How Humans Know

How Do We Humans Know What Is True?

However you cogitate over a range of potential replies to this deceptively complex query, one thing is fairly certain. You are unlikely to assert, at least not convincingly, that all we know comes from just one method. Knowledge does not issue only from, say, randomized controlled trials (RCTs).

Opinions about research methods, and their contributions to understanding, have taken on the same impassioned polarity as nearly all else in these fraught and divisive times. There is, indeed, a recurring campaign in the peer reviewed scientific literature to deride methods of inquiry other than RCTs, and disparage “knowledge” obtained by lesser methods.

Such “tyranny” of trial methodology is belied by public health history (e.g., the established harms of tobacco); substantially impertinent to the great imperatives of our time (e.g., pandemic control, responding to climate change); and fundamentally at odds with common knowledge born of basic sense. There are no RCTs to validate the merits of looking both ways before crossing a busy street; no RCTs to catalog the adverse outcomes when children run with scissors; no RCTs to establish the relative efficacy of water over gasoline for extinguishing a campfire; none, even, to demonstrate that touching fire results in burns. We recognize these truths just the same.

Perhaps no topic is more routinely subjected to the tyranny of the RCT than nutrition, where the claim recurs that little is known reliably for want of RCTs. ¹ Whatever the motivations for such contentions, they collectively provide cover for a massive pseudo-confusion about diet and health, ² with dire consequences. Mired in this unnecessary haze, diet has devolved into the single leading cause of premature death in the United States today. ³

There are, in fact, innumerable RCTs substantiating many fundamental truths about diet and health. But much of what we know as or even more reliably issues from other varieties of evidence. ⁴

In this issue of KWBW, we are provided a wider view of the inputs to human understanding than advocacy for one-method-fits-all. Martinez-Gonzalez and Trichopoulou characterize the population-level evidence supporting health benefits of the Mediterranean dietary pattern. Green and Allegrante highlight the need to source evidence in accord with its intended applications in practice. Gardner and colleagues show how readily the willful introduction of “strawmen” comparisons can corrupt the reliability of RCTs.

There is no single source of human understanding. Correlation—such as that between contact with fire and burns—can, in fact, equal causation when the pattern is unfailingly consistent, and mechanistically transparent. ⁵ Humble observation is the basis for much of what we humans know reliably, and apply routinely. RCTs have their valuable contribution to make, but like any given tool in a carpentry shop, they can be used well or badly. A saw is indispensable, but makes a poor hammer.

Human understanding has recourse to many modern tools. Knowing which to use when, and how best to combine their utilities, is itself a key consideration in the pursuit of sturdy, well-constructed truth.