Becoming invisible: The effect of triangulation on children’s well-being

The study explored children’s experience of triangulation in their families. In all, 15 children aged 11–16 years, who were attending an early intervention family therapy service, participated in the study. The children’s understandings and emotional experience of triangulation were explored by comparing their responses to pictures from the Separation Anxiety Test (SAT) and a set of pictures designed for the study depicting a variety of triangulation conflicts in families. An interview regarding the children’s personal family experiences of triangulation was also undertaken and clinical information about the children’s family contexts was also utilised. Statistical analysis was conducted based on eight of the children for whom a full data set was available. This indicated that children showed greater levels of anxiety in response to the triangulation as opposed to the separation scenarios. Qualitative analysis supported this finding and revealed that many of the children felt ‘invisible’ due to parents’ pre-occupation with marital conflict, felt caught in the middle of conflicts and coerced to take sides. Although able to describe their reactions and showing greater negative emotional responses to the triadic pictures, they were not consciously aware of the negative impacts of triangulation on their sense of well-being. Clinical implications are discussed with a focus on encouraging child-centred approaches to family therapy.

The concept of triangulation has been one of the conceptual cornerstones of systemic therapy (Dallos & Draper, 2009) and represented a fundamental shift from linear to systemic thinking and the suggestion that triads, rather than individuals or dyads, are the fundamental building blocks of family life. Bowen (1978) defined triangulation as potentially occurring in any pattern of family relationships and as a network of inter-locking triadic processes. Our focus here is on children’s experience of conflictual triadic processes in relation to their parents or carers. Specifically, triangulation is seen to occur when children become entangled in parental conflict, through coercion from parents to take sides, the child deciding to mediate or attempting to distract their parents from their conflict (Minuchin, 1974). Byng-Hall (1980) has suggested that the child can come to function as the ‘distance regulator’ of the couple’s marital relationship, which implies that the child behaves in ways to attempt to regulate not only the level of conflict but also closeness between the parents. This is an extremely important systemic idea since it suggests that the child becomes not simply the ‘victim’ of the parents’ conflict and distress but also can act to keep them apart and escalate negative processes. Research reveals that children can experience severe distress when ‘caught in the middle’ of their parents’ conflicts and tensions expressed, for example, as internalising symptoms, externalising problems, poor academic achievement, higher levels of depression/anxiety and weak parent–child relationships (Amato & Afifi, 2006; Buchanan, Maccoby, & Dornbusch, 1991; Buehler & Welsh, 2009; Gerard, Krishnakumar, & Buehler, 2006; Ng & Smith, 2006; Shelton & Harold, 2008). Some contemporary research has suggested that the patterns are developmentally very early and fundamental. For example, Fivaz-Depeursinge, Cairo, Lavanchy Scoila, & Favez (2012) have shown that differences in how consistent and co-ordinated parents’ responses are to the infant in turn shapes the infant’s responses to the parents, especially their ability to manage triadic interactions. Parents’ conflict and distress can distort their ability to function as attachment figures for their child in limiting their ability to offer care, support and guidance. However, attachment theory has essentially offered a dyadic picture of the development of security in children (Bowlby, 1969, 1988) and has not considered triadic attachment. There has been little exploration of how, for example, a mother’s ability to offer a secure attachment is influenced by the anxieties and tensions in her relationships with the child’s father (Dubois-Comtois & Moss, 2008). Not only the mother’s ability to respond empathetically but also the wider context of the security of the child’s family life is threatened by the conflict between the parents: the child runs the risk of losing not only the attachment figure but also their home, family and physical safety. Byng-Hall (1995) has suggested further that the parents’ own attachment needs can be mediated by their child who can function as marital ‘distance regulator’ in managing the nature of the emotional distance between them. This can involve considerable psychological cost to the child in terms of sacrificing their own attachment needs. We would suggest that a child has an attachment, not just with each parent but with the relationship between them. In addition, the conflicts between the parents may be related to permutations of matched and mis-matched attachment strategies employed by the parents (Dallos, 2014). Where both parents have secured attachment strategies, this is relatively unproblematic for the child. However, if both hold avoidant patterns, the child may struggle to gain attachment responses from either, or where both are anxious the child may struggle to gain some emotional space. However, adjusting to combinations of mis-matched patterns might be even more complex in that the child is required to develop ‘multiple-models’ to adapt to the differences. For example, an anxious parent may draw the child in to offer them emotional support against the other parent. In contrast, the avoidant parent backs off perhaps leaving the child feeling abandoned by them or feeling guilty that they have rejected their parent. Interestingly, according to Buchanan et al. (1991), being close to both parents is associated with low feelings of being caught in the middle, as both parents may then be more sensitive to their children’s feelings and therefore be less likely to behave in ways that triangulate them. In terms of strategies, a child may attempt avoiding strategies by attempting to try to withdraw from the parents’ conflicts although the parents may resist this so that the child feels compelled to side with one parent and avoid a relationship with the other. Children may also attempt to ‘mask’ or inhibit expressions of distress and reduce their presence in the conflict (Dallos & Denford, 2008; Davies & Forman, 2002) especially when they also feel to blame for parents’ arguments. Alternatively, involvement can consist of the child becoming drawn in and attempting to manage the parents’ conflicts such as trying to be a peacemaker, trying to calm the parents or being especially helpful.

