Psychoanalysis, Attachment Theory and the Inner World: How Different Theories Understand the Concept of Mind and the Implications for Clinical Work
From the 1930s onwards psychoanalysis has seen a paradigmatic shift away from Freudian drive theory towards a relational perspective. The work of Fairbairn, Melanie Klein, Winnicott and Balint provided momentum to this shift which gave rise to the development of object-relations theory (Holmes 1993). The British Psycho-Analytical Society was in a state of flux at this time, reflecting fierce disagreement between the Kleinian and Freudian camps about theoretical issues. These disputes were fuelled by the arrival in Britain of a group of psychoanalysts from Vienna which included Freud and Anna Freud. A compromise was agreed in 1944 following what became known as the Controversial Discussions, but significant theoretical disputes were left unresolved. Bowlby became increasingly disenchanted by the lack of scientific rigour characterizing psychoanalytic thinking. He took particular issue with Kleinian theory because of its emphasis on the role of unconscious phantasy in the aetiology of neurotic and psychotic symptoms at the expense of environmental factors, especially in relation to clinical issues of separation and loss.
In an attempt to provide psychoanalysis with scientific legitimacy, Bowlby turned to the newly emerging science of ethology. By linking the latter to neo-Darwinian evolutionary biology, Bowlby developed the idea that social as well as intrapsychic behaviour could be instinctive; that the child becomes attached to the person with whom he or she has the most interactions and not necessarily to the person who feeds him or her. Bowlby therefore suggested that human relationships could be the subject of empirical observation. His basic hypothesis, and that underpinning the development of attachment theory, is that the baby has a primary need to form an emotional bond with their primary attachment figure, often the mother (Bowlby, 1958). Moreover, his clinical experience led him to propose that when deprived of this relationship through separation and loss, the resultant fear, anxiety and distress has a deleterious and long-lasting effect on the infant’s overall physical and psychological development. For Bowlby, then, attachment is a goal corrected instinctual system and separation anxiety a purely instinctive reaction to an external danger which activates a distinct behavioural system.
In essence, the dispute that arose between attachment theory and psychoanalysis some 60 years ago focused on what Bowlby (1988) saw as a reluctance in analytic circles to examine the impact of real-life traumatic events in the genesis of pathology. Instead, classical thinking emphasised drive theory, unconscious phantasy and the death instinct
Quinodoz (1993) acknowledges that Bowlby’s approach poses a challenge to psychoanalytic theory but considers that his re-evaluation of this theory, with its introduction of control systems and instinctive behaviour, departs from the specific field of psychoanalysis and comes closer to experimental psychology.
From the very beginning, then, attachment theory was informed by a range of disciplines: object relations theory, ethology, evolution theory and developmental psychology. Contemporary attachment theory has continued in this vein, broadening its theoretical base to encompass cognitive science, cybernetics, social learning theory, linguistics, narrative theory, personality theory, philosophy and sociology. Holmes (1996) suggests that attachment theory has the potential to synthesize the most productive ideas from these various disciplines into a coherent new paradigm.
Psychological Theories of the Inner World
In broad terms, the inner world refers to the organization, structures and functioning of the mind and has been conceptualized in a variety of ways by different theorists. Psychological theories of the mind are inevitably linked to what philosophically is termed the mind-body problem. Here, the dichotomy is whether to approach the mind from an objective point of view, that is, the workings of the brain, or from a subjective point of view - as a psychology of personal experience (Hinshelwood 1991).
Freud (1911b), on discussing the topic of mental functioning in terms of drive theory, drew a distinction between primary and secondary processes. Primary process relates to the unconscious system and method of functioning. Here, infantile instinctual wishes and desires of a sexual and aggressive nature are repressed and become subject to mental defences of displacement, condensation and symbolism. However, repressed material returns in altered form and is expressed through, for example, dreams, parapraxes and psychosomatic symptoms. As the child matures and undergoes a process of socialisation, sexual and aggressive impulses are brought under the sway of the reality-principle. This principle characterizes the secondary process and incorporates the mature mental defence of sublimation whereby prohibited wishes and desires are channelled into culturally acceptable activities. For Freud, psychopathology is an external manifestation of unconscious neurotic conflict between an instinctual wish that is seeking discharge and a moral imperative, or between two contradictory emotions, such as ambivalent feelings of love and hate. Conflict is also given a central place in Freud’s later structural theory of mind, operating among the psychical agencies of id, ego, and superego (Laplanche and Pontalis 1988).
It may be seen, then, that Freud’s concept of mental life is both physiological, in terms of biological instincts and impulses, and psychological, in terms of the personal meanings that the developing infant comes to attribute to his or her instinctual life, and the way in which these meanings are assumed to motivate subsequent behaviour and activities. Hinshelwood (1991) therefore argues that Freud’s view on the mind-body problem reflects the philosophical position known as psychophysical parallelism. This position holds that there is both a mind and a brain, and that each work in their own particular ways. Hinshelwood (1991) goes on to suggest that Freud’s approach to this problem, as found in his writings, indicates a conflict between Freud the scientific neurologist and Freud the humanist psychologist, and that Freud never quite managed to extricate himself from the physiological psychology that was his starting point.
Melanie Klein, in her model of the mind, elaborates on Freud’s assumptions. Her theory posits that we live in two worlds consisting of inner psychic reality and external reality. The internal world is built up through the introjection of objects and comes to be experienced in as real and concrete a fashion as the outside world (Klein 1935, in Spillius 1988). Klein views the mind as also consisting of mental representations which include memories, ideas and, most significant of all, unconscious phantasies. These representations, together with relations with introjected objects, constitute the psychic structure of the Kleinian inner world.
