Stein et al1 provide a state-of-the-art summary of the philosophy-psychiatry field that centers around the mind-brain-world triumvirate. They do an astounding job summarizing 30 years of activity in the field, culminating in a present-into-the-future consideration of embodied cognition. Here I look forward in a different direction. Borrowing the internalist/externalist distinction from historiography, externalist philosophy of psychiatry might be paraphrased as a social philosophy of psychiatry. Externalist philosophy of psychiatry would consider the role of the social and cultural world in shaping conceptions of mental illness and its treatments. It would consider social phenomena as environmental contributors to the complex causalities that converge on psychopathological phenomena. It would reconsider concepts of free will in terms of structural features of the human organism alongside the structural features of our increasingly complex social environment. It would draw upon philosophy of technology, social epistemology, metaphysics, political philosophy, and philosophical anthropology in placing clinical diagnosis, research and theory into sociocultural contexts. Through understanding these social contexts, we can find new clarity in the mental health project. I can make these abstract generalizations more vivid by considering some examples from our current, and rapidly changing, era. I mention very briefly three areas ripe for development by social philosophers of psychiatry: the engulfment of clinical activity by electronic medical records; the role of domination by the State in mental health; and the psychopathology of artificial intelligence (AI). Back in 1977, Heidegger2 introduced the concept of "enframing" to the then-developing philosophies of technology. "Enframing" involves the tendency of technologies to appropriate resources for their own purposes, forming a "standing reserve" of resources increasingly remote from human goals. Heidegger's thinking helped to found the now-familiar philosophy of technology tropes such as "technologies solve problems which in turn generate new problems, requiring new technological solutions". This recursive expansion of technological imperatives can be seen playing out in clinics around the world through electronic medical records. Psychiatric training directors are already noting the deterioration of interviewing and diagnostic skills when young trainees are driven to checking box after box of clinical findings, being "enframed" into the metaphysical structure of those records, and missing the big narrative picture of the patient. What made "enframing" toxic for Heidegger was that technologies change the way we think, as in the above-mentioned check-box interviews. Fortunately, philosophers of technology have been working on this set of problems for a while, and philosophers of psychiatry can find a rich literature. The response to digital "enframing" has already begune.g.,3. The archeologist/anthropologist team of D. Graeber and D. Wengrow4 reconsider world history with a particular interest in the origin of social inequity and the origin of the State (e.g., government). As social scientists, they have access to and can interpret archeological and anthropological sciences freed from the limits of written text "traces". By studying cultural artifacts, they can provide evidence about the diversity of community living over the millennia. One of their major insights is the common threads that constitute domination of people by other people – e.g., social control. They identify three factors that are apparent singly, doubly or triply in virtually all the cultural forms of dominated communities: control of violence, control of knowledge, and control by charisma. Control of violence refers to the people who are authorized to apply violence to others, whether war-making or managing crime (as just two examples). Control of knowledge has to do with the people controlling what counts as truth and knowledge. Control by charisma has to do with the persuasiveness of the would-be empowered group; the ability to bring believers/followers into the fold. The authors note that the most successful domination occurs when all three factors of the triad are seized. They also note that the means to capturing the three factors are not determinative; one can have democratic, imperial or fascist dominations, for example. I need not detail the relevance of these concepts to our current world situation, but I can point to their relevance to the social philosophy of psychiatry. We have seen the rise of populist authoritarian rule over the past decade – the charisma element. We have witnessed the loss of confidence in "elites" – scientists, professors, philosophers, intellectuals and doctors. More importantly, talk of "post-truth" and the transformation of public lying from vice to virtue is eroding our prior forms of the domination of knowledge. Control of violence is also being shaken worldwide, as social media enable the challenging of conventional mechanisms of violence control (policing, the courts, civil liberties, rule of law, subordination) through familiar tropes such as #Icantbreathe, #metoo, and the unprecedented January 6 storming of the US Capitol building. For philosophers of psychiatry, the range of tasks deriving from these social changes is huge. Some are simple to conceive: how do clinicians communicate around these phenomena? Others are definitional: how many people must believe something before a false belief is not a delusion? Still others are insidious: what counts as mental health in a post-truth world where confidence in State institutions is diminishing? In recent months we have witnessed the explosion of AI technology proffered by the major info-tech companies. We have also witnessed gushing enthusiasm as well as apocalyptic worries from many users and commentators. Somewhere in-between these social reactions resides the systematic appraisal of these systems as they evolve. We have seen how human biases are entrained in AI, leading to perpetuation of bias in later use5. We have also seen both absurd and frightening results of natural language devices when asked ordinary questions6. Social philosophers of psychiatry can envision the development of looping effects of human-AI deviant interchanges online and worry about the social consequences. Some examples of inquiries in this area point to treating deviant AI responses as psychopathology analogues7. Others consider using AI tech as explanatory models of human psychopathologies and tools for therapeutic developments8. We can also envision the engagement of AI into disputes about medical and other misinformation in the social media environment, thus connecting psychopathology of AI to social-domination theory described above. Some preliminaries of this work have appealed to a "father" of philosophy of psychiatry: K. Jaspers9. All this potential work is wide open, deep and important.

