Genetic variation is a fundamental requirement for evolution, but many evolutionary mechanisms (such as selection and genetic drift) actually remove variation from populations. Therefore, evolution is entirely reliant on the formation of new genetic information, and without it, evolution would grind to a halt. Creationists often seize this fact, and erroneously claim that evolution is impossible because we have never found a mechanism that is capable of creating new genetic information. This claim is, however, completely false, because mutations do, in fact, create new genetic information. Nevertheless, many creationists respond to that fact by insisting that mutations simply “rearrange” existing genetic information, rather than creating “new” information. Therefore, I want to briefly explain why this argument is fundamentally flawed.

Hopefully everyone recognizes this, but if not, it’s Mr. DNA from Jurassic Park

Before I can explain the problems with this argument, we need to be clear about the basics, and you need to have at least a rudimentary understanding about how DNA works. DNA (or deoxyribonucleic acid if you prefer) consists of four bases adenine (A), guanine (G), thymine (T), and cytosine (C). These four bases get arranged into groups of three, and each group of three codes for an amino acid. Those amino acids then get strung together to form a protein, and those proteins combine to form tissues. Thus, your DNA is your body’s blue-print, and it tells your body which amino acids to make, and how to combine them to make proteins, tissues, organs, etc.

Please note that although what I have presented here is the most fundamental concept for you to understand, actual DNA also includes stop codons, start codons, and many other complexities that are irrelevant for this post.

Now that you understand the basics of DNA, we can talk about mutations. I explained them in far greater detail here, but in short, they are simply random changes to the genetic code (the ones that are important for evolution usually occur during the formation of egg and sperm cells). There are many different types of mutations such as inversions (which flip a segment of DNA), deletions (which remove base pairs), insertions (which add extra base pairs), substitutions (which insert the wrong base into a chain [e.g. a T instead of a C]), duplications (which duplicate a segment of DNA), and several others. In every case, however, they change the genetic code, and by changing the code, they can change the amino acids and ultimately the proteins that are produced (note: there is redundancy in the way that amino acids are coded, so not all mutations result in changes downstream).

At this point, we can examine creationists’ claim that mutations simply rearrange information rather than producing new information. The most obvious problem with this is simply that some mutations (like insertions) do actually insert entirely new base pairs. In other words, they don’t rearrange the code that is already there. Rather, they add new bases (i.e., new information) to that code.

The second problem is really the more important and fundamental one. Namely, this argument seems to be using the word “new” in a rather peculiar sense, because rearranging the existing bases does, in fact, produce a new code that often results in the production of different amino acids and new proteins. Let me illustrate. A coding strand of DNA that contains the sequence CTT would code for the amino acid Leucine. However, if a mutation rearranged those three bases so that they were TCT, that strand of DNA would code for the amino acid Serine instead of Leucine. Thus, by simply rearranging the existing bases, we created new information which produced a different amino acid. In other words, we added genetic variation to the population, because the individuals who receive that mutation will produce Serine, while the rest of the population is producing Leucine. Objecting to that by claiming that “no new information has been created” is really quite silly because the fact remains that individuals with that mutation are producing a different amino acid than everyone else. Something that codes for the production of an entirely different amino acid is, by any reasonable definition, “new information” (i.e., it is information that was not there before).

A useful analogy to help you conceptualize this is to think about letters in the alphabet. The English language has 26 letters (bases), and we combine those letters to from words (amino acids). We then arrange those words into sentences (proteins), and we arrange those sentences into paragraphs (tissues). Ultimately, we can use those paragraphs to make books, essays, etc. (organisms). Now, according to creationists’ reasoning, it should be impossible to make any new information by simply rearranging those 26 letters, but that is clearly absurd. We can arrange them one way and produce the works of Shakespeare. We can arrange them another way and produce “The Origin of the Species.” We can rearrange them yet again and produce the script to a Stargate episode. Indeed, there are a virtually infinite set of possibilities, each of which contains different information, and the situation is no different for DNA. We can arrange the bases one way and get a dinosaur, and we can rearrange that code (via mutations) and get a chicken. We can rearrange it yet again and get a whale, human, tree, bacteria, mushroom, etc. This notion that making new arrangements of the four existing bases doesn’t produce new information is absurd because everyone agrees that different arrangements of those bases produce very different organisms.

A big part of the problem here once again comes back to the definition of the word “new.” Creationists seem to think that evolution requires something that is completely and totally novel, such as a new base pair or, at the very least, an entirely new amino acid that has never existed anywhere before, but that is a straw man fallacy. Evolution does not require something that has never previously existed anywhere. Rather, it simply needs to have variation. Thus, any change to the genetic code is “new information” in an evolutionary sense, because it provides variation. A useful way to think about this is that evolution doesn’t need “new” information. Rather, it needs “different” information. In other words, all that it needs is a code that is different than the one that was there before.

