So You Want to be a Ninja...

Basics. Spelling, grammar, punctuation - these are your katas, the fundamentals. Any peasant can throw a punch or toss together a grammatically correct sentence. You must know why it is correct. You must be so familiar with the rules that even your Twitter updates are punctuated properly. Only then can you improvise, creating your own forms - doing so by intent, not laziness.

Words. Words are your weapons, and you must become familiar with as many as possible. More than familiar, you must become adept in their use. A simple farmer can pick up a sword and make a clumsy effort at wielding it. You must be its master. In addition, you must know which weapons are appropriate for a situation. A polearm is all but useless in assassination, as 'puissant' and 'scion' would find a poor home in the mouth of the common taxi driver.

With knowledge of weapons and katas, you would make a decent fighter, a writer of e-mails, a composer of persuasive essays. Any daimyo would be glad to have you among their common militia, but you would not be a ninja.

Style. Fighting is more than killing your opponent, and writing is more than words strung in the proper order. The samurai know this, and you can learn much from them. You must be aware of the clarity of your writing, the variation of sentence structure, the powerful techniques of imagery and metaphor. Writing is an art, not simply a means of communication.

With a knowledge of style, you could choose your own path. You could become a mercenary, writing for whomever would pay you. You could begin the path of the samurai, accepting their bushido and writing only the truth - news, non-fiction, and the like. If you seek a life of security and reputation, then perhaps the way of the samurai is for you.

Or you could begin the life of a ninja. To the samurai, bushido is life. To the ninja, it is a hindrance. The art of the ninja is one of lies and misdirection, surprise and subterfuge. To become a ninja, you must learn many techniques the samurai are not taught, master them, and make them your own.

You must learn the secrets of tension and plot, what drives a story forward and hooks the reader until the end. You must learn to create characters that are real, believable, and can gain or lose sympathy with the audience, as the situation dictates. You must understand the ways of dialogue to make your characters to speak without tearing down the lie you have constructed.

Once you have learned everything required to be a ninja, you will have only just begun. Millions have gone before you. Most do not survive. The shinobi masters whose names you've heard are the exception, not the rule.

It takes more determination than you've ever known to become a ninja, but you can do it. I believe in you.

And if I'm wrong, it won't matter. You'll be dead.

Subtasks and an Air Pirates Excerpt

I realized one of my problems with editing is that I can't keep track of my progress as easily as I can with writing the draft. The result is that I feel unproductive which, ironically, makes me unproductive.

I think I've solved that by cutting it into relatively bite-sized stages. I did that before, but these are a little more concrete (i.e. for each stage, I have about 10-20 specific items that need changing).

Stage 1: Strong Ending. (Includes all the changes to beginning and middle that will help improve the ending).
Stage 2: Continuity.
Stage 3: Global Changes (i.e. capitalizing officerial titles or being consistent with the definite article and ship names).
Stage 4: Other Revisions. (Mostly improvement notes I made while drafting).
Stage 5: Strong Beginning.
Stage 6: Full Read-Through. (The subtasks of this one are individual chapters).

I've finished stages 1, 2, and 3. I'm almost done with stage 4, but stage 5 will be hard and stage 6 will be super-rough. Hopefully if I can track my progress and check things off of a list, it will help me stay determined. (I do like lists).

The title of this post also promises an excerpt, so here you go. I had a hard time finding one, since the beginning has all the world explanations in it but is also the least-polished.

This is the beginning of chapter 3. The stone has shown Hagai a vision in which the Oleanna, a merchant airship, crashes into the docks. Unsure whether he's seeing the future, he runs to the docks to find everything exactly as it was in the vision prior to the crash, and he starts to get worried. (In the excerpt, a dak is like a goat).

Chapter 3

The
Oleanna hung in the sky like a child's lost balloon, no bigger than Hagai's thumb at this distance. For a moment, Hagai thought the string of coincidences might end there. It didn't look like anything was wrong. If anything, the Oleanna was getting smaller as it floated away.

That was before he heard the explosion.

It was really quiet for an explosion. Just a small, muffled pop like distant fireworks. Hagai thought he imagined it at first. For a while, he hoped maybe it wasn't an explosion after all, until he saw smoke issuing from the top of the
Oleanna above him. That was also when the daks started getting nervous.

"Did you hear that?" Hagai asked the dak counter.

"Nay."

"That explosion, up in the sky. You didn't hear it?"

