Paul Draper on God and the Burden of Proof

My friend Steve Hays (erstwhile classmate at Westminster Seminary in California back in the 1990’s!) passed along to me a link to an interesting two-page proof by Paul Draper (Professor of Philosophy at Purdue University) about “God and the Burden of Proof”. I wrote two emails in reply, and now Hays is coercing me to post them.

Draper provides four definitions:

1. Theist = a person who believes (that is, has a credence of .9 or higher) that a theistic God (that is, a personal creator of the physical world who is worthy of worship and unconditional allegiance) exists.
2. Atheist = a person who believes (that is, has a credence of .9 or higher) that no theistic God exists.
3. Naturalism is the statement that the physical world existed prior to any mental world and caused any mental world to come into existence.
4. Supernaturalism is the statement that the mental world existed prior to any physical world and caused any physical world to come into existence.

Draper claims:

“4.d Naturalism and supernaturalism are equally probable intrinsically because they are equally modest and coherent.”

… which we’ll concede for the sake of argument. Draper’s point is that these two ‘worldview-sized’ ideas — that the mental comes from the physical, and that the physical comes from the mental — are on a par with respect to modesty and coherence. There is no difference between the views except the direction of causation and explanation.

What does the most work in Draper’s arguments are premises 5.a and 5.b. They also strike me as the most dubious premises:

“5a. Theism is a very specific version of supernaturalism and so is many times (i.e. at least 10 times) less probable intrinsically than supernaturalism.”

“5b. Naturalism is a specific version of atheism and so is many times less probable than atheism.”

So on Draper’s view “theism is a very specific version of supernaturalism” but atheism isn’t a very specific version of naturalism. Draper needs this latter denial, or else the symmetry with respect to plausibility remains. Rather, he reverses the relation in the latter case: naturalism is a specific version of atheism.

So Draper is saying at least two things:

First, according to 5.a, believing in the existence of “a personal creator of the physical world who is worthy of worship and unconditional allegiance” further specifies the view that “the mental world existed prior to any physical world and caused any physical world to come into existence.” That is, theism is just one of many ways to stipulate the priority of the mental over the physical. The ‘mental’ that precedes the physical is personal, a creator, worthy of worship, etc. Presumably, other supernaturalists could speak of the mental in impersonal terms. They would be supernaturalists, but not theists. So far, so good.

Second, according to 5.b, believing that no theistic God exists doesn’t further specify the naturalistic view that “the physical world existed prior to any mental world and caused any mental world to come into existence.” Atheism doesn’t further specify naturalism. (Rather, naturalism further specifies atheism.)

So according to Draper, naturalism is one among many ways to be an atheist. Atheism is not one among many ways to be a naturalist. For instance, perhaps the mental and the physical are co-eternal brute realities, neither explaining the other. That would be a form of atheism (no personal creator of the physical), but it wouldn’t be naturalism (as defined by Draper).

Given these four definitions, as well as 4.d, 5.a, and 5.b, we can infer Draper’s conclusion:

“Thus, since naturalism and supernaturalism are equally probable intrinsically, it follows that atheism is many times more probable intrinsically than theism, which entails that atheism has a high intrinsic probability (certainly higher than .9) while theism has a very low intrinsic probability (certainly lower than .1). Therefore, theists but not atheists bear the burden of proof.”

One strength of Draper’s argument is that if naturalism and supernaturalism are on a probability par, then whatever the probability of supernaturalism, theism will be less probable than that. By way of contrast, whatever the probability of naturalism, atheism will be more probable than that. On this view you don’t need to know the prior probability of naturalism or supernaturalism. You just need a reason to think they’re equal, and Draper’s conclusion will follow.

Three problems as I see it. First, on that reasoning, supernaturalism could be one among many ways to be a theist, and so the symmetry would remain. For instance, perhaps the only things that exist are (immaterial) perceivers and their ideas, and there is no matter, and God is the most important Perceiver of all. That would be a form of theism, but it wouldn’t be supernaturalism (as defined by Draper). To make this reply work, we’d have to simplify Draper’s definition of “theism”. Rather than God being “a personal creator of the physical world who is worthy of worship and unconditional allegiance,” we drop the word ‘physical’ and just call God “a personal creator of the world who is worthy of worship and unconditional allegiance.” We don’t specify whether the world God has created is mental or physical. Theism is compatible with both. (This comports with the dualist and idealist traditions in Christian thought, so it isn’t ad hoc. Indeed, Draper’s definition of “theism” arbitrarily excludes the idealist tradition.)

(Another example: Mormonism seems to teach that a physical god created the physical universe. If so, this would seem to be a variety of theism as Draper defines it (“a personal creator of the physical world who is worthy of worship and unconditional allegiance”) but it wouldn’t be supernaturalism as Draper defines it (“the mental world existed prior to any physical world and caused any physical world to come into existence”). So supernaturalism (the mental brings about the physical), Christian idealism (the mental brings about the mental), and Mormonism (the physical brings about the mental) are all ways to be a theist (there is a personal creator of the world), and so supernaturalism is just one among many ways to be a theist.)

