Show Summary Details
Page of

(p. 282) Conclusions 

(p. 282) Conclusions
(p. 282) Conclusions

Robbie Duschinsky

and Sarah Foster

Page of

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

date: 06 December 2021

Constructive appraisal

Throughout this book, we have attempted to offer reflections that have been friendly and constructive, looking to understand the underlying coherence of the theory as well as potential limitations. We acknowledge that it can be hard for a paradigm in psychological theory and practice to metabolize criticisms, because there is much invested in the current state of affairs, including ‘secondary gain’ from the fuzzy use of concepts and their lack of operationalization. As discussed in the Introduction, Sandler identified that there can be resistance to efforts to appraise or refine concepts that seem to work well enough for pragmatic purposes, and help hold a coalition of interested parties together, even if their use has had associated disadvantages:

The fact that they work well may in turn lead to an undue resistance to the progressive integration and modification of our concepts, so necessary for scientific development. This resistance can partly be overcome by the cultivation of a critical attitude towards our ideas, by discussion with colleagues, and by honest reading of the literature, but unless we are directly confronted with contradictions in our thinking, resistance to change due to secondary gains may prove too great.1

An aspect of this book that we hope may help circumvent some of this resistance is that we have tried, whenever possible, to be constructive and specific with suggestions to accompany appraisal. A number of such reflections from across the book can be particularly highlighted, without intending to be exhaustive. In Chapter 3, we expressed concern about confusion in the use of the concept of disorganized attachment by Fonagy and colleagues. We proposed that processes such as i) fear of attachment figures, ii) dissociation, iii) conflicting feelings, iv) difficulties modulating feelings, and v) the absence of help-seeking in times of need should be addressed more explicitly, given that they have at times been lost or outright mistaken for one another when disorganized attachment has been invoked.2 This may help resolve the ambivalence Fonagy has shown (see Chapter 7) regarding whether or not to include disorganized attachment in his developmental model of mentalizing and epistemic trust, and help retain attention to elements that otherwise fluctuate in and out of view, such as dissociation.

In Chapter 4, we identified both strengths and weaknesses in the ‘four poles’ model of mentalizing (inner/outer, automatic/controlled, affective/cognitive, self/other), which has (p. 283) been prominent in the theory and clinical guidance of Fonagy and colleagues since 2009. We offered a number of friendly suggestions of matters that might be addressed in future. First, the account of affective versus cognitive mentalizing is troubled by a lack of clarity regarding whether this refers to the object of mentalizing (e.g. mentalizing one’s thoughts or feelings) or the process of mentalizing (e.g. mentalizing using one’s thoughts or feelings). Second, the conditions that facilitate high-quality automatic mentalizing are left unaddressed. Third, ‘automatic’ mentalizing and ‘outer’ mentalizing are generally treated as non-mentalizing, leading to conceptual and terminological muddle between mentalizing and non-mentalizing. The treatment of automatic mentalizing as non-mentalizing also leads to the excluded middle of slow, semi-intentional ‘feeling out’, as in therapy. Fourth, even after a decade, no attempt has been made to operationalize the model. Perhaps in part this reflects difficulties with the constructs, but equally in turn the lack of attempts at operationalization has insulated the four poles model from scrutiny, empirical validation, or refinement.

The four poles model was introduced in 2009 to respond to commentators such as Jeremy Holmes, who had identified that the term ‘mentalizing’ was serving as something of a catch-all, impeding communication between and among researchers and clinicians, and hindering scientific discussion. In Chapter 4, we identified 28 different uses of the term ‘mentalizing’ over time by Fonagy and colleagues. In part, this represented the evolution of theory over time; in part, it represented a desire to encompass different processes within a single umbrella; in part, it expressed a degree of haziness in the theoretical conceptualization. Our reflections go beyond identification of these problems. On the basis of a comprehensive analysis of every usage of the term across the writings of Fonagy and colleagues, we synthesized a new definition of mentalizing:

The capacity to conceive of and make available for reconsideration the thoughts and feelings implicated in motivations and intentions in order to account for and explain the observable social behaviour and present and past perceptual experience of oneself and others.

This definition makes it easier to distinguish between full and specific forms of mentalizing, articulate the relationship between mentalizing and non-mentalizing, and specify the relationship between mentalization and mental states. It situates more exactly within one framework the dimensions that Fonagy and colleagues had described using the ‘four poles’, while resolving some of the difficulties with that account.

