© Mike Harris; reproduced by permission. First published in The Handbook of Creative Writing edited by Steven Earnshaw, published by Edinburgh University Press, 2007; website at: ......http://www.eup.ed.ac.uk

Introduction to Scriptwriting

Mike Harris


The Argument

Beckett, Tarantino, Churchill, Shakespeare, Ponte, Bunuel, Wagner and the writers of The Archers all have something essential in common: if they want an audience to understand what they say, they have to keep us in our seats till they’ve finished saying it (even home-recording facilities have not as yet affected the basic structure of adverts-and-audience driven TV drama). The poet doesn’t have to do that, and although the novelist needs the reader to keep turning pages, her task doesn’t have to be completed in one sitting. Only the scriptwriter has to deal with the problem of passing time in this immediate and visceral way

Scriptwriters create the interest and attention of an audience mainly through narrative. Even scriptwriters who subvert narrative can’t do so without using its conventions. This is not surprising given that we’re soaked in stories from the earliest age. What is surprising is that when we come to write we often focus on anything but narrative: the ideas or the language or the characters or our experience, as if narrative were a lower form of aesthetic life which can either be dispensed with or left to itself to evolve slime-like while we concentrate on these spiritually higher matters; and it’s the main reason why scripts don’t work. So this chapter is about dramatic narrative and its practical uses. 


The Basics

An exercise that establishes the basics is to give a group five minutes to write a simple story or anecdote (no dialogue or descriptions and no longer than a paragraph). Collect them in, un-named, re-distribute them and invite people to read out the one they got. Each will inevitably involve one or more of the essential building blocks (a specific world, a protagonist, an incident that starts the story off, a chain of events, plot twists, a theme, etc.) which can be identified and form the basis of further discussion.


In a drama things have to happen. Yes, really. Small children who write stories that are simple chains of events (‘I got up, had my breakfast, walked to the bus-stop, got scooped up by a passing Pterodactyl and dropped my homework’) have grasped the essentials. In King Lear our protagonist gives up his kingdom, divides it up on the basis of which daughter says she loves him best, disinherits the one who won’t lie, marries her off to a French-man and then banishes the best friend who points out to him the folly of all this, and that’s only scene one. What the plot of Lear has, which the child’s story above doesn’t, is plausibility.

Chains of plausibly-connected events keep people in their seats not only because they create anticipation and suspense but also because in them character is explored and thoughts provoked. Novelists editorialise and write down the thoughts of their characters at will. A scriptwriter can of course use soliloquy and narration and in radio drama all forms of direct address can be especially effective (see the chapter on ‘Radio Drama’) – but performances composed entirely of these put a heavy strain on audience attention: which is probably why Ancient Greek and Medieval priests added incident and dialogue to their ceremonies and thus gave rise to drama in the first place (See Thomson, Aeschylus and Athens (1980) and Chambers (1963), The Medieval Stage for detailed discussions of this).

In the main, therefore, Drama explores character and ideas through the medium of events. Generally in the following way: When characters confront obstacles they have to act (or not act) in one way or another. The choice they make reveals them. We think Macbeth is a loyal subordinate but when ambition overcomes the obstacle of his conscience, he kills the King and we realise he’s not so loyal after all. Robert Mckee calls this a ‘story event’ (McKee 1999: 33-5). And that’s a useful term because it makes a distinction between events in life, which are more often than not meaningless, and events in a play, which can’t be.

The more ‘story events’ in a script, the more obstacles we see a character dealing with and the more various, the more we are likely to understand (or be puzzled by) him, and the more liable we are to ask questions: why do Vladimir and Estragon persist in waiting for Godot? Why do they not leave, or commit suicide? Why don’t we?


The mechanics of this can be explored by inventing aloud a simple collective story in which the group can employ only physical obstacles. Here’s one: a powerful woman dresses in a rush because she’s late for the most important business meeting in her life but still pauses to hug her beloved poodle. She goes to the front door – and? Physical obstacle only, please. It won’t open. O.k. so she looks for the key – and? She can’t find it/it breaks/whatever. Fine, so she tries the back door all the windows and? The windows won’t break, the phone won’t work, people don’t see her when she waves. Time-shift: a week later. She’s eaten all the food in the fridge. What’s she going to do? Eat the dog! shouts a cynic. Not yet, you reply, because she loves it, doesn’t she? Time-shift three more days. Is she hungry enough to eat the dog now? Yes! Cry all but the most hardened pet-lovers. But she’s got to catch it first and kill it, hasn’t she? Stop the story and discuss how we’ve turned this sophisticated woman into a primitive hunter and tested the limits of her values; mostly by keeping the doors shut.


