Conscious life and voodoo death

Abstract

Philosophical zombies—entities identical to humans but lacking conscious experience—have long served as the go-to thought experiment for illustrating the ‘hard problem of consciousness’: why subjective experience should arise at all from physical processes. This essay introduces another Haitian concept to counter the zombie intuition that mind can be separated from life processes:‘voodoo death.’ Voodoo death refers to the death of a person caused by the expectation of dying rather than organic causes. Drawing on studies of this phenomenon and interpreting it through two contemporary biogenic theories of the mind—enactivism and the free-energy principle—this essay argues that ‘voodoo death’ is an extreme manifestation of a broader phenomenon wherein consciousness plays an active role in the dying process. Consciousness, the argument suggests, is intrinsic to biology—not only because it is intrinsic to life but also because it is inextricably tied to death.

Introduction

“Why should physical processing give rise to a rich inner life at all?” asks philosopher David Chalmers. “It seems objectively unreasonable that is should, yet it does.” (Chalmers 1995:201). Zombies became the go-to thought experiment to illustrate what is now known as the ‘hard problem of consciousness’: if we can conceive of someone who is physically and functionally identical to us but lacks subjective experience, then subjective experience must not be reducible to purely physical processes. This suggests that there is something about consciousness that physicalist theories alone cannot fully explain. We are left with an explanatory gap.

As an anthropologist interested on how people think across different cultures—about consciousness, for example—I’m sometimes struck by the cultural specificity of the ‘hard problem’. The presence of consciousness in a material world has seemed “objectively unreasonable” to Chalmers and other thinkers within the Western philosophical tradition, but it hasn’t been so for the great majority of traditions outside the orbit of scientific naturalism—including, it should be noted, the Haitian one from which the very word ‘zombie’ comes from. In Haitian folklore, a zombi is a corpse externally animated and controlled by a bokor, a sorcerer with the power to enslave it (the concept is rooted in cultural fears of loss of freedom with connections to the history of slavery; Coates 2006). Existing in a state of mindless servitude, it is fundamentally different from a philosophical zombie: since it is devoid of agency, it is not truly alive. Its lack of consciousness is entirely expected.

The main reason the ‘hard problem’ is not ‘a thing’ in most traditions is that mind is almost universally assumed to be continuous with the rest of living nature. It is seen as an expression of life itself, its animating force. Without a bifurcation between life and subjective experience, thinkers within these traditions do not perceive an explanatory gap to fill (pace Chalmers 2018).

To be sure, cultural variation in how people approach this issue does not dissolve the hard problem per se, but it demonstrates that its salience is subject to framing effects that are historically shifting and perhaps ephemeral. Perhaps one way to overcome the ‘hard problem’ is simply the passage of time.

This essay does not address the ‘hard problem of consciousness’ but rather another problem whose explanation also requires accepting a deep continuity between mind and life, along with theories of consciousness that accommodate this continuity—one that, as it happens, also derives from Haitian folklore. Not zombies this time, but “voodoo”—specifically, “voodoo death.” The term refers to the death of an individual caused by the expectation of death rather than by any direct physical cause. In Haiti, this takes the form of a deadly curse cast by a bokor (sorcerer) on a victim who, believing they are doomed, succumbs to death shortly after. The term has since

been generalized beyond the Haitian context to describe psychologically induced deaths that result from frightening, culturally ingrained events. Synonyms include “psychogenic death,” “nocebo-induced death,” or “hex death.”

Cardiologist Bernard Lown recounts a vivid case retrieved from the Archives of Neurology and Psychiatry:

“A Hindu physician was authorized by prison authorities to conduct an astonishing experiment on a criminal condemned to death by hanging. The doctor persuaded the prisoner to permit himself to be exsanguinated —bled to death — assuring him that death, though gradual, would be painless. The convict, on agreeing, was strapped to a bed and blindfolded. Vessels filled with water were hung at each of the four bedposts and set up to drip into basins on the floor. The skin on his four extremities was scratched, and the water began to drip into the containers, initially fast, then progressively slowing. By degrees the prisoner grew weaker, a condition reinforced by the physician's intoning in a lower and lower voice. Finally the silence was absolute as the dripping of water ceased. Although the prisoner was a healthy young man, at the completion of the experiment, when the water flow stopped, he appeared to have fainted. On examination, however, he was found to be dead despite not having lost a single drop of blood. (Lown 1999:31, citing Yawger 1936)

The secular reader may approach this anecdote with some degree of disbelief. It seems objectively unreasonable to think that an individual could die solely from expectation, that consciousness alone possesses the power to bring about its own demise. If it did, we would need to reconsider a thing or two of what we currently believe about consciousness and its relationship to the body and to life itself. After all, voodoo death would not feel counter intuitive otherwise.

What I will do in the essay is 1) present evidence that ‘voodoo death’ is indeed a genuine phenomenon but far from a special one, for it represents a limit-case of a broader phenomenon wherein consciousness plays an active role in the process of dying, and 2) propose that accommodating this phenomenon requires theories of consciousness that assume a deep continuity between mind and life, where these exist in a relationship of ‘circular causality.’ I will

discuss two main theories that support this continuity: enactivism and the free-energy principle. As I explore the insights these theories throw into the apparent mystery of ‘voodoo death’, I also show how ‘voodoo death’ brings up unexplored implications for these theories, and for theories of consciousness more broadly. Ultimately, I contend that the counterintuitiveness of ‘voodoo

death’ and the ‘hard problem of consciousness’ arise from one and the same source: the analytic separation between mind and life. Just as the idea of deep continuity demystifies the former, it may go some way towards dissolving the ‘objective unreasonableness’ of the latter.

Voodoo death: evidence and explanations

A phenomenon “so foreign to the experience of civilized people that it seems incredible; certainly if it is authentic it deserves careful consideration”. So wrote American physiologist Walter Cannon about ‘voodoo death’ in an eponymous landmark paper in American Anthropologist, which first popularized the term (1942:169). Cannon surveyed a number of extreme ethnomedicogenic cases of this kind among so-called ‘primitive’ societies and found that they all followed a similar pattern: death appeared to result either from deadly curses or the breach of specific taboos. He discussed, for example, the case of a Maori woman who died after eating fruit she later discovered had come from a tabooed place. He observed that in Australia,

among Aboriginal people, pointing a bone at someone could lead to sudden death. Likewise, in Latin America and Africa, in societies where witchcraft was a major source of social distress, being cursed could also lead to ‘voodoo death.’ Cannon argued that this was likely caused by the fear of supernatural consequences for culturally sanctioned actions, combined with a deeply rooted conviction in the inevitability of death. He also believed that the phenomenon was far more likely to occur among “primitive people, because of their profound ignorance and insecurity in a haunted world, than among educated people living in civilized and well-protected communities” (1942:174).

As it turns out, the phenomenon might not be so exotic after all. What a bokor can allegedly accomplish in Haiti, a medical doctor may replicate in a modern clinic to similar effects. Consider the case of Sam Shoeman, a patient from Tennessee whose death was chronicled by physician Clifton Meador (who would later write a series of papers on psychogenic death; Meador 1992; 2005). Diagnosed with metastatic carcinoma of the esophagus in October 1973, Shoeman was told by Meador that he had only a few months to live. Despite the prognosis, Sam managed to survive through Christmas and even to get married shortly after. When he eventually passed away a few weeks later, no one was surprised. Sam knew he was going to die, and so did his physician and family. However, the autopsy revealed something unexpected: doctors found only a 2cm nodule of cancer in his liver (with no trace in the esophagus) and a small patch of pneumonia, neither of which was sufficient to cause death. “He died with pneumonia and with cancer,” Meador wrote, “but I do not believe that he died of cancer. What is certain is that he died thinking he was dying of cancer, a belief shared by his wife, her family, his surgeons, and me, his internist” (1992:244). Meador ultimately concluded that he might have played a role in sending Sam to an early grave.

