Attachment Theory and psychoanalysis
[The] early formulations of psychoanalytic theory were strongly influenced by the physiology of the day … cast in terms of the individual organism, its energies and drives, with only marginal reference to relationships. Yet, by contrast, the principal feature of the innovative technique for treating patients that Freud introduced is to focus attention on the relationships patients make with their therapist. From the start, therefore, there was a yawning gulf between the phenomena with which the therapist was confronted, and the theory that had been advanced to account for them.
(Bowlby, 1990)
Bowlby was primarily a theorist rather than a therapist. Although trained psychoanalytically, and active in the Psychoanalytical Society from the late 1930s until the late 1950s, he saw himself mainly as a researcher and administrator. Case reports and clinical illustrations are to be found throughout his writings, but, with the exception of his earliest papers, these are almost all based on the work of colleagues or on published articles by other authors. Dreams are nowhere to be found in his work, and he is, for the most part, concerned with observable behaviour rather than the inner world. Nevertheless, Attachment Theory is a child of psychoanalysis and has much to contribute to the theory and practice of the various modalities of psychotherapy, incuding psychoanalysis.
Towards the end of his career Bowlby (1991) wrote, ‘my theoretical work has always been directed primarily to my colleagues in the International [Psychoanalytical] Association’. The aim of this chapter is to consider in some detail the relationship between Attachment Theory, psychoanalysis and contemporary psychoanalytic psychotherapy.
Bowlby's reservations about psychoanalysis come under four main headings: neglect of real experience and environmental influence in favour of overemphasis on endogenous fantasy; an atmosphere of dogmatism inimical to scientific enquiry; outmoded metapsychology; lack of experimental observation to underpin its unbridled theorising. All of these objections may seem to the contemporary observer somewhat overstated, and we must first place them in an historical context.
The atmosphere in the British Psychoanalytical Society when Bowlby started training in the mid–1930s was one of ferment and controversy. The heady excitement of a new science of mind that went straight to the heart of twentieth century men's and women's discontents, also generated a hotbed of intrigue, back-biting, gossip and jockeying for position. The climate resembled less that of a scientific society than of a family in which a patriarch was nearing the end of his life with the terms of his inheritance still undecided.
Ernest Jones had invited Melanie Klein to practise in London and had entrusted her with his own two children for child analysis based on her new technique of play therapy. Until the advent of Klein, the focus of analytical work was predominantly the Oedipus complex. She insisted on the importance and the analysability of much earlier stages of development, and in particular on the phantasies and anxieties of the infant in its first two years of life. As we saw in Chapter 1, Freud tended to regard Klein's views with some disfavour (Steiner, 1985), especially as her ideas about child analysis differed from those of his daughter Anna, who saw splitting and other primitive defence mechanisms proposed by Klein – such as projective identification – as belonging to a much later stage of development.
At the time of the arrival of Sigmund and Anna Freud in London in the late 1930s, the British Society comprised a group of talented and intelligent people, including James Strachey, Edward Glover, Sylvia Payne, Melanie Klein (Bowlby's supervisor) and Joan Riviere (Bowlby's analyst). A leadership struggle broke out with an increasing schism between the Kleinian camp, who were accused of dogmatism and attempts to win students exclusively onto their side, and the more orthodox Freudians, represented by Anna and her followers, together with a third group, who remained non-aligned. Eventually, in 1944, a compromise was reached with the ‘gentleman's agreement’ (in fact made between three women, Klein, Payne and Anna Freud) between the parties to form separate ‘streams’ of training and scientific discussion, while remaining united within one society.
Although Bowlby's organisational and intellectual talents were recognised early on – he was appointed Training Secretary of the Society in 1944 (Melanie Klein opposed this on the grounds that he was not a Training Analyst) – he was even then at variance with the mainstream of the analytic milieu. His strong commitment to the scientific method, his quintessential Britishness and reserve, his decision to work in child guidance clinics rather than in private practice, set him apart. These very qualities, as well as the fact that he was the son of a famous surgeon, may also have given him credibility with the medical establishment. He was later able to use this to popularise psychoanalytic ideas about the importance of infancy and the mother–child relationship as determinants of later mental health.
