Frequently, when cooking, doing the dishes, or other household tasks, I will listen to YouTube debates between theists and atheists (here is a YouTube channel that has a huge number to choose from). Nearly always when these debates turn to evolution vs. creationism, the facts and theories of evolution are horribly misrepresented. While a thorough explanation of evolution would require a full university-level course, certain aspects of it that are commonly misrepresented in such debates are amenable to short-form explanation. That is the purpose of this document.
You might want to start your exploration of these topics with my earlier document, Evolution: A Primer. Here, in FAQ format, I will address a number of common creationist claims. If there are any particular other claims that I am not addressing that you would like me to, please let me know in the comments and if they are appropriate, I will add them to this document. This list will inevitably be incomplete (see here for an attempt at a relatively comprehensive list of debunked creationist claims to see just how extensive such a list would have to be), but I am happy to address any specific claims for which there is demand.
- Mutations are nearly always detrimental
- Getting just the right mutation is ridiculously unlikely
- Mutations cannot add information
- Mutations are insufficiently frequent to account for the genetic differences between species
- There are no transitional forms, a.k.a., missing links
Mutations are nearly always detrimental
Actually, mutations are nearly always silent, meaning that they make no difference at all to the organism. It has been estimated that 90-97% of human DNA is noncoding. This noncoding DNA, formerly called “junk DNA,” does not code for proteins. Some of this noncoding DNA has been found to have some function other than coding, but that is but a small fraction of the total.
Further, the code that is used to convert DNA sequences to proteins (amino acid sequences) is redundant. DNA is made up of three “letters,” chemicals that are abbreviated T, C, A, and G (for RNA, T is replaced with U). A codon is a sequence of three of these letters that code for a particular amino acid, of which there are twenty. But there are multiple codons that code for the same amino acid. For example, the amino acid alanine can be coded by GCT, GCC, GCA, or GCG. The amino acid arginine can be coded by CGT, CGC, CGA, CGG, AGA, or AGG. On average, one third of point mutations to a DNA coding sequence will result in no change at all to the resulting protein… it will just be spelled differently in the DNA code. Combine the spelling differences and the noncoding DNA regions, and 93-98% of point mutations have no effect at all on the organism.
Most of the remaining 2-7% of point mutations are also neutral. A typical protein consists of about a thousand amino acids which fold up around an active site that consists of about fifty amino acids. Changes to the active site can have a strong effect on the properties of the protein, but changes elsewhere in the protein usually do not unless they substantively change the folding pattern. This means that the fraction of point mutations that are subject to evolution’s selective pressures is tiny.
So not only are most mutations not detrimental, most mutations make no difference at all.
Getting just the right mutation is ridiculously unlikely
There are two problems with this claim. First, it assumes that there is a particular mutation that evolution is targeting under a certain set of conditions. This is not the case. There are lots of ways to skin a cat, as it were. Protein families, where mutational changes to proteins produce slightly altered functionality, are quite common, with over 13,000 such families now known. Different species have, for example, different forms of hemoglobin, all of which more-or-less perform the same biological function. What is more, the dominant forms in each species can be compared across species to build a tree of relatedness, a tree that matches what is expected based on other means of determining species relatedness. Evolution doesn’t decide that a particular variant is necessary for a new species… it just tries a bunch of things and some of them end up surviving.
But the bigger problem with this claim is that the likelihood of a particular mutation depends on the population size. Let’s consider humans. Our DNA contains approximately three billion base pairs. If we pick a typical estimate of how many point mutations there are per individual, 64, and look at the population of the earth, a little over seven billion, it becomes clear that every single possible point mutation exists somewhere in the Earth’s human population.
So getting any given point mutation is not unlikely, it is inevitable.
Mutations cannot add information
As I discussed in my primer on evolution, there are numerous mutation mechanisms that result in the addition of information. To pick one example, gene duplication followed by subsequent modification of one of the duplicates adds information. There are thousands of scientific articles documenting examples of this (try a PubMed search on “gene duplication”).
Mutations are insufficiently frequent to account for the genetic differences between species
Let’s look at the human/chimpanzee difference as an example. The difference in DNA between humans and chimpanzees is 1.2% (of total DNA, not just coding DNA). Our most recent common ancestor was between six and seven million years ago. Using an average generation time of 25 years (accurate for both humans and chimpanzees), there were between 240,000 and 280,000 generation between that common ancestor and today. Using the same estimate of point mutation rates used above of 64 per generation, that gives us between 15.4 and 17.9 million point mutations for us compared to that common ancestor. But chimpanzees have as many generations between them and that common ancestor as we do, so we expect between 30.8 and 35.8 million base pair differences between us and chimpanzees. Since the human genome consists of 3 billion base pairs, this difference is between 1.02% and 1.19% of the genome, in remarkable agreement with the experimental difference in DNA listed above.
This calculation is something of an oversimplification, since there may have been an original split between the lines before this time, but with some cross-breeding since. It also only takes into account point mutations, whereas other mutation mechanisms, which are rarer but contribute multiple base-pair differences per incidence, certainly contribute some. Still, even with those caveats, it is clear that mutations are easily sufficient to account for the observed genetic differences between species.
There are no transitional forms, a.k.a., missing links