The Architecture of Association
Principal Investigators
Bill Seaman, Duke University Visual Studies Program
Daniel C. Howe, New York University Media Research Lab
Project Description
I. The Architecture of Association
The Architecture of Association is a large-scale, generative artwork
that draws associative links between media elements to form an evolving
visual collage. A distributed flow of image, video and poetic text is
“intelligently” distributed over a span of 12 large, vertically
oriented monitors. As the work is emergent in nature, it does not
repeat sequences of images or texts but instead dynamically generates a
continuously recombinant network of associations. In 1995, Seaman
coined the term ‘Recombinant Poetics’ to articulate a set of generative
virtual worlds.
II. The Architecture of Association v2.0
The Architecture of Association (v2.0) develops the original concept to
create rich media landscapes from real-time associative processes. The
work will use keywords, meta-data and custom clustering algorithms to
make “informed” selections from the databases, bringing associative
material into proximity for a particular duration. These informed
selections will create a network of changing relations, stimulating
thought and reflection on key concepts poetically related to
“communication.” Thus this “collage” of associations functions as a
generative “idea bank,” with sets of relationships dissolving slowly in
across the set of screens, then dissolving away only to be replaced by
a new set of media materials drawn from the databases.
The work seeks to provide a changing set of “intelligent” associations to continually provoke thought, drawing viewers into active contemplation of their own associative processes. Where other works might diminish in interest over time, AoA attempts to be provocative in an ongoing manner, continually renewing itself in an emergent fashion. An ambient audio track, synced to the pace of the recombination engine, plays simultaneously in the space.
Joyce Rudinsky, domain scientist for the arts and humanities at
the Renaissance Computing Institute, has facilitated this project in
collaboration with RENCI technologists and coordinated the display of
the project in the UNC/RENCI Engagement Site Visualization
rooms.

