Why Interactive Films Will Never Work

I believe that enforced conscious interactivity between film and audience (e.g. pressing buttons) reduces the viewer’s sense of being drawn into a narrative — of experiencing the narrative from within. I’ll examine this from the point of view of database cinema, spectatorship and suture theory.

Lev Manovich was the first to analyse how the concept of database influences cinema and narrative (Manovich 1998). A database is a set of objects used in information technology — I spent some time working as a database programmer. The structure of a database conforms to two principles: (i) data is broken down into the smallest possible re-useable types; (ii) data is labelled and related to other data so it is not abstract but relational. Once this is achieved, an incredible flexibility emerges that was not possible in the analogue days of filing cabinets.

Lev Manovich

For example, suppose that a production office enters all of their important data for a planned movie shoot into a database format. This data could include all cast and crew, their availability, their per day costs, their travel availability, locations, cost of locations per day, equipment, the rental cost of equipment and so forth. It is then possible to make what is known as a database “query”. For example: “all crew available in this date range”, “all cast available in this date range”, “total cost of cast and crew over this period”, and so forth. Once the data has been entered and appropriately labelled and linked, then all that is required is a flexible interface.

In my work as a database programmer I used SQL (pronounced sequel) (Petkovic 2005) — the Symbolic Query Language — to query trades on the German stock exchange. Suppose the data was simply stored as a long list of trades containing date, time, amount, traders, and stock on a piece of printed-out paper. Then re-using the data to develop new perspectives would be a lengthy process. However, suppose that all trades were stored in a relational database. You could enter SQL queries such as “SELECT * FROM Trades where Stock = ‘SIE’;”. This would return all trades done in the stock Siemens AG. Whereas “SELECT MIN(Price) FROM Trades where Stock = ‘SIE’;” would return the lowest priced trade in Siemens AG stock in the database. Both of these are from the same database but are very different outputs with different functions.

Manovich (1998) makes the point that the result of the production phase of movie making generates a database consisting of footage and sound and any other production-generated assets and notes. This database may sometimes be analogue, and thus not strictly a database in the IT sense. But just like the database of trades — which are not ordered in time but only labelled in time — the results of film production are normally shot out of time-order compared to the scripted narrative. Then, as Manovich says, “during editing the editor constructs a film narrative out of this database, creating a unique trajectory through the conceptual space of all possible films which could have been constructed.” The editor performs a query on the shooting database. Manovich’s focus is on how new types of narratives could be created if we think of film from a database point of view, rather than a fixed and ordered point of view. For example, he considers Peter Greenaway’s “The Draughtman’s Contract” (Greenaway 1982) to be influenced by database narrative, being broken down into the sections of the multiple drafts done by the protagonist — and these not in any particular order. Manovich extends his argument by observing that Greenaway went on later to create installations and exhibitions. Whereas traditional narrative puts database elements in a specific order, moving elements into museums spatialises them into a database form of narrative, reducing temporality.

Location for The Draughtman’s Contract

My interest in this article is on another effect of database narrative. The focus is Manovich’s quote regarding the editor selecting one of a number of possible routes through the production asset database. Thinking of the editor as an interface to the database also opens up the possibility of involving the audience in the “query” process that creates the narrative. Thus the editor ceases to be the definitive interface. One instantiation of this possibility is interactive cinema that involves forced conscious decisions by the audience during exhibition. Filmmakers have been intuitively aware of the possibility for at least 60 years. It is perhaps a more surface-level implication of Manovich’s ideas, but nonetheless has important implications for the media of digital streaming.

Currently digital streaming is almost exclusively limited to allowing the user to query at the level of which film to view, not what happens in the film. This will surely not remain the case. It was not long ago that broadcasters thought the only interface to their library was the TV channel with fixed schedules. Netflix quickly realised that the move from analogue to digital allowed the audience to create their own interface to query the library — leading to such phenomena as binge-viewing. A short history of interactive cinema should be read here before continuing.

My interest in interactive cinema came about in 2013 when I premiered a short film called ‘many worlds’ (Kirke et al. 2018). I do not consider it interactive film in the normal sense because it does not require conscious decision making from the audience. However it still has multiple possible plots (routes through the production asset database). Four members of the audience wore a biosensor each (brainwave, perspiration, muscle tension and heart rate). These values were used to determine implicitly using an automated computer AI (i.e. without an explicit decision from the audience) the story direction at two decision points in the narrative. This was actually the first such film shown to the public in a cinema that did this. A couple of years earlier, “Unsound” was premiered at SXSW (Bristow 2011) in which perspiration monitors were used to adjust a soundtrack in real-time.

