Show Summary Details
Page of

(p. 216) Mentalization-based therapies 

(p. 216) Mentalization-based therapies
(p. 216) Mentalization-based therapies

Robbie Duschinsky

and Sarah Foster

Page of

PRINTED FROM OXFORD CLINICAL PSYCHOLOGY ONLINE ( © Oxford University Press, 2021. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a title in Oxford Clinical Psychology Online for personal use (for details see Privacy Policy and Legal Notice).

date: 06 December 2021


From early in his career, Fonagy was adamant that ‘it is inconceivable’ that ‘if the ideas proposed by the group have substance’, that they would have no implications for the technical priorities of therapeutic work with patients.1 As we saw in Chapter 1, Fonagy’s clinical work in the 1980s with patients diagnosed with borderline personality disorder (BPD) played an important role in the development of his thinking. Clinical work as a psychoanalyst working with adults and children directly informed his concern with mentalizing, which grew out of technical innovations in ‘developmental help’ offered at the Anna Freud Centre. In turn, the theory developed by Fonagy and colleagues fed into the emergence of mentalization-based therapy (MBT) as a treatment modality. Initially, in the 1990s, the target of this intervention was patients with BPD, a group who were often regarded as ‘untreatable’. However, MBT has subsequently been used with patients facing a variety of mental health symptoms, including common symptoms like anxiety and depression. In this chapter, we will describe the development of MBT over time, including variants such as dynamic interpersonal therapy. We will also consider some remaining questions about MBT, including the potential for MBT to harm particular groups of patients.

The development of mentalization-based therapy

Fonagy has recalled the origin of MBT in clinical discussions with George Moran in the late 1980s. Moran was the charismatic director of the Anna Freud Centre, and had a practice with adolescents with poor control of their diabetes. Fonagy and Moran used to meet on Saturdays to discuss and seek advice on their adolescent clinical cases: ‘We gradually realized that all the interpretations we were giving didn’t make any difference to these young people. But once we started talking to them about their lives and what their thoughts and feelings were in very simple language—their minds changed and their diabetic control got better.’2 Moran’s approach aligned with aspects of Hurry’s own clinical technique in her analysis of (p. 217) Fonagy as an adolescent, and with Fonagy’s thinking about adult patients such as Mr S. (see Chapter 1). These experiences revealed the potential benefits, including relief of symptoms, that seemed to stem from patients’ gaining perspective on the sources and scope of their feelings and thoughts.3

In the early 1990s, Fonagy had become convinced of the value of manualized approaches to therapeutic intervention: for the sake of conducting research evaluation; to facilitate transparency to stakeholders and commissioners of services; in order to establish an explicit model for teaching and further improvement; and to help buffer the effects of clinicians’ own anxieties and the interpersonal pressures generated in the consulting room.4 He perceived a gap in available treatment modalities for BPD. The condition was regarded by many as untreatable.5 And cognitive behavioural therapy (CBT), otherwise the dominant approach in UK mental health services, was struggling to demonstrate effectiveness with this group of patients.6 Bateman and Fonagy regarded mentalization as the ingredient within the psychoanalytic approach central to its treatment efficacy. Psychoanalytic therapies are especially concerned with exploring thoughts and feelings, comparing experiences in the present with constructions about the past, and reducing the potential for overwhelming experiences that might put at risk the capacity to reflect—for instance, through use of the couch.7 However, the mentalization-focused elements of psychoanalytic therapies could be refined, and also extracted from those elements that distract from or hinder mentalizing. As discussed in previous chapters, these latter include the use of speculative interpretations, use of sexuality as a metaphor for other challenges and conflicts, and clinicians’ opacity and reticence with the patient.

In a study published in 1999, Bateman and Fonagy compared the outcomes for 19 patients with BPD treated with an early version of MBT, and 19 matched controls who received treatment as usual under the NHS. The key elements of the treatment were weekly individual MBT and thrice-weekly group analytic psychotherapy. At both 6 months and 18 months after the end of treatment, the patients who received the new therapy had fewer depressive symptoms, engaged in fewer suicidal and self-harming acts, spent fewer days as inpatients, and reported better social and interpersonal functioning.8 In fact, the patients who had received MBT continued to improve over time, whereas the controls did not.9 Though the early version of MBT was initially more expensive, this money was ultimately recouped by follow-up (p. 218) in terms of a large reduction in the patients’ use of community and hospital services.10 Even eight years after the initial intervention, the researchers found that the patients who received MBT showed fewer symptoms, less suicidality, and were more likely to be in employment or education. They had less crisis service use, less use of long-term outpatient mental health services, and were less likely to have social services involvement.11 Many patients who received MBT still faced substantial difficulties requiring forms of professional support, but this was at a lower frequency and over shorter duration than for control participants.

In the early 2000s, Bateman and Fonagy developed a manualized MBT approach, published as a book in 2004.12 The process by which this manual was developed is relatively opaque.13 However, in texts from the period, Bateman and Fonagy reflected that ‘the overall goals of treatment are to stabilize the self-structure through the development of stable internal representations, formation of a coherent sense of self, and capacity to form secure relationships. But the self-structure is destabilized in the context of emotional turmoil and so a further goal is identification and appropriate expression of affect.’14 These goals were sought in the new modality by carefully supporting patients to articulate and consider their thoughts and feelings and those of others, without becoming overwhelmed, impatient, or entering into pretend mode.

In general terms, psychoanalytic approaches to therapy often centre attention on unconscious conflicts and fantasies, and the role of defences in holding disturbing thoughts or feelings out of consciousness. By contrast, MBT prioritizes present-day experience and near-conscious thoughts and feelings that can, perhaps with help, be identified by the patient themselves and considered together with other experiences.15 The training requirements for an MBT therapist are very modest compared with psychoanalytic training, contributing to greater scalability.16 And whereas classical psychoanalytic technique encouraged reticence, (p. 219) MBT adopted a more collaborative relationship between therapist and patient, with greater attempt by the therapist to make his or her thoughts and feelings legible for discussion and consideration. This collaboration is initiated through active pursuit of the inquisitive stance, expressions of curiosity, and the elicitation of detailed descriptive accounts by the patient of their experiences. The tone of interactions should aim to support the patient in achieving some degree of emotional equilibrium—sufficient to find the therapeutic encounter safe enough to consider thoughts and feelings.

Bateman and Fonagy argued that consideration of thoughts and feelings should place them as fully as possible in context, in order to understand their applicability in accounting for the patient’s behaviour and experience, as well as the behaviour and experience of others. Together, the patient and therapist attempt to elaborate upon descriptions, identifying causal sequences and the role played by thoughts and feelings within them. Within this process, ‘simply looking at a feeling and its antecedents and consequences is not enough. The patient must be helped to consider who engendered the feeling and how, to explore whether the feelings have occurred or are connected to events either in the recent or longer-term past, to assess the appropriateness of the feeling to any given situation in terms of others’ understanding of the patient, and to establish the appropriate locus of these feelings within current relationships, either past or present.’17 As part of this process, patients are asked to compare their thoughts and feelings at different times, as a step towards recognizing their causes and consequences.

In the course of therapy, the patient should be supported to articulate and consider mental states—even when, and especially when, intense emotions are aroused. This is part of the specific importance attributed to group-based aspects of MBT, because group conditions arouse strong emotions for patients and provide a supported opportunity to attend to their own and others’ mental states. Likewise, the relationship with the therapist is also used as a prompt for patients to identify and examine their thoughts and feelings, especially in the context of considering their assumptions about other people and characteristic patterns of relating to them. Patients’ experiences of the group and of the therapist should first be validated, then explored and elaborated further in collaboration. Alternative or additional accounts can then be considered, as well as the patient’s reaction to these perspectives.

Dynamic Interpersonal Therapy

In the late 2000s, opportunities arose for embedding psychotherapies within primary care in the UK, thanks to the Improving Access to Psychological Therapies initiative.18 This led to a large expansion of the availability of low-intensity CBT. Several psychodynamically inspired approaches also sought to be adopted within UK primary care. This included cognitive analytic therapy (CAT),19 and the Tavistock model.20 Fonagy and colleagues saw an opportunity (p. 220) for MBT, if it could be successfully adapted for delivery in primary care. In 2011, Lemma, Target, and Fonagy published the manual for a brief form of MBT, focused specifically on depression and on ‘anxiety with low mood’.21 They termed this approach ‘dynamic interpersonal therapy’ (DIT). In the model of clinical action underpinning DIT, depression and ‘anxiety with low mood’ are prompted by problems in current close relationships.22 In particular, experiences of impending or actual separation, rejection, loss, or failure were identified as particularly important threats to close relationships if they hinder the capacity of an individual to use those relationships as a secure base and safe haven. The interaction of relationship problems and negative affect may also prompt use of non-mentalizing modes—though generally not to the same degree or pervasiveness as in personality disorders:

DIT’s starting point is rooted in the common clinical observation that patients who present as depressed and/or anxious almost invariably also present with difficulties and distress about their relationships. Although the patient may well experience his problem as ‘I cannot sleep or concentrate’ or ‘I can’t face going into crowded places, or going to work’, the DIT therapist reframes such symptoms of anxiety and depressions as manifestations of a relational disturbance, which the patient cannot understand, or understands in a maladaptive way, attributing to himself or others motivations which are unlikely or unhelpful. Once the patient is helped to make some changes in the way he approaches relationship difficulties, depressive and anxious symptoms are typically alleviated.23

DIT does not attempt to improve all aspects of mentalizing. The focus is on the patient’s capacity to conceive of and reconsider the thoughts and feelings implicated in their present-day perceptual experience and social behaviour. As a brief treatment, the intervention focuses on improving ‘the patient’s capacity to reflect on his own states of mind’, and the patient’s understanding of ‘the connection between his presenting symptoms and what is happening in his relationships’.24 This is achieved over five steps. In the first step, the patient and the therapist identify a problem in one or more of the patient’s relationships, perceived as contributing to depression and/or anxiety. By the end of the early sessions of therapy, the therapist and patient should have at least a tentative understanding of the kinds of relationship the patient tends to create and how this relates to their presenting symptoms. In the second step, the therapist works collaboratively with the patient to create a picture of the thoughts and feelings raised by the problem. In the third step, the therapist encourages the patient to explore alternative ways of considering these thoughts and feelings. The patient–therapist relationship is utilized to help the patient identify their characteristic assumptions about relationships, and ways of responding to others. Fourth, the therapist supports the patient to reconsider these assumptions, and their associated thoughts and feelings. Finally, the (p. 221) therapist presents the patient with a written summary of the collaboratively held view of the shift in perspective achieved over the previous sessions.

Lemma, Target, and Fonagy gave self-representations (see Chapter 6) a fundamental place in DIT. They proposed that short-term work with patients should be oriented by an ‘interpersonal and affective focus’.25 This focus comprises some representation about the self, some representation about others, the emotion that links them, and the psychological function of the configuration. So, for instance, a relevant self-representation may be the feeling that ‘I always ask for too much’. The representation of others may be ‘They are not there when I need them’. The linking emotion may be worry about abandonment. And the psychological function of the configuration may be a pre-emptive rejection of and distrust in others, which protects the individual from hurt and also expresses frustration with others, while also itself contributing to the frustration-evoking situation.

