Editorial NexoScience
Is the universe actually speeding up, or have astronomers been fooling themselves for almost thirty years?
That question sits at the center of one of the biggest recent controversies in cosmology. Back in 1998, two independent teams studying a special class of exploding stars called Type Ia supernovae found that distant explosions looked fainter than expected. The only way to explain it was if the universe’s expansion was accelerating, pushed by a mysterious force that later got the placeholder name “dark energy.” The discovery reshaped physics and won its leaders the Nobel Prize.
Type Ia supernovae work as cosmic mile markers because they tend to explode with very consistent brightness, which lets astronomers use them to measure distance the same way you’d judge how far away a lightbulb is from how dim it looks. That consistency is the entire foundation of modern dark energy research, so any crack in it matters enormously.
Early in 2025, a paper by Son and colleagues put a very visible crack in that foundation. They argued that the brightness of these supernovae quietly drifts with the age of the star system that produced them, and that older, more distant explosions have been systematically misread. Correct for that drift, they claimed, and the case for acceleration weakens dramatically, an effect they measured at nine standard deviations away from the accepted model, essentially screaming that something fundamental was wrong with three decades of cosmology.
A large international team, including Adam Riess, one of the astronomers who first discovered the acceleration, went back through that claim line by line. Their answer, published this year, is that the supposed crisis mostly evaporates once standard corrections already used across the field are applied properly.
The rebuttal makes three main points. First, the original analysis skipped a routine adjustment for the mass of a supernova’s host galaxy, a correction that already captures much of what looks like an age effect. Once the team added it back in, the apparent link between a supernova’s brightness and its host’s age disappeared. Independent observations of nearby galaxies backed this up too. Supernovae in galaxies several billion years apart in age showed no meaningful difference in brightness once mass was accounted for.
Second, if the original claim were correct, the effect should have grown steadily worse at greater distances, since the universe held systematically younger galaxies in its past. Data from the Dark Energy Survey found essentially no such trend, a shift measured at just -0.028 magnitudes per unit of redshift, statistically indistinguishable from zero and far too small to meaningfully change the measurement of dark energy’s behavior.
Third, and perhaps most tellingly, the original paper’s own central claim, that nearby and distant supernovae come from progenitor stars roughly five billion years apart in age, turned out to be inflated. The new analysis found the real gap was three to five times smaller, largely because the original study had confused the age of an entire host galaxy with the age of the specific star that actually exploded inside it, two related but distinct things.
None of this means the case for dark energy is closed or beyond question. Cosmology is still actively debated, and other independent surveys have found hints of their own that dark energy’s strength might not be perfectly constant over cosmic time. What this particular paper shows is narrower and more mundane: the specific statistical argument for scrapping supernova cosmology altogether does not hold up once it is checked against the corrections the field already relies on. Extraordinary claims still need evidence that survives scrutiny, and this one largely did not.
The study, “Still Accelerating: Type Ia supernova cosmology is robust to host galaxy age evolution,” by Phil Wiseman, Brodie Popovic, Mark Sullivan, Adam G. Riess and colleagues, was published in Monthly Notices of the Royal Astronomical Society, 2026 (DOI: 10.1093/mnras/stag797).