The list of story events in a play are its most basic structure and if we haven’t got enough we haven’t got a play. The approximate number of significant story events a script needs depends on taste, length, genre and medium. Full-length films tend to have more than full-length stage or radio plays. Action thrillers more than European art films. Writing by numbers is idiotic but given that most of us generally under-estimate just how many narrative ideas we need to tell our stories effectively, it’s worth quoting (sceptically) a guesstimate. Mckee reckons you need 40-60 full blown story events in a feature film and 40 or so in a full-length stage play. Most first drafts I come across, including my own, certainly have far fewer, which is one reason why they aren’t working.


Count the number of times in a great play or film when characters confront obstacles. Then count the number of times your own protagonist confronts obstacles. It will almost certainly be far fewer.


Obstacles in story events aren’t just physical. They can also be the conflicting desires of other people: for example the younger son who resents his returning father in Andrei Zvyagintsev’s The Return; or social: for example The racism of the 60s deep South, against which Sidney Poitier’s sophisticated New York cop has to struggle in order to conduct a murder investigation in In The Heat Of The Night; or psychological. In Hiroshima Mon Amour, we gradually realise that The Woman’s rejection of The Man is a deep-seated response to the death of a lover when she was 18. The arena of such a script is primarily interior; its gladiators the competing impulses within a character: not Arnie with an Uzie, but that memory you don’t want to walk over. Radio Drama is particularly suited to dramatising psychological conflicts.


Most story events in drama occur because Character A wants something different than Character B. This can be reduced to an equation, if you are so inclined.  A wants x, B wants y and therefore z (where z is what happens as a result). If there is ever a moment in a script when A wants x and B wants x too, and they both get it, it’s either the end, a third antagonistic force is about to ruin the idyll (for example, Martians, a psychopath, a Bad Thought, etc.) or your story just died. It may also suggest that you have characters who are too similar. A cast of characters should be as diverse as possible in order to maximise the range of possible conflict. The equation above is a crude but useful way of testing whether story events in an outline or in a scene are working, and what’s wrong with them if they’re obviously not. What does your A want? Is it different than what your B wants? If it isn’t, why not? If it is, what happens as a result, if anything? If you don’t know the answer to all these questions, why have you started writing?



Scenes and Sequences

Whenever the narrative shifts in time or space, the scene changes, with the exception of single-room dramas. In these the location is fixed so the scene changes whenever the script shifts in time and, if you like, whenever a character enters or leaves thus affecting the balance of power, for example see 12 Angry Men.

The essential core of any scene on stage, and a long scene (or a sequence of  quick-cutting very short scenes) on radio or screen, is at least one story event. Most will have more. For example the two-page first ‘scene’ of Harold Pinter’s The Dumb Waiter, (between lights up and Gus retiring to the toilet) contains a couple of dozen small story events. There are long scenes that don’t have any and most need to be dumped or re-written. Some might be giving or establishing essential information but they’re a lower order of life and are altogether better if they contain a story event as well. There are lots of short scenes in film which simply give information. For example establishing shots which give us an external view of the location in which the next scene will take place, and travelling shots like the many we get of John Mills and co trudging through blizzards in Scott of the Antarctic. Scenes in film can convey a simple single piece of information very quickly. More complicated meanings are often made by editing together lots of these simple shorter bits. This can also happen in radio drama, but its less common. The basic narrative unit in film is frequently the scene sequence, for example the Polish wedding in the first act of The Deer Hunter, and then it’s this that always needs at least one story event. 

To all intents and purposes a scene or scene sequence is the story event that controls it. When we list story events, in the order in which we think they might occur, we have our most basic plan.


Prose descriptions of what’s going to happen in a script are a good way of forcing one-self to think and plan. This is not a late intrusion of Hollywood into serious writing: ‘The poet should first plan the general outline and then expand by working out appropriate episodes’ (Aristotle’s Poetics, 4th century B.C.)  Our reasons for not planning enough have nothing to do with seriousness and everything to do with the fact that thinking is much harder than writing screeds of bad dialogue; which is why we generally start writing dialogue before we know properly what we’re writing it for and then run out of ideas. It’s possible of course to finish a draft and then try to decide what it’s all about. The problem is you might just decide it’s not about anything and have to dump months of work, which is demoralising, wasteful and not viable on a serious writing course or with a commission because both have deadlines attached. An outline, or something like it, is therefore essential. You can discuss an outline with others and discover quickly that there’s only three story events in it and then profitably wonder what’s going to be going on for the other hour and three quarters? An outline can be re-drafted in far less time than a whole script and then repeatedly pulled apart and put back together until it’s got enough plausible story events in it, so that if you start writing and get stuck on one scene you can drop it, write an easier one, and come back to the hard one later when, if you’re lucky, you’ll have figured out the first problem whilst you weren’t looking.



