System Engineering

In this issue of “Intractable Problems” we wish take a close look at systems engineering. In many cases, proper system design requires that parts of the system be prototyped or tested through controlled experiments.

Effective experimental design was first pioneered by Finagle. He is the person who published methods for introducing fudge factors, rounding errors, and various other artifacts into our experiments. Without Finagle’s methods we would still be struggling to design many of the more complex systems that we use today. Some of Finagle’s most important contributions to experimental knowledge are summed up in his four laws, listed below.

Finagle’s First Law: If an experiment works, something has gone wrong.

Finagle’s Second Law: No matter what the experiment’s result, there will always be someone eager to: (a) misinterpret it. (b) fake it. or (c) believe it supports his own pet theory.

Finagle’s Third Law: In any collection of data, the figure most obviously correct, beyond all need of checking, is the mistake.

Finagle’s Fourth Law: Once a job is fouled up, anything done to improve it only makes it worse.

Finagle relied quite heavily on the previous work by Chaintok in developing his theory of change. Chaintok examined data on over 164 different engineering projects to understand what happens when you change the design specifications. Chaintok summarized the results of his work by proposing the various laws of revision, stated below.

First Law of Revision: Information necessitating a change of design will be conveyed to the designer after, and only after, the plans are complete.

Second Law of Revision: The more innocuous the modification appears to be, the further its influence will extend and the more plans will have to be redrawn.

Corollary to the First Law of Revision: In simple cases, presenting one obvious right way versus one obvious wrong way, it is often wiser to choose the wrong way, so as to expedite subsequent revision.

Some other results, usually published by students and thus not verified, are given below.

Wyszkowski’s Law: No experiment is ever reproducible.

Wyszkowski’s Second Law: Anything can be made to work if you fiddle with it long enough.

Fett’s Law: Never replicate a successful experiment.

Law of Research: Enough research will tend to support your theory.

Maier’s Law: If the facts do not conform to the theory, they must be disposed of.

Peer’s Law: The solution to the problem changes the problem.

Once again, if any of you have any insight into the inner meaning behind all this, please enlighten me.

Next issue, we will have a more serious look at language and some of the problems involved in understanding ourselves.