From early in his psychoanalytic career Bowlby had had misgivings about the way in which analysts downplayed the importance of the environment in the origins of psychological disturbance. Although Freud has been accused of a deliberate retreat from his original hypotheses about the adverse effects of childhood seduction (Masson, 1985), there is no doubt that he continued to believe that childhood trauma was important. As the pioneer of a new ‘science’, however, Freud emphasised the primacy of the inner world as the proper domain of psychoanalytic discourse, and this lead was certainly adhered to by his followers. Bowlby writes:
During … 1936–39 I was slowly waking up to the fact that my ideas were developing in a direction very different from those that were accepted truths in the British Psychoanalytical Society … under the influence of Ernest Jones and Melanie Klein it was held that an analyst should concern himself only with the patient's internal world and that to give attention to his real life experiences could only divert attention from what really matters. My experiences in the Child Guidance Clinic … were leading me to an opposite conclusion … that one can only understand a person's internal world if one can see how [it] has come to be constructed from the real-life events to which he has been exposed.
(Bowlby, 1991; Rayner, 1992)
A marked copy of a paper by Bowlby's analyst Joan Riviere contains the following passage:
Psychoanalysis is Freud's discovery of what goes on in the imagination … it has no concern with anything else, it is not concerned with the real world … it is concerned simply and solely with the imaginings of the childish mind.
(Quoted in Rayner, 1992)
In the margin Bowlby has pencilled ‘Role of the environment = zero’ (Rayner, 1992). Bowlby was particularly distressed when the mother of his first training patient in child analysis, a hyperactive little boy of three, was admitted to mental hospital:
When I reported this to Melanie Klein [his supervisor], however, her only concern seemed to be that, since it was no longer possible for me to continue the boy's analysis, another patient must be found for me. The probability that the boy's behaviour was a reaction to the way his mother treated him seemed altogether to escape her.
(Bowlby, 1991)
Bowlby consistently stressed the range of environmental traumata to which a developing child can be exposed: actual separations and disruptions in care; threats of separation or suicide by parents; being unwanted, or the ‘wrong’ sex; suppression of the true facts about parentage (for example, if the grandfather or uncle is the true father, or the sister the true mother); role reversal and the ‘parentification’ of children. His views have been thoroughly vindicated by the disclosure of the extent of physical and sexual abuse of children. There is, as we have seen, also strong evidence that more subtle forms of environmental failure such as parental unresponsiveness and mis-attunement underlie childhood and probably adult psychopathology.
These findings suggest that in characterising ‘the Kleinians’ (to homogenise what in fact is a group with a number of different voices) as uninterested in the environment and exclusively concerned with phantasy, Bowlby was driven by polemic, and the need for strawman-ism. The principal difference between Bowlby's account and more conventional psychoanalytic models, is that Attachment Theory postulates a clear picture of a variety of possible psychological developmental pathways, some healthy, some less so. Which of these pathways children find themselves on depends largely, but not exclusively, on the nurturing environment in which they find themselves. The Kleinian account is a phenomenological description of mental states found in adult patients – buttressed by some play therapy material – particularly those with borderline pathology; from this is inferred what may have gone on in the minds of infants and small children. Bowlby and the post-Bowlbians offer the outlines of an explanation of how such pathological states come about.
If there are such differing developmental pathways, a causal model is needed to account for one individual ends up a psychologically healthy states, while others succumb to misery and depression. Here too there is a degree of convergence between the attachment and neo-Kleinian approaches. Bion (1978) saw the capacity of the mother to hold and ‘detoxify’ her infants' negative affect, and eventually to put unhappiness and mental pain into words as a vital stepping stone towards the growing child's psychological health, and of course his mission was to argue that that is what analysts do too. The emphasis in Attachment Theory on parental sensitivity and appropriate responsiveness, while more ‘behavioural’ and less quasi-philosophical than Bion's account, is clearly compatible with his model, especially now that the capacity to accept and process negative affect is seen as a key quality in security-providing parents (Strathearn et al., 2009).
A second area of difficulty about psychoanalysis for Bowlby was its atmosphere of dogmatism and authoritarianism. Peterfreund (1983), who is approvingly cited by Bowlby in several places, decries what he calls the ‘stereotyped’, dogmatic, ‘algorithmic’ approach of traditional psychoanalytic formulation and interpretation. He compares this with the ‘heuristic’ approach which he and Bowlby advocate, in which patient and therapist find things out for themselves, rather than imitating Talmudic scholars burrowing in the obscure texts of the psychoanalytic testament. At its worst, and especially in the era in which Bowlby received his psychoanalytic training, psychoanalysis can degenerate into a mouthing of clichéd formulas by an omniscient analyst who, faced with the pain and complexity of suffering, offers some certainty, however ill-founded, to a confused patient who has no choice but to grasp at straws. The relentless interpretation of the transference may hypnotically open the patient up to layers of regression and dependency which make such interpretations self-fulfilling prophecies. There has been a move towards a much more negotiated and relational approach to interpretation in contemporary psychoanalysis (see Casement, 1985; Wright, 2010; Safran, 2012b), in which Keats's ‘negative capability’ – the capacity to allow oneself to be ‘in uncertainties, mysteries, doubts, without any irritable reaching after fact and reason’ – is valued, and indeed is seen as the hallmark of the ‘depressive position’ with its emphasis on compromise and reconciliation rather than splitting and false certainties.
