Home » Library » Modern Library » Richard Carrier Finetuning

Richard Carrier Finetuning

Response to James Hannam’s ‘In Defense of the Fine Tuning Design Argument’ (2001)

Richard Carrier


In his essay “In Defense of the Fine Tuning Design Argument” (2001) published here on the Secular Web, I do not believe Hannam has addressed the full range of issues and problems with the Fine Tuning Argument as discussed in our essays against the Design Argument. As but one example, he dismisses Victor Stenger’s “Monkey God” program as little more than “a bit of fun,” but this is a serious research product, defended at length in a technical article, “Natural Explanations for the Anthropic Coincidences,” Philo, 3:2 (Fall-Winter 2000) [click here for Acrobat/PDF version]. Hannam does not interact with this work at all. He thus does not consider that Stenger’s model varies all constants dynamically, not just one while keeping the others the same, and thus Monkey God generates a far more inclusive information space from which to judge relative probabilities. Hannam also criticises it for only addressing four rather than six constants, but in fact only four constants are relevant for generating long-lived stars, whose existence makes the conditions for life highly probable, regardless of what the other constants turn out to be, as Stenger argues in the above-mentioned paper. There are certain unjustified assumptions in Stenger’s argument (e.g. see point 3 below), but they are the very same assumptions Hannam and other creationists build upon.

However, Hannam correctly dismisses a great many theistic fallacies that our other essays have also refuted in more detail, and thus his position is more sound than most, and worthy of careful attention. But rather than pick on more specific examples like the Stenger case above, this essay will focus on some of the general, sweeping problems that Hannam’s essay does not resolve.

(1) An Insufficient Case

I do not believe Hannam has correctly framed the debate. I agree with him that the datum in need of explanation is the fact that “the laws of physics are such that life can appear.” The problem is in deciding whether theistic hypotheses are either necessary, or substantially superior as explanations. It is not enough for a theistic hypothesis to be merely “better,” for that would leave nontheistic hypotheses as sufficient, and there may be other prior reasons to disbelieve in god or to believe in metaphysical naturalism that make the nontheistic hypotheses more believable.

Thus, the theistic hypothesis must be necessary or substantially superior.[1] For any theistic hypothesis to be necessary, all nontheistic hypotheses must be inconceivable, i.e. logically impossible, or falsified by contrary evidence. For any theistic hypothesis to be substantially superior, all nontheistic hypotheses must be very improbable, or the theistic hypothesis must have significantly more direct evidential support than all nontheistic hypotheses.

Framed this way, we see that Hannam has not made an adequate case for theism from the proposal of ‘fine tuning’, since we have no direct evidential support for any theistic hypothesis (all we have is the datum in need of explanation) and no evidence falsifying the nontheistic alternatives, which are not illogical, and whose improbability cannot be observed or known. Even if we reject the claim (made by many an expert) that we have direct evidential support for some nontheistic hypotheses, agnosticism (i.e. the conclusion that we do not know who is right) is still the only justified position, and that means Fine Tuning cannot support any argument for the existence of a Creator without commiting the fallacy of argumentum ad ignorantiam.

(2) Confusing Hypotheses with Evidence

Hannam rejects the objection “that there is only one possible set of physical laws” on the grounds that we can imagine others, but this is irrelevant. The fact that we can imagine other things does not establish that they are all physically possible. The atheist who posits that the existing laws of physics are the only possible laws that can exist is stating a hypothesis that explains why we have the laws we do (and thus, a fortiori, why life exists). We cannot reject a plausible hypothesis simply because we can imagine others. This confusion is probably not Hannam’s fault, however, since many an atheist has treated the proposition “there is only one possible set of physical laws” as contrary evidence when in fact it is a contrary hypothesis. The fact that such a hypothesis begs the question is not fatal, since “God just exists” or “God just is the way he is” also begs the question, bringing both hypotheses to the very same status: neither is proven over the other, which is why this is a valid objection to a Fine Tuning argument for God.

(3) Where are the Independent Variables?

