David Hume

Hume is the right philosopher to put in conversation with psychiatry, and not just because his epistemology happens to be useful.

Unlike Descartes or Kant, Hume was — by any reasonable clinical reading — a patient. His philosophy of mind was written by a man who had broken down, recovered, relapsed in a new key, and recovered again, and who left behind two of the best first-person documents of depression in the eighteenth century. So the connection runs in both directions: his system illuminates psychiatry, and his case history illuminates his system.

The system in brief

Hume's project in the Treatise of Human Nature (1739–40) was a "science of man" built on observation rather than metaphysics — Newton for the mind. Everything in the mind is a perception: vivid impressions (sensations, passions) and their fainter copies, ideas, assembled by association. From this austere starting point come his famous demolitions. Causation: we never perceive necessary connection, only constant conjunction; after enough repetitions the mind acquires a habit of expectation and projects that habit onto the world as "necessity." Induction therefore has no rational foundation, only a natural one — "Custom, then, is the great guide of human life." The self: when Hume introspects, he finds heat, cold, love, grief, perceptions in flux — never a perceiver. The self is "a bundle or collection of different perceptions," and personal identity is a fiction stitched together by memory and imagination. Motivation: "Reason is, and ought only to be the slave of the passions" — reason calculates means and matters of fact, but only passion supplies ends and motive force. Ethics: you cannot derive an ought from an is; morality rests on sentiment, above all sympathy, his term for the transmission of feeling between minds. And the endpoint of all this corrosive doubt is not nihilism but what he called mitigated skepticism: radical doubt is unlivable, nature compels belief anyway, so the wise course is modest, evidence-proportioned inquiry. "A wise man proportions his belief to the evidence."

Hume as patient

Around 1729, at eighteen, after years of obsessive solitary study, Hume collapsed. A Scottish physician told him, half-laughing, that he had the Disease of the Learned. His 1734 letter to a prominent physician (probably Arbuthnot, possibly Cheyne — and possibly never sent) is a startlingly modern case report: loss of "ardor" for the work he loved, inability to sustain concentration, scurvy spots on his fingers, watering mouth, and a state he compared to the "coldness and desertion of the spirit" described by the mystics. The regimen: bitters, anti-hysteric pills, a daily pint of claret, riding eight to ten miles a day — and, on his own initiative, occupation. He left philosophy for Bristol to clerk for a sugar merchant, deliberately choosing what he called a more active life. He partially recovered, went to France, and wrote the Treatise.

Then, at the end of Book 1, the crisis recurs in philosophical register. His own skepticism has left him "in the most deplorable condition imaginable, environ'd with the deepest darkness," fancying himself "some strange uncouth monster" expelled from human commerce.

His diagnosis of his own recovery is the crucial line: "since reason is incapable of dispelling these clouds, nature herself suffices to that purpose, and cures me of this philosophical melancholy and delirium."

Notice what happened and what didn't. The intellectual problems were never solved — Hume never refuted skepticism, then or ever. The despair lifted anyway. And he didn't quit philosophy; he returned to it in measured doses, alternating it with company and backgammon for the rest of his life. Hence his lifelong maxim: "Be a philosopher; but, amidst all your philosophy, be still a man."

The lessons

Causal humility is the founding discipline. If all we ever have is regularity plus a habit of inference, then medicine's entire causal apparatus — randomization, Bradford Hill's 1965 considerations — is best understood as engineering around the Humean gap, not closing it. (Ghaemi makes this lineage explicit: Hill is the pragmatic answer to Hume.) Psychiatry has been burned precisely where it forgot this. The "chemical imbalance" story converted a loose conjunction (monoaminergic drugs sometimes help) into a felt necessity (low serotonin causes depression), and the field paid for it in credibility when the projection collapsed. The same discipline applies prospectively to our current enthusiasms. Hume's standard from the miracles essay — extraordinary claims require testimony whose falsehood would be more improbable than the claim itself — is a perfectly serviceable filter for breakthrough-of-the-quarter psychiatry, ketamine and psychedelics included. But proportioning belief to evidence cuts both ways: it forbids the nihilist move too. That our causal story was wrong does not license concluding the treatments do nothing. (One modern echo worth noting: Hume's "custom" is the direct ancestor of the Bayesian prior, and a predictive-processing account of depression as pathologically rigid negative priors — with psychedelics as prior-relaxation — is about as Humean as twenty-first-century neuroscience gets.)

