A Measure of Passing

A thermometer allows you to read the temperature of a system or body, but the temperature you read is not the fundamental property of that body, but a proxy to the underlying average energy state and kinetic energy of the constituents of that body.

Similarly, I argue, that a clock allows us to read time, or the passing of time, but that time itself is not a fundamental property of reality, but rather a proxy of what is truly present underneath.

In The Absent Operator I made this case structurally. Time has no Hermitian operator and has had none since Pauli proved it in 1926; the formalism of quantum mechanics has been marking time as non-foundational for a century, noted and set aside. The operator that does the work we expected of time is the modular Hamiltonian, K = −log D — entropy read one microstate at a time rather than averaged across all of them. That essay was the demolition. This one is the quieter question it leaves on the table. If time is not something the universe contains but a reading taken of something underneath it, then what, exactly, is a clock doing when it ticks? What is it counting?

The thermometer is worth dwelling on. You press it against a surface and it returns a number. That number is real — it constrains what is possible, decides whether water will boil or freeze, governs the direction in which heat will flow. But the number is a compression. Behind it lies the full distribution of molecular velocities: an enormous population of constituents moving in every direction at wildly varying energies, colliding and redirecting one another in a chaos too dense and too fast to track. The thermometer reports none of that. What it does instead is collapse the whole distribution into a single value — its mean — and hand you that. Temperature is what you get when statistical averaging is applied to something too large and too microscopic to describe directly. One legible number, standing in for a staggering complexity beneath it.

This is why a single molecule has no temperature. It surprises people the first time they meet the idea, but it follows directly from what temperature is. Temperature is a property of a distribution, not of a particle. You need the many before the concept takes hold at all. Below a certain scale of description it simply dissolves, and what remains is the distribution itself — real, but no longer summarisable in the way the thermometer requires.

Now press the same question against the clock.

The clock is built to measure precisely, and it does. Every tick is a measure, mechanically exact, calibrated to be exactly that. The question is not whether it is measuring. The question is what the measurement is of.

Consider the supposition directly: if time is discrete rather than continuous — if between one second and the next there is no half, no subdivision, no smaller part — then what the clock records on each tick is not the length of an interval but the passage of a moment. One state of the world has given way to another. Something that was the case is no longer, and something else has taken its place. The tick marks that transition. It does not measure the space between ticks, because on this supposition there is no space between them. There is only the succession: one moment arriving as the last departs, hard against it, with no continuous medium threading through.

The clock, then, is a transition-counter. It is calibrated to that task and performs it faithfully. But what we call time is not the transitions. It is the name we give to their ordering. The succession is real; the counting is faithful; the transitions the count is a count of are real in their own right, prior to any counting at all. Time is the label we lay over the whole arrangement — and somewhere in the laying of it, the faithful counting of real transitions becomes the sensation of something flowing.

This is a different and stronger claim than saying the clock measures irreversibility. In The Arrow of Time is a Perception I separated two things we habitually run together: the entropy gradient, which is structural and observer-independent, and the felt flow of the present, which is the mind's rendering of it. The cut I am making here is finer still. Irreversibility describes the character of the transitions — that they run one way only — and it sits one level up. What the clock records, tick by tick, is something more basic than direction: the bare fact that a transition has occurred at all, that a moment has passed from being present to being past. The arrow follows from that. It is not the primary thing. The primary thing is the succession of passages; the direction is a consequence of counting them.

The thermometer parallel now holds more exactly than it did at the outset. The thermometer averages over a distribution of velocities and returns one number. The clock counts a succession of transitions and returns another. In both cases the instrument is real, the reading it returns is real, and the concept we reach for — temperature, time — is the name we have given to what the instrument is doing, not a thing that exists behind the instrument independently of it. Both numbers are genuinely informative. Neither is the bottom of the description.

The uncomfortable consequence arrives at the same depth in both cases. If a single molecule has no temperature — if temperature becomes definable only once the distribution is large enough to average — then at the level of description before the averaging, there is no temperature. Not faint temperature, not approximate temperature: none. The structure that will be averaged into a reading is present; the reading is not, yet. Apply the same logic to the clock. Before the transitions are counted and ordered, before the succession has been compressed into a coordinate, there is no time. Not slow time, not thin time. The transitions are there. The count of them is not, and the word we fasten to the count even less so. This is the same relationship the last essay named in operator language — entropy as the average of the modular Hamiltonian's eigenvalues, the one object read at two resolutions. The clock works at the coarse resolution. It reports the average and calls it duration.

I have leaned on the Wheeler-DeWitt equation more than once — in The Distance Between Events, and in both of the last two essays — so I will not re-walk it, only state the one line that bears on the clock. The central equation of canonical quantum gravity carries no time term at all. The universe written at the deepest resolution we possess is a condition of balance, not a sequence — a constraint, not an unfolding. Time appears only further up, when physical subsystems compare their states against one another and name the comparison duration. It is relational in the most literal sense: it exists between things, as a measure of their mutual transitions, and not as a background they transition through. The clock is one of those subsystems. It does not stand outside the succession to measure it. It is a piece of the succession, arranged to count the rest.

What I find compelling about both instruments is that their usefulness depends entirely on their being proxies. The compression is not a flaw. It is the whole point of them. The full microscopic description of a gas in a cylinder is useless for building an engine; you need temperature. The full account of every transition in a physical system is useless for keeping an appointment; you need time. The proxy strips away what cannot be acted on and hands you what can. The interface is what makes the structure underneath it habitable for creatures of our scale and kind.

But habitability has a cost worth naming, and it is the cost I keep circling back to. We are built from the same structure we are measuring, which means we never measure it from the outside. The human body is a thermometer — it holds an internal temperature, senses heat, runs its metabolism on thermal gradients. It is also a clock: circadian rhythms keyed to the solar cycle, repair mechanisms that degrade one way only, memory laying down the past as a structured residue of transitions that cannot be undone. We are not observers of time standing apart from it, watching it go by. We are successions of transitions that have acquired the capacity to count themselves, and the self-count is what we experience as duration.

The felt passage of time is therefore not a fiction, and it is not a clear window onto the base of things either. It is the rendering produced by an organism made entirely of transitions, reading its own passage from one state to the next and finding in that reading a sense of forward motion. The rendering is accurate. What it renders is real. But what it renders is not time as a primitive dimension of the world; it is the succession of moments that "time," once we name it, is our name for.

The thermometer does not deceive you about temperature. It simply cannot tell you about the individual molecules, and it cannot tell you why heat flows the way it does and not the reverse. The clock does not deceive you about the order of events. It simply cannot tell you what a transition is, down at the level where transitions are all there is — before the counting starts, before the label is fastened, before the compression returns its number. Both instruments earn their keep. Both hide exactly what they cannot show. And what is hidden underneath them is not less real for being hidden. If anything, it is more.