'the data and tension' is the working title for a composition for two or more untrained vocalists. it is a realization of the composition i wrote for two or more computers connected by a local area network: blog-entry, recording. the participants read sections from a shared body of source text related to networks (as political, aesthetic, and communicative entities). each participant has her own rule set, which relates certain 'trigger words' to various places in the text. when a participant hears another participant say one of her 'trigger words,' she must jump to that section and begin reading. in reading the section, she may choose to start at a point other than the beginning of the triggered section, so long as she goes back to the top of the section and finishes the remainder. the piece begins with one participant reading, and i have not seen this ruleset yield any ending scenarios.
the participants generally found the composition fun and engaging. while observing the unfolding process was equally fun and engaging for the composer, i imagine the true audience of this piece to be the performers, since they are the only ones who get to experience the process first-hand. that said, there is room for performative components if they happen to emerge.
the source text was compiled from notes i had taken from classes, edited together by hand. the version i used in trials 0 and 1 may be found here. the version for trials -1 and prior were broken (hence their negative status) and are not available. i will be adding to this text as time goes by.
i divided the text into four sections of equal lines: s0, s1, s2, s3. four keywords were selected for each performer, one for each section. to facilitate this i wrote a simple word frequency counter not unlike the one found here. my version can be found here. to find the most frequent words from the list my scripts generates, i used the 'head' command in unix. then i chose 'trigger word' lists by hand for each participant. i chose these lists so the average frequency of all their words would be close to equal. the lists are available here: t0, t1, t2, t3.
to realize this version, give each participant a copy of the four sections of the source text (the 's' files). also give each participant a unique 'trigger word' list (the 't' files).
we found it was best to have everybody share a single copy of the source text pages, printed and laid out in front of them. we actually used two copies, so they could sit in a circle. we found this formation improves the group dynamics.
a larger source text would be nice because it would allow me to use more meaningful 'trigger words'. since my sample was so small, the frequency drop-off between words like 'a' and 'the' to words like 'data' or 'tension' was too great to omit the less meaningful words. this also made the realizing process more difficult, as one is trained to pay less attention to words of the first category.
while i was still in the planning phase, i had considered applying a delay to the vocalizations, to make the piece more difficult to perform. it was quickly discovered that the composition needed no help being difficult, and the idea was abandoned.
after a few trials, we started talking about putting more constraints on how long the triggered reading section should be. perhaps if, on hearing a 'trigger word,' the participant could read some number of contiguous (wrapping) lines from the appropriate section, instead of the entire section. we have yet to experiment with this parameter, although it should introduce a little more space into the realizations as participants wait for their 'trigger words.'
overall, the source material seemed appropriate and did not alienate those less familiar with networking protocols. instead, the group dynamic allowed for an engaging experience and participants seemed able to synthesize the content with the act of performing, which caused a few laughing fits.
an excerpt of the composition can be found here.