Hypertext '08: Literatronica. Adaptive Digital Narrative :: Juan B. Gutierrez, and Mark C. Marino

Literatronica:

Juan: Empower the computer to be part of the literary transaction… mathematical literature.

Mathematical literature — not the syntatic approach, using mathematical language to describe a story. Not a semantic approach, using theorems to define stories.

Lexicographic Hypertext — basic HTML with nodes connected via links.  We navigate through the network to get from the beginning to an ending.

Tree fiction — constructing a narrative through choosing options.

Adaptive Fiction — the computer delivers chunks based on what the user knows about the story world.   Same lexia, but the links to continue reading is different based on what the reader knows.

Temporal development of the plot. Juan agrees that we should not really find a plot as the authors… the reader will, however, find a plot.  The reader finds a plot that was rendered on the fly during one sitting.   In a printed book, you measure by pages… in electronic text, Juan suggest measuring by time invested.  [Meaningful time invested?]

Story changes to flashback create a greater cognitive gap than a change of scene. 

The computer should keep track of what the user has done. Shortest path algorithm.  System will calculate the links between every pair of lessons.  The author only has to build a limited number of links, let the computer handle the hard work of tracking and managing the connections.

Introduced a twist — the hypertextural attractor (lexias that attract a lot of traffic, connected from many lexias, could be repeated).  In traditional hypertext, repetition of these key nodes can be a problem — IBM researcher examined the liklihood that a reader will lose interest during a navigation event; repition increases significantly the probability of losing the reaer’s attention.  

Hypertextual Friction – if a narration has too many narrative jumps, or a sequence of lexias that’s disruptive, we increase the liklihood of losing the reader’s attention.

Author, reader, computer are three actors.  We can create a dynamical system that adjusts the text according to the relationships.

Juan invoked Aarseth’s Cybertext, ergodic, noetic… Within a system, when a consciousness outside that of the reader acts this is extranoematic.

We will discover essentially the same narrative space with new tools.  If we all play Oblivion we will eventually visit the same places.  In a book, readers will acquire all the contents of the narrative space. The difference is how they are collected.

Showed graphs of interest versus time (steep drop off, slow increase.)If readers pass a certain number of lexias, they are more likely to keep reading.

Firction increases as the reader experiences the space… as we become familiar with the world, we have fewer options.

Juan says the results will be published in a few months…. maximize interest, minimize friction, recalculate each time the user clicks, which requires a supercomputer.

Implementation — we start with a lexia, adaptive links displayed based on the probability that the next link ensures narrative continuity. [How are these calculated — not by keywords, but by the shortest-path algorithm?  What would the applications of a similar tool to keep track of what page a reader clicks off of a walled website like Facebook, or on a given newspaper’s website?]

Juan notes that each lexia could include video… number of links per page could change. [I wonder… are inline links possible?] 

Mark Marnio spoke of his experience using Literatronica to write A Show of Hands

Says that Literatronica solves several important problems facing writers of hypertext.

Proglems include the sense of aporia when Aarseth is lost in a hypertext… “I’m not sure that Aarseth likes literary hypertext” — that concern of not knowing when it ends, encountering repitition, etc.  Mark differentiated his characters from Joyce and Jackson’s (which he characterized as white and upper-middle-class), saying he hoped to bring in more types of stories from a broader array of stories. 

His work is inspired by Chicano literature from the southwest of the US.   Sees Joyce drawing on “the new novel,” and wants to draw from the Telenovella — soap opera. Noted how Charles Dickens reworked melodrama; has in mind a goal to move towards a more popular form. Open up electronic literature to a wider reading audience; center on families, move away from theoretical and historical and literary texts.

Mark noted that he labeled his threads with the characters that feature in each.   “I got trapped by my own storytelling.”  Showed a graph in which different threads converge in a lexia that ends a chapter, moves through a liminal section, then opens back up again.  [Reminds me of The Heist, which describes an event from multiple perspectives, and after the reader has seen a few perspectives,then offers the reader to click on a different level to advance to the next main story event.]

Mark noted his desire to communicate initial conditions, so that the reader can track the changes in that character.  There’s a challenge in getting even experienced readers to move beyond about 50 lexias. Mark walked the audience through the process of scripting out the relationships between lexias… you really only need to add one link to each lexia (which suggests a linear relationship… the story depends on chronology, so that makes sense).