Less than 2 billion years after the Big Bang, a galaxy roughly three times the mass of the Milky Way had already stopped forming stars, settled into a bloated shape, and lost something astronomers did not think could vanish that quickly: its spin. The galaxy, cataloged as XMM-VID1-2075, is the first ever confirmed to have stars drifting in random, directionless orbits at such an early point in cosmic history. Until now, that trait had only been measured in the oldest, most massive elliptical galaxies nearby, systems that have had more than 10 billion years to reach that state.
The discovery, published in Nature Astronomy in May 2026 by a team led by Ben Forrest of UC Davis, challenges a basic assumption in galaxy evolution: that it takes billions of years of collisions and mergers to strip a galaxy of organized rotation. “This is the first direct kinematic confirmation of a dispersion-dominated massive galaxy at this redshift,” Forrest said, describing the result as evidence that the universe could build dynamically mature systems far earlier than models predict.
What “no spin” actually means
Most galaxies that are still forming stars, including our own Milky Way, are disk-shaped and spinning. Their stars, gas, and dust orbit the center in roughly the same direction, like water circling a drain. That coherent rotation is what gives spiral galaxies their flat, pinwheel structure.
Elliptical galaxies are different. Their stars swarm in all directions, more like bees around a hive than cars on a racetrack. Astronomers call this “pressure-supported” motion, because the random velocities of the stars, rather than orderly spin, are what keeps the galaxy from collapsing under its own gravity. Galaxies dominated by this kind of motion are classified as “slow rotators.”
In the nearby universe, slow rotators are almost always massive ellipticals that stopped forming stars long ago. The prevailing explanation is that they got that way through repeated dry mergers, collisions between gas-poor galaxies that gradually scrambled any organized rotation over billions of years. Finding a slow rotator when the universe was barely 1.8 billion years old upends that timeline.
How the team measured it
Forrest’s team used the James Webb Space Telescope’s NIRSpec integral-field unit, an instrument mode that captures a separate spectrum for each tiny patch across a target rather than blending the whole galaxy into a single reading. The result is a three-dimensional data cube: two spatial dimensions and one spectral dimension. From that cube, the researchers extracted a map of how fast stars are moving and in what directions across the face of XMM-VID1-2075.
The kinematic maps showed that the galaxy is dominated by velocity dispersion, the spread of random stellar speeds, rather than by any coherent rotational pattern. The team quantified this using a standard dimensionless spin parameter. XMM-VID1-2075 scored roughly 0.12, placing it squarely among the slow rotators cataloged in large kinematic surveys of nearby elliptical galaxies. The difference is that those nearby systems are observed at the present epoch, after 13 billion-plus years of evolution. XMM-VID1-2075 was already there when the universe was a fraction of its current age.
The galaxy was originally flagged in deep near-infrared imaging from the VISTA Deep Extragalactic Observations (VIDEO) survey, which covers fields including XMM-LSS. Its unusually red colors and high estimated redshift marked it as a strong candidate for a massive, quiescent system at early times. JWST follow-up observations under program GO-02913 confirmed the redshift at 3.449 and delivered the spatially resolved kinematics that made the slow-rotator classification possible.
What the data cannot yet explain
The observation that XMM-VID1-2075 lacks organized spin is grounded directly in the NIRSpec measurements. But the reason it looks this way so early is an open question, and none of the current answers are fully satisfying.
One possibility is that the galaxy formed through a rapid, monolithic collapse in an exceptionally dense region of the early universe, producing a dispersion-dominated system almost from birth. Another is that a small number of major mergers happened very early and efficiently destroyed whatever disk-like rotation existed initially. In the Nature Astronomy paper, Forrest and colleagues write that current cosmological simulations struggle to produce a slow rotator this massive at such high redshift without fine-tuned conditions. That assessment reflects the authors’ own comparison of their observations against existing simulation outputs rather than a separately established community consensus, and it signals that models of galaxy assembly may be missing a key ingredient. The same paper is also available as an arXiv preprint.
There are also measurement uncertainties to consider. Building data cubes from NIRSpec’s integral-field unit involves complex processing steps, including resampling and correcting for slice-to-slice variations, that can introduce systematic effects. The authors describe mitigation steps, but independent teams have not yet re-reduced the raw data. The qualitative conclusion that the galaxy is dispersion-dominated appears robust, though the precise spin value could shift slightly with different reduction approaches.
The classification of XMM-VID1-2075 as quiescent, meaning it had already stopped forming stars, rests partly on spectral energy distribution modeling. The galaxy’s red continuum and absence of strong emission lines strongly support that interpretation, but the exact timing of when star formation shut down depends on which stellar population templates and fitting codes are applied. Those details carry more uncertainty than the kinematic result itself.
Why one galaxy could reshape the models
A single object does not overturn a field, but it can expose where the models break. In standard galaxy evolution theory, slow rotators are an endpoint: the product of a long merger history that gradually erases spin. XMM-VID1-2075 suggests that endpoint can be reached far faster than expected, or that some galaxies never had much spin to lose in the first place.
The next step is finding out whether this galaxy is a freak or a forerunner. The selection criteria that pulled XMM-VID1-2075 from the VIDEO catalog have not been described in exhaustive detail, leaving open the question of how many similar candidates might be sitting in the same survey fields, waiting for JWST spectroscopy. If more early slow rotators turn up, theorists will need to rethink how quickly massive galaxies can assemble and settle into their final dynamical state. If XMM-VID1-2075 remains alone, it becomes a puzzle that models must at least be able to accommodate as a rare but real outcome.
What comes after a galaxy with no spin at redshift 3.4
Either way, the discovery marks the first time astronomers have caught a galaxy looking this dynamically old, this early. The universe, it turns out, could build something resembling a retired giant well before most of its galaxies had even finished forming.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
