Who's an Authority on Nutrition Science?

Harvard science historian Steven Shapin says modern technical expertise about what we ought to eat is seriously divided. He considers who is an expert and who captures the public's attention when it comes to dietetic advice.

In the 1890s, Leo Tolstoy criticized modern science for not having an answer to the question "What should we do?" and in 1918 Max Weber celebrated modern science for just the same reason: "Science is meaningless because it gives no answer to our question, the only question important for us: 'What shall we do and how shall we live?'" But, in fact, quite a lot of current scientific practices are in the business of supplying answers to that kind of question.

I want to talk about one version of this sort of modern expertise, one that illustrates different strands of technical expertise, enormous lay interest in the general field, enthusiastic lay take-up of some kinds of expertise, and vigorous lay rejection of other sorts. The case is dietetics and such cognate practices as nutrition science—anyway, expertise about what you ought to eat, and not eat, to avoid a range of diseases and to live a long and healthy life. Dietetics is now, and always has been, an expert practice that has the possibility of uniquely tight engagement with quotidian life, lay knowledge, and the texture of morality. If you are lucky, you get to eat three times a day, seven days a week, so there are at least twenty-one occasions per week when the fork approaches the mouth. Each one of these is a possible occasion for the intercalation of expert knowledge between fork and mouth. In our sort of society, it is just not possible for food to be bought, prepared, and ingested unaccompanied by thoughts flowing from bodies of technical expertise: the relationship between dietary cholesterol and heart disease, fiber and colon cancer; the boundary between the amount of wine that is good for you and the amount that is not; and, of course, calorie content in relation to weight gain and loss. Every meal now comes polysaturated with technical expertise, and, while you can, at cost, decide to ignore it on this occasion, even ignoring expertise is a decision you probably have to take. In the main, however, the decision seems to be which type of expert commentary gets pride of place at this meal.

Dietetics is about exceedingly complex phenomena, and attempts to understand complex phenomena typically generate endemic expert disagreement. The British Astronomer-Royal Sir Martin Rees five years ago drew an instructive contrast between the security of scientific knowledge of simple things—the properties of celestial objects, for example—and the fragility of our knowledge of complicated things like human nutrition: "Scientists still can't agree on what food is good for us. There is a real sense in which dietetics is harder than cosmology." Questions about which expertise to invite to the table can be, and now often are, vexed decisions for the average eater. Almost needless to say, modern technical expertise about what you ought to eat is seriously divided. The heterogeneity of expertise on a very large number of occasions of public concern is a notable feature of our culture.

Most laypeople probably know that expert dietetic claims are various and change rapidly, though it would be good to have more securely grounded understandings of what people do and do not know. Sometimes the diversity of conflicting expert voices encourages the laity to suspect that genuine expertise is not available in the area. Dietetics seems to be a domain characterized by both very great expert heterogeneity and marked lay faith that authentic expertise must be available somewhere in the welter of conflicting claims. The laity assert their freedom to pick and choose which expertise is credible, while giving few signs that they find the whole domain of dietary expertise wanting.

This is one further indicator of the way in which the body of expertise has become entrenched in our culture and what a long way we have come from the late 16th century, when Montaigne thumbed his nose at the very idea of such expertise and its authority over his daily patterns of eating and drinking: "The arts that promise to keep our body and our soul in health promise us much; but at the same time there are none that keep their promise less." Expert physicians disagreed among themselves, and, so, if you did not like the dietary advice that one offered, you could always pitch one doctor's favored rules against another's: "If your doctor does not think it good for you to sleep, to drink wine, or to eat such-and-such a food, don't worry: I'll find you another who will not agree with him." For that reason alone, you might as well do what you thought best: you were, and ought to be, your own dietary physician.

The USDA Food Pyramid, initially published in 1992 with wide academic consultation, and much industry feedback, still seems to represent the condensation of academic expert knowledge among the nutrition scientists. Marion Nestle, one of the leading academic nutrition scientists, assures us that two-thirds of the American public know the original USDA Pyramid; that there is "almost universal" expert consensus about its recommendations; and that genuine expert counsel has long been and will remain stable: it is a high-carbohydrate, high-fiber, low-protein, low-fat diet—pretty close to "the Mediterranean diet" associated with the 1950s arid 1960s work of the pioneering nutrition scientist Ancel Keys.

