Keith Devens .com

How about this: evolution and creation are both theories, neither of which are indisputable. Why don't we set aside this debate until scientists on either side demonstrate a clear victory. Otherwise, this polarizing debate can become pretty tedious. Hey, it's your blog...

My bias is towards evolutionism (with a fairly large grain of salt, mind you), but that's actually a very convincing argument. Being able to address the genetic shifts we find in ourselves over time without assuming that it is necessarily forward progress is an interesting direction to take things.

But I also find myself on occasion thinking that intelligent design from a creator god by way of evolution would be far more impressive than the smoke-and-mirrors magic of snapping one's fingers. In that case, an evolutionist could be a creationist at the same time, and could have a much deeper awe and respect for the complexity of the design.

During those times when I feel that there must be a god, it is a direct result from thinking 'what if the theory of evolution is completely right?'.

Being able to address the genetic shifts we find in ourselves over time without assuming that it is necessarily forward progress is an interesting direction to take things.

Absolutely. There's tons of variation that can be attributed to non-evolutionary (in the sense of increasing information spontaneously arising) processes. Think of all the breeds of dogs and all the different kinds of people there are. All of these are explained by causes well within the purview of genetic variation. I've always believed in evolution in the sense of natural selection and genetic variation (like in the quote above). That's one of the reasons I thought the passage above was really neat -- it clearly explained what I'd been trying to explain before but didn't seem able to get across.

In that case, an evolutionist could be a creationist at the same time, and could have a much deeper awe and respect for the complexity of the design.

I don't view it as "smoke and mirrors" (which would imply that the creation was somehow illusory), and I wouldn't see it as "more impressive" if God took millions of years making infinitesimal changes to what He started with to finally "get it right", rather than showing His awesome creativity and power in a short flash of creative energy.

Keep in mind also that it's not just biology that He set up, but all the laws of physics, chemistry, etc. It seems arbitrary to say that it would be more impressive if God had created all the complexity of the universe, but stopped at living systems and did that slowly.

I'd argue, however, that it's inconsistent to be a Christian and believe in evolution. Not, mind you, that one couldn't be a sincere Christian and mistakenly believe in evolution, but just that it'd be inconsistent for him to do so. If you're talking about a non-Christian version of creationism, then I guess so.

During those times when I feel that there must be a god, it is a direct result from thinking 'what if the theory of evolution is completely right?'.

Weird, I get that sense, rather, when I look around and think "Wow, God made all this!!"

The scientific observations support the creationist interpretation that the changes we see are not creating new information. The changes are all within the originally created pool of information of that kind; sorting, shuffling or degrading it.

What I'd like to know is, simply: what form would this new information take? And why is natural selection incapable of creating something new?

why is natural selection incapable of creating something new?

That's simple. Natural selection only selects from the existing gene pool. Natural selection is not a creative mechanism. Mutation is the only "creative" mechanism in evolution, and it's never ever been shown to create new information. In fact, we know from information theory that randomness cannot produce new information. In addition, we know that things tend toward disorder, that adding randomness to order destroys that order, and that mutation is almost always harmful, or at best neutral. So it's not only very unlikely that mutation is a viable creative mechanism for evolution, but from everything else we know from every other science, it's downright impossible. Not only is mutation not a constructive or creative mechanism, but to the contrary it's known to be a destructive mechanism.

what form would this new information take?

I'll use the analogy from computer programs again, since that's the best real-world analogy I can think of, and I think it's very appropriate. New information in a computer program would be instructions for a new feature. And a new feature in a computer program can't be made with a simple flipping of a bit. In fact, if you flip a bit (which is analagous to a point mutation), you're likely to make the program crash -- the analogy of computer programs to living organisms is very apt in this way as well. So, a new feature in a computer program needs an intelligent programmer to specifically design and program new instructions into the computer.

We naturally understand this for computer systems -- that it takes an intelligent designer to create things. But for some reason people think a random process can create something more complex than anything we've built with the best of our intelligence. That doesn't make any sense to me.

In fact, we know from information theory that randomness cannot produce new information. In addition, we know that things tend toward disorder, that adding randomness to order destroys that order, and that mutation is almost always harmful, or at best neutral.

Do you have anything to back up this assertion? And please don't cite the second law of thermodynamics: it states that entropy in closed system cannot decrease, yes, but the Earth is not a closed system. It's bombarded with energy by the sun every day.

