There is a branch of engineering/business/computer science that deals with very complex systems and their failure modes. Systems that are so complex that we can't really measure every part of them or understand every interaction can generally be described along two dimensions:
1) Complex systems - these are systems where there are simply many different processes or participants
2) Tightly coupled systems - these are systems in which actions are linked in long chains, meaning that a small change in one place might affect something else, far away.
A university might be a complex but loosely coupled system. There are many complicated interactions going on in different classrooms, but they are more or less independent of one another.
The Berlin Airlift, for example, was both complex and tightly coupled.
The body is generally a very complex, but fairly loosely coupled system. It is very redundant. Typically a problem in one cell doesn't affect too many other cells. One problem with MS is that it attacks one of the points within the system-that-is-the-human-body in which the coupling is tighter. The nerve cells affect many other things directly and without as much redundancy.
Maybe, like cancer or so many other conditions, everyone "has MS" to some extent. Maybe everyone has a natural balance between the reactions which, when unchecked, cause the destruction of the myelin sheathing around nerve cells. The basic problem is that some people have a biochemistry, or DNA, or a lifestyle, or whatever, that makes these reactions more common or likely. So, why do women in the 20's get MS 4x as often as men in their 20's?
Perhaps another dimension to the problem is that people have variable levels of coupling of the immune system and the nerve tissue. Maybe something about being a women ties the immune system more closely to the nerve cells. There is a precedent for women's bodies ability to differentiate between foreign and normal cells being quite different than mens'. It might make sense that women's immune systems "mature" faster than men's. If there is a tighter coupling between their immune systems and the MS-like problems, perhaps they, with the same natural change in biochemistry or rate of nerve tissue development, or a host of other factors, are more likley to exhibit the problems.
Essentially, I am proposing that things are easier to "mask" when they are loosely coupled. Loosely coupled things don't manifest themselves until they are very common. One doesn't refer to a college as a "Spanish-speaking" college until many of the interactions which go on inside it independently start happening in Spanish. But what if you were playing a huge game of telephone on campus? Then, you would expect that once a critical mass of people started speaking Spanish, the whole chain might happen in Spanish.
My dad said that he thought he had read that pregnancy triggered the onset of MS in young women. Maybe pregnancy is the key difference that increases the tightness of coupling of the changes in body chemistry to the immune system to the nerve cells to the MS condition.
Subscribe to:
Post Comments (Atom)
2 comments:
An interesting take on the "failure" model of MS, one which I will think about more.
MS is currently thought of as an immune disease and with the incidence of late pregnancy or post natal MS symptoms, there would appear to be a hormonal component to the disease as well. There are currently a number of studies going on to investigate the hormonal aspect.
At best, we can only say it is an auto-immune illness with possible links to genetics and environment.
Tightly coupled system failures has peaked my interest though.
S.
Shauna,
Are you a researcher on this issue?
I like to think of complexity as, essentially, a count of how many nodes are in a system. I think of coupling as how many connections there are between those nodes, and more specifically, the length of chains of connections that are formed.
I think some other good examples for tight coupling vs. complexity (things we often confound in our minds) might be:
Tight coupling: Making a movie. A lot of things have to be done in sequence. A script has to be written. Directors and producers must sign on. You have to get a cast. Then, once you have a cast, you need a set... and on and on. The process is very tightly coupled because things must follow one another more or less serially. The whole venture is risky because you have to be able to work the whole chain. On the other hand, the process isn't necessarily that complex. A checklist of generic steps could be written down that was more or less correct for just about any movie out there.
An example of something that's tightly coupled and complex, though, is a nuclear reactor. Not only are there many systems tied together (if pressure goes up to high, open the steam valve. If steam valve fails, open the backup valve, etc...) but there are also very complicated reactions which are nearly unpredictable. No one know exactly how each piece will perform.
One thing that I've experienced in looking at systems, is that both complexity and coupling can be a source of confusion, error, or failure. If you have a very complex system, and you build in a "safeguard," you have a new node in the system which can now interact with every other node, increasing complexity a lot.
Tight coupling, on the other hand, can actually make it easier to fix errors, because tight coupling essentially means that the sequence of interactions between nodes is less variable.
Post a Comment