In Chapter 5, our new definition of mentalizing was used to address questions in the account of non-mentalizing offered by Fonagy and colleagues. The researchers had claimed that non-mentalizing represented a hijacking of mentalizing capacities, but without explaining how or why non-mentalizing deploys the very elements that comprise mentalizing. Our definition allowed us to address this question and specify more precisely what is being hijacked in non-mentalizing.

Pretend mode was characterized as using the capacity to conceive of thoughts and feelings implicated in motivations or intentions.

Hypermentalizing was characterized as a form of pretend mode using the capacity to conceive of thoughts and feelings implicated in motivations or intentions in order to account for and explain the observable social behaviour of others.

Psychic equivalence was characterized as using the capacity to account for and explain thoughts and feelings and observable social behaviour in terms of immediate experience. (p. 284)

Teleological mode was characterized as using the capacity to account for and explain observable social behaviour as implicated in the motivations or intentions of others and of oneself.

In articulating how non-mentalizing constitutes elements of mentalizing, this account helps specify when and how mentalization-based therapy (MBT) may contribute to harm to patients by entrenching non-mentalizing. An example, discussed in Chapter 8, is when MBT refines knowledge of the feelings of others, but where this is then used by the individual as a resource for relational aggression. A further advantage to our account of non-mentalising here is that it makes sense of an ambiguity in the work of Fonagy and colleagues: whether mentalizing is merely the absence of non-mentalizing, and vice versa. Our definition helps us resolve this question, because it highlights the centrality for mentalisation of the generative doubt that underpins inquisitive stance.

Chapter 6 drew together the threads of discussions of the ‘self’ across the writings of Fonagy and colleagues. Readers of the work of these researchers have generally been segmented, each audience having access to only a ‘part object’. Fonagy and colleagues have also used the concept in different ways across these contexts. While perhaps inevitable for the sake of communication, there has been a price to pay for the theory—for instance, in a lack of clarity about how ‘mental states’ occur, and whether mentalizing one’s own mental states is the same as or different from mentalizing others’ mental states. Bringing together discussions of the self from across the writings of Fonagy and colleagues helped clarify the extent to which, for these researchers, the ‘self’ is a confabulation, as well as the role of two distinct kinds of unconscious processes in producing the content of this confabulation. We also raised the question about the status of ‘externalization of the alien self’, which at times shuttles between a description of experience and an explanation of experience. Certain versions of the concept would appear to represent a candidate for another form of non-mentalizing, a question Fonagy and colleagues may wish to address. However, we also note that the concept of ‘externalization of the alien self’, so prominent in their work in the 2000s, appears to have dropped out of favour in recent years, without explanation and while formally remaining on the books as part of the paradigm.

Chapter 7 identified that Fonagy and colleagues have increasingly sought to emphasize that non-mentalizing is an adaptive response to adversity, drawing ideas from attachment theory. However, this claim has not been supported by explanation of what is meant by adaptation. Using an unpublished interrogation of the concept of adaptation by Ainsworth, one of the founders of attachment theory, we identified three different meanings to the term: species-level natural selection, individual responsiveness to the environment, and individual thriving. Natural selection may have facilitated the development of mentalizing and forms of non-mentalizing as a repertoire of responses available to humans. And both mentalizing and non-mentalizing may—or may not—reflect responsiveness to contextual cues. As such, non-mentalizing is theorized by Fonagy and colleagues to contribute to short-term positive outcomes, especially in harsh environments. In turn, mentalizing is theorized to contribute to long-term positive outcomes, especially—or perhaps solely—in non-harsh environments. The question of whether contextual adversity moderates the effect of mentalizing and non-mentalizing on outcomes is an important one, and should certainly be addressed in future research.

