Show Summary Details
Page of

(p. 166) Adaptation and mental health 

(p. 166) Adaptation and mental health
Chapter:
(p. 166) Adaptation and mental health
Author(s):

Robbie Duschinsky

and Sarah Foster

DOI:
10.1093/med-psych/9780198871187.003.0008
Page of

PRINTED FROM OXFORD CLINICAL PSYCHOLOGY ONLINE (www.oxfordclinicalpsych.com). © 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

Introduction

Dominant models of mental illness often treat it as a discrete pathology, distinct from ordinary human psychology. One of the more subversive aspects of psychoanalysis was the desire to look at typical and atypical psychological phenomena together, and examine variation between them in terms of common mechanisms.1 This included speculations about neurological mechanisms relevant to both the makings of ordinary life and the formation of symptoms. Dreams, jokes, and the slips of everyday life were included by Freud within the same framework as obsessional symptoms, somatic disorders, and psychosis:

The theoretical importance of this conformity between dreams and symptoms is illuminating. Since dreams are not pathological phenomena, the fact shows that the mental mechanisms which produce the symptoms of illness are equally present in normal mental life, that the same uniform law embraces both the normal and the abnormal and that the findings of research into neurotics or psychotics cannot be without significance for our understanding of the healthy mind.2

There was a price to this strategy for psychoanalytic theory. Many of the concepts used to link typical and atypical phenomena—such as ‘defence’—were forced to become over-absorptive of different meanings, to the point that their meaning became lost. Yet at the same time, Freud intended that this approach would reduce the stigma attached to mental illness and support preventative work with people who were not yet unwell.

Very much the same situation has played out for the concept of ‘mentalizing’. This concept likewise was treated by Fonagy as relevant to typical and atypical mental development, of transdiagnostic significance, and not wholly conscious. Though Fonagy and colleagues would regard their claims about neurology to be less speculative than Freud’s, these claims have likewise addressed mechanisms relevant to both ordinary life and the formation of symptoms. Like Freud, the simultaneous attention to typical and atypical development by Fonagy and colleagues has been pursued in part as a way to combat the ‘unfounded, crass, insuperable prejudice’ associated with mental illness.3 There have been theoretical gains associated with this approach. Insight into successful mentalizing offered ideas for conceptualizing mental ill health; and insight into forms of non-mentalizing offered ideas for conceptualizing the structure of the self, the nature of perceptual experience, and the specific (p. 167) qualities of psychological insight. However, like several of Freud’s concepts, along the way, uses of the idea of mentalization expanded, and expanded.

This prompted the work of Chapters 4 and 5 to attempt to pin down more precisely the definition and constituent elements of mentalization and non-mentalization, and to ask whether they are simple opposites. It also prompted effort in the previous chapter to draw out the complex picture of how Fonagy and colleagues have discussed the role of unconscious processes in everyday life, including in experiences of sexuality and aggression. In this chapter, we turn to consideration of Fonagy and colleagues’ conceptualization of the nature of mental illness, predominantly in papers written for researchers in the field of developmental psychopathology, though with some eye to clinical audiences. We start with criticisms of Fonagy and Target’s work from the 1990s, raised by allies such as Jeremy Holmes and Otto Kernberg who alleged that mentalization risked becoming a disempowering, deficit-focused model of mental ill health. One resource available to Fonagy and colleagues in responding to this concern was attachment theory, which had developed a theory of attachment strategies as evolutionarily primed responses to adverse conditions. We will survey the reflections Fonagy and colleagues have offered on three attachment theorists—Mary Main, Pat Crittenden, and Jay Belsky. The synthesis proposed by Fonagy and colleagues will then be described. The position of Fonagy and colleagues gives prominence to attachment relationships in calibrating an individual’s sense of vigilance or trust in the perspectives of others, as the precondition of learning from them. In recent years, this concern with vigilance, trust, and learning has been central to the theoretical work of Fonagy and colleagues, laying the ground for very important advances in specifying the role of mentalization in both ordinary life and the alleviation of mental ill health.

Adaptation and attachment

Through the 1990s and at the start of the 2000s, Fonagy, Target, and collaborators had a tendency to denigrate non-mentalizing. They described non-mentalizing as an ‘unholy perversion of reflection’,4 and as a ‘stunted or twisted’ version of mentalization.5 The ambition in these passages appears to have been to offer lively description of the way in which forms of non-mentalizing seem to utilize many of the same component elements as mentalization—for instance, that pretend mode utilizes the imaginative capacity to conceive of mental states that is required, too, for mentalization. However, the language is also stigmatizing, perhaps an unintended reflection of the authors’ clinical frustration with recalcitrant cases. In the mid-2000s, important criticisms were raised of the assumptions about mentalization that such language seemed to express. Jeremy Holmes and Otto Kernberg are psychoanalytic psychiatrists, and friends and allies of Fonagy and Target. Both were early advocates for mentalization in the 1990s,6 and have subsequently co-authored work with Fonagy.7 Yet, (p. 168) they criticized Fonagy and Target for tending to characterize non-mentalizing as a ‘deficit’ or ‘impairment’, without considering whether it may have benefits or adaptive aspects.8 This was out of keeping with a growing critique of deficit thinking in mental health in the 2000s: it was held that mental health theory and practice should recognize the strengths of users of services, acknowledge their agency beyond that ‘granted’ by professionals, and differentiate between short-term and long-term advantages and disadvantages.9 For instance, substance use might be reframed as, in certain contexts, a form of self-medication against other difficulties.10

Fonagy and colleagues accepted the criticism of Holmes and Kernberg as partially justified. There was a tendency towards deficit-thinking in their theorizing.11 The criticism was not wholly justified, however. Fonagy and Target had conceptualized symptoms of mental ill health as intelligible responses to adverse situations. One of the characteristic features of the Anna Freudian tradition, within which Fonagy and Target had trained, was a concern with ‘lines of development which lead from the individual’s state of infantile immaturity and dependence to the gradual mastery of his own body and its functions, to adaptation to the object world, reality and the social community’ (pp. 47–53).12 Anna Freud identified this adaptation with the ego.13 She opposed the ego to the id, conceptualized as continually seeking ‘fulfilment of drive derivatives which are unacceptable to the environment, bringing with it the threat of punishment or loss of love, and, instead of serving adaptation, disturbing it.’14 Defence mechanisms were regarded by Anna Freud as compromises between the ego and the id, representing the best available adaptation of the child to his or her environment. This adaptation would be shaped by the extent of supports available in that environment, and the extent of effective dominance of the ego over the id. In this subtle perspective, depending (p. 169) on development and context, exactly the same behaviour could be in the service of the ego and adaptation, or represent defence and the compromise of ego and id:

A five-year-old may ask his mother for an extra cardigan ‘so as not to catch cold’; or a six-year-old may anxiously wash his hands before dinner since ‘there might be germs’. Should that be understood as early compliance with adult rules, i.e. a welcome sign of adaptation, or, rather, as an ominous sign of hypochondriacal tendencies and of pathogenic defence mechanisms used in the struggle against anal messing? Obviously, only an approximate timetable for normal advance towards compliance with hygiene and cleanliness can supply the answer.15

In the 1990s, when Fonagy and colleagues were working on a manual for child psychoanalysis at the Anna Freud Centre, they adopted an aligned perspective, arguing that ‘the child needs his defences and has good reason for erecting them’.16 Speaking on BBC television, Fonagy criticized therapeutic approaches that do not take into account the difficult feelings or the problems in a child’s environment against which defences may have been elaborated: ‘if you are systematically undermining that very fragile, that very vulnerable sense of who that child is you could end up in the situation where the child becomes really very much more depressed and hopeless and helpless. My concern about it is that you are destroying something in the child that is the child’s own. However distorted and however maladaptive it is, it is the child’s own.’17

The concept of ‘adaptation’ would be a piece of the inheritance of Fonagy and his collaborators from the Anna Freudian tradition.18 Yet a problem with the Anna Freud’s concept of ‘adaptation’ was its multiple meanings, a problem raised with Freud late in her life by Joseph (p. 170) Sandler, and which she seemed to accept.19 What may be adaptive in the short run may not be adaptive in the long run, and vice versa. What may be adaptive for a species may not be adaptive for an individual. Furthermore, the term ‘adapt’ could mean simply the extent to which an individual can respond to a challenge, or the extent to which they can thrive in the face of it.20 There can also be a subtly normative aspect, in which adaptation to an environment may mean capitulation to its norms and existing power relations.21 By the time Fonagy was developing his ideas in the 1980s and 1990s, he was in discussion with psychoanalysts such as Sandler who were acknowledging these various meanings of the concept of adaptation, and the threat that the concept might smuggle unacknowledged assumptions about what counts as the good life. But psychoanalytic discourse frequently lost track of the distinctions. And there was no strong theory available for how individual responsiveness and thriving might relate to the evolutionary level of adaptation.

In this context, Fonagy and colleagues took particular interest in developments in attachment theory in conceptualizing adaptation, defences, and mental health and ill health. These developments had been prompted by an influential, but unpublished, conference paper presented by Mary Ainsworth at the International Conference on Infant Studies in April 1984.22 In this paper, Ainsworth expressed concern that the term ‘adaptation’ represented a twig-thicket of different meanings; attention to these different meanings offers clarification on the relationship between mentalizing and ‘adaptation’. Like Sandler, Ainsworth acknowledged that ‘adaptation’ could refer to processes at a species level, in identifying a behavioural system or trait as contributing to survival or reproduction. Secondly, ‘adaptation’ could refer to an individual level, identifying a behaviour or trait as responsive to the available rewards and punishments of the immediate environment.23 Ainsworth also observed a third meaning of the concept: ‘In the developmental mental health sense the focus is on how individual differences in development, and on evaluation of how well or how poorly such development equips the individual to cope with the impact of the environment in which he lives.’24 What distinguished this third meaning of the concept from the second was that an evaluation was entailed. The second meaning was merely an acknowledgement that an individual may (p. 171) ‘adapt’ to their circumstances. Ainsworth’s third meaning was to identify ‘adaptation’ as the capacity to thrive in the long term within those circumstances.

Ainsworth’s argument was terminologically complex. Indeed, the subtlety of Ainsworth’s argument may have contributed to her decision not to attempt to publish the article, despite its influence on her students and collaborators. Her proposal was that individual adaptation (long-term thriving) may result from adaptation (changing oneself in order to respond) to the environment. However, there are forms of adaptation (long-term thriving) where refusal to adapt (change oneself in order to respond) is optimal—for instance, in depleting or punitive environments that can be changed or exited. Some forms of adaptation (thriving) may come at the expense of other forms of adaptation (thriving), as in the familiar case in which the demands of one area of life—family, work—come at the expense of others—diet, exercise, self-care. A further complexity lies in the fact that there are forms of adaptation (responding and/or thriving) that are based very directly on adaptation (species-level natural selection), such as the deployment of conditional strategies as evolutionary-based behavioural repertoires. However, there are forms of adaptation (responding and/or thriving) that are more based on social learning or other processes based more on human plasticity than responses directly grounded in adaptation (species-level natural selection).

Working closely with Mary Ainsworth was her former student Mary Main. Ainsworth and Main had observed that in the Strange Situation procedure a sizeable minority of infants directed attention away from their caregiver on reunion. These were infants whose primary caregiver had more frequently rebuffed their attempts to seek closeness when distressed (see Chapter 3). Main argued that, from an evolutionary perspective, avoidance could be interpreted as a proactive response by the infant that ‘paradoxically permits whatever proximity is possible under conditions of maternal rejection’.25 Evolutionary processes would have selected for avoidance as one part of the infant repertoire for responding to caregivers, because infants who are able to avoid antagonizing their caregivers or making demands that their caregiver will rebuff are more likely to have survived. An infant successfully utilizing an avoidant strategy maintains an indirect but real proximity to their caregiver, as well as the regulatory control to continue to be responsive to the environment.26 Main argued that avoidance represented a behavioural repertoire, selected by evolutionary processes, and available to infants to respond to less sensitive care. Following use of the term in studies of animal behaviour, Main termed avoidance a “conditional strategy”, since it would be drawn on when the primary attachment strategy of proximity-seeking is unavailable to an infant, and it would provide a conditional form of access to proximity.

Main anticipated that there would be two possible conditional strategies. One of these, avoidance, directed attention away from cues that might activate attachment behaviour. The other, appearing as ambivalent/resistant attachment in the Strange Situation, entailed vigilant attention to signs of the caregiver’s potential unavailability. Both conditional strategies could be predicted, on average in human evolutionary history, to have provided sufficient indirect proximity with caregiver to have facilitated survival. In her 1984 paper, reflecting on Main’s theory, Ainsworth stated that it may be that avoidance hinders adaptation (long-term thriving), or could be beneficial even in the long-run if conditions continue to be difficult. (p. 172) Ainsworth offered her conviction that this was essentially an empirical matter: the question of whether ‘avoidant attachment may be adaptive according to ultimate criteria in the mental health sense is clearly a researchable proposition’.27

Main’s stance was that, even if short-term benefits could be identified, avoidance would hinder long-term thriving. Yet a greater threat was represented by the potential for breakdown or disruption of attachment strategies, disorganized attachment, which were argued to be elicited by a child’s experiences of frightened, frightening, or dissociative caregiving. Another student of Ainsworth’s, Patricia Crittenden, held a different position. She argued that there may be ecological niches where conditional strategies are simply superior, especially under conditions—most of human evolutionary history—where significant danger is prevalent. To characterize conditional strategies as a second-best option then would be both overgeneralized and potentially ethnocentric.28 Crittenden agreed with Main that there are two general forms of conditional strategy. She argued that Ainsworth’s avoidant attachment classification represented a local case of the broader strategy of inhibiting or distorting information about negative emotions. Ainsworth’s ambivalent/resistant classification represented a local case of the broader strategy of inhibiting or distorting information about the temporal and causal sequencing of others’ availability, permitting the maintenance of vigilance. Crittenden proposed that with maturation, and responding to experiences of threat or danger, additional forms of these two strategies become available.29 To the extent that symptoms are prompted by close relationships and/or experiences of threat, she suggested that many forms of mental illness may be regarded as effects of conditional strategies. For instance, one kind or component of post-traumatic stress disorder (PTSD) is avoidant symptoms, numbness, and other inhibitions of negative emotions. Another kind or component of PTSD is hyperarousal and vigilance for threats or potential separations from attachment figures.30

In the 1990s, Fonagy was impressed by Main’s account of conditional strategies as behavioural repertoires made available by human evolutionary history. This model aligned with Anna Freud’s suggestion that many forms of mental illness begin as a local response to adverse circumstances, and that they remain relevant and perhaps even helpful so long as those adversities continue. However, Main’s account added to Freud’s an underpinning ethological–evolutionary framework, in which certain profiles of symptoms have an underpinning logic and connection in their likelihood of contributing to survival in the face of adversity. Yet, in recent years, Fonagy has increasingly advocated for Crittenden’s position, which he has described as ‘more inspiring’.31 Crittenden’s account of the conditional (p. 173) strategies aligned with the Fonagy and Luyten characterization of affective and cognitive ‘poles’ of mentalization (see Chapter 4), and with Luyten’s distinction between avoidant and anxious forms of depression.32 Fonagy appreciated Crittenden’s claim that, when adversities continued, a conditional strategy may have distinct advantages for an individual, as well as significant costs. This aligned with Anna Freud’s approach to thinking about defences, and appeared to Fonagy to be a destigmatizing perspective.33 Fonagy was also interested by the idea that even symptoms of PTSD such as hypervigilance may, in certain ways, serve as coping strategies when they become woven into day-to-day living, part of how an individual works out their characteristic response to challenges.34

Another attachment theorist, Jay Belsky, had likewise argued that conditional strategies may have advantages within certain ecological niches. Belsky speculated that humans evolved to treat the early care we receive from attachment figures as a ‘signal’ about the safety or danger of the environment. He suggested that sensitive care prompts a secure attachment relationship, reflecting confidence in others and their availability in the expectation of a more favourable environment. Insensitive care prompts an insecure attachment relationship, reflecting caution about trust in others and their availability in the expectation of a less favourable environment. An individual’s endocrinology, behaviour, and forms of information processing may then be calibrated by a concern with survival and short-term benefits, even if these come with a long-term price. Following developmental adversities, on average, individuals may be anticipated to be more impulsive, anxious, distractible, and more ready to engage in sexual and aggressive behaviours. Seeking to examine this hypothesis, Belsky conducted secondary analyses on the data from large cohort studies. He found that attachment insecurity was associated with earlier date of menarche, and harsh early care associated with more sexual risk-taking behaviours.35 Belsky interpreted these findings as suggesting that the conditional strategies may have conferred benefits under adversity in human evolutionary history. They prompt early puberty and earlier sexual behaviour, increasing the likelihood in evolutionary history that a woman would survive to reproduce, though this strategy is also costly. For instance, early menarche is associated with long-run disadvantages (p. 174) in terms of physical health. Early sexual debut may be linked to less discrimination of sexual partners and less safety in the relationship.

While Crittenden and Belsky frame their claims as criticism of Main’s stance, Fonagy felt that they reflected different levels of analysis: ‘To my mind, in the same way that light can be seen as either waves or particles, the consequences of attachment trauma can be seen as an adaptation that also reflects the absence of an organised strategy. I see no loss of meaning coming from this admittedly heuristic or rather deeper integration of these models.’36 Main’s primary concern was with motivational systems, their modulation or disruption; Belsky’s was with factors that might have helped past humans survive to the age of reproduction and see offspring survive; Crittenden’s was with the potential benefits of the inhibition or distortion of emotion or cognition under conditions of threat. Fonagy agreed with Main that childhood experiences of maltreatment or trauma can result in breakdowns or interruptions of emotion regulation and attentional control. These are important contributors to mentalization, so problems in these areas are likely to have significant ramifications for an individual in their experience of self and future social interactions. However, Fonagy agreed with Belsky that ‘natural selection may have solved the problem of environmental variability by using the attachment system as an “early warning system” to indicate to an infant the degree to which violent conduct may be required later in life’.37 A whole variety of social, genetic, and endocrinal calibrations maybe made as a result of the early warning provided by childhood, priming later responses.38 Finally, Fonagy agreed with Crittenden that disruptions of emotion regulation and attentional control may be patterned in ways that have specific advantages under conditions of adversity.39

Even emotions like fear can contribute to the development of habitual ways of responding that benefit an individual in certain ways.40 PTSD symptoms of hyperarousal offer a good illustration. Fonagy and colleagues held that Main was right that these symptoms can be disruptive for an individual’s capacity to regulate their feelings and maintain attention and responsiveness to their immediate environment. These symptoms may also be psychologically and physiologically very costly. Yet the potential for hyperarousal may have contributed to human survival in our evolutionary history, where dangers may have been frequent and pressing. If the present likewise contains significant potential dangers then, Fonagy agreed with Crittenden that hyperarousal can have localized advantages for an individual in cuing a rapid response to potential threats. If the present does not contain relevant dangers, then (p. 175) indeed, he agreed with Main, hyperarousal is likely overwhelmingly counterproductive. In making appeal to the concept of adaptation, Fonagy has been accused by critics of absorbing the problems of earlier ego psychology in identifying mental health with subjection to hegemonic social norms.41 There is an extent to which this criticism holds, but it does not have full purchase. Fonagy’s use of the concept of adaptation has been qualified by acknowledgement, especially in recent years, regarding the harms of social norms that institutionalize non-mentalizing and prompt adaptations from individuals that hinder their long-term creativity and thriving (Chapter 9).

From the perspective of Fonagy’s approach to theorizing adaptation, even mental health symptoms that reduce environmental responsiveness, like dissociation, may be appraised for the extent to which they developed in response to past adversity, and the extent to which they are reinforced by helping an individual respond to their current challenges.42 Fonagy has argued that both appraisals are important in order to understand what is maintaining mental health problems. Clinicians who intervene without this understanding may risk increasing their client’s suffering and reducing their capacity to cope with their life. Fonagy likewise argued that the forms of non-mentalizing may similarly have evolved as repertoires because they contributed to survival under conditions of adversity. According to this logic, where circumstances would specifically penalize mentalizing, it may risk harm to patients to encourage its development. However, Fonagy and colleagues also argue that, when the environment is safe enough to mentalize adversity and trauma, this can prove grist to the mill. Skills gained in processing these experiences can lead to the specific development of strengths in reconsidering the thoughts and feelings of oneself and others, prompting the phenomenon of post-traumatic growth.43 For instance, Fonagy and colleagues have documented the potential for elevated self-reported empathy in adults who have experienced childhood trauma.44 However, they caution that this elevated empathy is not yet well understood, and could reflect hypermentalization rather than mentalization of others, or both may be making a contribution.

In their 2017 paper, ‘What We have Changed our Minds About’ (see Chapter 4), Fonagy, Luyten, Allison, and Campbell applied these ideas about adaptation in thinking about borderline personality disorder (BPD). They argued that many mental disorders, BPD among them, are ‘nested in the context of the evolutionary priorities of the human condition’.45 (p. 176) Following Belsky, Fonagy and colleagues argued that caregiving experiences represent a form of social communication about the most effective way for a child to function. They speculated that:

In certain situations, for example, an emergency milieu characterized by high levels of interpersonal aggression, the heightened and immediate sensitivity and seemingly instinctive and physically charged form of appraisal characteristic of BPD might in fact be adaptive, at least in the short term. In such an environment, extreme vigilance is a potential advantage, and similarly, the ability to form intense emotional relationships quickly might elicit resources or protection … we postulate that this mentalizing profile may be a response to cues suggestive of an unreliable and potentially threatening social environment. We thus should be wary of seeing apparent dysfunctions of the clinically ‘hard to reach’ as indicative of a deficit or any kind of sub-optimal functioning (as, indeed, we have done previously). We would now consider that what may appear to us as dysfunction is an evolutionarily primed adaptation.46

BPD is, here, regarded in a way that integrates the conclusions of Main, Belsky, and Crittenden. In part, it reflects breakdown of stabilizing and organizing aspects of emotion and attentional regulation, often but by no means always, following from adverse experiences of care in childhood and/or childhood trauma. This then contributes to non-mentalizing social interactions with others, which helps sustain difficulties in modulating emotion and attention, giving the impression of a disorder of ‘personality’. In part, BPD reflects a repertoire pre-primed by human evolution, and elaborated on the basis of experience, in which early adversity and trauma function as signals that hyperarousal and a focus on short-term concerns will facilitate survival. The slow work of generating second-order representations of mental states and reappraising them is impatiently downplayed. And, in part, BPD reflects a profile of symptoms that may be reinforcing for an individual if it helps them cope with present adversities.47

For example, in environments characterized by high levels of violence, it may be expected that security, trust, and the mentalization of oneself and others would result in risk of exploitation and increased danger—for instance, through a failure to access aggression when needed.48 Functioning in teleological mode, by contrast, could be expected to achieve a more immediate response, with greater effectiveness at least in the short term. Bateman and Fonagy have claimed that it may be beneficial to ‘prioritise an external, nonreflective, rapidly responding focus on the control of others’.49 In fact, it should be acknowledged that longitudinal (p. 177) evidence to date has generally shown the opposite: that under conditions of ongoing adversity, secure attachment in childhood is an asset rather than a deficit for later thriving.50 However, it may be that studies have not been designed to identify the specific benefits that might stem from non-mentalizing.51 Or that the studies have not been conducted in contexts with the levels of adversity and violence Fonagy and colleagues think are relevant.