The DIT therapist should not take as their direct aim to improve the accuracy, coherence, or consistency of the patient’s self-representation—for instance, by helping the patient calibrate how much exactly they should ask for from others. Instead, the therapist should begin by asking the patient to describe and characterize the representation of self and other in detail, to produce a shared vocabulary for talking about these.26 For Lemma, Target, and Fonagy, the ‘self’ of the patient is not treated as the primary target of therapeutic intervention. Nonetheless, representations of the self are considered for the contribution they make to characteristic forms of relating, the individual’s perceptual experience is taken as a key object for reflection, and support for mentalizing seeks to sustain the patient’s capacity for attention to their own thoughts, feelings, and/or intentions.

DIT is in several regards closer to cognitive behavioural therapy or cognitive analytic therapy than to MBT.27 This reflects its development for use in primary care mental health delivery and patient groups. Four particular contrasts can be drawn between MBT and DIT. In DIT, the focus of the work is localized to depression or anxiety with low mood; therapy is always provided individually rather than sometimes using a combination of individual and group work; the focus is on mentalizing the self and there is, in principle, little explicit focus on supporting the patient to mentalize the thoughts and feelings of others; and the therapeutic work is less concerned with managing the potential for outbursts and lability of sadness and anger, or helping the patient mentalize in the context of distress. Nonetheless, Fonagy has argued that DIT still bears the defining marks of a mentalization-based therapy:28

  1. 1) Maintain and, when it is lost, regain mentalizing (in both parties).

  2. 2) Active, curious, inquisitive; don’t feign understanding.

  3. 3) Direct joint attention to mental states.

  4. 4) Ordinary/non-expert: avoid guise of privileged knowledge about patients’ mind.

  5. (p. 222) 5) Emphasis on perspective-taking and marking discrepancies between perspectives (and exploring their sources).

  6. 6) ‘Not knowing’ stance: eschew certainty, mark what is not obvious but is presented that way, mark when you do suspect you know.

  7. 7) Model active, intentional effort to find out about opaque mental life.

  8. 8) Humility—acknowledge one’s own non-mentalizing errors, model interest in being corrected and having one’s mind changed.

  9. 9) Doggedness around exploring misunderstandings.

  10. 10) Transparency and authenticity about confusion, puzzlement, and self-reflection.

Many of these are generic aspects of relational psychotherapy, as Fonagy and colleagues have readily acknowledged. To the extent that there is a distinctively MBT quality to DIT, it perhaps can be identified in the proposed focus on sustaining and modelling generative doubt and the inquisitive stance, and avoidance of certainties that can inhibit these. This includes certainties prompted by holding one perspective to the exclusion of others. It also includes certainties prompted by the status of the therapist, and the assumption that he or she has privileged knowledge. The clinician’s mental processes, in their potential wildness, are not made wholly transparent to the patient. However, the clinician does seek to articulate experiences of generative doubt and reconsideration so that the patient can see how this works and why it may be valued.

Though CBT remains the dominant approach, DIT has become a widely adopted treatment modality within UK primary care, as have some other low-intensity psychodynamic therapies such as CAT. Whether the focus on sustaining and modelling generative doubt, which could be argued to distinguish DIT at a theoretical level, leads to actual differences in clinical practice between DIT and other low-intensity psychodynamic therapies remains unknown. Until clear evidence to the contrary is published, we suspect that there is reason for gentle scepticism on this front.

Mentalization-Based Therapy today

Over time, MBT has become a widely used treatment modality in the treatment of personality disorders, with DIT capturing a further share of the market for primary care mental health services.29 By 2014, over two and a half thousand clinicians had been trained in a form of MBT.30 Versions of MBT have been developed and delivered for patients referred on the basis of various diagnoses, including eating disorders, depression, conduct disorder, and anti-social personality disorder. MBT has also been adapted for delivery with children, and as a form of family therapy.31 MBT programmes have become institutionally embedded in (p. 223) many services. For instance, an MBT programme for antisocial personality disorder (ASPD) has been rolled out across the whole of the Irish prison service.32 Even before the COVID-19 pandemic initial attempts had been made to offer MBT via phone or video-link, and in general the modality has proved relatively straightforward to adapt for remote delivery. One challenge has been that Fonagy and colleagues feel that the pandemic has spurred greater epistemic hypervigilance, making it more difficult for many patients to trust and learn from encounters with the clinician. Another challenge has been that many of the therapist’s non-verbal ostensive cues are hindered or blocked by remote communication. However these challenges have been met through an intensified focus on what ostensive cues remain available for signalling recognition of mental states and modulating the patient’s level of arousal.33

Writing in 2016, Bateman and Fonagy remarked that ‘MBT has been more successful than we ever anticipated’.34 They identified several reasons for this growing popularity over the past two decades.

One has been that the modality had a research base. The 1999 study by Bateman and Fonagy, showing clinical benefits and cost-effectiveness despite its small sample, has proven an especially potent resource for the approach over subsequent decades in arguing for use of the modality in cash-strapped and hard-pressed services. Later studies have shown aligned findings, even if reported effects have been weaker.35 If CBT certainly remains dominant within contemporary mental health services, its credentials as an evidence-based treatment have helped MBT find several relevant niches. For instance, ‘clinicians find that some families are not able to benefit fully from the strategies they are offered because they do not (p. 224) have, or struggle to maintain, the capacity for reflective function and emotional regulation that are a prerequisite for their use’.36 In such contexts, MBT is attractive as an alternative evidence-based modality. A second factor identified by Bateman and Fonagy as contributing to the popularity of MBT is that MBT overlaps considerably with other treatment modalities and does not require a high level of specialism to be delivered in an adherent manner. It requires hard work, skill, and practice for a therapist to deftly check their understanding of the patient’s mental state, and whether this corresponds with the patient’s understanding. But it is an elaboration of the skills required by ordinary life, and a more sharply developed version of skills common to most therapeutic modalities.

A further reason Bateman and Fonagy provided for the popularity of MBT is that ‘clinicians easily understand the ideas underpinning the model and recognise that promoting mentalising is something they are already doing in their clinical work’. Clinicians appear to experience MBT as a ‘map through the woods’, helping clinical problems make sense and providing guidance for how to approach them. Though, at the same time, Bateman and Fonagy admit that ‘over the past few years it has become apparent that we were not specific enough about some of the core components of the model’.37 True, as Sandler observed (see Introduction), blurry concepts can be helpful for handling blurry phenomena. Nonetheless, there may also be a price in terms of coherence, and articulation of causal processes. Indeed, part of the appeal of ‘mentalizing’ may have stemmed from the way that the concept can magnetize various investments and meanings, so that the interpretations of different clinicians have at most a family resemblance to one another. One clinician may have supported patients to conceive of feelings implicated in the motivations and intentions of others; another may have supported patients to reconsider their own thoughts in making sense of their past experiences; another may have supported patients to gain skills in reconsidering the observable social behaviour of others and its possible meanings. Any of these, and many others, could legitimately say that they were implementing one of the definitions of mentalizing offered by Fonagy and colleagues (see Chapter 4). And any of these, and many others, would be judged to be showing fidelity to the modality if assessed by Fonagy’s 10 defining markers. Similarly, these different forms of clinical practice would be judged as adherent on the MBT Adherence and Competence Scale.38

A final contribution to the popularity of MBT, listed by Bateman and Fonagy, is that the approach has broad possible applications because it integrates both developmental psychology and social cognition. Again, though, it should be noted that broad applicability has to an extent been purchased at the expense of—mostly inadvertent—construct variance across domains. As we saw in the previous chapter, for example, Fonagy’s 2018 model integrated developmental psychology and social cognition, but in doing so included the concept of epistemic trust twice, and with different meanings on each occasion. So, for instance, one clinician may encourage young children to engage in symbolic play and then help the patient identify feelings implicit in these narratives; another clinician may help adolescents re-evaluate their sense of others’ thoughts about them; a third clinician may focus on helping (p. 225) a patient build trust in others in order to facilitate mentalization; a fourth may focus on helping a patient reduce non-mentalizing in order to reduce epistemic mistrust. All of these can legitimately claim to be mentalization-based approaches to therapy, given that the constructs of mentalization and epistemic trust both serve as switch-points for different and non-overlapping groups of meanings.

The latest version of the manual for delivering MBT was published in 2016.39 It draws on the mature model of mentalizing and non-mentalizing (see Chapters 4 and 5), as well as elements of the account of the self (see Chapter 6). The latest manual also shifts from a deficit discourse about non-mentalizing towards a greater emphasis on the potentially adaptive basis and role of these modes of mental processing in the context of adversities (see Chapter 7). MBT now begins with a 10-week group-based programme, which comprises psychoeducation about mentalization and support for patients in establishing their expectations for therapy. This preliminary psychoeducation programme appears to be experienced by many patients as relevant and useful.40 However some individuals do decide at this point to drop out, rather than initiating individual MBT or individual + group MBT.41 A recent study by Jørgensen reported that patients with BPD who reported low reflective functioning were at increased risk of dropping out of group MBT but not treatment as usual.42 Such findings agree with Batemen and Fonagy’s suspicion that MBT, and especially group MBT, may require some threshold degree of mentalizing skills, and highlights the importance of the introductory psychoeducation to the extent that it can lay the ground for these skills.

Bateman and Fonagy proposed that individual therapy should begin with an assessment of the specific forms of mentalizing that appear to be areas of existing strengths and weaknesses, as part of the development of a shared formulation:

The clinician should draw up a mentalising profile. This involves locating the individual’s style of mentalising on each of the different dimensions, and then considering the relationship between these different ways of functioning on each dimension: that is, do they cause mentalising difficulties to snowball, or do they compensate for each other?43

(p. 226) The weaknesses that should form the particular target of intervention should be those that contribute to such snowballing, as well as those that are especially pertinent to the patient, contributing to jointly agreed goals for the work together. Existing mentalizing strengths can be drawn upon as assets and resources in the therapy, though it should also be considered how they may be used as substitutes or compensation for aspects of mentalizing that have been left underdeveloped. Bateman and Fonagy indicated that mentalization should not be taken to be a unitary capacity, and that discrepancies between the different capabilities under the umbrella of mentalization are very much to be expected. A formulation of the patient’s strengths and weaknesses in areas of mentalizing, and how these relate to current relationships and behaviour, should be written down and shared with the patient. It should then be subsequently revised periodically.

If the patient comes to therapy already with a diagnosis assigned to them, whether by a clinician or self-ascribed, the therapist should help the patient to specify the particular symptoms that cause them trouble, and the history of these symptoms. This will allow the therapist and patient to generate a developmental narrative about how the symptoms emerged. The diagnosis should not in itself be treated as an account or explanation of mental states.44 Building on their recent thinking about adaptation (see Chapter 7), Bateman and Fonagy argued that strategic aspects of symptoms should be recognized before seeking to effect change, in order to minimize potential iatrogenic effects of intervention. In the psychoeducation that begins MBT, patients are taught about the conditional attachment strategies, and supported to reflect on the ways in which their behaviour when hurt, anxious, or distressed resembles one of these patterns. Patients are told that at times they may use one or the other conditional strategy, and sometimes they may feel that they use both together in relatively more or less coherent ways. Patients are not taught about the disorganized classification, though no reason is provided by Bateman and Fonagy for this absence from the curriculum. Perhaps they anticipate that the disorganized attachment classification would be experienced as disempowering by patients.

