My Doubts about the Fine-Tuning Argument

Bentham's Bulldog, hopefully you can take care of these problems.

A (probably AI-generated) image of a pretty-looking cosmic structure.

If you haven’t read it, go check out Matthew Adelstein’s article, ‘The fine-tuning argument simply works’. It’s an excellent reply to common objections to the fine-tuning argument (FTA), the claim that various features of the laws of physics are evidence for God’s existence (roughly). But it made me reflect on my outstanding issues with the FTA. That thesis has been my research focus the last few years, and I think that it may succeed. But I have some doubts. So, here they are.

#1 We don’t have a final theory of physics.

General relativity and quantum mechanics, the two pillars of modern physics, are inconsistent with one another. So, unless reality contains a contradiction, the final ‘complete’ theory of physics that resolves the inconsistency between these two paradigms - and thereby represents the physical world at all scales - is beyond our grasp. There are quantum theories of gravity on the table, like string theory or loop quantum gravity, that attempt unification, but none has emerged as victorious. So, we do not know what physics will look like when a unified theory comes into view. Thus, even if we conclude that various constants within current physics can vary across a range of values, that fact is no guarantee that a final physics will permit variable constant values. For all we know, a complete theory will allow us to derive the values of the constants and initial conditions from more fundamental principles. If this outcome is a live possibility, then we cannot infer that because current physics is fine-tuned, a complete physics will also be fine-tuned. Another way of making the point is that while our present theories are fine-tuned, physical reality itself may not be.

Matthew replies by asking us to “[i]magine that a person gets 10 royal flushes in a row in poker. You accuse them of cheating. ‘I’m not cheating,’ they cry, ‘there is more fundamental physics that we don’t know about that explains why I got 10 royal flushes.’ Clearly, they have gone wrong.” This is a good response, but it does not entirely resolve my misgivings. Our prior experience grounds the fact that getting ten royal flushes indicates cheating. We’ve seen this outcome loads of times, and we always infer (correctly) that the person with these extravagant hands is cheating. Not only do we make this inference, but we often verify that our inference is correct (e.g., the guilty party makes a confession). That experience filters into our background knowledge when we encounter a specific instance of someone’s getting amazing poker hands again and again. (Part of) the reason we find appeal to physical necessity unreasonable is that we know these kinds of poker hands often result from cheating. We do not have anything analogous to this prior experience when we examine nature’s constants.

Or have I made a mistake here? Maybe the Matthewian¹ could reply that we have merely pushed the relevant appeal to future physics back a step. Maybe the move now is to suggest that the single instance under consideration - say, Miles getting five royal flushes in a row whilst playing poker on May 26th, 2024 - somehow features into a fundamental theory of physics such that this occurrence, and no others like it, is physically necessary. I take it that this suggestion has a very low epistemic probability, despite the fact that we don’t know what a final theory will look like. Could we then say the same about the suggestion that the values of the constants are physically necessary? That is not clear to me. The burden of proof is on the fine-tuning advocate to show that the values of the constants are analogous to five royal flushes, such that the epistemic improbability of ascribing the latter to physical necessity carries over to the former. It is not the burden of the skeptic to prove that they are not analogous. Indeed, take the suggestion that the speed of light c - a constant in current physics with no known deeper explanation - will turn out to be derivable from more fundamental physics. That suggestion seems far more reasonable to me than the corresponding suggestion for a particular poker hand. If so, then the more general claim that the constants might be physically necessary is just not akin to suggesting that various poker hands might follow from future physics.

#2 A vague epistemological worry that is hard to properly formulate.

Be patient with me here because imma² give it my best shot. Here it is:

Why think that we are in the right epistemic position to evaluate hypotheses concerning the fine-tuning of the universe for life?

The most common version of the FTA - such as the one appearing in Barnes (2019) - attempts only to show that fine-tuning is evidence for theism over naturalism. That is, we narrow our focus to two hypotheses and ask which one the evidence favors. But that conclusion is enormously modest, because we know that there are many other possible explanations for fine-tuning aside from theism and naturalism. It is a massive simplification to focus only on those two explanatory options.³ But my present problem is even deeper: even if these were the only two explanations we could think of, why think that they are the only two explanations that there are?

This issue may be nothing more than the more general problem of unconceived alternatives. That is, in any inference to the best explanation, we only show that a particular hypothesis is the best that we could think of, not that it is the best period, among both known and unknown explanations. Insofar as we operate and reason fine in other areas of life, the problem of unconceived alternatives notwithstanding, maybe we should not worry about it in this context, either. It’s kind of like Hume’s problem of induction (e.g., how do we know that the future will resemble the past?); we may not have a good answer to it, but we usually assume induction is reliable regardless.⁴

But inferences about the universe as a whole or to realties beyond the universe, strike me as particularly acute instances of human beings operating well outside their depth. Though the point is difficult to specify in a philosophically sophisticated way, my worry is that when we as finite agents turn to reasoning about what realities may or may not even possibly explain the universe and that exist beyond it, there may be numerous possible realities beyond our comprehension.

