I'm going to talk about evolution.

Before I start, two things: First, I'm not out to convince anyone who might disbelieve in it. The Talk.origins archive, and science blogs Pharyngula and The Panda's Thumb, are full of better (and more knowledgeable) writers who have the information for arguments. Second, please be advised that I'm not interested in receiving comments from well-meaning people who want to enlighten me. I'd be stunned if you could come up with an argument I hadn't heard, and already rejected, long ago. So save the outraged email.

That said, I don't hide where my personal beliefs are on the matter. But as I said, this isn't about convincing. What it is about is my firm belief that if you're writing about something, whether you personally believe in it or not, you have a responsibility to know enough about it to get it right or, if you're changing something, to be aware that you are changing it. I'd make the same argument about politics, economics, religion, or current belief in alien abductions.

If you're writing about something involving evolution, somewhat common in the transformation genre, there are some basics you need to know. The first is, obviously, what evolution is:

Evolution is a process that results in heritable changes in a population spread over many generations.

In other words, an organism can be different from an ancestor. Those changes typically occur because an organism that has a problem dealing with its environment will tend to have fewer surviving descendants than a similar organism that has less of a problem.

And that's about it. No, really -- that's it.

You usually hear 'survival of the fittest', but that's not true. The correct formulation is 'survival of the barely adequate'. There's an old joke that makes it easy to understand:

Two men are walking through the woods when they are confronted by an obviously hungry bear. As the bear eyes them, the men argue about what they should do. One says they should climb a tree. The other says they should run.

"What!? You know you can't outrun a bear!"

"I don't have to outrun the bear. I just have to outrun you."

If one man outruns the other, it doesn't matter if he barely goes faster or if he takes off like a juiced 100-meter sprinter on a caffeine jag. He'll survive in either case. Nature uses a pass/fail basis for deciding who wins the evolutionary arms race, and you don't get higher marks for passing the test with a bigger margin than some of the others who just squeaked by.

With the basics out of the way, let's get into parts that are important for writers. The easiest way is to give the rules and then an explanation.

Evolution does not have a goal. Contrary to popular belief, it's not a struggle to the top, with the 'more evolved' superior to the 'less evolved'. We like to think of it that way because we're the center of our own little world and the end result of the particular evolutionary process that resulted in us... but in evolutionary terms, we just aren't that special. The history of life on Earth was not a process designed to culminate in us. It isn't a process meant to culminate in anything. Evolution does not plan ahead (or plan at all), nor does it have a memory. So if you, the writer, are planning on having a character 'evolve' into some superhuman thing, or 'devolve' into a fish or whatever, please be advised that the way it's usually shown is crap.

Let's look at 'evolving up'. Well, into what? You cannot predict what humans will look like in X number of years, any more than you can predict the exact number of centimeters of rain that will fall on a particular spot ten years from today. But just as you can be confident that that rainfall will be in the normal range for that spot at that time of year, so you can make some educated guesses about the future state of humankind. Ignoring technology, it's reasonable to assume that the genes that prevent wisdom teeth from forming (genes which my dentist and his bank account are happy I do not have) will spread in the population, since we really have too many teeth for the size of our mouths. Reduction in the number of toes is possible, since they're well on their way. Other than that, though, I wouldn't venture to guess. The reason is because I can't predict the two things that count: (1) Changes in the environment, and (2) Changes (due to combinations of genes or the rare mutations) that hang around and get an opportunity to express themselves.

Simple analogy again. Imagine a group of sexist men get together and decide they want to create, through natural (in this case, sexual) selection, a population of 'perfect women'. They'll create an artificial environment where women with the desired characteristics are allowed to breed (with men of the appropriate genetics themselves) while those who have mostly negative characteristics are not. This is, pretty much, feasible. It's no different than what breeders do with roses, dogs, cattle or horses.

So the population that results would be 'more evolved', from the point of view of the idiots in charge of this exercise in eugenics. The problem is that there are multiple outcomes depending on the environmental conditions (that is, the restrictions in this case that are artificial). Do they want lower intelligence, big breasted blondes? Tall, athletic and intelligent redheads? Medium intelligence but submissive brunettes? Any are possible. So, without knowing what environment those men will create, it's impossible to postulate what evolution (in this case artificially directed) will produce. And that's without throwing in the possible random variables.

Some people have streaky hair (that is, some follicles have different colours than the primary). Suppose one women has daughters who have this feature, and the director of the project finds this desirable and adds this to the project goals. This random factor is even less predictable. Even the originators of the project didn't consider this a desired feature at the beginning.

What in this case is done by artificial selection is, essentially, merely an accelerated version of what could occur in nature. So, from our point of view right now, before this rather abominable project started, we can't predict what will evolve due to the project. And if you can't make this kind of prediction for artificial selection in a mostly controlled scenario, how can you predict what will eventually result from mostly uncontrolled natural selection?

If this is true now, it was also true in the past. You could not take a Homo erectus and put it into your magical evolution machine (which 'accelerates natural evolution', or some such nonsense) and have a modern human walk out -- not unless you duplicated every single environmental effect, genetic combination and random mutation that took place in nature. Well, I suppose with sufficient historical knowledge and supertech you theoretically could, but the only way is to have all the necessary knowledge. And quite honestly, if you could do something like that, it's probably just easier to change the erectus into a modern human directly.

Thus, the way SCABS is supposed to work in the Blind Pig stories (at least when it comes to animal forms) can't work the way it's described as basically 'devolving' an organism's genome into a common ancestor, and then 'progressing forward' into another form. The first part isn't completely bogus, because each organism's DNA does carry traces of its ancestry going back an arbitrarily large number of generations. But in order to do the second part, the genome would have to know what changes it would have to undergo in order to change into the new form. Which, as just shown, it can't. You can't take my DNA, go back to the common ancestor of me and the canines, and then go forward to a wolf. The DNA of that common ancestor didn't know it was going to lead to humans and wolves and lions and tigers and bears, oh my. There's nothing in the DNA to tell it what changes it has to make to itself in order to do it.