Of course, planning and organising aren’t just a matter of listing in roughly the right order all the story events one can think of, although that’s a good start. It’s well known, for example, that every script has to have a beginning, a middle and an end. So why not start at the beginning?

Problem is, stories don’t start at the beginning. They start when we know everything we need to know in order to understand the story when it does start. Take the panto script of Cinderella. Before we can make any sense of the main storyline involving a fairy godmother, a ball and a glass slipper test, we have to know that Cinders is badly treated by her step mum and sisters, that she is deserving and beautiful, lives in an undemocratic state run by a bachelor prince who is looking for a bride, and also it’s the kind of world in which magic can happen. That’s the set-up, without it the story can’t work and it’s generally sensible to do it as quickly and as economically as possible. In Die Hard we need to know that Bruce Willis is an ordinary-Joe New York cop visiting his estranged and more successful wife in California, and that he still loves her. We get Cop when a fellow plane passenger sees his gun and asks about it. We get the backstory on his marriage when a nosy chauffeur badgers it out of him. We understand some of the complexity of his feelings when he keys his wife into a computerised reception and discovers, resentfully, that she’s reverted to her maiden name.

When the set-up’s completed the story needs something to make it start. This is a very particular story event or a series of proximate and linked events – call it the Inciting Incident or Plot Point 1 or what you like (McKee prefers ‘inciting incident’, Syd Field ‘Plot Point 1’ – it doesn’t matter) – but it’s the big bang that disturbs the status quo of the fictional world, creates volatility and makes narrative life possible. ‘The inciting incident’ in the panto Cinderella is of course the fortuitous arrival of a Fairy Godmother after Wicked Step-mother has banned Cinders from the ball.

Robert McKee insists that ‘the inciting incident’ has to happen ‘in the story’ (McKee 1999: 189-94) and Syd Field that it happen no later than page thirty (Field 2003: 119) but the fact is that in drama outside mainstream Hollywood it occurs almost everywhere. In Chekhov’s Uncle Vanya it’s the arrival of the professor and his young wife months before the play begins; in the Oresteia of Aeschylus it happens in the mythic past when Tantalus feeds the Gods on his own son’s flesh and so brings down a curse upon The House of Atreus. And in Christopher Nolan’s back-to front memory-thriller, Memento it happens, quite remarkably, at the very end of the film. But it’s normally a lot closer to the beginning of a script and tends to need to be if you want to avoid your audience turning off before you’ve even started. In Die Hard it’s the classic 20 minutes in, when Alan Rickman’s gang capture the building and Bruce Willis’s wife but not Bruce himself. Elizabethan theatre-goers were as familiar with the conventions of stage kingship as we are with cop-show procedurals so the inciting incident of King Lear can occur early (before the end of Scene One). However, in Brazil, Terry Gilliam’s quirky, complicated dystopia about a state dominated by a cruel but comically inefficient bureaucracy (and ventilation ducts), it takes nearly 50 minutes to arrive. In a long set-up it helps if there’s a sub-plot in it to tide us over till the real story begins. In Brazil it’s the wrongful arrest of the innocent Buttle instead of the renegade ventilation engineer Tuttle. It’s even better if the sub-plot is a thematic variation on the main one, which Tuttle and Buttle is, so it not only grips us but also helps us understand the rules of the world we’re in.

No matter where it occurs, however, the inciting incident obliges the protagonist to act, or leads us to expect that he will, eventually. In Die Hard, Bruce Willis has to because he’s trapped and Alan Rickman’s just kidnapped his wife. Hamlet thinks he has to because the ghost of his murdered father has just grassed up his step-dad.



... and then what? There’s a whole series of obstacles in the way, of course. What keeps things going is the protagonist’s persistence in seeking to overcome them and the insights we gain as that happens. Protagonists persist long after most of us would have given up and gone to the pub. That’s how they get to be protagonists. Even Hamlet – the greatest prevaricator in world drama – doesn’t actually give up, and Macbeth fights on even when he realises his death is inevitable because Macduff was Not of Woman Born.

There are of course dramas that don’t have a single main protagonist. A good example is Paul Thomas Anderson’s magnificent Magnolia in which the stories of half a dozen equally important characters are inter-cut throughout. Several things are instructive here: each character’s story is plotted as carefully as if he or she were the sole protagonist; we then discover, to our great narrative satisfaction, that the stories gradually link up, and that they all are subtle variations on a common theme. Being unable to decide who is one’s main protagonist, is not the same thing at all.

