The Galaxies That Shouldn't Be There: How Webb's Discoveries Are Testing the Foundations of Cosmology

In 2023, a team led by astronomer Ivo Labbé published findings from the James Webb Space Telescope that landed like a grenade in cosmology departments worldwide. Six candidate galaxies, detected at redshifts between 7 and 10 — meaning they existed just 500 to 800 million years after the Big Bang — appeared to contain stellar masses comparable to the modern Milky Way [1]. The implied baryon-to-stellar-mass conversion efficiency exceeded 80%, a figure so high that prior theoretical work had considered even 57% implausibly extreme [2]. The galaxies weren't just unexpected. According to the reigning cosmological framework, they shouldn't have been there at all.

Three years later, the picture has grown more complicated rather than simpler. Some of those initial mass estimates have been revised downward. Others have been spectroscopically confirmed. And the broader pattern — an overabundance of luminous, structured galaxies at the earliest observable epochs — continues to accumulate in survey after survey, pushing researchers toward an uncomfortable question: does the standard model of cosmology need patching, or replacing?

What ΛCDM Predicts — and What Webb Found

Lambda-Cold Dark Matter, or ΛCDM, has been the consensus cosmological model since the late 1990s. It describes a universe composed of roughly 5% ordinary matter, 27% cold dark matter, and 68% dark energy (represented by the cosmological constant Λ). Under this framework, the first galaxies form gradually: baryonic gas cools and collapses into dark matter halos, producing small, irregular, low-mass progenitor galaxies in the first 500 million years after the Big Bang. Massive, structured systems assemble later through mergers and steady accretion over billions of years [3].

Webb's observations have challenged this timeline at multiple redshift ranges. At redshifts above 10 — corresponding to less than 500 million years after the Big Bang — JWST surveys detected a stellar mass density of approximately 10^6 solar masses per cubic megaparsec among galaxies exceeding 10^10.5 solar masses [4]. One candidate galaxy at redshift 7.4–9.1 showed a possible stellar mass of roughly 10^11 solar masses — 100 billion times the mass of our sun, rivaling the present-day Milky Way at a time when the universe was less than a billion years old [4].

The JADES (JWST Advanced Deep Extragalactic Survey) program has been particularly productive in mapping this territory. Its Data Release 5 catalogued photometrically selected galaxy candidates at redshifts above 8, while spectroscopic follow-up confirmed a high abundance of galaxies beyond redshift 9, with very few cases of interlopers or redshift misidentifications [5]. The survey confirmed JADES-GS-z14-0 at a spectroscopic redshift of 14.32 — the most distant galaxy currently known — corresponding to a universe less than 300 million years old [6]. This galaxy is 1,600 light-years wide, luminous, and shows spectroscopic evidence of ionized oxygen, implying at least one prior generation of stars had already formed, lived, and died [6].

Source: JADES, CEERS, COSMOS-Web surveys (compiled)

Data as of Dec 1, 2025CSV

Beyond individual record-breakers, the aggregate numbers tell the story. Webb data consistently show roughly twice as many galaxies at high redshifts as the standard model predicts [3]. A 2025 Cambridge-led survey of more than 250 galaxies observed between 800 million and 1.5 billion years after the Big Bang found that most were turbulent, gas-rich, and still merging — but their sheer number strained the model's predictions [3]. A sample of around 70 dusty galaxies identified less than a billion years after the Big Bang posed an additional puzzle: dust is manufactured by stars that have already completed their life cycles, implying star formation began even earlier than their observed epoch [3].

The Defenders: Why ΛCDM May Still Hold

The initial panic has given way to a more measured assessment among many mainstream cosmologists. Steven Finkelstein, a University of Texas astronomer who leads the CEERS (Cosmic Evolution Early Release Science) survey, has been among the most prominent voices arguing against a crisis narrative. "The bottom line is there is no crisis in terms of the standard model of cosmology," Finkelstein stated, while acknowledging the unexplained factor-of-two overabundance [7].

The strongest defense has come from reanalysis of the most extreme mass estimates. Several of Labbé's original "impossible" galaxies turned out to harbor active galactic nuclei — central supermassive black holes that were rapidly consuming gas. The friction of infalling material generates enormous heat and light, mimicking the signatures of a much larger stellar population and inflating mass estimates by significant factors. When these AGN contributions were removed from the analysis, the remaining galaxies fell within the range that ΛCDM can accommodate [7].