Chapter 7 also offers some distinctions we hope will be constructive regarding the concept of epistemic vigilance, which has seen special prominence in the work of Fonagy and colleagues in recent years. A significant problem faced by their account is ambiguity between two concepts of epistemic vigilance: one in which the communication of others is subject to (p. 285) scrutiny, and another in which the communication of others is outright distrusted. We recommended that the term ‘epistemic vigilance’ reverts back to the original usage of Sperber and colleagues, who introduced the concept, to mean a suite of cognitive mechanisms that help calibrate the credence assigned to the inferred experiences and claim others. This would be distinguished from ‘epistemic mistrust’ 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, or relevant to the self. Epistemic mistrust may be a transitory state; however, it can be stabilized by the non-mentalizing modes, and in turn contribute to non-mentalizing. The distinction between epistemic vigilance and mistrust is helpful in making sense of the claim of Fonagy and colleagues that epistemic vigilance is actually beneficial for mentalizing. Epistemic vigilance, but not epistemic mistrust, can support generative doubt and the potential for the reconsideration of the experiences of oneself and others, though the rigid and inflexible subjection of experiences to epistemic vigilance may contribute to hypermentalizing, rather than mentalization.

Chapter 9 offers reflections and suggestions relating to the account by Fonagy and colleagues of wider social systems extending beyond the individual. This is clearly an area of particular interest and theoretical development, and we hope that our distinctions can help facilitate future thinking and research in this area. One contribution was our efforts to excavate the concept of ‘collective mentalizing capacity’: the properties of the community that facilitate mentalization. In discussions of schools, family therapy, and group therapy, Fonagy and colleagues have offered partial acknowledgment of this process, and it is hoped that with attention drawn to collective mentalizing capacity the concept may see greater prominence, including further elaboration and ideally operationalization. The question of whether groups themselves can mentalize and non-mentalize, or only their individual members, has also been left in abeyance, and may be worth raising more focally in future because the stakes seem quite high. Relevant to this question, in Chapter 9, we identified an unacknowledged hinge in the use of the concept ‘culture’ by Fonagy and colleagues, and difficulties this was causing their account of social systems. In a first sense, they have been discussing ‘anthropological culture’, meaning a collective form of life that shapes the perceptions and intentions of its members, with particular priority given to shared myths, prohibitions, presuppositions, and taken-for-granted practices and artefacts. In a second sense, they have been discussing ‘cognitive informational culture’, the net information held by members of a group. This distinction is important because in focusing on cognitive information culture in their work on epistemic trust to date, the researchers have tended to neglect the role of anthropological culture, including the operation of oppression. It has also made it difficult for them to distinguish the social conditions that facilitate different forms of non-mentalizing, and that make individual responsibility feasible rather than a persecutory ideal.

Some remaining questions: mentalizing

One of the most important areas for future trials of MBT will be to examine whether improvements in mentalizing mediate the success of the modality, and how or whether improvement is moderated by social systems. Fonagy and colleagues have recently acknowledged this as a limitation in the current evidence base, and highlighted the importance of questions about working mechanisms.3 As such questions are pursued in future research, it would be good to (p. 286) see the potential for iatrogenesis from MBT explicitly explored, ideally based on a means of measuring forms of non-mentalizing.

In terms of theory, one of the most important areas for future development will be to acknowledge and address limitations with the four poles model. In particular, the elements that appear to have particular value, such as mentalizing self versus mentalizing others, should be operationalized and/or incorporated into existing measures. Ensink’s work adapting the reflective functioning scales has been a valuable step in this direction (see Chapter 4).4 The account of automatic mentalizing has been widely criticized by commentators, and might be revisited. In terms of future work, thinking about ‘poles’ of mentalizing might also address the difference between mentalizing a determinate mental state versus mentalizing inchoate or conflicted mental states. This might pick up on a difference relevant to clinical efforts to help patients enrich their mentalizing capacities. It may also help counter a tendency in the theorizing of Fonagy and colleagues to treat mental states as relatively determinate rather than conflicted or inchoate, though certainly they acknowledge that conflicted and inchoate mental states exist. In part as a result of this tendency, there are outstanding questions about the role of proto-thoughts and proto-feelings.5 It is not clear that facilitation of the capacity to conceive of the determinate mental states of others or oneself will have similar correlates to the capacity to conceive of the inchoate mental states of others or oneself. Clinical efforts to help the former capacity, when this is at the expense of the latter, may inadvertently contribute to teleological mode rather than mentalizing.

