Sunday, September 09, 2007

[Process] - The Scene

For the last two and a half years, I've kept notes about what I learned while writing the Limit. Now I'm creating posts that will cover each phase of my script-writing process. They'll be added to as I trawl through the blog, and copy material from previous entries into this one. I'm also putting links up on the sidebar.

This entry is about the process of writing scenes.

-- -- --

I’m also getting a lot of experience treating drama scenes as action scenes – a little dodge I wanted to try out because action is easy for me to write and visualise, whereas drama and emotions, quieter stuff, is opaque.


The first 3 days of the rough redraft have gone fast and slow. Fast: I’ve already broken all the scenes, figured out what I want to do with them. Slow: even though it’s just transcribing for the most part, putting scenes on paper for the first time, I’m still tweaking them to make them play.


I'm treating this primarily dialogue-driven scene as an action scene.

You see, action scenes are easy to write. I'm not sure why I find that yet - but I build a clear visual image in my mind of what happens, and I find it easy to see where the gaps are and edit accordingly. I also find it easy to create and maintain the point of view (sympathy for the hero) in an action scene, and to increase tension and put the people I like under stress.*

Drama scenes lack that clarity for me. I feel they should build tension and maintain interest in the same way that action scenes do, but because the (opposing force?) is non-physical, that clear visual image is harder to create in my head.

How to represent an intangible (not physical or visual) form of jeopardy. What I'm trying is:

Clearly determine the main characters motivations.
Break the scene down into acts and turning points.
Visualise how the characters will move within those acts and turning points.
Use those movements to inspire deeper tensions and oppositions.

Boy, thinking along these lines may start me questioning exactly what a scene is.

* For me to write it effectively, an action scene has a person in jeopardy and something putting them in jeopardy. In the case of The Limit, that something is usually a person. Typically the person IN jeopardy as the hero or the person we have sympathy with in the scene. Because The Limit is a vigilante thriller, the person causing pain is the hero. So I'm constantly finding I have to tweak the scene to keep the hero sympthetic.


Finished the early section of The Limit. Those first 4 pages took 2 days to edit, I think because it’s very first-drafty (IOW, not worth paying $13 to see)*. But the second four pages took 15 minutes. So, out of drudgery, building momentum.


Had an insight while rewriting an interrogation scene. I was giving all the dialogue to the main character. Doing that forced her through sudden changes in her motivation and emotions. For instance, Trace was disbelieving everything Forster said, and then she'd clarify a point he’d made. An odd shift from hatred to logical.

So the Insight was: I’d use a third person in the scene to ask all of Trace's rational questions. And extrapolating from that: put as many players in a scene as there are positions to adopt.

[theory] Characters can shift positions, just not constantly. I mean, one big shift in position per scene is probably good. When I put more than 3 or 4 into a scene, I get actors complaining that they can't see where their character’s head is at. Of course, the longer the scene the more changes a character could go through – see Richard’s seduction of Lady Anne in King Richard the Third, Act 1, Scene 2.[/theory]


[from Freaks and Geeks] One line from the directors' commentary I thought was true is that on a TV show, ideally you're going to be working with the actors you've cast for 5 or more years. So you adjust the script towards them. You want to have fun with the actors, explore the little moments in their character relationships and most of all make them feel comfortable so they can bring a bit more realism to their roles.


What’s the most important line of a scene?
The last line.
It raises a question. Creates a tension we want released. Primes us for what we expect to see in the next scene.


Yesterday's scene was too on-the-nose. So I've brainstormed up a lot of ways to make it work.


A lot of money gets spent on famous actors and CG. Flashy things are supposed to keep our interest, our eyes glued to the screen (a pretty disgusting visual now I think about it). However, with a no-budget movie or Season 1 of a normal budget TV series, there’ll be no cash to spend on those things. In this case, emotions are your special effect.

Emotions can grip the audience. They can be complex and spectacular. You can find emotions that haven’t been tackled before. Best of all, they are cheap. And if they’re cool enough, maybe you’ll attract the funding to get those famous actors and flashy CG effects.

Some random thoughts on emotion

The first principle in the first book on screenwriting I ever read was that the primary goal of any screenplay is to elicit emotion from the audience.[1]

Scripts deal in emotions and motivations. As script-writers, those are our 2 basic tools. We can make characters behave in plausible and fascinating ways.
We can aim to make the reader (and hopefully the eventual viewer) feel a certain way.

An well-drawn action setpiece can illustrate these 2 things working together just as effectively as a moving on-screen declaration of love.

