Whole (14 page)

Baked into reductionist science is the assumption that the world operates in a linear way—that it operates on simple causality. What exactly do I mean by this? The classic conditions for proving that A causes B are three-fold:

1. A always precedes B.
   
2. B always follows A.
   
3. There is no C that could also cause B.
   

Not much wiggle room there. Certainly no room for messy, unpredictable, and complex interactions. No room for acknowledging systems that are too complicated to map out. No room for uncertainty of any kind. That’s why tobacco companies were able to get scientists to say that smoking doesn’t cause lung cancer: not all smokers develop lung cancer and not all lung cancers are attributable to smoking. In a reductionist universe, the statement “Smoking doesn’t cause lung cancer” is perfectly accurate. But it’s woefully inadequate when it comes to the practical issue of understanding the profound effect of tobacco on lung cancer, thus convincing people to stop smoking.

In the simple-causality reductionist view, the universe, ultimately, is as mechanical as a clock. Some reductionist philosophers of science have gone so far as to claim there’s no such thing as free will, since our very thoughts, emotions, and impulses are simply the result of chemical reactions that themselves were triggered by other chemical reactions, going back to the Big Bang itself.

As psychologist Abraham Maslow wisely observed, “If you only have a hammer, you tend to see every problem as a nail.” And if your only way of seeing assumes that the world operates on simple causality,
you’ll see simple causality everywhere, even where it doesn’t exist; we see the world, not as it is, but as we expect it to be. Reductionist research naturally produces reductionist findings. It can be no other way. The flip side is also true: since reductionist research assumes that simple causality is the way the world works, if we can’t find simple causality in our research subject it just means we must not be looking at it the right way, or we don’t have sufficient observational or computing power to reveal it. The only way to see the miraculous complexity of nature is to allow ourselves to do so.

But looking for complexity is a much harder task. Single-factor causality is much easier to measure, and gives much more satisfying (if ineffective) answers, since no matter how complex the system and its interactions are in reality, a good reductionist scientist still assumes that just one factor among the hundreds, thousands, or billions in the system is necessary and sufficient to cause the end result under study. Smokers get more cancer? That proves nothing to reductionists until you can isolate the single chemical in the cigarette that invariably causes cancer. When the effects of smoking are mitigated by lifestyle, nutrition, or whether the cigarette is a pleasurable interlude or a guilt-raising addiction, reductionist research must steadfastly ignore these complexities.

In one way, though, looking for complexity is actually easier than seeking rigid causality. Reductionism may work from simple models of causation, but those models often provide unexpected and unexplained findings, eventually suggesting complex and confusing (and sometimes totally implausible) solutions. Wholism, on the other hand, presumes complex models of causation in a way that suggests simple solutions. (You can’t get much simpler than, “Solve most of our health problems by eating more whole, plant-based foods”!)

In other words, reductionist research often requires the invention of
new
complexities—especially more complicated methods of study and explanation. There’s an old joke about a dairy farmer who could not get his cows to produce enough milk. He asked the local university for advice, and they sent a team of professors, headed by a theoretical physicist. After weeks of intensive study, the team returned to the university, where they pondered potential solutions. Finally the physicist returned to the farm with an answer to the production issue. But he prefaced his presentation
with a caveat: “This solution assumes spherical cows in a vacuum.” The physicists’ work, like that of reductionist nutritionists, is a whole lot of academic labor for a solution that doesn’t work in the real world. (No wonder one definition of the word
academic
is “moot”!)

Because I grew up on a real dairy farm, the study of spherical cows in a vacuum never occurred to me. When I entered academia, I tried to embrace the staggering complexity of biochemistry as the point and the challenge of my research. What could possibly be gained by trying to simplify it just to fit a theoretical framework?

I don’t want you to think that all of science is mired in reductionism. Particle physics, for example, chased and ultimately abandoned the reductionist dream of finding the “monad,” the elementary particle that could not be divided into anything smaller.

First physicists discovered atoms. Then the big subatomic particles that we learned about in school: protons, electrons, and neutrons. Then things started getting weird. Neutrinos, quarks, muons, bosons, fermions—each was anointed the elementary particle until theory or observation pointed toward yet another division. The closer the physicists looked, the more solid matter looked like mostly empty space with a tiny particle at its core. Now cutting-edge physicists see matter as simply a dense form of energy. It’s no accident that the recently discovered Higgs boson is nicknamed the “God particle.” Particle physicists realize that a comprehensive wholism underpins even the most reductionist mode of observation.

Many physicists point out in wonder the self-similarity between atoms, cells, planets, galaxies, and the universe as a whole (self-similarity among different levels is one of the hallmarks of a wholistic system). And the emergence of quantum theory in the twentieth century dealt a body blow to the reductionist paradigm by inserting uncertainty into what were supposed to be purely mechanical events. Theoretical physicist and popular author Stephen Hawking has written about subatomic particles that travel backward in time. The effect, known as retrocausality, suggests that certain effects can precede their causes. Talk about putting a nail in the coffin of cause-and-effect reductionism!

Yet many scientists still operate with both feet firmly planted in a seventeenth-century Newtonian universe—especially the ones (like nutritional scientists) responsible for studying human health and disease.

Scientists can argue philosophy all day long, but what really counts is evidence. This begs the question: What counts as evidence? What ways of looking for answers are considered good or bad science? Which methods are appropriate for what subjects of exploration?

