Wednesday, July 26, 2006

The Cameron Structure

I'm going to reveal the one secret bit of useful screenwriting knowledge that I think I posssess: the (James) Cameron Structure.

You've probably heard of various 8-sequence, or 12 sequence, or whatever structures, I assume this is like those books, but really short.

My first screenplay was an action-horror piece and I wanted to write something that had pace and energy, so I decided to take a more careful look at Terminator 2, which I remembered felt like non-stop action.

Surprisingly, T2 wasn't non-stop action at all. There's a humorous teaser, then a build as all the characters come together at the mall, then a big fight and chase scene.

Then more talking and setup bringing us to the set piece where they rescue Sarah Conner from the mental institution.

Finally more prep and character development leading into the final, very long set-piece where they blow up Cyberdyne and face down the T1000 at the factory.

That's just three set pieces, but it feels like you're constantly moving and still gives you plenty of room to fit in character development. The basic structure looks like the following:

The Cameron Structure
  • First Act: Intro characters and their issues, build towards -
  • First Action Sequence: Try to get to this fairly early, before page 30 usually. Doesn't need to be long (T2's is only five minutes).
  • Second Act, Part One: You'll probably need to do some explaining of what's happening in the film here, but this is also a good time to have characters relax a little and just show us who they are -- this means they can do things unrelated to the plot or their arc.
  • Middle Action Sequence: In T2 (and Mission:Impossible) it was a bit longer, but you might make it short if you've chosen a long first action sequence.
  • Second Act, Part Two: Here you're mainly setting everything up for the third act, because...
  • Third Act/Final Action: The third act is almost entirely an action set piece. Because of this, you'll need to somehow slip any unfinished character arcs into the action (watch Neo learn to believe when he thinks he can stop the helicopter from going over the edge of the building), or into a short interlude (the part where Claire gets killed in the baggage car in Mission: Impossible).

This sort of structure has also worked well for me in horror films, which I think supports my theory that horror films are just action films with weaker heroes and stronger villains.

Mission: Impossible clearly has the Cameron Structure -- the only difference being that the first action sequence occurs a bit earlier (but is also tamer, so the writer could fit some nice banter in). I think Aliens had the Cameron Structure too, but since I've only got the extended version on DVD, I'm not sure.

The key feature of this structure for me is that the story centers around three largish set pieces, roughly distributed as above, and this somehow makes it feel like you've got non-stop action, even though there's quite a bit of time just devoted to characters.

I'll typically work off the structure during the outlining process, since it gives me a place to start, then be willing to change stuff if cool ideas present themselves.

Undoubtedly, there are an infinite number of other structures that work as well -- but one interesting feature of infinity is that even though there an infinite number of ways to do things right and an infinite number of ways to do things wrong, discovering a way to do things right can still be much more valuable because the ratio of right ways to wrong ways is so horribly low.

Postscript: I call it the Cameron Structure because I learned it by watching James Cameron movies. There's a pretty decent chance it's been around for a lot longer so if anyone can identify earlier films with this structure I'd be interested to hear them.

Tuesday, July 18, 2006

Stross on Gygax, VR, and World Domination

Charles Stross has a post titled Gary Gygax, World Dictator up -- here's part of it:

Sad to say, the political landscape of the early to mid 21st century has already been designed -- by Gary Gygax, inventor of Dungeons and Dragons.

Gary didn't realize it (D&D predates personal computing) but his somewhat addictive game transferred onto computers quite early (see also: Nethack). And then gamers demanded -- and got, as graphics horsepower arrived -- graphical versions of same. And then multi-user graphical versions of same. And then the likes of World of Warcraft, with over a million users, auction houses, the whole spectrum of social interaction, and so on.

Which leads me to the key insight that: our first commercially viable multi-user virtual reality environments have been designed (and implicitly legislated) to emulate pencil-and-paper high fantasy role playing games.

Sure, Second Life shows a pattern for Ludic environments that is non-RPG based, more user-directed -- after the pattern of LambdaMOO and similar -- but again, the LambdaMOO experiment fell out of dissatisfaction with the fantasy RPG limits that the earlier MUDs imposed on social interaction, and the MUDs were basically networked multiuser implementations of the Colossal Cave Adventure and friends, which all came back to Gary Gygax.

There's no bloody escaping it. The gamers have given rise to a monster that is ultimately going to embrace and extend the web, to the same extent that TV subsumed and replaced motion pictures. (The web will still be there -- some things are intrinsically easier to do using a two dimensional user interface and a page-based metaphor -- but the VR/AR systems will be more visible.)

I'm not sure we've reached the equivalent of Netscape's 1.0 release. New interaction mechanisms are going to come along, especially once the VR experience moves away from the desktop computer paradigm and goes mobile, using PDAs or smartphones, head-up displays, and ubiquitous location services (and speaking of the latter, it is reported that the Galileo PRN codes have been cracked). But VR will be the first medium where the primary path to commercialization will be through game-play.


