Snuggle you have what I believe is an expansive magic system. What do you think makes a proper magic system? Doesn’t matter if its a heavily personal view. I’m willing to listen
There’s a lot of basic advice you can find easily, about limitations, consistency, the usual suspects.
For me, the key to developing a speculative element is to first associate, then extrapolate. Or, to put it metaphorically, I find a sapling of an idea, uproot it, graft it, then replant it in new soil to let it grow until even its branches have branches.
For an example, let’s take the design of the races in Black Nerve:
- In real life, ambrosia beetles are a niche of tree-boring bugs that survive by eating a symbiotic fungi that rots the wood for them. In Black Nerve, the weevils of ambrosia grow a hallucinogenic magic fungus, and by carefully heating or even burning the wood it grows in, the fungi’s metabolism accelerates: reactions are catalyzed and its magical emissions interact with the hallucinogenic enervate in the weevils’ brains. A few million years of evolution, and you have a psychic beetle hivemind centered around giant fungal supercomputers. And that’s half of what’s weird about the weevils.
- In real life, spiders weave webs to catch prey. Lichen exist because algae is willing to share some sugars with a fungus that’s willing to shelter it. In Black Nerve, the therids weave their webs into velaria in which photosynthetic microbes grow. Those microbes secrete a sweet, nutrious liquid the spiders can eat, and they can leave the liquid there to attract further prey.
- In real life, some ants carry leaves back to a fungus-garden and some of them use their silk to weave their nests. The intelligence of eusocial colonies is sometimes theorized not to lie in the individual ants, but in their aggregate interactions. In Black Nerve, the spinner ants use fungus to grow a kind of fabric, and can weave it into complex patterns. An ant naked and alone is dumb, but by encoding and expressing its thoughts in fabric and manipulating it, its capacity for thinking is expanded remarkably.
Those examples were a bit indulgent, but I hope you can see the underlying logic. I take a starting idea (“Ambrosia beetles…”), focus on one or two aspects of it (“…grow fungus…”), then free associate into a related idea (“hallucinogenic fungus?”), then I start extrapolating ever further. (The weevils adapt to the fungus; the fungus adapts to grow more potently hallocinogenic; their culture becomes centered on visions, and so on.) I’m omitting how the enervate stuff works here, but it’s better I dont get into all that.
The key thing here is that while I have a unique combination of ideas, it’s not just slapping random things together — I’m not even slapping unrelated things together! Weevils grow fungus, and some fungi are hallucinogens, so what if they grew hallucinogens? Spiders trap prey, and it’s easier to trap prey with a lure, so what if they had a lure?
I’d hazard no one else has thought of social spiders that grow moss-webs like artificial flytraps, but once you pose the idea, there’s a certain compelling thread of logic tying it all together, isn’t there? It’s weird, sure, but contrast this with like, trying to toss together balloons, tap-dancing and the platonic form of matrix multiplication. You could probably think of a magic system to tie that all together, and it would also be weird, but it’d be missing something I think I have here.
To pull back the curtain some, it may be instructive to note that the genesis of the weevils of ambrosia wasn’t quite as freewheeling as it may sound: an unspoken term in this question is a vague inclination I had that I wanted weevils to have something of an ‘elf’ or ‘fae’ vibe, which guided the specific association I called upon; I didn’t pick hallucinogens completely out of a hat. Even then, there’s still a compelling link between these concepts — both are integrally tied to trees, to forests.
If the exercise is association, you can see why metaphor and analogy would be powerful forces. This is a bit like that, so what if it was a lot like that? This is a bit like that in figurative way, what if it was also like that in a literal way?
The goal, to begin with, is to find a combination of ideas that are novel, yet ramified. Ingenous, creative, strangely cohesive. It doesn’t stop there, though. Next you extrapolate the idea by these same principles. Don’t stop until your branches have branches.
The best illustration of what this means is to ask a seemingly simple question: what actually is the magic system of Black Nerve?
A vesperbane can use spells, that seems like an easy place to start. A vesperbane molds the arete for a spell, and then casts it as prepared. But to understand how molding and preparing a spell works, you need to understand how their grafts and the vespers work. And to understand how those work, you need to understand bat blood and black nerve. Systems built on top of systems, just like how you couldn’t have the internet without computers.
The powerful thing, though, and what makes Black Nerve deep in a literal sense, is that when you back up like this, you can roll down to different endpoints. To understand a spell, you need the graft system, but through grafts, you can get other spells. To understand grafts, you need the fundamentals, but those fundamentals can manifest in other ways. The therids have bloodweaving and the euvepids have inkvenom inscription, and those rely on blood and black nerve respectively — these are in one sense different magic systems, but in another, they are sides of the same coin. A strange coin that has way more than two sides, but I digress.
