Friday 22 March 2019

Screenwriting manuals and three-act structures

When it comes to screenwriting manuals the rule "if you've read one, you've read them all" is close to being true, especially since 1979 when Syd Field's Screenplay: The Foundations of Screenwriting was first published. (And later revised and updated regularly until 1995.) There had been manuals before Field but he opened the floodgates, as it coincided with general changes in filmmaking praxis in Hollywood (to which most of these manuals are aimed) that made it easier for outsiders to sell their scripts. Almost all screenwriting manuals that have been published after Field are saying more or less the same as Field does in his book. This includes such books as David Trottier's The Screenwriter's Bible: a Complete Guide to Writing, Formatting and Selling Your Script (1994, last edition 2014); Robert McKee's Story: Substance, Structure, Style and the Principles of Screenwriting (1998); Neill D. Hicks's Screenwriting 101: The Essential Craft of Feature Film Writing (1999); Michael Tierno's Aristotle's Poetics for Screenwriters (2002); David Howard's How to Build a Great Screenplay (2004) and Writing Movies: a Practical Guide from New York's Acclaimed Creative Writing School (2008). Some are better and some are worse than Field in terms of how they say what they say, but they are close to interchangeable with regards to what they say. The films discussed are the same, the advice given is the same, the terminology likewise and almost all adhere to the basic three-act structure. All name-check Aristotle and Joseph Campbell.

Field is often criticised for the three-act structure which he popularised and made almost ubiquitous. It is sometimes claimed he invented it but that is not the case, it had been around for decades, if not centuries. William Archer discussed it in his book Play-Writing (1912) for what might be the first time in the modern era. (See end note.) Some think the three-act structure might be relevant for mainstream blockbusters but not other kinds of films, or that the structure is even the opposite of art, which is supposedly free from all rules and structures. But this is not Field's view.

Field describes the three acts for a two-hour film like this: act 1 should be the first 30 minutes and end with a turning point, act 2 is the following 60 minutes and also ends with a turning point and act 3 is the last 30 minutes. According to him, this is the basis (or, in his word, "paradigm") for all forms of storytelling, not just mainstream blockbusters. In his book he speaks of how Jean Renoir mentored him, he speaks of his friendship with Sam Peckinpah, and how they both inspired him, and he is as happy using the films of Michelangelo Antonioni as examples for how to create characters and story as he is of using Rocky (John G. Avildsen 1976) or The Usual Suspects (Bryan Singer 1995). Field further claims that even our lives follow this structure. Unfortunately, he is unconvincing. "Birth? Life? Death? Isn't that a beginning, middle, and end?" he says (p. 29). It is, but how is this comparable to a film, unless he had been arguing that the three acts of a film are the opening shot, the film and the last shot (or end credits)? A more accurate comparison with life would be childhood, adulthood, and retirement, or something like that. But best would be not to place the three-act structure on life at all. Field then says "Spring, summer, fall, and winter - isn't that a beginning, middle, and end?" No, it is not. It is a circular event, with no end and no beginning. And it is four acts, not three. "Morning, afternoon, and evening. It's always the same, but different." He seems to have forgotten night. I will say more on three-act structures further down but suffice to say now that Field's book is surprisingly underwhelming in its structure and arguments, considering its fame, and a good editor would have removed the many repetitions and redundancies that fill the book. He has his moments but there are other manuals that are better, both in terms of the writing and of the richness of the examples.


As I said, the advice these books give are more or less the same. Writing Movies for example has this to say: 
Once you know the most important conflict of a scene, you can apply a key screenwriting technique: Enter a scene late and leave it early. Start the scene not when the conflict is brewing on the horizon, but - bam! - right when it's staring us in the face, maybe even when it has already reached fever pitch. End the scene as soon as that conflict has been resolved, and not a moment later. (p. 165, italics in original)
Trottier and others (including William Goldman in Adventures in the Screen Trade (1983)) say the exact same thing and it can be good advice. But not as a rule. Only when it improves a scene. Sometimes it is the build-up in the scene that makes the scene. This is one of the things that is frustrating with these books, the certainty with which things are said. "This is how you must do." or "In all films this is how it is done." But they are usually wrong in their firm proclamations. McKee is especially keen on nailing down opinions. He says for example that every scene should change something for the characters or, to quote, and with his italics, "No scene that doesn't turn. This is our ideal." (p. 36) This is not how scenes or films work. There are many kinds of scenes, some involve changes for the characters and some do not. There are comic interludes, there are pauses after something dramatic has happened, there are scenes of characters relaxing, scenes when vital information for the audience is given but without changing anything for the characters, and so on. Such scenes "do not turn" and nor should they.