Bowlby wrote that:
I was dissatisfied with much of the [psychoanalytic] theory … being a somewhat arrogant young man … I was in no mood to accept dogmatic teaching. My analyst was not altogether happy with my critical attitude and complained on one occasion that I would take nothing on trust and was trying to think everything out from scratch, which I was certainly committed to doing.
(Bowlby, 1991)
Bowlby's re-thinking of psychoanalytic metapsychology and terminology has been discussed in the preceding chapters. Attachment is today perhaps best seen as a distinct motivational system – which includes drive, affect, cognition and behaviour – that parallels and complements other affective systems, which include sexuality, aggression, and exploration (see Panksepp, 2010). The main differences between classical Freudian theory, the Object-Relations Theory of Klein, Fairbairn and Winnicott, and Attachment Theory are summarised in Table 6.1.
The basic attachment credo vis-à-vis psychoanalysis can be summarised as follows. For Bowlby the important issue is not, as the orthodox Freudians thought, sex, but security. Attachment is primary, not a derivative of orality. The organism is not an isolated drive-driven creature in search of an object on whom to discharge his accumulated tension, but a whole person relating to other whole persons. Homeostatic and other cybernetic control systems govern his behaviour, just as they do that of other mammals. His relationship to the world is determined not just by unconscious phantasy but also by internal working models which include affective, cognitive and behavioural components, whose lineaments reflect the real-life experiences of their possessors. Aggression is a response to frustration and loss, not an intrinsic property of an individual dominated by the death instinct.
Bowlby's fourth cavil at psychoanalysis was its neglect of direct observation of normal and abnormal children. He felt that reconstructions based on childhood recollections of disturbed patients, while valuable in themselves, did not qualify as scientific accounts of what really goes on in real families. He therefore set out to study systematically the effects of separating infants and children from their parents; it was on the basis of those findings that Attachment Theory was born.
To continue with our historical account, Bowlby was of course not alone in his dissatisfaction with the state of psychoanalysis in the 1940s and early fifties. During the post-war period several divergent responses can be found within psychoanalysis and psychotherapy in response to the difficulties – dogmatism, obsolete meta-psychology and anti-empiricism – with which Bowlby was struggling. The first was the development of Object-Relations Theory, epitomised by the work of Winnicott, Fairbairn and Balint, all of whom were influenced by the Kleinian emphasis on the early infant- mother relationship, but, taking the decisive step of abandoning drive theory altogether, posited relationships as primary. Mahler et al.'s (1975) direct observations of mothers and infants from a psychoanalytic perspective combined object relations with a degree of empiricism.
A quite different tack was to reject the pseudo-scientific determinism of classical Freudianism altogether, seeing psychoanalysis as a hermeneutic discipline concerned with meanings rather than mechanisms and emphasising the importance of the creative imagination and spontaneity as the wellspring of the psychoanalytic process (Rycroft, 1985). Meanwhile, neo-Kleinian developments concentrated on delving deeper and deeper into the mysteries of the infant-mother relationship in the early stages of life and attempting to relate these to the phenomenology of psychosis (Bion, 1978). Finally, there were moves away from the psychoanalytic paradigm altogether, adopting either a family systems approach derived from cybernetics and anthropology (Bateson, 1973), or a ‘cognitive’ approach, based on Piaget and Kelly, in which the logical operations of the mind and the way in which they are organised hierarchically form the basis of psychotherapeutic theory and practice (Beck et al., 1979). The latter tack has of course become the dominant psychotherapeutic paradigm through the rise and rise of the Cognitive Behavioural Therapy movement, and its endorsement by UK Government funding.
The Bowlby-Ainsworth approach meant that Attachment Theory and research was underpinned by observable behaviours. Attachment, whether secure or insecure, avoidant or ambivalent, can be observed, rated, measured, and correlated – whether in the Strange Situation, using picture or story completion methods, or through the Adult Attachment Interview. By basing his ideas on ethology, Bowlby sidestepped the dehumanisation and absurdities of stimulus-response behaviourism, while remaining within the framework of conventional science. Winnicott, an outstanding clinician but an elusive theorist, was wrestling with the same problems but from the perspective of the inner world, developing in his idiosyncratic but highly original way a language of experience directly applicable to the therapeutic situation.
