Dark matter gave galaxies their spin 13 billion years ago — now confirmed at 7-sigma

Galaxy spins were set by dark matter 13 billion years ago — measured at 7-sigma

Every galaxy you can see in the night sky is rotating. The Milky Way turns. The Andromeda galaxy turns. The thousands of galaxies visible through any amateur telescope are all spinning, each on its own axis, each at its own rate. For decades, astrophysicists believed that rotation was seeded by gravity in the earliest moments of the universe — dark matter halos tugging on neighboring halos, transferring angular momentum before galaxies had even formed their first star. That theory has now been confirmed, at a statistical significance of 7-sigma, using the sharpest reconstruction of the primordial universe ever applied to real galaxy data.

The implication is direct and difficult to overstate: galaxy rotation is not a random accident of cosmic assembly. It is a fossil record, readable today, of the gravitational architecture of the universe as it was 13.6 billion years ago.

Recommended ReadingJWST found a galaxy cluster 10 billion years old that shouldn’t be this dense

Tidal-torque theory — often abbreviated TTT — was proposed in its modern form in the early 1970s. Its central claim: the angular momentum of proto-galaxy halos, the dark matter clouds that would eventually collapse into galaxies, was set by the gravitational tidal forces of their surroundings. If two neighboring proto-halos were elongated and slightly misaligned with the surrounding matter distribution, they exerted a torque on each other. That rotational push was imprinted in the halos and, through the collapse and accretion that followed, passed on to the visible galaxies inside them. The theory predicted that today’s galaxy spins should still carry a detectable statistical alignment with the primordial tidal field — the pattern of matter distribution in the universe before it formed any structure at all.

How the team built the proof

Ming-Jie Sheng and colleagues at Xiamen University, collaborating with researchers across China and Canada, used two complementary datasets. The first was the Sloan Digital Sky Survey, which provides precise measurements of the positions, velocities, and morphologies of hundreds of thousands of nearby galaxies. From those measurements, the team extracted the angular momentum vectors of the gas components in central massive elliptical galaxies — the rotation axes of the gas that orbits inside the most massive galaxy systems in the local universe.

The second ingredient was ELUCID — the Exploring the Local Universe with reConstructed Initial Density field simulation. ELUCID works backwards from the observed distribution of galaxies in the SDSS to reconstruct what the primordial density field must have looked like for today’s large-scale structure to have assembled the way it did. It is, in effect, a constrained reconstruction of the universe’s initial conditions in our cosmic neighborhood, run forward to verify its fidelity against what we observe.

The comparison was straightforward in concept, technically demanding in execution. The team asked: does the direction in which gas rotates inside massive elliptical galaxies align with the direction TTT predicts, given the ELUCID reconstruction of where dark matter was tugging 13.6 billion years ago? The answer, in the gas of central massive elliptical galaxies, was yes — at a directional correlation with a significance of approximately 7-sigma.

What 7-sigma means

In most sciences, a result reaching 5-sigma — a probability of roughly one in 3.5 million that it arises by chance — is treated as a discovery. Particle physics adopted 5-sigma as its standard precisely because the field was bruised by false positives at lower thresholds. A 7-sigma result pushes the probability of a chance alignment into territory that is, for practical purposes, ruled out. It does not mean the theory is complete or that every detail of TTT is correct. It means the directional signal between the primordial dark matter tidal field and present-day galaxy gas rotation is not a statistical artifact.

To appreciate how much this matters, consider that previous attempts to detect the TTT imprint had found only hints — 2- or 3-sigma correlations that sat uncomfortably between suggestive and inconclusive. The central difficulty is that galaxy formation is a messy process. Gas cools, falls in, gets disrupted by mergers, blown back out by supernova feedback, and reaccreted. Over 13 billion years, enough chaos accumulates that most of the primordial signal gets washed out. The fact that a 7-sigma imprint survives at all tells us something specific: the angular momentum printed by the primordial tidal field is robust enough to persist through the full violence of galaxy assembly, at least in the gas of the most massive elliptical systems.

What the confirmation leaves open

The signal is clear in the gas of central massive elliptical galaxies. It is not yet confirmed in their stellar components, which have different dynamical histories and are harder to disentangle from merger debris. Spiral and disk galaxies present their own challenges — their rotation is tied up in disks that formed through different accretion paths, and the ELUCID reconstruction, while sophisticated, carries model dependencies that become harder to control as one moves to lower galaxy masses or higher redshifts.

