In their quest to understand the first stars and galaxies that lit up the cosmos, astronomers are still in the dark—but getting closer to enlightenment one discovery at a time.

That’s the almost inescapable conclusion from initial observations by the James Webb Space Telescope (JWST), the $10-billion observatory that began science operations in July. Designed to glimpse the faint infrared glow of the universe’s earliest luminous objects, JWST’s vision reaches back into the first few hundred million years after the big bang, allowing it to obtain more and better data about newborn galaxies than any other facility yet built. But its haul of galactic “baby pictures” has proved more bountiful than most researchers dared to dream. Simply put, candidate galaxies in the early universe are popping up in numbers that defy predictions, with dozens found so far. Explaining this excess may require substantial revisions to prevailing cosmological models, changes that could involve the first galaxies forming sooner, their stars shining brighter—or perhaps the nature of dark matter or dark energy being even more complex and mysterious than previously thought.

Now two of JWST’s most tantalizing candidate early galaxies have stood up to further scrutiny, strengthening scientists’ suspicions that our knowledge of cosmic history is crucially incomplete. Dating back to 350 million and 450 million years after the big bang, at the time of their discovery, both galaxies were older than any others known before. They were found independently by two teams, one led by Rohan Naidu, now at the Massachusetts Institute of Technology, and the other led by Marco Castellano of the Astronomical Observatory of Rome in Italy. Initially posted on the preprint server, the two discovery papers have now cleared the key hurdle of peer-reviewed publication, each appearing in the Astrophysical Journal Letters in late November and October, respectively. This is more than a ceremonial milestone—early calibration issues with JWST’s instruments had fueled concerns among astronomers that such findings had potentially miscalculated the true distance to these galaxies, making them more modern imposters only appearing to be part of the early cosmic coterie. But after thorough peer review, “we can say with very good confidence that calibration is not an issue for these galaxies,” Castellano says. “They are very robust candidates. We have finally put to the rest the issues with calibration.” Follow-up observations will be needed, however, to absolutely confirm their record-breaking distances.

Astronomers have meanwhile since found several other early galaxy candidates, some seemingly as far back as 200 million years post–big bang. Prior to the launch of JWST, no one knew if galaxies could even form so early in the universe’s 13.8-billion-year history, at a time when matter was thought to still be sedately coalescing into the gravitationally bound clumps required to give birth to large groups of stars. “And so we’re wondering, ‘Do we really understand the early phases of the formation of these galaxies?’” said Garth Illingworth, an astronomer at the University of California, Santa Cruz, at a press conference held by NASA to announce the peer-reviewed validation of the first two candidates. “This has posed a lot of questions for the theorists.”

Chief among them is how, exactly, dark matter guided the emergence of galaxies. For the first few hundred thousand years after the big bang, the cosmos was so hot that gravity could not pull normal matter together to form large protogalactic clumps. Yet this was “not an issue for dark matter,” says Jorge Peñarrubia, a cosmologist at the University of Edinburgh in Scotland, “because dark matter does not interact via electromagnetic forces.” Instead gravity alone is this invisible substance’s master—meaning that in mere moments after the big bang, when primordial chaos otherwise reigned, gravity immediately began glomming together dark matter into large clumps known as halos. These dark matter halos are believed to have acted as gravitational sinks for normal matter, seeding the subsequent formation of galaxies in the early universe. The telltale motions of the stars they shepherd betray their endurance to this day. Such halos still surround galaxies like our own, majestic-but-invisible sculptors of the modern cosmos.

JWST’s rapid discovery of early galaxies might be testing our understanding of how these halos form, perhaps suggesting they reached an immense bulk earlier than expected. One explanation might involve the very nature of dark matter itself. Theorists have found that simple treatments of dark matter, in which it only interacts with itself and normal matter via gravity, can accurately replicate large-scale cosmic structure. But nature has no guarantee of simplicity. In reality, dark matter could interact with itself because of an as yet unknown force, perhaps via a particle that’s not in the current Standard Model of physics—possibly increasing the speed at which these halos grew and explaining how big, bright galaxies were able to arise so quickly.

Perhaps instead, however, these halos were simply more efficient at drawing in regular matter to feed star formation. “I think this is probably telling us something about how stars form in dark matter halos early on,” Peñarrubia says. Today our galaxy produces roughly one new star per year, but Castellano’s paper suggests that star-formation rates must have been at least 20 times higher in his and Naidu’s two candidate galaxies. Another JWST-derived preprint paper posits that Milky Way–sized galaxies could have arisen just a half-billion years after the big bang—a scenario that would demand star-formation rates 10 times higher still than Castellano’s estimates. But according to Michael Boylan-Kolchin, a cosmologist at the University of Texas at Austin, such outsize rates of star formation stretch the boundaries of what is physically possible. “If those values are correct, you’d need to have [galaxies] turning all their mass into stars and forming stars as fast as they could,” he says.

