I frequently get asked about the argument for evolution from the similarities observed in the genetic makeup of living things. Recently, the specific question was what to make of Richard Dawkins' claim that the heirarchies of similarities in the genetic record of living things provide 'undeniable proof of evolution.' Below, in part, is my answer.
The heirarchy of similarity among living things is a neutral fact and not in any way supportive of evolution or common ancestry. The understanding that Darwin sought to replace was a view known as typology. Typology said that the heirarchies of similarities we see among living things (extant and extinct) are grounded in necessity. Bats and rodents (on one level) share a suite of (mammalian) characteristics that sets them apart from all birds. On another level, bats share a smaller suite of characteristics that sets them apart from rodents. And these distinctions are adaptive and holistic. What we expect to see when we look at each part of a bat (and not just anatomically but also physiologically, behaviorally, and such) is that each contributes to bat-hood, that the membranous wings, the echolocation ability, the reproductive strategy, and such are all part and parcel of what it means to be a bat. The creature is well-designed for its role and niche. Moreover (on this view--which still remains the most reasonable view), the reason we don't see creatures that are half bat and half rodent (Darwin's predicted but yet undiscovered transitional forms) is because such a creature is non-functional and nonsensical.
This typological (and design-oriented) view did (and does) a very good job of accounting for the various levels (heirarchies) of similarity among living things. Before Darwin, biologists recognized that all living things were made of the same elements. In fact, Scripture clealy claims this, indicating that humans (in Gen. 2:7) and other animals (in Gen. 2:19) are alike made of "the dust of the Earth." This would simply be the best way of conveying to the ancient Hebrew mind that the physical components (the elements) of which all living things are made are the same elements as are found in the abiotic portions of the Earth. So the discovery (since Darwin's time) of the further similarities among living things at the level of proteins and (more basically) DNA does not in any way distinguish between the competing alternatives of theistic design (typology) and naturalistic evolution. In this regard, it is (at best) disingenuous of modern evolutionists to appeal to similarities among living things as evidence for their view. And this is especially true since what evolution was meant to explain--but has singularly failed to explain--was not the similarities but the differences.
We now know, for example, that there is more similarity in the DNA of humans and chimps than even evolutionists expected. What does this tell us? It tells us that relatively minor differences in DNA do not explain why chimpanzees are (like every other species of life on Earth) naked animals surviving from day to day in loose extended family groups while humans are civilized, uber-intelligent animals able to exploit every aspect of the Earth and to explore even the distant reaches of the universe. In fact, these deep DNA similarities should lead us to recognize that no strictly materialist explanation will ever satisfactorally account for the vast differences between chimps and humans, that there is something non-material going on here, and that scientific naturalism is false.
But perhaps I should be more to the point. Richard Dawkins finds in the DNA heirarchies common ancestry and evolution. But he finds these things not in the fact of the heirarchies but in the interpretive assumptions that he brings to them. In this way, his argument is (as all arguments from similarity are) circular. He begins by (wrongly) assuming that any similarities can be construed as evidence for common ancestry, and then when he perceives similarities, he's proved his assumptions. This, as any logician could tell you, is a fallacious way of reasoning. Simply put, taken at face value (that is, without beginning with biased assumptions), the discovery that living things exhibit a heirarchy of DNA similarities is just as (or more) amenable to a common-design inference as to a common-ancestry conclusion.
This was (inadvertently) illustrated many years ago by evolutionist Tim Berry. Frustrated by creationists' inability to understand evolution's dogma of "descent with modification," he asked his readers to picture a series of Corvettes. We see that the '56 Corvette is slightly different than the '55, and that the '57 is slightly modified from the '56, and so on. Unfortunately (for him and other evolutionists) the illustration shows just the opposite of what he wanted it to show, because we all know that each of those Corvettes was separately designed and manufactured, that the differences between them did not arise through their passing on, one to the other, slight variations.
All arguments from similarity are fallacious because they involve arguing in a vicious circle. And yet, this is really all that can be offered as support for evolution, whether of the atheistic form of Richard Dawkins or the theistic version of Francis Collins and others,