grammar

It was only after I began working with Robert Coover in the Brown Literary Arts program in 1998 that I remembered my father commenting years earlier on Coover's book Pinocchio in Venice. As a foremost Scholar of the Pinocchio story and its appearances throughout history in literature and media, he was impressed with Coover's handling of the archive. My father went on to write about Coover's treatment in a co-authored book, Pinocchio Goes Postmodern: Perils of a Puppet in the United States. RC_AI consists of texts composed by myself and Dr. Thomas J. Morrissey, my father, along with several generative algorithms and loose grammars in collaboration with a substantial portion of Robert Coover's Pinocchio in Venice. The panoramic text is a printed array (approximately 380,000 pixels long - or 422 feet) of variable content generated by parsing through approximately 1/2 of Coover's novel using the author's name as a search string. RC_AI was created specifically for ELO_AI: Archive and Innovate the Electronic Literature Organization conference and arts program. The overall event was in part a celebration of Robert Coover who will soon retire from teaching. RC_AI was performed in the auditorium of List Art Center at Brown University with my father on June 4, 2010. For RC_AI, I utilized tesseract, an open-source tool for optical character recognition, and then created a system for text processing using python's natural language toolkit. As this is my first experiment with both tools, the implementation is basic: the former accounts for bad spelling, the latter for poor grammar (as though the puppet sold his schoolbooks for a tree of ass ears). 

The third part of Jason Nelson's artgame trilogy.

Video games are a language, a grammar or linguistics of various texts. The sounds, the movement, the graphics, the rules or lack of rules, everything about a video game is a component of language.  A digital poetry game must combine all these elements, strange and interactive stanzas, crossed out and obstructed lines, sounds and texts triggered and lost during the play.  Indeed the game interface becomes a road to inhabiting the digital poem, to coaxing the reader/player into living and creating within the game/poetry space.

Using a top down, platform engine (without gravity) Evidence of Everything Exploding is a game driven digital poem exploring various historical and contemporary texts. Each level’s poetic content is built from the document’s sub-sub texts and curious consequences. With Bill Gates’ letter to the Computer Brew Club about monetizing hobby computing, we find the seeds of an empire, James Joyce is caught in an infinite loop of changing texts, Fidel Castro’s boyhood letter to the US president praising America and asking for money signals an opportunistic future.

 With ten levels, each with video poem prizes, and strange poetic narratives, this game is a sequel to “game, game, game and again game” and “I made this. You play this. We are enemies”.

The Great Wall of China is conceived for simultaneous realisation across media, including a Website (1995-96), a CD-ROM with portfolio of prints (1997-99) and an interactive installation (1999). The foundation of The Great Wall of China is a real-time interactive language machine. This uses the metaphor of the actual Great Wall of China as a navigational device. The system is capable of creating an endless stream of ever evolving and changing texts.

The inspiration for this project began with the short story of the same name by Franz Kafka. The database for the work consists of all the individual words in the original Kafka story. There are no linguistic structures stored in the system beyond the individual words. All sentences and grammar structures are formed "on the fly" through object oriented and behavioural programming techniques, based on pattern recognition, redundancy algorithms and Chomskian Formal Grammars. Formal Grammars are used at the sentence level to generate individual sentences and ensure a degree of correctness in syntactical formation. This basic grammar system is augmented with many small ad hoc functions for dealing with plurality, conjugation, tense, etc. Most of these functions operate at the word level, but depend on "self-reading" texts and backtracking techniques. Pattern recognition techniques are used at the higher level of content generation and contextualisation. This strategy has been employed as it was the objective to avoid having any form of "story-telling" model in the system. The artist also wished to avoid using behavioural (Artificial Life) or Agent (for example, modelling a "story-telling" agent) based techniques, as the intention has been to create a system where the story, its subjects, actions and context, would emerge from the formation of the language itself, as something simultaneously written and read. Although at this point this technology is still in early development it does lead to a prose form that is very open, unexpected in its results and poetic.

(Source: Artist's statement by Simon Biggs)