Home » Library » Modern Library » Ryan Stringer Fine Tuning

Ryan Stringer Fine Tuning


Two Fatal Problems with the Fine-Tuning Argument (2015)

Ryan Stringer

I. Introduction

Some theists think that scientific findings provide evidence for God’s existence. For instance, some of them have taken the beginning of the universe in a “Big Bang” to be evidence that God created it ex nihilo. More recently touted findings emphasize the uniquely life-permitting values of the universe’s fundamental physical constants. If any of six constants had been “set” at even a slightly different value from its actual value, then life as we know it would have been impossible. The remarkable fact that the universe’s physical constants happen to have these uniquely life-permitting values suggests, according to some theists, that they were “fine-tuned” by God to have these values. More specifically, it is the extreme improbability of the constants having, by chance, the uniquely life-permitting values they do have that suggests that God “fine-tuned” the constants to these values. This is the so-called “fine-tuning argument,” which I regard as the most promising theistic teleological argument.

Despite its initial promise, however, there are grounds for rejecting the fine-tuning argument as a justification for theism (or even supernaturalism). As the title of this paper suggests, there are (at least) two fatal problems with the argument. The first is that it contains an internal conflict, and the second is that it misuses improbability to conjure up support for supernaturalism. Once these problems are brought to light, it is evident that the fine-tuning argument supports neither theism nor supernaturalism.

II. A More-Detailed Formulation of the Fine-Tuning Argument

Before presenting my two fatal problems with the fine-tuning argument, it is worth spelling out a more detailed, formal version of the argument:

(P1) Life as we know it requires that the universe’s physical constants have their actual values instead of some other set of values.

(P2) The actual values of the universe’s physical constants are uniquely life-permitting. (from P1)

(P3) The uniquely life-permitting values of the universe’s physical constants are due to necessity, chance, or supernatural fine-tuning.

(P4) The uniquely life-permitting values of the universe’s physical constants are not due to necessity.

(P5) Given the enormous (perhaps infinite) number of possible values for the universe’s physical constants, it is extremely improbable for these constants to have, by chance, the uniquely life-permitting values instead of some other, non-life-permitting set of values.

(P6) The uniquely life-permitting values of the universe’s physical constants are probably not due to chance. (from P5)

(P7) The uniquely life-permitting values of the universe’s physical constants are probably due to supernatural fine-tuning. (from P3, P4, & P6)

(C) Therefore, it is defeasibly reasonable to attribute the uniquely life-permitting values of the universe’s physical constants to God.

P1 is the key scientific finding that grounds the fine-tuning argument. Once again, if any of the universe’s physical constants had been “set” at even a slightly different value, then life as we know it could not have arisen. It could only have arisen on the condition of the universe’s physical constants being their actual values. P2 then follows from P1: if the actual values of the universe’s physical constants are the only ones that permit life to arise, then they are uniquely life-permitting.

P3 seems to be an obvious fact: the uniquely life-permitting values of the universe’s physical constants are due to necessity, chance, or supernatural fine-tuning—there is nothing else that could be responsible for them. P4 also seems to be obvious: the uniquely life-permitting values of the universe’s physical constants are not due to necessity because they need not have been these values; they surely could have been some other set of values.

Then we have the crucial premise P5. Not only could the universe’s physical constants have been “set” to some other, nonactual set of values, but there is an enormous, if not an infinite, number of possible values that they could have been. If you think of the actual values as a set of numbers, it becomes obvious that we can get a new set of numbers—and thus a new set of possible values for the universe’s physical constants—just by slightly tweaking one of the numbers of the actual values. And since we can slightly tweak an enormous, if not an infinite, number of times, it follows that there is an enormous, if not an infinite, number of possible values for the universe’s physical constants—hence the first part of P5. And from that part the second part trivially follows. For if there is an enormous, if not an infinite, number of possible values for the universe’s physical constants, all of which have an equal shot of obtaining by chance, then it is extremely unlikely for the constants to have, by chance, the one and only life-permitting set (or very narrow range) of values instead of one of the other, non-life-permitting sets of values.

That P6 then follows from the second part of P5 should seem obvious: if it is extremely unlikely that the one and only set (or very narrow range) of life-permitting values will obtain by chance, then it certainly seems to follow that the obtaining of these values is probably not due to chance. But if the obtaining of these values is probably not due to chance—and it is not due to necessity per P4—then the only option left, according to P3, is that it is probably due to supernatural fine-tuning, as deduced in P7 (from P3, P4, and P6).

