Evolution is a non-random directed process

[ProfMoriarty]

The notion of Intelligent Design is often raised as a potentially acceptable alternative to evolution. The claim is often made based upon the following supposition:

– evolution is a random process, so how could it have produced the complex organisms we see today - but an Intelligent Designer could have intervened and directed evolution to produce all the world's creatures and man.

However, evolution is not a random process.

The issue I want to raise for discussion is:

"Darwinian evolution is a non-random directed process. "

This is strictly in accord with both Darwin's original discussions in On The Origin Of Species, and also with current evolutionary theory, and is what enables evolutionary theory to be able to explain adaptive change and speciation.

[ProfMoriarty]

However, I was more questioning the entire process of evolution being non-random. If evolution (mutation + selection) is not random, what are the implications?

We do need to keep in mind that by "directed" we are only talking about very localised direction towards a local optimisation of the genotype. Optimisation doesn't imply perfection, just the best combination of genes for the local environment. So it isn't that we are saying that evolution is aiming at global perfection for all organisms.

There does seem a slight confusion here over the mutation part. Mutation is most definitely RANDOM and is not directed by anything (at least according to current theory). However by mutation we are not just talking about actual genes changing, which is rare, but the combination of existing genes changing, which is normal and happens every meiosis. The combinations of random recombination and rare random mutations give a random set of genes in the offspring (all offspring in the population, we are not discussing individuals here). But the spread (range) of the mutations is limited by the initial genotype of the parents. As I mentioned in an earlier post, a population of cow-size animals would be expected to have cow size offspring. So selection limits the available mutations to be randomly ranged around the central trait of cow-size. Not mouse-size to elephant size.

Also it is important to remember this is occurring in retrospect. Selection does not say "I want 'bigger' genes" and they appear; it simply allows 'bigger' genes to succeed better than 'smaller' genes so the next generation will be slightly bigger. In their offspring the biggest will be bigger yet again than the parent, and the smallest could be slightly smaller than the parents.

Hypothetical case

Gen 1 - range average 4.0ft +/- 1 inch.
Gen 2 - range average 4.1ft +/- 1 inch
Gen 3 - range average 4.2ft +/- 1 inch
Gen 4 - range average 4.3ft +/- 1 inch

In each case the upper AND lower range is increasing by 0.1ft per generation, because the bigger side of the range is more successful than the lower side, even if only by a very small amount.

I don't think there are any particular implications to this, other than that it contrasts to the creationist claim that evolution is entirely random. As it is this supposed randomness that a lot of less open-minded and rigorous creationists hook on to as a straw man argument as to the invalidity of evolution, I thought it was sensible to bring it up. I also thought that this forum would be a good place as it appeared that there were a good number of christians here who could debate the facts rather than the received wisdom, and this appears to have been the case, as the discussion so far has shown.

Personally I think that evolution is compatible with a non-literal creation; i.e. thinking that God started it all. Not my personal belief but I am not averse to others holding this view. I do find Intelligent Design a bit lame, as it would be supposed that if God was a good designer he would have set up evolution to run all by itself, which is exactly what it does, as opposed to tweaking it every so often. I'm sure Paley's wristwatch would have kept very good time.

So this argument basically says: evolution is viable because it can be shown to direct variation and adaptation towards optimisation with the environment. And also it it not incompatible with a rational creationism as it is in fact the scientific explanation of what a good designer would have put into place.

Just because I and others do not believe that there is a need for a designer or that certain christians do believe that God designed/created it does not change the fact that evolution is occurring, so the most rational position is one that can explain both sides views with the minimum of massaging. It might even allow this overall rather destructive debate (creation/evolution, not this particular thread) to die down if common ground could be found.

[Jose]

Unfortunately, I cannot read it since my firewall blocks port 16080. The firewall only allows the standard http port 80.

That's weird. Oh well...let's use your coin analogy.

Let's say you are given coins randomly. You end up with a bunch of different coins.

Now, you go somewhere, and want to set up a business. In the new place, they only accept quarters. As you set up your business, you use the quarters. Your business is successful, and you earn more quarters. Pretty soon, you've forgotten all about those other coins--you've probably stopped carrying them around, even.

On the other hand, if you went somewhere else, they might use only dimes. Here, your business would be based on dimes, and you'd earn more dimes. You'd forget about the quarters.

Of course, if you had no quarters in the first location, and no dimes in the second, you wouldn't be able to set up a business at all, and you'd have to leave.

So: your diversity of coins was determined randomly.

In one environment, quarters were useful. As time went on, it got to the point that just about all you had was quarters.