Another area where future work may be fruitful is in relation to anxiety. As we saw in Chapter 3, Fonagy and Target found that anxiety was a positive prognostic factor for children receiving therapy at the Anna Freud Centre. Yet there has been minimal subsequent attention to anxiety, despite the fact that it was a core concept for the Anna Freudian tradition of psychoanalysis inherited by Fonagy, such as in the work of Sandler.6 Even though dynamic interpersonal therapy is ostensibly an adaptation of MBT for people with depression or depression and anxiety, there is no sustained published discussion of how the treatment is presumed to address anxiety. Indeed, Fonagy appears to recommend cognitive-behavioural (p. 287) therapy rather than a mentalization-based approach for children with anxiety.7 One reason for the relative lack of concern with anxiety for Fonagy and colleagues and other scholars working on mentalizing is that it does not align with their assumption that mental states are intentional (see Chapter 4). This assumption works well enough for some feelings like fear and disgust. But anxiety—and some other affects—can be free-floating, corresponding to a general sense of threat rather than a specific object. In general terms, anxiety can be conceptualized as the perception of present or future threat (general or specific), plus uncertainty (especially when extensive or poorly modulated), plus difficulty modulating uncertainty, in the absence of perceived help. The theory offered by Fonagy and colleagues is relevant to each of these elements. To sketch some illustrative directions for thinking about mentalization and anxiety: as we saw in Chapter 4, mentalization helps modulate threat appraisal and may help sustain access to help from others, whereas non-mentalizing reduces the modulation of uncertainty. In turn, high levels of anxiety hinders mentalizing. By contrast, reflection of the opacity of one’s own or others’ minds, or reflection on one’s own or others’ destructive feelings, may contribute to both anxiety and successful mentalizing. Drawing on the reflections in Chapter 5, it may be supposed that the modes of non-mentalizing contribute to relatively distinct and identifiable forms of anxiety:

  • Pretend mode anxiety may be persecutory, because it is not tied down by actual states of affairs and the ordinary limitations of their ramifications.8

  • Psychic equivalence anxiety may centre around the sense of crushing reality without alternatives or openings.

  • Teleological mode anxiety may be seen as a sense of panicked urgency, in the desire to concretely resolve perceived problems and meet standards.

Further theory and research on the relationship between anxiety and mentalization may help contribute to further advancement of the paradigm, and to its clinical efficacy in planning and delivering treatment.

Some remaining questions: non-mentalizing

The most important issue facing the account of non-mentalizing, and perhaps the most significant limitation of the paradigm as a whole, is the lack of operationalization of pretend mode, psychic equivalence, and teleological mode. In our view, this has profoundly hindered the validation and advancement of these ideas over time. The attempt by another research group to operationalize the modes of non-mentalizing appears not to have been registered by Fonagy and colleagues.9 Whereas the modes of non-mentalizing have been central to the theorizing and advice on clinical technique offered by Fonagy, in fact this has all occurred without specific supporting evidence—and places the theory and clinical guidance subject (p. 288) to the criticisms Fonagy levelled at psychoanalytic theory in the 1990s. If we compare the account of non-mentalizing to the six changes to psychoanalytic practice recommended by Fonagy as the basis for scientific credibility (see Chapter 2), this raises the following concerns:

  • We do not actually know empirically that the three modes of non-mentalizing are negatively associated with mentalizing, or how strongly.

  • We do not know that the reduction in non-mentalizing is associated with reduced symptoms or greater quality of life.

  • We do not know whether MBT reduces all three modes of non-mentalizing. We do not know what components of MBT are effective for what forms of non-mentalizing.

