For the time being, a generally accepted definition of “narrative” (and related terms like “story” and “storytelling”) seems elusive, and thus any hope of a simple solution on that end might be naive. At the same time, this does not mean that there is no chance for a comprehensive understanding. The key here is to move beyond the binary property of definitions (what is/is not a narrative) towards a relational approach: how can we describe the relationship between different definitions? On which dimensions do they differ? For this purpose, spatial mappings are promising. Provided that the respective dimensions of such a mapping are carefully chosen, they can offer novel insights into the relationship between different positions.
Janet Murray’s position in Hamlet on the Holodeck
) can be categorized as foregrounding the impact of the affordances of the digital medium (her categories of procedural, participatory, spatial, and encyclopedic) on narrative, as well as on the player’s agency, which for her represents a fundamental aesthetic quality. Murray also associates an additional aspect of player agency with transformation, which is the dual quality of affecting change in the narrative, as well as being affected by the experience. This position will be ranked as high (6) on both dimensions.
Espen Aarseth in Cybertext
) introduced the notion of the “ergodic”—the non-trivial effort required by the audience to traverse an interactive narrative. In the present context, this translates to a high ranking on the user agency dimension (6). In respect to media specificity, Aarseth indeed champions novel narrative forms in mostly digital manifestations, for example adventure games. Yet, his insistence on a transmedial perspective on narrative (according to him, textual representations on paper can also qualify as cybertextual machines) results in a lower score on this dimension (2).
Similarly, Jesper Juul’s initial rejection of narrative in games, when he says “the computer game is simply not a narrative medium” (Juul 1998
), is in line with traditional perspectives in the humanities with regards to media specificity (narrative is not affected by the digital medium), as Juul understands narrative as immutable while emphasizing the difference to player agency in games. This perspective will be ranked low (0) on the media specificity dimension and high on player agency (6). Juul’s later position from 2001 onward (Juul 2001
) moves the position on the dimension of media specificity to 1, in line with Aarseth’s later position (Aarseth 2001
), while Eskelinen’s pronouncement (Eskelinen 2001
) of the incompatibility of narrative with games ranks lower (0). Juul’s 2005 book Half-Real
) sees a role for narrative (in the guise of “fiction”) in games, and thus might initially appear to open up towards a more media-specific understanding of narrative. However, Juul still preserves a dichotomic perspective, in which games represent an open structure and narrative a fixed, closed one. For him, the role of narrative in games is mostly ornamental to provide context, and thus is increasingly superfluous the more the player progresses. From the perspective of media specificity, the mapping of this positions thus remains at 1, and player agency at 6.
In 2004, Henry Jenkins published his essay “Game Design as Narrative Architecture” (Jenkins 2004
), in which he argues for a media-specific perspective. He describes four specific modes for interactive digital narrative: “evocative”, “enacted”, “embedded”, and “emergent”. The evocative mode refers to narratives that reference prior stories in other media, e.g., a Harry Potter game. Enacted narratives allow the user to act out specific roles within an existing narrative universe, for example that of a Hobbit from Tolkien’s Lord of the Rings
. Embedded narratives convey information by means of spatially distributed narrative-infused encounters, as exemplified in the adventure game Myst
). Finally, emergent narratives appear in rule-based game worlds that provide players with the tools to construct stories of their own (e.g., The Sims
)). With respect to media specificity and player agency, the different modes require individual assessment—the evocative and enacted modes cannot be ranked as fully media-specific, since Jenkins describes them as dependent on more traditional print literary or cinematic narratives. Consequently, they are ranked at 4 on this dimension. Player agency is medium (3) in the evocative mode, as it is restricted to the “satellite narratives” that cannot affect the main traditional narrative core, and higher in the enacted mode, but does not reach full agency for its dependence on the main narrative (5). For embedded narrative, media specificity ranks higher (6), yet user agency is limited to movement (2). Finally, emergent narrative ranks highly on both dimensions, as a media-specific form with extensive user agency (6/6).
Celia Pearce, in her paper “A Game Theory of Game” (Pearce 2004
), foregrounds a media specific perspective and argues for a different understanding of narrative—no longer as “storytelling”, but in a “play-centric context”. Pearce rejects literary and film theory and proposes a path towards a specific theory of games and narrative within games that casts narrative as a mostly experiential quality. Narrative games are successful, Pearce tells us, if they apply what is engaging and interesting about narrative and “use it to enhance the play experience”. On this foundation, Pearce identifies six different modes of narrative in games. The first, “experiential”, is the emergent narrative that stems from the conflict of the games as experienced by the player. According to Pearce, this mode exists in every game; it is media-specific (6), and the user has agency (6). The second mode, “performative”, describes the same emergent narrative, but from the spectator position. This mode still ranks high on media specificity (6), but the spectator has no agency on their own (0). “Augmentary” is Pearce’s term for supplemental “layers of information, interpretation, backstory, and contextual frameworks” that enhance the experience. This mode can include media-specific elements, but could also use traditional fixed-linear narrative modes, and is thus scored at the midpoint (3) for media specificity, while user agency is low (1). “Descriptive” denotes the retelling of game experiences to third parties, either in person or through platforms such as Twitch. This narrative mode will not be represented in the mapping, as it constitutes a “second order narrative” that happens outside of the direct relationship between player/interactor and the work. “Metastory” is similar to “augmentary”, and is ranked in the same way (3/1), but instead describes a fully developed narrative framework. Finally, a “story system” allows the player to build her own narrative content from generic parts. This mode is similar to Jenkins’ understanding of emergent narrative (6/6).
Similar to Pearce, Salen and Zimmerman in Rules of Play
(Salen and Zimmerman 2004
) see narrative mostly as an experiential quality, but focus on the designer’s task in creating said experience. They describe two possible narratives modes: embedded and emergent. Their understanding of emergent narrative aligns with both Jenkins’ and Pearce’, and thus is ranked the same (6/6), yet their concept of embedded is broader than Jenkins’. For Salen and Zimmerman, an embedded narrative can be experienced through all kinds of interactions (not just spatial as with Jenkin’s), therefore user agency is ranked high (5); however, this narrative mode misses the highest mark for the lack of control over the global narrative. Media specificity is high in this case too (6).
Marie-Laure Ryan, in her 2006 book Avatars of Story
), approaches the question of interactive digital narrative from a narratological perspective. While she clearly acknowledges some of the changes brought about by the digital medium, and identifies some resulting opportunities and particular structures, she ultimately considers interactivity to be in conflict with narrative: “[…] interactivity is not a feature that facilitates the construction of narrative meaning”. Therefore, on the media specificity dimension, her position is ranked at the midpoint (3). When it comes to player agency, Ryan’s position requires a closer examination, as she offers a categorization of different types of interactivity along two different trajectories, internal–external and exploratory–ontological, which result in four different modes:
External–exploratory: the interactor makes choices in a game of discovery; e.g., text-based hypertext fiction,
Internal–exploratory: the interactor as a visitor with no ability to change; e.g., 360 video,
External–ontological: the interactor is a god; e.g., The Sims,
Internal–ontological: Interactor is a character in the virtual world, determining the fate of the avatar and the virtual world; e.g., Façade
(Mateas and Stern 2005
For the mapping, these modes equate to different levels of player agency. External-exploratory is at 4 because the interactor has control over choices, internal-exploratory is at 2 with limited agency (only in movement), and finally external-ontological and internal-ontological are at 6, as the interactor has considerable impact on the narrative in both modes.