The Whiteness of Lies

The body’s capacity to respond to “mere” symbols of medical treatment as if they were “real”—what we call the placebo effect—has become a hot topic. A couple of years ago, a high-profile article in the New York Times trumpeted that “placebos prove so powerful even experts are surprised,”¹ setting off many copycat media examinations of the issue. A few months earlier, two psychologists had tossed a minor bombshell into the psychiatric community with their analysis of 19 placebo-controlled clinical trials for antidepressant drugs, from which they concluded that as much as 75 percent—perhaps even more—of the reported effects of antidepressants are actually the placebo effect. In their ironic formulation, they had “listened to Prozac but heard placebo.”²

The passion, particularly in America, for alternative medical treatments, from herbal remedies to Chinese acupuncture, also has focused attention on the placebo effect. Critics warn that customer loyalty to specific treatments is no guarantee that these unorthodox treatments “really” work. Perhaps the customer’s satisfaction is being purchased with phony money: that is, the placebo effect.

The mainstream and alternative medical fraternities disagree on many things, but they both resent and resist any charge that one of their treatments may owe its apparent efficacy not to “real” factors, but “just” to the placebo effect. The open secret here is that the accusers in these rhetorical wars are always right, at least in part, even if we may quarrel with the implications they draw. Whether we like it or not, the placebo effect is tangled up in judging the efficacy of virtually all treatments, alternative and mainstream alike, because human subjectivity and human sociability are always part of the equation of healing.

As human beings who acquire a culture, and make meaningful symbols from our experiences, we do not passively receive experiences; we actively create them. We judge an experience before we have it, and our incorrigible sociability sways our judgment. Our need for belonging is so powerful that when we are sick we may let pills and crucifixes become proxies for other human connections perhaps not available. We too narrowly frame the problem of placebos when we see it as a question of monitoring the effects of “dummy” pills.

What placebo effects challenge us to ask is: How does the force of human cultural imagination, on the one side, and the energy of human sociability, on the other, interact to affect our biology and literally make us sick or well? How does the brain function as a translator between events we experience supposedly outside ourselves, in our society and culture, and events we experience inside our bodies, in our physiology and biochemistry?

Asking these questions pushes us up against the walls of today’s house of science. The observable realities of placebo effects demand that brain science and medicine finally start to take the measure of human beings with a generosity and nuance that have been more typically the province of the humanities and social sciences. Thus, a future neuroscience of placebo effects tantalizes as a potential bridge across C. P. Snow’s “two cultures” of the academy. Not a bad destiny for a phenomenon that has historically irritated both alternative and mainstream medicine as noise in the system—a lie, a sham, something fake posing as something real.

The Dishonest Body

Because the stakes in this project are so profound, the resistance we are likely to face in taking it on are likely to be no less so. A brief historical vignette may illustrate how this can happen and why. It is a story well-known to historians, unfolding in Paris in the late nineteenth century, and featuring the leading neurologist of Europe at that time, Jean Martin Charcot.

Charcot was a neurologist famous for his visual identification and systematic analysis of neurological disorders like multiple sclerosis. In the 1870s, he turned his attention to a particularly elusive nervous disorder then common in the great hospitals of Paris. This disorder was called hysteria. For generations, hysteria’s changeable symptoms—convulsions, paralyses, tunnel vision, color blindness, patches of anaesthesia, incessant coughing, tics, and feelings of choking—had frustrated clinicians with their infuriating tendency to shift around the body and replace one another. In addition, the disorder was often marked by periodic crises, involving epileptic-like convulsions and states of delirium marked by violent emotions and hallucinations, that made the patients (predominantly female) difficult to handle.

Charcot’s plan was to penetrate the baroque, bewildering world of hysteria’s symptoms. He would make hysteria lie down and reveal its secrets. And, to everyone’s amazement, in his hands, the plan worked. Underlying patterns of permanent symptoms emerged. The crises associated with the disorder turned out to be far less chaotic than people had previously thought, unfolding with complete predictability in four stages that—as Charcot put it—followed one another with “the regularity of a mechanism.” In fact, the fits were so reliably linked to supposed mechanisms that the doctors said they could bring one on by squeezing particularly sensitive spots near the ovaries of these patients—the so-called “hysterogenic” zones.