Importantly, Grych (1998) discussed how feeling responsible for marital conflict provides children with a sense of coping efficacy and perceived control over conflict, which increases the likelihood of involvement. The child may also adopt a more angry, coercive strategy towards the parents or distract them by engaging in dangerous or disruptive behaviours. Overall, both these strategies of avoidance and involvement have costs as they can cause psychological maladjustment. The word ‘strategy’ is used here with caution, since in our clinical experience these are not conscious choices and, to the contrary, children typically appear largely unaware of the distressing and disorienting effect that triangulation has on them. The patterns of triangulation can be seen to follow a developmental pathway. An infant or young child may express the tensions, conflicts and distresses between his parents in terms of a variety of somatic states, such as agitation, crying, sleeplessness or vomiting. These symptoms may have the effect of distracting the parents from their conflicts and temporarily reduce the emotional tension. With the development of language, in addition to this process, the child may be verbally persuaded to take sides in the conflict between the parents, and older age groups may use avoidance more effectively, perhaps due to the autonomy they have to leave the house or call their friends (Shelton & Harold, 2008). Older children may also be able to see multiple points of view and parents may confide in them or have them relay messages to one another, possibly increasing feelings of being entrapped (Buchanan et al., 1991). Buehler and Welsh (2009) argue that adolescents may need parents to reassure them that they do not expect them to take sides in order to help them avoid becoming overly concerned and pre-occupied with their parents’ conflict, which can lead to negative consequences for the children’s mental health. Ringer and Crittenden (2007) found triangulation patterns were often related to a range of problems, including anorexia. Additionally, Schindler, Thomasius, Sack, Gemeinhardt, and Kustner (2007) found that in drug-dependent adolescents (DDAs), a ‘triangulated pattern’ was found in two-thirds of the sample, characterised by pre-occupied strategies in mothers and dismissing ones in fathers. Despite the centrality of the concept of triangulation to clinical practice, there has been surprisingly little research exploring the nature and experience of triangulation. Arguably, most of what we know comes from clinical observation of families engaged in family therapy.

This study

The purpose was to explore the children’s experience of triangulation and in particular to

1. Compare the attachment dilemmas and their severity resulting from exposure to triadic as opposed to dyadic attachment conflicts;
2. Explore children’s understandings of triangulation;
3. Gain further understanding of the attachment dilemmas that the experience of triangulation raises for children.