Kirkbride et al1 provide convincing evidence of a causal relationship between social determinants and poorer mental health outcomes. The extent, complexity and prevalence of these determinants could lead to hopelessness at the prospect of addressing the inequalities and inequities involved. However, the authors instil hope and optimism by providing examples of interventions delivered at the family, school, neighbourhood or societal level that have been effective in primary prevention. They then offer seven key recommendations for action by mental health professionals, policy makers and researchers to prevent or reduce mental health problems that arise from social determinants. All these recommendations warrant further discussion. However, I will focus on two: the need to prioritize interventions providing positive outcomes across multiple domains, and the demand to utilize the causal architecture approach in psychiatric epidemiology. Kirkbride et al propose that mental health outcomes should be evaluated in any policy, programme or intervention targeting social determinants. This would be a highly informative and beneficial endeavour. However, while achievable, it would require a significant shift in how research is conducted, and would involve establishing and maintaining relationships and collaborations across multiple disciplines, faculties and departments. The benefits would be reciprocal, as the functional outcomes of education, employment, housing and diversion from the criminal justice system are likely to be very meaningful from the viewpoint of several stakeholders. This action would also entail that mental health clinicians and researchers have a role in advocacy at a government level, as any policy that affects the key social determinants will have a knock-on effect on mental health at a population level. However, clinicians are already overburdened with the high prevalence of mental disorders, and academics need to be continuously productive in a highly competitive environment for grants and fellowships. The benefits of this approach would take years to realize and can be challenging to evaluate directly. Therefore, those time-consuming endeavours of across-discipline collaborations and advocacy would need to be acknowledged by the authorities and institutions that award grants and promotions, as some are already doing. Kirkbride et al call for the causal architecture approach to be used in psychiatric epidemiology, as this would lead to an understanding of the pathways between one or more exposures and the disease/disorder. This approach could result in the identification of modifiable risk factors and inform where interventions should be targeted. The mechanisms by which social determinants contribute to the aetiology of mental disorders are likely to be complex and could also differ amongst individuals. For example, in the case of childhood adversity and trauma, the experience itself of being the victim of abuse or adversity could lead to subsequent mental health problems, but, additionally, the manner by which individuals are supported (or not supported) by their family and loved ones, as well as by the legal or medical system, could also be involved in the pathway2. Another example of a likely complex relationship is that between low income and poor mental health, since socioeconomic differentials as well as psychological perceptions and self-esteem, in addition to an absolute lack of material resources, may lead to a higher risk of experiencing a mental disorder3. Another key aspect of the causal architecture approach is that the sampling should be from a representative population. As the authors rightly point out, there is a relative lack of research in lower- and middle-income countries, despite these countries representing over 80% of the world's population4. There have been recent endeavours to undertake methodologically robust epidemiological studies in the Global South. In one such study, the incidence of untreated psychotic disorders was found to be three times higher in Northern Trinidad compared to both the Kancheepuram district in India and Ibadan in Nigeria5. The incidence rate observed in Northern Trinidad (59.1 per 100,000) would have ranked as the third highest if considered in a previous meta-analysis including 44 estimates of the international incidence of psychotic disorders6. Both substance use and levels of community violence and crime have increased markedly in Trinidad in recent years. This suggests that there are settings in which the potential for primary prevention of psychotic disorders by targeting social determinants could be particularly high. It is also worth considering the proportion of mental disorders that interventions targeting social determinants could prevent. Population attributable fractions