Here is a diagram to explain two different types of fine tuning; narrow and robust (if the image is too small to read, click on it for a larger version):
If narrow fine tuning is the kind of fine tuning there actually is, theists claim, then the probability of life coming about naturally is low; and this motivates the fine tuning argument for theism. This seems to carry at least some argumentative force. But this tweet from Justin Schieber (#FF) got me to thinking:
To say that such-and-such combination of cosmological constants doesn’t permit life is not to say anything about what’s logically necessary, but about what’s physically necessary (or so it seems – more on this in a moment). In other words, if the theist is right and God exists, fine tuning would be robust – and yet, the theist will argue that fine tuning is narrow as evidence for God.
But the theist cannot have it both ways. If we’d expect to see robust fine tuning were God to exist, then narrow fine tuning is actually evidence against God. From this, we can construct the following argument:
1. If God is omnipotent, then he can create life under any cosmological conditions.
2. If God can create life under any cosmological conditions, then fine tuning is robust.
3. Fine tuning is not robust.
4. Therefore, God cannot create life under any cosmological conditions. (2,3 modus tollens)
5. Therefore, God is not omnipotent. (1,4 modus tollens)
Premise 1 is just a consequence of omnipotence – it’s logically possible that God create life under any cosmological conditions – that some conditions are impermissible to life is a physical claim, not a logical one. Premise 2 is definitional, and premise 3 is the claim that motivates fine tuning arguments.
The theist clearly wants to keep 3 – if he does not, then he loses the fine tuning argument. He likewise wants to keep 1, as denying it would be to deny God’s omnipotence. So the theist must attack premise 2. How could this be done?
One response is to argue that the argument equivocates between different senses of robustness. When we talk of narrow fine tuning, the theist can say, what we mean is that a small set of values is physically necessary for life; but it’s still true that these values aren’t logically necessary. Thus the argument should look like this:
1'. If God is omnipotent, then he can create life under any cosmological conditions.
2'. If God can create life under any cosmological conditions, then fine tuning is logically robust.
3'. Fine tuning is not logically robust.
4'. Therefore, God cannot create life under any cosmological conditions. (2,3 modus tollens)
5'. Therefore, God is not omnipotent. (1,4 modus tollens)
The theist can then easily say that premise 3′ is false – fine tuning is logically robust, and it just so happens that God decided that fine tuning should be physically narrow. Furthermore, the theist might say, God wanted fine tuning to be physically narrow in order for it to act as evidence that He exists; a sort of divine signature.
Initially, I found this objection convincing. But after some reflection, this is no longer the case. Consider what things would look like if we admitted the logically/physically necessary distinction:
The problem here is that making a distinction between logical and physical robustness brings in modal considerations. This allows a premise like this:
6. There is a possible world W1 in which all values of cosmological constants are logically life-permitting, and yet life can never arise in W1 because it is physically impossible.
This highlights a difficulty for the theist. Given that, regardless of what the constants are when this world’s universe forms, life is logically possible; what does it even mean to say that life is physically impossible in W1? When discussing our world, the point is often made that if the constants were different, then life would be physically impossible.
The problem for the theist is that this sort of reasoning implicitly invokes the idea of possible worlds – the only reason that life would be physically impossible if the constants were different is because of how the actual world works. This other possible world doesn’t have to play by the rules of the actual world, precisely because the constants are different. When we examine which values of the constants could permit life, we’re peering into modal space, and to do so is to ponder what’s logically possible. Consider also this premise:
7. There is a possible world W2 in which the constants are the same as the actual world's constants, but W2 is physically not life-permitting.
This doesn’t seem like a possible world at all; whether a world is life-permitting or not just is a function of the constants. So it seems that physical life-permittingness reduces to logical life-permittingness. Thus, the original argument stands. What options does the theist have now?
Well, the theist isn’t going to want to accept the argument’s conclusion. The only remaining option is to deny premise 3, and say that fine tuning is robust. But if this is the case, then the theist’s fine tuning argument for God fails. Thus, the theist’s confidence in God’s existence should drop accordingly.