In Klein’s formulation, splitting allows the ego to emerge out of the chaos of the paranoid-schizoid position and to order its emotional experiences and sensory impressions. This is a pre-condition of later integration and the basis of the faculty of discrimination - the capacity to differentiate between good and bad. The leading anxiety in the paranoid-schizoid position is that of persecutory objects getting inside the ego and overwhelming and annihilating both the ideal object and the self. This position is termed paranoid-schizoid because the leading anxiety is paranoid and the state of the ego and its objects is characterized by splitting, which is schizoid (Klein, 1945).
In terms of the child’s inner world, fears of persecutors and of the bad mother and the bad father lead him or her to feel unable to protect loved internal objects from the danger of destruction and death. Moreover, the death of good internal objects would inevitably mean the end of the child’s own life. Klein sees this situation as constituting the fundamental anxiety of the depressive position because the good internal object forms the core of the ego and the child’s internal world. The depressive conflict consists of a constant struggle between the child’s aggression and destructiveness (the death instinct) and his love and reparative impulses (the life instinct) (Klein, 1945).
In a contribution to the Controversial Discussions in 1943, Susan Isaacs argued that unconscious phantasy underlies every mental process and accompanies every mental activity. Hinshelwood (1991) suggests that this conceptual framework is informed by the philosophical position of psychophysical interactionism. From this position, the mind is viewed as emerging from the activity of the brain which, in turn, may be manipulated by the mind. This interactive process may be seen in Kleinian theory, which postulates that biological processes are mirrored in activities of the mind called unconscious phantasies. Equally, this theory holds that unconscious phantasies mould both the personality of the developing infant and his or her social world.
From this perspective, it is suggested that the infant exists in a world of so called ‘primitive’ emotions and, therefore, that introjected objects are initially experienced in an emotional rather than a physical way. The distinction between mind and body comes about in the course of development and is generated psychologically by a process of splitting. This process creates a psychical space within which the infant may experience the physical and the psychological. Hinshelwood (1991) therefore concludes that Kleinian theory assumes an interaction between physical events and psychological events and that each will influence the other.
Winnicott (1988), in a somewhat similar way to Klein, sees primary and secondary processes as complementary, rather than as being in opposition. He assumes the existence of a transitional space, viewing this as an intermediate area of experience between mother and child in which imagination and reality coincide. Under optimal conditions of good enough mothering within a holding environment the infant is gradually disillusioned of subjective omnipotent phantasies leading to an integration of personality and a sense of continuity within an “objective reality”. In Winnicott’s theory of mind, transitional objects, in the form of a favourite teddy bear or blanket, are used by the infant to bridge the space between inner and outer reality. Such phenomena provide a non-compliant solution to the loss of omnipotence and assist the child to separate from the merged state with the mother. A key aspect of the mother’s role is to mirror or reflect back the child’s own being, thereby facilitating the development of an authentic sense of self.
Winnicott’s thought would seem to resonate with the philosophical writings of Sartre from an existentialist perspective, and Husserl, Heidegger and Merleau-Ponty from a phenomenological perspective. Phenomenology is concerned with the description of pure subjective experience - the phenomena of consciousness. Consciousness is viewed as being the bearer of experience. These writers stress the significance of direct awareness of subjective experience and draw a distinction between authentic and inauthentic modes of being or existence. For example, Heidegger (1962) argues that a sense of self is accomplished through a process of “being-in-the-world”. This process is underpinned by a doctrine of intentionality and characterized by the subject’s active participation and involvement in the world. Similarly, Merleau-Ponty (1945, in Copleston 1979) suggests that the body-subject exists in a milieu in which its perceptual behaviour is in dialogue; that a dialectical relationship exists between the subject and his or her environment. For Merleau-Ponty perceptual experience is “the exceptional relation between the subject and its body and its world”.
Bollas (1994) argues that the infant’s inner world is structured, in part, by both Freud’s notion of primary process and Kleinian concepts of unconscious phantasy. Bollas (1994) suggests that during the early stage of development the baby experiences the mother as a transformational object associated with an intersubjective process. He subscribes to the view that no clear distinction exists between internal and external perception at this time. In using the concept of intersubjectivity, Bollas (1994), too, is emphasising the phenomenological element of experience; the process by which the infant-mother dyad participate in and identify with their respective inner subjectivities, thereby creating a shared psychological experience. For Bollas (1994), the process of intersubjectivity “instructs” the infant into the logic of being and relating. This is achieved by means of the mother’s countless exchanges with her child. However, the earliest levels of psychic experience are not readily available for mental representation or symbolic processing leading Bollas (1994) to call this form of “knowledge” the unthought known. This term stands for everything that on some deep level is known, such as moods, somatic experiences and personal idiom, but which has not yet been thought, in that the phenomena have remained unavailable for mental processing.
