Earlier this year, the Dark Energy Spectroscopic Instrument (DESI) completed observations that mapped 47 million galaxies across 11 billion light-years, allowing astronomers to better evaluate the large-scale structure of the visible universe. After studying these data, astronomers Francesco Sylos Labini and Marco Galoppo say the universe may not look the same in all directions. Their results, published in Nature, contradict a fundamental assumption in modern cosmology.
Long-held assumptions in cosmology
At the scale of a single galaxy or local groups of galaxies, the universe clearly appears to be anisotropic, meaning the structure is different depending on which direction you look. In one direction, there may be more void space, while another direction may have a cluster of galaxies.
However, the cosmological principle says that at larger scales, the universe consists of matter that is more or less distributed evenly in all directions. This is based on the Copernican principle, which states that there should be no “special observers” in the universe, meaning that at large scales, the universe should look the same from anywhere else in the universe.
For example, if you imagine the universe as a piece of cloth and zoom in to the scale of the individual fibers, you can clearly see areas of empty space and filament-like fibers that connect to make a larger structure. Yet when you zoom out to much larger scales, the cloth appears to be the same everywhere, with evenly distributed materials.
There has been debate over exactly what scale the universe should appear isotropic. Galactic surveys have shown a “cosmic web” of filaments, walls and voids, and scientists are unsure how quickly that structure should fade with scale. Research focusing on cosmic background radiation has provided some support for the cosmological principle, but other studies have shown that anisotropic structure still exists at scales of tens to hundreds of megaparsecs. However, the statistical significance of these studies is uncertain.
Unexpected structure at large scales
The authors of the new study say that past anisotropy probes tested for preferred directions instead of evaluating more general directional structure. To test more generally and measure how the distribution of matter varies with both distance and angle, they used the Angular Distribution of Pairwise Distances (ADPD). ADPD is a parameter-free statistic that measures directional correlations. They compared their results with a model based on expected isotropy.
They found that galaxy samples from DESI show persistent anisotropic structure in galaxy distribution out to roughly gigaparsec scales, meaning galaxies were clumping together more than they should at scales far larger than those previously examined. Taking previous studies suggesting anisotropy at megaparsec scales as an example, this study indicates anisotropy still exists at scales 1,000 times larger.
“These results provide direct evidence that directional coherence persists to larger scales than predicted in the standard framework, challenging the assumption of large-scale isotropy,” the study authors write.
Rethinking the cosmological principle
There are some limitations to the new study, and it does not by itself pinpoint the physical origin of the anisotropy. Even if it is correct, the universe may eventually become isotropic at even larger scales. Still, because the cosmological principle underpins many ideas in cosmology, the authors say some ideas may need revision if larger-scale anisotropy is confirmed.
They write, “As such, this detection of large-scale anisotropies contrasts the standard formulation of the cosmological principle, which assumes statistical homogeneity and isotropy about any point, while remaining compatible with the Copernican principle, which requires only the absence of privileged observing locations.
“Then, from a theoretical perspective, the existence of such large-scale anisotropies motivates the exploration of more general solutions to Einstein’s field equations that explicitly allow for large-scale inhomogeneities as cosmological models and/or the investigation of alternative sources of accelerated structure formation, for example, by the introduction of self-interaction in the dark matter component or backreaction effects from inhomogeneities.”
Written for you by our author Krystal Kasal, edited by Gaby Clark, and fact-checked and reviewed by Robert Egan—this article is the result of careful human work. We rely on readers like you to keep independent science journalism alive.
If this reporting matters to you, please consider a donation (especially monthly). You’ll get an ad-free account as a thank-you.
Publication details
Francesco Sylos Labini et al, Detection of anisotropic cosmic structures on a gigaparsec scale, Nature (2026). DOI: 10.1038/s41586-026-10702-5
The content is provided for information purposes only.