There is a certain irony in the fact that the team responsible for writing, directing and producing 'Thunder Road', one of the BBC's flagship experiments in interactive drama, were not primarily interest in making an interactive drama. Both writer/director John Godber and director/producer Nick Copus openly admit to being lured to the project by the �200k budget and the prospect of being able to make a feature film by stealth, on their own terms.

"It was guerrilla film-making in two senses," claims Godber. "First of all, we were doing it on a low budget, with local actors, amateur performers and a tight production schedule. Secondly, the BBC thought we were creating an interactive experiment, when we were secretly making a proper film."

The BBC trials much of its interactive output in Hull, where there is near-universal access to interactive TV, thanks to local telecoms provider Kingston Communications, which rolled out it's dsl-based KIT iTV platform in the city in 1999. When the BBC's experimental drama department Fictionlab (now BBC Creative Research & Development) decided to commission a local interactive drama for viewers in the city, local resident Godber seemed like a natural choice as writer and director. The prolific playwright sets many of his plays in Hull and East Yorkshire, and has been Artistic Director of Hull Truck Theatre Company since 1984.

At first Godber was reluctant to get involved. Speaking at '21st Century Soap', an NMK talk at the TAPS London Writer's Festival in September 2003, he explained that he found TV to be "patronising and tediously slow", with schedules full of "comedies which aren't funny and dramas which aren't comic". Years ago he worked on Grange Hill (with Anthony Minghella, no less), and on early episodes of Brookside, and that had been enough for him. Recently, Godber also directed the movie version of his own play 'Up 'n' Under', but was frustrated by the dramatic compromises and casting decisions forced upon him by the studio. Such experiences only reinforced Godber's commitment to regional theatre, where he believes you find the real talent and highest quality material.

'Thunder Road' was conceived as a drama consisting of 30 three-minute episodes, broadcast every day for a month, and incorporating interactive elements. What eventually persuaded Godber to get involved was the realisation that, taken together, the 30 episodes could add up to become a 90-minute film. He was also drawn by the knowledge that he would be able to work in relative autonomy, and work with local talent, including amateur performers and professional stage actors ? that is, "it didn't have to have David Jason in it".

Because of the episodic structure of the concept, it made sense to model it on a soap opera. As 'Thunder Road' evolved it began to acquire 'soapy' traits: set in a pub/nightclub, centred on a landlord and landlady with marital problems, and all underpinned by a central storyline ? will the club stay open or be closed down? The project was in pre-production for five months, during which the script was constantly edited and refined.

For Nick Copus, the brevity of each episode posed a challenge, in that each instalment had to have its own mini narrative arc, rather like a commercial. When viewed as a continuous 90 minute sequence, this gives 'Thunder Road' a distinctive but unconventional rhythm, which perhaps detracts slightly from the drama's aspirations as a feature film.

To keep costs down, 'Thunder Road' was set and filmed in a genuine working men's club rather than on sets and a soundstage, and filmed on Sony 25P DVCAM, which resembles film stock but is much cheaper. The shoot took 17 days, with a single camera and a crew of (at most) 25, including 10 days with a steadicam operator.

The team also enlisted members of the local community, who enthusiastically volunteered as extras. In addition, all of the acts who appear on-screen in the fictional club are local singers and performers, some of them amateurs. The community element would prove to be a major factor in the success of the show, and come to define the nature of its interactivity.

Interactivity

With a writer/director and producer whose priority was to make a conventional, linear drama on the sly, Gabriel Gilson found himself under a lot of pressure as interactive producer, responsible for conceiving, designing and implementing the interactive elements of 'Thunder Road'.

On the KIT platform, episodes of the drama were downloaded via DSL onto the viewers' set-top boxes. These were then displayed on-screen as a combination of streaming video and HTML overlays, which were based on standard BBCi templates. The look and feel of Gilson's original interface designs was based on the concept of a 'virtual club', with features such as colour-coded ashtray as navigational icons, but early designs were rejected on the grounds that the film was overwhelmed by the surrounding graphics. The final version was cleaner, and consisted of a bar of icons over-layed onto a full-screen TV picture.