However, when Nestle published her book Food Politics in 2002, the Pyramid was already considered ripe for revision. In February 2000, the USDA staged a public "Great Nutrition Debate" among nutrition experts as a preliminary to revising the Pyramid and in September 2003 announced a program for that revision, partly as a response to anxieties that two-thirds of Americans were now classified as overweight or obese. During the writing of the original draft of this paper, it was understood that the revised Pyramid was to be announced early in 2005, and the odds were that, while it would still be a relatively low-fat, low-protein diet, it would be substantially different from the existing form. Important distinctions might be made between "good" and "bad" carbohydrates, "good" and "bad" fats—for example, whole wheat bread is pretty good, white bread very bad; olive oil is good, Crisco very bad. As it turned out, the revision released in April 2005—and labeled as "MyPyramid"—was widely viewed as an incoherent compromise, the USDA avoiding the responsibility of giving blanket dietary advice to the population as a whole and certainly not counseling eating "less" of anything. "One pyramid does not fit all of us," said the Agriculture Secretary Mike Johanns, "so we created 12 different ones."

Among academic experts, Walter Willett of the Department of Nutrition at the Harvard School of Public Health was leading the charge: "The [old] Food Pyramid is tremendously flawed. It says all fats are bad; all complex carbohydrates are good; all protein sources offer the same nutrition; and dairy should be eaten in high amounts. None of this is accurate." The probable causes of this knowing dissemination of bad advice were, first, that "many nutritionists decided it would be too difficult to educate the public about [the] subtleties" involved in discriminating among carbohydrates and fats and, second, that the USDA Pyramid was a nutritional icon substantially shaped by the political and economic interests of the food industry. Nor was there any convincing evidence that individuals adhering most rigorously to the counsel of the USDA Pyramid had improved health outcomes attributable to their diet: true, they did have "a lower risk of major chronic disease," but, Willett found, these same individuals also smoked less and, in general, led "healthier lifestyles." Instead, Willett proposed a conceptually jumbled "Healthy Eating Pyramid" with "exercise and weight control" at the base, complex carbohydrates and mono- or polyunsaturated vegetable oils just above, and moderate alcohol and multi-vitamins hanging in space off to the side.

And then there are the experts who actually sell books—lots of books—the ones who are clearly winning in the marketplace of dietary credibility. And here a crucial name is Atkins, who turned the USDA Pyramid upside down and commended lots of protein and fat and tight control of the carbs. Since Robert Atkins's Diet Revolution first appeared in 1972, over 16 million copies of its various versions have been sold worldwide; 30 million Americans have said they have given it a go. Atkins was very big in Britain, in Scandinavia, and, amazingly, in Italy, where there was a certain amount of panic among the pasta makers. In its U.S. heartland, few academic experts were won over, but their lives were being made more and more uncomfortable as their constituencies increasingly voted with their gut instincts. The New York Times Magazine pronounced Atkins vindicated, declaring that USDA Pyramid-type academic expertise may all along have been "a Big Fat Lie."

And so the question is put: what are the grounds of dietetic credibility? Why did Atkins, and many others of the low-carb popular writers, do so well in the public credibility market, while academic expertise is struggling to make its voices heard? That is a complex question, and all I want to do is to suggest some features of the credibility economy that figure here and that have got some interesting connections with how we tend to think about science and common sense (or lay knowledge) more generally. There are all sorts of considerations I am going to set aside for these purposes, and which are doubtless very important: for example, the fact that Americans like to eat meat and that it is relatively cheap; the possible route that low-carb regimes offer for the masculinization of dieting; a dialectical oscillation in diet fads—low-carb, high-protein fashions alternating with the high-cab, low-protein orthodoxies of the recent past; a rhetoric of luxury, metabolic causation of obesity, and the disavowal of will power that resonates beautifully with current American fashion and academic theory?

First, there is the matter of who is an expert and, therefore, entitled to public authority. That is not a simple question either. Presumably, attending to any given writer implies the recognition of expertise: who would put their health in the hands of someone who they reckoned did not know what they were talking about or who did not possess special and superior knowledge? What is, however, crystal clear is that membership in an academic institution, or in a credentialed expert community, is not, in this specific case, much of a public warrant of expertise and may even count as a liability. One of the most pervasive tropes one finds in the popular dietetic literature is a studied disavowal of credentialed expertise. So while Atkins was a physician, he quickly assured his vast readership that he was a cardiologist and not a diet doctor, drawn ineluctably to his calling by simple, but vast, experience of what actually worked and did not work: "Looking back at this period in my life I realize what a fortunate thing it was that my training has been in cardiology and not in metabolism. If I had been trained in nutrition and metabolism, I'd be parroting the same classical misconceptions that so many of my colleagues still hang on to. Being free of these misconceptions, allowed my mind then—and since—to approach the observable facts without prejudice."

