No active ingredients whatsoever

Placebos and the voluntary suspension of disbelief

I have always been fascinated by placebos. Dualistic philosophy, individualistic medicine, and a strictly mechanistic understanding of how the world works all appear to break down in the face of what placebos accomplish through some mystical efficacy of belief and self-delusion. No mechanism has been definitively identified to explain how placebos work, and no conditions have been found that always prime or trigger their effects, in either positive or, in the case of “nocebos,” negative directions. With placebos, the routines of providing care prove to be as significant and effective as pharmaceutical substances in and of themselves; the substances become arbitrary props in therapeutic rituals that draw on some immaterial power of collective attention to reshape somatic experience.

Placebos lend themselves to dramatic, anecdotal tales of mind over matter. A recent review essay in the New York Review of Books by Gavin Francis, a medical doctor, cites several striking examples of the imagination’s role in how we experience pain and illness. He notes “the case of a 29-year-old builder brought to Leicester Royal Infirmary in 1995 with a six-inch nail through his boot.”

Any attempt to remove the nail caused him to scream in agony. A clinical team heavily sedated him, pulled out the nail, and removed the boot—only to find that the nail had passed harmlessly between his toes. His foot might have been unharmed, but there was no doubt among the team that he had been experiencing real pain.

The claim that “there was no doubt” about the “real pain” is raising some questions already answered by the claim. What makes pain “real” and do what extent does that depend on other people recognizing it? When is pain a social phenomenon, susceptible to being managed collectively even if it is unevenly distributed among the collective? What if even our pain doesn’t really belong to us?

Francis describes how research has found that describing medications’ side effects to patients makes them far more likely to experience them — this is especially true of impotence, “the placebo-responsive condition par excellence” — and that the patient’s perception of whether their doctor “exuded warmth and competence” can change how effective a placebo will be. He cites a statistic from one of the books under review, Jeremy Howick’s The Power of Placebos: How the Science of Placebos and Nocebos Can Improve Health Care, that “annually around £67 million in the U.K. and $5 billion in the U.S. are spent on knee washout arthroscopies for osteoarthritis, yet it has been shown that simply anesthetizing the patient and giving them a scar on the knee so that it looks as if they’ve had surgery is just as good at reducing subsequent pain.” Francis invites readers to consider whether the fake scar is more ethical than the genuine surgery.

Given that the fluid situation established in the theater of medical consultation can have such a significant impact on a placebo’s outcomes — and that these oucomes defy calculation and prediction — the scenario can be likened to playacting, a kind of improv. Every doctor-patient meeting is a highly contingent, singular encounter, mediated through an object, the placebo, whose particular qualities convey a variety of intentions and potential effects in indirect and unpredictable ways, which defy ordinary notions of causality. It appears to lend itself to a dramaturgical analysis in which the whole mise-en-scène proves to matter more than any particular isolated element; no single part is as important as the volatile relations between them. Each aspect contributes in some difficult-to-isolate way to the effect of the whole, complicating any simple divide between thought and material feeling.

Unlike with conventional medications, the purely formal qualities of the placebo matter more than their literal content: “We know that expensive placebos work better than cheap ones, capsules work better than tablets, and colored capsules work better than white ones. Blues and greens work better as sedatives, while pinks and reds work better as stimulants and painkillers,” Francis writes.

What makes matters even more complicated is that patients can still benefit from placebos even when they are being given one. They don’t necessarily work through straightforward deception or trickery. As Francis explains:

Placebos seem to work on the basis of expectation, and one of the most curious aspects of their function is that they continue to have benefits even when you know that what you’re taking is a dud. These “open-label” placebos are nevertheless commercially valuable: some I looked up recently retail at more than $100 a bottle, though the label reads, “No active ingredients whatsoever.”

Placebos have no active ingredients, or rather they shift the activity elsewhere, to a different plane where causality can’t be traced through empirical measurements.

One could perhaps call this plane “aesthetics,” and liken placebos to works of art, in that they require a certain set of social relations to “work” and they resist reductive explanations of their efficacy. Placebos have always seemed to me to offer a way to understand the experience of art, which can be construed as a means to accessing that social power through rituals of concentration and focus. Experiencing real relief from a sugar pill is not so different from crying real tears at the theater, or more to the point, the trials and tribulations of Pamela or Clarissa, which didn’t even appear on stage but in a book. Many commentators in the 18th century were perplexed or astonished at how this could be such a widespread phenomenon — how so much emotional reactivity could be experienced privately, and be contained in a portable commodity that was just made of words.

Years ago, as a student, I had an idea to write a placebo-based theory of the rise of commercial fiction that assessed novels as a kind of quack medicine, and treated the suspension of disbelief as a example of, if not a gateway to, placebo effects — a means of harnessing them if not entirely rationalizing them. I was especially interested in how they often tried to dramatize their own effects, with descriptions of what effects reading was supposed to have on people (epistolary novels especially play on the scene of reading), and how crying over various spectacles (or descriptions of spectacles) proved you were imbued with the appropriate quantum of “sensibility.” This sets the stage for a whole new genre of advertising, just in time for a rapidly expanding publishing industry: Being able to cry when a book cues you to is pretty similar to being able to feel desire when an ad cues you to.