Similar anecdotes have been accumulating in the medical archives in quite a substantial number over the years (see Lester 2009 for a review). There is a sequence of events common to many such cases. Sometime after the ‘curse’ (a taboo breach, a terminal diagnosis), the individual enters a period of social withdrawal, where they detach emotionally from their friends and relatives. This stage is typically followed by apathy and aboulia. As this apathic state deepens, they progress into a near-catatonic phase and resignation to their fate, which is soon followed by death. The psychological process accompanying this sequence has been termed “givingup/given-up complex” (Engel 1971) or “give-up-itis” (Leach 2018). While this progression is frequently observed in the literature on voodoo death, however, it is not universal. The manner of psychogenic demise is culturally patterned and has been shown to vary across social worlds.

An illustrative example is the case of Sudden Unexplained Nocturnal Death Syndrome (SUNDS), which afflicted Hmong refugees in California during the late 1970s. This term was coined while documenting the deaths of 117 Hmong individuals who succumbed to the syndrome between 1977 and 1983, all of whom had migrated to the United States following he CIA-backed Secret War in Laos (Center for Environmental Health & Injury Control 1988). All but one of the victims were men, with a median age of thirty-three. In some instances, the deaths were witnessed by relatives who were awakened by the sounds of gurgling and gasping, but the sleeper would die within a few minutes and could not be revived. Autopsies failed to reveal an identifiable cause. Medical scientists were initially puzzled, but it emerged that the phenomenon was intimately connected to that of ‘sleep paralysis’.

Sleep paralysis is a state of consciousness that occurs during sleep, in which a person feels fully awake but is unable to move or speak. They also feel, often, an oppressive weight on the chest, a sense of suffocation, and the perception of a shadowy figure approaching. Subjects report that these experiences become intensely frightening. Only in extremely rare cases, however, does sleep paralysis lead to death.

Henry Fuseli, The Nightmare, 1781. A pictorial representation of sleep paralysis.

Why was sleep paralysis so lethal among Hmong refugees? In a book entirely dedicated to these events, Shelley Adler (2011) pointed to something significant about the cultural background of these individuals. In their previous lives in Laos, the Hmong engaged daily with a pantheon of spirits through various ritual practices or by observing rules and taboos while working at home or in the forest. These spirits included untamed, malevolent entities capable of harming humans, one of which, dab tsog, was notorious for attacking sleepers by attempting to suffocate them. Back in their villages, the Hmong had methods to protect themselves from dab tsog. They could perform animal sacrifices to appease ancestors or call upon a shaman to perform protective rites. Forced migration to the U.S. made these practices unviable. Sacrifices ceased, ancestors were neglected, shamans lost their traditional roles. According to Hmong immigrants, this left them vulnerable to deadly attacks by the spirit, who could now kill them easily in their sleep. Adler argues that this new situation had all the prerequisites to produce a strong ‘nocebo effect’ (the opposite of a ‘placebo effect’), and that it was this heightened perception of threat that made the dab tsog deadly during sleep paralysis. Why were men affected more than women? Because in traditional Hmong

villages, men were the primary custodians of ritual lore and responsible for providing spiritual protection. Women were generally excluded (and protected) from such knowledge. Hence the gender specificity of the attacks. In short, culture can powerfully shape psychogenic death, as well as the ways people think about it, which, in turn, produces self-fulfilling effects on how one experiences it.

One might understandably take these cases with a pinch of salt. Can we truly rule out all potentially lethal physical factors and instead attribute such deaths to a hard-to-account-for psychogenic causes? For instance, there is evidence that SUNDS has a hereditary component, with a significant prevalence among Southeast Asia (Zheng et al. 2018).1 Also, in many of the Aboriginal Australian cases described by Cannon, the cursed individual undergoes what is effectively a ‘social death’, where they are denied food and water as everyone around them anticipates their demise (Eastwell 1982). Death, here, would be more plausibly explained by malnutrition or dehydration than by fear.

Still, aside from the fact that a sorcerer’s ability to induce a member of society to refuse food and die is interesting in itself (and could still be considered a form of death by suggestion), studies have concluded that “[malnutrition and dehydration] are not present in most of those cases” (Lester, 2009:9). The instances in which physical causes can be confidently ruled out altogether are now abundant. We also know that something akin to ‘voodoo death’ occurs in other animals, and that there are parallels with better-documented typologies of death (see below). Modern reviews now affirm that “evidence has accumulated to support [Cannon’s] concept that ‘voodoo’ death is, in fact, a real phenomenon” (Samuels 2007:8), prompting neuroscientists to speculate about its underlying mechanisms.

The mechanistic explanations that have been offered appeal to processes that occur in the brain under traumatic fear or stress. In his original paper, Cannon (1942) conjectured that voodoo death might result from prolonged hyperactivity of the sympathetic nervous system. Adrenaline release and stress-induced overstimulation could lead to hypovolemic shock, a shortage of blood supply to vital organs. Others disagreed. Curt Richter (1957) proposed that voodoo death could stem from hyperactivity of the parasympathetic, rather than the sympathetic, system. This would involve a rapid decrease in heart rate and eventual death due to cardiac failure and the characteristic “giving up” behavior. More recent reviews from the perspective of neurocardiology concur with Richter’s account, expanding it. It has been speculated that extreme stress can trigger an ‘autonomic storm’ in the brain (Samuels, 2007), a severe dysregulation of the nervous system involving a deadly cascade of processes, including toxic catecholamine release and calcium overload in cardiac cells, which ultimately result in heart failure. Others point to potential dopamine depletion and frontal-subcortical circuit dysfunction as contributing factors (Sternberg 2002), mechanisms that may also accelerate the aboulia that often precedes psychogenic death (Leach 2018).

All of these represent well-informed speculations about what happens to the neurocardiac axis during such frantic moments—and important speculations at that—but I propose that, even if correct, they remain inadequate as comprehensive explanations of the phenomenon.

To start, there are instances of severely stressful situations that do not lead to psychogenic death. Consider hostage situations or warfare: soldiers might die from bullets and shrapnel, but there is little evidence that these scenarios are breeding grounds for psychogenic death. This suggests, perhaps, that it is something specific to the idea of death—its perception or expectation—rather than mere fright, that contributes to a person’s demise in these circumstances. Most crucially, there are cases of psychogenic death that occur in situations where overwhelming stress is largely absent.

Consider the findings reported by Phillips et al. (1993) in their Lancet paper ‘Psychology and Survival’, which serves both as an illustration of how the anticipation of death can begin long before the event and of how culture can shape the process. Comparing the cause of death of half a million Americans and some 28,000 Chinese-Americans in California, Phillips and colleagues found that “Chinese-Americans, but not whites, die significantly earlier than normal (1,3 – 4,9 yr) if they have a combination of disease and birth year which Chinese astrology and medicine consider ill fated” (Phillips et al. 1993:1142). For Chinese-Americans who died of lung diseases (bronchitis, emphysema and asthma), those who were born in the “Metal years”—which, for the Chinese, are associated with the organ of the Lung—had an average age at death of 66.9 years; those who died of the same diseases but who were born in other years, the average age at death was 71.9 years. No difference was found for ‘whites’ who died of similar diseases in the same period. The authors concluded that he intensity of the effect was correlated with “the strength of the commitment to traditional Chinese culture” (1993:1143)—a likely phenomenon of psychogenic death.