A further set of questions for future work on non-mentalizing will be to consolidate an evolutionary perspective, attentive to mechanisms at individual and population levels. One way to do this would be to ask Tinbergen’s four questions of each of the modes of non-mentalizing:10 How did it develop in the individual? What causes it? What is its function? How did it evolve in the species? Fonagy and colleagues have certainly addressed function and development. The account of the evolutionary basis of the modes of non-mentalizing is currently a sketch and warrants further attention. Cross-species comparison may help specify the role of representational and non-representational processes more sharply, alongside the phylogenetic building blocks of the non-mentalizing modes. However, perhaps most importantly, the proximal cause of non-mentalizing would benefit from exploration, hypothesis generation and empirical scrutiny. For instance, subsequent inquiry may consider forms of pleasure/enjoyment attending the modes of non-mentalizing, and help them to become reinforcing. It may be that non-mentalizing can feel invigorating, vital, or a relief in certain localized ways that help prompt this state of mind.11 An obstacle that has repeatedly obscured effective examination of the cause of non-mentalizing has been the concept of ‘trauma’ as used by Fonagy and colleagues. Both recognition of the need for better specification and some of the ongoing problems with the concept of trauma in the work of Fonagy and colleagues can be seen in their recent chapter in the second edition of the Handbook of Mentalising in Mental Health Practice.12 These include: 1) an overencompassing (and unoperationalizable) definition of trauma as ‘experiences that go beyond the average expectable environment’;13 2) a characterization of the most important traumas as occurring, and healed, primarily by attachment relationships, without a specification of what an attachment relationship is; 3) the claim that ‘aversity becomes traumatic only when it is compounded by the individual’s sense that his or her mind is alone’,14 a proposal that certainly contains insight, but has aspects of (p. 289) a soundbite. It seems to assume that reframing of one’s experience in discussion with others is the sole means by which adversity does not become traumatic. This claim is difficult to evaluate given that the term ‘trauma’ is used so generally, and given that the account of ‘with one’s mind’ seems to slide about between metaphor, phenomenological description, and causal mechanism.

Questions remain also regarding the status of ‘alienation of the alien self’. Has the concept dropped out of the theory? Or is it in fact a fourth mode of non-mentalizing? At present, links between this concept and theoretical developments of the 2000s and 2010s (e.g. epistemic trust, the p-factor) remain poor. This is despite the strong potential for relevant connections—for instance, between the alien self and epistemic credulity. If it is decided that alienation of the alien self is to be rehabilitated as a live element of the theory, a question that could be valuable to pursue is whether there are forms of alien self that contribute positively to mentalizing under certain conditions. For instance, it might be considered that the ambitions for ourselves to which we aspire form a kind of ‘alien self’, which may contribute to rumination but which may also, or instead, contribute to efforts to achieve a life commensurate with our hopes and values.15 If the ‘alien self’ is reinvigorated as a concern, another question might be whether alien selves with the same correlates and clinical meanings are created through sustained use of pretend mode, psychic equivalence, and teleological mode. There are hints in the work of Fonagy and colleagues that the different forms of non-mentalizing populate different kinds of innerscape, but this has not been considered explicitly. For instance, Fónagy and Fonagy’s account of primary and secondary meanings (see Chapter 2) may suggest that teleological mode focuses on primary meanings at the expense of secondary meanings, and the intentional at the expense of the non-intentional, and so loses the richness and nuance these provide.

Finally, an exciting set of questions is opened up by the emerging ideas of Fonagy and colleagues regarding collective non-mentalizing. Among other things, their account of systematically distorted public communication undermining the deliberative basis of democratic societies can be regarded as an incipient political theory.16 This would be attended by an incipient theory of the psychosocial mechanisms of dehumanization. For instance, the political objectification of humans may be analysed in terms of how propaganda is sustained by pretend mode, how the emotive construction of threats is sustained by psychic equivalence, and how the replacement of the intentions of individuals with reifications of their labour or political ‘will’ is sustained by teleological mode.

Finally, as we saw in Chapter 9, Fonagy and colleagues identify a major shift in their current thinking on the basis of attention to the wider social system. We perceive great potential benefits of a ‘distributed model’ of mentalizing—and epistemic trust—rather than treating these as capacities lodged solely in the heads of individuals. This would help facilitate attention to the conditions that support mentalizing capacity and forms of epistemic trust. For instance, attention to the material conditions of mentalizing could open opportunities for dialogue and collaborative research with researchers in architecture, urban planning, and geography.

(p. 290) Some remaining questions: epistemic trust

Mentalizing and non-mentalizing has been a long-standing concern for Fonagy and colleagues, and the outstanding questions above arise in reflection on an established research programme and clinical modality. By contrast, their work on epistemic trust is more recent and emergent. We hope that future work will include efforts to operationalize epistemic trust, epistemic mistrust, epistemic vigilance, epistemic hypervigilance, and epistemic credulity. However, there are some theoretical questions that may have precedence. This includes the question of the object of ‘trust’. Their account tends to imply that trust is in the thoughts of others. It remains ambiguous whether they think it is possible to have epistemic trust or mistrust in the feelings of others. The capacity for epistemic trust or mistrust in the thoughts or feelings of oneself would also benefit from further attention, because this has barely been acknowledged but would seem to be of the utmost importance for linking the idea of epistemic trust to the capacity to reconceive thoughts and feelings as part of mentalizing.