James Cameron says, “Audiences don’t think in scenes. They think in a continuously dynamic and evolving force field of emotions and ideas.” Sometimes when I’m considering the overall script, I don’t think in terms of plot or character; I just go through it feeling what the audience will feel at every stage. Then I ask myself, “Is this a good journey to go on?”

“Do we buy it?” was a question often asked in the latter stages of writing lovebites. Do we believe in what we’re reading. Maybe you don’t buy it because the script hasn’t made you care enough … or maybe you don’t buy it because you care so much that you think what the character’s doing is (a) wrong, (b) against type or (c) just something you don’t want to see them do …

I think it’s important to know the emotion you’re trying to produce with each scene. That’ll go on the checklist.


While talking to DBS last week, he mentioned that the biggest thing he’d learned from writing on Insiders’ Guide was to ‘layer scenes’. To have at least 2 (but more like 3 or 4) things going on. Part of this is to do with subtext, part of it’s about having stuff going on in the background and still more is about keeping the scene alive, filled with energy and keeping the viewer engaged.

The idea is to make scenes multi-dimensional as well.

So now we’ve got the starts of a checklist:

  • What’s the scene about: Stakes and Conflicts.
  • What’s the scene really about: Subtext.
  • What else is going on: background stories – and secondary characters.
  • Off-screen stories: (see the sound design commentary track on the Se7en DVD for many examples of this).
  • How should the scene make us feel: Emotion

I’m really happy with where this in-the-car confrontation scene is going now. After doing all the prep work of looking at their motivations, I think I may have found a new technique to breathe life, shove energy into this (any?) scene.

The idea is to think of each character’s input into the scene as a Bang (I’ll provide a link to a fuller definition of this later), something that must provoke a reaction from at least one of the other characters. Obviously this is simplest when there’s only 2 people around. What I’m doing is drawing a step diagram down the page, taking it very methodically and asking “If Forster does this, what is Peter’s reaction? Okay, if Peter reacts like that, what would Forster say?”

Seems obvious. Except that the reactions have to be big and personalised and surprising. My intention is to keep all the characters in a scene off-balance. Force them to respond to things they’re not expecting. Trap them in a rapidly evolving situation that’s at least partly out of their control. Kind of like life.

So at a scene level, what I’m doing is:
1. Defining stakes and conflicts.
2. Building up motivations at a beat by beat level.
3. Reaching a point where those feel artificial, where I’m bored with them.
4. Building a Step Diagram and keeping the idea of Bangs in mind.


A. Disappointed because I've wasted a lot of time today.

B. Satisfied because I've kept asking myself what this scene is about until I realised it's about 'Peter facing the consequences of becoming a vigilante'. I think I'm ready to write and finish this thing.


Reached the 50% mark on Saturday. Took a couple of days off and now I’m back working on the scene where we turn Michelle into the hero. It’s good; it’s action again, so all I need to do to write a good scene is a) keep coming up with ways to block her from escaping and b) figure out how to make the situation worse.

As ever, I know the scene is good when I can see it playing like a film in my head.


The longer I write a scene, the more chance I have of fucking up what the scene’s about.

Action obeys Harold Ramis’ rule for writing comedy: it should take as long to read it as it would to see it.


Just talked to Jenni’s brother-in-law, Jason, who’s a school teacher. He gave me lots of insights into how fights between students work in primary schools, plus a neat idea about how to personalize this scene I’m working*.

What’s cool is that this scene – which I originally thought would just be a pause in the story - is starting to emerge as a thematic representation of the script as a whole.


So I’m heading back from Gino’s tomorrow. There’ll be a period of just resettling into my life and then a day where I’m working on my 24 game for this month’s competition, and then maybe a day off.

At the moment The Limit’s going really well – I’ve worked through the big rewrites at the start of the movie & now I’m at the point where I can get through vast swathes in a day. However, last time I took a break I kinda … forgot how to write.

So, for my mind-prodding, here’s some thoughts (I may have written something like this already) …

To write a scene, I – repeat, ‘I’ – need to set some stakes (What’s the question this scene’s going to answer? What do we care about?).

Then know what’s the conflict (If the question has 2 possible answers, then I need 2 characters/forces fighting or advocating for each side). As soon as one side’s one, it’s time to wrap up the scene. If at all possible, the sides of the conflict have something to do with the thematic conflict at the heart of the story. In The Limit, that’s Law vs Vigilantism (vs. Criminality). It’s all very Story by Bob McKee (c.f. Adapatation by Charlie Kaufmann).

What does each character want? These motivations need to naturally come out of each character’s previous scene.