"Nay," the man replied, but he looked up towards the smoking airship.

"I heard an explosion, and then that smoke started coming out of the Oleanna."

The man squinted. "Nay, they're breezy. That's just exhaust from her boiler."

"No, it was an explosion. I heard it. I think they're in trouble."

"She's fine, see? She's still going."

The daks were getting louder and starting to push against the fences. Only Hagai seemed to notice. "Okay, look," he said, "I know this is going to sound weird, but… I
know that ship is going to crash into the docks. We need to get out of here. We need to warn everybody to get out of here."

"What?" he laughed. "What are you on about?"

"I got this stone, see, and it shows me the future." He put his hand on the sling.

"What'd you get that at the circ or something?"

"No, look..." Hagai looked up again. The airship looked bigger. "It's gonna... it's gonna crash right
there." He pointed. "People are gonna get hurt. The… the daks there are gonna break out and run wild, and, and you break your leg, and - "

"Here, now. That some kind of threat?"

"No! No, I just… I mean that's just what I saw."

"I got work to do, dog. Take your circus toy home."

"But I..." Hagai started, but the man had already turned around. The daks were restless now, and the dak counter looked pretty upset about it. Hagai left him alone. He looked for someone else to tell. There were plenty of skylers and dock workers everywhere, but they all seemed real busy. He considered yelling out a warning to everyone, but decided against it. He tried a one-on-one approach.

To one man he said, "Excuse me, I - "

"Outta my way, boy."

To another: "I'm sorry to bother you, b- "

"Does this look
light to you?"

"Excuse me, is that ship on fire?"

"Boiler exhaust. Please to move."

"Hey, that ship's falling."

"Nah, it's just coming in to port."

"But it just took off!" Hagai pleaded to the back of the last man.

Nobody would hear him, and the airship was getting closer.

19 Months of Air Pirates

It took me 19 months from word 1 to word 99,675 on Air Pirates. I told you I'd talk about the process, so here it is.

I write these kinds of posts to analyze my own process, to see what works and what doesn't and hopefully to improve. I hope other people can glean something useful from this as well, but if not, no worries.

This post is also an excuse for me to make charts. I like charts.


I started writing Air Pirates officially in September 2007. I'd been thinking about it for a lot longer than that (I have documents dating back to 2003), and I'm still not finished with it (the editing phase is a lot harder than I remember), but the first draft took about a year and a half to write.

You can see I had a slow start, and a plateau in early 2008 that I'll explain in a second. This next chart shows that more clearly. It's a chart of how many words I wrote in each month.


Clearly there's an upward trend, but it's pretty unstable. I'd write well for 2-3 months, then stop for 2-3 months, and so on. Here's how it went down.

Aug-Oct 07 (Slow): It took me 3 or 4 drafts to get the beginning, and rewrites weren't counted towards the final word count. Once I figured out how the story started, it got easier.

Feb-Apr 08 (Slow): At this point, I'd written four chapters of Air Pirates and needed to outline the rest of the story if I was going to get anywhere. Travelers also came back from the betas and for the first time I started seriously editing and working on queries.

May-Jul 08 (Fast): I really think the reason for my increase here was that I started thinking like a pro. In working on my query, and trying to figure out what to do next, I'd begun reading blogs like Nathan Bransford and Query Shark, as well as author's blogs. I learned a lot about writing good query letters, but also the publishing business as a whole. More than anything, I learned that this was something I really wanted to do.

Aug-Sep 08 (Slow): I tried writing a short story in the Air Pirates' world. I'd never written a short story for the purpose of publication before, so it took a lot of time (and it still isn't published). Our trip to the US slowed things down a bit too.

Oct-Apr 09 (Fast, mostly): The increase in word count here is most likely due to my wife's commitment to me. She started helping me get away for 2 hours a day, most days, just to write. I still wrote in whatever stolen moments I could, but having 2 undistracted hours really helped me discipline myself. I want to blame the dip here on holiday visitors, but I think it was due more to my lack of self-discipline than anything. I'm trying to get better at that.

Final Analysis. On average, I wrote 5,000 words a month on Air Pirates. I know that, with discipline, I can do 10,000 a month pretty regularly. Hopefully that means I can draft a novel in 9-10 months, but we'll see. I won't be doing any drafting for a while yet until I can get through all this editing I'm supposed to be doing.

