Freedom had been hunted round the globe; reason was considered as rebellion; and the slavery of fear had made men afraid to think. But such is the irresistible nature of truth, that all it asks, and all it wants, is the liberty of appearing.
~ Thomas Paine, Rights of Man, 1791 ~

Thursday, February 9, 2017

Does DNA require God?

I recently had the misfortune of arguing with a creationist who insisted that DNA is (or contains) information, and that information requires an intelligent creator (God, of course). This appears to be the latest angle creationists pursue in order to cling to some form of perceived relevance in today's scientific world.

First, let me just point out that it's interesting and perhaps encouraging to see that certain other long-debunked arguments for creationism are now advanced less often than before. Progress of a kind. Creationism will be with us for a while yet though, because it is almost never purely the result of careful examination of the evidence. These people typically became creationists as children, and then invented (or learned) creative ways of defending that belief as adults. Meanwhile, it is estimated that over 99% of those who have studied the evidence most carefully, accept evolution.

Information requires intelligence?

When you first hear it, it almost sounds plausible. It has that ring of human intuition to it, notwithstanding an intuition influenced by our own feeling of self-importance. Intelligence is, after all, something only humans have (so goes the hubris), and since we are the only ones who produce information, therefore information requires intelligence.

But wait, wasn't the claim that information (i.e. in DNA) exists in nature? But that would mean humans aren't the only source of information, which would mean information can arise by some means other than (human) intelligence. Bummer.

But creationists dive in deeper instead. They insist that information can still only arise via intelligence, and therefore there must be some super-intelligence that created everything in nature. Of course, creationists are not arguing for deism. Oh no. Not only do they "know" the identity of this super-intelligence, they also basically know the colour of its underwear and its preferred breakfast cereal too!

How do they know this? Well, they have a book. I know, right? Everyone knows that if something is written in a book, it must be true. Also, it's not like anyone else has a different book that they claim was written by a god. Oh, wait.

Appearance of design

Perhaps you've already noticed the similarities with William Paley's Watchmaker analogy. The argument is very similar, so it's worth mentioning this in order explain what's going on with the above claim.

I've covered Paley's Watchmaker in a previous article, but I will outline the main points here for convenience.

Basically, Paley argued that if he found a watch lying on the ground, he would instantly recognise design in its features. He then argued that the same features are apparent in nature, and thus nature must also be designed.

The problem is that Paley argued for the complexity of the watch by contrasting it with the surrounding nature, and in particular a stone. But that assumes there is some distinction whereby we can conclude that the watch was designed, while the other thing (nature) was not. So how did Paley then conclude that nature was designed? It seems he just ignored the previous contrast and then argued that the watch and nature were similar instead of different. But by claiming that all of nature must be designed, he was effectively standing in a field full of "watches" and deciding that the one he picked up was special.

In case the similarity with the earlier claim is not readily apparent, the basic process is that you see some pattern in things humans created (having easily identified that it was not naturally occurring), then you find what you perceive to be a similar pattern in nature, and conclude that nature was likewise created by an intelligent being. You would of course hope to avoid the conclusion that humans copied nature. That might spoil the argument.

However, this does not address the full story, so let's move on to what information is.

What is information?

Wikipedia gives the definition "that which informs" and also the following summary:
At its most fundamental, information is any propagation of cause and effect within a system. Information is conveyed either as the content of a message or through direct or indirect observation of anything. That which is perceived can be construed as a message in its own right, and in that sense, information is always conveyed as the content of a message.
Google's definition is a bit simpler:
2. what is conveyed or represented by a particular arrangement or sequence of things.
Note that the "message" does not require that the source be conscious or intelligent. For example, the arrangement of the stars provides information to sailors that can help them to navigate, but it would be absurd to think the stars were either conscious or intelligent, or somehow conspiring to provide this information. Also note that it almost doesn't matter which arrangement they are in - there is still information that can be useful.

