bookmark_borderAn F-Inductive Argument from Consciousness for Theism, Revisited

Edited on 15-Feb-20
While some theistic arguments are “God of the gaps” arguments, many, including those defended by Christian philosophers, are not “God of the gaps” arguments. Before accusing a theist of trotting out another “God-of-the-gaps” argument, atheists should first verify that the argument actually is a “God-of-the-gaps” argument.
Here is the basic structure of a “God-of-the-gaps” argument:

  1. Some odd or puzzling thing, E, occurs or exists.
  2. Science is unable to offer a plausible, God-free explanation for E.
  3. Therefore, God is the best explanation for E.
  4. Therefore, God exists.

There are many, well-known problems with such arguments. I’ve written on this topic elsewhere, so I won’t repeat those points here. Instead, I want to sketch how a theistic argument can avoid appealing to a gap in scientific knowledge. Here is the structure of an F-inductive argument from consciousness:
Let E=consciousness exists; N=naturalism; T=theism; B=background information; Pr(|H|)=the intrinsic probability of H; and Pr(x|y)=the epistemic probability of x conditional upon y

  1. E is known to be true, i.e., Pr(E) is close to 1.
  2. N is not intrinsically much more probable than T, i.e., Pr(|N|) is not much greater than Pr(|T|).
  3. Pr(E| T & B) > Pr(E | N & B).
  4. Therefore, other evidence held equal, N is probably false, i.e., Pr(T | B & E) > 0.5.