Participants and design

The participants were 15 children aged between 11 and 16 years (mean age = 12.13; females = 8; males = 7). The qualitative findings are based on data from all 15 children, but a full research data set was only available for eight children who were included in the statistical analysis. There were no significant differences between the eight children featured in the quantitative analysis and the wider sample. Constraints were imposed by the timing of clinical sessions and the availability of families to attend together, which was affected by travel and cost issues. The children were approached through a family therapy clinic and identified by the family therapy team as having substantial triangulation-related difficulties. The sample included children of separated or divorced couples, adopted children, and children living with step-parents. Presenting problems included anxiety, anger and conduct-related issues, phobia, criminal behaviours, depressive symptoms, self-harm, and separation anxiety. Consent was obtained from both the children and their parents. All of the children were from a White sample, but there was considerable variation in class and economic circumstances, ranging from families relying on benefits to those with parents in full-time professional employment. Ethical approval was sought and granted by the Research Ethics Committee of the local university in which the authors were based.

Measures employed

The children were given a measure of attachment strategies and an exploration of feelings and experiences elicited by visual material. The materials comprised a set of photographs designed to assess attachment strategies and triangulation experiences. Each child responded to nine pictures, with 16 measures recorded for each picture (Table 1), and two sets of cards were produced featuring a boy or girl in each of the scenarios. The measures fell into the following groupings: attachment responses, avoidant or anxious preoccupied styles, types of solutions, reflective awareness, triadic processes, and evidence of traumatic intrusions. The qualitative data were gathered using a semi-structured interview, and the quantitative data were produced through scoring the responses found within the qualitative accounts.

The Separation Anxiety Test and triadic situation pictures

The Separation Anxiety Test (dyadic pictures) consisted of the first set based on the original Separation Anxiety Test (SAT) (Resnick, 1993) and adapted by Wright, Binney, and Smith (1995). These pictures included scenarios such as a child saying goodbye to both parents as he or she goes away on holiday, the mother going into hospital, and the father getting arrested. These images had been previously piloted in our clinical work and a preliminary study, which found that they elicited significant attachment responses. The pictures used in this study depict White children in Western dress. Although Wright et al. (1995) developed a set of pictures for ethnic minority families, this research was conducted in the South West of the United Kingdom, which has a very small ethnic minority population, so the original ‘Western’ pictures were used.

The triangulation scenarios (triadic pictures) comprised the second set of images, taken from Smalley (2013). These included scenes such as parents arguing over an inappropriate present for the child; parents waiting up as the child comes home late; a mother or father listening in to the child’s conversation with the separated parent; parents discussing the child’s school report; the father leaving after an argument; and the mother leaving to have an affair. These pictures were designed to elicit understandings and emotional responses related to typical triadic interactions and conflicts. Two sets of pictures were developed, featuring either male or female children, to facilitate identification with the young people depicted.

Quantitative findings

The analysis of the SAT and triadic pictures indicated that, overall, the triadic pictures generated higher levels of attachment distress than the dyadic ones. A one-way within-subject analysis of variance (ANOVA) was conducted for all nine pictures across the 16 dimensions (attachment sub-scales) for each picture. Each dimension was scored on a scale from 1 to 9, and average scores were calculated separately for the six triadic and three dyadic pictures across each of the 16 dimensions. This resulted in 16 repeated-measure scores for each child for each picture type (triadic vs. dyadic). A test for equality of variance—Levene’s test—indicated that variances were not equal across the two groups (dyadic and triadic). Attempts to correct this violation of homogeneity through data transformation were unsuccessful. Consequently, the original data were used with caution, and significance levels were made more stringent, adjusting the alpha threshold from .05 to .025.

Results revealed a highly significant main effect of triadic versus dyadic pictures (F(1, 6) = 28.64; p < .025). Further ANOVAs comparing all the pictures found significant differences between SAT picture 1 (Child goes on holiday) and triadic picture 5 (school report). Significant differences were also found between triadic picture 5 (school report) and both SAT picture 6 (Mother goes to hospital) and SAT picture 8 (Dad arrested).

Figure 1 shows a graph of average responses for SAT and triadic pictures, indicating that triadic pictures produced higher attachment anxiety scores on each of the attachment sub-scales—the higher the mean score, the more anxious the response—with the exceptions of unrealistically positive solutions, resisting the attachment scenario, dismissing the attachment scenario, and Parental Accurate Empathy (PAE), which reflects the child’s ability to consider that their parent figure understood how they felt.

