Defender's Guide to Science and Creationism

Mark Vuletic


The genetic distance between bacteria and humans is the same as that between bacteria and plants.


One of the many striking pieces of evidence for evolution is the way that molecular phylogenies based upon neutral mutations generally match evolutionary expectations based upon the fossil record and morphological and biogeographical analysis. While there is no reason to expect a creator to specifically arrange useless mutations to match evolutionary predictions, it is a key prediction of evolutionary theory. However, some creationists actually argue that the evidence refutes evolution. Why?

One of the reasons is because of anomalies: creationists seize upon occasional anomalous results within the realm of expected error, and portray them either as representative of the results in general, or as somehow fatal to evolutionary theory. The error with such reasoning is readily evident. However, more important is the following: suppose we are about to measure the genetic distance from protozoa to redwood trees, fruit flies, and humans. What should we expect the data to look like? You might reason as follows: since we seem to think we can arrange these four types of organisms from most primitive to most advanced, it should follow that the genetic distance from protozoa to the other organisms should get larger and larger as one moves towards humans. But this is not what we see. In fact, the genetic distance from protozoa to any of those organisms is approximately the same. This kind of analysis has been done with many different organisms, and the same pattern keeps on arising: a primitive organism has (approximately) the same genetic distance to any more advanced organism. Creationists argue that this flies in the face of evolutionary expectations. Does it?


I. Evolutionary predictions

We're going to use a schematic example here. Let's try to find out what evolution predicts the relative genetic distances to be between H. sapiens (that's us), D. melanogaster (a species of fruit fly), S. sempervirens (Pacific coast redwood tree), and P. caudatum (a protozoan).

The evolutionary expectations for the relationships between these organisms in terms of common ancestry is given as follows:

The exact intervals of time will not matter for our analysis (if you don't believe it, change the intervals however you want and see whether it affects my analysis). Now we would expect that the kind of genealogy we have here would be confirmed by molecular studies looking at differences in neutral mutations.

How do we determine the genetic distance between two species predicted by evolution? We start at the first species, go back in time counting intervals until we get to the first common ancestor of the two species, then we go forward in time to the second species, continuing to count intervals. So, for instance, if we want to figure out the predicted genetic distance between Homo sapiens and Paramecium caudatum we count the intervals from H. sapiens to A (there are three of them) and then add the intervals from A to P. caudatum (three more, for a total of six). For the distance between Drosophila melanogaster and Sequoia sempervirens, we go from D. melanogaster to C (one interval), from C to B (one more interval), and then from B to S. sempervirens (two more, for a total of four). This table shows the number of intervals that between our four representative examples:

Homo sapiens Drosophila melanogaster Sequoia sempervirens Paramecium caudatum
Homo sapiens 0 2 4 6
Drosophila melanogaster 2 0 4 6
Sequoia sempervirens 4 4 0 6
Paramecium caudatum 6 6 6 0

Note the moral of this story: the distance of a given species outside of Clade 1 to any species inside Clade 1 is the (all things equal) the same; the distance of a given species outside of Clade 2 to any species inside Clade 2 is (again, all things equal) the same. If we were to make a third clade consisting of all of the descendants of C, and compare them each to an organism outside of that clade (say, to a prokaryote), we should, again, see the same distance every time.

All members of a clade should (ceteris paribus) have the same genetic distance from a non-member of the clade. This is the critical point that creationists tend to miss; it is a prediction of evolution, not disconfirmation of it. The different genetic distances are expected when one compares an organism to creatures further and further outside of its clade, and these, too, are confirmed by actual molecular studies. The molecular studies should show the nested structure of the clades, and it does so.

Now, the data naturally will be much messier than this. First, there is the normal range of experimental error in individual tests. Second, the example I gave above presumes a constant mutation rate, regardless of species; usually this is true, but not always. Third, the example does not take into account mutations that reverse previous ones. Molecular biologists running the actual experiments, the referees of the peer-reviewed journals they submit their work to, and the peers who subsequently read their published work, are sensitive to all of these factors.  The point is that as one does more and more studies with a wider and wider set of molecular data, and all of the necessary factors are taken into account, the overall pattern should match evolutionary expectations. And it does.

Wesley R. Elsberry goes through all of this in a much more thorough, much less schematic way, in Sequences and Common Descent: How We Can Trace Ancestry Through Genetics. Elsberry shows an actual data set that lets you see both how messy the actual data is, and how it broadly confirms evolutionary expectations.

II. Creationist predictions

Let's go on, though, and consider what the sequences should look like it there were no common descent between all of the organisms in the example above. If H. sapiens, D. melanogaster, S. sempervirens, and P. caudautum were all created within a week, with identical noncoding regions of their genomes, then the clock would have started ticking for each at the same time, and the interval between any two organisms should be (ceteris paribus) exactly the same. The probability of the actual data coming to match evolutionary predictions by pure chance would be negligible; one would need the creator to intervene, and to guide the neutral mutations so that they would match evolutionary predictions. Aside from the theological questions raised by belief in such a deceitful creator (such as, why trust Genesis so much, if you believe God tells lies with nature?), such speculation is not testable, since the actions of a deceitful creator could look like anything: there are no definite predictions one can make from such a hypothesis.

Last updated: 28 Jan 2011

Pleased? Angered? Confused? Have something else you would like
me to write about? Please send in your questions and comments!