Psychiatry and its discontents

Critics of psychiatry emerged long before the rise of antidepressants. The dire state of psychiatric institutions had already drawn public ire in the 1940s,Footnote ²⁶ although their most famous critique was not published until 1961 by the sociologist Erving Goffman.Footnote ²⁷ By this time, large state-run hospitals had already begun to close, spurred on by patients’ rights activism,Footnote ²⁸ financial pressures, and the introduction of the first antipsychotic, Thorazine (chlorpromazine), in 1952.Footnote ²⁹

While the term ‘anti-psychiatry’ has been broadly applied to figures challenging institutional psychiatry in the 1950s and 1960s, few identified with the label.Footnote ³⁰ More precisely, it describes a group of British psychiatrists, such as R.D. Laing and David Cooper (who first coined the term in 1967).Footnote ³¹ They gained fame and a degree of notoriety as figures in Britain’s counter-cultural movements, although their early reputations rested on establishing alternative therapeutic communities.

As Michael Staub has argued, Laing and the other British anti-psychiatrists echoed an already widely held position at the time, that the causes of mental illness were to be found not in the biology of the brain, but in the social and relational fabric of everyday life.Footnote ³² Their success stemmed in part from their ability to blend this idea with an increasingly wide array of countercultural sources – from the Beat Generation to eastern mysticism – in a way that resonated with the left-wing liberation politics of the 1960s. Madness, in this view, was an unreasoned yet reasonable response to a schizophrenogenic world: one to be celebrated and even emulated.Footnote ³³

US psychiatry was at the time dominated by psychoanalytic thought, where much mental illness was believed to be the consequence of repressed unconscious conflict – a position the anti-psychiatrists implicitly shared.Footnote ³⁴ Yet the loose diagnostic formulations that Freudian thought produced soon invited another kind of critique. While often grouped with the anti-psychiatrists, Thomas Szasz rejected the label outright.Footnote ³⁵ A Hungarian émigré who had fled Nazi Europe, Szasz trained and practised as a psychiatrist in the USA, but his most enduring legacy came from questioning the very basis of psychiatric diagnosis. In The Myth of Mental Illness (1961), Szasz argued that medicine concerns itself with diseases with observable biological correlates, such as lesions or physiological dysfunction. Since psychiatric diagnoses lacked such correlates and were instead defined by behaviour, he claimed they could not be diseases at all but should be understood as ‘problems in living’.Footnote ³⁶ If a biological cause were found, the condition would cease to be psychiatric and be reclassified as neurological. Psychiatric illness, in this view, was a category error.

Szasz’s critique was motivated in part by his libertarian political beliefs; he shared a loathing of totalitarian regimes common among intellectuals who had fled Europe. He saw mental illness as a way for individuals to shirk their responsibilities and psychiatry as a fundamentally coercive institution at odds with individual liberty. This left little space to account for psychological distress or ‘pain’, and Szasz saw claims of ‘illness’ as a form of unconscious malingering.Footnote ³⁷

His critique of psychiatry, however, was a philosophical one. On the one hand, he reified all ‘real’ disease as based in an identifiable biological lesion, an essentialist criterion that many physical diseases could not fulfil, and on the other, claimed human ‘behaviour’ and experience more broadly to be outside of this realm. Some have argued that Szasz relies on a form of mind-body dualism, and as medicine applies only to one side of that division, psychiatry becomes a logical impossibility. This position – along with strong dualistic claims more generally – has been widely critiqued on philosophical grounds.Footnote ³⁸ Nevertheless, the simplicity of its logic, and its ethical appeal to autonomy, proved compelling to many American libertarians. His work was frequently cited by jurists in civil liberties cases involving involuntary psychiatric commitment during the 1970s, helping to shape case law that significantly curtailed psychiatry’s coercive powers.Footnote ³⁹

Unlike his British counterparts, who retained some standing within mainstream psychiatry, Szasz’s radicalism alienated him from many colleagues. After a confrontation at the APA’s 1964 conference where he was publicly challenged, he sought allies outside the psychiatric establishment.Footnote ⁴⁰ He eventually aligned with the Church of Scientology, which had itself become a vocal critic of psychiatry, and co-founded its activist organisation, the Citizens Commission on Human Rights, which promoted the abolition of psychiatry. Although he claimed it was a tactical alliance, he remained associated with them until his death in 2012, something many felt damaged his credibility.Footnote ⁴¹

Nevertheless, the alliance illustrates the wide ideological spectrum – from radical left to libertarian right – united in opposition to psychiatry by the 1970s. While the intentions of their critiques varied, both Szasz and the British anti-psychiatrists shared a common belief that psychiatry threatened individual autonomy. These critiques had real-world consequences. They influenced reforms to the legal frameworks governing coercive psychiatric care and forced the profession to confront the philosophical accusation of social constructionism. Psychiatry’s response was not one these critics were likely to endorse, however, as it increasingly discarded psychoanalysis and reoriented itself towards a new ‘biological’ conception of disease, searching for the physical underpinnings that it was criticised as lacking.Footnote ⁴²

The rise of antidepressants

Even within an era of Freudian dominance, psychiatrists had long used medications to treat their patients. Thorazine was quickly regarded as a ‘wonder drug’, both in terms of its clinical effects and its economic success.Footnote ⁴³ It was not the first psychiatric drug to be discovered, but it spurred on an eager search for others.Footnote ⁴⁴ Whilst looking for analogues to Thorazine, researchers discovered that the drug imipramine caused some depressed patients to become elevated and energetic. After promising trial results, it was released in 1958, closely followed by amitriptyline in 1961, the first two tricyclic antidepressants (TCAs). Around the same time, iproniazid, a drug originally developed to treat tuberculosis, was also found to have mood-lifting effects.Footnote ⁴⁵

The idea that a chemical substance could alter one’s mental experience pushed the new field of neurophysiology to sudden prominence. The discovery of neurotransmitters (the chemical messengers between nerve cells) in the brain prompted researchers to investigate their effects on behaviour,Footnote ⁴⁶ especially after LSD was found to be structurally similar to serotonin.Footnote ⁴⁷ One group showed that using a drug that lowered concentrations of serotonin caused a depression-like state in rats, which was then reversed when serotonin was restored.Footnote ⁴⁸ However, the Swedish pharmacologist Arvid Carlsson demonstrated that another class of neurotransmitters – catecholamines – also affected behaviour.Footnote ⁴⁹ Further studies showed that TCAs worked by preventing the reuptake of serotonin and catecholamines from the synapse,Footnote ⁵⁰ while iproniazid prevented their breakdown. Both drugs seemed to work by increasing synaptic concentrations of neurotransmitters.