Bateman and Fonagy claimed that the key task for the MBT therapist is to help the patient balance the four poles of mentalizing: internal and external; affective and cognitive; self and other; implicit and explicit. But they make little of the need to balance explicit with implicit mentalizing, and in fact as we have seen ‘active, intentional effort to find out about opaque mental life’ is one of the defining principles of MBT. A balance between internal and external, and between self and other, translates primarily into two implications. A first is that the clinician should seek to ensure that patients are capable of both. A second is that patients should be capable of withdrawing from one to pursue the other as needed. Rather than ‘balance’, Bateman and Fonagy seem more concerned with the flexible use of capacities.

By contrast, the issue of balance is more clearly evident in relation to affect and cognition. This is because the overarching clinical model of Bateman and Fonagy is that reductions in symptoms arise from supporting patients during clinical sessions to simultaneously mentalize and retain emotional equilibrium. Whereas humanistic approaches to therapy generally regard expressions of empathy for the patient as beneficial, an MBT perspective treats such expressions as appropriate only when they will facilitate mentalization and not (p. 227) contribute to excess intensity or drama in the therapeutic relationship. In fact, when a patient is upset and this prompts forms of non-mentalizing, the MBT clinician should prioritize helping the patient regain their calm, so that interest can then be sparked in mental states. This is termed a ‘contrary move’. Such a move may entail somewhat matter-of-fact responses, a more cognitive angle on the problem, a focus outward to the minds of other people, or moving the patient away from his or her current focus:

The clinician moves emotionally closer to the patient during a session only to the point at which he/she judges the patient is on the verge of losing mentalizing. At this point, he/she moves back, distancing him/herself from the patient, to reduce the level of emotional arousal. Here we encounter a clinically significant paradox—just when the clinician would naturally move emotionally closer to the patient, we ask that he/she moves away.45

This stance shapes the MBT approach to trauma. Many psychotherapeutic modalities regard discussion of past traumas as essential for increasing insight and present-day functioning. Several of Fonagy’s collaborators, such as Jon Allen, have made claims that move in this direction.46 However, for their part, Bateman and Fonagy have stated that traumatic past experiences are given attention in MBT either only when this will support the patient to gain skills in retaining the capacity to mentalize when distressed, or when it will support the development of a self-representation facilitative of mentalization.47 Compared with psychodynamic psychotherapy, in MBT ‘there should be … an avoidance of extensive discussion of past trauma, except in the context of reflecting on current perceptions of the mental states of maltreating figures and on changes in mental state from one’s past as a victim to one’s experience now’.48

Bateman and Fonagy expressed concern that if a story about specific traumatic past experience is created to make sense of present-day symptoms—for instance, a demanding parent as the source of a patient’s contemporary anxiety—this can flatten the complexity of both the past and the present.49 The result may stimulate pretend mode for patients, focusing attention on an origin myth rather than helping the patient to use attention to mental states (p. 228) to account for observable social behaviour and/or perceptual experience.50 This is not to say that memories of trauma recalled during therapy should be regarded solely as constructions or fantasy. As we saw in Chapter 2, child patients whose therapists treated their accounts of abuse and neglect as fantasy had especially poor long-term outcomes in the Anna Freud Centre Retrospective study.51 Fonagy has reflected that traumas recalled in therapy can also serve as important ‘crystallization of an individual’s struggles with his or her circumstances, identity, relationships, and life’, and so may have importance for formulation.52 When a patient experiences something as a felt truth in a therapeutic context—and the impact of a past trauma can certainly serve this function—the effect can be to move the patient closer to the therapist and increase epistemic trust.53

Nonetheless, too exclusive a focus on identifying or remembering a particular trauma as the source of current symptoms can work against the toleration of uncertainty necessary for supporting reflection on mental states and the different perspectives they entail.54 More generally, Bateman and Fonagy have argued, to inhibit entry into hypermentalizing or other forms of pretend mode ‘the clinician should avoid inadvertently creating situations where the patient is forced into talking about mental states that they cannot immediately link to subjectively felt reality’.55 Instead, the therapist should aim to keep conversations centred around whatever is felt to be current and emotionally salient ‘in the patient’s mind, in other words, in working memory’.56 As we will see in Chapter 9, Fonagy and colleagues have very recently questioned whether an individual’s working memory is the only site of mental states, or whether groups or institutions can also have beliefs and atmospheres that can be the target of mentalizing. However, in Bateman and Fonagy’s 2016 manual for MBT, it is the individual alone who can be the subject of mental states and the target of mentalizing.

One of the essential techniques advocated by Bateman and Fonagy is what they have called ‘mentalizing functional analysis’. In light of the clinician’s understanding of the patient’s strengths and weaknesses, the MBT therapist keeps a look out for times when the patient seems to drop out of mentalizing and enter into a non-mentalizing mode of mental processing. At that point, the clinician can ask the patient to ‘stop and re-wind’, and narrate the (p. 229) sequence again, so that the clinician and patient can work collaboratively to identify when and how mentalizing was lost and non-mentalizing began.57 Special attention is paid to the feelings that may stimulate forms of non-mentalizing. The therapist and patient may then reflect together about what strategies could be implemented in future to halt the decay of mentalizing or the advance of non-mentalizing. Bateman and Fonagy recognized, however, that clinicians too may enter non-mentalizing states. For instance, they may find themselves just going along with what is being said, having entered into pretend mode. They may find themselves caught up in feelings that seem so pressing that new information cannot enter, having entered a state of psychic equivalence. They may find their minds dominated by an urgent wish do to something constructive, having entered teleological mode.58 When the patient or clinician perceives that the clinician has started to non-mentalize, Bateman and Fonagy likewise advised use of ‘stop and re-wind’.

‘Stop and re-wind’ was developed as a clinical technique, but it also has theoretical stakes. Over the decades, Fonagy and collaborators have given less and less explicit notice to attentional processes in their thinking about mentalization; it seems to have dropped, unceremoniously, out of sight.59 Attentional processes are not described by Bateman and Fonagy as part of what MBT attempts to change. Nonetheless, ‘stop and re-wind’ can be considered above all as a technique for practising both attentional flexibility and attentional focus as component parts and contributors to mentalizing.

This is all the more surprising because Fonagy and Gergely both attribute the basis of the idea of ‘ostensive cues’ to Sperber and Wilson, for whom such cues are primarily a kind of way of directing the attention of a conversation partner. Sperber appears even to somewhat regret that in highlighting ostension he appears to have drawn away recognition from the direction of others’ attention as the underlying and wider issue.60 In the work of Fonagy and colleagues, concern with ostensive cues has rather supplanted the wider question of attentional processes, which they had considered in the 1990s and early 2000s. In the 2016 MBT manual, Bateman and Fonagy and colleagues highlighted the importance of ostensive cues provided by the therapist to the patient, and of validating the salience and relevance of the patient’s perspective as critical to the initiation of any further discussion of that perspective. For instance, marked mirroring is a powerful ostensive cue: conveying, in the interplay of (p. 230) primary and secondary meanings, both that the therapist acknowledges the patient’s emotion and intentionality, and that the therapist can cope with these responses. Use of these techniques is assumed to ‘increase the patient’s epistemic trust and thus acts as a catalyst for therapeutic success’.61 While these techniques are common to many therapeutic modalities, Fonagy and colleagues anticipated that MBT could give them particular weight. In fact, they might even form the basis in the future for forms of MBT delivered over the internet, because Fonagy and colleagues regard ostensive cues rather than therapeutic alliance as reigniting epistemic trust and the capacity to learn from experience.62

This concern with ostensive cues has also led to theoretical revision. As we saw in the previous chapter, Fonagy and colleagues had initially described ostensive cues as behaviours that acknowledge the i) mental states; ii) intentionality; and iii) individuality of the subject. Initially, these properties of ostensive cues were extrapolated from observations of infants to later development—for instance, to adult practices that entail mutual recognition of intent. However, in a 2019 chapter, Fonagy, Allison, and Campbell have argued that, after infancy, the perception of ‘understanding of the individual’s personal narrative by another person creates a potential for epistemic trust’.63 Above all, ‘recognition of subdominant narratives is a particularly potent way of establishing epistemic trust’.64 They define personal narrative as a person’s ‘imagined sense of self’ at a given moment, and identify that the clarity and specificity of this narrative is in significant part a function of mentalizing of the self.65 The imagined sense of self is nonetheless complex and heterogenous, inflected by a person’s developmental history, their cultural context, the pragmatic demands of their immediate circumstances, alien intentions from the psychoanalytic unconscious, and sometimes significant (p. 231) involvement of the primary unconscious (see Chapter 6). There is considerable scope, therefore, for subdominant threads that are knowable by an individual and by others, but that do not ordinarily feature as part of an individual’s espoused self-representation.66 Epistemic trust is facilitated by the felt acuity of the match between some part of our self-representation and the account of us offered by another. A lack of clarity and specificity in this narrative will render individuals vulnerable to missing accurate matches (epistemic hypervigilance), or incorrectly identifying accurate matches (epistemic credulity), or both (see Chapter 7).

The characterization of others’ perceived understanding one’s personal narrative as an ostensive cue is innovative. It is not implied by the previous concept of ostensive cues as behaviours that acknowledge the i) mental states; ii) intentionality; and iii) individuality of a subject. The ‘imagined sense of self’ depicted by a personal narrative is not the sum of thoughts and feelings, though it plays a role in generating them and partly takes its content from them. Nor is it merely the intentionality of the subject, though this may play a part. The claim that epistemic trust is generated by recognition of personal narrative rather than acknowledgement of mental states and intentionality appear to signal a wish by Fonagy and colleagues, not as yet fully pursued, to bring to bear more of the complex account of the self on matters of clinical technique. Given Fonagy’s criticisms of the Bakhtin-inspired model of dominant and subdominant personal narratives in CAT in the 1990s (see Chapter 6), it would be interesting to see how Fonagy and colleagues would now regard this model of personal narratives and their role in clinical technique.67 This would be of particular note given Fonagy and colleagues’ growing interests in culture. The Bakhtinian approach in CAT holds that personal narrative is woven out of cultural resources. From a mentalizing perspective, such resources may suggest or impose codes that relatively hinder or facilitate the construction, identification, modulation, and expression of mental states.

