Looking Behind the Façade
Playing and Performing an Interactive Drama

by Jörgen Schäfer

Looking Behind the Façade: Playing and Performing an Interactive Drama

Following the ongoing debates between ‘ludologists’ and ‘narratologists’, the "interactive drama" Façade is apparently a response to widespread unease with mainstream computer games. By balancing between features of interactivity and (neo-)Aristotelian theory of drama, the developers Michael Mateas and Andrew Stern aim at enabling hybrid aesthetic experiences that combine elements of gameplay and performance. My paper explores how digital media require hybridizations of literary genres as well as reconfigurations of the complex interplay of human and non-human ‘actors’ – and it tries to point at both the opportunities and problems of these hybrid forms from the perspective of literary and performance studies.

Façade by Michael Mateas and Andrew Stern apparently is a response to widespread unease with mainstream computer games. It aims at creating “interactive experiences about human relationships” by providing “an artificial intelligence system that uses knowledge about how stories are structured to construct new story-like experiences in response to the player’s moment-by-moment, real-time interaction.”[1] Thereby Mateas and Stern’s project is counter to some very fundamental objections that have been brought up against literature in computer-aided media, which, according to its critics, does not live up to man’s deep-rooted desire of mutually telling fictional stories or performing dramatic plays since, of all things, interactivity has become the biggest obstacle for successful literary communication.[2]

With these reservations in mind, I prefer to follow a different approach that tries to identify both common ground and differences of genres and media adaptations. Given that a migration of literary forms into computer-aided media apparently is taking place, digital literature must contain invariant structures of repetition that only—in spite of any differences caused by distinct media of production, transmission and reception—enable us to talk of “literature” as a single field. My main assumption is that for reading and analyzing digital literature as “literature” the semantics of literary concepts is to be more durable than the pragmatics of communicative acts. Even if we aim at developing new and more adequate methods, terminologies or categories, we should not disregard the literariness of our subject matter. Therefore we inevitably have to start our thinking from those concepts that have been developed in literary studies during a long period. Among those concepts—this is my second proposition—literary genres still play an important role since they reflect core aspects of literariness.

In the following précis, I only give a brief outline of the theoretical approach that is to provide the basis for my close reading of Façade in the full conference paper.

Rethinking Genre Theory: Games, Narrative, and Drama

Drama by definition is about a conflict between characters that is not narrated by a narrator but enacted as present action, which the audience is witnessing. It carries and conveys the plot through dialogue and—when performed—also through gesture and facial expression; and it is not suited primarily for silent reading but for theatrical performance in real time. Mateas and Stern propose a theory of interactive drama that explicitly adopts the principles of classical theatrical drama, which yet are to be “modified to address the interactivity added by player agency.”[3] This, however, has significant effect on the motivational logic of a drama as it is represented in the famous Aristotelian model. If the dramatic logic is being changed to a game logic, the actions of characters in the fictional world inevitably become mere options; their decisions then get independent from motivational logics and instead are at the player-recipient’s disposal.[4] Therefore the dramatic logic, which has always been dependent on genre structures and genre conventions, is under a sort of collaborative agency of author(s), player(s) and media system(s). For at least limiting the disponibility of the motivational logic and thus integrating a sort of tension arc into interactive drama, according to Mateas’ and Stern’s approach, additional formal constraints are required in order to frame the player’s actions with a dramatic logic.[5]

For securing the player’s influence on the behavior of a character, however, the system has to offer two different forms of agency: On the one hand, the player needs to have “local agency” with an impact on the emergence of the story in one particular game level. The main problem that has to be solved if Façade is to provide a motivational logic of the characters, is the question of “global agency,” i.e., the player’s moves have to be dramatically motivated and his actions need to have an effect on the overall story.