Epistemic trust

Drawing on these reflections about adaptation, Fonagy and colleagues have argued that ‘as clinicians, we may end up calling these individuals “hard to reach”, yet they are simply showing what may be a reasonable adaptation to a social environment where information from most attachment figures is “tagged” as likely to be misleading’. As such, ‘it may be more accurate to characterize BPD as an “emergency mode” form of social understanding’.52 The defining characteristic of this ‘emergency form’ of social understanding is a higher bar for trusting the claims and experiences of others. Trust has been a topic of rapidly rising interest in the social sciences since the 1990s, in a social context in which management of vulnerability and complexity has become a central concern.53 The question of trust in professionals—including psychological professionals—has reflected, in intensified form, this wider concern, because precisely part of what it is to be a professional is to provide a service that depends in part on trust.54 Fonagy and colleagues were also reflecting on empirical findings that showed that early abuse and neglect predicted later symptoms of personality disorder, with mediation by reflective function clearly in operation, but to less of an extent than anticipated.55 Other factors besides mentalization appeared to be important for the relationship between early adversity and personality disorder.

Drawing from the work of Sperber and colleagues,56 Fonagy and colleagues have termed ‘epistemic vigilance’ an attitude in which the claims and experiences of others are not felt to be dependable, generalizable, or relevant.57 They conceptualized epistemic vigilance as the (p. 178) default state of humans, a valuable tool to protect against misinformation stemming from malice or incompetence on the part of others: ‘All young humans are at the mercy of a knowledge differential, uncertain about the trustworthiness of the information they are about to receive … We are more likely to survive and thrive if we do not forget that not everyone will teach us things that are true or indeed in our interest to learn.’58 It should be noted that there are three important differences between Sperber’s original claims and the exposition of them offered by Fonagy and colleagues, which was mediated by Gergely’s interpretation:59

  1. 1) Whereas for Sperber, epistemic vigilance is a ‘suite of cognitive mechanisms’,60 the characterization of epistemic vigilance as an ‘attitude’ by Fonagy and colleagues conveys the impression of a more unitary phenomenon.

  2. 2) Whereas for Sperber, epistemic vigilance seeks to assign the right amount of credence to people and conversations, this meaning remains operative, but Fonagy and colleagues also give the term the sense of distrust. For instance, Fonagy and Allison proposed that disorganized attachment means that a ‘child seeks others to confirm or deny his/her own understanding, which he/she has little faith in, but, being unable to trust information received from others, remains in a state of uncertainty and epistemic vigilance.’61

  3. 3) Sperber insists that epistemic vigilance may characterize individuals, but can also characterize groups, institutions, and cultures. Yet even individual epistemic vigilance occurs on the basis of repertoires that may be suggested or imposed by the individual’s wider community. Whether at an individual or a collective level, epistemic vigilance is then achieved through what Hutchins calls ‘distributed cognition’, a network of processes that together create effects greater than the sum of their parts.62 For instance, Sperber identifies that epistemic vigilance may be sustained by the coordination of ‘cognitive artefacts (e.g. measuring instruments), techniques (e.g. statistical tests of confidence), (p. 179) and procedures (e.g. for cross-examination)’.63 By contrast, Fonagy and colleagues have tended, except in their most recent work (see Chapter 9), to limit epistemic vigilance to individuals. They have also neither acknowledged nor explored the distributed cognition that Sperber and colleagues believe underpins epistemic vigilance, with the exception of some initial remarks on the affordances of mobile phone technology (see Chapter 9).

Comparing the work of Sperber and colleagues with the use of it by Fonagy and colleagues, the differences are quite marked. Most importantly, the concept of epistemic vigilance as used by Fonagy and colleagues slides between two very different meanings, without this distinction being marked. Sometimes, Fonagy and colleagues use the term to mean ‘discrimination’, the opposite of blind trust.64 This occurs especially when epistemic vigilance is distinguished from ‘epistemic hypervigilance’ and ‘epistemic mistrust’. Perhaps more commonly, however, the term ‘epistemic vigilance’ is taken to mean, precisely, hypervigilance and mistrust. For Fonagy and colleagues, ‘epistemic vigilance’ often appears to mean a state in which information from others is not felt to have bearing or resonance. This reduces exposure to others’ potential malice or incompetence, at the price of knowledge that would indeed be accurate and useful. In epistemic vigilance then, the betrayal and disappointment we dread are already within us to an extent, in the way that our dread of being undermined by others ultimately undermines our own capacity to benefit from encounters with others.65

In the clinic, an inability to relax epistemic vigilance can be seen when, ‘no matter how true or accurate the therapist’s interpretations are, the patient will not be able to make use of them because they are not experienced as true’, at least in any genuine and personally relevant sense.66 The patient cannot learn from experience—or modify their thoughts, feelings, or behaviour on the basis of this learning—because they are unable to make use of potential inflows of information. Fonagy and colleagues therefore argued that ‘personality disorder is not a “disorder of personality” but an inaccessibility to cultural communication relevant to the self from the social context’.67

Conversely, epistemic trust has been defined by Fonagy and colleagues as ‘an individual’s willingness to consider new knowledge from another person as trustworthy, generalizable, and relevant to the self’;68 or as ‘openness to acquiring social knowledge that is regarded as personally relevant and of generalizable significance.’69 As such, epistemic trust delineates those whose claims and experiences we (feel we can) rely upon in order to understand the (p. 180) world and to change ourselves to effectively respond to it.70 According to Fonagy and Allison, writing in 2014, ‘epistemic trust is there to ensure that the individual can safely change their position; it triggers the opening of what we can think of as an “epistemic superhighway”—an evolutionarily protected mechanism that signals readiness to acquire knowledge.’71 It is not stated whether, like mentalization, this epistemic trust depends on imaginative capacities, as a kind of suspension of disbelief (see Chapter 5). Presumably it does, unless the knowledge transmitted is wholly procedural rather than subjective.72 As Winnicott argued, and Fonagy and colleagues have echoed in other contexts, cultural knowledge needs to be both subjectively conceived as well as objectively discovered if it is to be alive for a person, and usable.73

In articulating the developmental basis for epistemic trust, Fonagy and colleagues criticized psychoanalytic theories that sought to account for a child’s experience of the minds of others in terms of ‘projective identification’; the term is too overladen with different meanings to convey precisely what is taking place.74 They also criticized those, such as Trevathan, who account for a child’s experience of the minds of others in terms of ‘intersubjectivity’.75 Again an over-encompassing term, intersubjectivity implies that mental states are ‘shared’. Fonagy and colleagues have generally regarded mental states as only ever the properties of individuals, though in recent writing they have brought this assumption into question (see Chapter 9). Rather than appeal to projective identification or ideas of intersubjectivity, Fonagy and colleagues instead drew on the work of Gergely and Csibra to propose that epistemic trust tends to be activated by interactions in which we feel the other to recognize our i) mental states, ii) intentionality, and iii) individuality. The concrete behaviours through which this sense of recognition is conveyed were termed ‘ostensive cues’ by Gergely and Csibra.76 For Fonagy and colleagues, ‘ostensive cues indicate to the infant that the caregiver (p. 181) recognises him or her as an individual, and as a thinking and feeling (i.e. mentalizing) “agent” ’.77 Fonagy sometimes also refers to ostensive cues as offering acknowledgement of the other’s experience of ‘personal agency’, which is given no definition, but presumably signifies the felt state in which mental states, intentionality, and individuality have been acknowledged. Ostensive cues are characterized as diverse and may include joint attention, turn taking, social referencing, and marked mirroring.78 Kindness has been conceptualized as an ostensive cue.79 One form of kindness, gentle touch that acknowledges and responds to embodied cognition, is an especially powerful ostensive cue.80 Fonagy and colleagues argued that humans have evolved to treat receipt of ostensive cues as a basis for epistemic trust, since because they signal the presence of another mind interested in our intentions, thoughts, and/or feelings, and therefore likely reflect a situation of relative safety and social supports.81

In the developmental model proposed, parental reflective function leads to the display of ostensive cues towards a child. In turn, Fonagy and colleagues claimed that this process predisposes a secure attachment relationship, because recognition of the intentions and/or mental state of the child will facilitate secure base and safe haven provision. At times, Fonagy and colleagues appear to imply that parental reflective function leads to a quantitative increase in ostensive cues towards the child. However, this does not appear to be their intended position. Over-use or ill-judged use of ostensive cues such as touch and eye contact would rather be intrusive caregiving, which has been associated with insecure and disorganized attachment rather than secure attachment.82 Instead, it must be assumed that ostensive cues are used in a ‘sensitive’ way: responding accurately to the signals and communications of the child in a way that is experienced by the child as appropriate and prompt. This would imply a reconciliation of the opposition drawn by Fonagy and colleagues between reflective function (p. 182) and sensitivity discussed in Chapter 3, and account for their large shared variance in accounting for child attachment security.

Yet, at the same time as contributing to attachment security, Fonagy and colleagues argue that (appropriate use of) ostensive cues facilitate development of a capacity to lower epistemic vigilance under appropriate circumstances.83 Both secure attachment and the capacity for epistemic trust are, Fonagy and colleagues suggest, resources for social adaptation and mentalization. They suppose that the attachment system evolved to ensure the protection of juveniles by having them seek their familiar caregiver when alarmed. However, the attachment system may, in human evolutionary history, have also gained the additional function of calibrating the capacity for epistemic trust, to equip the individual for the relative safety or danger of the environment in which they will grow up.84 In relative safety, the capacity for epistemic trust is a great social asset. In relative danger, its risks increase.

These reflections raise the question of the exact relationship between attachment and epistemic trust, not least because one way that Ainsworth had characterized individual differences in the Strange Situation was precisely in terms of the infant’s trust in their caregiver.85 In recent years, Ainsworth’s remarks have been developed further by attachment researchers; there has been a growing trend for secure attachment to be re-described as a kind of trust in the caregiver’s availability.86 In 2014, Fonagy and Allison attempted a re-description of all the Strange Situation classifications in these terms. Secure attachment was conceptualized as well-calibrated trust in the other. Ambivalent/resistant attachment was interpreted as excessive credulity to the perspective of the other; avoidant attachment was interpreted as vigilance about the perspective of others. Disorganized attachment was characterized as a disruption in the foundational basis of trust in others, prompting a state of persistent epistemic vigilance.87 Fonagy and Allison claimed support for this stance from a study by Corriveau, Harris, Meins, and colleagues. The researchers gave children stimuli to categorize, with input from their caregiver and a stranger. Children from securely attached dyads relied (p. 183) flexibly on the advice of their caregiver and the stranger, depending on who was giving the most plausible information. Children from avoidantly attached dyads tended to ignore the input from their caregiver even when the advice was relevant and helpful. Children from anxious/ambivalent dyads preferred information from their caregiver when stimuli were 50/50 ambiguous. Children from dyads that received a disorganized attachment classification showed a preference for the mother’s advice when the stimulus was novel.88 The data, however, do not line up neatly with the position of Fonagy and colleagues. The account of disorganized attachment as disruption of trust and the generation of epistemic vigilance finds little support in the study by Corriveau and colleagues. If anything, the findings weakly point in the opposite direction. Furthermore, elsewhere Fonagy and colleagues have associated both ambivalent/resistant (or preoccupied) and disorganized (or unresolved) attachment with epistemic vigilance and BPD. This is poorly aligned with the finding by Corriveau and colleagues that ambivalent/resistant attachment was associated with apparent trust in the mother when stimuli could otherwise just as easily be classified one way as the other.

Another relatively unconvincing model of the relationship between attachment and epistemic trust was presented by Fonagy and colleagues in their ‘What We have Changed our Minds About’ paper (see Figure 7.1).

Figure 7.1 The Natural Pedagogy Model of Personality Disorder. Illustrates the interaction between social dysfunction, failure in social communication, epistemic mistrust, and imperviousness to social influence that underpin personality disorder. Emotion dysregulation, disrupted attachment histories and the disorganized insecure attachment system generate social/interpersonal dysfunction. This undermines accurate social communication, causing social disruption, the misinterpretation of social signals, and difficulty in recognizing ostensive cues from others. These difficulties in the area of social communication can give rise to epistemic mistrust in relation to the social environment. This is not inherently a maladaptive process: epistemic vigilance has a natural function. However, the absence of epistemic trust sets a limit upon social learning. This can render the individual potentially unable to function effectively within their social environment and can lead to further disruption in the social network, leaving the individual increasingly isolated and prone to further social/interpersonal dysfunction

Figure 7.1 The Natural Pedagogy Model of Personality Disorder. Illustrates the interaction between social dysfunction, failure in social communication, epistemic mistrust, and imperviousness to social influence that underpin personality disorder. Emotion dysregulation, disrupted attachment histories and the disorganized insecure attachment system generate social/interpersonal dysfunction. This undermines accurate social communication, causing social disruption, the misinterpretation of social signals, and difficulty in recognizing ostensive cues from others. These difficulties in the area of social communication can give rise to epistemic mistrust in relation to the social environment. This is not inherently a maladaptive process: epistemic vigilance has a natural function. However, the absence of epistemic trust sets a limit upon social learning. This can render the individual potentially unable to function effectively within their social environment and can lead to further disruption in the social network, leaving the individual increasingly isolated and prone to further social/interpersonal dysfunction

Source: Reproduced from Peter Fonagy, Patrick Luyten, Elizabeth Allison, and Chloe Campbell, ‘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(9): Figure 1, DOI: https://doi.org/10.1186/s40479-017-0062-8 Copyright © The Authors. Licensed under a Creative Commons Attribution 4.0 International License.

In this model, only the history of adversities feeds mentalizing problems, with difficulties in emotion regulation and attachment contributing only through this adversity. This does not seem plausible: the concepts of ‘adversity’ and ‘dysfunction’ are losing purchase, as they become overinvested. In the model presented in ‘What We have Changed our Minds About’, mentalizing problems are described as causing persistent epistemic vigilance through two routes. A first is through the failure to pick up on and use others’ ostensive cues as a prompt for epistemic trust in them. A second is through general loss of interest in social communication. In this account, there is no contribution of epistemic mistrust to difficulties in mentalizing; it is assumed that causality is only in the other direction.

A more recent, and more persuasive, characterization of the relationship between epistemic trust and attachment was presented by Fonagy and colleagues at the Epistemic Petrification Conference at Bristol University in 2018. In a keynote address, Fonagy and colleagues updated the 1998 and 2008 models (see Chapter 3) of the role of caregiving in the development of mentalization (see Figure 7.2).

Figure 7.2 The role of caregiving in the development of mentalization.

Figure 7.2 The role of caregiving in the development of mentalization.

Source: Reproduced from Fonagy, P. Allison, Luyten, P. and Campbell, C. (2018). ‘Epistemic Petrification’. Keynote address at the Epistemic Petrification Conference, Bristol University, 2–3 July.

First, Fonagy and colleagues described a pathway in which the capacity for epistemic trust was supported. Sensitive caregiving was defined by Ainsworth as the caregiver’s ability to ‘perceive and to interpret accurately the signals and communications implicit in her infant’s behavior, and given this understanding, to respond to them appropriately and promptly’.89 Fonagy and colleagues argued that sensitive care is in large part underpinned by parental reflective function, which allows the caregiver to give ostensive cues to the child through behaviours that perceive and respond to signals and communications suggesting the child’s mental states. This facilitates the child’s development of secure attachment and epistemic trust. In the 2018 keynote, Fonagy and colleagues also permitted a reinforcing pathway from secure attachment to epistemic trust. Use of the caregiver as a secure base may facilitate exploration of things and people. But to find learning in the social aspects of exploration, (p. 184) and for this to be experienced as having genuine bearing and importance, epistemic trust is needed. As such, in the updated 2018 model, learning about the (social) world and the self are routed through epistemic trust.

The relationship between learning about the self, learning about the world, and mentalizing in the 2018 model is curious. Fonagy and colleagues retain from Fonagy’s 1998 model the Hegelian position that learning about the self is mediated through learning about the world. But the 2018 model describes only partial mediation. This implies that an individual can engage in learning about the self through epistemic trust without learning about the world. This once more raises the unresolved question in Fonagy’s thinking about whether envisaging and conceptualizing the i) mental states and ii) behaviour of the self and the iii) mental states and iv) behaviour of the other can feasibly be characterized under the same umbrella, and whether one or another element has priority. In Fonagy and Luyten’s 2009 paper, the Hegelian account was still dominant, with experience of the mental states of (p. 185) others given developmental priority, and experience of the mental states of oneself occurring on the basis provided by the former. In the 2018 model, this mediation clearly remains important. But there seems to be a direct route to learning about the self as well. Furthermore, whereas the 1998 diagram suggested that a mentalized account of the caregiver’s behaviour contributes to the development of mentalizing capacity, in the 2018 diagram there is no line between learning about the world and mentalizing. The implication would be that learning about the self has some priority, then, for the initial development of mentalizing in early childhood.

However, as we saw in Chapter 6, the term ‘self’ is a complex one in the vocabulary of Fonagy and his collaborators. It may be suspected that the meaning of ‘self’ in the phrase ‘learning about the self’ is the individual’s experience as an embodied subject, since other aspects of ‘self’, such as the sense of personal agency, have clearly been theorized elsewhere as depending on mentalization (see Chapter 6). This is supported by the fact that the production of a self-representation that weaves together our experiences, wishes and plans, self-knowledge, and other facets of the ‘self’ is distinguished from ‘learning about the self’ and placed in a bidirectional relationship with mentalizing. Hegel returns here: learning about and from others is placed in a bidirectional relationship with mentalizing after early childhood, mediated by an individual’s narratively constructed sense of self and self-relevance. It is this learning that supports successful navigation of the social world, and mentalizing is important to the extent that it facilitates such learning, not in itself.

In the 2018 keynote at the Epistemic Petrification Conference, Fonagy and colleagues also presented a further diagram (Figure 7.3), representing a developmental model from problems with early care to persistent use of non-mentalizing modes.

Figure 7.3 Developmental model from problems with early care to persistent use of non-mentalizing modes.

Figure 7.3 Developmental model from problems with early care to persistent use of non-mentalizing modes.

Source: Reproduced from Fonagy, P. Allison, Luyten, P. and Campbell, C. (2018). ‘Epistemic Petrification’. Keynote address at the Epistemic Petrification Conference, Bristol University, 2–3 July.

Here, neglect or traumatic attachment relationships contribute to the maintenance of epistemic vigilance and to disorganized attachment.90 In the 2018 keynote, Fonagy and colleagues stated that they were not sure whether disorganized attachment contributed to the maintenance (p. 186) of epistemic vigilance. This was in contrast to an earlier version of the same diagram, presented in various places in 2017 including London and San Francisco, in which the pathway from disorganized attachment to epistemic trust was absent.91 Difficulties sustaining epistemic trust contribute to difficulties in social learning. Unlike in the model of the development of mentalizing, in which learning about the self and the world were distinguished as playing somewhat different roles, in the development of non-mentalizing their role is characterized as much the same. In turn, this can contribute to excessive credulity at times, because inaccuracies in the perceptions of others’ behaviour and one’s own mind can lead the misleading claims of others to be mistaken for accurate ostensive cues about one’s mental states.92 There are also bidirectional links placed between epistemic hypervigilance and problems with mentalizing: an important difference from the model in the ‘What We have Changed our Minds About’ paper, in which there was no path at all from epistemic vigilance to difficulties in mentalizing.93

(p. 187) The 2018 keynote by Fonagy and colleagues also proposed bidirectional relationships between credulity and problems with mentalizing. Non-mentalizing prompts feelings of isolation and loneliness, because the individual’s mind is not able to take nourishment from their interactions with others. This stimulates ‘epistemic hunger’, in which the strength of the desire to feel recognized by others may reinforce credulity when someone seems to offer the possibility of recognition.94 Elsewhere during the Epistemic Petrification Conference, Fonagy clarified how this credulity would occur:

If I do not have any clear view of myself then when anyone says something about me, I may experience this as a match, and excessive trust can develop … So somebody says ‘you’re a bad person’ and I have some image of myself as evil. I may trust that statement; that’s very common for the patients that I treat. An analogy comes in with extremist politics here: you have a popular set of ideologies that are clearly defensive. They say ‘I’m being maltreated. The world is unjust to me. I have been excluded.’ Then you have someone who wants to acquire their trust, who says ‘yes, you have been excluded. You have been maltreated. You should be rebelling and I’m surprised that you’ve put up with this for so long.’ A populist ideology, and they come to be trusted. That is a sinister pattern seen in politics today in many places.95

In the first case, Fonagy describes the experience of match between social denigration and feelings of self-denigration. This is a situation he regards as especially predisposed today by the circulation of idealized images of lifestyle and success in social media, which can offer a picture of the self as bad in comparison. On the one hand, difficulties learning about the self and forms of non-mentalizing may facilitate this experience of match. As may adolescence, as a developmental stage in which cognitive and social supports for achieving clarity and specificity in self-representation are not fully online. On the other hand, the experience of social denigration matching feelings of self-denigration can be anticipated to contribute to further difficulties in mentalizing about oneself—for instance, by reinforcing both distress and deployment of psychic equivalence in the form of certainty about the self’s worthlessness. In (p. 188) the second case, Fonagy describes the experience of political solicitation. This is, in Fonagy’s view, an extreme and nasty use of a skill characteristic of all effective influencers: the ‘massive difference in ability of individuals to influence (teachers, politicians, managers)’ may be regarded as ‘explicable in terms of varying capacity to generate epistemic trust’.96 In the political case under discussion, the politician’s solicitation creates the illusion of recognizing the listener as an intentional subject and their feelings of having been wronged. This may be facilitated by forms of non-mentalizing, such as psychic equivalence, in which feelings are conflated with political reality. In turn, the match can be anticipated to contribute to forms of non-mentalizing. This might include difficulties in experiencing apparent political opponents as having minds, or even being fully human like oneself; a requirement for urgent action in teleological mode; and/or aspects of pretend mode in which political ideology is treated as reality, without attention to its actual limitations in characterizing states of affairs.

In Fonagy’s earlier work, both affect regulation and attentional control had been described as requisites and/or component elements and/or effects of mentalization. He had also elsewhere theorized bidirectional relationships between mentalization and social support—for instance, in helping an individual reframe and digest difficult experiences.97 The 2018 diagram did not include affect regulation, attentional control, or social support. But, in presenting his keynote, Fonagy verbally situated that they contributed to mentalizing difficulties. The focus in the keynote was on another relevant component, which was difficulties learning about the self and the world. When these difficulties prompt non-mentalizing, this helps sustain a state of epistemic mistrust, blocking the capacity to experience the claims and experiences of others as trustworthy and relevant. In turn, problems in adapting to the social world are likely to ensue. This will depend in part on the extent to which epistemic mistrust becomes context-insensitive.98

An interesting point for the model, raised by Fonagy and colleagues elsewhere, is that it is not just abuse and neglect that can lead to epistemic distrust, though they regard it as an especially important route. Social and political forms of adversity may predispose the development of epistemic distrust, as may various genetic and temperamental factors.99 Fonagy and colleagues have also argued that epistemic distrust is not simply characteristic of pathology, but is a feature of all our lives at times. As we saw in Chapter 6, Fonagy and colleagues have argued that sexuality and aggression are intentions that caregivers will, to some extent, reject or ignore rather than acknowledge. Fonagy and Allison observed that ‘this leaves us all with a gap, a failed definition, a rupture of epistemic trust’.100 The cycle of epistemic hypervigilance and epistemic credulity can therefore be expected to occur with aspects of (p. 189) the alien self, and perhaps with particular frequency in relation to sexual desire and aggressive intentions. Given that these are among the major challenges of adolescence, it is perhaps then unsurprising that ‘an average adolescent is in a state of often quite acute epistemic mistrust’,101 though punctuated by epistemic credulity with peer groups and sometimes more generally.102 To the extent that epistemic hypervigilance and credulity come to dominate an individual’s experiences of sexuality or aggression, it can be anticipated that they will stop being able to take in and process information about themselves and the world.

