Note: The advice I’m offering here comes from the perspective of:
a) having read a lot of fantasy short story anthologies.
b) having read a lot of amateur fantasy on the Internet.
c) having helped to edit fantasy short stories for one fantasy e-zine.
There are very likely markets out there to which my advice doesn’t apply. I can only tell you what I’ve seen work and what hasn’t (and, of course, what I happen to like).
1) Use your story to do something other than introduce your created world. I get intensely bored when I realize a “story” has no plot; it’s really just a description of the world’s history, geography, politics, races, etc. There are appropriate places for that: at the beginning of your own short story anthology, in a game module or game handbook such as those sold with D&D, on a website or LJ describing the series. But trying to drum up interest in your world by sending someone else a purely descriptive piece usually doesn’t work.
For one thing, there is a difference between a “story,” which is usually understood to have a plot and characters, and a “vignette,” a scene-setting piece in which no action need occur. Scan the market you’re looking at before you decide to send in a story only describing your world. If they specifically state that they do not publish vignettes, they’re probably not going to get overwhelmed by your description of your elves and decide to publish it. They may sometimes accept vignettes under special circumstances, such as a contest, but then the terms are usually clearly laid out—and even then, they probably won’t accept every random world-building piece that comes their way.
For another, it’s really, really hard to get interested in just the world without characters to populate it and a story to take place there. I find myself usually skimming the prologues to fantasy novels where the author is more interested in telling me a history than telling me a story. Let your world grow popular, and an introduction might be welcomed or even necessary. But it’s not a good choice to assemble your worldbuilding notes and send them off.
2) Begin the story. A short story is far shorter than a novel in most cases; one definition used is between 1000 and 7500 words. That means that if too much of the story is taken up with the top-heavy description that’s a problem in fantasy, even if the piece as a whole has good plot and dialogue and action and characters, then the reader feels the blow substantially more than he does in a novel of several hundred pages. Three pages of sheer exposition to begin a three-hundred-page fantasy book, draggy as those can be, are only a hundredth of its length. A ten-page story with one page of sheer exposition has just sacrificed ten percent of itself to explanation/backstory/telling. It’s not a good idea.
How to tell if it’s too much? Ask other people to read through the story. Ask them (and ask them to be honest) where their attention wanders. Scan your writing for phrases or words that you use too often, such as constant use of “glided” for motion. Look for clichés, like “black as a starless night” or “abused home.” I’m of the opinion that a lot of authors churn out description or exposition consisting of clichés because they’re not actually that good at writing description or exposition; they just think it has to be there. Description and exposition that thrills the audience is a grand thing. But first, it has to be thrilling.
Remember, in a short story, a lot of the considerations of a fantasy novel are reversed. Economy becomes important. Economy doesn’t always depend on sheer word-count, but it can, and it’s amazing what you can chop down once you start forcing yourself to recognize that, no, you don’t need the lengthy paragraph of purple prose describing the forest.
3) Strike for a consistent mood. This doesn’t always hold true, since I’ve read stories that could slip me successfully from laughter to tears—but I’ve read many more that tried to slip me from laughter to tears, and didn’t succeed. The main problem with the author building up to a grand emotional transition is that the transition itself becomes the focus of her thoughts, not maintaining the different tones on either side of the transition. (This is a lot like the part I mentioned in the last rant, where an author can drop a revelation out of the pure blue sky because she’s been looking forward too much to the introduction of the revelation and not paid enough attention to all the preliminary steps). Your audience has to be able to believe that the story’s mood is changing, and something so inchoate that they can’t identify the dominant emotion in the first place is not going to be very effective.
One alternative to maintaining a completely dominant emotional tone is to start out with just a hint of what you want the story to suggest, then build on it. For example, a vaguely creepy story that gets steadily more creepy can be great. Sredni Vashtar bothers the fuck out of me, not because it’s the scariest story I’ve ever read, but because it builds up to its creepy climax and after that maintains more creepiness, especially because the story ends with such an ordinary action and somewhat detaches itself from the main character after the climax. The strength of a story may depend a great deal on safeguarding its emotional tone’s strength, always adding more, but not lapsing into a different mood because the author gets distracted.
But another point to be wary of, especially if you write creepy stories, is…
4) Understatement can be just as effective as overstatement. That point about economy I’ve mentioned above? Too many authors ignore it, to write 12,000-word stories that could easily be 8000 or less, because they want to linger on gory scenes/battle scenes/dream scenes/other dramatic or “frightening” scenes. (A lot of these stories don’t really scare me, not in the way that “Sredni Vashtar” or homegrown dark fantasy do). They forget that they have other goals to achieve, and that they don’t have the context of a novel to buffer those dramatic scenes. Let the scenes get out of hand, and they can easily take over the story.
So tone down the scenes a little. Have a character who is ordinarily boisterous get quiet and very tense because of something that someone said, or because they’re walking through a particular place that evokes bad memories in her. Have a philosophical conflict between two characters get argued in some other way than a boring debate that goes on for pages. Have someone possess one disturbing image from a recurring nightmare, not the whole set. Have a death occur off-stage and the consequences reverberate. (This is one that Lovecraft used a lot, since his narrators often only got to see the tattered, gibbering remains of his “heroes” and guess at what really went on). Have explanations that seem obvious, but set up discordant echoes in readers’ minds with a bit of poking and prodding. (This was one for Poe, who cheerfully uses narrators the reader only slowly comes to suspect aren’t sane, as in “The Cask of Amontillado”).
