Cybertext Theory and Literary Studies, A User's ManualMarkku Eskelinen
"Still, what would theory be worth if it were not also good for inventing practice?
- Gérard Genette, Narrative Discourse Revisited
There's at least one serious downside to Espen Aarseth's cybertext theory: it puts, or is very capable of putting, an end to hype in the rapidly expanding field of digital textuality, where it seems there are always newcomers who can't make a living without fashionable exaggerations and homebred buzzwords (like multicourse for the tel quelian text and genres always already blended). The golden age of media essentialism - confusing readers with writers, links with intertextuality and texts with rhizomes (among other equally monumental and influential misreadings) - has been over for a while now, along with some careers, and if that's the end of the world as we know it, I feel fine.
In contrast to the dead ends of hypertext theory and its posthuman derivatives, cybertext theory addresses the unique dual materiality of cybernetic sign production and gives us an accurate and heuristic description of how the textual medium works. It achieves this goal by approaching computers as computers, and not in the common montypythonesque way of defining networked and programmable media as something completely different, be that theatre, cinema, comics or (poorly read) continental philosophy. The elementary idea is to see a text as a concrete (and not metaphorical) machine consisting of the medium, the operator, and the strings of signs. The latter are divided into textons (strings of signs as they are in the text) and scriptons (strings of signs as they appear to readers/users). The mechanism by which scriptons are generated or revealed from textons is called a traversal function, which is described as the combination of seven variables (dynamics, determinability, transience, perspective, links1, access, and user function), and their possible values. This combinatory approach gives us nearly 600 (576 to be exact) different media positions, where every text (from I Ching to MUDs) could be situated based on how its medium functions. All this has consequences that don't seem to be well understood or even realized, and in what follows I'll try to make some of them more obvious and visible. I'll begin with traditional literary studies and hypertext theory and then move on to perhaps more interesting issues like "new media" studies and computer game research.
Cybertext focuses on ergodic literature, where the user has to do non-trivial work to traverse the text (instead of merely interpreting it). After a thorough critique of existing paradigms, Aarseth presents his own model and then applies it to hypertext fiction, adventure games, text generators, and MUDs. The relation of ergodic literature to non-ergodic is not very extensively discussed; the focus is understandably elsewhere. Still, it is and should be obvious that to import values and expectations from the latter to the former is not only premature and blinding, but also an indication of most conservative colonisation. In that sense this essay serves as an introduction to things an average English professor should know before trying.
But times change. The word digital doesn't carry much descriptive weight any more, as almost every aspect of our culture is more or less digital. In such a situation digital theorists are or could be migrating or even sucked back to their old disciplines that are more than willing to have their fair share of "new media". And then the crucial question becomes how to negotiate and renegotiate the relationships between these two literatures: in what terms and in whose. It's also a question of autonomy and authority and there's more to it than the inevitable changes in triviality, although I have to admit that when teaching cybertext theory nowadays in Finland, the hardest part is always to convince students that navigation is non-trivial.
ergodic and non-ergodic literature
Between 1982 and 1983 I was very unsatisfied by what I then considered as a blind alley of visual poetry. Aware of the multiple directions the genre had taken in the twentieth century, I experimented with different media ... billboards, Polaroid cameras, artists' books, fine graffiti, electronic signboards, video, mail art, photocopiers, videotex, and finally holography.
- Eduardo Kac, interviewed by Simone Osthoff
for Xenia 2
The fact that poets in the 20th century (and before) were keen on using whatever material alternative there was available to the printed page is not or at least should not be a secret. Still, such efforts seem to have completely escaped those that were inclined to contrast print to the digital or electronic, with paralysing consequences to the maturing of the field. The fuss about print versus digital seems thus to have been based entirely on narrative bias, to which the material diversity of experimental poetry was either a novelty or an insignificant fact to be ignored. In the context of the latter, however, that much hyped contrast reads more like a joke or a fairy tale: once upon time there were only two media, the old and the new, and they didn't live happily ever after.