Whatever problems may exist with that argument, being a “God of the gaps” argument isn’t one of them. The present inability of science to explain consciousness plays no role whatsoever in the argument. What’s doing the work in the argument is the fact that theism, as a version of supernaturalism, entails that consciousness exists, whereas naturalism has no such entailment.
Allow me to explain. “Naturalism” is really just short-hand for “source physicalism,” which says that the physical world exists and, if the mental world exists, the physical explains why the mental exists (or, to allow for eliminative materialists, appears to exist). “Supernaturalism” is really just short-hand for “source idealism,” which says that a mental world exists and, if a physical world exists, the mental explains why the physical exists (or, to allow for eliminative idealists, appears to exist). “Theism” is a specific version of supernaturalism; it says that the mental being or entity which explains why the physical exists is a perfect supernatural person.
N.B. While theism does not entail human consciousness exists, theism does entail consciousness exists because theism entails that God exists and God is conscious, by definition. In contrast, naturalism is compatible with the non-existence of consciousness. So the existence of human consciousness, while not entailed by theism, isn’t surprising on theism in the way it is on naturalism. In that sense, human consciousness is evidence favoring theism over naturalism.
Objections to the Argument
Objection to (1): “We have no idea what ‘consciousness’, ‘mental,’ and ‘physical’ mean. Science can’t explain some E if the E is poorly defined.”
Reply: By a “mental world,” I mean the existence of a private, subjective world. By a “physical world,” I mean the existence of a public, objective world. By “consciousness,” I mean sentience.
Objection to (2): “But intrinsic probabilities don’t appeal to the propositions included in our background knowledge, and so ignore prior probabilities.”
Reply: As we say in computer science, that’s a feature, not a bug. Intrinsic probabilities come before prior probabilities. As the name implies, intrinsic probabilities are probabilities determined solely by the intrinsic properties of a proposition. Draper has argued (convincingly, in my opinion) that intrinsic probabilities are determined by scope, modesty, and nothing else. In contrast, prior probabilities are determined by the propositions in our background knowledge, such as “A physical universe exists,” “The universe is life-permitting,” “So much of the physical world is intelligible without appeal to supernatural agency,” and so forth.
Objection to (3): “The claim that Pr(E | T & B) > Pr(E | N & B) is unfounded because generic or mere theism doesn’t contain enough information to predict or demystify E. One would have to appeal to a specific kind of theism to justify something like (3), but a more specific kind of theism would have a lower intrinsic probability than mere theism.”
Reply: This is false for the reason explained above. While theism does not entail human consciousness exists, theism does entail consciousness exists because theism entails that God exists and God is conscious, by definition. In contrast, naturalism is compatible with the non-existence of consciousness. So the existence of human consciousness, while not entailed by theism, isn’t surprising on theism in the way it is on naturalism. In that sense, human consciousness is evidence favoring theism over naturalism.
Objection to (4): “But consciousness depends upon a physical brain. That’s more probable on naturalism than on theism.”
Reply: Correct. We know much more about the mental than the fact that it exists. We also know that it is dependent upon the brain, a fact which is much more likely on naturalism than on theism. So, once the evidence about consciousness is fully stated, it’s clear that there is also evidence favoring naturalism over theism. That fact, however, does nothing to refute this argument, which contains an “other evidence held equal” clause in its conclusion.
Objection to (4): “But the naturalistic evidence of mind-brain dependence outweighs the theistic evidence from consciousness.”
Reply: I am not aware of anyone having offered a successful argument for that claim. It’s not clear to me how such an argument could be adequately defended.
Objection to (4): “But the history of science contains numerous examples of naturalistic explanations replacing supernatural ones and no examples of supernatural explanations replacing naturalistic ones. That gives us reason to expect that science will eventually explain consciousness without God.”
Reply: I agree that gives us some reason to expect that science will eventually explain consciousness without appealing to God. That doesn’t change the fact, pointed out by (3), that the content of “naturalism,” as I have defined it, gives us no antecedent reason to expect consciousness to exist if naturalism is true, whereas “theism,” as I have defined it, does give us an antecedent reason to expect consciousness. This promissory naturalistic ‘atheodicy’ has no logical relevance to the argument anyway.
Objection to (4): “But (4) must be false because theism is false.”
Reply: That would follow only if one assumes that there can never be true evidence for a false proposition, but why assume that? There can be circumstantial evidence that a defendant is innocent of murder, while at the same time there could be other evidence for the defendant’s guilt, such as DNA evidence, which completely outweighs the circumstantial evidence. Similarly, a theist might say, “Suffering, imperfection, poor design, and mind-brain dependence are evidence against God’s existence, but that evidence is completely outweighed by the evidence from the finite age of the universe, the life-permitting conditions of the universe, human consciousness, etc.” Similarly, even if one believes (as I do) that it’s extremely improbable that God exists, one can consistently allow that consciousness is evidence–even strong evidence–favoring theism over naturalism, while simultaneously believing that other evidence outweighs the theistic evidence. People in general need to stop taking a binary, “all-or-nothing” approach to evidence.

bookmark_borderHow Theists Can Avoid God-of-the-Gaps Arguments and Still Argue for God

Background: In the context of a review of Dan Barker’s book, Godless, Randal Rauser had a very brief, even cryptic, exchange in the combox for his about God-of-the-Gaps (GOTG) arguments. (See here and here.) That exchange led to his latest post, which you can read for yourself here. I’ve decided to post my response on my own blog here, with some edits for further clarification.


I haven’t read Barker’s book, so I can only comment on what you have quoted:

“Many of these [theistic] arguments are reduced to a ‘god of the gaps’ strategy. At most, the theists might prove the existence of a current gap in human knowledge, but this does not justify filling the gap with their god. After all, what happens when the gap closes someday? The gaps are actually what drive science–if we had all the answers there would be no more science.” (Godless, 104-5)