Fonagy and colleagues have been continually active in evolving and developing MBT. Meanwhile, trials have been accumulating that have assessed the effectiveness of the modality in its different incarnations. Most of these, though certainly not exclusively, have focused on its application with adult patients with BPD. True, the relationship between mentalizing difficulties and BPD may be less tight than initially anticipated by Fonagy. For instance, Bouchard, Target, Lecours, Fonagy, and colleagues have found only an association of r = -.24 between symptoms of personality disorder and reflective function scored on the Adult Attachment Interview.68 Nonetheless, the importance of mentalizing difficulties has (p. 232) been indicated by the finding that, in the large majority of studies, patients with personality disorders benefit from MBT and show fewer symptoms and better day-to-day functioning by the end of the trial. The same is also generally true of other comparison interventions when patients have BPD alone, without other conditions. Fonagy and colleagues have interpreted this finding with a speculative proposal: that even if the models are different, the use of any evidence-based model by therapists gives coherence to the way the patient is recognized as an intentional agent by the therapist, which serves as an ostensive cue and facilitates epistemic trust.69

However, for patients with complex personality disorders, MBT has proven the most effective psychological modality.70 This may imply that the difficulties with mentalizing identified initially by Fonagy and colleagues as especially characteristic of BPD are in fact less specific. This would seem supported by the Sharp and colleagues’ study from 2015 mentioned in the previous chapter, in which BDP loaded with the general p-factor. To untangle these questions, it would be necessary to understand the relative role of mentalization in mediating clinical and functional outcomes for patients, and whether this differs between treatment modality. Fonagy and colleagues have argued that mentalization is improved by many talking therapies, and should be treated as the central target for effecting therapeutic change. In support for this claim, De Meulemeester found that improvement on the Reflective Functioning Questionnaire scale for indiscriminate uncertainty (RFQu) among patients receiving hospital-based psychoanalytic treatment was associated to the highest degree with relief from symptoms (r = .89).71 The remarkable strength of this association is such that, if replicated, it suggests that RFQu and relief from symptoms may even be part (p. 233) of the same underlying construct. However, to the surprise of the researchers, there was no link with the RFQ certainty scale, though, in general, the RFQc scale has not performed as anticipated across several studies (see Chapter 7). It is also notable that a major recent meta-analysis found that mentalization-based parenting interventions result in little change in reflective functioning.72 Such findings raise the question of the extent to which mentalization in general is indeed the mechanism of change as anticipated by Fonagy, or whether further refinement might be needed in specifying and measuring how MBT is anticipated to work, perhaps with greater focus on indiscriminate uncertainty about mental states.

As evidence that the effectiveness of MBT is mediated by mentalization, Fonagy and colleagues have frequently referred to a trial of MBT for self-harming adolescents by Rossouw.73 To measure mentalization, Rossouw and Fonagy used the ‘How I Feel’ Questionnaire, which is a self-report assessment of emotion regulation. Participants report on the frequency, intensity, and control of feelings such as sadness, fear, and anger. It would be expectable that there would be shared variance between the ‘How I Feel’ Questionnaire with self-reported self-harm. For self-reported emotion regulation to partially mediate associations between treatment and self-reported self-harm cannot be regarded as proof that mentalization mediated the effectiveness of the intervention. Furthermore, the model tested included self-reported attachment avoidance as an additional mediator, making the mediation by self-reported emotion regulation difficult to interpret with confidence. The researchers did not report whether scores on How I Feel mediated effects of the intervention on symptoms of BPD or on depression, though both were measured, and such tests were indicated by Rossouw’s own developmental model of mediating factors.74 The Rossouw study also has among the highest attrition rates of any trial of MBT published to date, and no assessment was made of whether participants who left the study differed from participants who remained in therapy to completion. Overall, the Rossouw study cannot be treated as evidence of the role of mentalizing in mediating the effects of MBT on outcomes. This should be regarded as still yet to be tested.

The potential for harm

One outstanding question faced by MBT is the potential for harm to patients stemming from the therapy—iatrogenesis. Fonagy and colleagues have been vocal about the potential for other treatment modalities to cause harm to patients. As we saw in Chapter 1, Fonagy had particular concern with the ways that classical psychoanalytic technique may at times actively contribute to problems with emotion modulation for patients with mentalizing difficulties.75 Fonagy and colleagues have in principle acknowledged the potential for iatrogenesis in mentalization-based therapies. But this has generally been only with passing references, and they have not developed these reflections into hypotheses for testing.76 The (p. 234) sole apparent exception to the lack of explicit hypothesis generation about the potential for iatrogenesis from MBT is a proposal from 2005 by Fonagy and Target. They stated that declines in use of psychic equivalence lead to an ‘almost inevitable worsening of the patient’s symptoms’ so long as mentalizing skills remain weak, because the experience of personal coherence permitted by psychic equivalence is initially undermined.77 However, this hypothesis has not been operationalized and tested, or even revisited. In part, this is likely to have been due to the available measures, which generally treat non-mentalizing as the opposite of mentalizing, and so are not able to characterize a state in which i) non-mentalizing has declined but ii) mentalizing has yet to increase.78

Moving from consideration of theory to research findings, one empirical report suggesting a price attached to mentalizing appeared in the Anna Freud Centre retrospective study (Chapter 2). Using the Adult Attachment Interview, the researchers found that ‘earned secure people may be reflective and functional, but they tend to be somewhat more vulnerable to depression than their unearned secure peers. The children who are most reflective about their families are also more worried about them.’79 Other results suggestive of iatrogenesis have started to accumulate. For instance, Suchman and colleagues found that strengths in mentalizing the self in substance-using mothers were positively associated with depression (r = .41). Strengths in mentalizing their child (r = .25) were also associated with depression.80 The study was cross-sectional so it is difficult to interpret causality, but it may be hypothesized that greater awareness of mental states is painful for substance-using parents. In another study, Stacks and colleagues likewise found a positive association between mentalizing (p. 235) scores and symptoms of depression among mothers with childhood maltreatment histories (r = .29).81 Such findings raise questions about whether MBT has the potential for iatrogenesis related to depression for patients with historic or concurrent vulnerabilities.

An obstacle to understanding the potential for iatrogenesis from MBT has been the multiple and diffuse ways in which the term ‘mentalizing’ has been used by Fonagy and colleagues. In light of the specifications in Chapter 4, passing remarks by Fonagy and colleagues may be drawn together to suggest that MBT has three potential bases for iatrogenesis. A first is the potential for mentalization to make individuals more open and vulnerable to harmful environments. This will be dealt with at length in the next chapter. A second potential basis for iatrogenesis has been the particular concern of Target, perhaps stemming from her reflections on the Anna Freud Centre Retrospective Study. For instance, Rizq and Target have expressed concern that ‘where high levels of RF [reflective functioning] tip over into anxious and depressive ruminations, they may unhelpfully sustain a preoccupation with the self, rather than with another’s experience and needs’. This ‘suggests that personal therapy for some may sponsor an unhelpful dwelling on inter and intra-personal dynamics.’82 Fonagy has agreed: ‘The idea of long-term deep introspection, of really studying your own mind, might be good for the training of therapists. Is it helpful for people with mental health problems? I think it probably does harm.’83 The notion of mentalization as a ‘focus’ on mental states appears in some definitions, particularly towards the end of the 2000s.84 However, as we saw in Chapter 3, focus on mental states—mind-mindedness—is perhaps best regarded as a potential correlate of mentalization, rather than a necessary element.

Fonagy and colleagues have acknowledged that mind-mindedness ‘is likely to be one of those parental attributes that is most adaptive in moderation’.85 The potential for improved mentalizing may prompt anxious and depressive ruminations, whether about the self or about others. This may be in the form of hypermentalization, in which the content of the ruminations is speculative. However, the rumination could just as well be grounded in accurate attention to the specifics of present or past perceptual experience.86 The problem lies (p. 236) not in the inaccuracy of the mind-mindedness, but in the lack of the qualities of tentativeness and generative uncertainty, and in the ability to use mentalizing as a capacity, i.e. to turn it to simmer when it is not helpful.87 Within publications associated with MBT, Fonagy and colleagues offer guidance for responding to hypermentalizing and promoting mentalization. However, they do not attend to the potential for the mind-mindedness prompted by mentalization to contribute to realistic forms of rumination.

A third particular threat for iatrogenesis lies in the multiple uses of insight into motivations and intentions. In the early 2000s, Judy Dunn and colleagues had found that skills in social cognition were associated with heightened sensitivity to criticism among young children. Sutton and colleagues also documented strengths in social cognition among ringleader bullies.88 Citing these findings in a paper from 2006, Fonagy and Target observed that ‘the possession of the capacity to mentalize is neither a guarantee that it will be used to serve pro-social ends, nor a guarantee of protection from malign interpersonal influence. The acquisition of the capacity to mentalize may, for example, open the door to more malicious teasing, increase the individual’s sensitivity to relational aggression, or even mean that they take a lead in bullying others.’89 Though Fonagy and Target framed this as a concern about acquisition of the capacity to mentalize in general, in fact the research they are citing is about social cognition and theory of mind. The problem would appear to lie, not in mentalization in general, but specifically in the multiple uses of skills in interpreting the thoughts and feelings of others.

Bateman and Fonagy reflected further on these matters in thinking about the delivery of MBT for patients with ASPD. They observed that ASPD can be associated with strengths in understanding the thoughts and motivations of others, despite difficulties in mentalizing the self, and particularly their own feelings.90 Bateman and Fonagy offered the conjecture that ‘the sacrifice of certain types of mentalizing frees up capacity for other domains to develop’.91 This potential for trade-offs between cultivations of kinds of mentalization is important, though it has been masked by reification of mentalization as a unitary process, and inattention to the kinds of motivation and fantasy that prompt forms of mentalizing and non-mentalizing. Bateman and Fonagy encouraged the rebalancing of mentalization in cases where there have been trade-offs, but do not consider the specific potential for iatrogenesis when rebalancing proves challenging. They also do not consider the role of (p. 237) uncivilized motivations and fantasies in prompting mentalizing, and of beneficent motivations in prompting non-mentalizing.

There can be cases, then, where MBT may primarily serve to further facilitate understanding of the thoughts and motivations of others, honing existing skills in relational aggression. A recent study by Gillespie, Kongerslev, Sharp, and colleagues has followed up on this concern. This work builds from Sharp’s previous collaboration with Fonagy on hypermentalizing in adolescents with BPD. Gillespie and colleagues conducted a study with 80 male adolescents incarcerated mostly for violent offences. They assessed participants’ ability to recognize emotions in others based on images of the eye region, intended as a measure of skills in mentalizing their display of feelings. Participants also completed questionnaire measures of aggression, psychopathic tendencies, and personality disorder. Success in identifying emotions from just the images of the eye region was associated with self-reported proactive aggression against others—but not reactive aggression—even controlling for psychopathic tendencies. The authors theorized that, when an individual has weak access to, and knowledge of, their own feelings, they are likely to also ‘show problems in emotional resonance, that is, the ability to feel what another is feeling.’92 Despite previous reservations about the word (see Chapter 3), this capacity has recently been described by Bateman, Fonagy, and Campbell as ‘empathy’.93 Gillespie and colleagues argued that knowledge of the feelings of others can be used as a resource for relational aggression. The authors urged recognition that MBT may be actively harmful for patients with both conduct problems and psychopathic tendencies.94 A similar concern has been raised about MBT by other commentators, such as Crittenden and Landini.95


1 Fonagy, P. (1998). ‘Moments of Change in Psychoanalytic Theory: Discussion of a New Theory of Psychic Change’. Infant Mental Health Journal, 19(3): 346–353, pp. 351–352.