Prologues

I don't know why I'm thinking about prologues. Nothing I'm reading or writing has one. But I'm thinking about them, now you get to, too.

What is a Prologue?

It won't do much good to talk about prologues if we don't agree on what they are. In fiction, there are three things that make something a prologue: (1) it comes before the first chapter, (2) it is a part of the story (as opposed to an introduction, preface, or forward, which are about the story, but not part of it), and (3) it says "Prologue" at the top.

Simple, right? That's what makes something a prologue instead of, say, "Chapter One," but it doesn't explain what makes a good prologue. That's what this post is about.

When Not to Prologue

A lot of people don't like prologues. Some people skip them entirely (which, to me, is way wacky). That's because a lot of writers use prologues as a band-aid for a bad beginning. Which is to say:
  1. Don't use a prologue because you need a better beginning. Fix your beginning.
This is important. It's hard for a reader to get involved in a new story, with unfamiliar characters and situations. Adding a prologue requires the reader to start your story twice; when the prologue's over, your reader has to get into the rest of the story. So:
  1. Don't use a prologue just to suck the reader in. You'll only have to suck them in a second time when the prologue's over.
These prologues are trying to create artificial excitement. Some prologues have the opposite problem. Instead of providing an exciting false start, they begin with boring exposition because the author is afraid the reader will become lost without all the background.

Like every prologue, this creates two beginnings, but instead of Exciting followed by Flat, the expository prologue starts Flat, with the Exciting beginning buried beneath it. Sci-fi and fantasy are notorious for this. A good genre writer, though, is able to mix telling details into the story so they don't have to put it all up front in one big exposition. So:
  1. Don't use a prologue to explain the world or backstory or any other kind of telling exposition.
When to Prologue

Everybody has had a bad experience with prologues, but I don't think they're all bad. If used wisely, they can be quite effective. For example, sometimes a story is told entirely from one point of view, but you need to clue the reader into some event the protagonist never witnessed (and it needs the impact of being dramatized). In this case:
  1. Use a prologue to show a point of view that doesn't appear anywhere else, or doesn't appear until the end.
This can be especially effective in mysteries and thrillers, where there is tension behind the scenes that the protagonist is unaware of. Say the Villain shows up in a prologue, kills somebody (so we know he's bad), and says, "Where's Paul Protagonist? He's next!" Now, when we meet Paul in Chapter One, whatever he's doing will be flavored with this tension because we know someone's after him. So:
  1. Use a prologue to create tension that the protagonist is not immediately aware of.
Lastly, have you ever gotten into a story that was all dragons and swords and magic, only to discover that the evil villain is a space alien with his own spaceship? Genre blending like this can be done well, but if it's done poorly you end up sucker-punching the reader (helpful tip: readers don't appreciate being sucker-punched).

Orson Scott Card's Homecoming saga is about a low-tech society of people whose religious values are challenged by a boy that hears from God. This would be fine except it later turns out that the boy's God is an artificial intelligence orbiting the planet and watching over their society. That's the kind of thing that would make a reader throw the book across the room unless there's a prologue (in this case, from the AI's point of view) to show or hint at the truth of the situation. So:
  1. Use a prologue to manage the reader's expectations about your story.
Final Tip

The main point of all this is that a prologue isn't an easy way out of anything, least of all out of grabbing the reader's attention - that still needs to be done in Chapter One, whether there's a prologue or not. So how do you know if you need one, or if you're just being lazy? From Nathan Bransford:
  1. Take out the prologue. If the book makes sense without it, you don't need it.
Note it doesn't say, "If the book is boring without it, then put it back in." If the book is boring without the prologue, something's wrong with the book, and a prologue won't fix it. Remember, the reader will be spending most of their time in the book, not the prologue, so put most of your work there.

I have some additional examples in the comments. Feel free to add your own, good or bad (or even to contradict what I just said!).

Playing Agent for a Day

Nathan Bransford is running an interesting game on his blog called Agent for a Day. On Monday, he threw 50 queries up on his blog, at random times, to simulate what happens with his slush pile (3 of them are queries that led to actual, published books). Those who want to play need to read the queries and request or reject as if they were an agent, but we're allowed to request no more than 5.

After 4 hours (interrupted by toddlers and a meal or two), I finished all 50. I probably could've done it faster if I just said a quick "yes" or "no" (or better, if I just didn't respond if I wasn't interested), but I also wanted to help those whose queries got chosen. So I left a short suggestion on most of them.

