JWST Reveals Galaxies That Should Not Exist So Early

JWST Reveals Galaxies That Should Not Exist So Early

In the spring of 2023, a team of astronomers pointed the James Webb Space Telescope at a patch of sky no larger than a few full moons. They were looking for something that, by all prior logic, should not be there: galaxies from the universe’s first 500 million years.

They found 26 of them.

These are not faint smudges. They are compact, dense, and surprisingly bright. The authors of the CEERS survey, led by Steven L. Finkelstein of the University of Texas at Austin, reported that the number of galaxy candidates at redshifts 9 to 16 exceeded nearly every theoretical prediction (Finkelstein et al., 2023). In plain language: the early universe appears to have been far more crowded with galaxies than our best models allow.

The result is not a tweak to existing theory. It is a challenge to the timeline of cosmic structure itself. If these galaxies are real, something about how stars and galaxies formed in the first half billion years is fundamentally different from what we assumed.

What the Telescope Actually Saw

The CEERS Survey in Numbers

The Cosmic Evolution Early Release Science (CEERS) program is one of 13 JWST Early Release Science projects. The team used JWST’s NIRCam instrument to image a 35.5 square arcminute field, roughly one sixth the area of the full moon. They combined seven NIRCam filters with six Hubble broadband filters to measure colors and total fluxes across a wide wavelength range.

The goal: find galaxies at redshifts above 9, meaning light that has been traveling for over 13 billion years, from when the universe was less than 5% of its current age.

After a meticulous data reduction process, the team identified 26 candidate galaxies at redshifts between roughly 9 and 16. The median half light radius of these objects was about 0.5 kiloparsecs, or roughly 1,600 light years. That is compact. For comparison, the Milky Way’s half light radius is about 5 kiloparsecs. These early galaxies were ten times smaller, yet they were producing enough ultraviolet light to be visible across cosmic time.

The Luminosity Function Surprise

The team calculated the rest frame ultraviolet luminosity function at redshift 11. This is a measurement of how many galaxies exist at a given brightness. What they found was striking: the number density of galaxies at an absolute magnitude of around 20 appeared to change very little from redshift 9 to redshift 11 (Finkelstein et al., 2023).

That flatness is the problem. In standard models of galaxy formation, the number density of bright galaxies should drop sharply at higher redshifts. The universe was younger. There was less time for dark matter halos to grow and for gas to cool and form stars. Yet JWST saw roughly as many bright galaxies at redshift 11 as at redshift 9.

The authors put it plainly: “the abundance (surface density [arcmin 2]) of our candidates exceeds nearly all theoretical predictions” (Finkelstein et al., 2023).

Why This Breaks the Cosmic Timeline

The Standard Story of Galaxy Formation

The prevailing model of galaxy formation, Lambda Cold Dark Matter, works beautifully at late cosmic times. It explains the large scale structure of the universe, the distribution of galaxies, and the cosmic microwave background. But it has a built in expectation for how quickly galaxies can assemble.

In the standard picture, dark matter collapses into halos first. Gas falls into those halos, cools, and eventually forms stars. This process takes time. In the first few hundred million years, the halos are small. The gas is hot. Star formation should be inefficient. The first galaxies should be faint, sparse, and hard to find.

JWST found the opposite: bright, abundant, and compact.

What “Exceeds Theoretical Predictions” Actually Means

When Finkelstein and his colleagues say the abundance exceeds predictions, they mean that the observed number of galaxy candidates is higher than the upper bounds of most simulations. Some models predicted a handful of galaxies at redshift 10. JWST found dozens in a single small field.

This is not a minor discrepancy. It is a factor of several, not a few percent. The authors note that the surface density of their candidates is higher than even optimistic predictions from semi analytic models and hydrodynamical simulations (Finkelstein et al., 2023). Something is systematically missing from the models.

Three Ways to Explain the Impossible

1. Top Heavy Star Formation

One possibility is that stars in the early universe were different. In the local universe, most star formation produces stars with a standard distribution of masses, called the initial mass function. Most stars are small, with only a few massive stars per thousand.

But in the early universe, gas was metal poor. Without metals to cool efficiently, star formation might have favored massive stars. A top heavy initial mass function would produce more ultraviolet light per unit of stellar mass. A galaxy with the same number of stars would appear much brighter.

The authors write that a top heavy IMF “would result in an increased ratio of UV light per unit halo mass” (Finkelstein et al., 2023). That could make galaxies look brighter than they really are, inflating the apparent number density.

2. No Dust

Dust absorbs ultraviolet light and re emits it at longer wavelengths. In the local universe, galaxies are dusty. But in the early universe, there may not have been enough time for stars to produce and distribute significant dust. Without dust, the ultraviolet light from young stars would escape unattenuated.