Bollas’s concept of the unthought known is influenced by Bion (1984) who argues that we all have sense impressions and emotional experiences. Bion (1984) suggests that there is a specific function of the personality which transforms sense impressions and emotional realities into psychic elements. These then become available for mental work by such means as thinking, dreaming, imagining and remembering. Bion (1984) terms the latter alpha elements, and the process by which they are transformed, alpha function. This process requires the mother to enter a state of calm receptiveness - a state of mind Bion (1984) terms reverie. The mother is thereby amenable to containing the infant’s inchoate state of mind and thus able to give meaning to the anxiety and terror inserted into her in unconscious phantasy by means of projective identification. Beta elements, on the other hand, consist of untransformed sense impressions and emotional experiences. These elements are experienced as split off, unintegrated “things-in-themselves” and are therefore evacuated by means of projective indentification, leaving the self feeling depleted, fearful of persecution and in a state of “nameless dread”. Bion (1990) developed Klein’s concept of projective identification, arguing that it may function as a form of normal communication between subject and object, as well as being a sadistic pathological act of expulsion of split off, disowned parts of the self experienced as intolerable.
Ogden (1994), also writing from a Kleinian perspective and, like Bion, seemingly influenced by phenomenology, posits that the paranoid-schizoid position and depressive position are, in essence, states of being which coexist dialectically. These defensive organizations consist of constellations of phantasies, relations with objects and characteristic anxieties and defences (Joseph 1994). Ogden (1994), in line with Bion (1990), argues that there is a continuous interplay between these two defensive organizations, rather than one transcending the other, as postulated by Klein. Furthermore, Ogden (1994) suggests that this inter-penetrative process is also in operation between conscious and unconscious states of mind, and between the past and the present. He concludes that our experience of inner and external reality is vitally affected by this dialectical interplay, as this process indicates a coexistence of multiple states of consciousness.
Along a similar theoretical line of thought to Bollas (1994), Benjamin (1992, in Mitchell 1993) equates the development of mind or inner reality with the experience of the self as a subject in relation to the subjectivities of others. From this intersubjective perspective, it is assumed that a “shared reality” comes to be established by means of a subtle intertwining of both intrapsychic and interpersonal processes. For Mitchell (1993), interactional processes of the type described by Benjamin (1992), Ogden (1994) and Bollas (1994) give rise to a manifold organization of self or mind patterned around different self and object images and mental representations, as derived from different relational contexts. Psychological meaning is negotiated through interaction in the relational field, rather than regarded as universal and biologically inherent as in drive theory (Mitchell 1988). Indeed, Mitchell (1993) suggests that relational theories would seem to have been confirmed and validated by recent infant research on the communication of affect between mother and infant. These findings indicate that the mother’s emotions in some way “become” part of the baby’s emotional experience, thereby supplying the tone and contours that make up the world in which the baby lives (Beebe and Lachmann 1992, in Mitchell 1993).
Mitchell (1993), then, points out that the model of mind or self found in relational theories emphasises its multiplicity and discontinuity, with experience portrayed as being embedded in particular relational contexts. He therefore argues that psychic organization and structures are built up through and shaped by the interactions we have with different others, and through different interactions with the same other. From this position, Mitchell (1993) goes on to suggest that our experience of self is discontinuous and composed of different selves with different others, rather than consisting of a singular entity. Moreover, at times we may identify with an aspect of our self or with an aspect of the other, and this will affect the way in which we organize experience and construct our sense of meaning.
Somewhat paradoxically, Mitchell (1993) argues that despite the discontinuous aspect to our experience we nevertheless retain a sense of self as enduring and continuous. He depicts this in Winnicottian terms as consisting of an unbroken line of subjective experience which forms the core of the personality. However, if it is accepted that the self is both multiple and discontinuous and integral and continuous, a creative tension arises requiring a balance to be struck. As Mitchell (1993) puts it, where there is too much discontinuity there is a dread of fragmenting, splitting and dislocation. On the other hand, where there is too much continuity there is a dread of paralysis and stagnation. In formulating this paradox, Mitchell (1993) acknowledges Winnicott’s (1988) concept of the true and false self. Mitchell (1993) also refers to McDougall’s clinical description of normopathic characterological traits (McDougall 1990), and to Bollas’s (1994) concept of the normotic personality, and suggests that “pseudonormality is the clinical problem of our time”. The clinical picture depicted here would seem to mirror the bleak and desolate aspect of our own existentialist experience and constitute what Sartre (1966) refers to as “bad faith”, that is the turning away from an authentic form of existence and choosing, instead, to become a passive subject of external influences.
In a way that complements a relational understanding of the inner world, Stern (1998) argues that repeated patterns of interaction constitute the basic building blocks of psychic formation and structures the infant’s representational world. These patterns, in turn, build the perceptual, affective and cognitive schemas used to organize and construct subsequent life experience (Fosshage, 1992). A schema is said to consist of a procedural memory. The latter consists of non-conscious, non-reflective emotional and impressionistic information derived from, and arising out of, a dynamic series of micro-events. Micro-events refer to processes of mutual influence between the mother and the infant and constitute what Stern (1998) calls the relational mode, that is “a-way-of-being-with” or “a style of relating”. Infancy research demonstrates that mutual influence takes the form of the timing, tracking and matching of vocal exchanges and the duration of “movements” and “holds” in the changes of facial expression and direction of gaze. These interactive behaviours are employed for the specific purpose of regulating the exchange (Beebe, Jaffe and Lachmann, 1992).
Stern (1985) argues that procedural memories are internalized and represented as patterns of interactive behaviours that become generalized. Findings indicate that such interactive patterns are the means by which affect is communicated and, moreover, provide a behavioural basis for the mother and infant to perceive and enter into the temporal and feeling state of the other (Beebe, Jaffe and Lachmann, 1992).
