Can attachment theory help us to understand the reluctance which many therapists feel about closing a practice? The losses at the end of a career could be very significant, but does it make any sense to think of this an attachment wound? The therapist is clearly not losing her own attachment figure as, unless something has gone very wrong, patients do not function as attachment figures for their therapists. What the therapist is losing is her role as an attachment figure; the role which may have given her a sense of purpose, meaning and affirmation over decades and which may have been contributing to her own secure base. As with any loss, our way of managing ourselves will reflect our default attachment pattern and residual insecure traits will determine whether we lean towards a more avoidant/ dismissing or ambivalent/preoccupied pattern.
Basic principles of attachment theory
Bowlby (1969, 1973, 1980) proposed that, in the interests of survival, an infant develops an attachment bond with its primary caregiver. Attachment behaviour includes seeking proximity to a safe person who is viewed as āstronger and/or wiserā (Bowlby, 1977: 203) and protesting at separation from this attachment figure at times of fear, tiredness, illness and other stress. Ainsworth, Blehar, Waters and Wall (1978) observed that infants develop patterns of attachment in response to the caregiving style of their primary carer. These secure, avoidant and ambivalent patterns of relating have been found to be relatively stable through childhood and into adulthood (Fraley and Brumbaugh, 2004), and a further category of disorganised/disoriented was identified by Main and Solomon in 1990. An attachment pattern is the visible reflection of a set of internal working models ā a set of beliefs about the self and the self in relation to others. These are the mechanisms by which early attachment experiences continue to colour development through the whole lifespan (Bretherton and Munholland, 1999).
When parents are able to be sufficiently accessible, reliable and responsive an infant will develop a secure attachment pattern and the capacity to make use of this secure base (Bowlby, 1969, 1977). This securely attached child will able to use the parent to regulate distress: once attachment needs have been met and the child feels soothed, she will be able to launch back into playful exploration of the world. Gradually the secure base offered by the attachment figure becomes internalised so that the child can cope with an appropriate degree of separation. Key aspects of security are described by Fonagy and Target (1996): a secure parent can mirror the childās feelings (but not too closely), contain the childās arousal, make sense of her experience, recognise difficulties and deal with them effectively. In a similar way a good-enough therapist is able to field strong emotions without being overwhelmed and can provide the recognition and containment which supports integration of thinking and feeling in the client.
An infant will develop an avoidant strategy (Ainsworth et al., 1978) when their primary caregiver indicates reluctance to engage with their distress. The child internalises a sense of the other as uncaring and a sense of self as undeserving of love. These children achieve a modicum of proximity by apparently deactivating their attachment system and making few demands on parents, hiding their needs first from their parents and eventually from themselves. In contrast, an ambivalent strategy (Ainsworth et al., 1978) develops when a parent is intermittently available. The child finds that hyper-activation of attachment behaviour may elicit parental care but for these children and adults, even when care is forthcoming, it is difficult to take it in and to feel soothed.
The most troubled individuals will have an attachment pattern with disorganised qualities (Main and Solomon, 1990), and they lack the coherent strategies of those with avoidant and ambivalent patterns. In infancy these children may have experienced the impossible dilemma of feeling both drawn to and repelled by the same person because their primary caretaker has been a source of fear. For these infants whose parents are themselves very troubled, there is no single strategy which will provide a better outcome. A fallback position for these children may be to attempt to control their caregivers as a way to feel safer, and this might be the position they would adopt with a retiring therapist who is threatening abandonment.
Following this understanding of attachment patterns in children, the 1980s brought the application of this thinking to adults. Main, Kaplan and Cassidy (1985) demonstrated patterns in adult discourse that correspond to the attachment relationships which Ainsworth et al. (1978) had identified in children. Main and colleagues identified how in adulthood these patterns reflect particular states of mind in relations to attachment figures, and they named them as autonomous, dismissing and preoccupied. Any one of these three patterns may be overshadowed by unresolved grief if an individual has suffered loss or other trauma without the opportunity for working it through. The principal insecure, but organised, patterns are thus described as avoidant/dismissing or ambivalent/preoccupied, though from now on I will generally use a single word rather than the double-barrelled term.
One of the great contributions of attachment theory is its understanding of the tension we all manage between our need for closeness and for separation. The more secure a person is, the more able they are to manage that tension. For people with a disorganised pattern the oscillation between those two needs will be particularly disorderly. It is not that people with a dismissing style have any less need for closeness or that people with a preoccupied style have less need for autonomy. The difference is in how they manage these needs. At times behaviour will clearly reflect the underlying attachment pattern, at other times interactions are less easy to decode. When faced with a retiring therapist, a patient with either of these patterns might suddenly bolt, the difference between a preoccupied or a dismissing style being felt in the countertransference. For example, if a client with a preoccupied pattern left suddenly the manner of their going might convey more sense of āCome and get meā, and the therapistās feelings in relation to this patient would probably contain more guilt or frustration.
