An Illustrated Guide to Writing Scenes and Stories
Jeff VanderMeer explains the ins and outs of using scenes in imaginative fiction
If you enjoy reading Electric Literature, join our mailing list! We’ll send you the best of EL each week, and you’ll be the first to know about upcoming submissions periods and virtual events.
The writing workshop/lecture Wonderbook: Scenes is an edited version, using as its starting point the transcript of a version presented at the Arkansas Book Festival in 2014. Both before that event and after, I gave versions of this lecture in other locations, including Shared Worlds, Clarion, and the Yale Writer’s Workshop. While keeping the core of the Arkansas version, I have added in material from the other versions and also expanded some sections based on participant questions. The sometimes informal wording of the original lectures has been retained where possible to reflect the source.
Writers often argue about the difference between the art of writing and the craft of writing. They also argue about what can be taught and what can’t. Writers argue about almost everything — as well they should, since there are a thousand and one ways to reach the same point…and a thousand different points, besides. My position is probably close to that of Vladimir Nabokov, in the way he would combine complex organic discussions of literature with what seemed like a mechanical, rote adherence to the physicality of, for example, settings such as houses in Jane Austen, because this too is very important to the effects a writer can achieve.
Every presentation has a lifespan before it kind of dies in the speaker’s mind. At this point, I have given this lecture so many times, I’ve decided to more or less retire it, and thank Electric Literature for giving it a home. I am indebted to Victor LaValle, who raised the issue of how one would map the actions of the Gormenghast scene below to a different context, like a dinner party. All images below are from Wonderbook, copyright to me and the artist, Jeremy Zerfoss. Thanks to Kari Wolfe for cleaning up the original transcript.
Choosing What You’ll Dramatize and What You Won’t
Good morning. I’m Jeff VanderMeer. I’m the author of Wonderbook.
Most of what we’ll talk about today is going to be about the decisions you make more or less after your rough draft. I don’t want you to think that some of the stuff I’ll be telling you are things that I think should be material or ideas that you’re exploring while you’re writing a rough draft, because when you’re doing a rough draft, you should be in a state where you’re not overly editorial or critical of what you’re writing, usually, depending on your process.
The first thing I wanted to show you is this image, which is about how you decide what the story is in the first place. Basically, I thought it would be useful to take some very dramatic job that a character has — in this case, a dragon slayer– and demonstrate how it is that the average day of a dragon slayer is no different than the average day of an insurance salesman, in terms of not necessarily being of any interest to a reader.
When you’re first thinking about story and scenes, you have to choose what to dramatize, and what you won’t. A lot of beginning writers will think that the continuum of a day, a week, or whatever else in the life of someone is a story. It’s not necessarily. The first part of this is reminding you that all of these things do not need to be in your story. Maybe some of them are dismissed in a sentence or two.
Then the second dragon timeline says to you, “Here are some of the things that indicate that there’s a story present,” and ratchets it up a notch by also putting the character into a more unusual situation. The dragon’s actually destroyed the dragon hunter’s house at the beginning, and that sets off a chain of events.
There is one lie in this diagram in terms of the difference between a person’s day and a story. It’s a built-in defect of the diagram. The lie is that, yes, you can do a day in the life of a character as a story. However that’s not usually good to tell a beginning writer because a writer has to learn how to apply the right elements that sustain a character through a story in such a way that it’s not just a character going through their normal day. Often, when you get to the intermediate stage, or if you’re a particular kind of writer from the very beginning, you can make that kind of A to B work. So long as you know this is what you’re doing, that you’re not just defaulting to a drawing a particular pattern because you don’t know any other patterns.
Where to Begin and Where to End
Once you get to the point where you have a sense of your story elements — the general situations, the impetus or driving force — you still have some decisions to make. You have the shape of your story — in this case, depicted as a lizard — but you still have decisions as to where you’re going to begin and where you’re going to end, not just the story but also your individual scenes. Where you end or begin your scenes is not only a question of pacing. It’s also a question of what’s right for the story you’re telling, for the kinds of characters that you’re using, and in the context of their unique characteristics.