There are also many examples from nature. Flowers also contain information used by bees, which is then relayed to other bees. Are flowers intelligent? What about bees? However, examples from nature come with the added baggage of creationists insisting that nature itself is the product of intelligence. It's not clear whether the cart is pulling the horse or vice versa.

Information then, is a message "about something", rather than (necessarily) "from someone". To argue otherwise is to argue in a circle. I will return to this later.

The next thing I want to look at is the contrast between information and random numbers. This is where things really get interesting.

Random numbers

Before I get into this, I want to first declare that random numbers by definition cannot be the product of intelligence. If they were, then there would effectively be no such thing as random. This may heavily impact the creationist assertion, showing it to be fundamentally unfalsifiable, but I'll get to that later.

If information is to only be possible via intelligence, then it cannot occur via random processes. But it's trivially easy to show how the same sequence of numbers could result from either intelligence or random processes. By imposing additional processes on top of the existing random numbers, one can generate increasingly complex patterns, potentially indistinguishable from information created via intelligence. In other words, just observing some information is not enough to determine its source.

To demonstrate, consider the sequence ABCD. Is it information? It's trivially easy to produce this by randomly shuffling those 4 letters. We can add more letters and create a complete "sentence". As an aside, I use the analogy with language cautiously, because DNA is not a language in the same way that English is a language. There is no meaning encoded in DNA (so far as we can determine). There are similarities with language, but they are not identical. Analogies are therefore useful as a tool to explain a concept, but not as the basis of an argument.

Most people would agree that if they can understand a sentence, therefore it is information. It could still be information even if not human-readable, but I'm simplifying the example. However, we can clearly see that purely random collections of letters could produce the exact same sentence. Just observing the sentence is not sufficient to determine what produced it. So is information only the product of intelligence? The answer is no.

By now you're probably imagining DNA forming entirely by random chance and wondering how such complexity could arise this way. Monkeys on typewriters. But that's the wrong analogy. Nature is not (entirely) random. I used random numbers as an example to show at least one scenario where information can arise without intelligence. Now I'll look at another example.

Non-random processes

Take our earlier sequence of letters and assume that there is a pre-existing bias for some letters to appear in a certain pattern. Perhaps sometimes the letters appear in order. Perhaps in some conditions the letters conform to some fundamental rules. Perhaps some letters only ever appear together. Perhaps certain letters always appear next to a particular other letter. In the case of DNA, this last example is exactly what we find. Each of the letters A, C, T, and G have a corresponding letter that it is supposed to pair up with. A always pairs with T. C always pairs with G. If you think about it, now you only need one side of the pair in order to re-create the other side. This is basically how DNA error-correction works. Neat, huh? But intelligent? Hardly, it's just a simple rule.

So given these underlying rules, common patterns start to emerge naturally. Are these patterns information or not? Who gets to decide? We already determined that completely random numbers can form short patterns easily. If these patterns are combined with non-random processes, duplicated, modified, and if those processes are repeated over and over, it's not difficult to imagine very complex patterns emerging all without any intelligence behind it. Natural selection and random mutation are just two such processes. There are many more. In fact in computer software, genetic algorithms are often used to produce near-optimal solutions to various difficult problems relatively quickly. That is, without knowing the end result in advance, a computer program can start with random data and reach a sophisticated solution simply by applying the same rules over successive generations.

People often compare DNA with computer software, but as with language the analogy is imperfect. With computer software, if there's a syntax error, the entire program stops (or may not even compile). With DNA, an error may cause a single gene to stop working (or work differently), but that will have an effect on the organism either way. There is no CPU interpreting and running the code. It's literally just chemistry and physics doing their thing. The analogy simply breaks down at anything beyond a superficial level. As I mentioned earlier, analogies are useful when explaining various features, but we have to be careful not to take them too far.