Let’s start with the ‘god of the gaps’ (GOTG) strategy. What is a GOTG argument and why are such arguments so bad?
Theistic Argument Schema #1 (Focus on Gap in Scientific Knowledge)
GOTG arguments go like this.
(1) There is some fact, F, which science cannot explain today (in terms of naturalistic, mechanistic, unguided) causes
(2) [probable] Science will never explain be able to explain F. [inductive inference from 1]
(3) The existence of God does explain F.
(4) Therefore, the existence of God is the best explanation for F. [from 2 and 3]
(5) [probable] God exists. [inductive inference from (4)]
The key feature of schema #1 (and other schemas like it) is that “science cannot explain F today” plays a major role in the argument.
The move from (1) and (2) is weak. Science has been extremely successful in explaining a wide variety of phenomena in terms of naturalistic, mechanistic causes. Before we even get into the specifics of F, it’s already extremely likely that F has a naturalistic, scientific explanation. In Bayesian terms, “F has a naturalistic explanation” has a high prior probability. This is why I agree with the Barker quotation.
Theistic Argument Schema #2 (Focus on Content of Propositions)
(1) There is some fact F, we know to be true.
(2) The content of the proposition, “The mental exists and, if anything physical exists, explains why anything physical exists” (hereafter, “source idealism”), provides us with reason to expect F or, if it doesn’t provide a reason to expect F, makes F less surprising than it would be on “source physicalism.”
(3) The content of the proposition, “The physical exists and, if anything mental exists, explains why anything mental exists,” (hereafter, “source physicalism”), provides no reason to expect F (or it provides some reason, but less of a reason than what “source idealism” provides).
(4) Therefore, we’d expect F more on the assumption that source idealism is true than on the assumption source physicalism is true.
(5) [probable] Source physicalism is false.
The key feature of schema #2 (and other schemas like it) is that “science cannot explain F today” plays no role whatsoever in the argument. Although F might, indeed, be a fact that science has no explanation for, the lack of scientific explanation for F does zero work in the argument. (In fact, the lack of a scientific explanation for F isn’t even a premise in the argument!) What does do the work in the argument? The content of the propositions represented by the labels “source idealism” and “source physicalism.”
This is a major advantage of schema #2 over schema #1: because “science cannot explain F today” plays no role whatsoever in the argument, schema #2 makes an objection based on the history of science irrelevant. If I were a theist trying to make an argument for God’s existence based one some fact F–a fact which in principle could have a scientific explanation but currently does not (such as fine-tuning, origin of life, consciousness, free will, etc.)– I would use schema #2, not schema #1.
Example: An Argument from Consciousness Using Schema #2
For example, suppose we decide to adopt schema #2 as a “template” for theistic arguments (specifically, natural theology) and we want to try it out with consciousness. This would yield something like the following.
(1) Consciousness exists.
(2) The content of the proposition, “The mental exists and, if anything physical exists, explains why anything physical exists” (hereafter, “source idealism”), provides us with reason to expect consciousness or, if it doesn’t provide a reason to expect consciousness , makes consciousness less surprising than it would be on “source physicalism.”
(3) The content of the proposition, “The physical exists and, if anything mental exists, explains why anything mental exists” (hereafter, “source physicalism”), provides no reason to expect consciousness.
(4) Therefore, we’d expect consciousness much more on the assumption that source idealism is true than on the assumption source physicalism is true.
(5) [probable] Source physicalism is false.
In this case, source idealism not only ‘predicts’ that something mental exists, but it says that something mental explains the existence of everything physical. In other words, something irreducibly mental plays a ‘deep’ role in a theistic worldview. In contrast, source physicalism is logically compatible with the nonexistence of anything mental. If source physicalism is true, the only want to ‘get’ something mental is to have living organisms with bits of matter arranged in very specific and complex ways (e.g., organisms with brains or something very much like a brain). But source physicalism is logically compatible with all sorts of scenarios where such bits of matter never get into that kind of arrangement. For example, source physicalism is logically compatible with a possible world in which only one universe exists, the universe allows carbon-based life, carbon-based life arises through naturalistic abiogenesis mechanism, and then evolution never progresses past single-celled life. Source physicalism is also logically compatible with a similar possible world, but with no life whatsoever. And so on.
In a source physicalism world, mental phenomena like consciousness do not play the kind of ‘deep’ role that they play in a source idealism world. (That’s the whole point of source physicalism.) And so the existence of mental phenomena like consciousness–even if consciousness turns out to have a naturalistic, scientific explanation–is very surprising on source physicalism but expected on source idealism.
The ‘Catch’
If a theist decides to use schema #2, however, there is a catch: in order to maintain logical consistency, the theist is required to admit that there are good, parallel arguments against source idealism and for source physicalism.
For example, notice the symmetry in the definitions of source idealism and source physicalism: one is based upon the mental and the other based upon the physical. You might say that the argument from consciousness described above is a version of an argument family we can call ‘arguments from mentality.’ Source physicalists have a corresponding argument family of their own, what we can call ‘arguments from physicality.’ Similar to the argument I defend in this post, source physicalists can argue as follows.
(1) Matter exists.
(2) The content of the proposition, “The physical exists and, if anything mental exists, explains why anything mental exists” (hereafter, “source physicalism”), provides us with reason to expect matter or, if it doesn’t provide a reason to expect matter, it makes the existence of matter less surprising than it would be on “source idealism.”
(3) The content of the proposition, “The mental exists and, if anything physical exists, explains why anything physical exists” (hereafter, “source idealism”), provides us with no reason to expect expect matter.
(4) Therefore, we’d expect matter much more on the assumption that source physicalism is true than on the assumption source idealism is true.
(5) [probable] Source idealism is false.

