Writing  7 Ghosts

My latest play for BBC Radio 4 is a collection of ghost stories.

Some people have asked me how I write for radio, what the process is. I thought I might set it all out in some detail just so you can see how a project evolves.

The idea

It’s always seemed to me that the radio is a good medium for ghost stories. Often you listen to radio plays on your own, sometimes alone in a house; at other times, you might have your headphones on. Ghost stories in sound alone fill empty houses with noises, put raise unbidden spirits in your mind.

We first pitched this, I think, four years ago in 2019. At that point, the BBC were very keen on the idea but (from memory) they had another ghost project already commissioned so they asked us to pitch it again the following year; the next year, they bought You & Me instead, but implored us to pitch it again the year after; the next year, true to their word, they bought it.

The pitch

This picaresque journey of rejection and deferral is not unusual. In this instance, it was very much for the good of the project. Initially, the idea was a patchwork of ghost stories. This is one version of the initial pitch:

7 Ghosts

IDEA FOR AN AFTERNOON DRAMA.

The play has a cast of six and it is about seven ghosts.

The seven ghosts will all be of different kinds. Some will haunt a place. Some will intervene in the world of the living. Some will be unable to make contact and simply watch. Some we will be with as they become a ghost. We may also hear the death of a ghost. Some will be famous, some will not, Some will be old and some young. Some women, some men. One of the ghosts does not speak.

The play will continue the experiment of My Life is a Series of People Saying Goodbye. There will be a fluidity in the relationship between the scenes. The actors playing the seven ghosts will also play the living for the other ghosts. We will pass from scene to scene the way a ghost passes through a wall.

We will have seven lives and seven deaths to hear about.

The play will be about death and about love and the edges between our lives. It will sharply juxtapose stories and tones, sometimes funny, sometimes very melancholy, but it will feel continuous and fluid and complete.

The idea at this stage is largely formal - that is, I have ideas about the form but am quite vague about the content. The earlier play I mention, My Life is a Series of People Saying Goodbye, was written a decade ago and it told ten or so separate stories that blended into one another in an unusual way. My sense is that the current Radio 4 Drama Commissioner, Alison Hindell, is unusual in that she likes my writing because of its formal experimentation and not, as I have felt with other Commissioners, despite it. It seemed that ghosts could pass between lives and stories with unusual fluidity.

Why 7? I don’t know. That was plucked from the air. 7 is a good awkward number; a prime; the ideal size for a theatrical chorus according to Jacques Lecoq; the perfect size for a decision-making meeting; a lucky number in some cultures; unlucky in others. And, more prosaically, I thought you could probably tell 7 stories in the 44 minutes of an afternoon drama; it’s around 6 minutes per story.

Why did I not worry about the individual stories at this point? I think, to be honest, I was just pretty confident that I’d think of story ideas. In radio, unlike TV and film, you’re not paid to come up with treatments for story ideas, so I tend to be reluctant to work out things in too much detail. Also, I hate foreclosing on the creative process. It’s only with commission in hand, deadline looming, and I am fully immersed in the ideas that I really trust myself to make decisions about story, character, scene. Setting out a story a year ahead feels odd, like those staff Christmas meals where you have to decide what food you want to eat a month in advance. From experience, I know I can trust that ideas just will come when I need them and the blank page now feels exciting rather than terrifying.

The idea changes

Why was the delay in getting 7 Ghosts commissioned an opportunity? Because in that four year period the world changed. In March 2020, long before the play had been commissioned. Britain went into lockdown. Suddenly half a dozen projects I had were cancelled. Also, my son’s nursery closed its doors. So I was stuck at home, partly looking after our boy, but also with a sudden access of free time. I’d been meaning to read David Olusoga’s Black and British (2017), a new history of black people in Britain but hadn’t got round to it; this seemed like a chance. I got the audio book version, read by Kobna Holdbrook-Smith and over the course of a week or two listened to the 24 hours of the book, dazzled by what Olusoga had put together, the many stories I was ashamed to realise I didn’t know, and entirely seduced by the reader’s voice (more on that later). In particular, I was struck by one of his central claims (which I later realised was a widespread observation, but Olusoga articulates it particularly well) that we have a curious asymmetry in our cultural memory of the British slave trade: we commemorate and celebrate the ending of slavery, but have a far vaguer shared view of the actual practice of it.