Marinus van IJzendoorn and Marian Bakermans-Kranenburg, eminent attachment researchers, also presented a paper at the Epistemic Petrification Conference. They raised a variety of critical points following on from the 2018 keynote paper by Fonagy and colleagues.103 A first was that, naturally for emergent scientific work, the model remains highly speculative. There is, as yet, no validated measure of epistemic trust.104 Though there are many measures of trust and distrust, Fonagy and colleagues would need to decide whether they are articulating something more specific, and how it relates to existing constructs. Recently Bo, Bateman, and Kongerslev used the ‘trust in peers’ sub-scale of the self-report Inventory of Parent and Peer Attachment as a measure of epistemic trust. Yet, with items like ‘I trust my friends’, it is not clear that this addresses epistemic trust specifically.105 No measure has been developed to assess epistemic vigilance. No measure has been developed to assess epistemic credulity. Some of the elements of the model are also quite vague, and will be difficult to operationalize without further articulation—for instance, the idea of ‘difficulties learning about self and the world’. Specific predictions about the relationship between caregiving and epistemic trust will need to be offered, likewise the relationship between epistemic trust and mentalization.

A second issue raised by van IJzendoorn and Bakermans-Kranenburg was that Fonagy and colleagues may be underestimating the multiple causal implications of attachment (p. 190) relationships. Fonagy and colleagues were not sure whether or not disorganized attachment would contribute to epistemic dysfunction. However, this may be because they have assumed that disorganized attachment represents a single process (see Chapter 3). One process under the umbrella of disorganized attachment is dissociation and disorientation. Van IJzendoorn and Bakermans-Kranenburg asked what role epistemic hypervigilance and credulity might play in dissociative symptoms, and vice versa. As we saw in Chapter 4, in a 2017 paper Ensink, Fonagy, and colleagues criticized earlier work on mentalization for missing out the mediating role of dissociation. Van IJzendoorn and Bakermans-Kranenburg offered aligned concerns, but added that it would be valuable to assess dissociation and disorientation in developmental perspective and in light of the attachment relationship.

Van IJzendoorn and Bakermans-Kranenburg also wondered about the integrity of the construct of epistemic trust. They worried that the term ‘trust’ is ambiguous.106 For instance, they observed that Fonagy and colleagues had discussed trust in the claims of others, and trust in their experience and viewpoint.107 There could also be other objects of trust, such as trust in the other’s availability in an attachment relationship. Trust in others’ claims, trust in the usefulness or applicability of their experience and viewpoint, trust in their integrity, and trust in the other’s beneficence towards us may all vary relatively independently from one another.108 Indeed, the ‘Epistemic Trustworthiness Inventory’ developed by Hendirks and colleagues distinguishes i) expertise/competence, ii) integrity, and iii) benevolence as separate scales.109

To give an example of such differences in forms of trust: for an individual operating in teleological mode, a claim substantiated by hard evidence is persuasive; the other person’s experience and viewpoint are not (see Chapter 5). To offer another example: a child’s trust in their caregiver’s availability may well be somewhat helped by trust in the adult’s claims—for instance, about how best to make friends at school. But even if these claims turn out to have been mistaken, this need not have bearing for whether the child will feel that they can return to the caregiver as a safe haven when alarmed or distressed. For Van IJzendoorn and Bakermans-Kranenburg, it is the availability of the caregiver as a safe haven, rather than (p. 191) their provision of trustworthy communication about the social world, which can be anticipated to have the more significant role in facilitating affect regulation, attentional control, and future social competence, though trustworthy communication may no doubt contribute to the experience of a caregiver as a safe haven.

Similarly, mentalization may not always contribute to epistemic trust. Fonagy and colleagues specifically argued in 2007 that part of the reason mentalization evolved was to facilitate our capacity for deception: ‘Mentalisation may serve competition: as the antlers of a reindeer are there to fight other reindeer rather than to catch prey or ward off predators. Self-awareness enables us to modify the way we wish to present ourselves, and to mislead.’110 However, this statement was from a period when they assumed mentalization was one thing, rather than a cluster of capacities. It may be that particular forms of mentalization are less facilitative of social deception (for instance, reconsideration of one’s feelings in order to account for behaviours discrepant with our self-representation). The overarching concern for the 2018 model is that the relationship with experiences of care and with mentalization may be different, depending on the kind of trust and the kind of mentalization.

In the 2018 model of non-mentalizing, Fonagy and colleagues situated mentalizing difficulties as causing persistent epistemic mistrust, which in turn causes problems learning from others. But this claim could mean any number of things depending on what is meant by ‘mentalizing difficulties’, ‘epistemic trust’, and ‘learning’. Consider, for instance, a situation in which mentalizing difficulties are taken to imply problems in conceiving of the mental states of others, epistemic mistrust is taken to imply problems in trusting the experiences and viewpoints of others, and learning from others is taken to mean gaining benefits from the thoughts of others. In this situation, we have done no more than state the logical implications of a problem in conceiving of the mental states of others. Consider, by contrast, a situation in which mentalizing difficulties are taken to mean problems in reconsidering one’s own feelings as implicated in our behaviour, epistemic mistrust is taken to imply difficulty trusting the claims of others, and learning from others is taken to mean non-verbal imitation of valuable qualities in others. That mentalizing (in this sense) would cause epistemic mistrust (in this sense) would be a surprising, interesting, and falsifiable claim.

A further question left open by the 2018 keynote by Fonagy and colleagues is the relationship between ‘absence of epistemic trust’ and ‘persistent epistemic mistrust’. Part of the problem, as has been widely acknowledged in the literature, is that the term ‘trust’ has various different meanings. Indeed, Lyon and colleagues have observed that ‘trust is one of the most fascinating and fundamental social phenomena yet at the same time one of the most elusive and challenging concepts one could study.111 In Fonagy’s 2018 diagram, ‘absence of epistemic trust’ is prompted by neglect or attachment trauma (and perhaps insecure/disorganized attachment), whereas ‘persistent epistemic mistrust’ is prompted by mentalizing difficulties and leads to problems learning from others.112 They are not simply different names for the opposite of epistemic trust.

(p. 192) The 2018 diagram suggests that the relationship between absence of epistemic trust and persistent epistemic mistrust is mediated by mentalizing difficulties. Likewise, absence of epistemic trust, epistemic hypervigilance, and excessive credulity are all linked to problems learning from others only via the same route. There is no direct path from epistemic hypervigilance and/or excessive credulity to problems learning from others except via mentalization. This suggests differences between absence of epistemic trust and persistent epistemic mistrust in terms of their expected correlates. Though Fonagy and colleagues discuss epistemic trust as if it were a single construct, they seem to be describing two distinct processes—or at least two distinct facets of an umbrella process (see Table 7.1).

Table 7.1 Distinction between absence of epistemic trust and persistent epistemic mistrust

Psychological process

Absence of epistemic trust

Persistent epistemic mistrust

Apparent meaning

Inability to relax ordinary concern to assign the right amount of credence to people and conversations, blocking fast learning

The experiences and claims of others are categorically regarded as not to be trusted, generalized or regarded as relevant

Underpinned by

Epistemic vigilance (in Sperber and colleagues’ original sense, and one of the senses used by Fonagy and colleagues)

Epistemic vigilance (in Fonagy and colleagues’ distinct sense)

Key anticipated correlates

Causal contribution to i) non-mentalizing and to ii) difficulties in mentalizing

Caused by i) non-mentalizing, and ii) difficulties in mentalizing

Disciplinary context

Linguistics (specifically pragmatics), plugged into developmental psychology

Cognitive science (specifically natural pedagogy), plugged into personality and clinical psychology

What Fonagy and colleagues term ‘absence of epistemic trust’ is described as contributing to mentalizing difficulties, whereas ‘persistent epistemic mistrust’ is fed by mentalizing difficulties. ‘Absence of epistemic trust’ is treated as causing difficulties learning about the self and the world; ‘persistent epistemic mistrust’ causes difficulties learning from others. ‘Absence of epistemic trust’ relates to epistemic vigilance in Sperber’s original sense: the concern to assign the right amount of credence to people and conversations is unable to be relaxed when needed. This leads to problems with mentalizing, by contributing to epistemic hypervigilance. ‘Persistent epistemic mistrust’ relates to epistemic vigilance in Fonagy’s revised sense: the experiences and claims of others are categorically distrusted. In both cases, the well-modulated and constructive doubt required for reconsideration of thoughts and feelings is hindered. For this reason, it is rather surprising that persistent epistemic mistrust is only ever an effect of problems with mentalizing, rather than also making a reciprocal contribution.

One explanation may be that, whereas the account of ‘absence of epistemic trust’ by Fonagy and colleagues seems to be primarily in dialogue with developmental psychology, his account of the ‘persistent epistemic mistrust’ is primarily in dialogue with personality and clinical psychology. The researchers’ characterization of ‘persistent epistemic mistrust’ suggests a state in which the thoughts (and perhaps feelings) of others as implicated in their motivations and intentions are not felt to be trustworthy, generalizable, and relevant to the (p. 193) self. Fonagy and colleagues situate mentalizing difficulties as mediating the effect of absence of epistemic trust on the capacity to learn from others. This may be because, in itself, extreme doubt or the absence of doubt about knowledge beyond perceptual experience do not represent a total block on learning from others, where this learning can figure within perceptual experience. What does present such a block is when others are not experienced as having thoughts (and feelings) distinct from our own perceptual experience, or when these are fabricated in pretend mode. Then there is nothing external from which to learn, even if others present claims about the credibility and relevance of their thoughts for us. Perceptual experience is experienced as immediate and immutable (psychic equivalence, teleological mode) or irrelevant (pretend mode), rather than as a contingent state that may be refined or modulated by the capacity to conceive or reconsider the thoughts and/or feelings of others. In general, mistrust is the expectation that others lack the competence or willingness to be reliable in acknowledging and meeting what we need or wish from them.113 Difficulties in conceiving or, especially, reconsidering the thoughts (and/or feelings) of others can be regarded as causing a specific form of mistrust, one in which there is little that can be learnt from the mental states of these others, since precisely these mental states are foreclosed or prejudged by non-mentalizing.

Fonagy and colleagues describe the effects of difficulties mentalizing others on the capacity to learn from others. Factors that have a predictable impact on mentalizing others, such as in-group/out-group dynamics, can be expected to likewise shape whether we are open to learning from these others. Missing, however, is attention to the effects of difficulties mentalizing the self on epistemic trust.114 Likewise, Fonagy and colleagues have as yet not been clear about whether epistemic trust/mistrust of feelings operates in the same way as epistemic trust/mistrust in thoughts. Asen, Campbell, and Fonagy have recently claimed that ‘critical thinking—some form of epistemic vigilance—is a highly valuable social cognitive resource’.115 This implies that the capacity to reconsider thoughts implicated in motivations and intentions of oneself can be facilitated by epistemic vigilance (this may link to the ‘unbinding’ function of meaninglessness discussed in Chapter 6). It is not clear whether there are forms of mentalization that are helped or hindered by epistemic distrust of others. However, Jurist has offered persuasive arguments that, with non-mentalizing and dehumanizing modes institutionalized and ascendant, epistemic distrust in contemporary culture and political discourse might facilitate mentalizing because it will facilitate reconsideration of the thoughts and feelings facilitated by this culture and discourse.116

Another outstanding question is what happens to mentalization when the self is subjected to persistent, as opposed to discriminated, epistemic mistrust. Luyten, Lemma, and Target have implied that this state characterizes depression, but this possibility has been raised only (p. 194) in passing.117 Another matter of remaining ambiguity is whether a capacity to effectively conceive of thoughts facilitates epistemic trust, epistemic vigilance, or either as needed in a discriminated way.118 Even with these questions still outstanding, the value ascribed by Asen, Campbell, and Fonagy to epistemic vigilance about thoughts as an asset to mentalization clearly disturbs any simple assumption that these are two ‘good things’ which must therefore come together. As we saw in Chapter 4, Bateman and Fonagy have described good mentalization of others in terms that seem to privilege considering their different thoughts and perspectives, whereas the qualities that characterize good mentalizing of oneself seem to privilege feelings and awareness of change in perspectives over time. This suggests that epistemic vigilance about the self’s present feelings would also be an asset to this form of mentalizing, in facilitating the reconsideration of inner sensations, affects, and moods, and the potential for feelings to be different across time. This vigilance may be especially what the Steeles and Fonagy were referring to in their initial concept of the ‘internal observer’, which predated the more general concept of ‘reflective function’ (see Chapter 3).

As discussed in Chapter 3, Fonagy and colleagues have headlined vaguely or variously defined terms like ‘mentalization’, which can help coalition-building among groups who can make various investments in the same buzzword, and thereby see themselves as holding the same stance and allegiance. The language of epistemic trust certainly appeals for this function: ‘trust’ is an undeniably potent term in contemporary society, and especially when addressed to clinical audiences for whom various kinds of trust will be salient and important.119 Trust is a language outlined in light, conveying shape more than contour and contrast. As such, the underspecification of epistemic trust may come at the price of conceptual confusion, especially regarding causal relationships.

To avoid this threat, firmer conceptual clarity is also needed on the distinction between epistemic vigilance, absence of epistemic trust, epistemic hypervigilance, and epistemic mistrust. Fonagy and colleagues have repeatedly emphasized that their thinking about epistemic trust is work in progress. In future work, it may be helpful to follow the usage by Fonagy and colleagues that maximally differentiates the concepts:

  • Epistemic vigilance’ would be returned to the usage of Sperber and colleagues, to mean a suite of cognitive mechanisms that help calibrate the credence assigned to the inferred experiences and claims of others. It might be added that such calibration is also relevant to our own inferred experiences and claims, given that these are rarely transparent to us in any simple way. Application of epistemic vigilance to our own thoughts and feelings entails generative doubt, and the potential for their reconsideration.

  • When experience has led this suite of mechanisms to be applied too rigidly and conservatively, the result may be identified as ‘absence of epistemic trust’. This is no longer a suite of mechanisms but a quality: the inflexibility of those mechanisms. This quality may be more or less enduring. The corresponding attitude can be termed (p. 195) ‘epistemic hypervigilance’. This attitude has a bidirectional causal relationship with non-mentalizing.

  • Epistemic mistrust’ can be distinguished as a chronic state in which the thoughts (and perhaps feelings) of others as implicated in their motivations and intentions are not felt to be trustworthy, generalizable, and relevant to the self. This may appear as a fixed aspect of an individual’s ‘personality’, but in fact should be recognized as caused by non-mentalizing and the absence of mentalizing, which stabilize the state.

There is also further conceptual work to be done in specifying the relationship between forms of epistemic trust and forms of mentalizing, including in resolving problems that have been obscured by over-general uses of the two concepts.120 Bateman, Fonagy, and Campbell have specified that teleological mode may prompt epistemic hypervigilance in a focus on moment-to-moment behaviour in lieu of trust in others’ stable intentions (see Chapter 5).121 And in earlier work Fonagy and Higgitt have described how a vicious cycle can arise, whereby psychic equivalence creates states of high arousal, which lead to reliance on pre-reflexive assumptions, which hinders the use of mentalizing, which sustains psychic equivalence.122 However, besides this, the relationship between epistemic hypervigilance and forms of non-mentalizing remains to be spelt out. Readers have tended to assume that Fonagy and colleagues believe that epistemic trust is an effect, not a cause, of mentalizing.123 But the relationship is actually bidirectional and more complicated. The urgency of specification of this theory lies partly in the fact that over the coming years Fonagy and colleagues will be wishing to develop a measure of epistemic trust, to empirically test anticipated relationships. If this measure incorporates the present obscurity in the notion of epistemic (p. 196) trust—for instance, whether it can apply to the self, whether it can apply to feelings—this will hinder subsequent empirical work.

However, perhaps the greater urgency for theoretical specification lies in the fact that, in the past few years, Fonagy and colleagues have reoriented the goal of mentalization-based therapy. They have drawn on Bion’s emphasis on curiosity and ‘learning from experience’ as essential for psychological health.124 On this basis, mentalization is now considered only a means to an end: ‘Mentalising in itself is only an intermediate step, not the ultimate therapeutic objective’ because ‘true and lasting improvement, we believe, rests on … learning from experience beyond therapy.’125 In general, ‘mental disorder can be conceptualised as impairments in the capacity of the individual for social learning expressed in terms of epistemic trust’.126 The question of the meaning and boundaries of epistemic trust and/or mistrust is therefore of the utmost importance. The way in which this question is answered will functionally delineate the goals of therapeutic work and the model of how to achieve them. It is also of great potential significance for conceptualizing the nature of mental illness, and how patients’ symptoms should be grouped and assessed. A concern with epistemic trust cross-cuts existing diagnosis-focused approaches to mental health, suggesting potential revisions to this approach.

Diagnoses and the p-factor

To a large extent, the delivery, resourcing, and administration of clinical services for mental health—as well as the curriculum and structure of clinical training—have been oriented by discrete diagnoses. Diagnoses are functionally attractive for bureaucratic systems, because they permit judgements about prognosis and qualitative distinctions to be drawn about treatment need and delivery.127 There have been long-standing critiques, however, of a diagnosis-centric approach to mental health. For instance, in his 1940 book, Personality and Mental Illness, John Bowlby described the assumption that mental illnesses ‘were separate disease-entities each with its specific pathology’ as wholly ‘untenable’. He argued that more important than this or that symptom was ‘the total mental state, a correct estimation of which will be the only reliable guide to prognosis.’128 Following the publication of the third edition of the Diagnostic and Statistical Manual of Mental Disorders by the American Psychiatric Association in 1980, Bowlby criticized the ‘categorists’ for their search ‘for diagnostic criteria that distinguish the mental ill from the normal’. He urged collective critique of such a stance by ‘those others who, like myself, believe continuity to be a more fruitful perspective’.129 For instance, he argued that ‘personality disorders’ did not represent a discrete (p. 197) kind of illness but an intensified and persistent form of ordinary difficulties humans may experience in recognizing, contributing to, and making use of the good that can come from social relationships. They ‘reflect an individual’s impaired ability to recognise suitable and willing figures and/or an impaired ability to collaborate in rewarding relationships with any such figure when found.’130

Anna Freud was also a critic of diagnosis-focused mental health practice. In her essay, ‘The Symptomatology of Childhood’, she argued that ‘symptoms are no more than symbols, to be taken merely as indications that some mental turmoil is taking place’. Excessive emphasis should not be given to any individual symptom or set of symptoms, because ‘their place may be taken almost instantaneously by other pathological formations which, although overtly different, express the same latent content and may be no less aggravating for the individual’s life’.131 Part of Freud’s concern was that exactly the same set of symptoms, corresponding to a particular diagnostic profile, could reflect utterly different psychological processes, and vary in their clinical implications:

Symptoms may be no more than the child’s answer to some developmental stress and as such transitory, i.e., liable to pass away together with the maturational phase which has given rise to them. Or symptoms may represent a permanent countercathexis against some threatening drive derivative and as such be crippling to further development. Or symptoms, though pathological in origin, may nevertheless be ego-syntonic, and merged with the structure of the child’s personality to a degree which makes it difficult to distinguish between such manifestations as outward evidence of ongoing pathological involvement or as more or less normal, stable features of the individual’s character.132

Freud thus advocated redirection of attention away from diagnostic categories and towards mental turmoil conceptualized as a latent factor. Within her clinic (latterly the Anna Freud Centre) she ‘developed a system of five diagnostic categories for classifying children’s psychopathology: category 1 (essentially normal), 2 (transient symptomatology or developmental strain), 3 (neurotic, with “permanent regressions, fixations and symptom formation”), 4 (atypical, distorted personality development), and 5 (destructive processes disrupting mental growth)’.133 This quasi-interval scale was among the outcome measures used in the Anna Freud Centre retrospective study (see Chapter 2). Another was the Hampstead Child Adaptation Measure developed by Target and Fonagy, in which ‘the general adjustment of a child’ was measured on a 100-point scale.134

For their part, Fonagy and colleagues have acknowledged the pragmatic functions of diagnostic-based clinical practice and have argued against the abandonment of diagnoses. For instance, Fonagy was a co-author on a public letter advocating that diagnoses (p. 198) can facilitate understanding the ‘interrelation of psychological processes in an individual patient’, by offering a picture of how symptoms may cohere.135 Bateman and Fonagy have specifically defended the diagnosis of BPD, despite the fact that the category may be stigmatizing, and the fact that BPD is probably better conceptualized as a spectrum rather than a distinct clinical category.136 Fonagy would prefer a label such as ‘persistent distress disorder’ rather than BPD.137 But whatever the name he regards, on balance, some label as necessary to identify a group of patients with relatively distinct needs:

Mental health professionals have expressed a great deal of anxiety about giving a patient a diagnosis of personality disorder. Fears are rightly expressed about pejorative overtones, judgemental attitudes, blaming the patient, attacking the very ‘soul’ of the individual … Despite these potential drawbacks, we firmly believe that it is both necessary and constructive.138

Discussing use of the BPD diagnosis for adolescents, Fonagy has argued that ‘it is time for clinicians to stop hanging on to outdated ideas about the alleged impossibility of diagnostics, saying that the diagnosis may be iatrogenic, that the disorder is “only a life phase” that “they will grow out of” or that there is insufficient data. These are all unsubstantiated concerns that do not work in the interest of young patients.’139

To take another example: Fonagy has been willing to use the term ‘gender dysphoria’, because the diagnosis can help people questioning their gender identity to access specialist services. However, he has simultaneously expressed concern that ‘if we label gender dysphoria a mental disorder, and that disorder is stigmatizing, then that can be unhelpful and may actually reduce their willingness to explore the complexity of their feelings’.140 Against those who characterize transgender as an identity necessitating swift movement to medical interventions, Fonagy has characterized anyone professing total certainty about gender identity as likely to be in a non-mentalizing state. Their statements about themselves should therefore be treated with a degree of doubt by clinicians. However, Fonagy equally described anyone who thinks that gender identity can be reduced to chromosomal differences to equally be in a non-mentalizing state (presumably teleological mode).141 There are people, both children (p. 199) and adults, who will need specialist services to respond to discrepancies between their gender identity and societal gender norms. In some cases, these discrepancies may stem from non-mentalizing, which is why a chance to talk to a psychological professional should be a first step. However, in other cases, discrepancies may remain for individuals who are mentalizing well, and then medical intervention may be helpful. While gender identity, like all self-representation, is always a confabulation for Fonagy, if a diagnosis like gender dysphoria can be made non-stigmatizing, then it will have value in allowing individuals to access specialist psychological and medical services.

In the case of BPD and gender dysphoria, Fonagy can be seen defending diagnostic categories. However, like Bowlby and Anna Freud, Fonagy and colleagues have more generally come down against diagnosis-focused mental health practice.142 In light of their account of non-mentalizing and epistemic mistrust as transdiagnostic risk factors, they have expressed concern at the limitations of a category-based system. They regard this system as neglecting five features of mental health that are of particular consequence for theoretical and clinical work, as well as health policy:

  1. 1. Continuities between ordinary difficulties and mental illness.