To a certain extent, this is also a diagram of a scene. You’re making more specific or micro-level decisions on where to start and what to leave out, what moment to stop on with what character or what emphasis, what action or reaction or thought. Again, all of these things are not just about making a scene seem to move for the reader, but they’re also about how to frame character and a bunch of other elements. You can literally make a huge change of emphasis just by ending with a different line of dialogue. Stopping a scene with what the person speaking says, rather than how the other person responds, can be a crucial change in emphasis.
There’s also the question of targets. This may depend on how you see fiction, because for me, before I can start a story, before I can know that I’m going to complete a story, I have to know how it ends. I have to have a character in a situation in my head, and then I have to more or less know where the ending is. I often don’t end up where I think I’m going to end up. But I have to have that theoretical, “That’s my end point,” when I start, or I never finish the story.
When you finish a rough draft, there are a lot of things you have to think about — how your decisions about scenes and summary and everything else have supported what you want the story to be. I would consider this, in a way, “reverse outlining” or outlining after the fact. After you’ve written your rough draft, you say, “Well, this was my target. I wanted to write this kind of story with these kinds of things in it. Now, what’s actually there on the page and what’s missing that needs to be there? What scenes are missing? Is emphasis wrong in the scenes? Do I need all the scenes I wrote? Have I thought I created one kind of story, and I actually created another kind of story? Have I gone off into something that’s so esoteric that maybe there’s one or two readers for it — and if so, was that my actual intent?”
By the way, the story gopher in the diagram means nothing. It was something put in there as a joke for the illustrator. He illustrated it. I left it in. It’s an enigmatic story gopher.
Now, the idea of a target in fiction may seem silly, because fiction is multidimensional. It has many targets, many things going on. All the arrows going all of the places they go, and whomever they hit. Let fly, let fall…but at a certain point, when you’re examining how you put your scenes together, it’s useful to, at least for a while, do the thought experiment of, “How is this stuff working?” at some specific level. Not only that, but the question, “How do you create unity and focus, but not be so on point that everything feels mechanical or forced or unsurprising?” Because if it’s mechanical or forced for you, then it’s probably not going to be that exciting for the reader either.
Thinking about Structure Through Character Arcs
What is a scene doing? How is it expressing its own integrity? Then how does that scene fit into a larger picture? Another way of looking at those questions is through character arcs. Every writer is different. When I do a more intensive workshop and I’m engaged in one-on-ones with writers, I try to find out exactly how their mind works. Do they work through structure? Do they think of structure as purely through character, or do they see it as a separate construct? Do they think of plot as separate from structure? These are words that serve as approximations of concepts about which we all think of a little differently. You have to, when you’re an instructor, try to look at what the writer is trying to do, and how it is different from your thought process — and inhabit that thought process, to be of the most use.
So if you’re not attracted to story dragons or lizards or the idea of targets, another way of thinking about structure (or plot if you prefer) is through the character arc, which is basically about where does the character start, and where does the character end up. That creates a different sense of scene, as well. Authority, the new book that I have out, is very interior to the character. As a result, even though it’s in third person, you have almost stream of consciousness in there. For my purposes, even though it’s a thriller — a spy thriller, more or less, with some weird elements — it was more useful to me to think of it as this personal journey of this character and be so tight in on him that I was thinking of the character arc, not thinking about the reveals or the mystery, or anything like that. Those things just slotted in naturally.
There was also a lot of improv involved in writing the scenes because I had only a very rough outline for these scenes: “I know that he’s going to have a meeting with so-and-so,” or, “This is going to probably happen here,” but I didn’t know what was going to actually occur in those scenes until I wrote them. Now, was this my process for the other novels in that series? No. From novel to novel, you shouldn’t restrict yourself in how you try to think about or conceptualize your approach, because each novel may be different. It may require a different process or a different approach. It’s helpful to realize that the story lizard that was so generous to you one year may be a stony sentinel impediment the next.
The other thing I just wanted to say briefly — and this image has less to do with my point than some of the others, but I like the image a lot — character arcs at the scene level, the decisions you make there, define character. They define character emphasis. The commitment to character that you have defines what you can get away with in a scene. You can sometimes get away with an action scene that doesn’t have much commitment to character because there seems to be something colorful going on, some movement generated that catches the eye. It’s less likely you can get away with that in a dinner conversation scene.