Since then, the same team has uncovered further correlations between death and cultural beliefs. For instance, they found that cardiac deaths among Chinese and Japanese individuals with chronic heart disease in California increased by up to 27% on the 4th of the month (the number “4” is ominous in both cultures, as its pronunciation is nearly identical to the word for “death”) (Phillips et al. 2001), or that mortality rates temporarily decrease before a symbolically meaningful occasion (e.g. Passover for Jewish believers), followed by a spike in deaths immediately after the event (Phillips and King 1988; Phillips and Smith 1990; see also Dein and George 2001). These are not sudden deaths due to fear, but in all likelihood, psychogenic deaths nonetheless. If so, then “voodoo death” can be said to occur on different timescales, underpinned by expectations of dying that can vary in intensity, levels of conscious reflection, and arguably mechanisms.

A strong continuity also exists—or so I will argue—between this body of work and the growing research on the association between ‘self-rated heath’ (SRH) and mortality (Jylhä 2009). Over the past half-century, epidemiological studies have discovered that one’s perception of health can, in and of itself, be a powerful predictor of age at death. There seems to be no better way of predicting mortality—not even objective factors like knowledge of the subject’s medical history or that of their family—than the answer to this simple five-point question: “In general, would you say your health is: excellent, very good, good, fair, or poor?” The association applies especially to older people but has been confirmed across different populations and demographics. As the authors of one of the earliest serious studies on the subject concluded, SRH “appears to have a unique, predictive, and thus far inexplicable relationship with mortality” (Idler and Kasl 1991:55, in Moerman 2002:120).

One might expect this association to weaken as more sophisticated controls for objective measures of health are introduced. In fact, the association has remained consistent. To mention a recent study, Wuorela et al (2020) examined the relationship between SRH and mortality among 70-year-old community-dwelling individuals in Finland over 5-, 10-, and 27-year follow-ups, with the objective of comparing SRH’s predictive power with that of a thorough assessment of ‘objective health’ (OH). This was evaluated using the Rockwood Frailty Index, a method that measures the cumulative effect of 36 health deficits across physical, cognitive, and social domains, incorporating a wide range of symptoms, signs, disabilities, diagnoses, and laboratory measurements. It represents one of the most thorough and well-rounded objective assessments of an individual’s vulnerability to date. As the results came in, researchers found that SRH outperformed OH as a predictor of short-term mortality (5 and 10 years). Only in the long term (27 years) did OH perform slightly better. From a biomedical point of view, these results are counterintuitive: why should this single bit of subjective information on a Likert scale hold such strong predictive power?

This question has led many to ponder on what exactly SRH is, or what it expresses. What we know is that SRH requires individuals to actively assess their health, whether intuitively or through self-reflection. This process involves integrating subjective experiences, such as physical sensations and vitality, with objective information, like knowledge of diagnoses and medications, as well as their overall sense of well-being. Research into what SRH expresses suggests that its ratings most strongly reflect one’s sense of vitality (feeling energetic and full of life), followed by physical functioning and bodily pain. Other dimensions, such as mental health and social functioning, also contribute, though to a lesser extent (Au and Johnston 2014).

As for why SRH predicts mortality, Marja Jylhä (2009), whose work is the most influential on the subject, argues that it is because SRH effectively captures physiological dysregulations in the body (e.g., inflammatory markers like cytokine activity) that are missed by objective indicators. These dysregulations can play a critical role in chronic diseases such as cardiovascular conditions, diabetes, and even cancer, thereby contributing to mortality while eluding clinical assessments. SRH’s predictive power on mortality, in short, would be a testament to human interoceptive subtlety and accuracy regarding the body’s actual health (see also Kananen et al. 2021). If Jylhä is correct, the phenomenon would have little to do with ‘voodoo death,’ as the causal arrow leading to death appears to flow from the body to the mind, rather than the other way around.

I propose that this view is partially mistaken, and this is where theories of the mind and consciousness come into play. New theories of perception, which I will explore below, suggest that interoception is not simply a readout of sensory signals but a process aimed at fulfilling the organism’s expectations. In other words, one’s perception of health (which captures well-being across a variety of biopsychosocial dimensions) can, in and of itself, function as a self-fulfilling prophecy. If that’s the case, its effect on mortality falls into the same category as psychogenic death, albeit unfolding on different timescales. Exceptional as it may seem at first, ‘voodoo death’ might simply be an extreme case of a much more common phenomenon in which the mind plays an active role in the process of dying.

Before delving into the core of this contention, let me address the overdue point about why all this should be of interest to theories of consciousness.

Intermezzo: why ‘voodoo death’ matters to theories of consciousness

To many strands of philosophy of mind, voodoo death exemplifies a form of ‘mental causation.’ This is what the term ‘psychogenic’ refers to. Mental causation, as is widely agreed in philosophy, “is at the heart of the mind-body problem” (Shoemaker 2001:71), the challenge of understanding how the mental affects the physical. Numerous positions exist within this debate, with varying degrees of compatibility, ranging from epiphenomenalism (the idea that the body affects the mind but not the other way around), to functionalism and multiple realizability (the idea that the same mental function can be instantiated in different physical systems), to embodied cognition (the idea that the mind is fundamentally shaped by the body’s interactions with its environment), among others.

Debates about the nature of consciousness grow out of the mind-body problem. Or better, the mind-body problem provides the conceptual framework within which discussions about consciousness occur. Not coincidentally, the same theories (epiphenomenalism, functionalism, embodied cognition, etc.) are applied to consciousness as well, because the nature of consciousness cannot be meaningfully discussed without considering how it relates to a specific bodily substrate.

Here is something new, however: the possibility that mental causation results in death—the demise of a living organism—introduces an altogether new dimension to this debate. It is no longer merely about the relationship between mind and body but becomes one about the relationship between mind and life. Logically speaking, death is the negation of life; only living entities can die. It follows that if we accept that consciousness can be involved in the dying process, theories of consciousness must account for the connection between the latter and life itself.

This specification significantly narrows the pool of candidate theories because many of these, including Chalmers' (1997, but also e.g. Dennett 1993; Edelman and Tononi 2001; Bostrom 2017), regard life as just another class of physical process that can be reducible to its structural and functional properties and that is fundamentally distinct from the ‘mental’, whether they agree that this exists (dualists) or not (physicalists). In any of these accounts, living organisms are analogous to machines. Despite the fact that the unique properties of life remain unaddressed in these accounts, or perhaps because of it, it has been relatively easy for these theorists to claim that a functioning machine can, in some sense, be considered ‘alive’, or better, that a living organism is like a functioning machine. But how, then, would these same theories define death?

If we are to follow their premises, death would simply be the breakdown of structure and function, the cessation of processes that sustain consciousness. One issue with this, however, is that, unlike machines—which can be repaired and restored to function—living beings cannot be revived after death. Granted, there exists a murky grey zone between life and death (incidentally, actual Haitian zombies are a great illustration here, Davis 1986) but in the grander scheme, death is final: once one is dead is dead. Brain scans can reliably confirm it, putrefaction is the ultimate proof. Death, and by extension life, is unique to organisms. Non-living entities, by their very nature and definition, cannot die, if metaphorically.

This consideration alone arguably rules out theories that do not account for the distinctive properties of life from explaining the relationship between consciousness and death. But there is another, related problem. If death represents the breakdown of structure and function, one that ends the process that sustain consciousness, the account we are left with is implicitly epiphenomenalist. It assumes that body affects mind, but not vice versa. ‘Voodoo death’, which defies epiphenomenalism, would be hard to reconcile within this picture. I suggest that explaining it compels us to consider a close relationship between consciousness and life.

In what follows I will examine enactivism and the free-energy principle as theories that have elaborated on the continuity between mind and life more than any other, and call them to make sense of the evidence just presented. I will show that they provide an explanation for ‘voodoo death’—thereby elevating their value as theories of consciousness—but also argue that ‘voodoo death’ reveals implications for these theories, and theories of consciousness more generally, that have rarely been contemplated, at least within the Western tradition.

Let’s start with enactivism.