And then a month or so later, George Floyd was murdered by a police officer in full public view, which accelerated the Black Lives Matter protests across the world, which included the moment in June that year when crowds pulled down the statue of slaver Edward Colston in Bristol (see picture). It truly felt and still feels like this was a cultural watershed, where the history and legacy of slavery, systematic and structural racism, racial violence and inequality became impossible for the culture as a whole to minimise or ignore. (That said, the behaviour of right-wing activists, in parliament and beyond, in ‘defence’ of statues was precisely about doing that.)

And at some point it seemed to me that a ghost story might be an interesting way of thinking about the legacies of slavery, slavery haunting our culture like ghosts haunting a house, maybe, like Scrooge’s ghosts, urging us to change, or like the ghost of Hamlet’s father, demanding vengeance.

Research & Notes

So when the play was commissioned the following year, I had a sense of form and some kind of thematic content. I read some more things: Sathnam Sanghera’s Empireland: How Imperialism Has Shaped Modern Britain (2021), Padraic X Scanlon’s Slave Empire: How Slavery Built Modern Britain (2022), and Michael Taylor’s The Interest: How the British Establishment Resisted the Abolition of Slavery (2021) and a collection of first-person slave narratives, the most powerful of which was the life story of Olaudah Equiano, a slave under British rule who achieved his freedom, wrote a book about his life and became an anti-slavery speaker and icon. His book, written in flamboyantly elegant English, tells his unusual life story - he was a sailor for most of his enslaved life, rather than, say, a plantation worker - was a fascinating immersion into a life, a pattern of thought, a verbal style. And I read other things too, particularly by the literary critic Henry Louis Gates, whose alertness to the complexity of these narratives and black writing generally encouraged a sense that the layers of the different ghost stories could rub against each other in interesting ways.

Through the reading, the character of the house in which all the ghosts would have lived and died became clearer. It would have been built for a slave-owner in the 1780s and have been occupied by a series of owners right up to the present day. This meant my reading and thinking could start to populate the house with ghosts from different moments in the house’s history; in fact, at one point I wrote a chronology of that history, its conversions, its handovers, its changes of use. That chronology would have to adapt as I started to flesh out the characters and the characters would change as I made more sense of the chronology. Here’s a glimpse of the chronology in mid-August:

Several of the characters are fixed. One is fixed but her name will later change. The history continues to change after this. Much of the history will not be part of the play, just my private background information (a property speculator dividing the house in two after 1916 for instance). I got myself very confused over the chronology of the early nineteenth century (as you can see from this much-amended family tree and timeline) but through it characters emerged and with them flashes of dialogue, events, motifs.

The idea emerges clearly that the house is built on profits from slavery. Its checkered onward history is haunted by slavery, whether directly in the presence of a slave ghost, in the guilt of its later white inhabitants, or just in motifs like cotton, tobacco. And also water. In my book on Playwriting for the National Theatre, I recommend a ‘Golden Hour’; that is, a scheduled hour where you will work on ONE thing from your play. It might be you’ll spend an hour on the characters or the structure or thinking of titles or something else. I did this a lot with this play. At some point, when developing the story of the (in some ways) central ghost, I imagined him miraculously walking across the ocean. I thought that he might bring water with him and spent a useful golden hour writing down ways a house might obsessively contain water.

Structure & Shape

I finished the reading and thinking and note taking by 19 August 2022 and went through all my notes sticking every individual idea for a moment, a scene, a character, an event, a line of dialogue onto separate index cards, vaguely (but probably not consistently) colour-coded. As I write out the cards, further ideas inevitably occur and I write those too. A rule: I am usually standing up when I look at the final form of the cards so I don’t want to be writing huge amounts of information on the cards; they should be legible from a distance. An idea that needs 30 words is probably two ideas and needs to be split on different cards. I tend to use Sharpies so I have big clear cards. (Sometimes I have a string of dialogue and then I’ll write in big letters a summary of what that dialogue is doing or maybe, if the dialogue is clear in. my mind, a keyword. Occasionally, I’ll write the dialogue down on the back of the card. Whatever is most helpful.)

I then laid them out on the floor. They go down on two axes. Left to right is broadly the order of the whole play; top to bottom are the order of individual sequences (sometimes that means scenes).