This reservation is exactly the one that many people point to when evaluating William Lane Craig’s argument for thinking that if there is a cause of the universe, that cause must be personal. Craig will argue as follows:

1. The cause of the universe is either an unembodied mind or an abstract object.

2. It cannot be an abstract object.

3. Therefore, the cause of the universe is an unembodied mind.

Let’s grant premise (2). Why does Craig think (1) holds? Because we cannot conceive of something that is immaterial and timeless, aside from minds and abstract objects. Now, even if this point is true, why think that if there were possible realities in addition to minds and abstract objects, we would know what they are? This worry reflects my deeper inclinations towards agnosticism generally.

It may be that my epistemic worries are ill-founded; but the first step in recovering is to admit you have a problem.

---

#3 Unresolved problems of divine psychology.

How do we know that the likelihood of a fine-tuned universe given theism is higher than that likelihood given naturalism? With Barnes (2019), let us grant that the probability of fine-tuning (F) given naturalism (N) is 1 part in 10^136. Why think that the probability of fine-tuning given theism (T) is any higher? If it is not, then fine-tuning has no evidential significance whatsoever for adjudicating between naturalism and theism. The typical response - one I myself have developed in ‘the literature’ (Donahue, 2024) - is to claim that moral agents are objectively valuable, and therefore God, a morally perfect being, would not be all that unlikely to create a universe where moral agents could arise.

This argument is right so far as it goes - to demonstrating that God would not be terribly unlikely to create finite agents. But how do we argue for the further claim that God would create such agents within a fine-tuned, physical universe? After all, God can create non-physical life: unembodied souls, demigods, angels, or any manner of Cartesian minds free-floating without any physical substratum. Thus, even if we can show that God would likely create agents, we have not thereby shown that God would likely create physical agents in a fine-tuned universe.

Okay, stay with me - we have to go one level deeper into the dialectic to get at the issue that plagues me. In response to these concerns, I have suggested that we ought to include certain facts in our background information - take certain things as given - when we then turn to evaluate fine-tuning. So, I say that we should take as given (1) our laws of nature, and (2) our psycho-physical laws (those laws dictating which physical states can give rise to mental states). The proposal is that we ask the following question: given that God wants to create a universe with our laws of nature and our psychophysical laws, what kind of universe ought we to expect? In answer to that question, I think we can then say that life’s objective value makes it not terribly unlikely that God would create a fine-tuned universe, given that He want to create a universe from that specified subset of possible realities.

Now, this move is only licit if the laws of nature and psychophysical laws are equally likely or unlikely given naturalism (otherwise, we would be guilty of gerrymandering our evidence). I initially thought this condition held; I saw no reason to think that naturalism made it any more likely that our particular laws of nature would obtain, or that our psycho-physical laws would obtain (naturalism, broadly construed, is at least compatible with the claim that electrons can have fully developed lives, just as much as theism). Nevertheless, after my recent dialogue with James Fodor, I’m questioning that assumption. On theism, God exists as an unembodied mind. Thus, given theism, the range of possible psychophysical laws is wider than the range given naturalism (i.e., ‘a mind does not require any physical substratum to obtain’ is a possible psycho-physical law if theism is true, but not if naturalism is true). Thus, we cannot take psychophysical laws as background information in our efforts to show that fine-tuning has a higher likelihood given theism than naturalism.

There might be a way around this issue by stating psychophysical laws in a more qualified manner, but I don’t know how exactly to carry that task out.

---

There you go, there are my three worries. I pass the mic to you, Matthew.

---

Barnes, L. (2019). A reasonable little question: A formulation of the fine-tunign argument. Ergo: An Open Access Journal of Philosophy. 1220-1257.

Donahue, M. (2024). Divine psychology and cosmic fine-tuning. Religious Studies. 1-17.

1. “Matthewian” refers to all those that subscribe to Matthew Adelstein’s views on the issue at hand. We’ll workshop a better designation in the future.

2. “Imma” is a colloquial or slang contraction of “I am going to”, permitted in the context of blog posts but not in academic writing (I’ve been told). Most people are aware of this fact, but I find humor in over-explanation.

3. A reviewer for one of my articles soon to be published once told me that I heavily overuse italics in my writing, and that it comes across like I’m trying to bully the reader into agreeing with my argument. I pretty much agree and have tried to cut down on italicized words and phrases. But sometimes I can’t help myself.

4. This is only the roughest and readiest statements of the problem of induction. Humean scholars, don’t come for me.