2 Ezrati, O. (2014). ‘Freud Off: Giving New Meaning to Psychoanalysis’, Haaretz, 5 April. Accessed at: Another retrospective account is offered by Fonagy and colleagues in their REF 2014 Impact Case Study: ‘The research originated in the 1990s when Fonagy and colleagues discovered that juvenile diabetics with poor insulin control struggled to depict their social experiences accurately in mental state terms and that this lack of “mentalising” capacity meant they often could not predict the consequences of their own and others’ actions. Listening to them talk about their feelings and their understanding of feelings dramatically improved their diabetic control (measured through glycosylated haemoglobin levels). Clinical work both with these young people and BPD patients suggested that failure of mentalising often followed a combination of early neglect and childhood trauma.’ University College London (2014). ‘REF 2014: Psychology, Psychiatry and Neuroscience Impact Case Study’. Accessed at:

3 For recent evidence supporting Moran’s findings with patients with brittle diabetes, see Costa-Cordella, S., Luyten, P., Cohen, D., Mena, F., and Fonagy, P. (2020). ‘Mentalizing in Mothers and Children with Type 1 diabetes’. Development and Psychopathology, Early View.

4 Fonagy, P. and Target, M. (1996). ‘Should we Allow Psychotherapy Research to Determine Clinical Practice? Comments on Sol J. Garfield: “Some Problems Associated with “Validated” Forms Of Psychotherapy”’. Clinical Psychology: Science and Practice, 3: 245–250; Fonagy, P. (1999). ‘Achieving Evidence-Based Psychotherapy Practice: A Psychodynamic Perspective on the General Acceptance of Treatment Manuals’. Clinical Psychology: Science and Practice, 6(4): 442–444.

5 In some quarters, these attitudes towards personality disorder have continued. See e.g. Chartonas, D., Kyratsous, M., Dracass, S., Lee, T., and Bhui, K. (2017). ‘Personality Disorder: Still the Patients Psychiatrists Dislike?’. BJPsych Bulletin, 41(1): 12–17; Beryl, R. and Völlm, B. (2018). ‘Attitudes to Personality Disorder of Staff Working in High‐Security and Medium‐Security Hospitals’. Personality and Mental Health, 12(1): 25–37.

6 Leichsenring, F. and Leibing, E. (2003). ‘The Effectiveness of Psychodynamic Therapy and Cognitive Behavior Therapy in the Treatment of Personality Disorders: A Meta-Analysis’. American Journal of Psychiatry, 160: 1223–1232.

7 For further discussion of the mentalizing elements of psychoanalysis, see especially Fonagy, P. and Adshead, G. (2012). ‘How Mentalisation Changes the Mind’. Advances in Psychiatric Treatment, 18(5): 353–362.

8 Bateman, A. and Fonagy, P. (1999). ‘Effectiveness of Partial Hospitalization in the Treatment of Borderline Personality Disorder—A Randomized Controlled Trial’. American Journal of Psychiatry, 156: 1563–1569.

9 Bateman, A. and Fonagy, P. (2001). ‘Treatment of Borderline Personality Disorder with Psychoanalytically Oriented Partial Hospitalization: An 18-Month Follow-Up’. American Journal of Psychiatry, 158: 36–42.

10 Bateman, A. and Fonagy, P. (2003). ‘Health Service Utilization Costs for Borderline Personality Disorder Patients Treated with Psychoanalytically Oriented Partial Hospitalization versus General Psychiatric Care’. American Journal of Psychiatry, 160(1): 169–171.

11 Bateman, A. and Fonagy, P. (2008). ‘8-Year Follow-Up of Patients Treated for Borderline Personality Disorder: Mentalization-Based Treatment Versus Treatment As Usual’. American Journal of Psychiatry, 165(5): 631–638; Bateman, A., Constantinou, M. P., Fonagy, P., and Holzer, S. (2020). ‘Eight-Year Prospective Follow-Up of Mentalization-Based Treatment versus Structured Clinical Management for People with Borderline Personality Disorder.’ Personality Disorders: Theory, Research, and Treatment., Early View. A first empirical study of the role of mentalization within treatment was conducted by Chiesa and Fonagy, who studied 300 patients in personality disorder services: from the Cassel Hospital Residential inpatient programme; from a therapeutic outpatient programme; and management as usual by the Devon personality disorder services. None of these services were mentalization-based treatments. However, Chiesa and Fonagy found that patient reflective functioning on the Adult Attachment Interview was improved by the outpatient programme, which also had the best patient outcomes by a six-year follow-up. By this point, 62% of patients in the outpatients intervention were below the clinical cut-point for a personality disorder, compared with 26% for the inpatient intervention and 13% in treatment as usual. Chiesa, M., Fonagy, P., and Holmes, J. (2006). ‘Six-Year Follow-Up of Three Treatment Programs to Personality Disorder’. Journal of Personality Disorders, 20(5): 493–509.

12 Bateman, A. and Fonagy, P. (2004). Psychotherapy for Borderline Personality Disorder: Mentalization-Based Treatment, Oxford: Oxford University Press.

13 Some historical remarks are presented in Fonagy, P. and Bateman, A. (2009). ‘A Brief History of Mentalisation-Based Treatment and its Roots in Psychoanalytic Theory and Practice’, in M. Brownescombe Heller and S. Pollet, (eds), The Work of Psychoanalysts in the Public Health Sector, London: Routledge, pp. 156–176.

14 Bateman, A. W. and Fonagy, P. (2003). ‘The Development of an Attachment-Based Treatment Program for Borderline Personality Disorder’. Bulletin of the Menninger Clinic, 67: 187–211, p. 195.

15 Sandler can be regarded as an intermediary figure here, because his account has elements that correspond much more to a classic psychoanalytic account (e.g. drives, discussion of the ego), and some elements that seem to presage MBT, such as his central concern with the preconscious. Fonagy, P. (2005). ‘An Overview of Joseph Sandler’s Key Contributions to Theoretical and Clinical Psychoanalysis’. Psychoanalytic Inquiry, 25(2): 120–147.

16 This may be one contributing reason why the British Psychoanalytic Council have treated Fonagy’s research and teaching on MBT as invalid activity for the purposes of continuing professional development as a psychoanalyst: despite many points of commonality, MBT is not framed as operating on the same epistemic object as psychoanalytic approaches. Fonagy, P. and Allison, E. (2016). ‘Commentary on Kernberg and Michels’. Journal of the American Psychoanalytic Association, 64(3): 495–500.

17 Bateman, A. W. and Fonagy, P. (2003). ‘The Development of an Attachment-Based Treatment Program for Borderline Personality Disorder’. Bulletin of the Menninger Clinic, 67: 187–211, pp. 196–197.

18 Clark, D. M. (2018). ‘Realizing the Mass Public Benefit of Evidence-Based Psychological Therapies: The IAPT Program’. Annual Review of Clinical Psychology, 14.

19 Ryle, A. and Kerr, I. B. (2003). Introducing Cognitive Analytic Therapy: Principles and Practice, John Wiley & Sons.

20 Malan, D. and Della Selva, P. C. (2007). Lives Transformed: A Revolutionary Method of Dynamic Psychotherapy, Karnac Books.

21 Lemma, A., Target, M., and Fonagy, P. (2011). Brief Dynamic Interpersonal Therapy: A Clinician’s Guide, Oxford: Oxford University Press.

22 See also Luyten, P., Fonagy, P., Lemma, A., and Target, M. (2012). ‘Depression’, in A. W. Bateman and P. Fonagy (eds), Handbook of Mentalizing in Mental Health Practice, Washington, DC: American Psychiatric Publishing, pp. 385–418: ‘The basic assumption of the mentalisation-based approach to depression is that depressive symptoms reflect responses to threats to attachment relations, and thus threats to the self, either because of (impending) separation, rejection or loss; (impending) failure experiences; or a combination of both … Moreover depressed mood leads to further increases in arousal and stress levels, resulting in further impairments and distortions in mentalisation, which in turn lead to a loss of resilience’ (p. 389).

23 Lemma, A., Target, M., and Fonagy, P. (2011). Brief Dynamic Interpersonal Therapy: A Clinician’s Guide, Oxford: Oxford University Press, p. 52.

24 Ibid. 63.

25 Ibid. 107–108.

26 Ibid. 116.

27 Bateman and Fonagy argue that CAT, unlike MBT, lacks a developmental perspective and does not use psychoanalytic interpretations. These criticisms seem arguable as grounds of contrast, both for MBT and DIT. Bateman, A. W. and Fonagy, P. (2004). Psychotherapy for Borderline Personality Disorders: Mentalization Based Treatment, Oxford: Oxford University Press, pp. 129–132. In recent years, there has been some mention of cognitive mentalising therapy, as an explicit attempt to blend MBT and CBT. However, references have generally been passing, and there has been no justification offered regarding the intended benefits of cognitive mentalising therapy compared with its parent modalities. The fullest description to date appears in Bales, D. L. and Bateman, A. W. (2012). ‘Partial Hospitalization Settings’, in A. W. Bateman and P. Fonagy (eds), Handbook of Mentalizing in Mental Health Practice, Washington, DC: American Psychiatric Publishing, pp. 197–227.

28 Fonagy, P. (2019). ‘Antisocial PD and Mentalisation Based Treatment.’ Paper presented on 3 July to the International Congress of the Royal College of Psychiatrists, London.

29 Work is also underway to elaborate and test a longer version of DIT for chronic depression with attendant other mental health symptoms, especially relevant to cases where other treatments have been attempted without success: the researchers consider the ‘treatment-resistant features of the complex case of depression to be rooted in difficulties with mentalizing and epistemic mistrust’, Rao, A. S., Lemma, A., Fonagy, P., Sosnowska, M., Constantinou, M. P., Fijak-Koch, M., and Gelberg, G. (2019). ‘Development of Dynamic Interpersonal Therapy in Complex Care (DITCC): A Pilot Study’. Psychoanalytic Psychotherapy, 33(2): 77–98, p. 82.

30 University College London (2014). ‘REF 2014: Psychology, Psychiatry and Neuroscience Impact Case Study’. Accessed at:

31 Midgley, N., Ensink, K., Lindqvist, K., Malberg, N., and Muller, N. (2017). ‘Mentalization-Based Treatment for Children: A Time-Limited Approach’. American Psychological Association. Accessed at:; Asen, E. and Midgley, N. (2019). ‘Mentalization-Based Approaches to Working with Families’, in Anthony Bateman and Peter Fonagy (eds), Handbook of Mentalising in Mental Health Practice (2nd edn), Washington, DC: American Psychiatric Association, pp. 136–149.

32 Bateman, A., Fonagy, P. and Campbell, C. (2019). ‘Antisocial Personality Disorder in Community and Prison Settings’, in Anthony Bateman and Peter Fonagy (eds), Handbook of Mentalising in Mental Health Practice (2nd edn), Washington, DC: American Psychiatric Association, pp. 335–349, p. 347.