bookmark_borderMoreland on Consciousness

(redated post originally published on 14 November 2011)
Re: http://www.jpmoreland.com/2010/11/18/critique-of-graham-oppys-objection/
There have been some further developments in this discussion. See:
Graham Oppy “Critical Notice of J. P. Moreland’s Consciousness and the Existence of God” European Journal for Philosophy of Religion 3, 1, 2011, 193-212
J. P. Moreland “Oppy on the Argument from Consciousness: A Rejoinder” European Journal for Philosophy of Religion 3, 1, 2011, 213-226
Graham Oppy “Consciousness in not Evidence for Theism” in C. Meister, J. P. Moreland, and K. Sweis (eds.) Oxford Contemporary Dialogues Oxford: OUP, forthcoming. (Should be out early in the new year. Also contains a chapter by Moreland, defending his argument from consciousness, which I haven’t yet seen.)
Re the above link to Moreland’s blog: In Arguing about Gods, I discuss two arguments from consciousness. First, I (briefly) consider the argument in Locke’s Essay Concerning Human Understanding that is critiqued by Mackie in The Miracle of Theism. Second, I provide a fairly lengthy discussion of the argument in Swinburne’s The Existence of God. I do not think that the main criticism that I make of these argument in Arguing about Gods is that “the theist’s use of personal explanation regarding consciousness is a bogus form of explanation” (cf. the claim in Moreland’s blog). (See p.401 of Arguing about Gods for a summary of five of the criticisms that I make of Swinburne’s argument. The claim that Moreland attributes to me is not among these five criticisms ….)
The most important point to note — vis a vis this discussion — I think, is this: The worst case for the naturalist is one in which ‘conscious state’ is an ideological primitive, with an ideologically primitive connection to ‘neural state’ (or the like). But, for theists like Moreland, ‘conscious state’ is evidently an ideological primitive — for, of course, Moreland thinks that God is conscious, and does not suppose that God’s consciousness is explained in terms of something else — and the connection between consciousness and the rest of God’s ‘state’ is also ideologically primitive. So, on a proper accounting of theoretical costs, the worst case for the naturalist is no worse than par with the view that Moreland defends. (And, of course, if the naturalist can provide a ‘reduction’ of consciousness, then the naturalist has a theoretically more virtuous position.) But, if this is right, the considerations about consciousness cannot possibly favour theism (regardless of the outcome of attempts to provide a naturalistic ‘reduction’ of consciousness).