Winnicott and Bowlby had much in common. Both were very ‘English’ in their background and outlook, in contrast to the European/Jewish or in Jones’ and Fairbairn's case, Celtic atmosphere of the Psychoanalytical Society. Both had had a Cambridge scientific education and were deeply influenced by Darwin. They shared an analyst, Joan Riviere, who, despite her later Kleinian orthodoxy, was firmly interpersonal in her philosophy. With an echo of John Donne, she wrote:
There is no such thing as a single human being, pure and simple, unmixed with other human beings. Each personality is a world in himself, a company of many. That self … is a composite structure … formed out of countless never-ending influences and exchanges between ourselves and others. These other persons are in fact therefore part of ourselves … we are members of one another.
(Riviere, 1927; reprinted 1955)
Winnicott (1965) was therefore paraphrasing Riviere in his famous dictum, ‘there is no such thing as an infant … wherever one finds an infant one finds maternal care and without maternal care there would be no infant’. Like Bowlby, Winnicott was primarily concerned with the welfare of children, and wrote to an American enquirer about his wartime experiences:
I became involved with the failure of the evacuation scheme and could therefore no longer avoid the subject of the antisocial tendency. Eventually I became interested in the etiology of delinquency and therefore joined up quite naturally with John Bowlby who was at that time starting up his work based on the relationship that he observed between delinquency and periods of separation at significant times in the child's early years.
(Rodman, 1987)
When Winnicott was later offered the Presidency of the Psychoanalytical Society he accepted, on condition that he have a deputy who would take care of the detailed administrative work. The ever-efficient Bowlby was an obvious choice. They make sparse but polite references to each other's work in their writings. There are many similarities between their theoretical viewpoints, despite the radically different language which each uses. Rycroft's (1985) remark that ‘I've always had a phantasy that Bowlby and I were burrowing the same tunnel, but that we started at opposite ends’, would be equally true of Bowlby and Winnicott.
Winnicott and Bowlby's responses to the Kleinian domination of the Psychoanalytical Society can be seen in terms of avoidant and ambivalent attachment. Bowlby, in an avoidant way, distanced himself, expressing neither warmth nor anger, but having little to do with the Society after the 1960s. Winnicott clung perhaps more ambivalently to his alma mater, and, in his theory of healthy hate, emphasised how identity can be forged through opposition and reaction.
Bowlby and Winnicott's overall view of the infant-mother relationship, and what may go wrong with it, is very similar. Winnicott postulates a ‘holding environment’ provided by the mother, in which, on the basis of her ‘primary maternal preoccupation’, she can empathise with the needs and desires of the growing child. The main job of the holding environment is, like attachment, protection, although, in contrast to Bowlby, Winnicott describes this in existential rather than ethological terms: ‘The holding environment … has as its main function the reduction to a minimum the impingement to which the infant must react with resultant annihilation of personal being’ (Winnicott, 1965). Winnicott sees ‘handling’ and ‘general management’, equivalent to the Bowlbian concept of maternal responsiveness, as the framework within which needs can be met. The mother's actual physical holding and handling are primary:
The main thing is the physical holding and this is the basis of all the more complex aspects of holding and of environmental provision in general…. The basis for instinctual satisfaction and for object relationships is the handling and general management and care of the infant, which is only too easily taken for granted when all goes well.
(Winnicott, 1965)
‘Good-enough’ holding leads to integration of the infant personality, to a ‘continuity of going-on-being’, which prefigures Stern's (1985) idea of a ‘line of continuity’ that is the germ of the sense of coherent self. Where there is such continuity the growing child can cope with temporary separations without resorting to maladaptive defences. Like Bowlby, Winnicott sees the seeds of pathology in failures of the holding environment. Separations may provide the nucleus of later delinquency:
Separation of a one or two year old from the mother produces a state which may appear later as an anti-social tendency. When the child tries to reach back over the gap [that is, created by the separation] this is called stealing.
(Winnicott, 1965)
Although Bowlby and Winnicott are saying something very similar about juvenile theft there is a subtle difference in their language and focus. For Bowlby theft is a sociological phenomenon, which can be accounted for by the disrupted lives and maternal separations of the thieves' early childhood. Winnicott is reaching towards an understanding of the symbolism of the act of theft itself. He is suggesting that the stolen object stands in for the missing mother which the youth is using to bridge the emotional gap left by her absence. Bowlby is reaching for explanation, Winnicott for meaning. Both, incidentally, tend to ignore other possible aspects of the problem: Bowlby looks exclusively at the childhood experiences of his thieves, and ignores contemporary influences such as housing and unemployment, while Winnicott leaves little room for the many other possible symbolic meanings that an act of theft might represent.
