First of all, happy J. R R. Tolkien’s birthday!

Second, it’s a good thing that I did the rants in the order I did, or I would have been completely confused about which one people wanted next. (There are three that are absolutely equal in the poll).

1) Walk the line with figurative language and new vocabulary.

After reading Sarah Micklem’s Firethorn, I might not have been able to tell you exactly what a ‘jackman’ or a ‘cataphract’ was, but I sure as hell knew what they did. Micklem uses the terms again and again, repeating them without any obnoxious, paragraph-long explanations, characterizing both occupations in terms of their functions. If only more fantasy authors could do this, instead of going off into half-page-long spiels that identify terms they’ll only rarely return to! Since 9/10 of the action of Firethorn takes place in an army camp, both these occupations—one a nobleman’s servant in charge of everything from food to raiment, one an armored and armed nobleman of a specific number of clans devoted to the gods riding a similarly armored horse—matter and appear often. But Micklem doesn’t bother jerking the audience out of the story to present them in clinical detail. How they work for the people of her world is enough for her, and it’s that understanding that the reader comes to share.

Likewise with figurative language. Authors tread a fine line between the banal (“black as a starless night,” “blue as a summer sky”) and similes and metaphors so impossibly involved that the audience steps out of the story trying to understand them (“hair like a summer silk breeze,” “a sky that changed color like the Eiras Weintrop”). Both tend to make the readers aware of the language as language, rather than immersing them in the story. Whether irritation or confusion results, your reader’s attention wanders away from the characters and plot and world that should be enticing them.

Try to provide some context to go along with your unfamiliar terms—but only some context. In the example of the Eiras Weintrop, because I don’t know if it’s a place or a plant or a thing, I have no idea how to interpret the simile. On the other hand, neither should the author use this as an excuse to infodump on the Eiras Weintrop for a paragraph. Instead, he should quickly provide me with clues that will let me figure it out on my own: “The sky changed color like an Eiras Weintrop that had suddenly noticed it was spring.”

2) Explain laws, customs, history, backstory, and so on as they become relevant.

I admit it: I tend to read fantasy exposition by skimming. When the author attempts to make me swallow a whole list of names, a whole history of a royal line, a description of a city, or whatever, I skim past it, snatch what I can, and return to reread if I think I need to know something. If I’m on page 400 and I still don’t know what rank Lord Temblor holds or why everyone defers to him, I’ll return to the infodump and look it up. (On at least a few occasions, I’ve caught authors cheating this way; for example, they insisted that the nobleman wasn’t important at first, and then elevated him in others’ eyes without mentioning why).

An easier way on everyone involved is to mention these things as they become important. If Lord Temblor ascends in rank because he fought and won a desperate battle, then mention him casually at first, then with increasing insistence, and finally detail his relevant family history, hair color, and what he likes to drink with his morning tea when you bring him back to the capital. This has a number of salubrious effects:

  • It introduces the character naturally, as the other characters learn about him, without having to interrupt the story with omniscient burble. Or, if you do have to eventually go omniscient, it’s not at the beginning, when the reader’s trust in the author is still fragile and she can be knocked out of the book by a display of infodumping incompetence.
  • It enables readers to remember the character more easily. Multiple small mentions will always serve better to establish Lord Temblor by his all-important speech on page 400 than one giant text block about him on page 4 and then no mentions again until 400.
  • It enables the author to weave in hints and clues, especially if the story is supposed to be a mystery.
  • It gives hope that the reader will not be like naughty me, and skim past that essential worldbuilding information you’ve worked so hard on.

3) Exception to the above: Explain allies and saving graces long before the end of the story.

Got a mysterious country across the sea that will send ships to the aid of your king in the final battle? They better be there early on.

Have a reason that the hero can make a fourth wish, when every other wishmaker in existence has only had three? Hit it early, hit it hard.

Sending dragons with a bad allergy to spikenard against natives who live among the nard plants? That allergy should fucking show up before the dragons arrive on the spikenard fields in page 500, please.

The reason for this is simple. While laws and customs and backstory often add just a touch of color to the narrative, and the reader may not miss them if he misses them—that made sense, right?—or can go back and skim to find them, allies and saving graces are often the reason that the kingdom/hero/world survives. Fantasy is addicted to loopholes. I think it should stop being so addicted to them. But the least a fantasy author can do is make sure that they don’t appear to be deus ex machina devices, which have ruined more fantasies than I care to count.