There is also the underlying limitation of any reconstruction-based method: ELUCID infers the primordial density field from where galaxies are today. That inference depends on assumptions about how structure grows — assumptions rooted in the standard cosmological model. If those assumptions are even slightly wrong, the inferred primordial field shifts, and so does the predicted angular momentum alignment. Sheng et al. tested their results against these systematics and found them stable, but the conclusion rests on the validity of the reconstruction chain. Independent measurements — from future surveys with greater sky coverage and spectroscopic precision — will be needed to eliminate that dependency.

Astronomer Pablo López of the National University of Córdoba, who commented on the result for Science magazine, noted that the researchers “have convincingly detected a surviving imprint” of primordial spin, and added that knowing the link between today’s spinning galaxies and that early gravitational tussle should help researchers zero in on what was happening during that remote epoch. What it cannot yet tell us, he noted, is the exact mechanism by which that angular momentum was preserved and amplified through billions of years of star formation and merging.

What it changes for dark matter research

Dark matter is the dominant mass component of every galaxy halo, and it is invisible by every conventional detection method. It does not emit, absorb, or reflect light. What it does, we know only through gravity. The TTT confirmation adds a new constraint to that gravitational record: the rotational orientation of massive galaxies today is an imprint of dark matter’s distribution in the primordial universe. That is a testable prediction that alternative gravity theories — those that try to explain galaxy rotation without invoking dark matter — must now also account for. Whether those frameworks can produce the same 7-sigma directional alignment from primordial tidal fields, without actual dark matter, is an open question that this result sharpens.

Common questions about galaxy spin

How do galaxies get their spin in the first place?

Tidal-torque theory holds that proto-galaxy halos — the dark matter clouds that collapse into galaxies — acquire angular momentum from the gravitational tugging of neighboring halos in the early universe, before galaxies themselves have formed. That torque is stored in the dark matter halo and transferred to the gas and stars that later accumulate inside it. The Xiamen University result now confirms this at 7-sigma, at least for the gas in central massive elliptical galaxies.

Why does the signal survive in elliptical galaxies but not clearly in spirals?

Elliptical galaxies are thought to have assembled primarily through major mergers, a process that randomizes stellar orbits but may preserve some of the gas’s net angular momentum. Spiral galaxies, by contrast, built their disks through extended accretion of gas that had its own complex infall history. The primordial imprint in spirals may exist but is overlaid by later processes that are harder to disentangle with current data.

What is the ELUCID simulation and why does it matter here?

ELUCID is a constrained cosmological simulation that reconstructs the primordial density field — the initial distribution of matter — in the local universe by working backwards from where galaxies are today. It allows researchers to predict what the dark matter tidal field looked like 13 billion years ago at the locations of observed galaxies, making it possible to directly compare those predictions to measured galaxy rotation directions.

What does 7-sigma significance mean in practice?

Sigma, or standard deviations, measures how far a result sits from what would be expected if there were no real effect. A 7-sigma result means the observed alignment between galaxy gas rotation and primordial predictions would arise by chance in roughly one in 390 billion trials. In astrophysics, 5-sigma is the conventional threshold for declaring a detection. 7-sigma rules out random chance with extraordinary confidence.

The next direct test of the result is expected from the Dark Energy Spectroscopic Instrument, the Euclid space telescope, and eventually the Nancy Grace Roman Space Telescope — all of which will map galaxy kinematics and large-scale structure over volumes far larger than SDSS, enabling the same analysis at higher redshift and across a wider range of galaxy types.

Reference: Sheng et al., “Observational evidence of tidal torque in the primordial Universe,” Nature Astronomy, 2026. arXiv: 2512.11383

Tags: astronomy, Dark Matter, ELUCID, galaxy angular momentum, Ming-Jie Sheng, SDSS

More Like This

Webb caught a galaxy several times the Milky Way’s mass — and it isn’t spinning

James Webb mapped 164,000 galaxies to chart the cosmic web back a billion years

DESI wasn’t looking to break the standard model. Its galaxy maps did it anyway

Physicists caught a crystal’s atoms reversing their spin for the first time

The Milky Way’s black hole has been blowing a wind for 20,000 years

Webb measured a 13-billion-year-old black hole — it formed before its galaxy