In some respects, placebos reflect the idea from Kant’s aesthetics that artworks suspend the definitive assignment of concepts to perceptions, and purposes to actions, but instead hold open a space where the imagination can operate — making for a kind of healthy exercise for a faculty of the mind that ultimately makes possible the connections between what we experience with how we understand and communicate it — i.e. linking mind and matter, words and things. “There can be no rule by which someone could be compelled to acknowledge that something is beautiful,” Kant claims in Critique of Judgment — beauty, like placebos, can’t be entirely rationalized. But making a claim of beauty affirms what Kant calls the “sensus communis,” the sense that what we feel is not a private and sealed-off thing but something fundamentally shared and sharable, meaningful because social, even if it can’t be articulated as a concept or be put into specific words.

Placebos are reminiscent of this, in that they are a means by which we experience a social but irrational phenomenon that suggests something of our common capacities as humans. Their efficacy can’t yet be reduced to a set of glandular secretions or prefrontal-cortex activity or whatever. When placebos work, it points to the patient’s having had some inarticulate intuition of what “being healthy” requires that, at some indefinable level, finds support in the way others have treated them or how others behave and respond to the world. Placebos manifest a sense that human experience is commensurate only with itself but is shared nonetheless, and that mystifying fact, that other beings think and feel as we do, intimates that human experience can reorient itself in its own self-understanding toward a collective sense of what is good.

That probably sounds like a lot of metaphysical nonsense. But placebo effects (like the reality of art) compel us to recognize other kinds of causality that draw on social forces in obscure ways. Placebos aren’t dependent on some automatic response compelled by physics, or some form of medicinal manipulation that necessarily occurs behind their backs. They are instead bound up with the willing suspension of disbelief, a willingness to have an essentially aesthetic experience, an openness to being moved. If the patient is in the right frame of mind, and everybody involved performs convincingly, real catharsis can come from a staged activity, a known simulation.

Francis provides this example from Howick’s book:

Howick describes a personal communication from an Italian doctor who gave a wealthy woman with a backache an injection of distilled water, which cured her pain: “She was so satisfied that seven days later she called me for the same reason and demanded the same solution, ‘which had done me so much good.’” As he was filling his syringe with water, the woman called out, “Doctor, is it a placebo that you are giving me?” He told her it was, and she replied, “Thank goodness. It helped so much last week.”

Would the placebo have still worked if the woman started injecting herself with water? Or was the placebo just a convenient fiction around which to establish routines of care, a means through which concern could be communicated and materialized, a kind of secularized and medicalized communion wafer?

The ambiguity of placebos suggests some complications for the often-hyped future of AI-assisted medicine. Placebos prove that simulations can supplant “real medicine,” so maybe word-generator therapists can be just as good as “real” psychologists. But placebos also suggest that he the medicine may only be a pretense for enabling social relations of care that are required for the patient to heal.

Placebos cast doubt on the idea that there are abstract cures, or that decontextualized medical knowledge can be extracted by statistical or other means and simply be transmitted and applied to address health issues by whatever means are expedient. Placebo studies suggest that a range of social and situational factors play into outcomes, and that abstracting treatments away from those factors is futile. If placebos are effective because they concretize successful social relations between patients and doctors, other caregivers, and the general ambiance of the medical apparatus that helps establish the tenor of those relations, how could AI medicine not undermine all of that? It eliminates the social relations involved, abandoning people to an indifferent piece of software.

But at the same time, “AI” opens up a different avenue for a new encounter with fiction, presenting a new opportunity to willingly suspend disbelief and interpret a machine’s outputs as caring concern (if not raw, relentless sycophancy). If you choose to believe that AI therapy can work, does that very choice make it so? The entire interaction with a chatbot could become placebo-like if situated within a particular system of care delivery that a particular patient finds convincing, persuasive, illustrative of some “advanced knowledge” being marshaled for their benefit — perhaps even for their specific benefit alone.

That may lead to a soothing sense of being uniquely cared for, or it could lead to paranoid delusions of grandeur (or both!). In the New York Times, Kashmir Hill recently reported on ChatGPT users who become convinced that they are tapping into secret knowledge and have been appointed by an all-knowing entity to take on a messianic mission.

People claimed a range of discoveries: A.I. spiritual awakenings, cognitive weapons, a plan by tech billionaires to end human civilization so they can have the planet to themselves. But in each case, the person had been persuaded that ChatGPT had revealed a profound and world-altering truth.

Just like the guy screaming in pain from the nail that didn’t actually touch his foot, there is “no doubt” that these people are having “real” experiences with chatbots, feeling “real” feelings. In a sense, these kinds of delusions are placebo effects taken too far; they suggest how people, given a potent enough mirror, can talk themselves into the reality of all kinds of apparent insights and experiences. The automated chatbot responder — or as Hill aptly defines it, the “word-association machine” that pulls users “into a quicksand of delusional thinking” — allows for a hyper-stimulation of the willingness to suspend disbelief, overdeveloping the imagination until it no longer has any relation to the sensus communis. Ryan Broderick describes this sort of thing as “emotionally psyopping yourself with AI”:

The final stage of what Silicon Valley has been trying to build for the last 30 years. Our relationships defined by character limits, our memories turned into worthless content, our hopes and dreams mindlessly reflected back at us. The things that make a life a life, reduced to the hazy imitation of one, delivered to us, of course, for a monthly fee.

Without a social context that can contain the delusions that chatbots produce in their willing interlocutors, they end up engaged with boundary-less simulations that undermine the reality of whatever concepts or subjects they come into contact with, urging on us the suspension of disbelief with respect to everything. Maybe that means seeing the whole world as an entertainment product made for your lonely consumption; maybe that means treating your phone as a potential spouse; maybe that means believing that the “Guardians” have chosen you to receive “interdimensional” messages. None of it seems likely to be very healthy.