Some friends and I have been discussing evolution versus creationism (especially intelligent design theory). This article combines my thoughts, slightly re-edited for presentation here, in response to this argument for creationism:
Try this experiment: Take your computer apart completely, all the bolts, nuts screws, boards, chips. Everything that can come apart (hard drives included) without being ruined. Put it in the box it came in and shake it up and drop it out onto the floor. Graph how many times the computer falls together perfectly as it was before disassembled (so that it works as before) versus the number of times you shook the box and dumped its contents. If it does not, can you extrapolate a rough guess as to the possibility of that happening? Like one in a billion. Then, taking that, determine the potential number of parts to our solar system (I'll exclude the galaxy for a later assignment) and determine it's chances of falling together randomly to what we have today. Once you have that probability, compare that to the probability of there being a creator. Let me know what you come up with. The probability of God versus the probability of random chaos.
The example tries to debunk evolution but fails because it is seriously flawed. It argues that random chance instantiated us all at once -- essentially, that we came to exist in an instant of creation. Evolution doesn't work this way.
Evolution works on gradual change. If we took a 5-gallon bucket and mixed it with random size screws and nuts, and shook that up a while, we probably would find a few screws matching onto nuts within several shakes. That's the kind of baby-steps evolution took, not showing up one day and assembling fully operational mammals.
The second flaw is that the example assumes evolution is random, but evolution theory doesn't rely on randomness alone -- it also works on natural selection. When my cat, Jack, is out hunting, does he eat random squirrels in the neighborhood? Well, there's a degree of randomness in where Jack goes and where the squirrels are at a particular time. Nevertheless, the squirrel that has better hearing, better peripheral vision, better sense of smell, or better intelligence to tell him he's about to be eaten and better get up the tree, has a better chance of surviving, and therefore reproducing, versus the average squirrel. And the stupid, half-blind, deaf backward-evolutionary squirrel was in Jack's belly months ago.
The random mutations are happening all the time. If they're not agreeable, they just don't work -- about 15% of confirmed pregnancies spontaneously abort in the first trimester; most of these miscarriages are evolutionary dead-ends, although there are other causes too. If we look at total pregnancies, as many as 78% miscarry1 . (I'm not sure if this same figure applies to the animal kingdom, but whenever we grew baby chicks in grade school there were a handful that didn't hatch) Of the fetuses that make it, there's always some mutation in the DNA in addition to selecting half of each parents' DNA (there are 23 pairs of chromosomes, or 2^23 = 8,388,608 variations for each parents' contributions). Usually, mutations don't change much. If they do, it's usually something that is almost insignificant. Let's say, for example, that a mutation provides me with a 2% improvement in wrist flexibility resulting in a 0.1% improvement in not being eaten by tigers or killing myself falling out of a tree before I have kids. This does almost nothing for me, but consider the compounding as the generations go on:
Let's say the normal percentage chance of survival is 43.4%, so my chance is 43.5%, and there's an average of 5 kids/generation. Then the originals have an output of 2.17 kids/generation that survive to reproduce, and the mutations have 2.175. After 10 generations (~200 - 300 years), the mutations have the upper hand on population -- 54 more, or about 2.3% more population than the originals. After 20 generations, it's closer to 5%. Compound over 100 generations, the mutations have the upper hand by 25%, and after 1000 generations (20,000 - 30,000 years) the mutation is winning just shy of ten-fold.
That tiny, trivial change, which made no significant difference (0.1%) chance in my ability to survive, results in my heirs being the dominant breed in 30,000 years.
The last problem with the argument is the assumption the computer has to fit together perfectly, one way. For biological critters, there isn't One Way. Some people are built with their innards in backwards (situs inversus), in which internal organs normally on the left are on the right and vice-versa. Others have just some parts in backward (dextrocardia, for example), which can result in reduced lifespan -- it's not a good thing for evolution to try, and natural selection will eliminate these people from the gene pool over time. Furthermore, consider some of the conjoined twins that have existed: if things fit together precisely one way, they couldn't survive. But nature is very adaptive, over the short-term, and willing to give stuff a whirl -- and given the massive parallel processing power of the universe, or even our single planet, some of these mutations will result in new, successful stuff.
The crucial flaw with the argument is that it confuses development with construction.
Say I go out, buy parts, and assemble a PC. Did I build it in a day? Maybe it appears that way, but really, no: I just did the final assembly. Intel's been working on those chips for the last 30 years, evolving their complexity up from the 4004. That was based on AT&T's development of the PN junction, which was in turn an improvement on electronics which we'd been improving on since the invention of radio in the early 1900's. Before that, it was just a way of making gasless lights and motors, and originally just a way to make a bit of ferrous material twitch. So that PC that I assemble really took thousands of person-years to design and develop.
The software is no different: compiling might be the moment of making executable code, but the compiler doesn't just put random bytes together until something works2 , it uses the source code to know what to do -- sort of like our DNA is the program for building a creature. I can open up vi and write a "Hello, world!" program in 1 minute and compile it. But I'm not starting from scratch: I'm giving my existing UNIX system a new capability of saying "Hello, world!", building on top of the system libraries, the UNIX I/O system, the kernel. If we include all that stuff, writing "Hello, world!" took years of development and tens of thousands, maybe even millions, of incremental steps.
In the example, you're just talking about the build phase. Building a human being takes 9 months, but it isn't spontaneous - it works from the existing source code/blue print/database that is DNA. And according to the theory of evolution, developing that blue print took millions upon millions of years, working it's way up from the tiniest self-replicating chemical reaction.
That doesn't necessarily exclude God completely; the hand of God may have poked evolution in particular directions along the way. But here's where I'm going to invoke Occam's Razor: if all the evidence I've cited above is sufficient to explain our development, why add this God guy to it? He seems like an extraneous piece, put in there just to make us comfortable; he adds nothing to the explanation.
And, if God "designed" things, why reuse parts that don't make sense? I can understand sharing the notochord that a bunch of other mammals start from, similar metabolisms, same major organs, similar muscle and bone layouts, same bone materials, and a lot of re-used chemistries and parts. Why leave in crap we don't use all that much, like arrector pili muscles, the appendix (although it does have a small function in immunity), the instructions that give the occasional person a vestigal tail? And it's not just us: why does a whale's flipper (which is nothing at all like my arm) have the same bones as my arm? If you were designing creatures, wouldn't you use a different structure for an appendage with such different function? This is a contradiction to "intelligent design".
Thus, I accept that evolution happened, and natural selection is a firm scientific explanation for how.