In a different environment, dimes were useful. As time went on, it got to the poiint that just about all you had was dimes.

If you didn't happen to have the right diversity of coins when you went to the new environment, you couldn't make it--went extinct, so to speak.

In other words, a random diversity of coins is fine. The conditions of the environment you are in determine which coins are useful--that is, the environment selects what is useful from the overall diversity.

We see something similar here, actually. Occasionally, I get "mutant" coins--Canadian or Mexican or British or whatever. They don't work here, so they are "selected against." But, if I went to Canada or Mexico or England, these coins would be "selected for."

So it is with new mutations in living things. Usually, in the original environment, they don't confer any particular advantage, and are "neutral," or even selected against. If they aren't selected against too rapidly, they become part of the genetic diversity of the species, part of the "gene pool." Sometimes, though, when the environment changes, mutations (or particular genetic variants) might be advantageous, and be selected for. Particular genetic variants that were advantageous in the old enviroment might not be advantageous in the new environment, and be selected against.

The greater the genetic diversity, the more "choices" there are for selection to work with. It really doesn't matter if the diversity is random, because the environment selects those variations that work. The variations that don't work die out. Sometimes, the environment changes, and a species doesn't happen to have genetic variations that allow it to survive in the new conditions...and the species goes extinct. It seems to me, the fact that so many species have gone extinct is pretty good testimony to the fact that it's impossible to predict what's going to happen in the future, and impossible to plan ahead and get just the right genetic diversity. You're better off with a diversity of random coins than you are with just quarters, because you don't know whether the new place will be quarter-world or dime-world. With random coins, you'd have a chance either place, or even in nickel-world!

[perfessor]

I think the problem is a lack of a definition of random.

I think of random as not being able to predict the outcome with certainty. When something is random, you can only give the odds of a certain outcome. When something is not random, you can definitively predict the outcome.

I think I see our problem. This is from dictionary.com:

random: adj. 1. Having no specific pattern, purpose, or objective: random movements. See Synonyms at chance.
2. Mathematics & Statistics. Of or relating to a type of circumstance or event that is described by a probability distribution.
3. Of or relating to an event in which all outcomes are equally likely.

I think the key definition is #2, in which the outcomes have a probability. If you flip an honest coin, you get a random result. If the coin has 2 heads, the result is deterministic - only heads will result. If the coin is weighted, so that heads come up 60% of the time, we have a situation in between: All outcomes are not equally likely, so it doesn't fit definition 3. But neither is the result determined; there is a probability function associated with it, as in definition 2.

To go back to your keeping-the-coins example: suppose when your friend hands you a coin, you apply a rule: if it's a penny, nickel, or dime, you roll a die and keep the coin if you roll a 1 or a 2. If he hands you a quarter, you keep the coin if you roll a 3, 4, 5, or 6. At the end of the day, you will have a plurality of quarters, even though the decision to keep or not was random - subject to a probability function.

If you count up your coins, and tell your friend that the pool of tomorrow's coins will start out with the percentages from today, then the plurality of quarters will be even greater at the end of day 2. After some number of days, you will have a vast majority of quarters.

I think this matches natural selection rather well.

Is there an objective criteria for which the result is ordered by natural selection? How can one determine what would cause something to survive or not? If one can objectively do this, then I can see how it can remove the random component of mutations.

The objective criterion is: does the individual survive long enough to reproduce? This is still subject to a probability function; a hereditary feature may have a negative, neutral, or positive influence on the organism's chances for success. It is not deterministic - a very "clever" mouse may still get taken by a lucky owl. So - "random" by definition 2, but not by definition 3.

In order to do this, one must be able to show that only certain mutations will result in a chance of survivial in a given environment prior to a mutation occuring.

This sounds like a strawman. Why "prior"??? There is no "prior" knowlege or intent, nor does there need to be. Mutations do occur. The mutations are random. If the mutation affects the survival rate of individuals carrying the trait, it will be selected for or against. It will skew the probability function up or down.

However, I was more questioning the entire process of evolution being non-random. If evolution (mutation + selection) is not random, what are the implications?

For me, the implication is that random mutations can lead to ordered results. The implication is that no design or direction is necessary.

[otseng]

I don't think there are any particular implications to this, other than that it contrasts to the creationist claim that evolution is entirely random. As it is this supposed randomness that a lot of less open-minded and rigorous creationists hook on to as a straw man argument as to the invalidity of evolution, I thought it was sensible to bring it up.

I don't particularly see how if evolution is random that it can be used to argue against evolution. So, I can't get too worked up about arguing about evolution being random or not. Perhaps that explains why this discussion has been able to maintain composure.

