I have lately tried to rethink my position on Gods existence and tried to be more critical of the atheistic position. This has partly been spurned by my impression objections to theistic arguments often rely on the same errors atheist accuse theists of, for instance using the shotgun-method (I.e. picking 3 faults with all premises in the arguments) or building conclusion on a web of arguments and, when considering a particular argument, supporting it by other parts of the web or simply not addressing the argument very well.
As part of this I came across Collins 1999 article on fine-tuning. I would recommend the article to others since it does (in my opinion) make a very honest case for fine-tuning, it is quite easy to read and offer relevant criticism for many atheistic objections. Here I want to briefly review some points of the article, explain why I think some of the atheistic objections for the fine-tuning argument are weak, and try to offer my reasons for not accepting the conclusion. Finally I want to offer a refutation of the fine-tuning argument.
The article is available here: http://rintintin.colorado.edu/~vancecd/phil1000/Collins.pdf
The strongest point in the article is in my opinion that it makes its method very clear. Collins explain:
The prime principle of confirmation is a general principle of reasoning which tells us when some observation counts as evidence in favor of one hypothesis over another. Simply put, the principle says that whenever we are considering two competing hypotheses, an observation counts as evidence in favor of the hypothesis under which the observation has the highest probability (or is the least improbable). (Or, put slightly differently, the principle says that whenever we are considering two competing hypotheses, H1 and H2, an observation, O, counts as evidence in favor of H1 over H2 if O is more probable under H1 than it is under H2.) Moreover, the degree to which the evidence counts in favor of one hypothesis over another is proportional to the degree to which the observation is more probable under the one hypothesis than the other.
This is in my opinion a very clear description of ordinary scientific reasoning over evidence and I think any criticism of the fine-tuning argument which does not rely on reasoning along those lines is suspect. Do keep in mind that since it is often easy to come up with a wealth of theories (and hard to come by evidence) scientists are often more conservative, more about that in the conclusion.
Collins then lay out the fine-tuning argument:
For example, the fine-tuning is much, much more probable under the theism than under the atheistic single-universe hypothesis, so it counts as strong evidence for theism over this atheistic hypothesis
Collins then go about illustrating the principle of confirmation and argue (1) universe if fine-tuned (2) God would make such a universe.
Collins distinguish between a hypothesis in which there is only one universe and one where there is many. He then first address atheistic arguments against the fine-tuning argument under the single-universe hypothesis, and I largely agree with him in his criticism. Below is my loose list of arguments I don’t consider to be very good, even though I might have been guilty of using them myself.
- Universe is full of natural evils such as viruses: This is really the problem of evil; sure, we might reject god on this ground, but it’s a bit of a different argument. See the problem of arguing from a web of arguments.
- Arguing the universe is not really fine-tuned: Sure this might be the case, I don’t think it’s particularly reasonable; for instance most string theories does not allow formation of galaxies.
- Arguing God could “just as well” have made a universe full of rocks: Again, this might be possible, but if we were to put ourselves in Gods shoes I think it’s fairly reasonable to assume God wouldn’t be that interested in a cosmic rock collection on theological grounds.
- Who fine-tuned God: This is an interesting question, but I think Collins relying on his principle of confirmation gets around the objection
- Finally there is the strong antrophic principle. According to this we should reason conditionally on there being at least one member of ones reference class of observers. This has sometimes been applied to the problem of fine-tuning to the effect of “there is no problem of fine tuning, because since we are here to observe the universe, the universe must have the laws, values we observe”. The problem is I haven’t seen a good reason to accept the strong antrophic principle and I think there are very good reasons to think it do not work.
One thing which I do think should be raised because it affects Collins positive case is according to his principle of confirmation is (1) do Collins agree the expansion of the universe and the second law of thermodynamics point towards the universe eventually being devoid of life and (2) if so, is such a universe particular probable under the type of God Collins consider? I am going to assume Collins come up with a theistic explanation, either that sin made the universe expand, that God will fix any problems, or that it’s a part of Gods plan to destroy the earth.
Secondly Collins turn to the many-world hypothesis and reject it for 3 reasons I am going to address later. First let me try to recap:
- I agree with Collins principle of reason
- I agree the universe is fine-tuned for life
- I (suppose) Collins would agree the laws of physics appear to be such the universe will eventually be devoid of life:
Now consider the set of alternative hypothesis
- There is only a single universe. (H1)
- God fine-tuned the universe (God)
- The (natural) mechanism which made the universe work in such a way to produce many universes with different values of all variable parameters (H3)
Quite clearly, the observations has very high probability under both H3 and God, and low under H1, and so by Collins principle of confirmation we should favor God and H3 (God & multiverse) over H1 (single-universe) without being able to tell God and H3 apart.
I really think this should be the end of Collins article, that this is about as far as we are going to go purely on the principle of confirmation and the general program Collins laid out in the beginning. But realizing the bus do not go all the way to the destination, Collins opt for the taxi which, aside his last reason, involve finding faults with the multiverse theory. The problem is this is strictly speaking no longer operating within the the principle of confirmation and need to make sure Collins ideas do not contain the same flaws.
