On Standardization

In 1956, a fifty-three-year-old trucker named Malcolm McLean bought two oil tankers, cut the decks off them, and welded steel boxes on top. The boxes were thirty-three feet long, made of corrugated steel, and stackable. Nobody at the time thought it was a turning point. It was logistical improvisation by a man who had spent decades watching freight sit on docks because the loaders were on strike, or the trucks were the wrong size, or the cranes couldn't reach.

By 1970 the world's trade had quietly reorganized itself around his boxes.

What I find arresting about this story is not the scale of what followed, that has been told a hundred times. It is which direction McLean steered.

He could easily have made his container proprietary. The dimensions could have been his alone. The corner-fittings, the locking mechanism, the stacking pattern, all of it could have been patented and protected, so that any shipper using his line had to use his crates, on his terms, in his ports. That would have been the obvious move. It is what almost every entrepreneur, before and since, has done with a new format. Mine. Pay me to use it.

McLean did the opposite. He published the dimensions. He made the corner-fittings interoperable. He let competitors copy the format, knowing they would, and accepting that they would. He pushed for the standard to belong to the industry, not to him.

This was not idealism. McLean was a hard man. The decision was strategic, and the strategy was specific: he knew that a container is worth nothing until everyone uses it. A unique McLean-only container would have been a curiosity. A universal container that he happened to understand better than anyone else would be the basis of his fortune. So he gave the format away on purpose, and built his business on operating it better than the competitors who copied him.

The ISO standardization came twelve years later, in 1968. By then the container was already universal. The committee of national standards bodies did not so much choose dimensions as ratify what the world had already settled on, mostly McLean's. The official document was post-hoc legitimation. Concrete was already poured.

I am drawn to this pattern because I think it is closer to how most durable standards actually emerge than the alternative we are taught.

The alternative is: committee first, implementation later. Experts deliberate, choose carefully, publish a specification, and the world adopts. This produces some standards. It also produces a graveyard of well-designed specifications that no one ever used. They are not bad designs. They simply never crossed the gap between paper and practice, because nobody began with paper.

The other pattern, implementation first, ratification later, produced QWERTY, TCP/IP, the USB connector, HTTP, JSON. These are not necessarily the best designs available. They are the designs that someone built and others copied before the committee could catch up. In each case, the standardizing body found itself documenting an emergent shape rather than designing a deliberate one.

There is a moral here that I am cautious about pulling out too cleanly, because morals derived from history tend to flatter whoever is doing the deriving. But I notice that the people who shape durable infrastructure are often the ones willing to give away the part that everyone else holds tightly. The format. The protocol. The basic interface. Whatever it is that makes the system work, they release. They retain only what they alone can do well.

In our age this looks a lot like open-source software, and there are good reasons to draw the analogy. But it predates software by centuries. The Roman road network was widened to a standard gauge so that anyone's cart could use it, not just the legions'. The early printing press was given away as a technique within a generation of Gutenberg, because the printers needed each other's apprentices. The Phoenicians did not patent the alphabet.

McLean's container is one variation on an older shape: the deliberate release of the format, the retention of operational excellence, the long bet on the structure outlasting any one player who profits from it.

I think about this when I am tempted to make something I am building too much my own.

Not because there is virtue in giving things away, there isn't, particularly. But because there is, sometimes, a specific kind of cleverness that consists of seeing which part of the work is yours to do well, and which part is only valuable if it belongs to everyone.