That Explains It Well

The role and usefulness of explanatory power in the epistemology of science, history, and apologetics.

Being 100% certain about something is a rare luxury. More often, we can only know something beyond reasonable doubt, and that is enough. In those situations, the concept of explanatory power is handy, whether the topic is science, history, or the epistemic grounding of Christianity.

Science’s crown jewel is the scientific method. I definitely appreciate the certainty and verifiability of scientific laws. But I think that science’s real power lies in the ability to ruthlessly follow the evidence, and to evolve its paradigms and theories accordingly. A helpful concept when comparing or evaluating competing theories is explanatory power, the ability of these theories to effectively explain or describe their subject matter.

This definition is a bit lax. A key aspect of explanatory power is what criteria to use to better measure or define it. Some options include: which theory can account for more of the observed facts; which theory has better predictive power; which theory makes less assumptions; and which theory is more verifiable or even falsifiable 1.

One of my favourite examples of explanatory power is gravity. This may be shocking to you: isn’t gravity well established, being in fact a law? Yes, but it’s not without caveats. Gravity doesn’t work particularly well at the subatomic level, and it doesn’t work particularly well in some conditions at galactic scale. In fact, Einstein’s relativity made important changes to the definitions and equations of gravity; and there are still unaccounted situations, and competing theories or proposed changes, such as modified Newton dynamics.

But to me, the most fascinating aspect of gravity is that we can’t measure or observe it directly. Light is visible; electricity can be visible; several of their aspects can be directly measured, and their effects on matter can be separately measured and observed too. Imagine that I go out in the sun with a glass of water. I can measure the sunlight’s luminous intensity, flux and illuminance. (You might recognize these more readily by their measurement units: candelas, lumens, or lumens over surface unit, respectively.) On top of that, separately, I can measure the effect of solar energy on the temperature of the glass of water. A similar reasoning applies to electrical charge, voltage and current.

In contrast, we don’t measure gravity directly. Instead, we measure its effect on bodies – the force it exerts on them, manifested as acceleration under the right circumstances. And yet, gravity can be very well modeled using the same formulas we use for electrical charge (with different coefficients). Given the mass of two bodies, we can predict their mutual gravitational pull. We can look at a remote planet, and estimate its mass based on the effect of its gravity on the orbits of nearby objects. We can predict exactly how much thrust a spacecraft needs to take off from that planet. And gravity is trivially falsifiable. That is a lot of explanatory power for something we strictly cannot even observe directly!

The usefulness of explanatory power goes beyond science. In history, we don’t have objective measurements, and we can’t use experiments to reproduce situations and outcomes. We can’t empirically verify that Canada became a confederation in 1867. In this realm, epistemology – obtaining knowledge – relies on the quality and trustworthiness of testimony and facts. Sometimes, facts don’t point to a single unequivocal explanation; multiple different theories compete. Disasters exhibit this unpleasant trait; black boxes don’t always survive airplane crashes, and accidents in extreme circumstances, like alpinism or deep diving, aren’t always well attested. Several theories can attempt to explain what happened. The one that better accommodates most facts (has the greater scope), makes the evidence more probable, better fits with other pieces of knowledge (and in fact, conflicts with the least of these too) is likely the better explanation2.

Because of this, explanatory power is also useful when establishing the epistemic grounds of Christianity as a worldview. The Principle of Causality, and the current state of the art in cosmology, impart the theistic account of creation with the greatest explanatory power. The same goes for the fact that DNA is essentially code or information, which science shows can never increase in a self-contained system. Again, the account of an external Designer and provider of this information has the greatest explanatory power. And the archaeological, historical and literary analysis of the Gospels make the resurrection of Jesus the hypothesis with the most explanatory power.

What do these theories have in common? They all involve a supernatural Entity or behaviour disrupting an otherwise natural flow of events. This is a stumbling block to many, but at that point, we need to ask ourselves: is a strict limit to natural causes and phenomena an intrinsic attribute of science? A legitimate application of the scientific method? Or a philosophical presupposition?

When it comes to natural phenomena, explanatory power is an elegant and useful concept. I’ll let you ponder on the explanatory power of this idea: that elegance and usefulness also extend to supernatural explanations. In other words, the idea that “explanatory power” applies to the supernatural too seems to carry plenty of explanatory power itself.

  1. I adapted these criteria from the Wikipedia entry on explanatory power. ↩︎

  2. I adapted these criteria from this William Lane Craig article: (the article has nothing to do with disasters) ↩︎

Tags: #science , #epistemology

Published: December 8, 2021