Hannam completely misses a central problem with any Fine Tuning argument: we are not able to identify the independent variables in the equation. Hannam says “it does appear that small changes in constants result in rapid breakdown of many of the chemical and physical processes,” but Hannam has failed to show that it is even possible for one constant to change while the others remain the same. For altering one constant may irrevocably alter another, greatly modifying any conclusion we can draw from modeling possible universes. Thus, just as Hannam warns against arguments based on the improbability of life forming naturally on the grounds that science might discover the natural means of forming life, so should he be warned against using a Fine Tuning argument based on the assumption of independent constants when science may soon discover, for example, a Grand Unified Theory or a Theory of Everything that shows how all the constants are causally related to each other.

That this may be the case is made probable by two facts. First, it is implausible to suppose that the constants are just arbitrary parameters that can be adjusted on some cosmic control panel. This is not impossible, but it seems more likely that they are determined by some prior fact or event, as all things seem to be (and the First Cause argument for God presupposes that this must be so). This alone would be a mere possibility, until we notice the second fact: whereas in the 19th century there were some twenty to forty “physical constants,” there are now only around six. All the others have over the intervening century been proven to be causally determined by more fundamental factors. For example, the boiling point of water was once considered a physical constant, but is now known to be the result of quantum mechanical laws, and thus could not be any different than it is without also changing the laws of quantum mechanics. Since the trend has been steadily in this direction, it is reasonable to predict that all the constants will end up being explained in this fashion. For example, since Planck’s constant defines the smallest possible unit of space and time, it may be the case that the speed of light is inexorably tied to Planck’s constant, so that one cannot be changed without altering the other.

Therefore, we do not know which constants are independent of the others, or of other more fundamental features of the universe, and therefore we cannot assume the constants can be different without changing the others. So models based on such an assumption are useless: we do not, after all, know the ratio of possible universes that produce complex chemistry to possible universes that do not. Therefore, we cannot argue that this ratio is small. Of course, we cannot argue it is large, but that does not matter: without any basis for presuming it is small, a Fine Tuning argument has no basis for producing “God Exists” as a conclusion. Related to this is the equally-fatal problem that we do not know what range of values for the constants are possible, whether that range is a span in which every value is equally likely, or whether it fits a bell-curve distribution of relative probability, or something else, like a chi-square distribution, and we have no clue, consequently, where on any of these probability curves fall the actual parameters we observe to exist. In other words, to argue that the existing parameters are improbable requires adopting a huge array of blind assumptions, and since none of these assumptions can be established or justified in any way at all, neither can the conclusion derived from them.

(4) Anthropocentrism and Hannam’s ‘Specific Fallacy’

Hannam warns against the fallacy of specificity (which he calls the ‘specific fallacy’), yet commits this fallacy anyway when he concludes that the unlikelihood of our universe being the one to arise naturally entails that we are entitled to believe it did not. He uses the analogy of a non-mutant sperm arising from a mutated sperm bank, which would be so peculiar we would seek explanations for it apart from chance, but he is not correct about this: chance would remain a viable explanation so long as it was known that some of the sperm in the bank were not mutated. So it is with the universe: even if the ‘bank’ of all possible universes were such that only a minuscule proportion could produce life (and we do not even know this much), it is necessarily the case that one of those universes may be the one chosen by whatever random process of selection exists. Thus, the fact that one pattern was chosen out of a zillion patterns tells us nothing about whether it was selected randomly or not, for we already know that one of those patterns was going to be selected, and that it could be any one of them.[2]

To argue that “we were incredibly lucky it was one that would produce us” is a trivial observation, for it could be said of any universe that it was equally lucky that it was the universe that came out of the pool of possibilities. The only difference is that humans are the lucky outcome in this case, and to argue that this makes this universe special over all others is not to argue a fact independent of human thought, but to state a value judgement, a judgement that naturally all humans would make about themselves. This may be a true value judgement, but it is a statement of value nonetheless, which exists only after the fact, and thus cannot relate to what caused our universe. Without positive evidence for a specific cause, as against a random one (assuming a random cause is possible in the first place), there is no basis for concluding the cause of this universe was not random. Again, one is left with agnosticism: we simply do not know.[3]

[Fortunately, I think Hannam has avoided the Anthropic Fallacy, though several other Fine Tuning proponents have fallen victim to this, cf. My Initial Response to Warrick Walker, and my review of Beck’s Argument for God.]