Figure 2 displays the average scores on two attachment measures and avoidant solutions for all participants per picture, arranged in ascending order. This shows that the highest mean responses across these central dimensions were produced by triadic picture 5 (school report) and triadic picture 7 (argument, dad leaves). All SAT pictures produced the lowest responses across the dimensions.

Importantly, in some cases with the dyadic (SAT) pictures, children produced responses indicating secure attachment, characterized by a range of affect, good reflective functioning (RF), and a positive relationship with the carer. In contrast, triadic pictures—particularly triadic picture 5—elicited insecure responses marked by self-blame, anger, and very low PAE.

Gender differences were found to have a significant effect on responses in both picture types (SAT or triadic). Boys had on average higher scores for unrealistically positive solutions, resisting the attachment dilemma, dismissing and avoiding conflict, whereas girls score higher across all other dimensions. This supports finding by Minuchin (1974) and Davies and Lindsay (2004), predicting that girls are more affected by appraisals of self-blame and more likely to over-involve in conflict and feel responsible for restoring family harmony.

Qualitative analysis of attachment dimensions

The responses to the pictures suggested that children often felt very much caught in the middle of family conflict. For example, Rick clearly expressed his dilemma in the ‘ring dad’ scenario, indicating that both the character in the photograph and he himself would feel torn between loyalties and conscripted into a coalition with his father against his mother:

“He can’t just speak to his father without someone listening too. So like maybe he will say something different because someone is listening to his conversation. [...] He clearly wants to say something that he wants to express with his dad. So he could be saying some things like that his mum is being annoying.” (Rick: 136–142)

“She was probably shouting at her son because of the results. She was probably saying, ‘It is something to do with your dad isn’t it?’ Then he is probably to tell her that it wasn’t. But then she just doesn’t listen to him and phones him to go round and starts shouting.” (Rick: 212–215)

Similarly, Megan’s responses indicated that she felt drawn into taking sides with one parent against the other:

“You do not have to go. Because I’m more closer to you than dad.” (Megan: 554)

These examples illustrate the emotional strain on children who feel pressured to align with one parent, intensifying their sense of being caught in the middle of family conflict.

Avoidance and involvement in conflict

A common theme in the conflict situations was that children described either avoidance or involvement in the conflicts. The findings showed that the higher a child’s score on avoiding conflict, the lower their score was for catastrophic solutions, whereas the higher their involvement in conflict, the higher their score for catastrophic solutions. Regarding attachment strategies, involvement in conflict and catastrophic solutions predicted a pre-occupied attachment pattern, while avoidance of conflict and fewer catastrophic solutions were associated with avoidant strategies.

Mediating conflict and role reversal

This theme showed that children often involved themselves in conflicts to mediate or resolve them, for example, by trying to reason with the adults. In the inappropriate present scenario, the child (Megan) indicated that, imagining herself in the picture, she would adopt a parental stance toward her parents’ argument:

"I don’t know. I might like five minutes later – just let them cool down. Then I might come in and say, ‘Oh, just have it. Because I don’t want it to make you two fall out just over this little thing’. [...] Mum would probably ground me. But if dad argues with it, they have a big argument, so I’m just like, ‘Just ground me’." (Megan: 150–152; 201–202)

This reflects the child’s awareness of her role in mediating her parents’ conflict, employing thoughtful strategies like allowing time for cooling down, then returning to reason with them, and even accepting punishment as a way to prevent further arguments.

Fluid coping strategies

his theme highlighted that some children used flexible coping strategies; specifically, if one behavior didn’t work during a conflict, they would try another. For example, Megan said, “I would either say to them, ‘Right, if you want me to give it back, I will’. Or I would either just go in my room and just – yes, just get out of the situation. [...] Well, we just try and walk off and try to just let them calm down and then go back and just say, ‘Please stop it. We don’t like it. We don’t want you to fall out’” (Megan: 144–146/622–624). Similarly, Jack described, “No. I was trying to stop them from arguing and then just kept on arguing. So I got really upset and cried and just like, sort of walked off to my bedroom” (Jack: 338–340).