One telling contrast between the writings of not just Atkins and the academic nutritional establishment, but between virtually all the bestselling writers and the establishment, concerns the question of a science of the particular versus a science of the universal. Look up dictionary definitions of "science," and you will see an overwhelming disposition to define science as the search for, and accumulation of, knowledge of universal scope." That is why Simplicio lost the argument with Salviati in Galileo's Dialogue: when Simplicio noted that Galilean idealizations worked well in the abstract but rarely if ever fit concrete instances, Salviati recognized a peasant when he heard one. Salviati/Galileo wanted to know how bodies moved through ideal, frictionless media; Simplicio wanted to know how a particular body moved: "These mathematical subtleties do very well in the abstract, but they do not work out when applied to sensible and physical matters." In fact, there are many present-day sciences of the particular, though you would not know it from the definitions: all those people trying to work out when the next Cascade volcano will go off, or when the next earthquake in San Francisco will occur, are practicing sciences of the particular. I would guess—though I can provide no firm statistics to support the guess—that the majority of people categorized in census data as "scientists" work within sciences of the particular. Many expert practices can be viewed as hybrid: civil engineering, it is said, "depends upon" physics but every bridge is different; medicine "depends upon" physiology in the same way, but every sick body is different. And you cannot "idealize" away either the unique geology of San Francisco or the unique physiological state of Professor Shapin's liver.

You can eliminate Simplicia's voice from the body of expertise, but there is a price to be paid. That price may not be consequential public disapproval—the American public have almost always been willing to pay the substantial bill for theoretical physics, and, for all the talk of "public hostility to science," scientists continue to be held in exceptionally high esteem and science research budgets continue to go up—but it is likely to be public "unconcern." This is not a big problem for physics, but it is a very big problem for those expert practices that purport to engage with the question "What should be done?" Dietetics is certainly one of these practices. It would be quite wrong to say that academic nutritional experts are trying to shrug off the responsibility of giving practical advice. But they go about it in a significantly different way than Dr. Atkins.

First, consider the philosophically crucial divide between the objective and the subjective. As a global generalization, I do not think that more than a tiny fraction of academic scientific writing on practical dietetics ever addresses what food and eating mean. Eating is treated as an instrumental act: here is what you want to do to avoid diabetes, heart attacks, colon cancer, Alzheimer's, death. The assumption is that the recipients of advice are like the subjects in rational decision theory, maximizing their utilities, and expert advice just tells people—to the best of the expert's ability—how to achieve this maximization most effectively. This is how Harvard's Walter Willett starts his exercise in extending academic nutritional expertise to a wider audience: "We eat to live. It is a simple, obvious truth. We need food for the basics of everyday life—to pump blood, move muscles, think thoughts. But we can also eat to live well and live longer. By making the right choices, you will help yourself avoid some of the things we think of as the inevitable penalties of getting older."

I do not doubt that people want to live as long and as healthily as possible, but this is not the only thing that people do when they eat. They eat because they feel bad, or because they feel good. They eat to celebrate or to mourn, to establish claims to status, to affirm or deny group identity. They eat to be sociable, and sometimes they eat like other people eat in order to be sociable. You would not know any of this if you read the works of almost all academic scientific experts on nutrition. In the best-selling literature, however, you cannot miss engagement with the emotional and social meanings of eating. It should come as no surprise, then, that the laity might recognize in this literature a sense of pertinent aboutness that they may not see in academic writing.

Second, academic dietetic expertise treats human bodies in general and as populations. If you weigh this amount, or if you drink this amount, your chances of contracting or avoiding this or that dread disease is n per cent, or your risk of it is elevated or decreased n per cent. This can be useful information, most especially for policymakers, and let us grant for the moment that it is reliable, even though the domain of such academic statements remains notoriously conflicted. The problem, however, is that individual people tend to be interested in what is likely to happen to them. The answer for any specific individual will either be colon cancer or no colon cancer, die before age 75 or after age 75. What any individual wants to know is what it is going to be like for them, an answer to what one historian of medicine has called "the patient's fundamental question—why me?" Expertise about populations just cannot give that answer to any individual, but it does not follow that individuals may not want it anyway and look for expert writing that gives signs that it considers such a desire legitimate and intelligible. The difference should not be thought of as one between knowledge and ignorance, between cognitive virtue and cognitive error, but one between different kinds of concerns. You cannot use better logic or more evidence to refute a different kind of concern.

