Here I present a unitary biological explanation for two of humankind’s most costly and least understood afflictions: psychopathology and suicide.

Both are in urgent need of theoretical progress. Psychopathology (or mental/psychiatric disorders/illness, interchangeable terms for this discussion) presents in recurring patterns worldwide—major depression, substance use disorders, schizophrenia, and so on—directly affecting more than a tenth of the global population (Rehm & Shield, 2019). Suicide (the act of deliberately killing oneself; W.H.O., 2014), also ubiquitous, takes the lives of some 700,000 people around the world annually, accounting for 1.3% of human deaths (W.H.O., 2021). Despite decades of research, a consensus scientific explanation of either phenomenon remains elusive. The “failure of psychiatry to make significant progress in understanding the aetiology of mental disorders” (Abed & St John-Smith, 2021, p. 24) is matched by equally slow headway in suicide research, which remains “still in a preparadigmatic phase” (Franklin et al., 2017, p. 188).

That both psychiatry and suicidology face an impasse may be no mere coincidence: the two domains may be missing a shared meta-theoretical key, which this article provides.

Psychopathology and Suicide Connect

Beyond question, suicide and psychopathology are linked (Ferrari et al., 2014; Schechter & Goldblatt, 2020). The connection is close, systemic, and trans-diagnostic, as evidenced by at least three epidemiological findings. First, virtually all psychiatric disorders associate with an increased risk of suicide, raised more than tenfold for some diagnoses (Favril et al., 2022; Ferrari et al., 2014; Fu et al., 2021; Harris & Barraclough, 1997; Moitra et al., 2021; Yoshimasu et al., 2008). Too et al.’s (2019) meta-analysis of record linkage studies, for example, reports pooled risk ratios of 13.2, 12.3, 8.1, 4.4, and 4.1 for psychotic, mood, personality, substance use, and anxiety disorders, respectively. Second, psychological autopsy studies find that many (in most studies, most) people who die this way qualified for a diagnosable mental illness at the time of death (Arsenault-Lapierre et al., 2004; Cavanagh et al., 2003; Cho et al., 2016). Depression features prominently but, contrary to a common belief, is not associated with the majority of suicides; a meta-review found that major depression and other mood disorders accounted for fewer than 1 in 3 of diagnoses present (Bertolote et al., 2004). Third, although varying in strength, the link prevails the world over (Ferrari et al., 2014; Knipe et al., 2019). Cho et al.’s (2016) meta-analysis, for example, finds psychopathology present in 90.4%, 88.2%, and 69.6% of suicides in South Asia, USA, and East Asia, respectively.

How and why they are linked is a longstanding puzzle. In theory, suicide risk (S) and psychopathology (P) could join via several possible causal routes. Plausibly, P → S; or S → P; or another variable (X) could produce both, so that X → (S, P). Many ideas have been floated (Mishara & Chagnon, 2016; Tanney, 1992) and there could be a mix of relationships. Empirically, a confluence of methodological, epistemological, and other problems has stymied the determination of causation by experimental or other means. A notable epidemiological curiosity is that psychopathology associates with suicide almost entirely inasmuch as it associates with suicidal ideation—it only weakly predicts the transition of suicidal ideas into plans and attempts (Klonsky et al., 2021; Nock et al., 2016).

Despite the uncertainty, one causal path has gained widespread credence: supposedly, P → S. Beyond explaining suicide in a statistical sense, psychopathology is often presumed to be “responsible”—“underpinning suicide” (Ferrari et al., 2014, p. 1, 10; italics added). Granted, there is a folk-psychological tendency to attribute suicide to mental illness (Atkinson, 1978). Perhaps this intuition is not wholly unfounded: certain symptoms of severe psychopathology can become so unbearable as to make life feel questionably worthwhile, as some sufferers testify (Fredriksen et al., 2017). However, the folk-psychological stance may have been adopted less than critically in some scientific circles. P → S is an implicit assumption behind calls to reduce suicide via the diagnosis and treatment of purportedly “underlying psychiatric disorders” (Wasserman et al., 2012, p. 137; italics added). Occasionally, the assumption is explicit, as in “• suicide is caused by depression (premise 1)” (Isacsson et al., 2010, p. 429). Notwithstanding such confidence, the putative causal mechanism of P → S is seldom explained, and the presumption that there is such a mechanism is open to scientific challenge. At least 1 in 10 suicides occur with no discernible psychopathology (Cavanagh et al., 2003), and the great majority of people with even the most serious mental disorders do not take their own lives (Fu et al., 2021; Nordentoft et al., 2011). Diagnostic risk factors do not predict suicidal behavior much better than chance (Franklin et al., 2017), and psychiatric issues rarely emerge in qualitative studies as a prominent causal driver (Bennett et al., 2023; Hjelmeland & Knizek, 2016). There are also anomalies in responses to treatment: psychopathology can be reduced without a reduction in suicidality, and vice versa (Goldsmith et al., 2002). In sum, despite the notion’s widespread acceptance, the evidence does not support P → S as the primary line of causation.

The pain-brain theory of suicide and human evolution (Soper, 2018, 2021) advances an arguably better-evidenced connection, that suicide and psychopathology have a largely shared etiology: X → (S, P). X here combines two inputs: psychological pain (or psychache; Shneidman, 1996); subject to a level of developmental maturation linked to intellectual competence. To precis the argument, self-induced death is a highly effective way for cognitively capable humans to escape psychache. However, by extinguishing prospects of further reproduction, death is a disaster for biological fitness. A powerful reproductive advantage accrued to the offspring of individuals with heritable protections against that disaster, thereby driving the evolution by natural selection of special-purpose anti-suicide defenses. Because suicide was recurrently preceded by chronic psychache, conditional on cognitive maturity, an evolved defensive system uses these variables as inputs for assessing suicide risk and, at times of danger, to trigger emergency interventions that block suicidal actions. These countermeasures present as certain psychiatric symptoms, which combine with psychache and side effects to form familiar heterogeneous syndromes. In an ultimate sense, S → P: suicide, as a recurring threat to human survival, explains why distressed adolescents and adults experience mental disorders and why psychopathology characterizes our species. Proximately, the human organism’s need to avoid suicide underlies a mutual etiology: X → (S, P), with P substantially functioning to prevent suicidal deaths.

The evidence, theoretical and empirical, that suicides may happen despite, not because of, certain psychiatric phenomena is at least intriguing, arguably considerable, and to some researchers persuasive (Lester, 2019; Maris, 2022; Shackelford, 2021; Wong, 2022). It is not a new or outlandish proposal. By training and clinical experience, psychiatrists are wary of “an early increase in suicide risk as depressive symptoms begin to lift” (American Psychiatric Association, 2003, p. 61), a caution backed by epidemiological evidence that such symptoms may impede the acting out of suicidal ideas (Rogers et al., 2018; Rufino et al., 2022; Stanley et al., 2018). Anti-suicide functionality has long been attributed to diverse other psychiatric presentations, all strongly co-morbid with depression, including manic states (Himmelhoch, 1988), psychotic delusion and hallucination (Hundert, 1992; Murphy et al., 2018), eating disorders (Himmelhoch, 1988), substance addictions (Himmelhoch, 1988; Menninger, 1938; Solomon, 1982), and non-suicidal self-injury (Simpson, 1976). The last of these, an evidenced method of regulating negative affect, is sometimes done purposefully to suppress suicidal urges (Herzog et al., 2022; Miller et al., 2021).

So, could it be true? Could certain psychiatric symptoms obstruct the path from suicidal thought to suicidal deed—and, indeed, be displays of psychological mechanisms that evolved to perform that task? If so, it would potentially call for a paradigmatic reappraisal of research, policy, and practice in suicide prevention and mental health. Hence, this article presents the evidence for critical review. The following sections discuss the pain-brain theory and its relevance to psychopathology, the inputs and outputs of a hypothesized evolved anti-suicide system, and how those outputs may form psychiatric syndromes. Later sections offer predictions generated by the theory that are open to testing, make comparisons with other models of psychopathology, and review clinical implications. The concluding section highlights reciprocating opportunities for advancement in suicidology and psychiatry: for suicidology to better integrate evolutionary biology; and for evolutionary psychiatry to better integrate suicide.