It would be good to see attention to the relationship between epistemic mistrust and the forms of non-mentalizing. An obstacle to date has been the rather limited diet of examples of epistemic trust and mistrust used by Fonagy and colleagues, which focus on parent–infant and therapist–client interactions. Both these forms of interaction have quite specific power dynamics that are unrepresentative of most adult interactions, and which contribute to unusual configurations of trust and mistrust. One consequence, discussed in Chapter 9, is that the role of power dynamics in epistemic trust and mistrust can be left out of frame by Fonagy and colleagues. This has left significant open questions about how the epistemic mistrust of others, including in collective forms through epistemic oppression, can contribute to forms of non-mentalizing. This could help advance the interest Fonagy and colleagues have demonstrated in identifying the societal conditions that contribute to the prevalence of borderline personality disorder. Their current emphasis on the role of ‘fragmented, dehumanizing social systems’ across society runs some risk of aligning with Fonagy’s pessimism about social change. By contrast, attention to the ways that epistemic mistrust contributes to institutionalized forms of non-mentalizing may offer a more specific, testable, and optimistic locus for concern.17

Core strengths of the theory

Having described some particular reflections across the chapters of the book which we hope will be constructive for the paradigm, and identified some outstanding questions for future theory and research, we would like to close by highlighting what we see as core propulsive strengths of the theory of mentalization and epistemic trust. In closing the present book, we wish to draw attention to five features of the paradigm that we see as contributing especially to its tremendous strengths.

First, as documented in Chapter 2, Fonagy and colleagues have assembled a remarkable institutional and theoretical edifice, permitting the movement of knowledge across different domains. Perhaps most remarkably, their paradigm has helped sustain bi-directional transfer (p. 291) of knowledge between psychoanalysis and health services research and policy. Fonagy and colleagues have been unusually reflexive in their work, repeatedly sensing earlier than peers which way the winds are blowing in sociological pressures faced by psychological research and practice. As well as helping defend against threats and make use of opportunities represented by these pressures, this reflexivity has also helped their work in other ways—for instance, in helping them tailor their message to different constituencies.

Second, as Chapter 3 identified, their theory is intrinsically and insistently developmental.18 This has contributed to the acuity of research, such as the Anna Freud longitudinal study (see Chapter 2) and the pre-natal use of the Adult Attachment Interview (see Chapter 3). It has helped the theory incorporate insights from the developmental psychopathology movement, such as the p-factor and the idea of symptom networks (see Chapter 7). The developmental perspective has also underpinned adaptations of MBT to support children and adolescents, and parents and other caregivers (Chapter 8). Without a doubt, the theory of mentalization has contributed to a great deal of important empirical work with relevance across a range of disciplines.

Third, MBT offers a clinical modality with impressive transferability between settings and evidence of efficacy for many populations. Early indications suggest that the modality has adapted very fluently to remote delivery during the COVID-19 pandemic. Some elements of MBT are relatively generic, and the training burden for practitioners is comparatively light. However, MBT also offers compelling and distinctive clinical guidance—for instance, in the proposal that therapists pursue ‘contrary moves’ to help expand patients’ repertoire of mentalizing (Chapter 8). Likewise, there is a great deal of clinical acuity in the account of the modes of non-mentalizing. The distinction between hypermentalizing and mentalizing is important in its commitment to the therapeutic value of accurate integration objectively perceived with subjectively conceived reality. For all that they have not been operationalized and so remain scientifically underdeveloped, the forms of non-mentalizing are perspicacious characterizations of ways in which information may be filtered, obstructing generative doubt. They represent a distinctive contribution, irreducible to either existing cognitive models or psychoanalytic concepts of defence.

