Theistic Anthropic Principle Refuted
A Survey of Arguments Against the Theistic Anthropic Principle
by Victor Gijsbers
for Positive Atheism Magazine

August 2000

The Anthropic Principle is a widely debated subject in the theism-atheism discussions of the past decade. Increased insight into physics has shown us that many of the physical constants of the Universe are such that if they had been even slightly different, life would never have been able to evolve. At a casual glance it would appear that those constants have been designed for bringing forth life; and this is exactly what theists using the Anthropic Principle claim: God 'fine-tuned' these constants at the moment of Creation, since he wanted life to come into existence.

The argument which I shall refer to as the 'Theistic Anthropic Principle' is often presented in the following way: The combination of physical constants that we observe in our universe is the only one capable of bringing forth life as we know it. Since we can conceive of other combinations of physical constants, some explanation is needed why our actual combination of physical constants exists rather than a different one. That explanation is given by the theory that God, an all-loving, all-powerful and perfect being, created the Universe with constants 'fine-tuned' just so that (human) life could develop. No other theory is as good as this one, so we can safely assume that God exists.

The probabilities of the Universe being just so that humans could live in it have, by some of the proponents of Theistic Anthropic Principle (TAP), been calculated to be so absurdly low that they say we may safely assume this could never have happened by chance. However, TAP is not unchallenged, and I will now give an overview of the ways in which it has been attacked.

The concept of 'fine-tuning' begs the question

Some have suggested that this argument begs the question, since the phrase 'fine-tuning' presupposes that the physical constants can be fine-tuned. It thus introduces a 'fine-tuner', God, as one of the premises of the argument. In another sense, it has not been proven that there actually is a way to change the physical constants, and as long as that is not done the argument is insubstantial. 'A demonstration that the fundamental constants are in fact tunable is prerequisite to speaking of fine-tuning.' [4] Thus the first argument against TAP is that it uses the unproven assumption that this fine-tuning is possible.

TAP is a form of 'carbocentrism'

TAP assumes that only a very small number of possible universes would bring forth life. This statement is certainly true for the particular kind of carbon-based life that we know. But this does not mean that other universes might not bear other forms of life. 'Sufficient complexity and long life may be the only ingredients needed for a universe to have some form of life. Those who argue that life is highly improbable need to open their minds to the fact that life might be likely with many different configurations of laws and constants of physics.' [9] Thus, TAP has not shown that our Universe is special in any way. If most Universes would bring forth life, the fact that ours has done so needs no explanation.

TAP is a form of 'biocentrism'

TAP claims that our Universe is special since it brought forth life, and that other Universes would not be special since they would not bring forth life. However, why would 'life' be the only thing that could make a Universe special? 'Assuming that other combinations of physical constants are physically possible, I see no reason to believe that all of them would result in a universe with less variety and complexity than our universe.' [2] Even if life failed to come into existence, other things at least as interesting might happen. Once again, TAP has not proven its claim that our Universe is special.

TAP is based on faulty analogies

The most popular, and most convincing, way of giving TAP is in the form of an analogy. One such analogy is: 'Suppose that a prisoner would stand before a firing squad, consisting of 50 marksmen. When the sign is given, they all shoot at him. But look, he still lives! What is a better explanation, that all marksmen missed by accident, or that there was a conspiracy to save the prisoner?' The marksmen are supposed to be the physical constants, which all 'missed' so that life could develop. However, the only thing this analogy tells us anything about is firing squads, not universes. 'To argue for design is to make an inductive argument, an argument by analogy, about the universe as a whole (of which there is only one, which we know of at least), and about its origin in particular, about which we know very little. One should be hesitant to draw many conclusions from a comparison of such dissimilar entities.' p[4] But this is exactly what TAP does: it extrapolates our ideas of the designed 'things' to the Universe, which is so dissimilar that an analogy is not allowable.

God is not the best explanation for the Anthropic Coincidences

A central part of TAP is that a creation by God (G) is the very best explanation for the state of the Universe. This might be attacked in two ways: by showing that G is not a good explanation, or by showing that better explanations exist.

G has a number of failures that make it a less than adequate theory. It does not explain how God created the Universe, and it does not tell us how God fine-tuned the physical constants. Creation out of nothing (which is certainly not the same as creation by quantum fluctuations) is an idea which is hard to understand. G is both incomplete and incomprehensible. Then there are problems with the concept of 'God'. How can a supernatural entity 'exist'? How can one be both all-powerful and all-knowing? What did an all-loving being do before there was anything to love? How could God have existed forever? How can God influence the Universe if he is outside of time? Theists often claim that these are mysteries, but that is no good: we began with a mystery, and now we have a mystery again. Why, if God is interested in humanity, did he create a Universe that needed 10 thousand million years to produce us? Why isn't earth more hospitable to us? It seems unreasonable that an all-powerful God would not have created us right at the start. Why didn't God fine-tune the Universe better, so genetic mutations would generally be less harmful and more useful? Why didn't he fine-tune it better so that there would be fewer natural disasters on earth? 'For all of these reasons, G can be seen to be a very poor explanation for the fact to be explained. It is incomplete, incomprehensible, obscure, unreasonable, anomalous, and counter-intuitive. It also appeals to still greater mysteries than the fact to be explained, so it hardly qualifies as an adequate explanation. It fails to illuminate anything or to enlarge our understanding.' [2]

