As a food producer, one of my interests is the relationship between diet and health. In part, this is because I think much of the conventional wisdom concerning the healthiest way to eat is flawed, and iconoclastic positions are exciting things to hold. (Which, in my opinion, can be one of the risks of holding them.) But what I find most fascinating about human diet is the way it highlights the limits of science.
Any dietary change that can improve health outcomes, even slightly, can have a significant positive impact when practiced by a large population. The incentive for organizations, public and private, to promote healthy eating habits is understandable, and the impulse to do so based on incomplete or ambiguous evidence is also understandable, because virtually all evidence in this particular area is incomplete, ambiguous, or both.
A few examples:
1.) Many studies, even the ones that make headlines in the NYT science section, are performed on mice or rats. Often, these mice or rats have been selectively bred for a particular trait, such as a predisposition to weight gain or diabetes. Human metabolism may not be identical to mouse metabolism, particularly selectively bred mouse metabolism.
2.) It is very difficult to establish a good control, particularly with macro-nutrients. For example, in an effort to study diets with a varying amount of protein, scientists might limit total calories to 1600. One group would get 40% of calories from protein, 30% each from fat and carbohydrates. The second "control" group might get 20% from protein, and 40% each from fat and carbohydrates. This would be a particularly poorly designed study, in that all three macro nutrient ratios are different. But even if carbohydrates or fat was kept constant, the other would still have to change to keep the caloric total identical. In other words, there is no way to have a single variable.
3.) Epidemiological studies are always problematic in that the most they can establish is correlation, but with something as complicated as diet, particularly when the studies are based on recalled food intake, they approach meaninglessness.
4.) All humans, so far as I know, suffer to a greater or lesser extent from confirmation bias. Researchers and institutions that have been offering strongly held opinions are likely to discover information that supports these opinions. I certainly view all dietary writing and studies through my particular prism, try as I do to judge them based on their objective merits.
If you're like me, this is about the point in an article when you start wondering what it has to do with gardening.
As previously mentioned on this blog, we have been making an effort to amend the soil of our garden to more ideal ratios, and we would eventually like to remineralize our entire farm. In part, this is justifiable in terms of simple economics - there comes a point at which a lack of calcium will seriously inhibit the growth of a crop or pasture. Yet I still believe it is worth the time to go beyond assuring that the soil has no gross deficiencies, and to make it have as close to perfect as possible a balance of minerals.
There are numerous difficulties.
A few examples:
1.) Soil tests are notoriously variable, based not only on the lab conducting the test, but even due to changing soil characteristics from one month to the next.
2.) There is no universally agreed upon ideal soil profile.
3.) Different amendments, even when providing the same quantity of a mineral by weight, behave quite differently.
4.) Soil is constantly changing - microbiological activity binds up certain nutrients while making others available, roots tap deep mineral sources and draw them to the surface, precipitation leaches each mineral at a different rate, organic matter and mycorrhizal activity in part regulate which of the nutrients in the soil are actually accessible to the the crop or forage.
5.) In trying to translate soil amendment to health, a long, flimsy chain is forged - amend soil (was it properly amended? are the minerals truly available for plant uptake?), grow a crop or forage (what variety? how did the weather conditions alter the quality of the crop?), feed this to a human or animal (all of the issues touched on in the discussion above), and then try to extract some substantial information from the results. William Albrecht's work is the most compelling I've read, but any attempt to deal with issues of this complexity will never be as certain as the link between vitamin C deficiency and scurvy. (And even this clear-cut deficiency disease is significantly more complicated than losing your teeth if you don't eat citrus.)
To me, this is a wonderful state of affairs. Depending on your viewpoint, the world either is or is not completely mechanistic. We have certainly organized large sections of our society around the portions of the world that can easily be mechanized for the sake of huge improvements in efficiency this allows. But what I'm describing, though proposed mechanisms and causal relationships can be speculated upon and doubtless do exist for even so complex a subject as the soil/health relationship, for human purposes, is not mechanistic.
We can take our farm and try one protocol for grazing or soil amending and get great results. But there are so many variables - hundreds beyond those few I've touched on - that it is impossible to foster a delusion that we actually know in the rigorous sense of the term what is going on, to what, if any, extent our actions had a positive result, let alone whether they would have a positive result somewhere else. Certainly, scrutiny and measurement are critical to assessing the impact of an action. But so are intuition, guesswork, gut feeling - all those horribly inaccurate, quite rightly unscientific human traits - and this is immensely enjoyable. There are few areas, let alone critically important areas such as food production, in which specific, unique decision making is practiced so regularly.
I enjoy science a great deal, and I think it is singularly capable of illuminating aspects of how the world works. Even in areas as fraught as diet and soil health, the amount of useful information available is staggering, and I'm grateful to have it. But faith in science is so ingrained that it can easily lead to the assumption that answers exist when they don't. So I will keep testing the soil and amending it in ways that I believe will improve its characteristics, because I believe this will ultimately lead to healthier land, healthier plants, healthier animals, healthier, happier people. Just don't ask me to prove it.