Total Drek

Or, the thoughts of several frustrated intellectuals on Sociology, Gaming, Science, Politics, Science Fiction, Religion, and whatever the hell else strikes their fancy. There is absolutely no reason why you should read this blog. None. Seriously. Go hit your back button. It's up in the upper left-hand corner of your browser... it says "Back." Don't say we didn't warn you.

Thursday, January 15, 2009

The joy of emergent behavior.

As an undergraduate I took a lot of psychology and sociology classes. In some ways that has proven helpful to me as it gives me a broad perspective on human behavior. In other ways, however, it left me conflicted. I say "conflicted" because psychologists sometimes have a rather prickly attitude towards sociology- namely they have been known to claim that sociology is just applied psychology.

This perspective can be thought of as an extreme form of reductionism. For the uninitiated, "reductionism" can mean either an approach to understanding complex phenomena by reducing them to interactions of their parts or a philosophical position that complex phenomena are nothing but interactions of less complex parts. The former position is pragmatic, asserting that it's easier to understand each part in an engine on its own before trying to grasp how they all fit together. The latter position is much more aggressive, in effect claiming that there is no ineffable something about the engine as a whole that is not resident in its components. Or, put more simply, it is the perspective that the whole is not, and cannot be, more than the sum of its parts.

In sociology the phrase "reductionistic" is most often used in a derogatory manner, indicating that another researcher is trying to break things down to an absurd degree. Nevertheless, reductionism in its first meaning is effectively indispensable to the scientific enterprise. The reason is simple: if complex phenomena cannot be broken down into their sub-parts and understood as interactions between them, then we have no choice but to study the phenomenon as a whole. And since it is, as we specified earlier, complex, this is really difficult, if not impossible. So, science generally takes a chance and makes a pragmatic assumption that most complex things can be studied as interactions between simpler pieces.

Now, the reductionistic argument about sociology from psychology is that social behavior is nothing more or less than individual psychology writ large. Society is just an interaction of lots of individuals and can be understood by just applying psychological knowledge to large groups. In effect, we can understand a group by modeling every member. Sociologists, as you might guess from our apparent dislike of "reductionism" (despite the fact that we use it as much as anyone), argue that social phenomena are not entirely reducible to individual behaviors and, as such, sociology is more than just population psychology. Is this correct? Is society really defined by something that only becomes relevant at a level beyond the individual? And if so, what is this ineffable something that kicks in?

My own view is that the answer is a variant of "kinda sorta". On the one hand, I do think that social behavior is reducible to lots and lots of individual behavior. On the other hand, however, I think that modeling social behavior like that makes about as much sense as trying to predict the flow of a river by modeling each individual water molecule. In one sense, yes, if you modeled them all very precisely you would build up essentially the same picture, but it's simpler and more direct to model it more abstractly as a fluid. You still predict the eddies and whorls- the emergent behaviors of the system whose appearance is not intuitively obvious from the rules governing the sub-parts- but you don't waste your time on a lot of extraneous forces whose impact on overall flow is minimal. And this is the key to sociology: it is in many ways the study of behaviors that emerge at the group level, but aren't intuitively obvious from just studying the individual. A mob is not just a whole bunch of more or less reasonable people in the same place at the same time.

And oddly enough I'm reminded of this back-and-forth discussion by recent news that communicable disease has once more penetrated into a new corner of human life. I don't refer to sexual activity, or public restrooms, or even the handsets of public telephones.* No, I'm referring to something much more fascinating than that: World of Warcraft.

As reported by the BBC it appears that the popular MMORPG** has seen the emergence of something very much like a highly virulent, extremely lethal plague:

To give these powerful characters more of a challenge, Blizzard regularly introduces new places to explore in the online world.

In the last week, it added the Zul'Gurub dungeon which gave players a chance to confront and kill the fearsome Hakkar - the god of Blood.

In his death throes Hakkar hits foes with a "corrupted blood" infection that can instantly kill weaker characters.

The infection was only supposed to affect those in the immediate vicinity of Hakkar's corpse but some players found a way to transfer it to other areas of the game by infecting an in-game virtual pet with it.

This pet was then unleashed in the orc capital city of Ogrimmar and proved hugely effective as the Corrupted Blood plague spread from player to player.

Although computer controlled characters did not contract the plague, they are said to have acted as "carriers" and infected player-controlled characters they encountered.


Luckily the death of a character in World of Warcraft is not final so all those killed were soon resurrected.

Blizzard tried to control the plague by staging rolling re-starts of all the servers supporting the Warcraft realms and applying quick fixes.

However, there are reports that this has not solved all the problems and that isolated pockets of plague are breaking out again.

And this, to me, sums up the whole danger of relying too much on reductionism. Fundamentally, what's happening can be thought of as nothing more than the interaction of a variety of programs. Yet, at the same time, this interaction has produced effects far beyond what any designers anticipated and may be extremely difficult to correct. The plague-like behavior of this element of the game is real and, despite its origins in smaller interactions, it nevertheless has massive effects at a higher level. Moreover, modeling those effects is more appropriately done by infectious disease specialists than computer programmers, despite the fact that it is, ultimately, a software issue. Reductionism is important, without it we couldn't do our work, but often it pays not to reduce things too much.

And as long as we're on the subject, people, PLEASE don't handle strange pets in Ogrimmar. You do NOT know where they've been!

* If I recall, Douglas Adams sometimes mentioned telephone sanitation specialists as particularly undesirable occupations.

** MMORPG stands for Massively Multiplayer Online Role Playing Game which, in turn, can be taken as a code phrase for, "That thing that keeps me from finishing graduate school."

Labels: , , , , ,


Blogger Marf said...

I think a good comparison of sociology and psychology would be classical physics and quantum mechanics.

One tries to describe a large quantity of individuals, and the other tries to describe each individual. As you approach the line between them, it gets fuzzy and grey. They don't really mix, yet they both interact with each other in profound ways.

The need for a grand unified theory is felt even beyond physics, I think.

Friday, January 16, 2009 6:04:00 PM  

Post a Comment

<< Home

Site Meter