Winnicott goes on to describe how the good mother empathically understands what stage the child's object constancy has reached and so knows how to handle separations: ‘She knows she must not leave her child for more minutes, hours, days than the child is able to keep the idea of her alive and friendly’ (Winnicott, 1965). If this is unavoidable she will have to resort to therapeutic ‘spoiling’:
‘If she knows she must be away too long she will have to change from a mother into a therapist in order to turn the child back into a state in which he takes the mother for granted again.’
(Winnicott, 1965)
Like Winnicott, Bowlby is insistent in his opposition to the notion, prevalent in the 1950s and today making something of a comeback, that children can be ‘spoilt’ by too much love. He reminds therapists who are working with adults who weep and cling: ‘it is perhaps too often forgotten by clinicians that many children when they become distressed and weepy and are looking for comfort are shooed off as intolerable little cry-babies’ (Bowlby, 1988a). Winnicott visualises ‘two mothers’ in the early months of life. The first protects the child from ‘impingement’ and acts as an ‘auxiliary ego’ which enables him gradually to build up his own autonomous ego. He calls this the ‘environment mother’ who offers ‘affection and sensuous coexistence’. Within the ambiance created by the environment mother the child then relates to the ‘object mother’ who can be sucked and bitten, loved and hated. Her response will have far-reaching consequences: over-intrusiveness can, in a seductive way, be as traumatic as neglect, and both can lead to defensive moves such as ‘self-holding’, disintegration and the development of a false self.
For Bowlby there are also two mothers. The first is equivalent to Winnicott's ‘environment mother’, the provider of the secure base. The second mother is the companion with whom the child, once a secure base has been established, engages in exploratory play. This ‘companion mother’ is different from Winnicott's ‘object mother’ with whom the child engages in orgiastic play. Bowlby seems less interested in orgasmic activities, although the sexual foreplay of trusting adults can be seen as a form of mutual exploration (analogous to the sensuous intimacy of mother and child), which enables a greater build-up of intense pleasure than orgasm not preceded by exploration (see Holmes, 2010).
Winnicott's sophisticated theory of the origins of play starts from the empathic responsiveness of the mother. This helps to create a necessary ‘illusion of omnipotence’ in the infant so that, as a wish begins to form in the child's mind, so the mother is there to answer it: just as the baby begins to feel hungry, the breast appears, as though by magic. In Winnicott's model, in this transitional zone of overlapping phantasy lie the origins of playfulness, creativity and, ultimately, culture. Bowlby's ‘companion mother’ can be seen in similar, if less mystical terms. ‘Post-Bowlbians’ emphasise the collaborative nature of exploration, the ‘zone of proximal development’ (Vygotsky, 1962), where parent and child interact and in which learning takes place (see Holmes, 2010; Feeney and van Vleet, 2010). Stern (1985) sees the task of the mother as maintaining an internal ‘line of continuity’ for the child, so that she will unobtrusively stimulate the child when his imagination begins to flag, back off when he is playing happily, and dampen his excitement when it threatens to get out of hand.
The differing languages of Winnicott and Bowlby reflect the varying foci of their thought. For the Bowlbian, child play and exploration take place ‘out there’ in the world, while Winnicott's child is concerned with inner exploration, with the world of the imagination ‘in here’. In reality child of course the child engages in both at the same time. The toddler building and breaking and building again his tower of bricks is simultaneously acquiring Piagetian knowledge of physics – the properties of materials, the mathematics of cubes, the nature of gravity – and exploring potency and castration, the balance between agency and passivity, when to do things for oneself and when to recruit help, and the interplay of destruction and reparation in the inner world.