This little detail of the research approach provides an interesting window onto relations between the researchers and their patients, but for our present purposes, we should pay more attention to something else: that there was very little interest in these women’s subjective experiences, in their stories and feelings—all the things about which we might immediately wonder today. For Charcot, the truth of hysteria was to be found in the physician’s study of outward signs and symptoms. This was where objectivity lay, this is where you knew you would not be led astray. In the famous Salpêtrière Hospital, Charcot’s assistants were set to work sketching pictures of the hysterics in action. Soon after, a photograph studio was installed in the hospital (a major innovation) and elaborate photographic records of the patients were made. People said that the camera was as crucial to the study of hysteria as the microscope had been to the study of cells. The camera, it was said, did not lie; it provided the evidence Charcot wanted and needed to prove—as he put it—that the laws of hysteria that he had discovered were “valid for all countries, all times, all races,” and “consequently universal.”

Except that they were not. In fact, this was a physiology of hysteria whose laws, far from being universal, were in the end so local that they unfolded only inside the walls of Charcot’s Salpêtrière. It turned out that the subjective experiences of Charcot’s patients had had the upper hand after all, and had fooled the great master by masquerading themselves as objective displays of universal significance. Using hypnosis (which Charcot had also helped to rehabilitate), rivals of Charcot showed that one could reproduce all the symptoms of hysteria, or change them, or make them disappear.

Charcot became a target of ridicule, and his disciples, feeling tricked and humiliated, scattered. Charcot’s entire neurological edifice of hysteria, rooted in the visible, the objective, the universal, crumbled. All of its contours were now chalked up to some invisible and obscure psychological process that people were beginning to call “suggestion.” The final blow came in 1901, when one of Charcot’s former students, Joseph Babinski, read a terse, uncompromising paper on hysteria before the French Society of Neurology. The paper was a mea culpa on behalf both of himself and of his discipline. They had been wrong, he said; hysteria had nothing to do with the brain and biology; it was a “made-up” entity, nothing more than the sum total of symptoms that could be created by suggestion and removed by suggestion. Neurology need pay no more serious attention to it.

What happened next? Sigmund Freud and others reinterpreted hysteria not as a disease of the brain and body, but as a disease of biography and forbidden fantasies. Thus redefined, hysteria passed out of the high-status, impersonal world of visible and universally predictable biology, and into the slightly disreputable human domain of culture, interpersonal relationships, and personal biography.

This is often where historians telling this story stop; and yet I am not satisfied. Let us grant that Charcot’s particular approach to making hysteria “real” within the framework of the neuroscience of his time did not work. But how do we make sense of those very solid-looking and material displays of local physiology that did— “really” —unfold inside the Salpêtrière? There was Charcot, misled about the truth of the situation, using nothing but props, performance cues, and the force of his own charismatic personality, yet nevertheless managing to create a theater of illusions so solid, so attuned to collective cultural understandings of the time, so agreeably visible, that for a while, all of Europe would be captivated by its power and authority. There is something so telling about this story that, for just a moment, threatens to burst open our available categories of understanding, before it is swiftly contained. If we cannot make something act like sensible biology, we will turn it into an affair of subjectivity and biography—because we do not know how to think about phenomena that may be both these things, or neither.

Moments like these help us to appreciate this: the profound issues that still demand clarification about the relationship between how we actually function in the world and how neuroscience and the other life sciences have taught us to see ourselves.

On the Road to Placebo

One enduring legacy of Charcot’s fall and Babinski’s shame is the conviction within modern medicine that the body lies, that patients trick. In the decades after Charcot’s fall, one sees a sharp increase in doctors’ hand-waving use of terms like “hysteria” (or “hypochondria”or “neurosis”) to identify an increasingly wide circle of patients whose problems they had decided were “all in their heads.” Mostly these were patients who showed up repeatedly in their offices, complaining about cramps, vague pains, or shortness of breath and demanding satisfaction— patients they did not know how to cure.

