Some three billion years ago, when Earth was a sprightly ocean world dotted with protocontinents and inhabited solely by single-celled organisms, a pair of black holes spiraled together and collided in a far-off region of the universe, leaving behind a single black hole some 50 times heavier than our sun. Emitting no light, the entire affair should have remained forever lost to the void.

Instead, the invisible violence of the pair’s final moments and ultimate merging was so great that it shook the fabric of reality itself, sending gravitational waves—ripples in spacetime—propagating outward at the speed of light. In the early morning hours of January 4, 2017, those waves washed over our modern Earth and into the most precise scientific instrument ever built, the Advanced Laser Interferometer Gravitational-Wave Observatory (LIGO). There the waves shifted the positions of vacuum-insulated, laser-bathed mirrors by less than the radius of a single subatomic particle. Traveling at light-speed, the waves first perturbed LIGO mirrors set up in Hanford, Wash., before passing through a second set of mirrors in Livingston, La., some three milliseconds later. Synced together from each station’s moving mirrors and converted to audible frequencies, the cosmos-quaking gravitational waves sounded like a single, soft “chirp.” Analyzing it, researchers are teasing out remarkable and otherwise-inaccessible details about the hidden lives of black holes. Announced Thursday by members of the LIGO team, the findings are described in Physical Review Letters.

As inconceivable as it may seem, tuning in to such chirps is now becoming routine. First predicted by Einstein more than a century ago as a consequence of his theory of general relativity, gravitational waves were long thought to be beyond observational reach—if not entirely nonexistent. But the chirp from January 4, dubbed “GW170104,” is actually LIGO’s third and farthest-reaching detection of gravitational waves, coming from somewhere about 3 billion light-years away. It follows earlier chirps from two other events detected separately in late 2015 that each occurred closer by, yet still more than a billion light-years distant.

Other cosmic phenomena such as supernovae in the Milky Way and colliding neutron stars in our galactic neighborhood should also produce detectable gravitational waves, each with their own accompanying revolutionary insights, but so far all three of LIGO’s detections have been death-rattles from merging pairs of black holes in remote stretches of the universe.

Gravity Rainbows?

For the time being, thousands of scientists around the world are making the most of LIGO’s limited view and the project’s three confirmed detections. Whereas the “loudness” of each chirp has clearly conveyed each event’s distance from us, LIGO’s twin stations can at present only vaguely constrain their celestial sources, which may lie anywhere within huge swaths of the heavens containing thousands upon thousands of large galaxies. So thirsty are theorists for new insights into black holes and relativistic processes that, with each LIGO detection, observational astronomers have leapt into action to target those enormous patches of sky, hoping to see some afterglow or other emission of electromagnetic radiation—even though by definition the resulting larger black hole should emit no light.

Fortunately, even without light the merger’s gravitational waves reveal much. LIGO team members have already used the billion–light-year intergalactic traverses of the first two chirps to look for signs of “dispersion” in the propagation of gravitational waves—a phenomenon analogous to how rays of light traveling through a prism disperse based on their wavelength to form rainbows. According to Einstein’s theory of general relativity, gravitational waves should experience no dispersion at all—and any deviation from that prediction would suggest Einstein’s relativistic reckoning of the universe is somehow incorrect, potentially pointing the way to new breakthroughs in physics. Signs of any dispersion should have been obvious in LIGO’s third event, GW170104, because its gravitational waves traveled across three billion light-years, rather than the one billion of LIGO’s previous two events. But when researchers looked, they saw no gravitational rainbows. “We made very careful measurement of that effect,” said LIGO team member Bangalore Sathyaprakash of The Pennsylvania State University and Cardiff University. “But we did not discover any dispersion, once again failing to prove that Einstein was wrong.”

Using that same measurement, researchers also honed in on the mass of the graviton, the hypothetical particle that mediates the force of gravity. “Basically we are testing general relativity in a new regime,” says Laura Cadonati, a physicist at Georgia Institute of Technology and LIGO’s deputy spokesperson. “The fact that this event is twice as far as the previous two gives us a longer baseline to test the dispersion relation, and as a result we now have a limit on the mass of the graviton that is 30 percent tighter than the one we previously set. One could say we are putting general relativity to a tighter and tighter test—it is still holding, but with more signals we may find something that does not quite agree.”

Mysterious Middleweight Mergers

Although LIGO’s latest event may be a brick in the towering edifice of Einstein’s general relativity, it is also restructuring the foundations of our understanding of black holes. Before LIGO’s detections, astronomers only had definitive observations of two varieties of black holes: ones that form from stars that were thought to top out around 20 solar masses; and, at the cores of large galaxies, supermassive black holes of still-uncertain provenance containing millions or billions of times the mass of the sun. Both are thought to be important for understanding the formation and evolution of galaxies, and thus to some degree important for the formation and evolution of everything galaxies contain—including stars, planets and people. Most of the black holes in LIGO’s mergers have been middleweights, being heavier than that 20–solar mass limit but much lighter than the supermassive variety, raising questions about their origins and relationship to the two well-studied populations of black holes.

The prevailing explanation for LIGO’s bulky black holes is that they form from very massive stars that are also quite pristine, composed almost entirely of hydrogen and helium with scarcely any heavier elements at all. Most stars of such immensity would have more heavy elements, causing them to lose much of their mass via high-speed winds whereas “low metallicity” stars would have weaker winds and keep more of their star stuff, ultimately ending their lives by collapsing to become overlarge stellar black holes.

