Two Factories, One Dimension

Work was being delayed on Ford’s assembly lines. Workers blamed the dimensions of ball bearings supplied by the Swedish company SKF. The supplier disputed the complaints and proposed checking its parts with Johansson gauge blocks—precisely made pieces of metal. In Simon Winchester’s account, the check redirected the complaint: the error was on Ford’s side.

The episode in chapter 5 of The Perfectionists explains more than a general appeal to make things more accurately. A factory can successfully assemble its own parts and still misjudge somebody else’s. The fact that a system’s components fit one another does not establish how that system will connect to another one.

Winchester describes Ford’s realization in terms of that uncomfortable possibility: his production might have been consistent with itself. While all the parts came from within it, the problem could remain unnoticed. Another manufacturer’s bearing brought a test that internal consistency alone could no longer satisfy.

A Dimension You Can Assemble

A gauge block is not another car component. Its purpose is to supply a known length against which something made can be compared. Johansson’s sets allowed different lengths to be obtained by combining several precisely manufactured elements. Instead of needing a separate reference for every required dimension, a user could assemble the appropriate combination.

In this story, the blocks change the terms of the dispute. Ford workers see delays on their line. SKF sees a bearing it considers correctly made. Both observations can be sincere, but they concern different things: the performance of an assembly operation and a component’s dimensions. A shared measure allows a more precise question than whose factory is better.

In the preceding chapter, Winchester shows why even reading a dimension requires a carefully designed procedure. A ruler’s lines have a thickness; someone has to judge where the object ends in relation to them. Joseph Whitworth uses a different principle. An object is held between two plane surfaces, which are gradually moved apart until it comes free and falls. Their positions are tied to the movement of a measuring screw.

This is not measurement without a human operator or the magical removal of every error. But the basis of the decision changes: instead of judging an edge against a line, the operator watches for a particular event between two surfaces. Precision comes not merely from a craftsperson’s attentiveness but from a device and an action that can be repeated.

Checking the Connection

The bearing episode becomes especially revealing beside Winchester’s explanation of the assembly line. If a component will not fit and a worker starts trying to make it fit, the delay spreads along the line. Something that could be slowly adjusted by hand in a different production setting interrupts other people’s work in continuous assembly.

What matters is therefore not only how carefully an object has been made. It also matters when its dimensions were checked, and against what. Discovering only at the point of assembly that two parties understood correctness differently is already too late for uninterrupted production.

Consider a simpler, hypothetical case: a workshop orders a component that does not fit its product. Returning it with an instruction to make it more accurately is not enough. The parties need to establish which dimension differs, what procedure checks it and what each accepts as a permissible deviation. This is an illustrative application of the book’s question, not a technical standard or a report on present-day factories.

In Winchester’s account, Ford eventually brings in Johansson himself and his gauge-block business. The more interesting step for a reader comes before that decision: the customer agrees to examine not only the unfamiliar part but the system it fails to fit. What began as a single accusation against a supplier becomes two things that can be checked.