That’s a key process: I have a stack of maybe 100 cards, sometimes more sometimes less, and they all have to go down so I am forced to group them, which is where I realise that this character could meet this character in that scene and that’s where this event might happen. So this is a point where connections are being made and the whole play is becoming a coherent, knitted-together, consistent thing. I take my time with this; it pays dividends later. Putting the cards down might take a couple of hours, maybe more, but it’s an entirely creative process and it’s worth doing. Here’s what the first cards-down process looked like:

Here for instance is the second sequence, in which two ghosts caustically observe a seance. The cards written in black are facts of the scene (character information, locations, etc.), blue is what is happening, purple is dialogue, green seems to be conceptual (but inconsistent). It’s not complete: often I have a sense of what the scene will feel like that I don’t bother put on a card - in this case, there’s some humour in the ghosts’ contempt for the spiritualist and then some rivalry between themselves.

I might put the cards down a couple of times, as I did here; putting the cards down is a very creative moment, because it’s where the thing is structuring and you sense how the ideas, events and characters will develop, how the story will reveal itself to the audience. I often write new cards as I’m constructing the map of the play because putting two cards next to each other will dialectically generate a third idea and so on. This is one of the most difficult, important, and enjoyable bits of writing the play.

First draft

What is a first draft? It depends on what has gone before. The play map (the cards) will give be the shape of the play. That is sometimes the structure, but not always. Sometimes it’s just saying, this will happen then this will happen then this will happen. That isn’t a structure: in a play like Cavalry I had no play map: I just knew what the feel of it would be and I knew there was a vast revelation (ho ho) of what was really going on and it would build and build but I didn’t know how we’d get there and what the overall structure would be that would make sense of the whole project - that emerged in the writing. In a play like You & Me, which was a kind of thriller, and required clues, turns, revelations, the play map is much more structural and I trust that the texture of the characters will come out in the writing.

In this instance, it’s a bit of both: there are some revelations and I knew about them and knew when and how they’d come. But equally I was very clear that there would be textures and transitions that I didn’t want to plan. With each project there are things I must not know before I start writing, because they can only be accessed through writing. Very rarely do I start with nothing

It’s important (to me, anyway) that this play map doesn’t tell me everything. Though I might know some things about the characters, I don’t really know what they’re like: that will only emerge in the writing, when they meet, talk, do stuff. Similarly, while there may be key phrases and scraps of dialogue, the dialogue only really emerges in the writing and it can often go somewhere entirely different. Bits of dialogue that seemed terribly exciting in the notebook have no place in the scene (there are a couple of instances in the cards above). Exactly how we move from scene to scene, how each scene will crest and move on, that’s all in the writing. Writing the play is still a process of discovery; if it weren’t, it wouldn’t be worth doing, at least that how it feels to me.

Once I’m happy with that. I’m ready to write. I usually start from the beginning and work through the play. I guess this is because I’ve already got a sense of the whole and resources to support each moment, so can afford to wait till I get to a particular scene, even if I’m itching to write it. Also, for me, the flow of the play is easier to feel for if I write it chronologically.

So I wrote the first draft over the next month and here it is. The way I write a play, it’s quite hard to control how long the draft will be. I’ve got a rough sense of the length, but really I just need to write each scene and speech as long as it needs to be and at the end see where I’ve got. This is less of an issue on stage where it matters less whether a play is ten minutes longer or shorter, but for this play - a BBC Radio afternoon drama - it has to be 44 minutes and no more because the 3pm news summary is happening after that. So this is too long. It’s just over 9000 words. For a 44 minute play you should be looking at around 7500 words. (This isn’t a fixed rule: if you’re writing very short lines of dialogue it’ll need to be more; if you write extremely elaborate stage directions you can probably go over that, but in my view extremely elaborate stage directions are to be avoided anyway.) But anyway, this play is probably ten minutes too long, so that’s problem one.

Second draft

What’s a second draft? For me, it’s usually the unsatisfactory but necessary process of solving the problems of the first draft. I don’t much enjoy the second draft, to be honest. I hope this will change, but mostly I find it the most clumsy moment. I usually end up with a draft that has, technically, solved the problems of the first draft but not in a particularly pleasing way.