Finally, the conclusion from P7 seems reasonable: if there probably is a supernatural fine-tuner behind the uniquely life-permitting values of the universe’s physical constants, then it is reasonable to postulate God as that supernatural fine-tuner. And this reasonableness is not conclusive, but is instead defeasible: there could be other considerations that end up overriding this initial reasonableness, thereby making it overall unreasonable to postulate God as the supernatural fine-tuner. However, insofar as there probably is a supernatural fine-tuner behind the actual values of the universe’s physical constants, it is reasonable, the argument contends, to postulate God as this supernatural fine-tuner.

Before I turn to evaluating this fine-tuning argument, note that I have formulated it very modestly in order to increase its odds of success. Specifically, I have framed it as providing grounds for the defeasible rationality of theism, not as proof that theism is true or even probably true. Those who think that the uniquely life-permitting values of the universe’s physical constants prove theism are mistaken for at least two reasons. First, they assume that the supernatural fine-tuning of these constants must be due to God, which is patently false—one or more other supernatural beings could just as easily be responsible for the fine-tuning. Second, that the uniquely life-permitting values of the universe’s physical constants are due to chance is ruled out based on the improbability—not the impossibility—of them having these values by chance. The actual values of the constants could very well be due to chance despite this extreme improbability; so even if supernatural fine-tuning could only be due to God, it would only be probable (and thus not proven) that God exists since it is only probable that the constants were supernaturally fine-tuned to their actual values.

Those who think that the uniquely life-permitting values of the universe’s physical constants make theism probable are also mistaken. For even if these values were probably due to supernatural fine-tuning, it would not follow that God is probably the one responsible for the fine-tuning. As previously noted, one or more other supernatural beings could just as easily be responsible, and so the need for a supernatural fine-tuner does not imply that any particular option, such as God, is probably responsible for the fine-tuning. At best, then, the need for a supernatural fine-tuner of the universe’s physical constants makes it defeasibly reasonable to postulate that God is that fine-tuner. And that is why I have framed the fine-tuning argument as an attempt to establish the defeasible rationality of theism.[1]

III. The Problems with the Argument

Before presenting two fatal problems with the fine-tuning argument, I want to mention some ways to contest it that I will set aside.[2] The first targets the inference from P1 to P2: P1 entails that the values of the universe’s physical constants are uniquely life-as-we-know-it-permitting, not that they are uniquely life-in-general-permitting. For all we know, there could be sets of nonactual values for the universe’s physical constants that would permit other forms of life. This possibility is also problematic for the inference embodied in P5. For if there are sets of nonactual values that would permit other forms of life, it may not be extremely improbable for the universe’s physical constants to turn out to have life-permitting values by chance given the enormous range of possible outcomes for these constants. Despite the enormous range of possible outcomes, a significant number of them might very well permit life in some form, in which case it would not be extremely improbable for the universe’s physical constants to have life-permitting values by chance.

This first objection to the fine-tuning argument is not very strong. For even if there could be sets of other-kinds-of-life-permitting values for the universe’s physical constants, we certainly do not know that there are such things. As far as we know, there is only one set of life-permitting values for the universe’s physical constants; so the inference from P1 to P2, and the inference embodied in P5, both seem to hold. Furthermore, the proponent of the fine-tuning argument could sidestep this objection altogether by replacing each instance of the phrase “life-permitting” with the phrase “life-as-we-know-it-permitting.”

Another way to contest the fine-tuning argument is to contest P4, perhaps by maintaining, as Baruch Spinoza did, that (1) nature itself necessarily exists and (2) everything happens necessarily. This naturalistic necessitarianism would make the actual values of the universe’s physical constants necessary as well. One might also contest P4 on scientific grounds. Perhaps the universe somehow requires, for scientific reasons, that the values of its constants be their actual ones. Finally, at the very least one could complain that P4 begs the question by simply assuming that the actual values of the universe’s physical constants are not the only possible ones. But even though I think that P4 could be false, I cannot resist accepting it as true. Spinoza’s necessitarianism seems outlandish to me, and I am not sufficiently versed in science to know of any actual way to contest P4 on scientific grounds. It seems entirely possible for the values of the universe’s physical constants to have been otherwise; so I will assume, along with fine-tuning proponents, that they could have been otherwise, and thus that P4 is true.