The authors list “a complete lack of dust attenuation” as another possible explanation (Finkelstein et al., 2023). If early galaxies are cleaner, they will appear brighter, and we will overestimate their star formation rates.

3. Changing Star Formation Physics

A third possibility is that star formation itself was more efficient in the early universe. Gas may have been denser, or feedback from supernovae may have been weaker, or the conversion of gas into stars may have simply happened faster.

The authors mention “changing star formation physics” as a potential factor (Finkelstein et al., 2023). This is the most radical option, because it would require rewriting the rules of how galaxies form.

How the Team Avoided False Positives

Photometric Redshifts and Contamination

The galaxies in this study are candidates, not confirmed. The team used photometric redshifts, which estimate distance based on the object’s color in different filters. This is a standard technique but it is not foolproof. A dusty galaxy at lower redshift can mimic the colors of a pristine galaxy at high redshift.

The team designed a robust set of selection criteria to minimize contamination. They required detections in multiple bands, consistent colors, and no detection in bands that would indicate a lower redshift interloper. They also checked their candidates against known populations of low redshift galaxies and active galactic nuclei.

Despite these precautions, the authors acknowledge that “spectroscopic confirmation of these sources is urgently required” (Finkelstein et al., 2023). Without spectra, there is always a chance that some candidates are impostors.

The Sample Size

Twenty six candidates is not a huge number. The authors are careful not to overstate their case. They present the result as an early look, not a final answer. But the fact that even a small survey found so many high redshift candidates is itself remarkable. Deeper surveys will likely find many more.

What This Means for the James Webb Space Telescope

JWST Was Built for This

The James Webb Space Telescope was designed to see the first galaxies. Its infrared sensitivity, its large mirror, and its location far from Earth’s heat all make it uniquely suited to this task. Hubble could see back to about redshift 11, but only with extreme effort and long exposures. JWST can see farther, faster, and in more detail.

The CEERS results are an early demonstration of that capability. The team used only the first epoch of NIRCam imaging. More data is coming. The authors note that “the deeper views to come with JWST should yield prolific samples of ultrahigh redshift galaxies” (Finkelstein et al., 2023). If the early results are any guide, those deeper views will be spectacular.

The Spectroscopic Follow Up

The next step is to obtain spectra for these candidates. JWST’s NIRSpec instrument can measure the redshift of a galaxy by detecting the Lyman alpha break or other spectral features. Confirming even a handful of these candidates would be a major result.

If the candidates are real, the implications are profound. If they are false positives, the models survive but the search continues. Either way, the data is forcing the field to confront a genuine puzzle.

What the Study Does Not Prove

This is not a settled result. The authors are clear about the limitations.

First, the sample is small. Twenty six candidates in one field is not enough to rewrite cosmology. It is enough to raise serious questions.

Second, photometric redshifts are not as reliable as spectroscopic ones. Some of these candidates may be lower redshift galaxies masquerading as high redshift objects.

Third, the theoretical models that the result “exceeds” are themselves uncertain. They depend on assumptions about star formation, dust, and feedback that are poorly constrained at high redshift. The discrepancy may be in the models, not in the data.

Fourth, the result applies only to the brightest galaxies. The faint end of the luminosity function, which contains most galaxies, is still unexplored at these redshifts. JWST will probe that population in coming cycles.

Fifth, the authors do not claim to have found a problem with the standard cosmological model. They are not challenging dark matter or dark energy. They are challenging our understanding of how galaxies form within that model.

What This Actually Means

• ▸The early universe produced bright galaxies faster than our current models allow. This is not a minor adjustment. It suggests that star formation or galaxy assembly was fundamentally different in the first half billion years.

• ▸Top heavy star formation is the most plausible fix. If early stars were more massive, they would produce more ultraviolet light per unit mass, making galaxies appear brighter and more numerous than they truly are. This is testable with JWST spectroscopy.

• ▸Dust is not the answer. If early galaxies were dust free, that would explain their brightness but not their abundance. The number density problem remains.

• ▸The models need to be updated. Semi analytic models and simulations that matched Hubble data are failing to match JWST data. The field is now in a phase of rapid revision, which is exactly what a new telescope should produce.

• ▸Spectroscopic confirmation is the next milestone. If even a handful of these candidates are confirmed at redshift 12 or higher, the result will stand. If they are all lower redshift interlopers, the field will need to refine its photometric selection methods.

• ▸JWST is working exactly as intended. The telescope was built to find things that surprise us. It has succeeded in its first year. The deeper surveys to come will either confirm or refute this result, and either outcome will teach us something new about how galaxies form.