Enactivism: death as the vanishing of affordances

The central guiding principle of enactivism is the idea that when we are studying a living being or the mind of a living being, such as the human mind, we are dealing with the properties of an organism, which differ in fundamental ways from the properties of a machine. Unlike machines, whose functions are specified by an external designer, organisms are defined by their autonomy. They are sources of their own being. Their purpose is not explicable in terms of external laws but only in terms of their own organizational structure, with metabolism serving as the baseline sustaining process. Most crucially, organisms are networks of dynamic systems (primarily metabolic but also sensorimotor, neural, immune, etc.) that manage to uphold an identity: to distinguish themselves from the environment while incessantly exchanging matter and energy with it. This process of self-maintenance and self-production in the face of entropy goes by the name of ‘autopoiesis’ (Maturana and Varela 1980).

Any encounter with the environment acquires meaning to the extent that it links up, either positively or negatively, with the norm of maintenance of the organism. This means that from the organism’s point of view the environment is essentially a range of affordances, which are defined as opportunities for action relative to the organism’s physical and perceptual capabilities (e.g. a chair affords sitting for a human, a fruit affords eating). In enactivist terminology, the ability to regulate and respond to changing conditions and evaluate affordances in ways that preserve the organism’s autonomy is defined as adaptivity (Di Paolo 2005).

Enactivists argue that mind is instrinsic to processes of life. The explanatory gap between consciousness (seen as internal and subjective) and life (seen as external and mechanical) that underpins the ‘hard problem’ is artificial because life already embodies an interiority—a self-regulating, sense-making, and purposive process—that cannot be reduced to external mechanism. Incipient in the self-organizing dynamics of the simplest organism, it blooms into consciousness and self-consciousness with increasing biological complexity.

Crucially, the continuity between mind and life is also considered to be phenomenological in character (Jonas 1966). That is to say: some basic concepts needed to understand life turn out to be applicable not just to cognition but to subjective experience itself. Mind is intrinsically life-like. The very feeling of being alive itself is deeply tied to, and emerges from, life-regulating metabolic processes. Similarly, the experience of temporal transcendence is grounded in metabolic activity, which imbues the organism with a forward-looking character. Enactivists argue that framing the engagement with the environment in terms of possibilities of actions also bridges biological and phenomenological domains. As Merleau-Ponty (1962) vividly suggests, our way of navigating the world is grounded in a pre-reflective ‘I can,’ a bodily sense of agency and ability to act that comes prior to reflective thought (i.e., the Cartesian ‘I think’). This framing applies equally both as a description of life and of subjective experience.

Continuity between mind and life does not mean a strict correspondence between the two. This is largely due to the evolution of the brain, which has massively complexified mental life. From an evolutionary perspective, it has been advantageous for the nervous system to achieve a degree of autonomy from the rest of the body. This autonomy has allowed the development of agency and other sensorimotor capacities that are not strictly tied to metabolism. In fact, the nervous system and its dynamic structures (or habits) can, to various degrees, decouple and even enter into conflict with metabolic values (habits, of course, can be bad). Yet, at a more fundamental level, the brain remains tethered to the task of maintaining the organism’s homeostasis, and it is at this deeper, foundational level that we observe mind-life continuity (Barandiaran and Moreno 2006).

It should be clear from the above account that mind for the enactivist is not reducible to the organ of the brain. It is embodied. It emerges from, and is interdependent with, the organism’s interwoven processes of self-regulation, sensorimotor coupling and inter-subjective coupling with the environment. The enactivist’s take on the mind-body problem is to argue, emphatically, that ‘mind’ and ‘body’ are not two distinct ontological categories, but two aspects of this ‘sense-making’ process. Because ‘mind’ and ‘body’ are two sides of the same coin, it is deceptive to say that mind affects body or vice versa (Varela et al. 1991).

Where enactivists make an important distinction is at the level of causes of experience.

The enactivist account of causality goes by the name of ‘circular causality’ (Fuchs 2017, 2021), which differs from ‘linear causality’ because any effect on the system is assumed to depend on, and affect in turn, the overall global organisation of that system. Proponents argue that physiological causes should be understood as ‘local-to-global’ (or ‘upward’) causes, whose effect on consciousness depends on the overall state the organism finds themselves in. For example, the role played by MDMA in heightening the activity of serotonin and boosting the feeling of ecstasy is of such local-to-global type, but where the effects fundamentally depend on the state of the person as a whole, on their history and social setting. By contrast, the role that, say, winning a trophy might have in heightening the activity of serotonin and boosting the feeling of ecstasy is a global-to-local (‘downward’) cause, where the global sense-making conditions of the organism affects lower-level physiological structures. Psychosocial causes are of this type. Importantly, physiological and psychosocial causes are always engaged in a relation of circular causality, where, depending on the case, one cause might take the leading role in affecting experience, but cannot be taken in isolation from the overall organism-world system.

The framework of ‘circular causality’ has been usefully employed to explore the causal nature of illness (De Haan 2020; Ongaro 2024) and, as we will see, death. How is the experience of illness understood within this perspective? From the enactivist perspective, illness is seen as a moment where adaptivity diminishes and the organism experiences a shrinkage of possibilities of action—a moment where the pre-reflective ‘I can’ falters. There is a contraction of the environment, which may result from different classes of causes.

In illnesses with a clear organic basis, such as tumors, the primary driver of the experience is typically upward in nature. However, how one perceives the illness (e.g., whether cancer is met with fear or acceptance) is fundamentally shaped by global sense-making processes, including culture and the social environment. Simply being diagnosed with the disorder, along with the negative expectations that come with it, can induce strong nocebo effects that amplify pain. In most cases of ‘mental’ illnesses and chronic conditions, downward causation is more dominant. Research shows for example that culturally induced negative expectations can pack a powerful neurobiological punch, leading to increases in stress hormones, catecholamines, cholecystokinin, and more (Bernstein et al. 2024). Nevertheless, for enactivists, these neurobiological manifestations should be considered neither ‘mental’ nor ‘bodily.’ Instead, they are the manifestation of downward causes acting on the organism. Unsurprisingly, they can, in fact, exploit the very same neural biochemical pathways of physical bottom-up perturbation (e.g. Benedetti et al. 2014), just as the effects of placebos (read: any therapeutic ritual that induces positive expectations) can activate the same neurobiological mechanisms as analgesic drugs (Piedimonte and Benedetti 2016). It is significant that, from the patient’s perspective, the downward effects of nocebos or the upward effects of physical perturbations are typically indistinguishable—and equally real.

What this suggests, the enactivists would say, is that the brain is fundamentally a ‘mediating organ’, whose primary function is to integrate streams of information coming from various sources, both upward and downward (Fuchs 2011). Consciousness is a process (not a thing) that emerges out of this integration, as the embodied organism interact with the environment. More specifically, it is a process of dynamic co-emergence, where the whole (the organism’s interaction with the environment) and its parts (local neural and bodily processes) mutually create and specify each other, rather than being reducible to each other (Thompson 2007).

What do make, then, of death—the moment where this process ceases for good? Death would correspond to the situation where adaptivity reaches zero, where we do not have a shrinkage of affordances, like in illness, but a vanishing thereof (Di Paolo 2009). All environmental conditions become unviable. The pre-reflective ‘I can’ that guides us throughout life capitulates; it turns into an “I can’t”. Like with illness, death too can result from a combination of upward and downward causes.

Under the account developed here, ‘voodoo death’ emerges as a type of death that results primarily from downward causes. It is an ultimate form of the nocebo effect, in which higher-level event, a change in the social environment around the individual, exerts a top-down causal influence on lower-level physical and biological properties, a cascade of neurological, immunological, and cardiac processes that lead to dying. While it is worthwhile to investigate the range of biological mechanisms involved in such deaths, as Cannon and many others have done, we can no longer think of these as sui generis mechanisms that can be turned on or off like a switch. Since the organism is a dynamical system, we should expect biological processes to vary in duration, intensity, and biochemical pathways depending on the conditions—just as nocebo effects can manifest in a wide variety of ways (Rossettini et al. 2023).