To further illustrate what I mean by this, it is worth mentioning that even mutations that remove bases can actually produce new information in an evolutionary sense. Going back to the alphabet example, imagine that the document in question is a recipe, and imagine that the instructions get “mutated” by the random deletion of one of the steps. That actually provides new information because the end-product will be something different than the intended product (i.e., there will be variation for the trait). The same thing happens with DNA. If you remove the bases for a particular amino acid, then the final protein product will often be different from the one that was originally coded for, and that variation is all that evolution needs, regardless of whether or not you want to describe it as “new.” To give an actual example of this, the virus that causes HIV typically uses the CKR5 protein to enter macrophages and complete its lifecycle. However, people who have deletions (a type of mutation that removes bases) on both copies of their genes for CKR5 are resistant to HIV because those mutations alter the protein, and HIV cannot bind to this new, modified protein (Dean et al. 1996; Sullivan et al. 2001). Thus, a loss of genetic material results in a new set of instructions, which causes cells to produce a protein with a new modification, and that new modification provides a beneficial function. Creationists may try to say that this example doesn’t illustrate the formation of “new” information because an existing protein was simply modified, but that is, once again, a straw man fallacy, because evolution just needs variation, and this mutation provides that (thus, from an evolutionary perspective, it is new information).

Finally, you might be tempted to protest to all of this on the grounds that, “mutations are nearly always harmful,” but that claim is a myth. Most mutations are actually neutral at the time that they occur (i.e., they are neither beneficial nor harmful; Nachman and Crowell 2000; Eyre-Walker et al. 2007). Further, although harmful mutations certainly do exist, they are selected against, so they are really irrelevant. In other words, even if there were 10,000 times as many harmful mutations as beneficial mutations (which there aren’t), that wouldn’t preclude evolution, because nature would select against the harmful ones and for the beneficial ones. On that note, I want to be absolutely, 100% clear that beneficial mutations do exist and have been well documented (Newcomb et al. 1997; Dean et al. 1996; Sullivan et al. 2001; Shaw et al. 2002, 2003; Joseph and Hall 2004; Perfeito et al. 2007; see Halligan and Keightley 2009 for a good review). Indeed, we have done experiments with bacteria were we monitored populations for many generations, and observed the formation of novel, beneficial mutations that provided the bacteria with new information that allowed them to perform a novel function that they were previously incapable of (Blount et al. 2008; more detailed explanation in the citrate section of this post).

From the Star Trek TOS episode “I, Mudd”

In short, this notion that mutations can’t produce new genetic information is laughably absurd. DNA is simply the code that tells organisms what proteins and structures to make, and modifications to that code result in novel proteins and structures. Indeed, if you accept that dogs and cats are different because of differences in their DNA, then you have already accepted that rearranging genetic codes can, in fact, produce new information and result in vastly different organisms. Indeed, saying that rearranging DNA can’t produce new information is no different from saying that rearranging the letters of the alphabet can’t produce new information. Obviously it can, as is evidenced by this new blog post that I am writing. Additionally, when we say that evolution needs “new information,” we don’t mean that it needs something completely novel and totally different from anything else that has ever existed. Rather, we simply mean that there needs to be variation for traits. In other words, any slight modification to an existing trait qualifies as “new information” when talking about evolution. Finally, we have experimentally documented that mutations can produce that variation, and we have demonstrated that beneficial mutations do in fact occur and result in novel proteins that perform novel functions (i.e., the mutations created new information).

Literature Cited

• Blount et al. 2008. Historical contingency and the evolution of a key innovation in an experimental population of Escherichia coli. Proceedings of the National Academy of Sciences 105:7899–7906.
• Dean et al. 1996. Genetic restriction of HIV-1 infection and progression to AIDS by a deletion allele of the CKR5 structural gene. Science 273:1856–1862.
• Eyre-Walker et al. 2007. The distribution of fitness effects of new mutations. Nature Reviews Genetics 8:610–618.
• Halligan and Keightley. 2009. Spontaneous mutation accumulation studies in evolutionary genetics. Annual Review of Ecology, Evolution, and Systematics 40:151–172.
• Joseph and Hall. 2004. Spontaneous mutations in diploid Saccharomyces cerevisiae more beneficial than expected. Genetics 168:1817–1825.
• Nachman and Crowell. 2000. Estimate of the mutation rate per nucleotide in humans. Genetics 156:297–304.
• Newcomb et al. 1997. A single amino acid substitution converts a carboxylesterase to an organophosphorus hydrolase and confers insecticide resistance on a blowfly. Proceedings of the National Academy of Sciences 94:7464–7468.
• Perfeito et al. 2007. Adaptive mutations in bacteria: high rate and small effects. Science 317:813–815.
• Shaw et al. 2002. A comprehensive model of mutations affecting fitness and inferences for Arabidopsis thaliana. Evolution 56:453–463.
• Shaw et al. 2003. What fraction of mutations reduces fitness? A reply to Keightley and Lynch. Evolution 57:686–689.
• Sullivan et al. 2001. The coreceptor mutation CCR5Δ32 influences the dynamics of HIV epidemics and is selected for by HIV. Proceedings of the National Academy of Sciences 98:10214–10219.