William H. Jefferys, "Review of The Privileged Planet" (2005)

The Privileged Planet: How Our Place in the Cosmos Is Designed for Discovery
> by Guillermo Gonzalez and Jay W. Richards<br> Washington DC: Regnery Publishing, 2004. 444 pages
> <p>

reviewed by William H Jefferys, The University of Texas at Austin

The Privileged Planet is based upon the odd notion that the more unsuitable our universe is for producing intelligent life, the more likely it is that our universe was "designed" to produce intelligent life by a "designer" of indeterminate nature; put another way, supposedly the less likely it is that there could be a planet in our universe that supports intelligent life, then the more likely it is that the universe was "designed" to produce a particular intelligent life form -- us -- that can and will investigate the nature of the universe.

We know from experience that this is not how human beings, the only intelligent designers of which we have any experience, work. We know that a human designer of a factory does not design a factory so that it will only occasionally, if ever, produce a car, or a computer, or whatever the target object is; rather the factory is designed to produce the largest possible amount of product consistent with the constraints: cost, physical reality, whatever.

The fundamental error made by Gonzalez and Richards, as with most creationists (including "intelligent design" [ID] creationists), is that they imagine that they can prove the existence of their "intelligent designer" by merely alleging evidence against a particular strawman naturalistic scenario, and, without clearly specifying an alternative model, simply assert that the only other explanation possible is that everything was created by a "designer". Under this strategy, no details are specified about what we would expect to see if the "designer" existed, or why we would expect to see that and not something else. It is, as we shall see, not a scientific theory. It is instead nothing but the usual fallacious Argument from False Dichotomy.

Of course, we know why ID creationists don't want to talk about the nature of the "designer". If they were to do so, they would undermine their claim that ID creationism has nothing to do with religion. They do admit the nature of their designer in private, among friends, but not before school boards or state boards of education. Since the real point of ID is to slip religion surreptitiously into the public school classroom, they can't reveal the true nature of their "designer" in any arguments intended for public consumption (as this book is). In line with this political strategy, the authors of this book are similarly cagey about the nature of the designer (p 330).

But they are between a rock and a hard place. Gonzalez and Richards don't realize that unless they can show that what we actually see is more probable, given that an "intelligent designer did it," they have no case. This is because a basic rule of inference is that one has to compare the likelihood of observing evidence E under all relevant hypotheses H₁, H₂, ..., H_n. Then the hypothesis that has the greatest likelihood is the one best supported by the evidence. Obviously, if you don't say what your hypothesis is -- in this case by specifically describing the nature of the "intelligent designer" and the consequences for the real world if that entity exists, so that actual calculations can be made -- then it is impossible to compute the likelihood of observing E under your hypothesis, and your hypothesis never even gets to the starting gate.

One wonders what Gonzalez and Richards would say if the evidence were otherwise. They talk about the fantastically small probability that our universe would give rise to intelligent, inquisitive life, but what if it were the opposite? What if we had observed that the universe was actually quite conducive to the existence of intelligent, inquisitive life? Would Gonzalez and Richards then conclude that the probability of observing such a universe, given that it was designed by an "intelligent designer", was small? I hardly think so. In such a case they would surely be pointing to the fecundity of the universe as evidence for the existence of their "intelligent designer". In other words, the assertion of a "designer" is a no-lose position. Whatever evidence one observed would by this fallacious reasoning support their "designer."

But there's the rub. They can't have it both ways. An elementary rule of inference is that if evidence E supports hypothesis H, then observing that E is false would undermine H. In other words, if observing that the universe is fecund were to support the hypothesis that the universe is "designed", then observing that it is not fecund would necessarily support the hypothesis that it was not "designed" and would undermine the design argument.

Unfortunately, it means that the ancient argument from design (of which this book is just a modern example) is scientifically useless. There is no conceivable evidence that could, even in principle, refute the notion that everything happens as a result of an unconstrained, very powerful "designer". This is because such an entity can be invoked to explain any evidence whatsoever. Real scientific hypotheses have to be vulnerable to evidence. It must be possible to imagine evidence that would undermine them (see Pennock 1999, ch 6, for an extensive discussion). This is not the case for a mysterious "intelligent designer" of nature so unspecified that one cannot even make predictions about what one would expect to observe if it existed.

