In In Defense of Food, Michael Pollan proposes a radical new approach to diet: ignore everything you’ve ever heard about scientific nutrition, and just build a diet of only non-processed foods, with little or no meat. This advice is summed up in a short, pleasing slogan: “Eat Food. Not Too Much. Mostly Plants.” And you can kill yourself following it.

If you know a little bit about either Michael Pollan or In Defense of Food specifically, this summary is likely to come as a surprise to you. In Defense of Food looks like a respectable piece of trade non-fiction, as high-brow as you can get without going over to the academic genre. It’s the kind of thing you expect to find on the bookshelf of avid consumers of popular science. It even contains useful scientific information. But at the end of the day, Pollan’s basic attitude is one of deep and dangerous distrust of modern science.

The first of the book’s three sections is mostly a history of two things: the science of nutrition, and public confusion about nutrition, the latter in many cases being actively encouraged by food companies. But Pollan tries to explain this history in terms of an evil doctrine called “nutritionism,” initially defined as the belief that all you need to understand about food is its chemical constituents or nutrients.

In the second section, we get a discussion of the phenomenon of “diseases of civilization” or “Western diseases,” health problems like obesity and diabetes that consistently arise when people switch over from a traditional diet to the modern Western diet. Scientific explanations for this phenomenon are discussed: First, it seems that eating a lot of highly refined carbohydrates (not just sugar, but also white flour) messes with our metabolisms, encouraging fat storage and overeating. Second, processing foods removes micronutrients, and the variety of healthy things contained in whole foods is so great that it is difficult to add all of them back in. Third, the nutrients in processed foods may be more difficult for our bodies to absorb than nutrients as they occur in whole foods.

The second section contains lots of useful information, and seems largely on-target. It’s almost good enough to make me regret the tone of the rest of this review–almost. Unfortunately, Pollan seems incapable of invoking scientific findings without immediately apologizing for doing so. He likes the idea that carbs could be dangerous as a way to attack the idea that we should worry about fat in our diets. However, no sooner has Pollan launched that attack than he dismisses the idea of actively watching our carbs as another form of “nutrionism.” The third section begins with an apology for the amount of sciece in the second: “The undertow of nutritionism is powerful, and more than once over the past few pages I’ve felt myself being dragged back under.” Similarly, in the middle of part three Pollan recommends eating a variety of food to “cover all your nutritional bases”–but in the very next sentence, says, in effect, “oops, that’s nutritionism.”

The cleanest illustration of how awful Pollan’s recommended approach to eating is–and why you could kill yourself following it–comes from Pollan himself. At the end of the first part, in an attempt to show the what horrible effects nutritional science has had on the American approach to eating, he cites a study where people would ask what they would eat if they had to survive on a desert island with only water and one kind of food:

The choices were alfalfa sprouts, corn, hot dogs, spinach, peaches, bananas, and milk chocolate. The most popular choice was bananas (42 percent), followed by spinach (27 percent), corn (12 percent), alfalfa sprouts (7 percent), peaches (5 percent), hot dogs (4 percent), and milk chocolate (3 percent). Only 7 percent of the participants chose one of the two foods that would in fact best support survival.

So far so good, but Pollan neither explains why those two foods work, or thinks what his own advice would dictate for the dilemma. Milk chocolate, in fact, would break Pollan’s rule against processed foods, and hot dogs go against his “mostly plants” recommendation. The fact that only 7 percent of survey respondants picked either of these answers suggests people are already use a vague verson of Pollan’s principles–they have the idea that fruits and veggies are healthy, processed sweets and fast food meats are unhealthy. And that idea leads them astray here.

If you want to get the right answer to the desert island problem, you need the nutritional science Pollan tries to demonize. Quick lesson: the main way the human body does anything is with protein. Most people associate protein with muscle because that’s where most of the protein we eat ends up, but smaller amounts are used in literally every last thing the body does. Protein is made up of small molecules called amino acids. Most amino acids are things the body can synthesize, but there are eight so-called “essential” amino acids that can’t be synthesized. You need all of these in your diet to live, ideally in the ratio which your body uses them. Animal products are good sources of protein not just because they contain a lot of it, but because they have all the essential amino acids in the right ratios. Therefore, if you have to survive on a single food, pick an animal product.

The lesson of the desert island problem is not that Americans care too much about nutrition, but that they’ve been fed lots of nutrition buzz-words without any real understanding. Another case in point: I’ve begun to notice diet soda advertised as “zero calories! zero carbs! zero sugar!”–suggesting the soda is being marketed to people who know that carbs are bad, but not that sugar is a form of carbs and carbs are a form of calories. Both these cases are symptoms of a broader American scientific illiteracy, and Pollan’s anti-science attitude doesn’t help the problem.

The title of this post is a reference to Robin Hanson’s contrarian poem which begins “Food isn’t about nutrition.” The idea behind the poem is that much of human behavior isn’t driven by the things we want to think drive them, rather, they’re driven by our desire to signal status and group identities. And I think a review of In Defense of Food would be incomplete without mentioning this angle. The book’s claims would have fit neatly into a trashy supermarket-aisle diet book: The All-Natural Diet: The Revolutionary Cure for Today’s Worst Diseases that Scientists and Corporations Aren’t Telling You About. One reason for not taking this appraoch to marketing In Defense of Food is obvious: it would have set off the target audience’s bullshit detectors. The book’s actual marketing strategy makes it look enough like a pro-science book to keep those detectors from going off.

I suspect there’s more to it than skilled disguising of bullshit for sale to the relatively smart, though. A lot of the people who are apt to read In Defense of Food would be glad to have their bullshit detectors foiled this time around. If you live in a big, liberal city and a good chunk of your friends are fans of farmer’s markets or Whole Foods, taking up Pollan’s extremely strict dietary recommendations is a good way to feel good about yourself, as well as show what a health-conscious, pro-animal, environmentally friendly person you are. At a crasser level, if you can afford to make your life so very difficult in something so fundamental as your daily eating habits, it’s a good sign you are a relatively comfortable, upper-middle class white person, and not a poor person with better things to worry about. How sad that in our society, the desire to show off can beat out scientific thinking so easily.