Other evidence began to push researchers towards the idea of the brain as a chemical organ and led some to hypothesise that mental illnesses may arise from neurochemical imbalances.Footnote ⁵¹ Schizophrenia was linked to elevated catecholamine levels,Footnote ⁵² which could be blocked by Thorazine, and Carlsson later demonstrated that movement disorders could be treated with the catecholamine precursor, L-DOPA, paving the way for its use as the first effective medication for Parkinson’s disease.Footnote ⁵³ In 1965, low catecholamine levels were proposed as a cause of depression,Footnote ⁵⁴ while others suggested low serotonin.Footnote ⁵⁵ Both camps recognised that chemical imbalance theories were likely to be ‘simplifications’, as did prominent figures like Carlsson.Footnote ⁵⁶ Yet as the best fit for the limited evidence at the time, they were meant to help organise further research and push future discovery.Footnote ⁵⁷

Nevertheless, these theories motivated a growing chorus of biological psychiatrists who pushed against the Freudian establishment.Footnote ⁵⁸ Although depression had long been recognised as a psychiatric condition, it was generally considered to be severe and fairly rare, involving symptoms like suicidality or psychosis.Footnote ⁵⁹ Milder cases tended to be seen in psychoanalytic terms as anxious reactions. But the medications’ direct effects on mood prompted some researchers in the 1960s to argue that more common everyday symptoms were the manifestations of ‘true’ depression instead.Footnote ⁶⁰ This view was encouraged by early epidemiological findings that suggested depression was indeed both common and under-recognised.Footnote ⁶¹ But it also rode a wave of optimism in the science of neurophysiology, evidenced by the retrospective christening of drugs like iproniazid and TCAs as ‘antidepressants’, giving an after-the-fact precision to fit the emerging chemical narrative.Footnote ⁶²

The biological turn in psychiatry came to a head in the 1980s, as the profession battled both itself and its critics over how to define the conditions it treated. US regulatory changes in the 1960s meant drug companies had to prove the effectiveness of new drugs in clinical trials,Footnote ⁶³ and the pharmaceutical industry began to push for more standardised diagnostic criteria for use in these trials. This became an opportunity for some actors to finally displace the Freudian hegemony. When, in 1980, the APA published the third edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-III), it was a revolution – stripping away the vague psychoanalytic descriptors that had long occupied it. Diagnoses became sharply defined categories made up of specified lists of symptoms.Footnote ⁶⁴ Although it was supposedly atheoretical – and agnostic about the causes of mental illness – in reality, the redrafting committee had leaned heavily towards a biological approach.Footnote ⁶⁵ A new catch-all category, Major Depressive Disorder, subsumed previous classifications like reactive or endogenous depression, and arguably opened the way for simplistic chemical imbalance theories to become its dominant explanatory framework.Footnote ⁶⁶

Yet even by this point, evidence against monoamine theories had begun to mount. The catecholamine hypothesis was at first the more popular, but a 1984 study by the NIMH found no correlation between catecholamine metabolite levels and depression, challenging a straightforward deficit model.Footnote ⁶⁷ In Europe, researchers had focused on serotonin as the mood-altering chemical – especially after low levels of its precursor, tryptophan, were associated with depression.Footnote ⁶⁸ At Carlsson’s urging, pharmaceutical companies began to develop drugs that selectively inhibited the reuptake of serotonin.

When SSRIs came to market in the 1980s (Prozac was released in the US in 1988), they were the first class of psychiatric drugs to appear only after they had undergone evaluation in clinical trials. Studies had shown that they were more effective than a placebo pill,Footnote ⁶⁹ although no more effective than earlier drugs (and potentially less).Footnote ⁷⁰ Nevertheless, they were sold as the climax of years of scientific progress, a clean drug specifically designed to target a single neurotransmitter, meaning fewer side effects.

What remained unknown, however, was how they actually worked – what scientists refer to as their ‘mechanism of action’. By the time they were released, further research had challenged any simple relationship between low serotonin and depression. Scientists could induce a serotonin-depleted state by starving individuals of the amino acid and serotonin precursor, tryptophan. However, healthy participants did not become depressed when starved, although those with a prior history of depression experienced a relapse in their symptoms, and some with a family history also experienced a mood drop.Footnote ⁷¹

As SSRIs’ mechanism became more uncertain, another challenge emerged from neuroscience itself. By the 1980s, researchers began to conceptualise the brain as a complex network and recast neurotransmitters as modulators of these systems.Footnote ⁷² Serotonin was found to play a role in how animals learnt about threat and punishment.Footnote ⁷³ As the dominant metaphor for the brain began to shift from chemical organ to dynamic network, it no longer yielded a simple disease model for psychiatry to map clinical conditions onto. Increasingly challenged by new evidence, the serotonin hypothesis became the orphan of an abandoned theory of the brain.

Nevertheless, SSRIs proved extraordinarily popular. In 1992, Prozac became the most prescribed drug in the United States,Footnote ⁷⁴ driven in part by SSRIs’ favourable safety profile, which made them easier to prescribe for the primary care physicians who treated most cases of depression.Footnote ⁷⁵ However, their success also owed much to the growing cultural acceptance of new biological models of depression, pushed by aggressive marketing campaigns. In the United States, direct-to-consumer drug advertising was legalised in 1985, and some advertisements portrayed SSRIs as correcting ‘low levels’ of serotonin.Footnote ⁷⁶

The ethics of this approach have since been widely debated.Footnote ⁷⁷ Although many scientists initially supported the serotonin hypothesis as a plausible explanation, by the time these drugs were marketed, evidence had clearly begun to complicate this. In presenting an increasingly disputed theory as established fact, this advertising strategy created a misperception that would last for decades. As subsequent scandals soon began to mount, the use of an obsolete theory to promote SSRIs left the pharmaceutical industry open to accusations of bad faith.

