This book opens new vistas in the science and religion dialogue by awakening both sides of the aisle to the consistency and limits of materialistic naturalism. Elliott Sober is a gifted thinker and writer who has made important contributions, singly or with coauthors, to the clarification of many debates within evolutionary biology and the philosophy of science as it pertains to this field. He has a profound understanding of Darwin's writings (across editions of Origin and other works) and his methods of reasoning. Did Darwin consists of four chapter‐length essays: the first holding the same title as the book; the second “Darwin and Group Selection” is followed by “Sex Ratio Theory—Darwin, Before, and After.” I will focus on the fourth chapter because it seems to me to break important new ground. The book concludes with a lengthy postscript that provides a more technical exploration of several topics considered earlier in the book.

Chapter 4 “Darwin and Naturalism” is an elegant and very pointed discussion about what naturalism can and cannot claim. Specifically Sober shows that evolutionary theory, as it stands, cannot rule out supernatural activity, nor is evolutionary biology damaged by this recognition. Several quotes will give a clear sense of the nature of his argument as well as indicate the concise analyses that characterize this important book. All italicized words in what follows are italicized in the original; the ellipses are the reviewer's. Let us begin (passing over much that is very rich indeed) with the following (134):

What I want to consider … is the view that God supplements what happens in the evolutionary process without violating any laws. An intervention, as I'll understand the term, is a cause: it can trigger an event or sustain a process. Physicians do both when they intervene in the lives of their patients. Physician intervention does not entail any breakage in the laws of nature; neither does God's.

Endnote 14 sharpens Sober's stance, lest we misunderstand (201):

I therefore disagree with Pennock's (1999, 195) statement that ‘to say that some power is supernatural is, by definition, to say that it can violate natural laws.’ The ability to violate natural laws might be a defining property of an omnipotent deity, but it isn't part of the definition of a supernatural power.

Sober drives home this line of reasoning in Section 4.3 “Why evolutionary theory does not rule out an intervening God.” He writes (134–35):

Consider the thesis that there is more going on in the evolutionary process than is dreamt of in evolutionary biology…. If evolutionary theory were causally complete, there would be no room for this idea. However, we have no assurance that the theory covers all the facts that are causally relevant to what happens in evolution. Please note that I am not saying that there is evidence that such hidden variables exist; my claim is only that they are not ruled out by current theory. The case for thinking that evolutionary theory does not preclude the existence of hidden variables begins with the simple fact that the theory is probabilistic.

Later in the same section, in consideration of the fact that “mutations do not occur because they would be useful to the organism” (136), Sober carefully develops the role of probability of various mutations and states (137):

The hypothesis that the different mutations have the same probability in different environments does not rule out the possibility that there are hidden variables; perhaps each mutation that occurs is the result of its own suite of deterministic causes. If the data does not rule out hidden variables, they also do not rule out supernatural hidden variables.

Sober's reflections continue to build on this theme. He goes on to ask in Section 4.4, “Should scientific theories talk only about what exists in nature?” He begins by noting that some scientists and philosophers assert that we should accept methodological naturalism because (139) “science, by definition, eschews discussion of the supernatural.” He continues (139–40):

If the point is put by saying that natural science asserts nothing about the supernatural [sic], the claim sounds like it must be true. In fact, this definitional ploy accomplishes nothing…. A substantive reason is needed for thinking that methodological naturalism provides good advice for inquiry…. It is interesting to note, in this connection, that science does not, as a matter of fact, avoid postu‐lating…. a different sort of supernatural entity — numbers. Evolutionary theory entails that numbers exist, and numbers are supernatural entities…. Mathematical Platonism says that number and other mathematical objects exist but do not have spatiotemporal location…. Consider the claim that there are infinitely many prime numbers. This is a true statement, as any number theorist will tell you. But what are these things called numbers? What must they be like for this statement to be true? First, it is important not to confuse numbers and numerals; numerals are names for numbers. The statement about primes isn't about names; it's about the things those names name. The statement would still be true if there were no language users, and hence no names for the numbers. Indeed, the statement would still be true if there were no matter in the universe. This is what leads Platonists to claim that numbers are supernatural entities…. What has this to do with evolutionary theory? The answer is that many statements in mathematized evolutionary theory entail that numbers exist. Scientists hardly notice that their models have this implication, but such models are everywhere.

But now, Sober asks the converse question (141), namely, “Does accepting metaphysical naturalism oblige one to accept the methodological thesis?” His argument is profound and stimulating:

At first glance, the answer seems to be yes, if there are no supernatural entities, a true scientific theory cannot claim that such things exist. If the goal of science is to find true theories, then scientific theories should not assert that supernatural entities exist. The problem with this argument is that science needs mathematics, whether or not Platonic entities exist. If numbers do not exist, then mathematics is a useful fiction; indeed, it is an indispensable fiction. Scientific theories should include mathematical statements that entail that numbers exist whether or not this existence claim is correct…. My claim is that science would still be capable of explaining and predicting what we observe, and doing all the wonderful things that scientists prize, even if mathematical Platonism were true. The success of the scientific enterprise does not in any way depend on its eschewing mention of all supernatural entities. In fact, the success of science requires that science postulates some supernatural entities, if mathematical Platonism is true.

These observations and others throughout the book are extremely valuable in grasping what evolutionary biology qua science can and cannot say. Sober's development casts no aspersions on either science or religion. It does suggest that we would not be able to tell the difference between a DNA alteration caused by a cosmic ray or by the finger of God or by the finger of God directing a cosmic ray; we simply have no ways of distinguishing (see Figure 4.1, 133).

Lest Sober be misunderstood by virtue of the quotes I chose, it is important to note that he also clearly outlines the tremendous weaknesses of creationism—the primary one being that it fails to produce hypotheses testable by data that can be gathered. The first four chapters are eminently readable by anyone willing to engage; thinking with Sober is much easier than Thinking with Whitehead (Isabelle Stengers, 2011, Harvard University Press). Sober's book makes important and (for this reviewer) surprising and necessary contributions to the understanding of the strengths and limits of evolutionary biology. The answer to the book's title question is at the end of the first chapter.