Galaxies That Should Not Exist
When the James Webb Space Telescope began sending back its first deep-field images in 2022, astronomers expected to see faint, small, irregular blobs at the highest redshifts, representing the earliest galaxies forming from primordial gas. What they found instead sent shockwaves through cosmology: massive, well-structured galaxies existing just 300-500 million years after the Big Bang, far too large and too mature to fit standard models of galaxy formation. Three years of follow-up observations have confirmed and deepened the puzzle.
What JWST Is Seeing
The telescope’s NIRCam and NIRSpec instruments have identified dozens of galaxy candidates at redshifts above 10, corresponding to a universe less than 500 million years old. Several have been spectroscopically confirmed, ruling out the possibility that they are lower-redshift interlopers. Quantum Physics Explained: A Complete Guide to the Subatomic World and Quantum Technologies covers earlier JWST discoveries. The most startling examples have stellar masses exceeding 10 billion solar masses, comparable to the Milky Way, at an epoch when theoretical models predict galaxies should be 100 times smaller. Their light profiles show disk-like structures, spiral arms, and even evidence of supermassive black holes, features that were supposed to take billions of years to develop.
Why Standard Models Struggle
The Lambda-CDM model, the standard framework for cosmology, predicts that structure forms hierarchically: small clumps of dark matter merge into larger halos, which gradually accumulate gas that cools and forms stars. This process has built-in speed limits. Gas must cool before it can collapse into stars, and feedback from the first stars and black holes heats surrounding gas, slowing further star formation. JWST’s massive early galaxies seem to violate these speed limits. Either star formation was extraordinarily efficient in the early universe, or something about our understanding of the physics is incomplete.
Proposed Explanations
Theorists have offered several potential solutions. Some argue that the early universe’s gas was denser and cooler than models assume, allowing faster gravitational collapse. Others suggest that the initial mass function, the distribution of star sizes, was different in the early universe, with more massive stars forming per unit of gas, which would make galaxies appear brighter and therefore seem more massive than they actually are. A more radical proposal involves modifying the dark matter model itself: if dark matter can self-interact or dissipate energy, early structure formation would be accelerated. String Theory: The Quest to Unify All Forces of Nature examines the mysteries surrounding dark matter and dark energy.
Canadian Astronomers at the Forefront
Canada contributed the Fine Guidance Sensor and NIRISS instrument aboard JWST, giving Canadian astronomers guaranteed observation time. Teams at the University of Toronto, the National Research Council, and the Dunlap Institute have been among the first to analyze JWST data from the CANUCS (CAnadian NIRISS Unbiased Cluster Survey) program. Canadian researchers have published several high-profile papers on early galaxy evolution, including the discovery of one of the most distant galaxy clusters ever observed. The Canadian Space Agency’s investment in JWST is paying scientific dividends far beyond what was anticipated.
A Crisis or a Course Correction
Is cosmology in crisis? Most working cosmologists say no, at least not yet. The Lambda-CDM model has survived decades of increasingly precise tests, from cosmic microwave background measurements to galaxy clustering surveys. The JWST results are genuinely surprising, but they may be accommodated by adjusting parameters within the existing framework rather than overthrowing it. History offers a precedent: the discovery of the accelerating expansion of the universe in 1998 was deeply surprising but was absorbed into the model by adding a cosmological constant. The early galaxy problem may require a similarly elegant patch. Or it may be the first crack in a framework that needs replacing. Either way, the data is what it is, and that is exactly what great telescopes are for.
