The standard and the movement

What a clock actually reports

A clock counts repetitions of a movement somebody chose in advance - the swing of a pendulum, the vibration of a crystal, a transition between two energy levels in an atom - rather than measuring time itself. We call the count “time”, but what is measured is always a movement, and what makes the measurement possible is always a reference point that someone picked.

The claim is conceptual, and the question is what follows from taking the description seriously.

Movement first, time second

Movement in space is continuous. It was, it is, and it will go on being. At a given point we can measure it; we cannot stop it. Even if we calculate badly, the movement continues, because it belongs to existence itself.

So time, as we use the word, is the capacity to perceive the duration of a thing’s movement relative to a chosen point. It is our way of describing movement.

The idea is old and it is checkable. Aristotle, in the fourth century BCE, defined time as the number of motion with respect to before and after: without change there is nothing to count, and therefore no time. Anyone is free to disagree with that position, but it stands in the text and can be read there.

One consequence gets forgotten: time frames a process; it does not determine the outcome and does not determine the duration. Processes unfold in unforeseen ways. Time gives us a frame for describing them, not a mechanism that dictates them.

The reference point is a choice

Every measurement of time has two parts: a movement that is counted, and a point from which counting starts. The first part is in the world. The second is in us.

Before instruments, people organised their lives by cycles they could see. The sun rises and sets. The shadow of a fixed object changes consistently. “When the sun is overhead” and “at sunrise” were sufficient for coordination, and the reference point was entirely local: wherever you were standing.

Ancient Egypt used sundials and water clocks, and developed a civil calendar of 365 days. Babylon divided the circle and the hour in base sixty, and that division survives in our sixty minutes and sixty seconds. Rome adopted and corrected, and the Julian calendar was established in 46 BCE. In Europe, in the late thirteenth and the fourteenth centuries, the mechanical clock appeared. Galileo observed that a pendulum’s period is very nearly independent of its amplitude, and Huygens turned that into a pendulum clock in the middle of the seventeenth century. John Harrison completed his fourth marine timekeeper in 1759, and it was tested on the voyage to Jamaica in 1761-62.

Every one of those stages has the same structure. The movement being counted got steadier. The reference point stayed a choice.

1884, and the politics of a zero

In 1884 an international conference met in Washington to choose a zero meridian. Greenwich was selected. Twenty-five nations took part, San Domingo voted against, and France and Brazil abstained. France went on working from the Paris meridian for years afterwards.

Notice what happened there and what did not. No physical fact was discovered. Nothing was measured. A vote was held. The leading argument was entirely practical: most of the world’s shipping charts already used Greenwich, and the standard adopted was the one already in use.

That is a good decision. It is also a decision, and the distinction erodes: after a hundred and forty years, a standard settled by a vote comes to feel like a property of the world.

The chain continued. In 1955 Essen and Parry, at the National Physical Laboratory in Britain, built the first practical caesium clock, counting the transition frequency between two hyperfine levels of the caesium atom. In 1967 the second itself was redefined in terms of that frequency. The base unit of time stopped being a fraction of an astronomical day and became a count of atomic movement.

What a standard owes the world

Here is the interesting part, and it needs no arguing, because it has already happened.

The atomic second is extremely stable. The Earth’s rotation is less stable, and is gradually slowing. So Coordinated Universal Time, the civil standard everyone actually uses, has had to insert an occasional leap second to stay tied to the rotation as it is. In other words: the standard was tested against the world, found to be drifting, and corrected.

That is exactly what a standard owes the world. Not absolute correctness, but the willingness to bear a test.

And in 2022 the General Conference on Weights and Measures resolved to end the leap second by 2035, because those irregular corrections break systems that depend on continuity. So the standard changes again - this time to serve not astronomy but the infrastructure built on top of it.

Three times in the same chain: the counted movement was replaced, the zero was chosen, and the correction against reality was renegotiated. No step in the chain is a discovery. Every step is a decision that was examined.

What would make a better standard

If every time standard is an agreement, then the right question about it is “what is it for, and does it still serve that”.

Three questions you can ask of any standard, not only of time:

1. What does it count? Which movement, how stable, and who verifies that stability.
2. Who chose the zero, and what for? A reference point was always chosen for a reason. The reason is checkable, and it may no longer hold.
3. What happens when the world and the standard disagree? If the answer is that the standard gets corrected, it is a living standard. If the answer is that reality is logged as an error, it has stopped being a standard and become an assumption.

The third question is the only one that really separates them. A standard that cannot be refuted is not more accurate. It has simply stopped listening.

Agreement is enough to coordinate. Understanding is what lets you choose.