Metacognition, social cognition, and mentalizing in psychosis: are these distinct constructs when it comes to subjective experience or are we just splitting hairs?

The construct of metacognition as studied across cognitive, educational, developmental and abnormal psychology, refers to a set of activities which involve thinking about one’s thinking and responding accordingly to what is happening in the moment in one’s life [21]. The integrative model of metacognition has been applied to explain the alterations in self-experience thought to be at the root of the subjective experience of psychosis. In this model, metacognition has been proposed to refer to a spectrum of activities. These include awareness of discrete experiences which can have embodied, affective and cognitive aspects. This is to say that metacognition can involve being aware of one’s own specific thoughts, beliefs, emotions, desires and experiences within one’s body (e.g. tension in one’s forehead).

Metacognition also makes it possible for a person to see and reflect upon how different specific experiences may be related to one another (e.g. an awareness of how certain thoughts and emotions might be related the experience of bodily tension) and then synthesize those into a broader and evolving sense of oneself and others (e.g. an awareness of how certain thoughts are distressing and lead to tension in the moment within the larger context of a psychological conflict the person has been facing over the last several months) [23]. In this model, metacognition is thus integrative. It involves a process by which individual experiences are understood in terms of their relationship to one another and ultimately integrated into a larger sense of a person as more than the sum of single experiences. The larger sense a person might form of themselves or others, enabled by metacognitive processes, is importantly never static. The holistic sense a person has of themselves is likely to be influenced and changed by new discrete experiences, while the meaning assigned to a specific experience is likely to be influenced by one’s sense of self. For example, an experience of becoming visibly angry might mean something different to someone who thinks about themselves as unflappable and nonreactive as opposed to righteous and enjoying conflict. In parallel, the experience of being visibly angry, if it is remarkable or happens enough times, may lead one to form an idea about themself as different than unflappable and nonreactive.

There are several qualities that metacognition cannot be reduced to in the integrated model. First, the nature of metacognition is not necessarily the product of willful act. A person may decide to reflect upon a given experience or something about the nature of their life and consequently realize something important. Metacognition may also involve automatic processes such that an awareness of a given experience or of something larger going on in one’s life may simply occur to a person without purposefully directing their attention to those issues. Second, metacognitive acts cannot be said to be purely solitary mentalistic acts. A thought about oneself may occur when one is alone. However, regardless of whether another person is present, the ideas we form about ourselves and others are always constructed with a potential audience in mind. For any sense of ourselves we have, there is also a sense of how others, real or imagined, might respond. This is to say that metacognition is not about the formation of final ideas within the solitude of one’s mind but occurs in intersubjective contexts [19]. Third, metacognitive acts are not tied to just thoughts about oneself. A person can notice and form ideas about others, one’s larger community and the possible usages of that knowledge to make sense of and respond to longstanding and newly emergent challenges [7, 20, 24]. Finally, metacognition is not something that can be categorized as intact versus not intact nor compromised by one set of problems. Metacognitive capacity can vary between persons with different degrees of deficits that can be caused by multiple social, psychological and biological processes.

Applied to psychosis, deficits in metacognition have been suggested to result in a sense of oneself and others as relatively more fragmented, resulting in the subjective sense of diminished self-experience reported in qualitative research [6]. In a manner consistent with the initial formulations of Bleuler [25] and Jung [26], decrements in metacognition have been suggested to explain how persons diagnosed with those conditions experience fragmentation of thoughts, affect and will, withdraw from social interaction, and understand themselves and the world in ways that are often inaccessible to others [7].

To measure decrements in metacognition within psychosis, an interview was developed to elicit a sample of how persons think about their life and its possibilities and challenges. To date, the most commonly used method is a semi-structured interview, the Indiana Psychiatric Illness Interview, which asks participants to offer a general narrative of their lives and experiences of psychosocial challenges [27]. This in turn required the development of a quantitative scale to assess participant’s metacognitive capacity as reflected within that interview. To date, the most commonly used measure is the Metacognition Assessment Scale Abbreviated (MAS-A) [7]. Developed from the Metacognition Assessment Scale [20] (MAS), the MAS-A provides separate scales to assess metacognitive capacity for thinking about the self (self-reflectivity), others (awareness of others), one’s place in the community (decentration) and one’s ability to develop and enact plans based on that information (mastery). The MAS-A departs primarily from the MAS in that the items of each of the four scales reflect a specific metacognitive act that is ordered numerically by level of complexity. Lower scores reflect a more fragmented experiential field per domain whereas higher scores reflect greater degrees of integration. Adequate psychometric properties have been reported [28]. Though less studied, other methods exist for assessing these forms of metacognition including the Metacognitive Assessment Scale Revised [29].