Intertextuality and Genre Evolution: Marital Drama from Stage to Interactive Media

It is against this theoretical and historical background that I am interested in why Mateas and Stern do classify Façade as “drama” and how they justify this classification. Therefore it is necessary to introduce a model that enables us to elaborate both the structural differences and the common subject matter of Façade and “traditional” drama. According to Aristotle, a dramatic action implies personal agents “who must necessarily have their distinctive qualities both of character and thought; since it is from these that we ascribe certain qualities to their actions, and in virtue of these that they all succeed or fail.”[6]

Jürgen Link thus argues that each individual drama is characterized by a unique character constellation within the fictional world, which is yet always formed within particular historical genre (or sub-genre) conventions.[7] It is essential, however, for this approach that not the characters as such but the characters’ smallest semantic qualities are the elements within a drama’s “configuration.” In the case of Edward Albee’s Who’s Afraid of Virginia Woolf?, from which Mateas and Stern admit to have gotten the idea for a drama about a marriage in crisis, e.g., this could include “natural” attributes such as age/generation, temper, gender, physical appearance, or rather “social” attributes such as education, rank, professional success, ambition, alcoholism, etc.[8] According to Link, any drama can be understood “as a particular selection taken from a set of combinatorially possible intersections of qualities that can be arranged in linear succession according to explicit rules.”[9] Of course, in each drama only a marginal number of possible combinations are actually selected. This is because there usually are one or two protagonists designated as the leader(s) of the action whose qualities determine the main conflict and thus establish a specific dramatic logic. Consequently only those other combinations are selected that support this dominant conflict.

In principle, the configuration of marital drama provides the opportunity for a multitude of conflicts. But in stage drama only those conflicts are conducted that have been determined from the set of possible conflicts. In interactive drama, in contrast, player interactions make it at least difficult, if not ultimately impossible, for game designers to determine a preference for a particular conflict and thus trigger the plot into one predetermined direction. From the player’s viewpoint, this is undesirable insofar as a strictly determined plot would constrain his agency. Therefore the dramatic oppositions from the configuration, at least potentially, are to be playable options in interactive drama. In Façade, authorial control has deliberately been reduced in order not to over-constrain the player’s first-person engagement within the dramatic world. Yet, at the same time, for supporting the concomitant player/spectator’s “third-person reflection across multiple experiences in the world,” the plot needs to be structured “such that each run-through of the story has a clean, unitary plot structure, but multiple run-throughs have different, unitary plot structures.”[10]

Experiencing Façade: Player Agency and Drama Experience

In a game, there always is a conditional boundary between the factual and the fictional world solely defined by game-rules. In narrative and dramatic contexts, however, there is an unconditional boundary, which needs to be strictly accepted in order not to destroy the “willing suspension of disbelief” (Coleridge).[11] Computer-aided interactive media, in principle, make it impossible to frame the fictional world by setting such an unconditional boundary. Therefore the design goals of an interactive drama are conflicting: On the one hand, the recipient/player ought to have an aesthetic experience comparable to that of the audience of a classical drama, namely “enactment, intensity, catharsis, unity and closure”[12], which requires the boundary between the factual and the fictional to be strictly observed. On the other hand, unlike Aristotelian drama, the interactive drama also has to provide the player with a strong sense of first-person agency as character within the story, which cannot be brought into compliance with the first premise.[13] For that reason, the development and implementation of an efficient “drama manager,” i.e. an artificial intelligence plot system that contains a library of basic plot elements and uses knowledge about the structure of well-formed plot arcs to construct new experiences, is of key importance. This drama manager has to organize the interaction of the player-recipient with the AI system in a way that prevents the crossing of the “unconditional boundary” to disturb or even undermine the aesthetic experience. As regards the double role of player and recipient, this requires the first-person agency as character within the story to be semantically determined in accordance with a dramatic logic. Only if this is the case, the third-person reflection about the dramatic story becomes possible at all.[14]

To sum up, I hope to have made clear that we still depend on literary genre theory—in this case, on drama theory—if we want literary studies to contribute to the discussion about emerging aesthetic forms in computer-aided media systems. However, one essential media-induced difference of interactive drama and “traditional” stage drama needs to be stressed: The conflicting qualities in a drama’s configuration and therefore the motivational logics of characters are not only at the producer’s but, to a high degree, also at the computer system’s as well as at the player’s disposal.