Pain-Brain Theory: Suicide as a Biological Problem

The pain-brain theory highlights suicide as a severe and recurring fitness hazard in our species's evolutionary environment (Soper, 2018, 2019a, 2021). Deliberate self-killing poses a major biological problem for an animal that can mentally represent the possibility. For humans, as with mammals generally, death by any cause is disastrous for the organism’s genotype—its genetic code: death nullifies prospects both of further reproduction and of further investment in any existing offspring (Duntley, 2005). From this perspective, suicide is “the ultimate contraceptive” (Shneidman, 1985, p. 37). However, for the human phenotype—the physical living organism—suicide is at the same time an outstandingly effective answer to pain (Murray & Kluckhohn, 1948). Pain demands an answer: a necessary concomitant of animal life, enabling the organism to navigate its internal and external environment, pain is designed precisely to induce action to relieve or end the aversive stimulus (Klein, 2015). Since “[n]othing hurts less than being dead” (Humphrey, 2018, p. 3), any pained animal that could elect death would expectably seize the opportunity.

Not every animal has the opportunity. Suicide is as Baechler (1979) surmised, “a privilege of fully self-conscious human beings” (p. 38): only humans deliberately kill themselves, and only cognitively mature humans at that, because the act presupposes intellectual capability. Notwithstanding disconcerting talk, behavior, and occasional self-injury, suicide is vanishingly rare among very young children (Borges et al., 2012) and adults with even moderate intellectual disability (Patja, 2004). The behavior is non-existent among nonhuman animals, as far as can be known (Bering, 2018). These protected populations presumably lack the high-level self-awareness and other cognitive faculties needed to mentalize their own mortality and the actions needed to bring that mortality about (Soper, 2018). Human intelligence is nonetheless functional; it evolved because of the fitness benefit, probably social, that it gave our ancestors (Humphrey, 1976), despite the consequent cost of suicide.

Managing that cost is a severe and uniquely human problem. From the above account, suicide arises as a by-product of two primary adaptations that come together in our species and ours alone—pain and “brain.” Pain motivates escape, and human brainpower presents the means of effecting that escape via self-induced death (Humphrey, 2018; Soper, 2018, 2021). The two ingredients are separately benign but lethal in combination. If an actor has both the motivation to commit a certain act and the means to do so, then that act will expectably be committed, unless some obstacle blocks the way. In this light, suicide emerges not as a mystery or an aberration but as the unsurprising, default outcome of human sapience.

If this analysis is accepted, then its corollary has to be accepted too: solutions to the adaptive problem of suicide must have evolved, or we would not be here to discuss the matter. Special-purpose psychological machinery operates in species-typical humans, intervening to block suicides that would otherwise occur. The question is not whether such defenses exist, but what are they? How do they work, and what are their manifestations?

Multiple tiers of anti-suicide defenses likely operate, commensurate with the lethal severity of the threat, in much the same way that layers of defenses protect the organism from attack by pathogens. By close analogy, the physiological immune system mobilizes in emergencies, responding to infections with life-saving but highly costly interventions. Prevention being better than cure, permanently active defenses lie ahead of the immune system to stop mortal danger from arising in the first place: skin and mucus membranes keep most pathogens safely at bay. Conceptually, the thicker the skin, the less the burden on the immune system.

Analogous to skin, a bulwark against deadly crises, the human psyche is posited to feature forward lines of permanently active defenses against suicide. Some ("brain-type fenders"; Soper, 2018) propagate an encultured fear of death and a special aversion to the idea of suicide. Others, (“pain-type”; ibid) are evolved psychological devices that promote a subjective sense of wellbeing and meaning-in-life, and manage the organism’s exposure to potentially suicidogenic distress. Since only negative affect motivates escape, the organism faces an extreme asymmetry of risks. By this light, a suite of intriguing cognitive illusions—loss aversion (Kahneman & Tversky, 1979), confirmation or “myside” bias (Mercier, 2017), unrealistic optimism (Sharot, 2011), religiosity and spiritual belief (Rosmarin et al., 2020), and related phenomena—may be understood as on-average fitness-serving responses to a human survival necessity. Overriding more ancient animal fitness concerns, such buffers meet situations that threaten negative affect (or “unpleasure,” “dissonance,” and other near-synonyms) by avoiding, denying, or reconstruing reality in the interest of softening the emotional blow. These front-line shields usually fulfill their task, keeping most of us, most of the time, adequately happy, hopeful, and motivated to live (Humphrey, 2011, 2018; Soper, 2018, 2021; Wong, 2022).

However, there is a cost to being thick-skinned. Pain serves a purpose: it alerts the organism to fitness threats—in the case of psychological pain, mainly threats in the social environment—and guides the organism’s adaptive response. Insensitivity entails the forfeit of potentially useful information. The human organism has to reconcile the conflicting demands of suicide avoidance and accurate cognizance, walking a line between too much pain and too little. It is a formidable and complex challenge, and the arrangement of solutions will vary from person to person and at different times over the lifespan.

According to the theory, when life becomes so distressful that front-line defenses are overwhelmed—when other fitness concerns cannot be overridden, and the resulting protracted psychache may make the idea of lethal escape start to appeal—defenses of last resort (“keepers”; Soper, 2018) will mobilize. Their task is to stop the individual from acting on suicidal thoughts. This final line of defense does not switch off the front line: it takes over when the front line fails, trying to save the day. Analogous to the physiological immune system, the last line of anti-suicide defense is all that stands between life and death, and its countermeasures present as pathology.

The above proposal controverts a pervasive but scientifically groundless premise in suicidology, that suicide transgresses a so-called “survival instinct” or “instinct for self-preservation” (eg., Bennett et al., 2023; Joiner et al., 2016; Lengvenyte et al., 2021). Notwithstanding the idea's millennia-old heritage (Joiner, 2010), modern evolutionary biology offers no mechanism by which a general-purpose, free-floating “survival instinct” could come about (Kirkpatrick & Navarrete, 2006; Soper, 2019b). We will come back to this point. Nor will the human organism rely on pre-human adaptations that evolved to prevent accidental self-harm (Soper, 2022). The adaptations that evolved to prevent suicide, although uniquely human, will be like all other evolved cognitive and behavioral mechanisms—i.e., special-purpose systems that respond to specific fitness-relevant information with specific fitness-promoting outputs (Tooby & Cosmides, 2005).

Explanatory Scope of the Theory

The emergency safeguards that deploy to stop chronically distressed individuals from acting on suicidal ideas are posited to display in a suite of overlapping, strongly comorbid, mental disorders that make up much of the caseload of adolescent and adult psychiatry. On empirical grounds, included are adult-pattern major depression, non-suicidal self-injury, obsessive/compulsive disorders, eating disorders, substance and behavioral addictions, bipolar disorder, at least some personality disorders, and psychoses. This array encompasses the various psychopathologies previously hypothesized by Himmelhoch (1988), Hundert (1992), and other clinicians as having “last gasp” anti-suicide functionality ("Psychopathology and Suicide Connect"). The conditions fall into a loose category of “functional” (as opposed to “organic”) mental disorders, having “no diagnosable pathophysiology sufficient to account for the etiology of the symptoms” (Bell et al., 2020, p. 3).

Not all aspects of psychopathology are explained by this framework. Three riders should be noted.

First, generalized anxiety and other anxiety disorders, although often co-occurring with the above diagnostic categories, likely manifest a different class of response—an ancient mammalian strategy for dealing with severe-but-ambiguous fitness threats, as will be discussed further below.

Second, neurodevelopmental and other organic disorders, having their own distinct etiologies, are also excluded (Goldstein, 2001).