Using these methods, research has tested the assertion that deficits in metacognitive capacity are prominent elements in a network of disturbances that culminate in the interruption of the lives of persons diagnosed with psychosis. At the level of more basic biological processes, studies have linked metacognitive deficits with alterations in phenomena that underpin social behavior, including levels of oxytocin [30] and linguistic structures that allow for communication [31, 32]. Metacognition has also been linked to interpersonal functioning [33‐35] as well as the processes that support social connections, including empathy [36‐38]. Deficits in metacognition also have been linked to self-compassion [39, 40] and the richness of personal narrative [41], as well as being implicated in influencing the association between emotional distress and positive symptoms [42]. Other studies have found that alterations in metacognition are linked to changes in neurocognition and, in a formal network analysis, metacognition was found to be a central node in a network that included social cognition, neurocognition and symptoms [43]. Concerning clinical phenomena, a substantial number of cross-cultural studies have linked metacognition with negative symptoms [44‐46], in particular with expressive negative symptoms [47, 48], and issues related to clinical insight [8], motivation [49], and intrapsychic function [50].

As work on links with psychosocial functioning has moved forward, the potential of metacognition as a treatment target has attracted increasing interest. One such treatment is Metacognitive Reflection and Insight Therapy [51] (MERIT). MERIT is a form of integrative individual psychotherapy that seeks to enhance a person’s abilities to form an integrated sense of themselves and others and to use that sense to manage challenges and pursue the possibility within their lives. It conceptualizes the mechanisms of action in a manner analogous to physical therapy. Specifically, patients, within the context of a therapeutic relationship, exercise the ability to form increasingly integrated ideas about the self and others by steadily engaging in metacognitive actions in accord with their current, maximal capacity. Though still in its infancy, research on this approach has suggested that this treatment is acceptable to patients and leads to clinically meaningful gains in randomized trials [52‐54]. Qualitative and case studies have linked this approach to metacognitive capacity with growth in senses of agency, historicity, and greater potential for future action [55, 56]. Other work includes a metacognitive oriented social skills training program found to be effective in a randomized trial [57], and interventions focused on the ability to doubt one’s conclusions, which may impact a person’s larger sense of themselves and others [11, 58].

Research generated by the integrated model of metacognition has revealed that many diagnosed with psychosis experience themselves and others in relatively more fragmented ways. These experiences of fragmentation may represent a central node in an interactive network that underlies disability including impairments in other forms of cognition and behavior. It offers a promising model of how both negative symptoms and poor insight may emerge and has spurred the development of novel psychosocial approaches to treatment which may address some of the most subjective and debilitating aspects of psychosis.

Social cognition is a broad term that refers to a collection of mental operations that underlie social interactions, including the perception and interpretation of others’ intentions, dispositions, and behaviors [59]. In research on social cognition in psychosis, most studies have focused on four domains. These include i) theory of mind, or the ability to understand the thoughts and intentions of others; ii) social perception, or the ability to identify socially bound roles and rules; iii) attributional style/bias, or how one infers the causes of events; and iv) emotion processing, or the ability to perceive and use emotions, including identifying emotions in others [17]. Persons with psychosis tend to exhibit deficits in social cognition across all domains [60], and these deficits have been linked to numerous functional outcomes, including a strong connection to community functioning [61]. In addition, these deficits have been found to span the course of the illness, including clinical high risk, early psychosis, and more prolonged groups [62, 63].

Measurement of social cognitive domains can be broadly categorized into two approaches: social cognitive skills (i.e. the capacity to make correct determinations, e.g. theory of mind) and social cognitive biases (i.e. patterns of interpretation unresponsive to evidence, e.g. hostile attribution bias) [64]. Most commonly, social cognitive domains in each category are assessed with task-based assessments that require a respondent to engage or demonstrate each skill. Self-report and interview-based measures are used less commonly [65], in part because individuals may inaccurately rate their own skill [66]. Measures of social cognition have benefitted from increased emphasis on methodological rigor and reliability through the Social Cognition Psychometric Evaluation study, which aimed to identify measures of the core domains of social cognition and examine their psychometric characteristics [17]. Measures of social cognitive skills had the strongest psychometric support, including one measure of theory of mind, The Hinting Task [67], and two measures of emotion processing, the Bell Lysaker Emotion Recognition Task [68] and the Penn Emotion Recognition Task [69, 70]. To illustrate, the Hinting Task, historically cited as one of the most commonly utilized measures of theory of mind within the field of social cognition [65], is still frequently utilized and tests the capacity for persons to infer the intentions of indirect speech [71, 72]. In this task, participants listen to ten short passages of a verbal exchange between two characters, which concludes with one of the characters dropping a “hint.” Participants must then explain what the character intended by this hint and are scored on the accuracy of their response [67].