Klein's depressive position becomes in Winnicottian terminology the ‘stage of concern’. Here the ‘environment mother’ and the ‘object mother’ come together as one person. The environment is necessarily defective: the mother cannot always be perfectly responsive: there will be gaps and breaks and discontinuities of care. The child responds with aggression and rage directed at the ‘object mother’: she survives the attacks and continues to love her child, and the balance is restored. He now realises that the mother who lets him down is also the one he loves. Clouds of guilt and anxiety appear on his horizon, but also the seeds of gratitude and reparation. For Bowlby, too, the good mother can withstand her child's aggressive onslaughts, and these early experiences lead to a mental set in later life (encoded in internal working models) that feelings can be expressed and ‘processed’, conflicts successfully resolved. The anxiously attached child is caught up in a vicious circle (see Figure 6.1) in which
a) he lacks a secure base;
b) feels angry and wants to attack the attachment figure in the hope of discouraging future premature separations;
c) doesn't dare to do so for fear of retaliation or pushing the attachment figure even further away;
d) and so suppresses his feelings of anxiety and rage thereby increasing the sense of insecurity;
e) leading ultimately to an expectation of lack of care, and danger in emotional expression with potentially disastrous implications for self-esteem and intimate relationships.
There is, in this ‘Winnicott -type theory’ (Bowlby, 1988b) a massive block against expressing or even feeling a natural desire for a close trusting relationship, for care, comfort and love – which can be regarded as the subjective manifestations of a major system of instinctive behaviour.
Like Bowlby, Winnicott also repudiates the linear ‘monorail’ model of development in which the child progresses from oral to anal to genital phases of development:
Most of the processes that start up in early infancy are never fully established and continue to be strengthened by the growth that continues in later childhood and indeed in adult life, even in old age.
(Winnicott, 1965)
Bowlby's conviction that attachment needs continue throughout life, from immature to mature dependency has significant implications for psychotherapy. It means that the therapist inevitably becomes an important attachment figure for the patient, and that this is not necessarily best seen as a ‘regression’ to infantile dependence (the developmental ‘train’ going into reverse), but rather the necessary activation and coming to consciousness of previously suppressed attachment needs. Heinz Kohut (1977) based his ‘self psychology’ on a similar perspective. He describes ‘selfobject needs’ that continue from infancy throughout life and comprise an individual's longing for, and representation of empathic responsiveness from parents, friends, lovers, spouses (and therapists). This responsiveness brings a sense of aliveness and meaning, security and self-esteem to a person's existence. Its lack leads to narcissistic disturbances of the personality, characterised by the desperate search for selfobjects – for example, idealisation of the therapist or the development of an erotic transference. When, as they inevitably will, these prove inadequate (as did the original environment), the person responds with ‘narcissistic rage’ and disappointment, which, in the absence of an adequate ‘selfobject’ cannot be dealt with in a productive way. Van Vleet's attachment-derived triad of availability, non-controlling helpfulness and encouragement exemplify Kohut's model of the positive contribution therapists can bring to their patients, and stands in contrast to psychoanalytic models that are exclusively preoccupied with negativity, important though that is.
There is an inherent dualism in the Freudian project. Freud saw psychoanalysis as a science, and wanted his accounts of psychopathology to have the same status and explanatory power as those of physical medicine. At the same time, as Rycroft (1985) points out, he called his magnum opus The Interpretation, not The Cause of Dreams, and
It can indeed be argued that much of Freud's work was really semantic and that he made a revolutionary discovery in semantics, namely that neurotic symptoms are meaningful, disguised communications, but that, owing to his scientific training and allegiance, he formulated his findings in the conceptual framework of the physical sciences.
(Rycroft, 1985)
The semantic-hermeneutic and scientific-explanatory poles of Freud's thought are epitomised in contemporary psychotherapy on the one hand by Kleinian and Lacanian psychoanalysis and on the other by cognitive therapy. In this section I shall first give a brief account of these apparently irreconcilable approaches and then suggest that Attachment Theory provides a possible bridge between them.
Rustin (1991) has described the history of psychoanalysis as moving through the three Kantian categories of truth: scientific, ethical and aesthetic. Freud saw himself as a natural scientist, looking for general truths about normal and abnormal psychology; Melanie Klein's theories were essentially ethical – about destructiveness and splitting and the reconciliation of good and bad in the depressive position; Rustin sees current psychoanalysis as predominantly aesthetic in its orientation. Kant distinguishes aesthetic from scientific or moral judgement in having to do with ‘disinterested contemplation of objects of experience, related neither to the goal of interpersonal knowledge of causes, nor to issues of conformity with the moral law’ (Rustin, 1991). The discovery of meanings is central to this aesthetic sensibility, the prototype of which is to be found within the mother-infant relationship. The mother ‘contains’ or ‘binds’ (Bion, 1978) infantile sense experiences and mental images; she points, shapes and names, and so gives meaning to them. Maternal attunement (Stern), secure base provision (Bowlby) and holding environment (Winnicott) are all reaching towards the same idea. The function of the parent, of the therapist and of cultural objects can all be understood in this framework of containment and structuring and naming of inchoate experience.