Notably, from a life-mind continuity perspective, these processes are also expected to reflect the phenomenological state of the individual. While a protracted fatalistic mood can contribute to death in a gradual manner, as suggested by studies showing the association between self-rated health and mortality, in the most striking cases of ‘voodoo death’, the total breakdown of affordances is subjectively felt by the individual as intense hopelessness. Engel described it as a situation in which, following an event that is perceived as stressful, the subject finds himself “unable to cope”, “feels utterly hopeless” and “has no expectation that any change in the environment is possible or will help” (cited in Hahn and Kleinman 1983:17). Enactivism provides a framework for understanding how this subjective experience and death might be causally connected.

What about non-human animals, one might ask? According to the life-mind continuity thesis, something akin to ‘voodoo death’ should also be expected in other animals, albeit in radically different forms. Evidence seems to bear out this expectation. To mention one example, in the 1950s, Curt Richter’s experiments revealed that wild rats experienced premature death when restrained and deprived of escape options. Trimming their whiskers, which are key sensory tools, would amplify their fear also lead to sudden death. Even calm, domesticated rats exhibited the same fearful behaviour and susceptibility to sudden death as wild rats when their whiskers were removed (Richter 1957). In all organisms, death occurs as the opportunities for action within an organism's environment shrink and vanish, whichever the primary cause.

In the human social environment, the downward effect leading to death is mediated through language and symbols. This raises the interesting possibilities that the idea of ‘death’ might be an embodied concept, that is, a concept grounded in the body’s physical, perceptual, interoceptive and sensorimotor experience, rather than existing as an abstract entity. Enactivists have long argued that some concepts are not just mental representations but are shaped by bodily experiences (in fact not only enactivists: Lakoff and Johnson 1980; Barsalou 1999). For instance, the way infants learn concepts like ‘up’ and ‘down’ is tied to their experience of moving and orienting themselves in space. If so, could ‘voodoo death’ also be explained by the fact that ‘death’ represents an embodied concept, one that bridges language with (the end of) life processes?

The enactive account of ‘voodoo death’ leaves us with some lingering questions. How can death be an embodied concept if we have never experienced it before? More generally, while enactivism arguably accommodates the phenomenon of ‘voodoo death’ through the life-mind continuity thesis, what can be said about the kinds of processes or cognitive architecture that might bring it about?

To arrive at some tentative answers to these questions, I will now turn to the other biogenetic theory on the agenda: the free-energy principle.

The free-energy principle: death as self-fulfilling prophecy

Touted as one of the most elegant and unifying scientific theories of the 21^st century, the Free-Energy Principle (henceforth FEP) is a big idea involving philosophy, biology, neuroscience, engineering, machine learning, with roots in statistical physics. It is at its core a theory of life—of how living beings, unlike any other entity in the known universe, can temporarily defy the second law of thermodynamics to fend off entropy and maintain a boundary between themselves and the environment. To do so, the theory posits, organisms must be able to minimize variational ‘free-energy’. This is a term that, without going into detail, represents an upper limit on the difference between an organism’s predictions about sensory input and the actual input it receives (a difference also known as ‘surprisal’). According to the FEP, living beings maintain their survival by minimizing the likelihood of encountering highly surprising or unexpected states, and by doing so they avoid the dissipating conditions that threaten their existence (Friston 2010). [1]

Some organisms have evolved nervous systems to get better at doing this. Nervous systems are also posited to follow the FEP, or at least to closely resemble it (see debates in Mendonça et al. 2021). As such, the FEP, like enactivism, offers a unifying theory of biology and cognition, life and mind, for the same principle that govern life are also considered to be a version of those that govern mind. [2] While the FEP is not a theory of consciousness itself, it is a theory for consciousness, which suggests that to understand conscious experiences—how they arise, what they are like, and how they relate to the self and the world—we must acknowledge their deep roots in the physiological processes that organize life itself (Seth 2022). Just like with enactivism, we have here a theory that erodes the central premise of the ‘hard problem of consciousness’: that the latter can be somehow divorced from living processes. Accordingly, it is a theory that I suggest is uniquely positioned to explain the relationship between consciousness and the process of dying.

Like enactivism, the FEP comes with a conceptual armoury that is usually laid out at the outset of every paper in the field. Succinctly put, the theory posits that the brain helps us make sense of the world by constantly predicting the sources of the (otherwise meaningless) barrage of inputs that hit the senses (see Clark 2016 for an excellent primer). It does so not by passively receiving and processing input, as earlier models suggested, but by actively anticipating it, producing a cascade of downward-flowing probabilistic predictions about the world. This neurally encoded stream of top-down predictions meets the bottom-up sensory signals. Any mismatch between prediction and actual sensory signals is converted into ‘prediction error,’ which propagates upward in the hierarchy to update the predictions, allowing the brain to progressively improve its ability to aid perception of the world. Some of these ‘priors’ act as stable background knowledge, accounting for the most temporally extended and spatially diffuse features of the world, while others are continually refined through learning. The main goal of the brain, according to FEP—what typifies it as a unifying theory of brain function—is the minimization of prediction error. This process also occurs in bodily action, which for FEP is simply the other side of perception: we predict something and act in a way that fulfils the prediction, a process known as ‘active inference.’ Some theorists see the FEP providing “an implementation of enactivism” at the neural level (Ramstead et al. 2021:59), for it lends further substance to the enactivist conception of ‘circular causality’ and autopoiesis (though this remains much debated; see Di Paolo et al 2022).

The same principle of prediction error minimization holds true for interoceptive states such as pain (Barrett and Simmons 2015). FEP is credited with offering persuasive explanations of phenomena that are difficult to accommodate theoretically, like chronic pain. But it has done so, in the first instance, by upending the understanding of the phenomenon that we thought was easiest to grasp: acute pain. The latter, for the FEP, is not a direct readout of sensory signal, but of a process that is invariably mediated by top-down predictions. When we are not in pain, the brain is continuously predicting our state of well-being. When the sensory signal goes beyond the predicted range of what is defined by higher priors as ‘well-being,’ predictions are quickly updated to the perception of pain. Of course, in the context of acute pain there is a tight correlation between lesion and pain perception because the latter, though determined by priors, is dominated by the sensory signal. In chronic pain, the situation reverses: here, it is highly precise top-down prediction of pain, even in the absence of lesion, that is responsible for generating the experience (Kaptchuk et al. 2020).

The framework allows us to get a clear picture of how nocebo effects work. It leads us to appreciate the salient role played by the social context in shaping and reinforcing predictions of symptoms. Knowledge of a drug’s side effects, verbal information about imminent pain delivered by a physician, or culturally specific ways of attending to the body can all heighten the strength of predictions about impending symptoms, leading these predictions to dominate symptom perception, and even engender symptoms without any physical, bottom-up cause. What’s more, through active inference, the brain might initiate visceral sensations (e.g., stomach tension) that match the hypothesis of being in pain. Aligning with conclusions reached by enactivists, the FEP shows that all symptoms are products of an inferential process that is never strictly reducible to physiological dysfunction and is sometimes only loosely related or even unrelated to it. ‘Explained’ and ‘unexplained’ (or ‘functional’) symptoms thus lie on a continuum, differing only in the extent to which they are coupled to an organic disorder. Given that the same inferential process is implicated in both cases, the theory also explains why so-called ‘real’ and ‘imaginary’ symptoms seem phenomenologically indistinguishable from the patient’s point of view (Ongaro and Kaptchuk 2019).

