Do We Need a Finely-Tuned God?

I have recently had a most interesting e-mail exchange with Professor Jason Waller of Kenyon College. His new book Cosmological Fine-Tuning Arguments will be coming out soon from Routledge. While, of course, I do not accept the fine-tuning argument  (FTA), I think Prof. Waller’s treatment is, by far, the best defense of it that I have seen to date. Here is the blurb I offered for the book:

Jason Waller’s Cosmological Fine-Tuning is exactly what is needed by anyone interested in this fascinating topic. He has mastered the technical issues and presented them clearly and accurately. The argument is judicious, fair, and nuanced and the conclusions are solidly supported. Highly recommended.

In a recent book of my own, Polarized: The Collapse of Truth, Civility, and Community in Divided Times (Prometheus Books, 2019), written with my old friend the Rev. Dr. Paris N. Donehoo, in my introductory remarks I offered a brief statement of the reasons that I am an atheist. It included these lines:

As for “fine tuning,” the argument is that, of all the possible ultimate, uncaused physical realities that might have been, we were, given naturalism, impossibly lucky to have one that permits our existence. Therefore, there must have been an a supernatural intelligent designer to actualize life-friendly conditions out of those vastly numerous possibilities. My reply is that, by precisely the same reasoning, of all the possible ultimate, uncaused supernatural entities that conceivably could have been, we were impossibly lucky to get one that both wanted our existence and had the power to create and control a physical universe so that it would produce us. Any posited ultimate (logically) contingent fact—natural or supernatural—will necessarily be only one of infinitely many alternative realities that were equally logically possible. Sauce for the gander. The upshot, on my view, is that it is meaningless to speak of objective probabilities of ultimate posits. Whatever exists as the ultimate, uncaused reality is neither probable nor improbable. It just is.

I sent these lines to Prof. Waller and he kindly sent a detailed response. Below is part of his reply:

…[T]he claim that we were impossibly lucky to get the right kind of god is really interesting.  I have two general thoughts about this claim.  First, the fact that the supposed god wanted a fine-tuned universe is part of the hypothesis itself.  As an analogy, suppose that we found strange crop circles in a corn field and are developing hypotheses to explain the appearance of exactly these particular shapes.  One hypothesis is the Alien Hypothesis (AH) on this hypothesis there are otherwise unknown aliens who had some unknown reason for creating exactly this shape in the cornfield. This is not overfitting the hypothesis to the data because the only reason for supposing that such aliens exist is to explain this phenomenon.  So it would be a mistake, I think, to claim that we were impossibly “lucky” to get aliens that wanted exactly the shape we happen to find.  The hypothesis under consideration includes the claim that the aliens wanted this shape.  In much the same way, if there is some weird graffiti on my car one morning I suppose that there is some youth who had some unknown reason for making this exact design.  Simply positing a youth who vandalizes cars or an alien who makes odd shapes in cornfields is really not going to help because it is presumably more likely that such a person or being would NOT desire the outcome than would desire it.  So we have to include the claim that the unknown person or being desired this outcome in the hypothesis itself.  Second, I think it is a major mistake for the theist to claim that the god is a contingent brute fact reality.  I absolutely agree with you that if our two hypotheses are (B) a brute fact universe or (B+) a brute fact god who made a contingent universe, then the first is theoretically simpler and so is to be preferred (all else being equal.) Much like positing a brute fact universe creating machine that made only one universe offers no theoretical advantage.  The best approach for the theist, I think, is the claim that the god is metaphysically necessary meaning that it exists in all possible worlds.  That way the same fine-tuning question cannot be raised again at the level of the god thing.  Why does the god thing exist?  It has to–it exists in all possible worlds.  Now the naturalist can make the same move without giving up on the contingency of our universe.  Of course, the naturalist could say that this is the only metaphysically possible universe and so it exists in all possible worlds.  But then we would need to radically rethink our modal intuitions.  To get a contingent universe it seems as though we either need the universe to be a brute fact or we need a necessary being with free will.  Otherwise, its Spinoza.

Here I would like to offer my response back. Perhaps I am missing something, but I do not see how incorporating the power and desire to create a finely-tuned universe into the hypothesis will help.

Let FT be the hypothesis that the universe, with its fundamental physical constants, was created by a god (a supernatural person) with the power and desire to create a universe of that specific nature.

Let VC be the values of the fundamental physical constants.

Let K be background knowledge.

Defenders of the FTA claim that VC is evidence for FT. In that case:

p(VC/FT & K) × p(FT/K)

p(FT/VC & K) = ——————————————————————

p(VC/FT & K) × p(FT/K) + p(VC/~FT & K ) × p(~FT/K)

Now we might raise a question about  p(VC/~FT & K ). We could argue that we really do not know what this value is yet, basic physics being not yet complete, or we could argue that physical processes, such as a selection process among multiple universes could explain fine-tuning (Prof. Waller considers such hypotheses in his book).

It seems to me, though, that the real problem is with p(FT/K). What is the background knowledge here? If FT is an ultimate posit, that is, is posited as an ultimate brute fact lying at the end of every explanatory chain, then, in that case, K can only contain tautological information, i.e. necessary truths. Necessarily—by definition—an ultimate brute fact is logically contingent. It is the ultimate reality—as a matter of fact—but it does not have to be. As a contingent fact, what other possibilities might have been instead? Normally, we would limit those other possibilities by appeal to background knowledge, but here we have extremely exiguous background information, limited strictly to necessary truths. It appears, then, that any possible state of affairs might have been the ultimate, original reality. Borrowing from my ex-professor John Earman’s example about what might emerge from naked singularities, the original existent might have been a single argyle sock, a patch of brown liquid, or a TV set playing Richard Nixon’s “Checkers” speech on endless loop. We are presuming that we have NO information to rule out any such possibilities.

So, if FT is an ultimate, brute posit, and if K contains only necessary truths, how do we assess p(FT/K)? We have two choices here: (1) We say that FT, like all ultimate posits, has no objective probability. This is the option I support. However, the FTA is all about assigning probabilities to ultimate posits. That is the whole motivation for the argument. If ultimate posits do have objective probabilities then, option (2) is to assess the probability of FT given only K. Prima facie that probability would be zero since there are infinitely many other possible ultimates (note that a probability of zero does not mean impossibility when the sample space is infinite). Hence we do seem to face the problem of fine-tuning with respect to FT as for any posited ultimate. Why this with just these properties out of the infinitely many alternatives? To posit a super-fine-tuner would lead to an infinite regress. So, the need for a finely-tuned universe is replaced by the equally urgent need for a finely-tuned god.

Here many philosophers would appeal to alleged metaphysical necessities to make some ultimate posits more probable than others. Thus, Richard Swinburne appeals to simplicity to argue that the God of theism is the most probable ultimate posit. However, post-Kant we know that metaphysical assumptions are slippery things. If theists are free to invoke the assumptions that support their side, then non-theists have that same freedom. Thus, Paul Draper uses the criterion of “modesty” to argue that naturalism is a more plausible ultimate posit. If, in the end, the FTA comes down to whose metaphysical assumptions we accept, non-theists can confidently expect that theirs will be as good as any—by God!

Many theists would follow Prof. Waller in regarding God as existing in all possible worlds. As it so happens, this is an issue I addressed in my last post before this one, so I will just refer back to that.