One common criticism against books like these is that they are only relevant for mainstream commercialism and are unrelated to art, but anybody can benefit from a good manual and even non-commercial "high-art" films often follow some basic structure like the ones discussed in many of these books. Unless you are a wunderkind it is good to start with a solid foundation instead of being all over the place; to follow a certain structure so you have something to lean on. But the better the writer you are, the better you are of knowing when a rule is good or when it is diminishing your scene. In general, any book that says "This is how you must do it." is not as good or helpful as a book that says, as one of the writers in Writing Movies does, "In the final analysis, it's not fidelity to the 'rules' that makes a film great. Greatness comes when a film finds its true path." (p. 63) Personally I think this quote from Carrie Bradshaw in Sex and the City is good advice: "Why is it that putting a tie around a man's neck is sometimes even sexier than taking it off?" The point is that sometimes doing it one way is the best thing to do, sometimes doing the opposite is the best thing to do.

Another fundamental mistake almost all manuals make is that they mentioned some very successful films, describe the scripts of these films, and then argue that a script with such a structure and character development and story arcs and such is what will become a box office success. But that is not true. If you look at very unsuccessful films you will find they can have the same structure and character development and arcs but since those films had bad direction, or were miscast, or were released at an unfortunate time, or had terrible dialogue, or dodgy special effects, or ludicrous twists or some other things, they failed miserably. The script in itself is never ever enough for a film to be a hit, or even good, however slavishly it follows a template, and it is irresponsible to make such a claim. It is doing a disservice to the aspiring writers too. Well-written films are not necessarily successful at the box office, which these books often claim that you will be after having read them. Neither the Transformers films nor the superhero films that are drawing the masses to the cinema would be taught in these classes as examples of great writing. But since that is where the cash is, students might be better off engaging in bad writing to get ahead in life. I do not mean to be cynical, I only want to point out that these books pretend that great writing equals success even though there is no obvious correlation between great writing and great box office success. It was for example not Sunset Boulevard (Billy Wilder 1950) or All About Eve (Joseph L. Mankiewicz 1950) that were the most successful films of 1950. The big hit of the year was Samson and Delilah (Cecil B. DeMille 1949), which had more to do with biblical mayhem and Hedy Lamarr in tight, revealing outfits than great writing.


Almost all manuals are pretty insistent on the importance of the three-act structure, David Howard in How to Build a Great Screenplay says that "a story must have three parts, three acts." (p. 255) and adds for emphasis: "the story must, by definition, still adhere to a three-act underpinning." (p. 256) I find this incomprehensible. Sure, all films have a beginning, middle and end, but if that is all that is meant by three-act structure it is so banal and obvious there is no need to say it and it has no effect on the actual telling of the story. All individual sentences have a beginning, middle and end, so by that definition even this sentence has a three-act structure. But you do not teach school kids how to write by drilling into them that all sentences have a three-act structure because it would not be helpful for them.

But even though it is sometimes simplified as "beginning, middle and end," three-act structure means something else, and when you say that each act has a specific dramatic and narrative function, and that the beginning, middle and end must have certain specified lengths, then it stops being a basic rule and instead becomes something subjective and arbitrary and not something that is "by definition" true. Those who are keen on the three-act structure often divides the second act in two halves, which seems to me to turn it into two acts and the three-act structure becomes a four-act structure. If you pay attention you will also notice that the books are not in agreement on what those acts consists of and how long they are. Field's paradigm which I described above is the most well-known:

Act 1: the first 30 minutes
Act 2: the following 60 minutes
Act 3: the last 30 minutes

McKee has this setup:

Inciting Incident after 1 min
Act 1: climax after 30 min
Act 2: climax after 100 min
Act 3: climax after 118 min
Fade Out: the last two min.

That is ridiculously precise. But at the same time McKee says that different films are divided into varying numbers of acts. As an example, he says that Four Weddings and a Funeral (Mike Newell 1994) has five acts.

How is this helpful for screenwriters? What are they supposed to do with these acts when they are so arbitrary and negotiable? It is such a loose concept that I am not at all certain it serves a purpose.

I re-watched Jurassic Park (Steven Spielberg 1993) and by writing down the time whenever something shifted in the film, for us and the characters, I came down with what can be called three acts, like this:

Act 1: first 22 minutes, introducing the main characters.
Act 2: the next 38 minutes, introducing the dinosaurs and beginning the tour, everything still calm.
Act 3: the following 60 minutes, from the moment the T-Rex attacks the humans in the two cars and until the moment the T-Rex attacks again but this time saves the humans. That is 60 minutes of perpetual motion and basic survival.
Coda: the last two minutes of calm.

