The notion of building my own world was one that attracted me to writing fantasy in the first place. This wasn’t so much because I liked the idea of playing God but when I decided the time had come to see if I really could make any kind of career from writing, I was an at-home mum, with two small children. I might well have written crime fiction but I knew that would require specific and detailed research, if a plot depended on a train timetable, an underground connection, if there was some scientific detail I needed. With two small boys in tow, I’m lucky if I can find out what the library’s opening hours are; you can forget any idea of actually using the place to look things up. But I’d studied history and literature, I had a house full of books, so I reasoned that if I wrote fantasy, I could basically make it up myself and as long as I kept the internal logic consistent, no one could tell me I’d got it wrong.
It sounds easy when you put it like that, doesn’t it? Well, since then, I’ve learned different. Certainly keeping the internal logic intact is crucial and every writer I know agrees with that but you do have to have a solid basis for that logic to work with, so one way or another, you have to build a world.
Where do you find the bits? History is the obvious answer and I’d say most fantasy is rooted in history, one way or another. That begs the first question, why bother tailoring your own bespoke world when you can get one off the peg? Guy Gavriel Kay has said, about history, ‘everything that can happen, has happened, somewhere, sometime.’ He’s a writer who’s opted to take historical settings and use them directly, the reconquest of Mediaeval Spain, for example. Dennis Jones, an American author, uses Byzantium at about 800AD and another American writer, Martha Wells said she uses 19th Century France. But what all of these stress is that they use specific times and cultures as a setting, as a guideline, as a means of exploring themes and ideas that interest them. These authors are using history but they telescope events and warp reality to suit the needs of the story they wanted to tell. Because the story has got to be the most important thing, and it’s the plot and the characters who are the heart of the story. Kate Elliot uses early medieval Europe as her setting and I pick up plenty of echoes and correspondences with my own knowledge of the period. But the story is her own, the characters are vivid, the pace and the ideas are what really grabbed me. Using history as a background like this is just as much work than building your own world from the ground up, as far as I can see.
There are other issues as well. Using actual history can give you a framework, but that can also be a cage. Dave Duncan, mentions a story he’s written with an early Middle Eastern culture where the only form of writing was cuneiform and hardly anyone could read. That immediately cut out all manner of plot devices, had all kinds of implications for how information was held and passed between people. You can’t have some one picking up a scrap of paper that holds a vital clue, if there’s no paper, and nearly everyone’s illiterate. Obviously, the intelligent writer will use that as a spur to their imagination and work round it but the fact remains that taking on real history means you can be taking on real problems. You have to be wary of characters in 14th century dress coming out with 20th century ideas about the role of women or slavery for example.
This gets even more problematic if you take on an historical story or characters as well as a setting. If someone’s already told that story, people have already read it. How are you going to bring anything new to it? I personally have a big problem with Arthurian sagas, for this very reason. There are only so many ways you can tell that story and as far as I can see, they’ve pretty well all been done. Even if you find a new twist, everyone still knows how it’s going to end. The only exception I’ve found to this in a long while is John Whitbourne’s Royal Changeling which has King Arthur as an Undead revenant which I have to confess was a new one on me!
That’s not to say a writer can’t use historical characters effectively; I’d not come across Dave Duncan before I was invited to be part of EosCon and so I bought one of his books, The Gilded Chain, to get some idea of his work. Then I went to read up on Dave using the Internet and found an interview where he said his king was a pretty well straight lift from Henry VIII. Now once I read that, it was obvious, I could see it at once but it hadn’t been obvious as I’d read the book and I don’t think I was being particularly unobservant. I hadn’t noticed because the king is in the background, he’s integral to the plot in a number of ways but the plot isn’t just a rehash of Henry’s life. Events from Henry’s life are paralleled but within a fantasy context with magic and all that implies, so again, the history is being used as a resource, a starting point, not as a short cut to save the author actually thinking for himself.
The main thing that decided me against using a historical setting, is the fact that if you do, you have to deal with the reader who knows all about that period, so you have to make sure every detail is right, otherwise you can guarantee letters about it but you also have to write something accessible and understandable for someone who knows nothing about the period. That’s an extremely fine line to tread. I’m personally highly wary of any kind of novel set in ancient Rome or Greece because Classics was my degree subject. I’ve read so many where my entire train of thought has been completely derailed by a factual error that probably 99% of readers would pass straight over. On the other hand, I know almost nothing about Polynesian history, so when I picked up The Navigator Kings by Garry Kilworth, I needed a book that wasn’t going to assume knowledge I don’t have, but isn’t burdened by great long chunks of explanation, because again, the story and the characters have got to rise above their setting.
Given taking an established historical milieu for your fantasy world poses as many questions as it answers, it’s not surprising a lot of us opt for the pick and mix approach to history. There’s a huge amount of material out there; if any of you scientists think history is just a lot of all dates, think again. There’s social history, women’s history, military history, history of food, sex, politics, crime, travel, culture, music, medicine, science, population and settlement, town planning, you name it, I’ve probably got a book on it by now. I go to museums, I watch documentaries on the telly and I read National Geographic, newspaper articles and find bits and pieces from all over that I take and adapt to suit my own purposes. I imagine there’s a few of you thinking this sounds a real turn off but certainly as far as I’m concerned, it’s not like school work or even university because there’s no essay question, no deadline and you can go chasing off after any tangent that takes your fancy. None of which is the slightest use if you don’t have a good plot and well-defined characters, of course.
