If I hadn't enjoyed Michael Pollan's The Omnivore's Dilemma so much, I probably would have put this one down after the first chapter. Pollan uses so many of the buzzwords I usually rely on to detect anti-scientific cranks that I honestly felt dirty pushing ahead and finishing the book. Over the course of just a few pages, he coins the term "nutritionism" to denote what he describes as a "reductionist" science that relies on a specially educated "priesthood" to tell the rest of us what to eat, in defiance of the cultural traditions we should instead rely on.

It does get better, and Pollan's guidelines for choosing meals (summed up in the book's first sentence, "Eat food. Not too much. Mostly plants.") are useful, and do, after all rely on recent scientific findings. His theme is that that nutrition science, by focusing on which particular nutrients are good and which are bad, has missed the possibility that there may be other factors in a healthy diet such as, to take just one example, the ratio of one nutrient to another, and that cultural culinary traditions, developed over long periods of time, may be a better guide to healthy eating. To his credit, he acknowledges the contradiction in opening with an assault on what he calls "nutritionism" and proceeding to use the findings of nutrition science to support his arguments, but he doesn't address the question of why he chose to open with his rhetorical assault on science in the first place.

There's some interesting stuff here, but it's overshadowed, in my opinion, by his gratuitous anti-scientific rhetoric.