You can find a three-act structure there yes, but not at all in the way any of the manuals would tell you to arrange it. This is not a flaw in the film, it is a flaw in the manuals.

We can compare it to Bergman's Persona (1966), which is usually regarded as the antithesis of a conventional narrative film. It is 80 minutes long and, after the prologue, the first act is at the hospital. After 20 minutes they leave for the beach house and Act II begins. In the middle of the film, 41 minutes, the nurse Alma reads the letter from Elisabet and the dynamics of the film change, so that would be the plot point that divides the second act into two halves. After 57 minutes Elisabet watch a haunting photo from Nazi Germany and there is a fade-out to black. Then Act III begins, when Alma and Elisabet blend into one, and after 80 minutes Alma leaves and the film ends. That is three acts in accordance with Field's paradigm: the first act 1/4 of the film, the second act 2/4 (in two neat halves), and the third one the last 1/4. From this perspective, Persona has a more conventional structure than Jurassic Park but I doubt whether something meaningful is learned from this.

In short, you can look for three acts if you wish but do not confuse an arbitrary structure with objective rules and, while there is nothing wrong with having three acts, you do not need them.

One exception to the three-act structure consensus is John Truby in The Anatomy of a Story: 22 Steps to Becoming a Master Storyteller (2007). He claims the three-act structure ruins films and is responsible for why many films are badly written. But all the films he talks about in his book as examples of great writings are the same films that all three-act structure guys also celebrate as great writing. Truby might be disagreeing about the terminology but he still says the same things about the same films as all the others. However, instead of the three-act structure he is suggesting a different approach, and this approach is related to the title of the book. 22 steps. Considering how important you would think that they are it is somewhat surprising that they do not appear until page 268 but in any case, here they are:

1) Self-revelation, need, and desire
2) Ghost and story world
3) Weakness and need
4) Inciting incident
5) Desire
6) Ally or allies
7) Opponent and/or mystery
8) Fake-ally opponent
9) First revelation and decision: changed desire and motive
10) Plan
11) Opponent's plan and main counterattack
12) Drive
13) Attack by ally
14) Apparent defeat
15) Second revelation and decision: obsessive drive, changed desire and motive
16) Audience revelation
17) Third revelation and decision
18) Gate, gauntlet, visit to death
19) Battle
20) Self-revelation
21) Moral decision
22) New equilibrium

If you are wondering if it is true that all stories have these arbitrary steps, Truby says no. Some stories have as few as seven and some as many as 60 steps. Again, how is this helpful? That list is also a good example of how macho and gung-ho many of these books are. Those 22 steps look more like a PowerPoint presentation at Pentagon than a sensible guide for writing a romantic comedy. There is a strain of thinking within some of these manuals and among these "gurus" that has an unpleasant vibe to it, combined by all the talk of "the hero's journey", "the avenging angel" and other mythical stuff, which inevitably excite fanboys and online and offline extremists.

While Truby is vague about those 22 steps (the title of his book is rather misleading) he is unequivocal about these seven:

1 Weakness and need
2 Desire
3 Opponent
4 Plan
5 Battle
6 Self-revelation
7 New equilibrium

All good stories have them he claims. But this is not the case. Not at all. He just made them up, and since nobody knows about them unless they have read his book and since he has not exactly studied all stories ever told, it would be a miracle if he was right. It is not even clear whether the few films he uses in his book as examples have these seven steps that allegedly all good stories (and therefore all good films) have. Take the first and, for Truby, most important one, "Weakness and need". This is how he defines it, on pages 40-41:
From the very beginning of the story, your hero has one or more great weaknesses that are holding him back. Something is missing within him that is so profound, it is ruining his life. /.../ The need is what the hero must fulfil within himself in order to have a better life. It usually involves overcoming his weaknesses and changing, or growing, in some way.
This need is something the hero must be unaware of until the end of the film. ("If he is already cognizant of what he needs, the story is over.") One of the most well-known and successful heroes in contemporary culture is James Bond. What is his weakness and what is his need? Martinis? How about Indiana Jones? Sure, Indy is afraid of snakes, but this is not ruining his life and he is aware of this weakness from the beginning. Truby provides Clarice Sterling in Silence of the Lambs (Jonathan Demme 1991) as an example. But what weakness has she got that is ruining her life, and what need does she discover in the very end? She has her bad memories from childhood, the screaming lambs, but they are not a weakness ruining her life and she is aware of them long before the end. Some heroes, if by hero we simply mean main character, have such a weakness as Truby defines it, but whether they do or not have nothing to do with the quality of the writing but with the aims and needs of the individual script. It is not an essential aspect of storytelling.