Next I either brainstorm 20 things that could happen in the scene – issues, cool moments, motivations, lines of dialogue, things I want to see, random oddball ideas – in no particular order. It’s just stuff to inspire me.

Then I reorder that stuff into rough chronological order.

Otherwise, if I’ve got a clear idea of where the scene’s going I brainstorm a starting point, and then brainstorm again – what’s the worst they could do, to trigger a response from the other person in the scene? I keep swapping through each character’s perspective, trying to continually increase the tension in the scene.

I brainstorm 20 things because I read a book that recommended doing that.

I am a drone.

Seriously, I’ve always brainstormed multiple options for moments in my script. Off my own back though, I used to only devise about 7 different options for things – like punchlines when I was writing eps of lovebites. With 7, I found I came up with something that worked.

But with 20, I start getting oddball and insightful ideas towards the end of the process. If I don’t, I take a break and then keep going. I want to get the 'right' idea by the end of this process. Not some idealised 'perfect idea' - just have a decent range of good options to choose from, so I can move on.

Finally, I need to know the resolution to the scene. That means at some point, there needs to be a turning point in the scene where things head towards that resolution. And I need to bear in mind that that has an affect on the person who didn’t get their way.


A lot of the time, I was watching the film going "Where's the conflict in this scene?" And I realise that my recent experiences with Primetime Adventures have been subtly educating me in this screen-writing tool. Creating conflicts and having to decide which ones are meaningful up to 15 times a game is a really effective way of building up your chops.


Not so much procrastination in the re-starting of the writing this time. Maybe because I’m already three-quarters of the way through the script – because it’s in media res, I’m immediately interested, whereas starting from the beginning involves slowly relearning why I like the characters and then building up interesting situations for them.

Anyway, I took a crack at the confrontation scene between Tracy and Forster on Tuesday. By about halfway through Draft 1, I stopped. Here’s what I wrote about that:

Fuck, I really want to mine the subtext – but I think I have to write the scene out first, then figure out what it’s about – and how that adds to the About of the film (‘2 dads vying for the love of their son’), and then simplify the scene down. I’m talking about writing and editing, really … and then re-reading the whole thing in sequence, so I can see how it all fits together. So … no need to panic about ‘getting it right’.

All I was doing here was taking the pressure off myself - to remind myself that while I was doing the best I could with the scene, there were going to be plenty more opportunities before the script was finished to judge how it worked.

So I backed up, started a Draft 2, and stalled again.

Time to take a break, think through it again. In the middle of doing dishes, I realised what each character wanted, that it was actually a pretty simple conflict (and that I’d been circling around articulating it that simply for about 2 years). Tracy wants Forster to live and face justice, while Forster wants to die, in a very specific way.

I also realised that I wanted to create a connection between the 2 of them, to show that they had common ground, having been through the same stuff. It’s a technique I admired in Lost, in the Sayid/AnaLucia conversations after Shannon’s death.

So, I worked yesterday, mulled over the scene and came back to it this morning after a bit of internet procrastination.

First things first, what’s at stake? Well, I know that the lead character will live, and I know that she’s going to ‘get’ Forster. What’s at stake is how she’ll do it: will she use the law and reason, or anger and brute force. It help, having already written the brute force scene that I know I can make that plausible.

I wrote down the fundamental beats I want to see in the scene, and ID’d the next point to get to – which is Tracy’s realisation that Forster wants to die. I decided that Tracy had to be the one who realised, because (a) it makes her active, and (b) it shows her doing some detecting.

The dialogue came easily to start with, because the characters’ Wants are so diametrically opposed. I was handwriting everything, and if I couldn’t get an exact phrasing, then I just jotted down the idea behind the dialogue – to work it up later.

I came across a couple of things repeatedly:
1. An impression that some lines were either clich├ęd, or contained a repetitive subtext. (Have I talked before about how (for me) subtext has to be deliberately constructed and hammered into a scene? That was certainly the case here, with Tracy’s concern for her Dad.) Anyway, I decided to leave judging all that to the next readthrough.

2. If the audience knows some information is coming, withhold it. Create tension. It’s very natural to pop that stuff too early.

Toward the end of the scene's second act, Tracy learns some bad news. It's a Bang - I'm fascinated in how she'll react because I have no idea what she'll do. So I start brainstorming 20 ideas. At #8, I get to one I like - that she confronts the situation. I'm able to write out two or three more lines between Tracy and Forster, and then the dialogue dries up. After half an hour, I admit that to myself & back up, brainstorm some more ideas and adopt a subtler, softer approach. The scene flows pretty easily from that point to its end.

