On Temporal Causation and the Nature of Scientific Repeatability

A Thought Experiment

This is an idea that developed over late-night drives home from work. It's not a challenge to established science. More - an exploration of whether we might be slightly misunderstanding what scientific repeatability actually demonstrates.

The Question That Started This

I was thinking about what happens when the same situation occurs at two different times.

If a system could behave one way at time T1 and then the exact same system, in the exact same state, behaves differently at time T2, what changed? The only obvious difference is time itself. But if time is "just a parameter," the way we seem to treat it in physics then it shouldn't matter. Same initial conditions should give the same outcome regardless of when you run the experiment.

Yet we know from experience that quantum measurements give different results on repeated trials. Systems near black holes behave differently depending on their temporal relationship to the event horizon. Entropy increases over time, meaning no state ever truly repeats.

So maybe time isn't just a parameter. Maybe time is causally relevant in ways not actively or fully accounted for.

What Science Actually Does

We often say science is about repeatability. Run the same experiment, get the same result, and you've discovered a law of nature. But I think what science actually does is more subtle than that. Science minimizes the range of variables that can influence an outcome, thereby shortening the window in which temporal and environmental variations matter.

Think about what it means to design an experiment. You don't control all variables. That's nigh on impossible because of the almost infinite factors in play. What you do is narrow the range of the important ones. Temperature held to a tenth of a degree instead of ten degrees. Pressure held to a hundredth of an atmosphere instead of a full atmosphere. Time between trials measured in minutes instead of years. You're not proving things are identical. You're proving they're similar enough that differences don't matter within your measurement precision.

You're creating conditions where temporal drift, environmental fluctuation, and entropic change are negligible compared to the effect you're measuring. And within those conditions, science works beautifully. But that's a statement about local stability, not universal repeatability.

The Narrowing Window

When we say "this experiment is repeatable," we're really saying that within this narrow range of conditions, within this short temporal window, with this measurement precision, the outcomes are insensitive to small variations. We're shortening the window in which things can vary.

The tighter our controls, the smaller the window, the more "repeatable" things appear. Not because they're genuinely identical, but because differences are smaller than we can detect. Nature is locally stable. Small changes in conditions produce small changes in outcomes in most regimes and that local stability is what the scientific method captures.

But local stability is not the same as proving outcomes are independent of time.

Time, Entropy, and True Repetition

There is something that seems obvious once you notice it. You can't have entropy without time. Entropy measures disorder increasing, and increasing requires temporal direction. No time, no change, no entropy.

And you can't have repeatability with entropy. The Second Law of Thermodynamics says entropy always increases. Even if you carefully reset all observable variables to run an experiment again, entropy has still gone up. Microscopic states have changed. Heat has dissipated. The universe is different than it was. Every moment is unique because entropy has increased since the last moment.

So when we "repeat" an experiment, we're not actually recreating the same state. We're creating a similar state in a universe that now has slightly higher entropy. True repetition is thermodynamically implausible. What we achieve instead is approximation precise enough that the difference doesn't register in our measurements.

Where Time Clearly Matters

I don't know who decided time doesn't causally matter, but it's obviously central to understanding black holes. Near an event horizon, time dilation is extreme. What looks like seconds to a distant observer is years to someone nearby. Physical outcomes depend on when you measure, relative to the gravitational field.

If time were "just a coordinate" that didn't causally matter, this would make no sense. Time dilation changes outcomes because time is different near the event horizon. This suggests time isn't just a background parameter we measure things against. It's an active component of what determines how systems evolve.

Standard quantum mechanics runs into something similar, but here I need to be more careful about what I'm actually claiming. Physics does not ignore time in quantum mechanics. The Schrödinger equation evolves quantum states through time. Time is in the formalism. What's strange is not that time is absent but that the measurement problem produces outcomes that don't seem to follow deterministically from the time-evolved state. The wave function evolves smoothly and predictably right up until you measure it, and then you get a specific result that the evolution didn't uniquely determine. Run the double-slit experiment multiple times and you get different results. Photon goes through the left slit, then the right slit, then the left again. The standard explanation is fundamental randomness. But what if the outcomes vary partly because the temporal context matters in ways the formalism doesn't capture? Not that physics forgot to include time. More that the role time plays at the moment of measurement might be deeper than "which value t has in the equation."

I should be honest about what this runs into. Bell's theorem puts hard constraints on exactly this kind of thinking. Bell showed that if you try to explain quantum measurement outcomes through hidden variables, local deterministic ones that just aren't visible to us, the resulting predictions disagree with experiment. The correlations in entangled particle measurements are stronger than any local hidden variable theory can produce. That's not a theoretical objection. It's been tested, repeatedly, and the quantum prediction wins every time.

So if I'm suggesting that temporal dependence is a hidden variable that explains measurement outcomes, Bell says no. Or at least, Bell says it can't be a local hidden variable. There are nonlocal approaches that survive Bell (Bohmian mechanics being the most developed), but they come with their own costs, and I don't have the technical depth to evaluate whether temporal causation could fit into that kind of framework without violating the constraints Bell established.

What I'm left with is something more modest than where I started. I'm not claiming time is a hidden variable that resolves quantum randomness. I'm asking whether temporal structure deserves more attention in measurement theory, whether the moment of measurement carries more causal weight than "just a coordinate" suggests. That might turn out to be the wrong question. But it seems worth asking given that the measurement problem remains genuinely unsolved.

What This Suggests

If any of this holds, it means science doesn't discover timeless laws. It discovers locally stable approximations that work within narrow temporal and parameter windows. This is not a criticism. It's a clarification. Science is more modest than we sometimes claim.

We're not proving universal truths that hold for all time. We're mapping regions of parameter space where temporal effects are negligible. Where entropic drift is small and conditions are stable enough that small variations don't detectably change outcomes. Within these windows, our laws work. Outside these windows, at quantum scales, relativistic speeds, cosmological timescales, temporal effects might become significant, and our "laws" are revealed as approximations valid in limited regimes.

If time is causally relevant at fundamental scales, a few things follow. Physical "constants" could vary over cosmic time, not because they're changing arbitrarily, but because time itself matters. Quantum "randomness" might show temporal patterns, since if outcomes are time-dependent rather than truly random, experiments at regular intervals might reveal subtle correlations that true randomness would not produce. Poincaré recurrence, the idea that finite systems should eventually return to their initial states, might be impossible in practice, because even if configuration space is finite, time has changed, and entropy prevents exact repetition.

The Scope of the Claim

There is a difference between treating time as a parameter, where t appears in the function, and treating time as causally active, where outcomes depend intrinsically on when they occur. Current physics mostly does the first. I'm suggesting we might need to take the second more seriously.

This doesn't diminish science. It clarifies what science achieves and suggests where to look for deeper understanding. Scientific repeatability is more subtle than commonly described. Time might be causally relevant at scales where we usually ignore it. "Narrowing the window" is a better description of what experimental method accomplishes than "proving universality." And some of the strangeness we encounter at the boundaries of physics might be temporal effects in disguise.

Maybe the reason we struggle with quantum measurement, time's arrow, and the unification of relativity with quantum mechanics is that we're treating time as "just a parameter" when it's actually causally fundamental. If so, the path forward might involve taking temporal dependence seriously in quantum foundations, looking for temporal patterns in outcomes we currently call random, examining whether physical constants vary over cosmological time, and reframing the scientific method as demonstrating local stability rather than proving universal repeatability.

Or maybe I'm completely wrong. Even so, I wanted to write it down in case it's useful to someone.