Psychiatry’s biological reorientation, which began in part as a response to its critics, had been made possible by the success of these medications. As they became a source of controversy, it would hand those same critics new grounds to remake their case.

After the decade of the brain

In July 1990, President George H. W. Bush issued a Presidential Proclamation declaring the new decade the ‘Decade of the Brain’. It would turn out to be the high-water mark for public acceptance of biological explanations for mental illness.Footnote ⁷⁸ SSRIs became potent cultural symbols. Psychiatrist Peter Kramer captured the prevailing enthusiasm in his best-selling book Listening to Prozac, in which he famously suggested that SSRIs could make people ‘better than well’, coining the provocative term ‘cosmetic pharmacology’ to describe this phenomenon.Footnote ⁷⁹ First-person memoirs such as Prozac Nation by journalist Elizabeth Wurtzel and Prozac Diary by psychologist and writer Lauren Slater portrayed these drugs in a positive, although somewhat ambivalent, light.Footnote ⁸⁰

However, this wave of optimism quickly gave way to a growing body of criticism. Figures like R.D. Laing and Thomas Szasz had seen their influence wane in the 1980s, marginalised by the biological turn, but their critiques began to find new audiences in the wake of SSRIs. In 1994, the American psychiatrist Peter Breggin – a student and acolyte of Szasz who also questioned the biomedical model – published Talking Back to Prozac. Though seen as highly alarmist, Breggin’s book raised concerns that would persist in critiques of SSRIs: that their efficacy was overstated due to industry bias, that their risks were under-recognised, and that they could induce serious side effects, including suicidality and aggression.Footnote ⁸¹ Breggin went further, however, comparing SSRIs to recreational drugs and attracting controversy by attributing the violent behaviour of one of the perpetrators of the 1999 Columbine High School massacre to their use.Footnote ⁸²

Concerns about SSRIs and suicidality had, however, already begun to surface. As early as 1990, case reports suggested that fluoxetine might be associated with increased suicidal ideation.Footnote ⁸³ Initially met with scepticism – suicidal thoughts being closely linked with depression itself – the claims eventually prompted the APA to launch a task force. Eli Lilly, the manufacturer of fluoxetine, vigorously defended the drug, even as it quietly settled several legal cases out of court.Footnote ⁸⁴ Subsequent studies and meta-analyses did reveal a small but measurable increase in self-reported suicidal thoughts among individuals under eighteen, although no conclusive link with aggression was established.Footnote ⁸⁵ In 2004, the FDA issued a black-box warning regarding this risk, and in the UK, NICE restricted SSRI use in those under eighteen to fluoxetine alone.Footnote ⁸⁶

Similar concerns were echoed by David Healy, an Irish psychopharmacologist and medical historian, in his 2004 book Let Them Eat Prozac. Footnote ⁸⁷ Healy accused the pharmaceutical industry of manipulating clinical trial data to downplay safety concerns. Blurring the roles of historian and historical actor, Healy became a whistle-blower and described how his public criticisms had led to the withdrawal of a job offer, which he interpreted as industry retaliation.

Although some questioned Healy’s motives – he, like Breggin, began serving as an expert witness in SSRI-related legal casesFootnote ⁸⁸ – further evidence emerged to support his claims. Perhaps the most infamous example was Study 329, sponsored by SmithKline Beecham. Published in 2001, the study concluded that paroxetine was safe and effective in treating adolescent depression.Footnote ⁸⁹ Yet a subsequent re-analysis of the original data contradicted this conclusion.Footnote ⁹⁰ Moreover, the original article was revealed to have been ghostwritten by pharmaceutical copywriters, raising serious concerns about scientific integrity and misleading authorship.Footnote ⁹¹

Concerns extended beyond individual studies to encompass the wider issue of selective publication. A 2008 analysis of both published and unpublished trials held by the FDA found that, of 74 studies on 12 antidepressants, only 51 were published – 94% of which reported positive results, despite only 51% being deemed positive by the FDA. This clear publication bias substantially overstated the efficacy of antidepressants.Footnote ⁹²

In a series of papers, Irving Kirsch analysed unpublished FDA trial data and suggested that SSRIs were only marginally more effective than placebo – not enough, he argued, to be considered ‘clinically meaningful’.Footnote ⁹³ A researcher at Harvard’s Placebo Studies unit, he argued that their benefits were better explained by an ‘amplified placebo effect’.Footnote ⁹⁴ However, Kirsch was later criticised for misrepresenting the difference between drug and placebo effects: Placebos work quickly and then tend to wane; antidepressants only begin to take effect weeks after starting, and their benefits are sustained.Footnote ⁹⁵

As doubt regarding the effects of SSRIs mounted, it spilled into other debates. Many began to reconsider the very concept of depression that these drugs’ success was built on. There was a flourishing of scholarly critique from historians and philosophers, challenging the checklist diagnoses that the DSM-III made possible. Some argued that this had allowed the category to expand and include much of what was once thought of as ‘normal sorrow’.Footnote ⁹⁶ Historians of medicine began to reconstruct this transformation, emphasising how the meaning of depression had changed under the influence of psychiatry’s biological turn.Footnote ⁹⁷

Some argued that psychiatry had been fundamentally altered by this new conception of mental illness, leading to an over-reliance on prescribing at the expense of psychotherapy, a trend often attributed to cost-containment by health insurers.Footnote ⁹⁸ In the UK, concerns about antidepressant overuse led NICE to recommend psychotherapy as a first-line treatment for depression, and in 2008 the government launched the Improving Access to Psychological Therapies (IAPT) programme, a self-referral scheme designed to expand access to NHS-funded talking therapies.Footnote ⁹⁹