  2. 2. The interaction and patterns of co-occurrence between symptoms separated into discrete diagnoses.

  3. 3. The variability of potential symptoms within many diagnoses which may mask different aetiologies and trajectories.

  4. 4. The variability in day-to-day functioning for individuals with the same diagnosis—or with no diagnosis.

  5. 5. Factors that specifically contribute to thriving and resistance to mental illness, as distinct from vulnerability or risk factors.143

(p. 200) Fonagy and colleagues have also highlighted findings by Goldberg and colleagues, which show that, while there is much individual variation, on average there is a small decline in clinical effectiveness for every year of experience a therapist accrues.144 Fonagy and colleagues conclude that these findings ‘might arise from the fact that less experienced therapists are more able to see the individual patient in all his or her subjective complexity rather than as a “walking diagnostic prototype” ’.145 Such prototyping undermines the clinician’s recognition of the specificity of their patient and their mental states, which would form the basis for prompting epistemic trust by the patient. Fonagy and colleagues have acknowledged that the spread of manualized therapy, a development they otherwise regard as positive and have helped support, may nonetheless have facilitated this prototyping process.146 This has prompted the use of web-based user-adaptable treatment manuals at the Anna Freud Centre for developing mentalizing teams within organizations,147 though conventional manuals currently remain used in most MBT.

Fonagy and colleagues have seen support for their criticisms of diagnosis-focused mental health practice in work by Caspi and colleagues on the ‘p-factor’, a transdiagnostic latent construct in the formation of mental health symptoms.148 As we have seen, precursors of the idea of transdiagnostic latent construct can be found already in the work of Bowlby and Anna Freud. In fact, already in 1946, the outgoing President of the British Psychoanalytic Society, Ernest Jones, had theorized in his valedictory address that later researchers would discover a general psychopathology factor equivalent to the general intelligence factor assessed as IQ.149 Yet, until recently, this hypothesis has lacked empirical support. In 2014, Caspi and collaborators reported from a factor analysis of mental health symptoms shown by participants in the Dunedin Multidisciplinary Health and Development Study. The diverse symptoms initially clustered into three higher order factors: internalizing symptoms, externalizing symptoms, and disorders of thought. However, more variance was accounted for by a model assuming one general psychopathology dimension. Caspi and colleagues termed this the ‘p-factor’. They found that the p-factor was associated with three character (p. 201) traits: low Agreeableness, low Conscientiousness, and high Neuroticism. The researchers found that child maltreatment was associated with the p-factor, but not with the lower-level factors once the p-factor was included in the model.

In 2015, Fonagy and Campbell acknowledged that there may be diverse forms of mental illness with particular properties, and even predictable profiles of symptoms that give the impression of relative distinct disorders. Nonetheless, Fonagy and Campbell argued, ambitiously, that ‘the p factor is a measurement of epistemic trust’.150 In a later statement, Fonagy has been more cautious: ‘it might be helpful to consider the p factor as a proxy for impairments in epistemic trust’.151 Whether strongly or more tentatively presented, in the perspective offered by Fonagy and colleagues, the initiation and maintenance of symptoms can be accounted for in terms of epistemic mistrust and its relationship with non-mentalizing modes; and their prevention and reduction can be accounted for in terms of epistemic trust and its relationship with mentalization. Fonagy and colleagues argued that the character traits of low Agreeableness, low Conscientiousness, and high Neuroticism identified by Caspi and collaborators should be regarded as reflecting ‘emotion dysregulation, impulsivity and social dysfunction’, the key symptoms of BPD.152 While these features may be misrecognized as aspects of a person’s personality, Fonagy and colleagues proposed that all three should be regarded as stemming from dysfunctions in epistemic trust. For instance, epistemic mistrust leads to non-mentalizing modes, which in turn prompt impulsive action because the potential for reconsideration is taken offline.153

Collaborators of Fonagy have pursued this perspective through empirical studies. Sharp and colleagues conducted a factor analysis of the symptoms of 966 inpatients in a paper published in 2015. On the one hand, there were some distinguishing features for specific forms of personality disorder. Disregard for safety, aggression, and lack of remorse seemed to form a profile aligned with the antisocial personality disorder (ASPD) diagnosis. Odd behaviour and beliefs aligned with the schizotypal personality disorder diagnosis. Grandiosity and need for admiration aligned with the diagnosis of narcissistic personality disorder. Moral inflexibility aligned with the obsessive–compulsive personality disorder. And preoccupation with rejection aligned with avoidant personality disorder. On the other hand, most of the variance in symptoms could be accounted for on the assumption of a general psychopathology (p. 202) factor.154 Sharp and colleagues were interested that no specific profile emerged for BPD. Instead, the items associated with BPD fell almost exclusively with the general factor.155 This included identity disturbance, feelings of emptiness, self-harm, instability of social relationships, and social actions motivated by a fear of abandonment. From this, Sharp and colleagues concluded that the criteria for BPD do not reflect a ‘type’ of personality disorder, but ‘best reflect general impairments’.156 This appears to offer empirical support for a position first put forward by Fonagy’s friend Otto Kernberg in the 1980s, that BPD is not a discrete form of mental illness but an indication of general psychosocial difficulties.157 On this basis of both theory and empirical findings, Bateman and Fonagy offered the clinical observation that, when symptoms of various forms of personality disorder are present, ‘treatment of BPD is required if comorbid disorder is to be treated’, because it is the general difficulties characteristic of BPD that are the more basic, and fundamentally causal.158

In another study, also published in 2015, Patalay, Fonagy, and colleagues conducted a factor analysis of self-reported symptoms in a community-based sample of 23,477 early adolescents, with follow-up data after one year for 10,270 participants.159 They found that most of the variance in symptoms in their sample could be accounted for in terms of a general p-factor, with lower-level dimensions of internalizing and externalizing symptoms making a small additional contribution (see Figure 7.4). Without the general p-factor taken into account, internalizing and externalizing symptoms were clearly correlated (r =.45). However, after removing variance associated with the p-factor, internalizing and externalizing symptoms were negatively correlated. The associations between female gender and internalizing symptoms, and between male gender and externalizing symptoms, also increased. This suggests that the p-factor may have been masking the extent of the gender specificity of these symptoms in previous epidemiological research on mental health.

Figure 7.4 Bi-factor model with the item loadings onto the internalizing and externalizing dimensions and the general psychopathology bi-factor.

Figure 7.4 Bi-factor model with the item loadings onto the internalizing and externalizing dimensions and the general psychopathology bi-factor.

Source: Reproduced from Praveetha Patalay et al., ‘A General Psychopathology Factor in Early Adolescence’, British Journal of Psychiatry, 207(1): 15–22, Figure 1, DOI: https://doi.org/10.1192/bjp.bp.114.149591 Copyright © Royal College of Psychiatrists, 2015.

In terms of prediction, the general p-factor had a strong association with mental health symptoms a year on, with the internalizing and externalizing dimensions comparatively less effective as predictors. Patalay, Fonagy, and colleagues claimed that diagnosis-focused mental health practice focuses attention on specific pathways to specific problems, ignoring the general factors that provide the motive force behind mental illness. In particular, the researchers proposed that diagnosis-focused mental health practice fails to consider the way (p. 203) (p. 204) that mental illnesses of various kinds may be prompted and maintained by the breakdown of processes that usually offer protection, above all epistemic trust and, stemming from it, the capacity to make use of social relationships for responding effectively to adversities. The prediction from the general p-factor more than the internalizing and externalizing dimensions in the study is a finding with multiple ramifications. For instance, it implies that trials of psychological interventions may be radically underestimating effectiveness, if diagnoses are used as outcome measures rather than a general p-factor.

A later study by St Clair, Neufeld, Jones, Fonagy, and colleagues reported from a study of the self-reported symptoms of 2,257 volunteer participants aged between 14 and 25.160 Again, a general latent factor accounted for much of the variance. There were also five lower-level factors: self-confidence, antisocial behaviour, worry and fear, aberrant thinking (including obsessional/compulsive and psychotic-like experience items), and mood. One important finding was that states of anxiety and depression,161 as well as items related to somatic functioning (e.g. ‘moved and walked more slowly than usual’, ‘didn’t sleep as well as usual’), loaded exclusively on the general factor and not on any of the lower-level factors. This aligns with claims made by Luyten, Fonagy, and colleagues that somatic disorders should be regarded as expectable features of the continuum of mental distress, rather than a discrete diagnostic category.162

A second finding of particular theoretical importance from the St Clair study was that agreement or disagreement with self-report items about negative aspects of mental health only offered good discrimination at the high end of the general and specific factors. By contrast, agreement or disagreement with statements characterizing positive aspects of mental health offered discrimination at both ends of the spectrum. St Clair and colleagues argued that this underlines the necessity of concern with positive factors that protect against mental illness in understanding why people suffer more or less mental illness. This finding likewise aligns well with mentalization and epistemic trust situated by Fonagy and colleagues as transdiagnostic sources of resilience, which would otherwise be missed through a focus only on diagnostic pathology. However, it also suggests that risk factors may not simply be the inverse of protective factors: each may have distinct correlates. As we saw earlier, epistemic mistrust is not simply the opposite of epistemic trust, and they play rather different roles in Fonagy and colleagues’ developmental model. It may likewise be wondered whether mentalization is simply the opposite of non-mentalizing, or whether mentalization and non-mentalizing both respectively have distinct qualities and correlates that theory and measurement should seek to capture.163

(p. 205) A further notable finding from the study was that while psychotic experiences loaded in part with the aberrant thinking lower-level factor, they were also strongly associated with the general psychopathology factor. St Clair and colleagues interpreted the findings as showing that psychotic symptoms should also be regarded as expectable features of the continuum of the p-factor, perhaps characterizing especially its most comorbid and/or its most severe end, rather than a discrete diagnostic category.164 Some psychotic symptoms, as much as other mental health symptoms within the overall p-factor, may be sequelae of vulnerabilities such as early trauma.165 Developing ideas initially formulated by Fonagy in the treatment of Mr S (see Chapter 1), Debbané, Brent, Fonagy, and colleagues have proposed that, like other mental health difficulties, vulnerabilities to psychotic experiences will be exacerbated by emotion dysregulation and biases in interpersonal attribution characteristic of the p-factor. These will then remain uncorrected—and treatment will prove less effective—to the extent that modes of non-mentalization are used intensively, and epistemic vigilance locked in place.166 Teleological mode can hold individuals to a single way of interpreting others’ behaviour. Psychic equivalence can make passing thoughts or feelings seem characteristic of all reality. Pretend mode can sustain states of dissociation or hallucination, with hypermentalization contributing to improbable over-interpretation of others’ behaviour and motives. Externalization of the alien self can make an individual’s own negative feelings appear to be coming from outside. Fonagy and colleagues have regarded symptoms of psychosis either as a facet of the overall p-factor, or else as a subfactor on a level with internalizing and externalizing. In either case, they have conceptualized mentalization as protective against psychosis, and non-mentalization as predisposing and potentially exacerbating psychosis. However, there remain significant ambiguities regarding this claim. Not least, the precise basis of the conceptual distinction between pretend mode and psychosis remains under-elaborated: is pretend mode an independent contributor to (some forms of) psychosis, a component of (some forms of) psychosis, solely a reinforcer of psychotic symptoms, among the consequence of (some forms of) psychosis, or all four?

A final intriguing finding from the St Clair study was that higher scores on the p-factor, but not the specific factors, were associated generally with attending treatment for mental ill health. The exception here was that both the p-factor and the mood lower-level factor predicted attending treatment for depression. However, participants who only reported high symptoms of the p-factor and not the symptoms of any of the lower-level factors were no more likely to be receiving mental health services than individuals with few or no symptoms. There are various interpretations that could be offered of this finding. One would be that there is significant unmet need among the population of individuals with only symptoms of the p-factor. St Clair and colleagues offered a very different interpretation, proposing (p. 206) that these individuals have less need of clinical services. They therefore questioned the widespread use of measures of anxiety and depression in screening in primary and secondary care settings, because their interpretation was that patients with both specific and general mental health symptoms were the ones with the most need. Evidence in favour of this interpretation is that elevations of the p-factor alone, in the absence of any of the lower-level factors, was also not associated with an increase in use of alcohol, cannabis, illegal drugs, or non-suicidal self-injury. By contrast, the likelihood of report of these behaviours was elevated for participants scoring highly on both the p-factor and the anti-social lower-level factor, when compared with just those participants scoring highly on the anti-social factor.167

In a further study on the same sample, Polek and colleagues found that the p-factor contained high loadings on various sets of items.168 One set were directly suggestive of epistemic mistrust. This included endorsement of the item ‘I feel I have to be on my guard even with friends’, as well as low agreeableness and high antagonism. A second set indicated a perception of the views and norms of others as lacking relevance to the self. This included items such as ‘I do not care if I get into trouble’ and ‘I do not care about doing things well’. Another set were items suggestive of weak integration of self-representation, of emotion, and of social relationships, such as affirmative answers to ‘I am an odd, unusual person’ and ‘Do your moods change unpredictably?’. This was reflected in strong loading on self-report of feelings of social isolation, such as affirmative responses to the item: ‘Do you feel that you cannot get close to people?’ A further set were items from a measure of impulsivity, specifically those related to difficulties managing attention such as ‘I concentrate easily’ (reversed) and ‘I don’t pay attention’, as well as reporting acting without thinking about the consequences.169 Such findings are well aligned with the account by Fonagy and colleagues of close relations between epistemic trust, difficulties in social relationships, difficulties in the modulation of anger and sadness, instability and incoherence in self-representations, and disruption of attentional processes.

Recently, Gibbon, Nolte, and Fonagy examined the factor structure of self-report data from a large study. There were 338 participants from two sources: either patients with BPD, recruited from clinical services in London; or patients with ASPD recruited from probation services. Active psychosis was an exclusion criteria.170 The study also included 167 (p. 207) community volunteers to serve as controls. This study was the first to consider symptoms of personality disorder within the same analysis as other symptoms of mental ill health.171 The analysis found that the best fit for the data was a general p-factor plus four lower-level dimensions: internalizing, antisocial/hostility, thought disorder, and borderline. The symptoms that loaded most clearly on the general p-factor were interpersonal sensitivity, depression, anxiety, and psychoticism. Phobias, obsessive-compulsivity, post-traumatic stress, paranoid thoughts, and high scores on the ‘Difficulties in Emotion Regulation Scale’ were also, to an extent, characteristic of the general p-factor. This was much in line with the findings of Caspi and colleagues. The internalizing dimension was notable for having strong associations with the other lower-level dimensions. Symptoms of post-traumatic stress fell strongly with the internalizing dimension. In fact, there were significant indications that the internalizing dimension might be better modelled as part of the p-factor. By contrast, the antisocial/hostility lower-level dimension was notable for having few associations with the other lower-level dimensions, and items loaded only weakly on to the p-factor.

Gibbon, Nolte, and Fonagy found that self-report of childhood maltreatment was strongly associated with all four lower-level factors (rs = .53-.59). But these associations were very substantially attenuated (to rs = .20-.28) when the p-factor was included (r = .55), indicating that it was the associations with the p-factor that were driving the links to the lower-level factors. Self-report retrospective measures of childhood should not, however, be taken to be equivalent to the findings from prospective studies of maltreatment: a recent meta-analysis has shown that 52% of individuals with prospective observations of childhood maltreatment do not report it retrospectively, and 56% of individuals retrospectively reporting childhood maltreatment were not identified as maltreated in prospective studies.172 And McCrory and other colleagues at the Anna Freud Centre have specifically criticized the measure of childhood maltreatment used in the Gibbon study as overinclusive—for instance, identifying being called ‘lazy’ or ‘stupid’ by people in one’s family—including siblings—as child maltreatment.173

Gibbon and colleagues reported findings for reflective function, measured using the RFQ. For the certainty sub-scale (RFQc), associations with the four lower-level dimensions were moderate to strong (rs = .47-.55). However, these associations mostly disappeared entirely once the p-factor was included (r = .50), with the exception of the borderline lower-level dimension which retained a moderate correlation (r = .34). For the uncertainty sub-scale (RFQu), associations with the four lower-level dimensions were remarkably strong (rs = .62-.70). These associations were very substantially attenuated after inclusion of the p-factor.174 The RFQu had an association with the p-factor of r = .62, which is an unusually large effect size for a psychological study, to the point that it could suggest overlap between the constructs.175 Nonetheless, it should be pointed out that again the association between RFQu (p. 208) and the borderline dimension was not entirely attenuated when the p-factor was included, but remained moderate (r = .37), and the antisocial/hostile dimension also retained a meaningful association (r = .28). The strongest association between the RFQu and individual symptoms was with items from the ‘Difficulties in Emotion Regulation Scale’ (r = .69) and emotional instability (r = .61). The strongest association between the RFQc and individual symptoms was with items from the ‘Difficulties in Emotion Regulation Scale’ (r = .56), items understood to signal problems with identity176 (r = .51) and phobia (r = .47).

Overall, these findings offer clear support for the prediction by Fonagy and colleagues that the p-factor would be associated with non-mentalizing, though they align with other findings in recent years that generally suggest that RFQu has better associations with expectable correlates than RFQc. For instance, no association has been found for the RFQc with reported interpersonal difficulties or with the ‘Difficulties in Emotion Regulation Scale’, or with observer-reported measures of mentalizing such as the Parent Development Interview.177 The findings also suggest potential qualifications to theory. Antisocial/hostile behaviour was only weakly associated with the p-factor, and retained a significant association with RFQu even after the p-factor was included in the model. Yet more strongly, excessive certainty and uncertainty regarding mental states was associated with borderline symptoms even when the p-factor was taken into account. This qualifies the implication that the high end of the p-factor = borderline = low reflective function. There appear to be aspects of BPD—and conduct problems to an extent—predicted by non-mentalizing that are distinct from the p-factor. However, these findings should be treated with caution because they are likely to have been affected by the specific recruitment of patients with BPD and ASPD, which may have bolstered the apparent independence of these lower-level factors. It might be, for instance, that the thought disorder lower-level factor would have retained its association with the RFQ if the sample had purposefully included a group of patients with psychotic and other anomalous experiences, rather than excluding participants with active psychosis. Or that the internalizing lower-level factor might have had more independence from the p-factor if the sample had included a group of patients with dysthymia.178

(p. 209) Symptom contexts and networks

In diagnostic-centric practice within bureaucratically oriented healthcare systems or those structured by health insurance, symptoms may not be fully registered unless they can be understood as surface manifestations of an underlying diagnostic entity, with expectable causes, qualities, and prognosis. This is an analogy with physical illness, where a given symptom (e.g. coughing up blood) is understood to have a single cause, which will eliminate the symptoms when removed. This perspective is advocated by some mental health professionals and researchers. However, more structurally, it has been held in place in part by the wish of healthcare systems for an orderly and auditable response to individual needs and health trajectories.179

For their part, Fonagy and colleagues have advocated in favour of a focus on symptoms and their interactions as the primary level of analysis.180 This has led them to examine the role of the p-factor, as a transdiagnostic latent factor in the formation of mental health symptoms. They have, with varying degrees of caution, identified the p-factor with impairment in epistemic trust. As we saw epistemic trust, Fonagy and colleagues regard epistemic trust as a self-righting mechanism, which offers a means for learning, feedback, and adjustment in the face of challenges, as well as capitalization on available social supports—including the capacity to make effective use of mental health services. A reason for the occasional caution shown by Fonagy and colleagues in making this claim may well be that elsewhere they have considered various reasons why mental health symptoms might intersect and reinforce, which would contribute to a general latent factor. Epistemic trust may well be one contributor to the latent general dimension, but its identification with the p-factor would then be overhasty. Here we will identify six further factors in the writings of Fonagy and colleagues that might contribute to the existence of a latent factor of mental health symptoms.

A second reason why mental health symptoms may form a latent dimension is that the diverse conditions that prompt or maintain them may intersect and reinforce. This point has been a particular argument developed in Fonagy’s writings with his wife Anna Higgitt in the 1990s and 2000s, reflecting her longstanding interest in health policy and work as a senior policy adviser to the Department for Health.181 Through research with prospective studies in the 1980s, Michael Rutter had first shown that most mental health conditions are best predicted not by particular kinds of risks but the aggregation of adversities or gaps in social support.182 These findings have subsequently been well replicated.183 Fonagy and Higgitt (p. 210) were struck by these findings, which stood in contrast to the psychoanalytic models of the day, in which early psychological risks were privileged as causes almost to the exclusion of the wider ecology of a person in the course of their development. They took from Rutter the conclusion that ‘risk conditions occur simultaneously’, and their ramification means that the accumulation of adversities or gaps in social support result not in an additive, but in a multiplying, likelihood of mental health symptoms of one form or another. The same was true of other outcomes such as criminal behaviour and substance use. The implication was that ‘outcomes for most etiologies exist along a continuum’.184

A latent p-factor may find some of its basis, therefore, in the fact that the adversities that predispose or maintain mental illness in general can interest and reinforce one another. Fonagy and Higgitt offered the example of the intersection of social disadvantage and a very disappointing life event.185 This intersection will create much more risk of various forms of mental illness than either of the risks alone, in part because the adversities may hit harder, and in part because each may knock out factors that would otherwise buffer the effects of the other. While individual qualities may offer protection to an extent, Fonagy and Higgitt claimed that this protection is always and necessarily lost in the face of accumulating risks: ‘we may expect children to be resilient to one or more of such stressors, but as risk compounds even the strongest constitutionally must succumb.’ As adversities accumulate and supports decrease in the course of development, so too will various mental health symptoms. In recent work, Fonagy has highlighted that the association between ramifying developmental risk factors and later outcomes will have a diversity of mediators.186 Some of the effects of adversity and lack of support on mental illness may be mediated by impairments in epistemic trust. Some of the effects, however, may be direct or operate through other mechanisms. As such, the coincidence of mental symptoms may in part reflect impairment of epistemic trust, but may also in part reflect the intersection and mutual reinforcement of adversities and gaps in social support.

A third reason that mental health symptoms may form a general latent factor can be extrapolated from the discussion of attachment and adaptation earlier in this chapter. Proposals by attachment researchers suggested to Fonagy and colleagues that human evolution may have pre-primed a repertoire of strategies and dispositions, which respond to escalating adversities and lack of support with coping strategies with the predictable outcome of short-term pay-offs but with a long-term price. Crittenden argued that humans have evolved to respond to dangerous and adverse circumstances with forms of perception and behaviour that exclude information about relationships that would hinder coping. In one class of response, individuals exclude emotional information prompted within relationships, because this information is regarded as disrupting the capacity to soldier on. In the other class of response, individuals exclude information about the temporal and causal sequencing of others’ availability, permitting the maintenance of vigilance. In both cases, however, (p. 211) Crittenden theorized a spectrum of transformations of information about relationships, ranging through truth, error, omission, distortion, and deception. She held that diverse forms of mental health symptoms will increase, the further along this spectrum one goes. Reflecting on Crittenden’s work, Fonagy reinterpreted this dimension as ‘congruent’ with a spectrum of epistemic mistrust.187 Certainly, both models are centrally concerned with the extent to which information in relationships is processed accurately and experienced as relevant to the self. Nonetheless, it is not clear that all coping strategies accepting long-term costs for short-term pay-offs solely reflect epistemic trust.

