This is part of a series on the proof of God in DNA. See this index.

In part 1, I discussed information and DNA-as-software as evidence for the existence of a creator. I was responding to the claims against this position as laid out in this article by Bob Seidensticker. Even though Mr. Seidensticker discussed them at the end of his article, the nature of patterns, information, and intelligence are such important macro issues that they had to be addressed first. In part 2, I discuss some of the more micro issues that he raised.

Inside the cell is something remarkable: DNA. DNA is remarkably like a computer program. A computer program, at its lowest level, is a sequence of CPU instructions. These instructions retrieve, modify, and store values. Each instruction is encoded, in binary (base-2), into an operation code (“op-code”). DNA is a coded sequence of building instructions inside the the cell. Like software, DNA has discrete instructions. These instructions are encoded a quaternary (base-4) system into a set of 64 op-codes, called codons. These instructions code for the 20 standard amino acids that cells use to build proteins. Remarkably, the instruction set includes no-op op-codes, just like a human-built computer. The coding also allows for error detection and error correction.

As a programmer these similarities are hard to miss. How can a computer program, obviously designed by humans, be so similar to the “program” that runs in our cells and is responsible for the complexity of all life? Must this, therefore, be proof that God both exists and that he designed life? For some, this is compelling. Author, Mathematician, and Oxford Professor John Lennox has argued this point in many debates and speeches. Of course others disagree. Yet, in researching this post, I had to do a double-take, wondering if maybe the original computer had been inspired by someone who understood how DNA worked. But no, that was impossible. Yet the comparison is striking.

“One popular science-y argument for God is that DNA is information.”

Note the mockery.^[1] This is typical disdain for any evidence for God: because it’s ‘not science’, it can’t be true. This will be important later, but it sets the tone for what follows. Most of the arguments sound good to the listener because of confirmation bias. It will be hard to untangle the will to predetermine the outcome of the argument.

Before I continue, I need to correct a few subtle mistakes that have been made. In analyzing the DNA software and human software, Mr. Seidensticker has not been careful to distinguish between the encoded software (the way the information is stored and retrieved) and the actual program itself (the decoded software). Wherever possible I will try to speak in a general way so it doesn’t matter, but if there are any objections to my rebuttal, I may be required to make this distinction more explicitly. When I refer to this below, it will be as the “encoding” and “decoding”. Moreover, to do an apples-to-apples comparison of human software and DNA software, one has to look at machine code, not the higher level programming languages and compilers that are used to generate that machine code.

“With DNA, we again see the natural vs. manmade distinction. It looks like the kind of good-enough compromise that evolution would create, not like manmade computer software.”

Mr. Seidensticker precedes this with an quoted anecdote on how attempting to replicate nature is less efficient than trying to do things the way a human designed it. DNA, therefore, shows all the inefficiencies of natural processes and none of the hallmark of human design. There are many things wrong with this.

First, why would we expect a designed biological system to behave like a designed non-biological system? It’s strange to expect a biological computer to be ‘efficient’^[2] in the same way a non-biological computer is designed. Nor would we expect that copying a biological design into a non-biological context would yield good results. The underlying mechanisms of each system are just too different.^[3]  Moreover, creating the biological system is significantly more difficult: humans have not succeeded in creating even the most basic of lifeforms from scratch.

Second, there is no reason to expect a God-designed system to look the same as a human-designed system. Now, you could argue that a God-designed system should be more elegant, but it’s hard to make the claim that the DNA system isn’t elegant. It’s quite remarkable in its design. And it’s lasted, quite successfully, for millions of years. No human designed system comes anywhere close to that sort of longevity (mechanical and electrical systems experience rapid degradation) or resiliency (human designed software systems are full of critical errors that need patching).

