Rethinking Psychosis: Culture, Brain, and Context

Jan. 10–11, 2014

Workshop Summary

Part I

Constance Cummings and Kathy Trang

Last weekend, the FPR-CBDMH program held its inaugural winter workshop  for CBDMH students and affiliated faculty at UCLA, bringing together an informal group to share research and perspectives on a topic of general interest: “Rethinking Psychosis: Culture, Brain, and Context.” As participants put on their name badges and settled into comfortable chairs arranged in a square we were struck by the range of ages and backgrounds – graduate students, postdocs, and faculty from anthropology, neuroscience, psychiatry, psychology, social work, epidemiology, and public health, thirty in all, from UCLA, USC, McGill, University of Ottawa, Columbia, and Stanford. This interesting mix was the basis for highly stimulating discussions that lasted late into the evening on Friday over copious amounts of sushi and ended with some reluctant leave-taking after lunch on Saturday.


Psychosis is as old as human language and deeply entwined with it (the Furies in Euripides’ Orestes or the psychic decay of Sophocles’ Philoctetes come to mind). From a clinical perspective, its symptoms include delusions, hallucinations, disorganized thinking, grossly disorganized or abnormal motor behavior (including catatonia), and negative symptoms (e.g., diminished emotional expression and avolition). DSM-5 also includes an “attenuated psychosis syndrome” in a section entitled “Conditions for Further Study,” which lists some anomalous experiences putatively associated with later onset: “The individual may experience magical thinking, perceptual aberrations, difficulty in concentration, some disorganization in thought or behavior, excessive suspiciousness, anxiety, social withdrawal, and disruption in sleep-wake cycle. Impaired cognitive function and negative symptoms are often observed.” As discussions leading up to publication of the fifth edition of the DSM attest, the disorder’s etiology, course, and successful treatment remain complex, varied, and elusive (but see our recent blog post “Challenging the ‘Soft Bigotry’ of Low Expectations in Psychosis.” A key objective of the workshop was to try to understand cultural differences and cultural variability in psychosis and to build connections between what we know about the social world and what we know about neurobiological functioning.

Rethinking Psychosis: An Overview

The first session provided an overview by cultural psychiatrist Laurence Kirmayer (“Steps Toward a Cultural Neurophenomenology”) and neuroscientist Carrie Bearden (“Endophenotypes for Psychotic Bipolar Disorder in a Genetically Isolated Population”).

Laurence began by highlighting the importance of discussions like those encouraged at this workshop for tackling issues relating to mental health, particularly in the global context. The questions they address are particularly urgent, he emphasized, because beneath the basic scientific issues are ones relating to policy and practice; people are all too ready to assume universality of psychiatric conditions and their underlying mechanisms when much of psychiatry has been developed based on a bracketed population.

Psychosis, for instance, may not be a natural kind; grouped within this term are a family of diverse phenomena, some of which are ordinary and common. Though RDoC has tried to supplant the DSM with biological markers (e.g., endophenotypes), these markers do not necessarily map well onto existing categories. Similarly, he argued, when discussion moves away from a checklist (e.g., impaired reality testing, thought disorder, delusions, hallucinations, alterations in experience of self and agency, and negative symptoms )  and toward what constitutes a delusion (e.g., the degree of bizarreness, the amount of delusion), purely functional descriptions of neuronal activity, for instance, no longer seem adequate. To highlight the importance of the social/cultural milieu for these definitions, Laurence gave the example of the “Hindu milk miracle,” a 1995 incident in which millions of people in Hindu communities, in response to a news report observing that a Ganesha statue appeared to have drunk offered milk, made similar offerings. Taken out of context, this may seem odd; within context, however, it says nothing about abnormality, and everything about social processes.

What then is context? Laurence defined context as “a situation, setting, place, or time,” offering a “social system or local social world” and “a symbolic system of meaning and reference.” The contexts of psychosis are developmental stages, individual biography and experiences, family environment, and social environment. Across these contexts, Laurence argued that experience carries both social and biological weight; for instance, even though developmental stages (e.g., adolescence) may be defined by their biological milestones, they are also defined socially. Therefore, phenomenology is needed to illuminate these experiential qualities. As a method, Laurence argued, neurophenomenology – which ties experience to the mind and its neurobiological substrate – can “use knowledge of neural processes to devise new ways to examine experience” and “use phenomenology to identify new neural correlates or processes to look for.” As a broader theoretical project, neurophenomenology can “link neurophysiological findings to first-person perspectives” and offer a means of piecing together the dissected findings of neuroscience with the settings that give meaning to our experiences. Through embodiment, discursive processes, and participation in local worlds, cultural experiences can shape the form and content of psychotic experiences by influencing basic affective and cognitive processes.


