What Counts as Time Without a Master Clock? The Thermal Time Hypothesis

Imagine trying to describe the universe without first placing a master clock outside it. You can still compare changes: a hand moves, a signal arrives, a particle is detected. But which variable should play the role we usually call time?

That is the doorway into the thermal time hypothesis. It is a proposal about how a physical description might pick out a time flow. It is not a claim that ordinary clocks are imaginary.

Choosing a clock raises a different question

Our article on entropy and the arrow of time asked why familiar processes have a preferred direction. Here we ask how a time parameter might be selected in the first place.

In theories that include gravity, an external, universally preferred time is not generally available. Extending statistical mechanics to that setting is one motivation for thermal time. Rovelli’s later work develops definitions of equilibrium and temperature for this problem. See his paper on general relativistic statistical mechanics.

Start with the state, then look for a flow

In 1994, Alain Connes and Carlo Rovelli proposed identifying physical time flow with a flow associated with a system’s statistical state. A statistical state describes probabilities for physical quantities. Think of an account of possible microscopic configurations, rather than a single photograph.

The proposal also assumes a mathematical structure for the system’s observables. With suitable conditions, that structure and the state define a flow called modular flow. Identifying this mathematical flow with physical time is the hypothesis. Connes and Rovelli explain the construction in their original paper.

Why does temperature enter the story?

For an ordinary equilibrium Gibbs state, thermal time follows the familiar mechanical time flow, with a temperature-dependent rescaling. Temperature helps relate the two parameters. This is more specific than saying that heat “creates time.”

Rovelli and Matteo Smerlak applied this idea to equilibrium in a stationary gravitational field. They derived the Tolman–Ehrenfest relation, in which equilibrium temperature need not be uniform across space. Their paper sets out the assumptions and derivation. This is a theoretical application, rather than experimental confirmation of the whole hypothesis or a recipe for changing the past.

Does every part need one master clock?

A related theoretical study by Nicolas Menicucci, S. Jay Olson and Gerard Milburn explored clocks for subsystems of a finite quantum system. Their protocol constructs local descriptions instead of one clock for the entire system. Read their study of clocks and relationalism.

This makes the question concrete: what does a particular clock measure, and how do we compare it with another? A model for answering that question should still be distinguished from a demonstrated device.

An SF question: who gets to set the meeting time?

Here is a fictional example, independent of BOZWARZ. Two civilizations arrange a meeting at a remote research station. One counts cycles in a reference device. The other defines its schedule through a statistical description of its environment. Both send instructions saying, “Meet after one hundred units.”

The first problem is wonderfully practical: what has to be exchanged before those instructions mean the same thing? A list of numbers may be insufficient. The visitors may also need to explain their reference system, its state, and the rule connecting its readings with events.

This is a story prompt, not a working implementation of thermal time. It turns an abstract question into a negotiation: how do people agree on a clock before they can agree on a deadline?

If this topic interests you, you might also enjoy this novel.