[article]

Folie à deux: the madness of two

4 Jun 2026

Editors note: republished from 2020 to 2026, after a pretty substantial revision. Worth knowing that this was written in September of the COVID pandemic though. Wonder why this was on my mind…

Folie à deux, or the ‘madness of two’, is the kind of psychological phenomenon that occasionally captures the imagination of the media. Two people, otherwise normal, suddenly go insane.

A useful, and fairly famous example happened here a few years ago. Two Swedish sisters who recently, travelling through Ireland into the UK, first started behaving erratically, eventually ending in murder.

The twins, as they travelled south, were first abandoned by their coach for their odd behaviour—suspicious demeanour and a weird fixation on their bags. The service station manager at the place they were left was so concerned, he called the police. Released after some questions, they went on to wander down the motorway, occasionally trying to sprint across it, insensitive to the passing cars. Again, the police were called to talk to the pair. This time, without any apparent trigger, first one sister, then the other, sprinted again into traffic, getting badly injured enough to be hospitalised. As they were being treated, on the way to the hospital, they were recorded saying things like “I recognise you—I know you’re not real”, and “They’re going to steal your organs”. One of the pair recovered quickly enough to be released from hospital. She spent the night wandering the streets, before being taken in by a couple of passers-by. Friendly initially, though still behaving strangely—claiming their cigarettes were poisoned for example—she stayed the night. The following day, around dinner time, she stabbed one of the passers-by to death before running into the night, beating her own head with a hammer she’d obtained along the way.

No drugs. No apparent history of mental illness. No criminal convictions. Two ordinary women, one with a family. A few days of madness that ended in tragedy.

It’s a premise that we can ghoulishly enjoy from afar because it seems like it could never happen to us. But I’m not so sure. From intra-family murder to Theranos to our own odder moments, I think shared madness is something that is much more common than you’d think.

¶A brief history

In the 1860’s, neurologist Jules Baillarger initially used the term folie communiqué, or ‘communicated madness’, to describe a couple so caught up in each others’ delusions physicians were unable to tell who became psychotic first. By the last 1870’s this phenomenon had become known as folie à deux through the work of Charles Lasegue and Jean-Pierre Farlet, doctors of some renown at the time.

Typically, these doctors observed a relationship in which the more dominant figure would be experiencing delusions of some kind and the other person would come to believe the delusions, or even perhaps begin to share them. But occasionally, rather than folie imposée, where one person’s set of delusions are ‘imposed’ on the other, we might see folie simultanée, in which both parties may already be delusory individually, but as they interact, their delusions coalesce. And folie did not need to come in pairs. We also had folie à trois, the madness of three, or folie à famille, the madness of a family. Eventually, the romanticism was put aside, and all related cases were placed under the banner of ‘shared psychotic disorder’. In each of these cases, the delusions of one person comes to drive the behaviour of others.

¶Are we all at risk of folie?

One would kind of imagine that such a phenomenon is kind of uncommon. The sharing of some psychotic episode seems like something from TV. But I’m not so sure. The literature on the subject is not particularly busy. But we might expect that: there are fairly stringent requirements to conclusively diagnose a case. There needs to be a commonality of delusion (of course), intimacy in the relationship, and acceptance or support of each others’ delusions. The details of these features aren’t always apparent to an outsider. Nor are closely coupled people often willing to share the intimacies of their connection to another. The literature does indeed occasionally complain about this:

Patients with shared psychotic disorder can go undiagnosed because only the primary partner is registered for treatment in a classic presentation … Awareness of the nature of the dyad relationship dynamics is necessary.

Clinicians can’t identify cases if they aren’t even aware that others shared the thing in the first place. So the literature leaves us a little cold. We are necessarily biased by reporting, and reporting will often only ever present newsworthy cases.

But there is something interesting in the cases we do see. We see examples in places as far flung as Saharan Africa, in India, and Japan. Folie, then, is not the invention of the U.S. media. If not common, it is certainly diverse—there isn’t a cultural limit to the sharing of madness.

And indeed, there is a common thread; intimacy in the context of social isolation. This is indeed the original qualification for Lasegue and Falret’s syndrome. Folie à deux is perhaps the most intriguing example of the crucial nature of our need to be social. It exemplifies, in its own puzzling way, the lengths the mind will go to to feel connected to something bigger. It’s also emblematic of the myriad peculiarities of love.

There is always some madness in love.

Nietzche, Thus Spake Zarathustra

¶A ‘spectacularity bias’ hides how common it really is

A spectacular case of folie à famille occured in Australia recently. In this case, the family fled their farm for hundreds of kilometres in apparent terror of some following threat. Some family members escaped the chaotic week-long journey unaffected, but others seemed to be in various altered states. Like our Swedish sisters, this episode was preceded by no apparent drug use or history of mental illness. And the members of the family reporting afterwards seem as puzzled by the episode as anyone. For example:

It’s very confusing, I still feel confused

Or from another:

You do start thinking the same way… you can get sick in some way

Even more recent were the Burari family deaths in India. Eleven members of a family dead by apparent suicide, their faces swathed in tape and cotton and cloth. Indeed many suspected cases of shared folie regard people who died doing something baffling. These may not be cases of shared delusion, but the number of suspected cases does not line up well with the number of reported cases.