While developing ‘many worlds’, it was my implicit assumption and intuitive understanding that films which were consciously interactive (such as Bandersnatch) upset the normal process of cinematic engagement. However I had no theoretical understanding to underpin this intuitive belief. This essay is an opportunity for me to examine possible theoretical underpinnings for how consciously interactive cinema “upsets” the normal process of viewing.

A “Scene” from Dear Esther

First, consider an extreme form of interactive narrative: narrative games. One of the most famous is probably ‘Dear Esther’ (Pinchbeck 2008). Dear Esther differs from an animated film by requiring the viewer to motivate all movement (forwards, backwards, left, right). However — unlike in traditional computer games — there are no puzzles to be solved, or enemies to be dodged. And in fact the main cumulative effect of playing is not the defeating of challenges or discovering of required secrets, but the triggering of narrative voice-over when the viewer reaches certain areas of the “game” (travelling through a windswept island off the UK). Dear Esther is a useful place to move the discussion of convergence and interactivity into the theory realm. Specifically: the majority of people who start “watching” Dear Esther are doing so in the context of playing a game, and thus of constant enforced interactivity. Their psychological priming is entirely different to that of a person visiting the cinema, or watching a YouTube video.

All key analyses of cinema spectatorship — whether the original Freudian/Lacanian psychoanalytical versions, or Mulvey’s (Mulvey 1989) and Diawara’s (Diawara 1993) re-evaluations — focus on an essentially unconscious viewing mode. I use the word “unconscious” as distinct to “passive” (and not in the psychological sense) because of the theoretical implications of the words active and passive. The viewer is still active in the sense that they bring all the elements of spectatorship to bear on what they see: the social, cultural, private and desiring self (Moreno-Malgar 2013). However they do not do this by consciously choosing at every moment which attitudes to take in each of these four elements. They do not have a series of control sliders for each of the four selves that allow them to adjust the film’s narrative based on their choices.

Laura Mulvey

There is a complex and often unconscious interplay between the selves that adjusts the subjective experience of the same sensory data in different ways for different spectators at different times. Interactive movies and narrative “games” change this experience by changing the sensory data based on a viewer’s explicit conscious control key presses — in the spirit of database cinema. Each new experience of the underlying film is a new database query spread across a temporal interaction. A form of “Flow” (Csikszentmihalyi 1990) can be created in the playing of narrative games (Jegers 2007). But this is an entirely different context to the spectatorship elements discussed above.

An example of how different spectatorship is compared to conscious choice of narrative is supported by Suture theory. Oudart (1977) was a key developer of Suture theory in film. Amongst other things, this theory covers how we are drawn into the world of the film, as if we experience it from within the film. Oudart used the shot / reverse-shot process to give insight into how Suture works in film Any film frame (A) implies an observer — a “consciousness” — who is looking at A. However the viewer does not know who this observer is — Oudart calls them the “Absent One”. The effect of the reverse shot is to fill in the Absent One — thus comforting the viewer, answering the question. This second shot sutures the wound opened by the implication of the Absent One in the first shot. Of course films do not consist of a series of shots and reverse shots. However the concept of suture has been expanded beyond this. For example Heath (1981) examines a montage which he describes as being sutured by the music and the sound design. Although not all theorists agree on the precise process, many would agree that through the continued process of the suture the viewer is drawn into the film and experiences it from “within”.


Suture theory is based on the psychological theories of Lacan. As mentioned, Lacan’s theories have been re-evaluated from a feministic and race perspective. However they are a useful starting point for considering the effects of consciously interactive cinema. This interactive cinema requires the audience member to become conscious of a decision to be made. They have to step out of the largely unconscious process of suture and look at the film as a series of conscious decisions rather than unconscious experiences. The constant rhythms of suturing are interrupted, thus interrupting the sense of experiencing the film from “the inside”.

Consider the traditional suture process: the viewer sees the first shot, which implies the “Absent One” and this tension is then resolved by the second shot. Suppose after the thirtieth pair of first shot / second shots, a message comes up on the screen asking the viewer a question and the film is paused until the question is answered. The viewer will be taken out of the rhythms of the Absent One and the resolution. They will become aware of somebody asking a question. Not only will the subconscious rhythms be interrupted, but a conscious decision process will be triggered. (The same argument applies to Heath’s example of the montage sutured by music and sound design.) Furthermore, this process of enforced conscious decisions is usually much slower than that found in a game. Jegers (2007) argues that the Flow theory of Csikszentmihalyi (1990) explains how users become engaged in a game through repeated frequent interaction. But in interactive cinema, two possible processes of deep engagement: suture and flow, are both compromised. Suture, by the breaking of the shot pair rhythm — and Flow, by the fact that the questions happen so infrequently (usually every few minutes at the most). The viewer is dragged between a possible process of suture and a possible process of flow, but not allowed to remain in either.

