The illusion of the empty chemical tank

Give me a flipchart and two minutes and I can make almost anything look wonderfully simple.

In a few strokes of the pen, bewilderingly complex human experiences can be reduced to a few words. They can be explained in a training room, remembered after a consultation and repeated to somebody else over dinner. This is the generosity of metaphor. It gives us something we can feel proud we understand when the reality itself is too large, too technical or too frightening to grasp all at once.

Every metaphor is a small mercy. It takes something we cannot hold in our hands, a hundred billion neurons firing in the dark, the slow architecture of despair, a person's life unraveling one bet at a time, and gives it a shape we can grasp.

We reach for these shapes because the alternative, sitting with irreducible complexity, is almost unbearable. And there is nothing wrong with that. A metaphor isn't a shortcut we take instead of thinking; it's largely how thinking happens at all.

The danger is not that we use metaphors. It's that we forget we're using one.  A model proposed to make a hard idea teachable quietly hardens into "the way things are." The map slowly replaces the territory. And because the map is clearer than the territory, it is often the map that survives. Professionals build careers on it. Institutions print it in their training manuals. And by the time better evidence arrives, the metaphor has outlived the need it was built for, it has become furniture in the room, something nobody thinks to question.

This article follows the journey of one remarkably persuasive metaphor: from tentative scientific hypothesis, to convenient clinical explanation, to something millions of people came to accept as medical fact. It is also a story about what can happen when an explanation becomes simpler than the suffering it is meant to describe.

In 1965, the Harvard psychiatrist Joseph Schildkraut proposed that depression might result from a shortage of certain brain chemicals called monoamines. Two years later, the British psychiatrist Alec Coppen sharpened the idea into what became the serotonin hypothesis. The logic ran backward from treatment to cause: the early antidepressants seemed to raise serotonin activity, and patients on them seemed to improve, so, the thinking went, depression itself must be a serotonin deficiency. But a medicine’s action does not, by itself, reveal the cause of the condition it helps.  It's the same shape of reasoning as concluding that headaches must be caused by a shortage of aspirin. Schildkraut himself, later emphasised that even his original account described the hypothesis as a reductionist oversimplification of a much more complex biological state.

Despite the fact that these were attempts to generate research, not announcements that the mystery of depression had been solved, this admittedly thin hypothesis became a household phrase in the 1990s, when pharmaceutical companies marketing the new SSRI antidepressants (Prozac chief among them) built their advertising around the idea of correcting a "chemical imbalance." Official bodies echoed it. Physicians, working with ten-minute appointments, found it an efficient and humane way to explain a frightening diagnosis: not a moral failing, not weakness, just chemistry. The same kind of thing you'd say about diabetes. It served a genuinely good purpose, reducing the sense that depression was something to be ashamed of.

It also was never well supported, and the evidence against it kept accumulating slowly for decades before the public caught up. In July 2022, a team led by UCL psychiatrist Joanna Moncrieff published a comprehensive umbrella review in Molecular Psychiatry, pulling together every major line of evidence: serotonin and its breakdown products in blood and spinal fluid, receptor binding studies, transporter levels, tryptophan depletion experiments, genetic studies, and gene-environment research. Across the board, they found no consistent evidence that depression is caused by lower serotonin activity. If anything, the one reliable association ran the other way - people already taking antidepressants tended to have lower serotonin, suggesting the drugs were altering baseline chemistry rather than correcting a pre-existing deficit. The review has itself generated debate, particularly about what its conclusions do and do not imply, but one distinction is essential: questioning a simple serotonin-deficiency explanation is not the same as claiming that serotonin plays no role in mood, or that antidepressants cannot help. This wasn't a new revelation to people inside the field. The psychiatrist David Healy had already called the theory a myth in the BMJ back in 2015, and the psychologist Irving Kirsch had spent years re-analysing antidepressant trial data and consistently finding effects only marginally above placebo for mild-to-moderate depression.

What the evidence actually points to is something far messier and more interesting than a leaky faucet of one chemical. Depression looks like a systems-level condition shaped by modest genetic risk spread across many small-effect variants, structural and functional differences in stress-response circuitry, disrupted neural networks, and a growing body of research on neuroinflammation, where elevated immune signaling molecules like certain cytokines and activated glial cells appear to play a central, not incidental, role in depressive symptoms. Layered on top of all of that are social determinants (poverty, discrimination, isolation, loss, trauma) that shape both risk and recovery. These influences will not be present in the same combination, or carry the same weight, for every person.

The chemical-imbalance story nevertheless remains culturally powerful, including among professionals. A 2025 US study of 1,219 students found that 73 per cent had encountered the explanation. People heard it in classrooms, from media, friends and clinicians, but hearing it from a healthcare provider was the only source that uniquely predicted believing it. The authority of the messenger helped transform a familiar story into an accepted fact.

The myth persisted among professionals for reasons that have less to do with ignorance than with convenience and inertia. It told people that depression was real, that their suffering was not laziness and that asking for medical help was legitimate.  It was easy to say, it reduced perceived blame. We should not dismiss that achievement. But more than anything it was reinforced by an entire industry's marketing and continuing-education apparatus. Walking it back raises an uncomfortable question that the field is still arguing about: if not a chemical imbalance, then what, exactly, do we tell a frightened patient in the time we have?

The cost of not answering that question well is measurable, and it isn't small.

Multiple studies have found that people who believe their depression is caused by a chemical imbalance report significantly worse expectations about their own recovery, a phenomenon researchers call prognostic pessimism, the belief that there isn't much you can do to change your own state. That belief correlates with lower confidence that therapy will help and a stronger reliance on medication as the only "real" treatment. In a 2020 study with 279 people in residential psychiatric treatment, Hans Schorder and his colleagues found out that endorsing the chemical imbalance explanation actually predicted more depressive symptoms two weeks after discharge, independent of how depressed they were to begin with.

A metaphor built to lighten shame may, in a measurable number of people, have reinforced hopelessness instead.