Assessing attachment in families: Beyond the Dyads

INTRODUCTION

Attachment is often thought of a dyadic process involving two people, usually a mother and child, in a relationship. However these two individuals almost always exist in the context of some sort of family, often including a father, or step father and frequently other relatives. This chapter considers issues of the assessment of attachment in such larger units. 

A clinical vignette 

 Rebecca (6) was referred for treatment because her mother Rose was becoming very concerned that she was ‘morbidly’ preoccupied with death and dead things, continually demanded attention from her and asked her endless questions about death. When outside the home she was said to want to pause and explore dead animals on the road, such as hedgehogs, birds or badgers that had been killed by cars. She was also said to be very obsessed with anything on television that contained any mention of death. Her mother was scrutinising closely what she watched and was also overwhelmed by Rebecca constantly asking questions about death.  

On meeting Rose and Rebecca it turned out that Rebecca’s father, who had been a very fit and healthy fire-fighter, had six months previously died very unexpectedly of a heart attack.  Rose and Rebecca had been devastated and Rose’s mother, Aubrey moved in for a while to help Rose to look after Rebecca and her 2 year old brother. Aubrey had suggested that Rose needed to be ‘strong’ for Rebecca and to not show her daughter her intense feelings of sadness and loss. Therefore she had not spent time with Rebecca grieving her father, for example looking at photos, crying with her or compiling mementos. They had also decided that Rebecca should not attend the funeral as she might become too upset.  The consequence of this strategy was that the full focus was on Rebecca’s reactions and this was fuelling an increasing anxiety in her mother and grandmother that she was morbidly pre-occupied with death.

Rather than seeing Rebecca as developing a morbid, potentially psychotic, state an alternative formulation was that the continual focus and scrutiny of her behaviour was serving to exaggerate and distort her interests which, in part, were developmentally normal for a child of her age.  This focus on Rebecca also functioned to help distract Rose and her mother from the distress of the loss but was also blocking their grieving process.  Since Aubrey spent so much time being worried about Rebecca this shifted away any attention from talking with her about her father and her grief.  Also, while not central to the referral, Aubrey’s focus on Rebecca and Rose functioned to divert her attention from the death of her husband (also of course Rose’s father) and facing her own grief.

In attachment terms this was seen to be increasing the dismissive strategy (Type A) (see Chapter 1 in this volume) used by Rose and Aubrey and the pre-occupied/ anxious one used by Rebecca (Type C). It was also important to consider this as a trans-generational process in that Aubrey was responding to her daughter by using an avoidant attachment strategy and encouraging her to also use this strategy with Rebecca. Rose had responded cognitively with an anxious – ambivalent strategy towards Rebecca,   in the extreme focus on her daughter’s interest in death but affectively with an avoidant strategy in not being able to offer her direct comfort about the attachment loss.   

Attachment patterns in families  

This brief clinical scenario points to the importance of assessing attachment in families not just in individuals or dyads and of attempting to do this within a family systemic lens (Marvin and Steward 1990).  Rebecca was indicating an attachment problem (namely the emotional withdrawal of her mother, her ‘secure base’), disguised as a pre-occupation with death that ensured her continually finding a way to gain her mother’s attention and concern. However, because the core attachment component of this behaviour, loss and grief, was forbidden to be openly expressed through a multi-person, trans-generational process a serious potential emotional disorder was being constructed.  

Mary Ainsworth’s (1973)  naturalistic observation of children and their parents and subsequently the development of the Strange Situation protocol led to the view that some mothers were more ‘sensitive’ and attuned to the emotional needs of their infants and consequently fostered the  development of  ‘secure’ attachment representations in their children as opposed  to mothers who produced  ‘insecure’ attachments.  Arguably this line of research has come to dominate attachment theory research with an almost exclusive focus on dyadic relationships.  There have been attempts to include a consideration of attachment patterns with fathers but this likewise has been essentially dyadic.  

A range of questions, however present themselves: staying within a dyadic paradigm how does the combination of the parents’ attachment with the child function? In particular if the parents have differing attachment relationships with the child how does this play out in terms of the effects on the child?   Main et al (1985) referred to ‘multiple models’ in that the child has a complex task of trying to adjust their attachment patterns for each parent. There is also the related question of whether one parent’s strategy with the child comes to dominate. Typically, attachment research appears to assume that the pattern with the mother will trump that with the father.  However, the situation is more complex and we need to ask, for example about what happens when the parents are both together with the child?  How does the child then adjust her responses to her parents in their joint presence and do they adjust theirs?  The situation arguably is more straightforward if the parents have similar patterns with the child but not clear when the parents employ contrasting attachment strategies.