If the process that enables viewers to see a film from “the inside” is interrupted, a new experience is created. I do not claim this experience is of less or greater value — merely that it differs. The promise of interactive cinema is often marketed as “Cinema plus”, i.e. cinema but with an improvement. However it is not cinema in the suture sense. Nor is it gaming in the flow sense. More broadly, database cinema is capable of enabling both sutured experiences and flow experiences — depending on the interface chosen and the natures of the database query. But interactive cinema, as exemplified in Bandersnatch, does neither.

In a sense much of interactive cinema so far has been comparable to Duchamp’s ‘Fountain’ (Camfield 1989), which questioned the nature of art and art galleries through presenting a urinal as a gallery artwork. Interactive films shown in cinemas have mostly been getting the audience to ask themselves the question: what is it like to watch interactive films? In fact Bandersnatch — which I consider to be the most successful of the consciously interactive films — is successful because it explicitly asks this question as part of its plot. The plot discusses the existence of Netflix and includes a database narrative designer as a character at the end. Hence the interruption process becomes part of a broader meta-story. There is nothing artistically wrong with continuing to ask these questions through consciously interactive cinema. But we will never develop a way of discussing our answers unless we accept that interactive cinema is neither this nor that.

Bejan, R. (1992). I’m Your Man. New York: DVD International.

Bristow, N. (2011). Unsound. Belfast: Filmtrip.

Csikszentmihalyi, M. (1990) Flow: The Psychology of Optimal Experience. New York: Harper Perennial.

Camfield, W. A. (1989). Marcel Duchamp: Fountain (Vol. 137). Houston, TX: Houston Fine Art Press.

Cincera, R. (1967). Kinoautomat. Montreal: Czechoslovak Pavilion at the World Exposition.

Damiani, J. (2019). Black Mirror: Bandersnatch could become Netflix’s secret marketing weapon. The Verge, January 2 2009 [Online[. Available at: https://www.theverge.com/2019/1/2/18165182/black-mirror-bandersnatch-netflix-interactive-strategy-marketing [Accessed Sat 30 May].

Diawara, M. (1993). Black Spectatorship: Problems of Identification. Black American Cinema, 211.

Greenaway, P. (1982). The Draughtman’s Contract. London: Curzon Artificial Eye.

Heath, S. (1981). On suture. In Questions of cinema. London: Palgrave, p. 76–112.

Jegers, K. (2007). Pervasive game flow: understanding player enjoyment in pervasive gaming. Computers in Entertainment (CIE), 5(1).

Kirke, A., Williams, D., Miranda, E., Bluglass, A., Whyte, C., Pruthi, R., and Eccleston, A. (2018). Unconsciously interactive Films in a cinema environment — a demonstrative case study. Digital Creativity, 29(2–3), p.165–181.

Lund, L., and Lund, C. (2008). The Outbreak. Los Angeles: Silktricky LLC.

Manovich, L. (1998). Database as a symbolic form. In Parry, R. (ed) Museums in a Digital Age, New York: Routledge, p.64–72.

Manovich, L. (2013). Software takes command (Vol. 5). London: Bloomsbury.

Moreno-Malgar, I. (2013). FM4 Spectatorship. Slideshare.net, May 2013 [Online]. Available at: https://www.slideshare.net/ianmorenomelgar/a2-film-studies-fm4-spectatorship [Accessed: 29 May 2020].

Mulvey, L. (1989). Visual pleasure and narrative cinema. In Visual and other pleasures. Palgrave Macmillan: London, p.14–26.

Oudart, J. P. (1977). Dossier suture: Cinema and suture. Screen, 18(4), p.35–47.

Petkovic, D. (2008). Microsoft SQL server 2005: A beginner’s guide. New York: McGraw-Hill.

Pinchbeck, D. (2008). Dear Esther: an interactive ghost story built using the Source engine. In Joint International Conference on Interactive Digital Storytelling ), Berlin: Springer, p. 51–54.

Alexis Kirke is a screenwriter and quantum/AI programmer. He has PhDs from an arts faculty and from a science faculty. http://www.alexiskirke.com

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store