This is a reply to The Christian Worldview is the Best Explanation by Jim Wallace.
Wallace says that just as a detective tries to come up with a story that best explains all the facts in a murder case, we all try to come up with a story – a worldview – that best explains everything we experience and know. So far, we agree.
Jim then says that Christianity offers a better explanation for things than naturalism does. And what does he mean by “better explanation”? He says that a good explanation
is feasible (it possesses “explanatory viability”), that it is simple (it has the most “explanatory power”), that it is exhaustive (it has the most “explanatory scope”), that it is logical (it has the most “explanatory consistency”) and that it is superior (it possesses “explanatory superiority”).
This is an “explanationist” account of abduction, according to which one compares rival explanations by the “explanatory virtues” they possess. But Jim’s account of explanatory virtues is strange, to say the least.
I’m not sure what it means for an explanation to be “feasible” or to possess “explanatory viability” – does that just mean it’s likely? Also, “explanatory power” is not simplicity – the two are different explanatory virtues. And what does “explanatory superiority” mean? That’s quite vague, and sounds like some kind of global assessment of explanatory virtue, perhaps.
To give you an idea of what a list of “explanatory virtues” or “explanatory desiderata” usually looks like when philosophers attempt this form of inference to the best explanation, here’s a list of explanatory virtues I’ve compiled from some of the leading thinkers on the subject from the past half-century: Peter Lipton, Gilbert Harmann, Wesley Salmon, William Lycan, Paul Thagard, and others.
- Testability: better explanations render specific predictions that can be falsified or corroborated.
- Scope (aka “comprehensiveness” or “consilience”): better explanations explain more types of phenomena.
- Precision: better explanations explain phenomena with greater precision.
- Simplicity: better explanations make use of fewer claims, especially fewer as yet unsupported claims (“lack of ad-hoc-ness”).
- Mechanism: better explanations provide more information about underlying mechanisms.
- Unification: better explanations unify apparently disparate phenomena (also sometimes called “consilience”).
- Predictive novelty: better explanations don’t just “retrodict” what we already know, but predict things we observe only after they are predicted.
- Analogy (aka “fit with background knowledge”): better explanations generally fit with what we already know with some certainty.
- Past explanatory success: better explanations fit within a tradition or trend with past explanatory success (e.g. astronomy, not astrology).
Even more surprising than Jim’s odd account of explanationism is the fact that he doesn’t give a single argument as to why Christianity is a better explanation for things, given his criteria, than naturalism!
In the second half of my lecture Why the New Atheists Failed and How to Defeat All Religious Arguments in One Easy Step, I considered a more standard list of explanatory virtues and explained why theism scores so poorly on all of them.
Indeed, theism as an explanation has much in common with what we know to be really bad explanations from pseudoscience and superstition, and almost nothing in common with what we know to be really good explanations from the physical sciences. So why should we think theism is a good explanation like those from science, rather than a really terrible explanation like those from pseudoscience and superstition? That’s what I’d like to hear from Jim Wallace.
Here is the relevant section of the transcript from my talk:
Let’s look at a list of arguments for the existence of God:
You’ve got cosmological arguments: God is the best explanation for why something exists rather than nothing.
You’ve got design arguments: God is the best explanation for certain complex things.
You’ve got moral arguments: God is the best explanation for why some things are really right and wrong.
And here are some other arguments proposed by the most important Christian philosopher alive today, Alvin Plantinga:
He says: God is the best explanation for the existence of mathematical sets.
God is the best explanation for our experience of flavor and color.
God is the best explanation for our appreciation of Mozart.
God is the best explanation for our experience of nostalgia.
So by now you might have noticed a problem. How does saying “God did it” explain any of these things? How does “God did it” offer a solution to any of the problems that philosophers and scientists are working on? When you’re confronted with a difficult problem, you can’t just say “Well, I guess it was magic.” That doesn’t solve anything!
“Poof! Magic” is not an explanation.
But I can’t just say that “Poof! Magic” is not explanation. I have to argue for it.
Scientists and philosophers, when looking for a best explanation, have identified some qualities that are often associated with good explanations. What is it that makes something a best explanation? What is it that makes one thing a good explanation, and another thing a not-so-good explanation?
