Editors for fantasy novels: magic, names, prologues

A fantasy editor reads for six things: whether the magic keeps its promise, whether the invented names are legible, whether the prologue earns its place, whether book one works without the sequels, whether every point-of-view character wants something, and whether the description is doing any work. Fantasy is the genre where a manuscript can be beautifully written, genuinely imaginative and still unpublishable, because the problems are structural and they hide behind good prose. A copy editor will hand it back clean and an agent will pass on page two, and figuring out why is most of what a fantasy developmental edit is.

Here's what we actually look for.

1. Magic-system consistency is a promise, not a rulebook

Writers hear "hard magic versus soft magic" and assume the job is picking one. The job is making sure the book keeps whichever promise it made in the first fifty pages.

If your opening establishes that magic has a stated cost, the reader starts doing arithmetic and will notice when the cost isn't paid in the climax. If your opening presents magic as numinous and unexplained, as in much of Tolkien, the reader stops doing arithmetic and the climax can be strange. Both work. What doesn't work is running soft magic for three hundred pages and then resolving the plot with a mechanic, or running hard magic and then resolving it with a miracle.

So the note in a fantasy edit is rarely "your system is inconsistent." It's usually that your system is doing a different job in act three than the one you trained the reader to expect in act one. Readers treat a magic system the way a solver treats a logic puzzle: once the constraints are set, they check every later answer against them. If you want a feel for how satisfying a rule that holds can be, try one of the free daily logic puzzles we host.

Things an editor should be tracking across the whole manuscript:

Every stated limit, and whether it's ever violated (including violations the narrative treats as fine).
Every cost, and whether anyone we care about actually pays it on the page.
Whether the protagonist's power at the end could have solved the problem at the beginning. If it could, the middle is a delay, not an escalation.
Whether the antagonist is bound by the same rules. Villains cheat in more drafts than you'd think, and readers feel it even when they can't name it.

2. Invented names: the legibility budget

When a reader hits a name they can't pronounce, they don't sound it out. They give it a shape, a blur with a first letter, and move on. That's fine for one or two names. It's a problem when three characters in the same faction start with the same letter, or when place names and person names look alike.

Rules I apply on a pass:

Distinct first letters and syllable counts inside any group the reader has to keep straight at once. Kaelen, Kethra and Kaervan in one war council is a memory failure waiting to happen, however good the council is.
Different shapes for different categories. If gods, cities and people all follow the same apostrophe-and-diaeresis convention, the reader can't tell what kind of noun they're looking at, and can't parse the sentence on a first read.
Apostrophes usually cost more than they're worth. They add a stumble more often than they add exoticism.
Front-load legibility. A name's first appearance is where you can afford one clause of anchoring: a title, a relationship, a body. After that the name has to carry itself.

None of this dumbs anything down. It's about where the reader's attention goes, and attention spent decoding isn't spent caring.

3. The prologue problem

I won't tell you never to write a prologue. I will tell you what one costs, because most drafts haven't done the math.

A prologue asks the reader to invest in a viewpoint, a place and a stake, and then takes all three away. That's a reset, and you pay for it in momentum, so the prologue has to be worth more than the reset. The ones that are usually do one of these:

Establish a tonal contract the first chapter can't (this book contains real violence; this book contains real magic; this is a story with gods in it).
Plant a concrete object or event whose meaning lands three hundred pages later, and which the reader can still recall by then, so it has to be vivid rather than informational.
Pose a question the book spends its length answering.

The ones that aren't worth it are mostly history lessons: an omniscient voice recounting a war a thousand years ago, with names the reader has no reason to hold. That's worldbuilding in the least efficient container available. It's material you want on page two hundred, when the reader is invested and the names mean something.

The test: cut the prologue and start at chapter one. Does anything in the first hundred pages become incomprehensible? If not, the prologue was a warm-up and belongs in your notes file.

4. Series setup starves book one

This has gotten more common as writers plan trilogies before drafting. Book one withholds revelations "for later," ends on a hook rather than a resolution, and introduces a cast sized for three books.

Commercially, book one has to stand alone, because if it doesn't there may be no book two. Editorially that means:

Your protagonist's book-one want must be resolved in book one. Not the series question, the personal one. The world can stay on fire.
Withheld information only builds tension if the reader knows it's being withheld. Otherwise it's just absence, and absence reads as the author not having decided yet.
A cast of thirty introduced in act one is a cast of thirty the reader abandons. Introduce people when they act. A character whose only job in book one is to matter in book three is a book-three character.

5. Point-of-view discipline in a big cast

Multi-POV epic fantasy has a specific failure: viewpoints with unequal weight. Say you have six POVs and two of them are plot-delivery mechanisms, like a chapter from a character with no want who exists to witness the siege so the reader can see the siege.

Every viewpoint needs its own want, cost and arc, or it should be folded into somebody else's chapter. Readers skim POVs that don't want anything, and a skimmed POV is worse than a cut one, because it also teaches the reader that skimming is allowed.

6. Description that stalls instead of moving

Fantasy invites lavish description, which is fine as long as it's attributed. Description filtered through a character with an opinion about what they're seeing moves the story, and description from a neutral camera stalls it.

This is a line-edit-level issue that shows up as a pacing complaint, so writers often misdiagnose it as "my middle is slow." It isn't the middle. It's nine hundred words of scenery with no one inside them. Color-coding your prose types is the fastest way to see the imbalance, and we walk through it in how to save your writing with a pack of highlighters.

What we've worked on

We've run live development edits on fantasy and fantasy-adjacent drafts, including Demon Priestess and Prologue and Ozma of Oz and Stuck in Silence. On the writing side, our science-fantasy novel The Distributor was published by The Story Plant and is listed on the publications page.

Which edit do you need?

If your beta readers say "I got confused" or "I put it down around the middle," that's developmental. If they finished it happily but the prose feels flat, that's a line edit. The full comparison is here, and the Get Edits page covers how to send us something, including free Twitch edits if you're willing to have your draft discussed in public.