpless states of mind on the Adult Attachment Interview are associated with disrupted mother–infant communication and infant disorganization. Development and Psychopathology, 17, 1–23.

39.

Main

Goldwyn

(1984). Adult attachment scoring and classification system. Unpublished manuscript, University of California, Berkeley.

40.

Main

Goldwyn

Hesse

. (2003). The adult attachment interview: Scoring and classification system, version 7.2. Unpublished manuscript, University of California, Berkeley.

41.

Main

Hesse

E. D.

(1990). Parents’ unresolved traumatic experiences are related to infant disorganized attachment status: Is frightened and/or frightening parental behavior the linking mechanism? In Greenberg

Cichetti

Cummings

, (Eds.), Attachment in the preschool years (pp. 161–182). Chicago, IL: Chicago University Press.

42.

Main

Kaplan

Cassidy

(1985). Security in infancy, childhood, and adulthood: A move to the level of representation. In Bretherton

Waters

(Eds.), Growing points of attachment theory and research. Monographs of the Society for Research in Child Development (pp. 66–107), 50 (1-2, Serial No. 203). Chicago, IL: Chicago University Press.

43.

Mallinckrodt

(2011). Addressing the decline in counseling and supervision process and outcome research in the Journal of Counseling Psychology. The Counseling Psychologist, 39, 701–714.

44.

McClatchey

I. S.

Vonk

M. E.

Lee

Bride

(2014). Traumatic and complicated grief among children: One or two constructs? Death Studies, 38, 69–78.

45.

Midgley

(2004). Sailing between Scylla and Charybdis 1: Incorporating qualitative approaches into child psychotherapy research. Journal of Child Psychotherapy, 30, 89–111.

46.

Muscetta

Dazzi

Decoro

Ortu

Speranza

A. M

. (1999). “States of mind with respect to attachment” and change in a psychotherapeutic relationship: A study of the coherence of transcript in a short-term psychotherapy with an adolescent. Psychoanalytic Inquiry, 19, 885–921.

47.

Neimeyer

R. A.

Currier

J. M.

(2009). Grief therapy evidence of efficacy and emerging directions. Current Directions in Psychological Science, 18, 352–356.

48.

Neimeyer

R. A.

Holland

J. M.

Currier

J. M.

(2008). Meaning reconstruction in later life: Toward a cognitive-constructivist approach to grief therapy. In Gallagher

Thompson

Steffan

Thompson

(Eds.), Handbook of behavioral and cognitive therapies with older adults (pp. 264–277). New York, NY: Springer.

49.

Parkinson

Totterdell

(1999). Classifying affect-regulation strategies. Cognition & Emotion, 13, 277–303.

50.

Ribbens Mccarthy

(2007). “They all look as if they’re coping, but I’m not”: The relational power/lessness of “youth” in responding to experiences of bereavement. Journal of Youth Studies, 10, 285–303.

51.

Salvatore

Conti

Fiore

Carcione

Dimaggio

Semerari

(2006). Disorganized narratives: Problems in treatment and therapist intervention hierarchy. Journal of Constructivist Psychology, 19, 191–207.

52.

Salvatore

Dimaggio

Semerari

(2004). A model of narrative development: Implications for understanding psychopathology and guiding therapy. Psychology and Psychotherapy: Theory, Research and Practice, 77, 231–254.

53.

Steele

Fonagy

(1996). Associations among attachment classifications of mothers, fathers, and their infants. Child Development, 67, 541–555.

54.

Stroebe

M. S.

(2010). Bereavement in family context: Coping with the loss of a loved one. Family Science, 1, 144–151.

55.

Target

Fonagy

Shmueli-Goetz

(2003). Attachment representations in school-age children: The development of the child attachment interview (CAI). Journal of Child Psychotherapy, 29, 171–186.

56.

Thomson

(2010). Loss and disorganization from an attachment perspective. Death Studies, 34, 893–914.