I also thought that this forum would be a good place as it appeared that there were a good number of christians here who could debate the facts rather than the received wisdom, and this appears to have been the case, as the discussion so far has shown.

And I hope other discussions here on CvE can do the same. I always have a nagging fear that discussions here can go the way of other debate forums on CvE and fall into uncivil debates. My hope is to provide a place for logical and respectful discussions on CvE (as well as all topics pertaining to Christianity). But I also realize that this is a lofty goal and one that few, if any, has been able to achieve. But, I am pleased with the direction and atmosphere with the forum so far. And I hope that it can continue.

Personally I think that evolution is compatible with a non-literal creation; i.e. thinking that God started it all.

Actually, I would guess that the majority of Christians think the same.

However by mutation we are not just talking about actual genes changing, which is rare, but the combination of existing genes changing, which is normal and happens every meiosis.

This I think is more of an important area to debate than whether evolution is random or not. Could I suggest if you could start a thread on this and explain this further? Is it true that only gene mutation can result in new morphological features? Or can gene recombination also explain for it?

[ProfMoriarty]

I don't particularly see how if evolution is random that it can be used to argue against evolution. So, I can't get too worked up about arguing about evolution being random or not. Perhaps that explains why this discussion has been able to maintain composure

Because usual arguments are such things as "how can a random process make living organisms out of nothing, the odds against that are phenomenal". Or the junk-yard argument - "If a hurricane blew through a junk-yard full of random aircraft parts you wouldn't expect it to assemble a Boeing 747 from them". By showing that evolution is not only non-random but directed towards adaptation, it can be shown that the odds are in favour of adaptation rather than against it.

Is it true that only gene mutation can result in new morphological features? Or can gene recombination also explain for it?

I think a thread about this would be a good idea. As you just said, in fact most christians do not have a problem between evolution and faith. In the UK we have a highly non-religious population these days, and it is generally the case here that christians find no issue with evolution. The main problem I find is the die-hard fundamentalist creationists, most often from the US, who wilfully spread disinformation and muddy the waters for everyone, christian an non-christian alike, and this tends to sow distrust of the whole structure of science. Going into correct explanations of evolutionary theory is a useful way to counter this.

[otseng]

Because usual arguments are such things as "how can a random process make living organisms out of nothing, the odds against that are phenomenal". Or the junk-yard argument - "If a hurricane blew through a junk-yard full of random aircraft parts you wouldn't expect it to assemble a Boeing 747 from them". By showing that evolution is not only non-random but directed towards adaptation, it can be shown that the odds are in favour of adaptation rather than against it.

However, the Boeing 747-from-junkyard argument is addressing randomness from the mutation step, not the entire mutation+selection steps. So, I think that it is a separate issue.

Perhaps it can be addressed in the yet-to-be-created "what causes new morphological features?" thread (that I hope somebody will create).

[Jose]

However, the Boeing 747-from-junkyard argument is addressing randomness from the mutation step, not the entire mutation+selection steps. So, I think that it is a separate issue.

As I have heard it from creationists, it applies to the entire process. If they mean it to apply just to mutation, they are very good at hiding it. For example, here are just a few examples from a quick Google search:

"To believe natural processes assembled a living cell is like believing a tornado could pass through a junkyard containing the bits and pieces of a airplane, and leave a Boeing 747 in its wake, fully assembled and ready to fly!" ( Evolutionary Improbabilities )

"Hoyle comments, “The current scenario of the origins of life is about as likely as the assembly of a fully functional (Boeing) 747 by a tornado whirling about in a junkyard.” The Darwinian theory of evolution fails to predict what we actually currently observe. Schutzenberger, a mathematician writes, “There is a considerable gap in the neo-Darwinian theory of evolution, and we believe the gap to be of such a nature that it cannot be bridged by the current conception of biology." ( Pilgrim Tours )

" For example, the mathematician D.S. Ulam argued that "it was highly improbable that the eye could have evolved by the accumulation of small mutations, because the number of mutations would have to be so large and the time available was not nearly long enough for them to appear." [2] Famous astronomer Sir Fred Hoyle conceded that the chances that life just occurred on earth are about as unlikely as a "tornado sweeping thorough a junkyard and constructing a Boeing 747." [3]" ( Matt Perman's Evolution: Fact or Fiction? )

"Consider this fact: there are 2,000 complex enzymes required for a living organism, but not a single one of them could have formed accidentally. As Fred Hoyle has put it, "The chance that higher life forms might have emerged in this (accidental) way is comparable with the chance that a tornado sweeping through a junk-yard might assemble a Boeing 747 from the materials therein."" ( Samuel L. Blumenfeld's Theory of evolution: Fact or fairy tale? )

What Hoyle was actually discussing was the origin of life, not the evolution of life once it arose. Yet, it seems that his actual words have mutated with the telling. Those who use this argument are clearly indicating that evolutionists think that life, and higher life forms, and currently existing enzymes formed accidentally by random events. They lump all of evolution together, even with the origin of life, under the concept of randomness.