There is yet another way to contest P5: one could question its assumption that the probabilities of the possible values for the universe’s physical constants are equal. However, I know of no good grounds for doing so, and from an a priori perspective these probabilities are equal. Because this objection seems inadequate for contesting P5, I will assume that P5 is true.[3]

One could also challenge the inference from P5 to P6 by postulating the existence of many different universes. Though the obtaining of the actual values of the universe’s physical constants by chance is extremely improbable in a single universe, given the enormous amount of possible values for those constants, if a sufficiently large number of universes exist, then (1) we can expect the actual values of the universe’s physical constants to obtain by chance in some universe, and thus (2) we cannot rule out chance as responsible for the actual values of the universe’s physical constants. This strategy is similar to that for responding to a similar theistic argument based on the improbability of earthly conditions conducive to life. That argument goes something like this. Since it is extremely improbable for the earth to have, by chance, an environment hospitable to life given the enormous number of possible environments that could have obtained, most of which would have been inhospitable to life (e.g., if earth were closer or farther away from the sun), then it was probably designed by some supernatural intelligence such as God. But even though it may be extremely improbable for the earth to have, by chance, an environment hospitable to life given the countless number of possible inhospitable environments, once we factor in that there are billions of planets, attributing the earth’s life-friendliness to any supernatural being becomes completely unwarranted: since there are billions of planets in the universe, we can expect at least one of them to turn out, by chance, to have life-friendly conditions, which means that we cannot rule out chance as responsible for these conditions obtaining on earth.[4]

Unfortunately, postulating many different universes does not seem to work as well against the fine-tuning argument. While we know that there are billions of planets in the universe, we certainly do not know that there are multiple universes—let alone billions of them. In fact, it strikes me as sheer conjecture to postulate anything more than the single universe that we know exists. (Then again, I am not well versed in physical cosmology, so perhaps my intuition here is unwarranted.) In any event, I will treat the postulation of many different universes as an inadequate way to contest the inference from P5 to P6.

The last way to contest the fine-tuning argument that I will be setting aside targets the final inference from P7 to the conclusion. Even if the uniquely life-permitting values of the universe’s physical constants are probably due to supernatural fine-tuning, how is it even defeasibly reasonable to choose God as the supernatural fine-tuner? While God is a potential candidate for the supernatural fine-tuner, there are other individual candidates, or groups of them, that could just as easily do the job, and no reason to favor God over these other options.[5] As such, it is not even defeasibly reasonable to pick out God as the supernatural fine-tuner—that would in fact be completely arbitrary.

Although I am inclined to think that this objection is sufficient to undermine the fine-tuning argument, a proponent could respond by arguing that it can still be defeasibly reasonable to attribute the fine-tuning to God even though he is only one option among many. For even though there may be no reason to favor God over the others, such a reason is not necessary for it to be defeasibly reasonable to pick out God as our fine-tuner. Instead, it is defeasibly reasonable to pick out God because he is, at least within the present context alone, no worse than any other option. Put another way, even though we may have no reason for favoring one option over another, one of those options must be the right one, and it is defeasibly reasonable to pick any option that is no worse than any other within the context of considering the fine-tuning argument all by itself.

While this response strikes me as inadequate because it amounts to maintaining the defeasible reasonableness of arbitrarily picking one’s favorite option out of many that are no worse than any other, I will not press this point. Instead, I will finally present my two fatal problems with the fine-tuning argument. The first is the internal conflict between the argument’s conclusion and its premises, which can be seen by considering the parallel “divine-pruning argument” parody:

(Q1) The uniquely life-permitting values of the universe’s physical constants are probably due to supernatural fine-tuning. (P7 of the fine-tuning argument)

(Q2) The supernatural fine-tuning of the universe’s physical constants is either due to God, or to some other supernatural being or beings.

(Q3) Given the enormous (perhaps infinite) number of supernatural options that could account for the fine-tuning of the universe’s physical constants, it is extremely improbable that this fine-tuning is due to God instead of some other supernatural being or beings.

