Perth.
I'm spending the next few days at the PerthDAC [3] conference here in Western Australia; I'll be presenting a paper on Sunday afternoon as well... Right now, though, it's Saturday, and we're just about to get started. Jason Lewis is the first speaker, presenting on the NextText project from Obx Labs at Concordia University in Montréal. He begins by showing a video presenting a number of interactive installations which aim to visualise everyday spoken interactions, lending a visual quality to such ephemeral interactions. Much of this is inspired by the interrelation between the structure and content of poetry (the contribution of rhyme and rhythm to the meaning-making process of poetry), as well as the use of text in comics and urban graffiti, and the experiments with layout and formatting in early-20th century avantgarde art. This produces a tight coupling of text and structure, and highlights questions of how to represent text visually, how to make use of interactive possibilities in new media technologies, and how to blur the literal and aesthetic functions of written language.

In a new media context, then, Jason suggests, computation is semantic, and helps contribute to the meaning-making process users engage in. At the same time, to do interesting things with digital text remains a process of fighting the tools, too - overcoming a continuing print-centricity (which often also means moving to a pixelised or vectorised representation of text rather than focussing on a more basic ASCII representation). Obx Labs has developed a NextText programming architecture which provides for such advanced visual experimentation with text while still enabling text to be used and manipulated as text rather than visual data. (Jason now shows another video demonstrating the possibilities - looks interesting.) Such projects raise a number of interesting questions about the interrelations between code, text, and visual representation, and the people working on those layers.

Next up is Fox Harrell, whose interest is in African diasporic orature. Such orature contains established philosophies of interactivity and performativity which may connect well with computational narrative technologies - and more broadly, this points to the processes of culture and technology mutually informing one another, as well as the cultural grounding of our understanding of technology (which introduces its own biases). The idea of African diasporic tradition is also problematic because of its multiple origins, adaptations, and transformations, however. Additionally, African orature has both performative (involving space, time frame, oral mise-en-scène, and audience-performer relationships) and integrative dimensions (connecting performance to cultural beliefs, values, and other factors shared with the audience).

Fox now presents the GRIOT system for computational narratives, which has been used especially to generate interactive, polymorphic poetry. It builds on cognitive semantics models, and includes an algorithm for generating concepts on the fly in response to user input. This utilises pre-set ontologies and phrase templates according to given narrative and discursive structures, but can also take place in a much more abstract, improvised way.

The final speaker in this first session is Su Ballard. She comes from a background in art theory and art history, and here considers two artworks by New Zealand artist Douglas Bagnall as a pathway towards developing a aesthetics of digital media. Perhaps the digital renders our idea of aesthetics mutable, and may be understood as a set of emergences. Information theory in this context may be applicable to the field of art history. For the first work, Bagnall created a filmmaking robot which uses existing telecommunications networks to create films: he mounted a camera 'eye' on the front of a Wellington bus which would transmit its recordings to the robot's 'mind' each time the bus passed by a wireless access point, and in its downtime at night the robot would generate visual 'dreams' through computational algorithms. In a second piece called Cloud Shape Classifier [4], users could train the classifier algorithm by evaluating images of clouds to select for them sets of clouds which they would find aesthetically pleasing (a play on processes of collaborative tagging, in other words).

What's interesting in the latter case, then, is the transition from subjective choices through technology to an objective aesthetic rating; it presents the aesthetic experience as an emergent event. The Classifier does not work on media or in a medium, but supports a networked, emergent digital materiality informed by its users; it never arrives at the certainty of a 'perfect cloud', but always provides the promise of something better that is just around the corner. It also unsettles the very fixed notion of information and media which we sometimes still operate with. This shifts our aesthetics, as well as our spaces of aesthetic reception, in the process turning algorithmic problems into aesthetic ones. This also affects the medium itself - digital artworks can no longer be understood as singular and fixed, but constitute instead provisional forms.

Technorati : African diasporic orature [5], Douglas Bagnall [6], NextText [7], PerthDAC [8], aesthetics [9], emergence [10], interative poetry [11], new media art [12]
Del.icio.us : African diasporic orature [13], Douglas Bagnall [14], NextText [15], PerthDAC [16], aesthetics [17], emergence [18], interative poetry [19], new media art [20]