I regard the Fine Tuning Argument (FTA) as one of the more cogent arguments for the existence of God. I still don’t find it convincing, but I take it seriously. Examples of fine tuning can be readily found online, but the idea in general terms is that modern physics has discovered that the values of the fundamental constants of the universe (i.e. the expansion rate of the big bang, the speed of light, the mass of the electron, etc.) have to fall within an exceedingly narrow range if the universe is to permit life at any point in its history. Physicists also point out that the chances that these constants would have just the value that they have is astronomically improbable. Proponents of the FTA infer that this is evidence that these constants have been fine tuned by an intelligence, namely God.
Now, one could raise several objections to the FTA. The most powerful in my estimation are the multiverse objection and the normalizability objection. It’s unclear, however, whether both objections can be employed simultaneously for reasons I shall touch on below. First, the multiverse objection is popular. I suspect it is the weaker of the two because it essentially concedes the point that the ‘fine tuning’ is improbable; it simply tries to mitigate its improbability by increasing the sample size. The alleged fine tuning, the multiverse objection goes, is improbable only if the number of actual universes is one. However, if there are an infinite, or near infinite, number of universes, the fact that our universe won the cosmic lottery is no longer so surprising. After all, some universe or other had to have life permitting characteristics.
Proponents of the FTA typically respond by saying that such an objection commits the ‘gambler’s fallacy.’ If you were playing poker, and one player was consistently dealt improbably good hands, you might suspect the game was rigged. What if the ‘lucky’ player responded by saying, “There are an infinite number of universes with an infinite number of poker games being played, so the fact that I’m dealt winning hands is not very improbable after all”? As Alvin Plantinga humorously asks ‘Would that argument fly in Dodge City or Tombstone?’ I think there is something to the gambler’s fallacy. If the multiverse hypothesis were employed in an ad hoc way to explain away apparent fine tuning, the FTA proponent’s objection would have some real force. However, it’s my understanding that physicists employ the multiverse hypothesis to explain phenomena quite independent from the fine tuning data. I’m not a physicist, but it seems as though M theory, which entails the multiverse hypothesis, is favored in physics today because it stands the best chance of reconciling quantum mechanics and general relativity. If this is the case, it’s not merely an ad hoc hypothesis invoked to rebut the FTA.
Another response to the multiverse objection asserts that there is no empirical evidence in its favor and, thus, it’s no less metaphysical than belief in God. Granted, the multiverse hypothesis currently rests on educated guesswork and mathematical models that may or may not match reality. But it’s more than mere speculation. Time will tell whether or not it remains a viable theory. FTA proponents also contend that a multiverse is less simple an explanation than a single God. The question of simplicity is a difficult one. Generally speaking, simplicity is a virtue of scientific hypotheses, but it’s difficult, at least for me, to determine which hypothesis has the upper hand in this case. Some theologians say that God is metaphysically simple; a mind with powers of agency, even infinite powers, is more simple than a bloated multiverse. Of course, non-theists have difficulty seeing an omnipotent, omniscient agent as anything but highly complicated. I can’t do justice to this debate here. I’ll simply note in passing that there may be different understandings of simplicity that are relevant to this debate. We could talk about the number of entities invoked in a given hypothesis. We might call this quantitative simplicity. We could also talk about the nature of the entities invoked. We might call this qualitative simplicity. It seems that theism has the advantage with respect to quantitative simplicity whereas multiverse theories have the advantage with respect to qualitative simplicity. To elaborate briefly, theism only postulates the existence of one more entity, God, whereas multiverse theories postulate an infinite number of universes. However, these universes are of the same physical nature as the universe with which we’re familiar. God, by contrast, is a non-physical, supernatural being, very different in kind from the universe. Which definition of simplicity is most relevant in deciding which hypothesis to favor is a tricky question.