Laurence ended his talk with a series of questions for the group. First, he was interested in “how . . . cultural processes and neural systems interact to give rise to psychotic experiences to allow resolution or recovery?” Next, he concerned with “how [to] put together the reality of social/cultural processes that make us human to think about the evolution of psychosis” and how to “frame the relationship between causal factors and their mediating processes.” Laurence emphasized that endophenotype research does not eliminate the question on either end; the question remains how one goes from causal factors to endophenotypes to their manifestation in particular contexts. Lastly, Laurence wanted to know how “particular forms of social adversity (e.g., discrimination) contribute to the risk of psychotic illness?”



Georg Northoff had two comments. The first concerned extrinsic vs. intrinsic context re Laurence’s call for better theorization of the social world, which could help guide the search for neurobiological underpinnings in a more precise way. To really get at how culture “works” neurobiologically, Georg argued that we need to understand intrinsic context-dependence, how neuronal activity is generated. (“Philosophically put, the context is the necessary condition for the possibility of the construction of that phenomenon.”) In the course of the discussion that followed, Ian Gold used an example from vision. A visual neuron is predisposed to activate in the presence of light, motion, or more complexly, a certain color. An external stimulus, the presence of something red for example in the case of a “red” neuron, generates or induces change in that neuron’s activity – the neuron’s disposition to “see” red and an object of reddish color interact, so to speak, and a history (different from a function or disposition) is created.

His second comment concerned neurophenomenology, which projects 1st and 3rd person perspectives onto the brain itself but does not really explore the structurating context, which is prior to the distinction between the two perspectives. Georg and Laurence agreed that the approach is not ecological; culture is “stripped off,” or only “taken in secondarily via the content, but by then it’s too late.” Georg also commented that neurophenomenology still deals with/cannot get rid of the explanatory gap between neuronal and mental processes (or between phenomenal and neuronal processes).

Steve López expressed concern about the degree to which the kind of nuanced critique that Laurence presented can be paralyzing in advancing the field and wondered whether there was value in a “good enough” understanding key points. As an example, he spoke about his work teaching Spanish-speaking people in LA to identify psychosis; although these interventions never emphasized fine-grained analyses, they do seem to improve understanding and provide care to those who usually don’t receive it in a timely manner. In the subsequent discussion, both Steve and Laurence agreed that there is value in clinicians keeping their eyes open, acknowledging their doubts, and checking with the local community.

Two discussants raised fascinating points about the psychiatric language we use to describe another’s disordered mind and at the same time the non-psychiatric, simple human interactions that are often at the core of treatment. Linda Garro wondered about the way we shift between nuanced critique and the diagnoses or descriptions (e.g., a “discrete psychotic episode”) we use to label disorder, despite understanding that these diagnoses are not natural kinds. Majorie Kagawa-Singer provided an example of “good enough treatment” by a therapist who simply engaged with his patient’s hallucinations, which allowed her to function.

Carrie Bearden began her talk (“Endophenotypes for Psychotic Bipolar Disorder in a Genetically Isolated Population”) by arguing that genetic research is important for medicine because many people do not currently benefit from treatment, there are many side effects associated with these treatments, and there are limited ways to intervene early or to prevent disease. Furthermore, progress has been particularly slow for behavioral disorders because of the questionable validity of diagnostic categories and of the imprecision of defining and measuring phenotypes. Carrie contended that research on endophenotypes and on functional genomics may allow for a better understanding of the disease state and comorbidities by revealing the biology beneath the syndrome.

Carrie’s research focuses on genetically isolated populations in Antioquia, Columbia, and in the Central Valley of Costa Rica. According to Carrie, these groups are an admixture of genetically similar founder populations of Amerindian and Spanish immigrants who lived in isolation for several hundred years and often intermarried, leading to exponential (1000-fold) increase in their respective groups From a genetic standpoint, studying isolated populations can be particularly fruitful for identifying genes or genomic regions segregating with the disorder of interest because only a few founders introduced and passed on mutations; thus, it may be easier to “identify inherited genomic regions identical by descent in modern day descents . . . [which may] hopefully localize with the disease.” For her study, 26 large families (N = 738) – heavily loaded for bipolar disorder (N = 181) – were evaluated for phenotypes hypothesized to underlie the disorder. Such measures included clinical assessments (e.g., diagnostic interview), temperament and neurocognitive assessment (e.g., measure of risk-taking and of declarative memory), and neuroimaging of 527 subjects (structural MRI and DTI).

In response to Laurence’s question of whether these measures had been normed in the two populations, Carrie admitted that they had been normed in Spanish-speaking populations generally, but not necessarily in these two groups; nonetheless, because the study was exploratory and only looked at association and heritability within the families, some meaning could still be derived from detected differences.

