26.5 Evidence

a.

neo-Darwinian theory: random mutations, genetic drift and recombination, natural selection,

b.

theistically filtered evolution: random mutations, genetic drift and recombination, natural and theistic selection,

c.

theistically driven evolution: preselected and random mutations, genetic drift and recombination, natural and theistic selection.

Detailed predictions for the theistic selection theory B will depend on advances in theistic science, especially concerning the manner in which variable theistic influx influences the behavior of organisms and concerning the manner in which partial ‘images’ of the human form may function mentally and physiologically. The mental aspect of this functioning will be discussed further in Chapter 27.

The detailed predictions for the theistic mutations theory C are more difficult to make. They depend God’s ‘management policy’ for biological evolution. Probably, we will have to examine and make predictions for specific limiting cases for this policy, such as mutations confined to the DNA sequence or mutations concerning the gene expression mechanisms or mutations affecting other epi-genetic control mechanisms of cells.

Since the theistic theories are fully-fledged causal theories of evolution and therefore are not arbitrarily disqualified before evaluation, the questions to be answered are the relative probabilities of the three theories in light of the empirical evidence available. According to Sober (2002) and Sober (2009), this assessment should use Bayesian arguments in probability theory, using what is called the Law of Likelihood:

Observation O favors hypothesis H₁ over hypothesis H₂ precisely when
Pr(O | H₁) > Pr(O | H₂). And the strength of the favoring relation is to be measured by the likelihood ratio Pr(O | H₁) / Pr(O | H₂).

In order to compare the three hypotheses A, B and C above, we need to determine the three conditional probabilities p_A = Pr(O,H_A), p_B = Pr(O,H_B), and p_C = Pr(O,H_C), where O denotes the sum of the empirical evidence concerning the history of life. We determine whether either of the theistic hypotheses B or C is favored over the neo-Darwinian theory A by seeing whether either of the ratios p_B/p_A or p_C/p_A is larger than unity.

Most neo-Darwinists say that if evolution were repeated, it is very unlikely that humans as we know them would be produced again. They agree that the probability p_A of producing rational humans is extremely low because of the occurrence of the many fortuitous mutations, drifts and selections that occur without any oversight. Given Darwin’s theory A, therefore, we insist that the occurrence of human life, though not impossible, is still clearly an extraordinary event and hence one that must require an extraordinary explanation.

Our first estimates of these ratios p_B/p_A and p_C/p_A place them much larger than unity since the theistic additions in B and C compared with Darwin’s theory A do indeed make the likelihood of life on earth much greater than that predicted by theory A. In this case, both the theistic theories are judged to be favored over Darwinian theory. I will then be reminded, and admit, that this requires an independent and robust prediction of ‘theistic selection’ on independent grounds and not one based merely on observations of existing forms. Otherwise the prediction of the success of human forms would be entirely circular. Making such independent predictions is the task of theistic science.

A more detailed likelihood analysis should also consider the relative ‘prior probabilities’ that enter into Bayesian arguments of likelihood. I concede that the prior probability for theism is less than for any naturalistic account of the universe and may well be as low as 10%, 1%, or even lower, at least in the eyes of modern-day scientists. Such prior probabilities should be taken into account, therefore, in any full calculation of the likelihood of theistically-filtered or theistically-driven evolution in relation to observations and also in comparison with neo-Darwinian theories.