One reason authors often err on the side of too little explanation here is because they want a surprise ending. You aren’t supposed to be able to figure out that the hero can make a fourth wish until he makes it. Yet the line here is even thinner than the one between banal and incomprehensible figurative language. “No clues” is not the same thing as “clues that the reader can put together later.” When in doubt, put in another minor explanation of the ally or the saving grace.

4) Trust that your reader can remember and follow conversations.

While the author may not include enough explanation about the ending for the reader to find it in a paper bag, they tend to make sure the reader would know information the characters discuss in dialogue in a haystack. Blindfolded. With both hands cut off.

A common pattern in fantasy conversations is for the characters to discuss their problems, suggest solutions, and then set up reasons the solutions would never work. Say they are stranded out in the wilderness with knowledge of the evil uncle’s plan, but have no way to get back to the palace before the good princess’s coronation, when the uncle plans to strike, and no way to send a message, either. They talk about ways of fast travel, magic, convincing someone else to take a message for them, and so on. One by one, they discount the solutions, until they hit upon one that works.

The thing is, fantasy authors seemingly don’t trust the reader to remember all that. There will be a few paragraphs of ordinary conversation, then one devoted entirely to a character summing up what the others have just said. Like so:

“But I don’t see how we can reach the palace in seven days, never mind five,” said Elzara, tone sharp as flint. “So pardon me for disliking the idea of galloping our horses to death.”

“Pegasi could make the journey,” muttered Feltaran. His drawing in the dirt had grown quite complicated by now.

“Will you stop playing with that stupid stick and listen to what I’m saying?” Elzara yelled, knocking the twig out of his hand. “We don’t have pegasi, either!”

Orich sighed and leaned back against the tree. Our horses can’t make the journey without dying in seven days, he thought in dismay. And we don’t have any pegasi.

What the hell does Orich’s thought need to be in there for? I think the readers are capable of holding what Elzara and Feltaran said in their heads for longer than that. They just said it, after all.

Dialogue tends to explain things a lot more clearly than action or description or poetry. Conveying information is one of its primary functions, after all. So there’s no need for continual summing-up. One great explanation at the end, maybe.

But even that has its perils.

5) Avoid scenes of pure ‘explaino.’

Fantasy novels of conspiracy, politics, intrigue, mystery, and the like tend to inherit these from detective novels. There has to be a scene (or a chapter, or even two chapters) where the fantasy character currently playing detective sits everybody down and explains it all.

The main problem? It’s the same problem with the omniscient burbling in block text at the beginning. They’re boring.

Why? Because of too little explanation or overexplanation. Far, far too many fantasy political landscapes imitate the Byzantine. Even after the explaino scene, there are still lots of unanswered questions—oftentimes of pure logic, since that’s something the fantasy authors tend to ignore in favor of giving their heroes multiple chances for daring last-minute escapes. Or else, what the author reveals is what the reader knew already, because the reader’s been peering over the hero’s shoulder as he discovered it. So it serves no more function than the constant summing-up paragraphs in dialogue, except that this is the Mac Daddy of summing-up paragraphs.

One thing fantasy can do more easily than almost any other genre is have things happen. Big things, important things, things to make the reader laugh or cry or shout. So have things happen during the explaino scene. Have a villain attack the character. Have the character hog a bit of important information to himself until the last minute. (This works best in multiple-viewpoint stories, since otherwise it seems as if the author’s been having the narrator cheat or flat-out lie until the end). Have something render the explaino scene useless, my utter favorite. The character researches until he can self-righteously present the tale of the evil nobleman who’s been trying to foment rebellion…only to find out that the king knew all along, and doesn’t care, because that rebellion is the excuse he’ll use to crush the nobleman and take his lands. Why should the hero be the only one who gets to play dramatic last-minute cards? And it sets up the next book of the obligatory trilogy, too.

6) Try providing multiple character motivations.

And then not explaining all of them.