If one tries to find a theoretical framework within which to begin discussing a wide variety of poetic practices, let's say E.M Melo e Castro's video poetry, John Cayley's programmatology (Beyond the Codexspace), Eduardo Kac's holopoetry, there's no real alternative to cybertext theory (or if there is, it's certainly not hypertext theory). That's for the most part because Aarseth's theory focuses on functional differences within media instead of making essentialist claims. That paves a way for more detailed findings, as the parameters of the cybertext typology can be supplemented, changed, and removed or made more detailed if necessary. The latter alternative is probable if such an undertaking ever traces connections between established practices and traditions including those usually situated outside serious or literal art, like advertising and movie titling for starters. This is of course only one example of many ways in which it is possible to introduce new topics and tensions to the seemingly saturated field of traditional literary studies. It may also be high time to make some sense of how texts are dispersed and scattered around our media environments.
The useful inclusiveness of cybertext theory results from its almost standard deconstructive strategy. Aarseth's theory lays its emphasis on a seriously and curiously understudied and marginalized area of literary scholarship (despite the previous honourable efforts of Brian McHale and the Tel Quel): the materiality of texts and functional differences within textual media. The existing field of textuality is then expanded and dynamically rearranged by combinatorial textonomy (the study of the textual medium) and the previously dominant forms are reinscribed back into a considerably changed field of study as mere subsets of cybertexts. (There's an array of quasi-transcendental metaphors and concepts (see Gasché 295, 316) and dissemination patterns of randomness that should spring to mind here if one is already past the absences and presences in the Of Grammatology section of Derrida for Beginners.) What emerges this way is a medium-independent map or set of functional possibilities (or media positions) that is both empirical (all the values its parameters can have are already operational in existing textual objects) and of great heuristic value, as it soon becomes evident the history of print literature and curiously print-like hypertext has been able to utilize only about 2 or 3% of those 576 non-hypothetical possibilities2 Aarseth's theory is able to foreground.
Needless to say this same inclusiveness of cybertext theory makes it useful also in defending its objects of study from various colonising enterprises from traditional literary institutions, whenever they'd become desperate enough to try. Cybertext theory can justify the study of digital and electronic textualities in their own terms, instead of submitting or committing to the traditions of print literature(s), as the above mentioned 2 or 3 % is not too much to justify any attempts to assimilate or ignore multiple textualities in networked and programmable media.
Despite the fact that cybertext theory doesn't build essentialist barriers between textual media, it's still clear that almost all the knowledge we can gain from traditional literary studies is based on literary objects that are static, intransient, determinate, impersonal, random access, solely interpretative and without links. The same goes for literary values as well. I'm not downplaying the importance of this knowledge or the flexibility of the traditional print format; I'm just saying we can now see and describe its limits more clearly (to our own benefit).
From the broader perspective it will be extremely interesting to see what will happen in and to attempts to combine the realm of non-exhaustive interpretation (literary meaning) with the exhaustively and empirically describable realm of cybertextual functioning (textonomy). I think it's exactly the combinatory and dialogic interplay between these two systems that will both make and keep literary studies attractive for years to come, and make us all look small-minded in the end. At least that's something non-trivial to do for many anti-formalists who have nothing to lose but their ignorance.
One possible way of proceeding might be borrowed from Louis Hjelmslev's semiotics. In his terms the empirical textonomical dimension of cybertexts could perhaps be seen as the substance of the expression, or at least a highly useful specification of its manifestations. Then the empirical and functional textonomy could establish a reliable point of departure from where to begin discussions and transformations between traditions, aesthetics, poetics, practices, theories and media, and approach the stunning diversity of actual and potential literatures without the usual unifying interpretative violence. Of course, there are even more dangers in this for the content oriented posthuman scholars. In order to succeed in their hunt of supposedly transparent themes, they must move from being simply ignorant to being doubly ignorant, as they now have to turn a blind eye to both the form and the substance of the expression.