33 Fonagy, P., Campbell, C., Truscott, A., and Fuggle, P. (2020). ‘Debate: Mentalising remotely–The AFNCCF’s adaptations to the coronavirus crisis’. Child and Adolescent Mental Health, 25(3): 178–179. Ventura Wurman, T., Lee, T., Bateman, A., Fonagy, P., and Nolte, T. (2020). ‘Clinical management of common presentations of patients diagnosed with BPD during the COVID-19 pandemic: the contribution of the MBT framework’. Counselling Psychology Quarterly, Early View. Fisher, S., Guralnik, T., Fonagy, P., and Zilcha-Mano, S. (2020). ‘Let’s face it: video conferencing psychotherapy requires the extensive use of ostensive cues’. Counselling Psychology Quarterly, Early View. Lassri, D. and Desatnik, A. (2020). ‘Losing and regaining reflective functioning in the times of COVID-19: Clinical risks and opportunities from a mentalizing approach’. Psychological Trauma: Theory, Research, Practice, and Policy, 12(S1): S38.

34 Bateman, A. W. and Fonagy, P. (2016). Mentalization-Based Treatment for Personality Disorders: A Practical Guide (2nd edn), Oxford: Oxford University Press, p. v.

35 The findings from the 1999 study stand out in part against the relatively poor backdrop of services at the time. Later trials of MBT have found weaker effects, as have all other trials of treatments for personality disorders. Cristea, I. A., Gentili, C., Cotet, C. D., Palomba, D., Barbui, C., and Cuijpers, P. (2017). ‘Efficacy of Psychotherapies for Borderline Personality Disorder: A Systematic Review and Meta-Analysis’. JAMA Psychiatry, 74(4): 319–328. The most probable interpretation is that treatment as usual has converged with the components of many of the branded therapies, making it more difficult for branded interventions to demonstrate effectiveness, though other factors may certainly also be implicated in these declines. Fonagy has also speculated that diagnosis-focused and manualized therapies may have also hindered efficacy by reducing attention to the specificities of the patient and their mental states, perhaps even encouraging entry into teleological mode. Publication bias and selective reporting may also play a role. For discussions, see Young, N. S., Ioannidis, J. P., and Al-Ubaydli, O. (2008). ‘Why Current Publication Practices may Distort Science’. PLoS Medicine, 5(10): e201; Fonagy, P., Luyten, P., and Bateman, A. (2017). ‘Treating Borderline Personality Disorder with Psychotherapy: Where Do We Go From Here?’. JAMA Psychiatry, 74(4): 316–317; Fonagy, P. (2019). ‘Why is it So Hard to Learn to Do Things Differently? On Not Being Able to Learn from Experience’. GAP Call-In Series Podcast. Accessed at:; Weisz, J. R., Kuppens, S., Ng, M. Y., Vaughn-Coaxum, R. A., Ugueto, A. M., Eckshtain, D., and Corteselli, K. A. (2019). ‘Are Psychotherapies for Young People Growing Stronger? Tracking Trends over Time for Youth Anxiety, Depression, Attention-Deficit/Hyperactivity Disorder, and Conduct Problems’. Perspectives on Psychological Science, 14(2): 216–237.

36 Allison, E. and Campbell, C. (2019). Transforming Child Mental Health: Principles of Sustainable Development, London: Anna Freud Centre.

38 Karterud, S., Pedersen, G., Engen, M., Johansen, M. S., Johansson, P. N., Schlüter, C., … and Bateman, A. W. (2013). ‘The MBT Adherence and Competence Scale (MBT-ACS): Development, Structure and Reliability’. Psychotherapy Research, 23(6): 705–717. See also Muñoz Specht, P., Ensink, K., Normandin, L., and Midgley, N. (2016). ‘Mentalizing Techniques Used by Psychodynamic Therapists Working with Children and Early Adolescents’. Bulletin of the Menninger Clinic, 80(4): 281–315.

39 Bateman, A. W. and Fonagy, P. (2016). Mentalization-Based Treatment for Personality Disorders: A Practical Guide (2nd edn), Oxford: Oxford University Press.

40 Ditlefsen, I. T., Nissen-Lie, H. A., Andenæs, A., Normann-Eide, E., Johansen, M. S., and Kvarstein, E. H. (2020). ‘ “Yes, there is actually hope!”—A qualitative investigation of how patients experience mentalization-based psychoeducation tailored for borderline personality disorder’. Journal of Psychotherapy Integration, Early View.

41 Barnicot and Crawford report a 64% drop-out rate by the end of the group-based programme. It is not clear how representative this is of other implementations of MBT. Barnicot, K. and Crawford, M. (2019). ‘Dialectical Behaviour Therapy v. Mentalisation-Based Therapy for Borderline Personality Disorder’. Psychological Medicine, 49(12): 2060–2068. Other studies have tended to report attrition once individual MBT begins. When attrition is reported at all, rates vary between 7% (Brune, Dimaggio, and Edel, 2013) and 72% (Robinson et al., 2016): Brune, M., Dimaggio, G., and Edel, M. A. (2013). ‘Mentalization-Based Group Therapy for Inpatients with Borderline Personality Disorder: Preliminary Findings’. Clinical Neuropsychiatry, 10: 196–201; Robinson, P., Hellier, J., Barrett, B., Barzdaitiene, D., Bateman, A., Bogaardt, A., . . . and Kern, N. (2016). ‘The NOURISHED Randomised Controlled Trial Comparing Mentalisation-Based Treatment for Eating Disorders (MBT-ED) with Specialist Supportive Clinical Management (SSCM-ED) for Patients with Eating Disorders and Symptoms of Borderline Personality Disorder’. Trials, 17(1): 549.

42 Jørgensen, M. S., Bo, S., Vestergaard, M., Storebø, O. J., Sharp, C., and Simonsen, E. (2021). ‘Predictors of dropout among adolescents with borderline personality disorder attending mentalization-based group treatment’. Psychotherapy Research, Early View. On dropout from MBT, see also Andersen, C. F., Poulsen, S., Fog-Petersen, C., Jørgensen, M. S., and Simonsen, E. (2020). ‘Dropout from mentalization-based group treatment for adolescents with borderline personality features: A qualitative study’. Psychotherapy Research, Early View.

43 Bateman, A. W. and Fonagy, P. (2016). Mentalization-Based Treatment for Personality Disorders: A Practical Guide (2nd edn), Oxford: Oxford University Press, p. 115.

44 Addressing mental states merely as effects of diagnoses has been characterized as poor reflective functioning. Fonagy, P., Target, M., Steele, H. and Steele, M. (1998). Reflective Functioning Manual, Version 5, London: UCL/Anna Freud Centre: ‘The use of diagnostic terminology, or reference to mental illness, should be considered very carefully, and on the whole rated low, if this is the sole explanation offered for the caregiver’s behaviour, and the specific mental states of caregiver and other persons affected are not specified’ (p. 21).

45 Bateman, A. W. and Fonagy, P. (2016). Mentalization-Based Treatment for Personality Disorders: A Practical Guide (2nd edn), Oxford: Oxford University Press, p. 195. On contrary moves, see also Bateman, A. (2019). ‘Mentalizing, Mentalization-Based Treatment and BPD’. GAP Call-In Series Podcast. Accessed at:

46 Allen, J. G. (2013). Mentalizing in the Development and Treatment of Attachment Trauma, London: Routledge.

47 Fonagy, P. (2010). ‘Attachment Trauma and Psychoanalysis: Where Psychoanalysis Meets Neuroscience’, in M. Leuzinger-Bohleber, J. Canestri, and M. Target (eds). Early Development and its Disturbances: Clinical, Conceptual and Empirical Research on ADHD and other Psychopathologies and its Epistemological Reflections, London: Karnac Books, pp. 53–75: ‘I believe that reconstruction is essential to the therapeutic process because: i. it provides a means to bring the patient’s mind into contact with what it has previously found intolerable; ii. It provides a place where threat to the ego and therapeutic goal are reasonably balanced; iii. It generates a coherent self-narrative assuming a historical continuity of self which may itself be of therapeutic value (Holmes 1998; Shafer 1980; Spence 1994), and iv. Most importantly, it can help in the primary task of the recovery of mentalisation’ (p. 68).

48 Fonagy, P., Bateman, A. W., and Luyten, P. (2012). ‘Introduction and Overview’, in A. W. Bateman and P. Fonagy (eds), Handbook of Mentalizing in Mental Health Practice, Washington, DC: American Psychiatric Publishing, pp. 3–42, p. 39.

49 This clinical recommendation was already put forward by Bion, for whom too central a focus on a past traumatic event as the perceived origin of present difficulties ran the risk of ‘saturating’ the situation, blocking opportunities for further learning by giving the sense of explanatory satisfaction. Bion, W. R. (1970). Attention and Interpretation, London: Karnac Books, p. 29. See also Bion, W. R. ([1973] 1990). Brazilian Lectures, London: Karnac Books: ‘We bring into the open certain elements of an analysand’s past, not because we think they are particularly valuable, but because they are not valuable for him to have in his luggage. If we bring them to the surface then he can forget them. Those memories, past or future, which he does not know seem to have a great deal of power; they are what I would call feeble ideas but powerful emotions’ (p. 30).

50 Bateman, A. W. and Fonagy, P. (2016). Mentalization-Based Treatment for Personality Disorders: A Practical Guide (2nd edn), Oxford: Oxford University Press: ‘MBT has never been focused on the past or been insight oriented, although it is respectful of past experience and how it affects the present. It is our view that explaining a person’s current wish to please in terms of a continuing wish to satisfy a demanding image of a parent is a descriptor masquerading as an explanation. Furthermore, it potentially has the harmful side effect of stimulating pretend mode’ (p. 218).

51 See also Daly, E. (1997). ‘Women from a Broken Home’, Independent, 21 May. Accessed at: ‘Bernice Andrews, an undoubted opponent of the Society, believes there is strong evidence—not least from a survey she conducted—that patients can recover memories. She also believes some such memories are false. So does Professor Peter Fonagy, who has resigned from the advisory board of the British False Memory Society because, he says, “the more recent evidence that has been coming through … has been somewhat inconsistent with the position of the False Memory Society.” ’

52 Fonagy, P. (2016). ‘Ask the Experts: A Conversation with Peter Fonagy’, Trauma Matters, Spring, p. 3. Accessed at: As a clinician, ‘you will inevitably encounter traumatic experiences. What you find there is not necessarily the cause of an individual’s problems, but it is inevitably a crystallization of an individual’s struggles with his or her circumstances, identity, relationships, and life’ (p. 3).

53 Allison, E. and Fonagy, P. (2016). ‘When is truth relevant?’. Psychoanalytic Quarterly, 85(2): 275–303, p. 294.

54 Fonagy, P. (2003). ‘Rejoinder to Harold Blum’. The International Journal of Psychoanalysis, 84(3): 503–509: ‘While reconstruction of how things actually were in childhood can significantly contribute to therapeutic action, it is the process rather than the outcome of this reconstruction that is therapeutic, due to the opportunity thus afforded to rework current experiences in the context of other perspectives.’ (p. 503).

55 Bateman, A. W. and Fonagy, P. (2016). Mentalization-Based Treatment for Personality Disorders: A Practical Guide (2nd edn), Oxford: Oxford University Press, p. 208.