Are there better explanations? Indeed there are. The first of these, E, posits not the existence of an all-loving God, but of a more or less evil deity. This is a better explanation, for it explains why the conditions on earth are not better fine-tuned for the human race. If this deity is indifferent towards human life, it would also explain why it took so long for life to evolve. Because of these reasons E is a better explanation than G, for though both have a lot of problems, G has some additional ones E hasn't got. Another explanation, F, posits the existence of fine-tuners. These are beings who are interested in life and who have the power to adjust physical constants. They are wise, but not all-knowing, powerful, but not omnipotent, and good-natured, but not all-loving. These beings are the ones who have fine-tuned our Universe so life could come into existence. This theory has less problems than both G and E, for the questions about the nature of God (omnipotent vs. omniscient, etc.) are avoided. Also, the Universe is not better adjusted to make life possible because the power of these beings is finite. F is a better explanation than G, showing how improbable G actually is.

Serious explanations that are far better than E, F and G, include the hypothesis that it is all just a huge coincidence, the hypothesis that the constants cannot be changed, the hypothesis that some values for the constants are much more probable than others, the hypothesis that there are many universes and even the hypothesis that the universe merely exists in our imagination (classical solipsism). These hypotheses all are better than G, so TAP can readily be discarded.

TAP fails to take 'probability densities' into account

TAP claims that since the physical constants could have so many values, the chance that they would be exactly as we see them is very small. However, this fails to take the notion of 'probability densities' into account. Though many constants may be possible, some might be much more likely to be selected than others. Suppose you were to roll two dice; you could get any number between two and twelve, but getting seven would be much more likely than getting twelve, since the outcome has a nonuniform probability density. TAP wants us to believe that the physical constants have a uniform probability density, but cannot show this is true. Maybe our constants were the only ones that could possibly obtain. Maybe they were extremely more likely than most others. I do not know if they are, but neither do the people who use TAP. We haven't a clue. But unless we know this, TAP is based on an unwarranted premise, that this probability density is more or less uniform.

More unknown probabilities

The calculations needed to give TAP a reasonable foundation are made even more problematic because for all we know not only the physical constants may differ from universe to universe, but the very laws of nature may too. We cannot calculate how likely or unlikely life is unless we know what the probabilities are that a universe comes into existence where, say, the gravitational force becomes stronger the farther you get away from a particle; or for a universe with five space and two time dimensions; or for a universe where addition is not commutative; or where pi is 2. Unless we can calculate the probability of these universes coming into existence and the probability of any kind of life coming into existence in such a universe, we cannot know if life is likely or unlikely. Thus, with our present knowledge, TAP's statement that life is unlikely is merely a guess.

TAP is based on a misconception of probability

Suppose that a huge number of universes were possible, and that when ours came into being one of these universes had to be selected. The probability that that universe would be ours is, if the probability density was more or less uniform, extremely small. Therefore, TAP concludes, there has to be an explanation for the fact that this happened. But this is false. Imagine a lottery with billions of participants. The chance that any individual wins the main prize is extremely small; yet the chance that someone wins the prize is 100 percent. Suppose the lottery is won by person A. TAP would now say that the chance that person A would win the lottery is vanishingly small; therefore, there must have been a plot to let him win. 'Without any evidence, God is accused of fixing the lottery.' [9] The fact that person A won is not enough to warrant an accusation; the fact that our Universe won the cosmic lottery is not enough to warrant accusing God.

TAP contains a logical error in its use of probabilities

One of the premisses of TAP is that, if a universe comes into existence naturally, the probability of it being 'fine-tuned' is very low. In logical notations this statement can be written as P(F|N)<<1, with N = the universe came about in a natural way, F = the universe is fine-tuned, and P(a|b) meaning: the probability that a is true, given that b is true. A probability of 1 is a certainty, a probability much lower than 1 (<<1) means that its happening is very unlikely. Then TAP claims that since this is so, it must also be so that, if a universe is 'fine-tuned', the probability of it coming into existence naturally is very low. But this is an entirely different statement that does not logically follow from the first one. This second statement is P(N|F)<<1, which is evidently different from the first. Is it good reasoning to exchange N and F in such a probability statement? 'This ... is an elementary if common blunder in probability theory. One cannot simply exchange the two arguments in a probability like P(F|N) and get a valid result.' [5] As an example, take F = 'I have a Royal Flush' and N = 'I will win this poker hand'. If F is true, N will almost certainly be true too, so P(N|F) = 1. But if N is true, F will most likely be false (since a Royal Flush is very unlikely), so P(F|N)<<1. It is thus shown that interchanging F and N is not a valid logical operation.