In larger numbers, this makes one wonder how many of these members shared the delusion, and how many simply went along (or were forced along) for the ride? In the case of the Australian family, one member, Mitchell, explicitly reported that he did not share his family’s fear. He simply “couldn’t leave them”. This kind of half-hearted participation would surely make it hard to diagnose a shared madness.

Again, given the stringent criteria for diagnoses, one wonders how many examples simply escape our notice, particularly outside of the more focused dyadic cases. If shared madness doesn’t just include those who went along because they believed the delusions, but also people who just stayed, with no particular belief of their own, the possible pool grows quite large indeed.

The second feature to consider is that many, many cases of shared madness involve violence. Murder, suicide, and injury. We might even say that these cases are overrepresented. Psychoses run the full gamut of bizarre perceptions and beliefs, not simply those that end in pain. So why is shared psychosis so much more harmful?

The answer may be that perhaps it’s not.

¶Selection bias and mundane conformity

It took me a long time, researching this article, to come across cases of shared madness that weren’t obviously harmful. You have to dig into academic work for that. When you do, you find that non-violent shared delusions are actually pretty common in clinical settings. For example delusional parasitosis—where the belief is around some kind of infestation—is pretty frequent. Here, for example, a family of four all eventually came to share the psychotic hallucinations of the youngest son—an idea that:

“jelly-like substance or amoeboid materials come out from my mouth, and people around me react to those materials”

First the son, who did show some of the symptoms of schizophrenia, but then his mother, later his twin brother, and finally his father, all reported that they could see these ectoplasmic emissions, and had observed the effect they had on others in the form of coughing and drooling and so on.

Another type of folie I encountered was shared pseudocyesis—a belief in a phantom pregnancy. Apparently particularly common in developing countries, and particularly in couples suffering from infertility, shared phantom pregnancy can go so far as to cause pregnancy-related physical symptoms.

Violent cases of shared madness are, in fact, thought about as more of a subset than a feature. This paper, for example, is interested in shared religious delusions—the actual category—that led to fatal, or near-fatal outcomes.

The fact of the matter is that the only cases which enter the public consciousness are those which have some spectacular end. Something obviously newsworthy, rather than merely interesting, strange, or tragic.

What’s a little strange, though, is that the examples which dominate the googlable headlines, and thus the collective consciousness, are still weighted toward violence. I suspect that this is because psychosis is a fairly loaded term. It implies violence. When violence isn’t in the picture, we have trouble identifying it for what it is. And I think can show you.

Theranos was a multibillion dollar biotech startup headed by the now infamous Elizabeth Holmes. The company raised huge sums with claims of a revolutionary blood test. The bizarre antics of Elizabeth Holmes and her boyfriend/business partner Balwani have become jokes, after the downfall of the company. A pretend wolf permitted to roam the sterile halls of the company, freely shitting anywhere it likes. Holmes speaking in different voices. Rampant paranoia.

Theranos is typically reported as a straightforward case of a con enabled by a credulous, hype-drunk industry. Silicon-Valley hubris. But there might be another explanation.

On page 143 in the book Bad Blood, the author reports suggestions that this may have been a case of folie. The couple, Holmes and Balwani, may have been living some elaborate shared delusion. A delusion that sucked in thousands who, like Mitchell from the Australian family earlier, were not necessarily sharing the delusion but simply “couldn’t leave” for one reason or another. A couple of true believers, and a horde of ‘stayers’.

These kinds of examples aren’t easy to read as shared psychosis, because they’re messier. They don’t fit the pattern we associate with psychosis. But they could be. One of Holmes’ defences was the claim that their relationship was coercive and abusive. True or not, there’s certainly evidence of intense and emotional intimacy, and the dynamics at play around the couple are clearly an example of group polarisation. We know that groups can easily push their members to do more than each would do alone.

But a different pattern satisfies explanation—the antics of tech elites—and so, folie is ruled out in favour of mundane conformity.

¶Outro

There are many cases of shared delusions at scale which seem, at least superficially, related to folie. Internet phenomena like gang-stalking or Morgellons. Many of the reported cases of mass psychogenic illness. Many of these involve the kinds of social isolation and intense relational dynamics characteristic of folie. Rarely are these factors considered, because they fit other patterns more neatly.

Which leads me to the question that spurred my writing this article: how many cases of folie are missed, simply because they aren’t spectacular enough? Or put another, more interesting way, are the more spectacular nature of the reported cases distracting us from something more fundamental?

Our little survey alone shows us that cases can become invisible if any of three things aren’t spectacular enough. If the beliefs aren’t sufficiently strange; if the individuals aren’t clearly isolated enough; and if it never received any kind of clinical or legal attention. If any of these three don’t meet a high enough threshold, then we choose another pattern to explain things.

I said before that the common thread in cases of shared madness is intimacy in the context of social isolation. Social isolation is an unexpectedly dangerous thing for humans. I’ve said it before, but loneliness is emerging as one of the single greatest threats to human health, both mentally and physically. When we combine isolation with new intimacies, perhaps it shouldn’t surprise us that things can start to look a little strange. Given the widespread nature of social isolation, I think we can be sure that the faces of shared madness are more varied, and more subtle, than we can imagine.¹

It’s certainly a reason to be cautious. There’s every chance we, too, could become one of the spectacular cases.²

Anthologies: Connection, Successful Prophets, On Friendship, On Love, On Leadership, On Politics and Power

View on main site »