But most scripts do have a single main protagonist and they are very pro-active. Syd Field suggests that you can’t have one who isn’t. If this were true, drama would lose some of it greatest protagonists. For example, The Alienated Hero and an awful lot of Child Heroes. Alienated protagonists find it hard to act – because they are alienated. Hamlet decides he’d better kill his step-dad at the end of Act One and then spends the best part of the remaining four acts not doing it. Child protagonists in an adult world tend to be limited in what they can do. What happens therefore is that other characters act instead. In Hamlet virtually every other character in the play does what Hamlet says he ought to do: Ophelia commits suicide, Laertes acts to revenge the death of his father and so on. In Billie August’s Pele the Conqueror the child hero is primarily an observer of older lives but each is in some sense a trial run for what he might be when he grows up. As the alienated or child protagonist watches, or dithers, the narrative energy goes to the sub-plots of secondary characters, until he finally acts. Hamlet slaughters almost the entire cast, and Pele, now pubescent, leaves to make his own life.

King Lear shows how obstacles in The Middle get progressively tougher. When Goneril tries to strip him of some of his rude retinue of knights, Lear goes to Regan in the hope of better treatment but she takes all his knights away and Lear, unable to reconcile himself to this or to do anything about it, goes mad.

The course of the plot through the middle doesn’t have to be progressively Worse, it can get Worse then Better, or Better then Worse. But it does need to progress and the cost of action nearly always escalates. 

Matters are generally forced by one or two major story events that cause things to change (reversals) or which bring the audience or the protagonist to a very different understanding of what has been happening (recognitions). 1. Macbeth kills Duncan and there’s no going back, but Duncan’s heirs escape. 2. He’s consolidating his power at a banquet for the nobles when Banquo’s ghost appears and throws him into embarrassing mental torment. 3. He goes back to the witches for re-assurance, they confirm that Banquo’s descendants will inherit the throne but tell him his position is secure unless Burnham Wood comes to Dunsinane and he can’t be killed by any man of woman born so he’s happy, until….

When the consequences of one of these main story events, or ‘plot turns’, or ‘twists’, have run out and the next has not yet kicked in, the audience will start to fidget and it’s time to reach for sub-plot again. In Memento when the love-interest, Natalie, proves to our protagonist that the man we saw him shoot at the start was the one who killed his wife, the main story seems to be over. At this point, Nolan gives us in flashback the vital sub-plot/backstory of how our hero discredited another amnesiac whilst conducting an insurance investigation and the tragic consequences that ensued. This holds us in place until we shortly realise that Natalie is not to be trusted and the main-plot is on the road again.


Pick a genre and then list what always happens in it. Then pick another, and so on. If we know what generally happens in stories and therefore what the audience will be expecting, there’s a chance of coming up with something different. Make everyone decide what genre they are writing in. Some will claim they aren’t but they are and pretending not to be is the surest way to write the clichés of the genre in which we are, in fact, writing. Hamlet is a great play, in part, because Shakespeare uses the conventions of the Elizabethan revenge tragedy to do something different. Thelma and Louise works because it re-imagines the Buddy-Buddy adventure genre for two modern women. When we’ve worked out what genre or genres we’re writing in it’s possible to root out the clichés and make something different happen.



…and what happens in the end? The protagonist generally comes up against the greatest obstacle of all, which is usually in some form or other himself, and then he overcomes it, or not. In Krapps Last Tape, Beckett’s protagonist literally confronts his younger self in the form of a tape-recording, whilst the final obstacle The Woman in Hiroshima Mon Amour has to face is not whether she loves her new married lover enough to be with him, it’s whether she can abandon her love affair with Death and, in the end, we’re not sure whether, in choosing the lover, she’s done that, or its opposite. This is the Ambiguous Ending. There are also, of course, the Happy and Sad ones. Uncle Vanya ends with Vanya and his niece going back to their farm accounts as they return to their life of provincial obscurity (Unhappy). In Groundhog Day, Bill Murray gets the girl and time stops repeating itself (Happy). Hollywood films now end with such idiotically predictable happiness that all reasonable people leave the cinema wishing the entire cast had died in a bath of acid instead. But a Happy ending is perfectly viable if you’ve earned it after inflicting lots of misery, and its still plausible despite that.