If you subscribe to the theory that the structure of a language also structures the ways of thinking of its native speakers, then think about what effect the basis of the new hyper-reality will have on the future.

Monday, July 17, 2006

Learning By Seeing What Others Do Wrong

Occasionally I've seen (and even made) this sort of comment in regards to reading motley assortments of screenplays and films -- the premise being that in a film such as Voodoo Tailz you'll at least learn some ways not to make a film, or in reading the worst screenplays for a contest or company, you'll learn how not to write a screenplay.

I no longer buy into that theory. There are an infinite number of ways to write a bad story. Eliminating one of them still leaves you with an infinite number.

What you need to learn is ways to write a good story (or frame a nice shot). And that's why you should read even the bad screenplays or watch even the bad movies.

For instance, The Grudge, while financially successful has some serious story problems. However, it also has some amazing visual moments: a ghost floating up in the corner of a room while her impossibly long hair swarms down the walls. That sort of liquid image is something I've seen replicated in other stuff since then.

Tarantino has made a living off of seeing the terrific moments in less successful films then combining and using them in new ways.

Or the wonderfully budget-conscious Necropolis: Awakened, previously mentioned here. I'm still enamored with its idea that the zombie-king hires some hit men to come and take care of his pesky-protagonist hero-killing-all-the-zombies problem. That's just remarkably common sensical -- though, were I to steal it, I'd instead have it be the townsfolk ordering out for mob hitmen to come in and take care of a vampire, or similar monster -- something that they couldn't convince the government to take seriously but bad enough that they couldn't handle it themselves.

The reason this works is that everyone's got a few good ideas and if they write their one story, make their one film, or sing their one song, that's where those ideas end up -- well, except for Voodoo Tailz.

Saturday, July 15, 2006

Success Has Made A Failure Of Our Horror

One thing that I can't quite figure out, is that in comic books (Alan Moore, Neil Gaiman) and roleplaying games (John Tynes, Dennis Detwiller, Greg Stolze), the horror genres are where you go to see all the smart writing, but in films, while there's a few sharp pieces, there's also a lot of sloppy trash.

Moreover, it's not like filmmakers couldn't steal these guys away from their current job for the length of time it takes to make a film -- RPGs pay practically nothing and film budgets are such that even successful comic writers could be lured.

I have a theory on this, though. The success of the horror genre in film is what has lowered its average quality.

Eventually every small production company that's starting out figures that they ought to make a quick horror film or two to help drum up the cash for their real movies. So you have some people who actually don't like horror making decisions about how horror should be made.

Once they set their mind to it they'll convince themselves that they're horror fans and understand the horror genre because they liked Aliens and The Exorcist. I liked When Harry Met Sally, but I can guarantee you that I do not understand the romantic comedy genre.

In some cases the project will take the form of "we want to make sure that everyone knows that we're just having fun here, not taking it seriously, and you should look out for our real stuff down the line". That often results in campy horror, which in the 80s could make decent money, but today doesn't even warrant a theatrical distribution. And it sucked in the 80s too.

Otherwise, it'll just result in a series of mediocre decisions -- often partially saved nowadays by higher budgets, thus allowing talented cinematographers and production designers to at least make the visuals powerful.

Despite this, even mediocre horror films can turn a tidy profit. And this is another reason the average quality of horror films is low.

If you've put together a $1-3 million budget drama about the trials of family life or the emptiness of the human experience, it has to be fantastic to make it onto more than three screens.

But you do the same with a horror film, and as long as it doesn't actively stink, they'll toss it out onto at least a few hundred or a thousand screens and make some cash. Which, of course, drives more people who have no interest in horror to try their hand at making horror movies.

And the final link in the chain is me, and people like me. Horror fans desperate for the kinds of stories that they won't show on television, and who enjoy seeing a film in the theater -- so we're willing to buy a ticket to even the mediocre ones because they're all we got, and even a mid-range film usually includes a few excellent moments (which is often what makes it a mid-range film).

Friday, July 07, 2006

Comics, Hamster Poop, and Art

J'aime Wells offers a sharp, common sense introduction to some issues in the philosophy of art in this post. She knows something about this stuff, having worked with Professor Peter Kivy, one of the leading scholars in this field. Since people don't click on links, I'm quoting it all here:


Chapter Seven aims to define art, and anyone with a course in the philosophy of art is probably cringing just hearing that. You can see what he is thinking. One thing he wants to do in this book is to salvage the reputation of comics as an art form. So naturally, he thinks that he will define "art" and then show that comics fits within that definition. The problem being, of course, that hordes of philosophers of art have written hordes of books about their attempts to define art, and after Arthur Danto shot them all down, everyone pretty much gave up and started thinking about other questions for a while.

