Part I
Theory
1
The Mentalization-Focused Approach to Social Development
Peter Fonagy
Introduction
Reflective function refers to a quantified index of attachment-related mentalization, that is, the capacity to conceive of mental states as explanations of behavior in oneself and in others. We assume that the capacity to mentalize is a key determinant of self-organization that, along with contributory capacities of affect regulation and attention control mechanisms, is acquired in the context of early attachment relationships. Disturbances of attachment relationships will therefore disrupt the normal emergence of these key social-cognitive capacities and create profound vulnerabilities in the context of social relationships. Unusually, for what is fundamentally a psychoanalytic approach, we elaborated our model of social development on the basis of empirical observations as well as clinical work.
In the 1980s an extremely active research program in developmental psychology investigated when children begin to understand that people are capable of having false beliefs about the world (Perner & Lang, 2000; Wellman, 1990). Yet a number of researchers consider the resulting construct of theory of mind and its false-belief paradigm to be too narrow (Carpendale & Chandler, 1996) as it fails to encapsulate the relational and affect regulative aspects of interpreting behavior in mental state terms. Developmentalists have started to use the term mentalizing as an alternative, because it is not limited either to specific tasks or particular age groups (Morton & Frith, 1995; OâConnor & Hirsch, 1999).
We define mentalization as a form of mostly preconscious imaginative mental activity, namely, interpreting human behavior in terms of intentional mental states (e.g., needs, desires, feelings, beliefs, goals, purposes, and reasons). Mentalizing is imaginative because we have to imagine what other people might be thinking or feeling; an important indicator of high quality of mentalization is the awareness that we do not and cannot know absolutely what is in someone elseâs mind. We suggest that a similar kind of imaginative leap is required to understand oneâs own mental experience, particularly in relation to emotionally charged issues. In order to conceive of others as having a mind, the individual needs a symbolic representational system for mental states and also must be able to selectively activate states of mind in line with particular intentions, which requires attentional control.
The ability to understand the self as a mental agent grows out of interpersonal experience, particularly primary object relationships (Fonagy, 2003). The babyâs experience of himself as having a mind or self is not a genetic given; it evolves from infancy through childhood, and its development critically depends upon interaction with more mature minds, assuming these are benign, reflective, and sufficiently attuned. Mentalization involves both a self-reflective and an interpersonal component. It is underpinned by a large number of specific cognitive skills, including an understanding of emotional states, attention and effortful control, and the capacity to make judgments about subjective states as well as thinking explicitly about states of mindâwhat we might call mentalization proper. In combination, these functions enable the child to distinguish inner from outer reality and internal mental and emotional processes from interpersonal events.
This paper provides an overview of the mentalization-focused approach to social development. We address the complex relation of attachment and mentalization and summarize contemporary neurobiological research bearing on the cognitive, affective, and relational aspects of mentalizing. This biological perspective underpins our discussion of the role of mentalizing in the development of the agentive sense of self, followed by a broader consideration of the role of interpersonal relationships in the maturation of mentalizing capacities. Finally we discuss the contribution of attachment trauma to the development of psychopathology by virtue of undermining mentalizing capacity.
Evolutionary and Neurobiological Links Between Attachment and Mentalization
The Selective Advantages of Attachment
As our understanding of the interface of brain development and early psychosocial experience increases, we can see that the evolutionary role of the attachment relationship goes far beyond giving physical protection to the human infant. Attachment ensures that the brain processes that come to subserve social cognition are appropriately organized and prepared to equip the individual for the collaborative and cooperative existence with others for which the brain was designed.
In our view the major selective advantage conferred by attachment to humans is the opportunity to develop social intelligence that nearness to concerned adults affords. Alan Sroufe (1996) and Myron Hofer (2004) played a seminal role in extending attachment theory from a concern with the developmental emergence of a complex set of social expectancies to a far broader conception of attachment as an organiser of physiological and brain regulation. More recent work has begun to articulate the associated biological pathways (e.g., Champagne et al., 2004; Jaworski, Francis, Brommer, Morgan & Kuhar, 2005; Plotsky et al., 2005; Zhang, Chretien, Meany, & Gratton, 2005). This body of work illustrates how processes as fundamental as gene expression or changes in receptor densities are influenced by the infantâs environment. The brain is experienceâexpectant (Siegel, 1999).