Attachment theory has gone on to develop a wide range of assessments to describe and measure features of attachment, especially narrative and questionnaire approaches to assess individual attachment representations.  This has included procedures to explore attachment behaviours, attachment related cognitions and feelings and also biological features of attachment.  This research emphasis has given attachment theory considerable scientific credibility in contrast to some related approaches broadly derived from psychodynamic theory.  However, this strength has arguably come with limitations regarding the clinical applications of attachment theory and in particular its application to work with families.    

An essential feature of clinical practice is the recognition of the recursive link between assessment, exploration and formulation: The assessment/exploration leads to the accumulation of data which is employed to develop causal formulations regarding how the problems have developed and are being maintained. In work with families it has been recognised that the assessment framework needs to consider complex family process in terms of a circular versus a linear causal relationships:

Marvin and Stewart state: ‘Research on attachment at all ages requires detailed description of how various patterns of attachment behaviours and attachment – caregiving interactions affect further interactions within a family,  rather than just focusing on  what effects family interactions have on subsequent attachment interactions’.  (Marvin and Stewart 1990: 67)

Another clinical example

 In this family a mother and her daughter have an extremely enmeshed and anxious attachment relationships associated with symptoms such as tantrums, psychosomatic problems or eating disorders. The father is emotionally distant or even in conflict with the mother and stays distant to the mother daughter – dyad, perhaps joining with a son in a coalition which is critical of mother and daughter. However, this pattern reinforces the mother and daughter entanglement, since father does not assist mother and also keeps the couple as distant. Rather than starting to resolve their conflicts this organisation aggravates it by escalating the distance between the couple and also through their lack of co-operation about how to manage the child. This further serves to conceal the attachment problems in the couple, as their intimacy decreases the focus on the girl’s problems and the enmeshment between her and her mother in turn increases. 

Such examples suggest that the attachment strategies of each family member and the relational attachment patterns are related in circular, recursive patterns (Marvin and Stewart 1990, Dallos 2006, Byng-Hall 1980).  This raises interesting questions about how we approach the question of assessing attachments in families.

THE ASSESSMENT OF ATTACHMENT IN FAMILIES

Bearing these examples in mind the chapter describes the following approaches to assessing attachment in families:

1 Summative Approaches – combining individual assessments of attachment to provide a matrix of attachment strategies

2 Observational Approaches – mapping of patterns of attachment requests and responses in families

3 Shared family models – exploring shared Internal working models of attachment

4            Clinical case study of family attachment dynamics

1        Summative Approaches: combining individual assessment to provide a matrix of   

           attachment strategies

One approach to considering attachment in families is to explore patterns of dyadic attachment and their implications.  This approach in effect attempts a deductive syntheses.  The parents’ attachment patterns can be assessed, for example using the AAI or other attachment inventories (see Chapters 8 & 9). Likewise it also possible to assess the attachment patterns of the children using observational methods such as the Strange Situation or Q – sort (see Chapter 2), narrative attachment assessments (see Chapter 4) or the School-age Assessment of Attachment – SAA (see Chapter 5).  

The individual and dyadic assessments can then be synthesised to map the overall family attachment structures and consider the implications of the various constellations. This approach is being developed, for example by Crittenden (2006) who starts with such a synthesis to formulate the potential family attachment dynamics, problems and potential intervention strategies.

There have been a number of studies that have looked at the patterns of attachment in families and in particular the discrepancies between the parents’ individual attachment strategies and the dyadic ones with the child (Schindler et al. 2007, Dallos and Smart 2011).  These have also started to document the different challenges such discrepancies pose for the children and what possible symptoms may emerge as solution to the unsolvable dilemmas posed by the contradictory attachment strategies that they demand from the children. For example, Schindler et al (2007) found that where a mother holds a pre-occupied attachment strategy as opposed to the father holding a dismissing one, the child is likely to develop an enmeshed relationship with the mother. He may become overly entangled in meeting her emotional needs, acting as a form of substitute husband while her actual husband withdraws further from her and the child. At adolescence the competing pressures for the child to also form intimate peer relationships and sexual intimacy may come into conflict with the mother’s needs and the child attempts to escape the emotional dilemmas by turning to substance abuse. However, as dependency on substances escalates the child becomes less able to function independently and in turn becomes dependent on the mother for his welfare. This can maintain their mutual emotional dependencies on each other and delays or aborts the leaving home transition, development of peer relationships and leads to the young person becoming ‘stuck’ at home. Use of drugs can function to gain some separation from the mother and also to gain emotional relief from their increasing sense of isolation and failure.