By putting all of this together with the 747 argument, the creationists both mis-use Hoyle's statement, and mis-represent the role of randomness (mutation) and the role of direction (selection) in evolution. They say that it's all randomness and accident. Most likely, this results from the simple fact that the actual mechanisms of mutation and selection, and the distinction between them, are rarely understood very clearly. We who teach it are partly to blame, and can certainly do better. However, I think a larger share of the blame rests on those we might call "evangelical anti-evolutionists" who perpetuate the misconception expressly so that arguments like the 747 idea will resonate with their listeners and their readers.

[otseng]

However, the Boeing 747-from-junkyard argument is addressing randomness from the mutation step, not the entire mutation+selection steps. So, I think that it is a separate issue.

As I have heard it from creationists, it applies to the entire process. If they mean it to apply just to mutation, they are very good at hiding it.

Perhaps you are right, so I'll just argue this from my perspective.

Using a modified form of the 747 argument, evolution would be like the initial point being a bunch of random parts lying together. Over time, something caused the parts to form a complex object. A paper airplane. Once the first complex object got formed, mutations caused it to become more complex. A balsa wood glider. The flying conditions were such that the balsa wood gliders flew better than paper airplanes. The gliders, through mutations, formed a rubber band engine. This self-propelled airplane flew better than the glider. Then on and on the mutations occurred over successive generations until a 747 was formed.

The mechanism that formed each successive airplane was random mutations. The selection process only had a part in selecting the airplanes that got generated. So, the generation of the new airplanes itself was only a function of random mutations.

Certainly the step from a bunch of parts to 747 is a result of mutations and selection. But, each successive step in the process depended on random mutations to create a better plane.

Is the overall process from parts to 747 random? It can be argued no. But the generation of each successive airplane is random. And, as I mentioned before, I think that is where the real debate lies. How does each successive airplane (with new physical features) get generated?

[ProfMoriarty]

Is the overall process from parts to 747 random? It can be argued no. But the generation of each successive airplane is random. And, as I mentioned before, I think that is where the real debate lies. How does each successive airplane (with new physical features) get generated?

I don't think it can be called random, but it BLIND, which is an important feature of evolution. So you might not get a 747. You might get an Airbus, or a Lear Jet. You will get something, because selection directs evolutionary change to the most optimal form, but you could not predict the end result from the beginning.

[Jose]

Is the overall process from parts to 747 random? It can be argued no. But the generation of each successive airplane is random. And, as I mentioned before, I think that is where the real debate lies. How does each successive airplane (with new physical features) get generated?

I see the problem now. I also see how the 747 analogy makes it seem so improbable. There is no way that we can imagine that a "mutation" could turn a paper airplane into a balsa wood one. There just isn't any mechanism for it.

On the other hand, airplane engineers do use a form of random mutation in developing new designs. The idea is to give a computer all of the characteristics of an airplane, along with the mathematical parameters for how wind produces lift, drag, or turbulence. Then, allow the computer to produce zillions of alternative designs with random differences from what you started with. If any happen to improve the performance of the airplane, then save these--and repeat the process. After many iterations, we come up with new designs, some of which are fairly counter-intuitive to the designers, but work really well when models are built and tested.

The critical feature of the random changes is that they are small changes to what already exists. Each one is just a tweak here or there.

Rather than use analogies, it may be necessary to talk about actual organisms. There, at least, we can imagine changes in DNA as the mutations, and changes in protein function as the result of the changes. The mechanisms make sense.

How about considering the development of bat wings from the forelegs of a small mouse-like ancestor, and the development of whale flippers from the forelegs of a hippo-like ancestor. These are significant changes. How might they be possible if they are based in random mutation?

Here are some important points:

1. Mutations occur at some low frequency throughout the genome. They cannot be targeted to any particular place. This is the randomness that concerns us. How can pattern come from it?

2. Some of these random mutations may happen in the genes that control the patterns of limb development. These include the Hox genes, shh, fgf, and a bunch of others that establish the bone structure, and some other genes that are involved in the cell death program that separates the digits after they have been formed. Only mutations in these genes would be relevant to changes in limb morphology. Right here, we see that although mutations occur in DNA at random, there is already non-randomness showing up: only certain genes can give rise to certain kinds of changes.