Other systematic effects compound the uncertainty. Strong emission lines from ionized gas, dust reddening, and photometric redshift overestimation have all been identified as potential culprits. A study of 26 galaxies found that 21 showed photometric redshifts systematically higher than their spectroscopic measurements — a statistically significant bias that could make galaxies appear more distant, and therefore more anomalously massive, than they actually are [8]. The most famous example is CEERS-93316, initially estimated at a photometric redshift of approximately 16, which spectroscopy later placed at redshift 5 — a dramatic correction that moved the galaxy from the cosmic dawn to a comparatively unremarkable epoch [8].

As reported by Quanta Magazine, many experts came to suspect that tweaks to the initial mass function — the distribution of star sizes when they first form — and other astrophysical parameters could be sufficient to square the early galaxies with the standard model, without requiring new cosmological physics [9].

The Error Bar Problem

The steelman case against crisis requires confronting the question of how reliable Webb's measurements actually are. Photometric redshifts — estimated from the colors of galaxies across multiple filter bands — remain the primary method for identifying distant galaxy candidates. These estimates are efficient but imprecise. NIRISS (Near-Infrared Imager and Slitless Spectrograph) observations yield spectroscopic redshifts accurate only to approximately Δz ≈ 1×10^-3, and the process of deriving stellar masses from photometry involves stacking assumptions about star formation histories, dust attenuation laws, and metallicity that can each introduce factors of 2–3 uncertainty [10].

The critical question is how many of the anomalous galaxies have been confirmed spectroscopically rather than photometrically. The JADES Data Release 4 derived secure redshifts for 136 galaxies and tentative redshifts for an additional 109 that had previously lacked spectroscopic measurements [5]. Spectroscopic follow-up has substantially narrowed the field of candidates, confirming that some are genuinely at extreme distances while revealing others as lower-redshift interlopers. But the confirmed galaxies that remain are still more numerous and more luminous than pre-JWST models predicted.

A Research Ecosystem Built on ΛCDM

The stakes extend well beyond academic debate. ΛCDM is not merely a theory — it is the operating system on which modern cosmology runs. The U.S. Department of Energy has allocated $24 million in a single recent grant cycle for dark matter research, with an additional $6.6 million announced to capitalize on theoretical and technological advances [11]. The National Science Foundation maintains standing solicitations for direct-detection dark matter experiments [12]. Across the field, decades of effort and billions of dollars in cumulative investment have been directed toward experiments designed to detect dark matter particles whose existence is predicted by ΛCDM's assumptions [11].

The number of active research programs, graduate dissertations, and career trajectories built on these assumptions is difficult to quantify precisely, but the OpenAlex database offers a proxy. Since JWST's launch, peer-reviewed publications on "JWST galaxy formation" have surged from 593 in 2022 to 2,569 in 2024, before settling to roughly 2,287 in the current year [13]. This represents a field in active ferment, with thousands of researchers worldwide attempting to reconcile observations with theory.

Source: OpenAlex

Data as of Jan 1, 2026CSV

A major revision to ΛCDM would not invalidate all this work, but it would change the interpretive framework through which results are understood. Dark matter detection experiments, for example, are calibrated to specific particle mass ranges and interaction cross-sections derived from ΛCDM parameters. If the model's assumptions about early structure formation are wrong, the downstream implications for particle physics could be significant.

Alternative Frameworks and Their Advocates

The researchers most vocally arguing that Webb's data herald a new era in cosmology tend to be those who have long worked on alternatives to ΛCDM. The most prominent example is Modified Newtonian Dynamics, or MOND — a theory proposed by physicist Mordehai Milgrom in 1983 that modifies gravity's behavior at extremely low accelerations rather than invoking dark matter.

MOND's advocates argue that their framework predicted JWST's findings decades before the telescope launched. In the late 1990s, Stacy McGaugh and colleagues calculated that if MOND were correct, the universe's first galaxies would appear large and bright in their earliest stages — precisely the pattern JWST has revealed [14]. "The large and bright structures seen by JWST very early in the universe were predicted by MOND over a quarter century ago," McGaugh's group has argued [14]. Under MOND, gravity is stronger at galactic scales than Newton's equations suggest, allowing galaxies to assemble rapidly without needing dark matter halos as scaffolding.

However, MOND faces its own significant challenges. It has never been successfully integrated into a comprehensive cosmological framework that can explain the cosmic microwave background, large-scale structure, and galaxy cluster dynamics with the same quantitative precision as ΛCDM. Integrating MOND into a unifying framework that explains a wide range of cosmological observations has proven difficult [15].