That doesn’t mean you won’t need a dramatic battle-scene at some point in the story. It does mean that the scene should fit comfortably in the context of the story. Don’t suddenly turn to lavish description of the gory wounds if your style everywhere else is laconic. Your audience will notice, and if they have reason to think that you’re more interested in death and blood than ordinary character interaction, they might be faintly sickened.
5) Try a series of hooks instead of just one. You’ve probably heard that you need a “hook” in a short story to get the reader interested—a really great opening sentence, an intriguing situation that they can’t wait to see resolved, a character who carries herself well on the page. Problem is, too many authors use just one hook and think they’ve caught the reader like a fish, that she can’t swim away now. They have their interesting first sentence, and immediately following it is a page-long description of a fantasy inn that sounds exactly like every other fantasy inn, or they set up an intriguing situation and then have it solved by a dues ex machina, or they introduce a quirky character who vanishes once she’s served to introduce us to the “real hero” of the story.
Use hooks instead. Your reader might love the first page, like the second, and hate the third, after all. And you don’t have the luxury of novelists, who often tempt readers into going further by loading their books with chapters. How many times, as a reader, have you said, “Well, I’ll give this book until the end of this chapter to get good,” and then ended up reading all night? Short stories don’t, because they can’t. They’re supposed to move more quickly. Play that to your advantage, instead of against yourself. (And if you’re thinking right now of adding chapter headings instead of scene breaks to your 4000-word story, I would beg you for the love of gods above NOT to. It will be taken as a marker of pretentiousness, and/or thinking that your little story is really a novel).
How do you use hooks? Here’s some advice:
a) Always, always make the hook you begin with relevant to the story, not just a gimmick to get the reader to start.
b) At any point when the pace the story began with seems to be slowing down, as opposed to speeding up, throw in a new and intriguing bit. It might be necessary to show a grim, dramatic battle scene and then have the character reflect on it while watching outside her campfire, but introspective scenes don’t have to be boring just because they’re slower. Put some sounds in the darkness to remind the reader that she’s on watch, and perhaps the enemy will come back. If someone died in the battle, have her recall memories of things she did with them, rather than just conversations they had. If she’s had a reaction of extreme shock or anger to the battle, record the observations of other characters; perhaps they eye her askance or try to take her aside and have “talks” with her. And try not to let the introspection scene repeat itself, bringing up insights that the character already had two pages ago and presenting them as if they’re new. It’s remarkably easy to write conversation and introspection scenes that go in circles.
c) Don’t be self-indulgent with hooks. If a particular phrase that you’re fond of having the heroine repeat turns from an adorable quirk into an annoying dialogue mannerism, and that opinion holds true across the board of your readers, then tone it down or get rid of it.
d) Question everything. Does that scene you originally planned have to be in there? Why? What does it add to the story? Does it actually connect to the scenes that come before and after? Does it represent a needed lightening of tone for a grim story, or a sudden and silly break into toilet humor? Could it be taken out and something more interesting put in its place instead?
6) Short stories generally don’t do labyrinthine. They might if they’re mystery fantasy short stories dedicated solely to the solving of the mystery, or if the whole point is the unraveling of a Byzantine political plot. But trying to put too many characters in there is asking for trouble (or it’s a sneaky reversion to point 1, where the author really just wants to dump a bunch of people I don’t know in front of me and make me admire them by hook or by crook). So is too complicated a plot.
Some things that should be considered especially hard—much harder than they would be in a novel—before being included in a short story are multiple viewpoint characters, unreliable narrators, political or mystery plots needing a whole lot of explanation in the middle or at the end, non-English words that are not translated or are not obvious from context, gratuitous references to fantasy organizations and governments that are not relevant to the plot, tragic backstories, and anything where the author’s intent is to truly confuse the reader instead of make them read the story a second time and say, “Aha!” And then, of course, always, always remember…
7) You are not Frank R. Stockton. And you are not O. Henry, either. Even harder to write in a short story than the transition from laughter to tears is the trick ending. I’ve seen many, many amateur fantasy authors try to do it. All but perhaps two of them failed.
Frank R. Stockton is most famous for writing “The Lady or the Tiger?” where the reader is left to essentially guess the ending. Many people find it an intensely irritating story. Supposedly you can guess the outcome from the clues in the story itself, but there’s plenty of evidence to support either reading. Just about every attempt to duplicate it falls flat. This is a one-trick pony. Don’t expect the reader to like you very much if you build everything up to a grand climax and then ask the reader to guess who assumes the throne or if a character lives or dies. Ending the story with a question that’s not part of dialogue is generally a very bad sign.
O. Henry wrote stories with surprise or twisty or ironic endings, like “The Gift of the Magi.” Lots of people want to be him. Doesn’t work, either. Your surprise ending has to grow naturally out of the story yet still be a surprise, depend on clues yet not let the reader guess the clues beforehand, and be original while still not seeming like a random joke or a Supreme Copout (the “It was all a dream!” ending is an example of the Supreme Copout). It’s damn hard. Aiming for a twisty ending will almost certainly dump you in the dust.
There might be a third motive for writing a fantasy short story like this, namely that the writer intends to answer the unanswered question or explain the ending in another short story or novel. DON’T. Unless you’ve already gotten people interested in the next short story or novel, it generally doesn’t work. Prequels are much better when the audience can see that they’re actually prequels.
Again, not sure which one to do next.