Emergent Narrative in Interactive Media

The connections between the concepts of emergence and narrative are manifold, complex and significantly non-obvious, but in one context at least they come together explicitly: the term "emergent narrative" has an established currency in computer game studies as a potential (and desirable) effect of interactive media. Indeed for many it is the holy grail of contemporary computer game design, offering as it does the prospect of reconciliation between the conflicting values of narrative satisfaction and player autonomy. In the academic context of digital media studies, this same promise of synthesis has put emergent narrative in the front line of a long-running debate between ludologists and narratologists about the relative importance of game and narrative paradigms. My argument here suggests that emergent narrative is not the unifying concept it appears to be for computer game studies, though it does have interesting possibilities in that field; more fundamentally, though, I want to argue that this seemingly very specific concept helps to clarify the incommensurability of emergence and narrative and has implications for our larger understanding of the process of narrative sense making. The discussion begins with an introduction to emergence and some indication of its problematic relation to narrative. I then turn to emergent narrative itself, outlining the history of the concept and some difficulties of definition. I argue that these difficulties arise from confusions about the nature of simulation, and I make a case for understanding narrative and simulation as distinct and, in certain respects, antithetical modes of representation. This view of simulation undermines the narrative status of current notions of emergent narrative, despite the support those notions draw from the field of narrative theory; the argument has implications for both the concept and the field, which I expound in terms of the relation between simulations and fictional worlds. Following this critique, I make a positive case for the possibility of genuinely emergent narrative, understood as a particular use of simulation, and I conclude by extending the discussion to non-digital contexts and the lessons emergent narrative might offer to our larger understanding of narrative production.

The concept of emergence is used across a remarkably diverse set of disciplinary contexts, from theoretical physics to social science, economics to cognitive psychology. The essence of the concept is readily conveyed, but it becomes slippery upon closer examination, for reasons that will partly appear in the following discussion; while it does service as an explanatory term in many contexts, it is as likely to appear as a theoretical problem. Emergence is a feature of complex systems: the term refers to phenomena or behavior produced by a system but not apparent from an inspection of the elements of the system and the laws governing it. Aphoristic definitions of emergence commonly draw upon Aristotle's discussion of unity in the definition of substance, where he writes, "the totality is not, as it were, a mere heap, but the whole is something besides the parts" (Metaphysics 1045a); or, they cite Philip Anderson's influential argument against the "constructionist" interpretation of scientific reductionism, "More is Different"; or, perhaps more suggestively from a narrative point of view, they characterize an emergent system as "a pattern in time" (Holland 1; Johnson 20, 27, 104–105). It is more helpful for the purposes of this discussion, though, to think of emergence in relation to a distinction of levels: a system's emergent phenomena require description at a level of organization above that which provides the base-level description of the system itself. These two descriptions are non-continuous—the higher-level description cannot be reduced to the terms of the lower level—but an adequate account of the emergent phenomenon requires both. One of the most commonly invoked and accessible examples of emergence is Conway's automaton, otherwise known as the game of Life.¹ "Life," because the laws of the system are supposed to model basic parameters of survival; for this reason it is doubly suited to my purposes here, serving to illustrate not only the concept of emergence but also that of simulation, which will become important later. The game of Life is a cellular automaton...