How do I identify the problem with the first draft? I like to get away from a draft for a bit if I can so that I can come back to it relatively fresh. I was aiming for mid-November for a final draft of the play so assuming there will be two more drafts, I can’t really do that for very long but I think I got a couple of weeks where I didn’t look at it. I also sent it to two brilliant people: Polly Thomas, my producer/director, with whom I’ve worked in radio drama almost exclusively for 25 years and is my greatest collaborator; and Olly Emanuel, a fellow radio writer, an excellent writer for radio (and the stage), and someone that I’ve known I think for about 15 years but with whom I worked on the massive Zola adaptation and someone I feel a real creative kinship with. The three of us all read the first draft and all have thoughts. The consensus seems to be (a) it seems to be structurally sound; (b) some of the transitions, particularly in the second half, need work; (c) Ejikeme needs to be in earlier; (d) the exact nature of the Sir William/Lord Havers dispute needs clarifying; (e) we probably need more of a sense of the house; (f) the final scene needs to be a bit more frightening and less descriptive. And a few more things like that.

These are all very helpful notes. They are precise and practical. Sometimes people give notes like ‘this scene needs something else’ or ‘this character needs more oomf’. These notes are lazy and stupid, even if they come from some genuine perception of a problem in the play. I need a clear opinion about something specific. If I strongly disagree with a note, at least I understand it and it might be that I can see a way of fixing the problem in a completely different way. To offer a purely theoretical example, Scene 2 might be confusing, say, but rather than rewrite scene 2, as suggested, I realise I need to approach Scene 1 differently.

The other thing to say about rewriting is this: simple is good. That is, rather than solve a problem with something big, it’s very helpful if you can solve a problem with a few words, or, even better, by cutting - particularly because the draft is too long. After the second draft, Polly says that we might need to establish slightly more clearly that David and Sarah are Black (this is radio after all). This could be done with an impassioned discussion between them about the Black Lives Matter movement and how important it has been to their own lives. Or I can drop in a reference to them having attended a Black Voices Forum and trust that alongside the vocal quality of the actors and the wider context of the play, this will be enough for audiences to understand what they’re hearing.

In this draft, I’m looking for cuts. Polly, very sensibly, notes that there are two ghost stories told by David and Elliott in the middle of the play; they are both long so maybe one of then could go. I’m resistant to that, because I like the idea of two people, on a cold winter night, by a crackling fire, exchanging ghost stories. It feels like the moment where we deliver on the title: two proper ghost stories (that turn out to be three, though we don’t know that yet). But its clear that the first story could be trimmed extensively; he’s not telling the story, he’s describing a story he’s read, so we can shortcut that. Some of the scenes can trim some fat. There’s an awkward transition (around p.28) where I bring back some earlier characters briefly (and intend to bring back more - the scene is incomplete); it’s there for two reasons: one, because there are two scenes with David at different times so I needed to cut away from him, as it were, to get back to him in a different mood and time; second, I wanted to bring Ejikeme in earlier and this felt like a moment to do that. But it’s a placeholder and doesn’t really solve any of the problems. I realise it’s easier to make the two David scenes flow into one another without changing time and solve the Ejikeme issue in a different way.

So over the next month I write the second draft. Here it is. It’s much tightened (down to only a few words over 7500). The basic architecture hasn’t changed, except for removing that awkward transition scene and tying the two together. There’s lots of tweaking of dialogue, some smoothing out of sequences, some building of motifs. It’s not a huge rewrite. To be honest, the index card play map process means that I’ve usually got a shape I am pretty confident will work. I try to do the structural work there rather than have to do that kind of major rebuilding in the drafts. Lots of writers find the play through drafts - and that works for them - but I try to have a strong sense of the play’s architecture before I write.

Third draft

What is a third draft? The second draft is a bit crude. Some of the cuts are a bit raw. Some of the solutions to problems are a bit rough. It’s like you’re repairing structural weaknesses in a house by putting up the temporary supports. They’ve made the house structurally sound, but you don’t want to live in it that way. The third draft is where you replace the crude supports with something more elegant that blends into the form of the house.

I sometimes wonder why I don’t go straight from the crude rewrite to the elegant rewrite and call all of that the second draft. But, for some reason, I seem to need someone to see the second draft and approve in principle what I‘m doing. Polly’s particularly good at reading and not panicking, seeing where it’s going rather than fixating on where it is. And Polly thinks it’s going in the right direction, which is a relief.. She has some good notes and thoughts. David should respond more angrily to the revelation about the house’s origins, she says, which I’d also thought. She lists a few more things. These aren’t structural: for me, I get worried if there are fundamental structural things to do after the second draft.