We naturally understand this for computer systems -- that it takes an intelligent designer to create things. But for some reason people think a random process can create something more complex than anything we've built with the best of our intelligence. That doesn't make any sense to me.

Do a Google search for genetic algorithms. People have been evolving computer algorithms since the 1960s, and some of the resulting algorithms have proven as efficient, or even slightly more efficient, than human-designed programs.

Just to clarify my view more, my image of a god intelligently creating via evolution would include all the laws of physics, etc, and not be limited to setting up biological laws.

What I'd find impressive about it is not that he would be constantly trying to 'get it right' with incremental revisions, but that he'd be able to set up the initial state of the universe and sit back to watch it unfold, and that stars, planets, the primordial ooze, life, Snapple, and Ginger Snaps would simply be guaranteed to result from his initial design (not necessarily to be confused with determinism, though).

Sam:

About genetic algorithms, I think the point that a creationist might make about that is that the information that results from the algorithm, ie. the specific weights to apply to various nodes in a neural net, were already there in the initial generations, and therefore you really haven't created anything "new". A particular configuration (or a particular human's genes) are the emergent behavior of shuffling, trial, and error. A human could have come up with the right weights himself, but practical laziness leads us to have genetic algorithms do it all for us. That's more of a "directed brute force" than "evolution".

What I'd find impressive about it is not that he would be constantly trying to 'get it right' with incremental revisions, but that he'd be able to set up the initial state of the universe and sit back to watch it unfold

Oh, I took your meaning to be that God worked his creation through evolution intentionally -- that He intervened in the natural course of events to effect evolution, not that He set up a system to lead inexorably towards the emergence of Man.

I'd find that strange though, and since I see no evidence for evolution I wouldn't believe in it anyway. Plus, I think that's contrary to Christianity, so I personally couldn't believe that way.

I've actually been in the middle of writing a relatively large comment to respond to Sam, but since we're talking about heavy stuff I wanted to make sure I have all my facts correct first. But I'll say something about genetic algorithms right now... "directed brute force" seems like an interesting way to describe genetic algorithms. For many reasons, genetic algorithms are not analogous to evolution at all. I'll make a more specific argument in my comment later, but the processes are really not similar at all. I'm actually very interested in genetic algorithms, and I've been considering that if I go on to get a master's degree in computer science that that might be a very fruitful area for me to write my thesis on. But it seems to me that people who cite genetic algorithms as analagous to evolution just go on the words "genetic" and "evolution" and conclude that it's somehow related without really understanding how genetic algorithms work (or maybe how evolution supposedly works, for that matter).

But it seems to me that people who cite genetic algorithms as analagous to evolution just go on the words "genetic" and "evolution" and conclude that it's somehow related without really understanding how genetic algorithms work (or maybe how evolution supposedly works, for that matter).

All of the literature that I've read on genetic algorithms (caveat emptor: I'm not an expert by any means), suggests to me that these computing scientists are explicitly basing their endeavours on their understanding of evolution and natural selection.

Pretty much every account of genetic algorithms has read as follows:
1) We assume that evolution works, with the main factors being mutation of DNA, recombination and survival of the fittest. We'd like to apply that model to computer code.
2) We're going to try and model all of those in software
3) It takes some trial-and-error tweaking to work out how best to model evolution / natural selection.
4) Once you've got those numbers tweaked succesfully, the evolution / natural selection model is pretty damn powerful for evolving computer algorithms.

They are explicitly basing themselves on the standard understanding of evolution and natural selection, and finding that such an approach reaps significant dividends.

As I understand it. Please prove to me where I am wrong.

Keith wrote, regarding God influencing evolution:

since I see no evidence for evolution I wouldn't believe in it anyway. Plus, I think that's contrary to Christianity, so I personally couldn't believe that way.

How can you even dare to claim a scientific basis for your theories, when you a priori declare that you will not accept any conclusion, even logically proven, that contradicts your religious beliefs?

Patrik Lioi suggested that:

the information that results from the algorithm, ie. the specific weights to apply to various nodes in a neural net, were already there in the initial generations, and therefore you really haven't created anything "new". A particular configuration (or a particular human's genes) are the emergent behavior of shuffling, trial, and error. A human could have come up with the right weights himself, but practical laziness leads us to have genetic algorithms do it all for us. That's more of a "directed brute force" than "evolution"

First of all, I wasn't talking about neural nets. I was talking about successive generations of randomly-mutated, and then recombined, algorithms. But I accept your point that computing all of these combinations takes time.