Just taking the processes of random mutation and natural selection, we can easily see complex patterns arising in nature. Random mutation causes genes to behave differently. Creationists like to pretend that all mutations will just cause an organism to die. This is simply false. Most mutations are neutral. But even if 90% of non-neutral mutations were harmful, that still makes 1 in 10 mutations beneficial, and considering the relative frequency of mutations among a population, that's a lot of beneficial mutations in terms of raw numbers. Harmful mutations rarely survive, while beneficial ones often do. Thus over time, beneficial mutations accumulate, while harmful ones don't (in general). Natural selection simply says that organisms with beneficial mutations are more likely to survive and breed (which is demonstrably true), and thus over time we would expect to see organisms with more beneficial mutations and few harmful ones (and many neutral ones). This is exactly what we find in nature.

Digging deeper beneath DNA we have chemistry, which we know conforms to various rules. The result of any chemical reaction is never completely random. Various simple chemicals can be combined to form more complex ones (which is rather interesting considering creationists often claim that complexity cannot arise naturally from less complex things). This process is repeatable. Do these complex chemicals contain information? Could the simple chemicals be formed by random combinations of yet simpler chemicals? The answer to both questions, I think, is yes.

And of course, underlying chemistry is physics. Just like chemicals combine to form more complex chemicals, molecules can combine to form more complex molecules. Again the process is repeatable. The combinations are not random. There are again rules that describe how it works. Notice that I said "describe" rather than "govern". I tend to believe the laws of nature are descriptive, not prescriptive. That is, they are descriptions of what could be emergent behaviour from some processes occurring at a level beyond our current knowledge. I may discuss this someday in a future article. For now, I'll just mention that our current best understanding is that all of the elements in nature, including every atom in your body, were formed inside dying stars via nuclear fusion. I find that truly mind-blowing.

Muddying the waters

But what about information that is the result of intelligence? Does that mean there must be a super-intelligence out there somewhere? Or can intelligence itself arise naturally (i.e. as opposed to super-naturally)?

If intelligence can arise naturally, the entire creationist argument collapses. Of course, if evolution is true, then intelligence can arise naturally. Some people think that God was required somewhere along the way. That's fine, but unless you think God put something magic into humans (that we have not yet found), or you believe in mind-body dualism, whatever method God used must have involved natural processes, and thus intelligence can arise naturally. For what it's worth, if you do believe in mind-body dualism, note that the brain is made of atoms, and those atoms obey cause and effect. At every stage leading to you performing an action, there was a prior natural cause. In order to prove dualism, you need to insert a supernatural cause in there somewhere, but if you did have a natural effect without a natural cause, that would violate conservation of energy. That pretty much rules out dualism, unless you insist that the universe is not a closed system, but that has other much wider implications.

Consider that we are now living in a very exciting (and potentially disturbing) time when computers are now "intelligent" enough to defeat the world's best Chess player, the world's best Go player, and more recently even defeat the world's best poker players.

Most of these artificially intelligent machines have some way of learning (and thus producing) new information. They are performing tasks once thought to require human intelligence. Still, these machines were created by humans, which complicates the issue. Back when I was a believer myself I thought it would be fascinating if humans could produce artificial "life". I wasn't sure whether that would be evidence in favour or against "intelligent design". The same argument applies to the invention of intelligence itself. It muddies the waters somewhat.

Does information require intelligence?

The answer depends on the assumptions you start with.

Creationists by definition assume that everything that exists was created by an intelligent being. Thus, there is no way, even in principle, to demonstrate the existence of information that was not the product of intelligence. That is, the claim becomes unfalsifiable, and therefore meaningless.

On the other hand, if evolution is true (which it is according to over 99% of those who study nature as a profession), or if one can otherwise demonstrate that natural processes are all that is required to produce information (as I believe I did above) then information does not require intelligence, at least not directly. The only way to conclude that information does require intelligence would be to insist that God created those natural processes, for which there is no evidence, and which is again an unfalsifiable claim. You simply cannot know either way.

So why do creationists use this argument?

