The Tale-Spin effect has had a huge impact on previous interpretations of Tale-Spin, even when the interpreters have come from very different positions as scholars. Janet Murray’s Hamlet on the Holodeck (1997) and Espen Aarseth’s Cybertext (1997) provide helpful illustrations of this. In these cases, the Tale-Spin effect not only causes the authors to misinterpret Tale-Spin, but also to miss opportunities for making fruitful connections to their own areas of interest.
Murray, in her book, discusses Tale-Spin in the context of an argument that “for authors to create rich and satisfying stories that exploit the characteristic properties of digital environments ... [w]riters would need a concrete way to structure a coherent story not as a single sequence of events but as a multiform plot” (185). We might expect, given this, for Murray to criticize Tale-Spin for organizing its operations at the level of character, rather than at the level of plot. Instead, however, Murray seems to assume that Tale-Spin does operate at the level of plot, and simply does so defectively.
Murray reprints the famous mis-spun tale of Joe Bear forming the failed goal, over and over, of bringing Irving Bird a worm so that Irving will tell him where a worm is. She precedes the reprinting by saying that “stories told from an abstract representation of narrative patterns but without a writer’s relish for specific material can be incoherent” (200). After the story Murray writes:
The program goes into a loop because it does not know enough about the world to give Joe Bear any better alternatives. The plot structure is too abstract to limit Joe Bear’s actions to sequences that make sense. (200)
Actually, as discussed earlier, Tale-Spin looped because — at the partially-completed state it was in at the time this mis-spun tale was generated — its characters could reassert a goal that had already failed. In fact, Joe Bear’s problem had to happen at the character level — it could not happen at the level of “plot structure” — because Tale-Spin has no “abstract representation of narrative patterns” at all.
This problem does not cause Murray’s argument to derail, by any means. Hers is mainly a speculative argument, about the sorts of experiences that might eventually be possible with interactive story systems. In fact, Murray spends many more pages on her imagined system for interactive stories set in the Casablanca world than she does on Tale-Spin or other actually implemented examples.
No, rather than a derailing, Murray’s misinterpretation mostly leads to a missed opportunity. As the next chapter of her book demonstrates, she is very interested in systems that model the interior operations of fictional characters. And characters like Joe Bear and George Bird have quite complex interior operations, if one looks beyond the anemic events output by Mumble. The Tale-Spin model is, of course, problematic — and the cognitive science ideas that inspired it are now abandoned. But, nevertheless, its operations provide much that deserves the attention of writers such as Murray. The complex character behavior Tale-Spin produced in the 1970s is much more likely than an imaginary Casablanca system to make convincing fodder for an argument such as Hamlet on the Holodeck’s.
Espen Aarseth, in Cybertext, calls Tale-Spin “a cybernetic fiction device that does not work” (131). He concludes this based on a selection of its mis-spun tales — rather than those produced by the actual, completed Tale-Spin system. Aarseth does see fit to qualify his statement with the phrase “at least in the examples given here,” but the rhetoric of failure is important for the point he seeks to make. Tale-Spin is one of Aarseth’s three primary examples for the argument that machine narrators should not be “forced to simulate” human narrators (129). Tale-Spin is presented as a failed example of such simulation, with its mis-spun tales its only claim to interest.
From the viewpoint of AI, Aarseth’s is an exceedingly strange argument. As I will discuss in the next chapter, the primary critique of Tale-Spin in AI circles is precisely that it does not attempt to simulate a human narrator. Tale-Spin simulates characters — not narrators, not authors. This has been seen as a fundamental mistake, and was outlined by writers such as Natalie Dehn (in widely-cited papers) more than 15 years before the publication of Aarseth’s book (Dehn, 1981a, b).
However, as with Murray, we can still follow Aarseth’s argument even while acknowledging its troubles. When we see phrases such as “trying to create a surrogate author” we can substitute something like “ideological attachment to narrative ideals” (141). This is because Aarseth is arguing against simulating human narrators only as a proxy. He’s really arguing against focusing attention on attempts to use the computer to extend the pleasures of traditional fiction and drama. Instead, Aarseth seeks to turn our attention to literature based on such features as combinatorics, interaction, and play — on the new literary possibilities opened by the specifics of the networked computer. As Aarseth writes:
To achieve interesting and worthwhile computer-generated literature, it is necessary to dispose of the poetics of narrative literature and to use the computer’s potential for combination and world simulation in order to develop new genres that can be valued and used on their own terms. Instead of trying to create a surrogate author, efforts in computer-generated literature should focus on the computer as a literary instrument: a machine for cybertext and ergodic literature.... [T]he computer as literary agent ultimately points beyond narrative and toward ergodic modes — dialogic forms of improvisation and free play... (141)
It is the puzzle of Tale-Spin that we can diagnose, here, a problem very much like Murray’s. In Tale-Spin Aarseth has a great missed opportunity. The story structures Tale-Spin produces are almost never like those that a human storyteller would produce. Instead, it produces strange structures of plans within plans within plans. It produces what we might call, from the possible worlds perspective, “minimalist fictions” — made up almost entirely of possible worlds of planning, speculation, lies, and so on (without redundant emotions, movements, even geographical locations). It is a combinatory engine for spinning off possible worlds embodying an alien vision of humanity, driven by the temporary worldview of a research lab. In other words, Tale-Spin can be seen as an example of one of the types of literature for which Aarseth is calling.
Aarseth’s missed opportunity, combined with Murray’s missed opportunity, reveals something interesting. Tale-Spin, early as it was, stands at an important crossroads. If we choose to emphasize its continuities with traditional fiction and drama, via its characters, then it becomes a useful touchstone for views such as Murray’s. If we choose to emphasize its complicated strangeness, its computational specificity, then it becomes an important early example for views such as Aarseth’s. In either case, a close examination of the system’s operations reveals something much more intriguing than either author assumed — making a clear case for the necessity of developing an approach to reading what processes express.
Returning to the themes of this book’s introduction, we can see the processing of Tale-Spin as expressive in two ways. It is an authorial expression, an act of media-making, that creates fictional worlds through its processes — which are only partially visible on its surface. It also expresses a relationship with histories of artificial intelligence and cognitive science through the very design of its processes and data structures — which are completely invisible on its surface. Through the first of these it demonstrates the approach of story generation, moving beyond the assembly of fiction through pre-created and explicitly-connected chunks, as well as the difficulties of the Tale-Spin effect. Through the second type of expression it also provides a legible example of the limits of simulations of human behavior based on hand-authored rules. Altogether, it demonstrates that understanding processes in both these ways is important for the development of digital media and software studies — for creators and for scholars.
Powered by WordPress