"Unhappy Meals," says Michael Pollan. That's the title of his article published in The New York Times Magazine this past Sunday. (As Jonah has already pointed to.) After last year's The Omnivore's Dilemma, about what defines/describes different chains of food production, he is speaking still about food, but now more directly about nutrition. It's a push in the same direction to write about food and ecology, about how what we know of nutrition comes from a lot of scientific research that ignores the ecological relationships between and within food and human systems. Nutrition is a great example in this regard, and a quote Pollan uses from Marion Nestle (at New York University and author of Food Politics) sums it up pretty well: "The problem with nutrient-by-nutrient nutrition science is that it takes the nutrient out of the context of food, the food out of the context of diet and the diet out of the context of lifestyle."
The basic issue for Pollan is not a new one -- not new for him; not new in the history of social studies of science. He's worried that reductionist nutrition science takes individual entities out of their context and then exmaines them as if they ever exist as such, in isolation. The research then draws conclusions based on specific prodecures, the premises of which are antithetical to actual systemic situations. This strategy for research is part of the same philosophy of nature that treats the environment as a machine. It's the basis for the same industrial logic that Pollan discussed at length in The Omnivore's Dilemma. A quote from the article:
Scientific reductionism is an undeniably powerful tool, but it can mislead us too, especially when applied to something as complex as, on the one side, a food, and on the other, a human eater. It encourages us to take a mechanistic view of that transaction: put in this nutrient; get out that physiological result. Yet people differ in important ways. Some populations can metabolize sugars better than others; depending on your evolutionary heritage, you may or may not be able to digest the lactose in milk. The specific ecology of your intestines helps determine how efficiently you digest what you eat, so that the same input of 100 calories may yield more or less energy depending on the proportion of Firmicutes and Bacteroidetes living in your gut. There is nothing very machinelike about the human eater, and so to think of food as simply fuel is wrong.
For human sciences, of course, this is more tricky than for, say, studying astronomy or particle acceleration or whether a rock sample is a million years old or 100 million years old. That is to say, Pollan's concern for reductionist science is not a case against science. It's just a case against the use of a particular view of scientific inquiry that hasn't helped us understand food and nutrition.
People have a hard time dealing with complexity. If ever there was a cliche, right? He and I got into this subject very heavily in a (still-two-months-away-from-being-published) interview. The appeal for straightforward answers, for the one-to-one correspondences between single cause and single effect, is quite profound in journalist too, not science alone. I took it from our interview that this issue of strict causality is front-and-center with his currrent writings. I think this is why he threads throughout his article an appeal to journalists and scientists (and food eaters!) at the same time. (So, if you go and actually read the article, you'll find that he makes semi-coy overtures to the multiple audiences for his essay.) It's a common theme for Pollan the Journalist, and for ecologists, no doubt. And for Pollan, it's a theme that he's been working to make more clear in how we understand and then approach issues of food -- be it the production of it, consumption of it, or, as in his new writings, nutritional value of it.
So it goes. But now here's a post-script, since I started this post yesterday, but am only now geting back to actually posting it. In the intervening time, I saw and read Jonah's post.
I was intrigued by some of the replies over at Jonah's page. One reader blasted Pollan, suggesting with great confidence that "common statistical tests can tell you about the complex interactions between individual variables" (link). But that's just the purpose of the article. This is not entirely true. And, what is more, how woud you know?
Very defensive, these readers. Is all science good science then? If one critiques a scientific topic, is one then damning all science (as another replier said)? So strange to me. Why applaud reductionism if you don't have to?
- Log in to post comments
Awesome post. I sometimes worry about our polarized times. People feel the need to defend some "science" - whatever that may be - at all costs. Noting that reductionism has limits provokes the same sort of hostility as subscribing to intelligent design.
Good post. And here comes the 'but'... but, it seems to me that there is a bit of confusion about scientific reductionism. Reductionism does not demand that scientists study individual nutrients in isolation. It does, however, demand that something (the nature of which is not specified) be studied in isolation. That thing could be whole foods, drugs, homeopathic 'remedies', or a swift kick to the groin for example. I believe that scientists frequently study individual nutrients is because of methodological (read: measurement) issues in determining exposure to something like whole foods, not because they are beholden to reductivism per se.