The sense of scandal precipitated other changes. Concerns around bias in industry-funded trials were not unique to psychiatry and extended to the whole of academic medicine.Footnote ¹⁰⁰ Still, antidepressants loomed large in this broader debate, and they were frequently cited in congressional hearings leading to the 2007 FDA Amendments Act.Footnote ¹⁰¹ This marked a shift in government oversight in how trials were run, requiring drug companies to preregister trials and publish all results – not just positive ones. This legislation coincided with other grassroots science-activism movements that helped to establish a new ethical standard in scientific publishing, including authorship declarations and conflict-of-interest disclosures.Footnote ¹⁰²

As scientists began to restore faith in their methods, doubt about the effectiveness of antidepressants seemed to settle as well. Following Kirsch, other researchers reanalysed the same datasets and found that antidepressant efficacy remained significant – albeit with less hyperbole than in earlier decades. Jay Fournier and colleagues, for instance, found that antidepressants produced significant benefit, but primarily in cases of severe depression.Footnote ¹⁰³ Later meta-analyses confirmed modest but meaningful efficacy in milder cases as well.Footnote ¹⁰⁴ Between 2001 and 2006, the NIMH conducted the STAR*D trial, involving over 4,000 patients, many of whom had not responded to previous treatments or presented with complex comorbidities (unlike the simpler patients typically included in industry trials). The trial – then the largest and most expensive of its kind – found that remission could be achieved in 67% of cases when non-responders were switched to alternative antidepressants, although these findings were later scrutinised.Footnote ¹⁰⁵

The STAR*D trial used other, newer antidepressants that had emerged since SSRIs, like bupropion (Wellbutrin) and venlafaxine (Effexor). In the wake of the monoamine hypothesis, researchers had looked at other neurotransmitters like dopamine and noradrenaline and found that drugs targeting these (with or without serotonin) could relieve depression in some people as well.Footnote ¹⁰⁶ What remained unclear, however, was which drug to use in which person. The trial helped to establish the principle that depression was a heterogeneous condition: If one class of antidepressants didn’t help, a different one might.Footnote ¹⁰⁷

By the late 2000s, however, the hope of ever more treatment options was fading quickly. Faced with growing regulatory and financial burdens, and a lack of new neurotransmitter targets, the pharmaceutical industry largely retreated from psychiatric drug development.Footnote ¹⁰⁸ Aside from a few notable exceptions, psychopharmacology’s half-century of promise had stalled.Footnote ¹⁰⁹

Two decades after the so-called ‘Decade of the Brain’, a more sober scientific consensus had emerged about antidepressants. Peter Kramer summarised the two decades of scientific controversy following Listening to Prozac by concluding that they worked ‘ordinarily well’ – that is, they were moderately effective in a substantial proportion of cases, with effect sizes comparable to those of other common medications such as beta-blockers or NSAIDs.Footnote ¹¹⁰ As David Healy argued in 1997, surveying the history of twentieth-century psychopharmacology before SSRIs, ‘many drugs follow a pattern of being hailed as a panacea at first, only to be later dismissed as poisons before achieving pedestrian status years or decades later’.Footnote ¹¹¹ SSRIs too had become pedestrian.

Doubt had once again transformed psychiatry: It had driven the evolution of new methods in data analysis and a recalibration of its research ethics. It had also helped to force a reckoning with the reductionist overconfidence of biological psychiatry. These criticisms would continue to shape the discipline in the coming years. Yet alongside it, another, less productive form of doubt began to gain prominence. Buoyed by this reckoning, some critics turned to older philosophical critiques which no scientific answer seemed to satisfy.

The new critics

By 2010, psychiatrists were calling for ‘a paradigm shift’.Footnote ¹¹² Many researchers had recognised the flaws in the DSM checklist criteria and began to argue for new diagnostic frameworks that recognised both the porous boundaries of mental illnesses and the need to characterise their heterogeneous underpinnings.Footnote ¹¹³ There were renewed calls for pluralism both in research methods and underlying philosophy.Footnote ¹¹⁴ This latter, more reflexive stance came to be called ‘critical psychiatry’ – consciously absorbing insights proffered by historians and social theorists, and their critiques of biological reductionism.Footnote ¹¹⁵

Yet critical psychiatry was a diverse movement. The term originated in the late 1990s from a group of British psychiatrists and professionals called the Critical Psychiatry Network (CPN): a UK-based ‘lobbying, campaigning and mutually supportive body’ that aimed to challenge the ‘hegemony of the now dominant biological view’ in psychiatry.Footnote ¹¹⁶ The CPN started as a psychiatric trainee reading group established by Joanna Moncrieff and others in 1997 at the UK’s Institute of Psychiatry. They invited speakers critical of mainstream psychiatry, including Thomas Szasz, the historian Andrew Scull, and the sociologist Nikolas Rose.Footnote ¹¹⁷ Szasz’s influence in particular weighed heavily on the group. In 1999, they formalised as an activist group in reaction to new mental health legislation that allowed compulsory community-based treatment with medication.Footnote ¹¹⁸

An NHS psychiatrist and professor at UCL, Moncrieff published theoretical work that drew on a range of philosophical influences. Citing Nikolas Rose and others, she developed an explicitly Marxist critique of psychiatry, arguing that it reinforces neoliberal political economy by framing economic non-productivity and broader human distress as individual pathology to be medicated.Footnote ¹¹⁹ Yet, once again illustrating the unlikely alliances within this debate, she also drew heavily on libertarian figures like Szasz and Breggin. She frequently recapitulated Szasz’s view of psychiatry as biologically untenable,Footnote ¹²⁰ and adopted Breggin’s claims about antidepressants functioning as mood-blunting or numbing agents.

