A common myth that goes around in Intelligent Design (creationist) circles is the idea that DNA can only degrade over time, and thus any and all mutations are claimed to be harmful and only serve to reduce “information” stored in that DNA. The claim is specifically meant to suggest that evolution from a common ancestor is impossible by naturalistic processes because DNA wouldn’t have been able to form in the first place and/or it wouldn’t be able to grow or change to allow for speciation. Thus, the claim implies that either an intelligent designer had to intervene and guide evolution every step of the way (by creating DNA, fixing mutations as they occurred or preventing them from happening, and then ceasing this intervention as soon as scientists began studying genetics), or it implies that all organisms must have been created all at once by an intelligent designer with DNA that was “intelligently” designed to fail and degrade over time (thus questioning the intelligence of this designer).
These claims have been refuted a number of times over the years by the scientific community with a consensus that’s been drawn from years of research in evolutionary biology among other disciplines, and the claims seem to be mostly a result of fundamental misunderstandings of biology (or intentional misrepresentations of the facts) and also the result of an improper application of information theory to biological processes. What’s unfortunate is that these claims are still circulating around, largely because the propagators aren’t interested in reason, evidence, or anything that may threaten their beliefs in the supernatural, and so they simply repeat this non-sense to others without fact checking them and without any consideration as to whether the claims even appear to be rational or logically sound at all.
After having recently engaged in a discussion with a Christian that made this very claim (among many other unsubstantiated, faith-based assertions), I figured it would be useful to demonstrate why this claim is so easily refutable based on some simple thought experiments as well as some explanations and evidence found in the actual biological sciences. First, let’s consider a strand of DNA with the following 12 nucleotide sequence (split into triplets for convenience):
If a random mutation occurs in this strand during replication, say, at the end of the strand, thus turning Guanine (G) to Adenine (A), then we’d have:
If another random mutation occurs in this string during replication, say, at the end of the string once again, thus turning Adenine (A) back to Guanine (G), then we’d have the original nucleotide sequence once again. This shows how two random mutations could lead to the same original strand of genetic information, thus showing how it can lose its original information and have it re-created once again. It’s also relevant to note that because there are 64 possible codons produced from the four available nucleotides (4^3 = 64), and since only 20 amino acids are needed to make proteins, there are actually several codons that code for any individual amino acid.
In the case given above, the complementary RNA sequence produced for the two sequences (before and after mutation) would be:
UGA-CUG-ACU-GUC (before mutation)
UGA-CUG-ACU-GUU (after mutation)
It turns out that GUC and GUU (the last triplets in these sequences) are both codons that code for the same amino acid (Valine), thus showing how a silent mutation can occur as well, where a silent mutation is one in which there are no changes to the amino acids or subsequent proteins that the sequence codes for (and thus no functional change in the organism at all). The fact that silent mutations even exist also shows how mutations don’t necessarily result in a loss or change of information at all. So in this case, as a result of the two mutations, the end result was no change in the information at all. Had the two strands been different such that they actually coded for different proteins after the initial mutation, then the second mutation would have reversed this problem anyway thus re-creating the original information that was lost. So this demonstration in itself already refutes the claim that DNA can only lose information over time, or that mutations necessarily lead to a loss of information. All one needs are random mutations, and there will always be a chance that some information is lost and then re-created. Furthermore, if we had started with a strand that didn’t code for any amino acid at all in the last triplet, and then the random mutation changed it such that it did code for an amino acid (such as Valine), this would be an increase in information regardless (since a new amino acid was expressed that was previously absent), although this depends on how we define information (more on that in a minute).
Now we could ask, is the mutation valuable, that is, conducive to the survival of the organism? That would entirely depend on the internal/external environment of that organism. If we changed the diet of the organism or the other conditions in which it lived, we could arrive at opposite conclusions. Which goes to show that of the mutations that aren’t neutral (most mutations are neutral), those that are harmful or beneficial are often so because of the specific internal/external environment under consideration. If an organism is able to digest lactose exclusively and it undergoes a mutation that provides some novel ability of digesting sucrose at the expense of digesting lactose a little less effectively than before, this would be a harmful mutation if the organism lived in an environment with lactose as the only available sugar. If however, the organism was already in an environment that had more sucrose than lactose available, then the mutation would obviously be beneficial for now the organism could exploit the most available food source. This would likely lead to that mutation being naturally selected for and increasing its frequency in the gene pool of that organism’s local population.
Another thing that is often glossed over with the Intelligent Design (ID) claims about genetic information being lost is the fact that they first have to define what exactly information is necessarily before presenting the rest of their argument. Whether or not information is gained or lost requires knowing how to measure information in the first place. This is where other problems begin to surface with ID claims like these because they tend to leave this definition either poorly defined, ambiguous or conveniently malleable to serve the interests of their argument. What we need is a clear and consistent definition of information, and then we need to check that the particular definition given is actually applicable to biological systems, and then we can check to see if the claim is true. I have yet to see this actually demonstrated successfully. I was able to avoid this problem in my example above, because no matter how information is defined, it was shown that two mutations can lead to the original nucleotide sequence (whatever amount of genetic “information” that may have been). If the information had been lost, it was recreated, and if it wasn’t technically lost at all during the mutation, then it shows that not all mutations lead to a loss of information.
I would argue that a fairly useful and consistent way to define information in terms of its application to describing the evolving genetics of biological organisms would be to describe it as any positive correlation between the functionality that the genetic sequences code for and the attributes of the environment that the organism is contained in. This is useful because it represents the relationship between the genes and the environment and it seems to fit in line with the most well-established models in evolutionary biology, including the fundamental concept of natural selection leading to favored genotypes.
If an organism has a genetic sequence such that it can digest lactose (as per my previous example), and it is within an environment that has a supply of lactose available, then whatever genes are responsible for that functionality are effectively a form of information that describes or represents some real aspects of the organism’s environment (sources of energy, chemical composition, etc.). The more genes that do this, that is, the more complex and specific the correlation, the more information there is in the organism’s genome. So for example, if we consider the aforementioned mutation that caused the organism to develop a novel ability to digest sucrose in addition to lactose, then if it is in an environment that has both lactose and sucrose, this genome has even more environmental information stored within it because of the increased correlation between that genome and the environment. If the organism can most efficiently digest a certain proportion of lactose versus sucrose, then if this optimized proportion evolves to approach the actual proportion of sugars in the environment around that organism (e.g. 30% lactose, 70% sucrose), then once again we have an increase in the amount of environmental information contained within its genome due to the increase in specificity.
Defining information in this way allows us to measure degrees of how well-adapted a particular organism is (even if only one trait or attribute at a time) to its current environment as well as its past environment (based on what the convergent evidence suggests) and it also provides at least one way to measure how genetically complex the organism is.
So not only are the ID claims about genetic information easily refuted with the inherent nature of random mutations and natural selection, but we can also see that the claims are further refuted once we define genetic information such that it encompasses the fundamental relationship between genes and the environment they evolve in.