Trauma, Dissociation and the Inner World
From a relational/intersubjective perspective, psychological trauma is viewed in interpersonal terms and traumatic affect as playing a part in organizing mental functioning. Thus, painful affect may become a significant motivating force behind so called ‘pathological reactions’, such as violent behaviour and attempts at self-soothing through, for example, addiction, compulsive sexuality, and self-injury.
During the past four decades there has been renewed interest in psychological trauma. This interest has been fuelled by the experience of American veterans of the Vietnam War and by an acknowledgement of the widespread incidence of emotional, physical and sexual abuse in Western culture. This was particularly taken up by the women’s movement and in the emergence of self help groups where the medicalisation of trauma was questioned and survivors collaborated in their own support and recovery. These factors have, in turn, generated a renewal of interest in the mental defence of dissociation and to a debate about multiple personality disorder or dissociative identity disorder.
The term dissociation was first used in the mid 1890s by Janet to describe a changed state of consciousness in patients who had suffered traumatic experiences. Herman (1992) defines dissociation as an altered, detached state of consciousness which is automatically induced as a mental defence against psychological trauma involving pain, danger, terror, and helplessness. Herman (1992) argues that the traumatic event has the effect of overwhelming and disorganizing the person’s normal responses, with the consequence that part of mental functioning becomes separated from other activities. She suggests that in such situations events become disconnected from their ordinary meaning and that perception is distorted, with partial loss of memory occurring. Experience takes on a dream-like, unreal quality and the traumatized person feels as though the event is not happening to him or her.
Davies and Frawley (1994) view dissociation as existing on a continuum, with multiple personality disorder representing the most extreme form of this defence. From a psychobiological perspective, Herman (1992) suggests that altered states of consciousness result from the release of endogenous opioids within the central nervous system, and that this biological reaction is triggered by the traumatic event itself. De Zulueta (1993) thinks it probable that a similar psychobiological process is involved in attachment behaviour which comes into operation in reaction to separation trauma and bereavement. Herman (1992) points out that dissociated traumatic experiences may become frozen in time, and that a process of mourning within a secure, holding environment is needed to facilitate the integration of split off affect and cognitions, and of verbal and mental representations associated symbolically with the traumatic events.
Attachment Theory and the Inner World
The clinical issue of trauma and its disorganizing effect on mental functioning brings us back to attachment theory and the question concerning how writers from this approach portray the inner world.
Bowlby (1969) sets out his position by quoting approvingly the philosophical view of mind expounded by Hampshire. The latter asserts that “patterns of behaviour in infancy … must be the original endowment from which the purely mental states develop” (Hampshire, 1962, quoted in Bowlby, 1969, p.6). On developing his thinking about psychopathology in general, and aggression in particular, Bowlby (1969, 1979) pointed out that Freud’s major theoretical formulations consistently centre on trauma and on an understanding of how intrapsychic conflict between sexual and ego instincts and life and death instincts, expressed as the ambivalent conflict between love and hate, comes to be satisfactorily regulated (Freud, 1915c, 1923b). Following Klein (1940), who was his supervisor during his psychoanalytic training in the 1930s, Bowlby (1960, 1969, 1973, 1980) drew a connection between pathological childhood mourning and psychiatric illness in adulthood. He was particularly influenced by Klein’s view that certain mental defences in early childhood are directed against “pining” for the lost object (Klein 1940). However, Bowlby (1958, 1960, 1969, 1973, 1980) eventually discarded the dual-drive theory of sexuality and aggression, arguing that a biologically based “drive” for attachment was more compelling. In a direct and bold challenge to classical theory, he proposed that it is the particular quality of love and security provided by the caregiver that helps the child to regulate the basic conflict between love and hate. In addition the child’s so called ‘aggression’ far from being innate is a response to failures in attunement to the child’s distress.
Bowlby was particularly influenced by Fairbairn’s account of the way in which the infant’s actual experiences with people structures his or her internal world of object-relations. For Fairbairn (1996), feelings of security vitally influence the manner by which the infant affectively relates to internalized, split off idealized objects and rejecting objects. Furthermore, Fairbairn (1996) sees insecurity as stemming primarily from separation anxiety. He argues that this type of anxiety is a causative factor in the development of “schizoid” aspects of personality, engendering a sense of futility and hopelessness. Such a person lacks the capacity to differentiate self from other and thus is unable to attain a state of “mature dependence”. These hypotheses seem to have accorded with Bowlby’s clinical experience and are reflected in his theoretical statement that a secure attachment tie to the mother functions to integrate the child’s personality (Bowlby 1958).
Whilst initially empirical observation focused on the infant’s acute distress when separated from the mother, as expressed by protest, despair and detachment, extensive evaluation of Bowlby’s hypothesis of maternal deprivation by Rutter in the 1970s and 1980s suggested that a far more complex set of social and psychological factors were in operation. As Holmes (1993) points out, Rutter’s work prompted a move away from a simplistic event-pathology model to an appreciation of the subtle nature and quality of the child’s attachment to the mother or primary caregiver. Indeed, the work of Bowlby and his collaborators inspired wide interest in infancy research, a field of study that continues apace. Research findings would seem to confirm both the validity of Bowlby’s theory of attachment and the connection between the child’s interpersonal relations and his or her inner world.