Both McKee and Field require the penultimate sequence or section of a script to be in the opposite mood to the end. So a happy ending should always be preceded by an unhappy sequence and visa versa. This is one of the many reasons why it’s possible to watch almost any current Hollywood movie and predict exactly what will happen. But that up-down-up-down rhythm is common in good scripts too. Just when you think it couldn’t get any worse for Lear, Cordelia turns up with an army and you just know its going to be alright, and then she’s defeated and it gets much worse. There’s a good reason for the prevalence of this pattern. A string of dialogue, scenes, sequences, turns or big story events written in the same emotional register runs the risk of inappropriate audience reaction. One unhappy bit = sad audience; two unhappy bits in a row = not quite so sad audience; three unhappy bits in a row = audience stifling inadvertent guffaws. Scripts that repeat Sad-Sad-Sad or Happy-Happy-Happy and get away with it have to play half-tones in scenes or descend into bathos and worse. In Sarah Kane’s Blasted, eyes are sucked out, babies eaten and people try to strangle themselves to death in such quick succession that the tittering audience member begins to feel like he’s just stumbled into Benny Hill’s Theatre of Cruelty. Whereas, although scene after scene in Magnolia is unutterably sad, the bleakness is subtly interwoven with strands of dark comedy.


Literary critics merely describe plays. We have to make them better. To do that requires practical evaluative criteria. Academic debates about the impossibility of evaluation since it is always conditional on class, gender or history, or about the irrelevance of The Writer either because we’re all dead or because our intentions are always fallacious, are deeply damaging to any writer who takes them seriously. We do exist and we do have to try and make our work better. The evaluative terms described above and below work for me and have proved useful in developing scripts with students, but they may not work for you. Suck them and see. For example: Has a draft outline got an inciting incident, and does it start the story off convincingly? Has it got a couple of major ‘turns’ in the middle; if it’s only got one, or none, that’s why it feels flabby. Who’s the protagonist? If the writer doesn’t know (a surprisingly common phenomenon) that’s why the whole thing feels unfocused. Look at the cast of characters. What motivates each one, and is it incompatible with what motivates the others? If not, why not? How many of these characters have stories of their own, apart from the protagonist? None? Then the script is going to feel very thin and you’ve got no sub-plot to play with. Kill characters who do nothing or merge them with another cipher to create a more interesting composite. Are characters being pushed far enough? People in plays are much less dull than people in life and do much more extreme things, even when they’re supposed to be ordinary. For example Chekhov’s ‘ordinary’ characters are much given to suicide, and waving guns around. Most important of all, what ideas or values are being explored in the drama? If none are, why are you writing it? Life is very hard to change. Drama on the other hand is an entirely virtual reality that the writer can fiddle with at will until it seems to work.



Most people think Drama is dialogue but they’re wrong. Drama is about doing things. The word itself is ancient Greek meaning deed or action and in his Poetics, Aristotle lists in order of importance the constituent elements of Tragedy (the dominant dramatic form of his time): 1. Plot 2. Character 3. Ideas 4. Dialogue.  If you’re disinclined to credit dead foreigners (even when their compatriots invented what you’re trying to do) you might prefer to consult the Oxford English dictionary which defines drama as, ‘a series of actions or course of events leading to a final catastrophe or consummation’. It doesn’t mention dialogue.

The most common fault in scripts is Too Many Words, which is understandable because the most obvious way of externalising thoughts and feelings is by having one character tell another character what he’s really thinking and feeling. The problem is, if a character tells us everything about herself, what’s left for the audience to find out and why should we carry on watching? Experience teaches that people aren’t like that, anyway.  What we say is rarely what we think and what we think is not what really motivates us and what really motivates us is, more often than not, a complete mystery: what is Hamlet’s problem at bottom? Or Lear’s? We’re just not sure, so we keep thinking about them long after the play has ended and centuries after they were first written. Of course, fiction isn’t reality and need not be held hostage by it, but when an aesthetic strategy is neither true to life nor dramatically effective, why use it?

The principle way we explore character in drama, as in life, is by comparing what people say, with what they do (‘I love you more than word can wield the matter;/Dearer than eye-sight, space and liberty’ – Goneril to Lear, (I.i)). Which is why it makes sense, when planning a scene, to start not with a list of things we think should be said but with the main story event in it. Dialogue will then happen if A and B happen to use words to get what they want. They may also use violence of course, or silence or gestures, or the merest glance. Moments of silence in drama are more expressive than the noisiest dialogue. Cassandra stands silently to enormous theatrical effect for most of the 2nd Act of Aeschylus’s Agamemnon. Whole sequences on film can (and should) contain no dialogue at all. One of my favourites is the long speechless sequence in the middle of the Coen brothers Blood Simple, which culminates in a good man burying his lover’s husband alive because he can’t bring himself to kill him. Perhaps most remarkable of all is the almost total silence of Anthony Minghella’s protagonist in the radio script of Cigarettes and Chocolate.