But doctors often did find ways to offer satisfaction to these assumed make-believe patients. They treated their make-believe ailments with make-believe medicine, giving them pills made of sugar or milk lactose, without any “real” pharmacological action, pills whose efficacy was presumed (drawing on the lessons learned from the 19th century) to lie in the power of suggestion. By the 1940s, the term for these little packages of unreality was “placebo.” One could even buy them from mail catalogues: blue, yellow, or red tablets inside bottles with the word “placebo” written on them, although perhaps backwards or with some letters displaced, to disguise the true nature of the contents.

But here was the hitch. If some patients responded to the suggestive powers of a fake pill, they also might respond to the suggestive powers of a treatment that the pharmaceutical industry hoped might have some real pharmacological action. Thus any time researchers were trying to test the effectiveness of a new drug for the consumer market, they risked being misled by those gullible or suggestible patients.

What to do? People said you needed to protect yourself against what was beginning to be called the “placebo effect” by the use of placebo-controlled drug trials. Today, we assume all drugs have to be tested against a placebo version of themselves, a dummy drug. Why? Not to learn about the placebo response, but to subtract that response from our equation, skim its traces off the surface of our data like an impurity.

Placebo-controlled trials are the ritual of modern medicine that, perhaps more clearly than anything else, expresses our inability to deal with something in the equation of human healing that, at the same time, we know is omnipresent and powerful. We believe in the ubiquity and power of the placebo effect, even as we are committed to keeping that effect as “unreal.” It is a fascinating form of mirror-logic.

Making Room for Placebos

The social psychologist W. J. McGuire (1969) refers to “three stages in the life of an artifact”: First it is ignored, then it is controlled for its presumed contaminating effects, and finally it is studied as an important phenomenon in its own right. Our emerging knowledge of the placebo effect may now enable us to move them from stage two (mainstream status as a contaminant in the equation of human healing) to stage three (acknowledgment as an important phenomenon in our understanding of human healing). Several decades of trying to weed out the contaminating impact of the placebo effect in drug trials have taught us—inadvertently—about the scope and strength of that effect. We have learned, for example, that we probably are looking not at a single effect but at a whole family of effects that are unlikely to be elicited in identical ways, mediated by identical internal mechanisms, or follow identical courses. Much more research is needed, but we already know, for example, that there is a very high placebo response associated with depression, but a much lower one associated with obsessive-compulsive disorder. Pain is notoriously responsive to placebo effects; broken bones may be less so.

The table above illustrates my next point: mapping the topography of conditions and syndromes where there is a demonstrated placebo effect may lead us toward real pay dirt on the placebo effect, insight into its underlying mechanisms in varied conditions. Our advantage over Charcot and Freud is our knowledge, albeit still emerging, that the mind and brain exist in, affect, and are affected by the body. Few today imagine that mind is separate from brain, but too many still assume that the brain is its own universe, apart from the body. As this assumption is challenged, placebo effects begin to seem less spooky or unreal; they begin to look like something we might explain.

In explaining placebo effects, however, we may eventually have to retreat from an exclusive preoccupation with the brain basis of mind. We may have to conclude that the body is in truth a “minded” entity. For now, though, we can observe that all known clinical effects of placebos appear to be compatible with, and therefore potentially explicable within, an emerging view of our physiology: the view that the central nervous system exists in a subtle dance of communication with other systems in our body—autonomic, endocrinological, immunological—that are essential to health and healing.

We are far from a comprehensive neuroscience of placebo effect. We have not even fully mapped a research program on the neuroscience of placebo effects. Yet we do know something about how placebos work. We just do not see, as yet, how this knowledge fits into our other knowledge, and therefore cannot yet clearly see what still must be asked. The proverbial tale of the blind men feeling different parts of the elephant is useful here: We must first inventory the different “views” of the elephant in the published literature; then ask what kind of whole animal the placebo effect may be. Here, as listed in the table above, are those views.

1. The “Placebo Personality”

Since the 1940s, we have assumed that about one-third of all patients are placebo responders. Early on, doctors asked: Who are these people? What kind of person gets better from imaginary medicines? Tellingly, their explanations incorporated the ambivalence clinicians have historically felt about placebos. Much of this research has a distinctly condescending tone. Some researchers thought that so-called “placebo reactors” might be unusually suggestible (or, alternatively, unusually hypnotizable). Others looked for hysterical or other neurotic symptoms among placebo responders, or evidence that they were unusually submissive to authority. Still others suggested that these patients might be less intelligent than the rest of us.