Attachment theory views the mind as structured by patterns of attachment that are conceptualised as self-other internal working models. Internal mental models reflect states of mind in relation to attachment, vitally influencing our expectations and appraisals of others, and our capacity to regulate emotions, particularly in contexts involving interpersonal stress. Indeed, Peterfreund (1983), in advocating a “heuristic” approach to the process of psychoanalytic psychotherapy, suggests that different working models are active during different activities and in different situations making predictive calculation and adaptive behaviour possible. He emphasises the significance of information processing and error-correcting feedback in this process, as these are the means by which internal working models are modified, updated and fine-tuned. However, Main’s (1985) research findings indicate that once patterns of attachment are established they tend to become actively self-perpetuating. Thus, when an established attachment behavioural system receives potentially disruptive signals, these are actively countered by perceptual and behavioural control mechanisms. Main et al (1985) conclude, albeit tentatively, that internal working models derived from insecure patterns of attachment are resistant to modification and updating because error-correcting information is being defensively and selectively excluded from consciousness. From a more traditional psychoanalytic approach, it would be assumed that mental defences such as repression, denial and idealization are being evoked in response to signal anxiety with the purpose of maintaining psychic equilibrium. In line with the ideas of narrative theory and concepts of metacognition and self-reflexivity, Main and her colleagues (1985) hypothesize that the coherent integration and organization of information relevant to attachment may play a determining role in the creation of security in adulthood.
Fonagy and his colleagues have developed Main’s concepts of metacognition and self-reflexivity, labelling these processes as the capacity to mentalize one’s own and the other’s emotional and intentional states of mind. The terms reflective functioning and mentalization are often used interchangeably. Reflective functioning refers to a scale devised by Fonagy and Steele that operationalises the concept of mentalization for research purposes. Recent research suggests that mentalization is not a unitary capacity but one that may fluctuate within the individual across different relational contexts. Indeed, there is increasing evidence that mentalization is a multifaceted and variable process. Moreover, even when mentalization is reasonably well-established, particular emotions may have been defensively excluded. However, it is generally agreed that the capacity to mentalize crucially depends on the infant’s earliest emotional exchanges with the primary caregiver (Target 2008). In this context, Fonagy (2008) points to the vicious cycle created by attachment trauma in hyper-activating the attachment system and shutting down mentalization. As a consequence of being unable to mentalize the traumatic experience, the child relives it in the mode of psychic equivalence instead of in the pretend mode. Thus, defensive strategies and implicit procedures developed in infancy in response to attachment trauma and misattuned care-giving become aspects of character and relating that persist precisely because they are automatic and outside awareness. Stressful interpersonal contexts in later life may activate the attachment and fear behavioural systems and again compromise the capacity to mentalize.
Interestingly, Slade's research with parents and children in psychotherapy draws a distinction between parental reflective functioning - the caregiver's capacity to reflect on the current mental state of the child and upon her own mental states as these pertain to her relationship with her child, and her capacity to reflect upon her childhood relationships with her own parents, as classified by the Adult Attachment Interview (Slade 2008). Her research suggests that parental reflective functioning is more influential than parental attachment organisation in terms of predicting positive outcomes - such as secure attachment in the child. This is thought to account for the so-called "transmission gap" noted by van Ijzendoorn in 1995. He pointed out that researchers have consistently failed to clearly document that maternal sensitivity and responsiveness is what links adult and infant attachment. This led him to suggest that the mechanisms underlying such intergenerational processes have yet to be understood. It is thought that reflective functioning, as described by Grienenberger et al. (2005) and Fonagy (2008), helps to explain this transmission gap.
Although attachment security in the first year of life is predicted primarily by the security of the mother's attachment organisation and level of reflective functioning, Howard and Miriam Steele's longitudinal study at University College London found that the father's level of reflective functioning was significantly related to a number of interpersonal and personality factors in boys in middle childhood. Such factors include self esteem, identity formation, affect regulation and delinquency, and also the son's ability to give a coherent account of himself and others at age 11. Steele and Steele (2008) conclude that attachment is relationship specific, with representations or internal working models of mother and father developing separately, rather than as one overarching model of attachment. This fits in with Main et al's (1985) earlier finding showing that a child may be disorganised with one parent but not with the other. It also confirms the emerging concept of multiplicity – that our sense of self is represented by multiple states of mind.
More generally, attachment research suggests that the mind can continue to develop throughout the lifespan via changes in internal working models (Siegal 2001). Such findings are supported by neuroscience, which increasingly recognises that the brain retains plasticity throughout life, adapting to changes in environmental challenges and demands. Dissociation in reaction to trauma represents an uncontrolled and negative expression of neural plasticity which is reflected in the disruption of learning, memory and neural network organisation (Cozolino 2002). However, a more positive challenge is provided by the process of therapy, with new neural connections being reflected in updated internal working models and the attainment of “earned security” (Cozolino 2002; Main 1991). These developments point to the need for a long term attachment relationship for changes in internal working models to be accomplished.
Theoretical Integration?
From the foregoing it will be noted that the philosophical position informing attachment theory’s approach to the mind-body problem is that of psychophysical interactionism, as described on pages 3 and 4. As we have seen, this is also the position informing Kleinian and other relational and interpersonal perspectives. The latter approaches, however, tend to develop theory about the infant’s state of mind retrospectively via phenomena gleaned from clinical work with adults. Nevertheless, it would appear that theoretical assumptions drawn from clinical experience have much in common with empirical research supporting attachment theory. Indeed, Holmes (1996) argues that a truly interpersonal/intersubjective psychoanalytic psychotherapy is evolving out of the work of Klein, Bowlby, Bion, Winnicott, Kohut, and, more recently, that of Benjamin and Mitchell. The extent to which this integrative process is happening is a matter of debate. In the meantime, it would seem clear that theoretical assumptions are of central importance to any discussion of the inner world, as the very purpose of psychological theory is to provide a conceptual framework for the understanding of mental functioning. Bowlby, himself, makes a similar point by quoting Kurt Lewin’s dictum that “There is nothing so practical as a good theory” (Bowlby 1988, p.37).