The reader might already see where the argument is heading, but before we get to ‘voodoo death’ directly, let me discuss an important aspect of interoception under the FEP framework. While the goal of exteroception is primarily to figure out what’s out there (for obvious survival reasons), the task of the interoceptive brain is not so much to accurately perceive the body’s state as to control it. Its goal is to maintain essential variables, like body temperature or heart rate, within the narrow ranges necessary for survival. To achieve this, interoception relies on strong, stable predictions about the body. Exemplifying the principle of active inference, interoception minimizes prediction errors by acting to fulfil top-down predictions rather than constantly updating them (though updates do occur). As a result, the brain systematically misperceives the body to create an illusion of stability. This ‘self-change blindness’ ensures we hardly notice constant shifts in our physiology, making regulation easier. Emotions, within this framework, reflect predictions about the body’s state in specific moments, while moods arise from longer-term regulatory patterns (Tsakiris and Preester 2019). Anil Seth (2022), among others, has argued that the feeling of ‘being you’—the most stable and invariant sense of self—stems from this interoceptive purpose of keeping the body seemingly stable. The key implication is that the conscious self is inextricably tied to the state of being alive (contrary to the intuition behind the ‘hard problem’).

With this in mind, we can revisit the studies showing a strong association between self-rated health and mortality discussed earlier. According to conventional wisdom, self-rated health predicts mortality because humans are equipped with the capacity to accurately perceive the state of their own health. Poor self-rated health supposedly captures disease lurking beneath the surface, which will eventually kill us (Jylhä 2009). In light of the FEP account of interoception, this interpretation no longer holds. (In fairness, this explanation also fails to account for why certain diseases remain asymptomatic until the brink of death, but that’s another issue.) The brain does not accurately perceive the body; instead, it constructs a perception of it based on predictions that are aimed at maintaining stability and control. These predictions, however, can be susceptible to psychosocial change, and when that happens, can drive the body to follow suit.

At once, this interpretation places the predictive power of perceived health on a continuum with the self-fulfilling prophecy of voodoo death, just at a different time scale. It is a death influenced by mood, rather than by intense emotion, with predictions of deteriorating health that are pre-reflective and stretched out over time, and which subtly guide the body toward breakdown and death. In line with these predictions, the body sets the stage for its own deterioration (the range of physiological mechanisms involved in this process would be interesting to investigate but are not necessary for this theoretical account). In essence, ‘voodoo death’ is simply an extreme manifestation of a common phenomenon: the entanglement of consciousness and the process of dying.

Conclusion

Bringing together all the threads woven thus far, we arrive at the following causal picture regarding death and its relationship to consciousness. As with pain, there exists a continuum depending on the degree to which experience is dominated by sensory signal or downward predictions. At one end of the spectrum, we find deaths primarily caused by external perturbations: falls from heights, car crashes, actual executions. Here the ‘sensory signal’ is an attack on the vital functions of the organism and the breakdown of the very conditions that make enaction and prediction possible. In these situations, downward causation has virtually no effect on the timing of death.

Moving further along the spectrum, there are deaths that are still primarily induced by a bottom-up physical cause but exacerbated by top-down prediction, particularly when the individual is aware of the lethality of the cause. HIV patients during the height of the AIDS epidemic, who were burdened by hopelessness and social stigma (both known to significantly affect immune function), illustrate this point. Researchers at the time recognised that psychological condition in HIV patients could influence mortality, while also noting that patients with strong social supports tended to survive longer (Cohen 1988). Further along still, we have cases where predictions play a critical role despite organic deterioration. For instance, low self-rated health reflects a mood that may act as a causal factor that hastens death (Newquist 1985; Dramé et al. 2023). At the opposite end of the spectrum, we encounter the more spectacular deaths driven almost entirely by top-down prediction, even in the absence of any discernible physical cause. These are the ‘voodoo deaths’ proper.

For any typology of death, it becomes an empirical question to determine where, along this continuum, the needle stands with regard to the power of downward causation. This consideration can also be extended to a demographic level. We might ask: what is the influence of downward causes in shaping average longevity in a population? My hunch is that the answer, if we ever uncover it, will surprise many.

Let me make a couple of assumptive examples.

We know that life expectancy has dramatically increased over the last 100 years across the globe. While the exact causes are still debated, it is undeniable that a major role was played by public health measures: vaccination, hygiene, new medical treatments, i.e. physical, ‘bottom-up’ causes. However, it must have been the case that, as the average longevity increased due to unprecedented medical wonders, the expectation of living longer than before also spread throughout the population, and that this expectation, in and of itself, also contributed to extend life expectancy. Equally, it must have been the case that, in societies where the post-infancy life expectancy was around 40 years, the prevailing expectation of dying around that age played a role in holding longevity in check.

Consider, as second example, longevity guru Bryan Johnson. An entrepreneur and rejuvenation Olympian, Johnson has invested a significant portion of his wealth into developing data-driven approaches to achieve what he calls ‘Blueprint’, a lifestyle designed to extend biological youth and, in his words, to “not die” (Johnson 2024). It involves a smorgasbord of daily supplements, sleep-optimization techniques, carefully selected workouts, the use of a hypobaric chamber, and more. Visibly, this arsenal of healthful measures does have an effect on his outward look. It will likely have it on his lifespan. However, it must be the case that his awareness of adhering to such life-extending regime, in and of itself, also plays a role in extending longevity. The degree to which this is the case would, again, be a matter for empirical testing (something that Johnson may not have planned to take into account). [3]

My broader point, going back to consciousness, is that the expectation to die is an inherent aspect of life and, therefore, of consciousness. And the expectation to die is inextricably tied with its bringing it about. Could this mean that, following the FEP formulation, we are endowed with a form of “death prior” that inevitably condemns us to our demise? One approach might be to argue, rightly, that FEP only describes the behaviour of organisms at a given point in time. As such, even if we accept the story presented in this paper, active inference about death would only emerge in old age, when predictions of dying lead the organism to act accordingly, both behaviourally and interoceptively, or in particular situations where affordances break down (as in ‘voodoo death’ proper). Not, that is, at the stage infancy. Death then could only become a concern later in life rather than being an integral constituent of consciousness from the get-go.

But there is another dimension to consider. Organisms are products of ontogeny as much as phylogeny. We carry the imprints of our evolutionary past within us. Proponents of the FEP argue that it also offers a framework that unifies the short-term processes of development within an organism’s lifetime (ontogeny) with the long-term processes of evolution—how species adapt over generations (phylogeny). It achieves this by modeling both as processes of probabilistic inference: evolution (phylogeny) furnishes the generative models (encoded in genotypes) that constrain and shape the adaptive behaviors and learning of individuals (ontogeny). These generative models are refined over evolutionary timescales but are expressed and tested in real-time through an organism’s actions and interactions with its environment (Friston et al. 2023) From an evolutionary perspective, then, the fact that, past the age of reproduction, our body lowers its defenses and gives way to deterioration could be seen as a manifestation of a latent “death prior.” This perspective also lends some plausibility to the idea that the concept of ‘death’ could be a deeply ingrained ‘embodied concept,’ one we possess without needing direct experience of it. Could we not, then, conceive of death as an inherent dimension of life rather than merely the state that follows it?

In Beyond the Pleasure Principle (1975[1920]), Sigmund Freud argued that human behaviour is not solely driven by the pursuit of pleasure and the avoidance of pain but also by deeper, unconscious forces like the ‘death drive’, which invites individuals to a return to the inorganic state. There is a duality, Freud thought, between life instincts (Eros) and death instincts (Tanathos). Leaving aside issues of compatibility between FEP, enactivism and psychoanalysis (Holmes 2022), my argument, here, will serve as a analogous corrective to the common refrain, prevalent in both FEP and enactivist circles, that consciousness arises out of a ‘drive to survive’ (Nave 2025) or that, as Seth puts it, “as long as we live, the brain will never update its prior belief of expecting to be alive” (Seth 2022:194). This framing overshadows the idea that conscious life is inherently shaped by, and carries with it, the anticipation of death.

