Evolution for world builders

Look at you, flying over a marine desert. What might your children become?

Evolution. It’s one of those incredibly simple concepts that has such all-encompassing, mind-bending potential. That makes it a very useful tool for a world builder to add to their arsenal, and also an incredibly deep pitfall down which you can fall.

Firstly: Let’s have a quick chat about what evolution can do. An often asked kind of question is ‘Can my X evolve?’. The simple answer to that is always ‘Yes, if it’s got exactly the right stimuli’. Remember that whatever creature you want to put in your world, you’re allowed to. It is, after all, your world. As long as the creature abides by the laws of your world (I’m partial to using the same laws as ours, personally) then it is possible (not necessarily likely) that such a creature can rise out of the primordial soup because of evolution.

Another thing evolution can do is work on things that aren’t even strictly alive. Essentially any set of things that can pass ‘traits’ from one generation on to the next is capable of evolution, though it might get a bit abstract. Here are some examples: Viruses. Parameters for computer programs. Cars. Architecture. Shoes. If that sounds a little odd, don’t worry, in this article I’m exclusively going to use living organisms as examples, but bear in mind that when I say ‘creature’ I really mean ‘Thing that is in some way capable of passing its successful or unsuccessful traits on to a new generation of things’, and when I say ‘breed’ I mean ‘pass on successful or unsuccessful traits in such a way that the next generation can inherit them’. Phew.

Now. Let’s discuss what evolution doesn’t do.

1: Evolution doesn’t do individuals.

I take a Yellow Tailed Lightning Rat and put it through a series of grueling and cruel challenges. After it’s completed enough of them it evolves, accompanies by wailing guitar music, into a Thun-Degu.

No. No no no no.

That is not evolution. It would better be described as metamorphosis: ‘A change of the form or nature of a thing into a completely different one’. Anyone who describes such a process as evolution is fundamentally missing the point: In evolution individual creatures do not change, whole populations of creatures change, and they do so by dying a little. I’ll come back to that cheery thought in a minute.

Similarly creatures that change to better suit their surroundings are not evolving. If a superhero in the well known Y-Blokes (Let’s call them Wallace) has the power to grow gills underwater and develop heat resistant skin when they’re engulfed in flames the word you’re looking for is adaptation, not evolution. Evolution is the mechanism by which an entire population can adapt, but an individual adapting is not the same as evolution. Linking an individual that adapts to the process of evolution is an incorrect use of the term.

What does this mean for worldbuilders? Don’t use the word evolution if you’re referring to a single individual. More to the point, don’t have characters or organisations that should know better misuse the word. If a geneticist claims that your dinosaur-newt hybrids have ‘evolved’ the ability to breed then they really shouldn’t have been trusted with the DNA splicing kit.

2: Evolution doesn’t care about time.

Evolution takes millions and millions of years!


This is a common one, and the reason it’s so common is because while evolution doesn’t directly care about time it does care about other things that may rely on time. Buckle up. This is about to get conceptual.

Every creature has a lifespan. That lifespan can be roughly split into three parts: Can’t breed yet, I can breed, and past it. Evolution cares about the ‘I can breed’ part of a creature’s lifespan, because that roughly indicates the next generation of creatures can be made. In most cases there is a length of time between the ‘I can breed’ portions of the lives of parent and a child, which in turn means that evolution takes time. In the case of complex creatures where there are decades between one generation breeding and the next it can take a very, very long time. In the case of bacteria it’s much shorter, but evolution doesn’t really care. The thing to watch out for here is making sure that you don’t think of evolution in terms of time but instead in terms of generations. You can convert it to time later.

Of course, there is the possibility that the last generations can overlap with newer ones. This introduces another layer of complexity as it means all the generations blur together. This means that potentially fast breeding does not necessarily mean fast evolving, as an older generation may be capable of outcompeting the younger ones for a while. This will eventually settle down into a general pattern of generations though. The older generation has to die eventually. Like I said: cheery.

What does this mean for worldbuilders? Take into account the lifespans of your creatures to get a feel for how long it might take them to develop new features. You’ll need to think about how long it might take for one generation to stop being the dominant breeding population and the next to take over, but you should be able to get an idea. For most stories this isn’t an issue: The narrative happens in less than one creature’s lifespan and thus evolution isn’t a concern. If it is a concern then you need to watch out for how many generations are happening and remember that it takes many, many generations for changes to happen.

3: Evolution doesn’t do big changes

The Dean of Y university was born with the ability to read minds. None of his ancestors showed even the slightest hint of this. Some call it evolution taking a big leap forwards.

And while they might be right, evolution doesn’t do that kind of thing.