Well, the first thing is that they are testable. In fact, if a theory wasn’t testable, it wouldn’t make much sense to say it’s the best explanation of something, because there’s no way for you to test whether it’s true or not! These theories render specific predictions, so you can go out in the world and see whether those predictions are true or false.
And of course, it should be not only testable, but it should pass the test…
Our most successful explanations also tend to be consistent with our background knowledge. If your new theory requires that we throw out everything we know about gravity and light and animals and humans, then that’s probably not the right theory. Consistency with background knowledge is important for a best explanation.
Successful explanations also tend to be simpler than alternatives. If a cookie is missing from a cookie jar, that’s probably just because Timmy took the cookie. It’s possible that the FBI and a gang of poltergeists conspired to use a time-stopping machine to freeze time and walk right past you and steal the cookie and then get away and unfreeze time again. That’s possible. But it’s extremely unlikely. Why? Because every element in that story is by itself unlikely, and the theory requires that they all be true, which is even more unlikely. So don’t add things to your theory that don’t need to be there.
Successful explanations should also have good explanatory scope, meaning they should explain a wide variety of data. Take, for example, the theory that those puffy white lines you see in the sky sometimes are from government planes dropping mind-control gas on all of us. One problem with this theory is that it might explain why New Yorkers see them, but it does not explain why you see lots of puffy white lines in the sky over deserts and oceans, where there are no minds to control. So that theory might explain some of the data, but it doesn’t explain all the data. It has weak explanatory scope.
So this is just the start of what scientists and philosophers look for in a good explanation, but we can already see there are major problems for the “God did it” theory.
For example, is the God hypothesis testable? No. Saying “God did it” renders no specific predictions for us to test, because God is all powerful and he could be responsible for anything. And theologians are very insistent on this because if they started to make the God hypothesis more specific and he would render specific predictions, it usually turns out that he fails the test. So they’ve been very careful to make God this mysterious, all-powerful thing and we don’t understand his purposes and he could be doing just about anything and we wouldn’t understand why. So there are no specific predictions that come out of the God hypothesis; there’s no way to test it. And it doesn’t make sense to say God is the best hypothesis if there’s no way to test whether or not that hypothesis is true.
What about the second criterion? Is the God hypothesis consistent with our background knowledge? Not at all. God is an extreme violation of our background knowledge about how things work. God is a person but he doesn’t have a body. God thinks, but without the passage of time. He knows everything, but he doesn’t have a brain. God is a terrible violation of our background knowledge in many serious ways.
Is the God hypothesis simple? If you’re talking about the God of the Bible, definitely not. The God of the Bible is an extraordinarily complex person; a being with thoughts and emotions who loves and hates and condemns and forgives; a being who turns a staff into a snake and a woman into salt; a being who changes his mind; a being who starts fires and throws rocks from the sky; a being who kills and resurrects; a being who takes part in personal relationships and political struggles; and a being who incarnates himself as a complex biological organism known as Jesus of Nazareth. The God of the Bible is far from simple.
And even if you’re talking about a more generic kind of God, God is not simple. Christian philosopher C. Stephen Layman lists four ways that a hypothesis can be simple, and in all 4 ways he admits that the God hypothesis is more complex than the atheistic hypothesis. But I don’t have time to go into that here. [See Letters to Doubting Thomas.]
What about explanatory scope? Does the God hypothesis have good explanatory scope? Again, no. I’ll give just one example. If you invoke God as the explanation for apparent design in the universe, you immediately run into the problem of all the incompetent and evil “design” in the universe.
So why is “God did it” a bad explanation? It’s because “God did it” lacks all the virtues we look for in successful explanations, and instead has many of the qualities that appear in terrible explanations, like explanations from pseudoscience and superstition.
Jim Wallace’s lack of care with explanationism, along with his total lack of interest in articulating why theism is a better explanation for things than naturalism, looks really bad. Frankly, it looks like he’s trying to sound philosophically informed so that he can reassure the masses: “Don’t worry, kids, smart people who use philosophical-sounding words believe Jesus is magic, too!”
If Jim was really interested in figuring out whether or not theism is a good explanation for things, he would have taken more care in examining his account of explanation, and he would have offered at least a hint of why he thinks theism is a better explanation for things than naturalism.
Frankly, Jim’s essay is a hallmark of apologetic method: a blissfully empty pretense of sophistication and truth-seeking, all to prop up ancient faith and dogma.