The following permutations are possible and indicate the permutations of parent’s attachment strategies and the potential problems for the child and the responses they may elicit:
The above table summarizes the various permutations possible in terms of the parents/carers combinations of attachment styles and the likely response from the child.  However, the possibilities are even more complex in that gender issues may be relevant,   for example whether the mother as opposed to the father is dismissive, pre-occupied or secure or vice versa is a further relevant factor. Research on triadic family processes suggests that the pattern of an emotionally pre-occupied and enmeshed relationship between mothers and children is more common when the father displays an avoidant pattern and disengages from the family (Du Blois Comptoise and Moss, 2008). Given that in many cultures women still take more of the parenting role than fathers, spend more time with their children, and are expected to be more emotionally available and responsive the likelihood of them being more emotionally enmeshed is greater, This is especially likely since fathers are more likely to be engaged in work outside of the home and to be less emotionally engaged with their children.

The emotional consequences for the child

Where there is consistency between the styles offered by the two parents (carers) the task for the child may be seen as less cognitively complex in terms of differentiating between their behaviours and forming an internal representation of the parenting system.  However, the emotional /attachment task for the child remains potentially demanding:  For parents who are both dismissive (in Type A) the child will have to develop ways of getting her needs met. A range of attachment sub-strategies may be developed, for example the child may become increasingly self- reliant. An alternative is that they may develop a role as the emotional conduit between the parents sometimes this involves attempts to please both parents or become increasingly compliant to gain their affection. A further possibility is role reversal   whereby they come to meet the attachments needs of each parent by caring for them – all Type A strategies. Alternatively children may develop a contrasting strategy to their parents and use exaggerated displays of affect in order to elicit some emotional response from the parents (Type C) (see Chapter 1)

nger, tears and storming out of rooms.  Excessive emotionality shown by the child may help to distract the parents from their own conflicts even at times to the point where the child becomes seen as the ‘problem’ (Byng- Hall,1980)   Alternatively the child may attempt to withdraw from the family or attempt to try to calm the parents, offer reassurance and take on a reversed caring role.

The possible permutations can present further complication for the child when the parents employ contrasting strategies.  It has been assumed that where one parent offers a secure strategy towards the child this can help to protect the child and offset the insecure pattern of the insecure parents. In fact the situation may be more complex.  A study by Schindler and colleagues (2007) indicates that the child may nevertheless be concerned about the well-being of the insecure parent in contrast to more emotionally stable ‘well functioning’ secure parent.  However, the most challenging situation may be where one parent employs a dismissive and the other a pre-occupied pattern in relation to the child.    Main et al (1985) have described how this confronts the child with a complex cognitive and emotional task in learning to differentiate what is an appropriate strategy towards each parent. Especially for the young child this requires sophisticated mental operations which developmentally they do not yet possess  In fact this discrimination may be all the harder since the pre-occupied patterns contain within them  inconsistency  in that the parent may at different times be available, unavailable  or emotionally  intrusive. 

A problem with such a summative approach is that it tends to assume that the individual attachment strategies are fixed and dictate the family interactions rather than consider how the attachment strategies are shaped and maintained by the family dynamics in a recursive process.    

2       Observational Approaches: mapping of patterns of attachment requests and responses   in families 

Attachment theory has employed observational research extensively but mainly, to date, to observe dyadic processes only. Systemic family therapists have the opportunity to observe families interacting and typically use visual recording of sessions. This offers a rich pool of data regarding family dynamics and the interplay of attachment request and responses in families.  Marvin and Stewart (1990) suggested that the structural family typology developed by Minuchin (1974) has significant overlaps with the classificatory system of attachment developed by Ainsworth et al  (1978).