The logical error and universes

Supposing that the chance that a natural universe is fine-tuned is very small, how could it be that the chance that a fine-tuned universe is natural is still large? The chance that a universe is made by something supernatural (~N) is P(~N). The chance that it is made by something natural is P(N). Necessarily P(N) + P(~N) = 1, since a universe must be made either in a natural or a supernatural way. The chance that a natural universe is fine-tuned is P(F|N), the chance that a supernatural universe is fine-tuned is P(F|~N). Now the chance that any given universe is a fine-tuned, natural universe is P(N) P(F|N), the chance that it is natural times the chance that a natural universe is fine-tuned. The chance that any given universe is a fine-tuned, supernatural universe is P(~N) P(F|~N). Thus the chance that any given fine-tuned universe is natural is the first chance, divided by the first plus the second: P(N|F) = P(N) P(F|N) / ( P(~N) P(F|~N) + P(N) P(F|N) ). TAP tells us that P(N|F) is small because P(F|N) is small. However, we can now see that this is not true if P(N) is much larger than P(~N), or if P(F|~N) is small. Thus, if natural universes are more probable than supernatural ones, or if supernatural universes are not very probably fine-tuned, TAP breaks down completely. Evidently saying that P(~N) is not much less than P(N) is to presuppose that supernatural creation is quite likely; in fact, the existence of god is taken to be likely in order to show that it is likely! Also, a hidden premise is that if gods create universes, they necessarily create fine-tuned universes, as that is what a large P(F|~N) implies. Since it is very easy to imagine gods who would not do this, this premisse can be rejected too. We must conclude that TAP has no logical basis.

TAP leaves out relevant information: life exists

In order to make a valid conclusion one must take into account all information relevant to making that conclusion, and TAP fails to do that. We do not only know that our Universe is 'fine-tuned' (F), we also know that life exists (L). 'Therefore, it is invalid to make inferences about N if we fail to take into account the fact that L, as well as F, are already known to be true. It follows that any inferences about N must be conditioned upon both F and L.' [5] Now if life exists, and the Universe came about in a natural way, it must evidently be 'fine-tuned'. Thus P(F|N&L) = 1. For if life is not possible in a natural way in a natural universe, it will not exist.

In order to reach a pretty astonishing result, we must develop one line of probability logic. P(A|B&C) P(B|C) = P(A&B|C): the probability that, given C, A and B are true, is equal to the probabilty that, given C, B is true, times the probability that given B and C, A is true. This is fairly straightforward, but if you don't see it at once, please think about it for a while. It follows that: P(A|B&C) P(B|C) = P(A&B|C) = P(B|A&C) P(A|C), since in the middle equation there is no difference between A and B. Therefore it is true that P(A|B&C) = P(B|A&C) P(A|C) / P(B|C). We will use this rule.

The chance that, given a fine-tuned universe with life, that universe is natural is P(N|F&L). But this is equal to P(F|N&L) P(N|L) / P(F|L). Now P(F|N&L) is 1, as we saw above. Thus P(N|F&L) = P(N|L) / P(F|L). P(F|L) is a probability, and therefore must be a number between 0 and 1. So, P(N|L) is smaller or equal to P(N|L) / P(F|L). It follows, then, that P(N|F&L) >= P(N|L).

The chance that a universe in which we observe both life and fine-tuned constants is natural, is equal to or greater than the chance that a universe in which we observe life (and have not looked for fine-tuning) is natural. This result shows that seeing that our Universe is fine-tuned, in addition to the fact that there is life in it, can only increase the chance that it is natural! The fine-tuning argument should not be used by theists, but by atheists! [5]


The Theistic Anthropic Principle fails in many ways: it assumes that fine-tuning is possible though this is not proven; it is carbocentric and biocentric and thus fails to show that our Universe is special; it uses faulty analogies between things in the Universe and the Universe itself (which is the set of all things); it assumes that God is the best explanation for the Anthropic Coincidences, though this is not true; it fails to take probability densities into account, and thus cannot show that our Universe is unlikely; it is based on a 'lottery' misconception of probability and it uses wrong logic to come from 'A natural universe is unlikely to be fine-tuned' to 'A fine-tuned universe is unlikely to be natural'. Finally, it has been shown that observing fine-tuning can only increase the probability that our Universe is natural, and can never decrease it. I conclude that TAP is sufficiently refuted, and the last observation shows that it could even be used as an atheistic argument.

Graphic Rule

Bibliography & References:

1. Craig, William Lane "Barrow and Tipler on the Anthropic Principle vs. Divine Design"

2. Drange, Theodore M. "The Fine-Tuning Argument"

3. Hawking, Stephen W. *A Brief History of Time*

4. Hurben, Michael J. "On Universes and Firing Squads"

5. Ikeda, M. & Jefferys, B. "The Anthropic Principle Does Not Support Supernaturalism"

6. Kelly, Kyle "Is the Weak Anthropic Principle Compatible With Divine Design?"

7. Ross, Hugh "Design and the Anthropic Principle"

8. Smith, Quentin "The Anthropic Coincidences, Evil and the DisconWrmation of Theism"

9. Stenger, Victor J. "The Anthropic Coincidences: A Natural Explanation"

Graphic Rule