Making LIGO’s merging black hole pairs, one conventional theory goes, would then require the “binary evolution” of two massive, low-metallicity stars that form as a pair. If, for instance, the two stars are very close, over the courses of their lives they can swap gas from their atmospheres back and forth in a cyclic process that pulls their orbits even closer and eventually produces two tightly orbiting, supersize black holes. At the end of this process, the spins and orbits of both black holes would have become inextricably linked, so each black hole’s equator would be aligned with the plane of their shared orbit.

“Think of black holes as being like tornadoes that drag stars and matter around them,” Cadonati explains. “Now think of two going around each other, and each one spinning clockwise or counterclockwise,” aligned with the orbital motion. Two black holes with such an alignment would possess more rotational energy than an unaligned pair, and thus require ever-so-slightly more time to coalesce together in the final moments of their merger. The deepest mystery of GW170104, LIGO’s latest discovery, is that the merger happened too quickly for both of its progenitor black holes to be so aligned; in terms of Cadonati’s analogy, at least one of the orbiting “tornadoes” must have been paradoxically tilted near or on its side.

The most common explanation for black hole pairs with such “spin misalignment” is that they did not form from the binary evolution of isolated twin stars. Instead, each black hole must have formed independently, and somehow found its partner after millions or billions of years of wandering through the universe. Any eventual union through this “dynamical formation” channel would most likely take place in thick swarms of stars called globular clusters, says Fred Rasio, a physicist at Northwestern University who is not a member of the LIGO collaboration. “Imagine throwing a thousand black holes into a mosh pit where they kick each other around like crazy,” Rasio says. “Their spins will be randomized. The dynamics don’t care which way the holes are spinning, so when they are bound into a pair that merges, their spins have no correlation with how they orbit.”

Black Holes from the Big Bang?

According to some theorists, the best explanation for GW170104’s curious misalignment is that its black holes did not start out as stars at all. “Even in dense globular clusters, these black holes would not form in sufficient density to find each other in the age of the universe,” says Juan García-Bellido, a professor at the Autonomous University of Madrid who is not a member of the LIGO collaboration. García-Bellido is a leading proponent of the unorthodox idea that LIGO’s abnormally heavy, oddly misaligned merging black holes are actually part of a putative population of “primordial black holes.” Rather than arising from stars, such exotic objects could have emerged in the first moments after the big bang, coalescing from particularly dense regions of the fiery plasmatic fog that then suffused the universe. If grouped in clusters, primordial black holes could also form merging pairs with misaligned spins.

There is, however, an additional wrinkle to ascribing primordial origins to some or all of LIGO’s observed black holes—something that could be seen as either the theory’s most alluring feature, or a nasty bug. Clusters of primordial black holes dense enough to produce LIGO’s newfound population of merging ones, García-Bellido and others say, could also be a natural solution to the mystery of dark matter—the elusive and invisible 80 percent of the universe’s matter that astronomers see solely through its gravitational effects on glowing stars and gas in galaxies.

“The idea would be that [the primordial black holes] would be concentrated in halos around the matter we can see,” said Michael Landry, the head of LIGO’s Hanford Observatory, summarizing the speculative concept in response to a question at a recent press conference. “It’s not impossible that what we’re seeing are primordial black holes that form the dark matter.” On the other hand, Landry added, some teams of astronomers occasionally looking for halos of primordial black holes around the Milky Way have yet to find evidence they exist in sufficient numbers to account for the effects of dark matter. Whether black holes from the big bang explain dark matter—not to mention LIGO’s results—is an “open question,” Landry said.

Hearing the Black Hole Symphony

Whether born from binary evolution, dynamical pairing, the big bang or something else entirely, the true origins of LIGO’s mysterious black hole mergers could soon be revealed. The collaboration’s current best guess is that somewhere between 12 and 213 such mergers occur each year in a cubic volume of space a bit over three billion light-years on a side. This suggests LIGO—which is in the midst of upgrades to boost its sensitivity and planning for a new station in India—could eventually be detecting the chirps from black hole mergers at a rate of anywhere between once per day to once per week. Upgrades are also in progress for Virgo, a companion gravitational-wave observatory approaching LIGO’s sensitivity. As early as this summer both projects will simultaneously monitor the sky to better localize the origins of any new celestial gravitational grumbles. Beyond LIGO and Virgo, additional observatories are likely to debut in coming years around the world, creating a globe-girdling network for finer-grained gravitational-wave searches. By the 2020s, the chirps will come so fast and furious, from so many merging pairs of black holes, their sounds could form a symphony.

“It’s not a single one or two black hole binaries by which we can distinguish between different models,” Sathyaprakash said. “It’s only from a population of detections, which will give us distributions for spins and for masses. That’s where the differences between formation mechanisms will become clear.” Very heavy, misaligned black hole pairs could prove to be very rare, strengthening the case that most mergers come from isolated systems of binary stars—or they could prove common, suggesting denser, more dynamical origins. And if, García-Bellido says, any black hole in a LIGO merger proves to weigh less than our sun, this would be a “smoking gun” for primordial black holes, as such relatively minuscule black holes are thought impossible to form from stars.

“Before our discovery, we didn’t even know for sure that these [middleweight] black holes existed,” Cadonati said at the press conference announcing GW170104. “What we do know now is, first of all, they do exist, they may have played an important role in the early universe and we’re now starting to get a glimpse into how they behaved…. This has really opened a new window on the universe, and we’re learning more about where we’re coming from. That’s the big excitement.”