Another attachment theorist who, as we saw, has thought about attachment in terms of strategies and long-term/short-term transactions is Belsky. As we saw adaptation and attachment, Belsky identified that insecure infant attachment is associated with early menarche and with more risk-taking behaviours. This offers support for his theory that evolution has primed humans to treat adverse early experience as a prime for biological and behavioural responses that prioritize short-term survival and reproduction, even if this may come with a long-term price. A portion of this price may be physiological, in terms of health implications. However, other aspects may register as mental health symptoms, such as anxiety, attentional problems, impulsivity, and conduct problems. The idea of strategies with expectable short-term benefits but long-term disadvantages may in some regards align with the concept of epistemic trust, in the proposal that human evolution has offered various ways that, even without the conscious intention of the individual, we may be disposed to respond to developmental adversity with less faith that all will work out. However, it does not seem likely that all the features of a life history strategy focused on short-term pay-off can be reduced to epistemic mistrust. The metabolic, social, and psychological strain and ramifications of this strategy may be regarded as another reason why symptoms may form a latent p-factor.188

A fourth contributor to the p-factor as a latent dimension may stem from symptoms that arise to mitigate other symptoms, and so predictably co-vary. This was a classic argument of Winnicott’s, who argued that some symptoms—such as depression—should be examined carefully to see whether they counterbalance other symptoms and result, overall, in greater psychological health than if they had been absent.189 Though Fonagy and colleagues have not tended to follow Winnicott in this interpretation of depression, they have agreed on the general point. The example most often discussed by Luyten, Fonagy, and their collaborators is that overeating may function as a self-soothing or self-medicating strategy as a response to felt distress or emptiness. Prospective research has documented a sharp dose–response relationship between an individual’s exposure to various adversities and their likelihood of disordered overeating.190 This relationship may be underpinned in part by the role of serotonin and dopamine in both food and mood, but there is also clearly a wider context implicating (p. 212) attempts to regulate self–other relationships, feelings of privacy and self-ownership, as well as the cultural meanings of embodiment.191 Skårderud and Fonagy have argued that the symptoms of disordered eating should be regarded as stemming from a ‘need to drown out painful self-states’, and therefore ‘as attempts at recovering cohesion, vitality and self-regulation’.192 Equivalent arguments may be made for hypersomnia and insomnia, both of which may be initiated or sustained as a means to mitigate other mental health symptoms, and so will predictably co-occur with them.193 Likewise, worry may develop as a response to other symptoms, as a preoccupation with attempting to resolve them. In fact, Fonagy and colleagues have identified that the absence of worry in the context of other mental health symptoms may come with its own problems, such as maintenance of risky behaviour without concern for possible consequences.194 This aligns with the findings of anxiety as a positive prognostic sign in the Anna Freud Centre retrospective study (see Chapter 2).

A fifth contribution to the p-factor as a latent dimension may lie in reciprocally reinforcing relationships between mental health symptoms. The most salient of these for Fonagy and colleagues are the reciprocal relationships between problems in affect modulation, attentional difficulties, non-mentalizing, and a paucity of stable social relationships. In fact, Bateman and Fonagy observed that these are abstractions, and in reality each is likely a component of the others, with blurred lines between them. It is often the case that ‘one may end up triggering another’, especially towards the more severe end of mental illness.195 Reinforcing and intensifying interactions may in fact occur between symptoms that were initiated as attempts at coping. For instance over- or under-eating may begin as a coping strategy for mitigating rumination or feeling low, only to develop over time into a set of symptoms dominating the individual’s experience of food, or even of social interactions.196 To take another example, reduced hours of sleep may have developed as a tactic to mitigate rumination or low mood that would otherwise occur during attempts to sleep. However, insufficient sleep is likely to contribute further over time to both symptoms. In fact, Patalay, Fonagy, and colleagues found that poor sleep was the single strongest loading on the p-factor of any mental health symptom, higher even than distress or anger.197

Another set of reinforcing and intensifying relationships between symptoms have been discussed by Luyten, Fonagy, and colleagues in relation to severely depressed states. Severely depressed states may deplete the attentional focus required to clearly conceive of or reconsider the thoughts and feelings of oneself or others. This may prompt entry into a state of (p. 213) psychic equivalence, in which feeling low defines reality.198 Or even when mental states are considered, in a state of affective hypoactivation use of reflections to account for and explain behaviour or experiences may be felt as futile.199 In turn, modes of non-mentalizing will contribute to states of extreme certainty or extreme doubt about the meaning of relationships with other people or aspects of the self-representation relevant to action. This can directly prompt further feelings of depression, as well as responses such as social withdrawal that reinforce these symptoms and that hinder future mentalizing. Luyten, Fonagy, and colleagues have referred to such patterns as ‘dysfunctional interpersonal transaction cycles’.200 However, they reported empirical findings that scores on the reflective function scale are not associated with depression scores among patients with severe depression. Instead, reflective function was negatively associated with the chronicity of depression and the likelihood of inpatient admission.201 Such findings suggest that non-mentalizing may sustain depression, and mediate interpersonal transaction cycles, but is not itself its initial basis.202 In the model proposed by Bateman and Fonagy, difficulties in forming and making use of relationships prompts feeling low, which in turn takes mentalizing offline and reinforces epistemic vigilance.203 This cascade hinders recovery from depression and increases the risk of relapse following recovery.

To take another example of disadvantageous interpersonal transaction cycles: Luyten, Fonagy, and colleagues have also proposed that substance and alcohol use may begin as a ‘means of silencing painful thoughts’.204 However, in serving this function, and especially when drug and alcohol use are one of the few available tactics at the disposal of the individual for achieving short-term regulation, they may contribute over time to reciprocal relationships between problems with affect modulation, attentional difficulties, use of non-mentalizing modes, and a paucity of stable social relationships. St Clair and colleagues (p. 214) examined associations between drug and alcohol use with the p-factor. Daily alcohol use was positively associated with the p-factor, but weakly (r = .18). Frequent alcohol consumption is normative in England, which may mask associations between the p-factor and drinking to silence painful thoughts (and perhaps feelings) in a brief self-report measure. Or painful thoughts and feelings may prompt alcohol use in the general population without contributing to the p-factor.

By contrast, the association between the p-factor and various forms of substance use was much more substantial (r = .34 for cannabis and .37 for other illegal drugs).205 Skårderud and Fonagy have described the ‘polysemy’ of behaviour, its embeddedness in multiple systems of causation and potential meaning-making. They used this idea to argue against interpretations of particular mental health symptoms as having any ‘one’ meaning.206 In this light, there may be such polysemy in why substance use is associated with other mental health problems:

  • Epistemic mistrust may contribute to a sense of risks as irrelevant, credulity towards peer norms, difficulties learning lessons for future behaviour from experiences with negative consequences, and difficulties in seeking or making use of social or medical support for substance addiction.

  • The multiplier effect of risk may contribute to the availability of substances, the presence of substance-using peers, and contexts in which drug use is frequent.

  • Life histories primed for short-term benefits may contribute to substance use through impulsivity, feelings of urgency in response to distress, and intensified neuroendocrinal rewards.

  • Substance use may offer significant short-term benefits in self-medicating for other symptoms such as distress or rumination

  • Substance use may also contribute to the cause or maintenance of other mental health problems, as well as reducing the effectiveness of protective factors—for instance, by hindering mentalising.

As the illustration of substance use suggests, there may be a variety of processes that contribute to the existence of a general latent dimension for mental health symptoms. The claim by Fonagy and Campbell that the ‘the p factor is a measurement of epistemic trust’207 is over-strong. Evidence in favour of a link between the two is still emerging, and there seem to be a variety of other processes in play. However, the considerations elsewhere in the work of these researchers and their collaborators offer the basis for an account of symptom contexts and networks in clustering mental health symptoms together into a latent dimension. The ideas present in the work of Fonagy and colleagues includes insights from cognitive and (p. 215) information processing (epistemic mistrust), developmental psychopathology (multiplier effect of risk), evolutionary socio-biology (life history theory), strengths-based approaches (self-medication) as well as network approaches (reciprocal instigation and reinforcement). All these processes may also be relevant to how amenable symptoms are to mental health intervention. The work by Fonagy and colleagues to apply their ideas to therapeutic practice will be the focus on the next chapter.

Notes:

1 Freud, S. ([1901] 2001). The Psychopathology of Everyday Life, in James Strachey (ed.), The Standard Edition of the Complete Psychological Works of Sigmund Freud, Volume 6, London: Vintage.

2 Freud, S. ([1923] 2001). ‘Two Encyclopaedia Articles’, in James Strachey (ed.), The Standard Edition of the Complete Psychological Works of Sigmund Freud, Volume 18, London: Vintage, pp. 233–260, p. 241.

3 Luyten, P. and Fonagy. P. (2012). ‘The Multidimensional Construct of Mentalization and its Relevance to Understanding Borderline Personality Disorder’, in A. Fotopoulou, D. Pfaff, and M. A. Conway (eds), From the Couch to the Lab: Trends in Psychodynamic Neuroscience, Oxford: Oxford University Press, pp. 405–427, p. 406.

4 Bach, S., Mayes, L., Alvarez, A. and Fonagy, F. (2000). ‘Panel 1: Definition of the Self’. Journal of Infant, Child & Adolescent Psychotherapy, 1(3): 5–24, p. 21.

5 Fonagy, P. and Target, M. (2000). ‘Playing with Reality: III. The Persistence of Dual Psychic Reality in Borderline Patients’. The International Journal of Psychoanalysis, 81(5): 853–873, p. 859.

6 Kernberg, O. F. (1997). ‘The Nature of Interpretation: Intersubjectivity and the Third Position’. The American Journal of Psychoanalysis, 57(4): 297–312; Holmes, J. (1998). ‘The Changing Aims of Psychoanalytic Psychotherapy: An Integrative Perspective’. The International Journal of Psychoanalysis, 79: 227–240.

7 E.g. Shedler, J., Beck, A. T., Fonagy, P., Gabbard, G. O., Kernberg, O., Michels, R., and Westen, D. (2011). ‘Response to Skodol Letter’. American Journal of Psychiatry, 168(1): 97–98; Chiesa, M., Fonagy, P., Holmes, J., and Drahorad, C. (2004). ‘Residential versus Community Treatment of Personality Disorders: A Comparative Study of Three Treatment Programs’. American Journal of Psychiatry, 161(8): 1463–1470. See also Hepworth, M. (2019). ‘“Peripheral Visions”: A Conference Celebrating the Contribution of Jeremy Holmes and 20 Years of Psychoanalytic Studies at Exeter University, UK’. British Journal of Psychotherapy, 35(2): 215–216.

8 Holmes, J. (2005). ‘Notes on Mentalizing—Old Hat, or New Wine?’, British Journal of Psychotherapy, 22(2): 179–198; Kernberg, O. F., Yeomans, F. E., Clarkin, J. F. and Levy, K. N. (2008). ‘Transference Focused Psychotherapy: Overview and Update’. The International Journal of Psychoanalysis, 89: 601–620.

9 E.g. Blundo, R. (2001). ‘Learning Strengths-Based Practice: Challenging our Personal and Professional Frames’. Families in Society, 82(3): 296–304; Crossley, N. (2004). ‘Not being Mentally Ill: Social Movements, System Survivors and the Oppositional Habitus’. Anthropology & Medicine, 11(2): 161–180.

10 Khantzian, E. J. (1997). ‘The Self-Medication Hypothesis of Substance Use Disorders: A Reconsideration and Recent Applications’. Harvard Review of Psychiatry, 4(5): 231–244.

11 See e.g. Allison, E. and Fonagy, P. (2016). ‘When is Truth Relevant?’. Psychoanalytic Quarterly, 85(2): 275–303, p. 286. They now acknowledge ‘when we think about mental ill health, particularly in children and young people, inorganic metaphors of damage or even plasticity may be less helpful than ways of thinking that recognize the human capacity for growth and adaptation in conditions. Of course, this capacity is not infinite and can be constrained by both biological and environmental factors, but it is at times grossly underestimated both by those who tend to think of disorder in terms of irreparable damage and by well-meaning professionals who believe that recovery occurs only as a consequence of intervention’, Allison, E. and Campbell, C. (2019). Transforming Child Mental Health: Principles of Sustainable Development, London: Anna Freud Centre.

12 Freud, A. (1962). ‘Assessment of Childhood Disturbances’. The Psychoanalytic Study of the Child, 17(1): 149–158, p. 152–3. See also Joffe, W. G. and Sandler, J. (1979). ‘Adaptation and Individuation’. Bulletin of the Anna Freud Centre, 2(3): 127–161; Sandler, A.-M. (2012). ‘Anna Freud’s Influence on Contemporary Thinking about the Child’, in N. Malberg and J. Raphael-Leff (eds), The Anna Freud Tradition: Lines of Development, London: Karnac Books; Midgley, N. (2012). Reading Anna Freud, London: Routledge.

13 The work of Anna Freud and Heinz Hartmann would together prove influential for the central role of the concept of ‘adaptation’ within subsequent ego psychology. See e.g. Rapaport, D. and Gill, M. M. (1959). ‘The Points of View and Assumptions of Metapsychology’. The International Journal of Psychoanalysis, 40: 153–162. It is commonly pointed out that ‘adaptation’ receives special attention in the work of these central European Jewish émigrées, who had to acclimatize to America and England. Two generations later, there seem relevant analogies to the case of Fonagy.

14 Freud, A. (1981). ‘Insight: Its Presence and Absence as a Factor in Normal Development’. The Psychoanalytic Study of the Child, 36(1): 241–249, p. 243.

15 Freud, A. (1983). ‘The Past Revisited’. Bulletin of the Anna Freud Centre, 6(1): 107–113, p. 111.

16 Fonagy, P., Edgcumbe, R., Target, M., and Miller, J. (1999). Contemporary Psychodynamic Child Therapy: Theory and Technique, London: Anna Freud Centre and University College London: ‘Early in treatment, it is usually safer to begin with verbalisation of affect until the child’s emotional balance is clear enough for the analyst to predict how he might react to interpretation of unconscious wishes. Even so, it is also usually more readily acceptable to the child to interpret defences before the wishes or impulses against which the child is defending. Bypassing the defences with a direct interpretation of an unconscious wish ignores the fact that the child needs his defences and has good reason for erecting them’ (p. 103). See also Fonagy, P., Bleiberg, E., and Target, M. (1997). ‘Child Psychoanalysis: Critical Overview and a Proposed Reconsideration’. Child and Adolescent Psychiatric Clinics of North America, 6: 1–38: ‘For these children, the possibility of relinquishing their single, fixed, concrete model of self-other relatedness triggers terror and confusion. Therapists more profitably point out the advantages of not changing, remarking on the price children would pay if they were to give up their maladaptive, but often life-saving, defences and self-other internal models’ (p. 31).

17 BBC2 (2001). ‘Taming the Problem Child, 9 p.m. Horizon, Thursday, 8 March. Accessed at: http://www.bbc.co.uk/science/horizon/2000/problemchild_transcript.shtml.

18 Target, M. (2018). ‘20/20 Hindsight: A 25-year Programme at the Anna Freud Centre of Efficacy and Effectiveness Research on Child Psychoanalytic Psychotherapy’, Psychotherapy Research, 28(1): 30–46: ’Our effort in this direction, the HCAM [Hampstead Child Adaptation Measure], distinguished 14 domains of normal and abnormal development which could be tracked cross-sectionally and longitudinally, going well beyond symptoms to cover things like the child’s capacity to have relationships, to play, learn, look after himself physically, regulate moods, manage stress, and so on’ (p. 43). To take another example, a scale for the extent to which an individual’s behaviour appears adaptive would be developed by Steele, H., Steele, M., and Kriss, A. (2009). The Friends and Family Interview (FFI) Coding Guidelines. Unpublished manuscript: ‘This scale refers specifically to responses to the question asking what the respondent does when distressed or upset. An adaptive strategy may involve seeking comfort from others (e.g. parents, friends, or siblings), engaging in a favorite activity that relieves their unhappiness (e.g. listening to music, walking the dog), or simply thinking things through.’ See also Chow, C. C., Nolte, T., Cohen, D., Fearon, R. P., and Shmueli-Goetz, Y. (2017). ‘Reflective Functioning and Adolescent Psychological Adaptation: The Validity of the Reflective Functioning Scale–Adolescent Version’. Psychoanalytic Psychology, 34(4): 404–413.

19 Sandler, J. and Freud, A. (1981). ‘Discussions in the Hampstead Index on “The Ego and the Mechanisms of Defence”: V. The Mechanisms of Defence, Part 2’. Bulletin of the Anna Freud Centre, 4(4): 231–277; Sandler, J. and Freud, A. (1982). ‘Discussions in the Hampstead Index on “The Ego and the Mechanisms of Defence”: VIII. Denial in Word and Act’. Bulletin of the Anna Freud Centre, 5(3): 175–187. See also Joffe, W. G. and Sandler, J. (1968). ‘Comments on the Psychoanalytic Psychology of Adaptation, with Special Reference to the Role of Affects and the Representational’. The International Journal of Psychoanalysis, 49: 445–454. Early concerns about the term ‘adaptation’ were raised by Hartmann, one of the popularizers of the concept within the psychoanalytic community. See Hartmann, H. (1939). ‘Psycho-Analysis and the Concept of Health’. The International Journal of Psychoanalysis, 20: 308–321.

20 Winnicott, in turn, used the term ‘adaption’ in a different, slightly non-standard way, developing further the idea of ‘responding’. His predominant use of the term characterized the way that a caregiver moulds his or her body and interactions to fit around those of the child. Winnicott, D. W. ([1960] 1965). The Maturational Process and the Facilitating Environment, New York: International Universities Press.

21 See also Lee, N. N. (2014). ‘Sublimated or Castrated Psychoanalysis? Adorno’s Critique of the Revisionist Psychoanalysis: An Introduction to “The Revisionist Psychoanalysis”’. Philosophy & Social Criticism, 40(3): 309–338.

22 Ainsworth, M. (1984). ‘Attachment, Adaptation and Continuity’. Paper presented at the International Conference on Infant Studies, April, John Bowlby Archive, Wellcome Collections, PP/Bow/J.1/57. See Duschinsky, R. (2020). Cornerstones of Attachment Research, Oxford: Oxford University Press, Chapter 2.

23 Ibid.: ‘In the phylogenetic or evolutionary sense adaptation implies that in the course of natural selection those behaviours that yield survival advantage in the environment in which the evolutionary changes are taking place become part of the behavioural repertoire characteristic of a species … In the ontogenetic sense adaptation refers to the process through which an organism adjusts to its environment in the course of development.’

25 Main, M. (1979). ‘The “Ultimate” Causation of Some Infant Attachment Phenomena: Further Answers, Further Phenomena, Further Questions’. Behavioral and Brain Sciences, 2: 640–643, p. 643.

26 Main, M. (1981). ‘Avoidance in the Service of Proximity: A Working Paper’, in K. Immelmann, B. Barlow, L. Petrovich, and M. Main (eds), Behavioral Development: The Bielefield Interdisciplinary Project, Cambridge: Cambridge University Press, pp. 694–699, p. 686.

27 Ainsworth, M. (1984). ‘Attachment, Adaptation and Continuity’. Paper presented at the International Conference on Infant Studies, April, John Bowlby Archive, Wellcome Collections, PP/Bow/J.1/57.

28 Crittenden, P. M. (2000). ‘A Dynamic-Maturational Model of the Function, Development, and Organization of Human Relationships’, in R. S. L. Mills, and S. Duck (eds), Developmental Psychology of Personal Relationships, New York: Wiley, pp. 199–218.

29 Crittenden, P. M. (1997). ‘Truth, Error, Omission, Distortion, and Deception: The Application of Attachment Theory to the Assessment and Treatment of Psychological Disorder’, in S. M. C. Dollinger and L. F. DiLalla (eds), Assessment and Intervention across the Lifespan, Hillsdale, NJ: Erlbaum, pp. 35–76.

30 Crittenden, P. M. (1997). ‘Toward an Integrative Theory of Trauma: A Dynamic-Maturational Approach’, in D. Cicchetti and S. Toth (eds), The Rochester Symposium on Developmental Psychopathology, Volume 10, Risk, Trauma, and Mental Processes, Rochester, NY: University of Rochester Press, pp. 34–84; Crittenden, P. M. and Heller, M. B. (2017). ‘The Roots of Chronic PTSD: Childhood Trauma, Information Processing, and Self-Protective Strategies’, Chronic Stress, 1: 1–13.

31 Fonagy, P. (2013). ‘Commentary on “Letters from Ainsworth: Contesting the ‘Organization’ of Attachment”’. Journal of the Canadian Academy of Child & Adolescent Psychiatry, 22(2): 178–179: ‘Most would agree that, clinically, Crittenden’s approach is more inspiring, particularly in understanding the behaviour of children and young people whose life has been blighted by malevolence on the part of their carers’ (p. 179).

32 Luyten, P. and Blatt, S. J. (2012). ‘Psychodynamic Treatment of Depression’. Psychiatric Clinics, 35(1): 111–129. See also Rost, F., Luyten, P., and Fonagy, P. (2018). ‘The Anaclitic–Introjective Depression Assessment: Development and Preliminary Validity of an Observer‐Rated Measure’. Clinical Psychology & Psychotherapy, 25(2): 195–209.

33 Fonagy, P. (1999). ‘Points of Contact and Divergence between Psychoanalytic and Attachment Theories’. Psychoanalytic Inquiry, 19(4): 448–480: ‘Crittenden’s (1990) work has been particularly helpful in translating behaviors typical of avoidant and resistant attachment patterns into the language of the defensive behaviors of infancy’ (p. 452); Fonagy, P. (2016). ‘The Role of Attachment, Epistemic Trust and Resilience in Personality Disorder: A Trans-Theoretical Reformulation’. DMM News, 26 September. Accessed at: http://www.iasa-dmm.org/images/uploads/DMM%20%2322%20Sept%2016%20English.pdf: ‘The model’s emphasis on attachment as a means of protecting the self and one’s offspring from danger is then used to make a valuable—and in the process, valuably destigmatising—account of dysfunction as arising from knowledge acquired through environmental influence’ (p. 2).

34 See e.g. Fonagy, P. (2012). ‘The Neuroscience of Related Trauma and Evidence-Based Intervention’. Accessed at: https://smad6740.files.wordpress.com/2015/01/the-neuroscience-of-related-trauma-and-evidence-based-intervention1.pdf: Such ‘anomalies could confer a short-term advantage: Vigilance to threat; It is found in healthy soldiers exposed to combat. But they constitute a latent neural risk that predisposes to an increased likelihood of maladaptation in safe contexts (e.g. school) and of adult psychopathology.’

35 Belsky, J., Houts, R. M., and Fearon, R. P. (2010). ‘Infant Attachment Security and the Timing of Puberty: Testing an Evolutionary Hypothesis’. Psychological Science, 21(9): 1195–1201; Belsky, J., Steinberg, L., Houts, R., and Halpern-Felsher, B. (2010). ‘The Development of Reproductive Strategy in Females’. Developmental Psychology, 46(1): 120–128; Belsky, J. (2019). ‘Early-Life Adversity Accelerates Child and Adolescent Development’. Current Directions in Psychological Science, 28(3): 241-246. See also Szepsenwol, O. and Simpson, J. A. (2019). ‘Attachment within Life History Theory: An Evolutionary Perspective on Individual Differences in Attachment’. Current Opinion in Psychology, 25: 65–70.

36 Fonagy, P. (2013). ‘Commentary on “Letters from Ainsworth: Contesting the ‘Organization’ of Attachment”’. Journal of the Canadian Academy of Child & Adolescent Psychiatry, 22(2): 178–179, p. 179.

