Measure of emergence through smugness and faith

-April 09, 2014

If you were given a complete list of the elements that composed earth a thousand years ago along with their proportions, dynamics of how they interact, etc, could you have predicted Twitter?

Emergent systems are composed of parts that organize themselves in such a way that the system as a unit exhibits behavior wholly removed from the behavior of the components. An explicit form of the cliché: the system is greater than the sum of its parts. The behavior of hives emerges from their bee components. Traffic patterns emerge from combinations of cars, roads, and human behavior. That Python script you just wrote emerges from a bunch of transistors.

Perhaps the most fascinating example of emergence is consciousness itself. Throw together 86 billion neurons averaging 10,000 connections each, assign them the fairly simple tasks of exciting or inhibiting each other when they pass voltage spikes among themselves and…you’re reading this article, fully aware, chock full of wonder, skepticism, joy, and angst. Who saw that coming?

Descriptions of emergence separate the phenomenon into two categories: Strong and weak. The properties of systems arising from weak emergence can be traced back to the behavior of their components.

A strong-emergent system has the incredible quality that its behavior can't be predicted from the properties of the parts that compose it. Given a set of bees, you could never predict the behavior of hives. Given neural wetware, you could never predict a daydream. Impossible, not just in practice, but impossible in principal, like making a perpetual motion machine.

In other words, strong emergence is magic.

Okay, its proponents portray it as less than magic but more than physics which, to me, means magic.

The concept of strong emergence is laughable to an engineer or scientist, right?

When you were in school, your math and science—now called STEM (science, technology, engineering, and mathematics)—homework consisted of problem sets. You were given situations and asked to predict their outcomes. Science and math are still taught this way, for the most part. But not for long.

The dilemma with education through homework problems—as you know from your experience as a practicing professional—is that only a tiny fraction of science and engineering problems have solutions that can be written down in tidy equations.

Here’s your new-school homework problem: Given the complete set of atoms and molecules on earth, a reasonable estimate of their positions, and the complete laws of nature for how they interact, what is the probability that Lady Gaga will emerge?

In practice, solving the problem is impossible with current technology and science. But strong emergence sits on the smug pedestal of “impossible in principal” so we’re not limited by the practical or practicable.

If you had sufficient computing power and were provided the complete set of details promised in the homework problem description, you could determine the solution by simulating the system many times, each with slightly different initial conditions. Since the destinations of nonlinear systems vary radically with small variations in their initial conditions, you’ll need to perform many simulations. I mean, a lot of simulations, oodles and scads, but far fewer than an infinity. And in some of those simulations, Justin Bieber will emerge. Now keep running simulations until you get that result 20 or so times and you, too, will be a belieber.

Once you have reasonable statistical samples you can calculate the probability of Tom Cruise jumping up and down on Oprah’s couch, convert it to information entropy, and, voila, there’s your measure of emergence.

Two issues emerge (as well as this pun): first, the entire concept of science and mathematics education needs to shift from solving problems to simulating systems—again, few problems can actually be solved, but every situation can be simulated.

Second, my argument against the existence of strong emergence includes standard reductionist smugness/tautology: for any system that appears to be strong emergent, I can always say, "When you understand the physics better, your simulation will predict it."

All great arguments boil down to smugness vs. faith.

Loading comments...

Write a Comment

To comment please Log In