Some Systems Concepts



  • If you have a more developed interest in Systems, I think the book Introduction to Systems Philosophy by Laszlo is excellent. It is peripheral to most discussions in this site and probably is a bit much for most people to read, but I would suggest it for anyone who checks it out and finds it interesting.  Another book that I think is very important for people that like to investigate science and the basis of knowledge is by Rescher called Methodological Pragmatism.  Both of these books were written in the 1970s and while there may well be better references, I have yet to come across them.



Positive and Negative Feedback. How do I define these ? Remembering my brief experience in undergraduate philosophy courses I would define positive feedback as feedback that is not negative.

I'm sure you find that as satisfying as I do.

Rather than give abstract definitions I will try to define them with some examples and then try to find someone else's definitions that I can refer to later.

Think of yourself when you get up in the morning and drag yourself to the shower. You turn on the water and sort of aim for a familiar hot/cold setting. After the water runs for a little bit you stick your hand in the stream of water and fine tune it to your liking. Your interaction with the hot/cold water system in your house is a slightly bigger system including you. You are the controller. Initially you are using positive feedback within this human/water system when you turn the water on. There is some visual and memory feedback so that you get a reasonable initial setting. Then when you stick your hand in the stream to sense the temperature and adjust the settings you are using negative feedback to get a desired equilibrium.

Maybe another example will help. You turn on your car radio and it is playing your mom's favorite station. You gag momentarily and then reach over to the dial to change it to a station you like better. You listen for a second and then reach over to more finely tune it while you're listening. Your impulse of energy to change the station is positive feedback to the system. The negative feedback occurs when you fine tune the setting while listening to the radio to get a clear signal.

Finally, you get in an elevator on the ground floor and press the button for the floor you want. When the elevator moves towards the floor you want it is a sort of positive feedback to the overall system. While I don't know anything about the actual engineering inside elevators, assume this motive force is provided by a cable being moved by a motor from above. When you near the floor, some form of resistance is applied to the elevator car and cable so that the floor of the car is level with the floor of the building as the door opens. Consider this to be negative feedback to the system.


An Overdamped System. Think of all the times that you have ridden in an elevator and thought your grandmother could beat you to your floor by taking the stairs. This is an example of an overdamped system. The resistance might be so great to ensure that the car doesn't overshoot its goal that the whole car moves very slowly.

You want to know what school the engineers that designed the elevator attended so you can root against their sports teams.

The diagram on the left shows how the green line very slowly approaches the equilibrium goal of the yellow line.


An Underdamped System. You take the elevator to your desired floor and it moves real fast pressing you to the floor as it accelerates. When the floor indicator shows the floor you want you feel the car decelerate and you actually lift off the floor of the car slightly. Then you drop back to the floor as the doors open and you notice that the floor of the car is bouncing back and forth, above and below your desired floor of the building. You jump out of the car and remember to take the stairs on the way back down.

This is an example of an underdamped system. Eventually it will reach the desired equilibrium but it bounces around quite a bit before doing it. It moves towards the goal too fast to be able to smoothly converge.

The diagram on the right shows the diminishing oscillations towards the desired yellow line.


A Critically Damped System. This is an elevator you like to ride. It gets you there fast with no bouncing around.

The diagram on the left shows how the green line quickly approaches the equilibrium goal of the yellow line.


An Unstable System. You take the elevator to your desired floor and it moves real fast pressing you to the floor as it accelerates. When the floor indicator shows the floor you want you feel the car shoot on by. You reach your apex and then you drop back past your floor. You keep on oscillating like this. At least you didn't shoot through the roof !

This is an example of an unstable system.

The diagram on the right shows the oscillations that don't move towards the desired yellow line.


Systems Control. The diagram on the left describes three axes (plural of axis) of control. One's self, other people, and for lack of a better word, automation.

Think of a system as everyday as the human and hot/cold water system used for taking a shower. While specific aspects vary from place to place, there are many common features. The person taking the shower directly influences or controls the water temperature. But if you really think about the entire system, the person taking the shower doesn't exert as much control as they might take for granted. The hot water heater must be working effectively, an aspect of system control by "automation". All the pipes must be working and you need a steady and high quality water supply as well as many other things. Since these sorts of systems tend to operate with very little intervention by the user or "other people" that they hire for repairs they can largely be taken for granted.

At one point, while feeling rather exposed about my current lack of breadth in systems, I was running through the library trying to find books with titles relating to "autonomic control".  I thought I could remember this as an important concept when I was going through school twenty-some-odd years ago and more.  All I kept finding were books relating to "automatic control".  I think this says far more about the directions of this society than I can appreciate.  It's also probably related to why a "robot" can become "Person of the Year" in Time magazine.


The human/elevator system is quite a bit different. You push a button to gain access to the elevator and then you push a button that represents your desired floor. Almost everything else is done by automation. Though a rider must trust that the designers, builders and service people have done a good job.

I'm sure everyone has experience with elevators that don't quite work as desired or that require a lot of intervention. A lot of things must be working right in order for the entire system to run effectively.


Upside/Downside. Probably one of my greatest bugaboos concerning systems is when the people developing a system pretend there is only an upside to their approach. I think this is particularly problematic because the people developing a system should be the ones that are most capable of evaluating the downside and minimizing its influence or deciding an entirely different approach should be taken. If they are pretending there is no downside this is not going to happen. The problems with these situations are only compounded if they refuse to talk openly with others about potential downside problems or different approaches that could be taken.

I will start with my usual examples.

The human and hot/cold water system in your home is largely a stable design with very little downside compared to the upside. You can get leaky pipes, or if you live in a cold climate they can freeze and break in extreme circumstances. Regardless of this, almost everyone would agree the system and its risks are far preferable to pumping water from your own well and heating it for a shower, using an outhouse or many other things. While the system has other vulnerabilities in extreme situations, it is very difficult to evaluate the actual risks and the returns for trying to compensate for all these risks.

There can be other downsides. For example, when growing up, I had an aunt and uncle that had a retirement trailer on a lake in northern Michigan. Each year came the "blowing of the pipes" ritual before they would close up the trailer until next spring. The kids would run around the trailer and open all the faucets and our uncle would go outside and sidle up to each of the outside faucets and blow real hard to clear out any water in the pipes. It was funny, but I have to believe he and my aunt were just acting, though I have to admit I was never really sure. Whatever they actually did, it didn't always work and if it really ever worked I know where I got the genes for my big mouth.

Probably most everyone has heard some horror stories about people that bought a home where "new and improved" materials were used for pipes that weren't as good as claimed. But again, these are pretty isolated instances.