The hunt for the “placebo personality,” at least in the largely psychoanalytic or psychodynamic formulations people favored during the heyday of this approach, turned up empty-handed. That said, we should not ignore the possibility that stable individual differences in responsiveness may be important. Perhaps we just need better ways to think about them.

2. Props, Symbols, and Ritual Interactions

By the late 1960s, behavioral science had moved from an era of Freud to an era of B.F. Skinner. Attention turned to the role that environmental cues played in helping placebos to work. Doctors observed, for example, that placebos were moderately effective when given as little white tablets, but much more effective when given as big red capsules—and still more effective when the patient had to roll up his sleeve and get an injection. In other words, the visible signs or symbols of medicine seemed to have different powers. A white pill did not “mean” the same thing as an injection, and it therefore embodied a different kind of potency for the patient, triggering a different kind of placebo experience. The research here was never very systematic, but it has implications worth reexamining. They suggest that context, the whole “surround” of human therapeutic interventions, may be as significant in the chain of causes leading to a positive outcome as the specific intervention. The distinction between the picture and its frame begins to blur.

3. The Doctor as Placebo

The 1960s also saw a growing recognition that the doctor himself could also be a potent placebo. If a doctor seemed authentic, confident, enthusiastic about the treatment, if he listened with a caring attitude, or showed concern for the suffering of the patient, all this predicted a good placebo response. Were some people better at eliciting placebo responses than others? Were there reliable interpersonal techniques to maximize benefits? No one knew for sure, but understanding of the placebo effect was getting stretched to uncomfortable dimensions. We began by thinking only of the palliative effects of sugar pills; now, we had to make room for ideas of relationship, trust, and human connection.

4. Just Imagining You Feel Better

By the early 1970s, a more skeptical note was sounded. It arose within social psychology, particularly a branch of research known as “attribution theory.” The basic argument here was that placebo effects do not involve “real” (structurally discernible, clinically significant) changes in a patient’s physical state. Instead, patients just believe that they are improving; they just think the pain has gone away. Patients are eager to please their doctors. They want to get better, so they pay attention to the various naturally occurring and ambiguous changes in their clinical condition and interpret them in biased ways. They ignore negative changes, fixate on positive changes, and (falsely) attribute all the latter to the placebo effect—much as many people think they discover their life problems in daily newspaper horoscopes.

Social scientists challenged us to relate placebo effects to a larger insight from social psychology: We have such a poor tolerance for uncertainty that we will actively, almost compulsively, take ambiguous data and construct reassuring or self-affirming stories out of it. In this view, placebo effects were less lies than a creative triumph of the mind and brain over reality. We still need to pay attention to this insight.

5. Hope Heals, Hopelessness Kills?

Another school of thinking in the 1970s agreed with attribution theorists that placebos do not affect the course of a disease—at least not in any specific sense. They argued, however, that your belief that you are receiving effective therapy makes you feel more hopeful and less anxious, which has a positive effect on your physical state. We know, for example, that lowered anxiety seems to decrease subjective experience of physical pain. The converse of this claim about the effects of hope, supported by animal research, is that emotional states of helplessness and hopelessness may lead to sharp declines in health, maybe even a hastening of death. In humans, extreme cases came to be designated “voodoo death,” or death brought on by hexing.

Voodoo death existed for years on the margins of the already marginal debate on placebo efficacy. It seemed an exotic, probably anecdotal phenomenon associated with “primitives,” with dubious relevance to modern medicine. In the late 1980s, however, an article by Sanford Cohen on “Voodoo Death, the Stress Response and AIDS”³ took up the theme and dropped it squarely into our own world.

Cohen was interested in death that appeared to be hastened by extreme hopelessness and isolation—for example, displaced refugees in a new culture. Cohen also explored clinical reports on the apparent physiological effects of familial rejection and social stigma associated with AIDS. Cohen offers the example of a mother who:

learned on the same day that her son was gay and had AIDS. She reacted to this with hostility and openly maintained a prayer vigil outside the intensive care unit, praying that her son would die because of the shame he had caused her. The patient could hear his mother praying. One hour later the patient died, much to the surprise of his physician, since he did not appear to be terminal.