The reason is pretty simple: It’s the same reason you don’t expect to roll ten thousand dice and have them all come up 6. No, actually. It’s worse than that, because some dice can’t roll six if another set of dice haven’t rolled exactly six, other dice won’t roll six unless both of their neighbours are sixes and some dice require the moon to be full or they only roll sevens.

Evolution isn’t some magic process that occasionally drops horns onto people’s heads or bestows bipedalism; it’s the slow, gradual combination of tiny, tiny changes that lead to more pronounced and astounding changes over the course of many generations. We rely on the fossil record for our knowledge of long-dead species, and the fossil record is horrendously incomplete. That means that what we see are two distinct species, one with hips oriented like a bird, one with hips oriented like a lizard, but we miss all the tiny, tiny incremental changes in-between lizard-hipped and bird-hipped.

Since evolution works in tiny changes you won’t see much deviation once a good pattern has been established. If something works well enough then there will be lots and lots of tiny variations on the theme, but no one thing that powers forward and becomes a new feature. Why are so many creatures on earth four legged? It works; and because it would be a big change to somehow add on an extra pair of limbs (that don’t cause problems) we don’t see a preponderance of six legged dogs, just lots of variations on the four legged theme.

It also means that we get branches. An often-asked question by evolution skeptics is ‘If humans are better suited to survival, why are there still monkeys?’. A: Give it a few more decades, we’re doing pretty well at killing everything else on the planet, B: There’s no need for the monkeys to change. For the vast majority of monkeys the tiny changes haven’t added up to a new feature, and that means they don’t get any advantages, but also they haven’t felt any disadvantages. They still work, so why would they go away?

What does this mean for worldbuilders? Don’t say things like ‘suddenly they evolved’. Suddenly in evolutionary terms is still measured in tens of generations for very simple changes. Don’t have a new species just ‘pop’ out of nowhere. The first chicken egg came from a thing that was almost but not exactly like a chicken. Don’t just forget about an evolutionary lineage either. If you’ve worked out the evolutionary path for your 6 legged flying cat-bear remember that other things will have evolved out of the same lineage, many of which will still be around. Also remember that for the 6 legged flying cat-bear to have been successful, its lineage will have been a relatively successful one, so there will be a lot of six-legged things around. Whatever you do: Don’t have an entire biosphere of six legged creatures and one species of four limbed hominids.

4: Evolution doesn’t do predictable

A scientist stands in the desert and proudly proclaims ‘It’s hot here, of course all the lizards would evolve parasol tails!’.

Gah. No. There is no ‘of course’ when it comes to evolution. There is only ‘well, that seems to have worked’.

There are some truly wonderful fossils from back in the dim and distant past. Some have eye-analogues on their backs. Some have body plans not seen in any other species on the planet. Some have half internal and half external skeletons. These creatures aren’t primitive attempts at getting things ‘right’, because there is no concept of ‘right’, nor are they just failed attempts at life, because some are complex enough that they must have been plentifully successful in their own right. So why aren’t they around now?

Chance. Pure, bloody minded chance.

Similarly: There are creatures around today who are surviving, nay, thriving in situations that evolution did not and could not have prepared them for (though rapid breeding cycles mean that evolution will be a factor for some species, like pigeons or rats). Again: This is just chance.

There is no plan here. There is no greater power presiding over the proceedings and saying ‘ah, it makes sense to evolve X feature, so X feature will certainly evolve’. If something evolves Y feature first and it works well enough then X feature may never get a look in. Parasol tails are outcompeted by dancing on hot sand, not because dancing is necessarily better, but because pure chance means dancing happened first, and that was good enough to stop parasol tails from taking off. Dancing ‘seems to have worked’, not ‘dancing was inevitable’. Even given exactly the right stimuli any given feature is not guaranteed to evolve.

This is a mistake commonly made by worldbuilders, but it’s an important concept to wrap your head around. Just because something has happened does not mean that it must have happened. If we rewound even a hundred thousand years, kicked a monkey in the gonads and ran off then it’s entirely possible that the person writing this would be writing using tentacles. There’s nothing to say that monkeys must evolve sentience in the right circumstances, just that it becomes more likely because sentience might help out.

“But wait!” you may object “What about convergent evolution?”. You’re right. Some things are so fundamentally useful that they have evolved separately all over the world. Cuttlefish eyes are eyes, but they didn’t come from the same place as our eyes. The number of different types of lung in the world is frankly astonishing, as are the number of cardiovascular solutions. The thing is that for every plan that was successful there are a thousand other ways of doing things that could have worked, as evidenced by some creatures losing the use of their previously evolved eyes because it’s not worth the energy, or some really very complicated arthropods that hunt by vibration and can only really see their prey when they’re already on top of it. The evolution of the eye is likely because it helps keep things alive, that’s why it happens so often, but it’s not an inevitability.