Marvin and Stewart describe three styles of family interaction:

Adaptive:  mutually sensitive, open communication, and supportive while respectful of   developmentally and situationally appropriate autonomy

Disengaged:      avoidant or uninvolved, angry and insensitive

Enmeshed:       over-involved, intrusive, ambivalent, and disrespectful of appropriate autonomy and  

                         boundaries

The Adaptive family pattern can be seen to resemble Ainsworth’s secure classification and likewise the disengaged the avoidant (Type A) and the Enmeshed the anxious – ambivalent attachment (Type C) patterns.  An important question though is whether the attachment classification, initially derived from observations of dyadic interactions (Ainsworth et al, 1978) can be used to identify family patterns.  Related question are whether, for example in an anxious – ambivalent/ enmeshed family attachment dynamic,  all the family members also consistently  show anxious – ambivalent attachment strategies and internal working models ( as assessed by individual focused attachment measures).  This connects with an important finding in attachment theory, namely that different children in a family can have different qualities of attachment to the same figure. Crittenden (2006) has referred to the processes of ‘meshing’, with different but complementary attachment between a parent and child, and ‘matching’, where they display similar patterns. This challenges the basic assumptions of continuity across the generations which had been assumed by early attachment research but has become increasingly questioned (Van IzJendoorn 1995)

Observational studies to support Marvin and Steward’s eminently sensible suggestions are still waiting to be conducted. Family therapists have a rich data base of recordings of family interactions which support the relevance of their framework. However, it is likely that an overall family attachment typology may in many cases be too crude. As we saw in the previous section there may be permutations where all the family members share the same attachment strategies, but this may in fact not be the norm. It is more likely that a systemic framework may be more helpfully employed to consider family  sub-systems, for example that a mother – daughter dyad may be enmeshed/anxious – ambivalent and the father son/s dyad or triad  as disengaged/avoidant (Dallos 2006).  

Crittenden’s (2006) DMM attachment model helps in this context because it suggests that the so called disorganised attachment strategy may more accurately be seen as a mixed pattern with a person attempting to employ a combination of both insecure strategies, either with rapid or more gradual alternations between them.

Triadic Patterns and Triangulation

A move to extend attachment research to study triadic processes seems eminently desirable especially since triads are fundamental to family life.   Though this represents an obvious step for attachment theory in fact it has received relatively scant attention. Studies, such as the Strange Situation have offered important revelations about the links between relational patterns and the development of internal representations in children but with a few exceptions they have stopped short of looking at separation and re-union from a systemic, e.g. triadic perspective. 

Stewart (1977) observed mother, fathers and babies in modified Strange Situation and fathers were found to interact very little with the baby when the mother was in the room but as mother left the room they typically started to interact playfully with the baby as if to distract her from  any distress triggered by mother leaving.  When mother returned they tended to return to activities, such as reading a magazine and left mother to return to contact with the baby. Interestingly the babies displayed brief, almost casual greeting to the returning mother and returned to exploration. This suggested that a co-ordinated process was in place whereby the father anticipated the baby’s distress and helped alleviate if before the baby became over fearful.  Also it was apparent that this was a shared understanding in that the mother knew that the father would respond  and he knew that she knew he knew .

This important study also invites us to consider the family processes when the parents each have a different style of attachment relationships with the child, for example, how the child responds to a shift in attachment provision from an anxiously pre-occupied parent to avoidant/ dismissing from the other.   The child needs to develop different behavioural and emotional strategies in relation to each parent and this in turn requires them to develop a cognitive understanding of   their parents and themselves. Given that they also need their parent’s assistance to develop understanding this is all the more difficult since the parents respectively, either distort cognitive awareness or emotional information. The effective integration of emotional and cognitive representations required for the child to develop coherent attachment strategies may thereby be impeded.

Further our assessment needs to consider situations where the parents are distant from each other and are not meeting each other’s needs for closeness and intimacy.  A child may then step into the role of providing comfort   for one or both of them (Byng- Hall 1980.  This leads us to a further consideration that the child can be seen not only to have an attachment relationship with each parents but also with the relationship between them (see fig.1).    