Megan initially attempted to mediate or avoid the conflict, which may relate to her younger age since children around 11 years old are thought to have more fluid attachment strategies. Dean reflected on how his coping had evolved over time: “I just walk away. He just [self-discipline 00:13:25] or he’ll just talk to us ___[ and he] started arguing. [...] Do you think you’ve changed? Because you said you now kind of walk away. Do you think you’ve always done that? Or has it- Sometimes I ask like, why they did it and get annoyed. But I won’t do it.” (Dean: 318–319/331–334).

As children mature into adolescents, their coping strategies often become less flexible, tending to settle on one approach that seems effective in stressful situations.

Self-blame

Self-blame showed a positive relationship with reflective functioning (RF) and a negative relationship with resisting the attachment scenario—where resistance was indicated by responses that avoided the emotional core of the dilemma, such as focusing on superficial details like what people were wearing. Self-blame emerged as a recurring theme, particularly in the narratives of Megan, Rick, and Kate. For example, Megan demonstrated a tendency to take personal responsibility for the conflict between her parents, especially when it involved her, such as the inappropriate present. She expressed a willingness to sacrifice or give something up in order to resolve the situation, and repeatedly described herself as being at fault:

“They’re arguing over my grades and it’s my fault. It’s not their fault. Mum can’t blame it on dad and mum can’t blame it on herself. It’s her fault. She’s really sad. But angry because they’re just arguing when it’s her fault. It’s not fair.” (Megan: 284–287)

Megan appeared to search for someone to blame—often herself—as a means of resolving the conflict once a cause had been identified.

Regarding Parental Accurate Empathy (PAE), the analysis found that PAE had a significant positive relationship with awareness of the emotional impact of triadic processes, which itself had significant positive associations with triadic thinking and trauma. Overall, the children generally believed their parents were reasonably aware of how they felt. However, they did not typically expect their parents to act constructively to alleviate their emotional distress. This is a notable and somewhat paradoxical finding.

In scenarios involving parental conflict—such as arguments over an inappropriate present or a school report—children frequently reported feeling ‘invisible.’ They described their parents as so consumed by marital problems that they failed to notice the child’s presence or emotional needs. Supporting the findings of Ringer and Crittenden (2007), the children expressed that during these conflicts, they felt as though their parents were unaware of their existence and especially unaware of their internal experiences or distress.

I don’t think they will be thinking about him. (Dean: 44–45)

For some of the children, RF and PAE were repeatedly lower for fathers compared to mothers. Peter and Megan show this by the difference in the intimacy they would exhibit towards their mother and father:

Probably kiss my mum, shake my dad’s hand and then get in the car. (Peter: 33)

The dad, I wouldn’t have a clue what my dad would think, I can’t really say. (Peter: 174)

His dad is probably thinking . . . I don’t know what his dad is thinking. (Peter: 253)

[Parents would] Not [know how I am feeling] unless I have said. (Peter: 188)

Yes, my mum would definitely know. But my dad would be like, ‘I don’t know’. (Megan: 347–348)

One possible explanation for these findings is that the children had predominantly spent more time with their mothers, and in all cases of parental separation, the children had lived with their mothers. This dynamic may have made it more difficult for them to develop or express Parental Accurate Empathy (PAE) toward their fathers. Being more aligned with their mothers, whether by circumstance or expectation, may have created an obstacle in empathizing with the absent father—especially when they had limited opportunity to hear his perspective. This is reflected in Jack’s statement:

“Dad is probably quite upset that she’s leaving. He’s probably thinking he will have to look after the son now.” (Jack: 306)

Jack’s response implies that caregiving is primarily the mother’s role and that the father would be a reluctant or unprepared parent in this situation.

However, there were notable exceptions to this pattern. While some children expressed that their mothers understood them better, this was not consistent across all cases. In certain instances, informed by clinical work with the families, it was evident that the child’s relationship with the mother was conflictual. As a result, some children believed that their mothers did not understand them. Conversely, some children—particularly those who missed an absent father—held the belief or hope that their father did understand them better, despite the lack of regular contact. This complexity reflects the nuanced and sometimes ambivalent ways children navigate loyalty, empathy, and emotional connection within triangulated family dynamics.