So Draper has given us nothing that would break the ‘intrinsic improbability’ impasse between theism and atheism. He who affirms must prove, and that burden applies to both. Theists give arguments. So do atheists.

Second, once we (more modestly!) define “theism” in this way, it turns out that atheism is belief that there is no “personal creator of the world who is worthy of worship and unconditional allegiance.” And this seems right. Atheists do not just deny a personal creator of the physical world. Presumably, they would deny a personal creator of the world (whether physical or mental). For instance, if abstract entities or mental thoughts exist, atheists would deny that there is a personal creator of all of those as well.

And now we see the other side of the coin. We can now (contra Draper) regard atheism as one among many ways to be a naturalist. The naturalist says: “the physical world existed prior to any mental world and caused any mental world to come into existence.” Well, guess what? That’s compatible with pagan theistic eliminativism. Perhaps a physical pagan God (Zeus) caused the rest of the physical world, and there is no mental reality (just physical reality). Atheists won’t be happy about that view. So they will further specify naturalism and deny any personal creator of the world at all (no mental/spiritual creator, no physical creator like Zeus). Viola! Atheism is now a specific version of naturalism, which is the very thing Draper denies.

Third, Draper defines “coherence” as “The coherence of a hypothesis depends on how well its components fit together,” and then he claims in 4.d: “Naturalism and supernaturalism are equally probable intrinsically because they are equally modest and coherent.”

So Draper has committed himself to something: the idea that “the physical world existed prior to any mental world and caused any mental world to come into existence” is just as coherent as the idea that “the statement that the mental world existed prior to any physical world and caused any physical world to come into existence.” I don’t buy that. A physicalistic explanation of the mental is not as coherent as a mentalistic explanation of the physical, since there are severe difficulties in explaining the mental in terms of the physical.

To summarize: Draper only breaks the intrinsic probabilistic symmetry of theism and atheism given a dubiously narrow construal of ‘theism’ and an overly-optimistic understanding of the coherence of a physicalistic basis of the mental.

Here’s one way to visualize what’s going on. We can represent Draper’s reasoning by way of the following chart:

Column 1		Column 2		Column 3
	[<—less probable]		[more probable—>]
theism	[further specifies]	supernaturalism
		naturalism	[further specifies]	atheism

As you go from left to right (from column 1 to column 2 to column 3), positions become more probable. (As you go from right to left, positions become less probable.)

Draper’s claim about the central column is that supernaturalism and naturalism are on an intrinsic probability par; that’s why they’re in the same column. In addition, if position X further specifies (i.e., is just one of many versions of) position Y, then X is less probable than Y.

Given the two claims above, implicit in the two rows above, theism is far less intrinsically probable than atheism, and that’s why theism (and not atheism) bears the burden of proof, according to Draper.

To defeat Draper’s argument, we can argue that as a matter of fact supernaturalism further specifies theism, since supernaturalism (as Draper defines it, such that the physical comes from the mental) is just one view of how there could be a personal creator of the world (there are other theistic views, such as Christian idealistic views in which there are just perceivers and their ideas, or Mormonism in which the physical comes from a personal, physical creator).

If we were represent this point by reversing the relation in the first row, we’d get:

Column 1	Column 2	Column 3
[<—less probable]		[more probable—>]
supernaturalism	[further specifies]	theism
naturalism	[further specifies]	atheism

… and there is once again symmetry between theism and atheism. Both bear the burden of proof.

Alternatively, we can argue that as a matter of fact atheism further specifies naturalism, since there are at least two ways to further specify naturalism as Draper defines it: pagan theistic eliminativism (in which a physical creator creates a thoroughly physical world, and there are no mental events), and atheism (there is no personal creator of the physical world). Thus, atheism would just be one version of naturalism, and so further specifies naturalism.

If we were represent this point by reversing the relation in the second row of the first chart, we’d get:

Column 1	Column 2	Column 3
[<—less probable]		[more probable—>]
theism	[further specifies]	supernaturalism
atheism	[further specifies]	naturalism

… and here there is symmetry between theism and atheism as well. Both bear the burden of proof.

Finally, it’s dubious that positing a physicalism about the mental (the latter arises from the former) is “just as coherent” as positing an intentional mental source of the physical. (This is contentious, of course; I’m simply giving Thomas Nagel et al their due.)

As far as I can tell, each of these three points is independent of the others. But you might consider: add the first two points together (they are compatible, after all, and can be independently argued). What you get is:

Column 1		Column 2		Column 3
	[<—less probable]		[more probable—>]
		supernaturalism	[further specifies]	theism
atheism	[further specifies]	naturalism

So it turns out that atheism is far less intrinsically probable than theism, and thanks go to Draper for supplying this schema!

Let me close by saying that I don’t think Draper intends his argument as a reason to ignore evidence for theism. Rather, it’s supposed to be a reason for thinking that only theists have to provide evidence for their view in order to be rational. On Draper’s view, atheists are rational without providing evidence for their atheism. It’s compatible with this to say: but they should respond to theistic evidence if provided. I think Draper is committed to that.