Seven Red

A remarkable study was published yesterday in the Proceedings of the National Academy of Sciences. The researchers were seeking empirical explanations of collective animal behavior, specifically, the swarming and flocking behavior of starlings. How do starling flocks manage to maintain cohesion, even with rapid directional changes. How do the flocks self-organize? These sorts of studies have exploded over the last ten to fifteen years, as researchers in fields as various as biology, physics, economics, robotics, graphic design, and (humbly) rhetoric have sought to make sense of crowd and swarm phenomena.

The answer to the flocking problem has usually been presented as a “nearest neighbor” operation, which is to say, each bird functions according to a specific algorithm that “programs” responses based on the actions of several other birds in the vicinity. The key here is that flock behavior (and other sorts of crowd behavior, we might suggest, though serious scientists would tell loopy humanities people like me not to stretch empirical work into vague analogies) relies on metric distance: how close birds are to other birds. The current study overturns that long-standing position, arguing instead that flock behavior relies on “topological distance,” or the relative position of, and number of birds between, six to seven other birds, regardless of metric distance (within a given range, of course). This explanation – based on a method of empirical observation and 3D modeling – would seem to account for fluctuations in flock density in a way that metric distance would not.

I just read through the study, so I’ll have to turn over its implications before I irresponsibly and haphazardly apply them to rhetorical interactions (more motion, Mr. Burke!) , but I do find the notion of topological distance interesting. First, I think this causes big problems for prevailing cellular automata theory that continues to operate on nearest neighbor principles. Second, we can now see flocks as cut through with numerous dispersed packs; on a conceptual level, this makes more sense to me than a kind of jigsaw puzzle aggregation of nearest neighbor units. A new day in the study of immanence? Maybe!

For now, Starling Formations over Termini Station, Rome (the study took its starling data from this location):

Trackback URI | Comments RSS |