Second in series 20 Atheist answers to questions they supposedly can’t.
2. What explains the fine tuning of the universe?
Short answer: There’s no rigorous way to define fine tuning that’s hospitable to a fine tuning argument. Even if there were, a multiverse ensemble would be about as good an explanation as Theism.[ref] Another move atheist could make is to ‘trump’ fine tuning with an argument from evil. For even if it is the case that a fine tuned universe alone would support theism, it might be that further knowledge (it is fine tuned but has lots of evil in) will support atheism, in a similar way that knowing the accused was in the house at the time of the murder is some evidence for his guilt, but finding out he was locked in the basement is evidence for his innocence. (cf. Draper’s work on this). [/ref]
Long answer: The fine tuning or teleological argument has had a long and august history, and it has a new lease of life due to modern scientific discovery. For a variety of parameters (the gravitational constant, the cosmological constant, and so on) any slight deviation from their actual values would, as far as we can work out, lead to a universe where intelligent life is impossible. So surely this is a cosmological watch for want of a watchmaker?
Not quite. It is not straightforward to actually make an argument along these lines. The universe (and its constants) are old evidence, and it is obvious that the constants are life-permitting because we are here to see them (the Weak Anthropic Principle). We also don’t know what (if any) universe generator there is, and how (if at all) these constants/parameters/whatever can be said to vary.
Perhaps the best way forward is to follow Robin Collins and give a Bayesian argument, subtracting out knowledge we exist. For it seems in the space of the various constants, life permitting regions appear incredibly scarce. Happening onto one ‘by chance’ is unlikely – by contrast, Theism predicts such a universe, and gets some hypothesis confirmation. However, things still aren’t so simple. You need to rely on a slightly-squiffy principle of indifference to give probabilities over this unbounded space and (as always) there are normalization worries.
There’s a bigger problem. In the same way there’s nothing to suggest the particular values of the universe necessarily have to be, there’s nothing to suggest the particular parameters or particular physics has to be: alien physics with completely different laws are conceivable (and its remotely possible they are the actual universe – best science can be mistaken). If so, we have to look over more than the ranges we know about, but try and work how common life permitting universes are over the entire space of conceivable universes. That’s inscrutable, so we cannot say it unlikely for a universe to end up life permitting ‘by chance’, and so the fine tuning argument is moribund.[ref]It is a bit like the ‘looking under the lamp-post for the car keys example. We can illuminate what we know to be a minute fraction of the field we want to evaluate, and find in this region life permitting universes are rare. That is still evidence for life permitting universes being rare, and if you put a gun to my head and told me to guess my best guess of how common life permitting universes are is just the proportion I see in the illuminated region (straight rule) but it is minute evidence and a really unreliable estimate, because we next to no confidence our sample will be representative, so surely the right answer is to say it is basically inscrutable.
Collins accepts this, but argues it is reasonable to restrict ourselves to the epistemically illuminated range and not worry about what lies outside. I don’t think this works.[/ref]
But say I’m wrong and we can rescue a fine tuning argument from these worries. Has Atheist any other cards to play? Yes, the multiverse ensemble. On a suitable multivariate-generating principle (say there’s a universe for each and every set of values for the parameters/constants/whatever), then that, combined with the weak anthropic principle, means seeing ourselves in a (locally rare) life permitting universe isn’t evidence against Atheism at all.[ref]There’s a common ‘poker player parody’ of this view, but it just illustrates misunderstanding. The parody goes: “Suppose you’re playing against a dodgy poker player, and whilst playing he draws twenty royal flushes in a row. He reassures you: ‘Naw, I’m not cheating, it’s just that out of all the possible universes, we just happen to inhabit the one that’s really lucky for me. Now pay up.’”. This is cute, but really confused. Given some basic credences (P(He’s cheating) >>> P(20 royal flushes)), it’s the case there are far more universes in which you exist where he’s cheating you than the ones where he honestly just drew 20 royal flushes. In the fine tuning case, you are never in a universe where the values are not life-permitting. If we changed the thought experiment so you would spontaneously combust if 20 royal flushes weren’t drawn, then we get the ‘can’t decide he’s cheating or not’ result.[/ref]
The complaints made against these things is either they’re ad hoc or unfalsifiable. Neither is necessarily so. There are other reasons to posit multiverses besides ‘getting out’ from fine tuning (cf. Many worlds in quantum mechanics), and folks have argued that there are possible experiments which could tell us whether they are true. In short, it doesn’t seem multiverses are really implausible, nor are they really implausible given Atheism, so they’re sufficient to fend off the confirmatory force of fine-tuning.
Further reading
Gregory Lewis 