To redraft, I print the play off and annotate it. This isn’t a great process environmentally, but it still feels easier to me to write notes on a script than to annotate a PDF or something similar. The play feels more finite on paper, easier to amend and rework. Scribbling on paper is a freer exercise and - crucially, for some reason - different from the typing that I’m annotating. I make two kinds of annotation: (a) tinkering with dialogue, and (b) specific instructions about bigger things that need to change, preferably with some ideas about how to do that, definitely with details of where. The key is that both of these types of annotation are actionable: specific, precise, concrete. (Katie Mitchell uses these terms a lot and, on reflection, this comes from talking to her for a couple of decades.)

And here it is, the third draft, finished in November 2022. Mostly this redraft is nuancing, finessing, refining. Some of the characters’ psychology needed clarifying and strengthening or simplifying. Some moves in the dialogue need to be sharper. One issue was that Ejikeme was not built enough into the play. In the first two drafts, I had Ejikeme come to the door and meet David, but this doesn’t really work - too many people at the door, too many ghosts for David to cope with, too much implausibility. I have removed that and drawn Ejieme forward through the play. I decided that he should be part of the transitions, the layers, the water. I gave him a spectral appearance in the first Sir William/Lord Havers scene too. I also think I’ve picked up from Olaudah Equiano a too-deferential tone for Ejikeme that lasts almost to the end, so I decide the ending should be much more of a warning. I think it makes sense to bring back some of the other Black voices from the play (this was Polly’s suggestion too).

Happily, Polly thinks this is ready for production. We take this into studio in December and record it. We’re lucky with our cast; they are all wonderful. Fanta Barrie plays Florence and Isabelle; Don Gilet is Ejikeme and Lord Havers; Sarah Berger takes on Charlotte and Lady Caroline; Mac Runham is Robert and Elliott; Sir William is played by David Annen (who has been in a few of my plays); Ani Nelson plays Sarah and, for the part of David, we get the man whose voice took me into the world of this play, Kobna Holdbrook-Smith. Polly is directing and Eloise Whitmore will be sound designer. There will be original music by Jeremy Warmsley.

We record over two days. Radio drama is a very quick process (actors used to stage or tv are often shocked). A one-hour play will likely be recorded in two days, maybe three. A 45-minute play, like 7 Ghosts, Is maximum two days; in fact we wrap up around lunchtime on the second day; this is because this play is largely two-handers, which are simpler to record, and has some substantial sections that need to be created in the edit.

The speed of recording has some implications for the writer: first, your play really does need to be ready - there’s no time to fix a script in recording; second, your play should give lots for your actors to get their teeth in to, but nothing that will require a four-week rehearsal period to solve. I like to think this doesn’t mean the play must be simpler, but it means you don’t have the luxury of deep ambiguities - actors must be able to make decisions fairly quickly. There is a read-through at the beginning of the recording period, but then you go straight into recording. There are no rehearsals: the first take is your rehearsal. Generally, you will do 3-4 takes. Occasionally, if a scene is particularly delicate, maybe 5 or 6. But if you go much over that, you’ve got problems. Some producers create scenes out of snips of different takes. Polly prefer to choose a single take, because, even if it has moments that were better elsewhere, you’re getting the actors’ continuous work and it feels more authentic somehow. I think she’s right and sometimes I’ve heard radio plays where there’s just something psychologically awkward and discontinuous in it and I realise it’s been patchworked together from numerous takes.

In the recording there are some very minor rewrites, mostly a word here and a word there. We finesse one further transition - from the Lady Caroline scene - which I’ve silently folded into this third draft.

The edit

I have no idea about this. The writer is never in the edit and that’s fine. It’s the producer/director’s job, along with the editor/sound designer (in this case the excellent Eloise Whitmore). While, very occasionally, there are decisions I’d have like to see done differently, overwhelmingly I know that I just couldn’t have put the thing together with the skill, insight and invention that Polly and Eloise do it. It’s also a great privilege to leave the recording and then wait for an MP3 to turn up in my email and hear it fresh, like any other listener.