How three principles force the universe to build time out of events
Most of us, if pushed, will say that time is a kind of universal stage on which events play out. There’s a clock running somewhere underneath reality, ticking forward whether anything happens or not, and our lives — and the lives of stars and atoms — unfold against this background. This intuition is so deep that it survived even Einstein, who wove time into spacetime but kept it as a dimension — a feature of the fabric, something that exists prior to and independently of events.
A new VERSF paper, Discrete Commitment and the Necessity of Tick–Bit Structure, argues that this intuition is wrong. And not wrong as a philosophical preference, but provably wrong, given three principles that any observer-invariant theory of physics has to accept. Once you accept those three principles, you cannot also accept time as a dimension. You are forced into a picture in which the universe constructs time out of events, rather than locating events in time.
The three principles
The paper begins from three foundational commitments — what we call the kernel. None of them is exotic. Each is something any honest physical theory has to grant.
The first is observer-invariance: physics has to be the same for every observer. Two observers may use different coordinates, different units, different reference frames — but if something is a fact, it has to be a fact for both of them. No observer’s truths trump another’s.
The second is finite resolution: any bounded region of reality has only finitely many distinguishable possibilities. You can’t probe a finite system to infinite resolution. There is no measurement that registers an uncountably infinite list of distinct outcomes inside a finite chunk of space.
The third is irreversibility: facts, once produced, cannot be unmade. The past is fixed. Once a measurement has happened, once a record has been written, once an event has occurred, you cannot reverse it back into nothing.
These are the three. Almost everyone working in foundations of physics accepts all three. They are weaker than quantum mechanics, weaker than relativity, weaker than thermodynamics. They are the kind of thing you would put in the prologue, before the actual physics begins.
How the kernel rules out time as a dimension
Now ask: what does it actually take for time to be a fundamental dimension? It would mean that time is a continuous parameter along which events are positioned. Between any two moments, there is another moment. Inside any finite stretch of time, there are infinitely many distinguishable temporal positions.
But the second principle — finite resolution — forbids this. A bounded region of reality cannot have infinitely many distinguishable possibilities. So either the temporal positions inside a finite interval are not all genuinely distinguishable (in which case “time as a continuous dimension” was never doing the work; only the distinguishable events were), or they violate the principle directly. The continuous-dimension picture cannot survive the finite-resolution principle without becoming mathematical scaffolding rather than physical reality.
A defender of time-as-dimension might try to retreat: “Fine, time is a continuous parameter, but only finitely many facts are produced inside any interval. The dimension is real but the events on it are sparse.” This is the most sophisticated version of the proposal. The paper closes it too. The trick is to ask: where do those events sit on the dimension? If different observers can disagree about the temporal positions of facts, the positions aren’t observer-invariant, and the first principle is violated. If observers must agree, then the only physically meaningful structure is the discrete arrangement of events themselves — the continuous dimension underneath has become redundant. The events are doing all the work; the dimension is just a coordinate system on top.
That leaves the question of order. Couldn’t the order in which events happen come from a continuous time-parameter sitting underneath them, even if the events are discrete? Here the third principle — irreversibility — closes the last loophole. The order of facts on a worldline (which fact happened before which) is itself a fact. By the first principle, that order has to be observer-invariant. By the third, it is fixed once the facts are made. Put these together, and the order is constituted by the irreversible production of facts, not by their position along an external clock. The order is what irreversibility produces. It does not need a clock underneath.
That is what it means to say time is not a dimension. There is no clock underneath. The ticking of the universe just is the production of facts.
An analogy
Imagine you’ve been told that the order of frames in a film comes from the projector’s clock — there is a master timing mechanism, and each frame happens “at” some moment on that timing. Now suppose someone shows you that the projector’s clock is itself built by the frames running through it. The frames don’t happen at moments on a clock; the clock is built by counting how many frames have passed. The “time” the clock reads is just a tally of the events.
That is, structurally, the picture VERSF is forced into. What we call “time” is not the projector. It is the count of frames that have run through. Each frame is a moment of fact-production — a commitment event, in the paper’s language — and time is what you get when you string the commitments together along a worldline. Time is something the universe makes by producing events. There is no separate substance flowing past underneath.
What’s new
The relational view of time — the idea that time is constituted by relations between events rather than being a stage on which events occur — is not new. Leibniz argued for it against Newton in the 17th century. Mach pressed it again in the 19th. Reichenbach, Whitrow, Julian Barbour, Carlo Rovelli, and the causal-set programme have developed it in different ways across the last hundred years.
What this paper adds is that the relational view is no longer a philosophical preference. Earlier writers postulated it; they took it as a foundational stance and built theories on top. The present result derives it. Anyone who accepts the three kernel principles cannot also accept time as a dimension. The relational view is not one option among many compatible with the principles. It is the only option.
This is a substantial change in the modal status of the position. A philosophical preference can be argued with on the grounds of taste. A theorem cannot. If you want to keep time as a dimension, you have to give up at least one of observer-invariance, finite resolution, or irreversibility — and giving up any of those costs you more than time as a dimension was worth.
Why this matters
Foundational results of this kind don’t immediately give you new technology. Their work is to shrink the assumption-base of physics — to take things that were postulated and show them to be theorems. Every prediction the broader VERSF programme makes downstream, about the cosmological constant, about the fine-structure constant, about quantum measurement, about the arrow of entropy, now rests on a smaller foundation. There is one less thing to assume.
The result also tells us something about the world we already know. The next time you look at a clock, you can remember: that clock isn’t measuring something flowing past underneath. It is counting events. And every event it counts is a small irreversible step in the construction of what we call time. The clock isn’t reading time off a hidden parameter. It is making time, one tick at a time, by producing facts.
That is what the universe does. That is what time is. Not the stage. The play.