37 Bateman, A. W. and Fonagy, P. (2016). Mentalization-Based Treatment for Personality Disorders: A Practical Guide (2nd edn), Oxford: Oxford University Press, p. 62.

38 Fonagy, P. and Campbell, C. (2017). ‘What Touch can Communicate: Commentary on “Mentalizing Homeostasis: The Social Origins of Interoceptive Inference” by Fotopoulou and Tsakiris’, Neuropsychoanalysis, 19(1): 39–42: ‘The attachment relationship therefore serves as an indicator of the nature of the infant’s environment (Belsky, Steinberg, & Draper, 1991; Simpson & Belsky, 2016), indeed it appears to be a powerful communication mechanism which works at the level of gene expression as well as at the level of social cognition, as Meaney’s work (Meaney, 2010; Meaney & Szyf, 2005) and some epigenetic human studies have shown’ (p. 40).

39 E.g. Fonagy, P. and Sharp, C. (2008). ‘Treatment Outcome of Childhood Disorders: The Perspective of Social Cognition’, in C. Sharp, P. Fonagy, and I. Goodyer (eds), Social Cognition and Developmental Psychopathology, New York: Oxford University Press, pp. 411–470: ‘All behavioural-emotional problems of children are both a reflection of biological vulnerability and/or adversity and an effort to cope and adapt’ (p. 448).

40 Bateman, A. W. and Fonagy, P. (2016). Mentalization-Based Treatment for Personality Disorders: A Practical Guide (2nd edn), Oxford: Oxford University Press, p. 62: ‘The basic emotions [curiosity, fear, anger, lust, caring, sadness and joy] are survival and adaptive experiences, which are essential and inform action and reaction. Many patients will try to avoid them but in doing so will diminish their ability to appraise situations and themselves’ (p. 307).

41 Ferraro, D. ([2011] 2014). ‘The Other, Clinical and Empirical: A Review of Fonagy et al. On Affect Regulation, Mentalisation, and the Development of the Self’. Accessed at: https://melbournelacanian.wordpress.com/2014/11/24/the-other-clinical-and-empirical-a-review-of-fonagy-et-al-on-affect-regulation-mentalisation-and-the-development-of-the-self/.

42 Drinkwater, M. (2008). ‘Portraits of a Divided Self’. Guardian, 11 September. Accessed at: https://www.theguardian.com/society/2008/sep/11/mentalhealth: ‘Peter Fonagy, a psychology professor at University College London, said: “Consciousness is one of the remaining mysteries of neuroscience. We may be able to explain the features of [dissociative identity disorder] DID, but without understanding consciousness we cannot entirely explain how DID occurs.” Far from being considered pathological, Fonagy suggested that DID can be seen as a sign of resilience; a strategy for coping with extremely traumatic events. “The mind is capable of separating and we put traumatising events into little ‘boxes’. This can be useful as it can help people dissociate from these experiences”, he said.’ See also Fonagy, P. (2018). ‘Preface’, in Werner Bohleber (ed.), Destructiveness, Intersubjectivity and Trauma: The Identity Crisis of Modern Psychoanalysis, London: Routledge, pp. xi–xiv.

43 Fonagy, P. and Bateman, A.W. (2016). ‘Adversity, Attachment, and Mentalizing’. Comprehensive Psychiatry, 64: 59–66: ‘The experience of trauma, and understanding and overcoming adversity, could also be considered to entail the enhancement of mentalizing in the process of overcoming trauma’ (p. 60).

44 Greenberg, D. M., Baron-Cohen, S., Rosenberg, N., Fonagy, P., and Rentfrow, P. J. (2018). ‘Elevated Empathy in Adults Following Childhood Trauma’. PLoS One, 13(10): e0203886.

45 Fonagy, P., Luyten, P., Allison, E., and Campbell, C. (2017). ‘What We have Changed our Minds About: Part 1. Borderline Personality Disorder as a Limitation of Resilience’. Borderline Personality Disorder and Emotion Dysregulation, 4(1): 11.

47 Bateman, A. W. and Fonagy, P. (2016). Mentalization-Based Treatment for Personality Disorders: A Practical Guide (2nd edn), Oxford: Oxford University Press: ‘particularly in these individuals, subsequent adversity or trauma could further disrupt mentalising, in part as an adaptive manoeuvre on the part of the individual to limit exposure to a brutalising psychosocial environment, and in part because the high levels of arousal generated by attachment hyperactivation and disorganised attachment strategies serve to disrupt less well-practiced and less robustly established higher cognitive capabilities’ (p. 47).

48 Bevington, D., Fuggle, P., Cracknell, L., and Fonagy, P. (2017). Adaptive Mentalisation-Based Integrative Treatment: A Guide for Teams to Develop Systems of Care, Oxford: Oxford University Press: ‘In the face of very real social conflict, then a mentalising stance in relation to social relationships is likely only to leave the individual more vulnerable; for example, it may ultimately render the individual less able to make use of aggressive and violent social strategies, which might seem absolutely necessary for self-protection. In such an environment, more reactive, impulsive forms of mentalising and non-mentalising modes of behaviour, which may most conspicuously take the form of aggression or risky sexual behaviour, might be at one level an adaptive response’ (p. 63).

49 Bateman, A. W. and Fonagy, P. (2016). Mentalization-Based Treatment for Personality Disorders: A Practical Guide (2nd edn), Oxford: Oxford University Press, p. 440.

50 E.g. Sroufe, L. A., Egeland, B., Carlson, E. A., and Collins, W. A. (2005). The Development of the Person. New York: Guilford Press; Englund, M. M., Kuo, S. I. C., Puig, J., and Collins, W. A. (2011). ‘Early Roots of Adult Competence: The Significance of Close Relationships from Infancy to Early Adulthood’. International Journal of Behavioral Development, 35(6): 490–496. Short-run benefits of insecure attachment under adversity, in this case maternal depression, have been documented by Milan, S., Snow, S., and Belay, S. (2009). ‘Depressive Symptoms in Mothers and Children’. Developmental Psychology, 45(4): 1019–1033. However, the researchers found that there were longer-term advantages to security under conditions of adversity.

51 Ein-Dor, T., Mikulincer, M., Doron, G., and Shaver, P. R. (2010). ‘The Attachment Paradox: How can so Many of Us (the Insecure Ones) Have no Adaptive Advantages?’. Perspectives on Psychological Science, 5(2): 123–141.

52 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: 9.

53 Giddens, A. (1991). Modernity and Self-Identity, Cambridge: Polity Press; Sievers, B. (2003). ‘Against all Reason: Trusting in Trust’. Organizational & Social Dynamics 3(1): 19–39.

54 Gorman, E. H. and Sandefur, R. L. (2011). ‘“Golden Age,” Quiescence, and Revival: How the Sociology of Professions Became the Study of Knowledge-Based Work’. Work & Occupations, 38(3): 275–302. Already in 2000, Fonagy was arguing that ‘Therapy is based on trust, and it is hard for it to be effective without it.’ Interview with Fonagy, cited in Rice, M. (2000). ‘Therapy is the New Religion’. Guardian, 13 August. Accessed at: https://www.theguardian.com/theobserver/2000/aug/13/life1.lifemagazine11.

55 Chiesa, M. and Fonagy, P. (2014). ‘Reflective Function as a Mediator between Childhood Adversity, Personality Disorder and Symptom Distress’. Personality and Mental Health, 8(1): 52–66.

56 Sperber, D., Clément, F., Heintz, C., Mascaro, O., Mercier, H., Origgi, G., and Wilson, D. (2010). ‘Epistemic Vigilance’. Mind & Language, 25(4): 359–393.

57 A different formulation is offered by Asen, E. and Fonagy, P. (2020). ‘Mentalization in Systemic Therapy and Its Empirical Evidence’, in Matthias Ochs, Maria Borcsa, and Jochen Schweitzer (eds), Systemic Research in Individual, Couple, and Family Therapy and Counseling, New York: Springer, pp. 207–221. There epistemic trust is conceptualized as treating information as possessing ‘authenticity and personal relevance’ (p. 216), rather than as ‘dependable, generalisable or relevant’. This emphasis on authenticity is interesting, and materially distinct from the idea of dependable and generalizable information. Yet another definition is offered in 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: ‘Epistemic vigilance refers to the ability to be justifiably suspicious about socially transmitted information, in a manner that helps protect one against potentially deceitful or erroneous information.’

58 Fonagy, P. and Allison, E. (2018). ‘The Origin of Human Life: A Psychoanalytic Developmental Perspective’, European Psychoanalytical Federation, 31st Annual Conference, Warsaw, 24 March.

59 Gergely, G. (2008). ‘Learning “about” versus Learning “from” Other Minds: Natural Pedagogy and its Implications’, in P. Carruthers, S. Laurence, and S. Stich (eds), The Innate Mind. Foundations and the Future, Volume 3, New York: Oxford University Press, pp. 170–198. Speber became a colleague of Gergely at the Department of Cognitive Science at the Central European University in 2010, following retirement from CNRS, Paris. Gergely’s interpretation of Sperber and colleagues’ concept of epistemic vigilance pre-dates this, however.

60 Sperber, D., Clément, F., Heintz, C., Mascaro, O., Mercier, H., Origgi, G., and Wilson, D. (2010). ‘Epistemic Vigilance’. Mind & Language, 25(4): 359–393, p. 359. See also Sperber, D. and Wilson, D. (2015). ‘Beyond Speaker’s Meaning’. Croatian Journal of Philosophy, 15(44), 117–149.

61 Fonagy, P. and Allison, E. (2014). ‘The Role of Mentalizing and Epistemic Trust in the Therapeutic Relationship’. Psychotherapy, 51(3): 372–380, p. 374.

62 Hutchins, E. (1995). Cognition in the Wild, Cambridge, MA: MIT Press; Hutchins, E. (2006). ‘The Distributed Cognition Perspective on Human Interaction’, in Nicholas Enfield and Stephen C. Levinson (eds), Roots of Human Sociality: Culture, Cognition and Human Interaction, Oxford: Berg, pp. 375–398; Hutchins, E. (2010). ‘Cognitive Ecology’. Topics in Cognitive Science, 2: 705–715. For a distributed cognition perspective on playful exploration, for example, see Parker-Rees, R. (2014). ‘Playfulness and the Co-Construction of Identity in the First Years’, in L. Brooker, M. Blaise, and S. Edwards, (eds), The SAGE Handbook of Play and Learning in Early Childhood, London: SAGE, pp. 366–377.

63 Sperber, D., Clément, F., Heintz, C., Mascaro, O., Mercier, H., Origgi, G., and Wilson, D. (2010). ‘Epistemic Vigilance’. Mind & Language, 25(4): 359–393, p. 383.

64 The term ‘discrimination’ is from Bion, W. R. (1976). ‘Interview with Anthony G. Banet’, in Tavistock Seminars, London: Karnac Books, pp. 97–114: ‘To what extent is one to allow any idea to come in? One feels that there’s a need for a sort of discriminating screen’ (p. 113). A related distinction between ‘default’ and blind’ trust is made in Granqvist, P. (2020). Attachment in Religion and Spirituality: A Wider View, New York: Guilford Press.

65 In significant regards, the thinking of Fonagy and colleagues about epistemic mistrust echoes earlier thinking about the origins of prejudice in the breakdown of trust within attachment relationships as a prompt for ‘suspicion’ in internal working models. See Fonagy, P. and Higgitt, A. (2007). ‘The Development of Prejudice: An Attachment Theory Hypothesis Explaining its Ubiquity’, in H. Parens, A. Mahfouz, S. W. Twemlow, and D. Scharff (eds), The Future of Prejudice: Psychoanalysis and the Prevention of Prejudice, Lanham, MD: Rowman & Littlefield, pp. 63–79.

66 Allison, E. and Fonagy, P. (2016). ‘When is Truth Relevant?’. Psychoanalytic Quarterly, 85(2): 275–303, p. 285.

67 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: 9, p. 27.

68 Fonagy, P. and Allison, E. (2014). ‘The Role of Mentalizing and Epistemic Trust in the Therapeutic Relationship’. Psychotherapy, 51(3): 372–380, p. 373.

69 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.

70 There remains some instability in the account of epistemic trust as to whether it refers to felt experience or a causal mechanism or both. For Fonagy’s emphasis on the importance of this distinction for psychoanalytic theory, see Chapter 6.

71 Fonagy, P. and Allison, E. (2014). ‘The Role of Mentalizing and Epistemic Trust in the Therapeutic Relationship’. Psychotherapy, 51(3): 372–380, . p. 374.

72 The extent to which imagination is implicated in procedural information has been debated by philosophers. Ultimately, it hinges on the definition of ‘imagination’ and the model of perceptual experience. Matherne, S. (2016). ‘Kant’s Theory of the Imagination’, in A. Kind (ed.), Routledge Handbook of Philosophy of Imagination, London: Routledge, pp. 55–68.

73 Winnicott, D. W. (1971). Playing and Reality, London: Routledge. See the discussion of Winnicott in Fonagy, P. and Target, M. (1996). ‘Playing with Reality: I. Theory of Mind and the Normal Development of Psychic Reality’. The International Journal of Psychoanalysis, 77: 217–233.

74 Bateman, A. W. and Fonagy, P. (2016). Mentalization-Based Treatment for Personality Disorders: A Practical Guide (2nd edn), Oxford: Oxford University Press, p. 20.

75 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. See also Gergely, G. and Csibra, G. (2006). ‘Sylvia’s Recipe: The Role of Imitation and Pedagogy in the Transmission of Cultural Knowledge’, in Nicholas Enfield and Stephen C. Levinson (eds), Roots of Human Sociality: Culture, Cognition and Human Interaction, Oxford: Berg, pp. 229–255; Gergely, G. and Unoka, Z. (2008). ‘Attachment, Affect-Regulation, and Mentalization: The Developmental Origins of the Representational Self’, in C. Sharp, P. Fonagy, and I. M. Goodyer (eds), Social Cognition and Developmental Psychopathology, New York: Oxford University Press, pp. 305–342.

76 Csibra, G. and Gergely, G. (2011). ‘Natural Pedagogy as Evolutionary Adaptation’. Philosophical Transactions of the Royal Society B: Biological Sciences, 366(1567): 1149–1157; Egyed, K., Király, I., and Gergely, G. (2013). ‘Communicating Shared Knowledge in Infancy’. Psychological Science, 24(7): 1348–1353. Subsequent research has suggested that Csibra and Gergely overstated the necessary role of ostensive cues for joint attention and for learning: e.g. Gredebäck, G., Astor, K., and Fawcett, C. (2018). ‘Gaze Following is not Dependent on Ostensive Cues: A Critical Test of Natural Pedagogy’. Child Development, 89(6): 2091–2098, though the argument that ostensive cues may facilitate engagement with the mind of the other is unaffected by these qualifications. A further point to raise is that Csibra and Gergely ostensibly take the term ‘ostensive cues’ from their reading of Sperber and Wilson. However, comparison of the two groups of theorists suggests that the former have reified the notion, because for the latter it is only a designation for a set of a broader class of strategies through which attentional processes are directed within the dance of conversation. It would also appear that, for Sperber and Wilson, ostensive cues do not necessarily imply recognition of the other’s intentional agency: it depends on what is meant by the latter concept, which can be rather opaque (see Chapter 9).

77 Bateman, A. and Fonagy, P. (2019). ‘Introduction’, in Anthony Bateman and Peter Fonagy (eds), Handbook of Mentalising in Mental Health Practice (2nd edn), Washington, DC: American Psychiatric Association, pp. 3–20, p. 16.

78 Fonagy and colleagues have recently published interesting reflections on how ostensive cues may be achieved through video working platforms such as Zoom—for instance, using joint attention through the screen share function. 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; Fonagy, P., Campbell, C., Truscott, A., and Fuggle, P. (2020). ‘Mentalising Remotely: The AFNCCF’s Adaptations to the Coronavirus Crisis’. Child and Adolescent Mental Health, 25(3): 178-179. Attempts to deliver therapy remotely to young people in their homes also suggested to the researchers that another ostensive cue professionals may use is behaviour indicating that a message was for the recipient specifically rather than others in earshot.

79 Fonagy, P. (2020). ‘Kindness can Work Wonders. Especially for the Vulnerable’. Guardian, 17 May. Accessed at: https://www.theguardian.com/society/2020/may/17/kindness-can-work-wonders-especially-for-the-vulnerable.

80 Fonagy, P. and Campbell, C. (2017). ‘What Touch can Communicate: Commentary on “Mentalizing Homeostasis: The Social Origins of Interoceptive Inference” by Fotopoulou and Tsakiris’, Neuropsychoanalysis, 19(1): 39–42.

81 Anna Freud has been given as an example by Allison and Campbell: ‘in accounts of her interactions with children, one gets the strong sense that Anna Freud was tremendously gifted at eliciting epistemic trust in the children she worked with, and that she achieved this through her genuine recognition of and interest in the minds of the children she worked with’. Allison, E. and Campbell, C. (2019). Transforming Child Mental Health: Principles of Sustainable Development, London: Anna Freud Centre. See descriptions in Kennedy, H. (1983). ‘Anna Freud 1895–1982’. Psychoanalytic Quarterly, 52(4): 501–506; Solnit, A. J. and Newman, L. M. (1984). ‘Anna Freud: The Child Expert’, The Psychoanalytic Study of the Child, 39: 45–63; Blomfield, O. H. (1991). ‘Anna Freud: Creativity, Compassion, Discipline’. International Review of Psycho-Analysis, 18: 37–52.

82 Carlson, E. A. (1998). ‘A Prospective Longitudinal Study of Attachment Disorganization/Disorientation’. Child Development, 69(4): 1107–1128; Wang, F., Cox, M. J., Mills‐Koonce, R., and Snyder, P. (2015). ‘Parental behaviors and beliefs, child temperament, and attachment disorganization’. Family Relations, 64(2): 191–204.

83 Fonagy, P. and Allison, E. (2014). ‘The Role of Mentalizing and Epistemic Trust in the Therapeutic Relationship’. Psychotherapy, 51(3): 372–380: ‘While attachment may be a key mechanism for mediating epistemic trust, it is secondary to an underlying biological process preserved by evolution. In other words, secure attachment is unlikely to be necessary for generating epistemic trust but it may be sufficient to do so, and, further, it is the most pervasive mechanism in early childhood because it is a highly evolutionarily effective indicator of trustworthiness’ (p. 375).

84 For discussion of this claim, see Granqvist, P. (2020). ‘Attachment, Culture, and Gene-Culture Co-Evolution: Expanding the Evolutionary Toolbox of Attachment Theory’. Attachment & Human Development, 23(1): 90-113. Alongside praise for the advance represented by attention to epistemic trust in evolutionary perspective, Granqvist ultimately criticizes Fonagy and colleagues for neglecting the contribution of epistemic trust to cultural transmission and evolution. On theory of mind as an exaptation of the phylogenetically earlier attachment system, see also Chisholm, J. S. (2003). ‘Uncertainty, Contingency, and Attachment: A Life History Theory of Theory of Mind’, in K. Sterelny and J. Fitness (eds), From Mating to Mentality: Evaluating Evolutionary Psychology, New York: Psychology Press, pp. 125–153.

85 Ainsworth characterized ambivalent/resistant patterns as representing distrust in the caregiver. Ainsworth, M. D. S., Bell, S. M., and Stayton, D. J. (1974). ‘Infant-Mother Attachment and Social Development: “Socialisation” as a Product of Reciprocal Responsiveness to Signals’, in J. M. Richards (ed.), The Integration of a Child into a Social World, Cambridge: Cambridge University Press, pp. 9–135: ‘It may be viewed as advantageous for an infant whose mother seems to him to move unpredictably and inconsistently (and whom he has not been able to learn to trust) to monitor her movements with exceptional alertness and to evince disturbance whenever she moves off’ (p. 125).

86 E.g. Verhees, M. W., Ceulemans, E., Bakermans-Kranenburg, M. J., Van IJzendoorn, M. H., De Winter, S., and Bosmans, G. (2017). ‘The Effects of Cognitive Bias Modification Training and Oxytocin Administration on Trust in Maternal Support: Study Protocol for a Randomized Controlled Trial’. Trials, 18(1): 326: ‘Bowlby proposed that whether children are able to develop trust in parental support depends for a significant part on children’s experiences of care in response to distress during interactions with their parents. Through repeated interactions with sensitive and responsive parenting, children develop trust in the parent and become securely attached’ (p. 2).

87 Fonagy, P. and Allison, E. (2014). ‘The Role of Mentalizing and Epistemic Trust in the Therapeutic Relationship’. Psychotherapy, 51(3): 372–380, p. 374.

88 Corriveau, K. H., Harris, P. L., Meins, E., Fernyhough, C., Arnott, B., Elliott, L., … and De Rosnay, M. (2009). ‘Young Children’s Trust in their Mother’s Claims: Longitudinal Links with Attachment Security in Infancy’. Child Development, 80(3): 750–761.

89 Ainsworth, M. (1969). ‘Sensitivity vs. Insensitivity to the Baby’s Signals Scale’. Accessed at: http://www.psychology.sunysb.edu/attachment/measures/content/ainsworth_scales.html.

90 Chiesa and Fonagy had found that abuse and neglect made independent contributions to personality disorder. Chiesa, M. and Fonagy, P. (2014). ‘Reflective Function as a Mediator between Childhood Adversity, Personality Disorder and Symptom Distress’. Personality and Mental Health, 8(1): 52–66.

91 Fonagy, P. (2017). ‘The Concept of Epistemic Trust: Can it Help Integrate Attachment Theory?’. Paper presented to the San Francisco Centre for Psychoanalysis, Haskell Norman Prize Event, Monday, 9 January 2017. . The diagrams also featured in the keynote address, ‘Epistemic Trust and Attachment’, to the International Attachment Conference, London, 1 July 2017. There seems to be some differences among Fonagy’s collaborators on how attachment theory is conceived and the extent to which early attachment is regarded as critical for later development. For instance, we see claims such as Luyten, P., Malcorps, S., Fonagy, P., and Ensink, K. (2019). ‘Mentalising and Trauma’, in Anthony Bateman and Peter Fonagy (eds), Handbook of Mentalising in Mental Health Practice (2nd edn), Washington, DC: American Psychiatric Association, pp. 79–102: ‘the development of the stress system … is largely determined by the quality of attachment relationships’ (p. 82).

92 This argument appeared already in Fonagy, P. and Target, M. (2000). ‘Playing with Reality: III. The Persistence of Dual Psychic Reality in Borderline Patients’. The International Journal of Psychoanalysis, 81(5): 853–873: ‘The patient lacks an authentic, organic self-image built around internalised representations of self-states. The absence or weakness of such a self-image leaves the child, and later the adult, with affect that remains unlabelled and confusing, uncontained (Bion, 1962) … This will create a desperation for meaning, and a willingness to take in reflections from the other that do not map on to anything within the child’s own experience’ (p. 865).

93 A certain hesitancy regarding the directionality of relations between mentalizing and epistemic trust can also be seen in Luyten, P., Malcorps, S., Fonagy, P., and Ensink, K. (2019). ‘Assessment of Mentalising’, in Anthony Bateman and Peter Fonagy (eds), Handbook of Mentalising in Mental Health Practice (2nd edn), Washington, DC: American Psychiatric Association, pp. 37–62: ‘The assessor should determine to what extent mentalising impairments are linked to problems with epistemic trust’ (p. 50).