What does this mean for worldbuilders? Oddly: This is actually a bit of a boon. Because anything can happen we can describe pretty much any creature that’s still within the laws of physics, as long as we’re willing to sit down and work out exactly what convoluted series of bizarre steps led to its creation. Its also a bit of a pain because for every step we work out we also have to work out why a million other things that might have been more suited for the situation at hand didn’t happen. Sometimes this is easily explained as pure chance (parasols or dancing?); other times it’s a bit more tricky (why didn’t eyes work out?). The other thing to consider is exactly why you want to work out this set of complicated steps. If it’s purely so you can describe to a reader each blow-by-blow environmental change that led to your creature then you should stop worrying quite so much. If its so you can populate your world with other creatures that might have piggybacked on the same evolutionary plan then you don’t need to worry so much about the fine detail of the whys and wherefores. The latter is very worthwhile doing, and you might get some results (even entire new creatures or storylines) that you didn’t expect.

5: Evolution doesn’t stop

The creatures lurking in the bushes have evolved into the peak predator. They don’t need to evolve any more.


The creatures lurking in the bushes are still evolving. Even if the environment hasn’t changed; even if the basic body plan of the creature has remained the same for millennia: Those creatures are still evolving. There is still competition. Some creatures still live or die or get to breed over and above the others of their species. Some traits are handy, others not.

Now, I’ll grant that the rate of evolution slows right down when creatures get ‘comfortable’. As we said about small changes: Monkeys are still here, and they’re still basically the same as they were when humanity was a glimmer in an australapithicene eye, but they never stopped evolving. There will still be tiny changes that make the monkey of today subtly different than the monkeys of yesteryear.

You’ll note that I said ‘different’, not ‘better’ there.

The reason I didn’t say better is because the concept doesn’t exist in evolution. There is no such thing as ‘better’, because ‘better’ implies that theres a scale of goodness that remains the same from generation to generation. There isn’t. The monkeys of today may well lose out to the monkeys of yesteryear if they were just slapped together willy-nilly in a jungle. The monkeys of today are only better at surviving when compared to the monkeys of today and their parents, and the goalposts are constantly shifting as one generation subtly differs from the last. And that’s before we consider environmental changes!

So evolution doesn’t stop.

“But wait!” You may cry “What about humanity? We have medicine and technology and free society! Surely we have bypassed evolution?”

Nope. Not even a little bit. There are still things that will increase or decrease your chances of breeding, even if you don’t know it. Social norms, concepts of beauty or morality, the ways that we communicate, even the ability to allow older humans to breed via IVF. All of these things still fall under the envelope of evolution, along with a myriad of other subtle things that I’m probably not even aware of. These things may change rapidly, even within a generation, and the basic plan of the human being almost certainly won’t change enough that our distant ancestors will look back at the fossil record and pick out pre-and-post Kardashian human beings, but we are still changing, still evolving, still adapting to our own adaptations.

What does this mean for worldbuilders? If you’re working over long-time scales or far into the future, try and keep in mind that no species will remain perfectly the same, even if it’s so finely optimised that it doesn’t seem to change at all. Also bear in mind that some species (ants, for example) can hang around for a long period not because they’re very well optimised but because they remain competitive in pretty much every situation against pretty much every other creature. They’ve evolved to be just good enough to persist no matter what the environment or competition is, and even they are still evolving.

One final thought: Evolution doesn’t do simple

Someone on the internet writes a five point outline of what evolution does and doesn’t do and it’s entirely perfect.

Hah! Hahah! Nah.

Everything I’ve said in this article is quite simple, but evolution isn’t. The most well known attempt at describing evolution isn’t an elegant formula or a two-page thesis, it’s a tome. Since then our understanding has grown, expanded, been rewritten, expanded some more and then gotten crazy. Pretty much every creature alive today is a mind-bogglingly complex bag of chemicals put together and optimised by trillions of random changes, and every description of evolution is either woefully inadequate or similarly mind-boggling.

What does this mean for worldbuilders? Even if you’re already an evolutionary biologist, don’t try to use too much detail. Evolution is a broad, powerful brush that you can use to paint your world, but too much detail will turn your world from a canvas for storytelling into a treatise on hypothetical evolutionary paths. If something feels off about how you describe a creature evolving, don’t include it unless it’s essential for the story. Even then, it’s often better to describe creatures as they are than as they were ten thousand years ago (Does it really matter where in its genealogy your monster picked up its vulnerability to carrot juice?). Go, read a few decent articles on evolutionary theory, work out the rough evolutionary path of your sky-whale, maybe put in some offshoot species.

Then forget all about the evolutionary work you did and focus on the fact that your world has sky-whales.