What is evolution?
Evolution, in organisms, is the process by which inherited traits change over time.
What is natural selection?
Natural selection is the non-random survival of randomly varying, hereditary mutations.
There are several essential points here.
First, it’s the mutations which vary randomly, not the survival. It’s quite easy to see that if the survival were random, then evolution would be impossible. Any complexity would be the result of extreme luck, and nothing more.
Secondly, mutations must be heritable. If they are not, selection can’t occur since all mutations are simply lost.
What do you mean by mutation?
A common creationist quip is that monkeys can’t mutate into plants, or some variation on that. That’s of course complete hogwash; it’s simply not how evolution works. Mutations occur during the copying of DNA. This can come in different forms, but certainly it never comes in the form of a monkey morphing into a plant.
First of all, you have to understand two different processes: meiosis and mitosis. You might be familiar with these processes; you learned them in middle and/or high school, no doubt. These are two methods of cellular reproduction (i.e. replication).
There are two general ways cells replicate themselves in eukaryotic organisms. One way is the obvious way: just make (or try to make) an exact copy. This method is asexual; a single cell may replicate itself. It’s easy to see that all mitotic cells of a certain type will be overwhelmingly identical; they have the same set of genes. Thus a mitotic mutation is generally a local phenomenon, one which the body usually can deal with. I say generally, because of course we have things like cancer which are certainly global phenomena.
Meiosis, on the other hand, is at the heart of baby-making in sexual organisms. Meiosis is the process of creating sex cells, or gametes. Gametes are different from normal cells because they each have a unique set of genes. In humans, these gametes are of course the sperm and the egg, and each of them contain exactly half the number of chromosomes of a normal cell. When a sperm fertilizes an egg, it creates a new cell called a zygote which, given certain circumstances, will eventually develop into a new organism with its own unique genome (comprised of the chromosomes of the egg and the sperm).
As you can see, meiosis is at the root of genetic variation and genetic mutation. A mutated gamete will cause consequences far down the line for the organism. That is, of course, assuming the mutation doesn’t make the gamete malfunction in some way. That assumption is, of course, naturally selected. It seems obvious, but an organism that doesn’t develop can’t pass on its genes. Therefore those really bad mutations are weeded out at the get go.
So as you can see, the creationist line makes little sense. The mutations which contribute to evolution develop long before infancy. Monkeys don’t mutate into plants.
What selects beneficial mutations and how does it know they are beneficial?
You have to be careful not to overly interpret the words we use. When we say mutations are “selected” we don’t mean that any consciousness is involved.
That being said, it is the environment which selects mutations. But keep in mind that it selects the same way a sieve selects, that is without any conscious regards to the contents. A sieve lets smaller clumps of flour pass through, while leaving the bigger clumps. It has no consciousness of its selective process. Natural selection works upon a similar lack of consciousness.
Imagine that we are in the arctic, and we’re an animal that mutated a thinner fur coat. We would likely freeze to death more often than our competitors who have thicker coats. In this case, the environment selects against thin coats and for thick coats.
But take this same mutation and transplant it into the deep tropics. Now the environment would select for thin coats and against thick coats. So you see, speaking of evolution without regards to the environment, and hence natural selection, makes no sense.
How does evolution arise?
As we’ve seen, all evolution requires is hereditary mutation and natural selection. Therefore first we require some kind of molecule which can make copies of itself most of the time. That is, to get evolution rolling, a replicating molecule has to make some mistakes. No mistakes means that the molecule population remains homogeneous, which means no selection occurs. Of course, as these replicating molecules do in fact evolve, it would be a bad thing to make mistakes a lot of the time. Naturally, then, copying fidelity becomes something that will benefit you. Evolving complexity is a bad idea if you can’t replicate that complexity with very few errors.
So what I’m saying is, we need these molecules to exist first before evolution may occur. Evolution technically says nothing about the origins of such molecules. Now, the religiously-minded may take that as a cue to invoke God; God created these molecules, right? I’m an atheist, so I would say no, but you’re free to speculate, I suppose. Richard Dawkins, in The Selfish Gene, describes how such a molecule might arise naturally:
Actually a molecule that makes copies of itself is not as difficult to imagine as it seems at first, and it only had to arise once. Think of the replicator as a mould or template. Imagine it as a large molecule consisting of a complex chain of various sorts of building block molecules. The small building blocks were abundantly available in the soup surrounding the replicator. Now suppose that each building block has an affinity for its own kind. Then whenever a building block from out in the soup lands up next to a part of the replicator for which it has an affinity, it will tend to stick there. The building blocks that attach themselves in this way will automatically be arranged in a sequence that mimics that of the replicator itself. It is easy then to think of them joining up to form a stable chain just as in the formation of the original replicator. This process could continue as a progressive stacking up, layer upon layer. This is how crystals are formed. On the other hand, the two chains might split apart, in which case we have two replicators, each of which can go on to make further copies.
While conjectural, this scenario seems realistic. After all, we observe the random formation of different types of amino acids and other organic compounds all the time when we look at comets and other celestial bodies. This initial replicator may in fact be quite improbable; but as Dawkins says, it only has to happen once.
Can we observe evolution in progress?
It depends. Hypothetically, such a thing would require the observation of several hundreds, if not thousands and thousands, of generations. That’s the only way we’ll observe which mutations are selected for. In organisms such as ourselves, that would be a daunting task indeed. So we can witness (and have witnessed) evolution in those organisms which have short reproductive generations. If you’re interested, I urge you to visit this listing, which has some of those experiments.
If we switch from physical experiments to modeled experiments, we can certainly verify the veracity of these ideas. For instance, we can set up a digital environment using cellular automata and observe what kind of complexity arises. This kind of idea is used in adaptive computer-aided designs (CADs), where the engineer inputs parameters he wants to maximize or minimize (for example, minimize the amount of metal or wiring used, but maximize the efficiency), and the computer runs through testing generation after generation of design, selecting the ones that do minimize and maximize those things. (Of course, both of these scenarios replace natural selection with artificial selection; instead of the environment selecting, human beings program computers to select.) This kind of digital evolution has shown without a doubt that apparently irreducibly complex structures are in fact reducible (and there are physical experiments which show it, too).
Are mutations either beneficial or harmful?
Nope! Mutations can also be selectively neutral. In fact, most mutations are that way. This subset of evolutionary biology is what’s called the Neutral Theory. A selectively neutral mutation doesn’t affect an organism’s fitness. This means that such a mutation will dominate the gene pool only if its host organism becomes inordinately successful at reproducing. In other words, we get into the realm of luck. This is what is meant by genetic drift.
Why was this post so long?
That’s how it evolved.