Fig. 1  About here

This representation suggests that a child may feel secure and confident when in the company of both parents or alternatively anxious and concerned. It is also possible that she may feel secure with either of them alone but anxious when with both. For many families where there has been conflict, separation or violence, security for the child is clearly not going to be the case. However, even where the level of conflict does not constitute formal ‘abuse’ children may still feel extremely distressed by feeling emotionally embroiled in their parent’s conflicts:

 ‘They used to really hurt me because they used to play each other off…And they would be like “ Go on tell me all the bad stuff about the other one”. And I used to sit there and think to myself I am made up of half of each of these people and they hate each other and do they hate me? That used to play on my mind for ages when I was really young and that was the limit of my thought, I didn’t analyse it further’’.     Kate, (in Dallos 2006) 

 Kate’s family took part in a combined observational and clinical study Dallos (2006) which employed a modified version of the AAI to produce a family attachment interview (see appendix 1 – Family Interview)

The interview has also been developed to produce protocols which can be employed for research purposes or for clinical interventions. Appendix 2(a) summarises the sculpt protocol to prompt family enactments relating to changes in the emotional /attachment connections (Dallos and Vetere 2009).     This uses coins or objects, such as stones to represent family members and the distance between them to represent emotional closeness or distance. The sculpt can be employed to consider changes in attachments as a result of critical events, divorce,  losses, traumas and  other life cycle transitions, such as births, separations and exits, such as leaving home. Family members engage in a discussion to  clarify their individual  views, and to explore  disagreements and misunderstandings in an attempt to foster greater  insights into each other’s perspectives and  to form some shared views  of the impacts of their connections and changes. 

 As an example, the sculpt indicated that Kate frequently initiated the discussions and also stepped in to offer reassurance to her mother at points where she repeatedly appeared distressed or tearful. As splits in their relationships were revealed in the sculpt Kate also demonstrated that she was attempting to minimise the conflicts between her parents. She introduced humour but also was the one who took charge in revealing the core dilemmas and conflicts. It also became apparent in the way her parents approached the attachment dilemmas raised by the sculpt that Kate found herself triangulated between one parent who was employing a pre-occupied strategy and the other parent a dismissing one. She would at times switch her strategies when with either of them on her own or alternate between strategies when in the presence of both. The observations from the sculpt supported the clinical evidence that the parent’s relationship was in significant difficulty and the parents had turned to Kate to side with each of them against the other and to meet their own attachment needs. 

3        Shared family models - exploring shared Internal working models of attachment

The attachment responses in families can be seen to be governed by their internal working models. But In addition to each person’s individual representations, families need also to have shared representations including understanding of each other’s working models. This can consist of shared understandings, for example that one of the children is closest to the mother, that father is less strict than mother and is more easy going, that one of the children is most likely to get angry and one most easily upset . This may also include understandings of who provides more of the emotional comforting and who the guidance and ensuring order and discipline in the parents.  Hill et al (2011) similarly have developed a conceptual framework in terms of families being seen to have domains of shared understandings. Two of the central domains relate to that of attachment seeking and responses and discipline and maintenance of family rules.  Mutual understanding in the family is very important regarding these twin areas. Provision of safety involves both the provision of emotional comforting and reassurance when necessary but also guidance and sanctions regarding engaging in behaviour which is dangerous to one member of the family or endangers the others. More broadly this relates to a need to assess the family’s understanding and beliefs – their shared attachment representations:

‘The use of these shared working models probably accounts for  much of the smooth patterning of attachment interactions within the family, in that each individual can anticipate the outcomes of his and other’s behaviour and select his plans accordingly….  ‘This is .... equivalent to Bowlby’s notion of the ‘goal corrected partnership’. (Marvin and Stewart 1990: 65)

The assessment of these shared working models can utilise a family interview protocol (appendix 1 and appendix 2b ) to facilitate an exploration of their working models.  This can be employed to reveal shared constellations of family beliefs regarding attachment -family attachment themes. Dallos (2006) for example, found that a series of families with an anorexic child shared common themes of conflictual relationships, fragile sense of relational reality, triangulation and lack of giving and receiving of comfort. There was an over-riding expectation that family members could not trust the emotions that were being shown, for example that people were being false or deceptive. There was also a shared expectation that comfort would not be provided or expected.  At the same time the parents also held a shared wish – a corrective script – to try and do things better than had been the experience with their own parents. This combined with their relative lack of experience of comfort and provision of secure attachment appeared to leave them vulnerable to a sense of failure and frustration   regarding their good intentions.

Analysis of Family Dynamics / Conversations

It is important to consider that such shared internal models may be subject to processes of continual discussion and negotiation. This process also shapes the family dynamics and in turn the family dynamics serve to reinforce the internal working models.  Families in responding to the interview protocol (Appendix 1) engage in conversational dynamics and relational strategies that are essential to assessment of the nature of family attachment patterns.  The analysis can utilise the attachment discourse markers employed in the AAI (see chapters 8 & 9 and other narrative attachment measures (such as the Stem Stories, or the Separation Anxiety test), for example to reveal specific types of family patterns, such as derogation, idealisation and   dismissing.