In the CPN’s foundational paper, Moncrieff argued that, unlike the earlier anti-psychiatry movement, which ‘called for the abolition of psychiatry on what have been perceived as ideological grounds’, critical psychiatry’s goal was to offer ‘constructive criticism of clinical psychiatry’.Footnote ¹²¹ Although, as we have seen, self-identified anti-psychiatrists like Laing never called for psychiatry’s abolition but rather for its liberatory reform, it was Szasz (and Scientology) who declared it philosophically impossible and demanded its abolition.

Alongside her philosophical interventions, Moncrieff’s early work focused on challenging the evidence for antidepressant efficacy. Whilst at the Institute of Psychiatry, she was involved in a research unit that used new methods from evidence-based medicine. During this time, she published a meta-analysis suggesting that the efficacy of antidepressants was lower in trials with active placebos. She proposed that efficacy in other trials could be explained by ‘unblinding’, where participants realise they are taking a medication due to its side effects. Co-authored with the unit director Simon Wessely, the paper’s conclusions were highly cautious and inconclusive, reflecting the significant limitations of the data.Footnote ¹²² Later, however, Moncrieff adopted a more definitive stance, asserting that unblinding largely accounted for antidepressant effects in trials in her 2008 book, The Myth of the Chemical Cure. Footnote ¹²³ Studies subsequently challenged this view: trial data showed participants remained mostly unaware of whether they were taking the drug or not.Footnote ¹²⁴

In the same book, she also advanced a conceptual distinction between the ‘drug-centred’ and ‘disease-centred’ models of psychiatric medication. In the disease-centred model, drugs target an underlying pathology to relieve symptoms; in the drug-centred model, medications produce altered mental states – akin to recreational drugs or alcohol – that may mask or suppress symptoms. This was a reformulation of Szasz’s claim that psychiatry involved a categorical error: because mental disorders were not biologically based, drugs could not treat them in the conventional medical sense. Moncrieff’s argument was circular: If psychiatry lacked biologically defined disease entities, then psychiatric drugs could not treat disease, and could only produce drug effects. It also reified medications used elsewhere in medicine as always targeting pathology, even though many do not.

Her insistence on separating ‘mental states’ from biology also implied that any attempt to correlate conscious experience with brain functioning was untenable. In a later interview, she said: ‘we cannot pin down the biological mechanisms of complex human thoughts, feelings and behaviours … we cannot do this now and we have no indication that we will be able to in the future’.Footnote ¹²⁵ Once again, the Szaszian categorical dualism rendered psychiatry – and now cognitive neuroscience – essentially futile.

While her philosophical commitments were distinctive, Moncrieff’s scepticism about antidepressants was at that time reflective of the prevailing view. Rejecting the possibility of a biological basis for depression made her more inclined to doubt medication efficacy, but she was not alone in doing so. She continued to publish critiques of SSRIs and found common ground with Irving Kirsch. Together they co-authored several papers, combining her focus on unblinding with his on the clinical insignificance of antidepressant effect sizes.Footnote ¹²⁶

Other scholars had already begun to scrutinise how the mechanism of SSRIs had been communicated to the public. Research showed that many laypeople still believed that the ‘serotonin hypothesis’ was established science, due in part to aggressive marketing. By contrast, they found scientists and clinicians largely dismissed it as a simplification. Some commentators worried that this disconnect could cause harm, for instance, by fostering a reductive understanding of one’s condition.Footnote ¹²⁷ Moncrieff included these arguments in The Myth of the Chemical Cure, as well as an evaluation of the early research into the hypothesis. She argued that there was no evidence to support it and that its uncritical acceptance reflected a collusion between psychiatry and industry to distort public understanding.Footnote ¹²⁸

Meanwhile, other figures associated with the CPN began to gain public visibility. One was British psychotherapist James Davies, who had also been influenced by Peter Breggin. In The Importance of Suffering (2011), Davies argued that psychological distress should be seen as a ‘healthy call to change’ and not suppressed by medication.Footnote ¹²⁹ Like Moncrieff, he later adopted Marxian critiques of neoliberalism and psychiatry.Footnote ¹³⁰ In 2014, Davies and British aristocrat Luke Montagu co-founded the Council for Evidence-Based Psychiatry (CEP), with Moncrieff on its board. At the organisation’s launch in the House of Lords on 30 April 2014, they warned of rising antidepressant prescription rates, claiming that benefits were overstated, harms minimised, and scientific evidence lacking.Footnote ¹³¹ UK media coverage was extensive, with headlines such as ‘Antidepressants do more harm than good, research says’.Footnote ¹³² Figures from the CEP began making regular media appearances. Montagu shared his experience of withdrawal symptoms in The Times, whose editor he knew personally.Footnote ¹³³ In April 2014, Davies appeared on Sky News, asserting that antidepressants ‘work no better than placebo pills’ for most people, and that ‘most people taking antidepressants are not suffering a mental illness but rather understandable, albeit painful, human responses to difficult life factors that these pills may be exacerbating’.Footnote ¹³⁴

As the CEP’s media profile grew, there was concern that these figures were no longer offering critique but presenting themselves as experts to cast doubt on the effectiveness of antidepressants. Instead of focusing on legitimate uncertainty, they asserted with increasing certainty that antidepressants were ineffective and unsafe. As work by Cipriani’s Oxford group continued to reaffirm evidence of efficacy, this position diverged from most scientists’ understanding.Footnote ¹³⁵ Some observers argued that these activists were deliberately shifting the debate outside of scientific forums and into the public sphere.Footnote ¹³⁶ Despite this, the CEP received relatively little pushback from mainstream psychiatrists.Footnote ¹³⁷

Some figures associated with the CPN did continue to challenge evidence for efficacy in academic venues. Michael Hengartner, later a co-author on the 2022 ‘umbrella review’, published several studies and meta-analyses questioning antidepressant efficacy and safety.Footnote ¹³⁸ But by this point, the disagreement had become increasingly technical, with each side critiquing the other’s methodology, and decoupled from an intensifying public debate driven by media commentary and social media activism.