Getting rid of unnecessary words: Take a few bad or early draft scenes and set everyone to cut as many words as possible but still tell all the story. Very soon you can’t see the black for the red. Then get them to re-write a sequence of their own scenes with very little dialogue – no more than 10 or 20 words.


Dialogue in a scene based on conflicted desire is more dynamic and engaging because it involves a struggle for power the outcome of which is uncertain.

In a good scene or scene sequence power is held in contest and then shifts at least once from one character to another. In the Oresteia, Clytaemnestra urges Agamemnon to cross the blood-red carpet that will lead him to his death. He refuses, she persuades, he crosses and is lost. In the case of psychological drama, power passes between one aspect of a person’s psyche to another. In Tyrone Guthrie’s groundbreaking 1930 radio play Matrimonial News the thoughts of shop-girl Florence are personified and conflict with each other as she sits waiting for her blind date to arrive.

Conflicted dialogue can sound more like real speech with all its interruptions, half finished phrases, non-sequiturs and so forth as in Caryl Churchill’s Top Girls. This is not because writers set out to copy ‘real-life’ speech, because they don’t, being primarily interested in the exploration of character and ideas within a particular aesthetic structure. And most ‘real speech’ is too unutterably dull to be worth copying, anyway. A tiny portion of it, however, embodies interesting struggles for power or status and we remember this as ‘real speech’ because it sounds like good dramatic dialogue. 


Set people to tape-record some ordinary conversation in a pub or at home.  Each should select a section that interests them, transcribe it exactly and then read it out aloud. Underneath the naturalistic surface there will almost certainly be power-play. It was this that caused it to be selected from all the taped dialogue that didn’t contain conflict.


A legendary piece of advice on scene structure (probably emanating from William Goldman), is “go in late, get out early”. By which he means start the scene as close to the moment of conflict as possible (in mid-conflict even) and get out before it can resolve. The late start means we avoid the kerfuffle dialogue that infests so many first drafts (‘Hello come in, sit down, have cup of tea, lovely day isn’t it’) and the early departure ensures that audience interest will be maintained as they are made to wait to find out what happens next. Even the dullest information-giving can be lifted this way. At the start of Lear, Shakespeare offers one of his, ‘two courtiers doing the set-up’ dialogues that can feel drearily functional but here he starts half way through a debate:


Kent: I thought the King had more affected the Duke of Albany than Cornwall?

Gloucester:       It did always seem so to us; but now….


And the small disagreement makes a significant difference. Disagreements can of course make all the difference: contrast bad dialogue in which, for no good reason, characters tell each other everything that’s in their head and hearts, with those scenes in which a succession of climaxing disagreements force out secrets. In these we accept the revelations because they are first refused many times.  In Act 2 scene one of Lorca’s Blood Wedding a peasant Bride is visited by Leonardo, her former lover who is himself now married. The more she evades his advances the more we suspect that she is protesting too much, and the more Leonardo is forced to reveal. This scene also illustrates the importance of making things literally happen in scenes. During this scene the wedding guests are arriving, adding crucial pressure to the confrontation. Setting and physical action should be written into each scene.


It’s a good idea to write (on a sequence of cards or A-4 sheets, in an outlining programme, on the back of cigarette packets or whatever suits) the main story event in every scene, where it’s occurring and what is literally happening in it, apart from talk. After that scribble down anything else you think needs to go in a scene. For example any dialogue that is starting to emerge, any other information that has to be carried, and anything else that occurs to you.


But what of soliloquy, narration and choruses? There are plenty of occasions in Shakespeare, and in radio drama, when characters directly address the audience but such passages generally dramatise either psychological conflict (‘To be or not to be’), or are in conflict with what happens before or after it, for example Edmund’s ‘I’m the Bad Guy’ soliloquy in Act One Scene two of King Lear undermines Gloucester’s defence of him in Scene One. And in every case soliloquies must move the drama on or we need to ask, why are they there? When Mozart was trying to re-invent the moribund form of Opera Seria with Idomeneo, he bombarded his old fashioned librettist with demands that the lyrics embody the action and not simply repeat lyrically what the audience knows already (Anderson 1997: 662).