You also need to know your strengths and weaknesses. I tend not to dramatize dinner conversations because that’s not my strength. Many times, that will be the part that’s summary in a couple of sentences. There are other writers who don’t particularly like doing, for example, just battle sequences. You don’t necessarily have to commit to doing a type of scene that you just aren’t suited for, if you can find a way around it that isn’t just a work-around but serves to strengthen the story or novel in question.
Cutting Away from a Dramatic Scene
Digging down a bit further, here’s a more practical example. Airship disaster, as I call it: An actual example with three different ways of cutting away or not cutting away from the scene. Here, you’ve got a situation where, whether it’s a heated conversation or an airship blowing up, you have context at the beginning of the scene. You know where you are. You know who the characters are. You have some kind of focus. It’s embedded in the plot in a place where it makes sense for the reader.
Then you get to the point where there’s maximum drama or conflict. Either you follow the red example where you show everything through the downfall and the aftermath of that, or, at that point, you cut away, and you go right to the aftermath for the green version. Or, with the blue version, you go a little bit past the aftermath, past the point of crisis, so there’s a little bit of dying fall, and then you break away entirely, and you don’t really show this part of the action.
These approaches all create different effects. Again, you may not be comfortable with showing a certain kind of scene, or there may be so much drama up to this point that if you actually show what happens here, it’s going to be melodrama. The other thing about this approach is that maybe you want the reader to fill this in. Maybe it’s much more effective in the reader’s mind if you leave it at this point. You let them do the work here. Or, in showing the aftermath, it’s actually more emotionally effective for whatever reason than showing the moments after the explosion through to the aftermath.
One thing you may find in the middle of some of your scenes is that you can cut away and come back, and lose a paragraph or two, depending. I have a scene in Acceptance where two characters come to blows, and there’s something said that can’t be unsaid. It’s just anti-climax. I cut right there. I come back to a description of, basically, them both looking at out the window, and then come slowly back to both of them speaking again a little bit later. You get a sense of how they’ve become reconciled to what was said. Sometimes, you need to stay there and have all that messy stuff there. Sometimes you do need to cut away and that takes just as much skill.
This said, in the knowledge that the full through-line, where you leave nothing out, requires great, great skill of a different kind. An airship that blows up for too long with little reader commitment can be just as boring as something less dramatic.
A good example of how to be successful writing without leaving anything out is in Donna Tartt’s novel, The Goldfinch. She has, basically, a one-take scene set in a museum — where there’s a bomb explosion. She starts that scene from the aftermath, and she goes all the way through until the main character gets out of the museum. There are all kinds of places where she could have cut away — she could have done this, that, or the other — but instead, she shows you the entire thing.
It’s extremely masterful to hold a reader’s attention through all of that, because even situations that seem inherently dramatic can be rendered incredibly boring if you don’t do them the right way. To be honest, action scenes in particular are a great example of this, because if you’re not invested in whatever action’s going on, it’s just about as bad as a boring dinner conversation.
There’s also the issue of repetition across multiple scenes. George R.R. Martin mentions this with regard to his novels, about battle sequences, which he varies because he has so many. You can, again, substitute in any word for battle. “I have so many dinner conversations,” whatever. Sometimes he’ll cut away. Sometimes he will show the whole thing. Sometimes you just learn about what happened from a messenger galloping up.
It’s interesting, too, the acts of translation from film. A lot of films have sophisticated scenes that if studied properly are useful for fiction. Orson Welles has this great amazing scene in Chimes at Midnight — a battle scene that starts with a bird’s-eye view of this battle going on, and then slowly you zoom in and then zoom in again. You can’t even really tell what’s going on at the end, but in a good way. It’s just all these muddy boots and parts of bodies, and then Welles cuts away to show the aftermath from far above again…and there are all the soldiers, dead in the mud and it’s shocking. Highly effective. It’s something that would be effective in fiction to some degree, too, probably. I used a similar idea in a single-combat scene in one novel. I also think the writers for Game of Thrones on HBO used Chimes at Midnight for their “Battle of the Bastards” episode recently.
A Break Down of an Action Scene
What I wanted to do now is show you one particular scene and break it down, from the first novel in the Gormenghast Trilogy by Mervyn Peake. It’s kind of fantasy and kind of not. It’s set in this huge ancestral castle home with this dysfunctional family that’s been there for generations. You don’t have a really good idea of what’s in the outside world.