Then, in 2018, a new meta-analysis gained widespread public attention. Cipriani’s group, responding to criticisms of their earlier 2009 study, conducted their most methodologically rigorous analysis yet. Including 522 RCTs and over 116,000 participants, with extensive inclusion of unpublished trial data, they published their findings in The Lancet. Once again, they concluded that most antidepressants were modestly effective and well tolerated.Footnote ¹³⁹ It was the largest meta-analysis ever published in psychiatry. Despite some lingering criticisms, the scale and rigour of this article gave it particular weight among clinicians and researchers. It was regarded as a landmark contribution that further consolidated the consensus.Footnote ¹⁴⁰

In an interview, Moncrieff described the paper and its public reception as a ‘setback’ in a debate she felt ‘we were starting to win’.Footnote ¹⁴¹ Along with Davies, Montagu and other CPN associates, she published a response claiming there was ‘nothing new or clinically useful’ in the findings.Footnote ¹⁴² Their critique repeated earlier arguments – unblinding, publication bias – but neglected to comment on how Cipriani et al. had accounted for these concerns. Paradoxically, they argued that the paper showed antidepressants should be used ‘far less or perhaps not at all’. The certainty of this stance suggested they were unlikely ever to have been persuaded by its findings. The idea that antidepressants could be moderately effective was incommensurate with their belief that depression is not a meaningful disease category rooted in some element of biology.

The umbrella review

Shortly after the publication of Cipriani’s 2018 meta-analysis, Moncrieff assembled a team of like-minded researchers whom she said she ‘could trust’ to undertake a new meta-review on the serotonin hypothesis.Footnote ¹⁴³ They preregistered the protocol in 2020,Footnote ¹⁴⁴ although Moncrieff had already critiqued the early science underpinning it at length in her 2008 book, casting doubt on the idea that this was a disinterested scientific inquiry, as she would later claim in media appearances.

Why take up the issue again? As the position that antidepressants were effective became settled within the scientific community, Moncrieff sought to shift the debate towards the unresolved question of their mechanism of action. For most scientists, these were distinct issues. Many drugs are used clinically despite an unclear or evolving understanding of how they work. But for Moncrieff and her collaborators, the lack of an established mechanism was itself discrediting. There remained widespread public belief in the ‘low serotonin’ theory, and the review was intended to provoke debate about the paucity of its evidence. This created an opportunity to reiterate the group’s longstanding claims: that there is no meaningful biological basis for depression; that antidepressants are better understood as recreational drug-like substances rather than disease-targeting treatments; and that any perceived efficacy could be attributed to unblinding or bias.

Yet for most researchers and clinicians, the inconsistent evidence around serotonin’s role did not invalidate the possibility of a biological underpinning. Although no simple correlation between serotonin levels and depression had been found, broader evidence implicating neurotransmitter systems continued to motivate investigation. The consistent clinical efficacy of SSRIs was itself indirect evidence of serotonin involvement. Drawing on decades of research, scientists now framed depression as a disorder of brain networks, where negative self-beliefs and aversive behaviours became increasingly reinforced, producing distress and disability.Footnote ¹⁴⁵ By increasing levels of neurotransmitters, antidepressants were thought to work by promoting plasticity and the updating of these fixed beliefs.Footnote ¹⁴⁶ These ideas echoed neuroscience’s broader paradigm shift – from ‘chemical brain’ to ‘networked brain’ – and reflected a form of ‘normal science’: the positing of provisional models revised in response to new evidence. Moncrieff and her co-authors did not engage with these new theories.

This helps explain the disconnect between the scientific community’s puzzled reaction to the paper and its authors’ own assessment of its significance. When the ‘umbrella review’ was published in Molecular Psychiatry, a prominent Nature Portfolio journal, on 20 July 2022, it drew intense media attention. The article summarised 17 studies on serotonin levels and activity in the brain – including blood serotonin, tryptophan depletion studies, and genetic data – and concluded that these provided no support for the serotonin hypothesis.

But many scientists challenged the paper’s ‘adamantine certainty’.Footnote ¹⁴⁷ A group of 36 experts published a formal rebuttal, highlighting what they saw as serious methodological flaws.Footnote ¹⁴⁸ Responding to Moncrieff’s claim that the paper was an ‘objective’ systematic review, they argued that it was selective and incomplete, since it cited only 17 of over 360 potentially relevant studies. For example, it ignored several tryptophan depletion studies showing that previously depressed patients could relapse when serotonin levels were lowered. Critics also challenged Moncrieff’s public characterisation of the findings as ‘new’, when in fact the review relied largely on historical literature. And, as another response pointed out, the serotonin hypothesis comprised multiple related propositions, not a single falsifiable claim.Footnote ¹⁴⁹

As these methodological critiques were being voiced in academic forums, a very different – and disconnected – debate had begun to emerge in the media and online. Whilst the paper itself was careful to avoid making claims about efficacy, a press release written by Moncrieff and co-author Mark Horowitz explicitly linked the umbrella review to a lack of efficacy in antidepressants. Press releases, unlike academic articles, do not go through peer review. Even so, staff in the UCL media office were uneasy about including comments on efficacy as this diverged from the article’s content. They worked with the authors on a revised version. However, the authors pre-emptively circulated their own version that included comments on efficacy.Footnote ¹⁵⁰

The article was quickly picked up by media outlets across the globe, far outstripping earlier CEP efforts. Within days it had gone viral: it was read over 413,000 times in the first few weeks and became the 217th most-shared scientific paper ever recorded by Altmetric.Footnote ^{151 ,} Footnote ¹⁵² In media appearances, Moncrieff frequently suggested that the paper discredited the efficacy of antidepressants. She appeared on ITV’s This Morning and the right-wing news platform GB News, claiming the umbrella review showed there was ‘no biological cause for depression’ and that this meant that antidepressants had little or no benefit.Footnote ¹⁵³

When some scientists attempted to clarify the distinction between mechanism and efficacy, Moncrieff responded by claiming that researchers were backtracking, that they had long promoted the serotonin theory and were now dishonestly claiming it had been abandoned years earlier.Footnote ¹⁵⁴ She thus recast clarification as evidence of deception.