Intersubjectivity and the Inner World
As we have seen, relational and interpersonal theorists view the inner world as developing through a process of intersubjectivity. For Mitchell, all meaning is generated in “the symbolic textures of the relational matrix” (Mitchell 1988, p.62). He conceptualizes the relational matrix in broad, paradigmatic terms, seeing it as integrating the theories of Bowlby, Fairbairn, Klein, Winnicott, and Kohut. Mitchell (1993) argues that the dynamics and life history of the person in analysis are actually co-created by the analyst’s participation during the course of the therapeutic process. This assumption constitutes a shift in theory, with truth now being viewed in terms of narrative intelligibility and discourse coherence, rather than historical veracity. Thus, there is no singular correct version of reality and experience may be understood in various ways.
From this post-modern perspective, the past is not reconstructed, but constructed and given meaning in the here and now, with reality being mediated by personal narrative. It follows, therefore, that in the analytic situation the patient’s inner world of experiences, associations and memories can be integrated or organized in many different ways. Mitchell (1993) argues that the scheme arrived at is a dual creation, shaped partly by the patient’s material, but also inevitably moulded by the analyst’s patterns of thought, theory and systems of ideas. He goes on to stress the enormous importance that the analyst’s theory has on the analytic process, arguing that the theory itself influences what is seen in the clinical material and also shapes and organizes it.
From this standpoint, the therapeutic process is redolent of the subtle intersubjective process that takes place between the mother and the child, which is assumed to structure, shape and organize the child’s inner world. It is, therefore, perhaps not surprising that Mitchell (1993), in a similar way to Winnicott (1988), views the mother-infant relationship as the prototypical therapeutic model.
Hermeneutics versus Empiricism?
Mitchell’s thinking in regard to the therapeutic process is influenced by hermeneutic interpretive theory, a discipline closely associated with phenomenology. Recent exponents of this theory are Heidegger (1962) and Habermas (1972). In broad terms, a hermeneutic may be defined as a set of practices or techniques used for the purpose of revealing intelligible meaning (Shotter, 1986, in Harré and Lamb 1986). The task in the hermeneutic tradition is to understand the subjective inner reality of the mind by reference to the person’s historical and cultural context. This approach requires a splitting of reality into two: an outer reality to be explained causally, and an inner reality which needs to be understood, that is given meaning. Hermeneutics, then, may be seen as acting to reveal the hidden subjectivity, intentions and purposes of the inner world. Shotter (1986) suggests that the recent renewal of interest in hermeneutics reflects the view that empiricism is an inadequate approach to use in the understanding of mental phenomena.
Certainly, the inner world as portrayed by attachment theory would seem to have little space for the stuff of primary process such as dreams, fantasies, wishes and desires. Indeed, Bowlby’s view of hermeneutics in relation to psychoanalysis was trenchant and dismissive, as summed up by his terse comment that “There are people who think psychoanalysis is really a hermeneutic discipline. I think that’s all rubbish quite frankly” (Bowlby et al. 1986, quoted in Holmes, 1993, p.145). It is of interest to note, therefore, that attachment theory’s position would seem to have shifted in this respect, in that hermeneutic interpretive theory is now embraced in the form of narrative theory, metacognitive knowledge and self-reflexivity, as seen in Main’s work in respect of the Adult Attachment Interview.
The incorporation of narrative theory into attachment theory would seem to be a yet further example of the interplay between theory informed by prospective empirical infancy research, on the one hand (Ainsworth et al 1982), and theory developed retrospectively from clinical phenomena with adults, on the other. Therefore, hermeneutics, applied empirically in order to interpret or code a semi-structured Adult Attachment Interview, may be viewed as bridging the divide between empiricism and phenomenology. This paradigmatic shift would appear to lend weight to the respective claims of Mitchell (1993) and Holmes (1996) that integration is taking place between the different relational approaches. The convergence between postmodern theory, attachment theory, infant research and psychoanalysis is also noted by Teicholz (2009). Ogden (1994), however, is somewhat more cautious in this regard, viewing psychoanalytic theory as characterized by “an uneasy coexistence of a multiplicity of epistemologies” (p.193). Indeed, he suggests that the task of theoretical integration needs to be accomplished at the level of the individual practitioner. He therefore argues that clinicians need to develop their thinking within the context of different systems of ideas which together “in a poorly integrated way constitutes psychoanalysis” (ibid).
The Therapeutic Process Using an Integrated Attachment, Relational, Intersubjective and Neurobiological Approach
In summarising the ideas of the Boston Change Process Study Group (Stern et al. 1998), Fonagy (1998) points out that non-conscious schemata are thought to define how interpersonal behaviour is conducted. This view is based on the premise that memory consists of two relatively independent systems, namely autobiographical or declarative memory, which is partly accessible to awareness, and implicit-procedural memory. Research in this area has demonstrated that implicit memory is principally perceptual, non-declarative and non-reflective, being more dominated by emotional and impressionistic information than autobiographical memory. Fonagy (1998) considers it likely that “the schematic representations postulated by attachment and object relations theorists are most usefully construed as procedural memories, the function of which is to adapt social behaviour to specific interpersonal contexts” (p.348). Given this, Fonagy (1998) suggests that patterns of attachment are stored as procedural memories which themselves are organized as mental models. He argues that knowledge of these procedures is accessible only through behavioural performance, that is, by direct observation of the individual’s manner or style of relating. Previously it was assumed that such knowledge was accessed through the verbal description of ideas or memories.
