Why Time Flows Differently Between Galaxies

A December paper by Antonia Seifert and collaborators at the University of Canterbury argues that the universe’s apparent acceleration might be an artifact of how cosmological models treat time. In their “timescape” framework, time runs at different rates in different environments: regions of strong gravity (dense galaxy clusters and filaments) tick more slowly, while vast low-density voids tick faster. Because the universe’s large-scale structure keeps changing—voids occupy a growing fraction of space—photons traveling through the cosmos would accumulate extra redshift over time. That extra redshift could mimic what observers currently attribute to dark energy, potentially removing the need for a cosmological constant in the simplest interpretation.

The claim matters because the dominant cosmological model, Lambda-CDM, is built on a key simplifying assumption: matter is treated as smooth and evenly distributed on large scales. Under that approximation, the Friedmann equations describe expansion driven by smoothly distributed matter plus a constant dark energy term. Observations of Type Ia supernovae—standardizable explosions of white dwarfs—then map out an expansion history. In the late 1990s, those supernova surveys revealed that expansion was accelerating, which pushed cosmologists to add a cosmological constant (Λ) to general relativity. Since then, Lambda-CDM has become the best overall fit to a wide range of measurements, even though cosmic acceleration has not been “directly” measured so much as inferred from how well models match the data.

Seifert’s team leans on a known feature of general relativity—gravitational time dilation—but adds a second, more consequential effect: differential time flow should alter the effective expansion rate from place to place. In their picture, faster-ticking voids expand more than slower-ticking dense regions. Photons crossing those environments would be stretched (redshifted) differently depending on how much time they spend in voids versus denser structures. As the universe becomes more clumpy and voids take up more volume, the void contribution grows, producing a late-time redshift boost that can resemble acceleration without invoking global acceleration.

The recent attention comes from an updated comparison against the Pantheon+ supernova dataset. Using a Bayes factor approach, the analysis finds timescape preferred over Lambda-CDM, especially when nearby supernovae are included—precisely where local inhomogeneities should matter most because photons traverse fewer structures and the “averaging out” assumption is weakest. When only more distant supernovae are used, the two models converge, though timescape still edges ahead in the reported results.

Still, the case is far from settled. Lambda-CDM retains strong support from independent probes such as baryon acoustic oscillations, which provide an expansion measure less sensitive to void-induced photon effects, plus the observed evolution of large-scale structure and the universe’s near-flat geometry. The timescape model also faces a quantitative challenge: it requires very large age differences between voids and dense regions—on the order of billions of years—where some estimates suggest the expected differential time flow should be closer to hundreds or thousands of years. The timescape team also acknowledges that supernova distance calibration is complex and needs further refinement.

Bottom line: dark energy hasn’t been cleanly disproved. Lambda-CDM still offers the most consistent overall framework, but the timescape results are a meaningful stress test—especially for how precisely cosmologists model the “temporal landscape” created by cosmic structure. Any crack in Lambda-CDM’s armor, such as the ongoing Hubble tension and hints of possible time variation in the cosmological constant, keeps alternative models like timescape on the table.