As the media coverage continued, the paper was widely reported as showing that antidepressants do not work – a view encouraged by its authors. It fuelled a vast amount of negative discussion about antidepressants and their efficacy in US right-wing media outlets, podcasts, online blogs and social media (it was referenced nearly 9,000 times on Twitter by 2023).Footnote ¹⁵⁵ Even as the paper itself faded from view, it left a widespread public perception that scientists disagreed among themselves about the efficacy of these medications. Tarred by this obfuscating sense of controversy, psychiatrists struggled to offer convincing correctives. And the profession was soon challenged by Secretary of Health and Human Services Robert F. Kennedy Jr., who took up antidepressant use as a cause célèbre in his Make America Healthy Again campaign in 2026, referencing the review’s authors.Footnote ¹⁵⁶ By promoting this disconnect between the paper’s findings and its portrayal in the media – and between the scientific consensus and public understanding – the ‘umbrella review’ thus became a powerful tool for the manufacture of doubt.

Reasonable and unreasonable doubt

‘Doubt is crucial to science’, Oreskes and Conway argue: ‘healthy scepticism drives science forward’.Footnote ¹⁵⁷ Disagreement among scientists is extremely common, even when reviewing the same evidence. Do the debates around the efficacy and use of antidepressants represent healthy, reasonable doubt?

There was a great deal of scepticism about the usefulness of SSRIs both within and outside psychiatry after their early years of hype. This doubt provoked intensive scientific inquiry: it led to further trials, including the largest post-marketing clinical trial ever conducted in psychiatry (STAR*D). It generated many meta-analyses of these clinical trials that contributed to innovation in the methodologies of meta-analysis more broadly. It helped to establish new ethical standards about the pre-registering of trials that are now the benchmark across the whole of academic medicine. Despite their many controversies, SSRIs survived intensive scientific scrutiny and re-analysis. Today, every major expert body and evidence-based guideline endorses them as an effective and appropriate treatment for depression.Footnote ¹⁵⁸

Yet, recently, there has emerged a widespread public misperception that they do not work, that they are ‘no better than placebo’ for most of those who take them. Claims like this are regularly repeated by journalists and media commentators, and circulate widely on social media. This paper has argued that figures associated with the Critical Psychiatry Network have helped to manufacture this doubt. They rejected the growing scientific consensus and, rather than contesting the evidence for efficacy directly, shifted focus to mechanism of action, hoping that uncertainty in this domain would undermine confidence in the drugs’ broader use. The ‘umbrella review’ did not engage with the ongoing internal scientific debate about the aetiology of depression or recent theories of antidepressant action. Most of its authors have no prior expertise in these areas. Instead, they aimed their challenge at an outdated mechanistic theory and misrepresented the paper’s findings in the media by suggesting that this challenged the evidence of efficacy. By arguing their paper supported claims that it did not, they bypassed scholarly debate altogether. In doing so, they adopted the stance of scientific sceptics, but not the practices of scientific inquiry.

When should doubt raise suspicion? When it persists in the face of further scientific inquiry, when it ignores the consensus that this inquiry generates, when it is promoted primarily outside the normal sites of scientific debate, and when it obscures or denies the broader balance of evidence. The authors of the ‘umbrella review’ demonstrate many of the hallmarks of the manufacture of doubt, as articulated by Oreskes and Conway. Where these hallmarks are found, they prompt us to consider the underlying commitments that may motivate them.

In her writings, Moncrieff articulates a consistent ethical and political critique of psychiatry: that it is coercive, that it medicalises social suffering, and that it reinforces neoliberal logics of personal responsibility. Yet as we have seen, these concerns are not inherently anti-scientific. Many psychiatrists, researchers, and historians share them, and such critiques have played a valuable role in shaping both policy and clinical practice. Indeed, they are deeply intuitive if you hold the kind of liberal values common in Western cultures. What distinguishes Moncrieff is her philosophical stance, inherited from Szasz, that there is a fundamental and irreducible distinction between the realm of human experience – which encompasses all mental phenomena including mental illnesses – and the biology of the brain. My aim here is not to refute this position but rather to draw attention to its implications. Holding it seems to significantly impair one’s ability to engage with scientific inquiry into psychiatric conditions, especially the careful and open-minded testing and updating of hypotheses that Oreskes and Conway argue characterises normal science. Strongly endorsing this stance has meant that, throughout her career, Moncrieff has repeatedly rejected scientific findings that support any such biological underpinning. It has also meant she has rejected the idea that psychiatric medications can meaningfully treat mental diseases.

However, we cannot expect scientists (or indeed anyone) to be free of prior moral or political commitments. This issue has been widely discussed in the philosophy of science: The philosopher Heather Douglas has argued that it is neither possible nor desirable for scientists to exclude ethical or social values from their work. Scientific hypotheses often carry consequences that extend beyond the lab, and so decisions about which theories to pursue, accept, or reject will inevitably involve normative judgement.Footnote ¹⁵⁹ Psychiatry is an exemplary case of this principle: The consequences of its conclusions are far-reaching, and it benefits from scrupulous and value-driven practitioners to guide its research. Many scientists working in psychiatry have ethical critiques to make of it. Yet only for some do their values lead them to persistently discount new evidence in this way.

Others have elaborated this point with reference to Bayesian probability: that all empirical assessments we make necessarily take into account both current evidence and prior experience. Yet this position sets up a tension as Bayesian priors are also correlated with a degree of certainty: If one’s prior beliefs are held with a very high degree of certainty, then one is much less likely to allow novel data to challenge them. So whilst recognising the inevitability and normative benefit of our non-epistemic values and their involvement in our work, I would also suggest that the degree of certainty with which those priors are held is key in their evaluation. If one holds a philosophical position that means excluding the possibility of a certain kind of evidence – that depression might be a disease with a biological cause – then you will be highly predisposed to dismiss supporting evidence as weak, inconclusive or flawed. Holding such a prior strongly thus appears to impede one’s ability to engage with balanced scientific inquiry.

