The universe may be rougher than cosmologists assumed
For about a century, modern cosmology has relied on a simplifying assumption: that on the largest scales, the universe is roughly uniform and looks much the same in every direction. That idea sits inside the standard FLRW model, named for Alexander Friedmann, Georges Lemaître, Howard Robertson and Arthur Geoffrey Walker. Now, according to a New Scientist report on three recent preprints, evidence is building that this assumption may be wrong.
If the challenge holds up, it would not be a small adjustment. Nearly all cosmological observations are interpreted through the FLRW framework. A serious flaw in that foundation would force a rethinking of how distance, expansion and structure are modeled across the cosmos.
A test designed to fail if the standard model is incomplete
The first of the three preprints, by Timothy Clifton and Asta Heinesen, proposes a new test for whether the FLRW model can accurately describe the universe. The idea is elegant in principle. The researchers construct combinations of formulas for cosmic distances using observations of supernovae and fluctuations in matter density. Those combinations should evaluate to zero if FLRW is correct. If they do not, it would suggest the need for a different model.
The significance here lies not just in the proposal of a new test, but in the context around it. Other tests have been proposed before and have not produced a clear signal that something is wrong with FLRW. The promise of the new approach, as described in the report, is that it may be more decisive.
Applying the test without assuming the answer
The second and third preprints, by Heinesen and Sofie Marie Koksbang, apply that test to existing cosmological data. But the researchers first had to solve a methodological problem. Past analyses often extracted relevant distance measurements in ways that already assumed the FLRW framework. To test the model fairly, they needed a path that did not presuppose its correctness.
According to the report, the pair worked out how to obtain the necessary distance measurements from the data without building FLRW into the process. They then used an AI-based approach known as symbolic regression to identify formulas that fit those measurements and could be used in the test.
The result was a clear nonzero outcome. Under the logic of the test, that suggests the FLRW model is flawed.
Why a “lumpy” universe matters
The report frames the emerging picture as a “lumpier” universe than cosmologists had thought. That wording matters because the standard homogeneous and isotropic view was never a claim that matter is perfectly evenly distributed everywhere. It was a claim that once one zooms out far enough, the unevenness smooths away. The new work suggests that may not happen cleanly enough for the standard approximation to hold.
If so, the implications are profound. The FLRW model is not just one theory among many. It is the scaffolding used to interpret a vast range of observations. Weakening it could change how cosmologists infer the behavior of the universe from the data they collect.
Not settled, but consequential
It is important that the work described here consists of preprints, not yet a settled consensus. The report itself notes surprise from Heinesen, who said the result breaks with much of what came before. That alone is a reason for caution. Extraordinary challenges to long-standing models typically require extensive scrutiny and replication.
Still, the reason the report is notable is not that it proves a revolution has happened. It is that it identifies a plausible route by which one could occur. A carefully constructed test, a way of applying it without assuming the standard model in advance, and a nonzero result together create a concrete technical challenge rather than a vague philosophical objection.
A possible path through cosmology’s tensions
New Scientist notes that the emerging evidence could help address some of cosmology’s biggest mysteries. The source text does not enumerate those in detail, but the implication is clear: if a core approximation about large-scale uniformity is wrong, then some persistent mismatches in cosmological interpretation may partly reflect the framework rather than the data alone.
That is why the debate matters. When a field leans heavily on a single simplifying assumption, any credible sign that the assumption fails can ripple through everything built on top of it. For now, the case is still forming. But if these analyses survive scrutiny, cosmology may have to confront the possibility that the universe is not smooth enough for its most familiar model to remain the default description of reality.
This article is based on reporting by New Scientist. Read the original article.
Originally published on newscientist.com