To illustrate his point Rustin contrasts classical and contemporary psychoanalytic accounts of Hamlet. Ernest Jones (1949) saw the play as a quintessentially Oedipal drama in which Hamlet is wracked by his ambivalence towards the father-figures (the Ghost, the King, Polonius), and his simultaneous yearning and rage towards Gertrude, his mother, now married to the Uncle he wants to kill, and yet cannot. A contemporary reading however (Meltzer and Williams 1988) sees the problem of the play centring on Hamlet's search for a vehicle with which to express his grief, anger and ambition. The corrupt world of the court, and of institutional power, actively opposes its description – to do so would be to ‘expose’ the world of naked power greed and patriarchal domination. As the drama unfolds the famous play-within-a-play ‘catches the conscience of the king’. Throughout, Hamlet, like a patient in therapy (therapy is like a ‘soliloquy’ in which the patient muses aloud in the presence of the therapist/audience), struggles to tell his story, to achieve narrative competence, to find, in Williams' words, an ‘aesthetic correlative to image the idea of a new prince’ (that is, one not caught up in power and corruption). Dying, he enjoins his faithful Horatio to Absent thee from felicity awhile,
And in this harsh world draw thy breath in pain
To tell my story. (author's italics)
In this perspective, narrative becomes a key feature of the psychotherapeutic process, and especially to find a story which can express and contain painful experience. Therapy provides a setting in which thought rather than action can happen, and in which the patient can begin to tell himself his own story, undistorted by repression, splitting and affective distancing.
In the Lacanian (Bowie, 1991) account too, narrative is central, although a ‘story’, spoken in words, is seen as the imposition of the logos, of phallocentric culture on the primal, pre verbal unity of mother and child. For the Kleinians there is no such radical rupture with the onset of language: integration is achieved at the advent of the depressive position, rather than thwarted by the insertion of the paternal order.
The aesthetic perspective provides a cultural location for psychoanalysis, but what of its claim to be a science, and how do we evaluate one narrative account against another? Are all ‘stories’ equally valid, or are some more ‘true’ than others? And what of Bowlby's own comment:
I believe that our discipline can be put on to a scientific basis. A lot of people think you can't or don't know how to. There are people who think psychoanalysis is really a hermeneutic discipline. I think that's all rubbish quite frankly.
(Figlio and Young, 1986)
Bowlby wanted to make psychoanalysis more scientific, claiming always to be true to Freud's fundamental intentions, even if he was recasting them in different language.
It is here that the work initiated by Main (1991), Fonagy (1991) and Bretherton (1991a and b), and taken forward, among others, by Slade et al. (2005), Meins et al. (2001) and Lyons-Ruth (1999) are so important. As outlined in the previous chapter, the Adult Attachment Interview is a standardised instrument by which an individual's autobiographical narrative account of their childhood and attachment history can be linked with their behaviour as parents, and with the security of their children. Clear, coherent stories correlate with securely attached children. Narrative incompetence – inability to tell any sort of story, or embroilment in a muddled and incoherent one – is linked with insecure attachment. The narrative dimension in psychotherapy – helping patients to gain a clearer picture of their life and their early attachments – can therefore be supported on scientific as well as aesthetic grounds. Narrative competence, with the help of the AAI and other instruments, can be measured and tracked over time, both as it merges from childhood attachments, and as it changes in the course of therapy.
The polarisation between hermeneutics and science implicit in Bowlby's rather intemperate dismissal now looks a lot less clear-cut. Psychoanalysis provides a system of meanings for helping to decode patients' symptoms, but, if we step back from the specific meanings, we find good scientific evidence that narrative capacity, the ability to make meanings out of the inchoate flow of an ‘unstoried’ life – especially out of loss and disappointment – is associated with healthy psychological functioning.
If hermeneutics is ‘rubbish’ – a view which, had he lived long enough to consider the implications of Main and Fonagy's work, Bowlby might well have revised – what then of the opposing scientific tendency within psychotherapy? Cognitive therapy, devised by Beck et al. (1979), works primarily with cognitions, as opposed to the emotions that are the raw material of psychoanalysis. It is based on the idea that cognitions determine feelings (rather than vice versa), and that if the faulty cognitions which underlie neurotic states can be unearthed and corrected, then psychological health will ensue. There are strong echoes of Bowlbian metapsychology in this model. Mental structures are visualised in a hierarchy of expectations and assumptions, from specific assumptions in the psychologically healthy such as ‘When I am distressed I will receive help’, to core beliefs such as ‘I am lovable and can love’.