McCloud writes, "Art, as I see it, is any human activity which doesn't grow out of either of our species' two basic instincts: survival and reproduction." (Boldface is to approximate the look of the lettering in that panel.)

This is so clearly a hopeless definition, that I won't fill space listing counterexamples. Instead I'll take a quick break to go artistically clean out the hamster cages. Excuse me.

OK, I'm back. McCloud doesn't really mean what he just said anyway. He is haunted by the ghosts of aesthetic theories past. For example, R.G. Collingwood argued that (real) art was expression of emotion. Dwight MacDonald argued that (real) art was distinctive expression of the individual. There is a whole family of these theories, which McCloud seems to want to align himself with. Some of his Chapter Seven examples include a caveman blowing a raspberry, a bicycle messenger riding with "personal style," and an average person signing his name. These things are said to "have an element of art." That element is self-expression. McCloud also says that he does not believe in a dichotomy between Art and Not Art. So I think what he envisions is a continuum, ranging from acts with a small element of self expression (perhaps the signature) on one end, to acts that are almost all self expression on the other end, with no clear dividing line defining Art. This doesn't, of course, require the ridiculous definition of art that includes hamster cage cleaning. A theory with a continuum still allows one to say that one end of the line is definitely NOT art, even if one declines to say exactly where the cut-off is. There can be clear cases and unclear ones. There isn't THAT much self expression in cage cleaning, so it would fall on the low end of the spectrum, and he'd be free to say that my activity there was neither survival/reproduction-related NOR artistic.

McCloud's goal here is clearly to get the broadest possible definition of art, to cut off the rather boring discussions of "That's not art! Oh yes it is!" I have a lot of sympathy for that motivation. I'd just as soon include anything that wants to be art, myself. Unfortunately, the "art as expression" theories are completely barking up the wrong tree. Whatever the continuum is, it can't be about more vs. less self-expression.

For one thing, consider modernist or conceptual art. Some of it may involve emotive expression, but many pieces are purely intellectual, and not emotional at all. You could say that intellectual "self-expression" counts as art too, but then you'd need some way to separate conceptual art from the papers in academic journals. It seems highly implausible to count academic papers as art. But whatever separates them from conceptual art is the very thing that makes art art. No kind of self-expression can make that separation.

McCloud mixes this discussion up with considerations about the motives of the artist, which is another red herring. It's not the same red herring as in the above paragraph: you could make intellectual, nonemotive art for any motive. Lots of people believe that "pure" art has to be made from "pure" motives, like individual expression, and not for "dirty" motives like making money or getting famous. But this obsession with artistic purity only dates from the nineteenth century. Shakespeare wasn't concerned about it - he was aiming to please his audience, with jokes for the groundlings and flattery for the royalty. (Look at Henry V!) Velazquez made a living painting portraits of the royal family of Spain. Mozart composed on commission all the time. The notion of "selling out" would have made little sense to them. On the other hand, some of the crappiest teenage poetry is the most pure and sincere self expression. So "expression" neither defines art as opposed to non-art, nor defines good as opposed to bad art. Lots of art, good and bad, is characterized by self-expression. Some is not. That's all.





Sunday, July 02, 2006

Bryan Singer Teaches a Class on Doing Sequels Right

Caught Superman Returns last week and I didn't know what to expect. But as soon as the retro title sequence fired up I had a good idea, and thoroughly enjoyed the remainng 2.5 hours of it.

This is old time cinema, with a blend of action, romance, and humor -- and that makes sense because the film fit perfectly in as the third movie in the trilogy that should have been, instead of the trilogy that we actually got. When they sell box sets of Superman DVDs in the future they should just include this one and drop the original 3 and 4.

One impressive thing about this is the suppression of ego required to make a sequel that both builds on, yet is continuous with the themes and stories from the original films. This requires the writers and director to admit that someone else actually did something right, that someone else had a good idea and maybe I should shape my work around that -- not at all a common thing to see anywhere, let alone Hollywood.

James Cameron did a great job of this in Aliens. He took the Ripley character and used the movie to explore her in depth. Why was she such a cold-hearted hard-ass in the first film? Because she needed to keep distance from people. But her dealings with the marines, and particularly Newt, breaks through that.

If the people who made Alien 3 and Terminator 3 understood what Cameron and Singer know, I'd own two more DVDs.

To wrap up, one more thing I appreciated about Superman Returns was how immersed it was in Superman-as-boy-scout (to use the Dark Knight's term): helping cats out of trees, telling people that flying is still the safest way to travel. In a world of Jack Bauers, where the people in charge of America are saying that torture's actually got some things going for it, where cynicism seems to rule the day, it's nice to see a film where the hero aspires to a higher ideal -- despite the fact that it makes him a goofball.

Another point I thought interesting: when Clark Kent expresses those ideals and acts that way in the film, people see him as a quaint nerd. But when Superman expresses them, they're ideals to respected. I think the subtext here is that we should respect the Clark Kents of the world a little more.