94 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. 72. Fonagy, P. and Allison, E. (2018). ‘The Origin of Human Life: A Psychoanalytic Developmental Perspective’, European Psychoanalytical Federation, 31st Annual Conference, Warsaw, 24 March: ‘Epistemic hypervigilance and epistemic credulity, can co-exist and can oscillate rapidly. This can be confusing for the clinician whose understanding appears to be valued and idealized one day only to be scotomised, spat out, denigrated and spurned the next.’

95 Duschinsky, R., Collver, J., and Carel, H. (2019). ‘“Trust Comes from a Sense of Feeling One’s Self Understood by Another Mind”: An Interview with Peter Fonagy’. Psychoanalytic Psychology, 36(3), 224–227, p. 225. See also Luyten, P., Campbell, C., and Fonagy, P. (2019). ‘Borderline Personality Disorder, Complex Trauma, and Problems with Self and Identity: A Social‐Communicative Approach’. Journal of Personality, 88(1): 88-105: ‘We understand such excessive epistemic credulity as being triggered by a hyperactive or unmoored imagination generating a personal narrative that is too diffuse to provide an accurate sense of differential awareness of others’ capacity to perceive oneself. Excessive credulity results, as all personal narratives feel as if they “fit” sufficiently for trust to be generated, making the person vulnerable to exploitation.’ The most detailed case study of epistemic credulity offered by Fonagy and colleagues, though without using the term, is the case of Mr A. in 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.

96 Fonagy, P. (2013). ‘Mentalizing and the Attachment Process’. Paper delivered at the Menninger Clinic, 22 March. Accessed at: http://www.creatingconnections.nl/assets/files/2013.04.18%20Peter%20Fonagy%20Creating%20Connections.pdf.

97 E.g. Luyten, P., Van Assche, L., Kadriu, F., Krans, J., Claes, L., and Fonagy, P. (2017). ‘Other Disorders often Associated with Psychological Trauma’, in C. Dalenberg, S. Gold, and J. Cook (eds), APA Handbook of Trauma Psychology. Volume 1: Foundations in Knowledge, Washington, DC: American Psychological Association, pp. 243–280, p. 246.

98 A ‘preliminary’ distinction between state-like and trait-like problems with epistemic trust has been put forward in 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.

99 Luyten, P., Campbell, C., and Fonagy, P. (2019). ‘Borderline Personality Disorder, Complex Trauma, and Problems with Self and Identity: A Social‐Communicative Approach’. Journal of Personality, 88(1): 88–105.

100 Fonagy, P. and Allison, E. (2015). ‘A Scientific Theory of Homosexuality for Psychoanalysis’, in A. Lemma and P. E. Lynch (eds), Sexualities: Contemporary Psychoanalytic Perspectives, London: Routledge, pp. 125–137, p. 134. See also Lorenzini, N., Campbell, C., and Fonagy, P. (2019). ‘Mentalisation and its Role in Processing Trauma’, in Bernd Huppertz (ed.), Approaches to Psychic Trauma: Theory and Practice, Lanham, MD: Rowman & Littlefield, pp. 403–422.

101 Ezrati, O. (2014). ‘Freud Off: Giving New Meaning to Psychoanalysis’, Haaretz, 5 April. Accessed at: https://www.haaretz.com/life/books/.premium-giving-new-meaning-to-psychoanalysis-1.5243899.

102 See Fonagy, P., Target, M., Steele, M., Steele, H., Leigh, T., Levinson, A. et al. (1997). ‘Morality, Disruptive Behavior, Borderline Personality Disorder, Crime, and their Relationships to Security of Attachment’, in L. Atkinson and K. J. Zucker (eds), Attachment and Psychopathology, New York: Guilford Press, pp. 223–274, pp. 239–41. The extent to which adolescence is sufficiently different in this regard from adulthood as to require a different model of the relationship between mentalization and mental health has been raised by Battersby, S. (2018). Is Mentalising Ability Associated with Mental Health Difficulties in Adolescents? A Systematic Review: Understanding the Construct of Mentalising in Adolescence and its Association with Mental Health, Structural Equation Model. Unpublished doctorate in clinical psychology, University of Edinburgh.

103 Van IJzendoorn, M. H. and Bakermans-Kranenburg, M. J. (2018). ‘Disorganized Attachment, Entropic Care, Epistemic Trust, and “Unusual” Beliefs’. Paper presented at the Epistemic Petrification Conference, 2–3 July, Bristol University.

104 The Epistemic Trust Instrument was developed by Jennifer O’Connell in 2014. O’Connell, J. (2014). Can We Develop an Adult Assessment Tool for Measuring Epistemic Trust? Unpublished master’s thesis, London: University College London. However, this has yet to be used in any published study. Work is underway by the Heidelberg group to construct a measure of epistemic trust: Schröder-Pfeifer, P., Talia, A., Volkert, J., and Taubner, S. (2018). ‘Developing an Assessment of Epistemic Trust: A Research Protocol’. Research in Psychotherapy: Psychopathology, Process and Outcome, 21(3).

105 Bo, S., Bateman, A., and Kongerslev, M. T. (2019). ‘Mentalization-Based Group Therapy for Adolescents with Avoidant Personality Disorder: Adaptations and Findings from a Practice-Based Pilot Evaluation’. Journal of Infant, Child, and Adolescent Psychotherapy, 18(3): 249-262. The ‘trust in parents’ sub-scale of the self-report Inventory of Parent and Peer Attachment, taken at baseline, has also been found not to moderate reduction of symptoms in the course of therapy in a sample of adolescents. Such findings highlight the need for caution in assuming that existing measures of trust can be treated as measuring epistemic trust. Orme, W., Bowersox, L., Vanwoerden, S., Fonagy, P., and Sharp, C. (2019). ‘The Relation between Epistemic Trust and Borderline Pathology in an Adolescent Inpatient Sample’. Borderline Personality Disorder and Emotion Dysregulation, 6(1): 1–9. See also Gullone, E. and Robinson, K. (2005). ‘The Inventory of Parent and Peer Attachment—Revised (IPPA-R) for Children: A Psychometric Investigation’. Clinical Psychology & Psychotherapy, 12(1): 67–79.

106 For discussions of the heterogenous meanings of the concept of trust, see e.g. McKnight, D. H. and Chervany, N. L. (2001). ‘Trust and Distrust Definitions: One Bite at a Time’, in Rino Falcone, Munindar Singh, and Yao-Hua Tan (eds), Trust in Cyber-Societies, New York: Springer, pp. 27–54; Dietz, G. and Den Hartog, D. N. (2003). ‘Measuring Trust Inside Organisations’. Personnel Review, 32(5): 557–588; Joni, S.-N. (2004). ‘The Geography of Trust’. Harvard Business Review, 82(3): 7–11; Fink, M., Harms, R., and Möllering, G. (2010). ‘Introduction: A Strategy for Overcoming the Definitional Struggle’. International Journal of Entrepreneurship and Innovation, 11(2): 101–105; Driver, M. (2015). ‘How Trust Functions in the Context of Identity Work’. Human Relations, 68(6): 899–923; Duncan, G. (2019). The Problem of Political Trust: A Conceptual Reformulation, London: Routledge.

107 In fact, Bateman and Fonagy appear to want to encompass all objects of trust. They described epistemic mistrust as entailing ‘destruction of trust in social knowledge of all kinds’. Bateman, A. W. and Fonagy, P. (2016). Mentalization-Based Treatment for Personality Disorders: A Practical Guide (2nd edn), Oxford: Oxford University Press, p. 27. Likewise, Luyten and colleagues give the example of a study in which ‘participants aged 12–17 who had experienced physical abuse were less able than their peers, who had no history of maltreatment, to correctly learn which stimuli were likely to result in reward, even after repeated feedback’. In Luyten’s characterization of the study, epistemic distrust is generic—it entails difficulty learning from any feedback on the properties of any stimuli. Luyten, P., Campbell, C., and Fonagy, P. (2019). ‘Borderline Personality Disorder, Complex Trauma, and Problems with Self and Identity: A Social‐Communicative Approach’. Journal of Personality, 88(1): 88–105.

108 The most important aspect of trust for Fonagy and colleagues appears to be less our conscious feelings about the other’s claims, and more whether we are willing to work this knowledge into our plans. This distinction is discussed by Holton, R. (1994). ‘Deciding to Trust, Coming to Believe’, Australasian Journal of Philosophy, 72: 63–76.

109 Hendriks, F., Kienhues, D., and Bromme, R. (2015). ‘Measuring Laypeople’s Trust in Experts in a Digital Age: The Muenster Epistemic Trustworthiness Inventory (METI)’. PLoS One, 10(10): e0139309. Interestingly, the factor of ‘perceptions of the other as reliable/unreliable’ was excluded because it loaded on more than one dimension.

110 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. 297.

111 Lyon, F., Möllering, G., and Saunders, M. N. K. (eds) (2012). Handbook of Research Methods on Trust. Cheltenham: Edward Elgar, p. 1.

112 Only the epistemic vigilance that contributes to mentalizing difficulties, not the epistemic mistrust that results from mentalizing difficulties, figures in the diagram presented by Taubner, Gablonsky, and Fonagy in their account of the emergence of conduct disorder. This is despite the fact that it could readily be imagined that epistemic mistrust resulting from mentalizing difficulties could contribute to conduct problems. See Taubner, S., Gablonski, T.-C., and Fonagy, P. (2019). ‘Conduct Disorder’, in Anthony Bateman and Peter Fonagy (eds), Handbook of Mentalising in Mental Health Practice (2nd edn), Washington, DC: American Psychiatric Association, pp. 301–321, Figure 18.1.

113 Mollering, G. (2006). Trust: Reason, Routine, Reflexivity. New York: Elsevier; Barbalet, J. (2019). ‘The Experience of Trust: Its Content and Basis’, in Masamichi Sasaki (ed.), Trust in Contemporary Society, New York: Brill, pp. 11–30.

114 See e.g. Bion, W. ([1973] 1990). Brazilian Lectures, London: Karnac Books: ‘Some patients cannot … listen to what they themselves say. They have no respect for what they already know, so that their experience and knowledge are of no use to them. The question is not simply one of the relationship of the patient to the analyst, but the relationship of the patient with himself which may be so bad that he cannot even make use of what he already knows’ (p. 50).

115 Asen, E., Campbell, C., and Fonagy, P. (2019). ‘Social Systems: Beyond the Microcosm of the Individual and Family’, in Anthony Bateman and Peter Fonagy (eds), Handbook of Mentalising in Mental Health Practice (2nd edn), Washington, DC: American Psychiatric Association, pp. 229–243, p. 241.

116 Jurist, E. (2018). ‘Preface’, in Minding Emotions: Cultivating Mentalisation in Psychotherapy, New York: Guilford Press.

117 Luyten, P., Lemma, A., and Target, M. (2019). ‘Depression’, in Anthony Bateman and Peter Fonagy (eds), Handbook of Mentalising in Mental Health Practice (2nd edn), Washington, DC: American Psychiatric Association, pp. 387–401, p. 398.

118 The latter option—mentalization facilitating discriminated trust/vigilance—seems to be implied, without being fully spelt out, in Fonagy, P., Campbell, C., and Allison, E. (2019). ‘Therapeutic Models Mentalising’, in Anthony Bateman and Peter Fonagy (eds), Handbook of Mentalising in Mental Health Practice (2nd edn), Washington, DC: American Psychiatric Association, pp. 169–180.

119 Ellman, S. J. (2007). ‘Analytic Trust and Transference: Love, Healing Ruptures and Facilitating Repairs’. Psychoanalytic Inquiry, 27(3): 246–263.

120 One way of considering epistemic trust and mentalization is in terms of constituent elements and their interrelation. Given the different definitions and uses Fonagy and colleagues have made of the two concepts, it would appear that there are eight permutations of epistemic trust to consider, against eight relevant permutations of mentalization:

  • Epistemic trust of the self’s feelings.

  • Epistemic trust of the self’s thoughts.

  • Epistemic trust of others’ feelings.

  • Epistemic trust of others’ thoughts.

  • Epistemic mistrust of the self’s feelings.

  • Epistemic mistrust of the self’s thoughts.

  • Epistemic mistrust of others’ feelings.

  • Epistemic mistrust of others’ thoughts.

  • Conceiving of the other’s feelings.

  • Conceiving of the other’s thoughts.

  • Conceiving of the self’s feelings.

  • Conceiving of the self’s thoughts.

  • Reconsidering the other’s feelings.

  • Reconsidering the other’s thoughts.

  • Reconsidering the self’s feelings.

  • Reconsidering the self’s thoughts.

121 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. 328.

122 Fonagy, P. and Higgitt, A. (2007). ‘The Development of Prejudice: An Attachment Theory Hypothesis Explaining its Ubiquity’, in H. Parens, A. Mahfouz, S. W. Twemlow, and D. Scharff (eds), The Future of Prejudice: Psychoanalysis and the Prevention of Prejudice, Lanham, MD: Rowman & Littlefield, pp. 63–79), Figure 4.1.

123 E.g. Knox, J. (2016). ‘Epistemic Mistrust: A Crucial Aspect of Mentalization in People with a History of Abuse?’. British Journal of Psychotherapy, 32(2): 226–236: ‘Epistemic trust is essentially a process of trust in the other person’s mind and therefore depends on mentalizing’ (p. 226).

124 E.g. Bion, W. (1967). Second Thoughts, London: Karnac Books, p. 101: ‘lack of progress in any direction must be attributed in part to the destruction of a capacity for curiosity and the consequent inability to learn’.

125 Bateman, A. W. and Fonagy, P. (2016). Mentalization-Based Treatment for Personality Disorders: A Practical Guide (2nd edn), Oxford: Oxford University Press, p. 33. See also Allison, E. and Fonagy, P. (2016). ‘When is Truth Relevant?’. Psychoanalytic Quarterly, 85(2): 275–303: ‘Mentalizing in this context is not an end in itself. Mentalizing is the catalyst that activates the effective ingredient of therapy: learning from experience’ (p. 294).

126 Fonagy, P. (2016). ‘Epistemic trust and attachment: A fresh look at therapeutic processes in personality disorder’. AMBIT Conference, 2 July. Accessed at: https://www.altrecht.nl/wp-content/uploads/2016/11/Ambit-Conference-Fonagy.pdf.

127 Jablensky, A. (2016). ‘Psychiatric Classifications: Validity and Utility’. World Psychiatry, 15(1): 26–31.

128 Bowlby, J. (1940). Personality and Mental Illness, London: Kegan Paul, p. 5.

129 Bowlby, J. (1983). ‘Darwin: Psychiatry and Developmental Psychology’, Contribution to a Symposium on Darwin and Psychology held at the conference of the British Psychological Society, December. London: Wellcome Collection, PP/BOW/F.3/132. See Duschinsky, R. (2020). Cornerstones of Attachment Research, Oxford: Oxford University Press, Chapter 2.

130 Bowlby, J. ([1970] 1979). ‘Self-Reliance and Some Conditions that Promote it’, in The Making and Breaking of Affectional Bonds, London: Routledge, pp. 124–149, p. 126.

131 Freud, A. (1970). ‘The Symptomatology of Childhood: A Preliminary Attempt at Classification’. The Psychoanalytic Study of the Child, 25(1):19–41, pp. 19–20.

132 Ibid. 21.

133 Fonagy, P. and Target, M. (1994). ‘Who is Helped by Child Psychoanalysis? A Sample Study of Disruptive Children, from the Anna Freud Centre Retrospective Investigation’. Bulletin of the Anna Freud Centre, 17(4): 291–315, p. 296.

134 Ibid. Recently, the AMBIT Adolescent Integrative Measure has been developed as an abbreviated version of the Hampstead Child Adaptation Measure. Fuggle, P., Bevington, D., Cracknell, L., Hanley, J., Hare, S., Lincoln, J., … and Zlotowitz, S. (2015). ‘The Adolescent Mentalization-Based Integrative Treatment (AMBIT) Approach to Outcome Evaluation and Manualization: Adopting a Learning Organization Approach’. Clinical Child Psychology and Psychiatry, 20(3): 419–435.

135 Shedler, J., Beck, A. T., Fonagy, P., Gabbard, G. O., Kernberg, O., Michels, R., and Westen, D. (2011). ‘Response to Skodol Letter’. American Journal of Psychiatry, 168(1): 97–98, p. 97. See also Maldonado‐Durán, M., Helmig, L., Moody, C., Fonagy, P., Fulz, J., Lartigue, T., … Glinka, J. (2003). ‘The Zero‐to‐Three Diagnostic Classification in an Infant Mental Health Clinic: Its Usefulness and Challenges’. Infant Mental Health Journal, 24(4): 378–397.

136 See also Hutsebaut, J., Debbané, M., and Sharp, C. (2020). ‘Designing a range of mentalizing interventions for young people using a clinical staging approach to borderline pathology’. Borderline Personality Disorder and Emotion Dysregulation, 7(1): 1–10.

137 Kirby, T. (2019). ‘Peter Fonagy—Battling the Enemy of Loneliness’. The Lancet Psychiatry, 6(12): 987.

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

139 Fonagy, P. (2015). ‘Peter Fonagy and the Undermining of Old Ideas on Personality Disorder’. Accessed at: https://www.escap.eu/resources/resource-centre-disorders/undermining-of-old-ideas-on-personality-disorder.

140 Interview with Fonagy in Steidinger, S. (2018). ‘Trans-Actions: An Exploration of Gender Dysphoria’. Accessed at: https://vimeo.com/285555219: ‘Enhancing an individual’s capacity to represent their subjective experience in a more coherent and more fluid way would aid the resolution of gender identity issues in a child or young adult above other approaches that are more “radical”, shall we say. So I would strongly wish to support an initiative that aimed to enhance the capacity of an individual to conceptualize themselves and their subjective experience of their gender in a more complex way as a first stop in their journey to resolving gender dysphoria.’

141 See also Fonagy, P. and Allison, E. (2015). ‘A Scientific Theory of Homosexuality for Psychoanalysis’, in A. Lemma and P. E. Lynch (eds), Sexualities: Contemporary Psychoanalytic Perspectives, London: Routledge, pp. 125–137: ‘Drescher (2002) refers to this habit of thought as “binary thinking” … From a neuroscientific perspective, we might describe it as a result of failure of the mentalising capacity’ (p. 125).

142 Bateman, A. W. and Fonagy, P. (2016). Mentalization-Based Treatment for Personality Disorders: A Practical Guide (2nd edn), Oxford: Oxford University Press: ‘Even if we do not believe in categorical diagnostic systems, someone else in the mental health service is likely to have given the patient a diagnosis’ (p. 149).

143 Chiesa, M., Cirasola, A., Williams, R., Nassisi, V., and Fonagy, P. (2017). ‘Categorical and Dimensional Approaches in the Evaluation of the Relationship between Attachment and Personality Disorders: An Empirical Study’, Attachment & Human Development, 19(2): 151–169. See also St Clair, M. C., Neufeld, S., Jones, P. B., Fonagy, P., Bullmore, E. T., Dolan, R. J., … and Goodyer, I. M. (2017). ‘Characterising the Latent Structure and Organisation of Self-Reported Thoughts, Feelings and Behaviours in Adolescents and Young Adults’. PloS One, 12(4): e0175381; Bergen, L., and Grimes, T. (1999). ‘The Reification of Normalcy’, Journal of Health Communication, 4(3): 211–226: ‘Because normalcy, when it is clinically considered, actually implies a wide variation in mental states (DSM-IV American Psychiatric Association, 1994), we believe that one must assume that “normal” children can vary greatly in their reaction and susceptibility to violent television programming. The reason many ill or prodromal children (that is, children who do not manifest DSM-IV classifiable behaviors but who show early symptoms of potentially classifiable mental disorders) may be viewed as normal is because most lay people, perhaps communication researchers among them, consider normalcy to be characterized by the ability to operate within the world satisfactorily enough to “get along,” to interact passably or better with one’s friends, family, and colleagues (P. Fonagy, personal communication, January 8, 1998). This definition of normalcy does not preclude the presence of psychopathology, in either a clinical or prodromal stage (P. Fonagy, personal communication, January 8, 1998). Indeed, most of the people in the population who are ill are undiagnosed and are generally treated as being normal’ (p. 216). Most recently, see Wendt, L. P., Wright, A. G., Pilkonis, P. A., Nolte, T., Fonagy, P., Montague, P. R., … and Zimmermann, J. (2019). ‘The latent structure of interpersonal problems: Validity of dimensional, categorical and hybrid models’. Journal of Abnormal Psychology, 128(8): 823–839. Fonagy has also repeatedly drawn attention to lifetime prevalence data for mental illness as suggesting deficits in the assumption of health in the absence of a diagnosis: Schaefer, J. D., Caspi, A., Belsky, D. W., Harrington, H., Houts, R., Horwood, L. J., … and Moffitt, T. E. (2017). ‘Enduring Mental Health: Prevalence and Prediction’. Journal of Abnormal Psychology, 126(2): 212–224: ‘Only 17% of repeatedly assessed Study members managed to reach midlife (age 38) without experiencing the psychiatric symptoms and resulting functional impairment necessary to meet criteria for the diagnosis of a mental disorder’ (p. 220).

144 Goldberg, S., Rousmaniere, T., Miller, S., Whipple, J., Nielsen, S., Hoyt, W., and Wampold, B. (2016). ‘Do Psychotherapists Improve With Time and Experience? A Longitudinal Analysis of Outcomes in a Clinical Setting’. Journal of Counseling Psychology, 63(1): 1–11.

145 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. Fonagy has drawn comparison with ideas from Zen Buddhism here, citing, ‘In the beginner’s mind there are many possibilities, in the expert’s mind there are few.’ Suzuki, S. (1970). Zen Mind, Beginner’s Mind, New York: Weatherall, p. 21. 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: https://www.borderlinepersonalitydisorder.org/gap-call-in-series-podcast/.

147 Bevington, D. (2020). ‘Towards a Learning Stance In Teams: Developing a Community of Practice to Capture and Disseminate what Works for Whom’, in Ilana Crome and Richard Williams (eds), Substance Misuse and Young People: Critical Issues, London: Routledge, pp. 450–466: ‘each team inherits all the generic AMBIT content as the starting point for its local manual. Teams are trained to engage in regular brief discussions about elements of practice, described in more detail below, that end in their producing key bullet points that can be owned by the team (manualising is thus a collective activity, never a single worker’s ideas). Teams either overwrite their own local versions of pages inherited from AMBIT or can add new material alongside this inherited content. Thus, local manuals are co-constructions that draw together centrally curated evidence-based practice and locally generated practice-based evidence … The process of a team manualising its local practice is seen as analogous to the group mentalising itself because it inquires and reflects on why and how we do things in this way’ (p. 459).

148 Caspi, A., Houts, R. M., Belsky, D. W., Goldman-Mellor, S. J., Harrington, H., Israel, S., … and Moffitt, T. E. (2014). ‘The P Factor: One General Psychopathology Factor in the Structure of Psychiatric Disorders?’. Clinical Psychological Science, 2(2): 119–137. See also Laceulle, O. M., Vollebergh, W. A., and Ormel, J. (2015). ‘The Structure of Psychopathology in Adolescence: Replication of a General Psychopathology Factor in the TRAILS Study’. Clinical Psychological Science, 3(6): 850–860.

