Being Unique...

Just Like Everyone Else

Cars jam bumper to bumper on the freeway, birds fall into perfect formation as they wing their way north for the summer, fish mass together in a school as they swim about: the world about us is full of patterned behavior and activity. From whence do these patterns arise? Does a great sysadmin in the sky whisper flight control orders to geese?

*squawk* Rogue-6, you are out of formation. Come left to heading 110.76, 215.47, confirm. *zzzzt*

*Bzzzt* Copy that Rogue Leader, altering flight path to conform. *crackle*

The idea is amusing, but why then, do geese fly in their familiar vees? It may perhaps be within the realm of possibility that a dominant alpha-male goose takes charge and orders everyone into position, but that does seem to put a rather grievous strain on the mental capacity of a goose. Furthermore, insects, including termites and ants, show similar organized behaviors. It seems unlikely that such organisms possess sufficient initiative and intelligence to coordinate maneuvers, and it might be argued that the majority of Southern California drivers exhibit even less acumen during rush hour. The answer to our dilemma may lie in the simpler, but rather subtler, idea of emergence.

In July of 1987, Craig Reynolds published a paper entitled Flocks, Herds, and Schools: A Distributed Behavioral Model. More or less the seminal work in the field of decentralized thought, Reynolds' paper set forth the postulate that, in many systems, order emerged from chaos through the decisions of individuals following simple internal rule sets. This idea was seized upon by several innovative minds in the fields of mathematics and computer science, undergoing rapid evolution. Mitch Resnick, at the MIT Media Labs, Moshe Sipper, and Luigi Paglianari all contributed to the growth of this concept, which reached its pinnacle in the 1999 work by Kendra Knudtzon and Jude Battista, in their ArtiFishial Life project. In its most basic form, the theory of decentralized behavior states that complex group behaviors may arise, not from the leadership of any one individual, but rather from many local interactions among group members. Returning to our earlier example of a gaggle of geese, the bird at the point of the vee is not the gaggle's leader, but in the words of Mitch Resnick "it just happens to end up there" (Beyond the Centralized Mindset). Instead the geese may achieve their formation by each goose responding to simple desires concerning only its immediate neighbors. In a similar fashion, traffic jams may arise, not from that oft-cursed motorist who slowed down ninety minutes ago, creating a braking pattern that propagated though time to create the nightmare that is I-10 at about 5:00 in the afternoon, but rather through the individual responses of each motorist to commuters nearby. A hierarchy of levels is important to this concept as well: a traffic jam, though composed of cars, behaves entirely differently than its component automobiles. The traffic jam, or gaggle of geese, or formicary of ants, is known as an emergent object. Such an object has no guiding or motivating force integral to itself as an individual, but arises from and acts out of the decisions made individually by its component parts. One particularly ironic example of this involves the student body of most high schools. Unless teenagers have dramatically changed since the days when I numbered amongst their ranks, a disproportionately large number of adolescents are driven to step outside of social norms (in particular, those laid down by parental figures). One of the dominant ways this spirit of nonconformity is expressed, particularly amongst those who actively consider themselves free spirits of art, rebels, and the like, is in the choice of raiment. The wardrobe of counterculture has a look all its own, often generated in opposition to the prevailing standards of good taste. However, wandering the hallowed halls of the local edifice of education, one quickly notices that all these rugged individualists look exactly the same! Distinguishing one from the next is almost as bad as walking into corporate America and attempting to differentiate the suits from one another. I am not yet such a hardened cynic as to suggest that these rebellious souls have simply joined another establishment with all of the restraints and conformity that entails, but perhaps this aggregate similarity is another emergent pattern. By individually selecting attire in directconflict with social norms, these high school students have, albeit unwittingly, created an emergent style of dress.

The nature of decentralized behavior makes it hard to prove or disprove as the source of a pattern: termites are not known for their responsiveness to interviewers. Further compounding the problem, even intelligent organism, nominally including human beings, are generally unaware of any emergent behavior they engage in. So dominant in our society is the centralized world view, that Resnick initially encountered severe resistance to his ideas when presenting them to high school students engaged in his Starlogo program. The power of the cultural conditioning undergone by these children, and by extension society, to allow them to disregard suggestions by an authority figure of the magnitude of an MIT professor, is immense indeed. The most effective way that Resnick found to give credence to his ideas was to produce computer models, wherein local interactions between individual components gave rise to exactly the sort of complex behavior that emergence theories predicted. The idea of such modeling was not new: in his initial paper on the subject, Reynolds supplied the a set of rules for allowing directionally oriented computer particles known as boids, to emulate a bird flock. Reynolds managed to achieve a fairly realistic flocking simulation with only three simple rules each boid follows:

1. Avoid collisions with other boids

2. Match velocity with nearby boids
3.Attempt to stay close to other boids in the flock.

Naturally, the implementation of said commands became rather more complex than their formulation, but Reynolds' simulation did offer substantial support for the supposition that biological organisms could exhibit similar decentralized organization. Research along similar lines has uncovered a great deal, not only about the simulations, but about the modeled complexities as well. Reynolds, for example, found that bird flocks appear to have a constant complexity of interaction, no matter how many birds join up. Linear complexity adds credence to the idea that flocking is a decentralized phenomena; if there were one invisible hand guiding the flock, it would have to take into account each and every member of the flock at all times, creating at best an o(n) and more probably an o(n^2) complexity. If, instead, each bird merely has to keep track of its neighbors while abiding by as few as three rules, linearity becomes quite achievable. Since 1987, a great deal more work has been done with modeling. Projects' complexities range from Reynolds' creation of flight models for individual boids, and Terzopoulos' intricate modeling of the physical properties of fish, down to one page Starlogo programs created by Mitch Resnick's high school students.

The idea of decentralized control has been gaining adherents ever since its inception. Perhaps because of a superficial similarity between emergence and chaos theory in mathematics, the rise of the latter in the public imagination seems to have spurred popular interestin the former as well. While a centralized view of the world remains deeply ingrained in the public consciousness, emergence is beginning to receive both scholastic and public recognition as an alternative way to view the world we live in.