A new preprint released in April 2026 by McLeod, Dunlop, McLure and collaborators (arXiv:2604.16666) reports one of the widest-area searches yet for galaxies in the infant universe. Using JWST NIRCam imaging spanning more than 0.6 square degrees—roughly the area of 3,000 full moons—and more than 150 independent sight-lines, the team measured the ultraviolet luminosity function from redshift 12.5 to 18.5. This corresponds to looking back when the universe was between roughly 200 and 350 million years old.

Methodology: The researchers performed a photometric candidate search across 11 separate JWST fields, deliberately chosen to maximize independent lines of sight and thereby mitigate cosmic variance. They derived the UV luminosity function and converted it to star-formation rate density (ρ_SFR). Sample size is effectively large in area but sparse at the highest redshifts: they report a notable absence of robust candidates above z ≈ 14.5.

The headline result is a rapid decline in ρ_SFR between z=11 and z=13.5, with the drop above z=12 being approximately four times steeper than the trend seen at lower redshifts. At z≈15.5 the measured density sits well below extrapolations from earlier, narrower JWST surveys. The authors show that a piecewise log-linear fit describes the data better than a single power law, and they note consistency with a straightforward theoretical model in which the first substantial galaxy formation begins around z≈15, partially masked at slightly later times by very young stellar populations that are maximally luminous.

This finding goes well beyond the original preprint's technical abstract and reframes the narrative that dominated coverage in 2022–2024. Early JWST deep-field results from the CEERS, JADES, and GLASS programs (e.g., Finkelstein et al. 2023, arXiv:2305.15442; Robertson et al. 2023, Nature) repeatedly highlighted unexpectedly luminous galaxies at z>10. Media narratives often framed these as potential crises for ΛCDM cosmology, suggesting needs for exotic solutions: dramatically higher star-formation efficiency, top-heavy initial mass functions, bursty stochastic histories, or even departures from standard dark-matter physics.

What that coverage largely missed was the role of field-to-field variance. Most early JWST pointings targeted ultra-deep but tiny sky patches or lensing clusters, increasing the chance of sampling overdense regions. The new McLeod et al. study, by surveying 150 sight-lines, averages over that variance and reveals the bright early objects were statistical fluctuations atop a steeper underlying decline. It also synthesizes cleanly with theoretical work such as the FLARES simulations (Vijayan et al. 2021) and analytic models by Mason, Trenti & Treu (2023, arXiv:2302.07279), which showed that modest adjustments to star-formation duty cycles can reproduce bright high-z galaxies without breaking standard cosmology. The Edinburgh-led team demonstrates that even a vanilla model—with no evolving IMF or boosted efficiency—matches the new wider-area data once the progressive youth of stellar populations is accounted for.

Limitations must be stated clearly. This is a preprint, not yet peer-reviewed. At z>14 the candidate sample shrinks to near zero, so Poisson uncertainties and completeness corrections are large. Photometric redshifts, while carefully vetted, carry contamination risks that only spectroscopy can eliminate. Still, the convergence of multiple independent fields strengthens the conclusion.

The broader pattern now emerging is that cosmic dawn was neither as quiescent as pre-JWST models predicted nor as frenetically active as the first wave of JWST headlines implied. Instead, a relatively sharp onset around z≈15, followed by accelerating build-up, appears consistent with both observations and hierarchical structure formation. This tempers earlier hype, reduces tension with theoretical halo mass functions, and refocuses attention on the critical z=12–15 window where the first generations of stars and galaxies truly transformed the universe. Future wider JWST programs and ELT spectroscopy will test whether this piecewise evolution holds or if additional physics still lurks in the data.