A Chorus can perform many different functions. In Greek Tragedy it sometimes becomes a character in conflict with or a confidante of the protagonist, for example the Corinthian Women in Euripides Medea; at others it performs a holding operation at a moment of great suspense, making the audience wait whilst it reflects on what has gone before. Lorca re-invents this technique very effectively in Act 3 of Blood Wedding. Greek Choruses often related violent off-stage action or essential back-story and, Greek Tragedy being closer to opera than modern drama, the lack of dramatic action on stage at these points was compensated for by music, movement and heightened verse. Which is precisely what happens in Steven Berkoff’s Greek in which characters speak Mockney-Shakespeare and the sheer virtuosity of the language and the physical theatre (actors turning each other into motorbikes, etc.) almost blinds you to the fact that what you are watching is little more than a series of misanthropic music hall sketches. Narrators work like choruses. That is, best if they are part of the play’s conflicts and best of all when unreliable. Direct address written simply to cover an inability to tell the story through dramatic action, is best avoided.


Strange Worlds, Minimalism and Anti-narratives

It’s widely believed that lots of dramas work without using narrative, or with so little of it that it doesn’t signify. Richard Gilman, for example, thinks that Chekhov’s Three Sisters abandoned, ‘the usual linear development of a play ... from a starting point, to exposition and development ... to a denouement’, and instead, ‘worked toward the filling in of a dramatic field’ (Chekov 2002: ix). If this were true we would justifiably baulk at going to Chekhov to learn how to write minimalist dramas of ordinary life, for how does one teach, or learn, how to ‘fill in a dramatic field’?

Fortunately, it’s not true.

At the most basic narrative level, lots of things happen in Three Sisters: A weak, selfish woman marries their beloved older brother who becomes a wastrel and a gambler. His wife gradually takes over the family house, forces the sisters out and has an affair with his boss. Irene gives up her dreams of Moscow to marry the manager of a brick works, who is then killed in a duel by a thwarted admirer. Masha has an affair with an army officer who is unhappily married to a lunatic and then, when he leaves, finds out that her boring husband knew all along. Oh yes, and there’s a major conflagration at the start of Act Three. These events all occur in regular time-sequence (not back to front, at random or in flashback) and the multi-narratives are organised into inciting incidents, progressive developments, sub-plots and a dramatic climax.

But Chekhov is very different from the writers of melodrama against whom he was in revolt because he does successfully create a sense of quiet unchanging lives riven by regret and despair. He doesn’t do this by abandoning narrative but by using two of the oldest known narrative techniques to help give the impression that this is what he is doing. He sets his most melodramatic events (the duel, the fire and most of the adultery) off-stage and tells the rest, ‘on the cut’, that is, in time-shifts between acts, leaving him free to devote onstage action and dialogue to what may appear to be ‘ordinary life’, but which, amongst other things is getting the audience up to speed on what’s happened since the last act. We realise this soon into Act 2, and are thereafter partly hooked by the need to find out not so much what’s going to happen next, but what’s happened since.

But if we want to boldly go beyond the merely humdrum to the Utterly Purposeless and Absurd, the best pilot is that master story-teller and all-round entertainer, Samuel Beckett. But not if we think that his plays, ‘lack plot more completely than any other works’, because they adopt ‘a method that is essentially polyphonic ... [confronting] … their audience with an organised structure of statements and images that interpenetrate each other and that must be apprehended in their totality’ (Elsin 1976: 44). Try teaching that. ‘Everyone interpenetrate your meanings in their totality now please!’

Luckily Beckett no more does this than Chekhov fills in fields. Take Waiting for Godot. He gives us four very distinct characters (Vladimir, Estragon, Lucky and Pozzo) who vie for status throughout a play which is, in essence, a simple one-room, invasion of space drama. There’s an inciting incident: when Vladimir informs Estragon and us that they are waiting for Godot, which induces the wholly conventional audience anticipation that Beckett plays cat and mouse with for the next two acts. On several occasions in Act One he deliberately, Hitchcock-like, makes us think that Godot has arrived:


Vladimir: Listen!

(they listen, grotesquely rigid)

Estragon: I hear nothing

Vladimir: Hssst!

But of course he hasn’t. Eventually, instead of him, Pozzo and Lucky arrive to inject danger into the play just when it’s in danger of flagging. In between these conventional narrative tropes Becket occupies his waiting audience with stand-up patter and music hall slapstick routines. In the second half the pattern is often said to be identical but isn’t. There’s considerable development – the power relations between Pozzo and Lucky have been transformed, Vladimir and Estragon are more despairing and it is also, sensibly, much shorter than the first, because by then we know that Godot probably isn’t coming and Beckett is too wily a story-teller to push us over the brink. He gives us the illusion of a purposeless, story-less world by invoking narrative expectations that he never, completely, fulfils but which allow him to portray ennui without making his audience feel it (too much).