56 Ibid. 218–219.

57 Ibid. 228.

58 Ibid. 284.

59 A study by Perroud and colleagues, including Fonagy, examined self-reported mentalization and attention deficit hyperactivity disorder (ADHD). However, this paper addressed attention difficulties solely as an effect of mentalization, not as a contributing factor or component of mentalizing, as in Fonagy’s earlier work. Perroud, N., Badoud, D., Weibel, S., Nicastro, R., Hasler, R., Küng, A. L., … and Prada, P. (2017). ‘Mentalization in Adults with Attention Deficit Hyperactivity Disorder: Comparison with Controls and Patients with Borderline Personality Disorder’. Psychiatry Research, 256: 334–341. Curiously, the Swiss authors, in a paper without Fonagy as co-author, did later consider attention as a contributor to mentalizing, developing a model of bidirectional influence. Badoud, D., Rüfenacht, E., Debbané, M., and Perroud, N. (2018). ‘Mentalization-Based Treatment for Adults with Attention-Deficit/Hyperactivity Disorder: A Pilot Study’. Research in Psychotherapy: Psychopathology, Process and Outcome, 21(3): 149–154. A concern with attention has been re-integrated with mentalization in the AMAR Model proposed by Lecannelier, F. (2019). ‘A Transcultural Model of Attachment and Its Vicissitudes: Interventions Based on Mentalization in Chile’. In Clinical Handbook of Transcultural Infant Mental Health. New York: Springer, pp. 135–149.

60 E.g. Sperber, D. (2019). ‘Personal Notes on a Shared Trajectory’, in Kate Scott, Billy Clark, and Robyn Carston (eds), Relevance, Pragmatics and Interpretation, Cambridge: Cambridge University Press, pp. 13–20. See also recent blog posts by Sperber, accessed at A conceptualization of the direction of attention as fundamental to ostension and to guiding meaning-making appears already in Sperber’s earliest works. For instance, Speber, D. (1975). Cultural Symbolism, Cambridge: Cambridge University Press: ‘cultural symbolism focusses the attention of the members of a single society in the same directions, determines parallel evocational fields that are structured in the same way, but leaves the individual free to effect an evocation in them as he likes. Cultural symbolism creates a community of interest but not of opinion’ (p. 137).

61 Fonagy, P., Luyten, P., Allison, E., and Campbell, C. (2017). ‘What We have Changed our Minds About: Part 2. Borderline Personality Disorder, Epistemic Trust and the Developmental Significance of Social Communication’. Borderline Personality Disorder and Emotion Dysregulation, 4(1): 9.

62 Fonagy, P. and Clark, D. M. (2015). ‘Update on the Improving Access to Psychological Therapies Programme in England: Commentary on … Children and Young People’s Improving Access to Psychological Therapies’. BJPsych Bulletin, 39(5): 248–251: ‘The remarkable success of internet-delivered therapies strongly challenges the claim that a strong therapeutic alliance is essential’ (p. 249). See also Falconer, C. J., Cutting, P., Davies, E. B., Hollis, C., Stallard, P., and Moran, P. (2017). ‘Adjunctive Avatar Therapy for Mentalization-Based Treatment of Borderline Personality Disorder: A Mixed-Methods Feasibility Study’. Evidence-Based Mental Health, 20(4): 123–127. At best, they regard the concept of ‘therapeutic alliance’ as an umbrella term that points weakly at the kinds of interaction facilitative of epistemic trust. Fonagy, P. (2019). ‘Why is it So Hard to Learn to Do Things Differently? On Not Being Able to Learn from Experience’. GAP Call-In Series Podcast. Accessed at:

63 Fonagy, P., Allison, E., and Campbell, C. (2019). ‘Mentalising, Resilience and Epistemic Trust’, in Anthony Bateman and Peter Fonagy (eds), Handbook of Mentalising in Mental Health Practice (2nd edn), Washington, DC: American Psychiatric Association, pp. 63–77, p. 70.

64 Fonagy, P., Campbell, C. and Allison, E. (2019). ‘Therapeutic models’, in Anthony Bateman and Peter Fonagy (eds), Handbook of Mentalising in Mental Health Practice (2nd edn), Washington, DC: American Psychiatric Association, pp. 169–180, p. 171. The turn to attention to narrative may reflect Liz Allison’s particular interests in narrative and storytelling. E.g. Allison, E. (2017). ‘Observing the Observer: Freud and the Limits of Empiricism’. British Journal of Psychotherapy, 33(1): 93–104.

65 Fonagy, P., Allison, E. and Campbell, C. (2019). ‘Mentalising, Resilience and Epistemic Trust’, in Anthony Bateman and Peter Fonagy (eds), Handbook of Mentalising in Mental Health Practice (2nd edn), Washington, DC: American Psychiatric Association, pp. 63–77, pp. 72 and 74; Fonagy, P. and Allison, E. (2018). ‘The Origin of Human Life: A Psychoanalytic Developmental Perspective’, European Psychoanalytical Federation, 31st Annual Conference, 24 March, Warsaw, ‘It is perhaps helpful here to bear in mind a distinction that Sandler made between two different levels at which the concept of self-representation could be considered. On the one hand, he thought that the self-representation exists as a more or less enduring organization, schema or set of rules that is constructed out of a multitude of impressions and exists outside subjective experience even though it is in large part a consequence of subjective experience. On the other hand, there is the level of the phenomenal or experiential self-representation, which is “the subjective representation or subjective experience of ourselves that we have, at any given moment, in what we can refer to as the experiential realm.” It is this second sense that we wish to capture with the term personal narrative.’

66 Horne, among others, has previously criticized the tendency of Fonagy and colleagues to give insufficient attention to subdominant narratives. Horne, M. (2004). ‘The Universe of Our Concerns: The Human as Person in the Praxis of Analysis’. Journal of Analytical Psychology, 49(1): 33–48. There are some points of similarity between these recent remarks by Fonagy, Allison, and Campbell on acknowledgement of subdominant narratives as an ostensive cue, and earlier work by Feldman, M. (2007). ‘Addressing Parts of the Self’. The International Journal of Psychoanalysis, 88(2): 371–386. Though, unlike Fonagy and colleagues, Feldman also considers how recognition of subdominant narratives may precisely disturb epistemic trust, through being experienced as an intrusion or threat to the dominant self-representation.

67 Pollard, R. (2008). Dialogue and Desire: Mikhail Bakhtin and the Linguistic Turn in Psychotherapy, London: Karnac Books. It may also provide impetus for Fonagy and colleagues to clarify whether they regard ‘thinking’ as necessarily verbal-discursive, and more generally help situate the boundaries of what is meant by ‘thought’ as a ‘mental state’ compared with the building blocks of thoughts.

68 Bouchard, M. A., Target, M., Lecours, S., Fonagy, P., Tremblay, L. M., Schachter, A., and Stein, H. (2008). ‘Mentalization in Adult Attachment Narratives: Reflective Functioning, Mental States, and Affect Elaboration Compared’. Psychoanalytic Psychology, 25(1): 47–66. Another study found no association between reflective function and the severity of personality disorder symptoms. Fischer-kern, M., Schuster, P., Kapusta, N. D., Tmej, A., Buchheim, A., Rentrop, M., … and Fonagy, P. (2010). ‘The Relationship between Personality Organization, Reflective Functioning, and Psychiatric Classification in Borderline Personality Disorder’. Psychoanalytic Psychology, 27(4): 395–409. Researchers have also found a very weak (r = -.11) association between Theory of Mind and BPD symptoms. Belsky, D. W., Caspi, A., Arseneault, L., Bleidorn, W., Fonagy, P., Goodman, M., … and Moffitt, T. E. (2012). ‘Etiological Features of Borderline Personality Related Characteristics in a Birth Cohort of 12-Year-Old Children’. Development and Psychopathology, 24(1): 251–265, Figure 1a. On the interpretation of effect sizes in psychology, see Funder, D. C. and Ozer, D. J. (2019). ‘Evaluating Effect Size in Psychological Research: Sense and Nonsense’. Advances in Methods and Practices in Psychological Science, 2(2): 156–168.

69 Fonagy, P., Luyten, P., Allison, E., and Campbell, C. (2017). ‘What We have Changed our Minds About: Part 2. Borderline Personality Disorder, Epistemic Trust and the Developmental Significance of Social Communication’. Borderline Personality Disorder and Emotion Dysregulation, 4(1): 9. In fact, this claim warrants significant further scrutiny. Is it the accuracy of the model that is contributing to the patient’s sense of feeling recognized? Or the clinician’s confidence in the model? Or the coherence of the narrative the model permits the therapist and/or permits the patient? It does not seem fully plausible that Kleinian psychoanalysis and exposure therapy, though both could claim to have supporting evidence, have the same approach to recognizing the patient as an intentional agent. Or if recognition is achieved by both through some generic process (e.g. therapist belief that the patient has intentions and mental states; use of ostensive cues; adequate professional boundaries), why this recognition would not be achieved by non-evidence-based therapies.

70 Fonagy, P., Luyten, P., and Bateman, A. (2017). ‘Treating Borderline Personality Disorder with Psychotherapy: Where Do We Go From Here?’. JAMA Psychiatry, 74(4): 316–317. ‘The various evidence-based treatments included in the present meta-analysis may have studied different populations and thus may be differentially effective in different subtypes of BPD. For example, mentalization-based therapy was superior over control (structured clinical management) only in patients with BPD with multiple Axis II diagnoses’ (p. 317). See also Volkert, J., Hauschild, S., and Taubner, S. (2019). ‘Mentalization-Based Treatment for Personality Disorders: Efficacy, Effectiveness, and New Developments’. Current Psychiatry Reports, 21(4): 25; Kvarstein, E. H., Pedersen, G., Folmo, E., Urnes, Ø., Johansen, M. S., Hummelen, B., … and Karterud, S. (2019). ‘Mentalization‐Based Treatment or Psychodynamic Treatment Programmes for Patients with Borderline Personality Disorder—The Impact of Clinical Severity’. Psychology and Psychotherapy: Theory, Research and Practice, 92(1): 91–111.

71 De Meulemeester, C., Vansteelandt, K., Luyten, P., and Lowyck, B. (2018). ‘Mentalizing as a Mechanism of Change in the Treatment of Patients with Borderline Personality Disorder: A Parallel Process Growth Modeling Approach’. Personality Disorders: Theory, Research, and Treatment, 9(1): 22. Though not an intervention study, Morosan documented in a community-based cohort study of adolescents that self-reported indiscriminate uncertainty about mental states (RFQu) was associated with more externalizing behaviours at the start of the study; reductions in self-reported certainty about mental states (RFQc) was associated with sharper declines in externalizing behaviours over time. Morosan, L., Ghisletta, P., Badoud, D., Toffel, E., Eliez, S., and Debbané, M. (2019). ‘Longitudinal Relationships between Reflective Functioning, Empathy, and Externalizing Behaviors During Adolescence and Young Adulthood’. Child Psychiatry & Human Development, 51: 59–70 .

72 Lo, C. K. and Wong, S. Y. (2020). ‘The effectiveness of parenting programs in regard to improving parental reflective functioning: A meta-analysis’. Attachment & Human Development, Early View.