Third, to be clear, I do not argue that major depression, schizophrenia, or any other diagnosis is an evolved adaptation, whether for preventing suicide or any other biological purpose. Most diagnostic categories in psychiatry are pragmatic, heterogeneous constructs—more “practical kinds” (Zachar, 2014, p. 153) than natural kinds, thus likely unamenable to a purely biological explanation. The current theory’s primary explanandum is not diagnoses but certain constituent symptoms that recur under different diagnostic labels, either as shared diagnostic criteria or by comorbid association. Acknowledging that diagnostic schemas—the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) (American Psychiatric Association, 2013a) and the mental health section of the International Classification of Diseases (W.H.O., 2018)—are psychiatry’s current “law of the land” (Sadock et al., 2015, p. vii), this article borrows their terminologies and extends bridges to them. However, the connections are at best messy, a mismatch that lays the discussion open to cross-purposes. The pain-brain framework leads to a different understanding of psychopathology and a different basis for its nosology.

With the above caveats, it is argued that trans-diagnostic patterns of human psychopathology manifest emergency anti-suicide functionality with such precision that it would be an extraordinary coincidence if these responses to chronic psychache were not displays of adaptations that evolved to fulfill the hypothesized purpose.

The proposal combines elements of the new and old in psychiatric thought. New: the focus on functional components is conceptually not dissimilar to the US National Institute of Mental Health’s (2023) RDoC initiative, which calls for research into domains of neuro-behavioral processing instead of DSM-style diagnostic labels. Old: the umbrella notion of psychopathology recalls a time before the reign of the DSM, when “psychiatrists did not take diagnosis very seriously” (Paris, 2013, p. 45). Menninger et al. (1958), for example, advanced a unitary conception in which “mental illness” (in the singular) varies in severity, not kind. Paris (2022b) takes a comparable trans-diagnostic stance in current psychiatry, using the term “psychopathology” instead of “mental disorders.”

Likewise cutting across diagnostic categories, the suggestion made here is that the more costly an individual’s symptoms, the higher the organism’s assessment of suicide risk that warranted that response. Longer-lasting psychiatric conditions indicate that the organism detects a commensurately more persistent danger. The theory conceives anti-suicide mechanisms as discrete routines, variable in intensity but functionally indivisible, combining in familiar syndromic clusters.

The proposed emergency suicide-prevention system is visualized in Fig. 1 in three parts, from left to right: a symptom-general process, a symptom-specific process, and the formation of identifiable but heterogeneous psychiatric syndromes. The following three sections explain each part in turn.

Fig. 1
Fig. 1
Full size image

Pain-brain etiology of psychiatric phenomena. Left panel (Symptom general): suicide risk and proportionate anti-suicide defenses arise from a shared pain-brain etiology, setting up a general vulnerability to psychopathology. Centre panel (Symptom specific): separate factors relating to personal differences determine the response’s final psychiatric presentation, conceived as personalized blends of “pain-type” and “brain type” interventions. Right panel (Syndromes): heterogeneous psychiatric syndromes constitute often-observed combinations of anti-suicide responses, alongside psychache and sundry side effects. For full commentary, see "Symptom-General “Pain” and “Brain” Inputs", "Symptom-Specific “Pain” and “Brain” Outputs" and "Syndromes" sections in main text

Symptom-General “Pain” and “Brain” Inputs

Two adaptive inputs, “pain” and “brain,” when combined, produce two distinct but related outcomes (Fig. 1, left panel): a costly, maladaptive by-product—a heightened risk of suicide; and, to manage this cost, a secondary adaptive output—the activation of a proportionate protective response, which displays as non-specific psychopathology.

“Pain” Input

The “pain” input is quantitative. The more intense the pain, the greater the impetus to escape it; and the longer-lasting the pain, the greater the opportunity to plan and organize a suicidal escape. Hence, a singular input, algorithmically merging pain’s intensity and chronicity, is expected to operate as an internal regulatory variable (Tooby et al., 2008) for the assessment of suicide risk and the release of proportionate countermeasures. The system disregards pain’s qualitative particulars, all pain collapsing onto a single dimension of affective-motivational power (Auvray et al., 2010). Psychological pain figures more than physiological pain on the grounds of being not categorically distinct, but quantitatively more unbearable, (MacDonald & Leary, 2005). Thus, any configuration of adversities can lead to an anti-suicide response.

If this plank of the pain-brain theory is correct, life stressors in general should produce mental disorders in general. There is little empirical evidence to the contrary (Goldberg & Goodyer, 2005; Kessler, 2000; Paris, 2022a), with three provisos.

First, as noted earlier, anxiety symptoms are probably not uniquely a suicide-avoidance measure. Protracted anxiousness is a prototypical animal response to severe-but-indeterminable threats (Blanchard et al., 2001) of which, for the human animal, suicide may be viewed as an extreme example: mortal danger looms, but from no particular direction. There being nowhere to run, a good strategy is for the organism to put itself on high alert, and to remain in that state until it is sure the danger has passed. On this basis, lasting anxiety should co-occur with elevated suicide risk and consequent suicide-prevention responses. However, anxiety may be precipitated by other ambiguous hazards, unconnected with suicide and the current model. It may be for this reason that the etiology and epidemiology of generalized anxiety only partly overlap with those of other common mental disorders (Goldberg & Goodyer, 2005; Kendler et al., 2003; Kessler, 2000).

Second, certain depressive symptoms may have common-sense validity as plausible reactions to certain kinds of adversity, such as sadness following bereavement and fatigue from chronic stress (Keller & Nesse, 2006). When asked to explain their symptoms, people self-report reasons that usually align with these expectations, at least at a subsyndromal level (Keller et al., 2007). However, such intuitions are epistemically unreliable as a guide to the etiology of psychiatric syndromes. There is a long history of folk-psychological ascriptions of psychiatric complaints to special causes, going back to 19th-century “railway spine” and 20th-century “shell-shock,” which have not stood the test of time (Horwitz, 2018). With the aforementioned exception of anxiety states, there remains little epidemiological evidence of specific causation for most functional psychiatric disorders.

Third, the experience of psychache is not solely dependent on exposure to stressors: individual predisposition, linked to personality traits and other aspects of individual difference, moderates the aversive impact. Vulnerability to psychopathology is probably caused by a complex interplay of stress and diathesis factors (Eisenbarth et al., 2019; Healy et al., 2022). The p factor—a unidimensional measure of susceptibility to psychopathology (Caspi & Moffitt, 2018)—emerges in light of the pain-brain theory as a psychache factor: for reasons that may be genetic, environmental, or (most likely) an interaction of the two, some individuals’ lives are subjectively experienced as more distressful than others, therefore more suicidogenic, and therefore more prone to emergency suicide-prevention responses.

The hypothesized quantitative nature of the “pain” input implies a dose–response pattern, which is indeed seen (McKay et al., 2021). The more numerous and intolerable the preceding adversities, the greater the suicide danger and the more drastic the psychiatric effect, albeit sub-additively, the strength of the association weakening with increasing numbers of events (Jenkins et al., 2015).

The “pain” input works both ways. Mirroring activation, psychiatric disorders usually remit given time and benign conditions, with or without medical intervention (Brown, 2009; Goldberg & Goodyer, 2005). However, reflecting the severity of suicide as a fitness threat, deactivation is cautious: delayed, and then episodic and/or gradual. Remission from even severe mental illness is more the rule than the exception, but it is rarely prompt or sudden (Álvarez-Tomás et al., 2019; Lally et al., 2017; Whiteford et al., 2013).

Psychopathology associates with suicide, both outcomes precipitated by psychache. However, more narrowly, the predicted association is between psychopathology and psychache-induced suicidal ideation. Since psychopathology’s protective components are hypothesized to block the planning and enacting of suicidal ideas, psychopathology should not associate with such escalations (Soper, 2018)—nor empirically, as already noted, does it appear strongly to do so (Klonsky et al., 2021; Nock et al., 2008).