That life and death are two sides of the same coin is, of course, a leitmotif that we find in many philosophical traditions outside the West. If for Epicurus “as long as we exist, death is not present”, other traditions have seen things differently. Some have thought of death as an inherent aspect of living as a philosophical first principle. In Tibetan Buddhism, for instance, dying is thought to take place at every living moment, in the sense that the dissolution of every conscious moment represents a microcosm of our final dissolution. Death, in this view, necessarily belongs to the structure of consciousness itself (Thompson 2014). In the Zhuangzi, similarly, we read that life derives part of its value from its relationship to death, and vice versa (Liu 2022). In Haitian voodoo, too, though as different from both Tibetan Buddhism and Daoism as one can imagine, we encounter the idea, reflected in ritualistic practices, of life and death as entwined dimensions, where life isn’t interrupted by death but transformed by it (Metraux 1972). The Western tendency to view death as a discrete stage that follows life is a culturally and historically specific one.

There is an evident genealogical connection within the Western philosophical tradition between the ‘substance dualism’ about life and death and the ‘substance dualism’ about physical processes and subjective experience that underpins the ‘hard problem of consciousness’. Fleshing it out may be a matter for another time. The purpose of the paper has been to show that the life-mind continuity perspectives challenges both perspectives simultaneously. Regarding the latter—the dualism of the ‘hard problem’—it exposes philosophical zombies as a metaphysical impossibility. It supports the idea that “our conscious experiences of the world around us, and of ourselves within it, happen with, through, and because of our living bodies” (Seth 2022:180). It follows—to dabble into one of the hottest conversations of our times—that current AI systems are unlikely to ever support consciousness because they do not possess autopoietic system that typifies life (Seth 2025). On the former subject of mortality, I noted that if consciousness is part of our nature as living beings, it must also be tied to the fact that we are mortal beings—from which it follows that for AI to be conscious it will also have to die. ‘Voodoo death’ served as that anomaly whose explanation compels us to accept a form of life-mind continuity. ‘Voodoo death,’ however, is simply a boundary case. It is the tip of the iceberg that invites us to discover that, in the vast majority of cases, an embodied anticipation of dying plays a part in bringing it about. Consciousness may be intrinsic to life because it is also inextricably tied up with death.

Bibliography

Adler, S. R. (2011). Sleep paralysis: nightmares, nocebos, and the mind-body connection. New Brunswick, N.J: Rutgers University Press.

Andrews, M. (2021). The math is not the territory: navigating the free energy principle. Biology & Philosophy, 36(3).

Au, N., & Johnston, D. W. (2014). Self-assessed health: What does it mean and what does it hide? Social Science & Medicine, 121, 21–28.

Barandiaran, X., & Moreno, A. (2006). On What Makes Certain Dynamical Systems Cognitive: A Minimally Cognitive Organization Program. Adaptive Behavior, 14(2), 171–185.

Barrett, L. F., & Simmons, W. K. (2015). Interoceptive predictions in the brain. Nature Reviews. Neuroscience, 16(7), 419–429.

Barsalou, L. W. (1999). Perceptual symbol systems. Behavioral and Brain Sciences, 22(4), 577–660.

Benedetti, F., Durando, J., & Vighetti, S. (2014). Nocebo and placebo modulation of hypobaric hypoxia headache involves the cyclooxygenase-prostaglandins pathway. Pain, 155(5), 921–928.

Bernstein, M., Blease, C., Locher, C., Brown, W. A., & Mayo Clinic (Eds.). (2024). The nocebo effect: when words make you sick. Rochester, MN: Mayo Clinic Press.

Bostrom, N. (2017). Superintelligence: paths, dangers, strategies. Oxford: Oxford University Press.

Cannon, W. B. (1942). “Voodoo” death. American anthropologist, 44(2), 169–181.

Center for Environmental Health & Injury Control. (1988). Update: Sudden Unexplained Death Syndrome Among Southeast Asian Refugees—United States. JAMA: The Journal of the American Medical Association, 260(14), 2033.

Chalmers, D. (1995). Facing up to the problem of consciousness. Journal of Consciousness Studies, 2(3), 200–19.

Chalmers, D. (2018). The Meta-Problem of Consciousness. Journal of Consciousness Studies 25 (9-10):6-61

Chalmers, D. (1997). The conscious mind: in search of a fundamental theory. New York: Oxford University Press.

Clark, A. (2016). Surfing uncertainty: prediction, action, and the embodied mind. Oxford; New York: Oxford University Press.

Coates, C. F. (2006). Vodou in Haitian Literature. In Vodou in Haitian Life and Culture (pp. 181–198). New York: Palgrave Macmillan US.

Cohen, S. I. (1988). Voodoo death, the stress response, and AIDS. Advances in biochemical psychopharmacology, 44, 95.

Davis, W. (1986). The serpent and the rainbow. London: Collins.

De Haan, S. (2020). Enactive Psychiatry. Cambridge ; New York, NY: Cambridge University Press.

Dennett, D. C. (1993). Consciousness Explained.London: Penguin.

Di Paolo, E. (2009). Extended Life. Topoi, 28(1), 9–21.

Di Paolo, E. (2005). Autopoiesis, Adaptivity, Teleology, Agency. Phenomenology and the Cognitive Sciences, 4(4), 429–452.

Di Paolo, E., Thompson, E., & Beer, R. (2022). Laying down a forking path: Tensions between enaction and the free energy principle. Philosophy and the Mind Sciences, 3.

Dramé, M., Cantegrit, E., & Godaert, L. (2023). Self-Rated Health as a Predictor of Mortality in Older Adults: A Systematic Review. International Journal of Environmental Research and Public Health, 20(5), 3813.

Eastwell, H. D. (1982). Voodoo death and the mechanism for dispatch of the dying in East Arnhem, Australia. American Anthropologist, 84(1), 5–18.

Edelman, G. M., & Tononi, G. (2001). Consciousness: how matter becomes imagination. London: Penguin Books.

Engel, G. L. (1971). Sudden and rapid death during psychological stress: folklore or folk wisdom? Annals of internal medicine, 74(5), 771–783.

Freud, S. (1975[1920]). Beyond the pleasure principle. New York: Norton.

Friston, K. (2010). The free-energy principle: a unified brain theory? Nature Reviews Neuroscience, 11(2), 127–138.

Friston, K., Friedman, D. A., Constant, A., Knight, V. B., Fields, C., Parr, T., & Campbell, J. O. (2023). A Variational Synthesis of Evolutionary and Developmental Dynamics. Entropy, 25(7), 964.

Fuchs, T. (2011). The Brain -- A Mediating Organ. Journal of Consciousness Studies, 18, 196–221.

Fuchs, T. (2017). Ecology of the Brain: The phenomenology and biology of the embodied mind. Oxford, United Kingdom: OUP Oxford.

Fuchs, T. (2021). In defense of the human being: foundational questions of an embodied anthropology. Oxford, United Kingdom ; New York, NY, United States of America: Oxford University Press.

Hahn, R. A., & Kleinman, A. (1983). Belief as Pathogen, Belief as Medicine: “Voodoo Death” and the “Placebo Phenomenon” in Anthropological Perspective. Medical Anthropology Quarterly, 14(4), 3–19.

Holmes, J. (2022). Friston’s free energy principle: new life for psychoanalysis? BJPsych bulletin, 46(3), 164–168.

Idler, E. L., & Kasl, S. (1991). Health Perceptions and Survival: Do Global Evaluations of Health Status Really Predict Mortality? Journal of Gerontology, 46(2), S55–S65.

