Must Every Explanation Be Explained?

I don’t think we need to know the explanation for some proposed explanation in order for it to work as a provisional explanation.¹ But ultimately, there had better be an explanation for that explanation if it’s going to hold up as a real, genuine explanation. (FWIW, I touched on this point in my recent interview with Anthony Alberino).

Here’s what I mean: borrowed intelligibility only works if it eventually gets cashed out somewhere—that is, if there’s something that ultimately explains itself in some relevant respect. Otherwise, what seemed like an explanation turns out to be a sham. Take the classic example of how the earth is stable and the answer that it rests on the back of a turtle.² Let’s say, for argument’s sake, that we accept this as a provisional explanation. Fine—maybe the turtle explains the earth’s stability. But the obvious next question is: what explains the turtle’s stability?

If there’s no account for the turtle’s stability—if it’s just a brute fact—then the initial mystery we were trying to solve hasn’t really been resolved at all. In fact, it’s gotten worse! Because now we don’t just have the earth’s stability to explain; we’ve also got the turtle’s stability to account for.

Someone might push back here and say, “But the turtle is stabilizing the earth, so it’s still a valid explanation of the earth’s stability—even if the turtle’s stability itself is unexplained.” But this misses the point. If the turtle doesn’t account for stability by its own nature—if it requires further explanation for that particular feature—then it’s not truly resolving anything; or at least not in a way that satisfies the intellect along the relevant lines. What we really want to know is how some particular feature or caused property entered the—call it, metaphysical ecosystem—to begin with. In this case, the question simply starts with the earth, since that is (presumably) the earliest instance where we recognize a particular property that is not adequately accounted for by the immediate circumstances. But once a brute fact is admitted concerning that particular feature or property—whether early or late—we can ultimately see that the original question—how the earth is stable, meaning how it has that particular property, which it presumably does not possess inherently—hasn’t really been answered.

In other words, it’s no more answered than if we had resorted to a brute fact appeal earlier regarding the earth’s stability, without bringing the turtle into the picture at all. And if someone doesn’t think a brute fact is a good explanation in that instance—and, of course, they shouldn’t, because a brute fact is simply the claim that there is no explanation!—then they shouldn’t think they can achieve a good explanation simply by pushing the brute fact a little further out!

To give another example, imagine I’m wondering how my room is illuminated in the evening—that is, how there’s light coming in through the window. Then, imagine one guy—who just randomly happens to be standing in my bedroom (totally not weird)—tells me it’s just a brute fact and to go back to sleep. He’ll take care of everything. Including my children.

Another guy, who also just randomly happens to be there, tells me there’s a bean floating outside in the sky, reflecting light into my room. I then ask this second guy how the bean is reflecting light, and he just tells me it’s a brute fact and to go back to bed. He’ll also take care of everything.

Now, think about it. Is the mystery of my illuminated bedroom any less mysterious, ultimately, with the brute fact bean than it was without it? Would it be reasonable for me to respond to the second guy, “Oh, well, that takes care of it. Goodnight then!” Of course not! The original situation is no less mysterious—if anything, it’s even more mysterious. Now we’ve got this bizarre issue with the bean to contend with.

In other words, not only has the question of how my room acquired the property of illumination not been ultimately answered, but we now have even more questions—namely, how the bean is reflecting light and why it’s just randomly floating in the sky.

Unless there’s some legitimate explanation for how that bean is there reflecting light, we really don’t have an explanation for how my room is being illuminated—particularly since beans are not the sort of thing that can generate light by the principles of their own nature. (By contrast, if a bean were the sort of thing that could, by its nature, float and generate light, then I think we would have a legitimate explanation!³ And note: If one of the guys told me a bunch of fireflies were outside my window, that would satisfy my intellect in the relevant respect—though, of course, I’d immediately want to know why some random dude has entered my house, but that’s a question for another day!)

In short, just because the mystery is passed along doesn’t make it any less brute. Passing on bruteness is just passing on bruteness!

What’s the Point of All This?

Well, I guess there are several.

First, this point is both good for some theists and bad for others. In fact, it relates to an objection Richard Dawkins has raised about God: namely, that positing God as the “designer” of the universe immediately raises the question of who designed the designer. The claim, it seems, is that if we can’t explain why God exists (given his complexity), then there’s no use in positing God as the cause/designer/explanation of the universe.

Of course, many have taken Dawkins to task on this, pointing out that one doesn’t need to have an immediate explanation of the explanation for it to be a legitimate explanation. (Obviously, ignore for the moment the fact that God is not complex for classical theists, but absolutely ontologically simple.)

Again, I think that’s true—but only with certain qualifications. First, it’s true that one doesn’t need to immediately know the explanation of the explanation for it to be a legitimate explanation; however, I do think there had better be, ultimately, some explanation of the explanation to secure its legitimacy. (This is why the distinction between provisional and ultimate explanations matters.)

Indeed, if the theist is ultimately going to say that God has no explanation—that God is brute—then I think we run into the same sort of catastrophically unintelligible scenario raised earlier with the turtle and the earth. At this point, the atheist can, if they so choose, simply move the bruteness up a little and stop at the universe being a brute fact, rather than God. Just as someone skeptical of the turtle proposal might say the earth is stable as a matter of brute fact, the atheist could argue that the universe doesn’t need any further explanation and make additional claims about simplicity, parsimony, and so on.

Alternatively, the theist can do the better and smarter thing: argue that God is the only sort of thing that could be completely, intrinsically intelligible—that is to say, serve as the ultimate self-explained explainer of everything.⁴ This allows us (theists, I mean) to provide a truly principled ultimate explanation and escape brute facts altogether. Such an explanation would secure and legitimate all previously provisional explanations, grounding the complete intelligibility of reality. (The masterwork on this is Bernard Lonergan’s Insight).

Indeed, we don’t have to see exactly how God is self-explanatory; we can simply argue (as I’ve done in many places, including my book) that such an entity must exist and that when we consider what features it must have—or lack—it turns out to be pretty much what classical theists have always said about God.

1. Minimally, an explanation is that which removes mystery; that which somehow illumines the intellect.

2. For a unique development of this line of thought, particularly with respect to naturalism and scientific explanation, see Bogardus’s paper.

3. This intuition, I think, aligns with what Aquinas is getting at when discussing the difference between per accidens and per se causal series, and why the latter must necessarily terminate, while the former may not. In a per se causal series, the caused causes do not possess the relevant property in virtue of what they are—they are merely instruments, possessing the property derivatively. Thus, to achieve a legitimate explanation, one must trace back to something that does—that is, a primary cause (even if not temporally first).

4. Remember not all explanations are causal; I am not saying God is self-caused. Rather, I am saying God is the right sort of thing to be truly ontologically independent and uncaused.