When Winfrith of Crediton, not many miles from the Hump’s rural seat, went to Germany and cut down a celebrated pagan sacred oak around 723 AD, the lack of any resulting thunderbolt from Thor destroyed the entire basis of pagan belief. With the tree gone, there was literally nothing left.
I cite these papers at my peril, I guess, as the authors are proponents of Intelligent Design and the papers come from a symposium held at Cornell University on Biological Information which has attracted the usual scorn simply for having dared to happen.
But before briefly looking at what they say, I must comment that they only build on mainstream work that has already raised all the same issues. Two years ago I reviewed Eugene Koonin’s 2009 evolution overview paper, which downgraded natural selection mainly to an inefficient purifying machanism, with little adaptive role on the myriad of near-neutral mutations that swamp the genome. The clear implication is that poorly controlled random mutation, not natural selection, is the main driver of evolution.
Our authors, whilst not referencing Koonin, do cite a number of original research papers (such as those by Lynch) that tell the same story. They give the kind of detailed analysis of natural selection that I miss so much in the kind of articles and discussion on BioLogos, where the same plausible but vague arguments that Darwin used remain the norm. It’s obvious that a mutation that causes a purple lizard to become green and well-camouflaged will probably be selected over its purple peers – ergo multiply everything up to include every major or minor feature of every living thing and life’s wonders are simply explained.
Yet even a simple thought experiment shows this to be an oversimplification. Suppose the green gene also happened to give slightly poor vision and lower reproductive rates. The overwhelming camouflage issue would still ensure this strain survived preferentially, but general fitness would have decreased.
What our authors do that is new is to use a piece of software designed to simulate chosen levels of factors accepted to be significant in population genetics – population size, mutation rate and so on. In particular, they take a much more detailed look at the business of near-neutral mutations. It is now accepted that the vast majority of mutations are very minor, or “near-neutral”. At first they were called “neutral”, but those in the field like Kimura soon realised that true neutrality is rare – some mutations are slightly disadvantageous, and some slightly beneficial. But the former are in the vast majority – in fact beneficial mutations overall are so rarely found in nature that they cannot be accurately measured.
The important point is that below a certain threshold of “harm” or “benefit”, natural selection is completely blind to these mutations, which must therefore accumulate in the genome – and hence we have “neutral theory”. The “Mendel’s Accountant” software (which you can download should you have the background to use it) puts figures on all these parameters and calculates their effects.
The result is that, with any realistic parameters, natural selection proves totally unable even to keep the genome from progressive loss of fitness by the accumulation of slightly deleterious mutations. At some stage, extinction will result.
One old solution to this dilemma (which has been recognised for a long time) is to suggest that occasional significantly beneficial mutations compensate for this process of degradation. In fact, the simulations show the opposite – a bit like my green lizards, the selective advantage of moderately beneficial mutations actually increases the build-up of mildly deleterious mutations and exacerbates the overall loss of fitness.
Since Darwin’s time, it has been suggested that natural selection may be more of a stabilising force than an innovating one – here it seems it may actually be destructive. Clearly it is not a simple concept at all.
The authors, in their “possible criticisms” section, anticipate the accusation that their software was written just to provide such negative results. Such defensive writing is only necessary for political, rather than scientific reasons – critics are actually in a position to evaluate the program themselves, just as Ewart, Dembski and Marks did with the AVIDA program. In fact, as the authors point out, even AVIDA gives comparable results to Mendel’s Accountant when realistic parameters are used.
But I’ve focused on these two papers only because they quantify the problem so clearly – Koonin’s paper, papers referenced by the two articles themselves and related papers like this or this or this linked from those citations all tell a similar story. And that is that natural selection, as it is currently understood, seems unable to produce the infinitesimally gradual adaptive change posited by Darwin. It can only respond adaptively to quite major beneficial mutations (if they can be shown to occur), and at a rate too low to account for the vast variety of exquisite tailoring to the environment that is universally seen in nature.
The significance of this is quite obvious. In all discussions of why evolution is not simply pot luck against hyperastronomical odds, the quasi-intelligent design role of natural selection is the only conventional defence. Darwin’s Origin of Species sometimes personified selection as a benign overseer constantly watching, like a super-livestock breeder, to ensure progess towards perfection. For that Wallace criticised him and suggested the term “survival of the fittest” as avoiding the pretence of teleology, though later firmly embracing teleology himself, of course.
But however it’s worded, the sidelining of natural selection to an occasional adaptive tool and an imperfect purifier of the worst errors removes all the power of evolution to explain “endless forms most beautiful” without teleology. It is hard to conceive of any other process that would not be fundamentally based on some internal, or external, sense of purpose.
It’s for that ideological reason alone, it seems to me, that those who raise such issues will be quietly ignored, if they are mainstream, or loudly anathematized, if they have an ID background. What I’ve not seen is any detailed refutation of the evidence and its implications.