“Brain” Input

The “brain” input is qualitative. At a certain stage of cognitive maturation, usually reached in adolescence, the concept of intentional self-killing becomes intellectually conceivable, plannable, and actionable (Soper, 2018). This ontogenetic Rubicon is a living fossil of an ancient phylogenetic transition: suicide emerged as a recurring survival hazard when an ancestral hominin encephalized to the extent of being able to mentally represent its mortality (Stengel, 1964). The shift would have been marked and momentous because the suicide problem sets up a cliff edge in the fitness function of intelligence. Above a certain level of general intellectual competence—presumably about that of a modern prepubescent child—incremental intelligence for our ancestors would have been self-destructive (Soper, 2019a, 2021). Hence, what displays as “a cognitive ‘floor’ for suicide” (Perry, 2014, p. 110) in our species may impose a ceiling on others. Other animals are set well back from the hazard, as is generally the adapted response to fitness cliff edges (Vercken et al., 2012). How a population of ancestral Homo came to approach, and then overcome, the fitness cliff edge of suicide is a fundamental but neglected puzzle in human evolutionary science (Soper, 2019a, 2021); but clearly, overcome it they did.

The “brain” input is expected to control the developmental scheduling of anti-suicide responses, meeting the arrival of suicide as an intellectually accessible course of action. This dynamic plausibly explains why first onsets of adult-pattern psychopathology peak in the teen years (McGrath et al., 2023), approximately synchronous with peak first onsets of suicidal behavior (Borges et al., 2012). It does not suggest that younger children are immune from psychiatric problems, but it does predict a difference in the prevalence and nature of psychopathology with first onsets before, compared to after, this transitional stage (Weiss & Garber, 2003).

Suicides as Statistical Residuals

Despite the entwined etiology outlined above, the origins of suicide and psychopathology are expected to diverge in at least one important respect: deliberate self-killing will be precipitated by a conscious decision, psychopathology by a neurochemical activator, and the latter will only approximate to the former. The organism has to assess suicide risk based on imperfect information—on a correlational best fit, computed cumulatively over an evolutionary timescale, between suicidal outcomes and the preceding patterned information available to the organism in its internal and external environment. The central nervous system cannot “read” directly the plot content of human thoughts—whether these concern an anticipation of, and a determination not to tolerate, future suffering (“pain”), or the mentalizing of suicidal escape (“brain”)—except insofar as such representations carry affective associations. The organism’s suicide-prevention system is programmed to respond to the best available indicator of despair: an archetypal emotional “pain” precursor. It cannot intervene if this on-average prior is missing, as may be the case for impulsive suicidal acts and, at the other extreme, exits organized with meticulous calm. Psychopathology’s “brain” input will also be an actuarial approximation: an endocrine developmental switch, enabling anti-suicide defenses to activate, will serve as a proxy for the attainment of the intellectual threshold beyond which suicide is meaningfully thinkable.

The system’s reliance on imprecise data makes it vulnerable to being blindsided. At least 1 in 10 suicides occur with no associated psychopathology (Cavanagh et al., 2003) plausibly because these deaths are preceded by no sign stimulus that could have triggered protective countermeasures. Such non-activations, together with a larger number of deaths where defenses under-reacted, may explain a persisting level of suicide, a “basic biological substrate” (Goldney, 2003, p. 145), that characterizes human populations. This base rate can be taken as evidence not of the human psyche’s faulty operation but of the complexity of its task—to maintain an intellect so promiscuous that the individual can, unfortunately, mentalize self-induced death as a solution to pain, but usually without acting on that thought. Given the lethality of the threat, the system will more often err on the side of over-reacting to cues of danger, inflicting psychopathology on many people who were never going to take their own lives. These errors are not dysfunctions; they reflect the presence of a cost asymmetry and actuarial uncertainty (Haselton et al., 2009).

The evolved suicide-prevention algorithm is expected to exhaust the predictive utility of any “pain” and “brain” information that is perceptible to the organism. The system could not, and clearly does not, eliminate the possibility of suicide: costly to activate, defenses have evolved only up to a compromise position where the fitness benefit of an incremental reduction in suicide risk is cancelled out by the fitness cost of the measures required to obtain it. However, by acting on available informational inputs, the system will eliminate actionable risk factors. Suicidal deaths constitute statistical residuals, scenarios tagged with no utilizable correlates with which the organism could have anticipated and headed off the lethal outcome (Soper, 2019b). It is understandable in this light that no risk or protective factors in any known combination predict suicidal outcomes much beyond chance (Belsher et al., 2019; Franklin et al., 2017). Tellingly, more information does not help: adding complexity to risk assessment models does not result in a useful improvement in accuracy (Corke et al., 2021). The non-additivity observed indicates that the source of uncertainty is aleatory, rather than epistemic.

This biological assessment of suicide as an intrinsically out-of-the-blue phenomenon challenges another tenet of received wisdom in suicidology: that the behavior’s non-predictability can be explained as an emergent statistical property of its rarity (eg., Carter et al., 2017; Mulder et al., 2016; Rosen, 1954). Bolton et al. (2015), for example, state that “[t]he positive predictive value (PPV) of any risk tool that assesses an event with a low base rate in the population, such as suicide, is likely to be low even when sensitivity and specificity values are high” (p. 3). Although valid, this argument is circular—it is not a substantive explanation. It begs the question as to why assessment is conducted at the level of a gross population with a low base rate, instead of focusing on a sub-population with a high one (Soper, 2019b). The substantive explanation is that such a sub-population is not available. Intense natural selection over an evolutionary timescale is very unlikely to have passed over any selection criterion, perceptible to the organism, that indicated high risk. If the pain-brain analysis is correct, almost the opposite to “rare, therefore non-predictable” may be true. Suicide looks rare because it is by nature non-predictable: it eventuates as a rarified, stochastic dispersal across an undifferentiable population because selection has left no usefully predictive marker with which to narrow the field of vision to a sub-group with a high base rate.

The point can be viewed from another angle. If psychopathology and other evolved machinery regulates suicide risk homeostatically or homeorhetically, as Himmelhoch (1988) proposed, then we should not expect to find a direct mechanistic link between potentially suicidogenic inputs and suicidal behavioral outputs. This biological decoupling lends support to critics of “linear cause-and-effect thinking” (Hjelmeland & Knizek, 2016, p. 697) that they see dominating current suicide research (Marsh, 2016). Such thinking is far from new (Soper, 2019b): a scientific hunt for hypothetical special conditions that supposedly lead to suicide has gone on at least since the time of Morselli (1881) and Durkheim (1897)—as yet with no sign of success (Franklin et al., 2017). Pain-brain theory suggests that the long-standing null result is no mere technical issue. The action of evolved psychological defenses, a causal confound, gives principled grounds for anticipating that such conditions are very unlikely to exist. Wishful thinking aside, it may not be realistic to view the discovery of a usefully accurate predictor of suicide as a scientific “challenge”, as it is often depicted in the suicidology literature (eg., O’Connor & Portzky, 2018, p. 4). Predictably, despite population-level actuarial associations, very few people who display even the strongest risk factors (including so-called “warning signs” and “red flags”) will kill themselves, and the absence of such factors will not confer immunity. By dint of being human, virtually all of us carry more-or-less the same minimal, but above zero, risk of taking our own lives (Soper et al., 2022).

Symptom-Specific “Pain-type” and “Brain-type” Outputs

The symptom-general inputs of Fig. 1 (left panel) produce an anti-suicide response of a severity that is proportionate to the organism’s assessment of suicide risk. How that response presents in symptom-specific outputs (Fig. 1, center panel) is determined by a separate set of factors pertaining to individual difference—including age, sex, personality traits, learned inputs, and life history—the origin of which may be again genetic, environmental, or most likely, a gene–environment interaction. Much of the variation may be stochastic.

Each sufferer will be served with a customized cocktail of involuntary psychological interventions that are designed to stop suicidal ideas from materializing into lethal actions. Responses are matched to individual needs because multiple trade-offs are involved: the biological challenge is to obstruct lethally self-destructive trajectories while minimizing the inevitable collateral damage to the individual’s social and other fitness-serving faculties. The least-bad compromise solution will vary from person to person and at different times of life. Thus, not everyone who is equally susceptible to psychopathology will be equally susceptible to all of its forms.