(5) Multiple Universes

Hannam dismisses the “multiple universes” hypothesis because we have no evidence for other universes. This is of course the pot calling the kettle black: the evidence for a creator (or even the Christian’s ‘other universes’, i.e. heaven and hell) is hardly better. But there are certain merits to “multiple universes” that many forms of theism lack. The “multiple universes” hypothesis itself entails that we could not have direct evidence of its truth, owing to the fact that by virtue of being separate there is no way for information to pass from one universe to another (note that this separation may exist in time, i.e. a series of unique universes, or in space, i.e. an array of concurrent universes, or both). Yet it remains possible that through study of the nature of this universe we may be able to prove either that other universes are impossible, or that they and the mechanism that produces them are possible, or even inevitable, and thus “multiple universes” is both verifiable and falsifiable, and therefore a valid hypothesis. But we may never know if it or any other theory is correct, and that is precisely the atheist’s point in considering it.

In contrast, many God hypotheses entail that we ought to have evidence that in fact we do not, requiring the theist to build an unwieldly architecture of ad hoc hypotheses that explain away this lack of evidence, whereas the “multiple universes” proponent needs no such ad hoc hypotheses. This gives the theory greater merit. Of course, this particular merit is equal to a theism that does not require an ad hoc architecture, but we must then pay attention to what sort of theism we are proposing as an explanation for apparent Fine Tuning before we can dismiss “multiple universes” so casually.

Again unlike theism, “multiple universes” has another inherent merit that Hannam does not consider: we know that a universe exists, and Hannam himself is agreeing that different universes are in principle possible, so we have a ready explanation of what is unknown by appealing to a known entity–that universes exist. In contrast, the theist tries to explain the same unknown by appealing to a completely unknown entity, that is, an entity that has never been scientifically observed and could well not exist at all. How does it make more sense to appeal to such a strange and unobserved entity when we can explain the same things by appealing to an entity that everyone agrees exists? Since a universe exists, and other universes are possible, isn’t it plausible that other universes exist? Certainly, we cannot know they do. But we cannot know they do not and thus any argument for God that supposes they do not is an argument from ignorance. Once again, agnosticism is the only justified outcome of this line of reasoning.

(6) Incorrect Concluding Remarks

Hannam thinks the Fine Tuning argument would be stronger if the laws of physics in this universe were found to be such that even a slight change would make life impossible. But this is incorrect, since this would only return us back to the same issues explored above, and beyond agnosticism one cannot seem to go without any real evidence. He equally incorrectly thinks that discovering extraterrestrial life that was similar to our own would improve the force of a Fine Tuning argument for a Creator, when it could just as easily prove that this universe has an easy time of producing life, which through convergent evolution inevitably winds up along the same path, or that all life has a common extraterrestrial origin, as argued by Hoyle and others. This would tell us nothing about whether life would be possible in other universes, or whether this universe was designed to produce life (vs. life being merely another natural byproduct). For instance, what if the twenty or so amino acids on which earth life is based are the chemicals in this universe that most readily bind in long stable chains, and function most easily in the processes essential for forming life? If anything like this were true, it would mean that we ought to expect life to be similar all over the cosmos, even if nothing about it were finely tuned.

(7) Isn’t Creationism Vacuous?

Hannam’s essay did not even address the charge that Creationism is a vacuous theory: intelligent design does not make any real predictions. It doesn’t even attempt to explain why we have the specific laws we do instead of others, i.e. others that would support life. It is thus not much of an explanation, being empty of any substantial content. For instance, why is the speed of light just what it is instead of something else? The theist can only appeal to God’s whim, while the scientist pursuing methodological naturalism has a good chance of finding a natural reason for the value of the speed of light. And we can say this, because history proves that the odds are entirely on his side (cf. my “Prima Facie Presumptions vs. The Lessons of History“).

In contrast, what a Creationist theory ought to predict, which a naturalistic theory would not, does not turn up. The laws of physics proceed without any regard for right or wrong, good people or bad, and they proceed relentlessly and monotonously, never demonstrably deviating, much less with anything like a value-laden purpose; resources are arbitrarily limited and randomly distributed without regard for merit; and no clear supernatural events or messages are present in any of our lives–guns are not suddenly turned into flowers, churches are not protected by mysterious energy fields, True Bibles are not indestructible nor do they glow in the dark, preachers cannot regenerate lost limbs, and when we ask God a question, with all sincerity and earnest urging, we never receive a clear, reasoned answer that all can hear and agree upon. Thus, intelligent design is a rather poor explanation for the universe we have, whereas a naturalistic theory fits it like a glove. We are faced with what is ultimately a mindless, careless, silent and blind machine. Doesn’t it make more sense that it should have an ultimately mindless, careless, silent and blind cause? The point is that Intelligent Design really doesn’t bring anything to the table as far as explaining our universe, our actual universe. By focusing solely on one single feature, an “ability to produce life,” Fine Tuning proponents miss the forest for the tree.