Two case studies
The following two case studies offer an overview of the children’s responses to the pictures accompanied by the case material available from the family therapy.

Megan

Megan’s responses were predominantly characterized by anxiety, self-blame, and high reflective functioning (RF), which aligned closely with her clinical background of being treated for anxiety. Her narratives frequently referenced past family conflicts that may have resulted in trauma, reinforcing the connection to her clinical profile. For instance, she expressed intense worry about her parents arguing and the potential consequences:

“Worried, just in case they fall out and one of them decides to leave or something? . . . I want them to do that [smooth things over] so then they don’t get in a big argument and then fall out and then just leave.” (Megan: 120–121/133–134)

This response illustrates her catastrophizing tendencies, which may stem from her lived experience—her parents had separated, and her father had left. These anxieties surfaced again in her reflections on a scenario where the father leaves:

“Really upset and just saying, ‘Please, dad, don’t go. I don’t want you to go, because you’re my dad. I just hope it won’t ___ [0:20:39] our seeing each other still. Because I’m your daughter and it’s not really fair if this overtakes our relationship.’” (Megan: 417–420)

This emotionally charged dialogue suggests a re-living of past trauma, as Megan seems to be pleading directly with her father. Her response demonstrates heightened emotional arousal and vividly reflects unresolved distress. The emotional tone continues as she recounts the actual separation from her father:

“Because my dad left me and my mum when I was three. I don’t really know anything then. But it still upsets.” (Megan: 442–423)

These reflections not only highlight Megan’s preoccupation with loss and abandonment but also underscore the personal and clinical relevance of triangulation and attachment conflict in her family experience.

The high instances of emotional arousal were particularly evident in the picture scenarios that closely mirrored Megan’s real-life experiences. Her emotional engagement was intensified when the content evoked personal memories of conflict or loss, leading to deeply felt, and at times, distressing responses. In the follow-up interview, she openly expressed her discomfort with family conflict:

“I normally just say, ‘Stop it, please’. Because instead of saying my dad, my mum and my sister, let’s say that, and they were arguing, I try to stop them. I say, ‘Stop arguing. Because I hate it. I just hate it’. I just don’t like them arguing.”(Megan: 593–596)

This direct and emotional statement highlights not only her heightened sensitivity to conflict but also her desire to act as a mediator—an effort to restore emotional harmony, which is consistent with her anxious and preoccupied attachment style.

Megan demonstrated a strong capacity for understanding and mentalisation, reflected in her high reflective functioning (RF) scores. She frequently showed an ability to consider multiple perspectives, including insight into the emotional states and motivations of others, particularly her mother. For example, in the 'ring dad' scenario, although her initial emotional response was anger—saying she would “hate” her mum—she quickly moved toward empathy, offering a plausible rationale for her mother’s behaviour:

“But then my mum . . . having the right reasons though. Because he might have just upped and left.” (Megan: 258–259)

This progression from reactive emotion to reflective empathy demonstrates her advanced RF skills. Despite her own emotional distress and complex family background, Megan was able to generate nuanced interpretations of her mother’s actions, showing both emotional intelligence and relational awareness. These patterns underscore how past trauma and current emotional needs can coexist with a strong capacity for reflective thought, especially in children navigating triangulated family dynamics.

Rick

Rick scored very highly on reflective functioning (RF) and demonstrated significant awareness of triadic processes and their emotional impacts. His responses revealed a complex profile: he was simultaneously dismissive, anxious, avoidant, and involved in conflict. Though identified by his parents as the most "difficult" child, his siblings described him as the most "sensitive" and the most affected by family conflict. This contrast suggests a child who internalizes emotional tensions and may act out partly as a response to feeling overwhelmed by unresolved familial dynamics. In his narrative responses, Rick often added emotionally charged details that were not directly presented in the scenarios, such as recalling that arguments would start when he was already in bed: “I think that they could do it when I am like in bed, and they start arguing again or something. […] he wishes they didn’t like shout.” (Rick: 75–76/82). These additions likely reflected the impact of real-life experiences on how he processed the scenarios.