Only a limited range of ingredients are available for the cocktail because, as the pain-brain framework indicates, the organism has two, and only two, suicide-prevention strategies available: it can make the act unnecessary—weakening the power of psychache to motivate lethal escape (pain-type); and/or it can make the act difficult—withholding the cognitive and psychomotor resources needed to conceive, plan, and carry out deliberate self-killing (brain-type). Many, perhaps most, commonplace psychopathological symptoms can be interpreted as one or other of these two classes of response.

Nonetheless, within the two strategies, many tactical variations are possible. With pain-type defenses, the organism has multiple placebo-like neurological gates and psychological mechanisms available with which to adjust the motivational potency of pain to suit current contextual needs (Melzack & Wall, 1965; Wall, 1999). There may be a generalized emotional blunting, volitional deficit, and a loss of interest in activities, as often seen in major depression, bipolar disorder, and schizophrenia. Their effect, observed from clinical (Hendin, 1975) and epidemiological (Rufino et al., 2022) perspectives, is to dampen the impetus for suicidal escape—and, incidentally, for any other similarly demanding action. The organism can also put to involuntary use any of a range of psychological interventions that people suffering chronic pain conditions intuitively turn to for relief. By one such approach, pain may be endured (Eccleston, 2001): people soldier on with the aid of distractions, “pain-inhibits-pain” behaviors (easing one pain by applying another, unrelated stimulus; Moont et al., 2010), and/or by self-medicating with psychotropic substances and behaviors. This style of coping may account for the frequently comorbid symptoms of (respectively) obsessive–compulsive disorders, non-suicidal self-injury, and substance and behavioral addictions. At the extreme, when the real world becomes too painful to encounter, the human mind can construct an alternative reality—perhaps distressing in its own way, but at least provisionally habitable. The human capacity for delusion, apparently mapping universal existential dilemmas (Collin et al., 2023), may be as Hundert (1992) surmised, an evolved strategy for avoiding willful self-destruction. As long as these various pain-type defenses hold, there is no need for suicidal escape.

Manic states, posited by Himmelhoch (1988) as a different style of emergency anti-suicide response, plausibly manifest an alternative pain-relief strategy: activity engagement (Viane et al., 2004). Extreme activity can bring respite, perhaps analogous to the tolerance athletes achieve when pushing themselves beyond a pain barrier (Addison et al., 1998), sometimes to the point of euphoria (Siebers et al., 2021). It is difficult to see how an individual could be both manically active and depressively inactive at the same time, but the two methods of coping could be adopted at different times as a bipolar sequence. In diagnostic terms, it is rare for an individual to meet the full criteria for manic and major depressive episodes simultaneously (American Psychiatric Association, 2013b), although other, mainly anxiety-related, symptoms of hypomania (agitation, distractibility, etc.) often co-occur with depressive symptoms, constituting “mixed states” (Barroilhet & Ghaemi, 2020).

As for brain-type defenses, these will disrupt the individual’s ability to conceive and carry out a suicidal endeavor—and, incidentally, any other similarly involved undertaking. Exigent suicide risk will be met with targeted denials of memory, logical thinking, decision-making, and other high-level cognitive resources that are required for the planning and organizing of complicated tasks. Such deficits occur across psychotic and affective disorders (Millan et al., 2012). Psychomotor losses may play a role. With regard to psychotic hallucination, Murphy et al. (2018) suggest that the capacity to externalize suicidal thoughts, constructing the inner threat as a command from outside, may create distance from such thoughts and help to mobilize a defensive response.

Whether pain-type or brain-type, the anti-suicide measures outlined above are probably not an exhaustive list; the two strategies may well present in other ways.

The proposed functional connection between commonplace psychiatric symptoms and organismic suicide prevention is not proven, but it is a reasonable inference. Its plausibility is strengthened by cross-diagnostic features of psychopathology that are consistent with anti-suicide functionality and, as a set, are difficult otherwise to explain: involuntariness, association with anxiousness, abnormality, autoimmunity, universality, and species-specificity.

Involuntariness. Analogous to immune responses—commensurate with a threat that is too serious for the organism to permit defenses to be electively overridden—anti-suicide responses will be strongly compulsive. Blind to the instinctual processes involved (Cosmides & Tooby, 1994), we will have little if any scope to tamper with them.

Association with anxiousness. Symptoms frequently associate with protracted anxiety (Goldberg & Goodyer, 2005)—high vigilance that is in keeping, as discussed earlier, with the deadly-but-ambiguous nature of suicide as a fitness threat.

Abnormality. Overriding normal emotional (“pain”) and cognitive (“brain”) fitness-serving routines in the interests of survival, output behaviors are conspicuously costly. In the technical sense of the word, they constitute “abnormal” psychology (Plante, 2013). Unsurprisingly, for this reason, mental illness is easily construed as dysfunctional and harmful (Wakefield, 1992). The pain-brain analysis challenges this assessment on both counts. It suggests that common forms of mental disorder are usually not dysfunctions but displays of functional routines, biologically designed to preserve survival at times of grave danger. Nor are they usually harmful: that is, not harmful in comparison to the potentially fatal result had they not intervened. Analogous to the fever, sickness behavior, and other unwanted responses of the physiological immune system, anti-suicide interventions can be unpleasant, debilitating, and socially deviant. An easy mistake is to compare these highly visible states to those of “normals”—and thereby perceive costs, overlook benefits, and infer net harm. No doubt, when judged against the general population, psychopathology displays seriously maladaptive consequences, including elevated rates of suicide (Favril et al., 2022), higher all-cause mortality (Walker et al., 2015), and lower fertility (Keller & Miller, 2006). The paradox that heritable mental disorders persist despite these associations (Keller & Miller, 2006) resolves on recognizing the faulty basis of the comparison: if the current theory is correct, the countervailing benefit may reside, invisibly, in self-killings averted. The correct ceteris paribus comparator is a hypothetical control group that is equally at risk of suicide but not protected. The adaptive payoff is invisible because suicide and defenses against suicide co-evolved deep in human prehistory: a control group does not exist.

Autoimmunity. Beyond mere abnormality and the base-rate of actuarial errors discussed earlier ("Suicides as Statistical Residuals"), a certain level of harmful dysfunction proper may be inevitable in this domain. Grave danger calls for drastic measures, which may sometimes backfire. Comparable to autoimmune disease, suicide-prevention defenses may have iatrogenic effects, “producing other types of maladaptation” (Himmelhoch, 1988, p. 46). Psychopathology can sometimes induce vicious cycles of stress generation (Rnic et al., 2023). Usually, however, psychopathology should follow, rather than precede, adversities and only rarely be degenerative.

Universality. Suicide being a species-typical adaptive problem, similar organismic solutions in the form of human psychopathology should be observable in all parts of the world and all historical eras, as appears the case (Plante, 2013; W.H.O., 2018). However, universality does not imply uniformity. The discussion so far suggests that psychiatric presentations may be influenced by learned inputs, and by several possible routes: via the selection of pain-relieving distractions, for example, or via the goals pursued in activity engagement. The selection of psychotropic substances and behaviors for the compulsive management of psychache necessarily depends on prior exposure, presumably at least partly social. Analogous to ethological imprinting, a formative mood-altering experience from gambling, eating, dieting, drinking alcohol, ingesting narcotics, self-cutting, and suchlike encounters will guide later involuntary recourse to the same behavior when the organism assesses the danger of suicide to be pressing. Cultural patterns of psychopathology may arise from such mechanisms.

Species-specificity. Anti-suicide defenses should not be found in nonhumans. Other mammals, primates especially, may well display rudimentary components of human-pattern psychopathology, vestiges of phylogenetic raw material from which (human) anti-suicide adaptations evolved. Such ancient precursors might help to explain intriguing molecular analogues of psychiatric disorders found in other species (eg., Dutcher et al., 2023; Gyles et al., 2023). And, no doubt, nonhuman animals may be prone to their own forms of psychological ill-health (Troisi, 2003). However, the species-typical psychopathology of humans appears to be categorically different from that of other animals in prevalence, range, complexity, comorbidity, and other dimensions. Although translational researchers point to atomized “schizophrenia-like,” “depression-like,” and “mania-like” animal responses (Nestler & Hyman, 2010, pp. 1164, 1166, 1167), these analogues fall short of grounds for labelling a nonhuman clinically “schizophrenic,” “depressive,” or “manic.” Animal models have consistently failed to predict psychiatric treatment outcomes, not least because of acknowledged “difficulties of modelling disorders that often seem uniquely human” (p. 1161). Centrally to this discussion, there is no direct animal model of suicide (Gould et al., 2017).