Before getting into positive consequences and potential points of contact between theories (and practices) of ergodic and non-ergodic literature, let me briefly explain the two main reasons why hypertext theory can neither defend the autonomy of ergodic literature nor exercise much authority in studying it.
Firstly, hypertext theory was too much into creating hype (invidiously separating digital or electronic and print textuality), however necessary it might have seemed to be at some moment of recent history, and too much of its legitimacy and identity is still connected to that enterprise. The fierceness of the hype was understandable, as hypertext is much nearer to average print textualities than most other subsets of cybertexts: it had to differentiate at any cost because it wanted to resort to its two most print-like qualities: permanent signifiers and intransient time.3 These two values helped and will help it to continue any print tradition without causing too much trouble to future chroniclers of literary history. There was so much bull in that static labyrinth.
Secondly, for too many years it was en vogue to pretend and claim that the expanding field of digital textuality could be controlled by the shotgun wedding of hypertext theory and its forrestgumpian interpretations of post-structuralism. In retrospect this means hypertext theory will not be effective in defending that field because it is dependent upon its grave misreading and simplifications of the theories it borrowed. In short, serious print scholars will eat hypertext theory for breakfast sooner or later. And actually I can't wait for that to happen because whatever its merits elsewhere hypertext theory was an educational disaster in what comes to the level of sophistication in its attempts to apply and create literary theory. Witness Landow clarifying the supposed novelty of hypertext narratives by using Aristotle4, Douglas struggling with the dickensian or cartlandian expectations of closure5, or Hayles compromising her posthuman project by ignoring sophisticated theories of narrative (of focalization, regulation of narrative information, narrative situation etc.) in favour of outdated Henry James6 - in addition to quite amusingly reducing deconstruction to a dialectics of absence and presence (of all things).7
So in order to minimize the damage done to justifiable claims for scholarly autonomy and authority – and to clarify the current globally marketed but easily localizable conceptual mess - hypertexts should be seen as a subset of cybertexts, among many others the advocates of the former never managed to come into terms with. This solution also helps to save outstanding hypertext fictions like Afternoon, Victory Garden, and Patchwork Girl from hypertext theory.
points of contact and departure
Due to self-imposed constraints of time and space I'll limit myself to only seven examples of how to use cybertext theory in relation to other state-of-the-art practices and theories as well as to certain new disciplines or pseudo-disciplines.
Traditionalism (the OuLiPo). Marcel Bénabou could situate and present nearly 80 oulipian procedures in his heuristic scheme of objects and operations. The former were linguistic units (from a letter to a paragraph and beyond) and the latter such elementary ones as displacement, substitution, or deduction. If we run this system of practices through cybertext theory after supplementing it with the key concepts of Eduardo Kac's holopoetry, we could once again both considerably expand and transform the variety of intriguing options. In practice this means that we divide objects into textonic and scriptonic ones and specify the operations with the seven parameters of Aarseth's traversal function (dynamics, determinability, transience, perspective, links, access, and use functions).
The differences between the ideas and practices of the OuLiPo (and the ALAMO) and the literary hypertext tradition are not necessarily as great as they would seem to be at first glance. Nothing prevents us from defining conditional links of Afternoon, a story as a new type or even genre of constraint, not the traditional one situated between the author and the text, but between the text and the reader; an obvious third possibility and position would be between different phases or versions of the text. One needs only to read the user's manual to Storyspace to see how precisely formulated and flexible this system of constraints both is and could be. And from that point on one might wonder why it is that Afternoon had so few followers in its use of the link structure that allows the work to go right against the basic assumptions of previous literary theory – in the sense that because of its guard fields controlling the reader's access to some of its lexias, it can't be read in just one reading session however long that period of time would be in human or inhuman terms.