Third, the biological computer has error handling built in. To go into the mechanism of this would take another post, but it is quite sophisticated. Be careful not to commit the fallacy of equivocation here. A good-enough compromise is a remarkable achievement: balancing the difficulties of a biological computer surviving in a harsh biological environment. No, what is being implied by “compromise” is that concessions were made: a poorer than optimal outcome was achieved. Yet when research was done on this very topic, it was discovered that the specific coding used in DNA today (not millions of years ago) is very close to the optimal human calculated design.^[4]

Fourth, the origin of life is currently a scientific unknown. Evolution describes the development of life from other forms of life. The origin of the RNA/DNA replicating machine is not even close to settled science, and it is not clear what the role of evolution was in this process. It is dubious, therefore, to make the absolute claim that that DNA looks like a “good-enough compromise that evolution would create”, since DNA (or at least RNA) preceded evolutionary processes. Embedded in this argument is circular reasoning that presupposes that God could not create life.^[5] This is a sloppy argument that needs better science in order to make it properly.

Fifth, man-made computer software involves many design compromises at all levels (hardware, software encoding, and software decoding).

“As another illustration of the non-software nature of DNA, the length of an organism’s DNA is not especially proportionate to its complexity.”

First, software complexity can have significant correlative deviations from software size. Two programs that do the exact same thing can differ wildly in size by orders of magnitude.^[9] This is further accentuated when comparing DNA software with machine code. This is a terrible argument to make.

Second, as human software programs increase in size and complexity, they also dramatically decrease most quality metrics, including efficiency and clarity. This does not happen in the DNA computer. This demands an explanation.

Third, the complexity of an organism is based less on the size of its program than the code base that it is derived from. This makes sense. The software that a cell inherits from its parent has a much stronger correlation with complexity than any other factor. The same thing holds true for human software systems.^[10] DNA length is unlikely to be an independent variable in the complexity equation.

Fourth, the DNA computer is very old. It’s not clear if the argument is being made that DNA has always been disproportionate to its complexity or if it has developed that way over time. If the latter, then there is no problem at all with the notion that God created it one way and it developed changes over time. If it has always been disproportionate, that really throws a wrench in the idea that life originated from a single common ancestor. This poses no problem at all for the theistic position, but it could pose significant problems for the atheistic position. Science still has to figure out exactly how the disproportionate DNA lengths originated before we can decide if those processes are pro- or anti-design.

“The Craig Venter Institute encoded four text messages into synthetic DNA that was then used to create a living, replicating cell. That’s what a creator who wants to be known does. Natural DNA looks . . . natural.”

First, natural DNA looks biological, which is hardly surprising or unexpected. Don’t make the mistake of equating the mechanism for encoding, a biological process, with the encoding itself. The encoding is digital^[6]. It does not look digital, it is digital. Digital encodings do not look natural, they look unnatural^[7]. This is why they constitute evidence for God, not against.

Second, imagine that my friend calls me on the phone and says “Hi, Derek? This is Matt.” Do I shout back “Are you really Matt? What is your social security number?” That’s the same level of absurdity in claiming that the creator, who wants to be known, isn’t doing enough to make himself known. It’s faulty logic to reject the evidence and then claim that there is no evidence. Refusal to accept the evidence is not proof against evidence for God.

Third, the results of human genetic engineering looks… natural. Is that, therefore, proof against human intelligence?^[8] Of course not, the argument is invalid in either formulation.

“If God designed software, we’d expect it to look like elegant, minimalistic, people-designed software, not the Rube Goldberg mess that we see in DNA. Apologists might wonder how we know that this isn’t the way God would do it. Yes, God could have his own way of programming that looks foreign to us, but then the “DNA looks like God’s software!” argument fails.”

First, complex machine code is a huge mess. Appealing to a mess in DNA isn’t going to meaningfully alter the analogy.

Second, why does DNA look like God’s software? Well it doesn’t.^[11] It looks designed. For instance, aliens could, theoretically, have done it. If God designed software we’d expect it, first and foremost, to be functional and long-lasting. The core requirements would not be elegance or minimalism.