When it comes to how to extrapolate like this, it might sound really hard and complex, but just keep your eye on the ball. You really only need to ask two questions:
- “They can do that, but how does it really work?”
- “If they can do that, can they also use it to do this?”
Say you have a magic system where people can summon fireballs. Cool, how are they doing that? This doesn’t need to be anything special or rigorous for now, you can practically handwave it, just say they’re manipulating the latent flame energy within themselves. But what else can they do with that latent fire energy? Could they keep themselves warm in the cold? Could they also cast flamethower or fire aura spells in addition to fireball?
If they can manipulate flames with enough precision, could a blacksmith with these powers use the manipulation when forging to enhance their control and the quality of the result? If you have a land full of firebender blacksmiths, is the kingdom especially advanced in metallurgy because of it? Is their culture chiefly focused on the symbology and imagery of metals because of this — do they have a pantheon where every kind of metal gets its own god?
And that’s just what happens when we “build out”. I showcase that first mainly because I think it’s more productive for actually making stories, but in my own work I’m very fond of “digging in”, where I would ask, okay, but what is latent fire energy really? Which specific part of the body does it come from? How does it interact with nonliving matter? And so on.
For quick rules of thumb, I think that for an important, central feature of your setting or magic system, you should be able to ask “why?” twice and ask “so?” twice, and get satisfying answers each time. Which is to say, if magic users can do X, have an answer for “how does that work?”, and once you have that answer, figure out the answer to how that works. Then do the opposite. If the magic users can do X, so what? What effect does this have on society? How is the magic practiced and viewed? Then, once you’ve figured out it’s knock-on effects, so what? What effects do those effects have?
You can go crazy with this but you probably shouldn’t. This stuff goes quadratic, even combinatorial, and two levels will give you a lot of depth to play with. But you do get some bonus points in my book if you go on to “so?” the “why?”s. If the magic users can do X because Y, what other effects does Y have? Are there other schools of magic? Or perhaps monsters that draw from the same source?
Just keep extrapolating.
I could leave it there — this is in fact where the original chatroom ramble ended, but I like my posts to resound rather than peter out, and I have a bit more to say now, two and a half years on.
The mother of good art is good taste. That’s what nurses it, cooing at its cute little face and urging it onward when it still needs to grow. To do creative worldbuilding, you need to be able to recognize what’s lacking in mediocre worldbuilding — but that much is obvious, isn’t it?
More specifically, the key to creative worldbuilding is being studied and discerning. You need to be familiar enough with the medium and with craft, to be able to look at an idea and recognize that it’s been done before, so many times, or that it stinks of being the first or second thing the creator thought of.
I often get compliments for my prose and even my naming sense when coining fantasy terminology. Some of this comes down to specific skillsets But there’s a very specific eye I’ve developed, a way of parsing and probing sentences. When you learn how to write metered poetry, you learn about feet, and you begin to see each line mechanically, as this thing with “slots” that you can fit syllables into — it almost becomes an inventory management puzzle, how to arrange words so they’re still intelligible and beautiful while obeying the rules of meter.
You need three syllables with a stress in the middle, so you start to think of words that fit that pattern, how you can bend and conjugate and rearrange the line around it to fill that spot, satisfy that constraint.
But more instructive is the concept of drabbles (which I’d classify as a borderline poetic form). Formally, these are stories that are exactly 100 words long. I’ve written several[Cheating, I admit; mostly this comes from writing just one story structured as a chain of ten drabbles: “Tines of the Devil’s Fork”], but what you learn when writing them, is how to squeeze ideas, shave away what isn’t interesting. You look for spots where you can pack meaning into single words.
You learn how to write densely.
What this whole tangent is getting at, is the practice of amplifying density at the level of individual words is a microcosm of a general skill, which is identifying the limp, unsurprising, low entropy parts of an idea and fortifying it with a spice of novelty.
This is something that should stand in sharp relief, in the age of the LLMs and token predictors in general. Language models create a statistical model of a dataset — their fundamental reality is compression.
All of which is to say, the key to a good, creative worldbuilding, is to make it dense. “Could GPT-3 write this?” can be your canary. When building up concepts, you have inspirations and starting points, and — as I described above — you turn that into conceptual worldbuilding by associating and extrapolating. But you need to avoid superficial connections and developments.
There’s some amount of depth you can fake by knowing how and where to inject randomness and spontaneity, but the thing is, modelling randomness statistsically isn’t that hard either, if there’s no deeper structure to it.
So how do you get that depth consistently? You know the answer — it’s simple and unglamorous. You have to go out and read, enrich yourself with sources of inspriation and acclimate yourself to what’s in vogue, what’s already been done. You have to study the real world concepts your worldbuilding is drawing on, finding those nuggets and unlikely associations that shine with potential.
We know that great artists steal — but the art isn’t in what you copy, it’s in knowing what to copy.