Syndromes

Psychiatric syndromes (Fig. 1, right-hand panel)—familiar, discernible, heterogeneous constellations of symptoms—are conceived as commonly observed combinations of anti-suicide symptoms, psychache, and sundry side effects of both. In the symptomatology of major depression, for example, anti-suicide responses may contribute psychomotor retardation, a loss of interest in activities, and cognitive deficits; psychache may present as low mood and feelings of worthlessness or guilt; and side effects may include sleep disturbance, irritability, and weight change. Symptoms may interact, producing complex networks of cause-and-effect (McNally, 2016). However, limiting the interactivity, a syndrome’s anti-suicide elements will be in response to historic, chronic (rather than current) psychache (see "“Pain” Input").

The same protective components may reoccur in different syndromes: psychomotor retardation, for example, features in major depression, schizophrenia, and bipolar disorder (American Psychiatric Association, 2013a); delusion and hallucination across a similar range of disorders (Baethge et al., 2005). Cognitive impairment is a frequent concomitant of, inter alia, major depression and obsessive compulsive disorders (Millan et al., 2012). Symptoms will blend in familiar clusters for the practical reason that some suicide-prevention intercessions work better together than others. Sufferers may switch between equivalently effective, but mutually incompatible, configurations. These may display in bipolar sequences, as discussed earlier, or in the transfer of addictions from one pain-relieving substance to another depending on availability, or to behavioral substitutes (Hirschman, 1992).

Complexity is implicit in Fig. 1. Individual differences (Fig. 1, center panel) mediate how the organism’s assessment of suicide risk (left-hand panel) feeds through into final psychiatric presentations (right-hand panel), with many interacting factors and a vast array of possible permutations of outcomes. Syndromes are heterogeneous both between persons and within the same person over time (Oquendo et al., 2004). Although statistical patterns of inputs and outputs will be visible at a population level, the cost–benefit trade-offs and mechanistic processes of particular psychiatric presentations may be inscrutable even in principle. Nonetheless, the notion of recurring clusters of compatible anti-suicide responses offers a coherent way to understand why psychiatric syndromes occur, why they associate with suicide risk (Cavanagh et al., 2003), and why they link to dimensions of individual difference (Goldberg & Goodyer, 2005).

The Theory Is Overall Falsifiable

The above framework is work in progress. It appears to be internally coherent and broadly consistent with the empirical evidence, but it is not without problems.

One problem is that it is difficult to see how, beyond a probability judgment based on intuitive reasoning, the functional roles of the psychiatric phenomena concerned can be ascertained. Whether a particular symptom is a display of anxiousness, or a display of psychache, an anti-suicide response to psychache, a side effect, or several or none of the above, may not be determinable by experimental means. This uncertainty is not disproof; it may be an inconvenient facet of the complexity of the human psyche and pose a general problem for evolutionary analyses of psychopathology, perhaps by necessity “ahead of the data” (Kennair, 2011, p. 471). However, it lays this particular aspect of the theory open to a charge of unfalsifiability.

In other respects, the theory is falsifiable but (to my knowledge) not entirely evidenced. Put another way, it generates testable predictions. Contrary to a widespread misbelief, evolutionary explanations can be powerful tools not only for making post hoc sense of what is already observed but for predicting patterns yet unseen (Al-Shawaf, 2020). Four predictions are offered here.

  1. (i)

    Developmental correlates. Since hormonal variables are posited to constitute the organism’s proxy “brain” cue, indicating that the intellectual capability for suicide has arrived ("Suicides as Statistical Residuals"), suicide’s and psychopathology’s first onsets in adolescence should display distinct associations (Gunn et al., 2021). Earliest serious plans and attempts will associate with intellectual, rather than endocrine, development; earliest onsets of adult-pattern psychopathology will be the other way around. The differences may be fine, but in principle ought to detectible.

    As an aside: regarding the suicide-prevention system’s “pain” input, defenses are hypothesized to activate in response to a unitary, as yet unspecified, neurochemical cue. If a marker for this posited activator were identified, it might be possible to devise a test that predicts the onset and severity of psychopathology, albeit nonspecifically—exact presentations may not be open to usefully accurate prediction even in principle. Such a discovery would have major clinical and ethical implications.

  2. (ii)

    Sequencing of suicidal ideation and psychopathology. Because chronic (not just acute) psychache is likely a prerequisite for the triggering of anti-suicide responses, suicidal ideation should tend to precede, more than follow, first onsets of adult-pattern mental illness. There is tentative evidence along these lines with regard to psychosis (Murphy et al., 2018), a temporal pattern that should be visible generally across psychiatric disorders.

  3. (iii)

    Neurodegenerative conditions and suicidality, psychopathology. Although suicidality may well intensify around the time of diagnosis and early progression of Alzheimer’s disease and frontotemporal dementia, reflecting the distressing nature of these diseases, a precipitous fall in serious suicide attempts, approaching zero, should be discernible beyond a threshold of cognitive degeneration, mirroring suicide’s patterned emergence in childhood. If functional psychopathology is as suggested, driven by neurochemical rather than cognitive inputs, then it ought not to show a simultaneous sudden decline.

  4. (iv)

    Chronic pain conditions and psychopathology. A history of chronic pain, especially in early life, should be a generalized risk factor for onsets of psychopathology. As a reviewer of this article pointed out, if the organism’s assessment of suicide risk and the activation of psychopathology are driven by the experience of protracted pain, regardless of the pain’s qualitative details, then chronic physiological pain should be a non-specific risk factor, albeit a less strong one than psychological pain. There is only incomplete evidence of such a trans-diagnostic effect. Connections with chronic pain conditions are well documented for mood and anxiety disorders, associations that are likely complex and apparently bidirectional. However, as the pain-brain framework would anticipate, pain proceeds more than follows onsets and relapses of these disorders (Roughan et al., 2021). There is less evidence of a more generalized link, possibly because few studies have focused on other diagnostic categories (Birgenheir et al., 2013; Docherty et al., 2023).

If these predictions were confirmed, it would support the current theory while contraindicating the rival hypothesis, discussed earlier, that psychopathology causes suicide.

Is There a Better Theory of Psychopathology and Suicide?

With due reservations, evolved anti-suicide functionality as conceptualized by the pain-brain theory may be the most promising available framework for understanding mental disorders and their association with suicide, for at least five reasons.

First, since the existence of multiple connections between mental disorders points to a substantially shared etiology (Marshall, 2020), a unifying meta-theory such as pain-brain has the advantage of parsimony over the many diagnosis- and symptom-specific hypotheses that currently dominate evolutionary psychiatry (Brüne, 2016; Del Giudice, 2018). Mental disorders’ numerous common features, some already mentioned, require a simultaneous explanation: such as their peak first onset in the teen years (McGrath et al., 2023); non-specific precipitation by life adversities (Paris, 2022a); non-specific prodromal antecedents (Ortiz-Orendain et al., 2023); a common course, usually easing spontaneously given time and benign conditions (Goldberg & Goodyer, 2005); cross-diagnostic symptomatology (Baethge et al., 2005); via the p factor, a common genetic substrate (Selzam et al., 2018); shared absence of specific biomarkers (Munkholm, 2021); comorbidity—extreme, pervasive, and enduring (McGrath et al., 2020); consistent, only modestly effective, response to treatments (Leichsenring et al., 2022); “lack of treatment-specificity is the rule” (First & Pincus, 2009, p. 107); and so on. Regarding the last of these features, the placebo-like effectiveness of therapeutic “common factors” is a remarkable trans-diagnostic curiosity (Wampold & Imel, 2015), comprehensible through the lens of the pain-brain model: if psychopathology is understood as an anti-suicide response to psychache—usually social pain, stemming from relational setbacks—then almost any credible intervention that entails a therapeutic relationship and imparts hope could be expected to expedite recovery (Soper, 2021). In sum, as Nesse (2019) observes, “it is time to give up on the fantasy that each mental disorder has a specific cause” (p. 27).