A perhaps more plausible possibility is that stars were somehow more efficient at accumulating mass in the early universe. This would lead to bulkier, brighter stars, enhancing early galaxies’ visibility to JWST. “Maybe you just create a whole load of very, very massive stars,” says Stephen Wilkins, an astronomer at the University of Sussex in England. These could be so-called Population III stars, the hypothesized first stars in the universe. Although astronomers have yet to conclusively observe such stars, there is plentiful circumstantial evidence for their existence. Emerging from the primordial hydrogen and helium gas that pervaded the early universe, Population III stars would lack heavier elements, allowing them to reach humongous sizes—hundreds of times bulkier than our sun. But like the brightest, briefest candles, these stars’ immensity would limit their lifetime to no more than a few million years, making their detection today difficult.

It is possible, however, that some of the more remote galaxies already found by JWST—and those even more ancient that may still await discovery—could contain evidence for Population III stars. The brightness of these galaxies could be attributed to such stars, which would be much hotter and brighter than subsequent Population II stars and Population I stars, such as our sun, both of which fill our modern-day universe. “It’s very definitely possible,” says Daniel Whalen, a cosmologist at the University of Portsmouth in England. To find out for certain, JWST will need to perform spectroscopic follow-up of these more distant galaxy candidates—a time-consuming process of gathering a rainbowlike spectrum from a galaxy’s emitted light to work out which chemical elements are present in its constituent stars. One clear signature of Population III stars, Whalen says, could be a specific spectral feature of helium that could only arise within stars that are hotter than about 100,000 degrees Celsius. “That would be evidence for a massive Population III star,” he says.

Such follow-up observations are set to begin imminently. Jeyhan Kartaltepe of the Rochester Institute of Technology is part of a team that has been approved time on JWST to follow up a handful of early galaxy candidates found in the Cosmic Evolution Early Release Science (CEERS) Survey,  for which Kartaltepe is a leading investigator. Such candidates are distinguished by their high redshifts—a stretching out of the wavelengths of their light caused by the expansion of the universe across cosmic time. This makes Kartaltepe’s spectroscopic follow-up not only an important probe of the galaxies’ stellar populations but also yet another “reality check” of their cosmic vintage. The hope is the measurements will allow astronomers to “understand the star formation rates and the age of the stars,” Kartaltepe says. The program, expected to begin no sooner than late December, will use eight hours of JWST time to obtain spectra of three target galaxies. Many more such programs are expected in the future.

Other, more intriguing ideas abound. If JWST finds that the apparent early burst of massive galaxy formation suddenly ebbed in subsequent cosmic epochs, this could suggest the universe was expanding faster than expected back then—perhaps twice as fast as predicted by current consensus estimates, says Nicola Menci, an astronomer at the Astronomical Observatory of Rome. This could be linked to the influence of a particular (and so far entirely hypothetical) variety of dark energy, which is the enigmatic and mysterious force that appears to drive the accelerating expansion of the universe. So-called phantom models of dark energy allow its potency to fluctuate across cosmic time. If such models are valid, they suggest dark energy’s influence on the universe’s expansion could have been far greater shortly after the big bang than they are today. Initial results from JWST “seem to be in contrast with most logical models we have considered up to now,” Menci says, namely Lambda Cold Dark Matter (Lambda-CDM), the theoretical model incorporating cosmologists’ current best estimates for the properties of dark matter and dark energy and their resulting effects on cosmic evolution.

Such ideas, while seemingly far-fetched, cannot yet be entirely ruled out as astronomers continue to grapple with the prevalence of galaxy candidates in the early universe. Some will likely turn out to be mirages, much closer galaxies masquerading as more remote ones because they contain large amounts of dust, which also causes their light to be redshifted. Yet initial follow-up of one of Castellano’s and Naidu’s galaxies using the Atacama Large Millimeter/submillimeter Array (ALMA) in Chile suggested little evidence for such high dust content. “Despite the ALMA results being interesting, JWST is the only instrument that can give definitive answers on these galaxies,” however, Castellano says.

Additional follow-up observations of galaxies like these may be conducted in JWST’s first year of science, Cycle 1, which runs until June 2023. More interesting results may occur in its second year of science, Cycle 2, for which astronomers can now propose programs by a deadline of January 27, 2023. “Spectroscopic follow-up with JWST is essential and is likely to dominate the requests on distant galaxies in Cycle 2,” Illingworth says. “We have a problem, and it’s real: Where the hell did these bright things come from? They weren’t in the storybook. We really have to understand what’s going on here.”

Editor’s Note (12/8/22): This article was edited after posting to remove quotes that were retracted by a source.