Clinically, Rick’s situation was complex and distressing. His father was facing charges related to sexually inappropriate material, and his mother was accused by the father of alienating the children. These circumstances had clearly created a profound sense of confusion and distress for Rick. His responses to the ‘ring dad’ scenario illustrated how deeply embedded he was in the triangulated dynamics between his parents. He rationalized his mother’s behavior as protective, suggesting she was monitoring the phone call to ensure the father said only “good things,” while also empathizing with his father’s sense of being unfairly accused: “She would think that she was making sure that everything was safe, like the phone call and everything. That he is saying good things, so ___ negative things and that.” (Rick: 165–167) and “The mum is probably feeling annoyed with his dad, and the dad is probably feeling, ‘What have I done?’ Or something.”(Rick: 263–264).

These insights point to a child who was not only highly aware of the emotional undercurrents but also deeply conflicted about where his loyalties should lie. His defense of his father and distress over the situation placed him in a triangulated role between his mother and step-father, compounding his emotional burden. His mother’s response—perceiving him as disruptive—can be seen as an attempt to manage her own stress and conflicting roles, though it ultimately invalidated Rick’s emotional reality. This left Rick feeling both misunderstood and responsible, and his behavioral issues were likely symptomatic of his efforts to cope with an emotionally untenable situation.

Discussion

This study has been innovative in implementing the use of triadic pictures to generate attachment responses, based on Resnick’s (1993) SAT. The triadic conflict scenario pictures produced significantly higher attachment responses than the original dyadic (SAT) pictures, for children who are triangulated. This finding, coupled with the over-whelming self-blame, behavioural difficulties (John, Peter, Dean, Jack) and anxiety these children exhibit, indicated that they were distressed on a daily basis, demonstrating that the concept of familial conflict is much more arousing than concepts previously deemed as extreme by Resnick (1993), such as mother going into hospital. It was found that most children reported feelings of being caught in the middle, being conscripted to take sides or feeling invisible in the face of the parent’s pre-occupation with their marital problems, supporting past literature (Amato & Afifi, 2006; Grych, Raynor, & Fosco, 2004; Ringer & Crittenden, 2007; Shelton & Harold, 2008).

An interesting corollary was that the children displayed significant awareness of their parents’ feelings and were also confident that the parents were generally aware of their own thoughts and feelings, which has also been observed in a previous study of disadvantaged children in foster care (Dallos, Morgan, & Denman, 2014). It seems that children have not abandoned the idea that their parents understand their needs. However, they displayed less confidence that they would act upon their understanding. In fact, in this study, many of the children felt that they had to take the initiative to mediate and try to resolve the conflicts in the triadic scenarios.

Findings suggest that children may find conflicts between their parents to be extremely anxiety provoking, and the reasons for this may be that the ‘ultimate’ fear of their parents breaking up represents a massive attachment threat, namely, the loss of their family, home, stability and sources of affection. Importantly, some of the children also revealed that they saw their parents as emotionally unavailable and unaware of their needs when the parents were seen to be in conflict with each other. The children at these points felt ‘invisible’ and perhaps felt that they had lost their attachment figures. Alternatively, some of the children did not see their parents as conflictual with each other, but became very anxious that they may have caused conflicts to occur between the parents. Importantly, the pictures invited children to look at situations they may not have actually witnessed; yet, their negative emotional responses were elevated regardless of actual experience, a premise based on Resnick (1993). This suggests that triadic family conflict situations reliably provoke attachment anxieties in children. Importantly, it also indicates that children’s ‘internal working models’ (Bowlby, 1988; Crittenden, 2006) are not predominantly dyadic but triadic. In terms of development processes, the age of these children is early to mid-adolescence. This is an important period of the development of autonomy and moving to significant relationships outside of the family. It is important to understand how general representations about relationships start to develop. Theoretically, the findings suggest that this important transition is influenced by how much the children are emotionally entangled in the parents (triadic) relationships. For example, for Kate, her autonomy may be impeded by her continual concern and anxiety about her parent’s relationship and her role in attempting to mediate their conflicts.