These screenwriting manuals are like self-help books, or books about losing weight, and about as helpful. I do not think they have ruined cinema but I do not think they have done any good either. Sure, some who have read these books, or attended a "guru's" masterclass, have written some successful films afterwards but that is not proof of anything. Thousands of people have engaged with these theories and it is probably a statistical certainty that a few of these readers/students will become successful, regardless of the quality of the books and classes.

There is a scene in Paris When It Sizzles (Richard Quine 1964) which parodies cliched Hollywood screenwriting. These screenwriting manuals remind me of that scene, only they are not meant to be parodies. They take themselves seriously. Quine had better sense.

I have taught film writing and script development and I am not saying you cannot teach it. There are good and helpful things of a more general kind for a teacher to say and do. Usually though it is when working together with the students as they are writing a script that teaching is helpful, not by nailing down rules on the blackboard.

If you want to write a script you should write a compelling story with interesting characters. If you cannot do that I am not sure any manual can help you because they are not able to give you talent or imagination. They can only, at best, help you get started and, should you be stuck, help with some inspiration in the moment. They will not make you able to write a good or successful script, whatever they might claim on the back cover. The feeling I get after having read so many manuals over the years is that many are written by hucksters who cannot write themselves but are eager to earn a quick buck by fooling aspiring writers into buying their books and/or signing up for their workshops. The worst of them is Blake Snyder's Save the Cat! (2005), which I have discussed in an earlier article here. It is bewildering and absurd. But a few might be useful and the best I have read is David Howard's How to Build a Great Screenplay. It is by no means perfect but it is the most well-written and he covers more areas and is clearer and more vivid and helpful than any of the others I know. It is not a quick read, not a collection of bullet points, but 400 pages and you have to read it thoroughly and take notes. But if you are not prepared to do the necessary work then maybe you are not prepared to be a screenwriter.

In Play-Writing, Archer wrote: "Taken in its simplicity, this principle would indicate the three-act division as the ideal scheme for a play. As a matter of fact, many of the best modern plays in all languages fall into three acts /.../ many old plays which are nominally in five acts really fall into a triple rhythm and might better have been divided into three." (p. 106)

The most popular films to discuss in these manuals are Casablanca (Michael Curtiz 1942), Butch Cassidy and the Sundance Kid (George Roy Hill 1969), The Godfather (Francis Ford Coppola 1972), Chinatown (Roman Polanski 1974), Thelma and Louise (Ridley Scott 1991) and American Beauty (Sam Mendes 1999). Die Hard (John McTiernan 1988) and The Shawshank Redemption (Frank Darabont 1994) are also popular.

Friday 8 March 2019

Stanley Donen (1924-2019)

The last film Stanley Donen made for a cinema release was Blame It on Rio (1984) and in one sequence Michael Caine (who plays the male lead) is looking out the window of the airplane he is a passenger on. What he sees is Rio de Janeiro, but then the scenery changes from colour to black and white and instead of the present-day Rio it is a clip from the film Flying Down to Rio (Thornton Freeland 1933), with women dancing on top of airplanes flying over Rio.

This is a moving reference to the very film which made a then 9-year-old Donen want to get into the dance and film world after he saw the extravagant musical in the cinema. As soon as he was old enough, he got himself immersed in that world. At first he was primarily a choreographer but he also directed scenes, and two exceptionally fine examples of his art, with Gene Kelly, are from Cover Girl (Charles Vidor 1944) and Anchors Aweigh (George Sidney 1945). He conceived of them and choreographed them with Kelly and then directed them and did the post-production too.

A few years later he became director of whole movies and not just individual scenes. It was a very uneven career, with perhaps more failures than successes. Some consider him the greatest maker of musicals, even better than his rival Vincente Minnelli, but for me Minnelli is far superior both in individual films and as an artist, and with a much richer oeuvre. It is however interesting to compare them, as they are so different in style, vision, temperament and ideas.

Here are some favourite sequences:

Royal Wedding (1951)

Singin' in the Rain (1952)

It's Always Fair Weather (1955)

Funny Face (1957)

Charade (1963)

Charade is uncommonly delightful and an excellent blend of light comedy and thriller. Two for the Road (1967) is also good, and for a change Audrey Hepburn is there playing against a man younger than her, Albert Finney. It also has one of Henry Mancini's most beautiful scores.

For a comparison between Donen and Minnelli this clip from Minnelli's The Pirate (1948) is a good starting point, as it is more or less the same song as Donald O'Connor does in Singin' in the Rain, seen above. Cole Porter wrote the original, Be a Clown.

I shall not show any clips from Blame it on Rio because I cannot stand it. But here is the scene from Flying Down to Rio.