Authored by: Andrew Johnson.

The cosmological fine tuning argument is commonly cited as one of the most interesting and persuasive arguments in the theist’s arsenal. This argument focuses on the fact that the laws, constants, and initial conditions of the universe must all fall within a narrow range for life to be possible at all; and whether the fact that it does is more likely on theism, or on naturalism. I will present Robin Collins’ fine-tuning argument (Collins 2009, pp. 202-281), and show that, though it attempts to explain how the fine-tuning evidence favors theism, it fails.

Collins’ Argument Paraphrased

Collins’ cosmological fine-tuning argument is deceptively simple, yet requires an extensive vocabulary of abbreviations, various hypotheses, and scientific and philosophical terms. Here, I will attempt to offer a version of the argument that is more accessible, without losing any of the power or simplicity of the original¹.

1. Given the fine-tuning evidence, it is very, very surprising that a life-permitting universe (LPU) exists, under a naturalistic single universe hypothesis (NSU).

2. Given the fine-tuning evidence, it is not surprising at all that a life-permitting universe exists, under a theistic hypothesis (T).

3. The theistic hypothesis was advocated prior to the discovery of the fine-tuning evidence, and thus is not ad hoc.

4. Therefore, by the likelihood principle, LPU (the existence of a life permitting universe) strongly supports T (the theistic hypothesis) over the NSU (naturalistic single universe hypothesis).

(Collins 2009, p. 207)

The fine-tuning evidence of which Collins writes is the fact that the laws, constants, and initial conditions of our universe must fall within a very narrow range in order for a universe to be life permitting. A life permitting universe is one which can support the existence of, what Collins calls, embodied moral agents; complex physical beings possessing intelligence and the ability to make moral choices and affect one another. The naturalistic single universe hypothesis states that there is only one universe and that the values of the constants, laws, and initial conditions are a unexplainable brute fact, and could have had any value from a very wide range. The theistic hypothesis states that there is an omnipotent, omniscient, eternally existing, and free creator of the universe. And the likelihood principle states that the degree that a piece of evidence counts towards one hypothesis over another, is proportional to the ratio of how surprising the evidence is under each hypothesis.

The justification for the first premise comes from the fact that there is a very large range of values that each physical constant could have taken, but only a very small range of values which would have allowed the resulting universe to be life-permitting. And, given the naturalistic single universe hypothesis, there is no reason to expect each constant to take one value, as opposed to any other. That each constant took a value which allows the universe to be life-permitting, then, is very surprising under NSU.

The second premise is justified by appealing to God’s motivations, which we can know by reasoning from the attributes that Collins takes Him to have. Collins explains that the only reasons God would have to do anything would be to increase the moral and aesthetic value of reality (2009, p. 254). From that, it is reasonable to conclude that the existence of embodied moral agents does add significantly to the moral and aesthetic value of reality. Therefore, it is not surprising at all that God would create a universe (with aesthetic value in its own right) which could hold such creatures.

The third premise seems obvious, in that the theistic hypothesis was widely held well before the fine-tuning evidence came to be understood.

As for the conclusion, the likelihood principle is fairly uncontroversial and derivable from Bayes’ Theorem. If the first two premises are true, then LPU does serve as powerful evidence for T and against NSU.

The Understatement of the Century

Imagine a trial where the defendant stands accused of stabbing a man to death in the victim’s living room. The accused takes the stand, and admits to being in the victim’s house at the time of the murder. Does this evidence support the “defendant is guilty” hypothesis or the “defendant is innocent” hypothesis?

It seems easy to see that it would be unsurprising that the defendant would be in the house at the time of the murder, if he is guilty. It also seems quite surprising that the defendant would be in the house if he were innocent. Using the likelihood principle, this evidence counts toward the guilty hypothesis and against the innocent hypothesis.

Were this singular piece of evidence to be the extent of our knowledge of the events in question, this conclusion seems quite reasonable. But how would our evaluation change, were we to also know more specific facts about the case? If the defendant were locked in the basement of the victim’s house at the time of the murder, it would still be true that he was in the house at the time of the murder. But the additional, more specific evidence seems able to render the initial likelihood calculation irrelevant when determining which hypothesis is supported by the evidence.

What exactly is the relationship between more general pieces of evidence which seem to favor one hypothesis over another, and the more specific evidences which point in the opposite direction? And, most importantly, what does this have to do with the fine-tuning argument?

As for the first question, Paul Draper calls this the fallacy of understated evidence. “This fallacy (i.e., mistake in reasoning) is committed when one uses some relatively general known fact about X to support a hypothesis when a more specific fact about X (that is also known to obtain) fails to support that hypothesis. (Draper, 2008)” We can see that this applies to the case of the man accused of murder; the general fact is that he was in the victim’s house at the time of the killing, the specific fact is that he was locked in the basement while the murder took place.

As for the second question, Draper accuses Collins of unintentionally committing the fallacy of understated evidence by, “understating what we know about life, Collins makes the fine-tuning data appear to support theism more than it really does. (Ibid.)” He agrees with Collins that the mere fact that intelligent life, of some sort, exists is less surprising under theism than naturalism. But Collins ignores the more specific evidence that humans are the only form of intelligent life known to exist. Draper takes this more specific evidence to point strongly toward naturalism.