3. Most mutations in these particular genes would screw things up. Those mutations would be selected against (probably by producing individuals that can't use their limbs, or that have more serious problems with development). That is, individuals that are born with these deleterious mutations would be likely to die early. Here, we see another non-random aspect showing up: "bad" mutations are eliminated because the organisms that carry them don’t make it. So, evolution of limb morphology due to random mutation has become evolution by only "good" mutations in "just the right genes." This is an important point.

4. Some mutations would not screw things up too much. These mutations, once passed on to offspring, become part of the genetic diversity of the population. If they happen to be advantageous, then they may be selected for. One such mutation would be one that prevents the cell death program in tissue between the digits. (There is such a mutation kicking around in human populations, by the way, so we know that such mutations can occur.) For small mouse-like creatures or larger hippo-like creatures, webbed hands might work just fine. In fact, for hippos, they might make it easier to move around under water, because it wouldn't be quite so necessary to push off from the river bottom. The webbing might push against the water a bit. So, these particular mutations might be helpful, and be selected for. Out of all of the zillions of mutations that could have happened all over the genome (and may have happened) these are the only ones that provide this particular advantage so these are the only ones that are selected for in this particular way.

5. Another mutation could occur that changes the shape of the limb. In the hippo-like creatures, we might get a mutation that results in the repetition of the "digit-development" program. (Repetition mutations are well known in C. elegans, so the precedent exists). This would make more joints in the toes, making the webbed feet longer and more effective at paddling. In the mouse-like creature, we might get a mutation that makes the digits longer. This might be helpful when climbing trees. Again, lots of mutations happened, but only these were the ones that were helpful.

6. Let's add a couple of additional helpful mutations. The hippo-like creature has a mutation that shortens its forelimbs into flipper-like things. It's easier to paddle around in the water, so this is selected for. The mouse-like creature has a mutation that makes a bit more tissue between the digits. When individuals with this mutation jump from branch to branch, and open their fingers to grab the branch they land on, they get some aerodynamic lift from the webbing between their fingers. If they can jump better and farther, maybe they can escape predators better, or forage better.

7. By the sequential accumulation of mutations like these, that slightly alter the development of the limbs, changes would occur. Selection would enable these mutations to become common in the population. This is despite the fact that mutations occur at random throughout the DNA. "Bad" mutations are selected against; "good" mutations are selected for. Here, these few mutations that happened to affect limb development in these particular ways turned out to be helpful, and were selected for.

We can reconstruct this kind of scenario for evolutionary transitions that would lead to organisms that now exist. If we know the ancestor (a hippo-like creature) and the current animal (whale), we can propose a plausible series of events. Perhaps, if we find fossils of some of the intermediates, we can refine our proposal. When we make such a proposal, we are attempting to re-create what might have happened to achieve the endpoint we see today. Actual evolution didn't do it this way--it had no "endpoint" in mind. The mutations happened. Those that turned out to be helpful were selected for. Others were selected against. The particular environments of the animals determined which mutations were helpful and which were not. A different environment would have a different outcome. Just because we're "trying to make a whale" doesn't mean that evolution tried to make a whale. Mutations happened, selection happened, and whales are one of the results.

We can also imagine lots of other kinds of mutations that could occur, and that might have been selected for if conditions had been different. Maybe an ancient hippo-like creature might have acquired a mutation that made the lungs just a bit bigger, or that gave it just a bit more belly fat, so it floated better. Maybe this would have led to it floating on its back, instead of paddling along the river bottom. Maybe webbed fingers would have been helpful on windy days, so the animal could hold its hands up in the air and scoot along like a sailboat. As far as we know, however, these mutations didn't occur, or if they did, they were not selected for, so these morphological transitions didn't happen. That is: evolutionary change occurs if mutations happen to occur, and if those mutations happen to produce traits that are advantageous in the particular environment that the organism is in.

We also see, I hope, that mutations can only change existing features. The limbs in this scenario remained limbs, but acquired novel characteristics by subtle alterations of the developmental programs that construct limbs. The limbs did not turn into leaves or eyes. What is already present in an organism constrains what evolution can do to change that organism's descendents.

On the other hand, we aren't turning paper into balsa wood into metal, or rubber band into jet engine. In every life form, we're working with the very same molecular building blocks--nucleic acids, proteins, lipids, etc. There's no fundamental difference among different organisms; rather, it's just the way the basic building blocks are assembled. This actually gives real evolution a lot more flexibility than any of the analogies that we use to simplify particular concepts.