Book Review: God's Undertaker

It's easy to view creationists through the lens of caricature. From Dr Kent's belief that dinosaurs died out due to spontaneous nasal combustion, to the belief that the 300 mile Grand Canyon formed in about five minutes, we can find instances of claims that would have to work hard to be considered merely absurd. But still, it is perhaps unfair to judge a group based on the worst of its members. Though I have yet to hear good creationist arguments I do know smart creationists. With that in mind I was quite happy to receive a recommendation to read Professor John Lennox's book on the subject, God's Undertaker.

I went in with high hopes - my initial readings on evolution were prompted back in 2003, when I studied evolutionary computation. Lennox's background in mathematics would surely mean he could speak with authority on this area.

He was born on the same island as I, a coincidence of birth which admittedly does not make him more likely to be correct, but caused me to warm to him nonetheless.

He's also a talented linguist, speaking Russian, German, Spanish and French in addition to his native English. He has been published and has given lectures in many languages.


Why are languages important?

Some feel evolution cannot be true because it offends their interpretation of Genesis. Yet Genesis does not confine itself to writings on the diversity of life we see - it also gives extensive time to the diversity of languages we speak. For those who have not read Genesis 11 recently I will summarise - originally, after the flood, everyone spoke only one language. The Tower of Babel was constructed, so high it reached the heavens, then God came down, created separate languages, scattered the people to separate lands and confounded the common tongue.

Yet the modern, unchallenged explanation for the diversity of languages we encounter is that a gradual process of change led to new languages diverging slowly from an original base. It is not controversial in any circles that I know of to say that Spanish and Italian share a common ancestor in Latin. It is also quite obvious that no Latin speaking mother gave birth to a Spanish speaking son with whom she could not communicate; the process is incremental and not visible from generation to generation. True, this is not evolution but the comparison is apt - the accepted theory disagrees profoundly and irreconcilably with a creationist literal interpretation of Genesis. Given that he is not a proponent of intelligent etymology, I hoped his objections to evolution found their motivations elsewhere.

With that I started.

I think this book will seem very convincing to those who are unwilling to expend an inordinate amount of time doing further research. He writes with authority and gives a spirited defence of theistic evolution, but then some cracks begin to appear. I enjoy an occasional science book but there are gaps in my knowledge and when I read a tome that covers biology, abiogenesis, cosmology, theology, probability and computing I have to put a certain amount of trust in the honesty of the author. Luckily I can choose the areas with which I'm most familiar as a benchmark of the quality of the author's work.

Take the below quotes on Anthony Flew to warm up:

"‘My whole life has been guided by the principle of Plato’s Socrates,’ writes Anthony Flew, in connection with his recent turning from atheism to theism."

"...very recently philosopher Anthony Flew gave as the reason for his conversion to theism after over 50 years of atheism that biologists’ investigation of DNA..."

(Emphasis mine.)


These quotes would support Lennox's case if they were true. They are not - Anthony Flew is now a deist. Deism and creationism are mutually incompatible, which could explain Lennox's repeated slip of the keyboard.

Later, Lennox defines a transitional form:

"An intermediate form would only be transitional if it could be shown to have descended from A and was an ancestor of B. To establish those relationships, of course, some mechanism would have to be exhibited that was demonstrably adequate for the task."

Having thus defined them out of existence he goes on to bemoan their absence.

Read the below description of Dawkins' weasel program, a small demonstration of how randomly generated text can be mutated, then the offspring selected for how closely they match the text "METHINKS IT IS LIKE A WEASEL":

"Each time a monkey hits a letter, the letter it types is compared with its target letter – a highly non-random process. This comparison, of course, has to be done by some mechanism, a computer (or by a Head Monkey, as mathematician David Berlinski delightfully suggests). If the monkey has typed its target letter the comparison mechanism retains that letter – another highly non-random process – and the monkey stops typing, its job done."

Again, a strong point, casting Dawkins as dishonest, misleading, and bungling. Again it suffers from being flat-out wrong. For example:

"If the monkey has typed its target letter the comparison mechanism retains that letter – another highly non-random process – and the monkey stops typing, its job done."

This is not true - all characters are free to mutate at every copy event. Coders will find source code here in many languages. Non-coders will find a screenshot countering this claim here. If you'd like to see footage, there is a video here. (Skip to 5:30.)

I found Lennox's criticism of genetic algorithms all the more absurd when I recalled that a few chapters earlier he'd used their efficacy in defence of theistic evolution:

"...computer-implemented genetic algorithms are routinely used for sophisticated engineering optimization purposes – for example, to construct the best possible shape for an aircraft wing. It would be absurd to suggest that the fact that these evolutionary algorithmic optimization processes are themselves blind and automatic constitutes an argument that they do not have an intelligent origin."

If Lennox is confident that blind, automatic, goalless algorithms producing randomly generated designs, breeding with mutation and subjecting output to a fitness function is a sound method for producing the wings of the aircraft within which he regularly flies, and a method of greater efficacy than human intelligence alone, then where is his argument?

Let us leave the computer lab for the biology lab. When discussing the evolution of bacteria, Lennox writes:

"More recent work on the E. coli bacterium backs this up. In this research no real innovative changes were observed through 25,000 generations of E. coli bacteria. Biochemist Michael Behe points out that now more than 30,000 generations of E. coli have been studied, equivalent to about a million human years, and the net result is that evolution has produced: ‘Mostly devolution. Although some marginal details of some systems have changed during that thirty thousand generations, the bacterium has repeatedly thrown away chunks of its genetic patrimony, including the ability to make some of the building blocks of RNA. Apparently throwing away sophisticated but costly molecular machinery saves the bacterium energy. Nothing of remotely similar elegance has been built. The lesson of E. coli is that it’s easier for evolution to break things than to make things.’ - Lennox, 2009

Lennox fails to mention that around generation 31,500, just outside the bounds of his self-imposed generation restriction, there was a major, observed evolutionary leap - the bacteria evolved the ability to metabolise citrate. Later work on frozen samples revealed this real, innovative change began in generation 20,000.

This is the equivalent of saying that in thirty years I showed no evidence of producing blog posts. (I'm 31 at time of writing.)

What am I to make of this? Is he intentionally misleading? If so it is hard to take his other claims on faith. Is he merely behind on his research, or straying too far from his core competencies? Perhaps, but if so, what else has he misunderstood?

Addendum - feel free to  leave a comment on my Amazon review.