Another clinical example: Rick

An example of an analysis of such conversational sequences from a research study (Dallos et al, 2013) is illustrated below. The sequence below shows a family conversation relating to how Rick aged 17 (said to have ADHD and has self-harmed) is seen to be distancing himself emotionally from the family. One of the attachment issues for the family is how close they are and there are some anxieties about this changing in terms of Rick’s perceived withdrawal which appear to arouse his mother’s (Diane) attachment anxieties. He is seen as the ‘cause‘ of their problems  by his step-father and his younger brother (Andrew) whereas his mother continues to be concerned about his well-being and worries that he may be withdrawing from the family too quickly. The short piece of  interactions illustrates how the step-father (Peter)  enacts a more distancing and blaming position regarding the boy whereas his mother appears display her position of being emotionally caught  between Rick and his step-father: 

(Capital letters indicate raised voice and emotional arousal)

Diane (mother) to Rick:   you would I think, if you had your choice, you would have your dinner in your bedroom if you could  

Rick (inaudible, rubbing chin):  Well it’s good to have it with somebody   …….

Peter (step-father): but you NEVER WANT TO DO IT

Diane: Yea you never {laughter} so

Peter: You come down as late as possible gobble {Diane laughter} it and then BUGGER OFF AS QUICKLY AS POSSIBLE

Diane: a he he he he he  {Diane smiling and looking at both her sons in turn}

Peter:  turn around and say, yeah you can sit there and say it’s a really good thing that we have dinner together as a family BUT YOU’RE LYING THROUGH YOUR TEETH CAUSE ACTUALLY

Diane:  he he he he he  

Rick {quietly}:  Yeah I know it’s a good thing

Peter: But you object 

Rick: I object but 

Peter:  ok that’s honest that’s honest   {raised palms – accepting} 

Peter,.Rick’s step-father’s   attachment style in the family is typically to be dismissive and highly intellectually analytical (Type A – also indicated in his AAI) but here he displays a more angry response to which Rick’s   mother responds with defusing and laughter. Rick’s  mother’s lack of support looks towards him and laughter may indicate that she is anxious and wants Rick  to change his story or apologise.  Rick can be seen to respond to this twin communication from his parents of anxiety and anger to back off and negate his initial statement:  Rick:  {quietly} Yeah I know it’s a good thing.  This can be seen as an example of a pattern of disoriented and mixed attachment which was revealed in his individual AAI (Dallos et al, 2013).  Dianne appeared to display both A and C (also revealed in her AAI)I in  attempting to defuse (A) these interactions and also at times was emotionally enmeshed (C) with Rick. Peter would withdraw from such occasional emotional expression into a rational/analytic and emotionally distant position from the children (A).

Analysis of such conversational patterns indicates that family members both shape the dynamics by their attachment strategies but also that their strategies are in turn shaped and, in some cases exaggerated by the family dynamics. In effect attachment strategies and family dynamics are recursively maintained by such continual moment-to-moment relational dynamics. However, there is also scope for change and re-organisation so for example, in the conversations with Rick’s family new insights started to emerge and the parents made some new connections between their own childhood attachment histories and their current relationship dynamics.  These were discussed in the family session and in individual session with each family member. Analysis of family conversational process to reveal attachment processes in families is relatively under- explored but might offer rich insights into how attachment relationships both guide and are shaped by conversational processes in families.

4                     Clinical Case Studies

This chapter started with a clinical case study and arguably these best illustrate the interplay of attachment processes and family dynamics. The involvement with a family in the therapeutic process can be considered as a form of ‘action research’ which helps to  reveal the complexities of the family’s attachment patterns and processes by which the family responds to attempts to intervene and change their way of functioning.  Clinical work and research can complement each other and the family interview protocol, and the formats for exploration described earlier can be woven into clinical work to assist in  developing a systematic assessment, formulation and directions for intervention. Rather than being seen as separate processes the clinical interventions can be seen as forming a recursive loop with assessment in that they provide further data in terms of how the family members respond to intervention strategies (Dallos 2006).