Fourth, the ideas of Fonagy and colleagues have an underpinning architecture, reflected in the 18 key concepts explored in this book. The writings of Fonagy and colleagues are carefully crafted to be accessible to specific audiences, and their clarity can give the impression that they need only be read once to be fully understood. This has misled some critics, particularly from within psychoanalysis, who misapprehend Fonagy and colleagues as a simplified and superficial version of familiar psychoanalytic notions. Yet, when the works of Fonagy and colleagues are looked at in the round, an underpinning architecture comes into focus, and reveals the huge questions pursued by the researchers and, ultimately, the depth of their response. So for instance a measure such as the Reflective Functioning Questionnaire may appear to be a simple set of items to a psychoanalytic critic, but in fact possesses ingenious and theoretically-innovative aspects in the conceptualization of certainty, uncertainty, and mental health. In one of his papers, Derrida asks, ‘What is it in a “great” work, let’s say of Plato, Shakespeare, Hugo, Mallarmé, James, Joyce, Kafka, Heidegger, Benjamin, Blanchot, Celan, that resists erosion? What is it that, far from being exhausted in amnesia, increases its reserve to the very extent to which one draws from it, as if expenditure augmented the capital?’19 In the case of Fonagy and colleagues, a strength of their work is its (p. 292) underpinning architecture, which allows clinical, theoretical, and research activities to cross-fertilize and sustain one another. For instance, ideas from attachment theory about adaptation have been drawn upon to qualify the potential pathologization of patients associated with diagnosis and deficit models; and empirical findings from developmental psychopathology (e.g. symptom networks) have been used to qualify limitations with attachment theory.

Fifth, something frankly quite distinctive about the work of Fonagy and colleagues has been their openness to development and refinement of their theory, through the incorporation of ideas from different communities. Theory, research, and clinical practice have been conducted in a way that, whenever possible, maintains ‘inquisitive stance’: acknowledging the potential for influence, surprise, change, and learning with and from others. Key moments include:

  • The integration of ideas from within psychoanalysis: the Anna Freudian and Kleinian traditions of psychoanalysis, as well as the work of Winnicott (Chapters 1, 2 and 6).

  • The integration of ideas from within cognitive science: work by researchers on social cognition (Chapter 3) and Gergely and Csibra’s natural pedagogy theory (Chapter 7).

  • The integration of ideas from within developmental science: the tradition of empirical attachment research in the 1990s (Chapter 3), and subsequent advances in developmental psychopathology and evolutionary approaches to behaviour (Chapter 7).

  • A significant transition seems to be presently underway in growing concern with culture and the wider social system (Chapter 9).

It is our hope that the reflections offered in the present book may contribute to greater understanding of the work of Fonagy and colleagues. We also hope they will prove useful for further refinements of the paradigm.


1 Sandler, J. (1962). ‘Research in Psycho-Analysis—The Hampstead Index as an Instrument of Psycho-Analytic Research’. International Journal of Psycho-Analysis, 43: 287–291, p. 290.

2 See also Duschinsky, R. and Solomon, J. (2017). ‘Infant Disorganized Attachment: Clarifying Levels of Analysis’. Clinical Child Psychology and Psychiatry, 22(4): 524–538; Solomon, J., Duschinsky, R., Bakkum, L., and Schuengel, C. (2017). ‘Toward an Architecture of Attachment Disorganization: John Bowlby’s Published and Unpublished Reflections’. Clinical Child Psychology and Psychiatry, 22(4): 539–560.

3 Sharp, C., Shohet, C., Givon, D., Penner, F., Marais, L., and Fonagy, P. (2020). ‘Learning to Mentalize: A Mediational Approach for Caregivers and Therapists’. Clinical Psychology: Science and Practice, 27(3): e12334.

4 Ensink, K. (2004). Assessing Theory of Mind, Affective Understanding and Reflective Functioning in Primary School‐Aged Children. Unpublished doctoral dissertation, London: University College London; Ensink, K., Normandin, L., Target, M., Fonagy, P., Sabourin, S., and Berthelot, N. (2015). ‘Mentalization in Children and Mothers in the Context of Trauma: An Initial Study of the Validity of the Child Reflective Functioning Scale’. British Journal of Developmental Psychology, 33(2): 203–217; Bizzi, F., Ensink, K., Borelli, J. L., Mora, S. C., and Cavanna, D. (2019). ‘Attachment and Reflective Functioning in Children with Somatic Symptom Disorders and Disruptive Behavior Disorders’. European Child & Adolescent Psychiatry, 28(5): 705–717. A parallel distinction has been made in scaling the Parent Development Interview for mentalization by Suchman, N. E., DeCoste, C., Leigh, D., and Borelli, J. (2010). ‘Reflective Functioning in Mothers with Drug Use Disorders: Implications for Dyadic Interactions with Infants and Toddlers’. Attachment & Human Development, 12: 567–585.