Where do we draw the line? In this paper, I have tried to mark a distinction between two kinds of doubt in science. Many of the critiques made of psychiatry have been both effective and productive: earlier criticisms contributed to the de-institutionalisation movement and shaped the laws that regulate the use of coercive practices in psychiatry. In the last few years, legal debate around human rights law led to the WHO recently calling for an end to coercive practices as an infringement on human rights. Critiques of biological reductionism have led to changes in diagnostic frameworks and promoted calls for more philosophically grounded approaches.Footnote ¹⁶⁰ This productive kind of doubt allowed for a degree of eventual convergence between science and its critics: scientists had to respond – to change their methods, beliefs and attitudes.

Another kind of doubt appears, however, singularly unproductive: attempts to address concerns, through further research into efficacy, provoked only louder and more vehement critique. Their doubt appears unresolvable: to demand an unreasonably high burden of proof. This incorrigible form of doubt seems to be driven in this instance – and perhaps others too – by strongly held philosophical beliefs that are irreconcilable with the premise of scientific inquiry. It drove these critics to repeatedly question emerging evidence and, eventually, disengage from scholarly debate altogether and redirect their efforts towards the public sphere. In their ever-greater efforts to discount new evidence, they turned to the tactics of manufactured doubt.

What are the consequences of this position? Oreskes and Conway argue that unreasonable doubt can be harmful to science, as evidenced by the climate change denialism that was propagated by contrarian scientists in the twentieth century. Can we find examples of such harms here? After all, antidepressant prescription rates have continued to rise.

However, as the philosopher and disability-rights campaigner Robert Chapman has argued, insisting on the invalidity of psychiatric diagnoses and treatments risks epistemic harm to marginalised groups, including disabled and neurodivergent individuals who would otherwise find value in them as a means to overcome stigmatisation and source agency.Footnote ¹⁶¹ Drawing an equivalence between antidepressants and recreational drugs stigmatises those who use them. This unreasonable doubt has been remarkably effective at generating controversy, evidenced by the widely held belief that the science remains unsettled. In promoting a disconnect between scientific and public understanding, it has impaired public trust in psychiatry more broadly, and risks alienating individuals from the mental health services that support them.Footnote ¹⁶² Finally, unreasonable doubt may also impoverish psychiatry of the benefits of productive external critique which, given the legal powers psychiatrists can employ and the vulnerability of the patients they manage, may be of particular importance to it.

Figures from the CPN have responded to critiques by pointing out the harms that antidepressant medication can have, such as withdrawal effects, which they argue are under-recognised in psychiatry. They make an important point that continues to generate further inquiry. It is an example of how figures associated with the CPN have used such activism to spur on meaningful scientific inquiry, including the publishing of new guidelines on antidepressant withdrawal.Footnote ¹⁶³ However, whilst research to determine how common these experiences are is ongoing, there are also risks of over-emphasising the potential harms of antidepressants. Arguing that SSRIs can cause aggressive states that motivate school shootings, as Peter Breggin did, seems to be one such example. It is irresponsible not to represent the balance of risks and benefits that these medications offer.

And again, there is a question of motivation. Many figures in this debate, like Luke Montagu, are motivated by personal experiences of distressing side-effects, experiences they felt were under-recognised and mismanaged by the doctors who prescribed these medications. Yet, at the same time, others are motivated by the political position that psychiatric drugs are a coercive tool that suppresses individual selfhood and reinforces neoliberal hegemony. If one’s priors are that antidepressants can have no meaningful benefit as a result of these beliefs, then they will likely bias judgement towards over-emphasising the harms of antidepressants. So, whilst acknowledging the very real possibility of unrecognised and poorly managed withdrawal effects, I want to draw attention to the potential harms for these same patients in using the issue to further delegitimise antidepressant use when one is motivated primarily by a strong belief that they do not work.

This debate speaks to a broader tension between prior beliefs and scholarly inquiry. Many of the activist figures in this story are also published historians of psychiatry, including Moncrieff and David Healy. This reflects the broader critical stance that historians of psychiatry have taken towards their subject matter.Footnote ¹⁶⁴ As historian Jonathan Sadowsky has observed, the historiography has often framed psychiatry as a ‘special case’, repeatedly emphasising its ignorance or failures whilst obscuring its therapeutic benefits, with few engaging with the question of efficacy at all.Footnote ¹⁶⁵ We should note that the philosophical and political positions that drive the critics of psychiatry have influenced some historians at times, arguably leading them to recapitulate in their histories their own strongly held priors.

More broadly, historians of science and scholars of science studies are notable for the critical stance they take towards their subject matter, driven in part by a sense of unease at the normative authority that science has held for a long time. Yet as science has been seen to come under threat, prominent scholars in this field, such as Latour and Oreskes, have drawn attention to the risks of such critiques.Footnote ¹⁶⁶ Critique can go too far, and unreasonable critique should be called out, as science has a degree of fragility and can be undermined. There is then an essential tension between the position taken by scholars like Oreskes, who give a normative account of science as a privileged domain of inquiry by nature of its method and organisation, and others who argue that there is no such distinction to be made.Footnote ¹⁶⁷

Psychiatry and its critics are interesting examples to introduce into this debate. On the one hand, it points to some of the limitations of the analytic that Oreskes and others offer: it is far easier to critique unreasonable doubt where those promoting it are more obviously fringe, due to right-wing libertarian politics, for example, or through their association with financial interests. It is harder to critique it when the motivating ideologies are widely held – as they are among psychiatry’s critics who, as we have seen, are drawn from across the political spectrum by an underlying liberal suspicion of an institution that seems to threaten individual liberty (however that liberty is coloured politically).

As I have tried to show, liberalism continues to sit uncomfortably with any medically coercive practice, yet the critiques it motivates have both benefits and harms. One such harm has been to either obscure or romanticise suffering from mental illness, in particular severe mental illnesses which are profoundly debilitating and deeply stigmatised. Support for people with these conditions remains disproportionately under-resourced. We need critiques that will help this, not exacerbate it.