(Q4) The fine-tuning of the universe’s physical constants is probably not due to God. (from Q3)

(Q5) The fine-tuning of the universe’s physical constants is probably due to some other, non-God supernatural being or beings. (from Q2 & Q4)

(C) Therefore, it is not defeasibly reasonable to attribute the uniquely life-permitting values of the universe’s physical constants to God. (from Q5)

This argument begins with the last premise of the fine-tuning argument and then, by utilizing the same strategy employed to derive that premise, delivers a conclusion that flatly contradicts the conclusion of the fine-tuning argument. If the premises of the fine-tuning argument are good and their strategy actually rules out the uniquely life-permitting values of the universe’s physical constants being due to chance, then the divine-pruning argument shows that the fine-tuning argument’s conclusion is dubious since its strategy is equally effective at ruling out God as the one responsible for fine-tuning. Consequently, the fine-tuning argument fails to provide justification for theism.

Of course, fine-tuning proponents could try to undermine the divine-pruning argument without abandoning the strategies employed by both. They could contest Q3, for instance, by rejecting either (1) that there is an enormous number of supernatural options that could account for the fine-tuning of the universe’s physical constants, or (2) that the enormous number of supernatural options entails the extreme improbability of God being the one responsible for fine-tuning.[6]

However, both of these options fail. To legitimately reject (1), it must be shown that, contrary to appearances, most of the imaginable supernatural options are not genuine ones. This in turn requires demonstrating that these options are incoherent in some way; yet surely this cannot be done. To legitimately reject (2), one not only needs to show that (a) the fine-tuner options are not equally probable, but that (b) the probability of the God option is higher than that of each non-God option such that the God option is not extremely improbable. But then one must assess the probabilities of the fine-tuner options, which can be done most thoroughly by using both a priori and a posteriori considerations. But even if the fine-tuner options are not equally probable, the God option will not even emerge as more probable than each of its competitors—let alone so much more probable than them that it is not extremely improbable.[7] For starters, the God option will not emerge as more probable than each of its competitors from an a priori perspective. In fact, the God option cannot even be equally probable as its competitors unless we make two crucial assumptions. First, we must assume that incompatibility arguments attempting to show that the concept of God is incoherent are without force since these arguments threaten the very possibility of God, and thus the option of God as the fine-tuner. Second, we must assume, conceptually speaking, that God is not a necessary being. Otherwise the very possibility of God, and thus the option of God as the fine-tuner, would conflict with some very strong possibilities—e.g., the possibilities of gratuitous evil, of all minds being physically realized, and of there being some non-God creator(s) of the universe.[8] By making both assumptions we avoid these a priori grounds for rendering the probability of the God option lower than that of the countless supernatural options that are immune to similar threats.[9]

Yet even if we make these assumptions, the probabilities of these supernatural fine-tuning options are, for the most part, like the probabilities of the possible values for the universe’s physical constants—they are equal from an a priori perspective. For as long as an option is (1) free from incoherence, (2) not threatened by any competing possibility, and (3) a being or beings that could have a reason for particularly wanting life-permitting constants, then that option and the God option will be, like the numbers on fair dice, equally probable from an a priori perspective. And since at least most of God’s competitors will be such options, the God option is at best just as good as an enormous number of competitors from an a priori perspective. The probability of the God option, then, is not higher than each of its competitors from an a priori perspective.

This leaves only a posteriori considerations to demonstrate (b). These considerations fall into two categories: those that are illegitimate to use, and those that reduce the probability of the God option compared to at least some its competitors. The illegitimate a posteriori considerations are nonteleological theistic arguments and nonteleological atheistic arguments. Using them would require us to look beyond the independent merit of the fine-tuning argument, with an eye toward evaluating other theistic (and atheistic) arguments to determine their effects on the probability of the God option over the other fine-tuner options. Since this would take us far from evaluating the merit of the fine-tuning argument itself, it is illegitimate to use these considerations.

The considerations in the second category are a posteriori arguments based in our experience of design. For example, all of our experience of design is of finite, imperfect beings designing things. In addition, great and complex designs are typically, if not always, undertaken by multiple finite, imperfect beings. Our experience thus suggests that the alleged fine-tuning of the universe’s physical constants—which would be a very great and complex feat of design—was likewise undertaken by multiple finite, imperfect beings.[10] These considerations reduce the probability of any option that does not consist of such beings, which of course includes the God option, and increases the probability of the options that do consist of such beings.

Overall, then, the probability of the God option does not emerge as higher than each of its competitors, and so (2) cannot be legitimately rejected. And since (1) also cannot be legitimately rejected, Q3 and the divine-pruning argument itself still stand. Consequently, the internal conflict between the fine-tuning argument’s conclusion and its premises has not been dissolved, and thus the fine-tuning argument still fails to provide justification for theism.