The answers to these questions are themselves quite subjective, even if science believes itself to be an objective, value-free pursuit. They depend heavily on the questions being asked, and also on how the answers are sought. Epidemiologists, those scientists who study the causes of human health and disease, refer to the ways we explore scientific questions more formally as “study designs.” Let’s look at a few of the points on that continuum of study design, from highly wholistic to deeply reductionist. We’ll take a closer look at the difference between the two and the types of evidence they collect, as well as how they affect the kind of conclusions we draw from the resulting research—especially when it comes to nutrition.

One way to identify the optimal human diet, pretty obvious to all but fundamentalist reductionists, is to survey and compare populations as they already exist, and see what they eat and how healthy they are. Epidemiologists refer to this kind of study as ecological or observational. Its main characteristics include observation without intervention and looking at certain observable facts, like food intake and rates of disease, without trying to prove that one caused the other. Instead, researchers simply record the diet and disease characteristics of the populations as they are. If an ecological survey looks at those diet and disease rates in a group of people at more or less the same time, like a snapshot, it is called cross-sectional. The population under study can range in size from a small community of a few hundred people to a large country.

The results that ecological studies produce show associations between variables rather than proof that a particular input caused a particular output. These associations are often presented as correlations between input and output, the biological relevance and probable significance of
which are determined statistically. Hence a study like this is also known as correlational.

Since the data collected in these studies are averages for entire populations, it is not possible to conclude causality for individuals. If we try to read causality into the data, we make a mistake known as an ecological fallacy. We might observe for various populations, for example, that a higher concentration of cars, indicative of a richer society, is correlated with a higher risk of breast cancer, also present in richer societies. It doesn’t make sense to conclude that cars cause breast cancer, or to tell women fearful of breast cancer to avoid driving cars. Instead, it suggests that the two have something in common that warrants further study; the strength of an ecological study is its ability to highlight significant patterns and to compare the relative successes of different lifestyles. But because conclusions about specific causes cannot be made in this type of study, it is considered by reductionists to be a weak study design.

Our project in China (the main study highlighted in
The China Study)
was just such a cross-sectional, ecological study design. Using various kinds of evidence, we found that the higher the consumption of animal products in different regions of China, the greater the incidence of and mortality from a whole host of diseases, including various types of cancer, heart disease, stroke, and many others. Yet critics trumpeted that we could not claim that a plant-based diet had any effect on lowering disease rates based on that correlation, because our study design was not discriminating enough to make such a claim.

They’re right in one way, but they’re wrong in another. According to reductionist philosophy, it’s technically correct to say that we cannot claim that a WFPB diet reduces disease risk, any more than we could say that driving cars causes breast cancer. But on close examination, the analogy breaks down. We weren’t comparing one input (driving) with one output (breast cancer). Rather, we were looking at nutrition, which as we’ve seen is a staggeringly complex set of processes and interactions. There’s really no meaningful way to reduce nutrition to a single input. I constructed the China project on the hypothesis that the effects of nutrition on health are wholistic, not reductionist. In other words, I wasn’t interested in whether more vitamin C prevents the common cold; I wanted to determine, from a wholistic perspective, whether a particular diet led to markedly better health outcomes than other diets. One way to do that was to study the
people in an entire ecosystem—the rural population of China—who ate in a way markedly different from populations in the West. Using the rural population of China allowed us to consider a large-enough number and variety of lifestyle factors and health and disease conditions to see the big picture—the elephant, not just the trunk or tusk. We were able to investigate hypotheses that certain groups of foods are associated with certain diseases that share similar biochemical bases. That then let us assess whether there was something about those groups of foods that might be causing or preventing and remediating those diseases.

Another wholistic way of gaining insight into our “ideal” diet is to look at our nearest animal relatives—gorillas and chimps—and see what they eat, a strategy known as biomimicry. Primates’ diets haven’t changed much in tens of thousands of years, unlike those of humans. So we would expect a primate’s instinctual food choices to produce sustainably healthy outcomes. As well, primates in the wild haven’t been influenced by fast food commercials and government propaganda, so perhaps their instincts are more trustworthy than ours. Furthermore, wild primates don’t take drugs or undergo surgeries to deal with the effects of poor diets, so if a group of primates did eat unhealthy food, they probably would become too sick and obese to survive and reproduce.

According to Janine Benyus, author of
Biomimicry,
early humans probably used this wholistic research strategy to determine which plants were safe and which were toxic. After all, it makes evolutionary sense to let someone else serve as your taster!

While not conclusive, animal observation can give us a starting point for our own dietary explorations. For example, just noticing that chimps and gorillas have strong bones and muscles while eating WFPB undercuts the notion that humans need lots of animal protein to grow and maintain muscle mass. And of course we can point to the largest land animals in the world, elephants and hippos, whose 100 percent plant-based diets don’t seem to render them weak or scrawny.

In short, biomimicry reframes the issue of nutrition as one in which humans are seen as one species among many. Observing animals that resemble us can provide insight into diet in a way that observing human
eating habits, which have been affected by human technologies from agriculture to refrigeration to processing, can’t. It also identifies areas of current research where we may be wrong (i.e., by casting doubt) as well as suggesting areas of further reductionist inquiry.

A third wholistic approach is that of evolutionary biology, in which we examine our physiology and determine what our bodies have evolved to ingest and process. For example, we can look at the length of our digestive systems, the numbers and shape of our teeth, our upright postures, the shape of our jaws, and the pH of our stomachs, among many other characteristics, and compare those elements to known carnivores and herbivores. (We see, by the way, that we share almost all the characteristics of herbivores, and have almost nothing in common with carnivores.) By doing so, we can use reverse engineering to discover possibilities for the kinds of foods our bodies are “built” to eat.