have been estimated to be nearly 38% for childhood adversity and just under 10% for cannabis use in schizophrenia, and to be 13.4% for childhood sexual abuse in depression7. This is in keeping with the call by Kirkbride et al for social determinants to be fully integrated into the bio-psycho-social model of mental health and illness. It also highlights that primary prevention strategies aimed at social determinants could reduce the incidence of mental disorders but not eradicate them. Considering this, while Kirkbride et al focus on primary prevention, social determinants could also inform improvements and advancements in secondary prevention. To ensure an equitable allocation of resources, a secondary prevention strategy could be the determination of resourcing and funding of services on the basis of the geographic prevalence of social determinants, such as social deprivation, fragmentation and rates of migrants and minority ethnic groups. Indeed, it has been demonstrated that the incidence of psychotic disorders can be reliably predicted based on the prevalence of social determinants in geographically defined areas resulting from census data8. Yet, the majority of mental health services continue to be funded on a per capita basis. There are also inequalities and inequities for certain groups in accessing treatments; for example, individuals from racial and minority ethnic groups are less likely to be referred to or receive psychological interventions in the UK9. This unequal distribution of resources and these barriers to accessing services need to be addressed alongside the efforts for primary prevention. The call to action by Kirkbride et al is ambitious, but its goals are achievable. It can help to address the underlying inequalities and inequities within our societies that contribute to the development of mental health problems and may sustain them across further generations.

Borderline personality disorder (BPD) was introduced in the DSM-III in 1980. From the DSM-III to the DSM-5, no major changes have occurred in its defining criteria. The disorder is characterized by instability of self-image, interpersonal relationships and affects. Further symptoms include impulsivity, intense anger, feelings of emptiness, strong abandonment fears, suicidal or self-mutilation behavior, and transient stress-related paranoid ideation or severe dissociative symptoms. There is evidence that BPD can be reliably diagnosed and differentiated from other mental disorders by semi-structured interviews. The disorder is associated with considerable functional impairment, intensive treatment utilization, and high societal costs. The risk of self-mutilation and suicide is high. In the general adult population, the lifetime prevalence of BPD has been reported to be from 0.7 to 2.7%, while its prevalence is about 12% in outpatient and 22% in inpatient psychiatric services. BPD is significantly associated with other mental disorders, including depressive disorders, substance use disorders, post-traumatic stress disorder, attention-deficit/hyperactivity disorder, bipolar disorder, bulimia nervosa, and other personality disorders. There is convincing evidence to suggest that the interaction between genetic factors and adverse childhood experiences plays a central role in the etiology of BPD. In spite of considerable research, the neurobiological underpinnings of the disorder remain to be clarified. Psychotherapy is the treatment of choice for BPD. Various approaches have been empirically supported in randomized controlled trials, including dialectical behavior therapy, mentalization-based therapy, transference-focused therapy, and schema therapy. No approach has proved to be superior to others. Compared to treatment as usual, psychotherapy has proved to be more efficacious, with effect sizes between 0.50 and 0.65 with regard to core BPD symptom severity. However, almost half of the patients do not respond sufficiently to psychotherapy, and further research in this area is warranted. It is not clear whether some patients may benefit more from one psychotherapeutic approach than from others. No evidence is available consistently showing that any psychoactive medication is efficacious for the core features of BPD. For discrete and severe comorbid anxiety or depressive symptoms or psychotic-like features, pharmacotherapy may be useful. Early diagnosis and treatment of BPD can reduce individual suffering and societal costs. However, more high-quality studies are required, in both adolescents and adults. This review provides a comprehensive update of the BPD diagnosis and clinical characterization, risk factors, neurobiology, cognition, and management. It also discusses the current controversies concerning the disorder, and highlights the areas in which further research is needed.