Order and Disorder


IEEE Spectrum Magazine, July 2003


I love it when my little two-year old granddaughter visits my home. There is a small price to pay, however. In the space of about a half hour she can reduce the house to shambles. With great energy and enthusiasm she rushes about, pulling things from shelves and tables and throwing them on the floor.

Last time I got down on my knees and looked her in the eyes. I accused her of being an agent of the second law of thermodynamics. I think she gave me a knowing wink, but maybe it was my imagination. Patiently, I explained to her that she was increasing the entropy in my house. I said that the second law required that disorder should increase and spread in a closed system like my house. Perhaps it needed a little occasional help from someone like her.

It was hard holding her attention. There were so many more things that needed scattering, and so little time. She rushed off. But I started thinking about order and disorder. Apparently young children have no sense of the desirability of structure in the world about them. When do we acquire this sense, I wondered, and do we engineers have a special attraction to an ordered world?

An adult wouldn’t tolerate all this junk on the floor. With maturity we learn that it is better to collect it all and throw it in the bottom of a closet. Then the world that is visible would have order, while the disorder would be confined to the unseen world -- sort of like the registry in Windows, I was thinking.

I sometimes think that our job as engineers is to impose structure on a resisting natural world. We’re happy when we’ve reduced a complicated problem to a simple model with underlying equations. We see beauty in a printed circuit board with a pleasing rectangular layout of parallel lines. We like our graphs to be straight lines. We praise structured code, and the worst description is to liken code or wiring to spaghetti. We work relentlessly against the second law.

I think of the beauty of Steve Rice’s mathematical theory of noise or of Shannon’s classic derivation of a simple capacity equation for a random channel – the creation of virtual order from natural disorder. Lotfi Zadeh defined “fuzzy logic” and reduced it to mathematical analysis. Chaos theory – which seems like an oxymoron – became popular when we discovered strange attractors and predictable behavior.

Where other people see a mess, we see the possibilities for a descriptive underlying structure. I look at the floor of my living room, now strewn with debris. For a moment I think like the engineer I am. Perhaps I could model this. She seems to move at a fixed rate. At a given moment only so many objects are within reach. This number decreases with time as objects join their companions on the floor. She can’t throw things very far, and her throws are regular in length and random in direction. The most recently strewn objects lie on a circle about her present position. This is starting to take shape. Now I’m happier with the mess. It really obeys laws, so it’s ok.

I think we’re uncomfortable when problems don’t have neat solutions. When the real world frustrates us, we make assumptions and propose simple models that may or may not capture the true behavior. We grow up with courses where every problem has a solution. I imagine the heresy of a textbook where some of the problems don’t have solutions. You look in the answer book, and it says, “This problem cannot be solved with the information available – welcome to the real world.”

Of course, the second law of thermodynamics really only applies to a closed system, like the universe. So order can be achieved locally at the expense of disorder elsewhere. I think engineers are like this too. While our work seems to involve the creation and imposition of structure, our environment often disintegrates spontaneously around us. This makes the second law very happy.

I walk down the halls of my lab and peer into the offices. I can tell which are those inhabited by engineers. Those are the offices where you can’t find the desk for the piles of paper and equipment. Sometimes I can’t find the floor either. On the other hand, when I see a neat office with a clean desk, I know that this is an office housing a financial person, a human resources person, or a top executive.

Perhaps we can’t evade the second law after all – structure in our work, but chaos in our environment. Speaking of environment, someone has to pick up all this junk on my floor. Alas, it looks like a job for an engineer.


Robert Lucky