Depression is a disorder of the passions, and reason is its press secretary. If reason is the slave of the passions, then depressive cognition is what reason produces when enslaved to a darkened passion. Hume's associationism predicts mood-congruent cognition almost exactly: melancholy enlivens congruent ideas, which then masquerade as conclusions. This explains three clinical facts at once — why insight alone rarely cures, why "just think positive" insults the phenomenology (the patient's bleak beliefs feel like discoveries because the machinery that makes beliefs feel true sits downstream of affect), and why Hume's own cure worked. Change the passion — dinner, backgammon, friends — and the reasons lose their grip, exactly as he reported. He discovered behavioral activation two and a half centuries before Lewinsohn and Jacobson formalized it. It also reframes CBT charitably: cognitive work succeeds not by out-arguing the disease but by giving the passions different inputs.

The same lens lets you answer Hume with Hume on the hardest question. His posthumous essay Of Suicide defended its potential rationality. But a Humean psychiatrist can reply: in a depressive episode, the apparent reasons for dying are productions of the passion, not independent judgments — and like his own skeptical despair, they tend to evaporate when the passion lifts. That's a principled, non-paternalistic ground for treating in-episode suicidality as state rather than verdict, while leaving honest room for the genuinely difficult chronic and refractory cases where his essay still presses on autonomy debates.

The bundle self dissolves the depressive's metaphysics. The depressive cogito is an essence claim: I am worthless, broken, defective. Hume's point is that introspection never finds an essence available to be defective — only successive states. Whatever this does to metaphysics, therapeutically it is defusion avant la lettre: ACT's self-as-context versus self-as-content, MBCT's "thoughts are not facts," the observing self. The convergence with Buddhist anattā is real, and possibly more than convergence — Gopnik has argued Hume may have encountered Buddhist ideas through the Jesuit network at La Flèche, where he wrote the Treatise. The same caution applies to our own labels: "I am bipolar" reifies a description of course and symptoms into an identity. Useful shorthand; dangerous metaphysics.

Recovery does not require resolution. Hume got better twice without ever answering the questions that broke him. Patients do not need their existential or ruminative questions settled as a precondition of recovery; re-engagement with what he called common life often dissolves the questions' urgency rather than their content. This is squarely consistent with the rumination literature — self-focused analysis of distress prolongs it, engagement shortens it — and it has a contemporary edge: a culture of constant self-monitoring and ambient mental-health content arguably runs Hume's experiment in reverse at scale, which is roughly what the prevalence-inflation hypothesis worries about. But hold the balance he held. Hume didn't renounce inner life; he learned alternation and dose — the Treatise and the backgammon. The prescription is rhythm, not avoidance.

The is-ought gap runs straight through the DSM. Where ordinary sadness ends and disorder begins — the bereavement-exclusion fight is the canonical case — is not discoverable in nature alone; it's partly a judgment about which suffering warrants a medical response. Hume doesn't say values are illegitimate. He says don't launder them as facts. Both of psychiatry's chronic failure modes violate this in opposite directions: medicalizing normal misery pretends an ought-question was settled empirically, while antipsychiatric nihilism invalidly infers from "diagnosis involves values" to "there is no illness." Honest nosology keeps both ledgers visible.

Sympathy is the active ingredient. Hume made the transmission of sentiment between minds the foundation of moral life, and psychotherapy research keeps rediscovering the corollary: alliance and empathy carry a remarkable share of outcome variance across modalities. In Humean terms, the therapist's steadier passions are communicated, not merely their propositions — which is also a decent folk model of expectancy and placebo effects, nowhere larger than in psychiatry. And the mechanism is bidirectional, which makes it a theory of burnout: the clinician absorbs the conjunction of suffering daily.

Mitigated skepticism is the psychiatric temperament. The endpoint of Hume's epistemology is neither dogmatism nor paralysis but chastened, active modesty: act on the best regularities available while holding theory lightly. That is arguably the correct professional stance for a field whose categories are provisional, whose mechanisms are conjectural, and whose patients cannot wait for the metaphysics to be sorted. It's the disposition Ghaemi has spent a career defending under other names — Jaspersian pluralism, Hippocratic humility — and he's candid about the Humean debt.

Coda

Hume died in 1776 of abdominal cancer, cheerfully and without consolation — joking with a scandalized Boswell about annihilation, prompting Adam Smith to enrage Britain by writing that his friend had died about as well as a human being can. For a field now seriously engaged with existential distress and dying — psilocybin in terminal illness, palliative psychiatry — Hume stands as a proof of concept: equanimity at the end is achievable without metaphysical guarantees, built instead from temperament, habit, and friendship. Passions well-tended.

Which is his whole lesson for psychiatry, compressed: the mind runs on passion and custom, not on reason, so treat those; proportion belief to evidence, against both hype and nihilism; hold the self and the theory lightly, because neither is the fixed thing it pretends to be; and amidst all your science, be still a man.