Given that I cannot read minds, I don't know the exact reason why creationists use arguments like this. However, I can offer one plausible reason. Creationists may use this style of argument as a wall in order to protect their underlying beliefs, or as a hook to hang their faith coat on. They don't even need to look at other evidence so long as they've got this wall to rely on. It thus resolves cognitive dissonance by simply blanket avoiding all contrary evidence.  That conserves a lot of mental energy, so somewhat ironically you can see the adaptive benefit to that style of thinking. I think we can all admit to having done this sometimes. Not necessarily in this area, but perhaps with some deeply held belief that we didn't want to have challenged. It is a very human thing to do.

The problem with putting up such a wall arises when the wall is not falsifiable. That is, there is no possible scenario, even in principle, where some piece of evidence or reasoning would make your claim false. This is why it's always important to make sure our beliefs can be tested (at least in principle), lest we simply fool ourselves. We should also be prepared to change our beliefs if they are indeed falsified. Easier said than done.

As I mentioned earlier, the claim that "information requires intelligence" may well be unfalsifiable, especially if it rests on the assumption that everything that exists is the result of intelligence. Such assumptions then actually get in the way of discovering what is real, rather than being of any help to us.

Science denialism (all the experts are wrong but not me)

One thing that does confuse me about the creationist position is that it necessarily depends on believing that virtually an entire field of scientists, over 99%, are completely wrong without knowing it. This is such a strange and naive view that it's really difficult to wrap one's head around. If it were true that so many trained biologists and others have so utterly failed to understand the natural world they study daily, then on what basis should any of us (who have no such training or experience) think we could do better? Surely if the world was so difficult to understand that over 99% of those who tried got it completely wrong despite their best efforts, then the rest of us should probably just give up in despair!

As it happens, I definitely think we can understand the natural world, and I think the majority of scientists have discovered an awful lot about it. It is only blind arrogance (or perhaps religious indoctrination as a child) that would cause a person to reject the overwhelming consensus in the relevant disciplines in favour of a fringe group. It gets worse when most members of the fringe group were also indoctrinated as children and openly admit their religious bias.

To be clear, the argument is not simply that you should blindly believe whatever the majority believe. Else we should all be catholic. That's just a straw man often raised against this view. The argument is that if over 99% of the world's biologists (and other related fields) agree on how to interpret the mountains of evidence they've uncovered, and if almost all scientific papers published on the subject in respected, peer-reviewed journals also generally agree overall, then you'd want to have a damn good reason if you disagree with them. At the very least you should have a very good understanding of the subject. Not only that, but the onus would be on you to publish your own contrary evidence and seek to change the views of your colleagues. But creationists generally don't try to publish their views in respected journals. They would rather preach direct to the public, often to people who don't know enough to determine whether they're lying or not. You have to ask why that is.

Interestingly, most of those who deny climate change use exactly the same arguments as the creationists (i.e. "all the experts are wrong but somehow I stumbled on the truth"), and follow exactly the same tactic of preaching to the public instead of publishing their evidence.


Does information require intelligence? This question arose in the context of DNA, but it has wider implications than that.

My short answer is, "I don't know". I have provided examples where information can arise from random and non-random processes (and combinations of the two), which strongly suggests to me that information does not require intelligence. Further, I have no good reason to think that it does. I do have good reasons not to "just assume" that it does, because that would necessarily bias any conclusions based on that assumption. To assume it is true is to decide ahead of time how you think the world should work. I don't think we have that kind of power, just quietly. Rather, we'd just be fooling ourselves.

One objection creationists may raise is, "Where did the non-random processes come from?". The question is equivalent to asking where the laws of physics came from. The obvious answer is, "Nobody knows". As far as we can tell, that's just the way things are. That answer isn't likely to satisfy many people, including myself, but without some deeper understanding and/or evidence, both of which seem to be permanently beyond our reach, we're not likely to get any further. The multiverse hypothesis at least speculates that the laws of physics may be different in different universes, which suggests that we just happen to live in the only universe with characteristics that permit our kind of life. Again, it's not completely satisfying, but it may be the best we can do. At the end of the day, we don't know, and perhaps we can't know. That won't stop people from speculating wildly and staking their life on it though.