Anthropologist Robert Hahn proposes that voodoo death is, in fact, the dark counterpart of the placebo effect. He calls this cocktail of emotional and cognitive forces that undermine the body in the face of illness, or even under normal circumstances, the “nocebo” effect.

6. Enter the Brain: the Endorphins Story

For the first three decades of exploration, explanations of the placebo effect had had little to say about the brain. In the mid1970s, however, two new theories of placebo action brought the brain to center stage. In doing so, they promised to alter the stakes and status of the entire placebo research effort.

The first of these new theories arose from the discovery that the brain naturally secretes certain substances called endorphins that function in ways chemically similar to opioids. In the brain, endorphins attach themselves to the same receptor sites as morphine. They therefore appear to be the brain’s own natural painkillers.

In 1978, Jon D. Levine, N.C. Gordon, and Howard L. Fields reported that some pain relief from placebos was mediated by these same endogenous opioids.⁴ They gave patients placebo medication for postoperative pain, followed by a hidden infusion of naloxone, a substance that blocks the brain’s opiate receptors. The effects of placebo were inhibited; patients who had felt better under placebo now experienced a resurgence of pain.

Lively controversy arose over the details of the endorphin-placebo link, though the basic finding has held up. Equally important, you sense from the literature that this was a galvanizing moment in the history of investigating the placebo effect. For the first time, there was a mechanism that suggested how pain could “really” diminish following placebo. The “subjective” elephant of placebo effects started to feel like it had biological flesh on its bones after all. True, we did not know any mechanism that could account for placebos’ effects on conditions other than pain, but a door had opened. Others, people hoped, would follow.

7. Classic Conditioning: Placebos and Rats

The endorphin story took us a step closer to understanding how placebo outcomes could be mediated by “real” biochemistry, but how might the process of placebo response be initiated? Enter the second influential theory about placebos from the late 1970s: the conditioning model. We all know about conditioning: Pavlov’s dogs salivate when fed meat powder; a bell is rung at the same time; and pretty soon the dogs start to salivate when just the bell is rung. Since the 1960s, a few behavioral scientists had hypothesized that patients responding to placebos might be conditioned. They argued that placebos work because of people’s experience getting better from taking medications. From these experiences, the patients become conditioned to connect various cues associated with treatment—such as the shape of a pill or a doctor’s white coat—with their experience of feeling better. Soon, with appropriate repetitions, they may respond just to the props or incidental features of a medical intervention.

Although this idea had been around for some time, it barely registered on the broader radar screen of academic debate until the 1970s, when something happened that fundamentally changed the stakes. Psychologist Robert Ader and his colleagues at the University of Rochester reported a study with rats that were given saccharin-flavored drinking water accompanied by injections of cyclophosphamide, an immunosuppressive and nausea-inducing drug, often resulting in the death of the rats.⁵ At this time, the immune system was believed to be self-contained, not susceptible to influences from the central nervous system. Yet when a subgroup of rats who were not given further injections of cyclophosphamide continued to be fed drinks of saccharin water, they kept dying at a high rate. The otherwise harmless saccharin drink, because originally paired with the harmful injections, now seemed to depress the immune system of these rats in the same way as had the original active drug.

What was this, Ader reasoned, but an experimental placebo, one that got down into the immune system, and perhaps explained how placebos could start the healing process? That this process was observed in rats, rather than in humans, was a further block-buster. Perhaps you did not need to drag in all those funny, hard-to-quantify stories about personality and the doctor-patient relationship to explain placebo after all. These were just rats.

8. Swallowing Culture Along with Capsules

Yet maybe the last word on this score still had not been said. In 1926, the physiologist Charles Judson Herrick had closed his book, Brains of Rats and Men, with the firm assertion: “Men are not rats. Men are bigger and better than rats.”⁶ Important features of the human mind and brain make the processes of falling ill and recovering from illness more complex for us than for other animals.