A large body of research has demonstrated the relationship between social cognition and functional outcomes like social functioning, social skills, independent living skills, and community integration [61, 65]. More recent research has continued to corroborate these findings. For example, it has been demonstrated that deficits in thought, language and communication domains are associated with worse functional outcomes, including occupational functioning, ability to communicate effectively in social situations, and general community functioning [73]. Difficulties interpreting verbal cues like humor and emotional prosody [74] as well as visual cues like facial emotion [75] have been linked with poor social communication and functioning in psychosis. Further, deficits in affective prosody recognition have been found to be significantly associated with both role and social functioning in first episode schizophrenia [76]. It has also been demonstrated that affective prosody recognition is negatively correlated with disorganization symptoms, whereas facial emotion recognition is negatively associated with positive, negative, and disorganized symptom domains [76]. More broadly, social cognition, especially domains of theory of mind and emotion recognition, have been linked to elevated disorganization [77], however the relationships between positive and negative symptoms and social cognition are often more complex [78]. Recent results also suggest a role for social cognitive impairments in other meaningful outcomes for people with psychosis, such as suicidal ideation [79] and homelessness [80].

Work on social cognition in psychosis has led to increased interest in interventions focusing on social cognition. These treatments have generally focused on two methods of intervention: group-based treatments and computerized cognitive remediation. These treatments have shown promise in their impact on social cognition, but study results have been mixed in demonstrating the generalization of clinical benefits to in vivo social functioning. Social Cognition and Interaction Training (SCIT), a manual based group intervention, initially showed promising results with improvements on social cognition measures, including social and emotion perception [81, 82] and social functioning [83]. However, a more recent examination of SCIT found no significant differences between an intervention group and treatment as usual group on several social cognition outcomes, though it should be noted that significant improvements in quality of life, emotion recognition, and social skills were found within the SCIT group [84]. Another group intervention, Social Cognitive Skills Training, has demonstrated positive effects, including improvement in emotional processing [85], social functioning, and psychiatric symptoms [86]. When integrated, cognitive remediation approaches and social cognitive training demonstrate greater improvements in working memory, cognitive flexibility, verbal memory, and theory of mind than a waitlist group, but groups tend to not differ on measures of symptoms and social functioning [87]. These findings have raised questions as to whether improvements in social cognition from treatment may not generalize to long-term or in vivo social functioning. Some have attempted to address this issue by adding an intervention component that allows for practicing social skills; however, this approach appears to have not improved durability of in vivo functioning benefits [88].

Research exploring social cognition has revealed that many diagnosed with psychosis appear relatively less able to correctly identify the thoughts, feelings and intentions of others. Greater levels of these difficulties appear to be a significant source of social dysfunction and are correlated with symptom severity. Social cognitive deficits likely contribute to a vicious cycle in which difficulties correctly apprehending the internal states of others leads to isolation which then leads to exacerbation of other features of psychosis. This work has spurred the development of carefully defined interventions which improve social cognition and can be easily incorporated into existing treatment regimens.

Mentalizing can be defined as the “ability to understand actions by both other people and oneself in terms of thoughts, feelings, wishes, and desires” [89] (p3). First introduced as a higher-order mental process through which individuals engage in representation and symbolization, the term mentalization was used to characterize the process by which individuals make meaning of human behavior within a context of understanding the mental activities of oneself and other people [90, 91]. Fonagy and colleagues [92] then progressively outlined an attachment-based approach to the development of mentalizing on the basis of observed, early parent-child interactions in borderline personality disorder [93] and other severe mental disorders [94] in which there is a failure to understand oneself because of a lack of access to or a negative regard for others [95]. During this period, the construct of mentalization was relabeled as mentalizing to highlight its procedural character [96] and its explicative and uncertain, imaginative nature [97]. Bateman and Fonagy [89] focused on three non-mentalizing and potentially pathogenic modes: psychic equivalence (i.e. perceiving inner states to be directly equivalent with events in the external world); pretend mode (i.e. decoupling mental life from external reality); and teleological mode (i.e. recognizing mental states only if outcomes are physically observable). Mentalizing was further categorized to include four sub-domains: automatic/explicit, self/other, affect/cognition, and inner-focused/outer-focused [97, 98], illustrating the dynamic nature of mentalizing as nested within varying relational contexts [95]. At any one point, an individual may consider the experiences of others while at the same attending to their own thoughts or feelings. This process may involve explicit, effortful reflection or operate implicitly and prior to any reflection [90].