When one turns to Absurdists like Ionesco and Ominous-ists like Pinter, the case hardly needs making. Ionesco’s shtick is injecting highly conventional narrative with something ridiculous to see what effect this has on the audience. Thus the plot of Rhinoceros is Invasion of the Body Snatchers with rhinoceroses, and The Lesson is 10 Rillington Place with added gibberish. Pinter’s preferred mode is to put lots of pauses into highly conflicted, idiomatic dialogue whilst refusing ultimate explanation of his characters’ generally violent or oppressive behaviour. The Room, for example is, like Godot, a conventional invasion of space drama, except we never learn exactly why Riley has come back for Rose or why Bert brutally kills him in the end.

And all these writers also maintain attention and interest second by second with dialogue that contains almost as much blocking and gainsaying per-scene as a soap.

Scripts that inhabit strange worlds are sometimes conflated with these ‘anti-narratives’ but they are generally even more conventionally structured. So only two additional points need to be made. Firstly, the main plot in a weird world script ought to arise at least partly from the strange conditions of that world, otherwise what’s the point? So, in Delicatessen, which takes place in a post-holocaust Gallic No-Time in which food is scarce and cannibalism rife, the unemployed clown-protagonist is (of course) in danger of being turned into steak fillets by his butcher-landlord. Secondly, the stranger the world the more helpful to the audience will be at least one conventional narrative strand. It’s not an accident that Delicatessen, Brazil, and Groundhog Day (in which the protagonist experiences the same day over and over again) are all partly powered by a love-story.






Aeschylus (1977), The Oresteia: Agamemnon, The Libation Bearers, The Eumenides, London: Penguin.

Anderson, Emily (1997), ed., The Letters of Mozart and his Family, London: Macmillan.

Anderson, Paul Thomas (1999), Magnolia, screenplay: Paul Thomas Anderson.

Aristotle (2000), Poetics, London: Penguin Classics.

Beckett, Samuel (DATE?), Krapp’s Last Tape, Waiting for Godot, London: Faber.

Berkoff, Steven (1989), Greek, Decadence and Other Plays, London: Faber.

Caro, Mark & Jeunet, Jean-Pierre (1991), Delicatessen, screenplay : Gilles Adrien.

Chambers, E. K. (1963), The Medieval Stage Vol. 11, Bk 3, Oxford University Press.

Chekhov, Anton (2002), Three Sisters, Uncle Vanya, London: Penguin.

Churchill, Caryl (1993), Top Girls, Plays 2, London: Methuen.

Cimino, Michael (1978), The Deer Hunter, screenplay: Michael Cimino, Deric Washburn.

Coen, Joel (1984), Blood Simple, screenplay: Joel Coen, Ethan Coen.

Esslin, Martin (1976), The Theatre of the Absurd, London: Penguin.

Field, Syd (2003), The Definitive Guide to Screenwriting, London: Ebury.

Frend, Charles (1948), Scott of the Antarctic, screenplay: Walter Meade, Ivor Montague.

Gilliam, Terry (1985), Brazil, screenplay: Terry Gilliam, Tom Stoppard.

Goldman, William (1984), Adventures in the Screen Trade, London: Macdonald.

Ionesco, Eugene (1962), Rhinoceros, The Chairs, The Lesson, London: Penguin.

Jewison, Norman (1967), In the Heat of the Night, screenplay: Stirling Silliphant.

Kane, Sarah (2001), Blasted (Complete Plays), London: Methuen.

Lorca, Federico Garcia (1999), Blood Wedding, Four Major Plays, Oxford: Oxford University Press.

McKee, Robert (1999), Story: Substance, Structure, Style, and the Principles of Screen Writing, London: Methuen.

McTiernan, John (1988), Die Hard, Screenplay: Jeb Stuart.

Minghella, Anthony (1989), Cigarettes and Chocolate, Best Radio Plays of 1988 London: Methuen/BBC Publications.

Nolan, Chris (2000), Memento, screenplay: Chris Nolan.

Pinter, Harold (1978), The Dumb Waiter, The Room, London: Methuen.

Ramis, Harold (1993), Groundhog Day, screenplay: Danny Rubin.

Resnais, Alain (1959), Hiroshima Mon Amour, screenplay: Marguerite Duras.

Scott, Ridley (1999), Thelma and Louise, screenplay: Callie Khouri.

Shakespeare, William, King Lear, Hamlet, Macbeth (Arden Editions) London: Methuen.

Thomson, George (1980), Aeschylus and Athens, London: Lawrence and Wishart.

Zvyagintsev, Andrei (2003), The Return, screenplay: Vladimir Moiseyenko, Alexander Novototsky.

Back to top

Radio Plays
Wine Making
Cosby Methodist Church
Gokart Racing
Links to other sites
Contact Us