I wasn't sure exactly how Tracy would subdue Forster, so when I reached that moment I wrote from the heart. The end result is probably way over-long, but I was really INTO it while writing.

Anyway, I’ve roughed out the scene, I’m ready to move on now. Tracy has a real hero moment; I understand the characters better; their conversation NEEDED to happen; and it’s probably made the remainder of the film about 70% to 240% more interesting than it was before.


Sometimes the writing of a scene just flows – every line clicks into place; it’s easy to visualise the action – and that is awesome.

Other times, not so easy. That’s when I have to get a bit mechanical. I roughly outline the beats that I know will be in a scene, and then I B20 (brainstorm 20 options) for each character in the scene – for their overall motivation, all their reactions, and every line of dialogue.

When I do that, I’m looking to find truth about the characters, insight into them, and either originality or authenticity.

Every time I go through the B20 process, I tend to got through the same emotions and reactions. What I’m trying to do here is describe that pattern and then (hopefully) isolate some key questions to ask, that’ll speed up the whole thing up.

  1. First off, the obvious lines are the ones I write down.
  2. Then variations (sometimes very slight) on those obvious lines.
  3. Random lines, as they occur to me.
  4. Come up with a few arbitrary lines, that don’t really fit with what the beat’s trying to do.
  5. Write a line that hits the mark. Experience satisfaction, then slack off / consider giving up or settling. This (and every point here) can happen multiple times during a B20.
  6. Spell out the subtext behind the line.
  7. Play around with that.
  8. Try another subtext. Every subtext I find is a different area to explore and mine for possibilities.
  9. Realise that the line doesn’t exist in isolation and link it back to the previous one(s), so it flows.
  10. Imagine the actor who’s saying the line.
  11. I get exhausted towards the end, and struggle to come up with lines.
  12. That’s when I re-read it all and jot down any lines that occur to me from reading all the others.
  13. Towards the end, I almost always get a fresh insight (or two) into what’s really going on.
  14. And I usually write down some crazy, usually rude or sociopathic stuff just to get the thing finished off.

So, what can I distil from that?

Before I begin:
Bear in mind that the line doesn’t exist in isolation. It needs to flow from what has come before.
Imagine the actor who’s saying the line.

  1. First off, write down the obvious lines.
  2. Then spell out the subtext behind the beat, and play around with that.
  3. Once those lines dry up, try another subtext. Feel free to write down random lines, as they occur to me.
  4. Then re-read it all. Jot down any lines that occur to me from that.
  5. I always get a fresh insight (or two) towards the end.
  6. Finish off with some arbitrary stuff.

Then I go through the list of 20 options, circling the ones that appeal to me. Create a separate list of those options and choose the one that most appeals. The point is not to get it perfect; it’s to get it done. This is the point to rust my instinct and save ‘perfection’ for the rewrite, once I see how the line plays in the context of the whole show.

If, at any point, the scene just completely tries up for me, I use the Brian Johnson trick of tracking back a few lines or a page and seeing where it all started to go wrong. It’s usually quite obvious in hindsight.

I don’t get the scenes completely right using this process, but so far every time I’ve re-read one it’s been obvious where it works and where it doesn’t.


Man, this frickin' script. I had a great start to the rewrite (5 pages in one day, which is fantastic for my first day). Then I hit this big father-son scene at page 14 and get bogged down in it for 2 and a half weeks until I figure out what it's really about (clue: it's personal).

Anyway, once I figured out the heart of the scene, it flowed easily & the script has kept going at a nice rate.


The more important I think a scene is, the longer it seems to take me to write. Thinking of something as "important" makes me freak out, because I need to get it "right".

But at this stage of the script, every scene should be important. Every scene needs to contribute. Every moment, too.

That means I'm going to have to develop some way off de-freaking-out. Being a full-time writer would be good too (more dedicated time to solve problems and get into a groove). Unfortunately, I may have to let that ambition go for a while, and content myself writing on the bus.

(Example: It's taken me five days to work on a single line of the script. It's an important line -- it has to let us know what Peter's worst fear is, and hint at some of the oddness that is to come.)


With our recent changes this whole sequence has to be reconsidered. 50% of what’s there has to be scrapped. It has to be rebuilt at a motivational level, starting with what Peter and Forster want. Then moving between their heads, asking “What do I think he should do if I were him?”. The goal is to make each reaction something that boggles the other character.


A. Disappointed because I've wasted a lot of time today.

B. Satisfied because I've kept asking myself what this scene is about until I realised it's about 'Peter facing the consequences of becoming a vigilante'. I think I'm ready to write and finish this thing.


Post a Comment