Also, causation does not flow naturally from reductivism. Or, rather, determining causal associations from reductive models requires deep hand-waving (usually involving the estimation of fundamentally unobservable quantities). However, reductivism is not prevented from studying the 'causal' effect of one element on many outcomes. That is reductivism does not demand that we adopt a one-to-one causal scheme. Although, in my limited experience, it seems that people are more comfortable with diseases when they are perceived to have exclusive antecedents (low vit. C -> scurvy, smoking -> lung cancer, etc).
Having said that, I admit that I am a advocate for systems approaches to problems in my own field (infectious disease epidemiology) as an alternative to reductive approaches. However, if we accept that non-communicable disease are a function of the individual's biochemical processes over the course of the years and decades that it takes to develop many of the chronic conditions nutritionists are interested in, then systems approaches are simply too far off to be useful currently.
I would guess that, in the fullness of time, when we can accurately predict the effect of nutritional exposures using deductive, systems approaches, we will be left with powerful, rational justification for a mode of eating which our current intuition tells us is best: moderation, variety, whole foods.
Ethan, I appreciate your reply; it's a nice addition to the post above. You note that "reductivism does not demand that we adopt a one-to-one causal scheme," and this seems a well put point. I think that's right -- I wanted to suggest, coming at it from a slightly different angle, that the two grow from the same kind of philosophical backdrop, not necessarily that causation flows naturally from reductivism. Ben
When I said that common statistical tests can tell you about interactions between variables, I was referring to ANOVA tests that show whether the interaction terms are significant. They don't elucidate the mechanism that produces the interaction effect, but they tell you whether or not it's there (and more, which variables interact with how many other variables). If you followed Pollan's advice to "Let culture be your guide, not science," you'd never understand anything about interaction effects and main effects. His poo-poo-ing of expert advice is also off-base for the same reason: of course the average person doesn't know much about nutrition -- by Pollan's own admission, a substantial fraction of people slavishly follow whatever fad is cooked up by an industry's PR department and/or gorge themselves on processed foods.
The antidote to bad nutrition science isn't an overly skeptical and nihilistic view of that discipline but better science -- not relying on tradition. Another point he makes is that traditional food ways must be healthy since if they weren't, the people practicing them would've died off by now. Not so at all: the history of agriculture shows that agriculturalists suffered from the ravages of malnutrition and famine far more so than hunter-gatherers, yet the former have vastly outsurvived the latter. Yet all of the traditional diets Pollan mentions are from agriculturalists. I mean, the Irish were doing pretty well relying on the potato up until 1845. Almost all agriculturalist diets have persisted to the present, even though they leave their eaters less well nourished than H-Gs.
Also by Pollan's admission, "humans are omnivores, requiring somewhere between 50 and 100 different chemical compounds and elements to be healthy." Again, you would never know that, let alone which compounds are crucial, unless you investigated the matter scientifically, reductionistically -- since, unless Pollan is keeping a secret from us, he has no clue how these 50 to 100 compounds interact complexly, and so only understands their reductionistic contribution to health.
Lastly, there is an annoyingly persistent use of the "Who? Whom?" fallacy in Pollan's essay, whereby any scientific paradigm that is a boon to the food industry is put under a more powerful microscope than it would otherwise be. OK, sure, PR departments lie. But is not the more culture-centric approach a boon to Berkeley journalism professors like Pollan who can then make a comfortable living writing about it in the New York Times and bestselling books? Sure it is, but a hypothesis' truth or real-world relevance has nothing to do with whom it's a boon to. Every actor in the economy is a self-interested, self-deluded stooge who espouses certain beliefs because they support their position of status or power -- except for, conveniently enough, me!
Agnostic -- Where I part with Pollan is his view that culture and science are so clearly separate. And where I disagree with your take above is when you suggest "better science -- not relying on tradition." That too assumes that those are distinct spheres, which presumes that "science" is some clear, immutable "thing." Tradition is not so clearly enumerated, nor is science. But the human knowledge born of both is fascinating. BRC