Modernism and postmodernism. Cybertexts are capable of introducing their own sets of epistemological and ontological problems in addition to the already automated ones of modernism and postmodernism that Brian McHale was able to discern in his heuristic study of postmodernist fiction. We could undermine the user's predictable points of identification by changing the number and the content of scriptons, regulate the reader's possibilities to read in time, decide to what he may return or whether he may do so at all, circulate characters between scriptonic and textonic positions, or process the traces of configurative uses of a cybertext as toxic waste in the "fictive" world it is supposed to contain. From this viewpoint it is sometimes hard to understand the constant attraction to static hypertexts.
Narratology. Textonomy seen as the study of the substance of the expression is easy to combine with Chatman's story (the form of the content) and discourse (the form of the expression). The options of controlled access and transient time necessitate the shift from two systems of time to four: in addition to story and discourse times and their altered specifications we need two other registers to account for the controllable and measurable reading time and the life span of the work of art. And so on. I'm not too interested in expanding narratives to infect too many cybertextual possibilities, but those who are, should at least know their narratologies.
Transtextuality. If there's one horizon cybertext theory is particularly capable of exploding, it is the studies of transtextuality (in this Genettean model intertextuality is only one of its five subcategories (Genette, "Palimpsests" 1-8). The relations between two or several texts described both in textonomical and textological terms results in relations and types of relations that go far beyond those recognizable in print and hypertext fiction and poetry. The main reason for this is that the relations between texts have ceased to be merely interpretative (as in print), or interpretative and explorative (as in web hypertexts): cybertexts can be programmed to affect each other far more deeply than that. In this scenario the way I read and use my Hegirascope could affect someone else's options to read Reagan Library, or these works might assimilate or annihilate each other. More seriously, we could go beyond independent textual objects, which would be a logical continuation of self-variable and self-supplementing texts. This example of texts consuming and annihilating other texts is just one of many exciting and unforeseen possibilities that can be generated from Aarseth's heuristic study and typology of cybertexts. It's also capable of continuing the discussion of the architext and the self-contradictory mess called Western poetics from the point where Genette managed, in The Architext, to bring or advance it in the early 1980's.
Creative writing. Cybertext theory can be used for inventing new practices, and in its most banal form this means it could be a pedagogical tool of great heuristic value in the dubious business of teaching what is too often called creative writing. Cybertext theory is a very effective antidote to the well-known theories of literatures of exhaustion, or the almost senile laments of the passing of the golden age, or of the supposedly necessary or unavoidable multimedialisation – for the simple reason it can show roughly 570 fresh alternatives to what those other approaches try to bury.
"New media." The semiotic point of departure of cybertext theory gives this approach an advantageous edge that could be used and applied also to phenomena other than textual, such as digital cinema or various other forms of audiovisual presentations, representations, and transmissions. It's ironic that film scholars like Lev Manovich, busy in their attempts to colonise something regrettably feebly described as new media for film studies and thematic narratives, are at their best just reinventing parts of the cybertextual wheel when they finally discover and try to theorize the all-important difference between database and interface. If we describe the material and functional side of audiovisual (re)presentations and transmissions through cybertext theory, we can quickly find 575 alternatives to traditional cinema (that is to the system of static dynamics, determinate response, impersonal perspective, transient time, controlled access, no links and interpretative user function) – much more than Manovich was able to find in his efforts to define what digital cinema is. So maybe there's something that deserves to be called ergodic art (Aarseth, "Aporia"), if for once we could look beyond the differences in signs into the mechanisms that produce them.
Computer game studies. In Aarseth's typology of cybertexts there are four user functions (very useful in specifying the nature of so-called interactivity). In every form of literature the interpretative user functions dominates the other possible user functions: even though we have to work in order to enjoy John Cayley's Book Unbound, we do it for interpretative reasons. However, the typology suggests other arrangements not dominated by interpretative interests and goals, but only assisted by them. We can find such a situation in gaming and in computer games in particular. There we interpret in order to be able to configure and move from the beginning to the winning or some other situation, whereas in ergodic literature we may have to configure in order to be able to interpret. And here, finally, we'll confront the great unstudied and non-theorized of our culture: computer games.