The therapistās own attachment experience
Bowlby proposed that attachment needs remain significant āfrom the cradle to the graveā (1979: 129). The quality of our own internalised secure base will be important throughout life, but in old age, when our actual attachment figures (parents and partners) are more likely to have died, this internal capacity to sooth ourselves could arguably be even more crucial if we are to keep on exploring and engaging with the world. This kind of security would make it more possible for an older therapist to sustain critical and creative reflection on her own work and perhaps make it easier to resolve the dilemmas around retirement. Although I have presented these attachment patterns as though they were discrete categories, their value is not as a system of labels but, in Sladeās terms, to sensitise ourselves āto observing the functioning of the attachment system and to the internal and interpersonal functions of attachment processesā (2004: 269).
While we hope that practising therapists have achieved a good measure of earned security (Main and Goldwyn, 1998), we know anecdotally that we are a profession of wounded healers and there is some evidence that many of us have insecure attachment histories (Sussman, 2007; Adams, 2014). This means that at times of stress our original attachment patterns tend to impact both internally and on the people around us. A therapist who is contemplating retirement is therefore likely to be managing either a tendency to minimise the meaning of the transition, if dismissing is her core pattern, or a pull to hold on longer if she is inclined towards a more preoccupied pattern. Holmes (1997) observes how the fit between attachment patterns in the patient and the therapist can lead to ending a therapy too soon or too late. Perhaps something similar may happen in terms of bringing an end to a whole practice. Where there is residual disorganised attachment, in either client or therapist, there is scope for inversion in the relationship and confusion about who is looking after whom. In the worst cases clients have become carers to ailing therapists.
We may speculate that where there is a residual insecure pattern of avoidance in a retiring therapist, she may find it relatively easy to make a clean break. In this case the risk may be of underestimating and overlooking the impact on clients and minimising her importance in the clientās internal world. For preoccupied clients this could result in their distress being uncontained and the ending being insufficiently processed. For dismissing clients the risk would be of replaying an old relational script in which their pain goes unrecognised by themselves as well as by those around them, sometimes surfacing indirectly in the body.
Where the therapistās residual insecurity is preoccupied then the longing and yearning which is normal in grief may persist: ruminative thoughts and regrets may weigh heavily. These retirees may struggle more than their avoidant peers with ambivalence about taking the step to retire. It is possible that this more anxious style of caregiving may communicate itself unconsciously to the clients: there could be an underestimation of patientsā resilience and consequently an undermining of it, leading to a messy ending. A strength of therapists with this residual pattern is the ability to identify and empathise deeply with clientsā feeling of abandonment; there will be times where this means that the hard-to-reach grief of the dismissing client is affirmed in a way which finally helps them to acknowledge it. Some individuals with strongly ambivalent or disorganised patterns may try to get their attachment needs met through compulsive caregiving. Anecdotal evidence suggests that this type of insecure group, who had an inverted caring relationship with their own parents, may be highly represented among therapists. A possible example of this is given by one of Adamsās interviewees, who describes herself as a parenting child:
my mother needed care, as it were, not my father, but I was definitely a parenting child. Itās a sort of way of life really, and I think I just professionalised it.
(Adams, 2014: 74)
For these therapists, the challenge at the point of retirement would be to adequately process their own anxiety about separation, so that they can more effectively tune in to how each client is taking the news. One of the difficulties for the retiring therapist is that her own feelings about ending may interfere with her capacity for containment. Thus at the point when clients may have extra need of responsive attunement, therapists may struggle to supply it. If the therapist is retiring under pressure of illness she will be managing uncertainty about her own future as well as guilt about decamping. At the same time, negative transference from some clients may increase to unfamiliar levels.
Our own internalised secure base, formed from our attachment history as well as our current attachment network, will be a factor in how we come to the decision about retirement and how we carry it through. The context of our current attachment relationships will also impact. At the point of retirement some will be in satisfying long-term relationships, some in less fulfilling ones and some will be happily or unhappily single or bereaved. It may make sense to think of our work, our profession and our circle of colleagues as jointly providing a secure base through their validation and responsiveness. This constitutes a valuable package which may have compensated for disappointments in our own attachment relationships. It is also interesting to think in terms of our theoretical secure base. What is the impact of our chosen theory? The question of why we are drawn to train and work in one style rather than another seems reflective of our own internal worlds. We might justify our choice in intellectual terms but surely we are drawn to that theory which best helps us manage ourselves. The enmity between different schools suggests that we may then cling to that theory as though to a life raft because naturally we want to believe we are in the right lifeboat. Does our specific choice of theory impact on how we handle ourselves through the retirement? I see my own interest in the subject as arising from my own concerns about attachment and loss which, of course, influenced my original attraction to a training based on Bowlbyās ideas.
As well as loss of what we value in our work, the threat of ageing and dying carried by retirement may also contribute to a testing of our secure base. The principles of attachment theory might suggest that we would respond to old age as we would to other types of loss. We might thus expect those of us with dismissing traits to tend to deny the losses involved in ageing and those with an ambivalent pattern to be more preoccupied with their losses. Studies on this age group are rare and as yet inconclusive (Magai, 2008). What is clear is that the secure group at this age, as in younger years, are better equipped to face losses. They are both more resilient and more able to accept and use help.