In the end, what we learn is that it is not possible to use Fine Tuning to argue for the existence of God. If one could prove a god existed by some other means, one could then use that to argue for Fine Tuning, but that is not what theists like Hannam are saying. The only way, it seems to me, to verify (and thus justify belief in) the hypothesis that the universe was intelligently designed is to verify that a designer existed and that an act of design occured. But if one could do that, one would no longer need design as a proof of such a deity. It remains possible that all known natural explanations could be shown to be illogical or falsified by contrary evidence, or that some sort of empirical evidence of intelligent design could be found sufficient to make that theory believable on its own terms, but this can only happen in the future, and is thus of no use to anyone who wants to argue now that a God exists.

Hannam actually agrees with this conclusion. He says in the end that Fine Tuning does not prove that a god exists. He also correctly notes that even if intelligent design were proved it would not prove the Judeo-Christian God existed, for some other god or force, from the Deist’s Maker to the ineffable Tao, could be at work. In the end, I agree with Hannam that the truth of theism is a really good explanation for a universe, just as the truth of Taoism is a really good explanation for my mystical experiences of the Tao (see From Taoist to Infidel). But that no more means God exists than it means Taoism is true: for being a good explanation is not sufficient for anything to be true. In the end, you need evidence: evidence for your hypothesis and evidence against all contrary hypotheses. And as there is none here, the existence of Fine Tuning, even if we could prove it existed (which we cannot), can lead to nothing more than agnosticism, not theism.

[1] i.e. in order for it to defeat the atheistic hypothesis, i.e. to be more believable and thus to justify belief in God as opposed to agnosticism. I do not mean that this is required for the theistic hypothesis to be rational. It is possible for creationism to be a rational belief, but not more believable than alternatives. I agree that, for example, a Deistic Creationism is perfectly rational. But we cannot know it is true, certainly not enough to believe atheistic alternatives are false. We can admit we do not know if it is true and believe it on faith, but we would have to admit in the process that this belief is not obligatory on anyone, not even ourselves.

[2] I do not mean chance is always a viable explanation if there is at least some chance no matter how small. Some probabilities are indeed too low to credit. What I am saying is that there is nothing objectively specific in the selection of cosmic constants that allows categorizing the particular set we have as any more improbable than others. In other words, if the selection of constants was random, then the odds were 100% that we would get a set of constants, and that set would produce effects only their precise arrangement would produce. We cannot say “these effects would not happen with any other set, therefore this set was not randomly chosen,” since we already know every set will have unique effects. Far from being improbable, the result we are pointing to is guaranteed to happen. This is like the configuration of the stars: that the stars should happen to be arranged so as to form the body of Orion when viewed from Earth is vastly improbable, but since every arrangement of stars would form one shape or another, the odds that we would find them forming a shape like Orion are 100%. Thus, we cannot conclude from the incredibly improbable arrangement of the stars that the arrangement was designed.

[3] In contrast, consider evolution: that a series of living genomes of ever-increasing complexity would arise, and in their particular order of increasing complexity, is not at all a given but is indeed inherently strange and in need of causal explanation. Even though it is remotely possible that it is a chance outcome, the odds are so low that this is wholly unreasonable to believe: some causal force must be at work instead (e.g. natural selection or design of some kind). But unlike this state of affairs, a universe would have a very complex arrangement of constants no matter what. The error, then, is in assuming that life is an effect more objectively strange than any other unique effect. But there is no objective basis for such an assumption. Our only reason for regarding life as special is entirely subjective: we benefit from it being that way. However, I must note that I don’t regard this as a strong defeater of Creationism as a theory. That is, if this were the only argument I had against Creationism, i.e. Creationism had actual, strong positive evidence, I would believe Creationism. In other words, we need evidence to support Creationism. We cannot support it on the argument that “we’re lucky,” since that statement has no objectively-based truth-value.

all rights reserved