Menschliches, Allzumenschliches

Let us consider a second, more specific piece of evidence:

HE: The only intelligent life we know to exist is human.

Then, the following argument can be formed:

5. Given the fact that intelligent life exists, it is not surprising at all that the only intelligent life we know to exist is human, under a naturalistic single universe hypothesis (NSU).

6. Given the fact that intelligent life exists, it is very, very surprising that the only intelligent life we know to exist is human, under a theistic hypothesis (T).

7. The theistic hypothesis was advocated prior to the discovery of the fine-tuning evidence, and thus is not ad hoc.

8. Therefore, by the likelihood principle, HE (humans are the only form of intelligent life known to exist) strongly supports NSU (the naturalistic hypothesis) over T (the theistic hypothesis).

Under the naturalistic single universe hypothesis, it is no surprise at all that HE is the case. In evolutionary terms, intelligent life is expensive, requiring huge amounts of resources. Given that intelligent life exists, we should expect that it would be relatively unsophisticated, only incrementally better morally, aesthetically, in intelligence, and in emotional sophistication than its evolutionary ancestors, and far more common in the universe than more advanced life.

However, under the theistic hypothesis, HE is very surprising. If God’s motivation can be understood entirely in terms of adding moral and aesthetic value to the universe, that the most morally and aesthetically valuable intelligent life is human, seems very unlikely. An omnipotent God would have the power to create life which was not constrained by its evolutionary history, which was not limited in morally and aesthetically irrelevant ways, and which was “better” intellectually, physically, and emotionally. Given that intelligent life exists and the truth of the theistic hypothesis, HE is very, very surprising indeed.

The conclusion of this argument mirrors the conclusion of Collins’. As HE entails that a life-permitting universe exists, by using the more specific evidence, we should expect to obtain a more accurate conclusion. And, if this conclusion is true, it should dampen, if not extinguish entirely, the persuasive power of Collins’ fine-tuning argument.

Of Angelic Aliens…

One possible objection to this reasoning rests on the idea that, although humans are not the only kind of intelligent life we would expect God to create, our existence is a net-positive, all things considered. So the creation of humans could very well play a role in maximizing the moral and aesthetic value of reality.

I will agree with the first part of that objection; it does seem that humans, for all our flaws, are a net-positive in terms of moral and aesthetic value. But I would disagree with the second part; our existence cannot fit with a plan to maximize the moral and aesthetic value, carried out by an omnipotent being.

It seems very easy to conceive of a species of intelligent life, which was morally and aesthetically superior (if only marginally) to humans, able to live and thrive in exactly the same kinds of environments as humans, and which would be far more likely candidates for existence under the theistic hypothesis. It would be very surprising, if all of our intuitions of what kind of intelligent life is possible were false. Therefore, it would be very surprising were HE to be true, even if it were also true that humans play but a small role in a larger plan to maximize the moral and aesthetic value of reality.

…and Panglossian Possibilities

A different sort of objection to the probability calculation involving HE, is that it is possible that humans are much, much more morally and aesthetically valuable than we give them credit for, that it is possible that God has knowledge of some kind of moral consideration which makes the existence of humanity consistent with the maximization of moral and aesthetic value.

On the face of it, I would agree; this is certainly a possibility. God, with his perfect view of the moral landscape, would possess a better perspective on the relevant moral issues regarding the value of humanity. But what we are concerned with, for the purposes of this argument, is not mere possibility, but the likelihoods of those possibilities.

If the theist maintains that she would not expect to know the relevant moral considerations which would affect what actions God would take in order to maximize the moral and aesthetic value of reality, this does nothing to undercut how surprising it would be, to us, that morally and aesthetically superior beings would not raise the moral and aesthetic value of reality more than morally and aesthetically inferior beings.

So, without some kind of explicit justification for the idea that, contrary to our own moral intuitions and reasoning, God does have a good moral reason for creating humans, which is also non-ad hoc and does not fall victim to Collins’ probabilistic tension criticism for extended hypothesis, we are completely justified in continuing to believe that HE is surprising.

Conclusion

The fine-tuning argument does show that the existence of intelligent life is more likely if theism is true, rather than if naturalism is. Were the singular fact that intelligent life exists make up the totality of our evidence, this argument would require the rational person to significantly revise their degree of belief in theism upward, and their degree of belief in naturalism downward. Unfortunately for the proponents of the fine-tuning argument, LPU is not all the evidence we have available to us. If the mere specification of the kind of life which is known to exist is enough evidence to cancel out the epistemic effects of the fine-tuning argument, then there seems to be no repairing it.

Notes

[1] I have endeavored to reproduce the argument without introducing any subtle errors. To be sure, any that are found are due to my rendering of the argument, not the original. I would urge anyone interested in learning more about fine-tuning arguments to read the original, as it is a model of clarity and precision.

Works Cited

Collins, Robin. “The Teleological Argument: An Exploration of the Fine-Tuning of the Universe.” The Blackwell Companion to Natural Theology. Ed. William Lane Craig and J. P. Moreland. Oxford: Wiley-Blackwell, 2009.

Draper, Paul. “Collins’ Case for Cosmic Design (The Great Debate).” The Secular Web. Internet Infidels Inc., 2008. Web. 26 Apr 2010. .