If we really did respond to placebos in fundamentally the same way as rats— through a conditioned response—then our responses should mimic the effects of the “real” active substance that we think we have ingested. In fact, though, our responses often do not. In the 1980s, research suggested that the effects of at least some placebos are tied more closely to people’s expectation of a substance’s effect than to its “real” action. For example, people get drunk on placebo alcohol (an alcohol-free drink that looks and tastes like alcohol) in the way that their culture has “scripted” the drama of getting drunk, which may not have much to do with the actual sedative effects of alcohol.

Moreover, there is evidence that, in at least some people, an act of persuasion by another person can override “real” pharmacology. Jerome Frank, in his classic Persuasion and Healing, told of a severely nauseated pregnant woman given an emetic, ipecac. Although ipecac is given to induce vomiting, the patient was assured that she would soon feel better after taking this medication. What happened? She trustingly took the medicine and, within minutes, her nausea went away and a balloon in her stomach showed that her stomach began contracting normally.⁷

Reports liked these serve as a backdrop for another view of the placebo effect, championed in the 1980s by medical anthropologists. They suggested that the rituals and stories of our society (teaching us what it means to be sick, how we might expect to feel better, who are the people endowed with the power to help) are not neutral; they are important parts of a process with real physiological action that gets right “under our skin.” We swallow culture along with our capsules. In appreciating this, we may begin also to appreciate—to our possible intense discomfort— how much Western high-tech medicine might still share with the “superstitious” bone-pointing, faith healing, or shamanistic healing practices of various non-Western healing traditions—traditions studied by anthropologists but largely ignored, or dismissed, within modern scientific medicine.

The Next Questions

So we come full circle, back to features of the placebo effect that may be intellectually intriguing but, at the same time, professionally threatening. If placebo effects challenge us with the possibility that culture and context are within the body as well as without, then many assumptions guiding biomedical and neuroscience research will need rethinking. We may find ourselves beginning to think differently about the historical rationale for banishing meaning from biology. We may question some conventional distinctions that see natural facts as one thing (necessary, universal, and objective), and sociocultural facts as another (contingent, value-oriented, and subjective). These are exciting, radical possibilities.

Where do we go from here? We might start with the idea that, on the physiological level, an array of things go under the heading of “placebo effect.” Some biological processes, such as pain modulation, are notoriously susceptible to placebo effects, while others are not. These differences may not occur simply because different physiological systems in our bodies are differently porous or resistant to placebo action. It is more likely that different physiological systems in our bodies interact with the higher brain-centered directives of placebos in different ways. In other words, the placebo effect that produces pain relief arises when our brain communicates with specific systems in our body that regulate pain; the different placebo effect that may alter cancer cell growth arises when our brain communicates with specific systems in our body that regulate normal and abnormal cell division. Thus research on placebos should draw in experts on all body systems involved in healing. Then we can begin to understand how placebos interface differently with each of those systems.

Mapping differences in placebo action “downstream” from the brain, though, is just the first step. It sets up the big question: What mechanisms enable an idea that enters our minds as a culturally specific story, or image, or ritual drama to become a set of physiologically specific instructions to the body? Placebo effects are often defined as the “non-specific” effects of a particular treatment, but what seems most striking about them is the typical specificity of their actions. Can features of the placebo “drama” be manipulated to give us a better handle on the environmental, interpersonal, or contextual determinants of different specific placebo outcomes? Can we then imagine asking questions about the different kinds of contributions that perceptual, emotional, and interpretive brain mechanisms play in these different kinds of “placebogenic” events?

Candies or Sponges?

Consider two images, two metaphors. The first views people and their bodies as M&Ms that melt in your mouth, not in your hands. This metaphor suggests that each of us bounces around inside our meaningful life stories, our collective histories, and our cultures, but that our biology has little to do with these—because it is tucked safely inside its protective candy coating. This first metaphor, in other words, tells us that some things are “inside” us and others “outside.”

But what if we imagine people as sponges? It makes no sense to talk about the physiology of sponges, on the one hand, and the watery world in which they live, on the other. The water is integral to the way these animals function. As sponges, we are immersed from birth in an ambiance rich not only in oxygen, nitrogen, and other nutrients, but in symbols, stories, and human interactions. This metaphor asks if the final lesson of placebo research may be that our physiology is as porous to our cultural ambiance as it is to the material one that sustains us.