Cold atoms / relational time

What if change is the clock?

A 2026 experiment used a trapped cloud of ultracold atoms to test whether events can be ordered from entropy exchange inside the system, without using an external clock as the ordering variable.

No outside clock. Still a before and after.

The paper asks a relational question: can one part of an isolated system provide an ordering variable for another? Here, entropy exchange between an observed bright sector and an unobserved dark sector creates a monotonic internal parameter called entropic time.

The isolated analogue

A Bose-Einstein condensate sits in a conservative trap. A thin optical barrier partitions it into sectors while atoms can still exchange across the boundary.

BRIGHT
observed
OPTICAL BARRIER
DARK
unobserved

What provides the arrow?

Coarse-grained entropy can fluctuate locally, but the constructed entropic time robustly orders measured events across repeated expansion and recollapse cycles.

What changes the speed?

Adjusting the barrier changes entropy production. In the relational description, the same laboratory interval can correspond to faster, slower, or nearly stationary internal time.

Order events by the system itself.

Scroll through one conceptual cycle. The laboratory still records ordinary time, but the internal description labels events using accumulated entropy flow.

Localized seed

The observed bright sector begins narrow. A lab clock could label the frame, but the internal construction has barely accumulated entropy flow.

tau = integral |entropy flow| dt

Expansion

The bright density spreads and atoms exchange with the dark sector. Entropic time advances as the internally measured entropy distribution changes.

d tau / dt > 0

Turning point

The cloud reaches maximum spread. Expansion speed can vanish while the relational ordering remains defined by the history of exchange.

d width / dt = 0; tau keeps its order

Recollapse

Cloud width decreases, but events do not run backward. The internal parameter still places recollapse after expansion.

width decreases; tau increases

What the result does and does not say.

It is a controlled analogue experiment.

The work establishes a cold-atom platform where a relational construction of time can be quantitatively tested. It does not create a literal universe or prove that all physical time is emergent.

Not a Page-Wootters demonstration

The 2026 experiment constructs time operationally from entropy exchange. Page-Wootters uses correlations with a clock subsystem; it is a related relational-time idea, not the mechanism tested here.

Measured evolution survives

An effective Schrodinger equation parameterized by internal entropic time reproduced the measured evolution of the observed sector.

Expansion is an analogy

The bright sector repeatedly expands and recollapses in a structure reminiscent of minisuperspace cosmology. It is an atomic analogue, not a miniature spacetime.

The external clock still exists

Laboratory time is used to run and record the experiment. The key test is whether the data can also be ordered and modeled using only the internal variable.

Can change replace a clock?

Three checks distinguish the actual result from the headline.

What served as the internal ordering variable?

What happened during recollapse?

Which is the careful conclusion?

Enjoy this tool? Build your own with Super