While this internal conflict is sufficient to show that the fine-tuning argument fails to support theism, it does not show that the argument fails to support supernaturalism in general. All that this internal conflict means is that the conclusion is no good if the premises are all good, which is obviously compatible with the premises being all good. And the cogency of the premises is enough to show that supernaturalism in general is supported by the argument. So despite its internal conflict, the fine-tuning argument is still a potential threat to naturalism.

But the other fatal problem with the fine-tuning argument is its misuse of improbability to conjure up support for supernaturalism in general. The problem lies with the inference from P5 to P6—namely, the inference from

(P5b) It is extremely improbable for the universe’s physical constants to have, by chance, the uniquely life-permitting values instead of some other, non-life-permitting set of values

to

(P6) The uniquely life-permitting values of the universe’s physical constants are probably not due to chance.

Now this inference is a specific instance of the following general assumption:

(A) If it is extremely improbable for a unique outcome to occur by chance, then that outcome’s occurrence is probably not due to chance.

But this assumption does not hold in every context. For instance, suppose that there was a lottery involving every single person on earth. Only one name will be drawn, and that person will win billions of dollars. The probability that any individual person will be drawn by chance is roughly 1 in 7 billion; so it is extremely improbable that any particular person will win by chance.[11] However, some particular person will win—that outcome will occur with absolute certainly. And yet the extreme improbability of that particular person winning by chance does not at all suggest that it will not happen by chance. In fact, the default position here is that this is precisely what will happen—some particular person will win by chance despite the extreme improbability of that person winning by chance. The default position is not that the lottery is probably rigged, which would be the case if it were always true that the extreme improbability of a unique outcome occurring by chance probabilistically ruled out chance as responsible for the outcome’s occurrence.

Now the lottery could of course be rigged. But we would need some positive evidence that it was rigged in order to have evidence against the default position that chance is responsible for the outcome. However, the extreme improbability of the outcome occurring by chance does not count as such evidence. Since the default position is that some particular person will win by chance despite the extreme improbability of that outcome occurring by chance, the evidence required to doubt or reject the proposition that chance was responsible for the outcome has to be something other than the extreme improbability of the outcome occurring by chance. For example, if there were a significant connection between the winner and the person(s) responsible for drawing the winner’s name, then there would be evidence to doubt that chance was responsible. Or if there were some evidence of human artifice, then there would be evidence to reject that chance was responsible for the outcome. But again, the extreme improbability of the outcome occurring by chance does not itself count as evidence against chance being responsible for the outcome. So assumption A does not hold in lottery contexts.

Although the situations are not exactly the same, we should approach the actualization of the uniquely life-permitting values of the universe’s physical constants in the same way that we would approach some particular person winning my hypothetical lottery. In the lottery case, the unique outcome—the particular person winning—is explained by either chance or human artifice. And even though it is extremely improbable for that particular person to win by chance, the default position is that the outcome was nonetheless due to chance. The extreme improbability of the outcome occurring by chance neither counts as evidence against it being due to chance, nor as evidence in favor of it being due to human artifice. Again, we need other evidence to overcome the default position that chance is the explanation for the outcome here. So assumption A does not hold.

Turn now to the actualization of the uniquely life-permitting values of the universe’s physical constants. The unique outcome—the obtaining of those values—is explained by either chance or supernatural fine-tuning. And just like in the lottery case, it is extremely improbable for that particular outcome to occur by chance. Of course, we cannot say that the default position here is the same as in the lottery case—that some particular outcome will occur by chance despite the extreme improbability of the outcome happening by chance. There is no presumption in favor of chance here. Fortunately, this does not matter. What does matter is that, as with the lottery, the extreme improbability of the unique outcome occurring by chance does not count as evidence against the outcome being due to chance. In each case, the extreme improbability of the outcome happening by chance merely reflects the fact that the outcome is one among an enormous number of possible outcomes, each of which is equally likely. It in no way indicates that chance is probably not responsible for the outcome. Evidence that chance is not responsible must be found in something other than the extreme improbability of the outcome occurring by chance. Consequently, assumption A does not hold in this context, either; so the inference from P5 to P6 does not go through.[12]

