Dark Matter May Feel a Force of Its Own — and It Does the Opposite of What Cosmologists Assumed

Everything anyone knows about dark matter has been learned by watching what it does to things that are visible. It bends light. It holds galaxies together at rotation speeds that should fling them apart. It leaves a fingerprint in the cosmic microwave background. What it does to itself has been almost entirely a matter of assumption — and the standard assumption is that it does nothing at all.

A new paper in the Journal of Cosmology and Astroparticle Physics tests what happens if that assumption is wrong, and the result runs against intuition. Zachary J. Weiner of the Perimeter Institute for Theoretical Physics, with collaborators Marco Costa, Cyril Creque-Sarbinowski and Olivier Simon, worked through a model in which dark matter particles attract one another through a force of their own, mediated by a light particle that couples only to the dark sector. The naive expectation is straightforward: an extra attractive force on top of gravity should pull matter together faster and produce more structure, sooner.

The clumping part holds up. The extra force does make dark matter gather more effectively than gravity alone would manage. But the same interaction that produces the attraction also drains the particles' effective mass as the universe expands, so the growing clumps carry progressively less gravitational weight. The two effects fight each other, and across most of the parameter space the analysis explored, the second one wins. The combined result typically suppresses the growth of cosmic structure rather than accelerating it.

The consequence for observations is where this gets sharp. Enhanced clumping does not translate into a stronger gravitational imprint on the cosmic microwave background — the relic light from 380,000 years after the Big Bang that serves as cosmology's most precisely measured dataset. A model that clumps harder while weighing less can hide from the very measurement that would normally catch it. That severs a link cosmologists have leaned on: more clustering has generally been read as more gravitational signal, and here it is not.

"What we really know about dark matter has so far been learned only through its gravitational effects," Weiner said, which is precisely the opening the work exploits. Decades of direct-detection experiments have been built around the idea that dark matter must occasionally bump into ordinary matter. If the dark sector talks mainly to itself, those detectors can keep coming up empty without the hypothesis being wrong.

The paper does not claim a discovery. It maps how a self-interacting dark sector would register in the data cosmologists already have, and it narrows the range of interaction strengths that survive current constraints. The practical payoff is a sharper target for the next generation of surveys, which will measure the clustering of matter across cosmic time precisely enough to distinguish a universe where dark matter is inert from one where it has been quietly pulling on itself the whole time.