5 Thinking on embodied mentalizing has offered one way that questions about proto-thoughts and feelings are starting to be explored, though this has not seen much attention since Shai, D. and Fonagy, P. (2014). ‘Beyond Words: Parental Embodied Mentalizing and the Parent Infant Dance’, in M. Mikulincer and P. R. Shaver (eds), Mechanisms of Social Connections: From Brain to Group, Washington, DC: American Psychological Association, pp. 185–203. Another resource here may be Winnicott’s concept of ‘aliveness’. See Goldman, D. (2013). ‘Vital Sparks and the Form of Things Unknown’. Psychoanalytic Inquiry, 33(1): 3–20.

6 E.g. Freud, A. (1977). ‘Fears, Anxieties, and Phobic Phenomena’. The Psychoanalytic Study of the Child, 32(1): 85–90; Sandler, A. M. (1977). ‘Beyond Eight-Month Anxiety’. The International Journal of Psychoanalysis, 58: 195–207; Sandler, A. and Freud, A. (1982). ‘Discussions in the Hampstead Index on “The Ego and the Mechanisms of Defence”: VI. Orientation of the Processes of Defence According to the Source of Anxiety’. Bulletin of the Anna Freud Centre, 5(1): 5–35. See also ‘The problem of anxiety is a nodal point at which the most various and important questions converge, a riddle whose solution would be bound to throw a flood of light on our whole mental existence.’ Freud, S. ([1916–1917] 2001). Introductory Lectures on Psycho-Analysis (standard edition), London: Hogarth Press, 15–16, p. 393.

7 Fonagy, P. (2016). ‘Guest Post: It’s Important to Recognise When Childhood Anxiety is Becoming a Problem.’ Accessed at:

8 See e.g. Craib, I. (1994). The Importance of Disappointment, Bristol: Psychology Press.

9 Hausberg, M. C., Schulz, H., Piegler, T., Happach, C. G., Klöpper, M., Brütt, A. L., … and Andreas, S. (2012). ‘Is a Self-Rated Instrument Appropriate to Assess Mentalization in Patients with Mental Disorders? Development and first validation of the Mentalization Questionnaire (MZQ)’. Psychotherapy Research, 22(6): 699–709. See also Gagliardini, G. and Colli, A. (2019). ‘Assessing Mentalization: Development and Preliminary Validation of the Modes of Mentalization Scale’. Psychoanalytic Psychology, 36(3): 249–258.

10 Tinbergen, N. (1963). ‘On Aims and Methods of Ethology’. Zeitschrift Tierpsychology, 20: 410–433.

11 Cf. Bollas, C. (1996). ‘Borderline Desire’. International Forum of Psychoanalysis, 5(1): 5–10. ‘The borderline personality unconsciously seeks emotional turbulence because this complex of affect is the shape of the object of desire. Whether these people were intrinsically disturbed as infants, or, whether the early object world was itself disturbing, they knew the maternal object as disruptive effect. This effect then became the shape of the object, so, in seeking turbulence they are in fact constituting the primary object. As painful and disturbing an event as this is, it is nonetheless desired and finding themselves in states of distress is unconsciously gratifying. This person cultivates “borderline objects” which evoke turbulent frames of mind’ (p. 6).

12 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.

13 Ibid. 82.

14 Ibid. 86.

15 Cf. Blos, P. (1974). ‘The Genealogy of the Ego Ideal’. The Psychoanalytic Study of the Child, 29(1): 43–88.

16 See Habermas, J. (1970). ‘On Systematically Distorted Communication’. Inquiry, 13(1–4): 205–218.

17 This would align with advances in sociological theory. The concept of general societal fragmentation, which Fonagy and colleagues take from Durkheim, is generally regarded by sociologists today as limited; more specific causes tend to be identified in the rules, norms and routines that organise institutionalised practices. Baert, P. and Da Silva, F. C. (2010). Social Theory in the Twentieth Century and Beyond, Cambridge: Polity, Chapter 1.

18 See Sroufe, L. A. (2009). ‘The Concept of Development in Developmental Psychopathology’. Child Development Perspectives, 3(3): 178–183.

19 Derrida, J. (1989). ‘Biodegradables’. Critical Inquiry, 15: 812–873.