Consider, for example, the fine-tuning argument: The fact that "the constants are right" for our own existence is supposed to support the existence of an intelligent designer. Philosopher of science Elliott Sober (2003) has refuted this argument and, independently, Michael Ikeda and I have made similar points with some variations (Ikeda and Jefferys 1997). Sober points out that the usual design argument is that the probability that the "constants are right," given that design is true, is greater than the probability that "the constants are right," given a naturalistic universe. Notwithstanding the fact that we don't know whether this inequality is true or not in the ID creationist view -- because the ID community stubbornly refuses to specify the nature of the "designer" so that we can actually do the required calculations -- there is a deeper problem.

Sober and Ikeda and I pointed out that the relationship fails to take into account our own existence. In other words, we are here (we know this, and could not be making any arguments if it were not so), so any discussion must take this fact into account. Thus, the correct comparison is between (A) the probability that "the constants are right" given design and our own existence, versus (B) the probability that "the constants are right" given a naturalistic universe and our own existence. Since in a naturalistic universe our own existence implies that the constants must be right, this means that (B) is equal to 1. What about (A)? Clearly, since probabilities are always less than or equal to 1, (A) cannot be larger than 1, so the ratio of (B) to (A) must be at least 1. This means that observing that "the constants are right" cannot undermine the naturalistic hypothesis.

Sober says that (A) is also 1, but here he missed an important point. Since the nature of the designer is unspecified, and might be an omnipotent deity, for example, it would be possible for the designer to produce universes where the constants are not right, but in which we could still exist.

An example would be a universe where the constants are not right for producing carbon in stellar interiors. In their book, Gonzalez and Richards mention Fred Hoyle's remarkable 1954 prediction of special resonances in carbon and oxygen nuclei (p 198 and following). These resonances were predicted because without them, carbon and oxygen could not be synthesized in stars, and since they also cannot be synthesized by the Big Bang, our own existence implies that the resonances must exist, at least if the universe is naturalistic. This in turn leads to rather narrow predicted ranges for certain physical constants ("the constants are right"). Indeed, the resonances were found to exist, one of the earliest and possibly best examples of a prediction of a physical fact from the so-called weak anthropic principle, that sentient beings ought to observe that the universe they inhabit is consistent with their own existence.

But, if the universe had been designed by a sufficiently powerful designer, the constants would not have to be right in order for us to exist. For example, the designer could create a universe where the constants are not right for the production of carbon and oxygen in the interiors of stars, preferring instead (for whatever reason: whim, or the desire to accomplish other goals such as letting us know that he exists by means of a subtle scientific clue) just to manufacture the required carbon atoms and sprinkle them where needed throughout the universe. If we consider the possible existence of such a designer -- and remember, the ID creationists' intentional refusal to specify the nature of their designer leaves this possibility open -- then it is no longer the case (as Sober asserts) that (A) is equal to 1. Indeed, it is less than 1 and could be quite small, which means that our observing that "the constants are right" actually provides powerful evidence in favor of the naturalistic hypothesis. It would actually be our observing that "the constants are wrong" that would undermine, and in fact refute the naturalistic hypothesis. The ID creationists have the inequality backwards.

In another section, Gonzalez and Richards also attempt to refute the so-called "Many Worlds Hypothesis" (MWH), which postulates the existence of a very large or even actually infinite collection of universes called the multiverse (p 268 and following). I should first point out that they are simply wrong to think that the motivation for the MWH is to get around the fine-tuning problem. In fact, it is a consequence of the leading theory of cosmology -- the theory of chaotic inflation -- which is the theory best supported by the evidence (including that from the recent Wilson Microwave Anisotropy Probe, or WMAP). Chaotic inflation was invented to explain certain observed facts about our universe, for example its flatness and homogeneity. One consequence of inflation is that the universe is actually infinite in extent both in space and time, containing infinitely many regions that have each inflated into expanding universes much like ours, but perhaps with physical constants different from ours. Indeed, this multiverse is so vast that it would contain infinitely many universes exactly like ours, as well as infinitely many others that differ from ours in only subtle ways, for example ones in which I am an ID creationist and the authors are attempting to refute my pro-ID arguments, or ones where I have a long green tail, or ones in which a particular gene in my genotype has a C substituted for an A (see Seife 2004 for more on this).