However, a better objection to the FTA, in my judgment, is what’s called the normalizability objection (it really needs a sexier moniker). Ironically, this objection was raised by theistic philosophers Timothy and Lydia McGrew and Eric Vestrup. For a good, somewhat technical, summary of the argument and replies to it, click here. I’ll try to summarize it in as non-technical a way as possible. Essentially, the McGrews and Vestrup claim that the probabilities assumed by the FTA cannot be formally stated. In other words, probability calculations are only meaningful if the relevant probabilities add up to 1. For example, the odds that I’ll roll a number between 1 and 6 on a six-sided die are 1 in 1 (1/6 x 6 = 1). However, we can calculate the odds in this case because the range of outcomes is finite. In the case of the fundamental constants of the universe, there seem to be no logical limits on their values. The speed of light could have been anything. Since the range of possible values is infinite, there’s no way to make the range of probabilities add up to 1. In other words, we can’t meaningfully state that it’s highly improbable that the universe has just these values and not others. In an infinite probability space, the odds cannot be calculated because we can’t specify in advance the likelihood of one value over another. Incidentally, the reason it might be inconsistent to appeal to both the normalizability and the multiverse objection is that the latter also depends upon there being an infinite range of values; it just happens that these are all actually instantiated. Nonetheless, the problems attending the probability calculations remain.
There are various responses to this objection (see link). One objection says that it proves too much. That is to say, it would rule out even an obvious example of design. Plantinga imagines stars arranged to spell out a message that affirms God’s existence. Even most skeptics would recognize this to be an instance of divine design. However, the same problems that attend the FTA attend this hypothetical case. The distances between the stars would have to fall into a message permitting range to be visible and intelligible to us. But there’s an infinite number of values the distances between the stars could take, so we can’t make a probability judgment here either. Plantinga takes this to be a reductio of the normalizability objection. Robin Collins, a prominent defender of the FTA, also raises a counter-intuitive consequence that would follow if the objection were sound. As the range of possible values grows larger, approaching infinity, the improbability of these constants having exactly the values they do also grows. However, once the range of values reaches infinity, we must say that these values are no longer improbable. This seems counter-intuitive. However, infinity is a very counter-intuitive notion and our intuitions might not be a reliable guide in this case.
There’s a further objection to the FTA entailed by the normalizability objection known as the ‘coarse tuning argument.’ If we consider the range of cosmic constants to be infinite, we would have to conclude that even if the life permitting range were much larger than it in fact is, that the universe was still designed. After all, the odds against the particular life permitting range obtaining is infinite. However, this suggests that a so-called coarsely tuned universe would be just as improbable as a finely tuned universe. Consequently, a coarse tuning argument should be just as sound as a fine tuning argument. However, this certainly seems counter-intuitive. Surely, a coarsely tuned universe would not favor theism as decisively as a finely tuned universe. In fact, theists and non-theists seem to presuppose this point; it’s the shared assumption that gives the FTA force. Although a coarse tuning argument might not favor theism as decisively as a fine tuning argument, Plantinga thinks that a coarse tuning argument would still lend some support to theism. Thus, the coarse tuning argument does not work as a reductio of the FTA.
The FTA, despite the objections to it, is still probably the best version of the design argument. If one is a theist, there are interesting consequences to accepting or rejecting it. The McGrews are theists, but don’t accept the FTA. They do, however, hint at other types of design arguments that escape their objection. These would be appeals to design within the universe, as it were, for which we could specify the probabilities. Such examples would include planetary design arguments of the kind made by Gonzalez and Richards in Privileged Planet to the effect that Earth is finely tuned for life. Other examples would include biological design arguments such as those found in Meyer’s Signature in the Cell or Behe’s The Edge of Evolution. One could make such intelligent design (ID) arguments even if the normalizability objection to the FTA succeeds.
However, there may well be another consequence: if, unlike the McGrews and Vestrup, a theist already accepts the FTA, should he then a fortiori accept the planetary and biological design arguments because the probabilities in these cases are much easier to ascertain? I have commended Plantinga for his consistency in accepting ID arguments in addition to the FTA. I have also criticized as inconsistent theists who accept the FTA but reject ID arguments. However, recently I have begun to rethink the relation between the FTA and other versions of the argument from design. Although I think that the FTA is the most promising tack for theists to pursue, I no longer think consistency compels them to also endorse ID arguments. Some theists have pointed out to me that the opposite might be true. In other words, the FTA, if sound, should make us skeptical that there are any good ID arguments to be had. The reasoning goes something like this: if God set up the universe for intelligent life at the beginning, we should expect life to evolve without subsequent interventions. Indeed, such interventions would be beneath God. It would reduce him to a god of the gaps. I’m not entirely convinced by this move, because it seems like deism rather than theism. But it is not inconsistent to say that while the FTA is a good argument, ID arguments are not. In fact, if the FTA argument is cogent, we might very well expect ID arguments not to be. Thus, the success of the FTA might come at the expense of ID arguments. Rather than insist that theists endorse both, a better strategy might be to pit the arguments against each other and force the theist to choose between them.