Modern cosmology rests on a simple assumption: if we look on large enough scales, matter should be distributed evenly, with no preferred direction within the cosmos. This is known as the cosmological principle.
Now, as new telescopes both on Earth and in space, such as the Dark Energy Spectroscopic Instrument (DESI) and Euclid, deliver ever more detailed maps of the universe, this assumption can finally be properly tested.
In our new paper, we uncover evidence that the distribution of galaxies does not become uniform on the largest scales we can currently test. Using DESI data, we find directional patterns extending across distances of several billion light years.
If confirmed, our results would force physicists to rethink some basic ideas about the universe, including what dark matter is, and how gravity shapes matter on the largest scales.
A model that worked remarkably well
The cosmological principle underpins the standard cosmological model, which provides a recipe for the universe: roughly 5% ordinary matter, 25% dark matter and 70% dark energy (represented by the Greek letter Λ). This is known as the standard Lambda Cold Dark Matter (ΛCDM) model.
The model has been remarkably successful. For example, it describes the expansion history of the universe, the formation of light elements after the Big Bang, and the cosmic microwave background – ancient light released when the universe first became transparent – with impressive precision.
However, this success has also made the growing observational tensions harder to ignore.
The rate of cosmic expansion is known as the Hubble constant, but precise estimates of the present expansion rate of the universe do not all agree. This has led to a much debated challenge of the ΛCDM model – the Hubble tension.
Recent observations of ancient galaxies by the James Webb telescope also put into question our understanding of early cosmic structure formation.
However, many recognise the most perplexing puzzle is an anomalously large dipole – a “one direction versus the opposite direction” asymmetry in the sky – in the distribution of very distant quasars and radio galaxies. This is in stark contrast with the ΛCDM model.
Finally, last year data from DESI have challenged the very nature of dark energy, which may not be a constant as assumed. This shakes the foundation of modern cosmology.
Investigating large-scale cosmic structures
DESI is building one of the most detailed three-dimensional maps of the universe yet made, measuring galaxy positions in the sky and their redshifts, which tell us how far away they are.
Our work asks whether the matter distribution really is becoming smooth and directionless on the largest scales we can observe. In other words, is the cosmological principle supported by our best data?
To test this, we used a technique which measures the probability of finding a galaxy at a given distance and along a specific direction from another galaxy.
We computed this for all galaxy pairs and averaged the result.
If galaxies are distributed uniformly, those pair directions should be evenly spread. If galaxies sit in long filaments or walls, more pairs will line up along particular directions.
A persistent cosmic web
Applying this to DESI galaxies, we found a clear directional signal. Galaxy pairs were not randomly oriented but rather aligned, tracing coherent filaments and walls.
This would not be surprising if the signal weakened at larger scales. Instead, the patterns persisted over enormous distances, extending to several billion light years in the deepest samples.
The cosmic web did not appear to fade into a uniform, directionless distribution on the largest scales we could test.
Even on the largest scales, the universe seems closer to a tangled yarn rather than a misty fog.
We then compared the observations with simulated universes based on the standard ΛCDM model. The difference was striking. The simulated universes showed weaker and smaller directional patterns in the matter distribution. The real DESI data showed stronger structures, persisting across much larger distances.
What this means
Our results suggest that, within the standard model, there has not been enough time for structures this large to form.
If galaxies follow the overall distribution of mass, including dark matter, the pattern in galaxy locations calls into question our assumption that the universe is roughly uniform at large enough scales.
One possible explanation is that dark matter can interact in complicated, unexpected ways, beyond those included in the simplest models.
Another is that we need a more complex general description of the Universe, one that allows large-scale inhomogeneities to play a greater role.
Or the answer may be something else altogether.
Our results reveal coherent structures spanning billions of light years, much larger than expected in the standard cosmological model. If confirmed, they would directly violate the cosmological principle.
This would suggest that matter remains organised into large-scale patterns over much greater distances than currently thought.
The next step is not speculation, but measurement.
Future data from DESI, Euclid and other surveys will be crucial. If the evidence persists, cosmologists may need new models of structure formation and a revised picture of the Universe on the largest scales.