>>---> Scott Rettberg, N. Katherine Hayles, and Matthew Kirschenbaum respond.
^1. Links are then only one of these seven dimensions, and it is exactly this broader view provided by cybertext theory that is valuable, as it gives us more to think of than simple link and node-structures. Consequently, it doesn't make much sense to contrast links to the "computational" as they are both included in Aarseth's model. Hayles's cyberlliterature becomes then just a parental guidance version of Aarseth's cybertext: " In the case of cyber|literature, the set of statements are 1) the literary tradition is its parent, 2) the computer game is its parent, 3) the link is the essential feature, and 4) computation is the essential feature." And cyber|literature is just literature if we are past fetishising the digital.
^2.Aarseth's typology contains 576 different media positions for texts depending on the combination of their values in those seven parameters (dynamics, determinability, transience, perspective, links, access, and user function).
^3.Stuart Moulthrop seems to have moved from hypertext fiction to cybertext fiction during the 1990's – from static scriptons and intransient time of Victory Garden to transient time of Hegirascope and intratextonic dynamics of Reagan Library. In contrast to later hype, Michael Joyce's "Selfish Interaction" gives an interesting account of literary thinking behind and literary ideas beyond Storyspace.
^4.In his Hypertext 2.0, 181-192. Beginning with: "Hypertext, which challenges narrative and all literary form based on linearity, calls into question of plot and story current since Aristotle." Challenges to Aristotle have been numerous in theory and practice, and there is really no excuse to reduce the problems of narrative and narration to the petty problems of plot and story. Landow seems to be totally unaware of sophisticated structuralist and post-structuralist theories of narrative communication (Chatman), comprehension (Sternberg, Bordwell, Branigan), and construction (Genette, Prince).
^5.See her "How Do I End This Thing?" in Landow (ed) Hyper/text/theory, 159-188. For example: "While my reading of all these versions are logically possible, I cannot accept all of them simultaneously in my final understanding of the events described in Afternoon" (167). Why not accept if you know at least a little of postmodernist fiction? For the strategies of the latter see McHale, and in practice the novels of Alain Robbe-Grillet in the 1960's and 70's, and short stories like Robert Coover's "The Babysitter."
^6.At least from the early days of the Tartu School (Jurij Lotman et al.), it's been clearly articulated that narrative is a very effective way of packaging information. In order to find out how effective one needs to study narratives as narratives, hence the necessity of formal narratology and narratological sophistication to the highest degree if one uses narrative in relation to cybernetics and information theory as a circulatory system between the stories of science and those of literature. Sadly, in How We Became Posthuman N. Katherine Hayles neither has this kind of elementary knowledge nor seems to understand she should have. James doesn't carry very far, as his theories of point of view have been replaced a long time ago with more accurate ones of focalization, perspective, and narrative instance. At least in Europe, as the sad fact remains that sophisticated narratology is still not a shared knowledge base in the U.S. according to a recent interview with Brian McHale.
^7.In her response to Andrew Kurz's review.
Aarseth, Espen. "Nonlinearity and Literary Theory". Hyper/text/theory. Edited by George P.Landow. Baltimore: Johns Hopkins University Press, 1994. 51-86.
---. Cybertext. Perspectives on Ergodic Literature. Baltimore: Johns Hopkins University Press, 1997.
---. "Aporia and Epiphany in Doom and The Speaking Clock: Temporality in Ergodic Art". Cyberspace Textuality. Edited by Marie-Laure Ryan. Bloomington and Indianapolis: University of Indiana Press, 1998. 31-41.
---. "Allegories of Space". Cybertext Yearbook 2000. Edited by Markku Eskelinen and Raine Koskimaa. Saarijärvi: Publications of The Research Centre for Contemporary Culture, University of Jyväskylä, 2001. 152-171.
