Cobra Effect · People and organisations
Gall’s law
Complex systems that work grow out of simple ones that worked.
6 cards, read aloud in 2:29, with a test and sources.
A children’s doctor writes a book about why nothing works.
John Gall is a paediatrician. In 1975 he self publishes a small sardonic book called General Systemantics, after 30 publishers turn it down. It is about systems. Hospitals, armies, bureaucracies, machines. One of its lines has outlived the rest.
A complex system that works evolved from a simple system that worked.
That is Gall’s law. He goes further. A complex system designed from scratch never works, and cannot be patched to make it work. You have to start over with a working simple system. He wrote it as a joke. Engineers stopped laughing around the third failed project.
Look at anything big that works.
The internet began as a handful of university computers that could send each other a message. The Wright brothers flew gliders for three years before they added an engine. A city is not designed. It is a village that kept working. The complexity came later, one working step at a time.
Now look at anything big that was designed complete.
Governments have spent billions on grand computer systems meant to switch on all at once, and switched them off instead. The new headquarters that nobody can find their way around. The reorganisation with a 200 page plan. Each was complete on paper before it had ever worked at any size.
Why the shortcut fails.
A working simple system has already met reality and been corrected by it a hundred times. A designed complex system has met only its designer. Every part contains guesses, and the guesses interact. You cannot debug a thousand guesses at once. You can debug one.
So start with the smallest thing that works. Then grow it.
Before the plan, build the version that does one job for one person, and use it. Add the next piece only when the last one is working in real hands. If someone shows you a complete design for something that has never run at all, ask what the simple working version was. If there wasn’t one, there won’t be a complex working one either.
Sources
- Systemantics, Wikipedia. The book, its odd publishing history, and the full list of Gall’s laws, of which the one about simple systems is the most quoted.
- The Systems Bible, John Gall, third edition 2002. The current edition of General Systemantics, expanded and still funny. Short chapters, one law each, drawn from medicine, government and engineering.
- Worse is better, Wikipedia. Richard Gabriel’s 1989 essay arguing that simple, working and slightly wrong beats complete, correct and unfinished. Gall’s law from the inside of software.
Nearby ideas
- Brooks’s law. Adding people to a late project makes it later.
- Conway’s law. Systems end up shaped like the teams that build them.
- Antifragility. Some things get stronger from knocks, and fail without them.
- Cunningham’s law. The fastest way to get the right answer is to post a wrong one.
- Pournelle’s iron law of bureaucracy. In any organisation, those who serve the organisation end up in charge.
- The bus factor. How many people could leave before the work stops.
- The iron law of oligarchy. Even democratic groups end up run by a permanent few.
- The Shirky principle. Institutions tend to preserve the problem they exist to solve.