The other common mistake I see in writers is about 'devolution' or 'evolution going backward' or some such. It doesn't work, for the same reasons going forward doesn't. Things have changed, often so much there are limits to what can be pulled out. Yes, we share proteins and DNA sequences with other creatures but the genetic sequence that causes the liver to form or hemoglobin is only a part of what makes us us. Consider a house. A sequence of houses can have the same supplier provide the windows and doors and paint, can use the same framing wood and type of wire. In fact, two houses could come from piles of material that are 99% identical but look clearly, even radically, different. It's the overall plan that's important.

Anyway, as I was saying, for the same reason you can't predict the future, you can't precisely retrace the past. Those random combinations of genes or chance mutations which are the raw material of evolution sometimes do things that can't be undone. Again an architectural example. A stone arch is a freestanding structure, but the only way to build it is with a scaffolding to hold the stones up until it's complete. Remove a stone... any stone... and the entire thing collapses. Some biological processes can do the same thing, the 'scaffolding' having been eliminated from the genome through evolution, leaving something that can't be changed by removing a component lest it fail. So you can't go backward because the capability required in order to reach the primitive form isn't there for you to use.

Now, that isn't to say you can't have something that looks like the primitive form. Everyone should be familiar with convergent evolution (organisms occupying the same niche in similar environments will tend to look alike), so it's possible to get something that looks like something else without being exactly the same. But there's a limit. You have to work with what you have.

Marine animal tails provide an excellent example. Cetaceans have horizontal flukes, while fish and the extinct marine reptiles like ichthyosaurs had vertical tails. Why? Fish backbones are constructed to allow side-to-side motion. Reptiles inherited that, so when reptiles went back to the sea, the ones that used their tails to propel themselves developed flukes using the side-to-side motion. You can still see this in crocodiles, alligators and snakes in the water. Mammals, on the other hand, had since their separation from the early reptiles evolved backbones which flexed up and down instead of the side-to-side of their ancestors. As a result, when the whales returned to the sea, they adapted an up-and-down motion for propulsion, thus horizontal flukes.

Evolution used the materials available.

At this point, I'll toss out some useful references. They aren't about evolution per se, not entirely, but they show the results of it, some very speculative. And they show you the variety of things that are possible using only natural evolutionary processes.

Dougal Dixon has several books, After Man: A Zoology of the Future being the best. It's about what life on Earth looks like 50 million years from now, assuming humans go extinct in the near future and there's no replacement sophont. Birds that take the place of whales, rats that have evolved into top predators (including some that fill the seal and walrus niches), rabbits evolving into the equivalents of deer, antelope and giraffes, bats that have become flightless predators running on their strong arms, using their legs for grasping and using sonar to see and hunt. It's a great way of seeing what might happen in a completely scientifically plausible way.

The Future is Wild, the BBC/Discovery Channel series and companion book does the same sort of thing, looking at the world 5, 100, and 200 million years in the future. Some of us consider it a bit dodgy in the second and third segments -- it seems like the writers and consultants had a grudge against the vertebrates, and especially the mammals -- but it isn't totally ridiculous. Some of the things like the giant strolling forest squid were peculiar, but they're just unlikely, not impossible.

Of course we can't forget the three BBC Walking with... series, Dinosaurs, Beasts and Monsters (respectively dealing with the dinosaurs; the mammals up to the present time; and life on Earth before the dinosaurs). Although it doesn't engage in evolutionary speculation like the previous two references, it does show that life had far more variability than we usually think about and that those "primitive" organisms could be just as mean and nasty as anything today.

Dragons, a Fantasy Made Real, the Discovery Channel special that used as its central conceit the ideas that dragons really existed. It deals with the evolution of the creatures, but I really can't rate it that high. While it has interesting ideas, the way they pass off a change from a four-limbed ancestor to a six-limbed 'modern' dragon just to have a stereotypically European dragon with four legs and a pair of wings just beggars belief. They try and pass it off as a mutation, but, well, there's no known vertebrate where this has happened. It's been four limbs since forever. A chance mutation might create the odd monster (in the biological sense, as in something that deviates significantly from the norm for that species) but none have survived.

The short story The Dragons of Summer Gulch by Robert Reed (published in 2004) touches on something similar but does it better. In that alternate world, dragons used to be the dominant class (no dinosaurs), and humans find the fossils of all sorts of different types, but they didn't evolve from 'our' family of vertebrates. One character mentioned primitive fossils from their version of the Burgess Shale that has two types of chordate worms, one which led to everything else and one which led to the six-limbed dragons. That sort of thing is plausible. Unlikely, but plausible.

What I'm trying to point out with those examples is that you can have all sorts of wildly possible creatures without needing to resort to deities, magic, or magical science. Nature has given us a myriad of possibilities, both in creatures that have really lived and those that exist only in our imaginations of the paths not taken.

I'd finish this off by mentioning one type of transformation story that I've very rarely seen: That of changing an organism into another that it actually has the capacity to change into, that it might have been had the evolutionary history been a little different. This means a creature that doesn't exist but could have. Instead, everyone focuses on animals.

Bah! What about humans with tails? We have the bones to make the basis of it, and a rare child is born with a vestigial tail. That X-Files episode was one of the few that was actually plausible. Fur-covered humans? Easy. Fewer fingers or toes, same. In fact you could imagine all sorts of ways humans could have evolved differently (we'll assume the mind doesn't alter; this isn't realistic, but we're only concerned about physical changes here) while staying within the range exhibited by the primates.

Try it. Might be fun.