Johnson, B. (Ed.). (2024). Zeroism. Venice: Zero.

Jonas, H. (1966). The Phenomenon of Life: Toward a Philosophical Biology. New York: Harper&Row.

Jylhä, M. (2009). What is self-rated health and why does it predict mortality? Towards a unified conceptual model. Social Science & Medicine, 69(3), 307–316.

Kananen, L., Enroth, L., Raitanen, J., Jylhävä, J., Bürkle, A., Moreno-Villanueva, M., et al. (2021). Self-rated health in individuals with and without disease is associated with multiple biomarkers representing multiple biological domains. Scientific Reports, 11(1), 6139.

Kaptchuk, T. J., Hemond, C. C., & Miller, F. G. (2020). Placebos in chronic pain: evidence, theory, ethics, and use in clinical practice. BMJ, m1668.

Lakoff, G., & Johnson, M. (1980). Metaphors we live by. Chicago, IL: University of Chicago.

Leach, J. (2018). ‘Give-up-itis’ revisited: Neuropathology of extremis. Medical Hypotheses, 120, 14–21.

Lester, D. (2009). Voodoo Death. OMEGA - Journal of Death and Dying, 59(1), 1–18.

Liu, P. (2022). Death in the Zhuangzi : Themes, arguments, and interpretations. Philosophy Compass, 17(4), e12825.

Lown, B. (1999). The lost art of healing (First Ballantine ed.). New York: Ballantine Books.

Maturana, H. R., & Varela, F. J. (1980). Autopoiesis and cognition: the realization of the living. Dordrecht Boston London: D. Reidel Publishing Company.

Meador, C. K. (1992). Hex death: voodoo magic or persuasion? Southern Medical Journal, 85(3), 244–247.

Meador, Clifton K. (2005). Symptoms of unknown origin: a medical odyssey. Nashville: Vanderbilt University Press.

Mendonça, D., Curado, M., & Gouveia, S. S. (2021). The philosophy and science of predictive processing. London: Bloomsbury Academic.

Merleau-Ponty, M. (1962). Phenomenology of perception. Abingdon, Oxon ; New York: Routledge.

Metraux, A. (1972). Voodoo. Aylesbury, UK: Sphere Books Limited.

Moerman, D. E. (2002). Meaning, medicine, and the “placebo effect.” Cambridge ; New York: Cambridge University Press.

Nave, K. (2025). A drive to survive: the free energy principle and the meaning of life. Cambridge, Massachusetts London: The MIT Press.

Newquist, D. D. (1985). Voodoo death in the American aged. Cognition, stress, and aging, 111–133.

Ongaro, G. (2024). Outline for an Externalist Psychiatry (1): Or, How to Fully Realize the Biopsychosocial Model. Philosophy, Psychiatry, & Psychology, 31(3), 269–284.

Ongaro, G., & Kaptchuk, T. J. (2019). Symptom perception, placebo effects, and the Bayesian brain: PAIN, 160(1), 1–4.

Phillips, D. P., & King, E. W. (1988). Death takes a holiday: mortality surrounding major social occasions. Lancet (London, England), 2(8613), 728–732.

Phillips, D. P., Liu, G. C., Kwok, K., Jarvinen, J. R., Zhang, W., & Abramson, I. S. (2001). The Hound of the Baskervilles effect: natural experiment on the influence of psychological stress on timing of death. BMJ (Clinical research ed.), 323(7327), 1443–1446.

Phillips, D. P., Ruth, T. E., & Wagner, L. M. (1993). Psychology and survival. Lancet (London, England), 342(8880), 1142–1145.

Phillips, D. P., & Smith, D. G. (1990). Postponement of death until symbolically meaningful occasions. JAMA, 263(14), 1947–1951.

Piedimonte, A., & Benedetti, F. (2016). Words and drugs: Same mechanisms of action? Journal of Contemporary Psychotherapy: On the Cutting Edge of Modern Developments in Psychotherapy, 46(3), 159–166.

Ramstead, M. J. D., Kirchhoff, M. D., Constant, A., & Friston, K. J. (2021). Multiscale integration: beyond internalism and externalism. Synthese, 198(S1), 41–70.

Richter, C. P. (1957). On the phenomenon of sudden death in animals and man. Psychosomatic Medicine, 19(3), 191–198.

Rossettini, G., Campaci, F., Bialosky, J., Huysmans, E., Vase, L., & Carlino, E. (2023). The Biology of Placebo and Nocebo Effects on Experimental and Chronic Pain: State of the Art. Journal of Clinical Medicine, 12(12), 4113.

Samuels, M. A. (2007). “Voodoo” death revisited: the modern lessons of neurocardiology. Cleveland Clinic Journal of Medicine, 74(Suppl_1), S8–S8.

Seth, A. (2022). Being you: a new science of consciousness (Paperback edition.). London: Faber.

Seth, A. (2025). Conscious artificial intelligence and biological naturalism. Behavioral and Brain Sciences, 1–42.

Shoemaker, S. (2001). Realization and mental causation. In C. Gillett & B. Loewer (Eds.), Physicalism and its Discontents (Vol. 9, pp. 23–33).

Sternberg, E. M. (2002). Walter B. Cannon and “Voodoo Death”: a perspective from 60 years on. American Journal of Public Health, 92(10), 1564–1566.

Thompson, E. (2007). Mind in life: biology, phenomenology, and the sciences of mind. Cambridge, Mass.: Belknap Press of Harvard Univ. Press.

Thompson, E. (2014). Waking, Dreaming, Being: Self and Consciousness in Neuroscience, Meditation, and Philosophy. New York: Columbia University Press.

Tsakiris, M., & Preester, H. de (Eds.). (2019). The interoceptive mind: from homeostasis to awareness. Oxford, United Kingdom: Oxford University Press.

Varela, F. J., Thompson, E., & Rosch, E. (1991). The embodied mind: cognitive science and human experience. Cambridge, Mass.: MIT Press.

Wuorela, M., Lavonius, S., Salminen, M., Vahlberg, T., Viitanen, M., & Viikari, L. (2020). Self-rated health and objective health status as predictors of all-cause mortality among older people: a prospective study with a 5-, 10-, and 27-year follow-up. BMC Geriatrics, 20(1), 120.

Yawger, N. S. (1936). Emotions as a cause of rapid and sudden death. Archives of Neurology and Psychiatry, 36(4), 869–890.

Zheng, J., Zheng, D., Su, T., & Cheng, J. (2018). Sudden Unexplained Nocturnal Death Syndrome: The Hundred Years’ Enigma. Journal of the American Heart Association: Cardiovascular and Cerebrovascular Disease, 7(5), e007837.

[1] Some authors have questioned the compatibility between FEP and autopoiesis (Di Paolo et al. 2022). While I remain agnostic on this issue, there are some important parallels between the two, which I hope will become clear below. What I highlight here is FEP’s value in accounting for the anticipatory nature of life and mind.

[2] A caveat, here: born as a piece of mathematical philosophy, the FEP is not a scientifically falsifiable theory, which can be evaluated through hypothesis testing. It is better described as a formal ‘modelling framework’ (Andrews 2021). Nevertheless, there are theoretical offshoots of the FEP for which its tools take on epistemological and methodological force. Predictive processing (PP) is a family of models, nested within the FEP framework, that uses these tools to make predictions about brain function and cognition. While mindful of these distinctions, I hope I may be forgiven for using ‘FEP’ throughout this paper to refer to both the framework and its theoretical applications, for the sake of conciseness.

[3] Be as it may, this effect should never be interpreted as proof of the “power of the will”, as positive psychologists often claim, typically in favour of individual self-help remedies while oblivious to socio-structural factors. Top-down predictions are very heavily affected by social context and structural conditions.