The First Test of a Logical Explanation

Imagine you come home one day and find one of your windows is broken. Shards of glass are lying both inside and outside the broken pane, but you don’t see any other foreign materials or markings near the window. What might have happened?

Among the many possible scenarios we could invent to try explaining the broken window, a few might include:

• A burglar smashed it.
• A baseball went through it.
• Someone inside the house broke it by accident.
• A UFO hit it.

If you want to be Latin about it, the thing being explained—in this case, a broken window—is called the explanandum, and the candidate explanations are called explanans. How can you tell which of the explanans is most likely true? Let’s begin unpacking the logic of explanations to find out, as being able to narrow down explanans comes in handy for thinking about everything from historical science1 to conspiracy theories.

The Logic of Explanations

One distinction to keep in mind is that explanations are not arguments. Arguments present a set of reasons for why we should believe a conclusion is true. But explanations are themselves conclusions. For instance, we might conclude, “the window is broken because a burglar smashed it.” Such statements form the basis for a type of argument called inference to the best explanation. These arguments adopt the following structure:

• X is an observable phenomenon.
• Y is the best explanation for X.
• Therefore, Y is most likely true.

Inference to the best explanation involves a form of inductive reasoning2 called abduction, which reasons from little facts, which we know are true, to a broader explanation which may or may not be true. Because inductive processes cannot prove conclusions are true, abduction doesn’t guarantee the best explanation is the right explanation. It’s just the explanation which currently makes the most sense compared to all the other explanations we know about.

Evaluating Explanations

When it comes to evaluating explanations, we can avoid some mistakes by applying the critical thinking hack of asking, “is this message true or false because . . . ?” For example, is an explanation true because it exists, or false because it sounds weird and unbelievable? Not necessarily. To find out if an explanation is likely true, we need to examine the explanation itself in comparison to observable facts and competing explanations.

So, how do we compare explanations? Should we simply look for the one with the most “evidence?” Certainly, the amount of evidence for an explanation (that is, is the number of observable facts that are consistent with the explanation) is essential to consider. But, as my critical thinking textbook from university emphasized, the quantity of evidence is not the only factor to consider. This is especially true because the same facts may often be interpreted as “evidence” for different explanations. The same broken window might be called “evidence” for a UFO collision or for a break-in. The same fossil might be viewed as “evidence” for gradual permineralization over millions of years or for rapid burial in a global flood.

The First Test of a Reasonable Explanation

When multiple explanations exist for the same “evidence,” how can we narrow down which explanation is best? First off, a believable explanation must meet the minimum requirement of internal and external consistency. It cannot include or lead to contradictory statements, and it must fit what we know about observable reality.

For example, the Bible shows remarkable internal consistency, and a biblical explanation for life’s origins matches what we know of observable reality, including the observation that living entities only come from other living entities. But a strictly material origins explanation is inconsistent with observations, including life comes only from life and information comes only from intelligence. And it leads to internally inconsistent conclusions, such as believing in the necessary existence of immaterial laws of logic when such laws would have no foundation in a purely material universe. As these and many other examples illustrate, a biblical worldview, but not a naturalistic one, passes the internal and external consistency tests for explaining the world around us.

Summing Up

To recap, inference to the best explanation is a form of inductive reasoning that helps us identify the best explanans (interpretation) for an explanandum (observation)—whether a living cell, a fossil, or a broken window. The best explanation isn’t necessarily the right explanation; it’s just the most logical one based on what we currently know. For an explanation to be considered remotely logical, however, it must first meet the minimum requirement of internal and external consistency.

What happens if multiple explanations pass this internal and external consistency test? In that case, we can narrow the explanations down further with five other criteria, which part 2 of this article will unpack. Stay tuned!