Early Dark Energy (EDE) models represent another class of alternatives. These propose that a temporary burst of dark energy in the universe's first hundred thousand years could have accelerated the growth of density perturbations, producing more massive galaxies sooner. Research published in Monthly Notices of the Royal Astronomical Society has shown that EDE cosmologies "feature modifications in the growth of structure, potentially resulting in the formation of more massive galaxies" consistent with JWST observations [16]. EDE also has the advantage of potentially resolving the Hubble tension — the persistent disagreement between different measurements of the universe's expansion rate — making it an attractive two-for-one solution.

Other proposals include dark matter–baryon interactions, f(R) modified gravity theories, and dynamical dark energy models, each of which modifies ΛCDM in different ways to accommodate faster early structure formation [15].

Historical Parallels: The 1998 Supernova Surprise

The closest historical analogy to the current situation is the 1998 discovery that the universe's expansion is accelerating rather than decelerating. Two independent teams — the Supernova Cosmology Project and the High-Z Supernova Search Team — used Type Ia supernovae as standard candles to measure cosmic distances and found results that contradicted the simplest cosmological models [17]. Both teams had expected to confirm that gravitational attraction was gradually slowing the expansion. Instead, something was pushing the universe apart.

The response followed a pattern worth examining. The discovery was named "Breakthrough of the Year" by Science Magazine in 1998, received the Gruber Prize in Cosmology in 2007, the Breakthrough Prize in Fundamental Physics in 2015, and the Nobel Prize in Physics in 2011 (awarded to Saul Perlmutter, Adam Riess, and Brian Schmidt) [17][18]. The transition from anomaly to consensus took roughly 3–5 years, driven by the fact that two independent groups reached the same conclusion using different supernovae samples and analysis pipelines.

The key difference today is that JWST's findings do not point unambiguously toward a single new ingredient in the way that the 1998 data pointed toward dark energy. The accelerating expansion required adding a cosmological constant (or something equivalent) to the equations. The early galaxy anomaly could be resolved by any of several mechanisms — faster star formation, modified dark matter properties, early dark energy, or observational systematics — making the path to consensus less straightforward.

Who Gains, Who Loses

The sociology of the debate matters. Researchers like McGaugh and other MOND advocates have spent careers developing an alternative that mainstream cosmology largely ignored. If JWST data ultimately vindicate their predictions, they stand to gain significant credibility and influence. Conversely, researchers whose careers are built on ΛCDM-based simulations and dark matter detection experiments face the prospect of their foundational assumptions being weakened, though not necessarily overturned.

The loudest voices claiming a crisis are indeed disproportionately drawn from the community of alternative-model proponents. This does not make them wrong, but it does introduce a selection effect worth acknowledging. Meanwhile, the mainstream response has been to absorb the anomalies through refinements to astrophysical modeling — adjusting star formation efficiency, the initial mass function, and AGN feedback prescriptions — rather than discarding the cosmological framework itself.

The tension is real enough that it was the subject of a dedicated session at the 2025 APS Global Physics Summit, titled "Do JWST Early Galaxies Pose a Crisis for Cosmology?" [19]. The framing of the question — as a question rather than a declaration — reflects where the field actually stands: genuinely uncertain, with significant data still incoming.

What Comes Next

JWST is scheduled to continue operations through at least the early 2030s, and its survey programs are far from complete. The critical next steps involve three things: expanding the spectroscopic confirmation of high-redshift candidates, which will separate real anomalies from photometric artifacts; improving stellar mass estimates through better modeling of AGN contamination and dust; and testing whether the galaxy overabundance persists as survey areas grow and cosmic variance — the statistical noise from observing limited patches of sky — diminishes.

If the overabundance holds at a factor of two or more across large survey volumes with spectroscopic confirmation, the implications for ΛCDM become harder to dismiss as a problem of astrophysical tuning. The model's defenders would need to identify a specific mechanism — such as a top-heavy initial mass function in the early universe — that can be independently verified. If the anomalies are instead gradually explained away as measurement artifacts, the episode will be remembered as a lesson in the difference between a telescope breaking a model and a telescope stressing a model's least-constrained parameters.

Either outcome represents progress. The universe formed its first galaxies in ways we did not anticipate, and reconciling theory with observation — whether through new physics or better astrophysics — is exactly what a $10 billion telescope was built to enable.