The lord of the manor has a cook down, basically, in the basement who is kind of a brute. Then he has a guy named Flay who’s his servant. Flay and Cook hate each other, because Flay sees Cook as a possible usurper, and Cook is hatching a plot to overthrow Flay or undermine his influence. They have been at it for a very long time by the time this scene that I’m going to go through occurs in the novel. They have a long, long history. It’s a fascinating scene, a chapter entitled “Blood at Midnight.”
This is an image we couldn’t put in Wonderbook. It demonstrates the thing that’s really fascinating about this scene: it is one long action sequence. It could have been several different scenes, and you may even still say it is, but it’s really one long scene. He doesn’t cut away during the entire thing. He uses extremely long sentences. He goes back and forth between the point of view of both characters, which makes perfect sense in the context of the scene, because they’re so linked at a psychic, psychological level. Some other writer would have taken a totally different approach with this. Think about if, say, Lawrence Block had written this scene. Pithy versus verdant styles give an immensely different feel to violence. One may speed it up and make us “only” look at the actions and the other may slow it down and give us time to look for other things.
As an exercise, I suggest you look at this slide, and read the specifics of the scene later. Write your own version and compare it to what he did, and see what choices you made that you think are actually better for your context, and which things he did that were better for his.
I’ll be honest, too, in admitting that this scene Peake wrote should not work. Here’s this 25-page action sequence with long, convoluted sentences, and yet you are on the edge of your seat the entire time.
So the scene starts out with this initial stalk. Flay knows that Cook is planning to kill him. Flay has been stalking or haunting Cook as Cook goes around his daily business. When he sees Cook go upstairs towards his quarters with a cleaver, he knows that it’s probably about time to try to not die. Flay has already — and this is another complication right off the bat –prepared a fake Flay in his bed. When Cook goes to kill Flay, he is unable to, because it’s a dummy. At that time in the novel, there’s a storm going on during a dark night. There’s a flash of lightning — chance operating to make things worse, really — and Cook actually sees Flay behind him, just as Flay was about to stab Cook. I know it looks a bit like Spy vs. Spy on the screen. Pretend it looks fancy.
Then there’s this chase sequence that’s up all of these endless stairs, which should be boring as hell. This is a place where, in some novels, you might cut. You might cut to the next scene, where they’re actually fighting or something, or fighting each other. But Peake uses this opportunity to recap the incredibly convoluted relationship they’ve had in a really fascinating and new way. You get a different view of both characters because of this — because you learn something new, you become more invested in the outcome. Then, by the time that they get to the confrontation, it’s in this interesting place, this Hall of Spiders, with all of these bizarre artifacts. You have a location, in addition to the specifics of character, that is specific and very interesting.
Next, you have “further complications” as Jarvis Cocker might say. Again, this looks more comic in the image than it is in the actual scene, but Cook gets a spider in his eye, basically. This slows him down a bit, which is useful, because in the natural unfolding sequence of events, Cook is much better suited for killing Flay than Flay is for killing Cook. The detail that Peake uses throughout acts as a form of slow motion, but not like in a bad De Palma film, so that you’re focusing more on the how and not the what. This focus emphasizes character motivation. He whirls around. He’s trying very hard to get to Flay. In the process of this, because of “old spider eye,” he smashes his blade right into the side of the wall. That gives Flay enough time to get away from him for a while. There’s a sequence we couldn’t put in here where there’s a little bit more back and forth right after that. Then finally, Flay wins out and kills Cook.
In addition to what I’ve noted, there are a few other things that the scene does that are very important to note. Flay has a very specific plan to lure Cook to the Hall of Spiders. You know this beforehand, so there’s been setup. The visualizations by Flay anticipating Cook’s attack create further tension. The two combatants are evenly matched. Flay is smarter, but Cook is a lot stronger. Neither party is a professional killer, so there’s no expectation of a quick resolution. In fact, you expect a bit of inefficiency.
The scene occurs towards the novel’s end. It functions as part of the climax. The reader is expecting that there will be something like this at the end of the book, just not in this form. Peake has the luxury of being able to go in depth because at this point the reader is not expecting something really fast and short, as they might toward the beginning, before being as invested. It’s the climax, for god’s sake — who wants their climax over in a haiku?