As well as the summing-up paragraphs and the final explaino scenes, scenes of psychological analysis haunt fantasy. This is the one where the hero realizes that the villain has been stalking him because the hero murdered his father years ago, and “experienced in one horrible moment all the terrible pain of those years, all the grieving, all the muddy despair of days and nights.” Of course, because you are not allowed to get away with stalking the hero, and no one ever stabs him while he’s standing there experiencing all this muddy despair, the villain either has to renounce his vengeance or die by the hero’s sword. And that’s it. Another human being is reduced to a neatly painted, ultimately flat, jigsaw puzzle piece.

This is the part where fantasy secondary characters really get short shrift. They might seem to have lives apart from the heroes, nuanced characterizations, and more than one desire in life, but that collapses the moment the hero encompasses them with his Mighty Understanding. They’re not allowed to have multiple motivations, because that would be too difficult to explain.

Try not explaining them instead. The hero might think of several possible explanations when the villain tries to stab him and ultimately dies, but he’s not ever going to get the chance to ask, because afterwards the villain’s lying dead in the mud with a smile on her face. He might remember memories she confessed to him, and wonder if they were true or just part of the ploy. Don’t do this “Something told him they were real” bullshit. He might be able to come up with six or seven reasons that she waited until the book’s climax (which he doesn’t know is the book’s climax, after all; someday I must do a rant on authors giving their characters the traits of people who know they’re in a fantasy novel) to stab him. Which is right? Who knows?

This is not, oddly enough, something most readers will let you get away with on the heroes. There, they do usually want to understand, fully and completely, the people they’ve spent several hundred pages rooting for. It’s a trade-off, I suppose. Heroes get more time for their complicated motivations to be spread on the page. Secondary characters get less, but in return, you get to paint them with shadows that suggest depths, even though those depths might not actually be there.

7) Prepare well some commonly-used last-minute avenues of communication, or don’t use them.

This is small, but I will mention it because it bugs me so much.

If a variation on number 6 seems likely to occur, with a character dying with his motivations multiple or unexplored, the hero often gets some word that clears things up. The dead character’s ghost appears to him, or he finds a suicide note, or he “suddenly remembers” that he heard the character whisper she loved him the final night before she tried to kill him.

The main problem with these is that they’re so clumsily used. If you’ve worked speaking to ghosts into your story from the beginning, and perhaps even had the hero and the secondary character make a pact to come back from the dead and speak to each other if one of them died untimely, it’s one thing. But introducing a convenient necromancer at the tail end of the story just so that you can make sure no end is left intriguingly untied?

Quit it.

8) Put exposition in your character’s voice.

So you’ve done everything you could. You’ve woven in subtle hints, you’ve saved what you could until later in the story, and you’ve leavened scenes of pure explaino with what action and humor you can. It’s no good. You’ve still got a great big block of exposition to be digested.

Let your characters masticate it and feed it to your audience for you. (And I will now abandon that metaphor, as it’s gotten really disgusting).

First-person fantasy owns the ground here. A first-person narrator talking about his or her experiences tends to lull us into following along even if we would find it obnoxious in omniscient. This happens in Micklem’s Firethorn. The book starts with a fifteen-page prologue describing the heroine’s past, in which she was beaten by her mistress and raped by the lord of the manor, and the Kingswood where she’s fled to live for a year. I didn’t care. The details of the forest and so on didn’t bore me. I was seeing it through Firethorn’s eyes, as she related what had happened to her, not just what had happened, and that made it interesting.

Yet it can work even in third-person limited, and even in pure conversation where the character’s describing, say, the history of the world to another character. Yes, you can rescue your conversation between the Wise Old Mentor and the Dunderheaded Young Hero from its well-deserved oblivion. Make your Wise Old Mentor not a Wise Old Mentor. Make him Esaragaar, who’s been waiting for this damn prophecy to finish for three hundred years now, and has always had an eye for a pretty young woman even though his body hasn’t been holding up to most of the demands he’s placed on it. Having him mention while explaining that he knew those mountains when they were unknown, and that that dead queen who set the prophecy into motion was like a wildcat in bed, and that he hates living in his secret cave hideaway because his knees ache in the damp, can make the story his story. Far too many authors don’t do this. The information they give could come out of any character’s mouth; it really wouldn’t sound any different if it was the Dunderheaded Young Hero telling it to the Wise Old Mentor.

Your characters are usually what your readers will root for, cheer for, violently hate or love, and find fascinating. The exposition by itself is not, no matter how much you love it for its own sake. Filter it through the characters. Never forget that this should be a story, not a textbook.