Second, the few trans-diagnostic models that have been previously proposed fail to account for psychopathology’s characteristic symptomatology. Some models address certain dimensions of comorbidity, inspired by life history theory or invoking genetics and gene–environment interactions (Crespi & Badcock, 2008; Del Giudice, 2020; Uher, 2009). Others look to explain variations in general susceptibility to psychopathology (Albott et al., 2018; Belsky & Pluess, 2009; Caspi & Moffitt, 2018; Keller & Miller, 2006). None of them explain why, in or after adolescence, a chronically distressed human being should become motivationally blunted, compelled to ingest a pain-relieving substance or perform a mood-altering behavior, preoccupied with an arbitrary concern, delusional, hallucinatory, cognitively prevented from organizing complicated tasks, or frantically active. The pain-brain framework uniquely accounts for these peculiar responses to protracted psychache: singly or more usually in combinations, they plausibly manifest diverse ways in which human organism makes suicidal thoughts unnecessary and/or impracticable to act upon, at times of heightened risk. They often occur alongside the protracted restlessness, agitation, and mental rumination that speaks of the organism having detected a grave threat—which, given the prospect of suicide, may be a prudent assessment.

Third, the theory provides a framework for integrating other cross-diagnostic models. It aligns with an emerging consensus on the part-general, part-specific, polygenic origins of common mental disorders (Lynall et al., 2022; Selzam et al., 2018). Overarching genetic risk factors (Keller & Miller, 2006) connect to the left-hand panel of Fig. 1, while life history and other factors that may influence final presentations (Del Giudice & Haltigan, 2021) connect to the center and right panels. The theory meets all four requirements of an integrated evolutionary framework for psychopathology as set out by Del Giudice and Haltigan (2021), addressing (a) patterns of comorbidity and overlap between disorders, (b) heterogeneity within diagnostic categories, (c) connections between pathologies and aspects of individual differences, and (d) developmental patterns across the lifespan.

Fourth, the theory is biologically comprehensive. It is an improvement in that respect on other prominent theories of suicide (Soper, 2019b, 2022). Encompassing proximate and ultimate, dynamic and synchronic, modes of biological understanding, it explains the link between mental disorder and suicide from the perspective of each of Tinbergen’s (1963) “four questions”—adaptive value, phylogeny, ontogeny, and mechanism:

  • The adaptive value of psychopathology lies in preventing suicidal ideas from escalating into suicidal actions.

  • The phylogeny of psychopathology tracks ancestral human encephalization—the high-level cognitive intentionality and general intelligence that, in conjunction with psychache, by-produced suicide as an adaptive problem.

  • The ontogeny of psychopathology echoes that ancient phylogenetic transition; the scheduling of earliest onsets, in adolescence, meets the species-typical development of suicide as a cognitively representable escape.

  • The proximate mechanism of psychopathology operates in two parts: a general activation by the pain-brain cues of suicide risk, with other aspects of individual differences determining final psychiatric presentations.

Finally, although some other proposals address psychopathology’s link with suicide (Mishara & Chagnon, 2016; Tanney, 1992), the current theory explains this link’s specific configuration—why psychopathology associates strongly with suicidal ideas, but only weakly with those ideas’ progression into suicidal plans and attempts ("“Pain” Input" section; Klonsky et al., 2021; Nock et al., 2008). Not to address this remarkable pattern is to neglect one of the most important questions in psychological and medical science: how do humans live with “The Lure of Death” (Humphrey, 2018)—having to hand, but not using, an all-purpose literal pain-killer? To reiterate the point, an evolved emergency anti-suicide system almost certainly exists, blocking the path from ideation to action (Himmelhoch, 1988; Soper, 2018). If the pain-brain analysis is broadly right, then it is difficult to conceive of a plausible alternative to psychopathology as the system’s most likely manifestation. If not psychopathology, then some phenomenon very much like it would have to be found that matches the a priori specification; but there is no obvious better candidate. The configuration of mental disorders, their inputs and outputs, align so precisely with the expected anti-suicide design that it would be an extraordinary coincidence if they were not displays of adaptations that evolved to serve this purpose. Pain-brain, as an adaptationist theory, can be judged on this goodness of fit (Williams, 1966).

Clinical Implications

The pain-brain theory suggests conceptually robust directions for mental health research, practice, and public policy. Four can be noted.

First, the theory predicts that suicide will not be amenable to useful prediction even in principle ("Suicides as Statistical Residuals"). By stopping all self-killings that the organism can practicably stop, evolved defenses will expectably erase linearity from the causal connection between risk factors and suicidal outcomes. Being statistical residuals, suicides will eventuate stochastically (Soper, 2019b). On this basis, science’s continuing search for a usefully accurate method of predicting suicidal behavior, an effective way to assess or screen for individual risk, may be misguided. Treatment decisions that are based purely on suicide risk assessment, however the assessment is done, likely misallocate resources. Instead of concentrating on suicide, arbitrarily allotting distressed people to risk categories, and often unhelpfully exaggerating the danger in the process, clinicians would do better to focus on understanding and trying to ease patients’ distress (Soper et al., 2022).

Second, for the same reason, suicide prevention interventions will have the best prospects for success if aimed not at individuals but at whole cohorts and populations (Soper et al., 2022). Pointing to one opportunity, human sapience probably evolved in an environment where self-killing was generally not a quick or simple undertaking (Soper, 2021). It may be essentially for this reason that humans are vulnerable to suicide in settings where easy means of lethal self-injury are to hand. Restriction of lethal means may correct this noxious mismatch between evolutionary and modern conditions. Given a little time in safe surroundings, the organism’s homeostatic defenses will usually resolve suicidal crises spontaneously (Bryan, 2020).

Third, a broad therapeutic target emerges: the way to alleviate both psychopathology and suicidality is to relieve their shared stimulus—psychological pain. Lasting relief presumably entails changing some painful life situation, usually relational; or, if it cannot be changed, accepting it (Linehan, 2020). As noted earlier, the establishment of a therapeutic relationship is a generalizable aid to recovery, potentially deliverable via many, outwardly disparate, schools of therapy (Wampold & Imel, 2015). Since pain is especially open to placebo-like modulation (Eccleston, 2001), there may be scope for what Fish (1973) calls, provocatively, “placebo therapy”, at least as a provisional intervention. The finding that placebo accounts for much of the effectiveness of some prominent psychopharmacological treatments (Khan & Brown, 2015) could point to placebo effects constituting, not an experimental nuisance variable, but the primary therapeutic process (Davies, 2013; Frank, 1974). Once the precipitating psychache eases, both suicide risk and, in its own time, diverse forms of psychopathology will expectably stand down (Goldberg & Goodyer, 2005).

Fourth, as for more targeted interventions, realistic humility may be in order. A limitation of the pain-brain theory in the clinical domain is that, although it may aid a general understanding of the human tendencies for psychopathology and suicide, it offers no simple solutions to either. The mechanisms with which the human brain manages its security and health have been fine-tuned by intense selection since the dawn of our species, or earlier (Humphrey, 2018; Soper, 2019a). Some of the machinery may be so complex as to constitute a black box. It may not be possible to determine why, say, suicide rates vary across different psychiatric disorders, or why one patient presents with life-long psychiatric conditions and another with self-remitting brief episodes. The exact etiology of many outcomes may be unascertainable even in principle.

