In this book, John Lennox (updating an earlier work) attempts to address the age-old question of how one can honestly hold a serious religious belief in the age of modern science. Lennox starts out by discussing the scope and limits of science, including the limits of reductionism. This is followed by a discussion of some intriguing developments in cosmology, which may suggest that ours is a "designer universe".
Lennox then addresses that old bugaboo evolution. He first treads a number of well-worn paths, including (yes) Paley's example of the watch and a number of philosophical points as well.
It is at this point (Chapter 6: The Nature and Scope of Evolution) that Lennox's exposition breaks down. Here he quotes Michael Behe, the Intelligent Design scholar who has made a big deal of "irreducible complexity", and William Dembski, the ID scholar who has argued from information theory and probability. In one section, entitled "What say the mathematicians", for instance, Lennox claims that distinguished mathematicians find evolution to be fantastically improbable. This section is followed by a discussion of the fossil record, where Lennox claims that "punctuated equilibria" is in "complete contrast" to the gradualist approach of "ultra-Darwinians".
In Chapter 7, Lennox calculates, for instance, that the chances of getting some protein right involving 100 amino acids is one part in 10^(130), and the chances of getting all proteins right is one chance in 10^(40,000) -- ie far far too small to be taken seriously. This material is followed by discussions of information content and complexity, and, in conclusion, denies that Dawkins' "Mount Improbable" is climbable.
Sadly, all of these probability and information theory arguments are fallacious. The principal fallacy is to presume that a protein, for instance, can only form by an all-at-once chance assembly of a large number of amino acids or other components. But note that this is the *creationist*, not the scientific theory, of the formation of these proteins -- scientists hold that they are the end product of a very long process involving many steps that were useful in earlier contexts. Probability calculations that ignore the process by which the structure came to be are *not* meaningful and may easily deceive. Consider, for instance, the fact that Lennox's analysis applies equally well to snowflakes -- probability calculations applied to these intricate, symmetrical structures yield even more remote probabilities than arise in analysis of biomolecules, yet no one seriously suggests that supernatural action is required to explain their existence. Instead, they form over a period a several minutes, with each step governed by natural law.
Such errors would be merely amusing if not for the fact that thousands of readers, who are not well versed in the potential fallacies of probability calculations, are likely to read and be impressed by this book. Religious believers are *not* well served by fallacious arguments. As Paul warned the Corinthians, "For if a trumpet give an uncertain sound, who shall make himself ready to battle?" (1 Cor 14:8).
Also, one can ask why Lennox feels he must "prove" God by means of "scientific" arguments. Is faith no longer necessary in religion? After all, the notion that religion can and should be placed under the microscope is curiously in full concordance with the assertions of persons at the other extreme of the intellectual spectrum -- namely scientists such as Dawkins who claim that science proves religion to be utterly false and vain. Lennox would have done much better to have contested this (false) dichotomy, rather than to affirm it and then trying (in vain) to "prove" religion.