If we accept that the inner world of subjective experience is encoded and stored in the systems of implicit-procedural memory (Schacter 1996), as conceptualised as self-other internal working models of attachment, the theoretical understanding of implicit modes of interaction, particularly as this relates to verbal and non verbal emotional communication, becomes a central focus of the therapeutic process. This is particularly the case if we also accept that such non conscious experience is enacted in our most intimate relationships and is most readily accessed in the context of those relationships (Fonagy 1998; Shimmerlik 2008; Stern et al. 1998).
The implicit domain and mode of functioning differs from the Freudian concept of the dynamic unconscious. Whereas the dynamic unconscious consists of what was once known and has then been repressed, implicitly encoded non conscious strategies and procedures reflect the ways in which the mind and the brain have been shaped and sculpted by patterns of interaction in our earliest relational matrix.
It follows that what we communicate to others, and register from others, both in everyday life and in the therapeutic encounter, often occurs out of awareness and in an enactive mode of relating. This emphasises the fact that emotional communication is complex and at times highly ambiguous and that our experience is mediated by our non conscious internal working models of attachment.
More generally, attachment theory and infant research demonstrate that psychological organization is an adaptation aimed at preserving critical, life-sustaining relationships. As Slade (2004) points out, attachment classifications used for research purposes are simply ways of describing and organizing implicitly encoded attachment phenomena. These phenomena, and the processes and relational procedures they represent, are the focus of clinical work, not the classifications per se. A basic understanding of attachment theory and research sensitizes the therapist to the nature and functioning of the attachment system and aids in the observation and recognition of attachment phenomena, as revealed in the patient’s speech and behaviour (Slade 2004).
The initial interview provides an ideal opportunity to begin to listen for attachment phenomena, as manifested in the patient’s talk about his or her relationships with parents, partners and children. In my experience, many people commence therapy on becoming parents because they have become aware of repeating with their children the negative aspects of the relationship they had with their own parents. Despite this cognitive awareness, they seem unable to change the way they relate to their children in emotional (procedural) terms (Renn, 2008a). Familiarity with adult attachment research will guide the therapist to listen to the fluency, coherence, affectivity and flexibility in the patient’s narrative descriptions of early childhood attachment experiences. This provides the means of identifying his or her particular ways of regulating and defending against implicitly encoded attachment-related memories and feelings (Slade 2004).
Attachment research also alerts the therapist to listen for themes of attachment trauma in the form of loss, neglect, rejection, abandonment and abuse in the patient’s narrative. Such narratives, and the discrete discourse style in which they are communicated, can tell the therapist a great deal about the patient’s capacities to hold and reflect upon their own and the other’s mental states in making sense of behaviour and relationship patterns. By extension, the patient’s narrative also informs us about his or her early intersubjective experience and developmental trauma. These narratives and discourse styles also offer an opportunity to evaluate the patient’s attributions of the other – the nature and affective qualities of his or her inner world representations of the other.
Adults who have developed a dismissing attachment state of mind avoid intimacy and exploration of painful thoughts and feelings. By contrast, those who have created a preoccupied attachment state of mind are angrily enmeshed with their past and current attachment figures. Adults with an unresolved state of mind cannot maintain affective continuity in their inner worlds and become disorganized and disoriented when re-experiencing a traumatic event. These contrasting adult attachment states of mind are captured in attachment research utilizing the Strange Situation Test and the Adult Attachment Interview. Findings show that, while the avoidant infant and dismissing adult develop a state of mind that values emotional self-reliance and separateness, the ambivalent-resistant infant and preoccupied adult develop a state of mind that is angry, frightened and anxious about being separate and autonomous. The disorganized child and unresolved adult dissociate from the immediate environment and develop either a helpless or hostile/controlling state of mind. These states of mind, then, give rise to attachment strategies and phenomena that are communicated, in part, via the patient’s particular discourse style and interaction with the therapist. Being aware of our own predominant attachment state of mind may help us, as therapists, to recognize and understand the enactments that we inescapably get drawn into with our patients and inform how best to repair such inevitable ruptures to the attachment relationship or therapeutic alliance.
In clinical practice, then, attachment theory and research are used to conceptualize the developmental antecedents and interpersonal features of the patient’s difficulties in living, particularly his or her implicitly encoded procedures or strategies for managing closeness and distance and separations and reunions in intimate romantic relationships, and the influence of these phenomena on the formation of the therapeutic alliance (Lopez & Brennan 2000). Attachment theory and research provide both a particular way of listening to the patient’s story and of understanding the clinical process (Slade 1999). An aspect of this process involves identifying similarities in the complex dynamic interplay between the patient’s early relational matrix and his or her current intimate relationships, including that with the therapist. This facilitates an understanding of the way in which archaic, non conscious cognitive-affective internal working models of attachment are being perpetuated in the here and now, actively mediating and distorting the person’s attachment-related thoughts, feelings and behaviour, particularly at times of heightened emotional stress – how the relational past lives on in the interpersonal present (Renn, 2008b). These relational dynamics and defensive processes are most immediately apparent in couple therapy (Shimmerlik 2008).