Despite these analogical grounds for doubting the inference from P5 to P6, some may still accept it because they are so flabbergasted by the life-permitting values of the universe’s physical constants. That these constants have these life-permitting values is too special, significant, or remarkable for them to just have happened by chance. But as natural as it may seem to think this way, doing so is unjustified. Consider again my hypothetical lottery from above. One would not be justified in thinking that someone probably rigged the lottery for the winner just because the outcome is special, significant, or remarkable. Unlike finding a connection between the winner and the person who picked the winner, or a connection between the winner and whoever was in charge of the lottery, the outcome’s being special, significant, or remarkable does not at all suggest that someone rigged the lottery for him or her to win as opposed to him or her winning by chance. After all, there is nothing even remotely problematic or odd about chance happenings being special, significant, or remarkable. People win unrigged lotteries all of the time. Meaningful relationships can spring from chance encounters. On his way home from running an errand on a Saturday morning, a cat lover randomly finds a cat sitting out front of his apartment complex’s front office that he will decide to keep.[13] That an outcome is special, significant, or remarkable constitutes no evidence against its being due to chance and in favor of its being due to intentional design; indeed, we should expect at least some outcomes to be special, significant, or remarkable by chance. So, even if the life-permitting outcome of the values for the universe’s physical constants were significant, special, and remarkable, that would be no reason to think that supernatural fine-tuning fashioned these constants instead of them being due to chance.

IV. Conclusion

I have here argued that there are two fatal problems with the fine-tuning argument. The first is that there is an internal conflict between the argument’s premises and its conclusion—namely, that if the premises hold, then the conclusion does not. This was shown through my parody divine-pruning argument. Its first premise was the final premise of the fine-tuning argument and, by utilizing the same strategy employed by the fine-tuning argument to reach its final premise, it delivered a conclusion that contradicts the fine-tuning argument’s conclusion. The second problem with the fine-tuning argument is its misuse of improbability. Specifically, it assumes that the extreme improbability of a unique outcome occurring by chance implies that chance was probably not responsible for that outcome’s occurrence, which is very dubious when applied to lottery-like contexts like the present one. These problems show that the fine-tuning argument supports neither theism nor supernaturalism in general, and so we should reject the argument as a justification of either position.[14]

Notes

[1] It might be even more modest to formulate the fine-tuning argument as an inference to the best explanation (IBE). In that case the first two premises would remain the same, but the third premise would assert that the uniquely life-permitting values of the universe’s physical constants are best explained by necessity, chance, or supernatural fine-tuning. It would then rule out both necessity and chance as the best explanation of these values (on the same grounds used in the original formulation), which of course would leave supernatural fine-tuning as the remaining best explanation of them. From here the IBE could assert the reasonableness of accepting supernatural fine-tuning as responsible for the constants’ values, and then from there make the final inference to the reasonableness of postulating God as responsible for those values. But even if this alternate formulation is more modest than the original, the criticisms that I level against the original fine-tuning argument seem to apply, mutatis mutandis, to the one grounded in an IBE. So there is no advantage to that alternate formulation.

[2] I set these other potential problems aside because even if they do not undermine the fine-tuning argument, it still fails in virtue of the two problems I highlight in this paper.

[3] Even if a critic of the fine-tuning argument could somehow undermine this assumption, a proponent could salvage P5 by arguing that, although the probabilities of the possible values of the universe’s physical constants are not equal, there is still an enormous number of nonactual values, each of which is probable enough to render extremely improbable the actual values obtaining by chance.

[4] Perhaps an example with numbers would be helpful here. Suppose that there are 500 billion possible earthly environments, only one of which is hospitable to life. It would thus be very improbable for the actual earthly environment to turn out, by chance, to be hospitable to life (the odds are obviously 1 in 500 billion). But if there were also at least 500 billion planets, then, given the odds, we would actually expect one of them to turn out to be hospitable to life by chance. And since there is no reason to reject a life-friendly earth as instantiating this expected result, we cannot rule out that the earth’s life-friendly environment is due to chance.