Gonzalez and Richards's "refutation" of the MWH is unconvincing. It consists of a bland dismissal that an actual infinite set can exist (p 268 -- where did they learn their mathematics?) together with a claim that "we have no evidence to think that other universes exist," a claim that happens to be false, for several reasons. One reason is that it is a prediction of the best-supported theory in cosmology, one that is strongly supported by evidence. And the second is that under that model, our own existence evidentially supports the MWH (since under that hypothesis a selection effect is involved: we can only exist in one of the very small proportion of worlds in which "the constants are right," so our own existence implies the existence of these other worlds).

As Mark Perakh (2004) has pointed out in another context, there is nothing particularly unparsimonious about the multiverse hypothesis. For one thing, it is based on the observational fact that our own universe definitely exists, and since it does exist, it is reasonable to presume that naturalistic processes would produce other universes, just as different versions of our own. If physics can produce one universe, there is nothing in principle to prevent it from producing infinitely many. Indeed, it would be expected. By contrast, the hypothesis of an intelligent designer of universes is completely speculative; there is, as Perakh points out, not a single observational fact that points to the existence of such an entity other than ancient, conflicting legends.

In their discussion of the MWH, Gonzalez and Richards also repeat a fallacious argument (p 270) that has been made by John Leslie, concerning a hypothetical officer who survives a Nazi firing squad and concludes that this must be due to design (the firing squad intended to miss) rather than chance (the firing squad members all missed by accident). We are supposed to reason by analogy that since the officer concludes that design rather than chance was the reason for this particular low-probability event, we should infer the same as regards the universe. Notwithstanding the obvious differences between naturalistic universes that have no known intentions, alleged "designers" whose intentions cannot be clearly specified without undermining the political aspirations of ID creationists, and firing squads that have well-understood intentions, this argument is plain silly, and has been decisively refuted in Sober's paper (2003). Analogies are treacherous things.

Finally, I turn to Gonzalez and Richards's notion that our earth is uniquely designed for its inhabitants to do scientific exploration, and that the universe is similarly designed for us to do that scientific exploration. They point to a number of phenomena that have aided our scientific enterprise, such as the transparency of the earth's atmosphere, the fact that we have a moon that is just far enough from the earth to produce spectacular solar eclipses, and so on. Of all the arguments in the book, I find this the weakest. It puts the cart before the horse. For suppose it were not so; if we existed on another world very different from the earth, then we would surely be doing something. We would be doing whatever was possible for us to do under the circumstances in which we found ourselves. If we accepted the Whiggish reasoning of the authors, we would be just as justified in concluding that our planet -- and our universe, if we could see it in this alternative reality -- was designed so that we would do whatever we happened to be doing at the time or find interesting at the time (as diverse human cultures have always done). The authors could learn much by studying a little anthropology and a little history.

To summarize, the little that is new in this book isn't interesting, and what is old is just old-hat creationism in a new, modern-looking astronomical costume. It is the same old shell game. It's too bad that Guillermo Gonzalez (whom I know from his tenure as a postdoctoral fellow in the University of Texas's Astronomy Department) has allowed himself to be sucked in as an advocate for this ancient argument. The Argument from Design is 200 years old, if not older, and it has not improved with age. It hasn't resulted in any new knowledge in all of those years. Modern astronomy is constantly producing new knowledge and understanding of the universe. Guillermo is a promising young astrophysicist, and I hope that he doesn't throw away his career on such nonsense.

References

Ikeda M, Jefferys WH. 1997. The anthropic principle does not support supernaturalism. Available on-line at http://bayesrules.net/anthropic.html; last accessed January 4, 2005.

Pennock RT. 1999. Tower of Babel: The Evidence against the New Creationism. Cambridge (MA): MIT Press.

Perakh M. 2004. Paul Davies: Emergentist vs. reductionist. Available on-line at http://www.talkreason.org/articles/Davies.cfm; last accessed January 4, 2005.

Seife C. 2004 Jul 23. Physics enters the twilight zone. Science 305 (5683): 464-6.

Sober E. 2003. The design argument. In: Manson NA, editor. God and Design: The Teleological Argument and Modern Science. New York: Routledge. p 27-54. Also available on-line at http://philosophy.wisc.edu/sober/design%20argument%2011%202004.pdf; last accessed January 4, 2005.

About the author
> William H. Jefferys is the Harlan J. Smith Centennial Professor of Astronomy, Emeritus, at the University of Texas at Austin.<br>

Author's address
> William H. Jefferys<br> Department of Astronomy
> The University of Texas at Austin<br> Austin TX 78712-1020

[Forthcoming in Reports of the National Center for Science Education.]