Robert Oerter has offered an argument I recently ran across, according to which the popularly cited Fine-Tuning argument for the existence of God can be construed as an even stronger argument for Naturalism (he offers it in jest, though, only to demonstrate that the Fine-Tuning argument is not a good argument for Theism). Here are some excerpts:
given a naturalistic hypothesis (N) and general background knowledge (K), the probability of a fine-tuned universe is small:
P(FTU|N&K) < 1
On the other hand, given the theistic hypothesis (T), we would expect the universe to be suitable for life: P(FTU|T&K) is not small, or at least not as small as P(FTU|N&K).
But that is not true if God exists! Indeed, under theism, there is no reason to expect that the universe will be fine-tuned. Remember that God is, by hypothesis, omnipotent. That means that God could have caused life to arise by miraculous means, even in a universe that was not fine-tuned.
I want to point out that the probability envisioned in the fine tuning argument is a sort of prior probability that ignores some of our background information: namely, the fact that life actually exists. That is, we have to take (K) to mean “general background knowledge not including the knowledge that life exists.” But we actually do know that life exists (L), and it is perfectly legitimate to include this knowledge along with our other background knowledge.
Conclusion: given that we know that life exists, the probability of discovering we are living in a universe with parameters fine-tuned for life is much higher under the naturalistic hypothesis than under the theistic hypothesis.
This is an interesting argument, about which I’d like to offer some thoughts. First, can we really include the knowledge that we exist in our background knowledge without just misunderstanding the Fine-Tuning argument altogether? I think the answer is that we cannot. The background knowledge is supposed to be nothing other than our best working knowledge of physics – that is why this argument takes science as an authoritative source. It says ‘given our knowledge of physics, ought we have expected the Universe to look like it does?‘ The champion of this argument will argue that on Naturalism we have no reason to expect it to look this way, whereas on Theism we do have good reason to expect it to look this way.
Notice that Oerter counters by arguing that on Theism, given the background knowledge that life exists (along with our knowledge of physics), we do not have good reason to expect to find that our universe is fine-tuned to be life permitting. I’m not sure this is true (for instance, we might have good theological reasons for thinking that God would prefer, or strongly prefer, a universe in which life arises naturally, to one in which life ‘doesn’t belong’), but I think we can at least admit that life in a non-life-permitting universe is more likely on Theism than on Naturalism (on Naturalism it is impossible). That makes it trivially true that, given that our background knowledge includes the existence of life, Theism doesn’t make Fine-Tuning as likely as Naturalism does. Naturalism would be, all things being equal, negligibly more likely than Theism given Fine-Tuning.
However, the Fine-Tuning argument is not usually given in the form: P(FTU|T&K) > P(FTU|N&K). Instead, I think the more forceful and popular form of the argument is that, first, given our background knowledge of physics, our universe being Fine-Tuned can only be explained by reference to either chance, physical necessity, or design. However, since it so strains credulity to explain it either by chance or physical necessity it must be explained by design (which, it is suggested, does not strain credulity).
It seems like a fairly obvious point, yet I haven’t seen it mentioned much in the fine tuning discussions I’ve read. I’d be grateful if anyone can show me the glaring flaw in the argument that I’m missing.
I think the two difficulties I have identified are 1) that the existence of life is just not part of the background knowledge, because the background knowledge refers exclusively to the deliverances of our best physics, and 2) that the fine tuning argument, in any case, is not typically given in the form Oerter’s argument seems to think itself a parody of.
By way of offering a few final thoughts: I think perhaps we can make a parody of Oerter’s argument from Fine-Tuning for Naturalism, such that we can argue from the existence of life to the falseness of Naturalism. Imagine, by contrast to reality, that in another logically possible world some scientific model exists as the preferred model, according to which a broad range of universes are life-permitting (say the vast majority). That wouldn’t make Theism or Naturalism more or less likely, but at least then Naturalism wouldn’t seem so terribly unlikely. Moreover, suppose that on Theism life were extremely likely on any Universe, since God doesn’t need, nor does he have any reasons for preferring, a fine-tuned universe (but does, presumably, have good reason to prefer the existence of life); the theist would expect to find life, but the Naturalist cannot expect there to be life given our background knowledge (here meaning the deliverances of modern physics). Thus the Naturalist is stuck with the fact that given our background knowledge the probability of life existing on Naturalism is infinitesimally low, whereas the probability of life existing on Theism is, comparably, incomparably greater. Thus, we can articulate this insight as an argument from life to the falseness of Naturalism, given our knowledge of physics.
P(EL|T&K) > P(EL|N&K)
Perhaps Oerter, in his own defence, would respond that the background knowledge in the argument cannot be the scientific information alone, since then it might be thought include Fine-Tuning itself, but what the champion of the argument will respond is that the background knowledge includes the scientific information other-than the actual fine-tuning (eg. it includes the range of physically possible values, meaning constants and quantities, that our universe could have had, but does not include the actual values, etc.).