The popular literature acknowledges this circumstance. The style is overwhelmingly casuistical. That is, while there may be much metabolic science of general scope in any given text, the books are quite generally structured around particular cases. Indeed, they often are organized as a kind of counterpoint between the general and the populational, on the one hand, and stories about, say, Mrs. Finkelstein, a 49-year-old woman weighing 212 pounds, on the other. Case follows case, testimonial follows testimonial. Dr. Atkins barely gets his story underway before we are introduced to Stanley Moskowitz, "a vigorous 64-year-old sculptor who had survived three heart attacks," who was overweight, who was addicted to junk food, and who, under Atkins's tutelage, learned the delights of low-carb dieting. Then Mary Anne Evans gets to "tell her own story," a 42-year-old, 209-pound woman who had abandoned all hope before she achieved low-carb redemption.

You could describe this casuistical style as a deliberate attempt at credibility engineering, and that would almost certainly be correct, but it is also possible to interpret it as a response to the concerns which deliver the reader to the text. That is the sort of thing they expect to hear. True, Mrs. Finkelstein is not you, but she is, and is here presented as, a unique individual, not a population, and the metaphorical extension from aspects of her predicament to yours is available for you to make if you happen to feel her case to be sufficiently like your own. That is one way in which the popular writings are in a direct lineage of early modern medical traditions, which were focused on the idiosyncrasies of individual sick bodies whose workings the physician attempted intimately to understand. Enjoying special authority, both in early modern medicine and in the contemporary best-selling dietetic literature is "the case of the author"—a confession of how the author stood before undertaking the proposed dietary discipline and how it turned out for him or her—invariably, of course, very well indeed. In the mid 18th century, the most celebrated diet doctor in England, the enormously fat George Cheyne, devoted 30 pages of his celebrated The English Malady to "The Case of the Author," and in 1972 the most famous diet doctor of his day, Robert Atkins, told his readers how, as a medical student, he "had the reputation of being the biggest chow-hound in the hospital." So deep was he in denial of his own condition that it was not unti1 he was in his thirties that he realized he had "three chins." If you find dieting hard, Dr. Atkins assures you that so too did he: he understands your difficulties. So far as I can now see—though future reading may prove me wrong—there are no cases in the academic literature.

An early modern maxim—used as much by analogy outside of medicine as within it—was "Physician, cure thyself." You might well want to take the expert advice of healthy and slim physicians, just because they gave embodied warrants of sincerity as well as efficacy, and one thing you might well want to know about those attempting to give you expert advice is whether they took it themselves, and to what effect. That is what an interviewer asked Walter Willett, and, having told readers next to nothing about his own regimen in his expert writings, Willett responded to his interviewer directly. Interviewer: "What do you typically eat in any given day?" Willett: "For breakfast, I'll usually have something like cooked whole grains with fruit. Lunch may be leftover grains from breakfast as part of a salad, with vegetables, and nuts. Dinner is usually salad or vegetables with tofu, chicken, or fish." Interviewer: "What about secret indulgences?" Willett: "Sometimes I'll have some flavorful cheese or a bit of chocolate."

Having offered his own case to launch his 1972 text, when Atkins died in 2003, much attention focused on the circumstances of his death. From within the academic scientific point of view, it is hard to see why that should matter: probabilities being what they are, why shouldn't you "do as I say, not as I do," and why shouldn't someone following his own excellent advice nevertheless suffer a heart attack at age 71 and wind up dead on a Manhattan pavement at age 72? Assuming that Atkins followed his own diet, and assuming that he reduced his chance of a heart attack by, say, 50 per cent, the fact that he nevertheless suffered a heart attack should not affect the credibility of his diet: that's probabilities for you. But in this case it was rather harder to gloss away a New York death certificate which shows that at the time of his death Robert Atkins weighed 258 pounds and was, according to his own criteria, and those of the Centers for Disease Control, "clinically obese." So consider the contrast between regimes of credibility in which such a circumstance makes no difference and one in which it makes all the difference. Out of such a consideration, I suggest, might come an interesting new way of parsing the sciences and of describing the different conditions of their credibility. Less ambitiously, I have offered a sketch of some reasons—other than the simple distribution of unpalatable truth and appealing falsity—why the popular writers are popular and why the academic experts are not.

Steven Shapin is the Franklin L. Ford Professor of the History of Science at Harvard University. A version of this essay, excerpted with permission from Public Science in Liberal Democracy, where it originally appeared as "Expertise, Common Sense, and the Atkins Diet" (Toronto: University of Toronto Press, 2007), will be included in Dr. Shapin's forthcoming book about the history of ideas concerning eating and drinking. For the original essay including notes and references, see www.fas.harvard.edu/~hsdept/bios/docs/shapin-exp_com_sen.pdf.

Dr. Shapin will speak at the Academy in 2011 as part of the Science & the City Stone Age to Internet Age Series. He will present, "You Are What You Eat: The Long History of Knowing about Our Food, Our Bodies, and Ourselves," on January 26, 2011.