73 Rossouw, T. I. and Fonagy, P. (2012). ‘Mentalization-Based Treatment for Self-Harm in Adolescents: A Randomized Controlled Trial’. Journal of the American Academy of Child & Adolescent Psychiatry, 51(12): 1304–1313.

74 Rossouw, T. I. (2012). Self-Harm in Young People: Randomised Control Trial Testing Mentalisation Based Treatment against Treatment As Usual. Unpublished MD thesis, London: University College London. Diagram 2, p. 34.

75 E.g. Fonagy, P. and Bateman, A. (2006). ‘Progress in the Treatment of Borderline Personality Disorder’. British Journal of Psychiatry, 188: 1–3, p. 1.

76 In papers from 1992 and 1994, Fonagy and colleagues acknowledged that improved capacities for mentalizing will allow an individual to ‘conceive of his world in new, sometimes sadder and sometimes happier ways’. Fonagy, P., Moran, G. S., and Target, M. (1992). ‘Aggression and the Psychological Self’. Bulletin of the Anna Freud Centre, 15: 269–284, p. 282. Repeated in Fonagy, P., Steele, M., Steele, H., Higgitt, A., and Target, M. (1994). ‘The Emanuel Miller Memorial Lecture 1992: The Theory and Practice of Resilience’. Journal of Child Psychology and Psychiatry, 35(2): 231–257, p. 251. Likewise, the phrase ‘mind reading may not be an unequivocally positive experience’ first appeared in Fonagy, P., Steele, H., Steele, M. and Holder, J. (1997). ‘Attachment and Theory of Mind: Overlapping Constructs?’. Association for Child Psychology and Psychiatry Occasional Papers, 14: 31–40, p. 37. It then appeared in several subsequent papers. However, in none of these cases was the passing phrase elaborated. In a recent paper, Barbara Castro Batic and Daniel Hayes at UCL/Anna Freud Centre have developed hypotheses for testing regarding potential mechanisms through which harm can occur to patients in talking therapies. However they do not attend to MBT specifically. Castro Batic, B., and Hayes, D. (2020). ‘Exploring harm in psychotherapy: Perspectives of clinicians working with children and young people’. Counselling and Psychotherapy Research, 20(4): 647–656.

77 Fonagy, P. and Target, M. (2005). ‘Some Reflections on the Therapeutic Action of Psychoanalytic Therapy’, in J. Auerbach, K. Levy, and C. E. Shaffer (eds), Relatedness, Self-Definition and Mental Representation: Essays in Honor of Sidney J. Blatt, New York: Taylor & Francis, pp. 191–212, p. 207.

78 A new measure would be able to test this hypothesis: Gagliardini, G., and Colli, A. (2019). ‘Assessing Mentalization: Development and Preliminary Validation of the Modes of Mentalization Scale’. Psychoanalytic Psychology, 36(3): 249–258.

79 Target, M. (2008). ‘Commentary’, in F. N. Busch (ed.), Mentalization: Theoretical Considerations, Research Findings, and Clinical Implications (Psychoanalytic Inquiry Book Series, Volume 29), New York: Analytic Press, pp. 261–279, p. 277. See also Mark, I. L., IJzendoorn, M. H. V., and Bakermans‐Kranenburg, M. J. (2002). ‘Development of Empathy in Girls during the Second Year of Life: Associations with Parenting, Attachment, and Temperament’. Social Development, 11(4): 451–468: ‘Radke-Yarrow et al. (Radke-Yarrow, Zahn-Waxler, Richardson, Susman, and Martinez, 1994) emphasized the importance of recognizing interacting influences of both parent and child. The results of their study with 24–48-month-old children showed how important these interactions may be. The highest frequencies of empathic responses were from children with severely depressed mothers, problems of affect regulation, and a secure attachment relationship with their mother, whereas children of well or less severely depressed mothers with secure attachment relationships and without problems of affect regulation scored neither extremely high nor extremely low. It might be that these middle scores were actually the more optimal scores, and that the very high scores of the children of severely depressed mothers reflected the caregiving behavior that children with disorganized attachments display as toddlers’ (p. 453).

80 Suchman, N. E., DeCoste, C., Leigh, D., and Borelli, J. (2010). ‘Reflective Functioning in Mothers with Drug Use Disorders: Implications for Dyadic Interactions with Infants And Toddlers’. Attachment & Human Development, 12: 567–585.

81 Stacks, A. M., Muzik, M., Wong, K., Beeghly, M., Huth‐Bocks, A., Irwin, J. L., and Rosenblum, K. L. (2014). ‘Maternal Reflective Functioning among Mothers with Childhood Maltreatment Histories: Links to Sensitive Parenting and Infant Attachment Security’. Attachment & Human Development, 16: 515–533.

82 Rizq, R. and Target, M. (2010). ‘If that’s what I need, it could be what someone else needs.’ Exploring the role of attachment and reflective function in counselling psychologists’ accounts of how they use personal therapy in clinical practice: a mixed methods study’. British Journal of Guidance & Counselling, 38(4): 459–481, p. 475.

83 Fonagy, P. (2015). ‘Peter Fonagy on Psychoanalysis and IAPT.’ The History of Emotions Blog, posted on 14 May by Jules Evans. Accessed at: See also Lucassen, N., Tharner, A., Prinzie, P., Verhulst, F. C., Jongerling, J., Bakermans‐Kranenburg, M. J., … and Tiemeier, H. (2018). ‘Paternal History of Depression or Anxiety Disorder and Infant–Father Attachment’. Infant and Child Development, 27(2): e2070: ‘Father’s history of depression or anxiety disorder was not significantly related to infant–father attachment security in the total sample. Interestingly, daughters of fathers with a history of depression or anxiety had higher scores on attachment security than daughters of fathers without this diagnosis’ (p. 1); ‘A potential explanation for the unexpected direction of this effect is that fathers with a history of depression or anxiety might have developed better mentalizing capacities and are therefore better able to “read” their child’, potentially increased by experiences of psychotherapy’ (p. 7).

84 E.g. Bateman, A. and Fonagy, P. (2008). ‘Comorbid Antisocial and Borderline Personality Disorders: Mentalization‐Based Treatment’. Journal of Clinical Psychology, 64(2): 181–194: ‘Mentalizing simply implies a focus on mental states in oneself or in others, particularly in explanations of behaviour’ (p. 182).

85 Fonagy, P., Bateman, A. W., and Luyten, P. (2012). ‘Introduction and Overview’, in A. W. Bateman and P. Fonagy (eds), Handbook of Mentalizing in Mental Health Practice, Washington, DC: American Psychiatric Publishing, pp. 3–42, p. 12.

86 See e.g. Luyten, P., Fonagy, P., Lemma, A., and Target, M. (2012). ‘Depression’, in A. W. Bateman and P. Fonagy (eds), Handbook of Mentalizing in Mental Health Practice, Washington, DC: American Psychiatric Publishing, pp. 385–418: ‘Alongside the major evolutionary advantages, there is a potential shadow side to mentalisation. First, self-awareness and self-consciousness bring with them social emotions such as embarrassment, shame, and guilt … awareness of being unable to achieve one’s aspirations may lead to feelings of depression, loss, pain and fatigue’ (p. 396).

87 Asen, E. and Fonagy, P. (2017). ‘Mentalizing Family Violence. Part 2: Techniques and Interventions’. Family Process, 56(1): 22–44: ‘ “Effective” mentalizing is not the uninterrupted capacity to be reflective and to mentalize explicitly at all times: This would not only be completely unsustainable, but would also kill spontaneity. Instead, the aim of therapy is to establish mentalizing in a balanced way which involves all family members and adapts flexibly and creatively to the context as and when needed’ (p. 27).

88 Cutting, A. L. and Dunn, J. (2002). ‘The Cost of Understanding Other People: Social Cognition Predicts Young Children’s Sensitivity to Criticism’. Journal of Child Psychology & Psychiatry, 43: 849–860. Sutton, J., Smith, P. K., and Swettenham, J. (1999). ‘Social Cognition and Bullying: Social Inadequacy or Skilled Manipulation?’. British Journal of Developmental Psychology, 17(3): 435–450.

89 Fonagy, P. and Target, M. (2006). ‘The Mentalization-Focused Approach to Self Pathology’. Journal of Personality Disorders, 20: 544–576, p. 559. The passage is then repeated in Fonagy, P., Gergely, G., and Target, M. (2007). ‘The Parent–Infant Dyad and the Construction of the Subjective Self’. Journal of Child Psychology and Psychiatry, 48(3–4): 288–328, p. 307.

90 Bateman, A. W. and Fonagy, P. (2016). Mentalization-Based Treatment for Personality Disorders: A Practical Guide (2nd edn), Oxford: Oxford University Press, p. 56.

91 Ibid. 59.

92 Gillespie, S. M., Kongerslev, M. T., Sharp, C., Bo, S., and Abu-Akel, A. M. (2018). ‘Does Affective Theory of Mind Contribute to Proactive Aggression in Boys with Conduct Problems and Psychopathic Tendencies?’. Child Psychiatry & Human Development, 49(6): 906–916, p. 912. cf. Davidsen, A. S. and Fosgerau, C. F. (2015). ‘Grasping the Process of Implicit Mentalization’. Theory & Psychology, 25(4): 434–454.

93 Bateman, A., Fonagy, P. and Campbell, C. (2019). ‘Borderline Personality Disorder’, in Anthony Bateman and Peter Fonagy (eds), Handbook of Mentalising in Mental Health Practice (2nd edn), Washington, DC: American Psychiatric Association, pp. 323–334, p. 337. For a valuable recent contribution to specifying the relationship between RF and empathy, see Borelli, J. L., Stern, J. A., Marvin, M. J., Smiley, P. A., Pettit, C., and Samudio, M. (2020). ‘Reflective Functioning and Empathy among Mothers of School-Aged Children: Charting the Space between’. Emotion, Early View.

94 The Gillespie paper, and other such efforts to address the potential harm of MBT with patients with histories of relational aggression, remain hindered by problems in the conceptualization of affect within accounts of mentalizing. In particular, it remains unclear whether affective mentalizing refers to efforts to mentalize feelings or the use of feelings in mentalizing. Additionally, the definitions of mentalizing considered in Chapter 4 would appear—perhaps inadvertently—to exclude the use of feelings from the definition of mentalizing except insofar as they help individuals deploy, conceive of, or reconsider mental states; the use of emotional resonance to make sense of mental states is, currently, technically outside of the definition of mentalizing. An exception is the encompassing definition from Fonagy, P. (1996). ‘The Significance of the Development of Metacognitive Control over Mental Representations in Parenting and Infant Development’. Journal of Clinical Psychoanalysis, 5(1): 67–86: ‘The psychological processes underpinning the view of oneself and others as motivated by mental states’ (p. 74). However, what this definition gains in inclusiveness, it loses in precision, which is no doubt part of the reason it has been superseded over the decades.

95 Crittenden, P. M. and Landini, A. (2011). Assessing Adult Attachment: A Dynamic-Maturational Approach to Discourse Analysis, New York: W. W. Norton & Company: ‘we are wary of encouraging very disturbed individuals [to] use mentalization for self-protective functions because this may concurrently endanger others’ (p. 39).