The internal working models of Attachment Theory are similarly visualised as a set of guiding affective and cognitive models of the world that are more or less (depending on the extent of ‘defensive exclusions’) subject to revision and updating. Cognitive therapy assumes that in neurosis the normal process of testing and modifying assumptions about the world breaks down, so that, for example, if the core belief in depression is ‘I am unworthy of love and deserve rejection’, when a fortuitous rebuff occurs this serves to reinforce the faulty belief and to deepen the depression.
In Ryle's (1990) modification of cognitive therapy, cognitive analytic therapy (CAT), he considers that these underlying core beliefs have their origins in disturbed attachment patterns in infancy and early childhood, later perpetuated in adult relationships by a vicious circle of self-fulfilling negative assumptions about the self and the world. As in CBT, but also Relational Psychoanalysis (see Leffert, 2013; D. Stern, 2012), Ryle's model of therapy requires a much more active collaborative attitude on the part of the therapist than in traditional analytic therapy. The therapist sets tasks for the patients, encouraging them to keep a ‘mood diary’ and to rate their progress on visual scales, as well as offering the patient a written formulation of the problem and its dynamics and a farewell letter when therapy (which is brief – typically sixteen sessions) comes to an end.
CAT is ‘Bowlbian’ in three significant ways. First in its theoretical eclecticism: Ryle happily marries cognitive science with psychoanalysis in an information-processing model that is very similar to Bowlby's attempt to re-write psychoanalytic defence mechanisms in terms of control theory. Second, Ryle's active therapist is engaging in ‘companionable interaction’ with the patient just like the secure base mother who actively plays with her child, and meeting the need for affiliation postulated by Heard and Lake (1986). Third, CAT focuses on a primary therapeutic task of helping the patient to become self-reflective, that is, to foster mentalising skills.
These contemporary developments are consistent with Fonagy's account of narrative capacity discussed above; indeed his group have devised a brief therapy model, Dynamic Interpersonal Therapy (DIT) (Lemma et al., 2012) that shares many characteristics of CAT. In the ‘Fonagy model’ the good mother accurately reflects the moods and wishes of her infant. This mirroring is then internalised as self-reflexive capacity, as the child gradually comes to know about his own internal states. This in turn manifests itself, as development proceeds, in the capacity to verbalise these states, and to ‘tell a story’ about oneself. The main themes of this autobiographical skill are the history of one's attachments, separations and reunions. Being brief therapies, CAT and DIT highlight and try to accelerate the emergence of autobiographical competence in a deliberate way rather than assuming that it will spontaneously emerge as therapy proceeds. Attachment research provides a rationale for this in that there is a demonstrable link between the capacity to ‘tell one's story’ and the development of secure relationships (Fonagy et al., 1995), which, from an attachment perspective, is a key therapy goal.
In summary, Attachment Theory has shown that the emphasis on narrative and hermeneutics in contemporary psychotherapy can be justified on good developmental grounds. Good mothers help their infants towards personal meanings, which in turn are a basis and mark of secure attachment. Cognitive therapy, although apparently opposed to the narrative approach in its concern with here-and-now cognitions, is also a story about the internal world. Its ‘basic assumptions’ are not far removed from Bowlby's internal working models or the ‘representational world’ of psychoanalysis.
Freud (1911) delineated two principles of mental functioning, the primary and secondary processes – the visual and the verbal, the imaginative and the rational. Implicit in his weltanschauung is the view that healthy functioning required a balance between the two. In Humphrey's (1992) re-working of this model there are two channels of information available to the organism, sensation and perception, which tell it about its own internal states and the state of the world respectively. More recently McGilchrist (2009; Holmes, 2012c) has added a comparable neuroscience gloss in his Right Brain/Left Brain dichotomising of the analytic and creative components to our mental lives.
The post-war psychoanalytic schisms led to a malign split between the concerns of psychoanalysis with primary processes and the focus of attachment and cognitive theory on secondary processes. The narrative paradigm, blending sensation and perception, and in which the inner world can begin to be described objectively, while the subjective colouring of the outer world is also held up for inspection, provides a possible escape route from this dilemma (Spence, 1982; Shafer, 1976; Roberts and Holmes, 1999). The question arises whether a secondary-process type verbal encouragement towards self-observation and narrative capacity is likely in itself to be effective, or whether primary-process ingredients, especially the arousal of affect and its re-working through transference, are also needed. To consider this and other more practical questions we must now turn to a consideration of the specific implications of the attachment perspective for the practice of psychotherapy practice.