Regarding her results, Carrie said she was struck by the remarkable similarities between the intermediate traits identified in this unique cultural context and those observed in other cultural contexts (e.g., reduced corpus callosum volume and white matter integrity aligned with a Finnish twin study). Of the 169 quantitative phenotypes measured, 41 were both heritable and associated with BP-1. Several measures of temperament (delusion-proneness and perceptual creativity), cognitive (e.g., verbal memory and inhibitory control) and neuroimaging phenotypes (e.g., reduced cortical thickness), fell into this category, while no measures of cortical surface area did.

Carrie highlighted a few findings from her CBDMH pilot study of 17 children of bipolar parents in Columbia. She reported high rates of psychopathology (e.g., anxiety, ADHD, and some mood disorder) in the adolescents, which was “very consistent” with studies of children in other cultural contexts. Stress factors such as increased family conflict correlated with rates of anxiety; children with anxiety are more likely to develop bipolar disorder. The study was cross-sectional, however, so the key question is to determine the direction of causality.

Carrie concluded by saying that such research, which is very much in line with NIMH’s RDoC initiative, gives us a way of prioritizing phenotypes and identifying the most relevant traits for genetic mapping; however, many questions remain as to what the causal pathway is.

Brian Anderson asked Carrie about families’ coping mechanisms, given the cultural normativity of the disorder. In a followup question, Carole Browner wondered if affected individuals saw themselves as different. Both were terrific questions; but they concerned information that would be difficult to quantify, suggesting the importance of a mixed methods approach that could include qualitative data. Steve Lopez wondered how structural abnormalities related to mood fluctuation. Mary Helen Immordino-Yang asked about DTI (white matter) abnormalities in addition to abnormal cortical thinning, and whether it might be possible to triangulate between the two and neuropsychological functioning. More specific questions on method followed.

“Living Under the Description of a Psychotic Disorder”

The second session focused on an ongoing CBDMH project conducted by CBDMH co-director Steven López and USC neuroscientists Mary Helen Immordino-Yang and Vanessa Singh (“Family Socialization and Neurobiological Processes”). The main objective of the project is to examine the role of the social world – particularly the ways in which families of Mexican origin interact with their ill relatives – on the neurobiology as well as the behavior of persons living with schizophrenia.

Steve began by presenting a case study describing successful efforts by a middle-class family in Mexico to orient their ill relative to the social world. Largely through the efforts of the father (who described himself as his son’s “shadow” at one point), they were able to engage their college-educated son (“Alberto”) in meaningful work raising rabbits on a family-owned ranch. (“We consider this granja our salvation for Alberto,” the father said.) In an interview, Alberto said each day’s structure and responsibilities forced him to apply the criteria of logic to everyday social life and helped him deal with his feelings of sadness. This suggests a different process, in addition to families’ affective climate (which is essentially a “stress model,” Steve said). That is, families structure their lives in ways that keep the ill relative involved and less focused on the self. (As “Anna,” from a 2010 Harper’s article on attenuated psychosis syndrome observed, “the more I focus on my thoughts, the more it feels like they don’t actually belong to me.”) Steve compared this method of “treatment” with how persons living with psychotic disorders are treated in the US. Many reside in board and care facilities but spend most of their days drifting through public spaces (rather than being actively integrated into their social worlds), watching television, or sleeping.

Steve thought that these differences between social vs. self orientation might show up in brain activity. His second point was that (psychiatric) neuroscience is very limited in terms of how it studies or measures the social world. He said it was essential to ground research in social context, which his own group is doing via multiple methods: home observations, interviews, rating scales, and questionnaires, which can help measure social/self-orientation; ratings of social functioning; and clinical symptoms and their interrelationships. In a study of eighteen families with ill relatives in Mexico, for example, a higher degree of social orientation was negatively associated with self-report of daydreaming. It was also positively associated with a greater ability to assess others’ intentions and maintain friend/work relationships and negatively associated with clinical symptoms.

A second component of this project is an attempt to relate the research on self/other orientation to neurobiological functioning. In earlier research on healthy subjects, Mary Helen Immordino-Yang and colleagues examined the neural correlates of experiencing strong social emotions – a more abstract form of emotion associated with compassion for another person’s social pain, for example, or a more concrete, less inferential emotion associated with compassion for another person’s experiencing of physical pain (Immordino-Yang et al., 2009). Of particular interest to Mary Helen was the way in which, in the original research, “naturalistic” narratives designed to elicit compassionate responses would cause (video-taped) participants to pause after their initial, other-oriented attentional phase, avert their gaze, disengage, and reflect (“remove themselves from the current context”), which seems to suggest how continuously, fluidly, and dynamically the brain moves in and out of the social world. (“Feeling emotions about other people’s mental and physical situations involves the neural mechanisms for feeling and regulating your own body and for constructing your own sense of ‘self’,” she said.)