<![if !supportEndnotes]>

[1] Michael Mateas/Andrew Stern: “Façade: An Experiment in Building a Fully-Realized Interactive Drama,” 2003, 31 Aug. 2007 <http://www.lcc.gatech.edu/~mateas/publications/MateasSternGDC03.pdf>, p. 2.

[2] Cf. Michel Chaouli: “How Interactive Can Fiction Be?,” in: Critical Inquiry 31.3 (2005), pp. 599-617.

[3] Michael Mateas: “A Preliminary Poetics for Interactive Drama and Games,” in: Noah Wardrip-Fruin/Pat Harrigan (eds.): First Person: New Media as Story, Performance, and Game. Cambridge, MA: MIT Press, 2004, p. 19.

[4] Rainer Leschke: “Narrative Portale: Die Wechselfälle der Verzweigung und die Spiele des Erzählens,” in: ders./Jochen Venus (eds.): Spielformen im Spielfilm: Zur Medienmorphologie des Kinos nach der Postmoderne. Bielefeld: Transcript, 2007, p. 200.

[5] At first view, however, it seems paradoxical that the designers of Façade, on the one hand, attempt to utilize the latest media technologies for creating dramatic experiences, but, on the other hand, explicitly refer to the oldest and most traditional theoretical drama model. Thereby they also ignore exactly those participatory genres in 20th century theater that aimed at dissolving the boundaries between stage and audience. What takes place on stage then no longer only is a representation of a dramatic text but it is the product of the real-time interaction of actors and audience. Cf. Erika Fischer-Lichte: Ästhetik des Performativen. Frankfurt am Main: Suhrkamp, 2004.

[6] Aristotle: The Complete Works of Aristotle: The Revised Oxford Edition, vol. 2. Ed. By Jonathan Barnes. Princeton, NJ: Princeton UP, 1984, p. 2320 (my italics).

[7] Cf. Jürgen Link: “Zur Theorie der Matrizierbarkeit dramatischer Konfigurationen,” in: Aloysius van Kesteren/Herta Schmid (eds.): Moderne Dramentheorie. Kronberg: Scriptor, 1975, pp. 193-219.

[8] Cf. Edward Albee: The Collected Plays, vol. 1: 1958-65. Woodstock; New York: Overlook Duckworth, 2004, pp. 149-311.

[9] My translation from Link: “Zur Theorie der Matrizierbarkeit dramatischer Konfigurationen,” loc. cit., p. 197 (my italics).

[10] Mateas: “A Preliminary Poetics,” loc. cit., p. 27.

[11] Cf. Jochen Venus: “Teamspirit: Zur Morphologie der Gruppenfigur,” in: Leschke/Venus (eds.): Spielformen im Spielfilm, loc. cit., pp. 305ff.

[12] Mateas: “A Preliminary Poetics,” loc. cit., p. 28.

[13] In a strict sense, the differences between game, play and drama are even more complex. In modern drama, there has always been an inherent aesthetic interdependency of play/playfulness and seriousness, of the fictitious actions within the fictional world and its real-world counterparts. This aesthetic requirement is certainly more difficult to meet for an “interactive drama” like Façade, as it has to make the player reflect on the difference between the domestic argument in Façade and domestic arguments in general or even his/her own marital problems.

[14] It should, however, not be forgotten that this has been regularly brought up in drama history: In the tradition of the “play-within-play”—from Shakespeare’s Hamlet to Ludwig Tieck’s Puss in Boots or Pirandello’s Six Characters in Search of an Author—the “fictitious/imaginery spectator” had already been integrated into the dramatic action.