Collins first reasons for rejecting H3:
The first reason for rejecting the atheistic many-universes hypothesis, and preferring the theistic hypothesis, is the following general rule: everything else being equal, we should prefer hypotheses for which we have independent evidence or that are natural extrapolations from what we already know.
The first rule amount to saying we should take all evidence into consideration. Certainly, if we are convinced (perhaps by the moral argument) God must exist and love us, then I would agree he properly fine-tuned the universe as well, or on the other hand, if Collins was convinced by the argument of evil that God do not exist, he would properly not propose him as a likely explanation for fine-tuning. But this basically mean re-introducing the entire “web” of arguments and one wonder why Collins would have written on the fine-tuning argument alone in the first place.
Collins argument for the second rule:
we already know that minds often produce finetuned devices, such as Swiss watches. Postulating God—a supermind—as the explanation of the fine-tuning, therefore, is a natural extrapolation from of what we already observe minds to do. In contrast, it is difficult to see how the atheistic many-universes hypothesis could be considered a natural extrapolation from what we observe
Collins misses the crucial point that the many-universe theories are natural extrapolation of various theories for our universe, for instance inflationary theories. But again, these points pertain to our initial judgement of how probable God and H3 are, and so Collins need to argue a designer is inherently more probable than a multiverse.
Collins first reasons for rejecting H3:
A second reason for rejecting the atheistic many-universe hypothesis is that the "many-universes generator" seems like it would need to be designed. … . It stands to reason, therefore, that if these laws were slightly different the generator probably would not be able to produce any universes that could sustain life.
This amount to applying the design-argument a second time to whatever principle underly the a multi-verse theory. Firstly, this again shift the argument outside its original scope, secondly, if this is a “fair-game” question we might as well ask: “well who fine-tuned god to be able to think, since most changes to the human mind are detrimentary presumably the same hold for God”, but I doubt Collins would consider that very legitimate. Thirdly, I think fine-tuning is being used with the opposite sense. Suppose we have a theory for multiple universes that is maximally non-fine-tuned, wouldn’t that correspond to generating the broader range of universes, because otherwise we could ask: “who fine-tuned it to only generate THAT type of universes”?
Collins third reasons for rejecting H3:
…the atheistic many-universes hypothesis (nor the atheistic singleuniverse hypothesis) can at present adequately account for the improbable initial arrangement of matter in the universe required by the second law of thermodynamics.
This is properly the strongest argument and, as opposed to the two previous, it ties into Collins principle of confirmation by claiming that even if a multi-universe theory was true then it would be unable to account for the low-entropy initial state.
This is an important aspect of big-bang cosmology and a point which is often misunderstood by atheists as theists simply re-asking the fine-tuning argument. This is wrong in my opinion. According to most models, the early universe was in an almost uniform state of energy; this is often associated with maximal states of entropy (think of gas in a room), but since gravity is attractive, the maximal entropic states would intuitively correspond to black holes hence the problem: Even if we assume multi-universe theories, we would expect an early universe full of black-holes unless someone had “tuned” the early state of the universe to correspond to low states of entropy. As such, the argument is legit since it (unlike the other objections) tie directly into Collins principle of confirmation.
While the entropic state of the early universe is considered a very important open problem of big-bang cosmology, it is important to keep in mind it exist in an eco-system of “big problems of physics” which are unlikely to be resolved without a full account of gravity at those scales. There are also suggestions for how it could be resolved, in other words, theories where the entropic state of the early universe was actually maximal, see for instance: http://arxiv.org/abs/1212.1087. Lets denote such a theory for “Low-Entropy Solution” (LES) such that “H3+LES” is the hypothesis of a multiverse and a solution to the low-entropy problem. We can now summarize an answer to the design argument:
Conclusion:
- Assuming Collins principle of confirmation
- Assuming the parameters of the universe are indeed fine-tuned for life
If we stick to how science normally accept things, namely if novel predictions of a theory turns out to be true, then we should not accept the God-idea since it make no novel predictions beyond accounting for accepted facts (and it is hard to see it could in principle).
If we decide to be more optimistic about the principle of confirmation, then H3+LES would account for fine-tuning just as well of God.
Keep in mind H3+LES is preferable over God when we take into account the universe also appear to be fine-tuned for destruction.
This leave us with one thing, namely going outside the principle of confirmation and arguing H3+LES is intrinsically less likely than God. If we do this by referring to other arguments (moral argument, ontological, etc) then we have moved past the fine-tuning argument and should not have introduced it in the first place.
If we argue H3+LES is made implausible by the way they account for fine-tuning (perhaps because both make use of speculative ad-hoc physics, especially LES) then we should also criticize God for making use of speculative ad-hoc physics; namely no discernible physics at all. At any rate this would require operative characteristics for what make a theory initially more plausible than another and I simply don’t see how God could fare very well in that setting.
For those reasons I believe fine-tuning is not a very good argument for the existence of God.