Clinical studies also offer an extremely important additional perspective in that they permit the important focus on change and transitions. Attachment theory is essentially a theory of danger regarding how families mobilise their emotional resources to deal with stress, anxiety, distress and danger. Attachment and family systems theory share this emphasis, for example by recognising that many serious mental health difficulties are associated with difficulties in negotiating life cycle transitions. In the opening case we saw the attachment dilemmas presented by the unexpected death of the father in a family. A typical scenario for contemporary families is divorce and separation and this involves not only the potential loss of one or other attachment figure but also of the child’s relationship with their parent’s relationship.  Transitions, such as a birth, a child starting school, leaving home, divorce or a death in the family require families to engage in practical and emotional re- organisations. They often involve periods of emotional upheaval and sometimes conflicts and distress. Importantly, they also involve change in the attachment relationships and these can be opportunities for growth as well as potential for problems. It is important in assessing attachment in families to assess the changes through significant transitions and also to consider what these   trigger in terms of the activation of the attachment systems and responses. The sculpt formats in appendix 2 can be usefully employed to assess changes in relationships and attachment responses over times through transitional points.  Also family genograms can be employed to further explore transitions and trans- generational patterns of dealing with these changes and demands.

DISCUSSION 

The chapter has looked at four main approaches to assessing attachments in families. What they have in common is the idea that attachments in families are complex interactional and fluid processes.  It may be possible to consider a family’s overall attachment orientation as dismissing, secure or pre-occupied  but this needs to be held as a very broad overview which needs to incorporate how sub-systems within the family are organised and how their attachment processes relate to the wider family pattern.  The importance of attempting this task of moving from a dyadic to a triadic and wider focus is arguably now becoming an important issue for the progress of attachment theory. It is becoming increasingly clear that children can have multiple attachment strategies with different parents and a simple dyadic framework struggles to explain why and how this occurs. 

Adopting a family framework has considerable clinical implications: In terms of clinical formulation a key  starting point for a family attachment assessment will usually be the couple  relationship ( in some cases one parent and a carer or relative other than the other parent). This relationship shapes the attachment context into which the children are born and the nature of subsequent sub-systems. Regarding clinical interventions – it is suggested that the processes of assessment, for example the family sculpting and format for exploring comfort can also function to promote changes in the family’s understandings. Importantly both also promote family interaction (enactment) which also facilitates changes at the procedural and emotional representational levels.  As one set of parents said;

‘We were more able to explore and  discuss  with you areas of problems and difficult emotions without  becoming anxious  and cutting of  and retreating into ourselves  as we would have at home’

This also raises the important point that assessment of family attachment can, and should be regarded as a collaborative process of co-exploration with family members. 

REFERENCES

Ainsworth ( 1973) The development of infant-mother attachment. In. B.M. Caldwell and H. Riccuiti (Eds.) Review of Child Development research (Vol.3)  Chicago: Chicago University Press 

Ainsworth, M.D., Blehar, M.C., Waters, E. and Wall, S. (1978)  Patterns of attachment: A psychological study of the  strange situation. Hillside, NK: Lawrence Erlbaum

Byng-Hall, J. (1980) ‘ The symptom bearer as marital distance regulator: clinical implications’ .  Family Process, 19:355-65

Crittenden, P.M. (2006). A dynamic maturational model of attachment. Australia and New Zealand Journal of Family Therapy, 27(2), 105-115

Crittenden, P.M., & Dallos, R. (2009). All in the family. Clinical Child Psychology and Psychiatry, 14: 387-407.

Dallos, R. ( 2006) Attachment Narrative Therapy. Maidenhead: Open  University Press

Dallos, R and Vetere, A (2009) Systemic Therapy and Attachment Narratives: Applications across diverse settings. London: Routledge

Dallos, R.,  Denman,K., Stedman’s., & Smart.C. ( 2012). The Construction of ADHD:   Family Dynamics, Conversations and Attachment Patterns. Human Systems. 23(1): 

 Dallos, R. and Denford, S. (2008) A qualitative exploration of relationship and attachment themes in families with an eating disorder. Clinical Child Psychology and Psychiatry. 13: 305-322

Dallos, R., & Smart, C. (2011). An exploration of family dynamics and attachment strategies in a family with ADHD/conduct problems.  Clinical Child Psychology and Psychiatry, 16(3), 535-555

Dubois-Comtois, Karine and Moss, Ellen(2008) 'Beyond the dyad: do family interactions influence children's attachment representations in middle childhood?', Attachment & Human Development, 10(4): 415 — 431