The stakes are high, as well. After this scene, this action has consequences on all the other characters going forward. You could say there is an extreme ripple effect and we also understand after this scene that things may get very dark indeed.
Yet despite that kind of darkness, Peake is always very precise about where the characters are standing and the scene never feels murky. This is a quality I admire greatly — you can also find it in Mark Helprin’s A Soldier of the Great War. As a writer, I always want to know where the light is in the room and how it’s striking the characters. Even if that description doesn’t make it to the end — maybe because the viewpoint character isn’t that observant — the echo of it there means that there’s a little bit more reality to the situation. Always knowing the basic blocking, like the stage directions, is very helpful in terms of creating specificity in your scenes.
Thus, the details of the confrontation are choreographed pretty darn well. Peake has a very clear idea ahead of time. Because he was an artist, I would imagine he did little sketches of all these different bits before he actually sat down to write, which is another important thing to think about when doing scenes: visualizations.
An Action Writing Exercise
The exercise, then, that I want you to do here is take your short story and isolate the scene with most action in it. I want you to make a list of every action that occurs in the scene and I want you to add some context, almost as if footnoting or layering in annotations, about the emotional resonance that is brought from prior scenes or events not dramatized in the story. If you don’t know, make something up. Why does this action have significance? What off the page leaves a ghost of its presence in your scene? You’ll likely wind up with melodrama from thinking of too much, and have to scale back, but that’s just part of the process.
And then I want you to draw diagrams of the characters and their setting, thinking hard about where they should be standing and where the light is and where it isn’t. Once you’ve done that, you should rewrite the scene keeping both the emotional context and the physical context in mind.
Two additional take-home exercises: For the first, re-cut a problem scene in your story by removing the first paragraph and the last paragraph, rewriting to include any context lost, deleting anything in the middle no longer relevant, and adding more suggested by the cuts. The second is to re-map the beats and progressions of the Gormenghast scene in the context of a dinner party. This requires you to translate action into less physical acts — the jabs and reposts of conversation, perhaps. I would also think about clues about where everyone’s coming from in that moment, like where they immediately came from, for example. If some guy’s coming immediately from a crappy day at work, that guy’s totally different than the day when he didn’t have to come from work at all like the weekend. What is their immediate emotional state? Then who knows each other already? What is the history between them? How’s that going to possibly create conversation, conflict, or whatever else. And who might have come to the dinner party with a premeditated agenda too.
The Different Decisions You’re Making When You’re Writing a Story
These other slides are just a kind of dying fall of different ways of saying the same thing we’ve been talking about. You’ve all seen those choose-your-own-adventures. One of the images in Wonderbook is of an evil eye that wants a body. It gives me an opportunity to show all the different decisions that you’re making when you’re writing a story. It’s all the different ways the story can go. That’s another thing to test once you finish the story: “Have I made it too easy? Is this scene too easy? Should these characters or the outcome have been different? What would the outcome be?” In a couple cases, the evil eye comes to naught. In the others, the evil eye finally finds his father, which is a key to getting by in the world.
Sometimes this could be as simple as changing the relationship between the characters, and it radically changes the scene. There was one workshop where this writer had a really interesting story, but there was no tension in it whatsoever. My wife Ann, who’s a great editor, said, “Well, what if, in this first scene, instead of saying” — this is a very simple example, but — “instead of saying” — this person’s coming to somebody asking for something — “What if, instead of the person saying yes, they say no, and then it turns out these two people have known each other for 20 years, and they have this whole history? There’s a very good reason why this person’s saying no.”
Out of that one change, the story suddenly had tension. A lot of the same elements were there, but they were completely re-contextualized, and the story worked. She didn’t actually have to change all that much. She had to change dialogue, but she had the scenes she needed. They just weren’t doing the right things originally. Again, it’s all about thinking about your character interactions, what their histories are, what they’re bringing to the table, and so on and so forth. This, on the slide, being a fairly whimsical and ridiculous example.
This image is more theoretical — probably the most abstract in Wonderbook. I’ve had some musicians say this actually looks a little bit like how they map music. Thinking about scenes on a more abstract level, you have all these little interactions, these beats, which are more or less the pulse of the story. Beats, more or less, are the action/reaction going on at that micro-level in the scene — and you usually hear them in the really crude context of commercial screenplays even though a beat is a very complex organism and plays a complicated role in the health of all kinds of fiction.