'Thunder Road' featured a range of interactive elements, including an 'answerphone' enabling viewers to leave voicemail for Malc (the landlord), and message boards which allowed fans to communicate with their favourite characters via the set-top box. Gilson responded to most of the messages himself 'in character', but consulted Godber to make sure he didn't reveal any crucial plot points.

Originally, there had been plans to film short 'drop-ins' to insert into the pre-recorded footage in response to audience votes and feedback, but in the end the viewers seemed to feel a strong enough connection to the show without this feature. 'Thunder Road' was, however, broadcast in pre- and post-watershed versions, depending on whether viewers downloaded the show before or after 10pm.

Perhaps the most popular features, however, were the extra scenes which allowed viewers to find out more about different characters after they'd watched each episode. During production, a second crew armed with a hand-held digital video camera followed the actors around and continued to film them, 'documentary style', in between scenes. Remaining in character, the actors improvised to camera, talked about what was going on in their fictional lives, and fleshed out their parts. This footage was then edited and made available to viewers of the interactive version of the show, rather like the extra features on a DVD.

When Godber and Copus saw these improvised scenes, they were pleasantly surprised to discover that many of them were at least as good as the 'proper' film. For Copus, it was evidence our culture is now so visually literate, and the technology so accessible, that relatively untrained or amateur film-makers can produce quality content. He argues that such low budget material will bring about "the end of elitism in TV", and provide the only way of meeting the demand for content in multi-channel broadcasting.

The interactive features on 'Thunder Road' were a success, helping to drive up audience share for BBC interactive services in Hull, and increasing the amount of interaction taking place in the average household. For Gilson, however, one of the most satisfying aspects of the project was that not all of the interaction was virtual.

In the final episode of the series the club was full to the brim with genuine customers, all of whom had come along to visit the club they'd been following on-screen for a month. The club was staffed by the actors, still in character, and footage from the evening was broadcast later that night on TV. For most of those who took part, it was the first time they'd been able to see themselves on TV, or seen Hull represented in such a positive, entertaining light. This, perhaps, is one of the most exciting possibilities that digital technology is bringing to television: the opportunity to make quality, local programming that genuinely connects with and involves communities in innovative ways.

Branching Narratives

One thing that 'Thunder Road' did not include was interactive elements that enabled the audience to choose alternative endings or otherwise make decisions that influenced how the plot developed. "Branching narratives are an old model, we have to get beyond that," said Gilson. On 'Thunder Road', branching narratives would have been too expensive and time-consuming to produce, and would have been impractical to implement given the limitations of the platform.

Branching narratives were also rejected for artistic and dramatic reasons. Nick Copus felt that allowing viewers to choose what happened to Thunder Road's characters would have made the project too much like a computer game and detracted from the narrative experience. It is an argument that has been explored time and again (not least in previous Creative Alchemy talks) since the prospect of interactive narratives in TV first became a possibility. Nowadays few who are working in this field seem to seriously believe that dramas with branching narratives are a realistic or desirable prospect. This is particularly true of those coming from a broadcasting, rather than a games, background.

If you're a writer, dramatist or film-maker, you are motivated by the idea that your skill, talent and sensibility are what makes your work relevant and worthwhile. Through the process of revision and editing, producing a film, novel or play is as much an act of exclusion as it is one of creation. To embark on a many-stranded, branching narrative is to go against this idea, and potentially to compromise your own vision.

The last word goes to John Godber: "what would Samuel Beckett have said if some bloke from Birmingham had phoned up and said that Godot should turn up at the start of the play? Or Shakespeare if his audience had said 'bring back Ophelia, we liked her'?".

The 21st Century Soap event was produced in association with Taps, Bafta (www.bafta.org) and Business Link for London (www.bl4london.com).

Take a look at the original event.