book review

Mind The Gap
An artsie review of Chris Crawford on Interactive Storytelling

When Chris Crawford speaks many people stop and listen. When Chris Crawford speaks about interactive storytelling I rush to the nearest bookstore and get the book. As an Atari veteran from the early eighties, he published several games, articles and books on games. His Art of Computer Game Design (1983) is a fixture on the list of references of many academic books and articles. He is one of the few people that try to bridge the gap between the "techie" culture of the industry and the "artsie" culture of the academia and his name carries weight on both sides of the divide. Does he succeed in bridging that gap with his latest book On Interactive Storytelling (OIS)? Well, that is a difficult question to answer, but I am afraid he does not. His book is technically sound, but it fails to cater for the appetite of the artsie reader.

The "two cultures" problem is addressed by Crawford in chapter 4 of OIS. It is a problem that has been identified by C.P. Snow nearly 50 years ago. The two cultures are those of science ("techies") on the one hand and those of the arts and humanities ("artsies") on the other. These two fields grow increasingly apart and there is much misconception, disdain, and even hostility between the two. The gap between the techies and artsies is felt very strongly in the fields of computer games and interactive storytelling because there the two cultures meet as games and interactive stories are both technical and artistic objects. Crawford comes from the techie culture as he admits himself (p. 73). As a result he is aware that his perception of the artsie culture and their take on interactive storytelling is somewhat flawed. The perceptive witticisms of the fictional lectures on the "the semiotics of Mario Brothers" and "mimetics in text adventuters" aside, his work does suffer a little from Crawford's ignorance to the finer issues that occupy the attention of the artsie culture. At least it does for someone whose background stems from that side of the divide. Someone like me.

However, first and foremost OIS is an impressive work. This should come as no surprise as Crawford's experience and experiments of the last twenty-five years went into its making. His insights, logic and intuition is spot on most of the time. Furthermore, OIS is written in an accessible style. There are many examples, and the use of pseudo-code is easy enough to follow for anyone with an interest in games. Crawford uses an 'alter ego' that raises objections and asks questions at appropriate moments, and the dialogues that follow are an efficient way to deal with some of the more fundamental issues without being dragged into lengthy academic expositions. Although some would say that this method lacks academic rigour or does not do justice to the persons who raised the issues in the first place. I am not to bothered by it much. After all the book seems to be aimed more at game designers and developers more than at the game scholars.

The build up in OIS is sound. Crawford begins by discussing the core concepts of narrativity, interactivity and interactive storytelling. He makes a strong case for the notions of abstraction, mathematical representation and verb thinking which are pivotal in his implementation of interactive storytelling. In short, abstraction tackles the problem between artistic control over a plot and the freedom of the player by redirecting the artistic effort to the level of the 'metaplot'. At this level the author lays down the rules of engagement rather than directing every dramatic instance. Dramatic potential, and by extension the artistic expression is, realised by these rules instead of a fixed set of dramatic developments. This potential is better represented mathematically, as the mathematical pseudo-code indeed only defines the potential and while the details are left to be filled in by the player or a random generator. With verb thinking Crafword stresses the importance of the player's activity. Repeatedly he asks us to consider what it is the player does. With verb thinking it is easy to see the problem with many commercially produced games. Whereas the industry has revolutionised the technology, the things the player does have not changed that much at all. She is still only "shooting, moving, sneaking, hiding acquiring, and so forth" (p. 244). This way the resolution of conflicts, which is central to both games and stories, remains violent and primitive.

Crawford goes on to discus different strategies of interactive storytelling. He dismisses a large number of strategies that have been tried in the past. Branching tree plots, with or without 'foldback', storified games and emergent stories are all considered poor strategies. He sees more potential in 'data-driven' approaches such as the application of Vladimir Propp's universal functions of the fair-tale in a storytelling engine which yielded "impressive results" (p. 166). But in the end Crawford puts most faith in 'language-driven' strategies. This approach takes as his basis the language of interaction between the human and the computer. The shape and scope of this language ultimately determine the shape and scope of the stories told with it. If the expression of the player is limited to a handful of violent actions this is going to affect the narrative potential of the story.

