ISO Studies / What A Standard Actually Is

The World Runs On Documents Almost Nobody Reads

Standards are the least visible infrastructure we have, and the most load bearing. They are agreements, not laws, and that distinction explains almost everything about how they behave.

Begin with a bolt. Somebody in one country cuts a thread into a steel shank, and somebody in another country cuts a matching thread into a nut, and the two have never met, never spoken, never seen each other's workshop. Yet the nut turns onto the bolt and holds. That is not luck and it is not law. It is a standard: a written agreement that a particular thing will be done in a particular way, arrived at by the people who will have to live with it, and adopted because adopting it is easier than not. Strip away the committee rooms and the formal language and that is the whole of the idea. Everything else is machinery built to make the agreement stick.

The word carries baggage. To many people a standard sounds like a rule handed down, a hurdle placed in the path, a thing enforced by inspectors. That reading is understandable and mostly wrong. The overwhelming majority of published standards carry no force at all until somebody chooses to give them force, whether that is a purchaser writing them into a contract, a regulator citing them in legislation, or an industry adopting them because the alternative is chaos. Left alone, a standard is simply a description of an agreed way of doing something, sitting on a shelf, waiting to be useful. Its power is entirely borrowed from the people who reference it.

What makes that borrowed power so remarkable is how far it reaches. A standard is a coordination device. It solves the problem of many parties needing to converge on one option when no single party can compel the others. Which side of the road, which voltage, which paper size, which way round the threads go: these are questions where the specific answer matters far less than everybody choosing the same answer. Economists have a dry name for this, but anybody who has carried the wrong charger across a border understands it in the body. The value is in the convergence, not in the choice.

That has a strange consequence. A great many standards are, considered purely on their technical merits, unremarkable. Somebody could design a better thread form, a better connector, a better way of writing a date. Better designs exist, and they lose, because the value of a standard grows with the number of things already built to it and a newcomer starts at nothing. The incumbent does not have to be good. It has to be everywhere. This is the quiet tragedy sitting underneath a great deal of technical work, and it is also, viewed from another angle, the reason anything works at all.

The people who write standards know this. They are not, in the main, idealists chasing the perfect specification. They are practitioners who understand that a standard is a negotiated settlement between organisations with different equipment, different customers, different sunk costs and different views about the future. The document that emerges is a compromise, and it is supposed to be. A specification that pleased only its authors would be adopted only by its authors, and a standard adopted by its authors alone is not a standard at all. It is a house style.

This publication takes standards themselves as its subject: what they are, how they come into being, where they hide, and what they cannot do. Not the business of checking whether an organisation follows one, which is a different craft with a different literature. The document itself. The strange, slow, consensual, deeply human process by which a large number of parties who disagree about most things agree about one thing, write it down, and thereby make a piece of the world predictable. That process is worth understanding on its own terms, because it is one of the few genuinely effective answers we have found to the problem of getting a very large number of independent actors to move together without anybody being in charge.