Then you have these progressions going on in the scene sometimes, where there’s a progression of a topic or a progression of a theme, or whatever else driving things in the background. Sometimes novels fail because the progressions are in the wrong order or composed of the wrong beats. Sometimes they fail because the novelist doesn’t recognize the autonomous heartbeat they’ve created and they don’t perpetuate that sonorous yet almost imperceptible progression throughout the entirety of their story.
In the middle of all this, you have what I would call “time intrusions,” although there’s also a note there that not all scenes include time intrusions. Sometimes they just include the neighbor’s dog. The point is that people have memory. People have memories. People do not think in sequence. They’re not always moving forward. I’m not talking about flashbacks, necessarily, but even during a conversation with someone, what are you thinking about? You’re not just thinking about the conversation you’re having with the person. Some of that may be relevant to the scene. Some of it may get into the scene.
Then there’s something called contaminations. Joyce Carol Oates did this beautifully in a story called “The Corn Maiden.” It starts out in the mother’s point of view. You think that the scene is really about her, but it’s actually about the police in the investigation and a member of the police. As the scene progresses, there’s this contamination or intrusion of this other subject matter.
Then it allows Oates at the very end to immediately go into a sequence that’s really from no one’s point of view at all. It’s just an overview of all the different actions that the police are taking at this time. It starts out as an interrogation of her by the police, and it ends as something completely different.
I’ve included “Science of Scenes” because I like the idea of pushing the boundaries of what can be articulated in an image. I wasn’t sure that it was completely accurate in what it was showing, but it’s something for people at least to argue about, about how scenes work. Even if I wind up being wrong, it has caused you to engage with an interesting idea, and one of you can confront me five years from now and tell me how stupid I was, which will be fun for both of us.
Writing Multiple Character POVs
Then what do you do if you have character points of view that, again, you have more than one character point of view, and you’re trying to figure out the emphasis there? It’s a very commercial fiction thing. But this applies to a lot of “literary” fiction too despite the emphasis of my diagram. The person you spend the most time with — that’s your main character. When you’re sequencing scenes, remember that. If you think that one of these characters that you’re showing the lives of in these different scenes from different points of view is your main character, then we have to fairly regularly come back to them. Usually, they should be the longer, more substantive scenes.
One quick easy way to accomplish that is to simply not break up scenes with that character in them. Show the whole scene, and have the other character scenes be ones where you cut away a little bit more. These are tongue-in-cheek examples from Chive Muscle’s cult classic Monster Island Bloody Hellfest, where they all come to this island, and there’s this thing that’s after them, and two different ways of cutting. One, obviously, is emphasizing a particular character. One’s a more ensemble novel. You can see, though, that you’re not looking at something that’s that precise. It shouldn’t be this cookie-cutter thing where this character’s scenes are all going to be 20 pages or 5,000 words, or whatever else, but that you have a regular rhythm that you establish.
Does it apply to all novels? No. Adichie in Americanah doesn’t do this. She does this really wonderful thing where she sticks mostly with the female lead, and then she does these asides to the male character, who, at one point, is in the UK. She’s not necessarily consistent because it would be too mechanical, but there’s a weird consistency to her inconsistency in how she goes back between the points of view. If you go to the Wonderbook website, I have a breakdown, a diagram breakdown, of that novel, which I think is a really interesting one, because it looks so unstructured, in a way, but it’s really cleverly structured, if you really take a close look at it.
Now, have these diagrams and this lecture covered every possible way you can think about scenes? No — that would be impossible. But what I hope this lecture has done is given you enough general and specific examples that the information will mix with your own approach and imagination in a synergistic and organic way. You will need to experiment in mechanical ways too — plugging in these ideas to existing fiction you’ve written. But at the end of the day, you’re trying to get to the point where more and more technique becomes muscle memory. That’s when you begin to talk about the “art of fiction,” but it comes to you first through absorbing craft. You can do that without instruction — without taking classes but just studying fiction you admire — but sometimes a lecture like this one can shorten the time span it takes to internalize these topics. Thanks for your engagement and your attention.