The best way I found of building my world was to start with that plot and those characters and work out from there. I wanted a basically feudal yet fluid society for my heroine Livak to work in, so I’ve used ideas of medieval Germany, hints of medieval Italy and when putting together the reasons for this society’s development, I looked at the relation of empires like feudal Russia with the countries on their borders. When I’m thinking about the practicalities of magic and how wizards function within society, I look at medieval medicine and science and how universities regulated themselves. I look at how cities grow, at historical reasons for population shift and how Empires grow and power structures shift, so I put together a history for this part of my world, to explain how the current situation arose. I read something about the historical reasons for promiscuity on the South Pacific, I find something about Mogul Indian architecture that I like, I tie that together with bits from the early Arabian exploration of sub-Saharan Africa and some aspects of Japanese society and I’ve got a basis for the Aldabreshin culture in the south of my world.
But there are still problem you can come up against. You do your research and discover fascinating facts that no one is ever going to believe, even in fantasy fiction. Guy Gavriel Kay on EosCon says that his The Lions of Al-Rassan is based on historical fact – and that astounds everyone! My own personal favourite is something I came across in a very well researched book by two archaeologists on ancient inventions. There was a Chinese Emperor, Shih Huang Ti, the guy with the terracotta army, who had the gates to his palace made of naturally magnetic ore so that anyone trying to get in with weapons or armour stuck to them. I think this is a brilliant notion but I cannot see how I could use it in a book without readers laughing out loud. Back in my student days of Friday night D&D I might just have got away with it but only if the players were willing not so much to suspend disbelief as lynch it – which is a line you’ll find originally in Tom Holt’s work, by the way.
The next problem comes when you have a wealth of material for your wonderfully detailed world and you start to write the story. You find that you can get almost none of this in to your novel, without sending the plot into suspended hibernation for half a chapter. The temptation here really is something you have to fight. I’ve got a book on medieval travel; it’s got the earliest recorded Highway Code in it, which has some fascinating stuff in it. It tells you how much you can pack on a horse, as opposed to a mule, as opposed to a donkey, as opposed to a camel and what speed each of these will do over what kind of terrain. I haven’t yet discovered the in-flight ratio of the unladen sparrow (African or European) but it’s just a matter of time. You can bore your readers silly with your research; an American writer and academic, Sarah Smith, who writes on crime fiction and SF, described it as the attitude in an author which says ‘I’ve suffered for my art, now it’s your turn.’ You’ve always got to bear in mind that the world has to serve the story.
But you do have to get the detail in. To go back to the crime novel example, you have to be sure you get the detail right if your plot hinges on being able to get a Bakerloo line train at Paddington but you don’t have to explain what the Underground is. You don’t have to explain who the Police are and how they function as part of the Establishment. Using a specific historical setting helps you a bit with things like that but not a great deal and with an entirely constructed world, you have to make sure things like this are explained but, as I said before, without putting the plot on hold for so long your reader mentally hangs up. It’s another fine line, another balance you need to strike.
Victoria Strauss has talked about ‘transparency.’ You need to slide information in with bits and pieces here and there; in The Tales of Einarinn I open my chapters with a page or so of something supposedly quoted from a historical source. If I have a character sit down and tell another ‘this is how it is,’ I aim to make sure it’s integral to the plot, that the character being told has a good reason for not knowing that information and I also try to keep it as short as possible. For one thing, the more detail you give, the more you commit yourself and that limits what you can do in the future. I personally didn’t have any problem with the introduction of trader caravans in The Renegades of Pern, even though they’d not come up before, because the detail we got in the earlier books was always specific to the story being told, not an all-encompassing guide to the entire society.
So, I do my research, I write my story, painting the background in with as light a touch as I can and then I get reviews that say things like, ‘the fantasy world here is a familiar one.’ I read an author new to me, Pamela Belle and discover that her world has northern plains, an important city on a delta and a southern archipelago of dark skinned people. I read Diana Wynne Jones’s Tough Guide to Fantasyland (which should be required reading for all aspiring genre authors by the way) and find I can tick off a worrying number of clichés. Yes, we have a Vestigial Empire, City of Canals, City of Wizards and my characters do frequently eat Stew.
How does this happen, especially when I could bore for England on some aspects of medieval life? To quote Guy Gavriel Kay again, ‘we’re all mining the same seams’, of history, of myth and of tradition. This is why, when I showed a friend the artwork for The Swordsman’s Oath, his first response was, ‘Oh very Arthurian!’ You can imagine how pleased I was with that but as he explained, the more I realised that certain aspects of that story do echo various plot archetypes, including elements from Arthurian cycles. This common resource base is why elves, orcs and dwarves are often so very similar in so many stories. Basically, if you bring them into your world, they’re going come with all kinds of baggage. I personally see elves with tasteful sets of matching cases in Italian leather while orcs probably go for something in tatty nylon tied up with binder twine. Again, this is where the quality of the story and the characters is what makes the difference between a dull novel and a great one. Katherine Kerr uses both Celtic history and myth in her books but manages to make it both fresh and intriguing. It’s not what you do it’s the way that you do it that’s so important.