[5] Some theists might try to dispute my claim here on either of the following grounds. They might argue that God is a simpler option than the others, and that this is a reason to favor the God option over the other fine-tuning options. Or they could argue that since God is the kind of being that would especially want life to exist, the uniquely life-permitting values of the universe’s physical constants are particularly compatible with God fine-tuning those constants to have those values, and that this is a reason to favor him over the other fine-tuning options. But both objections fail here. As for the first one, God is at best a simpler option than the polytheistic ones; he is not a simpler option than the other single-being options (such as his indifferent and perfectly evil counterparts). So appeals to simplicity would not supply us with a reason to pick out God over all of the other fine-tuning options. As for the second objection, there are plenty of possible supernatural beings other than God that could especially want life to exist (for any number of reasons), and thus that are particularly compatible with the fine-tuning of the universe’s physical constants. So the particular compatibility of the life-permitting values of the physical constants with God fine-tuning those constants to have such values is no reason to favor him over the other fine-tuning candidates, either. If anything, we have reason not to pick out God over the other fine-tuning candidates. Given the numerous examples of design in the universe that are due to multiple designers, it is dubious to think that postulating a single-being option in this context is more reasonable than postulating a polytheistic one. In fact, it is arguably more reasonable, on empirical grounds, to postulate a polytheistic option over a single-being option since the alleged fine-tuning of the universe is an enormously impressive feat of design. For an argument to this effect, see Michael Martin, Atheism: A Philosophical Justification (Philadelphia, PA: Temple University Press, 1990), pp. 324-326.

[6] It would do no good to point out that, while we can generate an enormous number of possible values for the universe’s physical constants just by changing the numbers, we cannot generate an enormous number of possible supernatural fine-tuning options in the same manner. For even though we obviously cannot generate an enormous number of possible supernatural fine-tuning options just by changing numbers, we can do so in a similar way by modifying (a) the properties of the supernatural fine-tuners and (b) the number of supernatural beings responsible for the fine-tuning. So, for example, we might start with God as a possible supernatural fine-tuning option, and then generate more possibilities by modifying God’s defining set of properties in various ways. An obvious move here would be to replace God’s supreme goodness with supreme evilness, or replace it with utter indifference. These basic moves would immediately give us two more supernatural options, and of course there is room for more complex modifications still. And it would not be difficult to supplement this process by thinking of supernatural fine-tuning options that consist of multiple supernatural beings that are jointly responsible for the fine-tuning of the universe’s physical constants. Like our imaginations, the possibilities here are seemingly endless.

[7] The probability of the God option plus that of all of the non-God options combined is one, and so the probability of the God option is one minus the probability of all of the non-God options combined. So in order for the God option to fail to be extremely improbable, it would have to be so much more probable than each of its competitors that, when the addition of their probabilities is subtracted from one, the result is not an extremely small probability value.

[8] For a few more of these possibilities and the explanations for how they conflict with the possibility of God on the assumption that he is a necessary being, see my “Modal Arguments for Atheism” (2012) on the Secular Web.

[9] Although these assumptions might be controversial, they are perfectly legitimate to make in the present context because, in the first place, they must be made in order to evaluate the fine-tuning argument in its own right. On the one hand, if the incompatibility arguments do have force, then they undermine the very possibility of God and therefore any hope of the fine-tuning argument providing reasonable grounds for maintaining his actual existence. On the other hand, if God is assumed to be a necessary being, then the fine-tuning argument becomes completely superfluous. After all, the proponent of the fine-tuning argument must be assuming, prior to mounting the argument, that God could be the supernatural fine-tuner of the universe’s physical constants—otherwise there would be no point in attempting to mount the argument. But then, if the proponent is also assuming that God is a necessary being, he already has everything that he needs to make a swift case for theism before the fine-tuning argument even begins: from his assumptions that (a) God is a necessary being and (b) it is possible that God is the supernatural fine-tuner of the universe’s physical constants, it follows that (c) God necessarily exists. The proponent of the fine-tuning argument would then not even need that argument to defend his theism, especially since he can establish a stronger theistic conclusion in a much swifter fashion. Furthermore, as I argue in my “Modal Arguments for Atheism,” the assumption that God is a necessary being is highly problematic in its own right.

[10] For explicit formulations, explanations, and defenses of such arguments, see Michael Martin, Atheism: A Philosophical Justification, Chapter 13.

[11] One might be tempted to complain about the particular numbers used in my example, but nothing hinges on them. Jack them up all you want—you will still get the same results.

[12] I do not mean to suggest here that the actual values of the universe’s physical constants were in fact due to something akin to a chance lottery drawing. Instead, I am only arguing for the weaker claim that a chance explanation cannot be probabilistically ruled out by the extreme improbability of the values obtaining by chance.

[13] As you may have guessed, I am that cat lover.

[14] Many thanks to an anonymous reviewer for helpful comments on an earlier version of this paper.


Copyright ©2015 Ryan Stringer. The electronic version is copyright ©2015 by Internet Infidels, Inc. with the written permission of Ryan Stringer. All rights reserved.