All of us have a varied ability to “toggle” between outward attention (e.g., social orientation toward an interviewer) and the more self-related default mode network (DMN; but see Georg Northoff’s critique of our “self-related” characterization later in this summary), i.e., they appear to be anti-correlated. Mary Helen exemplified this toggling by showing the video-taped reaction of one subject to a story about a selfless mother meant to induce compassion. “I can almost feel the physical sensations,” he said, and then, after a long pause, commented that the story made him think about his own parents. The pause seemed to imply behavioral manifestation of DMN activity (not only the pause but also such factors as eye-gaze aversion, the slowing of speech), which was supported by neural and psychophysiological data gathered later in the scanner using the same stories. (Interestingly, college students in China and the US showed similar neural activity to the emotion stimuli and similar reporting of the strength and frequency of emotions, but, Mary Helen said, “the way in which the neural activity corresponds trial by trial to psychophysiological reactivity and to the person’s experience is strongly shaped by culture and strongly related to their natural behavior in the interview.”)

This toggling appears to be disrupted in persons with schizophrenia, who show on the one hand abnormal connectivity within the DMN and on the other decreased deactivation of DMN when the outer world demands attention via, e.g., the executive control network. Although Mary Helen’s research on a group of 11 subjects of Mexican origin with schizophrenia did not incorporate an interview phase, the subjects were exposed to somewhat simpler versions of emotionally compelling stories in the scanner. They were asked to determine whether the actors in the stories were feeling physical pain, e.g., from a broken limb, or some form of emotional pain. The question was whether the subjects’ toggling between stories composed of concrete details and those that would have more abstract, socio-emotional or DMN-activating impact would correlate with the social-orientation measures based on Steve’s home interviews (preliminarily, this appears to be the case).

Participants had several questions on methodology and the directionality of the findings. Other questions concerned the current strength of evidence for DMN disruption and the significance in terms of level of coupling between DMN and, e.g., the salience network, or inefficient toggling between DMN and the executive control network in schizophrenia. This work is relatively new, and the temporal resolution of fMRI is an issue, Georg Northoff said. Another issue he wrote about in Unlocking the Brain, is that “the resting state, metaphorically speaking, ‘has its hands’ in all kinds of neural processing related to different stimuli, tasks, and their respectively associated functions.” Schizophrenia is perhaps best characterized “by an overall presence of [what early psychiatrists referred to as] a ‘basic disturbance of self’” that is more deeply rooted than resting state networks (Northoff, 2013, p. 395). Also, the DMN is a lot fuzzier than our common characterizations suggest (“DMN = self”; non-DMN = non-self); he said that the self experiences both internal and external contents and should not be identified with certain contents, e.g., the DMN. (He also pointed out that although internal and external contents may be confused in schizophrenia, consciousness is not impaired.)

Ian Gold felt that the researchers could make the interpretation of the neural data easier by looking at component parts of a psychological state or activity (such as eye-tracking) rather than, say, “thinking about myself” (which may be less boring but is infinitely more complex). At this stage, that is essentially what we do, Mary Helen said (i.e., we correlate eye-gaze aversion with activation of the DMN). The current thinking, as Vanessa Singh observed, is that it’s harder to correlate a complex social behavior to a specific region of the brain; hence the burgeoning interest in networks and overall context.

Finally, as Laurence Kirmayer reminded us, if the neuroscience pans out (and Carrie Bearden pointed out many other purely physiological factors affecting toggling, such as faster/slower blood flow), correlation of a non-task related ecologically more interesting measure (e.g., the ability to toggle between networks) with psychological functioning and behavior would have great potential in terms of carry over to Steve’s work with families. Efforts at greater social integration create opportunities for ill relatives to practice switching, implying support for “a totally different kind of intervention.”

Stay tuned for Part II, which covers “Psychosis and the Environment” and “Psychosis and the Self.”



Immordino-Yang, M. H., McColl, A., Damasio, H., & Damasio, A. (2009). Neural correlates of admiration and compassion. Proceedings of the National Academy of Sciences of the United States of America, 106(19), 8021–8026.

Immordino-Yang, M. H., Christodoulou, J. A., & Singh, V. (2012). Rest is not idleness: Implications of the brain’s default mode for human development and education. Perspectives on Psychological Science, 7(4), 352–364.

Northoff, G. (2013). Unlocking the brain: Consciousness (Vol. 2). New York: Oxford University Press.