You have to be aware of the associations people are going to make but as with so things when you’re world building, it’s something that you can make work for you as well. The best example I can give is the Runes of Einarinn. Livak is a gambler but I didn’t want her to use cards or dice; I wanted something different to point out unobtrusively that this is a different world to our own. I decided to use rune bones and beyond a bit of description, making them three sided, I knew I could leave it at that because people will have a general understanding of what a rune bone means. But I didn’t want to use any of the existing systems, Tolkien’s runes, the Futhark, anything like that because those come preloaded with all kinds of assumptions for the reader which I don’t want to affect their view of my world. So I devised my own system and iconography and here I was able to use the common cultural and historical associations we all share, whether or not we realise it. Ask most people what the Wolf means, as a symbol and they’ll say hunger, cunning, sometimes courage, sometimes evil. We all have those associations, from hearing Little Red Riding Hood as kids or thinking about Oz in Buffy the Vampire Slayer. The Drum is a good one; even people who’ve never read a history book in their life will have vague memories of old westerns with the Indians’ tom-toms echoing in the hills or some old movie where one colonial type says to another, ‘It’s too damn quiet Carruthers,’ and the jungle drums start up on cue. Livak gets a full rune reading done for her in The Gambler’s Fortune and I don’t have to explain the reasons for each interpretation. All these kind of associations are working for me and that’s probably the only reason I could do it, without running into pages and pages and getting utterly bogged down in detail.
But the fact remains that writers are all working from the same basic information and readers share the same cultural associations. So why bother doing all that research to build your own world, given getting in things you find out can be so problematic anyway? Why not just put together a standard fantasy world from a bit of Tolkien, a bit of Dungeons and Dragons, a few old stories you heard as a kid and go from there? Why bother reinventing the wheel?
Because you’ll end up with something like a film set. Remember Blazing Saddles, where the honest homesteaders build a decoy town to lure the evil cowboys away from the real one? They have a main street with the hotel, the saloon, the goods store and the jailhouse but they’re just painted flats held up with props. A fantasy world that’s essentially nailed together out of clichés is going to end up a bit like that, one dimensional, superficially convincing but it’s going to sound very hollow when anyone actually tests it. You might get away with it if your story is moving fast enough but I personally wouldn’t want to risk it as a writer and I find it less satisfying as a reader.
If you’ve done your research, be it on a specific period or more generally, if someone comes up and says ‘why does every fantasy novel have a city of canals?’ you can explain the comparative ease of river transport compared to pre-industrial roads, the relative costs, the relative speed and labour until their eyes glaze over and they go away. Fantasy novels generally rely on having A Baddie who is trying to take over the world but if you’ve done your research and built your world up in three dimensions, he’ll have entirely logical and justifiable reasons for what he’s doing, rather than just, ‘because he’s evil!’ You won’t escape the reviews that say ‘this is the usual fantasy world’ but they will be outnumbered by the ones saying ‘the marvellously portrayed world,’ or ‘ a beautifully drawn world with a rich history.’ When you read the Tough Guide to Fantasyland, for every cliché you’ve got, you find three more that you avoided, not from conscious intent but because you did your homework.
We’re all mining the same seams but the more you dig, the more little gems, the more nuggets you find. These are the ones that are so much more effective when you use them with that light touch for detail that is so vital. Anne McCaffrey uses her study of Slavonic languages in the Pern books, using very few Latin-based names, to give a sense of difference about the world and the culture. My characters speak informally for fantasy fiction, they use a lot of slang and I read all kinds of books about this because being able to say ‘it’s raining darning needles out there,’ rather than ‘it’s raining cats and dogs’ adds so much to the atmosphere, the vividness and that indefinable ‘otherworldliness’ that you want in a fantasy novel.
If you do this kind of background research for one book, you’ve probably got half your next one written as well. You’ll come across things that spark plot ideas, you’ll find you’ve put a scenario together that just cries out for a story to explore it. I’ve got an on-going civil war going in part of my world and I’m already getting readers asking for more on that, how’s it going to be resolved, if at all. I developed the Aldabreshi culture for The Swordsman’s Oath and all kinds of ideas are jumping up and down and waving down south. It’s a culture that’s totally opposed to magic but in this world, magical talent is innate; so what happens when someone mageborn turns up in the Archipelago? I’ve got five books plotted in the Einarinn sequence so far and my problem after that isn’t what to do, it’s what idea to follow up first.
So I’ve discovered there’s an awful lot more to building a world than just keeping a card index handy and up to date. It’s not easy but then if it was, I personally wouldn’t find it nearly so much fun.
The earliest version of this essay I can find is the text for a talk written in early 2000, presumably for a visit to a university SF society. I don’t have any note to indicate printed publication elsewhere.