Bénabou, Marcel. "Rule and Constraint". OuLiPo – A Primer of Potential Literature. Edited by Warren J. Motte. Normal, IL: Dalkey Archive Press, 1998. 40-47.
Bordwell, David. Narration in the Fiction Film. Madison: University of Wisconsin Press, 1985.
Branigan, Edward. Narrative Comprehension and Film. London: Routledge, 1992.
Cayley, John. Book Unbound. London: Wellsweep, 1995.
---. "Beyond the Codexspace: Potentialities of Literary Cybertext". Visible Language 30:2, 1996. 164-183.
Chatman, Seymour. Story and Discourse. Ithaca: Cornell University Press, 1978.
---. Coming to Terms. Ithaca: Cornell University Press, 1990.
Douglas, Jane Yellowlees. "How Do I Stop This Thing?" Hyper/text/theory. Edited by George P. Landow. Baltimore: Johns Hopkins University Press, 1994. 159-188.
Gasché, Rodolphe. The Tain of the Mirror. Cambridge, Ma. And London: Harvard University Press, 1986.
Genette, Gérard. Narrative Discourse: An Essay in Method. Translated by Jane E Lewin. Ithaca: Cornell University Press, 1980.
---. Narrative Discourse Revisited. Translated by Jane E Lewin. Ithaca: Cornell University Press, 1988.
---. The Architext. An Introduction. Translated by Janet E Lewin. Berkeley and Los Angeles: University of California Press, 1992.
---. Palimpsests. Literature in the second degree. Translated by Channa Newman and Claude Doubinsky. Lincoln and London: University of Nebraska Press, 1997.
Hayles, N. Katherine. How We Became Posthuman. Chicago and London, The University of Chicago Press, 1999.
---. "Cyberlliterature and Multicourses: Rescuing Electronic Literature from Infanticide". EBR 11, 2001.
---. "Reply to Andrew Kurz". http://otal.umd.edu/~rccs/books/hayles.html#respond
Hjelmslev, Louis. Prolegomena to a Theory of Language. Madison: University of Wisconsin Press, 1961.
"Interview with Brian McHale: The Sense of Technology in Postmodernist Poetry". Cybertext Yearbook 2000. Edited by Markku Eskelinen and Raine Koskimaa. Saarijärvi: Research Centre for Contemporary Culture, 2001. 69-85.
Joyce, Michael. Afternoon, a story. Cambridge, Mass.: Eastgate Systems, 1990.
---. "Selfish Interaction". Of Two Minds: Hypertext Pedagogy and Poetics. Ann Arbor: The University of Michigan Press, 1995. 135-147.
Kac, Eduardo. "Key Concepts of Holopoetry". Experimental-Visual-Concrete. Avant-Garde Poetry Since the 1960's. Edited by K. David Jackson, Eric Vos and Johanna Drucker. Amsterdam and Atlanta, GA: Rodopi, 247-257.
Landow, George P. Hypertext 2.0. Baltimore: Johns Hopkins University Press, 1998.
Manovich, Lev. The Language of New Media. Cambridge, Ma.: MIT Press, 2001.
McHale, Brian. Postmodernist Fiction. New York: Methuen, 1987.
Melo e Castro, E.M. "Videopoetry". Visible Language 30.2, 1996. 138-149.
Moulthrop, Stuart. Victory Garden. Cambridge, Mass.: Eastgate systems, 1991.
---. Hegirascope. http://raven.ubalt.edu/staff/Moulthrop/hypertexts/HGS
---. Reagan Library. http://raven.ubalt.edu/staff/Moulthrop/hypertexts/rl
Prince, Gerald. Narratology: The Form and Functioning of Narrative. New York: Mouton, 1982.
Sternberg, Meir. Expositional Modes and Temporal Ordering in Fiction. Baltimore: Johns Hopkins University Press, 1978.