Anyway, here's what I learned about query letters:
  1. Most bad queries were vague with the details. Instead of saying, "Frodo must keep the ring from falling into the clutches of Sauron, the dark wizard," they'd say, "Frodo is up against the forces of evil." Instead of "Meg Ryan finds herself attracted to the arrogant bookstore owner who's running her out of business," they write "Meg Ryan finds love in the unlikeliest of places." This is bad for two reasons: (1) vague is boring, specific is interesting and (2) without specifics, your story sounds like every other story ever written.
  2. Many bad queries were vague with the ending. The premise sounded interesting, but I passed because I wasn't sure if the story delivered on the promise (and there were lots of other queries that did).
  3. The little mistakes that sites like Query Shark and Evil Editor rail against (e.g. mentioning you were a finalist in a writing contest, or putting word count/bio info first) were never a reason for my rejection. If the premise was good and the query well-written, I didn't care about anything else.
  4. Some little mistakes were the reason for my rejection however. For example, if a query, or even a paragraph, was too long, it could make a decent query hard to understand and the story hard to find.
  5. Almost everybody had good ideas. Not everybody knew how to write about them.
  6. Not a single query was perfect. Even the 5 I chose had points against them.
Nathan asked us to look for stories that were publishable, whether or not they were our favorite genre. Even so, it was really hard for me to be objective. Every time I saw a SF/Fantasy hook, I got really interested and gave the query more grace than I might have otherwise.

The game has me worried, though, because I could imagine what my query would look like amidst the slush. I don't know if I can write a query that would stand out, but these tips will help, I know.

The queries are still on the blog, and the game runs through to Saturday. So if you want to play you can (and you don't have to do it all at once, like I did).

Two Important Announcements

First announcement: As of this moment, I am officially a published author. Thaumatrope has published all 22 words of a science-fiction Easter story I wrote for them. You can read it here.

I have a second story due to be published on August 17, this year. You don't have to remember that date. I'll link to it here when it happens. Also, Thaumatrope is open for submissions again, including serials. If that sounds interesting, there's more information here.

Second announcement: The first draft of Air Pirates is finished, after 19 months. That's a long time, but... well, I'll give you a breakdown of that whole process one of these days. With graphs and everything. It'll be great.

You'd think the hard part is over, but it's just beginning. Brainstorming, outlining, plotting, writing... these are the fun bits, when the story's fresh and exciting in my mind, and nobody else is telling me the truth (i.e. that it's not fresh or exciting). So let's take a look at the next couple of months:
  1. Ending: Fix the resolution to meet the standards of my Beloved Alpha Reader and myself. This is currently the best ending I've written so far, but it is still merely mediocre - not as satisfying as it could be. Cindy's helping me make it better. ETC: 1-2 writing days.
  2. Continuity: Go through the revision notes I made while writing and repair the novel's continuity (I broke it a little during the later chapters because the way I'd planned things wasn't working). ETC: 1 writing day.
  3. Polish: Read through the whole novel, fixing everything I see. ETC: 1-4 weeks? I'm not really sure.
  4. Beginning: Go back one last time and give the opening special attention. ETC: 2-3 writing days.
  5. Beta Phase: Send the novel out to beta readers and await their response. ETC: 2-6 weeks.
  6. Freak Out: Getting feedback is the best and worst part of this process. I love hearing what people liked (especially when it's some scene taken the way I meant it). I hate hearing about things I need to fix.* ETC: Ongoing.
I'll talk more about the beta phase when I get there. For now, I need to work up the nerve to revise, and I need to figure out what daily stats mean for me in the revision phase. I think that's part of what this list is for; it's a means for me to track my progress and help me feel better about myself.

I have to admit, I'm kind of rabbit-in-the-headlights right now. I'm scared to take a break, and I'm scared not to. I'm afraid to touch it, but I know I can't leave it. I just need to sit down in front of it and make something happen, I guess.

* This is the part where I'm not thinking like a pro yet. I get critiques and think, "Crap, you mean it isn't finished yet?" Hopefully I'll be able to get over this attitude with some practice.

Cliffhangers

There's one problem with endings the way I described them. Like Dr. Manhattan said, "Nothing ever ends." Stuff happens before a story starts and stuff happens after it ends. The end is not an end, just a stopping point. So in a sense, something always has to be left hanging.