The Selective Advantages of Mentalization
Mentalization is arguably the evolutionary pinnacle of human intellectual achievement. But what has driven the selection processes of the two million or so years of human evolution toward a consciousness of mental states in self and others? Was it to meet the periodic challenges the physical environment presented to our ancestors who were presumably only somewhat more agile and strong than we are? Surprisingly, leaps forward in human brain size in the course of evolution do not correspond to what we know about ecological demands on our hominid ancestors (e.g., climatic variability, threat of predation, and availability of prey).
The evolutionary biologist Richard Alexander (1989) proposed a widely accepted model of how humans evolved their minds. He suggested that our exceptional intelligence evolved not to deal with the hostile forces of nature but rather to deal with competition from other people. This further evolution occurred only after our species had already achieved relative dominance over their environment. At that point we became our âown principal hostile forces of natureâ (Alexander, 1989, p. 469). And to meet this challenge to the survival of our genes, those with common genetic material had to cooperate.
All species face competition from conspecifics but humans are special in the role that social groups play in achieving success in this regard. A kind of evolutionary arms race probably took place among ever more effective social groups. Competition with intelligent conspecifics requires skill in understanding and outsmarting other people. As the intelligence of the opposition increased so too did the requirement for ever-greater mentalizing ability. The construction and manipulation of mental scenarios (of thoughts about thoughts and feelings) acquired a major reproductive advantage. The assumption that the mind governs actions and the possibility of interpreting and anticipating behavior permits cooperation, offers competitive advantage, and continually selects for increasingly higher levels of social interpretive capacity.
The Interpersonal Interpretive Function
The capacity to interpret human behaviour (see Bogdan, 1997) requires the intentional stance: âtreating the object whose behaviour you want to predict as a rational agent with beliefs and desiresâ (Dennett, 1987 p. 15). We label the capacity to adopt this stance the interpersonal interpretive function (IIF), an evolutionary-developmental function of attachment. Unlike Bowlbyâs internal working model concept, its function is not to encode representations of attachment experiences, nor is it a repository of personal encounters as in Sternâs (1998) concept of schemata-of-ways-of-being-with. Rather, the IIF is a cluster of mental functions for processing and interpreting new interpersonal experiences that includes mentalization and the cluster of psychological processes on which effective mentalizing depends (Fonagy, 2003).
The emphasis on interpretation is helpful because we are particularly concerned with the possibility of misinterpretations and misperceptions of othersâ thoughts, feelings, and intentions. Interpretive function also underscores the perspective-taking facet of mentalization that equips us to recognize how individuals can come to different conclusions with the same set of facts at their disposal (Carpendale & Chandler, 1996). We suggested, following Baron-Cohenâs (2003) distinction between theory of mind and empathy, that a cognition-oriented interpersonal interpretive function (IIF-C) is complemented by an emotion or affect oriented set of processes (IIF-A). Earlier, Henry Wellman proposed a related developmental transition from a desire psychology of toddlers to a beliefâdesire psychology of 3â4-year-olds (Bartsch & Wellman, 1995). We also include in IIF-A the notion of mentalized affectivity which refers to the simultaneous experiencing and knowing of a feeling.
Three Neural Systems of Social Cognition
Four emotional processing and control mechanisms contribute to the developmental unfolding of interpretative function: labelling and understanding affect, arousal regulation, effortful control, and specific mentalizing capacities (Fonagy & Target, 2002). We propose that these interpretive functions are subserved by three separate but interconnected and interacting nodes within the brain that are related to social detection, affect regulation, and cognitive regulation (Adolphs, 2003; Nelson, Leibenluft, McClure, & Pine, 2005).
The first node consists of a hard-wired set of structures that categorizes stimuli as social and deciphers or detects their social purpose. The brain regions that make up this social-detection node include the fusiform face area, the superior temporal sulcus, and the anterior temporal cortex. These regions have been shown to be involved in carrying out basic perceptual processes on social stimuli.
The second node is concerned with affect and encompasses regions of the brain engaged by reward and punishment. The generation of affect imbues social stimuli with emotional significance and modulates emotional arousal. The system has a significant role in mediating attachment experience and is activated by attachment-related stimuli. Brain regions that make up the affect-regulation node include the amygdala, hypothalamus, nucleus accumbens, and bed nucleus of the stria terminalis. These regions interact with the social-detection node to imbue social stimuli with emotional significance.