Third, a poor design is still a design. It would still constitute proof of a creator, even if that proof indicated that the creator was incompetent. Either way, it would be devastating to the atheist case, so it is surprising to see this argument being made here. Saying that it can’t be designed because the resulting design would be poor is hardly a foregone or obvious conclusion.^[12] But that isn’t what the atheist is doing. The atheist is arguing that there is no design, poor or otherwise. The standard for proof of this is much, much higher. For sake of argument, I’ll happily accept any atheist’s claim that DNA was designed and designed poorly.^[13]

“This is no evidence for a designer, just an unsupported claim that complexity demands one. And why think complexity is the hallmark of design? Wouldn’t it be elegance or something similar?”

First, there is evidence for a designer. If that evidence has been evaluated and rejected, that’s perfectly fine. But it can’t be said that there is no evidence. If one is going to take the logical high ground, then you can’t just discount the evidence presented out-of-hand and assume a particular conclusion.

Second, complexity and elegance are not mutually exclusive. Either could be evidence of design. Or neither.

Third, while the argument-from-design includes things other than DNA-as-software, I’ll keep my focus restricted there. Things that are designed can be complex or simple. Things that are not designed can be simple. Can things that are not designed be complex? The theist (and atheist) must be careful not to engage in circular reasoning by defining as complex those things that do (or don’t) match their concept of design.

^[1] The reliance of New Atheists on using ridicule is intentional. The goal is to portray themselves as rational while making use of aggressive emotional propaganda to cover over the inherent flaws in their arguments. Atheists are rarely interested in dialog and are usually especially hostile to it as they feel it legitimizes the beliefs of the religious.

^[2] It isn’t clear what is meant by efficient here. In fact, it’s hard to imagine a way to compare the efficiency of biological and non-biological systems. For example, the lens in a human eye is not perfect. It degrades with time. We can, of course, design ground glass fluorite lenses with superb optical properties that, with proper care, out last a human lifetime. If this is a backwards way of saying that God was stupid to use a biological system for life when a mechanical or electrical system would have worked more efficiently, that’s a bizarre way to make that argument (and it’s a weak one too).

^[3] This is an example of picking the wrong tools for the job. For example, running a high-level language interpreter on a small embedded system. In the software world this would be considered bad design, the opposite of Mr. Seidensticker’s claim.

^[4] I’ll post a link to this research when I can find it again.

^[5] It’s pointless to argue that DNA proves that God exists if you’ve already decided that God can’t exist and therefore life could not have come from God and must have come from a natural process. This is circular reasoning. The result is confirmation bias: any evidence is interpreted with the conclusion already in mind. The Christian is not required to appeal to a God-of-the-gaps here, as the Christian is putting both premises up for consideration: “God created life” and “Natural processes created life by chance” and examining the evidence.

^[6] That is, discrete, non-continuous values. In this case non-binary base-4 digital values.

^[7] The atheist thinks that digital data can be natural or unnatural. This is a central point of debate. Asserting that DNA is natural or unnatural is asserting the conclusion of the argument as true, obviously faulty logic. That’s why Mr. Seidensticker’s objection is not an objection, it’s just a restatement of his thesis. More needs to be said.

^[8] Flipped around: if humans can create natural-looking intelligence, then so can God.

^[9] See: here and here. With regards to the latter link, ‘Junk DNA’, if ever scientifically proven, would compare quite favorably with the human software analog.

^[10] A New version of Microsoft Word is going to be significantly more complex than a new version of GNU grep because that was true of the previous versions. Divergent, or forked, software (DNA or human designed) is not expected to maintain its size correlation.

^[11] Theists, understandably, get excited at the idea of proof of God and thus engage in hyperbolic language. That doesn’t make the arguments invalid, it just means that they need to calm down and be more precise in making those arguments.

^[12] Poorly written software is still software. For example, Windows Vista.

^[13] Why would a theist mind if DNA proved the existence of a creator but failed to conclusively prove that such a creator was God? That is still a remarkable discovery.