For standalone stories, that's often stuff we don't care about. The killer is caught, and we're content to know everybody goes back to their normal lives. The guy and girl get together and live happily ever after (or maybe not, but we're content to know they start the rest of their lives together).

But what if it's part of a series? How do you provide an ending that satisfies the reader, while leaving gaps open for a sequel? I think it's related to the promises I talked about last time. One of those promises is that all the reader's questions will be answered, or at least touched on.

As an example, let's look at another classic: The Empire Strikes Back. Going into the climax, what are the viewer's questions?
  1. Will Luke escape the trap?
  2. Will Luke rescue Han, Leia, and the rest?
  3. Will everyone survive?
  4. Who is the "other hope" Yoda mentioned when Luke left Degobah?
The first two are answered completely, the third partially, the fourth not at all. Luke escapes (barely). He is not able to rescue Han, but he rescues the rest of them. Everyone survives (meaning they are brought back to a place of safety), except Han.

The fact that we don't know who the "other hope" is doesn't bother us because it's not a Big Question. It's mysterious and intriguing, but it's not vital to this story. These are the kinds of questions that can (usually) be left unanswered without causing an uproar.

But what about Han? The question "Will everyone survive?" is a always a Big Question for action/adventure stories. So why is this okay? I can think of three reasons.

First, we know there will be a sequel. If Empire were pitched as a standalone, people would have been more angry about Han. Lesson: Don't leave big cliffhangers unless you're sure you'll be able to write the sequel.

Second, Empire gives us lots of clues as to the direction the story is headed after the ending. We've been assured that Han is alive and safe in the carbonite. Lando has earned our trust, and he and Chewie are going after Han. Luke says when he's well, he will do the same. Lesson: Give the reader clues so they can piece together their own ending.

Third, the story of Empire was not about Han. Empire was about Luke learning to be a Jedi, and what it meant for him to face Darth Vader. That story was resolved satisfactorily. But if Han had been frozen in carbonite near the beginning of the story, and most of the movie was about trying to save him, then we'd be pissed if at the end he wasn't either rescued or dead. Lesson: Determine what story you're telling, and resolve that story completely.

Finally, I want to say that ambiguity is dangerous. There probably were some people who were angry at the way Empire ended. With ambiguous endings, somebody will always be upset that not all the threads were tied up. Everyone has their own threads they care about more than anyone else, and if you happened to leave their thread hanging, they'll be upset.

So I guess the lesson is: Be ambiguous at your own risk.

Endings, Again

I'm only a couple of days out from finishing Air Pirates. It's exciting (obviously). On the other hand, this last chapter has been taking longer than the others because it's the end. It has to be good. To do that, as Natalie said in her post on endings, I have to take it slow.

So obviously I've been spending a lot of time thinking about what makes an ending good. I've mentioned before that I have a problem with endings. I think I'm starting to figure out why.

As a writer, one thing you have to realize is that while you're telling your story, you are making certain promises to your reader. Some of those promises are inherent in the genre you're writing: if you're writing a murder mystery, you promise the reader will learn who did it and why; if it's a romance, you promise the right people will get together in the end.*

But genre aside, every story makes promises, and it's your job to give the reader what they want. That doesn't mean you have to be predictable, but throw in the wrong kind of twist and your reader will throw your book across the room in frustration.

Let's look at an example. Halfway through Back to the Future, everything's set up for a big climax. The two major conflicts (will Marty get back to the future? can he get his parents together so he still exists when he gets there?) are set up so that Marty's only chance at both is at the same time: the night of the Enchantment Under the Sea Dance. How does this have to end? You might think there are a thousand ways it could end - after all, anything's possible - but the truth is that the viewer is expecting a very limited subset of what is possible.

In BttF, the viewer expects Marty to get home and his parents to get together. Why? Because it's a light-hearted, funny movie. From the very beginning, the movie sends subtle clues that this will be a fun story, which implies a happy ending. There are a number of twists that can happen, but if Marty dies in the end, or gets stuck in the past forever, the viewer will be upset.

BttF also sent signals about what kind of climax it would be. Because there are action scenes (the Libyans attacking Doc Brown, Biff and his goons chasing Marty), the reader expects not just similar, but bigger, action for the climax. Because the movie is funny, we expect a little comic relief from the climax (or at least aren't blind-sided when it happens).