Nonetheless, it may be taken as a reasonable working assumption that the organism is competent. It knows what it is doing, so to speak, even if its workings are opaque to us. Almost certainly, the psyche’s evolved algorithm can gauge the risk of suicide more reliably from the inside than any synthetic assessment tool can from the outside. Moreover, in light of the onboard assessment, it can usually be trusted to manage the risk, making whatever adjustments are necessary for survival while allowing social and other functions to continue as far as it deems safe to do so. There may be limits to how far medical science can improve on nature in this territory. Nonetheless, an informed professional can surely help by empathically countering the stigmatizing and fearful misconceptions that surround suicide and mental illness (Soper, 2021; Soper et al., 2022).

Concluding Comment: Challenges for Suicidology and Psychiatry

If this article’s thesis is correct, a long-sought key for progress in suicidology and psychiatry may be found in the dark of two overlapping shadows—evolution, and suicide. They point to reciprocal opportunities for illumination: suicidology stands to gain from an informed evolutionary stance, and evolutionary psychiatry would gain from paying closer attention to suicide.

Suicidology’s chief difficulty may not stem, as some critics suggest, from ongoing efforts to mold itself as a biological science: a “biologizing of suicide” (Shneidman, 2001, p. 73). The problem is that the biologizing has made insufficient reference to modern biology’s (evolutionary) first principles. The Darwinian puzzle of suicide is discussed at length outside of mainstream suicidology (Bering, 2018; deCatanzaro, 1981; Soper, 2018). Within suicidology, by contrast, evolution is rarely mentioned, even more rarely on an informed basis (Soper, 2022). Evolutionary issues are signally absent from most biological reviews, such as Lengvenyte et al.’s (2021) “broad overview of the biological bases of suicide” (p. 330) and discussions by van Heeringen and Mann (2014), Mann and Currier (2016), and others. At the same time, to return to a misconception raised earlier, the pre-Darwinian, folk-biological idea of a “survival instinct” is routinely cited as a premise—as a self-evident given. Lengvenyte et al. (2021), for example, introduce their review with the unsubstantiated claim, “In humans, suicide overrides the basic instinct of survival that is common to the vast majority of known life forms” (p. 330). Others assert, in similar vein,

Suicide is one of the most complex human behaviors to understand, transgressing our biological drive to survive and reproduce, (Bennett et al., 2023, p. 371).

It is also the case that self-harming behavior of any kind runs counter to a fundamental survival instinct. (Hooley et al., 2020, p. 101).

Given that suicidal behavior runs counter to the basic biological instinct to survive that is present even in single-celled organisms,” (Miller & Prinstein, 2019, p. 454, italics added).

This notion of a “basic biological instinct to survive” is erroneous. It has no place in 21st-century biology because modern evolutionary theory provides no mechanism by which an all-encompassing, superordinate drive to survive and/or reproduce could evolve (Kirkpatrick & Navarrete, 2006; Soper, 2019b). Nonetheless, precisely this chimera—“the most basic instinct of all; namely, self-preservation”—is advanced by Joiner (2005, p. 46) as footings for what may be suicidology’s currently most studied theoretical framework, and appears to be accepted uncritically by many other prominent researchers, such as Klonsky et al. (2016, p. 322), Nock et al. (2019, p. 258), O’Connor (2021, p. 80), Pompili (2022, p. 90), Wasserman (2016, p. 269), and Zalsman (2021, p. IX). The “survival instinct” fallacy is a systemic obstacle to scientific progress because it blocks sight of the possibility that humankind's primary protection against suicide resides in evolved, special-purpose, but not failsafe, anti-suicide adaptations. If Dobzhansky (1973) is right, that “[n]othing in biology makes sense except in the light of evolution”(p. 125), then this light is obligatory for making sense of the biology of suicide. An encouraging sign is a forthcoming suicidology handbook that (a first, it seems) dedicates a chapter to evolutionary perspectives (Soper & Shackelford, in press).

Psychiatry faces a different problem: a mid-level theoretical gap. On the strength of Darwinian macro-theory, evolutionists have long promised a solution to psychiatry’s pre-paradigmatic disarray (McGuire & Troisi, 1998; Wakefield, 1992), offering hope of “A New Beginning” (Stevens & Price, 1996). Fulfilment of this promise has yet to begin. As touched upon earlier, evolutionary psychiatry remains dominated by a miscellany of disjointed and often competing mini-theories (Brüne, 2016; Del Giudice, 2018)—at least 10 for depression alone (Nesse, 2023)—very few of which have the explanatory power to rally a consensus (Kennair, 2011; Kennair et al., 2021). To borrow one of Nesse’s (2019) axioms, “when so many smart people disagree, something is usually missing” (p. 86).

The missing theoretical “something” may be suicide. This extraordinary feature of our species is the direct cause of most premature mortality in the mental health domain (Woelbert et al., 2020) and, as we have seen, a pan-diagnostic phenomenon. It could justifiably take center stage in evolutionary analyses of psychopathology. Instead, it is usually confined to the wings. The section on suicide in McGuire and Troisi’s (1998) Darwinian Psychiatry (344 pp) spans a few paragraphs under “Other Conditions” (pp. 229–231). Brüne’s (2016) Textbook of Evolutionary Psychiatry and Psychosomatic Medicine: The Origins of Psychopathology (23 chapters) turns to the matter in Chapter 20. Del Giudice’s (2018) Evolutionary Psychopathology: A Unified Approach (560 pp) indexes “suicide” on just two pages. Sometimes the subject seems to be almost effortfully avoided. A landmark primer on evolutionary psychiatry by Nesse (2023) attends to metaphorical suicide among honeybees (“suicidally stinging intruders”; p. 179) and fictional suicide among lemmings (“Do lemmings commit suicide?”; p. 196)—but is silent about literal, actual suicide among humans. Nesse makes a passing reference to suicidal ideation as a “worse than useless” (p. 187) symptom of depression; why it is worse than useless and a symptom of depression (or, for that matter, why it associates with psychopathology across the board; Nock et al., 2008) are not discussed. The tangentiality suggests that suicide’s biological significance is sensed, even if it eludes frontal scrutiny.

Progress may require strong stomachs ahead of sharp minds. A society-wide reticence towards suicide—“something you do not talk about” (Goulah-Pabst, 2023, p. 17) and preferably not contemplate—affects the scientific community too, as generations of suicidologists report first-hand (Felix, 1965; Joiner, 2005; Nelson, 1968; Rogers, 2001). The taboo may be protective (Soper, 2018), but it can stifle scientific discourse. It appears to have consigned suicide’s profound biological implications to a scientific black hole—a mass that predictably exists but is so dark that it requires special exertion to behold. If this is true, the primary scientific challenge is more psychological than technical. The task entails, as Wakefield (2023) remarks, overcoming a “tension between… what we want to believe about ourselves… versus the scientific reality of human nature” (p. 144). Evolutionary psychiatry is far from alone in looking past the abyss. Psychology and anthropology’s grand theories of our species’s success—social intelligence (Haidt, 2012), cumulative cultural learning (Henrich, 2015), and others—shy away from considering how ancestral humans would have flourished if, on reaching reproductive maturity, they killed themselves: prior defenses had to be in place, channeling humankind’s psychic, moral, and socio-cultural evolution (Humphrey, 2011; Soper, 2018, 2019a, 2021). However, the onus to face suicide as an evolved human potentiality falls squarely on the science of mental health. If psychiatry is to move forward, it cannot continue to side-step this issue.

What may be the most fundamental, least comfortable, question for psychiatry, suicidology, and psychological science in general is posed by Menninger (1938): “Why does the wish to live ever, even temporarily, triumph over the wish to die?” (p. 5, original italics). To answer: the human psyche is equipped with evolved adaptations that are designed to motivate us to shun the idea of suicide. When the idea is in danger taking hold, other evolved adaptations are designed to keep us from putting the idea into action. These protective interventions may present as psychiatric disorders. The evolutionary biological problem of suicide may lie at the center of human psychopathology. Once the core’s presence and gravity are recognized, then the behavior of the orbiting material begins to make sense. The center, if seen as such, offers a much-needed basis for understanding our species’s susceptibility to mental illness, and a platform for developing practical ways to promote recovery, health, and wellbeing.