In terms of coping strategies, this study also found that children would generally either avoid or involve in conflict, with some younger children exhibiting more fluid coping strategies and engaging in both. Crittenden, Dallos, Koslowska, and Landini (2014) have similarly suggested that younger children are more likely to alternate between attachment strategies and come to settle on dominant patterns. However, in a more distressed and problematic family situation, such fluidity may become a fragmented or disorganised pattern where fluctuations are unpredictable and dysfunctional. This suggests that children settle on a coping strategy as they get older. This is again supported by findings in the past literature; for example, Lazarus and Folkman (1984) defined coping as engaging or involving with the stressor, and Buchanan et al. (1991) explain that as they get older children become increasingly able to distance themselves from parental dispute; however, older adolescents may also feel more caught, possibly due to increases in social and cognitive maturity. We found some indications of gender differences in that boys appeared to be more inclined to avoid and girls more willing to engage with the picture scenarios, but given the small size of sample this requires further investigation.

A relatively small sample was employed for the quantitative part of the study although the qualitative part is based on a larger sample. Hence, the quantitative findings need to be seen as exploratory and a wider study is required, for example, to explore further the differences between triadic and dyadic scenarios. Our sample was a White population, and further investigation of whether these findings hold for ethnic minority populations is required. Our clinical observations, though, are that our findings also apply to ethnic minority populations. For example, one family who were refugees from Albania showed very marked emotionally anxious responses to triadic situations.

This appeared to be related to their sense of threat and the need for family members to stay very closely connected and supportive of each other. Also, this was a clinical population and we have not conducted this study with non-clinical samples. Previously, we have conducted a study with a non-clinical sample of young adults which revealed that they had experienced conflicting emotional processes in response to the conflicts prior to and during their parents’ divorce. However, they did not report feeling coerced to take sides in disputes between their parents (Abbey & Dallos, 2004).

Clinical implications

In this study, PAE was related to higher scores on SAT pictures rather than triadic pictures, suggesting that the children felt that their parents understood them less when the parents were engaged in conflict situations with each other. This supports other findings that children display distressed or problematic behaviour when ‘caught in the middle’ of their parents’ conflict (Amato & Afifi, 2006; Buehler & Welsh, 2009; Gerard et al., 2006; Ng & Smith, 2006; Shelton & Harold, 2008). Arguably not being able to find feasible ways of resolving the conflicts between their parents triggers symptomatic behaviour in children which is associated with a loss of coherent cognition and RF.

The findings here support our experience of work with children in clinical contexts where children and parents are generally not aware of the processes of triangulation. Even when there is an awareness, the full impact that this has on children’s functioning is typically under-estimated by both the child and the parents. For example, in many cases, we have found that even when a child describes feeling ‘caught in the middle’, ‘made to take sides’ or compelled to intervene in parents’ conflicts, they do not connect this with the onset of their symptoms. Fostering awareness of the impacts of triangulation is therefore seen as an important goal in family therapy. However, it is also a potentially very sensitive and difficult one since parents are likely to feel blamed, guilty and may become defensive. Our clinical experience is that asking parents to think directly about their own families and children can be too threatening and working with the parents to discuss theory own childhood experiences can be more productive, for example, how they experienced conflicts and arguments between their own parents. This can lead into exploring their ‘corrective’ scripts in terms of how they have tried to be different with their own children, for example, to not involve them in their conflicts. Such exploration for a position of positively framing the parents’ intentions can then lead reflections on whether in fact their own children are influenced by triadic processes and their marital issues and conflicts.

Such reflection on their own childhood experiences can help parents to understand how they may also have inadvertently reproduced similar processes of entangling their own children in their marital conflicts. Importantly, many parents hold the erroneous view that since they do not deliberately involve the children in their conflicts and keep their problems from them, that the children are unaware. Reflecting on their own experiences can help parents to understand that despite their best intentions it is very difficult to hide their distresses from their children. At the same time, it is necessary to also discuss that revealing too much of their problems to their children and making them confidants can also serve to entangle and confuse them. The therapeutic process needs to involve a consideration of a balance between being open about the conflicts and how the child might be influenced by them and also of reassuring the child that despite conflicts they need not and will not be expected to take sides.