< A Game Divided Cannot Roundup
Loaf Is Weird Food >

[Comments] (3) Systemantics: The working epigram for my current project comes from a book called Systemantics:

A complex system that works is invariably found to have evolved from a simple system that worked.

(The first sentence of Gall's law)

I decided to buy this book and read it to make sure I wasn't quoting a book that said crackpot things about freemasonry or something. Of course, now that I've bought it I'm able to search on a key phrase and find the whole first edition online.

I had my hopes for this book. In my mind it was a romantic relic of the age of cybernetics. It would have feedback diagrams and equations vaguely remembered from CS112. It would have lots of other quotable bits backed up by distilled essence of case study. It would be the Mythical Man-Month or the Peopleware of systems design.

The truth is more prosaic: it's a pop-psychology book. The bibliography includes pop-psych classics I'm OK, You're OK and Games People Play. It mentions General Semantics on the third page. It talks about Murphy and Parkinson and Peter--pop psychologists all--as though they were Weiner and Taylor. There are no data and not even many anecdotes, just thought experiments.

Plus it's got dumb Watergate jokes. Hey, that's real funny to you guys in 1975 but I'm reading your book in the twenty-first century. You savvy my future-speak? To us, that happened decades ago, and we have higher standards for jokes about it. (Sample Watergate joke that's still funny. Non-broken image link for same.) Nothing wrong with trying to make a book funny, but the Poignant Guide to Ruby is pretty funny, and it's also got code you can run that works.

Okay, so it's a bunch of thought experiments. That's not too bad if you run the experiments and they jive with your experience, which is true here for the most part. The book has some more good quotes ("Systems work best when designed to run downhill") as well as good ideas I'd never thought of ("the confusion of Input and Output"). I'll probably pick up the revised edition of the book, The Systems Bible, which will surely have fewer Watergate jokes. But the pop psychology emphasis really decreases the authority of a quote from this book.

I could say "A complex system that works is invariably found to have evolved from a simple system that worked." myself, and it would have the same force as quoting John Gall in this book. I've probably got more hard evidence than the book does, because I live in the future. I quote him the way you quote Mark Twain: because you like the way he put it. So maybe I should just find a quote about systems design from Mark Twain.

Filed under:


Posted by Leonard at Wed Jul 12 2006 08:13

I have read a little Stafford Beer. That's where I formulated my own modest law of systems theory (nicely confirmed by Gall and his commentators): "a system always breaks down in a way that confirms your preexisting prejudices". It is good booze though. Did he come up with the idea that the most stable system is the one with the greatest diversity of internal states?

Posted by Zack at Wed Jul 12 2006 14:26

I don't know if he came up with it, but _Designing Freedom_ is certainly where I first saw it.

Incidentally, the "remember this information" tickbox isn't working.


Unless otherwise noted, all content licensed by Leonard Richardson
under a Creative Commons License.