149 Jones, E. (1946). ‘A Valedictory Address’. The International Journal of Psychoanalysis, 27: 7–12.

150 Fonagy, P. and Campbell, C. (2015). ‘Bad Blood Revisited: Attachment and Psychoanalysis, 2015’. British Journal of Psychotherapy, 31(2): 229–250, p. 243

151 Fonagy, P. (2016). ‘The Role of Attachment, Epistemic Trust and Resilience in Personality Disorder: A Trans-Theoretical Reformulation’. DMM News, 26 September. Accessed at: http://www.iasa-dmm.org/images/uploads/DMM%20%2322%20Sept%2016%20English.pdf: The most recent formulation is also more cautious: ‘many manifestations of mental disorder may be underpinned by an inability to benefit from social communication due to epistemic mistrust’. 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. 71.

152 Fonagy, P., Luyten, P., Allison, E., and Campbell, C. (2017). ‘What We have Changed our Minds About: Part 1. Borderline Personality Disorder as a Limitation of Resilience’. Borderline Personality Disorder and Emotion Dysregulation, 4(1): 11.

153 Bateman, A. W. and Fonagy, P. (2016). Mentalization-Based Treatment for Personality Disorders: A Practical Guide (2nd edn), Oxford: Oxford University Press: ‘We understand impulsivity in terms of imbalance among the poles of mentalising: it involves a heavy emphasis on the automatic pole of the automatic-controlled dimension … What reflection there is will most likely be disconnected from reality [pretend mode] … I will entangle myself and my clinician in extended analyses, but offer little compelling evidence for any of my assertions. I urgently seek validation for my view, but even when it is forthcoming, it is meaningless because I am simultaneously aware that I made up my explanation; thus confirming or elaborating it only increases my sense of emptiness and meaninglessness’ (p. 44); ‘Impulsivity can also result from the teleological stance … for such a patient, what does seem to help are physical actions that make him/her feel real’ (p. 45).

154 An exception was symptoms of narcissistic personality disorder, which turned out to be quite distinct: Sharp, C., Wright, A. G., Fowler, J. C., Frueh, B. C., Allen, J. G., Oldham, J., and Clark, L. A. (2015). ‘The Structure of Personality Pathology: Both General (‘g’) and Specific (‘s’) Factors?’. Journal of Abnormal Psychology, 124(2): 387–398: ‘Narcissistic PD criteria’s average loading on the general factor was rather weak (r = .31). What implications does this have for how we conceptualize, for example, narcissistic criteria and traits?’ (p. 396). However, the distinct qualities of symptoms of narcissistic personality disorder have not been found to replicate: Wright, A. G., Hopwood, C., Skodol, A., and Morey, L. (2016). ‘Longitudinal Validation of General and Specific Structural Features of Personality Pathology’. Journal of Abnormal Psychology, 125(8): 1120–1134.

155 These findings were subsequently replicated in Ibid. Here the relationship between BPD and the p-factor was even stronger than in the study by Sharp and colleagues.

156 Sharp, C., Wright, A. G., Fowler, J. C., Frueh, B. C., Allen, J. G., Oldham, J., and Clark, L. A. (2015). ‘The Structure of Personality Pathology: Both General (‘g’) and Specific (‘s’) Factors?’. Journal of Abnormal Psychology, 124(2): 387–398, p. 394.

157 Kernberg, O. F. (1984). Severe Personality Disorders: Psychotherapeutic Strategies. New Haven, CT: Yale University Press.

158 Bateman, A. W. and Fonagy, P. (2016). Mentalization-Based Treatment for Personality Disorders: A Practical Guide (2nd edn), Oxford: Oxford University Press: ‘treatment of BPD is required if comorbid disorder is to be treated; however, treating the comorbid disorder does not improve BPD’ (p. 82). See also 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.

159 Patalay, P., Fonagy, P., Deighton, J., Belsky, J., Vostanis, P., and Wolpert, M. (2015). ‘A General Psychopathology Factor in Early Adolescence’. The British Journal of Psychiatry, 207(1): 15–22.

160 St Clair, M. C., Neufeld, S., Jones, P. B., Fonagy, P., Bullmore, E. T., Dolan, R. J., … and Goodyer, I. M. (2017). ‘Characterising the Latent Structure and Organisation of Self-Reported Thoughts, Feelings and Behaviours in Adolescents and Young Adults’. PloS One, 12(4): e0175381.

161 St Clair and colleagues identified a lower-level factor of long-term, stable negative mood, which seemed to have different properties from more state-like components of depression, the latter loading exclusively on the general p-factor. St Clair, M. C., Neufeld, S., Jones, P. B., Fonagy, P., Bullmore, E. T., Dolan, R. J., … and Goodyer, I. M. (2017). ‘Characterising the Latent Structure and Organisation of Self-Reported Thoughts, Feelings and Behaviours in Adolescents and Young Adults’. PloS One, 12(4): e0175381.

162 Luyten, P., Van Assche, L., Kadriu, F., Krans, J., Claes, L., and Fonagy, P. (2017). ‘Other Disorders often Associated with Psychological Trauma’, in C. Dalenberg, S. Gold, and J. Cook (eds), APA Handbook of Trauma Psychology. Volume 1: Foundations in Knowledge, Washington, DC: American Psychological Association, pp. 243–280, p. 245.

163 The potential for mentalization to contribute to broaden-and-build cycles, but not the absence of non-mentalizing, has been discussed recently by Bo, S., Bateman, A., and Kongerslev, M. T. (2019). ‘Mentalization-Based Group Therapy for Adolescents with Avoidant Personality Disorder: Adaptations and Findings from a Practice-Based Pilot Evaluation’. Journal of Infant, Child, and Adolescent Psychotherapy, 18(3): 249–262.

164 See also Stochl, J., Khandaker, G. M., Lewis, G., Perez, J., Goodyer, I. M., Zammit, S., … and Jones, P. B. (2015). ‘Mood, anxiety and psychotic phenomena measure a common psychopathological factor’. Psychological Medicine, 45(7): 1483–1493.

165 See also van Nierop, M., Viechtbauer, W., Gunther, N., Van Zelst, C., De Graaf, R., Ten Have, M., … and OUtcome of Psychosis (GROUP) investigators (2015). ‘Childhood Trauma is Associated with a Specific Admixture of Affective, Anxiety, and Psychosis Symptoms Cutting across Traditional Diagnostic Boundaries’. Psychological Medicine, 45(6): 1277–1288.

166 Debbané, M., Salaminios, G., Luyten, P., Badoud, D., Armando, M., Solida Tozzi, A., Fonagy, P. and Brent, B. K. (2016). ‘Attachment, Neurobiology, and Mentalizing along the Psychosis Continuum’. Frontiers in Human Neuroscience, 10: 406; Debbané, M. and Toffel, E. (2019). ‘Mentalizing Through the Early Stages of the Psychosis Continuum’, in João G. Pereira, Jorge de Almeida Gonçalves, and Valeria Bizzari (eds), The Neurobiology-Psychotherapy-Pharmacology Intervention Triangle, Wilminton, DE: Vernon Press, pp. 141–160; Debbané, M. and Bateman, A. (2019). ‘Psychosis’, in Anthony Bateman and Peter Fonagy (eds), Handbook of Mentalising in Mental Health Practice (2nd edn), Washington, DC: American Psychiatric Association, pp. 417–429.

167 On suicidality as reflecting the p-factor, see Prabhu, G., Dolan, R. J., Bullmore, E. T., Fonagy, P., Stochl, J., Jones, P. B., and NSPN Consortium. (2020). ‘How do the Prevalence and Relative Risk of Non-Suicidal Self-Injury and Suicidal Thoughts Vary across the Population Distribution of Common Mental Distress (the P-Factor)? Observational Analyses Replicated in Two Independent UK Cohorts of Young People’. BMJ Open, 10: e032494. Fonagy and Bateman have contrasted BPD and ASPD on several grounds. These contrasts include that ‘people with ASPD are more likely to demonstrate over-control of their emotional states within well-structured, schematic attachment relationships, rather than the under-control in chaotic attachment relationships that are more commonly found in people with BPD. Their dyscontrol may be limited to volatile anger: Bateman, A. W. and Fonagy, P. (2016). Mentalization-Based Treatment for Personality Disorders: A Practical Guide (2nd edn), Oxford: Oxford University Press, pp. 377–378.

168 Polek, E., Jones, P. B., Fearon, P., Brodbeck, J., Moutoussis, M., Dolan, R., … and NSPN Consortium. (2018). ‘Personality Dimensions Emerging during Adolescence and Young Adulthood are Underpinned by a Single Latent Trait Indexing Impairment in Social Functioning’. BMC Psychiatry, 18(1): 23, online additional file, Table 1.

169 Other strongly loading items were: ‘Your emotions are shallow and fake’; ‘You use or con other people to get what you want’; ‘I seem very cold and uncaring to others’; ‘People sometimes find me aloof and distant’; ‘Some people find me a bit vague and elusive during a conversation’; ‘I feel very uncomfortable in social situations involving unfamiliar people’.

170 Gibbon, L., Nolte, T. and Fonagy, P. (2017). ‘Modelling Axis I and Personality Disorder Symptomatology and its Associations with Childhood Trauma and Mentalising’, in L. Gibbon, An Interpretive Thematic Analysis of the P-Factor Literature and an Empirical Investigation of the Relationship between the P-Factor and Childhood Trauma and Reflective Function Childhood Trauma and Reflective Function. Unpublished DClinPsy thesis, London: University College London. Accessed at: http://discovery.ucl.ac.uk/1574529/1/Thesis_final_volume1_Gibbon.pdf.

171 A previous study did include other relevant symptoms, but not symptoms of BPD specifically, in the analysis: Hoertel, N., Franco, S., Wall, M. M., Oquendo, M. A., Kerridge, B. T., Limosin, F., and Blanco, C. (2015). ‘Mental Disorders and Risk of Suicide Attempt: A National Prospective Study’. Molecular Psychiatry, 20(6): 718.

172 Baldwin, J. R., Reuben, A., Newbury, J. B., Danese, A. (2019). ‘Agreement between Prospective and Retrospective Measures of Childhood Maltreatment: A Systematic Review and Meta-Analysis’. JAMA Psychiatry, 76(6): 584-593 .

173 McCrory, E. J., Gerin, M. I., and Viding, E. (2017). ‘Annual Research Review: Childhood Maltreatment, Latent Vulnerability and the Shift to Preventative Psychiatry—The Contribution Of Functional Brain Imaging’. Journal of Child Psychology and Psychiatry, 58(4): 338–357.

174 However, the internalizing and thought disorder dimensions remained barely significant correlates of RFQu even after taking into account the p-factor (rs = .20 and .17 respectively).

175 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.

176 As measured with the Personality Assessment Inventory—Borderline Features. The ‘identity problems’ sub-scale comprises the following questions: ‘My attitude about myself changes a lot’; ‘Sometimes I feel terribly empty inside’; ‘I worry a lot about other people leaving me’; ‘I often wonder what I should do with my life’; ‘I can’t handle separation from those close to me very well’; and ‘I don’t get bored very easily’ (reversed). Criticism of the psychometric qualities of the RFQc scale has been offered by Spitzer and colleagues. For instance they observe that ‘Although RFQ_C is supposed to represent certainty about mental states, all items are geared towards a state of uncertainty with respect to their semantic content (e.g., item 1: “I find the thoughts of others confusing”, or item 3: “When I get angry, I say things without really knowing why I am saying them”) and are, ultimately, reversely scored. Thus, the certainty scale is based entirely on the denial of uncertainty’. Müller, S., Wendt, L. P., Spitzer, C., Masuhr, O., Back, S. N., and Zimmermann, J. (2020). A critical evaluation of the reflective functioning questionnaire. https://psyarxiv.com/5rhme/. Despite such problems with the scale, for recent recent evidence that parental RFQc at the start of family MBT can serve as a predictor of treatment outcome, see Jewell, T., Herle, M., Serpell, L., Eivors, A., Simic, M., Fonagy, P., and Eisler, I. (2020). ‘Attachment and mentalization as predictors of outcome in family therapy for adolescent anorexia nervosa’. PsyArXiv pre-print, https://psyarxiv.com/fgytk/

177 Euler, S., Nolte, T., Constantinou, M., Griem, J., Montague, P. R., Fonagy, P., and Personality and Mood Disorders Research Network (2019). ‘Interpersonal Problems in Borderline Personality Disorder: Associations with Mentalizing, Emotion Regulation, and Impulsiveness’. Journal of Personality Disorders, Early View; Handeland, T. B., Kristiansen, V. R., Lau, B., Håkansson, U., and Øie, M. G. (2019). ‘High Degree of Uncertain Reflective Functioning in Mothers with Substance Use Disorder’. Addictive Behaviors Reports, 10: 100193. However, rather than a problem with the concept of certainty about other minds as contrary to mentalizing, this more likely reflects a methodological limitation: such certainties come in to play under conditions of high arousal, not as much when completing pen-and-paper questionnaires. Luyten, P., Malcorps, S., Fonagy, P., and Ensink, K. (2019). ‘Assessment of Mentalising’, in Anthony Bateman and Peter Fonagy (eds), Handbook of Mentalising in Mental Health Practice (2nd edn), Washington, DC: American Psychiatric Association, pp. 37–62, p. 57.

178 Bateman, A. W. and Fonagy, P. (2016). Mentalization-Based Treatment for Personality Disorders: A Practical Guide (2nd edn), Oxford: Oxford University Press: ‘from a research perspective the precise nature of the relationship between depression and BPD requires further elucidation … there is uncertainty about the extent to which depression and BPD should be considered as part of the same spectrum’ (p. 81).

179 Rosenberg, C. E. (2002). ‘The Tyranny of Diagnosis: Specific Entities and Individual Experience’. The Milbank Quarterly, 80(2): 237–260; Greco, M. (2016). ‘What is the DSM? Diagnostic Manual, Cultural Icon, Political Battleground: An Overview with Suggestions for a Critical Research Agenda’. Psychology & Sexuality, 7(1): 6–22; Schnittker, J. (2017). The Diagnostic System: Why the Classification of Psychiatric Disorders Is Necessary, Difficult, and Never Settled, New York: Columbia University Press; Armstrong, D. (2019). ‘Diagnosis: From Classification to Prediction’. Social Science & Medicine, 237: 112444.

180 Cf. Borsboom, D. (2017). ‘A Network Theory of Mental Disorders’. World Psychiatry, 16(1): 5–13.

181 Higgitt, A. (2000). ‘Suicide Reduction: Policy Context’. International Review of Psychiatry, 12(1): 15–20.

182 Rutter, M. (1987). ‘Psychosocial Resilience and Protective Mechanisms’. American Journal of Orthopsychiatry, 57(3): 316–331.

183 See also McCrory, E. J. and Viding, E. (2015). ‘The Theory of Latent Vulnerability: Reconceptualizing the Link between Childhood Maltreatment and Psychiatric Disorder’. Development and Psychopathology, 27(2): 493–505; Caspi, A., Houts, R. M., Belsky, D. W., Harrington, H., Hogan, S., Ramrakha, S., … Moffitt, T. E. (2016). ‘Childhood forecasting of a small segment of the population with large economic burden’. Nature Human Behaviour, 1: 5; Most recently, the ‘Adverse Childhood Experiences’ measure has served partly as a poster-child and partly as a placeholder for acknowledgement of the multiplier effect of risk in the course of development. See Steptoe, A., Marteau, T., Fonagy, P., and Abel, K. (2019). ‘ACEs: Evidence, Gaps, Evaluation and Future Priorities’. Social Policy and Society, 18(3): 415–424.

184 Fonagy, P. and Higgitt, A. (2000). ‘Early Influences on Development and Social Inequalities: An Attachment Theory Perspective’, in A. R. Tarlov and R. F. S. Peter (eds), The Society and Population Health Reader, Volume 2: A State and Community Perspective, New York: New Press, pp. 104–130, p. 104.

185 Fonagy, P. and Higgitt, A. (2007). ‘The Early Social and Emotional Determinants of Inequalities in Health’, in G. Baruch, P. Fonagy, and D. Robins (eds), Reaching the Hard to Reach: Evidence-Based Funding Priorities for Intervention and Research, Chichester, UK: John Wiley & Sons, pp. 3–34, pp. 6–7.

186 E.g. Steptoe, A., Marteau, T., Fonagy, P., and Abel, K. (2019). ‘ACEs: Evidence, Gaps, Evaluation and Future Priorities’. Social Policy and Society, 18(3): 415–424.

187 Fonagy, P. (2016). ‘The Role of Attachment, Epistemic Trust and Resilience in Personality Disorder: A Trans-Theoretical Reformulation’. DMM News, 26 September. Accessed at: http://www.iasa-dmm.org/images/uploads/DMM%20%2322%20Sept%2016%20English.pdf.

188 The idea that sustained states of arousal may contribute to somatic symptoms without the involvement of epistemic mistrust is implied in Luyten, P., van Houdenhove, B., Lemma, A., Target, M., and Fonagy, P. (2012). ‘A Mentalization-Based Approach to the Understanding and Treatment of Functional Somatic Disorders’. Psychoanalytic Psychotherapy, 26(2): 121–140, though no doubt epistemic mistrust may have a powerful role in sustaining these symptoms. Somatic symptoms load strongly with the p-factor in most factor analytic studies of symptoms.

189 E.g. Winnicott, D. W. (1955). ‘The Depressive Position in Normal Emotional Development’. British Journal of Medical Psychology, 28(2–3): 89–100.

190 E.g. Kubzansky, L. D., Bordelois, P., Jun, H. J., Roberts, A. L., Cerda, M., Bluestone, N., and Koenen, K. C. (2014). ‘The Weight of Traumatic Stress: A Prospective Study of Posttraumatic Stress Disorder Symptoms and Weight Status in Women’. JAMA Psychiatry, 71(1): 44–51.

191 Duschinsky, R., Reisz, S., and Messina, S. (2019). ‘“Pulling the World in and Pushing it Away”: Participating Bodies and the Concept of Coping’. Medical Humanities, 45(2): 124–130.

192 Skårderud, F. and Fonagy, P. (2012). ‘Eating Disorders’, in A. W. Bateman and P. Fonagy (eds), Handbook of Mentalizing in Mental Health Practice, Washington, DC: American Psychiatric Publishing, pp. 347–384, p. 352.

193 Blatt, S. J. and Luyten, P. (2009). ‘Depression as an Evolutionarily Conserved Mechanism to Terminate Separation Distress: Only Part of the Biopsychosocial Story?’. Neuropsychoanalysis, 11(1): 52–61, pp. 54–55.

194 St Clair, M. C., Neufeld, S., Jones, P. B., Fonagy, P., Bullmore, E. T., Dolan, R. J., … and Goodyer, I. M. (2017). ‘Characterising the Latent Structure and Organisation of Self-Reported Thoughts, Feelings and Behaviours in Adolescents and Young Adults’. PloS One, 12(4): e0175381.

195 Bateman, A. W. and Fonagy, P. (2016). Mentalization-Based Treatment for Personality Disorders: A Practical Guide (2nd edn), Oxford: Oxford University Press, p. 41.

196 Luyten, P., Van Assche, L., Kadriu, F., Krans, J., Claes, L., and Fonagy, P. (2017). ‘Other Disorders often Associated with Psychological Trauma’, in C. Dalenberg, S. Gold, and J. Cook (eds), APA Handbook of Trauma Psychology. Volume 1: Foundations in Knowledge, Washington, DC: American Psychological Association, pp. 261–262.

197 Patalay, P., Fonagy, P., Deighton, J., Belsky, J., Vostanis, P., and Wolpert, M. (2015). ‘A General Psychopathology Factor in Early Adolescence’. The British Journal of Psychiatry, 207(1): 15–22: ‘Sleep disturbance emerges as a poor indicator of internalising problems and might be better conceived of as a generic indicator of vulnerability to psychiatric disorder’ (p. 19).

198 Fonagy, P. and Target, M. (2003). ‘Being Mindful of Minds: A Homage to the Contributions of a Child-Analytic Genius’. The Psychoanalytic Study of the Child, 58(1): 307–321: ‘Depression entails an over involvement with and concretisation of mood related ideation’ (p. 319); Luyten, P., Lemma, A., and Target, M. (2019). ‘Depression’, in Anthony Bateman and Peter Fonagy (eds), Handbook of Mentalising in Mental Health Practice (2nd edn), Washington, DC: American Psychiatric Association, pp. 387–401.

199 Bateman and Fonagy describe how ‘absence of arousal prevents the development of attachment-based affect, which is the area of sensitivity in interpersonal interactions for people with BPD. Treatment becomes cognitively organized and the patient is detached from relational process. Pretend mode is often associated with absence of affect and may become persistent. Working on areas of interpersonal sensitivity that lead to loss of mentalising becomes impossible because interpersonal meaning is absent.’ Bateman, A. W. and Fonagy, P. (2016). Mentalization-Based Treatment for Personality Disorders: A Practical Guide (2nd edn), Oxford: Oxford University Press, p. 212.

200 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, p. 407.

201 Fischer-Kern, M., Fonagy, P., Kapusta, N. D., Luyten, P., Boss, S., Naderer, A., … and Leithner, K. (2013). ‘Mentalizing in Female Inpatients with Major Depressive Disorder’. Journal of Nervous and Mental Disease, 201(3): 202–207.

202 See also Ensink, K., Bégin, M., Normandin, L., and Fonagy, P. (2016). ‘Maternal and Child Reflective Functioning in the Context of Child Sexual Abuse: Pathways to Depression and Externalising Difficulties’. European Journal of Psychotraumatology, 7(1): 30611. Recent work has suggested qualifications to the Fischer-Kern study, in showing that distinct forms of depression may have different trajectories and correlates. Rost, F., Luyten, P., and Fonagy, P. (2018). ‘The Anaclitic–Introjective Depression Assessment: Development and Preliminary Validity of an Observer‐Rated Measure. Clinical Psychology & Psychotherapy, 25(2): 195–209.

203 Bateman, A. W. and Fonagy, P. (2016). Mentalization-Based Treatment for Personality Disorders: A Practical Guide (2nd edn), Oxford: Oxford University Press, pp. 335–336.

204 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, p. 400. See also Arefjord, N., Morken, K., and Lossius, K. (2019). ‘Comorbid Substance Use Disorder and Personality Disorder’, in Anthony Bateman and Peter Fonagy (eds), Handbook of Mentalising in Mental Health Practice (2nd edn), Washington, DC: American Psychiatric Association, pp. 403–416.

205 St Clair, M. C., Neufeld, S., Jones, P. B., Fonagy, P., Bullmore, E. T., Dolan, R. J., … and Goodyer, I. M. (2017). ‘Characterising the Latent Structure and Organisation of Self-Reported Thoughts, Feelings and Behaviours in Adolescents and Young Adults’. PloS One, 12(4): e0175381. The strength and direction of this association may well be influenced by cultural factors, though see also Handeland, T. B., Kristiansen, V. R., Lau, B., Håkansson, U., and Øie, M. G. (2019). ‘High Degree of Uncertain Reflective Functioning in Mothers with Substance Use Disorder’. Addictive Behaviors Reports, 10: 100193.

206 Skårderud, F. and Fonagy, P. (2012). ‘Eating Disorders’, in A. W. Bateman and P. Fonagy (eds), Handbook of Mentalizing in Mental Health Practice, Washington, DC: American Psychiatric Publishing, pp. 347–384, p. 356.

207 Fonagy, P. and Campbell, C. (2015). ‘Bad Blood Revisited: Attachment and Psychoanalysis, 2015’. British Journal of Psychotherapy, 31(2): 229–250, p. 243.