Common measures of capacity for mentalizing include survey and transcript-based instruments. The Reflective Functioning Questionnaire [99] and Parent Reflective Functioning Questionnaire [100] are self-report measures designed to assess mentalizing capacity across two subscales, which, respectively, represent salient aspects of mentalizing and evaluate a parent’s ability to understand their child’s behavior in terms of underlying mental states. The Reflective Functioning Scale [101] is a measure used to code mentalizing capacity within an interview that provides opportunities for persons to reflect on their childhood experiences, particularly in the context of their relationships to caregivers (e.g. Adult Attachment Interview). In addition to assessing the patient’s capacity for mentalizing, instruments have been created to assess the therapist’s capacity to mentalize. Such measures include the Therapist Mental Activities Scale, Therapist Relationship Interview, and Mentalization-Based Treatment (MBT) adherence and competence scale, which assess the therapist’s own ability to mentalize as well as their ability to promote the mentalizing process within therapy [91]. A final approach to measuring mentalizing, the interactional approach, derives from the assumption that mentalizing occurs in relationships with others, or within intersubjective contexts [91]. An example of this is Conversational Analysis [102], a qualitative method that examines the structure and process of social interactions, such as a therapy session. Across these studies, it may be noted, the non-mentalizing modes (i.e. psychic equivalence, pretend mode, teleological mode) are only partially and indirectly assessed.

Deficits in mentalizing have been tied to a multitude of poor outcomes [103]. Research has found that deficits in mentalizing are linked to decreased community engagement, reduced vocational functioning, as well as poorer quality of life [90, 104]. Other studies, which have investigated whether aggression and violence are linked with impaired mentalizing, offer conflicting views. One study found that psychotically driven aggression was linked with poorer mentalizing capacity [105], while another found no difference in mentalizing between individuals with schizophrenia spectrum disorders who are violent versus those who are non-violent [106]. A recent study found that mentalizing deficits act as a mediator between childhood abuse and negative symptoms in that the effect of childhood abuse on the severity of negative symptoms was amplified by the absence of the ability to mentalize [107]. Most generally, a recent study conducted by Boldrini and colleagues [108] suggests that the presence of impaired mentalizing may predict the development of psychosis itself.

Inspired by research on mentalizing and the emerging recovery movement for individuals with psychosis, psychotherapeutic interventions have been developed to target mentalizing deficits in psychosis [90]. Mentalization-Based Treatment for psychosis (MBT-p) is an adaptation of a therapeutic intervention originally developed for borderline personality disorder. The primary premise of MBT-p is that psychosis entails core deficits in the areas of awareness of one’s own mind (e.g. failing to differentiate between one’s thoughts and external stimuli) and the minds of others (e.g. difficulty in interpreting others’ intentions) [109]. As a dynamic practice, MBT-p aims to promote mentalizing ability to strengthen the ability to make sense of interpersonal experiences and to better navigate social life by recognizing how various social situations affect thoughts, feelings, and motivations [90]. Interventions focus on the patient’s experience in the present through the therapist’s continued curiosity about the patient’s thoughts, feelings, and intentions as well as by encouraging the sharing of narratives of social interactions [103]. Case studies have reported that MBT-p is linked to improvements in social functioning [103] and the decrease of negative symptoms. A recent, randomized-controlled trial found MBT-p led to significant improvements in social functioning at post-treatment and at 6-month follow up. Results further revealed a trend toward better performance on measures of mentalizing ability in the MBT-p group and an increased effectiveness of MBT-p in early psychosis versus prolonged psychosis [110].

Research exploring mentalizing has revealed that many diagnosed with psychosis appear to experience less secure levels of attachment and as a result are relatively less able to consider the experience of others while attending to their own thoughts or feelings. This process may lead to reduction in persons forming coherent ideas about what others want, think and feel. As has been found in social cognition, greater levels of these difficulties appear to be a significant source of social dysfunction. This work has also spurred the development of interventions with psychodynamic roots which are less structured than social cognition interventions and focus on noticing and developing the ability to think about others and one’s own relationship with them.