Our primary concern, the third node, is devoted to cognitive regulation. Its key functions include inhibiting prepotent responses (effortful control), mediating goal-directed behavior, and mentalizing (as exemplified in perspective taking and theory-of-mind tasks). The brain regions that make up the cognitive-regulation node include the dorsomedial prefrontal cortex and the ventral prefrontal cortex. There are several systems within these structures that mediate different aspects of regulation and control, including integrating emotion with cognitive processing and making accurate social judgements. Each of these aspects of social intelligence subserves different aspects of interpersonal interpretation.
The foundations of mentalization are present in nonhuman species. Recent work on the mirror neuron system (Gallese, Keuseers, & Rizzolatti, 2004; Rizzolatti & Craighero, 2004) claims that understanding othersâ actions requires the activation of the mirror neuron system, and understanding othersâ emotions requires the activation of viscero-motor centres. Motor neurons, originally found in the ventral premotor cortex of the macaque monkey, respond both when the monkey performs a particular goal directed act and when it observes another individual performing a similar action (Gallese, Fadiga, Fogassi, & Rizzolatti, 1996). Action observation automatically activates the same neural mechanism triggered by action execution or even by the sound produced by the same action (Kohler et al., 2002). The mirror neuron system also encompasses communicative actions, both in monkeys (Ferrari, Gallese, Rizzolatti, & Fogazzi, 2003) and in humans (Rizzolatti & Craighero, 2004). In a recent fMRI study, participants observed communicative mouth actions in humans, monkeys, and dogs which led to the activation of different cortical foci corresponding to the different observed species; actions in the motor repertoire of the observer (e.g., biting and speech reading) were mapped accordingly on the observerâs motor system (Buccino et al., 2004).
Extrapolating from mirror neuron research, we might conceive of a two-level system underpinning mentalization: a frontal-cortical system that invokes declarative representations and a mirror-neuron system subserving a more immediate and direct understanding of the other. In the anterior insula, visual information concerning the emotions of others is directly mapped onto the same viscero-motor neural structures that determine the experience of that emotion in the observer (Wicker et al., 2003). This direct mapping can occur even when the emotion of others is merely imagined (Singer et al., 2004). Gallese and Goldman hypothesize a shared subpersonal neural mapping between what is enacted and what is perceived that can be used to predict the actions of others (Gallese, 2003; Goldman & Sripada, 2005). This automatically established link between agent and observer may not be the only way in which the emotions of others can be understood, but the simulation of actions by means of the activation of parietal and premotor cortical networks may constitute a basic level of experiential understanding that does not entail the explicit use of any theory or declarative representation.
Mentalization also involves the capacity to represent affects in others (perhaps the limbic circuits including the amygdala), to inhibit the prepotent response of assuming that others think exactly the same as we do (the anterior cingulate) and representing and reasoning about beliefs and also desires in others (the orbitofrontal and prefrontal areas of the cortex). We should remember that mentalizing pertains to interpreting mental states in both self and others. Representing the contents of oneâs own mind taps into the same meta-representational capacity required for representing the contents of anotherâs mind (Frith & Frith, 2003). Self-awareness and awareness of the mental states of others are closely linked in terms of the brain areas involved. Mentalization does not just facilitate collaboration and positive relationships but also facilitates social survival through competition. Self-awareness enables us to modify the way we present ourselves to others and to mislead them. The right prefrontal cortex may âallow us to see ourselves as others see us so that we may cause competitive others to see us as we wish them toâ (Alexander, 1990, p. 7).
The Evolutionary Psychology of Mentalization
Because the mind needs to adapt to ever more challenging competitive conditions, the capacity for mentalization cannot be fixed by genetics or constitution. The social brain must continuously reach higher and higher levels of sophistication to stay on top. Evolution has charged attachment relationships with ensuring the full development of the social brain. The capacity for mentalization, along with many other social-cognitive capacities, evolves out of the experience of social interaction with caregivers. Increased sophistication in social cognition evolved hand in hand with apparently unrelated aspects of development, such as increased helplessness in infancy, a prolongation of childhood, and the emergence of intensive parenting.
We have proposed a mechanism for this process rooted in dialectic models of self-development (Cavell, 1991; Davidson, 1983). Our approach explicitly rejects the classical Cartesian assumption that mental states are apprehended by introspection; on the contrary, mental states are discovered through contingent mirroring interactions with the caregiver (Gergely & Watson, 1999). Psychoanalysts have...