The latter part of OIS is largely devoted to the techniques of interactive storytelling Crawford developed during the creation of his storytelling engine the Erasmatron. These techniques which range from the attributes of the personality model, how to craft events, keep track of these and distribute knowledge are very interesting. They make me want to boot up my C++ compiler to try my own hand at them, especially as Crawford indicates points for improvement and problems he has not yet tackled. But this part and the last part that describes the applications of interactive storytelling techniques by Crawford and others reads a little as an appendix to the earlier parts.

All in all, OIS makes a good read and Crawford's voice remains one that is articulate, experienced and authoritative. But I find one thing lacking: I miss a spark of life that could lift this work beyond the technical level and to a level that testifies a clear and complete vision of the technological and artistic dimensions of interactive storytelling.

To be more precise, what I find missing is a more detailed theory of representation. How does a 'storyworld' make statements about the world? How is a interactive author supposed put in her vision of love, the world at large or her government in particular? These are all important aspects of stories Crawford acknowledges in the first chapters. The mathematical representation discussed half way through OIS falls short in this respect. The idea of mathematical representation is illustrated with the climactic scene of The Return of the Jedi. In this scene Darth Vader is shown to hesitate between his loyalty to the Emperor and his love for his son Luke Skywalker for a long moment. mathematical formula. Here is mathematical formula Crawford uses to express the same situation (p. 108):

competing force 1 = Loyalty[Darth, Emperor] + SelfInterest[Darth] Idealism[Darth]

competing force 2 = Love[Dart, Luke] + Empathy[Darth]

if (competing force 1 > competing force 2)
then WatchLukeDie
else TurnAgainstTheEmporer

However I think the formula fails to represent one crucial aspect of that scene: the difficulty Darth Vader has in making this discussion. This difficulty is represented by the length of the shot and the way Darth Vader alternately turns his head between Luke and Emperor. This difficulty adds considerable dramatic weight and significance to the moment. It is an indication of how far Darth Vader had gone over to the 'dark side', and of the greatness of the effort and risk of Luke's gambit to bring Darth Vader back. Moreover, it makes the decision a "closely balanced decisions that can reasonably go either way" that Crawford favours for interactive storytelling (p. 54). The scene in the film ties in with a lot of events of the same and the previous two films. In particular it echoes the climax of The Empire Strikes Back where Darth Vader tries to seduce Luke Skywalker to join the dark side. Can mathematical representations do the same? And if so, how does that mechanism work? Studies of cinema and literature have developed elaborate theories on metaphors, references, reversals, satire, and many more structures of meaning. Artsies often refer to such structures and theories as 'rhetorics' or 'poetics'. Without a poetics of mathematical representation, many of the examples Crawford offers remain trivial and lifeless.

The issue I raise is not an issue with mathematical representation itself, but with Crawford's discussion of it. Never does he describe how different mathematical formulae interact. If we take for example SimCity how we can claim that it advocates a left-wing or right-wing ideal, as has been done? What are the lessons taught to us by a game like Civilization? How can September 12th (Gonzalo Frasca's simulation of the war on terror) inspire so much debate? I am inclined to argue that it is the combination of the mathematical formulae, their interaction with each other and with external ideas or formats that plays a crucial role. Such ideas have been developed by many cultural inspired theories of representation such as semiotics and intertextuality, which are more about these and other (inter)relations than the individual signs or texts. True, these examples are more simulations than games, let alone interactive stories, but they are prime examples of mathematical representations.

Crawford does an admirable job in describing the elements of interactive storytelling. The technical detail maybe more than an artsie like me wishes for, but Crawford's easy and accessible style turns it into a valuable lesson for all. Crawford's approach to interactive storytelling is interesting and shows much promise for the future. The key notions of abstraction, mathematical representation and verb thinking are all valid. What is lacking, is a conclusive account of the combination and artistic configuration of such elements. But maybe expecting such an account would be to soon. After all, interactive storytelling is only a fledgling discipline. And in any case, a lot of scholars including me would be out of business.


Crawford, Chris (1983) The Art of Computer Game Design. Available at

Crawford, Chris (2005) Chris Crawford on Interactive Storytelling. New Riders, Berkeley.

Gonzalo Frasca's September 12th can be found at

I assume the reader is well familiar with the other films and games I mentioned.