There's more. Marty's dad didn't have to become confident, did he? Could Marty have gone home and found everything exactly as he left it - loser parents and all? He probably could have, but we're all glad he didn't, because the viewer expects the characters they care about will not only win, but win big (or, if it's a tragedy, lose big). It's not enough for George McFly to get the right girl, he has to become more than he was before Marty interfered. Marty doesn't just come back home, he comes back to something better (a new truck, Doc Brown lives and is a closer friend to him than ever).

I'm not saying all endings have to be happy and predictable, but they have to be satisfying. They have to be bigger and better than anything that's happened in the book so far. If you twist it, the twist should be better than the straight-forward ending would have been - don't twist just to be unpredictable (woah, that advice came out of nowhere, and I realize I need to follow it!).

This post is long already, and there was something else I wanted to say about cliffhangers. I'll put it off for another post. In the meantime, ask yourself, what has to happen in the end? Twists and details aside, where do the characters have to end up for me to be satisfied? That's where the ending needs to go.


* These rules aren't strictly true 100% of the time, of course. They can be broken, but if you don't know what you're doing, nobody will put up with you breaking them.

Language Problems

(For WAG #6: Overheard, in which the goal is to eavesdrop and notice how people really talk. I had some problems, as you'll see, and so shattered the rules into tiny, glittering pieces on the floor. Wear shoes.)

I can't understand you. I've lived here 4 years, even went to school to learn your language, but I can't understand you. You're not speaking your language.

I realize they didn't teach me about languages in the US. Oh, they teach the big ones - Spanish, French, maybe Chinese - but nobody is expected to actually use them. When someone, or some country, doesn't speak English, the typical American sentiment is, "Why don't they just learn English?"

I didn't want to be like that, so when I came here I determined to learn Thai the best I could. For you. I've done that, am still doing that, but though Thai is the national language, it's hardly the language everyone speaks.

Even in my own house, Thai is everyone's second language. My wife and I speak English. My oldest daughter grew up speaking Karen. My youngest daughter speaks Lisu with her mom, Kham Mueang with her school friends, and English with me; she only uses Thai with her teachers.

By your rhythms and sounds, I know you're speaking Kham Mueang. I recognize it, even know a word or two, but I don't understand you. In northern Thailand, everyone speaks Kham Mueang. Everyone but me. In the market, if my face didn't already mark me as an outsider, my use of the so-called "national language" would.

It's not really fair, you know? Why can't you all just learn English?

----------------

Other participants in WAG #6:

How to Join the Writing Adventure Group

Cora Zane

Christine Kirchoff

Nancy J Parra

Mickey Hoffman

Sharon Donovan

Iain Martin

Criss

Lulu

Jon Strother

Marsha

Carol

Book Trailers

I didn't know there was such a thing as book trailers until a few months ago. They're... weird. Let me explain.

First, some definition. A book trailer is like a movie trailer. It's an ad designed to get you excited about the book so you'll pay for it. The major difference, obviously, is that movie trailers have about 2 hours of existing footage to work with, every second of which accurately portrays the movie being advertised (because, obviously, it is the movie being advertised).

It's because book trailers don't have this that they're so weird. About the only visual both a book and its trailer share is the book cover. So they have to make do with stock pictures, movie soundtracks, actors playing out scenes expressly for the trailer... none of which are a part of the actual experience of reading the book.

There's another weird aspect to book trailers. Due to the ease of making videos and putting them online, there's a lot of amateur book trailers out there - both fan trailers for bestselling books and actual trailers done by midlist authors themselves. Some of them are pans and zooms of the book cover while the author reads the back cover blurb. Some use stock footage to help visualize the narration. Some piece together clips from actual movies.

Some of these amateur trailers are decent. Many are not.

I think the reason I find book trailers weird is because I'm comparing them to movie trailers. On the one hand, it's an unfair comparison - the mediums are very different. On the other hand, the comparison is demanded; most book trailers are trying to be movie trailers. They're even called "trailers."

Which brings me to my question: what should book trailers be? Should they be like movie trailers - a visual representation of the book? Or should they be something else (and what)?

I'm not sure there's one right answer to this. The better trailers I've seen are good more because of the production quality than because of any methods used.

I will say that I'm not sure about the movie-clips-as-trailer method. It looks cool, but I think it sends mixed messages to the viewer. Especially if the clips are from movies I'm familiar with, I get confused as to what's being advertised. Often I find myself, at the end, more excited about the movies than the book.