Hill, J., Wren, B., Alderton, J., Burck, C., Kennedy, E., Senior, R., Aslam, N., & Broydon, N. (2011). The application of a domains-based analysis to family processes: implications for assessment and therapy. Journal of Family Therapy, DOI: 10.1111/j.1467-6427.2011.00568.x

Main, K., Kaplan, N. and Cassidy, J. (1985) Security in infancy, childhood, and adulthood: a move to the level of representation. In: I. Bretherton and E. Water (eds), Monographs of the Society for Research and Child Development, Serial No. 209.50: Nos1-2. Chicago: the University of Chicago Press

Marvin,R. S. and Stewart, R.B. ( 1990) A family systems framework for the study of attachment. In:  M.T. Geenberg, D Cicchetti and M. Cummings ( Eds.) Attachment in the Preschool Years, London: University of Chicago Press

Minuchin, S. (1974) Families and Family Therapy.   Harvard: Harvard Univ.Press

Schindler, A , Thomasius, R , Sack, P’, Gemeinhardt, B. and Küstner,U.(2007) 'Insecure family bases and adolescent drug abuse: A new approach to family patterns of attachment', Attachment & Human Development, 9(2): 111 — 126

Stewart,R. ( 1977) Parent – child interaction in a quasi – natural setting. Unpublished master’s thesis, Pennsylvania  State University

van Ijzendoorn, M. H. (1995). Adult attachment representations, parental responsiveness, and infant attachment: A meta-analysis on the predictive validity of the Adult Attachment Interview. Psychological Bulletin, 117:387 – 403.

Appendix 1:  Family Interview/ Enactment protocol – eating disorders 

Collaborative research – what would they like to know about other families with similar problems?  What questions would they like us to ask? What would they like us to share with them at the end of the research?

Describe themselves as a family – how are they different/similar to other families? Similarities and differences between family members, nature of the relationships?

Genogram  -  3 generations , kinship, friends– losses , exits and entrances

Life cycle/closeness sculpt (use coins, buttons etc.) - current 2 generations (prompt – prior to , during and after the onset of the eating disorder) 

Emotional atmosphere in the family – how feelings are dealt with, expressed,  perceived similarities to other families – more or less expressive, restrained etc.  

If someone is upset:

how does each parent comfort a child

how does a child comfort a parent

how does a sibling comfort another sibling

History of eating problems in the family across the generations. How do they explain these – heredity , learning,  traumas etc ?

What were meal times like in mother’s family and dad’s family, Describe what typically happened, how people felt, habits etc.

APPENDIX 2:   Formats for Exploring Attachment Dynamics

(a)Sculpting with family members, sculpting with objects (coins, buttons, stones, figures  etc ) …

 Aim is to explore their understandings but also to prompt an enactment of how they communicate  regarding attachment needs

 PROMPTS  (can use direct or circular questions…) 

 Map the current attachment patterns, relationships – who looks after who etc.

How does it feel to be at the centre, on the edge, between your parents?

Now that you and your brother are closer how does that make you feel? 

If you were to get closer to your mother what would that be like?

How do you think your sister feels being that distant from your father? 

(b) Exploring Beliefs about Comforting

The main aim is to prompt an exploration of beliefs but the activity is also intended to prompt and enactment to provide an indication of ‘how ‘ the family members communicate regarding this core attachment topic

When you were ill,  upset or frightened as a child – what happened? 

How did you get to feel better? Who helped you to feel better? How did they do this?

What have you learnt from this for your own family?

What do you want to do the same?

What do you want to do differently?

How do people comfort each other in your own family/relationship?

How do you comfort your children?

How do they comfort you?

What do you want your children to learn about comforting?

Can be held as a family or couple interview or as a one-to-one conversation

Table 1

Attachment styles parents / caregivers and likely responses from the child

MATCHED	CONTRASTING
Parent                    Child  Secure                 secure Secure attachment	Parent                    Child  Secure –                    dismissive Switch strategies possibly cut off from or try to help/ make emotional contact with  the insecure parent
Dismissive – dismissive How to gain attachment responses, get needs met, withdraw or become bridge between parents – emotional conduit	Secure –                    pre –occupied Switch strategies possibly cut off from or try to help/calm  the insecure parent
Pre-occupied – pre-occupied How to gain individuation , emotional space,  try to withdraw or become entangled in continuing emotional arousal between parents	Pre-occupied – dismissive Switch between strategies neither of which are unproblematic and are contrasting