Friendly Humanist

This is a review of the second essay in the book Contending with Christianity’s Critics.

At home in the multiverse? by James Daniel Sinclair

Sinclair sets his sights on the multiverse, one of the leading contenders for a sound naturalistic explanation of apparent fine-tuning. I will pick out some highlights.

First, let me say that the fine-tuning argument – the latest and least ambitious incarnation of the ancient argument from design – has always seemed to me to be the strongest argument for the existence of a god. But, having read accounts of it from both sides, I’ve come to feel that its strength lies mainly in our anthropocentric biases rather than any logical superiority it possesses. (See Luke Muehlhauser’s discussion of Fine Tuning at Common Sense Atheism.)

Sinclair also commits some curious blunders. For example, he says that science flatly rejects gods as impossible. Certainly, few modern scientists consider gods as possible explanations. But that is largely because they’ve learned the lessons of history. Early scientists (such as the ancient Greeks, Newton, and Darwin in his youth) did believe – at least in some deistic lawgiver, if not a full-on personal god. But those beliefs got them nowhere in terms of explanation, so modern science tends to skepticism about the usefulness of gods as explanations. Also, look at Dawkins. Sure, he rejects the god hypothesis, but he does so only after evaluating it within a scientific framework. There are scientists, even atheistic ones, who assert that god is outside their purview, but that is not a universal belief among scientists.

As another example, I will share an interesting passage that presents a multiverse version of the ontological argument.

Jay Richards asks us to consider another refutation of an atheist Many Worlds: Christian Alvin Plantinga’s modal version of the ontological argument. In the strong version of the SAP, all possible worlds are considered actual. But if this is so, then if it is even remotely possible that God (the necessary being) has reality (i.e., He is in one possible world), then this necessity implies He must be present in all possible worlds. In essence, an atheistic attempt to produce a necessary universe produces God-as-computer-virus which propagates to “infect” every world! As Richards states, “Such can be the penalty for toying with notions such as possibility, necessity, and infinite sets.” (pages 22-23)*

This argument suffers not only from the linguistic defect of Anselm’s original ontological argument; it also commits a fatal equivocation. Anselm’s key error was to treat “existence” as the same sort of property as “redness”. That error is repeated here. The equivocation in the multiverse version above has to do with whether the god exists separately for separate universes, or exists transcendently, a single presence spanning them all. On the one hand, if the god’s existence in universe A is a different question from the god’s existence in universe B, then it is true that the probability of the god existing in some universe increases as the number of universes increases. On the other hand, if the god is equally present across all the universes by definition, then the probability of its existing is unaffected by contingent details like the number of universes. The above argument switches definitions at a crucial point. In a more valid form, the argument can give you either a probably-existing contingent god in a small subset of universes, or a very unlikely god that is present across all universes.

In truth, I don’t know if a multiverse approach is worth pursuing. I don’t know if it solves the apparent problem of fine-tuning. But then, after reading Luke Muehlhauser’s thoughts on the issue, I’m not convinced that fine tuning is a legitimate “problem” for naturalism that requires a solution.

At any rate, I don’t see that this essay gives any reason to shift my beliefs.

Footnote:

* Sinclair references this online paper by Richards as the source of this argument.

This is a review of the first essay in the book Contending with Christianity’s Critics.

Dawkins’s Delusion by William Lane Craig.

In the first essay, William Lane Craig outlines what appears to be Dawkins’ main argument against belief in god from The God Delusion. Here is the structure, as Craig presents it:

1. There is an appearance of design in the universe.
2. A designer is one way to try to explain the appearance of design.
3. Positing a designer raises the question of who designed the designer.
4. The best explanation we have for the emergence of complex things is evolution by natural selection.
5. We have no equivalent explanation for physics.
6. We should maintain hope that such an explanation may turn up.
7. Therefore, God almost certainly does not exist.

Craig correctly points out that this is a crashingly bad argument. The conclusion does not follow from the premises, and point 3 in particular seems to raise the spectre of an infinite regress of explanations. But is this a fair assessment of Dawkins’ argument?

No. It ignores the very important aspect of explanations that they be simplifying. That is, you have a simpler account of things after adopting the explanation than you had before. Dawkins harps on about this rather a lot in his book. Craig may not agree that simplicity is a key virtue of a successful explanation, or that a creator god fails the simplicity test; but he really should acknowledge that this is part of Dawkins’ argument. This answers, I think, the problem of the infinite regress of explanations. What I read Dawkins as meaning is that, if your explanation fails to simplify things, then the only reason we would have to adopt your explanation if, behind it, there isanother explanation that does simplify things.

Now, I realize that this may be me projecting rather than successfully reading Dawkins’ original intent. But that doesn’t really matter. The point here is not an atheist apologetic (“What is the true meaning of the text?”) but an attempt to get the best understanding of reality. So here is my reformulation of Craig’s version of the argument.

1. The universe exhibits the appearance of design.
2. A designer is one purported explanation of the appearance of design.
3. Generally speaking, appeals to a designer fail as explanations because:
   1. they fail to systematically predict actual observed phenomena and rule out phenomena we do not observe, and
   2. they fail the test of simplicity, relative to naturalistic alternative explanations.
4. In the past, comparable design arguments have been countered by the very powerful and well-evidenced theory of evolution by natural selection.
5. Although not yet as evidentially-supported as evolution, naturalistic explanations of the appearance of fine-tuning – such as the multiverse – are available and being explored.
6. We therefore have good grounds for optimism that naturalistic explanations will prove more empirically successful than theistic explanations for the appearance of fine-tuning.
7. Therefore, we should prefer the more parsimonious no-god hypothesis until substantial contrary evidence arises.

I know, it is not watertight. Nor is it a deductive argument. Like any scientific argument, it is inductive – seeking the most likely explanation of the observations available.

And despite my disclaimer, I do think that it is closer than Craig’s version to the actual arguments presented by Dawkins. (But again, I don’t want to pretend that my goal is to faithfully parrot Dawkins; please don’t attribute any of my errors to Dawkins. If you want to know what Dawkins says, read Dawkins.)

So much for the first essay in the series.