From an attachment/trauma perspective, the patient’s symptoms, destructive and self-destructive behaviours are understood as expressing unprocessed traumatic experience encoded in implicit-procedural memories, as represented in confused, unstable self-other working models (Renn 2006, 2007). These non conscious state-dependent memories and patterns of expectancies organize experience and emerge in the relational system or intersubjective field, being communicated directly to the therapist via the patient’s discourse style and expressive behavioural display. This, in turn, activates a matching countertransferential or psycho-physiological response in the therapist (activates the mirror neuron system), enabling the therapist to participate in the subjective experience of the patient in terms of shared attentional, intentional and affectional states of mind (Schore 1994; Stern 1985).
The developing attachment relationship with the therapist, then, provides a good-enough safe haven and secure base from which the patient can explore his or her self-states, as reflected in the mind of the therapist moment-by-moment, thereby unlocking the affective components of their unresolved trauma (Schore 1994). In addition to the repair of inevitable ruptures to the therapeutic relationship, crucial aspects of the therapeutic process consist of the interactive regulation of heightened affective moments, the provision of new perspectives, the re-organization of maladaptive patterns of expectancies, the transformation of implicitly encoded representations, and the promotion of reflective functioning or mentalization (Bateman & Fonagy 2004; Beebe & Lachmann 2002; Cozolino 2002, 2006; Fonagy, 2008; Schore 1994).
In terms of therapeutic action, Diamond and Kernberg’s longitudinal investigation of the treatment process and outcome of borderline patients in Transference Focused Psychotherapy suggest that improvements in the capacity to mentalize appear to be a function of the characteristics of the therapeutic relationship, including the level of the therapist’s reflective functioning about the particular patient (Diamond and Kernberg 2008). This aspect of the therapeutic relationship is emphasised by Wallin (2007), who argues that the therapist’s own internal world is transmitted to the patient, influencing the development of his or her internal working models. Indeed, he asserts that no factor influences our effectiveness as therapists more than our own attachment patterns.
With regard to clinical practice, these various findings emphasise that an emotionally meaningful therapeutic relationship facilitates a collaborative co-construction of the patient’s dissociated traumatic experience and promotes the recognition of the mental states that motivate human behaviour in various relational contexts (Davies & Frawley 1994). More specifically, the process of interactive regulation of affect facilitates the recognition, labelling and evaluation of emotional and intentional states in the self and in others (Bateman & Fonagy 2004; Fonagy 2008; Grienenberger et al. 2005; Slade 2008). This, in turn, engenders a coherent, secure and agentic sense of self as archaic internal working models are revised and updated and new relational models develop (Knox 2003; Peterfreund 1983). This, together with the patient’s growing realization that he or she can contingently influence the therapist and, by extension, others in everyday life, engenders a secure enough sense of self and recognition of other people as separate, differentiated subjects who can be related to in non coercive and non destructive ways (Benjamin 1992). As Herman (1992) points out, the antidote to the helplessness characteristic of trauma is the ability to exercise control and self-agency.
The enhancement of the patient’s ability gradually to organize and integrate error-correcting information consists, in significant degree, of the moment-to-moment micro-repair of misattunement or misaligned interaction - an intersubjective process operating at the level of implicit relational knowing (Beebe & Lachmann 2002; Stern et al. 1998; Tronick et al. 1978). The therapeutic process is informed by the tracking and matching of subtle and dramatic shifts in the patient’s mood-state as they narrate their story (Schore, 1994). This interactive process leads, in turn, to the recognition of the existence of the therapist as a separate person available to be used and related to intersubjectively within a shared subjective reality (Benjamin 1992).
By these means, the therapist’s facilitating behaviours combine with the patient’s capacity for attachment. Though operating largely out of conscious awareness, this process of mutual reciprocal influence or contingent reciprocity engenders a sense of safety and security and thus the development of an attachment relationship that facilitates a collaborative exploration and elaboration of painful, unresolved clinical issues and dissociated traumatic self-states underlying the patient’s difficulties in living. Key aspects of this intersubjective and reparative process are the dyadic regulation of dreaded states of mind charged with intense negative affect and the co-construction of a coherent narrative imbued with personal meaning.
Optimally, the therapist becomes a new developmental object, the relationship with whom provides a corrective emotional experience, thereby disconfirming the patient’s transference expectations (Alexander & French; Hurry 1998). This process enhances the patient’s capacities for affect regulation and mentalization, and also their sense of self-agency (Knox 2003). This, in turn, strengthens the insecure/unresolved patient’s ability to activate alternative mental models of interaction, enhances their capacity to empathize with others and so make more reasoned choices, and reduces their tendency to deploy mental defences of perceptual distortion, defensive exclusion and selective attention in stressful situations that generate a sense of endangerment to the self and a concomitant increase in the risk of destructive and self-destructive behaviour (Holmes 1996; Renn 2006).
From a neurobiological perspective, the process of affect regulation, so central to attachment theory and research, links non verbal and verbal representational domains of the brain. This process facilitates the transfer of implicit-procedural information in the right hemisphere to explicit or declarative systems in the left. Thus, body-based visceral-somatic experience is symbolically transformed into emotional and intentional states of mind that then become available for reflection and regulation (Damascio 2000; Schore 1994).
Again, this aspect of the therapeutic process reflects the philosophical position of psychophysical interactionism – that the workings of the brain are expressed in terms of subjective personal experience and that personal experience influences the workings of the brain. Moreover, recognition of the interaction between mind and body may be seen as a further example of the integration of a scientific, empirical methodology with a hermeneutic, phenomenological approach in order better to understand and give meaning to the inner world of subjective experience.