Secondly, what is this difference between "directed brute force" and "evolution"? I would object to the word "directed" - one of the major advantages of Darwin's theory of evolution is that it does not rely on any teleological forces. But I would embrace 100% the term "brute force" - how else is the survival of the fittest to be measured? (Bearing in mind that the fittest can also mean the smartest, or best adapted to a particular environment.)

> In fact, we know from information theory that randomness cannot produce new information. In addition, we know that things tend toward disorder, that adding randomness to order destroys that order, and that mutation is almost always harmful, or at best neutral.

Do you have anything to back up this assertion?

Which one? That randomness cannot produce new information? That seems uncontroversial. That adding randomness to existing information destroys that information also seems uncontroversial. That's how we destroy data from hard drives Both of these are basic parts of information theory, AFAIK. Mutation always being harmful (or neutral at best)? That's a fundamental aspect of mutation. Mutation always involves a loss of information, not a gain of any new information. Even times when it's "beneficial", as in sickle-cell anemia protecting against malaria, it's still through a destructive change in the organism from pre-existing information.

And please don't cite the second law of thermodynamics: it states that entropy in closed system cannot decrease, yes, but the Earth is not a closed system. It's bombarded with energy by the sun every day.

At the time I wasn't thinking of thermodynamics. I was primarily thinking of information theory, though I suppose my "we know that things tend toward disorder" came more from thermodynamics. But I've never really liked the open/closed distinction as an argument showing how the second law doesn't apply to living systems, because it seems to be arbitrary at what level you consider the system "open" or "closed". Sure, the earth gets energy from the sun, if you want to take your reference point to be such a large system, but things on the earth lose energy even though they're part of that larger system. In addition, I've seen quotes from scientists saying that the open/closed distinction doesn't make one bit of a difference and that the 2nd law always holds.

Even so, the important thing for this discussion is order and complexity, and simply adding energy to a system isn't sufficient to produce either. Think about it this way... I can add energy to your "system" by bombarding you with X-rays. But that won't increase the order in your body. Rather, that'll likely cause harmful mutations in your body, give you cancer or radiation poisoning, and kill you. An analogy I've often heard repeated is that of a bull in a china shop. It adds lots of energy, but is destructive, not constructive. Any time order and complexity are found, it took some kind of machine, intelligence, or ordered process to create, in addition to an input of energy.

As for genetic algorithms. I think they're very cool, and yes, they do try to model natural selection, though on a much larger and vastly accelerated scale then we find in nature. Though it's also very artificial -- what's "selected" isn't selected the same way as things in nature are. In fact, it takes an intelligent process to get something useful out of it. This example seems kind of silly, because it's built by intelligent programmers and directed by them, unlike in nature, and people want to claim that it's analagous to how nature works.

Keep in mind that even the "mutations" that take place in a genetic algorithm aren't similar to those that take place in a living system. As I understand it, the code of the simulation using the genetic algorithm doesn't change - it's rather the variables that they set the system up with. And the "creatures" in a genetic algorithm simulation don't have the same "survival pressures" that living creatures have.

How can you even dare to claim a scientific basis for your theories, when you a priori declare that you will not accept any conclusion, even logically proven, that contradicts your religious beliefs?

Well, first of all, I don't claim a scientific basis for my "theories". I understand that theories of origins, like creation and evolution, are outside the realm of scientific inquiry. Like the quote says above, we each have different "histories" through which we "interpret the science and facts of the present". In fact, there are many aspects of evolutionary theory that aren't scientific. Even natural selection isn't defined sufficiently... it's essentially defined tautologically. The ones that are fittest are the ones that survive, and the ones that survive are the ones that are fittest. Of course! Natural selection isn't a scientific theory either, but I think it's a fine interpretation and a useful framework to understand what goes on in nature.

In fact, I think belief in Christianity frees me to be more objective about this issue, since I don't have any vested interest in believing that we evolved from goo. Dust of the ground or primordial ooze, at least I'm explicit about my religious beliefs, while evolutionists deny that they're believing in their own religion.